9588
J. Phys. Chem. C 2010, 114, 9588–9595
Capturing Nitrosamines by Zeolite MCM-22: Effect of Zeolite Structure and Morphology on Adsorption Jing Yang,† Yu Zhou,† Jia Yuan Yang,† Wei Gang Lin,† Ya Jing Wu,‡ Na Lin,† Jun Wang,‡ and Jian Hua Zhu*,† Key Laboratory of Mesoscopic Chemistry of MOE, School of Chemistry and Chemical Engineering, Nanjing UniVersity, Nanjing 210093, China, and State Key Laboratory of Materials-Oriented Chemical Engineering, College of Chemistry and Chemical Engineering, Nanjing UniVersity of Technology, 210009, China ReceiVed: December 8, 2009; ReVised Manuscript ReceiVed: April 29, 2010
The activity of zeolite MCM-22 in trapping nitrosamines, a class of well-known carcinogenic environmental pollutants, is reported in this article for the first time. MCM-22 possesses a set of unique porous structures and morphologies, making it possible to trap both volatile nitrosamines and bulky tobacco specific nitrosamines. Liquid adsorption and instantaneous gaseous adsorption methods have been employed to study the impact of morphology on MCM-22’s ability in adsorbing nitrosamines in both gaseous and liquid media. As-synthesized MCM-22 was subjected to different treatments to induce morphological changes. SEM revealed a special rose-like appearance. The effects of these morphological modifications on MCM-22’s adsorption capacities was studied and compared to NaY and NaZSM-5. The results obtained seem to suggest that enhanced collision probability between adsorbate and adsorbent may have an important role to play for MCM-22. Furthermore, the treatments created mesopores in MCM-22 that enhance mass transport within its hierarchical structure. 1. Introduction Zeolites have been extensively used in the petrochemical and chemical industry as adsorbents, catalysts, and catalyst supports, and their application has recently been extended to life science and environment protection.1,2 However, some requirements of environment protection challenge the traditional zeolite materials because it is normal in applications to treat a large amount of gas or fluid flow within a short time in order to selectively capture the target chemicals. In other situations, zeolites have to face a group of targets with various molecular structures in a complex mixture. One typical example is to trap carcinogens like nitrosamines in environmental tobacco smoke (ETS).3 Nitrosamines are well-known to be potent carcinogens,4 and they can induce the cancer or tumor in the laboratorial animals; even a trace amount of nitrosamines is able to affect the lungs, larynx, oral cavity, pharynx, pancreas, kidneys, and bladder.5 There are two kind of nitrosamines, volatile N-nitrosamines (VNA) and tobacco specific N-nitrosamines (TSNAs) in tobacco smoke. The former includes N-nitrosodimethylamine (NDMA) and Nnitrosopyrrolidine (NPYR); they possess a high volatility and a relatively small molecular size. The latter contains N′nitrosonornicotine (NNN), N′-nitrosoanatabine (NAT), 4-methylnitrosarnino-1,3-pyridyl-1-butanone (NNK), and N′-nitrosoanabasine (NAB), and they have a relatively large molecular diameter. For example, the structure of NNN looks like one of the H atoms in the five-membered ring of NPYR is replaced by pyridine, consequently, its transect size (0.80 nm) is larger than NPYR (0.42 nm). Both NPYR and NNN have the functional group of N-NO in which the oxygen atom possesses the negative charge, and they can be adsorbed by zeolite Y,6 ZSM-5,7 and A8 due to the electrostatic interaction with the * To whom correspondence should be addressed. Tel.: +86-25-83595848. Fax: +86-25-83317761. E-mail:
[email protected]. † Key Laboratory of Mesoscopic Chemistry of MOE. ‡ State Key Laboratory of Materials-Oriented Chemical Engineering.
cation in the zeolite. This feature enables nitrosamines to be selectively removed by zeolite in gas stream, solution, and even cigarette smoke that contains thousands of components.3,6,8-11 Of the nitrosamines, some 40-50% could be eliminated from the mainstream smoke when a zeolite was placed directly on the tobacco fibers of cigarettes to catalyze the degradation of carcinogens in smoke,3 while a zeolite-like adsorbent could reduce the TSNA level of mainstream smoke in the range of 30∼60% once it was added into the cigarette filter.12 The adsorption isotherms of nitrosamines on a zeolite were able to be fitted to the Freundlich equation,9,13 and molecular modeling of nitrosamines adsorbed on H-ZSM-5 suggested the selective recognition of nitrosamine by the zeolite.14 Among the commercial zeolites assessed by laboratorial adsorption experiments, NaY exhibited the highest performance;7 and modification with copper oxide, cobalt oxide, or zirconia could further enhance the adsorptive capability of zeolite through strengthening the electrostatic attraction toward nitrosamines.15-19 TSNA are bulky nitrosamines with the diameter of around 0.8 nm so that it is difficult for most zeolites to efficiently adsorb TSNA,20 because the small pore opening of zeolite limits the mass transportation of TSNA within channel.10 Recently, Wei et al. tried to utilize mesoporous silica capturing the TSNA in solution11 and Zhou et al. synthesized the 3D net-linked SBA-15 for trapping the TSNA in tobacco smoke,21 but the high cost of mesoporous silica limited their potential application. Consequently, a new strategy has to be sought to remove both types of nitrosamines in the environment. So far a great deal of effort has been made to develop new functional materials through tuning their pore structures and surface functionality,22,23 but relatively little has been paid to explore the effects of morphology at both micro- and macrolevels on their adsorption properties. Common commercial zeolites such as NaY, NaZSM-5, and NaA have a hexagonal or cubic morphology. New adsorbents with fiber-like morphology have also been synthesized.12,21,24 To explore the effects of special
10.1021/jp9116497 2010 American Chemical Society Published on Web 05/10/2010
Capturing Nitrosamines by Zeolite MCM-22 morphology further in relation to their performance in trapping nitrosamines, zeolite MCM-22 has been chosen in this work because it possesses a rose-like morphology.25 There are many factors such as suitable pore sizes, ionic states as well as morphology governing the adsorption of nitrosamines,7,12,21 and it would also be desirable to optimize these factors in a systematic manner.24,26 Zeolite MCM-22 has two independent pore systems, one involves two-dimensional, sinusoidal channels (0.4 × 0.5 nm) with circular 10-membered rings like ZSM-5,27 and another consists of 12-membered large cylindrical supercages (0.7 × 0.7 × 1.8 nm) similar to that of NaY.28 In other words, MCM22 possesses a similar structure to ZSM-5 but with a large supercage, therefore, it may be possible for it to trap large molecules than medium aperture (10 MR) if the supercages are open and accessible. To modify its structure and pore size, desilication technique was chosen to create mesopore in MCM22 because it has been successfully utilized in NaZSM-5 zeolite to promote the liquid adsorption of nitrosamines.29 The development of the hierarchical porosity was obtained by preferential extraction of the Si framework by hydrolysis in the presence of OH- ions, while the aluminum in the framework positions could form controlled mesopores.30 2. Experimental Section There are five samples of MCM-22 zeolites used in this study. The sample M22a (which is the as-received MCM-22) is from China University of Petroleum. In the first treatment, the M22a was stirred in 1 M NaCl for six times at 353 K each lasting for 2 h, the material was then washed with deionized water to remove Cl- ion with the monitor of AgNO3 solution until no deposit of AgCl was observed filtration solution, and finally was calcined in air. The sample thus obtained is denoted as M22b. To make alkaline-treated sample, 1.5 g M22a was mixed in 150 mL of 0.1 mol L-1 NaOH and stirred at 358 K for 5 h. The resulting solid was filtrated, washed, and dried at 373 K. After calcinations, the sample is denoted as M22c. The Si/Al ratio of sample was determined by an ARL-9800XP+ X-ray fluorescence (XRF) spectrometer. On the other hand, the morphology control samples M22d and M22e were synthesized in situ as follows: 1.66 g of sodium aluminate (41% Al2O3, Shanghai Chem. Reagent Co., AR) and 0.41 g of sodium hydroxide (96%, Shanghai Chem. Reagent Co., AR) were dissolved in 110 mL of deionized water. Then, 13.78 mL HMI (98%, Jiangsu Fengyuan Biochemical Co., LR) was added dropwise to the stirred mixture. To the well-dispersed mixture, 30 g of colloidal silica (40% SiO2, Zhejiang Yuda Chem. Co., LR) was slowly added with vigorous stirring. The synthesis gel was aged at room temperature for 24 h. The composition of the above mixture in terms of oxide molar ratio was 0.075 Na2O/ SiO2/0.033 Al2O3/0.6 HMI/35 H2O. Syntheses were carried out in a Teflon-lined autoclave under static condition at 431 K for 8 days. The products were recovered, filtered, washed with deionized water, and finally dried at 373 K overnight to obtain the as-synthesized products. The M22d sample was obtained by calcining the as-synthesized materials for 20 h in air at 823 K. Sample M22e was synthesized in a similar procedure but using sodium metasilicate as the silica source.31 M22d and M22e had the Si/Al molar ratio of 16 according to the ICP analysis. NaY was obtained from Wenzhou Catalysts Factory and NaZSM-5 from Nankai University (China).7 For comparison, an activated carbon from Chemviron Carbon (CAS number 7440-44-0, Belgium) with 1.0 nm average pore size was also used in this study. N-Nitrosopyrrolidine (NPYR) was obtained
J. Phys. Chem. C, Vol. 114, No. 21, 2010 9589 from Sigma. N′-Nitrosonornicotine (NNN) was purchased from Toronto Research Chemicals. X-ray diffraction (XRD) patterns of the sample were recorded on an ARLXTRA diffractometer with Cu KR radiation in the 2θ range from 5 to 80°. Scanning electron microscopy (SEM) was performed with a HITACHI S4800 microscope. For SEM analysis, the samples were coated with Au film. Transmission electron microscopy (TEM) analysis was performed on a FEI Tecnai G2 20 S-TWIN electron microscope operating at 200 kV. Nitrogen adsorption and desorption isotherms were measured at 77 K by a Micromeritics ASAP 2000 volumetric adsorption analyzer.22 Room temperature 27Al MAS NMR experiments were performed with a 4.0 mm MAS probe on a Bruker Avance III spectrometer at 9.4 T and 104.263 MHz Larmor frequency. Powdered samples were packed inside zirconia MAS rotors and spun at 14 kHz. The 27Al MAS spectra were obtained using a single hard pulse with radiofrequency (rf) of 83 kHz. To obtain more quantitative spectra, an rf tip angle of less than one-third of the solid π/2 pulse duration was employed, resulting in 0.3 µs long pulses for single-pulse experiments. Recycle delays of 0.2 s were used for 27Al MAS experiments. 27Al chemical shifts were referenced to the external standard (0.1 M Al(NO3)3), which was set at 0 ppm. A standard triple-quantum MAS pulse sequence with two hard pulses followed by a z-filter was used to acquire 27Al MQMAS spectra. The optimized lengths of the first and second pulses with an rf power of 83 kHz were about 4.5 and 1.5 µs. The length of the third pulse was about 12.0 µs, with an rf power of 6.6 kHz. The MQMAS experiments were performed in a rotor synchronized fashion to improve sensitivity and line shapes. The data were collected with a recycle delay of 0.1 s. Liquid adsorption of nitrosamines was carried out using a batch method. A total of 20 mg of powder sample were added into a tube containing 1.1 mL of a dichloromethane solution of NNN (100 µL). After the static adsorption of 24 h at 277 K, the solid was separated and the residual NNN in solution was detected by an improved spectrophotometric method,6 in which the HBr in glacial acetic acid chemically denitrosated the nitrosamines to liberate gaseous product. The formed gaseous product was oxidized to NO2 and then converted to NO2- that was finally detected by a Digital Visible Spectrophotometer at 540 nm.8,9,13 Gaseous adsorption of volatile nitrosamines was performed in a stainless steel microreactor.7 Five milligrams sample (20-40 meshes) were added in the reactor and directly heated to 338 or 453 K in the flow of hydrogen, and the nitrosamine solution was pulse injected with the amount of 2 µL each time. After being gasified in the injector, the nitrosamines were pushed to pass through the zeolite layer then to packed column, and the decrement in the ratio of solute to solvent represented the amount of nitrosamines adsorbed by zeolite.7 3. Results 3.1. Characterization of Modified MCM-22. Figure 1 presents the XRD patterns of M22a and M22c samples. The M22c sample exhibited a near identical diffraction pattern to that of the parent zeolite, but the intensity for some of the peaks was slightly decreased as a result of the treatment. It was clear that the long-range ordering and the framework of MCM-22 were preserved. The two small peaks at 6.4 and 9° indicated the presence of a small quantity of zeolite mordenite in the assynthesized M22a sample. As shown in Figure 2A, the M22a sample exhibited an isotherm resembling type I in N2 adsorption/
9590
J. Phys. Chem. C, Vol. 114, No. 21, 2010
Figure 1. XRD patterns of MCM-22 and its alkaline-treated analogue M22c.
desorption at 77 K with a marked hysteresis at high p/p0 values, indicating a relatively high external surface area with little mesoporosity.32 Desilication of MCM-22 made the hysteresis extend to low p/p0 values (∼0.2), while the loop became much more obvious (Figure 2A). These observations mirrored the formation of slit-like mesopores on M22c,33 which is similar to that of desilicated ZSM-5 zeolite, where the newly formed, open mesopores are found to connect to the external surface.34 Figure 2B shows the pore size distribution of M22a and M22c samples. Because the inherent micropore of M22a was invisible, there was a weak peak around 3.8 nm reflecting the existence of some mesoporosity. The mesopore volume of 0.18 cm3 g-1 (Table S1) was attributed to the aggregates of plate-like particles giving rise to slip-shaped pores.33 A sharp peak emerged on the profile of M22c at 3.5 nm, indicating the generation of mesopores by the alkaline treatment. This is evidenced by the broad peak around 30 nm was enlarged on M22c in comparison with M22a (Figure 2B). Desilication of M22a enhanced the pore volume from 0.36 to 0.41 cm3 g-1, while the Si/Al molar ratio was lowered from 10 to 6 according to the analysis of XRF. This means that about 5.7 wt % of Si in MCM-22 was dissolved in the NaOH solution. Consequently, the microporous volume of M22a decreased from 0.18 to 0.03 cm3 g-1, while the mesoporous volumes increased from 0.18 to 0.38 cm3 g-1. 27 Al MAS NMR was used to determine the coordination, the local structure, as well as the geometry of specific Al species in the zeolites.35 Figure 3 displays the one-dimensional 27Al MAS NMR spectra of M22a and M22c samples in which the resonance in the 54-68 ppm region reflects the average environment of an Al atom in the tetrahedral framework,36 and
Yang et al.
Figure 3.
27
Al MAS NMR spectra of zeolite M22a and M22c samples.
the peak at about 0 ppm is ascribed to an extraframework octahedral (six-coordinated) Al species.37 The 56 ppm peak remained on the M22c sample, but the 0 ppm peak disappeared. This was not unexpected, because the six-coordinated Al could be redistributed into the framework of MCM-22 once the Si atoms were leached, and thus, the coordination changed from a 6-fold to a tetrahedral coordination. In addition, some sixcoordinated Al might be removed during the treatment. The asymmetric shape of the four-coordinated Al signal might originate from the overlap of different resonances and quadrupolar effects.38 To provide further evidence to the identity of the 56 ppm peak, we carried out two-dimensional 27Al MQ MASNMR experiments in which the second-order quadrupolar line broadening was removed and the chemical shift dispersion was amplified.39 By refocusing the second-order quadrupolar anisotropy, a high-resolution spectrum in the second dimension was achieved (Figure 4) to resolve the overlapped peak at about 56 ppm. Three distinct Al framework species, designated A, B, and C, were presented in the tetrahedral region, while the fourth species at 0 ppm (D) was in an octahedral environment. In fact, the signal corresponding to four-coordinated Al species consisted of a major resonance at 56 ppm (B) along with shoulders at about 49 (A) and 61 ppm (C). The structure of MCM-22 had eight crystallographically distinct T sites that could be divided into three groups;36 sites T6 and T7 corresponded to the signals at 49 ppm, and site T2 was attribute to 61 ppm; while sites T1, T3, T4, T5, and T8 gave rise to a signal at 56 ppm. Three fourcoordinated Al peaks at 49, 56, and 61 ppm were observed on sample M22c, while the six-coordinated peak at about 0 ppm disappeared, confirming the occurrence of Si extraction and Al coordination transformation aforementioned.
Figure 2. (A) N2 adsorption (open symbols) and desorption (full symbols) isotherms at 77 K and (B) pore size distribution of MCM-22 and its alkali-treated analogue M22c.
Capturing Nitrosamines by Zeolite MCM-22
Figure 4.
27
J. Phys. Chem. C, Vol. 114, No. 21, 2010 9591
Al MQMAS NMR spectra of zeolite M22a (A) and M22c (B).
Figure 6. HRTEM micrographs of M22a (A) and M22c (B) samples.
Figure 5. SEM images of (A) M22a, (B) M22b, (C) M22c, (D) M22d, (E) M22e, and (F) NaY zeolites.
M22a has a lamellar or platelet-like morphology, and these platelets are closely packed to form the rose-like structure25 and then aggregate to a ball, as seen in Figure 5A, which is quite different from commercial zeolites such as NaY (Figure 5F), which has a cubic morphology. After the sample was stirred in NaCl solution (M22b), its “rose petals” were scattered, and the platelets were loosely piled up (Figure 5B). Comparing Figure 5A and C, it was clear that the effect of the alkaline treatment on the morphology is significant, resulting in the multiple, partially collapsed zeolite grains. Figure 5D demonstrates the morphology of sample M22d, whose spherical particle was formed by a tight packing of many platelets. For M22e, its roselike structure was much more loose (Figure 5E). The TEM images of M22a and M22c samples in Figure 6 revealed some internal changes in the particle structure. M22a was a lamellar crystal with a regularly repeating interlayer (Figure 6A). After the alkali-treatment, the newly formed mesopores could be clearly viewed on the M22c sample. It appears that the formation of the mesopore preferentially initiated along the boundaries between platelets.30 This process has led to some dissolution of the parent framework.33
Figure 7. Liquid adsorption isotherms of NPYR by different porous samples in CH2Cl2 solution at 277 K.
3.2. Adsorption of Nitrosamines. Figure 7 describes the liquid adsorption of NPYR by four zeolites at 277 K, in which M22a shows the highest adsorption capacity but NaZSM-5 has the lowest one. M22a, M22b, and NaZSM-5 samples reached their maximum adsorption capacity around 5 min, while the NaY zeolite required 15 min due to the different pore size. The pore diameter of NaZSM-5 (0.54 × 0.56 nm) was close to the size of NPYR (0.54 × 0.42 nm); once the NPYR molecule was adsorbed in the channel of NaZSM-5, its movement was strictly limited so it had to diffuse frontward but could not diffuse outside of the channel. Therefore, this delicate geometric confinement to the nitrosamines will be characterized by the time required for achieving adsorption equilibrium, while the slightly wider channel of NaY made NPYR easily diffuse out. Based on these uptake curves, the distribution coefficient Kd was calculated from the equation: Kd ) rV/(100 - r)m,40 where V/m was the ratio of liquid volume to solid mass and r was the removal ratio of NPYR by zeolite. Kd expresses the ratio of NPYR to be adsorbed in the solid over that remaining in
9592
J. Phys. Chem. C, Vol. 114, No. 21, 2010
Yang et al.
TABLE 1: Distribution Coefficients and Lagergren Constants of NPYR Adsorption by Different Porous Samplesa
TABLE 3: Freundlich Constants of the Adsorption Isotherms of NPYR in Porous Materials at 338 Ka
adsorbent
adsorbed NPYR (mg g-1)
r (%)
Kd (cm3 g-1)
Kt (min-1)
M22a M22b NaY NaZSM-5
5.97 2.94 2.11 1.84
72.3 35.6 25.6 22.2
143.6 30.4 18.8 15.7
0.701 0.316 0.285 0.255
a Adsorption conditions: T ) 277 K; C0 (NPYR) ) 1.501 mmol L-1.
TABLE 2: Liquid Adsorption of NNN in CH2Cl2 Solution and Relevant Parameters by Porous Materialsa sample M22a NaZSM-5 NaY AC
S (m2 g-1) V (m3 g-1) Si/Al 481 325 766 1128
0.36 0.17 0.31 0.54
10 26 2.86
adsorbed NNN reduced (mg g-1) (%) 9.85 2.58 4.36 3.55
66 17 30 24
a Adsorption conditions: T ) 277 K; C0 (NNN) ) 1.53 mmol L-1.
solution, defining the selectivity of adsorbent toward NPYR; the larger the Kd value, the more the NPYR captured by the adsorbent. The maximum of Kd emerged on M22a, corresponding to its highest adsorption capacity (Table 1). The kinetic data of these curves were also calculated with first-rate Lagergren equation:41 log(qm - qt) ) log qm - (Kt × t)/2.303, where qm and qt (mg g-1) were the adsorption capacities at equilibrium and time t, respectively, and Kt (min-1) was the rate constant of pseudofirst order adsorption. The same zeolite with different morphologies, say, M22a and M22b, had different Kt values, and the former with rose-like morphology possessed the higher data (Table 1). It appears that adsorption rate constant Kt relates to the structure and morphology of zeolite. Table 2 lists the liquid adsorption of NNN in dichloromethane solution. For the larger molecule NNN, M22a trapped 66% (9.85 mg g-1) in the solution of 1.53 mmol L-1, while NaZSM-5 (2.58 mg g-1) and NaY (4.36 mg g-1) adsorbed much less under the same conditions. Activated carbon is a popular adsorbent; it adsorbed 3.55 mg g-1 of NNN, about two-fifths of M22a. Figure 8 illustrates the instantaneous adsorption of NPYR at 338 K, in which the same amount of the absorbents were used; therefore, when the materials all have different apparent densities, they will have different volumes, hence, different path lengths for the NPYR to flow through. The M22a sample trapped
a
adsorbent
KF
n
R2
M22a M22b M22c M22d M22e NaY NaZSM-5
0.973 0.992 0.904 0.870 0.807 0.946 0.500
1.06 1.02 1.24 1.48 1.75 1.12 3.05
0.996 1.000 0.980 0.945 0.918 0.986 0.949
Conditions: 1 atm; the rate of carrier gas was 20 mL min-1.
all the NPYR in a gas stream at 338 K until the accumulated amount of NPYR reached 1.5 mmol g-1. In comparison, zeolites NaZSM-5 and mordenite lost their ability at the amount of 0.58 and 0.43 mmol g-1, respectively. Zeolite NaY was able to adsorb all the NPYR, but inferior to M22a, once the amount of NPYR exceeded 1.5 mmol g-1. In the case of 2.28 mmol g-1 NPYR, NaZSM-5 adsorbed about 28%, while NaY trapped 80%, but M22a captured 88% (Figure 8A). Activated carbon was also found to adsorb all the NPYR in the gas stream (Figure 8A). As demonstrated in Figure 8B, sample M22b exhibited a lower adsorptive ability than M22a at 338 K, implying the possibility that more tightly packed platelets in M22a promoted the adsorption of NPYR. The M22b sample also showed a reduced ability in the liquid adsorption of NNN (Figure 2). For the desilicated MCM-22 sample, M22c, showed the highest adsorptive ability of NPYR at 338 K (Figure 8B). Its adsorption capacity was greater than that of M22a, which was also shown in Figure 8A. The formation of mesopores is the key factor (Figures 2 and 5) with the advantage to adsorb NPYR at low temperature.42 For the other two samples, M22d and M22e, which have the same Si/Al ratio but different morphology (Figures 5C, 5D and 8B), 1.5 mmol g-1 of NPYR was captured by M22d, while M22e trapped 1.2 mmol g-1, when 2.3 mmol g-1 of NPYR passed through the zeolite at 338 K. On the basis of these results, it is clear that the surface morphology can affect its performance in the adsorption of nitrosamines. On the other hand, the desilicated MCM-22 sample, M22c, showed a surprisingly high activity in the adsorption of NPYR at 338 K (Figure 8A). Its adsorption capacity was greater than that of M22a once the amount of NPYR exceeded 1.75 mmol g-1. To qualitatively assess the adsorption of zeolites, the experimental isotherms at 338 K were fitted with Freundlich equation. As shown in Table 3, all the adsorption isotherms can be described by the Freundlich equation, and the exponent of 1 < n < 10 indicates the favorable nature of the adsorption system.
Figure 8. Gas adsorption of NPYR by (A) zeolites MCM-22 (a and c), NaY, NaZSM-5, Mordenite, and activated carbon (AC), and (B) MCM-22 series samples with a 20 mL min-1 flow rate at 338 K.
Capturing Nitrosamines by Zeolite MCM-22
Figure 9. Gas adsorption of NPYR by zeolites M22a and M22b with different gas flow rate.
The smallest value of n, the strongest affinity between the zeolites and NPYR, was observed for the M22c sample. It appears that both the morphology and the Si/Al ratio govern the adsorption ability as based on the Freundlich constant KF. As the Si/Al ratio was reduced from 16 in M22d and M22e to 10 in M22a and M22b, the latter two samples contained more cations and, hence, displayed a larger adsorption amount than the other two samples. For the samples with the same Si/Al ratio, however, their morphology affected the adsorption significantly. Figure 9 depicts the gaseous adsorption of NPYR by M22a and M22b samples with various gas flow rates in which M22a captures more NPYR than M22b at the same gas flow rate. When the flow rate of carrier gas increased from 10 to 30 mL min-1, both samples trapped more NPYR, which differed from zeolite NaY but similar to NaZSM-5,7 involving the mass transfer of adsorbate. It is necessary for NPYR to penetrate the channel of zeolite at first and then diffuse to adsorptive sites.9 Nonetheless, the molecular size of NPYR is close to the pore diameter of NaZSM-5, hence, it is difficult for NPYR to enter or diffuse in the confined channel. Once the flow rate of carrier gas is increased, the kinetic energy of adsorbate becomes larger because it is pushed by the carrier gas moving quickly. That is to say, the NPYR has more energy to enter the channel of zeolite. As a result, the entering and diffusion of the molecules are relatively easier and more rapid. Of course, the number of the diffuse-in molecules is also increased, resulting in the elevated adsorption capacity of NaZSM-5 zeolite. MCM-22 has the pore size (0.5 nm) similar to ZSM-5, hence, the higher flow rate of carrier gas is also beneficial for its adsorption toward NPYR. However, a more obvious flow-effect was observed on the M22b sample. When the flow rate changed from 10 to 30 mL min-1, M22b adsorbed more NPYR of 0.57 mmol g-1, while the smaller value (0.29 mmol g-1) emerged on M22a under the same conditions. This phenomenon originates from the different morphologies of two samples. The close-packed slice morphology of M22a with various surface curvatures enables more favorable orientations of NPYR to be occurred during the collision onto the adsorbent, hence, it exhibits a relatively higher capability in the adsorption with lower flow rate of carrier gas. Contrarily, M22b lost this special morphology so that it showed the obvious flow effect in its adsorption, similar to common ZSM-5 zeolite. Figure 10 illustrates the adsorption of NPYR by the different adsorbents at 453 K. When the amount of NPYR passed through the adsorbent, reaching 1.0 mmol g-1, four samples (M22a, M22c, NaY, and activated carbon) captured almost all of the
J. Phys. Chem. C, Vol. 114, No. 21, 2010 9593
Figure 10. Gas adsorption of NPYR by porous materials at 453 K with a 20 mL min-1 flow rate.
NPYR (0.99 mmol g-1 or more), while NaZSM-5 trapped 0.52 mmol g-1 at 338 K. As the adsorption temperature rose to 453 K, the corresponding value declined to 0.92, 0.72, 0.86, and 0.52 mmol g-1 whilst NaZSM-5 zeolite still adsorbed 0.52 mmol g-1 at the same condition (Figure 10). Compared with M22a, M22c became less active in the adsorption of NPYR at 453 K. 4. Discussions Zeolite MCM-22 exhibited a higher adsorption capability than NaY and NaZSM-5 in the liquid adsorption of nitrosamines; this may be explained by the following three main factors. The first one is the specific porous structure of MCM-22. In general, the degree of geometrical match between the channel opening of a zeolite and the molecular diameter of the nitrosamines determines the adsorption manner by the zeolite.7,11,20 The relatively larger pore size of the zeolite enables the nitrosamine molecule to enter inside the channels, for example, zeolite NaY can adsorb more volatile nitrosamines such as NPYR than NaZSM-5 and NaA under the same conditions.7 On the other hand, the adsorption of nitrosamine will be strengthened if the channels of the zeolite can provide a geometric confinement for the target molecule, therefore, retarding the desorption of nitrosamines. For example, the desorption in zeolite NaZSM-5 and NaA is suppressed in comparison with NaY.6 Consequently, an ideal porous adsorbent to trap nitrosamines should have a bell-mouthed pore opening, connecting the slightly narrow channel to restrict the target, as reported by Reitmeier et al.24 The structure of zeolite MCM-22 partially meets this requirement, because it has side cups of half pockets on the exterior surface and the narrow channel similar to that of ZSM-5. The opened supercages of MCM-22 can trap the larger molecules, which are superior to ZSM-5 in trapping the bulky nitrosamines, while the narrow channels connected with the supercage provide a tighter geometric confinement than that of NaY. As a result, MCM-22 exhibited a higher adsorption capability of nitrosamines than other zeolites (Figures 7 and 8). The second factor is the hydrophilicity/hydrophobicity of zeolites. MCM-22 has an Al-content higher than ZSM-5, but lower than Y zeolite; hence, it has a suitable hydrophobicity, which does not seriously hinder the adsorption of nitrosamines in dichloromethane solution. The third factor is its rose-like morphology, which is behind the higher adsorption ability of M22a than that of M22b (Figure 7). M22a possessed close-packed slice morphology with abundant surface curvature, and the roughened surface is thought to allow more favorable orientations of adsorbate molecules during their collision with the surface, thus, promoting the adsorption of nitrosamines. This close-packed slice morphology
9594
J. Phys. Chem. C, Vol. 114, No. 21, 2010
SCHEME 1: Collision Occuring on a Solid Surface with Different Morphologies
was partially destroyed in preparing the M22b sample, where the platelets were only loosely piled up (Figure 5B). Consequently, this morphology resulted in the declined adsorption ability for M22b in both the solution and the gas stream (Figures 7 and 8). The effect of morphology emerged much more obviously in the instantaneous gas adsorption of NPYR, because the contact time was less than 0.1 s;7 hence, any step that could enhance the collision probability would in theory improve the adsorption. However, there are some fine differences in the adsorption ability among the MCM-22 samples, for example, M22a was better than M22b, and M22d surpassed M22e (Figure 8). The collision frequency with a rough surface cannot be exactly described by the equation. Instead, the key factor is the relationship between the frequency f and the fractal dimensions D. And the shape of the relation f versus D depends on the ratio between the diameter of the gas phase atom and the surface atom of the solid.43,44 Increasing the fractal dimension of the surface leads to an increasing number of strongly folded places on the surface and the increasing possibility of a collision. Unlike the collision that happened on a flat surface where the molecule usually rebounds, the molecule that collided onto the rough surface may be inserted into the gaps or defects of the surface (Scheme 1), which may be beneficial for adsorption. For M22a and M22b, it is clear that fa > fb, thus, ra > rb (assuming Ea ) Eb). The same phenomenon has been observed on M22d and M22e samples. For instantaneous adsorption, variation in the surface roughness or morphology of zeolite may cause considerable differences in the actual adsorption, though the fractal dimension of the surface is only one of the parameters.43 The close-packed slice morphology with various surface curvatures enables zeolite MCM-22 to show a high activity in the adsorption of nitrosamines because the roughened surface allows more favorable orientations of the molecules during the collision onto the adsorbent.12 Likewise, the loss of the close-packed morphology results in the declined adsorption ability in the M22b sample (Figures 1B and 2). Furthermore, the degree of slice stacking may also affect the adsorption (Figure 8B), as in the case M22d and M22e. The surface curvature of the zeolite may also change the pattern and reduced the flow rate airflow, further enhancing the contact time.12 Comparing Figure 8 and Figure 9 demonstrates that desorption of adsorbate is easier at a higher temperature so that both the geometric confinement and the strong electrostatic interaction are necessary to minimize desorption. Activated carbon lost about 40% of adsorptive ability at 453 K due to the lack of strong electrostatic interaction toward nitrosamines; zeolite NaY lost about 13% of the adsorption capability owing to its wide channel that failed to tightly hold the NPYR molecules at the relatively high temperature, while the narrow channel enabled zeolite NaZSM-5 to keep the adsorption capacity.6,7 M22a was the other sample maintaining a high adsorption at 453 K (Figure 9); however, this advantage could not simply be attributed to the lower Si/Al ratio of the M22a sample than that of NaZSM-5
Yang et al. (Table 2), because M22c has an even lower Si/Al ratio. Two factors should be taken into account to explain the deactivation of M22c. One was the formation of mesopores resulted from desilication (Figures 2 and 5), which adsorb NPYR at low temperature but lost this function at the higher temperature.42 Another was the different state of Al in M22c sample where six-coordinated aluminum species disappeared (Figure 3). Because the extra-framework Al species were mainly located on the external surface of the zeolite, they easily interacted with NPYR molecules, promoting the adsorption through the electrostatic interaction. Other evidence on the location-function relation of cations toward NPYR come from the modified amorphous silica and the mesoporous silica MCM-41,46,47 both of them were coated with alumina through postmodification, hence, almost all of the aluminum species were 6-coordinated; they all exhibited an excellent capability similar to zeolite NaY in the adsorption of NPYR at 453 K.47 The electrostatic interaction, however, is limited within 0.05 nm45 so that NaY is unable to pull the NPYR molecule at 453 K, though the zeolite has a relatively low Si/Al ratio (Table 2). Contrarily, NaZSM-5 has fewer cations than NaY zeolite but keeps its adsorptive ability at 453 K, which may be explained by the narrower pore size (0.5 nm) than that of NaY (0.7 nm). Based on these results, it is safe to infer that the absence of six-coordinated aluminum species in the sample of M22c should weaken the electrostatic interaction toward nitrosamines, resulting in the deactivation of M22c at 453 K. In general, the two steps, intercepting NPYR in a gas stream and holding them in the pore, are crucial for the zeolite to adsorb volatile nitrosamines in air flow. The former is important in the adsorption at the relatively low temperature such as 338 K because the influence of desorption is minor at this temperature; also, the generation of mesopores and the establishment of a suitable electrostatic field can further increase the efficiency of adsorption. Nonetheless, minimizing desorption becomes the determining step when the adsorption is carried out at a higher temperature, such as 453 K; hence, the confinement provided by the narrow channel along with the strong interaction originating from the cations in zeolite are decisive to catch adsorbed nitrosamines. Accordingly, MCM22 is the most competitive candidate to capture the nitrosamines in air flow in the laboratorial adsorption experiments. The generation of mesopores in MCM-22 zeolite through desilication is only beneficial for the adsorption of volatile nitrosamines at relatively low temperature, indicating the limitation of this postmodification method. 5. Conclusions Some conclusive remarks can be derived from these experiment results mentioned above. (1) MCM-22 is superior to zeolites NaY and NaZSM-5 in liquid adsorption of volatile nitrosamines such as NPYR and tobacco specific nitrosamines like NNN due to its advantageous pore structure, surface hydrophobicity, and close-packed morphology. (2) MCM-22 is able to capture more NPYR in a gas stream than NaY and NaZSM-5 at 338 K, and its adsorptive capability can be further elevated by the generation of artificial mesopores in the zeolite through desilication. (3) MCM-22 exhibits the highest ability in the instantaneous adsorption of NPYR at 453 K due to its inherent advantageous porous structure containing a bellmouthed pore mouth connected with the slightly narrow channel, which are beneficial to intercept nitrosamines in air flow. (4) The morphology of the zeolite considerably affects the adsorption of volatile nitrosamines, especially at the relative low temperature such as 338 K.
Capturing Nitrosamines by Zeolite MCM-22 Acknowledgment. The National Natural Science Foundation of China (20773061 and 20873059), Grant 2008AA06Z327 from the 863 Program of the Ministry of Science and Technology of China, Grant CX08B_009 from Jiangsu Province Innovation for Ph.D. candidate, the Scientific Research Foundation of Graduate School of Nanjing University, and Analysis Center of Nanjing University financially supported this research. The authors are grateful to Professor T. Dou and his group (China University of Petroleum) for providing the MCM-22 sample, and Dr. L. M. Peng (Nanjing University) for his helpful discussion on 27Al MAS NMR. Supporting Information Available: Elaboration of structure, morphology, and other adsorption properties of the porous materials in this study. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Wheatley, P. S.; Butler, A. R.; Crane, M. S.; Fox, S.; Xiao, B.; Rossi, A. G.; Megson, I. L.; Morris, I. E. J. Am. Chem. Soc. 2006, 128, 502–509. (2) Terzano, R.; Spagnuolo, M.; Medici, L.; Vekemans, B.; Vincze, L.; Janssens, K.; Ruggiero, P. EnViron. Sci. Technol. 2005, 39, 6280–6287. (3) Meier, W. M.; Siegmann, K. Microporous Mesoporous Mater. 1999, 33, 307–310. (4) International Agency for Research on Cancer Some N-Nitroso Compounds. IARC Monographs on the EValuation of Carcinogenic Risk of Chemicals to Humans; France Press: Lyon, 1978; Vol. 17, p 365. (5) Prokoczyk, B.; Hoffmann, D.; Bologna, M.; Cunningham, A. J.; Trushin, N.; Akerkar, S.; Boyiri, T.; Amin, S.; Desai, D.; Colosimo, S.; Pottman, B.; Leder, G.; Ramadani, M.; Henne-Bruns, D.; Beger, H. G.; EI-Bayoumy, K. Chem. Res. Toxicol. 2002, 15, 677–685. (6) Xu, Y.; Zhu, J. H.; Ma, L. L.; Ji, A.; Wei, Y. L.; Shang, X. Y. Microporous Mesoporous Mater. 2003, 60, 125–138. (7) Zhou, C. F.; Cao, Y.; Zhuang, T. T.; Huang, W.; Zhu, J. H. J. Phys. Chem. C 2007, 111, 4347–4357. (8) Yang, J.; Zhou, Yu.; Wang, H. J.; Zhuang, T. T.; Cao, Y.; Yun, Z. Y.; Yu, Q.; Zhu, J. H. J. Phys. Chem. C 2008, 112, 6740–6748. (9) Zhu, J. H.; Yan, D.; Xia, J. R.; Ma, L. L.; Shen, B. Chemosphere 2001, 44, 949–956. (10) Yang, J.; Dong, X.; Zhou, Y.; Yue, M. B.; Zhou, C. F.; Wei, F.; Zhu, J. H.; Liu, C. Microporous Mesoporous Mater. 2009, 120, 381–388. (11) Wei, F.; Yang, J. Y.; Gao, L.; Gu, F. N.; Zhu, J. H. J. Hazard. Mater. 2009, 172, 1482–1490. (12) Gao, L.; Cao, Y.; Zhou, S. L.; Zhuang, T. T.; Wang, Y.; Zhu, J. H. J. Hazard. Mater. 2009, 169, 1034–1039. (13) Zhou, C. F.; Zhu, J. H. Chemosphere 2005, 58, 109–114. (14) Pinisakul, A.; Kritayakornupong, C.; Ruangpornvisuti, V. J. Mol. Model. 2008, 14, 1035–1041. (15) Zhu, J. H.; Xia, J. R.; Wang, Y.; Xie, G.; Xue, J.; Chun, Y. Stud. Surf. Sci. Catal. 2001, 135, 320–327. (16) Xu, Y.; Liu, H. D.; Zhu, J. H.; Yun, Z. Y.; Xu, J. H.; Wei, Y. L. New J. Chem. 2004, 28, 244–252. (17) Xu, Y.; Yun, Z. Y.; Zhu, J. H.; Xu, J. H.; Liu, H. D.; Wei, Y. L.; Hui, K. J. Chem. Commun. 2003, 1894. (18) Wu, Z. Y.; Wang, H. J.; Ma, L. L.; Xue, J.; Zhu, J. H. Microporous Mesoporous Mater. 2008, 109, 436–444.
J. Phys. Chem. C, Vol. 114, No. 21, 2010 9595 (19) Cao, Y.; Zhuang, T. T.; Yang, J.; Liu, H. D.; Huang, W.; Zhu, J. H. J. Phys. Chem. C 2007, 111, 538–548. (20) Cao, Y.; Yun, Z. Y.; Yang, J.; Dong, X.; Zhou, C. F.; Zhuang, T. T.; Yu, Q.; Liu, H. D.; Zhu, J. H. Microporous Mesoporous Mater. 2007, 103, 352–362. (21) Zhou, Y.; Gao, L.; Gu, F. N.; Yang, J. Y.; Yang, J.; Wei, F.; Wang, Y.; Zhu, J. H. Chem.sEur. J. 2009, 15, 6748–6757. (22) Zhao, D.; Feng, J.; Huo, Q.; Melosh, N.; Fredrickson, G. H.; Chmelka, B. F.; Stucky, G. D. Science 1998, 279, 548–552. (23) Lee, J. S.; Joo, S. H.; Ryoo, R. J. Am. Chem. Soc. 2002, 124, 1156– 1157. (24) Reitmeier, S. J.; Gobin, O. C.; Jentys, A.; Lercher, J. A. Angew. Chem., Int. Ed. 2008, 48, 533–538. (25) Wu, Y. J.; Ren, X. Q.; Wang, J. Mater. Chem. Phys. 2009, 113, 773–779. (26) Zukal, A.; Sjiklova´, H.; Cjejka, J. Langmuir 2008, 24, 9837–9842. (27) Ravishankar, R.; Bhattacharya, D.; Jacob, N. E.; Sivasanker, S. Microporous Mater. 1995, 4, 83–93. (28) Daems, I.; Mthivier, A.; Leflaive, P.; Fuchs, A. H.; Baron, G. V.; Denayer, J. F. M. J. Am. Chem. Soc. 2005, 127, 11600–11601. (29) Dong, X.; Zhou, C. F.; Yue, M. B.; Zhang, C. Z.; Huang, W.; Zhu, J. H. Mater. Lett. 2007, 61, 3154–3158. (30) Groen, J. C.; Jansen, J. C.; Moulijn, J. A.; Pe´rez-Ramı´rez, J. J. Phys. Chem. B 2004, 108, 13062–13065. (31) Wu, Y. J.; Ren, X. Q.; Lu, Y. D.; Wang, J. Microporous Mesoporous Mater. 2008, 112, 138–146. (32) Corma, A.; Diaz, U.; Forne´s, V.; Guil, J. M.; Martı´nez-Triguero, J.; Ceyghtony, E. J. J. Catal. 2000, 191, 218–224. (33) Sing, K. S. W.; Everett, D. H.; Haul, R. A. W.; Moscou, L.; Pierotti, R. A.; Rouque´rol, J.; Siemieniewska, T. Pure Appl. Chem. 1985, 57, 603– 619. (34) Groen, J. C.; Peffer, L. A. A.; Moulijn, J. A.; Pe´rez-Ramı´rez, J. Chem.sEur. J. 2005, 11, 4983–4994. (35) Ma, D.; Deng, F.; Fu, R.; Han, X.; Bao, X. J. Phys. Chem. B 2001, 105, 1770–1779. (36) Kennedya, G. J.; Lawtona, S. L.; Funga, A. S.; Rubina, M. K.; Steuernagel, S. Catal. Today 1999, 49, 385–399. (37) Kolodziejski, W.; Zicovich-Wilson, C.; Corell, C.; Perez-Pariente, J.; Corma, A. J. Phys. Chem. 1995, 99, 7002–7008. (38) Corma, A.; Corell, C.; Forne´s, V.; Kolodziejski, W.; Pe´rez-Pariente, J. Zeolites 1995, 15, 576–582. (39) Ma, D.; Han, X.; Zhou, D.; Yan, Z.; Fu, R.; Xu, Y.; Bao, X.; Hu, H.; Au-Yeung, Steve, C. F. Chem.sEur. J. 2002, 8, 4557–4561. (40) Metaxas, M.; Kasseloui-Rigopoulou, V.; Galitsatou, P.; Konstantopoulou, C.; Oikonomou, D. J. Hazard. Mater. 2003, 97, 71–82. (41) Lagergren, S. Kungliga SVenska Vetenskapsakademiens; Handlingar; P.A. Norstedt: Stockholm, 1898; Vol. 24, pp 1-39. (42) Cao, Y.; Shi, L. Y.; Zhou, C. F.; Yun, Z. Y.; Wang, Y.; Zhu, J. H. EnViron. Sci. Technol. 2005, 39, 7254–7259. (43) Panczyk, T.; Warzocha, T.; Rudzinski, W. Appl. Surf. Sci. 2007, 253, 5846–5850. (44) Rudzinski, W.; Lee, S. L.; Yan, C. C.; Panczyk, T. J. Phys. Chem. B 2001, 105, 10847–10856. (45) Yang, J.; Ma, L. L.; Shen, B.; Zhu, J. H. Mater. Manuf. Processes 2007, 22, 750–757. (46) Xu, J. H.; Zhuang, T. T.; Cao, Y.; Yang, J.; Wen, J. J.; Wu, Z. Y.; Zhou, C. F.; Huang, L.; Wang, Y.; Yue, M. B.; Zhu, J. H. Chem. Asian J. 2007, 2, 996–1006. (47) Gu, F. N.; Zhou, Y.; Wei, F.; Wang, Y.; Zhu, J. H. Microporous Mesoporous Mater. 2009, 126, 143–151.
JP9116497
