Observational constraints on the formation of pre-galactic condensates in the universe.

Abstract

The distribution of galaxies in the Universe displays structure on a variety of scales, ranging from groups of a few galaxies to rich clusters containing thousands of galaxies. These clusters are in turn clustered to form superclusters that are often elongated, filamentary structures. The work presented in this thesis has provided observational and theoretical constraints on the evolutionary scenario that has resulted in the eventual formation of the large?scale structure we see in the present?day Universe. Observational constraints on the nature of the Universe during the “dark ages”—the period subsequent to recombination and prior to the formation of the first stars—have been lacking. The Universe is observable prior to “first visible light” through the 21 cm emission from the neutral hydrogen component of matter. We have used simple theoretical models to derive the characteristics of the emission expected to be received from a variety of structures that could be present at high redshifts. Considering the evolution of a top?hat density perturbation having a mass corresponding to clusters of galaxies, we expect the inhomogeneity to appear in absorption during its linear?evolution phase but in emission later on. The linearly evolving perturbations would be optically thin, with an increasing integrated flux density in the line profile that progressively becomes narrower. Close to maximum expansion, the inhomogeneity is optically thick and is seen as narrow spectral absorption lines. Subsequent to collapse and virialization, the condensates are again optically thin but display broad emission lines. The perturbations spend a very small fraction of their lifetimes in the phase where their emission is confined to narrow spectral spikes, making detection at this stage improbable. Post?virialization, the received line profiles from adjacent condensates along the line of sight could increasingly overlap, and hence the spectral fluctuations are severely reduced. This is true unless the condensates have a very large dispersion in the epoch of their formation compared to their lifetimes in the neutral gaseous phase, but in such a case the number of visible protoclusters in any search volume would be small. Sensitive searches conducted with radio telescopes at appropriate observing frequencies can distinguish between various morphologies in the matter distribution by a careful examination of the characteristics of the excess variance in the spectral?line cubes. A correlation spectrometer was built to conduct observations over a large fractional bandwidth (??/? ? 1/35) using the Ooty Radio Telescope. A search was conducted over a comoving volume of 10? h?³ Mpc³ in an attempt to detect proto?superclusters at the redshift z = 3.3. The search placed 2? limits in the range ~1 × 10¹³ to 3.5 × 10¹³ h?² M? on the HI mass in proto?superclusters having line?of?sight velocity dispersions between 600–2100 km s?¹. The observations indicate that if supercluster?mass condensates are the first objects to separate out from the Hubble expansion, as is expected in a universe dominated by hot dark matter, then their formation and subsequent fragmentation into galaxies must be completed at redshifts z > 3.3. A complementary search was conducted using the VLA, covering a comoving search volume of 10³ Mpc³ and attempting to detect protoclusters with line?of?sight velocity dispersions in the range 200–900 km s?¹. The observing frequency corresponded to a search redshift of z = 3.3. While the large survey volume and fractional bandwidth of the Ooty observations have constrained the evolution of superclusters, the higher sensitivity and spectral resolution of the VLA observations aimed at constraining the evolution of the smaller?mass clusters of galaxies. Within the search volume and over the velocity dispersions stated above, the VLA observations have placed 2? upper limits in the range 2 × 10¹³ to 1.2 × 10¹? h?² M? on the HI content of protoclusters at z = 3.3. The results indicate that present?day clusters did not exist in the neutral gaseous phase at z = 3.3. The observations also show no evidence for HI condensates at z = 3.3 corresponding to present?day rich (R ? 1) Abell clusters. There is also weak evidence (~1? significance) indicating that the low?density clusters in the present?day Universe, whose parent perturbations are expected to be approaching the virialized state at z = 3.3, are also not in the neutral gaseous state. The observations are consistent with the predictions of a cold?dark?matter?dominated universe, where galaxies form at redshifts z > 4 and structures on larger scales form subsequently by hierarchical clustering. In a neutrino?dominated (hot dark matter) universe, the proto?superclusters, as well as the protoclusters, must have completed their lifetimes in the neutral gaseous phase at epochs prior to z = 3.3. Another probe into the scenario leading to galaxy formation is the diffuse X?ray background (XRB), which originates at cosmological redshifts. We find that the spectral shape in the 3–80 keV energy band is well fit by bremsstrahlung radiation from a hot plasma at a temperature of ~60 keV. Our investigations suggest that the XRB finds a natural origin in the evolutionary scenario of a neutrino?dominated universe. Since perturbations in the neutrino dark matter commence growth prior to recombination, the baryonic matter is expected to fall into supercluster?scale potential wells created by the dark matter. Because these potential well depths are larger than the characteristic temperature of the XRB, infall heating and the consequent bremsstrahlung emission at redshift z ~ 3–5 would naturally be visible today as a redshifted bremsstrahlung spectrum. Considerations involving the characteristics—spectral shape, intensity, and isotropy—of the XRB and the Cosmic Microwave Background have led to constraints on the subsequent evolution of the infalling gas. It appears that around z ~ 5, infall results in the formation of high?density (n ? 0.3 cm?³) fragments, with a hot plasma phase constituting ~½ of the baryonic content of the Universe. Taken together, our investigations suggest a scenario in which proto?superclusters form first at z > 3.3. These fragment into galaxies (at z > 3.3), which subsequently cluster hierarchically to form the clusters of galaxies. Our theoretical investigations into the signatures of structures at high redshifts, as observed in the 21 cm emission from neutral hydrogen, have yielded various methods for examining observational data. The data we have obtained using currently operational telescopes do not yet have the sensitivity required to justify the application of many of these tests. Stronger constraints on the nature of the Universe during the “dark ages” will require observations that are an order of magnitude more sensitive, and which cover several frequencies corresponding to redshifted 21 cm emission from redshifts between z = 4 and 10. This capability is expected with the commissioning of the Giant Metrewave Radio Telescope (GMRT) in India.