The ALICE Electron Test Accelerator Challenges, Achievements, and Future Plans
Professor Jim Clarke
ASTeC, STFC Daresbury Laboratory & Cockcroft Institute
JAI Lecture, 17th March 2011
Contents
•
•
•
•
•
•
•
Introduction to ALICE
Major Subsystems
Experimental Highlights
EMMA
Free Electron Laser
Future Plans
Summary
ALICE
• Accelerators and Lasers In Combined Experiments
• An R&D facility dedicated to accelerator science
and technology
– Offers a unique combination of accelerator, laser and free-electron
laser sources
– Enabling studies of electron and photon beam combination
techniques
– Provides a range of photon sources for development of scientific
programmes and techniques
Reminder: 4GLS
ERLP Funded in 2003
• Energy Recovery Linac Prototype
• To develop skills and technologies for 4GLS:
– Operation of photo injector electron gun
– Operation of superconducting electron
linac
– Energy recovery from a FEL-disrupted
beam
– Synchronisation of gun and FEL output
ALICE
Parameter
Value
Gun Energy
350 keV
Injector Energy
8.35 MeV
Max. Energy
35 MeV
Linac RF Frequency
1.3 GHz
Max Bunch Charge
80 pC
Oct 10: IR-FEL First Lasing
Aug 10: EMMA Ring 1000s turns
Apr 10: First THz Cell Exposures
Mar 10: EMMA Injection Line Beam
Feb 10: IR-FEL Spontaneous Em.
Nov 09: CBS X-Rays
Feb 09: Coherently Enhanced THz
Dec 08: Full Energy Recovery
Oct 08: First Booster Beam
Aug 06: First Electrons
ALICE Milestones (Champagne Moments…)
ALICE parameters
Parameter
Design Value
Operating Value
Injector Energy
8.35 MeV
6.5 MeV
Total beam energy
35 MeV
27.5 MeV
RF frequency
1.3 GHz
1.3 GHZ
Bunch repetition frequency
81.25 MHz
81.25 MHz or 16.25 MHz
Train Length
0 - 100 ms
0 - 100 ms
Train repetition frequency
1 - 20 Hz
1 - 20 Hz
Compressed bunch length
<1 ps rms
<1 ps rms (measured)
Bunch charge (81.25 MHz)
80 pC
40 pC
Bunch charge (16.25 MHz)
80 pC
80 pC
Energy Recovery Rate
>99%
>99% (measured)
Photoinjector
Gun ceramic was major source of delay
(~1 year)
Alternative ceramic on loan from
Stanford was installed to get us started –
still in use today!
Limits gun voltage to 230 kV (cf 350 kV)
Original ceramic is on shelf waiting for
opportunity to be installed
First electrons August 2006
Photoinjector Vacuum
• XHV needed for good lifetime of cathode (GaAs)
– UHV is not good enough!
• A new in-situ bakeout procedure was developed which monitored the
ratio of gas species in the vacuum system during the bake.
• Evidence suggests that partial pressures of any oxygen containing
species (CO, CO2 and H2O) should be < 10-14 mbar.
Photocurrent (a.u.)
1.2
1.0
0.8
0.6
CH4
O2
CO2
CO
0.4
0.2
0.0
0.00
Standard Bake
Optimised Bake
0.02
0.04
0.06
Gas Exposure (L)
0.08
0.10
Photoinjector upgrade
• Never need to let up gun
vacuum
• Photocathode activated offline
• Reduced time for
photocathode changeover,
from weeks to mins
• Higher quantum efficiency
Photocathode
preparation facility
– Allows practical experiments
with photocathodes activated
to different electron affinity
levels
– 15% achieved in offline tests
(red light)
• Allows tests of innovative
photocathodes
• Installation?
Activation chamber
Loading chamber
Hydrogen rejuvenation chamber
Superconducting Linacs
• Both linacs were procured from
ACCEL (now Research
Instruments)
• They each contain two 9-cell
ILC type cavities (as used by
XFEL) – 1.3 GHz
• Linacs only designed to
operate in pulsed mode (20Hz)
• Would not be suitable for 4GLS
or NLS type, high-rep rate,
facilities
Linac Collaboration
• International initiative led by ASTeC to develop linac module
suitable for CW operation as required by a high rep rate
facility (eg NLS)
–
–
–
–
–
–
–
–
Higher power and adjustable input couplers
Higher beam currents, CW operation
Piezo actuators provide improved stability control
Improved thermal and magnetic shielding
Better HOM handling
7 cell cavities so space for HOM absorbers
Same footprint as ACCEL linac so can install in ALICE easily
Validation with beam
Linac Collaboration
Current Module
Will be installed into
ALICE in 2011
New Module
Linac Collaboration
7 cell cavity
Input coupler
testing
Outer
cryomodule
assembly
HOM absorber
Compton Scattering
800 nm pulses, ca. 70 fs duration,
500 mJ pulse power @ 10 Hz
X-rays
Camera:
DicamPro
Scintillator
Be window
To linac
and beam
dump
Laser beam
Horizontal beam size: 27 µm RMS
Generation of short x-ray
pulses by interacting a
conventional laser with a low
energy electron bunch
deflection and
focussing
mirrors
OTR Camera
Size of foil in the straight on is 47.5mm. When
turned through 45° the vertical height of the foil
is 47.5 but the horizontal is only 39.77mm
because of the clamping ring
small.
Dipole magnet
Vertical beam size: 39 µm RMS
Camera:
Pixelfly QE
135mm
200mm
Interaction region
Quadrupole-04
Size is not known because this
would depend on the lens and the
camera, but this should only be
E Beam
(50. 8mm mirror) when seen in
the holder in the straight on
position you can only see
46.8mmØ. When rotated
through 45° the vertical is
46.8mm and the horizontal is
41.14mm because of the mirror
holder
Correctors
DIAGNOST
ICS ROOM
Quadrupole-03
Electron beam
~40pC/bunch, 29.6 MeV
Head on Collisions
First data November 2009
Background:
Electron beam ON
Laser OFF
Time delay
Electron beam ON
Laser beam ON
DIAGNOST
ICS ROOM
Evidence points to mis-alignment
Only 2 days of actual experimentation
Use of THz
3.5
3
THz signal amplitude, V
• CSR generated in THz
region because bunch
length ~1 ps
• Output enhanced by many
orders of magnitude (N2)
• Dedicated tissue culture lab
• Effect of THz on living cells
being studied
• Source has very high peak
intensities but very low
power – so no thermal
effects!
2.5
2
1.5
1
0.5
0
0
2
4
6
Bunch charge, pC
8
10
EMMA
• Fixed Field Alternating Gradient accelerators are an old
idea (invented in 1950s)
• They use DC magnets with carefully shaped pole
profiles
• The beam orbit scales with energy so the magnet
apertures are large
EMMA
• Non-Scaling Fixed Field Alternating Gradient accelerators
are a new idea (invented in 1990s)
• They use simple DC magnets (eg quadrupoles)
• The beam orbit changes shape with energy enabling the
magnet apertures to be small
• EMMA is the first of this type – a proof of principle
Non-scaling FFAG
• Born from considerations of very fast muon acceleration
– Breaks the scaling requirement
– More compact orbits ~ X 10 reduction in magnet aperture
– Betatron tunes vary with acceleration (resonance crossing)
– Parabolic variation of time of flight with energy
• Factor of 2 acceleration with constant RF frequency
• Serpentine acceleration
• Can mitigate the effects of resonance crossing by:– Fast Acceleration ~15 turns
– Linear magnets (avoids driving strong high order resonances)
• Or nonlinear magnets (avoids crossing resonances)
– Highly periodic, symmetrical machine (many identical cells)
• Tight tolerances on magnet errors dG/G <2x10-4
Novel, unproven concepts which need testing
Electron Model => EMMA!
EMMA Goals
Graphs courtesy of Scott Berg BNL
Lattice Configurations
Understanding the NS-FFAG beam
dynamics as function of lattice tuning & RF
parameters
Tune plane
• Example: retune lattice to vary
resonances crossed during
acceleration
Time of Flight vs Energy
• Example: retune lattice to vary
longitudinal Time of Flight
curve, range and minimum
Graphs courtesy of Scott Berg BNL
ALICE Provides the Beam
EMMA
EMMA Parameters
Energy range
10 – 20 MeV
Lattice
F/D Doublet
Frequency
(nominal)
Circumference
16.57 m
No of RF cavities 19
No of cells
42
Repetition rate
1 - 20 Hz
Bunch charge
16-32 pC
single bunch
Normalised
transverse
acceptance
3π mm-rad
1.3 GHz
Diagnostics Beamline
Injection Line
EMMA Ring Cell
65 mm
Field55Clamps
mm
D
D
F
Low Energy
Beam
Cavity
Beam stay clear aperture
Long drift
210 mm
F Quad
58.8 mm
Short drift
50 mm
D Quad
75.7 mm
42 identical doublets
•No Dipoles!
•
High Energy
Beam
110 mm
210 mm Centre-lines
Magnet
Independent slides
Injection
Kicker
Septum
Power supply
Septum
Kicker
Realisation of EMMA August 2010
First Data ...
First Turn
Aug 2010 - First turns
Second Turn
September 2010 - beam
circulates more than 1000 turns
Extraction (07/03/11)
• Going clockwise towards
extraction
– Yellow = Inj. Kicker1
– Pink = Ext. Kicker1
– Green = Ext. Kicker2
– Blue = beam
• Action of injection kicker
too early to be seen
• Spikes = turns
• Effect of extraction clearly
visible
• Image seen on first YAG
screen in extraction /
diagnostic line 
CERN 07/10/10
Bruno Muratori
Trina Ng
Optical Clock Distribution Scheme
Highly stable clock distribution across large scale facilities is important for the synchronisation of beam
generation, beam manipulation components and end station experiments. Optical fibre technology can
be used to combat the stability challenges in distributing clock signals over long distances with coaxial
cable.
An actively stabilised optical clock distribution system based on the propagation of ultra-short optical
pulses has been installed on ALICE. Femtosecond pulses emerging at the far end are currently used to
implement a beam arrival monitor. However, the clock signals could also be integrated into other
diagnostic systems such as electro-optical beam diagnostics.
Link Operation
 60 fs pulses are distributed to BAM
sites around ALICE.
Mode-Locked
Fibre Ring Laser
(81.25 MHz)
 The other half of the timing
stabilized pulses will be used to
measure the arrival time of electron
bunches and other diagnostics.
Normalized power
 Half the pulse power will be
reflected back at the far end to enable
detection of optical path length
changes.
 Timing is actively stabilized with a
fibre stretcher and delay line.
Fibre Stabilization
Interferometer
Feedback
Loop
Circuitry
Fibre
Stretcher
Beam Arrival Time Calibration
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
-250
Single Mode
Distribution
Fibre (100m)
Accelerator Area
Beam Arrival Monitor
EOM
Zero crossing for
arrival time
measurements
Beamline
-150
-50
50
Delay (ps)
150
250
Detector
Faraday Rotating
Mirror (50:50)
RF
pickup
ALICE Electro-optic experiments
o Energy recovery test-accelerator
intratrain diagnostics must be non-invasive
o low charge, high repition rate operation
typically 40pC, 81MHz trains for 100us
Spectral decoding results for 40pC bunch
o confirming compression for FEL commissioning
o examine compression and arrival timing along train
o demonstrated significant reduction in charge requirements
S.P. Jamison
Laser-electron Beam Interactions
• New concepts & proof-of-principle tests
• Developing technique for direct phase-space
manipulation of electrons with longitudinally laser &
unipolar THz pulses.
• Aim to adjust phase-space without need for
modulators/chicanes
ALICE experiment in final stages
of preparation ...
EM Source development and testing
propagation
direction
Oscillator FEL Process
ALICE IR-FEL




Dec 2009/Jan 2010: FEL Undulator and Cavity Mirrors installed and aligned.
Throughout 2010: FEL/THz/CBS programmes proceeded in parallel with
installation of EMMA. One shift per day of beamtime for commissioning.
Of available beamtime, FEL programme gets ~15%.
Progress:

Feb 2010: First observations of undulator spontaneous emission. Stored in
cavity immediately.

But no lasing could be found. Problem was that we were limited to 40pC:
above 40pC @ 81.25Mz beam loading prevented constant energy along
100µs train.

On 17th October 2010 we installed a Burst Generator to reduce laser
repetition rate from 81.25MHz to 16.25 MHz and increased bunch charge to
60pC.

A week later, on 23rd Oct 2010 achieved first lasing @ 8µm

Shutdown Nov/Dec 2010

Jan/Feb 2011: Lasing from 8.0-5.7µm

Mar 2011: IR transported out of ALICE area to beyond shield wall
FEL SYSTEMS + Transverse/Longitudinal Alignment
ALIGNMENT
MIRROR
ALIGNMENT
MIRROR
TARGET
OPTICAL
(OPTICAL
TARGET)
LASER
TRACKER
ALIGNMENT
WEDGES
INFRA-RED
HeNe
FEL-M2
FEL-WIG-TRANS-01
DWN-LAM-01
MCT
DETECTOR
SPECTROMETER
MCT
DETECTOR
UNDULATOR
ARRAYS
CCD VIEWER
CAMERAS
UPS-LAM-01
UPS-LAM-02
HeNe
1.
2.
4. Alignment
Undulator
Upstream
Arrays
Mirror
Wedges
aligned
and
and
Optical
Downstream
using
Targets
Downstream
Mirror
surveyed
aligned
HeNe
onto
optically
Reference
using
Axis
Theodolite
with Laser
6.
Cavity
length
scanned
looking
for
enhancement
ofSystems
spontaneous
emission,
thenTracker
LASING.
SPECTROMETER
5.
Electron
Beam
steered
to
Alignment
Wedges
3.
Downstream
Mirror
aligned
using
Upstream
ALICE
FEL
Schematic
(OPTICAL
TARGET)
FEL-M1
POWER
METER
METER DOWNSTREAM
FELDWN-LAM-02
MIRROR
OPTICAL TARGET
REFERENCE AXIS
FEL Overview
DOWNSTREAM MIRROR
ELECTRON BEAM AT FEL
UNDULATOR
UPSTREAM MIRROR
Energy
27.5MeV
Bunch Charge
80pC
Bunch Length
~1ps
Normalised
Emittance
~12 mm-mrad
Energy Spread
~0.6% rms
Repetition Rate
16.25MHz
Macropulse
Duration
100µs
Macropulse Rep.
Rate
10Hz
BUNCH COMPRESSOR
FEL Undulator
UNDULATOR
On loan from JLAB where previously
used on IR-DEMO FEL
Now converted to variable gap
PARAMETERS
Type
Hybrid planar
Period
27mm
No of Periods
40
Minimum gap
12mm
Maximum K (rms)
1.0
FEL Resonator
RESONATOR
Mirror cavities on kind loan from CLIO.
Previously used on Super-ACO FEL
PARAMETERS
UPSTREAM MIRROR
DOWNSTREAM MIRROR
Type
Near Concentric
Resonator Length
9.2234m
Mirror ROC
4.85m
Mirror Diameter
38mm
Mirror Type
Cu/Au
Outcoupling
Hole
Rayleigh Length
1.05m
Upstream Mirror Motion
Pitch, Yaw
Downstream Mirror Motion
Pitch, Yaw, Trans.
FEL Local Diagnostics
LASER POWER
METER
DOWNSTREAM
ALIGNMENT HeNe
FEL BEAMLINE TO
DIAGNOSTICS ROOM
PYRO-DETECTOR
on Exit Port 2
SPACE FOR DIRECT
MCT DETECTOR
MCT (Mercury Cadmium
Telluride) DETECTOR on
Exit Port 1
SPECTROMETER
Based upon a Czerny
Turner monochromator
Spontaneous Emission as a Diagnostic
February 2010:
1st
1. Spectrum used to optimise steering in
undulator
Observation
5
12
x 10
x = -1.0 mm
x = 0.0
x = +1.0 mm
10
P( ) (a.u.)
8
Spontaneous emission a useful
diagnostic
6
4
2
0
-2
2. Coherent enhancement used to set
minimum bunch length
1.6
7.5
8
Wavelength  (mm)
8.5
9
4
Intensity
enhancement at
maximum bunch
compression
3.5
MCT Signal (V)
MCT Signal (V)
1.7
7
3. Interference of coherent SE used to set
correct cavity length
1.9
1.8
Shortest wavelength +
Narrowest Bandwidth when
beam on reference axis
1.5
1.4
3
Intensity
Oscillations at λ/2 in
cavity length
indicating round trip
interference
2.5
2
1.3
1.5
1.2
1.1
10
12
14
16
Linac Off-Crest Phase (Degrees)
18
1
40
50
60
70
Cavity Length Detuning ( mm)
80
ALICE IR-FEL: First Lasing
Simulation (FELO code)
14
Outcoupled Average Power (mW)
Outcoupled Average Power (mW)
First Lasing Data: 23/10/10
12
10
8
6
4
2
0
-5
0
5
10
15
20
Cavity Length Detuning (mm)
25
50
40
30
20
10
0
-5
0
5
10
15
20
Cavity Length Detuning (mm)
25
14
Pulse Energy (mJ)
1.5
10
8
6
FWHM B/W (%)
4
20
25
30
Cavity L (mm)
35
1
2
0.5
20
1.8
1.6
1.6
1.4
1.4
1.2
1
0.8
20
25
30
Cavity L (mm)
35
40
1.5
1
0.5
0
20
1.2
30
Cavity L (mm)
40
1
0.8
30
Cavity L (mm)
Peak Power (MW)
12
 T (ps)
Average Power (mW)
Results from First Lasing Period (23-31 October 2010)
20
30
Cavity L (mm)
40
Implies electron bunch length
≈1ps, in agreement with
previous EO measurements
of a similar ALICE setup
Results from First Lasing Period (23-31 October 2010)
MCT
NB: No optimisation done at
higher charges (just turned up the
PI laser power (to 11))
10
-1
Single Pass Gain (%)
25
10
20
3.54
15

10
5
20

40
60
80
100
120
Q (pC)

100
80
T sat (ms)
-2
60
40
20
20
40
60
80
Q (pC)
100
120
3.56
3.58
T (s)
3.6
3.62
3.64
x 10
-4
Gain determined from cavity rise time
From one pulse train to the next (@10Hz) the gain
jitters

Cause under investigation. Phase jitter in
pulsed RF? Laser jitter?....
On average the gain is lower than we want:

rms Energy spread of 0.6% is too big:
degrades the gain significantly

Aim to halve energy spread and double gain

Can then change to outcoupler with larger
hole

Can set up beam to achieve this (set injector
to deliver shorter bunch to linac) but haven’t
yet lased with this setup – still to be
understood! Should work, but doesn’t!
Results from February 2011: Gap Tuning
g
g
g
g
g
1
P( )(a.u.)
0.8
0.6
= 16 mm
= 15 mm
= 14 mm
= 13 mm
= 12 mm
0.4
0.2
0
5
5.5
6
6.5
7
7.5
8
8.5
1.8
900
1.6
1.4
1.2
1
800
700
600
6
7
8
Wavelength ( mm)
500
6
7
8
Wavelength ( mm)
2.5
3.5
2
3
PPk (MW)
1000
Pulse Energy (m J)
2
FWHM  t (fs)
Bandwidth (%)
 (mm)
1.5
1
0.5
2.5
2
6
7
8
Wavelength ( mm)
1.5
6
7
8
Wavelength ( mm)
ALICE FEL Future Plans




3
Simulation results
6
5
Ppeak (MW)

(MW)

Improved electron beam set-ups
with reduced energy spread and
jitter.
Transport of FEL beam to
diagnostics room, then full output
characterisation.
Slightly reduced Mirror ROC to
improve gain, plus selection of
outcoupling hole sizes to optimise
output power.
Plan to run at 27.5MeV (5-8µm) and
22.5MeV (7-12µm)
Beyond that depends on funding
being obtained for specific
6
exploitation programmes.
But ALICE itself5 will not run
indefinitely.
4
We are now thinking
beyond
ALICE….
peak

4
3
2
1
0
4
5
6
7
8
9
10
11
 (mm)
27.5MeV,
22.5MeV,
27.5MeV,
22.5MeV,
27.5MeV,
22.5MeV,
0.75mm Hole radius
0.75mm Hole radius
1.5mm Hole radius
1.5mm Hole radius
2.25mm Hole radius
2.25mm Hole radius
12
13
The Future ...



Concepts for post-ALICE future hundred-MeV-scale electron test
accelerators are currently under development in consultation with
other stakeholders (including JAI!).
Potential topics of interest:
 Ultra-Cold injectors (low emittance, low charge, velocity
bunching, fs bunches…..)
 Novel acceleration (laser plasma….)
 Compact FELs (short period undulators….)
 Attosecond FEL pulse generation (slicing, modelocking…)
 Novel FEL seeding schemes (HHG, self-seeding, EEHG….)
 FEL pulse diagnostics
Will be a national and international collaboration taking ~12
months to develop the plans in more detail.
Summary
• ALICE is an extremely versatile and flexible test accelerator
• We have gained practical experience/skills of several key accelerator
technologies
– Photoinjectors
– SRF & 2K cryo
– High power laser/electron interactions
– FELs
– Timing & Synchronisation
– Energy Recovery
– Coherent SR
– .....
• EMMA is currently being commissioned (using ALICE as the injector)
• Plans are being drawn up for future test facilities – please join in the
discussion!
Acknowledgements
• Thanks to the following for providing slides and
other material
–
–
–
–
–
–
–
–
–
–
Neil Thompson
Bruno Muratori
Elaine Seddon
Neil Bliss
Rob Edgecock
Steve Jamison
Peter McIntosh
Susan Smith
Keith Middleman
Trina Ng