1. Gas Heating Effect in High Power FEL Operations Y. Feng, J. Krzywinski, D. Schafer, and T. Raubenheimer 11/12/2014 LCLS-II Accele. Phys. Meeting

2. 2 High Power FEL Operations for XTES LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 Maximum FEL average power limited to 200 W* into FEE *by using low charge or lower rep rate to cap at 200 W

3. 3 FEE Design to ≤ 200 W; Instruments to ≤ 20 W LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 X-ray diagnostics -Beam imagers -Energy monitors -HXR spectrometer -K-monochromator X-ray components -Attenuators -Stoppers/collimators -Apertures/slits -Offset/focusing Mirrors ≤ 200 W ≤ 20 W New SXR instrument(s) Existing HXR instruments *LCLS-I SXR and HXR shared the same FEE beamline

4. 4 XTES General Requirements LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 XTES components must be capable of sustaining full un-focused FEL peak power be capable of dissipating a maximum of 200 W FEL average power for FEE components*, and 20 W for instruments preserve FEL characteristics to extent feasible by -minimizing FEL intensity reduction for “non-invasive” diagnostic devices to preserve extremely high peak power Enabling X-ray strong-field physics -minimizing FEL wavefront distortions and transverse coherence degradations to preserve coherence properties Enabling coherence-based experimental techniques -minimizing pulse stretching if monochromators are used to preserve the ultrafast temporal property Enabling ultrafast time-resolved studies

5. 5 LCLS-II FEL Thermal Effects from Impulse and Average LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 For LCLS-I low repetition rate operation Only damages from a single-shot impulse are considered, average heating effect is generally negligible due to low power < 1 W. For LCLS-II high repetition rate (thus high average power) operation Thermal effects from both impulse and average power are considered together. -Steady-state solution can be worked out by assuming CW input. -Single-shot impulse solution can be worked out separately using the steady-state as baseline -Complete solution is a superposition of those of steady-state and single-shot impulse*

6. 6 LCLS-II X-ray Attenuators LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 Performance requirements

7. Must Use Gas Based Techniques for SXR LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 7 Design concept similar to LCLS-I gas attenuator, but -Using Ar gas, 5 m long volume, up to 10 torr -Differential pumping w/ 1st variable (impedance) apertures to reduce conductance (beam size ~ 10 mm at 200 eV at location)

8. 8 Gas Filamentation Effect in Pulsed Laser LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 High repetition rate (and high power) operation “Hole burning” effect induced by femtosecond lasers in gas, leading to density depression with slow recovery time of milliseconds Y-H. Cheng, et al, Opt. Express 4, 4740 (2013) Ti:Sapphire 800 nm 40 fs 0.72 mJ/pulse at 20 Hz 100  m FWHM N 2 at 1 atm pressure 10-20% density depression recovery time ~ 1 ms Interferometer technique after one pulse

9. 9 “Photo-Thermal” Dynamics in Gas LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 “State of gas” irradiated by FEL Photo-, electronic processes (fs to ps) -Photoionization -X-ray fluorescence -Auger process -Impact ionization -Recombination Thermalization from electronic to mechanical degrees of freedom creating temperature/pressure gradient (< ns) -High energy deposition of 70 meV/particle/pulse (~ 1000 K) Pressure equilibration followed by thermal diffusion resulting in temperature/density gradients (ns to ms, and then  s) Other dissipation mechanism -(Atomic) emission process -Luminescence (vibrational, rotational) Ionization ratio (~ 2 mJ/pulse, ~ 500 - 1000  m beam size) Ar

10. 10 Pressure Equilibration LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 Upon thermalization: n=n 0, T=> T 0 +  T, p=> p 0 +  p Pressure equilibration: p=>p 0, T=>T 0 +  T, n=> n 0 -  n Thermal diffusion: p=>p 0, T=>T 0 +  T(t), n=> n 0 -  n(t)

11. 11 Steady-State Simulations LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 Assumptions -Post thermalization -Pressure equilibrated throughout gas volume -Use ideal gas law for thermal properties -5 m long, 200 eV, Ar gas, 20 mm diameter, 10 5 attenuation -p=1.43 torr for low power operation -Outer diameter cooled to 300 K

12. 12 Geometry Dependence LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 10 5 attenuation Use larger pipe, 75 mm radius vs. 10 mm -Increases diffusion time ~ L 2 /  is thermal diffusivity -Decrease diffusion length ~ sqrt(t  ) for a given time interval -Temperature gradient must go up

13. 13 Energy (In)Dependence LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 As X-ray energy change, so does attenuation length, thus different pressure/density is required for the same attenuation factor. -5 m long, 251 eV, Ar gas, 20 mm diameter, 10 5 attenuation -p=0.145 torr for low power operation -Outer diameter cooled to 300 K Pressure independence stems from the pressure independence of the thermal conductivity for an ideal gas.

14. 14 Dynamics: Initial Ramp-up & from Steady-State LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 Ramp from initial room temperature (long time constant) From steady-state temperature profile (short time constant) -Speed-up dynamics due to continuing energy transfer throughout gas volume 2 ms

15. 15 Time Dependence at Different z-Positions LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 Z = 0 mZ = 1 m Z = 3 m Z = 4 m Z = 2 m Z = 5 m “Hole burning” never get through

16. 16 Time Dependence at Different r-Positions LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 Steady-state at entrance z = 0 m 100 KHz, 200 W, 1  s steps Immediately after n th pulse Just before after n+1 th pulse Immediately after 1 st pulse Just before after 2 nd pulse Cold start at entrance z = 0 m 100 KHz, 200 W, 1  s steps No steady-state temperature gradient, energy transfer over large distance, thus long time constant (1 ms) Steady-state temperature gradient, energy transfer over short distance, thus short time constant (10  s)

17. 17 Time Dependence at Different r-Positions LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 Steady-state at entrance z = 0 m 100 KHz, 200 W, 1  s steps Immediately after n th pulse Just before after n+1 th pulse Immediately after 1 st pulse Just before after 2 nd pulse Cold start at entrance z = 0 m 100 KHz, 200 W, 1  s steps No steady-state temperature gradient, energy transfer over large distance, thus long time constant (1 ms) Steady-state temperature gradient, energy transfer over short distance, thus short time constant (10  s) 75 mm radius geometry

18. 18 Dependence on Average Power LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 Constant attenuation factor 10 5 Vary input power at entrance of gas attenuator

19. 19 Power Dependence LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 Why effect in LCLS-I low power operation is negligible?

20. 20 Length Dependence LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 Why it is not a concern for European XFEL using a long attenuator? Or should it be?

21. 21 Gas Detector Based on N 2 Photoluminescence LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 High operating pressure ~ 100 mtorr to a few torr Short interaction length ~ 100 mm -Equivalent to at entrance of gas attenuator at entrance of gas attenuator Simulated by maintaining constant pressure

22. 22 What Happens in Gas Monitor Detector (GMD)? LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 Low operating pressure 10 -3 Pascal Constant power, variable attenuation at entrance of gas attenuator

23. Summary LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 23 Gas Attenuators -Heating effect for high average power -Complex pressure-attenuation correlation -But independent of X-ray wavelength -Negligible for LCLS-I operations -Somewhat negligible for European XFEL Gas Detector -Heating effect when operating high pressure -Dependent on average power, making it unreliable if average power varies via rep rate or bunch charge Gas Monitor Detector -Effect completely negligible

25. X-ray Attenuator for HXR Branch 25 Re-purposing existing LCLS-I gas and solid attenuator -Using 4.3 m N 2 gas column up to ~ 50 torr, variable size aperture -For energies 1 to 2 keV at rep rate up to 1 MHz, use gas attenuator -For energies 2 to 5 keV at rep rate up to 1 MHz, use gas (pre- attenuation) and cooled solid (diamond) attenuators -For energies 5 to 25 keV at 120 Hz, use solid (diamond pre-attenuation and Si) attenuators Re-located to downstream

26. 26 Analytical Solution/Estimate LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 One-dimensional simulation by J. Krzywinski T [deg] X [cm] T = 0 °C 0.1 cm FEL Case : 100 KHz operation 1 mJ/pulse 5 keV normal incidence, 1 mm beam size After 2000 shots, time evolution shown in 500  s steps (50 shots) t=0 t=500  s Evolution between shots in steady state (between shot # 2000 and 2001). Time evolution shown in 1  s steps Just after shot #2000 1  s after the shot Just before shot #2001 “zoomed - in”

27. 27 Heating Effects in Gas Attenuator LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 Density “hole” burning L0L0 T0T0 T beam Gas injector ~ 40 cm for 10 5 attenuation 5 m

28. X-ray Attenuator for SXR Branch Note: higher energy X-rays require higher gas pressure, but can operate with smaller apertures LCLS-II Accel. Phys. Mtg, Nov. 12, 2014 28 Expected performance -Use Ar < 1.25 keV -1 st aperture 4 - 10 mm FEL beam size Ar Courtesy of D. Schafer