EUROPIUM s-PROCESS SIGNATURE AT CLOSE-TO-SOLAR METALLICITY IN STARDUST SiC GRAINS FROM ASYMPTOTIC GIANT BRANCH STARS

Measurements of the ¹⁵¹Eu isotope fraction [fr(¹⁵¹Eu) = ¹⁵¹Eu/(¹⁵¹Eu+¹⁵³Eu)] in stars are still scarce and only available for metal-poor stars (Sneden et al. 2002; Aoki et al. 2003a, 2003b; Roederer et al. 2008), with the exception of the solar system, where measurements on meteorites, as well as the solar photosphere, give fr(¹⁵¹Eu) = 0.48 (Lawler et al. 2001; Lodders 2003). Recent astronomical observations of the ¹⁵¹Eu isotope fraction in carbon-enhanced metal-poor (CEMP) stars enriched in the elements made by the slow neutron capture process (the s-process) have been shown to differ from the solar system value (Aoki et al. 2003b), but to be in good agreement with s-process predictions in low mass asymptotic giant branch (AGB) stars of one-half solar metallicity (Arlandini et al. 1999). The values reported by Arlandini et al. (1999), however, were obtained using a theoretical ¹⁵¹Sm(n, γ)¹⁵²Sm cross section from Toukan et al. (1995) that is approximately 40% lower than recent experimental measurements (Abbondanno et al. 2004; Marrone et al. 2006; Wisshak et al. 2006). The higher ¹⁵¹Sm(n, γ)¹⁵²Sm cross section results in an enhanced neutron-capture channel on ¹⁵¹Sm and a lower production of ¹⁵¹Sm. Since ¹⁵¹Eu is the radiogenic product of the decay of ¹⁵¹Sm (with a half-life of 90 yr), a higher ¹⁵¹Sm(n, γ) cross section results in a lower predicted fr(¹⁵¹Eu)_s-process value, outside the range observed in CEMP stars.

Analyses of presolar stardust grains provide an opportunity to explore the Eu isotopic compositions in stars with a different range of metallicities than those of metal-poor stars. This is because the vast majority of stardust SiC grains (the "mainstream grains") are believed to have originated in the outflows of ∼1.5–3 M_☉ carbon-rich AGB stars with close-to-solar metallicity (Hoppe et al. 1994; Zinner et al. 2006). Here, we present the results of Eu isotopic analyses carried out with a Sensitive High Resolution Ion Microprobe-Reverse Geometry (SHRIMP-RG). We analyzed 13 large mainstream SiC grains (LS+LU fractions) and a SiC-enriched bulk sample (KJB fraction) extracted from the Murchison carbonaceous chondrite (Amari et al. 1994).

Previous Eu isotopic secondary ion mass spectrometry (SIMS) analyses by Terada et al. (2006) in two mainstream stardust SiC grains indicated fr(¹⁵¹Eu) values lower than s-process predictions and astronomical observations. The difference observed between the fr(¹⁵¹Eu) values obtained for SiC grains by Terada et al. (2006) and those obtained for CEMP stars by Aoki et al. (2003b) may indicate that fr(¹⁵¹Eu) depends on metallicity. However, the values reported by Terada et al. (2006) are even lower than the fr(¹⁵¹Eu) shown by the bulk solar system. Since 98% of Eu in the solar system is expected to have been produced by the rapid neutron capture nucleosynthesis (r-process), fr(¹⁵¹Eu)_☉ is commonly used as a proxy for fr(¹⁵¹Eu)_r-process. Thus, it is surprising that the results presented by Terada et al. (2006) for grains believed to have condensed in the outflows of low mass carbon-rich AGB stars are closer to the fr(¹⁵¹Eu) expected for r-process nucleosynthesis than that expected for s-process nucleosynthesis. The results presented here will help to resolve these puzzles.

2.1. Carbon, Nitrogen, and Silicon Isotopic Measurements

2.2. Europium Isotopic Measurements

The LS+LU SiC grains studied here show Si, C, and N isotopic compositions in the range displayed by mainstream grains (Table 1) suggesting that these grains formed in the outflows of low mass C-rich AGB stars with close-to-solar metallicity. The ¹⁵¹Eu/¹⁵³Eu ratios measured in single SiC grains from the LS+LU fractions range from 0.95 to 1.62 (Table 1), compared to the solar system ¹⁵¹Eu/¹⁵³Eu of 0.916 (Lodders 2003). The fr(¹⁵¹Eu) values derived from our measurements range from 0.49 to 0.61 (Figure 1 and Table 1). The weighted mean ¹⁵¹Eu/¹⁵³Eu of the SiC-enriched bulk sample (KJB fraction), based on four measurements, is 1.20 ± 0.14 (95% conf.), which corresponds to a fr(¹⁵¹Eu) value of 0.54 ± 0.03 (95% conf.). All single SiC grains show fr(¹⁵¹Eu) values higher than the solar system value, however, only 1 out of 13 grains shows a fr(¹⁵¹Eu) value that differs from the solar system value by more than 2σ.

Zoom In Zoom Out Reset image size

Our results are clearly at odds with previous SIMS determinations in two mainstream stardust SiC grains from low-mass AGB stars that indicated lower-than-solar fr(¹⁵¹Eu) values (Terada et al. 2006). The fr(¹⁵¹Eu) values derived by Terada et al. (2006) for the two SiC grains are ∼0.43. We suspect that the inconsistency between our measurements and those of Terada et al. (2006) may be due to differences in the experimental approach used. We have shown previously that energy filtering, i.e., the selection of an appropriate energy window for the secondary ions on the basis of the momentum spectrum, is essential for suppressing unwanted interferences in the mass region of the Eu isotopes in a SiC matrix. The data of Terada et al. (2006) were obtained by using high-mass resolution alone, which we have found to be insufficient to remove molecular interferences that can significantly affect the Eu isotope measurements. Given the low value measured by Terada et al. (2006) and the similarly low ratio we have measured in the pure SiC standard, we suspect that their measurements are compromised.

Since the SiC-enriched bulk sample (KJB fraction) yields a fr(¹⁵¹Eu) value that is within the range observed in the single grains, but considerably more precise, we will focus the following discussion on this result.

In AGB stars, the s-process nucleosynthesis is responsible for the production of elements heavier than Fe. The main neutron source for the s-process in the He-intershell of AGB stars is the ¹³C(α, n)¹⁶O reaction which operates under radiative conditions at relatively low temperatures (T ∼ 0.9 × 10⁸ K) during the intervals between thermal pulses, and results in low neutron densities (∼10⁶–10⁷ neutrons cm⁻³). During thermal pulses in low-mass AGB stars, the ²²Ne(α, n)²⁵Mg reaction is marginally activated when the maximum temperature at the bottom of the He-burning shell reaches T ∼ 3 × 10⁸ K, producing a small neutron burst with a high neutron density peak (up to ∼10¹⁰ neutrons cm⁻³).

Branching points on the s-process path at ¹⁵¹Sm, ¹⁵³Sm, ¹⁵²Eu, and ¹⁵³Gd affect the production of ¹⁵¹Eu and ¹⁵³Eu isotopes in the He-intershell of AGB stars (Figure 2(a)). The competition between neutron capture and β-decay at these branching points can be expressed by a branching factor (f_n), calculated from

$$\begin{equation*} f_n = \frac{{\lambda _n }}{{(\lambda _n + \lambda _\beta)}}, \end{equation*}$$

where λ_n = N_n ν_T 〈σ〉 and λ_β = ln2/t_1/2 are the neutron capture rate and the β-decay rate, respectively. Here, N_n, ν_T, 〈σ〉, and t_1/2 are the neutron density, the thermal velocity, the Maxwellian averaged (n, γ) cross section (MACS), and the half-life, respectively. Branching factors were calculated for ¹⁵¹Sm, ¹⁵³Sm, ¹⁵²Eu, and ¹⁵³Gd as a function of neutron density for thermal energies of kT = 8 keV (T ∼ 0.9 × 10⁸ K) and kT = 23 keV (T ∼ 2.7 × 10⁸ K). The results are given in Figures 2(b) and (c). We used the β-decay rates reported by Takahashi & Yokoi (1987) and the latest accepted neutron-capture rates from Dillmann et al. (2006). Note that the branching at ¹⁵²Eu is peculiar, since it involves the competition between neutron capture, β-decay, and electron capture (EC). The branching at ¹⁵³Gd is governed by electron capture and neutron capture.

Zoom In Zoom Out Reset image size

During the interpulse periods most of the s-process flow proceeds via the sequence ¹⁵¹Sm(n, γ) ¹⁵²Sm(n, γ) ¹⁵³Sm(β, ν) ¹⁵³Eu bypassing ¹⁵¹Eu. After the neutron flux is extinguished, ¹⁵¹Eu is fed by the radioactive decay of ¹⁵¹Sm. During the thermal pulses, the β-decay rate of ¹⁵¹Sm is increased by a factor of ∼26 (Takahashi & Yokoi 1987) and the s-process path can branch toward ¹⁵¹Eu. However, the neutron-capture channel on ¹⁵¹Sm is also enhanced since higher neutron densities are produced together with higher temperatures. The branching points at ¹⁵³Gd and ¹⁵³Sm are largely open at these high neutron densities, feeding ¹⁵⁴Gd and ¹⁵⁴Sm, respectively, which result in a smaller s-process contribution to ¹⁵³Eu. The AGB model predictions shown in Figure 3 demonstrate that the final result is a very mild enhancement of fr(¹⁵¹Eu) values in the stellar envelope.

Zoom In Zoom Out Reset image size

In Figure 3, the stardust SiC-enriched bulk data are compared to the s-process AGB predictions from the FRUITY database (for details, see Cristallo et al. 2009, 2011). In all stellar models, the fr(¹⁵¹Eu)_envelope is not significantly different from the solar value. The close-to-solar metallicity models (Z = 0.01 and Z = 0.014, Figures 3(a) and (b), respectively) show fr(¹⁵¹Eu)_envelope approximately 2% higher than the fr(¹⁵¹Eu)_☉ value (Lodders 2003). The low-metallicity models (Z = 0.0001 and Z = 0.0003; Figures 3(c) and (d), respectively), on the other hand, show fr(¹⁵¹Eu)_envelope values that are about 4% lower than the fr(¹⁵¹Eu)_☉ value. It is important to note that Cristallo et al. (2009, 2011) adopted the experimental ¹⁵¹Sm(n, γ) cross section reported by Marrone et al. (2006), which is significantly higher than the value used by Arlandini et al. (1999). This higher cross section results in a lower fr(¹⁵¹Eu) as more ¹⁵¹Sm is destroyed by neutron captures. The fr(¹⁵¹Eu) derived for the SiC-enriched bulk sample (KJB) shows a higher fr(¹⁵¹Eu) value than those predicted by Cristallo et al. (2009, 2011).

The fr(¹⁵¹Eu) values derived from our measurements agree well with fr(¹⁵¹Eu) values derived from astronomical observations (Figure 1). Aoki et al. (2003b) reported fr(¹⁵¹Eu) for two CEMP stars, LP 625-44 and CS 31062-050, which show excesses of s-process elements. These stars are very metal-poor subgiants with [Fe/H] of −2.7 (LP 625-44; Aoki et al. 2000) and −2.4 (CS 31062-050; Aoki et al. 2002). Both stars show variations of their radial velocities, indicating that they belong to binary systems (Aoki et al. 2000, 2003b). According to the classification suggested by Jonsell et al. (2006), LP 625-44 and CS 31062-050 are CEMP-r+s stars, showing both s- and r-process enhancements. One possible scenario to explain the enrichment observed in CEMP-r+s stars is to assume that the parent cloud of the binary system was already enriched in r-process elements, while the s-process elements (and also carbon) were the result of mass transfer from an AGB star to the lower-mass companion (e.g., Bisterzo et al. 2010). Lugaro et al. (2012) instead propose that the composition of CEMP-r+s stars is the result of a neutron-capture process in-between the s-process and the r-process. The fr(¹⁵¹Eu) values derived by Aoki et al. (2003b) for LP 625-44 and CS 31062-050 are 0.60 and 0.55, respectively. These values are clearly higher than fr(¹⁵¹Eu)_☉ of 0.48 and current AGB model predictions from Cristallo et al. (2009, 2011).

To address this mismatch we performed several tests varying the neutron-capture cross sections and beta-decay rates that affect fr(¹⁵¹Eu) and found that the ¹⁵¹Sm(n, γ) reaction rate plays the major role in setting the predicted fr(¹⁵¹Eu) value. In Figure 4, we show fr(¹⁵¹Eu) in the envelope of a 2 M_☉ model with Z = 0.01 computed with different theoretical ¹⁵¹Sm(n, γ) reaction rates and their uncertainties. Models have been calculated with the FUll-Network Stellar Evolutionary Code (Straniero et al. 2006 and references therein). The results labeled as "M06" were obtained using the experimental data by Marrone et al. (2006), multiplied by the theoretical stellar enhancement factor (SEF) from Bao et al. (2000). Marrone et al. (2006) reported an experimental uncertainty on the order of 5%, which results in a narrow predicted range of fr(¹⁵¹Eu). In previous investigations, applying the SEF was thought to account for the contribution of excited nuclear states in stellar environments. It has been shown that this is not the case by Rauscher et al. (2011), and Rauscher (2012) presented an improved approach, also accounting for the uncertainties in the stellar rates. This leads to error bars larger than those given by the experiments due to the remaining theoretical uncertainties in the excited state contributions to the stellar rate. Using this approach (labeled "R12" in Figure 4), we obtained a larger allowed range of fr(¹⁵¹Eu), which easily covers the values observed both in CEMP stars and in SiC grains.

Zoom In Zoom Out Reset image size

We presented new Eu isotopic data obtained on stardust SiC grains with SHRIMP-RG and compared them to previous astronomical observations in CEMP stars enriched in s-process elements and with s-process AGB model predictions. Despite the difference in metallicity between the parent stars of the grains and the metal-poor stars, we found good agreement between the grains and the stellar data. Because of large uncertainties in the single grain data, the result of low Eu concentrations, only in one case is the fr(¹⁵¹Eu) value derived for a single grain higher than fr(¹⁵¹Eu)_☉ by more than 2σ. The fr(¹⁵¹Eu) value derived for the KJB aggregate is, on the other hand, very well constrained, and matches those of CEMP stars, but is approximately 12% higher than current s-process predictions. Our new data can only be matched when the uncertainties predicted with the model of Rauscher (2012) are applied and support this more accurate approach of computing neutron-capture cross sections in stellar environments.

J. N. Ávila acknowledges support by CNPq grants 200081/2005-5 and 150570/2011-2. T. R. Ireland acknowledges support by ARC grants DP0342772 and DP0666751. M. Lugaro acknowledges the support of the ARC via a Future Fellowship and of Monash University via a Monash fellowship. E. Zinner acknowledges support by NASA grant NNX11AH14G. S. Cristallo acknowledges financial support from the FIRB2008 program (RBFR08549F-002) and from the PRIN-INAF 2011 grant "Multiple populations in Globular Clusters: their role in the Galaxy assembly." T. Rauscher acknowledges support by the Swiss NSF, the EUROCORES EuroGENESIS research program, and the ENSAR/THEXO European FP7 program. We thank an anonymous referee for comments.