Ti Dual Rotating Disk Photocatalytic (PC) Reactor: Dual

Dr. Nicholas Petraco at John Jay College, The City University of New York and Ms. Karrie Radloff at Lamont-Doherty Earth Observatory of Columbia Unive...
0 downloads 0 Views 313KB Size
Download PDF
Environ. Sci. Technol. 2009, 43, 6289–6294

Cu-TiO2/Ti Dual Rotating Disk Photocatalytic (PC) Reactor: Dual Electrode Degradation Facilitated by Spontaneous Electron Transfer Y U N L A N X U , † Y I H E , ‡ J I N P I N G J I A , * ,† DENGJIE ZHONG,§ AND YALIN WANG† School of Environmental Science and Engineering, Shanghai Jiao Tong University, Shanghai 200240, P. R. China, Department of Sciences, John Jay College and the Graduate Center, The City University of New York, New York, New York 10019, and Department of Environmental Science and Engineering, Tsinghua University, Beijing 100084, P. R. China

Received April 29, 2009. Revised manuscript received May 21, 2009. Accepted July 2, 2009.

A Cu-TiO2/Ti dual rotating disk photocatalytic (PC) reactor has been developed based on our single rotating disk photoelectrocatalytic (PEC) reactor (Y. Xu, et. al, Environ. Sci. Technol. 2008, 42, 2612-2617), and successfully applied to the treatment of laboratory and industrial dye wastewater. Round TiO2/Ti and Cu disks of the same size are connected by a Cu wire and fixed parallel on an axis continually rotating at 90 rpm. High treatment efficiency is obtained due to direct photooxidation on the TiO2/Ti photoanode as well as additional degradation on the Cu cathode, which is speculated via indirect hydrogen peroxide (H2O2) oxidation and direct electro-reduction of dye on cathode. The mechanism of the Cu-TiO2/Ti dual rotating disk PC reactor was investigated. In a 20 mg L-1 Rhodamine B (RB) solution, approximately 100 mV of potential and 10 µA of current were measured between the Cu and TiO2/Ti electrode during PC treatment. Such phenomenon was explained by spontaneous electron transfer based on the same principle of establishing a Schottky barrier. On the Cu electrode surface, the photoelectrons either reduced dye molecules directly or reacted with dissolved oxygen (DO) to form H2O2. Rotation of electrodes out of the solution enhanced the mass transfer of target compound and kept the aqueous film fresh. The Cu-TiO2/ Ti dual rotating disk PC reactor is a simple and effective device for the treatment of RB dye wastewater.

Introduction TiO2-based photocatalytic (PC) and photoelectrocatalytic (PEC) oxidation processes have drawn increasing attention in wastewater treatment because they use nontoxic and stable materials and offer high treatment efficiency (1-3). There are two key factors affecting the overall treatment efficiency: (a) the TiO2 photocatalyst and (b) light utilization efficiency. In order to obtain satisfactory oxidation results, researchers have made efforts to improve the performance of TiO2 photocatalyst and to extend its working range from UV to * Corresponding author phone: +86-21-54742817; fax: +86-2154742817; e-mail: [email protected]. † Shanghai Jiao Tong University. ‡ The City University of New York. § Tsinghua University. 10.1021/es901269s CCC: $40.75

Published on Web 07/15/2009

 2009 American Chemical Society

visible spectrum using doping technology (4-6). Further, in order to enhance light utilization efficiency, new PC reactors have been designed to decrease the path length of solution that light has to penetrate before it reaches the TiO2 photocatalyst surface, thus reducing the loss of irradiation due to solution absorption (7-10). A rotating TiO2/Ti disk PEC reactor was developed and successfully applied to degrade dye wastewater in our previous work (11). This reactor combined a highly effective thin film and a conventional PEC process on a single TiO2/Ti electrode. As it rotated, the upper exposed part of the round TiO2/Ti disk photoanode was coated with a thin film of wastewater and was irradiated with UV light. The lower part of the disk electrode was immersed in the wastewater solution to perform conventional treatment. The average thickness of the wastewater film coated on the electrode was about 75 µm at 90 rpm, which significantly increased the utilization of photon energy and improved treatment efficiency. The rotation of the TiO2/Ti disk electrode refreshed the thin aqueous film on the exposed side of the electrode and promoted mass transfer of target compounds in solution. Positive holes generated in the TiO2 semiconductor photocatalyst oxidized the organic pollutants. To prevent the rapid recombination of photogenerated electrons and holes, a bias voltage was applied to the electrode to drive the electrons through an external circuit. In conventional electrochemical wastewater treatment systems, anodic oxidation is the principal driving force for degradation of organic pollutants. However, a small amount of hydrogen peroxide (H2O2) is produced at the cathode by reduction of dissolved oxygen, as shown in eq 1 (12, 13), but it is transferred to the anode and destroyed (12). O2 + 2H+ + 2e- f H2O2

(1)

In order to prevent the transportation of H2O2 to an anode, efforts have been made to separate the anode and cathode using a membrane (14) or an electrolyte bridge (12). Higher treatment efficiency can thus be obtained through simultaneous direct oxidation at the anode and indirect H2O2 oxidation at the cathode, a process termed dual electrode oxidation (12). A Schottky barrier can be formed at a metal-semiconductor interface at which an initial net photoelectron transfer from the semiconductor into the metal is followed by establishment of an equilibrium distribution of holes and electrons (15). However, if the holes and electrons are continuously consumed, the equilibrium distribution would not be established. This would result in continuous electron transfer from semiconductor to metal. Although the principle of the Schottky barrier has been widely used in diode design (16), there is to our knowledge no application of this principle to seminconductor-based photocatalytic wastewater treatment. In this study, we developed a PC reactor using Cu and TiO2/Ti rotating disk electrodes to achieve direct oxidation at the TiO2/Ti photoanode and additional organic degradation by reactions occurred on Cu cathode, which we found including direct electro-reduction and indirect H2O2 oxidation. Photoelectrons are continuously transferred from TiO2 to Ti and then to the Cu electrode to satisfy the consumption of electrons at the cathode. Compared with our rotating TiO2/ Ti disk PEC reactor (11), the design of the new reactor has the following features: (a) the Cu cathode is enlarged to have the same surface area as the TiO2/Ti electrode; (b) both Cu and TiO2/Ti electrodes are fixed on the same axis and rotated together; (c) the Cu disk is connected to TiO2/Ti by a Cu wire; VOL. 43, NO. 16, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6289

FIGURE 1. (a) Schematic diagram of the side view of the Cu-TiO2/Ti dual rotating disks PC reactor. The figure is not to scale. (1) speed controller; (2) motor; (3) axis, composed with tow Cu tubes connected by a glass rod; (4) Cu disk; (5) TiO2/Ti disk; (6) carbon brush; (7) reaction cell; (8) wastewater; (9) UV lamp; (10) aluminum foil. The reactor was placed in a wooden box during operation. (b) The front view of the reaction cell and disk electrodes. The cell is filled with sample solution. (d) the thin aqueous film on the exposed Cu cathode isolates H2O2 generated there from migrating to the anode, making it equivalent to a dual cell. Absence of an external bias applied to the electrodes reduces the overall energy consumption and operating cost of the system. Wastewater produced by the textile dyeing industry is of great concern because of its adverse impact on the natural environment and public health (17). Simple, fast, economical and effective methods to treat dye wastewater are needed. In this study, Rhodamine B (RB) solution was employed as a model system to evaluate the proposed Cu-TiO2/Ti dual rotating disk PC reactor. The results were compared with those obtained using our previously studied TiO2/Ti single rotating disk PEC reactor (11). The feasibility of the system was demonstrated by the treatment of actual industrial dye wastewater samples.

Experimental Section Source of Textile Plant Effluent. Textile plant effluent was collected from a textile factory (Shanghai, China). They were untreated raw effluent from the disposal of the textile furnishing and finishing process. The effluent sample was stored at 4 °C and used without any treatment. The physicochemical properties of the dye wastewater sample were pH 11.3, conductivity 208 µS cm-1 and TOC 277 mg L-1. Materials and Reagents. Round titanium disk sheets (99.6% purity, diameter 75 mm, thickness 1.5 mm, and surface area 43.5 cm2) were purchased from Shanghai Hongtai Metal Production Co. Ltd. (Shanghai, China) and employed as the substrates for the TiO2 film coating. Copper disk sheets (>98% purity, diameter 75 mm, thickness 1.5 mm, and surface area 43.5 cm2) were purchased from Shanghai Runze Metal Material Co. Ltd. (Shanghai, China). Tetrabutyl titanate (Sinopharm Chemical Reagent Co. Ltd., China) was used as the precursor for preparing TiO2 colloidal suspensions. Rhodamine B (Shanghai Jiaying Chemical Co. Ltd., Shanghai, China) was of commercial grade and used as received. Na2SO4 (Shanghai Chemical Reagent Co. Ltd., Shanghai, China) was employed as the supporting electrolyte. All other chemicals were of reagent grade or better quality and used as received. All solutions were prepared in doubly distilled water. Preparation of the TiO2/ Ti Disk. The TiO2/ Ti disk was prepared by the sol-gel and dip-coating method. The detailed procedure for electrode preparation is found in our earlier paper (11). 6290

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 16, 2009

Cu-TiO2/Ti Dual Rotating Disk Photocatalytic Reactor. The Cu disk and the TiO2/Ti disk were mounted parallel on an axis and fitted into a semicircular quartz cell (85 mm diameter) (Figure 1a). The TiO2/Ti and Cu disk electrodes were connected by a Cu wire for the PC process (closed circuit in Figure 1a). To perform the PEC process, the electrodes were connected to an external power supply (LW5J5, Shanghai Liyou electric Co. Ltd., Shanghai, China) (open circuit in Figure 1a). With the reactor filled with sample solution, both TiO2/Ti and Cu disks had an area about 24.75 cm2 exposed to the air and 18.75 cm2 immersed in the solution (Figure 1b). Unless specified otherwise, disks were rotated at the optimum speed, 90 rpm (11), driven by a motor (Outai Transmission Electromechanical Co. Ltd., Shanghai, China). The reactor was placed about 3 cm away from an 11 W mercury lamp (Philips, 254 nm) with the TiO2/Ti disk facing the lamp. The mercury lamp was backed by a piece of aluminum foil placed behind the lamp and opposite to the reactor, so that light could be reflected and more fully used to irradiate the TiO2/Ti disk. The radiation power was maintained constant at 15 mW cm-2 (Spectra physics power meter, model 407A, U.S.) during all experiments. The cell and lamp were placed in a wooden box to avoid the influence of natural light and provide protection for the operator. The top of the box can be opened for sampling. Degradation Experiments. Cu-TiO2/Ti PC experiments were carried out to treat 55 mL of RB solution at concentration levels ranging from 20 to 150 mg L-1 each containing 0.5 g L-1 of Na2SO4, and 55 mL of textile plant effluent. The initial pH value of the RB solution was adjusted to pH 2.50 with 1 mol L-1 H2SO4 prior to PC treatment. (1 mol L-1 NaOH might be needed for fine-tune the pH.) This pH value was found to give the highest degradation efficiency in previous parameter optimization experiments. Industrial effluent was treated as received. Samples of aliquot were taken from the cell at desired intervals and filtered through a 0.45 µm Millipore filter prior to chemical analysis. In order to investigate the effect of electron transfer from anode to cathode, two sets of experiments were performed in 55 mL of 20 mg L-1 RB solution each containing 0.5 g L-1 of Na2SO4 as electrolyte. The experiments were carried out under UV light irradiation. They are (a) a PC process by directly connecting Cu disk and TiO2/Ti disk electrodes with Cu wire, i.e., Cu-TiO2/Ti PC. Here, photoelectrons are spontaneously transferred from semiconductor to metal as

occurs when a Schottky barrier is established; and (b) a PEC process by applying + 0.4 V bias between Cu disk and TiO2/ Ti disk electrodes, i.e., Cu-TiO2/Ti PEC, so that the photogenerated electrons are driven by the applied voltage to the cathode. To investigate the role of dissolved oxygen in the PC process, four sets of experiments were performed using 55 mL of 20 mg L-1 RB solutions containing 0.5 g L-1 Na2SO4 under different solution purging conditions. In the first set, PC was carried out in ambient air without any purging. For the other three sets, the RB solutions were continuously purged with air, pure oxygen, or pure nitrogen. Purging was carried out for 10 min at 50 mL/min prior to irradiation without electrodes in the solution; then continued at the same flow rate after electrodes were put into solution and experiment started. The entire setup was placed in a second box. However, in order to take samples of the solution every 10 min, the outside box needed to be removed. Thus the average atmosphere composition in the box was in fact O2or N2-rich rather than pure O2 or N2. Dual reaction cell was built to investigate the potential reactions occurred on Cu cathode. The setup was same as single cell (Figure 1) except that Cu disk and TiO2/Ti disk were separately immersed in two individual semicircular quartz cells (85 mm diameter, filled with 55 mL of 20 mg L-1 RB solutions containing 0.5 g L-1 Na2SO4). The two cells were connected by a saturated KNO3 salt bridge. The four sets of experiments mentioned above, but only the Cu cell was purged, were repeated again in this setup. All experiments were carried out at 26.3 ( 0.2 °C. Each experiment was performed in triplicate. Analysis. Dye concentrations were determined using a Unico UV-vis spectrophotometer (UV-2102 PCS, UNICO, Shanghai). The absorbance was measured at λmax of 563 nm for the RB solution and 514 nm for the textile plant effluent. Concentrated dye solutions were diluted with doubly distilled water prior to measurement to bring the absorbance into the linear range. The removal of color was defined as % removal ) (A0-A)/A0 × 100%, where A0 and A were initial and measured absorbance of the solution. TOC was measured using a TOC/TN analyzer (Jena 3000, Germany) to evaluate the extent of mineralization of the dyes, because it is possible for dyes to be degraded under PC conditions into intermediate organic compounds rather than complete transformation into CO2 and H2O. Dissolved oxygen (DO) was determined with a HQ series portable DO meter (HQ40d18, Hach Company, U.S.). The pH value was measured using a PHS3C pH meter (Shanghai Leici Apparatus Manufactory, Shanghai, China) and solution conductivity was obtained with a DDS-307 conductivity analyzer (Shanghai Leici Apparatus Manufactory, Shanghai, China). The concentration of H2O2 generated at the cathode was determined by titrating an acidified 5 mL of sample aliquot with 1 mM potassium permanganate solution (18). Distilled water, instead of dye solution, was used to carry out the dual disk PC experiment for determination of the H2O2 concentration. Current and voltage difference between the TiO2/Ti and Cu disks were measured using a multimeter (VICTOR, VC890D, Shanghai Tongqin Trading Co. Ltd., Shanghai, China). A cyclic voltammogram of a 20 mg L-1 RB solution was measured using an Autolab 4.9 (Metrohm, Switzerland) system with the Cu electrode as the working electrode, and the TiO2/Ti electrode as the counter and reference electrode.

TABLE 1. Dual Disk PC and Single Disk PEC Degradation of RB at Different Concentration Levels in 1 h Treatment Cu-TiO2/Ti dual rotating disk PC

TiO2/Ti rotating disk PEC

C0 (mg L-1)

% color removal

% TOC removal

% color removal

% TOC removal

20 30 50 80 100 150

100 89.8 87.9 87.7 78.6 74.6

64.6 59.2 53.4 43.8 40.9 32.5

89.5 80.4 76.1 60.6 47.3 32.1

60.5 51.3 39.6 30.2 14.6 7.3

to 150 mg L-1, even higher treatment efficiency is achieved without external power supply by using the dual disk reactor in the PC mode. As shown in Table 1, for a 1 h PC treatment, the removal of total color ranged from 100% at the lowest RB concentration to 74.6% at the highest concentration. Similarly, TOC was reduced by 64.6% at the lowest concentration and 32.5% at the highest. Table 1 gives the corresponding data for the single disk PEC mode obtained under the same conditions except for the +0.4 V bias potential applied for PEC. It clearly shows that superior results are obtained for the dual disk PC mode. The absolute quantity of RB removed, defined as initial absolute quantity of RB in solution multiplies % color removal, in 1 h by these two processes from solutions of different starting concentrations was also calculated. The results are shown in Figure 2. The amount of RB removed by dual disk PC increases steadily with increasing starting concentration. For the single disk PEC process, however, the amount of RB removed increased for RB concentrations up to 80 mg L-1 and reached a plateau, indicating the full treatment capacity of the TiO2/Ti photoelectrode was reached. We propose that the superior treatment efficiency obtained by dual disk system is due to dual disk reaction. In addition to photo-oxidation on TiO2/Ti anode, on the Cu electrode surface, the photoelectrons reacted either with dye to result in direct electro-reduction or with dissolved oxygen (DO) to form H2O2, which was isolated in the solution film on the surface of the electrode and further oxidized dye molecules. Consumption of photoelectrons on Cu electrode leads to electron transfer from TiO2/Ti anode to Cu cathode, and, therefore, effectively separates photogenerated holes and electrons, which is in favor of increase of the overall treatment efficiency.

Results and Discussion Dual Rotating Disk PC versus Single Rotating Disk PEC. The single TiO2/Ti rotating disk PEC process was shown to be more effective than the conventional PEC process in wastewater treatment (11). However, it is found here that for a series of RB solutions with concentrations ranging from 20

FIGURE 2. Absolute quantity of RB removed using TiO2/Ti PEC and Cu-TiO2/Ti PC procedures to treat RB solutions at different concentration levels. Treatment conditions: 0.5 g L-1 Na2SO4, initial pH 2.50, disk rotation speed 90 rpm, and treatment time 1 h. VOL. 43, NO. 16, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6291

TABLE 2. Concentration of Hydrogen Peroxide and Dissolved Oxygen during the Dual Disk PC Process and Single Disk PEC Process (H2O2: mmol L-1 and DO: mg L-1) under Different Atmospheric Conditionsa Cu-TiO2/Ti dual rotating disk PC atmosphere

air purge

TiO2/Ti rotating disk PEC

O2 purge

N2 purge

atmosphere

time (min)

DO

H2O2

DO

H2O2

DO

H2O2

DO

H2O2

DO

H2O2

10 20 30

8.02 8.13 8.23

0.250 0.300 0.350

8.12 8.16 8.26

0.250 0.312 0.362

10.24 10.26 10.27

0.262 0.312 0.362

1.21 1.16 1.12

0.075 0.062 0.050

8.14 8.18 8.19

0.088 0.112 0.125

a Saturated DO concentration is 8.22 mg L-1 at 26 °C (21). Titration of blank solution yielded KMnO4 consumption equivalent to 0.050 mmol L-1 H2O2. H2O2 data reported here has subtracted the background value.

The mechanism of the Cu-TiO2/Ti dual rotating disk PC reactor will be discussed based on the following four aspects: (a) photogenerated electrons transfer from semiconductor to metal; (b) the role of rotating Cu disk electrode; (c) the role of the dissolved oxygen; and (d) reactions at Cu cathode. Photogenerated Electron Transfer from Semiconductor to Metal. In order to investigate the effect of different pathways on photogenerated electrons transfer, two processes were applied to treat 20 mg L-1 RB solutions and the results were compared with those obtained by a single TiO2/ Ti rotating disk PEC (TiO2/Ti PEC) in our previous work (11). The two processes were (a) dual disks with +0.4 V bias potential (Cu-TiO2/Ti PEC), in which photogenerated electrons were driven through external voltage; and (b) dual disks without bias potential (Cu-TiO2/Ti PC), in which photogenerated electrons were transferred from semiconductor to metal based on the principle of establishing Schottky barrier. Approximately 100 mV of potential and 10 µA of current were measured between Cu and TiO2/Ti disks in this PC reactor. With 20 min treatment, the decolorization efficiency of aforementioned processes is 92.1, 90.9, and 34.3% for Cu-TiO2/Ti PEC, Cu-TiO2/Ti PC, and TiO2/Ti PEC, respectively. The similar decolorization efficiency obtained by CuTiO2/Ti PEC and Cu-TiO2/Ti PC suggested that net transfer of photogenerated electrons from semiconductor to metal has a similar effect as applying external bias. The Role of the Rotating Cu Electrode. The enlarged rotating Cu disk cathode plays an important role in improving treatment efficiency. As mentioned above, the dual-disk PC and PEC systems had similar decolorization efficiencies, both significantly higher than that obtained using the single disk TiO2/Ti PEC process. Although photogenerated electrons in both Cu-TiO2/Ti dual disk PEC and TiO2/Ti single disk PEC were driven to the Cu electrode by the +0.4 V external bias, the effects were different. Because the laboratory dye solution was acidic, the electrons on Cu cathode surface can react with dissolved O2 to form H2O2 (eq 1) (13, 19), or react with H+ to form H2 (eq 2): 2H+ + 2e- f H2

(2)

However, because the standard electrode potential of forming H2O2 (φ0 ) 0.68 V) is higher than that of H2 (φ0 ) 0 V) (20), the reaction of electrons with O2 is easier than that with H+ when dissolved oxygen is sufficient. In a solution reaction, electrons on Cu cathode surface only react with O2 in its vicinity. In Cu-TiO2/Ti PC process, DO concentration in the vicinity of Cu disk electrode was kept saturated (Table 2) and it was the same as that in the bulk solution because the Cu disk was continuously rotated to enhance mass transfer and keep the solution on the Cu surface fresh. In this case, eq 1 occurred and the resulting H2O2 further participated dye oxidation. Additionally, because the Cu disk was kept rotating, the thin aqueous film on the upper Cu electrode could isolate H2O2 generated on Cu cathode surface and prevent it from being consumed by the anode, which was confirmed by H2O2 6292

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 16, 2009

FIGURE 3. I-V response of Cu electrode. Conditions: Cu electrode is the working electrode and the TiO2/Ti electrode is the counter electrode and reference electrode. Distilled water containing 0.5 g L-1 Na2SO4 was used for the blank I-V curve and 20 mg L-1 RB containing 0.5 g L-1 Na2SO4 for the others. Scan rate: 50 mV/s. The I-V response of a RB free blank solution is listed in insert. concentration measured in different processes (Table 2). For single rotating disk PEC system, H2O2 may be formed at the beginning of the reaction; however, with the process of the reaction, DO concentration was gradually decreased on the static Cu electrode surface and resulted in eq 2 becoming a dominant reaction. The Role of Dissolved Oxygen. It was found DO positively affected treatment efficiency. For the ambient atmosphere, air-, O2- and N2-purging conditions, the respective decolorization efficiencies were 90.1, 91.5, 92.2, and 66.4%, and the respective TOC removals reached 27.1, 27.6, 30.9, and 13.4%)Supporting Information (SI) Figure S1). Table 2 shows that DO concentrations in solutions with ambient atmosphere and air purging were saturated. O2-rich environment allows additional uptake of O2, and DO was lower in N2-rich environment, as expected. Similar color and TOC removal efficiencies were obtained with the higher DO solutions, which points to the importance of O2 and H2O2 on treatment. Since there is little difference among the results with ambient air, air purging, and O2 purging, ambient air is experimentally the simplest and is the recommended condition for Cu-TiO2/ Ti PC operation. Reactions on Cu Cathode. In order to investigate the possible reactions on the Cu cathode, we first investigated I-V response at Cu electrode in different gas conditions. A reduction peak was observed at ∼-450 mV in a RB free blank solution (Figure 3 insert), indicating the formation of H2O2 on the electrode surface. In a 20 mg L-1 RB solution, a more obvious reduction reaction occurred on Cu electrode surface at approximately -250 and -360 mV under atmosphere and N2 purge, respectively, suggesting that RB also participated in electrode reaction. According to the results shown in Figure

FIGURE 4. Color removal of 20 mg L-1 RB treated by Cu-TiO2/Ti dual cell PC process with different gas condition. Treatment condition: 0.5 g L-1 Na2SO4, initial pH 2.50, disk rotation speed 90 rpm, and treatment time 1 h. 3 and Figure 4 (discussed in detail below), we proposed that indirect oxidation due to the formation of H2O2 and direct electro-reduction have occurred on the Cu electrode. Such results did not surprise us since dye degradation is usually very complicated and often involves multistep reactions. The electrode reactions were summarized by following equations. (a) for H2O2 indirect oxidation nH2O2 + RB f nH2O + RB′

(3)

The overall O2-participating electrode reaction is expressed by eq 4 through combination of eq 1 and 3. nO2 + 2nH+ + RB + 2ne- f nH2O + RB′

(4)

Where RB′ is the oxidation product. (b) for electro-reduction reaction RB + me- f RB′′

(5)

Where RB′′ is the reduction product. The peak shift observed in Figure 3 under atmosphere and N2 purging condition was explained by Nernst equation written based on O2 participating eq 4 and reaction on TiO2/ Ti disk surface (20):

[

0 E ) φ+ - φ- ) φ+

RT n · aRB′ / · ln(aH 2O 2nF

]

2n n aH (6) + · aO · aRB) - φ2

Where φ+ and φ- are potential for positive (reaction on Cu disk surface, eq 4) and negative electrode (reaction on TiO2/ 0 is the standard electrode Ti disk surface), respectively; φ+ potential for positive electrode; and aH2O, aH+ and aO2 are the activity of H2O, H+ and DO; aRB and aRB′ are the activity of RB and its intermediate products. Under both atmosphere and N2 conditions, aH+is the same because the initial pH was adjusted to the same value; aH2O can be regarded as the same since H2O is greatly excessive; the difference of aRB and aRB′ can be ignored because the scan time was very short, and the consumption amount of RB and the generation amount of RB′ was very small. In accordance with Henry’s law (20), aO2decreased with the decrease of oxygen partial pressure, therefore, aO2 under N2 purge is lower than that under atmosphere, which led to a more negative reduction peak under N2 purge condition. The generation of H2O2 was further investigated by measuring the concentration of H2O2 and the results are listed in Table 2. When operated in atmosphere condition, DO was saturated in both Cu-TiO2/Ti dual-disk PC and single-disk

TiO2/Ti PEC process; however, H2O2 concentration in dualdisk reactor was found to be higher than that in single-disk system. Such results suggested dual-disk reactor was in favor of eq 1 because a saturated DO concentration could be maintained on the Cu rotating disk in that the solution on the surface of Cu electrode was continuously refreshed. On the contrary, the DO concentration on the surface of the small static Cu electrode decreased quickly. Dual cell experiments were also carried out under different gas conditions to investigate the RB involving electroreduction. Figure 4 shows that the respective decolorization efficiencies of Cu reaction cell were 59.3, 61.1, 64.5, and 41.5% in the ambient atmosphere, air-, O2-, and N2-purging conditions. Since the result of N2- purging approximately represents the direct RB electro-reduction, and that of O2purging is the overall efficiency of direct reduction and indirect H2O2 oxidization, the effect of H2O2 oxidation can be estimated through the difference of the data obtained between N2-rich and other conditions. We can see that although electro-reduction is very important for RB decolorization on Cu cathode, accounting for about two-thirds of the treatment efficiency, indirect H2O2 oxidation in no doubt contributes significantly as well. Application of Dual Disk PC to Treat Textile Plant Effluent. The dual rotating disk PC reactor was applied to treat the textile plant effluent sample to investigate its feasibility in practical dye wastewater treatment. The absorption spectra of effluent were obtained in the range of 250-700 nm (SI Figure S2a). The absorbance at 514 nm rapidly decreased in 135 min. The color and TOC removal efficiency reached 90 and 49% (SI Figure S2b), respectively, in 135 min, which was much higher than that obtained by single rotating disk PEC (70.7% color removal and 38.1% TOC removal). In addition to RB, we also used this system to treat other dyes such as Acid Black ATT, Allura Red, Reactive Brilliant Red X-3B, and Direct Copper Blue 2R. Very promising treatment results were obtained in the preliminary experiments as well. Such results demonstrated that the dual rotating disk PC reactor is an efficient method for the treatment of dye wastewater.

Acknowledgments Financial support from the Natural Science Foundation of China (Project No. 20477026 and 50878126) is gratefully acknowledged. We thank Dr. David C. Locke of Queens College, The City University of New York, for helpful discussions and assistance with the manuscript. Dr. Nicholas Petraco at John Jay College, The City University of New York and Ms. Karrie Radloff at Lamont-Doherty Earth Observatory of Columbia University are also thanked for their valuable inputs.

Supporting Information Available

Figures for color and TOC removal of 20 mg L-1 RB treated by Cu-TiO2/Ti PC process under different gas conditions, UV-vis spectral changes and decolourization based on 514 nm and TOC depletion of textile plant effluent treated by Cu-TiO2/Ti PC process as a function of treatment time. This material is available free of charge via the Internet at http:// pubs.acs.org.

Literature Cited (1) Li, X. Z.; Liu, H. L.; Yue, P. T.; Sun, Y. P. Photoelectrocatalytic oxidation of Rose Bengal in aqueous solution using a Ti/TiO2 mesh electrode. Environ. Sci. Technol. 2000, 34, 4401–4406. (2) Yang, J.; Chen, C. C.; Ji, H. W.; Ma, W. H.; Zhao, J. C. Mechanism of TiO2-assisted photocatalytic degradation of dyes under visible irradiation: photoelectrocatalytic study by TiO2-film electrodes. J. Phys. Chem. B 2005, 109, 21900–21907. VOL. 43, NO. 16, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6293

(3) Zanoni, M. V. B.; Sene, J. J.; Selcuk, H.; Anderson, M. A. Photoelectrocatalytic production of active chlorine on nanocrystalline titanium dioxide thin-film electrodes. Environ. Sci. Technol. 2004, 38, 3203–3208. (4) Wawrzyniak, B.; Morawski, A. W. Solar-light-induced photocatalytic decomposition of two azo dyes on new TiO2 photocatalyst containing nitrogen. Appl. Catal., B 2006, 62, 150–158. (5) Egerton, T. A.; Janus, M.; Morawski, A. W. New TiO2/C sol-gel electrodes for photoelectrocatalytic degradation of sodium oxalate. Chemosphere 2006, 63, 1203–1208. (6) Nagaveni, K.; Hegde, M. S.; Ravishankar, N. Synthesis and structure of nanocrystalline TiO2 with lower band gap showing high photocatalytic activity. Langmuir 2004, 20, 2900–2907. (7) Dionysiou, D. D.; Balasubramanian, G.; Suidan, M. T.; Khodadoust, A. P.; Baudin, I.; Lan ˆ e´, J.-M. Rotating disk photocatalytic reactor: development, characterization, and evaluation for the destruction of organic pollutants in water. Water Res. 2000, 34, 2927–2940. (8) Noorjahan, M.; Pratap Reddy, M.; Durga Kumari, V.; Lave´drine, B.; Boule, P.; Subrahmanyam, M. Photocatalytic degradation of H-acid over a novel TiO2 thin film fixed bed reactor and in aqueous suspensions. J. Photochem. Photobiol. A 2003, 156, 179–187. (9) Chan, A. H. C.; Chan, C. K.; Barford, J. P.; Porter, J. F. Solar photocatalytic thin film cascade reactor for treatment of benzoic acid containing wastewater. Water Res. 2003, 37, 1125–1135. (10) Shankar, M. V.; Anandan, S.; Venkatachalam, N.; Arabindoo, B.; Murugesan, V. Novel thin-film reactor for photocatalytic degradation of pesticides in an aqueous solution. J. Chem. Technol. Biotechnol. 2004, 79, 1279–1285. (11) Xu, Y. L.; He, Y.; Cao, X. D.; Zhong, D. J.; Jia, J. P. TiO2/Ti rotating disk photoelectrocatalytic (PEC) reactor: a combination of highly effective thin film PEC and conventional PEC process on a single electrode. Environ. Sci. Technol. 2008, 42, 2612–2617.

6294

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 16, 2009

(12) Shen, Z. M.; Yang, J.; Hu, X. F.; Lei, Y. M.; Ji, X. L.; Jia, J. P.; Wang, W. H. Dual electrodes oxidation of dye wastewater with gas diffusion cathode. Environ. Sci. Technol. 2005, 39, 1819–1826. (13) Li, X. Z.; Liu, H. S. Development of an E-H2O2/TiO2 photoelectrocatalytic oxidation system for water and wastewater treatment. Environ. Sci. Technol. 2005, 39, 4614–4620. (14) Kusvuran, E.; Gulnaz, O.; Irmak, S. Comparison of several advanced oxidation processes for the decolourization of Reactive Red 120 azo dye in aqueous solution. J. Hazard. Mater. 2004, 109, 85–93. (15) Meng, J.; Liu, H. B.; Meng, Q. H. Semiconductor-Device Physics; Science Publishing Company: Beijing, China, 2005. (16) Sedra, A. S.; Smith, K. C. Microelectronic Circuits; Oxford Univeristy Press: New York, 1991. (17) Uygur, A.; Ko¨k, E. Decolorisation treatments of azo dye waste waters including dichlorotriazinyl reactive groups by using advanced oxidation method. J. Soc. Dyers Colour 1999, 115, 350–354. (18) Oloman, C.; Watkinson, A. P. Hydrogen peroxide production in trickle-bed electrochemical reactors. J. Appl. Electrochem. 1979, 9, 117–123. (19) Leng, W. H.; Zhu, W. C.; Ni, J.; Zhang, Z.; Zhang, J. Q.; Cao, C. N. Photoelectrocatalytic destruction of organics using TiO2 as photoanode with simultaneous production of H2O2 at the cathode. Appl. Catal., A 2006, 300, 24–35. (20) Fu, X. C.; Shen, W. X.; Yao, T. Y.; Hou W. H. Physical Chemistry; Higher Education Publishing Company: Beijing, China, 2006. (21) Peavy, H. S.; Rowe, D. R.; George, T. Environmental Engineering; McGraw-Hill Book Company: New York, 1985.

ES901269S