Accelerator Distribution according to Scope New applications New technologies Basic Science >15000 1% EPS HEP2003 Frontier Accelerators -> Technologies & Methods -> Applications (Interconnection Scheme) TW Laser s Very Hi gh Acc.

Download Report

Transcript Accelerator Distribution according to Scope New applications New technologies Basic Science >15000 1% EPS HEP2003 Frontier Accelerators -> Technologies & Methods -> Applications (Interconnection Scheme) TW Laser s Very Hi gh Acc.

Accelerator
Distribution
according to
Scope
New applications
New technologies
Basic
Science
>15000
1%
EPS HEP2003
1
Frontier Accelerators -> Technologies & Methods -> Applications
(Interconnection Scheme)
TW
Laser s
Very
Hi gh
Acc. Fields
Fr on t ier
Col l i der s
Sy n c h r o t r o n
Ra d ia t io n
So u r c e s
Ap p l ie d
Su p e r
Co n d u c t iv i t y
Ve r y Hi g h
Fr e q u e n c y
Po w er
High Bea m
Quality , Fine
Bea m Contr ol
Ne u t r o n
Sp a l l a t i o n
So u r c e s
High Bea m
P owe r
EPS HEP2003
New Ac ce l. Te chniques,
Table -top Acc el.
VH P owe r Spallation,
Ra dio Ac tive Isotope s
Wa ste Tra nsm uta tion,
Ener gy P roduction
Fr e e El e c t r o n La s e r s
Medic al a nd
Industria l
2
High Power
Hadron Beams
Spallation neutrons
Radioactive Ions
Waste Treatment
Energy Production
Inertial Fusion
Materials Irradiation
Present Generation of HP p Accelerators
EPS HEP2003
4
Technology advances for High Power
Hadron Beams
 Sources: high current, high brightness
 RFQs : low b, high efficiency, brightness,
SC devices being developed
 SC RF : low b , optimization of warm-cold mix,
reliability of large high field systems tested in
HE install.
 RF Power Generators : new HF components
(Klystrons..) developed and tested for Colliders
EPS HEP2003
5
Progress
in SC RF
systems
Large SC
systems in
operation
0,35 GeV Linac
1 GeV re-circulated Linac
EPS HEP2003
6
Typical SC Linac Schematic Layout
n Factory CERN SC Linac design
LEP type SC Cavities
EPS HEP2003
7
ADTF
(Accelerator Driven Test Facility)
SC Linac Design
ANL
2-gap
Injector&
spoke
350 MHz
LEDA RFQ b=0.175
6.7 MeV
Courtesy C. Pagani
EPS HEP2003
350 MHz
3-gap
spoke
b=0.20
14 MeV
700 MHz
3-gap
spoke
b=0.34
44 MeV
5-cell elliptical
b=0.48
109 MeV
5-cell elliptical
b=0.64
211 MeV
600 MeV
13.3-mA
ANL
8
needs
high peak current,
high brightness p, Hor heavy ion beams
Neutron Spall sources - title
Neutron production energy window
EPS HEP2003
9
Science with neutrons
EPS HEP2003
ESS Scientific Case
1
10
Science with neutrons
EPS HEP2003
ESS Scientific Case
2
11
Major spallation sources
European Spallation Source (ESS)
2 x 5 MW
QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture.
Warm design
Superconducting
in evaluation
Ongoing study
1.4 MW
Superconducting
Fully operational
in 2007
EPS HEP2003
SNS (Oak Ridge)
12
Trans
Uranic
Elements
Accelerator
Transmutation of Waste
Fission
Fragment
s
EPS HEP2003
13
ccelerator
ransmutation of
aste
• A high power proton accelerator
produces spallation neutrons in a
heavy metal target.
• The target is at center of “blanket”
region (“transmuter”) filled with
chemically separated long-lived
transuranics and fission products
• Subcritical :
fissionable transuranics arranged
such that chain reactions cannot be
sustained without an external
n source
EPS HEP2003
14
Accelerator Transmutation of waste HOW
Pu &
long lived
Minor Act’s
& long lived
ransmutation of
0,15 %
Stable &
short lived
SPENT FUEL
aste
0,95 %
U
95,6 %
3,3 %
FISSION PRODUCTS
Transmuter
EPS HEP2003
15
ccelerator
TransmutTransmutation
initial demo of
confaste
j
Reliability !
10 to 40 MW proton beam
1 GeV , 10 to 40 mA, CW
EPS HEP2003
16
The 30 MW TRASCO linac
80 keV
5 MeV
100 MeV
Proton
Source
RFQ
Intermediate energy linac
Source
RFQ
ISCL
Microwave RF Source
High current (35 mA)
80 keV
90% transmission
30 mA, 5 MeV
5-100 MeV
Intermediate Energy
Super Conducting Linac
With reentrant (or QWR)
cavities
Single type cavity string
8bl focussing
EPS HEP2003
~200 MeV
3 sect.
~500 MeV
high energy
>1000 MeV
SC linac
High Energy SC Linac
3 section linac:
– 100-190 MeV, b=0.5
– 190-450 MeV, b=0.65
– 450-1600 MeV, b=0.85
Five(six) cell elliptical cavities
Quadrupole doublet focusing: multi–
cavity cryostats between doublets
– 352.2 MHz (CERN/LEP)
– 704.4 MHz
17
Cryomodule design
Cryomodule engineering studies
– Based on TTF/CEBAF/CERN experience
– Sliding cavity fixtures
– Single thermal shield with integrated
cooling pipes: “finger welds”
– Integration of cavity ancillaries
(e.g. Tuner)
CEBAF derived
TESLA derived
18
A Multi Purpose Facility
The CONCERT Complex
MW
m, n Factory
4
LP Spall. S.
5
Waste Transm. 5
1
R.Ion Facility
Irradiation
10
25
EPS HEP2003
19
High intensity Ion Driven Inertial Fusion
“The ultimate challenge of accelerator physics”
“The ultimate accelerator physics challenge”
Pav ~ 200 MW
SC Dipoles
Induction Linac Bunchers
1 per species, 24 beams each
mm
“the ultimate challenge for1-1,6
accelerator
physics “ ?
(to scale)
SC Dipoles
4 MJ/ pulse
spot size
Indirectly driven
target
10 ns
50 Hz
3x2
rings
2x12x3
synchronization
stages
EPS HEP2003
3x16
ion
sources
Main DTL Linac
3 Ion species
HIDIF Study Group
20
New High Power Accelerator Designs
Radioactive Ions
Project
RIA
EURISOL KEK/ JAERI AUSTRON
Based
ANL
Saclay
Tokai
Beam Power (MW)
0,4
5
Energy (GeV)
0,9
1
Rep Rate (Hz)
n Factory
Neutron Spallation
Multipurpose
SNS
ESS
SPL
Austria
Oak
Ridge
Jülich
CERN
Korea
Saclaly
0,4
0,5
1.4
2x 5
4
20
3
1,6
1
1,33
2,2
1
ATW
KOMAC CONCERT TRASCO
Ion
Tritium driven
Fusion
ATF
HIDIF
Italy
Los
Alamos
GSI
25
30
100
~200
1
1
1
10
25
50
60
50
50
CW
CW
CW
CW
50
synchro- synchroDT or SC
Main Accelerator SC Linac SC Linac
SC linac
SC Linac SC Linac SC Linac SC Linac SC Linac DT Linac
tron
tron
linac
Start constr. Ongoing
Ongoing Operation Ongoing Ongoing Ongoing Ongoing Ongoing Ongoing Ongoing
Status
Approved
2004
study
study
2007
study
study
study
study
study
study
study
EPS HEP2003
CW
21
Hadrons
for Therapy
Extremely precisely controlled beams
Therapy oriented optimization
Economics of running the facility
Hospital level reliability
Decades of accelerator
physics and engineering
know-how
Hadron
Therapy
EPS HEP2003
Planned
In Operation
23
GSI beam-scanning technique: any shape
Plastic sheets immersed in water
The ions deposit the bulk of their energy
in the target volume.
E and I
changed within
a second. An
independent
control system
monitors the
beam position
every 100 ms
and intensity
every 10 ms. If
either deviates
by > 2% from
spec, beam is
shut off within
0.5 ms.
TERA
CERN
INFN
Design
protons
ions
Hadron
Therapy
photons
EPS HEP2003
C ions
G
S
I
s
i
m
u
l
a
t
i
o
n
T
E
R
A
24
Synchrotron
Radiation Sources
“the most important spin-off of HEP storage rings”
20 000
Users
Worldwide
PM Linear undulator
Brilliance is the word
UB : photons/mm2/s/mrad2/ 0.1% BW)
1021
PM Elliptical undulator
Undulators
UB
1015
SPring8
APS
Hard
X-rays
109
Wigglers
ESRF
Bending
magnets
Rotating anode
1900
EPS HEP2003
1950
Courtesy of Lenny Rivkin, PSI
2000
J.-L. Laclare
26
PERFORMANCE OF 3th GENERATION LIGHT
SOURCES
~10 nm
~1 nm
~1 Å
DIFFRACTION LIMIT
Circ. acc
upper limit (?)
Medium E
3d Gen. LS
ESRF
APS
Spring-8
Because of
radiation,
in circular
machines
emittance,
bunch length,
energy spread
are determined
by the lattice,
PHOTON ENERGY [eV]
EPS HEP2003
Courtesy of Lenny Rivkin, PSI
27
Linear vs Circular
The brightness depends on the
geometry of the source, i.e., on
the electron beam emittance
In a storage ring, the
electrons continuously emit
photons. This “warms up” the
electron beam and negatively
affects its geometry
Controlling the electron beam
geometry is much easier in a
linear accelerator.
Thus, linac sources can reach
higher brightness levels
EPS HEP2003
courtesy of L.Rivkin, SLS
28
Energy-recovery LINAC sources
However, contrary to the
electrons in a storage ring,
the electrons in a LINAC
produce photons only once:
the power cost is too high
Solution: recovering energy
Accelerating
section
EPS HEP2003
courtesy of L.Rivkin, SLS
Energyrecovery
section
Superconducting
is ideal
29
New Sources
meV
• FEL oscillators (IR, VUV)
l
mm
• Inverse-Compton-scattering table-top
sources
• Energy-recovery Linac based sources
(ERL)
10
KeV
EPS HEP2003
hn
• Self-amplified spontaneous emission
(SASE) X-ray FELs
Å
30
IR, UV Free
Electron Lasers
SR or LINAC
DRIVEN OSCILLATORS.
ENERGY RECOVERY
A
P
P
L
I
C
A
T
I
O
N
S
Spontaneous
emission
FEL Oscillator
Principle of operation

 ~ 
Spontaneous radiation resonance
condition: the electron slips back
by l every lu because of
e - photon speed difference
l(=0) ~ lu/ 2
l~lu/2
>l <
u
Linear Accelerator
electron Beam
TUNABLE !
(B, )
Undulator
Magnetic Field
Resonator Mirror
EPS HEP2003
Stimulated
emission
32
EUFELE
Storage Ring driven
FEL Oscillator
Pave = 500 mW @ 250 nm => Bpeak ~ 3 ·1024 UB (*) , Full spatial coherence
Dll = 3 ·10-4 FWHM
Lc= 0.8 mm
DT= 10 ps FWHM-> sL= 3 mm
Rep. Rate 4.6 MHz (216 ns)
Ppeak = 10 kW
EUFELE
S R- FEL
l=190 nm
Mirrors !
EPS HEP2003
33
Courtesy B. Diviacco (ELETTRA)
FEL for Industrial Applications
Beam
Energy
recovery
> 90%
EPS HEP2003
E= 48 MeV
t=0,4 ps
e=8 10-6 m
1.7 KW average
34
FELs for Industrial Applications
T. Jefferson Laboratory High Power Demo FELs
10 KW
“Powerful, multipurpose
free-electron lasers (FELs) driven by electron SRF accelerators
prospectively represent substantial, cost-effective
new manufacturing capabilities for industry.” (G.R.Neil, TJ Lab)
Polymer surface processing:
Micromachining
Metal surface processing
“amorphization” to enhance adhesion,
fabric surface texturing, enhanced food packaging,
induced surface conductivity.
ultrahigh-density CD-ROM technology,
micro-optical components
Micro-Electrical Mechanical Systems (MEMS).
laser glazing for corrosion
resistance and adhesion pre-treatments.
Electronic materials processing
large-area processing (flat-panel displays)
laser-based“cluster tool” for combined deposition, etching, and insitu diagnostics.
and:
........medical isotope production, fusion, ..................
EPS HEP2003
36
IR-UV Free-electron laser for
environmental and atmospheric research
University
of
Hawaii
Pan-Oceanic
Environmental &
Atmospheric
R esearch
Laboratory
EPS HEP2003
37
PEARL - Free-electron laser applications in
environmental and atmospheric research
LIDAR
University
of
Hawaii
Pan-Oceanic
Environmental &
Atmospheric
R esearch
Laboratory
University of Hawaii
EPS HEP2003
38
Free-electron laser
and the Space Elevator
The space elevator is essentially a cable with
one end attached to Earth and the other end
above geosynchronous altitude. Once in place
the cable can be ascended with mechanical
climbers.
Major component of the SE is the power
delivery system.
“”.....for delivery of high power at great distances a
FEL appears to be the obvious choice ......”
pro:
con:
· Minimal impact on cable or climber
· High power available
· Primary system located on Earth
· Mature technology
· susceptible to clouds
EPS HEP2003
39
Eureka Scientific, Berkeley, CA)
Free-electron laser
( less exotic) Space application
Laser-power beaming to generate electricity in
satellites is being seriously considered.
Using the same solar panels now in operation, the
electrical power deliverable to a satellite is
expected to increase by as much as a factor of ten
by using a FEL.
Free-electron lasers
in Surgery (and other medical applications)
 Most used wavelength band : IR
Main features that make a FEL a unique tool for surgical and medical applications :
 Tunability
“Tunability is the most critical attribute of FEL technology.”
( wavelength, power and pulse duration )
 Coherence ( spot size ...)
 Time structure
Free-electron lasers
2
in Surgery and Medical diagnostics

Tunability
Soft tissue vaporization
l
6.45
mm
EPS HEP2003
42
CSX Source
Collimated, intense, quasi monochromatic X-ray beam
The electron beam is generated by a 75 MeV/m
NLC type Linac section
EPS HEP2003
43
Towards
Soft X-ray
Sources
SASE FELs
Self-amplified spontaneous emission. X-ray free-electron lasers
(SASE X-FEL’s)
No mirrors
X-ray lasers: no mirrors  no optical cavity 
need for one-pass high optical amplification
R.Bakker
SASE strategy:
Long Wiggler
electron bunch
LINAC (linear accelerator)
“Microbunching increases the local electron density
and the amplification and creates very short pulses
L.Rivkin, SLS
EPS HEP2003
45
LEUTL (ANL)
530 nm
TTF SASE FEL
EPS HEP2003
47
TTF2 (DESY)
Goal: ≤ 6 nm
2004
EPS HEP2003
48
Many SASE FEL projects are under way …
YEAR
NAME
INSTITUTE
l [nm]
2000
2000
2004
2006
2008
2008
2011
TTF1
LEUTL
TTF2
SCSS
LCLS
BESSY
X-FEL
DESY
ARGONNE
DESY
SPRING-8
SLAC
BESSY
DESY
90
530
24-6
30-20
0.15
100-20
0.1
Why
Hard X-ray
Sources
ERL Sources , SASE FELs
Protein Data Bank
X-rays:
NMR:
EPS HEP2003
82% Dominate!
18%
courtesy of L.Rivkin, SLS
51
...engineering too !
60 KeV X-ray Diffraction study of
residual stress profiles in a US
railway section.
Transverse (Y)
(MPa)
Vertical (Z)
EPS HEP2003
Courtesy of D.J. Hughes
FaME38 (Facility for Materials
Engineering), Grenoble
52
Protein crystallography
Typical crystal size:
50 mm by 50 mm
Protein Structure
Reconstruction
Low divergence e.g (0.2
x 0.2 mrad) required for
high resolution
very high brilliance
Diffraction pattern
N. Ban et. al.
EPS HEP2003
Part of a Ribosome
53
courtesy of L.Rivkin, SLS
What electron beam quality does it take to operate a
SASE X-ray FEL?
Electron beam requirements
from a) to c) a factor of ~ 1000 in Ipeak/e
c)
b)
a)
TTF
SASE FEL
operation
EPS HEP2003
54
Low
Emittance
Injectors
Evolution of injector electron gun emittance
TTF SASE FEL : First Saturation
Warning: depends on bunch peak current
M.Ferrario
for the SLC first LinColl
SLC (SLAC)
Thermionic injectors
With sub-harmonic bunchers
BOEING
Photoinjectors
BNL
LANL-APEX
LANL-AFEL
BNL/UCLA/SLAC
EPS HEP2003
Goal
56
CRITICAL COMPONENTS OF A SASE FEL
SMALL EMITTANCE
RF Photoinjector e Gun
SHORT BUNCH
ACCELERATING STRUCTURES SC AND CW
TESLA
(DESY)
SC Cavity developed for Lin Coll
RF photoinjector developed for FEL
PITZ
(Zeuthen)
adopted by FEL
adopted by LinColl
BNL/SLAC 1.6 cell S-band RF Gun
courtesy of D. T. Palmer
EPS HEP2003
58
Compensation of emittance blow up at low
energy by space charge forces
z= 0.23 891
z= 1.5
z= 10
0.0 5
0.0 4
0.0 4
0
pr_ [rad]
0.0 2
Pr
Pr
0.0 2
0
-0. 02
0
-0. 02
-0. 04
-0. 04
-0. 05
0
3.5
0.0 01 0.0 02 0.0 03 0.0 04 0.0 05 0.0 06
R [m]
0
0.0 008 0.0 016 0.0 024 0.0 032
0.0 04
0
0.0 008 0.0 016 0.0 024 0.0 032
R [m]
0.0 04
R_ [m]
3
rms normalized emittance [mm]
norm.
emittance [um]
rmsrms
beam
size [mm]
rms beam s ize [mm]
2.5
2
1.5
Final emittance = 0.4 mm
1
0.5
Gun
Linac
0
0
2
4
z= 1.5
zZ_[m]
(m)
6
8
10
Z= 10
0.0 04
0.0 035
0.0 035
0.0 03
0.0 03
0.0 03
0.0 025
0.0 025
0.0 025
0.0 01
0.0 005
0.0 04
0.0 035
Rs [m]
0.0 02
0.0 015
Rs [m]
Rs [m]
z= 0.23 891
0.0 04
S-band Linac: HOMDYN simulation from photo injector to 150 MeV,
(RF Gun + 2 Travelling Wave Accelerating Sections)
0.0 02
0.0 015
0
-0. 003 -0. 002 -0. 001
0
0.0 01
0.0 01
0.0 005
0.0 005
0.0 01 0.0 02 0.0 03
Zs -Zb [m]
0.0 02
0.0 015
0
-0. 003 -0. 002 -0. 001
0
0.0 01 0.0 02 0.0 03
Zs -Zb [ m]
0
-0. 003 -0. 002 -0. 001 0
0.0 01 0.0 02 0.0 03
Zs -Zb [ m]
M. Ferrario’s working point, to be used for the LCLS and TTF-FEL II injectors
(1997 - 2002)
Performance
of the TTF
Photo-injector
Laser
System
TESLA specs.
A far from trivial
development
EPS HEP2003
Max-Born-Institut Berlin & DESY
60
Cathode production at
INFN Milano-LASA
Photocathode for TTF
QE
%
Installed on a FNAL-TTF Gun at FNAL
to TESLA specifications
Quantum efficiency distribution
over cathode surface. As produced.
EPS HEP2003
Several months lifetime at high QE
61
OUTLOOK / NEXT GENERATION LIGHT SOURCES
USER DESIRES 
Single shot imaging of
single biomolecular complexes
Ribosome
MPG
Time resolved studies of structural
processes during chemical and
biological reactions
Light induced structural changes
during photocycle
EPS HEP2003
62
courtesy of L.Rivkin, SLS
4th Generation X-ray Sources
The Energy Recovery Linac (ERL)
≤ 0.3 mm emittance (80 pC)
5-7 GeV
Super conducting energy recovery Linac
Cornell
with a SC
RF system
and energy
recovery.
r =Pbeam/ PRF >
> 200
EPS HEP2003
I. Bazarov et al. , CHESS Techn.Memo 01-003
Extraordinary flux.
Extraordinary brilliance,
adjustable via the photo injector.
Picoseconds bunch lengths.
Great flexibility in the timing
of the bunch sequences.
TESLA SC Cavities
Quic kTime™ and a TIFF (Unc ompr es sed) decompr es s or are needed to s ee this pic ture.
63
ERL
ERL
Short
Pulse
Performance
ps
EPS HEP2003
64
THE UK ERL LIGHT SOURCE DESIGN
, UK
EPS HEP2003
65
SASE FEL
UNBEATABLE BRILLIANCE
27
10
26
10
25
10
24
10
23
10
22
10
21
10
20
10
19
10
18
10
17
10
16
10
15
10
14
2
Average Brightness [Photons/sec/mm /mrad /0.1%BW]
(1030 - 1033)
10
2
HIGH AVERAGE
BRILLIANCE
(1022 - 1025)
SHORT PULSES
(1 ps – 50 fs)
TESLA FEL
TTF-FEL (M)
SLAC LCLS
ESRF
SLS
Spring 8
AP S
BESSYII, ALS
0.01
Bends
0.1
1
10
100
Photon energy [keV]
EPS HEP2003
66
TESLA XFEL at DESY
user facility
0.85-60 Å
3 compressors
multiple undulators
EPS HEP2003
67
X-ray FEL Scientific Case
(LCLS)
TESLA TDR
EPS HEP2003
68
LCLS at SLAC
1.5-15 Å
2 compressors
LCLS
one undulator
X-FEL based on last 1-km of existing SLAC linac
EPS HEP2003
69
4th Generation
X-ray Sources
New Acceleration Techniques
“Advancing the Accelerator Art”
(A. Sessler )
The maximum achievable accelerating field determines not
only the accelerator cost per GeV, an all important parameter for
VHE LinColl, but also its physical dimensions, crucial for most
applications (e.g. medical instruments).

R&D on the next generation “warm” LinColl has therefore led to
the development of very high frequency, high field RF systems
and of their power drivers.
NLC/JLC: 11.4 GHz, 75 MV/m (unloaded) an its MW Klystrons
Application
CLIC: 30 GHz, >150MV/m (unloaded), and a novel two-beam
powering scheme
EPS HEP2003
71
New Acceleration Techniques
“Advancing the Accelerator Art”
(A. Sessler )
 New technology ( and a great deal of Physics ! ): Laser or e-beam driven
plasma wake fields are in the main R&D line (triggered originally by dev’pt
of table–top TW lasers). Field gradients > 150 GV/m can be reached
Plasma oscillation wavelengths and longitudinal field values can be estimated from
1015 cm3
lp 
[mm]
no
Ez  100 no
[V / m]
no =plasma density
Peak laser powers today reach up to 100 TW with focused intensities of  1020 W / cm2 .
Future 6 orders of magnitude higher intensities could produce, according to some possibly
optimistic authors, very bright, 100 TeV, sub-picosecond intense e- pulses in a few cm.
Applications would certainly quite revolutionary
EPS HEP2003
72
Maximum acceleration with intense beams: Plasma
Wakefield Accelerator (PWFA)
Defocusi ng
Focusi ng
- - ---
-- - --
- - -- - -- --- - --
- -- --
- - -- - -- --
Decelerat ing
- - --
z
e- driv e bunch
- - ---
r
elec tr on beam
Ez
Accelerat ing
• Plasma wave/wake excited by a single relativistic electron bunch
• Plasma e- expelled by space charge forces
=> energy loss, focusing
• Plasma e- rush back on axis, induction field
=> energy gain
• Plasma Wakefield Accelerator (PWFA) = Beam Energy Transformer
Booster for high energy accelerator?
Courtesy of J. Rosenzweig (UCLA)
73
Plasma Wakefield Accelerator (PWFA)
Scheme: a leading bunch generates the plasma wave to accelerate a smaller trailing bunch.
In the extreme non-linear, ‘blow-out’ regime, when all electrons are effectively swept
from the beam path, the positive ions provide transverse focusing.
Example: The “SLC afterburner” proposal(1) to double the SLC energy
3 nC drive bunches generated by the present SLAC linac are compressed by a factor of ~10,
to ≈60 mm length.
The trailing bunch acceleration is computed to be 8 GeV/m over a 7 m long plasma cell,
totalling ≈ 56 GeV and thus doubling the present SLC energy
The %6 GeV beam energy spread is 20% and the required transverse bunch size to
be obtained by means of plasma lenses is ≈ 1 mm ( well below the resolution of
present tracking programmes). Beam stabilty and operabilty of the plasma lenses are
open questions.
(1) S.Lee et al., Pys.Rev ST AB 5 01001 (2002), http://prst-ab.aps.org/pdf/PRSTAB/v5/i1/e011001
Courtesy of J. Rosenzweig (UCLA)
74