Apolline Leclef

The Uranian moons Ariel, Umbriel, Titania, and Oberon are enriched in CO ₂ mixed with CO, but the origin(s) of these carbon oxides, be they primarily native or radiolytic, remain(s) uncertain. Using data collected by NIRSpec on the James Webb Space Telescope (JWST), we measured the spectral signature of CO ₂ and other carbon oxides to help disentangle these hypotheses. Through comparison to laboratory data, we find that many of the detected spectral features are consistent with CO ₂ ice, including ¹² CO ₂ scattering peaks (4.15–4.26 μ m), multilobe ¹³ CO ₂ bands (4.35–4.43 μ m), and CO ₂ biphonon and triphonon modes (4.80–5.25 μ m). Our measurements show that CO ₂ and CO are concentrated on the trailing hemispheres of the inner moons Ariel and Umbriel, potentially supporting a radiolytic production hypothesis, consistent with prior ground-based results. However, many of the identified spectral features are only observed in thick crystalline ice deposits measured in the laboratory, which may be difficult to form via radiolysis of carbon-bearing material mixed in icy regoliths. Similarly, the data exhibit weak 4.02 and 4.40 μ m bands, hinting at the presence of carbonate minerals and ¹³ CO ₂ clathrates, respectively, possibly formed in the interiors of these moons. Furthermore, JWST has revealed that CO ₂ is widespread at Uranus, present in its system of rings, ring moons, and irregular satellites, consistent with its largest moons accreting CO ₂ and other carbon oxides from the Uranian subnebula. We conclude that exposed carbon oxides are potentially native, with their surface distributions shaped by charged particle irradiation and seasonal sublimation–condensation cycles.

Introduction: Radar observations of Mercury’s poles revealed bright features within its Permanently Shadowed Regions (PSRs), interpreted as water ice [1]. Subsequent neutron measurements by MESSENGER supported this hypothesis, indicating water-equivalent hydrogen-saturated soil and potential water ice layers a few meters thick [2]. The origin of this ice remains uncertain, with scenarios including endogenic (e.g., volcanism) or exogenic sources via hydrated asteroids/comets (containing ~50% H₂O in comets, up to 10-20% in asteroids [3]) or solar wind implantation [4]. Impacts could also deliver other volatiles, notably CO₂ [3,5], suggesting CO₂ ice may coexist with H₂O in Mercury’s PSRs [6,7]. Temperature maps and sublimation rates [6,8] even indicate that some northern PSRs have conditions conducive to CO₂ cold traps.Building on previous spectral simulations of H₂O/CO₂ mixtures [9], our work incorporates complementary laboratory experimental approach to also investigate possible interactions at molecular level between H₂O and CO₂.The SIMBIO-SYS instrument [10] aboard BepiColombo, and particularly its VIHI channel [11], will soon observe these regions. We thus further account for PSR illumination conditions and Signal-to-Noise Ratio (SNR) instrumental constraints to calculate Band Detection Limits (BDL) for selected diagnostic features, to evaluate the detectability of CO₂ ice in H₂O mixtures on Mercury’s surface. Method and results: VIHI/SIMBIO-SYS [10,11] is a Vis-NIR spectrometer (0.4–2.0 µm, 6.25 nm sampling, 125 m spatial resolution at 400 km altitude), partially covering CO₂ ice absorption features at 1.4 µm and 2.0 µm. We deposited H₂O/CO₂ ice mixtures using INGMAR [12,13], under conditions relevant to Mercury’s PSR, to measure the spectral reflectance in the 1 µm – 5 µm range. Thick ice layers were produced at 50 K under a chamber pressure of ~5 × 10⁻⁷ mbar. The substrate was an InfraGold surface coated with a residual organic film formed by irradiation of a CH₃OH:NH₃ mixture, with an estimated thickness below 1 µm. Spectra were acquired in bidirectional reflectance, with illumination, emergence, and phase angles of 20°, 15°, and 15°, respectively. Measurements were performed at a spectral resolution of 2 cm⁻¹, using 256 scans or 512 scans. Pure CO2 and amorphous H2O were produced as references, alongside 3 different mixtures : 20%, 33% and 50% of CO2. These proportions are set by the relative partial pressures of the injected gases in the chamber for each mixture.Figure 1: High Resolution experimental data of intimate mixtures between H2O/CO2 ices. Experimental deposition of H2O/CO2 ice mixtures for various CO2 ratio (0 to 100 %), for layers of ~100 µm thick (and ~20 µm for pure CO2). Black arrows show the position of the OH-dangling combinations bands of H2O ice at 1.39 and 1.89 µm.Fig. 1 shows experimental spectra for mixtures of H2O/CO2, with diverse CO2 ratio, at 50K. These spectra exhibit the behaviour of spectral features for various CO2 ratio in the mix (0 to 100%). Here, the intensity of CO2 features increase with the CO2 ratio in the mix. This trend is primarily evident in the 1.0–2.2 µm range, particularly for bands at 1.43 µm and 1.96 µm, which are of direct relevance for VIHI observations. The 2.01 μm band, slightly outside the spectral range of VIHI, is following the same behaviour. The 1.43 µm band is barely detected, while those at 1.96 µm and 2.01 µm offer stronger band depth. Beyond 2.2 µm, absorption features of CO2 become saturated due to the thickness of the deposited ice layers. In this spectral interval, strong OH-dangling overtone combinations bands are also detected at 1.39 µm and 1.89 µm (also seen in [14, 15] but weaker), varying alongside the CO2 ratio in the mix as well, as also seen in previous studies [14]. Their intensity varies consistently with CO2 content, as also seen in [16] for the fundamental OH-dangling, and they are both broader and stronger than nearby CO2 features. These bands are absent in pure H2O, indicating that their presence is directly associated with the incorporation of CO2 within the ice matrix.These spectral signatures were then modelled under realistic illumination conditions representative of Hermean PSRs, focusing on Kandinsky crater in the Northern Hemisphere, following the methodology described in [17]. The results indicate that OH dangling features remain detectable under expected environmental conditions, whereas CO2 absorption bands are significantly less robust. Conclusions: Our experimental simulations revealed the emergence of OH-dangling combination bands at 1.39 µm and 1.89 µm, which are directly linked to the presence of CO2 in the ice matrix. These bands are systematically deeper and broader than CO2 absorption features and increase in intensity with higher CO2 ratios in the mixture. While CO2 band may be challenging to detect under PSR conditions, the OH-dangling bands may offer a more reliable, even if indirect, indicator of CO2 presence in H2O ice, particularly with the detection limits anticipated for VIHI/SIMBIO-SYS instrument. Acknowledgments: The leading authors would like to thank all the people on SIDONIE who helped us during the experimentations, and the CNES for funding. References: [1] Harmon J. K. et al. (2011) Icarus, 211, 37–50. [2] Lawrence D. J. et al. (2013) Science, 339, 292–296. [3] Altwegg et al. (2019), Annu. Rev. Astron. Astrophys. 57:113–55. [4] Tucker O. J. et al. (2019) Journal of Geophysical Research: Planets, 124, 278-293. . [5] Neumann G. A. (2013), Science, 339, 296–300. [6] Ahrens et al. (2022) Geosciences, 12, 51. [7] Schorghofer N. et al. (2021) Geophysical Research...

Introduction: The planet Mars hosts "gullies" [1] — erosional structures resembling those shaped by liquid water on Earth — mainly at the mid-latitudes, with a higher density in the Southern Hemisphere. While most are dormant, some remain active [2], possibly linked to seasonal surface ice [3]. Initially attributed to the action of liquid water [4], gully formation is now thought to predominantly result from processes involving CO2 ice [5, 6]. CO2 sublimation may notably trigger erosion via mechanisms such as geysers formed in translucent slab ice [7]. Other possible contributing mechanisms may rather involve granular CO2 ice or water ice [6, 9, 10]. The use of infrared spectroscopy may help clarify the composition and physical properties of the involved ices.This study focuses on the Sisyphi Cavi region in the Southern Hemisphere (68.5°S, 1.5°E), a high-latitude outlier known for its active gullies. Previous studies [3, 5, 6], mainly based on HiRISE and CTX (Mars Reconnaissance Orbiter, NASA) imagery, have reported dark spots and erosive flows, though the mechanisms behind these features remain uncertain. Infrared data has been less frequently used [3]. Building on [8], this work incorporates data from CRISM (Mars Reconnaissance Orbiter, NASA) and OMEGA (Mars Express, ESA) to investigate the nature and behavior of CO2 ice, aiming to distinguish between granular and translucent forms and assess their potential roles in gully formation.Method: CO2 ice spectra from CRISM are modeled assuming a transparent slab. Retrieved thicknesses are compared with climate model predictions to test this hypothesis, following a method previously used at another site [13]. The novelty lies in using multiple slope orientations, allowing relative comparisons that reduce uncertainties in absolute thickness estimates due to orientation-dependent CO₂ ice accumulation — thicker on pole-facing, thinner on equator-facing slopes.A synthetic spectral database is constructed using CO2 optical constants [11], and resampled to CRISM’s spectral resolution (6.55 nm). A neutral slope spectrum models contaminants (e.g., dust, calibration noise), following a simplified approach from [8]. As shown in Fig. 1, spectra are fitted by minimizing the quadratic deviation, yielding estimates of the slab optical path, CO2 ice purity, and the contamination level. Prior to fitting, CRISM data are corrected for atmospheric effects following the methodology described in [12]. Optical path values are then converted into physical ice thicknesses by incorporating the full acquisition geometry, including slope orientation, slope steepness, and local solar incidence angle. The resulting thicknesses are compared with those predicted by a 1D climatic model [13]. If observed trends align with model predictions, the ice is likely transparent or translucent. In contrast, a mismatch suggests that photons do not traverse the full ice layer, implying the presence of a granular structure at the surface or within the entire ice deposit.Figure 1: Quadratic fit of a CRISM spectrum at Sisyphi Cavi (Ls 227.02°, MY 29 ; FRT00011935), north-facing slope. After atmospheric correction, the fit yields an optical path of 220 mm and ~ 40% contamination. Solid red: corrected CRISM data; dashed red: pure CO2 model; solid blue: best-fit contaminant; dashed black: final fit. Gray region excluded due to calibration issues.Discussion: Following the approach used in [8], we optimized the climate model for the region using CO2 ice detections from several datasets [6].We then compared CO2 ice thicknesses previously derived as a function of slope azimuth from a CRISM observation at Ls 227° [8] with two other periods (Ls 183° and Ls 192°). The new results at earlier Ls also show a trend with slope orientation that suggests translucent ice, similarly to that observed previously [8]. More precisely, these preliminary analyses suggest incomplete transparency: the ice appears to transmit some light but may be mixed with granular components.Figure 2: Dark spots and flows on a gullied crater wall in Sisyphi Cavi (Ls 183°, MY 29; ESP_011396_1115), indicative of CO2 geyser activity. Their potential role in gully modification remains debated.We propose that CO2 ice exists either as fractured translucent blocks or as a mixture of slab and granular structures, for the CRISM observations studied. At first, it could have been in a translucent state (a “slab”), allowing the occurrence of CO2 geysers. During and after the formation of geysers, the ice would break and may have a partly granular structure and/or become contaminated by mineral grains or small grains of CO2 ice. This is supported by the observed dark spots and flows at the same time (Fig. 2, [6]).To further validate our ice thickness retrieval, we plant to compare it with an independent method based on Bayesian Monte Carlo inversion of a radiative transfer model [14]. This approach is quantitative and already provide clues to demonstrate that ice is translucent in the Richardson crater. Combined with gully activity timing, these findings will help shed light on gully formation and evolution on Mars.References: [1] Malin, M.C. et al. (2000). Science 288, 2330–2335. [2] Dundas, C.M. et al. (2022) Icarus 386, 115133 [3] Raack, J. et al. (2015) Icarus 251, 226–243 [4] Costard, F. et al. (2002) Science 295, 110–113. [5] Raack, J. (2020). Icarus 350, 113899. [6] Pasquon, K. (2023). Planetary and Space Science 235, 105743. [7] Pilorget, C. et al. (2016), Nature Geoscience 9, 65–69. [8] Leclef A. et al. (2024), LPI Contributions 3007. [9] Vincendon M. (2015), J. Geophys. Res. Planets, 120, 1859–1879. [10] Forget, F. et al. (2024), LPI Contributions 3007.[11] Quirico E. and Schmitt B. (2004) SSHADE/GhoSST (OSUG Data Center), Da-taset/Spectral Data. [12] Vincendon M. et al. (2007), J. Geophys. Res. Planets, 112, E08S13 [13]...

Introduction: Radar observations of Mercury’s poles revealed bright features within its Permanently Shadowed Regions (PSRs), interpreted as possible water ice [1]. Subsequent neutron measurements by MESSENGER supported this finding, as they are consistent with a water-like-hydrogen saturated soil, potentially indicating layers of water ice about a few meters thick [2]. The probable detection of water ice at the poles of Mercury raises the question of its origin. Different scenarios relying on either endogenic or exogenic sources have been suggested, such as release through volcanism, hydrated asteroids or comets impacts, or solar wind implantation. These two latter mechanisms could have taken place during recent timescales. While the impact mechanism implies bodies that are already H2O rich (estimated average of 50% inside comets, up to 10-20% in asteroids[3]), the solar wind mechanism on the other hand will only bring H+, which is then implanted in the soil to form OH or H2O [4]. The possible presence of other species brought alongside water within small bodies was considered to distinguish between these two mechanisms [5]. The impact mechanism could actually introduce organic material and other types of volatiles, notably CO2 [3, 6]. CO2 ice could then be present within Mercury PSR [7], as suggested for the Moon too [8]. Based on temperature maps derived from MESSENGER's data and CO2 sublimation rates [9, 7], some PSRs on Mercury’s North Pole may actually exhibit temperatures potentially conducive to the presence of CO2 ice. The SIMBIO-SYS instrument onboard the BepiColombo’s mission (ESA/JAXA) will observe Mercury’s PSRs in 2026. In particular, the Visible Infrared Hyperspectral Imager Channel (VIHI) [10] will produce spectra that are expected to provide spectral evidence for water ice within the PSRs [11], as tentatively obtained on the Moon [12]. Our goal is to simulate PSR spectra of various water and CO2 ice mixtures to evaluate the detectability of putative CO2 ice with SIMBIO-SYS. Figure 1: Simulations of areal mix between H2O ice and CO2 ice, for a ratio of 50% each. In yellow and blue we show the high-resolution CO2 and H2O ice reference spectrum (obtained from [13]) respectively, measured at 179 K completed with 28 K data for CO2, and 140-145 K for H2O. In green we mix these two spectra downgraded to the spectral sampling of VIHI. Water ice grain size is fixed at 100 µm in all panels, while CO2 ice grain size decreases from 100 mm (A) to 1 mm (B) to 100 µm (C). Method: VIHI is a visible and near-infrared spectrometer ranging from 0.4 µm to 2.0 µm. With a spatial resolution down to 100 meters for a spacecraft altitude of 400 km, this instrument will provide data with a spectral resolution of 6.25 nm. The VIHI wavelength range includes several absorption features diagnostic of CO2 ice, in particular near 1.4 and 2.0 µm. Fig. 1 shows an initial simplified simulation of spectra corresponding to a 50% mixture of water and CO2 ices. Three combinations of optical path lengths within each ice are shown (this allows e.g. different grain size configurations to be represented). Panel A corresponds to a much larger path length within CO2 than in H2O: we can see that most CO2 ice features are clearly distinguishable. Panels B and C explore configurations that may be more plausible, with lower to no relative differences between CO2 and H2O path lengths. In such configurations, we only observe the presence of two remaining spectral bands diagnostic of CO2 ice, at 1.43 µm and 1.96 µm. These bands could serve as primary indicators of CO2 ice presence in a mixture with water ice. The 1.96 µm is stronger than the 1.43 µm (Fig. 1C), however, due to the spectral range of VIHI that terminates at 2 µm, distinguishing the 1.96 µm band may be more challenging. Next steps: We intend to enhance our simulations by incorporating the SNR conditions specific to VIHI observations of Mercury's poles. This will involve integrating the performances of VIHI [14, 15] and considering the illumination conditions of the PSR coldest areas possibly compatible with CO2 ice presence. References: [1] Harmon J. K. et al. (2011) Icarus, 211, 37–50. [2] Lawrence D. J. et al. (2013) Science, 339, 292–296. [3] Altwegg et al. (2019), Annu. Rev. Astron. Astrophys., 57, 113–55. [4] Tucker O. J. et al. (2019) Journal of Geophysical Research: Planets, 124, 278-293. . [5] Neumann G. A. (2013), Science, 339, 296–300. [6] Nakamura T. et al. (2023), Science, 379, eabn8671. [7] Ahrens et al. (2022) Geosciences, 12, 51. [8] Schorghofer N. et al. (2021) Geophysical Research Letters, 48, e2021GL095533. [9] Paige D. A. et al. (2013) Science, 339, 300–303. [10] Cremonese G. et al. (2020), Space Sci Rev, 216, 75. [11] Filacchione G. et al. (2020), MNRAS, 498, 1, 1308-1318. [12] Li S. (2018), PNAS, 115, 36, 8907–8912. [13] Quirico E. and Schmitt B. (2004) SSHADE/GhoSST (OSUG Data Center), Dataset/Spectral Data. [14] Altieri F. et al. (2017), Rev. Sci. Instrum., 88 (9), 094503. [15] Filacchione G. et al. (2017), Rev. Sci. Instrum., 88 (9), 094502.