Uncertainties in the production of iron-group nuclides in core-collapse supernovae from Monte Carlo variations of reaction rates

TL;DR

The problem addressed is the substantial Nuclear-physics uncertainties in predicting iron-group nucleosynthesis in core-collapse supernovae. The authors apply a Monte Carlo post-processing framework (PizBuin/MC-WinNet) to vary about 8,000 rates across $k=10^4$ iterations for 1D PUSH explosions of a $16\,M_{\odot}$ progenitor at multiple metallicities, tracking 615 nuclides up to $A=80$ and enforcing detailed balance for rate pairs. They find that most Fe-group production is NSE-dominated with modest uncertainties, but several nuclei show significant sensitivity to rate variations; key reactions identified via weighted Pearson correlations reduce uncertainties when removed (Level-2/3 analyses). The results yield a prioritized list of reactions affecting observables such as $^{44}{\rm Ti}$, $^{57}{\rm Ni}$, and $^{56}{\rm Co}$, informing experimental efforts to constrain nuclear inputs and providing a quantitative sensitivity map for explosive nucleosynthesis, while acknowledging limitations of 1D modelling and the need for future multidimensional validation.

Abstract

Core-collapse supernovae, occurring at the end of massive star evolution, produce heavy elements, including those in the iron peak. Although the explosion mechanism is not yet fully understood, theoretical models can reproduce optical observations and observed elemental abundances. However, many nuclear reaction rates involved in explosive nucleosynthesis have large uncertainties, impacting the reliability of abundance predictions. To address this, we have previously developed a Monte Carlo-based nucleosynthesis code that accounts for reaction rate uncertainties and has been applied to nucleosynthesis processes beyond iron. Our framework is also well suited for studying explosive nucleosynthesis in supernovae. In this paper, we investigate 1D explosion models using the "PUSH method", focusing on progenitors with varying metallicities and initial masses around $M_\mathrm{ZAMS} = 16 M_{\odot}$. Detailed post-process nucleosynthesis calculations and Monte Carlo analyses are used to explore the effects of reaction rate uncertainties and to identify key reaction rates in explosive nucleosynthesis. We find that many reactions have little impact on the production of iron-group nuclei, as these elements are primarily synthesized in the nuclear statistical equilibrium. However, we identify a few "key reactions" that significantly influence the production of radioactive nuclei, which may affect astrophysical observables. In particular, for the production of ${}^{44}$Ti, we confirm that several traditionally studied nuclear reactions have a strong impact. However, determining a single reaction rate is insufficient to draw a definitive conclusion.

Figures (7)

• Figure 1: The $615$ nuclei contained in the network (blue circles), overlaid on all nuclei within the drip lines (grey circles) as determined by the FRDM mass model 1995ADNDT..59..185M. Among them, the $90$ stable nuclei and $24$ selected radioactive species (see Table \ref{['tab:ri_nuc']}) are indicated by black and green squares, respectively.
• Figure 2: Weight factors of nuclei after decay for the models defined in Section \ref{['sec:models']} and summarised in Table \ref{['tab:models']}, shown for s16 (top), w16 (middle), and u16 (bottom). The left panels show stable nuclei and key radioactive nuclei with $A \lesssim 80$, while the right panels display results for 8 selected key radioactive nuclei. The final abundances of stable nuclei are taken from the values after all decays, while those of radioactive nuclei are taken at the end of the nucleosynthesis calculations ($t = 10^4~{\rm s}$). The order of adopted stable nuclides is p, d, ${}^{3,4}{\rm He}$, ${}^{7}{\rm Li}$, ${}^{10,11}{\rm B}$, ${}^{12,13}{\rm C}$, ${}^{14,15}{\rm N}$, ${}^{16,17,18}{\rm O}$, ${}^{19}{\rm F}$, ${}^{20,21,22}{\rm Ne}$, ${}^{23}{\rm Na}$, ${}^{24,25,26}{\rm Mg}$, ${}^{27}{\rm Al}$, ${}^{28,29,30}{\rm Si}$, ${}^{31}{\rm P}$, ${}^{32,33,34,36}{\rm S}$, ${}^{35,37}{\rm Cl}$, ${}^{36,38,40}{\rm Ar}$, ${}^{39,40,41}{\rm K}$, ${}^{40,42,43,44,46,48}{\rm Ca}$, ${}^{45}{\rm Sc}$, ${}^{46,47,48,49,50}{\rm Ti}$, ${}^{50,51}{\rm V}$, ${}^{50,52,53,54}{\rm Cr}$, ${}^{55}{\rm Mn}$, ${}^{54,56,57,58}{\rm Fe}$, ${}^{59}{\rm Co}$, ${}^{58,60,61,62,64}{\rm Ni}$, ${}^{63,65}{\rm Cu}$, ${}^{64,66,67,68,70}{\rm Zn}$, ${}^{69,71}{\rm Ga}$, ${}^{70,72,73,74,76}{\rm Ge}$, ${}^{75}{\rm As}$, and ${}^{79}{\rm Br}$.
• Figure 3: Nucleosynthesis yields of explosive nucleosynthesis at $10^4$ s after explosion for $16 M_\odot$ stellar models. The s16 and w16 progenitors have solar abundance, whereas u16 is based on a metal poor star (see Table \ref{['tab:models']}). Vertical gray bands labeled with zone numbers indicate the boundaries of the Si-, O-, and Ne-burning layers.
• Figure 4: Uncertainty ranges (90% intervals around the peak value $y_{\mathrm{peak}}$) obtained from the MC variations for explosion models: s16 (top), w16 (middle), and u16 (bottom), shown for Level 1–3 MC runs. Nuclei marked with squares indicate key products listed in Tables \ref{['tab:key_s16']}, \ref{['tab:key_w16']}, and \ref{['tab:key_u16']}. Note that the selection of nuclei is slightly different for each model.
• Figure 5: Uncertainty ranges obtained from the MC variations for s16 (left), w16 (middle), and u16 (right). Same as Figure \ref{['fig:y_unc']}, but for selected nuclei.