Common Occurrence of Explosive Hydrogen Burning in Type II Supernovae

As discussed by Liu et al. (2017a), the most likely stellar sources of AB1 grains are SNe that experienced explosive H burning in their outer layers, based on the fact that AB1 grains show both explosive H burning and n-capture isotopic signatures. In addition, the explosive H burning isotopic signatures of AB1 grains, such as elevated ¹³C/¹²C, ¹⁵N/¹⁴N, and inferred ²⁶Al/²⁷ Al ratios, depend strongly on the peak temperature achieved and the amount of H present during the explosion (Liu et al. 2016, 2017a). The fact that AB1 grains define a tight correlation between their ¹⁴N/¹⁵N and inferred ²⁶Al/²⁷Al ratios strongly indicates that they share a common stellar origin, namely SNe with explosive H burning. In contrast, another suggested stellar source for AB grains, born-again AGB stars like Sakurai's object (Fujiya et al. 2013), are less likely, based on the fact that the state-of-the-art models for Sakurai's object predict lower-than-solar ¹⁵N/¹⁴N ratios for the ¹³C-rich region of a born-again AGB star (Herwig et al. 2011), while, in contrast, AB1 grains are characterized by ¹⁵N/¹⁴N ratios in excess of the solar value. However, it is noteworthy that the H entrainment rate during the post-AGB phase depends on the initial progenitor stellar mass, which can lead to quite different N isotope ratios in the ¹³C-rich region of a post-AGB star, and more modeling efforts are therefore needed for comparison with presolar grain data.

The heavy-element isotope data provide further evidence against a born-again AGB stellar origin of AB1 grains. The model calculations by Herwig et al. (2011) for Sakurai's object predict that the intermediate n-capture process (i-process, Cowan & Rose 1977) takes place in born-again AGB stars at enhanced neutron densities (∼10¹⁵ cm⁻¹) relative to those for the s-process (∼10⁷ cm⁻¹) in AGB stars, resulting in an anomalous pattern of elemental abundances (relative to s-process elemental patterns) observed on the surface of the Sakurai's object (Asplund et al. 1999). Specifically, the Herwig et al. (2011) models predict extremely anomalous Sr, Mo, and Ba isotopic compositions for born-again AGB stars (model predictions for Ba isotopes are shown in Figure 10 of Liu et al. 2014, and the corresponding predictions for Mo isotopes are given in Figure 1(b)), e.g., large positive δ^97,98,100Mo values, which are inconsistent with the s-process isotopic signatures seen in the AB1 grains from this study. Thus, an origin in born-again AGB stars seems unlikely, but the effect of the initial progenitor stellar mass on Mo isotope production in post-AGB stars needs to be investigated for comparison with the grain data. In the following sections, we will discuss the Sr, Mo, and Ba isotopic signatures of AB1 grains in the context of SN nucleosynthesis.

In Figure 2, we show the model predictions for the He/C and He/N zones of a 25 M_⊙, Z_⊙ (Anders & Grevesse 1989) SN with two vastly different explosion energies. We have adopted the simulations of Rauscher et al. (2002) for the pre-SN stellar compositions. For reference, Rauscher et al. (2002) adopted an explosion energy of 10⁵¹ erg for the same pre-SN stellar model in their calculations. It is obvious from Figure 2 that the bottom of the He/C zone (at 7 M_⊙) is characterized by large excesses in the most neutron-rich isotopes, including ⁵⁰Ti, ⁸⁸Sr, ¹⁰⁰Mo, and ¹³⁸Ba as a result of the so-called neutron-burst process powered by the ²²Ne(α,n)²⁵Mg reaction during SN explosions (Meyer et al. 2000). In contrast, the outer region of the He/C zone (shaded area in Figure 2) remains unmodified by the neutron-burst process, and preserves the pre-SN n-capture signatures, because the peak temperature reached during the explosion is too low to activate the ²²Ne(α,n)²⁵Mg reaction in this region. With increasing explosion energy, the neutron-burst process occurs more efficiently, resulting in more anomalous isotopic signatures at the bottom of the He/C zone and affects the nucleosynthetic signatures of a larger fraction of the He/C zone. Explosive n-capture isotopic signatures were previously found in the main X subgroup, X1 grains (Pellin et al. 2000; Stephan et al. 2018), and one of the five AB1 grains studied by Savina et al. (2003). However, none of the AB1 grains in this study showed such neutron-burst isotopic signatures. The isotopic compositions of the isotopically anomalous AB1 grains are distinctly different from the predictions for the explosive n-capture process (Figure 2), and are more similar to the pre-SN signatures predicted by the model calculations except for δ⁴⁹Ti (Figure 1(c)).

Zoom In Zoom Out Reset image size

Pignatari et al. (2018, hereafter P18) recently conducted a detailed analysis of n-capture nucleosynthesis in outer SN zones in the context of Zr and Mo isotopic signatures of presolar X1 grains. The model calculations shown in Figure 2 agree generally with the P18 model calculations for s-process elements, except for the fact that we did not find the formation of a Si/C zone even by increasing the explosion energy to 5 × 10⁵¹ erg. In contrast, Pignatari et al. (2015, hereafter P15) were able to produce such a Si/C zone at the bottom of the He/C zone by adopting explosion energies of (4–7) × 10⁵¹ erg for a 25 M_⊙ SN. This discrepancy is related to differences in the details of the pre-SN evolutionary simulations adopted in the two sets of models. Nevertheless, our calculations confirm the general features of the P18 models (e.g., Figure 3 of P18) that the inner He/C zone is significantly modified by the neutron-burst process, while the outer region (shaded area in Figure 2, zone 4 of P18) remains unmodified. In addition, similarly to the P18 models, our models also predict large ¹⁰⁰Mo excesses and higher enrichments in ⁹⁴Mo relative to ⁹²Mo for the neutron-burst process, which are not observed in X1 SiC grains that sampled neutron-burst materials from their parent SNe (Pellin et al. 2000) and thus imply problems related to the productions of ⁹⁴Mo and ¹⁰⁰Mo in the current SN model calculations, as pointed out by P18.

P15 investigated the effect of H ingestion (up to 1.2% by mass) into the He/C zone in the pre-SN phase, on explosive nucleosynthetic isotopic signatures during SN explosions (see P15 for model details). The P15 models (25T and 25d models) showed that large amounts of ¹³C, ¹⁵N, and ²⁶Al are produced as a result of explosive H burning at the bottom of the He/C zone. Based on the P15 models, Liu et al. (2017a) were able to reproduce the ¹⁴N/¹⁵N and ²⁶Al/²⁷Al correlation seen in AB1 grains.

The pre-SN n-capture isotopic compositions of AB1 grains confirm the P15 model prediction that the presence of H at the bottom of the He/C zone during SN explosions suppresses the neutron-burst process there. We also further tested this effect by adding 1% H in the He/C zone in our pre-SN models, and verified the result of P15 that this amount of H is sufficient to completely deactivate the neutron-burst process in the He/C zone during SN explosions in our models by adopting energies from 0.75 × 10⁵¹ erg to 5 × 10⁵¹ erg. In turn, compared to the He/C zone with varying neutron-capture isotopic signatures in Figure 2, the H-ingested He/C zone in all the models after SN explosions exhibits the pre-SN isotopic signature (Figure 1(c)) across the zone. This suppression of the explosive n-capture process is due to the ²²Ne(α,n)²⁵Mg reaction becoming deactivated as a result of the competing ²²Ne(p,γ)²³Na reaction powered by protons (P15). Thus, material in the He/C zone that experiences explosive H burning is naturally expected to show pre-SN n-capture isotopic signatures, as seen in AB1 grains.

Heterogeneous H mixing was previously proposed to explain the AB1 grain data, based on the observation that ³²S and ⁵⁰Ti excesses and explosive H burning signatures were concomitantly incorporated into AB1 grains (Liu et al. 2016, 2017a). This is mainly because current SN models predict that ³²S and ⁵⁰Ti can only be abundantly produced by the neutron-burst process at the bottom of the He/C zone during SN explosions and that the neutron-burst process becomes completely suppressed by the presence of H during SN explosions. Since the overall Sr, Mo, and Ba isotopic patterns of the AB1 grains are consistent with pre-SN rather than explosive, n-capture isotopic signatures, the mismatch of the current SN model predictions with the grain data points to a problem in the current model predictions for the pre-SN n-capture process. In turn, the coexisting pre-SN n-capture and explosive H burning isotopic signatures of AB1 grains can be well explained by the P15 models as a natural consequence of explosive H burning without invoking heterogeneous H mixing.

Specifically, the AB1 grain data suggest that the pre-SN n-capture process occurred with enhanced neutron densities with respect to the current model predictions. Specifically, the lower ⁴⁹Ti/⁵⁰Ti ratios of the AB1 grains compared to the pre-SN n-capture model predictions (Figure 1(c); also see Liu et al. 2017a) cannot be due to errors in n-capture cross sections, since those for ⁴⁹Ti and ⁵⁰Ti are known to ∼10% accuracy (Karlsruhe Astrophysical Database of Nucleosynthesis in Stars version v1.0, Dillmann et al. 2014).¹¹ Rather, the data suggest that the pre-SN n-capture process occurred with enhanced neutron densities with respect to the current model predictions, since the predicted ⁵⁰Ti/⁴⁹Ti ratio increases with increasing neutron density (e.g., Zinner et al. 2007). Moreover, the most anomalous Mo isotopic pattern observed in an AB1 grain, M1-A7-G987, cannot be explained by the current model predictions nor can this grain's large ³²S and ⁵⁰Ti excesses. It is also worth mentioning that although most of the AB1 grain data can be matched by the current pre-SN model predictions in Figure 1(c), mixing of H-envelope material that has normal Ti, Sr, Mo, and Ba isotopic compositions with explosive H burning product from the He/C zone is required to explain the Al-N isotope trend observed in AB1 grains (Liu et al. 2017a). This is likely to further dilute the n-process isotopic signatures in the precursor SN mixtures of AB1 grains, and a stronger n-capture process is thus generally required to explain the AB1 grain data.

Our calculations show that, in addition to the bottom of the He/C zone, explosive H burning can also occur at the top of the He/N zone at relatively low temperatures, corresponding to the ¹⁵N-rich spike previously reported by Nittler (2008) and Lin et al. (2010). This burning occurs as a result of pre-SN mixing of a large amount of H (up to 40% by mass, Figure 3(a)) from the adjacent H envelope in the pre-SN phase (Figure 3) due to semiconvection and/or overshoot (Canuto 2000). In the 5 × 10⁵¹ erg case, the ¹²C/¹³C and ¹⁴N/¹⁵N ratios are predicted to reach minima of 1.9 and 0.01, respectively, and thus can well account for the C and N isotope ratios of all AB1 grains and most of the putative nova grains in the literature (e.g., Liu et al. 2016, 2017a). Figure 4 further illustrates the trends of different ratios varying with increasing explosion energy, and, of these, the ¹⁴N/¹⁵N ratio shows the strongest dependence on the explosion energy. Both the P15 H-ingestion models and the model predictions shown here strongly indicate that explosive H burning can occur at multiple locations within the He/N and He/C zones depending on the H mixing mechanism and the peak temperature achieved during an SN explosion. Also, the predicted N and Al isotope ratios in Figure 4 are similar to the explosive H burning endmember inferred by the N-Al isotope trend of AB1 grains (e.g., Figure 1 of Liu et al. 2017a), which points to the general similarity of explosive H burning products (He/N versus He/C) within SNe. Since the pre-SN isotopic signatures of heavy elements are predicted to be normal in the He/N zone, the normal Mo isotopic composition of the ∼30% "uncontaminated" AB1 grains can also be explained. Note that if explosive H burning only occurred in the He/N zone of the progenitor star of an SN SiC grain, the He/N zone material alone probably cannot provide a C-rich environment for SiC grains to condense, because, even in the 5 × 10⁵¹ erg case, the C/O ratio of the He/N zone barely reaches unity (Figure 4). In addition, mixing with the O-rich material from the H envelope as required to explain the N-Al isotope trend of AB1 grains is likely to further reduce the C/O ratio. In this case, material from the He/C zone needs to be mixed with the He/N zone material to raise the C/O ratio. Thus, if AB1 grains with normal heavy-element isotopic compositions inherited their H-ingestion signatures from the He/N zone, it requires their parent SNe to have low explosion energies (e.g., less than 5 × 10⁵¹ erg for a 25 M_⊙ progenitor star) so that normal heavy-element isotopic composition can be preserved in the outer part of the He/C zone (e.g., Figure 2).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The occurrence of explosive H burning in SNe is also supported by the isotopic compositions of X grains. Previous studies have shown that explosive H burning material from the ¹⁵N-rich spike in the He/N zone and/or from the H-ingested region in the He/C zone is required to explain the C and N isotopic ratios of X grains (Lin et al. 2010; Hoppe et al. 2018). Thus, SN X grains suggest that explosive H burning occurred in the outer C-rich He/C and/or He/N zones of their parent SNe, as was also argued for the progenitor SNe of C2 and AB1 grains (Liu et al. 2016, 2017a), implying that it is a common SN phenomenon.

Evidence that X1 and X2 grains have different n-capture isotopic signatures has recently been reported. While X1 grains exhibit clear neutron-burst signatures (Pellin et al. 2000; Stephan et al. 2018), pre-SN signatures were found in the only two X2 grains that have been studied for their heavy-element isotopic compositions so far (Stephan et al. 2018). As mentioned above, the C and N isotope ratios of X grains strongly suggest that explosive H burning generally occurs in the SN C-rich region. Thus, the pre-SN n-capture isotopic compositions of both X2 and AB1 grains are consistent with the P15 model prediction that explosive H burning deactivates the neutron-burst process in the He/C zone. The presence of neutron-burst signatures in X1 grains, however, contradicts this picture. While heterogeneous H ingestion into the He/C zone is one possible explanation, an alternative is that in the parent SNe of X1 grains, explosive H burning mainly occurred in the He/N zone, instead of in the He/C zone where the neutron-burst process takes place.