Research highlights
MASS TRANSFER IN TIDALLY HEATED STARS ORBITING SMBHs
Using analytic models coupled with MESA stellar-evolution calculations, we examine how intense tidal heating of stars on tight orbits around supermassive black holes drives their structural evolution, gravitational-wave– and tide-induced orbital decay, and Roche-lobe–overflow mass transfer. We find that, unlike comparable-mass binaries, the mass-transfer rate is regulated by the tidal-heating timescale—yielding stable rates of ∼10⁻³–10⁻⁵ Msun yr⁻¹—and that the transfer’s stability depends sensitively on where heating occurs in the stellar interior, with implications for low-luminosity AGN populations and the progenitors of partial TDEs and short-period QPEs.
Star-Disk Collisions & Implications for QPEs
We used Athena++ to perform 3D hydrodynamic simulations of repeated collisions between a star and an accretion disk near a supermassive black hole, quantifying the stellar mass loss per impact with simple fitting functions and showing that shock-heating from earlier encounters dramatically amplifies subsequent stripping. From these results, we demonstrate that observed quasi-periodic eruptions (QPEs) are unlikely powered by direct star–disk impacts but rather by “circularization shocks” as previously stripped debris re-encounters the disk, which naturally explains the regular timing of GSN 069 and eRO-QPE2, the large flare-to-flare variations in eRO-QPE1, and constrains the QPE-emitting phase lifetime to ≲1000 yr (with eRO-QPE2 lasting only a few decades).
Planetary Nebula Luminosity Function
We used the photoionization code CLOUDY to simulate planetary nebulae over a wide range of central-star and nebular parameters, finding that the peak [O III] 5007 Å luminosity depends only weakly on stellar temperature and ejecta mass but is strongly governed by the central stellar luminosity and the nebular dust-to-gas mass ratio—naturally explaining the near-constant bright-end cutoff of the planetary nebula luminosity function (PNLF) across galaxy types. Furthermore, we demonstrate that photoionized nebulae produced by surviving double white-dwarf merger remnants—lacking the hydrogen recombination lines typical of PNe—could populate the high-luminosity tail of the PNLF in both young and old stellar systems, offering a unified explanation for unusually bright PNe observed in early-type galaxies.
White Dwarf
Merger Remnants
We investigate the observational prospects for detecting surviving remnants of double carbon–oxygen white dwarf mergers by modeling their surrounding photoionized nebulae with CLOUDY, predicting weak hydrogen recombination lines and strong carbon and oxygen recombination and fine-structure lines in the UV, optical, and IR for the ∼10⁴ yr post-merger phase. From these models, we show that narrow-band imaging or integral field spectroscopy of high-stellar-mass, low-star-formation-rate regions (e.g., M31’s bulge, M87’s outskirts, and the Milky Way) should reveal multiple candidates, and we successfully reproduce the WISE nebula around IRAS 00500+6713 while predicting strong [Ne VI] and [Mg VII] lines detectable with JWST.
Negative Long Lag in Fairall 9’s Continuum
We report the first detection of a persistent, long-timescale negative lag (~–70 days) in the Seyfert 1 galaxy Fairall 9, where variations in bluer bands systematically lag those in redder bands—opposite to the short (<10 day) reverberation lags previously measured. By modeling the frequency dependence of this long lag with a thin-disk framework, we infer that the accretion disk’s scale height likely increases with radius, opening a new avenue for mapping AGN disk structure over a broad range of physical scales.
We performed 3D radiative general-relativistic magnetohydrodynamics (GRMHD) simulations of accretion onto M87’s supermassive black hole using the public code EBHlight, self-consistently evolving a frequency-dependent Monte Carlo radiation field in both SANE and MAD magnetic-flux regimes with subgrid electron heating models. From these, we found that the near-horizon photon energy density exceeds simple isotropic estimates by an order of magnitude, the photon momentum distribution is strongly anisotropic, and—by feeding the simulated fields into an analytic gap-acceleration model—we predict very-high-energy γ-ray luminosities (~10^41 erg s⁻¹ for MAD, ~2×10^40 erg s⁻¹ for SANE) comparable to M87’s observed VHE flares.