The E RL Injector Project at Cornell University The E RL Injector - - PowerPoint PPT Presentation

Bruce Dunham

For the Cornell ERL Team

Jefferson Lab Sem inar Decem ber 4 , 2 0 0 7

The E RL Injector Project at Cornell University The E RL Injector Project at Cornell University

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

2

CESR at Cornell

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

3

ERL at Cornell

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

4

Outline

• Project Overview
• Gun and Laser Progress
• Beam Experiments and Results
• SRF and RF
• Diagnostic Beamlines
• Construction and Commissioning

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

5

Overview

Our Charter . . .

Build an injector for an ERL to demonstrate we can produce a beam with the required properties Understand the limitations in the injector (both physics and technology) to allow for improved design in the future Develop a cost estimate for a full ERL

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

6

Injector Requirements

• up to 100 mA average current, 5-15 MeV beam energy
• norm. rms emittance ≤1 μm at 77 pC/bunch
• rms bunch length 0.6 mm, energy spread 0.1%
• Achieve gun voltage in excess of 500 kV
• Demonstrate photocathode longevity
• Cleanly couple 0.5 MW RF power into the beam without affecting

its transverse emittance.

• Control non-linear beam dynamics: over a dozen of sensitive

parameters that need to be set just right to achieve the highest brightness

• Instrumentation and tune-up strategy
• Drive laser profile programming (both temporal and spatial)

Many Challenges!

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

7

ERL Injector Layout – L0 Area

Diagnostic Beamlines 600 kW Dump Injector Cryomodule Photocathode Gun

• Limited diagnostics after the gun (before the cryo-

module)

• Full interceptive diagnostics capabilities at 5-15 MeV
• Limited full power diagnostics

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

8

Gun and Laser Photoemission Gun and Laser System

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

9

750 kV Gun

Electron beam out Cathode Entry Laser in

Max capabilities: 750 kV 100 mA

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

1 0

750 kV Power Supply

750 kV, 100 mA DC supply Kaiser Systems, Inc in Beverly, MA

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

1 1

Inside the SF6 Tank

Floating ammeter mounted on the processing resistor

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

1 2

GaAs Photocathode

GaAs is still our cathode of choice . . .

• good quantum efficiency
• low thermal emittance
• fast time response (@520 nm)

But . . .

• need extreme UHV
• limited lifetime
• minimum thermal emittance near

bandgap (lower QE)

• thermal emittance degrades at higher

QE . . . We’ll willing to try other cathodes

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

1 3

Load Lock System

• Load lock chamber with quick bakeout

capability

• Heater chamber
• Cathode preparation and transfer chamber

Can swap a fresh cathode into the gun in ~30 minutes

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

1 4

HV Performance

• For optimum emittance, need to operate between 500-600 kV
• So far, we have only reached 420 kV
• What are the problems and how to solve them?

Must have a way to control field emitted electrons near the insulator Prepare electrodes for minimal field emission

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

1 5

Ceramic Properties

The resistive coating on the first ceramic was not done well (R ~ 7000 G-Ohm). Experienced a vacuum leak due to punch-through at 330kV CPI made a second ceramic with a better coating ~100 G-

• Ohm. Good up to ~420 kV

Must have a way to bleed off any field emitted electrons

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

1 6

Third Try

Daresbury Lab purchased a ceramic with a bulk resistivity doped alumina, which has worked well (480 kV). We just ordered one for

• ur geometry.

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

1 7

HV Testing – Large Area Electrodes

3-4 mm

0 to -125 kV Pico-ammeter anode Test Electrode

150 mm

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

1 8

Initial Results

Field Emission Chamber Results

50 100 150 200 250 300 350 400 450 500 5 10 15 20 25 Field (MV/m) I (nA) hand polished titanium

Should be okay for gun electrodes too, but the ‘technology’ does not transfer

This titanium disk was hand polished and reached a field higher than the gun will see.

Max field on the cathode

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

1 9

1. Either hand polish or electro-polish metal electrode 2. After hand polished, ultrasound in hot soap and water, rinse in DI water and store in DI

• water. After electro-polishing, store under DI water

3. Transfer to a clean room environment 4. Mount the sample on the HPR system – (high pressure rinse system) 5. HPR for 2 hours 6. Remove and let dry in the clean room 7. Store in a clean, sealed container until ready for installation 8. Remove from sealed container in a clean room 9. Final cleaning using a commercial sno-gun 10. Install in system

SRF-like cleaning procedure

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

2 0

“SRF-clean” results

50 100 150 200 250 300 350 400 450 500 10 20 30 40 Field (MV/m) I (nA) System max is 35 MV/m

Following the new procedures, the ‘good’ Ti electrode improved from 20 MV/m to nearly 35 MV/m (pink to yellow curve) The blue curve shows the results for a SS disk that was electro- polished, then ‘SRF-cleaned’

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

2 1

HV Testing – Real Gun Parts

Pre-test gun parts before installing in the gun.

HPR equipment for large gun electrodes

Field Emission for Titanium Gun Stalk Piece

• 5.00

10.00 25.00 40.00 55.00 70.00 85.00 100.00 115.00 130.00 145.00 160.00 175.00 190.00 0.000 2.000 4.000 6.000 8.000 10.000 12.000 14.000 Electric Field (MV/m) Current (nA) Before HPR After HPR

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

2 2

New Gun Assembly Procedure

1. Clean all electrodes per the ‘SRF-clean’ procedure 2. Pre-test parts to the max field they will experience at 750kV 3. Enclose gun in portable clean room 4. Vent, maintaining a large nitrogen purge 5. Carefully remove old parts, wipe out any particles, wafer chips, etc 6. For each new part, clean with sno-cleaner before installing. 7. Pre-test all bolts and fasteners for particle generation before use (test on a bench with a particle counter). Avoid plated bolts unless you are sure they do not flake. 8. Do not use aluminum foil to cover parts or flanges – we have found tiny pieces of foil stuck to electrodes. 9. For surveying, cover all but one port at a time if possible. Use o-ring sealed covers instead of foil, or clear plastic wrap if you need to see through for alignment.

• 10. Pump out slowly to reduce the chance of stirring up dust that may be in the chamber
• r ion pumps.

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

2 3

Present Gun Status

• Reached 420 kV, beam experiments carried out at 250/350 kV
• 20 mA DC current obtained (no life-time measurements yet)
• 70 pC/bunch at low duty factor
• GaAs cathode performance typically 6-10% QE
• New ceramic on order
• Continue tests with electropolishing/HPR of electrodes
• HV modeling for field emission mitigation/future insulator

Next Steps: Currently:

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

2 4

Fiber Laser Description – 50 MHz

Pump diode

Yb fiber

Oscillator Pre-Amplifier Amplifier 15 mW 300 pJ 60 mW 1.2 nJ 4 W 80 nJ

5 4 3 2 1 Signal power [watts] 15 10 5 Coupled pump power [watts] efficiency: 60%

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

2 5

Pump diode WDM Mirror Mirror Grating Isolator

HWP QWP QWP PBS Mirror

Yb fiber

λ = 1040 nm pulse duration ~ 2.5 ps power ~ 15 mW fr ~ 50 MHz

The 50 MHz Laser Oscillator

8 Intensity [a.u.]

• 6
• 4
• 2

2 4 6 Time [ps] 2.6 ps

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

2 6

Laser Shaping

We use an ‘optical pulse-stretcher’ to get 20-40 ps flat-top pulses from a 2 ps laser (DPA – divided pulse amplifier) Gauss to flat top transformation using a commerical aspheric lens (Newport Corp) A ‘ b e e r

• c

a n ’ d i s t r i b u t i

• n

i s t h e g

• a

l

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

2 7

Laser Summary

The laser itself works as advertised, need work on

• Synchronization
• Beam shaping
• Pulse control
• Transport to the gun
• Stability control (position, power)
• Laser beam ‘halo’
• 1.3 GHz oscillator
• Sensitivity to acoustical noise

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

2 8

Data Beam Experiments and Data

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

2 9

Initial Beam Tests

• Goal: full understanding of the beam

phase space from the gun

• Gun & diagnostics line
• Full phase space characterization

capability after the gun

• Temporal measurements with the

deflecting cavity

• Lifetime studies

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

3 0

Thermal emittance

• kT⊥

= 121±8 meV at 520 nm

• r 0.49 mm-mrad per 1 mm rms
• GaAs still best overall perform.

GaAs GaAsP

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

3 1

Emittance measurement system

• Uses a pair of 20 μm precision

slits to sample the beam

• Instead of scanning the slits,

pairs of identical coils deflect and scan the beam across the slits

• armor slit intercepts most of the

beam

• kW beam power handling

measured phase space

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

3 2

70 pC/bunch

log scale

1 2 3 4 5 1 2 3 4 5

Good agreement with Astra prediction: 77 pC/bunch: about 2 mm-mrad data astra

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

3 3

Cathode Response Time

A deflecting cavity is used to transform bunch length into transverse spot on a viewscreen. This gives a direct measurement of bunch length

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

3 4 Pump diode

Yb fiber Fiber Stretcher PZT mirror Fast Photodiode

1300 MHz MO

Extract 50 MHz clock

50 MHz

• scillator

PID filter photocathode electron beam Deflection cavity

Cathode Response Time

Laser to RF synchronization has not worked well due to acoustical noise – phase jitter is too high

Shaped laser pulse E beam

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

3 5 Pump diode

Yb fiber Fiber Stretcher PZT mirror Fast Photodiode

50 MHz

• scillator

photocathode electron beam Deflection cavity

TWT

1.3 GHz filter Manually tune to 50 MHz viewscreen

Cathode Response Time

Now, use the laser signal to drive the cavity for better

• synchronization. The

RF controls still have ~2 ps of jitter, we need < 1 ps for.

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

3 6

Cathode Response Data

laser laser laser e beam e beam e beam 2 crystals 3 crystals 1 crystal

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

3 7

Upcoming Experiments

• Complete cathode response data
• Emittance measurements using 77 pC/bunch and

‘beer can’ laser shape, other shapes

• Push towards 100 mA

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

3 8

SRF and RF

SRF and RF Systems

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

3 9

ERL injector klystron

K3415LS S/N 03 Transfer Curve @ CORNELL

20 40 60 80 100 120 140 160 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Drive Power (W) Output Power (kW)

45kV 42kV 37kV 38kV 40kV Efficiency 51.1% Efficiency 54.8%

• e2 v designed a high CW pow er klystron.
• Param eters of this 7 -cavity tube: m ax. beam voltage 4 5 kV,

current 5 .8 7 A, full pow er collector, at m ax. output pow er of 1 3 5 kW the efficiency is > 5 0 % , gain > 4 5 dB, bandw idth is > ± 2 MHz @ 1 dB and > ± 3 MHz @ 3 dB.

• The first tube ( SN0 3 ) w as delivered and successfully tested

at Cornell on March 6 – 8 .

• Transfer curves w ere m easured for several HV settings.

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

4 0

Coupler design highlights

Compress Air Inlet for Bellows Cooling Compress Air Inlet for Window Cooling Air Outlets 300 K Intercept 80 K Intercepts 5 K Intercept

Design features:

• The cold part w as com pletely redesigned using a 6 2 m m , 6 0 Ohm coaxial line for stronger

coupling, better pow er handling and avoiding m ultipacting

• Antenna tip w as enlarged and shaped for stronger coupling
• “Cold” w indow w as enlarged to the size of “w arm ” w indow
• Outer conductor bellow s design w as im proved for better cooling ( added heat intercepts)
• Air cooling of the w arm inner conductor bellow s w as added

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

4 1

HOM Loads

Total # loads 3 @ 78mm + 3 @ 106mm Power per load 26 W (200 W max) HOM frequency range 1.4 – 100 GHz Operating temperature 80 K Coolant He Gas RF absorbing tiles TT2, Co2Z, Ceralloy

TT2-111R Co2Z Ceralloy

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

4 2

2 Cell Cavity

Frequency 1300 MHz Cells per cavity 2 R/Q 222 Ω Voltage 1-3 MV Gradient 5-15 MV/m Qo @ 2K >1010 Qext 4.6·104 – 4.1·105 Active length 0.218 m Total length 0.536 m

• 5 cavities tested, all meet specs, E > 15 MV/m

(most E > 20 MV/m) , Q > 1010 @ 2K

• Only BCP, no 800C treatment
• Two tested for H disease, no H disease

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

4 3

Horizontal Cavity Test

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

4 4

Horizontal Test Results

Low Q probably due to dust from broken

• tiles. Now cold testing all HOM loads

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

4 5

RF Power Distribution

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

4 6

Diagnostic Beamlines

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

4 7

L0 layout: 15 MeV straight-thru

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

4 8

L0 layout: merger & chicane

merger diagnostics chicane

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

4 9

Diagnostics overview

• Beam position resolution: 10 μm (spec)
• Energy spread resolution: 10–4
• Transverse beam profile resolution:

30 μm (viewscreens) 10 μm (slits) 30 μm (flying wire)

• Angular spread resolution: 10 μrad
• Pulse length (deflecting cavity&slits): 100 fs
• RF phase angle: 0.5°

Ability to take phase space snapshots of the beam, both transverse planes, and longitudinal phase space

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

5 0

Construction in Progress

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

5 1

Summary

• Construction and utilities finish Feb/March
• Move the gun in March
• Cryomodule complete March/April
• Commissioning begins in April

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

5 2

Acknowledgements

This work is supported by NSF. PHY-0131508

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

5 3

Cathode time response – Part 1

• Measured response time from GaAs

and GaAsP at different wavelengths

• GaAs response @ 520 nm on the
• rder of a picosecond
• Diffusion model correctly describes

fast response and a slow tail

measured temporal response GaAs diffusion model: fit to data response to a 100 fs pulse 800 nm: 15 ps 520 nm: 0.83 ps 50% emission point

50 % 18 %

expected temporal profile

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

5 4

ERL Injector Layout – L0 Area

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

5 5

Vacuum Performance

• Want extreme UHV levels to reduce ion back-bombardment (which is

the key for long lifetime)

• Many getter pumps (16 strips, 20000 L/s)
• Stainless steel parts fired at 400C in air to reduce hydrogen out-

gassing (similar to LIGO)

• System vacuum bake to 150C

Extractor gauge readings as low as 5x10-12 Torr

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

5 6

Buncher

Decem ber 4 , 2 0 0 7

• B. Dunham Decem ber 4 , 2 0 0 7

5 7

Injector Cryomodule

Cold Part of RF Power Coupler Motorized Frequency Tuner Cold He Gas Return Pipe Support Posts 2K Liquid Supply Line Piezo Tuners HOM Loads