The origin of life research has gradually shifted towards the identification of geochemical environments favourable for prebiotic synthesis (Hazen et al., 2008; Hazen and Sverjensky, 2010). An influential “Hot Spring hypothesis” proposes that the origin of life may have occurred in hot springs (Damer and Deamer, 2020). This hypothesis lends a central role to intermittent volcanic pools, which provide environment similar to that occurring in impact craters. Violent meteorite impacts represent potentially important energy and material sources deeply modifying the physical and chemical shape of the early Earth and other planets such as, e.g., Mars and Venus. For instance, estimates of the Earth’s impact flux prior to ∼ 4 Gya suggest an annual energy deposition larger than 10²⁰ J (Chyba and Sagan, 1992). Extrapolating such an estimate to Mars by using the impact rates and the velocity scaling taken from the literature (Ryder, 2002; Ivanov, 2008; Schlichting et al., 2012; Abramov and Mojzsis, 2016) leads to 10^–1 to 10^–2 times less deposited energy than is expected for Earth. Some contemporary scientists consider Mars an “archive” of planetary evolution. Almost all direct geological information from Earth’s Hadean eon is now long lost. Nowadays, the Yarrabubba Crater likely represents the oldest impact structure on Earth (dated back to 2,229 ± 5 Ma) (Erickson et al., 2020). On the other hand, important data is provided by the exploration of Mars, which lacks plate tectonics and at least two-thirds of its present surface is far older ( > 3.5 Gya) than Earth’s (McKay and Stoker, 1989). Martian surface therefore may preserve traces of processes relevant to prebiotic conditions (Oberbeck et al., 1990; Sasselov et al., 2020), even if life had never existed there.

Some evidence of this preservation comes from martian meteorites. As an example, the Nakhla meteorite contains niche environments exclusively formed during an impact that induced the circulation of hydrothermal fluids which altered the rock and the formed secondary clay and other phases (Chatzitheodoridis et al., 2014). Recent discoveries indicate a strong connection between impacts and the presence of rich organic inventories on the surfaces of early Earth and Mars (e.g., in Chyba and Sagan (1992); Jenniskens et al. (1998); Patel et al. (2015); Pastorek et al. (2019); Erickson et al. (2020); Ferus et al. (2020); Todd and Öberg (2020) and references therein). In particular, the exploration of the Gale crater by the Curiosity rover has confirmed the existence of a relatively complex organic inventory, including chlorobenzene and from C-2 to C-4 dichloroalkanes (Freissinet et al., 2015). Moreover, pyrolysis products of lacustrine mudstones at the base of the 3.5 Ga Murray formation contained thiophenic, aromatic and aliphatic compounds (Eigenbrode et al., 2018). S-bearing species can also be synthesized over Fe/Ni metals in the solar nebula (Llorca and Casanova, 2000, 1998). On Earth, electron paramagnetic resonance (EPR) analysis of 3.33 Ga fossil sediments from the Josefsdal Chert in South Africa shows an anomalous layer of organic matter formed from tiny dust particles. which originate from a flux of micrometeorites falling through the oxygen-poor Archean atmosphere (Gourier et al., 2019). Side processes directly connected with impacts, such as the delivery of possibly reprocessed meteoritic materials (Chyba et al., 1990; Kuwahara and Sugita, 2015), the chemical action of impact plasma (Ferus et al., 2017b; Ferus et al., 2019), hydrothermally-active craters and UV radiation may have played crucial roles in chemical reactions changing the organic inventory present in craters on longer timescales (Sutherland, 2016).

Impact and post-impact synthesis of organic molecules attributed to life’s origins might have started from a broad repertoire of prebiotic molecules. However, the apparent relevance of two major starting compounds has enormously increased in the last decade: HCN (Powner et al., 2011; Powner and Sutherland, 2011; Sutherland, 2016; Ferus et al., 2017a; Civiš et al., 2017; Sutherland, 2017; Xu et al., 2017; Ferus et al., 2020) and its hydration product formamide (HCONH₂) (Saladino et al., 2012, 2016; Ferus et al., 2015b; Rotelli et al., 2016; Šponer et al., 2016; Cassone et al., 2017; Bizzarri et al., 2020; Cabedo et al., 2021).

In this paper, we aim to provide an overview of the chemical mechanisms associated with the post-impact synthesis starting from formamide - a prebiotic feedstock or intermediate. Formamide contains all the essential biogenic elements: carbon, nitrogen, hydrogen, and oxygen, and it is the simplest representation of a peptide linkage. Several plausible endogenous sources of formamide have been identified in early planetary environments: the hydrolysis of HCN (Saladino et al., 2006), volcanic activity (Sponer et al., 2016), radioactivity (Adam et al., 2018) and reprocessing of either simple atmospheric mixtures [CH₄, NH₃, H₂O, H₂] (Miller, 1953), [CO, CO₂, N₂, NH₃, H₂, H₂O] (Abelson, 1956) or more complex molecules such as formaldehyde (Ferus et al., 2019) or HNCO (Ferus et al., 2018) as well as their various combinations (Rode, 1999; Saitta and Saija, 2014; Cassone et al., 2018; Sponer et al., 2020). Formamide has also been observed in the interstellar medium, in young stars (Schutte et al., 1999; Kahane et al., 2013) and in comets (Biver et al., 2014; Goesmann et al., 2015). In addition, extra-terrestrial synthesis (Cassone et al., 2018) and exogenous delivery from extra-terrestrial environments is a potential source of this compound on early planets - including young Earth, Mars and Venus.

We report a synergistic experimental (FTIR spectroscopy, GC-MS) and computational (ab initio molecular dynamics) investigation on the thermal treatment of formamide in presence of clays, meteoritic Fe-Ni grains and iron oxide, we summarise the mechanism and show the relevant reaction intermediates. Besides, with the aim of offering a general and unified chemical picture, our findings are discussed in the context of previous studies on the prebiotic synthesis of canonical nucleobases initiated either by direct chemical action of an impact plasma or by subsequent formamide-based thermal treatment.

Experiments were designed to prove the existence of stable intermediates inspired by the various mechanisms proposed by Saladino et al. (2007). Many other mechanistic variants have been published by Springsteen, Krishnamurthy and others (Hudson et al., 2012; Wang et al., 2013a,b). The intermediates are predicted to be formed from formamide or a formamide-clay mixture during thermal treatment and the main purpose of this experimental approach was their detection via an absolute detection method such as the Fourier Transform Infrared (FTIR) spectroscopy.

Impact plasma and residual post-impact heat are very important energy sources plausibly capable of changing the chemical inventory of early planets undergoing heavy bombardment. These factors may have created unique niches likely amenable also to energetically demanding reactions for prebiotic synthesis. Impact plasma chemistry and physics has become a subject of both laboratory and theoretical studies (Chyba and Sagan, 1992). Laser experiments summarized in our paper represent one of the approaches with their significant advantages as well as limitations, discussed in our recent studies and references therein, for example (Mohammadi et al., 2020; Ferus et al., 2020; Křivková et al., 2021). However, multiple other studies using different approaches have been also performed, such as target shock experiments (McCaffrey et al., 2014; Martins et al., 2013; Takeuchi et al., 2020) and references therein), shock tubes (Singh et al., 2020) and references therein) or even by quantum chemical computations (Kroonblawd et al., 2019; Koziol and Goldman, 2015) and references therein).

The present study draws a unified picture of complex mechanisms of prebiotic synthesis initiated by impact plasma and thermochemistry in the post-impact environment in presence of iron-rich clays. Among elements such as Fe, Ni, Co, Al, Mn and Cr (Jarosewich, 1990), iron is an ubiquitous meteorite component. Iron can be present in the form of pure metal, alloy, a derived compound or as a constituent of various minerals (see (Jarosewich, 1990; Rubin, 1997; Brearley et al., 1998) and references therein). Therefore, investigation of formamide prebiotic chemistry in presence of Fe-smectites is addressed in this study. Prebiotic synthesis of nucleobases and glycine related to impact reprocessing of formamide can be summarised by two independent mechanisms following each other, as displayed in Figure 5. The whole process starts with impact plasma-assisted chemistry, as shown in Figure 5A, and follows with thermochemistry summarised in Figure 5B. For the sake of clarity, the results are also quantitatively summarised in Table 1. Intermediates of impact (A) or post-impact (B) chemistry directly detected in our experiments by means of absorption or emission spectroscopy and GC-MS are highlighted by green rectangles in Figure 5. The current investigation, as well as some previous studies (the list of references is provided alongside each particular scheme), combines in this way impact and post-impact chemistry of formamide toward a unified picture experimentally and theoretically supported.

Investigations focused on prebiotic chemistry initiated by an impact plasma (Civis et al., 2012; Ferus et al., 2012, 2015b; Ferus et al., 2014; Ferus et al., 2019) have reported that the origin of nucleobases, glycine and urea might start with the formation of radicals such as . CN, .H, :NH and . NH₂. FTIR spectroscopy has also given evidence of the formation of two intermediate species associated with the radical-based mechanism displayed in Figure 5A (Ferus et al., 2012): 2-amino-2-hydroxy-acetonitrile (44, AHAN) and 2-amino-2-hydroxy-malononitrile (45, AHMN). In the direct impact synthesis, formamidine (5) - detected in this study also as an intermediate species in the formamide thermochemistry - results from the reaction of formamide (1) with ammonia (4) released by the decomposition of formamide in the plasma environment and suddenly created during the impact event. Formamidine (5) then reacts with HCN (2) to give rise to 2-iminoacetonitrile (6). 2-amino-2-hydroxy-acetonitrile (44, AHAN) and 2-amino-2-hydroxy-malononitrile (45, AHMN) hold a pivotal role in the thermal synthesis of nucleobases from formamide. For instance, formamide (1) may react with NH₃ to give formamidine (5), also detected in this investigation. Formamidine can then react with 2-aminomalononitrile (7) to AICN (12), which is a known precursor of adenine (13), hypoxanthine (15) and guanine (14). Formamidine can also react with HCN to give rise to 2-iminoacetonitrile (6). Besides, another way to produce 2-iminoacetonitrile (6) is from formamide and HCN. Then, 2-iminoacetonitrile (6) can either react with HNCO (21) and then urea (22), or can produce 2-aminoacetonitrile (16), which is a precursor of purine (20) and glycine (17).

Synthesis of nucleobases at ppmv (parts per million by volume) levels was observed not only starting from formamide, but also in experiments simulating impact plasma reprocessing atmospheres composed of NH₃:CO:H₂O (Ferus et al., 2017b) or formaldehyde aqueous solutions (Ferus, M. et al., 2019). In such a specific chemistry, formamide has been detected among other products and likely it does not serve as a starting compound but rather as a key prebiotic intermediate (Figure 5A). This aspect is quite important because impact plasma might have then served not only as an initiator of the nucleobases and glycine origin, but also as an endogenous source of formamide. The simplest approach to estimating the de novo impact synthesis yield is a linear scaling of experimental data (Curci et al., 2004; Rimmer et al., 2019; Mohammadi et al., 2020). We use the annual impact energy deposition estimated by Chyba et al. (Chyba and Sagan, 1992) (i.e., 10²⁰ J) for the end of the Hadean era. With this value, the yield of formamide per joule of delivered impact energy estimated by Ferus et al. (Ferus et al., 2017b) (3 × 10¹³ molecules/J) can then directly be translated into a global impact-formed formamide deposition equal to 2.24 × 10⁵ tons per year. Together with the extra-terrestrial delivery of formamide, such a de novo synthesis might have represented its significant source. The relevance of the direct formamide synthesis can be illustrated on a Tunguska-scale impact event. As an example, let us apply the simulated impact-induced yields (Sutherland, 2016; Ferus et al., 2017b; Ferus et al., 2019) summarised for different experiments in Table 1 for the estimation of formamide and nucleobases amounts produced by such large local impact event. It has been estimated that Tunguska provided energy on the order of 10¹⁷ J (Drobyshevski, 2009).

Then, it would be possible to expect that a few hundreds of tons of de novo formamide was deposited over an impact area of about 10³ km² (area presumed to have been affected by the Tunguska impact) and led to the local synthesis and spread of a few ppm of all the canonical nucleobases in that environment. Linear scaling of this rational to the case of the Gale impact on Mars estimates a six orders of magnitude larger synthesis yield (Asphaug, 2008; Schwenzer et al., 2012). Similarly with formamide, direct synthesis of canonical nucleobases initiated by a simulated impact plasma has been observed in the case of the HCN-based chemistry as well (Ferus et al., 2020) and, similarly to the formamide case discussed here, HCN might have also been deposited locally (Todd and Öberg, 2020), a fact that could have created particularly favourable circumstances for prebiotic synthesis by post-impact chemistry (Sutherland, 2016). These estimates are of course very rough. Proper investigation including a chemical model is, however, beyond the scope of this paper. Also, significant amount of formamide and/or nucleobases produced by impacts does not wholly overcome the classical problems in prebiotic chemistry, such as how the chemical complexity increased from prebiotic molecules through simple organics to biopolymers, how interfaces for selection and concentration of more dilute solutions were provided and what created physical and chemical gradients and supported fluxes or cycles in the environment (Hazen and Sverjensky, 2010; Powner and Sutherland, 2011). For example, (Oró, 1961) described synthesis of adenine under primitive Earth conditions and (Stern et al., 2015) gives evidence for indigenous nitrogen in sedimentary and aeolian deposits, which implies that the early martian atmosphere may have been N-rich and thus favourable for the here-described synthesis as well. On the other hand, it has been suggested that HCN might have been accumulated in ferrocyanides (Sasselov et al., 2020), and formamide - due to its low volatility and sufficient stability - might have been concentrated in the hot-dry-wet cycle of craters in lagoons et similia (Miyakawa et al., 2002). Both mechanisms are well related to the typical environments of impact craters. It is expected that the residual post-impact heat supported hydrothermal activity for several hundreds thousands years (Schwenzer et al., 2012) and may have harboured a rich thermochemistry associated with the clays (Ehlmann et al., 2011; Pastorek et al., 2019) and resulting from the crater alteration or with what remains of the original material from the impactor. Importance of impact craters has suddenly been recognized when considering the well-known Gale crater on Mars, where organic molecules have been discovered by the Curiosity rover (Freissinet et al., 2015; Eigenbrode et al., 2018). Our recent investigations show that iron-rich smectites formed by hydrothermal alteration of basalts (as discovered in the case of the Nakhla ovoid structure (Chatzitheodoridis et al., 2014)) increases the thermolysis yields starting from formamide (Pastorek et al., 2019). Previous investigations conducted by Saladino, Di Mauro et al. also highlight the beneficial role of original meteoritic material (Saladino et al., 2013). Laboratory experiments have shown that those materials assist, during the formamide-based synthesis of nucleobases, amino acids and urea (Pastorek et al., 2019; Saladino et al., 2013). Some recent results are listed in Table 1, which summarises the nucleobases yields in a range of thermolysis as well as plasma experiments mimicking all stages of a meteorite impact events triggering potentially relevant prebiotic chemistry.

In the current work, detection of the formamide dimer, formamidine (5) and of 2-aminoacetonitrile (16) has allowed us to build up a unified picture also for the post-impact thermally-driven formamide-based prebiotic synthesis in craters, as shown in Figure 5B. Although the formamide dimer was previously detected by FTIR inspection of gas-phase formamide samples treated by laser spark (Ferus et al., 2012), the present study provides the first direct evidence of its origin from formamide upon heating. The heating provides energy to allow the system to pass a reaction barrier and form a dimer. The presence of a dimer has not been observed in standard formamide without any heating or the effect of a laser spark. With the aim of discerning among the possible forms of the formamide dimer, quantum-mechanical calculations of the free energy of the covalent and non-covalent (i.e., hydrogen-bonded) dimer have been conducted at the hybrid meta PW6B95-D3(BJ)/def2-QZVP level at 160°C. Such a Gibbs free energy evaluation indicates that the non-covalent formamide dimer is quite more stable than the covalent one. In fact, the non-covalently bound formamide dimer is energetically preferred against the covalent one by a ΔG = 24 kcal⋅mol⁻¹. The conversion is schematically depicted in Figure 3B. In light of these fundamental quantum-mechanical information, we could speculate that the covalent dimer has not directly been detected possibly due, on the one hand, to the low volatility and limited occurrence of this intermediate in the liquid phase during the purine synthesis and, on the other, to its own supposed high reactivity with respect to the more stable hydrogen-bonded dimer. After all, indeed, dimerization of formamide to form a covalent bond represents the initial step in the synthesis of purines in a mechanism originally proposed by Yamada (Yamada and Okamoto, 1972; Yamada et al., 1978) and then revised by Saladino et al. (2007), as summarised in (c) and depicted in Supplementary Figure S10 (Supplementary Information), which shows a specific mechanism of an intermediate synthesis finally resulting into the formation of cytosine, uracil and thymine.

The unified mechanism here proposed is ultimately confirmed by the detection—obtained via FTIR spectroscopy—of formamidine (5), an important intermediate in the synthesis of purine nucleobases such as adenine (13), hypoxanthine (15) and guanine (14), as shown in Figure 5B. Formamidine (5) is expected to result from the reaction of formamide (1) with ammonia (4). Formamidine (5) then reacts with HCN (2) to give rise to 2-iminoacetonitrile (6). Besides, it should be noted here that 2-iminoacetonitrile can be produced directly from formamide (1) by reaction with HCN (2). 2-iminoacetonitrile (6) then is able to react with HCN (2) to produce 2-aminomalononitrile (7, AMN). This latter, by combining with formamidine, leads to aminoimidazole-carboxamide (12, AICN). Further reaction of (12) with (5) is capable to synthesise adenine (13) and guanine (15). AICN (12) can also react with formamide (1) and NH₃ (4) to produce adenine (13) as an alternative. If, instead, AICN (12) reacts with formamide (1) and then water (3), hypoxanthine (15) is produced. Another possible pathway to guanine (14) passes through AICN (12), whose cyano- group is hydrolyzed to produce aminoimidazole-carboxamide (AICA, 10). Further reaction with (CN)₂ (11) yields guanine (14). AICA represents also a possible precursor to hypoxanthine (15) (Ferus et al., 2014). 2-iminoacetonitrile (6) can also react with H⁺ to give rise to 2-aminoacetonitrile (16), which was also detected in the current experiment. 2-aminoacetonitrile (16) either hydrolyses to glycine (17) or reacts with 5-aminoimidazole (19) to produce purine (20). The mentioned 2-iminoacetonitrile (6) can also be converted into HNCO (21) which, upon reaction with 4, can yield urea (22) (Pastorek et al., 2019). Pyrimidine nucleobases can also be created from formamide, as described in (Saladino et al., 2007). Here, formamide (1) reacts with 2 HCN (3) and ammonia (4) to cytosine (26). Cytosine (26) then hydrolyses to uracil (25). Further reaction with formaldehyde (42) and formic acid (46) may even lead to thymine (38) (Barone et al., 2015).

Our experiments probe a particular scenario and mechanism in the direct impact and post-impact chemistry of formamide. Direct impact radical-based synthesis can be initiated by the chemical action of an impact plasma with temperatures exceeding 5000 K on surface liquid formamide lagoons, but also for instance by reprocessing of formaldehyde (Ferus, M. et al., 2019), of reducing NH₃:CO:H₂O atmospheres (Ferus et al., 2017b), or even of hydrogen cyanide (Ferus et al., 2020). This kind of chemistry results in the one-pot origin of a few ppm of nucleobases (Ferus et al., 2015b). As reported in the current investigation, on longer timescales, e.g., a few hundred thousand years, hydrothermal environments in impact craters characterised by the residual materials of meteorites (Saladino et al., 2013) or the formation of smectites (Pastorek et al., 2019)—especially when secondary iron-rich hydrous mineral phases are formed—can support a continuous nucleobase prebiotic synthesis by means of formamide thermochemistry and direct participation of 2-aminoacetonitrile and formamidine.