Electrochemical Dechlorination of 2-Chlorophenol on Pd/Ti, Ni/Ti and Pd-Ni Alloy/Ti Electrodes

The 2-chlorophenol has been treated by advanced oxidation processes (AOPs) and by electrochemical oxidation, but these methods show several disadvantages, such as: high-energy costs, hazardous substances production and electrode deactivation. To avoid the problems caused by oxidation, electrochemical reduction (dechlorination) was proposed as an alternative and promising method. Pd/Ti,Ni/TiandPd-Nialloy/Ticathodeswereusedforelectrochemicaldechlorinationof2-chlorophenol.Thecathodeswerepreparedbyelectrodepositionunderexperimentalconditionsdeterminedfrombothathermodynamicstudybyusingpredominanceexistenceanddistributionspeciesdiagramsandavoltammetricstudy.TheelectrodeswerecharacterizedbySEM,XRD,andXPS.SEMimagesshowedthatPd-Ni/TielectrodehasadifferentmorphologyfromthatofPd/TiandNi/Tielectrodesandthealloycompositionis45%Pdand49%Nimol.XRDanalysesshowedthatPd-Nialloy/TiexhibitsanetworkparameterdifferentfromthoseofPd/TiandNi/Tielectrodes.TheXPSprovedtheformationofPd-Nialloy.Theefﬁcientdechlorinationof2-chlorophenoltophenolwasachievedunderelectrochemicalconditionswhereprotonreductionandatomichydrogenadsorptiontookplace.ThePd-Nialloy/Tielectrodehadthehighestdechlorinationefﬁciency(100%removal)andphenolformation(100%formation)atapotentialof

2-Chlorophenol (2-CP) is a toxic, recalcitrant compound that has been classified as a priority pollutant by the United States Environmental Protection Agency (USEPA). 1,2 Nowadays, the accumulation of this chlorinated aromatic compound in the environment is an urgent problem that needs solving. 2-CP is very slowly consumed by microorganisms and is difficult to treat with conventional microbiological technologies. 3 One alternative is incineration that involves high costs and causes chemical transformations that produces hazardous substances such as dioxins and polychloro biphenyls. Advanced oxidation processes (AOPs), such as Fenton's reagent, ozone, UV, UV/H 2 O 2 , and UV/Fenton have been studied for the purpose of degrading chlorinated organic compounds; however, the efficiency of these processes requires choosing adequate, but high-cost, catalytic materials. 4,5 Another method explored for chlorinated organic compound degradation is electrochemical oxidation; however, the production of a polymer film at the electrode surface causes electrode deactivation. 6 In electrochemical oxidation, organic compounds can be oxidized up to carbon dioxide using a Boron Doped Diamond (BDD) electrode, but the electrode preparation is quite costly. 7 In order to avoid the problems caused by oxidation, electrochemical reductive dechlorination has been suggested as an alternative and promising method, because of several advantages, such as: rapid reaction, low-cost equipment, operating at room temperature and atmospheric pressure, and no production of secondary pollutants more toxic than the original. 8,9 Electrochemical dechlorination is an indirect reaction known as electrocatalytic hydrogenolysis (ECH). 10 In a ECH mechanism, the chemisorbed hydrogen atoms (H ads ) generated on the electrode surface by water electrolysis provide the driving force for chemical reduction causing the C-Cl bond cleavage from 2-CP. ECH involves several steps as described in reactions 1 through 6 11,12 :  [5] (H ) ads M + (H ) ads M → H 2 + M Tafel step [6] where M represents the electrocatalytic metal or alloy.
The key steps of the ECH process are proton reduction (H + ) and hydrogen adsorption (H ads ) (reaction 1). The adsorbed hydrogen reacts in two different ways: the 2-CP reduction to phenol (reactions 3 and 4) and the electrochemical/chemical H 2 formation (reactions 5 and 6). H 2 generates a layer on the electrode surface inhibiting mass transfer of 2-chlorophenol to the catalysts, decreasing the activity of the ECH. 13 Thus, nature and surface characteristics of the electrode material and potential at which the reduction takes place are of paramount importance in electrochemical dechlorination efficiency. 14 Palladium (Pd) is considered the ideal for electrochemical dechlorination due to its capacity in hydrogen adsorption, but the preparation of pure palladium electrodes are high costs. 15 An alternative that has been proposed is the use of alloys capable of reducing the proton (H + ) and adsorbing hydrogen (H ads ) on the surface. Sun et al. (2012) 12 observed that a Pd-Ni/Ti electrode could perform electrochemical dechlorination of 2,4-dichlorophenol, generating phenol and cyclic non aromatic hydrocarbons as final products. However, these authors neither described the experimental conditions for electrodeposition of Pd-Ni alloy nor the influence of reduction potential on dechlorination efficiency. Therefore, it is of special interest to investigate more on the electrochemically dechlorination capacity of alloy electrodes for degrading chlorinated organic compounds at high process efficiency.
In the present work, a thermodynamic study was included to establish experimental conditions, such as pH and NH 4 Cl concentration at which ammonia complexes of Pd and Ni are present in electrodeposition baths, and to prevent the formation of insoluble species that hinder the electrodeposition. The potentials to prepare Pd/Ti, Ni/Ti, and Pd-Ni/Ti electrodes were established as well. These electrodes were characterized by cyclic voltammetry, SEM, XRD and XPS, and their performance in 2-CP dechlorination was analyzed in a comparative study that allowed establishing the influence of nature surface electrode and reduction potential on the process efficiency.
Preparation and characterization of Pd, Ni and Pd-Ni electrodes.-Titanium plates were subjected to surface pretreatment by sandblasting (steel balls); afterwards, they were rinsed with Millipore-Q water in an ultrasound bath, washed with oxalic acid 10% w/w at 70 o C to remove oxides from the surface, rinsed again in an ultrasound bath with Millipore-Q water, and stored in ethanol for later use.
Three electrodeposition baths were prepared: palladium bath, nickel bath and palladium-nickel alloy bath. Palladium and nickel electrodeposition baths were prepared by dissolving separately PdSO 4 4 Cl under constant stirring. In all cases, the concentrations of Pd(II) and Ni(II) were 38 mM and 120 mM, respectively, and the baths' pH was adjusted at 8.5 with sodium hydroxide.
The deposits were prepared on titanium plates of 2.4 cm 2 geometric area in a glass cell under constant potential conditions: −0.95 V vs. Ag/AgCl (s) /KCl (sat) for palladium bath, −1.30 V for nickel bath and −1.25 V for bimetallic Pd-Ni bath. In all cases, the deposition time was 20 minutes and the deposition bath was stirred with a magnetic stirrer at 50 rpm. Then, the electrodes were rinsed with Millipore-Q water to eliminate the remaining electrolyte.
Voltammetric studies were carried out using PAR 273 A type potentiostat/galvanostat with a conventional three-electrode system to determine electrodeposition parameters and characterize the formed electrodes as well as for potentiostatic formation of the electrodes. A graphite bar was used as counter electrode and an Ag/AgCl (s) /KCl (sat) electrode as reference. Working electrodes for electrochemical studies were the previously prepared Pd/Ti, Ni/Ti and Pd-Ni/Ti electrodes. The H + reduction and hydrogen adsorption were performed using 0.1 M H 2 SO 4 solution. The electrodes were characterized in terms of morphology, composition and crystal structure. Surface morphology and elemental composition were observed and analyzed through scanning electron microscopy (SEM) in a JEOL ultrahigh resolution field emission electron microscope JSM-7800 F, with 15 kV accelerating voltage and 10.1 mm WD. The qualitative elemental analysis of electrodes' surface was performed with an integrated system for energy dispersive X-ray spectroscopy (EDX), EDAX/AMETEK Apollo X with a 30 mm 2 silicon drift detector (SDD). Crystal structure of electrodeposits was analyzed by X-ray diffraction (XRD) using a Siemens D-500 Kristalloflex apparatus. In addition, X-ray photoelectron spectroscopy (XPS) analysis was performed using a Thermo Scientific K-Alpha X-ray photoelectron spectrometer provided with a monochromatized AlKα X-ray source (1,487 V). XPS wide and narrow spectra were collected using an X-ray spot size of 400 μm 2 at 160 and 60 eV pass energy, respectively. The surface of the samples was ion beam etching (IBE) cleaned with Ar at ion acceleration potential of 3.0 keV for 3 min in order to remove the major part of the oxide layer before XPS testing.
Electrochemical dechlorination of 2-chlorophenol.-Electrocatalytic activities of the prepared electrodes were evaluated for 2-CP degradation. The experiments were carried out in a cell array split in two half-cells by a salt bridge ( Figure 1) that was utilized to prevent chloride ion (Cl − ), generated on the cathode during dechlorination process, from being transported to the anode surface to form Cl 2 or to prevent 2-CP oxidation. The 2-CP electrolysis at a controlled potential was performed at room temperature using Pd/Ti, Ni/Ti and Pd-Ni/Ti electrodes as cathode and a graphite rod as anode. During dechlorination, the electrolyte was stirred with a magnetic bar at 200 rpm. The electrolyte volume in the cathodic compartment was 60 mL in aqueous solution containing 0.1 M H 2 SO 4 and 50 mM Na 2 SO 4 , whereas the 2-CP concentration was 50 mg L −1 (0.388 mM). 2-CP and phenol were analyzed by high performance liquid chromatography (HPLC) (Perkin Elmer 200 UV series), with a C-18 column (Phenomenex), acetonitrile-water (60:40 v/v) as mobile phase and a flow rate of 1.5 mL min −1 . Phenolic compounds were detected at 274 nm.

Results and Discussion
Thermodynamic study.-For the purpose of determining the pH and NH 4 Cl concentration necessary for the formation of Ni(II) and Pd(II) complexes with ammonium that prevent formation of insoluble species in electrodeposition baths, a thermodynamic study was carried out using species predominance-zone diagrams (PZDs) and distribution diagrams. PZDs were built with The Chemical Equilibrium Software (MEDUSA), 16 (Figure 2b). PZDs only show predominant chemical species (more important fraction) in a given concentration zone; however, it is also necessary to know the fraction of other species in the solution that may affect the processes to be studied. Diagrams of the fraction of Ni(II) and Pd(II) chemical species were plotted at pH 8.5. The nickel fraction diagram (Figure 2c) shows that for a total Ni(II) concentration of 120 mM, the NH 4 Cl concentration must be above 2.4 M to prevent the formation of Ni(OH) 2 solid and at such concentrations, the ammonia complexes Ni(NH 3 ) 5 2+ and Ni(NH 3 ) 6 2+ present the greatest fractions, 50% and 30% respectively, followed by lower fractions of Ni(NH 3 ) 4 2+ and Ni(NH 3 ) 3 2+ complexes. For palladium (Figure 2d), the diagram shows that at concentrations above 0.25 M NH 4 Cl, the Pd(NH 3 ) 4 2+ complex with 100% fraction predominates. Taking into account that ammonia complexes of Pd(II) and Ni(II) ions allow controlling the process of electrodeposition, the conditions established for electrodeposition baths of Ni(II), Pd(II) and Ni-Pd Electrodeposition potential selection.-The selection of the potential for electrodeposition and formation of catalytic deposits was carried out using a voltammetric study in different electrodeposition baths containing Pd(II), Ni(II) and Pd(II)-Ni(II) under the following conditions: 38 mM Pd(II), 120 mM Ni(II) and 38 mM-120 mM Pd(II)-Ni(II), pH 8.5 and 2.5 M NH 4 Cl, on a titanium rotating disc electrode (A = 1.2 cm 2 ) with constant 100 rpm stirring. When the potential scan for Pd(II) and Ni(II) baths (Figures 3a and 3b) is initially conducted in the cathodic direction, reduction peaks c 1 and c 2 appear at potentials (E pc ) of −0.95 V and −1.30 V, respectively, which are attributed to the reduction process of Pd(II) and Ni(II) species, respectively. However, when the potential is reversed, oxidation peaks a 1 and a 2 appear at potentials of −0.20 V (E pa1 ) and −0.40 V (E pa2 ) that are attributed to the oxidation process of reduced species deposited on titanium surface. Figure 3c shows the voltammogram corresponding to the Pd(II)-Ni(II) alloy bath. When the potential scan is started in the cathodic direction, just one reduction peak c 3 appears at E pc3 = −1.25 V, and this potential is close to that obtained in Ni(II) bath (−1.30 V); however, no reduction peak of Pd(II) arises. Upon reversing the potential scan in the anodic direction, only one oxidation peak, a 3 , arises at E pa3 = −0.35 V. Such behavior may indicate the formation of Pd-Ni alloy, ruling out the independent deposition of Ni and Pd. In order to prove this hypothesis, deposits of each metal were prepared using reduction potentials at which the process is controlled by diffusion in the corresponding electrolytes (Figure 3). Once the deposits were formed on titanium plates (electrodes), they were characterized by different spectroscopic and electrochemical techniques. An important aspect of electrochemical dechlorination is the electrode surface because of its influence on H + reduction and H adsorption. Figure 4 shows SEM images of the different prepared electrodes. The Pd/Ti electrode (Figure 4a) shows a densely dendritic coating composed of large continuous agglomerates, whereas the Ni/Ti electrode ( Figure 4b) has a uniform deposit all over the surface. Lastly, the morphology of the electrode formed in the bath containing Ni(II) and Pd(II) (Figure 4c) is similar to that of Pd/Ti electrode and its coating composed of large continuous agglomerates, but in this case the agglomerates are much larger than in Pd/Ti. After elemental analysis by EDS made on Pd/Ti and Ni/Ti electrodes (Figures 4d and 4e), ZAF method was used to determine their composition consisting of 87% Pd and 85% Ni mol, respectively. The presence of oxygen was detected on both electrodes, 12% and 15% mol in the Pd/Ti and Ni/Ti electrodes respectively, because of the layer of palladium and nickel oxides formed on the surface of electrodeposits, corroborated by XPS analysis (see below). The electrode formed in the bath containing both Ni(II) and Pd(II) (Figure 4f) was estimated to have the composition of 45% Pd and 49% Ni mol of palladium and nickel, respectively. This proved that deposition of both metals was accomplished in the Pd(II)-Ni(II) bath at a potential of −1.25 V. Likewise, as in the previous case, the presence of the oxygen peak in EDS spectrum gave proof of the presence of a layer of oxides with 6.2% mol oxygen content.
Then, images obtained by back-scattered electron detector (SEM-BSE) were used to determine the distribution of oxygen, nickel and palladium in the different aggregates that make up the surface of the electrodes. Figure 5a shows the SEM-BSE image of Pd-Ni/Ti alloy exhibiting white and gray zones that were associated with the presence  of higher and lower average atomic numbers (Z), respectively. Thus, white-colored aggregates are mainly made up of a metallic phase (zone 1) whereas darker aggregates are richer in oxygen (zone 2). The above findings were corroborated by punctual EDS analyses carried out in each of these zones (Figures 5b and 5c). Pd-Ni/Ti) coincides with the orientation of (111) and (200) planes that are associated with a body-centered cubic structure (BCC). Diffractograms were used to calculate d-spacing for (111) plane of each electrode (Table I). Network parameter of Pd-Ni alloy is between the values of Pd and Ni, higher than that of Ni and lower than Pd's which demonstrates that a change in the crystal lattice occurs during formation of Pd-Ni/Ti electrode.
X-ray photoelectron spectroscopy.-In order to corroborate the formation and homogeneity of Pd-Ni alloy on Ti electrode, three measurements at different spots on the electrode surface were taken and employed to obtain XPS wide and narrow spectra. To remove the major part of metallic oxide formed by exposure of electrodes to air, the sample underwent an IBE in a vacuum chamber for 3 minutes. XPS wide spectra of Pd-Ni/Ti electrode before and after IBE are shown in Figure 7a, where besides different signal features, C1s, Pd3d, O1s and Ni2p main peaks are distinguishable and identifiable. 18 Figure 7b shows the evolution of Ni2p region after different IBE stages (level 0, without IBE) and how an appreciable decrease in oxide and satellite signals is observed while the metal peak increases in intensity.
For a more detailed determination of oxidation states and composition of Pd-Ni chemical species formed on the alloy electrode, analysis was performed by decomposition-based modeling of the highresolution XPS narrow scans for Pd3d. To detect and compensate the charge shift of the core level peaks, O1s peak position at 531.0 eV was used as an internal standard instead of C1s. The foregoing was done because carbon is neither the main component nor it forms continuous and homogeneous layers over the electrode surfaces. 19,20 Pd3d  doublet spectra were fitted using a Gaussian-Lorentzian mix function and Shirley background sub traction.
XPS narrow scan of Pd3d doublet spectrum ( Figure 8) shows that 3d 3/2 and 3d 5/2 were fitted with three double contributions. One is metallic palladium at 336.7 ± 0.2 eV and 341.0 ± 0.2 eV for Pd3d 5/2 and 286 Pd3d 3/2 peaks, respectively. The second contribution corresponds to palladium oxide where Pd3d 5/2 is located at 335.1 ± 0.2 eV and Pd3d 3/2 at 340.4 ± 0.2 eV. The third contribution is for Pd3d 5/2 and Pd3d 3/2 related to Pd-Ni alloy, which are located at 335.7 ± 0.2 eV and 341.0 ± 0.2 eV, respectively. These latter peaks show the major contribution of the total peaks (Figure 8), which confirms the formation of the Pd-Ni alloy formation.
Hydrogen adsorption.-The study of H + reduction and H adsorption on Ni/Ti, Pd/Ti and Pd-Ni alloy/Ti electrodes (2.4 cm 2 area) was  Figure 9). The potential scan started in the negative direction, Pd/Ti and Ni/Ti electrodes showed that the current increases as the potentials are more negative, and this increase corresponds to the process of proton reduction. Upon reversing the potential scan in the positive direction, peaks a 4 and a 5 arise at potentials E pa4 and E pa5 of 0.25 V and 0.38 V, respectively (Figures 9a and 9b); these peaks, absent in the case of the scan initially done in the positive direction, correspond to the oxidation of hydrogen (H ads ) that was adsorbed (H ads ) on the surface of the electrodes in the direct scan. 21 In the case of Pd-Ni/Ti alloy (Figure 9c), hydrogen reductionoxidation process is totally different from that occurred on separate metals, since independent deposition of Pd and Ni is ruled out, and interaction between the metals (alloy) is confirmed. In this case, two well-defined reduction peaks, c 4 and c 5 , appear at potentials E pc4 and E pc5 of −0.15 V and −0.30 V, respectively. These peaks are attributed to H + reduction to form H ads in two different crystal planes; whereas at a potential scanned to more negative values, the current continues to increase due to reduction of the medium and the evolution of H 2 . When the potential is reversed in the anodic direction, two oxidation peaks, a 6 and a 7 , appear at potentials E pa6 and E pa7 of −0.1 V and −0.028 V, respectively, which are attributed to the oxidation of adsorbed hydrogen (H ads ). According to the results obtained by XRD characterization, the Pd-Ni alloy exhibits two (111) and (200) planes in which the process of H ads adsorption is more important (higher reduction peaks) than in the same planes of pure metals. Thus, at −0.15 V and −0.30 V potentials or those more negative than −0.30 V, the presence of adsorbed hydrogen is ensured in order to carry out the dechlorination of 2-CP.  2H ads /M + 2 − chlorophenol → ← phenol + 2M + HCl [8] The efficiency of dechlorination and phenol formation is obtained in the following manner: Where: 2CP 0 = initial 2-chlorophenol concentration 2CP t = final 2-chlorophenol concentration P 0 = stoichiometric formed phenol concentration (theoretical) P t = final formed phenol Table II shows the results of dechlorination with the three electrodes at different electrolysis potentials. In the case of Ni/Ti electrode, the 2-CP removal, phenol formation was practically independent of the imposed potential. The 2-CP removal was 51% at the three potentials imposed, whereas phenol formation efficiency kept lower than 18% with respect to stoichiometry, Equation 8, so nickel was not a good catalyst to carry out electrochemical dechlorination of 2-CP. For Pd/Ti and Pd-Ni/Ti electrodes, the potential had an effect on the efficiency of 2-CP removal, being higher at more negative potentials. On the Pd/Ti electrode, 2-CP removal reached 60%, which could be associated with the fact that H ads is absorbed in the bulk of palladium 21 decreasing the amount of H ads available on the surface to carry out the complete dechlorination of 2-CP. In the case of Pd-Ni alloy/Ti electrode, the 2-CP removal reached 100% because the imposed potentials ensure the existence of H ads on the surface (as shown in Figure 9c), and the amount of H ads absorbed in the bulk of Pd-Ni alloy is lower as compared to Pd/Ti electrode, 21,22 remaining available on the surface for further dechlorination of 2-CP. The phenol formation appeared to follow the same behavior, but upon imposing a more negative potential (−0.60 V), the efficiency decreased on both electrodes reaching even 0% in the case of Pd-Ni/Ti alloy and 40% on Pd/Ti electrode. This was due to the fact that at this potential, the process of H 2 evolution becomes present affecting the mechanism of electrochemical dechlorination, transforming 2-CP to other not identified intermediates in HPLC. The Pd-Ni/Ti electrode turned out to be the best option to carry out the process of electrochemical dechlorination since an efficiency of 100% was obtained for both 2-CP removal and phenol formation at a potential of −0.40 V.
Phenol is a less toxic, inhibitory, and recalcitrant compound than 2-CP for microorganisms and it is more easily degradable in biological systems. The electrochemical process obtained in the present study might be coupled to a low-cost biological process where phenol could be totally mineralized. 23 Under the experimental conditions established in this study, the stoichiometric transformation of 2-CP into phenol was obtained, limiting the accumulation of others intermediates that could be toxic or inhibitory for the subsequent biological treatment. It has been reported that the limiting step in 2-CP mineralization by microbial processes is the reductive dechlorination of 2-CP to phenol, making very slow the overall process. In the present study, 50 mg/L of 2-CP was reduced into phenol after only 3 h of electrolysis. Results from the present study showed that the electrochemical dechlorination of 2-CP might constitute an alternative of pretreatment for the biological treatment of industrial wastewater polluted with chlorinated phenols.

Conclusions
2-chlorophenol is a compound difficult to degrade by biological methods; in chemical and electrochemical oxidation processes it presents several problems that need solving. Therefore, reduction by electrochemical dechlorination has been carried out in this study as an alternative process to provide a less toxic and more easily degradable compound for biological systems. Electrolytic bath conditions have been set for the preparation of three electrodes containing Pd/Ti, Ni/Ti and Pd-Ni alloy, which were later characterized by XRD, SEM, EDS and XPS. The voltammetric study helped to detect potential conditions (-0.40 V) to carry out electrochemical dechlorination of 2-CP on the Pd-Ni/Ti alloy electrode, where the dechlorination process was selective for phenol. The obtained dechlorination conditions will be used to operate an (ECCOCEL-type) electrochemical reactor and carry out 2-CP dechlorinationat a larger scale as pretreatment to feed a biological reactor (denitrifying process) for its complete mineralization.