2002ApJ...567..515Q

We determine the yields of the elements from Na to Ni for Type II supernovae (SNe II) and the yield patterns of the same elements for Type Ia supernovae (SNe Ia) and very massive (≳100 M_☉) stars (VMSs) using a phenomenological model of stellar nucleosynthesis and the data on a number of stars with -4≲[Fe/H]≲-3, a single star with [Fe/H]=-2.04, and the Sun. We consider that there are two distinct kinds of SNe II: the high-frequency SNe II(H) and the low-frequency SNe II(L). We also consider that VMSs were the dominant first-generation stars formed from big bang debris. The yield patterns of Na to Ni for SNe II(H), II(L), and Ia and VMSs appear to be well defined. It is found that SNe II(H) produce almost none of these elements; that SNe II(L) can account for the entire solar inventory of Na, Mg, Si, Ca, Ti, and V; and that compared with SNe II(L), VMSs underproduce Na, Al, V, Cr, and Mn, overproduce Co, but otherwise have an almost identical yield pattern. A comparison is made between the yield patterns determined here from the observational data and those calculated from ab initio models of nucleosynthesis in SNe II and VMSs. We show that the evolution of the ``heavy'' elements in the universe relative to Fe involves three distinct stages. The earliest stage is in the domain of [Fe/H]←3 and is governed by VMS activities with some small contributions from SNe II, all of which are dispersed in a dilution mass of M^VMS_dil∼10⁶-10⁷ M_☉. The beginning of the second stage is marked by the cessation of VMS activities and the onset of major formation of normal stars (with masses of ∼1-60 M_☉) at [Fe/H]~-3. The cessation of VMS activities causes the dilution mass for SN II contributions to drop sharply to M^SNII_dil~3x10⁴ M_☉. The subsequent quasi-continuous chemical evolution until [Fe/H]~-1 is governed by SNe II(H), which produce mainly the heavy r-process elements above Ba, and SNe II(L), which produce essentially all the other elements. The third stage starts with the onset of SN Ia contributions to mainly the Fe group elements at [Fe/H]~-1. The domain of [Fe/H]>-1 is then governed by contributions from SNe II(H), II(L), and Ia and low-mass stars. It is shown that the abundances of non-neutron capture elements in stars with [Fe/H]≤0 and those of r-process elements in stars with [Fe/H]←1 can be well represented by the sum of the distinct components in the phenomenological model. The proposed evolutionary sequence is directly related to the problems of early aggregation and dispersion of baryonic matter and to the onset of formation and chemical evolution of galaxies. It is argued that the prompt inventory governed by VMS contributions should represent the typical composition of dispersed baryonic matter in the universe, and that normal galactic evolution should begin in matter with [Fe/H]~3.