Optimizing Design of
SRF Electron Guns
Joe Bisognano
University of Wisconsin SRC
Starting Point:
What Do Light Source Users Want?
• Frontier is where physical, chemical, and
biological systems can be viewed on their
characteristic temporal, spatial, and energy
scales—femtoseconds, nanometers, millivolts
• Dynamics rather than statics (today’s 3rd
generation light sources) of fundamental
processes, diffractive imaging of nanoscale
structures, nonlinear phenomena
Where is Leverage
Lower energy per pulse: Signals for experiments limited by damage or space
charge. Giant pulses can be overwhelming
Higher rep rate: Could compensate for smaller pulses without loss of
average flux. Megahertz usable since pump lasers at megahertz now
Shorter pulses: Time resolutions of 0.1 ps to fs and lower are needed for
studying atomic and electronic motions or relaxations
Stability: Pulse to pulse variation of SASE unloved
Higher average flux: 2D imaging or photon in/photon out flux starved
For Example: Wisconsin Free Electron Laser (WiFEL)
Next Generation VUV/Soft X-ray Light Source
4
Cost Breakdown of a Soft X-ray FEL
• Conventional wisdom: ~ 2.5 GeV with few cm period undulators with
cost at least a good fraction of a billion dollars and probably a good
bit more
• Cost Breakdown
– Linac : 20-25% (less w/ pulsed RT rather than CW SRF)
– Injector, R&D, etc.: 5-10%
– Photon Generation: 20 % (fifty/fifty undulator and beamline; clearly
depends on number of beamline, say six)
– Maybe scalable stuff: civil and contingency: 50%
• Linac energy reduction and multiple users provides best value
• That is, high rep rate at lower charge and lowest normalized
emittance
5
Phased Approach to a Full Service FEL Facility
7
Electron Gun for CW WiFEL
Gun repetition frequency
I peak at a soft X-ray undulator
5 MHz or higher
1000 Amps
DE /E at a soft X-ray undulator
< few 10-4
Normalized eTransverse
<1 mm-mrad
Bunch length at undulator, rms
Charge/bunch
I average
70 fsec (seed jitter concerns)
200 pC
1 mA
At lower charge per bunch, higher rep rate (up to 200 MHz) and
lower emittance (tenths of mm-mrad) possible
Wisconsin SRF
Electron Gun Concept
Inherent Quarter Wave Advantages
Over Elliptical Gun Designs
•
•
•
•
•
Compact structure, so low frequency practical
Extremely high mechanical stability
BCS losses go as Freqency2 , so 4.2K operation possible
EPeak /ECath is less than elliptical, so Higher ECath
Bpeak / EPeak is less than elliptical, so higher quench threshold
UW Gun
BNL QWR
FZD Gun
EPeak /ECath
1.31
2.63
2.7
Bpeak / ECath , mT/MV/m
1.57
1.92
5.76
• Builds on work at BNL and NPS
A Brief Interlude
But Deemed Too Persnickety from
Fabrication Point of View
Blowout with Superconducting RF Electron Gun
• High gradient allows operation in so-called “blow out” mode
• SRF offers higher exit energy; less time for space charge to do evil
• Lower frequency for temporal field flatness (quasi-DC)
O.J. Luiten, et al., PRL 93, 094802-1 (2004).
S.B. van der Geer, Proc of Future Light Sources 2006,
Ellipsoidal bunch expansion
13
• Blow-out Mode Bunches Produce Uniform Charge
Distribution
•
Less susceptible to collective effects
Bunch with Initial
Longitudinal
Modulation
“Bad” laser
x vs z Z=0
Distribution in t
Distribution in t, Z=13 m
Bunch with Initial
Transverse
Modulation
Z=0
Histogram in x
“Bad” cathode
Histogram in x, Z=13 m
Key Gun Parameters
•
•
•
•
•
Electric field at cathode – up to 45 MV/m
Peak surface magnetic field – 93 mT
Dynamic power loss into He – 39 W at 4K
Q – 2.5E9
Frequency – 199.6 MHz
Key Bunch Parameters
•
•
•
•
•
•
RMS bunch length at gun exit – 0.18 mm
Cathode spot ~1 mm for 0.85 mm-mrad thermal emittance
At gun exit, dp/p ~ 2.5%, divergence – 7 mrad
Q – 200 pC
Kinetic energy – 4.0 MeV
With smaller spot, can be operated in lower charge modes with
lowered emittance; also more exotic cathode materials
15
Sequence of Events
for Wisconsin Electron Gun
•
•
•
•
Start of three year grant in August 2010
~FY 2011: final design, procurements, and vault prep
~FY 2012: fabrication and subsystem installation
~FY 2013: final integration, commissioning and beam tests
– Expect commissioning to start in April-May
• Total DOE program $4.125 million
Wisconsin
Superconducting
Electron Gun
• RF system uses Low level RF controls from
JLAB upgrade
•
•
Standard EPICS interface
Existing hardware base
20 kW
200 MHz RF
Harris Corporation
Broadcast Communications
Division
19
• Active tuner control
Cavity compression assembly
LLRF Controller
Mechanical Drive
350
Measured Delta Freq vs
Force
Displacement, microns
300
Calculated
250
200
150
100
50
0
-10000
0
10000
20000
Delta Freq., Hz.
30000
40000
RF Coupler and HPA and LLRF
• Power is introduced through a ceramic rf window and a tuned resonant
structure.
• Relatively low power, <10kW, at 1 mA of beam
21
•
Particle Free Cathode Holder and Transfer Arm
• Transfer mechanism and cathode holder specifically designed (and tested) to be
particle free in operation
• Support structure needs to be accurate from 10 to 20 microns in every axis and
linear direction. The cathode adjustment support is fixed to the vacuum vessel
• The cathode stem is designed to allow nitrogen to flow through a channel forcing it
near the exchangeable stalk insert
• Cavity provides rf short circuit and
thermal gap between the warm
cathode holder and the srf cavity
• The small gap region acts to minimize
the radial field across the cathode
holder face
• Bellows in filter allows final alignment
and tuning of filter
• Copper plated SS acts as to manage
RF heating
X position, mm
• Cavity Filter Design Details
Z position, cm
Ar:O Processing of SRF Cavity
• Need to clean cavity after receipt from Niowave, but
too large for conventional HPR facilities with He
vessel attached
• New technique demonstrated at SNS and JLAB using
plasma processing
• Uses RF driven Ar:O plasma to “ash” surface
contaminants
• Plasma process monitored spectroscopically
Plasma Glow
25
Spectrum Intensity vs Wavelength in Nanaometers
• Argon dominates spectrum; makes seeing contaminants hard.
• Use techniques from semiconductor industry for etching SiO using rf
plasmas;
• Look at 483 and 520 nm lines over time.
All major lines are Argon
CO lines
How semiconductor processing
determines the oxide is ‘done’1
Note amplitude of emission line drops to half initial value at completion.
1. John G. Shabushnig, Paul R. Demko and Richard Savage, Proceedings of Mat. Res. Soc., Vol 38,
Materials Research Society, 1985
CO Line Strength Before and After Plasma Processing
900
147mT -10.7db
147mT -14.3db
Initial intensity of CO emission lines at two levels of Rf power
148mT -14.3db
148mT -10.7 db
700
Intensity, Arbitrary counts
Final Intensity of CO emission lines after
Plasma processing
500
300
100
500
-100
505
510
515
520
525
530
Wavelength, nm
535
540
545
550
• High Temp Superconducting Solenoid and
Compensating Quad
•
•
Magnet can be closer to the cavity; Closer the focusing field is to cathode, the better
the emittance compensation
Field specified to minimize emittance dilution from quad and dipole terms
Downstream superposed skew and normal quad magnets to remove particle rotation
caused by quad terms in solenoid reduces final transverse emittance
2.10E-06
Effect of Downstream Correction Quad Rotation
Angle on Emittance
2.05E-06
Normalized emittance, mm-mr
•
Nominal emittance with no quad error term is 1.687e-6
Nominal emittance with no correction term is 2.04e-6
2.00E-06
1.95E-06
1.90E-06
1.85E-06
1.80E-06
1.75E-06
1.70E-06
150 mm Solenoid, using -7e-3 T/m for quad component. No Dipole moment.
Compensationg quad is at 0.6 m downstream of cathode and 150mm length.
0
10
20 Quad angle of
30rotation, deg 40
50
60
•
Synchrotron and Materials Physicists For Cathode
Research Integrated into Program
17.6
17.2
16.8
17.6
eV
Schematic view of the corrugated film geometry and the
wave interference or propogation patterns. The inset
shows the Fraunhofer single-slit diffraction pattern as a
function of Dkx.
G. Bian, T. Miller, and T.-C. Chiang, Phys. Rev. B 80, 245407 (2009)
16.0
16.8
eV
EXAMPLE:
• Bi thin film in the rombohedral phase.
The surface state ~0.4 eV below the
Fermi edge (blue spot) only has +2°
emission angle.
• Potential for prompt emitter with very
low thermal emittance
16.4
17.2
15.6
16.4
4
0
deg
-4
16.0
Spectra-physics Tsunami (oscillator) + Spitfire (amplifier) system
• Pulse duration: 100 femtoseconds
• Repetition rate: 1 kHz – 1 Hz
• Pulse energy
• Up to 4 mJ per pulse at the fundamental
(800 nm)
•~ 1 mJ per pulse at the second harmonic
(400 nm)
• ~ 300 microjoule per pulse at the third
harmonic (266 nm)
• Average power: 4 W
Current Scope
• Demonstrate single bunch beam dynamics
and operation of SRF gun
• Low repetition rate (kilohertz) drive laser
• Cu Cathode Used for Initial Operation
– Little chance of cavity contamination from
evaporated cathode material
– Cathode will not degrade over time like
semiconductor
– No cathode preparation chamber needed
33
Overall layout of SRF gun facililty
3D engineering drawing of Wisconsin electron gun hardware
Wisconsin SRF Electron Gun
Preparations for final e-beam
weld
Bake at JLab to prevent Qdisease
36
Frequency Map
• Map which starts with a cold cavity at the correct frequency
and moves back through the series of production steps
producing an expected resonant frequency at each step
• Goal is to understand any deviations from the calculated
frequency map and apply that knowledge to next generation
FEA to Evaluate Stress and Deformation
Freq,
MHz
State
Nominal, 4 K
199.58953
Remove 1600 lb
preload on tuner
199.65256
D Freq,
MHz
-
D volume,
in^3
6269.213
6267.753
-1.46
Warmed to 273 K
199.3704
-0.28216 6294.653
Skin depth vs temp at
200 MHz
199.3185945 -0.05180 6295.853
26.9
Remove vacuum load 199.2485945
-0.07
6300.243
Change in permitivity,
fvac/fair
199.1947645 -0.05383 6300.243
4.39
Undo BCP etch
Final weld shrinkage,
0.7 mm
0.06303
Volume,
in^3
199.3688075 0.174042 6282.793
199.280
-0.088
6294.87
TABLE 1. Steps from cavity blank to final frequency
37
1.2
0
-17.45
12.08
Tests at Niowave successful
38
Preliminary Tests Successful
• Initial cryogenic test at Niowave successful
– Low field Q of 3 109
– Gradients of about 7 MV/m obtained, limited by test configuration
– Demonstrated potential to reach design Q and design gradient (40
MV/m) after final processing at Wisconsin
•
•
•
•
39
Cavity installed in helium vessel and delivered to SRC
Cold shock test carried out
Plasma processing
Integration under way
Titanium Helium Vessel with Niobium
Cavity Inside
40
Cold Shock Test
Cryostat
Configuration of quarterwave cavity
superconducting RF electron gun.
Nitrogen Shield
Magnetic Shield
Phase II Proposal
• 3 years more years
• Key thrusts
– Detailed measurements as function of key
parameters, establishing technology reach
– Helium refrigerator for extensive testing program
– High repetition rate laser for high average current
operation (5-40 MHz, milliamp average current)
– High QE photocathodes and exotic photocathode
material
Acknowledgment
Wisconsin FEL Team
49