Geochemical Mobility of Elements in Antarctic Environments Affected by CO 2 -Rich Hydrothermal Fluids: Astrobiological Implications

Abstract

Hydrothermal systems are widespread in our solar system. Identification of alteration mineral assemblages on Mars and potentially in ocean worlds such as Enceladus suggests the existence of extensive hydrothermal fluid–igneous rock interactions of astrobiological interest in different planetary bodies. Here, we studied the terrestrial analog Cerro Caliente, a band of geothermal alterations located in the glaciovolcanic environment of Deception Island (Antarctica), with the aim of determining the mobility of major chemical elements (e.g., alkalis, phosphorus) and its implications in the habitability potential of such environments. We verified that the rock texture, particularly rich in volcanic glass, plays a major role in geochemical mobility, with permafrost delimiting the impact of hydrothermal activity by reducing the permeability of the lapilli tuff deposit. We studied the mineralogy and geochemistry of the alteration band by comparing borehole samples in different locations that represent different thermal regimes along the hydrothermal alteration band. The alteration products are characteristic of palagonitization processes, which favor the release of elements useful for life, such as phosphorus, although the basic alkalinity of the medium caused its precipitation in the form of tricalcium phosphate. In addition, lipid biomarker analyses were performed to assess the existence of possible potential ecological niches associated with these environments. On Mars, the circulation of low-temperature CO₂-rich hydrothermal fluids through glass-bearing volcanic rocks results in a loss of silica content and a secondary mineral assemblage composed of palagonite, phyllosilicates, and zeolites, which establishes Cerro Caliente as a valid Mars analog for understanding such environments. In addition, our results support the hypothesis of a hydrothermal origin of phosphorous for the formation of Enceladus’ phosphates recently detected in the plumes. We also determined that a fraction of the calcium in Cerro Caliente was sequestered as carbonates of biogenic origin, which produced a distinctive Raman signal that, together with the lipid content, would make it a relevant potential biosignature if similar findings were made in the search for life in such low-temperature hydrothermal environments. Key Words: Hydrothermal systems—Palagonitization—Phosphates—Lipid biomarkers—Mars—Ocean worlds. Astrobiology 25, 777–792.

1. Introduction

Hydrothermal systems are suitable environments capable of supporting life, as they meet the three essential requirements for habitability: the presence of water, an active element mobility, and energy generated through geochemical reactions between water and rock. Consequently, characterizing these environments is a critical astrobiological priority to understand the conditions necessary for the development of life and to identify potential biosignatures for the detection of life in other planetary bodies.

Hydrothermal systems arise from physicochemical disequilibria driven by the circulation of hot fluids through the rock substrate. This allows their development in a variety of geological settings prone to water–rock interactions. On early Mars, the existence of a hydrological cycle in a time of widespread magmatic activity impacts processes, and local amagmatic hydrothermal systems would have generated extensive hydrothermal activity across the planetary surface (Carr and Head, 2010; Fairén et al., 2010; Ojha et al., 2021). Low-temperature hydrothermal processes are also likely taking place currently in ocean worlds such as Enceladus and Europa, where the circulation of water oceans into geologically active and porous, rocky silicate interiors could lead to the formation of terrestrial-like hydrothermal vent systems on their seabeds (McCollom, 1999; Sekine et al., 2015).

The extensive hydrothermal activity on Mars is evidenced by the presence on its surface of both high-temperature hydrothermal alteration minerals (>230°C, Sanyal, 2005), such as pyroxenes, magnetites, and quartz (Osinski et al., 2005; Schmidt et al., 2009), and low-temperature ones (<190°C, Sanyal, 2005), such as phyllosilicates (smectites, vermiculites, illite, kaolinite, and serpentine), zeolites, Fe-oxides, sulfates, phosphates, carbonates, and opaline silica (Ackiss et al., 2018; Ehlmann et al., 2010, 2011; Ruff et al., 2020). Similarly, in ocean worlds such as Enceladus, hydrothermal reactions ranging from above 50°C (Sekine et al., 2015) to above 90°C (Hsu et al., 2015) may take place beneath the icy surface. At comparable temperature ranges to those of Mars, these three key minerals—Na-phosphates, carbonates, and silica nanograins—were detected in geyser-like ejections that came from the interior (Hsu et al., 2015; McBride et al., 2005; Postberg et al., 2008, 2011; 2023). However, the potential formation of other hydrothermal alteration minerals, such as Mg-rich phyllosilicates similar to serpentine, cannot be excluded, as their formation could explain the high levels of H₂ detected in the plumes, which would be indicative of ongoing serpentinization processes (Glein and Waite, 2020; Sekine et al., 2015; Vance and Daswani, 2020).

The compositional diversity of the low-temperature alteration mineral assemblages, unlike the high-temperature ones, would depend on the mineralogy/composition of the parent rock and the geochemistry and pH of the hydrothermal solutions. This is because, at lower temperature, the kinetics of hydrothermal reactions proceed more slowly, which makes primary minerals, whose formation conditions in terms of pressure and temperature differ significantly from the prevailing conditions, more susceptible to suffering alteration to reach the new equilibrium state (Browne, 1978; Pereira et al., 2024). Furthermore, the intensity of alteration correlates positively with the porosity and permeability of the rock (Browne, 1978; Frolova et al., 2010), which makes the fabric and texture of the rocks another important factor to be considered in the study of these alteration processes and analog validation. Each of these factors is summarized in the diagram presented in Fig. 1, which provides context for the cases of early Mars and Enceladus as they are developed throughout the text.

FIG. 1.

Diagram of the key factors (primary mineralogy, texture, and temperature) that influence the secondary mineral assemblage and chemical mobility resulting from the hydrothermal alteration by CO₂-rich fluids in the context of early Mars and Enceladus. This diagram is based on and synthesizes information from references cited in the text.

Geochemical models based on gas and icy grains identified in Enceladus’ plumes suggest that its rocky core probably has a mafic/ultramafic composition, with sufficient porosity and permeability to allow alteration by CO₂-rich fluids (Glein and Waite, 2020). In contrast, our understanding of the martian rock substrate—derived from the numerous data collected by orbital spectrometers, landers, and rovers—reveals that its composition is predominantly mafic. The dominant lithology consists of basalts, although more alkaline facies and rocks with lower silica content, such as tephrites and picrobasalts, have also been reported in the Gusev Crater by the Spirit rover (McSween et al., 2009). Nevertheless, outcrops of intermediate rocks, such as basaltic andesites, were identified via the Thermal Emission Spectrometer onboard the Mars Global Surveyor (Zimbelman et al., 2021). Rocks with a more felsic composition, including andesites at the Ares Vallis landing site by Mars Pathfinder (McSween et al., 2009) and diorites, trachytes, and trachyandesites analyzed near the alluvial fan in Gale crater by the Curiosity rover (Cousin et al., 2017; Wray et al., 2013), have also been described. This higher silica concentration was attributed to the weathering processes of the basaltic rocks, rather than to magmatic differentiation processes, due to the remobilization of dissolved cations (Mg²⁺, Fe²⁺, Ca²⁺, Na⁺, K⁺) from olivine, pyroxene, and feldspar minerals against Si (Cousin et al., 2017; McSween et al., 2009; Michalski et al., 2024; Zimbelman et al., 2021).

There is evidence for possible extensive explosive volcanism in the martian highlands during the Noachian (for a review on explosive volcanism on Mars, see Brož et al., 2021). Large depressions, volcanic in origin, have been described in Arabia Terra (Chu et al., 2023; Michalski and Bleacher, 2013), a region of Mars characterized by the presence of regional deposits of layered small grain-sized and friable materials. Although their origin is still an object of debate, there is mounting evidence for these deposits to be interpreted as regional basaltic pyroclastic deposits related to these large depressions or plain-style caldera complexes (Bates et al., 2023; Chu et al., 2023; Michalski and Bleacher, 2013; Whelley et al., 2021). Spatially associated with these layered deposits, we found extensive light-toned materials interpreted to be phyllosilicate-bearing materials (Noe Dobrea et al., 2010). Hypotheses for the formation of these materials include in situ weathering (Bibring et al., 2006), materials of sedimentary origin (Michalski and Noe Dobrea, 2007), hydrothermal processes (Ehlmann et al., 2011), and weathering of fine-grained materials in cold acidic environments (Michalski et al., 2013).

Located in the southwest margin of Arabia Terra, Oxia Planum is the selected landing site for the future Rosalind Franklin ExoMars rover mission (Quantin-Nataf et al., 2021; Vago et al., 2015). This area is characterized by a widespread presence of clay-rich deposits and volcanic materials (Mandon et al., 2021; Quantin-Nataf et al., 2021). Although the nature of this volcanism is not clear, to the south of the landing site there are some circular structures that resemble plain-style caldera complexes (Molina et al., 2017; Quantin-Nataf et al., 2021), so the presence of explosive materials in the area is a possibility.

In Oxia Planum, the Mars Reconnaissance Orbiter (MRO) multispectral CRISM data pointed to the presence of vermiculite and smectites such as saponite, nontronite, and montmorillonite (Brossier et al., 2022; Carter et al., 2016; Mandon et al., 2021). Spectral signatures related to kaolin, Fe, Ca-rich carbonates, and zeolites were also observed (Brossier et al., 2022; Dugdale et al., 2020). The interpretation of the origin of these phyllosilicate materials is still under debate. Although the influence of hydrothermal fluids is not ruled out, several hypotheses have been proposed that range from their formation in a subaqueous environment (palustrine, lacustrine, or marine) to their origin due to pedogenic processes and groundwater alteration (Altieri et al., 2023; Brossier et al., 2022; Quantin-Nataf et al., 2021).

Therefore, although phyllosilicates, carbonates, and zeolites are not formed exclusively in low-temperature hydrothermal environments, their association with certain geological settings and geochemical characteristics of the altered rocks could be related to these environments. This makes these materials a target for the in situ study to be carried out by the Rosalind Franklin ExoMars rover (Dugdale et al., 2023).

Analyses of nakhlite meteorites have inferred the composition of low-temperature hydrothermal fluids on Mars, which may have been CO₂-rich fluids with temperatures between 150°C and 200°C (Bridges and Schwenzer, 2012). The alkaline richness probably came from the altered host rock. Laboratory simulation experiments have shown how CO₂-rich hydrothermal fluids were enriched in cations such as Na⁺, Ca²⁺, Fe²⁺, Mg²⁺, and SiO₂ by the leaching of a martian synthetic glass rocky analog. Consequently, as the pH increased, dissolved Fe²⁺, Mg²⁺, and Ca²⁺ were sequestered from the liquid phase by precipitation of secondary minerals such as phyllosilicates, zeolites, and carbonates, to ultimately result in an alkaline composition of the hydrothermal fluid with high Si concentration (Noda et al., 2022). Therefore, on Mars, glass-rich volcanic deposits that have undergone low-temperature hydrothermal alteration would show a depletion of Si compared with the parental rock.

A secondary mineral that also results from the low-temperature hydrous alteration of basaltic glass is palagonite. Palagonite was discovered, for example, in the altered subglacial volcanic deposits of Sisyphi Montes (Ackiss et al., 2018; Henderson et al., 2021). Furthermore, it has been inferred that palagonite may be widely present in the composition of the martian soil due to the widespread distribution of glaciovolcanic environments throughout the history of the planet (Allen et al., 1981; Cousins and Crawford, 2011; Squyres et al., 1987). These glaciovolcanic environments would have generated glass-rich volcanic deposits and an associated low-temperature hydrothermal system as the result of the interaction of the volcanic system with water.

Palagonite, recognized visually for its earthy tones, has been defined as a hydrated mineral aggregate composed of a variety of smectites and minor amounts of zeolites and oxides (Stroncik and Schmincke, 2001). The transition from yellowish to brownish tones indicates the increasing degree of palagonitization of the basaltic glass, caused by the crystallization of the secondary phases that form it. During crystallization, surface absorption also occurs for those elements that were released during the initial stages of glass dissolution, which become immobile elements as the composition and pH of the fluid change (Stroncik and Schmincke, 2001, 2002).

These glassy materials that result from hydrous alteration are not just of astrobiological interest as indicative of the presence of potential low-temperature hydrothermal habitable environments. Palagonite may also be formed due to the bioalteration driven by microorganisms that colonize the glassy surfaces in search for nutrients and energy (Izawa et al., 2010; Staudigel et al., 2008). Jerome et al. (2022) have recently demonstrated the catalytic properties of mafic glass for the prebiotic synthesis of RNA from ribonucleoside triphosphates, which makes the presence of basaltic glasses possible catalysts in the origin of life. The astrobiological importance of these materials goes beyond the presence of low-temperature hydrothermal sites on Mars; their properties also have important implications for the habitability of ocean worlds such as Enceladus or Europa, as basaltic glass has been suggested to be present in their unconsolidated silicate cores, based on compositional similarity in the komatiite-basalt lavas of Io and Earth (Izawa et al., 2010). As is the case for basaltic glass, phyllosilicates are also catalysts for organic reactions that lead to the formation of increasingly complex organic molecules (Kloprogge and Hartman, 2022). They may therefore have played a crucial role in the origin of life. In addition, they have the ability to absorb organic matter onto their external surface or into the interlayer space, functioning as an impermeable barrier against weathering and the harsh martian UV radiation conditions. This makes phyllosilicates remarkable materials for the preservation of biosignatures (Broz et al., 2019; Orofino et al., 2010; Poch et al., 2015). However, as Fornaro et al. (2018a) pointed out, their preservation potential compared with that of other minerals depends on a combination of factors, such as the nature of the organic biomarker, the specific characteristics of the mineral, the presence of oxidants, and environmental and UV irradiation conditions. This complexity has been supported by additional studies (e.g., Ertem et al., 2017; Fornaro et al., 2018b) that demonstrated how variations in these factors can significantly impact the ability of phyllosilicates to preserve organic matter. Focusing on phosphorus, it is a crucial element in the assessment of an organic molecule as a biosignature, as it is required in the synthesis of the RNA molecule and is also part of other building blocks of life such as DNA and lipids (Palmer, 2013). On Mars, phosphorus is mainly concentrated in sedimentary and diagenetic phosphate deposits such as fluorapatite and merrillite/whitlockite, respectively (Hazen et al., 2023). However, it could also be of hydrothermal origin as a result of the alteration of phosphate crystals that would be part of mesostasis in igneous rocks (Hazen et al., 2023; Liu et al., 2022).

Hence, the low-temperature hydrothermal alteration of primary basaltic minerals would supply the elements essential for the synthesis of prebiotic chemistry and nutrients for a putative microbial life. In addition, the paragenesis of phyllosilicate-rich secondary minerals would provide a more favorable environment for the preservation of biosignatures compared with that of high-temperature hydrothermal systems (Broz et al., 2019; Orofino et al., 2010). It is crucial to understand how geochemical mobility affects the availability of elements such as phosphorus that are essential for life and the implications this has for the habitability of martian hydrothermal environments and ocean worlds in the outer solar system.

With these premises, we conducted our study at Cerro Caliente, a Mars glaciovolcanic analog site located on Deception Island volcano (Antarctica). The summit of Cerro Caliente is crowned by a 40-m-long geothermal fracture in which porous deposits of glass-bearing basaltic lapilli tuffs are altered by the confluence of the ascending hot CO₂-rich fluids (≤100°C) and the descending ice melt and meteoric waters. However, the extent and dynamics of this hydrothermal alteration could be constrained by the presence of permafrost. Lapilli tuffs could be considered textural analogs to pyroclastic materials in Mars and the porous inner rocky mantles of ocean worlds. Previous X-ray diffraction (XRD) analyses of ground samples have revealed a mineral assemblage similar to that found in low-temperature hydrothermal environments on Mars, comprising smectites, zeolites, and carbonates, while phylogenetic analyses detected an associated microbial community (Lezcano et al., 2019). In addition, palagonite might be a mineral component in these soils given its ubiquitous presence in the altered pyroclasts of the island (Álvarez-Valero et al., 2020; Heap et al., 2025). All these characteristics, combined with extreme Antarctic conditions that help preserve the textural and mineralogical integrity of the volcanic materials, make Cerro Caliente a unique natural laboratory for studying the interaction of low-temperature CO₂-rich hydrothermal fluids with volcanic substrates under more defined conditions.

We characterized the hydrothermal system of Cerro Caliente by analyzing the secondary mineral assemblages and geochemical data taken from 2-m-long drill-core samples of different thermal regimes along the fracture. In addition, we analyzed lipid biomarkers of potential microbial origin in carbonate fractions, as carbonates tend to be more stable than phyllosilicates in environments exposed to ongoing hydrothermal fluid interactions.

2. Materials and Methods

2.1. Cerro Caliente in the Deception Island volcanic setting

Deception Island is an active volcano located in the Bransfield Strait, between the South Shetland Islands to the northwest and the Antarctic Peninsula to the southeast (Fig. 2A). Its origin dates back to the Quaternary, linked to back-arc rifting of the Bransfield Basin because of the rollback subduction of the Phoenix plate beneath the Shetland plate (Aparicio et al., 1997; Baraldo, 1999; Gónzález-Ferrán, 1985; Smellie et al., 2002).

FIG. 2.

(A) Sketch map of the tectonic setting of the Bransfield Strait Region, showing the location of Deception Island. (B) Geological sketch map of Deception Island from Geyer et al. (2020). The lithostratigraphic units are represented in simplified form in five colors: post-caldera deposits (green), syn-caldera deposits (yellow), pre-caldera deposits (blue), alluvium, beach and scree deposits (brown), and permanent ice (white). The red cross marks the location of Cerro Caliente hill. (C) The geothermal fracture at Cerro Caliente and the location of the three drill points. (D) Electrical driller perforating in the alteration zone.

The origin and evolution of the volcanic activity of Deception Island are recorded in the six volcanic lithostratigraphic units described by Smellie (2001), who classified them into three groups according to whether they pre-, syn-, or postdate the caldera collapse event dated at c. 8300 years according to paleomagnetic data (Oliva-Urcia et al., 2016), and 3980 ± 125 calibrated years Before Present (cal yr BP) based on tephrochronology, sedimentological studies, and ¹⁴C dating (Antoniades et al., 2018), and c. 4.6 ka according to cosmogenic data (Álvarez-Valero et al., 2025) (Fig. 2B). The shoaling and subaerial emergence of the volcanic edifice took place in the pre-caldera stage (Smellie, 2001). The submerged volcano, then known as Foster volcano, because the main crater was localized over present Port Foster Bay (González-Ferrán et al., 1971), was born by the lava extrusion in Hawaiian-style eruptions followed by scoria and lapilli tuff ejection in Strombolian eruptions (Agusto et al., 2007; Smellie, 2001). This first phase of volcanism is represented by the Fumarole Bay Formation (Smellie, 2001). The same volcanic activity continued in the subaerial shield-building stage that ended with the development of the Basaltic Shield formation, characterized by a lavic lower section and a top section that is scoriaceous (Smellie, 2001; Smellie et al., 2002).

Once the Foster stratovolcano was entirely constructed, with the presence of parasitic cones on its southeastern flank (Torrecillas et al., 2013), a combination of rifting tectonics, seismic activity, and phreatomagmatic eruptions is thought to have contributed to the collapse of the volcanic edifice and the formation of the caldera (Agusto et al., 2007; Antoniades et al., 2018; Baraldo, 1999; Martí et al., 2013). This led to the ubiquitous emplacement of Outer Coast Tuff Formation, described as yellowish lapilli tuff due to a later palagonitization process (Smellie, 2001).

Since the caldera collapse, volcanic activity in Deception Island has mostly consisted of low-volume hydrovolcanic events in response to the asymmetrical subsidence of the caldera and flooding of its central zone by meteoric and melting water. However, the hydrovolcanic character of eruptions did not intensify until the southwestern wall of the caldera—affected by the coalescence of slope instabilities, explosive eruptions, and seawater infiltration—was breached by the formation of Neptune’s Bellows. This breach caused Deception Island to open to the Southern Ocean, which led to the formation of Port Foster Bay (Hopfenblatt et al., 2021). A large hydrothermal system was then established in the Port Foster area (Martí et al., 2013), while eruptive vents were mostly circumscribed to the caldera rim periphery (Smellie et al., 2002).

The locus of volcanic vents and the different degrees of interaction between magma and seawater/groundwater/ice melting caused eruptions to range in explosivity, from effusive phases to explosive phreatomagmatic eruptions (Pedrazzi et al., 2018; Smellie et al., 2002). Thus, lavas and scoria deposits of the Stonethrow Ridge Formation, generated during low-energy explosive eruptions, are found interlayed between consolidated lapilli tuffs and ashes of Baily Head Formation and the friable, more recent ones of the Pendulum Cove Formation (Smellie, 2001; Smellie et al., 2002).

The current manifestations of volcanic activity in Deception Island are subaerial geothermal fields related to the fault and fracture systems of the offshore part of the caldera (Caselli et al., 2004). Hot gas-bearing hydrothermal fluids circulate and alter the volcanic rock substrate, which results in associated mineralizations that are compositionally distinct from the composition of the pyroclastic deposits of the Pendulum Cove Formation and the alluvial and marine beach deposits. This is the case for Cerro Caliente, located in the far south of Fumarola Bay close to Irizar Lake (Fig. 2B). Cerro Caliente is a geothermal anomaly zone with fumarolic activity of 90–100°C locally constrained to the fracture that crosses its summit (Agusto et al., 2004; Caselli et al., 2004).

2.2. Sampling the Cerro Caliente hydrothermal environment

The hot CO₂-rich fluids associated with the fracture produce a hydrothermal alteration band developed over the Pendulum Cove Formation. The temperature may reach 100°C in some patches, which promotes colonization by green mats of bacteria (mainly cyanobacteria) and fungi (Lezcano et al., 2019). During the austral summer of 2012, three boreholes referenced as CCBH2, CCBH3, and CCBH4 were dug along the summit ridge using an electrical driller (Fig. 2C, D). The 1- to 2-m-drill-core samples were extracted from the central part of the alteration band (CCBH2) and from two sites adjacent to this band (CCBH3 and CCBH4). They registered superficial temperatures of 98°C, 7°C, and 2°C, respectively.

2.3. Mineral and chemical characterization

Each of the drill-cores was subdivided into intervals based on color and grain size changes. Two sample pieces were obtained from each predefined interval so that each piece was sent to the appropriate sample preparation method according to the analytical technique to be used. The first piece was for petrographical analysis, and the second one was pulverized to be analyzed by XRD, near- and mid-infrared spectroscopy, and inductively coupled plasma mass spectrometry. Profex-BGMN software that contains the COD-210114 database (Doebelin and Kleeberg, 2015) was used to analyze X-ray diffractograms and OMNIC software for infrared spectra.

During this phase of visual examination of the drill-core samples, we found that CCBH2 samples are distinguished from CCBH3 and CCBH4 samples by the presence of small white spots that correspond to carbonate minerals (Fig. 3). We selected a piece of the sample from the interval taken at a depth of 68–80 cm from the CCBH2 drill-core (CCBH2 68–80) to assess the possible biogenic character of these minerals (Fig. 3). We reduced it to a size of 10 mm in diameter for analysis utilizing a correlative Raman imaging system and a scanning electron microscope (Raman-SEM) at the FELMI-ZFE facility (Graz, Austria). The analysis procedure followed is detailed in de Dios-Cubillas et al. (2024). We conducted elemental mapping with energy-dispersive X-ray spectroscopy (EDX). The acquired Raman spectra were processed by OriginPro 2022b software. Raman peaks were fitted to Lorentzian functions. The remaining fraction of the piece CCBH2 68–80 was pulverized and split into two portions. The first one was sent to Laboratorio de Isótopos Estables de la Universidad de Salamanca (Spain) to measure the isotopic composition of organic and inorganic carbon (δ¹³C_org and δ¹³C_inorg) and oxygen (δ¹⁸O), and the last one was used to perform lipid biomarker analysis.

FIG. 3.

A drill-core section of the CCBH2 borehole showing a yellowish-red color resulting from the hydrothermal alteration process, with small carbonate mineralizations. In the bottom left, a fragment of the CCBH2 68–80 sample prepared for scanning electron microscope (SEM)—Raman analysis.

2.4. Lipid extraction, fractionation, and analysis

A piece of about 6 gr (d.w.) of sample CCBH2 68–80 was subjected to ultrasonication, extracted, and fractionated as described elsewhere (Megevand et al., 2022). A mixture of internal standards (tetracosane-D50, myristic acid-D27, and 2-hexadecanol) was added before the extraction for quantification. For analysis, the apolar fraction was injected in n-hexane, and the acid fraction was methylated with BF₃ in MeOH at 80°C for 30 min. The alcohol fraction was derivatized using BSTFA (Sigma Aldrich, Madrid, Spain) at 80°C for 60 min. Thus, acids and alcohols were detected as fatty acid methyl esters (FAME) and trimethylsilyl derivatives, respectively.

Organic compounds present in the sample were analyzed using gas chromatography-mass spectrometry (GC-MS) for identification and quantification. We used a gas chromatography system (8860 GC) coupled to a mass spectrometer (5977B MSD), with a single-axis detector (Agilent Technologies, Santa Clara, CA, US, USA), operating with electron ionization at 70 eV and scanning from m/z 50 to 650 (analytical details in Megevand et al., 2022). Compound identification was based on retention time and mass spectra comparison with reference materials and the NIST mass spectral database. Quantification was performed with the use of external standards of n-alkanes (C₁₀–C₄₀), FAMEs (C_12:0–C_24:0), and n-alkanols (C₁₄, C₁₆, C₁₈, C₂₀, and C₂₂), all supplied by Sigma-Aldrich (Madrid, Spain).

3. Results

3.1. Petrological and mineralogical analysis of drill-core samples

CCBH3 and CCBH4 samples were recovered from the alteration band edges (Fig. 2C). The rock fabric and the degree of alteration are similar in both drill-core samples despite the higher surface temperature recorded in CCBH3. They are composed of heterometric, poorly consolidated, and lapilli-sized pyroclastic deposits. Pyroclasts have subrounded and rounded shapes defined by ash rim and present a hypocrystalline texture. They consist mainly of subidiomorphic phenocrystals of plagioclase laths displaying preferential orientations and polysynthetic twinning surrounded by an unaltered greenish volcanic glass matrix known as sideromelane (Fig. S1A). The glassy matrix could also be tachylyte (Fig. S1B), which is more common in pyroclasts corresponding to a depth of 40–50 cm in CCBH3 (CCBH3 40–50) (Fig. S1C). Moreover, some pyroclasts with holocrystalline-seriate texture are found throughout the drill-core. They are constituted by lath-shaped plagioclase crystals with polysynthetic twinning and allotriomorphic olivines replaced by iddingsite (Fig. S1D). All plagioclases were identified as andesine-labradorites by XRD analysis, and some of them show a type of hydrothermal alteration known as saussuritization, as evidenced by the growth of clinozoisite–epidote crystals from them (Fig. S1E) (Lauri et al., 2003; Rasool and Ahmad, 2023). At the 73–84 cm depth in CCBH3 (CCBH3 73–84), calcareous biostructures were identified (Fig. S1F, S1G), and zeolites infilling volcanic vesicles were found (Fig. S1H). Furthermore, zeolites are present throughout both boreholes. All mid-infrared spectra revealed reflectance bands that match with mordenite (Klunk et al., 2020) and Na⁺-Y-zeolite (Shameli et al., 2011).

The same type of lapilli deposit described in CCBH3 and CCBH4 was observed at the bottom of the CCBH2 drill-core with some changes from the 88–102 cm depth (CCBH2 88–102) upward. The first change is the porosity reduction associated with the cementation of the ash groundmass by iron oxides (Fig. S1I), with the exception of the 50–58 cm depth (CCBH2 50–58) and 20–37 cm depth intervals (CCBH2 20–37) (Fig. S1J).

The second change is related to the palagonitization process that extends to the upper part of the borehole. The progressive hydration of the glassy matrix that started from the rims and vesicles of the pyroclasts caused sideromelane to transform into caramel-colored palagonite (Fig. S1K), which in turn evolved into smectite (Farrand et al., 2018; Stroncik and Schmincke, 2001), characterized as nontronite by XRD and near-infrared. Furthermore, hydration also affects zeolites according to the intensity increase of O-H stretching (3000–3800 cm⁻¹) and H-O-H bending band (1600–1700 cm⁻¹). The zeolites from CCBH2 exhibit greater hydration compared with those from the CCBH3 and CCBH4 boreholes, indicated by a deeper reflectance band (Supplementary Fig. S2) (Singh and White, 2020).

The third and fourth changes are the detection of montmorillonite and saponite by XRD and the observation of alternating layers of carbonate-phyllosilicates infilling the interparticle porosity of pyroclasts (Fig. S1L).

Aragonite is the carbonate mineral present based on SEM-Raman analysis of a sample taken from a 68 to 80 cm depth (CCBH2 68–80) (Fig. 4A). Diagnostic Raman bands of aragonite located at 157 and 210 cm⁻¹ are assigned to translational and librational lattice modes between calcium and CO₃²⁻ groups, and at 703 and 707 cm⁻¹ they are attributed to the ν₄ double degenerate in-plane bending of CO₃²⁻ groups (Nehrke and Nouet, 2011). Aragonite usually shows a characteristic intense sharp Raman band at 1085 cm⁻¹ (Kranz et al., 2010), which contrasts with our results, where the ν₁ symmetric stretching mode of CO₃²⁻ group appears shifted 4 cm⁻¹ to the higher wavenumber at 1089 cm⁻¹. The incorporation and substitution of Ca²⁺ by Sr²⁺ would not explain this bandshift, as it tends to cause the band to move to lower frequencies (Alía et al., 1997). Similarly, increasing temperatures tend to have a comparable effect (Gillet et al., 1993). Instead, this bandshift might be attributed to the sorption of organic matter onto aragonite (Brahmi et al., 2010), an occurrence that has also been reported for calcite (Vogel et al., 2017). In fact, the secondary electron (SE) image revealed numerous objects with a spherical morphology consistent with bacterial cell shapes, appearing to be attached to the aragonite surface (Fig. 4C).

FIG. 4.

(A) The energy-dispersive X-ray (EDX) map shows the overall distribution of calcium minerals (green) identified by Raman as aragonite, phosphorous minerals (purple) within the marked area (b), which corresponds to the magnified view in B, and carbonaceous material (light blue) within the marked area (c), which corresponds to the magnified view in C. (B) Secondary electron (SE) image and Raman spectrum of tricalcium phosphate. (C) SE image and Raman spectrum of microbial-shaped carbon. Red crosses refer to the acquisition point of Raman measurement in the SE image.

The Raman spectrum of these spherical objects showed that disordered (D) and graphite (G) bands correlated with a carbon composition at 1248 cm⁻¹ (the combined D4- and D5-bands), 1321 cm⁻¹ (D1-band), 1508 cm⁻¹ (D3-band), and 1576 cm⁻¹ (the combined D2- and G-bands) (Henry et al., 2019). The identification of the D4-band, an intensity ratio of the Raman D- and G-bands higher than 1.5, and the impossibility of decomposing the band around 1600 cm⁻¹ into two distinct bands (G-band and D2-band) would indicate an amorphous carbon structure. This structure is likely related to its exposure to low temperatures, approximately 150°C, within the range of low-grade metamorphism (Supplementary Table S1) (Kouketsu et al., 2014). The equation developed by Kouketsu et al. (2014), using the D1-FWHM parameter, was applied to determine the metamorphic temperatures within the low- to medium-temperature range (150–400°C), resulting in a temperature value of 159°C (Supplementary Table S1).

Interestingly, electron microscope images show phosphate minerals. In Fig. 4B, for example, a small tricalcium phosphate mineralization near the aragonite spot was observed (Fig. 4B), which is characterized by a needle-like texture and is only a few microns in size. Tricalcium phosphate was identified by Raman peaks at 967 cm⁻¹ corresponding to the ν₁ symmetric P-O stretching mode, 434 and 453 cm⁻¹ to the ν₂ O-P-O symmetric bending mode, 1048 and 1077 cm⁻¹ to the ν₃ (PO₄)³⁻ asymmetrical stretching mode, and 589 and 611 cm⁻¹ to the ν₄ (PO₄)³⁻ asymmetrical bending mode (Jillavenkatesa and Condrate, 1998). The Raman spectrum aligns better with the β-tricalcium phosphate structure instead of α-tricalcium phosphate (Carrodeguas and De Aza, 2011).

3.2. Lipid biomarkers and stable isotopic composition

The analysis of the oxygen and carbon isotopic composition of the aragonite revealed positive δ¹⁸O values (+12.22‰) and slightly depleted δ¹³C_inorg values (−2.37‰), similar to the isotopic signature of hydrothermal calcites formed from alkaline hydrothermal fluids (Shibuya et al., 2012). On the contrary, the low isotopic value of organic carbon (−27.08‰) would be reflecting the presence of microorganisms such as methanogenic bacteria, anaerobic photosynthetic bacteria, cyanobacteria, and algae (Vieth and Wilkes, 2010).

The GC-MS analysis of the organic extract yielded a variety of organic compounds among the three polarity fractions (Supplementary Tables S2–S4). In the apolar fraction, a series of linear and saturated alkanes (n-alkanes) from C₁₄ to C₂₉ carbon chains, alkenes (i.e., C₁₆ and C₁₈), and one isoprenoid (squalene) with a general dominance of low-chain compounds (i.e., C₁₆, C₁₈, and C₂₀) were found.

The acidic fraction was composed of n-alkanoic acids with a maximum peak at C_16:0 and C_18:0, unsaturated alkanoic acids (C_16:1[ω9], C_18:1[ω7], and C_18:1[ω9]), and terminally branched acids (iso/anteiso from C₁₂ to C₁₈ carbons). Finally, the polar fraction was dominated by saturated (C_16:0 and C_18:0) and monounsaturated (C_20:1, C_22:1, and C_24:1) alkanols and a minor presence of sterols (ergosterol and cholesterol).

3.3. Geochemical trends in the drill-cores

Petrological and mineralogical characterization of CCBH3 and CCBH4 drill-core rock samples showed a lower degree of alteration further away from the fracture where the fluids circulate. This was also reflected in the total alkali-silica diagram, as the andesitic basalt composition remains unaltered at depth (Supplementary Fig. S3). These lapilli tuff deposits would be the relatively fresh rocks that were being moderately affected by hydrothermal fluids in CCBH2, producing an increasing depletion in silica and alkalis toward the upper and shallower part of the CCBH2 borehole. In addition, this depletion trend was accompanied by a decrease in magnesium content according to the alkali-total-iron-magnesium diagram, which would indicate the shift from a calc-alkaline into tholeiitic composition during hydrothermal alteration (Supplementary Fig. S4). Instead, there is an enrichment in iron and also in aluminum, as observed in the Chemical Index of Alteration (CIA) and A-CN-K (Al₂O₃-CaO+Na₂O-K₂O) trivalent diagram (Supplementary Fig. S5).

The sequential decrease in CIA values from the CCBH2 borehole to CCBH3 and CCBH4 showed an alteration gradient that is developing from the fracture zone. CCBH2 rock samples have intermediate CIA values (82–56) compared with the low values for CCBH3 (57–56) and CCBH4 (56–55). CCBH3 and CCBH4 rock samples are weakly influenced by hydrothermal fluids, which is partly related to their proximity to the alteration band.

However, the geochemical diagrams of CCBH2 reported that the depletion of Na, K, Si, Mg, and Ca and enrichment of Fe and Al, as well as the increase of the CIA values, did not occur in linear correlation with depth. A possible reason for this observation may be related to local variations in porosity and, consequently, permeability. The CIA profile exhibited advanced hydrothermal alteration at intervals 88–102 (CCBH2 88–102), 50–58 (CCBH2 50–58), and from 32 to 37 (CCBH2 32–37) to the top of borehole. According to the petrological description, these are the most porous intervals with the lowest iron cement content.

To assess the extent of leaching of the major and trace elements that would be available for biological processes, we performed a geochemical mobility study based on the isocon method of Grant (1986) and classified these elements according to their biological functionality defined in the Astrobiological Periodic Table (Cockell, 2015; Remick and Helmann, 2023; Wackett et al., 2004). The Grant method consists of representing the correlation of concentration of each element (expressed in weight percent for main oxides and in parts per million for trace elements) between the fresh rock (X axis) and the altered rock (Y axis) by means of an isocon diagram (Fig. 5). CCBH2 20 sample was selected to represent the altered rock sample for two reasons: (i) it showed the highest degree of alteration in all geochemical diagrams, and (ii) it represents the shallowest permeable sample that would transfer elements to the surface, incorporating them into the biogeochemical cycle. On the contrary, CCBH4 was the least altered drill-core, and the selection of a sample as “fresh rock” was done by predefining the isocon line, which is the trend line that arises from fitting data by pair-wise comparisons of the rock samples from the CCBH4 drill-core. The slope of the isocon line must have a value of R² equal or close to 1 because it would thus represent compositional similarity between fresh rocks with no or very low alteration. Thus, CCBH4 47–52 sample was selected as the precursor rock.

FIG. 5.

Isocon diagram (after Grant, 1986) illustrates the chemical changes of volcanic rock during hydrothermal alteration from less altered rock (CCBH4 47–52) to altered rock (CCBH2 20). Both major elements, plotted in wt%, and trace elements, in ppm, are colored according to their biological use defined in the Astrobiological Periodic Table. Isocon represents a zero volume change, so positive and negative values indicate gains and losses, respectively. The same data are shown below as percentage changes relative to the fresh rock.

The isocon line of CCBH4 47–52 against CCBH2 20 displayed the compositional deviation of the altered rock from the source rock (Fig. 5). Elements plotted on the isocon line, known as immobile, do not change in concentration with alteration, but those above or below the isocon line do gain or lose concentration, respectively. We observed that the chemical exchange from these fluid–rock interactions would have a positive effect on planetary habitability, as more than half of the elements known to play a biological role were leached from palagonitized volcanic glass as their concentration is reduced in the CCBH2 20 sample. These bioelements would return to their respective (bio)-geochemical cycles and might be modified and/or incorporated into microorganisms (Bertrand et al., 2015) or reprecipitated in other mineral forms.

4. Discussion

We have analyzed the mineral and physico-chemical characteristics of the Cerro Caliente hydrothermal system to establish its potential as a terrestrial analog to glaciovolcanic-hydrothermal settings on Mars (e.g., Sisyphi Montes, Ackiss et al., 2018) and to gain insights into processes that occur in rocky interiors in contact with water on ocean worlds (Hsu et al., 2015).

At the outset, several key aspects make Cerro Caliente a useful terrestrial analog for Mars and ocean worlds. First, the host rock of Cerro Caliente is an unconsolidated lapilli of andesitic basalt composition. This is comparable with the mafic pyroclastic materials that seem to be widespread in the ancient martian highlands (Brož et al., 2021) and is likely analogous, in terms of composition and physical properties, to the rocky mantle of Enceladus, which would be characterized by porous and permeable mafic and ultramafic silicate rocks.

Second, mafic glass plays a role in influencing the mineral composition of secondary assemblages that result from the low-temperature hydrothermal alteration. This would have implications not only for martian explosive volcanic deposits, such as the possible pyroclastic ashfall deposit from the Nili Fossae Region and Arabia Terra, as well as distal impact melt deposits, which are also expected to contain mafic glasses (Cannon et al., 2017), but also for Enceladus and Europa, where Izawa et al. (2010) suggested the possible presence of basaltic glasses in the silicate mantle.

Third, water–rock interactions on early Mars were inevitable, driven by the close interrelationship between the hydrological cycle, volcanic, and impact processes. This is supported by the detection of alteration minerals such as phyllosilicates, zeolites, and palagonite (Ackiss et al., 2018; Ehlmann et al., 2009, 2011; Fairén et al., 2010). In particular, the presence of palagonite would indeed indicate the presence and alteration of basaltic glass through hydration (Stroncik and Schmincke, 2002). Similarly, on Enceladus, the proposed presence of basaltic glass, along with the existence of an unlocked subsurface ocean that could allow water–rock interactions, might create conditions favorable for the formation of palagonite (Izawa et al., 2010).

Building on these analogies, analysis of the mineralogy and geochemistry of the Cerro Caliente rock samples offers a better understanding of the hydrothermal processes that would have taken place in the early martian environments and could take place beneath Enceladus’ ice crust.

The high porosity and permeability that characterize the lapilli tuff deposits would suggest that in CCBH3 and CCBH4 boreholes the permafrost would act as an impermeable layer that would significantly limit the circulation of hydrothermal fluids. CCBH2 showed the highest CIA values, particularly in its most porous intervals (CCBH2 88–102, CCBH2 50–58, and from CCBH2 32–37 to the top of borehole), which are devoid of iron cement content. This low level of cementation might be related to changes in the composition and pH of hydrothermal fluids and possibly a lack of microbial biomediation. Both CIA and porosity would indirectly indicate the most permeable intervals, through which hydrothermal fluid circulation would be favored, which would intensify the alteration of andesitic basalt rocks. This is supported by the mineralogical assemblage of CCBH2, which evidences a greater degree of alteration due to palagonitization with smectite formation, in contrast to saussuritization of plagioclases observed in CCBH3 and CCBH4, where their porosity was affected by permafrost.

The geochemical disequilibrium that results from the interaction between hydrothermal fluids and the basaltic substrate might support ecological niches for potential life on Mars. However, the extent of these niches could be limited by the prevailing Noachian paleoclimatic conditions. Some studies propose a cold and wet scenario (e.g., Fairén, 2010; Valantinas et al., 2025), in which the presence of permafrost, such as that observed at Cerro Caliente, may have reduced the subsurface permeability of the volcanic rock deposits, thus hindering hydrothermal fluid circulation, except in localized thaws or fractures that potentially allow their circulation. Conversely, other studies suggest a warm and wet scenario for Mars (Craddock and Howard, 2002; Ramírez and Craddock, 2018), in which an active hydrological cycle could lead to high pore saturation in the shallow subsurface. Under such conditions, upward migration of hydrothermal fluids might be hindered due to reduced pressure gradients and competition for pore space, especially if permeability is low or mineral precipitation occurs.

The geochemical mobility study shows that palagonitized rocks of CCBH2 borehole are enriched in Fe₂O₃ and Al₂O₃, likely due to their concentration in phyllosilicate minerals. Conversely, they are depleted in Na₂O, K₂O, CaO, MgO, and SiO₂. Although a fraction of this concentration mobilized during alteration would have accumulated in secondary phases as phyllosilicates, zeolites, and carbonates, the rest would have been transferred to the environment because the altered samples are depleted in these elements. Accordingly, this same leaching signature would be expected in the glass-bearing rocks on Mars, as well as on Enceladus (Izawa et al., 2010), as authors such as Byers et al. (1986) reported similar chemical variations between subaerial and submarine palagonitization, with the exception of potassium, which is enriched in the submarine environment.

As observed at Cerro Caliente, palagonitization on Mars and Enceladus would constitute a source of Na⁺, Ca²⁺, Mg²⁺, and dissolved silicon ions, as well as other elements essential for life (Fig. 5) (Cockell, 2015; Remick and Helmann, 2023; Wackett et al., 2004), for example, light rare earth elements (La, Ce, Pr, and Nd). These lanthanides are crucial for some bacterial enzymes, including those in methylotrophs, to conduct their catalytic functions in biological reactions (Remick and Helmann, 2023). In fact, methylotrophs are considered potential life-forms that could inhabit the oceans of Europa and Enceladus, as well as the permafrost of Mars (Swathi and Sravanti, 2020).

In contrast, phosphorus, a required element for life’s building blocks, appears to exhibit low mobility. The concentration of P₂O₅ is approximately 0.28%, slightly above the average concentration range (0.18–0.21%) typically found in basic igneous rocks (Yudovich et al., 2023). This value remained invariable along the fracture zone, as similar values were registered in the sediments of all boreholes and near a fumarole site. This would suggest that the overall amount of phosphorus did not change despite varying degrees of hydrothermal alteration along the fracture.

However, the precipitation as tricalcium phosphate in the CCBH2 borehole revealed that P₂O₅ was released from the altered primary minerals. Despite this release, P₂O₅ exhibited low solubility, as its concentration remained unchanged in the altered rock. This behavior is likely attributed to the presence of aragonite and the basic pH of the medium, which favor its precipitation (Durucan and Brown, 2002). The pH analysis of the ground soil of Cerro Caliente revealed a value of 7.1 (Lezcano et al., 2019). Hence, volcanic rocks are a source of dissolved phosphate. Basalts in ancient Mars and the potentially present mafic–ultramafic rocks in the seafloor of Enceladus and other ocean worlds could have experienced the same leaching behavior and could have provided phosphorus to their respective environments.

The alkalinity of the hydrothermal solutions that may have developed on Mars and Enceladus would have constrained the bioavailability of phosphorus, which would have been enhanced under slightly more acidic conditions. However, the detection of phosphate minerals on Mars and in the plumes of Enceladus would indicate basic conditions in the hydrothermal environments developed on these two bodies (Adcock and Hausrath, 2015; Postberg et al., 2023). Similar to Cerro Caliente, the presence of carbonate ions, resulting from the reaction of CO₂ with calcium released during the alteration of calcic plagioclases and glass in the palagonitization process, might contribute to the limited solubility of phosphate minerals in these environments (Durucan and Brown, 2002). As a result, the reincorporation of phosphorous into its biogeochemical cycle would be restricted. However, this would not be an obstacle for hypothetical life-forms that could exist in similar hydrothermal environments on Mars and Enceladus. Phosphorus could be available through the metabolic activity of bacteria capable of metabolizing insoluble phosphate phases, such as the tricalcium phosphate found at Cerro Caliente. Although no study has reported the presence of these bacteria at Cerro Caliente, their potential activity is supported by the detection of phosphate-solubilizing bacteria in geothermal sediments extracted from the protected area (ASPA 140) beside the southeast bay of Deception Island (Vicente et al., 2021). This finding underlines the possibility of similar (bio)-geochemical processes occurring in extraterrestrial environments.

The chemical exchange that results from water–rock interactions proved to be a sufficient source of sustenance for the existing microbial niches in Cerro Caliente. The organic compounds found in the secondary aragonite formed by hydrothermal alteration would be related to the presence of microbial life, as demonstrated by the detection of low-chained n-alkanes, fatty acids, and n-alkanols with a typical even-over-odd preference (C₁₆, C₁₈, C₂₀) that suggested the presence of microbial biomass (Grimalt and Albaigés, 1987; Meyers and Ishiwatari, 1993; Volkman et al., 1998). Iso and anteiso C_15:0 and C_17:0 fatty acids would indicate the presence of sulfate-reducing bacteria (Li et al., 2011; Russell et al., 1997), while the content ratio of the sum of unsaturated alkanoic acids to the sum of n-alkanoic acids (ΣMUFA/ΣFA = 1.037) would reveal that some current bacterial activity is present. The ΣMUFA is associated with fresh organic matter as opposed to the ΣFA, which corresponds to degraded organic matter from a past activity (Summons et al., 2022). Thereby, extraterrestrial substrates that undergo low-temperature hydrothermal alteration, similar to that studied at Cerro Caliente, may also become potentially habitable environments by providing an aqueous medium with the essential ionic solutes and chemical energy for the development of microbial life over time.

5. Conclusions

Palagonitization dominates the alteration process of glass-bearing volcanic rocks when they interact with CO₂-rich hydrothermal fluids. During this process, not only is there a change in the mineral association in which sideromelane is altered to palagonite, olivine to iddingsite, and pyroxene and plagioclase to phyllosilicates, but also a cation exchange occurs that results in the precipitation of carbonates and phosphates.

In the glaciovolcanic environments of Cerro Caliente, permafrost has been observed to delimit the influence zone of hydrothermal fluids around the fracture by reducing the permeability of the affected lapilli tuff deposit. At the surface, this is evidenced by the yellowish-brown alteration band that contrasts the surrounding grayish deposit and by significant differences in the surface temperature recorded in the boreholes: 98°C (CCBH2 borehole) over the 7 and 2°C recorded in the zones adjacent to the band (CCBH3 and CCBH4 boreholes). The samples extracted from the CCBH2 borehole showed this compositional change toward alteration mineralogy composed of phyllosilicates such as nontronite, saponite, and zeolites, as well as carbonates and phosphates. Therefore, on Mars, permafrost terrains would have conditioned the propagation of hydrothermal fluids.

Analysis of geochemical data of CCBH2 revealed a considerable lixiviation of elements known to play a role in biological processes. This suggests that palagonitization favors the availability of these elements for sustaining microbial life. Phosphorus was released; however, its availability for life was limited, as it was sequestered as tricalcium phosphate due to its low solubility in an alkaline medium. Despite this, lipid biomarkers detected in the carbonate fraction demonstrated that microbial life on Cerro Caliente continues due to fluid–rock interactions.

The low-temperature hydrothermal system on Cerro Caliente showed that hydrothermal alteration of glass-bearing volcanic rocks may act as a source of phosphorus and silica and may indicate the possible origin of the sodium phosphates and also silicon cations discovered in Enceladus’ plumes. In addition, the high wavenumber of the CO₃²⁻ stretching band of carbonate would constitute a potential biosignature, as this shift would indicate the presence of organic matter. Moreover, carbonate would become a crucial mineral for the search for biosignatures due to its stability in such environments, which favors the preservation of lipid biomarkers, and its association with phyllosilicates, which, in the context of Mars, would provide protection for these lipids against organic degradation induced by ultraviolet radiation.