FabryPerot cavity for the Compton polarimeter Goal 10

Fabry-Perot cavity for the Compton polarimeter Goal: 10 -100 m. J/pulse @ 5 MHz repetition rate & small diameter ≈ 50 mm (c. f. P. Schuler’s talks)

Fabry-Perot cavity: Principle (HERA cavity, cw laser) Gain 10000 e beam L Polar. Circ. Polar. Lin. When n. Laser =n 0 c/2 L • But : Dn/n. Laser = 10 -11 resonance for Gain=104 laser/cavity feedback • Done by changing the laser frequency

Some of the advantages of using a FP cavity • Compact (& cheap) system compared to a laser of same power (500 W in average) • Laser power small outside the cavity: full power only at the electron-laser IP – no thermal effects producing parasitic birefringence & high quality frequency controlled beam accurate control of the laser beam polarisation

Proposal: Cavity filled with a pulsed laser for a Compton polarimeter at FLC ≈ 5 MHz / ≈10 n. J/pulse Electron beam Ti: sa oscillator 500 fs-1 ps Pulse laser Fabry-Perot cavity with Super mirrors • A priori impossible because the laser frequency width Dn ≈1/(1 ps)=1012 Hz for picosecond laser (c. f. 3 k. Hz cavity banwidth) • In fact possible with mode lock lasers Jones et al. Opt. Lett. 27 (2003) 1848, Jones at al. Phys. Rev. Lett. 15 (2001) 3288, Hood et al. Phys. Rev. A 64 (2004)033804, Potma et al. Opt. Lett. 28 (2003)1835

Mode lock laser Dt=1 ps ≈10 ns t Fourier transform →superposition of N longitudinal laser mode – in phase Dn~1012 Hz=1/(1 ps) n If F. P. cavity length = laser cavity length all modes are also resonant modes of the FP cavity Available laser pulse energy: 1 -10 n. J cavity Gain ≈104

• Pulse width limited by dispersion in the super-mirror coatings (Nb round trips=F/(2 p) ≈ 5000 for F=30000 Gain ≈10000): circulating pulse gets broader and broader power loss when overlapped to the incoming pulses (constructive interferences reduced) R. J. Jones et al. Opt. Lett. 27 (2003) 1848 Cavity gain Width : 300 fs-1 ps for gain=104

Reduction of the laser beam size at the IP • To get a 50 mm laser beam size at the electron-laser beam IP – Use of a quasi-concentric cavity (mirror curvature radius ≈ half cavity length) – BUT, mechanical tolerance mm & mrad needed on relative mirror positions – Active feedback on relative mirror position needed (c. f. LIGO & VIRGO where nm tolerances are reached)

Present status of FP cavities filled with fs pulses • Power amplification ≈ 120 and cavity Finesse ≈ 300 for pulse width 2 -3 ps (Potma et al. Opt. Lett. 28 (2003)1835 ) • Proposed R&D: – Reach a Finesse ≈ 30000 in a first step – And using a quasi-concentric FP cavity in a second step

Cavities in operation (for Compton polarimetry) • CEBAF (N. Falletto, NIM A 459(2001)412): F≈24000 • HERA (upstream the HERMES experiment): F≈30000 – Installation: 2003 summer – Laser & controllers dismounted after synch. rad. damages (huge, generated by 2 new dipoles in HERMES) – Presently: strong shielding and re-mounting – after 1 year of radiation, cavity finesse is still the same and locked again …

ellipsometer 4 motorised miroirs bellow Optique input ligne HERA CAVITY

2003 installation shielding (3 mm pb) HERA CAVITY

Conclusion • Proposal: a high finesse FP cavity filled with a pulse laser to produce 100 m. J/pulse @5 MHz – Will contribute to a high precision on the polarisation measurement • This proposition make sense if the polarisation is to be measured bunch by bunch – If not, commercial laser with low rep. rate & high pulse energy do exist – But, this R&D may also be useful for other applications related to FLC (e. g. polarised positrons)

• Laser/cavity feedback – similar to cw laser case (Jones et al. , Opt. Comm. 175(2000)409) • Stabilisation channels, e. g. MIRA (Coherent) Ti: sa oscillator – 3 channels: 2 PZT mounted 2 mirrors & output coupler mounted on translation stage • High frequency correction signal by an EOM if required • Phase velocity & group velocity must be matched to the cavity (both pulse-round-trip/pulse-repetition matching and frequency matching are required) – A priori not a problem for 0. 3 -1 ps pulse width but precise feedback techniques are known if needed

Aservissements