Synthesis and characterization of a novel cationic chitosan-based flocculant with a high water-solubility for pulp mill wastewater treatment


Раздел науки:

Читать онлайн: 
Текст статьи: 
Jian-Ping Wang a,b, Yong-Zhen Chen a, Shi-Jie Yuan a, Guo-Ping Sheng a, Han-Qing Yu a,* a Department of Chemistry, University of Science & Technology of China, Hefei 230026, China b Department of Agricultural and Biological Engineering, University of Illinois at Urbana-Champaign, 1304 W. Pennsylvania Avenue, Urbana, IL 61801, USA a r t i c l e i n f o Article history: Received 27 March 2009 Received in revised form 8 August 2009 Accepted 27 August 2009 Available online 2 September 2009 Keywords: Chitosan Flocculant Grafting Optimization Pulp mill wastewater Water-solubility a b s t r a c t In this work, pulp mill wastewater was treated using a novel copolymer flocculant with a high water-solubility, which was synthesized through grafting (2-methacryloyloxyethyl) trimethyl ammonium chloride (DMC) onto chitosan initiated by potassium persulphate. The experimental results demonstrate that the two main problems associated with the utilization of chitosan as a flocculant, i.e., low molecular weight and low water-solubility, were concurrently sorted out. The physicochemical properties of this flocculant were characterized with Fourier-transform infrared spectroscopy, 1H nuclear magnetic resonance spectroscopy, X-ray powder diffraction and field emission scanning electron microscopy. Reaction parameters influencing the grafting percentage, such as temperature, reaction time, initiator concentration and monomer concentration, were optimized using an orthogonal array design matrix. With an increase in grafting percentage, the water-solubility of the flocculant was improved, and it became thoroughly soluble in water when the grafting percentage reached 236.4% or higher. Its application for the treatment of pulp mill wastewater indicates that it had an excellent flocculation capacity and that its flocculation efficiency was much better than that of polyacrylamide. The optimal conditions for the flocculation treatment of pulp mill wastewater were also obtained. ª 2009 Elsevier Ltd. All rights reserved. 1. Introduction There are more than 10,000 papermaking mills in China, and more than 63% of the mills are using the straw slurry as the feedstock. It is estimated that for each ton of paper produced, about 10–15 m3 wastewater is generated. The large amount of base (such as alkali lignin and sodium salts of organic acids) in the wastewater increases the difficulty of treatment. In general, there are two methods for the treatment of pulp mill wastewater. One method is to be concentrated and then used as fuels to recover the energy and base. Another method is to be acidized at about 70–80 C firstly to precipitate the lignin, followed by flocculation and biological treatment. However, these two methods will expand a great deal of additional energy and are expensive. On the other hand, the abundant lignin existing in the wastewater could not be reclaimed efficiently and thus increases the burden of the biological treatment. Therefore, a convenient and cost-effective method, e.g., flocculation, for the pretreatment of pulp mill wastewater prior to biological treatment is necessary. In this case, an * Corresponding author. Tel.: þ86 551 360 7592; fax: þ86 551 360 1592. E-mail address: (H.-Q. Yu). Available at journal homepage: 0043-1354/$ – see front matter ª 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.watres.2009.08.040 w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 5 2 6 7 – 5 2 7 5 efficient and environmentally friendly flocculant that could be used under basic conditions is highly desirable for the pretreatment of pulp mill wastewater. Utilization of natural or modified natural polymers as flocculants for wastewater treatment has recently attracted increasing interests (Xiao et al., 1995; Yang et al., 2004; Franks, 2005). Chitosan, the second most abundant natural organic resource next to cellulose on the earth, has been extensively used as a flocculant for wastewater treatment and sludge dewatering, as it is nontoxic, biodegradable and environmentally friendly (Bratskaya et al., 2004). When chitosan is used as a flocculant, its molecular weight and water-solubility are both most concerned factors (Divakaran and Pillai, 2004). Its high molecular weight is favorable for improving the aggregation of colloids, thus promoting the separation of aggregates with water, but because of the inter-molecular and intra-molecular hydrogen bonding, chitosan can only be dissolved in an acidic solution through the interaction between Hþ and –NH2. However, at pH lower than 5, the acidic condition could accelerate the degradation of chitosan and consequently reduce its flocculation efficiency (Divakaran and Pillai, 2004). On the other hand, so far, all the modifications of chitosan could only improve one of the two factors, either molecular weight or water-solubility, but not both (Sashiwa et al., 2001; Li et al., 2004; Bratskaya et al., 2004). In our previous studies, two types of flocculants based on chitosan were prepared using the grafting method, which were initiated by gamma-ray radiation (Wang et al., 2007a, 2008). However, such a modification resulted in only an improvement of molecular weights, rather than water-solubility. Grafting has been proven to be an effective modification technique for chitosan, as there are abundant amino groups and hydroxyl groups in chitosan backbone, which could react with vinyl monomers under mild conditions (Lazaridis et al., 2007). A large amount of work has been carried out to perform grafting copolymerization of chitosan and vinyl monomers (Jenkins and Hudson, 2002; Neira-Carrillo et al., 2005; Lazaridis et al., 2007; Zhou et al., 2007). The graft copolymers could be biodegradable to some extent because of the presence of polysaccharide backbone, and also be stable against shearing because of the attachment of flexible synthetic polymers onto rigid or semi-rigid polysaccharide backbones (Rath and Singh, 1997). In addition, the flexible grafting chain would increase the possibility for the flocculant to approach to particles in wastewater, and thus improve the flocculation ability (Singh, 1995). Because colloids in the pulp mill wastewater are negatively charged, the cationic flocculant is suitable for charge neutralization, and thus favorable for the flocculation of the wastewater. However, synthetic cationic polymers, e.g., PDMC [poly (2-methacryloyloxyethyl) trimethyl ammonium chloride], are expensive and hydrophilic, which makes them difficult for transportation and storage, and thus holdbacks their application. Moreover, their molecular weights are not sufficiently high for a flocculant. Therefore, synthesis of a cationic flocculant without these disadvantages of the synthetic cationic monomer is highly desirable for the pretreatment of pulp mill wastewater. Taking into account all the factors above, a cationic monomer, (2-methacryloyloxyethyl) trimethyl ammonium chloride (DMC), was selected as the monomer grafted onto chitosan initiated by potassium persulphate to prepare a flocculant for the pretreatment of pulp mill wastewater in this work. The introduction of DMC onto chitosan backbone was expected to increase the molecular weight and cationic content, and to improve its water-solubility. The synthetic flocculant, chitosan-g-PDMC [chitosan-graft-poly (DMC)] was characterized, its flocculation ability was evaluated by jar tests with kaolin suspensions, and it was used for the treatment of the pulp mill wastewater. 2. Materials and methods 2.1. Synthesis of the chitosan-based flocculant The grafting reaction was initiated by a radical initiator, potassium persulphate. Since the deprotonated C-2 amino group in chitosan is a powerful nucleophile (pKa z 6.5) (Ashmore and Hearn, 2000), it could react with electrophilic reagents readily. Such a reaction process is illustrated in Fig. 1. The chitosan with a molecular weight of 520 kDa and a deacetylation degree of 95% concentration was kept at 0.115 mol/L throughout the experiment. First, chitosan and DMC solution were prepared with 1.0% acetic acid in a threenecked flask with constant stirring and bubbling of a slow stream of nitrogen at ambient temperatures. Then, the flask was placed in a preset water bath and a given amount of potassium persulphate was added into the solution as an initiator. After that, the mixture was continuously stirred with a mechanical stirrer under nitrogen atmosphere. At the end of reaction, the sample solutions were precipitated in acetone and separated through filtration. Thereafter, the homopolymer formed in the reaction was removed using ethanol in Soxhlet apparatus. The extracted products were dried in a vacuum oven at 50 C until a constant weight was obtained. The grafting percentage was calculated according to the following equation: % grafting ¼ W2  W1 W1  100 (1) where W1 and W2 are the weights of original and grafted samples, respectively. 2.2. Characterization of the chitosan-based flocculant Infrared spectra of chitosan and the flocculant were recorded with a Fourier-transform infrared spectroscopy (FTIR) spectrometer (Magna-IR 750, Nicolet Instrument Co., USA) using a potassium bromide disc technique (Lim et al., 2008). The 1H nuclear magnetic resonance spectroscopy (1H NMR) spectra were measured on a 300-MHz spectrometer (AVANCE) with standard pulse programs in D2O (Chang et al., 2002). X-ray powder diffraction (XRD) patterns of the polymers were obtained with an X-ray diffractometer (D/Max-rA, Japan) with graphite monochromatized Cu Ka radiation (l ¼ 1.54056 A ˚ ). Field emission scanning electron microscopy (FESEM) images of chitosan and chitosan-g-PDMC particles were obtained using a JSM-6700F electron microscope (JEOL CO., Japan). All samples were metalized with gold prior to the analysis. Flocculation ability of chitosan and chitosan-g-PDMC with different grafting percentages was evaluated using the same 5268 w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 5 2 6 7 – 5 2 7 5 method as described in one of our published papers (Wang et al., 2007a,b). The microscopic photographs of kaolin clay, the floccules of kaolin clay obtained from the suspensions treated with chitosan and the flocculant were taken with a phase contrast microscope (Olympus, U-5RE-2, Japan). The settling tests with chitosan and chitosan-g-PDMC respectively were conducted using the same method as described previously (Wang et al., 2008). 2.3. Pulp mill wastewater treatment tests Pulp mill wastewater was blending black liquor from the primary sedimentation tank of Guoyang Paper and Pulp Mill Co., China. The initial pH, chemical oxygen demand (COD) and turbidity were 6.99, 1358 mg/L and 1209 NTU, respectively. The average size of the colloid particles in this wastewater was 544 nm. Aluminum chloride was used as the coagulant in the coagulation–flocculation process. For comparison, the conventional synthetic polymeric flocculant, PAM, and chitosan-g-PDMCwererespectivelyusedastheflocculantinthetwo independent jar tests. A 23 full factorial central composite design (CCD) was used to evaluate the effect and interaction of such three experimental factors, i.e., coagulant dosage, flocculantdosageandpH(Ahmadetal.,2005).Removalefficiencies of turbidity, lignin, chemical oxygen demand (COD) and water recovery were chosen as the dependent output variables. The operating conditions were optimized using the response surface methodology (RSM). The experimental design and the respective levels of the variables are listed in Table 1. 3. Results and discussion 3.1. Physicochemical characteristics of the chitosan-based flocculant 3.1.1. FTIR spectrum The infrared spectroscopy confirms the occurrence of the graft copolymerization (Fig. 2). The bands of 2911 and 2866, 1400–1100 and 1070–1020 cm1 are attributed to the C–H stretching, C–O stretching of secondary alcohol and C–O stretching of the primary alcohol, respectively. The strong peak around 3400 cm1 could be assigned to the stretching vibration of O–H, the extension vibration of N–H, and intermolecular hydrogen bonds of the polysaccharide. The band at 1597 cm1, the characteristic peak of primary amine N–H CH 2 O CH 2OH OH O NH 2 O CH 2OH OH O NH 2 O CH 2OH OH O NH 2 O CH 2OH OH O NH-CO-CH 3 O CH 2OH OH O NH 2 O CH 2OH OH O NH 2 O CH 2OH OH O NH O CH 2OH OH O NH 2 O CH 2OH OH O NH O CH 2OH OH O NH-CO-CH 3 O CH 2OH OH O NH O CH 2OH OH O NH CH 2 C CH 3 COOC 2H4N(CH3)3Cl K 2S2O8 O CH 2OH O O CH 2OH OH OH O O CH 2OH NH O O CH 2OH NH OH OH OH NH-CO-CH 3 NH CH 2 CH2 H C 3C COOC2H4N(CH3)3Cl CH 2 CH 3C COOC2H4N(CH3)3Cl CH 3C COOC2H4N(CH3)3Cl CH 2 n HC CH 3C COOC2H4N(CH3)3Cl C CH 3 COOC 2H4N(CH3)3Cl C CH 3 COOC 2H4N(CH3)3Cl H H3C C 2C COOC 2H4N(CH3)3Cl Fig. 1 – Schematic representation of graft copolymerization of chitosan. w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 5 2 6 7 – 5 2 7 5 5269 vibration in chitosan, disappeared in the profile of graft copolymer, indicating the deformation of the primary amine in the grafting copolymer. This implies that the grafting had occurred at –NH2 groups. The new peaks appearing at 1729, 1481 and 954 cm1, should be assigned to the vibrating absorption of carbonyl groups in chitosan-g-PDMC, the methyl groups of ammonium and quaternary ammonium salt in PDMC, respectively. These spectra demonstrate that PDMC were successfully grafted onto chitosan backbone. 3.1.2. 1H NMR spectroscopy The 1H NMR spectroscopy of the prepared flocculant and the assignment of the signals are shown in Fig. A in Supplementary online materials. The two intensive signals observed at 3.2 and 4.7 ppm were attributed to the spectrum of protons of the methyl groups on the quaternary ammonium salt and that of protons of the hydroxyl groups in chitosan backbone, respectively. The peaks at 2.9, 1.1, 4.6 and 3.6 ppm, were respectively attributed to the proton signals of H-10, H-20, H-30 and H-40 in the grafting chain. The proton signals of chitosan backbone were observed at 1.1 ppm for H-2, at 2.0 ppm for H-4, at 4.7 ppm for H- 1 (overlapped by hydroxyl group) and at 3.6–3.7 ppm for H-2, H-5 and H-6 (overlapped by H-40 protons). The 1H NMR spectroscopy also confirms the grafting of DMC onto chitosan backbone. 3.1.3. XRD patterns and FESEM images The XRD profiles of the prepared flocculant present distinct crystalline peaks, compared with that of chitosan (Fig. 3). The profiles of chitosan showed two peaks around 2q ¼ 10.4 and 20.1, which corresponded to 020 and 110 reflections, respectively (de Vasconcelos et al., 2007). For the flocculant, the two peaks decreased drastically, suggesting a decrease in crystallinity. This might be attributed to the fact that the grafting of PDMC onto chitosan reduced the hydrogen bonding ability of chitosan. The introduction of DMC onto chitosan backbone destroyed its original ordered structure. The FESEM images of chitosan and the flocculant in Fig. 4 also indicate that the ordered laminated structure of chitosan was destroyed when PDMC was grafted onto the backbone. 4000 3000 2000 1000 Chitosan-g-PDMC Chitosan Transmittance Wavenumbers (cm-1) Fig. 2 – FTIR spectra of chitosan and chitosan-g-PDMC. 10 20 30 40 50 Diffraction intensity 2θ (°) chitosan chitosan-g-PDMC Fig. 3 – XRD patterns of chitosan and chitosan-g-PDMC. Table 1 – Levels of the variable tested in the 23 central composite designs. Variables Range and levels 2 1 0 1 2 X1, coagulant dosage (mg/L) 0 440 880 1320 1760 X2, flocculant dosage, PAM (mg/L) 0 10.0 20.0 30.0 40.0 X2, flocculant dosage, chitosan-g-PDMC (mg/L) 0 7.5 15.0 22.5 30.0 X3, pH 2.0 4.5 7.0 9.5 12.0 Fig. 4 – FESEM images of: (a) chitosan; and (b) chitosan-gPDMC particles. 5270 w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 5 2 6 7 – 5 2 7 5 3.1.4. Solubility of the flocculant The water-solubility of a flocculant is essential. Since chitosan is insoluble in water, it is usually dissolved into an acidic solution when it is used as a flocculant. At pH lower than 5.0, the acidic solution could accelerate the degradation of chitosan and thus reduce the flocculation ability. Table 2 summarizes the solubility of chitosan and chitosan-g-PDMC in 1% acetic acid and in water. All of them could be dissolved in 1% acetic acid. Chitosan could not be dissolved in water; the prepared flocculant, however, with an increase in grafting percentage, was initially insoluble, then became swollen, and ultimately soluble in water when the grafting percentage was 236.4% or above. Through grafting DMC onto chitosan backbone, the hydrophilicity of chitosan increased and then its water-solubility was improved. On the other hand, by grafting DMC onto chitosan backbone, the original inter-molecular and intramolecular hydrogen bondings were destroyed. This was confirmed by the XRD analysis as shown in Fig. 3. As a result, the water-solubility of the flocculant increased with the increasing grafting percentage. It became completely dissolved in water as the grafting percentage was greater than 236.4%. The solubility of the prepared flocculant with a grafting percentage of 236.4% was 1.08 g/(100 g H2O) at 293 K. Compared with another grafting polymer, i.e., chitosan-gPDMC (Wang et al., 2007a), the grafting polymer synthesized in the present study has a much better dissolvability in water. Chitosan-g-PDMC was initiated by gamma ray. In this case, the free radicals formed could cause serious degradation of the polymer, leading to a lower level of grafting percentage. On the other hand, the crystallinity of the grafting polymer initiated by gamma ray was reduced by a less degree than that of the grafting polymer initiated by potassium persulfate in this study. These two factors were likely to be responsible for the poor dissolvability in water for chitosan-g-PDMC. 3.2. Optimization for the flocculant synthesis In the flocculant preparation process, temperature, reaction time, initiator concentration and monomer concentration are important parameters influencing the grafting percentage (Sashiwa et al., 2001), which ultimately govern the flocculation ability of the copolymer. Thus, a L9 (3)4 orthogonal array design matrix was employed to evaluate the effects of these four factors on the grafting percentage. The range and level of these parameters are listed in Table A in Supplementary online materials, whereas the assignment of the experiments and the variance analysis are summarized in Table B in Supplementary online materials. In the light of variance analysis, temperature and monomer concentration were found as the most important factors that affected the grafting percentage. The optimal reaction conditions were found to be: temperature of 40 C, reaction time of 3 h, initiator concentration of 2.52 mmol/L and monomer concentration of 0.46 mol/L. Two additional parallel experiments were carried out under optimal reaction conditions to verify the optimization results. Under the optimal conditions, the grafting percentages obtained were 276.3% and 263.0%, respectively. This demonstrates that the L9 (3)4 orthogonal array matrix was appropriate for the optimization of the grafting reaction and flocculant synthesis. 3.3. Flocculation tests with kaolin suspension In this study, chitosan and the flocculant with grafting percentages of 46.5% and 144.8% were employed as examples for the flocculation tests. Fig. 5 illustrates their flocculation performance at various pHs. Under identical conditions, chitosan-g-PDMC with a higher grafting percentage always exhibited better flocculation ability. Table 2 – Solubility of chitosan-g-PDMC in water and acetic acid. Solvent Grafting percentage of samples (%) 0 (chitosan) 33.8 90.7 144.8 236.4 245.8 Acetic acid (1%) þ þ þ þ þ þ Water     þ þ þ, soluble; , partially soluble or high swollen; –, insoluble. Fig. 5 – Comparison of flocculation ability of chitosan and chitosan-g-PDMC using kaolin suspension at: (a) pH 4.0; (b) pH 7.0; and (c) pH 10.0. ---: chitosan; -C-: chitosan-gPDMC (grafting percentage 46.5%); -B-: chitosan-g-PDMC (grafting percentage 144.8%). w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 5 2 6 7 – 5 2 7 5 5271 Under acidic conditions, chitosan was positively charged as the amino groups bonded with hydrogen ions. However, the positive electric charge density of chitosan-g-PDMC was higher than that of chitosan because of the grafting of cationic monomer, DMC. Furthermore, chitosan-g-PDMC with a higher grafting percentage exhibited even better flocculation ability. At pH 7.0, a similar phenomenon was observed as that under acidic conditions. However, the grafting of DMC was relatively more importantincontributingto thepositivelychargedensity,asthe proton concentration was much lower in this case. On the other hand, under alkalineconditions, the transmittance of thekaolin suspension treated with chitosan-g-PDMC was high at the optimaldoses,whilechitosanhad negligibleflocculation ability. As mentioned above, because the amino groups on chitosan could bond with protons and thus chitosan became positively charged, chitosan had better flocculation ability under acidic conditions. Under alkaline conditions, however, there were few free hydrogen ions on chitosan and thus protonation could hardly be realized. As a result, almost no flocculation was observed when chitosan was used as a flocculant at pH 10.0 (Fig. 5c). Nonetheless, the cationic characteristics of the chitosan-g-PDMC, which originated from quaternary ammonium salts of grafted PDMC, were not influenced under alkaline conditions. Therefore, a significant flocculation could still be achieved under alkaline conditions. As seen from Fig. 5, the kaolin suspension transmittance initially increased, and then declined with an increase in flocculant dosage, regardless of pH and the type of flocculant. This might be attributed to two reasons. On the one hand, at a lower polymer concentration, its long chain adsorbed on the surface of one colloid particle may be adsorbed onto the surfaces of the other ones, and thus two or more particles aggregated together, resulting in flocculation through ‘‘bridging’’. However, when the polymer concentration was increased to a certain level, the adsorbed polymer would completely cover the particle surface and prevent them from flocculation. On the other hand, with an increase in flocculant dosage, the zeta potential of the particle gradually increased and the compression of electrical double layer was enhanced. When the zeta potential was increased up to zero, the optimal flocculation was achieved. After the flocculant dosage was further increased, the presence of the excessive flocculant will make the suspended particles positively charged, and thus cause mutual repulsion. In this case, a further increase in flocculant dosage would result in the re-dispersal of flocs and the reduction in transmittance. The sedimentation rates of the flocs from the suspension treated with chitosan and chitosan-g-PDMC are compared in Fig. B in Supplementary online materials. Under identical conditions, the kaolin flocs from the suspension treated by chitosan-g-PDMC with a higher grafting percentage settled more rapidly than that with a lower grafting percentage. This might be attributed to the fact that a higher positive charge was in favor of the compression of the electrical double layer and thus the formation of denser and larger kaolin flocs. However, forthesuspensiontreatedwithchitosan,evennoclearinterface between water and solid was observed under alkaline conditions, as chitosan had negligible flocculation ability in this case. The images of the flocs obtained from the suspensions respectively treated with chitosan-g-PDMC (grafting percentage 144.8%), chitosan and without flocculant are shown in Fig. 6. Dense and large flocs from the kaolin suspension treated with chitosan-g-PDMC are displayed in Fig. 6c, compared with those treated by chitosan in Fig. 6b, and the raw kaolin particles without flocculant dose in Fig.6a. These physical characteristics of the flocs were in well accordance with their corresponding sedimentation process as shown in Fig. B. Fig. 6 – Images of flocs from kaolin suspension treated with: (a) no flocculant; (b) chitosan; and (c) chitosan-gPDMC. 5272 w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 5 2 6 7 – 5 2 7 5 3.4. Treatment of pulp mill wastewater The experimental results of the pulp mill wastewater treatment with PAM and the prepared flocculant, chitosan-g-PDMC, are listed in Tables 3 and 4 respectively. The optimization of the four responses was individually achieved under different optimal conditions. Thus, a compromise among the conditions for the four responses is desirable. This could be achieved using the desirability function approach (Wang et al., 2007b). The regression functions for PAM and chitosan-g-PDMC as flocculants were obtained by Eqs. (2) and (3) respectively. Y ¼ 78:90 þ 1:40X1  1:16X2 þ 0:44X3  7:19X2 1  2:91X2 2 5:36X2 3  0:52X1X2  4:51X1X3  3:36X2X3 R2 ¼ 0:882; F ¼ 8:47 (2) Y ¼ 89:13 þ 5:82X1 þ 3:50X2  0:99X3  8:87X2 1  4:27X2 2 6:83X2 3  1:05X1X2 þ 2:88X1X3 þ 1:95X2X3 R2 ¼ 0:951; F ¼ 22:95 (3) The quadratic regression results show that the two models were significant because the values of Fstatistic (the ratio of mean square due to regression to mean square to Table 3 – CCD and response results for the study of three experimental variables in coded unites when PAM was used as the flocculant. Run no. X1 X2 X3 Turbidity removal (%) Lignin removal (%) COD removal (%) Water recovery (%) 1 0 0 0 94.8 72.3 73.7 80.0 2 0 0 0 95.2 72.8 72.8 80.0 3 1 1 1 87.0 35.1 52.4 80.0 4 1 1 1 87.8 40.6 50.5 58.0 5 1 1 1 91.5 42.3 56.2 80.0 6 1 1 1 89.7 40.4 48.6 80.0 7 0 2 0 91.5 71.9 60.0 60.0 8 2 0 0 35.9 19.9 69.3 60.0 9 1 1 1 84.8 46.9 79.0 50.0 10 1 1 1 79.9 70.4 52.3 65.0 11 1 1 1 85.2 60.3 78.5 75.0 12 1 1 1 83.3 31.2 62.2 40.0 13 2 0 0 84.3 42.0 63.5 64.0 14 0 2 0 91.7 72.7 52.2 60.0 15 0 0 2 86.1 39.8 52.7 62.0 16 0 0 2 84.6 41.8 47.8 76.0 17 0 0 0 95.0 71.1 75.1 80.0 18 0 0 0 94.9 71.1 74.3 80.0 19 0 0 0 94.5 71.5 75.8 75.0 20 0 0 0 94.7 72.4 74.5 75.0 Table 4 – CCD and response results for the study of three experimental variables in coded unites when chitosan-g-PDMC was used as the flocculant. Run no. X1 X2 X3 Turbidity removal (%) Lignin removal (%) COD removal (%) Water recovery (%) 1 0 0 0 98.0 76.7 93.5 86.0 2 0 0 0 99.2 78.9 93.0 84.0 3 1 1 1 96.9 52.8 54.4 84.0 4 1 1 1 85.9 34.4 79.2 94.0 5 1 1 1 96.1 52.8 67.3 92.0 6 1 1 1 89.3 38.8 63.4 94.0 7 0 2 0 85.0 49.7 70.2 60.0 8 2 0 0 68.4 26.6 51.5 20.0 9 1 1 1 99.1 70.5 65.5 70.0 10 1 1 1 97.7 54.3 69.4 80.0 11 1 1 1 79.5 67.1 53.9 30.0 12 1 1 1 96.4 79.4 72.9 73.0 13 2 0 0 96.9 59.3 52.5 70.0 14 0 2 0 98.1 64.1 68.6 80.0 15 0 0 2 81.8 19.0 75.1 84.0 16 0 0 2 89.3 38.0 59.1 90.0 17 0 0 0 98.8 79.7 93.5 88.0 18 0 0 0 98.4 78.9 93.0 84.0 19 0 0 0 99.1 76.7 92.9 86.0 20 0 0 0 98.7 79.9 92.4 88.0 w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 5 2 6 7 – 5 2 7 5 5273 real error) of 8.47 and 22.95 were greater than F0.001,9,10 (4.94). From the regression of Eq. (2), the optimal conditions for PAM as flocculant were obtained as follows: coagulant dosage of 915 mg/L, flocculant dosage of 17.4 mg/L, and pH 7.2. Under the optimal conditions, the removal efficiencies of turbidity, lignin and COD, as well as water recovery were estimated to be 95.7%, 70.6%, 75.0% and 77.2%, respectively, from their individual regression equations. The optimal conditions for chitosan-g-PDMC as flocculant obtained from Eq. (3) were: coagulant dosage of 1017 mg/L, flocculant dosage of 17.8 mg/L, and pH 7.1. The optimal conditions for PAM and chitosan-g-PDMC as flocculants were similar. However, when chitosan-g-PDMC was used as a flocculant, the estimated removal efficiencies of turbidity, lignin and COD, as well as water recovery of 99.4%, 81.3%, 90.7% and 89.4%, respectively, were better than the corresponding values for PAM. The grafting of cationic polymer chain onto chitosan backbone was in favor of the charge neutralization, compression of the double electric layer, and improvement of the sweep-floc effects. As a result, its flocculation ability was better than that of PAM, especially in terms of the removal efficiency of turbidity and water recovery. These results indicate that the flocculant prepared in this study was superior to PAM from the application point of view. To confirm the validity of the statistical experimental strategies, two additional confirmation experiments were conducted under each of the operating conditions. The chosen conditions for coagulant dosage, flocculant dosage and pH were the optimal conditions according to their respective desirability function regression equations. The four responses measured were close to those estimated using RSM. This also verifies that the RSM approach was appropriate for optimizing the operational conditions for the coagulation–flocculation treatment of pulp mill wastewater. 4. Conclusions Chitosan-g-PDMC, a novel flocculant with a high watersolubility, was prepared through grafting DMC onto chitosan backbone initiated by potassium persulphate in an acidic water solution. The characterization using various instruments demonstrates that DMC was successfully grafted onto chitosan backbone. Temperature and monomer concentration were found to be the most important factors influencing the grafting percentage. In addition, the watersolubility of chitosan-g-PDMC increased with the increasing grafting percentage, and became completely dissolved in water as the grafting percentage exceeded 236.4%. When the flocculant prepared was applicated for the treatment of pulp mill wastewater, it had excellent flocculation capacity and its flocculation efficiency was much better than that of polyacrylamide in terms of the removal efficiencies of turbidity, lignin and COD. The optimal conditions for the flocculation treatment of pulp mill wastewater were: coagulant dosage of 1017 mg/L, flocculant dosage of 17.8 mg/L, and pH 7.1. Acknowledgements We wish to thank the Key Special Program on the S&T for the Pollution Control and Treatment of Water Bodies (2008ZX07103-001 and 2008ZX07010-003) for the partial support of this study. Appendix. Supplementary material Supplementary material associated with this article can be found in the online version at doi:10.1016/j.waters.2009.08.040 r e f e r e n c e s Ahmad, A.L., Ismail, S., Bhatia, S., 2005. Optimization of coagulation–flocculation process for palm oil mill effluent using response surface methodology. Environ. Sci. Technol. 39 (8), 2828–2834. Ashmore, M., Hearn, J., 2000. Flocculation of model latex particles by chitosans of varying degrees of acetylation. Langmuir 16 (11), 4906–4911. Bratskaya, S., Schwarz, S., Chervonetsky, D., 2004. Comparative study of humic acids flocculation with chitosan hydrochloride and chitosan glutamate. Water Res. 38 (12), 2955–2961. Chang, L.L., Bruch, M.D., Griskowitz, N.J., Dentel, S.K., 2002. NMR spectroscopy for determination of cationic polymer concentrations. Water Res. 36 (9), 2255–2264. de Vasconcelos, C.L., Bezerril, P.M., Dantas, T.N.C., Pereira, M.R., Fonseca, J.L.C., 2007. Adsorption of bovine serum albumin on template-polymerized chitosan/poly(methacrylic acid) complexes. Langmuir 23 (14), 7687–7694. Divakaran, R., Pillai, V.N.S., 2004. Mechanism of kaolinite and titanium dioxide flocculation using chitosandassistance by fulvic acids? Water Res. 38 (8), 2135–2143. Franks, G.V., 2005. Stimulant sensitive flocculation and consolidation for improved solid/liquid separation. J. Colloid Interface Sci. 292 (2), 598–603. Jenkins, D.W., Hudson, S.M., 2002. Heterogeneous graft copolymerization of chitosan powder with methyl acrylate using trichloroacetyl-manganese carbonyl co-initiation. Macromolecules 35 (9), 3413–3419. Lazaridis, N.K., Kyzas, G.Z., Vassiliou, A.A., Bikiaris, D.N., 2007. Chitosan derivatives as biosorbents for basic dyes. Langmuir 23 (14), 7634–7643. Lim, S.F., Zheng, Y.M., Zou, S.W., Chen, J.P., 2008. Characterization of copper adsorption onto an alginate encapsulated magnetic sorbent by a combined FT-IR, XPS, and mathematical modeling study. Environ. Sci. Technol. 42 (7), 2551–2556. Li, H.B., Du, Y.M., Wu, X.J., Zhan, H.Y., 2004. Effect of molecular weight and degree of substitution of quaternary chitosan on its adsorption and flocculation properties for potential retentionaids in alkaline papermaking. Colloids Surf. A 242 (1–3), 1–8. Neira-Carrillo, A., Yazdani-Pedram, M., Retuert, J., Diaz-Dosque, M., Gallois, S., Arias, J.L., 2005. Selective crystallization of calcium salts by poly(acrylate)-grafted chitosan. J. Colloid Interface Sci. 286 (1), 134–141. Rath, S.K., Singh, R.P., 1997. Flocculation characteristics of grafted and ungrafted starch, amylose, and amylopectin. J. Appl. Polym. Sci. 66 (9), 1721–1729. Sashiwa, H., Shigemasa, Y., Roy, R., 2001. Chemical modification of chitosan. 10. Synthesis of dendronized chitosan–sialic acid 5274 w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 5 2 6 7 – 5 2 7 5 hybrid using convergent crafting of preassembled dendrons built on gallic acid and tri(ethylene glycol) backbone. Macromolecules 34 (12), 3905–3909. Singh, R.P., 1995. In: Prasas, P.N., Mark, J.F., Fai, T.J. (Eds.), Polymers and Other Advanced Materials: Emerging Technologies and Business Opportunities. Plenum Press, New York. Wang, J.P., Chen, Y.Z., Ge, X.W., Yu, H.Q., 2007a. Gamma radiation-induced grafting of a cationic monomer onto chitosan as a flocculant. Chemosphere 66 (9), 1752–1757. Wang, J.P., Chen, Y.Z., Ge, X.W., Yu, H.Q., 2007b. Optimization of coagulation–flocculation process for a paper-recycling wastewater treatment using response surface methodology. Colloids Surf. A 302 (1–3), 204–210. Wang, J.P., Chen, Y.Z., Zhang, S.J., Yu, H.Q., 2008. A chitosanbased flocculant prepared with gamma-irradiation-induced grafting. Bioresour. Technol. 99 (9), 3397–3402. Xiao, H.N., Pelton, R., Hamielec, A., 1995. Flocculation of polystyrene latex by polyacrylamide-copolyethylene glycol. J. Colloid Interface Sci. 175 (1), 166–172. Yang, W.Y., Qian, J.W., Shen, Z.Q., 2004. A novel flocculant of AI(OH)3-polyacrylamide ionic hybrid. J. Colloid Interface Sci. 273 (2), 400–405. Zhou, Y., Liedberg, B., Gorochovceva, N.A., Makuska, R., Dedinaite, A., Claesson, P.M., 2007. Chitosan-N-poly(ethylene oxide) brush polymers for reduced nonspecific protein adsorption. J. Colloid Interface Sci. 305 (1), 62–71. w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 5 2 6 7 – 5 2 7 5 5275