3.1. Ultraviolet Bombardment of a Slightly Reducing Atmosphere
 The beginning of the process starts with simple molecules in regions of the primordial upper atmosphere. These are constantly undergoing bombardment by ultraviolet solar photons, so that the atmospheric composition is described by steady state photochemistry. When in molecular form that atmosphere is now considered to have consisted primarily of N2 (∼90%), CO2 (∼10%), and H2O (∼1%) [Kasting, 1990].
 The molecule formaldehyde may be considered one of the crucial starting points for any speculation regarding ab initio synthesis of RNA, and in fact, other important biomolecules. For example, all sugars are based on the formaldehyde monomer with the formula (CH2O)x. Since a sugar molecule is contained in every RNA monomer, polymerized formaldehyde is a component of the biopolymers RNA and DNA.
 Many investigators have studied, experimentally, the formation of HCHO from H2O and CO in the vapor phase. Focusing on the widely accepted “soup theory” of biomolecule origins, Bar-Nun and Hartman  demonstrated experimentally that formaldehyde and other organic compounds synthesized in the vapor phase from UV- irradiated H2O and CO, can be simultaneously extracted by liquid water in significant quantities.
 Investigating the possibility of organic compound formation from a weakly reducing prebiotic atmosphere, it was shown theoretically [Pinto et al., 1980] that formaldehyde (CH2O) could be formed by photolysis of H2O in an atmosphere of N2, CO2 and H2O. Pinto et al. used a “standard model” in which the “abundances of the major atmospheric gases” are N2, 0.8; H20, 0.012; CO2, 2.4 × 10−4; H2, 8 × 10−4; and CO, 2.4 × 10−7 bar. In the absence of oxygen, the following reactions take place
where M is a third body. They add that most of the HCHO thus formed is destroyed by subsequent photolysis. However, “a small fraction is incorporated into rain droplets and delivered into the oceans.” For this (standard) model the rainout rate was calculated to be 2.8 × 108 molecule cm−2 s−1. This is sufficient to fill the oceans, at their present volume, to a 10−3 M solution in 107 years. The authors point out that, according to Ponnamperuma's  results, polymerization may take place in a 3 × 10−4 M solution of formaldehyde on irradiation by UV near the wavelength 280 nm (4.44 eV). The standard model rainout rate would produce a 3 × 10−4 M solution of formaldehyde in 3 × 106 years.
3.1.2. Hydrogen Cyanide
 Another small molecule [Kasting and Brown, 1998] vital to the synthesis of RNA is hydrogen cyanide, HCN. Adenine, one of the four bases present in RNA, has the empirical formula (HCN)5. Conditions for its derivation from the proposed original atmospheric mix, however, are somewhat more difficult to arrive at than those for formaldehyde. The obvious approach is by adding methane to the hypothesized original “slightly reducing” atmospheric components of N2, CO2, H2O and trace CO and H2. The possible methane path has been studied in detail by a number of investigators. Toupance et al.  report that far UV “irradiations of CH4-NH3 mixtures give rise, in addition to hydrocarbons, to important amounts of HCN (about 0.1%) and to lesser amounts of CH3CN and C2H5CN.” Furthermore, “These results evidence that UV irradiation may contribute largely to synthesis of HCN in CH4-NH3 atmospheres and, consequently, to the synthesis of many biochemical compounds that can be derivated from HCN.” However, as was pointed out above, both methane and ammonia were constituents of a reducing atmosphere, now pretty much discarded. Neither CH4 nor NH3 are present in significant amounts in the presently favored “slightly reducing” prebiotic composition.
 Zahnle [1986b], however, considers the production of HCN from CH4 in considerable detail, including the probable presence of an otherwise weakly reducing atmosphere. He points out that absorption of solar EUV (extreme ultraviolet) by N2 provides a large source of N(4S) in the thermosphere which then descends into the mesosphere, where it immediately reacts with the CH3 formed from methane photolysis
to form HCN. He concludes that HCN could have been formed “efficiently” from CH4 in an atmosphere that is weakly reducing.
3.1.3. Source of Methane in the Prebiotic Era
 The key question now, however, becomes, where did the methane come from, since it is not adequately present in the hypothesized “slightly reducing” prebiotic atmosphere. The answer of choice would seem to be it was present in a particular location at a particular time, that is, it was present where and when it was “needed” for the occurrence of that particular step of chemical evolution. This basic notion has been used before. For example, advocates of thermal activation of inanimate matter to bring about the first living systems make use of “thermal vents” to provide activation energy for the unknown mechanism of creating the bacterial cells which they feel constituted the first living systems. Once a living system has come about at any time and location, the hypothesis is that its autocatalytic capability will, or at least can, result in its widespread dissemination.
 Consider the introduction of small amounts of methane from some source, to an atmosphere consisting primarily of CO2, N2, and H2O, with smaller amounts of CO and H2. Steady state CH4 mixing ratios of 10–100 ppm could have been maintained by a ground level methane source of order 1011 molecules cm−2 s−1. The supply of “solar UV photons” then fixes the stable rate of CH4 consumption at 4 × 1011 to 1 × 1012 molecules cm−2s−1. According to Zahnle [1986a], “This is greater than the modern abiogenic rate.” Then, from
the globally averaged flux in the thermosphere of nitrogen atoms is 1011 to 1012 atoms per cm−2 s−1 in an anaerobic atmosphere. Adsorption of solar EUV at wavelengths <102.3 nm (corresponding to an energy of 279.8 kcal/mol), would have provided a “large thermospheric source of atomic nitrogen.” The atoms then flow downward and at atmospheric density
However, methane is easily dissociated by the strong Ly-alpha line at 121.6 nm.
Now, of course, the question still remains, where did the methane, not present in our presently assumed “weakly reducing atmosphere,” come from? The problem must be solved, as pointed out above, on the basis of effective local sources. An obvious suggestion of course is the emissions of volcanoes.
 Volcanoes, as they exist in modern times, give off CO2 and H2O, but no methane. This composition reflects the same oxidative history that resulted in the presence of CO2, rather than methane, in the “slightly reduced” prebiotic atmosphere discussed above. That atmosphere, however, represents a specific stage in overall atmospheric evolution. It is chosen on the basis of the considerations that (1) the first few hundred million years of Earth history experienced too much bombardment and possible internal upheavals to have been host to any biology such as we know it; (2) the subsequent history follows the formation of the iron core, whose isolation in the Earth's center removed the main force for chemical reduction (oxygen removal) from the surface and atmosphere; and (3) therefore the beginning of life had to occur in the absence of reducing molecules such as methane and ammonia.
 The above arguments however, do not by any means preclude pockets of primeval gas trapped here and there, that are not in equilibrium with the atmosphere, during the prebiotic window of time referred to. This would provide specific sites where methane might be available for HCN synthesis. In this connection, Kasting , after considering oxidation rates of the early planetary mantle, states that it is “premature to rule out a volcanic source of highly reduced gases at the time of life's origin.”
3.1.4. Present-Day Methane Sources
 Methane released to the atmosphere today, man-made or of microbial origin, has been estimated as roughly 535 ± 125 Tg per year globally [Etiope, 2004, 2005]. In addition, however, there is a geological release that has been ignored in Intergovernmental Panel on Climate Change (IPCC) reports [Etiope, 2005]. The geological methane is usually considered to be directly or indirectly organic, thereby presumably from an original biotic source. However, there are sources of geological methane that provide strong evidence of being abiotic. One of these is present as a large component of seepage from ocean ridges [Kasting and Catling, 2003]. Recent evidence indicates that, extrapolating back to prebiotic planetary conditions, it could be enough to satisfy Zahnle's [1986b] requirement of approximately 1011 molecules cm−2 s−1 on a global basis [Kasting and Catling, 2003]. However, the requirement of our model does not necessarily require global presence of methane in the atmosphere. If the source is on land, adequate presence of methane in the region is sufficient. There are a number of known land sources yielding small amounts. However, a dramatic recent finding [Hosgormez, 2007; Hosgormez et al., 2008] has measured a land source yielding more than 25 ton per year of abiotic methane. Assigned to an area of one square kilometer containing the seepage, this amounts to 3 × 1012 molecules s−1cm−2, easily satisfying the Zahnle requirement.
 While the presence or absence today of abiotic methane sources does not prove presence or absence of these sources in prebiotic times, a finding of such a presence tends to increase the plausibility of conjecture that these things were by no means absent in the prebiotic era.
3.1.5. Cyanomethanol Hypothesis
 Assuming the presence in the atmosphere, in a more or less steady state, at a given location, of a quantity of methane capable of satisfying Zahnle's requirements for the synthesis of “important” amounts of HCN, said HCN then, according to his model, is removed from the atmosphere by dissolving in raindrops. According to our hypothesis, upon hitting the ground, the HCN mixes with similarly deposited formaldehyde. The two molecules react
to form cyanomethanol. The raindrops continue to carry cyanomethanol through the surface into the Earth. The rainwater may act as an eluting agent and the cyanomethanol as the eluent in a functioning liquid chromatography-type column. Simultaneously, the Sun is intermittently beating down on the surface. Occasionally the cyanomethanol will become completely dehydrated, while concentrated in a pocket of the column, or actually at the surface. The heat is postulated to cause a microexplosion, and the cyanomethanol decomposes with simultaneous oligomerization.
The reaction (R14) provides a natural answer to the question of why does RNA (and DNA) contain only a pentose sugar and never a hexose, although the latter is the more stable of the two. The reason implied here is that the stoichiometry is controlled by the formation of the adenine fragment. There is no stable (HCN)6, only (HCN)5. Five molecules of cyanomethanol produce adenosine, containing (HCN)5 (adenine) and (HCHO)5 (pentose) minus a water molecule.
 Adenosine, having been formed on, or deposited on, a portion of the Earth's surface containing the mineral apatite, then reacts with the apatite as follows:
where AMP is adenosine monophosphate and CaAMP is calcium adenosine monophosphate.
 To initiate the process of polymerization to convert AMP to a single base analog of RNA, two AMP's must combine to yield the dimer plus calcium hydroxide (Ca1/2(OH).
We postulate that (R15) and (R16) occur simultaneously. Continuation to form the trimer and longer polymers is then a repetition of the combined step.
 Subsequent stages of the synthesis occur with participation of the substrate, which contains or is a phosphate. At this point we note that there are three bases in addition to adenine whose syntheses are as yet unaccounted for. We assume that as the AMP polymerizes a certain amount of oxidative damage took place, caused by the substrate, or column, which is most probably a mixture of phosphate and clay. The flexible stoichiometry of these oxygen-loaded minerals should provide ample opportunity for this oxidation to take place. Accordingly, slight but increasing oxidative damage yields, in turn, guanosine, cytidine and uridine (from the bases guanine, cytosine, and uracil). Interestingly, the most heavily oxidatively damaged of the bases, uracil, is present only in RNA, being replaced by thymine in DNA.
 Needless to say, a bit of reductive damage to the ribose component of the nucleoside, caused perhaps by an oxygen vacancy in the clay-phosphate column, could result in the DNA polymer. A channel of suitable dimensions in the substrate, or column, furthermore, might promote formation of the double spiral.
3.1.6. Comparison of Rates
 Assuming the “rainout” to be rate determining, comparison of the time of 3 million years to produce a useful concentration of formaldehyde [Ponnamperuma, 1965] soup, with the time for useful concentrations of CH2(CN)OH on solid earth, yield enormous differences. Assuming adenosine forming on the surface, then polymerizing downward, the rainout could produce roughly 90 units of RNA monomer a year, for a total of 33 years to produce a polymer of three thousand RNA monomer units by penetration downward. On the other hand, if the polymerization is assumed to take place horizontally, we can have a high molecular weight polymer formed almost instantaneously. The same UV that catalyzed the polymerization as calculated by Ponnamperuma may catalyze the surface polymerization.
3.2. Enthalpies of Reaction
 Table 1 lists the enthalpies of formation from which are calculated the key enthalpies of reaction involved in transformation of simple atmospheric components to RNA. These are of course net reactions, not necessarily mechanisms, the purpose of which is to investigate the thermodynamic feasibility of the chemical evolution of original atmospheric molecules and earth minerals to RNA. Enthalpy is chosen as the measuring stick, rather than free energy, so that what is dealt with is the basic stability, i.e., chemical bonds, of the reaction products. While of course it is the free energy which determines the equilibrium constant of a given reaction, its entropy component is very much dependent on the specific conditions, which are too variable for this approximation of thermodynamic feasibility. The enthalpies, although not the exclusive determiners of reaction spontaneity, are nevertheless regarded as more useful here as measures of molecular stability.
Table 1. Enthalpies of Formation and Reaction
|A||H2O + CO2 → HCHO + O2 −57.8 −94.05 → −27.7 +0||+124.15|
|B||CH4 + N2 → HCN(g) + NH3 −17.889 +0 → +31.2 −11.04||+38.05|
|C||HCHO + HCN → CH2(CN)OH −27.7 +31.2 → −6.5||−10.0|
|D||5CH2(CN)OH → Adenosine + H2O(g) −32.5 → −156.36 −57.8||−181.7|
|E||2 adenosine + Ca10(PO4)6(OH)2 → 2CaAMP + 2Ca3(PO4)2 + 2Ca(OH)2 (−312.8)cr − 3224 → −1048 − 1972.4 −471.8||+44.6|
|F||2CaAMP → (CaAMP)dimer + Ca1/2OH −1048 → −988.3 −118.0||−58.3|
 Reactions A and B in Table 1 are first examined, and it is noted that the production of formaldehyde and hydrogen cyanide from their elements in their standard states is accompanied by an increase in enthalpy, which is contrary to thermodynamic feasibility when entropy is ignored. However, these molecules are not produced by dark reactions from elements in their standard state. Rather, they are photochemically energized as shown above in reactions (R1)–(R7). The key components involved in the syntheses are produced directly or indirectly by ionization of atmospheric molecules, producing high-energy atoms. The energy provided by absorption of upper atmosphere UV is more than adequate to ionize any of the atmospheric small molecules. For example, a widely occurring absorption maximum in the EUV (extreme ultraviolet) region for these molecules is 92 nm. This translates to an energy of 311.1 kcal/mol, whereas the strongest bond to be found among these molecules of interest is N ≡ N, that of the nitrogen molecule, with a bond energy of 170 kcal/mol.
 The meaning of this here is that the relatively high values of enthalpy of formation of the formaldehyde and hydrogen cyanide molecules, have as their original source the high-energy atomic and molecular fragments formed in the upper atmosphere by the action of solar UV radiation. This energy is then utilized for subsequent decreases in enthalpy of reactions.
3.3. Lower-Energy UV Bombardment of the Surface
 In reaction E in Table 1, the adenosine penetrates the mineral apatite to form two molecules of adenosine monophosphate, in an endothermic reaction requiring an input of +44.6 kcal/mol. The dimerization of the adenosine monophosphate in reaction F in Table 1 is exothermic, yielding −58.3 kcal/mol. It is reasonable to assume that the dimer may be formed immediately as part of the original reaction. The sum of the reactions E and F in Table 1 is then the small negative enthalpy change of −13.7 kcal, in the range of supramolecular chemistry.
 Possibly the formation of two AMP molecules does not occur simultaneously with their dimerization. Possibly, also, an activation energy is required for the reaction of adenosine with hydroxyapatite mineral. Also, the approximations used, for example, those inherent in the bond energy method, the deliberate choice of crystal enthalpy of formation in Reaction E in Table 1, and the correction for substitution of Ca for Mg, yield uncertainties which could add up to, roughly, 25 or so kcal/mol. So, the postulate that an additional photochemical infusion of energy is necessary for continuation of the chemical evolutionary process for production of RNA should be considered.
 The energy can be provided by UV activation of the adenosine under the prebiotic atmospheric conditions of roughly 3.9 GYA (billions of years ago). The relevant UV absorption maximum of adenine nucleoside [Holm et al., 2007] is 260 nm. There is no ozone layer in the prebiotic atmosphere to absorb the activating solar UV radiation. In addition, there is no oxygen to aid in destroying the nucleotide (by UV + O2) once it is formed. Also, this photochemical source of energy can be used for catalytic purposes, in addition to its role in the thermodynamics, if needed. The photon energy at 260 nm is equivalent to 110.2 kcal/mol, far more than enough to overcome the uncertainties projected in the enthalpy required for reactions E and F in Table 1.
 The presence of thermodynamic feasibility does not of course guarantee that these reactions will indeed have taken place. There must exist one or more kinetically reasonable pathways. In light of that, nothing is as convincing as successful laboratory experiments, even if they were not the ones which actually occurred in the geological past. Among the more recent investigators, Ferris and coworkers [Ferris, 2002; Wang and Ferris, 2005; Ferris, 2005, and references therein] have been particularly active in investigating pathways looking for individual steps of the overall synthesis of RNA from small molecules, experimenting with pathways for the conversion of mononucleotide RNA to the polymer final product. The path they have been concentrating on recently involves the possible role of the clay montmorillonite as catalyst for the formation of the phosphodiester bond in RNA. In their scheme the catalyst acts on a phosphate group bound to an imidazole “activating group” to improve efficiency of the reaction. Progress of this sort along the lines of specific mechanisms, workable in the modern laboratory, brings us ever closer to an understanding of what happened in the early stages of the origin of life.
 An interesting experiment which does not fit into the framework of chemical evolution described here, but nevertheless points out the existence of relevant UV catalyzed solid state reactions, has been performed recently [Senanayake and Idriss, 2006]. Formamide, adsorbed to TiO2 under ultra-high vacuum conditions and irradiated with UV, was found to yield “high-molecular-weight compounds” which may have been nucleoside bases. It was found that the reaction proceeded as a result of excitation of the semiconductor surface by the 388 nm UV radiation.
 The present work provides a number of suggestions for additional pathways of experimental investigation. Foremost is the use of the chromatographic column, or surface, technique for exploration of reactions, even without any known catalytic function. The advantages over the soup concept are enormous in terms of orders of magnitude less time required to bring the reactants together, and decreased probability of hydrolysis which would destroy the desired products. In a large volume aqueous solvent a condensation-hydrolysis equilibrium is immediately shifted toward hydrolysis since the products of the latter immediately leave the scene of reaction. The “chromatographic” substrate however, provides an enclosure which greatly minimizes that effect.
 Overall, the Urey-Miller idea of reactions of atmospheric components to form bioprecursors and ultimately biomolecules is still viable and is shown here to be thermodynamically feasible. The transfer of energy from Sun to Earth, or, similarly, from other individual stars to their planetary satellites, is a ubiquitous phenomenon that can be regarded as part of programmed cosmology. While this does not guarantee that the exact conditions for bioevolution will be (or have been) present elsewhere in the universe, it does provide evidence that bioevolution can occur within the natural laws of physical chemistry as part of overall cosmological evolution.