Enhancement of hydrothermal stability of Cu-ZSM5 catalyst for NO decomposition

Article abstract of DOI:10.1134/S0023158406050119

RECu-ZSM5 samples (RE = La, Sm or Ce) have been prepared by ion-exchange of H-ZSM5 with rare earth and copper ions in an aqueous solution both simultaneously and consecutively (RE before copper), and their physico-chemical and catalytic properties in NO decomposition have been compared with those of a Cu-ZSM5 with a similar copper content. The catalysts have been characterized by N₂ adsorption at 77 K (BET) and XRD before and after aging cycles under wet (2-2.5 vol % H₂O) conditions at 450-500°C. NO decomposition tests have been carried out in a fixed bed reactor at 450°C under dry or wet conditions on both fresh and aged catalysts. NO adsorption tests have been performed at 120°C on the same samples. The addition of RE does not affect the crystalline structure, surface area, and porosity but strongly enhances the catalytic activity and the hydrothermal stability of Cu-ZSM5, which, on the contrary, is completely deactivated under wet conditions. The two-step exchanged catalysts provide better performances likely due to the larger RE exchange level obtained with this method. A unique linear correlation between the amount of N₂O produced in the NO adsorption tests, related to the reoxidation of prereduced Cu⁺ sites, and the TOF, estimated from the catalytic activity tests, has been found for both fresh and aged catalysts.

Full text of DOI:10.1134/S0023158406050119

ISSN 0023-1584, Kinetics and Catalysis, 2006, Vol. 47, No. 5, pp. 728–736. © MAIK “Nauka /Interperiodica” (Russia), 2006.
Published in Russian in Kinetika i Kataliz, 2006, Vol. 47, No. 5, pp. 751–759.
Enhancement of Hydrothermal Stability of Cu-ZSM5 Catalyst
1
for NO Decomposition
a
b
b
a
c
B. I. Palella , L. Lisi , R. Pirone *, G. Russo , and M. Notaro
a
Dipartimento di Ingegneria Chimica Universita di Napoli Federico II; P.le V. Tecchio, Napoli, 80-80125 Italy
b
Istituto di Ricerche sulla Combustione, CNR; P.le V. Tecchio, Napoli, 80-80125 Italy
c
Centro Elettrotecnico Sperimentale Italiano, Via R. Ribattino, 54-Milano, Italy
*
e-mail: pirone@irc.na.cnr.it
Received April 8, 2005
Abstract—RECu-ZSM5 samples (RE = La, Sm or Ce) have been prepared by ion-exchange of H-ZSM5 with
rare earth and copper ions in an aqueous solution both simultaneously and consecutively (RE before copper),
and their physico-chemical and catalytic properties in NO decomposition have been compared with those of a
Cu-ZSM5 with a similar copper content. The catalysts have been characterized by N adsorption at 77 K (BET)
2
and XRD before and after aging cycles under wet (2–2.5 vol % H O) conditions at 450–500°C. NO decompo-
2
sition tests have been carried out in a fixed bed reactor at 450°C under dry or wet conditions on both fresh and
aged catalysts. NO adsorption tests have been performed at 120°C on the same samples. The addition of RE
does not affect the crystalline structure, surface area, and porosity but strongly enhances the catalytic activity
and the hydrothermal stability of Cu-ZSM5, which, on the contrary, is completely deactivated under wet con-
ditions. The two-step exchanged catalysts provide better performances likely due to the larger RE exchange
level obtained with this method. A unique linear correlation between the amount of N O produced in the NO
2
+
adsorption tests, related to the reoxidation of prereduced Cu sites, and the TOF, estimated from the catalytic
activity tests, has been found for both fresh and aged catalysts.
DOI: 10.1134/S0023158406050119
1
INTRODUCTION
The catalytic decomposition of NO into N and O
ysis, concluded that the deactivation observed under
real conditions was not due to the degradation of the
zeolitic framework but to sintering of copper ions to
2
2
would be a very effective solution for nitrogen oxide
abatement since no addition of a reducing agent is
required [1, 2]. The breakthrough in the investigation of
NO decomposition occurred when Iwamoto et al. found
that copper-exchanged zeolites, in particular Cu-
ZSM5, exhibit a rather high activity for this reaction [3,
form CuO and Cu O, which grow primarily inside the
2
zeolite, with a partial loss of micropores volume.
According to the findings ofYan et al. [12], highlighted
by XRD and BET measurements, a significant destruc-
tion of the zeolite framework is not detectable even for
a catalyst that underwent 50% deactivation. Neverthe-
less, they showed that the aging of a fresh catalyst con-
4
]. Later on, it was found that Cu-ZSM5 is also very
2
+
active in other reactions of nitrogen oxide abatement,
taining Cu ions in two different coordination environ-
such as the selective reduction to N with hydrocarbons
2
ments resulted in the formation of CuO particles and
2
+
in the presence of oxygen [2, 5] and the decomposition
the exchange sites, thus vacated by Cu , were occupied
by protons with consequently easier framework dealu-
mination of H-ZSM5 [10].
of N O [6, 7].
2
Since then, the Cu-ZSM5 catalyst has been the
object of a lot of investigations mainly aiming to char-
The deactivation of Cu-ZSM5 in the presence of
acterize its properties and/or understand the reaction steam was also studied through TEM/EDX, XRD, and
mechanism, as recently reviewed [2, 8]. However, the XPS by Zhang and Flytzani-Stephanopoulos [13], who
strong limitation of Cu-ZSM5 as a potential catalyst for suggested a twofold mechanism: (I) the reversible
applications in de-NO processes is the very poor dura- hydration of copper sites; (II) the irreversible migration
x
bility in the presence of steam at high temperature. The and aggregation of copper sites to form inactive CuO
cause of the deactivation is still debated [9–17], being crystallites. The migration of copper ions, responsible
its phenomenology strictly related to the operating con- for the weak hydrothermal stability of Cu-MFI zeolites,
ditions of testing (temperature, water vapor pressure, was also investigated via DFT calculation by Rice et al.
time of exposition to wet mixtures). Kharas et al. [9], [18] who found that the genesis of CuO clusters is ther-
through a combined XRD, EXAFS, and catalytic anal- modynamically favored by the Madelung stabilization
energy characterizing the formation of the bulk oxide
phase. Moreover, they also demonstrated that the driv-
1
The text was submitted by the authors in English.
7
28

ENHANCEMENT OF HYDROTHERMAL STABILITY OF Cu-ZSM5 CATALYST
729
+
–
2+
– –
ing force for demetalation of Cu Z and Cu (OH) Z is we have investigated the effect of three different rare-
2
+
– –
significantly greater than that of Co (OH) Z , thus earth cations (Sm, Ce, and La) on ZSM5 in this work.
explaining the better stability of Co-ZSM5 with respect The preparation method, the nature of rare-earth cat-
to Cu-ZSM5 experimentally observed in the HC-SCR ions, the morphological features and the redox proper-
of NO [19–21]. The migration of copper ions without ties of copper sites before and after aging treatments
the destruction of the zeolite lattice upon hydrothermal were also studied in order to determine the influence of
treatment was also proposed by Kucherov et al. [11, 14], rare-earth cations on the catalytic activity and hydro-
Iwamoto et al. [15], Quincoces et al. [16] and Gomez thermal stability in the decomposition of NO.
et al. [17].
Further studies of Cu-ZSM5 in the NO decomposi-
EXPERIMENTAL
tion were focused to improve the catalytic perfor-
Preparation of Catalysts
mances. The addition of a second cation seems to be
one of the most promising ways to increase the catalyst
activity under dry conditions [22–29], but among all
these studies, only a few contributions were devoted to
the enhancement of the hydrothermal resistance [11,
The Cu-ZSM5 sample was prepared by ion-
exchange of a commercial H-ZSM5 zeolite (Zeolyst
CBU–5020, Si/Al = 25, and BET specific surface area =
2
4
00 m /g) in an aqueous solution at 50°C for 2 h with a
1
3, 29, 30]. A possible way to stabilize Cu-ZSM5
2
0 mM solution of copper(II) acetate monohydrate
towards the deactivating effect of water vapor at high
temperatures could be the insertion of large rare-earth
cations [11, 13, 29] following an approach used to
improve the resistance of Ni-Y zeolites in the cracking
of PE-derived waste oil [31, 32]. As a matter of fact,
Masuda et al. [32] demonstrated that the presence of
RE elements resulted in constant activity during regen-
eration and reaction cycles over Ni-Y zeolite, while
MFI zeolites are almost completely deactivated due to
the dealumination.
(
Aldrich purity 99.8%). After ion exchange, the sample
was centrifuged, washed twice with double-distilled
water, and dried at 120°C overnight. Finally, the Cu-
exchanged sample was calcined according to the fol-
lowing procedure: (a) heating under helium flow at
5
K/min up to 550°C, keeping this temperature for 2 h;
(
b) cooling to room temperature under He flow;
(
c) switching the gas flow to 1% O /He and heating
2
5
K/min up to 550°C, keeping this temperature for 4 h.
Two series of rare-earth containing Cu-ZSM5 were
Among rare earths, the role of samarium as a pro- also prepared:
moter for Cu-ZSM5 has been deeply studied by Par-
—
co-exchanged samples were obtained by simul-
vulescu et al. [25–27]. They found that Sm ions
strongly enhance the catalytic activity of Cu-ZSM5
when co-exchanged in small amounts (below 0.20 wt %)
with copper, increasing the T.O.F. (turnover frequency)
of NO decomposition by 100% with respect to un-
taneously exchanging the parent zeolite at 50°C with a
2+
solution containing both copper ([Cu ] = 20 mM) and
rare-earth cation with an atomic ratio of Cu/RE (RE =
La, Sm, and Ce) equal to 12 as proposed by P aˆ rvulescu
et al. for Sm co-exchanged Cu-ZSM–5 catalysts [26].
Drying and calcination were performed as described
above. These samples will be indicated as CuRE-Z
doped CuZSM5. Their NO-DRIFT, O -TPD, and XPS
2
results revealed that the presence of samarium
improves the dispersion of copper ions during the ion
exchange with consequent enhancement of both redox
(
where Z = ZSM5 and RE = rare-earth cation);
+
—two-step-exchanged samples were obtained by
properties of Cu ions and mobility of extra-lattice oxy-
first exchanging the rare-earth cation and subsequently
the copper cation. The experimental conditions for the
first ion exchange were chosen in order to maximize
rare-earth loading according to literature data [25–28,
gen whose release was proposed to be the rate-deter-
mining step in the decomposition of NO [33]. Never-
theless, the hydrothermal stability of the promoted cat-
alyst was not investigated.
3
4–36]. More specifically, the samples containing lan-
On the other hand, Zhang and Flytzani-Stephanopo-
ulos [13, 29] first claimed that cerium prevents the
thanum and samarium were exchanged at 95°C for 5 h
in a 20 mM solution of the rare-earth salt, while the
cerium-containing zeolite was obtained exchanging in
a 8-mM solution for 3 h at 80°C. These catalysts will be
indicated as RECu-Z. Lanthanum(III) nitrate hydrate
2
+
migration of Cu , which is responsible for the forma-
tion of inactive CuO clusters in the deactivated cata-
lysts. It was also shown that cerium ions inserted in the
zeolite provide a better copper dispersion, contributing
to the enhancement of the catalytic performance. In
agreement with these findings, the ESR results of
Kucherov et al. [14] revealed that the effect of rare-
earth cation as lanthanum or cerium resulted in a signif-
(
99.9%), cerium(III) nitrate hydrate (99.999%), and
samarium(III) nitrate hexahydrate (99.9%), all supplied
by Aldrich, were used as precursor salts for rare-earth
exchange.
2
+
icant slackening of the coordination of Cu ions after
severe steam-aging treatments (20 vol % H O at 630°C
Catalytic Tests
2
for 17 h).
A fixed-bed microreactor consisting of a 60-cm-
In order to improve the performances of long quartz tube (I.D. 1.0 cm), supplied with a porous
Cu-exchanged ZSM–5 catalysts in NO decomposition, disk supporting the catalyst particles (d = 200–
KINETICS AND CATALYSIS Vol. 47 No. 5 2006

7
30
PALELLA et al.
Table 1. Chemical composition, surface area, and pore volume (V , total; V , micropores) of fresh and aged samples
p
m
FRESH
AGED
SAMPLE
V ,
cm g
V ,
cm g
V ,
cm g
V ,
cm g
RE loading, Cu loading, BET surface
p
m
Cu loading, BETsurface
p
m
2
–1
2
–1
3
–1
3
–1
3
–1
3
–1
wt %
wt %
area, m g
wt %
area, m g
Cu-Z
–
1.69
1.82
1.75
1.63
1.81
2.16
1.93
365
369
347
383
368
362
362
0.32
0.35
0.39
0.47
0.36
0.41
0.34
0.17
0.17
0.17
0.18
0.17
0.18
0.17
1.69
1.76
1.58
1.61
1.89
2.01
1.91
334
354
359
336
367
360
358
0.31
0.30
0.43
0.43
0.35
0.46
0.42
0.17
0.17
0.18
0.16
0.16
0.18
0.17
CuLa-Z
CuSm-Z
CuCe-Z
LaCu-Z
SmCu-Z
CeCu-Z
0.09
0.11
0.08
0.16
0.75
0.12
2
+
3+
3+
3+
4
00 µm), inserted in an electrical furnace (LENTON) ratios of Cu /2Al and RE /3Al , respectively,
characterized by three heated and temperature-con- according to a procedure discussed elsewhere [37].
trolled zones, was used to perform the experiments. The
temperature of the catalytic bed was measured by a
Chromel-Alumel thermocouple placed in another
quartz tube coaxially inside the reactor. Mass flow con-
trollers (BROOKS 5850S) were used to control and
measure the flow rates of high purity gases: 1 vol %
XRD patterns at room temperature were collected
using a PW 1100 Philips XRD 6000 diffractometer
with CuK radiation.
α
BET surface areas were measured by N adsorption
2
at 77 K with a Carlo Erba 1900 Sorptomatic.
In a typical NO adsorption experiment [41, 42], a
NO/He (NO impurity about 150 ppm) and pure He
2
–
1
6
00 ppm NO/He gas mixture (20 l h ) was fed to the
(
99.995%). Before each experiment, the catalyst was
catalytic reactor at 120°C. The transient behavior of the
system was recorded by analyzing the inlet and the out-
let reactor gas composition (the concentrations of NO,
treated under He (30 l/h) for 2 h at 550°C in order to
2
+
reduce Cu [33, 38–41]. The inlet and the outlet
streams were analyzed by Hartmann and Braunn con-
tinuous analyzers specifically designed for NO, NO2
N O, NO , and O were continuously monitored). After
2
2
2
each adsorption test, the reactor was purged under an
He flow and then the TPD experiment was carried out by
heating the reactor 9 K/min up to 550°C in flowing He.
(
URAS 10 E), and O (MAGNOS 6G), while the NO
2
2
concentration was indirectly evaluated through a NO2
to NO catalytic converter (Hartmann and Braunn CGO-K).
The standard conditions used for the activity measure-
–
3
ments were W/F = 0.09 g s cm , T = 450°C, and NO
concentration = 5000 ppm. These conditions allowed
the NO conversion to be enough low (<15% in all
experiments) to assume a differential reactor behavior.
Wet gas feed was obtained by saturating a helium
stream at room temperature. The effect of water vapor
on the catalytic activity was determined by measuring
the activity under dry conditions after the two aging
treatments reported below: (a) 80 min at 450°C by add-
ing 2.0 vol % water to the reaction feed; (b) treatment
RESULTS AND DISCUSSION
Effect of the Ion-Exchange
The chemical composition of the catalysts shows
that the preparation procedure very strongly affects the
amount of rare earth loaded in the samples (Table 1).
However, it is worth noting that the catalysts prepared
via a two-step exchange were exchanged under more
favorable conditions (temperature, time of exchange,
and concentration of rare earth salt in the solution) with
respect to the co-exchanged samples (see experimental
section). These conditions, together with the absence of
copper ions in the first exchange step are probably
(
a) followed by the restoration of dry conditions and a
further aging treatment under wet helium flow (2.5 vol %
of water vapour) at 500°C for 15 h.
The catalytic performance is expressed as TOF responsible for the larger amount of rare earth loaded in
Turnover frequency = number of NO molecules con- the two-step-exchanged samples. Among the three rare
(
verted to nitrogen per second per copper atom) in order earth cations investigated, samarium is the more easily
to better compare samples with different copper con- exchanged ion, especially for the two-step preparation.
tents.
On the contrary, copper loading is only slightly
dependent on the preparation procedure: all samples
exhibit a copper content very similar to the unpromoted
Cu-Z, which corresponds to a level of Cu exchange
Characterization Measurements
Elemental analysis of Cu, La, Sm, and Ce was per- close to 100%, except for SmCu-Z, which contains an
formed by atomic emission spectroscopy with induc- amount of Cu about 30% greater than CuSm-Z. The
tively coupled plasma atomization (ICP-AES) after enhanced copper loading due to the large amount of
drying the samples overnight at 100°C. The ion samarium has been attributed by P aˆ rvulescu et al. [26]
exchange level was calculated on the base of the atomic to a better dispersion of Cu ions.
KINETICS AND CATALYSIS Vol. 47 No. 5 2006

ENHANCEMENT OF HYDROTHERMAL STABILITY OF Cu-ZSM5 CATALYST
731
Intensity, cps
Intensity, cps
(
‡)
(b)
1
0000
10000
8
6
4
2
000
000
000
000
0
8000
CuSm-Z
SmCu-Z
6000
CuCe-Z
CuLa-Z
CeCu-Z 4000
2000
Cu-Z
H-Z
LaCu-Z
1
0
20
2
30
40
10
20
2θ, deg
30
40
θ, deg
Fig. 1. XRD patterns of fresh co-exchanged (a) and two-step exchanged (b) samples in comparison with the parent H-ZSM5 zeolite.
Surface area measurements (Table 1) show that the slightly lower with respect to that of lanthanum (40 and
insertion of a large rare earth cation does not result in 18%, respectively for cerium and samarium two steps
any significant change of both specific surface area and exchanged samples).
porosity with respect to the unpromoted Cu-ZSM5.
As a general trend, the samples prepared via the
Also, the pore size distribution was almost unaffected
two-step exchange procedure exhibit activity higher
by the presence of rare earth, which is in agreement
than that of the co-exchanged catalysts. This stronger
with Parvulescu et al. [28] for Tl-promoted samples.
promoting effect of rare earths for the two-step-
XRD patterns of the parent H-ZSM5 and all
exchanged samples are shown in Fig. 1. In agreement
with the BET textural analyses, the diffractograms
reveal that all exchanged samples keep the MFI struc-
ture without appreciable loss of crystallinity with
respect to the parent zeolite.
exchanged samples could be simply related to the
higher rare earth content. Nevertheless, the very close
TOF, mHz
1
1
1
.2
.0
1
2
2
Catalytic Activity
1
2
3
Figure 2 reports the results of catalytic activity tests,
as NO turnover frequency (TOF), evaluated for the two
series of rare earth promoted Cu-ZSM5 samples com-
pared with that of the unpromoted Cu-ZSM5.
The experiments clearly show that the samples con-
taining rare earths show a higher activity in NO decom-
position than the un-promoted Cu-ZSM5. In particular,
the lanthanum two-step exchanged sample reaches a
TOF of 1.23 mHz, corresponding to an increase of
about 60% of intrinsic activity of copper in the unpro-
moted zeolite. Also samarium- and cerium-based sam-
ples are more active than the corresponding Cu-Z,
although the promoting effect of the RE cation is
0.8
0.6
0
0
.4
.2
0
Cu-Z
LaCu-Z
SmCu-Z
CeCu-Z
Fig. 2. NO decomposition rate (TOF) at 450°C evaluated
under dry conditions for unpromoted and rare earth-pro-
moted Cu-ZSM5. Two-step exchanged (1) and co-
exchanged RE-containing zeolites (2), Cu-Z (3).
KINETICS AND CATALYSIS Vol. 47 No. 5 2006

7
32
TOF, mHz
PALELLA et al.
TOF, mHz
.2
1
1
0
0
0
0
.2
.0
.8
.6
.4
.2
0
1
2
3
96%
1
1
2
3
9
7%
1.0
9
1%
83%
9
2%
81%
0
0
.8
.6
7
9%
4
8%
4
1%
4
6%
34%
0.4
0.2
0
2
8%
16%
6
%
LaCu-Z
SmCu-Z
CeCu-Z
Cu-Z
LaCu-Z SmCu-Z CeCu-Z
Fig. 3. Effect of the aging treatments on the TOF of NO
decomposition evaluated under dry conditions at 450°C for
CuRE-Z and Cu-Z samples. (Fresh sample (1), after 80 min
Fig. 4. Effect of the aging treatments on the TOF of NO
decomposition evaluated under dry conditions at 450°C for
RECu-Z samples. (Fresh sample (1), after 80 min under
under reaction mixture in the presence of 2 vol % of H O
reaction mixture in the presence of 2 vol % of H O (2), after
2
2
(
2), after further 15 h in the presence of 2.5 vol % at 500°C
3). Numbers on corresponding bars represent the residual
further 15 h in the presence of 2.5 vol % at 500°C (3). Num-
bers on corresponding bars represent the residual catalyst
activity as percentage of that of the fresh sample).
(
catalyst activity as percentage of that of the fresh sample).
values of TOF evaluated for the two Sm containing compared on the base of percentage of residual activity
ZSM5, despite the very different Sm exchange level, of the fresh sample; however, after the second aging
could suggest that the promotion of the rare earth starts
to decline at concentration levels that are too high, as in
the case of the SmCu-Z sample.
treatment, the Ce-containing catalysts are significantly
more active than Cu-Z. This difference with Zhang and
Flytzani-Stephanopoulos data can be likely attributed
to the different preparation technique (ion exchange for
3
h instead of three subsequent ion exchanges) and
Effect of Aging Treatments
lower rare earth content (≤1000 ppm in our samples vs.
Figures 3 and 4 report the TOF values measured
under dry conditions after the two aging treatments car-
ried out following the procedure described in the
1
.7% in [13]).
The results also show that the extent of deactivation
“
Experimental” section. Lower values were obtained of both co-exchanged (Fig. 3) and two-step exchanged
for all aged samples, indicating an irreversible partial samples (Fig. 4) is quite similar. For both classes of cat-
loss of the original activity of the fresh catalysts. How-
ever, the presence of rare earth cations results in an
enhancement of the hydrothermal stability of catalysts.
Indeed, for Cu-Z, even the first and shorter aging treat-
ment produces about a 20% irreversible loss of the orig-
inal activity, whereas CuLa-Z and CuSm-Z retained
alysts, the first aging treatment results in a weak deac-
tivation of samples that are much more severely
affected by the second longer treatment. Moreover, the
two-step exchanged zeolites show not only a better
activity of fresh samples but also a higher stability to
almost the same activity measured before aging. Upon severe wet treatments (Fig. 4). The recovery of activity
the second and more severe aging treatment, prolonged after the first aging treatment is almost similar for the
for 15 h at 500°C in the presence of water vapour
2.5 vol %), all samples were significantly deactivated.
However, in this case as well, the values of TOF esti-
mated at the end of the treatment are significantly
higher for the rare earth-containing samples. The aged
Cu-Z is about 20 times less active than the fresh one,
while after the two aging cycles, the rare earth pro-
two classes of catalysts, but a significant enhancement
of catalyst stability is clearly shown by the RECu-Z
samples at the end of the second aging treatment.
Therefore, the preparation procedure and, above all, the
higher rare earth loading promote not only a better dis-
tribution of copper and a higher activity of the catalysts
(
moted catalysts still keep about 40–50% of the original but also their hydrothermal resistance. At the end of the
activity. Nevertheless, unlike the findings of Zhang and second cycle, the following scale of stability among all
Flytzani-Stephanopoulos [13, 29], the cerium-doped
zeolite exhibits a lower hydrothermal stability.After the
first aging treatment, the resistance of both CuCe-Z and
catalysts investigated can be given:
LaCu-Z > SmCu-Z > CeCu-Z > CuLa-Z > CuCe-Z >
CeCu-Z is similar to that of the undoped Cu-ZSM5, if CuSm-Z > Cu-Z
KINETICS AND CATALYSIS Vol. 47 No. 5 2006

ENHANCEMENT OF HYDROTHERMAL STABILITY OF Cu-ZSM5 CATALYST
733
TOF, mHz
nitric oxide for the adsorption on the same active Cu
sites.
1
2
1
1
0
0
0
0
.2
Also in the presence of water, the two-step
exchanged samples were more active than coexchange
catalysts, as observed under dry conditions. However,
the effect of the kind of rare earth cation on the activity
under wet conditions is slightly different compared to
that observed for the corresponding dry tests. In fact,
under the wet conditions explored, CeCu-Z is the most
active catalyst.
.0
.8
.6
.4
.2
0
1
5%
11%
Characterization of the Aged Samples
1
0%
6
%
7%
3
%
The deactivation of the Cu-ZSM5 and of the rare
earth-containing samples upon water addition to the
feed was preliminarily investigated through structural
and chemical characterization. Data reported in Table 1
clearly show that the hydrothermal aging did not cause
any noticeable loss of surface area or micropore occlu-
sion. Moreover, XRD patterns of the steam-aged cata-
lysts, shown in Fig. 6 in comparison with the corre-
sponding fresh samples for Cu-Z and LaCu-Z, do not
reveal any significant difference, neither in the range
CuLa-Z
CuSm-Z
LaCu-Z
CuCe-Z
CeCu-Z
SmCu-Z
Fig. 5. TOF of NO decomposition evaluated at 450°C in the
presence of water for rare earth-promoted Cu-ZSM5 (2)
compared with those of fresh samples under dry condi-
tions (1). (Numbers on corresponding bars represent the
residual catalyst activity as percentage of that of the fresh
sample.)
2
θ = 35°–40°, where the presence of segregated phases
of copper oxides could be evidenced [13, 16, 43]. How-
ever, the absence of these peaks cannot exclude the for-
Catalyst Performance under Wet Conditions
The activity of catalysts was also measured under mation of very small particles of copper oxides with
wet feed conditions. Cu-Z does not have any detectable sizes smaller than 40 Å [43].
activity in the presence of steam, but the samples con-
Likewise, the results of elemental analysis of the
taining rare earth metals (two-step exchanged samples) aged samples (Table 1) indicate that the copper content
show stable residual activity, although the reduction of remains quite unchanged with respect to the fresh sam-
TOF with respect to dry conditions is quite large ples, suggesting that no copper is washed up under wet
(
Fig. 5). This dramatic reduction of activity in the experimental conditions.
decomposition of NO can be attributed either to an irre-
Therefore, all these results suggest that the catalyst
versible deactivation of catalysts, as evidenced by the deactivation is not caused by the collapse of the MFI
subsequent tests under dry conditions (Figs. 3 and 4), or framework, as proposed byYan et al. [12] and Iwamoto
to a reversible effect of competition between water and et al. [15], but could be better attributed to some modi-
Intensity, cps
0000
Intensity, cps
1
10000
(
‡)
(b)
8
6
4
2
000
000
000
000
0
8000
6000
2
1
2
4000
2000
1
0
1
0
20
θ, deg
30
40 10
20
2θ, deg
30
40
2
Fig. 6. XRD patterns of fresh and aged Cu-Z (a) in comparison with LaCu-Z (b).
KINETICS AND CATALYSIS Vol. 47 No. 5 2006

7
34
PALELLA et al.
Cu-Z
LaCu-Z
6
4
2
00
00
00
NO
0
50
00
1
1
N2O
5
0
0
10
20
30 150 300
0
10
20
30 150 300
Time, min
Time, min
Fig. 7. NO and N O outlet concentrations as functions of elapsed time after the addition of NO (600 ppm) at 120°C on pre-reduced
2
Cu-Z and LaCu-Z. (Continuous lines represent the signals of the fresh samples, dotted lines of the aged ones.)
fication of copper properties as proposed by Kucherov results of this experiment for the unpromoted Cu-Z and
14]. These modifications should be associated to the the lanthanum-containing LaCu-Z samples, chosen as
[
migration of active species to inactive positions and to representative of all investigated catalysts, are reported
the subsequent segregation of small CuO clusters in the in Fig. 7 showing the outlet reactor concentrations of
zeolite channels, as already proposed by others [13, 15, NO and N O during the transient NO adsorption.
1
2
6], copper content in the catalyst being unchanged.
For all samples, a transient evolution of N O was
2
observed during NO adsorption. Moreover, both the
fresh and aged rare earth-containing samples show a
larger NO adsorption capacity with respect to the
unpromoted zeolite. Since no modification of the
microporous structure occurred upon rare earth
exchange and the total copper content was the same for
either unpromoted and promoted catalysts, the
enhancement of NO adsorption capacity should be
attributed to modifications induced to copper by rare
earth cations. A better distribution of copper species in
the zeolite promoted by the presence of rare earth, as
already suggested for other rare earth cations [26],
could be responsible for this phenomenon. In addition,
it should be noticed that the transient production of
NO Adsorption Measurements
In previous works [41, 42], we showed that the
study of the interaction between NO and over-
exchanged Cu-ZSM5 at temperatures low enough to
completely inhibit a steady-state activity was a very
useful tool to investigate the redox chemistry of these
systems. In particular, we demonstrated that NO
adsorption experiments at low temperatures can be
used to titrate those copper ions that are “self-reduced”
at high temperature under inert flow or vacuum through
the evaluation of Cu sites selectively re-oxidized by
nitric oxide that are supposed to be active. The re-oxi-
+
dation of Cu sites at low temperatures was reported to
N O, associated to NO adsorption, is significantly
2
occur via the transient formation of gaseous N O dur-
2
larger for rare earth-containing zeolites than for unpro-
moted Cu-ZSM5, both as fresh and as aged samples, as
ing NO adsorption [33, 38–41, 44–49]. In the present
work, the experiments of NO adsorption on the pre-
reduced catalyst were hence carried out with the aim to
correlate the enhancement of the hydrothermal stability
induced by rare earths with the amount of these copper
sites.
shown in Table 2 where the amount of N O produced,
2
evaluated from the integration of the area of N O sig-
2
nal, is reported for each catalysts. In a previous paper
[
41], we proposed that under these conditions, NO is
partially reduced to N O leaving one oxygen atom per
2
NO adsorption tests on both fresh and aged catalysts two reacted NO molecules on the surface. The stoichi-
were performed at 120°C after a standard reducing pre- ometry of this gas-solid reaction is not very trivial,
treatment in He flow (see experimental section). The depending on the different hypotheses made on the
KINETICS AND CATALYSIS Vol. 47 No. 5 2006

ENHANCEMENT OF HYDROTHERMAL STABILITY OF Cu-ZSM5 CATALYST
735
Table 2. Amount of N O produced during the adsorption ing catalysts, as expected from the lower reduction of
2
of NO at 120°C for fresh and aged pre-reduced samples
activity after the aging treatments. Thus, the presence
of even small amounts of lanthanum, cerium, or samar-
ium (see Table 1) not only enhances the number of self-
reducible copper ions and, consequently, the catalytic
activity in the decomposition of NO under dry condi-
tions, but also partially preserves these features in the
action of water at high temperature.
N O produced
2
FRESH
AGED
SAMPLE
molecules/
Cu atom
molecules/
Cu atom
mmol g–1
mmol g–1
Cu-Z
62
89
84
71
100
82
0.23
0.31
0.30
0.28
0.35
0.24
0.30
3
29
14
21
49
40
28
0.01
0.10
0.05
0.08
0.17
0.12
0.09
In order to quantitatively demonstrate that the nature
of active copper sites does not change upon both rare
earth addition in the zeolite and aging treatment but that
only the number of these sites is affected, in Fig. 8, the
values of TOF estimated for all samples investigated in
this work (fresh and aged) is reported as a function of
CuLa-Z
CuSm-Z
CuCe-Z
LaCu-Z
SmCu-Z
CeCu-Z
91
the N O/Cu ratio calculated from the NO adsorption
2
experiments on pre-reduced catalysts. A very good lin-
ear correlation between the amount of nitrous oxide
evolved during the NO adsorption and the catalytic
activity has been obtained. Moreover, it is noteworthy
that such correlation passes through the origin, further
suggesting that the catalytic activity in the decomposi-
redox chemistry of copper in Cu-ZSM5 [45–51] and
can be expressed according to the following equation:
2NO + σred
N O + σ-O,
2
where σred indicates the general form of copper in the
reduced state and σ-O, the same site in the oxidized
form.
+
tion of NO is directly related to the presence of Cu ions
involved in the redox cycle.
An oxygen atom is then available for reoxidizing the
+
CONCLUSIONS
pre-reduced copper sites. As a consequence, if Cu con-
taining sites are supposed to be the active centers, the
The effect of the addition of small amounts of rare
earth cations (lanthanum, samarium, and cerium) on
the catalytic properties of copper-exchanged ZSM5 in
the decomposition of nitric oxide results in a significant
enhancement of both catalytic activity and catalyst
hydrothermal stability. Catalysts prepared by two-step
exchange show better performances compared to those
prepared by simultaneous exchange of copper and rare
earth cations, likely due to the higher rare earth content
obtained using the former method. The enhanced activ-
ity has been associated with a better copper dispersion
promoted by the rare earth cations, which leads to an
increase of copper reducible sites involved in the redox
mechanism of NO decomposition. The partial or total
deactivation of the catalysts upon aging treatment
under wet conditions at high temperatures has been
attributed to the migration of active copper into inactive
positions, likely forming CuO micro-clusters. This phe-
nomenon is highly inhibited by the presence of rare
earth cations which significantly preserve active copper
by limiting its mobility in the presence of water, thus
improving the hydrothermal stability of the catalyst.
amount of N O formed should be proportional to the
2
catalytic activity. The dramatic decrease of the amount
of N O formed over Cu-Z after the aging treatment is in
2
agreement with the hypothesis. This indicates that the
+
redox properties of Cu sites, or the nature itself of
these sites, are significantly modified by the aging treat-
ment. This effect is less marked for rare earth-contain-
TOF, mHz
1
1
1
0
0
0
0
.4
.2
.0
.8
.6
.4
.2
1
2
LaCu
CeCu
Cu
CuLa
CuSm
CuCe
SmCu
LaCu
CeCu
SmCu
0.2
CuLa
CuCe
ACKNOWLEDGMENTS
CuSm
Cu
This work has been financed by the Research Fund
for the Italian Electrical System established by the Min-
istry of Industry Decree DM 26/1/2000.
0
0.1
0.3
0.4
N O/Cu
2
Fig. 8. TOF of NO decomposition as a function of the
REFERENCES
amount of N O produced during the NO adsorption over
2
pre-reduced CuRE-Z and RECu-Z catalysts. (Fresh samples
1. Bosch, H. and Jansenn, F., Catal. Today, 1988, vol. 2,
(ꢀ), aged samples (ꢁ)).
p. 369.
KINETICS AND CATALYSIS Vol. 47 No. 5 2006

7
36
PALELLA et al.
2
3
4
. Parvulescu, V.I., Grange, P., and Delmon, B., Catal 26. P aˆ rvulescu, V.I., Oekler, P., Grange, P., and Delmon, B.,
Today, 1998, vol. 46, p. 369.
Appl. Catal., B, 1998, vol. 16, p. 1.
. Iwamoto, M., Yokoo, S., Sakai, K., and Kagawa, S., 27. Parvulescu, V.I., Centeno, M.A., Grange, P., and Del-
ˆ
J. Chem. Soc., Faraday Trans., 1981, vol. 77, p. 1629.
mon, B., J. Catal., 2000, vol. 191, p. 445.
. Iwamoto, M., Furukawa, H., and Kagawa, S., in New 28. P rvulescu, V.I., Grange, P., and Delmon, B., Appl.
aˆ
Developments in Zeolite Science and Technology,
Catal., B, 2001, vol. 33, p. 223.
29. Zhang, Y. and Flytzani-Stephanopoulos, M., in Environ-
mental Catalysis, Centi, G. et al., Eds., 1995, p. 271.
30. Chajar, Z., Denton, P., de Bernard, F.B., et al., Catal.
Lett., 1998, vol. 55, p. 217.
Amsterdam: Elsevier, 1986, p. 943.
5
6
7
8
9
. Iwamoto, M. andYahiro, H., Catal. Today, 1994, vol. 22,
p. 5.
. Kapteijn, F., Marbán, G., Rodriguez-Mirasol, J., and
Moulijn, J.A., J. Catal., 1997, vol. 167, p. 256.
. Dandekar, A. and Vannice, M.A., Appl. Catal., B, 1999,
vol. 22, p. 179.
. Yahiro, H. and Iwamoto, M., Appl. Catal., A, 2001,
vol. 222, p. 163.
31. Songip, A.R., Masuda, T., Kuwahara, H., and Hashim-
oto, K., Appl. Catal., B, 1993, vol. 2, p. 153.
32. Masuda, T., Kuwakara, H., Mukai, S.R., and Hashimoto, K.,
Chem. Eng. Sci., 1999, vol. 54, p. 2773.
3
3
3
3
3
3
3
4
4
4
4
4
4
3. Iwamoto, M. and Hamada, H., Catal. Today, 1991,
vol. 10, p. 57.
. Kharas, K.C.C., Robota, H.J., and Liu, D.J., Appl.
Catal., B, 1993, vol. 2, p. 225.
4. Li, R.S., Wen, R.W., Zhang, W.Y., and Wei, Q., Appl.
Spectrosc., 1993, vol. 13, p. 229.
1
1
1
1
1
1
1
0. Grinsted, R.A., Jen, H.-W., Montreuil, C.N., et al., Appl.
5. Van Kooten, W.E.J., Liang, B., Krijnsen, H.C., et al.,
Spectrosc., 1993, vol. 13, p. 602.
Appl. Catal., B, 1999, vol. 21, p. 203.
1. Kucherov, A.V., Hubbard, C.P., and Shelef, M., J. Catal.,
6. Van Kooten, W.E.J., Krijnsen, H.C., Bleek, C.M., and
1
995, vol. 157, p. 603.
Calis, H.P.A., Appl. Catal., B, 2000, vol. 25, p. 125.
2. Yan, J.Y., Lei, G.D., Sachtler, W.M., and Kung, H.H.,
7. Ciambelli, P., Corbo, P., Gambino, M., et al., Catal.
Today, 1995, vol. 26, p. 33.
J. Catal., 1996, vol. 161, p. 43.
3. Zhang, Y. and Flytzani-Stephanopoulos, M., J. Catal.,
8. Aylor, A.W., Larsen, S.C., Reimer, J.A., and Bell, A.T.,
1
996, vol. 164, p. 131.
J. Catal., 1995, vol. 157, p. 592.
4. Kucherov, V., Hubbard, C.P., Kucherova, T.N., and
9. Li, Y. and Armor, J.N., Appl. Catal., A, 1991, vol. 76,
Shelef, M., Kinet. Catal., 1997, vol. 4, p. 556.
p. L1.
5. Iwamoto, M., Wang, J., Sperati, K.M., et al., Chem. Lett.,
0. Henriques, C., Ribeiro, M.F., Abreu, C., et al., Appl.
Catal., B, 1998, vol. 16, p. 79.
1
997, p. 1281.
6. Quincoces, C.E., Kikot, A., and Balsadella, E.I., and
Gloria Gonzalez, M., Ind. Eng. Chem. Res., 1999,
vol. 38, p. 4236.
7. Gomez, S.A., Campero, A., Martinez-Hernandez, A.,
and Fuentes, G.A., Appl. Catal., A, 2000, vol. 197,
p. 157.
1. Pirone, R., Garufi, E., Ciambelli, P., et al., Catal. Lett.,
1
997, vol. 43, p. 255.
2. Palella, B., Ciambelli, P., Landi, G., et al., II Int. Cataly-
sis Workshop for Young. Scientists, Beijing, 2001, p. 43.
3. Dossi, C., Fusi, A., Recchia, S., et al., Microporous
Mesoporous Mater., 1999, vol. 30, p. 165.
4. Spoto, G., Zecchina, A., Bordiga, S., et al., Appl. Catal.,
B, 1994, vol. 3, p. 151.
5. Trout, B.L., Chakraborty, A.K., and Bell, A.T., J. Phys.
Chem. B, 1996, vol. 100, p. 17582.
1
1
1
2
8. Rice, M.J., Chakraborty, A.K., and Bell, A.T., J. Phys.
Chem. A, 1998, vol. 102, p. 7498.
9. Li, Y. and Armor, J.N., Appl. Catal., B, 1993, vol. 2,
p. 239.
0. Armor, J.N. and Farris, T.S., Appl. Catal., B, 1994,
4
4
6. Bell, A.T., Catal. Today, 1997, vol. 38, p. 151.
vol. 4, p. L11.
7. Konduru, M.V. and Chuang, S.S.C., J. Phys. Chem. B,
2
2
1. Armor, J.N., Catal. Today, 1995, vol. 26, p. 147.
1
999, vol. 103, p. 5802.
2. Iwamoto, M., Yahiro, H., Kutsuno, T., et al., Bull. Chem.
Soc. Jpn., 1989, vol. 62, p. 583.
4
8. Konduru, M.V. and Chuang, S.S.C., J. Catal., 1999,
vol. 187, p. 436.
2
2
2
3. Eränen, K., Kumar, N., and Lindfors, L., Appl. Catal., B, 49. Rakic, V., Dondur, V., Gajinov, S., and Auroux, A., Ther-
994, vol. 4, p. 213.
mochim. Acta, 2004, vol. 420, p. 51.
4. Kagawa, S., Ogawa, H., Furukawa, H., and Teraoka, Y., 50. Beutel, T., Sarkany, J., Lei, G.D., et al., J. Phys. Chem. B,
Chem. Lett., 1991, vol. 3, p. 407.
1996, vol. 100, p. 845.
1
5. P aˆ rvulescu, V.I., Grange, P., and Delmon, B., J. Phys. 51. Moden, B., Da, CostaP., Lee, D.K., and Iglesia, E.,
Chem. B, 1997, vol. 101, p. 6933.
J. Phys. Chem. B, 2002, vol. 106, p. 9633.
KINETICS AND CATALYSIS Vol. 47 No. 5 2006