Hydrothermal stability enhancement by sequential ion-exchange of rare earth metals on Fe/BEA zeolites used as NO reduction catalysts

Article abstract of DOI:10.1039/c1cc00108f

An addition of rare earth metals with trivalent cation radii of 1.05-1.15 A to Fe/BEA zeolites improves hydrothermal stability for selective catalytic reduction of NO by NH₃. The Royal Society of Chemistry.

Full text of DOI:10.1039/c1cc00108f

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2011, 47, 3966–3968
www.rsc.org/chemcomm
COMMUNICATION
Hydrothermal stability enhancement by sequential ion-exchange of rare
earth metals on Fe/BEA zeolites used as NO reduction catalystsw
Masaoki Iwasaki* and Hirofumi Shinjoh
Received 6th January 2011, Accepted 10th February 2011
DOI: 10.1039/c1cc00108f
An addition of rare earth metals with trivalent cation radii of
˚
1.05–1.15 A to Fe/BEA zeolites improves hydrothermal stability
these metals greatly alter the acidic properties of zeolites. In
contrast, RE metals have less of an effect on the acidic
properties of zeolites while still increasing hydrothermal
stability. For example, the addition of La to Cu or Co
exchanged ZSM-5 improved the hydrothermal stability of
the catalysts for NO reduction by hydrocarbons.9,10 This
increase in hydrothermal stability occurs because La decreases
the concentration of residual protons and enhances the stability
of [AlO4]À, as predicted theoretically.8 In spite of these
promising results, an exhaustive experimental survey of zeolites
exchanged with RE metals has not yet been conducted.
In this study, we investigated the hydrothermal stability of
secondary metal doped Fe/BEA zeolites prepared by sequential
ion-exchange. Previously, we demonstrated that Fe/BEA zeolites
showed the highest NH3-SCR activity before hydrothermal
aging of a range of Fe/zeolites with different pore structures.4
Notable deterioration of Fe/BEA activity was noted after
hydrothermal aging. Therefore, one goal of this study was to
improve the NH3-SCR activity of zeolite catalysts, even after
aging. In addition, this study also aimed to investigate the
catalytic efficiency of Fe/BEA zeolites ion-exchanged with
alkali, alkali earth and RE metals to increase our understanding
of zeolite chemistry.
for selective catalytic reduction of NO by NH3.
Nitrogen oxides NOx (NO + NO2) emitted from mobile and
stationary sources are directly harmful to human health and
are a major source of acid rain and photochemical smog. One
of the most efficient technologies for removing NO is selective
catalytic reduction (SCR) by ammonia (4NO + 4NH3 + O2 -
4N2 + 6H2O), a process that is currently widely used in power
plants and other stationary sources. Recently, there has been
significant interest in adapting this technology to mobile diesel
vehicles in order to meet stringent NOx emission regulations.1
Fe-loaded zeolites have been extensively investigated as SCR
catalysts because of their high activity and the harmless nature
of Fe.2 To effectively reduce NOx emissions from diesel
engines, SCR catalysts must exhibit high activity even at low
temperatures for cold start conditions in cold weather regions.
SCR catalysts must also possess high durability in high
temperature environments as under certain circumstances,
such as during the regeneration of the diesel particulate filter
(DPF), gas exhaust temperatures can reach over 973 K. Thus
far, most studies have focused mainly on improving catalyst
performance at low temperatures by upgrading catalyst
materials and optimizing preparation methods, or by controlling
reaction conditions.3–6 Efforts to improve the stability of catalysts
at high temperatures have been insufficient, in spite of the
importance of this aspect of thermal stability to catalyst longevity.
In general, the thermal stability of H-form zeolites is
improved by ion-exchange of alkali, alkali earth or rare earth
(RE) metals as these metals suppress conversion of Al sites
from framework positions to extra-framework positions,
a process known as dealumination.7,8 Since dealumination
occurs easily over proton sites in zeolites, ion-exchange of
protons with metal cations is an effective method for reducing
dealumination. However, basic metal exchange with alkali and
alkali earth metals sometimes reduces catalytic activity as
Fe/BEA catalysts were prepared by CVD of FeCl3 to
H-BEA (Si/Al2 = 37, Tosoh HSZ-940HOA) as described in
detail previously.3,4 Briefly, FeCl3 vapour was sublimed onto
dehydrated H-BEA under N2 flow at 873 K. The sample was
then washed twice and calcined at 823 K in air. The addition
of the secondary metals was conducted by a liquid ion-
exchange method. Fe/BEA (3.5 g) was suspended in a metal
ion solution (0.1 M, 250 ml), stirred for 12 h at room
temperature, then filtered and calcined at 673 K. Nitrate salts
were used for RE metal exchanges and acetate salts were used
for alkali and alkali earth metal exchanges. All samples were
aged at 973 K for 5 h under 3% H2O/air condition. Metal
contents were determined by inductively coupled plasma-
atomic emission spectroscopy (ICP-AES).
A catalytic activity test (NH3-SCR) was conducted using a
fixed-bed flow reactor equipped with a chemiluminescent NOx
analyzer (Best Sokki, Bex-5900C). The feed gas (0.1% NO,
0.1% NH3, 8% O2, 10% CO2, 8% H2O, and N2 as the
balance) was introduced at 5 l minÀ1 to 1.0 g of pelletized
sample. NO conversion was measured under steady-state
conditions at a pre-determined temperature.
TOYOTA Central R&D Laboratories, Inc.,
F41-1 Yokomichi, Nagakute, Aichi 480-1192, Japan.
E-mail: iwasaki@mosk.tytlabs.co.jp; Fax: +81 561 63 6712;
Tel: +81 561 71 7810
w Electronic supplementary information (ESI) available: UV-Vis
spectra of alkali and alkali earth metal doped Fe/BEA and XRD
spectra of Nd doped Fe/BEA. See DOI: 10.1039/c1cc00108f
c
3966 Chem. Commun., 2011, 47, 3966–3968
This journal is The Royal Society of Chemistry 2011

View Article Online
27Al-NMR, FT-IR and UV-Vis analyses were carried out to
characterize the catalysts. 27Al magic-angle spinning NMR
spectra were obtained using a Varian VNMRS-400 spectrometer
at a resonance frequency of 104.2 MHz. The sample rotation
speed was 15.0 kHz. Chemical shifts were referenced relative
to an aqueous Al(NO3)3 solution. Samples were fully hydrated
overnight in a controlled environment before NMR measure-
ments. Diffuse reflectance FT-IR spectra were obtained using
a Thermo Nicolet NEXUS 670 spectrometer at a resolution of
4 cmÀ1. Samples were pre-treated in situ at 773 K for 1.5 h
under O2/N2 (10 : 90) flow, and spectra were recorded at 323 K.
Diffuse reflectance UV-Vis spectra were obtained using a
JASCO V-560 spectrometer at a band width of 2 nm. The
spectra were recorded at ambient atmospheric pressure and
converted with a Kubelka–Munk function.
The metal content of all prepared samples is shown in Fig. 1.
The amount of Fe losses by the sequential ion-exchange were
very little (less than 10%), while the amount of added secondary
metals varied widely by metal type. The metal content
increased in the order alkali metals > alkali earth metals >
RE metals, which is in the reverse order of the metal cation
charges, +1 > +2 > +3. This result is consistent with the
fact that higher charged ions, such as RE metals, are more
difficult to exchange.11 In some of the samples studied here,
the sum of Fe and secondary metals added was larger than the
Al content. Our previous study revealed that a-Fe2O3 was a
major species in Fe/zeolites prepared by CVD and that the
amount of ion-exchanged Fe sites was lower than the theoretical
exchangeable value.3 Therefore, Fe/BEA should still have free
sites available for exchange. Additionally, a replacement of
ion-exchanged Fe with secondary added metals might occur
for alkali or alkali earth exchanged samples. This reverse-
substituted Fe was expected to aggregate and consequently
convert to Fe2O3 particles, which was verified by UV-Vis
spectra (Fig. S1 in ESIw).
Fig. 2a and b show the NO reduction activity for several
samples before and after hydrothermal aging. In fresh samples,
Fe/BEA without addition of secondary metals showed the
highest activity, whereas the addition of alkali and alkali
earth metals significantly reduced the activity. Therefore, the
addition of secondary metals did not improve the activity
of fresh Fe/BEA. After aging, the activity of all samples
decreased relative to their respective fresh activities. However,
it should be noted that the aged, Ce-exchanged sample showed
Fig. 2 Temperature dependence of NO conversion of fresh (a) and
aged (b) samples; comparison of NO conversion of all samples at
523 K and 573 K (c); and relationship between NO conversion and
cation radii of RE metals (d). Feed: 0.1% NO, 0.1% NH3, 8% O2,
10% CO2, 8% H2O, and N2 as the balance. Flow rate: 5 l minÀ1
.
Catalyst weight: 1 g.
higher activity than the aged sample with no secondary metal,
indicating that Ce addition increased the catalyst hydrothermal
stability.
Fig. 2c shows NO conversion at 523 and 573 K for all aged
samples. Although all the alkali and alkali earth samples were
less active, some of the RE samples (Ce, Nd, Sm, Gd and Tb)
exhibited higher activity than the aged sample with no secondary
metal. However, the order of activity of the RE samples did
not always correspond with the RE content (Fig. 1). Interestingly,
a local maximum was observed when NO conversion was
plotted as a function of RE cation radius (Fig. 2d), indicating
that there may be an optimum cation size for attaining higher
thermal stability.
To investigate the catalytic states, 27Al-NMR, FT-IR and
UV-Vis were applied to Fe/BEA and Fe–Nd/BEA as repre-
sentative high stability samples. Fig. 3a shows 27Al-NMR
spectra of fresh and aged samples. A distinct signal at
approximately 54 ppm, attributed to tetrahedrally coordinated
Al atoms in lattice positions, was detected for both samples.
The peaks of fresh samples were broader than those of aged
samples due to the paramagnetic nature of the iron nuclei
present in the zeolites. Additionally, a small signal assigned to
extra-framework Al with octahedral coordination was detected at
0 ppm on fresh samples.12 After aging, the sharp resonance at
0 ppm was transformed to a broad signal at approximately the
same value. Although difficult to ascertain from the results
Fig. 1 Loaded metal contents of Fe and secondary added metals
measured by ICP-AES. Parent zeolite was BEA type (Si/Al2 = 37).
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 3966–3968 3967

View Article Online
as RE metal loaded zeolites, such as Nd/zeolites, have absorption
bands in this region,15 it is difficult to directly compare
Fe/BEA and Fe–Nd/BEA spectra. Nevertheless, the peak
shape of Fe–Nd/BEA below 300 nm was similar to that of
Fe/BEA, suggesting that equivalent Fe states are maintained
even after Nd addition. An increase in absorption bands above
300 nm indicates that the oligomerization of Fe species
progressed after aging. It should be noted, however, that the
difference was larger for Fe/BEA than Fe–Nd/BEA, suggesting
that the addition of Nd may have inhibited the oligomerization
of Fe species. This is reasonable because dealumination results
in a loss of the ion-exchanged Fe species that lead to an
increase in the Fe oligomer. This hypothesis is also supported
by XRD data of the a-Fe2O3 peak intensity (Fig. S2 in ESIw).
In conclusion, an enhancement of the hydrothermal stability
of Fe/BEA was achieved by the sequential ion-exchange of rare
earth metals. The enhancement was dependent on the ionic
radius of the rare earth metals. When using metals with radii of
1.05–1.15 A, NO reduction activities after hydrothermal aging
were improved versus aged Fe/BEA. Spectroscopic character-
izations revealed that RE addition inhibited dealumination that
leads to the formation of terminal silanol and Fe oligomers.
Inhibition of dealumination in this manner may be applied
effectively to the practical development of automotive catalysts
in the future.
The authors acknowledge Dr Hideyuki Oka (Tosoh Analysis
and Research Center) for 27Al NMR measurements and
Mr Satoru Kosaka (Toyota Central R&D Labs.) for ICP-AES
measurements.
Fig. 3 27Al NMR (a), FT-IR (b), and UV-Vis (c) spectra of fresh and
Notes and references
aged Fe/BEA and Fe–Nd/BEA.
1 T. V. Johnson, Int. J. Engine Res., 2009, 10, 275.
2 (a) R. Q. Long and R. T. Yang, J. Am. Chem. Soc., 1999, 121,
5595; (b) A. Z. Ma and W. Grunert, Chem. Commun., 1999, 71.
3 (a) M. Iwasaki, K. Yamazaki, K. Banno and H. Shinjoh, J. Catal.,
2008, 260, 205; (b) M. Iwasaki and H. Shinjoh, J. Catal., 2010, 273,
29.
4 M. Iwasaki, K. Yamazaki and H. Shinjoh, Appl. Catal., B, 2011,
102, 302.
5 P. Forzatti, I. Nova and E. Tronconi, Angew. Chem., Int. Ed.,
2009, 48, 8366.
presented here, the possibility of the formation of extra-
framework mixed iron–aluminium–oxo species should be con-
sidered.12 The areas of the resonance at 54 ppm from tetrahedral
Al for the aged samples were 83% and 91% relative to the
respective fresh samples for Fe/BEA and Fe–Nd/BEA respec-
tively. This result indicates that less dealumination was observed
for Fe–Nd/BEA than Fe/BEA.
6 (a) M. Iwasaki and H. Shinjoh, Appl. Catal., A, 2010, 390, 71;
(b) M. Iwasaki, K. Yamazaki and H. Shinjoh, Appl. Catal., A,
2009, 366, 84.
Fig. 3b shows FT-IR spectra from the OH stretching region
for fresh and aged samples. For the fresh samples, three main
peaks were observed at 3735, 3682 and 3610 cmÀ1, assigned to
terminal silanols, OH on Fe ion species and bridging OH
7 F. E. Trigueiro, D. F. J. Monteiro, F. M. Z. Zotin and E. Falabella
Sousa-Aguiar, J. Alloys Compd., 2002, 344, 337.
8 (a) G. Yang, Y. Wang, D. H. Zhou, J. Q. Zhuang, X. C. Liu,
X. W. Han and X. H. Bao, J. Chem. Phys., 2003, 119, 9765;
(b) G. Yang, Y. Wang, D. H. Zhou, X. C. Liu, X. W. Han and
X. H. Bao, J. Mol. Catal. A: Chem., 2005, 237, 36.
9 P. Budi and R. F. Howe, Catal. Today, 1997, 38, 175.
10 J. Y. Yan, W. M. H. Sachtler and H. H. Kung, Catal. Today, 1997,
33, 279.
11 A. V. Kucherov, C. P. Hubbard, T. N. Kucherova and M. Shelef,
Appl. Catal., B, 1996, 7, 285.
12 E. J. M. Hensen, Q. Zhu, M. Hendrix, A. R. Overweg,
P. J. Kooyman, M. V. Sychev and R. A. van Santen, J. Catal.,
2004, 221, 560.
13 A. Vimont, F. Thibault-Starzyk and J. C. Lavalley, J. Phys. Chem.
B, 1999, 104, 286.
14 M. Schwidder, M. S. Kumar, K. Klementiev, M. M. Pohl,
A. Bruckner and W. Grunert, J. Catal., 2005, 231, 314.
15 W. Xiaoning, Z. Zhen, X. Chunming, D. Aijun, Z. Li and
J. Guiyuan, J. Rare Earths, 2007, 25, 321.
(Bronsted) sites, respectively.13 For the aged samples, the
¨
magnitude of the bands at 3610 cmÀ1 decreased significantly.
Additionally, new peaks at 3780 cmÀ1, assigned to OH groups
attached to a tri-coordinated aluminium atom partially
disconnected from the framework, were observed.13 The intensity
of the band at 3735 cmÀ1 for aged Fe/BEA samples was higher
than aged Fe–Nd/BEA samples. This result suggests that the
formation of terminal silanols is suppressed more for Fe–Nd/BEA
than Fe/BEA, which is consistent with the NMR results.
Fig. 3c shows the UV-Vis spectra of fresh and aged samples.
For Fe/zeolites, absorption bands below 300 nm have been
assigned to ion-exchanged Fe, while bands above 300 nm have
been assigned to oligomeric species and aggregated Fe2O3
particles located at the outside of the zeolite pores.14 However,
c
3968 Chem. Commun., 2011, 47, 3966–3968
This journal is The Royal Society of Chemistry 2011