Low-Mass X-ray Binaries and Globular Clusters in Early-Type Galaxies. I. Chandra Observations

Abstract

We present a Chandra survey of LMXBs in 24 early-type galaxies. Correcting for detection incompleteness, the X-ray luminosity function (XLF) of each galaxy is consistent with a powerlaw with negative logarithmic differential slope, beta~2.0. However, beta strongly correlates with incompleteness, indicating the XLF flattens at low-Lx. The composite XLF is well-fitted by a powerlaw with a break at 2.21(+0.65,-0.56)E38 erg/s and beta=1.40(+0.10,-0.13) and 2.84(+0.39,-0.30) below and above it, respectively. The break is close to the Eddington limit for a 1.4Msun neutron-star, but the XLF shape rules out its representing the division between neutron-star and black-hole systems. Although the XLFs are similar, we find evidence of some variation between galaxies. The high-Lx XLF slope does not correlate with age, but may correlate with [alpha/Fe]. Considering only LMXBs with Lx>1E37 erg/s, matching the LMXBs with globular clusters (GCs) identified in HST observations of 19 of the galaxies, we find the probability a GC hosts an LMXB is proportional to LGC^alpha ZFe^gamma} where alpha=1.01+/-0.19 and gamma=0.33+/-0.11. Correcting for GC luminosity and colour effects, and detection incompleteness, we find no evidence that the fraction of LMXBs with Lx>1e37 erg/s in GCs (40%), or the fraction of GCs hosting LMXBs (~6.5%) varies between galaxies. The spatial distribution of LMXBs resembles that of GCs, and the specific frequency of LMXBs is proportional to the GC specific luminosity, consistent with the hypothesis that all LMXBs form in GCs. If the LMXB lifetime is tau and the duty cycle is Fd, our results imply ~1.5 (tau/1E8 yr)^-1 /Fd LMXBs are formed per Gyr per GC and we place an upper limit of 1 active LMXB in the field per 3.4E9Lsun of V-band luminosity.