Abstract
What do we expect to see in thehydrogensignal at 21 cm, shifted to 2 mwavelength, for the redshift z = 9 ? Theory shows that if the temperature of the cosmological background varies as (1 + z), the temperature of gas atoms, under adiabatic expansion, decreases as (1 + z)2. Hydrogen will therefore become colder than the background. This could be seen in absorption ? For this to happen, the spin temperature of HI's hyperfine fundamental level would have to thermalize. This is possible at the beginning of the period, as the density of the Universe is quite high, and by collision Tspin = T. In this early period, known as the cosmic dawn, HI is expected to absorb in front of the CMB background. Then, as the expansion proceeds, the density is no longer sufficient, and Tspin = T(CMB), which implies no signal at all. Fortunately, as soon as there are enough stars and ionizing UV radiation, hydrogen recombines with Lyman-alpha lines, and the Wouthuysen-Field effect causes the atoms to fall back to the upper level, as in optical pumping. This allows Tspin = T to be thermalized again, and we should see absorption in front of the CMB background. Finally, star formation, supernovae and X-rayswill heat the HI gas to a temperature Thigher than the CMB, and emission will be detected, just at the end of the reionization period : then there will be no more atoms. Only simulations can predict in detail the evolution of the signal as a function of redshift, as reionization and heating depend on the environment. Heating by X-raysdepends on the number of Xbinariesin the galaxy, and also on active nuclei. Many free parameters can be deduced from observations, using a library of models.
