Linking impact melt redox with crustal weathering regime Open Access

Impact melting was a widespread process in the early solar system, when collisions wracked the young, predominantly basaltic crusts of the terrestrial planets (Pierazzo et al., 1997). This type of melting results chiefly from the enormous quantities of kinetic energy imparted by hypervelocity projectiles (Pierazzo et al., 1997; Dressler and Reimold, 2001). Redox-controlled outgassing—initially from possible magma oceans on Earth and Mars during their first 10–100 m.y. and then from impact melts and vapors generated by 100–1000-km-scale projectiles—likely dominated the composition of a succession of early atmospheres (Catling and Zahnle, 2020). Even after a gradual transition to the dominance of volcanic outgassing as the impactor flux waned, sporadic large (100-km-scale projectile) impacts in Earth’s Archean and Paleoproterozoic eras could have had important consequences for surface environments (Marchi et al., 2021).

Impact outgassing occurs through two main mechanisms. The first is through vaporization of target and projectile materials, which contain volatile elements that can remain in the atmosphere even after condensation of vaporized silicates. The second is through outgassing of volatiles including carbon-, sulfur-, and hydrogen-bearing species from impact melts. The speciation of gases released from molten target and projectile material depends strongly on the melt redox state (Burgisser et al., 2015). The fugacity of oxygen (fO₂), a measure of free oxygen available to mediate reduction-oxidation (redox) chemistry, is a useful tracker of the overall redox of a system. Through setting the proportions of oxidized and reduced forms of elements such as Fe, S, and V, the fO₂ recorded by rocks and magmas affects mineral stability, volatile solubility, and element partitioning (Cottrell et al., 2021).

The Fe micro–X-ray absorption near edge structure (μ-XANES) technique employs X-ray absorption spectroscopy to track variations in the oxidation state of Fe dissolved in natural glass samples at small spatial scales (~2–50 μm) and with high analytical precision (≤±0.01 Fe³⁺/ΣFe; Zhang et al., 2018; Berry et al., 2018). Because systematic variations in Fe oxidation as a function of fO₂, temperature, pressure, and melt composition have been calibrated experimentally (e.g., Kress and Carmichael, 1991; Borisov et al., 2018; Jayasuriya et al., 2004; O’Neill et al., 2018), assuming redox equilibrium, measurements of synchrotron source X-ray absorption spectra can yield a robust and spatially resolved portrait of fO₂.

We used Fe-μ-XANES to test hypotheses related to controls on redox in impact melts from Lonar crater, India. Specifically, we investigated the importance of redox shifts due to (1) interaction with oxidized atmosphere or impact vapor plume (Ebel and Grossman, 2005); (2) interaction with reduced projectile material (Schaefer and Fegley, 2007); (3) an increase (McCanta and Dyar, 2017) or decrease (Lukanin and Kadik, 2007) in fO₂ due to shock; and (4) fO₂ imposed by target materials (Ebel and Grossman, 2005).

Lonar crater is an ideal test case to explore impact melt redox conditions using Fe-μ-XANES because it is relatively young, with a ⁴⁰Ar/³⁹Ar age of 570 ± 47 ka (Jourdan et al., 2011) and still younger age assignments from other methods (Maloof et al., 2010; Jourdan et al., 2011), and consequently impact glass remains fresh. In addition, the Deccan Traps basaltic target rocks make Lonar crater a strong analog for impacts into the basaltic crusts of Mars and early Earth (Son and Koeberl, 2007).

Lonar samples were collected in 2010 by B. Weiss (see Fig. S1 in the Supplemental Material¹). Care was taken to collect impact glasses rather than glasses produced during ancient brickmaking in the area. Twenty glass beads were selected (Fig. S2), with an emphasis on smaller glass beads (1–3 mm in radius) that displayed splash forms with rounded habits, and larger glass beads (3–8 mm in radius) with partially rounded forms, along with a piece of a larger (~20 cm in size) block of frothy glassy material.

We analyzed polished thin sections mounted on Suprasil 2A quartz slides to determine Fe³⁺/Fe²⁺ ratios via Fe-μ-XANES at beamline 13-IDE at the Advanced Photon Source at Argonne National Laboratory in Argonne, Illinois, USA. We followed analytical procedures and used basalt standards described by Cottrell et al. (2009), recalibrated by Zhang et al. (2018) to determine unknown Fe³⁺/Fe²⁺ ratios (see Supplemental Material). A binomial regression fit to analyses of these basaltic glass standards yielded a calibration curve with R² = 0.9975 (Fig. S3).

Our Fe-μ-XANES measurements recorded a range in Fe³⁺/ΣFe from 0.21 to 0.49. Some glass beads were homogeneous, with variations in Fe³⁺/ΣFe of <0.06 (e.g., Lonar2–18). Other glass beads were heterogeneous, with variations in Fe³⁺/ΣFe of >0.1, including variations of ~0.07 across tens of microns (Fig. 1). Lonar target basalts have been geochemically correlated to the Deccan Traps Poladpur Formation (Chakrabarti and Basu, 2006). Major-element compositions of Lonar glass beads overlapped whole-rock compositions of Poladpur and other low-MgO Deccan lavas, including lavas from the Saurashtra region (Fig. S4; Krishnamurthy and Cox, 1977). Fe³⁺/ΣFe ratios are available for these Saurashtra lavas (where whole-rock FeO was determined by colorimetry and Fe₂O₃* was determined by X-ray fluorescence). The Fe³⁺/ΣFe ratios in these Deccan lavas (0.27–0.66) and in the Lonar glasses (0.21–0.49) span a strikingly similar range (Fig. 2; Fig. S4).

The calculated magmatic fO₂ values of Lonar glasses depend on the choice of Fe-μ-XANES calibration and oxybarometer (see Supplemental Material). We calculated fO₂ via multiple approaches (e.g., Kress and Carmichael, 1991; Borisov et al., 2018; Jayasuriya et al., 2004; O’Neill et al., 2018) and report these values in Table S1. Overall, these calculations yielded very similar fO₂, and therefore our choice of Fe-μ-XANES and fO₂ calibrations does not impact our conclusions. In the remaining discussion, we refer to Fe³⁺/ΣFe ratios calculated with the calibration of Zhang et al. (2018), and fO₂ calculated with the oxybarometer of Kress and Carmichael (1991), because this approach has been shown to yield fO₂ values consistent with furnace fO₂ and spinel oxybarometry in gas-mixing experiments (Cottrell et al., 2021). For this approach, the range in Fe³⁺/ΣFe we measured corresponds to a range in fO₂ of ~ΔQFM = +1.2 to +4.0 (where ΔQFM denotes fO₂ relative to the quartz-fayalite-magnetite buffer in log fO₂ units).

The fO₂ values calculated for the Lonar impact glasses (ΔQFM = +1.2 to +4.0) were higher than values recorded by global mid-ocean-ridge basalts (ΔQFM ~0; Zhang et al., 2018; O’Neill et al., 2018; Cottrell et al., 2021) and Reunion Island basaltic glasses (ΔQFM ~0; Brounce et al., 2022), i.e., in young lavas originating from the same mantle plume that sourced the Deccan Traps ca. 66 Ma.

We synthesized our data with prior work to discriminate among factors shaping redox in the Lonar impact melts. Globally, impact melts and tektites (their enigmatic cousins) span a large range in Fe³⁺/ΣFe and fO₂ (Fig. 2A). This range has been attributed to a combination of redox reactions during shock; interaction with reduced projectile material, atmosphere, or impact vapor; distinct conditions of formation for tektites versus impact melts; and properties of the target rock (e.g., Lukanin and Kadik, 2007; Schaefer and Fegley, 2007; McCanta and Dyar, 2017; Sekine et al., 2023).

For our Lonar data, Fe³⁺/ΣFe correlated with proximity to the edge of small glasses (Figs. 1 and 2). However, transects toward the rims of ~1-mm-diameter glasses (Fig. 1B) showed oxidation of only the outermost tens of microns (Fig. 3A), and the range spanned in these transects represented only a fraction of the overall Lonar range (Figs. 2 and 3B). We interpret this as evidence that interaction with the atmosphere or vapor plume led to oxidation of these thin rims, but that this process played a secondary role in driving the broad range and overall elevated values of Lonar Fe³⁺/ΣFe.

Prior work has found that proximal impact melts (like our Lonar samples) tend to align in redox with target materials, whereas far-traveled tektite glasses are often reduced relative to initial target redox conditions (Lukanin and Kadik, 2007; Giuli et al., 2013; Fudali et al., 1987). Reduction of tektite glasses has been attributed to more extreme pressure-temperature conditions experienced by tektites compared with proximal impact melts, possibly due to intense frictional heating (to >3000 K) during ejection (Sekine et al., 2023) and reduction driven by decompression at high temperatures (Lukanin and Kadik, 2007). Interaction with chondritic projectiles—with an intrinsic fO₂ of ΔQFM ~ –6 to –8 (Schaefer and Fegley, 2007)—could also cause reduction (Lukanin and Kadik, 2007).

We cannot exclude the possibility that some of the observed heterogeneity in Lonar impact melts resulted from projectile interaction or heterogeneous shock. However, the chemical composition of Lonar impact melts has revealed only faint hints of projectile influence (Osae et al., 2005; Misra et al., 2009). Moreover, our data showed that Lonar impact melts were strongly oxidized relative to the estimated fO₂ of Deccan lavas upon eruption (based on Reunion Island lavas; Brounce et al., 2022), whereas interaction with a chondritic projectile or decompression at high temperatures would be expected to lead to a reduction of material with initial fO₂ of ΔQFM ~0 (Lukanin and Kadik, 2007).

Oxidative weathering produces strong changes in Fe speciation. Since the time of their emplacement, Deccan Traps lavas have been weathered in the relatively oxidized and hydrous conditions prevailing at Earth’s surface, causing a variety of mineralogical changes to the rocks, such as hydration and breakdown of primary basaltic minerals to form phyllosilicates and clays (Babechuk et al., 2014). These mineralogical transformations include substantial increases in the Fe³⁺/ΣFe values of the bulk altered lavas, from values that may have been similar to that of pristine Reunion Island basaltic glasses at the time of eruption (~0.15; Brounce et al., 2022) to Fe³⁺/ΣFe ~0.27–0.66 in incipiently weathered lavas (Krishnamurthy and Cox, 1977) to Fe³⁺/ΣFe approaching unity during advanced weathering (Babechuk et al., 2014). Major-element compositions, Fe³⁺/ΣFe ratios, and fO₂ values for incipiently weathered Deccan Trap lavas overlap with the values of Lonar impact glasses (Figs. 2 and 3B; Fig. S4). When plotted alongside data from the Chhindwara weathering profile in the Deccan Traps, both Lonar glass beads and Deccan whole-rock Fe³⁺/ΣFe data fall between fresh Reunion Island lavas and the least-altered rocks from Chhindwara (Fig. 3B). This pattern suggests that incipient oxidative weathering of the target basalts could explain the measured range in Fe³⁺/ΣFe ratios.

Abundant field and petrographic evidence supports heterogeneous pre-impact alteration and oxidation of the Lonar target basalts. Crater walls and drill cores have revealed brecciated and commonly oxidized lava flow tops, and a lower sequence of more heavily weathered lavas (Kieffer et al., 1976). Interstitial glass in unshocked target basalts is altered to palagonite, and secondary minerals include hematite, calcite, zeolites, chlorite, and serpentine. Coesite, a high-pressure silica polymorph found within shocked Lonar basalts, has been related to shock transformation of secondary silica precipitates in amygdules within the target lavas (Jaret et al., 2017).

Heterogeneous pre-impact weathering provides a straightforward explanation for the substantial redox heterogeneity we measured across short spatial distances within Lonar glasses (Fig. 1). The interpretation that target material redox conditions strongly affect the redox of the impact melts is consistent with the correlation between Fe³⁺/ΣFe and Ca seen in Chicxulub glasses, which has been interpreted to reflect the influence of stoichiometrically oxygen-rich Ca-sulfate–and Ca-carbonate–bearing target lithologies (Giuli et al., 2008).

In summary, published data from other impact melts and tektites point to a suite of processes influencing redox. However, our Lonar data suggest that for basaltic targets, weathering history may play a critical role in determining the redox conditions of impact melts, particularly for impact partial melts that have not experienced the extremely high (>3000 K) temperatures inferred for tektites (Sekine et al., 2023).

Gas speciation calculations with the D-COMPRESS model (Burgisser et al., 2015) illustrate the strong sensitivity of gases—particularly sulfur species—to fO₂ (Fig. 4). At fO₂ more than 1 log unit below the QFM buffer, as expected for vapors in equilibrium with chondritic material, S₂ dominates at low pressures (or H₂S for increasing melt H₂O), and vapors contain nonnegligible CO. Under more oxidizing conditions, similar to those we report for Lonar glasses, SO₂ is the most abundant sulfur species. Extrapolation of these calculations to early Earth is subject to considerable uncertainties, but they serve to demonstrate how the large range in the fO₂ of impact melts—including substantial deviations from volcanic rocks—could affect outgassing.

Lonar crater records a relatively small impact compared with the hundreds of projectiles >100 km across that pummeled early Earth and Mars (Marchi et al., 2014). How shock, projectile interaction, atmosphere, and target properties—including pre-impact weathering—shaped impact melt redox conditions across a range of collisional scales remains an open question. The overall effect of impact outgassing on atmospheric composition further depends on the relative magnitudes of projectile vaporization and melt outgassing.

Basaltic volcanic rocks experience weathering when they interact with fluids near the surface of rocky planets. Our Fe-μ-XANES data from Lonar crater impact glasses suggest that the redox conditions of this weathering could carry over to alter the redox balance in impact melt bodies, influencing the gases released to the atmosphere.

In the case of Lonar, the Deccan target basalts have undergone weathering in a Cenozoic atmosphere with ~20% O₂. This contrasts with the evolving chemical weathering regimes expected for early Earth and Mars. Spectral data indicate that the redox conditions of chemical weathering on Mars have undergone at least one major shift, from anoxic weathering under reducing conditions on Noachian Mars to oxidative weathering (producing Mars’ distinctive red tint) in the Hesperian and Amazonian (Liu et al., 2021). While the details remain debated, broadly anoxic chemical weathering of the crust likely also prevailed on Earth during the Archean, with a transition to more oxidative weathering following the Great Oxidation Event ~2.4 b.y. ago (Catling and Zahnle, 2020).

If, as our data suggest, the alteration-driven redox state of target materials strongly influences the redox conditions in impact melts, we infer that the redox balance in impact melts on both Earth and Mars may have undergone a reversal. In the pre-Noachian and Noachian on Mars, and in the Hadean and Archean on Earth, impact melts derived from altered crust were likely to be reduced and would have released reduced gases, reinforcing reduced conditions in the atmosphere. The transition to more oxidative weathering regimes on Earth and Mars favored more oxidized impact melts, as observed, for example, at Lonar crater, leading to melt outgassing that reinforced and stabilized more oxygenated surface environments.

Portions of this work were performed at GeoSoilEnviroCARS at Argonne National Laboratory, which is supported by the National Science Foundation (EAR-1634415) and U.S. Department of Energy (DE-AC02–06CH11357). We thank T. Lanzirotti, A.H. Nava, and the Argonne National Laboratory Advanced Photon Source; E. Cottrell and the Smithsonian National Museum of Natural History; P. Burger and J. Gross; and B. Weiss and J. Shah. Black acknowledges support from NASA grant NNX16AR87G.