We have studied the kinematics of \sim 2800 candidate close pair galaxies at 0.1 < z < 1.2 identified from the Canada-France-Hawaii Telescope Legacy Survey fields .
Spectra of these systems were obtained using spectrometers on the 6.5m Magellan and 5m Hale telescopes .
These data allow us to constrain the rate of dry mergers at intermediate redshifts and to test the ‘ hot halo ’ model for quenching of star formation .
Using virial radii estimated from the correlation between dynamical and stellar masses published by Leauthaud et al .
( 2011 ) , we find that around 1/5 of our candidate pairs are likely to share a common dark matter halo ( our metric for close physical association ) .
These pairs are divided into red-red , blue-red and blue-blue systems using the rest-frame colors classification method introduced in Chou et al .
( 2011 ) .
Galaxies classified as red in our sample have very low star-formation rates , but they need not be totally quiescent , and hence we refer to them as ‘ damp ’ , rather than ‘ dry ’ , systems .
After correcting for known selection effects , the fraction of blue-blue pairs is significantly greater than that of red-red and blue-red pairs .
Red-red pairs are almost entirely absent from our sample , suggesting that damp mergers are rare at z \sim 0.5 .
Our data supports models with a short merging timescale ( < 0.5 Gyr ) in which star-formation is enhanced in the early phase of mergers , but quenched in the late phase .
Hot halo models may explain this behaviour , but only if virial shocks that heat gas are inefficient until major mergers are nearly complete .