ALP couplings to muons and electrons: a comprehensive analysis of supernova bounds

TL;DR

This paper investigates axion-like particles (ALPs) that couple to electrons or muons and are produced in core-collapse supernovae. It develops a complete leptophilic EFT, derives a loop-induced ALP–photon coupling, and computes six production channels in realistic SN cores, including plasma effects and massive-ALP kinematics. Using three independent ccSN models, it derives bounds from SN 1987A cooling, gamma-ray searches, explosion energy, the diffuse gamma-ray background, and the 511 keV line, while quantifying systematic uncertainties across models. The results exclude ALP couplings over seven orders of magnitude up to masses near 500 MeV, and demonstrate that SN modeling uncertainties can shift bounds by factors of a few to an order of magnitude. The study strengthens the role of ccSNe as probes of beyond-Standard-Model physics and outlines avenues for extending the framework to additional couplings and heavier leptons.

Abstract

We provide a comprehensive analysis of the phenomenology of axion-like particles (ALPs) produced in core-collapse supernovae (ccSNe) through interactions with electrons and muons, both of which have a non-negligible abundance in the SN plasma. We identify and calculate six significant ALP-production channels, two of which are loop-level processes involving photons. We then examine several observational constraints on the ALP-electron and ALP-muon parameter spaces. Those include the bounds on anomalous cooling, energy deposition, decay into photons, diffuse gamma rays, and the 511 keV line. Our results provide updated and robust constraints on ALP couplings to electrons and muons from an improved treatment of production and absorption processes. Furthermore, we quantify the uncertainties of the results by using three state-of-the-art supernova models based on two independent simulation codes, finding that constraints vary by factors of O(2-10).

Figures (4)

• Figure 1: Snapshots at $t=1$ s post bounce of plasma properties for the three SN models; from left to right and top to bottom: the temperature, the mass density, the electron chemical potential, the muon chemical potential, the plasma frequency, and the gravitational lapse factor. The solid blue lines correspond to the DD2-18.0 model, the dashed orange lines to the SFHo-18.8 model, and the dot-dashed green lines to the LS220-20.0 model (see main text for details).
• Figure 2: Processes contributing to ALP production and reabsorption in the SN plasma. From top left to bottom right: Compton scattering (\ref{['subsec:compton']}), lepton-pair annihilation (\ref{['subsec:annihilation']}), Bremsstrahlung (\ref{['subsec:bremsstrahlung']}), electron-positron fusion (\ref{['subsec:fusion']}), Primakoff process (\ref{['subsec:primakoff']}), and photon coalescence (\ref{['subsec:coalescence']}). The blobs in the latter two diagrams indicate the effective, loop-induced ALP-photon coupling. Note that crossed diagrams are not shown.
• Figure 3: Comparison between the different contributions to ALP production in the SN models in the case of an ALP-electron coupling of $\hat{g}_{ae} = 10^{-10} \, \mathrm{MeV}^{-1}$ (left panel), and ALP-muon coupling of $\hat{g}_{a\mu} = 10^{-10} \, \mathrm{MeV}^{-1}$ (right panel). The plots show the total number of produced ALPs, $N_a$, ignoring reabsorption and decay (see main text) as function of the ALP mass $m_a$. The transparent band shows the range of predictions from the three different SN models.
• Figure 4: Bounds on the coupling of ALPs to electrons (upper panel) and muons (lower panel), respectively. We show the bounds from the three SN models employed here (see \ref{['sec:SNmodel']}) with different levels of opacity. Mostly, the largest region of each color corresponds to LS220-20.0, the smallest to SFHo-18.8, and the one in the middle to DD2-18.0. Details on each bound and how they are influenced by the different SN models can be found in \ref{['subsec:cooling', 'subsec:decayBound', 'subsec:calorimetricBound', 'subsec:positronBound', 'subsec:diffuseGammaRayBound']}. In shades of gray, we show other constraints that are relevant in this part of the ALP parameter space, as described in the main text.