Detection of an NH$_3$ absorption band at 2.2 $μ$m on Europa

TL;DR

This study reports the first detection of an NH3-bearing component on Europa through a distinct absorption near $2.20 μm$ in Galileo/NIMS data. Using rigorous noise-cleaning, continuum removal, and areal linear spectral modeling that includes $NH_3\cdot H_2O$ and $NH_4Cl$, the authors show that ammoniated components are required to reproduce the observed spectrum, with a central band at $2.20 ± 0.02 μm$ and depth ~1.6%. The detected pixels spatially align with young geologic units (microchaos, linear, and band terrains), suggesting emplacement from the subsurface via cryovolcanism or related processes, consistent with a thinner ice shell and a thicker, high-pH subsurface ocean. This finding provides constraints on Europa’s interior chemistry and presents nitrogen-bearing material as a key astrobiological marker, motivating further high-resolution investigations by JUICE and Europa Clipper to identify specific NH3-bearing species.

Abstract

The presence of NH$_3$-bearing components on icy planetary bodies has important implications for their geology and potential habitability. Here, I report the detection of a characteristic NH$_3$ absorption feature at 2.20 $\pm$ 0.02 $μ$m on Europa, identified in an observation from the Galileo Near Infrared Mapping Spectrometer. Spectral modeling and band position indicate that NH$_3$-hydrate and NH$_4$-chloride are the most plausible candidates. Spatial correlation between detected ammonia signatures and Europa's microchaos, linear, and band geologic units suggests emplacement from the underground or shallow subsurface. I posit that NH$_3$-bearing materials were transported to the surface via effusive cryovolcanism or similar mechanisms during Europa's recent geological past. The presence of ammoniated compounds implies a thinner ice shell (Spohn & Schubert, 2003) and a thicker, chemically reduced, high-pH subsurface ocean on Europa (Hand et al. 2009). With the detection of NH$_3$-bearing components, this study presents the first evidence of a nitrogen-bearing species on Europa -- an observation of astrobiological significance given nitrogen's essential role in the chemistry of life.

Figures (12)

• Figure 1: An example of noise removal routine applied to the NIMS scan. Left panel: Spectral denoising using despiking approach applied to the Minnaert-corrected spectrum (black dots). The black arrow indicates a random noise spike near $\sim$2.3$\mu$m that was removed in the despiked spectrum (red line). Any black dot (Minnaert-corrected data point) not coinciding with the red line (despiked spectrum) is considered a noise artifact and was removed and replaced during the despiking step. Right panel: Radiation noise removal via spatial smoothing applied to the despiked spectrum (black dots). The black arrow indicates a single-band noise artifact near $\sim$2.2$\mu$m that was treated in the final smoothed spectrum (blue line). Any black dot (despiked data point) not aligned with the blue line (spatially smoothed spectrum) is considered radiation noise and was corrected during the radiation removal step. Inset is the figure with a zoom on the $\sim$2.05-2.40$\mu$m region. The x, y coordinates are the location of the pixel in NIMS observation 11ENCYCLOD01A of malaska2024updated.
• Figure 2: Upper panel: Locational distribution of detected pixels (red polygons) and some regions of non-detected pixels (indigo polygons) using 5th-degree polynomial continuum fit from Galileo/NIMS observation 11ENCYCLOD01A overlaid on the Galileo/SSI basemap malaska2018europamalaska2024updated. Middle panel: The continuum fits to the average reflectance spectra of detected pixels (left subplot) and some non-detected pixels (right subplot) in the NIMS scan. Lower panel: The continuum-removed average spectra along with associated standard errors against the continuum baseline for detected pixels (left subplot) and some non-detected pixels (right subplot). For detected pixels, the reflectance and standard errors values at 2.19-2.21$\mu$m fall clearly below the continuum baseline— confirms the feature exceeds the noise level.
• Figure 3: Upper panel: Average reflectance spectrum with corresponding standard error bars across all wavelengths for the pixels within the detected clusters from NIMS observation 11ENCYCLOD01A (cf. Fig. \ref{['fig:fig4']}) The broad absorption features near $\sim$1.5 and $\sim$2.0$\mu$m are characteristic of H2O ice grundy1998temperaturemastrapa2008optical. A distinct absorption feature centered at $\sim$2.20$\mu$m– indicative of NH3-bearing species (marked by the dashed blue vertical line). Note that the average reflectance of each cluster (4 in total; cf Fig. \ref{['fig:fig4']}) is given in Fig. \ref{['fig:fig10']}. Bottom panel: Gaussian band fit to the continuum-subtracted average spectrum shows a 2.20$\pm$0.02$\mu$m absorption feature with a band depth of 1.59$\pm$0.16%.
• Figure 4: Upper panel: Distribution of detected pixel clusters (red polygons) overlaid on the Galileo/SSI basemap malaska2018europamalaska2024updated. Inset: The yellow box shows the locational context of the investigated image on the USGS's Voyager and Galileo/SSI Global Mosaic at a resolution of 500 m/pixel. Note that the average spectrum of each pixel cluster is provided in Fig. \ref{['fig:fig10']}. Bottom panel: A portion of the geologic map of the same region on Europa leonard2024global. The violet color shaded feature represents the band unit, while linear features include cycloid (aqua blue line), linear band (navy blue line), depression margin (wider black line), and ridges/troughs (narrow black lines). The dark green asterisk symbol represents microchaos features. Refer to leonard2024global for details of the map elements. The detected pixel clusters in the upper panel correlate with microchaos, linear, and band geologic units in the bottom panel.
• Figure 5: Top panel: The average reflectance spectrum of the detected clusters (black circles) with standard error bars, along with the best-fit model spectrum (red line) derived using an areal (linear) mixing approach incorporating the optical constants of NH3·H2O. The reduced chi-square value– $\chi$2(Red.)– of the model fit is also provided in the plot. The inset shows a zoomed-in view of the model fit between 2.05–2.40$\mu$m, with the corresponding RMSE value. The model spectrum closely matches the observed absorption feature at 2.20$\mu$m, rendering a lower RMSE than the fit shown in Fig. \ref{['fig:fig6']}. Bottom panel: A plot of the residual (observation-model) spectrum over the entire wavelength region at $\sim$1.2 - 2.5$\mu$m.