1. Linac e+ source for ILC, CLIC Vitaly Yakimenko, Igor Pogorelsky Brookhaven National Laboratory Compton Sources for X/  Rays: Physics and Applications

2. Compton Sources for X/  Rays: Physics and Applications, September 11, 20082/33 Compton Experiment at Brookhaven ATF US-Japan Collaborations More than 10 8 of x-rays were regestered in the experiment N X /N e- ~0.35. 0.35 was limited by laser/electron beams diagnostics Interaction point with high power laser focus of ~30  m was tested. Nonlinear limit (more then one laser photon scattered from electron) was verified. Real CCD images Nonlinear and linear x-rays

3. Compton Sources for X/  Rays: Physics and Applications, September 11, 20083/33 Layout of the experiment at ATF

4. Compton Sources for X/  Rays: Physics and Applications, September 11, 20084/33 Nonlinear Compton Scattering at ATF Experiment Simulations

5. Compton Sources for X/  Rays: Physics and Applications, September 11, 20085/33 10 ns 200 ps 5 ps CO 2 oscillator 3-atm preamplifier 10-atm regen. amplifier 10-atm final amplifier Kerr cell Ge switch 5 ps YAG pulse ATF CO2 laser System delivers 1 TW, 5 ps pulses

6. Compton Sources for X/  Rays: Physics and Applications, September 11, 20086/33 Laser focus characterization mm 75  m100  m150  m Gaussian approximation Transmitted energy 75  m d =75  m 100  m150  m no pinhole High power (~3J, 5ps) CO 2 laser spot size of  =32  m was demonstrated using F#~4 parabola with a hole for e-beam transmission. Laser is circular polarized as is required for ILC.

7. Compton Sources for X/  Rays: Physics and Applications, September 11, 20087/33 Small focus generation and measurements In the vacuum, permanent magnet quadrupoles were installed approximately 20 cm from focal location to produce 10  m spot size. (  cm and  m ). Image of the 30  m wire taken with same optical magnification as 0.5 nC e-beam on the right

8. Compton Sources for X/  Rays: Physics and Applications, September 11, 20088/33 Numbers and conclusions 60-75 MeV e-beam 0.5nC, 30-50  m RMS (50  m to minimize background), 1-3.5ps CO2 laser, sub-terawatt at IP, a 0 ~0.3 5J 30ps, 30  m 2 10 7 x-rays @ up to 6.5KeV per pulse (>10 8 x-rays in high background mode) 2.5% in the second harmonic >10 keV after 10  m Ag foil Rotation of the laser polarization leads to rotation of the peaks in the second harmonic Increase of the laser pulse duration to 200ps with the same energy eliminated nonlinear part with small change in the linear signal Peak brightness is estimated as B~1.6 10 20 mm -2 mr -2 s -1 0.1%BW

9. Compton Sources for X/  Rays: Physics and Applications, September 11, 20089/33 N x /N e ~ 1 is not a limit From Free Space … Laser pulse duration shall match the interaction length that is defined by Rayleigh distance in free space. ….to Plasma Channel We brake the laser pulse duration constraint by extending the interaction length in a plasma channel. I.V. Pogorelsky, I. Ben-Xvi, X.J. Wang, T. Hirose, Nuclear Instrum. Methods in Phys. Res. A 455 (2000) 176-180

10. Compton Sources for X/  Rays: Physics and Applications, September 11, 200810/33 Channeling of CO2 laser beam in capillary discharge laser beam at the focal point laser beam 18 mm downstream from the focus in the free space plasma discharge with the capillary We were unable to claim enhancement on X ray production with the guiding. This is due to strong focusing / scattering / loss of the 60 MeV electron beam by the plasma channel (Enhancement on the detectors was ~10x the expected amount)

11. Compton Sources for X/  Rays: Physics and Applications, September 11, 200811/33 ILC Source requirements ParameterSymbolValueUnit Positrons per bunchnpnp 2x10 10 e+e+ Bunches per pulseNbNb 2820 Bunch Spacing* bb ~300ns Pulse rep. ratef rep 5Hz Positron Polarization**PpPp ~60% * 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  rays into polarized positrons. 70% level corresponds to ~2% efficiency.

12. Compton Sources for X/  Rays: Physics and Applications, September 11, 200812/33 Polarization

13. Compton Sources for X/  Rays: Physics and Applications, September 11, 200813/33 Polarized Positrons Source (PPS for ILC) Conventional Non- Polarized Positrons: In this proposal Polarized  -ray beam is generated in the Compton back scattering inside optical cavity of CO 2 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 CO 2 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 CO 2 laser is operated at Brookhaven ATF.

14. Compton Sources for X/  Rays: Physics and Applications, September 11, 200814/33 Compton based PPS with CO2 laser No positron accumulation is needed: –Efficient head-on collision due to higher divergence of CO2 beam. –10  m 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 (1  s @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.

15. Compton Sources for X/  Rays: Physics and Applications, September 11, 200815/33 Choice of parameters ~40  m 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  s. Train of ~10 nC electron bunches is required to produce 10 12 polarized gammas per bunch. (~1  -ray per 1 electron per laser IP) Conversion efficiency of polarized gammas into captured polarized positrons is estimated at ~2% and is subject of optimization. N , N e  and N   are the numbers of  -rays, electrons and laser photons, S is the area of the interacting beams and  C is the Compton cross sections

16. Compton Sources for X/  Rays: Physics and Applications, September 11, 200816/33 Laser system for PPS amplifier 24ns ring cavities (2 pulses x 12ns spacing) 1J / pulse sustained for 1.2 ms IP#1IP#10 CO2 oscillator 2 pulses, 5ps, 10mJ (YAG laser) 2 x 200ps Kerr generator 2 x 5ps 1  J Regenerative amplifier amplifier 2x5ps 10mJ 2x300mJ BS TFP PC 2x 30mJ 5ps 2 x 1J 5ps 2x30mJ 2 x 1J 1x150ns Ge optical switch amplifier 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.

17. Compton Sources for X/  Rays: Physics and Applications, September 11, 200817/33 Lasers from SDI WH20WH100WH350 WH500 Wavelength 9 – 11µm, Line Tunable Continuous 20 Hz100 Hz350 Hz500 Hz Repetition Rate Pulse Energy1.5 J Mode Type Multimode Optional:TEMoo, custom beam shapes, SLM Beam Size13 x 13 mm 2 Average Power30 W150 W525 W750 W Power Stability < 7 % http://www.lightmachinery.com/SDI-CO2-lasers.html

18. Compton Sources for X/  Rays: Physics and Applications, September 11, 200818/33 Linac Compton Source (LCS): Numbers e- beam energy 6 GeV e- bunch charge 10nC RMS bunch length (laser & e - beams) 3 ps  beam peak energy 60 MeV Number of laser IPS 10(5) Total N  /Ne - yield (in all IPs) 10(5) Ne + /N  capture 2% (4%) Ne + /Ne - yield 0.2 Total e + yield 2nC # of stacking No stacking Proposal numbers are in black, Optimistic numbers are in Red

19. Compton Sources for X/  Rays: Physics and Applications, September 11, 200819/33 LDRD – cavity tests Has a potential to increase average intra- cavity power ~100 times at 10.6 microns. Purpose of the test: Demonstration of 100- pulse train inside regenerative amplifier that incorporates Compton interaction point. Demonstration of linear- to-circular polarization inversion inside the laser cavity. Test of the high power injection scheme

20. Compton Sources for X/  Rays: Physics and Applications, September 11, 200820/33 “~100 times increase of the average intra- cavity power at 10  m” The required laser train format / repetition rate /average acting power at each IP: 100 pulses x 150Hz x 1J = 15 kW. Efficient interaction with electron beam requires short (~5ps) and powerful (~1-2J) laser beam. Such high-pressure laser does not exist. Non-destructive feature of Compton scattering allows putting interaction point inside laser cavity. We can keep and repetitively utilize a circulating laser pulse inside a cavity until nominal laser power is spent into mirror/windows losses. Assuming available 0.5 kW CO2 laser and 3% round-trip loss, 1-J pulse is maintained over 15,000 round trips/interactions (100 pulses x 150 Hz). Thus, 0.5 kW laser effectively acts as a 15 kW laser. Equivalent solid state ( 1  m) laser producing the same number of gamma photons should be 150 kW average power with ~10J, 5ps beam.

21. Compton Sources for X/  Rays: Physics and Applications, September 11, 200821/33 LCS LDRD: Simplified test setup First observations: Optical gain over 4  s Misbalanced gain/loss regime results in lasing interruption by plasma Single seed pulse amplification continues to the end

22. Compton Sources for X/  Rays: Physics and Applications, September 11, 200822/33 LCS LDRD: Simplified test setup Balancing gain/losses and plasma threshold results in continuous amplification with two characteristic time constants : –~200 ns due to CO2 inversion depletion –~2  s due to N2-CO2 collisional transfer. Early injection of seed pulse before the gain reaches maximum allows to control train envelope

23. Compton Sources for X/  Rays: Physics and Applications, September 11, 200823/33 3% over 1  s The best train uniformity achieved Very encouraging results obtained with simplified cavity test setup: ~200 ps pulse of the order of 100 mJ circulated for >1  s. Further test would require pulse length monitoring and high pressure or isotope mixture based amplifier (to sustain 5 ps beams).

24. Compton Sources for X/  Rays: Physics and Applications, September 11, 200824/33 Applications for high power  - source 30-60MeV: e+ source for ILC, CLIC or SuperB 15MeV: Rare isotopes production, radioactive waist management 2-10GeV High brightness polarized muon production

25. Compton Sources for X/  Rays: Physics and Applications, September 11, 200825/33 Pair production cross-section* or 0.5 10 -5 for hydrogen for the rest of the nuclei *above threshold and before the saturation

26. Compton Sources for X/  Rays: Physics and Applications, September 11, 200826/33 Muon beams The probability of the formation of a      pair by  in the field of nuclei is suppressed approximately by a factor of (m e /m  ) 2. Low emittance from the direct production of the     pair makes this approach competitive with the currently considered production scheme in which a high-power proton beam generates a pion shower, and the pions, in turn, decay into muons. Indirect muon production is orders-of-magnitude more efficient in terms of the number of the muons per incident beam power yet the brightness of the resulting beam is much less, so that very challenging and complicated cooling schemes must be incorporated into the system in the case of proton driver. High-efficiency normal conducting, SC energy-recovering linacs and high-average-power laser cavities offers the opportunity to generate extremely powerful high-energy  beams through Compton backscattering.

27. Compton Sources for X/  Rays: Physics and Applications, September 11, 200827/33 Probability of creating  pairs as a function of the incident photon energy for hydrogen- (solid line) and tungsten (dotted line) targets.

28. Compton Sources for X/  Rays: Physics and Applications, September 11, 200828/33 Emittances longitudinal and transverse normalized emittances of the  beams produced with a high-intensity 2 GeV  beam from the tungsten target are estimated as 0.3 mm lower transverse emitances are possible with focusing (plasma length…)

29. Compton Sources for X/  Rays: Physics and Applications, September 11, 200829/33 2GeV  beamPulsed LinacERL e-beam energy [GeV]3611 Laser wavelength [  m]101 Bunch charge [nC]101.5 Rep. rate [Hz]200CW Bunches per beam250 Average current [mA]0.530 / 300 e-beam power [MW]18330 / 3300 e-to-  convers. efficiency30.33  -beam power [MW]320 / 200 Total AC-to-  efficiency10%20% / 75% Peak  +  - [per bunch]10 6 3 10 4 Average  +  - [per second]5 10 10 3 10 11 / 3 10 12

30. Compton Sources for X/  Rays: Physics and Applications, September 11, 200830/33 High Brightness of muon beams The longitudinal and transverse normalized emittances of the     beams produced with a high-intensity 2 GeV  beam from the tungsten target are calculated as 0.3 mm. Transverse emittances can be further reduced by immersing the target into a focusing field thereby limiting the muon beam’s size during the gamma ray-beam interaction. Focusing will be necessary for a long, targets from light nuclei like liquid-hydrogen. It is estimated that the longitudinal and transverse normalized emittances of the captured muon beam produced with proton beams as 20cm and 18 mm, respectively.

31. Compton Sources for X/  Rays: Physics and Applications, September 11, 200831/33 Conditions for Bubble formation Laser power threshold: Bubble can be formed in a finite window of plasma densities: laser trapped e - cavity Accelerated charge scales as: Final energy : 10 micron laser unlikely to offer record gradient in this application, but it might solve problems for practical applications: lower plasma density, longer channel, more stable, controlled final energy and higher charge.

32. Compton Sources for X/  Rays: Physics and Applications, September 11, 200832/33 Compton back scattering – compact sub 100 fs x ray source 0.5 ps 50J 5ps 10J “Dream beam” accelerator