Orbital period changes are an important diagnostic for understanding low mass X-ray binary (LMXB) accretion-induced angular momentum exchange and overall system evolution. We present our most recent results for the eclipse timing of the LMXB EXO0748-676. Since its discovery in 1985 it has apparently undergone three distinct orbital period epochs, each characterized by a different orbital period than the previous epoch. We outline the orbital period behavior for EXO0748-676 over the past 18 years and discuss the implications of this behavior in light of current theoretical ideas for LMXB evolution.
We report our complete database of X-ray eclipse timings of the low mass X-ray binary EXO0748-676 observed by the Rossi X-Ray Timing Explorer (RXTE) satellite. As of this writing we have accumulated 443 full X-ray eclipses, 392 of which have been observed with the Proportional Counter Array on RXTE. These include both observations where an eclipse was specifically targeted and those eclipses found in the RXTE data archive. Eclipse cycle count has been maintained since the discovery of the EXO0748-676 system in February 1985. We describe our observing and analysis techniques for each eclipse and describe improvements we have made since the last compilation by Wolff et al. (2002). The principal result of this paper is the database containing the timing results from a seven-parameter fit to the X-ray light curve for each observed eclipse along with the associated errors in the fitted parameters. Based on the standard O-C analysis, EXO0748-676 has undergone four distinct orbital period epochs since its discovery. In addition, EXO0748-676 shows small-scale events in the O-C curve that are likely due to short-lived changes in the secondary star.
We have observed an unusually strong X-ray burst as a part of our regular eclipse timing observations of the low mass binary system EXO0748-676. The burst peak flux was 5.2x10^-8 ergs cm^-2 s^-1, approximately five times the normal peak X-ray burst flux observed from this source by RXTE. Spectral fits to the data strongly suggest that photospheric radius expansion occurred during the burst. In this Letter we examine the properties of this X-ray burst, which is the first example of a radius expansion burst from EXO0748-676 observed by RXTE. We find no evidence for coherent burst oscillations. Assuming that the peak burst luminosity is the Eddington luminosity for a 1.4 solar mass neutron star we derive a distance to EXO0748-676 of 7.7 kpc for a helium-dominated burst photosphere and 5.9 kpc for a hydrogen-dominated burst photosphere.
We present 7 eclipse timings of the low mass X-ray binary EXO0748-676 obtained with the USA experiment during 1999-2000 as well as 122 eclipse timings obtained with RXTE during 1996-2000. According to our analysis, the mean orbital period has increased by ~8 ms between the pre-RXTE era (1985-1990) and the RXTE/USA era (1996-2000). This corresponds to an orbital period derivative of P(orb)/(dP(orb)/dt)~2x10^7 years. However, neither a constant orbital period derivative nor any other simple ephemeris provides an acceptable fit to the data: individual timings of eclipse centers have residuals of up to 15 or more seconds away from our derived smooth ephemerides. When we consider all published eclipse timing data including those presented here, a model that includes observational measurement error, cumulative period jitter, and underlying period evolution is found to be consistent with the timing data. We discuss several physical mechanisms for LMXB orbital evolution in an effort to account for the change in orbital period and the observed intrinsic jitter in the mid-eclipse times.
The accretion behaviour in low-mass X-ray binaries (LMXBs) at low luminosities, especially at <E34 erg/s, is not well known. This is an important regime to study to obtain a complete understanding of the accretion process in LMXBs, and to determine if systems that host neutron stars with accretion-heated crusts can be used probe the physics of dense matter (which requires their quiescent thermal emission to be uncontaminated by residual accretion). Here we examine ultraviolet (UV) and X-ray data obtained when EXO 0748-676, a crust-cooling source, was in quiescence. Our Hubble Space Telescope spectroscopy observations do not detect the far-UV continuum emission, but do reveal one strong emission line, Civ. The line is relatively broad (>3500 km/s), which could indicate that it results from an outflow such as a pulsar wind. By studying several epochs of X-ray and near-UV data obtained with XMM-Newton, we find no clear indication that the emission in the two wavebands is connected. Moreover, the luminosity ratio of Lx/Luv >100 is much higher than that observed from neutron star LMXBs that exhibit low-level accretion in quiescence. Taken together, this suggests that the UV and X-ray emission of EXO 0748-676 may have different origins, and that thermal emission from crust-cooling of the neutron star, rather than ongoing low-level accretion, may be dominating the observed quiescent X-ray flux evolution of this LMXB.
We utilize multi-dimensional simulations of varying equatorial jet strength to predict wavelength dependent variations in the eclipse times of gas-giant planets. A displaced hot-spot introduces an asymmetry in the secondary eclipse light curve that manifests itself as a measured offset in the timing of the center of eclipse. A multi-wavelength observation of secondary eclipse, one probing the timing of barycentric eclipse at short wavelengths and another probing at longer wavelengths, will reveal the longitudinal displacement of the hot-spot and break the degeneracy between this effect and that associated with the asymmetry due to an eccentric orbit. The effect of time offsets was first explored in the IRAC wavebands by Williams et. al (2006). Here we improve upon their methodology, extend to a broad ranges of wavelengths, and demonstrate our technique on a series of multi-dimensional radiative-hydrodynamical simulations of HD 209458b with varying equatorial jet strength and hot-spot displacement. Simulations with the largest hot-spot displacement result in timing offsets of up to 100 seconds in the infrared. Though we utilize a particular radiative hydrodynamical model to demonstrate this effect, the technique is model independent. This technique should allow a much larger survey of hot-spot displacements with JWST then currently accessible with time-intensive phase curves, hopefully shedding light on the physical mechanisms associated with thermal energy advection in irradiated gas-giants.