1. New Light Source Design Studies Joint Accelerator Workshop, RAL, 20 th Jan 2009 Peter Williams ASTeC, STFC Daresbury Laboratory & Cockcroft Institute (on behalf of the NLS design team)

2. New Light Source Project STFC-led project to examine and propose a 4 th Generation synchrotron user facility for the UK with unique and world leading capabilities (NLS is a working title) Three stages: 1. Science Case 2. Technical Design Study 3. Funding and Location Science Case presented to STFC - PALS on 15 rd Oct (www.newlightsource.org). From executive summary… IMAGING NANOSCALE STRUCTURES: Instantaneous images of nanoscale objects can be recorded at any desired instant allowing, for example, nanometer scale resolution of sub-cellular structures in living systems. CAPTURING FLUCTUATING AND RAPIDLY EVOLVING SYSTEMS: Rapid intrinsic evolution and fluctuations in the positions of the constituents within matter can be characterized. STRUCTURAL DYNAMICS UNDERLYING PHYSICAL AND CHEMICAL CHANGES: The structural dynamics governing physical, chemical and biochemical processes can be followed by using laser pump- X-ray probe techniques. ULTRA-FAST DYNAMICS IN MULTI-ELECTRON SYSTEMS: New approaches to measuring the multi-electron quantum dynamics, that are present in all complex matter, will become possible. Outcome: Phase 2 Technical Design Study to go ahead

3. New Light Source Design Team STFC-led collaboration involving (pretty much) all players in the UK (named persons contributed directly to this talk) Julian McKenzie, Boris Militsyn, Neil Thompson, James Jones, Deepa Angal-Kalinin Riccardo Bartolini, Jang-Hui Han, Ian Martin Hywel Owen

4. NLS Technical Design Study NLS to be a coupled set of facilities including conventional lasers (high field/ultra-short pulse), long wavelength sources and at its core, short wavelength (<100nm) free electron lasers Facility requirements presented in Science Case Photon energies tuneable over the range from THz/IR through to soft X-ray Two-colours (i.e. UV/Vis or IR/THz synchronised with soft X-ray pulses) Ultra-fast pulses with duration down to sub-femtosecond range High temporal and transverse coherence

5. Transverse Brightness  Not a Storage Ring! Linac: 1 mm-mrad (state of the art @ 1 nC charge) TME 6-cell (100 m) TME 12-cell (200 m) TME 48-cell (800 m) TME 24-cell (400 m) Storage Ring Horizontal Emittances Storage Ring Vertical Emittances (0.1% Coupling) TME (theoretical minimum emittance) is the smallest emittance possible in a ring, based on minimising Courtesy Hywel Owen

6. Short Pulses  Strong Bunch Compression A method to compress the length of electron bunches to small values, e.g. less than 1 ps Chirp + Compression, similar to CPA in lasers The chirp is conveniently carried out at the same time as the bunch is accelerated – in a series of radiofrequency cavities in a linear accelerator The compression may be performed in a 4- dipole chicane

7. Compression Scheme Design Non-Trivial BC1BC2 L1L2L33H/4H 00  h =  h = 0 /h Main RF3 rd Harmonic CavityFinal Linearised Chirp Resistive wall wakes Collimator wakes Longitudinal cavity wake E.g. Trade off large CSR in transporting short bunch with jitter caused by large R56

8. 4 th Generation Machines Worldwide Blue – single-stage Red – multi-stage (inc. harmonic correction) Yellow – ERL (various methods) Bold – they have measured that bunch length Initial NLS SC Design Initial NLS NC Design Users want 1kHz rep. rate, 20fs photon pulses, 3 GeV ideally Our initial interpretation, a ~1 GeV SC linac, upgradable to 3 GeV Other initial approach, a 3 GeV NC linac (R. Bartolini) Decision last November to propose a high rep. rate machine ( >1kHz) based upon a Superconducting electron linac – high rep rate pushed by users.

9. Initial NLS SC Concept (Hywel Owen and PW) 735 MeV chosen as it corresponds to 1 nm, the limit for HHG seeding i.e. this is a possible extraction energy where we want short bunches Compression scheme must be carefully designed – linearisation, cavity wakefield compensation, CSR, LSC 200 pC bunch charge chosen, based injector on XFEL EPAC08: MOPC034, MOPC035 available at www.jacow.orgwww.jacow.org ParameterValue Bunch Charge200 pC Fundamental RF1.3 GHz Bunch Rate1 kHz to 1 MHz Gradient17 MV/m 3.9 GHz Total Voltage20 MV Transverse Slice Emittance< 2 mm-mrad rms Energy Spread4.1 MeV Bunch Length10 fs

10. Initial NLS SC Simulation – The Upright Bunch Peak currents (for FEL lasing) 16.6 kA 11.4 kA 11.8 kA CSRtrack 3D CSRtrack Projected Elegant Projected Central Bin Width Slice EmittanceChargeEquivalent Current 1 fs1.32 mm-mrad16.9 pC16.9 kA 5 fs1.43 mm-mrad57.5 pC11.5 kA 10 fs1.91 mm-mrad104 pC10.4 kA ParameterProjected (ELEGANT) 3D Method (CSRtrack) Projected Emittance2.98 mm-mrad4.95 mm-mrad Slice Emittance (5 fs)1.43 mm-mrad1.85 mm-mrad Slice Energy Spread (5 fs)0.29 %0.27 % Peak Current (1 fs)16.6 kA14.5 kA Numerical LSC microbunching CSR e-spread After first compressor After second compressor

11. Initial NLS NC Simulation – R. Bartolini (DLS) S01 GunX01 S02S03S04S05S06S07S08S09S10S11S12 undulators BC1BC2 DL Astra elegant genesis 3 GeV NC S-Band linac with 2 stage compression, 200 pC bunches chosen NC RF S-band gun, 0.21 mm mrad at injector exit – more later CSR – yes, LSC – no. Bin width ~1 fs Long. PS Energy Spread Emittance Current

12. Initial NLS Options Study OptionTechnologyMax. Cavity GradientAverage Bunch Rate 1Normal-conducting 2.856 GHz (pulsed) 26.7 MV/m400 Hz 2Super-conducting 1.3 GHz (CW) 20 MV/m1 kHZ 3Super-conducting 1.3 GHz (CW) 20 MV/m1 MHz OptionEnergy [GeV]Accelerator Length [m]Source Length (up to FEL) [m] 11113193 2158238 3203283 2/31141221 2229309 3317397 OptionEnergy [GeV] Transverse Slice Emittance (norm.) (1fs) [mm-mrad] Peak Current (1fs) [kA] Bunch Length (r.m.s.) [fs] Energy Spread [MeV] 11/2/30.410.2161.0 2/31/2/31.416.6104.1 Difference in emittance due to using CSRTrack 3-D in compressors (slice emittance increased ~30%), Longitudinal space charge and non-optimal injector Showed broadly similar bunch characteristics from both NC and SC linacs – SC chosen due to user demand for high rep. rate.

13. NLS Current Work: Recirculation vs Single Pass Users want high rep. rate ( > 1 kHz)  superconducting machine  capital expense Mitigation strategy – Recirculation Example: Build a 1 GeV SC linac and recirculate to 2 (3) GeV. Possible issues: Compression Scheme (no ~10 fs bunches at high energy – assume that we do NOT NEED electron bunches this short at this stage) Emittance Degradation (CSR, ISR, LSC) due to arcs Beam Break Up High Energy Diagnostics Linearisation Jitter due extra transport BUT – can build upon 4GLS experience Exercise – compare recirculation and single pass for 2.2 GeV, 200fs bunch length, 20MV/m linac @ 1.3 GHz feeding 3 FEL’s. Use identical gun – NCRF L-band gun by Jang-Hui Han (more later)

14. NLS Current Work: Recirculation (PW) Do some simple-minded 1-d longitudinal phase space transformations… an example… Assume a transport rather than dog-bone. Should we minimise CSR in arc by keeping bunch as long as possible in the first pass by putting first compression after all arcs? Answer is no! Cannot linearise. However microbunching MAY require BC1 @ > 250 MeV to combat microbunching resulting from relatively low energy compression Inject at ~200 MeV, 1 st linac pass = 1.2 GeV, 2 nd pass = 2.2 GeV

15. NLS Current Work: Recirculation

17. We need an arc! Regular FODO channel arc was eventually rejected for 4GLS-XUV (size, non-zero R56) Went to a zero R56 compact QBA design GA optimisation algorithm by James Jones, Daresbury Using as starting point of NLS recirculation arcs NLS Current Work: Recirculation in More Detail

18. Floor coordinates for ring – injection and extraction being worked on 3HC, BC1 and inject at ~200 MeV 7 modules take to 1.2GeV, recirculate to 2.2 GeV and extract Need an optics solution for this!! NLS Current Work: Recirculation Machine Model

19. NLS Current Work: Recirculation - Optics

20. NLS Current Work: Recirculation – More optics

21. NLS Current Work: Recirculation – Even More Optics

22. NLS Current Work: Recirculation – Optics Matching

23. NLS Current Work: Recirculation – No, not more optics

24. NLS Current Work: Recirculation – OK, now I’m bored

25. NLS Current Work: Recirculation – Aaargh, not more optics

26. NLS Current Work: Recirculation – Final Optics Slide!

27. NLS Current Work: Recirculation – To Do! Injection Design – think about R56, microbunching, use experience from ALICE & 4GLS designs, CEBAF etc. Extraction Design – ditto Tracking!! Working point optimisation (see single pass work) Additional components? – e.g. Path Length Corrector to Enable independent control of phase on second linac pass Spreader to 3 FEL’s (common to single pass design)

28. NLS Current Work: Single Pass (RB & IM) A01 GunA39 A02A03A04A05A06A09A10A11A12 undulators BC1 BC2 DL Astra elegant genesis A07A08A13 A14 At tracking stage - optimising the beam quality at the beginning of the undulators peak current, slice emittance, slice energy spread Parameters that can be used in the optimisation Accelerating section amplitude and phase 3HC amplitude and phase Bunch compressors strengths Used so far Phase of ACC2-3 Phase of ACC4-8 Amplitude and phase of 3HC BC1 BC2

29. NLS Current Work: Single Pass – Longitudinal Profiles before 3HCafter BC1before undulators One particular tuning: BC2 7.3 deg; best slices I peak  1.2 kA,  n  0.35  m,    5  10 –5 ; 47 fs (rms)

30. NLS Current Work: Single Pass – BC2 Optimisation 10 e - bunches superimposed

31. NLS Current Work: Single Pass – Clever Optimisation A Multi-objective multi-parameter optimisation GA parallel search algorithms -18000 runs with 100K particle each 2 objectives: minimise Xie Length and maximise and P sat 4 parameters: phase of ACC02; ACC4-7, BC1, BC2 Manual optimisation (red dot) xie length 1.24 (m) avg sat power 2030 (MW) (theta2, theta3, phi2, phi4) = (17.5, 6.1, 7.75, 25) Multi-objective optimisation xie length 1.15 (m) (–8%) avg sat power 2220 (MW) (+11%) (theta2, theta3, phi2, phi4) = (17.54, 4.92, 8.96, 29.91)

32. NLS High Rep. Rate (to GHz) Photoinjector Options 1.HV DC gun: Status - operational in user facilities. Experience with technology at DL. Lower emittance due to lower field strength (10MV/m) at cathode. Need XHV vacuum and have HV issues ie ceramic insulator. Can use GaAs photocathodes + others. 2.VHF NC RF gun: Status - design studies (LBNL). ~100MHz gun, similar beam transport/dynamics to DC gun but due to higher field strength (20MV/m) at cathode have lower emittance. NC-RF technology is well established. Cannot use GaAs so use multi-alkali photocathodes such as K2CsSb. 3.SRF gun: Status - under commissioning (ELBE). Don’t require a buncher/booster. Therefore less timing jitter but less tuneability. Perfomance limit is the amount of power you can couple in. Up to 50MV/m should be possible giving very low emittance beam. Cannot use GaAs, probably use Cs2Te DCVHFSRF Projected emittance (mm·mrad)1.951.080.84 Slice emittance (mm·mrad)1.20.80.4 Bunch length (mm)1.721.31.67 Longitudinal emittance (keV·mm)295115198 Beam energy (MeV)120117118 Linac 2008 TUP042: Boris Militsyn, Carl Beard, Julian McKenzie – Daresbury

33. NLS Low Rep. Rate (to kHz) Photoinjector Options Thanks to Jang-Hui Han, Diamond 1.NC RF gun at L-Band well proven at PITZ. Up to 50MV/m, upgradable to 1kHz. Transverse emittance: 0.68 mm mrad @ 1 nC and 0.33 mm mrad @ 0.2 nC 2.NC RF gun at S-Band scaled down from DESY L-band gun. Cooling-water channel redesigned  400 Hz rep. rate. Up to 120MV/m field strength at cathode. Transverse emittance: 0.42 mm mrad @ 1 nC and 0.21 mm mrad @ 0.2 nC. 3.SCSS style thermionic gun with CeB6 cathode. Only 60Hz at present.

34. Summary NLS project now in technical design phase Wide range of options for major systems being studied (injectors, linac, FELs, layout) Recirculation vs. Single Pass decision in February Other technology selections will be made in next few months  Detailed proposal by October 2009