New Light Source Design StudiesJoint Accelerator Workshop, RAL, 20th Jan 2009

Peter Williams

ASTeC, STFC Daresbury Laboratory & Cockcroft Institute

(on behalf of the NLS design team)

New Light Source Project

STFC-led project to examine and propose a 4th Generation synchrotron user facility for the UK with unique and world leading capabilities (NLS is a working title)

Three stages:

1. Science Case2. Technical Design Study3. Funding and Location

Science Case presented to STFC - PALS on 15rd 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

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

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

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 EmittancesStorage Ring

Vertical Emittances

(0.1% Coupling)

TME (theoretical minimum emittance) is the smallest emittance possible in a ring, based on minimising 22 ''2 xxxx DDDDH

dsH

I35 Courtesy Hywel Owen

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

Compression Scheme Design Non-Trivial

BC1 BC2

L1 L2 L33H/4H

0 h=

h=0/h

Main RF 3rd Harmonic Cavity Final 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

4th 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.

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.org

Parameter Value

Bunch Charge 200 pC

Fundamental RF 1.3 GHz

Bunch Rate 1 kHz to 1 MHz

Gradient 17 MV/m

3.9 GHz Total Voltage 20 MV

Transverse Slice Emittance < 2 mm-mrad

rms Energy Spread 4.1 MeV

Bunch Length 10 fs

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 Emittance

Charge

Equivalent Current

1 fs 1.32 mm-mrad

16.9 pC

16.9 kA

5 fs 1.43 mm-mrad

57.5 pC

11.5 kA

10 fs 1.91 mm-mrad

104 pC 10.4 kAParameter Projected (ELEGANT)

3D Method (CSRtrack)

Projected Emittance 2.98 mm-mrad

4.95 mm-mrad

Slice Emittance (5 fs) 1.43 mm-mrad

1.85 mm-mrad

Slice Energy Spread (5 fs) 0.29 % 0.27 %

Peak Current (1 fs) 16.6 kA 14.5 kANumerical LSC microbunching

CSR e-spread

After first compressor

After second compressor

Initial NLS NC Simulation – R. Bartolini (DLS)

S01Gun X01S02 S03 S04 S05 S06 S07 S08 S09 S10 S11 S12 undulators

BC1 BC2 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

Initial NLS Options Study

Option Technology Max. Cavity Gradient Average Bunch Rate1 Normal-conducting

2.856 GHz (pulsed)26.7 MV/m 400 Hz

2 Super-conducting1.3 GHz (CW)

20 MV/m 1 kHZ

3 Super-conducting1.3 GHz (CW)

20 MV/m 1 MHz

Option Energy [GeV] Accelerator Length [m] Source Length (up to FEL) [m]1 1 113 193

2 158 2383 203 283

2/3 1 141 2212 229 3093 317 397

Option Energy[GeV]

Transverse Slice Emittance(norm.) (1fs) [mm-mrad]

Peak Current

(1fs) [kA]

Bunch Length(r.m.s.) [fs]

Energy Spread[MeV]

1 1/2/3 0.4 10.2 16 1.02/3 1/2/3 1.4 16.6 10 4.1

• 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.

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)

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, 1st linac pass = 1.2 GeV, 2nd pass = 2.2 GeV

NLS Current Work: Recirculation

NLS Current Work: Recirculation

• 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

• 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

NLS Current Work: Recirculation - Optics

NLS Current Work: Recirculation – More optics

NLS Current Work: Recirculation – Even More Optics

NLS Current Work: Recirculation – Optics Matching

NLS Current Work: Recirculation – No, not more optics

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

NLS Current Work: Recirculation – Aaargh, not more optics

NLS Current Work: Recirculation – Final Optics Slide!

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)

NLS Current Work: Single Pass (RB & IM)

A01Gun A39A02 A03 A04 A05 A06 A09 A10 A11 A12 undulators

BC1 BC2 DL

Astra elegant genesis

A07 A08 A13 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

NLS Current Work: Single Pass – Longitudinal Profiles

before 3HC after BC1 before undulators

• One particular tuning: BC2 7.3 deg; best slices Ipeak 1.2 kA, n 0.35 m, 510–5; 47 fs (rms)

NLS Current Work: Single Pass – BC2 Optimisation

0

20

40

60

80

100

120

140

160

180

6.4 6.6 6.8 7 7.2 7.4 7.6 7.8

BC2 compressor angle (deg)

bu

nc

h l

en

gth

rm

s (

fs)

0

20

40

60

80

100

120

6.4 6.6 6.8 7 7.2 7.4 7.6 7.8

BC2 compressor angle (deg)

no

rma

lis

ed

H e

mit

tan

ce

(u

m)

0

1000

2000

3000

4000

5000

6000

7000

8000

6.4 6.6 6.8 7 7.2 7.4 7.6 7.8

BC2 compressor angle (deg)

pe

ak

cu

rre

nt

(A)

10 e- bunches superimposed

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 Psat

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)

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

DC VHF SRF

Projected emittance (mm·mrad) 1.95 1.08 0.84

Slice emittance (mm·mrad) 1.2 0.8 0.4

Bunch length (mm) 1.72 1.3 1.67

Longitudinal emittance (keV·mm) 295 115 198

Beam energy (MeV) 120 117 118

Linac 2008 TUP042: Boris Militsyn, Carl Beard, Julian McKenzie – Daresbury

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

0 5 10 15 20

z (m)

siz

e (

mm

) o

r e

mitta

nce

(m

m m

rad

)

beam size (0.2 nC)

emittance (0.2 nC)

beam size (1 nC)

emittance (1 nC)

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.

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