The Role of Carbonate in Catalytic Oxidations

Conspectus CO2, HCO3–, and CO32– are present in all aqueous media at pH > 4 if no major effort is made to remove them. Usually the presence of CO2/HCO3–/CO32– is either forgotten or considered only as a buffer or proton transfer catalyst. Results obtained in the last decades point out that carbonates are key participants in a variety of oxidation processes. This was first attributed to the formation of carbonate anion radicals via the reaction OH• + CO32– → CO3•– + OH–. However, recent studies point out that the involvement of carbonates in oxidation processes is more fundamental. Thus, the presence of HCO3–/CO32– changes the mechanisms of Fenton and Fenton-like reactions to yield CO3•– directly even at very low HCO3–/CO32– concentrations. CO3•– is a considerably weaker oxidizing agent than the hydroxyl radical and therefore a considerably more selective oxidizing agent. This requires reconsideration of the sources of oxidative stress in biological systems and might explain the selective damage induced during oxidative stress. The lower oxidation potential of CO3•– probably also explains why not all pollutants are eliminated in many advanced oxidation technologies and requires rethinking of the optimal choice of the technologies applied. The role of percarbonate in Fenton-like processes and in advanced oxidation processes is discussed and has to be re-evaluated. Carbonate as a ligand stabilizes transition metal complexes in uncommon high oxidation states. These high-valent complexes are intermediates in electrochemical water oxidation processes that are of importance in the development of new water splitting technologies. HCO3– and CO32– are also very good hole scavengers in photochemical processes of semiconductors and may thus become key participants in the development of new processes for solar energy conversion. In this Account, an attempt to correlate these observations with the properties of carbonates is made. Clearly, further studies are essential to fully uncover the potential of HCO3–/CO32– in desired oxidation processes.

CONSPECTUS: CO 2 , HCO 3 − , and CO 3 2− are present in all aqueous media at pH > 4 if no major effort is made to remove them. Usually the presence of CO 2 /HCO 3 − /CO 3 2− is either forgotten or considered only as a buffer or proton transfer catalyst. Results obtained in the last decades point out that carbonates are key participants in a variety of oxidation processes. This was first attributed to the formation of carbonate anion radicals via the reaction OH • + CO 3 2− → CO 3 •− + OH − . However, recent studies point out that the involvement of carbonates in oxidation processes is more fundamental. Thus, the presence of HCO 3 − /CO 3 2− changes the mechanisms of Fenton and Fenton-like reactions to yield CO 3 •− directly even at very low HCO 3 − / CO 3 2− concentrations. CO 3 •− is a considerably weaker oxidizing agent than the hydroxyl radical and therefore a considerably more selective oxidizing agent. This requires reconsideration of the sources of oxidative stress in biological systems and might explain the selective damage induced during oxidative stress. The lower oxidation potential of CO 3 •− probably also explains why not all pollutants are eliminated in many advanced oxidation technologies and requires rethinking of the optimal choice of the technologies applied. The role of percarbonate in Fenton-like processes and in advanced oxidation processes is discussed and has to be re-evaluated. Carbonate as a ligand stabilizes transition metal complexes in uncommon high oxidation states. These high-valent complexes are intermediates in electrochemical water oxidation processes that are of importance in the development of new water splitting technologies. HCO 3 − and CO 3 2− are also very good hole scavengers in photochemical processes of semiconductors and may thus become key participants in the development of new processes for solar energy conversion. In this Account, an attempt to correlate these observations with the properties of carbonates is made. Clearly, further studies are essential to fully uncover the potential of HCO 3 − /CO 3 3 This study is the most detailed study of the role of carbonate in electrocatalytic water oxidation.

INTRODUCTION
All aerated solutions contain a mixture of CO 2 /HCO 3 − /CO 3 2− , and their speciation depends on the CO 2 solubility under the partial pressure of CO 2 in the air and on the following equilibria: 5 5 Therefore, one commonly considers the role of CO 2 /HCO 3 − / CO 3 2− present in solution as a buffer and/or as a proton transfer agent. However, results in recent years have pointed out that the role of CO 2 /HCO 3 − /CO 3 2− in redox processes is often of major importance. This is due to three different reasons: (a) CO This reaction is clearly slow in neutral solutions and requires relatively high concentrations of H 2 O 2 in order to contribute to redox processes. However, the equilibrium in reaction 4 might shift to the right in the presence of metal cations because of reaction 5: (b) The redox potential of the CO 3 •− /CO 3 2− couple is 1.57 V vs NHE, 9,10 and that for the (CO 3 •− + H + )/HCO 3 − couple is clearly somewhat higher because of reaction 3. These potentials are considerably lower than those of the OH • /OH − and (OH • + H + )/H 2 O couples. Thus, at all pH values CO 3 •− has the potential to oxidize water, i.e., it is expected to be involved in oxygen evolution reactions (OERs). Therefore, it is thermodynamically easier to oxidize bicarbonate/carbonate than to oxidize water. Not surprisingly, adsorbed carbonates on semiconductors facilitate photocatalytic water oxidation. 11 The CO 3 •− anion radical is considerably less reactive than the OH • radical, and its reactions are more selective. 21 The CO 3 •− anion radical reacts in most systems via the inner-sphere mechanism. 9,22 The carbonate anion radical is also formed via the following reactions: 23 In biological systems it is formed via the following reactions: 25 25 The CO 3 •− thus formed is believed to be one of the sources of oxidative stress induced by superoxide. 26−28 CO 3 •− is also formed in several enzymatic processes, e.g., in superoxide dismutase. 29 The above-mentioned processes are shown graphically in Scheme 1.
(c) As a strong hard base, carbonate is a very good ligand for high-valent transition metal cations and therefore stabilizes transition metal complexes with higher valence, e.g., Mn III , 30 Fe IV,III , 19,31 Co V/IV/III , 3 Ni IV/III , 32 Cu IV/III , 4 Ru IV , 32 etc. In these complexes, the carbonate is a noninnocent ligand and is therefore involved in electrocatalytic OER processes and photoelectrocatalytic processes.
In the next sections these processes are discussed separately. Carbonate can act as a catalyst/cocatalyst in electrocatalytic water oxidation on the basis of the arguments raised in the Introduction. On the basis of the redox properties of carbonate, the reactions shown in Scheme 3, where M n is formed electrochemically, have to be taken into account to describe the participation of bicarbonate/carbonate in electrocatalytic water oxidation processes. In all of these reactions, the formed peroxide is easily further oxidized to form molecular oxygen at lower potentials. The involvement of bicarbonate/carbonate in homogeneous and heterogeneous electrocatalytic water oxidation, including mechanisms involving specific metal ions (e.g., Cu, Co, Ni), are discussed in the following sections.

Homogeneous Electrocatalysis
The first study reporting homogeneous electrocatalytic water oxidation in the presence of CO 2 /HCO 3 − /CO 3 2− was by Chen and Meyer. 17 They showed that Cu II (aq) with CO 2 /HCO 3 − / CO 3 2− in the medium acts as an efficient water oxidation catalyst on a variety of working electrodes. The catalytic current increases with [Cu II ] with a cathodic shift of the onset potential. The catalytic current depends linearly on [Cu II ] in neutral media, while under alkaline conditions it depends on [Cu II ] 2 . Therefore, at pH 10.8 a bimolecular mechanism involving the active intermediate was proposed, while at pH 6.7 a single copper site was suggested to be the active species. The redox potential of Cu III/II (H 2 O) n is >2.3 V vs NHE, 41 which is obviously shifted cathodically by the strong carbonate ligand. However, the authors did not clarify the involvement of either Cu III or Cu IV as the active species in the catalytic cycle. Later, a density functional theory (DFT) study at pH 8.3 suggested the possibility of a Cu IV complex as the active intermediate. 4,40 Pulse radiolysis is a useful tool to study the chemical properties of complexes in unstable oxidation states. 42,43 This technique was used to study the properties of [Cu III (CO 3 ) n ] 3−2n formed by the oxidation of [Cu II (CO 3 ) n ] 2−2n by CO 3 •− . The results suggested the following oxidation mechanism: DFT calculations of the NBO charges suggested that significant charge transfer from the CO 3 2− to the central metal ion in Cu III (CO 3 ) n 3−2n takes place. The Cu III (CO 3 ) n 3−2n thus formed decomposes in a process that obeys second-order kinetics irrespective of the pH of the medium: 4 The discrepancy with the electrochemical results in neutral solutions is probably due to the fact that in neutral media the electrocatalytic process proceeds via reactions 30−34 in Scheme 3. This hypothesis cannot be tested by the pulse radiolysis technique. 23 The analogous systems containing other divalent first-row transition metal cations in the presence of HCO 3 − /CO 3  3 2− is an excellent catalyst for electrocatalytic water oxidation. 3 During chronoamperometry, a precipitate is formed that serves as a heterogeneous catalyst for the same (vide infra). Electrochemically, three homogeneous processes are observed: (i) A relatively small wave is observed at E p,a ≈ 0.71 V vs Ag/ AgCl that is due to the Co III/II redox couple, i.e., the carbonate ligands shift the redox potential of the Co III/II couple cathodically by ca. 0.9 V. (ii) A second wave is observed at E p,a ≈ 1.10 V vs Ag/AgCl that has a considerably larger current (by a factor of >20) than the first peak. This process is attributed to the Co IV/III (CO 3 ) 3 2−/3− redox couple. The observed potential is in good agreement with the results of DFT calculations. The large current is attributed to catalytic oxidation of bicarbonate/carbonate by the Co IV complex to form HCO 4 − via reactions analogous to reactions 26−28 and 31 in Scheme 3. (iii) A process with very large currents, with a current density of 10.50 mA·cm −2 at [Co II ] = 0.50 mM at a peak plateau that starts at >1.2 V vs Ag/AgCl, is due to the formation of Co V (CO 3 ) 3 − , and the observed redox potential is in accord with that calculated using DFT for the Co V/IV (CO 3 −/2− complex oxidizes both water and bicarbonate/carbonate. The results suggest that the oxidation proceeds via reaction 20 and reactions 25−28 and 30−33 in Scheme 3. If one assumes, as these equations indicate, that the ratedetermining step in this catalytic process involves a twoelectron oxidation process, then k cat = 350 s −1 . During long-time electrolysis at these potentials, a precipitate forms that is a heterogeneous OER catalyst (vide infra). It should be noted that in order to calculate the redox potential, the simplest structures, [Co III/IV/V (CO 3 ) 3 ] 3−/2−/− with octahedral geometries (Figure 1), were considered.
In the system containing the NiL 2 2+ complex, the role of carbonate is due to the lowering of the redox potentials of the Ni III/II and Ni IV/III couples and to the reactions shown in Scheme 4. Recently the process of water oxidation by Ni II L 2 2+ was reinvestigated, and it was shown that the formation of nickel oxide on the electrode surface is mainly responsible for the catalysis. However, the large catalytic wave in the cyclic voltammogram and the isomerization of the complex were not addressed. Under long-term chronoamperometry, the oxidation of the organic ligand and the formation of nickel oxide and/or nickel carbonate as nanocomposites cannot be ruled out.
CuL 3 also acts as an active catalyst for water oxidation in the presence of carbonate at pH 9.0−11.0. The proposed mechanism is a "proton shuttle" mechanism, as shown in Scheme 5. 44 However, the possible oxidation of carbonate to HCO 4 − and C 2 O 6 2− was not considered 44 and cannot be ruled out. Carbonate was also shown to act as a cocatalyst in electrocatalytic water oxidation by an aluminum porphyrin (Al(TMPyP)). 45 This catalytic system forms H 2 O 2 as the major product. The suggested mechanism involves the formation of an Al III −percarbonate complex as the key intermediate. 45

Heterogeneous Catalysis
Bicarbonate and carbonate are involved also in heterogeneous electrocatalytic water oxidation processes via precipitates on the anode. Reactions analogous to reactions 20−34 are expected on the electrodes. Here the role of the carbonate is also dual: it lowers the redox potential of the central cation, and it has a lower oxidation potential than OH − /H 2 O. Processes analogous to reactions 22, 23, 29, and 33 in Scheme 3 require carbonate in the homogeneous medium only to replace the carbonate loss from the precipitate. Processes analogous to reactions 21, 24−27, 31, 32, and 34 in Scheme 3 can also proceed without any carbonate in the precipitate on the electrode: An example of the latter type of catalysis is the report that Ni II (aq) adsorbed on a SiO 2 sol−gel matrix and mixed with graphite acts as an OER electrocatalyst in solutions containing HCO 3 − /CO 3 2− with a current proportional to [HCO 3 − / CO 3 2− ]. 46,47 Clearly one cannot rule out that ligand exchange in the sol−gel matrix transforms the Ni II (aq) into carbonate   3 2− . This film was shown to be a good OER electrocatalyst for which the current increased with [HCO 3 − / CO 3 2− ]. 18 The detailed mechanism causing the electrocatalytic process was not discussed. Clearly reactions analogous to reactions 22 and 26−34 are probably responsible for the electrocatalytic properties of this precipitate. The observation that the film is stable during electrolysis in carbonate solutions that do not contain Fe II ions proves that the mechanism does not involve reduction of the Fe ions in the precipitate to Fe II and dissolution to the aqueous phase.
An analogous study involving nickel carbonate solutions 48 resulted in the formation of an analogous Ni III (HCO 3 − ) n 3−n / Ni III (CO 3 2− ) n 3−2n precipitate on the anode that is a good electrocatalyst for water oxidation in solutions containing Ni II and HCO 3 − /CO 3 2− . However, the precipitate dissolves during electrolysis when no Ni II ions are present in solution. This proves that the catalytic step involves the reduction of a Ni IV complex into a Ni II complex that is soluble, i.e. via reactions analogous to reactions 21, 22, 25, and 26−28 in Scheme 3. In another study, 49 an amorphic nickel carbonate nanowire array on a nickel foam (NiCO 3 /NF) anode was prepared. This modified electrode is a very good electrocatalyst for water oxidation in solutions containing only HCO 3 − /CO 3 2− . 49 The source of the discrepancy between these two studies might be that the nickel foam supplies the needed nickel to preserve the precipitate. The use of a mixture of salts of iron and nickel in carbonate solutions was also reported to form an oxide precipitate that acts as an efficient catalyst in carbonate media to oxidize water. 19  To sum up this section, the results suggest that precipitates of transition metal oxides and/or carbonates, where the central cation can be oxidized to a high oxidation state, on anodes serve as good electrocatalysts for water oxidation in solutions containing HCO 3 − /CO 3 2− . The detailed mechanisms of these catalytic processes depend on the properties of the central cation. The advantage of these catalytic processes is that all of the components are stable inorganic species that are not consumed during the catalytic process.

PHOTOCATALYSIS
The difference in the oxidation potentials of the CO 3 •− /CO 3 2− and OH • /OH − (aq) or (OH • + H 3 O + )/2H 2 O couples suggests that the holes formed photochemically in semiconductors will oxidize HCO 3 − /CO 3 2− faster and in higher yields than the oxidation of water. This was verified computationally in a DFT study of the photocatalysis by GaN (Figure 2). 13 The condition for this is that adsorption of HCO 3 − /CO 3 2− on the surface of the semiconductor, i.e., a relatively high point of zero charge (the pH at which the net charge of the adsorbent's surface is zero or positive), is favorable. The products of these oxidations are either CO 3 •− or C 2 O 6 2− . Indeed, several studies have pointed out that the presence of HCO 3 − /CO 3 2− in the system catalyzes photochemical water oxidation. 14,50,51 Furthermore, it has been shown that photochemical oxidation of SO 2 in aqueous media is enhanced by the presence of carbonate. 11 It was proposed that the formation of CO 3 •− is involved. In the previous section, the role of metal carbonate precipitates as efficient heterogeneous electrocatalysts for water oxidation was reviewed. It seemed to be of interest to check whether these precipitates are photoactive, and indeed, preliminary results 32 point out that precipitates of cobalt and nickel carbonate act as photoelectrocatalysts for water and methanol oxidations.

PERCARBONATE AS AN OXIDATION AGENT
In principle, percarbonates are involved in two types of processes: 1. An inorganic percarbonate salt, e.g., sodium percarbonate (Na 2 CO 4 ·1.5H 2 O 2 ), is used as the oxidizing agent. These percarbonates are used mainly in advanced oxidation processes (AOPs) (vide infra). Their mechanisms of oxidation are suggested to proceed via formation of H 2 O 2 upon dissolution followed by Fenton-like and/or photolytic processes and/or reaction with O 3 , often involving CO 3 •− anion radicals. 52−55 2. The percarbonate is formed in situ via the reaction of H 2 O 2 with a transition metal carbonate complex, analogous to reaction 4 followed by 5, or via reactions analogous to reaction 6 or reaction 69 below. The first peroxocarbonate complex of a transition metal was reported by Hashimoto et al. 56 4 ](ClO 4 ) 3 also involves a similar mechanism. 57 As reactions 53−55 probably do not involve an oxygen atom transfer, it is reasonable to propose that intermediates in which the percarbonate is ligated to the central Mn III are formed (Scheme 6).
Oxidation of calmagite (H 3 CAL) dye in aqueous solution using H 2 O 2 at pH 7.5−9.0 is accelerated considerably in the presence of HCO 3 − . 58 The proposed mechanism involves the following steps:  59 Oxidations of primary alcohols to carbonyl compounds using catalytic amounts of percarbonate and dichromate were also reported. 60

ROLE OF BICARBONATE IN FENTON AND
FENTON-LIKE REACTIONS The Fenton reaction, Fe(H 2 O) 6 2+ + H 2 O 2 , is of major importance. Thus, a search in SciFinder for "Fenton" limited to articles in English for 2019 yielded 2371 results. The major source of this importance is in its role in inducing oxidative stress 61−66 and its role in AOPs. 67 where RH is a substrate. The reactions always proceed via the inner-sphere mechanism.
Still, nearly all of the recent articles cite the Fenton reaction as proceeding via the formation of OH • . Furthermore, recent results point out that in systems where it is difficult to oxidize the central cation, the central cation is not oxidized in the process, and the reaction proceeds via the following mechanism: 72,73 These results point out the following: (a) No OH • radicals are formed. Though a percarbonate ligand is formed as an intermediate in the process, the active oxidizing agent formed is the CO 3 •− anion radical. (b) The central Co II cation is not oxidized during the process. The process is analogous to reactions 65 and 66 with k = 2. Thus, in AOPs with Co II (H 2 O) 6 2+ and H 2 O 2 , the active species is CO 3 •− and not OH • as usually assumed. 75 The mechanism is outlined in Scheme 7. ( The mechanism is shown in Scheme 7. In the Fe Fenton reaction, metal ion oxidation occurs, whereas in the Co Fentonlike reaction, the metal oxidation state does not change. One more difference is that for Co a cyclic percarbonate complex is formed, but this does not occur in the case of Fe. The (CO 3 2− )Fe II (OOH − )(H 2 O) 2 − thus formed decomposes via the following reactions: Thus, also in this system, the reactive oxygen species (ROS) formed under physiological conditions (i.e., in the presence of ∼1.0 mM HCO 3 − ) is not the OH • radical as commonly assumed.
(III) The results concerning the Fenton reaction in the presence of HCO 3 − suggest that under physiological conditions, the Fenton reaction yields CO 3 •− and not OH • radicals as commonly assumed, and this is of major importance. However, since in biological media Fe II ions are not present as Fe II (H 2 O) 6 2+ but appear in the mobile pool mainly as Fe II (citrate), the Fenton reaction was studied 15 where (citrate)Fe II (H 2 O 2 )(HCO 3 − ) 2− is formed in a reaction analogous to reaction 76. It was suggested that this mechanism fits the results better.
The results of these three studies of the mechanisms of Fenton and Fenton-like reactions point out that the role of bicarbonate present in biological systems and in wastewater cannot be overlooked. This conclusion was recently verified. 28 6. OTHER BIOLOGICAL SOURCES OF CO 3

•−
As stated in the Introduction, CO 3 •− is formed biologically via the reaction of peroxonitrite with CO 2 (reactions 11−13). 76 The formation of NO 2 • and CO 3 •− is believed to be the main mechanism via which O 2 •− causes oxidative stress. 77,78 CO 3 •− is also formed enzymatically by the enzymes superoxide dismutase (Cu,Zn-SOD, SOD-1) 79 and xanthine oxidase. 80 The CO 3 •− anion radicals formed in biological systems can oxidize nucleobases [26][27][28]77,81,82 and proteins 83,84 and peroxidize low-density lipoprotein. 85 7. ADVANCED OXIDATION PROCESSES/TECHNOLOGIES The treatment of polluted water and soil is of major importance. Many organic pollutants are treated by advanced oxidation processes/technologies. 86 There is still no optimal treatment for pollutants, and a variety of technologies have been studied. These include chemical oxidation by peroxides, Fenton and Fenton-like processes (including photo-Fenton and electrochemical Fenton), ozone, photochemical processes (including solar light using TiO 2 as a photocatalyst), electrochemical processes, microwave processes, ultrasonic processes, ionizing radiation, hydrodynamic cavitation, and various combinations of these techniques. 86 Carbonate is involved in these processes via the addition of percarbonate as an oxidizing agent, 52 to the system. 97 The role of the added percarbonate is clear. However, the role of the bicarbonate/carbonate present in the solution is more complicated. For many systems it has been proposed that bicarbonate/carbonate reacts with the OH • radical initially formed, thus decreasing the reactivity of the oxidizing species formed but increasing its lifetime and its selectivity in the choice of substrates. 52,94,95 On the other hand, for some systems involving Fenton and photochemical reactions, at least the carbonate is involved in the ROS formation and therefore increases the pollutant decomposition yield. 75 It should be pointed out that the Fenton reaction in neutral solutions in the absence of bicarbonate/carbonate forms Fe IV (aq) and not OH • radicals. 98 This is commonly not addressed in publications on the Fenton reaction discussed in AOPs. Thus, in these systems the presence of bicarbonate/carbonate exchanges Fe IV (aq) with CO 3 •− as discussed above. CO 2 , HCO 3 − , and CO 3 2− are present in all aquatic media at pH > 4 if no effort to remove them is made. Usually one considers their role only as buffers and/or proton transfer agents. The recent results discussed in this Account point out the important role of HCO 3 − /CO 3 2− in a variety of catalytic oxidation processes in aqueous media. This role is due to the following properties of carbonates:

CONCLUSIONS AND PERSPECTIVES
1. Carbonate is a strong hard ligand that stabilizes transition metal complexes in high oxidation states. These complexes are key intermediates in electrochemical water oxidation processes. 2. The redox potential of the CO 3 •− /CO 3  The results obtained thus far point out that the presence of HCO 3 − /CO 3 2− dramatically changes the mechanisms of the Fenton and Fenton-like reactions. These findings require reassessment of the major sources of oxidative stress in biological systems and of the role of carbonate anion radicals in oxidative stress. Furthermore, the results point out the activity of HCO 3 − /CO 3 2− as catalysts/cocatalysts for water oxidation electrochemically and photochemically. This is of importance in developing technologies for efficient water splitting processes.
Extremely little has been done to date on the application of HCO 3 − /CO 3 2− in catalytic oxidations of specific substrates performed electrochemically, photochemically, and photoelectrochemically. Also, the role of HCO 3 − /CO 3 2− in Fentonlike processes using other peroxides (e.g., S 2 O 8 2− , HSO 5 − , and ROOH, where R is an aliphatic residue) has not been studied. These studies might be of importance in the development of new advanced oxidation technologies. https://pubs.acs.org/10.1021/acs.accounts.0c00344

Notes
The authors declare no competing financial interest. He is now Professor of Inorganic Chemistry at Ariel University and Professor Emeritus at Ben-Gurion University. His research interests include reaction mechanisms, catalysis, radical processes, radiation chemistry, electrochemistry, and nanochemistry.