Polarized Electrons for Linear Colliders J E Clendenin

Polarized Electrons for Linear Colliders J. E. Clendenin, A. Brachmann, E. L. Garwin, R. E. Kirby, D. -A. Luh, T. Maruyama, R. Prepost, C. Y. Prescott, J. C. Sheppard, and J. Turner

Outline Charge 2. Polarization 3. Other 4. Conclusions 1.

1. Collider Charge Requirements Parameter at Source Ne, mpulse Dz Impulse, avg n. C ns A *Twice the IP requirement ILC SCRF ILC NCRF 6. 4* 2 3. 2 2. 4* 0. 5 4. 8

Generating Polarized Electrons from Ga. As Illuminate p-doped Ga. As (or its analogues) crystal with circularly polarized monochromatic light tuned to the band-gap edge. Absorbed photons promote efrom filled VB states to CB. CB eeventually reach surface. In p-doped materials, band-bending lowers work function by ~1/3 of the 1. 4 e. V band gap. Treating surface with Cs(O) lowers it several additional e. V, resulting in the vacuum level being lower than the CBM in the bulk (NEA surface). If the cathode is biased negative, CB electrons at the surface are emitted into vacuum.

Space Charge Limit (SCL) (Child’s Law) SLC DC gun: ● Cathode bias -120 k. V to keep max fields <8 MV/m ● Low fields necessary to minimize the dark current that degrades the QE ● Ga. As crystal 2 -cm dia. decreases je, but increases beam emittance at source

For SLC, low-energy beam transport (various apertures in the 3 -4 cm range) designed for 20 n. C in 3 ns with beam interception in first m <0. 1%, in first 3 -m <1%. [Eppley et al. , PAC 91, p. 1964] cath-dia -bias cm k. V Eppley 1. 5 160 SLC 2. 0 120 Laser 3 ns, thick Ga. As cathode dia=1. 5 cm, bias -160 k. V, 20 n. C, thus Impulse, avg =6. 7 A SCL A 10 17 11

Parameter at Source ILC ILC SCRF NCRF-Inj/ Design SCRF-Linac (2 -cm) ne n. C 6. 4 2. 4 6. 4 20 Dz ns 2 0. 5 3 Impulse, avg A 3. 2 4. 8 12. 8 6. 7 Impulse, peak A 11 (SCL) Conclusion: Space charge limit a problem for ILC source only if try to operate with NCRF injector S-band linac

Surface Photovoltaic (SPV) Effect [Clendenin et al. , to be published in NIM A (2004)] Elsevier B. V.

Higher doping solves the SPV problem: can be restricted to last few nm at surface (“gradient doping”) to avoid depolarization effects in bulk* [Clendenin et al, to be published in NIM A (2004)] Elsevier B. V. Four samples with different doping levels: 5 1018 cm-3 1 1019 cm-3 2 1019 cm-3 5 1019 cm-3 *Creates the practical problem of how to clean the surface at low T prior to Cs(O) activation

SLAC Experimental Results Using High-Polarization Gradient-Doped Cathodes and Long Pulse Laser [a] Maruyama et al. , NIM A 492 (2002), 199, Fig. 18 [b] Clendenin et al. , to be published in NIM A (2004)

Very high current densities achieved by reducing the laser spot diameter at the cathode

2. Polarization n Highest polarization from thin (~100 nm) epilayer having a biaxial compressive strain. Strain produced by lattice mismatch with substrate and/or by quantum confinement associated with short-period superlattice structures. n Strain breaks the degeneracy of hh and lh energy bands at the VBM. A separation of 50 -80 me. V now readily achieved.

On absorption of photon, VB electron promoted to CB. The hh-lh splitting sufficient to selectrons from hh band only, resulting in CB electrons of 1 spin state only. [Alley et al. , NIM A 365 (1995) 1] Elsevier B. V.

Accuracy of SLC Polarimeters n The CTS (Cathode Test System) Mott at SLAC is a compact n The GTL (Gun Test Lab) Mott at SLAC is a medium-energy (120 n SLC Compton polarimeter was located at the IP (50 Ge. V) low-energy (20 k. V) retarding-field polarimeter located in the Cathode Test Lab k. V) multiple-foil polarimeter located in the GTL Same Mott polarimeters in operation at SLAC today

For ’ 96 and ‘ 97/’ 98, error of SLC Compton polarimeter measurements ~0. 5%, dominated by systematic uncertainties. [Abe et al. , PRL 84 (2001) 5945] Run CTS-Mott GTL-Mott Compton ’ 97 -98 77% ’ 96 78 72. 92%± 0. 38% 79 76. 16%± 0. 40% Known depolarization in NDR and NARC: ~2% (NDR 0. 8%; NARC 0. 7% energy spread, 0. 3% synchrotron radiation, 0. 4% beam emittance). Thus, during SLC, the Mott measurements (in lab) were consistently ~2% higher than Compton corrected for known depolarization effects. Some of the difference may be spin detuning in NARC.

E-158 Results n E-158 an experiment (2001 -2003) to measure parity violation at 50 Ge. V in electron-electron scattering at SLAC n Moller polarimeter at 50 Ge. V, similar to JLab’s. Depolarization in A-line ~1%. n Runs 1, 2 used Ga. As. P/Ga. As strained-layer cathode n Run 3 used Ga. As. P/Ga. As superlattice (SL)

Ga. As. P/Ga. As Superlattice (SL) Data showing high polarization from MOCVDgrown version first presented by Nishitani et al. at the PESP 2000 Workshop in Nagoya. n SVT Associates and SLAC collaborated to explore parameter space for MBE-grown version. n – Results show an amazingly stable high polarization over a wide range of parameter space [Maruyama et al. , Appl. Phys. Lett. 85 (2004) 2640] while maintaining a high QE. – One of these SL wafers used for E-158 -III.

Comparison of 3 photocathodes representing 2 structures

a T. Nishitani et al. , in SPIN 2000, AIP Conf. Proc. 570 (2001), p. 1021 b T. Maruyama et al. , Appl. Phys. Lett. 85 (2004) 2640 c On line, preliminary value of Pemax. d T. Maruyama et al. , Nucl. Instrum. and Meth. A 492 (2002) 199, Fig. 13 e P. L. Anthony et al. , Phys. Rev. Lett. 92 (2004) 181602

“Spin Dance” at Jefferson Lab [Grames et al. , PRST-AB 7 (2004) 042802] American Physical Society Relative analyzing power for 5 JLab polarimeters operated simultaneously to measure polarization of common beam on pulse-to-pulse basis. Error bars represent fits to the data only, statistical (much larger) and systematic errors not included. The Moller A value reduced to 1. 04 if data set limited to within 25% of max measured polarization (but error bars increase).

Maximum Polarization of SVT SL n CTS Mott n E-158 -III Moller (corrected for source) n Average (86± 5)% (91± 5)% (88± 4)%

3. Other Issues n Cathode QE, QE uniformity, anisotropy, lifetime – QE determines required laser energy—the higher the QE the more reliable the laser system can be – QE non-uniformity affects low-energy beam optics, thus needs to be stable – QE anisotropy very low for SL – QE lifetime must be >100 h to ensure stable operating conditions § Cannot always compensate for low QE with more laser energy because of SPV effect § Can restore QE by re-cesiating, takes ~15 min. § SLC lifetimes typically >400 h [J. E. Clendenin et al, in AIP CP-421 (1998), p. 250]

n Source Vacuum – Critical for high QE and long lifetime – Affects ion back bombardment § JLab 10, 000 C/cm 2 equivalent to 1/e lifetime [C. Sinclair, PAC 99, p. 65] § ILC maximum 1000 C/cm 2 per year (SLC type source) n High voltage cathode bias – Beam loading effects: each mpulse ~1 m. J, 1 -ms pulse train ~3 J – Pulsed HV can be shaped n Laser system – Laser to be modulated at mpulse frequency, i. e. , at ~3 MHz – The pulse train envelope can be shaped

Next Generation Polarized Electron Sources n Higher voltage n RG guns

4. Conclusions High probability that required charge for ILC can be produced using SLC type PES. Numerous problems introduced if mbunch spacing is reduced to significantly <300 ns. n Pe 85% is assured using well-tested Ga. As/Ga. As. P SL structure. n Various relatively minor issues remain. n