Accueil du site > Thèmes de recherche > Magnéto-optique avancée > Biréfringence magnétique du vide > Photorégénération de bosons de faibles masse > 2006-2008 : Photoregeneration experiment at LULI


2006-2008 : Photoregeneration experiment at LULI

This experiment, complementary to the measurement of vacuum magnetic birefringence, begun in 2006 in response of results of the italian collaboration PVLAS [1] also working on the measurement of vacuum magnetic birefringence with an optical cavity. The most plausible interpretation of their ellipticity and dichroïsm measurements of vacuum inside magnetic field was the existence of a new bosonic, neutral, spinless and low mass particle : the axion. However PVLAS results were seriously inconsistent with searches of solar axions. Thus an independent and complementary experiment was fundamental to confirm or not their results.

1. Photoregeneration of light bosonic particles

1.1. Principle

Contrary to the BMV project which would only show indirectly the axion existence, this new experiment would allow to detect it directly. This particle can couple to two photons via the Primakoff effect. So, it should be possible to create an axion with a photon going through a transverse magnetic field.

The principle of the experiment is then very simple. It is shown in Figure 1. A laser beam travels through a coil giving a transverse magnetic field. Some photons are converted into axions with a probability P. A wall is placed after the coil in order to stop every photons, while axions cross since they hardly interact with ordinary matter. A second coil, identical to the first one, placed behind the wall converts back axions into photons. Those photons have the same energy as the initial ones and they can be detected with a suitable detector.

Fig. 1

1.2. Key elements of the experiment

The key elements of the experiment are :

- A powerfull laser : We need a number of initial photons as high as possible, in order to have the number of regenerated photons as high as possible. So we want a powerfull source, at a wavelength that can be efficiently detected. The experiment has been set up at LULI (Laboratoire pour l’Utilisation des Lasers Intenses) in Palaiseau, on the Nano 2000 chain. It can deliver more than 1.5 kJ over a few nanoseconds at 1,05 nm. This corresponds to about 8\times 10^{21} photons per pulse.
- An efficient detector : Our detector integrates a InGaAs Avalanche Photodiode. It is a commercially available songle photon receiver from Princeton Lightwave Instruments. We get a detection efficiency of almost 50 % with a reasonable dark count rate.
- An intense magnetic field : Since the conversion probability is proportional to the square of the magnetic field, we need a field as high as possible. The most suitable technology are pulsed coils. Coils have been developed at LNCMI following the same design as those developed for the BMV project.

2. Results

The experiment has been set up and tested from February to May 2007. We began laser pulses in May 2007. data acquisition was spread over 4 different weeks (in July 2007, September 2007 and January 2008).

During the 80 high energy pulses, no signal has been detected. Our results definitively invalidate the axion interpretation of the original PVLAS optical measurements (at least one regenerated photon per ulse should have been detected). These results have been published in :
- Physical Review Letters (C. Robilliard et al., Phys. Rev. Lett. 99, 190403 (2007)) for the first results, getting ahead of the other groups like the Fermilab and the Jefferson Lab in the US, CERN in Switzerland or DESY in Germany.
- Physical Review D (M. Fouché et al., Phys. Rev. D 78, 032013 (2008)) for the final results.

In Figure 2 are represented our limits of the inverse constant M for the coupling of axion to 2 photons as a function of the axion mass m_a (dark gray region). We also compare our results to other laboratory experiments.

Fig. 2

3. Conclusion

By combining the laser Nano 200 chain of LULI (Palaiseau), pulsed coils developed at LNCMI and a commercial single photon detector, we were able to be the first team to invalidate the italian results.

However, Figure 2 clearly shows that limits on the coupling constant given by purely experimental experiments (ours, BFRT, GammeV, PVLAS 2008 are several orders of magnitude lower than limits given by searches on solar axions (CAST) or galactic axions (ADMX) or than what is expected theoretically. Improving the sensitivity of our apparatus in order to test the axion model predictions seems rather unrealistic. In that respect, the vacuum magnetic birefringence experiment seems more promising. It can improve by one or two orders of magnitude the precision of purely terrestrial axion searches.

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