a study of the feasibility of a laser system for the clic photo-injector

Rutherford Appleton Laboratory

A STUDY OF THE FEASIBILITY OF A LASER SYSTEM FOR THE CLIC PHOTO-INJECTOR

Ian Ross, Central Laser Facility, RALSteve Hutchins, CERN


REPETITION TIME

PULSE TRAINDURATION

PULSEENERGY

TIME BETWEEN PULSES

STABILITY

PULSE TRAINS


PHOTO-CATHODE SPECIFICATIONS

CLICCTF3

UV energy per micropulse 5µJ 0.84µJ

Pulse duration ‹10ps ‹10ps

Wavelength ‹270nm ‹270nm

Time between pulses 2.13ns 0.67ns

Pulse train duration 91.6µs 1.4µs

Repetition Rate 100Hz 5Hz

Energy stability ‹ 0.5% ‹ 0.5%

Laser/RF synchronisation ‹1ps ‹1ps

Reliability 109 shots between servicing4 months at 100Hz


LASER SPECIFICATIONS

CLICCTF3

Energy per micropulse 100µJ 17µJ

Total pulse train energy 4.3J 32mJ

Pulse train mean power 47kW 25kW

Laser average power 430W 0.15W

Efficiency of IR to UV 5% 5%


BASIC LASER SYSTEM

PULSE GENERATOR/OSCILLATOR OPTICS AMP 1 OPTICS AMP 2

TARGET OPTICS HARMONIC FINAL AMP OPTICS CONVERSION


KEY ISSUES

•0.5% Stability and Controllability

•47kW pulse train power

•430W average power

•1ps synchronisation

•Uniform flat top beam profile

•Design for CLIC useable for CTF3


BASIC DESIGN FEATURES

Stability - CW or QUASI-CW laser DIODE-PUMPING

careful design FEEDBACK system with

rapid response

Pulse train power - diode pump power = COST47kW pump efficiency - AMP. DESIGN

storage efficiency - MATERIAL

Average power - thermal dynamics420W MATERIAL FRACTURE

OPTICAL DISTORTION

Pulse duration (10ps) - MATERIAL

Simple design - few amplifiers with high gain MATERIAL

UV generation - high efficiency gives min. COST

Beam quality - OPTICS


BASIC LASER SYSTEM

PULSE GENERATOR/OSCILLATOR OPTICS AMP 1 OPTICS AMP 2

TARGET OPTICS HARMONIC FINAL AMP OPTICS CONVERSION

cw MLNd:YLF1047nm

5ps

Nd:YLF Nd:YLF

Nd:YLF4ω BBO

262nm


QUESTIONS TO ANSWER

•Is it feasible?

•Is it affordable?

•Where are the uncertainties in the physics/technology?

•What programme will establish confidence and lead to an optimised design?


ND:YLF OSCILLATOR

Available commercially

Expected performance - 10W @ 0.5GHz (CLIC) 1.5GHz (CTF3)

- 5ps @ 1047nm

NUMBER OF AMPLIFIERS

Available input energy per pulse = 10nJRequired output energy per pulse = 100μJ

Required amplifier gain = 10,000

Need to limit gain per amplifier to about 20.

Simplest system has 3 amplifiers with average gain per amplifier of 22.


FINAL AMPLIFIER DESIGN - PHYSICS

Requirements - diode pump power › output power (47kW)

- efficient extraction of diode power - high stability along the pulse train

Simulations carried out for single and double pass amplifiers.

For maximum stability the trick is to operate in quasi-steady-state mode with continuous pulse train input.

Sensitivity to 1% changes in input energy and pump power

0 50 1000.99

1

1.01

(microseconds)

Out

put

Ene

rgy

(rel

)

0 50 1000.99

1

1.01

(microseconds)

Out

put

Ene

rgy

(rel

)


AMPLIFICATION SCHEME TOLERANCES FOR 0.5% STABILITY

AMP 1 X100 AMP 2DOUBLE X20 AMP 3PASS DOUBLE X4

PASS SINGLE PASS

100% 40% 2% 0.5%

QCW PUMP DIODE ARRAY MODULES

0.5%2%40%


FINAL AMPLIFIER DESIGN - LAYOUT

STACKED DIODE ARRAY

WATER COOLINGSILICA TUBE

AMPLIFIER ROD

MIRROR COATINGS

•All rays from diodes (20 x 80deg) collected by rod

•1cm diam. gives excellent absorption tolerant to pump wavelength and polarisation

•un-absorbed fraction reflected back into rod

•water blanket improves coupling efficiency


THERMAL EFFECTS

•Thermal fracture of rod at CLIC 420W - this may be a problem for Nd:YLF

•Thermal lensing in the amplifier rods - Nd:YLF gives both a spherical and a

cylindrical lens which must be compensated


FOURTH HARMONIC GENERATION

ω ω 2ω 2ω

4ω2ω

BBO BBO η = 50% η = 50%

•Predicts 25% efficiency overall

•Literature reports 25% efficiency

•Design exercise assumed 10% - achievement of say 20% would substantially cut the cost of the laser.


OPTICS DESIGN

REQUIREMENTS

•Stability requires generation of a single mode beam.

•Optics for beam size changes.

•For maximum efficiency the beam profile must be a flat top in both amplifiers and harmonic crystals.

•Beam profile to be flat-top on the photo-cathode.

•Compensation for thermal lensing in amplifiers.

•Minimise the effects of diffraction to keep intensity variations across the beam to less than 2:1.


OSC AMP 1

RELAY CYL PC LENS GATE

APODISER

RELAYCYL RELAYLENS

LENS

AMP 2 AMP 3

CYL LENS

RELAY DELAY PC FHG RELAY LINE PHASE RETARDER

PHOTOCATHODE

OPTICS SCHEME FOR PHOTO-INJECTOR LASER SYSTEM


CONCLUSIONS

•FEASIBLE

•AFFORDABLE Total pump power for CLIC ~75kW @ $7/W gives $0.5M for the diode arrays and a system cost of perhaps $1M

•CRITICAL ISSUES HAVE BEEN IDENTIFIED

•FUTURE PROGRAMME HAS BEEN PROPOSED


FUTURE PROGRAMME

Experimental Programme to:

Resolve uncertainties - Increased confidenceObtain more accurate data - Definitive design

• Oscillator operation at 0.5GHz and 1.5GHz CLIC and CTF3

commercial systems have lower repetition ratesCERN

• Feedback control of a) laser pump diode current b) fast optical gate

needs to be FAST (μs response)ACCURATE (0.1%)

CERN• Amplification - all aspects of design

- thermal effects (lensing/fracture limit) RAL

• Fourth harmonic generation - maximum efficiency?CERN

• Check laser damage thresholds


PROPOSED RAL PROGRAMME

Amplifier development - test as close to design parameters as possible - at minimum cost

STACKED DIODE ARRAY1.5kW each

CYLINDRICALLENS

AMPLIFIER ROD5 X 50 mm

MIRROR COATING

WATERCOOLING

SILICA TUBE

Scaled down version with short length - 4.5kW pumpGives measurable small signal and saturated gain

Good test of: pump efficiencygainsteady state saturated operationextraction of stored energythermal effects

Develop theory and simulations

a study of the feasibility of a laser system for the clic photo-injector

Documents