There are a number of proxies for calculating whether or not there was liquid water at the surface of the Earth: 1) fractionated felsic crust, i.e. mafic crust that has been altered in the presence of water and fractionated to produce more silica-rich (felsic) crust; 2) zircon crystals of Hadean age that have ?¹⁸O signature indicative of liquid water (i.e. formed in felsic melts); 3) sediments formed under aqueous conditions or specific mafic lava structures, such as pillow lavas, indicative of extrusion under water.

We will deal with felsic crust formation later because, owing to the absence of directly-preserved early crustal materials from the Hadean, we rely on inherited information from long-lived components, e.g., zircon crystals, that may have been formed either through fractionation of mafic mantle or during the production of aqueously-mediated felsic crust. These robust crystals generally have complicated histories often including overgrowths formed under different geological ages over the original crystal. Studies of the oxygen isotopic signatures preserved in Hadean-age zircon crystals (4.4–4.3 Ga, Wilde et al., 2001; Mojzsis et al., 2001; Valley et al., 2014) and combined oxygen and silicon isotope measurements (Trail et al., 2018) suggest that they formed in the presence of hydrothermal processing of the crust, implying the presence of water recycled into the crust from the surface of the Earth by 4.4 Ga. A number of more recent studies have questioned the age dating of the older zircons because of the fact that the crystals are long lived and exhibit multiple mantles of growth (Whitehouse et al., 2017), as well as the resetting of their U/Pb ages by later metamorphism or impact events (Griffin et al., 2014). What is important here is that the oldest, bona fide, oxygen isotope signature indicative of aqueous interaction with a fractionating felsic melt is revealed in a zircon aged 4.15 Ga (Muehlenbachs and Clayton, 1976; Whitehouse et al., 2017). This then is the baseline for liquid water at the surface, although, following the models described above, it is likely that liquid water at temperatures below ∼120°C existed previously.

Dating the oldest sediments known on Earth, i.e. detrital or chemical deposits formed in aqueous conditions, is highly controversial, as different dating techniques provide different ages. For example, the Nuvvuagittuq terrane in northern Quebec is purported to have a formation age of between 3.7 Ga according to U-Pb methods (Cates and Mojzsis, 2007) and 4.321 Ga according to the Sm-Ne method (O'Neill et al., 2012). This could be crucial because, if the terrane is Hadean in age, it holds the oldest known sediments. If not, it falls in the age range of the sediments from the 3.7–3.8 Ga Isua terrane in West Greenland (Appel et al., 1998), also containing metamorphosed sediments. In both cases, the sediments in question comprise banded iron formations, i.e. chemically (and possibly biologically) precipitated alternations of Fe-rich and Si-rich layers, as well as possible subaerial conglomerates. Volcanic sediments deposited under water are common in the younger, better preserved crustal remnants of the ∼3.5 Ga Barberton and Pilbara Greenstone Belts. Note that the Isua terrane contains the oldest pillow lava structures, i.e. evidence of direct lava extrusion under water. The rims of the pillow lavas are enriched in Fe₂O₃, MgO, MnO, K₂O, Rb, Ba, Ga, Y, and transition metals compared to the cores (higher concentrations of SiO₂, Na₂O, P₂O₅, Sr, Pb, U, Nb, and the light rare earth elements (REEs) than the rims, Polat et al., 2003). Given the recent successful experiments to convert ribonucleoside triphosphates to polyribonucleic acid when incubated with the glassy rims of Hadean-analogue basalts (Jerome et al., 2022), these transition metal-enriched, vitreous rock surfaces are certainly very interesting for prebiotic chemistry.

The oldest morphological signs of life are purported to be hosted at the 3.7–4.3 Ga Nuvvuagittuq terrane. Dodd et al. (2017) and Papineau et al. (2022) report hematite filaments and tubes in the jasper-carbonate banded iron formation. The hematite filaments are centimetre-size, pectinate-branching, parallel-aligned, undulated, and contain Fe²⁺-oxides. Morphologically, they resemble modern Fe oxidising filamentous microorganisms. Papineau et al. (2022) invoke metabolic pathways including Fe-oxidation and S-disproportionation, as well as anoxygenic photosynthesis. However, given the great age and metamorphic condition of the purported traces of life it may be difficult to definitively conclude their biogenicity, and as pointed out by Papineau et al. themselves, the structures could be simply abiotic in origin. Indeed, these structures have been alternatively described as chemical precipitates because of their hydrothermal affinities (McMahon, 2019). Other studies have also demonstarted their non-biogenicity (Greer et al., 2020; Lan et al., 2022), describing the structures as metamorphic phenomena. Nevertheless, the fact that the sediment was clearly a chemical deposit indicates precipitation in water.

On the other hand, the oldest potential chemical signs of life during the Hadean comes from analysis of the carbon isotope signature of graphite in a 4.1 Ga zircon crystal of—24‰¹³C, consistent with biological fractionation (Bell et al., 2015). Similarly, the sediments in the 3.8–3.7 Ga Isua terrane hold potential evidence of life, also in the form of carbon isotope signatures (Rosing, 1999) as well as, more importantly, remnant organic molecules with compositions and structures suggestive of a biogenic origin (Hassenkam et al., 2017). The latter in situ FTIR investigation of metastable carbonaceous inclusions in a garnet crystal within the metamorphosed sediments documents structural binding of nitrogen, oxygen and possibly phosphorous to organic molecules, strong signatures that the material derived from living organisms. On the other hand, purported stromatolites from the same Isua Greenstone Belt (Nutman et al., 2016) are clearly metamorphic artefacts (Allwood et al., 2018; Zawaski et al., 2020).

Definitive evidence of widespread and varied life forms is well preserved in the ∼3.5 Ga Pilbara and Barberton Greenstone Belts. Both chemotrophic (Westall et al., 2006a; Westall et al., 2011a) and photrophic life forms (Hofmann et al., 1999; Allwood et al., 2006; Noffke et al., 2013) have been interpreted from the fossil remains. In the case of the Pilbara and Barberton Greenstone Belts, their excellent preservation has conserved morphological remains of microorganisms, biofilms, and stromatolites, as well as various isotopic and organo-geochemical evidence. The widespread distribution and degree of evolution of early life by 3.5 Ga strongly suggests its appearance much earlier (for a review of evidence of life and its implications and evolution in Southern Africa, see Hickman-Lewis and Westall, 2021).

Thus, by 4.15 Ga there is evidence of water at the surface of the Earth, as documented by zircon crystals of that age formed in a mantle environment influenced by altered, supracrustal material of felsic origin (Cavoisie et al., 2005). Apparently, a mechanism for fractionating carbon that was similar to that used by life today was in operation by then or soon after (Bell et al., 2015). By Isua times (3.8–3.7 Ga) life was very likely present on Earth, and by about 3.5–3.45 Ga it was widespread.

An important question regarding some of the possible scenarios for the origins of life is the availability of exposed landmasses. Models suggest that there was up to 40% more water on the surface (Sim et al., 2016), which was eventually slowly removed through plate tectonic recycling. It had originally been thought that the abundance of Hadean zircons indicated widespread production of felsic crust (and exposed landmass) through modern-style plate tectonics. This is not the place for a detailed discussion on the timing and origin of plate tectonics (a recent review can be found in Westall et al., 2022), but we will provide a brief overview here.

During the Hadean, when life emerged, the planet was dominated by higher mantle temperatures (Franck, 1998; Schubert et al., 2001; O’Neill et al., 2007; Perchuk et al., 2020), possibly up to six times higher than present day mantle temperatures. Indeed, ultramafic lavas, such as Mg and Fe-rich komatiites, common during the Archaean, are testimony to higher early mantle temperatures (Figure 2) (Arndt et al., 2008). The tectonic regime is uncertain but the Hadean Earth was likely dominated by a stagnant lid with likely squishy lid/plume-lid and plate tectonics (Sizova et al., 2015; Rozel et al., 2017; Lourenco et al., 2018), as well as some proto continents. Structural and geochemical evidence indicates that plate tectonics initiated between 4.0 and 3.0 Ga (Griffin et al., 2014; Lammer et al., 2018; Dehant et al., 2019), with some contending that modern style plate tectonics did not occur until only about 1 Ga (Korenaga, 2018; Hawkesworth and Cawood, 2020). The process would have been gradual and likely took place in different locations around the globe at different times.

Was there emerged continental crust during the Hadean? As mentioned above, the apparent relative abundance of Hadean age zircon crystals has been interpreted to suggest the formation of a significant amount of fractionated (felsic) Hadean crust (cf. Arndt and Nisbet, 2012). However, on the basis of cathodoluminescence investigations of zircon crystals to determine their age and history, Whitehouse et al. (2017) interprets many of the Hadean-age crystals as being younger, and the majority of them formed in the mantle rather than the crust. These authors noted four pulses of zircon formation centred around 4.37, 4.15, 4.1, and 4.02 Ga, but whether they indicate pulses in felsic crustal formation, enclaves of fractionated crust in a mainly mafic protocrust, or melts formed by impacts or other processes, is uncertain. Based on modelling and geochemical analysis (U–Pb ages, and ¹⁷⁷Hf/¹⁷⁶Hf ratios) of the ancient zircons, as well as Re–Os model ages on sulphides and alloys in mantle-derived rocks and crystals, Griffin et al. (2014) also conclude that the Hadean/Eoarchaean Earth was highly volcanic with rocks of predominantly mafic composition. Recycling of the crust would have been provoked by impacts, with fractionation occurring due to massive melting associated with huge impacts, similar to the Sudbury impact. Indeed, they also noted episodic clumping of zircon ages recording possible peaks in felsic crustal formation at 4.5 Ga (?); 4.2–4.3 Ga; 3.8–3.9 Ga; ?3.3–3.4 Ga, possibly due to mantle overturns or major plume episodes, each followed by 150–300 Ma of quiescence.

Hadean protocontinents were not like modern continents. Modern continents are characterised by highly felsic cores, high elevations, and a thick, stabilising underplating “keel”. They are difficult to subduct. This was not the case during the Hadean. Since only highly metamorphosed remnants of Hadean protocrust survive, we will base our description on well-preserved crustal remnants dating from the Palaeoarchaean, the ∼3.5 Ga Barberton and Pilbara Greenstone Belts, that can be used, to a certain extent, as proxies of older crust. However, these ancient terranes record only remnants of the upper parts of protocontinents but no “deep sea crust” per se. They document abundant mafic and ultramafic effusions (massive and pillow basalts) and intrusions. Most of the volcanic lithologies are tholeiitic, calk-alkaline or komatiite-type basalts. The latter particularly are characterised by higher Mg contents than present day volcanics (by several orders of magnitude) (Table 2), and were formed from hotter melts (Arndt et al., 2008). The ultra/mafic volcanics may also be accompanied by more fractionated lavas or intrusions, including andesites, dacites and rhyolites. Table 3 documents the characteristics of Archaean basalts arand Arndt et al., 2008).

The Eoarchaean-Palaeoarchean continental crust was characterised by thick layers of mainly ultramafic to mafic lavas and intrusions and sediments derived from these materials, interspersed with rarer, more fractionated, felsic lavas and intrusions. Formation of the early granites (tonalite-trondjemite-granitoids, TTGs, less rich in Si and K compared to modern granites) intruding into this crust appears to have been related to lower crust melting rather than to plate tectonic subduction (Smithies et al., 2007). Indeed, structural and geochemical evidence of granites formed through plate tectonic subduction is not found in the Pilbara until after 3.2 Ga (Smithies et al., 2007).

Vanderhaeghe et al. (2019) suggest that the hotter Archaean crust was at least 40 km thick. Despite this, there is no evidence for widespread erosion of landmasses and exposure of the granitic cores of the early protocontinents, these were subdued landscapes with much of the continental crust submerged (Arndt and Nisbet, 2012). This means that the delivery of detrital material eroded from continents was lower than it is today. Thus, the early sediments comprised volcaniclastic detritus, either very locally derived from locally eroded rocks (the particles are generally euhedral to subhedral in shape, indicating lack of long transport by water, i.e. rivers), or ashfall into water bodies. These sedimentary materials were deposited into mainly shallow water basins and, although there is evidence of rare subaerial/deltaic fans, such evidence is rarely preserved, and one of the few examples is the 3.45 Ga Hoogenoeg Formation in Barberton (Lowe, 1999). Possible subaerial sediments have also been identified on the 3.75 Ga Isua and 3.8 Ga Nuvviagittuq terranes (Fedo, 2000; Bolhar et al., 2005; Cates and Mojzis, 2007; O’Neill et al., 2011).

In a sedimentological-structural investigation of the basinal structures in the Pilbara Craton, Nijman and De Vries (2004) and Nijman et al. (2017) describe these features as collapse basins (Figure 3), similar to the coronae structures on Venus or early Mars (although not all researchers agree with this interpretation). These collapse basins are roughly circular in shape and of the order of 50 to several hundreds of kilometres in diameter. These basins were generally not very deep, since most sedimentary horizons were deposited in shallow settings (offshore to onshore), although some sedimentary sequences were deposited below wave depth (less than about 100 m). The basins formed on top of mantle plumes in the weaker (because hotter) Palaeoarchaean crust. Figure 3 illustrates collapse basins in the Pilbara Craton where the continental crust is mostly submerged, and only low relief volcanic edifices and surrounding areas emerged from the sea. In the example from the Pilbara Greenstone Belt shown in this figure, the emergent land masses are felsic volcanoes.

As representative examples of the mineralogical composition of Palaeoarchaean mafic lavas, tholeite basalts from the Kromberg Formation in the Barberton Greenstone Belt, South Africa, are comprised mainly of plagioclase (Na-Ca feldspar) and clinopyroxene (a Ca,Mg,Fe, Ti alumina silicate) with minor olivine and chromite (Vennemann and Smith, 1999) (Tables 2, 3). Titano-magnetites occur in the more evolved tholeites, while cumulates comprise only pyroxene and olivine. Komatiitic lavas (even more mafic than tholeites) are composed primarily of olivine and chromite (chrome spinels). They are characterised by anomalously high Mg (and Fe) contents owing to their very high temperatures of formation and eruption, up to >300°C hotter than basalts (Arndt et al., 2008). (Note that the original mineralogy of these Palaeaoarchaean basalts has been affected by seafloor alteration and subsequent metamorphism to actinolite-tremolite, chlorite, albite quartz, and rare, relict pyroxene.) Other Palaeoarchean basalt types include calk-alkaline basalts (enriched in magnesium and calcium oxides) comprise plagioclase, clinopyroxene and various metal oxides, such as magnetite.

In the Pilbara Greenstone Belt, there is much felsic volcanic material interlayered with the ultramafic and mafic lavas (Smithies et al., 2007). Here, more evolved lavas include andesite to dacitic rocks with sodic ratios of K₂O/Na₂O between .05 and .45 (van Kranendonk et al., 2007), as well as more enriched felsic rocks (low K₂O < 1.0 wt%, high Fe, HREE and Y concentratinos positively correlated with SiO₂ and La/Yb ratios) derived from fractionated tholeiitic lavas. Other felsic rocks have higher sodic ratios and have been derived from tonalite-trondgehemite-granodiorite (TTG) granites, while some highly potassic rhyolites may have been formed by remelting of pre-existing continental crust (van Kranendonk et al., 2007).

Examples of the compositions of andesitic (fine-grained lavas with phenocrysts of plagioclase and the ferromagnesian minerals, pyroxene and amphibole) to dacitic lavas from the Duffer Formation of the Pilbara Greenstone Belt (fractionated from basalts) are shown in Table 4. They include plagioclase porphyry (i.e. large plagioclase crystals (phenocrysts) in a fine-grained groundmass. Even more fractionated, porphyritic rhyolitic lavas occur in the Palaeoarchaean terranes, comprising more silica rich minerals, such as quartz and feldspars.

Typical of continental crust are granitic intrusions formed through fractionation of hydrated ultramafic to mafic crust, and sometimes remelting of previous granitic crust. In fact, the most common early crustal remnants are comprised predominantly of the early granitic cores of the protocontinents. However, as with the ultramafic lavas, the Hadean/Palaeoarchaean granites (TTGs) differed from their modern counterparts in their compositions, being less rich in potassium feldspathic minerals than modern granites (Moyen and Martin, 2013).

While the early detrital sediments were mainly volcaniclastic in origin (including volcanic glass, protoliths, spherules, pseudomorphed feldspars and pyroxenes, as well as accessory minerals, such as chromites, zircon, rutile, quartz, carbonates, Fe oxides, and barite, see Lowe and Byerly, 2003), many of the sediments comprise varying amounts of carbonate minerals that formed diagenetically. All these lithologies have been significantly altered during early diagenesis to mostly Fe carbonate, sericite, chlorite, muscovite, and biotite followed by more or less complete replacement by silica, now forming the microcrystalline quartzitic rock, chert. With respect to the carbonates, given the diagenetic and metamorphic overprint, it is difficult to determine their original mineralogy, but Lowe and Knauth (1977) suggest local primary Fe-rich dolomite and ankerite for the Kromberg Formation in Barberton, now replaced by dolomite, siderite, ankerite and calcite. The Strelley Pool carbonates were Fe-dolomites (Shields, 2007).

Carbonates have also been described from the Isua Greenstone Belt in Greenland, but the calcites, Fe-dolomites and siderites are generally interpreted to be metasomatic deposits, although small amounts of carbonates associated with banded iron formations and metacherts may be original (van Zuilen et al., 2003).

Mineral surfaces, such as clays, basaltic glasses, pyrite, or other reactive materials, are an important substrate for prebiotic reactions, as demonstrated by the recent conversion of ribonucleoside triphosphates to polyribonucleic acid when incubated with basaltic glass (Jerome et al., 2022). In the hot, generally acidic waters of the Hadean oceans, inorganic surfaces of volcanic materials or meteorites, would have been rapidly altered. For example, in the preserved Palaeoarchaean rocks, most of the volcanic detrital particles were altered to phyllosilicate (probably some kind of smectite, now mostly sericite, chlorite or muscovite after metamorphism) and to anatase before being silicified (Lowe, 1999; Foucher et al., 2010; Westall et al., 2015b). Moreover, an experiment to alter basalt and obsidian surfaces in artificial Hadean seawater at 73°C showed that the surfaces were rapidly coated with a mixture of clay particles and organic matter, the former apparently forming in association with the organic film (Figure 4) (unpublished data, Westall). In this experiment, the mineral surfaces were coated by a “conditioning film” of organic molecules as soon as the material was exposed to seawater. This is a known, automatic chemical reaction caused by hydrostatic interaction between the hydrophobic organic macromolecules and the hydrophilic surface of materials immersed into water, and it takes place within minutes (Costeron et al., 1995; Westall et al., 2000). All natural water contains dissolved organic matter (DOM), all biologically derived, in the reduced form. Indeed, Catala et al. (2021) note that DOM in marine waters has one of the most diverse molecular compositions known, consisting of millions of individual, low mass compounds, including compounds that are alicyclic, organic acids with amphiphilic properties (Hertkorn et al., 2006; Dittmar et al., 2008; Zark et al., 2017). While today much of the DOM is rapid recycled by life, some molecules have life times of >1,000 years. The latter situation would have been the case before the advent of cellular life in the prebiotic Hadean oceans. Today, it is estimated that there is more than 1 Eg (exagram, 1,018) of DOM on the planet (662 ± 32 Pg (petagram, 1,015) carbon; Hansell et al., 2009). DOM in the Hadean oceans would have been sourced from the mantle, hydrothermal crustal processes, and particularly from extraterrestrial meteorites and micrometeorites (Maurette, 2006). DOM from hydrothermal fluids would have been more common in the Hadean than today, given the higher crustal temperatures and more vigorous hydrothermal recycling. DOM in such fluids has been thermally altered (Rossel et al., 2017) and comprises more aromatic molecules and with less carboxyl-rich alicyclic species.

The close association of clay minerals with organic matter for prebiotic reactions has long been investigated (Cairns-Smith, 1966; Hartman, 1975; Ferris et al., 1979; Ferris et al., 1996). Interactions between the inorganic and organic phases are facilitated by adsorption, intercalation and cation exchange. Furthermore, negatively charged organic ions bind to clay minerals by positive edge charges or by the exchange of structural OH groups (Lagaly, 1984). Kloprogge and Hartman (2022) review present understanding with respect to the role of clay minerals, in particular mixed layer, Fe-clays, such as smectites, in the origin and development of metabolism, noting 1) the formation of amino acids on the surface of clay minerals on carbonaceous chondrites from simpler molecules, e.g., CO₂, NH₃, and HCN; and 2) the catalytic role of small organic molecules, such as dicarboxylic acids and amino acids found on carbonaceous chondrites, in the formation of Fe-clays themselves. Importantly, the evolution of metabolism can be replicated and catalysed by clays that can synthesise monomers, such as amino acids and nucleotides, that will subsequently polymerise, an important reaction for RNA-peptide worlds (op.cit.).

One important component of these early sediments is chemically-precipitated silica (Dass et al., 2018; Westall et al., 2018; Ledevin et al., 2019). Hadean-Palaeoarchaean ocean waters were supersaturated in silica owing to global, widespread hydrothermal silica input (Hofmann and Harris, 2008), surficial weathering, and devitrification of the volcanic protoliths in water. Since chemical silica was deposited in rhythmic layers interspersed with other sediments, Ledevin et al. (2019) suggest that this was controlled by seasons, with silica precipitation occurring during winter months. The fact is that we have no idea what the yearly climate was like during the Palaeoarchaean era. Nevertheless, the rhythmic alternation of chemical silica layers with other sediment does indicate some kind of cyclical control, similar perhaps to that functioning for the banded iron formations.

One of the consequences of the early silica-saturated seawater was the pervasive silicification of all lithologies in contact with water, both sediments and volcanic rock surfaces. Silicification was extremely rapid in the vicinity of hydothermal activity in these facies (Westall et al., 2015b; Westall et al., 2018), and slower away from the direct influence of hydrothermal vents, where there is evidence of more advanced diagenesis. The degree of silicification and its rapidity has important consequences for prebiotic chemistry, as discussed below.

White smoker-type hydrothermal vents, with their associated carbonate deposits, are not recorded in the crustal record of Eo-Palaeaoarchaean cratons. They were either rare, did not exist, or were not preserved. They may have existed in the Hadean deep oceans. What do exist in the geological record are silica-rich vents that expulsed silica-rich fluids and, in the process, thoroughly silicified the surrounding lithologies (Hofmann and Harris, 2008; Westall et al., 2015a), forming pods of sediment around the conduits that are now >99.9% SiO₂ (Figure 5). Silicification implies permeation of all matter, particulate, solid rock or biogenic, by silica and replacement of the pre-existing feature by silica. Replacement may be more or less complete, leading to ghost structures; or partial, depending upon the degree of saturation of the seawater by silica. Other factors, such as temperature and pH, will have an effect on the rate of silicification.

One way to estimate the rate of silicification is the degree of diagenesis of the non-organic detrital and biogenic facies. Close to hydrothermal effluents, biogenic remains were so rapidly silicified that there was no time for diagenetic alteration that would lead to the precipitation of framboidal pyrite (sometimes associated with biogenic activity, such as degradation of pre-existing organic matter, e.g., dead cellular materials, by heterotrophic microbes, such as Sulphur reducing bacteria (SRBs), e.g., Duverger et al., 2020; Duverger et al., 2021). At further distance from hydrothermal activity, framboidal pyrite associated with degraded organic matter was common (F. Westall, personal observations). In an experiment to document the formation of pyrite in a pure microbial culture of sulphate reducing bacteria, Duverger et al. (2020) noted that spherules of pyrite a few hundreds of nm in size precipitated after 1 month of incubation. Indeed, pyrite framboids have not been described in association with well-preserved, cellular fossils from the Palaeaoarchaean sediments, confirming their rapid silicification (Westall et al., 2001; Westall et al., 2006a; Westall et al., 2006b; Westall et al., 2011a; Westall et al., 2011b). Superb cellular preservation by silicification is another means of estimating the time scales of silicification. The Palaeoarchaean cells mentioned above document preservation of intact cells undergoing cell division, as well as cells and colonies in varying stages of degradation.

Experimental silicification of the kinds of thermophilic Archaea that could have existed on the early Earth, at temperatures of 60°C documents initiation of fossilisation already within 24 h (Orange et al., 2009). Note that the Palaeoarchaean shallow seas were very warm, ∼50°C–76°C, water temperatures being estimated from oxygen and silicon isotopes, e.g., van den Boorn et al. (2007), Tartèse et al. (2017), and Lowe et al. (2014) In these cases, the organic substrates functioned as substrates for the chelation of silica, either directl.y onto the organic matter, or through cation bridging, e.g., with Fe. This was rapidly followed by direct chemical precipitation of silica in an acidic environment that formed a matrix cement around the organic substrate. Initially the silica would have presented a gel-like consistency that permitted permeation of fluids (Iler, 1979) and prebiotic reactions. As it polymerised and dehydrated, this permeable capacity would be lost. The rate of polymerisation of silica into silica gel depends on the saturation of silica, pH and temperature, as well as the presence of dissolved salts (Iler, 1979). We can conclude that, close to hydrothermal activity in the salty (Knauth, 2005, estimates early salinity to be 1.5–2.0 times present level because of the relative paucity of continents and enclosed water bodies in which halite could precipitate, consistent with estimates for salinity also on early Mars in Fairén et al., 2009), warm to hot environment of the early Earth, initial precipitation of silica was on the order of hours to days, while further polymerisation and dehydration leading to the cessation of prebiotic chemical reactions taking place within the gel, would have occurred on longer time scales. Further away from direct hydrothermal influence, the rates of precipitation, polymerisation and dehydration would have been slower (time scales of one to a few months, cf. Duverger et al. (2020)), as documented by the more prolonged diagenetic alteration of the organic matter leading to framboidal pyrite precipitation.

The silica formed initially by polymerisation of monomeric silica in solution, nucleating to form ever larger particles depending upon the availability of the silica. The supersaturated Palaeoarchaean seawater thus precipitated silica, both within the sediments as well as on top of them forming layers of chemically precipitated silica gel (cf. Ledevin et al., 2019). Evidence of this chemically precipitated gel formation is seen in where colonies of microbes grew in three-dimensional spicular colonies within a mixture of fine volcanic dust and silica gel (Westall et al., 2015a; Hickman-Lewis et al., 2020a). Again, in these cases, the rarity of pyrite associated with the degraded, silicified organic matter is testimony to the rapidity of silicification.

The fact that silica gel was pervasive in the Palaeoarchaean oceans has important consequences for prebiotic reactions. Silica gel is a porous medium. Incorporating reactive volcanic detritus (including primary minerals, such as (rare) olivine, pyroxene, plagioclase feldspar, oxides e.g., chromite, and pyrite, and secondary minerals (alteration phases), such as clays, carbonates, anatase), the silica gel could act as myriads of microscopic reactors, especially when influenced by hydrothermal effluent (Dass et al., 2018; Westall et al., 2018). Here, gradients in pH, ionic concentrations and temperature from the vents through the immediately surrounding sediments would have been important factors for prebiotic reactions taking place in these sediments. Indeed, Pollack (2001), Trevors and Pollack (2005) and Dass et al. (2018) note that gels represent an excellent mechanism to maintain concentration gradients, alter the structure of water, and influence ion-macromolecule interactions. Gradients of protons and cations are conferred because the ability of gels to exclude solutes maintains the overall system out of equilibrium by creating an imbalance of solutes inside and outside the gel.

The rate of silica gel formation and its subsequent dehydration and lithification would have been an additional control on rates of prebiotic reactions. For example, using an artificially prepared solution of silica (sodium silicate) in artificial Hadean seawater, Dass et al. (2018); Dass et al. (2018) showed that the gel can form within hours upon addition of glacial acetic acid (pH 2.5). Of course, in the warmer Hadean waters, where pH would have been variable at the surfaces of the volcanic rocks and sediments (initially a relatively high pH 8, then decreasing after a couple of days to pH 6, Westall et al., 2018), the rate of silica gel formation would also have been influenced by the saturation of silica in the ocean waters, which, as we noted above, appears to have been controlled also by volcanism and direct hydrothermal input. Rapid silica gel formation would have occurred during the Hadean-Palaeoarchaean over time spans of hours to possibly days. With pore sizes of the order of 100 s nm to a couple of micrometres, these pervasive gels were a critical physical and chemical component of the early rock/sediment interfaces.

While the silica gel was in a porous state that allowed the transport of fluids, small molecules and dissolved components, prebiotic reactions could take place within the precipitate. However, dehydration and solidification of the gel that would put an end to these kind of reactions would depend on factors, such as temperature and pH. For example, Gallo and Klein (1988) note that silica gel dehydration is constrained by the diffusion of hydrogen-bonded species at low temperatures (<600°C), as would have been the situation in the Hadean oceans. Gels formed under basic conditions are structurally different to those formed under acidic ones (Orcel et al., 1986), the former being characterised by a macroporous structure with surface silanols and the latter by a microporous one with internal silanols. Acid-formed gels have a smaller scale silica network, larger surface area, smaller pores and larger pore interconnection. However, dehydration of silica gels formed under basic conditions is more rapid than those formed under acidic conditions. Sleep et al. (2001) and Sleep (2010) consider that the Hadean oceans were probably as neutral as they are today because of buffering of the seawater chemistry due to reactions with basalt. However (Ueda et al., 2021), suggest that fluids close to hydrothermal vents would have been probably more acidic at low temperatures owing to high water/rock ratios during the hydrothermal reactions (although pH would have been more alkaline at higher temperatures). Also, the predominantly CO2 atmosphere would have resulted in slightly acidic waters, as equally proposed for early Mars (Fairén et al., 2004). As noted above, while more alkaline alteration of the surfaces of Palaeoarchaean subaqueous basalts occurred at high temperatures (e.g., Robins et al., 2010), the pervasive silicification of the sedimentary horizons and the tops of the basaltic sequences is testimony to a lower temperature, silica-rich, acidic medium.

In the Barberton craton, the situation is similar (Nijman and de Vries, 2004). As noted above, most of the sedimentary detritus derived from exposed volcanic lavas and volcanic ashfall. Evidence for riverine input into the shallow basins comes from geochemical data. Hickman-Lewis et al. (2020b) made in situ LA ICPMS measurements of microbial mat layers (i.e. relatively short-lived, ephemeral phenomena) that document flat, light REE-enriched REE + Y patterns and chondritic Y/Ho ratios indicating major contributions from terrigenous, riverine fluids, i.e. continental weathering, on a short temporal scale. This signature is superimposed on a long-term, background signature of predominantly hydrothermal origin (Figure 6).

Numerous different types of locations have been suggested for the emergence of life (see review in Westall et al., 2018). The most commonly considered are hydrothermal environments but locations including associated with impact craters (Osinski et al., 2020; Sasselov et al., 2020), pumice rafts (Brasier et al., 2011; Brasier et al., 2013), deep seated fault zones (Schreiber et al., 2012), and radioactive placer sands (Adam et al., 2018) have also been invoked. For many of these environments, solar irradiance at the surface of the Hadean Earth, both on exposed landmasses and in water bodies, could have played an important role in the emergence of life. Pascal et al. (2013) hypothesise on the possible importance that UV radiation may have had as a free energy requirement for the origin of life (Figure 7), although not all prebiotic chemists adhere to this theory. Modelling suggests that solar irradiation at the surface of the early earth was much higher than today.

Solar X-ray and UV radiation covers wavelengths from .1 to 320 nm. The luminosity of the Hadean/Palaeoarchaean Sun was about 74%–77% of its present value (Bahcall et al., 2001) but the solar X-ray and UV (EUV and FUV) luminosity reaching the top of the early Earth’s atmosphere was higher (Micela et al., 1985; Micela et al., 1988; Micela, 2002; Cnossen et al., 2007). Radiation at the surface of the early Earth would have been controlled by the composition of the atmosphere. Cnossen et al. (2007) modelled the spectral radiance of the young Sun on the basis of evolution of solar-like stars, taking into account factors, such as atmospheric composition, solar flares and activity cycles, as well as aerosol concentrations and cloud cover. They showed that shorter wavelength radiation (<200 nm) would have been attenuated by the atmosphere, while the flux of longer wavelengths (>200 nm) reaching the surface of the Earth would have been ∼10^–5 Wm⁻² nm⁻¹ at 200 nm, to ∼.4 W m^-2 nm⁻¹ at 320 nm for small concentrations of CO₂ in the atmosphere (.024 bar). Higher CO₂ atmospheric concentrations (e.g., 1.20 bar) would have significantly attenuated the surface flux of radiation (∼10–16 W m^-2 nm⁻¹ at 200 nm, to .02 W m^-2 nm⁻¹ at 320 nm. Cnossen et al. (2007) calculated the irradiance at the top of the atmosphere and the surface irradiance (180–320 nm because the shorter wavelengths do not reach the surface) for the early Earth (Archaean in this modelled case, i.e. 4.0–2.45 Ga) at different atmospheric CO₂ concentrations, compared with the DNA weighted spectra for different CO₂ concentrations (Cnossen et al., 2007). also demonstrated that small amounts of other gases in the atmosphere, such as N₂, CO₂, CH₄, H₂O, O₃, O₂, and SO₂ would have little effect on the attenuation of solar irradiation. They conclude that the amount of radiation reaching the early Earth’s surface would have been damaging for the DNA of early life. From the point of view of Pascal et al. (2013), this high radiation is essential for providing the necessary initializing energy for prebiotic reactions in order for the “multiplying entities (to be) associated with the dissipation of free energy” (Figure 7).

Solar radiation is, however, generally rapidly attenuated by water, reaching at maximum depths of a few metres, although note that blue-green light (UVA at 360 nm) can penetrate up to depths of 50–70 m in oceanic waters when the Sun is at its zenith (lee et al., 2013). Attenuation of radiation at all wavelengths will be increased depending upon water opacity. On the volcanically-active early Earth, water bodies would have carried much suspended and sedimenting particulate matter (volcanic debris, e.g., ashfall and detrital sediments) during eruptive phases, which would have significantly decreased solar radiation penetration depths. The fluvial contribution to the solid particulate load being deposited in shallow water basins adjacent to the exposed volcanoes would also have been higher during and after eruptions owing to the increased topography and erosion of the volcanic edifices. However, the intervening quiescent periods would have lower detrital input and hence suspended sediment load.

Recently, subaerial springs have been proposed as suitable locations for the emergence of life (van Kranendonk et al., 2018; Damer and Deamer, 2020; van Kranendonk et al., 2021). Today, subaerial springs are common in volcanic areas associated plate margin and intraplate, long-lived (10 s My) plume activity. During the Hadean, plumes on the hotter early Earth may have existed for longer, up to several 100 s Ma, judging by the ages of the Palaeoarchaean greenstone terranes, while individual volcanic eruptions and associated hydrothermal activity would have been more short-lived and episodic. For example, hot spring activity at Yellowstone over the North American mantle plume has been ongoing since at least the end of the last ice age (∼15,000y) with the construction of large geyser cones taking place over thousands of years (USGS, 2022). A modern analogue of Hadean/Palaeoarchaean protocontinents with associated volcanic and hydrothermal activity is the Kerguelen plateau with its volcanic island archipelago. Characterised by mainly basaltic lavas with some felsic components intruded by plutonic rocks, the volcanic rocks date back to 39 Ma (Gagnevin et al., 2003), with individual lava complexes have a longevity of a couple of million years.

Palaeoarchaean suberial springs have been described from the Pilbara Greenstone Belt, while much of the hydrothermal activity in the 3.48 Ga Dresser Formation occurred under shallow water submarine conditions. In one case, some of the venting was associated with relatively restricted (<10 m) geyserite and siliceous sinter deposits in the North Pole Chert, typical of a subaerial hot spring field (Djokic et al., 2017; Van Kranendonk et al., 2018; Djokic et al., 2021) (Figure 8). Van Kranendonk et al., 2021 and references therein) describe a volcanic caldera setting characterised by voluminous, contemporaneous hydrothermal fluid circulation that formed a dense network of hydrothermal chert and baryte veins cutting through the underlying komatiitic pillow basalts and lowermost layers of the North Pole Chert. Van Kranendonk et al. (op.cit.) conclude that the sedimentary succession records “a transition upward from open marine, through a shallow, evaporative marine basin (caldera phase), to subaerial conditions with hot springs and fluvio-lacustrine deposition, followed by a return to deep marine conditions that resulted from caldera collapse with/without active extension (Djokic et al., 2017; Van Kranendonk et al., 2018; Djokic et al., 2021)”.

The hydrothermal veins associated with the North Pole Chert succession are large, up to 50 m wide and also deep, penetrating down to depths of several kilometres. The tops of the veins are associated with the hot spring deposits and mineralised remnants of the hot spring pools (Van Kranendonk et al., 2007; Djokic et al., 2017; Van Kranendonk et al., 2018; Djokic et al., 2021). Fluid temperatures in the veins are estimated to have been about 350°C at depth and about 120°C at the palaeosurface (Harris et al., 2009). Interestingly, tourmaline crystals (containing boron) are associated with some of the hydrothermal deposits, and boron is of interest in prebiotic processes (cf. Ricardo et al., 2004), as noted above.

Although the individual Dresser springs were of limited dimension (Kranendonk et al., 2021), it is known that subaerial springs may occur in swarms of many hundreds or even thousands of individual vents, each with different physico-chemical parameters in terms of temperature, fluid chemistry (pH, ionic concentration, element composition), gas content, style of upwelling (passive flow, to bubbling, to geysers), that can vary widely over lateral scales of only a few meters (Campbell et al., 2015). These kinds of environments provide a wide range of possibilities for concatenating prebiotic reactions over periods of up to a couple of million years, according to the Kerguelen analogy.In an analysis of the chemical energy available at one subaerial spring (Obsidian Pool, Yellowstone), Shock et al. (1995) determined that the energy yields for different redox reactions involving reduced hydrothermal fluids depend largely on the electron acceptors. In this modern case, the electron acceptors include, O₂, nitrate, nitrite, elemental S, magnetite, hematite, goethite, sulfate, CO, and bicarbonate/CO₂ in order of the highest to lowest energy yields. On the Hadean Earth, O₂ would only have been available in very small quantities, formed by abiotic radiolysis of H₂O. Likewise, nitrate and sulfate were probably not present, and nitrite limited.

Scenarios for the emergence of life in the submarine realm are mainly related to hydrothermal activity (Baross and Hofman, 1985; Russell and Hall, 1997; Martin et al., 2008). The submarine realm on the Hadean/Palaeoarchaean Earth comprised a wide range of environments with differing physico-chemical characteristics, and hosted above differing types of igneous substrates. These include the shallow water, volcano flanking environments on top of the oceanic plateaux that were the protocontinents (i.e. at littoral to subwave base water depths), as well as environments in the deep ocean associated with plume-related hot spots and, if they existed in the Hadean when life emerged, early tectonic spreading ridges. Redox reactions, such as those described at the beginning of Chapter 4 involving dissolved gases (e.g., CO₂, and H₂) (Shock, 1993; Shock and Schulte, 1998; McDermott et al., 2015) or metal sulphides as catalysts and H2S as a reductant (He et al., 2019) to form reduced carbon compounds (e.g., formate and acetate in the latter case).

It is generally assumed that primordial life was based on organic molecules containing carbon and hydrogen atoms associated with oxygen, nitrogen, and sulfur, the basis of current biology. However, some scientists imagined an alternative scenario based on inorganic substrates. Schneider (1977), for example, suggested that the complex dislocation networks found in some crystals could follow the definition of living units, proposing the idea of a crystalline physiology. According to Cairns-Smith (1982), there is no compelling reason to relate the first living cell to a last common ancestor made of organic molecules. He proposed that the first living systems, and the chemical evolution preceding them, could have been based on clay chemistry, different from life as we know it today. Although each step of this hypothetical sequence of events was developed in detail, the scenario has not been subsequently supported by experimental facts. Weiss (1981) published data that appeared to experimentally support Cairns-Smith’s scenario but, unfortunately, the results could not be replicated. Since no satisfactory answers could be obtained, the clay-mediated replication hypothesis will not be further considered here.

In his historical experiment, Miller (1953) exposed a mixture of methane, ammonia, hydrogen, and water to spark discharges and silent electric discharge. In his initial experiment, he obtained three amino acids (glycine, alanine, and β-alanine) in the hot water within his experimental flask via the intermediary formation of hydrogen cyanide and aldehydes. The synthesis of amino acids occurred efficiently when using a reducing gas mixture containing significant amounts of hydrogen. However, the dominant view now is that the primitive atmosphere consisted mainly of CO₂, N₂, and H₂O, along with small amounts of CO and H₂ (Catling and Kasting, 2007). Only small yields of amino acids are formed in such a mixture (Schlesinger and Miller, 1983), though more recent studies show that the low yields previously reported could be the outcome of oxidation of the organic compounds during hydrolytic reprocessing by nitrite and nitrate produced in the reactions. The yield of amino acids is greatly increased when oxidation inhibitors, such as ferrous iron, are added prior to hydrolysis, suggesting that synthesis from neutral primitive atmospheres may have been more important than previously thought (Cleaves et al., 2008). Re-analyses of original sample residues from a previously unreported Stanley Miller electric discharge experiment, using a gaseous mixture of H₂S, CH₄, NH₃, and CO₂, were found to contain a total of 23 amino acids and 4 amines, including 7 organosulfur compounds (Parker et al., 2011). Miller also sparked a gas mixture of CH₄, NH₃ and H₂O while intermittently adding the condensing reagent cyanamide, and modern reanalysis of original sample residues of this experiment found a dozen amino acids, ten glycine-containing dipeptides, and three glycine-containing diketopiperazines (Parker et al., 2014).

The principal driver of organic synthesis is the disequilibrium between very different fluids. Thus, the reducing conditions in hydrothermal systems could have been also an important source of biomolecules on the primitive Earth (Holm and Andersson, 2005; Konn et al., 2015; Sojo et al., 2016; Damer and Deamer, 2020; Rimmer and Shorttle, 2019; White et al., 2020). The reducing environment results from substances dissolved in seawater interacting with inorganic compounds present in very hot crustal material (ultramafic rocks) that reduce them. These reduced compounds flow out of the hydrothermal system and the inorganic sulfides formed precipitate when they mix with cooler ocean water (4°C today but much warmer on the hotter, Hadean Earth, ∼60°C–70°C). For example, hydrocarbons containing 16–29 carbon atoms have been detected in the Rainbow ultramafic hydrothermal system, Mid-Atlantic Ridge (Holm and Charlou, 2001). Hydrothermal vents are often disqualified as efficient reactors for the synthesis of bioorganic molecules, because of their high temperatures, but this applies only to the very hot (>300°C) black smokers, and not vast amount of oceanic crust that provides a three-dimensional substrate through which fluids are flowing, including the cooler (120°C) white smokers and diffuse venting. However, the products that are synthesized in hot vents are rapidly quenched in the surrounding cooler water, which may preserve those organics formed. Recently, hydrothermal-sedimentary environments have been suggested as prebiotic reactors for the origin of life (Westall et al., 2018).

Intense bombardment by meteors and comets probably caused some chemical reprocessing of the Earth’s primitive atmosphere by impact shock chemistry (cf. Maher and Stevenson, 1988; Melosh and Vickery, 1989; Abramov et al., 2013; Mojzsis et al., 2019). This was demonstrated by Furukawa et al. (2009) who subjected a mixture of solid carbon, iron, nickel, water, and nitrogen to high-velocity impacts using a propellant gun. They recovered several organic molecules after the impact, including complex molecules, such as fatty acids and amines. Glycine, the simplest protein-building amino acid, was formed when the starting material contained ammonia, which is believed to have been formed during prior impacts on the early Earth. All the RNA canonical nucleobases and the simplest amino acid glycine were obtained when impacting a mixture of carbon monoxide, methane, and molecular nitrogen using a terawatt high-power laser system in the presence of montmorillonite (Ferus et al., 2020).

Figure 11 is a block diagramme showing the prebiotic products arriving on, or formed on, the early Earth, and illustrating the interactions between different components and different environments. Note that there is no significant transfer of compounds between the subaerial and shallow water hydrothermal environments and those of the deep ocean; any transported molecules into the deep would be so diluted that their impact would be negligible. We note here that some of the organic compounds found in the interstellar medium are fairly complex (up to 256 molecular species, Guérlin and Cernicharo, 2022) and may form from radicals made by photolysis (high energy light, up to UV, Öberg, 2016). Some of this material may be the source of meteoritic material (although in the carbonaceous chondrite meteorites this was subsequently reprocessed by aqueous activity); the “kerogen” found in carbonaceous chondrites may be originally of interstellar origin (as shown by the carbon isotopic composition, Swart et al., 1983).

A great number of organic molecules, including amino acids, have been found in carbonaceous chondrites. The Murchison meteorite, a CM2 type carbonaceous chondrite that fell in Australia in 1969, has been extensively analyzed (Shock and Schulte 1990; Pizzarello, 2007; Pizzarello and Shock, 2010; Pizzarello and Shock, 2017; Glavin et al., 2021). A combination of high-resolution analytical methods applied to the organic fraction of Murchison extracted under mild conditions documented its great chemical diversity of tens of thousands of different molecular compositions and likely millions of diverse structures (Schmitt-Kopplin et al., 2010; Hertkorn et al., 2015). The soluble organic compounds found in the meteorite represent a diverse and abundant group of organics that range from small water-soluble compounds, such as amino acids and polyols, up to 30 carbon-long hydrocarbons.

Vesicle-forming fatty acids have been extracted from different carbonaceous meteorites (Yuen and Kvenvolden, 1973; Deamer, 1985; Deamer, 1998). Nucleic acid bases, purines and pyrimidines, have also been detected in the Murchison meteorite (Stoks and Schwartz, 1982), and ribose was recently identified in the soluble organic matter of carbonaceous chondrites (Furukawa et al., 2019). The total number of meteoritic amino acids detected is now about one hundred, including all the possible α-amino alkylamino acids up to seven carbon atoms, as well as large abundances of N-substituted, cyclic, β, γ, δ, and ε-amino acids, and eight protein-building amino acids. Most of the amino acids detected in carbonaceous chondrites are chiral but are present as racemic mixtures, i.e. L- and D-enantiomers are present in equal proportions. However, Cronin and Pizzarello (1997) found L-enantiomer excesses in six α-methyl-α-amino alkanoic acids from the Murchison (2.8%–9.2%) and Murray (1.0%–6.0%) carbonaceous chondrites, and an enantiomeric excess up to 18% has been measured for isovaline. These amino acids are either unknown or rare in the terrestrial biosphere and cannot therefore be attributed to terrestrial contamination (Pizzarello 2007). In addition, the endogeneity of D- and L-isovaline enantiomers is supported by carbon and hydrogen isotopic data (Pizzarello et al., 2003; Pizzarello and Huang, 2005). Large enantiomeric excesses up to 60% have also been found in Antartica Renazzo-type (CR) chondrites (Pizzarello et al., 2012). The meteoritic enantiomeric excesses figure prominently among the many hypotheses put forward to explain the emergence of homochirality, the one-handedness, of life. Looking for homochiral entities beyond the Earth would probably be one of the best ways to detect extraterrestrial life (Brack, 2019; Gleiser, 2022; Lee et al., 2022).

From dust collections in the Greenland and Antarctica ice sheet (Maurette, 1998; Maurette, 2006), it appears that the Earth captures interplanetary dust as micrometeorites at a rate of about 20,000 tons per year. The value has been lowered to 5,200 tons per year by analyzing recent melts of large volumes of ultra-clean snow at Dome C in Antarctica (Rojas et al., 2021). This value is much higher than the most reliable estimate of the meteorite flux, i.e., about 10 tons per year. At least, ∼20 wt% of the micrometeorites survive unmelted upon atmospheric entry. As their kerogen fraction represents about 2.5 wt% of carbon, this amounts to a total mass of kerogen on the early Earth surface equivalent to a ∼30 m thick global layer (Maurette and Brack, 2006), although this represents an accumulation rate of about 10–6 mm/y over a 500 Ma period. One amino acid, α-amino isobutyric acid, has been identified in Antarctic micrometeorites (Brinton et al., 1998; Matrajt et al., 2004). These grains contain also a high proportion of metallic sulfides, oxides and clay minerals (Folco and Cordier, 2015), a rich variety of inorganic catalysts which could have promoted the reactions of the carbonaceous material which led to the origin of life. Many similarities can be found between Antarctic micrometeorites and the comet Wild 2 samples, in terms of chemical, mineralogical, and isotopic compositions, and in the structure and composition of their carbonaceous matter (Dobrica et al., 2009).

Nevertheless, one of the issues with imported organic matter is what happens to it upon interaction with the early oceans? Certainly carbonaceous chondrites and micrometeorites would have served as raw material sources. However, in the hot, slightly acidic conditions of the early seawaters (especially in shallow water regions where mechanical disruption would have contributed to disaggregation of the exogenic rocks), this material would have been broken down and incorporated into the sedimentary deposits. While numerous layers of Palaeoarchaean extraterrestrial spherule beds are testimony to impacts and the presence of impacted rocks (Lowe et al., 2003; Lowe et al., 2014; Lowe and Byerly, 2015), the recent documentation of finely particulate extraterrestrial organic matter disseminated within early terrestrial sediment (Gourier et al., 2019) is testimony to the fine-scale incorporation of extraterrestrial organic matter into hydrothermal sediments and its availability for reworking, transformation and availability for prebiotic processes.

Two distinct hypotheses have been put forward for the inception of life depending on the availability of organics. In a “metabolism-first” approach, the proponents of an autotrophic life call for the direct formation of simple molecules from carbon dioxide that rapidly evolve to life (Wächtershäuser, 2007). In this scenario, the energy is provided by chemical and thermal disequilibrium. Shock et al. (1995) and McCollom and Shock (1997) note that the geochemical constraints on organic formation and early metabolism are readily met in hydrothermal systems. The energy source required to reduce CO2 could have been provided by the oxidative formation of pyrite from iron sulphide and hydrogen sulphide, giving rise to a two-dimensional mineral surface metabolism. Along with the scenario proposed by Michael Russel (Martin et al., 2008), a laboratory set up simulating conditions prevailing in alkaline hydrothermal vents generated low yields of simple organics (Herschy et al., 2014). So far, the proponents of a “metabolism-first” approach have not been able to produce large enough precursor prebiotic molecules.

In the second hypothesis, the primeval soup scenario, also labelled “replication-first”, preformed complex organic molecules accumulated in a warm little pond, ‹‹ à la ›› Darwin. Targeting a primitive living cell-like system, great efforts are still being deployed in laboratories to produce boundary molecules providing compartmentalization, protein enzymes and RNA, using either hydrothermal conditions, wet-dry cycling or minerals. Nevertheless, we note that wet-drying cycles are not necessary for dehydration reactions.

For compartmentalization, amphiphilic compounds assemble into membranous vesicles in hydrothermal hot spring water (Monnard and Walde, 2015; Milshteyn et al., 2018; Damer and Deamer, 2020; Deamer et al., 2019). Experiments that demonstrate how different prebiotically-available building blocks can become precursors of vesicle-forming phospholipids have been reviewed elsewhere (Fiore and Strazewski, 2016). Mixtures of C10–C15 single-chain amphiphiles form vesicles in aqueous solutions at temperatures of ∼70 C in the presence of isoprenoids and under strongly alkaline conditions (Jordan et al., 2019). Vesicles functionalized with RNA and peptides could have been an interesting step towards the formation of early protocells (Izgu et al., 2016). The possible origins of a cellular life have been reviewed elsewhere (Schrum et al., 2010).

As for prebiotic peptides, seminal articles include the possible role of short peptides in the steps towards the formation of a protocell (Fishkis, 2007), the formation of peptides on oxide surfaces (Lambert et al., 2013; Kitadai et al., 2017), mineral surface chemistry control for origin of prebiotic peptides (Erastova et al., 2017), ester-mediated amide bond formation driven by wet–dry cycles (Forsythe et al., 2015), or the formation and self-assembly of long prebiotic polypeptides produced by the condensation of non-activated amino acids on oxide surfaces (Martra et al., 2014). Frenkel-Pinter et al. (2020) exhaustively reviewed the present state of the art in this question. More recently, abiotic synthesis and chain extension of unique peptide isomers from free amino acids have been obtained in aqueous microdroplets (Holden et al., 2022). Self-replicating catalytic amyloid peptides have been hypothesized as prebiotic informational and protometabolic entities (Maury, 2018; Rout et al., 2022).

RNA is generally claimed as being the most important molecule for the origin of life (Budin and Szostak, 2010; Bernhardt, 2012; Higgs and Lehman, 2015; Benner et al., 2020). In this respect, significant progress has been made in “one-pot” syntheses. Sutherland’s team, in a relevant example of system chemistry, showed that precursors of ribonucleotides, amino acids and lipids can all be derived by the reductive homologation of hydrogen cyanide and some of its derivatives, and thus that all the cellular subsystems could have arisen simultaneously through common chemistry (Patel et al., 2015). A systems chemistry approach has been recently developed (Strazewski, 2019). The synthesis of pyrimidine nucleosides driven solely by wet-dry cycles has also been reported (Becker et al., 2018). In the presence of phosphate-containing minerals, 5′-mono- and diphosphates also form selectively in one-pot reactions (Becker et al., 2019). Diamidophosphate efficiently phosphorylates a wide variety of potential building blocks, nucleosides/nucleotides, amino acids and lipid precursors, under aqueous conditions. Significantly, higher-order structures, oligonucleotides, peptides and liposomes, are formed under the same phosphorylation reaction conditions (Gibard et al., 2018). Abiotic synthesis of purine and pyrimidine ribonucleosides has been obtained in aqueous microdroplets (Nam et al., 2018). The chemistry of abiotic nucleotide synthesis has been reviewed elsewhere (Yadav et al., 2020).

Despite these encouraging results, it still seems unlikely that life could have started with RNA molecules, because whole RNA strands are not simple enough and yet too difficult to assemble under prebiotic conditions. Our preliminary approximation to answer the question asked is that a promising avenue will perhaps consist in far from equilibrium, open wet-dry cycling organic reactions, occurring on mineral surfaces under hydrothermal-like conditions, and generating a “replication-first” scenario in the primeval soup.

How long did it take for life to emerge? We do not know, but the fact that the process had to be driven by a kinetic momentum such that there could be no turning back, no undoing of the haphazard combination of molecules to form a protocell, means that it had to have occurred on shorter time scales rather than longer ones, because of the more unstable surface conditions reigning on the early Earth. However, given the uncertainty of our understanding of conditions on the early Earth, all that we can state is that, not knowing how RNA-based life appeared, it is not possible to estimate the time necessary for life to emerge (Orgel, 1998). This also depends on exactly which steps in the emergence of life are considered in the processes leading to the appearance of an auto-catalytic entity. Do we include the time required for the creation of the prebiotic building bricks of life, or simply the assemblage of the pre-formed molecules? In the latter case, the process could have started very rapidly or even instantaneously, like solutions that suddenly crystallize upon the addition of a crystal acting as a germ. There is also the possibility that life may have emerged more than once in different places on the planet but could have been extinguished by changing environmental parameters, such as a catastrophic meteoroid or asteroid impact (more common during the Hadean (Marchi et al., 2021), and re-emerged. Could life have emerged simultaneously in numerous locations, where one form was more efficient than the others and took over? These are questions we may not be able to answer on Earth, but perhaps future exploration of Mars, where rocks more than 4 Ga are preserved, may provide some insight into this possibility.

We take, as the context for these considerations, our initial premises, that life emerged after the Moon-forming impact about 4.51 Ga (Barboni et al., 2017) and was well-established and relatively evolved by about 3.5 Ga (Noffke et al., 2013; Hickman-Lewis et al., 2018) if not before (e.g., Hassenkam et al., 2017). Zircon crystals offer strong evidence for crustal water at about 4.19 Ga but habitable conditions conducive to prebiotic chemistry probably existed well before (Sleep et al., 2014). How much time did it take between the emergence of life and advanced life? This would have been a process, not an event, rather than individual, isolated cells, perhaps some kind of symbionts?

By demonstrating in 1953 that it was possible to form amino acids—the building blocks of proteins - from methane, Stanley Miller generated the ambitious hope that chemists will be able to create life in a test tube. Despite the tremendous efforts of chemists tackling the problem, the dream has not yet been accomplished. Obviously, experiments should adhere to geologically relevant environments pertinent to the early Earth. A rational methodology involving chemical systems on heterogeneous geological surfaces has been proposed, based on the relationship between prebiotic chemistry and the geological conditions of the Hadean Earth (Dass et al., 2016). A recently-created consortium of researchers seeking to understand the origins of life on Earth and in the Universe will hopefully advance the field (Müller et al., 2022). Novel apparatuses for incorporating natural selection processes have been also proposed (Root-Bernstein and Brown, 2022).

For future research, one approach could be far-from-equilibrium wet-dry cycling (either subaerial exposure or dehydration through chelation to mineral surfaces) of organic reactions occurring repeatedly and iteratively at mineral surfaces under hydrothermal-like conditions, an approach merging geology and chemistry (although we noted above that wet-dry cycling is not essential to concentrating organic molecules). Also, information from biology shows that early life flourished at relatively high temperatures (Woese, 1987; Miller and Lazcano, 1995; Gaucher et al., 2003; Camprubi et al., 2019 and references therein). Slight enantiomeric excesses of the reactants, as well as clays and other minerals, should be used. Concerning clays, they formed as soon as liquid water was present on Earth’s surface through aqueous alteration of the early volcanic rocks. Bernal (1949) listed the advantageous features of clays, i.e., their ordered arrangement, large adsorption capacity, shielding capacity against sunlight, ability to concentrate organic chemicals, and ability to serve as polymerization templates. Given the importance of catalysts for prebiotic chemistry, minerals, such as clays were probably the most efficient partners of the primitive game of life (Brack, 2013; Hansma, 2022; Kloprogge and Hartman 2022) but other minerals with reactive surfaces were present on the early Earth, such as spinels, could have been equally useful (Iqubal et al., 2016). The use of microdroplets represents also an interesting avenue for the synthesis of both peptides and nucleotides (e.g., Wang et al., 2021).