- Email: [email protected]
Assessment on the removal of dimethyl phthalate from aqueous phase using a hydrophilic hyper-cross-linked polymer resin NDA-702
Journal of Colloid and Interface Science 311 (2007) 382–390 www.elsevier.com/locate/jcis
Assessment on the removal of dimethyl phthalate from aqueous phase using a hydrophilic hyper-cross-linked polymer resin NDA-702 Weiming Zhang a,b,∗ , Zhengwen Xu a , Bingcai Pan a,∗ , Lu Lv c , Qingjian Zhang a , Qingrui Zhang a , Wei Du a , Bingjun Pan a , Quanxing Zhang a,b a State Key Laboratory of Pollution Control and Resources Reuse, School of the Environment, Nanjing University, Nanjing 210093, People’s Republic of China b Research Center for Engineering Technology of Organic Pollutants Control and Resources Reuse of Jiangsu Province, Nanjing 210038,
People’s Republic of China c Department of Chemical and Biomolecular Engineering, National University of Singapore, 10 Kent Ridge Crescent, 119260 Singapore
Received 8 January 2007; accepted 7 March 2007 Available online 12 April 2007
Abstract A hydrophilic hyper-cross-linked polymer resin (NDA-702) was synthesized, and the adsorption performance of dimethyl phthalate (DMP) on NDA-702 was compared with that on the commercial hydrophobic macroporous resin (Amberlite XAD-4) and granular activated carbon (AC750). The kinetic adsorption of DMP onto NDA-702 and AC-750 is limited mainly by intraparticle diffusion and obeys the pseudo-second-order rate model, while the uptake on XAD-4 is limited mainly by film diffusion and follows the pseudo-first-order rate model. All the associated adsorption isotherms are well described by the Freundlich equation, and the larger uptake and stronger affinity of NDA-702 than AC-750 and XAD-4 probably result from the microporous structure, phenyl rings, and polar groups on NDA-702 polymer matrix. An interesting observation is that in the aqueous phase all the adsorbents spontaneously adsorb DMP driven mainly by enthalpy change, but the hydrophilic nature of NDA-702 and AC-750 surfaces results in less entropy change compared to hydrophobic XAD-4. Dynamic adsorption studies show that the high breakthrough and the total adsorption capacities of NDA-702 are 388 and 559 mg per gram dry resin at 313 K. Nearly 100% regeneration efficiency for the resin was achieved by methanol at 313 K. © 2007 Elsevier Inc. All rights reserved. Keywords: Removal; DMP; Porous materials; Adsorption; Kinetics; Isotherms; Dynamics
1. Introduction Phthalates are a group of organic compounds widely used as plasticizers to improve the flexibility and durability of polyvinyl chloride (PVC)-based plastics. They are also used as additives in the fabrication of multilayer ceramic capacitors, paper coatings, cosmetics, inks, and paints. Since they are not chemically bound to the PVC polymer, a major part of the phthalates entering the environment is suspected to come from phthalatecontaining products during use or after disposal [1–3]. In addition, several industrial plants produce large amounts of waste* Corresponding authors. Fax: +86 25 83707304.
E-mail addresses: [email protected] (W. Zhang), [email protected] (B. Pan). 0021-9797/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2007.03.005
water containing a high concentration of phthalates which have negative effects on the water environment [4]. There are six phthalates included in the USEPA priority list because of their endocrine disrupting, carcinogenic, or xenoestrogenic properties [5,6]. How to effectively and economically remove these phthalates from the environment is critically important and interesting. Some kinds of removal methods for phthalates, such as biodegradation, coagulation, and adsorption, have been reported to date. The bioconversion of phthalates under both aerobic and anaerobic conditions has been investigated [7,8], and their biodegradation by activated sludge has also been demonstrated [9]. However, those methods require a long time to render the phthalates harmless, and microorganisms could hardly degrade them completely or remove them completely
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
from aqueous solution. Although coagulation including flocculation is useful for the removal of organic micropollutants and its removal mechanism has been reported [10], coagulation by ferric chloride was not effective. On the other hand, adsorptive removal by activated carbon and β-cyclodextrin was effective [11,12]. Usually the activated carbon used in water treatment is treated by a heat procedure or steam for recovery; therefore, it is difficult to reuse the materials adsorbed on the activated carbon. And according to its weak stability, cyclodextrin is not used widely as an adsorbent at industrial scale. In recent years, synthetic polymeric resins have been widely applied in industrial effluent treatment in China due to its efficient removal and feasible regeneration for repeated use, as well as the potential recovery of some valuable organic compounds from water [13,14]. For example, the commercially available Amberlite XAD-4 resin was reported as an ideal adsorbent for a wide variety of aromatic compounds. However, Amberlite XAD-4 resin has a hydrophobic surface and a lower adsorption capacity for most organic substances compared with activated carbon, which limits its wide application in the treatment of contaminated water [15]. Recently, some chemically modified polymeric resins have been synthesized and applied in organic pollution control for industrial waste streams in our previous studies [16,17]. Nevertheless, less work focusing on the removal of phthalates using polymer resins was reported [18]. Our understanding of factors controlling the adsorption processes of phthalates on the surface of resins is still limited. The primary objectives of this investigation are to synthesize a new hydrophilic hyper-cross-linked polymer resin (NDA-702) with high capacity and affinity for phthalates, and evaluate the removal and recovery efficiency of phthalates from water using NDA-702. Batch kinetic and equilibrium adsorption experiments were conducted, and thermodynamic analysis was used to elucidate the adsorption mechanism of phthalates on NDA702. Methanol was used to evaluate the regeneration efficiency of NDA-702 in fixed-bed column runs. In this study, NDA-702 (functional hyper-cross-linked polymer resin), XAD-4 (macroporous polymer resin), and AC-750 (granular activated carbon) were selected since they represent three typical kinds of widespread adsorbents in pollution control. Dimethyl phthalate (DMP) was selected as a target compound that represents the low alkyl chain phthalates, which are characterized by high solubility in water [1]. Furthermore, previous studies [19] showed that DMP and other low alkyl chain phthalates could be more toxic to microorganisms during the biotreatment process than the high alkyl chain phthalates. 2. Experimental 2.1. Materials CHA-101, a macroreticular polymer, was kindly provided by Langfang Electrical Resin Co. Ltd. (Hebei province, China). The spherical macroporous Amberlite XAD-4 resin was purchased from Rohm & Haas Company (Philadelphia, PA). Gran-
383
Scheme 1. Synthesis procedures for NDA-702.
ular activated carbon AC-702 is of analytical grade and was purchased from Shanghai reagent station (Shanghai, China). Acetone, methanol, iron chloride, and DMP are of analytical grade and were also purchased from Shanghai reagent station and used without further purification. The background solution used in the adsorption experiments contained 10−3 M KH2 PO4 for buffering the aqueous solutions at neutral pH. 2.2. Synthesis of NDA-702 resin CHA-101 is a macroreticular polystyrene adsorbent with some residual chlormethyl groups on its polymeric matrix during the synthetic process [16], and Friedel–Crafts reaction [20] can take place by the residual chloro atom. Procedures for synthesis of the hydrophilic hyper-cross-linked polymer resin NDA-702 are given in Scheme 1. The specific surface area and the pore distribution of the three adsorbents were respectively measured by BET and BJH methods using a Micrometics ASAP-2010 automatic surface area analysis instrument (Micromeritics Instrument, Norcros, USA). Infrared spectra of the adsorbents in the range of 650– 4000 cm−1 were collected by a Nexus 870 FT-IR spectrometer (USA) with a pellet of powdered potassium bromide. The chlorine content was measured according to the method of Volhard [21]. 2.3. Batch sorption experiments In the following experiments, hydrophilic NDA-702 and AC-750 can be used directly, while hydrophobic XAD-4 should first be wetted with 0.5 ml methanol and then rinsed three times with deionized water. Kinetics of adsorption was determined by analyzing the adsorptive uptake of DMP from its aqueous solution of required concentration 500 mg L−1 at different time intervals until the equilibrium is reached at 313 K. Equilibrium adsorption experiments were carried out at 283, 298, and 313 K in 100-ml glass flasks. All flasks contained a fixed mass of the adsorbent and a known volume of initial aqueous solutions with the desired different initial concentration (50, 100, 200, 300, 400, and 500 mg L−1 , respectively). The flasks were then transferred to a Model G 25 incubator shaker with thermostat (New Brunswick Scientific Co. Inc.) and shaken under 150 rpm until the adsorption process reached equilibrium.
384
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
For all the above systems, control experiments were conducted using flasks prepared similarly but containing no adsorbent for assessing loss of solutes to the flask components during adsorption tests. Results of triplicate flasks showed that the average solution phase concentrations were within 96–102% of the respective initial concentration of the same solution analyzed similarly. Hence, compound loss was negligible. 2.4. Fixed-bed column tests Dynamic adsorption was conducted using a 10 cm × 4.6 mm i.d. stainless-steel column packed with the newly synthesized hydrophilic adsorbent NDA-702, AC-750, and XAD-4, respectively (1.00 g dry adsorbent), and connecting with a 6672 reciprocating pump (Beijing Analytical Instrument Plant) at 313 K. DMP solution of 500 mg L−1 was passed through the column with the superficial liquid velocity (SLV) and the empty bed contact time (EBCT) equal to 0.30 m h−1 and 6 min, respectively. The effluents from the column were quantitatively analyzed. 2.5. Analysis All the solution concentrations of DMP were analyzed spectrophotometrically using a Helious Betra UV–vis spectrophotometer (UK) at a wavelength of 228 nm. 3. Results and discussion 3.1. Characterization of adsorbents Table 1 presents the typical properties of NDA-702, AC-750, and XAD-4. These adsorbents show a similar BET surface area, but quite different chemical composition and pore structures. The cross-linked pore structures of XAD-4 and NDA-702 are depicted in Fig. 1. The pore-size distribution plots in Fig. 2 indicate that the micropore (<2 nm) dominates the pore structure of NDA-702 and AC-702, while mesopore (2–18 nm) plays the main role in the XAD-4 resin. The strong bands of 1700 cm−1 (carbonyl group) and 3650 cm−1 (hydroxyl group) in the IR spectrum of NDA-702 (in Fig. 3) indicate the existence of polar groups on its polymer matrix.
3.2. Adsorption kinetics The influence of contact time on the removal of DMP by NDA-702, AC-750, and XAD-4 is presented in Fig. 4. A comparison made among them indicates that different adsorbents require different times for attaining equilibrium with different extents of uptake. In the case of XAD-4, with increase in shaking time, percentage uptake was found to increase gradually, approaching a maximum at around 100 min. For NDA-702 and AC-750, the required equilibrium times were shifted to higher values, 600 and 720 min, respectively. The faster rate of approach to equilibrium onto XAD-4 is reasonably compared with NDA-702 and AC-750 due to their abundant micropores. The kinetic adsorption data were processed to understand the dynamics of the adsorption process in terms of the order of the rate constant. The pseudo-first-order and pseudo-second-order models [22] were attempted to fit all the data: pseudo-first-order model: k1 t, 2.303 1 1 t = + t, pseudo-second-order model: qt k2 qe2 qe log(qe − qt ) = log(qe ) −
Sorbent
NDA-702
AC-750
XAD-4
Polarity BET surface area (m2 g−1 ) Micropore area (m2 g−1 ) Micropore volume (cm3 g−1 ) Average pore diameter (nm) Particle size (mm) Residual chloride content (%) Oxygen content (%)
Moderate polar 804 481 0.22 2.9 0.9–1.1 2.7 2.9
Polar 749 471 0.22 2.2 0.4–0.6 0 1.4
Nonpolar 868 54 0.004 5.6 0.5–0.7 0 0
(2)
where qe is the equilibrium adsorption capacity (mg g−1 ), qt is the adsorption capacity at contact time t (mg g−1 ), k1 is the pseudo-first-order rate constant (min−1 ), and k2 is the pseudosecond-order rate constant (g mg−1 min−1 ). The fitting results (in Fig. 4 and Table 2) show that the uptake of DMP on macroporous adsorbent XAD-4 follows the pseudo-first-order rate equation, while the uptake on microporous adsorbents NDA702 and AC-750 obeys the pseudo-second-order rate equation. Similarly, it can also be seen that the adsorption rate order of DMP onto these adsorbents is XAD-4 > AC-750 > NDA-702. The different uptake rates between AC-750 and NDA-702 show that not the average pore diameter but the particle size of the two adsorbents (in Table 1) may play a dominant role. In general, the adsorption process may be described as a series of steps: mass transfer from fluid phase to the particle surface across the boundary layer, diffusion within the porous particle, and adsorption itself onto the surface [23]. To determine which one (film diffusion or intraparticle diffusion) is rate limiting and also to find its rate parameters, adsorption kinetic data were further processed. The two diffusion models can be written as [24]: ln(1 − F ) = −kf t,
(3)
intraparticle diffusion model: qt = kp t 1/2 ,
(4)
film diffusion model:
Table 1 Main properties of sorbents
(1)
where kf is the diffusion rate parameter for film diffusion model, kp is the diffusion rate parameter for intraparticle diffusion model, and F is the fractional attainment of the equilibrium. The fitting results of Eqs. (3) and (4) listed in Figs. 5 and 6 show that for macroporous resin XAD-4, the film diffusion model yields better fit to the adsorption kinetic data; while, for microporous adsorbents NDA-702 and AC-750, the intraparticle diffusion model gives a perfect fit. However, all the straight
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
385
Fig. 1. Pore structures of cross-linked polymer resins, XAD-4 and NDA-702.
lines do not pass through the origin. It indicates that both diffusions are the rate-limiting steps for DMP adsorption [22], and the adsorption process on XAD-4 is controlled mainly by film diffusion while the adsorption processes on NDA-702 and AC750 are limited mainly by intraparticle diffusion. Furthermore, it can also be seen in Figs. 5 and 6 that DMP molecules diffuse more rapidly onto the XAD-4 surface than on NDA-702 and AC-750 surfaces. It is probably another creditable factor for the greater adsorption rate of DMP onto XAD-4.
3.3. Adsorption isotherms and thermodynamics Equilibrium data concerning the adsorption of DMP onto the test three adsorbents are presented in Fig. 7. Langmuir and Freundlich [25,26] parameters obtained by nonlinear regression of the experimental data with the Levenberg–Marquardt algorithm are listed in Table 3. The Freundlich model yields a better correlation for all the equilibrium adsorption curves. Fig. 7 and Table 3 show a significantly higher adsorbing ca-
386
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
Fig. 4. Effect of contact time on DMP removal by NDA-702, AC-750, and XAD-4 at 313 K. Table 2 Kinetic parameters of 500 mg L−1 DMP onto NDA-702, AC-750, and XAD-4 at 313 K Adsorbent NDA-702 AC-750 XAD-4
Pseudo-first-order model
Pseudo-second-order model
k1 (min−1 )
R2
k2 (g kg−1 min−1 )
0.932 0.907 0.992
50.4 207.9 307.4
0.032 0.080 0.155
±0.004 ±0.010 ±0.006
±7.62 ±20.03 ±30.28
R2 0.964 0.976 0.979
Fig. 2. Plot of pore-size distributions vs pore diameter for NDA-702, AC-750, and XAD-4.
Fig. 3. IR spectra of NDA-702, AC-750, and XAD-4.
Fig. 5. The straight lines of ln(1–F ) vs the contact time t from DMP adsorbed onto NDA-702, AC-750, and XAD-4 at 313 K.
pacity (KF ) and affinity (n) for DMP on NDA-702 than AC750 and XAD-4. It is probably attributed to three reasons: (1) the micropore (<2 nm) dominates the pore structure of NDA-702 and the micropore area of NDA-702 resin is much higher than that of XAD-4; (2) the polar carbonyl and hy-
droxyl groups on NDA-702 polymer matrix are also helpful for the higher adsorption affinity of the ester group on DMP molecules by the proposed hydrogen-bonding interaction than XAD-4; (3) the hydrophobic phenyl rings on NDA-702 polymer matrix also enhance the adsorption affinity of the aromatic
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
387
Fig. 6. The straight lines of qt vs t1/2 from DMP adsorbed onto NDA-702, AC-750, and XAD-4 at 313 K.
ring on DMP molecules by the proposed π–π interaction than AC-750. The adsorptive free energy change at equilibrium can be calculated with the Gibbs equation G = −nRT ,
(5)
where n is the characteristic constant in Freundlich equation, R is the universal gas constant (8.314 kJ mol−1 K−1 ), and T is the temperature in K. The isosteric adsorptive enthalpy changes with temperature expressed as [27] ln Ce = ln K + (−H /RT ),
(6)
where H is the isosteric adsorptive enthalpy change, and K is a constant. H was calculated from the slope of line plotted by ln Ce vs 1/T (in Fig. 8). The adsorptive entropy change was calculated with the Gibbs–Helmholtz equation [28] S = (H − G)/T .
(7)
Fig. 8 also shows the calculated thermodynamic parameters of DMP adsorption onto NDA-702, AC-750, and XAD-4. The negative value of G indicates that all the test adsorption processes are favorable and spontaneous. Compared to AC-750 and XAD-4, the largest absolute value of G for NDA-702 suggests the strongest adsorptive affinity of DMP to NDA-702. This is in agreement with their adsorption capacities. The negative values of the enthalpy changes indicate an exothermic adsorption process of DMP onto all the three adsorbents. In addition, the same negative values of the entropy changes present a regulation of molecules random movement. Both of them suggest that the adsorption processes of DMP onto all the three adsorbents were driven mainly by enthalpy change. Because in aqueous solution water molecules easily
Fig. 7. Adsorption isotherms of DMP on NDA-702, AC-750, and XAD-4 at different temperatures.
wet hydrophilic surfaces of AC-750 and NDA-702, DMP molecules must replace the water molecules and then they could be adsorbed on these adsorbent surfaces (in Fig. 9). The action is called “solvent replacement” [29]. For the greater hydrophilic property of the AC-750 surface (in Table 1) and the larger num-
388
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
Table 3 Isotherm parameters of DMP adsorption on the three sorbents at different temperatures Sorbent
T (K)
Freundlich
Langmuir
NDA-702
283 298 313
193.2 164.8 140.6
±0.801 ±0.570 ±0.003
5.778 5.203 4.900
AC-750
283 298 313
103.9 82.8 70.6
±0.351 ±0.279 ±0.972
XAD-4
283 298 313
101.3 66.4 48.0
±0.160 ±0.107 ±0.619
KF
R2
KL (L mg−1 )
±0.249 ±0.226 ±0.211
0.997 0.996 0.996
0.551 0.265 0.156
±0.362 ±0.152 ±0.085
447.7 436.3 411.5
±6.92 ±3.88 ±2.53
0.854 0.878 0.879
5.402 4.928 4.665
±0.130 ±0.187 ±0.236
0.999 0.997 0.994
0.129 0.160 0.113
±0.097 ±0.113 ±0.069
286.8 247.6 230.3
±7.15 ±1.10 ±8.98
0.762 0.767 0.784
3.864 3.210 2.912
±0.154 ±0.156 ±0.089
0.997 0.995 0.998
0.061 0.035 0.025
±0.031 ±0.015 ±0.009
424.7 398.5 365.0
±9.27 ±6.34 ±1.37
0.891 0.919 0.937
n
qm (mg g−1 )
R2
capacity and the total adsorption capacity are respectively 388 and 559 mg per gram dry resin, which are much larger than XAD-4 (378 and 430 mg per gram dry resin) and AC-750 (215 and 263 mg per gram dry resin). In addition, their breakthrough adsorption capacities are much larger than 1.29 mg per gram of β-cyclodextrin [12]. The organic solvent methanol was used to desorb DMP from the NDA-702 resin column. At the same flow rate as adsorption, SLV and EBCT equal to 0.30 m h−1 and 6 min, nearly 100% regeneration efficiency was achieved at 313 K (in Fig. 10). By further distilling, both DMP and methanol can be easily recovered. 4. Conclusion
Fig. 8. van’t Hoff plots (ln Ce versus 1/T ) for different types of isotherms presented in Fig. 5 (Qe = 1.0 mmol g−1 of adsorbent, T = 283 K).
ber of water molecules attached on the AC-750 surface than the NDA-702 surface, the number of water molecules on AC-750 replaced by DMP molecules was larger than that on NDA-702. So the solvent replacement process alters the entropy change order as (in Fig. 8) AC-750 > NDA-702 > XAD-4. 3.4. Dynamic adsorption and desorption Due to the satisfactory adsorption capacity and affinity for DMP on NDA-702 in our research, it is hopeful that NDA-702 will be developed as a polymeric adsorbent for the removal and recovery of phthalate pollutants from drinking water or industrial wastewater. It is necessary to test the dynamic adsorption and desorption. The results of minicolumn dynamic adsorption of DMP on NDA-702 are shown in Fig. 10, where C0 is the concentration of initial solution (mM), Cv is the concentration at different bed volumes of the effluent (mg L−1 ), and Cd is the concentration at different bed volumes of regenerative reagent (mg L−1 ). Breakthrough adsorption capacity and the total capacity were calculated on the basis of the total amount of DMP removed when the concentration of the effluent from the column reached 10% and nearly 100% of the initial concentration, respectively. The results of breakthrough adsorption
A new hydrophilic hyper-cross-linked resin NDA-702 for removing DMP from aqueous solution was synthesized. The batch kinetic studies show that the uptake of DMP onto NDA702 and AC-750 is limited mainly by intraparticle diffusion and obeys the pseudo-second-order rate model, while the uptake on XAD-4 is limited mainly by film diffusion and follows the pseudo-first-order rate model. The equilibrium adsorption capacity and affinity of NDA702 to DMP are markedly higher than those of XAD-4 and AC-750, which is attributed to its microporous structure, phenyl rings, and the polar functional groups. All the isotherm data can be fitted with the Freundlich equation satisfactorily. The spontaneous adsorption of DMP onto all the three adsorbents was driven mainly by enthalpy change. The experimental results of dynamic adsorption toward DMP by NDA-702 show that the high breakthrough capacity and the total capacity are 388 and 559 mg per gram dry resin. A 100% regeneration efficiency was achieved using methanol. By further distilling, both DMP and methanol can be easily recovered. Acknowledgments This study was funded by the National Nature Science Foundation of PR China (Grant 20504012) and the Nature Science Foundation of Jiangsu Province (Grant BK2006129).
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
389
Fig. 9. Schemes for the proposed adsorption process of DMP onto the hydrophobic surface of XAD-4 and the hydrophilic surface of NDA-702 and AC-750 in aqueous solution. [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] Fig. 10. Adsorption and desorption dynamic curves of DMP onto NDA-702 at 313 K.
[17]
References
[18]
[1] C. Staples, D. Peterson, T. Parkerton, W. Adams, Chemosphere 35 (1997) 667–749. [2] M. Vitali, M. Guidotti, G. Macilenti, C. Cremisini, Environ. Int. 23 (3) (1997) 337–347. [3] M. Bauer, R. Herrmann, A. Martin, H. Zellmann, Water Sci. Technol. 38 (2) (1998) 185–192.
[19] [20] [21]
J. Wang, K. Xu, L. Xu, Huanjing Kexue 9 (4) (1988) 46–52. C.S. Giam, H.S. Chan, G.S. Neff, E.L. Atlas, Science 199 (1978) 419–421. F.R. Knudsen, T.G. Pottinger, Aquat. Toxicol. 44 (1999) 159–170. R. Aftring, B. Chalker, B. Taylor, Appl. Environ. Microbiol. 41 (1981) 1177–1183. J. Ejlertsson, M. Alnervik, S. Jonsson, B. Svensson, Environ. Sci. Technol. 31 (1997) 2761–2764. R. Iturbe, G. Moreno, P. Elefsiniotis, Environ. Technol. 12 (1991) 783. P. Thebault, J.M. Cases, F. Fiessinger, Water Res. 15 (1981) 183. N. Adhoum, L. Monser, Sep. Purif. Technol. 38 (2004) 233–239. S. Murai, S. Imajo, Y. Takasu, K. Takahashi, K. Hattori, Environ. Sci. Technol. 32 (1998) 782–787. P. Li, A.K. SenGupta, Environ. Sci. Technol. 32 (1998) 3756–3766. Q.X. Zhang, in: K.D. Xu (Ed.), Science and Technology at the Frontier in China, Higher Education Press, Beijing, 2005, pp. 565–592. N. Masque, M. Galia, J. Chromatogr. A 771 (1997) 55–61. B.C. Pan, Q.X. Zhang, F.W. Meng, X.T. Li, X. Zhang, J.Z. Zheng, W.M. Zhang, B.J. Pan, J.L. Chen, Environ. Sci. Technol. 39 (2005) 3308–3313. W.M. Zhang, Q.J. Zhang, B.C. Pan, L. Lv, B.J. Pan, Z.W. Xu, Q.X. Zhang, X.S. Zhao, W. Du, Q.R. Zhang, J. Colloid Interface Sci. 3006 (2007) 216– 221. D. Zhao, A.K. SenGupta, in: Hazardous and Industrial Waster—Proceedings of the Mid-Atlantic Industrial Waste Conference, Lancaster, PA, 1996, pp. 63–70. N. Nielsen, P.B. Larsen, Environmental Review No. 6, Danish Environmental Protection Agency, Denmark, 1996. V.A. Davankov, M.P. Tsyurupa, React. Polym. 13 (1990) 27–42. Central China Normal University, Analytical Chemistry, first ed., People’s Education Press, Beijing, 1981, p. 418 (in Chinese).
390
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
[22] D.R. Lide, CRC Handbook of Chemistry and Physics, 76th ed., CRC Press, Boca Raton, FL, 1995, 9-5–9-37. [23] M. Dutta, R. Baruah, N.N. Dutta, Sep. Sci. Technol. 12 (2) (1997) 99– 108. [24] G.E. Boyd, J. Am. Chem. Soc. 67 (1947) 2836–2844. [25] I.J. Langmuir, J. Am. Chem. Soc. 39 (1917) 1848–1906.
[26] F.L. Slejko, Adsorption Technology: A Step-by-Step Approach to Process Evaluation and Application, Dekker, New York, 1985, p. 13. [27] R.A. Garcla-Delgado, L.M. Cotouelo-Minguez, J.J. Rodriguez, Sep. Sci. Technol. 27 (1992) 975–987. [28] J.P. Bell, M. Tsezos, Water Pollut. Control Fed. 59 (1987) 191–198. [29] V. Gokmen, A. Serpen, Food Eng. 63 (2002) 221–227.
Assessment on the removal of dimethyl phthalate from aqueous phase using a hydrophilic hyper-cross-linked polymer resin NDA-702 Weiming Zhang a,b,∗ , Zhengwen Xu a , Bingcai Pan a,∗ , Lu Lv c , Qingjian Zhang a , Qingrui Zhang a , Wei Du a , Bingjun Pan a , Quanxing Zhang a,b a State Key Laboratory of Pollution Control and Resources Reuse, School of the Environment, Nanjing University, Nanjing 210093, People’s Republic of China b Research Center for Engineering Technology of Organic Pollutants Control and Resources Reuse of Jiangsu Province, Nanjing 210038,
People’s Republic of China c Department of Chemical and Biomolecular Engineering, National University of Singapore, 10 Kent Ridge Crescent, 119260 Singapore
Received 8 January 2007; accepted 7 March 2007 Available online 12 April 2007
Abstract A hydrophilic hyper-cross-linked polymer resin (NDA-702) was synthesized, and the adsorption performance of dimethyl phthalate (DMP) on NDA-702 was compared with that on the commercial hydrophobic macroporous resin (Amberlite XAD-4) and granular activated carbon (AC750). The kinetic adsorption of DMP onto NDA-702 and AC-750 is limited mainly by intraparticle diffusion and obeys the pseudo-second-order rate model, while the uptake on XAD-4 is limited mainly by film diffusion and follows the pseudo-first-order rate model. All the associated adsorption isotherms are well described by the Freundlich equation, and the larger uptake and stronger affinity of NDA-702 than AC-750 and XAD-4 probably result from the microporous structure, phenyl rings, and polar groups on NDA-702 polymer matrix. An interesting observation is that in the aqueous phase all the adsorbents spontaneously adsorb DMP driven mainly by enthalpy change, but the hydrophilic nature of NDA-702 and AC-750 surfaces results in less entropy change compared to hydrophobic XAD-4. Dynamic adsorption studies show that the high breakthrough and the total adsorption capacities of NDA-702 are 388 and 559 mg per gram dry resin at 313 K. Nearly 100% regeneration efficiency for the resin was achieved by methanol at 313 K. © 2007 Elsevier Inc. All rights reserved. Keywords: Removal; DMP; Porous materials; Adsorption; Kinetics; Isotherms; Dynamics
1. Introduction Phthalates are a group of organic compounds widely used as plasticizers to improve the flexibility and durability of polyvinyl chloride (PVC)-based plastics. They are also used as additives in the fabrication of multilayer ceramic capacitors, paper coatings, cosmetics, inks, and paints. Since they are not chemically bound to the PVC polymer, a major part of the phthalates entering the environment is suspected to come from phthalatecontaining products during use or after disposal [1–3]. In addition, several industrial plants produce large amounts of waste* Corresponding authors. Fax: +86 25 83707304.
E-mail addresses: [email protected] (W. Zhang), [email protected] (B. Pan). 0021-9797/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2007.03.005
water containing a high concentration of phthalates which have negative effects on the water environment [4]. There are six phthalates included in the USEPA priority list because of their endocrine disrupting, carcinogenic, or xenoestrogenic properties [5,6]. How to effectively and economically remove these phthalates from the environment is critically important and interesting. Some kinds of removal methods for phthalates, such as biodegradation, coagulation, and adsorption, have been reported to date. The bioconversion of phthalates under both aerobic and anaerobic conditions has been investigated [7,8], and their biodegradation by activated sludge has also been demonstrated [9]. However, those methods require a long time to render the phthalates harmless, and microorganisms could hardly degrade them completely or remove them completely
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
from aqueous solution. Although coagulation including flocculation is useful for the removal of organic micropollutants and its removal mechanism has been reported [10], coagulation by ferric chloride was not effective. On the other hand, adsorptive removal by activated carbon and β-cyclodextrin was effective [11,12]. Usually the activated carbon used in water treatment is treated by a heat procedure or steam for recovery; therefore, it is difficult to reuse the materials adsorbed on the activated carbon. And according to its weak stability, cyclodextrin is not used widely as an adsorbent at industrial scale. In recent years, synthetic polymeric resins have been widely applied in industrial effluent treatment in China due to its efficient removal and feasible regeneration for repeated use, as well as the potential recovery of some valuable organic compounds from water [13,14]. For example, the commercially available Amberlite XAD-4 resin was reported as an ideal adsorbent for a wide variety of aromatic compounds. However, Amberlite XAD-4 resin has a hydrophobic surface and a lower adsorption capacity for most organic substances compared with activated carbon, which limits its wide application in the treatment of contaminated water [15]. Recently, some chemically modified polymeric resins have been synthesized and applied in organic pollution control for industrial waste streams in our previous studies [16,17]. Nevertheless, less work focusing on the removal of phthalates using polymer resins was reported [18]. Our understanding of factors controlling the adsorption processes of phthalates on the surface of resins is still limited. The primary objectives of this investigation are to synthesize a new hydrophilic hyper-cross-linked polymer resin (NDA-702) with high capacity and affinity for phthalates, and evaluate the removal and recovery efficiency of phthalates from water using NDA-702. Batch kinetic and equilibrium adsorption experiments were conducted, and thermodynamic analysis was used to elucidate the adsorption mechanism of phthalates on NDA702. Methanol was used to evaluate the regeneration efficiency of NDA-702 in fixed-bed column runs. In this study, NDA-702 (functional hyper-cross-linked polymer resin), XAD-4 (macroporous polymer resin), and AC-750 (granular activated carbon) were selected since they represent three typical kinds of widespread adsorbents in pollution control. Dimethyl phthalate (DMP) was selected as a target compound that represents the low alkyl chain phthalates, which are characterized by high solubility in water [1]. Furthermore, previous studies [19] showed that DMP and other low alkyl chain phthalates could be more toxic to microorganisms during the biotreatment process than the high alkyl chain phthalates. 2. Experimental 2.1. Materials CHA-101, a macroreticular polymer, was kindly provided by Langfang Electrical Resin Co. Ltd. (Hebei province, China). The spherical macroporous Amberlite XAD-4 resin was purchased from Rohm & Haas Company (Philadelphia, PA). Gran-
383
Scheme 1. Synthesis procedures for NDA-702.
ular activated carbon AC-702 is of analytical grade and was purchased from Shanghai reagent station (Shanghai, China). Acetone, methanol, iron chloride, and DMP are of analytical grade and were also purchased from Shanghai reagent station and used without further purification. The background solution used in the adsorption experiments contained 10−3 M KH2 PO4 for buffering the aqueous solutions at neutral pH. 2.2. Synthesis of NDA-702 resin CHA-101 is a macroreticular polystyrene adsorbent with some residual chlormethyl groups on its polymeric matrix during the synthetic process [16], and Friedel–Crafts reaction [20] can take place by the residual chloro atom. Procedures for synthesis of the hydrophilic hyper-cross-linked polymer resin NDA-702 are given in Scheme 1. The specific surface area and the pore distribution of the three adsorbents were respectively measured by BET and BJH methods using a Micrometics ASAP-2010 automatic surface area analysis instrument (Micromeritics Instrument, Norcros, USA). Infrared spectra of the adsorbents in the range of 650– 4000 cm−1 were collected by a Nexus 870 FT-IR spectrometer (USA) with a pellet of powdered potassium bromide. The chlorine content was measured according to the method of Volhard [21]. 2.3. Batch sorption experiments In the following experiments, hydrophilic NDA-702 and AC-750 can be used directly, while hydrophobic XAD-4 should first be wetted with 0.5 ml methanol and then rinsed three times with deionized water. Kinetics of adsorption was determined by analyzing the adsorptive uptake of DMP from its aqueous solution of required concentration 500 mg L−1 at different time intervals until the equilibrium is reached at 313 K. Equilibrium adsorption experiments were carried out at 283, 298, and 313 K in 100-ml glass flasks. All flasks contained a fixed mass of the adsorbent and a known volume of initial aqueous solutions with the desired different initial concentration (50, 100, 200, 300, 400, and 500 mg L−1 , respectively). The flasks were then transferred to a Model G 25 incubator shaker with thermostat (New Brunswick Scientific Co. Inc.) and shaken under 150 rpm until the adsorption process reached equilibrium.
384
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
For all the above systems, control experiments were conducted using flasks prepared similarly but containing no adsorbent for assessing loss of solutes to the flask components during adsorption tests. Results of triplicate flasks showed that the average solution phase concentrations were within 96–102% of the respective initial concentration of the same solution analyzed similarly. Hence, compound loss was negligible. 2.4. Fixed-bed column tests Dynamic adsorption was conducted using a 10 cm × 4.6 mm i.d. stainless-steel column packed with the newly synthesized hydrophilic adsorbent NDA-702, AC-750, and XAD-4, respectively (1.00 g dry adsorbent), and connecting with a 6672 reciprocating pump (Beijing Analytical Instrument Plant) at 313 K. DMP solution of 500 mg L−1 was passed through the column with the superficial liquid velocity (SLV) and the empty bed contact time (EBCT) equal to 0.30 m h−1 and 6 min, respectively. The effluents from the column were quantitatively analyzed. 2.5. Analysis All the solution concentrations of DMP were analyzed spectrophotometrically using a Helious Betra UV–vis spectrophotometer (UK) at a wavelength of 228 nm. 3. Results and discussion 3.1. Characterization of adsorbents Table 1 presents the typical properties of NDA-702, AC-750, and XAD-4. These adsorbents show a similar BET surface area, but quite different chemical composition and pore structures. The cross-linked pore structures of XAD-4 and NDA-702 are depicted in Fig. 1. The pore-size distribution plots in Fig. 2 indicate that the micropore (<2 nm) dominates the pore structure of NDA-702 and AC-702, while mesopore (2–18 nm) plays the main role in the XAD-4 resin. The strong bands of 1700 cm−1 (carbonyl group) and 3650 cm−1 (hydroxyl group) in the IR spectrum of NDA-702 (in Fig. 3) indicate the existence of polar groups on its polymer matrix.
3.2. Adsorption kinetics The influence of contact time on the removal of DMP by NDA-702, AC-750, and XAD-4 is presented in Fig. 4. A comparison made among them indicates that different adsorbents require different times for attaining equilibrium with different extents of uptake. In the case of XAD-4, with increase in shaking time, percentage uptake was found to increase gradually, approaching a maximum at around 100 min. For NDA-702 and AC-750, the required equilibrium times were shifted to higher values, 600 and 720 min, respectively. The faster rate of approach to equilibrium onto XAD-4 is reasonably compared with NDA-702 and AC-750 due to their abundant micropores. The kinetic adsorption data were processed to understand the dynamics of the adsorption process in terms of the order of the rate constant. The pseudo-first-order and pseudo-second-order models [22] were attempted to fit all the data: pseudo-first-order model: k1 t, 2.303 1 1 t = + t, pseudo-second-order model: qt k2 qe2 qe log(qe − qt ) = log(qe ) −
Sorbent
NDA-702
AC-750
XAD-4
Polarity BET surface area (m2 g−1 ) Micropore area (m2 g−1 ) Micropore volume (cm3 g−1 ) Average pore diameter (nm) Particle size (mm) Residual chloride content (%) Oxygen content (%)
Moderate polar 804 481 0.22 2.9 0.9–1.1 2.7 2.9
Polar 749 471 0.22 2.2 0.4–0.6 0 1.4
Nonpolar 868 54 0.004 5.6 0.5–0.7 0 0
(2)
where qe is the equilibrium adsorption capacity (mg g−1 ), qt is the adsorption capacity at contact time t (mg g−1 ), k1 is the pseudo-first-order rate constant (min−1 ), and k2 is the pseudosecond-order rate constant (g mg−1 min−1 ). The fitting results (in Fig. 4 and Table 2) show that the uptake of DMP on macroporous adsorbent XAD-4 follows the pseudo-first-order rate equation, while the uptake on microporous adsorbents NDA702 and AC-750 obeys the pseudo-second-order rate equation. Similarly, it can also be seen that the adsorption rate order of DMP onto these adsorbents is XAD-4 > AC-750 > NDA-702. The different uptake rates between AC-750 and NDA-702 show that not the average pore diameter but the particle size of the two adsorbents (in Table 1) may play a dominant role. In general, the adsorption process may be described as a series of steps: mass transfer from fluid phase to the particle surface across the boundary layer, diffusion within the porous particle, and adsorption itself onto the surface [23]. To determine which one (film diffusion or intraparticle diffusion) is rate limiting and also to find its rate parameters, adsorption kinetic data were further processed. The two diffusion models can be written as [24]: ln(1 − F ) = −kf t,
(3)
intraparticle diffusion model: qt = kp t 1/2 ,
(4)
film diffusion model:
Table 1 Main properties of sorbents
(1)
where kf is the diffusion rate parameter for film diffusion model, kp is the diffusion rate parameter for intraparticle diffusion model, and F is the fractional attainment of the equilibrium. The fitting results of Eqs. (3) and (4) listed in Figs. 5 and 6 show that for macroporous resin XAD-4, the film diffusion model yields better fit to the adsorption kinetic data; while, for microporous adsorbents NDA-702 and AC-750, the intraparticle diffusion model gives a perfect fit. However, all the straight
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
385
Fig. 1. Pore structures of cross-linked polymer resins, XAD-4 and NDA-702.
lines do not pass through the origin. It indicates that both diffusions are the rate-limiting steps for DMP adsorption [22], and the adsorption process on XAD-4 is controlled mainly by film diffusion while the adsorption processes on NDA-702 and AC750 are limited mainly by intraparticle diffusion. Furthermore, it can also be seen in Figs. 5 and 6 that DMP molecules diffuse more rapidly onto the XAD-4 surface than on NDA-702 and AC-750 surfaces. It is probably another creditable factor for the greater adsorption rate of DMP onto XAD-4.
3.3. Adsorption isotherms and thermodynamics Equilibrium data concerning the adsorption of DMP onto the test three adsorbents are presented in Fig. 7. Langmuir and Freundlich [25,26] parameters obtained by nonlinear regression of the experimental data with the Levenberg–Marquardt algorithm are listed in Table 3. The Freundlich model yields a better correlation for all the equilibrium adsorption curves. Fig. 7 and Table 3 show a significantly higher adsorbing ca-
386
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
Fig. 4. Effect of contact time on DMP removal by NDA-702, AC-750, and XAD-4 at 313 K. Table 2 Kinetic parameters of 500 mg L−1 DMP onto NDA-702, AC-750, and XAD-4 at 313 K Adsorbent NDA-702 AC-750 XAD-4
Pseudo-first-order model
Pseudo-second-order model
k1 (min−1 )
R2
k2 (g kg−1 min−1 )
0.932 0.907 0.992
50.4 207.9 307.4
0.032 0.080 0.155
±0.004 ±0.010 ±0.006
±7.62 ±20.03 ±30.28
R2 0.964 0.976 0.979
Fig. 2. Plot of pore-size distributions vs pore diameter for NDA-702, AC-750, and XAD-4.
Fig. 3. IR spectra of NDA-702, AC-750, and XAD-4.
Fig. 5. The straight lines of ln(1–F ) vs the contact time t from DMP adsorbed onto NDA-702, AC-750, and XAD-4 at 313 K.
pacity (KF ) and affinity (n) for DMP on NDA-702 than AC750 and XAD-4. It is probably attributed to three reasons: (1) the micropore (<2 nm) dominates the pore structure of NDA-702 and the micropore area of NDA-702 resin is much higher than that of XAD-4; (2) the polar carbonyl and hy-
droxyl groups on NDA-702 polymer matrix are also helpful for the higher adsorption affinity of the ester group on DMP molecules by the proposed hydrogen-bonding interaction than XAD-4; (3) the hydrophobic phenyl rings on NDA-702 polymer matrix also enhance the adsorption affinity of the aromatic
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
387
Fig. 6. The straight lines of qt vs t1/2 from DMP adsorbed onto NDA-702, AC-750, and XAD-4 at 313 K.
ring on DMP molecules by the proposed π–π interaction than AC-750. The adsorptive free energy change at equilibrium can be calculated with the Gibbs equation G = −nRT ,
(5)
where n is the characteristic constant in Freundlich equation, R is the universal gas constant (8.314 kJ mol−1 K−1 ), and T is the temperature in K. The isosteric adsorptive enthalpy changes with temperature expressed as [27] ln Ce = ln K + (−H /RT ),
(6)
where H is the isosteric adsorptive enthalpy change, and K is a constant. H was calculated from the slope of line plotted by ln Ce vs 1/T (in Fig. 8). The adsorptive entropy change was calculated with the Gibbs–Helmholtz equation [28] S = (H − G)/T .
(7)
Fig. 8 also shows the calculated thermodynamic parameters of DMP adsorption onto NDA-702, AC-750, and XAD-4. The negative value of G indicates that all the test adsorption processes are favorable and spontaneous. Compared to AC-750 and XAD-4, the largest absolute value of G for NDA-702 suggests the strongest adsorptive affinity of DMP to NDA-702. This is in agreement with their adsorption capacities. The negative values of the enthalpy changes indicate an exothermic adsorption process of DMP onto all the three adsorbents. In addition, the same negative values of the entropy changes present a regulation of molecules random movement. Both of them suggest that the adsorption processes of DMP onto all the three adsorbents were driven mainly by enthalpy change. Because in aqueous solution water molecules easily
Fig. 7. Adsorption isotherms of DMP on NDA-702, AC-750, and XAD-4 at different temperatures.
wet hydrophilic surfaces of AC-750 and NDA-702, DMP molecules must replace the water molecules and then they could be adsorbed on these adsorbent surfaces (in Fig. 9). The action is called “solvent replacement” [29]. For the greater hydrophilic property of the AC-750 surface (in Table 1) and the larger num-
388
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
Table 3 Isotherm parameters of DMP adsorption on the three sorbents at different temperatures Sorbent
T (K)
Freundlich
Langmuir
NDA-702
283 298 313
193.2 164.8 140.6
±0.801 ±0.570 ±0.003
5.778 5.203 4.900
AC-750
283 298 313
103.9 82.8 70.6
±0.351 ±0.279 ±0.972
XAD-4
283 298 313
101.3 66.4 48.0
±0.160 ±0.107 ±0.619
KF
R2
KL (L mg−1 )
±0.249 ±0.226 ±0.211
0.997 0.996 0.996
0.551 0.265 0.156
±0.362 ±0.152 ±0.085
447.7 436.3 411.5
±6.92 ±3.88 ±2.53
0.854 0.878 0.879
5.402 4.928 4.665
±0.130 ±0.187 ±0.236
0.999 0.997 0.994
0.129 0.160 0.113
±0.097 ±0.113 ±0.069
286.8 247.6 230.3
±7.15 ±1.10 ±8.98
0.762 0.767 0.784
3.864 3.210 2.912
±0.154 ±0.156 ±0.089
0.997 0.995 0.998
0.061 0.035 0.025
±0.031 ±0.015 ±0.009
424.7 398.5 365.0
±9.27 ±6.34 ±1.37
0.891 0.919 0.937
n
qm (mg g−1 )
R2
capacity and the total adsorption capacity are respectively 388 and 559 mg per gram dry resin, which are much larger than XAD-4 (378 and 430 mg per gram dry resin) and AC-750 (215 and 263 mg per gram dry resin). In addition, their breakthrough adsorption capacities are much larger than 1.29 mg per gram of β-cyclodextrin [12]. The organic solvent methanol was used to desorb DMP from the NDA-702 resin column. At the same flow rate as adsorption, SLV and EBCT equal to 0.30 m h−1 and 6 min, nearly 100% regeneration efficiency was achieved at 313 K (in Fig. 10). By further distilling, both DMP and methanol can be easily recovered. 4. Conclusion
Fig. 8. van’t Hoff plots (ln Ce versus 1/T ) for different types of isotherms presented in Fig. 5 (Qe = 1.0 mmol g−1 of adsorbent, T = 283 K).
ber of water molecules attached on the AC-750 surface than the NDA-702 surface, the number of water molecules on AC-750 replaced by DMP molecules was larger than that on NDA-702. So the solvent replacement process alters the entropy change order as (in Fig. 8) AC-750 > NDA-702 > XAD-4. 3.4. Dynamic adsorption and desorption Due to the satisfactory adsorption capacity and affinity for DMP on NDA-702 in our research, it is hopeful that NDA-702 will be developed as a polymeric adsorbent for the removal and recovery of phthalate pollutants from drinking water or industrial wastewater. It is necessary to test the dynamic adsorption and desorption. The results of minicolumn dynamic adsorption of DMP on NDA-702 are shown in Fig. 10, where C0 is the concentration of initial solution (mM), Cv is the concentration at different bed volumes of the effluent (mg L−1 ), and Cd is the concentration at different bed volumes of regenerative reagent (mg L−1 ). Breakthrough adsorption capacity and the total capacity were calculated on the basis of the total amount of DMP removed when the concentration of the effluent from the column reached 10% and nearly 100% of the initial concentration, respectively. The results of breakthrough adsorption
A new hydrophilic hyper-cross-linked resin NDA-702 for removing DMP from aqueous solution was synthesized. The batch kinetic studies show that the uptake of DMP onto NDA702 and AC-750 is limited mainly by intraparticle diffusion and obeys the pseudo-second-order rate model, while the uptake on XAD-4 is limited mainly by film diffusion and follows the pseudo-first-order rate model. The equilibrium adsorption capacity and affinity of NDA702 to DMP are markedly higher than those of XAD-4 and AC-750, which is attributed to its microporous structure, phenyl rings, and the polar functional groups. All the isotherm data can be fitted with the Freundlich equation satisfactorily. The spontaneous adsorption of DMP onto all the three adsorbents was driven mainly by enthalpy change. The experimental results of dynamic adsorption toward DMP by NDA-702 show that the high breakthrough capacity and the total capacity are 388 and 559 mg per gram dry resin. A 100% regeneration efficiency was achieved using methanol. By further distilling, both DMP and methanol can be easily recovered. Acknowledgments This study was funded by the National Nature Science Foundation of PR China (Grant 20504012) and the Nature Science Foundation of Jiangsu Province (Grant BK2006129).
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
389
Fig. 9. Schemes for the proposed adsorption process of DMP onto the hydrophobic surface of XAD-4 and the hydrophilic surface of NDA-702 and AC-750 in aqueous solution. [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] Fig. 10. Adsorption and desorption dynamic curves of DMP onto NDA-702 at 313 K.
[17]
References
[18]
[1] C. Staples, D. Peterson, T. Parkerton, W. Adams, Chemosphere 35 (1997) 667–749. [2] M. Vitali, M. Guidotti, G. Macilenti, C. Cremisini, Environ. Int. 23 (3) (1997) 337–347. [3] M. Bauer, R. Herrmann, A. Martin, H. Zellmann, Water Sci. Technol. 38 (2) (1998) 185–192.
[19] [20] [21]
J. Wang, K. Xu, L. Xu, Huanjing Kexue 9 (4) (1988) 46–52. C.S. Giam, H.S. Chan, G.S. Neff, E.L. Atlas, Science 199 (1978) 419–421. F.R. Knudsen, T.G. Pottinger, Aquat. Toxicol. 44 (1999) 159–170. R. Aftring, B. Chalker, B. Taylor, Appl. Environ. Microbiol. 41 (1981) 1177–1183. J. Ejlertsson, M. Alnervik, S. Jonsson, B. Svensson, Environ. Sci. Technol. 31 (1997) 2761–2764. R. Iturbe, G. Moreno, P. Elefsiniotis, Environ. Technol. 12 (1991) 783. P. Thebault, J.M. Cases, F. Fiessinger, Water Res. 15 (1981) 183. N. Adhoum, L. Monser, Sep. Purif. Technol. 38 (2004) 233–239. S. Murai, S. Imajo, Y. Takasu, K. Takahashi, K. Hattori, Environ. Sci. Technol. 32 (1998) 782–787. P. Li, A.K. SenGupta, Environ. Sci. Technol. 32 (1998) 3756–3766. Q.X. Zhang, in: K.D. Xu (Ed.), Science and Technology at the Frontier in China, Higher Education Press, Beijing, 2005, pp. 565–592. N. Masque, M. Galia, J. Chromatogr. A 771 (1997) 55–61. B.C. Pan, Q.X. Zhang, F.W. Meng, X.T. Li, X. Zhang, J.Z. Zheng, W.M. Zhang, B.J. Pan, J.L. Chen, Environ. Sci. Technol. 39 (2005) 3308–3313. W.M. Zhang, Q.J. Zhang, B.C. Pan, L. Lv, B.J. Pan, Z.W. Xu, Q.X. Zhang, X.S. Zhao, W. Du, Q.R. Zhang, J. Colloid Interface Sci. 3006 (2007) 216– 221. D. Zhao, A.K. SenGupta, in: Hazardous and Industrial Waster—Proceedings of the Mid-Atlantic Industrial Waste Conference, Lancaster, PA, 1996, pp. 63–70. N. Nielsen, P.B. Larsen, Environmental Review No. 6, Danish Environmental Protection Agency, Denmark, 1996. V.A. Davankov, M.P. Tsyurupa, React. Polym. 13 (1990) 27–42. Central China Normal University, Analytical Chemistry, first ed., People’s Education Press, Beijing, 1981, p. 418 (in Chinese).
390
W. Zhang et al. / Journal of Colloid and Interface Science 311 (2007) 382–390
[22] D.R. Lide, CRC Handbook of Chemistry and Physics, 76th ed., CRC Press, Boca Raton, FL, 1995, 9-5–9-37. [23] M. Dutta, R. Baruah, N.N. Dutta, Sep. Sci. Technol. 12 (2) (1997) 99– 108. [24] G.E. Boyd, J. Am. Chem. Soc. 67 (1947) 2836–2844. [25] I.J. Langmuir, J. Am. Chem. Soc. 39 (1917) 1848–1906.
[26] F.L. Slejko, Adsorption Technology: A Step-by-Step Approach to Process Evaluation and Application, Dekker, New York, 1985, p. 13. [27] R.A. Garcla-Delgado, L.M. Cotouelo-Minguez, J.J. Rodriguez, Sep. Sci. Technol. 27 (1992) 975–987. [28] J.P. Bell, M. Tsezos, Water Pollut. Control Fed. 59 (1987) 191–198. [29] V. Gokmen, A. Serpen, Food Eng. 63 (2002) 221–227.