Photonic Bandgap Fiber Wakefield
Experiment:
Focusing and Instrumentation for Dielectric Laser
Accelerators
R. Joel England
E. R. Colby, C. McGuinness, J. Ng, R. Noble, T. Plettner, C. M. S.
Sears, R. H. Siemann, J. E. Spencer, D. Walz
Advanced Accelerator Research Department
SLAC National Accelerator Laboratory
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
1
NLCTA E163 Test Beamline
E163: A facility for testing laser-driven accelerator structures.
Beam energy = 60MeV; st = 1ps to 400 attosec; sE = 0.1%
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
2
Dielectric Fiber Accelerator
conductor
hollow dielectric-lined waveguide
aperture ~ 0.26 l; Ez~ 2.5 GV/m
e
DF ; e / e  1  2
conductor lossy at
optical wavelengths
damage threshold for SiO2 ~ 5GV/m @ 1ps
Rosing & Gai, PRD 42, 1829 (1990)
e1,
e2
hollow Bragg waveguide
aperture ~ 0.3 l; Ez ~ 2.5 GV/m
DF ;
1  (2 a / l )2  2.1
Mizrahi & Schachter, PRE 70, 016505 (2004)
PBG fiber with central defect
aperture ~ 0.68 l; Ez ~ 2.5 GV/m
Ez  E pk / DF
s t  (1  vg / c)L fiber
X. E. Lin, PRSTAB 4, 051301 (2001)
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
3
Optimized PBG Fiber Geometry
a = 0.35 l ; R = 0.52 a
X. E. Lin “Photonic bandgap fiber accelerator,” PRSTAB 4, 051301 (2001)
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
4
The Road to a Fiber-based
Accelerator
• Manufacture/Prototyping
• Coupling
• e± beam: focusing, emittance, microbunching
• laser: mode-matching, coupling efficiency,
phase stability
• Fiber Characterization
• Proof-of-Principle Acceleration + Staging
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
5
Manufacturability
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
Courtesy Crystal-Fibre, Inc.
23-60 µm
Custom Fiber Manufacture
• prohibitively expensive for accel. prototyping
• SBIR or other funding for collaboration with
industry (e.g. Incom, Inc. - Charlton, MA)
Pre-made Telecom Commercial Fibers
• PBG telecom fibers exist (~$500/m)
• Thorlabs + Crystal-Fibre, Inc.
• Not designed for accelerator applications
Polymethylmethacrylate (PMMA) Fibers
• U. Sydney (A. Argyros, et al)
• drawing process less expensive
• technique could be used for geometrical
prototyping and tolerance testing
A. Argyros, et al., Optics Express 18, 5642 (2008)
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
6
Manufacturability
Air-core fibers from U. Southampton
matrix of 10µm holes, 1mm thick
Incom, Inc. SBIR proposal
hole dia = 10µm
courtesy J. E. Spencer, B. Noble
courtesy Dave Richardson
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
7
Commercial Fibers
fibers manufactured by Crystal-Fibre, Inc.
l
(telecom)
2R
(defect)
(µm)
a (pitch)
(µm)
lattice
dia. (µm)
cladding
dia. (µm)
1550
10.9
3.8
70
120
1060
9.7
2.75
50
123
633
5.1
1.77
33.5
101
830
9.2/9.5
2.3
40
135
BANDSOLVE
simulation of accelerating mode
for HC-1060 fiber
maximum gradient ~ 30 MV/m
courtesy B. Noble
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
8
Laser Coupling
TE mode
Bragg reflector
TM mode
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
9
Coupler Studies
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
10
Coupler Studies
PML
decompose excitation into normal modes
of the waveguide (including Lin mode):
perfH
ABC
1/12 section of fiber
HFSS Simulation
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
11
Laser Coupling from Free Space
Reflected Power
S11 (%)
Mode 1 = EH11, Mode 2 = TM01
45
40
35
30
25
20
15
10
5
0
Mode 1:1
Mode 1:2
Mode 2:2
Total
0
10
20
30
40
50
angle (deg)
max coupling ~ 35%
Coupling to fiber tip from
free space:
• shorter term solution
• HFSS model of simple
dielectric waveguide
• will extend to PBG lattice
type fiber
max coupling ~ 10%
options:
• radially polarized laser on
flat fiber tip
• linearly polarized laser on
angled fiber tip
Cleave angle = 0 deg
July 8, 2009
Cleave angle = 45 deg
Bob Siemann Symposium and ICFA
Workshop
12
E-Beam Focusing
New Halbach Magnet Design
Field Gradient ~ 500 T/m
Aperture = 6 mm
Adjustable z positions of magnets.
String encoder readback of magnet positions.
On slider stage for insertion/removal of assembly.
Magnets aligned on titanium rods.
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
13
Permanent Magnet Quadrupoles
PMQ 1: B’ = 483.0 T/m
PMQ 3: B’ = 462.4 T/m
PMQ 2: B’ = 440.5 T/m
PowerTrace Simulation
b*opt = Lfiber/2
b0 = 8m
a0 = 0
g2 = 11.85 mm
July 8, 2009
g1 = 23.46 mm
Bob Siemann Symposium and ICFA
Workshop
b* = 0.5 mm
14
Microbunch Washout
Initial Microbunched Beam
IFEL Interaction + Chicane Compression
Technique recently demonstrated by C.M.S.
Sears -> 400 attosec bunches
 f   0   sin(kL z0 ) Dominant washout terms
z f  z0  R56 [ 0   sin(kL z0 )]  T511 x0 2  T533 y0 2
llaser

I(z)  I 0 [1  2 bn cos(nkL z)]
After PMQ Focus
Primary culprits are the T511 and T533 of the PMQs
n1
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
15
Microbunch Washout
Possible Remedies
Radially Dependent Amplitude
z f  z0  R56 { 0  (x0 , y0 , z0 )sin(kL z0 )}  T511x0 2  T533 y0 2
T511 x0 2  T533 y0 2
(x0 , y0 , z0 )   
R56 sin(kL z0 )
this requires the IFEL modulation to increase quadratically with radial distance
Collimation
NO collimator
Qf = Q 0
July 8, 2009
400 µm diameter collimator
Qf = Q0/6
Bob Siemann Symposium and ICFA
Workshop
200 µm diameter collimator
Qf = Q0/21.5
16
Emittance Requirements
ELEGANT simulation of focal waist
Transmission vs. Normalized Emittance
b* = 0.5 mm
fiber aperture
b* = 0.5 mm
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
17
Emittance Preservation
Measured Emittance Growth in the NLCTA/E163 Beamline
D = random RMS steering error in quadrupoles
July 8, 2009
eN ~ 100 µm !
Bob Siemann Symposium and ICFA
Workshop
18
Improved Modeling Tools
Matlab-based
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
19
Improved Modeling Tools
ELEGANT-based
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
20
Improved Modeling Tools
DIMAD-based: Built into the Control System
Alternate definitions of bx and by
Matlab and ELEGANT models use
bx ~ 75 m
b x (s)  s x2 (s) / e x,rms (s)
2
If we instead use b x (s)  s x (s) / e x,initial
constant
then we (mostly) reproduce the SCP results:
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
21
Experimental Plan
800 nm
July 8, 2009
800 nm
Bob Siemann Symposium and ICFA
Workshop
22
Phase 1: Experiment Layout
Required Beam Parameters
Beam
9.6
µmCharge
50 pC
Normalized
Emittance
< 5 mm mrad
Energy
60 MeV
Bunch length
1 ps
Energy Spread
0.1 %
4 candidate commercial fibers
PMQs
Fiber Holder
fibers exit chamber for spectrographic analysis
FIBER HOLDER
4 candidate commercial fibers
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
beam passes through ~1mm of fiber
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
Newport MS 260i
Spectrograph
23
Summary
Issues to be addressed in developing PBG Fibers as Accelerators:
• Affordable (<$10k) manufacturing of Prototypes
• For injected test beam:
emittance
focusing and spot size
microbunch washout
• Laser coupling:
optimizing air-to-fiber coupling
developing high-efficiency advanced coupler designs
• Doing proof-of-principle experiments with single
and then multiple stages of acceleration.
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
24
Backup Slides
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
25
Coupler Studies
code: S3P
Advanced coupler design:
• in/out power couplers
• analogy to RF tw accelerator
• S11 = 0.1: power coupling
can be close to 100%
• how to manufacture?
courtesy of Cho Ng, SLAC
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
26
Motivation
S-Band RF
X-Band RF
Optical to IR
3D “woodpile”
structure
dielectric
gratings
July 8, 2009
PBG Fibers
P  RS
f
L
smaller RF structures:
• higher gradient
• machining tolerances
• transverse wakefields
• breakdown (Ez ≤ 100 MV/m)
Gradient 
laser-driven microstructures
• lasers offer high rep rates, strong
field gradients ( >0.5 GV/m),
commercial support
• dielectrics: high breakdown
threshold (~4 GV/m)
• manufacturability
• coupling of beam/laser
Bob Siemann Symposium and ICFA
Workshop
27
Photon Budget & SN Ratio
DEmode
e2 c b g Z c L
 kq 
4 1  b g l02
2
HC-1060 fiber: ZC  0.005 ;
DEmode
1 
DECherenkov
120
 0.12
l[nm]
k  1.42  1018 J / C 2 m  DEmode  3.6  10 23 J / electron
50 pC 1
)
 6162 photons
e h
 N transmission fiberspectrgraph  1150 photons
N  DEmode  (
N detector
50%
July 8, 2009
98%
ZC [] 
120
l[nm]
• Cherenkov background cannot couple to the
lattice modes at frequencies in the bandgap.
• Coupling to the defect modes would improve
the signal.
• Coupling to the cladding: can be reduced
through bend losses
38%
Bob Siemann Symposium and ICFA
Workshop
28
Search for Candidate Accel.
Modes
HC-1060 SEM image
RSoft BandSolve Model
courtesy of B. Noble
toward SOL line
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
29
Schottky vs Cherenkov
DEmode
e2 c b g Z c L
 kq 
4 1  b g l02
2
l0  Dl
DECherenkov
4 2 re Lmc 2
1
 
(1

)d l
3
fl
e
l0  Dl
f e 0 cb g / (1  b g )
DEmode
e
l0
[
]
Z
DECherenkov
e  1 4 (1  fLcladding / L fiber ) Dl c
vg  0.6;
Dl  0.48 nm;
L fiber  1 mm;
L cladding  164 µm;
e = 2.13 ; f  10;
DEmode
1 
DECherenkov
HC-1550 fiber: ZC  0.2 ;
Lin Fiber: ZC  19 ;
July 8, 2009
ZC [] 
212
l[nm]
212
 0.13
l[nm]
212
 0.2
l[nm]
Bob Siemann Symposium and ICFA
Workshop
30
Experimental Layout
NLCTA: design parameters
Beam Charge
50 pC
Normalized Emittance
1-2 mm mrad
9.6 Energy
µm
4 candidate commercial fibers
60 MeV
Bunch length
1 ps
Energy Spread
0.1 %
July 8, 2009
FIBER HOLDER
beam passes through ~1mm of fiber
fibers exit chamber for spectrographic analysis
Bob Siemann Symposium and ICFA
Workshop
31
Experimental Layout
e-beam
image of mounted fiber
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
32
Challenge: Small Spot Sizes
PMQ triplet with motorized gap
spacing and focal position.
420, 560, 560 T/m field strengths
modified Halbach design
C.M. Sears, “Production, characterization, and acceleration of optical
microbunches,” PhD dissertation, Stanford U. (2008)
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
33
Challenge: Small Spot Sizes
PowerTrace Simulation
QuickTime™ and a
TIFF (LZW) decompressor
are needed to see this picture.
ELEGANT simulation
of the final focus
transmission ~ 50%
(for 1060 fiber with
10 µm defect)
July 8, 2009
bx,by ~ 0.5 mm
QuickTime™ and a
TIFF (LZW) decompressor
are needed to see this picture.
Bob Siemann Symposium and ICFA
Workshop
sx,sy ~ 3 µm
34
Challenge: Small Spot Sizes
July 3, 2008 PMQ Scan
300
200
100
QuickTime™ and a
TIFF (LZW) decompressor
are needed to see this picture.
0
14
19
24
29
34
-100
-200
-300
QuickTime™ and a
TIFF (LZW) decompressor
are needed to see this picture.
QuickTime™ and a
TIFF (LZW) decompressor
are needed to see this picture.
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
35
Summary
optical to IR accelerating structures:
• offer high gradients (~ 1GeV/m),
• high rep rates, high damage threshold of dielectrics
• require micron-scale focusing, microbunching, and manufacturing
PBG fiber accelerators:
• permit large apertures
• commercial manufacturing capability;
• premade fibers are designed for telecom, not acceleration
• need to develop custom geometries: Lin fiber
E163
• near-term: focusing of e-beam through fiber cores + spectrally resolving
fiber modes from the emitted wakefield radiation
• long-term: coupling of structure to drive laser and observing net
acceleration of microbunched e-beam ---> multiple stages
July 8, 2009
Bob Siemann Symposium and ICFA
Workshop
36
Experimental Plan
Phase I: Wakefield Excitation (no laser)
Phase II: laser-driven scenario
July 8, 2009
• tightly focus beam through fiber
central defect (spot sizes < 10 µm)
• wakefield excitation of fiber
modes
• resolve accelerator-like
modes by spectral analysis
• few-100 attosec microbunched beam using
IFEL + chicane
• laser coupled to the fiber
accelerator mode
• measure net microbunch acceleration
with magnetic spectrometer
Bob Siemann Symposium and ICFA
Workshop
37