Experimental Evidence of the Mechanism of Selective Catalytic Reduction of NO with NH3 over Fe-Containing BEA Zeolites

Article abstract of DOI:10.1002/cssc.201801883

Various temperature-programmed techniques were used as tools in mechanistic studies of selective catalytic reduction (SCR) of NO with ammonia in the presence of Fe-containing BEA zeolites. Moreover, FTIR studies of adsorbed NH₃ and NO were conducted to determine the interactions of reactants with the catalyst surface. Iron was introduced into BEA zeolite by three different methods: i) two-step post-synthesis; ii) conventional wet impregnation; iii) ion exchange. The catalytic activity was dependent on the method used for iron introduction. The reactivities of NH₃ and NO adsorbed on iron-modified zeolites obtained by impregnation and ion-exchange methods were higher than those measured for the catalyst obtained by a two-step post-synthesis method. The activity of Fe-containing zeolites in SCR was related to the form of deposited iron species, as well as to the nature, strength, and concentration of acid sites. Possible reaction pathways of NO reduction over the FeBEA zeolite catalysts were presented and discussed.

Full text of DOI:10.1002/cssc.201801883

DOI: 10.1002/cssc.201801883
Full Papers
Experimental Evidence of the Mechanism of Selective
Catalytic Reduction of NO with NH3 over Fe-Containing
BEA Zeolites
[a, b]
Małgorzata Rutkowska,[a] Barbara Gil,[a] Bartosz Marszałek,[a]
´
Paweł Boron,
Lucjan Chmielarz,*[a] and Stanislaw Dzwigaj*[b]
Various temperature-programmed techniques were used as
tools in mechanistic studies of selective catalytic reduction
(SCR) of NO with ammonia in the presence of Fe-containing
BEA zeolites. Moreover, FTIR studies of adsorbed NH3 and NO
were conducted to determine the interactions of reactants
with the catalyst surface. Iron was introduced into BEA zeolite
by three different methods: i) two-step post-synthesis; ii) con-
ventional wet impregnation; iii) ion exchange. The catalytic
activity was dependent on the method used for iron introduc-
tion. The reactivities of NH3 and NO adsorbed on iron-modified
zeolites obtained by impregnation and ion-exchange methods
were higher than those measured for the catalyst obtained by
a two-step post-synthesis method. The activity of Fe-contain-
ing zeolites in SCR was related to the form of deposited iron
species, as well as to the nature, strength, and concentration
of acid sites. Possible reaction pathways of NO reduction over
the FeBEA zeolite catalysts were presented and discussed.
Introduction
Selective catalytic reduction (SCR) of NOx with ammonia, also
known as the deNOx process, is a widely used technology for
diesel engine vehicles to reduce NOx emissions (NO, NO2) to
N2.[1] In this method, NOx is catalytically reduced with NH3,
which is formed on board by urea hydrolysis.[2,3] The deNOx
process has also been used for conversion of NOx emitted by
stationary sources, such as electric power stations, for nearly
five decades. Unfortunately, the V2O5/WO3/TiO2-based catalysts
that are commercially used for this process have inadequate
stability at the higher temperatures encountered in heavy vehi-
cle operation and are not suitable for systems including a par-
ticulate filter that actively regenerates at high temperatures.
Hence, the focus of SCR research has now shifted to Fe- and
Cu-based zeolite catalysts, which were reported to be very
efficient in NOx reduction at high space velocities.[4–10]
catalytic activity in SCR of NO.[11–19] As reported by Li and co-
workers,[11,12] Al-free Fe-beta zeolite exhibited good catalytic ac-
tivity in NH3-SCR at temperatures higher than 573 K, compara-
ble with Fe-H-beta prepared by a solid-state ion-exchange
route, in the presence of SO2. Moreover, Al-free Fe-beta exhib-
ited better durability in NH3-SCR than Fe-H-beta and a NO con-
version of higher than 90% at the reaction temperature of
673 K was maintained for over 400 h. As suggested by the au-
thors, the remarkable durability of the Al-free Fe-beta zeolite
was due to the absence of framework aluminum species and
the stabilization of active iron species in the zeolite framework.
Despite the large number of reported studies on the SCR pro-
cess, its mechanism and the role of various forms of metallic
species in the catalytic process are still under debate. Several
studies on the mechanism and kinetics of the SCR process
over various Fe-zeolite catalysts have been already report-
ed.[9,20–27] Some reports have suggested that the SCR reaction
over FeBEA catalysts proceeds according to a redox mecha-
nism whereby NOx and NH3 are adsorbed on neighboring iron
sites and FeIII is reduced to FeIIÀNO. The re-oxidation of iron
sites by oxygen is thought to occur. Sun et al.[28] presented a
mechanistic model for the Fe-ZSM-5 catalysts, in which the
preferred path for NOx reduction with ammonia occurs
through formation of ammonium nitrite, which decomposes to
N2 and H2O. Some reports have suggested that reduction of
NO and NO2 by NH3 involves the formation of HNO2 and
HNO3.[29–32] Both NOx and NH3 are known to adsorb on ion-ex-
changed zeolites.[28,30,33,34] Ammonia molecules are known to
react with protons of H-form zeolites to generate ammonium
ions, which are stable at high temperatures. The reaction be-
tween NO and O2 on Fe-ZSM-5 can produce various species,
Extensive studies on zeolite-based SCR catalysts have been
conducted to determine the nature of active sites and their
´
[a] Dr. P. Boron, Dr. M. Rutkowska, Dr. B. Gil, Dr. B. Marszałek,
Prof. L. Chmielarz
Faculty of Chemistry, Jagiellonian University in Krakꢀw
Gronostajowa 2, 30-387 Krakꢀw (Poland)
E-mail: chmielar@chemia.uj.edu.pl
´
[b] Dr. P. Boron, Prof. S. Dzwigaj
Laboratoire de Rꢁactivitꢁ de Surface
Sorbonne Universitꢁ-CNRS, UMR 7197, 75005, Paris (France)
E-mail: stanislaw.dzwigaj@upmc.fr
The ORCID identification number(s) for the author(s) of this article can
be found under:
https://doi.org/10.1002/cssc.201801883.
This publication is part of a Special Issue focusing on “Celebrating 50
years of the Jerzy Haber Institute of Catalysis and Surface Chemistry of
the Polish Academy of Sciences”. A link to the issue’s Table of Contents
will appear here once it is complete.
À
such as NO, NO2, and nitrate (NO3 ).[28,33,34] Moreover, some re-
ChemSusChem 2018, 11, 1 – 15
1
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
ports have claimed that NO can be adsorbed on FeII sites,
The presented studies, related to the mechanisms of
NH3-SCR over FeSiBEA, FeAlBEA and Fe(IE)BEA zeolites, were
conducted with the application of two groups of methods:
spectroscopy, for identification of the possible species formed
on the catalyst surface during catalytic reaction, and tempera-
ture-programmed desorption/reaction, for determination of
the interactions of reactants with the catalysts and the reactivi-
ties of surface species under reaction conditions. The combina-
tion of these two complementary methods is important for
better understanding of reactant interactions and conversions
on the surface of the studied catalysts. A very important point
in the presented studies was simultaneous or sequential ad-
sorption of reactants (a rather unique method), which gave
new insight into the interaction of reactants adsorbed on the
catalyst surface and their reactivity. Recently, we reported a
similar study on BEA zeolite modified with cobalt species by
using three different preparation methods.[40]
À
whereas NO2 and NO3 species are adsorbed on FeIII sites.[33]
When NO is adsorbed in the presence of oxygen, NO2 may be
formed.[30,33]
À
The aforementioned studies suggest that NO, NO2, NO3
,
+
and NH4 species may play an important role in the SCR reac-
tion performed in the presence of the Fe-exchanged zeolite
catalysts. To obtain clearer information concerning the nature
of the active species and reaction intermediates involved in
the reaction, as well as the nature of the actual catalyst active
sites, spectroscopic and surface reactivity techniques were ap-
plied. It has often been reported that NO does not appreciably
adsorb on the catalyst surface, whereas ammonia is strongly
adsorbed.[35–39] Moreover, it was suggested that the SCR reac-
tion involves strongly adsorbed ammonia species and gaseous
or weakly held NO molecules. It was also suggested that the
rate of the SCR reaction depends on ammonia surface concen-
tration for low NH3 coverage, whereas a much weaker depend-
ence, if any, is evident at higher ammonia coverage.[35,36] In
spite of a host of experimental studies focused on mechanistic
issues and reaction pathways, the SCR mechanism on FeBEA
zeolite catalysts is still the subject of debate.
Results and Discussion
As reported previously,[17] all of the materials discussed herein
have high BET surface areas of 630–790 m2 gÀ1 and micropore
volumes of 0.21–0.24 mLgÀ1 (Table 1). These values are typical
of the BEA structure and indicate that textural properties of
BEA zeolite are preserved upon dealumination and introduc-
tion of iron into the zeolite structure by different methods. The
introduction of 1.0 wt% of Fe into SiBEA by the two-step post-
synthesis procedure leads to an increase in the d302 spacing of
FeSiBEA zeolite, indicating expansion of the unit cell as a result
of the reaction of iron ions with OH groups at vacant T-atom
sites (T=Si or Al) and their incorporation into the framework
of BEA zeolite, in agreement with an earlier report.[17] Impreg-
nation and ion-exchange methods do not lead to such signifi-
cant increase in the d302 spacing as observed for FeSiBEA zeo-
lite. A small increase in the d302 spacing for FeAlBEA and Fe(IE)-
BEA suggests the reaction of iron ions with OH groups of
vacant T-atom sites and their partial incorporation into the
framework of FeAlBEA and Fe(IE)BEA zeolites. In all obtained
catalysts, iron is present mainly as pseudo-tetrahedral FeIII.
However, in the Fe(IE)BEA sample obtained by ion-exchange
method, some amounts of FeIII in octahedral coordination are
present, as evidenced by the combined use of diffuse reflec-
tance (DR) UV/Vis, 57Fe Mçssbauer, and X-ray photoelectron
spectroscopy (XPS) techniques.[17]
This study is a continuation of the previous studies,[17] which
showed high catalytic activity of BEA zeolite modified with
iron by different methods (two-step post-synthesis, wet im-
pregnation, and conventional ion exchange) in the deNOx pro-
cess performed in the presence of oxygen. The main goal of
the presented studies is comparison of the catalytic activities
of various iron species and recognition of the mechanism of
the SCR process over FeBEA zeolite catalysts. Fe-modified zeo-
lites have been studied for decades. However, most past stud-
ies focused on Al-containing zeolites modified with iron by
ion-exchange and classical impregnation methods resulting in
deposition of iron in extra-framework positions. Herein, we
present another, new group of zeolite materials containing
iron incorporated into the zeolite framework. This type of
FeSiBEA materials was prepared in the previous work and com-
pared to the FeAlBEA and Fe(IE)BEA materials obtained by clas-
sical impregnation and ion-exchange method with respect to
properties and catalytic performance in NH3-SCR. Studies fo-
cusing on the application of BEA with iron incorporated into
the zeolite framework by a two-step post-synthesis method in
NH3-SCR are very limited.[11,12] This manuscript is focused on
the determination of the reaction mechanisms for Fe-zeolites
obtained by different methods or, more precisely, on the deter-
mination of differences in the reaction mechanisms.
Moreover, the characterization of the FeBEA zeolites pre-
pared according to different previously reported meth-
Table 1. The physicochemical properties of FeSiBEA, FeAlBEA and Fe(IE)BEA zeolite catalysts.
Sample
Si/Al Fe content
[wt%]
Specific surface area
Micropore area
Micropore volume
[cm3 gÀ1
]
FeIII geometry
NH3 desorp-
tion
[m2 gÀ1
]
[m2 gÀ1
]
HAlBEA
SiBEA
FeSiBEA 1000 1.0
FeAlBEA
Fe(IE)BEA
12.5 0
728
790
714
647
631
565
626
538
522
511
0.22
0.24
0.21
0.21
0.21
0.15
2.61
1.01
2.77
1000
0
pseudo-tetrahedral
pseudo-tetrahedral
12.5 1.0
12.5 1.5
pseudo-tetrahedral and octa- 3.78
hedral
&
ChemSusChem 2018, 11, 1 – 15
www.chemsuschem.org
2
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
ods[17,41–43] by using spectroscopic methods, such as DR UV/Vis,
EXAFS, XANES, 57Fe Mçssbauer, and XPS, allowed us to deter-
mine the state and distribution of Fe species in these materials.
In FeSiBEA zeolites prepared by a two-step post-synthesis
method, iron atoms were present as framework pseudo-tetra-
hedral FeIII. The DR UV/Vis spectra of FeSiBEA samples with low
iron content (0.5 and 1.0 wt%) included a band centered at
about 253–276 nm assigned to oxygen-to-metal charge trans-
fer involving framework pseudo-tetrahedral FeIII species.[42–47]
The absence of a broad DR UV/Vis band near 500 nm suggests
that FeOx oligomers are not present in these samples.[46–48] It
should be noted that during preparation of these samples by
two-step post-synthesis method, almost all iron was incorpo-
rated into the zeolite framework. Therefore, these samples are
white. The EXAFS and XANES results obtained for FeSiBEA sam-
ples with 0.5 and 1.0 wt% Fe confirm that, at low iron content,
these samples contain only pseudo-tetrahedral FeIII species. In
contrast, for FeAlBEA and Fe(IE)BEA samples, pseudo-tetrahe-
dral and octahedral FeIII species were detected in framework
and extra-framework positions.[44] It is important to mention
here that two kinds of pseudo-tetrahedral FeIII species were
detected in all studied FeBEA zeolites with low Fe content, as
evidenced by 57Fe Mçssbauer spectroscopy.[49]
tion methods were more efficient at elevated temperatures
(T>623 K).
We next studied the influence of water vapor on the catalyt-
ic performance of the zeolitic samples by sequential switching
from dry to wet reaction mixtures (Figure 1). For all samples,
introduction of water vapor into reaction mixture resulted in a
decrease in NO conversion by about 2.5–3.0%. However, in the
next step, after exchange from wet to dry reaction mixture,
the initial NO conversion level was fully restored for the
sample with iron incorporated into the zeolite framework
(FeSiBEA), whereas for the catalysts containing iron in extra-
framework positions [FeAlBEA, Fe(IMP)BEA], the NO conversion
was lower by about 1.0–1.5% compared to the initial level.
Very similar effects were observed after repetition of this pro-
cedure—the level of NO conversion was very similar to the ini-
tial one for FeSiBEA, whereas for FeAlBEA and Fe(IMP)BEA it
decreased by about 2.0–2.5% in comparison to the initial level.
This experiment shows that the zeolite sample containing iron
incorporated into framework positions is more resistant to
water vapor in the reaction mixture than zeolites with iron in
extra-framework positions. This could be explained by the
higher stability of iron ions in the zeolite framework than that
of iron species located in extra-framework positions, which is
characterized by higher surface mobility and therefore also the
tendency to undergo aggregation.[11] It should be also noted
that, after exchanging the flow from wet to dry reaction mix-
ture, the level of NO conversion was completely or partially re-
The FeAlBEA and Fe(IE)BEA samples have a significantly
higher acidity than FeSiBEA, as evidenced by FTIR spectroscopy
of adsorbed CO or pyridine and temperature-programmed de-
sorption (TPD) of ammonia. This effect could be assigned to
the presence of framework pseudo-tetrahedral ꢀAlIIIÀO(H)ÀSiꢀ stored. This effect is possibly related to the competition of am-
and ꢀFeIIIÀO(H)ÀSiꢀ sites, as well as extra-framework iron spe-
cies in the former and only ꢀFeIIIÀO(H)ÀSiꢀ in the latter
sample. The most important physicochemical properties of the
samples are given in Table 1.
monia and water molecules for the same adsorption sites. In
the case of the wet reaction mixture, the contribution of ad-
sorption sites occupied by water molecules is larger in compar-
ison to the dry reaction mixture and therefore the amount of
adsorbed ammonia is lower. The introduction of water vapor
into reaction mixture only slightly influenced the reaction se-
lectivity to N2.
The results of catalytic tests in the process of selective cata-
lytic reduction of NO with ammonia for FeSiBEA, FeAlBEA, and
Fe(IE)BEA are given in Table 2. The catalytic activities of all the
Fe-containing zeolites are high, although it should be noted
that the catalyst obtained by ion-exchange method is the
most active in the low-temperature range (T<553 K), whereas
those prepared by two-step post-synthesis and wet impregna-
The influence of SO2 on the catalytic performance of the
zeolite samples was studied by sequential exchange flow of
the reaction mixture for a gas mixture containing 2000 ppm of
SO2 diluted in helium (Figure 2). Treatment of the catalyst sam-
Table 2. The results of catalytic tests (XNO =conversion of nitrogen oxide, SN =selectivity to nitrogen) in SCR NO with NH3 over FeSiBEA, FeAlBEA, and
2
Fe(IE)BEA zeolite catalysts.
T [K]
SiBEA
HAlBEA
FeSiBEA
FeAlBEA
Fe(IE)BEA
XNO [%]
SN [%]
XNO [%]
SN [%]
XNO [%]
SN [%]
XNO [%]
SN [%]
XNO [%]
SN [%]
2
2
2
2
2
373
423
473
523
573
623
673
723
773
823
0
0
0
1.5
2.8
4.8
3.7
1.2
0
–
–
–
–
–
–
–
–
–
–
0
0
0
0
0
–
–
–
–
–
0
0
–
–
0
0
–
–
0
–
12
67
93
95
94
93
89
79
76
96
96
96
95
95
95
94
94
93
5.7
21
62
81
92
93
90
78
94
94
94
94
95
95
93
92
6.5
63
95
96
96
97
97
93
95
95
96
96
97
97
96
96
0
–
11
38
68
71
78
80
80
81
0
[a] Reaction conditions: 0.1 g of catalyst outgassed in a flow of pure helium at 823 K for 30 min. [b] Reaction mixture: [NO]=[NH3]=0.25 vol.%, [O2]=
2.5 vol.%, and [He]=97 vol.%. [c] Total flow rate: 40 mLminÀ1; WHSV: 24000 mLhÀ1 gÀ1
.
ChemSusChem 2018, 11, 1 – 15
www.chemsuschem.org
3
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Figure 1. Results of catalytic tests in the presence of dry and wet reaction
mixtures. Reaction conditions: 0.1 g of catalyst outgassed in a flow of pure
helium at 823 K for 30 min. Reaction mixture: [NO]=[NH3]=0.25 vol.%,
[O2]=2.5 vol.%, and [H2O]=0 or 3.0 vol.% diluted in helium; total flow rate:
Figure 2. Results of catalytic tests before and after treatment of the catalyst
with a gas mixture of 2000 ppm SO2 in helium (40 mLminÀ1; 30 min.). Reac-
tion conditions: 0.1 g of catalyst outgassed in a flow of pure helium at 823 K
for 30 min. Reaction mixture: [NO]=[NH3]=0.25 vol.%, [O2]=2.5 vol.%
40 mLminÀ1; WHSV: 24000 mLhÀ1 gÀ1
.
diluted in helium; total flow rate: 40 mLminÀ1; WHSV: 24000 mLhÀ1 gÀ1
.
ples with such gas mixture decreases NO conversion by about
2.0–3.0%. Partial recovery of NO conversion was observed only
for the FeSiBEA catalyst. Treatment of the catalysts with SO2
only slightly influenced selectivity to N2.
(1.1 mmolmÀ2) suggesting the presence of additional acid sites
generated by iron species.[16–19] The desorption profile of FeSi-
BEA consists of at least two poorly resolved maxima at 430
and 550 K (Figure 3). It should be noted that ammonia remains
chemisorbed on FeSiBEA at higher temperatures than on
SiBEA, which suggests the presence of relatively strong acid
sites. The NH3-TPD profile of HAlBEA consists of two maxima at
450 and 610 K (Figure 3) and the concentration of acid sites
determined for the sample is 2.61 mmolmÀ2. Introduction of
iron into this sample by wet impregnation method resulted in
an increase in the concentration of acid sites to 2.77 mmolmÀ2
and a shift of the desorption maxima to 470 and 630 K, which
could be explained by increased strength of the acid sites in
FeAlBEA.[16,17,19] In contrast, for the FeSiBEA sample the NH3 de-
sorption maxima are shifted to lower temperatures than for
HAlBEA, since all acid ꢀAlIIIÀO(H)ÀSiꢀ sites were removed by
acid leaching. The shape of the desorption profile of Fe(IE)BEA
is very similar to that recorded for FeAlBEA, but the acid site
concentration in this sample is significantly higher
(3.78 mmolmÀ2). The acidity of the FeAlBEA and Fe(IE)BEA cata-
lysts is related to the presence of Al and Fe in the zeolite struc-
The mechanism of NO reduction with ammonia over Fe-
doped zeolites depends on many factors, including the type of
zeolite support (zeolite structure, Si/Al ratio) and the form of
deposited iron species (location, coordination, and aggrega-
tion), and is still a subject of debate. In the presented study,
various temperature-programmed and spectroscopic methods
were used to investigate the SCR mechanism over BEA zeolites
doped with iron in various ways.
NH3-TPD
The ammonia desorption profile obtained for SiBEA zeolite
(Figure 3) gives a symmetric maximum at about 400 K. The
SiBEA sample has low acidity, as proven by the low concentra-
tion of adsorbed ammonia (0.15 mmolmÀ2). The introduction of
iron into SiBEA by a two-step post-synthesis method, resulting
in FeSiBEA, significantly increased ammonia uptake
&
ChemSusChem 2018, 11, 1 – 15
www.chemsuschem.org
4
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
NH3 coordinatively bonded to Lewis acid sites, which may orig-
inate from Al or Fe. The characteristic bands of adsorbed am-
monia are present in the spectra of the catalysts evacuated at
573 K, in full agreement with the results of NH3-TPD. Moreover,
in the case of Fe(IE)BEA, which was the most active catalyst, a
higher contribution of Lewis acid sites was found than for
other samples. Thus, a significant role of this type of acidity on
the activity of FeBEA catalysts cannot be excluded.
NO-TPD
Analysis of NO desorption profiles obtained for HAlBEA, FeSi-
BEA, FeAlBEA, and Fe(IE)BEA (Figure 5) shows that amounts of
chemisorbed NO, which, respectively are 0.011, 0.007, 0.021,
and 0.018 mmolmÀ2, are significantly lower than those of
chemisorbed NH3. The NO-TPD profile of nitrogen oxide evolu-
tion from HAlBEA includes two desorption peaks at tempera-
tures of 395 and 558 K. Nearly the same NO desorption
maxima (at T=395 and 560 K), albeit with significantly higher
intensity, were found in the profile of Fe(IE)BEA, whereas depo-
sition of iron into HAlBEA by impregnation method to give
Figure 3. NH3-TPD profiles obtained for SiBEA, HAlBEA, FeSiBEA, FeAlBEA,
and Fe(IE)BEA.
ture, which may form either Brønsted or Lewis acid sites, as
well as extra-framework Fe species that generate Lewis type
acidity.[16,17,19] Comparison of the ammonia desorption profiles
recorded for FeSiBEA (containing only iron) and FeAlBEA and
Fe(IE)BEA (containing both iron and aluminum) leads to the
conclusion that iron generates centers with lower acid strength
than aluminum.
FTIR of samples with adsorbed NH3
The FTIR spectra recorded for the FeSiBEA, FeAlBEA and Fe(IE)-
BEA catalysts with adsorbed NH3 are shown in Figure 4. Am-
monia adsorption in all the samples results in the appearance
of two IR maxima. The first, at 1450 cmÀ1, is characteristic of
+
NH4 species formed by the reaction between NH3 and
Brønsted acid sites. The second, at 1620 cmÀ1, is assigned to
Figure 5. NO-TPD on HAlBEA, FeSiBEA, FeAlBEA, and Fe(IE)BEA samples.
Figure 4. FTIR spectra of NH3 adsorption on Fe(IE)BEA, FeAlBEA, and FeSiBEA activated under vacuum at 723 K; all spectra at 373 K, normalized to 10 mg of
catalyst. Individual designations: d, e, and f signify NH3 adsorption on activated sample and consecutive desorption for 20 minutes at 373, 473, and 573 K,
respectively.
ChemSusChem 2018, 11, 1 – 15
www.chemsuschem.org
5
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
FeAlBEA also resulted in an increase in the intensity of NO de-
sorption peaks, which were at T=592 and 563 K, respectively.
The desorption profile of FeSiBEA consists of only one peak,
centered at about 420 K. The obtained results may suggest
that the presence of iron in pseudo-tetrahedral coordination is
responsible for weak NO adsorption (desorption below 500 K),
whereas aluminum present in the zeolite structure and possi-
bly also iron in the form of extra-framework species is attribut-
ed for both weaker (desorption below 500 K) and stronger NO
adsorption sites (desorption above 500 K).
NO+O2-TPD
For comparison, we studied the simultaneous adsorption of
both NO and O2. Desorption profiles obtained in this experi-
ment presented in Figure 6. Alongside NO, evolution of O2 was
also detected. Moreover, the amount of NO released from the
samples was significantly higher than for NO-TPD runs
(Figure 5). Desorption of small amounts of weakly bound NO
was detected at temperatures below 450 K for Fe(IE)BEA and
FeAlBEA. At higher temperatures, the studied samples gave
rise to intense and broad peaks related to the simultaneous
evolution of NO and O2. For Fe(IE)BEA and FeAlBEA, these
maxima were centered at about 580–590 K, whereas for the
FeSiBEA, this temperature was higher by about 160 K. Compar-
ison of the desorption profiles in Figure 5 and Figure 6 shows
a very important role of oxygen on the formation of surface ni-
trogen species during the adsorption process. To identify the
surface species formed under adsorption conditions, the FTIR
spectroscopy was carried out on samples with adsorbed NO or
both NO and O2.
FTIR spectroscopy of samples with adsorbed NO or NO+O2
The FTIR spectra recorded for FeSiBEA, FeAlBEA, and Fe(IE)BEA
with adsorbed NO are shown in Figure 7. There are some
common features present in the spectra of all iron-containing
samples. Nitrosyls (both mono and polynitrosyls) are believed
to be formed almost exclusively with the participation of Fe2+
ions, because NO adsorption on oxidized samples is negligi-
ble.[41] Moreover, Mul et al. claimed that interaction with Fe2+
causes stronger perturbation of the dipole moment of the ad-
sorbed NO, resulting in a much higher extinction coefficient of
the NÀO stretching maximum. The band at 1875 cmÀ1 may be
therefore assigned to the NÀO symmetric stretching of mono-
nitrosyl adsorbed on Fe2+. The band at 1845 cmÀ1 has been at-
tributed in some reports to NO adsorbed on isolated Fe2+ spe-
cies in ion-exchange positions of FeBEA.[34,42] This band may be
given another assignment, on the basis of our previous
work,[50] as well the report by Spoto et al.[51] on framework-sub-
stituted zeolites (ferrisilicates). Because this system is much
closer to that examined in the present work, this assignment
will be used here. The band at 1845 cmÀ1 is therefore assigned
to NÀO antisymmetric stretching and the maximum at
1765 cmÀ1 to antisymmetric stretching vibrations of the Fe2+
(NO)2 dinitrosyls. For each sample, the relative intensities of
the two bands remain nearly constant but may differ for differ-
Figure 6. NO+O2-TPD on HAlBEA, FeSiBEA, FeAlBEA, and Fe(IE)BEA samples.
ent samples. The relative intensities of symmetric and antisym-
metric components depend on the angle between two NO
molecules adsorbed on one iron site.
The examined samples differed not only in the intensities of
the adsorbed NO bands but also in the relative stabilities of
the dinitrosyls. The most intense NO bands were observed for
the ion-exchanged sample, whereas those for FeSiBEA were
the least intense. At moderate temperatures, for FeAlBEA and
Fe(IE)BEA the most stable NO species are mononitrosyls,
&
ChemSusChem 2018, 11, 1 – 15
www.chemsuschem.org
6
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Figure 7. FTIR spectra of NO adsorption on Fe(IE)BEA, FeAlBEA and FeSiBEA zeolites activated under vacuum at 723 K; all spectra at 373 K, normalized to
10 mg of catalyst. Individual designations: a, b, and c signify NO adsorption on activated sample and consecutive desorption for 20 minutes at 373, 473 and
573 K, respectively.
whereas for FeSiBEA dinitrosyls are more stable than Fe2+(NO).
FTIR spectroscopy showed that all forms of nitrosyls are com-
pletely removed at desorption temperatures near 600 K, which
is consistent with the results of NO-TPD experiments.
for the FeSiBEA, FeAlBEA and Fe(IE)BEA catalysts. It should be
noted that apart from NH3, NO, and O2 desorption, evolution
of N2, H2O and N2O, which are the products or side products of
the SCR process, also takes place. For the FeAlBEA and Fe(IE)-
BEA catalysts, adsorbed NO was nearly completely converted
into the SCR products N2 and H2O, with a very small contribu-
tion of N2O, which is a side reaction product. Desorption of sig-
nificant amounts of unreacted ammonia from the surface of
the FeAlBEA and Fe(IE)BEA catalysts was also detect-
ed. It shows that significantly larger amounts of am-
monia were adsorbed on the catalyst surface in com-
parison to NO and therefore only a fraction of the
adsorbed NH3 molecules were converted into the
SCR products. For both catalysts, the peak maximum
for evolution of reaction products is at T=510–
525 K, which is lower by about 60–70 K than the
temperatures of thermal decomposition of nitrates
formed on the surfaces of these samples (Figure 6).
Therefore, it could be supposed that ammonia reacts
Most of the peaks on the FTIR spectra after NO+O2 adsorp-
tion (Figure 8) are characteristic of nitrates, formed both on
the iron and on SiÀOÀSi sites.[52,53] The common peaks for all
three samples are at 1620, 1635 and 1575 cmÀ1, assigned to
with surface nitrate species with the formation of
the SCR products. Different results were obtained for
the FeSiBEA sample. In this case, the formation of ni-
trogen was not detected and desorption of ammonia
Figure 8. FTIR spectra of NO+O2 adsorption on Fe(IE)BEA, FeAlBEA and FeSiBEA zeolites
activated under vacuum at 723 K; all spectra were taken at 373 K and normalized to
10 mg of catalyst.
À
NO3 formed on the Fe centers. Their concentrations are the
highest for Fe(IE)BEA and the lowest for FeSiBEA. The band at
1400 cmÀ1 is characteristic only for the FeSiBEA sample and is
related to the presence of stable nitrates, which were not fully
desorbed even at 673 K (Figure 8). Similar results of FTIR stud-
ies were recently reported for iron-modified silica materials.[54]
The characteristic band for the Fe(IE)BEA sample at 1625 cmÀ1
may be assigned to Fe(ONO) species,[53] which are moderately
stable, and were completely desorbed at 523 K.
and NO occurred at similar temperatures to those observed in
NH3-TPD (Figure 3) and NO+O2-TPD (Figure 6) experiments.
The stability of nitrates formed on the FeSiBEA surface proba-
bly is too high for their conversion at temperatures of ammo-
nia desorption.
NH3-TPSR and NO-TPSR
The reactivity of pre-adsorbed ammonia was studied by tem-
perature-programmed surface reaction (NH3-TPSR) performed
in a continuous flow of helium-diluted NO and O2 (Figure 10).
The reaction between NO+O2 and adsorbed NH3 species re-
sults in formation of N2 and H2O. The nitrogen evolution profile
obtained for the FeSiBEA sample extends from 350 to nearly
500 K, with a maximum at 395 K. Water vapor evolution is ob-
served in a much broader temperature range (350–800 K) with
a maximum at 380 K. It should be noted that some of the
chemisorbed ammonia was not converted into the SCR prod-
ucts and desorbed from the catalyst surface over a broad tem-
In general, the FTIR spectra confirmed the trends observed
for NO+O2-TPD experiments. The low-temperature desorption
maximum observed for all the samples may be due to desorp-
tion of monodentate nitrito species [Fe(ONO)] and surface
nitrates.[53]
(NO+O2)-NH3-TPD
Figure 9 shows the results of (NO+O2)-NH3-TPD measurements
(first simultaneous adsorption of both NO and O2, then NH3)
ChemSusChem 2018, 11, 1 – 15
www.chemsuschem.org
7
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Figure 9. (NO+O2)-NH3-TPD on FeSiBEA, FeAlBEA, and Fe(IE)BEA samples.
perature range with maximum at 370 K. Thus, not all chemisor-
bed ammonia molecules are properly activated for the reaction
with NO and, according to the hypothesis postulated by Kama-
samudram et al.[22] and Klukowski et al.,[24] such inactive mole-
cules serve as an ammonia reservoir.
Figure 10. NH3-TPSR measurements on FeSiBEA, FeAlBEA, and Fe(IE)BEA
samples.
The results of NH3-TPSR studies for Fe(IE)BEA and FeAlBEA
were very similar (Figure 10). Evolution of nitrogen occurred in
the temperature range of 350–520 K with maximum at 490 K.
Profiles of NO consumption is a reflection symmetry of N2 evo-
lution profiles, which supports hypothesis that in this tempera-
ture range ammonia chemisorbed on the catalyst surface
&
ChemSusChem 2018, 11, 1 – 15
www.chemsuschem.org
8
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
reacts with NO from gas phase. The profiles of water vapor
emission are spread in the range of 350–800 K with a maxi-
mum at about 395 K. The NH3-TPSR results obtained for Fe(IE)-
BEA and FeAlBEA differ in the amount of ammonia desorbing.
In the case of the former catalyst, ammonia was totally con-
verted into the SCR products, whereas in the case of FeAlBEA
desorption of a small amount of ammonia was detected. This
is in full agreement with the results of the catalytic tests, in
which the Fe(IE)BEA sample showed the highest activity.
The reactivity of pre-adsorbed NO was studied by tempera-
ture-programmed surface reaction (NO-TPSR) performed in a
continuous flow of helium-diluted ammonia (Figure 11). It
should be noted that significant amounts of ammonia were
adsorbed on the catalyst surface just after the flow of NH3/He
gas mixture was switched on. However, during ammonia ad-
sorption, evolution of pre-adsorbed NO was not detected, sug-
gesting that the surface nitrates are highly stabile and are not
removed from the catalyst by chemisorbing ammonia. For all
the studied samples, the formation of N2 and H2O as products
of the SCR process was detected. In the case of FeSiBEA and
FeAlBEA, the evolution of nitrogen and water vapor is detected
in the range of 470–570 K, with a maximum at about 510 K.
Formation of the SCR products took place at temperatures
lower by about 80 K than the thermal decomposition of sur-
face nitrate species in the flow of helium (NO+O2-TPD;
Figure 6). Thus, it could be suggested that such surface nitrate
species are directly reduced by ammonia. It is difficult to deter-
mine whether chemisorbed ammonia or ammonia in the gas
phase is active in this specific reaction. However, comparison
of the locations of nitrogen evolution peaks in (NO+O2)-NH3-
TPD (515–518 K; Figure 9) and in NO-TPSR (507–509 K;
Figure 11) may suggest that surface nitrate species are possibly
converted by chemisorbed ammonia molecules. An additional
interesting effect related to the evolution of N2 and H2O at
temperatures above 600 K, was observed for the Fe(IE)BEA cat-
alyst. This effect is surprising because any form of surface NOx
species that are stable at such high temperatures were detect-
ed in NO+O2-TPD (Figure 6), as well as by FTIR spectroscopy
of the Fe(IE)BEA sample treated with NO+O2 (Figure 8).
For FeSiBEA, evolution of the SCR products—N2 and H2O—
was detected at temperatures above 600 K with maxima at
740 and 730 K, respectively (Figure 11). Moreover, the evolution
of NO at about 710 K shows that, in spite of such high reaction
temperatures, only some of the surface nitrate was effectively
converted by ammonia into the SCR products.
Stop-flow TPD
The type of species present on the catalyst surface under reac-
tion conditions was studied by stop-flow TPD experiments.
Firstly, isothermal catalytic testing was carried out for 1 hour at
473 K for FeAlBEA and Fe(IE)BEA or at 573 K for FeSiBEA. Sub-
sequently the sample was purged in a flow of helium and then
TPD measurements were taken. Only very small amounts of
adsorbed species were identified on the catalyst surface under
reaction conditions (Figure 12). These species were not con-
verted under reaction conditions into the SCR products and
Figure 11. NO-TPSR measurements on FeSiBEA, FeAlBEA, and Fe(IE)BEA
samples.
possibly play only the role of spectators. Chemisorbed ammo-
nia molecules were the main species present on the catalyst
surface under reaction conditions for FeSiBEA, FeAlBEA, and
Fe(IE)BEA. Ammonia desorption from FeSiBEA occurred in the
ChemSusChem 2018, 11, 1 – 15
www.chemsuschem.org
9
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
range of 600–730 K, whereas that for FeAlBEA and Fe(IE)BEA
only occurred above 630 K. These differences can be related to
the different acidities of iron and aluminum-based species. For
FeAlBEA and Fe(IE)BEA, both aluminum and iron species are
present, whereas for FeSiBEA only iron species were identified.
Evolution of NO for all the studied catalysts is detected at the
same temperatures as desorption of ammonia, although the
amount of desorbing NO is significantly lower. For FeSiBEA,
the chemisorbed NH3 and NO molecules presented low reactiv-
ity and, therefore, only small amounts of N2 were formed
during the experiment. Larger amounts of nitrogen were
formed by the reaction between NH3 and NO chemisorbed on
Figure 12. Stop-flow TPD experiments on FeSiBEA, FeAlBEA and Fe(IE)BEA samples.
&
ChemSusChem 2018, 11, 1 – 15
www.chemsuschem.org
10
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
FeAlBEA and Fe(IE)BEA. The results of the stop-flow TPD ex-
periments corroborate the results obtained by using other
methods.
activation of ammonia molecules on the catalyst surface. The
concentration of ammonia adsorbed on FeSiBEA zeolite is sig-
nificantly lower than that on zeolites containing iron in extra-
framework positions. Moreover, in the case of FeSiBEA, ammo-
nia is very weakly bound to the catalyst surface, which leads to
its desorption at relatively low temperatures. Thus, taking into
account ammonia chemisorption and activation on the catalyst
surface, again the catalyst containing iron incorporated into
the zeolite framework should be less active. Thus, according to
the proposed mechanisms, the predicted activity of the FeSi-
BEA catalyst should be lower than those of catalysts containing
iron in extra-framework positions. These predictions are in
agreement with the results of catalytic tests (Table 2).
Based on the above studies, it could be supposed that the
SCR process over the FeBEA catalysts has a very complex char-
acter and depends on the form of deposited iron species, as
well as the nature, concentration, and strength of acid sites in
the zeolite. The -FeAlBEA and Fe(IE)BEA samples with higher
acidity, assigned mainly to the presence of aluminum in the
zeolite framework and iron species in the extra-framework po-
sitions, are significantly more active in the SCR process than
the Al-free catalysts containing iron incorporated into the zeo-
lite framework (FeSiBEA). It seems that iron incorporated into
the BEA framework results in too weak acid sites, which are
not able to effectively activate ammonia molecule for the SCR
process. Moreover, the amount of ammonia molecules inter-
acting with the catalyst surfaces at higher temperature is very
limited.
Conclusions
The modification of BEA zeolite with iron resulted in active and
selective catalysts for NO reduction with ammonia (SCR,
deNOx process). Moreover, we have shown that the zeolite cat-
alyst containing iron in framework positions was more resistant
to poisoning by water vapor than BEA zeolites with iron in
extra-framework positions.
For the FeAlBEA and Fe(IE)BEA catalysts, which were found
to be more active in the SCR process comparing to FeSiBEA, at
least two possible parallel reaction mechanisms could be sug-
gested: i) reaction between chemisorbed ammonia and NO
from the gas phase (Eley–Rideal mechanism) and ii) reaction of
surface nitrates and possibly also NO2, formed by oxidation of
NO, with ammonia (probably in the chemisorbed state). It
seems that the second reaction pathway is possible only for
BEA zeolite samples containing iron in the form of extra-frame-
work species, which result in the formation of reactive nitrates
that can be relatively easily reduced by ammonia into the SCR
products. The formation of surface nitrate species by NO oxida-
tion was also observed over the FeSiBEA catalyst, although in
this case their stability was much higher than that of nitrates
formed on FeAlBEA and Fe(IE)BEA, and therefore the reaction
of such species with ammonia was possible only at higher
temperatures. It seems that such high stability and therefore
also low reactivity of nitrate species in FeSiBEA is related to
their interaction with iron ions in the zeolite framework. More-
over, it seems that a very important role, especially in the low-
temperature SCR process, is played by Lewis acid sites, possi-
bly related to the presence of extra-framework iron species.
This type of acidity is predominant in the Fe(IE)BEA catalyst,
which showed excellent activity in the low-temperature range.
Also, only in the case of this catalyst, co-adsorption of NO and
O2 resulted in the formation of FeÀONO species, which were
reported to be an intermediate species in the SCR pro-
cess.[28,29,33,34] The type of iron species present in the zeolites
plays an important role in the process of NO oxidation. It
seems that extra-framework iron species are more catalytically
active in this process than iron incorporated into the zeolite
structure. Moreover, nitrate species formed on the surface of
zeolites with extra-framework iron are less stable and therefore
more reactive than those on catalysts containing iron in frame-
work positions. Thus, taking into account the mechanism
based on the reaction between chemisorbed ammonia and
the surface oxidized NO species, the higher activity of Fe(IE)-
BEA and FeAlBEA in comparison to FeSiBEA could be expected.
Both suggested reaction mechanisms include adsorption and
The temperature-programmed studies and IR experiments
allowed for conclusions given below:
-
The incorporation of iron in the framework of BEA zeo-
lite results in the formation of weaker acid sites, which are not
able to effectively activate ammonia molecules for the SCR
process.
-
The iron species in BEA zeolite are able to catalyze oxi-
dation of NO to nitrates, which in the case of the samples con-
taining iron in extra-framework positions can be easily reduced
by ammonia to the SCR products. In the case of Al-free BEA
zeolite, which contains iron in framework positions, stable ni-
trate species with the limited reactivity in SCR, were formed.
Based on the presented studies, at least two parallel possible
SCR reaction pathways can be suggested: i) reaction between
chemisorbed ammonia molecules and NO from the gas phase
(Eley–Rideal mechanism); ii) reaction of surface nitrates
(possibly also NO2 as FeÀONO), formed by oxidation of NO,
with ammonia (probably in the chemisorbed state).
Experimental Section
Catalyst synthesis
Fe-containing zeolites were prepared by two-step post-synthesis
(PS), ion exchange (IE), and conventional wet impregnation (Imp).
FeSiBEA zeolite was prepared by a two-step post-synthesis method
reported earlier.[17] In the first step, HAlBEA zeolite (2 g, Si/Al=
12.5), obtained by calcination (823 K, 15 h) of tetraethylammonium
BEA (TEABEA) zeolite, provided by RIPP (China), was treated with
13m HNO3 solution under stirring (353 K, 4 h) to remove aluminum
ions from the zeolite structure. In the second step, the resulting
SiBEA (2 g, Si/Al=1000) obtained after filtration was dispersed in a
1.8mm aqueous solution (pH 2.5) of Fe(NO3)3·9H2O (200 mL) and
stirred at room temperature for 24 h. Then, the obtained suspen-
sion was stirred in an evaporator under suction by a water pump
ChemSusChem 2018, 11, 1 – 15
www.chemsuschem.org
11
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
in air at 353 K for 2 h until water was evaporated. The solid with
the iron content of 1.0 wt% is labeled as FeSiBEA (Si/Al=1000).
flow of SO2/He mixture was exchanged for the SO2-free reaction
mixture. Both reaction mixture and SO2/He gas mixture were sup-
plied to the reactor with a total flow rate of 40 mLminÀ1
.
FeAlBEA zeolite was prepared by a conventional wet impregnation
method. HAlBEA (2 g, Si/Al=12.5) was dispersed in a 1.8mm aque-
ous solution (pH 3.0) of Fe(NO3)3·9H2O (200 mL) and stirred at
room temperature for 24 h. Then, the suspension was stirred in
evaporator under suction by a water pump at 353 K for 2 h in air
until water was evaporated. The solid with iron content of 1.0 wt%
is labeled as FeAlBEA (Si/Al=12.5).
Temperature-programmed experiments
Temperature-programmed experiments were carried out in a fixed-
bed continuous-flow microreactor system. The reaction tempera-
ture was measured by a K-type thermocouple located in a quartz
capillary immersed in the catalyst bed. The flow of gas mixture
supplied to the microreactor was monitored by mass flow control-
lers (Brooks Instrument). Gases desorbing from the zeolite samples
were continuously analyzed by a quadrupole mass spectrometer
(RGA 200 Prevac) connected directly to the reactor outlet. Prior to
the TPD and TPRS experiments, the zeolite samples (50 mg) were
outgassed at 673 K for 1 h in a flow of pure helium (20 mLminÀ1).
Calibration of quadrupole mass spectrometer with commercial mix-
tures allowed recalculating the detector signal into NH3 or NO de-
sorption rates.
Fe(IE)BEA zeolite was prepared by an ion-exchange method. Firstly,
BEA zeolite containing organic template (TEA) was calcined (823 K,
15 h) to obtain HAlBEA (Si/Al=12.5). Organic-free BEA was treated
two times with a 0.1m NH4NO3 solution (400 mL) at 343 K for 3 h
to exchange H+ ions for NH4+. Then, the solid was washed with
distilled water and dried overnight at 363 K. The resultant sample
was labeled as NH4AlBEA (Si/Al=12.5). Then, it was dispersed in a
10mM aqueous solution of Fe(NO3)3·9H2O (400 mL, pH 2.5, T=
343 K). After filtration and washing with distilled water the sample
was dried. The obtained sample with iron content of 1.5 wt% was
labeled as Fe(IE)BEA (Si/Al=12.5).
To determine Si/Al molar ratio in all materials used in this work
and iron content in the Fe-loaded zeolites the X-ray fluorescence
chemical analysis was performed at room temperature on a SPEC-
TRO X-LabPro apparatus. Before physicochemical characterization
and catalytic tests, FeSiBEA, FeAlBEA, and Fe(IE)BEA were calcined
at 773 K for 2 h. Details of the catalyst preparation and characteri-
zation, as well as results of the catalytic tests, were reported previ-
ously.[17]
TPD of NH3, NO, and NO+O2
Prior to the NH3-TPD and NO-TPD runs, the sample was saturated
in a flow of gas mixture containing 1 vol.% NH3 or 1 vol.% NO di-
luted in helium (20 mLminÀ1) at 343 K for 1 h. Then, the catalyst
was purged in helium flow to evacuate ammonia molecules physi-
sorbed on the zeolite surface. This process was continued until a
constant baseline level of signals at m/z=16 and m/z=17 (charac-
teristic of NH3) or m/z=30 (characteristic of NO) was attained. De-
sorption was carried out with a linear heating rate of 10 KminÀ1 in
pure helium flow (20 mLminÀ1).
Catalyst testing
In the case of NO+O2-TPD experiments, the sample was saturated
in a flow of gas mixture containing 1 vol.% NO and 2.5 vol.% O2
(20 mLminÀ1) diluted in helium at 673 K for 1 h. Then, the reactor
was cooled down to 343 K also in a flow of the gas mixture con-
taining NO and O2 diluted in He. At 343 K the reactor was purged
in helium flow to evacuate NO and O2 molecules physisorbed on
the zeolite surface. This process was continued until a constant
baseline level of signals m/z=30 and m/z=32 (characteristic of
NO and O2, respectively) was attained. Desorption process was car-
ried out with the linear heating rate of 10 KminÀ1 in helium flow
(20 mLminÀ1).
Catalytic experiments of NO SCR with ammonia were performed in
a fixed-bed flow microreactor system. Prior to the catalytic run, the
catalyst sample (0.1 g) was outgassed in a flow of pure helium at
823 K for 30 min. The reaction mixture containing 0.25 vol.% NO,
0.25 vol.% NH3, and 2.5 vol.% O2 diluted in helium was supplied to
a quartz microreactor with a total flow rate of 40 mLminÀ1 (WHSV
ꢁ24 mLhÀ1 gÀ1), which was monitored and adjusted by mass flow
controllers (Brooks Instrument). The reaction was studied in the
temperature range of 373–723 K. The reactant concentrations were
continuously monitored by
a quadrupole mass spectrometer
(Prevac RGA 200) connected directly to the reactor outlet.
The stability of the studied zeolite catalysts in the presence of
water vapor was studied by sequential exchange from dry to wet
reaction mixture containing 0.25 vol.% NO, 0.25 vol.% NH3, and
2.5 vol.% O2, and 3.0 or 0.0 vol.% water vapor diluted in helium.
The reaction mixture was supplied with a total flow rate of
40 mLminÀ1. The isothermal catalytic test was performed with dry
reaction mixture for the first 25 min. Then water vapor was added
into reaction mixture and testing was continued for the next
25 min. These procedures were repeated once again before a final
change to the dry reaction mixture. All tests were performed with
0.1g of sample at 673 K.
TPD of both NO+O2 and NH3 adsorbed subsequently
The sample was saturated with a gas mixture containing NO and
O2 according to the procedure described in a previous paragraph.
After cooling the reactor to 343 K and purging in a flow of pure
helium, the sample was saturated in a flow of gas mixture contain-
ing 1 vol. % NH3 diluted in helium for 30 min. Then, the catalyst
was again purged in a flow of helium to evacuate ammonia mole-
cules physisorbed on the zeolite surface. Finally, the TPD measure-
ment was performed with a linear heating rate of 10 KminÀ1 in
helium flow (20 mLminÀ1).
A similar procedure was used to determine the effect of SO2 on
the catalytic performance of the studied zeolitic samples. The cata-
lytic test was performed with a reaction mixture containing
0.25 vol.% NO, 0.25 vol.% NH3, and 2.5 vol.% O2 diluted in helium
with 0.1 g of the catalyst sample at 673 K for 60 min. Then the
flow of the reaction mixture was exchange for a flow of gas mix-
ture containing 2000 ppm of SO2 diluted in helium. The sample
was treated with this gas mixture for next 30 min, and then the
NH3-TPSR and NO-TPSR
The zeolite sample was saturated with ammonia and then purged
in helium flow according to the procedures applied in NH3-TPD.
Then, the microreactor was heated up to about 773 K with the
linear temperature increase of 10 KminÀ1 in a flow of gas mixture
&
ChemSusChem 2018, 11, 1 – 15
www.chemsuschem.org
12
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
[3] I. Nova, C. Ciardelli, E. Tronconi, D. Chatterjee, B. Bandl-Konrad, Catal.
Today 2006, 114, 3–12.
containing 1 vol.% NO and 2.5 vol.% of O2 diluted in helium
(20 mLminÀ1). This experiment is denoted as NH3-TPSR.
[4] M. Colombo, I. Nova, E. Tronconi, Catal. Today 2010, 151, 223–230.
[5] H. Sjçvall, L. Olsson, E. Fridell, R. J. Blint, Appl. Catal. B 2006, 64, 180–
188.
[6] R. Delahay, S. Kieger, N. Tanchoux, P. Trens, B. Coq, Appl. Catal. B 2004,
52, 251–257.
[7] L. Olsson, H. Sjovall, R. J. Blint, Appl. Catal. B 2008, 81, 203–217.
[8] P. Metkar, N. Salazar, R. Muncrief, V. Balakotaiah, M. P. Harold, Appl.
Catal. B 2011, 104, 110–126.
NO-TPSR was performed in a similar way. However, in this case the
sample was saturated with a mixture of NO and O2 diluted in
helium and then purged in flow of pure helium. Finally, the sample
was heated up to 773 K with a linear temperature increase of
10 KminÀ1 in a flow of gas mixture containing 1 vol.% NH3 diluted
in helium (20 mLminÀ1).
[9] A. Grossale, I. Nova, E. Tronconi, Catal. Today 2008, 136, 18–27.
[10] M. Devadas, O. Krocher, M. Elsener, A. Wokaun, N. Soger, M. Pfeifer, Y.
Demel, L. Mussmann, Appl. Catal. B 2006, 67, 187–196.
[11] S. Song, G. Wu, W. Dai, N. Guan, L. Li, Catal. Sci. Technol. 2016, 6, 8325–
8335.
[12] E. Yuan, G. Wu, W. Dai, N. Guan, L. Li, Catal. Sci. Technol. 2017, 7, 3036–
3044.
[13] K. Krishna, M. Makkee, Catal. Today 2006, 114, 23–30.
[14] G. Delahay, D. Valade, A. Guzman-Vargas, B. Coq, Appl. Catal. B 2005, 55,
149–155.
Stop-flow TPD experiments
The isothermal catalytic tests were carried out for 1 h at 473 K for
FeAlBEA and Fe(IE)BEA and at 573 K for FeSiBEA, with catalyst mass
of 100 mg using a reaction gas mixture containing 0.25 vol. % NO,
0.25 vol.% NH3 and 2.5 vol.% O2 diluted in helium with a total flow
rate of 40 mLminÀ1. Subsequently, a flow of the reacting mixture
was exchanged for pure helium (20 mLminÀ1). Then, the catalyst
was purged in a helium flow until a constant baseline level of
quadrupole mass spectrometer signals characteristic of NH3 (m/z=
16 and 17), NO (m/z=30), N2 (m/z=28), N2O (m/z=44), and H2O
(m/z=18) was attained. Afterwards, to identify species adsorbed
on the catalyst surface under reaction conditions, the TPD run was
carried out with a linear heating rate of 10 KminÀ1 in helium flow
(20 mLminÀ1) to 823 K.
[15] R. Long, R. Yang, Catal. Lett. 2001, 74, 201–205.
´
[16] P. Boron, L. Chmielarz, J. Gurgul, K. Ła˛tka, T. Shishido, J.-M. Krafft, S.
Dzwigaj, Appl. Catal. B 2013, 138, 434–445.
´
[17] P. Boron, L. Chmielarz, J. Gurgul, K. Ła˛tka, B. Gil, J.-M. Krafft, S. Dzwigaj,
Catal. Today 2014, 235, 210–225.
´
[18] P. Boron, L. Chmielarz, J. Gurgul, K. Ła˛tka, B. Gil, B. Marszałek, S. Dzwigaj,
Microporous Mesoporous Mater. 2015, 203, 73–79.
´
[19] P. Boron, L. Chmielarz, S. Dzwigaj, Appl. Catal. B 2015, 168–169, 377–
384.
[20] M. Ruggeri, A. Grossale, I. Nova, E. Tronconi, H. Jirglova, Z. Sobalik,
Catal. Today 2012, 184, 107–114.
FTIR of samples pre-adsorbed with NH3 or NO
[21] S. Shrestha, M. Harold, K. Kamasamudram, A. Yezerets, Catal. Today
2014, 231, 105–115.
[22] K. Kamasamudram, N. Currier, X. Chen, A. Yezerets, Catal. Today 2010,
151, 212–222.
[23] P. Metkar, M. Harold, V. Balakotaiah, Appl. Catal. B 2012, 111 – 112, 67–
80.
[24] D. Klukowski, P. Balle, B. Geiger, S. Wagloehner, S. Kureti, B. Kimmerle, A.
Baiker, J. Grunwaldt, Appl. Catal. B 2009, 93, 185–193.
[25] R. Long, R. Yang, J. Catal. 2002, 207, 224–231.
[26] A. Grossale, I. Nova, E. Tronconi, D. Chatterjee, M. Weibel, J. Catal. 2008,
256, 312–322.
The zeolite sample in the form of self-supported wafers was acti-
vated under vacuum at 723 K. FTIR spectrum was recorded at
373 K with a Bruker Tensor 27 spectrometer. NH3 or NO was ad-
sorbed on the activated sample and then desorbed under vacuum
for 20 minutes at 373, 473, and 573 K. Additional FTIR experiments,
in which NO and O2 were subsequently adsorbed (NO first and O2
second) on this same sample at 373 K were done. For each desorp-
tion step, the sample was heated to the desired temperature in
the closed cell, then evacuated for 10 minutes and cooled down to
373 K (temperature of spectra recording) in the closed cell.
[27] A. Grossale, I. Nova, E. Tronconi, J. Catal. 2009, 265, 141–147.
[28] Q. Sun, Z. Gao, H. Chen, W. Sachtler, J. Catal. 2001, 201, 89–99.
[29] E. Choi, I. Nam, Y. Kim, J. Catal. 1996, 161, 597–604.
[30] J. Eng, C. Bartholomew, J. Catal. 1997, 171, 14–26.
[31] M. Richter, R. Eckelt, B. Parlitz, R. Fricke, Appl. Catal. B 1998, 15, 129–
146.
[32] Y. Yeom, J. Henao, M. Li, W. Sachtler, E. Weitz, J. Catal. 2005, 231, 181–
193.
[33] R. Long, R. Yang, J. Catal. 2000, 194, 80–90.
[34] L. Lobree, I. Hwang, J. Reimer, A. Bell, J. Catal. 1999, 186, 242–253.
[35] E. Tronconi, L. Lietti, P. Forzatti, S. Malloggi, Chem. Eng. Sci. 1996, 51,
2965–2970.
Acknowledgements
These studies have received funding from the Marian Smolu-
chowski Krakꢀw Research Consortium—a Leading National Re-
search Centre KNOW supported by the Ministry of Science and
Higher Education. Part of the research was carried out with
equipment purchased thanks to the financial support of the Euro-
pean Regional Development Fund in the framework of the Polish
Innovation Economy Operational Program (contract no.
POIG.02.01.00-12-023/08).
[36] L. Lietti, I. Nova, E. Tronconi, P. Forzatti, AIChE J. 1997, 43, 2559–2570.
[37] S. Brandenberger, O. Krçcher, A. Wokaun, A. Tissler, R. Althoff, J. Catal.
2009, 268, 297–306.
[38] A. Ates, A. Reitzmann, G. Waters, Appl. Catal. B 2012, 119, 329–339.
[39] R. Long, R. Yang, J. Catal. 2001, 198, 20–28.
Conflict of interest
´
[40] P. Boron, L. Chmielarz, B. Gil, B. Marszałek, S. Dzwigaj, Appl. Catal. B
2016, 198, 457–470.
The authors declare no conflict of interest.
[41] K. Hadjiivanov, Catal. Rev. 2000, 42, 71–144.
[42] G. Mul, M. W. Zandbergen, F. Kapteijn, J. A. Moulijn, J. Perez-Ramirez,
Catal. Lett. 2004, 93, 113–126.
Keywords: deNOx · heterogeneous catalysis · iron · reaction
mechanisms · zeolites
´
[43] P. Boron, L. Chmielarz, J. Gurgul, K. Ła˛tka, T. Shishido, J.-M. Krafft, S.
Dzwigaj, Appl. Catal. B 2013, 138, 434–445.
[44] S. Dzwigaj, L. Stievano, F. Wagner, M. Che, J. Phys. Chem. Solids 2007,
68, 1885–1891.
[1] P. Forzatti, Appl. Catal. A 2001, 222, 221–236.
[45] K. Hadjiivanov, E. Ivanova, R. Kefirov, J. Janas, A. Plesniar, S. Dzwigaj, M.
Che, Microporous Mesoporous Mater. 2010, 131, 1–12.
[2] C. Ciardelli, I. Nova, E. Tronconi, D. Chatterjee, B. Bandl-Konrad, M.
Weibel, B. Krutzsch, Appl. Catal. B 2007, 70, 80–90.
ChemSusChem 2018, 11, 1 – 15
www.chemsuschem.org
13
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
[46] J. Janas, T. Machej, J. Gurgul, R. P. Socha, M. Che, S. Dzwigaj, Appl. Catal.
B 2007, 75, 239–248.
[53] G. Spoto, A. Zecchina, G. Berlier, S. Bordiga, M. G. Clerici, L. Basini, J.
Mol. Catal. A 2000, 158, 107–114.
[47] J. Perez-Ramirez, J. C. Groen, A. Brꢁckner, M. S. Kumar, U. Bentrup, M. N.
Debbagh, L. A. Villaescusa, J. Catal. 2005, 232, 318–334.
[48] S. Bordiga, R. Buzzoni, F. Geobaldo, C. Lamberti, E. Giamello, A. Zec-
china, G. Leofanti, G. Petrini, G. Tozzola, G. Vlaic, J. Catal. 1996, 158,
486–501.
[54] K. Hadjiivanov, Catal. Rev. Sci. Eng. 2000, 42, 71–144.
[55] M. Rivallan, G. Ricchiardi, S. Bordiga, A. Zecchina, J. Catal. 2009, 264,
104–116.
[56] K. Mandel, M. Straßer, T. Granath, S. Dembski, G. Sextl, Chem. Commun.
2015, 51, 2863–2866.
[49] P. Wu, T. Komatsu, T. Yashima, Microporous Mesoporous Mater. 1998, 20,
139–147.
[50] H. Tippins, Phys. Rev. B 1970, 1, 126–135.
Manuscript received: August 17, 2018
´
[51] P. Boron, L. Chmielarz, J. Gurgul, K. Ła˛tka, S. Dzwigaj, Catal. Today 2014,
235, 210–225.
Revised manuscript received: September 15, 2018
Version of record online: && &&, 0000
[52] B. Gil, A. Adamski, Microporous Mesoporous Mater. 2010, 127, 82–89.
&
ChemSusChem 2018, 11, 1 – 15
www.chemsuschem.org
14
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

FULL PAPERS
´
BEA here now: Iron introduced to BEA
zeolite in various forms gave rise to dif-
ferent catalytic activities in the deNOx
process. The interactions of NH3 and NO
molecules with iron species, as well as
reactivity of adsorbed reactants and
possible reaction mechanisms, were
analyzed by temperature-programmed
methods and FTIR spectroscopy.
P. Boron, M. Rutkowska, B. Gil,
B. Marszałek, L. Chmielarz,* S. Dzwigaj*
&& – &&
Experimental Evidence of the
Mechanism of Selective Catalytic
Reduction of NO with NH3 over Fe-
Containing BEA Zeolites
ChemSusChem 2018, 11, 1 – 15
www.chemsuschem.org
15
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ