TENDL-astro: a new nuclear data set for astrophysics interest

TL;DR

The paper addresses the challenge of uncertain nuclear data in astrophysical models by building TENDL-astro, a comprehensive TALYS-based database that computes cross sections, reaction rates, MACS at $kT=30$ keV, and the partition function $G(T)$ for more than $8000$ nuclides. It systematically varies nine reaction models to quantify uncertainties, producing nominal values plus uncertainty ranges across 480 (non-fissile) or 960 (fissile) model combinations, and offers a recommended model set along with 10 preferred sets. The work demonstrates that model choices can cause large differences, especially for nuclei far from stability, and provides detailed MACS comparisons to KADoNiS showing generally good predictive power with quantifiable model-uncertainty impacts. The dataset, publicly available online, should enhance astrophysical reaction-network simulations and guide further improvements in TALYS modeling and uncertainty quantification for nuclear data.

Abstract

In this work, we are presenting a new database of astrophysical interest, based on calculations performed with the nuclear reaction code TALYS. Four quantities are systematically calculated for over 8000 nuclides: cross sections, reaction rates, Maxwellian Averaged Cross Sections (or MACS) at 30 keV and partition functions. For cross sections and reaction rates, nine reactions are considered, induced by neutron, proton or alpha. The main complement of this database compared to existing ones is that the impact of reaction models ({\it e.g.} level density, gamma strength function, and optical model) is estimated by varying 9 different models, and by proposing calculated values for each of them, together with averages, standard deviations and other statistical quantities. This new database, called TENDL-astro, version 2023, is available online (https://tendl.web.psi.ch/tendl\_2023/astro/astro.html) and linked to the well-known TENDL database, used in a variety of applications.

Figures (7)

• Figure 1: Examples of recommended radiative neutron-capture rates for $^{120}$Sn (left) and $^{160}$Sn (right) from this work, compared to the JINA calculations Cyburt_2010 (blue curves). The label "RR" indicates reaction rate. See text for details.
• Figure 2: Examples of recommended radiative neutron-capture rates for neutron-rich Sn isotopes (from $A=118$ to $A=161$, left panel) and neutron-poor Sn isotopes (from $A=97$ to $A=118$, right panel).
• Figure 3: Examples of radiative neutron-capture rates as a function of the temperature for neutron-rich Sn isotopes using four different model sets.
• Figure 4: $\epsilon_{\rm rms}$ and $f_{\rm rms}$ for (n,$\gamma$) MACS at 30 keV, for 480 models. Ratios are the model $i$ over the experimental MACS, as compiled in the KADoNiS database dillmann2006DILLMANN2014171, for $40 \le A \le 210$.
• Figure 5: Top: Ratio of the calculated (n,$\gamma$) MACS (either recommended (black circles) or JEFF-3.3 (red squares), being the quantity Theory) over the KADoNiS values (being the quantity Exp) for neutron numbers corresponding to $40 \le A \le 210$. Uncertainties in black vertical lines correspond to the standard deviations of the 10 most trusted model sets (Group B), and gray bands correspond to the experimental uncertainties. Dashed lines are median values and the full line indicates Theory/Exp.=1. Bottom: same with C being the average of 480 model variations (Group C).