809

Comparative Adsorption of Nitrophenols on Macroporous Resin and Newly-synthesized Hypercrosslinked Resin

Yuping Qiu*, Jinlong Chen, Aimin Li, Quanxing Zhang, Zhenghao Fei, Zhicai Zhai and Fuqiang Liu State Key Laboratory of Pollution Control and Resources Re-use, School of the Environment, Nanjing
University, Nanjing 210093, P.R. China. (Received 26 September 2002; revised form accepted 10 July 2003)

ABSTRACT: The comparative adsorption of o-nitrophenol, m-nitrophenol and p-nitrophenol onto a macroporous polymeric adsorbent (Amberlite XAD-4) and a hypercrosslinked polymeric adsorbent (NJ-8) was investigated. Adsorption equilibrium isotherms were measured by batch equilibration over the temperature range 283323 K, and two kinds of isotherm model (Langmuir and Freundlich) provided the best fit of the experimental data. The adsorption capacity for each adsorbate on NJ-8 was ca. twice that on Amberlite XAD-4, which is explained by the micropore structure and stronger donoracceptor interaction of NJ-8 resin. Mini-column experiments showed that NJ-8 resin exhibited the longest breakthrough time. The enthalpy of adsorption was estimated from the temperature dependence of adsorption.

INTRODUCTION Nitrophenolic compounds, which are toxic components in chemical industrial wastewater, cause serious pollution because their biological degradation is difficult (Chern and Chien 2002). Furthermore, o-nitrophenol and p-nitrophenol have been prescribed as priority pollutants by the US Environmental Protection Agency (EPA). In the past, activated carbon was used for removing these substances in the adsorption technology of wastewater treatment (Haydar et al. 2003; Nagamany and Richard 1990). However, because of the large number of chemicals identified as pollutants and the overall treatment process costs, other new adsorbents are now being used for this purpose. Non-ionic macroreticular polymeric resins have been used as a medium for removing organics from wastewater because of their wide range of pore structures and excellent physicochemical characteristics (Xu et al. 1997). However, although a polymeric adsorbent such as Amberlite XAD-4 possesses high chemical stability and better regeneration properties relative to activated carbon, it exhibits poor hydrophilic properties and a lower adsorption capacity (Tai et al. 1999). A new series of hypercrosslinking techniques have been developed in recent years by Davankov and Tsyurupa (1969, 1990) in an attempt to resolve such defects. Such techniques have led to the development of polystyrene adsorbents with exceptional hypercrosslinked structures and much higher adsorption capacities (Veverka and Jerabek 1999). The use of such hypercrosslinked polymers for the adsorption of organic compounds has been investigated (Tsurupa et al. 1995; Xu et al. 1999; Li et al. 2001).
*Author to whom all correspondence should be addressed. E-mail: qiubetter@yahoo.com.cn.

810

Yuping Qiu et al./Adsorption Science & Technology Vol. 21 No. 9 2003

In the present work, a type of hypercrosslinked polymeric adsorbent (NJ-8) with improved properties relative to activated carbon and a macroporous polymeric adsorbent was synthesized. The adsorption properties of the synthesized NJ-8 towards nitrophenols were compared with those of Amberlite XAD-4. EXPERIMENTAL Materials o-Nitrophenol, m-nitrophenol and p-nitrophenol were obtained from the Shanghai Third Chemical Reagent Plant (Shanghai, China) and dissolved in deionized water for adsorption studies without any pH adjustment. According to Yao et al. (1985), the solubilities (wt%) of o-nitrophenol, m-nitrophenol and p-nitrophenol in water are 0.208, 3.89 and 2.14, respectively. The spherical Amberlite XAD-4 adsorbent was obtained from Rohm & Haas Co. (Philadelphia, PA, USA) while the NJ-8 hypercrosslinked polymeric adsorbent was prepared via a post-crosslinking step involving the FriedelCrafts reaction between the chloromethylated copolymer of styrenedivinylbenzene using the following procedure. Briefly, zinc chloride (10.0 g) was added slowly to a mixture of 100.0 g of the chloromethylated copolymer of styrenedivinylbenzene swollen in 600 ml nitrobenzene. The mixture was stirred for 12 h at 388 K. The mixture was then poured into an acetone bath, the filtered polymer extracted for 8 h with ethanol in a Soxhlet apparatus and dried under vacuum at 333 K for 8 h. The synthetic method has been described in detail in our previous paper (Li et al. 2002). Prior to their initial use, the two adsorbents were extracted for 8 h with acetone and then dried under vacuum for 2 h at 333 K. Characterization of XAD-4 and NJ-8 The specific surface areas and the pore-size distributions of XAD-4 and NJ-8 were measured using a Micromeritics ASAP-2010 instrument (Micromeritics Instrument Corp., Norcross, GA, USA) applying the BET method. The typical properties of the XAD-4 and NJ-8 polymeric resins are listed in Table 1. The IR spectra of NJ-8 (Figure 1) obtained via a Nicolet 170 SX TR
TABLE 1. Typical Properties of XAD-4 and NJ-8 Resins Property Structure Polarity BET surface area (m2/g) Micropore area (m2/g) Micropore volume (ml/g) Mesopore + micropore area (m2/g) Average particle size (mm) Porosity (ml/g) Colour
a

XAD-4 polystyrenea non-polara 880c 3.1 0.005 645 0.50.6 1.0 white

NJ-8 polystreneb moderately polarb 906 530 0.25 210 0.50.6 0.52 brown

Juang and Shiau (1999). bLi et al. (2002). cValue measured in the present study. According to Juang and Shiau (1999), the value should be 730 m2/g.

Comparative Adsorption of Nitrophenols on a Macroporous Resin and a Hypercrosslinked Resin

811

Figure 1. IR spectra of the chloromethylated styrenedivinylbenzene copolymer before and after FriedelCrafts reaction: curve 1 corresponds to chloromethylated styrenedivinylbenzene copolymer while curve 2 corresponds to NJ-8.

spectrometer show that the band for CCl at 673 cm1 still existed after reaction. The strong new bands at 1700 cm1 and 3421 cm1 indicate the formation of carbonyl and hydroxyl groups during the reaction. Since the partial polarity of the network was favourable towards aqueous media, the product could be used directly without any prior wetting process. The pore-size distribution data obtained experimentally for NJ-8 and XAD-4 are shown in Figure 2. NJ-8 exhibited a microporous nature mainly concentrated in the micropore range (< 1.8 nm), while Amberlite XAD-4 had a dominant pore distribution within the mesopore range (13 nm).

Figure 2. Plots of pore-size distributions versus pore radius for Amberlite XAD-4 and NJ-8. Experimental points: n , XAD-4; l , NJ-8.

812

Yuping Qiu et al./Adsorption Science & Technology Vol. 21 No. 9 2003

Adsorption studies Equilibrium isotherms were obtained by adding 0.100 g of a specific adsorbent to a known volume (100 ml) of o-nitrophenol, m-nitrophenol or p-nitrophenol solution, respectively, at 283 0.5 K. Equilibrium experiments were conducted at 283 K, 303 K and 323 K in order to estimate the adsorption enthalpy. For such studies, the Amberlite XAD-4 resin was firstly wetted with 0.5 ml of methanol and then rinsed three times with deionized water before use. The flasks were capped and agitated in a G-25 model incubator shaker (New Brunswick Scientific Co. Inc.) at 130 rpm for 24 h. After adsorption equilibrium had been attained, the concentrations of the three adsorbates (Ce) were determined. The adsorbent-phase concentration of the adsorbates, qe (mmol/g), was then calculated via the equation: q e = V1 ( C 0 C e )/ MW (1)

where V1 is the volume of solution (l), W is the weight of dry resin (g) and M is the molecular weight of the corresponding nitrophenol. About 1.6 ml of the wet adsorbent was loaded at room temperature into a 10 cm 4.6 mm i.d. stainless-steel mini-column fitted with a 2-m stainless-steel frit connected to a 6672 reciprocating pump (Beijing Analytical Instrument Plant). The column adsorption tests for each nitrophenol were conducted using an adsorbate solution of 14.4 mmol/l initial concentration under identical operation conditions at the same flow rate of 10 BV/h (where BV is the resin bed volume). Analysis of the concentration of o-nitrophenol, m-nitrophenol and p-nitrophenol was conducted spectrophotometrically at a wavelength of 275 nm using an HPLC assembly consisting of a Waters 600 controller and a Waters Dual l Absorbance Detector (Waters Corp., Milford, MA, USA). RESULTS AND DISCUSSION Equilibrium adsorption The equilibrium adsorption isotherms at 283 K on Amberlite XAD-4 and NJ-8 for the three adsorbates are depicted in Figures 35 while the equilibrium adsorption isotherms for the three adsorbates on NJ-8 and Amberlite XAD-4 at 283 K, 303 K and 323 K are shown in Figures 611, respectively. It will be seen that the adsorption capacity on NJ-8 was higher than that on XAD-4 for each adsorbate studied. The adsorption equilibria were described by the Langmuir and Freundlich equations. The Langmuir equation may expressed mathematically as: C e / q e = 1/( K L q m ) + C e / q m (2)

where KL is a direct measure of the intensity of the adsorption process and qm is a constant related to the area occupied by a monolayer of adsorbate, and thus reflects the adsorption capacity (Fytianos et al. 2000). Similarly, the Freundlich equation may be written as: log q e = log K F + 1 n log C e (3)

where KF is taken as an indicator of the adsorption capacity and 1/n is an empirical constant related to the magnitude of the adsorption driving force (Jung et al. 2001).

Comparative Adsorption of Nitrophenols on a Macroporous Resin and a Hypercrosslinked Resin

813

Figure 3. Equilibrium adsorption isotherms measured at 283 K for o-nitrophenol onto Amberlite XAD-4 and NJ-8. Experimental points: n, XAD-4; l, NJ-8.

The values for the parameters obtained at 283 K for the adsorption isotherms of o-nitrophenol, m -nitrophenol and p-nitrophenol, i.e. KL, qm, KF, n and the correlation coefficient R, are listed in Table 2. From these, it is obvious that both the Langmuir and Freundlich adsorption laws are applicable to the adsorption of the three adsorbates on Amberlite XAD-4 and NJ-8 resin, with the value of qm on NJ-8 being ca. twice that on Amberlite XAD-4. The results indicate that the adsorption capacity of NJ-8 for each adsorbate was greater than that of Amberlite XAD-4. The values of the exponent n were greater than unity in all cases, indicating that the adsorption was favourable (Juang et al. 1999). Both equations are reliable for all correlation constants R greater than 0.99. The specific surface area of NJ-8 (906 m2/g) was similar in magnitude to that of Amberlite XAD-4 (880 m2/g), but the adsorption capacity of NJ-8 appeared to differ from that of Amberlite XAD-4. The surface area of the adsorbent available for the adsorption of p-nitrophenol is listed in

Figure 4. Equilibrium adsorption isotherms measured at 283 K for m-nitrophenol onto Amberlite XAD-4 and NJ-8. Experimental points: n, XAD-4; l, NJ-8.

814

Yuping Qiu et al./Adsorption Science & Technology Vol. 21 No. 9 2003

Figure 5. Equilibrium adsorption isotherms measured at 283 K for p-nitrophenol onto Amberlite XAD-4 and NJ-8. Experimental points: n, XAD-4; l, NJ-8.

Table 3. According to Furuya et al. (1997), the surface area of the p-nitrophenol molecule is 0.7 nm2, which is larger than that of NJ-8 (0.46 nm2). This suggests that the maximum adsorption observed on NJ-8 was less than that expected for monolayer coverage. Hence, it would be incorrect to apply the concept of surface area evaluation based on a monolayer adsorption capacity for NJ-8 resin. The micropore-filling mechanism (Mattson et al. 1969) and the capillary condensation theory (Nikolai et al. 2000) may be employed to explain the observed adsorption phenomenon. Relative to Amberlite XAD-4, NJ-8 exhibits a predominantly microporous distribution (Table 1). This indicates that strong adsorption of the adsorbates studied would occur in the micropores of NJ-8, because the fine pore structure of the adsorbent would encourage solutesolute interaction in accordance with the capillary condensation theory. This conclusion agrees well with those of other

Figure 6. Equilibrium adsorption isotherms measured at three different temperatures for o-nitrophenol onto NJ-8. Experimental points: n, 283 K; l , 303 K; and s , 323 K.

Comparative Adsorption of Nitrophenols on a Macroporous Resin and a Hypercrosslinked Resin

815

Figure 7. Equilibrium adsorption isotherms measured at three different temperatures for m-nitrophenol onto NJ-8. Experimental points: n, 283 K; l , 303 K; and s , 323 K.

Figure 8. Equilibrium adsorption isotherms measured at three different temperatures for p-nitrophenol onto NJ-8. Experimental points: n, 283 K; l , 303 K; and s , 323 K.

Figure 9. Equilibrium adsorption isotherms measured at three different temperatures for o-nitrophenol onto XAD-4. Experimental points: n, 283 K; l , 303 K; and s , 323 K.

816

Yuping Qiu et al./Adsorption Science & Technology Vol. 21 No. 9 2003

Figure 10. Equilibrium adsorption isotherms measured at three different temperatures for m-nitrophenol onto XAD-4. Experimental points: n, 283 K; l , 303 K; and s , 323 K.

Figure 11. Equilibrium adsorption isotherms measured at three different temperatures for p-nitrophenol onto XAD-4. Experimental points: n, 283 K; l , 303 K; and s , 323 K.

TABLE 2. Parameter Values for the Adsorption Isotherms of the Three Adsorbates at 283 K

Adsorbates

Adsorbents Langmuir model KL qm 3.4423 6.5445 1.9810 3.1686 1.9826 3.2637 R

2

Freundlich model KF 1.6331 4.6979 0.8449 1.9240 0.6976 1.7470 n 2.2252 2.1853 2.5126 3.3738 2.2411 2.9163 R2 0.9951 0.9821 0.9865 0.9954 0.9954 0.9985

o-nitrophenol m -nitrophenol p-nitrophenol

XAD-4 NJ-8 XAD-4 NJ-8 XAD-4 NJ-8

1.2104 2.8668 0.7437 1.9469 0.5043 1.3264

0.9901 0.9963 0.9991 0.9946 0.9832 0.9884

Comparative Adsorption of Nitrophenols on a Macroporous Resin and a Hypercrosslinked Resin

817

TABLE 3. Estimates of the Surface Area Available for the Adsorption of p-Nitrophenol on the Two Adsorbents Studied at 283 K qm (mmol/g)a 1.98 3.26

Adsorbent XAD-4 NJ-8

a

S/(qmN) (nm2/molecule)b 0.73 0.46

Based on the maximum qm at 283 K. bSurface area available for adsorption of the three adsorbates, based on the maximum qm obtained in our studies (S = specific surface area, N = Avogadros number).

reports (Kipling 1965). In addition, the donoracceptor mechanism between the adsorbate and adsorbent should be taken into consideration during an adsorptive process from aqueous solution (Gould 1959). Two factors influence the pp aromatic ring system. One is the availability of electron density on the donor molecule (the aromatic ring of the resin network) while the other is the electron affinity of the acceptor (nitrophenol). In comparison to Amberlite XAD-4, the surface of NJ-8 resin possesses carbonyl and hydroxyl groups as shown by the infrared spectra depicted in Figure 1. Thus, it would be expected that the aromatic ring of the NJ-8 resin network would act as a strong donor thereby making the adsorption capacity of NJ-8 greater than that of XAD-4. As shown above, o-nitrophenol possesses the least solubility of the three adsorbates because of the formation of an intramolecular hydrogen bond. For this reason, it would be expected to exhibit the highest adsorptive capacity. Previous results (Rojas and Voilley 1996) regarding the adsorption behaviours of organic compounds from aqueous solution onto non-polar and moderately polar polymeric adsorbents may be related to the hydrophobic properties of the solutes concerned (Russell et al. 1971). Thermodynamic adsorption behaviour of each adsorbate The enthalpy of adsorption at different adsorption levels can be estimated from the derivative Vant Hoff equation expressed as (Garcia-Delgado et al. 1992): log C e = log K 0 + ( D H 2.303RT) (4)

where DH is the enthalpy of adsorption, K0 is a constant, R is the gas constant and Ce (mmol/l) is the equilibrium concentration of the solution at an absolute temperature T (K). The equilibrium concentrations, Ce, correspond to the determined value of q. The values of Ce at different adsorption levels (with q = 1.0 mmol/g and 1.5 mmol/g) were obtained from the well-fitting Freundlich equation at 283 K, 303 K and 323 K, while D H was calculated from the slope of the linear plot obtained when log Ce was plotted against 1/T (Figures 12 and 13). A summary of the estimated values for the adsorption enthalpy at two adsorbent loadings is shown in Table 4. Values of the enthalpy changes onto NJ-8 and XAD-4 (which were always negative) are indicative of an exothermic process, with their magnitudes (< 40 kJ/mol) reflecting the nature of the adsorption process. The adsorption enthalpy of NJ-8 was higher than that of

818

Yuping Qiu et al./Adsorption Science & Technology Vol. 21 No. 9 2003

Figure 12. Determination of enthalpy change of adsorption for o-nitrophenol on XAD-4. Experimental points: n, q = 1.0 mmol/g; l , q = 1.5 mmol/g.

Figure 13. Determination of enthalpy change of adsorption of o-nitrophenol on NJ-8. Experimental points: 1.0 mmol/g; l , q = 1.5 mmol/g.

n,

q =

TABLE 4. Enthalpy of Adsorption, DH (kJ/mol), Evaluated at Different Adsorption Capacities Over the Temperature Range 283323 K Adsorbate o-nitrophenol o-nitrophenol m-nitrophenol m-nitrophenol p-nitrophenol p-nitrophenol Adsorbent NJ-8 XAD-4 NJ-8 XAD-4 NJ-8 XAD-4 q = 1.0 mmol/g 19.6 18.3 27.7 17.2 28.0 17.6 q = 1.5 mmol/g 18.4 16.5 20.9 16.6 20.4 13.1

Comparative Adsorption of Nitrophenols on a Macroporous Resin and a Hypercrosslinked Resin

819

XAD-4. Similarly, the adsorbent exhibiting the lowest adsorption capacity also showed the lowest adsorption enthalpy. At low adsorptive capacity (q = 1.0 mmol/g), the interactions between the adsorbate and adsorbent indicated mainly direct interaction between the solute and the adsorbent. With increased loading (q = 1.5 mmol/g), the surface of adsorbent was gradually covered with adsorbate and as a result the interaction between the adsorbate and adsorbent gradually reflected the interaction between the solute in solution and the solute adsorbed on the adsorbent. Hence, the enthalpy gradually decreased. Column adsorption of each adsorbate on XAD-4 and NJ-8 The breakthrough times for o-nitrophenol, m-nitrophenol and p-nitrophenol were measured at initial concentrations of 14.4 mmol/l at 10 BV/h and were 34.6 h, 4.7 h and 5.8 h, respectively, on NJ-8. These values were greater than those on XAD-4, which were 10.5 h, 4.1 h and 4.4 h, respectively. The results demonstrate that the adsorption capacity for each adsorbate on NJ-8 was greater than that on XAD-4, in agreement with the trend noted for the equilibrium adsorption results. CONCLUSIONS Studies were conducted of the adsorption of o-nitrophenol, m-nitrophenol and p-nitrophenol from aqueous solution onto XAD-4 and NJ-8. The Langmuir and Freundlich equations were found to fit the experimental data well. In general, the adsorption capacities on NJ-8 were ca. twice those on Amberlite XAD-4, a result attributed to the micropore structure of the former adsorbent and the stronger donoracceptor interaction. The adsorption enthalpies of the three adsorbates on NJ-8 resin (which were always negative) indicated an exothermic process, with their magnitudes (< 40 kJ/mol) corresponding to a physical sorption process. The breakthrough time for each adsorbate on NJ-8 was longer than that on XAD-4, in agreement with the equilibrium adsorption results.

ACKNOWLEDGMENTS The authors express their thanks to the Analytical Center of Yancheng Teachers College for the IR spectral analyses and the Analytical Center of Changzhou Petro-Chemical College for measurements of surface area and pore diameter. This work was funded by the National Natural Science Foundation of P.R. China (Project No. 20274017).

REFERENCES
Chern, J.M. and Chien, Y.W. (2002) Water Res. 36, 647. Davankov, V.A. and Tsyurupa, M.P. (1969) Pure Appl. Chem. 61, 1881. Davankov, V.A. and Tsyurupa, M.P. (1990) React. Polym. 13, 2. Furuya, E.G., Chang, H.T., Miura, Y. and Noll, K.E. (1997) Sep. Purif. Technol. 11, 69. Fytianos, K., Voudrias, E. and Kokkalis, E. (2000) Chemosphere 40, 981. Garcia-Delgado, R.A., Cotouelo-Minguez, L.M. and Rodriguez, J.J. (1992) Sep. Sci. Technol. 27, 975.

820

Yuping Qiu et al./Adsorption Science & Technology Vol. 21 No. 9 2003

Gould, E.S. (1959) Mechanism and Structure in Organic Chemistry, Holt, Rinehart & Winston, New York. Haydar, S., Ferro-Garcia, M.A., Rivera-Utrilla, J. and Joly, J.P. (2003) Carbon 41, 387. Juang, R.S. and Shiau, J.Y. (1999) J. Hazard. Mater. B 70, 171. Juang, R.S., Shiau, J.Y. and Shao, H.J. (1999) Sep. Sci. Technol. 34, 1819. Jung, M.W., Ahn, K.H., Lee, Y.H., Kim, K.P., Rhee, J.S., Park, J.T. and Paeng, K.J. (2001) Microchem. J. 70, 123. Kipling, J.J. (1965) Adsorption from Solutions of Non-electrolytes, Academic Press, London. Li, A.M., Zhang, Q.X., Chen, J.L., Fei, Z.F., Long, C. and Li, W.X. (2001) React. Funct. Polym. 49, 225. Li, A.M., Zhang, Q.X., Zhang, G.C., Chen, J.L., Fei, Z.F. and Liu, F.Q. (2002) Chemosphere 47, 989. Mattson, J.S., Harry, Jr., B.M., Michael, D.M., Walter, Jr., J.W. and John, C.C. (1969) J. Colloid Interface Sci. 31, 116. Nagamany, N. and Richard, E.S. (1990) Environ. Sci. Technol. 24, 575. Newcombe, G. (1999) Water Sci. Technol. 40, 191. Nikolai, V.C., Gerhard, S. and Jurgen, A. (2000) J. Colloid Interface Sci. 221, 246. Rojas, J.A. and Voilley, A. (1996) Sep. Sci. Technol. 31, 2473. Russell, S.D., Glenn, C.V. and Terence, E.N. (1971) J. Am. Chem. Soc. 17, 6014. Tai, M.H., Saha, B. and Streat, M. (1999) React. Funct. Polym. 41, 149. Tsyurupa, M.P., Maslova, L.A., Andreeva, A.I., Mrachkovskaya, T.A. and Davankov, V.A. (1995) React. Polym. 25, 69. Veverka, P. and Jerabek, K. (1999) React. Polym. 41, 21. Xu, Z.Y., Zhang, Q.X., Chen, J.L. and Wang, L.S. (1997) Chemosphere 35, 2269. Xu, Z.Y., Zhang, Q.X., Chen, J.L., Wang, L.S. and Anderson, G.K. (1999) Chemosphere 38, 2003. Yao, Y.B., Tao, X. and Gao, Y.M. (1985) Handbook of Chemistry & Physics, Shanghai Science Press, Shanghai, P.R. China, p. 741 (in Chinese).