Applied Catalysis A: General 222 (2001) 163181

Copper ion-exchanged zeolite catalysts in deNOx reaction

Hidenori Yahiro a , Masakazu Iwamoto b,
a b

Faculty of Engineering, Department of Applied Chemistry, Ehime University, Bunkyo-cho 3, Matsuyama 790-8577, Japan Chemical Resources Laboratory, Tokyo Institute of Technology, Nagatsuta 4259, Midori-ku, Yokohama 226-8503, Japan

Abstract Copper ion-exchanged zeolites have widely been studied as the catalysts for NOx emission control. Cu-MFI zeolites, especially over-exchanged Cu-MFI, are very active for the catalytic decomposition of NO. The reaction mechanisms and the active sites suggested are summarized. The newly developed selective catalytic reduction of NO with hydrocarbons in the presence of excess oxygen is then introduced. The role of oxygen and the deactivation of the zeolite catalysts are discussed. In the last chapter, the high ability of Cu-zeolites for NO adsorption is overviewed. Here, some fundamental aspects of the bonding and the molecular diffusion of NO or NO2 molecules in zeolites are also summarized. 2001 Elsevier Science B.V. All rights reserved.
Keywords: Copper ion-exchanged zeolite; deNOx ; HC-SCR; Decomposition; Adsorption; Diffusion

1. Introduction For the last 40 years in the 20th century, the growing need for reduction of nitrogen oxides (NOx ) in emissions from stationary and mobile sources has prompted a worldwide search for deNOx catalysts. A large number of early efforts paid in the 19601970 periods produced two current commercial catalytic systems: noble metal-based three-way catalysts for the purication of automobile emissions and vanadiumtitanium-based catalysts for the control of stationary nitrogen monoxide (NO) emissions by selective catalytic reduction with ammonia [14]. On the other hand, the catalytic decomposition of NO was continuously studied because it is the most ideal reaction among the catalytic methods for removing NO from engine exhaust gases. A few catalysts based on noble metals and metal oxides had been reported to be active for NO decomposition during the 19601970 periods; however, they did not show any consistently

Corresponding author.

high activity. The catalytic activities of copper ionexchanged FAU and MFI zeolites for the decomposition of NO were found by us [5,6] in 1981 and 1986, respectively, and conrmed by Hall and coworker [7] long afterward. This is the starting point of the subsequent intensive studies on Cu-zeolites. Later, the desire for improved fuel economy and lower emissions of carbon dioxide increased the demand for diesel and lean-burn gasoline engines throughout the world. It means the development of catalytic technologies that will allow NOx reduction/ decomposition in lean conditions, e.g. in oxygen-rich environment. For diesel-engine vehicles, the removal of NOx in the presence of SOx is further required. The traditional three-way catalyst, however, cannot control NOx emissions under such conditions. In 1990, one epoch-making result has been reported, which breaks through the deadlock in the removal of NO. It is the selective catalytic reduction of NO with hydrocarbons in an oxidizing atmosphere (HC-SCR). This also destroys the widely accepted concept that ammonia is the only selective reductant for NO in the presence

0926-860X/01/$ see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 6 - 8 6 0 X ( 0 1 ) 0 0 8 2 3 - 7

164

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181

of oxygen and hydrocarbons are not useful for the selective reduction. This novel HC-SCR process was rst reported on copper ion-exchanged zeolites by Iwamoto et al. [8] and by Held et al. [9] independently. The new selective reduction is remarkably enhanced by the presence of oxygen, and proceeds even in the presence of water and SO2 , and could overcome the disadvantages or problems of the present reduction systems. Since the discovery of HC-SCR, a vast number of papers dealing with HC-SCR on metal-containing zeolites, composite metal oxides, and precious metals have been published so far. This review consists of three parts. The rst is the overviewing of the catalytic performance of coppercontaining zeolite for the direct decomposition of NO and the reaction mechanism including the characterization of active sites. HC-SCR of NO is then summarized in the second part. The deactivation of the catalyst is described in this section. The last part deals with the adsorption, bonding, and diffusion of NOx in zeolite. Copper-containing zeolites are active in a wider range of reactions of nitrogen monoxide such as NO + CO reaction [10,11], selective catalytic reduction of NO with ammonia [1217], decomposition of N2 O [18,19], photocatalytic NO [20,21] and N2 O [22] reduction as well as the above decomposition and HC-SCR. Therefore, understanding of afnity of NO

with copper-containing zeolite is important for elucidating the mechanism of NO-related reaction. Thus, the last part sums up the afnity of NO with zeolites.

2. Catalytic decomposition of NO on Cu-zeolites 2.1. Activities of Cu-zeolites NO decomposition to molecular nitrogen and oxygen is the simplest, the most attractive, and the most challenging approach to NOx abatement. Iwamoto et al. have rst reported that Cu ions exchanged into the FAU [5] and MFI [6] matrix exhibit unique and stable activity among metal ions exchanged into zeolites (Table 1). In particular, the over-exchanged Cu-MFI (Cu2+ /Al > 0.5) shows very high decomposition activity [23,24]. This result was conrmed by Li and Hall [7]. Such a high catalytic performance of Cu-MFI will be introduced briey in the beginning of this section. Fig. 1A shows the temperature dependence of the decomposition reaction over Cu-MFI. It should rst be pointed out that the degrees of conversion of NO were not equal to those to N2 and O2 . We suggested that the remaining nitrogen and oxygen balances could be attributed to the formation of NO2 ; indeed, Li and Hall [25] conrmed that the reaction of NO with produced

Table 1 Adsorption, decomposition, and reduction of NO over cation-exchanged ZSM-5 (SiO2 /Al2 O3 = 23.3) zeolites Cation Adsorptiona EL (%)d Amount of NO adsorbed (cm3 g1 )
) qrev (qrev ) qirr (qirr

Decompositionb EL (%)d Conversion into N2 (%)

Reductionc EL (%)d Temperature (K) Conversion into N2 (%)

Na Ca Sr Cu Co Mn Ni Fe H

100 54 105 157 90 127 68 62 100

0.16 1.81 2.71 4.28 1.52 1.19 1.03 0.52 0.12

(0.006) (0.246) (0.159) (0.206) (0.131) (0.069) (0.112) (0.061) (0.004)

0.00 1.56 0.20 14.90 19.96 5.81 6.64 3.08 0.32

(0.000) (0.212) (0.014) (0.716) (1.693) (0.339) (0.727) (0.362) (0.011)

100 80 73 80

0 0 79 2

100 25 102 90 127 99 94 100

673 873 523 673 573 673 473 723

0 12 41 40 27 38 12 39

80

a NO = 1000 ppm, adsorption temperature = 273 K, adsorption weight = 0.5 g, ow rate = 100 cm3 min1 , adsorption time = 35 min, desorption time = 60 min. b NO = 1%, temperature = 823825 K, W/F = 4.0 g s cm3 . c NO = 1000 ppm, C H = 250 ppm, O = 2%, W/F = 0.2 g s cm3 . 2 4 2 d Exchange level of cation.

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181

165

Fig. 1. Temperature dependence of catalytic activity of Cu-MFI for decomposition of NO (A) and selective catalytic reduction of NO with hydrocarbons (B). (A) Copper exchange level = 143%, NO = 1.0% and W/F = 4.0 g s cm1 ; (B) copper exchange level = 137%, NO = 1000 ppm, C2 H4 = 250 ppm, O2 = 2%, and W/F = 0.2 g s cm1 .

O2 to give NO2 proceeds on the catalysts and in the homogeneous phase. Thus, it is clear that Cu-MFI has the ability to stoichiometrically decompose NO to N2 and O2 . Secondly, maximal activity was observed around 823873 K. The best temperature for the reaction was dependent on the catalysts used and the partial pressure of NO in the feed [26]. Several reasons have been suggested for appearance of the maximum activity. We considered it related with the desorption temperature of adsorbed/produced oxygen. Ganemi et al. [27,28] has recently pointed out that the temperature of maximum conversion coincides with the disappearance of surface nitrates which are presumed to be site blockers for NO decomposition. The decrease in the catalytic activity at higher temperature was not attributable to the deactivation of the catalyst since the activity did not change when the reaction temperature was raised and lowered stepwise between 773 and 923 K [24]. The change in the adsorption equilibrium of NO or the properties of copper ions at elevated temperatures would be the reasons of the decrements, though the further research is required in the future. The catalytic activities of copper-zeolites for NO decomposition were strongly dependent on the zeolite structures and the degrees of Cu loading. Fig. 2A shows the NO conversion on various Cu-zeolite catalysts. Clearly, Cu-MFI zeolite showed a good catalytic

activity. The catalytic activity for NO decomposition increased with copper loading, as shown in Fig. 2A. Note that the conversion into N2 over Cu-MFI increased even in the region where the Cu:Al ratio is >0.5 (the dotted line in Fig. 2A) [24]. Many workers have reported similar results. The over-exchanged Cu-MFI was generally prepared by a repeated ionexchange method. Alternative methods have also been proposed; addition of ammonia to the ion-exchange solution to adjust the pH of the solution at 78 [29] or heating of the mixture of CuCl and H-MFI zeolite at 573 K [30]. Although the discovery of exceptionally high activity of the over-exchanged Cu-MFI zeolite for NO decomposition is undoubtedly a milestone in the eld of catalytic deNOx technology, most of the experts believe that the activity of Cu-MFI is not sufcient yet in practical conditions in which the catalyst should work in very low NO concentration, high oxygen concentration, and high space velocity. From this viewpoint modication of Cu-MFI or development of new catalytic system has widely been studied to achieve higher performance in NO decomposition. In the case of zeolites or porous materials, various efforts have been reported. Wichterlov a and coworkers [31] have found that Cu-MeAlPO-11s (Me = Mg or Zn) exhibited constant conversion in NO decom-

166

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181

Fig. 2. Catalytic activity of Cu-MFI (), Cu-MOR ( ), and Cu-FAU () for decomposition (A) and selective catalytic reduction of NO (B). (A) NO = 4.0%, temperature = 823 or 873 K, and W/F = 4.0 g s cm3 (); NO = 1.0%, temperature = 723 K, and W/F = 4.0 g s cm3 (- - -); (B) NO = 880 ppm, C3 H6 = 800 ppm, O2 = 4%, W/F = 0.12 g s cm3 .

position and the turnover frequency values at 770 K were comparable to those of Cu-MFI with high silica matrix. Schay et al. [32] found a similarity in the catalytic activities of Cu-AITS-1 and Cu-MFI. The addition of the second cation to Cu-zeolites improved the activity for NO decomposition. Iwamoto et al. [5] found that Ni and Co are the positive promoters for NO decomposition over Cu-FAU. Kagawa et al. [33] and Zhang and Flytzani-Stephanopoulos [34] have tested the addition effect of a number of cations over Cu-MFI. Among the cations tested, cerium ion possessed a positive effect. Grange and coworkers [35] have pointed out that samarium favorably modied the activity of Cu-MFI when added in small amounts. It has been claimed that Co-MFI zeolite that contains Co in the framework has a considerably larger maximum activity for NO decomposition than Cu-MFI [36] though no data has been reported in a continuous ow system. In the case of metal oxides, Co3 O4 is the most active, single component, metal oxide for NO decomposition [37]. Its activity can be enhanced by the addition of Ag, presumably by modifying the extent of oxygen suppression [38]. It has also been shown that the modication of Co3 O4 by alkali metal ions, in particular, Na is very effective for the enhancement of decomposition activity [39].

YBa2 Cu3 Oy [40] and Sr2+ -substituted perovskite oxides [4143] have been reported as candidates for the catalyst. Lunsford and coworkers [44] have reported high activities of Ba/MgO. On the other hand, the decomposition activity of Pt metal has been established for a long time. Recently, the formation of Tb-nitrate intermediate was observed to be important in NO decomposition over Tb-promoted Pt catalyst [45]. 2.2. Structure of active sites for NO decomposition 2.2.1. Structure of copper ions exchanged into zeolites In order to elucidate the active sites in over-exchanged Cu-MFI, understanding of the mechanism of over-exchange reaction in zeolites is essential; however, it has not been fully claried. Schoonheydt et al. [46] has rst reported the over-exchange of Cu-FAU in the solution of CuCl2 and acetic acid. We [47] have proposed the following exchange schemes on MFI zeolite: Cu2+ (s) + 2Na+ (z) + H2 O Cu(OH)+ (z) + H+ (z) + 2Na+ (s), Cu2+ + H+ (z) + H2 O Cu(OH)+ (z) + 2H+ (s)

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181

167

where s and z indicate in the solution and in the zeolites, respectively. Vaylon and Hall [48] and Centi and Perathoner [49] have independently suggested the formation of Cu(OH)+ in MFI. One and more copper hydrates, such as Cu2 (OH)3+ , Cu(OH)+ , Cu2 (OH)2 2+ and Cu3 (OH)2 4+ , in which the valence of copper was 2, may take part in as copper source [46,50]; further spectroscopic studies on the geometrical structure of copper hydrates in the zeolite matrix are necessary. Delmon and coworkers [51] pointed out that surface enrichment of copper was observed on freshly prepared Cu-MFI, especially on over-exchanged samples, and that copper might migrate into the zeolite channels during heat treatments. The structure of a copper complex with water in Cu-MFI was characterized by ESEEM [52] and ENDOR [53] spectroscopies. Anderson and Kevan [52] have proposed the structure with six octahedrally coordinated water molecules at a distance of 0.28 nm. A corresponding ENDOR measurement provided complementary data including couplings of two groups of protons with anisotropic couplings, which were proposed to be due to water molecules placed along the principal axis or lying on the equatorial plane of the distorted octahedral structure [53]. In addition, the Cu2+ NH distances were estimated to be 0.285 nm for the protons of the two water molecules in the axial positions, and 0.258 nm

for the protons of the four molecules in the equatorial position. Since such a copperwater complex is larger than the diameter of straight channel of Cu-MFI, it is suggested that this complex exists in intersection positions of two channels. 2.2.2. NO species adsorbed on activated Cu-MFI The NO adsorbates on Cu-zeolites activated at high temperature have been extensively investigated through the infrared (IR) technique. We [54] and Hall and coworker [55] have reported IR bands associated with NO species formed on Cu-MFI and Cu-FAU, which are listed in Table 2. The NO stretching frequency in the gas-phase is 1876 cm1 and the bond order of NO is 2.5. This band shifts to higher and lower frequencies when the electron transfers occur from the level of NO to the d-orbital of the metal atom (M NO + ) and from an occupied d-orbital of metal atom to the empty antibonding orbital of NO (M + NO ). In the case of Cu-MFI, the band assigned to Cu+ NO+ (the sign has often been omitted in the literature) appeared at ca. 1905 cm1 . This species is formed on isolated copper ions having a square pyramidal conguration and have a straight conguration of CuNO [5658]. The band at 18981895 cm1 was also assigned to Cu+ NO+ , but it was formed on copper ions, which have an adjacent O anion [59,60], in a squareplanar

Table 2 IR adsorption peaks and their tentative assignment for surface species formed from NO on Cu-MFI and Cu-FAU Zeolite MFI Frequency (cm1 ) 22402230 21202140 19051904 1895 19061895 18271825 18151807 17341730 16301619 13401300 19511946 19121907 19021891 1825 17321740 17961802 1400 Adsorbed species N2 O NO2 + or NO+ NO+ NO NO (NO)2 (sym) NO (NO)2 (asym) NO2 (asym) NO2 (sym), NO3 NO NO NO NO2 (sym) NO NO2 (asym) NO3 Note

On On On On On On

an accessible isolated Cu2+ Cu2+ carrying extra lattice oxygen Cu2+ Cu+ Cu+ Cu+

FAU

On On On On On On

an accessible isolated Cu2+ Cu2+ carrying extra lattice oxygen isolated Cu2+ moved to accessible position Cu+ Cu+ Cu+

168

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181

conguration. On the other hand, the band assignable to Cu2+ NO was detected at ca. 1810 cm1 [20,54,61]. ESR results with well-resolved Cu and N hyperne structure revealed that the Cu2+ NO species possesses an end-on bent structure [6264] and quantum chemistry calculation supported the above geometries [65,66]. When the pressure of NO increased, two IR bands associated to dinitrosyl species, Cu2+ (NO)2 , were observed at 1827 and 1734 cm1 [54]. IR results by Zecchina and coworkers [67] demonstrated the reversible interconversion between Cu2+ NO and Cu2+ (NO)2 through the equilibrium Cu2+ O + NO = Cu2+ NO2 . The bands observed at 21202140 cm1 are discussed on the several papers. In many earlier works, the band at 2133 cm1 was assigned to NO2 in a similar manner to the assignment of Chao and Lunsford [68]. Several years ago, however, the assignment was open to question. Very recently, Hadjiivanov et al. [69] have attributed this band to NO+ by using isotropic tracer method. A few bands in the 16501550 cm1 region have been assigned to monodentate nitrito [35,61] and nitro [70] species formed on copper ions. Nitrate species was also formed on Cu-MFI that gave IR bands at 15751633 cm1 . The IR data have usually been discussed in connection with redox properties of copper-zeolites. The reduction of Cu2+ ions to Cu+ ions upon prolonged evacuation at high temperature has been recognized in Cu-MFI [10] as well as Cu-FAU [71]. As expected by the above IR results [54], the adsorption of NO on Cu2+ and Cu+ ions in zeolites gave Cu+ NO+ and Cu2+ NO (or Cu2+ (NO)2 ), respectively. The redox chemistry of copper species in zeolites has been well-established but the quantitative analysis of reduced copper ions was little reported so far. We [10] have investigated the distribution of Cu2+ and Cu+ ions in Cu-MFI evacuated at 773 K. The amounts of Cu2+ and Cu+ were estimated from ESR and CO adsorption measurements and are shown in Fig. 3. Although the distribution was changed with copper loading, approximately 40% of Cu2+ ions in over-exchanged Cu-MFI were readily reduced to Cu+ . Shelef and coworkers [72] have claimed that there was no thermal reduction of Cu2+ ions up to 773 K but the suggestion gains small support.

Fig. 3. Distribution of copper ions in Cu-MFI as a function of exchange level after oxygen treatment at 773 K. Amounts of Cu2+ and Cu+ were estimated by ESR and CO adsorption measurements, respectively. The dotted line indicates the amounts of total copper ions in Cu-MFI. Open symbols: repeated ion-exchange using aqueous copper (II) acetate solution. Closed symbols: ion-exchange by addition of ammonia into aqueous copper(II) nitrate solution.

2.3. Reaction mechanism of catalytic NO decomposition Several excellent reviews [49,7375] have been published for establishing reaction mechanism of NO decomposition over Cu-zeolites and it is apparent that no general consensus of opinion exists with respect to the nature of the active site involved or indeed the reaction mechanism occurring. The main points of dispute can be summarized as follows. 1. Considerable evidence has been provided to indicate that Cu+ species participate in the reaction [54,60,7680]. On the other hand, the reaction on Cu2+ ion with no contribution of Cu+ has also been postulated [81]. In our opinion, however, there is no doubt that the NO decomposition is a redox process. 2. The NO decomposition reaction is promoted on over-exchanged Cu-MFI catalysts and this behavior may correlate with the availability of extra-lattice

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181

169

oxygen (ELO) species. The identity of the ELO is not clear. We [54,77,78], Sachtler and coworkers [82,83] and Schmal and coworkers [84] have proposed that it is of the form Cu2+ O2 Cu2+ , whereas Bell and coworkers [85] have suggested that the ELO is associated with isolated Cu2+ sites and is of the structure Cu2+ O or Cu2+ O2 . Very recent investigations [75,86] have supported the presence of Cu2+ O or Cu2+ O2 species. 3. The mechanism for coupling of nitrogen species to form N2 is a topic of controversy. There are two kinds of problems to be solved. First signicant problem is that the number of copper ions working as active site is one or two. Second argument is the form of intermediates to produce N2 ; a nitrosyl species, a nitro species, a nitrate species, and dissociatively chemisorbed NO, etc. have been suggested. The rst point of the third item will be discussed here in more detail. It was demonstrated that the most active catalysts are those with the low Si:Al atomic ratios and Cu exchange levels in the range of 90150%. These results have leaded to two kinds of possibilities for copper active sites in Cu-MFI catalysts. One suggestion is that the active site responsible for the high catalytic activity is a unique dimeric Cu species which is stabilized by zeolite framework (Fig. 4). Adsorption of NO on this dimeric species to form a cuprous hyponitrite that decomposes to form N2 O and then N2 is proposed to be a possible reaction mechanism [10,54,80,82,83,8691]. The species Cu2+ O2 Cu2+ , Cu+ O2 Cu2+ , and Cu+ Cu2+ O are suggested [86,88,89]. In contrast, the monomeric Cu site was suggested as an active site by several researchers [30,67,85,92,93] (Fig. 4, [49]). Giamello et al. [61] and Spoto et al. [67] have proposed that oxygen released in the transformation of dinitrosyl species remains bound to the surface and preferentially reacts with a NO molecule to form nitrite/nitrate species. Valyon and Hall [55,94] assume that a Cu+ dinitrosyl complex decomposes, via a hyponitrite intermediate, to Cu2+ ions, nitrous oxide, and extra-lattice oxygen ion. Although Cu+ (NO)2 has been proposed as precursor for N2 O formation in the studies, the lack of correlation between Cu+ (NO)2 and N2 formation [95] and rst principles quantum mechanical calculations [65]

suggest that Cu+ (NO)2 is not formed under reaction conditions. Thus, Cu+ (NO)2 as precursor would be ruled out. The Cu2+ O or Cu2+ O2 species may form on the over-exchanged Cu-MFI and act as the active sites [93]. Detailed characterization of Cu-zeolites has now been carried out by Wichterlov a and coworkers [57,96103], Kuroda and coworkers [50,104112] and other researchers [113115] eagerly, being expected to solve the above controversial reaction mechanism. For example, very recently, the locations of Cu+ ions are proposed on the basis of experimental [103] and theoretical [113] studies and their conclusions are in good agreement with each other. In addition, Kuroda et al. have claimed that zeolite having an appropriate Si:Al ratio, in which it is possible for the copper ions to exist as dimer species, may provide the key to the redox cycle of copper ion as well as catalysis in NO decomposition [112]. This conclusion coincides with the results of theoretical calculation [114,115] in which bent CuOx Cu structures are found in Cu-MFI and these are suggested to be the part of a catalytic cycle.

3. Selective catalytic reduction of NO with hydrocarbons in the presence of excess oxygen 3.1. Development of HC-SCR and outline of catalytic performance of Cu-zeolites Selective reduction of NO with hydrocarbons in an oxidizing atmosphere (HC-SCR) over Cu-MFI has rst been reported by the present group [8]. At the same time, Held et al. [9] have reported similar ndings independently and Toyota Motor Co. also applied for the patents. The distinguishing characteristic of this new technology is that the presence of oxygen is indispensable for the progress of the reduction of NO. This new selective reduction of NO proceeds even in the presence of excess O2 , and has the possibility to overcome the disadvantages of the present reduction systems, NH3 -SCR and three-way catalytic systems. The catalytic activity of Cu-MFI for the selective reduction of NO with C2 H4 is shown in Fig. 1B as a function of the reaction temperature [116]. The temperature at which the conversion into N2 reached its maximum value corresponds to the temperature where

170

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181

Fig. 4. Proposed mechanism for NO decomposition. Details are described in the text.

hydrocarbon oxidation was just completed. At higher temperatures, the conversions into N2 decreased probably due to the more rapid oxidation of hydrocarbon with oxygen. It should be noted that the active temperature region of HC-SCR was lower than that of NO

decomposition. The catalytic activities for HC-SCR were compared among the samples prepared by different methods: mechanical mixture and ion-exchange methods [117]. Cu-MFI prepared from the mechanical mixture of H-MFI and CuCl2 , followed by heating at

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181

171

673 K, gave the comparable activity at 600800 K to that prepared by a conventional ion-exchange method. Fig. 2B shows the maximum catalytic activity of three catalysts, Cu-MFI, Cu-mordenite (Cu-MOR), and Cu-FAU, as a function of copper loading [118]. On Cu-MFI and Cu-MOR catalysts, the catalytic activities increased with the increment of the exchange levels, reached the respective maximums at 80100%, and then decreased gradually. It means that when too much amount of copper ions were incorporated into zeolite, the efciency of the catalyst tends to drop probably because of very high oxidation activity for the hydrocarbons. This dependency is in contrast with the fact that the activity of Cu-MFI for NO decomposition increased even above 100% of exchange level (Fig. 2A). On the other hand, the catalytic activity of Cu-FAU was almost constant independent of the degree of copper loading. Cu-beta zeolite was also reported to show an excellent activity [119,120]. Many hydrocarbons have been examined as reductants. On Cu-zeolites most of hydrocarbons were more or less active but methane was not effective for the SCR reaction. Later, Co- [121], Ga- [122], In- [122], and Pd-zeolites [123] are proposed to be potential candidates for the catalyst in the CH4 -SCR reaction. The efciency of the NO removal is also dependent on the gas composition and the gas hourly space velocity (GHSV) [74,124]. The hydrocarbons in diesel exhausts are better reductant than the trial mixtures in the laboratories, which is probably due to a higher concentration of radicals of hydrocarbons in real exhausts [125]. Water vapor of 710 vol.% in diesel exhausts has an inhibiting effect on the HC-SCR reaction over Cu-MFI. There are two kinds of suppression/ deactivation, fully reversible suppression by short exposure of the catalyst to water vapor and irreversible one after the long-term service of the catalyst at high temperature in water vapor. When Cu-MFI was applied to actual diesel engine exhaust for the short time, it gave high N2 conversions [125]. Cu-SAPO-34 [126] and Cu-IM5 [127] catalysts showed higher durability in water than Cu-MFI. It would be noteworthy that the HC-SCR over Cu-MFI was little inhibited by SO2 [128], which is favorable for the practical application. After the nding of HC-SCR on Cu-zeolites, a great number of metal ion-exchanged zeolites, metal

oxides, and precious metals have been suggested as the catalysts; several reviews [124,129135] have already summarized the progress of HC-SCR in 1996 or before. Here, the catalytic activities of metal ion-exchanged MFI zeolites examined under unied conditions by the present group are summarized in Table 1. In contrast to NO decomposition reaction, it is clear that Cu-MFI is one of the efcient catalysts for HC-SCR. The other zeolites, such as Co-MFI, H-MFI, and Pt-MFI, also showed the good activities for HC-SCR; the catalytic performances of other metal-containing zeolites have been mentioned in another excellent review [136]. 3.2. Mechanism of HC-SCR over Cu-MFI zeolites The reaction mechanism of HC-SCR is more complicated than that of NO decomposition, because the contribution of hydrocarbons and oxygen to reaction mechanism should be considered. Many types of reaction mechanisms have been suggested on Cu-zeolites, the majority of which are still controversial. The reaction schemes proposed over Cu-zeolites can be divided into three categories [74,133,136138]: (a) decomposition of NO proceeds to yield N2 and subsequent regeneration of the active site by the hydrocarbon occurs; (b) some reaction intermediates formed in the partial oxidation of hydrocarbons have the ability to reduce NO selectively and (c) nitrogen oxides generated from NO and O2 , for example, NO2 , which acts as a strongly oxidizing agent, can preferentially react with the hydrocarbons. We believe that the combination of the mechanisms (b) and (c) would proceed on Cu-MFI. In the discussion on the reduction mechanism of NO on Cu-zeolites, the change in the adsorbed species with the copper ion-exchange level should be carefully considered [70,139153]. For example, some types of adsorbed NO were observed on over-exchanged ones, while nitrosyl and nitritenitrate adsorbates were found on low-exchanged ones. The behavior of some surface N-containing intermediates such as nitrosopropane [154] was greatly dependent on the exchange level of copper and on the atmosphere of the catalysts. The role of N-containing surface species in HC-SCR has recently been summarized by Sachtler and coworkers [155].

172

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181

3.3. Deactivation of Cu-MFI As is widely known, Cu-zeolite system in various catalytic reactions including NO reduction has two critical problems, which have not been fully solved yet in the practical point of view. One is that Cu-zeolites are very sensitive to poisoning with H2 O, as stated previously. The other is the low thermal stability of zeolite lattice; treating the Cu-MFI catalyst at or above 873 K results in a deactivation even under dry conditions [156]. The reason why the deactivation occurs in Cu-MFI is debatable. Recent efforts of many researches have been focused on development of catalysts stable for long times in practical conditions. The mechanism of gradual deactivation under relatively mild conditions has not been identied. Formation of CuO particles [156] or clusters [157,158] and migration of Cu2+ ion into inert sites [159,160] have been suggested as the causes. The fresh Cu-MFI samples pretreated at 673773 K usually show two kinds of ESR spectra with g = 2.312.33 and A = 140155 G (CuA ) and g = 2.272.29 and A = 155175 G (CuB ). The spectra have been assigned to the Cu2+ species in squarepyramidal and squareplanar coordinations, respectively. A few research groups [159163] have independently reported that the treatment of Cu-MFI at 1073 K causes the elimination of the CuA and CuB species, the formation of new CuC species with g = 2.302.32 and A = 155160 G, and the simultaneous dealumination of the zeolite lattice. It has been suggested that the dealumination brought about the change in the location of Cu ions and the resulting migration of Cu ions to inert sites is the origin of the deactivation under the mild conditions [160163]. On the other hand, Tabata et al. [157] have not found any dealumination under the similar conditions but observed the formation of Cu Cu bonds by EXAFS. The formation of CuO clusters has been suggested for the deactivation. There is another report [158] in which the CuAl2 O4 formation is associated with the deactivation. Iwamoto et al. [164] have independently compared the ESR, IR, XRD, and 27 Al MAS NMR spectra and the surface areas of the hydrothermally treated Cu-MFI with those of the fresh one. The results have indicated that the migration of Cu ions to inert sites without dealumination resulted in the

deactivation and the change in zeolite lattice occurred under more severe reaction conditions. There are many reports for improvement of the stability of Cu-MFI. Cu-containing silicate has been reported to show better stability than Cu-MFI [165]. The coloading of La or Ce [34,166170], Cr [171], or P [172] stabilized the catalytic activity of Cu-MFI. In particular, the addition of P was very effective. The catalyst treated at 923 K for 50 h in water vapor has still possessed the reduction activity though the active temperature region became higher. The addition of Ca onto the CuP zeolite was reported to be effective for the further improvement of durability.

4. Adsorption and molecular diffusion of NOx in zeolites 4.1. Adsorption of NO on metal ion-exchanged zeolites 4.1.1. Amounts of reversibly and irreversibly adsorbed NO on zeolites Adsorption is the primary step in the catalytic reaction and, therefore, the elucidation of adsorption property is indispensable for understanding the catalytic reactions such as direct decomposition and selective reduction of NO. In the present review, the NO adsorption behavior of metal ion-exchanged zeolites, in particular, Cu-zeolites will be discussed on the basis of our results. Although metal oxides and supported oxides have been reported to show interesting properties for NO adsorption and absorption, detailed discussion of the properties is beyond the scope of this review; please see the other reviews [173]. We have systematically examined the reversible and irreversible adsorption capacities of NO on metal ion-exchanged zeolites (denoted as qrev and qirr , respectively) [174], a part of which are summarized in Table 1. The values in parentheses are the amounts of reversible and irreversible adsorption of NO per and q ). The q cation (qrev rev and qirr greatly changed irr with the kind of metal ions. With MFI zeolites, the order of qrev was transition metal ion alkaline earth metal ion > rare-earth metal ion alkali metal ion proton. Among the adsorbents listed in the table, Cu-MFI showed the largest qrev . Note that qirr of Cu-MFI was also considerably large among those

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181

173

() and q () with Al content in the parent zeolites: (a) MFI; (b) FER; (c) MOR; (d) OFF/ERI; (e) LTL; (f) Fig. 5. (A) Change in qrev irr () and q () of Cu-MFI as a function of copper loading. The concentration of NO and the adsorption FAU(Y); (g) FAU(X). (B) qrev irr temperature are 1910 ppm and 273 K, respectively.

of the metal ions-exchanged MFI zeolites. It is clear that Cu-MFI zeolite has the signicant afnity to NO. The amounts of reversible and irreversible adsorption of NO were strongly dependent on silica:alumina and q decreased ratio of the parent zeolites. Both qrev irr with increment of the aluminum content in zeolites regardless of zeolite structure, as shown in Fig. 5A. Similar results were observed for Co-zeolites [175] and Ag-zeolites [176]. These correlations indicate that the absorbability of NO was mainly controlled by the aluminum content and not by the zeolite structure. The fact that the chemical and physical natures of zeolites depend on the Al content in zeolite was reported in several cases: acid strength in proton-exchanged zeolite [177] and the binding energy of each constituent element of the zeolites [178]. Consistent with these facts, it was concluded that the dependences of the amounts of NO adsorption on the Al content of zeolite reect the change in the electronic state of the zeolite. We pointed out that the catalytic activity of NO decomposition was also dependent on the Al content of zeolite [26]. This suggests that the activity was closely related to the absorbability of NO. The other interesting point is that the qrev and qirr on MFI zeolite were proportional to the exchange level of copper ion as shown in Fig. 6b, showing the

and q are constant, approximately 0.23 and 0.64 qrev irr NO molecule Cu1 , respectively [174]. It follows that the effectiveness of each copper ion in the MFI zeolite for NO adsorption is independent on its loading level or that the ratio of the effective Cu ions for NO adsorption to ineffective ones is constant. Note that both dependencies of qrev and qirr on the copper loading, which show the proportional correlation, are different from that on the catalytic activity of NO decomposition, which gave the rapid increase in catalytic activity at or above approximately 80% of copper loading, as shown in Fig. 1A. One cannot accept the simple explanation that a copper ion, which can adsorb NO, works as an active site for NO decomposition. One possible interpretation can be proposed to explain this phenomenon; the NO decomposition proceeds only with the cooperation of two or more adjacent active sites, in which the copper ions can adsorb NO and, therefore, the catalytic activity emerges at high copper loading, in particular, on over-exchanged Cu-MFI. In real exhaust gases, there coexist various gases such as NO2 , O2 , CO2 , SO2 , CO, and H2 O and, therefore, it is important from a practical point of view to clarify their inuence on adsorption properties. The effect of each gas on amount of NO adsorption was examined on Cu-MFI [174]. The pre-adsorption of NO2

174

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181

sites for the reversible adsorption of NO. When O2 , CO2 , or SO2 was pre-adsorbed, qrev was little reduced (4.26, 4.25, or 3.92 cm3 g1 , respectively). CO or H2 O poisoned the adsorbability (1.39 or 0.22 cm3 g1 , respectively). On the other hand, qirr is always decreased by the pre-adsorption of these gases though the degree of decrement is dependent on the pre-adsorbed gas. The amounts of NO adsorption over metal ionexchanged zeolites were compared with the catalytic activities for NO decomposition and HC-SCR in Table 1. One can found that Cu- or Co-MFI show the great abilities for NO adsorption. In addition, these zeolites show the high catalytic activities for the selective reduction NO. It is likely that the adsorbability of NO might be closely related to the catalytic activity; further investigations are required to clarify this correlation. 4.1.2. ESR studies on NO molecules adsorbed on zeolites The state of NO adsorbed on metal ion exchanged into zeolites is investigated by several spectroscopic techniques. The most popular technique is infrared method as mentioned in the Section 2.2.2 and a review in [179]. On the other hand, ESR and related techniques sometimes provide the useful information about geometrical structure of NO adsorbed on metal ion exchanged into zeolite [180183]. As an example, ESR spectrum of NO adsorbed on Na-A zeolites, recorded at 5 K, is illustrated in Fig. 6 [184,185]. The resolved three different g-tensor components and the resolved y-component hyperne coupling with the 14 N nucleus (I = 1) were observed in this spectrum, indicating that NO molecules are attached to the Na cations in a bent conguration. The gzz tensor component is very sensitive to the cation introduced into the zeolite, reecting the degree of separation of the orbital. The ordering of gzz (=ge gzz ) was Na+ > Ba2+ > Zn2+ H+ for the FAU zeolite, Na+ > Zn2+ for the LTA zeolite, and Na+ > H+ for the MFI zeolite [186]. It is clear that Na+ -exchanged zeolites showed larger gzz than the divalent cationor proton-exchanged zeolites, indicating the weaker electrostatic eld associated with Na+ ions. This result is consistent with earlier reports that the electrostatic eld in the vicinity of divalent and trivalent cations and protons exchanged into the zeolites is stronger than that of monovalent cations [187]. gzz is also

Fig. 6. Experimental () and simulated ( ) Q-band ESR spectra of NO (13.2 kPa) introduced in Na-LTA at 5 K. The theoretical spectra (c) for NO (I) and (d) for the NO biradical were calculated using Lorentzian lineshape with anisotropic linewidth of 40, 75, and 60 G for the x-, y-, and z-components, respectively. The parameters used are gxx = 2.0019, gyy = 1.9961, gzz = 1.8856, Axx = 0, Ayy = 33, and Azz = 0 for NO monoradical and |D | = 331 G and |E | = 28 G for NO biradical. Trace (b) is a superposition of (c) and (d) with the weighted ratio 1:1.

on Cu-MFI-147 resulted in the enhancement of qrev (from 4.35 cm3 g1 without pre-adsorbed NO2 to 7.14 after the pre-adsorption of NO2 ). At low temperature, N2 O3 is known to be in equilibrium with NO and NO2 . NO2 irreversibly adsorbed can work as new active

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181

175

sensitive to the zeolite structure, Na-MOR > NaMFI > Na-LTA, suggesting a stronger electrostatic eld associated with Na+ ions in LTA zeolite [188]. Another study by Kasai and Gaura [189] has revealed an interesting observation that the ESR spectrum of NO in Na-LTA consisted of two signals, one due to the NO monomer and the other one due to an unusual NONO triplet species (see the inset in Fig. 6). The facts that only the NO monomer was detectable when the NO pressure was low, while the triplet became dominant with higher NO pressure, and that the half-eld signal due to the Ms = 2 transition was detected when the corresponding triplets were observed at the normal eld of g 2, secured the assignment of the NONO triplet species. The D parameter of zero-eld splitting, as shown in Fig. 6, depends on the average distance between two radicals, R, according to the relation D = 3g/2R 3 . The values of R evaluated from the experimental D-value (331 G) were in the range 0.45 nm [184]. This unusual NONO triplet was observed in the spectrum of NO adsorbed on Na-MOR [186]. The presence of such NONO triplet species in zeolite is likely to become the important information to evaluate the reaction mechanism of NO decomposition. 4.2. Molecular diffusion of NOx in zeolites Many studies not only on the removal of NOx but also on some catalytic reactions using zeolites lead to some critical unanswered questions; why are the zeolite-supported catalysts more active than metal and metal oxide-supported catalysts and which zeolite structure is the best and why? For example, the MFI structure is the most efcient support for the direct decomposition of NO as mentioned in Section 2.1, but the explanation is still obscure despite several plausible attempts [190]. A large stride to solve the above problems would be to recognize the diffusion of reactants (or products) in zeolite channels and the interaction between the host zeolite (or the catalytically active site) and the guest molecule in the channel and/or cavity of zeolites. Hereinafter, these points will be discussed. It is well-known that zeolite catalysts have more advantages such as providing the coordinately unsaturated metal ions, enrichment of reactants in zeolite pore and cavity, coexistence of metal cations and

acidic protons, and so on than metal and metal oxide catalysts. On the other hand, the most plausible negative effects on catalytic performances may be the high resistance of mass transfer in zeolite channel, that is, severe intracrystalline diffusion. Tabata and Ohtsuka [191] have found the dependence of HC-SCR activity on primary crystal size of Co-MFI; the reaction rate on Co-MFI with large crystal size (1.3 m, mean crystal diameter) was less than that from small one (0.1 m). A similar result was observed on Co-MOR [192]. These results indicate that the rate of HC-SCR reaction on Co-zeolites is controlled by intracrystalline diffusion. In addition, the presence of SO2 inhibited the diffusion of reactants with Co-BEA because the micropores in BEA are clogged or narrowed by the adsorbed SO2 [193]. Shichi et al. [194] have found that the kind of reductant inuenced on the intracrystalline diffusivity in HC-SCR over Cu-MFI catalysts; the reaction rates of NO conversion was dependent on the crystal size in the NOC2 H4 O2 reaction, but not in the NOC3 H8 O2 reaction. This is caused by the stronger interaction of C2 H4 with copper ions than that of C3 H8 , that is, the adsorption properties of hydrocarbon on Cu-MFI play an important role in determining intracrystalline diffusivity. The spectroscopic investigations of the physical and chemical properties of the intracrystalline void systems of zeolites have been carried out. NMR study on 129 Xe adsorbed on zeolites is of this case [195]. This is a convenient and sensitive tool and widely used in the eld of zeolite study. An alternative spectroscopic investigation, which will be described in this section, is ESR measurement using nitrogen dioxide, NO2 , as a probe molecule at the gassolid interface. In earlier works, molecular dynamics of NO2 adsorbed on Na-FAU zeolites were analyzed by temperature-dependent ESR method [196,197]. At low temperature, the ESR spectrum showed an anisotropic triplettriplet with respect to the coordinate system given in the inset in Fig. 7. This spectrum is very close to the rigid limit in the ESR time scale (average rotational correlation time > 106 s). 1 Motional effects on the spectra became pronounced with
R R R = (6 R R )1 for simple is dened as R = Brownian diffusion, where R and R are the principal values of axial and perpendicular rotational diffusion tensor and R and R are the corresponding rotational correlation times.
1

176

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181

Fig. 7. ESR spectra at 77 K (a), 160 K (b), and 240 K (c) of NO2 adsorbed on Na-MOR. The NO2 pressure at adsorption was 1.3 kPa. Dotted lines indicate ESR spectra of NO2 simulated to t the experimental ones, with different Heisenberg exchange rates: (a ) 0.5 s1 ; (b ) 70 s1 and (c ) 340 107 s1 . The spectra were obtained with a linewidth of 1/T20 = 0.05 mT and were convoluted with a Gaussian linewidth.

increasing temperature, resulting that the spectra consist essentially of an isotropic and equally spaced hyperne triplet. The line shape simulations were done by adopting the Brownian rotational diffusion model in order to evaluate the associated (average) rotational correlation time, R , and its degree of anisotropy, N = R /R . It was found that the value of R decreased from 1.7 109 (230 K) to 7.5 1010 s (325 K) with increasing temperature and N was very close to 1 (N = 1.25) in the motional narrowing region. The Arrhenius plots gave 5.9 kJ mol1 for the activation energy, which was evaluated for the nearly isotropic rotational diffusion of NO2 in Na-FAU zeolite. On the other hand, temperature-dependent ESR spectra of NO2 adsorbed on Na-FAU(Y) zeolites were observed to be somewhat different from that of Na-FAU(X) zeolites [196]. The most remarkable difference from ESR spectra of NO2 adsorbed on Na-FAU(X) is that the x- and z-components at the M I = 1 and 0 bands have collapsed into one broad line in the spectrum measured even at 4.8 K for the Na-FAU(Y) zeolite. This suggests that the rotational diffusion of NO2 is axially symmetric about the molecular y-axis and is fast enough to average the Azz and Axx values. The activation energy of the

rotational diffusion about the x- or z-axis was estimated to be 0.4 kJ mol1 in the temperature range 13209 K, which is much smaller than that for NO2 /Na-FAU(X) zeolites. The ESR line shapes of NO2 adsorbed on Na-MOR and Na-MFI cannot be well-simulated by adopting the Brownian rotational diffusion model [198,199]. Instead, the best agreement with the experimental lineshapes was obtained with a Heisenberg type of spin exchange (Fig. 7). The result leads to the conclusion that the main cause for the reversible spectral change with temperature is due to the Heisenberg type of exchange. This conclusion is in agreement with the observation that the spectral resolution was lower in samples exposed to high NO2 pressure (13.3 kPa) than in samples exposed to low pressure (0.131.33 kPa). The rotational diffusion may occur to some extent but its effect on the lineshapes is hidden by the dominating exchange interaction. Recently, the analysis utilizing the Heisenberg spin exchange model was improved by adopting a rate distribution [200]. The dynamics of NO2 was strongly dependent on the kind of zeolites [201], Si:Al ratio [199], and kind of cations [202]. These results and detail of analytical technique have been described in original papers.

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181

177

Fig. 8. Schematic view of molecular diffusion of NO2 in various zeolites.

A preferential rotation about the y-axis takes place in several systems, especially at low temperature. This has been attributed to the low moment of inertia about the axis through the center of mass parallel to the OO direction. The rotational dynamics tend to be more isotropic at higher temperature. The diffusion may preferentially take place for NO2 in a vacancy, a pore, and a cage. The FAU zeolites are of this type. The effect of Heisenberg spin exchange was pronounced for the zeolites with channel structure. We tentatively assign the Heisenberg exchange dynamics to the interaction between NO2 molecules undergoing translation motion in the same channel. This cannot easily take place in vacancies and cages with room for one or a few NO2 molecules (Fig. 8). Temperature at which the rigid limit spectra were observed was dependent on the kind of zeolite with channel structure in the following order: MFI (<3 K) BEA (<4.2 K ) < FER (<20 K) < LTL (30 K) < MOR (77 K); e.g. the Heisenberg spin

exchange rate increased in the following order: MFI BEA > FER > LTL > MOR [201]. From this ordering, we conclude the following: (1) the rate is faster in multiple-channel structural zeolites (MFI, BEA, and FER) than in the single-channel zeolites (LTL and MOR) and (2) in zeolites with similar channel structure, the exchange rate is proportional to the channel size. Provided that the order prevails also at high temperature, this indicates that the NO removal in zeolites might be a diffusion-controlled reaction. Unfortunately, such a study dealing with NO2 diffusion was not applied for the catalysts active for NO reduction, such as Cu-MFI and transition metal ion-exchanged zeolite. It is expected to be done in near future.

5. Conclusions In this review, the removal of nitrogen monoxide over copper ion-exchanged zeolites has mainly been

178

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181 [18] G.I. Panov, V.I. Sobolev, A.S. Kharitonov, J. Mol. Catal. 61 (1990) 85. [19] Y. Li, J.N. Armor, Appl. Catal. B 1 (1992) L21. [20] M. Anpo, M. Matsuoka, Y. Shioya, H. Yamashita, E. Giamello, C. Morterra, M. Che, H.H. Patterson, S. Webber, S. Ouellette, M.A. Fox, J. Phys. Chem. 98 (1994) 5744. [21] M. Anpo, M. Matsuoka, K. Hanou, H. Mishima, H. Yamashita, H.H. Patterson, Coord. Chem. Rev. 171 (1998) 157. [22] K. Ebitani, M. Morokuma, A. Morikawa, Stud. Surf. Sci. Catal. 84 (1994) 1501. [23] M. Iwamoto, H. Yahiro, Y. Mine, S. Kagawa, Chem. Lett. (1989) 213. [24] M. Iwamoto, H. Yahiro, K. Tanda, N. Mizuno, Y. Mine, S. Kagawa, J. Phys. Chem. 95 (1991) 3727. [25] Y. Li, W.K. Hall, J. Phys. Chem. 94 (1990) 6145. [26] M. Iwamoto, H. Yahiro, K. Tanda, Stud. Surf. Sci. Catal. 44 (1988) 219. [27] B. Ganemi, E. Bj onbom, J. Paul, Appl. Catal. B 17 (1998) 293. [28] B. Ganemi, E. Bj onbom, B. Demirel, J. Paul, Micropor. Mesopor. Mater. 38 (2000) 287. [29] M. Iwamoto, H. Yahiro, Y. Torikai, T. Yoshioka, N. Mizuno, Chem. Lett. (1990) 1967. [30] G. Spoto, A. Zecchina, S. Bordiga, G. Ricchiardi, G. Martra, G. Leofanti, G. Petrini, Appl. Catal. B 3 (1994) 151. [31] J. Dedecek, J. Cejka, B. Wichterlov a, Appl. Catal. B 15 (1998) 233. [32] Z. Schay, L. Guczi, Zs. Kopp any, I. Nagy, A. Beck, V. Samuel, M.K. Dongare, D.P. Sabde, S.G. Hegde, A.V. Ramaswamy, Catal. Today 54 (1999) 569. [33] S. Kagawa, H. Ogawa, H. Furukawa, Y. Teraoka, Chem. Lett. (1991) 407. [34] Y. Zhang, M. Flytzani-Stephanopoulos, J. Catal. 164 (1996) 131. [35] V.I. P arvulescu, P. Oelker, P. Grange, B. Delmon, Appl. Catal. B 16 (1998) 1. [36] Y. Chang, J.G. McCarty, J. Catal. 178 (1998) 408. [37] G.K. Boreskov, Discuss. Faraday Soc. 41 (1966) 263. [38] H. Hamada, Y. Kintaichi, M. Sasaki, T. Ito, Chem. Lett. (1990) 1069. [39] P.W. Park, J.K. Kil, H.H. Kung, M.C. Kung, Catal. Today 42 (1998) 51. [40] H. Shimada, S. Miyama, H. Kuroda, Chem. Lett. (1988) 1797. [41] T. Uchijima, Hyomen 18 (1987) 132. [42] Y. Teraoka, H. Fukuda, S. Kagawa, Chem. Lett. (1990) 1. [43] H. Yasuda, N. Mizuno, M. Misono, J. Chem. Soc., Chem. Commun. (1990) 1094. [44] S. Xie, M.P. Rosynek, J.H. Lunsford, J. Catal. 188 (1999) 24. [45] S.S.C. Chuang, C.-D. Tan, J. Phys. Chem. B 101 (1997) 3000. [46] R.A. Schoonheydt, L.J. Vondamme, P.A. Jacobs, J.B. Uytterhoeven, J. Catal. 43 (1976) 292. [47] M. Iwamoto, N. Mizuno, H. Yahiro, T. Yoshioka, New Developments in Ion-Exchange, Kodansha, Tokyo, 1991, p. 407.

introduced. It is widely accepted that copper-zeolites show the catalytic activity for NO decomposition and that they are useful models for the investigation of the fundamental aspects in the chemistry of the interaction and surface transformation of nitrogen oxides. The discovery of HC-SCR over copper-zeolites is one of the main topics in the development of environmental catalysis in the last century. Such environmental catalysts, in general, have to work under very severe conditions; very wide temperature ranges, high space velocities, low concentrations of target materials, very high concentrations of coexisting gases and poisons, and considerable changes in the reaction conditions. Namely, environmental catalysts must have extremely high activities, selectivities, and durabilities. We expect much progress in the near future, both with respect to the development of environmentally benign technology and in the scientic understanding of the catalytic action of deNOx . References
[1] J.W. Hightower, D.A. van Leirsberg, in: R.L. Klimish, J.G. Larson (Eds.), The Catalytic Chemistry of Nitrogen Oxides, Plenum Press, New York, 1975, p. 63. [2] H. Bosch, F.J.J.G. Janssen, Catal. Today 4 (1988) 1. [3] A. Crucq, A. Ferennet, Catalysis and Automotive Pollution Control, Elsevier, Amsterdam, 1987, p. 1. [4] K.C. Taylor, in: J.R. Anderson, M. Boudart (Eds.), Catalysis, Springer, Berlin, 1984, p. 119. [5] M. Iwamoto, S. Yokoo, K. Sakai, S. Kagawa, J. Chem. Soc., Faraday Trans. 1 (77) (1981) 1629. [6] M. Iwamoto, H. Furukawa, Y. Mine, F. Uemura, S. Mikuriya, S. Kagawa, J. Chem. Soc., Chem. Commun. (1986) 1272. [7] Y. Li, W.K. Hall, J. Catal. 129 (1991) 202. [8] M. Iwamoto, H. Yahiro, Y. Yu-u, S. Shundo, N. Mizuno, Shokubai 32 (1990) 430. [9] W. Held, A. Konig, T. Richter, L. Puppe, SAE Transaction, Section 4, No. 900469, 1990, p. 209. [10] M. Iwamoto, H. Yahiro, N. Mizuno, J. Chem. Soc. Jpn. (1991) 574. [11] C.M. Alvarez, G.S. McDougall, A.G. Ruiz, I.R. Ramos, Appl. Surf. Sci. 78 (1994) 477. [12] G. Centi, C. Nigro, S. Perathoner, G. Stella, Catal. Today 17 (1993) 159. [13] T. Komatsu, M. Nunokawa, I.S. Moon, T. Takahara, S. Namba, T. Yashima, J. Catal. 148 (1994) 427. [14] W. Wang, S.-J. Hwang, Appl. Catal. B 5 (1995) 187. [15] G. Delahay, S. Kieger, B. Neveu, B. Coq, Surf. Chem. Catal. (1998) 229. [16] S. Kieger, G. Delahay, B. Coq, Appl. Catal. B 25 (2000) 1. [17] A.V. Salket, W. Weisweiler, Appl. Catal. A 203 (2000) 221.

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181 [48] J. Vaylon, W.K. Hall, Catal. Lett. 19 (1993) 109. [49] G. Centi, S. Perathoner, Appl. Catal. A 132 (1995) 179. [50] Y. Kuroda, A. Kotani, H. Maeda, H. Moriwaki, T. Morimoto, M. Nagao, J. Chem. Soc., Faraday Trans. 88 (1992) 1583. [51] T. Curtin, P. Grange, B. Delmon, Catal. Today 35 (1997) 121. [52] M. Anderson, L. Kevan, J. Phys. Chem. 91 (1987) 4174. [53] H. Li, D. Biglino, R. Erickson, A. Lund, Chem. Phys. Lett. 266 (1997) 417. [54] M. Iwamoto, H. Yahiro, N. Mizuno, W.-X. Zhang, Y. Mone, H. Furukawa, S. Kagawa, J. Phys. Chem. 96 (1992) 9360. [55] J. Valyon, W.H. Hall, J. Phys. Chem. 97 (1993) 1204. [56] J. Szanyi, M.T. Paffett, J. Catal. 164 (1996) 232. [57] J. Dedecek, Z. Sobalik, Z. Tvaruzkova, D. Kaucky, B. Wichterlov a, J. Phys. Chem. 99 (1995) 16327. [58] V.I. P arvulescu, P. Grange, B. Delmon, J. Phys. Chem. B 101 (1997) 6933. [59] A. Aylor, S. Larsen, J. Reimer, A.T. Bell, J. Catal. 157 (1995) 569. [60] H. Jang, W.K. Hall, J. dItri, J. Phys. Chem. 100 (1996) 9416. [61] E. Giamello, D. Murphy, G. Magnacca, C. Morterra, Y. Shioya, T. Nomura, M. Anpo, J. Catal. 136 (1992) 510. [62] C. Naccache, M. Che, Y. Ben Taarit, Chem. Phys. Lett. 13 (1972) 109. [63] C.C. Chao, J.H. Lunsford, J. Phys. Chem. 76 (1972) 1546. [64] Z. Sojka, M. Che, E. Giamello, J. Phys. Chem. B 101 (1997) 4831. [65] B.L. Trout, A.K. Chakraborty, A.T. Bell, J. Phys. Chem. 100 (1996) 17582. [66] Y. Yokomichi, T. Yamabe, H. Ohtsuka, T. Kakumoto, J. Phys. Chem. 100 (1996) 14424. [67] G. Spoto, S. Bordiga, D. Scarano, A. Zecchina, Catal. Lett. 13 (1992) 39. [68] C.C. Chao, J.H. Lunsford, J. Am. Chem. Soc. 93 (1971) 71. [69] K. Hadjiivanov, J. Saussey, J.L. Freysz, J.C. Lavalley, Catal. Lett. 52 (1998) 103. [70] T. Beutel, B. Adelman, W.M.H. Sachtler, Appl. Catal. B 9 (1996) L1. [71] P.A. Jacobs, W. De Wilde, R.A. Shoonheydt, J.B. Uytterhoeven, Trans. Faraday Soc. 5 (1976) 1221. [72] A.V. Kucherov, J.L. Gerlock, H.-W. Jen, M. Shelef, J. Phys. Chem. 98 (1994) 4892. [73] M. Iwamoto, N. Mizuno, H. Yahiro, J. Jpn. Petrol. Inst. 34 (1991) 375. [74] M. Iwamoto, H. Yahiro, Catal. Today 22 (1994) 5. [75] G.J. Millar, A. Canning, G. Rose, B. Wood, L. Trewartha, I.D.R. Mackinnon, J. Catal. 183 (1999) 169. [76] D.-J. Liu, H.J. Robota, Catal. Lett. 21 (1993) 291. [77] Y. Hoshino, M. Iwamoto, Chem. Lett. (1996) 631. [78] M. Iwamoto, Y. Hoshino, Inorg. Chem. 35 (1996) 6918. [79] D.-J. Liu, H.J. Robota, Appl. Catal. B 4 (1994) 155. [80] J. Valyon, W.K. Hall, J. Phys. Chem. 97 (1993) 7054. [81] M. Shelef, Chem. Rev. 95 (1995) 209. [82] J. Sarkany, J.L. dItri, W.M.H. Sachtler, Catal. Lett. 16 (1992) 241. [83] G.D. Lei, B.J. Adelman, J. Sarkany, W.M.H. Sachtler, Appl. Catal. B 5 (1995) 245.

179

[84] N.C.N.A. Carvalho, F.B. Passos, M. Schmal, Appl. Catal. A 193 (2000) 265. [85] S.C. Larsen, A.W. Aylor, A.T. Bell, J.A. Reimer, J. Phys. Chem. 98 (1994) 11533. [86] G. Moretti, C. Dossi, A. Fusi, S. Recchia, R. Psaro, Appl. Catal. B 20 (1999) 67. [87] J. Sarkany, W.M.H. Sachtler, Zeolites 14 (1994) 7. [88] G. Moretti, Catal. Lett. 23 (1994) 135. [89] M.C. Campa, V. Indovina, G. Minelli, G. Moretti, I. Pettiti, P. Porta, A. Piccio, Catal. Lett. 23 (1994) 141. [90] K.C.C. Kharas, Appl. Catal. B 2 (1993) 207. [91] D.C. Sayle, C. Richard, A. Catlow, M.-A. Perrin, P. Nortier, Micropor. Mesopor. Mater. 20 (1998) 259. [92] B. Wichterlov a, J. Dedecek, A. Vondrov a, J. Phys. Chem. 99 (1995) 1065. [93] M.V. Konduru, S.S.C. Chuang, J. Catal. 187 (1999) 436. [94] J. Valyon, W.K. Hall, Stud. Surf. Sci. Catal. 75 (1993) 1333. [95] M.V. Konduru, S.S.C. Chuangs, J. Phys. Chem. 103 (1999) 5802. [96] J. Dedecek, B. Wichterlov a, J. Phys. Chem. 98 (1994) 5721. [97] B. Wichterlov a, Z. Sobalik, A. Vondrov a, Catal. Today. 29 (1996) 149. [98] B. Wichterlov a, J. Dedecek, Z. Sobalik, A. Vondrov a, K. Klier, J. Catal. 169 (1997) 194. [99] J. Dedecek, B. Wichterlov a, J. Phys. Chem. B 101 (1997) 10233. [100] B. Wichterlov a, Z. Sobalik, J. Dedecek, Catal. Today 38 (1997) 199. [101] Z. Sobalik, J. Dedecek, I. Ikonnikov, B. Wichterlov a, Micropor. Mesopor. Mater. 21 (1998) 525. [102] Z. Sobalik, Z. Traruzkova, B. Wichterlov a, J. Phys. Chem. B 102 (1998) 1077. [103] J. Dedecek, B. Wichterlov a, Phys. Chem. Chem. Phys. 1 (1999) 629. [104] Y. Kuroda, S. Konno, H. Maeda, Y. Yoshikawa, Jpn. J. Appl. Phys. 32 (1993) 493. [105] Y. Kuroda, Y. Yoshikawa, S. Konno, H. Hamano, H. Maeda, R. Kumashiro, M. Nagao, J. Phys. Chem. 99 (1995) 10621. [106] Y. Kuroda, H. Maeda, Y. Yoshikawa, R. Kumashiro, M. Nagao, J. Phys. Chem. B 101 (1997) 1312. [107] Y. Kuroda, S. Konno, Y. Yoshikawa, H. Maeda, Y. Kubozono, H. Hamano, R. Kumashiro, M. Nagao, J. Chem. Soc., Faraday Trans. 93 (1997) 2125. [108] Y. Kuroda, Y. Yoshikawa, R. Kumashiro, M. Nagao, J. Phys. Chem. B 101 (1997) 6497. [109] R. Kumashiro, Y. Kuroda, M. Nagao, J. Phys. Chem. B 103 (1999) 89. [110] Y. Kuroda, Y. Yoshikawa, S. Emura, R. Kumashiro, M. Nagao, J. Phys. Chem. B 103 (1999) 2155. [111] Y. Kuroda, T. Mori, Y. Yoshikawa, S. Kittaka, R. Kumashiro, M. Nagao, Phys. Chem. Chem. Phys. 1 (1999) 3807. [112] Y. Kuroda, Y. Kumashiro, T. Yoshimoto, M. Nagao, Phys. Chem. Chem. Phys. 1 (1999) 649. [113] D. Nachtigallova, P. Nachtigall, M. Sierka, J. Sauer, Phys. Chem. Chem. Phys. 1 (1999) 2019. [114] B.R. Goodman, K.C. Hass, W.F. Schneider, J.B. Adams, Catal. Lett. 56 (1998) 183.

180

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181 [150] H. Takeda, M. Iwamoto, Catal. Lett. 38 (1996) 21. [151] H. Kato, C. Yokoyama, M. Misono, Catal. Lett. 47 (1997) 189. [152] A. Satuma, A.D. Cowan, N.W. Cant, D.L. Trimm, J. Catal. 181 (1999) 165. [153] S.A. Beloshapkin, E.A. Paukshtis, V.A. Sadykov, J. Mol. Catal. A 158 (2000) 355. [154] B.J. Adelman, T. Beutel, G. Lei, W.M.H. Sachtler, Appl. Catal. B 11 (1996) L1. [155] H.-Y. Chen, T. Voskoboinikov, W.M.H. Sachtler, J. Catal. 186 (1999) 91. [156] K.C.C. Kharas, H.J. Robota, D.J. Liu, Appl. Catal. B 2 (1993) 225. [157] T. Tabata, M. Kokitsu, O. Okada, T. Nakayama, T. Yasumatsu, H. Sakane, Stud. Surf. Sci. Catal. 88 (1994) 409. [158] J.Y. Yan, G.D. Lei, W.M.H. Sachtler, H.H. Kung, J. Catal. 161 (1996) 43. [159] S. Matsumoto, K. Yokota, H. Doi, M. Kimura, K. Sekizawa, S. Kasahara, Catal. Today 22 (1994) 127. [160] T. Tanabe, T. Iijima, A. Koiwai, J. Mizuno, K. Yokota, A. Isogai, Appl. Catal. B 6 (1995) 145. [161] A.V. Kucherov, A.A. Slinkin, S.S. Goryashenko, K.I. Slovetskaja, J. Catal. 118 (1989) 459. [162] A.V. Kucherov, C.P. Hubbard, M. Shelef, J. Catal. 157 (1995) 603. [163] A.V. Kucherov, J.L. Gerlock, H.-W. Jen, M. Shelef, J. Catal. 152 (1995) 63. [164] M. Iwamoto, J. Wang, K.M. Sperati, T. Sajiki, M. Misono, Chem. Lett. (1997) 1281. [165] T. Inui, S. Iwamoto, K. Matsuba, Y. Tanaka, T. Yoshida, Catal. Today 26 (1995) 23. [166] A.V. Kucherov, C.P. Hbbard, T.N. Kucherova, M. Shelef, Appl. Catal. B 7 (1996) 285. [167] J.Y. Yan, W.M.H. Sachtler, H.H. Kung, Catal. Today 33 (1997) 279. [168] P. Budi, R.F. Howe, Catal. Today 38 (1998) 175. [169] M. Guyon, V. Le Chanu, P. Gilot, H. Kessler, G. Prado, Stud. Surf. Sci. Catal. 116 (1998) 297. [170] W.E.J. van Kooten, J. Kaptein, C.M. van den Bleek, H.P.A. Calis, Catal. Lett. 63 (1999) 227. [171] R.L. Keiski, H. Raisamen, M. Harkonen, T. Maunnula, P. Niemisto, Catal. Today 27 (1996) 85. [172] T. Kesen, Y. Kawashima, H. Akama, M. Kamikubo, Shokubai 39 (1997) 119. [173] H. Arai, M. Machida, Catal. Today 22 (1994) 97. [174] W.-X. Zhang, H. Yahiro, N. Mizuno, J. Izumi, M. Iwamoto, Langmuir 9 (1993) 2337. [175] W.-X. Zhang, H. Yahiro, M. Iwamoto, J. Izumi, J. Chem. Soc., Faraday Trans. 91 (1995) 767. [176] W.-X. Zhang, M.-J. Jia, J.-F. Yu, T.-H. Wu, H. Yahiro, M. Iwamoto, Chem. Mater. 11 (1999) 920. [177] D. Barthomeuf, J. Phys. Chem. 83 (1979) 249. [178] Y. Okamoto, M. Ogawa, A. Maezawa, T. Imanaka, J. Catal. 112 (1988) 427. [179] K.I. Hadjiivanov, Catal. Rev. 42 (2000) 71. [180] J.H. Lunsford, J. Phys. Chem. 72 (1968) 4163. [181] C.L. Gardner, M.A. Weinberger, Can. J. Chem. 48 (1970) 1317.

[115] B.R. Goodman, K.C. Hass, W.F. Schneider, J.B. Adams, J. Phys. Chem. B 103 (1999) 10452. [116] S. Sato, Y. Yu-u, H. Yahiro, N. Mizuno, M. Iwamoto, Appl. Catal. 70 (1991) L1. [117] F.-S. Xiao, W. Zhang, M. Jia, Y. Yu, C. Fang, G. Tu, S. Zheng, S. Qiu, R. Xu, Catal. Today 50 (1999) 117. [118] C. Torre-Abreu, C. Henriques, F.R. Riberio, G. Delahay, M.F. Ribeiro, Catal. Today 54 (1999) 407. [119] A. Corma, V. Forn es, E. Palomares, Appl. Catal. B 11 (1997) 233. [120] G. Delahay, B. Coq, L. Broussous, Appl. Catal. B 12 (1997) 49. [121] Y. Li, J.N. Armor, Appl. Catal. B 1 (1992) L31. [122] K. Yogo, S. Tanaka, M. Ihara, T. Hishiki, E. Kikuchi, Chem. Lett. (1992) 1025. [123] Y. Nishizaka, M. Misono, Chem. Lett. (1993) 1295. [124] M. Iwamoto, N. Mizuno, J. Automobile Eng. 207 (1993) 23. [125] M. Konno, T. Chikahisa, T. Murayama, M. Iwamoto, SAE Paper 920091, Vol. 1, 1992. [126] T. Ishihara, M. Kagawa, F. Hamada, Y. Takita, Stud. Surf. Sci. Catal. 84 (1994) 1493. [127] A.E. Palomares, F. M arques, S. Valencia, A. Corma, J. Mol. Catal. A 162 (2000) 175. [128] M. Iwamoto, H. Yahiro, S. Shundo, Y. Yu-u, N. Mizuno, Appl. Catal. 69 (1991) L15. [129] M. Iwamoto, Stud. Surf. Sci. Catal. 54 (1990) 121. [130] M. Iwamoto, H. Hamada, Catal. Today 10 (1991) 57. [131] M. Iwamoto, Stud. Surf. Sci. Catal. 84 (1994) 1395. [132] M. Iwamoto, Catal. Today 29 (1996) 29. [133] A. Fritz, V. Pitchon, Appl. Catal. B 13 (1997) 1. [134] H. Akama, H. Kanesaki, S. Yamamoto, K. Matsumoto, Jidoushagijutu 54 (2000) 77. [135] M.D. Amiridis, T. Zhang, R.J. Farrauto, Appl. Catal. B 10 (1996) 203. [136] Y. Traa, B. Burger, J. Weikamp, Micropor. Mesopor. Mater. 30 (1999) 3. [137] H. Hamada, Y. Kintaichi, M. Sasaki, T. Ito, Appl. Catal. 70 (1991) L15. [138] N.W. Cant, I.O.Y. Liu, Catal. Today 63 (2000) 133. [139] Y. Li, J.N. Armor, J. Catal. 150 (1994) 388. [140] M. Shelef, C.N. Montreuil, H.W. Jen, Catal. Lett. 26 (1994) 277. [141] E. Kikuchi, K. Yogo, Catal. Today 22 (1994) 73. [142] M.H. Kim, I.-S. Nam, Y.G. Kim, Appl. Catal. B 6 (1995) 297. [143] T.E. Hoost, K.A. Laframboise, K. Otto, Appl. Catal. B 7 (1995) 79. [144] D.B. Lukyanov, G. Still, J.L. dItri, W.K. Hall, J. Catal. 153 (1995) 265. [145] A.D. Cowan, R. Dumpelmann, N.W. Cant, J. Catal. 151 (1995) 356. [146] K.A. Bethke, C. Li, M.C. Kung, H.H. Kung, Catal. Lett. 31 (1995) 287. [147] M. Iwamoto, H. Takeda, Catal. Today 27 (1996) 71. [148] M. Guyon, V.L. Chanu, P. Gilot, H. Kessler, G. Prado, Appl. Catal. B 8 (1996) 183. [149] K. Hadjiivanov, D. Klissurski, G. Ramis, G. Busca, Appl. Catal. B 7 (1996) 251.

H. Yahiro, M. Iwamoto / Applied Catalysis A: General 222 (2001) 163181 [182] P.H. Kasai, R.J. Bishop Jr., J. Am. Chem. Soc. 94 (1972) 5560. [183] P.H. Kasai, R.J. Bishop Jr., in: J.A. Rabo (Ed.), Zeolite Chemistry and Catalysis, ACS Monograph 171, Washington, DC, 1976, p. 350. [184] H. Yahiro, A. Lund, R. Aasa, N.P. Benetis, M. Shiotani, J. Phys. Chem. B 104 (2000) 7950. [185] H. Yahiro, A. Lund, N.P. Benetis, M. Shiotani, Chem. Lett. (2000) 736. [186] D. Biglino, H. Li, R. Erikson, A. Lund, H. Yahiro, M. Shiotani, Phys. Chem. Chem. Phys. 1 (1999) 2887. [187] R.M. Barrer, J. Colloid Int. Sci. 21 (1966) 415. [188] H. Yahiro, N.P. Benetis, A. Lund, M. Shiotani, Stud. Surf. Sci. Catal. 135 (2001) 348. [189] P.H. Kasai, R.M. Gaura, J. Phys. Chem. 86 (1982) 4257. [190] V.I. P arvulescu, P. Grange, B. Delmon, Catal. Today 46 (1998) 233. [191] T. Tabata, H. Ohtsuka, Catal. Lett. 48 (1997) 203. [192] A. Shichi, A. Satsuma, M. Iwase, K. Shimizu, S. Komai, T. Hattori, Appl. Catal. B 17 (1998) 107.

181

[193] T. Tabata, H. Ohtsuka, L.M.F. Sabatino, G. Bellussi, Micropor. Mesopor. Mater. 21 (1998) 517. [194] A. Shichi, K. Katagi, A. Satuma, T. Hattori, Appl. Catal. B 24 (2000) 97. [195] T. Ito, J. Fraissard, J. Chem. Phys. 76 (1982) 5225. [196] H. Yahiro, M. Shiotani, J.H. Freed, M. Lindgren, A. Lund, Stud. Surf. Sci. Catal. 94 (1995) 673. [197] S.A. Goldman, G.V. Bruno, C.F. Polnaszek, J.H. Freed, J. Chem. Phys. 56 (1972) 716. [198] M. Nagata, H. Yahiro, M. Shiotani, M. Lindgren, A. Lund, Chem. Phys. Lett. 256 (1996) 27. [199] H. Li, A. Lund, M. Lindgren, E. Sagstuen, H. Yahiro, Chem. Phys. Lett. 271 (1997) 84. [200] H. Li, H. Yahiro, M. Shiotani, A. Lund, J. Phys. Chem. B 102 (1998) 5641. [201] H. Li, H. Yahiro, K. Komaguchi, M. Shiotani, E. Sagstuen, A. Lund, Micropor. Mesopor. Mater. 30 (1999) 275. [202] H. Yahiro, M. Nagata, M. Shiotani, M. Lindgren, H. Li, A. Lund, Nukleonika 42 (1997) 557.