DEEP NEAR-INFRARED SPECTROSCOPY OF PASSIVELY EVOLVING GALAXIES AT z ≳ 1.4

We present the results of new near-IR spectroscopic observations of passive galaxies at z greater than or similar to 1.4 in a concentration of BzK-selected galaxies in the COSMOS field. The observations have been conducted with Subaru/MOIRCS, and have resulted in absorption lines and/or continuum detection for 18 out of 34 objects. This allows us to measure spectroscopic redshifts for a sample that is almost complete to K-AB = 21. COSMOS photometric redshifts are found in fair agreement overall with the spectroscopic redshifts, with a standard deviation of similar to 0.05; however, similar to 30% of objects have photometric redshifts systematically underestimated by up to similar to 25%. We show that these systematic offsets in photometric redshifts can be removed by using these objects as a training set. All galaxies fall in four distinct redshift spikes at z = 1.43, 1.53, 1.67, and 1.82, with this latter one including seven galaxies. SED fits to broadband fluxes indicate stellar masses in the range of similar to 4-40 x 10(10) M-circle dot and that star formation was quenched similar to 1 Gyr before the cosmoic epoch at which they are observed. the spectra of several individual galaxies have allowed us to measure their H delta(F) indices and the strengths of the 4000 angstrom break, which confirms their identification as passive galaxies, as does a composite spectrum resulting from the co-addition of 17 individual spectra. The effective radii of the galaxies have been measured on the COSMOS HST/ACS i(F814W)-band image, confirming the coexistence at these redshifts of passive galaxies, which are substantially more compact than their local counterparts with others that follow the local effective radius-stellar mass relation. For the galaxy with the best signal-to-noise spectrum we were able to measure a velocity dispersion of 270 +/- 105 km s(-1) (error bar including systematic errors), indicating that this galaxy lies closely on the virial relation given its stellar mass and effective radius.