The Neutron Energy Spectra of Lunar Meteorites Evaluated from Sm and Er Isotopic Compositions

Thirteen lunar meteorites, Dhofar 081, Dhofar 910, Dhofar 911, Northwest Africa (NWA) 482, NWA 2995, NWA 2996, NWA 3136, NWA 3163, NWA 4472, NWA 4734, NWA 4884, NWA 4932, and NWA 4936, whose abundances of several cosmogenic nuclides, such as ¹⁰Be, ²⁶Al, ³⁶Cl, and ⁴¹Ca, were partly reported (Nishiizumi & Caffee 2001, 2006; Nishiizumi et al. 2004, 2016), were used to collect additional new information on the CRE records from their Sm and Er isotopic compositions. Although the Sm isotopic data of 2 of the 13 meteorites, NWA 482 and NWA 2995, were already reported in our previous work (Hidaka et al. 2017), their Er isotopic data were newly measured in this study. The data of cosmogenic nuclides together with the elemental abundances of some major elements for the 11 samples are compiled in Table 1.

Five of the thirteen samples, Dhofar 081, Dhofar 910, Dhofar 911, NWA 3163 and NWA 4932, show feldspathic lithologies (Cahill et al. 2004; Korotev 2006; Korotev et al. 2009 Hudgins et al. 2011; Nagaoka et al. 2014; McLeod et al. 2016). Dhofar 081 is known to be paired with Dhofar 911 (Korotev et al. 2006). These two include fragmental highland breccias.

Among the other eight samples, NWA 2995, NWA 2996, NWA 3136, and NWA 4936 are felspathic breccias bearing basaltic components (Korotev et al. 2009; Mercer et al. 2013; Mészáros et al. 2016). These four have higher REE contents than the feldspathic materials because of the presence of the basaltic components. NWA 2995 is considered to be paired with NWA 2996 (Korotev et al. 2009). NWA 4472 is a KREEP-rich regolith breccia (Joy et al. 2011). NWA 4734 is an unbrecciated basalt that originated from low-Ti mare volcanism around 3.0 Ga ago (Chen et al. 2019; Wu & Hsu 2020). NWA 4884 is classified as a basaltic regolith breccia and originated possibly from the southern Lacus Veris within the Orientale basin and the western mare unit (Cao et al. 2019).

Individual samples were divided into two; one for the measurements of SRLs using AMS, and the other for the determinations of Sm and Er isotopic compositions using thermal ionization mass spectrometry (TIMS) and of REE elemental abundances using inductively coupled plasma-mass spectrometry (ICP-MS).

For SLRs measurements, 30–40 mg of each powdered sample was dissolved with a HF/HNO₃ mixture and a carrier solution containing Be and Cl. After separating Cl by AgCl precipitation, Be was separated and purified by hydroxide precipitation, cation exchange, and solvent extraction with acetylacetone. Then, Al was separated from the Al–acetylacetone complex by cation exchange. Following the separation of Cl, Be, and Al, Ca was separated from the remaining solutions by anion and cation exchange. The separation methods of Be, Al, Cl, and Ca as the analytical protocols for SLRs using AMS are given in previous studies (Nishiizumi et al. 1984a, 1984b, 1997).

For TIMS and ICP-MS analyses, 30–40 mg of each powdered sample was decomposed by a mixed acid of HF–HClO₄. Then, the residue was redissolved in 1 ml of 1.7 M HCl. The sample solution was divided into two portions: approximately 90% of the solution was used for the conventional resin chemistry to separate Sm and Er for their isotopic analyses using TIMS (Triton Plus), and the other 10% of the solution was used for the determination of the elemental abundances of REEs using ICP-MS (Agilent 7500cx).

As standard reference materials for the TIMS analyses, chemical reagents of Sm and Er (1000 mg l⁻¹ of single-element standard solutions for the ICP-MS and ICP-AES analyses provided by SPEX CertiPrep) were used for the determinations of their isotopic compositions.

For the chemical separation of Sm and Er, a two-step resin chemistry technique was used; the first step by cation exchange resin-packed column for the rough separation of REEs from major elements, and the second step by LN resin-packed column for mutual separation of Sm and Er from other REEs. The total procedural blank was less than 10 pg for each of Sm and Er, which is negligibly lower than the absolute amount of the element recovered from each sample through the chemical procedures. Detailed procedures of the resin chemistry were reported in our previous study (Mizutani et al. 2020).

Isotopic analyses of Sm and Er were performed using a TIMS machine equipped with nine Faraday cup collectors at the National Science Museum of Nature and Science, Tsukuba. All analyses were done in static mode with the amplifier rotation system. ¹⁴⁷Sm/¹⁵²Sm = 0.56081 and ¹⁷⁰Er/¹⁶⁶Er = 0.44547 were used as normalization values for the correction of instrumental mass fractionation. The analytical protocols for the conditions of the isotopic measurements of Sm and Er using TIMS, including the cup configurations, filament conditions, integration times, and so on, were described in our previous studies (Hidaka & Yoneda 2014; Mizutani et al. 2020; Hidaka et al. 2020).

AMS measurements were performed at the Lawrence Livermore Laboratory and the Purdue Rare Isotope Measurement Laboratory (PRIME Lab) at Purdue University.

The lithological properties of individual samples can be characterized from their REE abundances, as shown in Figure 1. The feldspathic species, Dhofar 081, Dhofar 910, Dhofar 911, NWA 482, and NWA 3163, show low contents of REEs with relatively flat REE patterns and clearly positive Eu anomalies. NWA 2995, NWA 2996, NWA 3136, NWA 4734, and NWA 4932, classified as intermediate-iron lunar meteorites with low to moderate concentrations of incompatible elements (Korotev et al. 2009), show intermediate REE contents in their REE abundances in the range of 8 to 50 × CI chondrite (CI). Because NWA 4472, NWA 4884, and NWA 4936 are regolith breccias including high-REE components such as REE-rich merrillite, KREEP, or mare basalt (Jolliff et al. 1993; Korotev et al. 2009), their patterns show high REE contents over the range of 100 × CI.

Zoom In Zoom Out Reset image size

Samarium has seven stable isotopes with mass numbers, 144, 147, 148, 149, 150, 152, and 154. Because ¹⁴⁹Sm has a very large thermal neutron capture cross section (σ_th = 4.2 × 10⁴ barn at E = 0.0253 eV), an isotopic shift from ¹⁴⁹Sm to ¹⁵⁰Sm caused by the neutron-capture reaction of ¹⁴⁹Sm(n,γ)¹⁵⁰Sm is expected in the presence of certain amounts of thermal neutrons arising by spallation reactions in association with the cosmic-ray irradiation. Isotopic data of Sm for the 13 lunar meteorites used in this study are listed in Table 2. Although the Sm isotopic data of NWA 482 and NWA 2995 have already been reported in our previous study (Hidaka et al. 2017), they are also shown in the table to compare with other data. All of the samples show significant variations of ¹⁴⁹Sm/¹⁵²Sm and ¹⁵⁰Sm/¹⁵²Sm in the range of −82 < ¹⁴⁹Sm < −3.5 and +6.2 < ¹⁵⁰Sm<+152, respectively. In particular, the isotopic variations of ¹⁴⁹Sm/¹⁵²Sm and ¹⁵⁰Sm/¹⁵²Sm found in NWA 4884 exceed those in any lunar surface materials recovered by the Apollo mission (Hidaka & Yoneda 2007), which correspond to a neutron fluence over 1 × 10¹⁷ neutrons cm⁻². Figure 2(A) is a three-isotope plot diagram of Sm to reveal the isotopic variations caused by neutron-capture reactions from the correlation between ¹⁴⁹Sm/¹⁵²Sm and ¹⁵⁰Sm/¹⁵²Sm isotopic ratios for the individual lunar meteorites used in this study. Almost all data points in the figure are plotted on the line with a slope of −1 in the figure, because the depletion of the isotopic abundance for ¹⁴⁹Sm should correspond to the enrichment of that for ¹⁵⁰Sm caused by the neutron capture reaction of ¹⁴⁹Sm(n,γ)¹⁵⁰Sm. It is easily understandable that the variations are due to the isotopic shift from ¹⁴⁹Sm to ¹⁵⁰Sm caused by the neutron capture reaction of ¹⁴⁹Sm(n,γ)¹⁵⁰Sm.

Zoom In Zoom Out Reset image size

Erbium has six stable isotopes with mass numbers 162, 164, 166, 167, 168, and 170. Considering the RI for the neutron-capture cross section of ¹⁶⁷Er in the epithermal energy region (0.5 eV < E < 0.5 MeV), ¹⁶⁷Er can react with not only thermal neutrons but also epithermal neutrons. Then, the isotopic shift of ¹⁶⁷Er–¹⁶⁸Er caused by ¹⁶⁷Er(n,γ)¹⁶⁸Er can be used to quantify the epithermal neutron fluences (Hidaka et al. 2020). Although the neutron-capture cross section of ¹⁶⁷Er is 2 orders of magnitude smaller than that of ¹⁴⁹Sm, the isotopic variations of ¹⁶⁷Er in most samples used in this study are expected to be detectable by consideration of the neutron fluences evaluated from the Sm isotopic variations. Isotopic data of Er are listed in Table 3. Nine of the thirteen samples, except Dhofar 911, NWA 4472, NWA 4932, and NWA 4936, show small but significant isotopic variations of ¹⁶⁷Er/¹⁶⁶Er and ¹⁶⁸Er/¹⁶⁶Er in the range of −4.0 < ¹⁶⁷Er < −0.4 and +0.8 < ¹⁶⁸Er <+4.3, respectively. It is evident from the isotopic data points shown in Figure 2(B) as the Er three-isotope plot diagram that the isotopic variations of ¹⁶⁷Er/¹⁶⁶Er and ¹⁶⁸Er/¹⁶⁶Er are caused by the neutron-capture reaction of ¹⁶⁷Er(n,γ)¹⁶⁸Er. All data points are plotted on the line with a slope of −1 in the figure, indicating that the isotopic decrements of ¹⁶⁷Er correspond to the isotopic increments of ¹⁶⁸Er resulting from the neutron-capture reaction of ¹⁶⁷Er(n,γ)¹⁶⁸Er.

The data of cosmogenic nuclides ¹⁰Be, ²⁶Al, ³⁶Cl, and ⁴¹Ca were partly reported in related meetings and conferences but have not been published yet (Nishiizumi & Caffee 2001, 2006; Nishiizumi et al. 2004, 2016; Hidaka et al. 2019). These data are compiled and listed in Table 1. The abundances of individual cosmogenic nuclides show the spread, such as 0.045 to 10.83 for ¹⁰Be, 0.334 to 57.5 for ²⁶Al, 0.10 to 9.74 for ³⁶Cl, and 2.57 to 215 for ⁴¹Ca, in dpm kg⁻¹ unit, as shown in Table 1, because their production rates vary with the chemical composition and the depth of the target material. Although the comprehensive collection of the SRLs data provide further constraints to understand the exposure histories of the meteorites, the low ⁴¹Ca activities in some samples make this difficult. The combination of the data set for cosmogenic nuclides provides information on the depth location of the meteorite species from the surface of their parent bodies (2π irradiation), and/or the preatmospheric size in space (4π irradiation) from a comparison of the measured data with the calculated data. There are two major models for the calculation of the production rates of SLRs based on the Los Alamos High Energy Transport Code System (LCS model by Masarik & Reedy 1994) and on the statistical estimation for a multielement target system (semiempirical model by Honda 1988; Nagai et al. 1993). Although the calculated data for the production rates of SLRs are somewhat model-dependent between LCS and semiempirical models (Welten et al. 2003), it is known that the measured data of SLRs for the Apollo-15 (A-15) lunar regolith core materials are in good agreement with those of calculated data by the LCS model (Nishiizumi et al. 1984a, 1984b, 1997) in the range of depth from the near surface to around 400 g cm⁻² on the Moon. In this study, estimations of the burial depths of individual lunar meteorites on the Moon were performed from a comparison of the measured SLR data with the calculated data by the LCS model. The estimated depths for individual samples are also listed in Table 1.

Because the production rates of SLRs such as ¹⁰Be, ²⁶Al, ³⁶Cl, and ⁴¹Ca caused by spallation and neutron-capture reactions vary along with the depth in the target materials for Galactic cosmic-ray (GCR) irradiation, the depth information of the target material can be estimated from the combination of the abundances of SLRs. It is known that most lunar meteorites have long CRE durations despite their short transition time from the Moon to the Earth (Nishiizumi et al. 1996). This suggests that most of them had experienced long CRE durations on the surface of the Moon. Because their residence durations on the surface of the Moon are considered to be generally longer than the lifetime of SLRs, their CRE ages cannot be estimated only from the SLR data. On the other hand, the ¹⁴⁹Sm–¹⁵⁰Sm isotopic shift is proportional to the CRE ages, because the neutron-capture reaction of ¹⁴⁹Sm(n,γ)¹⁵⁰Sm produces and accumulates the isotopic variations of the stable isotopes, ¹⁴⁹Sm and ¹⁵⁰Sm. The data set from a comprehensive combination between the abundances of SLRs and ¹⁵⁰Sm/¹⁴⁹Sm isotopic ratios in the same fragmental sample provides useful information on the CRE situations of individual lunar meteorites that resided on the surface of the Moon.

¹⁵⁰Sm/¹⁴⁹Sm isotopic ratios of the lunar surface materials vary with the depth, simply showing the frequencies of interactions between the atoms consisting of the materials and cosmic rays that penetrated the layer. The regolith materials of a 2.4 m length drilling core collected from the A-15 landing site are the best samples to understand isotopic evidence for the depth-dependence of the interaction of cosmic rays with planetary materials, because the samples had preserved the irradiation record without stratigraphical disturbance for around 450 Ma (Russ 1972; Hidaka et al. 2000b). Because the ¹⁵⁰Sm/¹⁴⁹Sm isotopic ratios in the A-15 regolith core show a smooth curve along with the depths, they are used as a function of the depth. Then, the fitting curve prepared from the data set of the A-15 regolith core can be used as a reference to calibrate the cosmic-ray exposure duration of individual lunar meteorites buried in a regolith layer on the moon (Hidaka et al. 2017).¹⁵⁰Sm/¹⁴⁹Sm isotopic variations corresponding to neutron fluences that are proportional to the GCR exposure duration are considered to be a function of the depth for the A-15 drill core samples. The ¹⁵⁰Sm/¹⁴⁹Sm isotopic ratios are proportional to the exposure duration.

First, the model for the single-stage GCR irradiation on the Moon would be applied for the individual lunar meteorites. While being exposed by cosmic rays, the lunar meteorites are assumed to have resided at certain depths for a long time before transportation to the earth. In this case, the depth on the Moon for each meteorite can be determined from the combination of the SLR data set, as shown in Table 1. As well as the SLRs abundances, ¹⁵⁰Sm/¹⁴⁹Sm isotopic variations related to the neutron-capture reaction show a function of depth (Russ 1972; Hidaka et al. 2000b). Based on the fitting curve of the ¹⁵⁰Sm/¹⁴⁹Sm isotopic variations as a function of depth for the A-15 regolith core, the CRE duration can be determined from the adjustment of the fitting curve by extension or contraction along with the direction of ¹⁵⁰Sm/¹⁴⁹Sm isotopic ratios. Figure 3 shows a correlation diagram between ¹⁵⁰Sm/¹⁴⁹Sm isotopic ratios and depth on the Moon. All data of individual lunar meteorites are from Tables 1 – 3.

Zoom In Zoom Out Reset image size

In our previous study, the data from NWA 482 and NWA 2995 are already reported, and support the single-stage irradiation model on the Moon (Hidaka et al. 2017). In this study, the data from 9 of the other 11 meteorites, Dhofar 081, Dhofar 910, NWA 2996, NWA 3136, NWA 3163, NWA 4472, NWA 4734, NWA 4884, and NWA 4936, can be explained by the single-stage irradiation model. On the assumption of the single-stage irradiation model, these nine meteorites had resided at depths from 50 to 330 g cm⁻² for 140 to 870 Ma. On the other hand, the data from two meteorites, Dhofar 911 and NWA 4932, cannot be explained by the single-stage irradiation model. The burial depths of these two meteorites are assigned at over 1000 g cm⁻² from the combination of the SLR abundances. However, it is unlikely to produce a significant variation of ¹⁵⁰Sm/¹⁴⁹Sm by neutron capture at the deeper level of over 1000 g cm⁻² in the Moon even for longer than 4500 Ma, considering the depth-dependence of Sm isotopic shifts observed in the lunar regolith drill core. It is reasonable to consider a multistage, at least two-stage, irradiation model to interpret the GCR irradiation situations for Dhofar 911 and NWA 4932. We propose the model, as a simple case, that these two meteorites had once resided at a shallower depth, preserved the Sm isotopic shifts, and then migrated to a deeper level over 1000 g cm⁻² before ejection from the Moon. In the shortest case at the first stage, the original materials for Dhofar 911 and NWA 4932 had been irradiated at a depth of 155 g cm⁻² for 38.1 Ma and 51.4 Ma, respectively, and then moved into a deeper level above 1000 g cm⁻² as the second stage before ejection from the Moon (see the right of Figure 3). The SLR records at the first stage disappeared because of their shorter half-life relative to the total irradiation duration but were preserved after migration to the deeper site at the second stage. On the other hand, the Sm isotopic variations had been accumulated since the first stage, although they were little affected at the second stage. In general, however, it is unlikely that the materials had migrated to a deeper level at the second stage. It is conceivable for Dhofar 911 and NWA 4932 that new regolith layers covered on the original layer after the first-stage irradiation by the gardening process and that two species apparently changed the depth to deeper levels.

Considering the numerous variations of neutron-capture cross sections along with the neutron energy, ¹⁶⁷Er reacts with both thermal and epithermal neutrons and changes its isotopic abundance by the neutron-capture reaction of ¹⁶⁷Er(n,γ)¹⁶⁸Er because of the large RI of ¹⁶⁷Er in the epithermal neutron energy region (0.5 eV < E < 2000 eV). On the other hand, ¹⁴⁹Sm sensitively reacts with thermal neutrons because of its large cross section in the thermal neutron energy region and changes its isotopic abundance by the neutron-capture reaction of ¹⁴⁹Sm(n,γ)¹⁵⁰Sm. In our previous study (Hidaka et al. 2020), thermal and epithermal neutron fluences at the lunar surface were estimated from the combination of Sm and Er isotopic variations. As a result, the neutron fluences in the epithermal region are more than 10 times larger than those in the thermal region in the regolith layer at the A-15 landing site (Hidaka et al. 2020). It is known that the energy balance of thermal and epithermal neutrons arising at the lunar surface depends upon the chemical composition of the material (Lawrence et al. 2006). Since the lunar meteorites show long CRE experience on the lunar surface, they are also expected to have similar neutron energy spectra to the lunar regolith. The volume of epithermal neutron fluence is enough to affect the change in the isotopic abundance of ¹⁴⁹Sm. The capture reaction for epithermal neutrons is not negligible for the estimation of thermal and epithermal neutron fluences from Sm isotopic data. In this study, the estimation of both thermal and epithermal neutrons can be performed from the combination of ¹⁵⁰Sm/¹⁴⁹Sm and ¹⁶⁸Er/¹⁶⁷Er isotopic ratios (Hidaka et al. 2020). We simply use the following equation for the calculation of the thermal and epithermal neutron fluences from the variations of ¹⁵⁰Sm/¹⁴⁹Sm and ¹⁶⁸Er/¹⁶⁷Er isotopic ratios:

$$\begin{eqnarray*}&&\sigma {\rm{\Psi }}={\sigma }_{{\rm{t}}{\rm{h}}}{{\rm{\Psi }}}_{{\rm{t}}{\rm{h}}}+{\sigma }_{{\rm{e}}{\rm{p}}{\rm{i}}}{{\rm{\Psi }}}_{{\rm{e}}{\rm{p}}{\rm{i}}}={\rm{l}}{\rm{n}}\left(\displaystyle \frac{1+R}{1+{R}_{0}}\right),\end{eqnarray*}$$

where σ_th is the cross section for thermal neutron capture, σ_epi is the cross section for epithermal neutron capture, Ψ_th is the thermal neutron fluence, Ψ_epi is the epithermal neutron fluence, R₀ and R are the isotopic ratios before and after neutron capture reactions (¹⁵⁰Sm/¹⁴⁹Sm for Sm and ¹⁶⁸Er/¹⁶⁷Er for Er). In this paper, ¹⁵⁰Sm/¹⁴⁹Sm = 0.533988 and ¹⁶⁸Er/¹⁶⁷Er = 1.17986, resulting from the measurements of the standard reference materials, can be used as R₀.

The results of thermal and epithermal neutron fluences for individual meteorites estimated from Sm and Er isotopic ratios are listed in Table 4. The epithermal neutron fluences of 4 of the 13 meteorites used in this study, Dhofar 911, NWA 4472, NWA 4932, and NWA 4936, cannot be estimated because of their small variation in the ¹⁶⁸Er/¹⁶⁷Er isotopic ratios within the analytical uncertainties (see Table 3). Because the thermal neutron-capture cross section of ¹⁴⁹Sm is very large, the estimation of thermal neutron fluences from the Sm isotopic variations is possible to quantify the fluence up to 10¹⁴ neutrons cm⁻². On the other hand, the isotopic variation of ¹⁶⁷Er caused by epithermal neutrons is less sensitive than that of ¹⁴⁹Sm. The detection limit of epithermal neutron fluence is at most 10¹⁷ neutrons cm⁻² roughly estimated from the analytical precisions of ¹⁶⁸Er/¹⁶⁷Er isotopic ratios in this study.

It is known that the energy spectrum of neutrons caused by GCR irradiation largely depends on the chemical composition of the target materials (Lingenfelter et al. 1972; Spergel et al. 1986). The bulk of target material is often treated as an effective macroscopic cross section (Σ_eff) evaluated from the elemental abundances of Si, Al, Fe, Ca, Mg, and Ti as major elements and Sm, Eu, and Gd as minor elements in the individual materials (Lingenfelter et al. 1972). Figure 4 shows a correlation diagram between the variation of neutron fluence ratio Ψ_epi/Ψ_th and the macroscopic cross section Σ_eff for individual meteorites. The variations of Ψ_epi/Ψ_th depending on Σ_eff are unclear for the data of the lunar meteorites. For comparison of the data between the meteorites and the A-15 regolith samples, all of the data points are plotted in the same figure. The data set in Figure 4 seems to have a trend in which larger Σ_eff result in larger energy balance between thermal and epithermal neutrons, considering that the A-15 regolith samples have higher _Er/_Sm and Σ_eff relative to the meteorites.

Zoom In Zoom Out Reset image size