1. PAC2001 J. Rossbach, DESY 1 New Developments on Free Electron Lasers Based on Self-amplified Spontaneous Emission J. Rossbach, DESY Why SASE FELs? How does it work? What are the challenges? Where are we? Where do we want to go?

2. PAC2001 J. Rossbach, DESY 2 Why SASE FELs? State of the art: Structure of biological macromolecule Needs  10 15 samples Crystallized  not in life environment reconstructed from diffraction pattern of protein crystal: LYSOZYME, MW=19,806 The crystal lattice imposes restrictions on molecular motion Images courtesy Janos Hajdu

3. PAC2001 J. Rossbach, DESY 3 Why SASE FELs? SINGLE MACROMOLECULE, Planar section, simulated image courtesy Janos Hajdu Resol. does not depend on sample quality Needs very high radiation power @  1Å Can see dynamics if pulse length < 100 fs

4. PAC2001 J. Rossbach, DESY 4 We need a radiation source with · very high peak and average power · wavelengths down to atomic scale λ ~ 1Å · spacially coherent · monochromatic · fast tunability in wavelength & timing · sub-picosecond pulse length Why SASE FELs? For wavelengths below ~150 nm: SASE FELs.

5. PAC2001 J. Rossbach, DESY 5 How does it work? Q = N e ·e, N e = # electrons Point charge radiates coherently P  N e 2 ! Radiation power of oscillating point-like charge Q: P  Q 2   2 „Point“ means above all: bunch length < radiation Synchrotron radiation of an incoherent electron distribution: P  N e  Potential gain in power N e = 10 9 – 10 10 !!

6. PAC2001 J. Rossbach, DESY 6 How does it work? Coherent motion is all we need !!

7. PAC2001 J. Rossbach, DESY 7 How does it work? Idea: Start with an electron bunch much longer than the desired wavelength and find a mechanism that cuts the beam into equally spaced pieces automatically Free-Electron Laser (Motz 1950, Phillips ~1960, Madey 1970) Special version: starting from noise (no input needed) Single pass saturation ( no mirrors needed) Self-Amplified Spontaneous Emission (SASE) (Kondratenko, Saldin 1980) (Bonifacio, Pellegrini 1984)

8. PAC2001 J. Rossbach, DESY 8 How does it work? Spectrum of amplified spontaneous radiation Resonance wavelength:

9. PAC2001 J. Rossbach, DESY 9 How does it work? FIREFLY microbunching; Ricci,Smith/Stanford

10. PAC2001 J. Rossbach, DESY 10 How does it work? 10 5 by FEL gain 10 3 by improved beam quality, long undulators

11. PAC2001 J. Rossbach, DESY 11 What are the challenges?Overview Electron beam parameters needed for Self-Amplified-Spontaneous Emission (SASE) Energy: für em = 1 Å: E  20 GeV Energy width: Narrow resonance   E /E ≤ 10 -4  Small distortion by wakefields  super conducting linac ideal! Gain Length: Beam size:  r  40  m  high electron desity for maximum interaction with radiation field Emittance  ≤  need special electron source to accelerate the beam before it explodes due to Coulomb forces Peak current inside bunch: Î > 1 kA feasible only at ultrarelativistic energies, otherwise ruins emittance  bunch compressor Straight trajectory in undulator: ultimately < 10  m over 100 m

12. PAC2001 J. Rossbach, DESY 12 Why a linear accelerator? X-ray SASE FEL needs: energy width σ E /E ≤ 10 -4 and bunch length σ l  25  m (~100 fs)  σ E  σ l  60 eV m storage ring is limited to >1000 eV m electron emittance  ≤   10 -11 m LEP (20GeV) (!):  x > 10 -10 m several kA peak current wakefields tolerable for single pass, BUT not in storage ring

13. PAC2001 J. Rossbach, DESY 13 What are the challenges?RF gun TESLA FEL photoinjector for small and short electron bunches

14. PAC2001 J. Rossbach, DESY 14 What are the challenges?Injector Layout of integrated injector/compressor for TTF2 and TESLA FEL

15. PAC2001 J. Rossbach, DESY 15 What are the challenges?Bunch compression Beware of coherent synchrotron radiation (CSR) Magnetic bunch compression Beam dynamics simulation must take into account combined space charge and e.m. radiation in near-field. see: TRAFIC4 by A. Kabel/SLAC

16. PAC2001 J. Rossbach, DESY 16 What are the challenges?Bunch compression y-z streak generated by deflector P. Krejcik et. al., WPAH116 P. Emma, J. Frisch, P. Krejcik, G. Loew, X.-J. Wang f = 2856 MHz V 0  15 MV  z  22  m f = 2856 MHz V 0  15 MV  z  22  m eeee zzzz 2.44 m cccc pppp   90° V(t)V(t)V(t)V(t) xxxxRF‘streak’ S-band Structures built at SLAC in 1960’s  now installed in linac for testing ‘slice’-  and ‘slice’ energy spread measurements also possible

17. PAC2001 J. Rossbach, DESY 17 What are the challenges?Bunch compression Interferometry of coherent synchrotron radiation Projection from longitudinal phase space tomography Longitudinal electron bunch profile at the TESLA Test Facility measured with two different methods

18. PAC2001 J. Rossbach, DESY 18 What are the challenges?Bunch compression Bunch compression down to few 20-30  m is a technical requirement (and complication) to achieve kA peak current for sufficiently small gain length. It is a lucky coincidence, that the ultra-short pulse length is exactly what users are calling for. From the user point of view, bunch length should be even  10  m !  try harder! 

19. PAC2001 J. Rossbach, DESY 19 What are the challenges?Wakefields Wakefields from surface roughness: Test at TTF FEL Smooth surface Rough surface, same diameter E E See Markus Hüning, Wed. afternoon

20. PAC2001 J. Rossbach, DESY 20 Where are we? Beam parameters TTF FEL nowTESLA FEL (LCLS similar) Normalized emittance from gun (Q = 1 nC)3.5 mrad mm0.8 mrad mm Norm. emittance at undulator entrance8 mrad mm1.6 mrad mm Beam size in undulator 100  m40  m Bunch length (rms) 1 ps0.1 ps Peak current500 A5000 A Long. emittance σ E  σ l 100 eV m60 eV m In all key beam parameters, the extrapolation from proven technology is a factor 2 – 10 We know what to do and how We will take further steps at TTF getting even closer to TESLA FEL parameters

21. PAC2001 J. Rossbach, DESY 21 Where are we? Progress with SASE FELs: VISA see: Tremain,Murokh WPPH118/122 Wed. afternoon

22. PAC2001 J. Rossbach, DESY 22 Where are we? Progress with SASE FELs: LEUTL

23. PAC2001 J. Rossbach, DESY 23 Where are we? Progress with SASE FELs: LEUTL 530 nm Energy vs. Distance along the Undulator Exponential Growth Region Saturation of SASE Flash of UV light (385 nm) near saturation. The expected wavelength as a function of angle (radial offset) is clearly seen. The darker “lines” are from shadows of secondary emission monitors in the vacuum chamber. Stephen Milton/ANL Tuesday 13:30h

24. PAC2001 J. Rossbach, DESY 24 Where are we? Progress with SASE FELs: TESLA Phase 1 of the SASE FEL at the TESLA Test Facility at DESY, Hamburg. The total length is 100 m.

25. PAC2001 J. Rossbach, DESY 25 Where are we? Progress with SASE FELs: TESLA TTF FEL undulator

26. PAC2001 J. Rossbach, DESY 26 Where are we? Progress with SASE FELs: TESLA SASE gain >1000 Spontaeous Emission x100 TTF FEL gain at 108 nm vs. bunch charge By now observed gain >10 5

27. PAC2001 J. Rossbach, DESY 27 Where are we? Progress with SASE FELs: TESLA

28. PAC2001 J. Rossbach, DESY 28 Where are we? Progress with SASE FELs: TESLA FEL wavelengths reached at TTF FEL

29. PAC2001 J. Rossbach, DESY 29 Where are we? Progress with SASE FELs: Summary where wavelength year Livermore ~1 mm 1986 LURE/Orsay 5-10  m 1997 UCLA/LANL 12  m 1998 LEUTL/Argonne 530 nm 1999 385 nm & saturation 2000 TTF FEL/DESY 80-180 nm 2000 VISA/ BNL/LLNL/SLAC/UCLA 845 nm saturation 2001 (+2 nd +3 rd Harmon.) All observations agree with theoretical expectations/computer models

30. PAC2001 J. Rossbach, DESY 30 Where do we want to go? SASE FEL projects under progress: min. wavelength APS/LEUTLPhase2120 nm APS/LEUTLPhase3 51 nm DESY: TTF FEL Phase2 6 nm2003/2004 SPring8: ~ 5 nm - 2005 SASE FEL projects proposed: SLAC: LCLS  0.15 nm  2006 DESY: TESLA XFEL  0.085 nm  2010

31. PAC2001 J. Rossbach, DESY 31 Where do we want to go?Brilliance Peak brillianceAverage brilliance LCLS multibunch LEUTL TTF FEL

32. PAC2001 J. Rossbach, DESY 32 Where do we want to go?LCLS

33. PAC2001 J. Rossbach, DESY 33 Where do we want to go?LCLS SLAC linac tunnelundulator hall Linac-0 L  6 m Linac-1 L  9 m  rf  38° Linac-2 L  330 m  rf  43° Linac-3 L  550 m  rf  10° BC-1 L  6 m R 56  36 mm BC-2 L  24 m R 56  22 mm DL-2 L  66 m R 56 = 0 DL-1 L  12 m R 56  0 undulator L  120 m 7 MeV  z  0.83 mm    0.2 % 150 MeV  z  0.83 mm    0.10 % 250 MeV  z  0.19 mm    1.8 % 4.54 GeV  z  0.022 mm    0.76 % 14.35 GeV  z  0.022 mm    0.02 %...existing linac new RF gun 25-1a 30-8c 21-1b 21-1d 21-3b 24-6d X Linac-X L  0.6 m  rf =  Producing short bunches for LCLS

34. PAC2001 J. Rossbach, DESY 34 28 GeV Existing bends compress to <100 fsec ~1 Å Add 12-meter chicane compressor in linac at 1/3-point (9 GeV) Damping Ring (   30  m) 9 ps 0.4 ps <100 fs 50 ps SLAC Linac 1 GeV 20-50 GeV FFTB RTL 30 kA 80 fsec FWHM 1.5% Short Bunch Generation in the SLAC Linac Compress to 80 fsec in 3 stages P. Emma et. al., P. Emma et. al., FPAH165 New proposal for:  LCLS accelerator optics R&D  Ultra-short x-ray science program at SLAC

35. PAC2001 J. Rossbach, DESY 35 Where do we want to go?TESLA TESLA scheme

36. PAC2001 J. Rossbach, DESY 36 Where do we want to go?TESLA Beam switchyard distributing the electron bunch trains to various undulators

37. PAC2001 J. Rossbach, DESY 37 Where do we want to go?TESLA

38. PAC2001 J. Rossbach, DESY 38 Where do we want to go?TESLA A potential site for TESLA near Hamburg

39. PAC2001 J. Rossbach, DESY 39 Conclusion SASE FELs clearly demonstrated for wavelengths far below the visible. Full agreement with theory User facilities in the VUV/soft X-ray range just around the corner User facilities in the Angstrøm range are feasible with only moderate extrapolation of present state-of-the-art; Computer simulations and mechanical design are available Accelerator physics & technology will play major role Fun guaranteed!