Simulating the late stages of WD-BH/NS mergers: an origin for fast X-ray transients and GRBs with periodic modulations

TL;DR

This study investigates the late-stage evolution of white dwarf–black hole/neutron star mergers when residual orbital eccentricity drives repeated tidal disruptions (RPDs). Using high-resolution SPH simulations with 16 WD-BH/NS setups across WD masses $0.8$–$1.3\,M_\odot$ and companion masses $2$–$10\,M_\odot$, the authors show that RPDs modulate accretion rates with the orbital period, producing peak rates from $4\times 10^{-4}$ to $0.2\,M_\odot\,\text{s}^{-1}$ and RPDs lasting $\sim 10$ s to $10^3$ s. They classify outcomes into three categories by mean accretion rate, linking more compact systems to shorter, more violent RPDs and stronger modulation. The work further translates these accretion histories into observational predictions for X-ray transients and gamma-ray bursts, including cases where GRBs (as in GRB 230307A and 211211A) are simultaneous with or delayed from the X-ray emission, and it discusses potential identifications with events like EP250702B. Overall, the results provide a physically motivated framework for WD-BH/NS mergers as progenitors of long-merger GRBs with quasi-periodic prompt signatures and distinct high-energy transients identifiable by upcoming wide-field X-ray surveys.

Abstract

Recent studies indicate that mergers of a white dwarf (WD) with a neutron star (NS) or a stellar-mass black hole (BH) may be a potential progenitor channel for certain merger-kind, but long-duration $γ$-ray bursts (GRBs), e.g., GRBs 230307A and 211211A. The relatively large tidal disruption radius of the WD can result in non-negligible residual orbital eccentricity ($0 \lesssim e \lesssim 0.2$), causing episodic mass transfer, i.e., repeated tidal disruptions (RPDs) of the WD. We perform smoothed-particle-hydrodynamics simulations of RPDs in sixteen WD-BH/NS systems, capturing the subsequent mass transfer and accretion. The WD undergoes RPDs near the orbital periastron, modulating the ensuing accretion process, leading to variations of the accretion rate on the orbital period. Across all simulations, the peak accretion rates range from $4 \times10^{-4}$ to 0.2 $M_{\odot} \rm \ s^{-1}$, while the RPD duration spans from $\sim$ 10 s to an hour. More compact systems, i.e., those with a higher mass ratio (higher WD mass and lower accretor mass), tend to undergo fewer RPD cycles, resulting in shorter durations and higher accretion rates. If such events can launch relativistic jets, three categories of non-thermal X/$γ$-ray transients are predicted, in decreasing order of their mean accretion rates: (1) an X-ray transient with a simultaneous GRB, both lasting for $10^{1-2}$ s; (2) a longer X-ray transient lasting up to $10^{2-3}$ s that has a GRB appearing only at its later phase ; (3) an ultra-long X-ray transient lasting for $\sim 10^{3}$ s without a GRB. A generic feature of these transients is that their prompt emission light curves are probably periodically modulated with periods of a few to tens of seconds.

Figures (8)

• Figure 1: Density evolution of the WD over one orbital period. We present density‐projection snapshots in the $x$–$y$ plane from a simulation using $10^{6}$ SPH particles to model the late merger phase of a $1\,M_{\odot}$ WD interacting with a $10\,M_{\odot}$ BH (for simulation runs No. 2a). Each panel covers a region of $0.14 \times 0.14\,R_{\odot}$.
• Figure 2: Repeating Partial Disruption Process of a WD. We present density‐projection snapshots in the $x$-$y$ plane from a simulation using $10^{6}$ SPH particles to model the late merger phase of a $1\,M_{\odot}$ WD interacting with a $10\,M_{\odot}$ BH (for simulation runs No. 2a). Each panel corresponds to distinct evolutionary times, with all panels spanning a physical scale of $0.2\times0.2\,R_{\odot}$. The main panel shows the WD after complete disruption (the end of the RPD is defined as $t=620$ s), where the black dashed circle marks the accretion radius $R_{\rm acc}= R_p/3$. A video showing the density evolution of this simulation can be found at https://youtube.com/shorts/o1G_qp9Fnpg?feature=share.
• Figure 3: Late-stage parameter evolution of a 1 $M_{\odot}$ WD merging with a 10 $M_{\odot}$ BH (for simulation runs No. 2a). Panels (a), (b), (c), (d), and (e) show the evolution of WD’s mass $M_*$, radius $R_*$, pericenter distance $R_p$, semi-major axis $a$, and penetration factor $\beta$, respectively.
• Figure 4: Evolution of the accretion rate during the late stage of the WD–BH merger for simulation runs No. 1a, 8a, and 10a. The red vertical dashed line indicates the termination of the repeated partial disruptions of the WD. Each panel corresponds to one of the three categories described in § \ref{['sec4_4']}.
• Figure 5: The relationship between the initial orbital eccentricity $e_0$ and the modulation amplitude $A = \dot{M}_{max}/\dot{M}_{min}$ obtained from simulation runs No. 0a, 2a, 6a, and 7a. The error bars represent the statistical standard errors of our measurements.