Physicochemical properties of lunar regolith simulant for in situ oxygen production

Abstract: Permanent lunar settlements will rely on in situ oxygen production from regolith for life support and propulsion. While oxygen is abundant in lunar materials, it is chemically bound within metal oxides whose extractability depends strongly on regolith composition and processing strategy. In this study, we validate and characterize high-fidelity lunar regolith simulants representative of the lunar highlands and south pole using scanning electron microscopy with energy-dispersive X-ray spectroscopy, X-ray diffraction, Brunauer-Emmett-Teller surface area and pore structure analysis, and hydrogen temperature-programmed reduction. The simulants exhibit strong mineralogical and compositional fidelity to returned Apollo and Chang'e samples, with ilmenite confirmed as the most readily reducible oxygen-bearing phase. However, despite low ilmenite abundance, bulk highland simulants display favorable reduction behavior arising from distributed Fe-bearing silicate and glassy phases, as well as surface and microstructural properties that influence gas-solid interactions. Adsorption experiments with gases (H2, CH4, and CO2) and water indicate that mineralogical heterogeneity and pore accessibility influence their uptake in simulants. These results indicate that oxygen extraction behavior in realistic lunar regolith is governed by whole-regolith response rather than ilmenite content alone, supporting the option of whole-regolith processing strategies for oxygen production in lunar in situ resource utilization architectures.

• The paper validates Exolith simulants as accurate representations of lunar regolith and explores whole-regolith processing for efficient oxygen extraction, despite low ilmenite content.
• The study highlighted that oxygen extraction from highland and polar simulants LHS-2 and LSP-2 generates oxygen yields at approximately 0.02 wt% and 0.01 wt%, respectively, showing oxygen release from dispersed Fe-bearing silicates, glassy phases, and not specifically from ilmenite.
• The Voltage-Energy gravimetric profiles have broad ranges and steps indicating various energy barriers present a good strategy to optimize the extraction process

Motivation and scope

Oxygen constitutes roughly 45 wt% of returned lunar samples, but it is chemically bound in silicates and oxides whose reducibility depends strongly on mineralogy. Mare regolith, enriched in ilmenite (FeTiO$_3$), is well matched to hydrogen reduction; highland and polar regolith, however, is feldspathic and typically contains less than 2 vol% ilmenite, so extraction strategies predicated on ilmenite beneficiation may be poorly suited to the regions where most lunar infrastructure is likely to be deployed (2601.14719). This paper validates the Exolith simulants LHS-1, LHS-2 (lunar highlands) and LSP-2 (south pole) as experimental analogs, and uses them to ask whether whole-regolith processing can yield oxygen despite low ilmenite content.

The characterization suite combines SEM-EDX elemental mapping, powder XRD benchmarked against Apollo 16 sample 64501, BET/N$_2$ physisorption at 77 K, single-component gas adsorption (CO$_2$, N$_2$, H$_2$, CH$_4$) at 25 °C up to 1 bar, water adsorption isotherms, and hydrogen temperature-programmed reduction (TPR) at 900 °C.

Simulant validation

SEM-EDX confirms oxygen-dominated compositions (~74 at.% O in both LHS-2 and LSP-2), consistent with silicate-rich regolith. The two simulants differ in a geologically meaningful way: LHS-2 is silicon-rich (16.4 at.% Si), matching plagioclase/pyroxene-bearing highland soils, whereas LSP-2 is aluminium-rich (17.8 at.% Al, 25.9 wt%) with low Si (4.8 at.%), mimicking feldspathic terrain. Grain-scale maps show heterogeneous, fine-grained Fe–Ti phases dispersed in silicate/glassy matrices rather than large continuous ilmenite domains — a point that becomes central to the reduction results.

XRD patterns for LHS-1 and LHS-2 overlap strongly with anorthosite and pyroxene references and correspond closely in peak position and relative intensity to Apollo 16 sample 64501; LSP-2 aligns with anorthosite plus basaltic mafic phases, consistent with the mixed feldspathic-basaltic composition reported for the South Pole–Aitken region by Chang'e-6 analyses. Ilmenite is confirmed in LHS-1/LHS-2 by XRD despite its low bulk abundance (~0.4 wt% per manufacturer specifications). Overall, the validation supports these simulants as reasonable laboratory stand-ins for ISRU testing.

Reduction behavior: ilmenite versus whole regolith

Hydrogen TPR yields an apparent oxygen release of 1.10 wt% from pure ilmenite, consistent with literature values of 1–2 wt% at ~900 °C via FeTiO$_3$ + H$_2$ → Fe + TiO$_2$ + H$_2$O. By contrast, the bulk simulants give far lower apparent yields: approximately 0.02 wt% for LHS and 0.01 wt% for LSP. These numbers should be read cautiously: the authors note that TCD baseline drift, gas cross-sensitivities, heating-rate effects, and particle/crystallite size dependence all limit TPR quantification.

The substantive claim of the paper is not the absolute yield but its distribution: reduction occurs across broad, multi-step profiles consistent with distributed Fe-bearing silicates and glassy phases rather than ilmenite alone. On this basis the authors argue that oxygen extraction behavior in realistic regolith is governed by whole-regolith response rather than ilmenite content, supporting whole-regolith processing schemes (e.g., molten regolith electrolysis, plasma methods) over ilmenite-selective beneficiation for highland and polar deployments. It is worth noting the tension here: the measured bulk yields are two orders of magnitude below pure ilmenite, so "favorable" applies to the qualitative reduction signature and process robustness, not to per-mass efficiency. The volumetric availability of highland material is offered as compensation, but no system-level mass/energy analysis is performed in this work to substantiate that trade-off.

The paper also discusses the strategic implications of hydrogen reduction producing water rather than O$_2$0 directly: water is dual-use (electrolysis to LH$_2$1/LOX, life support, shielding), enables closed-loop hydrogen recycling, and is easier to handle than cryogenic oxygen — though electrolysis efficiency degrades under reduced gravity, which may constrain mission power budgets.

Textural properties and gas adsorption

BET analysis shows uniformly low surface areas: ilmenite is highest at 3.195 m²/g (21 nm average pore size); LHS-2 is lowest among the measured solids at 0.349 m²/g but retains mesopore-scale porosity (26.8 nm), attributed to aggregation and packing in mixed-phase powders. Anorthosite and pyroxene have low areas (~0.4–0.6 m²/g) with large ~135 nm voids.

Gas adsorption at 25 °C reveals that uptake does not scale with surface area. Notably, LHS-2 exhibits the highest CO$_2$2 uptake (0.044 mmol/g), marginally exceeding anorthosite (0.042 mmol/g) despite its lower BET area, indicating control by surface chemistry and microstructural heterogeneity. Anorthosite's high CO$_2$3/N$_2$4 ratio (22.1) is an artifact of very low N$_2$5 uptake rather than high CO$_2$6 capacity — the authors correctly flag that their selectivities are simple single-component uptake ratios, not mixture or thermodynamic selectivities. Pyroxene shows comparatively high H$_2$7 uptake (0.0145 mmol/g), suggesting preferential accommodation of small molecules. These results bear on volatile capture, plasma processing, and gas handling steps in ISRU flowsheets more than on oxygen yield itself.

Limitations and open questions

The authors are explicit that neither current simulants nor returned samples replicate space-weathered surfaces shaped by solar wind irradiation, micrometeoroid bombardment, and nanophase iron formation — factors known to alter reactivity and volatile retention. Additional caveats include the semiquantitative nature of the TPR-derived yields, the use of uptake-ratio "selectivities," and the absence of reactor-scale or system-level performance modeling. Open questions left by the study include: how space weathering modifies the reduction and adsorption behavior measured here; whether whole-regolith processing can be closed energetically against ilmenite-selective routes at mission scale; and whether regolith's heterogeneous surface chemistry confers a catalytic or kinetic role in gas-solid reactions relevant to emerging plasma-based extraction methods.

Conclusion

This work provides a coherent multi-technique validation of LHS-1, LHS-2, and LSP-2 against Apollo and Chang'e reference materials and demonstrates that reduction and adsorption behavior in feldspathic simulants arises from the collective contribution of dispersed Fe-bearing phases and microstructure rather than ilmenite abundance. Its principal contribution is empirical support for treating lunar regolith as a whole feedstock in highland and polar ISRU architectures, with the important qualification that quantitative yield comparisons and system-level trade-offs remain unresolved.