Decolorization of Kraft Bleaching Effluent by Advanced Oxidation

Environ. Sci. Technol. 2007, 41, 2510-2514

Decolorization of Kraft Bleaching Effluent by Advanced Oxidation Processes Using Copper (II) as Electron Acceptor M A R IÄ A C . Y E B E R , * , † KATHERINE P. ON ˜ ATE,† AND GLADYS VIDAL‡ Faculty of Science, Universidad Cato´lica de la Santı´sima Concepcio´n, P.O. Box 297, Concepcio´n, Chile, and Environmental Science Center EULA-Chile, Universidad de Concepcio´n, P.O. Box 160-C, Concepcio´n, Chile

Two advanced oxidation processes (AOPs), TiO2/UV/O2 and TiO2/UV/Cu (II), were used to remove color from a Kraft bleaching effluent. The optimal decoloration rate was determined by multivariate analysis, obtaining a mathematical model to evaluate the effect among variables. TiO2 and Cu (II) concentrations and the reaction times were optimized. The experimental design resulted in a quadratic matrix of 30 experiments. Additionally, the pH influence on the color removal was determined by multivariate analysis. Results indicate that color removal was 94% at acidic pH (3.0) in the presence of Cu (II) as an electron acceptor. Under this condition, the biodegradation of the effluent increased from 0.3 to 0.6. Moreover, 70% of COD (chemical oxygen demand) was removed, and the ecotoxicity, measured by Daphnia magna, was reduced. Photocatalytic oxidation to remove the color contained in the Kraft mill bleaching effluent was effective under the following conditions: short reaction time, acidic pH values, and without the addition of oxygen due to the presence of Cu (II) in the effluent. Moreover, residual Cu (II) was a minimum (0.05 mg L-1) and was not toxic to the next biological stage. The experimental design methodology indicated that a quadratic polynomial model may be used to represent the efficiency for degradation of the Kraft bleach pulp effluent by a photocatalytic process.

Introduction The Kraft bleach pulp effluent is strongly colored. Lignin derivatives, such as chlorinated organic compounds with a low biodegradability, emerge in the bleaching stage process. These compounds are classified as environmentally hazardous because of their toxicity and slow biodegradation. Recent advances in the advanced oxidation processes (AOPs) using TiO2 indicate that AEROXIDE (TiO2 Degussa P-25) is the most effective process to improve the biodegradation of recalcitrant compounds. In the AEROXIDE process, oxygen is used as an efficient electron trap, preventing electrons from returning to photogenerated holes. Additionally, the biodegradation efficiency could be improved by using metallic ions as * Corresponding author phone: 56-41-735250; fax: 56-41-735251; e-mail: [email protected]. † Universidad Cato ´ lica de la Santı´sima Concepcio´n. ‡ Universidad de Concepcio ´ n. 2510

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 7, 2007

electrons acceptors. Previous works have shown that synthetic phenols, chlorophenols, and colored compounds can be degraded by using photocatalytic technology, although the Kraft bleaching process effluent has not been studied (1-4). On the other hand, conventional biological treatments (e.g., aerated lagoon or activated sludge) were unable to remove either recalcitrant compounds or color contained in the Kraft bleaching mill effluent (5, 6). As a result, biological treatment did not biodegrade compounds of high molecular weight (greater than 1000 Da), and thus, to precipitate the recalcitrant compounds, the implementation of a tertiary treatment, such as a physicochemical process, is required. Furthermore, the treated effluent with a high chemical load could environmentally impact aquatic ecosystems with discharges into receiving waters (7, 8). Therefore, to improve the biodegradability, an oxidative pretreatment of the effluent using AEROXIDE technology is necessary to modify the structures of recalcitrant compounds prior to biological treatment. Moreover, from an economical point of view, replacing oxygen as the electron trap with copper (II) contained in the Kraft mill effluent could be beneficial. For these reasons, the goal of this study is to investigate the use of copper (II) as an electron trap inducing color removal from the Kraft bleaching effluent.

Experimental Procedures Advanced Oxidation. The effluent was obtained from the first extraction step E1 of the elementary chlorine free (ECF) bleaching pulp of radiata pine wood. The experiments were carried out in a continuously operated 2 L cylindrical reactor and irradiated inside with a HPL 120 W mercury lamp (λ > 254 nm), where the lamp’s radiation was measured with radiometer PMA 2200 solar light and the detector PMA 2120 UV. Titanium dioxide was provided by Degussa P-25 (80% anatase and 20% rutilo) and used in suspension. Effluent decolorization was determined by UV-vis spectrophotometry at 465 nm, and the color was expressed in mg of Pt L-1. The copper solution was obtained from Merck, and the ion concentration was determined by atomic absorption spectrophotometry. Toxicity Determination. The acute toxicity assays were performed before and after photocatalytic treatment, with Daphnia pulex. To assess the LC50 values, five dilutions were performed for every sample step and the respective control. In all dilutions, 20 nematodes were placed in four replicates with 10 mL of solution and five specimens of each one. The test was maintained for 24 and 48 h at a temperature of 20 °C, assessing the individual mortality (%) in each dilution. The mortality was analyzed with the statistical method Probit, and the LC50 values were calculated with the program TOSTAD. Experimental Design. The experimental design consisted of a statistically significant model of the phenomenon with a minimum set of well-chosen experiments. The mathematical model of multivariate methods was used to optimize processes, to simultaneously modify the changeable operations of a reaction, to identify the weight and the relation among variables, and to determine the synergies and antagonisms. Once response surface methodology is tested, the resulting model of response surface methodology can be used to elaborate contour plots that predict the values of the response at any point in the experimental region of interest. This experimental design to evaluate of interaction effects among variables minimizes the number of experiments, 10.1021/es062544s CCC: $37.00

 2007 American Chemical Society Published on Web 02/27/2007

TABLE 1. Characterization of E1 Stage Effluent from Bleach Kraft Pulp parameter

value

pH color (mg of Pt L-1) total phenols (mg L-1) Cu (II) (mg L-1) COD (mg L-1) BOD (mg L-1)

10.6 1500 5.42 0.5 1600 449

identifying the statistically most significant experiment to perform. This methodology is used in the present research to study the influence of TiO2 and copper (II) concentrations with the reaction time on the color removal from Kraft bleaching wastewater. The experimental matrix for three variables is a quadratic model with 30 experiments. Before determining the optimal response, the starting matrix model with 11 experiments was performed to determine the influence of pH in the model. The general polynomial response (eq 1) is a quadratic polynomial model representing the associated response function

Y1 ) b0 + b1X1 + b2X2 + b3X3 + b11X12 + b22X22 + b33X32 + b12X1X2 + b13X1X3 + b23X2X3 (1) where b0 is the average value of the experimental response; b1 is the main effect of the coded variable X1; b2 is the second effect; b3 is the third effect; b11, b22, and b33 are the quadratic effects of the coded X1, X2, and X3, respectively; and b12, b13, and b23 are the interaction effects between the respective coded variables. Analytical Methods. The COD (mg of O2 L-1), TOC (mg of O2L-1), and total phenol compounds (mg of O2 L-1) were determined following standard method ISO 8466--1 and DIN38402 A51, ISO 8466--1 and DIN 38402 A51, and 5220-D and ISO 15705, respectively. The biological oxygen demand (BOD5 mg of O2 L -1) was determined using partial oxygen pressure (OXITOP) sensors, standard methods 5210-D. The samples were inoculated with 1 mL of the activated sludge taken from a continuously operating laboratory reactor that was grown in a pulp bleaching effluent. The pH was measured by an electrode bifunction SenTix WTW Inolab, and the effluent color was determined at 465 nm and referenced to a Pt-Co standard solution (EPA 00080). The acute toxicity assays were performed before and after photocatalytic treatment with D. pulex. To obtain the LC50 (lethal concentration) values, five dilutions were performed for every sample step and the respective control. In all dilutions, 20 nematodes were placed in four replicates with 10 mL of solution and five specimens of each one. The test was maintained for 24 and 48 h at a temperature of 20 °C, assessing the individual mortality (%) in each dilution, and the results were analyzed with the statistical method. Probit and the LC50 values were calculated with the program TOSTAD.

TABLE 2. Experimental Matrix for Natural Variables and Corresponding Experimental Response to Optimize Color Removal at pH 3.0 and 20 °C expt no.

order no.

TiO2 (g)

Cu (II) (mg L-1)

time (min)

color removal (%)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

22 1 4 7 20 13 29 12 9 16 27 18 3 17 25 28 26 23 24 14 11 5 21 6 15 30 10 19 8 2

0.5 0.75 1 0.5 0.75 1 0.5 0.75 1 0.5 0.75 1 0.5 0.75 1 0.5 0.75 1 0.5 0.75 1 0.5 0.75 1 0.5 0.75 1 0.75 0.75 0.75

0.5 0.5 0.5 1 1 1 1.5 1.5 1.5 0.5 0.5 0.5 1 1 1 1.5 1.5 1.5 0.5 0.5 0.5 1 1 1 1.5 1.5 1.5 1 1 1

30 30 30 30 30 30 30 30 30 90 90 90 90 90 90 90 90 90 150 150 150 150 150 150 150 150 150 90 90 90

88 94 95 87 87 95 86 99 95 89 92 95 90 94 96 91 89 96 88 83 87 88 89 89 87 94 91 93 90 92

mill wastewater. The initial effluent pH was adjusted at pH 3.0 and remained constant during the first modeling. Table 2 shows the experimental matrix for three variables, the number of the experiments, and the corresponding experimental response. All experiments were performed under identical conditions, pH 3.0 and room temperature, which remained constant in all experiments. In the same table, it can be observed that 99% of the color was removed with the experiment order number 12, with 0.75 g of TiO2, 1.5 ppm copper (II), and 30 min reaction time. Nevertheless, the response surface for the experimental design to evaluate the synergism among variables (Figure 1) indicates that the time variable is independent and that the polynomial

Results and Discussion The effluent was obtained from a modified continuous cooking (MCC) process of a local Kraft mill that uses Pinus radiata wood. The ECF effluent was generated after the first alkaline extraction of the bleaching sequence with 100% chlorine substitution (D0Eop), and the global parameter values for the initial ECF effluent are shown in Table 1. A quadratic model was used to determine the optimal experimental variables in photocatalytic oxidation of pulp

FIGURE 1. 1. Surface response to the quadratic experimental design for the color removal. VOL. 41, NO. 7, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2511

TABLE 3. Experimental Matrix of Natural Variables and Corresponding Experimental Response to Optimize pH for Color Removal Using the Optimal Point Determined in the Quadratic Matrix expt no.

order no.

TiO2 (g)

Cu (mg L-1)

time (min)

pH

color removal (%)

1 2 3 4 5 6 7 8 9 10 11

4 3 9 6 11 5 2 7 8 1 10

0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75

1 1 1 1 1 1 1 1 1 1 1

30 90 30 90 17.58 102.42 60 60 60 60 60

2 2 5.5 5.5 3.75 3.75 1.27 6.22 3.75 3.75 3.75

94 94 53 47 81 84 94 25 76 81 74

response is conditioned by only two variables, TiO2 and the copper (II) concentration. Indeed, it can be observed that when the catalyst and copper ion concentration increases, color removal also increases, indicating synergism between these two variables. In the contour plot, the difference in the color reduction can be observed, where 1.0 mg L-1 or more of copper ions is not significant, and the optimal point was determined for the experimental values of 0.75 g of TiO2 and 90 min reaction time with 1 mg L-1 copper (II). At this optimal point, the optimal pH was tested, building a new matrix with the variables of pH and reaction time. Table 3 presents the resulting model with 11 experiments for two variables and the same response, color removal, as a percentage. The response analysis indicates that the decoloration is not important at a pH below 3 and that color removal does not increase at higher pH values. Therefore, the color removal kinetics and enhanced biodegradability were performed using the results obtained since both experimental matrices found optimal response at pH 3. The objective of the experimental design was to determine the influence of each independent variable on conversion yield. The constant value represents the yield for the color removal, and the coefficients indicate the importance of the factor in the equation, where the catalyst concentration variable is found to be the most influential and the reaction time is an independent variable. As a result, eq 1 was transformed into eq 2. Using the parameters given by the polynomial response and replacing the coded for the real values in the equation, the total color removal achieved was 94% (eq 3). This value is in agreement with the optimal response, order number 17.

Y2 ) b0 + b1X1 + b2X2

(2)

Y3 ) 92.07 + 1.98(0.75) + 0.75(1)

(3)

The optimal conditions identified by the models for bleaching effluent color removal were experimentally validated as shown in Figure 3, where the kinetics indicate system efficiency after 20 min of reaction, with constant values after that time. Figure 3 shows color reduction over time for TiO2/ UV/oxygen and TiO2/UV/copper (II) systems applied to the effluent. As can be observed, color reduction in the effluent is achieved after 20 min of reaction, resulting in 94% reduction. As observed in the model’s response surface, time is an independent variable. Nevertheless, to verify if more time improved the biodegradability of the remaining organic matter, the reaction was followed up to 90 min of reaction time. In the same figure, it is important to emphasize that copper (II) acts efficiently as an electron acceptor, reducing the 2512

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 7, 2007

FIGURE 2. 2. Surface response to the starting experimental design for the color removal.

FIGURE 3. Color removal kinetics using optimized AOP treatment: TiO2 0.75 g L-1 and Cu(II) 1 mg L-1 and pH 3: (]) O2/UV, (0) Cu(II)/UV/O2, (4) TiO2/UV/O2 pure, (b) TiO2/UVCu II, and (O) TiO2/UV/ air.

FIGURE 4. Kinetics of the COD and BOD5 removal using optimized AOP treatment: (O) COD (mg of O2 L-1) and (b) BOD5 (mg of O2 L-1). copper ions by 98% with the photocatalytic treatment. This result is important when the treating effluent is contaminated with metal ions and organic matter because it avoids the need for additional oxygen. At the same time, as can be observed in Figure 4, the reduction of the chemical demand of oxygen COD (mg of O2 L-1) followed the same profile as color reduction, reaching 70% after 20 min of treatment. A similar response can be observed in the profile of the reduction of the biochemical demand of oxygen BOD5 (mg of O2 L-1), where the reduction also reached 70%. At the same time, the BOD5/COD ratio increased from 0.3 to 0.6,

FIGURE 5. TOC (white) and total phenol (gray) compound removal using optimized AOP treatment. FIGURE 7. Biotoxicity reduction as a function of remaining TOC: (white) initial, (light gray) TiO2/UV/Cu (II), and (medium gray) TiO2/ UV/O2.

FIGURE 6. Toxicity reduction by the optimized AOP treatment: (white) initial, (dark gray) TiO2/UV/Cu (II), and (light gray) TiO2/UV/ O2. indicating that the biodegradability of the effluent improved considerably by the end of the treatment for the optimal point studied. Similar studies (9-11) pointed out that ZnO and TiO2 photoassisted catalysis degraded organic matter dissolved in paper mill effluents, reduced the toxicity, and significantly improved biodegradability. However, effluent decolorization only reached 40% at pH 10, and the initial alkaline pH decreased rapidly (3 units) during the first minutes of the reaction with a profile similar to that decolorization, maintaining both parameters without changes over time. It is important to indicate that in this study, the acidic pH changed the catalyst’s superficial characteristics, charging positively the photocatalyst’s surface at pH < pHzpc, prioritizing adsorption of anionic compounds, and producing greater effluent decolorization. Similar evidence was found by others authors in photocatalysis treatment of colored solutions (12, 13). Figure 5 shows that the improvement of effluent biodegradability corresponds with mineralized organic matter, reducing total organic carbon (TOC) by 50% and practically eliminating total phenols derived from the dissolved lignin with the optimal treatment. Together with organic matter mineralization and phenol compound elimination, the effluent’s initial toxicity was notably improved. This result can be observed in Figure 6, where the initial toxicity has a LC 50 of 30%, which is considered to be toxic, while the effluent’s toxicity after photocatalytic treatment is reduced by about 50%, reaching a LC50 of 60%. Others authors have reported the formation of toxic intermediates as degradation products of bio-recalcitrant organic contaminants treated by AOP systems, specifically chlorine pesticide used as a model and other problematic pollutants where the toxicity of treated samples increased with treatment time, indicating the processes’ inefficient detoxification capacity (14). It is important to emphasize that pulp and paper wastewater

contains significant amounts of non-biodegradable organic compounds, particularly polychlorinated compound derivates from the bleaching stage, and that the AOP system used was able to detoxify it after a few minutes of treatment without the formation of more toxic intermediates as can be observed in Figure 6, where the detoxification reaches 70% with the same TOC maintained in both cases. On the other hand, treatment efficiency should be evaluated by calculating the electrical energy per order (EE/O) value (6). The electrical energy per order is directly related to the treatment cost and is defined as the electrical energy in kilowatt hours (kWh) required to degrade a contaminant by 1 order of magnitude in 1 m3 of contaminant water. For both the AOP pulp and the paper wastewater treatment, the EE/O was 65.8 Kwh/ m3, which is lower in comparison with a similar treatment using model compounds (1), where a lower EE/O indicates a more efficient and less expensive treatment. The low biodegradability of the organic matter and the high color in the bleach Kraft pulp effluent made this residual water toxic and inhibitory for the microorganisms and organisms in the environment. Nevertheless, in a few minutes of treatment with the photocatalytic system, the effluent conditions can be considerably improved, resulting in a more energetically efficient process and treated wastewater that does not produce problems in the ecosystem.

Acknowledgments This research was supported by The Science Faculty of Universidad Cato´lica de la Santı´sima Concepcio´n.

Literature Cited (1) Lizama, C.; Yeber, M.; Freer, J.; Baeza, J.; Mansilla, H. Reactive dye decoloration by TiO2 photo-assisted catalysis. Water Sci. Technol. 2001, 44, 197-203. (2) Parra, S.; Malato, S.; Pulgarin, C. New integrated photocatalyticbiological flow system using supported TiO2 and fixed bacteria for the mineralization of isoproturon. Appl. Catal., B 2002, 36, 131-144. (3) Gernjak, W.; Glaser, T.; Malato, S.; Caceres, J.; Bauer, R.; Ferna´ndez, A. Photo-Fenton treatment of water containing natural phenolic pollutants. Chemosphere 2003, 50, 71-78. (4) Farre, M.; Franch, M.; Malato, S.; Ayllo´n, J.; Peral, J.; Dome´nech, X. Degradation of some biorecalcitrant pesticides by homogeneous and heterogeneous photocatalytic ozonation. Chemosphere 2005, 58, 1127-1133. (5) Chamorro, S.; Xavier, C.; Vidal G. Behavior of aromatic compounds contained in the Kraft mill effluent measurements by UV-vis. Biotechnol. Prog. 2005, 21, 1567-571. (6) Bankian, G. T.; Mehrvar, M. Integration of advanced oxidation technologies and biological processes: Recent developments, trends, and advances. J. Environ. Sci. Health 2004, 39, 30293081. VOL. 41, NO. 7, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2513

(7) Muslow, S.; Grandjean, M. Incompatibility of sulphate compounds and soluble bicarbonate salts in the Rio Cruces waters: An answer to the desappearance of Egeria densa and blacknecked swans in a RAMSAR sanctuary. Ethics Sci. Environ. Politics 2006, 5-11. (8) Moraes, S. G.; Dura´n, N.; Freire, R. S. Remediation of Kraft E1 and black liquor effluents by biological and chemical processes. Environ. Chem. Lett. 2006, 4, 87-91. (9) Mansilla, H.; Yeber, M.; Freer, J.; Rodrı´guez, J.; Baeza, J. Homogeneous and heterogeneous advanced oxidation of a bleaching effluent from the pulp and paper industry. Water Sci. Technol. 1997, 35, 4, 273-278. (10) Yeber, M.; Rodrı´guez, J.; Freer, J.; Baeza, J.; Dura´n, N.; Mansilla, H. Advanced oxidation of a pulp mill bleaching wastewater. Chemosphere 1999, 39, 1679-1688. (11) Yeber, M.; Rodrı´guez, J.; Freer, J.; Dura´n, N.; Mansilla, H. Photocatalytic degradation of cellulose bleaching effluent by supported TiO2 and ZnO. Chemosphere 2000, 41, 1193-1197.

2514

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 7, 2007

(12) Machado, A.; Miranda, J.; Freitas, R.; Duarte, E.; Ferreira, L.; Alburquerque, Y.; Ruggiero, R.; Sattler, C.; Oliveira, L. Destruction of the organic matter present in effluent from a cellulose and paper industry using photocatalysis. J. Photochem. Photobiol., A 2003, 155, 231-241. (13) Chung-Hsin, W.; Chung-Liang, C. H.; Chao-Yin, K. Decolorization of amaranth by advanced oxidation processes. React. Kinet. Catal. Lett. 2005, 86, 37-43. (14) Evgenidou, E.; Fytianos, K.; Poulios, I. Semiconductor-sensitized photodegradation of dichlorvos in water using TiO2 and ZnO as catalysts. Appl. Catal., B 2005, 59, 81-89.

Received for review October 23, 2006. Revised manuscript received January 19, 2007. Accepted January 19, 2007. ES062544S