2.1 Generation and Observation of Water Radical Cations

Gas-phase monomeric water radical cation, H₂O⁺⋅ is generated readily by direct ionization of water vapor. This was done in vacuum in 1923, when Smyth observed an unexpected ion in a mass spectrum at m/z 18 and attributed it to H₂O⁺⋅, resulting from electron impact on water vapor.¹⁷ Subsequent research using mass spectrometry has characterized H₂O⁺⋅ in vacuum including its ionization energy and ion/molecule reactivity.^18-23 However, spectroscopic evidence for this species remained scarce until 1973 when Heiber et al. reported both absorption and emission features in the range 350–680 nm.²⁴

Studies of water radical cations under ambient conditions are more relevant to microdroplet reactivity. Recently, Chen et al. developed a corona discharge-based device for the generation of water radical cations in air in quantity (Figure 1a).²⁵ By introducing a moist inert gas into the discharge volume, direct ionization of water was achieved in air and this allowed detection of water cations by mass spectrometry. Intact H₂O⁺⋅ ions were accompanied by a dominant mono-solvated ion, H₂O⁺⋅(H₂O) and by higher solvated forms (Figure 1b). The relatively high abundance of the water dimer radical cation hints at its exceptional stability compared to the polysolvated species. Various computational studies have proposed two contributing structures for the charged water dimer.^26-28 One isomer exhibits characteristics of hydrogen O−H bonding (H₃O⁺…⋅OH structure A in Figure 1c). The other isomer, H₂O⁺⋅…OH₂, contains a one-electron O−O bond (structure B in Figure 1c). Computations indicate that structure A is more stable by 8.8 kcal/mol vs. isomer B. Moreover, the calculation highlights a 15.1 kcal/mol energy barrier to their interconversion (Figure 1c).²⁶ While experimental evidence is limited and somewhat indirect; the existence of both structures, A^{25, 27, 28} and B^{25, 29, 30} has been reported, consistent with their related but distinctive reactivities.²⁵ Further, the hemi-bonded isomer was verified as the intermediate complex in the proton transfer between H₂O⁺⋅ and H₂O.²⁸ Notably, recent research further validated the interconversion using a linear ion trap mass spectrometer.³¹

One noteworthy property of (H₂O)₂⁺⋅ is its lifetime, which is intimately linked to the chemical environment. For instance, (H₂O)₂⁺⋅ can persist in an ion trap for at least 200 ms, and the monomeric form exhibits an even longer lifetime, lasting for at least 800 ms.³¹ This differs notably from the lifetime of water radical cation in aqueous solution (less than a few hundred femtoseconds).^{28, 32, 33} Meanwhile, in solution, the solvation behavior of water radical cation differs between monomeric and dimeric forms. Hence, a thorough exploration of water radical cation's chemical reactivity in different scenarios is indispensable for understanding its role as an oxidant and for harnessing the potential of this highly reactive species.

2.2 Chemical Reactions of Water Radical Cations

Ion-molecule reactions between water radical cation and polyatomic molecules (M) were first investigated in vacuum.^{18, 19} Three types of reactions are involved (Figure 2, top panel), namely (I) electron transfer (also called charge transfer) from the neutral molecule to H₂O⁺⋅ to form a molecular ion (M⁺⋅) and seen for NO, O₂ and CH₂O, (II) proton transfer from H₂O⁺⋅ to the neutral, generating the protonated molecule ([M+H]⁺) and hydroxyl radical as products, as seen for neutral water itself, and (III) hydrogen atom transfer from the neutral to H₂O⁺⋅, e.g. CH₄, to give H₃O⁺⋅. It is notable that the three reaction channels are not mutually exclusive, for example, the reaction between H₂O⁺⋅ and H₂S generates H₂S⁺, H₃S⁺ and HS as products.¹⁹

Under humid ambient conditions, the water dimer radical cation is the dominant species (Figure 1b). The types of reaction of (H₂O)₂⁺⋅ are similar to those of the monomer in that electron transfer and proton transfer occur.³⁴ Most notably, its chemistry revolves around the direct oxidation of unsaturated chemical bonds (Figure 2, middle panel), including alkene,³⁵ acetone,³⁶ ester,²⁵ and benzene,^{25, 37} as confirmed by exact mass measurements^{35, 37} and ¹⁸O-labeled experiments.^{25, 35-37} It is worth noting that the true structure of the intermediate (M+H₂O)⁺⋅ in the oxidation of carbonyl compounds remains undetermined, although a reasonable case is made for a geminal diol radical cation. In the case of benzene, oxygen insertion occurs to give phenol as oxidation product, with an proposed intermediate, (C₆H₆…H₂O⁺⋅).^{25, 37} The reaction of disulfides with (H₂O)₂⁺⋅ also results in the formation of (M+H₂O)⁺⋅_, presumably adopting a triangular conformation, as illustrated in Figure 2, reaction VII.²⁵ The gas-phase chemistry of water radical cation has been thoroughly reviewed.³⁸ An unresolved question regarding the interconversion between the proton-bonded dimer (A) and the hemi-bonded dimer (B), concerns which of these structures accounts for the observed oxidation. Notably, benzene oxidation is attributed to structure A, while disulfide oxidation is attributed to structure B.²⁵ It is notable, too, that water radical cations at various solvation levels, such as monomer, trimer or those with a higher water content as depicted in Figure 2b, may also be involved in the oxidation process. This is particularly relevant considering that the dimer is not isolated in the experimental environment.

The solution-phase chemistry of the water radical cation in solution is illustrated in Figure 2, bottom panel. Although the generation of water radical cations is considered as the initial step in the radiolysis of water,³⁹ direct observation of the oxidation in aqueous solution presents a substantial challenge due to the ultrafast decay of water radical cations upon interaction with the solvent, water (Figure 2, reaction X).^{28, 32, 33} The lifetime of the monosolvated water radical cation is not known, although it is expected to be considerably longer. Hence, water-deficient environments proximate to the water radical cation are pivotal in decreasing undesired reactions. This expectation is consistent with the observation that the oxidation of H₂SO₄ and H₃PO₄ by water radical cations, generated through water radiolysis, is only seen in concentrated solutions (Figure 2, reactions VIII and IX),^{40, 41} where the water radical cation primarily encounters the reactants (H₂SO₄ and H₃PO₄) rather than water itself. Encounters with water result in the generation of ⋅OH, a milder oxidant incapable of oxidizing either of these acids. The slow proton transfer between the water radical cation and these strong acids further facilitates oxidation processes by the water radical cation.

3.1 Unique Interfacial Properties of Microdroplets

Water radical cation is considered the strongest oxidant in aqueous solutions, with an estimated standard redox potential (H₂O⁺⋅/H₂O) exceeding 3 V.^{33, 42} However, its practical application is hampered by two key limitations: (i) the high energy required for its generation and (ii) the rapid reaction of the water radical cation in the presence of proton acceptors, like water itself. A recent perspective by Ben-Amotz has shed a light on this dilemma, proposing that the charged water pair (H₂O⁺…H₂O⁻) exists fleetingly in bulk water at a concentration of 3 M.² The pair is generated through initial electron delocalization along the hydrogen bond between two adjacent water molecules (which gives the high concentration but a short lifetime), followed by the electron transfer event. A calculation suggests that this process can be significantly enhanced at the interface where an asymmetric hydrogen bond breaks, facilitating charge delocalization.⁴³ Small aqueous microdroplets, with their relatively large interfacial areas, are therefore ideal for fostering the generation of water radical cations. Furthermore, microdroplet surfaces have unique and interconnected features including (i) strong electric field, (ii) highly acidic environment, (iii) partial solvation of reagents, (iv) low water content, and (v) high local concentration of redox species (Figure 3), which support the generation of water radical cations and promote the desired oxidation reactions, while simultaneously reducing the undesired decay of these reactive species.

Microdroplets can function as electrochemical cells, owing to the presence of strong electric fields at their surfaces. Electric field formation relies on (i) alignment of water molecules, with dangling O−H bonds oriented perpendicular to the surface,¹⁰ and/or (ii) electron transfer leading to a water radical pair. An overall net charge may occur and it may be the result of (i) ionic gradients, where ions such as bicarbonate accumulate near the surface, resulting in the buildup of an electrical double layer,⁴⁴ or (ii) it may be produced during electrospray^{45, 46} or sonic spray microdroplet generation.⁴⁷ Spectroscopic techniques have been employed to quantify the electric field strength, and they reveal a remarkably high field of approximately 10⁹ V/m at aqueous microdroplet surfaces (although the interface is water-oil).¹¹ It seems that the strong electric field can facilitate the generation and persistence of water radical cations through the separation of charged water pair (H₂O⁺…H₂O⁻) induced by the electric field. It is notable that direct ionization of water can also form the water radical cation. Note too, that this latter process requires far less energy than direct ionization of isolated water, which has the relatively high ionization energy of 12.6 eV.⁴⁸

Meanwhile, consideration of this equation prompts a novel inquiry: Could the water radical cation serve as a potential chemical source for the lower pH at or near the interface? Although the answer is still under investigation, we believe eq. 3 is the key to connecting Brönsted's equation with Ben-Amotz's equation (eq. 2).

The rapid decay of water radical cation (eq. 3) raises a new question: how can the water radical cation survive in aqueous droplets? The answer is simple: surfaces of charged aqueous microdroplets are relatively dry. The local environment is rich in charged species (e.g. proton, water radical cation, hydronium ion) but is relatively water-poor. This situation offers an intriguing solution to a longstanding enigma in the field of the origin of life—the water paradox.⁵¹ Specifically, it addresses the question of how primordial biomolecule precursors, such as amino acids, can undergo dehydration to give rise to biopolymers like peptides.⁵² The “dry” surface conditions also play a crucial role in preserving water radical cations at the surface, as they experience reduced interactions with water, thereby diminishing their loss by hydration reactions that ultimately lead to hydronium ions and hydroxyl radicals as well as further secondary products like hydrogen peroxide.

The qualitative concept that the surfaces of aqueous microdroplets are relatively dry is closely linked to another concept, namely the partial solvation hypothesis, which is now discussed. The overwhelming importance of solvation in reaction kinetics is well-known from the epochal work of Brown, Brauman and others.^{53, 54} It is generally understood to account for the orders-of-magnitude difference in reaction rates between the gas phase and the bulk solution.⁵⁵ This fact is particularly significant in explaining the difference in bulk solution and microdroplet reaction rates. Partial solvation of reactants at the droplet interface explains why reaction rates exhibit a remarkable acceleration, reaching values of 10⁴ to 10⁶.^{4, 9} Much like the molecular mechanism for surface tension,⁵⁶ the active species located at the surface experience limited interactions with a reduced number of solvent molecules. This phenomenon leads to the incomplete envelopment of reactants by their solvation shells, thereby altering the reaction energy profile and producing a reaction rate intermediate between those in the gas phase and bulk solution. This general mechanism is believed to accelerate oxidation by water radical cation as well.

By analogy to the oxidation of H₂SO₄ and H₃PO₄ in solution, a high concentration of reactants proves to be essential for their oxidation by water radical cation. In microdroplet synthesis, solvent evaporation naturally boosts reactant concentrations,⁵⁷ while the accumulation of surface-active compounds further contributes to increased surface concentrations.^{58, 59} These dual mechanisms function in tandem to ensure that surface-generated water radical cations promptly engage with substrate M in its oxidation.

3.2 Spontaneous oxidation in microdroplets

Similar to ambient gas phase chemistry, reactions between the water radical cation and substrate (M) in microdroplets begin with direct addition (Figure 2, reactions IV to VII). When ions of organic compounds are generated via electrospray (with accompanying microdroplet formation), the protonated form of the neutral molecule (M+H⁺) typically predominates. Nevertheless, when the compound contains a heteroatom double bond, conditions favoring the dominance of the (M+18) ion become apparent, with two possibilities: either the simple ammonium adduct (M+NH₄⁺) or water radical cation adduct (M+H₂O⁺⋅). In fact, in an extensive study involving approximately 21,000 multifunctional compounds, a notable 18 % of all compounds exhibited abundant (M+18) ions, with this proportion increasing to 65 % within the subgroup featuring X=Y bonded compounds.⁶⁰ This preference agrees with the observed trend in the gas phase (Figure 2, middle panel), supporting the existence, or at least, partial existence, of (M+H₂O⁺⋅) in those cases, although further experiments to discriminate between the two possible species are required. What is particularly interesting is the possible close connection of the (M+H₂O⁺⋅) ion to the spontaneous redox reactions. The ion H₄O₂⁺⋅ has been successfully detected in spray-based microdroplets using mass spectrometry.^{14, 60, 61} Although the direct evidence to show (M+H₂O⁺⋅) conversion to the final oxidation products is missing, the addition of water dimer radical cation was shown to improve the oxidation conversion (see section 3.3).^{60, 62}

The true identity of the oxidizing agent is a widely-discussed yet controversial subject in microdroplet chemistry. Table 1 summarizes the classification, reaction conditions, and oxidants suggested in various studies. Generally, two types of oxidation have been observed in microdroplets: (i) simple oxidation of substrates, where a single reactant is oxidized by an oxidant to yield the final product,^{13, 14, 16, 60, 62-67} and (ii) oxidative coupling reactions, in which one reactant is oxidized to generate a radical species (as intermediate) that subsequently couples with another reactant to produce a coupling product.^{15, 68-70} As shown in table 1, spontaneous oxidation encompasses a wide array of compounds, from organic compounds (most cases), through inorganic compounds (reaction 1.6) to large biomolecules (reactions 1.12 and 1.13).

Table 1. Summary of spontaneous oxidation in microdroplets.

It is notable that spontaneous reduction processes have also been observed in microdroplets,^71-75 encompassing small molecule reduction and the formation of metal nanoparticles. In some instances, these reduction occur concurrently with oxidation,^{14, 76} indicating the coexistence of both oxidants and reductants. Much like the ongoing debate surrounding oxidation, the identity of reductants and their sources remain under investigation. The hydroxide at the droplet surface might be an electron donor, while the water radical anion (viz. the hydrated electron) is possibly a more plausible reducing agent. Although this is consistent with the importance of the redox pair discussed above (eq. 2), reduction is not a focus of this review.

3.3 Primary Oxidants in Microdroplet Reactions

An intriguing finding shown in Table 1 is the diversity of suggested oxidants, including (i) hydrogen peroxide and/or hydroxyl radical, often hypothesized to arise from the oxidation of hydroxide at the microdroplet‘s interfacial electric field, (ii) water radical cation (in a monomeric or dimer form), believed to originate from water radical pairs separated by the electric field, (iii) oxygen atoms, derived from molecular oxygen in the surrounding air due to the electric field, and (iv) direct electron transfer by the electric field itself. There is a broad consensus that the interfacial electric field serves as the primary source of oxidative activity. However, discriminating between direct electron detachment by the electric field from indirect oxidation via chemical oxidants generated within (or responsible for) the field remains a significant challenge. Even the investigators who suggest that reactions 1.13 and 2.4 are directly electric field driven, as supported by evidence of the inability of hydrogen peroxide to oxidize the substrates and the ability of electric field to direct oxidize the substrates (as estimated based on oxidation potentials), do not conclusively eliminate alternative chemical oxidants. Another interesting finding emerges from comparison of reactions 1.3 and 1.9, where both reactants possess with the same functional group but yield different products—ester and carboxylic acid, respectively. The different starting material structures between these cases and dissimilar conditions (solvents, droplet sizes, and spray distances) will contribute to this outcome. Additionally, the conversion of the ester to its corresponding acid via hydrolysis might also play a role in this process.

The most frequently suggested oxidizing agent in microdroplet-driven spontaneous oxidation is hydrogen peroxide/hydroxyl radical. Ongoing discussions continue regarding the production and yield of these species. In 2019, Zare and colleagues first reported the spontaneous generation of hydrogen peroxide in aqueous microdroplets produced by sonic spray (no applied electrical potential), with an observed concentration of H₂O₂ of 30 μM.¹³ Subsequently, they measured H₂O₂ levels in microdroplets formed by water vapor condensation, reporting concentrations in the range of 50–100 μM.⁷⁷ However, these findings have been challenged by Mishra et al., who quantified H₂O₂ in water droplets and reported no detection of H₂O₂ in condensation microdroplets, with only 1 μM of H₂O₂ detected in microdroplets generated by sonication.⁷⁸ This same research group later explored the impact of the surrounding air, suggesting that O₃ in the air serves as the source of H₂O₂, rather than water itself, with the function of the water interface being to aid in mass transfer (note that in this study, control experiments were performed, showing no detectable H₂O₂ generation without O₃).⁷⁹ In response to this argument, Zare et al. re-evaluated the quantity of H₂O₂ generated in aqueous microdroplets in the absence of O₃ (finding it to be less than 3 ppb), revealing the potential production of 0.3 to 1.5 μM of H₂O₂.⁸⁰ Separately, Nguyen and colleagues proposed that ultrasonic cavitation is responsible for H₂O₂ generation, with the sonication process being involved in microdroplet formation.⁸¹

It is important to highlight two crucial considerations: (i) microdroplet reactions commonly involve solvent evaporation, droplet fission, ion ejection, and/or impurity adsorption, making the quantification of reactive species within microdroplets under ambient conditions a formidable practical challenge. (ii) The question of whether hydrogen peroxide is generated and to what extent microdroplets can undergo spontaneous oxidation reactions are interrelated yet distinct inquiries, given the likely existence of other oxidants. Given the intricate nature of identifying authentic oxidants, meticulous control experiments are necessary, although regrettably, they have been conducted in only a limited number of studies.^{13, 14, 16, 60, 62, 63, 67}

Adding H₂O₂ into microdroplets is a simple way of testing evidence for H₂O₂ as the primary oxidant in spontaneous oxidations. However, only two studies have demonstrated an increase in oxidation yield with H₂O₂ addition, as observed in reactions 1.1 to 1.3.^{13, 63} Conversely, several studies have reported no change in oxidation yield when H₂O₂ is added, as evidenced in reactions 1.8, 1.11, 1.12, and 1.13.^{14, 16, 62, 67} Instead, alternative oxidants are suggested in the latter studies and supported by additional experimental evidence. In the oxidation of sulfones and sulfonamides (reactions 1.8 and 1.10), water radical cations were externally generated and added to the reaction system, and shown to result in increased conversion (Figure 4a).^{60, 62} In the examination of lipid oxidation, oxygen isotopic labeling experiments utilizing H₂O¹⁸ revealed that the oxygen originates from the surrounding air (here, oxygen atom is the proposed oxidant), rather than from water itself, since ¹⁸O was not inserted into lipids when H₂¹⁸O was used as solvent (Figure 4b).¹⁶ More interestingly, even though hydrogen peroxide was found incapable of oxidizing these substrates, its production was still detected.^{16, 67}

Variations in reaction conditions also hold significance. We observe that in pure water droplets, H₂O₂/⋅OH is often suggested as the primary oxidant, while in organic droplets containing traces of water, water radical cation is proposed. Considering that in the solution phase, water radical cations can react with water molecules, particularly in water-rich environments, to generate hydroxyl radicals (eq. 3), we propose that water radical cations play a central role in all cases and serve as the source of H₂O₂/⋅OH. It is worth noting that this assumption harmonizes with the observation of decreased oxidation yields when additional water was added in sulfone or phosphonate oxidation (where water radical cation decays to H₂O₂/⋅OH, incapable of oxidizing these substrates),^{14, 62} and aligns with the outcomes of oxygen isotope experiments in lipid oxidation.¹⁶ As mentioned earlier, spontaneous reduction can coexist with oxidation, and O₂ has been demonstrated to be captured and reduced to O₂⁻ (m/z 32) by either electrons or hydrated electrons (viz. water radical anions).^{8, 14} The resulting O₂⁻ also functions as an oxidant, facilitating the incorporation of oxygen from the surrounding air into substrates.

3.4 Mechanistic Overview of Spontaneous Oxidation in Microdroplets Driven by Water Radical Cation

The strong interfacial field is considered to be the driving force behind spontaneous oxidation and reduction in microdroplets. While direct evidence supporting the idea that the substrate (M) is oxidized by the electric field is lacking, this possibility cannot be dismissed (Figure 5a). As previously proposed, the electric field can facilitate the separation of charged water molecules, resulting in the formation of water radical cation and anion pairs (Figure 5b). In water-poor environments, such as at the droplet surface or in organic solvent droplets containing traces of water, the generated water radical cation may directly oxidize the substrate M, yielding the product (M=O or M⁺⋅). Conversely, in water-rich conditions, for instance, in the core of water droplets, the water radical cation would have a short lifetime and rapidly react with water, liberating hydroxyl radicals and hydrogen peroxide. This process may lead to the oxidation of M (if M has a low enough oxidation potential to be oxidized by H₂O₂/⋅OH) (Figure 5c) or the termination of the whole process. The latter scenario results in the observation of H₂O₂ production but an unchanged degree of oxidation when external H₂O₂ is added. Additionally, species in the surrounding gas, such as O₂, may also participate in spontaneous oxidation, relying on the reduction of O₂ to superoxide by the water radical anion liberated from the radical pair (Figure 5d). This general mechanism provides a rational explanation for all the observed experimental outcomes, highlighting the water radical cation/anion pair as the authentic chemical source for spontaneous redox chemistry in microdroplets.