
==== Front
Heliyon
Heliyon
Heliyon
2405-8440
Elsevier

S2405-8440(24)12258-X
10.1016/j.heliyon.2024.e36227
e36227
Research Article
Alkaline saline lakes: A chemical evolution experiment evaluating the stability of formaldehyde in an aqueous saline environment
Fuentes-Carreón Claudio Alejandro claudio.fuentes@ciencias.unam.mx
ab⁎
Meléndez-López Adriana Leticia ac
Cruz-Castañeda Jorge Armando a
Negrón-Mendoza Alicia negron@nucleares.unam.mx
a⁎⁎
a Instituto de Ciencias Nucleares, Universidad Nacional Autónoma de México, Cd. Universitaria, México City, Mexico
b Universidad Nacional Autónoma de México, Cd. Universitaria, México City, Mexico
c Escuela Nacional de Ciencias de la Tierra, Universidad Nacional Autónoma de México, Cd. Universitaria, México City, Mexico
⁎ Corresponding author. Instituto de Ciencias Nucleares, Universidad Nacional Autónoma de México, Cd. Universitaria, México City, Mexico. claudio.fuentes@ciencias.unam.mx
⁎⁎ Corresponding author. negron@nucleares.unam.mx
13 8 2024
30 8 2024
13 8 2024
10 16 e362274 6 2024
9 8 2024
12 8 2024
© 2024 The Authors
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Formaldehyde condensation in the presence of a mineral catalyst and under alkaline conditions is considered to be a "messy" chemical system due to its dependence on the complex chemical equilibrium between the reaction intermediates, which has a significant impact on the final products.

This chemical system is extremely important in prebiotic chemistry and has been proposed as a potential pathway for carbohydrate formation in the early Earth. Saline and soda lakes are alkaline systems that could concentrate and accumulate a wide variety of ions (such as phosphate) and clay minerals, which can catalyze prebiotic chemical reactions. These geological environments have recently been suggested as ideal environments in which prebiotic chemical reactions could have occurred.

This study uses Lake Alchichica in Mexico as a physicochemical analog of an early Archean saline lake to examine the stability of formaldehyde in these aqueous saline environments. Formaldehyde decomposes into sugar-like and CHO molecules in alkaline, high-salinity environments depending on the minerals phases present. As phosphate ion (HPO42−) is available in the aqueous medium, the results of our experiments also imply that phosphorylation processes may have occurred in these natural settings.

Keywords

Prebiotic chemistry
Aldehydes
Clay minerals
Chemical evolution
==== Body
pmc1 Introduction

The term “chemical evolution” refers to the suite of hypotheses that attempt to explain how the first components of life (i.e., biomolecules) were abiotically generated from multiple organic compounds. This idea, proposed independently by Oparin and Haldane in the 1920s [1], was verified experimentally using the spark discharge experiment designed by S. Miller [2]. Following this, several prebiotic chemical pathways have been proposed in an attempt to explain the formation of the aforementioned biomolecule precursors.

Chemical evolution experiments (i.e., prebiotic chemistry studies) involve the simulation of primitive Earth environments, such as the primeval ocean and ancient hydrothermal systems. Many of these environments share some common characteristics, such as the presence of an aqueous medium and a mineral phase. These factors play an important role in prebiotic chemical reactions, either as catalysts or as initiators of organic synthesis [3]. One clear example is the formose reaction [4], which involves the condensation of formaldehyde into aldehydes and carbohydrates under basic conditions [5]. This reaction requires alkaline conditions, heat, and the presence of carbonate minerals, which function as inorganic catalysts. These carbonate minerals can also influence the stability of the formed sugars, allowing for the formation of complex compounds [6]. Additionally, the active surfaces of some mineral phases, such as clays or iron hydroxide minerals, can further promote these aldehyde condensation reactions [7]. However, it should be noted that specific sugars, such as ribose or glucose, are merely by-products of these reactions; the main condensation products of these reactions comprise a complex mixture of straight-chain and cyclic pentoses and hexoses [8]. Therefore, studies on the reaction mechanisms of these prebiotic reactions in simulated early Earth environments are of great importance, as they are necessary to correctly describe how the building blocks of life could have emerged on the early Earth.

Aldehydes are of particular interest in the field of chemical evolution as they play a wide variety of roles in abiotic synthesis reactions. Formaldehyde can yield sugars when heated under alkaline conditions in the presence of a mineral catalyst [8]. Acetaldehyde has been proposed as a source of alanine [9], while multiple intermediaries of the formose reaction, such as pyruvaldehyde, glyceraldehyde, and glyoxal, have been shown to play an important role in carbohydrate formation. Besides their ability to form sugars, the phosphorylation of aldehydes is of particular interest, as several phosphate compounds, such as glyceraldehyde 3-phosphate, are essential metabolites in multiple metabolic pathways [10]. Consequently, the study of how these compounds could have formed in prebiotic environments has become a central topic in modern chemical evolution studies.

As previously mentioned, solid surfaces may have played an important role in the chemical evolution of minerals by serving as catalysts in the synthesis of organic compounds [3]. Of the mineral surfaces readily available on the early Earth, clay minerals have been highlighted as a crucial component for surface chemical processes. These minerals can selectively concentrate specific organic molecules. Also, the adsorption of organic molecules onto mineral surfaces can protect them from degradation by ionizing radiation [11]. Mineral surfaces can also catalyze the polymerization of organic molecules [12]. Clays have been ubiquitous mineral phases ever since their formation during planetary accretion [13]. Consequently, clays could have played an important role in the prebiotic synthesis of organic compounds due to their abundance on the early Earth.

However, phosphorylation reactions highlight the difficulties with prebiotic chemical reactions in aqueous mediums, where the formation of the final product is not thermodynamically favored. As phosphorylation is a dehydration reaction (i.e., releases water), this process could not have occurred readily in an aqueous medium, as any reaction products would rapidly hydrolyze. Furthermore, dissolved phosphate is relatively rare in natural environments [14]. Consequently, phosphate is not expected to have existed in high concentrations on the early Earth because its concentration is buffered by the dissolution of phosphate minerals [15].

In light of this “phosphate problem”, alternative mechanisms have been proposed for the phosphorylation reactions and the occurrence of phosphate concentrations in natural environments. One novel idea proposes that carbonate-rich saline lakes could have been a natural environment in which up to 1 mol L−1 of phosphate could have accumulated [16]. Specifically, the precipitation of phosphate with calcium to form low-solubility apatite minerals would have been inhibited by the sequestration of Ca2+ into carbonate minerals, increasing the availability of dissolved phosphate in the medium. Therefore, carbonate-rich saline lakes could have been an important geological environment in which sugar-forming prebiotic reactions could have occurred, especially due to their favorable physicochemical characteristics, such as their high pH values (>9) as well as the abundance of dissolved metallic cations (Ca2+, Na+, Mg2+, etc.), which have been shown to catalyze formose-like reaction of aldehydes [5,6,8]. As clays are a common mineral phase that are present in lacustrine sediments [17]; these minerals could have further catalyzed sugar-forming reactions.

1.1 Lake Alchichica as an analog of an ancient saline carbonate-rich lake

Lake Alchichica is a crater lake located in the Cuenca de Oriental in central Mexico. The crater is a maar formed by a phreatomagmatic eruption and is surrounded by tuff deposits [18]. Alchichica is a closed basin lake, with no surface water tributaries or outlets. The water balance is maintained by precipitation and evaporation. The major ions present in the lake waters include Na+, Cl−, Mg2+, and SO42−, with K+, Ca2+, Li+, Br−, and F− existing as minor constituents. The hardness of the lake waters has been measured to be 18.72 mmol L−1, which indicates that the lacustrine system is naturally saturated in CaCO3 [19]. Phosphate can be found in concentrations of up to 5.16 μmol L−1 [20]. A few studies have been conducted on the mineralogical composition of the lake sediments [21]. reports that the sediment comprises mainly carbonate minerals and silt deposits (predominantly clays). The geochemical characteristics of this crater lake make it an ideal analog for ancient saline lakes (Fig. 1). These ancient aqueous surface environments could have been formed due to the chemical weathering of volcanic rocks in a CO2-rich atmosphere [16]. The high amounts of volcanic and tectonic activity could have formed lake basins due to the continuous generation of young crust [22], which, in turn, would be filled by the surface water enriched in weathering products. As Lake Alchichica is a saline lake rich in calcium carbonates, this environment is an ideal candidate for the accumulation of phosphate in accordance with previously proposed models [16]. Furthermore, the alkaline pH of the lake waters as well as its ionic content represent chemical characteristics that would have been ideal for the base-catalyzed aldolic condensation reactions of aldehydes and ketones dissolved in the medium. This work aims to study the stability of formaldehyde and its decomposition products in an aqueous saline environment that is analogous to an ancient saline lake; it will also focus on the presence of clay minerals in these environments and how they could have influenced the stability of the aldehyde in the system.Fig. 1 Schematic representation of an ancient saline lake. The atmospheric composition denoted in the illustration follows the model proposed by Ref. [23]. These aqueous environments may have been abundant during the Archean (∼3.9 Ga) in crustal zones with a high incidence of tectonic activity, which would have resulted in the formation of basins in which saline water could have accumulated.

Fig. 1

2 Materials and methods

2.1 Reagents

DL-glyceraldehyde (90 %) paraformaldehyde (99.9 %), glyoxal (40 wt % in H2O), glycolaldehyde (>90 %), D-glucose (99.9 %), D-ribose (99.9 %), Sucrose (99.9 %), NaOH (99 %), HCl (37 wt % in H2O), D2O (99.8 %), NaCl, MgSO4 •7H2O, and Na2HPO4 • 7H2O were obtained from Sigma-Aldrich®, St. Louis, MO, USA. Methanol-free formaldehyde was prepared from paraformaldehyde according to the method described by Ref. [24]. Acetonitrile (HPLC), ethanol (HPLC), H2SO4 (95–97 %), α-D-Glucose 1-phosphate disodium salt hydrate (97 %), and 2,4-dinitrophenylhydrazine (DNPH) (97 %) were obtained from Merck Co.®, Kenilworth, NJ, USA. Ca-Montmorillonite (STx-1; Gonzales County, Texas, USA) and Na-Montmorillonite (SWy-2; Crook County, Wyoming, USA) were obtained from the Source Clay Minerals Repository, Purdue University, IN, USA. All the reagents were of the highest available commercial purity.

2.2 Preparation of samples

To simulate the physicochemical conditions of Lake Alchichica, saline water with an ionic content analogous to the Alchichica system was prepared. This analog consisted of an aqueous solution of 0.02 mol L−1 NaCl, 1 × 10−3 mol L−1 MgSO4 •7H2O, and 0.05 mol L−1 Na2HPO4 •7H2O. These salt concentrations were used to simulate the major ions present in the waters of Lake Alchichica [20], in which the total water column concentration of Na+, Cl−, Mg2+, and SO42− were reported to be approximately 100, 87, 35, and 16 meq L−1, respectively [20]. An exception was made with MgSO4 •7H2O, whose concentration was increased to increase the solubility of the salt at the desired volumes (100 mL). The total ion content, pH, ionic force, and temperature of the saline water samples are presented in Table 1. This water was used for the preparation of concentrated (1.5 mol L−1) aqueous methanol-free formaldehyde solutions. The samples were degasified with Ar to remove any dissolved O2. The final pH of the formaldehyde standard solution in saline water was 9.13.Table 1 Physicochemical characteristics of simulated Lake Alchichica saline water.

Table 1Ion	Concentration (mol L−1)	Temperature (K)	System pH	Ionic Force (I)	
Na+	0.12	298	9.2	0.173	
Cl−	0.02				
Mg2+	1 × 10−3				
SO42-	1 × 10−3				
HPO42-	0.05				

2.3 Sorption experiments

100 mg of clay mineral phases were mixed with 5 mL aliquots of the degasified formaldehyde standard solution dissolved in saline water in 15-mL polyallomer centrifuge tubes (Beckman Coulter®, Brea, CA, USA) to simulate the interaction between formaldehyde and clays in a saline, alkaline environment. Two different sets of clay suspensions inside polyallomer tubes were done, using different clays in each one (Na-Montmorillonite (SWy-2) and Ca-Montmorillonite (STx-1)) The polyallomer tubes with clay suspension were saturated with Ar to create an anoxic atmosphere and the polyallomer tubes were sealed to preserve the atmospheric integrity of the system.

Once prepared in the polyallomer tubes, the clay-formaldehyde mixtures were subjected to constant agitation. The minerals were separated from the aqueous phase using an Allegra XL-90 centrifuge (Beckman Coulter®, Brea, CA, USA) at 25,000 rpm at room temperature for 15 min at defined time intervals (1, 24, 48, 168, and 336 h). The mineral powder was desiccated and preserved for further analysis. The collected supernatants were filtered using 12-μm Acrodisc syringe filters (Whatman®, Chicago, IL, USA) and stored at 0 °C for the subsequent High pressure liquid chromatography (HPLC) measurements. The formation of new organic compounds in the sample supernatants was monitored by HPLC; analyzing each sample after the previously described sorption time intervals elapsed. Following this analysis, the aqueous phase was evaporated under a constant air flow until the dissolved organic compounds precipitated. The precipitate—a white powder—was stored in a vacuum desiccator for 24 h to remove excess water in the samples and stored for further analysis.

2.4 Carbonyl-containing compound measurements

The aldehydes were identified in the form of their 2,4-dinitrophenylhydrazine (DNPH) derivatives using HPLC–UV. The carbonyl compounds were derivatized by reacting 4 mL aliquots of the sample supernatants with 4 mL aliquots of the DNPH reagent (0.4 mg DNPH dissolved in 2 mL H2SO4, 3 mL H2O, and 25 mL of ethanol) for 12 h. The carbonyl derivatives precipitated as yellow-orange crystals, which were filtered, dried, recrystallized, and redissolved in acetonitrile for subsequent analysis. The DNPH derivatives of formaldehyde, glyoxal, pyruvaldehyde, and DL-glyceraldehyde were prepared and used as standards for the detection of carbonyl compounds in the experimental samples.

2.5 Sample analysis: sorption experiments

2.5.1 Raman spectroscopy analysis

Raman spectroscopy was used to identify the chemical structure of the organic standards and the reaction products. The organic solids precipitated from the experimental samples were pressed between a NaCl pellet and their Raman spectra were collected using an Optosky ATR 3000 portable Raman spectrometer. The Raman probe, which utilized a Class IIIB laser, was positioned at an operating distance of 6 mm from the sample. The laser was maintained at a constant power of 400 mW to avoid damage to the sample. The spectra were recorded in the interval of 3000 to 150 cm−1 with a resolution of ±5 cm−1.

2.5.2 IR-ATR spectroscopy

The infrared spectra of the precipitated organic solids (a dry organic powder) as well as the clays before and after sorption (as a dry mineral powder) were collected using attenuated total reflectance–Fourier transform infrared (ATR-FTIR) spectroscopy. Spectra was collected with a PerkinElmer® Spectrum 100-FTIR-ATR spectrometer, coupled to a Horizontal Attenuated Total Reflectance (HATR) accessory with a reflection element consisting of a ZnSe crystal (PerkinElmer®). The angle of incidence into the crystal cell was 45°. Prior to each batch sample, a background spectrum of the environment and empty cell was collected. The spectra were recorded in the interval of 4000 to 650 cm−1 with a resolution of 4 cm−1 using 25 scans per sample.

2.5.3 Analysis of remnant phosphate after sorption with STx-1 and SWy-2 clays

The remnant ion phosphate (HPO42−) in solution was determined by conducting an acid-base titration of the aqueous supernatant of the samples after sorption with STx-1 and SWy-2 clays. 0.01 mol L−1 NaOH was used as a titrant, which allowed us to quantify the formation of Na2HPO4 in the system, which possessed an equivalence point of pH 9.4. The full description of the method used can be found in Ref. [25]. The pH of the sample supernatants after sorption was 7.8 and 8.12 for the STx-1 and SWy-2 samples, respectively. Consequently, all samples were acidified with 0.1 mol L−1 HCl to a pH of 6.0. NaOH was subsequently added to 5 mL aliquots of the supernatants, with pH constantly monitored until the equivalence point was reached. The concentration of phosphate at the equivalence point was calculated based on the total titrant volume. This procedure was also conducted on the original saline solutions to determine the phosphate concentration in the system before sorption. The tritation curve of the experimental samples was monitored using a ThermoScientific® VersaStar Pro pH meter coupled to an Orion ROSS 8157 UWMMD pH/ATC triode electrode.

2.5.4 HPLC analysis

A Waters® ACQUITY UPLC evaporative light scattering (ELS) detector coupled to a Waters® 600E multisolvent delivery system was used to characterize the sugars and CHO compounds present in the samples using their retention times. The compounds were separated in a GL Sciences® Inerstil NH2 5 μm column (4.6 × 250 mm; Torrance, CA, USA) using isocratic elution at 1.0 mL/min with a mobile phase comprising 60 % acetonitrile and 40 % water.

An HPLC coupled to an electrospray ionization-mass spectroscopy (ESI-MS) detector was used to analyze the m/z of the detected sugar-like compounds. This analysis was conducted using a Waters® 515 HPLC pump coupled to a Waters® SQ-2 Single Quadrupole Mass Detector system, with electrospray ionization in negative (ESI-) mode. The cone voltage was 20 V, the capillary voltage was 2.55 kV, and the desolvation temperature was 350 °C. The sugar-like compounds were separated using the aforementioned GL Sciences® Inerstil NH2 5 μm column (4.6 × 250 mm) using isocratic elution at 0.8 mL/min with a mobile phase comprising of 60 % acetonitrile and 40 % water; this method was specifically designed for the separation of carbohydrates.

A Knauer®, Berlin, Germany Azura P 4.1s HPLC pump equipped with a Knauer® Smartline 2300 refraction index (RI) detector was used to analyze the phosphorylated compounds. The stationary phase was an Alltech® Wescan 10 μm Anion Exclusion column (7.5 × 300 mm; Illinois, USA); the procedure utilized isocratic elution at 0.5 mL/min with a mobile phase of 1.5 mmol L−1 H2SO4 in 10 % acetonitrile. All chromatographic measurements were done at 588 mm Hg pressure and 25 °C.

2.5.5 31P Nuclear magnetic resonance (31P NMR)

31P NMR spectra of the precipitated organic solids formed after sorption with SWy-2, and STx-1 clays was recorded using a Bruker®, Fällanden, Switzerland 400 MHz Avance III HD spectrometer. 5 mg of the precipitated organic solids were dissolved in 0.7 mL of D2O, and then transferred to 5 mm tubes to collect the 31P NMR spectra. The NMR measurements were done at 588 mm Hg and 25 °C.

2.6 Sample analysis: carbonyl-containing compounds

2.6.1 HPLC-UV

HPLC-UV was used to analyze the DNPH derivatives of the carbonyl-containing compounds. A Varian® 9010 Solvent Delivery System HPLC pump (California, USA) coupled to a Varian® 9060 Variable Wavelength UV–Vis detector was used. The derivatives of the carbonyl compounds were monitored by measuring their absorbance at 350 nm. The derivatives were separated in a SUPELCO® Supelcosil LC-18 C-18 column (250 × 4.6 mm) using isocratic elution at 1.0 mL/min with a mobile phase comprising 70 % acetonitrile and 30 % water. The results were collected using an HP® 3396A Integrator (California, USA).

2.7 Analysis of chemical equilibria in the aqueous solution

Following the equilibration of salts in the system, the total ionic concentration and saturation index (SI) of all possible solids in solution were modeled using the Chemical Equilibrium Modeling System MINEQL + Version 5.0 computer program, which has been distributed as free software.

MINEQL+ is designed for the modelling of chemical equilibria and uses a numerical engine to solve mass balance equations using equilibrium constants. The theoretical considerations and equations that underpin the MINEQL engine are discussed in detail in Ref. [26]. The equilibrium constant values were referenced from the USEPA MINTEQA2 thermodynamic database [27]. The following values were used for the initial concentration of ionic species in solution: 0.12 mol L−1 Na+, 0.02 mol L−1 Cl+, 1 × 10−3 mol L−1 Mg2+, and 1 × 10−3 mol L−1 SO42−. Phosphate was introduced into the system in the form of the HPO42− ion (0.05 mol L−1). All equilibrium equations were solved using a pH of 9.13, an ionic force (I) of 0.174 (calculated using the Debye-Hückel Davies equation [28]), a temperature of 288 K, and a pCO2 of 3.5 atm. The temperature and pCO2 values were selected based on the environmental conditions used for the sorption experiments. The theoretical calculations were repeated using a pCO2 of 0.3 bar and a temperature of 288 K to simulate the equilibria expected under early Archean environmental conditions [29].

3 Results

3.1 Formaldehyde and clay analysis

The IR pattern of both Na-montmorillonite (SWy-2; Fig. 2) and Ca-montmorillonite (STx-1; Fig. 3) exhibit vibrational bands associated with Si–O bonds at 797 and 697 cm−1 for SWy-2 and 797 and 693 cm−1 for STx-1. Vibrational bands associated with OH groups were present in both clays (Fig. 2, Fig. 3a). Based on the literature, the 883 cm−1 and 863 cm−1 bands in SWy-2 (Fig. 2b) were attributed to the vibration of AlFe3+OH and AlMgOH, respectively [30], while the 843 cm−1 band in STx-1 (Fig. 3b) was interpreted to be due to the vibration of MgOH groups.Fig. 2 IR spectrum of the natural Na-montmorillonite (SWy-2) samples. a) 4000–2000 cm−1 region. b) 2000–650 cm−1 region.

Fig. 2

Fig. 3 IR spectrum of the natural Ca-montmorillonite (STx-1) samples. a) 4000–2000 cm−1 region. b) 2000–650 cm−1 region.

Fig. 3

The IR spectrum of paraformaldehyde (the precursor from which the methanol-free formaldehyde was prepared) exhibits vibrational bands associated with sp3 carbons at 2981, 2921, and 1468 cm−1 (Fig. 4). The 1088 cm−1 vibration band is associated with the asymmetric C–O stretch. These observations suggest that the sample consists of pure crystalline paraformaldehyde. A weak band around 3000 cm−1 was also observed: These vibrations could be associated with OH groups present in the sample. However, the lack of the C–OH stretch vibration band suggests that the main C–O bond on the organic compound is an ether bond.Fig. 4 IR spectrum of paraformaldehyde powder.

Fig. 4

3.2 Chemical equilibrium of the simulated Lake Alchichica saline Water

The stability of formaldehyde in an aqueous saline environment, such as the chemical analog of an ancient saline lake, is influenced by the complex equilibrium in the saline solution. Consequently, it is important to consider the relative concentration of species in such solutions.

We determined the total ion concentration of the dissolved species in the simulated Lake Alchichica saline solution at 25 °C in an O2-free atmosphere with a pCO2 of 3.5 atm. The total concentration of non-precipitated ionic species in the saline water solution is presented in Table 2. The amount of free phosphate (PO43−) in solution was less than 0.05 mol L−1 due to the presence of additional chemical species that removed this ion from the medium. Due to the high Na concentration of the waters (0.12 mol L−1), NaHPO4− was one of the major phosphate species present in solution, accounting for up to 26 % of the total P in the system (Table 3). Soluble phosphate in the form of HPO42− was present at concentrations of 3.56 × 10−2 mol L−1, representing 71.3 % of the total P. The discrepancy in the concentrations of the phosphate ions in the calculated experimental values can be attributed to dispersion in the pH measurements; specifically, pH electrodes lose precision at high Na+ concentrations due to the saturation of the exchange membrane. An analysis of the SI of all possible solids in solutions reveals the complex equilibrium of the potential mineral phases. Positive SI values are associated with mineral phases that are likely to precipitate in the system, while negative values are associated with mineral species that are presented but are undersaturated and unlikely to precipitate. At the present temperature, pH, and chemical conditions, we found that the only mineral species in equilibrium with the ion concentration was Mg3(PO4)2 (Table 4). It should be noted that an SI of 0 suggests an unstable equilibrium: changes in the concentration of dissolved Mg2+ could favor the precipitation of this mineral phase.Table 2 Total concentration of dissolved ionic species in the simulated Lake Alchichica saline water.

Table 2Ion	Concentration (mol L−1)	
[Na+]	0.12	
[Cl−]	0.02	
[Mg2+]	5.68 × 10−4	
[SO42−]	1 × 10−3	
[PO43−]	4.97 × 10−2	
[CO32−]	1.12 × 10−2	

Table 3 Chemical species derived from PO42− in the simulated Lake Alchichica saline water.

Table 3Chemical species	Concentration (mol L−1)	Total %	
PO43-	8.979 × 10−5	0.18	
MgH2PO4+	4.252 × 10−7	0.00085	
MgHPO4(aq)	3.535 × 10−4	0.707	
NaHPO4−	1.342 × 10−2	26.8	
H2PO4−	1.708 × 10−4	0.342	
HPO4−2	3.566 × 10−2	71.3	
H3PO4	1.322 × 10−11	0	
MgPO4−	1.935 × 10−5	0.0387	
Mg3(PO4)2 (s)	1.439 × 10−4	0.576	
Total P	0.04971	100	

Table 4 Saturation indices of all possible mineral phases in the simulated Lake Alchichica saline water.

Table 4Mineral phase	SI	
Hydromagnesite (Mg5(CO3)4(OH)2•4H2O)	−7.956a	
Artinite (Mg2(CO3)(OH)2•3H2O)	−3.301a	
Mg(OH)2	−4.821a	
Periclase (MgO)	−7.611a	
Brucite (Mg(OH)2)	−2.871a	
MgHPO4•3H2O	−0.452a	
Nesquehonite (MgCO3•3H2O)	−3.004a	
Natron (Na2CO3•10H2O)	−4.289a	
Epsomite (MgSO4•7H2O)	−5.747a	
Mirabilite (NaSO4•10H2O)	−4.686a	
Halite (NaCl)	−4.58a	
Magnesite (MgCO3)	−0.214a	
Mg3(PO4)2	0	
Thenardite (NaSO4)	−6.122a	
a Negative values indicate the solid is undersaturated in the aqueous system and is thus unable to precipitate.

To investigate how the chemical equilibrium of the system would change in a CO2-rich atmosphere, we repeated the previous calculations using a pCO2 of 0.3 bar and a temperature of 288 K, which are the partial pressures and global average temperatures characteristic of the early Archean [29]. The total concentration of dissolved ions under these conditions is shown in Table 5. The amount of soluble phosphate was found to have increased by almost 30 % compared to the equilibrium concentrations in the system at modern pCO2 values, increasing the availability of this ion in saline waters (Table 6). The SI analysis also highlighted the rather complex equilibrium between the mineral phases (Table 7). Magnesite was the only mineral species in equilibrium with the system, likely due to the increase in the atmospheric pressure of CO2 (∼1000 times more than CO2 preindustrial values).Table 5 Total concentration of dissolved ionic species in the simulated Lake Alchichica saline water under early Archean conditions.

Table 5Ion	Concentration (mol L−1)	
[Na+]	0.12	
[Cl−]	0.02	
[Mg2+]	2.58 × 10−5	
[SO42−]	1 × 10−3	
[PO43−]	0.05	
[CO32−]	1.07 × 10−2	

Table 6 Chemical species derived from PO42− in the simulated Lake Alchichica saline water under early Archean conditions.

Table 6Chemical species	Concentration (mol L−1)	Total %	
PO43-	9.63 × 10−5	0.19	
MgH2PO4+	7.08 × 10−10	0	
MgHPO4(aq)	5.77 × 10−7	0.0011	
NaHPO4−	1.25 × 10−3	2.5	
H2PO4−	2.4 × 10−4	0.491	
HPO4−2	4.8 × 10−2	96.8	
H3PO4	1.71 × 10−11	0	
MgPO4−	2.51 × 10−8	0	
Total P	0.05	100	

Table 7 Saturation indices of all possible mineral phases in the simulated Lake Alchichica saline water under early Archean conditions.

Table 7Mineral phase	SI	
Hydromagnesite (Mg5(CO3)4(OH)2•4H2O)	−11.770a	
Artinite (Mg2(CO3)(OH)2•3H2O)	−6.795a	
Mg(OH)2	−7.676a	
Periclase (MgO)	−11.385a	
Brucite (Mg(OH)2)	−6.419a	
MgHPO4•3H2O	−3.256a	
Nesquehonite (MgCO3•3H2O)	−3.059a	
Natron (Na2CO3•10H2O)	−3.099a	
Epsomite (MgSO4•7H2O)	−8.459a	
Mirabilite (NaSO4•10H2O)	−6.288a	
Halite (NaCl)	−5.592a	
Magnesite (MgCO3)	0	
Mg3(PO4)2	−8.465a	
Thenardite (NaSO4)	−8.263a	
a Negative values indicate the solid is undersaturated in the aqueous system and is thus unable to precipitate.

3.3 Stability of formaldehyde in simulated saline aqueous media

The remaining amount of formaldehyde in the experimental samples was determined by measuring the absorbance of formaldehyde–DNPH at 350 nm as a function of the sorption time, using the Beer–Lambert law to calculate the concentration of the compound in solution. The calibration curve was constructed using formaldehyde–DNPH standards with concentrations ranging between 1 × 10−2 to 1 × 10−5 mol L−1. Besides these determinations, HPLC-UV analysis shows that formaldehyde is labile under the experimental conditions. Formaldehyde decomposes in the presence of clays (Fig. 5). The total percentage of remnant formaldehyde was found to decrease proportionally to sorption time. The decomposition rate of the compound was independent of the type of clay present in the system, though there was a slight increase in decomposition time when exposed to SWy-2 clay: at the maximum sorption time, 94.16 % formaldehyde had decomposed when exposed to STx-1 compared to 98.01 % when exposed to SWy-2. An exhaustive analysis of the aqueous supernatants by HPLC-ESI-MS as well as the organic solids precipitated after the evaporation of the organic sample yielded additional chemical information about the formaldehyde decomposition products.Fig. 5 Decomposition curve of formaldehyde under the experimental conditions as a function of sorption time.

Fig. 5

The precipitated organic solids included contributions from decomposition products, remnants of paraformaldehyde, and traces of Na2HPO4 salt. The IR spectrum of the precipitated organic solids formed from sorption onto SWy-2 exhibited vibrational bands associated with sp3 carbons (2894, 2922 and 1474 cm−1) and C–OH stretching (1084 and 1065 cm−1), as well as a vibrational band at 1622 cm−1 (Fig. 6). The precipitated organic solids formed from sorption onto STx-1 exhibited vibrational bands associated with sp3 and sp2 carbons as well as C–O bonds (Fig. 7); this contrasts with the products identified in SWy-2, in which the region corresponding to C–O bond vibration was attributed to ether groups.Fig. 6 The IR spectrum of the precipitated organic solids formed after sorption onto SWy-2 clay.

Fig. 6

Fig. 7 The IR spectrum of the precipitated organic solids formed after sorption onto STx-1 clay.

Fig. 7

Raman spectra yielded additional information about the functional groups present in the samples derived from sorption onto the STx-1 and SWy-2 clays (Fig. 8, Fig. 9, respectively). The most notable vibrational bands were found at 1090, 1075, 1052, 910, and 865 cm−1, which correspond to C–O and C–H bond vibrations. The vibrational band at 1600 cm−1, which corresponds to sp2 carbon (C=C), is present in both systems. The samples derived from sorption onto SWy-2 also exhibited a vibrational band at 1700 cm−1 associated with carbonyl bonds (C=O). This vibrational band was notably absent in the samples derived from sorption onto STx-1.Fig. 8 Raman spectrum of the precipitated organic solids formed after sorption onto the clays used in this study (1150–800 cm−1 region).

Fig. 8

Fig. 9 Raman spectrum of the precipitated organic solids formed after sorption onto the clays used in this study (3000–1150 cm−1 region).

Fig. 9

Due to the presence of HPO42− ion in the saline solution, additional IR and Raman analysis were done, aimed to study the interaction of this ion with the organic compounds present in solution. Specifically, the formation of organophosphates was monitored. Diagnostic peaks of P–O stretching modes typically appear between 800 and 1300 cm−1; both samples exhibited vibrational bands at 1087, 930, and 862 cm−1, which could correspond to P–O stretching modes of organophosphates (Fig. 10) [31]. Therefore, the band at approximately 975 cm−1 can be assigned to the antisymmetric C–O–P stretching vibration of phosphorylated organic compounds [32,33] (Fig. 9).Fig. 10 The IR spectrum of the precipitated organic solids formed after sorption onto SWy-2 and STx-1 clays in the 1200–800 cm−1 region compared to α-D-glucose 1-phosphate standard. The 1088, 930, and 862 cm−1 bands are associated with the vibrational frequencies of phosphate groups [31].

Fig. 10

HPLC-ESI-MS was used to detect sugar-like compounds in the sample supernatants, which were analyzed following sorption onto SWy-2 (Fig. 12) and STx-1 (Fig. 11, Fig. 12). The retention times of standard carbohydrates are shown in Table 8. Both samples were monitored using the Total Ion Count (TIC) mode with negative ionization. The elution times of standard carbohydrates were used as a reference for the identification of detected compounds (Table 2). The TIC chromatogram exhibited two signals at retention times of 9.2 and 16.4 min; neither of these times matched any of the standard carbohydrates. Furthermore, each signal exhibited multiple m/z values (Fig. 11, Fig. 12), suggesting the fragmentation of the separated organic compounds.Fig. 11 a) Total ion count (TIC) HPLC-ESI-MS chromatogram of the sugar-like compounds detected in the sample supernatants following formaldehyde sorption onto SWy-2 in the simulated ancient saline lake solutions.

Fig. 11

Fig. 12 a) Total ion count (TIC) HPLC-ESI-MS chromatogram of the sugar-like compounds detected in the sample supernatants following formaldehyde sorption onto STx-1 in the simulated ancient saline lake solutions.

Fig. 12

Table 8 Elution times of standard carbohydrates in a GL Sciences® Inertsil NH2 5 μm Column.

Table 8Compound	Elution time (min)	
D-Glucose (C6H12O6)	9.8 ± 0.03	
D-Ribose (C5H10O5)	7.97 ± 0.08	
Sucrose (C12H22O12)	11.37 ± 0.2	

In addition to these sugar-like compounds, aldehydes were detected by HPLC-UV (Fig. 13a and 13b). The chromatograms of the derivatized aqueous supernatants exhibited several peaks that were identified by comparing their retention times to standard carbonyl–DNPH derivatives that were calibrated under the experimental conditions used in this study (Table 9). The aldehydes observed after sorption onto SWy-2 and STx-1 were identified as glycolaldehyde and glyoxal (Fig. 13c and 13d). The maximum concentration of these aldehydes in solution was reached at 336 h, although their formation was detected even at the lowest sorption times.Fig. 13 HPLC chromatogram of sample supernatants derivatized with DNPH at different sorption times. a) Sample supernatant after 60 min of sorption onto STx-1. b) Sample supernatant after 60 min of sorption onto SWy-2. c) Sample supernatant after 336 h of sorption onto STx-1. d) Sample supernatant after 336 h of sorption onto SWy-2. Signals at retention times of 3.5, 4.18, and 7.6 min correspond to glycolaldehyde-DNPH, formaldehyde-DNPH, and glyoxal-diDNPH, respectively (Table 9). The small peak at 4.18 min corresponds to a glycolaldehyde impurity.

Fig. 13

Table 9 Elution times of standard carbonyl-DNPH derivatives in a SUPELCO® Supelcosil LC C-18 Column.

Table 9Compound	Elution time (min)	
Glyceraldehyde-DNPH	3.0	
Glycolaldehyde-DNPH	3.4	
Formaldehyde-DNPH	4.16 ± 0.3	
Glyoxal-diDNPH	7.623 ± 0.7	

3.4 Clays in simulated ancient saline aqueous media

Following the sorption experiments, the Ca- and Na-montmorillonite (referred to as STx-1 and SWy-2, respectively) were analyzed using IR-ATR spectroscopy. An analysis of the 4000–2000 cm−1 and 2000–650 cm−1 regions revealed the presence of additional vibrational bands that were not originally present in the clay standards (Fig. 14). These bands were attributed to the vibrations of sp3 carbons (Fig. 14a and b). The Si–OH vibrational band was notably deformed in the clay samples following sorption, which implies the presence of organic C–OH bonds (Fig. 14a and c). This data suggests that organic compounds were sorbed into the surface of the SWy-2 clays. In contrast, the spectra of the STx-1 clay before and after sorption did not exhibit any notable differences, suggesting that no organic matter sorbed onto these samples (Fig. 14c and d).Fig. 14 IR spectra of the SWy-2 and STx-1 clay samples after sorption. Red circles indicate changes in the IR spectra compared to the original standards (black line). a) 4000–2000 cm−1 region of the SWy-2 sample. b) 2000–650 cm−1 region of the SWy-2 sample. c) 4000–2000 cm−1 region of the STx-1 sample. d) 2000–650 cm−1 region of the STx-1 sample.

Fig. 14

3.5 Quantification of the remaining phosphate in solution

The total amount of phosphate in solution following sorption was calculated using an acid-base titration with 0.01 mol L−1 NaOH (Fig. 13). The initial phosphate concentration before sorption was calculated by titrating the initial formaldehyde saline solutions with the same reagent: this value was found to be 0.040 mol L−1, representing the total amount of H2PO42− available in the system for reaction with NaOH (additional phosphate species, formed by equilibrium with Mg and Na ions, were not accounted for in this study). The total percentage of phosphate (HPO42−) in solution diminishes proportionally to sorption time, up to a maximum of 94 % and 80 % when exposed to SWy-2 and STx-1, respectively. At a pH of 6.0, the initial value used for the titration of the sample, the total concentration of phosphate ion (HPO42−) in our experimental system was 0.040 mol L−1; i.e., final phosphate concentrations in solution were 0.037 and 0.032 mol L−1 for the systems exposed to SWy-2 and STx-1, respectively. The final concentration of phosphate in solution was reached at 48 h and remained stable up to the maximum sorption time (336 h)

4 Discussion

Formaldehyde and the formose reaction are chemical systems that are widely studied in prebiotic chemistry [6,[34], [35], [36]]. Most, if not all, of these studies focus on the role that formaldehyde plays as the main monomer in the aldol condensation reaction, which ends with the production of carbohydrates. The novelty of the present work is the insights that our experiments provide on the behavior of aldehydes in a simulated ancient saline system (specifically, alkaline ‘soda’ lakes). Soda lakes are considered to be promising settings for the occurrence of prebiotic reactions as they allow for the concentration of major elements as well as anions of biological relevance [37]. These lakes have been proposed as ideal environments in which phosphate could accumulate due to their concentration in water by the precipitation of its main sequestrant, Ca2+, in the form of carbonates [16]. Lake Alchichica was selected as an analog of these ancient alkaline lake systems, as the hydrogeochemistry of the Alchichica system allows for the concentration of phosphate in a manner similar to the chemical equilibrium proposed in the cited references.

This was validated by studying the chemical equilibrium of the proposed experimental system. At modern temperature and CO2 concentrations (25 °C and pCO2 = 3.5 atm, respectively), the proposed hydrogeochemical model could accumulate up to 3.56 × 10−2 mol L−1 of phosphate in the form of HPO42−. However, this accumulation was dependent on two factors: 1) the absence of microorganisms that could consume phosphate as nutrients [20], and 2) the absence of chemical cations with strong affinities for phosphate. The first condition is readily met in a prebiotic environment, as these systems would have been devoid of life. However, the second condition depends entirely on the chemistry of the system, as the presence of ionic species that precipitate alongside phosphate, such as Ca2+, Mg2+, CO32−, SO42−, and F−, could alter the precipitation equilibria of the systems [38,39]. The SI of all the possible mineral species under the Mg2+ and SO42− concentrations of the simulated system can help determine if phosphate minerals are likely to precipitate. Table 4 shows that Mg3(PO4)2 had an SI value of 0 under these conditions, indicating that it is in equilibrium with the system (i.e., near saturation).

Modifying the temperature and CO2 concentrations of the simulated saline waters to match the conditions proposed for the early Archean [29] resulted in changes in the phosphate and Mg2+ concentrations in the system (Table 5): dissolved phosphate concentrations were found to increase, while the concentration of Mg2+ decreased. This was due to the increase in the saturation of dissolved carbonate in the system as a result of the increase in pCO2 due to the abundance of atmospheric CO2. An analysis of the SI of the early Archean system reveals that magnesite (MgCO3) was in equilibrium with these saline waters. Thus, the presence of a CO2-rich atmosphere could have increased the chemical weathering rates of silicates, which, in turn, would have increased the amount of magnesium in solution. Although magnesite is the most thermodynamically stable form of magnesium carbonate, its precipitation is hindered at low ambient temperatures [40]. In contrast, the solubility of calcium carbonates would have been lower in highly alkaline systems. However, a high pCO2 would have rapidly increased the precipitation rate of calcium carbonates, increasing calcite saturation in the system [41]. Consequently, the near saturation of magnesite in the system is primarily due to the absence of dissolved Ca2+, which was likely sequestered by the precipitation of calcium carbonates.

In the context of phosphate sequestration, the high affinity of magnesium for carbonate [40,41] would have favored the precipitation of magnesium carbonates over magnesium phosphate (Table 7), especially due to the high concentration of dissolved carbonate. In the simulated Archean alkaline water, the carbonate concentration was found to be the main control on phosphate concentration. Although magnesium weathering (and mobility) would have been higher, phosphate would still have accumulated because the precipitation of magnesium carbonates is thermodynamically favored, especially at increased pCO2 values. Indeed, despite the near equilibrium of Mg3(PO4)2 in modern simulated saline water, an increase in dissolved carbonate could lead to an increase in phosphate concentration in the system. This phenomenon has already been observed in Lake Alchichica, which possesses waters saturated in calcite, dolomite, and magnesite due to the high pCO2 values in the environment. Phosphate is only undersaturated in the medium due to its consumption by local microorganisms [20]: this would not have been the case in a lifeless, early Archean environment in which only inorganic chemical species would have been capable of sequestering phosphate.

Formaldehyde was found to be labile under these experimental conditions, readily decomposing when exposed to SWy-2 and STx-1 clays (Fig. 5). The decomposition rate of the formaldehyde was independent of the type of clay. Previous studies have reported on the role that mineral surfaces play in the aldolic condensation of aldehydes; specifically, the mineral surfaces allow for the binding of the OH groups of diols, enols, and enolates of aldehydes in solution onto clays under acidic and alkaline pHs [6,7,42,43]. Our experimental system was expected to exhibit similar behavior. An exhaustive analysis of the IR spectra of the SWy-2 and STx-1 clays before and after sorption yielded additional information about the chemical species bound to their mineral surfaces. Before sorption (Fig. 2, Fig. 3), both clays exhibited bands associated with silica and surface OH groups (693, 797, 1633, and 3700–3400 cm−1). A Fe3+ vibration band was detected (885 cm−1) in the SWy-2 clay while vibrational bands corresponding to Al–OH (913 cm−1) and Mg–OH (843 cm−1) were detected in the STx-1 clay. An analysis of the chemical composition of both clays revealed that the primary difference between the SWy-2 and STx-1 clays was their Fe content [44], with SWy-2 possessing a maximum iron content of 4.37 % (in the form of Fe2O3) while STx-1 possessed a total Fe2O3 content of 1.20 %. The iron content of SWy-2 was thus high enough for the IR vibration band to overprint the Al–OH and Mg–OH bands. Following sorption, new IR bands appeared on the SWy-2 samples. Specifically, two absorption bands in the 2800–3000 cm−1 region, with an additional band appearing at approx. 700 cm−1 after 168 and 336 h of sorption with formaldehyde (Fig. 14). These vibration bands are associated with sp3 carbons, which suggests that the decomposition products of formaldehyde had been bound to the surface of the clay. However, the mechanisms behind this binding process remain unknown. One potential mechanism for the sorption process is the binding of newly formed aldehyde enediolates, such as glycolaldehyde (Fig. 13), to Fe3+ cations [7], which, as shown earlier, are known to be present in higher concentrations in the SWy-2 clay compared to STx-1. However, further work is needed to correctly determine the mechanisms by which organic compounds are bound to the SWy-2 clay.

An exhaustive analysis of the IR and Raman spectra of the organic solids precipitated following the evaporation of water from the experimental supernatants yielded useful information on the chemistry of the formaldehyde decomposition products (Fig. 6, Fig. 7). Both IR spectra indicated the presence of C–OH, CH3, and CH2 functional groups. The IR spectra of the precipitated solids obtained after sorption onto STx-1 revealed the presence of a vibrational band associated with C–O stretching, in contrast, the IR spectra of the sample exposed to SWy-2 exhibited bands associated with the C–OH stretching of secondary alcohols. These results indicated that the precipitated solids from both samples were primarily composed of a mixture of molecules containing R–OH and C–O–C functional groups. Raman spectroscopy and ESI analysis were conducted to further investigate the structure of these compounds.

The Raman spectra of the samples exhibited notable differences compared to the paraformaldehyde standard used for sample preparation (Fig. 8, Fig. 9); specifically, they contained evidence of vibrational bands associated with the vibrational modes of C–C–O, C–O C, C=O, C–C–C, and C–H bonds. By analyzing the Raman spectra of commercial analytical grade D-Glucose (Fig. 15, Fig. 16), we attributed the detected vibrational bands to the presence of glucose isomers such as α or β-glucose (Fig. 16a and c) [45]. The vibrational bands between 910 and 848 cm−1 (Fig. 16b) are often associated with glucose isomers. Most of the bands associated with the vibrational modes of glucose functional groups are present in the precipitated solids of both samples (Fig. 17). In the 1600-1200 cm−1 region (Fig. 17a), the CH2, δ(O–C–H), δ(C–C–H), and υ(C–O) stretching modes exhibited similar spectra profiles and wavenumber values to the glucose standard as well as to values reported in the literature [45]. Bands at 848 and 916 cm−1 (Fig. 17b) are commonly associated with υ(C–C) and δ(C1–H1) vibrations in α-glucose, while these bands are shifted to 868 and 916 cm−1 for β-glucose [45]. A shift from 848 cm−1 to 865 cm−1 was observed in both precipitates, consistent with the υ(C–C) vibration of β-glucose. However, the vibration band associated with the υ(C–H) vibration (894 cm−1) was notably absent, suggesting that the precipitated solids were neither α- nor β-glucose, but instead a complex mixture of sugar-like cyclic and linear polyalcohols and ethers with chemical structures similar to those of glucose. This was further supported by the absence of the endocyclic δ(C–C–O) vibration at 450 cm−1 (Fig. 17c), implying the presence of linear pentoses and hexoses in the sample. The lack of this band, as well as the low intensity of the vibration bands in the 1200–1000 cm−1 region, where the signals of the δ(C–O–C) angle bending mode typically appear, further suggests the absence of cyclic sugar-like compounds.Fig. 15 Remaining phosphate in solution after sorption with SWy-2 and STx-1 clays.

Fig. 15

Fig. 16 The Raman spectrum of glucose standards. a) 1500–1225 cm−1 region; b) 1200–800 cm−1 region; c) 800–150 cm−1 region.

Fig. 16

Fig. 17 Raman spectrum of precipitated solids after sorption with SWy-2 and STx-1 clays. a) 1600–1200 cm−1 region; b) 1200–800 cm−1 region; c) 800–150 cm−1 region.

Fig. 17

A new band at approximately 975 cm−1 appeared after sorption (Fig. 18). This band was not present in the Raman spectrum of formaldehyde before sorption but was observed in the α-D-glucose 1-phosphate spectrum. This band was thus identified as being associated with the antisymmetric C–O–P stretching vibration of phosphorylated organic compounds [32,33]. An analysis of the IR spectrum of the samples revealed more information about the nature of the phosphate groups present in the experimental samples. As previously mentioned, diagnostic phosphate peaks typically appear between 1300 and 800 cm−1 [31]. These peaks are shown in a non-aqueous physical mixture of glucose with disodium hydrogen phosphate heptahydrate, where the unreacted phosphate peak values are shown (Fig. 19b). The spectrum of disodium hydrogen phosphate heptahydrate (Fig. 19c) highlights the presence of monoprotonated phosphate (HPO42−) at 1050 and 869 cm−1, which can also be found in the SWy-2 and STx-1 organic precipitates (Fig. 19d and e). These bands were shifted to higher (1088 cm−1) and lower (845 cm−1) values, respectively, compared to the vibration bands present in the α-D-glucose 1-phosphate spectrum (Fig. 19a), which also exhibited a band of 930 cm−1, associated with a separate phosphate vibration mode. The absence of the 930 cm−1 signal in the pure disodium hydrogen phosphate heptahydrate spectrum suggests this is a vibration band unique to organically bound phosphate. Organophosphate bands could also be observed in the SWy-2 and STx-1 organic precipitates, indicating that these samples likely contained some amount of phosphorylated compounds as well as a significant quantity of unreacted disodium phosphate. The exact structure of these compounds is unknown; however, we can assume that they are composed of a mixture of both linear and cyclic sugar-like compounds based on the Raman analysis. If phosphorus-containing compounds were indeed formed, this suggests that the decrease in the dissolved phosphate concentration was due to phosphorylation reactions (Fig. 15). In this case, the maximum concentration in solution would have been 3 × 10−3 and 8 × 10−3 mol L−1 for the SWy-2 and STx-1 clays, respectively. It should also be assumed that not all of the available phosphate would have reacted with the available organic material, as some of it could have been sorbed by the clays. However, the spectroscopic analysis of the clay samples did not reveal the presence of any phosphate groups bound to the mineral surface, suggesting that most of the phosphate must have reacted with the available organic compounds in solution.Fig. 18 Comparison between Raman spectra in the 1125–800 cm−1 region, in which the vibrational band characteristic of υ(C–O–P) appears.

Fig. 18

Fig. 19 IR spectra (1200–750 cm−1 region) of precipitated organic solids after sorption. a) IR spectra of α-D-glucose 1-phosphate. b) IR spectra of the homogenized mixture. c) IR spectra of disodium phosphate heptahydrate. d) IR spectra of precipitated organic solid after sorption onto SWy-2 clay. e) IR spectra of precipitated organic solid after sorption onto STx-1.

Fig. 19

As a complementary analysis, possible phosphorylated compounds in the experimental samples were identified using HPLC-RI (refraction index) and 31P NMR. The chromatogram of the supernatants from the experimental samples after 336 h of sorption with SWy-2 and STx-1 clays (Fig. 20c and d, respectively). Two different compounds with a retention time of 8.5 and 10.1 min were identified: the retention time of the first compound matched that of α-D-glucose phosphate, a phosphorylated standard (Fig. 20a), indicating that this compound was a phosphorylated molecule with chemical characteristics that closely resembled that of this standard. This analysis also confirmed the presence of remnant HPO42− (Fig. 20b) that had been previously detected by IR spectroscopy. When a 1:1 mixture of the phosphorylated standard and the experimental samples was analyzed through co-injection (Fig. 20e and f), the intensity of the compound increased, while its peak shape and retention time remained constant for both clay samples. 31P Nuclear magnetic resonance (31P NMR) of the precipitated organic solids (Fig. 21) confirms the presence of one phosphorylated compound in each sample analyzed by this technique. The chemical shifts at 2.5 (Figs. 21a) and 2.2 (Fig. 21b) ppm suggests that the phosphorylated compound detected could be a phosphate ester [[46], [47], [48], [49]].Fig. 20 HPLC chromatogram of phosphorylated compounds. a) α-D-glucose 1-phosphate. b) Na2HPO4 7H2O salt. c) Supernatant of samples after sorption onto SWy-2. d) Supernatant of samples after sorption onto STx-1. e) Co-injection of phosphorylated standard (α-D-glucose 1-phosphate) with supernatant of samples after sorption onto SWy-2. f) Co-injection of phosphorylated standard (α-D-glucose 1-phosphate) with supernatant of samples after sorption onto STx-1.

Fig. 20

Fig. 21 31P NMR spectra of precipitated organic solids after sorption with a) SWy-2 clay and b) STx-1 clay. A single phosphate peak in each sample is located at 2.5 and 2.2 ppm, which are associated with the presence of a single phosphate ester in the precipitated organic solids.

Fig. 21

Electrospray ionization was used on the supernatants obtained from the SWy-2 (Fig. 11) and STX-1 (Fig. 12) samples to correctly determine the molecular weight and likely chemical formula of the decomposition products. Both chromatograms revealed the presence of two different compounds with retention times of 9.21 and 16.35 min. The product ion mass spectrum of [M − H]- (derived from formaldehyde) for each chromatographic peak revealed a rather complex ion spectrum. The analysis of the sugars and sugar-like compounds using negative ion ionization modes revealed a significant amount of fragmentation products [50]. Monosaccharides and oligosaccharides tend to fragment when analyzed without derivatization [51]. IR and Raman spectra show that the main functional groups of the decomposition products are C–O–C, C–OH, C–C, C=O, and C=C. Consequently, we can infer that the product ion mass spectrum is composed primarily of polyethers and polyalcohols with similar polarities that are not easily separated using silica-bonded NH2 HPLC columns, as well as the fragmentation products of large linear or cyclic sugar-like molecules.

The most abundant ions [M − H]- detected at 9.2 and 16.3 min in the SWy-2 and STx-1 samples, as well as their proposed chemical formulas, are presented in Table 10, Table 11, respectively. It should be noted that the proposed chemical formulas are not definitive, as the complex nature of the samples as well as the presence of multiple isomers means that the m/z of the observed ions can be described by more than one chemical formula. When analyzed by electrospray ionization in negative mode, sugars (monosaccharides and polysaccharides) fragment into ions that are characterized by the loss of H2O (m/z 18) and CH2O (m/z 30) [[50], [51], [52]]. Consequently, we can assume similar fragmentation patterns and chemical behavior in sugar-like compounds (including both cyclic and linear molecules) as they are isomers of sugars. We propose that the ions at m/z 371 and 185, which were detected in both samples at 16.3 min, are the result of the loss of H2O and CH2O from C13H26O13. m/z 371 corresponds to the loss of a single H2O fragment, while m/z 185 corresponds to the loss of two CH2O fragments and one H2O fragment. The ion at m/z 233 could be a fragment resulting from the loss of two H2O fragments from C9H18O9. These ions eluted at high retention times, suggesting that they are the products of cyclic or linear sugar-like compounds of high molecular weight since it is well-known that monosaccharides are rapidly eluted in silica-bonded NH2 columns [53]. Ions with m/z < 340 could be the products of several different polyalcohol or polyether compounds (i.e., CHO compounds). These compounds are expected to elute first as they are not strongly retained in NH2 columns and are thus more abundant at lower retention times.Table 10 m/z of the most abundant ions in the samples sorbed onto SWy-2 clays.

Table 10RT 9.2 min	RT 16.35 min	
Observed m/z ([M − H]-)	Molecular weight (g mol−1)	Probable chemical formula	Observed m/z ([M − H]-)	Molecular weight (g mol−1)	Probable chemical formula	
255	256	C9H20O8	75	76	C3H8O2	
265	266	C12H26O6	193	194	C7H14O6	
325	326	C14H30O8	371	372	C13H24O12	
339	340	C15H32O8				

Table 11 m/z of the most abundant ions in the samples sorbed onto STx-1 clays.

Table 11RT 9.2 min	RT 16.04 min	
Observed m/z ([M − H]-)	Molecular weight (g mol−1)	Probable chemical formula	Observed m/z ([M − H]-)	Molecular weight (g mol−1)	Probable chemical formula	
255	256	C9H20O8	75	76	C3H8O2	
325	326	C14H30O8	175	176	C7H12O5	
			233	234	C9H14O7	
			371	372	C13H24O12	

HPLC-UV analysis revealed the presence of both glycolaldehyde (C2H4O2) and glyoxal (C2H2O2) in solution after sorption onto both clay species. Both molecules were present in relatively low concentrations (compared to formaldehyde) after the minimum amount of sorption time (60 min); a sharp increase in concentration was observed after 336 h. The detection of both aldehydes suggests that the main mechanism by which these sugar-like and CHO compounds were formed could be base-catalyzed aldol condensation [54] via formaldehyde polymerization. Indeed, glycolaldehyde has been characterized as an important intermediary in this type of reaction [55]; consequently, its detection in the experimental samples supports aldol condensation as the primary mechanism by which CHO compounds are formed. Glycoaldehyde (and other aldehyde intermediates) is initially produced by formaldehyde polymerization (Fig. 13a and b) and consumed until almost all of the initial formaldehyde is transformed into CHO and sugar-like compounds (Fig. 13c and d). Although glycolaldehyde is principally an intermediary, it remains in solution due to its stability; furthermore, its resonant form, glyoxal, reduces its reactivity to enol addition. The detailed mechanisms of the aldol condensation reaction can be found in multiple studies [6,34,36,43]. Scheme 1 presents a theoretical reaction mechanism through which sugar compounds could form in this experimental system under these conditions. In this proposed scheme, formaldehyde condenses into glycolaldehyde which subsequently condenses into glyceraldehyde. The keto-enol equilibrium favors retro-aldol condensation, resulting in the formation of sugar molecules: the sugar-like compounds observed in this study could have followed a similar formation pathway. In addition, the alkaline conditions could have catalyzed the condensation of formaldehyde-diol (CH2(OH)2) into the CHO compounds [56] as shown in Scheme 2.Scheme 1 Formation of sugar-like compounds.

Scheme 1

Scheme 2 Formation of CHO compounds.

Scheme 2

4.1 Implications for prebiotic chemistry

The analysis of the decomposition products of formaldehyde at pH 9.13, 298 K, and in solution with Na+, Cl−, HPO42− rich waters with no dissolved O2 revealed that the primarily chemical species formed are composed of CHO-containing compounds and sugar-like molecules. The rate of formation of these compounds is slow, requiring up to 24 h to synthesize detectable organic compounds. After 60 min of sorption, the remnant formaldehyde in solution is approximately 50 %. Following this initial period, the decomposition rate slows significantly, reaching a maximum of 90 % at 336 h. Reactions in the presence of SWy-2 and STx-1 clay resulted in the same products. There was no discernible difference between the reaction products formed across the different clay species; this was expected as both clays possessed similar chemical compositions.

Although the type of clay did not influence the type of CHO compounds and sugar-like molecules formed in the simulated saline waters, it did affect the speed of formation of these chemical species. Fig. 22 shows the detection of sugar-like compounds by HPLC-ELSD on the experimental samples. When a solution of formaldehyde and saline water was left to react without clays, no chemical species were detected (Fig. 22a). In contrast, sugar-like compounds were detected when formaldehyde was dissolved in MiliQ grade water and allowed to adsorb onto the SWy-2 and STx-1 clays (Fig. 22b and c). Additional sugar-like and CHO chemical species were observed when clay was added to the system (Fig. 22d and e). This data suggests that the clays and the dissolved ions in solution act as catalysts for aldol condensation reactions. Under alkaline conditions, both the OH− ion and the R–OH mineral groups can act as catalysts for formaldehyde by deprotonating the α-carbon and favoring the condensation of formaldehyde units. Divalent cations can also act as catalysts [43]. The combination of both divalent cations (Ca2+, Mg2+) and the hydroxyl groups on the mineral surfaces of the clays would thus have catalyzed aldol condensation reactions. The apparent sorption of organics onto the SWy-2 clays (and lack thereof, in the case of STx-1) suggests that hydroxy groups catalyze the condensation reaction when formaldehyde binds to SWy-2 clay. Since STx-1 is slightly richer in divalent cations, condensation could have occurred in the aqueous medium. However, it is important to note that although aldol condensation reactions could have occurred, the formation of biologically relevant sugars, such as ribose or pentose, was not detected. None of the ions observed in the ESI-MS spectra were associated with the known m/z of biologically relevant sugars (e.g., 149 and 179). This is because the formation of ribose and glucose by aldol condensation reactions requires extremely high initial concentrations of formaldehyde as well as high-temperature conditions. Even then, without the presence of specific mineral phases—such as borate and molybdate (which inhibit the retro-aldol degradation of formed sugars [6])— these important biomolecules are unlikely to accumulate in an experimental system. Consequently, the formation and accumulation of ribose and pentose seems unlikely in our laboratory-based simulated ancient saline water samples, especially when clays as the only mineral phase present.Fig. 22 HPLC chromatogram of sample supernatants analyzed after sorption onto SWy-2 and STx-1 clays. a) Formaldehyde in saline water after 336 h. b) Formaldehyde dissolved in MiliQ water after sorption onto SWy-2. c) Formaldehyde dissolved in MiliQ water after sorption onto STx-1. d) Formaldehyde dissolved in saline water after sorption onto SWy-2. e) Formaldehyde dissolved in saline water after sorption onto STx-1. The compound at a retention time of 5 min represents an impurity present in MiliQ water.

Fig. 22

Our theoretical calculations show that HPO42− is freely available to react in the system. When the pCO2 values increase to levels proposed for the early Earth, the phosphate ion is capable of accumulating in ancient saline lakes that have physicochemical characteristics similar to those found in our simulated system. The spectroscopic and chromatographic analysis of the samples suggests that phosphorylated organic compounds could have formed in this experimental system. However, the mechanism behind their formation remains elusive. Previous studies have suggested that systems with low water activity (aw < 1) may drive the phosphorylation of organic compounds [15,57,58]. aw diminishes when solute concentrations increase [59]. When a mixture of salts is present, the total loss of aw in the system is equal to the sum of the loss of water activity that can be attributed to each solute in the system [60]. Our experimental system contains multiple solutes (e.g., Na+, Mg2+, Cl−, SO42−, HPO42− and CH2O), with the most highly concentrated solutes being Na+ (0.12 mol L−1) and CH2O (1.5. mol L−1) [61]. reported that solutions with Na and organic solutes (sucrose/D-glucose) at concentrations close to those reported in our system result in aw values of 0.997 and 0.975, respectively. Consequently, our system is expected to have an aw value of <1, which could be the main factor that drives phosphorylation in early Archean waters. Alternatively, water evaporation could be the main driver of phosphorylation. As water is removed from the system, the equilibrium of the phosphorylation reactions is displaced towards its products in response to the removal of water. Water evaporation could occur in saline lakes, as their hydrological regime is dependent on cycles of evaporation and precipitation [38]. If phosphorylation reactions are truly occurring in our system, this would suggest that ancient saline lakes could not only be environments in which phosphate could have accumulated but also systems in which the high solute concentration would have driven the sequestration of available phosphate by organic compounds, if any were available.

5 Conclusion

The condensation of formaldehyde under alkaline conditions is often considered to be a “messy” system as it results in the formation of a wide variety of sugars and sugar isomers. It is also affected by several factors, including temperature, the presence of specific mineral phases, solute concentrations, and pH [55]. Our work shows that the following phenomena occur in a system that closely resembles the physicochemical conditions that can be encountered in a natural ancient environment (i.e., an alkaline saline lake).1. Formaldehyde readily decomposes in the presence of Ca and Na montmorillonite into sugar-like and CHO-containing compounds that are formed by base-catalyzed aldol reactions and formaldehyde-methanediol condensation, respectively.

2. The complex nature of the aldol condensation reactions does not allow for the formation of sugars of biological importance (e.g., ribose, glucose) under these experimental conditions. Although clays could catalyze the decomposition and condensation of formaldehyde, they do not allow for the “chemical selection” of those sugars. Consequently, in a theoretical, primitive environment whose characteristics match those of the proposed experimental system, sugar formation and accumulation appear to be unlikely.

3. Our analysis (31P NMR, Raman spectroscopy, IR spectroscopy and HPLC) suggests that phosphorylation reactions could have occurred in ancient saline lakes. These reactions could have occurred due to the low aw of the environment (due to high salinity), water evaporation, or both.

4. Lake Alchichica can be considered an analog of ancient saline lakes as its chemical and mineralogical composition allows for the accumulation of phosphate, especially at atmospheric and temperature conditions that match those associated with the early Archean (3.8 Ga). Saline and soda lakes are interesting and novel environments in which prebiotic reactions could have occurred in the early Earth.

Ethical statement

This work does not involve the use or animal of human subjects.

Funding

This work was supported by CONACHyT fellowship 319818 and PAPIIT grant IN114122.

Data availability statement

The data generated and analyzed during the current study is included in the body of this paper. Additional data, if required, will be made available upon request.

CRediT authorship contribution statement

Claudio Alejandro Fuentes-Carreón: Writing – review & editing, Writing – original draft, Visualization, Methodology, Formal analysis, Conceptualization. Adriana Leticia Meléndez-López: Writing – review & editing, Methodology, Investigation, Formal analysis, Data curation. Jorge Armando Cruz-Castañeda: Writing – review & editing, Writing – original draft, Formal analysis, Data curation. Alicia Negrón-Mendoza: Writing – review & editing, Supervision, Resources, Project administration, Investigation, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors acknowledge the support of the Instituto de Ciencias Nucleares at the Universidad Nacional Autónoma de México (UNAM) and Posgrado en Ciencias de la Tierra. The authors acknowledge the funding CONACHyT fellowship 319818 and PAPIIT grant IN114122 . The authors wish to thank M. Sc. Gerardo Cedillo Valverde for his technical assistance. The authors are grateful to C. Claudia Camargo Raya for her technical assistance in the Laboratorio de Evolución Química, ICN-UNAM; Dr. Karina Elizabeth Cervantes de la Cruz, Dr. Alejandro Heredia Barbero, Dr. Maria del Carmen Virginia Ortega Alfaro and Dr. Saul Alberto Villafañe Barajas for their invaluable academic assistance and guidance throughout the experimental process. We are also grateful to Mr. José Rangel Gutierrez, Mr. Martín Cruz, and Mr. Enrique, M.Sc. for their contributions. Finally, the authors are grateful to the reviewers involved in the revision process, as their suggestions helped improved the quality of the present work.
==== Refs
References

1 Yamagishi A. Kakegawa T. Usui T. Astrobiology: from the Origins of Life to the Search for Extraterrestrial Intelligence 2019 Springer Singapore Singapore 10.1007/978-981-13-3639-3
2 Miller S.L. A production of amino acids under possible primitive earth conditions Science 117 1953 528 529 10.1126/science.117.3046.528 13056598
3 Dalai P. Kaddour H. Sahai N. Incubating life: prebiotic sources of organics for the origin of life ELEMENTS 12 2016 401 406 10.2113/gselements.12.6.401
4 Butlerow A. Bildung einer zuckerartigen Substanz durch Synthese Justus Liebigs Ann. Chem. 120 1861 295 298 10.1002/jlac.18611200308
5 Botta O. Bada J.L. Extraterrestirial organic compounds in meteorites Surv. Geophys. 23 2002 411 467 10.1023/A:1020139302770
6 Kim H.-J. Ricardo A. Illangkoon H.I. Kim M.J. Carrigan M.A. Frye F. Benner S.A. Synthesis of carbohydrates in mineral-guided prebiotic cycles J. Am. Chem. Soc. 133 2011 9457 9468 10.1021/ja201769f 21553892
7 Weber A.L. Prebiotic sugar synthesis: hexose and hydroxy acid synthesis from glyceraldehyde catalyzed by iron(III) hydroxide oxide J. Mol. Evol. 35 1992 1 6 10.1007/BF00160255 11536502
8 Cleaves H.J. Prebiotic chemistry: what we know What We Don’t, Evo Edu Outreach 5 2012 342 360 10.1007/s12052-012-0443-9
9 Miller S.L. Cleaves H.J. Prebiotic chemistry on the primitive earth Rigoutsos I. Stephanopoulos G. Systems Biology 2006 Oxford University PressNew York, NY 3 56 10.1093/oso/9780195300819.003.0001
10 Reece J.B. Campbell N.A. Biology, 9 student ed 2011 Cummings Boston, Mass. Munich
11 Elsafi M. Koraim Y. Almurayshid M. Almasoud F.I. Sayyed M.I. Saleh I.H. Investigation of photon radiation attenuation capability of different clay materials Materials 14 2021 6702 10.3390/ma14216702 34772229
12 Negron-Mendoza A. Ramos-Bernal S. The role of clays in the origin of life Seckbach J. Origins 2004 Springer Netherlands Dordrecht 181 194 10.1007/1-4020-2522-X_12
13 Hazen R.M. Sverjensky D.A. Azzolini D. Bish D.L. Elmore S.C. Hinnov L. Milliken R.E. Clay mineral evolution Am. Mineral. 98 2013 2007 2029 10.2138/am.2013.4425
14 Pasek M.A. Rethinking early Earth phosphorus geochemistry Proc. Natl. Acad. Sci. U.S.A. 105 2008 853 858 10.1073/pnas.0708205105 18195373
15 Pasek M.A. Gull M. Herschy B. Phosphorylation on the early earth Chem. Geol. 475 2017 149 170 10.1016/j.chemgeo.2017.11.008
16 Toner J.D. Catling D.C. A carbonate-rich lake solution to the phosphate problem of the origin of life Proc. Natl. Acad. Sci. U.S.A. 117 2020 883 888 10.1073/pnas.1916109117 31888981
17 Nichols G. Sedimentology and Stratigraphy second ed. 2009 Wiley-Blackwell Chichester, UK ; Hoboken, NJ
18 Chako Tchamabé B. Carrasco-Núñez G. Miggins D.P. Németh K. Late pleistocene to holocene activity of Alchichica maar volcano, eastern trans-Mexican volcanic belt J. S. Am. Earth Sci. 97 2020 102404 10.1016/j.jsames.2019.102404
19 Vilaclara G. Chávez M. Lugo A. González H. Gaytán M. Comparative description of crater-lakes basic chemistry in Puebla State, Mexico SIL Proceedings, 1922-2010 25 1993 435 440 10.1080/03680770.1992.11900158
20 Kaźmierczak J. Kempe S. Kremer B. López-García P. Moreira D. Tavera R. Hydrochemistry and microbialites of the alkaline crater lake Alchichica, Mexico Facies 57 2011 543 570 10.1007/s10347-010-0255-8
21 Caballero M. Vilaclara G. Rodríguez A. Juárez D. Short-term climatic change in lake sediments from lake Alchichica, Oriental, Mexico Geo Int. 42 2003 529 537 10.22201/igeof.00167169p.2003.42.3.942
22 Van Kranendonk M.J. Two types of Archean continental crust: plume and plate tectonics on early Earth Am. J. Sci. 310 2010 1187 1209 10.2475/10.2010.01
23 Zahnle K. Schaefer L. Fegley B. Earth's earliest atmospheres Cold Spring Harbor Perspect. Biol. 2 2010 10.1101/cshperspect.a004895 a004895–a004895
24 Castillo-Rojas S. Landeros J.C. Negrón-Mendoza A. Navarro-González R. Radiolysis of aqueous formaldehyde relevant to cometary environments Adv. Space Res. 12 1992 57 62 10.1016/0273-1177(92)90154-P
25 Orozco D F. Análisis Químico Cuantitativo 20a ed 1994 Porrúa México
26 Westwall J.C. Zachary J.L. Morel F.M.M. MINEQL, A Computer Program for the Calculation of Chemical Equilibrium Composition of Aqueous Systems, Tech Note 18, Dept. Of Civil Eng 1976 Mass. Inst. Technol. Cambridge, MA
27 Brown D.S. Allison J.D. MINTEQA1, an Equilibrium Metal Speciation Model: User's Manual 1987 U.S. Environmental Protection Agency Athens, GA EPA/600/3-87/012
28 White W.M. Geochemistry 2013 John Wiley & Sons Inc Hoboken, NJ
29 Feulner G. The faint young Sun problem Rev. Geophys. 50 2012 2011RG000375 10.1029/2011RG000375
30 Madejová J. Gates W.P. Petit S. IR spectra of clay minerals Developments in Clay Science 2017 Elsevier 107 149 10.1016/B978-0-08-100355-8.00005-9
31 Olsson R. Giesler R. Loring J.S. Persson P. Adsorption, desorption, and surface-promoted hydrolysis of glucose-1-phosphate in aqueous goethite (α-FeOOH) suspensions Langmuir 26 2010 18760 18770 10.1021/la1026152 21087005
32 Passauer L. Bender H. Fischer S. Synthesis and characterisation of starch phosphates Carbohydrate Polymers 82 2010 809 814 10.1016/j.carbpol.2010.05.050
33 Tang Y. Cheng F. Feng Z. Jia G. Li C. Stereostructural elucidation of glucose phosphorylation by Raman optical activity J. Phys. Chem. B 123 2019 7794 7800 10.1021/acs.jpcb.9b05968 31335146
34 Cleaves Ii H.J. The prebiotic geochemistry of formaldehyde Precambrian Res. 164 2008 111 118 10.1016/j.precamres.2008.04.002
35 López-Islas A. Colín-García M. Negrón-Mendoza A. Stability of aqueous formaldehyde under γ irradiation: prebiotic relevance Int. J. Astrobiol. 18 2019 420 425 10.1017/S1473550418000368
36 Benner S.A. Kim H.-J. Biondi E. Prebiotic chemistry that could not not have happened Life 9 2019 84 10.3390/life9040084 31739415
37 Haas S. Sinclair K.P. Catling D.C. Biogeochemical explanations for the world's most phosphate-rich lake, an origin-of-life analog Commun Earth Environ 5 2024 28 10.1038/s43247-023-01192-8
38 Hardie L.W.A. Eugster H.P. The evolution of closed-basin brines Mineral Society of America 3 1970 273 290
39 McConnell D. Apatite: its Crystal Chemistry, Mineralogy, Utilization, and Geologic and Biologic Occurrences 1973 Springer Vienna Vienna
40 Power I.M. Kenward P.A. Dipple G.M. Raudsepp M. Room temperature magnesite precipitation Cryst. Growth Des. 17 2017 5652 5659 10.1021/acs.cgd.7b00311
41 Rinder T. Dietzel M. Leis A. Calcium carbonate scaling under alkaline conditions – case studies and hydrochemical modelling Appl. Geochem. 35 2013 132 141 10.1016/j.apgeochem.2013.03.019
42 Fuentes-Carreón C.A. Cruz-Castañeda J.A. Mateo-Martí E. Negrón-Mendoza A. Stability of DL-glyceraldehyde under simulated hydrothermal conditions: synthesis of sugar-like compounds in an iron(III)-Oxide-Hydroxide-Rich environment under acidic conditions Life 12 2022 1818 10.3390/life12111818 36362973
43 Delidovich I.V. Simonov A.N. Pestunova O.P. Parmon V.N. Catalytic condensation of glycolaldehyde and glyceraldehyde with formaldehyde in neutral and weakly alkaline aqueous media: kinetics and mechanism Kinet. Catal. 50 2009 297 303 10.1134/S0023158409020219
44 Mermut A.R. Baseline studies of the clay minerals society source clays: chemical analyses of major elements Clay Clay Miner. 49 2001 381 386 10.1346/CCMN.2001.0490504
45 Söderholm S. Roos Y.H. Meinander N. Hotokka M. Raman spectra of fructose and glucose in the amorphous and crystalline states J Raman Spectroscopy 30 1999 1009 1018 10.1002/(SICI)1097-4555(199911)30:11<1009::AID-JRS436>3.0.CO;2
46 Godinot C. Gaysinski M. Thomas O.P. Ferrier-Pagès C. Grover R. On the use of 31P NMR for the quantification of hydrosoluble phosphorus-containing compounds in coral host tissues and cultured zooxanthellae Sci. Rep. 6 2016 21760 10.1038/srep21760
47 Badertscher M. Bühlmann P. Pretsch E. Structure Determination of Organic Compounds: Tables of Spectral Data fourth ed. 2009 Springer Berlin Heidelberg Berlin, Heidelberg 10.1007/978-3-540-93810-1
48 Clark L.L. Ingall E.D. Benner R. Marine organic phosphorus cycling; novel insights from nuclear magnetic resonance Am. J. Sci. 299 1999 724 737 10.2475/ajs.299.7-9.724
49 Gorenstein D.G. Luxon B.A. 31P NMR Encyclopedia of Spectroscopy and Spectrometry 1999 Elsevier 2204 2212 10.1016/B978-0-12-374413-5.00242-6
50 Taylor V.F. March R.E. Longerich H.P. Stadey C.J. A mass spectrometric study of glucose, sucrose, and fructose using an inductively coupled plasma and electrospray ionization Int. J. Mass Spectrom. 243 2005 71 84 10.1016/j.ijms.2005.01.001
51 Chai W. Piskarev V. Lawson A.M. Negative-ion electrospray mass spectrometry of neutral underivatized oligosaccharides Anal. Chem. 73 2001 651 657 10.1021/ac0010126 11217777
52 Shen X. Perreault H. Characterization of carbohydrates using a combination of derivatization, high-performance liquid chromatography and mass spectrometry J. Chromatogr. A 811 1998 47 59 10.1016/S0021-9673(98)00238-6 9691300
53 El Rassi Z. Carbohydrate Analysis: High Performance Liquid Chromatography and Capillary Electrophoresis 1995 Elsevier Amsterdam ; New York
54 McMurry J. Organic Chemistry 8e ed. 2012 Brooks/Cole, Cengage Learning Belmont, CA
55 Omran A. Menor-Salvan C. Springsteen G. Pasek M. The messy alkaline formose reaction and its link to metabolism Life 10 2020 125 10.3390/life10080125 32731352
56 Darvell B.W. More chemistry Materials Science for Dentistry 2018 Elsevier 771 789 10.1016/B978-0-08-101035-8.50030-4
57 Wächtershäuser G. Pyrite Formation, the first energy source for life: a hypothesis Syst. Appl. Microbiol. 10 1988 207 210 10.1016/S0723-2020(88)80001-8
58 Russell M.J. Hall A.J. The emergence of life from iron monosulphide bubbles at a submarine hydrothermal redox and pH front JGS 154 1997 377 402 10.1144/gsjgs.154.3.0377
59 Chen C.S. Water activity – concentration models for solutions of sugars, salts and acids J. Food Sci. 54 1989 1318 1321 10.1111/j.1365-2621.1989.tb05982.x
60 Dinane A. Guendouzi M.E. Mounir A. Hygrometric determination of water activities, osmotic and activity coefficients of (NaCl+ KCl)(aq) at T= 298.15 K J. Chem. Therm. 34 2002 423 441 10.1006/jcht.2001.0845
61 Mazurkiewicz J. Tomasik P. Zapłotny J. Relationships between water activity and viscosity of solutions Food Hydrocolloids 15 2001 43 46 10.1016/S0268-005X(00)00048-5
