1

Fabry-Perot cavity & pulsed laser

J. Bonis, V. Brisson, J.N. Cayla, R. Chiche, R. Cizeron, J. Colin, Y. Fedala, G. Guilhem, M. Jacquet-Lemire,

D. Jehanno, L. Losev, R. Marie, K. Moenig, V. Soskov, C. Sylvia, A. Variola and F. Zomer

LAL-Orsay & Lebedev Inst.-Moscow

Outline •Introduction: laser needs for Compton scattering at ILC•Fabry-Perot cavity, in pulsed regime•R&D : description & status

2

e laser(E≈1eV) E

Photondetector

Photondiffusé

Electron detector

Ee, e

dipole

ediffusé

Polarimeter

e+ polarised source : KEK scheme

Compton scattering e + laser → e +

Fit Eand/or Ee, e → e beam polarisation

3

Laser characteristics

• Polarimeter– 1ps pulsed laser – ~ 10-100J/pulse @ ~5MHz repetition rate– small beam waist 0≈ 100m (→ spot size 0≈50m)

• e+ polarised source (KEK + K. Moenig schemes)– 1ps pulsed laser – ~ 0.5J/pulse @ 100-350MHz repetition rate– beam waist as small as possible: few tens of m

• Solution: Fabry-Perot resonator in pulsed regime [cw laser off by oders of magnitudes…]

4

Fabry-Perot cavity: Principle (CBAF & HERA cavities,cw laser)

When Laser =0 c/2L resonance

e beam

L

•But: /Laser = 10-11 for Gain=104 laser/cavity feedback

•Done by changing the laser frequency

Gain 10000

5

Existing FP cavities in HEP (Compton experiments)

• Continuous laser beam – CEBAF/polarimeter - gain ≈104

• Falleto et al. (NIMA459(2001)412)

– HERA/polarimeter - gain ≈104

• Pulsed laser beam– 25ps pulses & gain ≈ 3000

• Loewen (Compton X-ray source, Slac-R-632)

– 7ps @350MHz (laser wire)• R&D in progress (Compton X-ray source & e+

polarised source)– Hiroki Sato and Kazuyuki Sakaue 

(POSIPOL06: http://posipol2006.web.cern.ch/Posipol2006/Links.html)

6

HERALaser cavity

Cw Nd:YaG laser Cavity gain 104

In operation even after 3 yearswith strong radiations conditions

7

1ps Pulsed laser

Fabry-Perot cavitywith Super mirrors

Electron beam

•A priori impossible because of the laser frequency width: ≈1/(1ps)=1THz for picosecond laser (c.f. 3kHz cavity banwidth for a gain of 104)•In fact possible with mode-locked lasers

Fabry-Perot cavity filled with a pulsed laser

8

t

t=1ps

Fourier transform →superposition of N longitudinal laser mode – in phase

~1THz=1/(1ps)

T=1/frep

Mode-locked laser

If F.P. cavity length = laser cavity length → all modes are also resonant modes of the FP cavity

= frequency combfrep

9

f0=frep

Jitter f0 ≈ 1 MHz → [f0 or ] & frep must be controlledeven for 1ps pulses if the cavity finesse is very high Feedback bandwidth : subHz-40kHz

[Femtosecond optical frequency comb technology, Ye&Cundiff, Springer 2005]

Feedback for mode-locked laser beam

10

Eurotev R&D at LAL/Orsay:pulsed cavity for a polarimeter

[1ps, 100J/pulse@76MHz]

• Locking of a Ti:sa laser (MIRA-Coherent pumped by a 6W VERDI) to a high finesse cavity (=2 spherical mirrors): – Feedback difficult & never done for 1ps pulses +

very high finesse cavity (F=30,0000-300,000)

• Schedule – STEP 1: Gain=104→105

• Start: Sept. 20052007

– STEP 2: Reduction of the beam waist • Start (thanks to LAL/IN2P3 PhD. & postdoc):

Sept. 20062007(2008)

11

3.5 m

Status: optics Optics in mounting process•Acquisition of a large class 10 air flow•Up to now main activities: laser noise measurement

•Necessary for the feedback

12

Laser amplitude noise measurement

Just received (LAL financial support)

13

Use of 23 GHz bandwidth photodiode+spectrum analyser

Laser phase noise measurement (of the periodic pulsed)

Quartz oscillator ordered

14

Feedback status : electronics

8 ADC:14 bits 105Msps

8 DAC : 14bits 125 Msps

fpga latency=60ns

Tests of hardware & programming tools : almost finished

Ready to be inserted in the opticalbench

15

STEP 1 : 2m long stable cavity (confocal)→ optomécanics (LAL): GIMBAL miroir mount in vacuum

Status: end of cleaning processInstallation: end of may

Mechanics: status

16

optical axis

c

c1mm

10 mm

10 rad

10 rad

R = 1m

laser

R ≈ > L/2 ±1/e laser enveloppe

R ≈ >L/2

0ω →0 when

R →L/2

L

STEP 2: Reduction of the laser beam waist with 2 mirrors

Concentric cavity

BUT for: 0≈200m

High qulaity coatingsare good at the centre …

17

X

Y YX

Z

1

2

1

2

plane mirror

spherical mirror

plane mirror

spherical mirror

=0 2D cavity.0 3D cavity reduction of astigmatism & light polarisation changes

V1

V2laser

00 whenRL 2D Ring cavity:

Tolerance: spot size shift of 1mm for100 rad, 100mL

RR

•2m long concentric cavity: IF 0=50m, =800nmspot size shift of 30mm on the mirrors !!!for an axial and angular mirror misalignment of 1m and 1rad.

Mechanical constraints very strong …A mechanical solution: Four mirrors cavity

BUT problems due to:•Non zero incident angles

•Astigmatism (spherical mirrors)•Polarisation (Fresnel coeffs.)

18

Numerical description: done 1rst experimental setup: low finesse with a cw laser

•beam width•mode shape•resonance frequencies

Start mechanical setup in autumn 2006

High finesse cavity (inside vacuum) with pulsed laser in 2007

19

Summary• R&D EUROTEV at LAL/Orsay : 2005-2007

– Fabry-Perot cavity for a polarimeter1. 1ps/100fs pulses of energy 100 J/pulse@76MHz

– Moderate input Ti:sa laser beam power but very high cavity finesse 30000-300000

– Development of a numeric feedback system The experiment is being mounted2. Ring cavity to reduce the laser beam size 1rst exp. studies will start in september 2006

(thanks to an extra support from LAL/IN2P3)

• If this R&D gives satisfaction– After 2007 a new laser source (e.g doped fiber

amplifiers) should be considered to match the laser power required for an e+ polarised source study (100mJ/pulse@~300MHz…)

• Collaboration started (april 2006) with KEK (ATF2)

20

Maximum Cavity Gain achievable in pulsed regime: •limited by the dispersion (=pulse time width broadening) & chromatic dependence of the reflection coefficient of the cavity mirror coatings

No effect for a pulse width of 1ps: gain up to 105 can - a priori - be envisaged

Multilayer coating model