1

XFEL Oscillator in ERLs

R. Hajima

Japan Atomic Energy Agency

March 4, 2010.

2

X-ray FEL Oscillator

K-J. Kim et al., PRL (2008), PRST-AB (2009)

Typical electron beam parameters for XFELO

Energy, charge, emittance, repetition are similar to ERL beams.

R. Hajima

3

ERL Injector Performancedesign example

I.V. Bazarov et al., PRST-AB (2005)

Ultimate Injector

8 pC, 0.1 mm-mrad will be feasibly obtained.

500-kV DC gun, 3-D pulse shaping, and 10-MeV SCA.

80 pC, 0.1 mm-mrad will be obtained.

750-kV DC gun, 3-D pulse shaping, and 10-MeV SCA.

R. Hajima et al.Compact ERL CDR (2008)

3R. Hajima

4

Integration of XFELO and ERL

straight section of the loop additional branch

Beam dump

Location

share an ERL injector use a FEL injector

Injector

independent concurrent

Operation mode

From the ERL side,

XFELO adds a new feature with minor modificationERL and XFELO provides complimentary X-rays

ERL Gun

FEL Gun

4R. Hajima

5

Growth of emittance and energy spread in a loop

pm4.13

25

5 ICr

qe

x

E=5 GeV, =25m, “half arc” = TBA x 15

)(103.2 155

mI

)(100.5 233

mI

5

2/1

35

108.13

2

ICr

pqep

incoherent SR

coherent SR

keV8CSRE

tuning of cell-to-cell phase advance negligible emittance growth

for 20pC/2ps

FEL gain is preserved

energy spread after FEL lasing

%05.0~1

uNE

E

Energy recovery is preserved

5R. Hajima

XFELO lasing at 5-GeV?

66

7 GeV, w=1.88cm, gap=5mm, aw=1 =0.1nm

5 GeV, w=1.43cm, gap=5mm, aw=0.59 =0.1nm

11][

16

1gain D-1

32223

A

pww I

IJJa

)1(2),()(

2

2

10w

w

a

aJJJJ

area mode current,peak kA,17 pA II

assuming same peak current and same mode area,

65.0)7(

)5(3

3

GeV

GeV

1-D gain for 5 GeV beam is “0.65 x 1-D gain for 7 GeV”

further gain reduction due to the emittance effect.

0.1nm XFELO with a 5-mm gap Halbach-type undulator

R. Hajima

7

XFELO with 5 and 7-GeV ERLs

0

0.05

0.1

0.15

0.2

0.25

0.3

0.1 0.15 0.2 0.25 0.3

normalized emittance (mm-mrad)

smal

l-sig

nal F

EL

gain

7 GeV

5 GeV

analytical estimationof small-signal FEL gain Energy 5 GeV 7 GeV

charge 20 pC

t 2 ps

E/E 1e-4

aw 0.59 1.0

u 1.43 cm 1.88 cm

Nu 3000

*=ZR 10 m

n 0.1 mm-mrad

gain 14 % 22 %

1Å X-FELO

The above calculations are based on a Halbach-type undulator.DELTA undulator gives 1.4 times larger FEL gain.

R. Hajima

8

Velocity bunching in an ERL main linac

Injector

Main Linac

SCA #2

SCA #1

SCA #3

Velocity bunching for a SASE-FEL injector

Velocity bunching for an ERL light source

Velocity bunching for an X-FELO

(1) no additional component is required (2) only 2-3% SCAs are used for the velocity bunching (3) residual energy spread is smaller than magnetic compression (4) moderate emittance growth for low bunch charge

H. Iijima, R. Hajima, NIM-A557 (2006).

L. Serafini and M. Ferrario, AIP-Porc. (2001)

R. Hajima, N. Nishimori, FEL-2009

R. Hajima

9

Gain reduction by bandwidth mismatch

growth rateof the m-th mode

gain loss cavity lengthdetuning

bandwidth mismatch

bandwidth of the Bragg mirrors = 12 meV

elMelM or

fs100M

sel f100

In the following calculations, we choose sel f400

12 meV

reflectivity and phase shiftfor a cavity round trip

K-J. Kim et al., PRL 100, 244802 (2008).

R. Hajima

1010

Example of the velocity bunching PARMELA simulation

bunch charge q = 7.7 pC

velocity bunchingbunching in 8 cavitiesinjection 10.9 MeV, 1.3 ps, -85 deg.gradient Eacc = 8.5 MV/m

emittance growth bychromatic aberration

x

y

z (m)norm

. e

mitt

ance

(m

m-m

rad)

y

xbe

am

siz

e (

mm

)

z (m)

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20 25 30 0

0.4

0.8

1.2

1.6

2

0 5 10 15 20 25 30

t E

Ene

rgy (MeV

)

bun

ch le

ngt

h (

ps)

z (m)

velocity bunching

0

1

2

3

4

5

6

7

8

9

10

0 5 10 15 20 25 30 0

5

10

15

20

25

30

injector merger SCA #1 SCA #2

4 cavities 4 cavities

R. Hajima

1111

Optimum design of the velocity bunching

0

0.2

0.4

0.6

0.8

1

0 5 10 15 20 25 30

x

y

z (m)norm

. e

mitt

ance

(m

m-m

rad)

0123456789

10

0 5 10 15 20 25 300

5

10

15

20

25

30

tE

Ene

rgy (MeV

)

bun

ch le

ngt

h (

ps)

z (m)

0

0.4

0.8

1.2

1.6

2

0 5 10 15 20 25 30

y

x

bea

m s

ize

(m

m)

z (m)

injector merger SCA #1 SCA #2

bunch charge q = 7.7 pC

velocity bunchingbunching in 6 cav. + on-crest 2 cav.injection 10.9 MeV, 1.3 ps, -90 deg.gradient Eacc = 8.5 MV/m

at the SCA#2 exitE = 27.7 MeV, t = 380 fs, E = 250 keVx = 0.16 mm-mrad, y = 0.13 mm-mrad

velocity bunching

R. Hajima

12

Phase plot at the SCA #2 exit

R. Hajima

13

Enhancement of the FEL gain by velocity bunching

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.1 0.15 0.2 0.25 0.3

normalized emittance (mm-mrad)

smal

l-sig

nal F

EL

gain

20 pC, 2 ps, E/E=10-4

7.7 pC, 0.38 ps, E/E=5x10-5

without velocity bunching

with velocity bunching

for 5-GeV, 1-Å X-FELO

Significant enhancement of the FEL gain by velocity bunching.Gain~40% is possible even with emittance growth during the bunching.

R. Hajima

14

Simulation of XFELO (5 GeV with velocity bunching)

After the saturation:

pulse duration=1.2 ps (FWHM)

photons/pulse (intra cavity) Np = 2x1010

photons/pulse (extracted) Np = 7x108

saturation

R. Hajima

15

2-Loop Design of 5-GeV ERL

HOM-damped cavity

high threshold current of HOM BBUallows 2-loop configuration.

TESLA (HOM:5×2)

ERL (HOM:6×2)

R. Hajima

16

Possible Scheme for Combining ERL and XFELO

We can switch two operation modes by introducing an orbit bump having rf/2= 11.5 cm.

5 GeV ERL for SR use : accelerate 2 times 7.5 GeV recirculating linac for XFELO : accelerate 3 times

Optional

S. Sakanaka, talk at PF-ISAC (2010)

R. Hajima

17

Growth of emittance and e-spread for 3-pass 7.5-GeV

n E

1st loop (2.5 GeV) 8pC/400fs 20pC/2ps

incoherent SR 0.029 mm-mrad 34 keV 34 keV

coherent SR assumed to be compensated 37 keV 11 keV

2nd loop (5 GeV)

incoherent SR 0.027 mm-mrad 130 keV 130 keV

coherent SR assumed to be compensated 53 keV 16 keV

1st-loop: E=2.5 GeV, =8.66m, 2x14-cell FODO

2nd-loop: E=5 GeV, =25m, TBAx30-cell

)(108.2 135

mI)(104.8 223

mI

)(106.4 155

mI)(100.1 223

mI

...22 cin

...2,

2, cEiEE

mradmm11.01.0 n

E/E = 2 x 10-5acceptable for FEL

17R. Hajima

18

Stability of SRF

Cornell LLRF SystemCornell LLRF System

A/A< 2·10-5 P < 0.01 deg

Demonstrated:

• Exceptional field stability at QL = 106 to 108

• Lorentz-force compensation and fast field ramp up

•Piezo microphonics compensation with ~20 Hz bandwidth

M. Liepe, ERL-09

energy stability << FEL gain band width

This requirement is fulfilled by current LLRF technology.

3000for107.12

1 4 u

u

NN

18R. Hajima

19

Conclusions

Hard X-ray ERL can accommodate XFELO. we can extend the frontier of X-ray beam parameters

0.1nm-XFELO is feasibly realized at 5-GeV ERL with velocity bunching 7.5-GeV beam from a 2-loop 5-GeV ERL an ERL injector is shared, no major modification is needed

XFELO can be installed either at a loop or a branch. however, beam loss in a long narrow duct might be a problem for

a XFELO in a loop

In the Japanese collaboration, XFELO is considered as a part of 5-GeV hard X-ray ERL.

19R. Hajima