Environ. Sci. Technol. 2004, 38, 3203-3208
Photoelectrocatalytic Production of Active Chlorine on Nanocrystalline Titanium Dioxide Thin-Film Electrodes
unavoidable side reaction present in most anodic processes
MARIA VALNICE B. ZANONI,† JEOSADAQUE J. SENE,‡ HUSEYIN SELCUK, AND MARC A. ANDERSON* Environmental Chemistry and Technology Program, University of Wisconsin-Madison, 660 North Park Street, Madison, Wisconsin 53706
These anodic reactions simultaneously take place with the following primary cathodic reaction of hydrogen evolution:
2H2O f O2 + 4H+ + 4e-
(2)
2OH- f O2 + 2H2O + 4e-
(3)
or
2H2O + 2e- f 2OH- + H2
The production of chlorine and hypochlorite is of great economical and technological interest due to their largescale use in many kinds of commercial applications. Yet, the current processes are not without problems such as inevitable side reactions and the high cost of production. This work reports the photoelectrocatalytic oxidation of chloride ions to free chlorine as it has been investigated by using titanium dioxide (TiO2) and several metaldoped titanium dioxide (M-TiO2) material electrodes. An average concentration of 800 mg L-1 of free chlorine was obtained in an open-air reactor using a TiO2 thin-film electrode biased at +1.0 V (SCE) and illuminated by UV light. The M-doped electrodes have performed poorly compared with the pure TiO2 counterpart. Test solutions containing 0.05 mol L-1 NaCl pH 2.0-4.0 were found to be the best conditions for fast production of free chlorine. A complete investigation of all parameters that influence the global process of chlorine production by the photoelectrocatalytic method such as applied potential, concentration of NaCl, pH solution, and time is presented in detail. In addition, photocurrent vs potential curves and the reaction order are also discussed.
(4)
Therefore, the appropriate selection of the electrode material is a prerequisite for optimizing this electrolysis processes. Thus, there has been a great deal of research to find an anodic material with optimal electrocatalytic properties: high selectivity, easy availability and low cost, long mechanical and chemical stability, a desirable health safety record, and other characteristics (17). Several reports (13, 15, 4) describe the use of anodic materials such as graphite, platinum, and metal oxides having good current efficiency for producing hypochlorite ions. Nevertheless, the oxidation of chloride ion and oxygen evolution occurs only at a high positive potential on all of these electrodes. Others have used dimensionally stable anodes (DSA), based on thermally prepared oxide electrodes with a titanium substrate covered by metallic oxides such as TiO2, IrO2, RuO2, and Ta2O3, for the oxidation of chloride ions (4, 5, 13-18). These DSA electrodes presented an improvement in catalytic activity, decent mechanical and chemical stability under high positive potential, good resistance to corrosion, and high current efficiency. However, most of these investigations have been conducted for brines in weak alkaline solution at a high constant current density that promotes a large potential generation, which diminishes the lifetime of the electrodes. The generation of hypochlorous acid and hypochlorite ion is the main side reaction in the anodic production of chlorine, due to hydrolysis of dissolved chlorine in the bulk solution (4, 5) according to the following equation:
Cl2(aq) + H2O f Cl-(aq) + HClO + H+(aq)
Introduction The electrolytic production of chlorine and hypochlorite from chloride is widely used in industry (1-3) and for the disinfection of drinking water (4-8). Accordingly, oxidation of the chloride ion on a variety of anodic materials has been extensively investigated (1, 9-18). The main anodic reaction is the formation of chlorine:
2Cl- f Cl2 + 2e-
(1)
Although chlorine production is one of the largest technological applications of electrochemistry based on direct anodic discharge of chloride ion, oxygen evolution is an * Corresponding author phone: (608)262-2674; fax: (608)262-0454; e-mail:
[email protected]. † Present address: Instituto de Quimica-UNESP, Caixa Postal 355, 14800-901, Araraquara-SP, Brazil. ‡ Present address: Fundac ¸ a˜o Educacional de Barretos, Av. Prof. Roberto Frade Monte, 389, CEP 14783-226, Barretos-SP, Brazil. 10.1021/es0347080 CCC: $27.50 Published on Web 04/23/2004
2004 American Chemical Society
(5)
Hypochlorous acid can dissociate to form hypochlorite ion:
HClO f ClO- + H+
(6)
In the water treatment field (19-21), the overall concentration of dissolved chlorine available after the chlorination process is called active chlorine and is given by the summation of three possible species: chlorine (Cl2), hypochlorous acid (HOCl), and hypochlorite ion (OCl-). The relative amount of each of these free chlorine forms is pH and temperature dependent. At room temperature and 1.0 mmol L-1 of chloride, Cl2(aq) has an important contribution under very acidic conditions and the HClO species predominates due to hydrolytic disproportionation at 3.3 < pH < 7.5, while OCl- is the main species at pH > 7.5. Control over pH can be a critical factor in determining the degree of disinfection achieved by a certain level of free chlorine. This has more than academic interest because the disinfecting ability of hypochlorous acid is generally regarded to be larger (80-100 times) than that of hypochlorite ion. VOL. 38, NO. 11, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3203
In some water treatment processes, the use of electrochemical disinfection has been based on the use of active chlorine produced as a main disinfecting agent from naturally (10-250 mg L-1) occurring chloride ions (4, 5, 22, 23). This is a favorable use of electrochemical methods in that there is no need for addition of any chemical. For other systems that do not have significant quantities of chloride ions, the disinfection typically requires the use of gaseous chlorine or concentrated hypochlorite solution, which is hazardous with respect to handling, storage, and transportation. The performance of photoelectrodes consisting of thinfilms of titanium dioxide coated on a titanium substrate has been evaluated previously (24-26). The nanostructured semiconductor has proven to be electrochemically stable, resistant to corrosion, conductive, and has been successfully applied in the photoelectrocatalytic degradation of organic pollutants. The process combines the photocatalytic activity of TiO2 under UV light with an electrochemical applied potential aiming to enhance the efficiency of the overall process. The function of the bias potential in this case is in promoting photogenerated charges as to increase the lifetime of electron-hole pairs. The influence of chloride ions on the photoelectrocatalytic degradation of organic contaminants has been observed previously (24, 27, 28). Highest rate of azo dye degradation was found to occur in chloride solution, where fast dye mineralization was attributed to the active chlorine production identified after photoelectrocatalysis measurement. However, no systematic studies have been reported dealing with the possibility to use the photoelectrocatalytic conversion of chloride ion. In addition, despite the frequent use of electrochemical disinfection in water treatment and the importance of the electrode material in the efficiency of chlorine production along with the advantage of the photoelectrocatalytic technique, reports dealing with the application of titanium dioxide in the photoelectrocatalytic oxidation of chloride ions to generate active chlorine are not found in the specialized literature. Accordingly, the present study is aimed at evaluating the production rate of active chlorine and the most important parameters affecting the performance of these novel photoanodes on chloride oxidation.
Experimental Section Catalyst and Photoelectrodes. Titanium(IV) isopropoxide (Aldrich) was used as a precursor for preparing TiO2 colloidal suspensions. Typically, 20 mL of titanium isopropoxide was added to a nitric acid solution keeping the ratio Ti/H+/H2O at 1/0.5/200. The resulting precipitate was continuously stirred until completely peptized to a stable colloidal suspension. This suspension was dialyzed against milli-Q water to a pH 3.5 by using a Micropore 3500 MW cutoff membrane (29). Photoelectrode thin-films were cast onto titanium foil back contacts (0.05 or 0.5 mm thick, Goodfellow Cambridge Ltd.) following a sequence of dipping, drying, and firing at 400 °C for 3 h. Further details are available in earlier reports (30, 31). Preparation of Doped Catalysts. Titanium(IV) ethoxide (Gelest) and the correspondent metal alkoxide precursor (Alfa Aesar) were used for preparing the M-doped TiO2 colloidal suspensions. Typically, the required volume of the doping reagent (enough for 1.0% atom ratio) was added under argon to 10 mL of titanium ethoxide, which was diluted with 10 mL of anhydrous ethanol. The mixture was stirred for 48 h and then hydrolyzed in enough 0.15 mol L-1 HNO3 aqueous solution to keep the ratio of H2O/Ti/H+ at 200/1/0.5. Immediately after adding the alkoxide mixture to the nitric acid solution, the alcohol was boiled off at 80 °C, and the sol was stirred until a stable colloidal suspension was obtained, which typically required about 2 days. This suspension was 3204
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 11, 2004
then dialyzed against ultrapure water to pH 3.5 by using a Spectra/Por 3 regenerated cellulose membrane (Spectrum Medical Industries, Inc.) with a molecular weight cutoff of 3500 daltons. These processing steps removed most of the organics from the suspension. The thin-film photoelectrodes were prepared by using the same procedure described for pure TiO2 and then heated to 400 °C for 3 h to completely eliminate all organics traces. Photoelectrochemistry. Electrochemical and photoelectrochemical measurements were conducted in a twocompartment open-air reactor isolated by a Nafion 117 membrane. Each compartment holds a 100 mL volume. A Princeton Applied Research (PAR) potentiostat model 6310 was used to bias the working photoanode electrode against a perforated platinum foil serving as auxiliary electrode. A saturated calomel electrode was used as a reference. All experiments were carried out utilizing 100 mL test solutions in each compartment. These test solutions typically consisted of 0.005 to 1.0 mol L-1 NaCl or other specified conditions. The photoactive area of the anode was 20 cm2 and was illuminated by a 450 W Xe-Hg arc lamp Oriel, model 6262 UV light source. The light intensity on the electrode surface was 50 mW cm-2 as measured with an International Light Inc. photometer, model IL 1400A. Measurements of photocurrent vs potential curves were performed using a single compartment rectangular Teflon cell with a 25 mL capacity that has been described previously (32). The cell contained a borosilicate glass window 0.25 mm in diameter that allowed the UV illumination to reach the titania-coated working electrode. An area of 4.6 cm2 of the working electrode was exposed to solution and UV irradiation. A platinum foil with approximately the same area was used as counter electrode. A bridge tube with a Vycor frit was connected to a saturated calomel reference electrode (SCE) and placed 5 mm in front of the working electrode. All the potentials given are referred to the SCE. Oxygen was bubbled into the solution during all experiments. The total time of electrolysis was kept short, less than 1 h, to obtain maximum performance without a significant change in the composition of the electrolyte. The pH of the solution was measured with a double-junction combination electrode (Orion Model 8172BN) connected to a Fisher Scientific Accumet 50 pH meter. In experiments where pH was to be kept constant, 0.1 M NaOH solution was added to the cell in order to control the pH. Since the complete differentiation of chlorine species was not required, the content of free chlorine dissolved in solution was determined by the DPD (N,N,-diethyl-p-phenylenediamine) colorimetric method (21) calibrated previously to a minimum detectable concentration of chlorine of 10 µg/L. Aliquots of 150 µL of the photoelectrolyzed solution were collected in a tube containing phosphate buffer and DPD solution. The product of the instantaneous reaction between free chlorine and DPD indicator produced a red color which was analyzed immediately by measuring the absorption spectra in the visible range of 400-800 nm with a HewlettPackard spectrophotometer, model HP 8452A in a 10 mm quartz cell. The concentration of chlorine (mg/L) in solution was determined by measuring the absorbance at 512 nm using a calibration curve prepared previously with potassium permanganate solution as recommended (21).
Results and Discussion The photoelectrochemical oxidation of 0.025 mol L-1 NaCl at pH 4 on a TiO2 thin-film photoanode under UV light and a potential of +1 V is shown in Figure 1, curve 1. These data were obtained by evaluating the amount of active chlorine generated as a function of time, in aliquots of the electrolyzed sample removed each 5 min during the electrolysis and
FIGURE 1. Phoelectrocatalytic generation of active chlorine as a function of the electrode material: (1) TiO2; (2) Ni(II)-TiO2; (3) Co(II)-TiO2; (4) V(V)TiO2; and (5) Cr(III)-TiO2 thin-film electrodes in 0.025 mol L-1 NaCl at +1.0 V vs SCE under UV illumination. analyzed instantaneously by the DPD colorimetric method. The results obtained show that active chlorine production is faster during the initial stage of the process and then reaches a maximum value after a longer period of time, suggesting that maximum saturation has been attained in addition to the expected limitation imposed by the geometric area of the electrode. The electrochemical oxidation under these experimental conditions, i.e., without using UV illumination, results in no production of chlorine that is detectable by the colorimetric method. Similar results were obtained when the solely photocatalytic treatment was tested, showing that the combination of an external anodic bias with the UV illumination leads to a remarkable enhancement in the photocatalytic process thereby greatly improving the efficiency of chloride oxidation. To improve upon our process of generating chlorine by photoelectrochemical methods, we studied the effect of charge injection into the TiO2 photoanode. The photoelectrocatalytic activity of undoped TiO2 electrodes was compared with those of TiO2 thin-film electrode doped with 1%V5+; 1%Cr3+; 1%Ni2+; and 1%Co2+ (atom-rate) with respect to their abilities to produce active chlorine. Curves 2-5 in Figure 1 show a comparison of these electrodes on the photoelectrocatalytic active chlorine production. The data were obtained using 0.025 mol L-1 of NaCl at pH 4 electrolyzed during 30 min under UV light and a potential of +1 V, i.e., a more positive potential than that of the flat band potential for all electrodes. At all time during electrolysis, the active chlorine production rate is decreased for all doped TiO2 thinfilm electrodes. Nevertheless, the most dramatic effect occurs for V5+-doped and Cr3+-doped TiO2 (curves 4 and 5 in Figure 1, respectively) where the suppression in chlorine production is bigger than in the other doped catalysts. Taking into consideration that the lifetime of the photogenerated holes must be long enough to reach the surface of the photoanode and promote the oxidation of chloride ion into active chlorine, we believe that the incorporation of transition metal ions into the photoelectrode adds new trapping sites into the catalyst (32). The increase in the charge doping decreases the photoelectrocalytical chloride production due to fewer charge carriers being able to reach the surface. Although these are preliminary results, it is possible to conclude that the pure TiO2 thin-film is the best of these electrode materials for the production of active chlorine. Hence, other parameters that affect the suitability of the process such as pH of the solution, applied voltage, and NaCl concentration were further investigated by using such an electrode. Effects of Applied Potential. Figure 2 shows a typical voltammogram of photocurrent vs potential under dark conditions and under UV illumination for the Ti/TiO2
FIGURE 2. Current vs potential curves for TiO2 thin-film electrode in 0.5 mol L-1 NaCl solution, pH 3.0; scan rate ) 10 mV s-1: (1) photocurrent and (2) dark current. photoanode in NaCl 0.50 mol L-1, pH 3. This curve is similar to those previously reported for titania-based photoelectrodes (24-26). It can be seen that an anodic photocurrent (curve 1) arises only under UV illumination when the potential range is more positive than -0.38 V, and this potential can be correlated with the flat band potential of the photoelectrode (30). As the positive potential increases, the resulting gradient within the titania films separates holes and electrons, decreasing their rate of recombination. As a result, electrons in the conduction band flow via an external circuit to cathode where reduction takes place. Positive charges photogenerated in the valence-band participate in the reaction by forming free-radicals with the electrolyte species present in the solution. The photocurrent reaches a steady-state condition after potential of 0 V vs SCE, in which case a maximum separation of the pairs electron/hole is obtained by the electric field. As expected, no current is observed in dark conditions (curve 2). Taking into consideration the characteristic of curve 1 presented in Figure 2, the dependence of active chlorine production on the applied potential was investigated in 0.025 mol L-1 chloride at potentials ranging between -0.4 V to +1 V. The active chlorine production was monitored each 5 min during 30 min of photoelectrolysis. The chlorine generation is zero when the applied potential is -0.4 V. This potential is more negative than the flat band potential for this electrode as measured by the onset potential (-0.34 V at pH 4.0 in 0.1 mol L-1 NaCl). The dependence of the initial chlorine production rate on the applied potential was evaluated from the initial increase of active chlorine concentration (mol L-1 s-1) with time, and the results are presented in Figure 3. At lower potentials -0.2 V < Eapp < +0.4 V, there is a linear increase in active chlorine production rate, which appears to be constant at higher applied potentials. These results are important because they indicate that it is possible to generate active chlorine even at a potential as low as -0.2 V, despite the lower generation rate. In addition, in the practical use of photoelectrochemical chlorine production this linear increase of the chlorine amount with applied potential could easily be used to adjust the chlorine demand related with water quality. An applied potential higher than +0.4 V (vs SCE) indicates that maximum chloride conversion has been obtained, since the maximum charge separation has also been attained. Effect of pH. The influence of pH on active chlorine production was investigated by comparing the electrolysis results obtained in the pH range of 2-12 conducted on 0.50 and 0.025 mol L-1 NaCl on TiO2 thin-film electrodes under UV illumination at +1 V. The amount of chlorine produced during 40 min of electrolysis was monitored as a function of time after removing aliquots each 5 min. VOL. 38, NO. 11, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3205
FIGURE 3. Effect of applied potential on the initial rate of active chlorine production. CNaCl ) 0.025 mol L-1, pH 4.0.
FIGURE 5. Photocurrent vs potential curves for TiO2 thin-film electrodes in 0.5 mol L-1 NaCl solution, in different pHs; UV illumination; scan rate ) 10 mV s-1: (1) pH 3; (2) pH 5; (3) pH 9; (4) pH 11. the hydrogen evolution by eq 4 generates a concomitant increasing of the pH at the cathode. The experimental results have shown a final pH 11.3 at the cathode even using original acidic medium at the start of the photoelectrolysis. Therefore, our photoelectrochemical results seem to point to the fact that high active chlorine production is generated at the lowest pH values. This seems to indicate that probably there is a competion between chloride ion and water for the positive photogenerated holes. Therefore, under acidic conditions chlorine production is prevalent.
FIGURE 4. Comparison between photoelectrocatalytic active chlorine production as a function of time (1) and change in pH (2) during photoelectrolysis of a 0.025 mol L-1 NaCl solution. The correlation between pH variation and active chlorine production is exemplified in Figure 4. Curve (1) displays the pH changes that occur during a photoelectrocatalytic experiment where active chlorine is generated using 0.025 mol L-1 at an initial pH of 10 during a time period of 60 min under Eapp ) +1 V and UV illumination and the curve (2) shows the concomitant chlorine production. When one compares both curves in Figure 4, it is easy to see that there is a good correlation between the pH decreasing (curve 1) as a function of photoelectrolysis time with the consequent active chlorine generation (curve 2). At pH values higher than 7 (10 min of electrolysis) the active chlorine production is neglected, since the initial pH is reasonably constant. Nevertheless, it is noticed that the fastest active chorine production is obtained at the lowest pH values. The pH variation can be explained by the following equations:
TiO2 + hν f TiO2 - ecb- + TiO2 - hvb+
(7)
TiO2 + hvb+ + H2Os f TiO2 - OHs• + H+
(8)
TiO2 - hvb+ + OHs- f TiO2 - OHs•
(9)
TiO2 - hvb+ + Cls- f TiO2 - Cls•
(10)
This behavior is an indication that the pH of the solution in the anode compartment is decreased due to the proton evolution from the anode surface as indicated by eq 8 and hydrolytic disproportionation of chlorine in solution as shown previously by eqs 5 and 6. As these reactions acidify the anode, 3206
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 11, 2004
The influence of the pH on the photoelectrocatalytic process was further investigated using NaCl 0.5 mol L-1 solutions bubbled with oxygen over a pH range of 2-12. Figure 5 illustrates a typical dark current as well as photocurrent vs potential curves for the TiO2 thin-film electrode in both acidic and basic conditions. These curves in acidic medium show a cathodic and anodic wave. The small wave produced under the most reducing potential can be attributed to the reduction of dissolved oxygen into hydrogen peroxide (26). This occurs even under dark conditions and disappears after degassing the solution with nitrogen. The main anodic wave is present in the voltammogram only when the TiO2 electrode is illuminated by UV light as previously mentioned, and its intensity decreases as pH increases. Under dark conditions no photocurrent is observed since this electrode is made of an n-type semiconductor (33). In addition, in very alkaline medium a small cathodic wave attributed to oxygen reduction rises at less negative potentials under UV illumination. This behavior suggests that when the TiO2 thin-film electrodes are immersed in a solution of chloride ions under UV illumination the larger photocurrent is observed in acidic conditions because a higher flux of holes can reach the electrode surface, so becoming available for oxidation of chloride ions to chlorine. This larger photocurrent and the resulting oxidation of chloride ions are a consequence of the lower recombination rate of the photogenerated charge (electron/hole) obtained by the biased potential. Therefore, as verified previously by electrochemical methods, the anodic process involves simultaneously two reactions photoinduced: oxidation of water into oxygen and the oxidation of chloride to active chlorine. Therefore, these results indicate that, in acidic solutions, higher photocurrents are obtained because the adsorption of chloride ions on the electrode surface is preferential, which in turn favors the reaction of these species with the photogenerated holes and as consequence a great number of electrons are made available to reach the external circuit. The decreasing in the photocurrent
FIGURE 6. Effect of pH on the active chlorine production curves as a function of time. CNaCl ) 0.025 mol L-1, E ) +1 V vs SCE and UV illumination. (1) pH 2.0; (2) pH 4.0; (3) pH 6.0; (4) pH 7.0; (5) pH 8.0; (6) pH 10.
FIGURE 7. Effect of initial NaCl concentration on the photocurrent intensity measured at +1 V from photocurrent vs potential curves recorded at pH 3.6 under UV illumination. under basic conditions is fairly explained by the increase in the competitive recombination rate of the photogenerated charges as well as by the preferential adsorption of OH- on the electrode surface under these basic conditions. To further check this reasoning, new photoelectrocatalytic experiments were carried out to investigate the active chlorine production under controlled pH conditions. The concentration of chloride was 0.025 mol L-1, and the pH of the electrolyte was kept constant at each point (pH 2-12) during the entire 30 min of electrolysis at +1 V with the active chlorine generated at each pH being monitored in intervals of 5 min. The results are shown in Figure 6. As expected, chlorine production is much higher under acidic conditions (pH e 6). At higher pH, the active chlorine generation decreases sharply, and the process is completely suppressed at pH g 11. This behavior indicates that probably in basic media, when the pH is maintained constant, the photogenerated holes are involved mainly in the hydroxyl radical formation, which less the amount of active chlorine production. Then, for analytical purposes, pH 4 was chosen to obtain fast conversion of chloride into active chlorine without the need of correcting the pH of the solution. Effect of Chloride Ion Concentration. The effect of the initial NaCl concentration on the photocurrent intensity evaluated from photocurrent vs potential curves was obtained at pH 3.6 under UV light measured at Eapp ) 1 V, which is shown in Figure 7. The results confirm the data obtained previously, where a higher photocurrent intensity is seen
FIGURE 8. Effect of the initial concentration of NaCl on the phoelectrocatalytic production of active chlorine on TiO2 thin-film electrodes biased at +1.0 V vs SCE, test solution pH 4 and UV illumination: (1) 0.005 mol L-1; (2) 0.01 mol L-1; (3) 0.025 mol L-1; (4) 0.10 mol L-1. when the chloride concentration is increased. In agreement with the literature (25, 26, 33) the shape of the photocurrent vs potential curves depends only on the energy distribution of the incident photons, the adsorption coefficient of the semiconductor, the diffusion distance of the excited hole and electron and the recombination rates. Therefore, it is possible to suggest that chloride ion acts as a hole scavenger, which minimizes charge recombination. This results in an indirect effect on the anodic photocurrents. At lower concentrations of chloride, the electron/hole pairs generated at a steady-state rate can recombine preferentially since chloride ions are weakly adsorbed. However, at higher chloride concentrations, it is likely that the adsorption effect would predominate, avoiding charge recombination. The photoelectrocatalytic production of active chlorine as a function of the concentration of NaCl (0.005 mol L-1 to 0.25 mol L-1) was investigated using solutions at pH 4.0 and an applied potential of +1 V. Results are shown in Figure 8. It can be seen that higher active chlorine production is obtained by an increase in chloride ion concentration. Nevertheless, further increasing the chloride concentration above 0.1 mol L-1 does not promote any change in the rate of conversion, thereby showing a limiting rate in the oxidation of chloride ion. These results indicate that the area of the electrode at higher concentration limits the active chlorine production rate. In addition, because we operated an openair reactor, the solubility of chlorine in the solution may impose a limiting concentration as well. Effect of Competing Anions. Taking into account that sulfate and chloride are the most common anions present in natural water and wastewater, the influence of sulfate on the formation rate of active chlorine was studied. The amount of free chlorine produced during 30 min of electrolysis on a TiO2 thin-film electrode at Eapp ) +1 V under UV light was monitored for solutions containing, respectively, 0.025 mol L-1 NaCl; 0.025 mol L-1 NaCl + 0.050 mol L-1 Na2SO4; 0.025 mol L-1 NaCl + 0.100 mol L-1 Na2SO4; and 0.010 mol L-1 NaCl + 0.250 mol L-1 Na2SO4. The formation of chlorine was practically not affected by the addition of sulfate up to 0.10 mol L-1. However, the addition of sulfate at relative concentrations 12 times higher as shown in the last experiment promoted a suppression of almost 50% on the active chlorine generation. Reproducibility of Results. To obtain insight into the reproducibility of the experimental results, five experiments were carried out under the same conditions, and the active chlorine produced was compared after 30 min of electrolysis of 0.025 mol L-1 of NaCl in pH 4.0 at applied potential of +1 VOL. 38, NO. 11, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3207
V under UV illumination. Average values of 780 ( 25 mg L-1 of free chlorine with a standard deviation of 12.50% between measurements were obtained thus indicating a good reproducibility.
Conclusion Our results indicate that nanocrystalline TiO2 thin-film electrodes, which are irradiated with UV light simultaneously with a biased potential, may have practical suitability as electrode materials to be used in technological applications of chlorine production. The use of photoelectrocatalysis offers a simple and precise methodology to convert chloride ions into chlorine with large economical potentiality.
Acknowledgments Financial support from Brazilian funding agencies Capes, CNPq, and Fapesp are gratefully acknowledged.
Literature Cited (1) Burney, H. S. In Modern Aspects in Electrochemistry; White, R. E., Conway, B. E., Bockris, J. O. M., Eds.; Plenum Press: New York, 1993; p 393. (2) Rudolf, M. Rousar, I.; Krysa, J. J. Appl. Electrochem. 1995, 251, 155. (3) Rengarajan, V.; Sozhan, G.; Narasimham, K. C. Bull. Electrochem. 1996, 12, 327. (4) Kraft, A.; Stadelmann, M.; Blaschke, M.; Kreysig, D.; Sandt, B.; Schroder, F.; Rennan, J. J. Appl. Electrochem. 1999, 29, 861. (5) Kraft, A.; Blaschke, M.; Kreysig, D.; Sandt, B.; Schroder, F.; Rennan, J. J. Appl. Electrochem. 1990, 29, 895. (6) Patermarakis, G.; Fountoukidis, E. Water Res. 1990, 24, 14911496. (7) Schoberl, M. Eur. Pat. EP 0 515 628 B1 1991. (8) Kanekuni, N.; Shono, N.; Kyohara, M.; Tabata, K.; Kono, S.; Hayakawa, M. Eur. Pat. Appl. EP 0 711 730 AL, 1995. (9) Ibl, N.; Vogt, H. In Comprehensive Treatise of Electrochemistry; Bockris, J. O. M., Conway, B. E., Yeager, E., White, R. E., Eds.; Plenum Press: New York, 1981; Vol. 2, p 167. (10) Bennett, J. E. Chem. Eng. Prog. 1974, 70, 60. (11) Arikado, T.; Iwakura, C.; Tamura, H. Electrochim. Acta 1978, 23, 9. (12) Harrison, J. A.; Hermijanto, A. D. J. Electroanal. Chem. 1987, 225, 159.
3208
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 11, 2004
(13) Do, J. S.; Yeah, W. C. J. Appl. Electrochem. 1995, 251, 483. (14) Czarnetzki, L. R.; Janssen, L. J. J. J. Appl. Electrochem. 1992, 22, 315. (15) Do, J. S.; Yeah, W. C.; Chao, I. Y. Ind. Eng. Chem. Res. 1997, 36, 349. (16) Rudolf, M.; Rousar, I.; Krysa, J. J. Appl. Electrochem. 1995, 25, 155. (17) Trasatti, S. Electrochim. Acta 1984, 29, 1503. (18) Mozota, J.; Conway, B. E. J. Electrochem. Soc. 1991, 128, 2142. (19) Snoeyink, V. L.; Jenkins, D. In Water Chemistry; John Wiley & Sons: New York, 1980; p 386. (20) Stum, W.; Morgan, J. J. In Aquatic Chemistry, 3rd ed.; John Wiley & Sons: New York, 1996; Chapter 8, p 490. (21) Callaway, J. O. In Standard Methods for the Examination of Water and Wastewater, 17th ed.; Clesceri, L. S., Greenberg, A. E., Trussel, R. R., Eds.; Alpha Awwa-WPCF, Washington, DC, 1989; Part 4000, p 62. (22) Stoner, G. E.; Cahen, G. L.; Sachyam, M.; Gileadi, E. Bioelectrochem. Bioenerg. 1982, 9, 229. (23) Patermarakis, G.; Foundtoukidis, E. Water Res. 1990, 24, 1491. (24) Candal, R. J.; Zeltner, W. A.; Anderson, M. A. J. Adv. Oxid. Technol. 1998, 3, 3, 270. (25) Kim, D. H.; Anderson, M. A. J. Photochem. Photobiol. A: Chem. 1996, 94, 221. (26) Candal, R. J.; Zeltner, W. A.; Anderson, M. A. Environ. Sci. Technol. 2000, 34, 3443. (27) Kim, D. H.; Anderson, M. A. Environ. Sci. Technol. 1994, 28, 479. (28) Zanoni, M. V. B.; Sene, J. J.; Anderson, M. A. J. Photochem. Photobiol. A: Chem. 2003, 157, 55 (29) Xu, Q.; Anderson, M. A. J. Mater. Res. 1991, 6, 1073-1081. (30) Candal, R. J.; Zeltner, W. A.; Anderson, M. A. M. A. J. Adv. Oxid. Technol. 1998, 3, 270-276. (31) Candal, R. J.; Zeltner, W. A.; Anderson, M. A. J. Environ. Eng. 1999, 125, 906-912. (32) Nelson, B. P.; Candal, R.; Corn, R. M.; Anderson, M. Langmuir 2000, 16, 6094. (33) Finklea, H. O. In Semiconductor Electrodes; Elsevier: New York, 1988.
Received for review July 3, 2003. Revised manuscript received February 26, 2004. Accepted March 11, 2004. ES0347080
