1

Progress in ultra-fine birefringence measurements

Optical Search for QEDvacuum magnetic birefringenceAxion and photon Regeneration

the 94th Meeting of the SPSC – CERN – 2009-November 24

STATUS REPORT

Mathieu Durand(1), Jérôme Morville(1), Daniélé Romanini(2)

(1 )Laboratoire de Spectrométrie Ionique et Moléculaire (LASIM)UMR CNRS 5579, Université Claude Bernard-Lyon1

(2 )Laboratoire de Spectrométrie Physique (LSP)UMR CNRS 5588, Université Joseph Fourier - Grenoble

NEW

collaborator

CERN, Geneva, Switzerland P. Pugnat (now at LNCMI-CNRS), M. Schott, A. Siemko

Charles University, Faculty of Mathematics & Physics, Prague, Czech Republic M. Finger Jr., M. Finger

Czech Technical University, Faculty of Mechanical Engineering, Prague, Czech Republic J. Hošek, M. Kràl, J. Zicha, M. Virius

ISI, ASCR, Brno, Czech Republic A. Srnka

IMEP/LAHC - INPG, 38016 Grenoble Cédex-1, France L. Duvillaret, G. Vitrant, J.M. Duchamp

IN, CNRS – UJF & INPG, BP 166, 38042 Grenoble Cédex-9, France B. Barbara, R. Ballou, Y. Souche

LASIM , UCB Lyon1 & CNRS, 69622 Villeurbannes, FranceM. Durand, J. Morville

LSP, UJF & CNRS, 38402 Saint-Martin d'Hères, France R. Jost, S. Kassi, D. Romanini

TUL, Czech Republic M. Šulc

Warsaw University, Physics Department, Poland A. Hryczuk, K. A. Meissner

The OSQAR Experiments at CERN to probe QD & Astroparticle Physics from the Photon Interaction with a Magnetic Field

2/45

The OSQAR Collaboration at present ► 24 Members from 10 Institutes (Cz, Fr, Po & CERN)

Ln//

Ln

LΔnλ

2π

LaserPP

Laser

2

P2

P

Matter anisotropy induces different answer with the electric field vector orientation

Entering linearly polarized Light exits elliptically polarized

Malus law for small dephasing

What is birefringence ?

3

4

The vacuum embedded in a magnetic field should also be anisotropic

20

24 B104n

22106,3n

Is predicted @ l=500 nm with the CERN dipole magnets

If T5,9B0

rad107,0 13

m15L

giving

Using High Finesse Optical Cavity

effLn2

LaserPP

R+T+Loss=1(Loss~few ppm)

R~99,999%

The medium is contained between two high reflective and low losses dielectric mirrors

outP

L

If light resonates inside the cavity (constructive interference at each round trip),Light transmitted trough the exit mirror has an effective optical pathlength of

LF2

Leff

510R1

F

with the finesse 5

Polarizer Crossed Polarizer

Voltage

Gas

Recent Kerr anisotropy measurementsPhD thesis : Mathieu Durand (07/23/2009)

At this level of sensitivity, residual mirrors birefringence entirely conceals the effect of gas

birefringence

6

Dn=K.E2

PolarizerCrossed polarizer

Modulated Voltage W

7

xqin

qout

yz qM

x

Using residual mirrors birefringence as an optical bias for homodyne detection

fin

fout

fM(fin,fout,qin,qout)

8

DC W component negligible high order terms

The lowest gas dephasing measurable : g

Sensitivity given by :

gσ

π

2F

I

Inoise M

2

Ω//

Gain induced by the optical bias : 410

g

MG

Experimental set-up with N2 as a gas test

l/4PolarizerModulated voltage W

DiodeLASER

IT

Lock’in detection at W

9

E=0 V/mm

E=6,4 V/mm

[V] IT

I┴

[mV]

I┴ WTime (sec)

trd=147 ms F =277 000

(L=50 cm)

Time (µs)

[10-12 rad]sfgDfg

0 V/mm

2,7 V/mm

6,4 V/mm1,3 V/mm

Measured sensitivity with N2 as a gas test

10

sfg = 4.10 -12 rad/Hz1/2

Allan variance plot :

9,3 V/mm

Measured sensitivity with N2 as a gas test as a function of integration time

11

4

0,3

sfg = 3.10 -13 rad

Measurement of Kerr constants

KN2 = 1,438.10-25 ± 0,006 m2/V2

-Kerr constant K:

12

L = 50 cml = 810 nm

Molecular gases alignmentAtomic gases distortion of electronic wavefunction

13

Kerr constants of several gases

20% 80%1,64 ± 0,04

·

CO2

O2

N2

Ar

Ne

He

14

(199

7)

(198

5)

(196

7)

(200

0)

(200

5)

(200

8)

Comparison with published results

Thi

s wor

k

15

Achievement

a table top optical system for Kerr birefringence measurement in gas phase is demonstrated

sfg = 4.10 -12 rad/Hz1/2

Only the PVLAS system reaches comparable performance, based on noise analysis (4-times better, in development since 1994)

sfg = 3.10 -13 rad By averaging over 800 s, limited by the laser-cavity servo-lock system

Our result constitutes the present state-of-the-art validated on a physical signal.

a factor two above the shot noise level with 3 mW incident laser power

outlooks

Recall : rad107,0 13

Accommodation to the OSQAR experiment

With B0=9,5T and L=15m

Measuring in transmission (instead of reflection) :g 5

g

Detection at the shot noise levelg 2

g

An increase of the laser beam transverse dimension (a factor 40 with a confocal cavity of 20 m length) increases the maximal incident power and thus the shot noise level

g

All taken into accountg rad101,160 152

g

6g

Integration time of 1 hour :g 3600

g

16

17

outlooks

Magnetic field and homodyne detection scheme

xqin

qout

y

fin

fout

B

All the setup rotates

Two possibilities

8 Hz demonstrated with the new version

Only mirrors rotates around their optical axis

maintaining high finesse properties

1818

nn nn+1nn-1

n

Spectral transmission

Constructive interferences => frequency comb nresonance

%10PP Laserout

<Dn 1kHzDn=c/2L

The finesse is defined as :

5

roundtrip

ringdown 10R1

2F

LF2

Leff

Cavity lifetime

effLn2

LaserPP

R+T+Loss=1(Loss~few ppm)

R~99,999%

outP

L

The cavity more in detailed

19

Optimum : fM ~ 0,1·p/F

19

Optical detection

Laser-cavity locking residual

instability

Mirror birefringence as optical bias and noise analysis