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 , 18 systems have both nuclei detected in the X-rays , 39 have one nucleus detected in the X-rays , whereas 10 have no X-ray detection .
The X-ray detection rate , 75/134=56 % ( \pm 6 % 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 ~ { } { 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 ~ { } { 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 ~ { } { 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 .