Over decades, scientists have been trying to understand how life as we know it today emerged from simpler molecules. From a chemical evolution perspective, the molecular complexity of life is daunting. Building on the key Miller experiment, (Miller 1953) scientists have attempted to rationalize how simple building blocks (e.g. HCN, NH3, CH4, aldehydes, etc.) have led to more complex structures such as: purines, pyrimidines, amino acids, carbohydrates and eventually to self-replicating machinery. By definition, these structures must form through intermolecular reactions, for which there is an entropic penalty. Intermolecular reactions are also inherently slow at low concentrations, which is a key problem that scientists have addressed. Indeed, reactions proceed faster at higher concentrations and this could have dictated how chemical evolution occurred. Several approaches have been proposed and validated experimentally in order to address this low concentration issue. Methods such as: eutectic freezing leading to more concentrated pockets (Sanchez et al. 1966; Miyakawa et al. 2002; Bada 2004; Cleaves 2008; Menor-Salván and Marín-Yaseli 2012), clay surfaces templating chemical reactions (Bernal 1949; Lahav and Chang 1976; Lahav et al. 1978; Cairns-Smith and Hartman 1987; Cleaves et al. 2012), aerosols (Shah 1972; Dobson et al. 2000; Tuck 2002; Donaldson et al. 2004; Ruiz-Bermejo et al. 2007;), molecular crowding (Mayer et al. 2015), and evaporating pools resulting in more concentrated aqueous solutions are generally accepted hypotheses within the scientific community (Nelson et al. 2001). However, to our knowledge catalytic reactions using simple organic molecules to induce temporary intramolecularity have not yet been explored in the context of the low concentration issue (Bols and Skrydstrup 1995; Fensterbank et al. 1997; Gauthier et al. 1998; Diederich and Stang 2000). Herein, we show that formaldehyde, a key prebiotic molecule, is able to catalyze an intermolecular reaction in water, operating solely through temporary intramolecularity.

Interestingly, formaldehyde is also a well-known crosslinking reagent, used to immobilize cells (Sutherland et al. 2008; Karmakar et al. 2015), promote the association of biomolecules (e.g. chromatin immunoprecipitation assay) (Poorey et al. 2013; Carey et al. 2009), and to form polymers (e.g. urea-formaldehyde resin) (Dunky 1998). Furthermore, its crosslinking ability is demonstrated by its industrial use in textiles, plywood, insulation, adhesives, etc. (ca. 32 million tons in 2006) (Salthammer et al. 2010). Formaldehyde is also uniquely effective as a racemization agent, used in industry preferentially when screened against an onslaught of other aldehydes (Beaver et al. 2016). Encouragingly, formaldehyde and related compounds have been shown to hasten HCN oligomerization in the prebiotic synthesis of purines (Schwartz and Goverde 1982). However, reactions in which formaldehyde acts as a catalyst are rare.

There are several examples of carbonyl groups being exploited for their ability to induce temporary intramolecularity. Relevance from a prebiotic chemistry context is Commeyras, Pascal and Taillades’ finding that temporary intramolecularity results in significantly faster hydrations of α-aminonitriles, which are precursors to α-aminoacids (105 rate accelerations; Scheme 1a). Broadly, these and other reports suggest that the intramolecularity induced by aldehydes or carbon dioxide can accelerate a variety of hydration and hydrolysis reactions (Pascal 2003; Tan 2011). Our interest in using aldehydes as catalysts arose when we noticed that simple aldehydes catalyze difficult intermolecular hydroamination reactions by exploiting the temporary intramolecularity present in a mixed aminal formed under the reaction conditions (Scheme 1b) (MacDonald et al. 2011). In this work, we also showed that chiral aldehydes can be used as efficient asymmetric catalysts, leading to preferential formation of one enantiomer of chiral allylic amines (MacDonald et al. 2013; Hesp et al. 2015). Surprisingly, the simpler catalysts also proved the most efficient ones, with formaldehyde showing the greatest catalytic activity and tolerance towards substrate scope (Guimond et al. 2012; Hesp et al. 2015). In contrast, the uncatalyzed reaction does not proceed under the same reactions conditions, and studies indicated that in the absence of the catalyst the reaction can only occur at temperatures of 80 °C and above, at high concentrations or in the absence of solvent (Zhao et al. 2012). The finding of an aldehyde catalyst structure-activity relationship clearly favoring prebiotic aldehydes recently led us to re-visit the work shown in Scheme 1a, study the catalytic efficiency of several aldehydes and develop a broadly applicable protocol using catalytic quantities of both formaldehyde and NaOH (Chitale et al. 2016).

Scheme 1
scheme 1

a Example of work exploiting temporary intramolecularity in prebiotic chemistry: Commeyras, Pascal and Taillades’ report that carbonyl compounds accelerate the hydration of α-amino nitriles. b Aldehyde induced temporary intramolecularity leads to catalysis of difficult hydroamination reactions

Full size image

While both approaches in Scheme 1 are exploiting temporary intramolecularity, they differ in that for hydration (A) the substrate benefiting from temporary intramolecularity is formed directly upon reaction between the α-amino nitrile and the carbonyl compound. In contrast for hydroamination the aldehyde pre-catalyst must successfully recruit and assemble both reaction partners in a temporary tether facilitating the desired hydroaminaton step (B). While a variety of difficult intermolecular reactions have been enabled or controlled through temporary intramolecularity, this has relied on stepwise assembly and cleavage of a temporary tether (Bols and Skrydstrup 1995; Fensterbank et al. 1997; Gauthier et al. 1998; Diederich and Stang 2000). Given the finding that formaldehyde and other aldehydes are efficient tethering catalysts (B), we were interested to see if this catalysis could operate under more dilute aqueous conditions. Arguably, this could guide further efforts directed at prebiotically relevant reactions. Herein, we show that formaldehyde is able to catalyze a difficult intermolecular reaction in water, at relevant prebiotic formaldehyde concentrations. Estimations place formaldehyde concentrations on primitive earth to be as high as 0.02 M (Taillades et al. 1998; Cleaves 2008).

The use of relevant model reactions to validate concepts of relevance in prebiotic chemistry is a well-established strategy (e.g. Soai’s systems of asymmetric amplification of chirality) (Soai et al. 1995; Blackmond 2004). This study was thus performed using our hydroamination reaction as a model reaction to explore the ability of aldehydes to act as tethering catalyst in water. This constituted a significant departure from previously reported reaction conditions, which relied on organic solvents and high concentrations (1 M in substrate, 0.05–0.1 M in aq. CH2O) to obtain high yields within 24 h at 20–50 °C. A critical issue using formaldehyde, which is present as its hydrate in water, was that it leads to the formation of two molecules of water upon aminal formation (Scheme 2). Thus under aqueous conditions, the large amount of water present could severely hinder its formation through LeChâtellier’s principle. Conversely, from an entropic perspective, this temporary intramolecularity is not disfavored: two reagents and one catalyst react to form the mixed aminal and two molecules of water (3 → 3). Interestingly, with more stable aldehydes, which do not exist as a hydrate, formation of the aminal would be entropically disfavored (3 → 2). Consequently, we embarked on a study of this intermolecular hydroamination using water-soluble reagents, with formaldehyde as a focus point.

Scheme 2
scheme 2

Temporary intramolecularity: Formation of mixed aminal with formaldehyde hydrate (a), vs normal aldehydes (b)

Full size image

We initially began by searching for a suitable concentration that would allow this investigation of the efficiency of formaldehyde as a tethering catalyst in water. Though it was possible to monitor the reactions by 1H NMR at concentrations lower than 0.1 M, we ultimately decided against this because of experimental challenges (e.g. oxidation induced by trace amounts of oxygen and slow reactions) that made it difficult to obtain conclusive data due to long reaction times. A reaction concentration of 0.1 M proved ideal: this allowed both high reproducibility while allowing to work at prebiotically relevant formaldehyde concentrations (0.02 M). Consequently, all further investigations were performed at 0.1 M using 20 mol% of formaldehyde as catalyst (0.02 M).

The formation of the hydroamination product was monitored by 1H NMR using dimethyl sulfone as a water-soluble internal standard. The experiments were repeated three times and were performed in rigorously degassedFootnote 1 aqueous solutions to ensure reproducibility; the averages of this are depicted in Fig. 1. In the presence of formaldehyde there is a 6-fold increase in the amount of measured product. A maximum yield of 66 ± 2% was observed for the case of N-methylallylamine in the presence of formaldehyde after a period of 8 days, whereas the background with no formaldehyde produced 13 ± 3% of the hydroamination product (Zhao et al. 2012 showed that this background reaction is promoted by a strong hydrogen bond, allowing this uncatalyzed hydroamination to occur). A similar result is observed for allylamine where the catalyzed reactions produced an average yield of 45 ± 10% after 8 days compared to 6 ± 6% for the uncatalyzed reactions. This result validates the hypothesis that formaldehyde is capable of tethering two molecules and catalyze difficult intermolecular reactions even at low concentrations in water.

Fig. 1
figure 1

Hydroamination reaction in the presence and absence of 20 mol% CH2O. Reactions were done with two different substrates and each trial was repeated a minimum of three times. The reaction yields were determined using dimethyl sulfone as an internal standard. Reactions contained: 2.5 equivalents of N-methylallylamine, 1 equivalent of dimethyl sulfone with respect to N-methylhydroxylamine. Reactions were done with a concentration of 0.1 M in rigorously degassed distilled water, under an atmosphere of argon. See supplemental information for a detailed procedure

Full size image

As predicted, other key prebiotic carbohydrates did not display the same catalytic efficiency as formaldehyde (Fig. 2). These results are consistent with the increased stability of these carbohydrates, and a less favourable tether formation process as shown in Scheme 2b. Indeed, the carbon chain present in all aldehydes (except formaldehyde) causes destabilizing steric interactions in the temporary tether (Scheme 2b, interaction between R’ and Me). Despite this, acetaldehyde and butyraldehyde acted as modest tethering catalysts, suggesting that these aldehydes are able to induce termporary intramolecularity over extended periods of time. Unexpectedly, glycolaldehyde and 1,3-dihydroxyacetone actually inhibited the reaction, which could be due to stabilization of a symmetrical aminal intermediate [e.g. 1: 2 glycolaldehyde: hydroxylamine adduct, which has been shown to be a resting state in related studies (Guimond et al. 2012)] or due to the presence of stable dimers (Fratzke 1985; Kua et al. 2013) that could limit the ability of these carbohydrates to act as tethering catalysts. In contrast, other aldoses and ketoses showed similar yields as the control experiment. Overall, these results show that formaldehyde is a superior tethering catalyst and provide calibration on the ability of simple aldehydes and carbohydrates to induce temporary intramolecularity at low concentrations.

Fig. 2
figure 2

Screening of other aldehydes and carbohydrates for catalytic activity. The reaction yields were determined after 8 days using dimethyl sulfone as an internal standard. Reactions contained: 2.5 equivalents of N-methylallylamine, 1 equivalent of dimethyl sulfone, and 20 mol% equivalents of catalyst with respect to N-methylhydroxylamine. Reactions were done with a concentration of 0.1 M in rigorously degassed distilled water, under an atmosphere of argon at room temperature. See the supplemental information for a detailed procedure and NMR yields after 1, 3 and 5 days

Full size image

It is generally accepted that the pH of prebiotic oceans were in the range of 5–11, and most commonly the pH is argued to have resembled the modern value of pH ≈ 8 (Cleaves 2008). When examining the effect of pH on the reaction, it was noticed that under acidic conditions (pH 5), the reaction did not proceed for either the catalyzed reaction (20 mol% CH2O) or the background reaction. This is logical seeing as how under acidic conditions, protonated N-methylallylamine is less likely to form the mixed aminal species in which temporary intramolecularity induces formation of the expected hydroamination product. However under buffered neutral conditions, the reaction proceeded normally, providing 65% of the hydroamination product after 8 days. Interestingly, the background (no CH2O) was substantially diminished (2% after 8 days) under neutral conditions, suggesting that the hydrogen bonding required for this background reactivity (Zhao et al. 2012) is disrupted under buffered conditions. When the reaction was buffered to a pH of 10, there was no change in reactivity; the catalyzed reaction and the background performed the same as in distilled water (59, and 9%, respectively). This similar result is in line with the observation that the pH of the non-buffered reaction in distilled water remains at 10 throughout the experiment. Experimental details of this pH study can be found in the supplemental information.

Overall, this study shows that the hydrate of formaldehyde is able to act as a tethering catalyst in water, and accelerate a difficult intermolecular reaction by induced intramolecularity. While the catalytic activity is modest, turnover was observed. Thus, this data suggests that select small molecules could accelerate slow intermolecular prebiotic reactions; a strategy complementary to other templating approaches based on non-covalent interactions that have been reported to accelerate reactions (e.g. the ligation of short oligonucleotides, Jain et al. 2004; Cafferty et al. 2016). Additional studies on the catalytic activity of simple carbohydrates in reactions of prebiotic interest are ongoing (Chitale et al. 2016), and will be reported in due course.