Sources of electrochemical energy are widespread in Nature, including hydrothermal gradients, corrosion of metal-rich rocks and minerals, concentration gradients of redox active species, and streaming currents caused by the flow of electrolyte solutions thorough porous media. Only recently, however, we began to realize that this energy is also widely used by Nature in both biological and abiotic processes (Nakamura et al., 2010b; Revil et al., 2010) and that electrochemical redox reactions are important contributors to biogeochemical cycles (Nielsen and Risgaard-Petersen, 2015). The passage of electric currents through conductive minerals has been long-appreciated (Wells, 1914), and has gained a renewed interest due to its relevance to the biochemistry of hydrothermal vent systems and the origin of life (Karato and Wang, 2013; Malvankar et al., 2014; Yamamoto et al., 2017). Most of the redox processes across electrochemical gradients were previously associated with bacterial activity (Müller et al., 2016; Malkin et al., 2017) or human-made devices (Du et al., 2007). A detailed understanding of the extent and mechanisms of abiotic electron flows can complement our understanding of natural electrochemical systems and will be important for more accurate modeling of global biogeochemical cycles.

Many oxide and sulfide minerals are electrical conductors or semiconductors, with resistivity varying from 0.1 to 10⁵ Ω × m (Karato and Wang, 2013), capable of conducting electrons over centimeter-long distances between environments of different redox potentials (Nakamura et al., 2010a; Nielsen et al., 2010). Conceivably, electrical currents transported through conductive minerals could couple chemical and biochemical reactions in physically separated environments, form natural “fuel cells” and provide the energy necessary to maintain rich biological communities on the border of oxic and anoxic zones (Jelen et al., 2016).

The efficiency of electrochemical reactions depends not only on the magnitude of an applied redox potential, but also on the properties of the electrode. In general, a good electrode must be electrically conductive, chemically stable under selected conditions and provide a catalytically active surface for the reagents. There is a growing interest in using sustainable mineral electrodes for chemical reactions (Konkena et al., 2016) and a growing awareness of the contribution of electrochemistry to natural processes (Nielsen and Risgaard-Petersen, 2015). This study was restricted to electrochemistry of common iron oxides and sulfides assuming that due to their abundance they will dominate electron flow across different geochemical settings. Iron sulfides have been used as electrodes for electrochemical CO₂ reduction in systems that simulate chemistry of hydrothermal vents related to the origin of life (Herschy et al., 2014; Yamaguchi et al., 2014; Roldan et al., 2015), while iron oxides have been extensively studied as terminal electron acceptors for bacterial respiration (Orsetti et al., 2013).

While filamentous and cable bacteria use electrochemical energy directly (Aklujkar et al., 2009; Seitaj et al., 2015), a wide range of bacteria are known to exchange electrons indirectly, using external electron mediators such as water-soluble flavines (Marsili et al., 2008), quinones (Newman and Kolter, 2000; Qiao et al., 2008) or, quinone-rich humic acid substances (Cervantes et al., 2000; Klüpfel et al., 2014). The addition of external quinones to microbial communities modifies interactions between microbes (Scheller et al., 2016) and changes the mode of respiration in bacteria (Cervantes et al., 2000). Interactions between organic mediators and conductive minerals may contribute to the overall flow of energy that defines biogeochemical cycles, being especially important in the environments where oxygen and solar radiation are absent (D'Hondt et al., 2004; Borch et al., 2010).

As model organic molecules, I selected quinones (Q) and their reduced form, hydroquinones (H₂Q), which are key electron mediators across all domains of life (Soballe and Poole, 1999) (Figure 1). Quinones are found everywhere from interstellar clouds (Bernstein, 2006) to environmental pollutants (Mousset et al., 2016). Selection of the quinones for the study broadly represented different environmental roles that these molecules have: 1 is the simplest known hydroquinone molecule and often used for model studies of quinone-mineral interactions (Uchimiya and Stone, 2006; Klüpfel et al., 2014), 2 is 2-methoxyhydroquinone, a common component on non-soluble organic matter and a plant messenger molecule (Lynn and Chang, 1990; Yuan et al., 2016), while 3 is an oxidation product of polycyclic aromatic hydrocarbons and is both a common pollutant and a likely component of prebiotic mixtures of organic molecules delivered by comets and meteorites to early Earth (Chyba and Sagan, 1992; Bernstein et al., 1999).

Electrochemical techniques are often used to study the participation of organic molecules in redox processes of redox active clays and minerals (O'Loughlin, 2008; Sander et al., 2015). Here cyclic voltammetry (CV) was used to detect hydroquinone oxidation on the surface of electrodes made from common iron oxides and sulfides. CV is an analytical technique used to monitor trace metals in marine sediments, where an increase in current is observed when a redox process takes place on the electrode's surface at a specific applied redox potential (Luther et al., 2008; Moore et al., 2009).

Differentiating the chemical contribution of electron redox cycling from usually observed microbial redox activity would improve our understanding of global biogeochemical carbon cycles. Electron transfer between conducive minerals and soluble organics can potentially generate out-of-equilibrium systems that impact environments dependent on the abiotic redox gradients, such as microbial communities around hydrothermal vent chimneys and in anoxic marine sediments (Hedrick and White, 1986; Unden and Bongaerts, 1997; Kim et al., 2012). Quinone redox cycling can affect distribution of redox active species in water columns (Nielsen et al., 2010; Nielsen, 2016) and natural bioremediation of organic molecules (Jiang et al., 2015). Finally, redox gradients have been long considered important energy source for the origin of life (Martin et al., 2008), and a mechanism of transferring energy stored in redox gradients into chemical reactions is an important step toward better understanding processes that may lead to life on Earth and other planets.

The reactivity between the mineral and redox species in solution depends on the amount of energy available in the system and the electrocatalytic properties of mineral surfaces. Overall, the energy available in the redox system depends on redox potential as defined by equation (1): (1)ΔG =-nEF E is a measurable potential difference between reacting molecules, described by the Nernst equation: (2)E=E°+RTnFln([Ox][Red]) Here E° is a standard redox potential of the redox reaction of interest, n is the number of electrons transferred in the reaction, T is the absolute temperature in K, and R and F are the gas and Faraday constants respectively. Naturally occurring sustained redox potentials vary from 0.2 V for marine sediments (Komada et al., 2004), 0.4 to 0.7 V for hydrothermal vents (Nakamura et al., 2010b; Ryckelynck et al., 2012) and up to 1.2 V for sulfide-rich mineral deposits (Kruger and Lacy, 1949). The naturally plausible range of redox potentials in aqueous systems is defined by the chemistry of iron and manganese, spreading from 0.7 V to −0.3 V [against a standard hydrogen electrode (SHE)] at pH 7 (Baas Becking et al., 1960). Here Ag/AgCl electrodes were used as a reference (+ 0.22 V redox potential compared to SHE), with the −0.5 to 0.5 V (Ag/AgCl) window used here broadly corresponding to the naturally occurring potential range available for redox reactions of quinone molecules.

This work shows that long distance electron transfer can be coupled to abiotic organic redox reactions. To simulate processes occurring in the environments where electrochemical gradients are present, reactions on electrode surfaces made of common minerals were studied at the potential range of −500 to 500 V (vs. Ag/AgCl) expected to be found in natural environments (Baas Becking et al., 1960). Electrodes made of reduced metal sulfides are chemically unstable under the selected conditions and cannot transfer electrons to model organic molecules hydroquinones. This result is surprising: Reduced iron sulfides are usually considered to be the major catalyst of biogeochemical processes in hydrothermal vents (e.g., White et al., 2015). Observed small rises of current on the surface of metal oxides hematite and magnetite are consistent with the observation of very small effect of quinones on chemical dissolution of Fe(III) oxides (O'Loughlin, 2008). In contrast with other studied minerals, pyrite showed a significant current rise in the presence of low concentrations of quinones (0.2–1 mM), within ranges of concentration of organic matter in natural waters (Uchimiya and Stone, 2009). Electrical currents on pyrite surfaces are at least an order of magnitude higher than the maximum oxidation and reduction peak currents on other minerals, and occur at lower redox potentials, suggesting that the electron transfer between quinones and pyrite is a very efficient process that can dominate electrochemistry in natural environments.

The redox stability of pyrite in the presence of quinones can result in long term quinone cycling driven by potential differences between oxic and anoxic zones. Given the ubiquity of both pyrite (Vokes, 1993) and quinones (Scott et al., 1998), the out-of-equilibrium system generated by their interactions may have multiple impacts on the environment.

It has been shown recently that cable bacteria can form millimeter-long conductive filaments that couple sulfide oxidation in anoxic sediments with oxygen reduction in the oxygen-rich areas (Müller et al., 2016). Given that electrical current through conductive minerals can be coupled to hydroquinone oxidation, and that quinones are common external electron mediators used by a wide variety of bacteria (Hedrick and White, 1986), it is possible that other microorganisms apart from cable bacteria may also participate in biogeochemical processes involving long distance electron transfer. In this case our current models underestimate the role of abiotic electron transfer across the sediments (Seitaj et al., 2015) (Figure 8A).

Many naturally occurring hydroquinones are readily oxidized by dissolved oxygen to quinones. Continuous electrochemical reduction of the quinones back to the hydroquinones in the oxygen-rich areas may increase the ratio of quinone cycling and result in accelerated oxygen depletion of the solution (O'Brien, 1991) as well as abiotic generation of high local concentrations of hydrogen peroxide (Roginsky and Barsukova, 2000) (Figure 8A). It is possible that quinone cycling coupled to the long-range electron transfer is a complex reaction-diffusion system that can contribute to sharp seasonal oxygen variations.

Finally, hydrothermal vents and other naturally occurring zones where electrochemical gradients are found have long been considered as out-of-equilibrium systems relevant to the origins of life due to an apparent similarity between biological and geological electrochemical gradients (Martin and Russell, 2003; Amend and McCollom, 2009; Nitschke and Russell, 2009). In general the hydrothermal vent scenarios of the origins of life suggest that reduced metal sulfides were the precursors to the biological iron-sulfur redox cofactors and the catalysts of prebiotic reactions. The exact nature of the metal sulfide catalysts as well as the range of the relevant prebiotic reactions remains a matter of debate (Orgel, 2008; Wächtershäuser, 2016).

In Wachterhauser's “pyrite pulled metabolism,” a model proposed almost 30 years ago, pyrite served as a catalytic surface able to adsorb and concentrate negatively charged organic molecules, but reduced metal sulfides, such as pyrrhotite and amorphous FeS, were expected to participate in the redox reactions of the system (Wachterhauser, 1988). With few exceptions (Wang et al., 2012), reduced iron and niquel sulfides have been generally proposed as plausible minerals to catalyze prebiotic redox reactions instead of pyrite (Cody et al., 2000; Huber et al., 2012; Novikov and Copley, 2013).

Electrochemical systems have been recently suggested as a relevant approach to study prebiotic chemistry related to hydrothermal vents scenario (de Aldecoa et al., 2013; Barge et al., 2014, 2015; Herschy et al., 2014; Yamaguchi et al., 2014; Roldan et al., 2015). The focus of these studies is usually on abiotic CO₂ reduction and reduced iron sulfides and sometimes iron oxides (green rust) are considered as possible electrode candidates. Focus on CO₂ reduction as a main model system relevant to the origin of life has been chosen because of the common assumption that life started simply and evolved toward larger synthetic complexity (Martin et al., 2008). The opposite idea that life started from complex chemical mixtures (de Duve and Miller, 1991; Burton et al., 2012) is gaining more support with the growing realization that a large amount of organic material is formed abiotically in space and delivered to Earth with comets and meteorites (Chyba and Sagan, 1992). Under this scenario quinones, which have been produced abiotically under conditions simulating exoplanetary comet ices (Bernstein et al., 1999), are plausible first molecules relevant to the origin of life. The ubiquitous occurrence of many different quinones used by bacteria, fungi and plants suggests that quinones have a critical role in maintaining life on Earth (Szent-Gyorgyi and McLaughlin, 1983), but the role of quinones in early prebiotic reaction networks is presently underexplored.

The observed affinity between pyrite and quinones loosely resembles reactions in electron transport chain systems found across all domains of life, where redox and pH gradients across cell membranes are maintained, in part, by electron transport between metal sulfur clusters and quinone cofactors (Sun et al., 2013). Electron transport chain has a modular structure (Yankovskaya et al., 2003) and it is possible that a first primitive fragments of a biological electron transport chain appeared from abiotic electrocatalytic interactions between minerals and redox active molecules. In this case, an ability of pyrite surfaces to adsorb and concentrate amino acids and short peptides (Wachterhauser, 1988; Sanchez-Arenillas and Mateo-Marti, 2016) could have provided the link between the early redox systems and prebiotic reaction networks based on peptides and nucleic acids.

The redox processes of the living cells, with a few exceptions (Bar-Even, 2013), take place in the relatively narrow potential zone between −0.6 and 0.1 V (vs. SHE) (Weber, 2002; Bar-Even et al., 2012). If electrochemical reactions were important components of prebiotic redox systems, then minerals that acted as electrodes should have been stable in the potential window corresponding to prebiotic redox reactions, such as reactions involving redox cofactors, components of the citric acid cycle, thiols and thioesters. Side to side comparison of the redox potentials of these reactions to the stability of iron minerals at the same pH and temperature are shown on (Figure 8B). While CO₂ reduction to organic molecules can take place on surfaces of FeS minerals, pyrite stability overlaps with a large part of biochemical redox potentials. Ubiquitous pyrite surfaces likely served as natural mineral electrodes in primordial reaction networks.

The simple system presented in this paper is a model that does not yet address many questions relevant to the natural settings. The selectivity and the stability of pyrite electrodes during the quinone redox cycling in presence of other organic and inorganic compounds will be studied in the next phase of the project. To understand the global impact of abiotic electrochemical reactions a model involving electrical currents across conductive sediments coupled with the diffusion processes at the sediment interphases has to be developed. Reduced iron sulfides are more electrically conductive than pyrite and magnetite (Telford et al., 1990), and are found at higher concentrations in the anoxic zones (Risgaard-Petersen et al., 2012). Possibly the electron flow across the sediments occurs mostly through the reduced iron sulfides, the electron transfer takes place at the pyrite interphase, and there is a competition between mineral corrosion and electrons transfer pathways. The complex pattern of the long-range electron transfer needs to be elucidated.

Present study of non-biological interactions between the organic molecules and the minerals suggests an existence of complex abiotic out-of-equilibrium systems driven by naturally occurring electrochemical gradients. These systems might contribute to global biogeochemical cycles and have an impact on different phenomena ranging from chemical pollution to the origin of life.

OT designed and performed the experiments, analyzed the data, and wrote the paper.