An improved limit on the axion-photon coupling from the CAST experiment

TL;DR

The paper presents a refined search for solar axions via the CERN Axion Solar Telescope (CAST) using improved detectors and a modern solar model to predict the solar axion flux from the Primakoff process. By searching for coherent axion-to-photon conversions in a 9.26 m, 9 T magnetic field and analyzing data from an X-ray telescope with CCD plus TPC and MM detectors, CAST sets a new upper limit on the axion-photon coupling of $g_{a\gamma} < 8.8\times10^{-11}\ \mathrm{GeV^{-1}}$ at 95% CL for $m_a \lesssim 0.02$ eV, improving on prior laboratory limits and, for this mass range, beating the HB-star energy-loss bound. The analysis carefully accounts for the solar flux, axion coherence, detector responses, and systematic uncertainties, with a combined result from phase I/III data confirming no significant axion signal. These results strengthen constraints on axion-like particles and set the stage for CAST phase II to probe higher masses and potentially QCD axion models using gas to maintain coherence over a broader $m_a$ range.

Abstract

We have searched for solar axions or similar particles that couple to two photons by using the CERN Axion Solar Telescope (CAST) setup with improved conditions in all detectors. From the absence of excess X-rays when the magnet was pointing to the Sun, we set an upper limit on the axion-photon coupling of 8.8 x 10^{-11} GeV^{-1} at 95% CL for m_a <~ 0.02 eV. This result is the best experimental limit over a broad range of axion masses and for m_a <~ 0.02 eV also supersedes the previous limit derived from energy-loss arguments on globular-cluster stars.

Figures (8)

• Figure 1: Comparison of the solar axion flux calculated from a modern solar model Bahcall:2004fg ($\full$) and an older solar model published in 1982 Bahcall:1981zh ($\broken)$. An axion-photon coupling of $1\times 10^{-10}$ GeV$^{-1}$ is assumed.
• Figure 2: Left: Solar axion surface luminosity depending on energy and the radius $r$ on the solar disk. The flux is given in units of $\text{axions}\,\text{cm}^{-2}\,\text{s}^{-1}\,\text{keV}^{-1}$ per unit surface area on the solar disk. Also shown is the radial distribution of the axion energy loss rate of the Sun (${\rm d}L_{\text{a}}/{\rm d}R$) as well as the energy distribution of the solar axion flux (${\rm d}\Phi_{\text{a}}/{\rm d}E$). Right: Differential solar axion spectrum, derived by integrating the model shown on the left up to different values of $r$ in units of the solar radius $R_\odot$. The peak of the spectrum moves towards lower energies if integration radius moves towards the outer rim of the solar disk.
• Figure 3: Expected "axion" image of the Sun as it would be observed by the CCD detector assuming the axion surface luminosity shown in figure \ref{['fig:axionspec-radius-dependance']} and assuming zero detector background. To determine the axion spot on the CCD, the PSF of the mirror system and the total effective area of the X-ray telescope were taken into account. The count rate integrated over the region of the spot is normalized to unity.
• Figure 4: Left: Signal-to-noise ratio for the X-ray telescope depending on the radius of the signal-spot. The axion-photon couplings vary from $g_{\text{a}\gamma}=2\times 10^{-10}$ GeV$^{-1}$ for the lowest curve, through $g_{\text{a}\gamma}=5, 9$ to $16\times 10^{-10}$ GeV$^{-1}$ for the curve with the largest peak value. Right: The axion flux expected in the signal-spot area on the CCD relative to the axion flux for an spot radius of the size of the solar disk (encircled flux). Two cases are shown: the encircled flux for a perfect linear optics (red line) and the encircled flux for a realistic X-ray optics taking into account the point spread function of the CAST X-ray mirror system (black line).
• Figure 5: Left: Spatial distribution of events observed under axion sensitive conditions by the CAST X-ray telescope during the 2004 data taking period. The intensity is given in counts per pixel and is integrated over the tracking period of 197 h. Right: Background spatial distribution as observed by the CAST X-ray telescope during the 2004 data taking period. The intensity is given in counts per pixel and integrated over the full observation period of $1890\,\text{h}$.