Sm and Gd Isotopic Shifts in Eucrites and Implications for Their Cosmic-Ray Exposure History

Eight basaltic eucrites, Dar al Gani (DaG) 380, DaG 391, DaG 411, DaG 443, DaG 480, Juvinas, Millibillillie, and Stannern, were used in this study. The DaG series meteorites were found in the Libyan Desert and were weakly shocked: DaG 380 is a monomict eucrite, whereas DaG 391, DaG 411, and DaG 443 are polymict eucrites (Grossman 1999; Grossman & Zipfel 2001). The other three eucrites, Juvinas, Millibillillie, and Stannern, are well known as basaltic eucrites and have often been used in chronological and geochemical studies to further understanding of the early evolution processes of the solar planetary materials (e.g., Bouvier et al. 2015; Bermingham et al. 2016; Brennecka & Kleine 2017).

As standard reference materials without cosmic-ray irradiation, chemical reagents of Sm and Gd (1000 mg L⁻¹ of single-element standard solutions for the inductively coupled plasma mass spectrometer (ICP–MS) and ICP atomic emission spectroscopy analyses commercially obtained from SPEX CertiPrep) were used for the determination of their isotopic ratios.

Each sample, weighing 0.3–0.4 g, was powdered and then decomposed completely by HF–HClO₄ with heating. After evaporation to dryness, the residue was redissolved in 5 mL of 2 M HCl. This sample solution was divided into two portions: the main portion, over 90% of the total solution, was for isotopic measurements of Sm and Gd using a thermal ionization mass spectrometer (TIMS), and the minor portion was for the determination of elemental abundances of rare earth elements (REEs) using ICP–MS.

For the isotopic study, the major portion of each sample solution was loaded onto a cation exchange resin-packed column (Bio-Rad AG50WX8, 200–400 mesh, H⁺ form, 50 mm length × 4.0 mm diameter). The column was washed with 1.7 M HCl for the elution of major elements and then washed with 6 M HCl for REE elution. The REE fraction was evaporated to dryness, redissolved in a drop of 0.25 M HCl, and loaded onto a column packed with LN-resin (Eichrom, LN-resin, particle size of 20–50 μm, 100 mm length × 2.5 mm diameter). The column was washed with 0.25 M HCl. Then, Sm and Gd were eluted with 0.35 and 0.5 M HCl, respectively. The details of the chemical procedures for the REE mutual separation are given in Mizutani et al. (2020).

The minor portion of each sample solution was evaporated to dryness and redissolved using 5 mL of 2% HNO₃. A 0.01 g quantity of a 1 ppb mixture solution of indium and bismuth was added to the sample solution as an internal standard to optimize the analytical conditions for the determination of the elemental abundances of the REEs. An Agilent 7500cx ICP–MS instrument was used for the analyses.

Isotopic analyses of Sm and Gd for individual samples were performed using a TIMS (Triton Plus) at the National Museum of Nature and Science in Tsukuba, Japan. Sm and Gd were measured on double Re filaments without any activators. ¹⁴⁷Sm/¹⁵²Sm = 0.56081 and (¹⁵⁵Gd + ¹⁵⁶Gd)/¹⁶⁰Gd = 1.61290 were used as the normalization factors for the correction of instrumental mass fractionation of Sm and Gd isotopic ratios, respectively (Hidaka et al. 1995; Hidaka & Yoneda 2007, 2014). All analyses were performed in static mode with the amplifier rotation system. The detailed analytical protocols for the Sm and Gd isotopic measurements were described in our previous studies (e.g., Hidaka & Yoneda 2014).

Figure 1 shows CI chondrite-normalized REE abundance patterns of eucrite samples measured in this study. All eight eucrites exhibit REE abundances similar to Stannern-trend eucrites. Judging from the pronounced negative Eu anomalies and heavy REE depletions (i.e., (Dy/Lu)_n < 1) of DaG 391 and Stannern, these two can be classified as Stannern-trend eucrites. Only Millbillillie shows a positive Eu anomaly, which is a typical characteristic of residual eucrites (e.g., Yamaguchi et al. 2009; Roszjar et al. 2011). However, the Millbillillie sample in this study does not show light-REE depletion, which is another typical characteristic of residual eucrites (e.g., Yamaguchi et al. 2009). Previous studies on the chemical measurements of Millbillillie provide several types of REE patterns showing a clear light-REE depletion with more pronounced positive Eu anomaly (Makishima & Masuda 1993) and a relatively flat REE pattern with a negative Eu anomaly (Barrat et al. 2000), which are obviously different from our data. It can be concluded that classification of Millbillillie based on REE abundance patterns is complicated. The other five eucrites can be classified as Main Group–Nuevo Laredo trend eucrites. High incompatible element abundances of Stannern-trend eucrites are interpreted as a consequence of contamination of Main Group eucritic magmas by melts derived by partial melting of the asteroid's crust. Millibillillie has a positive Eu anomaly (Eu/Eu* = 1.2), whereas other eucrites show negative Eu anomalies (Eu/Eu* = 0.7–0.9). These data are almost the same as those in typical basaltic eucrites shown by previous studies (Consolmagno & Drake 1977; Makishima & Masuda 1993; Barrat et al. 2000; Yamaguchi et al. 2009).

Zoom In Zoom Out Reset image size

It is reported that terrestrial weathering also causes significant REE mobilization of meteorites. The REE patterns of some Antarctic eucrites show positive or negative Ce anomalies (Floss & Crozaz 1991; Mittlefehldt & Lindstrom 1991). These REE abundance variations are interpreted as being caused by the dissolution of phosphates, which are important REE carriers in water equilibrated with the atmosphere (Mittlefehldt & Lindstrom 1991), and the fractionation of Ce from the other +3 charged REE due to oxidization of Ce to the +4 state. Cold desert weathering can also cause similar REE disturbance in meteorites (Shimizu et al. 1983; Floss & Crozaz 1991; Mittlefehldt & Lindstrom 1991), but the state in the hot desert is more complex than the situation of Antarctic weathering. Many secondary minerals filling cracks, such as carbonates and sulfates, were found in hot desert samples (Barrat et al. 2003; Lee & Bland 2004). However, among those meteorites, light-REE-poor samples have light-REE enrichment patterns, and significant REE mobilization has not been detected in Saharan eucrites, which are rich in REE abundances, except for one eucrite (NWA 047) showing obvious light-REE enrichment (Barrat et al. 2003). As shown in Figure 1, the samples used in this study display no positive Ce anomalies and no light-REE enrichments, suggesting no clear traces of cold or hot desert weathering on REE abundances.

The Sm and Gd isotope ratios of the eight eucrites measured using TIMS are listed in Table 1. The data show clear isotopic deficits in ¹⁴⁹Sm and ¹⁵⁷Gd and excesses in ¹⁵⁰Sm and ¹⁵⁸Gd for seven of the eight eucrites, excepting DaG 443. Because the neutron capture reaction of ¹⁴⁹Sm(n,γ)¹⁵⁰Sm produces ¹⁵⁰Sm from the equivalent number of ¹⁴⁹Sm, the isotopic decrement of ¹⁴⁹Sm quantitatively corresponds to the isotopic increment of ¹⁵⁰Sm. Thus, if Sm isotope composition is affected predominantly by this reaction, Sm isotope data points yield a linear trend with slope –1 in the ¹⁵⁰Sm/¹⁵²Sm–¹⁴⁹Sm/¹⁵²Sm three-isotope diagram. Similarly, if a Gd isotope is affected by a neutron capture reaction of ¹⁵⁷Gd(n,γ)¹⁵⁸Gd, Gd isotope data yield the same trend on the ¹⁵⁷Gd/¹⁶⁰Gd–¹⁵⁸Gd/¹⁶⁰Gd three-isotope diagram. The correlation diagrams between ¹⁴⁹Sm/¹⁵²Sm and ¹⁵⁰Sm/¹⁵²Sm and between ¹⁵⁷Gd/¹⁶⁰Gd and ¹⁵⁸Gd/¹⁶⁰Gd for the occurrence of neutron capture reactions are shown in Figures 2(a) and (b), respectively. Except for DaG 443, all seven eucrites measured in this study show significant isotopic shifts of Sm and Gd from those of terrestrial standard materials (STD). As shown in Figure 2, the Sm and Gd isotope data of these eucrites are consistent with the straight lines with slope –1 drawn from STD compositions, which represent isotopic shifts caused by the neutron capture reactions. These trends strongly suggest that the Sm and Gd isotope shifts of these eucrites are caused by neutron capture effects associated with exposure to cosmic rays. The fluence of neutrons generated by cosmic-ray irradiation can be quantified by the following equation based on previous studies (e.g., Hidaka et al. 1995, 2012):

$$\begin{eqnarray*}&&\begin{array}{l}{{\rm{\Psi }}}_{\mathrm{sample}}=\displaystyle \frac{{\left({}^{150}\mathrm{Sm}{/}^{149}\mathrm{Sm}\right)}_{\mathrm{sample}}-{\left({}^{150}\mathrm{Sm}{/}^{149}\mathrm{Sm}\right)}_{\mathrm{STD}}}{{\left({}^{150}\mathrm{Sm}{/}^{149}\mathrm{Sm}\right)}_{\mathrm{STD}2}-{\left({}^{150}\mathrm{Sm}{/}^{149}\mathrm{Sm}\right)}_{\mathrm{STD}}}\\ \ \ \times \,\displaystyle \frac{{{\rm{\Sigma }}}_{\mathrm{sample}}}{{{\rm{\Sigma }}}_{\mathrm{STD}2}}\times {{\rm{\Psi }}}_{\mathrm{STD}2},\end{array}\end{eqnarray*}$$

where Ψ and Σ represent neutron fluences and effective total macroscopic neutron capture cross sections, respectively. STD2 is a standard material that is experimentally irradiated by thermal neutrons of 5.94 × 10¹⁵ neutrons cm⁻² in a laboratory (Hidaka et al. 1995), and STD is a standard material that represents the natural composition of Sm isotopes without irradiation. Comparisons between Sm isotopic shifts of meteorite samples and experimentally irradiated material provide estimates of neutron fluence. According to this equation, the neutron fluences of eucrite samples measured in this study are calculated as (0.28–2.38) × 10¹⁵ neutrons cm⁻² (see Table 2), which are 10–10² times lower than those of lunar regolith and aubrites (>10¹⁶ neutrons cm⁻²; e.g., Hidaka & Yoneda 2007; Hidaka et al. 2012).

Zoom In Zoom Out Reset image size

Figure 3 shows a comparison between neutron fluences measured in this study and CRE ages measured by Miura et al. (1998). In general, these two quantities are proportionally correlated because they are proportional to amounts of certain nuclides that are produced by cosmic-ray irradiation. However, some meteorite samples do not show such a linear correlation owing to the inconsistency between the neutron fluences and the CRE ages because of complicated rather than simple CRE processes. Neutron fluences estimated from a stable nuclide like ¹⁵⁰Sm record the whole CRE history of meteorites after their formation as a part of the parent body. By contrast, the CRE ages, estimated from a short-lived nuclide like ⁸¹Kr with a half-life of 0.23 Ma, record the CRE duration of recent–100 Ma (mainly after the ejection from the parent body). Hence, if meteorite samples experienced irradiation on the surface of their parent body earlier than recent–100 Ma before ejection, they show high neutron fluences relative to their CRE ages. This is often the case for lunar meteorites (e.g., Hidaka et al. 2017). As shown in Figure 3, three fall-eucrites measured in this study, Juvinas, Millibillillie, and Stannern, show a positive correlation between their neutron fluences and CRE ages, with no excesses of their neutron fluences. This suggests that neutron capture reactions in these three eucrites occurred predominantly by 4π irradiation during their transport from the parent body to Earth and that 2π irradiation at the surface of their parent body is negligible. A slight deficit of the neutron fluence found in Juvinas is probably due to the different samples used for the estimation of neutron fluences in this study and CRE age in a previous study. Considering that the recovered mass of Juvinas was over 90 kg, many Juvinas fragments were originally located in various depths of the large meteoroidal body in space. The Juvinas sample used in this study may have been located at a deeper part of the whole Juvinas meteorite relative to that used for determining the CRE age (Miura et al. 1998) during 4π irradiation. This hypothesis is supported by a previous Sm isotope study on Juvinas conducted by Andreasen & Sharma (2006). Their isotope measurement and calculation of Juvinas resulted in neutron fluence of (0.2–0.4) × 10¹⁵ neutrons cm⁻² (Figure 3), which are in good agreement with our Juvinas data (0.28 × 10¹⁵ neutrons cm⁻²). The slight difference of the neutron fluence between this study and Andreasen & Sharma (2006) probably reflects a difference of sampling location from the 90 kg sized Juvinas clast as discussed above. Assuming that the four desert-found eucrites (DaG 380, DaG 391, DaG 411, and DaG 480; DaG 443 was excluded because of its low neutron fluence) are exposed to the same flux of neutrons as the three fall-eucrites, their CRE ages can be estimated using linear regression of neutron fluences and CRE ages of the three fall-eucrites. The estimated CRE ages of the desert-found eucrites are also shown in Figure 3, together with those of the three fall-eucrites measured by Miura et al. (1998). The gray bars in the figure display the CRE age clusters of eucrites reported by Strashnov et al. (2013). As shown in Figure 3, the estimated CRE ages of DaG 391, DaG 411, and DaG 380 can be attributed to the previously reported age clusters, whereas DaG 480 has an apparently greater CRE age than these age clusters. This indicates that DaG 480 was ejected from the parent body by an impact event that differs from other studied eucrites, or that DaG 480 experienced additional irradiation on the surface of the parent body.

Zoom In Zoom Out Reset image size

The energy of neutrons generated by cosmic-ray irradiation depends on the temperature, size, and composition of the irradiated target material (e.g., Spergel et al. 1986). In this study, the energy distribution of neutron fluence in the thermal region (E < 0.1 eV) in eucrites is investigated from the combination of the Sm and Gd isotopic shifts. Using the difference of the neutron capture resonance energies between ¹⁴⁹Sm and ¹⁵⁷Gd, 0.0973 and 0.0314 eV, respectively, a comparison of the magnitudes between ¹⁴⁹Sm–¹⁵⁰Sm and ¹⁵⁷Gd–¹⁵⁸Gd isotopic shifts provides information on the neutron energy profile in a thermal region. For the quantitative assessment of the neutron energy profile in individual eucrites, an index ε_Sm/ε_Gd given by Russ et al. (1971) is defined as follows:

$$\begin{eqnarray*}&&\displaystyle \frac{{\varepsilon }_{\mathrm{Sm}}}{{\varepsilon }_{\mathrm{Gd}}}=\displaystyle \frac{\displaystyle \frac{{\left({}^{150}\mathrm{Sm}{/}^{149}\mathrm{Sm}\right)}_{\mathrm{sample}}-{\left({}^{150}\mathrm{Sm}{/}^{149}\mathrm{Sm}\right)}_{\mathrm{STD}}}{1+{\left({}^{150}\mathrm{Sm}{/}^{149}\mathrm{Sm}\right)}_{\mathrm{sample}}}}{\displaystyle \frac{{\left({}^{158}\mathrm{Gd}{/}^{157}\mathrm{Gd}\right)}_{\mathrm{sample}}-{\left({}^{158}\mathrm{Gd}{/}^{157}\mathrm{Gd}\right)}_{\mathrm{STD}}}{1+{\left({}^{158}\mathrm{Gd}{/}^{157}\mathrm{Gd}\right)}_{\mathrm{sample}}}}.\end{eqnarray*}$$

As listed in Table 2, the seven eucrites measured in this study (DaG 443 is excluded because of its low neutron fluence) have similar ε_Sm/ε_Gd values (0.61–0.76), which are identical within the 2σ error range. This indicates that these eucrites experienced almost identical irradiation conditions, predominantly 4π irradiation in space, as discussed in Section 3.2. Hence, the ε_Sm/ε_Gd ratios of the eucrites measured in this study can be interpreted to represent the neutron energy distribution generated by the 4π irradiation. Although the ε_Sm/ε_Gd values of the eucrites are identical within the 2σ error range, it should be noted that DaG 480 shows a slightly higher ε_Sm/ε_Gd value (0.76) than those of the six other eucrites (0.61–0.67). Interestingly, DaG 480 also shows the highest neutron fluences among the seven eucrites (Figure 3), suggesting that DaG 480 might have experienced another irradiation period in addition to the simple 4π irradiation. One of the hypothetical ideas is that DaG 480 experienced 2π irradiation on the surface of the parent body. Assuming that this idea is correct, the difference of ε_Sm/ε_Gd values between DaG 480 and the six other eucrites may be interpreted as the difference in the energy distribution of neutrons generated by 2π and 4π irradiation, implying that 4π irradiation produces more thermalized neutrons than does 2π irradiation.

For a better understanding of the CRE conditions of the eucrites, their ε_Sm/ε_Gd ratios are compared with those of lunar samples (Hidaka et al. 1999, 2000; Hidaka & Yoneda 2007). The ε_Sm/ε_Gd ratios of different types of meteorites cannot be compared simply because it is considered that the parameter ε_Sm/ε_Gd depends on the chemical composition of the target materials. In this study, for characterizing the chemical compositions of meteorite samples from the perspective of their influence on the energy distribution of neutrons, an effective macroscopic neutron capture cross section Σ_eff, which is generally used for an index of chemical composition, is defined as follows:

$$\begin{eqnarray*}&&{{\rm{\Sigma }}}_{\mathrm{eff}}=\displaystyle \sum _{i}{\sigma }_{i}\times {A}_{i},\end{eqnarray*}$$

where i represents individual nuclides and σ_i and A_i are neutron capture cross sections and atomic abundances of the nuclide i, respectively. Figure 4 is a correlation diagram between Σ_eff and ε_Sm/ε_Gd for the individual meteorite samples. In the figure, dashed lines are theoretical prediction lines of ε_Sm/ε_Gd along with Σ_eff at two different temperatures, 0 and 400 K, given by Lingenfelter et al. (1972). Two data sets from eucrites and lunar samples are clearly distinguished in the figure because there is a clear difference in their individual Σ_eff values (0.0085–0.011 for A-12, A-15, and A-17 lunar regolith samples; 0.004–0.005 for A-16 lunar regolith samples; and 0.0065–0.007 for eucrites). The data points of eucrites measured in this study, together with those of lunar samples, are plotted between the two theoretical prediction lines in the figure. This indicates that the irradiation conditions of individual meteorites in the figure are roughly similar to each other. It is difficult to present a statistically critical discussion on the irradiation conditions of individual eucrites only from the ε_Sm/ε_Gd values because of their large analytical uncertainties. However, the irradiation conditions of eucrites measured in this study seem to show more thermalized effects than those of most lunar samples. The data points of eucrites are plotted close to the lower line showing the temperature of 0 K, whereas most of the data points of lunar samples are plotted relatively close to the upper line (400 K) in the figure. This implies that neutrons in eucrites generated by cosmic rays were well thermalized compared with those in the lunar samples. Because 2π irradiation is dominant for the lunar samples, whereas 4π irradiation is dominant for the eucrites, the difference in neutron energy between the eucrites and lunar samples may reflect this difference in the irradiation condition, implying that 4π irradiation produces more thermalized neutrons than does 2π irradiation. Interestingly, the discrepancy of ε_Sm/ε_Gd ratios among the eucrites (discussed above) also implies the same interpretation. Both the comparisons of the Sm and Gd isotopic shifts among the eucrite samples and those between eucrites and lunar samples consistently imply that 4π irradiation produces thermalized neutrons. However, considering the influence of target size on neutron energy, it is well known that larger targets can produce better-thermalized neutrons than can smaller targets (e.g., Spergel et al. 1986). Then, the thermalized neutrons make the ε_Sm/ε_Gd value low. In other words, the higher the contribution of 2π irradiation, the lower the value of ε_Sm/ε_Gd. Our results from the isotopic shifts of Sm and Gd for eucrites are inconsistent with the common sense for the mechanism of neutron production induced by cosmic-ray irradiation. Although DaG 480 seems to show a slightly higher ε_Sm/ε_Gd value than do the other eucrites, the precision of the ε_Sm/ε_Gd values for eucrites is not sufficient to distinguish the difference between individual eucritic species. By contrast, the isotopic shifts of Sm and Gd for most lunar samples are tens to hundreds of times larger than those for eucrites. As a result, the precision of the isotopic variations of Sm and Gd of lunar samples is sufficient to discuss the thermalization degree from the ε_Sm/ε_Gd value. Then, the model of Lingenfelter et al. (1972) is generally known to provide a lower trend of the ε_Sm/ε_Gd value relative to the experimental data (e.g., Russ et al. 1971, 1972; Sands et al. 2001). However, in the case of eucrites, in the current analytical situation, it should be concluded that there is no clear evidence for the pre-irradiation record of DaG 480 before ejection from the parent body. For the further investigation of this issue, isotopic study of Er may be helpful. Because ¹⁶⁷Er has a large resonance integral in the epithermal region (0.5 eV < E < 2000 eV), the combination of the isotopic variations between ¹⁴⁹Sm and ¹⁶⁷Er may provide a hint for the difference in CRE histories of individual meteorites. However, considering the simple comparison of the cross section for neutron capture between ¹⁴⁹Sm and ¹⁶⁷Er, a neutron fluence above 10¹⁶ neutrons cm⁻² is required to detect clearly the isotopic shift of ¹⁶⁷Er–¹⁶⁸Er. Although the combination of the isotopic shifts of Sm and Er was effectively used for lunar meteorites showing thermal neutron fluences of (2–14) × 10¹⁶ neutrons cm⁻² to determine the energy balance between thermal and epithermal neutrons (Hidaka et al. 2020), it would not be expected to find clear evidence from the same application to eucrites having thermal neutron fluences of (0.28–2.38) × 10¹⁵ neutrons cm⁻² in this study.

Zoom In Zoom Out Reset image size