Experimental Limit on Optical-Photon Coupling to Light Neutral Scalar Bosons

We report on the first results of a sensitive search for scalar coupling of photons to a light neutral boson in the mass range of approximately 1.0 meV (milli-electron volts) and coupling strength greater than ${10}^{-6}{\mathrm{GeV}}^{-1}$ using optical photons. This was a photon regeneration experiment using the “light shining through a wall” technique in which laser light was passed through a strong magnetic field upstream of an optical beam dump; regenerated laser light was then searched for downstream of a second magnetic field region optically shielded from the former. Our results show no evidence for scalar coupling in this region of parameter space.

Figure 1

Photon regeneration using the “light shining through a wall” technique. The incident light ($\gamma$) couples to photons in the magnetic field ($\mathit{B}$) creating the hypothetical light neutral boson ($\mathit{A}$). Because it is weakly interacting, the LNB passes through the optical barrier (the “wall”) while no incident photons do so. Regenerated photons having the same characteristics as the original photons result from the second magnetic field region downstream of the wall as shown.

Figure 2

The LIPSS experimental setup. Laser light from the FEL is directed to the LIPSS beam line by Turning Mirror 1 (TM1) and a collimator. Turning mirror 2 (TM2) directs the incident light through the generation magnet (GV) and to an optical beam dump (the wall) as shown. A second, identical magnet (RV) is used to regenerate any photons that would result from a hypothetical particle (a LNB) that passes through the wall; no incident FEL light passes into RV. These regenerated photons would be detected by the detector system in the light tight box. Details of the light tight box are shown in the insert.

Figure 3

Data from the LIPSS runs in the signal region [(a), top] and a representative background region [(b), bottom] after proper normalization. The latter was used to determine the contribution from backgrounds in signal region. All pixels not in the signal region described in the text are used to determine the backgrounds. No excess of events above background is seen in the signal region.

Figure 4

The new LIPSS limits on scalar coupling, $g$, of photons to a hypothetical LNB (in inverse GeV) versus the LNB mass in meV. The curves show the LIPSS result for a significance of 2.2 (full) along with the results from previous measurements at different photon energies: GammeV-1 (using magnet lengths 5.0 and 1.0 m), GammeV-2 (using magnet lengths 3.1 m and 2.9 m), and BFRT, as indicated. The data point is the region claimed (now disclaimed) by the PVLAS collaboration [12, 13]. The LIPSS photon energy was 1.33 eV while those of GammeV and BFRT were 2.34 and 2.44 eV, respectively.