The recent years have seen a breakthrough in the domain of high redshift galaxies, and in particular in the knowledge of their molecular gas and their star formation with ALMA. While the uv/optical/infrared domains give information on the star formation density, and its evolution in the Universe (e.g. Madau and Dickinson, 2014), the efficiency of star formation requires knowledge of the gas content. Molecular gas is the fuel of star formation, and its observation is necessary to understand galaxy formation.
ALMA with its 66 dishes (54 antennae of 12m and 12 antennae of 7m, located in a unique high and dry site, has increased the power of previous (before 2010) millimetric arrays by an order of magnitude. Baselines from 20m to 16km, at wavelengths between 3mm to 0.3mm, provide spatial resolutions up to 15mas. The large bandwidth of 7.5 GHz/polar ensures a high sensitivity for continuum observations, and allows to search and determine redshifts. ALMA is well adapted for deep fields, but not for big surveys, the field of view is from 1 arcmin (at 3mm) to 6 arcsec (at 0.3mm). Mapping small regions with mosaics is very efficient.
The main advantage for high redshift galaxies is that the peak of dust emission usually around 100 microns, for star forming objects, is redshifted to the submm and mm domain. This produces a negative K-correction, i.e. continuum emission from dust is as easy to detect at z = 10 than at z = 1 (e.g. Blain et al, 2002). Already in the pre-ALMA era, it was possible to detect hundreds of high-z galaxies with L(IR) > 1012 L⊙, up to z = 6 (e.g. Omont, 2007). The large derived dust masses of ∼ 108 M⊙, mean that dust forms early in the universe. These sub-millimeter galaxies (SMG) contribute significantly to the sub millimetre background, their redshift distribution peaks at z = 2-3 (Chapman et al, 2005).
For the CO lines, there is no negative K-correction, although the flux is higher at the upper levels of the ladder (high J), when the gas is dense enough (e.g. Combes et al, 1999). Distant galaxies have started to be explored in molecular lines in 1992, with lensed objects (Brown and Vanden Bout, 1992, Downes et al, 1995), and line detections followed at a high rate, about 50 objects up to z = 6.4 with the quasar J1148+5251 (Cox, 2005, Maiolino et al, 2005).
With ALMA, it is now possible to detect CO lines in a large amount of high-z galaxies, even not amplified by gravitational lensing. It is possible to discover obscured objects in deep fields, from their dust emission, and search for their redshift, when it is not possible in the optical domain. For z > 6 galaxies, the high-J CO lines (J > 7) are observed at low frequencies (3mm) with a field of view of 1 arcmin, and a bandwidth of 2x 8GHz ∼ 16%, or 50 000km/s. At z = 6, the spacing between the various CO lines of the rotational ladder is of 16 GHz, so that the redshift may be obtained with 2 tunings only.
This review highlights the main results of ALMA observations of distant galaxies, from the dust emission to molecular lines. The CO are the most usual gas tracers, but at very high z, the [CII] and CI lines bring important information, in a domain where the CO lines are not or little excited. Dense gas tracers (HCN, HCO+, CS, etc) and isotopes, bring complementary knowledge on gas properties, and are frequently observed simultaneously, thanks to the wide bandwidth of ALMA.
The review emphasizes only the recent results since the previous reviews in the domain, pre- or post-starting of ALMA (Solomon and Vanden Bout 2005, Carilli and Walter 2013).