Fe-rich X-ray amorphous material records past climate and persistence of water on Mars

CheMin X-ray diffraction on Curiosity shows that 15–73 wt% of Gale crater rocks and sediments are X-ray amorphous, with variable Fe and Si and typically low Al, likely from mafic sources. Although amorphous materials can form via multiple processes (e.g., glass from melt, hydrothermal alteration), SAM-detected volatiles suggest a significant fraction are secondary weathering products. The high abundance of amorphous material on Mars is surprising because amorphous phases are generally metastable and expected to transform to more crystalline phases over time. On Earth, amorphous aluminosilicates can be precursors to crystalline clays and are favored under kinetically limiting (colder) conditions. Variably Fe- and Si-rich but Al-bearing amorphous materials have been observed in terrestrial glacial sediments, Hawaiian soils, and paleosols, but climatic controls on Al-poor yet Fe-rich amorphous material—common at Gale—are not well constrained. This study tests how time and climate control the formation and persistence of Fe-containing but Al-poor amorphous material using analog ultramafic (serpentine) soils across climatic gradients (Mediterranean, subarctic, desert) spanning ~12 to >50 ka. The goal is to interpret Martian environmental conditions consistent with the observed amorphous component, hypothesizing that cool, wet conditions promote formation and cold, dry conditions promote persistence.

Prior work established widespread amorphous material at Gale crater and elsewhere on Mars and linked volatile-bearing amorphous components to secondary weathering. Classical soil science indicates amorphous aluminosilicates can precede formation of crystalline clays and are stabilized/form under colder, kinetically limited conditions. Recent terrestrial studies reported Fe- and Si-rich amorphous materials in glacial sediments, Hawaiian soils, and John Day paleosols, but these generally include Al. There is limited study of Al-poor but Fe-rich amorphous materials akin to Gale crater compositions. Serpentine-derived soils are relevant terrestrial analogs due to Mg/Fe/Si-rich and Al-poor chemistries and typical secondary products (Fe-oxides and smectites), with previous suggestions of secondary amorphous phases in such settings. Serpentine laterites and serpentinite rocks, as well as partly serpentinized lake sediments, are established analog environments for Mars-related aqueous alteration and paleolake contexts.

Field analog sites and climate gradient: Four serpentinite bodies across three climate regimes were studied to assess climatic effects on Fe-rich amorphous formation and persistence: Klamath Mountains, CA (Mediterranean; multiple sites including high-altitude cirques and valleys); Gros Morne Tablelands, Newfoundland, Canada (subarctic; multiple glaciated valleys); Pickhandle Gulch, NV (desert). Site climates and vegetation cover were characterized using NOAA/Environment Canada normals. Soil ages: Klamath cirques deglaciated ~12.1 ka; Swift Creek moraines Late (~15–23 ka) and Middle Wisconsinan (~25–50 ka); String Bean Creek undated and likely older; Tablelands sites dated or constrained to ~13–>20 ka; Pickhandle exposure age unknown, soils thin and likely dominated by detrital inputs. Sampling design: Hand-excavated soil pits to C horizon/bedrock/point of refusal or water contact. At least one sample per horizon; increased depth resolution at visually undifferentiated profiles (Tablelands) and selected Klamath sites. Collected ~4 kg bulk soil/gravel per depth interval; bedrock parent material from buried clasts. Bulk <2 mm soil fraction isolated; gravel fraction weighed. Clay-size fraction separation: Modified Edwards & Bremner method. Disperse ~10 g soil in DI water, sonicate, settle, pipette suspended load; flocculate with 1 M NaCl, centrifuge, wash, freeze-dry. Clay fraction (<2 µm) used for XRD, chemistry, and TEM. Mineralogical analyses: XRD of randomly oriented bulk soils (Cu Kα) to identify primary/secondary phases (using RRUFF/COD databases). Randomly oriented clay-size fraction XRD done on PANalytical X’Pert Pro MPD with Co Kα after spiking 20 wt% α-Al2O3 internal standard. Oriented mount XRD of clay fractions after cation saturation and treatments to diagnose expandable 2:1 clays: (1) Mg2+ saturation air-dried; (2) ethylene glycol vapor solvation; (3) K+ saturation then 550 °C heating. Smectite expansion/collapse behavior used to confirm presence/absence. Serpentine polymorphs distinguished where possible; overlapping peaks noted as limitation. Quantification of amorphous contribution: Rietveld refinements (Profex/BGMN) on spiked clay fractions to fit crystalline phases and obtain fitted background. The amorphous hump estimated by subtracting an SNIP-derived baseline from the fitted background; integrated area under hump used as a relative abundance proxy (not absolute quantification due to unknown densities and disorder of amorphous phases). Transmission electron microscopy (HRTEM): Selected clay-size fractions (Eunice Bluff BC horizon; Devil’s Punchbowl C horizon) prepared on C-coated Cu grids and imaged on FEI Titan 300/80 at 300 kV. Fast Fourier transform (FFT) used (SAED unavailable) to distinguish amorphous vs nanocrystalline packets; EDS for elemental characterization. Geochemistry: Bulk major/trace elements by ICP-MS after HF-HNO3 digestion (NPFL, UNLV); SiO2 by difference from 100% after accounting for LOI950 due to SiF4 loss. Na2O excluded from clay-fraction due to NaCl use; small Na2O in most parents minimizes bias except possible overestimation at Deadfall Lake. Selective dissolutions on powdered bulk soils: - Hydroxylamine HCl + HCl extraction (FeH and Si in Fe- and Si-containing amorphous material). - Citrate-dithionite extraction (FeD) for Fe in amorphous + crystalline (oxyhydr)oxides. - Na-pyrophosphate extraction (Fep) for organically complexed Fe. AAS used for Fe (and Si for hydroxylamine leachates). Ratios FeH/FeD used as a crystallinity index (substituting hydroxylamine for oxalate). Soil properties: pH measured in 1:1 soil:water; LOI550 for organic matter; LOI950 for volatiles. Field descriptions included Munsell color, structure, ped features, textures. Parent material and dust input analysis: Compared soil bulk and clay-fraction mineralogy/chemistry to parent rocks to assess in situ formation vs dust inputs (identifying non-formed phases like quartz, muscovite, calcite, plagioclase as dust where absent in parent material).

• In situ aqueous alteration is required to concentrate Fe in the clay-size fraction: Fe is enriched in clay fractions relative to parent material at wetter sites (Klamath, Tablelands) while Mg is depleted, indicating dissolution of Mg-silicates (e.g., olivine, pyroxene) and precipitation of Fe-rich secondary products (including amorphous materials and Fe-(oxyhydr)oxides).
• Climate control on amorphous formation/persistence: Subarctic Tablelands soils contain substantially greater amorphous Fe and Si than Mediterranean Klamath and desert Pickhandle soils. Hydroxylamine-extractable amorphous Fe (FeH) in Tablelands soils is ≥1.8× parent material values; FeH/FeD > 0.95 indicates that nearly all secondary Fe is poorly crystalline to amorphous. In Klamath soils of similar age, FeH is equal to or less than parent values and FeH/FeD is lower, indicating conversion to crystalline Fe-(oxyhydr)oxides. Pickhandle Gulch shows Fe values similar to parent and low FeH/FeD, indicating high crystallinity and minimal in situ amorphous formation.
• Quantitative ranges: Tablelands mean amorphous Fe is high (e.g., ~25.77 mg Fe/g soil) vs Klamath String Bean Creek (6.02–15.96 mg/g; mean 11.34 mg/g) and parent materials (Tablelands 12.11 mg/g; String Bean Creek 8.68 mg/g). Tablelands thus exhibit roughly double the amorphous Fe of the most amorphous-rich Klamath site.
• XRD amorphous hump positions: All clay-size fractions show amorphous humps centered 26°–30° 2θ (Co Kα), similar to Gale crater amorphous humps (generally 22°–26° 2θ with Fe-richer, Si-poorer at ≥26° 2θ). This supports chemical/mineralogical similarity between terrestrial Fe-rich amorphous phases and Gale crater amorphous materials.
• Composition of dissolved clay-size fractions overlaps Gale crater amorphous component: Al-poor (0.27–11.67 wt% Al2O3), Fe-rich (12.87–23.19 wt% Fe2O3T), Si-rich (13.08–50.20 wt% SiO2) compositions match Gale amorphous ranges.
• Parent mineral persistence indicates climatic impacts: In Klamath, olivine disappears from soils after ~12–15 ka (dissolved), while at the colder Tablelands olivine persists in bulk soils to >20 ka (absent in clay fraction), indicating slower dissolution at colder temperatures. Magnetite present in some parent materials but absent in soils suggests dissolution or below detection. Pyroxene dissolves within ~12.1 ka at one Klamath site.
• Dust inputs: Minor dust (quartz, plagioclase) in Klamath and Tablelands clay fractions; substantial dust-derived phases at Pickhandle (calcite, muscovite, plagioclase, quartz, smectite) and Al enrichment indicate dominance of eolian contributions in the desert site.
• Soil morphology mirrors climatic effects: Klamath shows increasing reddening and clay films with age (Fe-oxide and smectite formation), Tablelands show limited horizon development likely due to freeze-thaw mixing, and Pickhandle soils are thin with vesicular horizons and minimal in situ alteration.
• TEM confirms presence of truly amorphous gel and nanocrystalline Fe-rich domains in Klamath and Tablelands clays; nanocrystallite d-spacings (~2.45, 2.10 Å) consistent with ferrihydrite/goethite nanoparticles.

Findings demonstrate that water availability is required for in situ formation of Fe-containing secondary phases and that cooler temperatures favor formation and long-term persistence of Fe-rich amorphous material, whereas warmer conditions favor more crystalline Fe-(oxyhydr)oxides and, with time, smectites. The analog soils reproduce key attributes of Gale crater’s amorphous component: Al-poor, Fe- and Si-rich compositions, similar XRD amorphous hump positions, and concentration within fine (clay-size) fractions. Consequently, the abundance of Fe-rich amorphous material at Gale crater is consistent with wet conditions near freezing mean annual temperatures during formation, followed by prolonged cold and dry conditions that inhibited crystallization and preserved amorphous phases. The climatic framework explains spatial and temporal distributions of Fe-rich amorphous phases on Mars and supports interpretations of a cold and icy climate regime with intermittent wet episodes during the Noachian–early Hesperian, with subsequent arid, cold preservation. The conceptual model indicates minimal in situ amorphous formation under hot, arid conditions (desert analog), amorphous-dominated secondary products under cold, wet conditions (Tablelands analog), and progressive crystallinity and smectite formation with time in warmer, wetter settings (Klamath analog).

This study shows that Fe-rich, Al-poor X-ray amorphous material forms and persists preferentially in cool, wet climates and transforms toward crystalline phases under warmer conditions. Terrestrial serpentinite-soil analogs across climate and age gradients reproduce the chemical compositions and XRD amorphous signatures observed at Gale crater. The results imply that Gale’s abundant Fe-rich amorphous material likely formed under wet, near-freezing conditions and was subsequently preserved during long-term cold, dry periods. These climate constraints strengthen the interpretation of persistent, albeit cold, aqueous environments on early Mars. Future work should include absolute quantification and structural characterization of amorphous phases, expanded climate and lithology analogs, and direct analysis of returned Martian samples to definitively link amorphous compositions and structures to specific formation pathways and paleoenvironments.

• Amorphous abundance was estimated relatively via integrated area of Rietveld-fitted background humps rather than absolute quantification; unknown densities and complex disorder of amorphous phases precluded internal-standard quantification.
• Identification of specific amorphous phases remains non-unique; amorphous hump positions are influenced by composition and do not uniquely define phases.
• TEM SAED was unavailable; FFT analyses were used as a proxy to distinguish amorphous vs nanocrystalline materials.
• Selective dissolution methods have known selectivity/overlap issues (e.g., dithionite can extract minor structural Fe from smectites/vermiculites; pyrophosphate may include colloidal Fe-oxides), potentially biasing reservoir partitioning and crystallinity indices.
• SiO2 was calculated by difference due to HF digestion (SiF4 loss); Na2O was excluded from clay-fraction chemistry, potentially overestimating SiO2 in samples with higher Na-bearing phases (e.g., Deadfall Lake).
• Dust inputs, especially at the desert site, complicate attribution of secondary phases to in situ alteration.
• Age constraints at some sites are approximate; String Bean Creek and Pickhandle exposure ages are uncertain, and some glacial features lack direct cosmogenic dates.
• Serpentine polymorph discrimination by XRD is challenging due to peak overlap; chrysotile identification is not definitive from XRD alone.