Effects of hydrothermal aging on NH3-SCR reaction over Cu/zeolites

Article abstract of DOI:10.1016/j.jcat.2011.12.025

The effects of hydrothermal treatment on model Cu/zeolite catalysts were investigated to better understand the nature of Cu species for the selective catalytic reduction of NO_x by NH₃. After hydrothermal aging at 800 °C for 16 h, the

Full text of DOI:10.1016/j.jcat.2011.12.025

Journal of Catalysis 287 (2012) 203–209
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Effects of hydrothermal aging on NH3-SCR reaction over Cu/zeolites
⇑
⇑
⇑
Ja Hun Kwak , Diana Tran, Sarah D. Burton, János Szanyi, Jong H. Lee , Charles H.F. Peden
Institute for Integrated Catalysis, Pacific Northwest National Laboratory, Richland, WA 99352, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The effects of hydrothermal treatment on model Cu/zeolite catalysts were investigated to better under-
stand the nature of Cu species for the selective catalytic reduction of NOx by NH3. After hydrothermal
aging at 800 °C for 16 h, the NOx reduction performance of Cu-ZSM-5 and Cu-beta was significantly
reduced at low temperatures, while that of Cu-SSZ-13 was not affected. When the zeolite framework alu-
minum species were probed using solid state 27Al MASNMR, significant reduction in the intensities of the
tetrahedral aluminum peak intensity was observed for Cu-ZSM-5 and Cu-beta, although no increase in
the intensities of the octahedral aluminum peak was detected. When the redox behavior of Cu species
was examined using H2-TPR, it was found that Cu2+ could be reduced to Cu+ and to Cu0 for Cu-ZSM-5
and Cu-beta catalysts, while Cu2+ could be reduced only to Cu+ in Cu-SSZ-13. After hydrothermal aging,
CuO and Cu-aluminate species were found to form in Cu-ZSM-5 and Cu-beta, while little changes were
observed for Cu-SSZ-13.
Received 1 August 2011
Revised 14 November 2011
Accepted 31 December 2011
Available online 31 January 2012
Keywords:
Cu-SSZ-13
NH3-SCR
Hydrothermal aging
H2-TPR
Solid state 27Al NMR
Ó 2012 Elsevier Inc. All rights reserved.
1. Introduction
a wide range of temperatures. However, these catalysts have been
found to deactivate readily during high-temperature filter regener-
Various catalyst technologies have been developed to remove
pollutants from engine exhaust. For example, a three-way catalyst
(TWC) is used to remove hydrocarbons (HC), carbon monoxide
(CO), and nitrogen oxides (NOx) from stoichiometric gasoline en-
gines. For diesel engines, a diesel oxidation catalyst (DOC) is used
to control HC and CO emissions, while particulate matters (PM)
are removed by a diesel particulate filter (DPF). NOx can be re-
moved by either a lean NOx trap (LNT) catalyst that can store
NOx under lean conditions and reduce NOx under rich conditions,
or by a selective catalytic reduction (SCR) catalyst that can selec-
tively reduce NOx with a reducing agent (e.g., NH3, hydrocarbons)
under oxidizing conditions. A typical emission control system for
diesel engines consists of DOC, SCR, and DPF, which are placed in
a specific order to achieve a desired level of emission reduction
performance. As soot is often removed from the DPF at high tem-
peratures (>650 °C), high thermal durability is required for the
SCR catalyst to remain effective for NOx emission control [1].
Among the various catalysts developed for NH3 SCR, zeolite-
based base metal (e.g., Cu, Fe) catalysts are currently being used
for meeting diesel NOx emission standards in North America. Sig-
nificant research efforts have concentrated on Cu2+ ion-exchanged
ZSM-5 (Cu-ZSM-5) zeolites to study both its NO decomposition and
SCR activity [2]. Early development efforts have also focused on
Cu2+ exchanged beta zeolite (Cu-beta) for its excellent activity over
ation. Recently, we reported that Cu2+ ion-exchanged SSZ-13
(Cu-SSZ-13), a zeolite with the Chabazite (CHA) structure and con-
taining small radius (ꢀ3.8 Å) eight-membered ring pores, is more
active and selective in reducing NO with NH3 compared to
Cu-ZSM-5 and Cu-beta [3].
It was also found that Cu-SSZ-13 is less prone to deactivation by
hydrocarbon inhibition or thermal degradation [4,5]. Interestingly,
some improvement in high-temperature activity has also been pre-
viously reported for some Cu/zeolite catalysts after hydrothermal
aging, which may suggest that activity can sometimes improve
during hydrothermal aging processes before the catalyst becomes
fully deactivated [6]. Thus, in order to better understand the origin
of high activity and thermal stability of Cu-SSZ-13, we investigate
here the effects of high-temperature hydrothermal treatment on
Cu species and zeolite framework structure, and on NO reduction
activity using model Cu/zeolite catalysts, namely, Cu-SSZ-13,
Cu-beta, Cu-ZSM-5, and Cu-Y.
2. Experimental
2.1. Catalyst preparation
ZSM-5 (CBV-3024, Si/Al2 = 30), beta (CP-814C, Si/Al2 = 38), and
Y (CBV-100, Si/Al2 = 5.2) zeolites were obtained from Zeolyst Inter-
national Co. SSZ-13 (Si/Al2 = 12) was prepared using the method
reported previously [3,7]. Cu/zeolite catalysts were prepared by
aqueous ion exchange using Cu(NO3)2 as precursor. Cu solutions
contained twice the amount of Cu2+ ions needed for complete ion
⇑
Corresponding authors.
E-mail addresses: kwak@pnnl.gov (J.H. Kwak), jong.lee@pnnl.gov (J.H. Lee),
chuck.peden@pnnl.gov (C.H.F. Peden).
0021-9517/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.12.025

204
J.H. Kwak et al. / Journal of Catalysis 287 (2012) 203–209
exchange. After ion exchange over 1 day at room temperature, the
catalyst samples were filtered, washed, and dried overnight at
100 °C. In order to maximize the ion exchange level under the gi-
ven experimental conditions, this entire process was repeated
twice for each catalyst sample. The results of elemental analysis
(ICP) for the thus prepared samples are summarized in Table 1.
All catalysts were then calcined in an oven at 500 °C for 2 h prior
to testing and characterization. For the hydrothermal aging stud-
ies, some of the catalysts were further treated in 10% H2O in air
at 800 °C for 16 h.
amount of sample was used and the same number of scans was
acquired for each experiment. A 10° flip angle was used during
the single pulse experiment to obtain quantitative spectra.
Prior to H2 temperature-programmed reduction (TPR) experi-
ments, 0.05 g of Cu–zeolite catalyst was calcined at 500 °C for 2 h
in flowing air (1.0 ml/s). The sample was then cooled down to room
temperature in the same air flow, and then purged in 2% H2/Ar
(1.0 ml/s) for 1 h at RT. Once the TCD signal of a Hewlett–Packard
7820 gas chromatograph (GC) stabilized, a TPR experiment was
carried out in 2% H2/Ar (1.0 ml/s) at a heating rate of 10 K/min.
The amount of H2 consumed was determined from TCD signal
intensities, which were calibrated using a 10% CuO/SiO2 reference
sample.
2.2. Catalytic activity measurement
Catalytic activity was measured using a packed bed micro reac-
tor system. Catalyst samples (ꢀ250 mg) of 60–80 mesh size were
loaded in a 3/800 OD quartz tube, which was then placed inside
an electric furnace. The reaction temperature was monitored by a
thermocouple located at the inlet position. The feed gas contained
350 ppm NOx, 350 ppm NH3, 14% O2, 10% H2O, and balance N2. All
the lines were heated to over 100 °C to avoid condensation inside
the reactor system. The total gas flow rate was 210 ml/min (STP),
and the gas hourly space velocity (GHSV) was estimated to be
ꢀ30,000 hꢁ1. Concentrations of reactants and products were mea-
sured by a Nicolet Magna 760 infrared (FT-IR) spectrometer with a
heated 2-m gas cell.
3. Results and discussion
3.1. Catalyst activity measurements
The NOx reduction activity of each Cu ion-exchanged zeolite
catalyst was examined using the feed gas containing 350 ppm
NOx, 350 ppm NH3, 14% O2, and 10% H2O at the GHSV of
30,000 hꢁ1 between 150 and 550 °C. As shown in Fig. 1a, excellent
NO reduction activity was obtained for all fresh Cu/zeolite cata-
lysts. For example, NO reduction activity increased with increasing
temperature, reaching 85–95% at 250 °C. At higher temperatures,
NO reduction activity was limited by the availability of ammonia
as the ammonia oxidation reaction to NOx becomes significant.
Interestingly, a significant amount of N2O was produced over
Cu-Y (shown in Fig. 2a). Because N2O is a greenhouse gas, it is
The catalysts were evaluated for their activity for NOx reduction
to N2, and for NH3 oxidation. The NOx and NH3 conversion efficien-
cies were calculated based on the differences in their concentra-
tions measured before and after the catalyst in the following way:
% NOx conversion ¼ fðNO þ NO2Þinlet ꢁ ðNO þ NO2
þ 2ꢂN2OÞoutletg=ðNO þ NO2Þinlet ꢂ 100
(a)
% NH3 conversion ¼ ðNH3inlet ꢁ NH3outletÞ=NH3inlet ꢂ 100
Note that for calculating NOx conversions, N2O is not considered to
be a product, but rather unconverted NOx.
2.3. Catalyst characterization
X-ray diffraction (XRD) measurements were made on a Philips
PW3040/00 X’Pert powder X-ray diffractometer using Cu Ka radi-
ation (k = 1.5406 Å), with scans run between 2h values of 5° and
50° at 0.02°/s. Diffraction patterns were analyzed using JADE
(Materials Data, Inc.) as well as the Powder Diffraction File data-
base (International Center for Diffraction Data, 2003 release).
Solid state 27Al nuclear magnetic resonance (NMR) spectra were
acquired on a Varian Chemagnetics CMX Infinity 300 MHz instru-
ment, equipped with a Varian Chemagnetics 7.5 mm HX magic an-
gle spinning (MAS) probe operating at the spectral frequency of
78.2 MHz. About 0.2 g of each catalyst sample was transferred into
a gastight rotor (7.5 mm OD), which was then promptly placed in
the probe and NMR magnet. All 27Al MAS NMR spectra were exter-
nally referenced to an aqueous solution of Al(NO3)3 at 0 ppm. All
spectra were obtained at a sample spinning rate of 5 kHz and using
1 s recycle delay. For comparison of signal intensities, the same
(b)
Table 1
ICP analysis results of the Cu ion-exchanged zeolite samples before hydrothermal
aging.
Cu-Y, FAU
Cu-beta
Cu-ZSM-5
Cu-SSZ-13
Si/Al2
Cu/Al
Cu loading (wt.%)
Cu I.E. level (%)
5.3
0.35
7.2
70
39.0
0.34
1.73
69
32.9
0.53
2.83
106
12.4
0.40
4.3
Fig. 1. NO conversion to N2 over Cu/zeolites (a) fresh catalysts and (b) after
hydrothermal aging (Feed: 350 ppm NO, 350 ppm NH3, 14% O2, 10% H2O in balance
N2).
79

J.H. Kwak et al. / Journal of Catalysis 287 (2012) 203–209
205
Table 2
(a)
Turnover rates of Cu ion-exchanged zeolite samples before and after hydrothermal
aging.
Samples
Turnover rate (/s)*
150 °C
200 °C
Cu-Y
Fresh
HTA
2.23 ꢃ 10ꢁ3
1.18 ꢃ 10ꢁ2
0
0
Cu-beta
Fresh
HTA
3.69 ꢃ 10ꢁ2
1.74 ꢃ 10ꢁ2
7.36 ꢃ 10ꢁ2
5.24 ꢃ 10ꢁ2
Cu-ZSM-5
Fresh
HTA
1.32 ꢃ 10ꢁ2
1.12 ꢃ 10ꢁ2
2.87 ꢃ 10ꢁ2
2.13 ꢃ 10ꢁ2
Cu-SSZ-13
Fresh
HTA
7.60 ꢃ 10ꢁ3
7.60 ꢃ 10ꢁ3
2.30 ꢃ 10ꢁ2
2.88 ꢃ 10ꢁ2
(b)
*
Estimated based on the number of Cu ions.
reaction). However, it is well known that NH3 SCR reaction kinetics
can be improved in the presence of equimolar amounts of NO and
NO2 in the feed gas via the ‘‘fast’’ SCR reaction [8]. For this reason, a
typical emission control system for diesel engines often includes a
DOC ahead of the SCR catalyst to convert some of the exhaust NO
to NO2 [1]. However, when NO2 is included in the feed, N2O is often
produced at higher levels via a reaction mechanism that involves
the catalyzed formation and decomposition of NH4NO3 [9]. As
shown in Fig. 3, various amounts of N2O were obtained over both
fresh and aged Cu/zeolite catalysts under the conditions used in
Fig. 2. N2O formation over Cu/zeolites (a) fresh catalysts and (b) after hydrothermal
aging (feed: 350 ppm NO, 350 ppm NH3, 14% O2, 10% H2O in balance N2).
(a)
not considered to be a desirable product. Thus, when N2 was con-
sidered as the only product for the SCR reaction, very little NO
reduction activity was seen for fresh Cu-Y at temperatures
>350 °C. These results are consistent with our previous findings
that showed high NOx reduction activities for Cu ion-exchanged
ZSM-5, beta, and SSZ-13 zeolites [3].
As mentioned earlier, SCR catalysts are often exposed to very
high temperatures (>650 °C) during, for example, soot oxidation
processes in upstream DPFs. As such, high hydrothermal durability
is essential for practical implementation of the NH3 SCR technol-
ogy. To probe hydrothermal stability, the Cu/zeolite catalysts were
further treated in 10% H2O in air at 800 °C for 16 h to investigate
the effects of hydrothermal aging on the catalytic activities and
the catalyst structures. After this hydrothermal treatment, signifi-
cant loss of NO reduction activity, in greatly varying extent, was
observed for most of the Cu/zeolite catalysts (shown in Fig. 1b).
While Cu-SSZ-13 was found to show essentially no change in NO
reduction activity, Cu-Y lost its activity completely. Both
Cu-ZSM-5 and Cu-beta were found to lose NO reduction activity
primarily at low catalyst bed temperatures, but maintain reason-
able activity at high (>350 °C) temperatures. Since these Cu/zeolite
catalysts contain different amounts of Cu ions, turnover rates were
estimated based on the amount of Cu present in the sample. As
shown in Table 2, the loss of low-temperature NO reduction activ-
ity in high-temperature aged samples can be clearly observed for
all Cu/zeolite catalysts, except Cu-SSZ-13. The amount of N2O pro-
duced was low over all aged Cu zeolites, except Cu-beta (shown in
Fig. 2b). In fact, N2O formation over the aged Cu-beta was much
higher than on the fresh Cu-beta at all temperatures.
(b)
Fig. 3. N2O formation over Cu/zeolites (a) fresh catalysts and (b) after hydrothermal
aging (feed: 175 ppm NO, 175 ppm NO2, 350 ppm NH3, 14% O2, 10% H2O in balance
N2).
Data in Figs. 1 and 2 were for reaction mixtures that contained
NOx entirely in the form of NO (the so-called ‘‘standard’’ SCR

206
J.H. Kwak et al. / Journal of Catalysis 287 (2012) 203–209
this study. Interestingly, Cu-SSZ-13 was found to reduce NOx to N2
almost exclusively even in the presence of NO2. Because very little
N2O formation was seen over other small pore Cu/zeolite catalysts
[5], it is possible that the observed high selectivity to N2 may be re-
lated to the zeolite structure (type), and this aspect is currently un-
der investigation.
degradation of the crystalline zeolite structure can be followed
by powder XRD, while the removal of tetrahedral Al3+ ions from
the zeolite framework (i.e., dealumination) can be monitored by
27Al MAS NMR. On the other hand, changes in the chemical envi-
ronment of the Cu ions may be reflected in their reducibilities as
measured by H2-TPR.
Because of the enhanced NOx reduction kinetics in gas mixtures
containing equimolar NO/NO2 (fast SCR), high NOx reduction activ-
ity is typically reported for many of the Cu/zeolite catalysts, espe-
cially at lower temperatures. Note, however, that when only N2 is
considered as the N-containing reaction product (i.e., N2O is still
defined as NOx), our results show that improved NOx reduction
activity is obtained for Cu-SSZ-13 only (Fig. 4a). Following the
hydrothermal treatment at 800 °C for 16 h, some loss of NOx reduc-
tion activity was observed in Cu-SSZ-13 at low and high tempera-
tures (shown in Fig. 4b). On the other hand, total loss of activity
was observed for Cu-Y, and both Cu-ZSM-5 and Cu-beta were
found to lose NOx reduction activity, in part because of increased
N2O formation following hydrothermal aging.
First, XRD measurements were performed over both fresh and
aged samples to probe possible structural changes. XRD patterns
recorded for the fresh (panel A) and hydrothermally aged (panel
B) Cu/zeolite samples are displayed in Fig. 5. Comparison of the
XRD patterns from the fresh and aged samples clearly show the
complete collapse of the zeolite structure for Cu-Y as a result of
hydrothermal aging, consistent with the total loss of its SCR activ-
ity. For this catalyst, instead of crystalline zeolite diffraction peaks,
a broad feature representing an amorphous phase with CuO peaks
is evident in the diffraction pattern following the hydrothermal
aging. On the other hand, little or no changes were observed for
Cu-SSZ-13, Cu-ZSM-5, and Cu-beta, which indicate that these zeo-
lite structures remained largely intact during the hydrothermal
aging. For these Cu/zeolite catalysts, despite some changes in SCR
activity, no formation of CuO is evident in the XRD data.
3.2. Catalyst characterization
Most zeolites can dealuminate under high-temperature hydro-
thermal conditions without exhibiting significant damage to their
crystal structures. During this process, Al3+ ions are removed from
their tetrahedral positions in the zeolite framework, and extra-
framework AlOx species can be produced. As the Al3+ in the AlOx
Previously, adverse effects of hydrothermal aging on NOx reduc-
tion activities of zeolite-based catalysts have been associated pri-
marily with the degradation of the zeolite structure or the
deactivation of catalytically active sites by cation migration [10].
In order to better understand the changes in the NOx reduction
activity induced by hydrothermal aging, changes in the physico-
chemical properties of the Cu/zeolite catalysts were examined
using XRD, 27Al MAS NMR, and H2-TPR techniques. In particular,
(A)
Cu-Y
(a)
Cu-SSZ-13
Cu-ZSM-5
Cu-beta
5
15
25
35
45
(B)
(b)
Fig. 4. NOx conversion to N2 over Cu/zeolites (a) fresh catalysts and (b) after
hydrothermal aging (feed: 175 ppm NO, 175 ppm NO2, 350 ppm NH3, 14% O2, 10%
H2O in balance N2).
Fig. 5. XRD patterns of Cu/zeolites before (A) and after (B) hydrothermal aging.

J.H. Kwak et al. / Journal of Catalysis 287 (2012) 203–209
207
species possesses octahedral coordination, the dealumination can
be conveniently monitored by 27Al MAS NMR [12–15] by a peak
at ꢀ0 ppm. For example, a decrease in the intensity of the tetrahe-
drally coordinated Al3+ ions and concurrent increase in the inten-
sity of the octahedrally coordinated Al3+ ions are expected for
dealuminated zeolite samples.
H2-TPR experiments were performed to examine the effects of
hydrothermal aging on the nature of Cu species. As shown in
Fig. 7, H2-TPR spectra collected over fresh Cu/zeolites could be di-
vided into two groups based on the reducibility of Cu ions. For
fresh Cu-ZSM-5, three reduction peaks were observed at 155 °C,
207 °C, and 315 °C. In the literature, it was suggested that Cu2+ is
reduced to Cu+ at ꢀ200 °C, while Cu+ is reduced to Cu0 at 315 °C
for Cu-ZSM-5 catalysts [10,11]. The observed peak at 155 °C peak
may be a result of the reduction of oxygen in Cu–O–Cu structures,
which can be formed at high Cu ion exchange levels (the elemental
analysis results shown in Table 1 suggest that the Cu ion exchange
level in our Cu-ZSM-5 sample slightly exceeds 100%, i.e., this sam-
ple is regarded as over-exchanged). Similar to Cu-ZSM-5, two
reduction peaks were observed at 200 °C and 390 °C for Cu-beta,
which can be attributed to the reduction of Cu2+ to Cu+, and that
of Cu+ to Cu0, respectively. On the other hand, the two reduction
peaks observed at 195 °C and 310 °C for Cu-Y catalysts are assigned
only to the reduction of Cu2+ to Cu+ based on the literature, which
suggested that Cu2+ ions inside supercages of faujasite (FAU) zeo-
lites are reduced to Cu+ at 195 °C, while reduction of Cu2+ ions to
Cu+ inside sodalite cages occurs at 310 °C [16].
Interestingly, only one H2-TPR reduction peak at 230 °C with a
broad shoulder at ꢀ300 °C was obtained for Cu-SSZ-13. During
the TPR, Cu-SSZ-13 appeared white even at 700 °C, which suggests
the formation of Cu+, a result consistent with those reported by
Kieger et al. [16]. Notably, in this prior work, the authors report
that the color of Cu-Y powders turned white due to Cu+ formation
and then became purple when Cu+ was further reduced to Cu0 at
higher temperatures. A purple color due to Cu0 formation was con-
firmed in the present work for Cu-beta and Cu-ZSM-5 after TPR up
to 700 °C. Based on these results, we assign the 230 °C peak and
300 °C shoulder on Cu-SSZ-13 as both arising from Cu2+ to Cu+
reduction with different ionic positions in the zeolite structure.
As shown in Fig. 6, all fresh Cu/zeolite catalysts show a single
tetrahedral aluminum peak at ꢀ50–60 ppm chemical shift range
(e.g., Cu-ZSM-5 = 53 ppm, Cu-beta = 53 ppm, Cu-Y = 59 ppm, and
Cu-SSZ-13 = 57 ppm). The peaks recorded for these Cu/zeolite cat-
alysts are broader than those of Na-exchanged zeolites due to line
broadening effects of paramagnetic Cu2+ ions. A broad peak at
ꢀ0 ppm appears to be due to either an artifact from the instrument
or data processing, and less likely associated with octahedral alu-
minum ions in the fresh samples. Following the hydrothermal
aging, the intensities of these tetrahedral aluminum signals were
reduced significantly for three of the four zeolite samples studied,
remaining mostly unchanged for Cu-SSZ-13. Compared to the fresh
samples, the integrated peak areas of the tetrahedral Al3+ ions in
the aged samples, which are proportional to the number of frame-
work aluminum ions, were estimated to be only ꢀ57% in Cu-ZSM-5,
ꢀ31% in Cu-beta, and ꢀ23% in Cu-Y. In contrast, the peak area for
aged Cu-SSZ-13 was very high, around 95% of the fresh sample. The
observed high thermal stability for the Cu-SSZ-13 catalyst is con-
sistent with the minor change in its SCR activity. It is well known
that tetrahedral aluminum ions in the zeolite framework are chan-
ged to octahedral aluminum upon dealumination under high-
temperature hydrothermal treatment. However, no new peaks
associated with octahedral aluminum were observed in any of
the samples used in this study. The lack of octahedral aluminum
ions in the aged zeolite catalysts strongly suggests that paramag-
netic Cu ions may interact more strongly with the forming octahe-
dral aluminum than zeolytic Cu ions with framework aluminum.
Fig. 6. Solid state 27Al-NMR spectra of Cu/zeolites before and after hydrothermal aging.

208
J.H. Kwak et al. / Journal of Catalysis 287 (2012) 203–209
Following the hydrothermal aging at 800 °C, significant differ-
ences in the H2-TPR patterns were observed, as shown in Fig. 8.
For Cu-Y, sharp peaks at 275 and 295 °C are attributed to the
reduction of bulk-like CuO structures formed during the total
collapse of the zeolite structure. For Cu-beta, two broad reduction
peaks are now present at 305 and 670 °C. As most of the zeolite
structure was still intact after the hydrothermal aging, these H2-
TPR features may still be attributed to the reduction of Cu2+ and
Cu+ ions. Over Cu-ZSM-5, a sharp reduction peak at 270 °C for
bulk-like CuO is observed, as well as broad peaks at 280 and
470 °C, similar to those for Cu-beta. In trying to rationalize these
changes, a couple of points are likely to be relevant. First, prior lit-
erature has reported that H2-TPR reduction peaks are shifted to
higher temperatures when the ion exchange level of Cu is de-
creased [18]. Furthermore, the NMR results described above indi-
cate a moderate loss of Cu ions from the ion exchange sites for
both Cu-beta and Cu-ZSM-5 catalysts with a subsequent formation
of new Cu-aluminum-oxygen complexes where Cu strongly inter-
acts with the Al ions, therefore rendering the 27Al NMR peaks asso-
ciated with these new complexes ‘‘invisible’’ in the spectra. Thus,
the observed shift in peak temperatures and reduction in peak
intensities may represent the reduced occupancy of the zeolite
ion exchange sites by Cu and/or the reduction of Cu ions in the
newly formed Cu-AlOx species.
Fig. 7. H2-TPR profiles of fresh prepared Cu/zeolites; Cu-beta (black), Cu-ZSM-5
(blue), Cu-SSZ-13 (red), and CuY (green). (For interpretation of the references to
color in this figure legend, the reader is referred to the web version of this article.)
Recently, it was reported that there are several exchange sites in-
side chabazite (CHA) zeolites available for Cu ions [17]. Among
them, it has been proposed that Cu cations can occupy two of these
sites: ‘‘site I,’’ located inside the CHA cavity, and ‘‘site IV,’’ inside the
eight-membered ring connecting the CHA cavities. Note that site I
is similar to the exchange sites inside the sodalite cage of FAU zeo-
lite where the Cu2+ to Cu+ reduction has been proposed to occur at
ꢀ300 °C. Based on these previous assignments, it seems reasonable
to propose that the prominent H2-TPR peak at 230 °C is due to the
reduction of Cu2+ to Cu+ at the IV site, while the broad shoulder
peak is perhaps due to Cu2+ to Cu+ reduction for Cu ions at site I.
However, these assignments must be considered as tentative at
best at this point, as they contradict the conclusion of Fickel and
Lobo [7], who proposed the presence of Cu ions exclusively in
one type of cationic positions. Current studies in our laboratory
are aimed solely at understanding the locations of Cu ions in our
in-house synthesized SSZ-13 sample.
On the other hand, for Cu-SSZ-13, Fig. 8 shows that the reduc-
tion peak at 230 °C is still present and that the broad shoulder at
ꢀ300 °C has increased after hydrothermal aging. Furthermore, we
find that essentially the same amount of H2 is consumed for both
the fresh and aged Cu-SSZ-13 samples, suggesting relatively small
changes in the distribution of Cu ions between the likely two zeo-
lytic sites in SSZ-13. This result is consistent with the small ob-
served changes in both the solid state 27Al NMR spectra and SCR
activity for Cu-SSZ-13 catalysts after hydrothermal aging.
Considering the changes observed in 27Al NMR and H2-TPR, it is
likely that some of the Cu2+ ions in aged Cu-ZSM-5 and Cu-beta are
still located in ion exchange positions within zeolite structure and,
thus, responsible for the SCR activity albeit severely suppressed.
On the other hand, both aged Cu-ZSM-5 and Cu-beta exhibited rel-
atively good NOx reduction activity at higher temperatures. Re-
cently, we demonstrated that isolated Cu species on
c-alumina
can be effective in NH3 SCR [19]. The somewhat different struc-
tures for Cu species in these alumina supported catalysts and Cu
ions exchanged into zeolites likely explain the considerable differ-
ences in the optimum temperature ranges for NH3 SCR activity for
these two types of catalysts. Based on these recent results, we pro-
pose that new Cu/AlOx structures that form upon hydrothermal
aging and which exhibit strong interactions between Al ions and
paramagnetic Cu are, at least partially, responsible for the mainte-
nance of higher temperature NH3 SCR activity. Such Cu/AlOx struc-
tures may be thought of as small, isolated Cu-aluminate-like
species which remain dispersed in the intact zeolite channels of
the hydrothermally aged Cu-ZSM-5 and Cu-beta catalysts.
4. Conclusions
The effects of hydrothermal aging on the materials properties
and NH3 SCR activity of Cu-ZSM-5, Cu-BEA, Cu-Y, and Cu-SSZ-13
catalysts were studied here. After hydrothermal treatment at
800 °C for 16 h, Cu-SSZ-13 was found to show essentially no
change in NOx reduction activity, while Cu-Y completely lost its
NH3 SCR activity. Both Cu-ZSM-5 and Cu-BEA were found to lose
NOx reduction activity primarily at low temperatures (<350 °C).
In the presence of equimolar amounts of NO and NO2 in the feed
gas, significant amounts of N2O were produced over the aged Cu-
ZSM-5 and Cu-BEA at all temperatures.
Fig. 8. H2-TPR profiles of Cu/zeolites after hydrothermal aging; Cu-beta (black), Cu-
ZSM-5 (blue), Cu-SSZ-13 (red), and CuY (green). (For interpretation of the
references to colour in this figure legend, the reader is referred to the web version
of this article.)

J.H. Kwak et al. / Journal of Catalysis 287 (2012) 203–209
209
XRD measurement indicated that the zeolite structure re-
References
mained largely intact for all the Cu/zeolite catalysts following the
hydrothermal aging, except for Cu-Y. When the zeolite framework
aluminum species were probed using solid state 27Al-NMR, a
significant reduction in the intensities of tetrahedral aluminum
peak was observed for the aged Cu-ZSM-5 and Cu-beta catalysts
suggesting some de-alumination, while little changes were seen
in the spectra for Cu-SSZ-13. 27Al-NMR peaks due to octahedral
aluminum ions are expected for the aged and partially de-alumina-
ted zeolite catalysts but were not observed in their spectra. This
suggests that paramagnetic Cu ions are strongly interacting with
the aluminum ions that are removed from the zeolite framework.
When the redox behavior of Cu species was examined using H2-
TPR, it was found that Cu2+ could be reduced to Cu+ and to Cu0 in
Cu-ZSM-5 and Cu-beta, while Cu2+ could be reduced to Cu+ only
in Cu-SSZ-13. After hydrothermal aging, CuO and Cu-aluminate-
like species were found to form in Cu-ZSM-5 and Cu-beta, while lit-
tle changes were observed for Cu-SSZ-13, consistent with the min-
or changes in the SCR activity of this latter catalyst.
[1] J.H. Lee, M.J. Paratore, D.B. Brown, SAE 2008-01-0072.
[] M. Iwamoto, H. Furukawa, Y. Mine, F. Uemura, S. Mikuriya, S. Kagawa, J. Chem.
Soc., Chem. Commun. (1986) 12723.
[3] J.H. Kwak, R.G. Tonkyn, D.H. Kim, J. Szanyi, C.H.F. Peden, J. Catal. 275 (2010)
187.
[4] I. Bull, W.-M. Xue, P. Burk, R.S. Boorse, W.M. Jaglowski, G.S. Koermer, A. Moini,
J.A. Patchett, J.C. Dettling, M.T. Caudle, US Patent, 7610,662, 2009.
[5] P.J. Andersen, J. Casci, H.Y. Chen, K. Collier, J.M. Fedeyko, R. Foo, R. Rajaram. In:
The 22nd North American Catalysis Society Meeting, June, 2011 (no
publication yet).
[6] G. Cavataio, H.W. Jen, J.R. Warner, J.W. Girard, J.Y. Kim, C.K. Lambert, SAE 2008-
01-1025.
[7] D.W. Fickel, R.F. Lobo, J. Phys. Chem. C 114 (2010) 1633.
[8] K. Rahkamaa-Tolonen, T. Maunula, M. Lomma, M. Huuhtanen, R. Keiski, Catal.
Today 100 (2005) 217.
[9] M. Colombo, I. Nova, E. Tronconi, Catal. Today 151 (2010) 223.
[10] J.Y. Yan, W.M.H. Sachtler, H.H. Kung, Catal. Today 33 (1997) 279.
[11] J.Y. Yan, G.-D. Lei, W.M.H. Sachtler, H.H. Kung, J. Catal. 161 (1996) 43.
[12] E. Bourgeat-Lami, P. Massiani, F. Di Renzo, P. Espiau, F. Fajula, T. Des Courieres,
Appl. Catal. 72 (1991) 139.
[13] S.M. Campbell, D.M. Bibby, J.M. Coddington, R.F. Howe, R.H. Meinholdz, J. Catal.
161 (1996) 338.
[14] J.-H. Park, H.J. Park, J.H. Baik, I.-S. Nam, C.-H. Shin, J.-H. Lee, B.K. Cho, S.H. Oh, J.
Catal. 240 (2006) 47.
[15] O. Kröcher, M. Devadas, M. Elsener, A. Wokaun, N. Söger, M. Pfeifer, Y. Demel,
L. Mussmann, Appl. Catal. B: Environ. 66 (2006) 208.
[16] S. Kieger, G. Delahay, B. Coq, B. Neveu, J. Catal. 183 (1999) 67.
[17] J. Dedecek, B. Wichterlova, P. Kubat, Microporous Mesoporous Mater. 32
(1999) 63.
[18] R. Bulanek, B. Wichterlova, Z. Sobalik, J. Tichy, Appl. Catal. B: Environ. 31
(2001) 3.
[19] J.H. Kwak, R. Tonkyn, D. Mei, S.J. Cho, L. Kovarik, J.H. Lee, C.H.F. Peden, J. Szanyi,
ACS Catal., submitted for publication.
Acknowledgments
We gratefully acknowledge the US Department of Energy (DOE),
Office of Energy Efficiency and Renewable Energy/Vehicle Technol-
ogies Program for the support of this work. The research described
in this paper was performed at the Environmental Molecular
Sciences Laboratory (EMSL), a national scientific user facility spon-
sored by the DOE’s Office of Biological and Environmental Research
and located at Pacific Northwest National Laboratory (PNNL). PNNL
is operated for the US DOE by Battelle Memorial Institute under
contract number DE-AC05-76RL01830.