Pyrrhotite-driven early-stage terrestrial alteration in Ryugu grains

Abstract

Samples returned from the carbonaceous asteroid Ryugu preserve a rare record of water- and organic-rich materials from the early Solar System, but this record can be modified rapidly after arrival on Earth. Here we show through controlled atmospheric exposure experiments that Ryugu grains undergo measurable terrestrial alteration within weeks to months. Pyrrhotite develops Fe–O-rich amorphous rims through oxidation and sulfur loss, while adjacent phyllosilicates become amorphized and carbonaceous materials form nanoscale bubbles and carbon–oxygen–rich surface precipitates. X-ray absorption spectroscopy reveals partial oxidation of Fe and S and chemical modification of associated organic matter. From the thickness of pyrrhotite alteration rims, we estimate an early-stage alteration rate of ~0.1 nm per day under low-humidity laboratory conditions. These results identify pyrrhotite as a key trigger of coupled mineral and organic alteration and show that even moderate terrestrial exposure can compromise volatile-rich, fine-grained astromaterials. Our findings provide an empirical basis for cold curation of returned samples.

Experiments show that grains returned from asteroid Ryugu can start changing within weeks on Earth. Iron-sulfur minerals alter first, damaging nearby minerals and organic matter. The results show why returned samples must be kept cold and protected.

Introduction

Sample-return missions provide unprecedented opportunities to study pristine extraterrestrial materials, offering insights into the evolution of the Solar System and the delivery of volatiles and organics to early Earth. Unlike meteorites, which are inevitably altered during terrestrial residence, returned samples can preserve delicate mineralogical and chemical signatures essential for reconstructing parent body processes. However, the fidelity of these records depends critically on how well they are protected once on Earth. Even brief exposure to the terrestrial atmosphere may induce rapid chemical changes, altering or obscuring the very features that make these samples unique^1,2.

The Hayabusa2 mission returned material from the carbonaceous asteroid (162173) Ryugu, a volatile-rich body linked to CI-like meteorites^3–5. Initial studies revealed a complex assemblage of hydrated phyllosilicates, carbonates, sulfides, magnetite, and diverse organic matter^6–8. These phases are central to understanding the incorporation of water and organics into planetesimals, yet they are also highly reactive under terrestrial conditions. For example, sulfides, carbonates, and organic matter are generally reactive phases, regardless of whether they occur in meteoritic or terrestrial environments, but their susceptibility to alteration is greatly enhanced in fine-grained, porous, and hygroscopic CI-like materials under terrestrial conditions. The potential for such transformations raises an urgent question: how stable are Ryugu samples once exposed to Earth’s atmosphere, and what strategies are necessary to preserve their pristine state?

Previous studies have documented terrestrial weathering in meteorites over a wide range of timescales. Early quantitative work by Bland et al.⁹ demonstrated oxidation rates in hot desert ordinary chondrites, while subsequent studies have characterized the formation of secondary phases such as sulfates and carbonates^10,11. More recent work has shown that alteration can occur on much shorter timescales. Observations of Ryugu grains¹², the CI1 meteorite Oued Chebeika 002¹³, the Tarda meteorite¹⁴, and the Winchcombe CM chondrite¹⁵ demonstrate that significant mineralogical and chemical changes can occur within days to months of exposure. Oued Chebeika 002 is particularly important because material collected soon after its inferred fall shows mineralogical and magnetic properties close to those of returned Ryugu samples, providing a valuable terrestrial analog for minimally altered CI-like material¹⁶. Experiments by Lee and Glazer¹³ show that CI1 materials can rapidly disintegrate upon contact with liquid water, highlighting their extreme sensitivity to terrestrial environments and helping to explain the rarity of fresh CI1 meteorite finds.

Here, we report controlled atmospheric exposure experiments on Ryugu grains from sample plate C0105-042. This study aims to quantitatively constrain the rate and mechanism of early-stage terrestrial alteration, with a particular focus on the role of pyrrhotite as a driver of coupled mineral and organic transformations. The Ryugu grains were monitored over several months using high-resolution electron microscopy and synchrotron-based X-ray spectroscopy. These analyses provide quantitative constraints on the nanoscale rate of pyrrhotite breakdown and show that this reaction drives coupled alteration of surrounding minerals and organic matter. Our observations indicate that terrestrial weathering begins within weeks and rapidly progresses, producing alteration rims on sulfides, amorphization of phyllosilicates, bubble textures in organic matter, and precipitate layers on grain surfaces. Together, these results show that pristine asteroidal materials are much more vulnerable to terrestrial alteration than previously recognized, with important implications for planetary science, sample curation, and the preservation of prebiotic chemical signatures in returned samples.

Results

Rapid onset of surface alteration

Most Ryugu grains on sample plate C0105-042 initially exhibited rough or bumpy surfaces (Fig. 1a). These grains contained flake-shaped Mg–Fe phyllosilicates (saponite and serpentine), framboidal magnetite clusters, euhedral dolomite, Ca phosphate, and pyrrhotite. The euhedral pyrrhotite grains displayed pristine crystal facets, and no secondary precipitates were detected on any mineral surface (Fig. 2a, b). After 26 days of storage in a vacuum desiccator, only minor cracking was observed, confirming that the samples had remained largely unaltered (Fig. 1b, S1a, b).

Fig. 1. SE images of C0105-042_No.

6 Ryugu grain. a Immediately after allocation by JAXA. b–d Before atmospheric exposure experiment. e–f After atmospheric exposure experiment. Particle precipitates and ameboid material are indicated by a dark triangle and a white arrow, respectively. The area of the FIB-excavated TEM slice is indicated by a numbered white box. A.A., after allocation, A.E., after atmospheric exposure experiment. FMg: framboidal magnetite, Po: pyrrhotite. SE: secondary electron, FIB: focused ion beam, TEM: transmission electron microscope.

Fig. 2. SE images of C0105-042_No.

15_5 Ryugu grain. a Immediately after allocation by JAXA. b Magnified image of a white box in a). c Before atmospheric exposure experiment. d Magnified image of a white box in c. e After atmospheric exposure experiment. f Magnified image of a white box in e. Particle precipitates and ameboid material are indicated by a dark triangle and a white arrow, respectively. The area of the FIB-excavated TEM slice is indicated by a numbered white box. A.A., after allocation, A.E., after atmospheric exposure experiment. FMg: framboidal magnetite, Po: pyrrhotite. SE: secondary electron, FIB: focused ion beam, TEM: transmission electron microscope.

However, after reallocation of the same plate 497 days later, clear modification had already begun, even though the grains had been stored in a nitrogen desiccator. Some cracks had widened, and particle-like precipitates less than 100 nm in diameter appeared across grain surfaces (Fig. 1c, d). Ameboid materials were particularly visible on pyrrhotite surfaces (Fig. 2c, d).

Following 243 days of controlled atmospheric exposure (temperature: 20–23 °C; relative humidity: 30–40%), surface alteration accelerated dramatically. Grains became completely coated with precipitates exceeding 200–300 nm in size (Fig. 1e). Pyrrhotite facets had vanished beneath amorphous, irregular films, and magnetite framboids were fully obscured (Figs. 1f, 2e, f, S4). In several cases, grains detached entirely from the substrate, indicating loss of cohesion due to near-surface alteration (Fig. S1c, d).

Microscopic alteration textures

Cross-sectional transmission electron microscope (TEM) observations revealed pervasive modification in both mineral and organic components (Figs. 3–4). Bright-field (BF) TEM images showed abundant nanometer-scale bubbles within carbonaceous matter and in the surrounding saponite–serpentine assemblages (Fig. 3a, S5). A continuous particle-precipitate layer, 107 ± 28 nm thick (n = 83), formed on exposed grain surfaces (Fig. 3b, c). This layer consists of amorphous material rich in C and O.

Fig. 3. TEM images of C0105-042_No.

6 Ryugu grain after atmospheric exposure experiment (TEM_1). a Bright-field (BF)-TEM image of the grain surface. Serpentine and saponite are partially amorphized. Bubbles are present in carbonaceous materials filling the interstices of serpentine and saponite assemblages. b Particle precipitates composed of amorphous materials are seen on the grain surface. The inset shows the SAED pattern obtained from the amorphous materials. c Representative EDS spectrum of the amorphous materials. d Magnified image of the white-boxed region in a. The rims of pyrrhotite grains show depletion in Fe and S. e, f Elemental mapping of the area marked in d. Ser: serpentine, Sap: saponite, Po: pyrrhotite. Particle ppt: particle precipitates, C-prot: carbon-protection. TEM: transmission electron microscope, EDS: energy-dispersive X-ray spectroscopy.

Fig. 4. TEM images of C0105-042_No.

15_3 Ryugu grain after atmospheric exposure experiment (TEM_4). a BF-TEM image showing the cross section of the pyrrhotite grain surface. The inset is SAED patterns taken from the pristine and altered pyrrhotite regions. b Representative EDS spectrum taken from pristine and altered pyrrhotite regions. Po: pyrrhotite. Particle ppt: particle precipitates, C-prot: carbon-protection. TEM: transmission electron microscope, EDS: energy-dispersive X-ray spectroscopy.

At the margins of pyrrhotite, thin (18 ± 8 nm, n = 30) depletion rims in Fe and S were detected (Fig. 3d–f), whereas single-crystal pyrrhotite exhibited broader amorphous rims up to ~500 nm thick (Fig. 4a, b). These altered zones contain oxygen and lack crystallinity, indicating oxidation or hydration. In contrast, magnetite remained structurally intact despite being covered by the same precipitate layer, and calcium phosphate grains showed no measurable change (Fig. S6–S7).

Changes in chemical states

X-ray absorption near-edge structure (XANES) analyses reveal that oxidation and sulfur loss progressed at the nanoscale during atmospheric exposure. At the Fe L₃-edge (Fig. 5a), spectra from pyrrhotite-dominated areas show a pronounced absorption at ~708 eV corresponding to Fe²⁺, while the surrounding saponite–serpentine matrix displays a stronger Fe³⁺ feature at ~710 eV. This shift indicates progressive oxidation of Fe-bearing phases, extending from sulfide rims into adjacent phyllosilicates.

Fig. 5. Representative XANES spectra.

a iron (Fe) L₃-edge, b sulfur (S) L-edge, and c carbon (C) K-edge. FeS dominated: pyrrhotite-dominated region, Mg-Fe phyls dominated: saponite-serpentine-dominated region, Carbon in Mg-Fe phyls: carbonaceous materials embedded in saponite-serpentine-dominated region, Particle ppt: particle precipitates, C-prot: carbon-protection. a.u.: arbitrary units.

Sulfur (S) L-edge XANES spectra obtained from pyrrhotite-dominated regions show the development of oxidized sulfur species at grain margins (Fig. 5b). The dominant spectral features are consistent with sulfur in intermediate oxidation states, most plausibly sulfite (S⁴⁺)^17,18, indicating partial oxidation of sulfide sulfur during atmospheric exposure. Clear evidence for complete oxidation to sulfate (S⁶⁺) is not observed in the present dataset. Minor spectral features that may correspond to reduced sulfur species (S²⁻ and/or S⁰) are weak and close to the detection limit, and their presence should therefore be regarded as tentative.

Carbon K-edge spectra (Fig. 5c) obtained from carbonaceous and phyllosilicate-rich regions exhibit distinct absorptions at ~285 eV (aromatic C = C), ~287 eV (amidyl –CO–NH₂), ~288 eV(carboxyl –COOH)¹⁹, and at ~290.5 eV (carbonate CO₃²⁻)²⁰. Enhanced aromatic peak and attenuated carbonate peak near the grain surfaces indicate chemical modification of carbonaceous material.

Collectively, these spectral signatures document the coupled oxidation of FeS minerals and transformation of associated organic phases, revealing that chemical alteration in Ryugu grains advances rapidly even under limited atmospheric exposure.

Discussion

Rapid oxidation of pyrrhotite and transformation of organic matter

Exposure experiments demonstrate that sulfides within Ryugu grains are among the most reactive components once exposed to Earth’s atmosphere. In contrast to previous studies that reported acicular Mg–Fe–Ca sulfates and epsomite (MgSO₄·7H₂O) crystals on altered CI chondrites^10,11 and other Ryugu grains¹², the pyrrhotite grains in sample C0105-042 developed amorphous, ameboid alteration products (Figs. 1f, 2). These textures indicate a distinct solid–gas reaction pathway rather than the water-mediated oxidation typically responsible for sulfate formation (see Supplementary Discussion 1).

Pyrrhotite’s crystallographic surface is intrinsically unstable in the air, oxidizing tens of times faster than pyrite due to Fe-vacancy disorder within its lattice^21–24. TEM images show that oxidation proceeds through selective sulfur loss promoted by adsorption of atmospheric water vapor, producing Fe-O-rich amorphous rims up to ~500 nm thick (Figs. 3d–f, 4). Prior to atmospheric exposure, SEM observations did not reveal comparable depletion rims or amorphous alteration layers on pyrrhotite surfaces (Figs. 1, 2, S1–S4), indicating that these features developed during the exposure experiment. In addition, previous studies of space weathering on Ryugu grains, based on observations of several hundred particles that had not been exposed to the terrestrial atmosphere, reported no comparable alteration features on pyrrhotite²⁵. This further supports that the observed rims formed during terrestrial exposure rather than representing pre-existing structures. As detailed in Supplementary Discussion 1.1, this process involves coupled oxidation and hydration reactions that generate iron hydroxide on exposed surfaces. The iron hydroxide absorbs moisture from the air and expands, thereby forming ameboid precipitates (Fig. 1f).

X-ray absorption spectroscopy confirms incomplete oxidation (see Supplementary Discussion 1.2). S L-edge spectra are dominated by intermediate-valence sulfur species, most plausibly sulfite (S⁴⁺) (Fig. 5b), indicating partial oxidation of pyrrhotite under the experimental conditions. In partially shielded grains embedded within the matrix, only nanometer-scale depletion rims formed (Fig. 3d–f), indicating that restricted access to atmospheric moisture slows the reaction front. The rapid kinetics implied by these textures suggest that volatile sulfur species such as SO₂ and H₂S can form within weeks of exposure (see Supplementary Discussion 1.3), potentially accounting for the characteristic sulfurous odor of freshly fallen meteorites^26,27.

Carbon K-edge XANES spectra from near-surface regions display enhanced aromatic peaks and attenuated carbonate peak, consistent with oxidation and partial hydrolysis of carbonaceous material (Fig. 5c). Morphologically, a carbon- and oxygen-rich precipitate layer approximately 100–200 nm thick coats the grains (Fig. 3a–c). Ryugu grains are known to host both soluble organic matter (SOM) and insoluble organic matter (IOM)^7,8. This amorphous, swollen layer likely represents a hydrolyzed fraction of organic matter. IOM is predominantly composed of aromatic carbons⁷ and is generally resistant to degradation under exposure to atmospheric moisture or acidic conditions. However, structurally less stable fractions of IOM, along with SOM, may undergo hydrolysis in the presence of atmospheric water, yielding simpler molecular species. Alternative explanations, such as adsorption of volatile organics from epoxy adhesives, were evaluated and found to be negligible (see Supplementary Discussion 2). The spatial correlation between the layer and indigenous carbonaceous domains supports an intrinsic, rather than contaminant, origin.

Absence of gypsum

Gypsum (CaSO₄ · 2H₂O) is among the most common secondary minerals produced by terrestrial weathering of meteorites²⁸. In a previous atmospheric exposure experiment on Ryugu grains, fine-grained euhedral gypsum crystals were observed along grain surfaces and within fractures, interpreted as products of reaction between sulfate ions (SO₄²⁻) released from oxidized pyrrhotite and calcium ions (Ca²⁺) leached from dolomite¹².

In contrast, no gypsum was detected in our experiment. Instead, the surfaces of Ryugu grains were entirely coated by an amorphous particle-precipitate layer (Figs. 1–4). This coating likely acted as a diffusion barrier, restricting the outward migration of dissolved SO₄²⁻ and Ca²⁺ and thereby inhibiting gypsum precipitation. The desiccator environment used here employed a simple desiccant system that maintained lower relative humidity (30–40%) than in the previous study (30–55%)¹², reducing the availability and mobility of aqueous species. Furthermore, the grains used in this experiment contained fewer carbonate minerals, providing less Ca²⁺ for sulfate reactions.

Together, these factors—limited humidity, restricted ion transport, and lower Ca abundance—explain the absence of gypsum and highlight the strong environmental dependence of terrestrial weathering pathways in volatile-rich asteroidal materials. This interpretation is further supported by comparison with Oued Chebeika 002, in which natural desert exposure produced abundant sulfate-bearing alteration products, including gypsum, within only a few months¹³. In that case, repeated exposure to higher humidity followed by evaporation likely promoted aqueous ion transport and sulfate precipitation. By contrast, our Ryugu experiment was conducted under lower relative humidity, and the surface precipitate layer likely restricted ion mobility. Under these conditions, pyrrhotite oxidation can produce Fe–O-rich amorphous rims and ameboid precipitates without progressing to extensive crystalline gypsum formation.

These comparisons indicate that the reaction pathway is controlled not only by bulk humidity and Ca availability, but also by nanoscale microstructural relationships among sulfides, carbonate phases, and transport pathways^3,12–14. Where oxidized sulfur species remain spatially separated from nearby Ca-bearing phases, Fe–O-rich amorphous alteration layers may develop without significant gypsum precipitation. By contrast, where pyrrhotite oxidation products are closely associated with carbonate-derived Ca at the microscale, sulfate minerals such as gypsum may form more readily. Thus, the difference between amorphous alteration layers and gypsum-bearing assemblages may reflect local chemical connectivity as much as bulk environmental conditions.

Secondary alteration due to the oxidation of pyrrhotite

The oxidation of pyrrhotite initiates a cascade of secondary alteration processes that modify both mineral and carbonaceous material in Ryugu grains. The oxidation of pyrrhotite is highly exothermic and can locally elevate temperatures in fine-grained assemblages. In laboratory self-heating experiments, mixtures containing only 6 wt% pyrrhotite reached peak temperatures of ~140 °C at heating rates of ~5 °C h⁻¹²⁹, demonstrating the energetic potential of this reaction. While self-heating may locally contribute to the dehydration of interlayer H₂O in saponite, the dominant driver of alteration in Ryugu grains is inferred to be chemical reactions, as indicated by Fe–S–O redistribution around pyrrhotite, partial oxidation in X-ray absorption spectra, and localized amorphization of adjacent phyllosilicates.

The oxidation of pyrrhotite produces protons, thereby acidifying its microenvironment (see Supplementary Discussion 1). Even under ambient conditions, where water vapor equilibrates with atmospheric CO₂ at pH ≈ 5.6, the additional proton flux from sulfide oxidation further decreases pH. Under these mildly acidic conditions, the principal phases in Ryugu grains—saponite, serpentine, and dolomite—become increasingly soluble^30–33. TEM images reveal partial amorphization of phyllosilicates surrounding pyrrhotite grains (Fig. 3a, b, S5, S6), consistent with localized acid attack driven by oxidation reactions.

In addition to chemically driven dissolution, hydration and dehydration of phyllosilicate-rich phases may also have contributed to structural weakening. Recent work on clay-rich carbonaceous meteorites has proposed a breathing-like behavior, in which repeated uptake and release of atmospheric moisture causes expansion and contraction of fine-grained phyllosilicate-rich materials¹⁴. Consistent with this possibility, some Ryugu grains in our exposure experiment developed cracks and split into smaller fragments during atmospheric exposure (Figs. S1, S2), suggesting that physical stress may have accompanied chemical alteration. Such stress may have promoted the formation and propagation of microcracks and enhanced permeability, thereby increasing access of atmospheric water vapor and oxygen to reactive sulfide-bearing domains. This process may have acted together with pyrrhotite oxidation to accelerate progressive alteration.

Carbonaceous materials exhibit bubble-like textures (Figs. 3a, b, S5, S6) that likely formed through oxidation or decomposition under these same acidic conditions. The IOM in Ryugu grains occurs mainly as aromatic-rich nano-globules and as diffuse carbon intimately associated with Mg–Fe phyllosilicates⁷. The latter hosts molecular carbonate components that lack long-range crystallinity. C K-edge XANES spectra from air-exposed grains resemble those of this diffuse carbon, suggesting that molecular carbonate (CO₃²⁻) species were oxidized or decomposed, releasing CO₂. The bubble textures thus record localized degassing associated with carbonate breakdown and organic oxidation.

Collectively, these observations indicate that pyrrhotite oxidation not only drives mineral dissolution but also destabilizes adjacent carbonaceous phases through acidification and oxidation. The resulting loss of structural water, CO₂, and organic integrity highlights the vulnerability of prebiotic signatures in returned samples upon exposure to Earth’s atmosphere.

Benchmarks for terrestrial weathering rate

The alteration rates derived from our Ryugu exposure experiment provide a quantitative framework that can also be applied to evaluating terrestrial weathering in meteorites. Traditionally, the degree of alteration in meteorites has been described qualitatively using mineralogical indicators such as sulfate veins, carbonate dissolution, or iron-oxide veins. However, by linking alteration textures to experimentally constrained rates, it becomes possible to estimate the duration and intensity of terrestrial residence more robustly. The alteration rate was estimated from the mean thickness of depletion rims in matrix-embedded pyrrhotite grains (18 ± 8 nm, n = 30) divided by the duration of controlled atmospheric exposure (243 days), yielding ~0.07 nm/day, which we report as ~0.1 nm/day. Our estimate of ~0.1 nm/day for pyrrhotite alteration should be regarded as a first-order early-stage rate under relatively dry, curation-relevant conditions (30–40% relative humidity), rather than a constant long-term growth rate.

Previous studies show that alteration can accelerate dramatically under more humid conditions. For example, Garvie et al.¹⁴ reported gypsum formation within minutes at very high relative humidity, and Lee and Glazer¹³ showed that CI1 material from Oued Chebeika 002 developed extensive sulfate products after only a few months of natural desert exposure. The month-scale alteration observed in Oued Chebeika 002, therefore, provides an important natural comparison with our controlled Ryugu exposure experiment¹³. Although the alteration products differ, the timescale is consistent with the weeks-to-months alteration documented here under lower-humidity laboratory conditions. Together, these comparisons indicate that humidity acts as a major kinetic switch controlling both the rate and pathway of terrestrial alteration in CI-like materials.

The Orgueil CI chondrite, recovered in 1864²⁷, has experienced more than 160 years of terrestrial exposure. Our estimate of ~0.1 nm/day for pyrrhotite grains embedded within a phyllosilicate matrix provides a first-order indication of rapid early-stage alteration. However, this estimate should not be interpreted as implying strictly linear rim growth over long periods. As alteration rims thicken, diffusion of atmospheric reactants through the rim may become rate-limiting, and surface passivation may further reduce reaction rates. The experimentally derived value, therefore, most likely represents an early-stage alteration rate rather than a constant long-term growth rate. Nevertheless, the near absence of pristine pyrrhotite in Orgueil¹² is consistent with substantial sulfide modification during prolonged terrestrial exposure, although the exact kinetics likely evolved over time.

This interpretation is also consistent with recent magnetic studies of Oued Chebeika 002, which reported magnetic properties similar to Ryugu samples and suggested that older CI1 meteorites such as Orgueil, Ivuna, and Alais may have experienced transformation or loss of the finest-grained pyrrhotite during terrestrial alteration¹⁶. Our observations provide a direct nanoscale mechanism for this inference: pyrrhotite alteration begins at grain margins (Figs. 2–5), and the smallest grains or high-surface-area domains would be expected to lose their primary sulfide signature most rapidly during early terrestrial exposure. Although the measured alteration rate should not be extrapolated linearly over long timescales, it provides a quantitative basis for evaluating how fine-grained pyrrhotite progressively loses its primary mineralogical and magnetic signatures during terrestrial residence.

Implications for sample curation

Samples derived from C- and D-type asteroids—including CI chondrites, the Tarda and Tagish Lake meteorites, and grains from Ryugu and Bennu—commonly contain pyrrhotite^6,34–37. Pyrrhotite also occurs in other meteorite groups, including CM and CR chondrites^35,38,39, Rumuruti chondrites^40,41, nakhlites^42,43, and shergottites^44,45, suggesting that sulfide-driven terrestrial alteration may affect a broader range of extraterrestrial materials, although likely at different rates depending on lithology, grain size, and hygroscopic properties. C- and D-type asteroids are central to understanding the origin of water, volatiles, and prebiotic organic matter in the early Solar System because they accreted under low-temperature conditions and largely escaped melting or differentiation. As a result, they preserve hydrated minerals, sulfides, and complex organic matter that record primordial chemical processes. At the same time, this exceptional scientific value is coupled with extreme fragility. Our results show that key phases in Ryugu-like materials can be altered on weeks to months timescales when exposed to Earth’s atmosphere, even under conditions currently regarded as standard for curation, such as storage in dry air or nitrogen atmospheres at or near room temperature. In this respect, the present study differs fundamentally from previous work that has focused primarily on long-term terrestrial weathering, by demonstrating that significant chemical modification can occur during the early stages of post-recovery handling and storage.

The oxidation rate of pyrrhotite increases sharply with temperature. Experimental studies demonstrate that reaction rates double with a temperature rise from 25 °C to 35 °C and increase by a factor of three to five for a 20 °C rise when oxygen acts as the oxidant²³. Although oxidation slows as temperature decreases, it does not cease even below 0 °C⁴⁶. In practice, most meteorites and returned samples are stored at or near room temperature, conditions under which pyrrhotite oxidation remains kinetically active. Because this oxidation releases both heat and protons, pyrrhotite serves as a killer inclusion, a highly reactive phase that can drive secondary alteration of surrounding minerals and organic matter. Even modest temperature fluctuations under ambient conditions may therefore accelerate chemical degradation in volatile-rich asteroidal materials. Importantly, the incomplete nature of sulfide oxidation implies that chemically reactive intermediate sulfur species can persist even in the absence of extensive sulfate formation, allowing sulfide-driven alteration to continue during storage and handling.

These temperature-dependent kinetics underscore the necessity of cold curation under inert atmospheres. Because pyrrhotite oxidation remains active at ambient conditions and continues even below 0 °C⁴⁶, suppressing sulfide-driven alteration requires not only cooling but also strict control of reactive gases. Storage at low temperature under nitrogen or argon atmospheres (cold curation) can substantially reduce reaction rates, limit proton generation, and inhibit coupled oxidation and hydrolysis processes^47,48.

In practice, however, inert-gas storage alone is insufficient. Commercial high-purity nitrogen (99.999%) typically contains trace oxygen at the ppm level and owing to its high reactivity and large effective surface area, pyrrhotite can undergo measurable oxidation on timescales of months to years even under such conditions. Moreover, maintaining completely oxygen-free environments during sample transfer and analytical procedures is inherently difficult. Once initiated, pyrrhotite oxidation becomes self-sustaining, as protons generated in the early stages accelerate subsequent reactions, making the process difficult to arrest (see Supplementary Discussion 1). The importance of strict curation extends beyond suppression of sulfide oxidation. Ryugu and Bennu samples are also known to contain highly soluble salts, including Na carbonates, chlorides, and phosphates^1,2, which may be readily dissolved, redistributed, or transformed by even limited exposure to atmospheric moisture. Preserving such phases is equally critical, because they represent fragile records of early Solar System fluid chemistry that may be lost even more rapidly than sulfide-derived alteration textures.

These constraints show that preserving volatile-rich extraterrestrial materials requires cold curation that integrates low temperatures, inert-gas atmospheres, and ultra-low levels of oxygen and humidity. Crucially, such measures must be applied not only during long-term storage but across all stages of sample handling, including transfer and analytical procedures. In this sense, cold curation represents a reaction-kinetic requirement rather than a procedural preference, defining the conditions under which sulfide-driven alteration can be most effectively suppressed.

At the same time, cold curation should be regarded as an upper bound on preservation performance rather than a universally deployable solution. In realistic curation settings, trace oxygen exposure, logistical constraints, and the demands of sample distribution and analysis make complete suppression of chemical modification unlikely. Accordingly, a central outcome of this study is not only to outline idealized preservation strategies, but to provide a quantitative framework for understanding alteration processes operating under current standard curation conditions. This framework enables damage-aware interpretation of returned samples and meteorites and is directly relevant to ongoing and future missions, including Bennu (OSIRIS-REx), Phobos (MMX), and Mars Sample Return, particularly given that pyrrhotite occurs in some Martian meteorites and may also be present in surface materials on Mars and Phobos^{37,42,44,45,49–52}, although the extent and rate of sulfide-driven alteration are expected to depend strongly on lithology, texture, and hygroscopic properties.

Methods

Sample history

The sample plate C0105-042_000_00 (hereafter C0105-042), previously analyzed by the Hayabusa2 Initial Analysis Sand Team, was used in this study. A detailed history of C0105-042 is provided in Table S1. The Ryugu small grains on C0105-042, with an average diameter of approximately 100 µm, were sourced from chamber C, which contains C0105 aggregate grains collected during the second touchdown. Following allocation from JAXA, more than 24 Ryugu small grains (Nos. 1–24) were affixed to C0105-042 using a minimal amount of epoxy glue inside an N₂-filled glove box at Kyoto University. The sample plate was subsequently shipped to Hiroshima University in nitrogen-sealed packages.

The first and second electron microscopy observations were performed without any coating. After the second observation, the Ryugu grains on C0105-042 were coated with carbon and further with osmium to reduce damage from focused ion beam (FIB) processing. Grains Nos. 20, 23, and 24 were scanned with a gallium (Ga) ion beam, and several transmission electron microscope (TEM) foils were prepared from these grains to analyze the space weathering layer, mineralogical characteristics, and slickenside features^1,6,25,53,54. The third observation was conducted on all Ryugu grains except Nos. 20, 23, and 24 before the sample plate C0105-042 was returned to JAXA following the completion of the Hayabusa2 Initial Analysis period. Throughout the study, C0105-042 was stored in a vacuum desiccator, except when it was used for observations.

After our Announcement of Opportunity on terrestrial weathering was approved, the sample plate C0105-042 was once again shipped to Hiroshima University in nitrogen-sealed packages. The fourth observation was conducted approximately two weeks after the sample plate arrived. Following this observation, C0105-042 was exposed to the atmosphere inside a special desiccator located in a clean room. The temperature and humidity within the desiccator were maintained at approximately 20–23 °C and 30–40%, respectively. After the start of atmospheric exposure, the fifth and sixth observations were conducted 102 and 243 days later, respectively. After the sixth observation, the sample plate C0105-042 was placed in a vacuum desiccator, marking the completion of the atmospheric exposure experiment.

Analytical techniques

To examine the surface morphology, the Ryugu grains on C0105-042 were observed using secondary electron (SE) imaging at a low accelerating voltage (3.0–5.0 kV) with JEOL JSM-7100F and Hitachi S-5200 field emission scanning electron microscopes (FE-SEM). These analyses were conducted at the Faculty of Engineering, Tohoku University, and the Natural Science Center for Basic Research and Development (N-BARD), Hiroshima University. After being coated with osmium and carbon, portions of the sample were excavated using a FIB system, HITACHI SMI4050, at the Kochi Institute for Core Sample Research, Japan Agency for Marine-Earth Science and Technology (JAMSTEC). This system is equipped with a microprobe system for TEM observation and scanning transmission X-ray microscopy (STXM) analysis. During FIB sputtering, a Ga ion beam was accelerated to 30 kV, and the resulting foils were 100–130 nm thick.

For conventional TEM observation and selected area electron diffraction (SAED) pattern analysis, we used a JEOL JEM-2100F field-emission transmission electron microscope at Tohoku University, operating at 200 kV and equipped with a JEOL energy dispersive X-ray spectroscopy (EDS) detector system. The chemical composition of each mineral was determined using scanning TEM (STEM) mode equipped with the EDS detector. The obtained compositions were corrected using experimentally determined k-factors, with San Carlos olivine, pyrope, and albite as reference standards.

STXM analysis was conducted for the K-edge absorption energy of carbon (C) and the L-edge absorption energy of sulfur (S) and iron (Fe) using the BL4U, UVSOR Synchrotron at the Institute for Molecular Science⁵⁵. In this study, we conducted all STXM measurements using the energy stacking method. In the energy stacking method, the images of the region of interest in the samples were obtained by changing the energies of the X-ray from 0.2 to 0.5 eV steps (S L-edge: 158–190, 189 images, C K-edge: 280–300 eV, 299 images, Fe L-edge: 700–715 eV, 52 images). The point resolution was ~130 nm. The dwell time at an individual point was 2–10 milliseconds. By using the off-the-sample spectrum as a reference (I0) for the on-the-sample spectrum (I), we obtained 2-dimensional near-edge X-ray absorption fine structure (NEXAFS) spectra as optical density (OD) = ln(I0/I) from the Lambert-Beer law. Also, we obtained an elemental distribution image from the difference of OD images at X-ray energies between post- and pre-absorption edges. The normalization, peak fitting, and NEXAFS spectrum analysis were performed by the dedicated processing software REX 2000 (Rigaku, Japan) and aXis2000 (http://unicorn.mcmaster.ca/aXis2000.html).

Author contributions

M.M. conceived the study and led the project. M.M. designed the atmospheric exposure experiments, performed the electron microscopy observations, analyzed the data, and wrote the first draft of the manuscript. T.N., T.M. and N.T. contributed to the interpretation of mineralogical and petrological data and to discussion of the results. T.A. performed the STXM/XANES measurements and contributed to interpretation of the spectroscopic data. A.M., Y.I. and Y.S. contributed to data interpretation and discussion of the results. A.Y. contributed to discussion of storage and curation strategies for meteorites and astromaterials. All authors contributed to revising the manuscript and approved the final version.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available

Funding

This study was supported by a Grant-in-aid for Scientific Research, Nos. 18H01269 from MEXT to M.M.

Data availability

The XANES spectral data underlying Fig. 5 and the measurement data for pyrrhotite depletion rim thickness and surface precipitate thickness generated in this study have been deposited in figshare under 10.6084/m9.figshare.32143750⁵⁶. Additional image data are available from the corresponding author upon request.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-73875-7.