CLARA: A new particle accelerator test facility for the UK

CLARA: A new particle accelerator test facility for the UK Jim Clarke STFC Daresbury Laboratory and The Cockcroft Institute on behalf of the CLARA & VELA Project Teams RHUL Particle Physics Seminar, 25 th February 2015

2003 1997 2006 2012 2009

Science and Technology of Future Light Sources: A White Paper. SLAC-R-917 10 8 10 10

10 2 10 4 10 3 10 7 Science and Technology of Future Light Sources: A White Paper. SLAC-R-917

Chapman et al, Nature 470, 73, 2011 Example FEL experiment

What is an FEL? (Self Amplified Spontaneous Emission, SASE)

P (W) P(E ph ) (arb. units) Typical FEL Output D:\Documents\Genesis_Shift\CLARA\08-Jan-2015_14-37-28_SASE_100nm_\SASE_100nm_12.out.h5 x 10 8 D:\Documents\Genesis_Shift\CLARA\08-Jan-2015_14-37-28_SASE_100nm_\SASE_100nm_12.out.h5 x 10 11 7 6 2 5 1.5 4 3 1 2 1 0.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 s (m) x 10-4 Distance along photon pulse 12.1 12.2 12.3 12.4 12.5 12.6 12.7 12.8 12.9 13 E ph (ev) Photon Energy

Free Electron Lasers FELs have made huge advances in the past few years First X-ray FEL in 2009 (LCLS) then SACLA in 2011 More X-ray facilities are under construction Advanced soft X-ray facilities are also now operating routinely for users as well (FLASH & FERMI) The potential for improvements is enormous Increased wavelength and intensity stability Tighter synchronisation between FEL & external lasers Ultra-short pulses of light (sub-fs) Increased temporal coherence (monochromaticity) Increased power Two-colour or Multi-colour output ( Tailored Pulses )

The need for an FEL Test Facility There are many areas where FEL output can be improved: Shorter Pulses Improved Temporal Coherence Tailored Pulse Structures Stability & Power There are many ideas for achieving these aims, but these ideas need to be tested Beamtime on FELs is over subscribed by users and so little time for R&D

CLARA CLARA will be a purpose built dedicated flexible FEL Test Facility Capable of testing the most promising new schemes We have strategically decided to focus on stability, synchronisation, and ultra short pulse generation We are focussing on the longer term capabilities of FELs, not short term incremental improvements Taking FELs into a new regime By demonstrating these goals we will have to tackle all the challenges currently faced by state of the art FELs (and a few more!)

CLARA Compact Linear Accelerator for Research and Applications An upgrade of the existing VELA Photoinjector Facility at Daresbury Laboratory to a 250MeV Free-Electron Laser Test Facility Proof-of-principle demonstrations of novel FEL concepts and development of future accelerator technologies Emphasis on Stability, Synchronisation and Ultra-short Pulse Generation Strathclyde INFN Frascati SwissFEL CERN INR Moscow DLS Oxford Liverpool Lancaster Imperial

Other Goals and Benefits of CLARA The opportunity for R&D on advanced technologies: New photoinjector technologies Novel undulators (short period, cryogenic, superconducting.) New accelerating structures: X-Band etc... Advanced diagnostics The enhancement of VELA beam power and repetition rate, enabling additional industrial applications. ASTeC CERN FERMI@Elettra RHUL INFN Frascati SwissFEL DLS Lancaster INR Moscow The possibility to use the electron beam for other scientific research applications: ASTeC RHUL Oxford ASTeC Strathclyde Manchester INFN Frascati ASTeC York Swansea UCL Manchester ASTeC RHUL Liverpool Lancaster Manchester Oxford

FEL Concepts CLARA will be a flexible test facility enabling the broad range of accelerator and FEL R&D necessary to ensure a future UK FEL facility is world leading. Many of the FEL research topics are in two main areas which are intended to demonstrate improvement of FEL output beyond that available from SASE The generation of ultra-short pulses Our emphasis for the short pulse schemes is to generate pulses with as few optical cycles as possible. For these schemes we will lase at 400 250 nm, where suitable nonlinear materials for single shot pulse profile characterisation are available. Improvement of temporal coherence For these schemes we will lase at 266-100nm because here only spectral characterisation is required. In all cases, we aim to study the essential physics of the schemes which is independent of the FEL wavelength. Using a hybrid planar undulator, with minimum gap 8mm, and tuning range of 400 100 nm, the required electron beam energy is ~230 MeV.

FEL Layout + Operating Modes 1.1m GAPS, with QUAD, BPM, SCREEN + DELAY CHICANE 7 x 1.5m RADIATORS, 27mm PERIOD, ON-AXIS FIELD HORIZONTAL Seeding Mode is for Short Pulse Schemes FEL lasing: 400-250nm (Seed: 30-120µm) + Temporal Coherence Schemes FEL lasing: 266-100nm (Seed: 800nm)

power (MW) Short Pulse Scheme Example: Sliced Chirped Beam + Taper* Principle of scheme Few-cycle laser interacts with electron beam to generate strong energy chirp in short region of bunch Radiator taper is matched to energy chirp to maintain resonance as FEL pulse slips forwards to electrons with different energies FEL gain strongly suppressed in remainder of bunch 500 400 average 300 200 * E. Saldin et al., PRST-AB. 9, 050702, (2006) Slide courtesy Ian Martin, DLS 100 0 0 500 1000 1500 s (fs)

Short Pulse Scheme Example : Mode-Locking* A Recipe for Mode-Locking 1. Take a NORMAL FEL with a long sectional undulator 2. Add electron beam delays to produce axial mode structure: Mode Coupling 3. Add a periodic electron bunch modulation which gives each mode sidebands which then overlap neighbouring modes, causing phases to lock: Mode Locking! * N. R. Thompson & B. W. J. McNeil, Mode-Locking in a Free-Electron Laser Amplifier, PRL 100, 203901 (2008)

Short Pulse Scheme Example : Mode-Locking on CLARA CDR Lattice, 30um energy modulation CDR lattice, 120um energy modulation

Short Pulse Scheme Example : Mode-Locked Afterburner* Hard X-Ray @0.1nm 700zs Pulse Duration (rms) CLARA Parameters @100nm 700as Pulse Duration (rms) * D. J. Dunning, B. W. J. McNeil & N. R. Thompson, Few-Cycle Pulse Generation in an X-Ray FEL, PRL 110, 104801 (2013)

How short is an attosecond? An electron bound to a hydrogen atom takes 100 attoseconds to orbit the atom In 1 second it would orbit the Earth In other words: 1 attosecond is to 1 second as 1 second is to 32 billion years

CLARA Short Pulse Schemes: Predicted Pulse Durations Typical FEL output is ~100,000 cycles

Temporal Coherence Scheme Example: High-Brightness (HB) SASE* As in the mode-coupled FEL, delays are used between undulator modules Each delay is different to prohibit growth of modes Increased slippage gives increased communication length between radiation and electrons, delocalising the collective FEL interaction and allowing coherence length to grow exponentially by up to 2 orders of magnitude (compared to SASE) In contrast to other schemes for improving temporal coherence: No seed laser or photon optics are required Applies at Any Repetition Rate and Any Wavelength. Was demonstrated (over a limited parameter range) on LCLS, using detuned undulators as delays, and shown to reduce linewidth in inverse proportion to the increased slippage they can t test optimum scheme * B. W. J. McNeil, N. R. Thompson & D. J. Dunning, Transform-Limited X-Ray Pulse Generation from a High-Brightness SASE FEL, PRL 110, 134802 (2013)

P(E ph ) (arb. units) P(E ph ) (arb. units) Temporal Coherence: HB-SASE CLARA, 100nm output, 100A peak current bunch D:\Documents\Genesis_Shift\CLARA\14-Jan-2015_16-02-11_SASE_I_100_\SASE_I_100_22.out.h5 x 10 10 16 14 12 10 8 6 SASE 4 2 0 12 12.1 12.2 12.3 12.4 12.5 12.6 12.7 12.8 12.9 13 E ph (ev) D:\Documents\Genesis_Shift\CLARA\14-Jan-2015_12-09-07_S_5_D_1_I_100_\S_5_D_1_I_100_14.out.h5 x 10 10 12 10 8 6 HB-SASE 4 2 0 12 12.1 12.2 12.3 12.4 12.5 12.6 12.7 12.8 12.9 13 E ph (ev)

Old layout FEL Layout Revised 1.1m GAPS, with QUAD, BPM, SCREEN + DELAY CHICANE 7 x 1.5m RADIATORS, 27mm PERIOD, ON-AXIS FIELD HORIZONTAL Better optimised for mode locking and HB-SASE Radiator undulators reduced to 0.75m, spacing reduced to 0.5m Two modulators instead of one Total length the same Radiator undulators New layout 25m

CLARA Layout FEL OUTPUT STUDIED The existing VELA RF Photoinjector Facility Notes: 1. CLARA uses European RF frequencies 2. X-band cavity is used for phase space linearisation 3. Total length ~90m

CLARA Layout We would like to replace Linac 4 with X-band linac as test bed for CLIC technology on FELs collaboration with CERN Linac 1 is 2m long, S-band on order from Research Instruments Linacs 2 to 4 are 4m long S-band devices previously installed on SwissFEL Injector Test Facility just delivered to Daresbury

Feb 2015

CLARA Accelerator Schematic S EU gun, S EU linac, X EU linearizing cavity Space reserved for laser heater Dedicated diagnostics sections Extraction to VELA at <50 MeV Extraction at full energy for other uses, eg Plasma Acceleration

CLARA Simulation - codes ASTRA (including wake fields) ELEGANT ELEGANT+CSRTrack+ELEGANT ASTRA+CSRTrack+ASTRA --- ongoing* Genesis Puffin *Collaboration with Bas van der Geer on implementing CSR and wakefields in GPT, will use GPT to benchmark when complete.

Beam Optics

Linearisation and Compression Reserved space for laser heater pending microbunching studies XEU cavity for phase space linearisation Variable Bunch Compressor

Tracked Bunch - Entrance of FEL Optimised for 250 fs flat region to suppress effect of seed to bunch jitter

TDC after FEL Transverse Deflecting Cavity after FEL enables diagnostics of bunching and energy modulation Can be correlated with photon output Simulated beam image with deflecting voltage of 7.5 MV.

High Repetition Rate Gun for CLARA 400Hz gun (ASTeC, Lancaster, INR Moscow) 1.5-cell gun to operate at fields of up to 120 MV/m with a 100 Hz repetition rate or 100 MV/m at up to 400 Hz repetition rate. Peak power of up to 10 MW and an average power of 10 kw. Max beam energy ~5 MeV Photocathode exchange system to accept cathode plugs of the INFN/DESY type An RF probe is included for active monitoring and feedback Symmetric H-shaped coupler for precise control of the phase of the incoming RF along both feeds. Gun cavity is being manufactured now

CLARA Status JINST 9 (2014) T05001 The Conceptual Design was published in July 2013 SwissFEL have provided the required 3 linacs, together with a number of quadrupoles and solenoids The project has now been split into Two Phases PHASE 1 Front End, 50 MeV This is happening now, with procurement progressing, and installation in 2015. First electrons February 2016. Will enable access to bright, short, up to ~50 MeV electron bunches for UK accelerator science and technology community Will enable new 400 Hz photoinjector to be characterised with beam whilst VELA/CLARA Phase 1 still operational (i.e. two guns) Potential for early exploitation of 20 TW laser PHASE 2a 150 MeV, up to bunch compression section Funded, procurement starting this year PHASE 2b 250 MeV FEL Test Facility Not Yet Funded CLARA is a stated priority of STFC and additional funds are being requested from BIS

VELA VELA Specification Beam Energy 4-6 MeV Bunch Charge 10-250 pc Bunch length (σ t,rms ) 1 10 ps Normalised emittance 1-4 µm Beam size (σ x,y,rms ) 0.5-5 mm Energy spread (σ e,rms ) 1 5 % Bunch repetition rate 1 10 Hz *Not all parameters achievable simultaneously Notes: 1. VELA gun is from Strathclyde (ALPHA-X) 2. Max rep rate is 10 Hz but laser and RF capable of 400 Hz 3. 400 Hz gun being manufactured now

VELA Beam Momentum Measurement Power@klystron (Old) 8.7MW, at gun window = 6.7 MW (~23% loss) Measured Beam Momentum = 4.9 MeV/c (simulations predict 5.7 MeV/c) Power@klystron (New) 10 MW, at gun window = 8.7 MW (~13% loss) Measured Beam Momentum = 5.1 MeV/c (simulations predict 6.5 MeV/c) Missing power in the gun?

VELA Transverse Deflecting Cavity S-band Transverse Deflecting Cavity installed and conditioned with RF Will enable much more detailed diagnosis of bunches (slice properties, bunch length with fs resolution)

VELA TDC Commissioning A transverse kick of around 5 MV is required to achieve 10 fs resolution hence a 9-cell design has been chosen. First beam streaking 5 th Dec 14 TDC-off Amp10 Amp30 Amp40

Cavity BPM development Slot-coupled cavity C-band (6.5 GHz), later X-band Single CBPM + reference cavity next run 3 CBPMs later First scans using Single CBPM

Ultrafast Electron Diffraction Studying structural evolution in fs regime Synergy with FEL Structural Science Sample Chamber Structure at t=0 Pre-pump Monitor change to single shot diffraction pattern Faraday Cup Image intensifying camera with single photon sensitivity to capture diffraction pattern This sensitivity will allow single shot diffraction pattern to be recorded with a 1 pc ultrashort bunch Structure at t=t1 Post pump

Ultrafast Electron Diffraction Sample Chamber Faraday Cup Image intensifying camera with single photon sensitivity to capture diffraction pattern This sensitivity will allow single shot diffraction pattern to be recorded with a 1 pc ultrashort bunch

Faraday Cup chamber and Lanex screen detection chamber. Screen located 3.4 m downstream of sample First Results: September 2014 Charge at detector <<1 pc Au sample 1000 shots Pt sample Single shot Reconfiguring VELA in future will allow <100 fs time resolution.

Rapiscan: First VELA Industrial Users

First VELA Users Scintillator Detector Compton Photon X-ray Sample Checking feasibility of utilising Time of Flight information to create 3D X-Ray Compton Scatter Imaging. Allows for the development of a cargo scanner which can reconstruct 3D images using a scanner and detector mounted on only one side of the container. 45

VELA + CLARA Phase 1 (2015) ~5 MeV ~50 MeV ~50 MeV ~5 MeV ~50 MeV Phase 1 parameters: Max Energy ~50 MeV Max Charge 250 pc Norm. Emittance <1 mm mrad Min Bunch Length 50fs (rms), (10 MeV) Max Peak Current 2kA Bunches/RF pulse 1 Pulse Rep Rate 10 Hz (400Hz later)

VELA + CLARA Phase 1 (2015) ~50MeV ~50MeV ~5 MeV

CLARA to VELA Dogleg 1.25 m offset between CLARA and VELA lines Lozenge second dipole for merge to VELA and straight on for spectrometer line for CLARA and VELA guns. CLARA VELA

Beam Parameters in VELA with CLARA Phase 1 Beam properties in the VELA user areas when operating with CLARA Phase 1 (example: 100 pc bunch at ~50 MeV) There will be a possibility of using the beam straight on from Linac 1 before the rest of CLARA is constructed (example: 100 pc bunch at ~15 MeV)

Multi User Station Flexible end station being installed now to allow rapid reconfiguration of experiments on optical breadboard inside large vacuum vessel Combining electron beam with up to 20 TW laser is also a new capability

Feb 2015

Summary CLARA will be an FEL Test Facility for the UK Accelerator Community Normal Conducting RF Linacs (up to 250 MeV) Emphasis on stability, synchronisation, and ultra short pulse generation Enabling other electron beam applications Major upgrade to VELA VELA is an operational RF Photoinjector with two user areas Generating ~4.5 MeV bunches for use by industry and academia Electron diffraction station taking data now Other accelerator experiments in progress CLARA Phase 1 (50 MeV) will be installed in 2015 Phase 2 not yet fully funded UK FEL aspiration to hard X-ray implies Normal Conducting RF CLARA ideal for technique and technology tests and could effectively duplicate front section of multi-gev FEL CLARA will be a unique facility for the UK Accelerator Community and is an essential stepping stone towards a UK FEL

2018? Thanks for your attention