This uniform matter formed atoms of hydrogen and helium when the universe was 300000 years old, and the temperature was 3000 K. At this temperature, electrons had slowed down enough for them to combine with protons, forming hydrogen. This is called the epoch of recombination: matter made a transition from a plasma, made of charged particles, to an atomic medium (a better name would be the epoch of combination, because the charged particles were combining for the first time, but for historical reasons the process of combination of charged particles to form atoms is called recombination). At the same time, the universe became transparent to its own radiation as the photons stopped being scattered by charged particles, and at the present time we can see the Cosmic Microwave Background as it was released from the grip of the plasma at the epoch of recombination.
However, we know that at the present time the diffuse intergalactic matter (matter that is left over in the vast space between galaxies) is fully ionized. We know this from observations of the spectra of quasars (and other luminous sources) at very high redshifts, which show absorption by hydrogen that is known as the Lyman alpha forest. This absorption is actually produced by a tiny density of neutral hydrogen, compared to the mean density of the universe. If atomic hydrogen were present in the line of sight to a quasar at its average density, all the light of the quasar would be completely blocked (with an optical depth of about 100000). However, we observe the light is not all blocked, which means the amount of neutral hydrogen present is very small. Most of the hydrogen is ionized, and only a very small fraction is in atomic form, determined by an equilibrium between the rate of recombinations and the rate of photoionization by a background of ultraviolet light produced by all galaxies and quasars in the universe.
Hence, since the epoch of recombination, there must have been another epoch of reionization, when matter was ionized again. In fact, this process of reionization is expected to have taken place as the density fluctuations in the universe became non-linear and produced gravitationally collapsed objects. The theory that explains the observed large-scale structure of galaxies and clusters in the universe is the Cold Dark Matter theory, which predicts that the first objects in which gas was able to cool and form stars collapsed at redshift about 30 (when the universe was around 100 million years old), and had masses of ~ one million solar masses. These first objects are thought to have produced single, very massive stars in their centers, which were metal-free and extremely hot, and emitted copious amounts of ionizing photons. They started the reionization of the universe. Later, more massive objects continued to form by mergers of previously existing collapsed halos, and accretion of intergalactic matter. The more efficient cooling of gas in more massive objects gave rise to fragmentation into small clouds and the formation of disk galaxies containing many stars.
Over the last few years, the WMAP satellite has measured the electron scattering optical depth to the Cosmic Background radiation. The measured value, tau = 0.09 +/- 0.03, is consistent with the expectation of the Cold Dark Matter theory, under reasonable assumptions for the efficiency with which stars and quasars were formed in different objects. This efficiency of the emission of ionizing radiation from collapsed objects, however, cannot be calculated from first physical principles and remains the main uncertainty for modeling reionization.
My research in this area has centered in the way diffuse matter is ionized, and how we can gain an understanding of the sources of reionization and the way their HII regions overlapped, from observations of Lyman alpha spectra in quasars. Other interesting probes to this epoch that I have investigated are the damped absorption from a fully atomic intergalactic medium that could be observed in the spectra of high-redshift gamma-ray bursts, and the 21 cm emission and absorption against the Cosmic Background radiation.
This latter method for probing reionization has been the subject of recent research I have been engaged with in collaboration with Xuelei Chen, from Beijing, where we have studied the signature of the first stars in redshifted 21 cm. At very high redshift, when the medium was still atomic, hydrogen could emit or absorb the Cosmic Background photons at 21 cm, via transitions of the hyperfine structure levels. Initially, the hydrogen was colder than the background radiation owing to the adiabatic expansion. This hydrogen should have left a special 21 cm absorption signal around a first star, where the Lyman alpha photons generated by the star and scattering in the atomic medium around it coupled the spin and kinetic temperature of the gas. Later, the atomic medium was gradually heated by X-rays emitted from the same first stars, which eventually turned the 21 cm signal into emission. This 21 cm emission can be used to observe the gradual overlap of HII regions produced by different sources as reionization advanced.
Publications related to this project
See the astro-ph preprint online.