`ARTICLE NO. CA971681
`
`Copper !on-Exchanged SAP0-34 as a Thermostable Catalyst
`for Selective Reduction of NO with C3H5
`
`Tatsumi Ishihara, 1 Masaru Kagawa, Fumiaki Hadama, and Yusaku Takita
`Department of Applied Chemistry. Faculty of Engineering, Oita University, Dannoharu 700, Oita, 870-11, ]a pan
`
`Received September 11, 1995; revised December 19, 1996; accepted February 28, 1997
`
`Selective reduction of NO with C3H6 in the presence of oxygen
`was studied over Cu ion-exchanged SAPO-n (n= 5, 11, 34), and
`Cu ion-exchanged Zeolites {3, USY, and ZSM-5. All the Cu ion(cid:173)
`exchanged catalysts exhibited high activity for NO reduction with
`C3Hs in a large excess of 02; however, the temperature for maxi(cid:173)
`mum NO conversion depended on the kind of molecular sieve. Al(cid:173)
`though the maximum conversion of NO was attained at a slightly
`higher temperature in comparison with Cu-ZSM-5, Cu-SAP0-34
`exhibited the highest activity for NO reduction among the catalysts
`studied under the conditions examined. Furthermore, high NO con(cid:173)
`version was attained over a wide temperature range, from 623 to
`873 K. SAP0-34 has high thermal stability. High activity for NO
`selective reduction on Cu-SAP0-34 was sustained for more than
`60 hat 673 Kin an atmosphere containing 15 vol% H20. After ther(cid:173)
`mal treatment at 1073 Kin humidified atmosphere, the decrease in
`activity for NO reduction was also small. Redox behavior of Cu ions
`in SAP0-34 between monovalent and divalent states occurs during
`the selective reduction of NO, and the reaction seems to proceed via
`formation of adsorbed nitrate species followed by the formation of
`organic nitro compounds. © 1997 Academic Press
`
`1. INTRODUCTION
`
`Aluminophosphates (AlP04-n) are a new family of
`molecular sieves that possess the characteristics of a molec(cid:173)
`ularsized pore structure and solid acidity. In particular,
`AlP04-n exhibits extremely high thermal stability as com(cid:173)
`pared with synthetic zeolites. It was reported that the crystal
`structure of AlP04-5 is retained up to 1473 K (1). Simi(cid:173)
`larly, the crystal structure of SAP0-34 is· sustained up to
`1273 K, even in the presence of humidity (2). Silicoalumi(cid:173)
`nophosphates (SAPO-n) exhibit cation-exchange proper(cid:173)
`ties as a result of the isomorphous substitution ofP in AlP04
`by Si. Therefore, AlP04-n and SAP O-n have great possibil(cid:173)
`ities as thermostable catalysts; however, the advantages in
`thermal stability of SAP O-n or AlP04-n have not yet been
`thoroughly exploited in conventional studies.
`Selective reduction of NOx under an oxidizing atmo(cid:173)
`sphere with hydrocarbons has attracted attention as a new
`
`1 To whom correspondence should be sent. Fax: +81-975-54-7979.
`
`process for the catalytic removal of NOxin the exhaust gas
`of diesel or lean-burn engines (3-6). It is reported that Cu
`ion-exchanged ZSM-5 is highly active for the selective re(cid:173)
`duction of NO (7); however, the high activity ofCu-ZSM-5
`to NO reduction decreases gradually with time-on-stream
`at temperatures higher than 973 K. One reason for the de(cid:173)
`crease in activity seems to be the poor thermal stability of
`ZSM-5 (8-10). In addition, dealumination easily proceeds
`on ZSM-5 under a humidified atmosphere, resulting in ac(cid:173)
`celerating decreases in the activity for NOxreduction. We
`have investigated Cu ion-exchanged SAP O-n for the selec(cid:173)
`tive reduction of NO and have found that Cu-SAP0-34
`exhibits high activity for NO reduction with C3H6 (11).
`The state of ion-exchanged Cu in SAP0-34 and its cata(cid:173)
`lytic properties are also being studied by other groups
`(12, 13). In this paper, the activities of Cu-SAPO-n (n= 5,
`11, and 34) for NOx reduction with C3H6 in the presence
`of 0 2 are further studied as a thermostable NOx removal
`catalyst. Moreover, the oxidation states of the copper ions
`in SAP0-34 on treatment with C3H5, NO, and 02 in the
`working state have been studied to explore the mechanism
`of the selective reduction of NO with C3H5.
`
`2. EXPERIMENTAL
`
`Catalyst Preparation
`
`SAP0-5, 11, and 34, (Si, Al, and P contents: 1.77,
`12.09, and 10.03 mmol g-1, respectively) and zeolite
`fJ (Si02/Al203 = 26) were synthesized according to U.S.
`patents (14, 15). To prepare SAPO-n, colloidal Si02,
`Al compound [pseudo-boehmite phase (Cataloid AP,
`Catalysis & Chemical Ind.) for SAP0-5 and SAP0-11,
`Al[OCH(CH3)2b for SAP0-34]. phosphoric acid (85%),
`and template amine (triethylamine for SAP0-5, di-n(cid:173)
`propylamine for SAP0-11, and 10% tetraethylammonium
`hydroxide for SAP0-34) were mixed at room temper(cid:173)
`ature for a few hours. The precursor of SAPO-n ob(cid:173)
`tained in this way was heated (453 K, 12 h for SAP0-5
`and SAP0-11, 483 K, 24 h for SAP0-34) in autoclaves
`(Taiatsu Glass TAF-150) in which all parts were coated
`with Teflon. The synthesized SAPO-nand zeolite fJ, as well
`
`93
`
`0021-9517/97$25.00
`Copyright© 1997 by Academic Press
`All rights of reproduction in any form reserved.
`
`Umicore AG & Co. KG
`Exhibit 1112
`Page 1 of 10
`
`
`
`94
`
`ISHIHARA ET AL.
`
`as the commercial ZSM-5 (Mobil, Si02/ Alz03 = 30) and
`USY (Catalysis & Chemical Ind., Si02/Alz03 = 17), were
`ion-exchanged with Cu2+ in a 0.01 M Cu2+ acetate aque(cid:173)
`ous solution. At the final stage of Cu ion exchange, am(cid:173)
`monia water was added to adjust the pH to 7.5 to control
`the amount of Cu ion-exchanged. Exchanged amounts of
`Cu2+ for each type of SAPO-n, {3, USY, and ZSM-5 were
`estimated to be about 3 wt% from ICP analysis. In the case
`of SAPO-n, this amount of Cu corresponds to ca. 75% of
`the formula ion-exchange capacity, which is estimated by
`assuming that all the Si forms ion-exchange sites. Before
`measurement of the activity for NO reduction with C3H5,
`the catalysts were calcined at 773 K for 4 h in a He stream.
`
`AQA-100R) was connected to the IR cell for qualitative
`analysis of the gas phase during the adsorption treatment.
`The crystal structure of Cu ion-exchanged SAP0-34 was
`analyzed by X-ray diffraction and NMR measurements.
`X-ray diffraction analysis (Rigaku, 2013CN) was performed
`with the Cu Ka line and NMR spectra for 27 AI, 31P, and
`29Si were recorded with a Bruker APR-300 spectrometer
`operating at a field of 7 T using the magic-angle-spinning
`(MAS) technique. Spinning speeds of 5.5 kHz were used
`and the chemical shifts of AI. P, and Si were referred to the
`external standard of 1.5 M Al(H20)~+ in Al(N03)3 aque(cid:173)
`ous solution, H3P04 (85%), and 4.5 Mtetramethylsilane in
`benzene solution, respectively.
`
`NO Reduction with C3H6
`
`3. RESULTS AND DISCUSSION
`
`The catalytic activities for NO reduction were measured
`with a fixed-bed microflow reactor. A gas mixture consist(cid:173)
`ing ofN0(5000 or 1000 ppm), C3H6 (1000 ppm), 02 (5%),
`and He (remainder) was fed to the catalyst bed at VWF=
`0.3 g-cat · s cm-3 (Space velocity= ca. 8500 h-1), where W
`and F stand for catalyst weight and flow rate, respectively.
`Conversion to N2 was estimated from analysis of the N2
`concentration in the flue gases by gas chromatography. The
`heat resistances of SAPO-nand ZSM-5 were determined
`by calcining at a prescribed temperature for 2 h in air con(cid:173)
`taining 3 vol% H20.
`
`Characterization of Cu Ion-Exchanged SAP0-34
`
`The valence of the Cu ions in SAP0-34 was determined
`by ESR (JEOL JEX-FE1X) and XPS (Shimadzu ESCA-
`850). Before the ESR measurement, the catalyst was evac(cid:173)
`uated at 773 K for 5 h, calcined in 02 (100 Torr) at 673 K
`for 4 h, and then evacuated at 573 K for 4 h. The ESR
`measurement was performed at room temperature and
`diphenylpicrylhydrazyl (DPPH) was used as an external
`standard for the calibration of the g value. XPS measure(cid:173)
`ment of the ion-exchanged Cu in SAP0-34 was performed
`after a calcination in oxygen (1 atm) at 623 K for 3 h in
`the pretreatment chamber. Magnesium Ka (1254.6 e V) was
`used as the X -ray source for measurement of the XPS of Cu
`2[13!2 and the Auger line of Cu LMM. The binding energy
`of Cu 2[1312 and the position of the LMM Auger peak were
`corrected by taking the binding energy of deposited Au 4fm
`level as 84 e V
`Adsorption states of NO or C3H5 on Cu-SAP0-34 were
`investigated by infrared absorption spectroscopy (Hitachi
`270 spectrometer). A sample disk (ca. 0.03 g) was heated at
`773 K for 5 h in vacuo to remove water and other adsorbed
`gases. After calcination in 02 (100 Torr) at 673 K for 4 h, the
`sample disk was exposed to the adsorbed gas (100 Torr) in
`an IR cell with KBr single-crystal windows. The background
`spectrum was subtracted from the IR spectra measured af(cid:173)
`ter adsorption treatment. A mass spectrometer (Anelva,
`
`Selective Reduction of NO with C#6 on Metal
`Ion-Exchanged SAPO-n
`
`Figure 1 shows the catalytic activity of Cu ion-exchanged
`SAPO-n, {3, USY, and ZSM-5 as a function of reaction tem(cid:173)
`perature. Since it is reported that Cu-ZSM-5 is active for
`direct decomposition of NO (2NO = N2 + 02), an excess
`amount of NO in comparison with that of C3H5 was fed in
`this experiment. Note, therefore, that the reaction condi(cid:173)
`tions in Fig. 1 deviate strongly from the actual flue gases
`from vehicles. As reported, Cu-ZSM-5 exhibits high activ(cid:173)
`ity for NO reduction at 573 K and the activity decreases
`with increasing temperature (7). Since the Si02/ Al203 ra(cid:173)
`tio of the examined ZSM-5 is 30, which is higher than that
`of ZSM-5 examined by Hosose eta!. (7), the NO reduction
`activity of Cu-ZSM-5 examined in this study was slightly
`lower than that of Cu-ZSM-5 previously reported (7), albeit
`
`Q
`
`* i
`...... .s
`~ .... Q
`= Q
`.....
`~ ..
`= Q
`
`Q
`
`so
`
`40
`
`30
`
`20
`
`'10
`
`t:;. SAP0-34
`D SAP0-11
`0 ZSM-5
`e SAP0-5
`•
`fl
`.l USY
`
`' '' .. o
`
`QL-~~~-L~~~~~~L-~~~
`
`400
`
`500
`
`800
`700
`600
`Temperature IK
`
`900
`
`1 000
`
`FIG. 1. Temperature dependence of catalytic activity of molecular
`sieves ion-exchanged with Cu for NO reduction with C3Hs (PNo =
`5000 ppm, Pc3H 6 = 1000 ppm, Po2 = 5%, vv.tF= 0.3 g ·S cm-3).
`
`Umicore AG & Co. KG
`Exhibit 1112
`Page 2 of 10
`
`
`
`Cu-SAP0-34 FOR THE REDUCTION OF NO WITH C3Hs
`
`95
`
`with different reaction conditions. This may result from the
`different acidity of ZSM-5 resulting from the difference in
`the Si0z/Alz03 ratio. Zeolite f3 and USY ion-exchanged
`by Cu also exhibit a high activity to NO reduction; how(cid:173)
`ever, a temperature higher than 673 K was required to
`attain the high conversion to Nz. Although the tempera(cid:173)
`ture at which each Cu-SAPO-n attains maximum conver(cid:173)
`sion to Nz is higher than that of Cu-ZSM-5 by about 50 K,
`the activity of Cu-SAPO-n for NO reduction is compara(cid:173)
`ble to that of Cu-ZSM-5, except for SAPO-S. In particu(cid:173)
`lar, Cu-SAP0-34 exhibits a higher activity to NO reduc(cid:173)
`tion than Cu-ZSM-5 over the entire temperature range,
`although the space velocity of the reactant is low. Further(cid:173)
`more, the high activity was sustained up to 873 K. There(cid:173)
`fore, conversion to Nz is twice higher on Cu-SAP0-34 than
`on Cu-ZSM-5 at 873 K. Since Cu-SAP0-34 exhibited the
`highest activity to NO reduction with C3H6 among the cata(cid:173)
`lysts examined, catalytic activity for NO reduction and ther(cid:173)
`mal stability were investigated on SAP0-34 in this study.
`It has been reported that the activity ofZSM-5 is strongly
`affected by the ion-exchanged metal cation (16). Figure 2
`shows the effects of the metal cation on the NO reduc(cid:173)
`tion activity of SAP0-34. Reduction of NO with C3H6 pro(cid:173)
`ceeded on all metal cation-exchanged SAP0-34 and almost
`the same maximum conversion of NO to Nz was exhib(cid:173)
`ited; however, the temperature at maximum NO conver(cid:173)
`sion occurred depended strongly on the metal cations ion(cid:173)
`exchanged. Fe-SAP0-34 exhibited high activity in a low
`temperature range from 4 73 to 873 K. On the other hand,
`high activity was attained at temperatures higher than 873 K
`on Ag-SAP0-34. Among the metal cations examined in
`this study, copper is the most suitable as an ion-exchanged
`
`50
`
`A Cu
`0 Fe
`D Co
`
`Q
`
`~ 40
`z
`.... . s 30
`~ ... Q
`= 20
`.....
`~ ,..
`= 10
`Q u
`
`Q
`
`Temperature I K
`
`FIG. 2. Effects of different metal cations ion-exchanged in SAP0-34
`on the catalytic activity for NO reduction with C3H6. The loading of
`metal is 3.0 wt% (PNo = 5000 ppm, Pc3H 6 = 1000 ppm, Po2 = 5%, WI F =
`0.3g·scm-3).
`
`100
`
`Q
`
`0 60
`
`* i 80
`.... . s
`z ... Q
`= c 40
`.....
`~ Q u 20
`
`o~~~~QL~~~~~
`400
`500
`600
`700
`800
`900
`1 000
`Temperature I K
`
`FIG. 3. Effects of different hydrocarbons as reductant on the activ(cid:173)
`ity of Cu-SAP0-34 for NO reduction (PNo = 1000 ppm, ?hydrocarbon=
`1000 ppm, Po2 =5%, IMF=0.3 g·s cm-3).
`
`cation for SAP0-34, considering the activity and tempera(cid:173)
`ture range of high NO conversion.
`The activity of Cu-SAP0-34 for NO reduction is strongly
`dependent on the kind of hydrocarbon used as a reductant
`(Fig. 3). Although high activity for NO reduction was ob(cid:173)
`tained over a wide temperature range by using C3H6 for
`reductant, the reaction temperature at high NO conversion
`was shifted to a higher temperature range and became nar(cid:173)
`row by using C3Hs. On the other hand, selective reduction
`of NO to N2 hardly proceeded when using CH4 for reduc(cid:173)
`tant. The different influences of hydrocarbons as reductants
`on the activity of Cu-SAP0-34 for NO reduction seem to
`result from the differences in combustibility, since Nz began
`to form on commencement of the oxidation of the hydro(cid:173)
`carbons. It is expected that high conversion in NO reduction
`can be attained on Cu-SAP0-34 by using for reductant var(cid:173)
`ious kinds of hydrocarbons, except hydrocarbons with low
`combustibility such as CH4.
`Figure 4 shows NO conversion to Nz as a function of
`the amount of Cu ion-exchanged. Since the number of ion(cid:173)
`exchange sites is not completely correlated with the amount
`of Si in the case of SAP O-n, the amount of Cu is expressed
`on the basis of weight percent. It is also noted that 3.5 wt%
`Cu corresponds to almost 100% ion-exchange level.
`Clearly, conversion of NO to Nz increases with increas(cid:173)
`ing amount of Cu ion-exchanged and attains a maximum
`around 4 wt%. Although maximum NO conversion was
`obtained at 4.5 wt%, the activity gradually decreased and
`the color of the catalyst changed from clear blue to black
`after reaction. This suggests that some copper ions are re(cid:173)
`duced and aggregate during reaction. This may result from
`the fact that the amount of Cu ion is in excess of the ion- ·
`exchange sites, because Si in the framework of SAP0-34
`
`Umicore AG & Co. KG
`Exhibit 1112
`Page 3 of 10
`
`
`
`96
`
`ISHIHARA ET AL.
`
`thermal stability of SAP0-34 is expected to decrease in the
`presence of H20; however, the extent ofthe decrease in the
`activity of Cu-SAP0-34 for NO reduction was relatively
`small, even after calcination at 1073 Kin wet air, as shown
`in Fig. Sa. In contrast to Cu-SAP0-34, Cu-ZSM -5 became
`almost inactive by the same heat treatment, as shown in
`Fig. Sb. The thermal stability of SAP0-34 far exceeds that
`of the aluminosilicate, ZSM-5, since the crystal structure
`of synthesized SAP0-34 was sustained up to 1273 K. even
`in a wet atmosphere (2). Therefore, the thermal stability
`of Cu-SAP0-34 is satisfactorily high enough that it can be
`used as catalyst for automotive exhaust gases.
`The changes in the crystal structure of SAP0-34 under
`NO reduction in a wet atmosphere were studied by XRD,
`as shown in Fig. 6. The XRD pattern in Fig. 6 indicates that
`the synthesized SAP0-34 is a single phase with good crys(cid:173)
`tallinity (18). It is clearly shown that no changes can be rec(cid:173)
`ognized in XRD patterns of Cu-SAP0-34 prior to and after
`the NO reduction at 973 K for 5 h in an atmosphere contain(cid:173)
`ing 15 vol% H20. Moreover, degradation ofcrystals usually
`decreases the intensity ofXRD peaks, and no such decrease
`was observed after NO reduction in a wet atmosphere. This
`suggests that the long-range order of SAP0-34 structure is
`stably sustained under NO reduction at temperatures as
`high as 973 K. even in an atmosphere containing H20.
`Since no changes were observed in XRD patterns of
`SAP0-34 before and after reduction of NO at 973 K in
`a wet atmosphere, possible changes in the local structure of
`SAP0-34 ion-exchanged by Cu were investigated by NMR.
`Figure 7a shows the 27 Al-, 31P-, and 29Si-MAS NMR spectra
`of Cu-SAP0-34 prior to reduction of NO at 973 K for 5 h
`in an atmosphere containing 15 vol% H20. As-synthesized
`SAP0-34 exhibits MAS NMR spectra for Al, P, and Si sim(cid:173)
`ilar to those previously reported (2). The resonance at 42
`and -13 ppm in the 27 Al-NMR spectra seems to be that of a
`tetrahedral and an octahedral coordinated Al, respectively
`
`b
`
`80
`
`60
`
`40
`
`20
`
`l:;'<
`
`£
`0 .5
`0 z
`'0
`" ·~
`"' :> " 0 u
`
`0
`
`oL_~l_~~~-L~-L~-L~~
`
`0.0
`
`1.0
`
`5.0
`4.0
`3.0
`2.0
`Ion-exchanged Cu /wt%.
`
`6.0
`
`FIG. 4. NO conversion at 623 K as a function of the amount of Cu
`ion-exchanged (PNo=lOOO ppm, Pc3H 6 =1000 ppm, Po2 =5%, IMF=
`0.3g·scm-3).
`
`does not always form ion-exchange sites. This means that
`isomorphous substitution of Al with Si has also occurred in
`the case of SAPO-n (17). Consequently, the catalytic per(cid:173)
`formance of Cu-SAP0-34 was studied in more detail with
`Cu-SAP0-34 at 3.0 wt%.
`
`Thermal Stability of Cu-SAP0-34
`
`Exhaust gases from engines contain humidity at high con(cid:173)
`centration. The influence of calcination in a humidified at(cid:173)
`mosphere on NO reduction activity is shown in Fig. 5. Al(cid:173)
`though the activity of Cu-SAP0-34 to NO reduction was
`unaffected by the calcination up to 1073 Kin a dry atmo(cid:173)
`sphere, calcination at 1073 Kin a humidified atmosphere
`decreased the activity for NO reduction. Therefore, the
`
`50
`
`a
`
`0 30
`
`~ i 40
`0 :s
`z .... 0
`=
`.s 20
`l:
`" ~
`8 10
`
`QL---~~~~~~~--~~~~~~~~~--~--~~-L~~
`400
`500
`600
`700
`BOO
`900
`1 000
`900
`Temperature IK
`
`Temperature IK
`
`FIG. 5. Effects of heat treatment under an atmosphere containing 3 val% H20 on the activity for NO reduction with C3Hs (PNo = 5000 ppm,
`Pc3H 6 = 1000 ppm, Po2 = 5%, IMF= 0.3 g · s cm-3). (a) Cu-SAP0-34: (0) 773 K. (t,) 973 K. (D) 1073 K. (b) Cu-ZSM-5: (e) 773 K, (A.) 973 K.
`<•J 1073 K.
`
`Umicore AG & Co. KG
`Exhibit 1112
`Page 4 of 10
`
`
`
`Cu-SAP0-34 FOR THE REDUCTION OF NO WITH C3Hs
`
`97
`
`a
`
`10
`
`20
`30
`28/ degree
`
`40
`
`FIG. 6. X-ray diffraction patterns of Cu-SAP0-34 prior to or after
`NO reduction with C3H6 at 973 K for 5 h in an atmosphere containing
`15 val% H20. (a) Before reaction, (b) after reaction.
`
`(2, 19). On the other hand, the 31P-MAS NMR spectrum
`consists of a strong resonance line at- 26 ppm and a broader
`line at -19 ppm, which could be assigned to a tetrahedral
`P without additional coordination or with water coordi(cid:173)
`nation, respectively (2, 20). Since the amount of Si is far
`smaller than that of Al or P, the intensity of the 29Si-MAS
`NMR spectrum was weaker than those of P orAl; however,
`
`only one resonance due to tetrahedral Si was observed at
`-89 ppm in 29Si-MAS NMR (2, 21, 22). This peak observed
`for Si-MAS NMR could be assigned to Si(4Al) according to
`the reports of Barthomeuf and co-workers (22). At the low
`concentration of Si in SAPO-n, added Si is isomorphously
`substituted for lattice P atoms; however, it is reported that
`the substitution of Si in the Al sites in addition to the P
`sites occurs at high Si content (23). In this study, the Si con(cid:173)
`tent in synthesized SAP0-34 was as low as 1.77 mmol g-1.
`Consequently, all Si atoms added seem to substitute iso(cid:173)
`morphously at the lattice position of the P sites but not the
`Al sites. This is because only one kind of Si bonded with
`4 Al atoms was recognized in 29Si-MAS NMR spectra.
`Figure 7b shows the MAS NMR spectra of 29Si, 27 Al, and
`31P in Cu-SAP0-34 after NO reduction at 973 K, 5 h in a
`wet atmosphere. Although the relative intensity of peaks in
`27 Al-NMRspectra was slightly changed, significant changes
`could not be observed in the MAS NMR spectra of Si,
`Al, and P. These MAS NMR studies clearly indicate that
`SAP0-34 ion-exchanged with Cu retained the crystal struc(cid:173)
`ture in short-range order as well as long-range order, even
`after NO reduction at a temperature as high as 973 k in
`coexistence with water vapor.
`temperature dependence of
`Figure 8 shows
`the
`Cu-SAP0-34 for NO selective reduction with C3He prior
`to and after treatment at 973 K, 5 h in an atmosphere con(cid:173)
`taining 15 vol% HzO. It is clearly shown that the activity
`of Cu-SAP0-34 to NO reduction hardly varied after this
`heat treatment. These unchanged activities to NO reduction
`confirm the high thermal stability of the SAP0-34 crystal
`lattice which is shown by the NMR results.
`Conversion of NO to Nz was studied as a function of
`time-on-stream in the atmosphere containing 15 vol% HzO
`(Fig. 9). Compared with the NO conversion in a dry atmo(cid:173)
`sphere shown in Fig. 1, the presence of water decreased the
`NO conversion. This mayresultfrotn the suppression of NO
`or C3He adsorption by the coadsorption of water; however,
`
`ppm
`
`PP;Di
`
`~pm
`
`FIG. 7 .. MAS NMR spectra of Cu-SAP0-34 prior to or after NO reduction with C3Hs at 973 K for 5 h in an atmosphere containing 15 val% H20.
`(a) Before reaction, (b) after reaction. *Spinning side band.
`
`Umicore AG & Co. KG
`Exhibit 1112
`Page 5 of 10
`
`
`
`98
`
`100
`
`ISHIHARA ET AL.
`
`100
`
`Q
`
`~ 80 z
`.... .s 60
`~ .....
`Q
`~ 40
`.....
`Q
`"' ~
`~ Q 20
`
`0
`
`e Fresh
`0 Afterheat
`treatment
`
`"Ci
`
`80
`
`·Go
`
`~ ..... = "' 0
`= 0 z
`.... Q
`"' "' ~ 0
`
`40
`
`~
`.....
`Q
`
`20
`
`0
`
`···----·----t---·
`
`••
`" "
`"
`
`OL-~~~~~~~~~~~
`400
`500
`600
`700
`800
`900 1000
`Temperature IK
`
`0
`300
`
`400
`
`500
`
`900 1000
`800
`700
`600
`Temperature /K
`
`FIG. 8. Temperature dependence of Cu-SAP0-34 for NO selective
`reduction with C3Hs prior to and after calcination at 973 K for 5 h
`6 =
`in an atmosphere containing 15 vol% HzO (PNo = 1000 ppm, Pc3
`H
`1000 ppm, Po2 = 5%, 1-WF= 0.3 g · s cm-3).
`
`Cu-SAP0-34 exhibits high activity for NO reduction with
`C3H 6 even in an atmosphere containing 15 vol% HzO. Al(cid:173)
`though the conversion to Nz was slightly decreased within
`the initial10 h, decreases in conversion to Nz as well as C3H5
`conversion were negligibly small over the examined 70 h. It
`is also noted that the constancy in the local and long-range
`order of Al, P, and Si in SAP0-34 could be recognized by
`the NMR and XRD measurements, respectively, after NO
`reduction for 70 h in a wet atmosphere. Therefore, the high
`activity of Cu-SAP0-34 for NO selective reduction with
`C3H 6 was stably sustained for a long period, even in an at(cid:173)
`mosphere containing a fairly large amount of water vapor.
`
`Activity of Cu-SAP0-34 for N02 Selectiv(:!
`Reduction with C3H6
`
`Figure 10 shows the temperature dependences of the
`activity of Cu-SAP0-34 for selective reduction of NO or
`
`100 ~ A
`80
`
`6.--
`
`i:J.
`b. C3Ha conversion
`0 Conversion into N2
`
`ll<
`
`..... = 60
`.....
`Q
`Ill
`lol
`~ 40
`
`0
`0
`
`0
`
`20
`
`0
`0
`
`20
`
`60
`40
`Time on stream I h
`
`80
`
`FIG. 9. Conversion into Nz on Cu-SAP0-34 as a function of time on
`stream (PNo= 1000 ppm, Pc3H 6 = 1000 ppm, Po2 = 5%, ll-r2o = 15%).
`
`FIG. 10. Activity of Cu-SAP0-34 for NO and NOz reduction with
`C3Hs (PNoarNOz = 1000 ppm, Pc3H6 = 1000 ppm, Po2 = 5%). NO-C3Hs:
`(0) NO conversion, (e) C3Hs conversion. NOz-C3Hs: (6) NOz conver(cid:173)
`sion, (.A) C3Hs conversion.
`
`N02 with C3H 6• Yogo and Kikuchi reported that Ga ion(cid:173)
`exchanged ZSM-5 exhibits a higher activity for NOz reduc(cid:173)
`tion than for NO reduction (24). Similarly to this result with
`Ga-ZSM-5, Cu-SAP0-34 exhibits a higher activity for NOz
`reduction than for NO reduction over the whole tempera(cid:173)
`ture range examined. In particular, conversion ofNOz to Nz
`attained a value as high as 80% at 573 K, as showninFig.10.
`
`Oxidation State of Cu in SAP0-34 after Heat Treatment
`in C3H6, NO, and 02
`
`It has been reported that the valence of Cu in ZSM-5
`changes between monovalent and divalent during direct
`decomposition of NO and that the adsorbed NO is reduced
`to N2 by the redox behavior of the Cu ion (25). Reduc(cid:173)
`tion of Cu(II) to Cu(I) during the selective reduction of
`NO with hydrocarbon has also bei:m pointed out in the case
`of Cu-ZSM-5 (26). Therefore, it is expected that the re(cid:173)
`dox behavior of the Cu ion has an important role also for
`the selective reduction of NO. ESR gives useful informa(cid:173)
`tion on the oxidation state of copper (27). Figure 11 shows
`the ESR spectra of Cu in SAP0-34 after heating at 623 K,
`1 h in C3H 6, NO, and Oz. ESR signals assignable to Cu(II)
`were observed on Cu-SAP0-34 after evacuation at 523 K
`(Fig. lla). Furthermore, two kinds of hyperfine structure
`possessing differentgvalues were observed, suggesting that
`Cu(II) is dispersed at the atomic level in two different en(cid:173)
`vironments. Two kinds of Cu (II) in SAP0-34 were also re(cid:173)
`ported by Zamadics et al. (28) and assigned to Cu (II) exist(cid:173)
`ing near the plane of the six-membered ring and Cu(II) in
`the hexagonal prism. These ESR signals assigned to Cu (II)
`almost disappeared after heating in C3H5 at 623 K for 1 h
`(Fig. llb). Although it is reported that the ESR spectra of
`
`Umicore AG & Co. KG
`Exhibit 1112
`Page 6 of 10
`
`
`
`Cu-SAP0-34 FOR THE REDUCTION OF NO WITH C3Hs
`
`99
`
`DPPH
`
`g=2.0036 , _____ _
`
`g11=2.304
`
`vt
`vr
`
`b)
`
`c)
`
`~)
`
`FIG. 11. ESR spectra of Cu-SAP0-34 after various treatments.
`(a) 773 K, 5 h evacuation--+ 673 K, 4 h, in 02 (100 Torr)--+ 623 K, 3 h
`evacuation. (b) 623 K, 1 h in C3Hs (30 Torr)--+ 298 K evacuation. (c)
`623 K, 5 h in NO (200 Torr)--+ 298 K evacuation. (d) 623 K, 5 h in Oz
`(200 Torr)--+ 298 K evacuation. (e) Quenched in He during NO reduction
`(PNo = 1000 ppm, Pc3H 6 = 100 ppm, Po2 = 5%, 623 K, 5 h).
`
`Cu (II) ions reappear by exposure to NO in the case of Cu-Y
`zeolite reduced with CO (29), no changes were observed af(cid:173)
`ter subsequent exposure of Cu-SAP0-34 to NO at 623 K
`(Fig. 11c). This suggests that the reduced state of Cu formed
`by C3H5 treatment is stable in SAP0-34 in contrast to that
`in Y zeolite; however, the ESR signals assigned to Cu(II)
`as well as the hyperfine structure were partially restored by
`calcination at 623 K for 5 h in Oz (Fig. 11 d) and almost fully
`restored after treatment with 0 2 for a further long period.
`In addition, it is noted that no significant amount of carbon
`was deposited during C3H5 treatment as judged by the XPS
`measurement for the same treatment. Since the ESR sig(cid:173)
`nal of Cu(II) was restored by the oxidizing treatment, the
`initial disappearance by C3H5 treatment (Fig. llb) is not
`due to the aggregation of Cu ions but to the reduction of
`Cu(II). Similar changes in ESR spectra on heat treatment
`of C3H 6 have been reported for Cu ion-exchanged KY (30)
`and ZSM-5 zeolite (31), where it was also concluded that
`C3H5 readily reduces ion-exchanged Cu(II). Thus, copper
`ions in SAP0-34 are also easily reduced on exposure to
`C3H5. Figure lle shows the ESR spectra of Cu-SAP0-34
`quenched in He during NO selective reduction with C3H5
`at 623 K. Since Kucherov et al. reported that no measur(cid:173)
`able change of cupric cations occurs in a flow of pure He up
`to 773 Kin the case of ZSM-5 (31), it is expected that no
`
`TABLEl
`
`XPS Binding Energy of Cu 2p3/2 Electrons in Cu-SAP0-34
`after Heating in Oz and C3Hs
`
`Binding energy (eV)
`
`Cu 2[13,2 Shake-up line
`
`Remarks
`
`Cu-SAP0-34
`Cu-SAP0-34
`Cu-SAP0-34
`Cu
`Cu20
`CuO
`
`934.0
`932.5
`932.6
`932.2
`932.8
`934.3
`
`942.9
`
`943.0
`
`Heated in 0 2 at 623 K, 1 h
`Heated in C3Hs at 623 K, 1 h
`After reaction
`
`reduction occurred during the quenching in He. Although
`the ESRsignals assigned to Cu(II) were still observed, the
`double integral of the signal in Fig. 11e was far smaller than
`that of as-prepared Cu-SAP0-34 in Fig. 11a. This suggests
`that part of the Cu(II) in SAP0-34 remains in a reduced
`state under the selective reduction of NO with C3H 6•
`The valence changes of Cu ions were further studied by
`XPS (Table 1), and typical Cu 2PJt2 spectra after the various
`treatments are shown in Fig. 12. The oxidation state of Cu
`
`Cu 2p3/2
`
`Shakeup line
`
`Binding energy I eV
`
`FIG. 12. XPS spectra of Cu 2[13t2 line of Cu-SAP0-34. (a) 723 K.
`1 h in 02 and then evacuation. (b) 723 K, 1 h in C3Hs and then evacua(cid:173)
`tion. (c) Quenched in He during NO reduction (PNo = 1000 ppm, Pc3H
`6 =
`100 ppm, Po2 =5%, 623 K, 5 h).
`
`Umicore AG & Co. KG
`Exhibit 1112
`Page 7 of 10
`
`
`
`- - - - - - - - - - - - - - - - - - - -
`
`100
`
`ISHIHARA ET AL.
`
`in SAP0-34 seems to be Cu (II) after heating in Oz at 723 K
`followed by evacuation to 10-7 Pa at room temperature,
`since the binding energyofCu 2p3/2 (Table 1) and the Auger
`Cu LMM energy (335.3 e V) of Cu in SAP0-34 were almost
`consistent with those of CuO, and the shake-up satellite
`peaks were also observed in the binding energy range 940
`to 950 eV as shown in Fig. 12a. Therefore, XPS measure(cid:173)
`ments confirmed the results of ESR measurements and it is
`clear that the oxidation state of Cu in as-prepared SAP0-34
`is divalent. C3H5 treatment at 723 K decreased the bind(cid:173)
`ing energy of the Cu 2p3/2 peak and slightly increased the
`Cu LMM energy; moreover, the XPS satellite peaks dis(cid:173)
`appeared (Fig. 12b). The XPS spectrum was now similar
`to that of Cu20, but not that of metallic Cu or CuO, so
`Cu (II) in SAP0-34 is reduced to Cu (I) by C3H5 treatment.
`Although the oxidation number of Cu after the heating, in
`C3H5 could not be determined by ESR, the reduced state
`shown by ESR seems to be Cu(I) from the above XPS
`measurement. On the other hand, the XPS spectrum of
`Cu-SAP0-34 quenched during NO reduction at 623 K, 5 h
`closely resembled that of Cu-SAP0-34 treated with C3H5,
`although a weak shoulder was observed as shown in Fig. 12c.
`Considering the results of ESR, Cu (II) species also exist
`in SAP0-34 with reduced species of Cu on the quenched
`sample. It is most likely that the valence state of copper in
`SAP0-34 is a mixed state of monovalent and divalent under
`the selective reduction of NO. Facility in the monovalent(cid:173)
`to-divalent redox of Cu ion also plays a key role in the
`reduction of NO in the case of direct NO decomposition on
`Cu-ZSM-5 (25) and NO selective reduction (32).
`
`Chemisorption and Transformation of NO and C3H5
`on Cu-SAP0-34
`
`The adsorbed states of NO and C3H5 on Cu-SAP0-34
`were investigated by infrared spectroscopy. Figure 13 shows
`the IR spectra of Cu-SAP0-34 after heating in NO, C3H5,
`and Oz at 723 K. After adsorption of NO at 673 K, a
`strong absorption band at 1890 cm-1 and some weak ab(cid:173)
`sorption bands around 2200 and 1600 cm-1 were observed
`(Fig. 13a). The 1890 cm-1 band can be assigned to adsorbed
`NO, formed by donation of an electron from the antibond(cid:173)
`ing orbital of NO, so adsorption of NO on Cu(II)-SAP0-
`34 is assumed to be reductive (33). The bands at 2240
`and 2125 cm-1 are assigned to adsorbed NzO and NOt
`species, respectively, in accordance with IR studies of ad(cid:173)
`sorbed NO on Cu-ZSM-5 (33, 34) and CuO/SiOz (33, 35).
`Although bands around 1600 cm- 1 were not observed in the
`IR spectra of adsorbed NO on Cu ion-exchanged ZSM-5
`(34), absorption bands at 1600 and 1510 cm-1 can be as(cid:173)
`signed to the nitrate ion (NO:J) from the assignment of
`adsorbed NO on CuO/SiOz (35). The mass spectra of the
`gas phase taken at this stage consisted of a strong peak
`with mass number at 30, which was assigned to NO, and
`two weak peaks at 28 and 44, which corresponded to Nz
`
`No+
`1890
`
`j
`..0 <
`
`0
`fll
`
`3000
`
`2500
`
`2000 1800
`1600 1400
`Wav~number /cm- 1
`
`IR spectra of Cu-SAP0-34 after exposure to NO, C3H6, and
`FIG. 13.
`Oz. (a) Heating at 623 K, 3 h in NO (100 Torr). (b) Evacuation at 298 K.
`(c) Heating at 623 K, 6 h in C3Hs (30 Torr). (d) Calcination at 623 K, 1 h
`in Oz (100 Torr). (e) Evacuation at 298 K.
`
`and NzO, respectively. Therefore, the disproportionation
`of NO (for example, 3NO + cu+ = Nz +Cu2+ + NO:Jact
`or 4NO + 2Cu+ = NzO + 2Cu2+ + 2NO:Jact) seems to have
`occurred on Cu-SAP0-34 at 723 K to form adsorbed NO:J
`and gaseous N20 and Nz. All absorption bands exceptthose
`assigned to nitrate species disappear after evacuation ofthe
`gas phase at robm temperature (Fig. 13b). Therefore, the
`adsorption of NO is extremely