Type I X-ray bursts from low-mass X-ray binaries result from a thermonuclear runaway in the material accreted onto the neutron star. Although typical recurrence times are a few hours, consistent with theoretical ignition model predictions, there are also observations of bursts occurring as promptly as ten minutes or less after the previous event. We present a comprehensive assessment of this phenomenon using a catalog of 3387 bursts observed with the BeppoSAX/WFCs and RXTE/PCA X-ray instruments. This catalog contains 136 bursts with recurrence times of less than one hour, that come in multiples of up to four events, from 15 sources. Short recurrence times are not observed from so-called ultra-compact binaries, indicating that hydrogen burning processes play a crucial role. As far as the neutron star spin frequency is known, these sources all spin fast at over 500 Hz; the rotationally induced mixing may explain burst recurrence times of the order of 10 min. Short recurrence time bursts generally occur at all mass accretion rates where normal bursts are observed, but for individual sources the short recurrence times may be restricted to a smaller interval of accretion rate. The fraction of such bursts is roughly 30%. We also report the shortest known recurrence time of 3.8 minutes.
Type I X-ray bursts are thought to result from thermonuclear flashes of hydrogen and/or helium in the envelope of neutron stars (Woosley & Taam 1976;Maraschi & Cavaliere 1977;Lamb & Lamb 1978). This material is accreted through Roche-lobe overflow from a lower-mass companion star (low-mass X-ray binary, LMXB). Current one-dimensional models successfully explain burst features such as the peak flux, the fluence, decay time and recurrence time (e.g., Woosley et al. 2004;Heger et al. 2007; see Wallace & Woosley 1981;Fujimoto et al. 1981;Fushiki & Lamb 1987 for earlier work). During the flash, over 90% of the accreted hydrogen and helium is expected to burn to carbon and heavier elements (e.g., Woosley et al. 2004). For the next flash to occur, a fresh layer of hydrogen/helium must first be accreted. At typical accretion rates of up to approximately 10 -8 M ⊙ yr -1 this takes at least a few hours.
X-ray bursts have been observed since the 1970’s (Grindlay et al. 1976;Belian et al. 1976) from approximately 90 sources in our Galaxy, with recurrence times of hours up to days (e.g., Lewin et al. 1993;Strohmayer & Bildsten 2006). Lewin et al. (1976) reported the detection with SAS-3 of three bursts that were separated by only 17 and 4 minutes. These bursts originated from a crowded region and, therefore, source confusion cannot be ruled out. In the 1980’s similar recurrence times as short as 10 minutes were observed from both 4U 1608-522 with Hakucho (Murakami et al. 1980) and from EXO 0748-676 with EXOSAT (Gottwald et al. 1986(Gottwald et al. , 1987a)). This rare phenomenon implies that hydrogen and helium is left over somewhere on the star after the initial flash, because the recurrence time is too short to accrete enough fuel for the subsequent burst(s). This is at odds with the current models, that predict an almost complete burning of the available hydrogen and helium on the entire star surface. Boirin et al. (2007) analyzed 158 hours of XMM-Newton observations of EXO 0748-676, which revealed short recurrence time bursts in groups of two (doubles) and three (triples). This relatively large burst sample revealed that on average bursts with a short recurrence time (8 to 20 minutes) are less bright and energetic than bursts with ’normal’ recurrence times (over 2 hours). The fit of a black body model to the burst spectrum shows a lower peak temperature, while the emitting area is the same. The profiles of short recurrence time bursts seemingly lack the long 50 s to 100 s tail caused by rp-process burning, which indicates that the burst fuel contains less hydrogen. After a double or triple it takes on average more time before another burst occurs, suggesting a more complete burning of the available fuel. Galloway et al. (2008) showed that there are more sources that show this behavior, with bursts occurring in groups of up to four bursts, and with recurrence times as short 6.4 minutes. The short recurrence times were observed predominantly when the persistent flux is between approximately 2% and 4% of the Eddington limited flux. Furthermore, indications were found for the association of short recurrence times and the accretion of hydrogenrich material. The shortest recurrence time previously reported is 5.4 minutes (Linares et al. 2009).
Different ideas have been put forward to explain this rare bursting behavior. As most of the models only resolve the neutron star envelope in the radial direction, it is possible that short recurrence time bursts are due to multi-dimensional effects, such as the confinement of accreted material on different parts of the surface, possibly as the result of a magnetic field (e.g., Melatos & Payne 2005;Lamb et al. 2009). Boirin et al. (2007), however, found that the different bursts originate from an emitting area of similar size. Furthermore, the indication of a different fuel composition for the bursts with short recurrence times, argues against any scenario where accreted material of the same composition burns on different parts of the surface.
The idea of a burning layer with an unburned layer on top has been investigated (Fujimoto et al. 1987). After the first layer flashes, the second layer could be mixed down to the depth where a thermonuclear runaway occurs. Mixing may be driven by rotational hydrodynamic instabilities (Fujimoto 1988) or by instabilities due to a rotationally induced magnetic field (Piro & Bildsten 2007;Keek et al. 2009). The mixing processes take place on the correct time scale of approximately ten minutes. Although this scenario is able to explain many of the observed aspects of short recurrence time bursts, it has not been reproduced with a multi-zone stellar evolution code that includes a full nuclear burning network. Taam et al. (1993) created models that exhibit ’erratic’ bursting behavior, reminiscent of short recurrence time bursts. Later versions of the employed code, however, no longer produce this, most likely because of the inclusion of a more ext
This content is AI-processed based on open access ArXiv data.
