Paper detail

Constraints on the low-mass end of the mass-metallicity relation at z=1-2 from lensed galaxies

We present multi-wavelength imaging and near-IR spectroscopy for ten gravitationally lensed galaxies at 0.9<z<2.5 selected from a new, large sample of strong lens systems in the Sloan Digital Sky Survey (SDSS) DR7. We derive stellar masses from the rest-frame UV to near-IR spectral energy distributions, star formation rates (SFR) from the dust-corrected Ha flux, and metallicities from the [N II]/Ha flux ratio. We combine the lensed galaxies with a sample of sixty star-forming galaxies from the literature in the same redshift range for which measurements of [N II]/Ha have been published. Due to the lensing magnification, the lensed galaxies probe intrinsic stellar masses that are on average a factor of 11 lower than have been studied so far at these redshifts. They have specific star formation rates that are an order of magnitude higher than seen for main-sequence star-forming galaxies at z~2. We measure an evolution of 0.16+/-0.06 dex in the mass-metallicity relation between z~1.4 and z~2.2. In contrast to previous claims, the redshift evolution is smaller at low stellar masses. We do not see a correlation between metallicity and SFR at fixed stellar mass. The combined sample is in general agreement with the local fundamental relation between metallicity, stellar mass and SFR from Mannucci et al. (2010, 2011). Using the Kennicutt-Schmidt law to infer gas fractions, we investigate the importance of gas inflows and outflows on the shape of the mass-metallicity relation using simple analytical models. This suggests that the Maiolino et al.(2008) calibration of the [N II]/Ha flux ratio is biased high. We conclude that both an absolute metallicity calibration and direct measurements of the gas mass are needed to use the observed mass-metallicity relation to gain insight into the impact of gas flows on the chemical evolution of galaxies.