On the Gamma-ray Efficiency of Superluminous Supernovae: Potential Detections and Population-Level Constraints

Abstract

Superluminous supernovae (SLSNe) are among the most energetic stellar explosions, yet their central power source remains uncertain. Models invoking magnetar spin-down or circumstellar interaction predict GeV gamma-ray emission once the ejecta becomes transparent to high-energy photons. We search for such emission from 223 hydrogen-poor SLSNe using 17 years of Fermi-LAT data, defining source-specific search windows based on the Bethe--Heitler transparency time. We find no significant ($\geq5\sigma$) GeV emission. A joint-likelihood analysis constrains the GeV-to-optical efficiency to $\eta < 1.3\times10^{-3}$, two orders of magnitude below the predictions for weakly magnetized magnetar nebulae. A hierarchical population analysis shows that fewer than $0.7\%$ of SLSNe-I can have $\eta > 10^{-2}$. SN 2017egm, however, shows a suggestive excess ($\sim$4 $\sigma$). In the 0.1--500 GeV band, the observed $L_\gamma/L_{\rm opt} \sim 0.68$ for SN 2017egm exceeds hadronic expectations by over an order of magnitude, favoring a magnetar origin. The non-detection of the similarly nearby SN 2018bsz disfavors simple uniform-efficiency scenarios, or potentially points to diversity in the underlying powering mechanisms. We also note a possible excess from SN 2024jlc, though continued Fermi-LAT monitoring is needed because the source may still be within its transparency window.