A Chandra X-ray Survey of Optically Selected AGN Pairs

We present a Chandra archival study of optically selected AGN pairs at a median redshift $\bar{z} \sim$ 0.1. Out of 1286 AGN pairs (with projected separations $r_p < 100$ kpc and velocity offsets $\Delta v < 600$ km s$^{-1}$) optically identified from the Sloan Digital Sky Survey Seventh Data Release, we find 67 systems with archival Chandra observations, which represent the largest sample of optically selected AGN pairs studied in the X-ray. Among the 67 AGN pairs, 21 systems have both nuclei detected in the X-rays, 36 have one nucleus detected in the X-rays, whereas 10 have no X-ray detection. The X-ray detection rate, 78/134 = 58% ($\pm 7$% 1$\sigma$ Poisson errors), is significantly higher than that (23/134 = $17\% \pm 4\%$) of a comparison sample of star-forming galaxy pairs, lending support to the optical AGN classification. In the conservative case where X-ray contamination from star formation is removed, the X-ray detection rate becomes $27\% \pm 4\%$, consistent with predictions from latest galaxy merger simulations. The 2-10 keV X-ray luminosity $L_{2-10~{\rm keV}}$ increases with decreasing projected separation in AGN pairs for $r_p \gtrsim 15$ kpc, suggesting an enhancement of black hole accretion even in early-stage mergers. On the other hand, $L_{2-10~{\rm keV}}$ appears to decrease with decreasing projected separation at $r_p \lesssim 15$ kpc, which is contradictory to predictions from merger simulations. The apparent decrease in $L_{2-10~{\rm keV}}$ of AGN pairs at $r_p \lesssim 15$ kpc may be caused by: (i) enhanced absorbing columns from merger-induced gas inflows, (ii) feedback effects from early-stage mergers, and/or (iii) small number statistics. Future X-ray studies with larger samples are needed to put our results on firmer statistical ground.