Here we argue that life emerged on Earth from a redox and pH front at c. 4.2 Ga. This front occurred where hot (c. 150 degrees C), extremely reduced, alkaline, bisulphide-bearing, submarine seepage waters interfaced with the acid, warm (c. 90 degrees C), iron-hearing Hadean ocean. The low pH of the ocean was imparted by the ten bars of CO2 considered to dominate the Hadean atmosphere/hydrosphere. Disequilibrium between the two solutions was maintained by the spontaneous precipitation of a colloidal FeS membrane. Iron monosulphide bubbles comprising this membrane were inflated by the hydrothermal solution upon sulphide mounds at the seepage sites. Our hypothesis is that the FeS membrane, laced with nickel, acted as a semipermeable catalytic boundary between the two fluids, encouraging synthesis of organic anions by hydrogenation and carboxylation of hydrothermal organic primers. The ocean provided carbonate, phosphate, iron, nickel and protons; the hydrothermal solution was the source of ammonia, acetate, HS-, H2 and tungsten, as well as minor concentrations of organic sulphides and perhaps cyanide and acetaldehyde. The mean redox potential (delta Eh) across the membrane, with the energy to drive synthesis, would have approximated to 300 millivolts. The generation of organic anions would have led to an increase in osmotic pressure within the FeS bubbles. Thus osmotic pressure could take over from hydraulic pressure as the driving force for distension, budding and reproduction of the bubbles. Condensation of the organic molecules to polymers, particularly organic sulphides, was driven by pyrophosphate hydrolysis. Regeneration of pyrophosphate from the monophosphate in the membrane was facilitated by protons contributed from the Hadean ocean. This was the first use by a metabolizing system of protonmotive force (driven by natural delta pH) which also would have amounted to c. 300 millivolts. Protonmotive force is the universal energy transduction mechanism of life. Taken together with the redox potential across the membrane, the total electrochemical and chemical energy available for protometabolism amounted to a continuous supply at more than half a volt. The role of the iron sulphide membrane in keeping the two solutions separated was appropriated by the newly synthesized organic sulphide polymers. This organic take-over of the membrane material led to the miniaturization of the metabolizing system. Information systems to govern replication could have developed penecontemporaneously in this same milieu. But iron, sulphur and phosphate, inorganic components of earliest life, continued to be involved in metabolism.

For life to have emerged from CO₂, rocks, and water on the early Earth, a sustained source of chemically transducible energy was essential. The serpentinization process is emerging as an increasingly likely source of that energy. Serpentinization of ultramafic crust would have continuously supplied hydrogen, methane, minor formate, and ammonia, as well as calcium and traces of acetate, molybdenum and tungsten, to off-ridge alkaline hydrothermal springs that interfaced with the metal-rich carbonic Hadean Ocean. Silica and bisulfide were also delivered to these springs where cherts and sulfides were intersected by the alkaline solutions. The proton and redox gradients so generated represent a rich source of naturally produced chemiosmotic energy, stemming from geochemistry that merely had to be tapped, rather than induced, by the earliest biochemical systems. Hydrothermal mounds accumulating at similar sites in today's oceans offer conceptual and experimental models for the chemistry germane to the emergence of life, although the ubiquity of microbial communities at such sites in addition to our oxygenated atmosphere preclude an exact analogy.

We propose that life emerged from growing aggregates of iron sulphide bubbles containing alkaline and highly reduced hydrothermal solution. These bubbles were inflated hydrostatically at sulphidic submarine hot springs sited some distance from oceanic spreading centers four billion years ago. The membrane enclosing the bubbles was precipitated in response to contact between the spring waters and the mildly oxidized, acidic and iron-bearing Hadean ocean water. As the gelatinous sulphide bubbles aged and were inflated beyond their strength they budded, producing contiguous daughter bubbles by the precipitation of new membrane.[FeeS2] +/0 or [Fe4S4 ]2+/+ clusters, possibly bonded by hydrothermal thiolate ligands as proferredoxins, could have catalyzed oxidation of thiolates to disulphides, thereby modifying membrane properties.We envisage the earliest iron sulphide bubbles (proCorrespondence to: M. J. Russell Glossary: Hollow pyrite botryoids: hollow hemispheres of cryptocrystalline pyrite (FeS2) 0.1-1 mm across. Fischer-Tropsch syntheses: the highly exothermic catalytic hydrogenation of CO to hydrocarbons and aliphatic oxygenated compounds using finely divided iron. Greigite (Fe3S4): metastable iron sulphide precipitated from aqueous solution in a gel at 100°C and containing two-thirds of its iron as the high-spin ferric ion. Haber-Bosch process: the exothermic catalytic hydrogenation of nitrogen to yield ammonia. Probotryoid: a hydrostatically inflated colloidal iron monosulphide bubble; precursor to hollow botryoids and the progenitor to protocelts. Proferredoxins: [F%S2] and [F%MS4] clusters (M = Fe, Mo, W, Ni, etc.) ligated by abiogenic thiols and thiolates. Protocell: a cell comprised mainly of abiogenic organics including thiols with subordinate iron sulphides, partly as proferredoxins; growth results from catabolism and osmotic pressure botryoids) first growing by hydrostatic inflation with hydrothermal fluid, but evolving to grow mainly by osmosis (the protocellular stage), driven by (1) catabolism of hydrothermal abiogenic organics trapped on the inner walls of the membrane, catalyzed by the iron sulphide clusters; and (2) cleavage of hydrophobic compounds dissolved in the membrane to hydrophilic moieties which were translocated, by the proton motive force inherent in the acidic Hadean ocean, to the alkaline interior of the protocell. The organics were generated first within the hydrothermal convective system feeding the hot springs operating in the oceanic crust and later in the pyritizing mound developing on the sea floor, as a consequence of the reduction of CO, CO e, and formaldehyde by Fe z+-and se--bearing minerals. We imagine the physicochemical interactions in and on the membrane to have been sufficiently complex to have engendered auto-and cross-catalytic replication. The membrane may have been constructed in such a way that a "successful" parent could have "informed" the daughters of membrane characteristics functional for the then-current level of evolution.

Iron monosulphide globules and tubes grown in the laboratory have similar morphologies to the fossil pyrite botryoids and chimneys found in the Silvermines exhalative sedimentary ore-body of Carboniferous age in Ireland. We envisage analogous fine structures growing at hot springs (100-200°C) in the earliest oceans as having provided the culture chambers and flow reactors for life to originate by phosphorylation and growth of organic molecules on the iron sulphide surfaces. Such sulphide structures grown in the laboratory could be used in origin-of-life experiments. We: a hypothesis, System. Appl. Microbiof., 10,207-210. Wachtershauser G. (1988b) Before enzymes and templates: theory of surface metabolism, Minobiol. Reviews, 52,452-484. Wachtershauser G. (1988c) Deutsches Patentamt 3812158, FRG. Wark E.E. and Wark I.W. (1935) The physical chemistry of flotation. VI the adsorption of amines by sulfide minerals, J. Phys. Chem., 39,1021-1030. Windley B.F., S i p s o n P.R. and Muir M.D. (1984) The role of atmospheric evolution in Precambrian metallogenesis, Fortschr. Miner., 62, 99,712-720. 253-267.

Because of the continuous focusing of thermal and chemical energy, ancient submarine hot springs are contenders as sites for the origin of life. But it is generally assumed that these would be of the acid and high-temperature 'black smoker' variety (Corliss et al., 1981). In fact today the greater part of the ocean circulates through off-ridge springs where it issues after modification at temperatures of around 40 degrees C or so but with the potential to reach 200 degrees C. Such offridge or ridge-flank springs remind us that there are other candidate sites for the origin of life. Although there is no firm indication of the pH of these off-ridge springs we have argued that the solutions are likely to be alkaline rather than acid, We test the feasibility of this idea using EQ geochemical water-rock interaction modelling codes (Wolery 1983) and find that for a range of possible initial chemistries of Hadean seawater, the pH of issuing solutions at around 200 degrees C is around one or more units alkaline. Such pH values hold for interaction with both basaltic and komatiitic crust. The robustness of this result suggests to us that alkaline submarine springs of moderate temperature, carrying many hundreds of ppm HS to the ocean basins, are also serious contenders as sites for the origin of life, particularly as Hadean seawater was probably slightly acid, with a dissolved iron concentration approaching 100 ppm. On mixing of these solutions, supersaturation, especially of iron sulphide, would lead to the precipitation of colloidal gels. In our view iron sulphide was the likely substance of, or contributor to, the first vesicle membranes which led to life, as the supply organic molecules would have been limited in the Hadean. Such a membrane would have bid catalytic properties, expansivity, and would have maintained the natural chemiosmotic gradient, a consequence of the acid ocean and the alkaline interior to the vesicles.