Download
compton based polarized positrons source for ilc n.
Skip this Video
Loading SlideShow in 5 Seconds..
Compton based Polarized Positrons Source for ILC PowerPoint Presentation
Download Presentation
Compton based Polarized Positrons Source for ILC

Compton based Polarized Positrons Source for ILC

86 Vues Download Presentation
Télécharger la présentation

Compton based Polarized Positrons Source for ILC

- - - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

  1. Compton based Polarized Positrons Source for ILC V. Yakimenko Brookhaven National Laboratory September 28, 2006 Oxford Positron Meeting

  2. ILC Source requirements * The length of the bunch train in ILC is 2820x300ns = 0.85 ms or 250 km. Bunch spacing has to be reduced in the dumping ring. ** Polarization level defines conversion/capture efficiency of polarized g rays into polarized positrons. 60% level corresponds to ~1.5% efficiency.

  3. Polarized Positrons Source (PPS for ILC) Conventional Non-Polarized Positrons: In this proposal • Polarized g-ray beam is generated in the Compton back scattering inside optical cavity of CO2 laser beam and 4-6 GeV e-beam produced by linac. • The required intensities of polarized positrons are obtained due to 5 times increase of the e-beam charge (compared to non polarized case) and 10IP CO2 laser system. • Laser system relies on the commercially available lasers but need R&D for the new mode of operation • 5ps 10J@0.05 Hz CO2 laser is operated at Brookhaven ATF.

  4. Compton Experiment at Brookhaven ATF(record number of X-rays with 10 mm laser) • More then 108 of x-rays per pulsewere generated in the experiment PR ST 2000. NX/Ne-~0.1. • (0.35 as of April 2006- limited by laser/electron beams diagnostics) • Interaction point with high power laser focus of ~30mm was tested. • Nonlinear limit (more then one laser photon scattered from electron) was verified. PRL 2005. Real CCD images Nonlinear and linear x-rays

  5. Compton based PPS with CO2 laser • No positron accumulation is needed: • Efficient head-on collision due to higher divergence of CO2 beam. • 10 mm CO2 beam has 10x number of photons per laser energy. • 750 W average power industrial laser. • Easier target and efficient positron capture: • Beam format changed in the injector from 3000 bunches @5 Hz to 100 bunches @150 Hz (1ms @150 Hz is more natural for warm RF and laser). • 40MV/m gradient in post target linac is possible. • Efficient collimation due to strong energy / divergence correlation of the gamma beam in the Compton scattering. • Doable laser: • Commercially available components (designed for 750W@500Hz; needed 750W@150Hz). • Low repetition rate model (10Hz) is operational at ATF as an amplifier of 5ps beam (laser cavity mode with 5ps pulse is needed). • Can be add-on option for non-polarized linac source. • Two Step R&D is needed: • Optimize conversion target and capture for this scheme. Test of the laser cavity at low repetition rate with ATF laser. • Assemble and test seed system and one laser cavity at full power.

  6. Choice of parameters Ng ,Ne and Nfare the numbers of g-rays, electrons and laser photons, S is the area of the interacting beams and sC is the Compton cross sections • ~40 mm laser focus is set by practical considerations of electron and laser beams focusing and requires ~5 ps long laser pulses • Nonlinear effects in Compton back scattering limit laser energy at ~2J • Pulse train structure of 2820 bunches @ 5 Hz is set by main linac. We change it to 100 bunches at 150Hz. This mode is more natural for warm RF and lasers. • ~300ns bunch spacing in the main linac will be changed in the dumping ring in any design. 6-12 ns bunch spacing is selected for optimal current in the drive linac and to match the inversion life time of the laser 12ns *100 bunches = 1.2 ms. • Train of ~10 nC electron bunches is required to produce 1012 polarized gammas per bunch. (~1 g-ray per 1 electron per laser IP) • Conversion efficiency of polarized gammas into captured polarized positrons is assumed at ~1.5% and is subject of optimization.

  7. Kerr generator 2 x 200ps CO2 oscillator Laser system for PPS 1x150ns Ge optical switch 2 x 5ps 1mJ 2 pulses, 5ps, 10mJ (YAG laser) Regenerative amplifier PC TFP TFP 2x5ps 10mJ PC amplifier 2x300mJ BS 2x30mJ 2x 30mJ 5ps • Train of 2 pulses spaced by 12 ns and 5 ps long sliced with a YAG beam from a 150 ns CO2 oscillator pulse • This train is seeded inside a regenerative amplifier cavity that has a round-trip time (12ns x 2=24 ns) • Amplified 2 pulses are dumped from the regenerative cavity with a Pockels cell and, after amplification, split with partial reflectors in 10 beams. • After amplification to 1 J/pulse, each 2-pulse train is injected into a ring cavity individual for each IP • An intracavity amplifier serves just to compensate optical losses during 1.2 s time interval needed for interactions with 100 electron bunches. amplifier amplifier 2 x 1J 5ps 2 x 1J 24ns ring cavities (2 pulses x 12ns spacing) 1J / pulse sustained for 1.2 ms amplifier amplifier IP#1 IP#10

  8. Laser system for PPS • Optical slicing and amplification of 5 ps CO2 pulses has been demonstrated and utilized in routine ATF operation for user experiments. • CO2 oscillator and initial amplifiers are commercially available lasers from SDI and operate at rep. rate up to 500Hz. • Final intra-cavity amplifiers shall operate at average power 750W (100 bunches x 1J x ~5% roundtrip intra-cavity loss x 150Hz). • High pressure CO2 laser is available at 750W average power at 500Hz. • Isotope mixture option is currently considered by the SDI to increase extracted energy. • R&D subjects: • Roundtrip intra-cavity loss optimization • Injection and stabilization energy during 1.2 ms • Operation of laser in nonstandard regime at high rep. rate

  9. Issues (uncertain items) • 1.5% gto e+conversion efficiency: • Assumed from previous JLC/NLC Compton design. Might change number of laser IPs. • Injection/intensity stabilization: • Ge switch is tested at low rep rate only. Cooling and high power 1mm laser/electron beam to control switch might be needed • Laser cavity: • Extensive intra-cavity loss would lead to insufficient laser energy. Amplifier with isotopic mixture or optical pumping might be needed. • Alignment of 10 laser IPs with 40 mm electron beam is doable engineering task that requires testing. • Compton photons in visible (large angles) are valuable diagnostic for feedback (From ATF experiment.)

  10. Conclusion • We propose Polarized Positron Source based on Compton back scattering inside optical cavity of CO2 laser beam and 4-6 GeV e-beam produced by linac. • The proposal utilizes commercially available units for laser and accelerator systems. • The proposal requires high power picosecond CO2 laser mode of operation developed at ATF.

  11. R&D at ATF with LDRD support • We will use existing equipment and attempt to demonstrate 100-pulse intra-cavity train. • Output from our regen will be sent into preamplifier arranged into Compton configuration. • We will need to solve a problem of self-lasing using saturable absorber cell. • The most of tests can be done with 200 ps pulses. • We may try finally 5 ps with isotopes. (Botha at SDI operated with isotopes and did not see a problem with it. ) • This, together with simulations of the train regime, will be the postdoc job. • Our announcement of postdoc position will be in November issue of Laser Focus.

  12. R&D without support • We will buy high-repetition (150 Hz) lasers, add laser room space, • build optical transport, and demonstrate one operational IP. • We are not sure about a real Compton with e-beam at ATF. It does not have sense for several reasons: • difficulty to place amplifier with the cavity in experimental hall • e-beam does not have necessary repetition rate • e-beam does not have right energy to test sequential IP’s (divergence is too high) • single-pulse interaction has been tested sufficiently