Photoelectrochemical Oxidation of Ethanol under Visible Light Irradiation on TaON-Based Catalysts

The ethanol photoelectro-oxidation reaction was investigated under visible irradiation on TaON-based catalysts modiﬁed with Pd and IrO 2 . Prepared catalysts were characterized by energy dispersive X-ray analysis, X-ray diffraction, X-ray photoelectron spectroscopy and transmission electron microscopy. Electrochemical investigations were conducted in 0.2 mol L − 1 NaOH solution, by voltammetric multi-cycling experiments for dark and illuminated electrodes. Product and intermediate distribution analyses were carried out by differential electrochemical mass spectrometry and high-resolution liquid chromatography aiming at establishing the photoelectrocatalytic reaction pathways. The presence of small amounts of palladium (ca 2 atom% Pd in TaON) is essential to obtain enhanced ethanol photoelectro-oxidation under visible light in alkaline medium. Photocatalytic reaction promotes formation of molecular hydrogen, carbon dioxide, acetate ions, and acetaldehyde (the last one subsequently converted into crotonaldehyde and acetaldol), while the main product formed by the ethanol photoelectro-oxidation is only acetate ion. the terms of the work 10.1149/2.0131803jes]

The conversion of solar into electric energy has gained great attention in the last decades due to the environmental degradation and health issues caused by the use in a large scale of fossil fuels, besides the concern regarding their shortage in the future. These aspects leaded to an intense search for renewable energy sources, focused in finding new alternatives for electricity generation. In this context, Photoelectrochemical (PEC) cells are very attractive devices due to their capacity to convert light into electric power or chemical fuel (renewable hydrogen) by photo-electrochemical decomposition of a target substance using solar light. Several works have investigated PEC cells for the production of electricity and hydrogen from renewable resources, mainly water 1-4 and biomass [5][6][7][8][9][10][11][12][13] including ethanol, glycerol and ethylene glycol.
Nowadays, the main photocatalysts investigated for solar energy conversion have been based on TiO 2 and ZnO due to their high oxidation power when illuminated. [14][15][16][17] Nonetheless, these semiconductors have large band gaps, 3.0-3.2 eV and 3.35 eV, respectively, 18,19 and do not work under visible light, therefore just utilizing a small fraction, around 4%, of the solar spectrum. Lianos and co-workers 5,20,21 have investigated the photoelectrochemical decomposition of ethanol for hydrogen and electricity production using nanocrystalline titania deposited on FTO in a PhotoFuelCell at a high pH. They observed an increase in the faradaic current and hydrogen production when ethanol is employed, indicating occurrence a more efficient process compared to photocatalytic water splitting. Other researches [22][23][24][25][26][27] have investigated materials no based in titanium oxide as photoelectrodes (CdS, CdSe, WO 3 , etc), or differentiated systems of the conventional cells that employing high temperature as the solid oxide photoelectrochemical cell 28 in order to obtain higher performances.
Thereby, efforts have been focused in photocatalysts that respond to the visible light for different applications, such as water molecule splitting, organic compounds oxidation, inorganic ions reduction, disinfection and production of electricity. Tantalum oxynitride has been studied in the last years as a potential photocatalyst for water splitting into O 2 and H 2 under solar visible radiation. [29][30][31][32][33][34][35] TaON has been considered a photocatalyst with good absorption in the visible light region (absorption band-edge 500 nm), 30 good stability under photochemical conditions (acidic and basic solutions) 35 and a relatively high activity for O 2 evolution and moderate activity for H 2 evolution, but this can be further improved by combinations with other species, like ZrO 2 . 36-38 . Studies using this photocatalyst 35,39,40 have shown the evolution of N 2 when the oxynitride is photo irradiated, and this was associated to the oxidation of nitrogen anions (N 3− ) by photogenerated holes (2N 3− + 6h + → N 2 ). However, some works [39][40][41][42][43][44][45][46] have demonstrated that the nitrogen evolution can be reduced by loading the oxynitride with a co-catalyst able to oxidize water efficiently, such as CoO x 39 and IrO 2 . [40][41][42] As mentioned above, photoelectrochemical cells have been used to oxidize a fuel to produce electricity or storable chemical energy, mainly hydrogen, and ethanol has been mentioned as one of the possible fuel. 47 Such a photo-electrochemical anode, where the ethanol oxidation reaction (EOR) takes place, may be combined with a hydrogen generating cathode, so that the electrolyser cell voltage would be strongly reduced as compared to that with a regular water oxidation cathode system. On the other hand, if this anode is combined with an oxygen reduction cathode it may allow production of electricity by using a renewable fuel (in a so-called direct ethanol fuel cell). Of course, to be advantageous, it should be expected that the photocatalytic-assisted ethanol oxidation process is more efficient, as compared to that on regular anodes.
The EOR proceeds by a large number of steps, producing intermediates that can be more easily photo electrochemically oxidized than ethanol, eventually culminating with the complete fuel mineralization (eg. ethanol → acetaldehyde → acetic acid → CO 2 + H 2 O). In this way, a big challenge in the investigation of such reaction is the complexity of its mechanism, which may involve a large variety and reaction intermediates and products. In alkaline media and in non-photocatalytic anode materials such as Pd and PtSn alloys, the most abundant product is acetate ions, while in photocatalytic materials particularly involving TaON, nothing is known about the reaction products.
In this study, we investigated the ethanol photoelectro-oxidation reaction on TaON alone or with Pd and IrO 2 under visible light irradiation in 0.2 mol L −1 NaOH solution. Despite that this class of materials has shown promising results for the oxygen and hydrogen evolution from water, no works are found regarding their use for electricity generation and/or hydrogen production from ethanol. Products and intermediates distribution analyses were carried out by differential electrochemical mass spectrometry (DEMS) and high-resolution liquid chromatography (HPLC), aiming at establishing the photoelectrocatalytic ethanol decomposition pathways.

Experimental
Photocatalysts preparation.-TaON was prepared employing a previous developed method, 32,34,35,38 where tantalum (V) oxide, Ta   (Sigma-Aldrich Resistance 7 /cm 2 ), by electrophoretic deposition, adapting a previously published method. 41,42 Two FTO electrodes (1.5 × 5 cm 2 ) were dipped parallel to each other at a distance of 8 mm in a solution containing acetone (70 mL) + iodine (10 mg) + TaON (40 mg) dispersed by sonication (5 min). The coated area was limited by a Teflon tape (1.5 × 1.5 cm 2 ), and 5 V bias was applied for 30 min; to sustain the species under suspension, the solution was kept under magnetic stirring at slow speed (100 rpm). This was the optimized conditions for preparing stable working electrodes with minimal contact resistances, and in which the TaON film remains firmly adhered on the FTO substrate during all experiments. For obtaining the Pd and/or IrO 2 TaON-based photocatalysts, solutions containing 10 mmol L −1 of PdCl 2 in methanol (Alfa Aesar, Palladium (II) chloride, 99.9% (metal basis), Pd 59.91% Lot # F07Q26) and/or 10 mmol L −1 of IrO 2 in water (Alfa Aesar, Iridium (IV) oxide, 99.9% (metals basis), Ir 84.5% Lot # S10C040) were prepared by sonicating for 15 min. Pd/TaON was made immerging TaON/FTO for 5 min in a solution containing only Pd and then dried at room temperature (25 • C). This procedure was repeated 5 times and then the electrode was heated under NH 3 flow at 673 K for 1 h, to improve the contact among TaON nanoparticles. IrO 2 /Pd/TaON/FTO was prepared dipping Pd/TaON/FTO into the IrO 2 solution for 30 min, followed by washing with water and drying at room temperature. The IrO 2 was deposited on TaON (IrO 2 /TaON) in the same way as for Pd/TaON. Deposition of a Pd film on the FTO occurred immerging the FTO into the 10 mmol L −1 PdCl 2 solution and drying at room temperature (repeated 5 times) and then the palladium was reduced under hydrogen atmosphere for 1 h at 400 • C.
Physical characterization.-The XRD patterns of the photocatalysts were obtained with a RIGAKU Ultima IV diffractometer with Cu Kα radiation (λ = 1.54056 Å), generated at 40 kV and 40 mA, in the range of 2θ from 10 • to 100 • , with a scan rate of 0.3 • min −1 . The refinement of TaON diffractograms (for determination average crystallite sizes and the phases present in the tantalum oxynitride) was carried out by the Rietveld method 48,49 using the TOPAS 5 software (Bruker AXS). Crystal structure, morphology and composition of the photocatalysts were examined using high-resolution transmission electron microscopy (HR-TEM), scanning electron microscopy (SEM) with a Link Analytical micro-analyzer (Isis System Series 200), using a Zeiss-Leica/LEO 440 model (LEO, UK) and highangle annular dark-field scanning TEM-energy-dispersive X-ray spectroscopy (HAADFSTEM-EDS) elemental mapping images, using a FEI TECNAI G 2 F20 HRTEM. The photocatalyst powders were mixed with isopropanol under sonication until samples homogenization, and then dropped on standard TEM copper grids and dried in air. The palladium particle diameters were measured from the TEM images using the ImageJ software. X-ray Photoelectron Spectroscopy -XPS characterizations were carried out with a Scientia Omicron Model ESCA for the analysis of the atomic ratios among the elements, and the oxidation state of them. Some characteristics of the XPS analyses are: pressure: 10 −10 mbar; emission sources: Al kα (1486.7 eV); angle: 50 • ; depth of analysis: 5nm. The samples required the use of a charge neutralizer in order to compensate the charge effect while obtaining spectra, it was used a Cn10 Omicron Charge neutralizer with a beam energy in 1.6 eV. The C 1s peak (284.8 eV) was employed to calibrate the binding energy determined by XPS.
Photoelectrochemical experiments.-Measurements were carried out at room temperature (25 • C) in a conventional glass cell with three electrodes, with one compartment under light and dark conditions. The cell was kept in a container without natural illumination, and thus the only irradiation source for the photocatalysts was from Short Arc Xenon lamp. The electrolyte employed was a 0.2 mol L −1 NaOH aqueous solution (25 mL prepared from Sigma-Aldrich, sodium hydroxide 99.99% Lot # MKBJ8595V) using high purity water from a Milli-Q Millipore system, 18.2 M cm.
The photoelectrochemical experiments were carried out in a one compartment borosilicate glass cell, in Ar purged solution (0.2 mol L −1 NaOH or 0.2 mol L −1 NaOH + Ethanol 5% vol.), with the photocatalysts employed as working electrodes, a platinum foil electrode as the counter electrode and Ag/AgCl, KCl (sat.) as reference electrode. The duration of the experiments were short enough for not having contamination with corrosion products coming from the attack of the alkaline electrolyte to the glass cell and FTO electrode. Electrochemical measurements were carried out using an Autolab PGSTAT 30. All cyclic voltammograms were recorded at 10 mV s −1 , while the electrode potential was cycled in the range of −0.93 to 0.3 V vs. Ag/AgCl, which correspond to the region where the ethanol electro-oxidation reaction occurs in alkaline media, as shown in many previous works. [50][51][52][53] The photocatalytic reactions were promoted in a closed gas circulation system under visible-light irradiation using a Short Arc Xenon lamp 75 W (OSRAM). Radiation intensity was measured with a Power Meter Model 843-R (Newport), which showed that the total radiation intensity on the photocatalysts was around 100 mW cm −2 .
The gases and volatile products and/or intermediates produced during the ethanol photodegradation reaction were instantaneously detected by on line Differential electrochemical mass spectrometry (on line DEMS) with a Pfeiffer Vacuum QMA 200 quadrupole mass spectrometer. [54][55][56] The ethanol photodegradation products monitored by on-line DEMS were m/z = 22 (CO 2 : doubly ionized CO 2 2+ ), m/z = 2 hydrogen (H 2 : H 2 + ), and m/z = 29 (acetaldehyde, COH + ). 55,57 The species present in the liquid phase after the ethanol photoelectronoxidation reaction (60 CVs light off (dark) and 60 CVs light on, at scan rate 10 mVs −1 ) were analyzed by high performance liquid chromatography (HPLC Shimadzu Model Prominence) using a Aminex HPX -87H Biorad Column 300 × 7.8 mm. Since the reaction in the working electrode was always anodic, the corresponding reaction in the counter electrode only involves molecular hydrogen releasing to the electrolyte, which does not interfere neither in the course of the working electrode process nor the transformation of the reaction products. This was confirmed by an additional experiment of the ethanol electro-oxidation in the presence of hydrogen (not shown here).

Results and Discussion
The X-ray diffraction patterns of the as prepared photocatalysts deposited on FTO and the diffractogram refinement made by the Rietveld method for the TaON substrate are shown in Figs. 1a and 1b, respectively. The hkl peak positions of TaON and Ta 2 O 5 (TaON PDF # 71-178 / Ta 2 O 5 PDF # 89-2843) were included for comparison. These results indicate that TaON is the major phase in the material, corresponding to 95.4% (mass %) of a beta monoclinic structure (β -TaON) 58,59 with lattice parameters of a = 4.964 Å, b = 5.033 Å and c = 5.181 Å, while unreacted Ta 2 O 5 (2θ = 22.8 • and 28.2 • ) is a minor phase, corresponding to 4.6% of the sample. Previous work have shown that the position of the conduction and the valence bands of TaON-based catalysts can be modified by reducing the oxygen and increasing the nitrogen contents in TaO 1-x N 1+x . 60 Here, the synthesized photocatalyst contains some Ta 2 O 5 but it is forming a segregated phases, as evidenced by the XRD peaks at 23 • and 28 • . In this way, it results clear that this non-stoichiometry does not affect significantly the overall photocatalyst properties.
The average crystallite sizes of TaON and Ta 2 O 5 determined by the Rietveld method were 50 nm and 88 nm, respectively. No peaks related to palladium or IrO 2 are apparent in the IrO 2 /Pd/TaON and Pd/TaON diffractograms, because they are present in small amounts probably forming amorphous phases mixed on the tantalum oxynitride substrate. X-ray diffraction patterns of the photocatalysts after the potentiodynamic experiments under light off/on are shown in the inset of Figure 1a. These results show the presence of some Pd features after these cycling, indicating some crystallization and growth of the Pd particles. As shown in previous works 61, 62 involving Pt/C catalysts, this may occur during the cycles through a dissolution/re-precipitation mechanism, where Pd dissolved from small particles is redeposited onto larger particles (Ostwald ripening) leading to the particle growth. Another possibility is the migration and collision of the small particles causing coalescence and growth. Thereby, the cycling and the irradiation processes can favor the growth of the palladium particles deposited on the substrate (TaON) producing the Pd peaks seen in the XRD patterns. No features of silicate coming from the glassware are evidenced by these XRD responses. Fig. 2 shows high resolution XPS spectra of (a) Ta 4p 2/3 and N 1s, (b) Pd 3d, represented by spin-orbit split components associated with the Pd 3d 3/2 and Pd 3d 5   with Pd and IrO 2 in the diffractograms of Pd/TaON and IrO 2 /Pd/TaON, as shown in Fig. 1a.
The morphology, structure and composition of the photocatalyst particles were examined using scanning electron microscopy (SEM), high-resolution transmission electron microscopy (HR-TEM) and high-angle annular dark-field scanning TEM-energy-dispersive X-ray spectroscopy (HAADFSTEM-EDS) elemental mapping images, as show in Figs. 3 and 4. As can be seen, the SEM image of TaON shows that the tantalum oxynitride has a porous structure, as previously reported. 39,41,42 Fig. 4 presents the elemental mapping images of IrO 2 /Pd/TaON and Pd/TaON with the corresponding transmission electron microscopy images of the mapping area. These images show a uniform distribution of all elements in the samples (Ta, Pd, Ir, O and N), with some segregation/agglomeration of the palladium particles, as highlighted in Figs. 4a-4b and 4g-4h. This fact can be associated to the thermal treatment employed for the reduction of this metal precursor, which may be responsible for the segregation/agglomeration of some of the palladium particles.
However, it is clearly seen in TEM images for Pd/TaON and IrO 2 /Pd/TaON that the palladium deposition on the tantalum oxynitride surface generated highly dispersed particles (diameters of 2-10 nm); a similar distribution was also observed for iridium oxide (Fig. 4d). The HR-TEM image of Pd/TaON shown in Fig. 4m indicates that β-TaON has a lattice space of 0.36 nm, which corresponds to the (011) planes of TaON, and a lattice fringe of 0.23 nm corresponding to the (111) planes of palladium.
CVs obtained just in the electrolyte (0.2 mol L −1 NaOH) for the electrodes in the dark are shown in Fig. 5a. Results evidence complete absence of CV redox features for the bare TaON electrode, and that addition of IrO 2 does not introduce any change to this behavior. In the case of Pd/TaON, the observed peaks in the CVs are related to the oxi-reduction features of adsorbed/desorbed hydrogen on Pd at −0.93 V, OH − adsorption in the anodic scan (∼ −0.2 V), adsorbed oxides conversion into higher valence oxides (PdO x ); and the cathodic peak at (∼ −0.2 V) is associated with the reduction of PdO x produced in the anodic scan. [65][66][67] Fig. 5b shows the cyclic voltammograms (CVs) for the ethanol electro-oxidation reaction in the dark catalyzed by TaON, Pd/TaON, IrO 2 /TaON, and IrO 2 /Pd/TaON, and the result for the Pd film deposited on FTO was included for comparison; these measurements were carried in 0.2 mol L −1 NaOH + ethanol 5% vol. As observed by comparing the CV profiles of the bare TaON and IrO 2 /TaON with those of the Pd/TaON and IrO 2 /Pd/TaON electrodes, the presence of Pd is essential to improve the CV electrochemical response and consistently, to promote the ethanol electro-oxidation; it is also seen that TaON and IrO 2 /TaON are not electrochemically active for this reaction. Because of the uncertainty related to the surface areas, discussions on the effect of TaON on the Pd catalyst cannot be made.
Palladium is reported in the literature 52,53,68 as the metal that presents the highest activity for the ethanol electro-oxidation in al- kaline medium, and such fact was associated to its higher oxophicility and smaller surface poisoning by intermediate reaction species, turning the ethanol oxidation more selective to acetate formation, as compared to platinum. According to the results in Fig. 5b, the Pd film deposited on FTO is as active as Pd/TaON for the investigated reaction, while, IrO 2 /Pd/TaON presents higher performance for the Ethanol Oxidation Reaction (EOR). Iridium and its oxides has been described as good co-catalyst, and when combined with Pt, PtRu and PtSn it increases the alcohols oxidation in acidic medium. [69][70][71][72] The addition of iridium and its oxides improves the catalytic activity and stability of these materials due to the preferential adsorption of hydroxyl groups on these species at lower potentials, favoring the oxidation of the ethanol intermediates adsorbed in the catalyst surface.  inset of Fig. 6a. All results with the light-on were obtained in the same electrode, after finishing the light-off experiments and washing it with pure water. As seen in Fig. 6, the faradaic currents initially increases with the number of cycles for all catalysts, in both conditions (light and dark), achieving the highest current values around the 30 th cycle and then an electrode deactivation occurs. The voltammograms present two anodic peaks, one in the anodic scan around −0.5 V and another in the cathodic scan close to −0.2 V. As observed previously, 65 the peak in the anodic scan is related to the oxidation of chemisorbed ethanol species, which is progressively blocked by the electrode coverage with Pd oxides, while the anodic peak in the reverse scan occurs after the reduction of these oxides into metallic palladium, allowing again ethanol adsorption and oxidation.
For the Pd/TaON and IrO 2 /Pd/TaON catalysts, the currents for ethanol oxidation are higher under light on conditions, for which the forward oxidation peaks become broader, particularly for Pd/TaON. This fact can be associated with the oxidation of species produced from ethanol photodegradation on the photocatalyst, which occurs at the Pd sites. Despite of the increase of activity on IrO 2 /Pd/TaON under irradiation, the differences between the currents for both condition (light/dark) are not as high as for Pd/TaON, indicating that palladium is the active metal for the ethanol/intermediates electro-oxidation. In this case, the IrO 2 layer present on Pd/TaON probably leads to a decrease the number of Pd active sites for the electro-oxidation of photo generated species. These results show that the tantalum oxynitride photocatalyst when combined with Pd produces a higher amount of reactive species by ethanol photodegradation, resulting in a consistent increase in the faradaic currents. Results also show that the Pd/TaON photocatalyst exhibited more stable electrochemical response, especially when under irradiation. In the case of the Pd/FTO electrode, the currents in the presence of light are smaller than those in the dark, in agreement with a larger deactivation of the Pd layer. As reported in the literature 73 the electrode deactivation in the course of the ethanol electro-oxidation reaction may be ascribed to the poisoning of the electrocatalyst surface by strongly adsorbed species, such as CO and CH x . Also, these results clearly confirm the photocatalytic aid provided by the TaON substrate for the ethanol oxidation, for both Pd/TaON and IrO 2 /Pd/TaON electrodes.
On-line DEMS were employed to characterize the main products formed during the ethanol photodegradation on TaON, Pd/TaON and IrO 2 /Pd/TaON electrodes at open circuit-potential, and the results are shown in Figs. 7a-7c. These results clearly denote that acetaldehyde (m/z = 29), H 2 (m/z = 2) and CO 2 (m/z = 22) are produced under irradiation on the investigated photocatalysts and that tantalum oxynitride, when combined with Pd and IrO 2 , generates higher amounts of CH 3 CHO and H 2 than pure TaON. On the other hand, the intensity of the ionic signals for CO 2 remains almost the same for all the three materials, indicating that the ethanol photodegradation favors the formation of species containing two atoms of carbon. In other words, the C2 pathways where the C-C bond remain intact (leading to acetaldehyde and H 2 ) upon ethanol oxidation is the predominant step as compared to the C1 pathway (leading to CO 2 ). 50 Figs. 8a-8b presents chromatograms obtained for the ethanol photodegradation (light on) in the TaON, Pd/TaON, and IrO 2 /Pd/TaON electrodes maintained under open-circuit, and the cell initially containing 0.2 mol L −1 NaOH + C 2 H 5 OH 5% vol. solution. The exposition time was 4 h 10 min, after which the HPLC analyses of the electrolyte were carried out. Besides the peaks for unreacted ethanol and acetate, the only other observed peaks appear at high retention times at 26.5 and 37.1 min, and associated with 3-hydroxybutanal (acetaldol) and crotonaldehyde, respectively. All these products were completely absent when corresponding experiments were made with the electrode at open-circuit but with the light off.
It is well known that the semiconductor (SC) electrode photoexcitation produces photogenerated holes in the valence band (h vB + ) and electrons in the conduction band (e CB − ) (Reaction 1). 47,74,75 Following the results at open-circuit potential in Figs. 7a-7c and 8a-8b, here it is observed that the photogenerated holes act intermediating the ethanol oxidation producing oxidized species, such as acetaldehyde, acetic acid and carbon dioxide (Reactions 2-4), while the photogenerated electrons present in the conduction band (e CB − ) reduce water generating H 2 molecule (Reaction 5 The acetaldol and crotonaldehyde observed in the chromatograms of Fig. 8b is surely generated from the acetaldehyde molecules produced in Reaction 2. Acetaldol is formed by combination of two acetaldehyde molecules (Reaction 6), while crotonaldehyde may be formed by acetaldehyde aldol condensation (Reaction 7). Results clearly denotes that TaON alone is only active for production of acetate and crotonaldehyde; in the presence of Pd, some formation of acetaldol is apparent, while the presence of IrO 2 is important to  In summary, in the photodegradation process, ethanol is converted into hydrogen, acetate ions, acetaldol and crotonaldehyde. Possible practical applications of these products would involve the use of hydrogen in fuel cells, and sodium acetate in industry as a food additive. 76 Acetaldol and crotonaldehyde can be used as fuel in combustion engines and in various synthetic chemistry applications, such as, production of n-butyraldehyde, and crotonic acid. 77 Chromatograms of the solutions obtained after 60 CVs at a scan rate of 10 mV s −1 under light off and light on are presented in Fig. 9, for all investigated electrodes. The result for the ethanol photodegradation on pure TaON was included for comparison, since this photocatalyst is active to degrade ethanol under irradiation, although it does not catalyze the electro-oxidation of this molecule (Fig. 5b). Peaks at low retention times are related to acetate ions and un-reacted ethanol, so that it is concluded that acetate is an important product present in the liquid phase in all cases; acetaldehyde was not detected as a final product in none of the investigated conditions. In some cases, additional peaks are observed at higher retention times at 26.5 and 37.1 min, and these are related to acetaldol and crotonaldehyde, as also shown from Fig. 8b.
When the Pd/TaON electrode with or without irradiation is considered, results evidence that the major product of electrochemical oxidation is acetate, and the peaks of acetaldol and crotonaldehyde are absent. For IrO 2 /Pd/TaON these products are present but the signals are much smaller compared to those seen at open-circuit potential (Fig. 8b). This result indicates that Pd/TaON is active to electrooxidize acetaldol and crotonaldehyde to acetate, thus increasing the acetate production (red curve). Results for IrO 2 /Pd/TaON demonstrate a lower activity of this material compared to Pd/TaON to oxidize these intermediates, probably because iridium oxide partially blocks the Pd active sites, as mentioned before.
Concentrations of acetate in these experiments were determined by means of calibration curves recorded for the HPLC. These data are presented in Table I in terms of the fraction of ethanol conversion into products (obtained comparing the ethanol peaks (RID detector), before and after each measurement) and the concentration of acetate present in the solution after the experiments involving 60 CVs cycles with or without irradiation. It is evident that the acetate production is significantly increased when the photocatalysts (Pd/TaON and IrO 2 /Pd/TaON) are irradiated, and this confirms the electro-oxidation of the species produced in large amounts during the ethanol photodegradation. It is worth mentioning that Pd/TaON leaded to the highest acetate concentration, and this is due to the total electrooxidation of crotonaldehyde and acetaldol to CH 3 COO − , as indicated by the absence of HPLC peaks of these products. Moreover, Pd/TaON presented the higher efficiency to convert ethanol into products under photoelectrochemical conditions, as observed from the acetate formation, whose concentration corresponds to the double of the value observed for the same electrode in the dark condition. Based in these observation, the electrochemical steps can be proposed as:    17 the photodegradation of organic compounds usually produce intermediates with a relatively long lifetime, making possible the identification of these intermediates, and allowing to determine all structures/species for establishing of the protodegradation mechanism. Antoniadou, Ciambelli and co-workers 75,78 have shown that the ethanol photoreforming reaction in titanium dioxide electrodes in alkaline medium produces hydrogen and compounds with molecular weights higher than those of the usual products from ethanol electro-oxidation. 75 Our present results obtained in the TaON-based catalysts had confirmed occurrence of similar phenomena.

Conclusions
XRD results indicate that composition of the prepared TaON material refers to the beta monoclinic structure (β -TaON, 96 wt%), with average crystallite size of 50 nm. Elemental analysis by XPS showed that the photocatalysts composition are TaO 1.5 N 0.80 (Pd/TaON) and TaO 2.4 N 0.76 (IrO 2 /Pd/TaON), with atomic percentages of Pd and Ir around to 2.03 and 0.13 at.%, respectively. (HAADFSTEM-EDS) elemental mapping images evidenced uniform distributions of all elements in the samples (Ta, Pd, Ir, O and N), while TEM images for Pd/TaON and IrO 2 /Pd/TaON showed that Pd and IrO 2 depositions on the tantalum oxynitride surface generated highly dispersed particles with diameter ranging from 2 to 10 nm for Pd.
Cyclic voltammetric experiments showed that the presence of Pd is essential to improve the electrochemical response and to promote the ethanol electro-oxidation. These results show that the tantalum oxynitride photocatalyst, when combined with Pd, produces a higher amount of reactive species by ethanol photodegradation. Results also show that the Pd/TaON photocatalyst exhibited more stable electrochemical response especially when under irradiation. Hydrogen, acetate ions, acetaldol and crotonaldehyde are the main final products of ethanol photodegradation.
Acetate ions and hydrogen are produced from ethanol and water photodegradations, respectively, while acetaldol and crotonaldeyde result from the coupling of acetaldehyde intermediate molecules.
When the Pd/TaON electrode with or without irradiation is considered, the major product is acetate ion, showing that this catalyst is active to electro-oxidize acetaldol and crotonaldehyde to acetate. For IrO 2 /Pd/TaON, results demonstrate a lower activity to oxidize these intermediates, probably because iridium oxide partially blocks the Pd active sites.