Nucleosynthesis of Elements Beyond Fe in C-O Shell Mergers

Abstract

Carbon–oxygen (C–O) shell mergers in the final evolutionary stages of massive stars play a critical role in shaping the pre-supernova structure and the resulting nucleosynthesis. In this work, we investigate the impact of such a merger on the production of elements beyond the Iron peak, focusing on an extremely metal-poor ($[\mathrm{Fe}/\mathrm{H}]=-5$) rotating 15 ${\mathrm{M}}_{\odot }$ stellar model. The results show that the merger favors the synthesis of weak s-process seeds and light p-nuclei, such as ⁸⁸Sr, ⁹⁴Mo, and ⁹⁸Ru, via photodisintegration of heavier nuclei previously produced by rotational-induced nucleosynthesis. By simulating the subsequent core-collapse supernova explosion with a thermal bomb approach, we demonstrate that these chemical signatures are largely preserved, as the expanded structure of the merged shells significantly modifies the impact of the shock wave. These findings suggest that C–O shell mergers in early-generation stars could provide a primary-like source for intermediate and heavy elements, with important implications for the chemical evolution of the early Universe.

1. Introduction

Massive stars ($M\gtrsim 9$–10 ${\mathrm{M}}_{\odot }$[1,2]) in their advanced evolutionary stages are governed by a non-linear interplay between nuclear burning, neutrino cooling, and the growth of convective regions. In stars less massive than ∼25 ${\mathrm{M}}_{\odot }$, efficient O-burning shells often extend into the outer C- and Ne-rich layers, resulting in the creation of a single extended mixed zone during the final days prior to the core-collapse supernova (CCSN). This occurrence, known as a C–O shell merger, has significant implications for both the nucleosynthesis and the innermost structure of the star leading up to and during the CCSN stage [3,4,5,6,7].

The C–O shell merger leaves a peculiar signature in the chemical composition of the star that is preserved even after the CCSN, due to the extended radius of the newly formed mixed region [8]. This signature consists of several components: an enhanced production of the elements synthesized by O-burning (Si, S, Ar, Ca), and the production of odd-Z elements (P, Cl, K, Sc) via radiative proton captures during SPAr burning [4,7,9]. Radioactive nuclei such as ³⁶Cl, ⁴⁰K, ⁴¹Ca, and ⁴⁴Ti can be efficiently produced as well ([10,11,12,13], and references therein). Furthermore, efficient photodisintegrations at the base of the O shell ($T\sim$ 2.5–2.8 GK) can activate the “cold” $\gamma$–process and produce the p–nuclei with $A>110$ [4,14,15].

In this paper, we discuss how C–O shell mergers affect the nucleosynthesis of elements beyond the Fe-peak in the case of a progenitor enriched in s-process elements.

2. Heavy Nuclei Nucleosynthesis in a C-O Shell Merger

We explore the heavy nuclei nucleosynthesis during a C–O shell merger in a 15 ${\mathrm{M}}_{\odot }$ stellar model at extremely low metallicity ($[\mathrm{Fe}/\mathrm{H}]=-5$), with an initial equatorial rotation velocity of ${\mathrm{v}}_{\mathrm{ini}}=600\mathrm{km}{\mathrm{s}}^{-1}$ from [6]. The model is calculated with the FRANEC code [16,17], in which the structure, transport, and nucleosynthesis equations are solved simultaneously. The nuclear network is optimized to describe the s-process nucleosynthesis up to a neutron density of the order of ∼${10}^{14}$ cm⁻³ and includes 524 isotopes and more than 3000 reactions. We note, however, that this configuration lacks the resolution required to fully resolve the reaction flows of the $\gamma$-process nucleosynthesis. Therefore, our analysis regarding this channel remains preliminary, leaving a detailed investigation for future studies. Notwithstanding this limitation, this model is particularly suitable to discuss how the nuclei beyond Fe behave in a merger, as the fast rotation allowed for an efficient s-process nucleosynthesis that produced a chemical composition enriched in both the main and heavy components. This result stems from two main features: the enhanced production of neutrons favored by rotational instabilities and the very low neutron-to-seed ratio due to the low metallicity [6,17,18,19]. The main features of this model, as well as the nuclear network construction and the s-process nucleosynthesis across the other models in the set, are discussed in detail in the original paper [6].

It is worth noting, however, that the key conditions (chemical composition, temperature, and density in the O shell) in this model are comparable to those found in other models with initial masses between roughly 12 and 30 ${\mathrm{M}}_{\odot }$ across different sets [2,4,17,20,21]. Therefore, this model is suitable to discuss the general properties of in-shell O-burning before and after the C–O shell merger.

Figure 1 shows the evolution of the convective zones and the major fuels in the considered model via a Kippenhahn diagram. Following Si exhaustion, the stellar core begins its final contraction toward collapse. This phase lasts approximately two days and raises the temperature at the base of the convective O-burning shell, which develops at ∼2 ${\mathrm{M}}_{\odot }$ about ${10}^{-3}$ yr before the core collapse. As the core contracts, the bottom of the shell recedes in radius while the layers above expand due to efficient convection, allowing the convective zone to extend outward into the surrounding Ne, C, and O-rich layers (purple and light-blue zones in the Figure). Once the outer boundary of the convective zone starts mixing in these materials, a convective-reactive event is triggered. The ingested nuclei (C, Ne, and extra O) provide fresh fuel, generating the energy necessary to sustain further expansion and promote additional ingestion from the surrounding layers. Consequently, the O shell progresses in mass until it merges with the C convective shell a little before ${10}^{-4}$ yr from the collapse, reaching an extent of ∼4 ${\mathrm{M}}_{\odot }$. This C–O shell merger eventually freezes once an active Si-burning shell forms outside the Fe core, which accelerates the final evolutionary stages of the massive star.

Let us now focus on the internal chemical composition of the star before and after the C–O shell merger and how this event impacts the tracers of the weak, main, and heavy s-process components: ⁸⁸Sr, ¹³⁸Ba, and ²⁰⁸Pb. Figure 2 shows the internal chemical composition before (dashed lines) and after (solid lines) the occurrence of the C–O shell merger. As the Figure shows, before the merger, the zone between ∼2.3–4.2 ${\mathrm{M}}_{\odot }$ contains the ashes of both central He burning and C shell burning. In this region, ¹³⁸Ba is the most abundant among the three s-process tracers, followed by ⁸⁸Sr and ²⁰⁸Pb. This is due to the low initial Fe (with a mass fraction of ∼${10}^{-8}$) and the large amount of ²²Ne (with a mass fraction of ∼7 $\times {10}^{-3}$) produced during He burning by the rotational mixing, which allowed for a very large neutron-to-seed ratio that favored the production of the main s-process component. Below this region, in the O burning shell (∼1.8–2.3 ${\mathrm{M}}_{\odot }$) efficient photodisintegrations have already started destroying ¹³⁸Ba and ²⁰⁸Pb, while slightly contributing to the abundance of ⁸⁸Sr and two typical p-process nuclei, ⁹⁴Mo and ⁹⁸Ru [22]. The shell merger reprocesses the material from the bottom of the former O shell (∼1.8 ${\mathrm{M}}_{\odot }$) up to the zone containing the ashes of the central He burning (∼3.8 ${\mathrm{M}}_{\odot }$). It is clearly visible that the merger mixes the heavy nuclei (¹³⁸Ba and ²⁰⁸Pb) inward, towards higher temperatures, leading to their destruction. At the same time, ⁸⁸Sr, ⁹⁴Mo, and ⁹⁸Ru are produced by the photodisintegrations at the bottom of the O shell and mixed outward. They are effectively preserved from further reprocessing because their nuclear destruction timescales are much longer than the convective turnover timescale (which is of the order of 60–70 s).

The internal chemical composition at the pre-supernova stage is subsequently modified by the passage of the supernova shock wave. In the present model, the explosion is simulated as a thermal bomb using the Lagrangian hydrodynamic code HYPERION [23]. Specifically, the explosion is triggered by depositing a minimum amount of thermal energy at a mass coordinate of $0.8$ ${\mathrm{M}}_{\odot }$ (well within the Fe core), ensuring the full ejection of the mantle. The remnant mass is then determined by requiring the ejection of $0.07$ ${\mathrm{M}}_{\odot }$ of ⁵⁶Ni. While different collapse and explosion treatments can affect the explodability or alter the peak temperature of the shock and the extent of the explosive burning regions [12,24], the impact on a progenitor with a C–O merger is minimal. This is because the shock enters the expanded mixed region very early, and the peak temperature scales as $T_{peak}\propto R_{*}^{-3/4}$, where $R_{*}$ is the radial coordinate. Figure 3 shows the chemical composition before (dashed lines) and after (solid lines) the core-collapse supernova explosion. As is well established, s-process nuclei are destroyed by most explosive burning stages up to explosive Ne burning ($T\lesssim 3$ GK). As shown in the Figure, some ⁸⁸Sr, along with the p-nuclei ⁹⁴Mo and ⁹⁸Ru [25,26], are produced by photodisintegrations at ∼2.4 ${\mathrm{M}}_{\odot }$, whereas ¹³⁸Ba and ²⁰⁸Pb are destroyed by the shock up to $2.5$–$2.6$ ${\mathrm{M}}_{\odot }$.

3. Discussion and Conclusions

Nucleosynthesis in C–O shell mergers has been a subject of discussion for several years [3,7,27]. It represents a particularly powerful mechanism because, for many components, it involves primary production (as in the case of the elements between Si and Sc). Regarding the s- and $\gamma$-processes, these are typically considered secondary processes, as they depend strongly on the availability of seeds (Fe for the s-process, and heavy s- and r-nuclei for the $\gamma$-process). However, while remaining secondary in nature, we have shown that at low metallicity, the rotationally-enhanced s-process production combined with a C–O shell merger can lead to a significant rearrangement of the stellar material. This favors the weak component (⁸⁸Sr) and the lighter, more neutron-rich p-nuclei, such as ⁹⁴Mo and ⁹⁸Ru. This result aligns with the findings of Ref. [28], which suggested that rotating models can produce $\gamma$-process seeds in situ and synthesize p-nuclei during the supernova stage. In the case of C–O shell mergers, this process is almost independent from the explosion; furthermore, the reprocessing of s-process-rich material in extremely metal-poor stars could have a significant impact on galactic chemical evolution of p-nuclei [29].

While it is possible to identify tracers of C-O shell mergers in extremely metal poor stars among intermediate-mass elements, such as K and Sc [7], this does not hold for heavy elements. The merger modifies the Sr/Ba ratio, but this signature is degenerate with the contribution of other s- and neutron capture process sources. Thus, single-star yields alone are insufficient for a direct comparison with observations. Future galactic chemical evolution modeling is necessary to determine the actual contribution of these mergers to the chemical enrichment of the galaxy.

At low metallicity, rotation is the mechanism responsible for the production of a significant s-process distribution from the available Fe. Moreover, this occurs only upon exceeding a rotational velocity threshold that depends on the initial mass, as demonstrated in Ref. [6]. As metallicity increases, the pristine chemical composition of the star reflects contributions from an increasing number of stellar generations, eventually reaching the solar distribution, where Sr exceeds Ba. Consequently, the impact of C–O mergers evolves with metallicity: regarding s-nuclei, photodisintegrations tend to favor the destruction of the weak component rather than its production at the expense of the main and heavy components. Regarding the $\gamma$-process, we simply note that varying the metallicity results in different s-process distributions, which act as seeds for both the cold $\gamma$-process occurring in C–O mergers and the classic $\gamma$-process in core-collapse supernovae. As discussed in Section 2, a comprehensive analysis of the $\gamma$-process nucleosynthesis requires a more extensive reaction network than that employed here and, consequently, we defer a detailed investigation of this process in fast-rotating, low-metallicity models to future work. Finally, we note that in population III stars, at zero low metallicity, the lack of Fe seeds hinders the production of heavy nuclei through neutron captures, rendering the impact of C-O mergers negligible regarding the production of both s- and p-nuclei.