PREGRESSO

We model the evolution of the abundances of several neutron capture elements (Ba, Eu, La, Sr, Y and Zr) in the Milky Way and then we extend our predictions to some dwarf spheroidal galaxies of the Local Group. Two major neutron capture mechanisms on iron seeds are generally invoked: the slow process (s-process) and the rapid process (r-process), where the slow and the rapid are defined relative to the timescale of the β-decay. Nucleosynthesis calculations for r-process are very few, owing to the difficulties in modelling the physics the r-process and the lack of knowledge about the sites of productions of these elements. For s-process elements instead some calculations are available but the sites of production are also uncertain. By adopting a chemical evolution model for the Milky Way already reproducing the evolution of several chemical elements (H, He, C, N, O, α-elements and iron peak elements), we compare our theoretical results with accurate and new stellar data of neutron capture elements and we are able to impose strong constraints on the nucleosynthesis of the studied elements. We can suggest the stellar sites .of production for each element. In particular, the r-process component of each element (if any) is produced in the mass range from 10 to 30 Mʘ, whereas the s-process component arises from stars in the range from l to 3 Mʘ. Using the same chemical evolution model, extended to different galactocentric distances, we obtain results on the radial gradients of the Milky Way. We compare the results of the model not only for the neutron capture elements but also for α-elements and iron peak elements with new data of Cepheids stars. For the first time with these data, it is possible to verify the predictions for the gradients of very heavy elements. We conclude that the model, with an inside-out scenario for the building up of the disc and a constant density distribution of the gas for the halo phase, can be considered successful; in fact, for almost all the considered elements with our nucleosynthesis prescriptions, the model well reproduces the observed abundance gradients. We give a possible explanation to the considerable scatter of neutron capture elements observed in low metallicity stars in the solar vicinity, compared to the small star tostar scatter observed for the α-elements. In fact, we have developed a stochastic chemical evolution model, in which the main assumption is a random formation of new stars, subject to the condition that the cumulative mass distribution follows a given initial mass function. With our model we are able to reproduce the different features of neutron capture elements and α-elements. The reason for this resides in the random birth of stars coupled with different stellar mass ranges from where α-elements and neutron capture elements originate. In particular, the site of production of α-elements is the whole range of the massive stars, whereas the mass range of production for neutron capture elements has an upper limit of 30 Mʘ . Finally, we test the prescriptions for neutron capture elements also for the dwarf spheroidal galaxies of the Local Group. We use a chemical evolution model already able to reproduce the abundances for α-elements in these systems. We conclude that the same prescriptions used for the Milky Way well reproduce the main features of neutron capture elements also in the dwarf spheroidal galaxies for which we have observational data. In dwarf spheroidal galaxies for which we do not have observational data we only give predictions. We predict that the chemical evolution of these elements in dwarf spheroidal galaxies is different from the evolution in the solar vicinity. This is due to their different histories of star formation relative to our Galaxy and indicates that dwarf spheroidal galaxies (we see nowadays) cannot be the building blocks of our Galaxy.

This thesis work is based on the study of the dynamical and chemical evolution of the interstellar medium (ISM) in Blue Compact Dwarf (BCD) galaxies as a consequence of single or multiple episodes of star formation. After single or multiple star bursts, stars release energy into the ISM through stellar winds and supernova explosions. This energy can drive a galactic wind and thus eject most of gas out of the parent galaxy. The development of galactic winds strongly depends on what fraction of the mechanical energy released by the supernova is available to thermalize the ISM and thus power the galactic wind (a parameter often called Thermalization efficiency). During the expansion of the Supernova Remnant, radiati ve losses can occur, thus most of the initial blast wave energy can be lost. This is a controversia! point, thus we dedicateci part of our work to the determination of the Thermalization efficiency and we found that this efficiency is low for Type II SNe, whereas we expect higher thermalization efficiencies in Type la SNe, because this kind of explosions occur in a medium already heated and diluted by the previous activity of Type II SNe. We then used both purely chemical evolution models and a 2-D hydrodynamical code, coupled with detailed chemical yields, originating from Supernovae of Type II and Type la and from single intermediate-mass stars, to simulate the dynamical and chemical evolution of this kind of objects. There are a lot of hydrodynamical simulations concerning the evolution of the ISM in dwarf galaxies after a starburst episode in the current literature (see e.g. MacLow & Ferrara 1999; D'Ercole & Brighenti 1999; Silich & Tenorio-Tagle 1998). These simulations generally agree on the fact that the ISM is rather robust and only a tiny fraction of the ISM is expelled as a consequence of the galactic wind. However, none of these works consider the effect of Type la, whose role is certainly fundamental in the late evolution of these objects. Given the importance of this kind of Supernovae for our models, a significant part of this work has been devoted to this topic. We found analytical solutions for the rate of Type la SNe under different star formation regimes and for the most popular progenitor models, and we applied these results on the study of the chemical evolution in the salar neighbourhood. We found that the best prescription to obtain SNela rates in agreement with the observations seems to be the so-called single-degenerate scenario. In this scenario, SNela are thought to originate from the thermonuclear deflagration of a White Dwarf reaching the Chandrasekhar mass after accretion of mass from a non-degenerate companion star. We found also that the typical time-scale for chemical enrichment from Type la SNe, often claimed in literature to be of the arder of l Gyr, strongly depends on the star formation history of the galaxy, ranging from 40 -50 Myr for BCD galaxies and 4 - 5 Gyr for spirai galaxies, like the Milky Way. We then performed numerica! hydrodynamical simulations of the evolution of the ISM in BCD galaxies as a consequence of single or multiple instantaneous star bursts. The aims of this kind of study are: • evaluate the impact of a single or multiple instantaneous star bursts on the dynamics of the ISM and study under what conditions a galactic wind could develop. • Follow the evolution, in space and time, of some chemical elements of particular astrophysical interest (namely H, He, C, N, O, Mg, Si, Fe). Most of chemical evolution models do not take into account a dynamical treatment, thus evaluating the mass of gas which escapes from the parent galaxy in a simplistic manner. With our model instead we are ab le t o follow the dynamical evolution of the concentration of several chemical elements, by taking into account stellar evolution and nucleo synthesis prescriptions as well as stellar lifetimes. Moreover, the mass of gas ejected from the galaxy in the form of the various chemical elements is computed in detail. • Apply these simulations to a galaxy model whose structural parameters are similar to IZw18, the most metal-poor galaxy locally known, in arder to put constraints on its age and its past star formation history. Main results of this kind of work can be summarized as follows: • As a consequence of the energy injected during the star formation activity ( either single or multiple instantaneous starbursts), a galactic wind develops. This galactic wind is differential, in the sense t ha t the newly formed metals are ejected more easily t han the pristine gas. In particular, metals produced by Type la supernovae are ejected more easily than those produced by SNeii, because this kind of explosions occurs in a medium already heated and diluted by the previous activity of Type II SNe. As a consequence of this, the [a/Fe] abundance ratios outside the galaxy are found to be lower than inside, an important prediction for galactic evolutionary models. • We are able to reproduce the chemical abundances found in literature for IZw18, either with a single burst or with two instantaneous bursts, separated by a quiescent period. The predicted age for the single-burst model is rv 30 Myr, but we found better agreement for models with two bursts of star formation, the first with an age of 300-500 Myr and the second with an age between 4 and 70 Myr, depending on the adopted IMF and nucleo synthesis prescriptions. A first burst with an age of 300 Myr and a second, more intense o ne, with an age of few Myr, is particularly appealing since it agrees with the derived ages from stellar population studies (Aloisi et al. 1999; Òstlin 2000) and with the spectral energy distribution of this galaxy (Mas-Hesse & Kunth 1999). • W e find t ha t the majority of metals are in a col d phase, especially for the single-burst model. This is due to the fact that most of energy produced by the supernovae is quickly radiated away, thus the hot hubble of gas created after the star burst (the so-called super bubble) evolves slowly, having time to significantly cool before break-aut. This result is the first of this kind, since in previous works i t was concluded that most of the metals should reside in a hot gas phase, virtually undetectable with the optical spectroscopy. Some recent studies of Color-Magnitude diagrams in BCD galaxies seem to indicate that these objects experienced star formation episodes of non-negligible duration. We are not able at the moment to simulate this kind of star formation regimes with our chemo-dynamical code, thus we ran simulations by means of a purely chemical evolution code. We assumed the hypothesis of the "Differential wind", namely, when a galactic wind develops, it expel mostly metals. The fraction of various chemical elements ejected by the parent galaxy are assumed in agreement with the results of our chemo-dynamical simulations. Main results of this kind of simulations are that models of bursts of star formation with duration shorter than 100 Myr do not develop a galactic wind. We then concentrate mainly on the be haviour of nitrogen. A large spread of N/0 as a function of 0/H is observed in a large sample of BCD galaxies, whereas for a small subsample of metal-poor BCD (i.e. with log(O/H) < -4.5) the N/0 abundance ratio seems to be almost constant (Izotov & Thuan 1999). W e can explain the spread in the N/ O vs. O /H diagram as due to different star formation efficiencies, which means also different wind efficiencies, different burst ages or differences in the burst durations. The nearly flat trend of N/0 at low 0/H (if true) could be explained by primary production of N in massive stars (a suggestion already introduced by Matteucci & Greggio 1986), or alternatively by a couple of bursts separated by a short quiescent period, or more extended burst, in which a differential wind develops and, in the following evolution, both the O abundance and the N/0 abundance ratio decrease with time.

The thesis work is focused on the analysis of the galaxy clusters ABCG 209, at zrv 0.2, which is characterized by a strong dynamical evolution. The data sample used is based mainly on new optical data (EMMI-NTT: B, V and R band images and MOS spectra), acquired in October 2001 at the European Southern Observatory in Chile. Archive optical data ( CFHR12k: B and R images), and X-ray (Chandra) and radio (VLA) observations are also analysed. The n1ain goal of this analysis is the investigation of the connection between internal cluster clynamics and star formation history, aimed at understanding the complex mechanisms of cluster formation and evolution. The internal dynamics of the cluster was studied through a spectroscopic survey of 112 cluster Inembers. The dynamical analysis has pointed out that ABCG 209 is characterized by a very high value of the line of sight velocity dispersion: av = 1250-1400 km s- 1that results in a virial mass of M = 1.6-2.2 x 1015 h-1 M0 within Rvir· A preferential SE-NW direction is indicated by: a) the presence of a velocity gradient in the velocity field; b) the elongation in the spatial distribution of colour-selected cluster members; c) the elongation of the X-ray contour levels in the Chandra image; d) the elongation of the cD galaxy. T h ere is evidence of substructure, as shown by i) significant cleviation of the velocity distribution from a Gaussian, with evidence for two secondary clumps at z = 0.199 and z = 0.215, which appear spatially segregated from the main cluster, ii) the Dressler & Schectman test and iii) the two-dimensional distribution of the colour-selected members shows a strong luminosity segregation: bright galaxies R < 19.5 are centered around the cD galaxy, while faint galaxies R > 19.5 show some clumps. The main one, Eastern with respect to the cD galaxy, is well coincident with the secondary X -ray peak. ' The study of the galaxy luminosity function (LF) in B, V and R bands has pointed out that ABCG 209 is a cluster with intermediate characteristic between a dynamically-evolved, rich clusters and clusters with central dominant galaxies having bright characteristic luminosities and shallow faint-end slopes and less evolved clusters, characterized by steep faint-end slopes, thus reconciling the asynnnetric properties of X-ray emission with the non flat-LF shape of irregular systen1s. This observational scenario suggests that ABCG 209 is undergoing a strong dynan1ical evolution with the Inerging of two or more subclumps along the SE-NW direction in a plane which is not parallel to the plane of sky. The merging might be in a more advanced status, where luminous galaxies trace the remnant of the core-halo structure of a pre-merging clump hosting the cD galaxy. The elongation and asymmetry of the galaxy distribution (of the X-ray emission) and the shape of the LFs show that ABCG 209 is not yet a fully relaxed system. The effect of cluster environment (as measured in terms of the local surface density of R < 23.0 galaxies) o n the global properties of the cluster galaxies is examined through the analysis of the L Fs, colour-magnitude relations, and average colours by using wide field (30'x42') B- and R-band images. The faint-end slope of the LF, a, shows a strong dependence on environment, becoming steeper at > 3a significance level from high- to low-density environments. The red sequence is found to be 0.022 ± 0.014 mag redder in the high-density region than for the intermediate-density region by fixing the slope. In contrast no correlation between the slope of the red sequence and environment was observed. Studying the effect of the cluster environment on galaxy star-formation, we find that the blue galaxy fraction decreases monotonically with density, in agreement with other studies. The observed trends of steepening of the faint-end slope, faintening of the characteristic luminosity, and increasing blue galaxy fraction, from high- to low-density environments, are manifestations of the morphology-density relation, where the fraction of early-type galaxies decreases smoothly and monotonically from the cluster core to the periphery, while the fraction of late-type galaxies increases in the same manner. The observed trends in the composite LF reflect this morphology-density relation: the galaxy population in the cluster core is dominateci by early-type galaxies and so the cornposite LF resembles that of this type of galaxy, with a shallow faint-end slope and a bright characteristic lun1inosity; whereas in lower density regions the fraction ~f late-type galaxies increases, and so the composite LF increasingly resembles that of the late-type, with a steep faint-end slope and a fainter characteristic magnitude. The analysis of the mean colour of luminous (R < 21) cluster galaxies as a function of their spatial position shows clearly the complex effects of the cluster environment and dynamics on their constituent galaxies. The reddest galaxies are concentrateci around the cD galaxy (main cluster) and a more diffuse region 5 arcmin to the north is coincident with the structure predicted from weak lensing analysis. The effect of the preferential SE-NW direction for ABCG 209 is apparent in the presence of bright blue galaxies near the cD galaxy perpendicular to the axis and hence unaffected by the cluster merger, and an extension of red galaxies to the SE which may indicate the irifall of galaxies into the cluster along a filament. This preferential SE-NW direction appears also related to the large-scale structure in which ABCG 209 is embedded, with two rich (Abell class R=3) clusters ABCG 222 at z = 0.211 and ABCG 223 at z = 0.2070 are located 1.5° (15M pc) to the NW along this preferential axis. Cluster dynamics and large-scale structure clearly have a strong influence on galaxy evolution, so we have performed a detailed study of spectroscopic properties of 102 luminous member galaxies. W e find five different galaxy types: i) passive evolving galaxies (E), which exhibit red colours and no emission lines, ii) emission line galaxies (ELG), which are blue and have prominent emission lines, iiiiv) strong Hc5 galaxies, that are characterized by the presence of strong Hc5 equivalent width and can be divided into blue (HDSblue) and red (HDSred), according to the break at 4000 A and the B-R colours, v) and finally anemie spirals (Ab-spirals), that have spectral properties equal to passive evolving galaxies, but are disk-dominateci systems. These different spectral classes are strongly segregated in the phase-space, as indicated by the two dimensionai Kolmogorov-Smirnov test. Passive evolving galaxies represent the I"V 74% of the cluster members. This population formed very early, during the initial collapse of the cluster. They lie mainly in high density regions and ha ve a velocity dispersion fully consistent with those of the whole cluster. This result is understandable in terms of cosmological models of structure formation, in which early-type galaxies form in the highestdensity regions corresponding to the cores of rich clusters. HDSred galaxies are distributed along the elongation of the cluster mainly in intermediate density regions and have a significant low velocity dispersion, suggesting that this population could be the remnant of an infalling group. According to the evolution models, the presence of a strong Hc5 absorption line in their spectra indicates that these galaxies have experimented a short starburst of star formation in the past few Gyrs. In the starburst nwdel [Hc5] and Dn(4000) declines on a timescale of I"V 2Gyr after the burst has ceased, irregardless of their SFR before the burst. Thus the fact that we see these galaxies, detecting [Hc5] > 3.0 A implies that the burst has occurred no more than 2 Gyr ago. HDSblue galaxies are found in intermediate density regions in a direction perpendicular to the cluster elongation, dose to the secondary peak of the X-ray flux andina region where the intracluster Inedium (ICM) is dense. Moreover galaxies belonging to this class seem to be aligned Thus the burst in the star formation seen in these galaxies could be due to the interaction with the hot dense ICM. They have high velocity dispersion. ELGs lie in low density regions and have high line-of-sight velocity dispersion. Both the spatial position and the velocity dispersion suggest that these two populations of galaxies have recently fallen into the cluster from the field. All these results support an evolutionary scenario in which ABCG 209 is characterized by a sum of two components: an old galaxy population, formed very earlier (zJ ;:G 3), and a younger population of infalling galaxies. Moreover this cluster may have experimented l or 2 Gyrs ago a merging with an infalling galaxy group, as indicated also by the previous dynamical analysis. This detailed study has showed clearly the importance to have multi-band data and to perform a n1ulti-directional analysis, in order to precisely characterize the different cluster components. It is now fundamental to extend this kind of analysis to other clusters at higher redshift and with different dynamical properties. To address the issue if clusters are generally young or old one needs to have measurements of subclustering properties of a large sample of clusters and at the same time it is fundamental to precisely characterize cluster components belonging to different structures and environments inside a single cluster.

The electronic structure and the magnetism of transition metal single atoms and magnetic impurities on metal surfaces have been studied by means of x-ray absorption dichroic techniques (XMCD, and magnetization measurements) in high magnetic fields and at low temperatures. The unique capabilities of XMCD to probe the both spin and orbital magnetic moments with element selectivity have been applied to diluted systems with low surface concentration (1012 atoms cm-2 ) down to the limit of isolated atoms. This kind of measurements have been applied to surface impurity systems only in the last few year thanks to the availability of high flux and brilliance synchrotron radiation sources. At the present these measurements represent the state of the art in the study of the magnetism of surface impurities. Results are reported for three different classes of systems: Mn and V impurities on the surface of alkali metals (Cs, K, N a); M n, V, Fe, Co single atoms on the Cu(1 00) surface an d Co single atoms on the Pt(111) surface. Furthermore results have been presented, regarding the magnetic moment distribution of a Mn monolayer coupled to the Fe(1 00) surface. Manganese on potassium has been shown to possess an atomic like magnetic d5 electronic configuration. This electronic configuration is particularly stable for small hybridization changes induced by alkali metals of increased electronic density. lf the hybridization is further increased as on Al(100) the spectroscopic features related to atomic multiplets are suppressed together with magnetism. Vanadium impurities on alkali metal surfaces also show atomic multiplet features. The electronic configuration is not an atomic-like configuration as d3 or d4 . The orbital moment is found to be small, less than 0.5 !-ls with respect to the purely atomic configurations and to be antiparallel to the spin. lt decreases along the alkali metals column (group 1A) of the periodic table, as the electronic density is increased, i.e. going from Cs to Li. This effect is related to an increased hybridization with the substrate due to the larger electronic density. The measured magnetic moments are of the order of 3)-ls, and cannot be explained with simple atomic parameters. Two possible mechanisms have been proposed to explain the partial quenching of the orbital moment, the effect of a weak crystal field and the effect of hybridization through the larger width of Friedel resonant bound state for the early elements of 3d transition metal series. Fe single atoms on Cu(100) surface have a large ot1f of plane magnetic anisotropy of 1.5 meV/atom and enhanced orbital magnetic moment with the respect to the bulk. The Cu substrate does not contribute considerably to the anisotropy as in the case of Co o n Pt. At the apposite Co single atoms on Cu(1 00) surface do not show any dichroism an d hence magnetism. This result ca n be explained on the basis of the Kondo effect, since Co single atoms have been found to be a Kondo system with T K=88 K. A comparison of the experimental data for V, Mn, Fe, Co impurities on Cu(100) with the calculated magnetic behavior of adatoms along the 3d series, gives an overall reasonable agreement, with important deviations for V and Co impurities. Co single atoms on the Pt(111) surface have an extraordinary large out of plane magnetic anisotropy of about 9.2 meV/atom due to the interplay between an enhanced orbital moment, consequence reduced coordination of the single atom, at the surface, and the effect of Pt hybridization, consequence of a d-d mixing between Co and Pt orbitals. This result is of particular relevance since this magnetic anisotropy is the highest measured, by now, for any system.

The present work is aimed at the study of the evolution of "cosmic quantities", such as the luminosity density, star formation rate density, supernova rate and metal production rate density. In some cases, these quantities are calculated per unitary comoving volume, hence they are called "densities". All these quantities are investigated in detail by means of chemical and spectrophotometric models for galaxies of different morphological types, namely ellipticals, spirals and irregulars. The chemical evolution models allow one to calculate the time evolution of the production rate of any chemical element, along with the element abundances in the interstellar gas and in the stars. Starting from the matter reprocessed by the stars and restored into the ISM through stellar winds and supernova explosions, it is possible to model the chemical enrichment of the universe in all of its components, namely the interstellar medium (ISM), the stars and the intergalactic medium (IGM). It is possible to calculate the time evolution of the abundances of several chemical species, such as C, O, N, Fe and Zn. In this thesis, the predicted abundances are compared to the observed ones: such task can provide precious hints on galactic evolution and element production both in the local and in the distant universe. The light emitted by stars in galaxies is studied by means of population synthesis models, which allow one to calculate the galaxy spectra, magnitudes and colors. The cosmic evolution of the luminosity density in various optical bands has been calculated starting from the luminosity function observed in the local universe, which allows us to determine the number of galaxies per comoving volume. We have assumed that the galaxy densities remain constant throughout the whole cosmic epoch and that the galaxies evolve as isolated systems only in luminosity. This scenario represents a pure luminosity evolution (PLE) one and it is equivalent to assume that merging has had no dominant effect on galaxy evolution throughout the Hubble time. Hence, we have developed a galaxy evolution picture which is opposed to the galaxy formation models based on the current "concordance" cosmological paradigm, which assume a cold dark matter universe whose energy is dominated by the cosmological constant. Such models indicate that the formation of large scale structures occurs in a hierarchical fashion: the first objects to form are the smallest ones, which then merge and build larger and larger objects. The baryonic matter, which builds up the galaxies, has the same behaviour of the dark matter, so that the first galaxies to form are the dwarf ones, which then by means of several merging processes give place to the giant galaxies. Thus, in contrast to our scenario, these models predict a strong galaxy number evolution. The strength of the PLE scenario described in this thesis is that, at variance with other approaches such as semi-analytical galaxy formation modeling or hydrodynamical/SPH cosmological simulations, it allows to study the evolution of the galaxy morphological types and the contributions brought by each type to cosmic star formation and element production in the universe. This point represents the first main novelty of my thesis. The PLE picture has also been used to predict the evolution of the cosmic Supernova rate which, thanks to the next generation space telescopes, will be soon measurable up to the highest redshifts, providing other fundamental hints to constrain galaxy formation and evolution. Furthermore, the comparison between the predictions obtained by the PLE scenario described in this thesis and the hierarchical represents an important benchmark to study galaxy evolution. When compared to the observations, it is possible to infer which picture is better to describe galaxy evolution. This is another aim of the present thesis, along with the suggestion of new observational strategies to disentangle between the hierarchical and the PLE picture if the current data do not allow us to perform such task. The predictions on the cosmic element production rates are then used to calculate the metal and baryonic budget in the local universe. Part of the elements synthesized so far remains locked up in long living stars and remnants, part is restored into the galactic ISM through supernova (SN) explosions and stellar winds while another fraction is expelled into the IGM through galactic winds and outflows. The formalism described in this thesis allows us a detailed computation of the element fractions present in each of the three main components of the local universe. The investigation of the amounts of different elements locked up in various phases (stars, IGM, ISM) and in different galaxies represents the second main novelty of this thesis. Finally, the chemical evolution models for galaxies of different morphological types allow us to perform an analysis of the chemical abundances measured in high-redshift objects. Of particular importance are the Damped-Lyman Alpha (DLA) systems, often referred to as the most likely progenitors of the local spiral and irregular galaxies. With the present work, it has been possible to study first the chemical evolution of the DLA population as a whole, in order to infer which morphological galactic types can be associated to DLAs and which are to be ruined out. Secondly, by focusing on individual systems and by reproducing at the same time as many observed abundances as possible, it has been possible to infer the star formation history of some single DLA systems and to have important hints on their ages. This kind of analysis, never performed so far, represents the third major novelty of this thesis. The present thesis is organized as follows: in chapter l a general overview on galaxy evolution is performed. The observational properties of the different galactic types are described, along with their interpretation and current theoretical approaches to galaxy formation and evolution. Particular emphasis is put on the most intriguing debate in galaxy evolution, namely: how and when did the formation of spheroids occurred in the universe occurred? Chapter 2 is a description of the chemical evolution models used in the present work: the main physical assumptions at the basis of the models for ellipticals, spiral and irregular galaxies and some results concerning the predictions which these models provide. Chapter 3 is a description of the spectro-photometric models used to study galaxy spectra, magnitudes and colors, with a comparison between results coming from different models and a description of the method used to investigate the dust extinction effects. In chapter 4 the results concerning the cosmic star formation history, the galaxy luminosity density and cosmic supernova rate are presented, along with a comparison between the predictions provided by the PLE model developed with this thesis and the hierarchical model by Mencietal. (2002). In chapter 5 we present our detailed calculations of the metal production rates. We perform a budget of the metals in the local universe and investigate the metal abundances in stars, ISM and IGM. In chapter 6 we apply our chemicai evoiution models to the study of damped Lyman alpha systems. Finally, in chapter 7 some conclusions are drawn.