Search for Axion-Like Particles from Nearby Pre-Supernova Stars

TL;DR

The paper conducts a multi-source search for axion-like particles (ALPs) produced in the cores of 18 nearby pre-supernova stars and converted to photons in the Galactic magnetic field, using 22 years of INTEGRAL/SPI data in the $20$–$2000$ keV range. A Bayesian hierarchical model jointly constrains ALP mass $m_a$, ALP-photon coupling $g_{a\gamma}$, and ALP-electron coupling $g_{ae}$ by leveraging a shared spectral template across all stars and accounting for uncertainties in stellar parameters and distances. No ALP signal is detected; the study sets the strongest to-date bounds on the product $g_{a\gamma} g_{ae}$ and on $g_{a\gamma}$ for $m_a \lesssim 10^{-11}$ eV, with conservative scenarios broadening the allowed parameter space. The results underscore the importance of soft γ-ray observations and multi-source analyses for testing ALP scenarios and informing massive-star evolution, with implications for future γ-ray missions and stellar modelling.

Abstract

Axion-like particles (ALPs) are hypothetical pseudoscalar bosons that arise in many extensions of the Standard Model and are well-motivated dark matter candidates. Nearby massive stars in the late stages of stellar evolution provide a promising environment for enhanced ALP production due to their high core temperatures and densities. We search for a combined signal of ALP-induced hard X-ray and soft $γ$-ray emission from 18 nearby pre-supernova stars using the full public 22-year INTEGRAL/SPI dataset, construct individual stellar spectra and link them in a coherent analysis. A maximum-likelihood approach is used to extract fluxes in the 20--2000 keV energy range. Stellar evolution models are employed to obtain the expected spectral shapes of ALP production processes peaking between 50--500 keV, depending on stellar mass and evolutionary stage. We construct a joint likelihood that incorporates uncertainties in stellar parameters to derive combined constraints on the coupling constants $g_{aγ}$ and $g_{ae}$ as a function of the ALP mass $m_a$. The hard X-ray and soft $γ$-ray fluxes of all selected stars are consistent with zero within uncertainties. We provide upper limits on the continuum emission and on the 511 keV and 1809 keV line fluxes. The combined upper limit on $g_{aγ} \times g_{ae}$ is $(0.008 - 2) x 10^{-24}$ GeV$^{-1}$ (95% C.I.) while the ALP-photon coupling is constrained to $g_{aγ} = (0.13 - 1.26) x 10^{-11}$ GeV$^{-1}$ (95% C.I.) for $m_a\leqq10^{-11}$ eV, depending on the time to core collapse and magnetic field assumptions. Conservative limits of $(0.27 - 1.25) x 10^{-24}$ GeV$^{-1}$ (95% C.I.) are obtained assuming all but one star are in the early He-burning phase. These results rank among the strongest limits on ALP couplings to date and demonstrate the importance of soft $γ$-ray observations for probing ALPs and massive star evolution.

Figures (11)

• Figure 1: Feynman diagrams for ALPs production: Primakoff, Compton, and bremsstrahlung.
• Figure 2: The orange stars show the location of all the detected sources with SPI so far bouchet2008integral. The circular dots are the red super giants used in this work. The colored boundaries are the exposure regions for each dataset.
• Figure 3: Expected ALP fluxes from bremsstrahlung (black), Compton (red), Primakoff (blue), and total (green) using $g_{ae} = 10^{-13}$, $g_{a\gamma} = 10^{-11}~\mathrm{GeV}^{-1}$, and a time to core collapse of $\mathrm{t_{cc}} = 6900~\mathrm{yr}$. The blue dashed line shows the energy upper limit of NuSTAR ($79$ keV).
• Figure 4: Betelgeuse spectrum as obtained from SPI for the energy range $20$ -- $2000$ keV. The dot with the error bar shows the flux value in that energy bin. The downward arrows show the $3\sigma$ upper limit for bins where the flux significance is less than $2\sigma$. The 511 keV bin is systematically large because of incomplete modelling of the diffuse emission in the Crab/Orion region and is therefore not taken as a detection of 511 keV in Betelgeuse. The red shaded region shows the flux from the $g_{a\gamma} \times g_{ae}$ values allowed by NuSTAR that can now directly be excluded from the ALP parameter space since the flux prediction from them is larger than the $3\-\sigma$ flux limits from SPI. The excluded limit is $g_{a\gamma} \times g_{ae} \geqq 3\times 10^{-24}~\mathrm{GeV}^{-1}$. The blue shaded region shows the flux from the $g_{a\gamma} \times g_{ae}$ values that were already disallowed in the NuSTAR study xiao2022betelgeuse.
• Figure 5: Bayesian hierarchical model used to constrain the ALP parameters $m_a$, $g_{a\gamma}$, $g_{ae}$. The model assumes that these global parameters are shared across all 18 sources and govern both ALP production in stellar interiors and their conversion to photons in the Galactic magnetic field. Each star contributes a predicted photon flux based on its luminosity, distance, and a shared ALP spectral shape. This flux is convolved with the instrument response to yield the expected counts which are compared to the observed data. A joint likelihood analysis is performed across all sources using the 3ML framework to obtain constraints on the ALP parameter space.