**3. Minerals promote the prebiotic synthesis**

### **3.1 Carbon fixation and C-C bond formation**

Miller-Urey experiments open up the whole field of prebiotic chemistry [38]. At that time, the early Earth atmosphere was considered to be reducing and mainly composed of reducing species, such as methane, H2, CO, NH3, etc. However, this has been questioned, and the current consensus is that the early Earth atmosphere was oxidizing, with CO2, N2, and H2O as the major components, with a trace amount of H2 [39]. Based on this, a chemoautotrophic scenario pioneered by Wächtershäuser was proposed [27–31]. The chemoautotrophic origin of life scenario relies on primordial carbon fixation within a sulfide-rich hydrothermal vent. Driven by the reducing energy and activation ability of carbon monoxide (CO), many types of reactions were demonstrated. For example, at 100°C, C-C bond formation with the generation of acetate proceeds on FeS and NiS from a mixture of CO and CH3SH, or from a mixture of CO and H2S alone with the addition of Se [27]. This reaction resembles the reductive acetyl-coenzyme A (acetyl-CoA) pathway, where the key enzyme, acetyl-CoA synthase, contains a Ni-Fe-S active center and forms acetyl-CoA from coenzyme A, CO, and a methyl group. CO was also demonstrated to promote the reductive polymerization of HCN to α-hydroxy acids and α-amino acids at 80 ~ 120°C on FeS or NiS precipitates [29], where glycine and alanine were formed accompanied by glycolic and lactic acid. The polymerization of amino acids into short peptides was also demonstrated by CO activation [28]. In these experiments, high-pressure CO gas was used (1 ~ 75 bar) [29].

Based on the modeling of the atmosphere in the late Hadean period [39], the most abundant abiotic carbon feedstock on the early Earth is carbon dioxide (CO2), with a trace amount of CO. Therefore, for the autotrophic origin of life scenarios, CO2 was a more preferable and primary carbon source for primordial biosynthesis. CO2 dissolved in the ocean and resulted in a mildly acidic ocean (pH ~ 5.5) [40]. Compared to CO, CO2 is a chemically inert molecule that requires high activation energy. The

acetyl-CoA (AcCoA, or Wood-Ljungdahl) pathway is considered to be the most ancient autotrophic CO2 fixation pathway in nature [41]. To answer the question of how CO2 reduction occurs before the evolution of proteins, CO2 reduction has been demonstrated by Varma et al. [42], that native transition metals (Fe0 , Ni0 , and Co0 ) can reduce CO2 to acetate and pyruvate in millimolar concentrations. Moreover, in the AHV theory proposed by Russell and colleagues [9, 15, 18], the carbon fixation was driven by the direct redox coupling of CO2 and H2 on metal sulfides or oxyhydroxides [16]. Later, Lane and Martin [43, 44] also discussed the plausible relevance of the pH gradients in membrane-separated alkaline hydrothermal vent systems with the H<sup>+</sup> gradients across the cell membranes that drive ubiquitous chemiosmotic coupling in all life forms. This scenario was approved in recent work by Sojo et al., that in a microfluidic system with a freshly precipitated thin Fe-, or Ni-sulfide mineral membrane, CO2 in simulated sea water side was reduced to formate at several micromolar yield [45]. The reaction was likely promoted by pH gradient as evidenced by the boosted yield with increased pH gradient. This shows that CO2 was reduced by Fe or Ni sulfide, probably through an electrochemical process coupled with the oxidation of H2.

Inspired by Russell's AHV origin of life theory, in the past decade, an alternative scenario "geo-electrochemical driven carbon fixation" has been explored by Nakamura, Yamamoto, Kitadai, and their colleagues [46–55]. This theory is based on the pH, redox, and thermal gradients between the alkaline hydrothermal vent and seawater. Those gradients thermodynamically drive the redox conversions by coupling H2 oxidation in the hydrothermal fluid/mineral interface with CO2 reduction at the seawater/mineral interface. The experimental results show that CO2 was effectively reduced to CO on certain types of sulfide, such as Ag2S and CdS [53], with much higher efficiency than FeS and NiS despite their higher geological abundances [48, 53]. The product selectivity highly depends on the identity of the metal in the sulfide minerals. Using the CO gas generated by electrochemical reduction, many reactions that were reported in Wächtershäuser's experiments were confirmed [53]. Since the disequilibrium and gradients in the deep-sea hydrothermal vent system can be maintained throughout the Earth's early history, the geo-electrochemical CO2 reduction provided a stable and sustainable source of CO which could have fueled the prebiotic synthesis. By simulating the geo-electrochemical conditions of alkaline hydrothermal vents, other researchers also reported CO2 reduction to a variety of products, including formate [49, 53], acetate [49], methane [48, 52], pyruvate [49], C2H6 [52], methanol [49] on Fe- or Ni-containing sulfides.

Regarding the reaction mechanisms, relevance with biological enzymes has been suggested. In biology, the enzyme catalyzing the reduction of CO2 to CO is carbon monoxide dehydrogenase (CODH) with a [NiFe4S4] cluster [56]. The reaction is considered to be the oldest pathway of biological carbon fixation and therefore may have been involved in the origin of life [9, 10]. Yamaguchi et al. first studied two metal sulfides greigite (Fe3S4) and violarite (FeNi2S4) and found that Ni-bearing sulfides show higher efficiency in reducing CO2 [48]. Further, Lee et al. reported the in-situ FTIR spectroscopic analyses of the surface intermediate during electrochemical CO2 reduction on these two minerals [52]. Intermediate species assignable to surfacebound CO2 and formyl groups were found to be stabilized in the presence of Ni, lending insight into its role in enhancing the multistep CO2 reduction process. These researches suggested an evolutionary link between mineral-catalyzed prebiotic reaction and enzyme-catalyzed biochemical reaction.

### **3.2 Nitrogen fixation**

The most geologically abundant N source on the early Earth is dinitrogen (N2). This molecule is chemically inert because of the stable N ≡ N triple bond. Under high-temperature conditions, N2 can be reduced hydrothermally to ammonia, where reductants were considered to be abundant H2S [57] or sulfide minerals. The yield of ammonia using H2S as the reductant at low temperature (120°C) is relatively low even with iron monosulfide as the catalyst and is considered to be insufficient for providing ammonia for prebiotic synthesis. On the other hand, there have been accumulating reports on electrochemical reduction of N2 on Fe2O3 [58–60], or FeS [61], CuS [62–64], Mo sulfides [65, 66] at ambient temperature and pressure. These types of reactions could contribute to the prebiotic synthesis of ammonia, following the geo-electrochemistry-driven prebiotic synthesis scenario.

Another chemically more active form of inorganic N species on the early Earth is nitrogen oxyanions including nitrate (NO3 − ) and nitrite (NO2 − ). These compounds were formed by lightning and photochemical processes of atmospheric N2 and CO2 with subsequent hydration during rainfall. This could lead to the accumulation of these nitrogen oxyanions in the early ocean with a concentration of micromolar level that is expected to be sufficient for serving as high-potential electron acceptors for the emergence of life in the oceanic environment [67]. NO3 − and NO2 − are highpotential electron acceptors (E0 (NO3 − /NO2 − ) = 0.835 V vs. NHE (normal hydrogen electrode), E0 (NO2 − /NO) = 1.202 V vs. NHE) [68]. These electron acceptors are invoked to participate in redox coupling with the oxidation of reducing species, such as methane [17], for the synthesis of active methyl-bearing species such as Acetyl-CoA-like molecules.

Despite their relatively higher reactivity, the reduction of NO3 − is still kinetically demanding due to the low chemical affinity and low complexation ability with metal sites. Therefore, industrial reduction of NO3 − typically requires relatively harsh conditions, such as very acidic pH, UV-photolysis, or high temperature [69]. Fe-based species have been extensively studied for ammonia synthesis via reduction of NO3 − and NO2 − . These include mackinawite (FeS) [70, 71], Fe2+ [72, 73], pyrite (FeS2) [74, 75], and green rust (FeII 4FeIII 2(OH)12SO4·yH2O) [76]. Fe(II) ions and green rust can reduce nitrite to ammonia at neutral to alkaline pH (pH ≥ 7) [73, 77]. Although Fe(II) cannot reduce nitrate, the addition of a trace amount of Cu2+ enables the generation of ammonia at pH 8 [77]. Green rust and pyrite can also reduce nitrate into ammonia at pH ~8, which however requires an anion-free environment due to the strong competing adsorption effect from many types of anions [74, 76]. Moreover, the reduction ability of Fe(II) and green rust decrease upon decreasing the solution pH to acidic pH [73, 76]. Therefore, this reaction could have consumed ferrous ions in the ocean. In high temperature (300°C), high pressure (50 ~ 500 MPa) hydrothermal setting, a variety of minerals (Fe-, Ni-, Cu-sulfides, and magnetite) can reduce nitrate into ammonia [78] and the ammonia is maintained stably in contact with these minerals. Accordingly, it was argued that the hydrothermal vent system could have supplied sufficient ammonia for the prebiotic synthesis of biomolecules, such as amino acids. However, at the same time, these mineral-promoted reactions are stoichiometric and strongly affected by the presence of other ions and low pH [73, 76].

Metals other than iron have rarely been considered to account for the geochemical reduction of nitrogen oxyanions. Nevertheless, biological nitrate reduction is

catalyzed by nitrate reductase enzyme. All types of nitrate reductases exclusively utilize mononuclear molybdenum as the active center which is bounded by one or two dithiolene groups (-S-C-C-S-) ligated to a pterin group and other ligands (oxo, water, sulfur, etc.) [79, 80]. Inspired by the enzyme structures, the bio-inspired mineral catalysts provide another approach to tackle the kinetic problem. An oxo-bearing molybdenum sulfide as a structural analog of nitrate reductase was synthesized using the hydrothermal method. Notably, this mineral catalyzes both nitrate and nitrite reductions at a wide range of pH, with the generation of a variety of products, including NO, N2O, NH4 + , and N2 [51, 69]. The reaction mechanism of nitrate reduction resembles that of the enzyme, relying on a redox-active, pentavalent [(MoV=O)S4] species as the active intermediate. This species was likely generated by a concerted proton-coupled electron transfer step, as evidenced by the near Nernstian behavior revealed by the pH dependence [69]. During nitrite reduction, this mineral show ability to decouple the proton transfer with electron transfer, facilitating a pH-regulated reaction selectivity towards the N-N coupling process [81, 82]. Therefore, this study shows that minerals can not only catalyze a similar reaction with the enzyme but also share a similar reaction mechanism, therefore reinforcing the evolutionary link between geo- and biochemistry.
