2006ApJ...644..709G

We present 4-7 ks Chandra observations of 35 broad absorption line (BAL) quasars from the Large Bright Quasar Survey, the largest sample of sensitive, 0.5-8.0 keV X-ray observations of this class of quasars to date. The limited ranges in both redshift (z=1.42-2.90) and ultraviolet luminosity (a factor of ~12) of the sample also make it relatively uniform. Of 35 targets, 27 are detected for a detection fraction of 77%, and we confirm previous studies that find BAL quasars to be generally X-ray-weak. Five of the eight nondetections are known low-ionization BAL quasars, confirming reports of extreme X-ray weakness in this subset (∼10% of optically selected BAL quasars). Those BAL quasars with the hardest X-ray spectra are also the X-ray weakest, consistent with the interpretation that intrinsic absorption is the primary cause of X-ray weakness in this class of quasars as a whole. Furthermore, the observed trend is not consistent with simple neutral absorption, supporting findings from spectroscopic observations of individual targets that BAL quasars typically exhibit complex X-ray absorption (e.g., partially covering or ionized absorbers). Assuming normal quasar X-ray continua and using the hard-band (observed-frame 2-8 keV) X-ray flux to ``correct'' for the effects of intrinsic absorption at softer energies increases the relative X-ray to optical flux ratios to much closer to the range for normal quasars, further indicating that typically neither are high-ionization BAL quasars intrinsically X-ray-weak, nor do they suffer from Compton-thick absorption. In general, we find no evidence for correlations between X-ray weakness and ultraviolet absorption-line properties, with the exception of a likely correlation between the maximum outflow velocity of C IV absorption and the magnitude of X-ray weakness. We discuss the implications of our results for disk-wind models of BAL outflows in quasars.