Switching on the first light in the Universe

摘要

Star formation (SF) is still an unsolved problem in astrophysics.Gas cooling is the principal mechanism leading tothe condensation of gas and consequently to star and structureformation.In a metal-free environment, the main available coolants are H, He,H$_2$ and HD; once the gas is enriched with metals, these alsobecome important.In this work, in order to properly determine the SF in the earlyUniverse, we compute fine-structure transition metal cooling andimplement and test molecular chemistry.Moreover, we investigate its redshift ($z$) evolutionand compare different modeling running very high-resolution,three-dimensional, N-body/hydrodynamic simulations includingnon-equilibrium, atomic and molecular chemistry, SFprescriptions and feedback effects.We also study how the primordial SF changes accordingly todifferent semi-analytical approaches, cosmological parameters, initialset-ups and critical metallicity ($Z_{crit}$) for the transition froma metal-free SF regime to a standard enriched one.Our main findings are:the H$_2$ molecule is the most relevant coolant in early times;inclusion of HD cooling results in a $sim 10%-20%$ higher gasclumping;metal cooling at low temperatures can have a significant impact on theformation and evolution of first objects;typical numerical ``sub-grid'' models fail in following the coolingof primordial gas and predict too early SF ($zsim 30$);considering molecular cooling, we get a postponed epoch ($zsim 15$)for the same initial conditions;rare, high-density peak can host SF even at $zgtrsim 40$;metal-free SF regime is completely negligiblewith respect to the global SF rate, because of the veryshort first star life-times;it has some relevance only for $Delta zsimeq 1$ (at $zsim 16$);primordial pollution up to $sim 10^{-3},Z_odot$ootnote[1]{The solar metallicity is $Z_odotsimeq 0.02$.}, or higher, isextremely rapid and allows for a very fast transition to standard SFregimes;the different SF rates and metal enrichment got fordifferent $Z_{crit}$ are well distinguishable and span about one orderof magnitude for $Z_{crit}/Z_odot in [10^{-6}, 10^{-3}]$.Given the importance of the initial stellar mass distributionfunction, an analytical model describing its possible derivation fromturbulent dissipation is presented.