H. Maesaka*, K. Togawa, T. Inagaki, K. Onoe, T. Tanaka, A. Higashiya, H. Baba, H. Matsumoto,

measlyincompetentΠολεοδομικά Έργα

29 Νοε 2013 (πριν από 3 χρόνια και 8 μήνες)

77 εμφανίσεις

H. Maesaka*, K.
Togawa
, T. Inagaki, K.
Onoe
,

T. Tanaka, A.
Higashiya
, H. Baba, H. Matsumoto,

H. Tanaka, Y.
Otake

and T.
Shintake

FLS 2010, SLAC

1

Outline


Introduction


X
-
ray FEL

Project at SPring
-
8


SCSS Test Accelerator


Electron gun design


Performance


Emittance

measurement


Coherent OTR Search


Operational experience


Summary

2

3

X
-
ray FEL
Project at SPring
-
8


X
-
ray

wavelength:
< 0.1 nm


Self
-
amplified spontaneous emission (SASE) process


Beam energy:
8 GeV


Key technologies


Low
-
emittance

thermionic electron gun
:
0.6

p

mm
mrad


High
-
gradient C
-
band accelerator
: 35 MV/m


Short
-
period in
-
vacuum

undulator
:
l
u

= 18 mm, K < 2.2


First XFEL light

will be delivered in 2011.

4

SCSS Test Accelerator


To check the feasibility of XFEL


Extreme

ultraviolet (EUV) FEL facility


Wavelength:

50


60 nm
for saturated output


Beam energy:
250
MeV


Electron gun is identical to that of XFEL


Saturated EUV laser light
is stably generated.

Trend graph of FEL intensity

5

6

Thermionic
-
Gun
-
based Injector


Simple


Thermionic cathode with heater and high
-
voltage power source


No laser system


Stable


Stable emission rate


Long life time


But low current and long pulse length


Use fast chopper and velocity bunching


No grid electrode, because it causes emittance degradation.


7

Cathode

500kV

3
m
s

1A

1ns

Pulse deflector

(
Stripline

electrodes)

B

Collimator

(
f
5mm)

1ns, 1A

f
3mm

238 MHz

Pre
-
buncher cavity

Gun Design


Requirements from XFEL


Normalized emittance:
< 1
p

mm
mrad


Beam current:
> 1 A


Bunch length:
1 ns

(after the chopper)




Bunch charge:
1
nC


Voltage:
500 kV
(as high as possible)


To reduce space
-
charge effect


Cathode


CeB
6

single crystal


3mm
diameter


~

> 1400
deg.C

for 1A beam


K.
Togawa

et al., Phys. Rev. ST
Accel
. Beams 10, 020703 (2007)

Thermal Emittance

8

Cathode Assembly


CeB
6

cathode is mounted in a graphite sleeve


Heated to
~ 1400
deg.C

by graphite heater


500 kV
pulse voltage (3
m
s) is applied to
50 mm gap
(
10 MV/m
)

9

3mm

10MV/m

High Voltage Power Supply


50 kV Inverter power supply


Voltage stability:
~10
ppm

(
rms
)


Oil
-
filled compact pulse modulator


~ 3
m
s
pulse


Same as klystron power source


Gun high
-
voltage tank


Pulse transformer steps up the voltage to
500 kV
.


Dummy tube absorbs extra power


Stable impedance and easy cooling

10

50kV

Inverter
power
supply

AC 420V

Oil
-
filled

compact

pulse modulator

Beam

1:21

Gun high
-
voltage tank

500kV

Dummy tube

Cathode

Pulse forming
network

Thyratron

1A

~200A

Pulse
transformer

11

Normalized Emittance


0.6
p

mm
mrad

(
rms
)


Without 10% tail component


Energy: 500
keV


Current: 1 A


Measured by double
-
slit method

Charge density profile

Phase space profile

12

Estimation of Slice Emittance from
SASE
-
FEL Energy Curve


FEL energy curve was compared with simulation

13

Slice emittance of
SCSS test accelerator:

~ 0.7

p
mm
mrad

Gun emittance is
conserved after
x300 bunch
compression

14

Coherent OTR Search


15

1mm

Superposition of 5 shots

Very stable

Observed point

SCSS test accelerator

Bunch compression ratio was changed by the RF phase of S
-
band accelerator.

Courtesy of K.
Togawa

250MeV

OTR Image

OTR Intensity
v.s
. Bunch Compression


OTR intensity as a function of S
-
band RF phase.


Compression factor (calculation) is also plotted.

16

Lasing point

Total
compression factor ~ 300

SCSS test accelerator

250
MeV

~0.25
nC
/bunch

Intensity was normalized
by charge and
g
2

We didn’t observe non
-
linear amplification of OTR.

The gun emits a temporally and spatially smooth beam.

Courtesy of K.
Togawa

17

Operation at the SCSS test accelerator


Cathode heater is always on.


Except for shutdown periods


CeB
6

surface is cleaned by evaporation (10nm/hour)


Sometimes cathode temperature is adjusted to keep
the emission. (once per three month)


High voltage is turned on at 9:00 and off at 19:00
on weekdays


Fault rate


Less than once per day.


Mainly caused by
thyratron

misfire.


No spark around the cathode.


Down time due to the gun fault is only 0.4 %.

18

History of the SCSS test accelerator


2001
-
2003: Gun R&D


2004
-
2005: Construction of the SCSS test




accelerator


2005 Oct.: Started operation


2006 Jun.: First lasing


2006 Oct.: Dummy tube trouble


2007 Oct.: FEL saturation


2008 Jan.: Cathode replacement after 20,000




hour operation (heating time)


2010 Aug.: Cathode replacement?

19

The cathode
surface became concave of 0.2 mm deep from the initial flat surface.

It corresponds to evaporation speed of 10 nm/hour ( 10 nm/h x 20,000 h = 200 micron
-
meter)

Concave geometry made beam slightly focusing, but did not break emittance.

Electron microscope study showed (1) Surface is fairly smooth, (2) covered by carbon
contamination (lowered electron emission).

CeB6 cathode after 20,000 hour
heating.

20

Courtesy of
Shintake
-
san

Future Plan


New gun for XFEL


Identical to the SCSS test accelerator


Delivered soon


High
-
voltage test will be done in the coming April


Possible R&D areas


More stable power source


Thyratron



Solid
-
state switch


Dummy tube


more stable and longer life dummy load


Multi
-
bunch operation


Done by longer pulse generator for the chopper


Low charge operation with better emittance


Need thin collimator

or thin cathode


Higher current


We can do that with higher cathode temperature.


But cathode life time is shortened.


Higher voltage


Need improvements to prevent the cathode and the insulator from arcing


Higher repetition rate


Cathode itself is OK for kHz order


Depending on the high voltage power supply

21

Summary


Thermionic gun


CeB6 single crystal


3mm

diameter


~> 1400
deg.C


500 kV, 1 A


Low emittance and smooth beam


0.6
p

mm
mrad


No coherent OTR


No spiky modulation


Operation experience


Stably operated for four years.


Fault rate
: less than once per day


Down time
: only
0.4 %

(2008


2009)


Cathode lifetime:

~ 20,000 hours
(heating time)

22


23

24

Chopper


1ns bunch is extracted by pulse deflector from
3
m
s beam.


No grid electrode

25

Cathode

500kV

3
m
s

1A

1ns

Pulse deflector

(
Stripline

electrodes)

B

Collimator
(
f
5mm)

1ns, 1A

chopped

f
3mm

238 MHz

Pre
-
buncher cavity

Beam Profile and Emittance after the Chopper


Round, flat and smooth beam is generated.


Emittance was measured by slit scan method.


1.1
p

mm
mrad

(Without 10% tail)


This value can be larger than the true emittance

because of the space charge effect and the
poor resolution of the fluorescent screen

26

Fluorescent screen image

Emittance

Beam trajectory


27

Emission Curve


28

Longitudinal Bunch Profile Measured
by RF Zero Phasing Method


29

Peak Current vs. S
-
band TWA Phase

(Calculation)

(Data)

Courtesy of K.
Togawa

OTR Measurement Setup


30

Courtesy of K.
Togawa

Gold target (t~250nm)

Transmission of vacuum window

Kovar

glass

Quartz

Sapphire

Sensitivity of CCD camera

Sapphire window

Wavelen`gth

[
m
m]



We replaced CeB6 crystal in SCSS accelerator, after 20,000 hour
heating.

2008/01/28
First experience, but team did nice work.

Anode flange had color change.
31

Courtesy of
Shintake
-
san

Fault Statistics


Gun faults


Mainly comes from the trip of the high
-
voltage power
supply


Down time is only 0.4 %.


Trip rate: less than once per day

32

Operation time

Total down time

Gun down

time

#

of Gun faults

FY 2008

116444

min.

(~200 days)

3861 min.

3.3 %

640 min.

0.55 %

147

(~0.7 /day)

FY 2009

83552 min.

(139 days)

3193 min.

3.8 %

227 min.

0.27 %

43

(0.3 /day)

Total

199996 min.

7054

min.

3.5 %

867 min.

0.43 %

190

(~0.6 /day)

Dummy Tube Trouble


Occurred in Oct. 2006.


Cathode of the dummy tube is shorted to the
ground.


Inside of the vacuum vessel.


Insulation of the heater cable was insufficient.


Now, it’s improved.

33

Choice of Thermionic Electron Source


We chose the thermionic gun followed by chopper and velocity bunching
system.



Using the thermionic emitter, the system becomes
simple, eliminating use of
laser system

(for photo
-
cathode gun), which requires more frequent
maintenance than the thermionic cathode gun.



We believe “electron cloud emitted from the thermionic cathode has
smooth
temporal and
spatial
distribution
. There is no any internal structure”. It is the
nature of the thermionic electron emission (statistics).


In the chicane bunch compressor, CSR effect will amplify the density
modulation in the incoming electron bunch. Fluctuation of laser beam (fine
structure of temporal power profile) might be a source of

CSR instability
.
Quiet electron beam from the thermionic emitter is desirable to cure CSR
instability. (need to see X
-
ray lasing)


Thermionic gun system does not require “
the laser heater
” to smear the fine
structure in the bunch.


For future upgrade,
seeding scheme require smooth beam
.

34

Emittance Record in SCSS to XFEL/SPring
-
8


2001~2003
SCSS

R&D
CeB
6

thermionic gun



2004~2005
SCSS Test Accelerator

Construction



2006 June

First Lasing 49 nm at test accelerator.


2007 Oct.

Saturation at 50~ 60 nm




2006 April
XFEL/SPring
-
8 Construction

was funded.


Beam optics design. Technical design.

2007 Technical design, contract.


2008 Mass
-
production of hardware components.


2009 March.
Linac
, Undulator hall building completed.


Hardware installation.



2010 Oct. High power processing 8
GeV

accelerator
.


2011 April
~
Beam commissioning. First lasing at 1 A.

0.6
p
.mm.mrad@1ADC,500kV

0.7
p
.mm.mrad@300A,0.7pse挬

250䵥V,0.3nC

X300C潭pressin

X10Cmpr敳sin

0.8
p
.mm.mrad@3kA,8GeV

35


1
p


-
mrad敭ittan捥cb敡mfrmth敲mini挠捡thd攠was捨all敮e攬r捲azy?

this picture is copied from LCLS “
Conceptual Design Report
”,

SLAC
-
R
-
593 UC
-
414

Eliminating control
grid

from cathode.

Smaller size cathode,
from
8 mm to 3 mm

diameter.

Higher gun voltage, 150
kV to 500 kV.

Using single crystal CeB6
cathode.


36

Thermionic Gun looks better.

The non
-
linear space charge field

runs with bunch.


emittance increases

and energy spread increases.

RF
-
Photocathode gun.

Thermionic gun

Electrons run in static field.

Uniform beam


linear field



no emittance dilution



no energy spread

Time dependent rf
-
kick can be cured

by solenoid.

Fluctuations will be copied from the laser to the electron.

37

and, we need a quiet beam.


Intensity modulation on laser pulse
will be copied on electron density
coming out from RF photocathode




It might be a source (seeding) of
CSR instability.

38

Source of Modulation in Photocathode RF
-
Gun


Laser two color mixing

1
l
1
P
1 2 B
1 W, 10 mW 200 mW
P P P
   
Even a small power cause intense
modulation, due to heterodyne
amplification.

2 1
max min
max min 2 1
2/
1/
I I
I I
M
I I I I

 
 
This modulation is copied on electron beam and cause CSR instability.

2
l
2
P
multi
-
line oscillation

in high gain medium

THz laser source

uses this.

1 2
1 2
1 2
2 1
B
B
f f f
c c c
f
l l l
ll
l
l l
  
   


2 1 B
501 nm, 500 nm 250 m
l l l m
   
39

Modulation transfer (Smearing effect)

40

Modulation Transfer


RF
-
Gun carries modulation wavelength 1 um or longer to downstream, and
amplified
with CSR and convert into visible light in chicane through “wavelength compression”


T. Shintake, "Focal Point Laser Field

as Optical Seeder", Proc. FEL2006,


Berlin Germany

41

Use Small Size Cathode


First Strategy for smaller thermal emittance

Thermionic cathode

3
mm

diameter cathode (CeB6)

is used in a low emittance injector.

(SCSS SPring
-
8/RIKEN)

Operating Temperature 1500
°
C

3
223 meV
2
e B
w k T
 
Thermal Emittance

2
0
0.4
π
mm-mr
2
ad
c
B
xN
r
k T
m c
g

 
RF photo
-
cathode injector.

Assume ~ 2 mm laser spot size.

2
0
0.35
π
mm-mr
2
ad
c B e
xN
r k T
m c
g

 
T
e

is “measured” effective electron temperature of copper
cathode using 266 nm laser (ref. 2). k T
e

= 0.27 eV (2360
°
C).

Same order!

42

CeB6 Thermionic Electron Gun

K.
Togawa

43

1 A, 10 MV/m, 500 kV


1500
deg.C
. Highest temp. in system


Evaporating CeB6 at 1 nm/hour


Self Surface Cleaning


Surface Smoothing



Very stable e
-
emission


1 % decay per month


trim heater power each
3
-
months


Long life ~20,000 hour


four slots provide spring
action.

Graphite Heater 1700 deg. C

Target image of pyro
-
temp
measurement.

SUS316 (clean
-
Z)

Edge does not case high voltage
break down

Cathode Holder (graphite)

forms parallel field

CeB6 Single Crystal


3 mm diameter

44

Operational Experience of 500 kV Gun


HV processing is a few hours to
reach 500 kV


No HV breakdown at 500 kV for
4 years, daily operation.


SUS316 (clean
-
Z)

Electro
-
polished SUS

No baking



Gun sits inside HV

pulse tank, filled with oil.



Applying 500 kV pulse.



3 micro
-
sec pulse

driven by klystron

modulator.

45

Measured Emittance at the Gun

46

Q&A on the CeB6 electron gun


Why don’t we use LaB6?


CeB6 has twice longer lifetime than LaB6.


Can we use needle shape cathode like electron microscope for the FEL?


No. We need Ampere class beam. Available beam current density from CeB6
cathode is < 30 A/cm2. To obtain 1 A, we need at least 2 mm diameter. The current
is also space charge limited.


For longer lifetime, it is better to keep cathode temperature lower
,


and lower current density, thus we use 3 mm diameter cathode.


Can we apply DC voltage on the gun?


No. DC high voltage will cause a large dark current flow from cathode electrode and
HV breakdown.


At pulse width shorter than 10 micro
-
sec, HV breakdown limit becomes a few times
higher than DC limit. Thus we use 3 micro
-
sec, 500 kV pulse driven by pulse power
modulator.


Do we need to apply mechanical polish the cathode electrode?


No, polishing with diamond powder causes HV breakdown. Electro polishing is the
best choice for stainless steel.

47

CeB6 Crystal Cathode has reached its lifetime

after 20,000 hour operation.

January 2008, right after the new year holiday, the emission current
decreased quickly by 50%.

It has been nicely running for user experiment till end of December
2007, (60 nm full saturation)

This crystal was installed in October 2005.

Lifetime is
20,000 hours
, which is shorter than we expected. But, it is
fairly long enough, we decided to regularly change the crystal every
year at summer shutdown.

To replace cathode and tuning beam to recover full saturated lasing
was two
-
week job.

48