Environ. Sci. Technol. 2005, 39, 2848-2855
Laboratory Studies of Electrochemical Treatment of Industrial Azo Dye Effluent SANJAY S. VAGHELA, ASHOK D. JETHVA, BHAVESH B. MEHTA, SUNIL P. DAVE, SUBBARAYAPPA ADIMURTHY, AND GADDE RAMACHANDRAIAH* Central Salt and Marine Chemicals Research Institute, Gijubhai Badheka Marg, Bhavnagar 364 002, Gujarat, India
Removal of color and reduction of chemical oxygen demand (COD) in an industrial azo dye effluent containing chiefly reactive dyes were investigated under singlepass conditions at a dimensionally stable anode (DSA) in a thin electrochemical flow reactor at different current densities, flow rates, and dilutions. With 50% diluted effluent, decolorization was achieved up to 85-99% at 10-40 mA/ cm2 at 5 mL/min flow rate and 50-88% at 30-40 mA/ cm2 at high (10-15 mL/min) flow rates. The COD reduction was maximum (81%) at 39.9 mA/cm2 or above when solutionelectrode contact time (Ct) was as high as 21.7 s/cm2 and decreased as Ct declined at a given current density. Cyclic voltammetric studies suggesting an indirect oxidation of dye molecules over the anode surface were carried out at a glassy carbon electrode. The effect of pH on decolorization and COD reduction was determined. An electrochemical mechanism mediated by OCl- operating in the decolorization and COD reduction processes was suggested. The effluent was further treated with NaOCl. The oxidized products from the treated effluents were isolated and confirmed to be free from chlorinesubstituted products by IR spectroscopy. From the apparent pseudo-first-order rate data, the second-order rate coefficients were evaluated to be 2.9 M-1 s-1 at 5 mL/ min, 76.2 M-1 s-1 at 10 mL/min, and 156.1 M-1 s-1 at 15 mL/ min for color removal, and 1.19 M-1 s-1 at 5 mL/min, 1.79 M-1 s-1 at 10 mL/min, and 3.57 M-1 s-1 at 15 mL/min for COD reduction. Field studies were also carried out with a pilot-scale cell at the source of effluent generation of different plants corresponding to the industry. Decolorization was achieved to about 94-99% with azo dye effluents at 0.7-1.0 L/min flow costing around Indian Rupees 0.020.04 per liter, and to about 54-75% in other related effluents at 0.3-1.0 L/min flow under single-pass conditions.
Introduction Dyes are organic colorants used in textile, pharmaceutical, cosmetic, food, and other industries for imparting different shades of colors (1). Dye manufacturers and users, particularly the textile industries, release wastewaters in massive quantities containing dye to the extent of 0.001-0.7% (w/v), often with dissolved inorganic salts, dispersing agents, surfactants, * Corresponding author phone: +91-278-2568694, 2568114, or 2567760; fax: +91-278-2567562; e-mail:
[email protected] or
[email protected]. 2848
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 8, 2005
and organics washed out from the materials (2-4). Reports also suggest that dyehouse effluents contain 0.1-2% (w/v) dye resulting in 2-9% of total global annual dye production, which is equivalent to nearly 50 000 tons (5-11). This situation will be further aggravated in the ensuing years as the demand for cotton and other fabrics increases exponentially (9). The pH values of these discharges vary between 2 and 12, depending upon the dye and its application (12-14). The deleterious aspects of dye effluents are their unacceptable color and high chemical oxygen demand (COD) content. Further, the dye components are hardly degradable by physicochemical or biological methods and degradation becomes highly difficult as the textile dyes are constantly being replaced with modern dyes, which are resistant to chemical, photochemical, and biological degradation (2, 9, 15). Once released, these untreated discharges may deteriorate soil, water resources, and the environment (16-18). Conventional biochemical oxidative treatment of wastewaters containing azo dye often results in colored water unfit for reuse, while other methods such as coagulation, absorption, chemical oxidation, reverse osmosis, ultrafiltration, photo decomposition, etc., alone or in combination, are found to be ineffectual due to their cost, regeneration or reusability, and secondary pollution (10-29). Chemicals such as hypochlorite, ozone, and hydrogen peroxide, in the absence and in the presence of UV light and hydrogen peroxide with ferrous ions, have been used for pretreatment of dye-bearing wastewater (30-36). The electrochemical process is yet another rewarding competitive method. However, only a few studies have been attempted (37-46). Operational difficulties associated with the implementation of electrochemical methods by entrepreneurs in industry involve a huge investment in equipment and power. These methods become user-friendly provided the process limits objectionable color while leaving the COD associated with organics left in the effluent for natural degradation. In the present paper, as a sample case, the electrochemical decolorization and COD reduction of an orange-red effluent containing a mixture of reactive azo (0.4-0.5%) dyes (Red 2, CI 18200; Red 20, CI 25810; Reactive Orange 4, CI 18260; and Reactive Orange 13, CI 18270) corresponding to one of the local leading textile industries, has been studied under singlepass flow conditions employing a dimensionally stable catalytic anode vs a stainless steel cathode mounted in a thin flow cell. The percent of decolorization based on the absorption spectral data and the percent of COD removal by analytical methods have been studied at various possible current densities and flow rates. The observed rates for decolorization and COD reduction have been worked out and the second-order rate constants with respect to applied current density per unit area have been reported. A commensurate mechanism for the indirect oxidative decomposition of the dye has been discussed. The decomposed products have been studied by 1H NMR, FTIR, and elemental analysis. The results of cyclic voltammetry and chemical treatment of the effluent with hypochlorite and hydrogen peroxide are also presented.
Materials and Methods Dye Effluent. The deep orange-red effluent was obtained from M/s Atul Industries Limited, Atul, a leading textile industry situated in Gujarat, India. It chiefly contained a mixture of reactive azo dyes viz. Red 2 (CI 18200), Red 20 (CI 25810), Reacive Orange 4 (CI 18260), and Reactive Orange 13 10.1021/es035370c CCC: $30.25
2005 American Chemical Society Published on Web 03/01/2005
FIGURE 1. View of the flow reactor. (CI 18270) (42). It possessed about 0.74 g of combustible and 16.88 g of noncombustible dissolved solids in 100 mL, while the absorption spectra showed a broad group of bands between 550 and 600 nm centering at 576 nm, characteristic of the dyes. The pH, TDS, chemical oxygen demand (COD), Hazen value, and NaCl composition of this effluent were 7.3, 17.6% (w/v), 5957 ppm, 75 000, and 16.7% (w/v), respectively. The effluent was diluted to 25% and 50% with deionized water and used for further studies. The pH, TDS, chemical oxygen demand (COD), and electrical conductivity of the 50% diluted effluent as determined by dichrometric method (39) were 8.1, 1.64% (w/v), 2978 ppm, and 135 mmhos/cm, respectively. Instrumentation. Decolorization and COD reduction in the dye effluent were carried out at a dimensionally stable catalytic anode (DSA) (43) against a stainless steel cathode in an undivided flow cell under galvanostatic conditions. An Aplab (India) model L 1288 DC-regulated power supply was employed as a constant current source. The absorption spectra were recorded on a Shimadzu UV-160 spectrophotometer coupled to a temperature-controlling unit employing 1-cm quartz cuvettes. The solution pH was recorded using an Adair Dutt digital pH meter, sensitive to 0.01 pH units. 1H NMR spectra in D O on a Bruker Avance DPX-200-FT 2 NMR-200 MHz spectrometer, IR spectra on a Perkin-Elmer Spectrum (GX FT-IR system) and CHN analysis on a PerkinElmer CHNS/O 2400 were obtained for the dried residues of the treated and untreated effluents. Cyclic voltammetric responses were recorded with an EG&G PAR model 273 A Potentiostat/Galvanostat coupled with a three-electrode-cell assembly and a Gateway 2000 (4DX2-66) computer. A glassy carbon electrode (0.0314 cm2) and its potentials were measured with reference to Ag/AgCl (0.222 V vs NHE) in 3 M NaCl. The platinum wire separated from the analytical solution by a Vycor tip bridge served as a counter electrode (44). Flow Reactor Experiments. A rectangular (18 cm × 15 cm × 5 cm) PVC cell (Figure 1), consisting of an expanded DSA (10.5 cm × 6 cm) with an effective area of 50.16 cm2 as anode, a thin stainless steel plate (11 cm × 8 cm) as cathode, one inlet and one outlet, served as a flow reactor with 90-mL internal volume (43). The dye effluent was allowed to flow from the bottom of the cell at a regulated rate. A current between 0 and 100 mA/cm2 was applied across the two current-carrying electrodes while the effluent passed steadily through the cell such that the average solution-electrode contact time (Ct) per unit area varied between 5 and 22 s/cm2. The cell potential across the cathode and the anode and the pH of the solution at the outlet were monitored. The effluent after the electrochemical treatment was collected and analyzed for percent decolorization from the absorption data at 576 nm, and the percent COD removal from the COD data were calculated after comparing with that of the untreated
FIGURE 2. Absorption spectra of (A) original effluent after treatment at (a) 0, (b) 39.9, (c) 59.8, (d) 79.8, and (e) 99.7 mA/cm2; and (B) diluted effluent after treatment at 39.9 mA/cm2 with (a) 0%, (b) 25%, and (c) 50% dilution. Flow rate in both cases was 5 mL/min.
TABLE 1. Decolorization and COD Reduction in 0, 25, and 50% Diluted Effluent at Different Applied Current Densities and 5 mL/min Flow mA/cm2
% dilution
% decolorization
% COD reduced
99.7 79.7 59.8 39.9
0 0 0 0 25 50
98.8 98.0 92.7 87.1 94.6 99.2
52.8 42.5 42.8 40.1 63.1 80.8
effluent employed. The chemical oxygen demand of the treated and untreated effluents was evaluated by the standard method involving potassium dichromate (39).
Experimental Results Original Effluent. The effluent as such was passed through the flow reactor (Figure 1) at a current density in the range between 0 and 100 mA/cm2 while fixing the flow rate at 5 mL/min. The electrolysis between 0 and 30 mA/cm2 was less effective, as the color and the COD value of the effluent changed negligibly after the treatment. However, at higher current densities, the color and COD of the treated effluent changed remarkably despite the copious evolution of chlorine by the side reaction. Figure 2 shows the spectral changes in the effluent before and after the treatment at different current densities. The percent decolorization (eq 1) and percent COD reduction (eq 2) are presented in Table 1.
% decolorization )
% COD reduction )
[(OD)B - (OD)A] × 100
(1)
(OD)B
[(COD)B - (COD)A] × 100 (COD)B
(2)
where (OD)B and (OD)A are the absorbance at 576 nm, and (COD)B and (COD)A are the chemical oxygen demand of the pre- and post-treated effluent, respectively. The effluent was nearly colorless at high current densities exceeding 60 mA/cm2 with 42-53% reduction in COD (Figure 1 and Table 1). To determine the electrochemical reactions instrumental for these changes, the dye effluent was also electrolyzed at 59.8 mA/cm2 in a thin two-compartmental membrane (cation) flow reactor (43). The effluent was passed through the anode or cathode compartment while deionized water was circulated through the other. The flow of the counterion across the membrane from higher concentration VOL. 39, NO. 8, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2849
TABLE 2. Color Removal and COD Reduction in 50% Diluted Effluent (COD 2978 ppm) as a Function of Current Density and Flow Rate or Solution-Electrode Contact Time, Cta % decolorization
% COD reduction
mA/cm2
5 mL/min
10 mL/min
15 mL/min
21.5 s/cm2
10.8 s/cm2
7.2 s/cm2
10. 19.9 29.9 39.9
85.3 98.2 99.1 99.2
3.8 34.8 65.4 88.0
3.1 18.5 50.2 78.5
26.1 39.7 53.0 80.8
12.7 16.7 24.8 32.2
1.3 13.4 14.0 20.8
a Cell voltage was 3-4 V at 9-41 mA/cm2. In all cases, the pH of the original solution was 8.10 while that of the electrolyzed solution varied between 6 and 7.60 depending on the current density applied.
to the lower one supports the reaction at the electrode. The effluent was discolored while the COD reduced when it was passed through the anode compartment as effectively as in the case of a single compartment cell. There was no appreciable change in color intensity and COD value when it was passed through the cathode compartment accounting for the oxidative decomposition of dye molecules at the anode either by direct or indirect pathway. Diluted Effluents. Figure 1B compares the changes in spectra of the treated 0, 25, and 50% diluted effluents at 39.9 mA/cm2. The absorbance at 576 nm in the case of 50% dilute solution was tending to zero. The decolorization (Table 1) was increased nominally by 7.5 and 12.1% at 5 mL/min flow, while the COD reduction was affected to an extent of 23% in the case of 25% diluted effluent and 40.7% in the case of 50% diluted effluent. The liberation of elemental chlorine was less in the case of 25% and negligible in the case of 50% diluted effluent. Further evaluation studies were conducted with 50% diluted effluent to understand more the relative effects of applied current density, flow rate, pH, and electrical conductance on the decolorization and COD reduction and shed light on decomposition rates and its mechanism. Effect of Current Density and Flow Rate. The 50% diluted effluent was passed through the electrochemical flow reactor at three different flow rates viz. 5, 10, and 15 mL/min to get an average solution residence time (Rt) in the cell between 20 and 4 min (eq 3). The electrolysis was carried out under single-pass conditions by varying the current densities between 9 and 40 mA/cm2 at a given flow rate. During the experiments, the potential drop at the two current-carrying electrodes was steady and it varied between 3 and 4 V when the magnitude of the current changed. The pH was decreased from 8.1 to 6.0-7.6 depending on the current density applied and the flow rate employed.
Rt )
cell capacity flow rate
(3)
The variance between the absorption spectra of the original and the electrolyzed effluent flowing at 5, 10, and 15 mL/min at 10, 19.9, 29.9, and 39.9 mA/cm2 was studied. At 15 mL/min, the intensity of the 576-nm band gradually faded with the increase in current density. Accordingly, the color of the solution gradually diminished. Unlike at 15 mL/min, the color intensity of the effluent as well as the absorbance at 576 nm remarkably decreased at 10 mA/cm2 and diminished further when the current density was increased at 5 mL/min. Similar results were observed at 10 mL/min flow. Further, the intensity of the 576 nm band declined as the flow rate decreased from 15 to 5 mL/min at a given current density. The decolorization data measured at four different current densities and at three different flow rates are presented in Table 2. The decolorization was maximum (99.2%) at 39.9 mA/cm2 at 5 mL/min flow and decreased by about 14% when current density was reduced to 10 mA/cm2 while the flow rate was maintained constant, or by about 21% when the 2850
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 8, 2005
flow rate was increased to 15 mL/min while the current density was kept constant. The decolorization at 29.9 mA/ cm2 was 99.1% at 5 mL/min, reduced correspondingly by about 34% and 49%, and at 19.9 mA/cm2 was 98.2%, reduced by about 63% and 80% when the flow rate was increased to 10 and 15 mL/min, respectively. Wherein, the decolorization at 10 mA/cm2, which was 85.3% at 5 mL/min, decreased abruptly by about 81-82% when the flow rate was increased to 10 or 15 mL/min. On the other hand, the COD reduction (Table 2) was maximum (81% at 39.9 mA/cm2 ) when the contact time (Ct) was 21.5 s/cm2 (eq 4). It gradually decelerated to 26% as the current density decreased to 10 mA/cm2 and further plummeted to 21% with the increase in Ct. The COD reduction was maximum (81%) at 39.9 mA/cm2 at Ct ) 21.5 s/cm2 as compared to other values at other current densities.
Ct )
residence time anode area
(4)
The electrical conductance of the effluent after treatment at the above said four current densities and three flow rates was measured constant between 87 and 90 mmhos/cm. The negligible change in the solution conductance during the electrolytic treatment may partly account for the intact ionic chloride concentration in the effluent and for the absence of chlorine involved substitution reactions of dye molecule. Influence of pH. The pH of effluent greatly influences the treatment process as it effects the protonation/deprotonation of some of the basic sites present in the dye or the formation and stability of active intermediates responsible for the decomposition. To investigate these effects, the pH of the above said effluent was adjusted to pH 2.8 and 11.6 from 8.1 employing minimum quantities of concentrated hydrochloric acid or sodium hydroxide solutions and then treated in the flow reactor. At 39.9 mA/cm2 and 5 mL/min flow, the color removal was 99% at pH 2.8 which is identical to that at pH 8.1, but it was only 45% at pH 11.6. The inhibition in the desired reaction at higher pH > 11 may be attributed to the stabilization of the species responsible for the decomposition of dye molecules. In contrast, the COD reduction steeply decreased at both the pH values: 50.7 at pH 2.8 and 27.4 at pH 11.6, compared to 80.8 at pH 8.1. This is obvious when a dye decomposes through electro-generated intermediates. Chemical Treatment. Oxidative decomposition of dyes leading to decolorization of effluents may be carried out with inexpensive and easily available oxidants. These chemical oxidants vary in their chemical potential and some of them may need a catalyst or co-reactant. Sodium hypochlorite and hydrogen peroxide are some such commonly available oxidants. In this study, 50 mL of the effluent was reacted variously with 5, 15, and 20 mL of 4% NaOCl solution or 10 mL of the effluent was mixed with H2O2 (1 mL of 30%). In the former case, the color of the solutions faded slowly with time. After 48 h of reaction, the decolorization was found to be 90, 98, and 99%, respectively. Concomitantly, the COD reduction was 65, 85, and 90%, respectively, which is in good
FIGURE 3. FT-IR spectra of the extracted samples of (a) untreated effluent, (b) electrochemically treated effluent, and (c) chemically treated effluent with 4% NaOCl. agreement with the results in Table 2. Little change in either color or COD was observed even after a week in the second case. Similar results were obtained when HCl (1 mL of 12 N) was added as a co-reactant with H2O2. Analysis of Effluent Residues. Treatment of dye effluents with strong oxidants such as hypochlorite, or electrochemically at anodes in the presence of large concentrations of chloride, may give rise to doubts of generating more harmful chlorine substitutes which later on cause secondary pollution. The products of the process call for suitable and strong analytical methods. The IR spectra of the solid residues isolated from the effluent before and after the electrochemical and chemical treatments are shown in Figure 3a-c. Considerable changes were observed in the regions 400-700 cm-1 and 1000-1700 cm-1. Figure 3a shows two bands at 619 and 676 cm-1 probably due to C-Cl stretch present in all the dye molecules. The former band appeared at 622 cm-1 while the later one tends to disappear in Figure 3b and c. Similarly, the band which appeared at 1197 cm-1 in Figure 3a appears as a broad and relatively low intensive peak in Figure 3b and c, indicating that the corresponding group is the potential site of oxidation. However, the bands at 1400, 1429, and 1465 cm-1 seen in Figure 3a were absent in Figure 3b and concomitantly a new sharp peak appeared at 1384 cm-1. In contrast, the bands at 1400, 1429, and 1465 cm-1 were observed as two intense peaks at 1430 and 1458 cm-1 in the hypochlorite treated sample, Figure 3c, indicating that the corresponding functional groups are sensitive to oxidation. However, an intense band at 1384 cm-1 was observed in the electrochemically treated sample, replacing the three bands at 1400, 1429, and 1465 cm-1. Besides, an additional peak at 970 cm-1 that appeared in the chemically treated sample (Figure 3c) may be assigned to the ClO3- ion which is generally present in the hypochlorite solution. The peak identical to 970 cm-1 in Figure 3c was absent in 3b which suggests that ClO3- did not form as the side product during the progress of electrolysis. Since the band intensities in the region 505-760 cm-1 were not intensified due to C-Cl stretch (35) in Figure 3b abd c, the possibility of formation of chlorine-substituted byproducts, which may lead to secondary pollution, was ruled out in both electrochemical and chemical treatment of the effluent. Cyclic Voltammetry. Electroactive species reacting at the electrode give peak-type responses as a mark of electron exchange during anodic- and cathodic-potential scans. In the present case, the dye decomposition may occur directly or indirectly or both. Figure 4a-c shows the cyclic voltammetry responses of 0.1 M NaCl before and after the addition of 0.2 and 0.4 mL of dye effluent at a glassy carbon electrode
FIGURE 4. Cyclic voltammetric responses of 0.1 M NaCl with (a) 0.0, (b) 0.2, and (c) 0.4 mL of 50% diluted effluent. Scan rate in all cases was 100 mV s-1. (0.0314 cm2), Ag/AgCl (3 M NaCl), in the potential range of 1.5 to -1.5 V. No major difference between the electrode response before (Figure 4a) and after (Figure 4b and c) the introduction of the effluent into the cell was noted owing to the electro-inactive nature of the dyes in the potential range investigated here. However, at anodic potentials between 1.0 and 1.5 V, a considerable decrease in the anodic currents was observed and it further decreased with the increase in dye concentration denoting the adsorption of dye molecules over the anode surface prior to its decomposition. The large anodic currents at and above 1.0 V may be due to the displacement of chlorine and oxygen from the medium with or without the simultaneous decomposition of dyes. Calculation of Kinetic Constants. The rates of color removal and COD reduction in the effluent depend on the dye concentration [dye] and the oxidizing strength of the anode, i.e., the concentration of active species (discussed in the latter part of this paper) denoted as [O] by which the color and COD in the effluent are reduced. Thus, the kinetics of the electrochemical effluent treatment at the anode are expressed as
-d[color]/dt ) k′[dye][O]
(5)
-d[COD]/dt ) k′′[dye][O]
(6)
where k′and k′′ are the second-order rate constants for the color removal and the COD reduction reactions at the anode, respectively. During electrolysis, the value of [O] in eqs 5 and 6 will remain constant under a given set of experimental conditions, but it varies as the applied current density is altered. Presumably, under stationary conditions there is no loss in the oxidizing strength of the anode or [O], the rate of active species [O] is equal to the rate of its consumption in both of the electrode processes and at a given current density, the eqs 5 and 6 are modified as
-d[color]/dt ) k′obs[dye]
(7)
-d[COD]/dt ) k′′obs[dye]
(8)
where k′obs and k′′obs are the apparent pseudo-first-order kinetic constants for the electrochemical processes of color removal and COD reduction, respectively, at anode under a given set of experimental conditions. The values of k′obs and k′′obs for the treatment of 50% diluted effluent at different current densities and flow rates were calculated by substituting the data in Table 2 in eqs 9 and 10, respectively, and the results are summarized in Table 3. The k′obsvalue at a given flow rate increased linearly with the hike in the current density, except at 5 mL/min where it was VOL. 39, NO. 8, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2851
TABLE 3. Apparent Pseudo-First-Order Rates of Decolorization (k′obs) and COD Reduction (k′′obs) of Effluent as a Function of Applied Current Density at Different Flow Rates k′obs × 102 (s-1)
k′′obs × 102(s-1)
mA/cm2
5 mL/min
10 mL/min
15 mL/min
5 mL/min
10 mL/min
15 mL/min
10.0 19.9 29.9 39.9
7.11 8.18 8.26 8.27
0.63 5.80 10.90 14.67
0.78 4.63 12.55 19.63
2.18 3.31 4.42 6.73
2.12 2.78 4.13 5.37
0.33 3.35 3.50 5.20
Discussion of Experimental Results
FIGURE 5. Plot of pseudo-first-order rate vs [O] at (a) 5, (b) 10, and (c) 15 mL/min. almost independent of current density. At a given current density, it decreased to a negligible value with the increase in flow rate at 10 mA/cm2, but it decreased uniformly at 19.9 mA/cm2 due to a regular decrease in the liberation of species reacting with the dye molecules. In contrast, k′obs increased uniformly with high flow rates at 29.9 and 39.9 mA/cm2, maintaining status quo in the liberation of chemically active species at the electrode.
k′obs ) % decolorization × flow rate
(9)
k′′obs ) % COD reduction × flow rate
(10)
Similarly, the k′′obs value at a given flow rate increases correspondingly with the increase in current density. Expectedly, the liberation of chemically active species increases with the increase in current density. Unlike the k′obs, k′′obs decreased marginally with the increase in flow rate at all current densities indicating that the chemical steps responsible for the color removal and COD reduction are two independent processes. The observed rates k′obsand k′′obs at a given flow rate were plotted against the applied current density. The plots k′obs vs current density at 10 and 15 mL/min were linear and showed an intercept on the current (threshold current for decolorization) axis between 8 and 9 mA/cm2. Assuming that there is no consumption of [O] in the reactor and the rate of its production and the rate of its consumption are equal, the value of [O] (mole equivalents of active [O] species produced per mole of electron discarged at the anode) at a given current density and flow was obtained by eq 11. Plots k′obs against [O] (Figure 5) at a given flow rate were linear and showed a negative intercept at high flow (10 and 15 mL/min) rates. From slopes of these plots, the second-order rate constants k′ was calculated to be 2.9 M-1 s-1 at 5 mL/min, 76.2 M-1 s-1 at 10 mL/min, and 156.1 M-1 s-1 at 15 mL/min. Unlike the plots of k′obs vs [O], the plots of k′′obs vs [O] were linear and passed through the origin at all flow rates. The values of k′′ obtained from the slopes of these plots remained more or less constant (1.19 M-1 s-1 at 5 mL/min, 1.79 M-1 s-1 at 10 mL/min, and 3.57 M-1 s-1 at 15 mL/min).
[O] ) 2852
9
1000 × current density × time 96487 × flow rate
(11)
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 8, 2005
Azo dyes are synthetic organic colorants (1). They consist of a functional group possessing two doubly bonded nitrogens (-NdN-) attached to a trivalent (sp2) carbon atom of aromatic or heterocyclic nucleus on one side and an unsaturated molecule of the carboxylic, heterocyclic, or aliphatic type on the other side. Usually they are classified as acid, base/cationic, direct, disperse, mordant, or reactive types. The chromophoric nature of these dye effluents is due to the electronic transitions between the nonbonding electrons of N-atoms of the azo group and the π-molecular orbital energy levels of conjugal systems attached to it. Partial or complete cleavage of NdN bond forming simple amines by a reductive mechanism or C-NdN-C bonds leading to the liberation of N2 by an oxidative mechanism results in the decolorization of the effluent. The reductive cleavage of Nd N double bond is a difficult process in aqueous or near aqueous solutions and demands drastic conditions and/or a catalyst (51). On the other hand, the oxidative cleavage of C-N bonds in C-NdN-C is relatively easier. It may be achieved either at the anode or by employing a potentially equivalent oxidizing agent. Quintessentially, the chemical oxygen demand (COD) of a dye effluent is the requirement of total oxidizing power able to annihilate the dye molecule and the organics. In the process of COD reduction, the effluent may also be reduced in color and hence these methods are preferred. In electrochemical methods, the reactions at the anode depend on the nature of dye, the type and composition of organic and inorganic ingredients, and the nature of electrode materials used. Of the two possible ways of anodic reactions, the direct oxidation of the effluent molecules depends on the potentials of the dye and the organic ingredients. If these potentials are greater than the decomposition of water or the chloride which is excessively present in all effluents, then the direct oxidation of the dye is extremely difficult unless electrocatalytic electrodes such as platinum are employed, or it is done by indirect ways. In all the indirect methods, an electroactive species already present or externally added to the effluent oxidizes itself to its nearest high valent state at the electrode surface and immediately converts back to its original state in the subsequent chemical reactions with the proximate dye molecules. This process continues till the COD of the effluent at the electrode-solution interface, thereby, the bulk solution, reaches a minimum value. Most of the dye effluents contain sodium chloride as the major constituent. Thus, the method of treating such solutions electrochemically is easy as they involve no addition of chemicals for supporting the electrolysis. When dye molecules are inactive, the chemical changes at the anode are water (eq 12) and/or chloride ion oxidations (eq 13) which generate active OH and/or Cl radicals leading to O2 and/or Cl2 liberation, respectively. O2 is a relatively weak oxidant and hence not useful in the effluent treatment, whereas the Cl atom and Cl2 are equally robust oxidizing agents, hence Cl2 and OH are considered for the sake of simplicity. The counter reactions at the cathode would be the reduction of only water when no other reducible species are present.
2H2O f O2 + 4H+ + 4e-
(12)
2Cl- f Cl2 + 2e-
(13)
Reactions 12 and 13 occur in dilute chloride solutions equally. However, the latter reaction predominantly occurs in concentrated ([Cl-] > 1 M) solutions. The gaseous chlorine dissolves in aqueous solutions due to ionization as indicated in eq 14. It is less in acidic solution due to HOCl instability and considerably more in basic solutions due to the ready formation of OCl- (pKa 7.44) (52) ions (eq 15) implying that the basic or neutral pH conditions are more favorable for conducting reactions involving Cl2.
Cl2 + H2O f HOCl + HCl
(14)
HOCl + OH- f OCl- + H2O
(15)
The 0, 25, and 50% diluted effluent studied here contains considerable amounts of dye with high COD (5957-2979 ppm) and inorganic (16.88% w/v) salts, mainly chlorides having pH 7.35-8.1 favorable for the study. The dye molecules are electroinactive in the permissible potential range +1.5 to -1.5 V as shown in Figure 4. Thus, the decomposition that led to the decolorization and COD reduction in the effluent is thus attributed to the anodic oxidation of dye molecules mediated chiefly by OCl- and to a lesser extent by OH radicals considering the direct degradation of dye molecules at the electrode is negligible. The active oxygen (OH) and the active chlorine (OCl- at pH > 7, HOCl at pH < 5) species are taken as the mediators generated at the anode. Electrodes such as DSA employed in the present investigation preferentially work as a catalytic electrode to liberate chlorine from inorganic chloride solutions (53). As evidenced in Table 1, the treatment of (∼50%) diluted effluent is beneficial as far as the maximum reduction in COD, decolorization, and the complete utilization of liberated chlorine in the normal current density range (e40 mA/cm2) at the anode. The data in Table 2 and the plots shown in Figures 6 and 7 reveal that the highest decolorization occurs at low current densities and low flow rates, or high current densities and high flow rates, whereas the highest COD reduction occurs at high current densities and slow flow rates, indicating that the mechanisms of the two processes are different wherein decolorization is a simple step that needs less electrical energy as compared to the process of COD reduction. The decolorization and COD reduction are correlated as a function of applied current density in Figure 8 at 5 mL/min flow. Interestingly, it shows that decolorization is a rapid process and reaches a maximum at low current density at 10 mA/cm2, while the COD reduction increases almost linearly throughout. About 85-99% of decolorization could be achieved between 10 and 20 mA/ cm2 under the present set of experimental conditions with a 25-30% reduction in overall COD. The negligible changes in pH and the minor change in electrical conductance of the effluent after treatment show that the chlorine derivatized products are not liberated during the electrolysis. This is further supported by the absence of characteristic IR (C-Cl) stretches in Figure 3b. The proximate similarities in Figure 3b and c reveal that the decomposed products in both chemical and electrochemical treatments are nearly the same. However, the decrease in COD in acidified (pH 2.8) effluent shows the low cell-efficiency at low pH which may be accounted for by the low stability of the active chlorine (HOCl) in acidic solution. Similarly, the low performance of the cell at high pH 11.6 is again attributed to the loss of active chlorine (OCl-) due to its disproportionation as shown in eq 16. The effluent treatment by
FIGURE 6. Plot of applied current vs percent decolorization at (a) 5, (b) 10, and (c) 15 mL/min.
FIGURE 7. Plot of flow rate vs percent decolorization at (a) 10.0, (b) 19.9, (c) 29.9, and (d) 39.9 mA/cm2.
FIGURE 8. Effect of applied current on (a) decolorization, and (b) COD reduction in the effluent at 5 mL/min flow. electrochemical method seems to be more appropriate at pH 7-8.
3OCl- f ClO3- + 2Cl-
(16)
Among all the sites in the dye molecule, the nitrogen atoms of azo and hetrocyclic aromatic ring amines are easy to oxidize. Earlier investigators (54) have suggested that the oxidants such as OCl- and per acids act as electrophiles and initiate the oxidation process by theirreaction with the nitrogen atoms. We believe that the high valent chlorine end of OCl- first attacks the electron rich N-atom and is reduced to Cl- and corresponding N-oxide intermediate which rapidly dissociates to give colorless oxides (RdO and R′dO) and N2 as shown in Scheme 1. The kinetic data in Table 3 showed that the pseudo-firstorder rates with respect to decolorization and COD reduction VOL. 39, NO. 8, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2853
SCHEME 1
TABLE 4. Decolorization Data Obtained in Pilot Scale Studiesa effluent
pH flow rate current % (L/min) (A) decolorization before after
NBD plant outlet NBD plant gutter Azo plant 1 Azo plant 2 Sulfur black 1 Sulfur black 2 Shed D2 Amal plant
1.0 0.7 1.0 0.7 0.67 1.0 0.3 0.3
98 100 97 99 98 99 100 99
94.3 99.0 94.3 95.5 54.1 75.0 55.5 Nil
7.2 7.2 8.2 6.0
6.1 6.7 6.4 4.9
9.6 7.0
8.6 7.9
a The experiments were carried out on the premises of M/s Atul Industries Ltd with a pilot-scale cell having two DSA (10 in. × 21 in.) and three cathodes (10 in. × 25 in.) placed alternatively.
are in the order of 102 s-1. Szpyrkowicz et. al. (27), who studied the electro-oxidation of pollutants in synthetic textile wastewater containing partially soluble disperse dyes in 0.1 M NaCl solution with different anode materials, have suggested a mechanism similar to that discussed above. They have reported a 39% COD reduction in a span of 40 min. However, the apparent pseudo-first-order rate for the color removal is only 2.54 × 10-4 s-1. The very low rate of color removal in their investigations may be attributed to the dispersive nature of dye and the low concentration Cl-. The present study also suggests that the electrochemical method of treatment does not always give rise to secondary pollutants such as chlorinesubstituted byproducts. However, an independent study on each one of the dyes constituting the effluent is essential for a close look at the site of oxidation and the nature of the end products (Scheme 1) in the process of decolorization and COD reduction by chemical or electrochemical methods. Table 4 depicts the results obtained in the field studies conducted with a pilot-size electrochemical cell with different plant discharges of the above said industry. Interestingly, 94-99% decolorization was achieved with the azo dye effluents and less than 75% in other cases. Calculations revealed that the electrochemical treatment of azo dye effluents for 94-99% decolorization is economical; costing around Indian Rupees (Rs.) 0.02-0.04 under the present set of experimental conditions, assuming power cost at the rate of Rs. 5 per KW.
Acknowledgments We are grateful to Dr. Prakash Bhate, Vice President, Research, M/s Atul Industries Limited, Atul, Gujarat, India for providing the effluent and the details of dyes present in it. We are also thankful to Dr. P. K. Ghosh, Director, for invaluable suggestions during this work.
Literature Cited (1) Kirk, I.; Othmer, D. F. The Encyclopedia of Chemical Technology, 4th ed.; Howe-Grant, M., Ed.; J. Wiley & Sons: New York, 1997; Vol. 3, p 821. (2) Laing, I. G. The impact of effluent regulations on the dyeing industry. Rev. Prog. Color. Relat. Top. 1991, 21, 56-71 (3) Koprivanac, N.; Jovanovic-Kolar, J.; Bosanac, G.; Meixner, J. Studies on wastewater decolorization by the precipitation/ flocculation process. Microchem. J. 1992, 46, 379-384. (4) Koprivanac, N.; Bosanac, G.; Grabaric, Z.; Papic, S. Treatment of waste water from dye industry. Environ. Technol. 1993, 14, 385-390. 2854
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 8, 2005
(5) Easton, J. R. The maker’s view, in colour in dye house effluent. Colour in Dyehouse Effluent; Cooper; P., Ed.; Society of Dyers and Colourists; The Alden Press: Oxford, 1995; pp 9-21. (6) Shelley, T. R. Dye pollution clean up by synthetic mineral. Int. Dyer 1994, 79, 26-31. (7) Pierce, J. Colour in textile effluents - the origins of the problem. J. Soc. Dyers Colour. 1994, 110, 131-133. (8) Gahr, F.; Harmanutz, F.; Oppermann, W. Ozonation - an important technique to comply with new German laws for textile wastewater treatment. Ozonation Wat. Sci. Technol. 1994, 30, 255-263. (9) Neill, C. O.; Hawkes, F. R.; Hawkes, D. L.; Lourenco, N. D.; Pinheiro, H. M.; Dele´e, W. Colour in textile effluents - souces, measurement, discharge consents and simulation: a review. J. Chem. Technol. Biotechnol. 1999, 74, 1009-1018. (10) Correia, V. M.; Stephenson, T.; Judd, S. J. Characterisation of textile wastewaters. Environ. Technol. 1994, 15, 917-929. (11) Brown, D. Effects of colorants in the aquatic environment. Ecotoxicol. Eniron. Saf. 1987, 13, 139-147. (12) Industrial Waste Control; Gurnham, C. F., Ed.; Academic Press: New York, 1965. (13) Lin, S. H.; Lin, C. M. Treatment of textile wastewater effluent by ozonation and chemical co-agulation. Water Res. 1993, 27, 1743. (14) Szpyrkowicz, I.; Zilio, G. F.; Canepa, P. Performance of a fullscale treatment plant for textile dying wastewater. Toxicol. Environ. Chem. 1996, 56, 23. (15) Michelsen, D. L.; Fulk, L. L.; Woodby, R. M.; Boardman, G. D. Adsorptive and chemical pretreatment of reactive dye discharge. In Emerging Technologies in Hazardous Waste Management III; ACS Symposium Series; American Chemical Society: Washington, DC, 1993; pp 119-136. (16) Churchley, J. H. Ozone for the dye waste colour removal: four years operation at Leak STW. Inte. Ozone Assoc. 1998, 20, 111120. (17) Clarke, E. A.; Anliker, R. Safety in use of organic colorants: health and safety aspects. Rev. Prog. Color. 1984, 14, 84-89. (18) Loyd, C. K.; Boardman, G. D.; Michelsen, D. L. Anaerobic/aerobic treatment of a textile dye wastewater. Hazard. Ind. Wastes 1992, 4, 593-601. (19) Rao, B. V.; Sastry, C. A. Removal of dyes from water and wastewater by coagulation and biological treatment. Indian J. Environ. Prot. 1987, 7, 441-450. (20) Rao, B. V.; Sastry, C. A. Removal of dyes from water and wastewater by adsorption. Indian J. Environ. Prot. 1987, 7, 363373. (21) Nowak, K. M.; Winnichi, T. Decolorization of dye solutions by continuous ultra-filtration. Desalination 1986, 60, 59. (22) Reemtsma, T.; Jakobs, J. Concerted chemical and microbial degradation of sulfophthalimides formed from sulfophthalocyanine dyes by white-rot fungi. Environ. Sci. Technol. 2001, 35, 4655-4659. (23) Arslan-Alaton, I.; Dogruel, S. Photodegradation of hydrophobic disperse dyes and dye-bath effluents by silicadodecatungstate (SiW12O404-/5-) nanoparticles. Water Sci. Technol. 2004, 49 (4), 171-176. (24) Feigelson, L.; Muszkat, L.; Bir, L.; Muszkat, K. A. Dye photoenhancement of TiO2-photocatalyzed degradation of organic pollutants: the organobromine herbicide bromacil. Water Sci. Technol. 2000, 42, 275-279. (25) Duarte, C. L.; Sampa, M. H. O.; Rela, P. R.; Oikawa, H.; Cherbakian, E. H.; Sena, H. C.; Abe, H.; Sciani, V. Application of electron beam irradiation combined to conventional treatment to treat industrial effluents. Radiat. Phys. Chem. 2000, 57, 513-518. (26) Brown, N. W.; Roberts, E. P. L.; Garforth, A. A.; Dryfe, R. A. W. Treatment of dyehouse effluents with a carbon-based adsorbent using anodic oxidation regeneration. Water Sci. Technol. 2004, 49, 219-225. (27) Pereira, M. F. R.; Soares, S. F.; Orfao, J. J. M.; Figueiredo, J. L. Adsorption of dyes on activated carbons: Influence of surface chemical groups. Carbon 2003, 41, 811-821. (28) Netpradit, S.; Thiravetyan, P.; Towprayoon, S. Adsorption of three azo reactive dyes by metal hydroxide sludge: Effect of temperature, pH, and electrolytes. J. Colloid Interface Sci. 2004, 270, 255-261. (29) Lorimer, J. P.; Mason, T. J.; Plattes, M.; Phull, S. S. Dye effluent decolourization using ultrasonically assisted electro-oxidation. Ultrason. Sonochem. 2000, 7, 237-242. (30) Lin, S. H.; Chen, M. L. Treatment of textile wastewater by chemical methods for reuse. Water Res. 1997, 31, 868.
(31) Sarasa, J.; Roche, M. P.; Ormad, M. P.; Gimeno, E.; Puig, A.; Ovelleiro, J. L. Treatment of wastewater resulting from dyes manufacturing with ozone and chemical coagulation. Water Res. 1989, 23, 2721. (32) Wu, J.; Witeman, M. A.; Law, S. E. Evaluation of membrane filtration and ozonation process for treatment of reactive dye wastewater. J. Environ. Eng. 1998, March, 272. (33) Kuo, W. G. Decolorizing dye wastewater with Fenton’s reagent. Water Res. 1992, 26, 881. (34) Carriere, J.; Jones, J. P.; Broadbent, A. D. Decolorization of textile dye solutions. Ozone Sci. Eng. 1993, 15, 189. (35) Szpyrkowicz, L.; Juzzolino, C.; Kaul, S. N. A comparative study on oxidation of disperse dyes by electrochemical process, ozone, hypochlorite and fenton reagent. Water Res. 2001, 35, 21292136. (36) Lin, S. H.; Chen, M. L. Treatment of textile wastewater by chemical methods for reuse. Water Res. 1997, 31, 868-876. (37) Szpyrkowicz, L.; Juzzolino, C.; Kaul, S. N.; Daniele, S.; De Faveri, M. D. Electrochemical oxidation of dyeing baths bearing disperse dyes. Ind. Eng. Chem. Res. 2000, 39, 3241-3248. (38) Lin, S. H.; Peng, C. F. Treatment of textile wastewater by electrochemical method. Water Res. 1994, 28, 277. (39) Gaur, P. M.; Prajapati, M. N.; Dasare, B. D. Decomposition of dyes in the wastewater by electrolytic method. Indian. J. Technol. 1993, 31, 663-665. (40) Vlyssides, A. G.; Israilides, C. J. Detoxification of tannery waste liquors with an electrolysis system. Environ. Pollut. 1997, 97, 147-152. (41) Naumczyk, J.; Szpyrkowicz, N. J.; Grandi, Z. F. Electrochemical treatment of textile wastewater. Water Sci. Technol. 1996, 34, 17-24. (42) Luo, G. S.; Jiang, W. B.; Lu, Y. C.; Zhu, S. L.; Dai, Y. Y. Two-phase electrophoresis separation of dyestuffs from dilute solution. Chem. Eng. J. 1999, 73, 137-141. (43) Fernandes, A.; Morao, A.; Magrinho, M.; Lopes, A.; Goncalves, I. Electrochemical degradation of C. I. Acid Orange 7. Dyes Pigm. 2004, 61, 287-296. (44) Baklan, V. Yu.; Radygin, G. V.; Gonopol’skij, S. N.; Shuster, N. A.; Rukitskij, A. I.; Makordej, F. V.; Khitrich, V. F.; Ponomarenko,
(45)
(46)
(47)
(48)
(49)
(50) (51) (52)
(53) (54)
T. I. Development of electrocoagulation procedure for treatment of wastewater. Khimiya i Tekhnologiya Vody 1994, 16, 334337. Stephen, J.; Khader, K. Y. H.; Bino, M. Electrooxidation of dyestuffs in wastewaters. J. Chem. Technol. Biotechnol. 1995, 62, 111-117. Can, O. T.; Bayramoglu, M.; Kobya, M. Decolorization of reactive dye solutions by electrocoagulation using aluminum electrodes. Ind. Eng. Chem. Res. 2003, 42, 3391-3396. Vaghela, S. S.; Jethva, A. D.; Gohil, M. S. Adimurthy, S.; Gour, P. M.; Susarla, V. R. K. S.; Ramachandraiah, G.; Ghosh, P. K. Proceedings of Indo-Italian workshop on Emerging Technologies for Industrial Wastewater Treatment and Environment. NEERI: Nagapur, India, Sept. 2002; pp 71-82. Vasudevan, D.; Vaghela, S. S.; Ramachandraiah, G. Electrosynthesis of dimethyl sulfone from dimethyl sulfoxide at a dimensionally stable anode. J. Appl. Electrochem. 2000, 30, 1302. Prakash, R.; Tyagi, B.; Ramacandraiah, G. Electrometric Studies on the reduction of Ru(III)-EDTA and its Hydrazinium Complexes. Indian J. Chem. 1997, 36, 201. Socrates, G. Infrared Characteristic Group Frequencies; John Wiley: New York, 1980. Chatt, J.; Dilworth, J. R.; Richards, R. L. Recent advances in the chemistry of nitrogen fixation. Chem. Rev. 1978, 78, 589. Kumar, K.; Margerum, D. W. Kinetics and Mechanism of generalacid assisted oxidation of bromide by hypochlorite and hypochlorous acid. Inorg. Chem. 1987, 26, 2706-2711. Vasudevan, D.; Anantharaman, P. N. Electrochemical synthesis of eosin from flouorescein. J. Appl. Electrochem. 1993, 23, 808. Oakes, J.; Gratton, P.; Wilkes, I.; Gordon-Smith, T. Kinetic and spectroscopic study of heterogeneous oxidation of azo dyes using hypochlorite. Color. Technol. 2001, 117, 76-81.
Received for review December 8, 2003. Revised manuscript received January 10, 2005. Accepted January 19, 2005. ES035370C
VOL. 39, NO. 8, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2855
