Substrate oxidation enhances the electrochemical production of hydrogen peroxide

Graphical abstract


Introduction
Hydrogen peroxide (H 2 O 2 ) is often used as an eco-friendly oxidant because it is reduced to water as an electron acceptor and readily decomposes to water and oxygen (O 2 ) [1]. The high oxidation potential (E 0 = +1.76 V NHE ) of H 2 O 2 allows for the direct oxidation of certain organic and inorganic electron donors and for the indirect oxidation by hydroxyl radical produced via the UV photolysis of H 2 O 2 or by Fentonreagent activation [2]. Hydrogen peroxide is also used in organic syntheses, liquid fuel rocket propulsion, disinfection, and environmental remediation [1]. Moreover, H 2 O 2 is frequently used in advanced oxidation processes for water or air [3][4][5].
Hydrogen peroxide is normally produced by the anthraquinone method involving a multistep oxidation of 2-ethyl-9,10-dihydroxyanthracene and the subsequent hydrogenation of 2-ethylanthraquinone [6]. However, this method is not environmentally benign because hydrogen (H 2 ) gas, organic solvents, and high energy inputs are required. The direct reaction between the H 2 and O 2 gas using metal-based catalysts (e.g., Au or Pd/Au alloys) in acid or methanolic solutions has been investigated as an alternative method for H 2 O 2 production but this method is also not environmentally and economically viable [7,8] due to the explosion potential of the H 2 and O 2 gaseous mixture [9,10]. In contrast, the electrochemical production of H 2 O 2 via a two-electron transfer to O 2 is relatively benign synthetic method since it takes place at low temperatures and pressures [11].
In order to improve the overall yield of H 2 O 2 during the electrochemical production, the decomposition of H 2 O 2 should be substantially reduced. For example, carbon-based cathodes have been used since they inhibit the cathodic decomposition of H 2 O 2 [18]. However, the strategy for preventing the anodic decomposition of H 2 O 2 has received little attention.
In this study, we investigate the effect of organic electron donors on the electrochemical production of H 2 O 2 as a method to inhibit the anodic decomposition of H 2 O 2 . Our hypothesis is that the production of H 2 O 2 should be enhanced in the presence of organic electron donors at a constant cathodic potential. Organic substrates should prevent the anodic decomposition of H 2 O 2 with respect to further anodic oxidation to superoxide and oxygen. The impact of experimental variables including applied voltage, pH, and probe reagents on the production of H 2 O 2 in the presence of specific organic electron donors is explored.

Materials and chemicals
A dimensionally-stable anode consisting of Ir 0.7 Ta 0.3 O 2 formed during in situ spray pyrolysis of precursor reagents on a heated titanium metal substrate and over-coated with TiO 2 was used (TiO 2 /Ir 0.7 Ta 0.3 O 2 / Ti). This composite anode formulation has been shown to be active with respect to both the chlorine and oxygen evolution reactions [19]. The aqueous-phase precursor solutions were composed of 3.5 mM IrCl 3 and 1.5 mM TaCl 5 in isopropanol for formation of the Ir 0.7 Ta 0.3 O 2 layer, while the TiO 2 overcoating layer was formed using a 25 mM titaniumglycolate solution for deposition of the overcoating TiO 2 layer was deposited by spray coating of the solution directly on to Ti foil heated to 300°C. The resulting film was annealed at 500°C for 10 min. These procedures were repeated to reach a targeted mass loading. Upon achieving the desired mass loading, the final composite was annealed at 500°C for 1 h. Chemical reagents used in this study were as follows: sodium sulfate (Na 2 SO 4 , Sigma-Aldrich), bisphenol A (BPA, Aldrich), phenol (J. T. Baker), 4-chlorophenol (4-CP, Sigma-Aldrich), coumarin (Sigma), potassium bis(oxalato)-oxotitanate (IV) dihydrate (K 2 [TiO (C 2 O 4 ) 2 ]•2H 2 O, Alfa Aesar), sulfuric acid (H 2 SO 4 , J. T. Baker), hydrogen peroxide (H 2 O 2 (30 wt%), Sigma-Aldrich). All chemical reagents were used as received without any purification. Deionized water was used as solution and prepared by a Millipore system (≥18 MΩ % cm, Milli-Q).

Electrochemical experiments
A three-electrode configuration including a working electrode (graphite rod, diameter 6 mm), a counter electrode (TiO 2 /Ir 0.7 Ta 0.3 O 2 / Ti), and a reference electrode (Ag/AgCl) was employed in a single compartment cell with a working volume of 25 mL. The background electrolyte was a 60 mM aqueous solution of Na 2 SO 4 . The optimum concentration of Na 2 SO 4 was found to be 60 mM in terms of electrochemical efficiency (Fig. S1). As a consequence, the electrochemical reactions were primarily run in the 60 mM Na 2 SO 4 electrolyte solution. The distance between the anode and cathode was 13 mm. During testing for the simultaneous electrochemical production of H 2 O 2 and the concomitant degradation of organic substrates, a constant cathodic potential was applied to the electrodes using a computer-controlled potentiostat (SP-50, BioLogic). An aliquot of a substrate stock solutions (BPA, phenol, and 4-CP) was added to the electrolyte to give an establish a pre-set concentration of the target substrate. The initial pH was adjusted to a set value using either 1.0 M HClO 4 or 1.0 NaOH solutions. Oxygen was purged in to the reactor for 30 min before application of a constant potential and then continuously purged during the course of electrolysis. Nitrogen (N 2 ) gas purging of the aqueous solutions was carried out when low concentrations of dissolved oxygen were required. Aliquots of 1 mL were intermittently withdrawn from the reactor using a 1-mL pipet and were transferred into a glass vial without filtration for the analysis of the concentration of H 2 O 2 and organic pollutants. Cyclic voltammetry (CV) data were collected in the Na 2 SO 4 solution in the potential range of −0.8 to 0.0 V at a scan rate of 50 mV s −1 .

Analysis
The concentrations of BPA, phenol, and 4-CP were quantitatively analyzed using a high performance liquid chromatograph (HPLC, Agilent 1100 series) equipped with a Zorbax XDB column. The HPLC measurement was carried out using a binary mobile phase of acetonitrile and phosphoric acid (30%:70% for BPA and 10%:90% for phenol and 4-CP). Chloride produced by 4-CP degradation was monitored using an ion chromatograph (IC, Dionex, USA) with an anion-exchange column (Ionpac AS 19). The total organic carbon (TOC) was analyzed using a TOC analyzer (Aurora TOC). The production of % OH was monitored using coumarin as a chemical trap of % OH. Coumarin is oxidized by hydroxyl radical to form 7-hydroxycoumarin [20]. The hydroxylated product, 7-hydroxycoumarin, was quantified by measuring the fluorescence emission intensity at λ em = 456 nm after excitation at λ ex = 332 nm·H 2 O 2 was determined spectrophotometrically using potassium titanium (IV) oxalate [21]. The absorbance at 400 nm (ε = 9351 mol −1 cm −1 ) was measured using a UV/Visible spectrophotometer (Nanodrop 2000c).  [22,23]. The reduction peak at −0.4 V disappeared in the N 2purged solution as observed previously [11,22]. The degradation of BPA was also reduced in the absence of O 2 (Fig. 1a). Fig. 1b shows the production of H 2 O 2 coupled with BPA degradation as a function of the applied potential. The efficiencies for the production of H 2 O 2 and degradation of BPA were increased with increasing the applied potential. H 2 O 2 was not produced in the absence of an external potential bias, whereas the [BPA] was slightly reduced (Fig. S3). This result is most likely due to the adsorption of BPA on to the surface of anode at pH 3. Fig. 1c compares the production of H 2 O 2 as a function of the BPA concentration. The electrochemical generation of H 2 O 2 was increased with an increasing concentration of BPA. In particular, H 2 O 2 was continuously produced in the presence of BPA, whereas its generation reached an apparent steady-state level in the absence of BPA after 1 h of electrolysis. This steady-state is achieved due to the in situ decomposition of the H 2 O 2 [23]. During repeated electrolytic cycles, a loss of the activity for BPA degradation and H 2 O 2 production was not observed four catalytic cycles. However, cycling for more than four cycles resulted in a small loss of activity (Fig. 1d). This result can be ascribed to active site blocking on the electrode surface due to the adsorption of BPA and its reaction product intermediates generated during BPA degradation.

Influence of BPA on H 2 O 2 decomposition
The effects of BPA on the kinetics of decomposition of H 2 O 2 were determined by following the change in concentration of 5 mM of hydrogen peroxide in the electrolyte solution in the presence and absence of BPA (Fig. 2a). The decomposition of H 2 O 2 in the absence of BPA was faster than that observed in the presence of BPA. Although the decomposition of H 2 O 2 was significantly reduced at E app = 0.0 V compared to E app = −0.5 V, it was also found to be faster in the absence of BPA as shown in Fig. S4. The rate constant for H 2 O 2 formation and decomposition were treated in terms of zero-order kinetics for production and first-order kinetics for decay, respectively [17]. The formation rate was increased and the decomposition rate was reduced in the presence of BPA compared to the absence of BPA (Fig. 2b). To further clarify the effect of BPA on the production of H 2 O 2 , excess BPA (1 mM) was added into the electrolyte during the course of electrolysis (after 1 h). The electrochemical production of H 2 O 2 was enhanced by 55% (188 → 418 µM at 2 h) and the cathodic current was slightly increased (Fig. 2c)   hydroperoxo species (^Ti-OOH) through the adsorption on the protective overlayer of TiO 2 of the anode (reaction (2) in Scheme 1a) [17,25]. We were able to confirm this possibility as shown in Fig. S4. However, we did not observe any noticeable difference and characteristic peaks related with hydroperoxo species in either the Raman or the XPS spectra of the anode before and after electrochemical reactions (data not shown). This may be due to the fact that surface-bound hydroperoxyl species are rapidly converted into HO 2 % and O 2 (Eqs. (5) and (6) and reaction (3) in Scheme 1a) [24,26]. Furthermore, H 2 O 2 is decomposed by % OH generated on the surface of the anode (Eq. (7) and reaction (4) in Scheme 1a) [27]. On the other hand, the anodic decomposition of H 2 O 2 appears to be significantly inhibited in the presence of BPA (Scheme 1b). The formation of hydroperoxo species is most likely prevented due to the sorption of BPA is instead of H 2 O 2 on the outer surface of the anode (reaction (1) in Scheme 1b and Fig. S3).
In addition, BPA acts as a scavenger of % OH (reaction (2) in Scheme 1b) given the second-order rate constant for % OH + BPA ⟶ k = 6.9 × 10 9 M −1 s −1 . In comparison, the corresponding rate constant for % OH + H 2 O 2 ⟶ k = 3.2 × 10 7 M −1 s −1 [28]. The competition between the two decomposition pathways in the presence of BPA at the anode surface has net effect of allowing for the solution phase concentration of H 2 O 2 to increase with time until reaching a steady-state condition.

Effect of pH and electrolytes
The rates of H 2 O 2 production and BPA degradation as a function of the initial pH of the electrolyte solution are shown in Fig. 3a. From this data, it is clear that the rate of H 2 O 2 production decreased with increasing pH. These results may be due to the role of proton coupled electron transfer (PCET) to O 2 (Eq. (1)) [29] leading on the electrochemical formation of H 2 O 2 via O 2 reduction at the cathode. The decrease in the rate of BPA degradation was minimal over the pH range of 3-7. However, the degradation rate decreased between pH 9 and 11. The increased in reaction rate with pH can be directly correlated with pK a of BPA and the corresponding surface charge distribution of the TiO 2 layer of the anode. The deprotonation of BPA (pK a1 = 9.6 and pK a2 = 10.2 yields the conjugate bases (HBPA − ) and (BPA 2− )) [30] results in an electrostatic repulsion of the anionic BPA species from the negatively charged TiO 2 surface (pH pzc (point of zero charge) ≈ 6.0 [31]) at higher pH. The electrostatic repulsion thus inhibits the adsorption of BPA on the anode surface at higher pH and thus the increased decomposition of H 2 O 2 . On the other hand, the adsorption of BPA on the anode surface facilitates the degradation of BPA under acidic and circum-neutral pH compared to alkaline pH. Furthermore, competitive adsorption of BPA on the anode surface inhibits the adsorption of H 2 O 2 on the anode at lower pH resulting in the reduced anodic decomposition of H 2 O 2 (see Fig. 2b). The pH-dependent results indicate that low pH conditions are more favorable for the proton-assisted electrochemical production of H 2 O 2 coupled with the degradation of BPA.
The impact of the background electrolyte on the rates of H 2 O 2 production and BPA degradation was examined as shown in Fig. 3b. The rates of H 2 O 2 production and BPA degradation were slightly decreased in the NaClO 4 electrolyte solution compared to our reference electrolyte Na 2 SO 4 . In contrast, the degradation of BPA was enhanced in the NaCl electrolyte solution compared to Na 2 SO 4 [32]. The enhanced BPA degradation in a NaCl electrolyte is due to the anodic production of reactive chlorine species (RCS) (e.g., chlorine radical (Cl % ), dichloride radical anion (Cl 2 %− ), hypochlorous acid (HOCl), and hypochlorite (OCl − )) oxidatively generated on the surface of anode in the presence of NaCl (Eqs. (8)-(12)) [33,34].
HOCl ⇌ OCl − + H + (12) Given a sufficient applied potential, electron-hole pairs are formed and migration of a hole (h + ) to a surface titanol group (> TiOH) leads to the formation of surface bound hydroxyl radical. However, the production of RCS has a negative effect on the electrochemical production of H 2 O 2 . The reactive chlorine species contribute collectively to the decomposition of H 2 O 2 (Eqs. (13)-(16)) [27]. In experiments described herein, the decomposition of H 2 O 2 was accelerated at a high pH compared to low pH (Fig. S5). It is clear that H 2 O 2 reacts faster with RCS (e.g., − OCl) under alkaline conditions compared to those at lower pH. For example, the bimolecular rate constant for reaction of H 2 O 2 with hypochlorite (HOCl; pK a = 7.6) (Eq. (12)) is substantially higher at high pH (7.5 × 10 3 M −1 s −1 ) compared to circum-neutral pH (196 M −1 s −1 ) [35]. Furthermore, the RCS generated on the anode surface are reduced back to chloride on the cathode surface (Eqs. (17)-(20)) [33]. Cathodic chloride reduction is competitive with the reduction of O 2 leading to H 2 O 2 production. Therefore, the rate of H 2 O 2 production is significantly reduced in the presence of NaCl compared to Na 2 SO 4 (Fig. 3b) consistent with the following set of reactions. HClO

Mechanism of BPA degradation
Direct electron transfer to a surface-trapped hole may contribute to BPA degradation. In order to confirm this possibility, the cathodic current was measured in the presence and absence of BPA under oxic O 2 and then under anoxic N 2 conditions (Fig. S6). Under these conditions, the cathodic current was slightly increased in the presence of BPA compared to the absence of BPA under both O 2 and N 2 . These results imply that direct electron transfer to a surface-trapped hole provides a minor pathway for BPA degradation. The rate of BPA degradation in an NaClO 4 electrolyte solution was found to be slightly reduced compared to the same reaction conditions in the Na 2 SO 4 electrolyte (see Fig. 3b). This result suggests that SO 4 %− may have been produced via anodic sulfate oxidation. Another possible oxidant in the system is the surfacebound hydroxyl radical ( % OH) that is generated via surface titanol group oxidation (i.e., > Ti-OH + h + ) on the hydrated TiO 2 surfaces of anode (Eq. (8)) [27]. To confirm the role of % OH, BPA degradation was tested in the presence of tert-butanol (t-BuOH) and methanol (MeOH) as preferential radical scavengers of % OH and SO 4 %− , respectively (Fig. 4a) (Fig. 4a). This result demonstrates that BPA is mainly degraded by % OH or surface-bound hydroxy radical, > TiOH % , which is consistent with the finding that BPA degradation in a Na 2 SO 4 solution was similar to that in a NaClO 4 solution (see Fig. 3b). If BPA is degraded by SO 4 %− , the quenching effect with MeOH should be greater than that with t-BuOH. However, the BPA degradation kinetics were not completely quenched. This result can be attributed to direct electron transfer to a  surface-trapped hole by BPA leading to its degradation. The hypothesis is consistent with results showing a slight increase in the cathodic current in the presence of BPA compared to the absence of BPA under both O 2 and N 2 (see Fig. S6).
The generation of % OH was further confirmed by using coumarin as a selective probe reagent for % OH trapping. The hydroxylated products (7-hydroxycoumarin) generated by the reaction of coumarin with % OH ( % OH + coumarin → 7-hydroxycoumarin) was quantified by monitoring the fluorescence emission [38]. Fig. 4b shows the pH-dependent electrochemical production of 7-hydroxycoumarin. The electrochemical production of 7-hydroxycoumarin was increased with decreasing pH, which demonstrates that the electrolytic degradation of BPA can be mainly ascribed to the facile production of % OH as a primary oxidant produced on the surface of the anode at lower pH. This observation agrees with the data presented in Fig. 3a, which shows higher electrochemical activities for H 2 O 2 production and BPA degradation under acidic and neutral pH compared to alkaline pH conditions. Even though a graphite rod normally has a low activity for catalyzing the decomposition of H 2 O 2 , the decomposition of H 2 O 2 via this pathway (see Eq. (3)) cannot be ruled out. To test this possibility, the carbon cathode was replaced by a stainless steel cathode that has a lower activity for the electrochemical production of H 2 O 2 than a graphite rod [39]. However, in spite of a significant reduction in the rate of H 2 O 2 formation on the stainless steel cathode, the concomitant degradation of BPA was only slightly reduced compared to case of graphite rod cathode (Fig. S7). This result suggests that BPA degradation is initiated by surface-bound > TiOH % radicals produced at the anode surface not by free % OH radicals produced via H 2 O 2 reduction at the cathode.
The pH change observed during electrolysis in the presence of BPA was completely different from that observed in the absence of BPA (Fig.  S8). After applying an external bias potential (−0.5 V), the pH immediately increased from 5.8 to 7.5 in the absence of BPA due to the consumption of protons required for H 2 O 2 production and then very slightly decreased. On the other hand, the pH was continuously reduced during the oxidation of BPA. This result is due to the formation of organic acids such as lactic, oxalic, fumaric, and glutaric acid [40]. Quinones and catechol were formed as reaction intermediates (Fig.  S9a), which were, in turn, oxidized into organic acids. Despite the almost complete removal of BPA in 2 h, the TOC removal was only 66% after 4 h, although complete mineralized was achieve in 6 h (Fig. S9b).
Phenol (PhOH) and 4-chlorophenol (4-CP) were electrolytically oxidized under identical conditions (Fig. 5) and were found to have similar rates degradation and H 2 O 2 production compared to BPA. Even though reactive chlorine species were produced in the oxidation of 4-CP (Fig. S10), their effects on H 2 O 2 production was minor compared to electrolysis in the NaCl electrolyte (see Fig. S5). Thus, we conclude that the RCS concentration produced during the electrolysis of 4-CP was low (< 1 mM) [35].

Conclusions
Herein, we clearly demonstrate that positive effect of organic electron donors on the electrochemical production of H 2 O 2 . The organic substrates are preferentially adsorbed on the anode surface preventing the anodic oxidation of H 2 O 2 formed on the cathode. Furthermore, the organic electron donors actively scavenge surface-bound hydroxyl radical ( % OH), which also reacts competitively with H 2 O 2 . As a result, the oxidative decomposition of H 2 O 2 is reduced resulting in the net accumulation of H 2 O 2 during the electrolysis of organic pollutants.