Alternative Solvents for Life: Framework for Evaluation,Current Status,and Future Research

3.1.1. Liquid water: Occurrence

Water is cosmically abundant. It is composed of the first and third most abundant elements in the Universe and is by far the most abundant gas emitted by terrestrial volcanoes (see Supplementary Data S1). Its photochemical breakdown to H and OH is only a net loss to the planet if the H is subsequently lost to space [as has been proposed was the fate of water on Mars (Jakosky, 2021) and potentially Venus (Way et al., 2016)]. On Earth, water lost to space could be effectively replaced through volcanism. This abundance and stability mean that water is by far the most common liquid (other than magmas) in rocky solar system bodies, being present in liquid form on the surface of Earth, on the surface of early Mars and possibly Venus, and in the interior of a number of the larger icy moons.

Obviously, water’s viscosity enables biochemistry at all the temperatures at which water is a liquid on Earth. This sets conservative limits on the viscosity of candidate solvents for life, discussed in more detail in Supplementary Data S1.

Water also shows little change in its physical, chemical, and solvation properties with temperature (Pohorille and Pratt, 2012). This allows a single biochemistry to function in water over a wide range of temperatures, which means that water is a suitable solvent for life over that range of temperatures (compare this characteristic with the substantial changes in physical properties of liquid sulfur or supercritical CO₂ over relatively small changes in temperature, discussed below). On the criterion of occurrence, water is one of the strongest candidate solvents for life in our list.

3.1.2. Liquid water: Solvation

Water is highly favored as a solvent for life because, as well as being an excellent solvent for polar molecules and salts, many molecules are highly insoluble in water, allowing for a “hydrophobic” force that drives the folding of proteins, assembly and stability of membranes, and so on (Tanford, 1978; Pratt and Pohorille, 1992; Pohorille and Pratt, 2012). However, this is not a unique property of water. Of the solvents considered, only water, ammonia, HF, and sulfuric acid have the dense hydrogen bonding networks in the liquid state needed to drive a solvophobic effect, and micelle formation has been observed in all four liquids (discussed below).

The ability of protonating solvents—paradigmatically water—to enable proteins and nucleic acids to adopt the unique, functional 3D structures essential for their activity is considered prima facie evidence that nonprotonating solvents cannot be solvents for life. This is even true of intrinsically disordered proteins, which have to be poised to adopt ordered structure as required. The ability of water to satisfy the solvation needs of life need not be discussed further.

3.1.3. Liquid water: Solute stability

An enormous range of chemicals are stable in solution in cold water, but stability depends on temperature, and supercritical water is widely used as a method to comprehensively break down complex organic materials (Daniel et al., 2004; Brunner, 2009; Bains et al., 2015; Yakaboylu et al., 2015). As with all solvents, stability of dissolved molecules depends on temperature. However, water can exist as a liquid at temperatures where many organic chemicals are stable.

Water’s example illustrates the range of stability that life requires, or can tolerate. This includes the very short, almost transient water stability of certain crucial metabolites to the very long lifetimes of some molecules that are essentially completely water-stable. For example, carbamoyl phosphate, a key metabolite in the urea cycle, has a half-life to hydrolysis of between 0.5 and 1.5 s at 122°C (Bains et al., 2015), the temperature at which Methanopyrus kandleri can grow (Takai et al., 2008), and NADH, a central enzyme cofactor in metabolism, has a half-life of less than 10 min. On the contrary, DNA in bone is hydrolyzed at a rate of around 5.5·10⁻⁶ bases/year (Allentoft et al., 2012), and DNA in bacterial spores can remain sufficiently intact to support life for 20 million years or more (Cano and Borucki, 1995). Stability therefore requires that a substantial chemical space be stable over a range of timescales. Liquid water provides this range of stabilities.

3.1.4. Liquid water: Solvent chemical functionality

Water is also an essential reagent in terrestrial biochemistry, and as such water is not just a solvent but is a participant in biochemistry. It is widely stated that water’s chemical properties are important in the chemistry of life, and specifically for selectivity at the OOL (National Research Council, 2019; Hoehler et al., 2020). Moreover, water’s ability to solvate protons and hydroxyl ions is central to a wide range of biochemistry, including proton gradient-based bioenergetics. We have touched on the chemical role of water in biochemistry above and will not elaborate on it further here.

3.1.5. Liquid water: Conclusion

It need not be stated that water is a solvent for biochemistry. However, the analysis above suggests that it is indeed an optimal solvent for all four of the criteria presented in this article.

3.2.1. Liquid ammonia: Occurrence

Ammonia is unlikely to fulfill the occurrence criterion, for the following two reasons.

First, ammonia is photochemically labile, being broken down to form H₂ and N₂ as photolysis end-products (Strobel, 1975; Huang et al., 2022a). N₂ is extremely stable and unreactive, unlike OH and O₂, so the nitrogen from photochemically broken ammonia is permanently trapped as N₂. In the absence of life, ammonia is only observed in a planetary context in giant planets, where NH₃ is regenerated from N₂ and H₂ at high temperatures and pressures deep in the atmosphere (Moeckel et al., 2023). Ammonia can be maintained in the atmosphere of a rocky planet if it is regenerated by life (Seager et al., 2013; Huang et al., 2022a), but as discussed above, life that relies solely on ammonia as a solvent that is regenerated in this way is walking a perilous ecological tightrope.

The photochemical lability of ammonia is not a direct issue for a subsurface ocean such as that under the surface of Jupiter’s moon Europa. However, a Europa-style moon can only form with a water/ammonia ocean if the ocean is at least 80% water. The density of solid ammonia and of the water ammonia eutectic is greater than that of the corresponding liquid phases up to at least 2000 bar (Hogenboom et al. 1994). An “ammonia Europa” would therefore tend to lose liquid ammonia from its subsurface ocean to the surface over geological time, there to be irreversibly removed by photolysis. An ammonia/water ocean below an ammonia/ice shell is only stable with more than ∼40% water in the mixture, the chemistry of which is more similar to that of water than of anhydrous liquid ammonia.

The photochemical lability of ammonia is mirrored by its thermal lability. Ammonia is rarely present in volcanic gases, and when detected there (probably, at least in part, from biogenic sources), it is present at an abundance that is at least 10-fold lower than N₂ gas. Ammonia is readily broken down by thermal processing, even in a high pressure, strongly reducing environment (Larson and Dodge, 1923). Therefore, even in the absence of photochemical breakdown, volcanic or tectonic processing will reduce the ammonia content of an ammonia water ocean.

Second, ammonia and water are expected to co-occur on planets and moons, and water is expected to be more abundant due to the higher cosmic elemental abundance of oxygen over nitrogen [and hence higher abundance of water over ammonia in interstellar clouds (Wooden et al., 2004)] and due to the photochemical lability of ammonia compared with water in environments exposed to UV radiation. As water and ammonia are completely mutually miscible (Leliwa-Kopystyński et al., 2002), any ammonia will end up in an ammonia–water mixture, probably one dominated by water, and the chemistry of such a mixture will essentially be the chemistry of alkaline water. Freezing an ammonia–water mixture cannot generate pure ammonia, as ammonia hydrates melt at a lower temperature than ammonia itself (Chua et al., 2023). In principle, a mixture of >80% ammonia and <20% water could be frozen to generate pure ammonia ice, the remaining eutectic liquid could then be removed before it froze, and the ammonia ice thawed to create nearly anhydrous liquid ammonia (a scenario similar to one postulated for formamide below). However, this is a highly contrived scenario and unlikely to happen outside the laboratory. Thus, although a sufficiently cold, massive rocky planet might retain its ammonia against photolysis over geological time, ammonia is still very unlikely to be a solvent in its own right, rather than a solute in water oceans.

In conclusion, while ammonia may play a bigger role in exoplanet biochemistry than it does on Earth, as speculated by Seager et al. (2013) and Huang et al. (2022a), it is unlikely to be a solvent in its own right (Seager et al., 2013; Bains et al., 2014; Huang et al., 2022a). Protecting the ammonia from photolysis in a subsurface ocean or on the surface of a “rogue planet” (Scholz et al., 2022) would make ammonia as a solvent more plausible, but to be a solvent in its own right ammonia would have to be at least 10 times as abundant as water, which seems implausible. We note that in such an environment, liquid ammonia would have similar viscosity to water at 100°C, which is known to be compatible with life (see Supplementary Data S1).

3.2.2. Liquid ammonia: Solvation

As a solvent, ammonia is almost as good a solvent for life as water. It has a high dipole, a hydrogen bonding network that facilitates solvophobic interactions and the structures they enable (Griffin et al., 2015). It is a good solvent for a wide range of compounds, including inorganic salts (Hunt, 1932), and has similar chemical reactivity as water (Franklin, 1905; Griffin et al., 2015; National Research Council, 2019). A wide range of nitrogen analogs of terrestrial biochemicals that might be favored in an ammonia solvent or an ammonia-rich environment have been suggested (National Research Council, 2019).

3.2.3. Liquid ammonia: Solute stability

In general, solutes are more stable in liquid ammonia than in water simply because liquid ammonia is colder than water, with a freezing point of −77°C and a boiling point of −33°C at 1 bar. Ammonolysis reactions are well known but refer to the breakdown of substances in aqueous ammonia solutions rather than in pure ammonia (Stevenson, 1948).

The chemical diversity available to chemistry in liquid ammonia is higher than that in water. Classes of compounds that are unstable to hydrolysis by water, such as silanes, germanes, and arsanes, are stable in liquid ammonia (Fernelius and Bowman, 1940), so the chemical space available to life in liquid ammonia would be larger than that available to life in water. Compounds that would be extremely unstable at terrestrial temperatures, such as diazomethane and azidomethane, are potentially stable at the temperature of liquid ammonia (or ammonia/water mixtures) (Raulin et al., 1995). Nitrogen-containing analogs of phosphates and carbonyl groups are known and likely to be favored over their oxygen equivalents in liquid ammonia (Aspinall et al., 2002; Bains, 2004; Benner et al., 2004). Ammonia analogs of peptides and sugars are known or have been plausibly modeled (Firsoff, 1963; Raulin et al., 1995; Oliveira et al., 2014), and nucleic acids with one or more oxygen replaced by an NH group are well known from terrestrial biology and pharmacology (Oliveira et al., 2014). The idea of an ammonia analog of terrestrial biochemistry is therefore well established.

The stable solvation of metals in liquid ammonia allows the formation of organometallic compounds that would be extremely reactive in water, such as alkyl sodium and potassium compounds (Kraus, 1940). The relatively facile reaction of these chemicals with each other and with other compounds to make, rearrange, and break carbon–carbon and carbon–heteroatom bonds could compensate for the inherently slower reactivity of all chemistry at liquid ammonia temperatures.

3.2.4. Liquid ammonia: Solvent chemical functionality

Ammonia is a protonating, hydrogen bonding solvent similar to water but also has significant chemical differences from water. Ammonia does not solvate protons as well as water, and dry ammonia does not ionize to the same degree (K_eq of the reaction NH₃↔NH₄ ⁺ + NH₂ ⁻ is ∼10⁻³³ at 220K) (Greenwood and Earnshaw, 1997). In dry ammonia, proton gradients are therefore less likely than in water, but electron gradients could be formed instead, as dry liquid ammonia solvates electrons and elemental alkali metals (Greenwood and Earnshaw, 1997). The ability to solvate electrons makes liquid ammonia a much more flexible solvent for redox chemistry than water (Lagowski, 2007). The solvation of electrons as NH₂ ⁻ ions raises the intriguing idea that bioenergetics in liquid ammonia could be powered by electron gradients rather than proton gradients.

3.2.5. Liquid ammonia: Conclusion

We conclude that ammonia is an excellent candidate solvent for life on the criteria of solvation, stability, and chemical functionality. However, it fails as a plausible candidate solvent on occurrence, as liquid ammonia is unlikely to be present on a planetary surface as a solvent in its own right distinct from an aqueous solution of ammonia in water.

3.3.1. Sulfuric acid: Occurrence

Ballesteros et al. (2019) predicted that sulfuric acid could be a common liquid on exoplanets, where it would be formed from photochemical oxidation and the subsequent hydration of volcanic SO₂ (Titov et al., 2018) or directly emitted by volcanoes (Zelenski et al., 2015). In very water-poor planetary environments, the result would be an accumulation of concentrated sulfuric acid rather than dilute acid in aqueous solution.

The example of sulfuric acid illustrates that the stability of a solvent and its eventual abundance depend on its planetary environment, for the following two reasons.

First, sulfuric acid is extremely hygroscopic, and water and sulfuric acid are completely mutually miscible. Therefore, concentrated sulfuric acid will only accumulate as a solvent in its own right (rather than as a solute in water) in an environment that is very substantially depleted of water. If a planet loses almost all its water, sulfuric acid could come to be the dominant liquid. Such a scenario has been suggested for Venus and may occur on worlds orbiting in the “habitable zone” of mature red dwarf stars if their orbits complete any planetary migration before the star cools to its equilibrium, fusion-supported temperature. Such worlds could be common around low mass red dwarf (M dwarf) stars. Such planets that are initially too hot to allow for surface liquid are sufficiently cool after tens or hundreds of millions of years and could fall into the belatedly habitable zone of that star (Tuchow and Wright, 2023). The Kelvin–Helmholtz thermal radiation from a premain sequence red dwarf can be 1000 times that of its steady-state main sequence luminosity (Hayashi and Nakano, 1963), which means that any planet close enough to the star to be in the habitable zone when the star is at its main sequence luminosity would have been baked dry in the first few tens to hundreds of millions of years of the star’s life. Subsequent volcanic emission of SO₂ and H₂SO₄ would then generate sulfuric acid, which could condense into surface liquid. Sulfuric acid’s abundance in planets around M dwarfs will depend on the relative timing of planetary migration compared with cooling of the primary. The occurrence of sulfuric acid on planets around Sun-like stars will depend on how common planets such as Venus are (Arney and Kane, 2020).

Another scenario for dehydrating a small planetary body is through tidal heating, as illustrated by Jupiter’s moon Io. Io’s surface “ices” are almost exclusively sulfur compounds, primarily SO₂ and elemental sulfur, with only trace amounts of H₂S and H₂O (Nash and Betts 1998). With surface temperatures ranging from −143°C to −130°C (Spohn et al., 2014), Io is too cold for even 94% w/w sulfuric acid to be a liquid. However, a larger satellite (sufficient to retain a 6 millibar atmosphere similar to Mars) orbiting a planet that was nearer to its host star than Jupiter (and hence had a higher black body surface temperature of more than −33°C) could evolve to become a tidally heated super-Io, which could then support surface (or subsurface) liquid sulfuric acid.

The scenario envisaged here is different from any that could give rise to a water-depleted ammonia ocean. The water-stripping mechanisms that apparently operated on Venus have not depleted atmospheric and crustal sulfur reservoirs from which sulfuric acid can be generated. By contrast, any process that depletes an atmosphere and crust of water is likely to also deplete it of ammonia. Thus, a “dry ammonia” world is unlikely, while a “dry sulfuric acid” world is not only plausible but has a solar system precedent.

Second, sulfuric acid will probably react with some minerals. Weathering of rocks by aqueous sulfuric acid solutions is well known (e.g., Golden et al., 2005; Hausrath et al., 2013). How concentrated sulfuric acid interacts with minerals has not been explored, but it seems likely that reaction will occur with some minerals. The unknown is how much: after all, water weathers rocks, but this does not preclude the presence of oceans on Earth. Therefore, it is unknown whether stable surface lakes or oceans of sulfuric acid could exist on a “dry” planet, but possibly they can if the surface layers of the crust are fully converted to sulfates or are made of minerals resistant to attack by concentrated sulfuric acid, such as silica.

Sulfuric acid can exist stably as clouds, which do not come into contact with the surface. Again, Venus provides an example of an environment in which clouds of sulfuric acid have probably existed for geological time.

Sulfuric acid is more viscous than water at any given temperature. Comparison of the viscosity of sulfuric acid as a function of temperature shows that above ∼30°C, its viscosity is less than the maximum seen in the cytoplasm of some Earth life (see Supplementary Data S1), so above 30°C, sulfuric acid is a plausible solvent on viscosity grounds. If 30°C is a genuine lower limit for sulfuric acid as a solvent for viscosity reasons, then if the temperature dropped below 30°C, a sulfuric acid-based organism’s cytoplasm would effectively “vitrify,” as does water in supercooled terrestrial organisms. This change would not necessarily kill the organisms, but they would have to be warmed up again to continue replication. However, it is not firmly established that more viscous solutions cannot allow biochemistry.

We conclude that liquid sulfuric acid fulfills the occurrence criterion for clouds, and probably fulfills it for surface liquid, but that further work on rates of reactions with plausible surface rocks is needed to confirm this conclusion.

3.3.2. Sulfuric acid: Solvation

Recent modeling (Bains et al., 2021b) and experimental studies (Seager et al., 2023; Spacek et al., 2023; Petkowski et al., 2024; Seager et al., 2024b) have shown that sulfuric acid can stably dissolve a wide range of substances, including the bases from DNA and many biological amino acids. Extensive research shows that diverse surfactants can form micelles, and may be able to form liposome-like structures, in concentrated sulfuric acid (McCulloch, 1946; Miron and Lee, 1963; Steigman and Shane, 1965; Menger and Jerkunica, 1979; Müller and Miethchen, 1988; Müller, 1991a, Müller, 1991b; Müller, 1991c; Müller and Giersberg, 1991; Müller and Giersberg, 1992; Müller and Burchard, 1995; Torn and Nathanson, 2002). Some of these are compounds that are uniquely stable in concentrated sulfuric acid, such as carbonium compounds (Menger and Jerkunica, 1979). Interestingly, water is found to be a chaotrope in concentrated sulfuric acid, disrupting fatty acid micelles (Steigman and Shane, 1965); this illustrates the unexpected and complex behavior of this liquid. A range of polymers are known that are stable in sulfuric acid, some of which are soluble, others insoluble (Bains et al., 2021b). Sulfuric acid also dissolves a wide variety of metal ions. We conclude, as summarized before (Bains et al., 2021b), that sulfuric acid has good solvation properties for life.

3.3.3. Sulfuric acid: Solute stability

The physical and chemical properties of sulfuric acid change dramatically with the addition of 20% w/w water (Liler, 2012); hence, the stability of solutes in that acid changes with acid concentration as well. Eighty percent w/w sulfuric acid acts similar to a very strong aqueous acid, its chemistry dominated by H₂O and H₃O⁺. Sulfuric acid of >90% w/w concentration acts as an oxidizing, protonating solvent, with very different reaction mechanisms. (We note that this is different from the ammonia example above, where even 1% of water in ammonia dramatically changes its ability to perform some chemistries such as dissolving alkali metals.) The consequence of these characteristics is that reaction kinetics can change in a complex way as acid concentration increases, and some substances can actually be more stable in concentrated sulfuric acid than in dilute acid (see examples in Supplementary Data S1). Many components of terrestrial biochemistry would be rapidly and completely destroyed in pure sulfuric acid solvent, notably any biochemicals containing sugar moieties such as ATP, NADH, or RNA, so any sulfuric-acid-based biochemistry would be completely different from terrestrial water-based biochemistry.

However, a range of diverse chemicals are predicted to be stable in sulfuric acid at concentrations of >90% (Bains et al., 2021b), and reactive organics have been found to form specific stable products in sulfuric acid (Spacek et al., 2023), and not “tar” as might be expected. Surprisingly, some components of terrestrial biochemistry are stable for periods of months in 81% w/w and 98% w/w sulfuric acid, including DNA bases (Seager et al., 2023) and 19 of the 20 proteinaceous amino acids (Seager et al., 2024a). These include glutamine, serine, and cysteine as well as hydrophobic amino acids such as valine, alanine, and leucine, showing that amino acids with chemically diverse side-chains are stable, although some side-chains are reversibly modified by sulfation. Diverse polymers are stable in sulfuric acid (reviewed in Bains et al., 2021b), some of which are soluble, some not. Carbonyl groups are readily protonated in concentrated sulfuric acid, and as a result, some carbonyl compounds such as aldehydes are highly reactive. The carbonyl group is central to terrestrial biochemistry. However, Benner has pointed out that the vinyl group (C = CH₂) has similar reactivity in concentrated sulfuric acid as the carbonyl group (C = O) has in water (Benner et al., 2004) and so could provide an equivalent function.

If an environment contains a range of sulfuric acid concentrations, then organisms could adapt to this in either way. Organisms could use biochemistry that was appropriately stable in all concentrations of sulfuric acid, or they could adopt a dormant, protective form in lower concentration acid and only maintain active metabolism in concentrated acid >95%. The latter of these hypothetical strategies is analogous to terrestrial organisms that adapt to transiently low water environments by forming dormant forms, which can then germinate and return to active metabolism when the water activity rises (Clegg, 2001). The former would restrict the chemical space available to life, as only a subset of molecules that are stable in 98% w/w sulfuric acid are also stable in sulfuric acid containing significant water (see Supplementary Data S1 for some examples of this effect, and further discussion in Section 5).

We conclude that diverse chemistry is stable in sulfuric acid. However, sulfuric acid is an aggressive solvent, the reactivity of which toward organics has not been systematically explored, so whether sufficient chemical diversity exists to support a biochemistry has not yet been experimentally demonstrated.

3.3.4. Sulfuric acid: Solvent chemical functionality

Sulfuric acid is a polar, protonating solvent similar to water, with a strong hydrogen bonding network and a dipole moment similar to that of water. Sulfuric acid provides as much chemical functionality as water. The solvent self-ionizes to form positive and negative ions HSO₄ ⁻ and H₃SO₄ ⁺ (Cox, 1974) more readily than water forms OH⁻ and H₃O⁺ (∼10⁻³ M ions in 100% acid vs. 10⁻⁷ M ions in 100% water at room temperature), which can provide ions for charge neutralization, ion solvation, polymer stabilization, and charge separation for energy capture. Concentrated sulfuric acid shows solvophobic effects, as noted above. It can reversibly sulfate a range of chemical functional groups such as alcohols and thiols and can drive reversible protonation of aromatic systems (as well as sulfonation and degradation reactions). Sulfuric acid therefore has a similar potential to participate in biochemistry to water.

3.3.5. Sulfuric acid: Conclusion

We conclude that sulfuric acid is a surprisingly plausible candidate solvent for life on the criteria of solvation, chemical functionality, and occurrence. On the criterion of stability, it presents some challenges, as it is quite reactive, but preliminary work suggests that it can satisfy this criterion as well.

3.4.1. Liquid sulfur: Occurrence

Elemental sulfur is the third most common liquid erupted by volcanoes (after water and magma) on Earth (Mora Amador et al. 2019) and can form extended pools of liquid sulfur on the surface (Oppenheimer and Stevenson, 1989) or below a water layer (Takano et al., 1994; De Ronde et al., 2015; Sydow et al., 2017; Malyshev and Malysheva, 2023) in subaerial and submarine volcanoes. Liquid sulfur is also erupting on Io (Schneider and Spencer, 2023) and could be present in volcanic systems on Venus. However, such volcanic systems are problematic as the basis for sulfur-based life, for the following two reasons.

First, they are transient, so at least life based in volcanic liquid sulfur would have to consist exclusively of organisms that were occasionally “wetted” with solvent, rather than being bathed in it continuously. While this is not a complete barrier to life, as illustrated by the existence of desert flora on Earth that are only occasionally wetted by water, it makes the habitat more perilous. Transient volcanic sources of liquid sulfur might include subsurface reservoirs, but terrestrial precedent suggests that subsurface reservoirs of liquid sulfur do not persist over geological time.

Second, liquid sulfur’s properties vary substantially with temperature. Between 120°C and ∼155°C, it consists mainly of S₈ rings, but above 155°C, the predominant molecular species changes to polymeric chains. As a result, the viscosity of the liquid increases more than 10,000-fold, dielectric constant increases, and other chemical properties change significantly (Powell and Eyring, 1943). Such dramatic changes in liquid sulfur’s properties mean that in reality it is effectively composed of “two solvents” that interchange over a relatively narrow temperature range. Specifically, the viscosity of sulfur above ∼170°C far exceeds that in which any biochemistry is observed to occur on Earth. We note that quite small amounts of impurity can change this behavior dramatically. For example, the addition of 0.25% elemental iodine can reduce the viscosity of liquid sulfur at 180°C 1000-fold (Powell and Eyring, 1943). The presence of such impurities, however, is still insufficient to bring the viscosity of liquid sulfur within the ranges seen in terrestrial organisms (see Supplementary Data S1 for more on viscosity constraints). This complex temperature-dependent behavior makes liquid sulfur an effective solvent for life only between 115°C and ∼170°C, a narrow temperature range for volcanic systems. The only plausible habitat in which liquid sulfur would be present for extended periods would be a planetary surface that was between this narrow temperature range. Such surface conditions would have to be also free of oxidants or reductants that would convert sulfur to SO₂ or H₂S, respectively. Liquid sulfur also reacts with water at >120°C (Ellis and Giggenbach, 1971), in the reaction the overall stoichiometry is as follows: 4S + 4H 2 O → 3H 2 S + H 2 SO 4 , so an environment hosting liquid sulfur as a solvent would also have to be extremely dry.

In summary, we conclude that the occurrence criterion is hard to meet for liquid sulfur, but we do not rule it out.

3.4.2. Liquid sulfur: Solvation

Almost nothing is published on the solubility of organic materials in liquid sulfur. The only published material is on the solubility of hydrogen sulfide and sulfur dioxide (Fanelli, 1949; Touro and Wiewiorowski, 1966; Marriott et al., 2008). Sulfur is not a polar or a protonating solvent, so it would be expected to be a hydrophobic solvent and to dissolve polar molecules poorly. However, calcogenic metals complex sulfur very well (e.g., Dravnieks, 1951), and elements such as As, Sb, Se, Te, Hg, and Cu are found to be concentrated in fumarolic sulfur (Shevko et al., 2018). Beyond this, we cannot speculate.

3.4.3. Liquid sulfur: Solute stability

Solute stability in liquid sulfur has not been explored. Anecdotally, solidified sulfur from volcanic liquids is almost always coated in black precipitate, suggesting reaction with environmental material (e.g., Sydow et al., 2017; Malyshev and Malysheva, 2023). Liquid sulfur contains open chains at all temperatures, with sulfur radicals or ions at the ends, which would be expected to be reactive toward all molecules dissolved in it. All chemical reactions happen faster with elevated temperature, so the high melting point of sulfur mitigates against the potential thermal stability of organic compounds dissolved in it. This combination of factors leads us to expect that few classes of organic compounds will be stable in liquid sulfur. However, we thought the same of concentrated sulfuric acid and have been proven wrong, so the expectation of the instability of organic compounds in liquid sulfur awaits experimental verification.

3.4.4. Liquid sulfur: Solvent chemical functionality

Liquid sulfur is quite chemically active, as noted above. It is likely to be a good mediator of redox reactions, as sulfur can stably exist in many redox states. Sulfur can also mediate reductive photochemistry (Li et al., 2022). However, the chemistry of liquid sulfur has not been explored widely, so we cannot draw solid conclusions on its chemical functionality.

3.4.5. Liquid sulfur: Conclusion

We conclude that liquid sulfur is not a promising candidate solvent for life. Evidence is very limited, but what evidence there is suggests that it will perform poorly on the criteria of occurrence and solute stability, and performance on solvation is essentially unknown.

3.5.1. Liquid hydrogen sulfide: Occurrence

Hydrogen sulfide is the fourth most common gas emitted by terrestrial volcanoes (after H₂O, CO₂, and SO₂). The fraction of sulfur outgassed by terrestrial volcanoes as H₂S depends on the redox state of the volcanic vent system and can be as high as 95%, although SO₂-rich, H₂S-poor volcanic outgassing predominates in Earth’s oxidized crust (see Supplementary Data S1, Section 1). On a planet with a more reduced crust, H₂S would be the dominant outgassed sulfur species. Despite its potential abundance on rocky planets, H₂S is a reactive chemical unlikely to accumulate as a liquid on the surface of a planet. We discuss these limitations on the occurrence of hydrogen sulfide in detail below.

Unlike ammonia, hydrogen sulfide is not miscible with water (Selleck et al., 1952), it is poorly soluble in water (Chapoy et al., 2005), and water is relatively insoluble in liquid hydrogen sulfide (Santoli et al., 1999). It is therefore plausible that liquid hydrogen sulfide could exist on the surface of a rocky body where water is present (as ice) if the temperature and pressure conditions allow. However, liquid H₂S and liquid CO₂ are miscible (Francis, 1954), and CO₂ is likely to be much more abundant than H₂S, even in the atmosphere of a planet with a reduced crust or mantle. Liquid H₂S is therefore only likely to be the dominant solvent component in an environment where CO₂ is much rarer than H₂S, either because sulfur was much more abundant than carbon (which given the cosmic abundance of both elements is unlikely) or because carbon was overwhelmingly present as methane; hydrogen sulfide is readily soluble in liquid methane, and methane is readily soluble in liquid hydrogen sulfide, but the two liquids are immiscible (Kohn and Kurata, 1958). A methane-dominated environment would require a Titan-like planet with a surface temperature above −60°C—a planet that might be termed a “Cool Titan” by analogy with the reducing “Warm Titan” atmosphere postulated for a postimpact early Earth (Zahnle et al., 2020). We note, however, that the terrestrial “Warm Titan” model converts essentially all its CH₄ to CO₂ on a timescale of 2–20 MY (Zahnle et al., 2020) through atmospheric photochemistry.

More critically, H₂S is photochemically short-lived in almost any anoxic atmosphere, whether dominated by N₂, H₂, or CO₂ (Hu et al., 2013). The vapor pressure over H₂S even at its freezing point of −86°C is 0.21 bar (by comparison the vapor pressure of water at 0°C is 0.005 bar), so if an H₂S ocean or lake is not to rapidly “dry out,” then the atmosphere above it must have at least 0.21 bar of H₂S. The required atmospheric pressure of 0.21 bar of H₂S implies a high concentration of H₂S in the atmosphere where it can be photochemically destroyed. The photolysis rate of H₂S is 100 times that of water (Hu et al., 2013). The products of the photolysis of H₂S are hydrogen and elemental sulfur; they do not recombine to re-form H₂S.

As in the case of ammonia discussed above, the photochemical reactivity of H₂S is mirrored by its thermal lability (Gautam et al., 2024). Processing of the H₂S reservoirs through volcanoes would result in the removal of H₂S by breakdown to H₂ and elemental sulfur.

Similarly to ammonia, the accumulation of hydrogen sulfide would require a minimal volcanic activity and a UV-shielded environment to prevent rapid photolysis. Such an environment would also shield methane from photolysis, so a UV-shielded, highly reduced “Cool Titan” world could in principle accumulate stable liquid H₂S lakes or oceans. The “Cool Titan” surface rocks would also have to be largely devoid of transition metals or transition metal oxides, as H₂S reacts readily with these chemicals to form sulfides or elemental sulfur (Poulton et al., 2002; Georgiadis et al., 2020). While such a planetary environment can be imagined, no precedent for it is known, so we must consider the occurrence of liquid H₂S as very unlikely.

There are no reported measures of the viscosity of liquid H₂S, but as its density, boiling point, dipole moment, and hydrogen bonding potential are intermediate between those of methane and ammonia, we would expect a viscosity intermediate between these two solvents, hence at the lower end of the range known to be compatible with life. We note, however, that to confirm this hypothesis the viscosity of H₂S would need to be tested experimentally.

3.5.2. Liquid hydrogen sulfide: Solvation

Many organic and inorganic compounds stably dissolve in liquid hydrogen sulfide. These include alkanes, aromatic hydrocarbons, such as anthracene and naphthalene, alcohols, thiols and thioethers, and trichloroacetic acid (Quam, 1925). Other compounds are stable to reaction with liquid H₂S but insoluble. Examples of such insoluble compounds include citric acid, several sugars, cystine, and glycine. However, Quam (1925) gives no quantitation of solubility. Most inorganic compounds appear to be insoluble in liquid H₂S (Wilkinson, 1931). For example, elemental sulfur itself dissolves only sparingly in liquid hydrogen sulfide (Smith et al., 1970).

Whether soluble and insoluble moieties can be combined to form H₂S-amphipaths, a crucial characteristic of any biochemistry, has never been explored.

3.5.3. Liquid hydrogen sulfide: Solute stability

As mentioned above, Quam (1925) reports that many organic and inorganic compounds that are soluble in H₂S are also stable. However, their stability is not quantified. A number of compounds, both soluble and insoluble in liquid H₂S, react with it. Many of the tested compounds are also reactive with water, such as metalloid halides, PCl₃, and SeCl₄, although SiCl₄, which reacts almost instantly with water, is stably soluble in liquid hydrogen sulfide (Wilkinson, 1931).

We are not aware of any studies of amphipathic or detergent properties of substances in liquid H₂S, but both the solvation properties listed above and the physical properties of liquid H₂S suggest that liquid H₂S amphipaths should exist. Liquid H₂S does not show the dense, strong hydrogen bond network seen in water, but H₂S molecules in liquid are nevertheless weakly orientated with respect to each other (Santoli et al., 1999), which will provide a weak entropic drive to inserting solutes into the solvent structure analogous to the hydrophobic interaction in water. The specific heat capacity of liquid H₂S is similar to that of water (52–63 J/mol/K, depending on temperature, vs. 75.6 J/Mol/K for water at Standard temperature and pressure), which also supports significant internal structure in liquid H₂S.

In summary, the limited experimental data suggest that diverse compounds will be stable when dissolved in liquid hydrogen sulfide.

3.5.4. Liquid hydrogen sulfide: Solvent chemical functionality

Liquid H₂S can participate in a wide range of chemistry at Earth surface pressures and temperatures of −60°C (e.g., Quam, 1925; Wilkinson, 1931). H₂S is expected to be a much better medium for redox reactions than water, both because sulfur is a very redox-active element and because H₂S can readily, and reversibly, form chains in liquid H₂S. Such polymerization reactions allow for a variety of chemical reactions that are unique to this solvent, including a polythiol-mediated redox and radical chemistry (e.g., Dogadkin, 1958; Huang et al., 2020). Liquid hydrogen sulfide self-ionizes to form H₃S⁺ and HS⁻ ions to a much smaller extent than water or ammonia (Satwalekar et al., 1930; Wilkinson, 1931) but can dissolve ionic compounds to provide for ion gradients; some organic thioacids such as dithioacetic acid and dithiobenzoic acid and some base analogs such as triethylammonium hydrosulfide ionize in liquid hydrogen sulfide (Wilkinson, 1931; Smith, 1951). Such ionization implies that chemistry that relies on proton gradients or proton currents could be compatible with liquid hydrogen sulfide, which adds to its potential as an active participant of biochemistry.

3.5.5. Liquid hydrogen sulfide: Conclusion

We conclude that, similar to ammonia, liquid hydrogen sulfide has chemical properties, for example, in terms of solvation, that make it a potentially promising solvent for life. However, H₂S is very unlikely to occur as a liquid in any stable body on the surface of a rocky planet.

3.6.1. Liquid HF: Occurrence

It is hard to imagine an environment where substantial amounts of HF would be present as liquid for geological periods of time, for the following two reasons. First, while volcanoes outgas HF, they outgas far more water (see Supplementary Data S1). Therefore, for an HF ocean to form, as opposed to a water ocean containing HF, volcanic gases would have to have negligible oxygen and at the same time have at least some hydrogen. If they were completely depleted of water, then there would be no hydrogen to form HF, but if hydrogen was present in the presence of oxygen, water would form. It is not clear if an oxygen-depleted, hydrogen-containing planetary crust is realistic; the only possible scenario would be an extreme carbide planet, where carbon is more abundant than oxygen and oxygen is correspondingly outgassed exclusively as CO₂. Such planets have been postulated (Kuchner and Seager, 2005), but whether they exist is unknown.

The second reason is that HF reacts with silica, so unless the putative HF-ocean planet had a crust composed of nonsilicate, nonmetallic material, any HF ocean would react with crustal rocks to form fluorides and water. On a carbide planet, silica would be replaced largely by silicon carbide, which is also attacked slowly by HF (Habuka and Otsuka, 1998). This reaction may be very slow at low temperatures, so a cold planet with a surface temperature near the melting point of HF (−83°C) with an excess of HF and no water could, in principle, accumulate HF lakes or oceans.

3.6.2. Liquid HF: Solvation

HF has been used experimentally as a solvent and catalyst in organic chemistry, and proteins and amino acids are surprisingly stable in liquid HF below 0°C (Lenard, 1969; Norell, 1970; Polazzi et al., 1974). HF is likely to dissolve metals as well. We are not aware of any systematic study of solvation of organics in HF. Gore (1869) provides a long list of inorganic materials’ interaction with HF (over half react violently, most of the rest do not dissolve), and a sample of organics that stably dissolve, including mono- and disaccharides, caffeine, indigo, and nitrocellulose. Gore (1869) reports that moss and sponge were little affected by liquid HF, although the article did not report if they actually survived as living material. HF forms very strong hydrogen bonds, which form chain structures in liquid, rather than branching networks such as the hydrogen bonds in water, sulfuric acid, and ammonia (Maybury et al., 1955; McLain et al., 2004). However, the hydrogen bonding structure of liquid HF is sufficient to drive a solphophobic effect, and micelle formation has been observed experimentally in liquid HF (Zeiseler et al., 1992; Peters and Miethchen, 1993; Roth et al., 1995).

3.6.3. Liquid HF: Solute stability

As noted above, HF can stably dissolve some substances, in part, because it remains liquid to −83°C, at which temperature all chemistry will be very slow. (We note that the solubility of substances in liquid HF at these low temperatures has not been explored.) The functional diversity of stable solutes in HF is not known.

3.6.4. Liquid HF: Solvent chemical functionality

HF is a protonating solvent but does not self-ionize. HF is a powerful oxidizing agent and hence chemically active. It may therefore meet some of the chemical functionality criterion.

3.6.5. Liquid HF: Conclusion

We conclude that HF is an implausible solvent for life. While it passes on the criterion of solvation and can plausibly be argued to meet chemical functionality, it is questionable on the criterion of solute stability and fails on occurrence.

3.7.1. Formamide: Occurrence

Pure formamide would have required a very specific set of circumstances to form at all. A scenario suggested by Bada et al. (2016) requires the formation of a concentrated solution of formamide, cooling to between 0°C and 2.6°C to precipitate out pure formamide, removal of the water, and then rewarming to form liquid formamide. Other scenarios that lead to the formation of abundant formamide are likely to be equally convoluted. Thus, while liquid formamide might have occurred in rare, specific circumstances, it is extremely unlikely to provide for long-term lakes or oceans of formamide as a solvent in the planetary context. Formamide is also photochemically labile (Boden and Back, 1970), breaking down to CO, H₂, and NH₃, the NH₃ itself is photochemically labile. Thus, even if formamide lakes did form, they would not be stable to long-term photochemical destruction of the formamide from stellar UV.

3.7.2. Formamide: Solvation

Formamide is a powerful solvent, dissolving a wide range of materials. It has a high dielectric constant, which allows it to effectively dissolve polar molecules, but it is not hydrogen bonding and so can dissolve more hydrophobic molecules than can water. While it dissolves a larger breadth of molecules than water, it does not dissolve everything across the board. Some amphipaths do form micelles in formamide (e.g., Couper et al., 1975; Akhter and Alawi, 2000), and some substances, for example, the complex hydrocarbons in pitch, are virtually insoluble in formamide (Papole et al., 2010) as are some polymers (e.g., Jousset et al., 1998). Many metal salts are soluble in formamide, although the patterns of solubility differ from water; for example, ammonium chloride is very soluble in water but practically insoluble in formamide (Magill, 1934). Formamide therefore fulfills the solvation criteria of being able to dissolve a range of compounds, but not all compounds.

3.7.3. Formamide: Solute stability and solvent chemical functionality

Formamide is relatively unreactive and is not seen to react with a wide range of solutes. It is a participant in some chemistry, as noted above in the OOL literature. Materials such as hexamethyldisilazane, which would be rapidly hydrolyzed in water, are stable (although insoluble) in the presence of formamide (Grunert, 2002). Formamide therefore can stably dissolve a larger chemical space of molecules than water, but itself provides only limited chemical functionality.

3.7.4. Formamide: Conclusion

We conclude that formamide is an unlikely solvent for life. It can provide for solvation, solute stability, and some chemical functionality but is extremely unlikely to occur on planetary surfaces.

3.8.1. Carbon dioxide: Occurrence

CO₂ is a common and widely distributed compound. It is the dominant component of the atmospheres of Mars and Venus, a significant component of Earth’s atmosphere, and the second most common volatile emitted from terrestrial volcanoes (see Supplementary Data S1, Section S1). CO₂ is liquid over a range of temperatures and pressures, although not a range that is found on Earth’s surface today. Liquid CO₂ can be found in Earth’s ocean floor sediments at 4°C and tens of bar pressure, resulting from accumulation of volcanic CO₂ (Sakai et al., 1990; Konno et al., 2006; de Beer et al., 2013). Submarine liquid CO₂ is quite a rare find, as it is less dense than water, so it tends to leak upward as “bubbles” to higher in the ocean, where the lower pressure allows it to boil into CO₂ gas. It has been suggested that the cold, dense atmosphere of early Mars could have resulted in liquid CO₂ flow, and that liquid CO₂ could form at limited sites in the recent past (Hecht et al., 2024). Graham et al. (2022) studied whether a “Cold Venus” planet, with a dense CO₂ atmosphere but a low surface temperature, could have a liquid CO₂ ocean under its CO₂ atmosphere. Unlike Earth, where the boiling point of the constituents of the ocean and the atmosphere is quite distinct, the Cold Venus’ ocean would expand or contract substantially with changes in insolation, pressure, and “weather.” However, Graham et al. (2022) considered that the ocean could be stable over geological time. The “Cold Venus” would form through the same scenario envisaged for a desiccated red dwarf planet above (Section 3.3.1), but with a more distant orbit and without the requirement to remove all the planet’s water, as the solubility of water in CO₂ is quite low.

The other phase in which CO₂ has been discussed as a solvent for life is as a supercritical fluid (Budisa and Schulze-Makuch, 2014). Supercritical CO₂ occurs in the atmosphere of Venus in our own solar system, but we should be aware that Venus’ ground-level atmosphere is not a solvent for life, for the following two reasons. The first, an obvious reason, is that it is too hot to allow the stable existence of organic compounds. The second is that “supercritical” simply means above the critical temperature and pressure. There is no phase change in such fluids on compression, so a supercritical fluid under sufficient pressure can have a high density without being a liquid. The term “supercritical fluid” is often taken to mean substances that have a density that approaches that of the liquid phase because they are near their critical temperature and have been compressed to high pressure. Supercritical CO₂ in this sense has been suggested as a solvent for life (Budisa and Schulze-Makuch, 2014). Venus’ ground-level atmosphere, while technically supercritical, does not meet this criterion. Supercritical CO₂ needs to be dense supercritical CO₂ to dissolve large molecules, especially polymers. Such a scenario is discussed further below.

The link between density, temperature, pressure, and solubility of substances is expanded on in Supplementary Data S1. In this study, we note that, to be a plausible solvent for life, supercritical CO₂ must be under at least 100 bar pressure and be at moderate temperatures, probably below 100°C, but above the critical temperature of 30.98°C. The powerful greenhouse effect of CO₂ makes this a difficult scenario to fulfill in any realistic planetary environment.

3.8.2. Carbon dioxide: Solvation

CO₂ has no permanent dipole and low polarizability and does not form hydrogen bonds, so its solvent behavior is similar to organic solvents such as toluene, dissolving hydrophobic, nonpolar molecules (Hyatt, 1984). Liquid CO₂ is known to be a good solvent of a wide range of substances, including hydrocarbons, esters, crown ethers, substituted benzenes, and pyridines, but not sugars or metal salts (Francis, 1954; Gouw, 1969; Hyatt, 1984). Chelated metals, including Co, Cu, Fe, Mn, and Zn, are known to be soluble in supercritical CO₂ (Yazdi and Beckman, 1994; Ashraf-Khorassani et al., 1997; Smart et al., 1997). Some materials are amphipathic in liquid CO₂ (van Roosmalen et al., 2004) and can be used as detergents to solubilize polar molecules into liquid CO₂ (Cooper et al., 1997).

Supercritical CO₂ is widely used in industry as an extraction solvent, and the solubility of over 1600 substances, including ionic solids in supercritical CO₂, has been measured (Gupta and Shim, 2006).

Liquid CO₂ has a density of more than ∼800 kg/m³ (depending on temperature and pressure—see Supplementary Data S1, Section 4). The minimum density needed for CO₂ to effectively solubilize molecules of the sizes typically found in biological materials has been explored in the industrial use of liquid and supercritical CO₂ as a dry cleaning fluid and as an extractant (e.g., Banerjee et al., 2012). Densities of at least 400 kg/m³ (half that of liquid CO₂ at 1 bar pressure, 40% that of water) are generally required, and >800 kg/m³ for solubilization of polymers such as proteins. Liquid CO₂ fulfills these criteria, but supercritical CO₂ only fulfills them at high pressures at temperatures below 200°C; for example, at the average surface temperature of Venus (460°C), the density of supercritical CO₂ only exceeds 400 kg/m³ at ∼660 bar (7 times the surface pressure of Venus) and exceeds 800 kg/m³ at ∼1500 bar. As noted above, only a very specific set of planetary environments would allow these conditions to occur.

In addition, the ability of near-critical CO₂ to dissolve substances changes dramatically with changes in pressure (unlike liquid CO₂, or liquid water), as illustrated in Supplementary Data S1, Section 4. Dramatic changes in solvation with relatively small changes in temperature would make it effectively impossible to build structures that depended on differential solubility of their components, such as lipid bilayers or globular proteins, in an environment where the pressure or temperature can change.

3.8.3. Carbon dioxide: Solute stability

A wide range of materials can be stably dissolved in liquid and supercritical CO₂. These include chemically sensitive materials such as proteins that are likely to be damaged by other extractive or processing methods (Quirk et al., 2004; Woods et al., 2004). Supercritical CO₂ can be a medium in which biological materials can be hydrolyzed by water at elevated temperatures (e.g., Brunner, 2009; Yakaboylu et al., 2015; Zhao et al., 2019), but at low temperatures or in the absence of corrosive substances such as water, CO₂ is a very chemically benign solvent.

3.8.4. Carbon dioxide: Solvent chemical functionality

CO₂ as a solvent is unlikely to participate in the chemistry of life. CO₂ is quite chemically inert and does not self-ionize. Therefore, it is unlikely to participate in biochemistry, beyond its role to solvate life’s component molecules.

3.8.5. Carbon dioxide: Conclusion

We conclude that liquid carbon dioxide is a potential solvent for life if a planetary environment can be found in which it occurs stably as a liquid. It fulfills the criteria of solvation and solute stability, although it fails on solvent chemical functionality. By contrast, supercritical CO₂ appears less plausible as a solvent, due to the quite restricted range of conditions under which it could solubilize complex molecules and have stable, unchanging solvation properties.

3.9.1. Sulfur dioxide: Occurrence

SO₂ has a critical temperature of 157°C and could form a nonprotonating polar liquid under that temperature with sufficient pressure (Burow, 2012). On Earth, SO₂ is the third most common gas in volcanic outgassing (see Supplementary Data S1), and it may be the second most common volcanic gas on Venus. A “Cold Venus” scenario as outlined above could therefore result in sufficient atmospheric pressure and SO₂ abundance to accumulate liquid SO₂.

However, such a planet would have to be even more thoroughly dehydrated than the case suggested for sulfuric acid above. Even trace amounts of water would result in SO₂ forming sulfurous acid, which would disproportionate to sulfuric acid. Photochemical oxidation of SO₂ to SO₃ would also result in sulfuric acid if there was even a trace of water in the atmosphere; on Venus photooxidized SO₂ forms sulfuric acid even though the mesosphere has only less than 10 ppm water (Bains et al., 2021a). It is not clear if a complete stripping of all water (which means all hydrogen atoms) from the atmosphere and crust of a planet is possible.

3.9.2. Sulfur dioxide: Solvation

Liquid SO₂ has been widely studied as an industrial solvent and is known to dissolve a range of organic compounds and inorganic ions, including dimethylsulfide, benzene, naphthalene, and carbon tetrachloride. However, propane and butane are almost insoluble in liquid SO₂, which suggests that amphipaths in liquid SO₂ may exist (Ross et al., 1942; Burow, 2012).

3.9.3. Sulfur dioxide: Solute stability and solvent chemical functionality

The chemical functionality of liquid SO₂ is expected to lie between those of water and CO₂, in that SO₂ can engage in redox chemistry and addition reactions to carbonyl groups, but it is not a protonating solvent. SO₂’s ability to complex carbonyls is the reason for its consideration as a key player in some OOL scenarios (Sydow et al., 2023). SO₂ can also form solvent complexes with thioethers, such as dimethylsulfide, and with amines and ethers. This could point to liquid SO₂ having quite rich chemistry aside from its role as a solvent.

3.9.4. Sulfur dioxide: Conclusion

We conclude that sulfur dioxide is not attractive as a candidate solvent for life. While it passes the solubility and solute stability criteria and has promise for the solvent chemical functionality criterion, it is very likely to fail the occurrence criterion as a pure solvent in its own right, in any realistic planetary scenario.

3.10.1. Cryogens: Occurrence

Ballesteros et al. (2019) predict that liquid ethane should be more abundant in planetary environments than any other liquid, and liquid ethane, methane, and nitrogen between them should have 13 times the abundance of liquid water on the surface of planets and moons, including those around diverse star types. Ethane and methane are subject to photochemical destruction so will need a constant source or a regeneration mechanism to maintain them (Coustenis, 2021). Nitrogen is extremely stable, as noted above, but has a critical temperature of −145°C, so for a planet to maintain significant surface nitrogen as a liquid (unlike Triton’s transient N₂ geysers), it would need an atmosphere that remains gaseous at below −145°C, that is, one composed of hydrogen (or neon or helium). In addition, the planet would have to be sufficiently far from its star, and the internal radiogenic heat budget be sufficiently small, that the surface remained below nitrogen’s critical temperature despite hydrogen’s substantial greenhouse effect. None has explored whether such a scenario is plausible.

3.10.2. Cryogens: Solvation

All the cryogenic solvents are both nonprotonating and apolar. They are also, by definition, very cold. As the solubility of solids in liquids declines with temperature, it is unsurprising that very few molecules are soluble in cryogens. Thus, for example, liquid methane, despite being sometimes considered a possible solvent for life, is unlikely as a solvent because very few molecules dissolve in it (Petkowski et al., 2020). No polymer systems have been found that are soluble in cryogenic liquids (McLendon et al., 2015). Diverse small molecules and polymers can dissolve in higher hydrocarbons (McLendon et al., 2015), such as butane, which has a similar boiling point to ammonia, or octane. However, we do not consider higher hydrocarbons such as octane further here, as we are aware of no scenario that would provide a planet with stable liquid octane reservoirs in the absence of life.

In principle, biochemistry in cryogens could use a much wider range of chemical bonds than chemistry in water. The reactive chemistry of water substantially limits the chemical bonds that life can use; in cryogens, the low temperature would render dissolved water much less reactive, and water itself would be expected to be poorly soluble [although experimental measurements suggest that water is more soluble than expected in cryogens (Rebiai et al., 1984)]. Thus, exotic chemistry, including silicon, germanium, and selenium bonded to each other and to nitrogen, sulfur, and halogens, is expected to be stable in cryogens, greatly expanding the chemical space available to life in cryogens. However, it is doubtful whether the larger chemical space available in cryogens can compensate for the much lower solubility of all but the smallest molecules (Petkowski et al., 2020). Early work suggesting that bilayers-type multimolecular structures can be formed in cryogens from small nitrogen-containing molecules (Stevenson et al., 2015) has since been disputed (Sandström and Rahm, 2020).

3.10.3. Cryogens: Solute stability and solvent chemical functionality

At cryogenic temperatures, ethane, methane, and nitrogen are completely inert. To the limited extent that they can act as solvents, they are purely a support for the molecules they dissolve and will not participate in those molecules’ chemistry.

3.10.4. Cryogens: Conclusion

We conclude that the cryogenic solvents methane, ethane, and nitrogen are very unlikely to be solvents for life. While they may be common, they fail on the solvation, as well as on chemical functionality, criterion.