`
`343
`
`EFFECT OF FRAMEWORK CHARGE DENSITY ON CATALYTIC
`ACTIVITY OF COPPER LOADED MOLECULAR SIEVES OF CHABAZITE
`STRUCTURE IN NITROGEN(II) OXIDE DECOMPOSITION
`
`Jiří DĚDEČEK1,*, Alena VONDROVÁ2 and Jiří ČEJKA3
`J. Heyrovský Institute of Physical Chemistry, Academy of Sciences of the Czech Republic,
`182 23 Prague 8, Czech Republic; e-mail: 1 dedecek@jh-inst.cas.cz, 2 vondrova@jh-inst.cas.cz,
`3 cejka@jh-inst.cas.cz
`
`Received December 15, 1999
`Accepted February 23, 2000
`
`NO decomposition over Cu-loaded molecular sieves of chabazite structure with different
`framework negative charge density was investigated. Cu-ZnAlPO-34 with framework charge
`density (Al + P)/Zn = 15 exhibited high and stable catalytic activity in NO decomposition
`while Cu-chabazite with Si/Al = 2.7 was inactive. This evidences that the number of alumi-
`num atoms balancing Cu ions in cationic sites controls the catalytic activity of Cu ion. Ions
`balanced by single framework aluminum atoms exhibit high activity in NO decomposition
`while ions balanced by two framework aluminum atoms are inactive. Topology of cationic
`sites is not the limiting parameter for NO decomposition over Cu ions located in molecular
`sieves.
`Key words: NO decomposition; Cu ion; Chabazite; ZnAlPO-34; Molecular sieve; Zeolites;
`Heterogeneous catalysis.
`
`Nitrogen oxides (NOx) produced in high-temperature combustion processes
`are major air pollutants at present. Decomposition of NO into molecular ni-
`trogen and oxygen seems to be a very attractive approach to control the
`NOx pollution. Despite the substantial effort being currently devoted to de-
`velopment of a suitable catalyst exhibiting high and stable activity in NO
`decomposition, this challenge has not yet been solved. High catalytic activ-
`ity has been found only for Cu ions implanted at the cationic sites of zeo-
`lite matrices, namely with Cu-ZSM-5 (refs1–6). Nevertheless, for industrial
`application of this process, the conversion is still low; moreover, Cu-ZSM-5
`activity is sensitive to the presence of water vapor and SOx. Despite this
`fact, Cu-loaded zeolites attract high attention as the most important model
`system for the investigation of NO decomposition.
`Isolated Cu+ ions, adjacent Cu and O centers and two neighboring Cu
`ions (Cu pairs), were suggested to be active in NO decomposition7–22. By a
`
`Collect. Czech. Chem. Commun. (Vol. 65) (2000)
`
`Exhibit 2015.001
`
`
`
`344
`
`Dědeček, Vondrová, Čejka:
`
`detailed spectral analysis, we have shown19–21 that the Cu site active in NO
`decomposition possesses a low positive charge on the divalent Cu cation, is
`easily reduced, exhibits open, close to planar coordination sphere and is ad-
`jacent to a single Al framework atom. Recently, we have suggested one type
`of six-member ring, located in the straight channel of ZSM-5, to accommo-
`date Cu ions responsible for NO decomposition over Cu-loaded ZSM-5
`(ref.22). Nevertheless, the framework topology and Al distribution in ZSM-5
`do not allow to distinguish between the effect of local framework topology
`of the cationic site and its local framework charge density (the number of
`framework aluminum atoms forming the cationic site) on the catalytic ac-
`tivity of the Cu ion. Also, investigation of new systems active in NO de-
`composition, based on Cu-containing MeAlPO-5 and -11 materials23,24, did
`not lead to elucidation of this problem. Detailed investigation of these cata-
`lysts is complicated by a lack of detailed knowledge on cationic sites in
`these molecular sieves and their low thermal stability. Nevertheless, high
`catalytic activity of Cu-MeAlPO matrices indicated that high and stable cat-
`alytic activity in NO decomposition is not restricted only to aluminosilicate
`matrices and one type of framework topology.
`This contribution presents identification of parameters of cationic sites,
`necessary for catalytic activity of Cu ions in NO decomposition. Catalytic
`activity of Cu-loaded molecular sieves of chabazite structure with different
`framework charge density is reported. Only Cu ions located in matrix with
`a low density of the framework negative charge (Zn(Al)PO-34) exhibited an
`activity in NO decomposition while Cu-chabazites (Si/Al = 2.7) were inac-
`tive. Thus, the vicinity of only one framework negative charge is substantial
`for the activity of Cu ion. On the other hand, local topology of the Cu
`cationic site is not a limiting factor for the Cu ion activity, as follows from
`different framework topologies of chabazite and ZSM-5.
`
`EXPERIMENTAL
`
`Molecular Sieve Preparation
`
`Synthesis of chabazite was performed using the procedure of Gaffney25. Zeolite Y (Si/Al =
`2.7) in the ammonium form was used as a source material after calcination at 550 °C and
`was mixed with a solution of potassium hydroxide. The batch composition was 0.17 Na2O :
`2.0 K2O : Al2O3 : 5.4 SiO2 : 224 H2O. The synthesis took place in a polypropylene bottle
`with a screw-top lid at temperature of 368 K for 96 h without agitation.
`Synthesis of ZnAlPO-34 was performed in the following way. First of phosphoric acid
`(20.17 g) and pseudoboehmite (11.87 g) (Catapal B, Vista) were added to water (50.5 g) and
`stirred for 2 h. Then triethylamine (32.4 g) used as organic template was added and stirred
`for another 2 h. Finally, zinc acetate (1.26 g) in water (7.0 g) was added and the final mix-
`
`Collect. Czech. Chem. Commun. (Vol. 65) (2000)
`
`Exhibit 2015.002
`
`
`
`Molecular Sieves of Chabazite Structure
`
`345
`
`ture was thoroughly stirred for another 2 h. The synthesis was carried out in Teflon-lined
`stainless steel autoclaves at 155 °C for 22 h under agitation.
`In both syntheses, the products were recovered by filtration, washed repeatedly with
`deionized water and dried at ambient temperature. The crystallinity and phase purity were
`determined using X-ray powder diffraction. The diffraction pattern of chabazite is depicted
`in Fig. 1a, for ZnAlPO-34 in Fig. 1b.
`Isostructural chabazite and ZnAlPO-34 (structural code CHAB) were equilibrated
`four times with 0.5 M NaCl (20 ml of solution per 1 g of azeolite) for 12 h. After ion ex-
`change, Na-CHAB sieves were washed with distilled water and dried at room tempera-
`ture. Cu2+-CHAB sieves with Cu concentration varying from 0.4 to 12.5 wt.% were
`prepared by ion exchange of Na-CHAB sieves with aqueous solution of Cu acetate. pH of
`Cu acetate–CHAB solutions varied during the ion exchange procedure from 5.1 to 5.6. Sam-
`ples were carefully washed with distilled water, dried at ambient temperature and grained.
`Detailed conditions of the sample preparation and chemical composition of Cu-CHAB sam-
`ples are given in Table I.
`
`Characterization
`
`Powder X-ray diffraction patterns of synthesized molecular sieves were recorded on a
`Siemens D5005 diffractometer in the Bragg–Brentano geometry arrangement, using CuKα ra-
`diation with a graphite monochromator and scintillation counter. To verify that the treat-
`ment of as-synthesized MeAlPO material did not change the crystallinity of the molecular
`sieves, XRD patterns were recorded after synthesis and drying of the sample at room temper-
`ature, after calcination, and after ion exchange.
`
`Catalytic Activity
`
`Catalytic activity of Cu-exchanged molecular sieves was tested in a down-flow glass
`microreactor with inner diameter 5 mm, introducing 4 000 ppm of NO in He, total feed of
`
`3 000
`
`Intensity
`
`2 000
`
`1 000
`
`0
`
`0
`
`a
`
`b
`
`10
`
`20
`
`30
`
`40
`
`2Θ, °
`
`50
`
`FIG. 1
`XRD patterns for chabazite (a), ZnAlPO-34 (b)
`
`Collect. Czech. Chem. Commun. (Vol. 65) (2000)
`
`Exhibit 2015.003
`
`
`
`346
`
`Dědeček, Vondrová, Čejka:
`
`100 ml/min, a catalyst weight of 300 mg and in the temperature range 620–750 K. The cata-
`lysts were activated in a He stream (99.996%) with a temperature increase of 5 K/min up to
`680 or 720 K, held for 1 h, and cooled down to the reaction temperature. To achieve the
`temperature of the next measurement, the catalyst was heated in a NO/He stream with a
`temperature increase of 5 K/min. The temperature step between two temperatures of mea-
`surements was 30 K. The NO conversion reached in 30–60 min after stabilization of the
`working temperature was taken as a characteristic value. NO and NO2 were analyzed with an
`accuracy of 0.5% at the inlet and outlet of the reactor with a chemiluminescence analyser
`Vamet 138 (Czech Republic). No NO2 was detected in the products (detection limit 5 ppm).
`Only traces of N2O were observed by mass spectrometry (Hewlett–Packard, 5971A).
`
`RESULTS AND DISCUSSION
`
`Cu-ZnAlPO-34 molecular sieves activated at 680 K exhibited catalytic activ-
`ity in NO decomposition in the temperature range 620–700 K while
`Cu-ZnAlPOs-34 activated at 720 K exhibited significantly lower activity
`(maximum 50% of the activity of samples activated at 680 K, not shown in
`figures) and Cu-chabazites were inactive. Figure 2a depicts the dependence
`of NO concentration at the reactor outlet on the reaction time-on-stream.
`At first a sharp decrease in NO concentration was observed, followed at
`lower reaction temperature by leveling off to a constant conversion,
`reached within 30 min. This dependence of the NO conversion on
`time-on-stream evidences that the initial behavior of the NO conversion
`over Cu-MeAlPO-34 is similar to the well-known transient behavior of NO
`
`TABLE I
`Chemical composition and conditions of preparation of Cu-CHAB molecular sieves
`
`Molecular
`sieve
`
`Cu/Al
`or
`Cu/Zn
`
`Si/Al
`or
`(Al+P)/Zn
`
`Na/Al
`or
`Na/Zn
`
`Cu concentra-
`tion in solution
`(M)
`
`Solution/
`zeolite, ml/g
`
`Time of
`exchange, h
`
`Cu-CHAB
`
`Cu-CHAB
`
`Cu-CHAB
`
`Cu-CHAB
`
`Cu-ZnAlPO-34
`
`Cu-ZnAlPO-34
`
`0.08
`
`0.15
`
`0.22
`
`0.32
`
`0.13
`
`0.97
`
`Cu-ZnAlPO-34
`
`4.5
`
`2.7
`
`2.7
`
`2.7
`
`2.7
`
`15
`
`15
`
`15
`
`0.84
`
`0.68
`
`0.51
`
`0.26
`
`0.82
`
`<0.05
`
`<0.05
`
`0.005
`
`0.01
`
`0.01
`
`0.01
`
`0.01
`
`0.10
`
`0.10
`
`65
`
`60
`
`110
`
`1
`
`3
`
`12
`
`110 + 110
`
`3 + 15
`
`50
`
`90
`
`16
`
`16
`
`100 + 100
`
`12 + 24
`
`Collect. Czech. Chem. Commun. (Vol. 65) (2000)
`
`Exhibit 2015.004
`
`
`
`Molecular Sieves of Chabazite Structure
`
`347
`
`decomposition observed on reduced Cu-ZSM-5 or Cu-MeAlPO-5 and -11
`molecular sieves23,24,26. Thus, the nature of reaction centers in these matri-
`ces is similar. At reaction temperatures lower than 670 K, the catalyst activ-
`ity was stable for hours, as it is documented in Fig. 2b. At reaction
`temperatures higher than 680 K, the catalyst activity significantly decreased
`with time and samples became inactive, as it is shown in Fig. 2a. This deac-
`tivation at higher temperatures was also observed for other Cu-alumino-
`phosphate catalysts (Cu-MeAlPO-5 and -11) and should be associated with
`lower thermal stability of MeAlPO molecular sieves23,24. This suggestion is
`supported by observed lower catalytic activity of samples activated at 720 K
`because the dispersion of Cu ions in molecular sieves is not usually affected
`by the treatment at this temperature18–20.
`In Fig. 3, the temperature dependence of stabilized conversion of NO
`over Cu-ZnAlPO-34 is depicted. The activity increased with temperature,
`but this increase was cut after reaching the temperature of catalyst deactiva-
`tion at 680 K. The maximum values of the stable conversion of NO were
`taken as characteristic values of NO conversion over these catalysts and are
`summarized, together with corresponding turn-over frequency (TOF), in Table II.
`T-O-S, min
`80
`100
`
`4 000
`
`NOppm
`
`3 000
`
`0
`
`20
`
`40
`
`60
`
`2 000
`
`1 000
`4 000
`
`3 000
`
`2 000
`
`FIG. 2
`Dependence of NO concentration at
`the reactor outlet on the reaction
`time-on-stream (T-O-S) (N Cu--chaba-
`zite Cu/Al 0.32, Cu-ZnAlPO-34, Cu/Zn 4.5
`650 K, ❒
`700 K)
`
`Collect. Czech. Chem. Commun. (Vol. 65) (2000)
`
`a
`
`b
`
`0
`
`1
`
`2
`
`3
`
`4
`T-O-S, h
`
`5
`
`Exhibit 2015.005
`
`❍
`
`
`348
`
`Dědeček, Vondrová, Čejka:
`
`Cationic sites in chabazite are well known25–32. Cations in site I (for nota-
`tion see refs26–32) are located in the chabazite cavity and coordinated to
`three framework oxygens of the regular six-member ring above the
`six-member ring plane. Site II of the chabazite cavity is occupied only in
`partially hydrated chabazites. Site III is inside the hexagonal prism. Ions at
`this site can hardly play the role of catalytic center as their coordination
`sphere is fully occupied. Site IV corresponds to the cations located in the
`eight-member ring connecting chabazite cavities. Cu+ ions in dehydrated
`form occupy only sites I and IV (ref.33) and only these sites are shown in
`
`TABLE II
`Catalytic activity of Cu-CHAB molecular sieves
`
`Catalyst
`
`Cu, wt%
`
`Cu/Al or Cu/Zn
`
`Conversion, %
`
`TOF ⋅ 10–4, s–1
`
`Cu-ZnAlPO-34
`
`Cu-ZnAlPO-34
`
`Cu-ZnAlPO-34
`
`Cu-CHAB
`
`Cu-CHAB
`
`Cu-CHAB
`
`Cu-CHAB
`
`12.5
`
`2.4
`
`0.4
`
`7.6
`
`5.2
`
`3.6
`
`1.9
`
`4.5
`
`0.97
`
`0.13
`
`0.32
`
`0.22
`
`0.15
`
`0.08
`
`14.2
`
`6.5
`
`2.4
`
`0
`
`0
`
`0
`
`0
`
`0.9
`
`2.0
`
`4.5
`
`0
`
`0
`
`0
`
`0
`
`630
`
`660
`
`690
`
`720
`
`750
`
`T, K
`
`780
`
`FIG. 3
`Temperature dependence of NO de-
`composition over Cu-ZnAlPO-34 cata-
`lyst (Cu/Zn 4.5, 0.30 g of catalyst)
`
`Collect. Czech. Chem. Commun. (Vol. 65) (2000)
`
`16
`
`NO,%
`
`12
`
`8
`
`4
`
`0
`
`Exhibit 2015.006
`
`
`
`Molecular Sieves of Chabazite Structure
`
`349
`
`Fig. 4. MeAlPO-34 is isostructural with chabazite structure34 and the local
`topology of the cationic sites of Cu-ZnAlPO-34 structure is identical. Thus,
`Cu species active in NO decomposition in Zn-AlPO-34 are coordinated to
`the regular six-member or eight-member rings. However, these types of
`rings are not present in the ZSM-5 or MeAlPO-5 and -11 structures, in
`which ions are assumed to be coordinated to various deformed six-member
`rings22,24,35. This indicates that the local topology of cationic site is not a
`crucial parameter controlling the catalytic activity of Cu ions in NO decom-
`position.
`The only possible difference in Cu siting between chabazite and
`ZnAlPO-34 has to be associated with the change in the framework negative
`charge density, reflected in the Si/Al ratio of chabazite and the (Al + P)/Zn
`ratio of ZnAlPO-34. In the case of chabazite with Si/Al ≈ 2.7, at least 93% of
`cationic sites contain two aluminum atoms, as follows from simple statistic
`calculation. This assumption is supported by full ion exchange of divalent
`in chabazite (Me2+/Al
`ions
`ratio 0.5) observed for Co-, Mn- and
`Ca-chabazites26–32. On the other hand, in the case of ZnAlPO-34 with a sig-
`nificantly lower density of framework negative charge ((Al + P)/Zn = 15),
`the probability of the presence of two framework aluminum atoms at
`cationic sites formed by six- or eight-member rings is significantly lower.
`Thus, we can suppose that sites containing only one Al atom are negligible
`in chabazite and, in contrast, they prevail in ZnAlPO-34. The low density of
`the framework negative charge can also affect the relative number of
`cationic sites formed by six- or eight-member rings. In the chabazite frame-
`work, all rings contain aluminum atoms forming cationic sites and the ra-
`
`FIG. 4
`Cu sites occupied in ion-exchanged dehydrated chabazite and their notation according to
`refs26–32
`
`Collect. Czech. Chem. Commun. (Vol. 65) (2000)
`
`Exhibit 2015.007
`
`
`
`350
`
`Dědeček, Vondrová, Čejka:
`
`tio of sites formed by six- and eight-member rings is given by their
`frequency in the framework. In ZnAlPO-34 with low density of negative
`charge, rings without Al atom can be present. In this case, the ratio of sites
`formed by six- and eight-member rings is given by the Al distribution in the
`framework.
`Because all Cu-chabazites contain Cu ions located both in six- and
`eight-member rings, as reported in our previous study33, the inactivity of
`Cu-chabazites in NO decomposition is caused by the presence of two frame-
`work aluminum atoms in every cationic site of the matrix. The difference
`in the turn-over frequency values in NO decomposition by Cu-ZnAlPO-34
`samples with different Cu loading can be explained by the difference in the
`occupation of sites with different catalytic activity or by formation of
`CuxOy species in highly loaded Cu-ZnAlPO-34. The presence of these spe-
`cies is reflected in high Cu exchange level (Cu/Al = 1.4); these species do
`not exhibit catalytic activity36.
`
`CONCLUSIONS
`
`Cu ions exchanged to ZnAlPO-34 molecular sieves exhibit high and stable
`catalytic activity in NO decomposition, while Cu ions exchanged to chaba-
`zite are inactive.
`The density of the framework negative charge has been found to be the
`crucial parameter controlling the activity in NO decomposition of Cu ions
`located in molecular sieves. Cu ions balanced by two framework negative
`charges (two framework aluminum atoms) are inactive; however, Cu ions
`balanced by a single framework negative charge exhibit catalytic activity.
`Catalytic activity of Cu ions is not restricted to only one type of local to-
`pology of the cationic site. Both Cu ions located in regular six-member or
`eight-member rings (CHAB structure) and in elongated or deformed
`six-member rings (ZSM-5, MeAlPO-5 and -11) exhibit catalytic activity in
`NO decomposition.
`
`This work was supported by VW Stiftung project No. I/72937. The authors thank Dr T. Grygar for
`chemical analysis of zeolites.
`
`REFERENCES
`
`1. Iwamoto M., Yokoo S., Sakai K., Kagawa S.: J. Chem. Soc., Faraday Trans. 1 1981, 77, 1692.
`2. Iwamoto M., Yahiro H., Tanda K., Mizuno N., Mine Y.: J. Phys. Chem. 1991, 95, 3727.
`3. Iwamoto M.: Stud. Surf. Sci. Catal. 1994, 84, 1395.
`
`Collect. Czech. Chem. Commun. (Vol. 65) (2000)
`
`Exhibit 2015.008
`
`
`
`Molecular Sieves of Chabazite Structure
`
`351
`
`4. Li Y., Hall W. K.: J. Catal. 1991, 129, 202.
`5.IwamotoM.,YahiroH.:Catal.Today1994,22,5.
`6.CentiG.,PeranthonerS.:Appl.Catal.,A1995,132,179.
`7. Liu D.-J., Robota H. J.: Catal. Lett. 1993, 21, 291.
`8.SchneiderW.F.,HaasK.C.,RamprasadR.,AdamsJ.B.:J.Phys.Chem.1996,100,6032.
`9.HaasK.C.,SchneiderW.F.:J.Phys.Chem.1996,100,9292.
`J.Phys.Chem.B1997,101,1940.
`10.RamprasadR.,SchneiderW.F.,HaasK.C.,AdamsJ.B.:
`11.YokomichiY.,YamabeT.,OtsukaH.,KakumotoT.:J.Phys.Chem.1996,100,14424.
`12.BlintR.J.:J.Phys.Chem.1996,100,19518.
`13.TroutB.L.,ChakrabortyA.K.,BellA.T.:J.Phys.Chem.1996,100,4173.
`14.TroutB.L.,ChakrabortyA.K.,BellA.T.:J.Phys.Chem.1996,100,17582.
`15.SpotoG.,ZecchinaA.,BordigaS.,RicchiardiG.,MartraG.,LeofantiG.,PetriniG.:Appl.
`Catal.,B1994,3,151.
`16. Anpo M., Matsuoka M., Shioya Y., Yamashita H., Giamello E., Morterra C., Che M.,
`Patterson H. H., Weber S., Oullette S., Fox M. A.: J. Phys. Chem. 1994, 98, 5744.
`17. Teraishi K., Ishida M., Irisawa J., Kume M., Takahashi Y., Nakano T., Nakamura H.,
`MiyamotoA.:J.Phys.Chem.B1997,101,8079.
`18. Wichterlová B., Dědeček J., Vondrová A.: J. Phys. Chem. 1995, 99, 1065.
`19. Dědeček J., Sobalík Z., Tvarůžková Z., Kaucký D., Wichterlová B.: J. Phys. Chem. 1995,
`99, 16327.
`20.WichterlováB.,DědečekJ.,SobalíkZ.,VondrováA.,KlierK.:J.Catal.1997,169,194.
`21.DědečekJ.,WichterlováB.:J.Phys.Chem.B1997,101,10233.
`22. Dědeček J., Wichterlová B.: Chem. Phys. Phys. Chem. 1999, 1, 629.
`23.DědečekJ.,ČejkaJ.,WichterlováB.:Appl.Catal.,B1998,15,233.
`24.DědečekJ.,VondrováA.,ČejkaJ.:Collect.Czech.Chem.Commun.1998,63,1755.
`25. Gaffney T. R.: Micropor. Mater. 1998, 22, 551.
`26. Dědeček J., Wichterlová B.: Unpublished results.
`27.PluthJ.J.,SmithJ.V.,MortierW.J.:Mater.Res.Bull.1977,12,1001.
`28. Calligaris M., Nardin G., Randaccio L.: Zeolites 1984, 4, 251.
`29.MortierW.J.,PluthJ.J.,SmithJ.V.:Mater.Res.Bull.1977,12,103.
`30. Calligaris M., Mezzetti A., Nardin G., Randaccio L.: Zeolites 1986, 6, 137.
`31. Calligaris M., Mezzetti A., Nardin G., Randaccio L.: Zeolites 1984, 4, 323.
`32. Calligaris M., Mezzetti A., Nardin G., Randaccio L.: Zeolites 1985, 5, 317.
`33.DědečekJ.,WichterlováB.,KubátP.:Micropor.Mesopor.Mater.1999,32,63.
`34. Lok B. M., Messina C. A., Patton R. L., Gajek R. T., Cannan T. R., Flanigen E. M.: J. Am.
`Chem. Soc. 1984, 106, 6092.
`35. Dědeček J., Kaucký D., Wichterlová B.: Micropor. Mesopor. Mater., in press.
`36. Dědeček J., Bortnovski O., Vondrová A., Wichterlová B.: Appl. Catal., B, submitted.
`
`Collect. Czech. Chem. Commun. (Vol. 65) (2000)
`
`Exhibit 2015.009