Composite catalysts for selective catalytic reduction of NOx and oxidation of residual NH3

Article abstract of DOI:10.1134/S0965544116030087

A comprehensive study has been performed on the catalytic properties of composite catalysts [Mn/support + FeBeta] and their individual components (Mn/support and FeBeta) in the selective catalytic reduction of nitrogen oxides and the ammonia oxidation reaction. It has been shown that mixing the oxide component with the zeolite not only leads to an increase in NO_x conversion, but also improves the selectivity in the oxidation of residual ammonia, thus making it possible to conduct both processes in a single catalytic unit.

Full text of DOI:10.1134/S0965544116030087

ISSN 0965ꢀ5441, Petroleum Chemistry, 2016, Vol. 56, No. 3, pp. 211–216. © Pleiades Publishing, Ltd., 2016.
Original Russian Text © A.I. Mytareva, D.A. Bokarev, G.N. Baeva, D.S. Krivoruchenko, A.Yu. Belyankin, A.Yu. Stakheev, 2016, published in Neftekhimiya, 2016, Vol. 56, No. 3,
pp. 228–233.
Composite Catalysts for Selective Catalytic Reduction
of NOx and Oxidation of Residual NH3
A. I. Mytareva, D. A. Bokarev, G. N. Baeva, D. S. Krivoruchenko,
A. Yu. Belyankin, and A. Yu. Stakheev
Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Moscow, Russia
eꢀmail: aim@ioc.ac.ru
Received November 16, 2015
Abstract—A comprehensive study has been performed on the catalytic properties of composite catalysts
[Mn/support + FeBeta] and their individual components (Mn/support and FeBeta) in the selective catalytic
reduction of nitrogen oxides and the ammonia oxidation reaction. It has been shown that mixing the oxide
component with the zeolite not only leads to an increase in NOx conversion, but also improves the selectivity
in the oxidation of residual ammonia, thus making it possible to conduct both processes in a single catalytic
unit.
Keywords: MnOx, FeBeta, Nx, selective catalytic reduction, ammonia oxidation
DOI: 10.1134/S0965544116030087
Nitrogen oxides (NOx) are highly toxic substances
formed primarily during the fuel combustion process
[1]. Their negative environmental impact, as well as
the introduction of stricter regulations on NOx emisꢀ
sions, necessitates the development of new, highly
effective processes for neutralizing these compounds.
To date, one of the technologies used for the removal
of nitrogen oxides from both industrial (nitric acid
production, thermal power plants, etc.) offꢀgases and
motor vehicle exhaust gases is selective catalytic
reduction (SCR NOx) [2, 3], which is carried out on
vanadium (V–W–TiO2) (I), and Cuꢀ or Feꢀpromoted
zeolite catalysts. Ammonia (or urea solution) is typiꢀ
cally used as a reducing agent (I), which is introduced
with special feeders upstream of an SCR NOx catalyst
system.
In [6] we proposed the idea of a catalytic NOx neuꢀ
tralization system consisting in that an SCR compoꢀ
nent and a component responsible for the oxidation of
residual ammonia are combined in a single unit (by
mechanical mixing). Using the composite catalyst
[FeBeta + Fe (Mn) MCMꢀ48] as an example, we
showed that it is possible to achieve effective removal
of nitrogen oxides and residual ammonia in the single
catalytic unit by optimizing the ratio between the
activities of the SCRꢀmediating component and the
oxidative component.
In this work, we continued the study of the catalytic
properties of [oxidative component + FeBeta] comꢀ
posite systems in the SCR NOx process and the selecꢀ
tive oxidation reaction of residual ammonia. Mangaꢀ
neseꢀcontaining oxide catalysts (Mn/Al2O3 and
Mn/TiO2) having high oxidative activity [7] were used
as an oxidizing agent. It was shown [8, 9] that the
mechanical mixing of the Mn/support component
with the zeolite leads to a substantial increase in NOx
conversion over a wide temperature range, but the
activity of the [Mn/support + zeolite] composite catꢀ
alysts in the ammonia oxidation reaction was not studꢀ
ied yet. The catalyst Mn/FeBeta studied in detail [10],
in which the oxidizing component (MnOx) is supꢀ
ported directly onto the outer surface of zeolite microꢀ
crystals, was used as a reference and for evaluating the
effect of the degree of contact between the oxide and
zeolite components on the catalytic properties of the
composite system.
4NH3 + 4NO + O2
(I)
→
4N2 + 6H2O – standard SCR.
Despite the fact that the SCR NOx process is very
efficient and is widely used in practice, it nevertheless
has some drawbacks. One of these shortcomings is the
possible ammonia “slip” due to an overdose of the
reducing agent needed to compensate for the abrupt
change in the concentration of nitrogen oxides in the
effluent gas. To solve this problem, a ammonia slip
catalyst is mounted downstream of the NOx SCR catꢀ
alyst, a unit that must ensure high selectivity in the
oxidation of ammonia to molecular nitrogen [4, 5].
2NH3 + 3/2O2
→
N2 + 3H2O
(II)
211

212
MYTAREVA et al.
EXPERIMENTAL
Catalyst Preparation
the NH3 oxidation reaction was measured in a flow
unit with a quartz reactor (internal diameter 4 mm).
The reaction mixtures had the following composiꢀ
tions:
Supported manganese catalysts containing 8 wt %
Mn were prepared by incipient wetness impregnation
of support with a solution of manganese nitrate
NH3SCR: 580 ppm NH3, 500 ppm NO,
10 vol % O2, 6 vol % Н2О in nitrogen.
(
Mn(NO3)2
·
x
H2O, Aldrich, 98%). Preliminary calꢀ
NH3 oxidation: 580 ppm NH3, 10 vol % O2,
cined (550
°
С, 4 h) samples of commercial FeBeta (Fe
6
%
Н2О in nitrogen.
content of ~0.9 wt %, Si/Al = 12.5, Zeolist Internaꢀ
Excess of ammonia (NH3/NO > 1) was used to
tional), Al2O3 (SBET = 150 m2/g, Sasol), and TiO2
provide ~80 ppm residual NH3 at the reactor outlet
during the SCR reaction (for ammonia “slip” detecꢀ
tion).
(
SBET = 40 m2/g, SaintꢀGobain) were used as a supꢀ
port. After impregnation, the samples were dried in air
at room temperature for 24 h and then calcined in a
dry air flow (~300 mL/min) at 550°С for 4 h.
The measurements were made in the temperature
range of 100–500°С
at a GHSV of 270000 h–1 with a
The composite catalysts [Mn/Al2O3 + FeBeta] and
Mn/TiO2 + FeBeta] were obtained by thorough
catalyst loading of 0.04 g (fraction 0.2–0.4 mm),
unless otherwise indicated. The feed gas mixture and
the reaction products were analyzed using a Gasmet
FTIR analyzer (Temet Instruments Dxꢀ4000). To
avoid condensation of water vapor, all gas lines were
[
grinding in a mortar of the oxide (Mn/Al2O3 or
Mn/TiO2) and zeolite (FeBeta) components in a
weight ratio of 3 : 1.
heated (180°С).
The conversions of NOx (XNO
)
and NH3 (XNH
)
Instrumental Study of Catalysts
x
3
were calculated as follows:
Examination of the samples by temperatureꢀproꢀ
grammed reduction with hydrogen (H2ꢀTPR) was
performed on a semiautomatic flow unit with a therꢀ
mal conductivity detector. Prior to reduction, the
sample (100 mg) was heated under argon at 325°С for
1 h and cooled to room temperature. The reduction
XNO
x
,
)
(1)
(2)
Cin,NO − (Cout,NO + Cout,NO + 2Cout,N O
2
2
=
Cin,NO
C
− C
out,NH3
was carried out in a flow of 5 vol % H2/Ar gas mixture
in,NH3
,
XNH
=
3
(30 mL/min) with a temperature rise to 820
of 10°C/min. To remove water formed upon reducꢀ
tion, a trap cooled with dry ice in ethanol to –70
°С at a rate
C
in,NH3
where Cin and Cout are the concentrations of the
respective gases at the inlet and outlet of the reactor,
respectively.
°С
was placed between the reactor and the detector. The
detector was calibrated using data on the temperature
programmed reduction of CuO (AldrichꢀChemie
GmbH, 99%) taken in an amount of 0.8 to 11 mg.
Deconvolution of TPR peaks was performed using the
program Ekokhrom.
RESULTS AND ITS DISCUSSION
Scanning Electron Microscopy
The microstructure of the samples was studied by
fieldꢀemission scanning electron microscopy (SEM)
with a Hitachi SU8000 electron microscope. The anaꢀ
lytical measurement conditions were optimized in
accordance with the approach described previously
[11]. Before measurements, the samples were
mounted on a 25 mm aluminum specimen stub, fixed
with a conductive adhesive tape, and then coated with
a conductive metal layer (Au/Pd, 60/40) of a 7 nm
thickness by magnetron sputtering deposition. Images
were acquired in the secondary electron detection
mode at an accelerating voltage of 2 kV and a working
distance of 4 to 5 mm. The morphology of the samples
was studied taking into account possible influence of
metal coating on the surface [12].
The character of distribution of the oxide and zeoꢀ
lite components in the catalytic systems prepared on
their basis was evaluated using scanning electron
microscopy. The micrographs in Fig. 1 clearly show
that a very homogeneous distribution of the compoꢀ
nents can be achieved by mechanical mixing.
The images obtained before (Fig. 1a) and after
(Figs. 1b, 1c) mixing FeBeta with the oxide compoꢀ
nents distinctly display large (~0.4 mm), regular
shaped zeolite microcrystals. The micrographs of the
composite catalysts (Figs. 1b, 1c) show that the zeolite
microcrystals are in contact with smaller particles of
the oxide component (Mn/TiO2 or Mn/Al2O3).
In the case of FeBeta with 8 wt % manganese supꢀ
ported by impregnation (Mn/FeBeta sample), it was
possible to achieve more intimate contact between
MnOx and FeBeta, as well as uniform distribution of
Catalytic Experiments
The activity of the composite catalysts and the the oxide and zeolite components. Comparison of
individual components in both the SCR process and the micrographs of the FeBeta and Mn/FeBeta
PETROLEUM CHEMISTRY Vol. 56
No. 3
2016

COMPOSITE CATALYSTS FOR SELECTIVE CATALYTIC REDUCTION
213
500 nm
500 nm
(а)
(b)
500 nm
500 nm
(c)
(d)
Fig. 1. Scanning electron microscopy micrographs of (a) FeBeta, (b) [Mn/TiO + FeBeta], (c) [Mn/Al O + FeBeta], and
2
2 3
(d) Mn/FeBeta samples.
(Figs. 1a, 1d) shows that MnOx clusters decorate which is comparable with the value of hydrogen uptake
the surface of the FeBeta zeolite microcrystals for the Mn/FeBeta catalyst and suggests the prevaꢀ
(Figs. 1a, 1d).
lence of MnO2.
TemperatureꢀProgrammed Reduction
In the H2ꢀTPR spectra of the Mn/FeBeta,
Mn/TiO2, and Mn/Al2O3 catalysts (Fig. 2), two proꢀ
nounced peaks of hydrogen absorption at low (330–
5 µmol/g
360°С) and high temperatures (425–525°С) are disꢀ
5
4
3
tinguishable. In the literature, the presence of these
two peaks is associated with the consecutive reduction
of manganese oxides, wherein the first step is the
reduction of MnO2 to Mn2O3 and the second step is
the reduction of Mn2O3 to MnO [13, 14]. The total
hydrogen uptake is 0.93, 0.99, or 0.68 mmol/gcat for
Mn/FeBeta, Mn/TiO2, or Mn/Al2O3, respectively.
These values indicate that most of the manganese in
the Mn/FeBeta and Mn/TiO2 samples occurs as
MnO2, whereas a significant portion of the manganese
in the Mn/Al2O3 sample is in the form of Mn2O3
2
1
and/or nonstoichiometric MnOx
.
100
200
300
400
500
600
, °C
In the case of composite catalysts [Mn/Al2O3
FeBeta] and [Mn/TiO2 + FeBeta], in addition to the
above two peaks, the H2ꢀTPR spectra exhibit a lowꢀ
+
T
temperature signal at 200°С (Fig. 2), which can be
explained by reduction of finely divided MnO2 to
Mn2O3 [15]. At the same time, the total absorption of
Fig. 2. Profiles of temperatureꢀprogrammed reduction
with hydrogen for Mn/FeBeta, Mn/TiO ,
[Mn/TiO
(
1)
(2)
2
hydrogen turns to be 0.92 mmol/gcat for [Mn/Al2O3
+
(
(
3
) Mn/Al O ,
(
4)
+
FeBeta], and
2
3
2
FeBeta] and 0.96 mmol/gcat for [Mn/TiO2 + FeBeta],
5
) [Mn/Al O + FeBeta].
2
3
PETROLEUM CHEMISTRY Vol. 56
No. 3
2016

214
MYTAREVA et al.
mechanism, close contact between the oxide and zeoꢀ
X
NO , %
x
lite components facilitates rapid transfer of reaction
intermediates from one component to another,
thereby enhancing the reaction. During the twoꢀstep
process, NO2 produced on the oxide component (IV),
diffuses to the zeolite surface, on which it is reduced
along with NO by ammonia via fast SCR reaction (V).
100
1
2
3
4
5
6
80
60
40
20
0
2NO + O2
2NO2 on oxide component
(IV)
NO + NO2 + 2NH3
(V)
→
2N2 + 3H2O on zeolite component.
The decrease in the NOx conversion (71% on
Mn/FeBeta and 36% on the composite catalysts at
500°С) observed at Т > 350°С can be due to side reacꢀ
tion of ammonia oxidation (III), as evidenced by the
absence of ammonia in the gas stream at the reactor
outlet (residual NH3 concentration did not exceed
10 ppm).
100
200
300
400
500
, °C
T
Fig. 3. Temperature dependence of NOx conversion on
) Mn/Al O , ( ) Mn/TiO , ( ) FeBeta, ( ) [Mn/Al O +
) [Mn/TiO + FeBeta], and (
(
1
2
3
4
2
5
3
2
2 3
FeBeta], (
6) Mn/FeBeta
2
catalyst samples.
Activity of the Catalysts in the NH3 Oxidation Reaction
Testing the samples in the ammonia oxidation
reaction revealed that the individual components
(Mn/support) and the composite catalysts based on
them ([Mn/support + FeBeta]) almost do not differ in
Catalytic Measurements
Catalytic properties of the individual components
Mn/Al2O3 and Mn/TiO2) and the composite catalysts
on their basis were tested in both the selective catalytic
reduction of NOx and the ammonia oxidation reacꢀ
tion.
(
activity (Fig. 4a). The oxidation reaction begins at
Т
≈
300 and the NH3 conversion at > 400 is 94–
°С
Т
°C
99% on all of the Mnꢀcontaining samples, including
the reference sample Mn/FeBeta. The only exception
is FeBeta, on which the ammonia conversion does not
exceed 12%. These data lead to the conclusion that the
activity of the catalysts in the oxidation of NH3 is priꢀ
marily determined by the Mn component and the
contribution of the zeolite component is insignificant.
However, the introduction of the zeolite compoꢀ
nent results in a drastic change in selectivity of the
process. The main products of the reaction on the
individual oxide components are NO, NO2, and N2O
(Fig. 4b), whereas mixing the Mn component with
FeBeta zeolite leads to substantial inhibition of the
formation of undesirable nitrogen oxides at temperaꢀ
Activity of Catalysts in NOx SCR
Samples of the individual components Mn/Al2O3
and Mn/TiO2 have a low NOx SCR activity (Fig. 3).
The NOx conversion in both cases increases with the
increasing temperature; however, after reaching the
maximum (27% for Mn/TiO2 or 51% for Mn/Al2O3),
it sharply decreases because of the reaction of comꢀ
plete oxidation of ammonia and the formation of
additional amounts of NOx
4NH3 + 5O2 4NO + 6H2O
The activity of the zeolite component is relatively
low (<75% at 500 ), since the mass of FeBeta used in
.
→
(III)
tures above 350
trations of NOx on the composite catalysts
Mn/Al2O3 + FeBeta] and [Mn/TiO2 + FeBeta] at
500 are about ~50 and ~130 ppm (versus ~320 ppm
°С. For example, the maximal concenꢀ
°С
the individual test was the same as in the
composite catalysts (space velocity, ~1075000 instead
of 270000 h–1).
The mechanical mixing of the components
Mn/support and FeBeta leads to a sharp increase in
the catalytic activity (Fig. 3) as compared to the indiꢀ
vidual components. Thus, the composite catalysts
[
°С
on Mn/Al2O3 and ~410 ppm on Mn/TiO2), respecꢀ
tively.
The increase in the ammonia oxidation selectivity
is apparently due to the involvement of NO, produced
on the oxide component (III), in standard SCR reacꢀ
tion (I) on the zeolite [8]. The general scheme of the
process can be described by the following combination
of the reactions discussed above:
[
Mn/Al2O3 + FeBeta] and [Mn/TiO2 + FeBeta] at low
temperatures (
225°C for Mn/FeBeta, ~255
FeBeta], or ~240
T
< 350°C) are almost as good (T50
=
+
°C
for [Mn/Al2O3
°C
for [Mn/TiO2 + FeBeta]) as the
reference sample Mn/FeBeta in catalytic properties.
The observed synergistic effect can be explained by the
occurrence of the SCR reaction via the bifunctional
mechanism proposed in [8, 9]. According to this
4NH3 + 5O2
(III)
→
4NO + 6H2O on Mn component
2NO + O2
2NO2 on Mn component
(IV)
PETROLEUM CHEMISTRY Vol. 56
No. 3
2016

COMPOSITE CATALYSTS FOR SELECTIVE CATALYTIC REDUCTION
215
X
NH , %
CNO , ppm
(a)
(b)
3
x
100
80
60
40
20
0
500
400
300
200
100
1
2
3
4
5
6
1
2
3
4
5
6
0
100
100
200
300
400
500
, °C
200
300
400
500
, °C
T
T
Fig. 4. (a) Temperature dependence of NH conversion on (
1
) Mn/Al O , (
2
) Mn/TiO , (
3
) FeBeta, (
) Mn/FeBeta samples; (b) formation of NOx during the ammonia oxidation reaction on
) FeBeta, ( ) [Mn/Al O + FeBeta], ( ) [Mn/TiO + FeBeta], and ( ) Mn/FeBeta samples.
4) [Mn/Al O + FeBeta],
3
2
3
2
2 3
(5
) [Mn/TiO + FeBeta], and (6
2
3
(1
) Mn/Al O , (
2
) Mn/TiO , (
3
4
5
6
2
2
2
3
2
decline in the NOx conversion on [Mn/Al2O3
+
NO + NO2 + 2NH3
2N2 + 3H2O on zeolite component
FeBeta] and [Mn/TiO2 + FeBeta] is due to the beginꢀ
ning of domination of the ammonia oxidation reaction
over the composite catalysts at temperatures above
(V)
→
.
It should be noted that the lowest NOx concentraꢀ
tion was obtained on the Mn/FeBeta sample
(<20 ppm), thereby indicating that the prevalence of
the zeolite component in the catalyst increases the
probability of reaction (V).
350°С to form nitrogen oxides, most of which are not
reduced to nitrogen via the standard SCR reaction
because of a small zeolite content (three times below
that in the Mn/FeBeta sample).
The data obtained by studying the oxidation of
NH3 also allow the following explanation of the higher
NOx conversion observed on the Mn/FeBeta catalyst
CONCLUSIONS
It has been shown that composite catalysts consistꢀ
ing of an oxidative component (MnOx) and a zeolite
(FeBeta) have high activity in both the selective cataꢀ
lytic reduction of NOx and the selective oxidation of
ammonia. The feasibility to integrate the two catalytic
functions in one unit will make it possible to reduce
not only costs, but also the size of the nitrogen oxides
neutralization system as a whole, which is especially
important for addressing environmental challenges of
the automotive industry.
at high temperatures (330–500°C, Fig. 3). Ammonia
participates simultaneously in two reactions on the
composite systems [oxide component + FeBeta] or
the Mn/FeBeta catalyst. The selectivity of the ammoꢀ
nia oxidation process in this case will depend on the
ratio of the rates of reactions (III) and (V), which in
turn are directly related to the ratio between the oxide
and zeolite components in the composite catalyst.
Thus, the higher the proportion of the oxide compoꢀ
nent in the catalyst ([Mn/Al2O3 + FeBeta], [Mn/TiO2
+
FeBeta]), the greater the amount of ammonia oxiꢀ
dized via reaction (III) and the greater will be the conꢀ
centration of NOx in the products. An increase in the
zeolite content (Mn/FeBeta) will lead to involvement
of NOx in fast SCR NOx reaction (V).
In addition, it should be taken into account that the
standard catalytic reduction reaction (I) on the
FeBeta component, whose proportion is much higher
in Mn/FeBeta, plays a significant role in the overall
ACKNOWLEDGMENTS
This work was supported by the Russian Foundaꢀ
tion for Basic Research, project no. 15ꢀ03ꢀ07802 A.
The authors thank the Department of Structural Studꢀ
ies of Zelinsky Institute of Organic Chemistry for elecꢀ
tron microscopy study of the samples.
SCR process in the temperature range of 300–500°C.
REFERENCES
Thus, the comparison of the experimental data on
the catalytic activity of the catalysts in the SCR NOx
process and the NH3 oxidation reaction shows that the
1. WHO Air Quality Guidelines for Particulate Matter,
Ozone, Nitrogen Dioxide and Sulfur Dioxide: Global
PETROLEUM CHEMISTRY Vol. 56
No. 3
2016

216
MYTAREVA et al.
Update 2005, Geneva: World Health Organization,
2006).
9. A. Yu. Stakheev, A. I. Mytareva, D. A. Bokarev, et al.,
Catal. Today, No. 258 (Part I), 183 (2015).
10. D. S. Krivoruchenko, A. V. Kucherov, N. S. Telegina,
et al., Izv. Akad. Nauk, Ser. Khim., No. 2, 389 (2014).
11. V. V. Kachala, L. L. Khemchyan, A. S. Kashin, et al.,
Usp. Khim. 82, 648 (2013).
12. A. S. Kashin and V. P. Ananikov, Izv. Akad. Nauk, Ser.
Khim., No. 12, 2551 (2011).
2. A. Alessandro and R. D. Ana, Energy 34, 348 (2009).
3. T. Johnson, Platinum Met. Rev. 52, 23 (2008).
4. M. Jab
lon' ska and R. Palkovits, Appl. Catal., B 181,
332 (2016).
5. S. M. Lee and S. C. Hong, Appl. Catal., B 163, 30,
(2015).
13. X. Lou, P. Liu, J. Li, and K. He, Appl. Surf. Sci. 307
,
6. A. Yu. Stakheev, D. A. Bokarev, A. I. Mytareva, et al.,
Mendeleev Commun., No. 24, 313 (2014).
382 (2014).
14. J. Liu, C. Y. Yu, P. Q. Zhao, and G. X. Chen, Appl.
Surf. Sci. 258, 9096 (2012).
7. Z. Wu, N. Tang, L. Xiao, Y. Liu, H. Wang, J. Colloid
Interface Sci. 352, 143 (2010).
15. A. Zhang, Z. Zhang, J. Chen, et al., Fuel Process.
Technol. 135, 25 (2015).
8. A. Yu. Stakheev, G. N. Baeva, G. O. Bragina, et al.,
Translated by S. Zatonsky
Top. Catal. 56, 427 (2013).
PETROLEUM CHEMISTRY Vol. 56
No. 3
2016