Skip to main content
NCBI home page
NASA Author Manuscripts logoLink to NASA Author Manuscripts
. Author manuscript; available in PMC: 2020 Dec 2.
Published in final edited form as: Meteorit Planet Sci. 2020 Apr 9;55(4):771–780. doi: 10.1111/maps.13464

Origin and Age of Metal Veins in Canyon Diablo Graphite Nodules

Connor D Hilton a,*, Richard D Ash a, Philip M Piccoli a, David A Kring b, Timothy J McCoy c, Richard J Walker a
PMCID: PMC7709732  NIHMSID: NIHMS1647461  PMID: 33273799

Abstract

Previous studies attributed the origin of metal veins penetrating graphite nodules in the Canyon Diablo IAB main group iron meteorite to condensation from vapor or melting of host metal. Abundances of 16 siderophile elements measured in kamacite within vein and host meteorite are most consistent with an origin by melting of the host metal followed by fractional crystallization of the liquid. The presence of the veins within graphite nodules may be explained by impact, as peak shock temperatures, and thus the most likely areas to undergo metal melting, are at metal-graphite interfaces. The origin of the veins is constrained by Re-Os chronometry to have occurred early (>4 Ga) in Solar System history.

1. Introduction

The IAB iron meteorites are a large group of non-magmatic iron meteorites referred to as the IAB complex (Wasson and Kallemeyn, 2002). Some meteorites within this complex have disparate chemical and isotopic compositions, which has led to the division of the complex into several, possibly unrelated, sub-groups (Worsham et al., 2016, 2017). Of these, the “main group” (IAB-MG) is the largest with over 100 members. The IAB-MG consists of irons with diverse mineralogy and textures, which indicate a complex history. For example, IAB-MG metal hosts abundant silicates and graphite nodules, despite the strong density contrasts. The presence of the lower density phases has been interpreted to reflect impact-induced mixing of nonmetal and liquid metal components on the parent body, followed by the rapid cooling of metal inhibiting the gravimetric separation of these phases (e.g., Wasson and Kallemeyn, 2002).

The IAB-MG iron Canyon Diablo is noteworthy in that it contains massive graphite nodules, weighing as much as 10 kg (Garvie, 2016). These nodules are sometimes crosscut by metal veins (e.g., Buchwald, 1975; Garvie, 2016), the origins of which are debated. Characterization of the abundances of highly siderophile elements (HSE; Re, Os, Ir, Ru, Pt, Rh, Pd) within the vein by Hirata and Nesbitt (1997) revealed that parts of the vein metal are depleted in Re, Os, Ir, and Pt, compared to the host metal. This was interpreted to mean that the veins formed by the melting of host metal, followed by injection of the melt into the graphite nodules. Kurat et al. (2000) reported that thick veins (1 to 3 mm) are characterized by abundances of Re, Os and Ir that are depleted relative to the host metal, while thin veins (<0.2 mm) are enriched in these elements. Based on this chemical heterogeneity, they proposed that the veins formed by a “low temperature” process, such as chemical vapor deposition. This interpretation was subsequently supported by study of the noble gas compositions of the enclosing graphite nodules (Matsuda et al., 2005). That study concluded the graphite never reached temperatures above 900 K, based on its retained noble gas signatures.

In order to provide fresh insight to the origin of the metal veins in Canyon Diablo graphite nodules, we determine the abundances of 16 siderophile elements among different metal grains within the vein and host metal. The chemical compositions of these grains are then used to evaluate different potential vein origins. In addition, we apply the Re-Os chronometer (187Re → 187Os + β−, T1/2 = 41.6 Ga; Smoliar et al., 1996) to the vein to constrain the timing of formation. This chronometer has been broadly applied to determine the age of crystallization of iron meteorites and metallic phases in chondrites, as well as evaluate open-system behavior of HSE for metal in meteorites.

2. Methods

2.1. Sample preparation

A ~6 gram sample (Fig. 1) of Canyon Diablo (USNM 676), consisting of a graphite nodule with crosscutting metallic veins (hereafter referred to as vein metal), was obtained from the Division of Meteorites, Department of Mineral Sciences, Smithsonian Institution (SI) and prepared for separate in situ and bulk sample analyses. Pieces of the vein were cut using a water-cooled Leco ‘Vari-cut’ saw and a 12.7 cm diamond wafering blade. The blade was cleaned with carborundum before cutting each of the vein pieces. The surface of each piece of vein metal was polished using a range of coarse- to fine-grit SiC sandpaper and Al2O3 powder. The pieces were then sonicated multiple times in high purity water to remove polishing residue and then dried with ethanol. An in house thin section of Canyon Diablo (hereafter referred to as host metal), which was devoid of graphite and vein metal, was also used in this study. It was assumed to represent bulk Canyon Diablo metal that was unaffected by any process related to the formation of the vein metal.

Fig. 1.

Fig. 1.

Open in a new tab

Photograph of the graphite nodule fragment, from Canyon Diablo, hosting the metal veins analyzed in this study. A 2-cm scale in shown for reference. Also shown is a backscatter electron (BSE) image of the metal vein with 500 μm scale bar. K = kamacite. The BSE image is rotated relative to the photograph.

2.2. Canyon Diablo mineralogy

Kamacite (α-Fe,Ni), which is the dominant phase in the vein (Kurat et al., 2000) and host metal (Vdovykin, 1973), was analyzed for Fe, Ni, and Co concentrations using a JEOL 8900 electron probe microanalyzer (EPMA) at the Advanced Imaging and Microscopy Laboratory at the University of Maryland (UMd). The following operating conditions were utilized: accelerating voltage of 15 kV, sample current of 25 nA, and a beam diameter of 1 micron. A series of pure metals (Fe, Ni, Co) were used as primary standards. Raw x-ray intensities were corrected using a ZAF algorithm. Two-sigma absolute percent (here after abbreviated %) uncertainties were determined from counting statistics and range from 0.62–0.72 % for Fe, 0.70–4.6 % for Ni, and 6.5–27 % for Co. In situ analysis of siderophile element abundances was achieved by laser-ablation inductively coupled plasma mass spectrometry (LA-ICPMS), using a New Wave UP213 ultraviolet laser coupled to a Thermo Finnigan Element 2 at UMd. A 7 Hz, 55 μm laser spot size was used, and absolute concentrations were obtained from comparison with in house laboratory reference iron meteorites Hoba and North Chile, as well as the steel standard SRM 1263a, using Fe (determined by EPMA) as the element of reference.

2.3. Isotope dilution

Three pieces (0.159, 0.167, and 0.253 g) of the Canyon Diablo bulk vein material, each containing a range of graphite and metal, were prepared for isotope dilution measurement of Re and Os. These pieces, a Re-Os spike (185Re and 190Os), about 5 ml of concentrated HNO3 and 2.5 ml of concentrated HCl were combined in chilled Pyrex® Carius tubes. The tubes were sealed and heated at 220 °C for 72 hours. The Carius tubes were opened and their contents were transferred to centrifuge tubes, which were then centrifuged to remove graphite residue. Osmium was extracted from the solutions using a CCl4 solvent-extraction method (Cohen and Waters, 1996), and purified using the microdistillation technique of Birck et al. (1997). Purified Os samples were loaded onto outgassed Pt filaments in concentrated HBr, activated with Ba(OH)2, and analyzed as OsO3− by a Thermo-Fisher Triton thermal ionization mass spectrometer (TIMS) at UMd (Walker et al., 2008). Osmium isotopic data were corrected for instrumental and natural mass-fractionation by normalizing 192Os/188Os to 3.08271 (Allègre and Luck, 1980).

Rhenium was purified using anion exchange chromatography (Rehkämper et al., 1997). Each sample was dissolved in 1 N HCl and loaded onto disposable 2 ml Biorad columns filled with 1.7 ml of AG1X8 100–200 mesh resin. Sample matrix (Fe, Ni, Co) was eluted using 0.8 N HNO3 and 1 N HCl – 1 % HF, and Re was eluted in 6 N HNO3. The Re aliquots were purified further using an additional anion column (Walker et al., 2008), dried, dissolved in 0.8 N HNO3, doped with W in order to correct for instrumental mass bias, and analyzed on Faraday cups using Neptune Plus multi-collector inductively coupled plasma mass spectrometry (MC-ICP-MS) at UMd. The total analytical blanks for Re and Os are 0.6 and 1 pg, respectively, which constitute <<1 % of the quantity of these elements processed per sample and, therefore, have an inconsequential effect on the measured concentrations. The uncertainty for 187Os/188Os, Os concentrations, and Re concentrations are estimated to be ±0.1 %, and the corresponding uncertainty for 187Re/188Os is estimated to be ±0.15 % (Walker et al., 2008).

3. Results

The average Fe, Ni, and Co concentrations of kamacite grains (1 to 5 analyses per grain) of vein metal (41 total analyses) and host metal (12 total analyses) are reported in Table 1. The siderophile element compositions of kamacite grains (1 to 2 analyses per grain) in the vein (14 total analyses) and host (11 total analyses) metal, normalized to abundances in CI chondrites (Lodders, 2003), are shown in Fig. 2. The average vein kamacite and host kamacite siderophile element compositions are reported in Table 2. The enrichment/depletion factors of siderophile element abundances in average vein kamacite, compared to average host kamacite, are reported in Table 2. Rhenium, Os, Ir, and Pt are depleted in the average vein kamacite compared to average host kamacite by ≥60 %, while Rh, Ni, Co, and Cu are depleted by 6 % to 16 %. By contrast, Au and Ga are enriched in the average vein kamacite by 10 % to 20 %. Abundances of W, Mo, Ru, Pd, As, and Ge are identical, within uncertainty, in vein and host kamacite. Previously reported siderophile element abundances of bulk pieces of Canyon Diablo are reported in Table 2. The abundances of Ir, Pt, Rh, Ni, As, Ga, and Ge determined in the average host metal kamacite are within 5 % of the bulk meteorite abundances (Jochum et al., 1980; Wasson and Kallemeyn, 2002; Worsham et al., 2016). Abundances of Re, Ru, Co, Pd, and Au are within 11 % and abundances of Os, W, Mo, and Cu differ between 21 to 62 %.

Table 1.

Iron, Ni, and Co concentrations (wt. %) of kamacite in the vein and the host Canyon Diablo metal.

n Fe Ni Co Sum
Vein kamacite 41 93.3(0.2) 6.4(0.1) 0.64(0.01) 100.3 (0.2)
Host kamacite 12 91.3(0.4) 6.4(0.1) 0.68(0.02) 98.4 (0.4)
Open in a new tab

n - number of microprobe analyses of phases in the vein metal and the host metal. Uncertainties are reported as 2 standard error.

Fig. 2.

Fig. 2.

Open in a new tab

Comparison of siderophile element abundances, normalized to CI chondrites (Lodders, 2003), for vein kamacite grains (grey lines) compared to host kamacite (pink lines). The red and purple lines reflect the average host kamacite and vein kamacite compositions, respectively, and the blue line reflects a hypothetical initial kamacite composition that can produce the vein kamacite composition by melting and crystallization. Elements are listed from left to right in decreasing 50 % condensation temperatures (Lodders, 2003).

Table 2.

Siderophile element concentrations of kamacite in the vein and the host metal determined by LA-ICP-MS.

Vein Metal Kamacite (n = 14) Host Metal Kamacite (n = 11) Bulk Meteorite* Enrichment Factor
Re (ng/g) 80 (20) 210 (60) 231.7 0.4 (0.1)
Os (ng/g) 590 (130) 3650 (630) 2780 0.16 (0.05)
W (ng/g) 710 (130) 580 (100) 1060 1.2 (0.3)
Ir (ng/g) 460 (60) 2690 (270) 2560 0.17 (0.03)
Mo (μg/g) 4.17 (0.43) 3.71 (0.71) 9.8 1.1 (0.2)
Ru (ng/g) 4100 (660) 4980 (620) 5579 0.8 (0.2)
Pt (ng/g) 2530 (240) 6910 (460) 6959 0.37 (0.04)
Rh (ng/g) 1030 (80) 1230 (80) 1300 0.84 (0.08)
Ni (wt. %) 6.5 (0.2) 6.9 (0.2) 6.93 0.94 (0.04)
Co (wt. %) 0.48 (0.02) 0.51 (0.02) 0.463 0.94 (0.05)
Pd (ng/g) 3160 (270) 3150 (520) 3494 1.0 (0.2)
As (μg/g) 15.0 (1.8) 13.24 (1.02) 13 1.1 (0.2)
Au (ng/g) 1670 (110) 1430 (140) 1534 1.2 (0.1)
Cu (μg/g) 98.6 (4.8) 116.6 (6.8) 148 0.85 (0.06)
Ga (μg/g) 90.4 (3.1) 81.9 (2.3) 82.1 1.10 (0.05)
Ge (μg/g) 337.8 (14.6) 326.7 (9.9) 323 1.03 (0.05)
Open in a new tab

n represents the number of analyses of each phase. Enrichment factors are the ratio of siderophile element abundance in vein kamacite to host kamacite. Uncertainties represent 2 standard error.

*

Data are from Worsham et al. (2016) (Re, Os, Ir, Ru, Pt, Pd), Jochum et al. (1980) (Mo, Rh), and Wasson and Kallemeyn (2002) (other elements).

Rhenium-Os isotopic data obtained for pieces of the bulk vein metal (metal and included graphite) are reported in Table 3. Isotopic data for the metal vein pieces plot within −0.1 to +9.3 per mil of a 4.568 Ga reference isochron, calculated from an initial Solar System 187Os/188Os = 0.09517 and λ = 1.666 × 10−11 yr−1 (Archer et al., 2014; Smoliar et al., 1996).

Table 3.

Rhenium and Os concentrations (ng/g) and isotopic data determined by isotope dilution. Host metal data are from Worsham et al. (2016).

Mass Re Os 187Os/188Os 187Re/188Os ΔOs
Vein
Vein metal 1 0.167 102.1 951 0.13612 0.5180 −0.1
Vein metal 2 0.159 117.8 1123 0.13517 0.5059 −0.1
Vein metal 3 0.253 74.6 669 0.13868 0.5385 +9.3
Host
Host metal 0.071 231.7 2780 0.12723 0.4016 +3.0
Open in a new tab

Two-sigma uncertainties are ±0.1 % for Re and Os concentrations and 187Os/188Os, and ±0.15 % for 187Re/188Os. ΔOs is the deviation of 187Os/188Os of a sample from a 4.568 Ga reference isochron in units of per mil, with an uncertainty of ±2 per mil (Cook et al., 2004).

4. Discussion

4.1. Composition of vein metal

The kamacite grains in the metal vein are characterized by a range of siderophile element compositions (Fig. 2), consistent with the conclusion by Kurat et al. (2000) that the metal veins are chemically heterogeneous. However, unlike Kurat et al. (2000), we observed no vein metal enriched in Re, Os, Ir, and Pt, compared to host metal. The kamacite in the host metal is also characterized by a range of siderophile element compositions (Fig. 2). Overall, the average kamacite composition of the host metal is similar to published compositions of the bulk Canyon Diablo meteorite. The notable deviations between host metal kamacite and bulk meteorite abundances may be explained by the preferential partitioning of certain elements into other phases present in Canyon Diablo, such as W and Mo into cohenite ([Fe,Ni,Co]3C) (see supplemental materials). Differences for Cu and Os may be explained by higher abundances in troilite (FeS) (Vdovykin, 1973) and taenite (γ-Fe,Ni) (see supplemental materials), respectively, compared to kamacite.

4.2. Origin of vein metal

Previously proposed origins of the metal veins in Canyon Diablo graphite nodules include formation by condensation processes (Kurat et al., 2000) and formation by melting processes (Hirata and Nesbitt, 1997). Kurat et al. (2000) proposed an origin of the metal veins through a “low temperature” condensation event, such as chemical vapor deposition. This formation mechanism was favored based on the chemical heterogeneity observed among different veins from a single graphite nodule. It was argued that a “high-temperature” process, such as the melting of metal, would lead to a homogeneous vein composition. A “low-temperature” origin, and thus support for chemical vapor deposition, was furthered by Matsuda et al. (2005), in which it was argued that the retention of noble gas signatures by the graphite nodules provided evidence that the bulk nodules did not exceed temperatures above ~900 K.

To account for the presence of the metal-rich graphite nodules within Canyon Diablo host metal, Kurat et al. (2000) proposed that the graphite nodules and metal veins formed first, and then host metal was deposited around the nodules. The authors observed no correlation between the abundances of siderophile elements in the vein and their 50 % condensation temperatures (Tc), which was interpreted to suggest that the abundances of elements must instead reflect the volatilities of certain carrier phases for these elements. The results of the present study confirm that there is no correlation of siderophile element abundance in vein metal with 50 % condensation temperature. For example, W (Tc 1789 K) and Au (Tc 1060 K) are not depleted in the vein compared to the host metal, but Pt (Tc 1408) is. However, the results of this study do not support an origin of metal veins through the condensation of certain carrier phases. The major host phases of Re, Os, W, Ir, and Pt in a gas with a Solar System composition are refractory metal alloys (Lodders, 2003), yet there are no depletions observed for W in the vein compared to the host metal while there are depletions for Re, Os, Ir, and Pt. Further, if the vein metal formed before the host metal, as proposed by Kurat et al. (2000), then the vein metal should be characterized by either enrichments or equal abundances of refractory siderophile elements to the host metal, as the parental vapor to Canyon Diablo would have become depleted in refractory elements as the condensation sequence progressed. This is not observed, making the interpretation of the origin of the veins before the host metal by condensation problematic.

The origin of metal veins by condensation processes has also been proposed for ordinary chondrites (Widom et al., 1986; Rubin, 1999). This type of origin, however, differs from that proposed by Kurat et al. (2000) in that the veins present in ordinary chondrites formed after the host meteorite. Widom et al. (1986) and Rubin (1999) observed depletions in Re, Os, Ir, and Pt abundances in metal veins relative to fine grained ordinary chondrite metal, while no depletions were observed for other siderophile elements, including W and Mo. Based on these observations, these authors proposed that the metal veins in ordinary chondrites formed by fractional condensation of a vapor, possibly formed by impact. Silicates adjacent to metal in ordinary chondrites were proposed to have also vaporized, leading to the oxidation of W and Mo. In this scenario, refractory elements, including Re, Os, Ir, and Pt condensed from the vapor near the site of vaporization, while other elements, including volatile Mo and W oxides, were transported further and deposited into cracks. Similar chemical trends, including depletions in Re, Os, Ir, and Pt, with no depletions in W or Mo, are observed for the Canyon Diablo vein metal compared to host metal. Silicates are relatively rare in Canyon Diablo, however, compared to ordinary chondrites, limiting the oxidation potential of any vapor and making this proposed origin improbable.

Hirata and Nesbitt (1997) proposed an origin of the metal veins by melting of host metal. These authors observed a positive correlation between the ionic radii of HSE and the enrichment of HSE abundances in vein kamacite compared to kamacite in the host metal, which was interpreted to reflect the incompatibility of HSE in solid metal relative to liquid metal. As such, Hirata and Nesbitt (1997) proposed that in situ shock melting of the host metal resulted in HSE with larger ionic radii (Rh, Ru, and Pd) entering the melt phase more readily than those with smaller ionic radii (Ir, Os, Pt, and Re), which were largely retained in the solid residue. The melt phase was then injected into graphite nodules and cooled rapidly. Considerable experimental work to constrain the solid metal-liquid metal partitioning behavior of these elements conducted since the results of Hirata and Nesbitt (1997) allow for a re-examination of a model of host metal melting.

In order to test an origin by melting of host metal, we model melting of host Canyon Diablo metal and subsequent fractional crystallization of the melt (see supplemental materials for full model description). Experimental results show that the presence of S, P, and/or C in metal systems strongly affect the solid metal-liquid metal D values of siderophile elements (e.g., Willis and Goldstein, 1982; Jones and Drake, 1983; Malvin et al., 1986). Therefore, it is important to consider the effects of these elements when modeling metal behavior during melting of light element-rich systems such as Canyon Diablo. Partitioning studies by Jones and Malvin (1990), Chabot and Jones (2003), and Chabot et al. (2017) provide experimental constraints and methods for calculating the D values of different siderophile elements for various S, P, and C contents. Parental melt concentrations of 1 wt. % S, 0.26 wt. % P, and 1 wt. % C are assumed here, based on abundances of these elements previously determined for the bulk Canyon Diablo meteorite (Buchwald, 1975). Since kamacite is the dominant host phase of most siderophile elements analyzed in the host metal, we use the average siderophile element abundances in the host metal kamacite as the composition of melted Canyon Diablo.

Calculated liquid metal and solid metal compositions at 10 % increments of fractional crystallization are shown in Fig. 3a and 3b, respectively. The vein metal kamacite composition is broadly similar to the composition of liquid metal calculated after 60 % fractional crystallization. However, this model does not account for the chemical composition of all siderophile elements observed in the vein metal. Notable deviations appear for Ru (27 ± 15 %), As (66 ± 24 %), Cu (25 ± 10 %), and Ga (31 ± 3 %) between the vein kamacite composition from this study and the calculated liquid composition after 60 % fractional crystallization, even when considering uncertainties on the respective D values (Chabot et al., 2017). While no variations between Ru and As abundances are resolved between the average vein kamacite and average host kamacite, the results of the model predict that a liquid after 60 % fractional crystallization would be depleted in Ru and enriched in As. A minor depletion of Cu and a minor enrichment of Ga in the vein kamacite compared to host kamacite are observed, yet the model predicts that Cu should become enriched and Ga depleted in the vein. One possible explanation for the deviation of four of the 16 siderophile element compositions from the model is that a model of metal melting and fractional crystallization of the melt is incorrect. Another possible explanation is that the correct initial melt composition is not used in the model. As shown in Fig. 2, there is a range of siderophile element compositions measured in host kamacite. Adjusting the initial melt composition to reflect the range of host kamacite compositions results in better fits of the model to Ru, As, Cu, and Ga abundances, while causing worse fits for other elements. As such, a fractional crystallization model can account for all 16 abundances of siderophile elements determined within the vein, but only for an initial host metal composition that was not sampled in this study. Such a composition is plausible given the range of host kamacite compositions observed (Fig. 2). In addition, localized heterogeneities of S, P, and C concentrations for the melt source compared to the bulk Canyon Diablo meteorite may also contribute to deviations of the fractional crystallization model from the vein composition, as the fractional crystallization model is sensitive to the initial abundances of these elements. For example, the potentially unlimited source of C from the graphite nodules may affect the fractional crystallization of melted host metal in a way unaccounted for by the model presented here. The effect of changing the initial C content from 1 wt. % to 5 wt. %, for example, on the D values of the 16 siderophile elements is highlighted in Table S4.

Fig. 3.

Fig. 3.

Open in a new tab

(a) Liquid and (b) solid metal compositions, normalized to CI, calculated from 0 % to 70 % fractional crystallization of a parental liquid with the composition of the host kamacite (red line; initial liquid). Black lines represent the solid composition at 10 % intervals of fractional crystallization from 10 % crystallization to 70 % crystallization.

An alternative pathway to produce a liquid with a composition identical to that after 100 % host metal melting and 60 % fractional crystallization is by 40 % partial melt of host metal. Given the chemical similarity between the vein kamacite and this liquid composition, the data from this study allow for either model. However, if the chemical compositions of veins measured by Kurat et al. (2000) are also considered, then a model of 40 % partial melt of host metal can be eliminated, as it fails to account for the enrichments of Re, Os, Ir, and Pt observed by Kurat et al. (2000) for some vein metal compared to host metal. These enrichments can be explained, however, by a model of 100 % host metal melting and 60 % fractional crystallization. As shown in Fig. 3b, initial solids produced by fractional crystallization are enriched in refractory elements compared to the host metal. Thus, this model can produce metal compositions with enrichments and depletions of Re, Os, Ir, and Pt compared to host metal. Alternatively, the metal veins analyzed by Kurat et al. (2000) may have formed by a similar but independent event to the veins from this study. If this is correct, it suggests that a single graphite nodule could contain metal veins from multiple melting and crystallization episodes. This possibility could be addressed by future work through comparison of multiple metal veins present in a single graphite nodule.

An origin of the metal veins by host metal melting and fractional crystallization of the melt is most consistent with the abundances of 16 siderophile elements measured in this study and Kurat et al. (2000). As melting kamacite requires temperatures >1700 K (Swartzendruber et al., 1991), this temperature must be reconciled with the conclusions of Matsuda et al. (2005) that the surrounding graphite nodules did not experience temperatures >900 K. One possible explanation for these different temperature requirements is that graphite and kamacite have different specific heats. At 1 atm and 25°C, graphite has a specific heat of 709 J K−1 kg−1 while Fe has a specific heat of 449 J K−1 kg−1 (Lide, 2005). As such, it is plausible that host kamacite could melt and be injected into graphite nodules without raising the temperature of the graphite to that of the liquid. One way to cause metal melting within an iron meteorite is by impact. Shock waves induced by impact can result in significant heating at the boundary between phases with significant density contrasts (Stöffler et al., 1991). As such, the highest temperatures from shock pressures in Canyon Diablo would likely be expected at the interface of graphite and metal. Therefore, it is possible that an impact to Canyon Diablo could cause increased temperatures at the graphite-metal boundary, resulting in melting of host metal and injection of melt into the nodules. The validity of this model could be addressed by future studies through the textural and chemical study of host metal immediately surrounding graphite nodules.

4.3. Timing of vein formation

This study concludes that the metal veins likely formed by host metal melting, possibly induced by impact, and fractional crystallization of the liquid. As such, this interpretation requires that the metal veins formed after the host metal, which formed ~4.56 Ga (Worsham et al., 2016), either on the IAB-MG parent body or after Canyon Diablo was separated from this body. If the former, then multiple proposed impact events to the IAB-MG parent body could be related to the formation of the veins. Past studies have proposed that different IAB subgroups formed by impact events at ~4.56 Ga (Wasson and Kallemeyn, 2002; Worsham et al., 2017; Hunt et al., 2018). An impact event at ~4.55 Ga to the IAB-MG parent body, causing metamorphism of silicates and resetting of the Pd-Ag chronometer, has also been proposed (Schulz et al., 2009; Theis et al., 2013). If vein formation occurred after Canyon Diablo separated from the IAB-MG parent body, then one of three known, comparatively recent impact events to Canyon Diablo may be responsible, including at ~0.54 Ga and ~0.17 Ga (Heymann et al., 1966), and the Earth impact event around 0.00005 Ga (Sutton, 1985; Phillips et al., 1991).

Reference isochrons assuming melting of the host metal occurred at 4.0 Ga, 2.5 Ga, and 1.0 Ga are shown in Fig. 4. If formation of the vein had occurred at one of these ages, data for the pieces of vein would be expected to plot along the appropriate reference isochrons. Instead, the data for the three pieces of vein metal plot within ±10 per mil of a 4.568 Ga reference isochron (Fig. 4), which indicates formation >4 Ga when considering corresponding uncertainties on 187Re-187Os values. The ±10 per mil scatter of the vein metal 187Re-187Os isotopic systematics around this isochron suggests that the vein metal experienced variable, minor open-system behavior, resulting in the partial redistribution of Re and/or Os.

Fig. 4.

Fig. 4.

Open in a new tab

(Top) 187Re/188Os vs. ΔOs plot, where ΔOs is the parts per 10,000 deviation of the 187Os/188Os value of a sample from a 4.568 reference isochron, for vein metal pieces (this study) and a host metal piece of Canyon Diablo (Worsham et al., 2016). (Bottom) 187Re/188Os vs. 187Os/188Os plot for vein metal pieces and a host metal piece of Canyon Diablo, which fall along a 4.568 Ga reference isochron calculated from the isotopic systematics of chondrites (Archer et al., 2014). Reference isochrons for 4.0 Ga, 2.5 Ga and 1.0 Ga are shown for comparison.

Since the timing of vein formation is not resolved from the timing of host metal formation, it is currently impossible to determine if the vein formed on the original IAB-MG parent body or after a smaller piece of the parent was separated from the main body. To determine this, future studies could examine similar features in other IAB-MG irons, such as silicate-graphite-metal inclusions described in Campo del Cielo (Kurat et al., 2009) and troilite-graphite inclusions described in Odessa and Toluca (Benedix et al., 2000). If such features can be related to a similar event by chemical compositions and chronometers, it may provide evidence for vein formation on the IAB-MG parent body. Either way, multiple impact events for other meteorites are recorded around this time as well, including for ordinary chondrites (Weirich et al., 2011; Swindle et al., 2013), HED achondrites (Kennedy et al., 2013), IIE iron meteorites (Bogard et al., 2000), and IVA iron meteorites (Yang et al., 2007). Evidently, this was a period of abundant impacts for early planetary bodies, which is permissive of metal vein formation through impact-induced shock heating.

Conclusions

Metal veins present in graphite nodules from Canyon Diablo are dominated by the occurrence of kamacite, which are most notably characterized by depletions in Re, Os, Ir, and Pt abundances, compared to host metal kamacite. An origin of the vein by condensation is not consistent with the relative abundances of refractory siderophile elements in the vein. By contrast, abundances of most siderophile elements examined can be accounted for by a model that invokes the melting of host metal and subsequent fractional crystallization of the liquid. Melting of host metal may have been induced by impact, as the highest shock temperatures in the metal are expected at the graphite-host metal interface. Results from Re-Os chronometry indicate that the vein formed early (>4 Ga) in the Solar System, during a period of frequent impacts.

Supplementary Material

1

Acknowledgements

We gratefully acknowledge the Division of Meteorites, Department of Mineral Sciences, Smithsonian Institution for providing access to the graphite-metal vein sample for this study. Dr. Igor Puchtel, Dr. Katherine Bermingham, and Dr. Emily Worsham are also thanked for their assistance in the laboratory. We also thank Katherine Joy, Thomas Harvey, and two anonymous reviewers for helpful comments that improved the manuscript. We acknowledge the support of the Maryland NanoCenter and its AIMLab. This study was partially supported by NASA SSERVI grant NNA14AB07A and NASA Emerging Worlds grant NNX16AN07G.

References

  1. Allègre CJ and Luck J-M (1980) Osmium isotopes as petrogenetic and geological tracers. Earth Planet. Sci. Lett 48, 148–154. [Google Scholar]
  2. Archer GJ, Ash RD Bullock ES, Walker RJ (2014) Highly siderophile elements and 187Re-187Os isotopic systematics of the Allende meteorite: Evidence for primary nebular processes and late-stage alteration. Geochim. Cosmochim. Acta 131, 402–414. [Google Scholar]
  3. Benedix GK, McCoy TJ, Keil K, Love SG (2000) A petrologic study of the IAB iron meteorites: Constraints on the formation of the IAB-Winonaite parent body. Meteorit. Planet. Lett 35, 1127–1141. [Google Scholar]
  4. Berkey E and Fisher DE (1967) The abundance and distribution of chlorine in iron meteorites. Geochim. Cosmochim. Acta 31, 1543–1558. [Google Scholar]
  5. Birck J-L, Roy-Barman M, Capmas F (1997) Re-Os isotopic measurements at the femtomole level in natural samples. Geostand. Newslett 20, 9–27. [Google Scholar]
  6. Bogard DD, Garrison DH, McCoy TJ (2000) Chronology and petrology of silicates form IIE iron meteorites: evidence of a complex parent body evolution. Geochim. Cosmochim. Acta 64, 2133–2154. [Google Scholar]
  7. Buchwald VF (1975) Handbook of Iron Meteorites. University of California Press, Berkeley, CA. [Google Scholar]
  8. Buchwald VF and Clarke RS Jr. (1989) Corrosion of Fe-Ni alloys by Cl-containing akaganéite (β-FeOOH): The Antarctic meteorite case. American Mineralogist 74, 656–667. [Google Scholar]
  9. Chabot NL and Jones JH (2003) Parameterizing iron meteorite partitioning experiments. Meteorit. Planet. Sci. 37, 1425–1436. [Google Scholar]
  10. Chabot NL, Wollack EA, McDonough WF, Ash RD, Saslow SA (2017) Experimental determination of partitioning in the Fe-Ni system for applications to modeling meteoritic metals. Meteorit. Planet. Sci 52, 1133–1145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Cohen AS and Waters FJ (1996) Separation of osmium from geological materials by solvent extraction for analysis by thermal ionization mass spectrometry. Anal. Chim. Acta 332, 269–275. [Google Scholar]
  12. Cook DL, Walker RJ, Horan MF, Wasson JT, Morgan JW (2004) Pt-Re-Os systematics of group IIAB and IIIAB iron meteorites. Geochim. Cosmochim. Acta 68, 1413–1431. [Google Scholar]
  13. Garvie LAJ (2016) Observations on a 10-KG graphite nodule from the Canyon Diablo (IAB-MG) iron meteorite. Lunar Planet. Sci. Conf. XLVII, 2567 (abst). [Google Scholar]
  14. Heymann D, Lipschutz ME, Nielsen B, Anders E, 1966. Canyon Diablo meteorite: metallographic and mass spectrometric study of 56 fragments. J. Geophys. Res 71, 619–641. [Google Scholar]
  15. Hirata T and Nesbitt RW (1997) Distribution of platinum group elements and rhenium between metallic phases of iron meteorites. Earth Planet. Sci. Lett 147, 11–24. [Google Scholar]
  16. Hunt AC, Cook DL, Lichtenberg T, Reger PM, Ek M, Golabek GJ, Schönbächler M (2018) Late metal-silicate separation on the IAB parent asteroid: Constraints from combined W and Pt isotopes and thermal modelling. Earth Planet. Sci. Lett 482, 490–500. [Google Scholar]
  17. Jochum KP, Seufert M, Begemann F (1980) On the distribution of major and trace elements between metal and phosphide phases of some iron meteorites. Zeitschrift für Naturforschung A 35, 57–63. [Google Scholar]
  18. Jones JH and Drake MJ (1983) Experimental Investigations of Trace Element Fractionation in Iron Meteorites, II: The influence of sulfur. Geochim. Cosmochim. Acta 47, 1199–1209. [Google Scholar]
  19. Jones JH and Malvin DJ (1990) A nonmetal interaction model for the segregation of the trace metals during solidification of Fe–Ni–S, Fe–Ni–P, Fe–Ni–S–P alloys. Metall. Trans. B 21B, 697–706. [Google Scholar]
  20. Kennedy T, Jourdan F, Bevan AWR, Gee MAM, Frew A (2013) Impact history of the HED parent body(ies) clarified by new 40Ar/39Ar analyses of four HED meteorites and one anomalous basaltic achondrite. Geochim. Cosmochim. Acta 115, 162–182. [Google Scholar]
  21. Kurat G, Sylvester P, Kong P, Brandstatter F (2000) Heterogeneous and fractionated metal in Canyon Diablo (IA) graphite-metal rock. Lunar Planet. Sci. Conf. XXXI, 1666 (abst). [Google Scholar]
  22. Kurat G, Zinner E, Varela ME, Ntaflos T (2009) SIGRMET05: a silicate-graphite-metal inclusion from the Campo del Cielo (IAB) iron. Lunar Planet. Sci. Conf. XL, 1536 (abst). [Google Scholar]
  23. Lide DR, ed., CRC Handbook of Chemistry and Physics, Internet Version 2005, CRC Press, Boca Raton, FL, 2005. [Google Scholar]
  24. Lodders K (2003) Solar System abundances and condensation temperatures of the elements. Ap. J. 591, 1220–1247. [Google Scholar]
  25. Malvin DJ, Jones JH, Drake MJ (1986) Experimental investigations of trace-element fractionation in iron-meteorites. III. Elemental partitioning in the system Fe-Ni-S-P. Geochim. Cosmochim. Acta 50, 1221–1231. [Google Scholar]
  26. Matsuda J, Namba M, Maruoka T, Matsumoto T, Kurat G (2005) Primordial noble gases in a graphite-metal inclusion from the Canyon Diablo IAB iron meteorite and their implications. Meteorit. Planet. Sci 40, 431–443. [Google Scholar]
  27. Phillips FM, Zreda MG, Smith SS, Elmore D, Kubik PW, Dorn RI, Roddy DJ (1991) Age and geomorphic history of Meteor Crater, Arizona, from cosmogenic 36Cl and 14C in rock varnish. Geochim. Cosmochim. Acta 55, 2695–2698. [Google Scholar]
  28. Rehkämper M, Halliday AN, Barfod D, Godfry FJ and Dawson BJ (1997) Platinum-group element abundance patterns in different mantle environments. Science 278, 1595. [DOI] [PubMed] [Google Scholar]
  29. Rubin AE (1999) Formation of large metal nodules in ordinary chondrites. J. of Geophys. Res 104, 30799–30804. [Google Scholar]
  30. Schulz T, Münker C, Palme H, Mezger K (2009) Hf-W chronometry of the IAB iron meteorite parent body. Earth Planet. Sci. Lett 280, 185–193. [Google Scholar]
  31. Scott ERD and Goldstein JI (2012) Occurrence of carbides and graphite in iron meteorites and origin of C-rich irons. Lunar Planet. Sci. Conf. XLIII, 2671 (abst). [Google Scholar]
  32. Smoliar MI, Walker RJ, Morgan JW (1996) Re–Os ages of group IIA, IIIA, IVA, and IVB iron meteorites. Science 271, 1099–1102. [Google Scholar]
  33. Stöffler D, Keil K, Scott ERD (1991) Shock metamorphism of ordinary chondrites. Geochim. Cosmochim. Acta 55, 3845–3867. [Google Scholar]
  34. Sutton SR (1985) Thermoluminescence measurements on shock-metamorphosed sandstone and dolomite from Meteor Crater, Arizona 2. Thermoluminescence age of meteor crater. J. Geophys. Res 90, 3690–3700. [Google Scholar]
  35. Swartzendruber LJ, Itkin VP, Alcock CB (1991) The Fe-Ni (iron-nickel) system. J. Phase Equilibria 12, 288–312. [Google Scholar]
  36. Swindle TD, Kring DA, Weirich JR (2013) 40Ar/39Ar ages of impacts involving ordinary chondrite meteorites. Geological Society, London, Special Publications 378, 333–347. [Google Scholar]
  37. Theis KJ, Schönbächler M, Benedix GK, Rehkämper M, Andreasen R, Davies C (2013) Palladium-silver chronology of IAB iron meteorites. Earth Planet. Sci. Lett 361, 402–411. [Google Scholar]
  38. Vdovykin GP (1973) The Canyon Diablo Meteorite. Space Science Reviews 14, 758–831. [Google Scholar]
  39. Walker RJ, McDonough WF, Honesto J, Chabot NL, McCoy TJ, Ash RD and Bellucci JJ (2008) Modeling fractional crystallization of group IVB iron meteorites. Geochim. Cosmochim. Acta 72, 2198–2216. [Google Scholar]
  40. Wasson JT and Kallemeyn GW (2002) The IAB iron-meteorite complex: A group, five subgroups, numerous grouplets, closely related, mainly formed by crystal segregation in rapidly cooling melts. Geochim. Cosmochim. Acta 66, 2445–2473. [Google Scholar]
  41. Weirich JR, Wittmann A, Isachsen CE, Rumble D, Swindle TD, Kring DA (2011) The Ar-Ar age and petrology of Miller Range 05029: Evidence for a large impact in the very early solar system. Meteorit. Planet. Sci 45, 1868–1888. [Google Scholar]
  42. Widom E, Rubin AE, Wasson JT (1986) Composition and formation of metal nodules and veins in ordinary chondrites. Geochim. Cosmochim. Acta 50, 1989–1995. [Google Scholar]
  43. Willis J and Goldstein JI (1982) The effects of C, P, and S on trace element partitioning during solidification in Fe-Ni alloys. Proceedings, 13th Lunar and Planetary Science Conference pp. A435–A445. [Google Scholar]
  44. Worsham EA, Bermingham KR, Walker RJ (2016) Siderophile element systematics of IAB complex iron meteorites: new insights into the formation of an enigmatic group. Geochim. Cosmochim. Acta 188, 261–283. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Worsham EA, Bermingham KR, Walker RJ (2017) Characterizing cosmochemical materials with genetic affinities to the Earth: Genetic and chronological diversity within the IAB iron meteorite complex. Earth Planet. Sci. Lett 467, 157–166. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Yang J, Goldstein JI, Scott ERD (2007) Iron meteorite evidence for early formation and catastrophic disruption of protoplanets. Science 446, 888–891. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1647461-supplement-1.pdf (502.1KB, pdf)

RESOURCES