Meteoritics & Planetary Science 1–17 (2017) doi: 10.1111/maps.13027
The Chelyabinsk meteorite: New insights from a comprehensive electron microscopy and Raman spectroscopy study with evidence for graphite in olivine of ordinary chondrites 2 € David KAETER 1,2,7*, Martin A. ZIEMANN1, Ute BOTTGER , Iris WEBER3, Lutz HECHT4, 5 5 Sergey A. VOROPAEV , Alexander V. KOROCHANTSEV , and Andrey V. KOCHEROV6 1
Institute for Earth and Environmental Sciences, University of Potsdam, 14476 Potsdam, Germany 2 Institute of Optical Sensor Systems, German Aerospace Center, 12489 Berlin, Germany 3 Institute of Planetology, University of M€ unster, 48149 M€ unster, Germany 4 Museum f€ ur Naturkunde, 10115 Berlin, Germany 5 Vernadsky Institute of Geochemistry and Analytical Chemistry, Russian Academy of Sciences, Moscow 119334, Russia 6 Chelyabinsk State University, Chelyabinsk 454001, Russia 7 Present Address: iCRAG, School of Earth Sciences, University College Dublin, Dublin D04 N2E5, Ireland * Corresponding author. E-mail:
[email protected] (Received 16 March 2017; revision accepted 09 November 2017)
Abstract–We present results of petrographic, mineralogical, and chemical investigations of three Chelyabinsk meteorite fragments. Three distinct lithologies were identified: light S3 LL5, dark S4–S5 LL5 material, and opaque fine-grained former impact melt. Olivine– spinel thermometry revealed an equilibration temperature of 703 23 °C for the light lithology. All plagioclase seems to be secondary, showing neither shock-induced fractures nor sulfide-metal veinlets. Feldspathic glass can be observed showing features of extensive melting and, in the dark lithology, as maskelynite, lacking melt features and retaining grain boundaries of former plagioclase. Olivine of the dark lithology shows planar deformation features. Impact melt is dominated by Mg-rich olivine and resembles whole-rock melt. Melt veins (500 lm are rare. Typical Raman spectra of major phases are shown in Fig. 3: Olivine shows its characteristic doublet peak (DB) between 818–825 cm 1 (DB1) and 847–856 cm 1 (DB2). Its exact position depends on Fe-Mg concentrations. Clinopyroxene is rare in the primary lithology, but common in melt and friction veins. The structural
difference of ortho- and clinopyroxene results in a doublet for orthopyroxene between 660–680 cm 1 and a weak peak at approximately 235 cm 1, where clinopyroxene shows only a single peak and no peak around 235 cm 1, respectively. Feldspar shows a characteristic doublet at ~480 and ~510 cm 1 and only broad bands at ~790 cm 1 and 950–1150 cm 1. The spectra resemble those of hightemperature plagioclase of intermediate endmember composition. Reference spectra can be found in Freeman et al. (2008). The spinel is of the chromite group and differs mainly in its Al contents, which is expressed by a shift of the characteristic peak from 680 to 690 cm 1 of pure chromite to higher wave numbers with increasing substitution of Cr3+ by Al3+ (Wang et al. 2004). Chromite occurs mainly as xenomorphic, rounded grains (50–200 lm) commonly associated with troilite and Fe-Ni metal. Small idiomorphic grains (80 mole%) than olivine of the light lithology (~72 mole%). Orthopyroxene is Mg-rich as well (~En76). Olivine compositions calculated after Kuebler et al. (2006) from DB1 and DB2 Raman peak positions are shown in Fig. 6. The dark and light lithologies overlap in their average olivine compositions. Olivine compositions derived from Raman spectroscopy are in agreement with EMPA data (Fig. 7) as well as SEM-EDS data by Morlok et al. (2017) for primary olivine (~Fo71) and olivine crystallized from impact melt (Fo80–87). Fusion Crust The vesicular fusion crust of C-FL is 50–100 lm thick (Fig. 8a) and dominated by idiomorphic Mg-rich olivine (Fo89) ≤5 lm in a glassy matrix. Dendritic crystals of Fe-Cr oxides are common. Residual chromite overgrown by magnetite can be observed (Fig. 8b, inset).
Neither native metal nor sulfides were identified. Opaque shock veins and fractures in the adjacent primary lithology were annealed over an area of approximately 100 lm (Fig. 8c). Shock Metamorphism C-FL was only weakly deformed. Olivine and pyroxene often show undulose extinction and most irregular fractures are present in most minerals. Planar fractures and weak mosaicism in olivine are rare. Troilite is polycrystalline with approximately 120° grain boundaries (Fig. 9). Three parallel very fine-grained, opaque friction veins with minor offset cross the thin section. Their shear direction is indicated by arrows in Fig. 9. They are similar to terrestrial pseudotachylites and appear in weakly shocked rocks (Van der Bogert et al. 2003). Sulfidic-metallic shock veins occur near the margins (Fig. 9, upper right). Most plagioclase resembles former melt pockets as it lacks fractures, is often amorphous, and forms the aforementioned chromite– plagioclase assemblages. The light colored parts of C-LK show the same shock features, with planar fractures in olivine being more common. Shearing can also be observed (Fig. 10a). The transition to the dark lithology is gradual instead of abrupt. The dark lithology in C-FD and -LK is heavily brecciated and individual clasts are separated by impact melt veins. It appears nearly opaque due to fine, 900 °C (Ostertag and St€ offler 1982). A higher crystallinity of graphite in the dark lithology compared to the light lithology would be consistent with recrystallization at elevated postshock temperatures, but more grains need to be analyzed for statistical conclusions. The partial annealing of opaque shock veins observed near impact melt veins indicates that the impact melt may have prevented the material from fast cooling or even reheated it. Additionally, younger shock events can overprint mineral assemblages of earlier shock events. Considering the complex impact history with at least eight different shock events over the last 3.7 Ga (Righter et al. 2015), it is likely that only the most recent events were texturally preserved. It is possible that all shocked lithologies we see in Chelyabinsk fragments today were produced by only one impact event at 1.7 Ga, as it was recently proposed (Badyukov
Chelyabinsk: Metamorphism and evidence for graphite
11
Fig. 12. Photomicrographs of maskelynite (a) and regular glass (b) in reflected light. Fractures and shock veins are lacking in both phases. Grain boundaries of the regular glass are typical for former melt. Ol = olivine; Px = pyroxene; Spl = Cr-spinel; Tro = troilite.
et al. 2015; Korochantseva et al. 2017). The authors suggest that the Chelyabinsk meteorites are impactites comparable to rocks formed in impact craters on Earth. However, this does not exclude that other impact events happened to the Chelyabinsk parent body, which were not necessarily preserved. The weak shock deformation of C-FL may be explained by the pseudotachylitic friction veins, where shear stress was locally concentrated, while the rest of the material experienced only low stress. In addition, material of different shock stages is produced by different distances to the actual impact, and ejecta material of different shock stages and petrologic types can be deposited closely together. Furthermore, shock waves do not propagate homogeneously through meteorites because of variable sound velocities in different minerals as well as fractures and cavities (Sharp and DeCarli 2006). This is a likely explanation for the occurrence of weakly shocked light lithology right next to moderately to strongly shocked dark lithology in C-LK. Impact Melting The impact melt lithology is not overprinted by younger impact events and was therefore probably produced by one of the youngest shock events. Badyukov et al. (2015) also concluded that the dark lithology and the impact melt were produced at the same
event. They estimated melt temperatures to be as high as 2330 °C at peak pressures and over 1400 °C after pressure release. The latter is consistent with our observations: only partially molten residual olivine within the melt veins and absence of chromite melting suggest melt temperatures below the whole-rock liquidus temperature for LL chondrites of 1620 °C as calculated by Yamaguchi et al. (1998). However, it is still possible that higher temperatures were reached locally. Structural zoning of the larger melt veins is similar to structures often observed in granitic pegmatites, which commonly show a fine-grained boarder zone, a zone of inward-growing elongated crystals, and central zones with idiomorphic crystals and miarolitic cavities (e.g., London [2014] and references therein). It is remarkable that these textures can be produced at such different scales (micrometer versus meter) and compositions (ultramafic versus granitic). The textures were probably produced by fast crystallization from a supercooled melt. Liquidus undercooling of pegmatites was proposed by London (2009), who reproduced similar textures experimentally. Abundant sulfide-metal droplets were produced by liquid immiscibility of metal-rich sulfidic melt and silicate melt. During cooling, metal segregated from the liquid forming metal droplets inside sulfide droplets. Vesicles were likely produced by the release of sulfur as gaseous species, indicating that the impact melt is more oxidized than the primary lithology.
12
D. Kaeter et al.
Fig. 13. BSE images of impact melt veins separating the brecciated dark lithology of C-FD. a) Melt vein with rounded olivine clasts, sulfide-metal droplets, cavities, and chilled margins. The arrow shows Fe-depletion of residual olivine. b) Sulfide-metal droplets in impact melt vein. White Fe-Ni metal inclusions in gray sulfide spheres. Vesicles (arrow) were formed by a fluid phase. c) Narrow veinlet connected to larger melt vein. FeNi = Fe-Ni metal; Ol = olivine; Pl = plagioclase; Px = pyroxene; Spl = spinel; Tro = troilite.
The Fall: Melting in Earth’s Atmosphere It was estimated that the surface of the Chelyabinsk meteorite was heated to over 2000 °C (Petrova et al. 2016). This is supported by our observation of high-Mg
olivine and the melting of chromite within the fusion crust. Because of these extreme temperatures, a possible effect on primary mineral composition by frictional heating in the Earth’s atmosphere was taken into account when choosing measurement locations for
Chelyabinsk: Metamorphism and evidence for graphite
13
heat conductors and their inner temperatures are near interplanetary temperatures, the frictional heating only affects the surface and they cool down again shortly after the fall (McSween 1999). The occurrence of magnetite and the lack of sulfides and metal in the fusion crust reflect the strong oxidation during melting. Vesicles were likely produced by the incorporation of air and additionally due to oxidation and release of sulfur as gaseous species. Graphite: Traces of Nebular Processes in Ordinary Chondrites
Fig. 14. Schematic structure of impact melt veins as observed in C-FD (Fig. 13a).
EMPA. Examination of the fusion crust revealed that material 100 lm below the molten surface lacks any obvious alteration. Therefore, minerals can be safely analyzed if they are far enough away from the fusion crust (i.e., >200 lm). The affected volume seems quite small, regarding the high temperatures needed to produce a mafic whole-rock melt, but as rocks are poor
Fig. 15. BSE image of sulfide-metal melt veins in C-LK.
In carbonaceous chondrites, total carbon contents can reach up to 5.4 wt% (Pearson et al. 2006), while ordinary chondrites show 35 GPa) than previously thought. Although Morlok et al. (2017) reported a feldspar-normative groundmass in the impact melt lithology, these authors cannot proof unequivocally the presence of maskelynite. The heterogeneous, brecciated nature of the material and especially the large amounts of chondritic melt also argue for intense shock deformation. The observation of liquid segregation of silicate and sulfide-metal melt indicates that impact melting is probably an important factor in the differentiation of primitive planetary bodies. Combining our results and the results of former studies, it is likely that the Chelyabinsk parent body was a rubble pile asteroid compromised of material of different shock stages with low to moderate shock degree penetrated by impact melt. However, it is possible that the history of the Chelyabinsk parent body was less complex than proposed earlier and that all observed shock and melt lithologies were produced by a single impact event (Badyukov et al. 2015; Korochantseva et al. 2017). While this might be the case, it does not argue against a rubble pile structure of the parent body. Graphite in melt inclusions of olivine was so far only described for carbonaceous chondrites. Our description provides a starting point for the search of similar inclusions in other ordinary chondrites. Geochemical studies of such inclusions can provide important insights into their origin and the environment of chondrule formation. Carbon isotopic studies of graphite could be employed to test a possible presolar origin.
16
D. Kaeter et al.
Confocal Raman spectroscopy is a reliable method for the analysis of meteorites, because of its high spatial resolution and distinct spectra of each mineral. To date, there is no other method capable of characterizing inclusions and identifying carbonaceous material without exposure on the sample surface or destruction of the entire sample. Comparison to EMPA data from this study and quantitative SEM-EDS data by Morlok et al. (2017) shows that endmember compositions of olivine inferred from Raman spectra can be a good first estimate for average compositions, if good calibrations exist. Acknowledgments—We thank U. Heitmann, University of M€ unster, Germany, and H. Busemann, University of Manchester, England, for sample preparation; A. Greshake, Museum f€ ur Naturkunde Berlin (MfN), Germany, for helpful discussions; and P. Czaja, MfN, for his technical assistance at the museum’s electron microprobe facilities. We also thank P. Heck and M. Ivanova for their helpful reviews and insightful comments as well as A. J. T. Jull for the editorial handling of the paper. Editorial Handling—Dr. A. J. Timothy Jull REFERENCES Anders E. and Zinner E. 1993. Interstellar grains in primitive meteorites: Diamond, silicon carbide, and graphite. Meteoritics 28:490–514. Badyukov D. D., Raitala J., Kostama P., and Ignatiev A. V. 2015. Chelyabinsk meteorite: Shock metamorphism, black veins and impact melt dikes, and the Hugoniot. Petrology 23:103–115. Bennett M. E. and McSween H. Y. 1996. Shock features in iron-nickel metal and troilite of L-group ordinary chondrites. Meteoritics & Planetary Science 31:255–264. Bischoff A., Horstmann M., Vollmer C., Heitmann U., and Decker S. 2013. Chelyabinsk—Not only another ordinary LL5 chondrite, but a spectacular chondrite breccia (abstract #5171). Meteoritics & Planetary Science 48(Suppl.). Borovicka J., Spurny P., Brown P., Wiegert P., Kalenda P., Clark D., and Shrbeny L. 2013. The trajectory, structure and origin of the Chelyabinsk asteroidal impactor. Nature 503:235–237. Brearley A. J. 1990. Carbon-rich aggregates in type 3 ordinary chondrites: Characterization, origins, and thermal history. Geochimica et Cosmochimica Acta 54:831–850. Brown P. G., Assink J. D., Astiz L., Blaauw R., Boslough M. B., Borovicka J., Brachet N., Brown D., Campbell-Brown M., Ceranna L., Cooke W., De Groot-Hedlin C., Drob D. P., Edwards W., Evers L. G., Garces M., Gill J., Hedlin M., Kingery A., Laske G., Le Pichon A., Mialle P., Moser D. E., Saffer A., Silber E., Smets P., Spalding R. E., Spurny P., Tagliaferri E., Uren D., Weryk R. J., Whitaker R., and Krzeminski Z. 2013. A 500-kiloton airburst over Chelyabinsk and an enhanced hazard from small impactors. Nature 503:238–241.
Busemann H., Alexander M. O’D., and Nittler L. R. 2007. Characterization of insoluble organic matter in primitive meteorites by micro Raman spectroscopy. Meteoritics & Planetary Science 42:1387–1416. Chen M. and El Goresy A. 2000. The nature of maskelynite in shocked meteorites: Not diaplectic glass but a glass quenched from shock-induced dense melt at high pressures. Earth and Planetary Science Letters 179:489– 502. Evans B. W. and Frost B. R. 1975. Chrome-spinel in progressive metamorphism—A preliminary analysis. Geochimica et Cosmochimica Acta 39:959–972. Fabries J. 1979. Spinel-olivine geothermometry in peridotites from ultramafic complexes. Contributions to Mineralogy and Petrology 69:329–336. Freeman J. J., Wang A., Kuebler K. E., Jolliff B. L., and Haskin L. A. 2008. Characterization of natural feldspars by Raman spectroscopy for future planetary exploration. The Canadian Mineralogist 46:1477–1500. Fritz J., Greshake A., and St€ offler D. 2005. Micro-Raman spectroscopy of plagioclase and maskelynite in Martian meteorites: Evidence of progressive shock metamorphism. Antarctic Meteorite Research 18:96–116. Galimov E. M. 2013. Chelyabinsk meteorite—An LL5 chondrite. Solar System Research 47:255–259. Galimov E. M., Kolotov V. P., Nazarov M. A., Kostitsyn Y. A., Kubrakova I. V., Kononkova N. N., Roshchina I. A., Alexeev V. A., Kashkarov L. L., Badyukov D. D., and Sevast’yanov V. S. 2013. Analytical results for the material of the Chelyabinsk meteorite. Geochemistry International 51:522–539. Kocherov A. V., Korochantsev A. V., Lorenz C. A., Ivanova M. A., and Grokhovsky V. I. 2014. Recovery, laboratory preparation and current state of the main mass of the Chelyabinsk meteorite (abstract #2227). 45th Lunar and Planetary Science Conference. CD-ROM. Kohout T., Gritsevich M., Grokhovsky V. I., Yakovlev G. A., Haloda J., Halodova P., Michallik R. M., Penttil€ a A., and Muinonen K.. 2014. Mineralogy, reflectance spectra, and physical properties of the Chelyabinsk LL5 chondrite— Insight into shock-induced changes in asteroid regoliths. Icarus 228:78–85. Korochantseva V., Buikin A. I., Hopp I., Ivanova M. A., Korochantsev A. V., and Trieloff M. 2017. Formation time of Chelyabinsk breccia and its subsequent thermal history (abstract #1567). 48th Lunar and Planetary Science Conference. CD-ROM. Kuebler K. E., Jolliff B. L., Wang A., and Haskin L. A. 2006. Extracting olivine (Fo–Fa) compositions from Raman spectral peak positions. Geochimica et Cosmochimica Acta 70:6201–6222. London D. 2009. The origin of primary textures in granitic pegmatites. The Canadian Mineralogist 47:697–724. London D. 2014. A petrologic assessment of internal zonation in granitic pegmatites. Lithos 184–187:74–104. Makjanic J., Touret J. L. R., Vis R. D., and Verheul H. 1989. Carbon in glass inclusions in the Allende meteorite. Meteoritics 24:49–55. Makjanic J., Vis R. D., Hovenier J. W., and Heymann D. 1993. Carbon in the matrices of ordinary chondrites. Meteoritics 28:63–70. McSween H. Y. 1999. Meteorites and their parent planets, 2nd ed. Cambridge: Cambridge University Press. 324 p.
Chelyabinsk: Metamorphism and evidence for graphite
17
Moore C. B. and Lewis C. F. 1967. Total carbon content of ordinary chondrites. Journal of Geophysical Research 72:6289–6292. Morlok A., Bischoff A., Patzek M., Sohn M., and Hiesinger H. 2017. Chelyabinsk—A rock with many different (stony) faces: An infrared study. Icarus 284:431–442. Ostertag R. and St€ offler D. 1982. Thermal annealing of experimentally shocked feldspar crystals. Journal of Geophysical Research: Solid Earth 87(Suppl.):A457–A463. Ozawa S., Miyahara M., Ohtani E., Koroleva O. N., Ito Y., Litasov K. D., and Pokhilenko N. P. 2014. Jadeite in Chelyabinsk meteorite and the nature of an impact event on its parent body. Scientific Reports 4:5033. Pearson V. K., Sephton M. A., Franchi I. A., Gibson J. M., and Gilmour I. 2006. Carbon and nitrogen in carbonaceous chondrites: Elemental abundances and stable isotopic compositions. Meteoritics & Planetary Science 41:1899–1918. Petrova E. V., Grokhorvsky V. I., and Muftakhetdinova R. F. 2016. Heat treatment of the different structure zones in the Chelyabinsk meteorite (abstract #6487). 79th Annual Meeting of the Meteoritical Society. Popova O. P., Jenniskens P., Emel’yanenko V., Kartashova A., Biryukov E., Khaibrakhmanov S., Shuvalov V., Rybnov Y., Dudorov A., Grokhovsky V. I., Badyukov D. D., Yin Q.-Z., Gural P. S., Albers J., Granvik M., Evers L. G., Kuiper J., Kharlamov V., Solovyov A., Rusakov Y. S., Korotkiy S., Serdyuk I., Korochantsev A. V., Larionov M. Y., Glazachev D., Mayer A. E., Gisler G., Gladkovsky S. V., Wimpenny J., Sanborn M. E., Yamakawa A., Verosub K. L., Rowland D. J., Roeske S., Botto N. W., Friedrich J. M., Zolensky M. E., Le L., Ross D., Ziegler K., Nakamura T., Ahn I., Lee J. I., Zhou Q., Li X.-H., Li Q.-L., Liu Y., Tang G.-Q., Hiroi T., Sears D., Weinstein I. A., Vokhmintsev A. S., Ishchenko A. V., Schmitt-Kopplin P., Hertkorn N., Nagao K., Haba M. K., Komatsu M., Mikouchi T., and the Chelyabinsk Airburst Consortium. 2013. Chelyabinsk airburst, damage assessment, meteorite recovery, and characterization. Science 342:1069–1073. Reddy V., Vokrouhlicky D., Bottke W. F., Pravec P., Sanchez J. A., Gary B. L., Klima R., Cloutis E. A., Galad A., Guan T. T., Hornoch K., Izawa M. R. M., Kusnirak P., Le Corre L., Mann P., Moskovitz N., Skiff B., and Vrastil J. 2015. Link between the potentially hazardous Asteroid (86039) 1999 NC43 and the Chelyabinsk meteoroid tenuous. Icarus 252:129–143. Reisener R. J. and Goldstein J. I. 2003. Ordinary chondrite metallography: Part 2. Formation of zoned and unzoned metal particles in relatively unshocked H, L, and LL chondrites. Meteoritics & Planetary Science 38:1679–1696. Righter K., Abell P., Agresti D., Berger E. L., Burton A. S., Delaney J. S., Fries M. D., Gibson E. K., Haba M. K., Harrington R., Herzog G. F., Keller L. P., Locke D., Lindsay F. N., McCoy T. J., Morris R. V., Nagao K., Nakamura-Messenger K., Niles P. B., Nyquist L. E., Park
J., Peng Z. X., Shih C.-Y., Simon J. I., Swisher C. C. III, Tappa M. J., Turrin B. D., and Zeigler R. A. 2015. Mineralogy, petrology, chronology, and exposure history of the Chelyabinsk meteorite and parent body. Meteoritics & Planetary Science 50:1790–1819. Rubin A. E. 2003. Chromite-plagioclase assemblages as a new shock indicator; implications for the shock and thermal histories of ordinary chondrites. Geochimica et Cosmochimica Acta 67:2695–2709. Scott E. R. D., Taylor G. J., Rubin A. E., Okada A., and Keil K. 1981. Graphite-magnetite aggregates in ordinary chondritic meteorites. Nature 291:544–546. Sharp T. G. and DeCarli P. S. 2006. Shock effects in meteorites. In Meteorites and the early solar system II, edited by Lauretta D. S. and McSween H. Y. Tucson, Arizona: The University of Arizona Press. pp. 653–677. St€ offler D., Keil K., and Scott E. R. D. 1991. Shock metamorphism of ordinary chondrites. Geochimica et Cosmochimica Acta 55:3845–3867. Van der Bogert C. H., Schultz P. H., and Spray J. G. 2003. Impact-induced frictional melting in ordinary chondrites: A mechanism for deformation, darkening, and vein formation. Meteoritics & Planetary Science 38:1521–1531. Varela M. E., Metrich N., Bonnin-Mosbah M., and Kurat G. 2000. Carbon in glass inclusions of Allende, Vigarano, Bali, and Kaba (CV3) olivines. Geochimica et Cosmochimica Acta 64:3923–3930. Voropaev S. A., Sevast’yanov V. S., Eliseev A. A., and Petukhov D. I. 2013. Raman identification of calcite grains in the Chelyabinsk meteorite. Geochemistry International 51:593–598. Wang A., Kuebler K. E., Jolliff B. L., and Haskin L. A. 2004. Raman spectroscopy of Fe-Ti-Cr-oxides, case study: Martian meteorite EETA79001. American Mineralogist 89:665–680. Weber I., B€ ottger U., Pavlov S. G., Jessberger E. K., and H€ ubers H. W. 2014. Mineralogical and Raman spectroscopy studies of natural olivines exposed to different planetary environments. Planetary and Space Science 104:163–172. Wlotzka F. 1993. A weathering scale for the ordinary chondrites (abstract). Meteoritics 28:460. Wlotzka F. 2005. Cr spinel and chromite as petrogenetic indicators in ordinary chondrites: Equilibration temperatures of petrologic types 3.7 to 6. Meteoritics & Planetary Science 40:1673–1702. Yamaguchi A., Scott E. R. D., and Keil K. 1998. Origin of unusual impact melt rocks, Yamato-790964 and -790143 (LL-chondrites). Antarctic Meteorite Research 11:18–31. Zinner E., Amari S., Wopenka B., and Lewis R. S. 1995. Interstellar graphite in meteorites: Isotopic compositions and structural properties of single graphite grains from Murchison. Meteoritics 30:209–226.
SUPPORTING INFORMATION
Table S2. Chromite composition for each olivine– chromite pair for thermometry as determined by EMPA in wt%.
Additional supporting information may be found in the online version of this article: Table S1. Average olivine composition in each olivine–chromite pair for thermometry as determined by EMPA in wt%.
