Home page > Research topics > Advanced magneto-optics > Vacuum magnetic birefringence > Vacuum Magnetic Birefringence (BMV) > Vacuum Magnetic Birefringence


Vacuum Magnetic Birefringence


- January 2015 : Second generation setup under construction.

- 2014 : ANR financial support obtained.

-October 2013 : New measurements in vacuum : kCM =(5.1 +/- 6.2)x10-21 T-2 at 3 sigma. Our value is one of the most precise value ever realized as shown in the followin figure [1], [2] [3]. It definitely validates our experimental method based on pulsed fields. We reach a noise floor lower than the one of PVLAS collaboration in 2012 obtained with an integration time of 8192 s [3].

For more details, see our article published in EPJD :

Vacuum magnetic linear birefringence using pulsed fields : the BMV experiment, A. Cadène, P. Berceau, M. Fouché, R. Battesti, and C. Rizzo, Eur. Phys. J. D 68, 16 (2014)

- July 2013 : Measurements of Cotton-Mouton and Faraday effects in helium gas. Mesures de l’effet Cotton-Mouton et de l’effet Faraday dans l’hélium gazeux. For more details, see our article on arXiv :

Faraday and Cotton-Mouton Effects of Helium at λ= 1064 nm, A. Cadène, D. Sordes, P. Berceau, M. Fouché, R. Battesti, and C. Rizzo, Phys. Rev. A 88, 043815 (2013)

- January 2012 : See the article recently published :

Magnetic linear birefringence measurements using pulsed fields, P. Berceau, M. Fouché, R. Battesti, and C. Rizzo, Phys. Rev. A 85, 013837 (2012)

- July 2011 : improvement of our sensitivity with pulses applied inside vacuum giving : \Delta n_{\mathrm{vacuum}} = (7 \pm 48)\times10^{-20} T^{-2}

- July 2011 : First observation of the ellipticity on the whole applied magnetic pulse.

Nitrogen pressure = 32.1 matm.

As expected, the measured ellipticity (red) follows the square of the magnetic field filtered by the cavity (blue) .

Corresponding measured value : \Delta n_u = 2.1x10^{-15} T^{-2}atm^{-1}

- November 2010 : The highest finesse obtained up to now with LMA mirrors is 529 000 corresponding to a cavity linewidth of 124 Hz. Our cavity is thus the sharpest cavity in the world.

- September 2009 : A new coil has been tested in September 2009. This coil has exploded at more than 30 Tesla which corresponds to more than 300 T²m. So our goal to obtain 25 Tesla is reached !

Our scientific program, the BMV project (Vacuum Magnetic Birefringence), is an ambitious experimental project whose goal is to check in laboratory what is predicted for vacuum energy in quantum electrodynamics. This project is based on intense pulsed magnetic fields and a sensitive optical apparatus for the detection of effects induced by this field on a laser beam.

A vidéo is available : "Magnetism and quantum vacuum".

1. Goals and context

Classical electrodynamics, modified during the beginning of the XXth century to take into account quantum mechanics, gave rise to quantum electrodynamics (QED). Among the new phenomenon predicted by QED but never observed, we find the magnetic birefringence of vacuum, also called the Cotton-Mouton effect of vacuum.

Fig. 1

In vacuum, the velocity of light propagating in the presence of a transverse magnetic field depends on the polarization of light. The index of refraction n_{par} for light polarized parallel to the magnetic field is different from the index of refraction n_{perp} for the light polarized perpendicular to the magnetic field. The difference \Delta n = n_{par} - n_{perp} is proportional to B^2. Thus an incident linearly polarized light beam exits elliptically polarized from the magnetic field region (Fig. 1) [4]. Euler et Kochel in 1935 and then Heisenberg et Euler in 1936 [5] established the effective Lagrangian and calculated that a fiel of 1 T should induce an anisotropy of the index of refraction of about \Delta n = 4\times 10^{-24}. That’s a real challenge ! Observation of such an effect would be the first experimental evidence of non linear propagation of light in quantum vacuum.

This fundamental prediction has not yet been experimentally confirmed. Nevertheless, the Italian collaboration PVLAS [6] of professor Zavattini has announced in 2006 the observation of an optical activity due to vacuum. This phenomenon can’t be explained by QED but it could be the first observation of one of the particle constitutive of dark matter. Our photoregeneration experiment has been built motivated by this publication. Finally our null measurement allowed us to definitely exclude the Italian interpretation.

2. Project

2.1 Magnetic field

For our experiment, we need a transverse magnetic field as high as possible. Thus the best choice is to use the pulsed magnets of which the LNCMI is specialist. Furthermore the region of interaction with the magnetic field should be long since the measured ellipticity is proportional to B^2 L, where L is the length over which the magnetic field is applied. The final goal is to realize pulsed magnets which could deliver a transverse magnetic such that B²L>600 T²m.

The developed magnet is based on a X geometry, called X-coil. This geometry allows to have a high transverse magnetic field while keeping the room for the necessary optical access at both ends in order to let the laser in. As for usual pulsed magnets, the coils are immersed in a liquid nitrogen cryostat to limit the consequences of heating during the magnetic pulse. Images of the cryostat containing the coil are shown in Figure 2.

Fig. 2

Fig. 2 Left : Cryostat. Up and right : Coil inside its cryostat. Down and right : Cryostat over the optical table in the clean room.

2.2 Optical apparatus

The second part of the project concerns the conception of a 2 m long Fabry-Perot cavity, the entire vacuum system and optics with more particularly the laser locking to the cavity. High finesse mirrors are made by the LMA of IN2P3 in Lyon and they should provide a finesse up to 1 000 000.

2.3 Experimental setup

The experiment is at LNCMI since may 2006. A clean room has been built (see images below). Its access is limited to fully equipped personnel in order to minimize pollution over the mirrors of the very high finesse cavity. The laser, the vacuum system, the coils and their cryostats are placed inside the clean room. The capacitor bank is outside. Supervision of the experiment is realized in the room to reach the clean room.

For more details : see our article published in European Physical Journal D (R. Battesti et al., Eur. Phys J. D 46, 323 (2008)).

3. Current status


To date our apparatus is inside the clean room and is composed of
- two coils which deliver magnetic field higher than 11 T over 50 cm with a pulse duration of a few milliseconds.
- a 2,2 m long Fabry-Pérot cavity with a finesse of 130 000 (mirrors manufactured by Layertec). A Nd:YAG (\lambda = 1064 nm) is locked to this cavity.


A new coil has been tested in September 2009. This coil has exploded at more than 30 Tesla which corresponds to more than 300 T²m. So our goal to obtain 25 Tesla is reached !

Fabry-Perot cavity

- The highest finesse obtained up to now with LMA mirrors is 529 000 corresponding to a cavity linewidth of 124 Hz. Our cavity is thus the sharpest single cavity in the world.

- Birefringence measurements of our mirrors manufactured by Laytertec have been realized. A study based on a review of existing data on interferential mirror birefringence together with our new measurements and the development of a computational code allow us to indicate that the origin of the mirror birefringence can be ascribed to the reflecting layers closed to the substrate. For more details, see our article published in Applied Physics B (F. Bielsa et al, Appl. Phys. B 97, 457 (2009)).

- Since measurements are performed using pulsed magnetic field, understanding of the cavity dynamical behaviour is required. We have studied this behaviour both experimentally and theoretically. We have shown that the cavity acts as a first-order low pass filter for the ordinary beam, but as a second-order for the extraordinary beam. These results are published in Applied Physics B (P. Berceau et al, Appl. Phys. B 100, 803 (2010)).


- First measurements have been performed during summer 2008. We currently measure the Cotton-Mouton effect in gases like nitrogen and helium. Results agree with expected values.

- About 100 magnetic pulses have been applied in vacuum at the end of 2012. Our procedure of data acquisition and our analysis take into account the symmetry properties of raw data with respect to the orientation of the magnetic field and the sign of the cavity birefringence. Our current value of vacuum magnetic linear birefringence k_\mathrm{CM} was obtained with a maximum field of 6.5 T. We get kCM =(5.1 +/- 6.2)x10-21 T-2 at 3\sigma confidence level. Our result is a clear validation of our innovative experimental method.

[1] R. Cameron et al, Phys. Rev. D 47, 3707 (1993).

[2] E. Zavattini et al, Phys. Rev. D 77, 032006 (2008).

[3] G. Zavattini et al, Int. J. of Mod. Phys. A 27, 1260017 (2012).

[4] G. L. J. A. Rikken and C. Rizzo, Phys. Rev. A 63, 012107 (2003).

[5] H. Euler and B. Kockel, Naturwiss. 23, 246 (1935) ; W. Heisenberg and H. Euler, Z. Phys. 98, 714 (1936).

[6] E. Zavattini et al., Phys. Rev. Lett., 96 (2006).