Light Illuminated α−Fe2O3/Pt Nanoparticles as Water Activation Agent for Photoelectrochemical Water Splitting

The photoelectrochemical (PEC) water splitting is hampered by strong bonds of H2O molecules and low ionic conductivity of pure water. The photocatalysts dispersed in pure water can serve as a water activation agent, which provides an alternative pathway to overcome such limitations. Here we report that the light illuminated α−Fe2O3/Pt nanoparticles may produce a reservoir of reactive intermediates including H2O2, ·OH, OH− and H+ capable of promoting the pure water reduction/oxidation half−reactions at cathode and highly photocatalytic−active TiO2/In2S3/AgInS2 photoanode, respectively. Remarkable photocurrent enhancement has been obtained with α−Fe2O3/Pt as water activation agent. The use of α−Fe2O3/Pt to promote the reactivity of pure water represents a new paradigm for reproducible hydrogen fuel provision by PEC water splitting, allowing efficient splitting of pure water without adding of corrosive chemicals or sacrificial agent.

T he storage of solar energy in chemical bond of H 2 through water splitting under sun2light presents the most promising strategies to develop a solar2based energetic model in view of the abundant and renewable nature of solar and water resources [1][2][3][4][5] . Since the pioneering studies of Fushijima and Honda in the early 1970s 6 , which demonstrated oxidation of water on n2type TiO 2 single2crystal electrode by band2gap excitation, photoelectrochemical (PEC) water splitting is regarded as the simplest solar to hydrogen (STH) conversion scheme [7][8][9][10] . In a typical PEC water splitting reaction, oxygen is produced on light2excited semiconductor electrode via water oxidation half2reaction 2H 2 O 1 4h 1 (hole) R O 2 1 4H 1 , and hydrogen is generated on Pt counter electrode by water reduction half2reaction 2H 2 O 1 2e 2 R 2OH 2 1 H 2 . Thus, sun light plus water gives us clean hydrogen plus oxygen. It sounds good, but it is not all that easy because the water splitting reaction is an uphill reaction in which the Gibbs free energy increases by 237 kJ mol 2111 . Particularly, splitting of pure water is extremely difficult due to its prohibitively low ionic conductivity. A great deal of effort has been put to overcome the difficulty of splitting of the pure water. Electron donors (sacrificial reagents), including organic compounds (hydrocarbons) 12,13 , weak acids 14,15 , inorganic ions [16][17][18][19] , etc., are widely used for photocatalytic hydrogen production as they enhances the photocatalytic electron/hole separation by scavenging the photo2generated valence band (VB) holes 20 , resulting in higher quantum efficiency. However, since the electron donors are consumed in this photo2catalytic reaction, the product is only hydrogen and the reaction is not an overall splitting of water. Another way to increase the reactivity of the water splitting is to use the alkaline solutions, which enhance the forward photo2catalytic reaction and suppress backward reaction (recombination of hydrogen and oxygen into water) by scavenging of the photo2generated holes 20,21 . It is demonstrated that both of the hydrogen and oxygen production can be increased. The limits of this strategy are low STH conversion efficiency and performance degradation due to the corrosive environment for the electrodes. Thus, it is highly desirable that the PEC water splitting technique which is aimed at providing a clean and renewable fuel can efficiently split water into hydrogen and oxygen without adding of corrosive chemicals or sacrificial agent.
It is generally accepted that the photo2illuminated photocatalysts provide extremely reactive intermediates in water, such as superoxide anion (O 2 tants [22][23][24][25][26] . In effect, the reactive intermediates generated from photocatalysts are expected to enhance the PEC water splitting efficiency by promoting the water oxidation/reduction half2reaction at photoanode and cathode, respectively. However, the important and unique role of the light2illuminated photocatalysts in water and their consequent ability to serve as water activation agent by generating reactive intermediates for PEC water splitting has not been considered previously. Here, a new strategy of water activation by generation of various reactive intermediates using the photo2illuminated a2Fe 2 O 3 /Pt nanoparticles (NPs) has been demonstrated for PEC water splitting. a2Fe 2 O 3 was chosen in our experiment because it has a band2gap of 2.0 to 2.2 eV corresponding to the absorption of 564 to 620 nm light, allowing it a promising photocatalyst for harvesting solar energy for hydrogen production 27,28 or degradation of organic pollutants and toxics 29,30 . Furthermore, because the conduction band (CB) bottom (E CB ) and VB top (E VB ) of a2Fe 2 O 3 is more positive than the hydrogen and oxygen evolution potential, respectively, only reactive intermediates can be generated into water, rather than evolution of hydrogen through photocatalytic water splitting. The application of a2Fe 2 O 3 as a potential photocatalysts is mainly limited by its short lifetime of photogenerated charge carriers (,10 ps) and short hole diffusion length (,2 to 4 nm) 27,31 . To address these issues, Pt NPs were decorated on a2Fe 2 O 3 by a polyol reduction method, which can serve as cocatalysts to enhance the photocatalytic activity and increase the lifetime of the photogenerated charge carriers 32,33 . Combined with the high photocatalytic active TiO 2 /In 2 S 3 /AgInS 2 photoanode, remarkable photocurrent of ,0.788 mA cm 22 at 1.5 V vs. Ag/AgCl has been obtained with a2Fe 2 O 3 /Pt as water activation agent, more than ten times as large as the values without a2Fe 2 O 3 /Pt (0.075 mA cm 22 at 1.5 V vs. Ag/AgCl). Figure 1a shows a schematic diagram of the reaction vessels, where TiO 2 /In 2 S 3 /AgInS 2 photoanode pressed on the inner wall of the quartz vessel with conducting side facing the reaction solution serves as the working electrode, Pt2foil as the counter electrode, and Ag/ AgCl in saturated KCl as the reference electrode. The photoanode is the central to the PEC cell, whose material and structure both play critical roles in the device performance. An ideal photoanode requires fast water oxidation kinetics at the semiconductor/water interface, fast electron transport and suitable band gap large enough (.1.6 eV) to split water and small enough (,2.2 eV) to absorb a wide range of the solar spectrum. TiO 2 /In 2 S 3 /AgInS 2 core2shell structure is one of the high2performance photoanodes that satisfy these requirements simultaneously, which is crucial to evaluate the ability of our strategy to promote the water splitting efficiency. Figure 1(b1) and (b2) shows the top and cross-sectional view of the TiO 2 /In 2 S 3 /AgInS 2 photoanode, respectively. After growth of In 2 S 3 and AgInS 2 , the products inherit the morphology of the TiO 2 NW arrays, showing an average diameter of ,116 nm and a length of ,3.36 mm. A typical transmission electron microscopy (TEM) image of a single TiO 2 /In 2 S 3 /AgInS 2 NW demonstrates that the surface of the TiO 2 NWs appears to be very coarse, and many NPs are coated over the surface of TiO 2 NWs, as illustrated in Figure 1(b3). Further insight into the structural information was obtained by high resolution TEM (HRTEM) taken from the TiO 2 /In 2 S 3 /AgInS 2 interface, Figure 1(b4) and (b5). The In 2 S 3 buffer layer and TiO 2 NW form a core2shell structure and the AgInS 2 NPs decorated on surface of In 2 S 3 . The resolved spacings between the two parallel neighboring fringes are 0.325 and 0.246 nm, corresponding to the [311] plane of cubic In 2 S 3 and [122] plane of orthorhombic AgInS 2 . The crystal phase of the three materials was further investigated by XRD spectrum; the diffraction peaks in Figure S1 are well indexed with the rutile TiO 2 , cubic In 2 S 3 and orthorhombic AgInS 2 , respectively.

Results
Pt NPs were deposited on a2Fe 2 O 3 NPs by the polyol reduction method (see Experimental section and also Ref. 32). As illustrated in Figure 1(c1), the average size of the hybrid a2Fe 2 O 3 /Pt NPs was ,38 nm. Pt NPs with average size of ,3 nm were decorated on a2Fe 2 O 3 NPs, as confirmed by the TEM image in Figure 1 information, Figure S2). Figure 2 compares the UV-vis absorption spectra of a2Fe 2 O 3 /Pt and a2Fe 2 O 3 NPs dispersed in DI water with that of DI water. The DI water exhibits an intrinsic absorption of water with weak absorption intensity in the visible range; while a broad absorption with the absorption edge at about 600 nm is found for both a2Fe 2 O 3 /Pt and a2Fe 2 O 3 NPs dispersed in DI water, which originates from the intrinsic absorption of a2Fe 2 O 3 .
To test the ability of the reactive intermediates produced from light illuminated a2Fe 2 O 3 /Pt NPs to enhance the performance of PEC water splitting, we measured the current density versus applied voltage (J-V) curves of the TiO 2 /In 2 S 3 /AgInS 2 working electrode in different electrolytes of 15 MV DI water, 15 MV DI water with a2Fe 2 O 3 /NPs and 15 MV DI water with a2Fe 2 O 3 /Pt NPs in the dark and under illumination. The photocurrent in DI water is only 0.075 mA cm 22 at 1.5 V vs. Ag/AgCl, which increases sharply to 0.345 mA cm 22 with a2Fe 2 O 3 NPs added to DI water. When tested a2Fe 2 O 3 /Pt as a water activation agent, as expected, the J improvement is more pronounced for a2Fe 2 O 3 /Pt than for a2Fe 2 O 3 , to ,0.788 mA cm 22 at 1.5 V vs. Ag/AgCl, as shown in Figure 3a. What is more, the J value shows a a2Fe 2 O 3 /Pt concentration2dependent behavior, which increases with increasing of the concentration of a2Fe 2 O 3 /Pt, reaching a maximum value at 0.1 mg/mL, followed by decreasing with further increasing concentration, as shown in Figure 3b. Excessive a2Fe 2 O 3 /Pt in water decreases the light penetration depth, which reduces the rate of photo2catalyzed reaction of water and consequent generation of reactive intermediates by a2Fe 2 O 3 /Pt, resulting in the decrease of J. It is noteworthy that a minute amount of a2Fe 2 O 3 /Pt NPs were remained in the photoelectrode ( Figure S3) after PEC measurement, indicating that the a2Fe 2 O 3 /Pt NPs adsorbed on photoelectrode during the PEC measurement can be neglected.
To compare the effect of the addition of a2Fe 2 O 3 /Pt on the pure water splitting reaction rate to that of the addition of generally used chemicals, two control experiments were conducted by replacing the a2Fe 2 O 3 /Pt suspension with 1 M NaOH aqueous solution and 0.5 M K 2 SO 4 aqueous solution containing H 2 SO 4 (adjust the pH to 1.7). As shown in Figure S4, the J value is 1.07 mA cm 22 at 0.9 V vs. Ag/AgCl for NaOH electrolyte and 0.981 mA cm 22 at 1.5 V vs. Ag/  www.nature.com/scientificreports AgCl for K 2 SO 4 electrolyte, which are slightly larger than that of a2Fe 2 O 3 /Pt suspension electrolyte. This result indicates that our strategy of using photocatalysts to promote the reactivity of pure water provide a promising approach for high efficiency PEC water splitting, and the pure water splitting performance could be greatly improved by using more promising semiconducting materials with novel nanostructures in the follow-up works. Figure 3c shows a representative J-t curve of TiO 2 /In 2 S 3 /AgInS 2 in DI water containing 0.1 mg/mL a2Fe 2 O 3 /Pt. The measurement was conducted under the illumination of simulated solar light (AM 1.5 G, 100 mW cm 22 ) at 1.5 V vs. Ag/AgCl. Prior to the measurement, the newly synthesized TiO 2 /In 2 S 3 /AgInS 2 were illuminated under simulated solar light for 200 s, allowing stabilization of the performance of TiO 2 /In 2 S 3 /AgInS 2 electrode. From the result one can see that the instantaneous photocurrent density with turning the light on reaches the constant photocurrent density and remains constant until the light is turned off, where the current immediately decays to the dark value of the current. This reproducible rapid rise and decay behavior implies the fast hole scavenging from the surface of the In 2 S 3 /AgInS 2 heterostructure to the solution and rapid transferring of photoelectrons from In 2 S 3 /AgInS 2 to current collector via the interior TiO 2 NWs 34,35 . Additionally, the photocurrent was steady for 1700 s, indicating stable photo2stability of both Fe 2 O 3 / Pt and TiO 2 /In 2 S 3 /AgInS 2 photoanode.
To reveal the differences in the interfacial charge-transfer characteristics of both half reactions in the PEC cell with and without a2Fe 2 O 3 /Pt, electrochemical impedance spectroscopy (EIS) measurements were carried out in a two electrode configuration PEC cell 36 . The Nyquist plots of the obtained EIS data measured at open-circuit conditions under simulated solar-light illumination are shown in Figure 4. According to recent analysis on the EIS spectra of the PEC cell for water splitting 36,37 , the first semicircle in the high2frequency region (.10 3 Hz) represents the charge transfer (R ct ) at the TiO 2 /In 2 S 3 /AgInS 2 /electrolyte interface; and the other arc in a frequency range of 100 mHz-10 3 Hz corresponds to the reduction reaction (R rr ) at the Pt counter electrode. The fitting curves fitted by EIS Spectrum Analyser software using an equivalent circuit shown in the inset match well with the measured EIS data. The fitted R ct and R rr values for the cells with DI water as electrolyte is as large as 112 kV and 527 kV, respectively; while R ct and R rr for the cells with a2Fe 2 O 3 /Pt as water activation agent decreased to 0.985 kV and 2.5 kV, respectively. The EIS analysis revealed that the presence of a2Fe 2 O 3 /Pt in DI water can greatly promote the activity of water reduction/oxidation half2reactions at counter electrode and TiO 2 / In 2 S 3 /AgInS 2 photoanode, respectively.

Discussion
On the basis of the above experiments, it is reasonable to ascribe the significant improvement of the PEC water splitting efficiency to the generation of the reactive intermediates from the light illuminated a2Fe 2 O 3 /Pt NPs. Some potential reactions that could be initiated by photo electron2hole pairs generated in a2Fe 2 O 3 /Pt and the consequential process in relation to the water splitting can be depicted as in Figure 5. Under light illumination, the electrons in the VB of a2Fe 2 O 3 are promoted to the CB of a2Fe 2 O 3 by photo excitation (c), and electron (e 2 ) 2 hole (h 1 ) pairs are generated. The E CB of a2Fe 2 O 3 (0.38 V vs. NHE) is more negative than the reduction potential to form OH 2 (0. 40    conduction band electron through H 2 O 2 1 e 2 R ?OH 1 OH 239,40 ; and the resultant ?OH may be further reduced to OH 2 through ?OH 1 e 2 R OH 238 . The resulting H 2 O 2 , OH 2 and H 1 in the above reactions are highly active for the water reduction/oxidation half2reaction at the cathode and photoanode in a PEC cell. Thus, the a2Fe 2 O 3 /Pt dispersed in pure water can serve as a water activation agent under light illumination, which could produce a reservoir of reactive intermediates (H 2 O 2 , ?OH, OH 2 , H 1 ) capable of promoting the water splitting reaction.
According to the above mentioned reaction mechanism, ?OH is a key intermediate relating to generation of substances (H 2 O 2 , OH 2 and H 1 ) that can directly promote the water splitting reaction at photoanode and counter electrode. It is widely accepted that the fluorescent probe method using terephthalic acid (TA) as the ?OH capture is a highly sensitive technique, in which the TA reacts with ?OH and generates luminescent 2-hydroxyterephthalic acid (TAOH) with a characteristic peak at ,426 nm [41][42][43] . Figure 6 shows the fluorescence spectral changes observed during illumination of a2Fe 2 O 3 / Pt suspension containing 0.5 mM terephthalic acid at various irradiation periods. Gradual increase in the fluorescence intensity at ,428 nm with increasing illumination time implies that fluorescent TAOH was formed via the specific reaction between ?OH and TA during illumination of a2Fe 2 O 3 /Pt suspension, which is a direct evidence of the presence of ?OH. In addition, H 2 O 2 is another important intermediate, the existence of which can be verified by hydrogen peroxide indicator strip, as illustrated in Supporting Information, Figure S5. The observation of ?OH and H 2 O 2 in the water splitting reaction provide a weighty evidence to the reaction mechanism.
Because the E CB of a2Fe 2 O 3 (0.38 V vs. NHE) is less negative than the hydrogen evolution potential (0.00 V vs. NHE), it is not able to reduce the H 1 to give H 2 directly by a2Fe 2 O 3 . Therefore, in the PEC cell with a2Fe 2 O 3 /Pt activated pure water as reaction solution, the reduction of H 1 takes place only on Pt cathode, which requires efficient oxidation of the water or reactive intermediates including H 2 O 2 , OH 2 by the photoanode. Thus the water oxidation ability of the photoanode plays a key role for water splitting in this PEC cell. In this work, we investigated three TiO 2 NW based electrodes including TiO 2 NW, TiO 2 NW/CdS and TiO 2 /In 2 S 3 /AgInS 2 . Identical behavior has been observed for the three photoanodes, as illustrated in Figure 3, Figure S6 and S7. Due to the excellent light harvesting (see Supporting Information, Figure S8) and photocatalytic activity property 44 , the TiO 2 /In 2 S 3 /AgInS 2 working electrode shows the highest photocurrent value. To double check the reasonability of this conclusion and the reaction mechanism described in Figure 5, we compare the J-V curves of a PEC cell with a Pt2foil as working electrode, another Pt2foil as cathodic electrode and a2Fe 2 O 3 /Pt NPs dispersed in DI water as electrolyte in dark and under light illumination. As shown in Figure 7, identical behavior has been observed for J-V curves with and without light illumination. Furthermore, no instantaneous photocurrent was observed on the chopped2light current density versus time (J-t) curve of the cell. These results indicate that oxidation of water at the working electrode ( Figure 5) plays a key role for water splitting in our PEC cell.
In conclusion, an alternative pathway to activate the pure water for PEC water splitting by introducing photocatalysts into water has been developed. The light illuminated a2Fe 2 O 3 /Pt NPs may produce a reservoir of reactive intermediates including H 2 O 2 , ?OH, OH 2 and H 1 capable of promoting the water reduction/oxidation half2reactions at cathode and TiO 2 /In 2 S 3 /AgInS 2 photoanode, respectively. Remarkable photocurrents of ,0.788 mA cm 22 at 1.5 V vs. Ag/AgCl has been obtained with a2Fe 2 O 3 /Pt as water activation agent, more than ten time as large as the values without a2Fe 2 O 3 / Pt (0.075 mA cm 22 at 1.5 V vs. Ag/AgCl). The present results provide a fertile base for further investigation. The strategy of using photocatalysts to generate reactive intermediates in pure water for PEC water splitting demonstrated by a2Fe 2 O 3 /Pt NPs can be leveraged to other, more promising semiconducting materials with novel nanostructures to greatly improve their efficiencies and application areas. The approach could also be extended to other energy and artificial photosynthesis applications.

Methods
Preparation of a2Fe 2 O 3 /Pt NPs. a2Fe 2 O 3 NPs with diameter ,30 nm were purchased from Aladdin Industrial Inc. (Shanghai, China). Deposition of Pt onto a2Fe 2 O 3 NPs followed procedures outlined previously 32 . Typically, a2Fe 2 O 3 powder (0.5 g) was dispersed in a mixed solution (40 mL) containing H 2 PtCl 6 aqueous solution (1 wt.%) and ethanol under ultrasonication for 30 min. Then the slurry was dried at 60uC. Ethylene glycol (40 mL) was added to the dry powder followed by stirring and ultrasonication to form a homogenous suspension. The suspension was kept at 100uC in dark for 6 h. At last the a2Fe 2 O 3 /Pt powder was collected by centrifugation, washed with distilled water for several times, dried at 60uC and sintered at 400uC for 20 min.
Preparation of TiO 2 /In 2 S 3 /AgInS 2 core2shell electrodes. At first, a TiO 2 polymeric sol was prepared by the sol gel process according to the previous reports 45 . Then the TiO 2 sol was spin2coated on the fluorine2doped SnO 2 (FTO) substrates followed by annealing at 450uC for 2 h. The TiO 2 NW arrays were grown directly on seeded FTO substrates by using the hydrothermal method reported previously 46 . In a typical synthesis process, titanium (IV) butoxide (0.5 g) was added into an aqueous HCl solution (25 mL of deionized water and 25 mL of concentrated HCl (38%)) under  magnetic stirring. The solution was stirred for another 10 min and then poured into a Teflon2lined stainless steel auto2clave (100 mL capacity). Six pieces of the seeded2FTO (0.8 cm 3 2 cm, with seeded area of 0.8 cm 22 ) were placed at an angle against the wall of the Teflon2liner with the conducting side facing down. The autoclave was sealed, heated to 170uC and held at the temperature for 6 h. After cooling down to room temperature, the obtained products were washed successively by DI water and ethanol and finally annealed at 500uC for 2 h. In 2 S 3 /AgInS 2 were deposited on TiO 2 NWs by sequential chemical bath deposition (S-CBD) method according to a previous report but with a modified recipe 44 .
Typically, the TiO 2 NWs on FTO substrate were successively dipped into InCl 3 ?4HO 2 ethanol solution (3 mM) for 4 min, ethanol for 1 min, Na 2 S?9H 2 O water-methanol solution (3 mM) (151 volume ratio) for 4 min and water-methanol (151 volume ratio) mixture for 1 min at 25uC. The desired deposition of In 2 S 3 was achieved after 12 cycles with the white TiO 2 NW film gradually became pale yellow. Subsequently, the TiO 2 /In 2 S 3 film was immersed in AgNO 3 ethanol solution (2 mM) at 25uC for 2 min. The resultant brown TiO 2 /In 2 S 3 /AgInS 2 films were washed with ethanol and sintered at 400uC for 30 min in N 2 atmosphere.
Photoelectrochemical measurements. All the PEC measurements were performed in a quartz reaction vessel containing DI water (20 mL, 15.0 MV, Elix Advantage 10, Merck Millipore) and a2Fe 2 O 3 /Pt NPs. The PEC measurements were performed in a three electrode configuration with TiO 2 /In 2 S 3 /AgInS 2 as the working electrode, Pt2foil (surface area of 1.0 cm 2 ) as the counter electrode, and Ag/AgCl in saturated KCl as the reference electrode. To prevent suspended a2Fe 2 O 3 /Pt NPs from screening the photo2absorption of the photoelectrode, the TiO 2 /In 2 S 3 /AgInS 2 electrodes were pressed against the inner wall of the quartz vessel with conducting side facing the reaction solution. The TiO 2 /In 2 S 3 /AgInS 2 electrode was connected to the measuring instrument by pressing a Pt foil on the FTO layer of the TiO 2 /In 2 S 3 / AgInS 2 electrode. The PEC performances were measured using an Electrochemical Workstation (Bio-Logic SAS, VSP-300). Illumination was from a solar simulator with a Xe arc lamp as light source and the spectrum was matched to the AM 1.5 G spectrum. Before the measurement, the solar intensity (100 mW cm 22 ) was calibrated with a reference silicon solar cell. The illuminated area of the working electrode was 0.8 cm 2 .
Hydroxyl radical formation was studied by means of terephthalic acid (TA) fluorescence probe method as follows. An aqueous solution containing 0.5 mM TA was prepared, and then a2Fe 2 O 3 /Pt NPs (0.1 mg/mL) was suspended in this solution in a quartz reaction vessel. Prior to irradiation, the suspension was magnetically stirred for 30 min in a dark box to establish an adsorption2desorption equilibrium. The excitation light source was the same as that in PEC water splitting measurements. To sediment a2Fe 2 O 3 /Pt NPs from the suspensions and get rid of light scattering for the subsequent measurement of the fluorescence spectra, the samples for different irradiation periods were centrifuged at 10000 rpm for 2 min. Fluorescence spectra of 2-hydroxyterephthalic acid (TAOH) were measured on a fluorescence spectrophotometer (Omni-pR-PL, Beijing Zolix Instruments CO., LTD) with an excitation at 325 nm light.
Characterizations. The morphology and microstructure of the TiO 2 /In 2 S 3 /AgInS 2 electrode and a2Fe 2 O 3 /Pt NPs were characterized by a field emission scanning electron microscopy (FE2SEM, Hitachi S24800) and transmission electron microscopy (TEM, FEI Tecnai F30). Elemental analysis was performed on an energy2dispersive x2ray (EDX) spectroscopy attached to the FE2SEM. X2Ray diffraction spectra (XRD) was collected on a Bruker D8 Advance X2ray diffractometer using a Cu Ka source (l 5 0.154056 nm). The optical absorbance spectra were acquired using a UV-visible spectrophotometer (TU21901). Electrochemical impedance spectroscopy (EIS) was measured with the Electrochemical Workstation in a two electrode configuration within a frequency range from 0.1 Hz to 800 kHz at open-circuit voltage with a potential pulse of 100 mV in amplitude under simulated solar-light illumination (AM 1.5 G, 100 mW cm 22 ). Prior to the recording of EIS data, the PEC cell was illuminated for 10 min at an applied bias of 1.5 V to establish equilibrium of the system. The EIS data were fitted by EIS Spectrum Analyser software. Hydrogen peroxide indicator strips (Quantofix Peroxide 25, MACHEREY-NAGEL, Germany) were used to test the existence of H 2 O 2 .