Life on Earth takes a dizzying array of forms, and fills a dizzying array of ecological niches and environments. However, all (known?) life on Earth is based on carbon chemistry operating in a solvent of water. Speculation on what types of stars are suitable for hosting life has been dominated by this fact; the vast majority of attempts to whittle down the range of target stars from billions to something managable use likelihood of supporting liquid water and organic chemicals as basic criteria.
There is, however, speculation on alternate forms of biochemistry. If life using such alternatives is found, or even if we just accept some exist for the purposes of this thought experiment, the range of environments where life can bootstrap and flourish increases.
No reasonable alternative for carbon as the backbone of organic molecules has been proposed. Silicon, frequently featured in science fiction, has an affinity for oxygen that has both reduced to trace amounts the quantities of silicon compounds equivalent to carbohydrates, such as silane — silicon's analog to methane — and makes for an energy storage problem as oxydizing the silicon equivalents to carbohydrates produces silicon dioxide — quartz — which forms solid crystals rather than an easy to remove waste gas like carbon dioxide. The other alternate biochemistry basis is various combinations of nitrogen, phosphorous, sulfur, or boron, all of which suffer from being etremely rare except nitrogen, which has the tendency to revert to simple, molecular nitrogen (N₂).
Alternative solvents are more promising. One popular one is ammonia, which offers a system of chemical reactions as rich as water, but may suffer from a narrow range of temperatures at which it is liquid. Another possibility is a mixture of liquid hydrocarbons, such as in the polar lakes on Saturn's moon Titan, with the caveat that this would be a non-polar solvent, putting into doubt the likelihood that pre-biotic membranes (proto-cells) would form. Another possibility, specifically advanced to explain some of the odd results of the Viking probe's tests for life on Mars, is a mixed water/hydrogen peroxide (H₂O₂) solvent, although this is more a case of hydrogen peroxide working as an antifreeze.
Galaxies come in a wide range of sizes and shapes. The best for long-term star formation seem to be the spiral galaxies, like our own Milky Way Galaxy. These consist of spherical central bulges dominated by old, red stars, surrouned by flattened disks of mixed stellar populations. Spiral arms, dense regions of active star formation wrap their ways through the disks. Other galaxy types are either dominated by old, red stars, indicating a low rate of star formation and a probably lack of dust for planet formation, or they're currently undergoing intense star formation, possibly as the result of colliding with another galaxy, which makes for frequent supernovae.
In their infancies, spiral galaxies are composed of just hydrogen and helium. As time goes by, stars form, fuse hydrogen to helium in their cores, then, eventually, if massive enough, successively heavier elements. As the stars make their final transitions — either the relatively gentle transition from carbon stars to white dwarfs for low-mass stars or as spectacular supernovae as supergiants transition to neutron stars or black holes — the outer layers of the star are pushed out into interstellar space. This gas, enriched in heavy elements, mixes with leftover gas, enhancing future genrations of stars.
The earliest generations of stars, lacking heavier elements that solidify when cold enough, lack planets. As the heavy element content of successive generations of stars increases, dust grains begin to form in the circumstellar disks that form around protostars. The amount of dust determines both how quickly any planets form, and how large they get. A generation of purely hydrogen/helium stars, lacking any secondaries outright, would give way to a generation with just enough dust to form debris belts similar to gas giant rings, the Asteroid Belt, or Kuiper Belt, which would give way to a generation with dozens of "microplanets", and so on. Eventually, stars would form with enough dust for terrestrial planets, which could host life.
Giant planet formation is also an important factor. The four giants in the Solar System are believed to have formed in a relatively narrow band between five and fifteen times as far from the Sun as Earth, and then jostled each other (gravitationally, that is) into their current orbits from five to thirty times Earth's orbital distance. In the process, Neptune ended up plowing through the Kuiper Belt — a belt of small, icy bodies left over from the formation of the Solar System — which deflected many icy bodies into the inner system, where they collided with the rocky planets, delivering chemicals necessary to life. However, many of the recently detected planets orbiting other stars are giants which orbit close to their host stars, where they would disrupt the orbits of Earth-like planets.
Giant planet formation begins the same as terrestrial planet formation. Dust grains form as the circumstellar disk cools, then begin bumping as they circle the protostar. The dust grains stick together when they bump due to electrostatic forces (the same that cause static cling), and steadily grow into asteroid-sized bodies. At this point, the gravitational pull becomes strong enough to start drawing in nearby debris, instead of having to wait for it to bump into the planetesimal, leading to a period of rapid growth. Once Moon to Mars-sized bodies are produced, the rapid growth stops as available debris has been all been collected, but these closely-spaced bodies' gravitations disrupt each other's orbits, leading to a period of chaotic mergers. If this results in bodies getting big enough, fast enough — before the star starts shining and evaporates the gaseous portion of the circumstellar disk — the planet's gravitational pull draws in hydrogen and helium gas. This gas forms the vast majority of the circumstellar disks' mass, causing another period of rapid growth, and the resulting drag causes the giant planet to slowly spiral in until the disk evaporates.
Systems with too little dust won't form gas giants, which seem to be essential to kickstarting life by triggering Late Heavy Bombardment-style events to deliver essential chemicals. In systems with too much dust, gas giants get a big headstart, and end up migrating in to the inner system, where they eject terrestrial planets. (There is a chance that some of the more moderate of these "Hot Jupiters" could have moons that substitute for the inner planets.) However, some of the alternate biochemistries would favor "failed cores" — icy planets that didn't get big enough to make the transition to giants.
Intense star-formation periods in the central regions of spiral galaxies make for frequent supernovae, but also quickly enhance the supply of heavy elements. Eventually, a region just outside the supernova "danger zone" builds up enough dust for the formation of hospitable planets, begining a "galactic habitable zone". Over time, this narrow band widens as material in the interior of the galaxy is used up, ending the period of rapid star formation and frequent supernovae, and dust levels rise in the outer regions of the galaxy.
Regardless of the chemical base of a biosphere, they all have one thing in common: time. It takes time for a planet or moon to form and life to bootstrap, and some stars just aren't long-lived enough. The most massive stars, those in the OBA range of spectral types (blue, blue-white, and white) run out of core hydrogen "fuel" and evolve into giant stars, which would cook any life-bearing planets, before life could actually get started. Fortunately, these stars collectively make up less than one percent of all stars. The lower mass limit is much fuzzier, especially in light of (speculative) alternatives to Earth-like (warm carbon-water) biospheres.
A strict set of stellar criteria, one designed for Earth-like planets, indicates that the FGK range of spectral types (yellow-white, yellow, and orange) — and not all orange stars, at that — is the range most likely to host habitable planets. (Collectively, this makes up about ten to fifteen percent of stars.) Since brightness is dependent on mass, stars less massive than this would require Earth-like planets to snuggle in so close to the star that they would be guaranteed to become tidally locked, with one side in permanent day, the other in permanent night, as well as having a habitable zone — the range of orbits that would permit liquid water — so narrow that probability of a planet forming just there would become infinitesimal. The low-mass stars also have a tendency to be UV Ceti variables — aka flare stars — which produce, at irregular intervals, intense, X-ray heavy "superflares" (actually about the size of flares emitted by the Sun, but on a star with less than half the mass) which would sterilize any planets close enough to be Earth-like.
Alternate biochemistries offer more hope, however. Red Dwarfs, thanks to their low temperatures (temperature is also dependent on mass for Main Sequence stars), emit very little UV light in comparison to the FGK "prime range", making it possible for them to host planets with liquid ammonia. (Water, methane, and ammonia, the raw materials for biochemistry, are all broken down by UV light, but ammonia is especially sensitive to it.) They are also equally as likely to host planets with liquid hydrocarbons and ice planets with subsurface liquid water as FGK stars. All three would occur on planets that orbit farther from the star than Earth-like planets, making superflares less of a problem.
Besides the type of star and some basic properties like dust and low variability, the number of stars in a system also plays a part. Some star systems form with multiple stars — up to six — orbiting a common center of mass. If two of the stars are extremely close — with orbital periods of ten days or less — they tend to make each other variable. Slightly further apart, and their mutual gravitation prevents planets from forming in the inner system. Slightly further still, and their mutual graviation prevents gas giants or Kuiper Belts from forming — leaving a system with barren, rocky planets lacking the thin layer of organic "paint" that Earth has. It is only solitary or very widely separated multi-star systems that could host life.
All things considered, as much as one-tenth of stars (within the GHZ) are suitable for hosting Earth-like planets. If we allow alternate biochemistries in our speculations, this could increase to as much as one-third.
For the purpose of my speculations on extraterrestrial life, I assume that carbon-water, carbon-hydrocarbon, and carbon-ammonia life are all possible. This produces four basic planet types that could host life. One is the carbon-water system we're all familiar with from life here on Earth, and assumes a rocky planet with a thin veneer of organic, hydrate volatiles. Another is an alternate carbon-water system living far below the surface of ice planets, which have rocky cores with a thick shell of ice. Liquid water could exist in a layer between the two if the rocky core is warm enough, and would basically be equivalent to magma on Earth. The other two are variations on this, with the ice planet having an atmosphere. If ultraviolet levels are moderately high, such as the FGK range of stars normally considered optimal for hosting Earth-like planets, such an ice planet would become a "super Titan", with ammonia broken down to become molecular nitrogen, and a thick haze of hydrocarbons formed as methane molecules were "cracked" by UV light, working like an ozone layer, and raining out if the temperatures were right. In a low-UV environment, such as M-type stars — Red Dwarfs — ammonia could remain intact and rain out if temperatures are right.
In our own Solar System, the ice planets that could host three of the four types of life-bearing planet do not exist, but equivalent types of ice moon exist for two of the three: the icy carbon-water type (Jupiter's moons Europa and Callisto, and Saturn's moon Titan, where combinations of radioactive decay and tidal heating from interactions with other moons heat their interiors) and the carbon-hydrocarbon type (Titan again, with liquid hydrocarbon lakes known to exist in the polar regions, and a thick nitrogen/methane atmosphere). Ice planets would likely be "failed cores" — bodies that formed in systems with low levels of dust, causing them to not get big enough, or to get big enough too late, to trigger gas giant formation. Bodies with roughly the mass of Earth would likely still be warm enough for a layer of liquid water between the core and crust for the icy carbon-water type, and possibly enough for that water — loaded with organic molecules in solution — to feed cryovolcanoes on the surface, releasing the ammonia and methane needed to sustain the carbon-hydrocarbon and carbon-ammonia types.
4.5 billion years ago, the Earth was a very different place. Freshly formed, the cooling, rocky planet lacked a significant atmosphere, having formed in a region of the circumsolar disk where temperatures were too high for anything except metals and rocks (silicate compounds) to solidify. 400 million years later, orbital migrations in the Outer Solar System resulted in Neptune shifting outward a lot, into its current orbit, plowing through the Kuiper Belt. This caused a period of chaotic orbital changes amongst the many, icy bodies there, including many migrating into the Inner Solar System. Entering this much warmer region caused their thick layers of hydrate ices — dominated by methane, ammonia, and water — to vaporize, forming "megacomets". Many were also on collision course with the four, rocky, inner planets.
Being closest to the Sun, Mercury didn't change much during this Late Heavy Bombardment, other than becoming pockmarked with thousands of craters. The difference in angular momentum between it and the megacomets was so high that essentially all of the hydrate ices were vaporized and blown into space when they collided, leaving behind rocky and metallic fragments in the resulting craters. Venus, Earth, and Mars all fared better, having low enough angular momentums (and, therefore low enough differences between those of the megacoments) to retain most of these icy volatiles.
Although retained, these volatiles evaporated, forming a reducing atmosphere of methane, ammonia, and water vapor. Ionizing radiation — where a single photon stores enough energy to knock an electron out of its orbit, which breaks molecules, as shared electrons are what holds them together — especially UV radiation "cracked" these molecules. These moderate-mass planets lacked the gratitational pulls needed to retain hydrogen, so hydrogen atoms freed from this photodissociation were lost into space, preventing the recreation of these molecules. As a result, the composition of the atmosphere evolved over time to molecular nitrogen (N₂), with nitrogen left over from ammonia, and carbon dioxide (CO₂), with carbon and oxygen left over from methane and water, respectively. This atmosphere was still very dense, with a surface pressure around 100 times the current amount, very similar to Venus.
Thanks to the global warming controversy, it's widely known that carbon dioxide is a greenhouse gas. Sunlight shining on the surface of a planet during the day warms it up, and the heat is partially released as infrared. What makes greenhouse gases special is that they absorb infrared radiation, slowing the release of excess heat. Methane, ammonia, and water vapor are all greenhouse gases, as well. This atmospheric conversion eliminates ammonia (and, therefore, its contribution to the greenhouse effect), reduces water (and its contribution), and replaces methane (which is over sixty times as potent as carbon dioxide) with carbon dioxide. This strong reduction in the greenhouse effect caused temperature to drop enough that the leftover water condensed and rained out to form oceans.
Within these oceans, a dizzying array of minerals and organic substances were dissolved and began reacting, setting off a complex process that eventually produced single-celled organisms. As evolution marched on, some of those organisms developed photosynthesis, then photosynthesis using water as an electron donor — which produces molecular oxygen (O₂) as a byproduct. Oxygen is a highly reactive substance, and quickly degrades organic material (if the organism hasn't evolved ways of dealing with it) so it was exhaled. Over a few hundred million years, the atmosphere was converted from an extremely thick, carbon dioxide one to a moderate, oxygen-rich one. At the same time, the amount of organic material increased, thanks to the carbon absorbed from the atmosphere, minerals dissolved in the oceans were scrubbed, by reacting with oxygen to form metal oxides that precipitated out, and life got a much more potent energy source, once oxygen respiration evolved.
Studies indicate Venus had a period of about 600 million to 2 billion years when conditions allowed liquid water to exist before the Sun's gradual brightening raised temperatures enough to trigger a runaway evaporation cycle. (Once all that water evaporated, photodissociation broke it into hydrogen, which escaped into space, and oxygen, which oxidized surface rocks.) It is possible that life could have bootstrapped on Venus, since conditions were nearly identical to Earth at the same time. There are hints that life may even still exist on the planet, although evidence is far from conclusive. Microbes have been found living in water droplets in clouds on Earth, and the thick, mostly sulfurous venereal cloud deck is known to contain water droplets in a region where pressure and temperate is similar to Earth today. Additionally, the clouds contain both hydrogen sulphide and sulfur dioxide — which readily react with each other, requiring some mechanism to continously produce them — carbonyl sulfide — difficult to produce with non-organic means — and a relative lack of carbon monoxide — which should exist in bulk as ultraviolet light and lightning "crack" carbon dioxide molecules, requiring some mechanism to consume it. While certainly not a guarantee of venereal life, life would neatly explain all these oddities.
Mars was also similar to Earth in the distant past, with a thick atmosphere dominated by carbon dioxide and liquid water on the surface. Unlike Earth, and Venus when it had liquid water, Mars lacks a tectonic cycle. This is due to the planet's low mass — eleven percent Earth's — which makes for less energy stored in the interior from the planet's formation to drive tectonic processes. On Earth, carbon dioxide reacts with silicate rocks to form carbonate rocks, removing it from the atmosphere until it is subducted when its tectonic plate collides with another. As it subducts, it heats to 300 °C, causing the reaction to reverse, and the freed carbon dioxide outgasses from volcanoes. On Mars, the lack of this process has resulted in most of the planet's original atmosphere being locked up as carbonate rock. As the atmosphere thinned, temperatures fell and the range of temperatures that water would remain liquid under shrank, eventually causing the planet to freeze over. As with Venus, there are oddities associated with Mars, particularly excess methane and ammonia and anomalous results of the tests for life conducted by the Viking probes in the 1970s, both of which could be explained by the presence of single-celled life. In the case of the anomalous test results, the explanation involves cells using a mixed water/hydrogen peroxide solvent, with the hydrogen peroxide functioning as an antifreeze. As with Venus, the evidence is not conclusive, and life is merely a hypothesis to explain the anomalies.
If life is found on Venus, living in the water droplets in the high clouds and using sulfur compounds as sunscreen or even in UV-driven photosynthesis, or on Mars, living in the soil and using hydrogen peroxide as antifreeze and releasing methane and ammonia into the atmosphere, it would revolutionize how stellar habitable zones — the distance from a star at which a planet could support life — are defined. Currently, the HZ ranges are defined to allow abundant liquid water to exist on an Earth twin's surface. If life, no matter how simple, is proven to be so tenacious as to have hung on in the etreme environments on Venus and Mars, the HZ will not just have to be extended, but redefined as a gradient, with a "core" section where complex life might be found, and two sections inferior and superior to that where simple life might be found.
The thick, carbon dioxide atmospheres that existed on all three, and probably exemplifies the primitive state for planets that support the warm carbon-water biosphere type, especially widens the habitable zone. The intense greenhouse effect obviously extends the HZ outward from the star, keeping planets warm enough for liquid water in wide orbits where it would freeze in an Earth-like atmosphere. The high surface pressure also raises the boiling point of water, for example from 100 °C/212 °F at Earth's sea level pressure to 302 °C/576 °F at 100 times that, pushing the HZ inward, as well.
Reconstructing ancient climates is a branch of science fraught with error, but there is some (controversial) evidence that Earth may have gone through a Snowball Earth period — total glaciation from pole to equator — several hundred million years ago. Liquid water (and life) would have been restricted to deep in the ocean, especially around volcanic vents. Volcanic vents are also one of the theoretical places where abiogenesis (the emergence of living organisms from non-living material) could have occured. Further, several of the ice moons in the Outer Solar System appear to have layers of liquid water between their rocky cores and thick, ice crusts. Depending on the mechanism and amount of heat, volcanic vents on the surfaces of the rocky cores could exist, possibly even hosting life.
Within our own Solar System, Europa, Ganymede, Callisto (all jovian moons), and Titan (a saturnian moon) all probably have liquid water mantles beneath their ice crusts, and Enceladus (another saturnian moon), Oberon (uranian moon), and Triton (neptunian moon) might have liquid water layers or pockets of liquid water internally, possibly with ammonia serving as an antifreeze. This (hypothetically) greatly increases the odds that life could be quite common in the universe, although it's doubtful that complex life could exist on these moons.
In some ways, ice planets or moons with subsurface oceans are more attractive hosts for life than rocky planets with thin layers of biological materal on their surfaces. The thick ice crust provides excellent protection qualities. The top ten centimeters would provide excellent shielding from radiation, both particle (from charged particles trapped in the intense magnetic fields of gas giants) and electromagnetic (X-rays from superflares or UV light normally), while visible light can still penetrate to a depth of a few meters, possibly allowing near-surface caverns to host photosynthetic life. Incliment weather would be a non-issue, and the danger of meteoric impacts would be greatly reduced. Water would be maintained in a liquid state due to internal (geologic) heat, rather than heating from a star, effectively eliminating the need for the notion of a habitable zone.
Raw material for life might be in relatively short supply, limiting the degree of complexity of both organisms and the biosphere as a whole. (The largest of the candidate moons in the Solar System is Ganymede, at just 2.5% Earth's mass.) Life on Earth might have hit such a limit early on, before photosynthesis added atmospheric carbon dioxide to the menu, but ice planets would lack such an atmosphere. For Sun-like stars, after all the hydrogen is consumed in the core, the core will collapse, and a hydrogen-fusing shell will form around it, causing the star to brighten, swell, and cool, becoming a Red Giant. During the Red Giant phase, the star could become bright enough to thaw ice moons, giving them global oceans rich in organic material.
In the inner system, where just metals and rocky silicate compounds — collectively just fifteen percent of the potential dust — can solidify, small to moderate-sized rocky planets form, lacking in the hydrate volatiles essential to life. In the outer system everything that can solidify does, greatly increasing the dust density and amount of planet-building material. This increases both the speed at which planets form, and the final size. If they form large enough, quickly enough, they can draw in hydrogen and helium gas (which makes up ninety-odd percent of available mass) directly from the proplyd, rapidly growing to giant-planet size. If there are multiple giant planets, they may jostle each other, pushing one out into the Kuiper Belt, which would direct many of the small, icy bodies left over in the outermost system inwards to deliver organic material to the rocky, inner planets.
Theoretically, in low-metallicity systems lacking the amounts of dust needed to seed gas giants, "failed core" ice planets may form, instead. These would be rich in the hydrate ices need by life — the textbook case being about seventy percent. If large enough, say an Earth mass or larger, they may have enough geologic energy stored up to maintain liquid water mantles with enough organic material to host even complex life. The low-metallicity stars that could host them are found in the outer regions of the galaxy, and old stars in the inner regions, and any long-lived star type, from yellow-white F-type to Red Dwarf, can host them. This may make these cold carbon-water planets several times as common as Earth-like, warm carbon-water planets. Their local planetary systems would consist of small, dead rocky planets in the innermost system, slightly larger, but still dead sooty planets in the mid-inner system, large, icy planets potentially hosting life in the mid-outer system, and an outer, Kuiper Belt of tiny iceballs.
If an ice planet/moon has enough internal heat, possibly from being especially massive, having unusually high levels of radioactive material, or from tidal interactions with another body, it won't just have a layer of subsurface liquid water. Volcanoes and thermal vents will allow the liquid water, with methane, ammonia, and various organic compounds in solution, to reach the surface. If temperatures are right, methane and possibly ammonia will vaporize, producing an atmosphere.
If the local star is warm, it will put out a fair amount of UV light, which will eliminate the ammonia by photodissociation, leaving behind molecular nitrogen. Depending on the final temperatures, the methane may condense, forming bodies of liquid methane, or it may remain gaseous. If it remains gaseous, it will be photodissociated itself, but much more slowly than ammonia, catalyzing the formation of more complex hydrocarbon chains. The main advantage that hydrocarbon solvents have over water is that hydrocarbons are a large class of chemicals, so just about any combination of surface pressure and temperature range would allow one to be liquid.
Collectively, liquid hydrocarbons have on big difference from water that puts their ability to serve as a solvent for biochemistry in doubt: they are non-polar. In water molecules, the oxygen end has a slight, negative charge, while the hydrogens have slight, positive charges. This polarity is key to the origin of life on Earth, as only polar substances can be dissolved by polar solvents, including the all-important amino acids. The non-polarity of most liquid hydrocarbons requires an alternative to amino acids to jumpstart life. No lesser authority than Isaac Asimov — famous for his science fiction, but also a holder of a doctorate in biochemistry — proposed poly-lipids as a non-polar alternate to amino acids.
The non-polarity of hydrocarbons also puts into doubt the ability for cells to spontaneously form. On Earth, glycerophospholipids, large molecules which have a hydrophilic (polar) "head" and a hydrophobic (non-polar) "tail", spontaneously organized themselves into bilayer membranes. These membrances consisted of a layer of lipids with the hydrophilic heads facing out, the other layer had the heads facing in, sequestering a microscopic pocket of organic material (including the self-replicating molecule DNA) dissolved in water. Lipids are not hydrocarbon-phobic, making them useless as the building blocks for cell membranes in hydrocarbon solvents, but non-polar, hydrocarbon-phobic molecules, such as acetonitril and hexane, exist, indicating that suitable alternatives could exist.
The enormous variety of hydrocarbons makes it impossible to present an exhaustive list of potential hydrocarbon biosolvents. The two most common seem to be methane (CH₄) and ethane (C₂H₆), which is the most common product of photodissociation of methane in the absence of oxygen. (The lack of oxygen being caused by all the water being frozen.)
Methane is liquid across a very narrow temperature range — just 21 K/°C (38 °F) at 1 atm of pressure (Earth sea-level pressure) — so it would likely require higher atmospheric pressures, in order to have a reasonably broad liquid temperature range. At 40 atm, the range widens to 100 K (180 °F), the same as water has on Earth, and carbon-methane worlds could tolerate pressures as high as 184 atm before methane's liquid range starts to overlap with water's — with a wopping 181 K (326 °F) liquid window.
Ethane is much better in this respect, as it already has a 95 K/°C (171 °F) liquid window at 1 atm. Raising pressure to just 1.25 atm is enough to widen the window to 100 K (180 °F), and carbon-ethane worlds could tolerate atmospheric pressures as high as 8 atm, with a liquid window of 183 K (329 °F).
Something especially interesting about these planets is that, given than the heat needed to drive the cryovolcanoes all but guarantees a liquid water mantle, they might host two separate biospheres: a subsurface, cold carbon-water one, and a surface, carbon-hydrocarbon one.
An ice planet or moon with sufficient internal heat to drive cryovolcanoes, and orbiting a cool star (with low-UV output), such as a Red Dwarf, could have conditions conducive to surface bodies of liquid ammonia (NH₃). Another potential environment for liquid methane are high-mass ice planets, those with sufficient gravitational pull to retain hydrogen, but not draw it in; the hydrogen would not be lost when ammonia is photodissociated by UV rays, but stay in the atmosphere, ready to react with other compounds, some of which produce ammonia to create a closed cycle. Unlike liquid hydrocarbons, ammonia is polar (like water), a big point in its favor in regards to being able to host nascent cells. As with liquid hydrocarbons, however, ammonia-based life could not use lipids as cell membranes, as they are not ammonia-phobic, but some hydrocarbons are ammonia-phobic, providing an alternate for cell membranes. Similarly, amino acids would not be useful biochemical building blocks in an ammonia solution, but amides would.
At 1 atm pressure, ammonia only has a liquid window of 45 K (81 °F). The optimal pressure for carbon-ammonia planets seems to be 4 atm, at which point ammonia has a 77 K (139 °F) liquid window, with a boiling point just shy of water's melting point — water ice being the crust of such a planet.