Applied Catalysis B: Environmental 134–135 (2013) 55–59

Contents lists available at SciVerse ScienceDirect

Applied Catalysis B: Environmental

journal homepage: www.elsevier.com/locate/apcatb

Kinetics of the NO/NO2 equilibrium reaction over an iron zeolite catalyst

V. Bacher a , C. Perbandt b , M. Schwefer b , R. Siefert b , T. Turek a,∗
a
Institute of Chemical Process Engineering, Clausthal University of Technology, 38678 Clausthal-Zellerfeld, Germany
b
ThyssenKrupp Uhde GmbH, 44141 Dortmund, Germany

a r t i c l e i n f o a b s t r a c t

Article history: The steady-state kinetics of NO oxidation and its reverse reaction were studied by systematic measure-
Received 30 July 2012 ments in an integral reactor at a range of temperatures, reactor pressures and reactant concentrations on
Received in revised form a commercial iron zeolite catalyst. A Langmuir–Hinshelwood type rate equation was found to describe
21 December 2012
the experimental results with very good accuracy. The reaction orders for NO and O2 were found to be
Accepted 29 December 2012
1 and 0.75, respectively, while the reaction was inhibited by both NO2 and H2 O. For the first time, the
Available online 4 January 2013
kinetics developed allow the influence of water vapor on the rate of the NO/NO2 equilibrium reaction
both in the concentration range above and below 1 vol% H2 O to be described.
Keywords:
NO oxidation © 2013 Elsevier B.V. All rights reserved.
Iron zeolite
Kinetics

1. Introduction The kinetics of NO oxidation over iron-containing zeolites have

been studied in a few investigations only. Giles et al. [10] conducted
Iron exchanged zeolites have received much attention in recent NO oxidation on a Fe-MFI catalyst prepared by solid-state ion
years as catalysts for the removal of nitrogen oxides. A special fea- exchange. High NO oxidation conversions were achieved at tem-
ture of this catalyst system is its activity for both the reduction of peratures around 300 ◦ C in the absence of H2 O and SO2 , while both
NOx (NO and NO2 ) with ammonia (SCR) [1] and the decomposition species caused a quite strong inhibition of the reaction rate. Čapek
and reduction of N2 O [2]. There are first commercial applications in et al. [11] studied NO oxidation using copper and iron exchanged
automotive pollution control through SCR and for the simultaneous MFI catalysts in the temperature range 200–350 ◦ C. In these exper-
abatement of nitrous oxide and NOx in the EnviNOx® process [3]. iments, NO oxidation was studied in the absence of water vapor,
The EnviNOx® process is a catalytic tail-end exhaust gas system for where a strong inhibition of the reaction rate by the adsorption
nitric acid plants that combines SCR of NOx with either decompo- of NO2 was observed over both catalyst types. Schuler et al. [12]
sition or reduction of N2 O [4]. Depending on the specific nitric acid studied the ammonia SCR on iron zeolite catalysts. In some of their
technology, removal of the nitrogen oxides takes place at pressures experiments, the NO oxidation was investigated at different NO
between atmospheric and about 10 bar. inlet concentrations in the presence of 5 vol% H2 O at temperatures
The rate of the SCR reaction over iron zeolites is strongly depend- between 150 and 450 ◦ C. The results were kinetically modeled using
ent on the relative amounts of NO and NO2 with the fastest reaction the equation developed by Hauptmann et al. [13] for NO oxidation
occurring in the presence of equimolar mixtures [5]. The decom- over Pt/Al2 O3 . Sjövall et al. [14] conducted NO oxidation experi-
position of N2 O, on the other hand, is significantly enhanced in the ments at 150–650 ◦ C in the presence of water vapor (5 vol%) using a
presence of even traces of NO [6,7]. Moreover, the oxidation of NO commercial washcoated iron zeolite monolith. Markatou et al. [15]
in diesel engine exhaust gases is highly desirable, as the produced studied the NO oxidation over a commercial iron zeolite SCR cata-
NO2 is highly reactive toward soot trapped in diesel particulate fil- lyst under transient and steady-state conditions with gas streams
ters [8,9]. Hence, it is obvious that the oxidation of NO (1) is an containing 4.5 vol% CO2 and 4.5 vol% O2 in the absence of water
important step for a variety of technically important reactions. vapor. Chatterjee et al. [16] did not investigate the NO oxidation
individually, however, they have fitted the kinetics of this reac-
2 NO+O2  2 NO2 (1) tion to SCR experiments on an iron zeolite catalyst in the presence
of ammonia. Brüggemann and Keil [17] studied the NO oxidation
theoretically over Fe-MFI catalysts. Their results strongly support
∗ Corresponding author. Tel.: +49 5323 722184; fax: +49 5323 722182. the experimental findings of Giles et al. [10] and Devadas et al. [5]
E-mail addresses: [email protected] (V. Bacher), that water vapor has an inhibiting effect on the rate of NO oxida-
[email protected] (C. Perbandt),
[email protected] (M. Schwefer),
tion. Recently, Metkar et al. [18] have investigated NO oxidation
[email protected] (R. Siefert), [email protected] (T. Turek). over iron and copper zeolite catalysts. The influence of water was

0926-3373/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcatb.2012.12.042
56 V. Bacher et al. / Applied Catalysis B: Environmental 134–135 (2013) 55–59

studied in the concentration range typical for automotive applica- cNO2 ,eq
Keq = 0.50
(3)
tions (2–5 vol%). Reaction rates were determined by measurements cNO2 ,eq · cO ,eq 2
in a differentially operated tubular reactor and a kinetic model
based on elementary steps was developed that included the inhi- This equation was developed for temperatures between 100 and
bition of reaction rates by nitrogen dioxide and water vapor. 370 ◦ C, NO feed concentrations from 75 to 450 ppm and a constant
A deficiency of the existing experimental work on NO oxidation O2 concentration of 6 vol%.
is the limitation to water vapor concentrations above 1 vol%. Con- Chatterjee et al. [16] used a similar kinetics with reaction orders
sequently, no kinetic model describing the effect of water vapor of 1 for NO and 0.5 for oxygen Eq. (4 and 5). The parameters are
at low concentrations is available. This is justified by the fact that valid for temperatures from 150 to 450 ◦ C and feed concentrations
the reactions relevant for ammonia SCR under automotive exhaust of 500–1000 ppm NO, 2–10 vol% O2 , and 1–10 vol% H2 O.
conditions, where high H2 O concentrations are always present, are  −E   cNO2

0.5
hardly influenced by the water concentration. However, the cata- r = k0 · exp · cNO · yO − (4)
R·T 2 Keq
lysts employed in the EnviNOx® process [3,4] for the removal of
 
nitrogen oxides from nitric acid exhaust gases usually encounter Eeq (T )
much lower water concentrations. This is because the catalyst is Keq = exp (5)
R·T
located after the adsorption tower, which is operated close to ambi-
ent temperature with corresponding low equilibrium water vapor In contrast, Čapek et al. [11] observed an inhibition of the reac-
concentrations. Therefore, we have investigated NO oxidation over tion rates by adsorption of NO2 (6).
a commercial iron zeolite catalyst in a broad range of temperatures,  
 −E y2 ·yO −y2
pressures and reactant compositions taking into account varia- k0 · exp · Pt3
NO 2 NO2
R·T 2
Pt ·Keq
tions of NO, NO2 , O2 , and H2 O concentrations. The measured data
r= (6)
were described with a global Langmuir–Hinshelwood type kinetic (1 + KNO2 · Pt · yNO2 )3
model.
Here, yi is the mole fraction of species i and Pt is the total
pressure. Typical reaction mixtures consisted of 1000 ppm NOx
2. Experimental
(1000 ppm NO or 800 ppm NO and 200 ppm NO2 ) with 6 vol% O2 .
Recently, Metkar et al. [18] have developed an elementary step
A commercial iron zeolite catalyst (Clariant Süd-Chemie AG,
based model for NO oxidation over iron and copper exchanged zeo-
EnviCat® -NOx) was used for the experiments [19]. The catalytic
lite catalysts taking into account the adsorption of oxygen, nitrogen
material employed had been exposed to typical reaction temper-
dioxide and water vapor. In preliminary investigations, reaction
atures and reactant concentrations in a nitric acid plant exhaust
orders of 0.9–1 for NO, 0.55–0.59 for O2 , and 0.42–0.49 for NO2
gas stream for 3 months and exhibited stable catalytic proper-
were determined. For the case of excess water (at least 2 vol%),
ties. The as-received catalyst bodies were crushed and sieved to a
the following simplified rate equation (7) was developed, which
315–500 ␮m fraction which was packed in an 18 mm stainless steel
has a reaction order of −1 with respect to the normalized H2 O
tube mounted vertically in a cylindrical oven. The packing height
concentration:
for the 6.0 g of catalyst employed amounted to approximately  −E 
3.7 cm. k0 · exp R·T
· XNO · XO0.5 − XNO2 /Keq
The reactant mixtures (2000 cm3 /min (NTP), down flow) were r= 2
(7)
XH2 O
obtained by feeding NO, NO2 , O2 and N2 via electronic mass flow
controllers (Bronkhorst High-Tech) from pressurized gas cylinders, During these experiments, different concentrations of NOx
while water was dosed by means of a liquid mass flow controller (500 ppm, NO2 /NOx = 0.25, 0.5 and 0.75), O2 (5 vol%) and water
(LIQUI-FLOW, Bronkhorst High-Tech) through a tube into an evap- (2 and 5 vol%) were employed over a range of temperatures
orator system (CEM, Bronkhorst High-Tech). The absolute reactor (150–550 ◦ C). All cited studies relate to automotive catalysis and
pressure could be maintained with a regulating valve (HI-TEC, were therefore carried out at atmospheric pressure.
Bronkhorst High-Tech) in the range between 1 and 15 bar. The inlet
and outlet reactor streams were analyzed using a FTIR spectrome- 3.2. New model
ter (Nicolet 5700, Thermo Scientific), equipped with a 2 m gas cell,
as well as a thermomagnetic oxygen analyzer (Oxymat 6, Siemens). As some of our experimental results, especially at water con-
The NO oxidation was studied at temperatures from 200 to centrations below 1 vol% and at higher reactor pressures, could not
500 ◦ C. The feed contained 200–1000 ppm NO, 0–200 ppm NO2 , be described with one of the previously published models, a new
0.2–3.7 vol% O2 and 0.11–1.07 vol% H2 O balanced in N2 . The total kinetic equation had to be developed. This rate equation is of the
pressure was varied between 1.55 and 6.63 bar. Steady-state con- Langmuir–Hinshelwood type taking into account the inhibition by
ditions were observed after about 15 min, while measurements at water vapor and nitrogen dioxide (8)
 
increasing and decreasing temperatures yielded identical results.  −E n −
cNO ·c n−0.5
2 O2
k0 · exp R·T
· cNO · cO 0.5
2 Keq
3. Kinetic model r=  2
(8)
m
1 + KNO2 · cNO2 + KH2 O · cH
2O
3.1. Literature models
with the equilibrium constant given by:
The NO oxidation kinetics of Hauptmann et al. [13] developed for 2
cNO
precious metal catalysts is a power-law equation (2) with reaction Keq = 2 ·c
2
(9)
cNO O2
orders of 0.28 for NO and 0.49 for O2 , taking the reverse reaction
into account via the equilibrium constant (3). The concentration-based equilibrium constant Keq is obtained
  from the thermodynamic equilibrium constant
 −E 
r = k0 · exp 0.28
· cNO 0.49
· cO · 1−
cNO2
(2)
 − G(T, p) 
R
R·T 2 0.50
cNO · cO · Keq K 0 = exp (10)
2 R·T
 V. Bacher et al. / Applied Catalysis B: Environmental 134–135 (2013) 55–59 57

(a) (b)

Fig. 1. Measured and calculated NO (left) and NO2 (right) outlet concentrations as a function of temperature at different NO inlet concentrations (reactor pressure 1.55 bar,
2.5 vol% O2 and 3200 ppm H2 O at reactor inlet), solid lines: equilibrium, broken lines: simulations.

with the following relationship only. According to classical theory, the effectiveness factor is a
function of the Thiele modulus  (12).

Keq = K 0 · (R · T )− vi
(11) =
tanh()
(12)

The Thiele modulus can be expressed with the following equa-
where the change in the number of moles vi amounts to −1
tion
according to the stoichiometric equation (1). The thermodynamic
data for the calculation of the equilibrium constant (10) were taken kV
from the NIST databank [20]. =L· (13)
Deff
The laboratory reactor could be regarded as isothermal due to
the very small reaction enthalpy. At the given length of the cat- where L is the characteristic diffusion length (assuming one
alytic packing, the Bodenstein number was larger than 100, thus sixth of the equivalent average particle diameter) and kv the vol-
plug flow behavior could be assumed. The kinetic parameters k, ume specific pseudo first-order rate constant. The intrinsic rate
E, KNO2 , KH2 O as well as the reaction orders n and m were deter- constant kv was calculated with the apparent mass specific rate
mined by fitting measured and calculated outlet concentrations constant km,app which was obtained from the measured NO con-
in the integral reactor. The simulations were conducted using the version XNO in the plug-flow reactor, and the apparent density of
software package Presto-Kinetics. Ordinary differential equations the catalyst particles p (1110 kg/m3 ):
were solved applying the MEC (2) solver operating at an accuracy km,app · p
of 10−6 . Simulated annealing and box searches were used to obtain kV = (14)

starting values for parameter estimations. Optimized values were
determined applying a damped Gauss–Newton algorithm [21]. The effective diffusivity was based on a particle porosity of 0.38
and a tortuosity of 2.9 taking into account both Knudsen and binary
diffusion coefficients of NO [22]. Thiele modulus and effectiveness
factor are then finally calculated iteratively with Eqs. (12)–(14).
4. Results and discussion For calculation of the overall effectiveness factor, the mass trans-
fer coefficient, which can be obtained from correlations for the
The parameter fit yielded the kinetic parameters summarized Sherwood number, has to be taken into account additionally. The
in Table 1. The activation energy is in good agreement with Chat- following equation for ov has been used:
terjee et al. [16], while Čapek et al. [11] observed a higher value of
1
42 kJ/mol. The inhibition constant for NO2 is of the same order of ov = (15)
kv · L/ˇ + 1/
magnitude as the value of 80 m3 /mol (recalculated with the ideal
gas law) given by Čapek et al. [11]. The inhibition of water is not The results for the highest NO conversion rate during our mea-
as strong as in the kinetics developed by Metkar et al. [18] with a surements are summarized in Table 2.
reaction order of −1. It can be clearly seen that the calculated catalyst effectiveness
The possible influence of mass transfer phenomena on the factors are very close to 100%. This means that the influence of both,
observed reaction rates was estimated by calculation of catalyst internal and external mass transfer phenomena on the measured
effectiveness factors using a simplified first-order approach. The reaction rates can be neglected, which is in agreement with Metkar
overall effectiveness factor ov , which is the ratio of observed rate et al. [18].
and reaction rate for bulk concentrations, includes both internal The results in Fig. 1 show that the equilibrium concentrations
and external diffusion limitations. On the other hand, the effec- of NO and NO2 are reached at temperatures around 400 ◦ C, with a
tiveness factor  comparing observed rate and reaction rate for tendency to higher temperatures at higher nitric oxide inlet con-
catalyst surface concentrations, takes into account pore diffusion centrations. Experimental and calculated values for both NO and
58 V. Bacher et al. / Applied Catalysis B: Environmental 134–135 (2013) 55–59

Table 1
Kinetic parameters of Eq. (8).

k0 E (kJ/mol) KNO2 (m3 /mol) Water inhibition nO2

0.5
10.35 m 5.25 0.75
/kg/s/mol 31.0 27.9 KH2 O = 1.98 m1.5 /mol 0.75
mH2 O = 0.5

Table 2
Calculation of the catalyst effectiveness factors.

T (◦ C) XNO km,app (m3 /kg/s) Deff (m2 /s) L (m)   ˇ (m/s) ov

270 0.86 0.0146 5.00 × 10−6 6.75 × 10−5 0.148 0.993 0.57 0.991

NO2 are in very good agreement. In all measurements, the sum

of both nitrogen oxides was equivalent to the inlet concentrations
with a deviation of less than 3%. For that reason, only the NO con-
centrations are depicted in the following diagrams. Fig. 2 reveals
that the measured concentrations using a mixture of NO and NO2
at reactor inlet can also be described with good accuracy.
The effect of oxygen concentration on equilibrium and reaction
rate of NO oxidation is shown in Fig. 3. It can be seen that the equi-
librium is shifted in the expected way and that the reaction rate
is accelerated at rising O2 concentrations. In contrast to former
research, a reaction order of 0.75 with respect to oxygen was found
to be most suitable for description of the measured results.
The influence of the H2 O concentration on the rate of NO
oxidation is depicted in Fig. 4. During these measurements, the
equilibrium conditions remain constant. It is evident that the kinet-
ics are strongly inhibited by water vapor in the concentration range
studied. With the new kinetic equation, reaction rates and the
resulting reactor outlet concentrations can be described over a
range of water vapor concentrations between 0.11 and 1.07 vol%. It
must be noted that the rate equation does not hold in the complete
absence of water, where the measured NO outlet concentrations
are significantly lower than the calculated values (Fig. 4, lower bro-
ken line). Attempts to describe the water inhibition with a single Fig. 3. Effect of oxygen concentration on measured and calculated NO outlet con-
reaction order over the whole range of concentrations were not centrations as a function of temperature (reactor pressure 1.55 bar, 400 ppm NO and
successful. Obviously, the inhibition of reaction rates is very strong 3200 ppm H2 O at reactor inlet), solid lines: equilibrium, broken lines: simulations.

Fig. 2. Measured and calculated NO outlet concentrations as a function of tempera- Fig. 4. Measured and calculated NO outlet concentrations as a function of temper-
ture for 400 ppm NO and a mixture of 200 ppm NO and 200 ppm NO2 at reactor inlet ature at different H2 O concentrations (reactor pressure 1.55 bar, 2.5 vol% O2 and
(reactor pressure 1.55 bar, 2.5 vol% O2 and 3200 ppm H2 O at reactor inlet), solid line: 400 ppm NO at reactor inlet), solid line: equilibrium, broken lines: simulations.
equilibrium, broken lines: simulations.
 V. Bacher et al. / Applied Catalysis B: Environmental 134–135 (2013) 55–59 59

Overall, the present kinetics allow the rate of NO oxidation and

the corresponding reverse reaction over a commercial iron zeolite
catalyst over a broad range of concentrations to be described with
unprecedented accuracy. The parity plot depicted in Fig. 6 reveals
that the average standard deviation amounts to only ±2.6%.

5. Conclusions

Systematic measurements of NO oxidation with O2 and the cor-

responding reverse reaction were carried out in an integral reactor
operated at steady state using a commercial iron zeolite catalyst at
different temperatures, concentrations and reactor pressures. The
kinetic model developed yielded a very good fit of the experimental
data over a broad range of reaction condition. The reaction orders
with respect to NO and O2 were found to be 1 and 0.75, respec-
tively. The reaction rate is inhibited by both NO2 and H2 O. For the
first time, the influence of water vapor on the rate of NO oxidation
could be successfully described over a broad concentration range
between about 0.1 and 2.1 vol% H2 O.

Acknowledgments

Fig. 5. Effect of total pressure on measured and calculated NO outlet concentrations The authors would like to thank Clariant Süd-Chemie AG for
as a function of temperature (2.5 vol% O2 , 3200 ppm H2 O and 400 ppm NO at reactor kindly providing the catalyst samples.
inlet), solid lines: equilibrium, broken lines: simulations.
References

[1] S. Brandenberger, O. Kröcher, A. Tissler, R. Althoff, Catalysis Reviews-Science

and Engineering 50 (2008) 492–531.
[2] A. Guzmán-Vargas, G. Delahay, B. Coq, Applied Catalysis B: Environmental 42
(2003) 369–379.
[3] M. Groves, M. Schwefer, R. Siefert, TCE (2006) 30–31.
[4] M.C.E. Groves, A. Sasonov, Journal of Integrative Environmental Sciences 7
(2010) 211–222.
[5] M. Devadas, O. Kröcher, M. Elsener, A. Wokaun, N. Söger, M. Pfeifer, Y. Demel,
L. Mussmann, Applied Catalysis B: Environmental 67 (2006) 187–196.
[6] M. Kögel, B.M. Abu-Zied, M. Schwefer, T. Turek, Catalysis Communication 2
(2001) 273–276.
[7] J. Pérez-Ramírez, F. Kapteijn, G. Mul, J.A. Moulijn, Journal of Catalysis 208 (2002)
211–223.
[8] A. Setiabudi, M. Makkee, J.A. Moulijn, Applied Catalysis B: Environmental 50
(2004) 185–194.
[9] B.R. Stanmore, V. Tschamber, J.F. Brilhac, Fuel 87 (2008) 131–146.
[10] R. Giles, N.W. Cant, M. Kögel, T. Turek, D.L. Trimm, Applied Catalysis B: Envi-
ronmental 25 (2000) L75–L81.
[11] L. Čapek, L. Vradman, P. Sazama, M. Herskovitz, B. Wichterlová, R. Zuker-
man, R. Brosius, J.A. Martens, Applied Catalysis B: Environmental 70 (2007)
53–57.
[12] A. Schuler, M. Votsmeier, P. Kiwic, J. Gieshoff, W. Hauptmann,
A. Drochner, H. Vogel, Chemical Engineering Journal 154 (2009)
333–340.
[13] W. Hauptmann, A. Drochner, H. Vogel, M. Votsmeier, J. Gieshoff, Topics in Catal-
ysis 42-43 (2007) 157–160.
[14] H. Sjövall, R.J. Blint, A. Gopinath, L. Olsson, Industrial & Engineering Chemistry
Research 49 (2010) 39–52.
Fig. 6. Parity plot for all measured NO concentration at reactor outlet. [15] P. Markatou, J. Dai, A. Johansson, W. Klink, M. Castagnola, T.C. Watling, M.
Tutuianu, SAE Technical Paper 2011-01-1304.
[16] D. Chatterjee, T. Burkhardt, M. Weibel, I. Nova, A. Grossale, E. Tronconi, SAE
at low water vapor concentrations. However, this phenomenon Technical Paper 2007-01-1136.
was not investigated further since all exhaust gases of technical [17] T.C. Brüggemann, F.J. Keil, Journal of Physical Chemistry C 115 (2011)
2114–2133.
relevance contain a certain water concentration.
[18] P.S. Metkar, V. Balakotaiah, M.P. Harold, Catalysis Today 184 (2012) 115–128.
Finally, a variation of reactor pressure between 1.55 and 6.63 bar [19] A. Tissler, T. Turek, M. Kögel, W. Schwieger, M. Rauscher, R. Mönnig, K. Kesore,
(Fig. 5), during which all concentrations are changed accordingly, European Patent EP 0955080 (B1), 1999.
[20] http://www.nist.gov/index.html (retrieved 15.04.09).
shows that the new kinetic equation is appropriate to describe
[21] M. Wulkow, Handbook Presto-Kinetics, Computing in Technology GmbH, Rast-
the NO oxidation kinetics over a broad range of temperatures and ede, 2005.
concentrations. It should be noted that the water concentration of [22] C. Perbandt, Kinetische Modellierung der Zersetzung von N2 O an Eisen-Zeolith-
3200 ppm at a total pressure of 6.63 bar would be equivalent to Katalysatoren in Abgasen der Salpetersäureproduktion, Dissertation, Clausthal
University of Technology, 2011.
about 2.1 vol% at atmospheric pressure.