Galaxy pairs in the Sloan Digital Sky Survey - VII. The merger-luminous infrared galaxy connection

We use a sample of 9397 low-redshift (z ≤ 0.1) galaxies with a close companion to investigate the connection between mergers and luminous infrared (IR) galaxies (LIRGs). The pairs are selected from the Sloan Digital Sky Survey (SDSS) and have projected separations r p ≤ 80 h kpc, relative velocities ΔV ≤ 300 km s−1 and stellar mass ratios within a factor of 1:10. A control sample consisting of four galaxies per pair galaxy is constructed by simultaneously matching in stellar mass, redshift and environment to galaxies with no close companion. The IR luminosities (L IR) of galaxies in the pair and control samples are determined from the SDSS - Infrared Astronomical Satellite (IRAS) matched catalogue of Hwang et al. Over the redshift range of our pairs sample, the IRAS matches are complete to LIRG luminosities (L IR ≥ 1011 L), allowing us to investigate the connection between mergers and luminous IR galaxies. We find a trend for increasing LIRG fraction towards smaller pair separations, peaking at a factor of ∼5-10 above the median control fraction at the smallest separations (r p < 20 h kpc), but remaining elevated by a factor ∼2-3 even out to 80 h kpc (the widest separations in our sample). LIRG pairs predominantly have high star formation rates (SFRs), high extinction and are found in relatively low-density environments, relative to the full pairs sample. We also find that LIRGs are most likely to be found in high-mass galaxies which have an approximately equal-mass companion. We confirm the results of previous studies that both the active galactic nucleus (AGN) fraction and merger fraction increase strongly as a function of IR luminosity. About 7 per cent of LIRGs are associated with major mergers, as defined within the criteria and mass completion of our sample. Finally, we quantify an SFR offset (ΔSFR) as the enhancement (or decrement) relative to star-forming galaxies of the same mass and redshift. We demonstrate that there is a clear connection between the ΔSFR and the classification of a galaxy as a LIRG that is mass dependent. Most of the LIRGs in our merger sample are relatively high-mass galaxies (log (M /M) > 10.5), likely because the SFR enhancement required to produce LIRG luminosities is more modest than at low masses. The ΔSFR offers a redshift-independent metric for the identification of the galaxies with the most enhanced star-forming rates that does not rely on fixed L IR boundaries.