FeUSY Zeolite Catalyst for Effective Decomposition of Nitrous Oxide

AND G. Q. MAX LU ‡. Research Center for Eco-Environmental Sciences, Chinese. Academy of Sciences, Beijing 100085, PR China, and. Australian Research...
0 downloads 0 Views 389KB Size
Download PDF
Environ. Sci. Technol. 2007, 41, 7901-7906

Fe-USY Zeolite Catalyst for Effective Decomposition of Nitrous Oxide LAN DONG LI,† QUN SHEN,† J U N J I E Y U , † Z H E N G P I N G H A O , * ,† Z H I P I N G X U , * ,‡ A N D G . Q . M A X L U ‡ Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, PR China, and Australian Research Council (ARC) Centre for Functional Nanomaterials, School of Engineering, The University of Queensland, Brisbane, QLD 4072, Australia

Zeolite matrix FAU is applied as an effective support that can be readily exchanged with ferric ions simply by wet ion exchange. The high exchange degree in this zeolite (USY) probably benefits from its larger channel dimension, which makes the diffusion of hydrated ferric ions into the channels easier. The as-prepared and subsequently calcined Fe-USY samples contain various kinds of iron species, which enable Fe-USY to efficiently decompose N2O to N2 and O2. The presence of O2 (20%), H2O (2%), or both reduces the N2O conversion only to a small extent at the same temperature. To test the feasibility and the catalytic activity in a practical situation, we prepared a monolithic Fe-USY/cordierite sample and investigated the N2O conversion under similar conditions. Such a cordieritesupported Fe-USY catalyst (∼9 wt % USY and 0.5 wt % Fe) shows the catalytic performance in N2O decomposition similar to the pure Fe-USY catalyst. Remarkably, both the pure Fe-USY and Fe-USY/cordierite catalysts demonstrate a very good durability because there is no activity lost after 100 and 144 h tests. Thus, the Fe-USY zeolite shows its potential as a cost-effective catalyst for N2O elimination in future applications.

Introduction Nitrous oxide (N2O) had been considered as a relatively harmless gas and did not receive much attention from the environmental point of view. However, N2O has been found to be harmful to our environment since the past decade because it greatly contributes to the greenhouse effect, as well as severe destruction of the ozone in the stratosphere (1). N2O has a very long atmospheric lifetime (∼120 years), and its global warming potential (GWP) per N2O molecule for the 100 year time is about 310 times that of carbon dioxide (CO2) (2). As reported by IPCC, anthropogenic N2O emissions contribute about 6% to the global greenhouse effect and 6% to the ozone layer depletion (3). The major sources of anthropogenic N2O emissions come from the manufactures of nitric acid and adipic acid in industry. Many efforts have been devoted to develop end-of-pipe technologies to mitigate * Address correspondence to either author. Phone: +86-1062849194 (Z.P.H.); 61-7-33463809 (Z.P.X). Fax: +86-10-62923564 (Z.P.H.); 61-7-33463973 (Z.P.X.). E-mail: [email protected] (Z.P.H.); [email protected] (Z.P.X.). † Chinese Academy of Sciences. ‡ The University of Queensland. 10.1021/es071779g CCC: $37.00 Published on Web 10/11/2007

 2007 American Chemical Society

N2O emission from these two sources. For example, heterogeneously catalytic decomposition of N2O to N2 and O2 is proposed as the most economical method to eliminate N2O from the manufacture exhausts. N2O catalytic decomposition requires catalysts with sufficiently high activity and long durability. In the past years, several types of catalysts, for example, zeolites, mixed oxides, and supported noble metals were investigated for N2O catalytic decomposition (4). Currently, iron zeolites, especially Fe-ZSM-5, have been shown to be good catalysts for N2O thermal decomposition because of their good catalytic performance even in the presence of O2, H2O, SO2, etc. (5). Fe-zeolite catalysts can be prepared with several different ways, including solid-state ion exchange (6), chemical vapor decomposition (CVD) of FeCl3 (7), and isomorphous substitution, followed by extraction (8). Compared with these methods, wet ion exchange is simpler and more feasible for industrial applications, because it involves fewer steps, and the process parameters are easily controlled. One way to prepare iron zeolites is ferrous ion exchange (9). However, the oxidation of Fe2+ to Fe3+ during the ion-exchange process renders the process non-repeatable (10). Ferric ion exchange seems to be ideal method for the preparation of iron zeolites; however, ion diffusion is very limited, and the ion-exchange degree is very low because of the long diffusion path and the relatively larger size of hydrated ferric ions (0.68 nm) in comparison with the channel size of the normal zeolites. To achieve high ion exchange, one can shorten the diffusion path, choose a zeolite with larger channel size, or use more severe exchange conditions. The first method was demonstrated by Melie´n-Cabrera et al. to decrease the diffusion path upon a mild alkaline treatment (11). It has been reported that heating at 353 K can achieve the full iron loading (12, 13). The current study indicates ferric ions can readily diffuse into the large channels of zeolite by exchange. A special zeolite matrix (FAU) with a larger channel dimension ([111] 0.74 × 0.74 nm) was selected as the host of the ferric ions. We have found that the wet ion exchange of FAU with FeCl3 solution at pH 2 can readily lead to a highly exchanged Fe-USY with reliable reproducibility. The calcined Fe-USY shows high activity and long durability for N2O decomposition in the presence of O2 and H2O. Moreover, we have particularly prepared monolithic Fe-USY/cordierite catalyst and found the catalytic activity to be similar to that of the pure Fe-USY catalyst, showing a great potential as a cost-effective catalyst for N2O elimination.

Experimental Section Materials Preparation. Commercial H-ZSM-5, H-Beta and H-USY (Sinopec Co.) with similar Si/Al ratios (∼12) were used as the parent zeolites in this study. Iron species was introduced to zeolites by wet ion exchange using ferric salt as iron source. Typically, 5 g of parent zeolite was added to 500 mL of 0.05 M FeCl3 aqueous solution at room temperature and then stirred for 48 h under air. For comparison in N2O thermal decomposition, Fe-FER was prepared from NH4FER by wet ion exchange using 0.05 M FeCl3 aqueous solution as iron source (12, 13). The ion exchange process was carried out at 353 K for 24 h. After ion exchange, the sample was filtered, thoroughly washed, dried at 353 K overnight, and then calcined in air at 873 K for 4 h. Cordierites (Corning Inc., 60 cells/cm2, 0.3 mm average wall thickness) were used as support materials after ultrasonic cleaning with diluted HCl solution and then distilled water. For Na-Y to grow on cordierite substrate, tetraethylorthosilicate (TEOS), aluminum sulfate, sodium hydroxide, and VOL. 41, NO. 22, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7901

TABLE 1. Ion Exchange Results of Iron Zeolites sample

iron source

Si/Ala

Fe/Alb

Fe (wt %)b

SBET (m2/g)a

Fe-USY Fe-FER Fe-Beta Fe-ZSM-5

FeCl3 FeCl3 FeCl3 FeCl3

11.6 11.2 10.8 11.8

0.42 0.28 0.06 0.04

3.38 2.31 0.52 0.30

512 358 516 423

a For parent zeolites. b For obtained iron-zeolite and for full exchange Fe/Al ) 0.33.

water were mixed at the proportion of 10 SiO2/1 Al2O3/5 Na2O/ 200 H2O. After it was vigorously stirred, the mixture was placed into a Teflon-lined stainless steel autoclave for heat treatment at 373 K for 8 h. The as-synthesized Na-Y/cordierite sample was obtained by ultrasonically washing away the unreacted adherents with distilled water. The Na-Y/cordierite sample was then exchanged to NH4Y/cordierite with 0.1 M NH4Cl solution. The dealumination of NH4-Y/cordierite was performed with heat treatment at 1023 K for 12 h, and the dealuminated sample was further treated in diluted HNO3 solution to remove the nonframework Al to obtain H-USY/cordierite. Fe-USY/cordierite was finally prepared from H-USY/cordierite with the above wet ion exchange method. Materials Characterization. The Si, Al, and Fe contents of these iron zeolites were determined by ICP-AES using an Optima 2000 spectrometer. Low-temperature N2 adsorption/ desorption experiments were carried out in a Quantachrome NOVA-1200 gas absorption analyzer, and the specific surface area was calculated with the BET equation. UV-vis diffuse reflectance spectra (UV-vis/DRS) were recorded in the air against BaSO4 in the region of 190-800 nm on a Hitachi UV-3000 spectrometer. The X-ray diffraction (XRD) patterns of all samples were measured on a Rigaku powder diffractometer (D/MAX-RB) using Cu KR radiation (λ ) 0.15418 nm) at a scanning rate of 4°/min in 2θ ) 5-35°. Transmission electron microscopy images of samples were acquired on a JEOL 3010 transmission electron microscope at an accelerate voltage of 300 kV. Catalytic Activity Test. N2O decomposition experiments were performed in a fixed-bed flow microreactor at atmospheric pressure. Typically, 0.1 g of sample (sieve fraction, 0.25-0.5 mm) was placed in a quartz reactor (4 mm i.d.) and pretreated in He at 873 K for 1 h. After it was cooled to 573 K, the reactant gas mixture (N2O, O2; H2O, He; or both balance) was fed to the reactor. The total flowrate of the gas mixture was 60 mL/min , that is, GHSV ) 30 000 h-1. The steady-state tests were conducted isothermally every 25 K from 573 to 873 K, and the gas products (after 1 h of reaction time) were analyzed on-line using a gas chromatograph (Agilent 6820 series) equipped with a TCD detector and two serial columns (a Porapak Q column served for the separation of N2O and N2/O2 and a molecular sieve 5A column for the separation of N2 and O2). N2O decomposition on Fe-USY/cordierite monolith was performed under similar conditions. In particular, 2.0 g of monolithic catalyst was used, and the reactant gas mixture (10% N2O, 20% O2, 2% H2O, He balance) was fed at 500 mL/ min (GHSV ) 15000 h-1, referred to the total volume of the monolith catalyst). The steady-state tests were conducted isothermally at 673, 723, and 773 K for 48 h, respectively, and the gas products were analyzed online after 1 h at each temperature.

Results and Discussion Exchange Degree and Ferric Ion Size. The atomic ratios, the iron loading in zeolites, and the surface area of parent zeolites are listed in Table 1. The Si/Al ratios for the three 7902

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 22, 2007

zeolites (10.8-11.8) are very similar, and their surface areas (423-516 m2/g) are close to one another. However, when ferric salt was used as the iron source, the iron loading in USY (3.38 wt %) is 6-10 times higher than that in ZSM-5 and Beta (0.30 and 0.52 wt %). The Fe/Al ratio (Table 1) indicates a high Fe exchange degree in Fe-USY (0.42) but is very much under-exchanged in Fe-ZSM-5 (0.04) and Fe-Beta (0.06). Because the ion exchange is carried out under the same conditions, the exchange degree is thus suggested to relate to the peculiar structure of the parent zeolite and the properties of ferric ions. First, we can determine the ferric ion species available for the ion exchange in FeCl3 solution. The two main hydrolysis reactions at room temperature are

Fe(H2O)63+ + H2O S Fe(H2O)5(OH)2+ + H3O+ (K )

10-2.13)

2Fe(H2O)63+ + H2O S Fe(H2O)5OFe(H2O)54+ + 2H3O+

(K ) 10-2.91)

If the reactions reach the equilibrium in 0.05 M FeCl3 aqueous solution, the pH is estimated to be 1.6, thus the main ferric ion species are hydrated Fe(H2O)63+ (∼63%), primarily hydrolyzed Fe(H2O)5(OH)2+ (∼24%), and condensed dimer Fe(H2O)5OFe(H2O)54+ (∼13% Fe). Any other kind of ferric ion species, if any, can be neglected. It is our belief that the size of these ferric ions and the zeolite channel dimension largely determine the ion exchange degree in the current 3 cases. As shown in Figure 1S of the Supporting Information, the effective van der Waals diameter of the monomer ions (Fe(H2O)63+ and Fe(H2O)5(OH)2+) in aqueous solution is ∼0.68 nm, and the dimer ion (Fe(H2O)5OFe(H2O)54+) is ∼0.9 nm (14). We know that H-USY has a channel size of [111] 0.74 × 0.74 nm, so that the monomer ions can readily diffuse into USY channels and achieve the maximum exchange. Since the dimer ion can enter the channel only in one direction, we thus guess that only a small proportion of the dimer ions diffuse into the USY channels. Even though, the high-degree of exchange can be achieved via the diffusion of monomer ferric ions. After the monomer ions diffuse into the USY channels and replace the protons, some may react with the surface Al-OH to give Al-O-Fe that can further induce the formation of oligonuclear Fe3+xOy clusters, which seemingly results in an over-exchange (Table 1). In contrast, PerezRamı´rez et al. found that the loading of ferric ions over H-USY was not so high (15). Their preparation conditions ([Fe(NO3)3] ) 0.30 mM, and the estimated pH was >3.5 during the whole exchange process) may cause deep hydrolysis, more severe condensation, and easier precipitation of ferric ions into clusters that may block the channels and prohibit the further exchange, leading to a lower exchange degree. On the other hand, H-ZSM-5 and H-Beta have channel dimensions of [100] 0.51 × 0.55 nm T [010] 0.53 × 0.56 nm and [010] 0.66 × 0.67 nm T [001] 0.56 × 0.56 nm, respectively. The sizes of these channels are similar to or smaller than that of both monomer and dimer ions; thus the diffusion of these ferric ions to the channels is not easy, and the exchange may mainly occur in the surface of zeolites, resulting in a low exchange degree. We note that the full exchange can be also achieved on ferrierite (12, 13, 16) through exchange at 353 K, although the zeolite has smaller channel dimensions ([100] 0.42 × 0.54 nm T [010] 0.35 × 0.48 nm). This has been also demonstrated by Fe-FER prepared with the method reported (12, 13), with Fe/Al ) 0.28. In our opinion, heating dissociates the hydrated ferric ions to lose some coordinated water, so that the hydrated ion size can be reduced and the steric

FIGURE 1. UV-vis/DRS spectra of Fe-ZSM-5, Fe-Beta, and Fe-USY samples (A). Deconvolution of UV-vis/DRS spectra of ion-exchanged (B) and calcined (C) Fe-USY. hindrance of the ion diffusion into ferrierite channels becomes smaller. Meanwhile, heating makes both the ion diffusion and exchange quicker; thus the ferric ions exchanged on the channel mouth may swiftly diffuse and exchange with the adjacent protons and “jump” into the channels. In addition, heating may facilitate the formation of some species, such as Al-O-Fe, to enhance the exchange degree. In this connection, the exchange process at 353 K and room temperature may be very much different. Physical Properties of Catalysts. The nature and distribution of iron species in zeolites are assessed by UV-vis/ DRS spectra. Because the iron loading in ZSM-5 and Beta is much smaller than that in USY (Table 1), the absorbance of these two samples for the UV-vis light is much weaker, as can be seen in Figure 1A. After suitable deconvolution of the UV-vis/DRS spectra of as-ion-exchanged and calcined Fe-USY (Figure 1B and 1C), several characteristic UV-vis absorption bands can be identified and attributed to the specific ferric ion species. For example, the two bands at 215 and 285 nm can be assigned to isolated Fe3+ ions in tetrahedral and octahedral coordination, respectively (6, 8, 17). The two bands between 300 and 400 nm (centered at 335 and 385 nm) are caused by small oligonuclear Fe3+xOy clusters in the zeolite channel and surface (8), and the bands at above 400 nm (centered at 460 and 540 nm) are assigned to Fe2O3 nanoparticles at the external surface of zeolite crystal (8, 17). It is assumed that the isolate Fe3+ ions are the monomer ions (Fe(H2O)63+/Fe(H2O)5(OH)2+) that diffuse into the channels and the oligonuclear Fe3+xOy clusters in zeolite channels may come from the condensation on the surface Al-O-Fe species. In comparison with Fe-Beta and Fe-ZSM-5, Fe-USY exhibits

much stronger absorbance from oligonuclear Fe3+xOy clusters, which supports our above postulation that oligonuclear Fe3+xOy clusters are formed via condensation mainly in the larger channels of USY. As shown by the TEM images in Figure 2, the Fe2O3 nanoparticles on the external surface of uncalcined Fe-USY are small (∼1 nm). Such Fe2O3 nanoparticles are possibly grown from oligonuclear Fe3+xOy clusters on the zeolite external surface during drying process (353 K, overnight). After calcination, the Fe2O3 nanoparticles in calcined FeUSY are larger (1-4 nm), suggesting the dislodgement, migration, and aggregation of isolated Fe3+ species and oligonuclear Fe3+xOy clusters to form Fe2O3 nanoparticles upon calcination. The XRD patterns (see Supporting Information Figure 2S) indicate that the structure of parent zeolite is well preserved after wet ion exchange and calcination. Moreover, no diffraction peaks corresponding to iron oxides are observed, revealing that Fe2O3 nanoparticles are in a very small proportion in Fe-USY sample. Catalytic Activity of Iron Zeolites. As discussed above, we can readily prepare highly loaded Fe-USY catalyst with multiple Fe species by wet ion exchange. For the thermal decomposition of N2O, both isolated Fe3+ species and oligonuclear Fe3+xOy clusters are active, while the latter is more active (18-21). Fortunately, our Fe-USY has much more oligonuclear clusters, which may be the reason for its highest catalytic activity (Figure 3A). N2O decomposition starts at ∼600 K, and the 50 and 100% N2O conversion occurs at 680 and 750 K, respectively. Compared with the Fe-USY (1.28 wt % Fe) prepared by Perez-Ramı´rez et al. (15), our Fe-USY (3.38 wt %) exhibit a higher activity, largely caused VOL. 41, NO. 22, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7903

FIGURE 2. TEM images of ion-exchanged and calcined Fe-USY samples.

FIGURE 3. (A) Steady-state catalytic activities for N2O decomposition on Fe-USY and Fe-FER. Reaction conditions: 0.1 g of catalyst, 10% N2O, balance He, GHSV ) 30000 h-1. (B) Dependence of N2O conversion on initial N2O concentration at 698 and 723 K. Reaction conditions: 0.1 g of catalyst, 0-12% N2O, balance He, GHSV ) 30 000 h-1. (C) Effect of O2 and H2O in the stream on N2O conversion over Fe-USY catalyst. Reaction conditions: 0.1 g of catalyst, 10% N2O, 0 or 20% O2, 0 or 2% H2O, balance He, GHSV ) 30 000 h-1. (D) Comparison between Fe-USY and Fe-USY/cordierite. Conditions: 10% N2O, balance He, Fe-USY ) 0.1 g, GHSV ) 30 000 h-1; Fe-USY/cordierite ) 2 g, GHSV ) 15 000 h-1. by the higher iron loading. It is worthy mentioning that the activity of our Fe-USY for N2O thermal decomposition is comparable with most of the best Fe-zeolite catalysts (such as Fe-FER) reported elsewhere (12, 13, 16, 21-24) to our knowledge. For direct comparison, we prepared Fe-FER following the method reported (12, 13) and tested the activity under the same conditions. As shown in Figure 3A, Fe-USY (3.38 wt %) is more active than Fe-FER (2.31 wt %) for N2O catalytic decomposition. Effects of Experimental Conditions. Because in a real manufacturing situation, there is some O2 and water vapor 7904

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 22, 2007

in the stream with a varied concentration of N2O, it is necessary to clarify their effects on N2O decomposition. First of all, we investigated the effect of initial N2O concentration in the mixed gas stream on the N2O conversion at temperature of 698 and 723 K, respectively. As shown in Figure 3B, the N2O conversion is kept constant, 91-93% at 723 K and 6668% at 698 K, not affected by the initial N2O concentration (2-12% in He stream). This suggests that N2O decomposition is a first-order reaction. The presence of 20% O2 in the reaction system (as in the air) shows some negative effect on N2O decomposition. As

FIGURE 4. (A) Time-on-stream behavior of Fe-USY/cordierite catalyst for the decomposition of N2O at different temperatures. Reaction conditions: 2 g of catalyst, 10% N2O, 20% O2, 2% H2O, balance He, GHSV ) 15 000 h-1. (B) Time-on-stream behavior of Fe-USY catalyst for N2O decomposition. Reaction conditions: 0.1 g of catalyst, 10% N2O, 20% O2, 2% H2O, balance He, GHSV ) 30 000 h-1. shown in Figure 3C, the difference is 3-5% in the N2O conversion in the range of 650-750 K. The inhibition effect of O2 on N2O decomposition over Fe-USY catalyst is presumably related to the N2O decomposition mechanism, in which the stream O2 occupies the active sites to reduce N2O chemsorption and also inhibits the release of chemsorbed oxygen formed during N2O decomposition (25). Water vapor in the reaction stream presents a more negative impact on N2O decomposition. As shown in Figure 3C, the presence of 2% H2O reduces the N2O conversion by 10-20% at 650-750 K. The inhibition effect caused by water vapor is supposed to be the result of the competitive adsorption of water molecules on the active sites of the catalyst, which does not essentially affect the N2O decomposition mechanism. When oxygen (20%) and water vapor (2%) coexist in the stream (simulated conditions at a manufacture), the total inhibition effect is approximately accumulative. This may imply that O2 and H2O molecules occupy two different kinds of active sites or they occupy the same kind of active sites but do not affect each other. Nevertheless, Fe-USY still appears as a good catalyst for N2O decomposition in the presence of both oxygen and water vapor, and the 50 and 100% N2O conversion can be achieved at ∼700 and 780 K, respectively. Monolithic Test and Stability of Catalysts. In the above sections, Fe-USY catalyst has proven to be a promising catalyst for N2O decomposition, even in the presence of oxygen and water vapor. For a practical application, the catalyst is normally fixed on structured substrate (26, 27). In this study, we used honeycomb cordierite as the substrate because of its superior mechanical stability, hydrothermal stability, and plasticity (26), and we directly grew the USY on this substrate to make Fe-USY/cordierite, as described in Experimental Section (28, 29). The XRD pattern of Fe-USY/ cordierite composing the typical diffractions of both cordierite and zeolite USY (see Supporting Information Figure 3S) indicates that the zeolite USY has been successfully prepared on the cordierite substrate. The SEM images (see Supporting Information Figure 4S) clearly show that zeolite USY crystals cover the whole surface of cordierite substrate. The asprepared Fe-USY/cordierite contains ∼9.0 wt % USY and 0.5 wt % Fe with an Si/Al ratio of ∼9.5. The nature and distribution of iron species in Fe-USY/cordierite monolith are essentially the same as in Fe-USY from the UV-vis/DRS spectrum (see Supporting Information Figure 5): isolated Fe3+ ions in tetrahedral and octahedral coordination, oligonuclear Fe3+xOy clusters in zeolite channels, and Fe2O3 nanoparticles on the external surface.

The Fe-USY/cordierite catalyst was tested as the monolithic catalyst for N2O decomposition. The catalyst shows a comparable activity for N2O decomposition with Fe-USY in spite that the space velocity is halved (Figure 3D). Under simulated conditions, that is, 10% N2O, 20% O2, 2% H2O, and balanced He, Fe-USY/cordierite exhibits a high activity with the N2O conversion of 18, 63, and 98% occurring at 673, 723, and 773 K, respectively (Figure 4A). In comparison, the pure Fe-USY catalyst (Figure 4B) gives 67 and 98% N2O conversion at 723 and 773 K, respectively, under slightly different conditions. However, when the fact that there is only 9.0 wt % USY and 0.5 wt % Fe contained in the cordierite-supported catalyst is considered, the activity of the monolithic catalyst appears very impressive. Remarkably, both pure Fe-USY catalyst and Fe-USY/ cordierite monolith demonstrate high durability under the simulated conditions. As shown in Figure 4, there is no catalytic activity lost over 144 (monolith) and 100 h (pure Fe-USY) at the selected temperatures. The long durability is comparable with or even superior to the most stable Fezeolite catalyst (19) to our knowledge. The good durability of Fe-USY and Fe-USY/cordierite is in sharp contrast to the case of Fe-ZSM-5 catalysts that were also prepared by similar wet ion exchange. Fe-ZSM-5 catalyst performs well for N2O decomposition, but it suffers from severe activity lost in the presence of water vapor (30). The deactivation of Fe-ZSM-5 catalysts is caused by the dealumination of zeolite framework and subsequent clustering of iron species. As for Fe-USY and Fe-USY/cordierite catalysts, the exchange of most protons in USY matrix with ferric ions reduces the proton-induced dealumination at elevated temperatures. The dealumination could be further prevented by the formation of the specific Al-O-Fe species on the surface. Therefore, a high durability is achieved in this way. In conclusion, Fe-USY and Fe-USY/cordierite catalysts demonstrate a superior catalytic activity and a high durability in N2O thermal decomposition, which can be attributed to the larger channel dimension of H-USY that makes it easier to exchange most protons with ferric ions and to form multiple iron species. The simple preparation procedure and the very low cost present a huge commercial potential of Fe-USY-based catalysts from the current research that could be applied to the end-of-pipe treatment of N2O emissions from, for example, nitric acid, adipic acid, and glyoxal manufacture (31).

Acknowledgments This work was financially supported by the National Natural Science Fund of China (20725723, 20703057) and National VOL. 41, NO. 22, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7905

Basic Research Program of China (2004CB719500). The support from the ARC Centre for Functional Nanomaterials funded by the Australia Research Council under its Centre of Excellence Scheme is also appreciated.

Supporting Information Available Estimation of van der Waals diameter of various hydrated/ hydrolyzed ferric ions, XRD patterns of all samples, SEM images, and UV-vis/DRS spectrum of Fe-USY/cordierite. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Prather, M. J. Time scales in atmospheric chemistry: Coupled perturbations to N2O, NOy, and O3. Science 1998, 279, 13391341. (2) Pe´rez-Ramı´rez, J. Prospects of N2O emission regulations in the European fertilizer industry. Appl. Catal. B 2007, 70, 31-35. (3) Houghton, J. T. IPCC Third Assessment Report: Climate Change, http://www.ipcc.ch/pub/reports.htm, 2001. (4) Kapteijn, F.; Rodrı´guez-Mirasol, J.; Moulijn, J. A. Heterogeneous catalytic decomposition of nitrous oxide. Appl. Catal. B 1996, 9, 25-64. (5) Pe´rez-Ramı´rez, J.; Kapteijn, F.; Mul, G.; Moulijn, J. A. Highly active SO2-resistant ex-framework FeMFI catalysts for direct N2O decomposition. Appl. Catal. B 2002, 35, 227-234. (6) Sun, K.; Xia, H.; Hensen, E.; van Santen, R.; Li, C. Chemistry of N2O decomposition on active sites with different nature: Effect of high-temperature treatment of Fe/ZSM-5. J. Catal. 2006, 238, 186-195. (7) Pirngruber, G. D. The surface chemistry of N2O decomposition on iron-containing zeolites. J. Catal. 2003, 219, 456-463. (8) Pe´rez-Ramı´rez, J.; Groen, J. C.; Bru ¨ ckner, A.; Kumar, M. S.; Bentrup, U.; Debbagh, M. N.; Villaescusa, L. A. Evolution of isomorphously substituted iron zeolites during activation comparison of Fe-beta and Fe-ZSM-5. J. Catal. 2005, 232, 318334. (9) Feng, X.; Hall, W. K. FeZSM-5: A durable SCR catalyst for NOx removal from combustion streams. J. Catal. 1997, 166, 368376. (10) Marturano, P.; Drozdova´, L.; Kogelbauer, A.; Prins, R. Preparation of overexchanged Fe-ZSM-5 zeolites using the ferrous oxalate method: Why does it fail? J. Catal. 2000, 190, 460-468. (11) Melia´n-Cabrera, I.; Espinosa, S.; Groen, J. C.; v/d Linden, B.; Kapteijn, F.; Moulijn, J. A. Utilizing full-exchange capacity of zeolites by alkaline leaching: Preparation of Fe-ZSM5 and application in N2O decomposition. J. Catal. 2006, 238, 250259. (12) Guzma´n-Vargas, A.; Delahay, G.; Coq, B. Catalytic decomposition of N2O and catalytic reduction of N2O and N2O + NO by NH3 in the presence of O2 over Fe-zeolite. Appl. Catal. B 2003, 42, 369-379. (13) Neveu, B.; Hamon, C.; Malefant, K. Catalytic reduction of nitrous oxide content in gases. Fr. Patent WO 99/34901, 1999. (14) Martin, R. L.; Hay, P. J.; Pratt, L. R. Hydrolysis of ferric ion in water and conformational equilibrium. J. Phys. Chem. A 1999, 102, 3565-3573. (15) Pe´rez-Ramı´rez, J.; Kapteijin, F.; Mul, G.; Moulijn, J. A. NO-assisted N2O decomposition over Fe-based catalysts: effects of gasphase composition and catalyst constitution. J. Catal. 2002, 208, 211-223.

7906

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 22, 2007

(16) Melia´n-Cabrera, I.; Mentruit, C.; Pieterse, J. A. Z.; van den Brink, R. W.; Mul, G.; Kapteijn, F.; Moulijn, J. A. Highly active and stable ion-exchanged Fe-Ferrierite catalyst for N2O decomposition under nitric acid tail gas conditions. Catal. Commun. 2005, 6, 301-305. (17) Marturano, P.; Drozdova´, L.; Pirngruber, G. D.; Kogelbauer, A.; Prins, R. The mechanism of formation of the Fe species in Fe/ ZSM-5 prepared by CVD. Phys. Chem. Chem. Phys. 2001, 3, 5585-5595. (18) Kondratenko, E. V.; Pe´rez-Ramı´rez, J. Mechanism and kinetics of direct N2O decomposition over Fe-MFI zeolites with different iron speciation from temporal analysis of products. J. Phys. Chem. B 2006, 110, 22586-22595. (19) El-Malki, El-M.; van Santen, R. A.; Sachtler, W. M. H. Active sites in Fe/MFI catalysts for NOx reduction and oscillating N2O decomposition. J. Catal. 2000, 196, 212-223. (20) Pe´rez-Ramı´rez, J.; Kapteijn, F.; Bru ¨ ckner, A. Active site structure sensitivity in N2O conversion over FeMFI zeolites. J. Catal. 2003, 218, 234-238. (21) Pirngruber, G. D.; Luechinger, M.; Roy, P. K.; Cecchetto, A.; Smirniotis, P. N2O decomposition over iron-containing zeolites prepared by different methods: a comparison of the reaction mechanism. J. Catal. 2004, 224, 429-440. (22) Melia´n-Cabrera, I.; Espinosa, S.; Garcı´a-Montelongo, F. J.; Kapteijn, F.; Moulijn, J. A. Ion exchanged Fe-FER through H2O2assisted decomplexaion of organic salt. Chem. Commun. 2005, 1525-1526. (23) Pieterse, J. A. Z.; Booneveld, S.; van den Brink, R. W. Evaluation of Fe-zeolite catalysts prepared by different methods for the decomposition of N2O. Appl. Catal. B 2004, 51, 215-228. (24) Groen, J. C; Bru ¨ ckner, A.; Berrier, E.; Maldonado, L.; Moulijn, J. A.; Pe´rez-Ramı´rez, J. Iron site modification upon alkaline treatment of Fe-ZSM-5 zeolitessOpportunities for improved N2O decomposition activity. J. Catal. 2006, 243, 212-216. (25) Obalova´, L.; Fı´La, V. Kinetics analysis of N2O decomposition over calcined hydrotalcites. Appl. Catal. B 2007, 70, 353-359. (26) Heck, R. M.; Gulati, S.; Farrauto, R. J. The application of monoliths for gas phase catalytic reactions. Chem. Eng. J. 2001, 82, 149156. (27) Jansen, J. C.; Koegler, J. H.; van Bekkum, H.; Calis, H. P. A.; van den Bleek, C. M.; Kapteijn, F.; Moulijn, J. A.; Geus, E. R.; van der Puil, N. Zeolitic coatings and their potential use in catalysis. Micropor. Mesopor. Mater. 1998, 21, 213-226. (28) Li, L. D.; Xue, B.; Chen, J. X.; Guan, N. J.; Zhang, F. X.; Feng, H. Q.; Liu, D. X. Direct synthesis of zeolite coatings on cordierite supports by in situ hydrothermal method. Appl. Catal. A 2005, 292, 312-321. (29) Li, L. D.; Chen, J. X.; Zhang, S. J.; Guan, N. J.; Wang, T. Y.; Liu, S. L. Selective catalytic reduction of nitrogen oxides from exhaust of lean burn engine over in-situ synthesized Cu-ZSM-5/ cordierite. Environ. Sci. Technol. 2005, 39, 2841-2847. (30) Pieterse, J. A. Z.; Pirngruber, G. D.; van Bokhoven, J. A.; Booneveld, S. Hydrothermal stability of Fe-ZSM-5 and FeBEA prepared by wet ion-exchange for N2O decomposition. Appl. Catal. B 2006, 71, 16-22. (31) Pe´rez-Ramı´rez, J.; Kapteijin, F.; Mul, G.; Xu, X.; Moulijn, J. A. Ex-framework FeZSM-5 for control of N2O in tail-gases. Catal. Today 2002, 76, 55-74.

Received for review July 19, 2007. Accepted August 27, 2007. ES071779G