RuIII(edta) complexes as molecular redox catalysts in chemical and electrochemical reduction of dioxygen and hydrogen peroxide: inner-sphere versus outer-sphere mechanism

The reduction of molecular oxygen (O2) and hydrogen peroxide (H2O2) by [RuII(edta)(pz)]2− (edta4− = ethylenediaminetetraacetate; pz = pyrazine) has been studied spectrophotometrically and kinetically in aqueous solution. Exposure of the aqua-analogue [RuII(edta)(H2O)]2− to O2 and H2O2 resulted in the formation of [RuIII(edta)(H2O)]− species, with subsequent formation of the corresponding RuV 
<svg xmlns="http://www.w3.org/2000/svg" version="1.0" width="13.200000pt" height="16.000000pt" viewBox="0 0 13.200000 16.000000" preserveAspectRatio="xMidYMid meet"><metadata>
Created by potrace 1.16, written by Peter Selinger 2001-2019
</metadata><g transform="translate(1.000000,15.000000) scale(0.017500,-0.017500)" fill="currentColor" stroke="none"><path d="M0 440 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z M0 280 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z"/></g></svg>
 O complex. A working mechanism for the O2 and H2O2 reduction reactions mediated by the RuII(edta) complexes is proposed. The role of the coordinated water molecule (by its absence or presence in the primary coordination sphere) in controlling the mechanistic pathways, outer-sphere or inner-sphere, is discussed.


Introduction
The electrochemical oxygen reduction reaction (ORR) proceeds by two-electron two-proton (2e À /2H + ) partial reduction of O 2 to produce H 2 O 2 (eqn (1) in Scheme 1) or direct four-electron fourproton (4e À /4H + ) reduction of O 2 to 2H 2 O (eqn (3) in Scheme 1). Although selective reduction of O 2 directly to H 2 O is of continued interest in regard to its application in energy conversion, particularly in the eld of fuel cells and metal-air batteries, 1-4 production of H 2 O 2 via two-electron two-proton reduction of O 2 is also of considerable importance for environmental application like waste water treatment and chemical feedstocks. 5,6 Use of transition metal complexes as molecular catalysts as a redox mediator to affect the reduction of oxygen is well documented in the literature. [7][8][9][10][11][12] Many schemes of catalytic processes that affect reduction of O 2 in combination of 2e À /2H + and/or 4e À /4H + pathways, either selectively or sequentially, have been reported. Noteworthy here, is that mechanistic details and kinetic parameters that control the efficiency of the fourelectron reduction of O 2 to H 2 O versus two-electron partial reduction of O 2 to H 2 O 2 are still lacking, even though exhaustive efforts have been devoted for more than the last two-decades in mimicking and understanding the enzymatic activity of cytochrome c oxidase (which catalyses the direct four electron reduction of O 2 to H 2 O during the nal stage of respiration 13 ). Another important aspect is, that although the reduction of hydrogen peroxide to water (eqn (2)) is seemingly easier thermodynamically than the reduction of dioxygen to hydrogen peroxide (eqn (1)), it is kinetically very difficult as it involves the cleavage of the O-O bond.
While mononuclear as well as binuclear complexes of copper, iron, cobalt and manganese, have been exhaustively studied pertaining to the oxygen reduction reaction, 7-12 the use of the ruthenium complex in this context is scanty in the literature. 14 In the present work, we set out to examine the ability of the Ru II (edta) complex (edta 4À ¼ ethylenediaminetetraacetate) to affect the reduction reaction of molecular oxygen in aqueous medium. The feature that dominates the chemistry of the Scheme 1 Pictorial presentation of proton-coupled electron transfer reactions for dioxygen reduction. SHE ¼ Standard Hydrogen Electrode.
[Ru III (edta)(H 2 O)] À complex is its lability towards aquasubstitution reactions, 15 which affords an advantage of facile and straightforward binding of substrate molecules to the metal centre. In addition, a range of accessible and stable oxidation states made Ru(edta) complexes abidingly important to the catalytic studies for the past two decades. The signicance of the Ru(edta) complexes in mimicking enzymatic redox reactions and small molecule activation, have been well established in very recent review articles. 16,17 We for the rst time explore, that [Ru II (edta)(pz)] 2À (pz ¼ pyrazine) and its aqua-analogue [Ru II (edta)(H 2 O)] 2À , can efficiently mediate the sequential 2e À /2H + reduction of O 2 to H 2 O 2 and further reduction of H 2 O 2 to H 2 O. We report herein the results of the detailed spectral and kinetic investigation of the reduction of O 2 and H 2 O 2 by the above referred Ru II (edta) complexes.

Materials
The starting complex K[Ru III (Hedta)Cl]$2H 2 O was prepared and characterized as described elsewhere. 18 The complex K[Ru III (-Hedta)Cl] instantaneously converts into [Ru III (Hedta)(H 2 O)] when dissolved in water. 19,20 The pK a values associated with the acid-dissociation equilibria of the pendant carboxylic acid arm and the coordinated water molecule, are 2.4 and 7.6, respectively, at 25 C. 19,20 All other reagents and buffer components used were of the highest grade commercially available and were used as received. Doubly distilled H 2 O was used to prepare all solutions.

Instrumentation
The reactions were studied applying UV-Vis spectroscopy. Fast kinetic measurements were performed using a stopped-ow (Applied Photophysics SX20) equipped with a rapid-scan diode-array spectrometer with a multi-wavelength J&M detector. The solution temperature was maintained at the desired temperature within AE0.1 C using a Colora thermostat. The time courses of the slow reactions were followed spectrophotometrically, adopting a conventional mixing technique using the HP 8453 diode-array spectrophotometer. A tandem cuvette was used for this purpose. This instrument was thermostated at the desired temperature (AE0.1 C) using a HP 89090 Peltier Temperature Controller. Since, Ru II (edta) complexes are very air sensitive, all experimental solutions of the Ru IIcomplexes were prepared strictly under Ar atmosphere. Gastight Hamilton syringes were used to transfer these solutions throughout the studies. Preparation of oxygen saturated solutions was accomplished by bubbling pure oxygen through the deaerated double-distilled water for 30 min. This solution was carefully diluted by mixing with deaerated water in desired proportion. The pH of the buffer solution was measured with an Elmetron CP-505 pH meter. Acetate buffer (0.1 M) was used to control the pH of the experimental solutions. Several kinetic runs (at least three) were performed to obtain reproducible results within AE4%.

Results and discussion
The 'edta' ligand functions as a pentadentate ligand towards Ru(III) with a pendant acetate arm. 21 The [Ru III (edta)H 2 O] À complex reacts with pyrazine (pz) to form the [Ru III (edta)(pz)] À complex, through a rapid and straightforward water displacement reaction (k f ¼ 2 Â 10 4 M À1 s À1 and k r ¼ 2 s À1 at 25 C) 19 as outlined in eqn (4).
The electronic absorption spectrum of the [Ru III (edta)(pz)] À complex in aqueous solution is featureless in the entire visible range, whereas its Ru(II)-analogue, [Ru II (edta)(pz)] 2À (Fig. 1) displays a strong band in the visible range (l max ¼ 462 nm, 3 max ¼ 11 600 M À1 cm À1 ) which was assigned to a metal to ligand charge transfer (MLCT) band. 19 This huge spectral difference thus offers an amenable way to follow the kinetics of the electron transfer reactions spectrophotometrically.

Reduction of O 2 by [Ru II (edta)(pz)] 2À
Addition of an oxygen-saturated aqueous solution to the deaerated red solution of the [Ru II (edta)(pz)] 2À complex (see S1 in ESI †) at pH 5.0 (acetate buffer), resulted in a gradual disappearance of the red colour.
The overall spectral changes recorded immediately aer mixing the solutions of [Ru II (edta)(pz)] 2À and aqueous solution of dissolved oxygen, are shown in Fig. 2a. The spectral changes are attributed to the oxidation of [Ru II (edta)(pz)] 2À to [Ru III (edta)(pz)] À , and a typical kinetic trace recorded at 462 nm (decay) is shown in Fig. 2b.
Effect of the concentration of the dissolved oxygen on the rate of the reaction was studied at 25 C and pH 5.0 (representative kinetic traces recorded at 462 nm are shown in  in ESI †). Under the specied conditions, the rate of the reaction estimated by the maximum slope, increases linearly with increasing concentration of dissolved oxygen ( Fig. S2 in ESI †). The spectral and kinetic observations can be accounted for in terms of the reaction sequence outlined in Scheme 2. The ratedetermining step (5) proposed in the mechanism, involves a one-electron transfer from [Ru II (edta)(pz)] 2À to the O 2 to yield O 2 À c radical species in an outer-sphere manner. In the subsequent and kinetically inconsequential step (eqn (6)), the O 2 À c radical rapidly reacts with another molecule of [Ru II (edta)(pz)] 2À to produce the peroxide ion (O 2 2À ), protonation of which (eqn (7)) results in the formation of hydrogen peroxide (H 2 O 2 ). The reduction of O 2 with [Ru II (edta)(pz)] 2À under the speci-ed conditions can be accounted for in terms of the rate-law expressed by eqn (8).
The value of the second-order rate constant (k 1 ) estimated from the slope of the plot of rate versus [O 2 ] shown in Fig. S2 (in ESI †) is 0.14 AE 0.01 M À1 s À1 at 25 C. Addition of fresh ascorbic acid to the resultant solution obtained at the end of the aforementioned reaction (experimental conditions given under Fig. 2), resulted in the formation of the [Ru II (edta)(pz)] 2À almost quantitatively as evidenced by the spectral measurements ( Fig. S3 in ESI †). The above observations clearly indicate the existence of a catalytic process in the overall reactions, wherein dioxygen (O 2 ) is reduced to hydrogen peroxide (H 2 O 2 ) via an electron transfer reaction, and the [Ru III (edta)(pz)] À complex acts as a redox relay for electron transmission from ascorbic acid to O 2 .

Reduction of H 2 O 2 by [Ru II (edta)(pz)] 2À
In order to understand the reaction of [Ru II (edta)(pz)] 2À with H 2 O 2 (formed during the reduction of O 2 by [Ru II (edta)(pz)] 2À as shown in Scheme 2), we performed a detailed kinetic study of the reaction of [Ru II (edta)(pz)] 2À with H 2 O 2 discretely under similar conditions of pH (5.0) and temperature (25 C).
In Fig. 3a typical UV-visible spectral changes with time are shown (recorded by using stopped-ow rapid scan, diode array spectrophotometer) that occurred upon mixing an aqueous solution of [Ru II (edta)(pz)] 2À with the solution of H 2 O 2 (in acetate buffer). The overall kinetic trace ( Fig. 3b) derived from the time-resolved spectral changes, exhibited three clear steps (two decay and one growth at higher [H 2 O 2 ]) marked as I, II and III, respectively). The rst decay step involves a small decrease in absorbance, not of enough signicance (less than 5% as compared to the total absorbance change in the overall reaction time). Kinetic traces (see Fig. S4 in ESI †) pertinent to step I analysed on a shorter time scale as a function of [H 2 O 2 ], are seemingly exponential in nature and could be tted with a single exponential function within the precision of experimental data (R > 0.99). The values of the observed rate constant (k obs ¼ 0.18 AE 0.02 s À1 at 25 C and pH 5.0) so obtained were found to be independent of the H 2 O 2 concentration. The spectral changes involved in step I may be attributed to the formation of a very weak [Ru II (edta)(Ac)] 3À (Ac ¼ acetate) complex by the reaction of [Ru II (edta)(pz)] 2À with buffer component    (Fig. 3b). Nevertheless, the red [Ru II (edta)(pz)] 2À is not inactive for this apparent dormant period, but rather at a steady-state concentration under the employed conditions. The [Ru III (edta)(pz)] À complex (produced via oxidation of [Ru II (edta)(pz)] 2À by H 2 O 2 ) is rapidly reduced with ascorbic acid 23 (present in excess; see S1 in ESI †) to reform the [Ru II (edta)(pz)] 2À species back in the reaction mixture. Such reaction cycles were continued until the ascorbic acid present in the reacting system is fully consumed. The effect of [H 2 O 2 ] on the rate of the disappearance of the [Ru II (edta)(pz)] 2À complex (estimated from the maximum slope of the absorbance vs. time plots given in Fig. 3b), is shown in Fig. 4.
Based on the above experimental facts, particularly the attainment of a limiting rate at higher [H 2 O 2 ] (Fig. 4), the following working mechanism involving a rapid preequilibrium, is proposed in Scheme 3 for the reaction of [Ru II (edta)(pz)] 2À with H 2 O 2 .
As seen in Fig. 3a, aer complete disappearance of the peak at 462 nm (attributed to the oxidation of the [Ru II (edta)(pz)] 2À to [Ru III (edta)(pz)] À ) under the specied conditions (see Fig. 3a), the reaction is followed by a step that involves the formation of a band at 390 nm. This new band is characteristic of the [Ru V (edta)O] À complex (l max ¼ 390 nm; 3 max ¼ 8 Â 10 3 M À1 cm À1 ). 24 The observed increase in the absorbance at 390 nm with time at higher H 2 O 2 concentration (step III in Fig. 3b), recorded aer 95 s delay, is shown separately for clarity in Fig. 5  In Fig. 6 the UV-vis spectral changes with time that occurred upon mixing aqueous solutions of [Ru III (edta)(pz)] À and H 2 O 2 , are shown. The observed spectral changes (Fig. 6) are attributed to the oxidation of [Ru III (edta)(pz)] À to [Ru V (edta)O] À under the specied conditions. The kinetic traces at 390 nm, generated from the recorded spectra, are presented in Fig. S6 (16)) in a concerted pathway. However, the metal-pyrazine bond at a higher oxidation state of the metal in [Ru V (edta)(pz)O] À becomes less stable, and the pacidic ligand pyrazine, thereby dissociates allowing the dangling carboxylate group to bind to the metal centre again to produce the [Ru V (edta)O] À product complex in a kinetically inconsequential step (eqn (17)). The following rate-law (eqn (18) and (19)) can be derived for the reactions in Scheme 4 on the basis of the rate-determining formation of the [Ru V (edta)O] À complex.
The plot of 1/k obs versus 1/[H 2 O 2 ] is linear (Fig. 7b), and the values of k 3 and K 2 from the intercept and slope are 0.23 AE 0.03 s À1 and 82 AE 2 M À1 , respectively, at 25 C and pH 5. The product  O as ultimate reaction products, proceeds through an inner-sphere pathway demonstrating similar kinetic features. Although the common suggested mechanism involves heterolytic cleavage of the O-O bond, the ability of H 2 O 2 to bind to the Ru(III)-centre through a ligand substitution process, governs the efficiency of the H 2 O 2 reduction process. In this regard, [Ru III (edta)(H 2 O)] À due to its unusual lability towards a substitution reaction, has an advantage over [Ru III (edta)(pz)] À .

Reaction of O 2 and H 2 O 2 with [Ru III (edta)(H 2 O)] À
Noteworthy here, is that the [Ru III (edta)(H 2 O)] À complex also exhibited a metal based one-electron transfer reaction, and the E 1/2 value reported for the [Ru III (edta)( 19 which is much more negative than that reported for the [Ru III (edta)(pz)] À /[Ru II (edta)(pz)] 2À couple (E 1/2 ¼ 0.252 V vs. SHE). 19 Though [Ru II (edta)(H 2 O)] 2À is a stronger reductant thermodynamically than [Ru II (edta)(pz)] 2À towards O 2 reduction, detailed kinetic investigations were practically not feasible because of very insignicant spectral differences between aquaanalogues of Ru(III) and Ru(II)-edta complexes. The [Ru III (edta)(H 2 O)] À in water is almost featureless over the entire visible range of the spectrum, but exhibits a strong absorption band at 280 nm (3 max ¼ 2800 AE 50 M À1 cm À1 ) and a shoulder at 350 nm (3 max ¼ 680 AE 60 M À1 cm À1 ) in the UV region. 19 The Ru(II) analogue, [Ru II (edta)(H 2 O)] 2À displays almost similar spectral features exhibiting an intense band at 282 nm (3 max ¼ 2900 AE 100 M À1 cm À1 ) and a weak shoulder at 427 nm (3 max ¼ 260 AE 15 M À1 cm À1 ). 19 Nevertheless, the formation of the [Ru V (edta)O] À complex was evidenced in the spectral changes ( Fig. 8) that occurred in the reaction of [Ru II (edta)(H 2 O)] 2À with oxygen. Time resolved spectral changes recorded (Fig. 8) aer mixing of the aqueous solution of [Ru II (edta)(H 2 O)] 2À (0.025 mM) with an oxygen saturated aqueous solution, clearly revealed the gradual build-up of the peak at 390 nm (inset of Fig. 8) characteristic for the [Ru V (edta)O] À complex. 24 The above spectral observations may be explicable in terms of the following reactions as outlined in Scheme 5.
In the above proposed mechanism (admittedly speculative), the reduction of O 2 to H 2 O 2 with concomitant formation of the [Ru III (edta)(H 2 O)] À takes place in an outer-sphere pathway (eqn (20)). The appearance of the peak at 390 nm in Fig. 8  , is given in Fig. 9. As it is seen, the kinetic trace exhibits a clear initial induction period. Above observations may be explicable in terms of the following reaction scheme (Scheme 6) proposed for the reduction of The observed initial induction period (Fig. 9) (22)) rapidly undergoes reduction by the ascorbic acid present in excess in the reacting system (see S1 in ESI †), regenerating the [Ru II (edta)(H 2 O)] 2À species in the reaction mixture. Such reaction cycles sustained until the ascorbic acid present in the reacting system are completely   Fig. 11a. Constant potential electrolysis of the solution of [Ru III (edta)(pz)] À leading to the formation of the [Ru II (edta)(pz)] 2À species was evident spectrophotometrically by the appearance of its characteristic peak at 462 nm (Fig. 11b). The typical current versus time plot pertaining to the above mentioned electrochemical process is shown in the inset of the Fig. 11. Aer withdrawal of the potential, O 2 was bubbled through the electrolysed solution for 300 s. The spectral changes thereaer (Fig. 11c) show appreciable collapse of the peak at 462 nm, which is consistent with the re-oxidation of [Ru II (edta)(pz)] 2À to [Ru III (edta)(pz)] À by dioxygen (O 2 ). Electrolysis of the solution (at À0.05 V) for 500 s again regenerates the [Ru II (edta)(pz)] 2À species as evident spectrophotometrically (Fig. 11d). However, the intensity of the band at 462 nm as noticed in the spectrum (Fig. 11d) is signicantly smaller in comparison to that observed in the spectrum (Fig. 11b). This may plausibly be associated with the fact that the unconsumed dissolved oxygen present in the reacting system prior to the second run of electrolysis, may compete with the electrochemical reduction process for which complete formation of the [Ru II (edta)(pz)] 2À could not take place within the time period the voltage was on (500 s). Observation of the band (at 462 nm), with a higher absorbance in the spectrum of the electrolysed solution, which was deoxygenated through argon purging prior to the electrolysis (see Fig. S7 in ESI †), supports our above argument.  Based on the above observations, taken together with that reported for the reaction of [Ru II (edta)(pz)] 2À with molecular oxygen (O 2 ) in the preceding section, the role of [Ru III (edta)(pz)] À as an electron transfer redox catalyst in the electrochemical reduction of O 2 may be outlined in Scheme 8.

Conclusions
In conclusion, the results of the present study reveal that in the overall reactions, whether chemical or electrochemical, dioxygen (O 2 ) is reduced to hydrogen peroxide (H 2 O 2 ) via electron transfer reaction, and the [Ru III (edta)(pz)] À complex acts as a redox relay for electron transmission. In case of the chemical process, it takes electrons from ascorbic acid to reduce O 2 to H 2 O 2 , whereas in case of the electrochemical process, it uses the electrons from the working electrode to effect electrochemical reduction of O 2 to H 2 O 2 in aqueous acidic solution. The results of our studies further ascertain that both the [Ru II (edta)(pz)] 2À and [Ru II (edta)(H 2 O)] 2À complexes in presence of electron donors can reduce O 2 to H 2 O 2 and H 2 O 2 to H 2 O efficiently in a sequential manner. The redox mediating properties of the aforesaid Ru(edta) complexes along with their wide range of chemically accessible oxidation states (II to V), and their durability in the redox processes, are indeed intriguing and prospective. Our results may shed light towards mechanistic understanding of the homogeneously catalysed reduction of O 2 and H 2 O 2 and provide pointers for future research pertaining to the application of such metal complexes in the eld of fuel cells and metal-air batteries.

Conflicts of interest
There are no conicts to declare.