Design Challenges in PEP-X: An Option for Soft X-Ray FEL

Design Challenges in PEP-X:An Option for Soft X-Ray FEL

R. Hettel for Y. CaiSLAC National Accelerator Laboratory

3rd Low Emittance Ring Workshop July 8-10, Oxford, England

Rep rate100 MHz100 Hz

106

1012

Phot

ons p

er p

ulse

DLSR

FEL-

-

- -

Motivation for enhanced DLSR performance

FELs becoming more ring-like: higher rep rate, reduced photons/pulse

Can rings become more FEL-like: lower rep rate, increased ph/pulse?

Acknowledgements• FEL & Beam Physics Department: Karl Bane, Yuantao Ding, Zhirong Huang, Alexander

Novokhatski, Lanfa Wang • SSRL: Xiaobiao Huang, James Safranek, John Schmerge• Accelerator Design Department: Yuri Nosochkov, Min-Huey Wang• Advance Computing Department: Cho-Kuen Ng, Liling Xiao• SLAC Senior Management: Jerome Hastings, Robert Hettel, Norbert Holtkamp, Chi-Chang Kao

4

PEP-X DLSR (“PEP-Hex”)– Diffraction limited ring for

1.5 Å (e=l/4p = 12 pm)– Good beam lifetime 3

hours– Good injection with 10 mm

acceptance– Achievable machine

tolerances 20 microns – Off-axis injection

Y. Cai, K. Bane, R. Hettel, Y. Nosochkov, M-H. Wang, M. Borland,Phys. Rev. ST Accel. Beams 15, 054002 (2012)

E = 4.5 GeV I = 200 mA ex,y = 12/12 pm-rad54 7BA cells off-axis injection

[Note: most recent version of PEP-X is 6-GeV PEP-X (“PEP-Xtra”)]

• Convert PEP-II hexagonal tunnel to circular geometry

• 7BA lattice (72 cells), 5-m straight sections

• 48 ea 5-m straights available for IDs; 17600 m2 hall needed for 16 beam lines; east and west arcs not practical for beam lines

• 120-m straights available for injection and bypass options

• Cost of larger ring offset by reduced cost for experimental halls

E = 6 GeV I = 200 mA ex,y = 5/5 pm-rad72 7BA cells off-axis injection

6

PEP-X achromats

e ~ 3 ~ C-3

= dipole bend angleC = ring circumference

e= 29 pm at 4.5 GeV (no damping wigglers)

Cell phase advances: mx=(2+1/8) x 3600, my=(1+1/8) x 3600.

Good dynamic aperture reached by cancelation of 3rd- and 4th- order lattice resonances

(Developed from MAX-IV design)

Legend: 0.2km/2GeV: ALS-II, 52 pm0.8km/3GeV: NSLS-III, 30 pm1.1km/6GeV: APS-II, 80 pm2.2km/6GeV: PEP-X, 5 pm6.2km/9GeV: tauUSR, 3 pm

Brightness and Coherence of Future Rings

Current US Rings

Overseas Projects& Plans

US Projects

DLSR Designs• Competitive pressure drives optimization• Upgrades & greenfield facilities possible• 2-3 orders of magnitude improved brightness

over existing rings

DLSR Science complements FELs, offering• Similar high transverse coherence• Long pulses with high repetition rate • High average / low peak power• High stability• High capacity

Current US Rings

Overseas Projects& Plans

US Projects

M. Borland for BESAC Meeting, July 2013

FEL lasing in a long switched bypass

Electron bunch is recycled for 3 damping times in ring after lasing in the bypass

Bunch switched into FEL bypass (10-100 kHz)

Pellegrini et al., 1992

Reduce bunch length from 10 ps to 1 ps without reducing bunch current

Calculation of microwave instability threshold

An illustration using 4.5-GeV PEP-X nominal parameters: frf = 476 MHz, Vrf=8.3 MV, frev = 136.312 kHz, sz=3 mm, Ib=0.067 mA.

Parameter PEP-X (USR) PEP-X (FEL)Beam energy [GeV] 4.5Circumference [m] 2200Current [mA] 200 10*Betatron tune (H/V) 113.23/65.14Momentum compaction 4.96x10-5

Emittance [pm-rad] (H/V) 12/12 160/1.6Bunch length [mm] 3 0.3Energy spread 1.55x10-3

Energy loss per turn [MeV] 2.95RF voltage [MV] 8.3 282.0RF frequency [MHz] 476 1428**Damping time [ms] 18Length of ID straight [m] 5 Use also long straights

(> 100 meter)Beta at ID center (H/V) [m] 4.9/0.8

* Limited by SRF HOMs ** 2- or 3-frequency RF to provide long and short bunches is possible

Transverse Gradient Undulator (TGU)Generate a linear x-dependence of the undulator fields:

For a large energy-spread beam, disperse the beam according to its energy:

Choose dispersion and transverse gradient:

low gain FEL: T. Smith et. al., J. Appl. Phys. 50, 4580 (1979).N. Kroll et. al., IEEE Journal of Quan. Electro. QE-17, 1496 (1981).high gain FEL applying to laser-plasma driven accelerator: Z. Huang, Y. Ding and C. Schroeder, Phys. Rev. Lett. 109, 204801 (2012).

Betatron beam size << dispersed beam size rotate TGU to take advantage of very small vertically coupled emittance in DLSR

IBS growth for 1ps bunch

Flat beam option. Vertical emittance is 1% of the horizontal one.

Energy Spread Horizontal Emittance

0 20 40 60 80 10030

40

50

60

70

z (m)

e-be

am si

ze (m m

)

sy

sx

Electron beam and radiation size

FEL radiation far-field pattern, (Full transverse coherence!)

Dispersion dominated size

Emittance size

Almost round beam

Energy spread along the undulator

s E/0

.511

[MeV

]

Simulation using a modified Genesis

DsE/E=5.5x10-5

Radiation power and spectrum

Saturation is reached with > 200 MW FEL power

For a bunch with sz=1ps, FEL pulse energy is estimated about 0.2 mJ (~2x1012 ph/pulse)

@100m

@100m

PEP-X(FEL) at 1.5nmParameter PEP-X(FEL) LCLS (150pC case)Undulator lu = 3cm, K=3.7 (TGU) lu = 3cm, K=3.5Radiation wavelength 1.5 nm 1.5 nmPulse energy 0.2 mJ (1.6x1012 photons) 2 mJ (1.6x1013 photons)Peak power 200 MW 20 GWPulse length 1 ps 50-150 fs Saturation length 90 m 40 mRepetition rate nb x 100 Hz 120 Hz

Electron norm. emittance 1.45(x)/0.0145 (y) mm mrad 0.5 mm mrad

Electron peak current 300 A 1000-3000 AElectron energy 4.5 GeV 4.3 GeV

Electron energy spread 1.55x10-3 3.1x10-4

Note that nb is number of bunches. Bunches are recycled after three damping times in PEP-X.

SCRF in 12 GeV CEBAF Upgrade

We assume 20 MV/m for accelerating gradient. Three modules are necessary to reach 1ps bunch length.

8 cavities in a cryomodule produce 108 MV

7-cell SCRFPerformance of SCRF (1497 MHz at 2 K)

Measured dipole HOMs in JLAB SRF cavity

Multi-bunch instability driven by 24 JLab SCRF cavities

=76µs

0 500 1000 1500 2000 2500 30000

1

2

3

4

5

x 104

m

grow

th ra

te (1

/sec

)

Horizontal CBI in PEPX with 24 JLAB 7-cell SRF cavities, the fastest mode has =18.6516(ms)

0 500 1000 1500 2000 2500 30000

2

4

6

8

10

12

14

x 104

mgr

owth

rate

(1/s

ec)

Vertical CBI in PEPX with 24 JLAB 7-cell SRF cavities, the fastest mode has =7.4277(ms)

=7µs=18µs

• Beam current: 200mA• Bunch Number: 3492• Beam filling pattern: uniform• Beam Energy: 4.5GeV

Horizontal Vertical

200 bunches with 20mA beam current seems stable based on the current SCRF technology. That provides us 20 kHz repetition rate for the FEL.

R&D plan• 1ps bunch

– Improve design of a 1.5 GHz SCRF cavity similar to those built for CEBAF upgrade

– Build a prototype to demonstrate its performance in terms of accelerating gradient and sufficient damping of HOM

– Install the prototype in SPEAR3 to further quantify its performance with electron beam

• Study how to drive an x-oscillator– TGU in low-gain region with 20A peak current– 10,000 bunches to reach 1MHz repetition rate– Increase beam energy to 6 GeV (PEP-Xtra) and further

reduce emitttance (5/5 pm) or use harmonic lasing

• DLSRs are capable to drive SASE FEL in soft x-ray region to saturation within 100 meter using 1ps bunch in transverse gradient undulators.

• Based on the CEBAF upgrade, three cryomodules with 300 MV accelerating gradient are sufficient to reduce bunch length to 1ps and retain stability of 200 bunches with 20 mA beam current.

• TGU can be applied to accommodate large energy spread in storage rings. It is necessary to rotate the undulator 900, taking advantage of a very small vertical emittance in the ring.

• A feasible SASE FEL at radiation wavelength of 1.5 nm– Achieves full transverse coherence – Provides 0.2-mJ energy or 2x1012 photons per pulse– Has pulse length of 1 ps – Reaches a repetition rate of 10 kHz

• DLSR-based hard X-ray XFELOs might be possible (under study)

Conclusion