Measurements of the LCLS Laser Heater and its

Measurements of the LCLS Laser Heater and its impact on the X-ray FEL Performance Z. Huang for the LCLS commissioning team FEL 2009, Liverpool, UK August 26, 2009 Zhirong Huang zrh@slac. stanford. edu

Outline • Why a Laser heater? LCLS setup and measurements Effects on FEL performance Anomalous heating effects FEL 2009, Liverpool, UK August 26, 2009 2 2 Zhirong Huang zrh@slac. stanford. edu

Introduction Ø Beams from photoinjectors have extremely small initial slice energy spread (2 -3 ke. V). Ø Such a “cold’’ beam can undergo a microbunching instability* in linacs and compressors that may ruin beam quality. Ø “Heating” slice energy spread within the FEL tolerance can suppress the instability and make the beam more stable. Ø Many FEL designs call for a laser heater in the injector. Ø First of its kind was installed and commissioned at LCLS * Borland et al. , NIMA (2002); Saldin, Schneidmiller, Yurkov, NIMA (2002); Heifets, Stupakov, Krinsky, PRST-AB (2002); Huang, Kim, PRST-AB (2002). FEL 2009, Liverpool, UK August 26, 2009 3 3 Zhirong Huang zrh@slac. stanford. edu

m-Bunching Landau Damped by Laser Heater Ti: saph 758 nm <15 MW Injector at 135 Me. V 0. 5 -m undulator ‘Laser heater’ suggested by Saldin et al. , NIMA, 2004; independently by J. Galayda 14 Ge. V Without Laser Heater FEL 2009, Liverpool, UK August 26, 2009 LCLS design study: Z. Huang et al. , PRST 2004 (chicane suggested by T. Smith) 14 Ge. V With Laser Heater 4 4 Elegant simulations Zhirong Huang zrh@slac. stanford. edu

How does a laser heater work? Dg. L after chicane after undulator Modulation smeared by half chicane l. L R 52 and angular spread laser peak power laser rms spot size The last half chicane time-smears the energy modulation leaving an effective “thermal” energy spread increase FEL 2009, Liverpool, UK August 26, 2009 5 5 Zhirong Huang zrh@slac. stanford. edu

Outline • Why a Laser heater? LCLS setup and measurements Effects on FEL performance Anomalous heating effects FEL 2009, Liverpool, UK August 26, 2009 6 6 Zhirong Huang zrh@slac. stanford. edu

Layout of the Laser Heater Optics Two cameras for monitoring laser beam MH 2 & MH 3 provide laser pointing control 758 nm IR laser from laser room transport to the tunnel Two OTR screens for spatial alignment Fast photodiode for timing within ~10 ps OTR screens shutter photo diode e. H rt to L o p s n m tra vacuu VHC camera MH 4 wave plate MH 3 power meter FEL 2009, Liverpool, UK August 26, 2009 MH 2 Sasha Gilevich et. al. , CH 1 camera 7 7 Zhirong Huang zrh@slac. stanford. edu

Laser Heater Spatial Alignment IR Calculate and re-alignpoor laserheating? time energy e- good heating Slide from P. Emma One button click (~1 minute) After three iterations, Dave approves it! H. Loos FEL 2009, Liverpool, UK August 26, 2009 8 8 Zhirong Huang zrh@slac. stanford. edu

Laser Heater Temporal Overlap 16 ps OTR 22 L 2 -linac FEL 2009, Liverpool, UK August 26, 2009 BC 2 COTR signal after BC 2 bends (OTR 22 inserted) Laser pulse 10 -20 ps, electron bunch 5 -7 ps Laser timing done by minimizing the COTR signal after BC 2 Laser energy 230 µJ COTR from microbunching suppressed when laser timed Laser delay line in mm (1 ps = 0. 15 mm) 9 9 Zhirong Huang zrh@slac. stanford. edu

Heating Measurements YAGS 2 RF deflector ON energy Laser OFF σE/E < 12 ke. V FEL 2009, Liverpool, UK August 26, 2009 YAGS 2 Laser: 40 µJ σE/E 45 ke. V Laser: 230 µJ σE/E 120 ke. V time 10 10 Zhirong Huang zrh@slac. stanford. edu

Slice Energy Spread YAGS 2 Laser energy 230 µJ σE/E 120 ke. V Laser OFF FEL 2009, Liverpool, UK August 26, 2009 Double-horn when laser spot ~50% larger than transverse electron beam size 11 11 More uniform heating when laser spot matches transverse size of electron beam Zhirong Huang zrh@slac. stanford. edu

Slice-E Spread vs. Laser Energy Anomalous heating(? ? ) No heating < 10 ke. V (limited by resolution) FEL 2009, Liverpool, UK August 26, 2009 Operating point (~7 u. J) 20 ke. V compression ratio 75/14 Ge. V = 1. 1 10 -4 12 12 Zhirong Huang zrh@slac. stanford. edu

No Emittance Growth from Heating • Laser-induced E-spread in chicane does not introduce emittance growth (< 2%) Heater from 45 m. J to 250 m. J, no change in slice emittance Heater off, COTR after LH chicane biases emittance results 6 ps bunch length FEL 2009, Liverpool, UK August 26, 2009 13 13 Zhirong Huang zrh@slac. stanford. edu

Effects on COTR • Unwanted microbunching leads to many COTR problems in LCLS Laser heater suppresses COTR but not to incoherent level OTR 22 L 2 -linac BC 2 OTR 22 (heater off) OTR 22 image, heater full Camera 10 bit depth Camera 5 bit depth laser-heated energy profile non-Gaussian small part of beam escapes heating Microbunching restart from shot noise after a long linac and some R 56 Alternative diagnostics (wires) used for beam profiles after compressors FEL 2009, Liverpool, UK August 26, 2009 14 14 Zhirong Huang zrh@slac. stanford. edu

Outline • Why a Laser heater? LCLS setup and measurements Effects on FEL performance Anomalous heating effects FEL 2009, Liverpool, UK August 26, 2009 15 15 Zhirong Huang zrh@slac. stanford. edu

Laser Heater Improves FEL Power 250 p. C, 3 k. A, 1. 5 Å FEL Laser optimal • FEL saturation length 20 m shorter w/ heater OFF FEL 2009, Liverpool, UK August 26, 2009 Scan unsaturated FEL w/ 12 undulators vs. heater • FEL saturation power improves 2 X w/ heater 16 16 Zhirong Huang zrh@slac. stanford. edu

FEL Gain Length vs. Laser Heater 250 p. C, 3 k. A, 1. 5 Å FEL Heater OFF bunching effectively increases E-spread to 0. 03% ~ 0. 04% level 0. 01% at 14 Ge. V with compression factor 90 FEL 2009, Liverpool, UK August 26, 2009 1. 5 Å 0. 4 - m emit 0. 5 - m emit Heater OFF Heater ON 17 17 Zhirong Huang zrh@slac. stanford. edu

Outline • Why a Laser heater? LCLS setup and measurements Effects on FEL performance Anomalous heating effects FEL 2009, Liverpool, UK August 26, 2009 18 18 Zhirong Huang zrh@slac. stanford. edu

“Trickle” Heating Anomalous heating? ? Trickle heating peaks at ~ 1 m. J FEL 2009, Liverpool, UK August 26, 2009 19 19 Zhirong Huang zrh@slac. stanford. edu

Laser Modulation Hidden in 4 D Phase Space • Laser heater at 0. 5 m. J (Elegant simulation by Y. Ding) Longitudinal phase space Longitudinal after heater chicane after (emittance 0. 5 0. 01 mm) Emittance mm after undulator x-t correlation x’-t correlation (R 52 of half chicane) after heater chicane FEL 2009, Liverpool, UK August 26, 2009 20 20 Zhirong Huang zrh@slac. stanford. edu

Modulation Recovered in Real Space x’-t. correlation spectr YAGS 2 dipole heater x-t correlation FEL 2009, Liverpool, UK August 26, 2009 21 21 Zhirong Huang zrh@slac. stanford. edu

• Longitudinal Space Charge (LSC) x Consider a beam as shown R tilt angle R = Dz/Dx, R is related to transfer matrix z • When x. R > , longitudinal density modulation is suppressed strongly (exponentially if Gaussian distribution in x) • In a 1 D approach (as used in Elegant), longitudinal modulation multiply by 1 D LSC impedance to compute energy modulation amplitude FEL 2009, Liverpool, UK August 26, 2009 22 22 Zhirong Huang zrh@slac. stanford. edu

LSC for Tilted Microbunching Beam clearly modulated for tilted microbunching in 3 D density • If beam size x is much larger than , 3 D calculation shows G. Stupakov, private communications D. Ratner, A. Chao, Z. Huang, FEL 08 LSC depends weakly on R when R<1/ Tilted microbunching does not suppress LSC exponentially. 1 D approach underestimates LSC by a large factor Take L=758 nm, =264, x=60 m, R ~2, 1 D underestimates LSC by a factor ~250! Integrating 3 D LSC over heater downstream beamline for total energy modulation FEL 2009, Liverpool, UK August 26, 2009 23 23 Zhirong Huang zrh@slac. stanford. edu

Comparison with measurements Trickle heating subtract inquadrature Normal heating LSC induced by laser modulation (theory) LSC + normal laser heating (theory) FEL 2009, Liverpool, UK August 26, 2009 24 24 Zhirong Huang zrh@slac. stanford. edu

Summary • Laser heater is used in the LCLS to improve and optimize xray FEL performance at the nominal operating condition. • Microbunching is suppressed but not all COTR is gone alternative diagnostics necessary (wire scanner etc…). • Unexpected “trickle” heating explainable by a 3 D LSC model (Preliminary estimation of 1 D CSR effect is small, any 3 D effect at play? ) • Trickle heating does not affect LCLS operation but may have implications to other laser heater designs as well as lasermanipulations of high-brightness beams. FEL 2009, Liverpool, UK August 26, 2009 25 25 Zhirong Huang zrh@slac. stanford. edu

Thanks to: LCLS project director J. Galayda, all commissioning team members, and collaborators and visitors FEL 2009, Liverpool, UK August 26, 2009 26 26 Zhirong Huang zrh@slac. stanford. edu