We propose to use the CO emission lines and the signature that metals and ions leave on the CMB during reionization as a complementary window to H21 cm observations of that cosmological epoch. Since all those effects are associated to a given resonant frequency, different observing frequencies probe different redshits, and a tomographic study of reionization can be performed. Furthermore, these signals show a definite behavior versus the observing spectral resolution which is different from all potential foregrounds (which remain essentially constant), and this should allow for component separation. The CO and metal lines should arise in regions where star forming activity has taken place, and therefore are likely to give a *complementary* picture of reionization. In contrast, H21cm observations probe the neutral regions of the Universe, not yet affected by the UV radiation of the first stars. We propose to study all these effects in a common theoretical frame, validated by the analysis of state-of-the-art numerical simulations, and investigate the astrophysical and cosmological constraints that future observations will be able to set on reionization. This project becomes particularly timely with the advent of radio, millimeter and centimeter data from missions like LOFAR, ALMA, Planck, HERSCHEL, SPT or ACT. This is a field of great future perspective, and both the candidate and the host institution are ideally suited for this project, and both would largely benefit from it in a medium and large term.