Constraints on the22Ne(α,n)25Mgs-process neutron source from analysis ofnatMg+ntotal and25Mg(n,γ)cross sections

Abstract

The ${}^{22}\mathrm{Ne}(\ensuremath{\alpha}{,n)}^{25}\mathrm{Mg}$ reaction is thought to be the neutron source during the s process in massive and intermediate mass stars as well as a secondary neutron source during the s process in low-mass stars. Therefore, an accurate determination of this rate is important for a better understanding of the origin of nuclides heavier than iron as well as for improving s-process models. Also, the s process produces seed nuclides for a later p process in massive stars, so an accurate value for this rate is important for a better understanding of the p process. Because the lowest observed resonance in direct ${}^{22}\mathrm{Ne}(\ensuremath{\alpha}{,n)}^{25}\mathrm{Mg}$ measurements is considerably above the most important energy range for s-process temperatures, the uncertainty in this rate is dominated by the poorly known properties of states in ${}^{26}\mathrm{Mg}$ between this resonance and threshold. Neutron measurements can observe these states with much better sensitivity and determine their parameters (except ${\ensuremath{\Gamma}}_{\ensuremath{\alpha}})$ much more accurately than direct ${}^{22}\mathrm{Ne}(\ensuremath{\alpha}{,n)}^{25}\mathrm{Mg}$ measurements. I have analyzed previously reported ${}^{\mathrm{nat}}\mathrm{Mg}+n$ total and ${}^{25}\mathrm{Mg}(n,\ensuremath{\gamma})$ cross sections to obtain a much improved set of resonance parameters for states in ${}^{26}\mathrm{Mg}$ between threshold and the lowest observed ${}^{22}\mathrm{Ne}(\ensuremath{\alpha}{,n)}^{25}\mathrm{Mg}$ resonance, and an improved estimate of the uncertainty in the ${}^{22}\mathrm{Ne}(\ensuremath{\alpha}{,n)}^{25}\mathrm{Mg}$ reaction rate. For example, definitely two, and very likely at least four, of the states in this region have natural parity and hence can contribute to the ${}^{22}\mathrm{Ne}(\ensuremath{\alpha}{,n)}^{25}\mathrm{Mg}$ reaction, but two others definitely have non-natural parity and so can be eliminated from consideration. As a result, a recent evaluation in which it was assumed that only one of these states has natural parity has underestimated the reaction rate uncertainty by at least a factor of 10, whereas evaluations that assumed all these states could contribute probably have overestimated the uncertainty.