A Low Abundance of ¹³⁵Cs in the Early Solar System from Barium Isotopic Signatures of Volatile-depleted Meteorites

Short-lived radioisotopes such as ²⁶Al, ⁶⁰Fe, and ¹⁸²Hf that were present in the early solar system can be used for high-precision relative chronometry of early solar system events, but also provide evidence regarding the type of astrophysical environment in which the solar system formed. Such pursuits often lead to a better understanding about the origin and evolution of our solar system; however, both of these themes require accurate and precise knowledge of the abundance of the chosen radioisotope at the start of the solar system. There are two primary methods for determining the initial abundance of short-lived radionuclides in the early solar system. First, the isotope systematics are measured in the first solids to form in the solar system, calcium–aluminum-rich inclusions (CAIs), thus directly establishing an initial abundance in these original solids (e.g., Jacobsen et al. 2008; Kruijer et al. 2014; Marks et al. 2014). The second common method involves measuring the chosen isotope systematics of later-formed solids—such as differentiated bodies like angrites or eucrites with a known absolute age—and back-calculating to the initial abundance at the time of CAI formation (e.g., Boyet et al. 2010; Tang & Dauphas 2012). Whereas both of these methods have been used successfully, determining the initial abundance of some radioisotopes thought to have been present in the early solar system, such as ¹³⁵Cs, have remained problematic.

The short-lived radioisotope ¹³⁵Cs (t_1/2 = 2.3 Myr) decays to ¹³⁵Ba and, given the much higher solubility of Cs in aqueous environments compared to Ba, the Cs–Ba system has the potential to be a useful tool to investigate aqueous processes that occurred in the early solar system. Yet, perhaps more importantly than its potential as an early solar system chronometer, knowledge of the initial abundance of ¹³⁵Cs is important for understanding the stellar environment in which the solar system formed. Interestingly, of all of the commonly discussed short-lived radionuclides that were present the early solar system, ¹³⁵Cs is the only one produced primarily by s-process nucleosynthesis. This production could have occurred in asymptotic giant branch (AGB) stars with the abundance of ¹³⁵Cs following galactic chemical evolution (GCE), although ¹³⁵Cs may also be produced in SNe II (Hidaka & Yoneda 2013). Thus, understanding the abundance of ¹³⁵Cs at the start the solar system provides critical information about the stellar input and possible supernovae activity in the star-forming region of solar system formation. For instance, higher initial amounts of ¹³⁵Cs (¹³⁵Cs/¹³³Cs > ∼10⁻⁵) in the early solar system would require an additional stellar source of ¹³⁵Cs since the amount of ¹³⁵Cs would be higher than that expected by GCE. For lower initial solar system values, where ¹³⁵Cs/¹³³Cs < 10⁻⁵, the abundance of ¹³⁵Cs could be explained by simple GCE, and no extraneous input of stellar material would be required (i.e., Harper 1996; Nittler & Dauphas 2006, pp. 127–146; Young 2014). Recent estimates of the initial ¹³⁵Cs/¹³³Cs ratio range from 1 × 10⁻⁵ to 2 × 10⁻³ (Hidaka et al. 2001; Nichols et al. 2002; Srinivasan et al. 2011; Hidaka & Yoneda 2013), producing a huge amount of uncertainty in the initial ¹³⁵Cs/¹³³Cs of the solar system, severely restricting both its potential usefulness as a chronometer and its position as an informant about the solar system's protostellar neighborhood.

Determination of the initial solar system ¹³⁵Cs/¹³³Cs using CAIs is challenging because Cs is volatile and is therefore strongly depleted in refractory inclusions. This, coupled with nucleosynthetic excesses of ¹³⁵Ba and ¹³⁷Ba in CAIs (Harper et al. 1992; Brennecka et al. 2013; Bermingham et al. 2014), makes determining the initial ¹³⁵Cs/¹³³Cs directly in CAIs extremely problematic for the Cs–Ba system. A second possible approach to establish the initial abundance of ¹³⁵Cs is to seek samples with a known age after CAI formation with a variety of phases exhibiting a range of parent/daughter ratios. This type of work generally targets phases that are enriched in the parent isotope relative to the chondritic ratio (given as supra-chondritic ¹³³Cs/¹³⁶Ba) to have measurable excesses in the daughter isotope, ¹³⁵Ba. Given this, a correlation of radiogenic ¹³⁵Ba with the measured Cs/Ba ratio in the investigated phases can produce an isochronous relationship, allowing the calculation of the initial ¹³⁵Cs/¹³³Cs of the solar system. However, the exceedingly low Cs/Ba ratios of many early solar system objects along with the high mobility of Cs creating disturbances in the Cs/Ba have to this point hampered establishing a robust initial ¹³⁵Cs/¹³³Cs ratio.

Given the previous difficulties in locating samples with undisturbed supra-chondritic Cs/Ba (e.g., Hidaka et al. 2001; Hidaka & Yoneda 2013; Bermingham et al. 2014), here we employ an alternative approach in determining the initial ¹³⁵Cs/¹³³Cs of the solar system. Whereas previous works have searched for samples with high parent–daughter (Cs/Ba) ratios to calculate the initial ¹³⁵Cs/¹³³Cs ratio, early solar system material with a sub-chondritic abundance of Cs would conversely result in a deficit in the daughter isotope compared to the chondritic value. Provided the timing is independently known, it is possible to use a radiogenic deficit caused by an elemental fractionation event as a means of calculating the initial abundance of the short-lived radioisotope. The major process leading to fractionation of Cs and Ba in meteorites is volatile depletion, because Cs is a moderately volatile element, whereas Ba is refractory, with 50% condensation temperatures of 799 K and 1447 K, respectively (Lodders 2003). Thus, if the short-lived ¹³⁵Cs was present in significant quantities in the early solar system, as postulated by previous workers, then volatile-depleted bodies of the early solar system would lack the majority of the ¹³⁵Cs decay to ¹³⁵Ba, and thus have a sub-chondritic ¹³⁵Ba/¹³⁶Ba, as demonstrated in Figure 1.

Zoom In Zoom Out Reset image size

In this work, we investigate samples with decidedly sub-chondritic Cs/Ba ratios in an attempt to better constrain the initial ¹³⁵Cs/¹³³Cs ratio of the solar system. Angrite and eucrite meteorites derive from ancient and differentiated planetary bodies with a well-established chronology from a variety of short- and long-lived chronometers (e.g., Glavin et al. 2004; Amelin 2008; Brennecka & Wadhwa 2012; Kleine et al. 2012; Hans et al. 2013; Sanborn et al. 2015). Both angrites and eucrites have been long known to be some of the most volatile-depleted samples in the solar system (e.g., Papanastassiou & Wasserburg 1969; Wasserburg et al. 1977; Lugmair & Galer 1992; Nyquist et al. 1994) and the loss of volatile elements for these two bodies has been shown to occur within 1 Ma of CAI formation (Hans et al. 2013). Because the timing of the volatile depletion is known from the Rb–Sr system (Hans et al. 2013; Sanborn et al. 2015), a system directly above the Cs–Ba system on the periodic table and therefore chemically analogous, the Ba isotopic signatures of angrites and eucrites can be used to robustly establish the initial ¹³⁵Cs/¹³³Cs of the solar system.

In this work, we investigate the Ba isotope systematics of a variety of volatile-depleted meteorites from the early solar system, including five eucrites (Stannern, Millbillillie, Bouvante, Petersburg, and Juvinas), four angrites (NWA 4801, NWA 6291, Angra dos Reis, and D'Orbigny), two unique achondrites (GRA 06129 and Bunburra Rockhole), and an aubrite (Shallowater). Sample specifics are given in Table 1. Samples and the geologic standard BHVO-2 were dissolved using common mineral acids under clean lab conditions (details further described in Goldmann et al. 2015). Aliquots of each dissolved sample were taken for trace element measurement prior to any chemistry and were measured using a Thermo X-series quadrupole ICP-MS. For isotopic measurement, Ba was isolated from the matrix following the chemical procedure outlined by Carlson et al. (2007). In short, this procedure utilizes AG50W-X8 cation resin and various strengths of HCl and HNO₃ to separate Ba from the matrix elements. Once chemically purified, the Ba isotopic composition of the samples and standards were measured on a Thermo TRITON Plus instrument at the Institut für Planetologie at the University of Münster. Due to the setup of the specific instrument, the alignment of all Ba isotopes and required interference monitors was not possible on a single line and therefore no data are reported for ¹³²Ba. However, all other isotopes of Ba, along with La and Ce, which have direct isobaric interferences with Ba, were monitored during the measurements. The interference corrections (reported in Table 2) were found to be at baseline levels for all samples of this study, and thus no correction was applied. All Ba isotopes are reported relative to ¹³⁶Ba and are internally normalized to ¹³⁴Ba/¹³⁶Ba = 0.3078 (Carlson et al. 2007). Each sample was run multiple times, and each run consisted of 300–600 ratios with 16 s integration times. The reported reproducibility is based on 2× the standard deviation (2SD) of multiple measurements of the geologic standard BHVO-2. Data from all runs for samples and standards of this study are given in Table 2.

The trace element data obtained on the meteoritic samples of this study confirm the volatile-depleted nature of the chosen groups, with Cs/Ba values of more than two orders of magnitude below the chondritic value (Table 1). This Cs depletion is analogous to the depletion of Rb in angrites and eucrites, and the correlated behavior of the Rb/Sr and Cs/Ba ratios are displayed in Table 1 and Figure 2. The unique achondrites GRA 06129 and Bunburra Rockhole as well as the aubrite Shallowater also have lower-than-chondritic Cs/Ba ratios (Lodders 2003); however, they are less volatile depleted than the angrites or eucrites of this study and are not plotted in Figure 2.

Zoom In Zoom Out Reset image size

No Ba isotopic variations outside of analytical uncertainty are reported for any samples of this study. Isotopic results are reported in ε-notation, or deviations from the standard in parts per 10,000, and are shown for ¹³⁵Ba in Table 1 and Figure 3. All other measured isotopes of Ba are reported in Table 2.

Zoom In Zoom Out Reset image size

Previous studies have reported Ba isotopic compositions for the eucrites Juvinas (Andreasen & Sharma 2007; Bermingham et al. 2014, 2016), Stannern (Bermingham et al. 2014, 2016), and Piplia Kalan (Srinivasan & Ali 2010), which are indistinguishable from terrestrial standards and in agreement with our findings. Taken together, later-formed, less primitive solar system samples do not appear to have ¹³⁵Ba anomalies at the current level of precision of 0.15ε.

Unlike CAIs that have well-documented nucleosynthetic anomalies in Ba (Harper et al. 1992; Brennecka et al. 2013; Bermingham et al. 2014), and unlike chondritic meteorites that contain isotopically anomalous presolar grains (Savina et al. 2003; Ávila et al. 2013), angrites and eucrites represent melted (and thus likely homogenized) material with no recognized nucleosynthetic anomalies in Ba (Andreasen & Sharma 2007; Srinivasan & Ali 2010; Bermingham et al. 2014, 2016). This lack of resolvable nucleosynthetic anomalies allows for a more straightforward interpretation of the Cs–Ba isotope systematics discussed below.

Owing to their volatile-depleted nature and, hence, low Cs/Ba ratio, angrites and eucrites would potentially have deficits in the ¹³⁵Ba/¹³⁶Ba ratio, where the magnitude of this deficits depends on the timing of volatile loss and the initial ¹³⁵Cs/¹³³Cs of the solar system. Thus, if the timing of volatile loss is known independently, the size of the ¹³⁵Ba deficit measured in angrites and eucrites, or lack thereof, can be used to constrain the initial ¹³⁵Cs/¹³³Cs of the solar system. Prior work has shown that the initial ⁸⁷Sr/⁸⁶Sr of angrites and eucrites is indistinguishable from that of CAIs, indicating that the low Rb/Sr ratio of these meteorites and their parent bodies were established within 1 Ma after CAI formation (Hans et al. 2013). This timing is consistent with evidence from the short-lived ¹⁸²Hf-¹⁸²W system, indicating that accretion of the angrite and eucrite parent bodies occurred within <1 Ma after solar system formation (Kleine et al. 2012; Touboul et al. 2015). Combined, these observations indicate that the volatile-depleted nature of the angrite and eucrite parent bodies reflects their early accretion from volatile-poor dust (Hans et al. 2013), meaning that the low Cs/Ba of the angrites and eucrites were established within <1 Ma after CAI formation.

If the initial solar system value of ¹³⁵Cs/¹³³Cs = 4.8 × 10⁻⁴ (Hidaka et al. 2001) is assumed, samples at the start of the solar system (T₀) would have a markedly lower ¹³⁵Ba/¹³⁶Ba value at T₀ due to the lack of radiogenic ¹³⁵Ba. If a major volatile depletion event occurred <1 Ma after T₀ on the angrite and eucrite parent bodies—as demonstrated using the Rb–Sr system (Hans et al. 2013)—this event would deprive angrite and eucrite parent bodies of further significant radiogenic ¹³⁵Ba and one would predict a measured ¹³⁵Ba of approximately 4.3ε (shown by the red arrow in Figure 4). Instead, this study found that the measured average ¹³⁵Ba/¹³⁶Ba of angrites and eucrites is indistinguishable from that of terrestrial standards or the rest of the inner solar system with a ε¹³⁵Ba = 0.04 ± 0.06 (2SD, N = 9). When this value and its associated uncertainty (i.e., ¹³⁵Ba > −0.02ε) is combined with the volatile depletion event at <1 Ma after CAI formation (Hans et al. 2013), an upper limit of 2.8 × 10⁻⁶ for the initial solar system ¹³⁵Cs/¹³³Cs can be calculated (Figure 4).

Zoom In Zoom Out Reset image size

The latest time at which the volatile depletion (or low Cs/Ba) of the angrites and eucrites could have possibly been established is given by the crystallization age of the oldest sample. The quenched angrite D'Orbigny represents the most ancient sample of the suite and has a well-established crystallization age of ∼5 Ma after CAI formation (Amelin 2008; Brennecka & Wadhwa 2012). In a hypothetical endmember scenario, if one assumes Cs/Ba fractionation in the angrites occurred at 5 Ma after CAI formation, and until that point the angrites evolved with a CI-chondritic Cs/Ba ratio—i.e., the highest possible ratio—a solar system initial ¹³⁵Cs/¹³³Cs = 4.8 × 10⁻⁴ (Hidaka et al. 2001) would result in angrites exhibiting a deficit of ∼1.4ε in ¹³⁵Ba. Thus, even when assuming an unrealistically late volatile depletion and an unrealistically high Cs/Ba ratio, a significant and easily resolvable ¹³⁵Ba deficit in the angrites would be expected if the solar system initial ¹³⁵Cs/¹³³Cs were as high as ∼4.8 × 10⁻⁴. Consequently, the absence of resolvable ¹³⁵Ba deficits in angrites and eucrites require a much lower initial ¹³⁵Cs/¹³³Cs of the solar system, which most likely was <2.8 × 10⁻⁶ based on our data and a calculated time of volatile element depletion on the angrite and eucrite parent bodies.

Previous studies have reported slight correlations between the measured Ba isotopic composition and ¹³³Cs/¹³⁶Ba of various samples, leading both to higher estimations of the initial ¹³⁵Cs/¹³³Cs ratio (Hidaka et al. 2001; Nichols et al. 2002; Hidaka & Yoneda 2013), as well as model age estimations for the aqueous alteration on ordinary chondrite parent bodies (e.g., Hidaka et al. 2001). Based on multiple data sets, ¹³⁵Cs was present at a level high enough in the early solar system to produce measurable variations in samples by concentrating Cs in certain phases that are known to have high amounts of aqueous alteration, like Samaya chondrites (Hidaka & Yoneda 2013), but these works do not appear to produce robust isochrons with chronologically meaningful information. The scatter in such diagrams is most easily explained by altering the Cs/Ba ratio due to the extreme mobility of Cs in aqueous environments, or perhaps more likely by the leaching process itself during sample preparation.

Our finding of a much lower solar system initial ¹³⁵Cs/¹³³Cs than previously inferred has important implications for understanding the sources of short-lived radionuclides and the stellar environment of solar system formation. Importantly, the previously published values (¹³⁵Cs/¹³³Cs = 4.8 × 10⁻⁴ and higher) would require an influx of ¹³⁵Cs beyond that which would be inherited from GCE models, as shown in Hidaka & Yoneda (2013), and would thus require an outside source of short-lived radionuclides, such as input from an AGB star or recent material from SNe II. However, the solar system initial value of ¹³⁵Cs/¹³³Cs < 2.8 × 10⁻⁶ determined in the present study is significantly lower than previous estimates, and as such no longer requires extraneous input of stellar material to explain the value beyond basic GCE. As keenly pointed out by Young (2014, 2016), when corrected for input from Wolf–Rayet (WR) winds in the dense phases of star-forming regions, radionuclides of the early solar system align along a single trend and are consistent with a residence time in the molecular cloud of 200 ± 100 Ma, which is similar to the present-day rate of molecular cloud conversion into stars in the Milky Way. With a decrease in the initial ¹³⁵Cs/¹³³Cs of greater than two orders of magnitude, ¹³⁵Cs now joins this list of radionuclides that are simply obtained from the existing molecular cloud, fitting along the previously established trend (Figure 5). Thus, ¹³⁵Cs is no longer an outlier requiring a unique stellar source; it is simply another radionuclide present in the early solar system that was inherited from the surrounding molecular cloud—the most likely statistical scenario of solar system formation.

Zoom In Zoom Out Reset image size

This work was supported by a Sofja Kovalevskaja award from the Alexander von Humboldt Foundation (GAB).