Inverse free electron laser acceleration for compact light sources J. Duris (SLAC), P. Musumeci, N. Sudar, I. Gadjev, Y. Sakai, O. Williams, J. B. Rosenzweig (UCLA) I. Pogorelsky, M. Polyanskiy, M. Fedurin, M. Babzien, K. Kusche, C. Swinson (BNL ATF) V. Yakimenko and R. K. Li (SLAC) A. Murokh (Radiabeam) AAC Thursday, August 4, 2016

Overview IFEL background Rubicon helical IFEL experiment at BNL ATF GeV/m IFEL designs Applications to compact light sources TESSA for efficient radiation production

IFEL Interaction In an FEL, energy is transferred from an electron-beam to a radiation field In an IFEL the electron beam absorbs energy from a radiation field. High intensity laser An undulator magnetic field couples high power radiation with relativistic electrons Significant energy exchange between the particles and the wave occurs when the resonance condition is satisfied.

Why are we interested in IFELs? IFEL well suited for mid-high energy ranges (50 MeV – up to few GeV) High power lasers available (10 mm, 1 mm, 800 nm) Mature permanent magnet undulator technology (cm periods) Plane wave or far field accelerator: minimal 3D effects. Transverse beam dimensions can be mm-size for mm-scale accelerating wavelengths. Vacuum-based accelerator Efficient mechanism to transfer energy from laser to electrons Simulations show high energy, high quality beams with large gradient ~GeV/m achievable with current technology! Preserves e-beam quality/emittance Stable energy output: static undulator field sets resonant energy. Potential for compact GeV-class accelerators for light sources 1:05, 1:13, 1:14

IFEL milestones Lasers may have large electric fields but direct energy transfer to electrons limited by Lawson-Woodward theorem 1972 – Palmer proposed breaking an assumption of LW by using an undulator to propagate particles at an angle with respect to a laser for sustained acceleration Early ‘80s – Courant, Pellegrini, Zakowicz, Sprangle, Tang studied the possibility of using IFEL for high energy physics Showed synchrotron losses limit acceleration to 10s of GeVs Large gradients at lower energies Robert B. Palmer, J. Appl. Phys. 43, 3014 (1972) E. D. Courant, C. Pellegrini and W. Zakowicz, Phys. Rev. A. 32, 5 (1985)

IFEL milestones 1992 - Columbia IFELA 1998 – BNL ATF IFEL Staged 5 MW FEL and IFEL Accelerated 750 kV electrons to 1 MeV (700 keV/m gradient) Absorbed 40% of FEL radiation (1.6 mm wavelength) 1998 – BNL ATF IFEL 2 GW CO2 laser in a 2.8 mm diameter sapphire waveguide 2.3 MeV energy gain and 5 MeV/m gradient Observed microbunching at 2.5 um with CTR Wernick and Marshall, Phys. Rev. A 46, 6 (1992) Y. Liu, et al., PRL 80, 20 (1998)

IFEL milestones 2004 - STELLA2 at BNL ATF 33 cm gap tapered planar undulator 100 GW CO2 laser Staged prebuncher + IFEL accelerator 80% of electrons accelerated >20 MeV/m 2005 - UCLA Neptune IFEL Strongly tapered period and field amplitude planar undulator 400 GW CO2 laser 15 MeV -> 35 MeV in ~25 cm Accelerating gradient ~75 MeV/m W. Kimura et al. PRL, 92, 054801 (2004) P. Musumeci et al. PRL 94, 154801 (2005)

Helical undulator Planar undulator: energy exchange stops twice per period Helical undulator: electrons always moving transversely in helix so always transferring energy Helical yields at least factor of 2 higher gradient. Especially important for higher energy (high K) IFEL's.

Rubicon IFEL experiment Helical geometry high gain high gradient IFEL First strongly tapered helical Halbach undulator Two different tapers used Demonstrate control of the final beam properties by undulator tuning Input e-beam energy 50 MeV Average accelerating gradient 100 MeV/m Laser wavelength 10.3 μm Laser power at interaction point 500 GW Laser focal spot size (w) 980 μm Laser Rayleigh range 30 cm Undulator length 54 cm Undulator period 4 – 6 cm Magnetic field amplitude 5.2 – 7.7 kG 0:47, 1:05

High gradient acceleration 52 to 106 MeV in 54 cm =>100 MeV/m average accelerating gradient A beamline aperture necessitated relaxed laser focusing, reducing available accelerating gradient This motivated an undulator retune to improve capture performance

High quality accelerated beams No laser 93 MeV – 1.8 % energy spread Very reproducible (mean energy std < 1.5 %) despite 30% rms laser power fluctuations Laser intensity 5 orders of magnitude lower than LWFA 1.3% sigma fit width Laser on shots

IFEL + prebuncher run UCLA permanent magnet based prebuncher ~30 cm UCLA permanent magnet based prebuncher Permanent magnet chicane with adjustable R56 Achieved > 50% capture IFEL acceleration preserves emittance GPT First experiment using a CPA CO2 laser 0:43, 0:44 93 MeV peak 2% rms energy spread 90 pC >90 MeV Unaccelerated emittance 2.3 um Accelerated emittance 2.4 um

Improved bunching schemes 40 cm Harmonic prebuncher b = 0.95 @ 800 nm Linearize ponderomotive gradient using harmonics Undulator 1 Chicane 1 Bucket to inject into Experiment planned: see Nick Sudar’s talk tomorrow morning Cascaded prebuncher 0:47 Undulator 2 Chicane 2 Stack two prebuncher stages to increase capture and reduce phase space dilution E. Hemsing and D. Xiang, PRSTAB (2013)

Applications Inverse Compton Scattering-based compact gamma-ray source Mode-locked soft-X-ray FEL Take advantage of current increase Use chicane to realign radiation spikes with e-beam modulation IFEL 0:31 Recirculate drive laser for IFEL to increase repetition rate and average flux of ICS photons SBIR with UCLA and Radiabeam to go forward

IFEL + Compton scattering experiment Experiment ongoing Two CO2 laser pulses separated in time One for IFEL acceleration One for head on collision with accelerated beam Power split between two pulses Plans to improve acceleration by retapering undulator No acceleration Acceleration >8 keV photons

0.5 GeV IFEL at ATF2 Demonstrate GeV/m gradients Demonstrate GeV-class energy gain Design aiming at 0.5 GeV output energy Flexibility : undulator can be retuned Parameter Value Laser power 25 TW Laser pulse length > 0.5 ps M2 1.5 Gap 10 mm Input energy 90 MeV Output energy 500 MeV Energy spread 2 % Undulator length 75 cm Gradient > 0.5 GeV/m 1:05 ATF2 stage 1 upgrade parameters E-beam energy 50-150 MeV Laser beam power 10-100 TW

LLNL Ti:Sa IFEL accelerator 100 fs First TW-class laser driven IFEL Strongly tapered undulator for diffraction-dominated interaction Short pulses (sub-ps) interaction 77 MeV – 122 MeV in 22 cm > 200 MV/m peak accelerating gradient! Sub-ps synchronization and timing Laser Electric Field Courtesy of J. Moody EOS measurement IFEL signal Design Parameters Initial Final Period 1.5 cm 5.0 cm Peak K parameter 0.2 2.8 Neptune Energies 14 MeV 52 MeV LLNL 50 MeV 200 MeV 1:10 Laser off Laser on Simulations

GeV IFEL concept Prebunched beam Achieve GeV/m gradient with commercially available Ti:Sa laser 20 TW, 800 nm laser 100 MeV prebunched e-beam 1m long permanent magnet undulator 1.1 GeV energy gain in 1 m Initial resonant phase –π/4 80 % capture 0.6 % energy spread 1:10

Beam loading effects Efficient optical to electrical power conversion Compensate laser power absorption in taper design by matching resonant energy gradient with ponderomotive gradient >70% power conversion possible in principle Laser depletion limits acceleration Compensate laser power absorption in taper design 80% conversion! 0 A 20 kA

Lessons from Inverse FEL FEL beam-laser energy exchange is usually < 1 MeV/m IFEL demonstrated energy exchange rate ~ 100 MeV/m Design studies indicate possibility of GeV/m gradients Beam loading compensation: 10 kA beam absorbs 50% power Can we run IFEL in reverse? High power laser In an IFEL the electron beam absorbs energy from a radiation field. 0:40 UCLA results from prebunched RUBICON

Tapering Enhanced Stimulated Superradiant Amplification Reversing the laser-acceleration process, we can extract a large fraction of the energy from an electron beam provided: A high current, microbunched input e-beam An intense input seed Gradient matching to exploit growing radiation field 0:53 IFEL deceleration

TESSA afterburner at 13.5 nm 4 kA @ 1 GeV = 4 TW beam power available Refocusing seeded FEL amplifier (~GW) to recreate high intensity condition 45% efficiency in 23 meters High rep rate => high average power 1:01 FEL undulator (saturated) Re-focusing optics TESSA afterburner Prebuncher

NOCIBUR IFEL deceleration experiment Can use RUBICON IFEL set up in reverse at BNL ATF Reversed and retapered the 0.5 m undulator for deceleration Potentially extract 40% of energy from a relativistic electron beam in half a meter Undulator Prebuncher Parameter Value E-beam energy 65 to 34 MeV E-beam current 100 to 400 A Laser focal intensity 4 TW/cm2 Laser wavelength 10.3 μm Rayleigh range 30 cm Laser waist 1.0 mm Input peak power 100 GW Output peak power 130 GW 1:12 Resonant energy Undulator parameters

NOCIBUR IFEL deceleration experiment Maximized capture with variable field chicane 45% of the 100 pC beam captured and decelerated 30% energy extraction efficiency (2 mJ) Spectra agree with simulation Fraction captured Phase delay (rads)

Towards high average power oscillator >30% efficiency * high average power e-beams => high average power laser High average powers useful for EUV lithography and atmospheric power beaming Nocibur is not so useful as a single-shot amplifier (need a strong, high rep-rate seed) But it could be useful in an oscillator configuration 1 um design study: Nocibur prebuncher >50% electrical to optical energy conversion efficiency (see Pietro Musumeci’s talk in WG7 today at 10:30)

Conclusion Experiments to date point to IFEL as mature and reliable laser-based high gradient accelerator technology Push to GeV/m gradients and GeV energy gain Enable laser-driven compact accelerator applications High gradient IFEL deceleration: TESSA can achieve ~50% electrical to optical energy conversion efficiency IFELA ‘92 MIFELA ‘01 BNL IFEL ‘98 STELLA2 ‘04 NEPTUNE IFEL ‘05 LLNL IFEL ‘15 RUBICON ‘13 20 TW TiSa ATF2 IFEL 1:00

IFEL collaboration The Rubicon IFEL and related experiments could not have happened without P. Musumeci (PI), N. Sudar, I. Gadjev, Y. Sakai, O. Williams, J. B. Rosenzweig (UCLA) I. Pogorelsky, M. Polyanskiy, M. Fedurin, M. Babzien, K. Kusche, C. Swinson, P. Jacob, R. Malone, M. Montemagno, G. Stenby (BNL ATF) V. Yakimenko and R. K. Li (SLAC) A. Murokh (Radiabeam) Funding agencies: DOE, DTRA, DNDO