- Email: [email protected]
Enhanced adsorption and desorption of VOCs vapor on novel micro-mesoporous polymeric adsorbents
Journal of Colloid and Interface Science 428 (2014) 185–190
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Enhanced adsorption and desorption of VOCs vapor on novel micro-mesoporous polymeric adsorbents Shuangshuang Wang, Liang Zhang, Chao Long ⇑, Aimin Li State Key Laboratory of Pollution Control and Resource Reuse, School of the Environment, Nanjing University, 163 Xianlin Road, Nanjing 210023, China
a r t i c l e
i n f o
Article history: Received 27 December 2013 Accepted 25 April 2014 Available online 2 May 2014 Keywords: Hypercrosslinked resin Micro-mesopore Adsorption Desorption VOCs
a b s t r a c t To enhance adsorption and desorption efficiency of volatile organic compounds from gas streams, we synthesized a series of well-developed micro-mesoporous hypercrosslinked polymeric adsorbents (MM-1, MM-2, and MM-3). The adsorption and desorption performance of dichloromethane and 2-butanone on newly synthesized adsorbents was investigated and compared with commercial microporedominated hypercrosslinked polymeric adsorbent (JT-001). The contributions of micropore and mesopore to adsorption and desorption of dichloromethane and 2-butanone on polymeric adsorbents were well elucidated. Consequently, the adsorbent MM-3 had been sorted out with high BET surface area (1606 m2/g), large micropore and mesopore volumes (0.562 mL/g and 1.046 mL/g, respectively). The MM-3 exhibited the similar adsorption capacities with JT-001 for dichloromethane and 2-butanone at regions of p/p0 < 0.2, but had higher adsorption capacities than JT-001 at high relative pressures. The largest adsorption capacities of MM-3 for dichloromethane and 2-butanone at 308 K were 1345.3 mg/g and 853.5 mg/g, respectively, which are about 1.78 and 1.88 times those of JT-001 under the same condition. Furthermore, the MM-3 exhibited higher desorption efficiencies than JT-001, especially for 2-butanone with a higher boiling point. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction Nowadays, the emission of volatile organic compounds (VOCs) has gained more and more attention not only for their damage to human health and environment but for economic interests as well. Adsorption technology has been recognized as a preferred method for the removal of VOCs because it provides additional benefits from the recovery of the valuable VOCs for reuse [1,2]. The heart of an adsorption process is usually a porous solid adsorbent. Although an explosion has been seen in the development of new nanoporous adsorption materials in the past two decades, the activated carbon is still the major commercial sorbent used for gas stream processing. The activated carbon provides excellent adsorption capacity, but it has been recognized that it owns some disadvantages such as fire risk, pore blocking and hygroscopicity [3–6]. In addition, the activated carbons are predominantly consisted of micropores (<2 nm in size) with the specific surface area and pore volume on the order of 800–1500 m2/g and 0.2–0.6 mL/g, respectively [7,8]. Such microporous adsorbents are desirable for the adsorption of low concentration of VOCs, but inefficient to adsorb
⇑ Corresponding author. E-mail address: [email protected] (C. Long). http://dx.doi.org/10.1016/j.jcis.2014.04.055 0021-9797/Ó 2014 Elsevier Inc. All rights reserved.
medium–high concentration of VOCs owing to the limitation of pore volume. Furthermore, it is very difficult to regenerate adsorbed VOCs due to the strong dispersive force resulting from the overlap of attractive forces from opposite pore walls of micropore [9,10]. Hypercrosslinked polymeric adsorbent represents a class of predominantly microporous organic materials [11–14] and has gained increasing interest for removing the VOCs from gas steam currently. It is found that hypercrosslinked polymeric adsorbent had comparable adsorption capacities with activated carbon for VOCs [15–17]. However, it should be noted that commercial hypercrosslinked polymeric adsorbent is typical of adsorbent with a predominant microporous structure in the regions of pore size 0.5–2 nm [16,18–20], and has similar properties to activated carbons such as low regeneration efficiency and insufficient adsorption capacities for medium–high concentration VOCs. In addition, although the commercial hypercrosslinked polymeric adsorbent contains a certain amount of mesopore and macropore, these pores are mostly in the range of 30–70 nm in diameter, which do not have a considerable contribution to adsorption. In comparison with hypercrosslinked polymeric adsorbent, macroporous polymeric adsorbent has the considerably low adsorption capacities for VOCs at the ranges of low relative pressure. However, it can exhibit high adsorption capacities for VOCs at high relative pressure due to the
186
S. Wang et al. / Journal of Colloid and Interface Science 428 (2014) 185–190
existence of mesopore [21,22]. Therefore, the control of pore structure of adsorbents is strongly required in fields of application to obtain the good adsorption and desorption efficiency for VOCs. Recently, we reported a novel hypercrosslinked polymeric adsorbent with micro-mesoporous structure in a brief communication [23]. Compared with commercial hypercrosslinked polymeric adsorbent, the improved adsorption and desorption performance of n-hexane was found at higher relative pressure. However, the equilibrium isotherms of VOCs on the micro-mesoporous polymeric adsorbents were not investigated; the contributions and mechanism of micropore and mesopore to adsorption and desorption of VOCs on the adsorbents had not been elucidated yet. Especially, the available information on the role of mesopore in adsorption and desorption of VOCs vapors is very limited. In this work, we successfully prepared a series of controllable micro-mesoporous hypercrosslinked polymeric adsorbents by optimizing the synthetic process reported in the previous communication [23]. The adsorption–desorption performance of dichloromethane and 2-butanone onto a series of micro-mesoporous hypercrosslinked polymeric adsorbents was investigated. Moreover, the effect of micropore and mesopore on adsorption and desorption of the selected VOCs was clarified. 2. Experimental 2.1. Synthesis of micro-mesoporous polymeric adsorbents The micro-mesoporous hypercrosslinked polymeric adsorbents were prepared by optimizing the synthetic process reported in the previous brief communication [23]. Briefly, first, low-crosslinked copolymer was synthesized by suspension polymerization using vinylbenzyl chloride as the monomer, divinylbenzene as the cross-linking agent, and a mixture of toluene and n-heptane as the inert porogen. Three different mass ratios of toluene and nheptane listed in Table S1 in the Supporting Information (SI) were used to control the pore structure of polymers. Then, micro-mesoporous hypercrosslinked polymeric adsorbent was prepared via a post-crosslinking of low-crosslinked poly (vinylbenzyl chloridedivinylbenzene). The obtained samples in this study were named as MM-1, MM-2, and MM-3, respectively. The commercial hypercrosslinked polymeric adsorbent (JT-001) was supplied by Jiangsu N&G Environment Co. 2.2. Characterization of adsorbents Nitrogen adsorption–desorption isotherms were measured using an ASAP 2020 (Micromeritics Instrument Co., USA) at 77 K. The specific surface area was determined using the N2 isotherms data by means of the BET equation. The total pore volume (Vtotal) was estimated to be the liquid volume of nitrogen at a relative pressure of about 0.98. The micropore volume (Vmicro) and mesopore volume (Vmeso) were calculated from the N2 isotherms data by Dubinin–Radushkevich (DR) and BJH methods, respectively. The pore size distributions were calculated by applying the density functional theory (DFT) to N2 isotherm data. 2.3. Adsorption and desorption experiments The adsorption equilibrium of dichloromethane and 2-butanone onto MM-1, MM-2, MM-3 and JT-001 has been conducted using a gravimetric method. The detailed experimental apparatus and adsorption procedure have been described previously [19]. In brief, samples were precisely weighed out and then charged into the adsorption column made of glass. The carrier gas containing a known concentration of VOCs vapor was passed through the column until
the VOCs concentration becomes constant and stable. The change of VOCs concentration in the effluent steam from the adsorption column was measured by using gas chromatography (GC, 9790, FULI, China) with a flame ionization detector (FID) and recorded by a computer. The adsorption equilibrium was attained when the exit concentration became equal to the inlet concentration. The equilibrium adsorption amount of VOCs onto adsorbent was equal to the weight change of adsorbent before and after adsorption process. Here, a high precision microbalance (BS224S, Sartorius, Germany) was adopted as the weighing device. After the adsorption procedure, desorption behavior was investigated by connecting the adsorption column to a vacuum pump under vacuum of 0.005 MPa. The desorbed amount was equal to the weight change of adsorbent before and after the desorption process. 3. Results and discussion 3.1. Characterization of adsorbents The N2 adsorption–desorption isotherms at 77 K of four adsorbents MM-1, MM-2, MM-3 and JT-001 are demonstrated in Fig. S1 in Supporting Information. It is observed that at lower relative pressure (p/p0) below 0.05, the nitrogen uptake increases sharply with the increment of relative pressure, proving the existence of micropore structure in the four polymeric adsorbents. Moreover, visible hysteresis loop means that all the adsorbents contain mesopore. According to IUPAC classification, the adsorption–desorption isotherms of JT-001 and MM-1 are close to type I, reflecting the domination of micropores in the pore structure; however, the accelerated uptake at p/p0 1 means that JT-001 contains the larger-sized mesopores or macropores. The N2 adsorption–desorption isotherms of the newly synthesized adsorbents MM-2 and MM-3 are the combination of type I and IV with remarkable hysteresis loops, suggesting the distinct micro-mesoporous structures. In order to understand more clearly the pore structure of adsorbents, the pore size distributions of four adsorbents are shown in Fig. 1; the textural properties of all the adsorbents studied are included in Table 1. It is significantly learned that the four adsorbents have nearly similar micropore volume. Although the mesopore volume of JT-001 is 0.528 mL/g, the mesopores are mainly distributed in the regions of 20–50 nm near the macropores. In contrast, the mesopores of newly synthesized adsorbents are in pore size of 2–10 nm near the micropore. Therefore, the newly synthesized adsorbents MM-1, MM-2 and MM-3 possess higher mesopore surface area (317 m2/g, 716 m2/g and 1087 m2/g, respectively) than JT-001. Overall, all three newly synthesized adsorbents contain well-developed pore structure in both regions of micropore and mesopore. 3.2. Adsorption performance Fig. 2 shows the adsorption isotherms of dichloromethane and 2-butanone on MM-1, MM-2, MM-3 and JT-001 at temperatures of 308 K and 323 K, respectively. To clearly clarify the adsorption mechanism, volumetric adsorption capacity qv is used by the unit of mL/g, which is equal to the ratio of equilibrium adsorption capacity (mg/g) to the adsorbate density in the liquid phase (g/ mL). The adsorption amounts of both dichloromethane and 2-butanone rose as the adsorption temperature decreases, suggesting that the adsorption is an exothermic process. Compared with dichloromethane, the higher uptake of 2-butanone at the same temperature and vapor pressure may result from the stronger polarizability of 2-butanone (16.424 and 20.681 mL/mol for dichloromethane and 2-butanone, respectively), which is the first and usually most important contribution to dispersion forces
S. Wang et al. / Journal of Colloid and Interface Science 428 (2014) 185–190
187
Fig. 1. Pore size distribution of MM-1, MM-2, MM-3 and JT-001.
Table 1 Textural properties of MM-1, MM-2, MM-3 and JT-001. Adsorbent
JT-001
MM-1
MM-2
MM-3
BET surface area (m2/g) BJH surface area (m2/g) Vmicro (mL/g) Vmeso (mL/g) Vtotal (mL/g) Average pore diameter (nm)
1315 259 0.573 0.528 1.075 2.757
1373 317 0.539 0.241 0.744 2.166
1379 716 0.517 0.863 1.261 3.657
1606 1087 0.562 1.046 1.424 3.548
between the adsorbate and the surface of adsorbent [24]. Meanwhile, three micro-mesoporous polymeric adsorbents exhibited the comparable adsorption capacities with JT-001 for dichloromethane and 2-butanone at regions of p/p0 < 0.2, which may be attributed to almost the same micropore volume of the four adsorbents. However, it is clearly seen that the three newly synthesized micro-mesoporous polymeric adsorbents had larger adsorption capacities than JT-001 at high relative pressure. Among all the adsorbents, the MM-3 showed the largest adsorption capacities of 1.014 mL/g and 1.130 mL/g for dichloromethane and 2-butanone at 308 K, respectively, which are about 1.78 and 1.88 times those of JT-001 obtained under the same experimental conditions. On the other hand, although JT-001 had larger mesopore volume, it should be noted that JT-001 kept an almost constant volumetric adsorption capacities similar to its micropore volume for dichloromethane and 2-butanone at the regions of medium–high relative pressure. However, MM-2 and MM-3 can obtain higher adsorption capacities in excess of their micropore volumes. The results demonstrated that the mesopores with suitable pore size play a key role in the improvement of the adsorption capacity for medium– high concentration of VOCs. 3.3. Modeling of isotherms The equilibrium data were correlated by Dubinin–Astakhov (DA) equation. DA equation can be expressed as follows.
qv ¼ q0 exp½ðe=EÞr
ð1Þ
qv ¼ q=q
ð2Þ
e ¼ RT lnðp0 =pÞ
ð3Þ
where qv is volumetric adsorption capacity (mL/g), q0 is the maximum adsorption volume (mL/g), q is the equilibrium amount (mg/g), q is the adsorbate density in the adsorbed phase assumed to be the same as that in the liquid phase (g/mL), e is the adsorption potential (J/mol) written by Eq. (3), E is the adsorption characteristic energy (kJ/mol), R is a gas constant (8.314 J/molK), T is the absolute temperature (K), p0 is the saturation vapor pressure (kPa), p is the equilibrium vapor pressure (kPa). The fitting parameters of DA equation for adsorption data at each temperature are summarized in Table S2 (Supporting Information). It is shown that the DA model fits the adsorption data reasonably with large correlation coefficient (R2 > 0.993). According to potential theory, it is certain that plots of adsorbed volume (qv) versus the adsorption potential at different temperature should yield a unique characteristic curve that is independent of temperature [25], which can be used to further examine whether Polanyi theory mechanistically captures the adsorption process of compounds by adsorbent. Adsorption characteristic curve of dichloromethane and 2-butanone on JT-001, MM-1, MM-2, and MM-3 is shown in Fig. S2 in SI. As the Polanyi potential theory would predict, they all fell essentially onto a single curve with high correlation coefficient R2 (Table S3 in SI), indicating the availableness of Polanyi theory to describe the adsorption of dichloromethane and 2-butanone onto polymeric adsorbents. In addition, plots of the ratio of the adsorption potential (e) to molar polarizability (pe) versus adsorbed volume (qv) of dichloromethane and 2-butanone onto four adsorbents are shown in Fig. 3. It can be seen clearly from Fig. 3 that plots of e/pe versus qv for different adsorbents had a good superposition at low relative pressure (high values of e/pe), while a big deviation was observed at high relative
188
S. Wang et al. / Journal of Colloid and Interface Science 428 (2014) 185–190
Fig. 2. Equilibrium adsorption data of dichloromethane and 2-butanone onto MM-1, MM-2, MM-3 and JT-001 at 308 K and 323 K.
pressure (low values of e/pe) and this is also affirmed by residual analysis. Such results are directly related to the pore size distribution of four polymeric adsorbents. As we all know, the adsorbate molecules were preferentially adsorbed in micropores by pore-filling at low relative pressure. From Fig. 1 and Table 1, it is learned that four adsorbents had a very similar micropore size distribution and basically the same micropore volume. Therefore, the good superposition of characteristic curves of four adsorbents can be found in the regions of high values of e/pe in Fig. 3. However, at high relative pressure, the adsorbates are adsorbed mainly by multilayer adsorption and capillary condensation; the pore size distribution in the regions of mesopore has significant influence on adsorption capacity, which will be explained in the ensuing para-
Fig. 3. Plots of adsorbed volume (qv) versus normalized adsorption potential density (e/pe) for dichloromethane and 2-butanone onto MM-1, MM-2, MM-3 and JT-001.
graphs. Consequently, the characteristic curves of four adsorbents at high relative pressure (low values of e/pe) had a big discrepancy, resulting from the different pore size distribution and specific surface area of four adsorbents in the regions of mesopores.
3.4. Contribution of mesopores to adsorption The fitting parameters of DA equation calculated from all the experimental data for four adsorbents are given in Table S3. As the data presented in Table S3, the calculated values of q0 of four adsorbents were higher than the micropore volume but lower than the sum of micropore and mesopore volume except for adsorption of dichloromethane on JT-001, implying that the mesopores make a certain of contribution to the adsorption capacity of dichloromethane and 2-butanone. Although the commercial hypercrosslinked polymeric adsorbent JT-001 has larger mesopore volume (0.528 mL/g) than micro-mesoporous adsorbent MM-1, the values of q0 for dichloromethane and 2-butanone are less than MM-1. This result implies that the mesopores on JT-001 made very limited contribution to dichloromethane and 2-butanone uptake, which was confirmed by the equilibrium data of JT-001 and MM-1 at larger relative pressure. The result is probably ascribed to the different mesopore size distribution of two adsorbents. The mesopore diameter of JT-001 falls largely into the region of 30–50 nm near the macropore (Fig. 1). These larger mesopores made a small contribution to adsorption and primarily acted as transport pores. The mesopores of MM-1 were mainly distributed in region of 2–10 nm. Therefore, the multilayer adsorption and following capillary condensation in the mesopore of MM-1 may take place at lower relative pressures. Among the four adsorbents, MM-3 has the largest mesopore volume and suitable pore size distribution, whose mesopores are mainly distributed in region of 2–10 nm approaching to micropore. Therefore, MM-3 had the largest values of q0 (1.305 mL/ g for dichloromethane and 1.526 mL/g for 2-butanone), which are close to the sum volume of micropore and mesopore (1.608 mL/g).
S. Wang et al. / Journal of Colloid and Interface Science 428 (2014) 185–190
189
Fig. 4. Plots Vmeso and Smeso versus the difference value (DV) for adsorption of dichloromethane and 2-butanone adsorption on four adsorbents.
Fig. 5. Desorption curves for dichloromethane and 2-butanone from MM-1 MM-2, MM-3 and JT-001 (desorption temperature: 293 K).
In order to understand the contribution of mesopore of adsorbents to adsorption more clearly, plots of Smeso and Vmeso versus DV (Here DV is the difference value between the maximum adsorption volume q0 and micropore volume Vmicro) are depicted in Fig. 4, respectively. Smeso, Vmeso and Vmicro are determined by N2 adsorption using BJH and DR methods and listed in Table 1, q0 is model parameter of DA equation, the difference value DV is
served to estimate the contribution of mesopore to dichloromethane and 2-butanone adsorption. Fig. 4a indicates that there is a big deviation between DV and Vmeso and all data are below 45° diagonal line; in addition, it is found in Fig. 4b that DV is directly proportional to Smeso. The results demonstrate that although the mesopores play a significant role in improving the adsorption capacities for medium and high concentrations of VOCs vapor,
190
S. Wang et al. / Journal of Colloid and Interface Science 428 (2014) 185–190
the proper pore size distribution should be considered when designing the adsorbents. The mesopores near to the macropore in size have little contribution to the increase of the surface area of adsorbent. Hence it is less efficient for enhancing adsorption capacity. Taken together, the mesoporous surface area has more advantage to improve the adsorption capacity than mesoporous volume.
improve the adsorption capacity. Since the interaction force in mesopore is weaker than that in micropore, the micro-mesoporous adsorbents exhibited better desorption performance than JT-001, especially for 2-butanone with higher boiling point. Generally, the hypercrosslinked polymeric adsorbent MM-3 with proper microporosity and mesoporosity could have a promising future in VOCs emission controlling for its excellent adsorption and desorption performance.
3.5. Desorption behaviors Acknowledgments It is clearly shown in Fig. 5 that the desorption efficiency of dichloromethane and 2-butanone adsorbed at the relative pressures of 0.1 or 0.8 followed the order: JT-001 < MM-1 < MM2 < MM-3, which was consistent with the mesoporosity of the adsorbents. This result is mainly attributed to the different adsorption mechanism on the four adsorbents. It is well-known that the adsorption energy in the micropores is much stronger compared with that in the mesopores because of the overlapping of adsorption forces from the opposite walls of the micropores. The isotherms in Fig. 1 have shown that the adsorption mechanism of dichloromethane and 2-butanone on JT-001 is predominantly pore-filling in micropore, while that on newly synthesized polymers is the coexistence of pore-filling in micropore and multilayer adsorption and capillary condensation in mesopore. Therefore, compared with micro-mesoporous adsorbents, the interaction forces between the adsorbate and JT-001 are stronger, which is confirmed by the fitting values of parameter E in DA equation. The parameter E is a measure of the adsorption strength between adsorbate and adsorbent. The larger the value of E, the stronger the adsorption force between adsorbate and adsorbent. It is learned in Table S2 that the adsorption characteristic energy E had the following order: JT-001 > MM-1 > MM-2 > MM-3, which was in good agreement with desorption efficiencies of all the adsorbents. Therefore, the mesopore with suitable pore size distribution can enhance the adsorption capacity and desorption efficiency of VOCs. 4. Conclusion A series of well-developed micro-mesoporous hypercrosslinked polymeric adsorbents (MM-1, MM-2, and MM-3) were prepared by suspension polymerization and post-crosslinking process. Adsorption and desorption characteristics of dichloromethane and 2butanone on three micro-mesoporous hypercrosslinked polymers were investigated. The three micro-mesoporous polymeric adsorbents exhibit comparable adsorption capacities with commercial micropore-dominated hypercrosslinked polymeric adsorbent JT001 for dichloromethane and 2-butanone at regions of p/p0 < 0.2. However, they had larger adsorption capacities than JT-001 at higher relative pressure, indicating that the mesopores play a key role in improving the adsorption capacity for medium–high concentration VOCs. Further analysis revealed that compared to mesoporous volume, mesoporous surface area was favorable to
This research was financially funded by National Natural Science Foundation of China (Grant No. 51078180), New Century Excellent Talents in University (Grant No. NCET-11-0230), Qing Lan Project of Jiangsu Province, and Program for Changjiang Scholars Innovative Research Team in University. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcis.2014.04.055. References [1] A.K. Ghoshal, S.D. Manjare, J. Loss Prevent. Proc. Ind. 15 (2002) 413–421. [2] US EPA, Choosing an adsorption system for VOC: carbon, zeolite, or polymers?, vol. 5, 456/F-99-004, 1999. [3] R.A. Zerbonia, C.M. Brockmann, P.R. Peterson, D. Housley, J. Air Waste Manage. Assoc. 51 (2001) 1617–1627. [4] J.K. Brennan, T.J. Bandosz, K.T. Thomson, K.E. Gubbins, Colloid Surf. A: Physicochem. Eng. Aspects 187–188 (2001) 539–568. [5] F. Cosnier, A. Celzard, G. Furdin, D. Bégin, J.F. Marêché, Adsorpt. Sci. Technol. 24 (2006) 215–228. [6] R.H. Bradley, Adsorpt. Sci. Technol. 29 (2011) 1–28. [7] R.C. Bansal, M. Goyal, Activated Carbon Adsorption, CRC Pres, Taylor & Francis Group, LLC, Boca Raton, 2005. [8] M. Inagaki, New Carbon Mater. 24 (2009) 193–232. [9] J.H. Yun, D.K. Choi, H. Moon, Chem. Eng. Sci. 55 (2000) 5857–5872. [10] M.J. Lashaki, M. Fayaz, H. Wang, Z. Hashisho, J.H. Philips, J.E. Anderson, M. Nichols, Environ. Sci. Technol. 46 (2012) 4083–4090. [11] V.A. Davankov, M.P. Tsyurupa, React. Polym. 13 (1990) 7–42. [12] M.P. Tsyurupa, V.A. Davankov, React. Funct. Polym. 66 (2006) 768–779. [13] S.J. Xu, Y.L. Luo, B. Tan, Rapid Commun. 34 (2013) 471–484. [14] R. Dawson, A.I. Cooper, D.J. Adams, Prog. Polym. Sci. 37 (2012) 530–563. [15] C. Long, Q.F. Li, Y. Li, Y. Liu, A.M. Li, Q.X. Zhang, Chem. Eng. J. 160 (2010) 723– 728. [16] C. Long, Y. Li, W.H. Yu, A.M. Li, J. Hazard. Mater. 203–204 (2012) 251–256. [17] J. Wu, L. Zhang, C. Long, Q.X. Zhang, J. Chem. Eng. Data 57 (2012) 3426–3433. [18] C. Long, Y. Li, W.H. Yu, A.M. Li, Chem. Eng. J. 180 (2012) 106–112. [19] C. Long, P. Liu, Y. Li, A.M. Li, Q.X. Zhang, Environ. Sci. Technol. 45 (2011) 4506– 4512. [20] M.H. Tai, B. Saha, M. Streat, React. Funct. Polym. 41 (1999) 149–161. [21] V.V. Podlesnyuk, J. Hradil, E. Králová, Sorption of organic vapours by macroporous and hypercrosslinked polymeric adsorbents, React. Funct. Polym. 42 (1999) 181–191. [22] E.J. Simpson, W.J. Koros, R.S. Schechter, Ind. Eng. Chem. Res. 35 (1996) 4635– 4645. [23] L. Zhang, X.F. Song, J. Wu, C. Long, A.M. Li, Q.X. Zhang, Chem. Eng. J. 192 (2012) 8–12. [24] H. Grajek, J. Chromatogr. A 1145 (2007) 1–50. [25] G.O. Wood, Carbon 39 (2001) 343–356.
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Enhanced adsorption and desorption of VOCs vapor on novel micro-mesoporous polymeric adsorbents Shuangshuang Wang, Liang Zhang, Chao Long ⇑, Aimin Li State Key Laboratory of Pollution Control and Resource Reuse, School of the Environment, Nanjing University, 163 Xianlin Road, Nanjing 210023, China
a r t i c l e
i n f o
Article history: Received 27 December 2013 Accepted 25 April 2014 Available online 2 May 2014 Keywords: Hypercrosslinked resin Micro-mesopore Adsorption Desorption VOCs
a b s t r a c t To enhance adsorption and desorption efficiency of volatile organic compounds from gas streams, we synthesized a series of well-developed micro-mesoporous hypercrosslinked polymeric adsorbents (MM-1, MM-2, and MM-3). The adsorption and desorption performance of dichloromethane and 2-butanone on newly synthesized adsorbents was investigated and compared with commercial microporedominated hypercrosslinked polymeric adsorbent (JT-001). The contributions of micropore and mesopore to adsorption and desorption of dichloromethane and 2-butanone on polymeric adsorbents were well elucidated. Consequently, the adsorbent MM-3 had been sorted out with high BET surface area (1606 m2/g), large micropore and mesopore volumes (0.562 mL/g and 1.046 mL/g, respectively). The MM-3 exhibited the similar adsorption capacities with JT-001 for dichloromethane and 2-butanone at regions of p/p0 < 0.2, but had higher adsorption capacities than JT-001 at high relative pressures. The largest adsorption capacities of MM-3 for dichloromethane and 2-butanone at 308 K were 1345.3 mg/g and 853.5 mg/g, respectively, which are about 1.78 and 1.88 times those of JT-001 under the same condition. Furthermore, the MM-3 exhibited higher desorption efficiencies than JT-001, especially for 2-butanone with a higher boiling point. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction Nowadays, the emission of volatile organic compounds (VOCs) has gained more and more attention not only for their damage to human health and environment but for economic interests as well. Adsorption technology has been recognized as a preferred method for the removal of VOCs because it provides additional benefits from the recovery of the valuable VOCs for reuse [1,2]. The heart of an adsorption process is usually a porous solid adsorbent. Although an explosion has been seen in the development of new nanoporous adsorption materials in the past two decades, the activated carbon is still the major commercial sorbent used for gas stream processing. The activated carbon provides excellent adsorption capacity, but it has been recognized that it owns some disadvantages such as fire risk, pore blocking and hygroscopicity [3–6]. In addition, the activated carbons are predominantly consisted of micropores (<2 nm in size) with the specific surface area and pore volume on the order of 800–1500 m2/g and 0.2–0.6 mL/g, respectively [7,8]. Such microporous adsorbents are desirable for the adsorption of low concentration of VOCs, but inefficient to adsorb
⇑ Corresponding author. E-mail address: [email protected] (C. Long). http://dx.doi.org/10.1016/j.jcis.2014.04.055 0021-9797/Ó 2014 Elsevier Inc. All rights reserved.
medium–high concentration of VOCs owing to the limitation of pore volume. Furthermore, it is very difficult to regenerate adsorbed VOCs due to the strong dispersive force resulting from the overlap of attractive forces from opposite pore walls of micropore [9,10]. Hypercrosslinked polymeric adsorbent represents a class of predominantly microporous organic materials [11–14] and has gained increasing interest for removing the VOCs from gas steam currently. It is found that hypercrosslinked polymeric adsorbent had comparable adsorption capacities with activated carbon for VOCs [15–17]. However, it should be noted that commercial hypercrosslinked polymeric adsorbent is typical of adsorbent with a predominant microporous structure in the regions of pore size 0.5–2 nm [16,18–20], and has similar properties to activated carbons such as low regeneration efficiency and insufficient adsorption capacities for medium–high concentration VOCs. In addition, although the commercial hypercrosslinked polymeric adsorbent contains a certain amount of mesopore and macropore, these pores are mostly in the range of 30–70 nm in diameter, which do not have a considerable contribution to adsorption. In comparison with hypercrosslinked polymeric adsorbent, macroporous polymeric adsorbent has the considerably low adsorption capacities for VOCs at the ranges of low relative pressure. However, it can exhibit high adsorption capacities for VOCs at high relative pressure due to the
186
S. Wang et al. / Journal of Colloid and Interface Science 428 (2014) 185–190
existence of mesopore [21,22]. Therefore, the control of pore structure of adsorbents is strongly required in fields of application to obtain the good adsorption and desorption efficiency for VOCs. Recently, we reported a novel hypercrosslinked polymeric adsorbent with micro-mesoporous structure in a brief communication [23]. Compared with commercial hypercrosslinked polymeric adsorbent, the improved adsorption and desorption performance of n-hexane was found at higher relative pressure. However, the equilibrium isotherms of VOCs on the micro-mesoporous polymeric adsorbents were not investigated; the contributions and mechanism of micropore and mesopore to adsorption and desorption of VOCs on the adsorbents had not been elucidated yet. Especially, the available information on the role of mesopore in adsorption and desorption of VOCs vapors is very limited. In this work, we successfully prepared a series of controllable micro-mesoporous hypercrosslinked polymeric adsorbents by optimizing the synthetic process reported in the previous communication [23]. The adsorption–desorption performance of dichloromethane and 2-butanone onto a series of micro-mesoporous hypercrosslinked polymeric adsorbents was investigated. Moreover, the effect of micropore and mesopore on adsorption and desorption of the selected VOCs was clarified. 2. Experimental 2.1. Synthesis of micro-mesoporous polymeric adsorbents The micro-mesoporous hypercrosslinked polymeric adsorbents were prepared by optimizing the synthetic process reported in the previous brief communication [23]. Briefly, first, low-crosslinked copolymer was synthesized by suspension polymerization using vinylbenzyl chloride as the monomer, divinylbenzene as the cross-linking agent, and a mixture of toluene and n-heptane as the inert porogen. Three different mass ratios of toluene and nheptane listed in Table S1 in the Supporting Information (SI) were used to control the pore structure of polymers. Then, micro-mesoporous hypercrosslinked polymeric adsorbent was prepared via a post-crosslinking of low-crosslinked poly (vinylbenzyl chloridedivinylbenzene). The obtained samples in this study were named as MM-1, MM-2, and MM-3, respectively. The commercial hypercrosslinked polymeric adsorbent (JT-001) was supplied by Jiangsu N&G Environment Co. 2.2. Characterization of adsorbents Nitrogen adsorption–desorption isotherms were measured using an ASAP 2020 (Micromeritics Instrument Co., USA) at 77 K. The specific surface area was determined using the N2 isotherms data by means of the BET equation. The total pore volume (Vtotal) was estimated to be the liquid volume of nitrogen at a relative pressure of about 0.98. The micropore volume (Vmicro) and mesopore volume (Vmeso) were calculated from the N2 isotherms data by Dubinin–Radushkevich (DR) and BJH methods, respectively. The pore size distributions were calculated by applying the density functional theory (DFT) to N2 isotherm data. 2.3. Adsorption and desorption experiments The adsorption equilibrium of dichloromethane and 2-butanone onto MM-1, MM-2, MM-3 and JT-001 has been conducted using a gravimetric method. The detailed experimental apparatus and adsorption procedure have been described previously [19]. In brief, samples were precisely weighed out and then charged into the adsorption column made of glass. The carrier gas containing a known concentration of VOCs vapor was passed through the column until
the VOCs concentration becomes constant and stable. The change of VOCs concentration in the effluent steam from the adsorption column was measured by using gas chromatography (GC, 9790, FULI, China) with a flame ionization detector (FID) and recorded by a computer. The adsorption equilibrium was attained when the exit concentration became equal to the inlet concentration. The equilibrium adsorption amount of VOCs onto adsorbent was equal to the weight change of adsorbent before and after adsorption process. Here, a high precision microbalance (BS224S, Sartorius, Germany) was adopted as the weighing device. After the adsorption procedure, desorption behavior was investigated by connecting the adsorption column to a vacuum pump under vacuum of 0.005 MPa. The desorbed amount was equal to the weight change of adsorbent before and after the desorption process. 3. Results and discussion 3.1. Characterization of adsorbents The N2 adsorption–desorption isotherms at 77 K of four adsorbents MM-1, MM-2, MM-3 and JT-001 are demonstrated in Fig. S1 in Supporting Information. It is observed that at lower relative pressure (p/p0) below 0.05, the nitrogen uptake increases sharply with the increment of relative pressure, proving the existence of micropore structure in the four polymeric adsorbents. Moreover, visible hysteresis loop means that all the adsorbents contain mesopore. According to IUPAC classification, the adsorption–desorption isotherms of JT-001 and MM-1 are close to type I, reflecting the domination of micropores in the pore structure; however, the accelerated uptake at p/p0 1 means that JT-001 contains the larger-sized mesopores or macropores. The N2 adsorption–desorption isotherms of the newly synthesized adsorbents MM-2 and MM-3 are the combination of type I and IV with remarkable hysteresis loops, suggesting the distinct micro-mesoporous structures. In order to understand more clearly the pore structure of adsorbents, the pore size distributions of four adsorbents are shown in Fig. 1; the textural properties of all the adsorbents studied are included in Table 1. It is significantly learned that the four adsorbents have nearly similar micropore volume. Although the mesopore volume of JT-001 is 0.528 mL/g, the mesopores are mainly distributed in the regions of 20–50 nm near the macropores. In contrast, the mesopores of newly synthesized adsorbents are in pore size of 2–10 nm near the micropore. Therefore, the newly synthesized adsorbents MM-1, MM-2 and MM-3 possess higher mesopore surface area (317 m2/g, 716 m2/g and 1087 m2/g, respectively) than JT-001. Overall, all three newly synthesized adsorbents contain well-developed pore structure in both regions of micropore and mesopore. 3.2. Adsorption performance Fig. 2 shows the adsorption isotherms of dichloromethane and 2-butanone on MM-1, MM-2, MM-3 and JT-001 at temperatures of 308 K and 323 K, respectively. To clearly clarify the adsorption mechanism, volumetric adsorption capacity qv is used by the unit of mL/g, which is equal to the ratio of equilibrium adsorption capacity (mg/g) to the adsorbate density in the liquid phase (g/ mL). The adsorption amounts of both dichloromethane and 2-butanone rose as the adsorption temperature decreases, suggesting that the adsorption is an exothermic process. Compared with dichloromethane, the higher uptake of 2-butanone at the same temperature and vapor pressure may result from the stronger polarizability of 2-butanone (16.424 and 20.681 mL/mol for dichloromethane and 2-butanone, respectively), which is the first and usually most important contribution to dispersion forces
S. Wang et al. / Journal of Colloid and Interface Science 428 (2014) 185–190
187
Fig. 1. Pore size distribution of MM-1, MM-2, MM-3 and JT-001.
Table 1 Textural properties of MM-1, MM-2, MM-3 and JT-001. Adsorbent
JT-001
MM-1
MM-2
MM-3
BET surface area (m2/g) BJH surface area (m2/g) Vmicro (mL/g) Vmeso (mL/g) Vtotal (mL/g) Average pore diameter (nm)
1315 259 0.573 0.528 1.075 2.757
1373 317 0.539 0.241 0.744 2.166
1379 716 0.517 0.863 1.261 3.657
1606 1087 0.562 1.046 1.424 3.548
between the adsorbate and the surface of adsorbent [24]. Meanwhile, three micro-mesoporous polymeric adsorbents exhibited the comparable adsorption capacities with JT-001 for dichloromethane and 2-butanone at regions of p/p0 < 0.2, which may be attributed to almost the same micropore volume of the four adsorbents. However, it is clearly seen that the three newly synthesized micro-mesoporous polymeric adsorbents had larger adsorption capacities than JT-001 at high relative pressure. Among all the adsorbents, the MM-3 showed the largest adsorption capacities of 1.014 mL/g and 1.130 mL/g for dichloromethane and 2-butanone at 308 K, respectively, which are about 1.78 and 1.88 times those of JT-001 obtained under the same experimental conditions. On the other hand, although JT-001 had larger mesopore volume, it should be noted that JT-001 kept an almost constant volumetric adsorption capacities similar to its micropore volume for dichloromethane and 2-butanone at the regions of medium–high relative pressure. However, MM-2 and MM-3 can obtain higher adsorption capacities in excess of their micropore volumes. The results demonstrated that the mesopores with suitable pore size play a key role in the improvement of the adsorption capacity for medium– high concentration of VOCs. 3.3. Modeling of isotherms The equilibrium data were correlated by Dubinin–Astakhov (DA) equation. DA equation can be expressed as follows.
qv ¼ q0 exp½ðe=EÞr
ð1Þ
qv ¼ q=q
ð2Þ
e ¼ RT lnðp0 =pÞ
ð3Þ
where qv is volumetric adsorption capacity (mL/g), q0 is the maximum adsorption volume (mL/g), q is the equilibrium amount (mg/g), q is the adsorbate density in the adsorbed phase assumed to be the same as that in the liquid phase (g/mL), e is the adsorption potential (J/mol) written by Eq. (3), E is the adsorption characteristic energy (kJ/mol), R is a gas constant (8.314 J/molK), T is the absolute temperature (K), p0 is the saturation vapor pressure (kPa), p is the equilibrium vapor pressure (kPa). The fitting parameters of DA equation for adsorption data at each temperature are summarized in Table S2 (Supporting Information). It is shown that the DA model fits the adsorption data reasonably with large correlation coefficient (R2 > 0.993). According to potential theory, it is certain that plots of adsorbed volume (qv) versus the adsorption potential at different temperature should yield a unique characteristic curve that is independent of temperature [25], which can be used to further examine whether Polanyi theory mechanistically captures the adsorption process of compounds by adsorbent. Adsorption characteristic curve of dichloromethane and 2-butanone on JT-001, MM-1, MM-2, and MM-3 is shown in Fig. S2 in SI. As the Polanyi potential theory would predict, they all fell essentially onto a single curve with high correlation coefficient R2 (Table S3 in SI), indicating the availableness of Polanyi theory to describe the adsorption of dichloromethane and 2-butanone onto polymeric adsorbents. In addition, plots of the ratio of the adsorption potential (e) to molar polarizability (pe) versus adsorbed volume (qv) of dichloromethane and 2-butanone onto four adsorbents are shown in Fig. 3. It can be seen clearly from Fig. 3 that plots of e/pe versus qv for different adsorbents had a good superposition at low relative pressure (high values of e/pe), while a big deviation was observed at high relative
188
S. Wang et al. / Journal of Colloid and Interface Science 428 (2014) 185–190
Fig. 2. Equilibrium adsorption data of dichloromethane and 2-butanone onto MM-1, MM-2, MM-3 and JT-001 at 308 K and 323 K.
pressure (low values of e/pe) and this is also affirmed by residual analysis. Such results are directly related to the pore size distribution of four polymeric adsorbents. As we all know, the adsorbate molecules were preferentially adsorbed in micropores by pore-filling at low relative pressure. From Fig. 1 and Table 1, it is learned that four adsorbents had a very similar micropore size distribution and basically the same micropore volume. Therefore, the good superposition of characteristic curves of four adsorbents can be found in the regions of high values of e/pe in Fig. 3. However, at high relative pressure, the adsorbates are adsorbed mainly by multilayer adsorption and capillary condensation; the pore size distribution in the regions of mesopore has significant influence on adsorption capacity, which will be explained in the ensuing para-
Fig. 3. Plots of adsorbed volume (qv) versus normalized adsorption potential density (e/pe) for dichloromethane and 2-butanone onto MM-1, MM-2, MM-3 and JT-001.
graphs. Consequently, the characteristic curves of four adsorbents at high relative pressure (low values of e/pe) had a big discrepancy, resulting from the different pore size distribution and specific surface area of four adsorbents in the regions of mesopores.
3.4. Contribution of mesopores to adsorption The fitting parameters of DA equation calculated from all the experimental data for four adsorbents are given in Table S3. As the data presented in Table S3, the calculated values of q0 of four adsorbents were higher than the micropore volume but lower than the sum of micropore and mesopore volume except for adsorption of dichloromethane on JT-001, implying that the mesopores make a certain of contribution to the adsorption capacity of dichloromethane and 2-butanone. Although the commercial hypercrosslinked polymeric adsorbent JT-001 has larger mesopore volume (0.528 mL/g) than micro-mesoporous adsorbent MM-1, the values of q0 for dichloromethane and 2-butanone are less than MM-1. This result implies that the mesopores on JT-001 made very limited contribution to dichloromethane and 2-butanone uptake, which was confirmed by the equilibrium data of JT-001 and MM-1 at larger relative pressure. The result is probably ascribed to the different mesopore size distribution of two adsorbents. The mesopore diameter of JT-001 falls largely into the region of 30–50 nm near the macropore (Fig. 1). These larger mesopores made a small contribution to adsorption and primarily acted as transport pores. The mesopores of MM-1 were mainly distributed in region of 2–10 nm. Therefore, the multilayer adsorption and following capillary condensation in the mesopore of MM-1 may take place at lower relative pressures. Among the four adsorbents, MM-3 has the largest mesopore volume and suitable pore size distribution, whose mesopores are mainly distributed in region of 2–10 nm approaching to micropore. Therefore, MM-3 had the largest values of q0 (1.305 mL/ g for dichloromethane and 1.526 mL/g for 2-butanone), which are close to the sum volume of micropore and mesopore (1.608 mL/g).
S. Wang et al. / Journal of Colloid and Interface Science 428 (2014) 185–190
189
Fig. 4. Plots Vmeso and Smeso versus the difference value (DV) for adsorption of dichloromethane and 2-butanone adsorption on four adsorbents.
Fig. 5. Desorption curves for dichloromethane and 2-butanone from MM-1 MM-2, MM-3 and JT-001 (desorption temperature: 293 K).
In order to understand the contribution of mesopore of adsorbents to adsorption more clearly, plots of Smeso and Vmeso versus DV (Here DV is the difference value between the maximum adsorption volume q0 and micropore volume Vmicro) are depicted in Fig. 4, respectively. Smeso, Vmeso and Vmicro are determined by N2 adsorption using BJH and DR methods and listed in Table 1, q0 is model parameter of DA equation, the difference value DV is
served to estimate the contribution of mesopore to dichloromethane and 2-butanone adsorption. Fig. 4a indicates that there is a big deviation between DV and Vmeso and all data are below 45° diagonal line; in addition, it is found in Fig. 4b that DV is directly proportional to Smeso. The results demonstrate that although the mesopores play a significant role in improving the adsorption capacities for medium and high concentrations of VOCs vapor,
190
S. Wang et al. / Journal of Colloid and Interface Science 428 (2014) 185–190
the proper pore size distribution should be considered when designing the adsorbents. The mesopores near to the macropore in size have little contribution to the increase of the surface area of adsorbent. Hence it is less efficient for enhancing adsorption capacity. Taken together, the mesoporous surface area has more advantage to improve the adsorption capacity than mesoporous volume.
improve the adsorption capacity. Since the interaction force in mesopore is weaker than that in micropore, the micro-mesoporous adsorbents exhibited better desorption performance than JT-001, especially for 2-butanone with higher boiling point. Generally, the hypercrosslinked polymeric adsorbent MM-3 with proper microporosity and mesoporosity could have a promising future in VOCs emission controlling for its excellent adsorption and desorption performance.
3.5. Desorption behaviors Acknowledgments It is clearly shown in Fig. 5 that the desorption efficiency of dichloromethane and 2-butanone adsorbed at the relative pressures of 0.1 or 0.8 followed the order: JT-001 < MM-1 < MM2 < MM-3, which was consistent with the mesoporosity of the adsorbents. This result is mainly attributed to the different adsorption mechanism on the four adsorbents. It is well-known that the adsorption energy in the micropores is much stronger compared with that in the mesopores because of the overlapping of adsorption forces from the opposite walls of the micropores. The isotherms in Fig. 1 have shown that the adsorption mechanism of dichloromethane and 2-butanone on JT-001 is predominantly pore-filling in micropore, while that on newly synthesized polymers is the coexistence of pore-filling in micropore and multilayer adsorption and capillary condensation in mesopore. Therefore, compared with micro-mesoporous adsorbents, the interaction forces between the adsorbate and JT-001 are stronger, which is confirmed by the fitting values of parameter E in DA equation. The parameter E is a measure of the adsorption strength between adsorbate and adsorbent. The larger the value of E, the stronger the adsorption force between adsorbate and adsorbent. It is learned in Table S2 that the adsorption characteristic energy E had the following order: JT-001 > MM-1 > MM-2 > MM-3, which was in good agreement with desorption efficiencies of all the adsorbents. Therefore, the mesopore with suitable pore size distribution can enhance the adsorption capacity and desorption efficiency of VOCs. 4. Conclusion A series of well-developed micro-mesoporous hypercrosslinked polymeric adsorbents (MM-1, MM-2, and MM-3) were prepared by suspension polymerization and post-crosslinking process. Adsorption and desorption characteristics of dichloromethane and 2butanone on three micro-mesoporous hypercrosslinked polymers were investigated. The three micro-mesoporous polymeric adsorbents exhibit comparable adsorption capacities with commercial micropore-dominated hypercrosslinked polymeric adsorbent JT001 for dichloromethane and 2-butanone at regions of p/p0 < 0.2. However, they had larger adsorption capacities than JT-001 at higher relative pressure, indicating that the mesopores play a key role in improving the adsorption capacity for medium–high concentration VOCs. Further analysis revealed that compared to mesoporous volume, mesoporous surface area was favorable to
This research was financially funded by National Natural Science Foundation of China (Grant No. 51078180), New Century Excellent Talents in University (Grant No. NCET-11-0230), Qing Lan Project of Jiangsu Province, and Program for Changjiang Scholars Innovative Research Team in University. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcis.2014.04.055. References [1] A.K. Ghoshal, S.D. Manjare, J. Loss Prevent. Proc. Ind. 15 (2002) 413–421. [2] US EPA, Choosing an adsorption system for VOC: carbon, zeolite, or polymers?, vol. 5, 456/F-99-004, 1999. [3] R.A. Zerbonia, C.M. Brockmann, P.R. Peterson, D. Housley, J. Air Waste Manage. Assoc. 51 (2001) 1617–1627. [4] J.K. Brennan, T.J. Bandosz, K.T. Thomson, K.E. Gubbins, Colloid Surf. A: Physicochem. Eng. Aspects 187–188 (2001) 539–568. [5] F. Cosnier, A. Celzard, G. Furdin, D. Bégin, J.F. Marêché, Adsorpt. Sci. Technol. 24 (2006) 215–228. [6] R.H. Bradley, Adsorpt. Sci. Technol. 29 (2011) 1–28. [7] R.C. Bansal, M. Goyal, Activated Carbon Adsorption, CRC Pres, Taylor & Francis Group, LLC, Boca Raton, 2005. [8] M. Inagaki, New Carbon Mater. 24 (2009) 193–232. [9] J.H. Yun, D.K. Choi, H. Moon, Chem. Eng. Sci. 55 (2000) 5857–5872. [10] M.J. Lashaki, M. Fayaz, H. Wang, Z. Hashisho, J.H. Philips, J.E. Anderson, M. Nichols, Environ. Sci. Technol. 46 (2012) 4083–4090. [11] V.A. Davankov, M.P. Tsyurupa, React. Polym. 13 (1990) 7–42. [12] M.P. Tsyurupa, V.A. Davankov, React. Funct. Polym. 66 (2006) 768–779. [13] S.J. Xu, Y.L. Luo, B. Tan, Rapid Commun. 34 (2013) 471–484. [14] R. Dawson, A.I. Cooper, D.J. Adams, Prog. Polym. Sci. 37 (2012) 530–563. [15] C. Long, Q.F. Li, Y. Li, Y. Liu, A.M. Li, Q.X. Zhang, Chem. Eng. J. 160 (2010) 723– 728. [16] C. Long, Y. Li, W.H. Yu, A.M. Li, J. Hazard. Mater. 203–204 (2012) 251–256. [17] J. Wu, L. Zhang, C. Long, Q.X. Zhang, J. Chem. Eng. Data 57 (2012) 3426–3433. [18] C. Long, Y. Li, W.H. Yu, A.M. Li, Chem. Eng. J. 180 (2012) 106–112. [19] C. Long, P. Liu, Y. Li, A.M. Li, Q.X. Zhang, Environ. Sci. Technol. 45 (2011) 4506– 4512. [20] M.H. Tai, B. Saha, M. Streat, React. Funct. Polym. 41 (1999) 149–161. [21] V.V. Podlesnyuk, J. Hradil, E. Králová, Sorption of organic vapours by macroporous and hypercrosslinked polymeric adsorbents, React. Funct. Polym. 42 (1999) 181–191. [22] E.J. Simpson, W.J. Koros, R.S. Schechter, Ind. Eng. Chem. Res. 35 (1996) 4635– 4645. [23] L. Zhang, X.F. Song, J. Wu, C. Long, A.M. Li, Q.X. Zhang, Chem. Eng. J. 192 (2012) 8–12. [24] H. Grajek, J. Chromatogr. A 1145 (2007) 1–50. [25] G.O. Wood, Carbon 39 (2001) 343–356.