Selective catalytic reduction of NO with ammonia over FE3+-exchanged mordenite (FE-MOR): Catalytic performance, characterization, and mechanistic study

Article abstract of DOI:10.1006/jcat.2002.3521

Fe-exchanged zeolites, i.e., mordenite (MOR), clinoptilolite (HEU), beta, ferrierite (FER), and chabazite (CHA), are studied as catalysts for selective catalytic reduction (SCR) of NO with NH₃. It is found that SCR activity decreases in the following order: Fe-MOR>Fe-HEU>Fe-FER>Fe-beta>Fe-CHA. Fe-MOR and Fe-HEU are much more active than the commercial vanadia catalyst. For Fe-MOR, SCR activity increases with a decreasing Si/Al ratio. Moreover, SO₂ and SO₂+H₂O improve SCR activity. On Fe-MOR, nearly 100% NO conversion is obtained at a high space velocity (GHSV=4.6×10⁵ h^-1) in the presence of SO₂. The Fe-MOR catalysts are also characterized by H₂-TPR (temperature-programmed reduction) and Fourier transform infrared (FT-IR) spectroscopy. TPR profiles indicate that iron cations in the catalysts are present as approximately 73% Fe³⁺ and 27% Fe²⁺. FT-IR spectra show that NO can be oxidized by O₂ to N₂O₃, NO₂, and nitrate adsorbed species, and that they are bonded to the iron cations. NH₃ molecules adsorb on Bronsted acid sites of the zeolite to form NH₄⁺ ions.NO+O₂ is very active in reacting with NH₄⁺ ions on Fe-MOR at 300°C, but it is less active with those on H-MOR. This is in good agreement with their SCR activities and is probably related to the fast formation of NO₂ on Fe-MOR. A possible reaction scheme for SCR reduction is proposed. NO reduction involves the reaction between NO₂ and a pair of NH₄⁺ ions to generate an active intermediate, which then reacts with NO to produce N₂ and H₂O. NO oxidation to NO₂ is the rate-determining step.

Full text of DOI:10.1006/jcat.2002.3521

Journal of Catalysis 207, 274–285 (2002)
doi:10.1006/jcat.2002.3521, available online at http://www.idealibrary.com on
3+
Selective Catalytic Reduction of NO with Ammonia over Fe -Exchanged
Mordenite (Fe–MOR): Catalytic Performance, Characterization,
and Mechanistic Study
R. Q. Long and R. T. Yang1
Department of Chemical Engineering, University of Michigan, Ann Arbor, Michigan 48109-2136
Received August 31, 2001; revised January 9, 2002; accepted January 9, 2002
mixed oxides and ion-exchanged molecular sieves (1–3).
Fe-exchanged zeolites, i.e., mordenite (MOR), clinoptilolite The commercial catalysts are vanadia supported on tita-
(
HEU), beta, ferrierite (FER), and chabazite (CHA), are studied nia with WO and/or MoO as promoters. Ion-exchanged
3 3
as catalysts for selective catalytic reduction (SCR) of NO with
molecular sieves have received much attention for the SCR
reaction recently because they function in a broad temper-
ature window and do not have the disposal problem that is
associated with vanadia.
Various transition metal ion-exchanged molecular sieves,
such as Ce–mordenite (MOR) (4, 5), Ce–ZSM-5 (6), VO–
ZSM-5 (7), Cu–MOR (8, 9), Cu–ZSM-5 (10), Cu–Y (11),
Fe–Y (12, 13), Fe–MOR (8), Fe–ZSM-5 (14–18), and pil-
NH . It is found that SCR activity decreases in the following
3
order: Fe–MOR > Fe–HEU > Fe–FER > Fe-beta > Fe–CHA. Fe–
MOR and Fe–HEU are much more active than the commercial
vanadia catalyst. For Fe–MOR, SCR activity increases with a de-
creasing Si/Al ratio. Moreover, SO and SO + H O improve SCR
2
2
2
activity. On Fe–MOR, nearly 100% NO conversion is obtained at
5
−1
a high space velocity (GHSV = 4.6 × 10 h ) in the presence of
SO . The Fe–MOR catalysts are also characterized by H -TPR
2
2
(
temperature-programmed reduction) and Fourier transform in- lared clays (PILC) (19–24), have been investigated for SCR
frared (FT-IR) spectroscopy. TPR profiles indicate that iron cations reaction. They were reported to have various activities.
3
+
in the catalysts are present as approximately 73% Fe and 27% Among them, Fe-exchanged molecular sieves (PILC and
2
+
Fe . FT-IR spectra show that NO can be oxidized by O to N O ,
2
2
3
ZSM-5) were of special interest because their SCR activi-
ties were improved by the presence of H O and SO and
NO , and nitrate adsorbed species, and that they are bonded to the
iron cations. NH molecules adsorb on Brønsted acid sites of the
2
2
2
3
also they showed much higher SCR activities than the com-
mercial vanadia catalysts (15, 16, 22, 23). The improvement
of SCR activity by SO2 at high temperatures was attributed
to an increase in surface acidity due to sulfation of the iron
sites (16, 22, 23, 25). Sulfated Fe2O3 has been known as a
superacid (26). This would also increase NO reduction by
+
zeolite to form NH ions. NO + O is very active in reacting with
2
4
+
◦
NH ions on Fe–MOR at 300 C, but it is less active with those on
4
H–MOR. This is in good agreement with their SCR activities and
is probably related to the fast formation of NO on Fe–MOR. A
possible reaction scheme for SCR reduction is proposed. NO re-
duction involves the reaction between NO and a pair of NH4 ions
to generate an active intermediate, which then reacts with NO to NH3 at high temperatures. In addition, the sulfation of the
produce N and H O. NO oxidation to NO is the rate-determining catalysts decreased NH3 oxidation by O2, a competitive re-
step.
2
+
2
2
2
2
ꢀc 2002 Elsevier Science (USA)
action of the SCR reaction (25). Therefore, it is of interest to
Key Words: selective catalytic reduction; SCR of NO with NH ; extend the SCR investigation to other iron ion-exchanged
3
Fe3 -exchanged zeolites; Fe–MOR: Fe–HEU; Fe-beta; Fe–FER; zeolites, such as mordenite, cliniptilolite (HEU), ferrierite
+
Fe–CHA.
(
FER), beta-zeolite, and chabazite (CHA).
Great effort has been put into understanding the kinetics
and mechanism of the SCR reaction. The mechanism has
been intensively studied for vanadia catalysts and several
different mechanisms were proposed (1–3, 27–39). Most re-
searchers agree that the SCR reaction on vanadia catalysts
follows an Eley–Rideal-type mechanism, i.e., a strongly ad-
sorbed ammonia species reacting with a gaseous or weakly
adsorbed NO molecule to form molecular N2, although
which ammonia adspecies (Brønsted or Lewis) is involved
in the reaction is still under debate (2). Temperature-
programmed desorption and Fourier transform infrared
(FT-IR) experiments showed that NO does not adsorb on
INTRODUCTION
Selective catalytic reduction (SCR) of NOx (x = 1 and 2)
with NH3 is the most efficient technology for NOx abate-
ment from power plant exhaust gases. Many catalysts have
been reported to be active for this reaction. They include
1
To whom correspondence should be addressed. E-mail: yang@umich.
edu.
274
0
ꢀ
021-9517/02 $35.00
c 2002 Elsevier Science (USA)
All rights reserved.

SELECTIVE CATALYTIC REDUCTION OF NO WITH AMMONIA OVER Fe–MOR
275
the surface of oxidized vanadia under reaction conditions
2). Despite interest in the ammonia SCR over zeolite-type
catalysts, there have been few studies of the mechanism
over this type of catalyst. It was reported that both NH3 and
EXPERIMENTAL
Preparation of Catalyst
The starting materials used for preparation of the
(
NOx could adsorb on molecular sieves (5, 13, 23–25, 40–47). catalysts are as follows. Mordenites (MOR, Si/Al = 6.4,
−
In the presence of NO + O2, adsorbed NO2/NO species 10, and 45), NH4-beta (Si/Al = 12.5), and NH4–ferrierite
3
were observed by FT-IR on Fe-exchanged TiO2–PILC (45) (FER, Si/Al = 10) were obtained from Zeolyst Interna-
and ZSM-5 (40–43, 46). On Cu–ZSM-5, Kamatsu et al. (10) tional Co. Clinoptilolite (HEU, Si/Al = 5) and chabazite
reported that the reaction order was first-order with re- (CHA, Si/Al = 2) were provided by Steelhead Minerals and
spect to NO, half-order for O2, and zero-order for NH3. GSA Resources, respectively. The non-NH4-form zeolites
They proposed a mechanism for NO reduction involving were first transformed to NH4–zeolites by exchanging with
the reactions among NO, NO2, and NH3 (that was adsorbed 0.5 M NH4Cl solution (four times) at room temperature.
on copper dimer). Over H–ZSM-5, Eng and Bartholomew
Fe-exchanged zeolites were prepared by using the con-
(
48, 49) detected NO2 formation and concluded that it was ventional ion-exchange procedure. FeCl2 · 4H2O (99%)
an intermediate for the ammonia SCR reaction. In our pre- was obtained from Aldrich. Typically, 2 g of NH4-form ze-
vious work on the Fe–TiO2–PILC and Fe–ZSM-5 catalysts olite was added to 200 ml of a 0.05 M FeCl2 solution with
(
45–47), we also proposed a mechanism for NO reduction constant stirring. After 24 h, the mixture was filtered and
+
in which a pair of NH3 adsorbed species (NH4 or coor- washed five times with deionized water. The obtained cata-
◦
2+
dinated NH3) react with NO2 + NO to generate N2 and lysts were first dried at 120 C in air for 12 h, then calcined
◦
H2O. Therefore, NOx species (NO + NO2) may also play at 500 C for 6 h. Most Fe in the catalysts was oxidized
3
+
an important role in the reduction of NO by ammonia on to Fe (46). Finally, the obtained samples were pressed
molecular sieve catalysts. and ground to 60–100 mesh. The Fe and Al contents in the
Inthiswork, weinvestigatedFe-exchangedzeolites(mor- samples were measured by neutron activation analysis. The
denite, cliniptilolite, ferrierite, beta-zeolite, and chabazite) iron-exchange extent was calculated using 3× (number of
for the SCR of NO by ammonia. Effects of iron content, iron ions)/(number of aluminum atoms). Preparation of the
Si/Al ratio, H2O, and SO2 on SCR activity were stud- catalysts and the resulting Fe contents are summarized in
ied. The Fe–MOR catalysts were also characterized by Table 1. For catalyst designation, the number in parenthe-
H2-TPR (temperature-programmed reduction) and FT-IR ses after Fe indicates the Fe-exchange level and the other
spectroscopy. The adsorption of NO + O2 and NH3 on the number in parentheses shows the ratio of Si to Al; e.g.,
catalysts and the reactions among them were studied. The Fe(56)–MOR(6.4) indicates the Fe-exchange level is 56%
reaction mechanism for NO reduction is discussed.
and the Si/Al ratio is 6.4.
TABLE 1
Preparation Conditions and Iron Contents of Fe-Exchanged Mordenites
Fe content
(wt%)
Samplea
Si/Al
6.4/1
Preparation condition
Fe(56)–MOR(6.4)
2.1
Exchanging NH4–mordenite with FeCl2 solution
for 2 h at room temperature (once)
Fe(59)–MOR(6.4)
Fe(84)–MOR(6.4)
Fe(73)–MOR(10)
Fe(94)–MOR(45)
Fe(81)–HEU(5)
Fe(72)-beta(12.5)
Fe(21)–FER(10)
Fe(56)–CHA(2)
6.4/1
6.4/1
10/1
45/1
5/1
2.4
3.1
Exchanging NH4–mordenite with FeCl2 solution
for 24 h at room temperature (once)
Exchanging NH4–mordenite with FeCl2 solution
for 24 h at room temperature (3 times)
Exchanging NH4–mordenite with FeCl2 solution
for 24 h at room temperature (once)
Exchanging NH4–mordenite with FeCl2 solution
for 24 h at room temperature (once)
Exchanging NH4–clinoptilolite with FeCl2 solution
for 24 h at room temperature (once)
Exchanging NH4-beta with FeCl2 solution
for 24 h at room temperature (once)
Exchanging NH4–ferrierite with FeCl2 solution
for 24 h at room temperature (once)
Exchanging NH4–chabazite with FeCl2 solution
for 24 h at room temperature (once)
1.8
0.68
3.5
12.5/1
10/1
2/1
1.6
0.56
6.0
a
For example, Fe(56)–MOR(6.4) indicates the Fe-exchange level is 56% and Si/Al ratio is 6.4.

2
76
LONG AND YANG
Catalytic Activity Measurement
trum was recorded in flowing He and was subtracted from
the sample spectrum that was obtained at the same tem-
perature. The IR spectra were recorded by accumulating
The SCR activity measurement was carried out in a
fixed-bed quartz reactor, and 50 mg (0.065 ml) of catalyst
was used. The flue gas was simulated by blending differ-
ent gaseous reactants. The typical reactant gas composi-
tion was as follows: 1000 ppm NO, 1000 ppm NH3, 2% O2,
−
1
32 scans at a spectral resolution of 4 cm
.
During the FT-IR experiments, the gas mixtures had the
same concentrations as that used in the activity measure-
ments, i.e., 1000 ppm NO (when used), 1000 ppm NH3
5
00 ppm SO2 (when used), 2.5% water vapor (when used),
(
when used), 2% O2 (when used), and balance He. The
and balance He. The total flow rate was 500 ml/min (am-
total gas flow rate was 500 ml/min (ambient conditions).
bient conditions) and thus a very high GHSV (gas hourly
5
−1
space velocity) was obtained (4.6 × 10 h ). The premixed
gases (1.01% NO in He, 1.00% NH3 in He, and 0.99%
SO2 in He) were supplied by Matheson. Water vapor was
generated by passing He through a gas-wash bottle con-
taining deionized water. The tubings of the reactor system
were heat traced to prevent formation and deposition of
ammonium sulfite/bisulfite and ammonium nitrate. The NO
and NO2 concentrations were continually monitored by a
chemiluminescent NO/NOx analyzer (Model 42C, Thermo
EnvironmentalInstrumentsInc.). Toavoiderrorscausedby
the oxidation of ammonia in the converter of the NO/NOx
analyzer, an ammonia trap containing phosphoric acid so-
lution was installed before the sample inlet to the chemilu-
minescent analyzer. Since N2O formation was not detected
with Fe-exchanged molecular sieves in our previous work
NO Oxidation to NO2
The experiment of NO oxidation to NO2 was performed
in a fixed-bed, quartz flow reactor. A 50-mg sample was
used and the conversion at each temperature was obtained
after 1 h of steady state reaction. The reactant gas com-
position was as follows: 1000 ppm NO, 2% O2, and bal-
ance He. The total flow rate was 500 ml/min (ambient
conditions). NO concentration was continually monitored
by the chemiluminescent NO/NOx analyzer. NO conver-
sion to NO2 was obtained by using the equation NO con-
version to NO2 = (([NOx ]−[NO])/[NOx ])×100%, where
NOx represents NO + NO2.
RESULTS
(
15, 16, 22, 23), N2O or N2 was not analyzed in this work.
NO conversions were obtained using the difference of NOx
concentrations before and after reactions. All of the data
were obtained after 20 min when the SCR reaction reached
steady state.
SCR Activities on Fe-Exchanged Zeolites
The SCR activities on different Fe-exchanged zeolites
(
Fe–MOR, Fe–HEU, Fe–FER, Fe-beta, and Fe–CHA) are
summarized in Table 2. Under the conditions with a high
5
−1
space velocity (GHSV = 4.6 × 10 h ), Fe–MOR and
TPR Experiment
Fe–HEU showed high SCR activities. NO conversions (95–
H2-TPRwasperformedinafixed-bedquartzreactorwith
an inner diameter of 10 mm. In the experiment, 0.1 g of
◦
9
8%) were obtained at 400–450 C. By comparison, much
lower NO conversions were observed on Fe–FER, Fe-beta,
and Fe–CHA under the same reaction conditions. It is clear
that the structure of molecular sieves has a strong effect on
SCR activity. SCR activity can also berepresented quantita-
tively by turnover frequency (TOF) and the first-order rate
constant (k) (since the reaction is known to be first-order
with respect to NO under stoichiometric NH3 conditions
on molecular sieves (45)). TOF is defined as the number of
NO molecules converted per Fe per second. By assuming a
plug flow reactor (in a fixed bed of catalyst) free of diffusion
limitation, the rate constant (k) can be calculated from the
NO conversion (X) using
catalyst was pretreated in a flow of He (50 ml/min, purity
◦
9
9.9998%) at 400 C for 30 min to remove adsorbed species
(
e.g., H2O). After the sample was cooled to room temper-
ature in He, the reduction of the Fe–MOR was carried out
◦
from 20 to 600 C in a flow of 5.34% H2/N2 (27 ml/min)
◦
at 10 C/min. The consumption of H2 was monitored con-
tinuously with a thermal conductivity detector. The water
produced during reduction was trapped in a 5A molecular
sieve column.
FT-IR Study
Infrared spectra were recorded on a Nicolet Impact 400
FT-IR spectrometer with a TGS detector. Self-supporting
wafers of 1.3 cm in diameter were prepared by pressing
F0
k = −[
ln(1 − X),
[1]
NO]0W
2
0-mg samples and were loaded into a high-temperature
IR cell with BaF2 windows. The wafers could be pretreated where F0 is the molar NO feed rate, [NO]0 is the molar NO
in situ in the IR cell. Unless otherwise indicated, the wafers concentration at the inlet (at the reaction temperature),
◦
were first treated at 450 C in a flow of He for 15 min, and and W is the catalyst amount (g). From the NO conversions
then cooled to desired temperatures, i.e., 400, 350, 300, 200, and Fe contents or catalyst amounts in the above catalysts,
◦
1
00, and 30 C. At each temperature, the background spec- TOFs and first-order rate constants were calculated; they

SELECTIVE CATALYTIC REDUCTION OF NO WITH AMMONIA OVER Fe–MOR
277
TABLE 2
on Fe–MOR (Si/Al = 45). This result indicates that the
Fe–MOR with low Si/Al ratio is favorable to a high NO
conversion.
SCR Activities on Fe-Exchanged Zeolites
b
−2
TOFc
Temperature NO conversion k × 10
Catalysta
( C)
◦
(%)
(cm /g/s) (×10 )
3
3
Effects of H2O and SO2 on SCR Activity
Fe(59)–MOR(6.4)
350
400
85
98
92
6.6
15
10
7.7
11
12
13
15
14
BecauseresistancetoH2OandSO2 isanimportantfactor
for SCR catalysts, we studied the effects of H2O and SO2 on
catalytic performance over Fe–MOR (Fig. 2). When 2.5%
H2O was added to the reaction gases, NO conversions were
(mordenite)
450
Fe(81)–HEU(5)
clinoptilolite)
350
400
89
94
95
9.6
10
10
(
◦
decreased only slightly at 300–400 C and such decreases di-
450
minished at higher temperatures. However, the presence of
SO2 and SO2 +H2O increased NO conversion significantly
and also widened the temperature window, suggesting a
Fe(21)–FER(10)
ferrierite)
350
400
48
47
44
2.4
2.4
2.3
16
16
15
(
450
Fe(72)-beta(12.5)
350
11
30
20
0.41
1.3
0.90
2.6
7.1
4.8
promoting role by SO on Fe–MOR. Near 100% NO con-
versions were obtained at 350–450 C in the presence of
2
◦
4
4
00
50
SO2 and SO2 + H2O (Fig. 2). This observation is similar
Fe(56)–CHA(2)
350
400
12
21
27
0.45
0.89
1.3
0.75
1.3
1.7
to that on Fe-exchanged pillared clay (22, 23) and ZSM-5
(chabazite)
(
15, 16).
450
Note. Reaction conditions: 0.05
g
of catalyst; [NO] = [NH3] = H2-TPR Profiles of Fe–MOR
1
000 ppm; [O2] = 2%; He = balance; total flow rate = 500 ml/min; and
5
−1
H -TPR profiles of Fe–MOR are shown in Fig. 3. Three
GHSV = 4.6 × 10 h
.
2
a
◦
For catalyst designation, the number in parentheses after Fe indicates reduction peaks were seen, at 56, 304, and 377 C, on the
the Fe-exchange level and the other number in parentheses shows the ra-
tio of Si to Al.
b
First-order rate constant, as defined in the text, calculated using
TABLE 3
Eq. [1].
c
Overall TOF (turnover frequency) is defined as the number of NO
SCR Activities on Fe–MOR(6.4) Catalysts
molecules converted per Fe per second.
b
−2
c
Temperature NO conversion k × 10
TOF
Catalysta
( C)
◦
(%)
(cm /g/s) (×10 )
3
3
are compared in Table 2. Both NO conversion and the rate H–MOR(6.4)
constant on Fe-exchanged zeolites decreased in the follow-
ing order: Fe–MOR > Fe–HEU > Fe–FER > Fe-beta > Fe–
CHA, whereas TOF was also affected by iron content
300
4
5
5
3
2
0.13
0.18
0.19
0.12
0.087
—
—
—
—
—
3
4
4
5
50
00
50
00
(
Table 2).
Fe(56)–MOR(6.4)
Fe(59)–MOR(6.4)
Fe(84)–MOR(6.4)
300
50
400
34
85
97
92
88
1.3
6.6
13
10
9.2
5.9
15
17
16
15
3
Effect of Fe Content on SCR Activity over Fe–MOR
4
5
50
00
The effect of Fe content on SCR activity is summarized in
Table 3. H–MOR showed very low activity for the SCR re-
action. After iron was exchanged to H–MOR, SCR activity
was enhanced significantly (Table 3). NO conversion and
the rate constant increased with increasing Fe content at
low temperatures. TOF decreased with increasing Fe con-
tent in the Fe–MOR catalysts.
300
50
400
35
85
98
92
86
1.4
6.6
15
10
8.5
5.3
3
13
15
14
13
4
5
50
00
300
49
95
98
93
89
2.2
10
15
11
9.5
5.8
3
4
4
5
50
00
50
00
11
12
11
11
Effect of Si/Al Ratio on SCR Activity over Fe–MOR
It is known that the Si/Al ratio in zeolites affects their
acidity–basicity and cation exchange capacity. The effect of
Si/Al ratio on SCR activity over Fe–MOR was studied. As
shown in Fig. 1, the ratio of Si/Al had a strong effect on NO
conversion. With an increase in Si/Al ratio, NO conversion
decreased rapidly at the temperature range of 300–500 C
(
(
Note. Reaction conditions are the same as those in Table 2.
a
For catalyst designation, the number in parentheses after Fe indicates
the Fe-exchange level and the other number in parentheses shows the ra-
tio of Si to Al.
b
First-order rate constant, as defined in the text, calculated using
◦
Eq. [1].
Fig. 1). HighNOconversionswereobtainedoverFe–MOR
Si/Al = 6.4), but very low NO conversions were obtained molecules converted per Fe per second.
c
Overall TOF (turnover frequency) is defined as the number of NO

2
78
LONG AND YANG
FIG. 1. Effect of Si/Al ratio on NO conversion over Fe (59%
exchange)–MOR (Si/Al = 6.4), Fe (73% exchange)–MOR (Si/Al = 10),
and Fe (94% exchange)–MOR (Si/Al = 45). Reaction conditions: 50 mg
of catalyst, 1000 ppm NO, 1000 ppm NH3, 2% O2, balance He, and
5
−1
FIG. 3. H -TPR profiles of (a) Fe (84% exchange)–MOR (Si/Al =
GHSV = 4.6 × 10 h
.
2
6
.4), (b) Fe (59% exchange)–MOR (Si/Al = 6.4), and (c) Fe (56%
exchange)–MOR (Si/Al = 6.4).
◦
three Fe–MOR samples. The peak at 56 C is very weak,
which is probably due to reduction of adsorbed species on
the surface (e.g., O2 adsorbed species). The strong peak
◦
◦
at 304 C and its shoulder at 377 C can be attributed to the
reduction of iron species at two different sites. The intensity
of the shoulder increased significantly with iron content in
the catalysts. The TPR profiles are similar to that of Fe–
2+
ZSM-5 (46). On the Fe–ZSM-5, we found that, when Fe
ions were exchanged to ZSM-5 and calcined at 500 C in air,
◦
2
+
3+
most Fe ions were oxidized to Fe ions in the zeolite,
which was proved by an electron spin resonance signal at
g = 4.29 and a X-ray photoelectron spectroscopy band at
3
+
3+
7
11.7 eV for Fe (46). The Fe in the Fe–ZSM-5 could
2
+
◦
be partially reduced to Fe by H2/N2 at 300–600 C, but
2
+
the Fe could not be further reduced to Fe metal below
◦
7
00 C. Hence it is reasonable to deduce that most iron in
Fe–MOR was in the ferric form. Integration of the area
of the TPR peaks for Fe–MOR yielded H2/Fe molar ratios
of 0.36–0.37. These are close to the value of 0.5 required
3
+
2+
for the reduction of Fe to Fe . The lower ratio of H2
FIG. 2. Effects of H2O and SO2 on NO conversion over Fe (59%
exchange)–MOR (Si/Al = 6.4). Reaction conditions: 50 mg of catalyst,
2
+
consumption to Fe suggests that some Fe ions (≈26–28%)
◦
also existed in the Fe–MOR that was pretreated at 400 C
1
000 ppm NO, 1000 ppm NH3, 2% O2, 500 ppm SO2 (when used), 2.5%
5
−1
.
H2O (when used), balance He, and GHSV = 4.6 × 10 h
in He.

SELECTIVE CATALYTIC REDUCTION OF NO WITH AMMONIA OVER Fe–MOR
279
indicate that there are more Brønsted acid sites than Lewis
acid sites on Fe–MOR at room temperature. With increas-
−
1
ing temperature, the intensities of the 3360- and 3267-cm
bands increased first (at the expense of the 2950- and 2741-
−
1
◦
cm bands), passing through a maximum at 200–300 C,
and then decreased at higher temperatures (Figs. 4b–4e).
+
This suggests a NH transformation from a 2H structure to
4
a 3H structure with increasing temperature. The intensities
of all other bands decreased as temperature was increased
(
Figs. 4b–4e), indicating desorption of NH3. The IR bands
+
due to NH ions with the 3H structure (at 3360, 3267, and
1
This result indicates that NH ions with 3H structures are
4
−
1
◦
463 cm ) were still detected at 400 C in He (Fig. 4e).
+
4
more stable at high temperatures than the other ammonia
adsorbed species.
NH3 adsorption was also studied on H–MOR and other
◦
Fe–MOR at 300 C. As shown in Fig. 5, the IR bands due
+
+
to NH were observed. The intensities of the NH bands
4
4
decreased with an increase in iron content in the catalysts,
+
suggesting the same sites for NH4 and iron ions. A simi-
lar observation was also reported on Fe–ZSM-5 by Lobree
et al. (42).
FIG. 4. IR spectra of Fe (59% exchange)–MOR (Si/Al = 6.4) treated
◦
in flowing 1000 ppm NH3/He at 30 C for 30 min and then purged by He
◦
at (a) 30, (b) 100, (c) 200, (d) 300, and (e) 400 C.
FT-IR Studies
The IR spectra of ammonia adsorbed on Fe(59)–
MOR(6.4) at different temperatures are shown in Fig. 4. Af-
ter the sample was treated in flowing 1000 ppm NH3/He for
◦
3
0 min and then purged with He for 15 min at 30 C, a strong
−1
band at 1463 cm and weaker bands at 3360, 3267, 2950,
741, 1696, 1602, 1293, and 2600–2900 cm were observed
Fig. 4a). The broad band between 2600 and 2900 cm
−1
2
(
−
1
is attributed to physisorbed ammonia. The bands at 1696
−
1
and 1463 cm are due to the symmetric and asymmet-
+
ric bending vibrations, respectively, of NH ions that were
4
chemisorbed on the Brønsted acid sites, while the bands at
−1
1
602 and 1293 cm can be assigned to asymmetric and sym-
metric bending vibrations, respectively, of the N–H bonds
in NH3 coordinately linked to Lewis acid sites (45, 50). The
bands at higher wavenumbers are attributed to the N–H
+
stretching vibration of NH ions located at the AlO4 tetra-
4
hedra of the mordenite framework. The bands at 3360 and
−1
+
3
267 cm can be assigned to NH ions with three hydrogen
4
atoms bonded to three oxygen ions of AlO4 tetrahedra (3H
◦
FIG. 5. IR spectra (at 300 C) of (a) H–MOR (Si/Al = 6.4), (b) Fe
−1
structure), while the bands at 2950 and 2741 cm were due
(
59% exchange)–MOR (Si/Al = 6.4), and (c) Fe (84% exchange)–MOR
+
to the NH ions with two hydrogen atoms bonded to AlO4
◦
4
(Si/Al = 6.4) which were treated in flowing 1000 ppm NH3/He at 300 C
tetrahedra (2H structure) (46, 49, 51). The above results for 30 min.

2
80
LONG AND YANG
to adsorbed NO (1878 cm 1), NO2 (1617 cm−1), and ni-
−
−
1
trate (1570 cm ) species, were detected on the Fe–MOR.
Their intensities increased with iron content. By compari-
son, these bands were not observed on the H–MOR. This
result indicates clearly that the NOx species were adsorbed
on the iron sites of the Fe–MOR.
+
The IR spectra of the reaction between NH4 and
NO + O2 on Fe(59)–MOR(6.4) are shown in Fig. 8. The
fresh Fe–MOR was first treated with 1000 ppm NH3/He for
◦
3
0 min followed by a He purge at 300 C. NO +O2 was then
introduced into the sample and IR spectra were recorded as
+
a function of time. As noted above, NH ions with the 3H
4
−
1
structure (3354, 3260, and 1435 cm ) were generated af-
ter Fe–MOR was treated with NH3/He and their IR bands
did not decrease in flowing He for 30 min. After 1000 ppm
NO + 2% O2/He was passed over the sample for 1–3 min,
+
the bands attributed to NH ions decreased. At the same
4
−
1
time, two new bands were observed, at 1875 and 1622 cm
Fig. 8c). As mentioned above, they are due to NO and
NO2 adsorbed species, respectively. The 1622 cm band
may also come from adsorbed H2O generated during the
(
−
1
reaction. These results indicate that the reaction between
+
NH ions and NO + O2 occurred. With increasing time,
4
FIG. 6. IR spectra of Fe (59% exchange)–MOR (Si/Al = 6.4) treated
◦
in flowing 1000 ppm NO + 2% O2/He at 30 C for 30 min and then purged
◦
by He at (a) 30, (b) 100, (c) 200, (d) 300, and (e) 400 C.
Figure 6 shows the IR spectra of Fe(59)–MOR(6.4)
after it was treated with NO + O2 for 30 min followed by
purging with He for 15 min at room temperature. When
1
000 ppm NO + 2% O2 was introduced to the sample, a
−1
strong peak was observed at 1878 cm . This band is close
−
1
to the vibration frequency of gaseous NO (1880 cm ) and
has been assigned to Fe2 mononitrosyls [Fe (NO)] (41–
+
2+
4
3, 46). After the NOx -treated sample was purged with He,
−1
the band at 1878 cm disappeared and four bands were
observed at 1684, 1623, 1570, and 1371 cm (Fig. 6a). The
bandat1684cm isprobablyduetoadsorbedN2O3 species
because it is close to the IR band at 1690 cm for gaseous
−1
−1
−1
−1
N2O3 (52). The band at 1623 cm and those at 1570 and
−1
1
371 cm can be assigned to adsorbed NO2 and nitrate
species, respectively (41–43, 45, 46, 53, 54). The above ob-
servation suggests that some NO molecules were oxidized
to NOx by O2. When increasing temperature, the bands at
−
1
1
684 and 1570 cm decreased significantly, whereas those
−
1
at 1623 and 1371 cm increased first, and then decreased at
◦
−1
above 100 C. The N2O3 band at 1684 cm and the nitrate
−1
◦
band at 1371 cm disappeared at 200 C (Fig. 6c). All of
FIG. 7. IR spectra of (a) H–MOR (Si/Al = 6.4), (b) Fe (59%
exchange)–MOR (Si/Al = 6.4), and (c) Fe (84% exchange)–MOR
◦
the other bands vanished at 400 C (Fig. 6e).
The IR spectra of H–MOR and Fe–MOR in flowing
(
Si/Al = 6.4) which were treated in flowing 1000 ppm NO + 2% O2/He
◦
◦
NO + O2 at 200 C are shown in Fig. 7. Three bands, due at 200 C for 30 min.

SELECTIVE CATALYTIC REDUCTION OF NO WITH AMMONIA OVER Fe–MOR
281
DISCUSSION
The present work indicated that Fe-exchanged zeolites
were active for the SCR of NO with NH3. The structure of
zeolites had a strong effect on SCR activity. Under the same
reaction conditions, SCR activity decreased in the seque-
nce of Fe–MOR > Fe–HEU (clinoptilolite) > Fe–FER >
Fe-beta > Fe–CHA. The Fe–MOR and Fe–HEU showed
very high SCR activities, whereas the other catalysts were
less active. For the Fe–MOR catalysts, NO conversion in-
creased with iron content (Table 3). Since H–MOR showed
very low NO conversions under the same conditions (at a
high space velocity), it is clear that iron ions play a crucial
role in the SCR reaction. Moreover, SCR activity on Fe–
MOR was improved by the presence of SO2 and SO2 +H2O
(
Fig. 1). This observation is in agreement with our previ-
ous results obtained on Fe-exchanged pillared clay (22, 23)
and ZSM-5 (15, 16). Nearly 100% NO conversion was ob-
◦
tained at 350–450 C with SO2. This was close to the SCR
performance obtained on Fe–ZSM-5 and was much higher
than that of the commercial V2O5 + WO3/TiO2 catalyst
◦
(
(
67% NO conversion at 400 C) under the same conditions
15, 16). In the presence of SO2, the surface of the catalysts
was sulfated and the sulfate species bonded to iron sites
◦
FIG. 8. IR spectra taken at 300 C on passing 1000 ppm NO + 2%
O2/He over NH3 presorbed on Fe (59% exchange)–MOR (Si/Al = 6.4)
for (a) 0, (b) 1, (c) 3, and (d) 5 min.
+
the NH4 bands were further decreased and vanished at
5
min, while only NO and NO2 adsorbed species were de-
+
tected (Fig. 8d). In contrast, the NH ions on H–MOR was
4
much less active in reacting with NO + O2 than those on
Fe–MOR. When 1000 ppm NO + 2% O2 was passed over
◦
the H–MOR that was treated with NH3 at 300 C, the inten-
+
sities of the NH bands decreased and vanished at 30 min
4
(
Fig. 9). This result is consistent with their relative SCR
activities.
To identify the species present on the catalyst under
reaction conditions, IR spectra were recorded when Fe–
◦
MOR was heated from 300 to 450 C in a flow of 1000 ppm
NO + 1000 ppm NH3 + 2% O2 at steady state. As shown in
+
Fig. 10, the bands due to NH ions were observed at 3357,
4
−1
◦
3
258, and 1433 cm at 300 C. Raising the temperature re-
sulted in a decrease in the intensity of the NH bands. The
NH ions were still observed at 450 C. In the temperature
+
4
+
◦
4
range studied, the IR bands due to NOx adsorbed species
were not detected. This suggests that the NOx consumption
rate(byNH3)wasmuchhigherthantheNO2 formationrate
from NO + O2 under the reaction conditions. The observa-
◦
FIG. 9. IR spectra taken at 300 C on passing 1000 ppm NO + 2%
+
tion of strong NH bands also indicates that NH3 was in
4
O2/He over NH3 presorbed on H–MOR (Si/Al = 6.4) for (a) 0, (b) 1,
excess on the surface.
(c) 5, (d) 15, and (e) 30 min.

2
82
LONG AND YANG
+
the main cage) first, and then exchanged with the NH in
4
another cage when this cage was full. The reduction tem-
3
+
peratures of the Fe ions in the two cages were different,
possibly due to the different diffusion rates in the two cages
for H2 (with a kinetic diameter of 0.29 nm). The iron cations
provide sites for NOx adsorption, as discussed below.
The FT-IR spectra indicated that adsorbed NO, N2O3,
NO2, and nitrate species were formed after Fe–MOR cata-
lysts were treated with NO + O2 (Figs. 6 and 7). This is
similar to the results obtained on Fe–ZSM-5 (41–43, 46),
but different from that obtained on vanadia-based catalysts.
Most researchers agreed that nitrogen oxides do not adsorb
on fully oxidized vanadia (1–3). Since NOx species were
not detected on H–MOR (Fig. 7), they were bonded to iron
sites in the Fe–MOR. NO was reported to adsorb on the
2
+
Fe sites and the other NOx species were adsorbed on the
3
+
Fe sites (43, 46, 57, 58). NO adsorption was very weak.
It disappeared when the sample was purged with He. The
NO2 and nitrate species came from the reaction between
oxygen adsorbed species and gaseous NO (43). The NO2
◦
was the dominant species at 200–300 C. The concentrations
of the NOx adsorbed species increased with iron content in
the catalysts (Fig. 7). Since Fe–MOR catalysts were much
more active than H–MOR for the SCR reaction (Table 3), it
is clear that the NOx adsorbed species played an important
role in NO reduction.
FIG. 10. IR spectra of Fe (59% exchange)–MOR (Si/Al = 6.4) in
When Fe–MOR was treated in flowing NH /He, NH3
molecules were adsorbed on the Brønsted acid sites to
3
flowing 1000 ppm NH3 + 1000 ppm NO + 2% O2/He at (a) 300, (b) 350,
◦
(
c) 400, and (d) 450 C.
were formed under the reaction conditions (23, 46). The sul-
fation increased the surface acidity (mainly Brønsted acid-
ity) and decreased the activity for ammonia oxidation by O2
so that more NH3 could be used to reduce NO rather than
react with O2 (25). Therefore, SCR activity was enhanced.
◦
At 350–400 C, the Fe–MOR catalyst was stable and did
not decrease its SCR activity during 9 h on stream in the
presence of SO2 (8).
Although iron cations were exchanged with NH4–MOR
2
+
as Fe form, the TPR results indicated that most iron
3
+
cations (≈72–74%) in the Fe–MOR were present as Fe
ions. Oxygen oxidized some Fe2 to Fe during calcina-
+
3+
tion of the catalysts. Two TPR peaks were observed on Fe–
MOR (Fig. 3), suggesting that Fe3 ions occupied two sites
in the catalysts. As shown in Fig. 11, mordenite has a one-
dimensional channel system bounded by twelve-member
rings (main cage, 0.65 × 0.70 nm) with cross-links bounded
by eight-member rings (0.26 × 0.57 nm) (55). The positions
of cations are located mainly at the two cages (55, 56). For
instance, one is near the center of the eight-member ring
that connects the main cage with the smaller channel and
the other is in the main cage far from the four-member
+
ring. When Fe2 ions were exchanged with the NH in the
+
+
4
mordenite, they preferred to occupy one cage (probably
FIG. 11. Framework topology of mordenite.

SELECTIVE CATALYTIC REDUCTION OF NO WITH AMMONIA OVER Fe–MOR
283
+
form NH4 ions (Fig. 4). Although coordinated NH3 was activity for oxidizing NO to NO2 under the conditions of
5
−1
also detected, its concentration was much lower than that 1000 ppm NO, 2% O2, and GHSV = 4.6 × 10 h . When
+
+
4
of NH . The NH ions are present in two forms: a 3H iron ions were exchanged to H–MOR, the NO conversions
4
structure (three hydrogen atoms bonded to the AlO4 tetra- to NO2 were increased significantly, from 2–3% to 10–13%
◦
◦
hedra) and a 2H structure (two hydrogen atoms bonded at 300–400 C. At above 450 C, NO conversion was close
+
to the AlO4 tetrahedra). NH ions with the 3H structure to the thermal equilibrium value. We have verified that the
are more stable than those with the 2H structure at high formation of NO2 will promote the reaction between NH4
temperatures and they were still observed at 400 C in He ions and NO on Fe–ZSM-5. The reaction between NH and
4
+
◦
+
4
+
+
(
Fig. 4). This also indicates that NH ions were more stable NO + NO2 is faster than those between NH + NO2 and
4
4
+
than NOx adsorbed species at high temperatures because between NH + NO under the conditions with the same
4
◦
NOx desorbed below 400 C (Fig. 6). As expected, the re- NOx concentration (59). In addition, under the reaction
+
placement of NH by iron ions decreased the concentration conditions, i.e., with NO, O2, and NH3 at high tempera-
of NH (Fig. 5). The NH ions on the Fe–MOR were active tures, the surface of Fe–MOR was covered mainly by NH4
4
+
+
+
4
4
in reacting with NO+O2. They were consumed completely ions. NOx adsorbed species were not observed (Fig. 10).
◦
+
when introducing NO + O2 in 5 min at 300 C (Fig. 8). By This indicates that the reaction of NH ions with NOx was
4
comparison, when passing NO + O2 over the ammonia ad- much faster than that of NOx formation from NO + O2.
+
+
sorbed H–MOR, the NH bands vanished after 30 min. As soon as NOx was generated, it was consumed by NH4
4
This suggests that Fe3 in Fe–MOR promoted the reaction ions so that the steady state concentration of NOx was be-
+
+
between NO + O2 and NH ions, probably due to an in- low the detection limit of our IR spectrometer. This sug-
4
crease in NOx formation (Fig. 7). This phenomenon was gests that NO oxidation to NO2 is a slow step for the SCR
in good agreement with our SCR results showing that Fe– reaction.
MOR was much more active than H–MOR. The presence
According to the above results, it seems that the NO
of iron cations in the zeolite increased NO oxidation to reduction path on the Fe–MOR is similar to that on Fe-
NO2 by O2. As shown in Fig. 12, H–MOR showed very low exchanged pillared clay (45) and Fe–ZSM-5 (59). The SCR
reaction most probably takes place according to the follow-
ing steps:
+
+
,
4(a)
2
NH3(g) + 2H –2NH
[2]
[3]
NO(g) + 1/2O2(a)–NO2(a),
NO2(a) + 2NH4 –NO2(NH )2(a),
+
+
4
[4]
(a)
+
+
NO2(NH )2(a) + NO(g)–· · · → 2N2(g) + 3H2O(g) + 2H .
4
[
5]
This is a somewhat simplified mechanism. During the SCR
reaction, gaseous NH3 molecules are adsorbed quickly
+
on the Brønsted acid sites to form NH ions. Also, NO
4
3
+
molecules are oxidized to NO2 on Fe sites by O2. Then
+
one molecule of NO2 diffuses to two adjacent NH ions
4
+
to form an active complex NO2(NH )2. The active com-
4
plex subsequently reacts with one molecular NO to pro-
duce N2 and H2O and thus completes the catalytic cycle.
The oxidation of NO to NO2 is probably the rate-
determining step. H–MOR showed strong surface acidity
but very low activity for the oxidation of NO to NO2;
its SCR activity was therefore very low. After it was ex-
changed with iron, the iron ions increased the oxidation
rate of NO to NO2 significantly and hence a much higher
SCR activity was expected on Fe–MOR. It is noted that al-
−
1
though the nitrate species with an IR band near 1570 cm
FIG. 12. Oxidation activity of NO to NO2 by O2 on H–MOR (Si/Al =
.4) and Fe (59% exchange)–MOR (Si/Al = 6.4) under the following con-
was also observed on the fresh Fe–MOR when introduc-
ing NO + O2, it might not be an intermediate for the SCR
reaction. This is because the nitrate species comes from ox-
idation/disproportionation of NO2, but NO2 is consumed
6
ditions: 50 mg of catalyst, 1000 ppm NO, 2% O2, and 500 ml/min total flow
rate. The empty-tube conversion under the same conditions was below
◦
1
.5% at 300–450 C.

2
84
LONG AND YANG
+
very quickly by NH4 ions during the SCR reaction. It is
difficulttofurtherconvertNO2 tonitratespeciesinthepres-
ence of NH3. Additionally, mordenite with a low Si/Al ratio
will have strong Brønsted acidity (which is favorable for
NH3 adsorption) and large ion-exchange capacity (which
7. Wark, M., Bruckner, A., Liese, T., and Grunert, W., J. Catal. 175, 48
1998).
. Ham, S. W., Nam, I. S., and Kim, Y. G., Korean J. Chem. Eng. 17, 318
2000).
. Mishima, H., Hashmoto, K., Ono, T., and Anpo, M., Appl. Catal. B 19,
19 (1998).
(
8
9
(
1
increases iron content in Fe–MOR and thus enhances NO 10. Komatsu, T., Nunokawa, M., Moon, I. S., Takahara, T., Namba, S., and
Yashima, T., J. Catal. 148, 427 (1994).
oxidation to NO2). Hence, SCR activity increased with a
1
1
1. Kieger, S., Delahay, G., Coq, B., and Neveu, B., J. Catal. 183, 267
1999).
decrease in the Si/Al ratio of Fe–MOR (Fig. 1). In ad-
(
+
dition, NO reduction needs NH4 pairs (reaction [2]). A
2. Kiovsky, J. R., Koradia, P. B., and Lim, C. T., Ind. Eng. Chem. Prod.
Res. Dev. 19, 218 (1980).
small pore diameter in zeolites will promote the formation
+
of NH pairs and thus be beneficial to NO reduction. How- 13. Amiridis, M. D., Puglisi, F., Dumesic, J. A., Millman, W. S., and Topsøe,
4
N.-Y., J. Catal. 142, 572 (1993).
ever, a small pore size clearly decreases pore diffusion rate
1
4. Komatsu, T., Uddin, M. A., and Yashima, T., in “Zeolites: A Refined
Tool for Designing Catalysis Sites” (L. Bonneviot and S. Kaliaguine,
Eds.), p. 437. Elsevier, Amsterdam, 1995.
5. Long, R. Q., and Yang, R. T., J. Am. Chem. Soc. 121, 5595 (1999).
16. Long, R. Q., and Yang, R. T., J. Catal. 188, 332 (1999).
and thus decreases SCR activity. Therefore, an appropriate
pore diameter in zeolites, e.g., 0.5–0.7 nm in ZSM-5, mor-
denite, and clinoptilolite, seems to be good for high SCR
activity.
1
1
1
7. Ma, A. Z., and Grunert, W., Chem. Commun. 71 (1999).
8. Sun, Q., Gao, Z. X., Chen, H, Y., and Sachtler, W. M. H., J. Catal. 201,
CONCLUSIONS
89 (2001).
1
2
9. Yang, R. T., Chen, J. P., Kikkinides, E. S., Cheng, L. S., and
Cichanowicz, J. E., Ind. Eng. Chem. Res. 31, 1440 (1992).
0. Chen, J. P., Hausladen, M. C., and Yang, R. T., J. Catal. 151, 135
Based on the above results, it can be concluded that
Fe-exchanged mordenite and clinoptilolite were highly ac-
tive for the ammonia SCR reaction, whereas Fe-exchanged
beta, ferrierite, and chabazite were less active. SCR activity
was improved by the presence of SO2 and SO2 + H2O. Near
(
1995).
2
2
2
1. Cheng, L. S., Yang, R. T., and Chen, N., J. Catal. 164, 70 (1996).
2. Long, R. Q., and Yang, R. T., Catal. Lett. 59, 39 (1999).
3. Long, R. Q., and Yang, R. T., J. Catal. 186, 254 (1999).
1
00% NO conversion was obtained on Fe–MOR at a high 24. Long, R. Q., and Yang, R. T., J. Catal. 196, 73 (2000).
5
−1
2
5. Long, R. Q., Chang, M. T., and Yang, R. T., Appl. Catal. B 33, 97 (2001).
space velocity (GHSV = 4.6 × 10 h ). The iron cations in
the catalysts were present mainly as Fe3 ions. Adsorbed
NO, N2O3, NO2, and nitrate species were observed when in-
troducing NO + O2 to Fe–MOR. They were bonded to the
iron sites. NH3 molecules were adsorbed on Brønsted acid
+
26. Yamaguchi, T., Appl. Catal. 61, 1 (1990).
2
2
2
3
7. Miyamoto, A., Kobayashi, K., Inomata, M., and Murakami, Y., J. Phys.
Chem. 86, 2945 (1982).
8. Janssen, F. J. J. G., van den Kerkhof, F. M. G., Bosch, H., and Ross,
J. R. H., J. Phys. Chem. 91, 5921 (1987).
9. Odriozola, J. A., Heinemann, H., Somorjai, G. A., Garcia de la Banda,
J. E., and Pereira, P., J. Catal. 119, 71 (1989).
+
+
sites to form NH ions. The NH ions on Fe–MOR were
4
4
active in reacting with NO + O2 at high temperatures, but
0. Ramis, G., Busca, G., Bregani, F., and Forzatti, P., Appl. Catal. 64, 259
those on the H–MOR were less active. The present results
(
1990).
+
indicate that NH , NO, and NO2 species play an important
4
31. Schramlmarth, M., Wokaun, A., andBaiker, A., J. Catal. 124, 86(1990).
role in the SCR reaction. A possible reaction scheme for 32. Chen, J. P., and Yang, R. T., J. Catal. 124, 411 (1990).
3
3. Went, G. T., Leu, L. J., Rosin, R. R., and Bell, A. T., J. Catal. 134, 492
NO reduction was proposed involving NO2 as an interme-
diate.
(
1992).
34. Ozkan, U. S., Cai, Y., and Kumthekar, M. W., J. Catal. 149, 375 (1994).
5. Ozkan, U. S., Cai, Y., and Kumthekar, M. W., J. Catal. 149, 390
3
(
1994).
ACKNOWLEDGMENT
3
6. Odenbrand, C. U. I., Bahamonde, A., Avila, P., and Blanco, J., Appl.
Catal. B 5 117 (1994).
This work was supported by the NSF under Grant CTS-0095909.
3
3
7. Topsøe, N.-Y., Topsøe, H., and Dumesic, J. A., J. Catal. 151, 226 (1995).
8. Topsøe, N.-Y., Dumesic, J. A., and Topsøe, H., J. Catal. 151, 241
(1995).
REFERENCES
39. Ramis, G., Yi, L., Busca, G., Turco, M., Kotur, E., and Willey, R. J.,
1
2
. Bosch, H., and Janssen, F., Catal. Today 2, 369 (1988).
. Busca, G., Lietti, L., Ramis, G., and Berti, F., Appl. Catal. B 18, 1 40. Chen, H.-Y., Voskoboinikov, T. V., and Sachtler, W. M. H., J. Catal.
J. Catal. 157, 523 (1995).
(1998).
180, 171 (1998).
3
4
. Forzatti, P., and Lietti, L., Heterogeneous Chem. Rev. 3, 33 (1996).
. Ito, E., Hultermans, R. J., Lugt, P. M., Burgers, M. H. W., Rigutto,
41. Chen, H.-Y., Voskoboinikov, T. V., and Sachtler, W. M. H., J. Catal.
186, 91 (1999).
M. S., van Bekkum, H., and van den Bleek, C. M., Appl. Catal. B 4, 95 42. Lobree, L. J., Hwang, I.-C., Reimer, J. A., and Bell, A. T., J. Catal. 186,
(1994).
242 (1999).
5
6
. Ito, E., Mergler, Y. J., Nieuwenhuys, B. E., Calis, H. P. A., van Bekkum,
H., and van den Bleek, C. M., J. Chem. Soc., Faraday Trans. 92, 1799
43. Lobree, L. J., Hwang, I.-C., Reimer, J. A., and Bell, A. T., Catal. Lett.
63, 233 (1999).
44. Feng, X., and Hall, W. K., J. Catal. 166, 368 (1997).
45. Long, R. Q., and Yang, R. T., J. Catal. 190, 22 (2000).
46. Long, R. Q., and Yang, R. T., J. Catal. 194, 80 (2000).
(1996).
. Kooten, W. E. J., Kaptein, J., van den Bleek, C. M., and Calis, H. P. A.,
Catal. Lett. 63, 227 (2000).

SELECTIVE CATALYTIC REDUCTION OF NO WITH AMMONIA OVER Fe–MOR
285
4
4
4
5
5
7. Long, R. Q., and Yang, R. T., J. Catal. 198, 20 (2001).
8. Eng, J., and Bartholomew, C. H., J. Catal. 171, 14 (1997).
9. Eng, J., and Bartholomew, C. H., J. Catal. 171, 27 (1997).
0. Topsøe, N.-Y., J. Catal. 128, 499 (1991).
1. Teunissen, E. H., van Santen, R. A., Jansen, A. P. J., and van
Duijneveldt, F. B., J. Phys. Chem. 97, 203 (1993).
54. Valyon, J., and Hall, W. K., J. Phys. Chem. 97, 1204 (1993).
55. Szostak, R., “Handbook of Molecular Sieves.” Van Nostrand,
Princeton, NJ, 1992.
56. Barthomeuf, D., Catal. Rev.–Sci. Eng. 38, 521 (1996).
57. Segawa, K. I., Chen, Y., Kubsh, J. E., Delgass, W. N., Dumesic, J. A.,
and Hall, W. K., J. Catal. 76, 112 (1982).
5
5
2. Laane, J., and Ohlsen, J. R., Prog. Inorg. Chem. 27, 465 (1980).
3. Li, Y., and Armor, J. N., J. Catal. 150, 388 (1994).
58. Guglielminotti, E., J. Phys. Chem. 98, 9033 (1994).
59. Long, R. Q., and Yang, R. T., J. Catal. 207, 224 (2002).