📝 SummarXiv

Abstract

Ultra-long gamma-ray bursts (ULGRBs), a rare class of high-energy transients with durations $>10^3$,s, remain poorly understood. GRB,250702B ($z=1.04$) is notable for its multi-hour prompt emission, an X-ray precursor $\sim$1 day earlier, off-nuclear host position, and hard, rapidly variable gamma-rays. This combination is difficult to explain with standard ULGRB progenitors such as blue-supergiant collapsars, magnetar engines, or white-dwarf tidal disruptions by intermediate-mass black holes. We interpret the event as a micro-tidal disruption event ($\mu$TDE), where a stellar-mass black hole or neutron star partially or fully disrupts a main-sequence star. Three $\mu$TDE pathways can reproduce the observed precursor-main flare delay: (i) a \emph{dynamical (partial/repeating)} disruption, in which a grazing passage yields a faint precursor and the core returns after $\sim$day for a deeper encounter; (ii) a \emph{natal-kick} disruption, where the delay reflects the ballistic motion of a newborn compact object relative to its companion, leading to full disruption; and (iii) a \emph{hybrid natal-kick + partial} case, in which the kick seeds the close encounter but the first passage is only partial, with the core returning on the day-scale period. Cross-section scalings imply comparable rates for partial and full outcomes in both dynamical and natal-kick scenarios. The highly variable, hard $\gamma$-ray emission supports association with a stellar-mass compact object. Fallback and viscous accretion naturally explain the ultra-long duration, energetics, and ks-scale X-ray variability. We outline observational discriminants between the three channels and argue that $\mu$TDEs offer a compelling framework for ULGRBs such as GRB 250702B.