The ALICE Electron Test Accelerator - Challenges, Achievements, and Future
Plans
Professor Jim ClarkeASTeC, 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
• 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
ERLP Funded in 2003
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
ALICE Milestones (Champagne Moments…)
Au
g 0
6:
Firs
t Ele
ctro
ns
Oct
08
: Fi
rst
Boost
er
Beam
Dec 0
8:
Full
Energ
y R
eco
very
Feb
09
: C
ohere
ntl
y E
nhance
d T
Hz
Nov 0
9:
CB
S X
-Rays
Feb
10
: IR
-FEL
Sponta
neous
Em
.
Mar
10
: EM
MA
Inje
ctio
n L
ine
Beam
Ap
r 1
0:
Firs
t TH
z C
ell
Exposu
res
Au
g 1
0:
EM
MA
Rin
g 1
00
0s
turn
s
Oct
10
: IR
-FEL
Firs
t La
sing
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)
PhotoinjectorGun 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.
0.00 0.02 0.04 0.06 0.08 0.100.0
0.2
0.4
0.6
0.8
1.0
1.2
Pho
tocu
rren
t (a.
u.)
Gas Exposure (L)
CH4
O2
CO2
CO
Standard Bake Optimised Bake
Photoinjector upgrade• Never need to let up gun
vacuum• Photocathode activated
offline• Reduced time for
photocathode changeover, from weeks to mins
• Higher quantum efficiency– Allows practical
experiments with photocathodes activated to different electron affinity levels
– 15% achieved in offline tests (red light)
• Allows tests of innovative photocathodes
• Installation?
Photocathode preparation facility
Loading chamber
Hydrogen rejuvenation chamber
Activation 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
Current Module
Linac Collaboration
New Module
Will be installed into ALICE in 2011
Linac Collaboration
7 cell cavity
Input coupler testing
HOM absorber
Outer cryomodule
assembly
DIAGNOSTICS
ROOMElectron beam
Laser beam
X-rays
Camera:Pixelfly QE
Camera:DicamPro
ScintillatorBe window
Dipole magnet
Quadrupole-04
Quadrupole-03
Correctors
Interaction region
To linac and beam dump
deflection and focussing mirrors
Vertical beam size: 39 µm RMS
Horizontal beam size: 27 µm RMS
~40pC/bunch, 29.6 MeV
800 nm pulses, ca. 70 fs duration, 500 mJ pulse power @ 10 Hz
(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
200mm
135mm
Size is not known because this would depend on the lens and the camera, but this should only be
small.
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
E Beam
OTR Camera
Compton ScatteringGeneration of short x-ray pulses by interacting a conventional laser with a low energy electron bunch
DIAGNOSTICS
ROOM
Background:Electron beam ONLaser OFF
Electron beam ONLaser beam ON
First data November 2009
Time delay
Evidence points to mis-alignment
Only 2 days of actual experimentation
Head on Collisions
Use of THz
• 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!
0
0.5
1
1.5
2
2.5
3
3.5
0 2 4 6 8 10
TH
z si
gnal
am
plit
ude,
V
Bunch charge, pC
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 ConfigurationsUnderstanding the NS-FFAG beam dynamics as function of lattice tuning & RF parameters
Graphs courtesy of Scott Berg BNL
Time of Flight vs Energy
• Example: retune lattice to vary longitudinal Time of Flight curve, range and minimum
• Example: retune lattice to vary resonances crossed during acceleration
Tune plane
EMMA
ALICE Provides the Beam
EMMA Parameters
Injection Line
Diagnostics Beamline
Frequency (nominal)
1.3 GHz
No of RF cavities
19
Repetition rate 1 - 20 Hz
Bunch charge 16-32 pC single bunch
Energy range 10 – 20 MeV
Lattice F/D Doublet
Circumference 16.57 m
No of cells 42
Normalised transverse acceptance
3π mm-rad
EMMA Ring Cell
• 42 identical doublets
• No Dipoles!
Long drift 210 mm
F Quad 58.8 mm
Short drift 50 mm
D Quad 75.7 mmFD
Cavity
210 mm
110 mm
Beam stay clear aperture
D
65 mm
55 mm
Magnet Centre-lines
Low Energy Beam
High Energy Beam
Field Clamps
Independent slides
Injection
SeptumKicker
Kicker
SeptumPower supply
Realisation of EMMA August 2010
First Data ...
First Turn Second Turn
September 2010 - beam circulates more than 1000 turns
Aug 2010 - First turns
Bruno MuratoriCERN 07/10/10
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
Optical Clock Distribution Scheme
Mode-LockedFibre Ring Laser
(81.25 MHz)
Link Operation 60 fs pulses are distributed to BAM
sites around ALICE.
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.
The other half of the timing stabilized pulses will be used to measure the arrival time of electron bunches and other diagnostics.
Feedback Loop
Circuitry
Fibre Stretcher
Fibre Stabilization Interferometer
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.
Beam Arrival Monitor
BeamlineRF
pickup
Single Mode Distribution Fibre (100m)
Faraday Rotating Mirror (50:50)
EOM Detector
Accelerator Area
-250 -150 -50 50 150 250
0
0.5
1
1.5
2
Beam Arrival Time Cali-bration
Delay (ps)
Norm
ali
zed
pow
er
Zero crossing for arrival time measurements
Trina Ng
ALICE Electro-optic experimentso 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 buncho confirming compression for FEL commissioningo examine compression and arrival timing along traino 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 ...
propagationdirection
EM Source development and testing
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
ALIG
NM
EN
T
MIR
RO
R
ALIG
NM
EN
T
MIR
RO
R
(OP
TIC
AL
TA
RG
ET)
(OP
TIC
AL
TA
RG
ET)
POWER METER
MCT DETECTOR
SPECTROMETER
ALIGNMENT WEDGES
INFR
A-R
ED
ELECTRONS
DWN-LAM-02 DWN-LAM-01 UPS-LAM-02UPS-LAM-01
HeNeHeNe
FEL-M1
FEL-M2
CCD VIEWER CAMERAS
FEL-WIG-TRANS-01
ALICE FEL Systems Schematic
OP
TIC
AL
TA
RG
ET
OP
TIC
AL
TA
RG
ETUNDULATOR
ARRAYS
DOWNSTREAM FEL
MIRROR
REFERENCE AXIS
LASER TRACKER
1. Undulator Arrays and Optical Targets surveyed onto Reference Axis with Laser Tracker
ALIGNMENT WEDGES
OP
TIC
AL
TA
RG
ET
OP
TIC
AL
TA
RG
ETUNDULATOR
ARRAYS
DOWNSTREAM FEL
MIRROR
2. Alignment Wedges and Downstream Mirror aligned optically using Theodolite
ALIG
NM
EN
T
MIR
RO
R
HeNeCCD VIEWER
CAMERAS
3. Downstream Mirror aligned using Upstream HeNe
CCD VIEWER CAMERAS
HeNe
ALIG
NM
EN
T
MIR
RO
R
4. Upstream Mirror aligned using Downstream HeNe
ELECTRONS
5. Electron Beam steered to Alignment Wedges
POWERMETER
MCT DETECTOR
SPECTROMETER
6. Cavity length scanned looking for enhancement of spontaneous emission, then LASING.
FEL Overview
UPSTREAM MIRROR
UNDULATOR
DOWNSTREAM MIRROR
ELECTRON BEAM AT FEL
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
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.
UPSTREAM MIRROR
DOWNSTREAM MIRROR
FEL Local Diagnostics
LASER POWER METER
FEL BEAMLINE TO DIAGNOSTICS ROOM
SPACE FOR DIRECT MCT DETECTOR
MCT (Mercury Cadmium Telluride) DETECTOR on
Exit Port 1
SPECTROMETER Based upon a Czerny
Turner monochromator
PYRO-DETECTORon Exit Port 2
DOWNSTREAM ALIGNMENT HeNe
Spontaneous Emission as a Diagnostic
February 2010: 1st Observation
Spontaneous emission a useful diagnostic
7 7.5 8 8.5 9-2
0
2
4
6
8
10
12x 105
Wavelength (m)
P(
) (a
.u.)
x = -1.0 mmx = 0.0x = +1.0 mm
10 12 14 16 181.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
Linac Off-Crest Phase (Degrees)
MC
T S
ign
al (
V)
40 50 60 70 801
1.5
2
2.5
3
3.5
4
Cavity Length Detuning (m)
MC
T S
ign
al (
V)
1. Spectrum used to optimise steering in undulator
2. Coherent enhancement used to set minimum bunch length
3. Interference of coherent SE used to set correct cavity length
Shortest wavelength + Narrowest Bandwidth when
beam on reference axis
Intensity enhancement at maximum bunch
compression
Intensity Oscillations at λ/2 in
cavity length indicating round trip
interference
First Lasing Data: 23/10/10 Simulation (FELO code)
-5 0 5 10 15 20 250
2
4
6
8
10
12
14
Cavity Length Detuning (m)
Out
coup
led
Ave
rage
Po
wer
(m
W)
-5 0 5 10 15 20 250
10
20
30
40
50
Cavity Length Detuning (m)
Out
coup
led
Ave
rage
Po
wer
(m
W)
ALICE IR-FEL: First Lasing
Results from First Lasing Period (23-31 October 2010)
20 25 30 354
6
8
10
12
14
Av
era
ge
Po
we
r (m
W)
Cavity L (m)
20 30 400.8
1
1.2
1.4
1.6
1.8
FW
HM
B/W
(%
)
Cavity L (m)
20 25 30 350.5
1
1.5
Pu
lse
En
erg
y (
J)
Cavity L (m)
20 30 40
0.8
1
1.2
1.4
1.6
T
(p
s)
Cavity L (m)
20 30 400
0.5
1
1.5
2
Pe
ak
Po
we
r (M
W)
Cavity L (m)
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)
20 40 60 80 100 1205
10
15
20
25
Q (pC)
Sin
gle
Pa
ss
Ga
in (
%)
20 40 60 80 100 12020
40
60
80
100
Q (pC)
Tsa
t(s
)
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!
NB: No optimisation done at higher charges (just turned up the
PI laser power (to 11))
3.54 3.56 3.58 3.6 3.62 3.64
x 10-4
10-2
10-1
T (s)
MC
T
Results from February 2011: Gap Tuning
5 5.5 6 6.5 7 7.5 8 8.50
0.2
0.4
0.6
0.8
1P
()(
a.u
.)
(m)
g = 16 mmg = 15 mmg = 14 mmg = 13 mmg = 12 mm
6 7 81
1.2
1.4
1.6
1.8
2
Ba
nd
wid
th (
%)
Wavelength (m)6 7 8
500
600
700
800
900
1000
FW
HM
t
(fs
)
Wavelength (m)6 7 8
0.5
1
1.5
2
2.5
Pu
lse
En
erg
y (
J)
Wavelength (m)6 7 8
1.5
2
2.5
3
3.5
PP
k (M
W)
Wavelength (m)
ALICE FEL Future Plans
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 exploitation programmes.
But ALICE itself will not run indefinitely.
We are now thinking beyond ALICE….
4 5 6 7 8 9 10 11 12 130
1
2
3
4
5
6
(m)
Pp
eak (
MW
)
27.5MeV, 0.75mm Hole radius22.5MeV, 0.75mm Hole radius27.5MeV, 1.5mm Hole radius22.5MeV, 1.5mm Hole radius27.5MeV, 2.25mm Hole radius22.5MeV, 2.25mm Hole radius
4 5 6 7 8 9 10 11 12 130
1
2
3
4
5
6
(m)
Pp
eak (
MW
)
27.5MeV, 0.75mm Hole radius22.5MeV, 0.75mm Hole radius27.5MeV, 1.5mm Hole radius22.5MeV, 1.5mm Hole radius27.5MeV, 2.25mm Hole radius22.5MeV, 2.25mm Hole radius
Simulation results
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
Top Related