Observationally inferred superburst ignition depths are shallower than models predict. We address this discrepancy by reexamining the superburst trigger mechanism. We first explore the hypothesis of Kuulkers et al. that exothermic electron captures trigger superbursts. We find that all electron capture reactions are thermally stable in accreting neutron star oceans and thus are not a viable trigger mechanism. Fusion reactions other than 12 C + 12 C are infeasible as well since the possible reactants either deplete at much shallower depths or have prohibitively large Coulomb barriers. Thus we confirm the proposal of Cumming & Bildsten and Strohmayer & Brown that 12 C + 12 C triggers superbursts. We then examine the 12 C + 12 C fusion rate. The reaction crosssection is experimentally unknown at astrophysically relevant energies, but resonances exist in the 12 C + 12 C system throughout the entire measured energy range. Thus it is likely, and in fact has been predicted, that a resonance exists near the Gamow peak energy E pk ≈ 1.5 MeV. For such a hypothetical 1.5 MeV resonance, we derive both a fiducial value and upper limit to the resonance strength (ωγ) R and find that such a resonance could decrease the theoretically predicted superburst ignition depth by up to a factor of 4; in this case, observationally inferred superburst ignition depths would accord with model predictions for a range of plausible neutron star parameters. Said differently, such a resonance would decrease the temperature required for unstable 12 C ignition at a column depth 10 12 g cm −2 from 6 × 10 8 K to 5 × 10 8 K. A resonance at 1.5 MeV would not strongly affect the ignition density of Type Ia supernovae, but it would lower the temperature at which 12 C ignites in massive post-main-sequence stars. Determining the existence of a strong resonance in the Gamow window requires measurements of the 12 C + 12 C cross-section down to a center-of-mass energy near 1.5 MeV, which is within reach of the proposed DUSEL facility.

Superbubbles surrounding OB associations provide ideal laboratories in which to study the stellar energy feedback problem, because the stellar energy input can be estimated from the observed stellar content of the OB associations, and the interstellar thermal and kinetic energies of superbubbles are well defined and easy to observe. We have used DEM L192, also known as N51D, to carry out a detailed case study of the energy budget in a superbubble, and we find that the expected amount of stellar mechanical energy injected into the interstellar medium, ð18 AE 5Þ Â 10 51 ergs, exceeds the amount of thermal and kinetic energies stored in the superbubble, ð6 AE 2Þ Â 10 51 ergs. Clearly, a significant fraction of the stellar mechanical energy must have been converted into other forms of energy. The X-ray spectrum of the diffuse emission from DEM L192 requires a power-law component to explain the featureless emission at 1.0-3.0 keV. The origin of this power-law component is unclear, but it may be responsible for the discrepancy between the stellar energy input and the observed interstellar energy in DEM L192.

We carry out a general relativistic global linear stability analysis of the amassed carbon fuel on the surface of an accreting neutron star to determine the conditions under which superbursts occur. We reproduce the general observational characteristics of superbursts, including burst fluences, recurrence times, and the absence of superbursts on stars with accretion ratesṀ < 0:1Ṁ Edd , whereṀ Edd denotes the Eddington limit. By comparing our results with observations, we are able to set constraints on neutron star parameters such as the stellar radius and neutrino cooling mechanism in the core. Specifically, we find that accreting neutron stars with ordered crusts and highly efficient neutrino emission in their cores (due to direct Urca or pionic reactions, for example) produce extremely energetic (>10 44 ergs) superbursts that are inconsistent with observations, in agreement with previous investigations. Also, because of pycnonuclear burning of carbon, they do not have superbursts in the range of accretion rates at which superbursts are actually observed unless the crust is very impure. Stars with less efficient neutrino emission (due to modified Urca reactions, for example) produce bursts that agree better with observations. Stars with highly inefficient neutrino emission in their cores produce bursts that agree best with observations. Furthermore, we find that neutron stars with large radii (R $16 km) produce very energetic superbursts that conflict with observations, even if the core neutrino emission mechanism is highly inefficient. Superburst characteristics are quite sensitive to several other parameters as well, most notably the composition of the accreted gas, concentration of carbon in the ignition region, and degree of crystallization of the crust. All systems that accrete primarily hydrogen and in which superbursts are observed show evidence of H-and He-burning delayed mixed bursts. We speculate that delayed mixed bursts provide sufficient amounts of carbon fuel for superbursts and are thus a prerequisite for having superbursts. We compare our global stability analysis to approximate one-zone criteria used by other authors and identify a particular set of approximations that give accurate results for most choices of parameters.

We investigate the latitude at which type I X-ray bursts are ignited on rapidly rotating accreting neutron stars. We find that, for a wide range of accretion rates , ignition occurs preferentially at the equator, in accorḋ M with the work of Spitkovsky et al. However, for a range of below the critical above which bursts cease,Ṁ M ignition occurs preferentially at higher latitudes. The range of over which nonequatorial ignition occurs iṡ M an increasing function of the neutron star spin frequency. These findings have significant implications for thermonuclear flame propagation, and they may explain why oscillations during the burst rise are detected predominantly when the accretion rate is high. They also support the suggestion of Bhattacharyya & Strohmayer that non-photospheric radius expansion double-peaked bursts and the unusual harmonic content of oscillations during the rise of some bursts result from ignition at or near a rotational pole.