A bright year for tidal disruptions

Abstract

When a star is tidally disrupted by a supermassive black hole (BH), roughly half of its mass falls back to the BH at super-Eddington rates. Being tenuously gravitationally bound and unable to cool radiatively, only a small fraction f_in << 1 of the returning debris will likely be incorporated into the disk and accrete, with the vast majority instead becoming unbound in an outflow of velocity ~1e4 km/s. This slow outflow spreads laterally, encasing the BH. For months or longer, the outflow remains sufficiently neutral to block hard EUV and X-ray radiation from the hot inner disk, which instead becomes trapped in a radiation-dominated nebula. Ionizing nebular radiation heats the inner edge of the ejecta to temperatures of T > few 1e4 K, converting the emission to optical/near-UV wavelengths where photons more readily escape due to the lower opacity. This can explain the unexpectedly low and temporally constant effective temperatures of optically-discovered TDE flares. For BHs with relatively high masses M_BH > 1e7 M_sun the ejecta can become ionized at an earlier stage, or for a wider range of viewing angles, producing a TDE flare which is instead dominated by thermal X-ray emission. We predict total radiated energies consistent with those of observed TDE flares, and ejecta velocities that agree with the measured emission line widths. The peak optical luminosity for M_BH < 1e6 M_sun is suppressed due to adiabatic losses in the inner disk wind, possibly contributing to the unexpected dearth of optical TDEs in galaxies with low mass BHs. In the classical picture, where f_in ~ 1, TDEs de-spin supermassive BHs and cap their maximum spins well below theoretical accretion physics limits. This cap is greatly relaxed in our model, and existing Fe K-alpha spin measurements provide suggestive preliminary evidence that f_in < 1.