Applied Catalysis A: General 275 (2004) 207–212
`
`www.elsevier.com/locate/apcata
`
`Deactivation of La-Fe-ZSM-5 catalyst for selective catalytic reduction
`of NO with NH3: field study results
`
`Gongshin Qia, Ralph T. Yanga,*, Ramsay Changb, Sylvio Cardosob, Randall A. Smithc
`
`aDepartment of Chemical Engineering, University of Michigan, Ann Arbor, MI 48109, USA
`bAir Pollution Control, Power Generation, Electric Power Research Institute, Palo Alto, CA 94304-1395, USA
`cFossil Energy Research Corporation, Laguna Hills, CA 92653, USA
`
`Received 20 May 2004; received in revised form 10 July 2004; accepted 22 July 2004
`Available online 11 September 2004
`
`Abstract
`
`Results are summarized for a study on the effects of poisons on the La-Fe-ZSM-5 catalyst activity for the selective catalytic reduction of
`NO by ammonia. The deactivation of La-Fe-ZSM-5 honeycombs was studied in field tests. A honeycomb catalyst containing 25%La-Fe-
`ZSM-5 had an overall activity similar to that of a commercial vanadia honeycomb catalyst. Long-term activity test results show that the
`25%La-Fe-ZSM-5 catalyst activity decreased to 50% after 300 h and 25% after 1769 h of on-stream flue gas exposure. The deactivation is
`correlated to the amounts of poisons deposited on the catalyst. Poisons include alkali and alkaline earth metals, As and Hg. Hg was found to be
`ion-exchanged from HgCl2 to form Hg-ZSM-5, and Hg was found to be among the strongest poisons. The poisoning effects of these elements
`appeared to be additive. Thus, from the chemical analysis of the deactivated catalyst, the deactivation of Fe-ZSM-5 can be predicted.
`# 2004 Elsevier B.V. All rights reserved.
`
`Keywords: Catalyst deactivation; Deactivation of SCR catalyst; Acid catalyst deactivation; Selective catalytic reduction
`
`1. Introduction
`
`Vanadia-based catalysts (V2O5 + WO3 (or MoO3)/TiO2)
`are being used as the commercial catalysts for selective
`catalytic reduction (SCR) of NO with ammonia for power-
`plant emission control. However, many problems remain
`with the use of these catalysts (e.g., high SO2 oxidation
`activity with high-sulfur fuels). For this reason, new and
`more active catalysts are of considerable research interests.
`A number of zeolites have been found to have substantially
`higher activities than the vanadia-based catalysts and are
`also devoid of
`their problems [1–11]. Fe-ZSM-5,
`in
`particular, is among the most promising new catalysts [6–8].
`Because of the severe environments in which the SCR
`reaction is conducted, long-term deactivation has been an
`important practical problem. Although the causes for
`deactivation are many and complex, chemical deactivation
`
`* Corresponding author. Tel.: +1 734 936 0771; fax: +1 734 764 7453.
`E-mail address: yang@umich.edu (R.T. Yang).
`
`0926-860X/$ – see front matter # 2004 Elsevier B.V. All rights reserved.
`doi:10.1016/j.apcata.2004.07.051
`
`is a major cause and it is directly related to the mechanism of
`the SCR reaction. The SCR mechanism on vanadia catalysts
`has been studied extensively and several different mechan-
`isms have been proposed [12–20]. Most researchers agree
`that the SCR reaction on the vanadia catalysts follows an
`Eley–Rideal
`type mechanism,
`i.e., a strongly adsorbed
`+) reacts with a gaseous or weakly
`ammonia species (NH4
`adsorbed NO molecule to form one molecule of N2, which
`requires dual sites of V5+=O and V4+–OH.
`Despite the large amount of studies on the reaction
`mechanism and the importance of the deactivation, little
`work has been done on deactivation. A systematic study of
`the deactivation mechanism of the vanadia catalyst has been
`performed in our laboratory [21,22]. Our results showed that
`deactivation was caused by the basicity of the poisons,
`which decreased the Bronsted acidity of vanadia thereby
`decreasing the activity. Alkali metals are among the
`strongest poisons. Moreover,
`the extent of deactivation
`has been correlated with the M/V atomic ratio (where M is
`the poison metal) for alkali/alkaline earth metals as well as
`
`BASF-2044.001
`
`

`
`208
`
`G. Qi et al. / Applied Catalysis A: General 275 (2004) 207–212
`
`As [21,22]. For deactivation of the zeolite catalyst, a number
`of studies have been reported for the deactivation of Cu-
`ZSM-5 and Ce-ZSM-5 in the SCR reaction using hydro-
`carbons as the reducing agents [23–26].
`In this work, we report results from a field study of the
`long-term deactivation of the Fe-ZSM-5 catalyst for NH3
`SCR. The study was undertaken in a coal-burning power
`plant with a honeycomb catalyst under commercial
`operating conditions. In addition,
`the poisoning of the
`SCR catalyst by Hg was measured in our laboratory and
`reported here for the first time. We show that the extent of
`deactivation could be predicted from a chemical analysis of
`the poisons that were deposited on the catalyst.
`
`2. Experimental/field test
`
`Kilogram quantities of the La-Fe-ZSM-5 catalyst were
`first prepared by Eltron Research, Inc. (Boulder, CO). The
`preparation procedure was essentially the same as that
`developed in the laboratory [7]. The catalyst was subse-
`quently extruded into honeycombs by Cormetech, Inc.
`(Durham, NC). Field tests of the honeycomb catalysts were
`performed by Fossil Energy Research Corporation (Laguna
`Hills, CA) at plant Scherer Station, owned by the Oglethorpe
`Power Corporation (Tucker, GA). The test reactor was
`designed and built by Fossil Energy Research Corporation,
`and fed with a slip stream of the flue gas from Scherer Unit 1,
`which was firing a low-sulfur Eastern coal at the time of the
`tests. Sootblowing is the terminology used for the action of
`blowing air or steam to remove deposits off the water walls
`in a boiler and is used in utility SCR reactors to clean the
`catalyst. Since the mini reactor is made for long-term
`exposures, FERCo uses the same terminology to the action
`of blowing instrument air at the same temperature as the flue
`gas to clean off any buildup of ash deposition around the
`inlet and catalyst paths. The continued operation of these
`sootblowers maintains the mini reactor from plugging up
`due to ash deposits.
`
`2.1. Preparation of catalysts
`
`The optimized La-Fe-ZSM-5 catalyst formulation and
`+-ZSM-5
`preparation involved exchanging 1 or 3 kg of NH4
`with 0.05 M lanthanum nitrate in water at 20 8C. Following
`recovery, the La-ZSM-5 was exchanged in 0.1 M ferrous
`chloride solution at 50 8C. During the exchange procedures,
`the solutions were purged with helium to exclude oxygen.
`The resulting material was dried and then calcined at
`500 8C for 8 h. Monolithic honeycombs were prepared by
`Cormetech Corporation. For cost considerations,
`the
`honeycomb catalysts were prepared with different propor-
`tions of La-Fe-ZSM-5 and Na-ZSM-5/NH4-ZSM-5 plus
`small amounts of silica binder. Thus, honeycombs contain-
`ing 25 and 8%La-Fe-ZSM-5 were extruded and tested
`(wt.%, balance being Na-ZSM-5). The samples were
`
`prepared in extruded blocks nominally 768 mm square by
`152 mm long. The pitch and wall thickness of the catalysts
`were 7 and 1 mm, respectively [27].
`In order to study the effect of Hg on Fe-ZSM-5 catalysts,
`we also prepared the Hg-ZSM-5 and Fe-Hg-ZSM-5 catalysts
`using ion exchange method. Two grams of NH4-ZSM-5 (Si/
`Al = 10, obtained from Air Products and Chemicals Inc.)
`was added to 200 ml of 0.06 M HgCl2 solution with constant
`stirring. The ion exchange was carried out in air at room
`temperature for 24 h. In order to increase the content of Hg,
`the ion exchange was
`repeated several
`times. After
`exchange,
`the mixture was filtered and washed with
`deionized water. The obtained solid was first dried at
`120 8C for 12 h and then calcined at 400 8C for 6 h in air.
`The Hg-Fe-ZSM-5 catalyst was prepared in a similar
`manner as was the Hg-ZSM-5, except by using Fe-ZSM-5
`instead of NH4-ZSM-5. The Hg contents in the samples were
`measured by EDX. EDX can detect
`the elements
`quantitatively. The higher the atomic weight, the higher is
`the accuracy. So it is particularly useful for Hg and As, both
`of which are among the strongest poisons for the SCR
`reaction. A quantitative analysis can be performed by a
`standard or standardless analysis. Here we used the
`standardless analysis. This method quantifies the elements
`by calculating the area under the peak of each element
`combined with appropriate sensitivity factors.
`
`2.2. Field evaluation of extruded monolith catalysts
`
`Field tests were conducted on a slipstream at Plant
`Scherer Unit 1 located in Juliette, GA. Scherer burns a blend
`of low-sulfur, bituminous and subbituminous coals. The flue
`gas NOx levels were controlled in the combustion zone and
`ranged from 150 to 250 ppm. Evaluation of selected catalyst
`samples was conducted using an in situ mini SCR test
`reactor designed and fabricated by Fossil Energy Research
`Corporation (FERCo). The reactor consisted of a 6-in.
`(15.25 cm) diameter stainless shell housing that was fully
`inserted into the flue gas stream. The catalyst was installed in
`an inner housing that could be removed from the outer
`reactor shell. The inner catalyst holder provided full access
`for inspection and cleaning. The details of the mini SCR
`reactor is shown in Fig. 1. Anhydrous ammonia was injected
`through a multiple orifice nozzle at the probe inlet and the
`ammonia injection rate was controlled by a mass flow
`controller. Fig. 1 also shows a photograph illustration of how
`the catalyst samples were loaded. NOx was analyzed by an
`electrochemical-based NO and O2 emission analysis system.
`The NO/O2 sampling and analysis system was not operated
`continuously but only during short periods for the catalyst
`activity test.
`
`2.3. Catalyst activity determination
`
`SCR catalyst’s activity depends on a series of physical
`transport and chemical reaction steps. The chemical reaction
`
`BASF-2044.002
`
`

`
`G. Qi et al. / Applied Catalysis A: General 275 (2004) 207–212
`
`209
`
`Fig. 1. Field test reactor (upper) and monolith catalyst assembly (lower).
`
`V2O5/TiO2), 8%Fe-ZSM-5 and 25%Fe-ZSM-5, were inves-
`tigated. NO removal was measured for each catalyst sample
`at three NH3/NO ratios. The catalyst activity constant for
`each La-Fe-ZSM-5 formulation can be calculated from Eq.
`(1). The 25%Fe-ZSM-5 catalyst had an activity constant k
`similar to the commercial V2O5/TiO2. This is consistent with
`the activity of the pure La-Fe-ZSM-5 catalyst powder. The
`intrinsic activity of pure La-Fe-ZSM-5 catalyst powder was
`about three to four times that of V2O5/TiO2 [6,27]. Thus, a
`monolith extruded using 25%La-Fe-ZSM-5 would have a
`catalyst activity similar to standard V2O5/TiO2. The activity
`of 8%La-Fe-ZSM-5 catalyst was about 30% lower than that
`of 25%La-Fe-ZSM-5 catalyst.
`
`is first order with respect to ammonia concentration (under
`ammonia lean conditions, i.e., NH3/NO < 1, as used in this
`work). The overall or integral catalyst activity can be
`represented by an apparent first-order rate constant, k
`[21,22]. For this study, k is defined by (i.e., assuming being
`free from mass-transfer resistances, [21,22]):
`
` 
`lnð1 DNOxÞ
`k ¼ SV
`As
`
`(1)
`
`where DNOx is the fractional NOx removal, SV is the catalyst
`space velocity in 1/h and As is the specific surface area of the
`catalyst sample in m2/m3. The quantity SV/As is known as
`the catalyst area velocity and is often used by the catalyst
`industry in the calculation of k to accommodate difference in
`catalysts.
`
`2.4. XPS characterization
`
`The XPS experiment was carried out on a Perkin-Elmer
`8–
`PHI 5400 ESCA system at room temperature under 10
`9 Torr, using Mg Ka (1253.6 eV) radiation. The binding
`10
`energy of Hg 4f was calibrated relative to the carbon
`impurity with a C 1s at 284.5 eV.
`
`3. Experimental, field test results and discussion
`
`3.1. Initial catalyst activity comparison
`
`The initial catalyst activities as functions of the NH3/NO
`ratio are shown in Fig. 2. Three different monolithic
`samples, a commercial vanadia catalyst (designated by
`
`Fig. 2. Pilot honeycomb catalyst evaluation results for fresh catalysts.
`Reaction conditions: gas velocity in honeycomb = 4.6–5.5 m/s, temperature
`= 370–390 8C.
`
`BASF-2044.003
`
`

`
`210
`
`G. Qi et al. / Applied Catalysis A: General 275 (2004) 207–212
`
`3.3. Deactivation by poisons (k/k0 versus amounts of
`poisons)
`
`Much effort has been focused on the understanding of the
`kinetics and mechanism of the SCR reaction. It has been
`shown that there are some similarities in the mechanism of
`the SCR reaction between the V2O5/TiO2 and the Fe-ZSM-5
`catalysts, both of which involve the formation of ammonium
`ions on the Bronsted acid sites [19–22]. Weakening of the
`Bronsted acidity leads to deactivation. So, it is reasonable to
`use the deactivation data obtained from the V2O5/TiO2
`catalyst on the Fe-ZSM-5 catalyst. The deactivation of the
`SCR catalyst was correlated as functions of the amount of
`poisons that are deposited on the catalysts, expressed as k/k0
`versus M/V (where M/V is the ratio of the poison metal over
`vanadium)
`[21,22]. The correlations of
`the effects of
`different metal oxides on the relative rate constants are
`shown in Fig. 5.
`
`Fig. 3. Long-term NOx removal results for 25%La-Fe-ZSM-5 catalyst.
`Reaction conditions: gas velocity in honeycomb = 4.6–5.5 m/s, temperature
`= 370–390 8C, NH3/NO = 2.0.
`
`3.2. Long-term activity results
`
`3.4. Poisoning by Hg
`
`The activity test results are presented as k/k0, which is the
`ratio of the activity measured at different exposure periods to
`that measured when the catalyst was fresh. The reactor was
`operated at an NH3/NO ratio of 0.9 on a day-to-day basis
`during the long-term activity tests. The periodic activity tests
`were run at a higher injection rate (NH3/NO of nominally
`2.0), in order to properly assess k/k0. The activity test results
`for the 25%La-Fe-ZSM-5 are summarized in Figs. 3 and 4.
`The results of the long-term tests indicate that the catalytic
`activity decreased rapidly in the first 300 h of flue gas
`exposure, with k/k0 values of nominally 0.5 indicating that
`the catalyst had lost 50% of its original activity. After
`1769 h, k/k0 decreased to 0.25. Examination of the catalyst
`samples showed that some erosion had occurred over the
`course of
`testing, which would have contributed to
`deactivation. However, the major contribution to deactiva-
`tion was due to poisons deposited on the catalysts.
`
`Since our chemical analysis of the deactivated catalysts
`showed significant amounts of Hg, and the effects of Hg
`were not included in our earlier studies, such effects are
`studied here.
`It is known that mercury is difficult to control because it is
`present in flue gas as a vapor (either in the elemental or ionic
`form, i.e., HgCl2), rather than as particulate matter. In the
`field test, the 25%La-Fe-ZSM-5 contained a large amount of
`H-ZSM-5. During the SCR reaction (300–400 8C), vapor-
`phase ion exchange took place readily between H-ZSM-5
`and HgCl2, similar to the results of subliming FeCl3 on H-
`ZSM-5, reported by Chen and Sachtler [5]. The substitution
`
`Fig. 4. Long-term activity ratio (k/k0) results for the 25%La-Fe-ZSM-5
`catalyst. Reaction conditions: gas velocity in honeycomb = 4.6–5.5 m/s,
`temperature = 370–390 8C, NH3/NO = 2.0.
`
`Fig. 5. Relative rate constant (k/k0) for SCR reaction on 5%V2O5 (wt.)/TiO2
`doped with various amounts of alkali metal oxides. The effect of As was
`from reference [25] and the others were from reference [26]. Reaction
`conditions: [NH3] = [NO] = 1000 ppm, [O2] = 2%, He = balance, T =
`1.
`300 8C, GHSV = 15,000 h
`
`BASF-2044.004
`
`

`
`G. Qi et al. / Applied Catalysis A: General 275 (2004) 207–212
`
`211
`
`Fig. 6. XPS spectra of Hg 4f (Mg Ka) for different Fe-ZSM-5 samples.
`
`of proton by Hg could significantly decrease the activity of
`H-ZSM-5. In addition to replacing proton, the replacement
`of Fe by Hg could occur, thereby further decreasing the
`activity of Fe-ZSM-5.
`In order to determine the state of Hg (and to verify that
`ion exchange for Hg had indeed occurred during the SCR
`reaction),
`the deactivated catalyst and a FeHg-ZSM-5
`sample prepared by ion exchange in our laboratory were
`analyzed by using XPS. The XPS spectra of these two
`samples are shown in Fig. 6. From Fig. 6, it is clearly seen
`that only one peak at around 104 eV was present for both
`samples. The two spectra are very similar, although there is a
`small difference in the peak binding energy. The binding
`energies at 104.3 and 104.6 eV indicate that there was no Hg
`metal (as Hg0) on these two samples, because the component
`of Hg 4f7/2 associated with metallic mercury is at 99.85 eV
`[28,29] and such a peak was not detected by XPS on these
`two samples. The binding energy of HgO is at 100.7 eV,
`which is also very different from the peaks shown in Fig. 6.
`In our case, the binding energies of mercury were 104.3 and
`104.6 eV, which are higher than that of the pure HgO.
`Ehrhardt et al. [30] studied the sorption of Hg2+ onto pyrite
`FeS2, and found a peak at 100.7 eV and an additional peak at
`103.8 eV for Hg 4f level. They assigned these two peaks to
`the adsorption of Hg2+ onto both patches of pyrite (Eb for Hg
`4f = 100.7 eV) and islands of Fe3+ oxyhydroxides (Eb for Hg
`4f = 103.8 eV) [30]. In this work, we detected two peaks at
`around 104 eV, which is close to the value reported by
`Ehrhardt et al. [30]. For the deactivated catalyst and the
`HgFe-ZSM-5 prepared by ion exchange, it is reasonable to
`assign these two peaks to exchanged Hg2+. This result also
`showed that vapor-phase ion exchange of mercury onto
`ZSM-5 occurred during the SCR reaction.
`Next, the SCR activities of partially Hg-exchanged Fe-
`ZSM-5 catalysts were measured at 350 8C. The NO
`conversion data were converted into k/k0. The relative k/
`k0 was plotted against Hg/Al, shown in Fig. 7. It is seen that
`
`Fig. 7. Relative rate constant (k/k0) for SCR reaction on Hg-ZSM-5 with
`various amounts of Hg. Reaction conditions: [NH3] = [NO] = 1000 ppm,
`[O2] = 2%, He = balance, T = 350 8C, flow rate = 500 ml/min, catalyst 0.2 g.
`
`Hg is a very strong poison, similar to the strongest poison
`among the alkali metals, Cs (Fig. 5).
`
`3.5. Analysis/prediction of deactivation from
`amounts of poisons
`
`The EDX analysis of the deactivated monolithic Fe-
`ZSM-5 catalyst is given in Table 1. It is noted that EDX can
`provide very good elemental analyses for zeolites because of
`their chemically uniform crystal structures.
`From Table 1, on the deactivated Fe-ZSM-5 sample, Na,
`S, K, Ca and As were detected by the EDX analysis. The
`corresponding ratios of metal to aluminum are also included
`in Table 1. From these ratios, and using the deactivation data
`given in Figs. 5 and 7, one could predict the extent of
`deactivation. At Na/Al = 0.22, from Fig. 6, the relative rate
`constant would decrease from 1.0 to 0.7 so the rate constant
`should decrease to 70% of the original rate due to Na. For
`K/Al = 0.045, the rate constant should decrease to 93.3% of
`
`Table 1
`Composition of the deactivated catalyst from field test
`
`Element
`
`Fe-ZSM-5
`
`O
`Na
`Al
`Si
`Hg
`S
`K
`Ca
`Fe
`As
`
`wt.%a
`
`45.3
`0.5
`2.9
`45.0
`0.8
`1.8
`0.18
`0.15
`1.25
`2.0
`
`at.%b
`
`60.4
`0.51
`2.33
`34.2
`0.09
`1.20
`0.11
`0.08
`0.47
`0.58
`
`Metal/Alc
`
`0.22
`
`0.04
`0.51
`0.045
`0.034
`
`0.25
`
`a Weight percent.
`b Atom percent.
`c Molar ratio of metal/aluminum.
`
`BASF-2044.005
`
`

`
`212
`
`G. Qi et al. / Applied Catalysis A: General 275 (2004) 207–212
`
`the original rate. For Ca/Al = 0.034 and As/Al = 0.25, the
`rate constant should drop to 97 and 74%, respectively. From
`Fig. 7, at Hg/Al = 0.042, the rate constant would decrease to
`60% of the original value. By combining/adding the effects
`of all metal poisons, the rate constant should have decreased
`to 28% of the original value. The experimental value for the
`deactivation, from the long-term field test data, was 25%.
`This is indeed excellent agreement.
`Form the analysis above, it can be concluded most of the
`poison effect of the Fe-ZSM-5 catalyst was due to As and
`Hg; especially for Hg which reduced 40% of the overall
`activity.
`
`4. Summary
`
`The deactivation of La-Fe-ZSM-5 was studied by field
`test
`in a coal-fired power plant burning a low-sulfur
`bituminous coal. Based on the field test, a honeycomb
`catalyst containing 25%La-Fe-ZSM-5 had an overall
`catalyst activity similar to a commercial vanadia catalyst.
`Long-term activity test results show that the La-Fe-ZSM-5
`catalyst activity decreased significantly over time. The
`relative activity decreased to 0.5 after 300 h and 0.25 after
`1769 h of on-stream flue gas exposure. The deactivation is
`correlated to the amounts of poisons deposited on the
`catalyst. Poisons include alkali and alkaline earth metals, As
`and Hg. Hg was ion exchanged from HgCl2 to form Hg-
`ZSM-5, and was found to be among the strongest poisons.
`From the chemical analysis of the deactivated catalyst, the
`deactivation of Fe-ZSM-5 can be predicted.
`
`Acknowledgments
`
`We thank the assistance of Eltron Research, Inc. (Boulder,
`CO), Fossil Energy Research Corporation (Laguna Beach,
`CA), Cormetech Corporation and Oglethorpe Power Corpora-
`tion for invaluable help and cooperation in this study.
`
`References
`
`[1] J.R. Kiovsky, P.B. Koradia, C.T. Lim, I&EC Prod. Res. Dev. 19 (1980)
`218.
`[2] S. Sato, H. Hirabayashi, H. Yahiro, N. Mizuno, M. Iwamoto, Catal.
`Lett. 12 (1992) 193.
`[3] X. Feng, W.K. Hall, J. Catal. 166 (1997) 368.
`[4] W.K. Hall, X. Feng, J. Dumesic, R. Watwe, Catal. Lett. 52 (1998) 13.
`[5] H.-Y. Chen, W.M.H. Sachtler, Catal. Today 42 (1998) 73.
`[6] R.Q. Long, R.T. Yang, J. Am. Chem. Soc. 121 (1999) 5595.
`[7] R.Q. Long, R.T. Yang, J. Catal. 188 (1999) 332.
`[8] A.Z. Ma, W. Grunert, Chem. Commun. (1999) 71.
`[9] J. Valyon, W.S. Millman, W.K. Hall, Catal. Lett. 24 (1994) 215.
`[10] Q. Sun, Z.-H. Gao, H.-Y. Chen, W.M.H. Sachtler, J. Catal. 201 (2001)
`89.
`[11] M.D. Amiridis, F. Puglisi, J.A. Dumesic, W.S. Millman, N.-Y. Topsøe,
`J. Catal. 142 (1993) 572.
`[12] H. Bosh, F. Janssen, Catal. Today 2 (1988) 369.
`[13] G. Busca, L. Lietti, G. Ramis, F. Berti, Appl. Catal. B 18 (1998) 1.
`[14] M. Schramlmarth, A. Wokaun, A. Baiker, J. Catal. 124 (1990) 86.
`[15] G.T. Went, L.J. Leu, R.R. Rosin, A.T. Bell, J. Catal. 134 (1992)
`492.
`[16] U.S. Ozkan, Y. Cai, M.W. Kumthekar, L. Zhang, J. Catal. 142 (1993)
`182.
`[17] U.S. Ozkan, Y. Cai, M.W. Kumthekar, J. Catal. 149 (1994) 375.
`[18] U.S. Ozkan, Y. Cai, M.W. Kumthekar, J. Catal. 149 (1994) 390.
`[19] N.-Y. Topsøe, H. Topsøe, J.A. Dumesic, J. Catal. 151 (1995) 226.
`[20] N.-Y. Topsøe, J.A. Dumesic, H. Topsøe, J. Catal. 151 (1995) 241.
`[21] J.P. Chen, M.A. Buzanowski, R.T. Yang, J. Air Waste Manag. Assoc.
`40 (1990) 1403.
`[22] J.P. Chen, R.T. Yang, J. Catal. 125 (1990) 411.
`[23] J.L. d0Itri, W.M.H. Sachtler, Appl. Catal. B. 2 (1993) 7.
`[24] A.V. Kucherov, C.P. Hubbard, T.N. Kucherova, M. Shelef, Appl. Catal.
`B 7 (1996) 285.
`[25] Y. Zhang, M. Flytzani-Stephanopoulos, J. Catal. 164 (1996) 131.
`[26] W.E.J. van Kooten, H.C. Krijnsen, C.M. van den Bleek, H.P.A. Calis,
`Appl. Catal. B 25 (2000) 125.
`[27] R. Chang, S. Cardoso, Advanced NOx Catalyst Development Update,
`EPRI, Palo Alto, CA, and, Oglethorpe Power Corporation, Juliette,
`GA, EPRI Report 1006657 (2003).
`[28] C.J. Powell, Appl. Surf. Sci. 89 (1995) 141.
`[29] C.D. Wagner, W.M. Riggs, L.E. Davis, J.F. Moulder, G.E. Mullerberg
`(Eds.), Handbook of X-Ray Photoelectron Spectroscopy, Perkin–
`Elmer Corp., Eden Prairie, 1979.
`[30] J.-J. Ehrhardt, P. Behra, P. Bonnissel-Gissinger, M. Alnot, Surf. Interf.
`Anal. 30 (2000) 269.
`
`BASF-2044.006