ARTICLE Received 29 Sep 2016 | Accepted 21 Apr 2017 | Published 1 Jun 2017 DOI: 10.1038/ncomms15702 OPEN Tracing the oxygen isotope composition of the upper Earth’s atmosphere using cosmic spherules Andreas Pack1, Andres Ho¨weling1,2, Dominik C. Hezel3, Maren T. Stefanak1, Anne-Katrin Beck1, Stefan T.M. Peters1, Sukanya Sengupta1, Daniel Herwartz3 & Luigi Folco4 Molten I-type cosmic spherules formed by heating, oxidation and melting of extraterrestrial Fe,Ni metal alloys. The entire oxygen in these spherules sources from the atmosphere. Therefore, I-type cosmic spherules are suitable tracers for the isotopic composition of the upper atmosphere at altitudes between 80 and 115 km. Here we present data on I-type cosmic spherules collected in Antarctica. Their composition is compared with the composition of tropospheric O2. Our data suggest that the Earth’s atmospheric O2 is isotopically homogenous up to the thermosphere. This makes fossil I-type micrometeorites ideal proxies for ancient atmospheric CO2 levels. 1 Universita¨tGo¨ttingen, Geowissenschaftliches Zentrum, Goldschmidtstrae 1, 37077 Go¨ttingen, Germany. 2 Karlsruher Institut fu¨r Technologie, Institut fu¨r Angewandte Materialien - Werkstoffprozesstechnik, Hermann-von-Helmholtz-Platz 1, 76344 Eggenstein-Leopoldshafen, Germany. 3 Universita¨tKo¨ln, Institut fu¨r Geologie und Mineralogie, Greinstrae 4-6, 50939 Ko¨ln, Germany. 4 Universita´ di Pisa, Dipartimento di Scienze della Terra, Via Santa Maria 53, 56126 Pisa, Italy. Correspondence and requests for materials should be addressed to A.P. (email: [email protected]). NATURE COMMUNICATIONS | 8:15702 | DOI: 10.1038/ncomms15702 | www.nature.com/naturecommunications 1 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms15702 ree molecular oxygen (O2) is released by photosynthesis into are due to evaporation and do not reflect the isotope composition the atmosphere and is essential for all breathing animals. of the upper atmosphere; a conclusion that was supported by FWith exception of data for the last 800,000 years from air further measurements24,26–29. From the oxygen and iron isotope inclusions in polar ice, little direct information is available about composition, Engrand et al.24 modeled evaporative mass losses concentration and isotope composition of ancient atmospheric for I-type spherules of 54–85%. O2. This is due to the limited interaction between air molecular Because evaporative fractionation is strictly mass-dependent, oxygen and the lithosphere. I-type spherules still provide unique information about the Among the rare rocky materials that contain atmospheric mass-independent anomaly in D’17O of their upper mesospheric oxygen1–3 there are particular types of micrometeorites oxygen source. The reconstruction of variations in atmospheric (microscopic extraterrestrial dust particles) called cosmic D’17O from fossil cosmic spherules4,30–33 would be a new 4 17 spherules . Roughly 10 tons of small extraterrestrial particles paleo-CO2 proxy. The only published D’ O data on I-type are deposited onto the Earth’s surface per day5. The particles spherules by Clayton et al.4 and Engrand et al.24, however, have collide with the Earth’s atmosphere at velocities of 11–70 km s À 1 intrinsic uncertainties that are too large (0.1 to 41%) to provide (ref. 6) and are visible as shooting stars when they are decelerated reliable information on the composition of the upper atmosphere. and at altitudes up to B80–115 km7,8. A portion of these We present new high-precision oxygen isotope data of extraterrestrial particles totally melts during the atmospheric tropospheric O2 and compare these data with new high-precision entry and is termed cosmic spherules. Cosmic spherules that are oxygen and iron isotope data from Antarctic I-type spherules. composed of Fe,Ni oxides are termed ‘I-type cosmic spherules’ These data are combined with results of oxidation and (in the following, we use the short version ‘I-type spherules’9–11. evaporation experiments to test if the Earth atmosphere is These I-type spherules formed by oxidation of extraterrestrial isotopically homogenous and if isotope ratios of fossil cosmic Fe,Ni metal alloys, which are ubiquitous components of spherules are suitable paleo-CO2 proxies. meteorites. Because oxygen in I-type spherules originates entirely from Results the atmosphere, they are excellent probes for the isotopic Oxygen isotope composition of tropospheric air. The mean composition of upper atmospheric oxygen. The isotopic compo- composition of air oxygen from our study (series B; sition of atmospheric oxygen, in turn, is a proxy for the global 18 ± 1,2,12–14 Supplementary Table 1) is d O ¼ 24.15 0.05% with primary production (GPP) and atmospheric CO2 levels .It D’17O ¼0.469±0.007%. An earlier protocol (series A; is not clear, however, if the atmospheric oxygen is isotopically Supplementary Table 1) gave identical D’17O, but slightly lower homogenous up to the meso- and thermosphere, where cosmic d18O. The datum from this study is within the range reported in spherules interact with air. 17 17 16 the literature and agrees with D’ O values of Thiemens et al. The stable isotope composition of tropospheric O2 (99.8% O, and Young et al.14. The measured data of Young et al.14 have 0.04% 17O, 0.2% 18O) is controlled by the steady state between 15 been corrected relative to the San Carlos olivine value reported by photosynthesis and respiration (mass-dependent Dole effect ), Pack et al.34. The corrected measured datum of evapotranspiration and mass-independent fractionation in the D’17O ¼0.467±0.005% for air oxygen14 is then identical to stratosphere14,16. For the composition of the modern troposphere ± 18 17 our measured datum of À 0.469 0.007% (Supplementary values of 23.4rd Or24.2% and À 0.566rD’ Or À 0.430% % 14,17–20 Table 1) and agrees with the model datum of À 0.469 have been reported in the literature (for definitions, see presented by Young et al.14 The non-application of Methods) with little variations up to 61 km (ref. 17). The high 34 D 17 18 VSMOW2-SLAP2 scaling to our air data shifts the ’ O d O of tropospheric O2 is caused by the Dole effect, whereas the B 17 down to À 0.50%, which would be closer to the values of low D’ O value reflects mass-independent fractionation effects in Barkan and Luz18 and Kaiser and Abe20. For this study, we adopt the stratosphere. The higher the atmospheric CO2 levels, the 18 17 17 12,14 d O ¼ 24.15% and D’ O ¼0.47% for the troposphere. lower the D’ O values ; a relation that was used as paleo-CO2 barometer1,2,13. No experimental oxygen isotope data are available for the Oxygen and iron isotope composition of cosmic spherules. The upper atmosphere at altitudes 461 km. To obtain information isotope composition of the I-type spherules and the oxidation 17 experiment run products are listed in Supplementary Table 2. The about the D’ O heterogeneity of the atmosphere, we measured 18 the oxygen (d17O, d18O) and iron (d56Fe, d57Fe; see Methods for d O of the spherules ranges from 36 to 42%. The corresponding D’17O ranges from À 0.72 to À 0.62%. The d18O and the D’17O definition) isotope composition of I-type spherules from the 4 Transantarctic Mountains that have ages o2 Ma (ref. 21). For is within the range reported by Clayton et al. and 17 Engrand et al.24 The d56Fe values are high for all spherules, this time interval, atmospheric CO2 levels as well as D’ OofO2 did not deviate much from the modern level22,23 and data can be ranging from 22 to 32% (Supplementary Table 2). used to test whether the atmosphere is isotopically homogenous. B Because of the 1,000 years residence time of atmospheric O2 Discussion (ref. 14), no effect on the man-made increase in CO2 is yet visible The interaction between cosmic Fe,Ni metal and the Earth in decreasing D’17O. atmosphere during deceleration is considered to proceed in two The oxygen isotope composition of I-type spherules is consecutive steps. The first step is the atmospheric heating and controlled by the composition of the oxidizing species oxidation of the infalling Fe,Ni metal alloy (fractionation in (for example, atmospheric O2), the fractionation during oxidation oxygen isotopes only). The second step is the melting and of the Fe,Ni alloys, and the fractionation during atmospheric evaporation of the Fe,Ni oxides (fractionation in both, oxygen evaporation. and iron isotopes). Cosmic spherules have higher d18O values (up to 56%; ref. 24) Information about the oxygen isotope fractionation that is than any terrestrial material reported so far. The high d18O led associated with the oxidation step is obtained from experiments Clayton et al.4 to propose a heavy oxygen isotope reservoir in the (this study; see Methods) and from iron meteorite fusion crust upper atmosphere. Davis et al.25, however, showed that I-type data4,35,36. The products of the high-T metal oxidation spherules are also enriched in heavy iron isotopes with extreme experiments have d18O values that are 4% lower than air d56Fe values of up to 45%. They concluded that high d18O values oxygen (Supplementary Table 2). Clayton et al.4 reported d18O 2 NATURE COMMUNICATIONS | 8:15702 | DOI: 10.1038/ncomms15702 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/ncomms15702 ARTICLE 0.0 50 Experiments40 –0.1 40 = 0.506 ± 0.003 Oxidation + 1) (‰) –0.2 experiments (this study) 30 Iron meteorite –0.3 fusion crusts4 3 O (‰) 10 17 sample ' starting material 20 Δ O –0.4 18 Evaporation trend δ = 0.5305 In( 10 –0.5 Air O2 3 (slope = 1.18 ± 0.05) 10 –0.6 0 10 15 20 25 0 10203040 18 δ56 sample δ O Festarting material 103 × In(VSMOW2 + 1) (‰) 103 In( + 1) (‰) 103 103 Figure 1 | Triple oxygen isotope fractionation during metal oxidation.