Fundamental Studies of Catalytic Systems For Diesel Emission Control

T HESIS FOR THE D EGREE OF D OCTOR O F P HILOSOPHY
Fundamental studies of catalytic

systems for diesel emission control
Aging and sulfur treatment of a DOC, spatial investigation
and kinetic modeling of NO x reduction by ammonia
X AVIER A UVRAY
Department of Chemical and Biological Engineering

CHALMERS UNIVERSITY OF TECHNOLOGY
Gteborg, Sweden 2013
Fundamental studies of catalytic systems for diesel emission control

Aging and sulfur treatment of a DOC, spatial investigation and kinetic modeling of
NOx reduction by ammonia
XAVIER AUVRAY
ISBN: 978-91-7385-942-4
XAVIER AUVRAY, 2013
Doktorsavhandlingar vid Chalmers tekniska hgskola

Ny Serie nr. 3623
ISSN: 0346-718X

Competence Centre for Catalysis
Chalmers University of Technology
SE-412 96 Gteborg
Sweden
Telephone +46 (0)31 772 1000
Chalmers Reproservice
Gteborg, Sweden 2013
Fundamental studies of catalytic systems for diesel emission control

Aging and sulfur treatment of a DOC, spatial investigation and kinetic modeling of NOx reduction by
ammonia
XAVIER AUVRAY
Chalmers University of Technology
Abstract:
Due to global lean exhaust gas and new emission regulations, exhaust aftertreatment systems of
diesel engines are more and more sophisticated and composed of a series of catalytic units. In the
present work, two of these catalytic systems were studied with different approach. A model diesel
oxidation catalyst (DOC), used to convert nitric oxide into nitrogen dioxide and hydrocarbons and CO
into CO2, was examined in flow reactor experiments. A Cu-exchanged zeolite catalyst, devoted to the
lean NOx reduction by ammonia was studied with SpaciMS in operating conditions.
Since longevity and resistance to poisoning are two major challenges for automotive catalysts, the
effect of thermal aging in reactive atmosphere on NO oxidation activity was addressed and
correlated to platinum dispersion. Our experiments revealed the promotion of Pt sintering by SO2 as
well as the improvement of oxidation ability. Sintering in argon to obtain similar Pt dispersion did not
result in similar performance indicating the important role of aging atmosphere in subsequent
activity. The DOC was subjected to SO2 treatment in order to characterize the sulfur species formed
during SO2 poisoning and their impact on the oxidation of NO and C3H6. Two types of sulfur species
that differ in stored amount and impact on the activity were distinguished by TPR experiment.
However, both have a detrimental effect on the DOC performance. Finally, modification of the DOC
formulation by incorporation of acidity enhancer groups was carried out. The introduction of chlorine
and sulfate to increase the acidic nature of the support yielded suppression of catalyst deactivation
due to platinum oxide formation. However, this effect disappeared after aging and subsequent TPR
to 800C, suggesting a loss of these acidity-promoters.
The study of the NOx reduction catalyst was performed to evaluate its activity, NH3 storage capacity
and NH3 and NO oxidation ability. Intra-catalyst measurements were achieved with SpaciMS at Oak
Ridge National Laboratory. This technique provides insight of the reaction evolution throughout the
monolithic catalyst and showed the diminishing of the zone used for SCR reaction as the temperature
increased from 200 to 400C. The intra-catalyst concentration profiles are valuable data acquired in
realistic flow and temperature conditions and was utilized to develop a kinetic model for standard
NH3-SCR. The model accounts for the N2O production according to two routes and predicts well the
transient phenomena resulting from changes in gas composition.
Keywords: NO oxidation, DOC, NH3 SCR, aging, sulfur, SpaciMS, hysteresis, Pt/Al2O3, kinetic
modeling, catalyst acidity
ii
List of publications:
Paper I
The Effect Gas Composition During Thermal Aging On The Dispersion And NO Oxidation Activity
Over Pt/Al2O3 Catalysts.
Auvray X.; Pingel T.; Olsson E.; Olsson L.
Applied Catalysis B: Environmental, 129 (2013) pp. 517-527.
Paper II
Sulfur Dioxide Exposure: A Way To Improve The Oxidation Catalyst Performance
Auvray X.; Olsson L.
Industrial & Engineering Chemistry Research, 52, 41, (2013) pp. 14556-14566
Paper III
Effect Of Enhanced Support Acidity On The Sulfate Storage And The Activity Of Pt/-Al2O3 For NO
Oxidation And Propylene Oxidation
Auvray X.; Olsson L.
Catalysis Letters, accepted for publication (2013)
Paper IV
Local Ammonia Storage And Ammonia Inhibition In A Monolithic Copper-Beta Zeolite SCR Catalyst
Auvray X.; Partridge W. P.; Choi J. S.; Pihl J. A.; Yezerets A.; Kamasamudram K.; Currier N. W.;
Olsson L.
Applied Catalysis B: Environmental, 126 (2012) pp. 144-152.
Paper V
Kinetic Modeling Of NH3-SCR Over A Supported Cu Zeolite Catalyst Using Axial Species Distribution
Measurements
Auvray X.; Partridge W. P.; Choi J. S.; Pihl J. A.; Coehlo F.; Yezerets A.; Kamasamudram K.; Currier N.
W.; Olsson L.
In manuscript
iii
Contribution to the papers:

Paper I
I synthesized the catalysts and conducted all the reactor experiments. I interpreted the results with
my co-authors and was responsible for writing the manuscript.
Paper II
I synthesized the catalysts and conducted all the experiments. I interpreted the results together with
my co-author and was responsible for writing the manuscript.
Paper III
I designed and synthesized the catalysts and conducted all the experiments. I interpreted the results
with my co-author and was responsible for writing the manuscript.
Paper IV
I performed the experiments, analyzed the results together with my co-authors and wrote the first
manuscript.
Paper V
I performed the experiments and simulations, analyzed the results together with my co-authors and
was responsible for writing the manuscript.
iv
Contents
1. Introduction emission control catalysis........................................................................................ 1
2. Objectives ........................................................................................................................................ 5
3. Diesel Oxidation Catalyst ................................................................................................................. 7
3.1. NO oxidation over Pt .............................................................................................................. 7
3.1.1. Chemistry and thermodynamics .................................................................................... 7
3.1.2. Deactivation factors ....................................................................................................... 8
3.1.3. Oxygen adsorption and platinum oxidation ................................................................ 10
3.1.4. Interaction with other compounds .............................................................................. 11
3.2. Hydrocarbon oxidation ........................................................................................................ 12
3.2.1. Kinetics and reaction parameters of propene oxidation ............................................. 12
3.2.2. Particle size effect on propene oxidation .................................................................... 13
3.2.3. Thermal behavior during C3H6 oxidation ..................................................................... 13
3.2.4. Inhibition of C3H6 oxidation.......................................................................................... 14
4. Selective catalytic reduction of NOx by ammonia ......................................................................... 15
4.1. NH3-SCR ................................................................................................................................ 15
4.2. Catalysts ............................................................................................................................... 15
4.3. Side reactions ....................................................................................................................... 16
5. Catalyst synthesis and flow reactor measurements ..................................................................... 19
5.1. Catalyst preparation............................................................................................................. 19
5.2. Catalyst characterization...................................................................................................... 20
5.2.1. BET specific surface area measurement ...................................................................... 20
5.2.2. Platinum dispersion measurement .............................................................................. 20
5.2.3. Transmission electron microscopy .............................................................................. 21
5.2.4. Reactor setup ............................................................................................................... 22
5.2.5. Temperature control.................................................................................................... 23
5.2.6. NO oxidation ................................................................................................................ 23
5.2.7. C3H6 oxidation .............................................................................................................. 23
5.2.8. Aging under reactive atmosphere ............................................................................... 23
5.2.9. Advanced analytic technique: SpaciMS ....................................................................... 23
5.2.10. Reactor setup ............................................................................................................. 25
6. Challenges specific to automotive catalysts: aging and sulfur poisoning ..................................... 27
6.1. Aging of catalyst ................................................................................................................... 27
v
6.1.1. Thermal aging .............................................................................................................. 27

6.1.2. Effect of reactive atmosphere ..................................................................................... 29
6.2. Interaction of sulfur compounds with oxidation catalysts .................................................. 32
6.2.1. Adsorption and storage of sulfur species .................................................................... 32
6.2.2. Stability of sulfur adsorbates ....................................................................................... 33
6.2.3. Interaction of surface sulfur species with a DOC in operating conditions .................. 34
7. Effect of the support and modification of the acido-basic properties .......................................... 39
7.1. Support modification by alkaline additives .......................................................................... 39
7.2. Enhanced acidity by addition of chlorine ............................................................................. 40
7.3. Sulfation to increase the catalyst acidity ............................................................................. 40
7.4. Oxidation ability of an acidity-enhanced DOC ..................................................................... 41
7.5. SO2 aging of chlorinated and sulfated catalysts ................................................................... 43
7.6. SO2 storage of chlorinated and presulfated Pt/Al2O3 .......................................................... 44
8. Intra-catalyst insight of SCR with NH3 on Cu-BEA ......................................................................... 47
8.1. Experimental measurements ............................................................................................... 47
8.1.2. Selective catalytic reduction of NO by NH3.................................................................. 49
8.1.3. NH3 storage capacity .................................................................................................... 52
8.2. Kinetic modeling simulations ............................................................................................... 53
9. Concluding remarks ....................................................................................................................... 59
List of abbreviations .......................................................................................................................... 61
Acknowledgments ............................................................................................................................. 63
Bibliographic references.................................................................................................................... 65
vi
Chapter 1
1. Introduction emission control catalysis
The massive emission of CO2 started with the industrial era has led to a threatening global warming
due to CO2 greenhouse effect. Decisions were taken in order to decrease the global CO2 emissions
and, as a major emitter, the transportation sector has been targeted. In addition, with the oil
resources depletion and the always-increasing oil demand, a great effort has been devoted to
develop more fuel-efficient engines like diesel and lean-burn gasoline engines. For a more complete
reaction, these engines work at higher air-to-fuel ratio which lowers the fuel consumption and also
reduces carbon dioxide release. As a consequence, the nature of the exhaust gas is also modified and
is now globally oxidizing due to excess oxygen in the engine feed. This characteristic has a major
impact on the catalytic treatment of the exhaust. Pollution caused by combustion emissions has been
a concern for a long time and led to the introduction of emission regulations. In 1967, the United
States adopted the Clean Air Act to reduce vehicle emissions. To satisfy more stringent regulations,
oxidation catalysts, responsible for the carbon monoxide (CO) and hydrocarbons (HC) emission
control, were commercialized in the mid-1970s. In the 1980s, the three-way catalyst technology
(TWC) was developed to simultaneously reduce the emission HC, CO and NOx, i.e. NO and NO2 for
gasoline engine applications. TWC utilizes CO and HC to reduce NOx and is efficient within a narrow
air-to-fuel ratio window centered on 14.6, which corresponds to the stoichiometric operation of the
engine. This technology requires an accurate control of the fuel injection and air intake, which is
achieved by measuring the oxygen concentration of the exhausts, thus providing feedback to the
injection control unit. Diesel and lean-burn engines do not operate on this critical window, and
therefore the TWC can only oxidize HC and CO but cannot handle NOx in these oxidizing conditions.
The release of NOx in the atmosphere represents a health hazard in urban area. NO can react with
other chemical species to form very toxic compounds such as NO2 and ozone. NOx is one of the main
components of the brownish haze called smog that also contains particulate matters,
hydrocarbons and SO2. Under the action of sunlight, the extremely active ozone forms by reaction
between NOx and HC with severe consequence on air quality. The reactive nature of ozone can
provoke respiratory problems such as asthma, attack lung tissue and it inhibits the growth of plants.
On a more global level, NOx are also able to react with water to form nitric acid yielding acid rains.
Therefore, more restrictive emission regulations have been introduced and forced the design of new
catalytic systems to reduce NOx produced by diesel and lean-burn vehicles. Two processes aiming at
reducing vehicle NOx emissions have emerged in the 90s.
The lean NOx-trap (LNT) technique is based on cycling the operating regime of the engine between
fuel-lean and fuel-rich. During the lean period, NOx are stored on the catalyst designed to trap NOx in
form of surface nitrates thanks to an alkaline metal like barium. When the catalyst cannot trap 100 %
of the NOx, the regime is switched to rich mode by increasing the ratio of fuel/oxygen for a few
seconds. Thus the excess of unburned fuel in the exhaust is used as a reductant to convert stored
NOx into N2 and H2O. This technique is sensitive to sulfur and requires a very precise control of the
engine and the catalyst and the rich periods penalize the fuel economy.
1
Chapter 1 __________________________________________________________________________
The second method is the selective catalytic reduction (SCR) of NOx by injecting a reductant in the
gas stream upfront the catalyst. The mixture containing NOx and the reducing agent flows to a
catalyst, on which the reduction of NOx into N2 and H2O occurs. Many types of catalysts and
reductants have been studied. The fuel can be used as a source of hydrocarbons for HC-SCR, which
led to many studies [1, 2]. NH3 is also able to reduce selectively NO [3] and can be produced in situ by
the decomposition of a urea solution. NH3-SCR does not involve extra fuel consumption but demands
a tank to store the urea solution onboard as well as an accurate dosing system. The addition of a
subsequent ammonia slip catalyst (ASC) to control the excess NH3 that may slip out of the vehicle is
also a necessity.
As for gasoline engines, the after-treatment system of diesel engines includes an oxidation unit, but
with a slightly different function. Besides oxidizing CO and HC, the specific role of a diesel oxidation
catalyst (DOC) consists in converting NO into NO2, which is a suitable oxidant to burn the soot
trapped by the diesel particulate filter (DPF). NO2 is not only used to regenerate the DPF, it plays an
important role in the two aforementioned NOx abatement processes. Indeed, NO2 adsorbs more
readily than NO, enhancing the trapping performance of the LNT catalyst. The SCR is also improved
by the presence of NO2 up to a NO2/NO ratio of 1.
Figure 1: Combination of catalysts to reduce emissions of diesel vehicles.

In operating conditions, catalysts are exposed to variable temperature and gas composition that may
alter their structure and performance. The exhaust gas temperature attained downstream the DPF
can be very high due to the heat generated by the soot combustion. High temperature exposure
causes metal agglomeration and structural transition of the support material leading to the decrease
in catalytic surface area and deactivation. Fuel and lube oil contain small amount of inorganic
elements that end up in the exhausts and affect the catalyst. Despite the progress made in desulfurization of fuel, sulfur remains a major source of catalyst poisoning. It is oxidized in the engine
chamber to form SO2, which has a great affinity with conventional catalytic materials and blocks the
active sites by covering them. SO2 poisoning is particularly hindering on LNT because it forms stable
sulfates with the NOx-adsorbing material, thus gradually decreasing the NOx-storage capacity of the
LNT [4]. Phosphorous, sodium, zinc, magnesium and calcium figure also among potential poisons for
2
_________________________________________________ Introduction emission control catalysis

the catalyst. A great interest has been shown in studies of thermal and chemical deactivation in order
to understand the underlying mechanism and design more durable catalysts.
Chapter 2
2. Objectives
The major part of this thesis has been dedicated to the study of a typical oxidation catalyst in aging
conditions (Paper I-III). The effect of thermal treatment coupled to chemical exposure on the NO
oxidation activity was investigated in Paper I. The goal was to characterize the platinum sintering of a
1 wt.% Pt/Al2O3 catalyst aged in various atmosphere in a flow reactor. The Pt dispersion was
measured by CO chemisorption and the particles were directly observed by transmission electron
microscopy (TEM). The characterization was completed by the BET surface area measurement and
the activity test for NO oxidation.
SO2 is a typical compound contained in the vehicle exhausts and showed remarkable ability to
promote platinum sintering. The catalyst treated in this way interestingly showed an increased
activity towards NO oxidation in Paper I. A more detailed study focused on the effect of SO2 exposure
was then conducted in order to evaluate the influence and stability of the sulfur adsorbates formed
during aging and their interactions in subsequent reactions. Since the DOC mission is to oxidize both
NO and hydrocarbons, the activity test was extended to propene oxidation. The results of this work
are summarized in Paper II.
The formation of platinum oxides during NO oxidation has been noted in Paper I and II and credited
with the catalyst deactivation and inverse hysteresis phenomenon. The acidity enhancement of the
catalyst as a mean to suppress the Pt oxide formation was investigated in close relationship with the
effect of sulfation treatment. The results are reported in Paper III.
The second part of this thesis deals with the selective catalytic reduction (SCR) of NOx using NH3 as
reducing agent. Intra-catalyst gas analysis was performed with a SpaciMS to determine the spatial
concentration profiles of the components involved in key reactions occurring on the SCR catalyst. The
activity of a copper-exchanged zeolite catalyst was measured for NO oxidation, NH3 oxidation and
NH3-SCR. Ammonia storage was also characterized since ammonia storage capacity is a crucial
parameter for the NH3-SCR technology. Indeed the dosing of NH3 injection in the exhausts must take
into account the stored ammonia, which also participates to the NOx reduction. The results and the
conclusions drawn from the spatially-resolved data are reported in Paper IV.
SpaciMS is an excellent tool to measure intra-catalyst gas composition and to sample the gas flow at
several positions along the monolithic sample, giving a picture of the reaction rates. This type of data
is therefore invaluable to study the kinetics of a reaction and is appropriate to use when developing
kinetic models. The objective of the study presented in Paper V was to use the spatial data collected
with SpaciMS in Paper IV in order to develop a global kinetic model for the NH3-SCR process.
Chapter 3
3. Diesel Oxidation Catalyst
Oxidation catalysts were first used on spark-ignited gasoline engines to oxidize carbon monoxide and
hydrocarbons. Diesel vehicles did not need such technology at that moment due to their already low
CO and HC emissions. However with more stringent emission standards, the utilization of oxidation
catalyst was extended to diesel applications. Diesel oxidation catalysts (DOC) have the additional task
to convert NO into NO2, which participates in particulate filter regeneration and improves the NOx
reduction process. DOCs are noble metal based catalysts; they typically contain Pt and Pd dispersed
on high surface area supports like metal oxides. Commercial catalysts often contain additives that
improve their thermal stability and durability or provide a promoting function such as oxygen storage
capacity.
3.1. NO oxidation over Pt
3.1.1. Chemistry and thermodynamics
The oxidation of NO is an exothermic reaction, the standard enthalpy of which is -57.19 kJ/mol. The
stoichiometry of the global reaction is given by equation (3.1):
(3.1)
Concentration (ppm)
600
500
400
NO2
NO2
300
NO
200
100
0
0
200
400
600
Temperature (C)
800
Figure 2: NO and NO2 concentrations at thermodynamic equilibrium for a gas containing initially
8% O2 and 500ppm NO.
As shown in Figure 2, at low temperature, the most stable species is nitrogen dioxide and the
reaction (3.1) is favored. Conversely, the reaction is thermodynamically limited at high temperature.
The catalyst should therefore be particularly active at low temperature. The gas composition plays a
significant role in the conversion. Desprs et al. [5] noted the increase of NO conversion when the
7
Chapter 3: Diesel oxidation catalyst _____________________________________________________

concentration of oxygen was increased. In contrary, high NO concentration led to low NO conversion.
Kinetic studies [6, 7] report a global reaction order near 1 for O2 and NO and near -1 for NO2
indicating significant NO2 inhibition [5, 8]. The value of activation energy and the catalytic mechanism
used to model the reaction differ in the literature. The model developed by Olsson et al. [9, 10]
describes the experiments equally well with the Langmuir-Hinshelwood model, which considers the
surface reaction between adsorbed atomic oxygen and adsorbed NO, as with the Eley-Rideal model
[10], which implies reaction between adsorbed oxygen and gas phase NO. However, to describe and
include the NO2 decomposition, which requires two adjacent surface sites, the LangmuirHinshelwood model was preferred. Due to a low NO adsorption above 200C, Crocoll et al. used the
Eley-Rideal mechanism for NO oxidation [11]. In contrary, Bhatia et al. [8] found that NO coverage
was significant at low temperature and therefore should be included in kinetic models. In the model
by Mulla et al. [6], the adsorption of molecular dioxygen was considered to be the rate-determining
step. In this model, the apparent activation for NO oxidation was found to be 82 kJ/mol.
The particle size affects greatly the reaction rate and the conclusions of numerous studies [6, 12-18]
suggest that larger particles have a higher intrinsic activity, evidenced by a higher turn-over rate (rate
of moles of NO converted per Pt site). The reason commonly invoked is the greater resistance
towards oxidation of the larger particles. Indeed, Wang and coworkers [19] demonstrated that the
NO oxidation activity was lower on platinum oxides than on metallic platinum, supported by a higher
energy barrier for NO adsorption and O2 dissociation on PtO2. Platinum oxidation is discussed in more
detail in section 3.1.3. Matam et al. [20] suggested that particle morphology played an even more
important role than particle size in the NO oxidation activity. They determined the rates for two aged
catalysts; the one with higher dispersion presented the higher normalized rate constant, owing to
more active particle morphology. They concluded that cuboctahedral morphology was the best
configuration compared to truncated cubic and spherical Pt particles.
3.1.2. Deactivation factors
During NO oxidation, Pt/Al2O3 catalysts are prone to deactivate. This fact has been evidenced by the
formation of an inverse hysteresis when the catalysts are successively heated and cooled during NO
oxidation conditions [21-23]. The conversion obtained during the cooling was lower than the one
measured during initial heating (Figure 3), indicating significant deactivation. Time-on-stream
experiments (Figure 4) have also shown the decrease of NO conversion during isothermal step [14,
21]. Most of the authors who have investigated this phenomenon assigned the deactivation of the
reversible formation of platinum oxide in operating conditions [6, 21]. The production of NO2, known
to be a powerful oxidant, as well as the high oxygen concentration is a factor enabled to explain the
platinum reversible oxidation. Hauff et al. [21] showed that NO2 had no influence on catalyst
deactivation in presence of 12 % O2, suggesting that oxygen has a significant responsibility in
platinum oxidation.
_____________________________________________________________ 3.1. NO oxidation over Pt
NO conversion (%)
60
Increasing T
Decreasing T
50
40
30
20
10
0
0
100
200
300
400
Catalyst temperature (C)
500
600
Figure 3: Hysteresis phenomenon during NO oxidation over Pt/Al2O3 (500ppm NO + 8% O2 + 5% H2O).

The kinetic study of Mulla et al. [7] pointed out the negative effect of NO2 evidenced by the increase
of the apparent activation energy, from 39 kJ/mol to 82 kJ/mol, due to the presence of NO2 in the
feed. They also determined by CO titration that the oxygen uptake of the catalyst was higher when it
had been exposed to the reaction mixture (NO + O2) than to oxygen [6]. The oxygen uptake was even
higher upon exposure to NO2 + O2 which indicates the oxidation of platinum during activity
measurement. The XPS results by Olsson and Fridell [14] depicted the formation of platinum oxide,
enhanced by the presence of barium. Desprs et al. [5] also detected platinum oxide after
pretreatment with O2 and NO2. Several studies demonstrated the negative effect of NO2 on the
reaction rate. The catalytic oxidation of NO on Pt-based catalysts is therefore self-inhibited. The
interaction between dioxygen and the platinum is a crucial area in NO oxidation studies on platinum
because oxygen dissociation on Pt surface is an important step in the catalytic process but, on the
other hand, the strong oxygen-Pt interaction and oxygen diffusion in platinum particle cause
deactivation. This is discussed in the next section.
300
NO conversion (%)
70
250
60
200
50
40
150
30
100
20
NO conversion
T cat
10
0
0
5000
10000
50
80
0
15000
Time (s)
Figure 4: Catalyst deactivation with time on stream (3h) at 260C in activity test conditions
(500ppm NO, 8% O2, 5% H2O).
3.1.3. Oxygen adsorption and platinum oxidation
NO oxidation requires weakly bound oxygen on the surface [24]. However, on a clean Pt surface,
oxygen forms stable bond with Pt. The increase of oxygen coverage increases the repulsive
interaction between adsorbed O, creating more reactive oxygen. The weakening of Pt-O bond can
also be achieved by an elevation of the temperature. Mudiyanselage et al. [24] evaluated the
minimum oxygen coverage at 0.28 monolayer in order to produce this weakly bound oxygen.
Ovesson et al. [25] calculated that repulsive lateral interactions between oxygen atoms and NO were
necessary to render the NO oxidation exothermic.
Adsorbed oxygen is necessary for the reaction but in oxidizing environment, platinum oxides can
form and deactivate the catalyst. The studies by Wang and Yeh [26, 27] on the oxidation of platinum
nanoparticles give a valuable insight on the platinum dynamics in presence of O2. Their calorimetric
and gravimetric measurements revealed the different phases in the oxidation mechanism. At low
temperature (<300 K), oxygen adsorbs on the surface with a high adsorption enthalpy. At higher
temperature (300 < T < 750 K), the oxygen uptake per surface Pt increases and the heat of adsorption
decreases, indicating greater oxidation and formation of surface Pt oxide. According to Wang and
Yeh [26, 27], the stoichiometry of the surface platinum oxide depends on the particle size. For
smaller particles (d<1.3 nm), the observed O/Pts ratio indicates that PtsO2 was formed while PtsO was
created on larger particles (d>2.0 nm). Finally at T>800 K, oxygen and/or platinum dioxide desorption
occurs. Briot et al. [28] formulated a relationship between Pt particle size and adsorbed oxygen
reactivity. Their calorimetric measurements showed a lower heat of oxygen adsorption on sintered
catalysts accompanied with a higher TOR for methane oxidation. Furthermore, the reaction between
H2 and adsorbed oxygen occurred at lower temperature on the sintered catalyst implying higher
reactivity of oxygen on larger particles. McCabe et al. [29] studied the passivating oxidation of
platinum. The independence of oxygen uptake with oxygen pretreatment duration and temperature
and the presence of both metallic Pt and Pt oxide signature in EXAFS spectra led the authors to
conclude that only the platinum surface layer was subjected to oxidation. McCabe et al. also noted a
10
_____________________________________________________________ 3.1. NO oxidation over Pt

relationship between Pt oxidation and particle size. It was found that small particles were reduced at
higher temperature than larger particles, suggesting the higher stability of platinum oxide on small
particles. It has been later confirmed that the heat of oxygen adsorption on Pt linearly decreases with
the increase of particle size [26]. The size-dependent stoichiometry of Pt oxide measured by McCabe
et al. [29] and Wang and Yeh [26, 27] are in agreement. The formation of less stable oxide on larger
particle can explain the higher activity of sintered catalysts discussed in the section 3.1.1 and
evidenced in Paper I and Paper II.
3.1.4. Interaction with other compounds
The NO oxidation, like most of catalytic reactions, is sensitive to chemical compounds present in the
reactive gas mixture due to the more intense competition for adsorption sites on the catalyst. Since
vehicle exhausts contain a wide variety of components, it is necessary to study their individual effect
on the reaction of interest. The effect of the most common chemicals in automotive application such
as water, HC, CO and CO2 has been investigated. Other compounds such as SO2, P and Ca, present in
smaller quantities in automotive exhausts, have raised interest due to their major effect on catalytic
systems.
The presence of propene inhibits the NO oxidation due to competitive adsorption and consumption
of the produced NO2 as oxidant for the propene oxidation reaction [30-32]. As a consequence NO
oxidation is suppressed at temperature below the C3H6 combustion light-off temperature. Propane
and dodecane hinder the oxidation of NO over Pt/Al2O3 as reported by Olsson et al. [33]. NOx
consumption was also observed when the gas flow contained dodecane, suggesting the occurrence
of lean NOx reduction by HC. The absence of this reaction in the case of propane indicates that the
hydrocarbon chain length plays a selective role in the NOx reduction. In the same manner, CO
completely inhibits NO oxidation on Pt/Al2O3 until its total oxidation into CO2 [21], which then
facilitates the adsorption of NO. In a simulation, Crocoll et al. [11] calculated a high CO coverage and
measured with DRIFT a strong adsorption of CO on Pt, resulting in a severe inhibition of NO
oxidation. However, on Pt/TiO2 catalyst, Irfan et al. [34, 35] observed the NO oxidation promotion by
CO, which was attributed to the reduction of oxidized Pt by CO.
The influence of the most abundant compounds, namely CO2 and H2O, on NO oxidation has been
investigated [11, 33]. CO2 does not affect NO oxidation, which is supported by a theoretical low CO2
coverage [11]. The addition of H2O in the typical NO oxidation gas mixture decreases the NO2
production [11, 33]. In addition, a puff of NO was detected when water was switched on [33];
indicating the replacement and release of adsorbed NO by H2O. The detrimental effect of 5 % H2O
and 500 ppm C3H6 on the NO oxidation activity of Pt/Al2O3 catalyst is evidenced in Figure 5.
11
90
NO conversion (%)
80
70
60
50
40
30
20
10
0
0
100
200
300
400
500
NO only
5% H2O
H2O
5% H
H2O
+ 500ppm
500ppm CC3H6
2O +
3H6
C3H6
500ppm C3H6
600
Figure 5: Effect of H2O and C3H6 on NO oxidation over Pt/Al2O3 (500ppm NO, 8% O2, 5C/min).
3.2. Hydrocarbon oxidation
The first function of an automotive oxidation catalyst is to oxidize hydrocarbons and carbon
monoxide that can remain in the exhausts into CO2. Owing to its strong toxicity, CO must not be
produced by HC combustion, which means that catalysts used must be selective to CO2.
Hydrocarbons represent a large group of molecules presenting different stability. Methane, for
instance, is highly stable because its C-H bonds are difficult to break up. This is also the case for
saturated HC (alkanes or paraffins) with short carbon chain. Unsaturated HC (alkenes or olefins), such
as propene (called also propylene), are characterized by a C=C double bond that confers them a
higher reactivity than alkanes. The high electron density of propene favors the interaction and
adsorption on a catalytic surface. After exposure of Pt/Al2O3 to C3H6 and C3H8, Yao et al. [36]
observed no propane chemisorption but a strong propene chemisorption. The results reported also
showed that propene is oxidized on Pt/Al2O3 catalysts at lower temperature than the corresponding
alkane, namely propane. In this thesis, C3H6 combustion was used as a probe reaction to measure the
influence of SO2 aging on a DOC.
3.2.1. Kinetics and reaction parameters of propene oxidation
In order to determine the mechanism and the rate determining step of the catalytic combustion of
propene, many kinetic studies have been conducted [36-39]. The Langmuir-Hinshelwood model,
implying surface reaction between propylene and oxygen, is the preferred mechanism. TPD of
propene and oxygen evidenced that propene adsorbed associatively on Pt/Al2O3 and is more strongly
bound to Pt than oxygen [38]. This is in agreement with the negative partial order found for propene
and the positive partial order calculated for oxygen [37, 38]. Based on these findings, it was
suggested that propene is self-inhibiting by covering the platinum sites and that the dissociative
adsorption of O2 was the rate determining step of the reaction on Pt/Al2O3. The reported activation
12
___________________________________________________________ 3.2 Hydrocarbon oxidation

energy varies in the literature: 85.8 kJ/mol [38], 20-50 kJ/mol [37] and 41.9 kJ/mol [39] have been
estimated.
3.2.2. Particle size effect on propene oxidation
Carballo and Wolf [39] early noted that, at low propylene concentration, the reaction rate increased
with propylene concentration until a maximum. The rate then decreased with further propylene
concentration increase and became propylene independent. They also measured a higher specific
rate on aged catalyst which indicated a higher intrinsic activity of larger Pt particles. The global rate
was, however, higher on well-dispersed catalysts. Denton et al. [17] found that the dispersion was
the critical factor in the C3H6 combustion. They examined Pt catalysts on different supports with
different porosity, impurity content and a wide range of dispersion and reported that low-dispersion
catalysts of all categories were the most active for propene oxidation. In contrary, Haneda et al. [40]
recently observed the decline of C3H6 oxidation with the dispersion loss, in agreement with Marcot
et al. [41].
3.2.3. Thermal behavior during C3H6 oxidation
The propene combustion is exothermic with a standard enthalpy of -2058.4 kJ/mol. In a monolithic
catalyst, the heat released by the reaction generates a temperature gradient that, in its turn,
influences the local reaction rate. Figure 6 shows the temperature rise due to combustion light-off
and the temperature gradient existing between the inlet gas and the intra-catalyst gas at 100 % C3H6
conversion. The propene combustion ignition has been investigated using spatially resolved
technique by Epling et al. [42, 43]. It was shown during a temperature ramping that the conversion
increases from front to back, at low conversion. As a consequence, a temperature rise appeared first
in the monolith back. The temperature rise propagated to the upstream positions, as the 100 %
conversion point shifts closer to the inlet, describing a back-to-front heat wave. For a given axial
position, the temperature peaked when the conversion reached 100 % and then decreased as the
ignition front moved to the inlet. At the end, only a short part of the catalyst was used to oxidize
propene and an important temperature gradient developed: the front temperature, where the
reaction occurs, was much higher than the back temperature. Finally, thermal aging yielded a slower
back-to-front propagation of the light-off [42], whereas sulfation increased the light-off temperature
but increased the speed of the ignition wave [43].
13
600
250
500
200
400
150
300
100
0
1500
100
Propylene
T cat
T reactor
200
4500
7500
Time (s)
50
10500
Temperature (C)
C3H6 concentration (ppm)
0
13500
Figure 6: Temperature profiles measured at catalyst inlet (T reactor) and inside the monolith (T cat)
during propylene oxidation. Propylene concentration profile showing the correlation between
temperature variation and combustion light-off is also added.
3.2.4. Inhibition of C3H6 oxidation
Propene oxidation in presence of other components has been studied, especially for the purpose of
lean NOx reduction by HC [31, 44]. As previously mentioned, propene has a detrimental effect on NO
oxidation due to its strong adsorption on Pt. This inhibition is reciprocal since the increase of the
light-off temperature due to the presence of NO has been observed [30-32] and the explanation
proposed invoked the occupation of active sites by NO. Propene oxidation is also self-inhibited by the
strong propene adsorption and low concentration is therefore recommended to obtain a high
conversion, as noted by Carballo and Wolf [39]. Haneda et al. [40] ascribed the low rate on lowdispersion catalysts to the slow reaction of formed acrylate intermediates. Thus, it was suggested
that the accumulation of acrylate species on the active sites inhibited the C3H6 oxidation. During CO
oxidation, the presence of C3H6 induced the building of more hindering surface intermediates,
suggesting surface poisoning by C3H6 [45].
14
Chapter 4
4. Selective catalytic reduction of NOx by ammonia
4.1. NH3-SCR
The reduction of NOx with ammonia has proved to be efficient when applied to stationary sources
and does not require extra fuel to reduce NOx, unlike NOx storage and reduction (NSR) and
hydrocarbon-SCR. On the other hand, ammonia, usually in the form of urea dissolved in solution,
must be stored and carried onboard in an additional tank which is also constraining the application
on passenger cars but acceptable for heavy-duty trucks.
The stoichiometry for NH3 SCR in oxygen-containing conditions involves as much NO as NH3 and gives
N2 and H2O (Eq. 4.1)
4 NO + 4 NH3 + 3 O2 4 N2 + 6 H2O
(4.1)
The reaction (4.1) is often referred to as Standard SCR and takes place when the NO content of NOx
is 100 %. Most of the NOx released from the engine is on the form of NO. However, the DOC
upstream of the SCR catalyst produces significant amount of NO2. When a mixture of both NO and
NO2 is available, the two following reactions can occur:
4 NH3 + 2 NO + 2 NO2 4 N2 + H2O
(4.2)
4 NH3 + 3 NO2 7/2 N2 + 6 H2O
(4.3)
Reaction (4.2) is called Fast-SCR because it occurs faster than the Standard SCR and shows a
maximum rate with equal amounts of NO and NO2 [46-49]. The relative rate of the third reaction,
denoted NO2-SCR, depends on the catalyst and the conditions under which it is used. The NO2 SCR
can be faster than the Standard SCR on a Fe-zeolite catalyst [50] but slower on a vanadia-based
catalyst [47]. In general, the rate of reaction of SCR increases with concentrations of O2 and NOx but
NH3 inhibits the reaction by blocking the active sites for NO adsorption and oxidation [50]. This
inhibition is more severe on iron-containing catalysts.
4.2. Catalysts
Commercial catalysts based on vanadium oxide, like V2O5-WO3/TiO2, have been extensively studied
for their excellent activity and selectivity [47, 48] for ammonia SCR. Their major drawbacks are the
volatility and toxicity of vanadium oxide and a lack of efficiency at low and high temperatures. The
ideal catalyst for an automotive application would cover a wide range of temperature since it is low
after starting the engine (cold start) and high during the exothermic regeneration of the particulate
filter, located upstream. Ion-exchanged zeolite catalysts have also been investigated extensively [46,
51-53]. Iron and copper ion-exchanged zeolites have been especially well-studied and have
evidenced a good ability to reduce NOx. The recent Cu-chabazite formulation also shows good
hydrothermal and HC resistance [54-56]. Owing to their storage ability, zeolites can store unused NH3
that can be subsequently utilized for SCR. This feature can facilitate the NH3 dosing into the catalytic
system.
15
Chapter 4: Selective catalytic reduction of NOx by ammonia __________________________________

4.3. Side reactions
In general, several reactions other than SCR involving either NO or NH3 occur on the catalyst. These
reactions have a detrimental impact on the selectivity and on the catalyst efficiency. NH3 oxidation
might result in over-consumption of NH3.
At low temperature (<200C), the main side reaction is the formation of ammonium nitrate [47, 48,
51] onto the catalyst (see reaction (4.4)). NH4NO3 can form a deposit on the catalyst or sublimate to
exit the catalyst in the gas flow.
2 NH3 + 2 NO2 NH4NO3 + N2 + H2O
(4.4)
The NH4NO3 can subsequently decompose into N2O and water (eq. 4.5), which is suggested in many
studies and is problematic since it affects directly the selectivity.
NH4NO3 N2O + 2 H2O
(4.5)
An alternative proposed by Koebel et al. [47] is the decomposition in NH3 and nitric acid to oxidize
NO to NO2 which enhances the SCR reaction. In that case, ammonium nitrate is an intermediate
species that builds up in presence of NO2 and then decomposes according to the SCR reaction. The
second undesirable reaction is the formation of N2O which decreases the selectivity. One route is the
direct oxidation of NH3 according to the following reaction (4.6).
4 NH3 + 4 O2 2 N2O + 6 H2O
(4.6)
This reaction is the major source of N2O over vanadia-based catalysts. Since NO2 is a stronger oxidant
than O2, it also participates in the oxidation of NH3. This reaction was found to be predominant over
Fe-zeolite SCR catalyst by Devadas et al. [46] whereas no N2O was formed when only NH3 and O2
flowed through the catalyst suggesting that the direct NH3 oxidation (reaction (4.6)) is not
responsible for N2O production on this type of catalyst. This is also the case for Cu zeolites. At high
temperature, N2O decomposes into N2 and O2 and can also be reduced in parallel by NH3 yielding an
over-consumption of NH3. With intra-catalyst measurements, Luo et al. showed that N2O builds up in
the front part of Fe/zeolite catalysts and that a fraction subsequently decomposes in the rear part of
the catalyst at 500C [50]. At low temperature N2O generation can be explained by the
decomposition of ammonium nitrate as previously mentioned (reaction (4.5)).
At high temperature, direct oxidation of ammonia is observed leading to an over-consumption of
NH3. This reaction lowers the NOx reduction due to the competition for ammonia. It has been
observed over Cu-ZSM-5 that this reaction starts above 325C when only NH3 and O2 are in the gas
flow [52] and the decrease of SCR observed at higher temperature can be ascribed to ammonia
oxidation [52, 57]. Moreover the product of ammonia oxidation observed was N2 [52] and not NO
suggesting the following reaction equation:
4 NH3 + 3 O2 2 N2 + 6 H2O
(4.7)
Cu-exchanged zeolite catalysts show a high level of activity for SCR with ammonia and also limit side
reactions. The NH3 conversion and the NO conversion were found to be equal up to around 600 K,
indicating a NH3/NO stoichiometry of 1 [52, 57]. At higher temperature, NH3 starts to be oxidized by
O2 and a too high copper content produces more N2O.
16
__________________________________________________________________ 4.3. Side reactions

To summarize, NO2 enhances the NH3 SCR but is also responsible for NH4NO3 formation and N2O
generation depending on the temperature. Thus, the NO2/NOx ratio should be well adjusted and, for
optimal operation of the NOx emission control catalytic system, it should not exceed 0.5.
17
18
Chapter 5
5. Catalyst synthesis and flow reactor measurements
5.1. Catalyst preparation
The catalyst synthesized and studied for diesel oxidation catalyst application was a 1 wt.% Pt/Al2O3,
prepared by wet impregnation and freeze-drying technique. The deposition of noble metal onto
alumina surface was performed by mixing an aqueous solution of Pt precursor, i.e. Pt(NO3)2, with an
alumina slurry, under constant stirring. The alumina used had been calcined at 750C in air, and the
pH of the slurry was adjusted to 4 before the drop-wise addition of the Pt solution. The water used
for synthesis was milliq water dispensed by a Millipore water purifier. After 1h under stirring, the
obtained blend was poured in a special flask to be frozen by dipping the flask in liquid nitrogen. The
flask containing the frozen catalyst was immediately mounted on a freeze-dryer apparatus in order
to sublimate the water contained in the sample in solid form. At the end of the drying step, a yellow
powder was collected and calcined in air at 500C for 2h. The calcination step was carried out to
decompose the nitrates originating from the precursor. The powder color turned to dark grey during
this step.
Monolithic cores of cordierite were cut, calcined 2h at 600C and impregnated with a blend of
powder containing 80 wt.% of the catalyst described above and 20 wt.% of a binder to optimize the
attachment of the washcoat to the channel walls. The Pt/Al2O3 powder was dissolved in a solution
containing 50 wt.% ethanol and 50 wt.% water. The liquid-to-solid ratio is an important practical
parameter since too dense slurry may clog the channels of the monolith while diluted slurry allows
the deposition of a smaller amount of catalyst and requires too many dipping and drying steps. After
soaking the monolith in the solution and homogeneous wetting of the monolith with the slurry,
drying was carried out to evaporate water and ethanol at 90C. Calcination of the deposited catalyst
layer was then performed using a hot air gun at 600C for 1min. The evaporation of the liquid phase
was carried out thoroughly to deposit a uniformly thick layer of active phase throughout the
monolith. The operation of impregnation and drying was repeated until the correct amount of
washcoat was deposited on the walls of the monolith channels. The final mass of washcoat was
around 20 wt.% of the total mass of the monolithic catalyst.
In paper III, the effect of support acidity on oxidation reactions and SO2 storage was investigated and
required the synthesis of Cl-Pt/Al2O3 and SO4-Pt/Al2O3 catalysts. Both new catalysts were prepared in
the same manner, by mixing a standard Pt/Al2O3 catalyst with a concentrated acidic solution
containing Cl- and SO42- ions, respectively. The exchange between OH- groups from alumina and Cl/SO42- anions occurred, which was depicted by the increase of the pH of the solution (Figure 7).
The chlorination of alumina is a well-known process to increase the acidity of alumina and used for
instance in the alkylation and dehydration of alkanes. The usual chlorination method is based on
pretreatment of the catalyst with a gas flow containing chlorine species such as tetrachloromethane
(CCl4), hydrochloric acid (HCl) and phosgene (COCl2). Figure 8 represents the substitution reaction of
surface hydroxyle by chlorine in HCl.
19
Chapter 5: Catalyst synthesis and flow reactor experiments __________________________________
3
pH
pH
(a) sulfation
(b) chlorination
0
0
50
100
Time (min)
150
200
100
200
300
Time (min)
400
500
Figure 7: Evolution of the pH during preparation of sulfated (a) and chlorinated (b) Pt/Al2O3.
OH
OH
Cl
Cl
Al
Al
+ HCl
Al
O
Al
O
Hydroxylated Al2O3
+ H2O
Chlorinated Al2O3, higher

acidity
Figure 8: Mechanism of alumina chlorination by substitution of hydroxyle surface groups.

The incorporation of sulfate can be done by impregnation of the catalyst with sulfate salts like
ammonium sulfate [58]. Exposition of the catalyst to a gas flow containing H2S [59, 60] or SO2 [61] is
also widely used for catalyst sulfidation. Wet impregnation with sulfuric acid was encountered in
literature [62], which was applied also in our work.
5.2. Catalyst characterization
5.2.1. BET specific surface area measurement
The samples prepared in Paper I were characterized using adsorption of probe molecules to
determine their available surface area before and after aging. The surface area was measured by
physisorption of N2 at 77 K and calculated according to the Brunauer-Emmett-Teller Theory (BET).
This measurement gives a value representing the physical surface area of the catalyst without
differentiating between surfaces of different chemical properties, like metal particle and metal oxide
support. An automatic instrument Micromeritics ASAP2010 was used for this measurement. The
measurements were carried out on the coated monoliths after drying in vacuum for 2 h at 225C.
5.2.2. Platinum dispersion measurement
CO chemisorption experiment was performed in the flow reactor described in section 5.2.4 in order
to estimate the platinum dispersion, i.e. the fraction of platinum particle surface exposed. A 20 min
pulse of gas containing 100 ppm of CO was first sent over the catalyst where CO adsorbed. A flow of
20
__________________________________________________________ 5.2. Catalyst characterization

pure argon was then fed to remove weakly bound CO and a second pulse similar to the first one was
conducted for 10 min. During the whole experiment, the total flow was 1L/min and the carrier gas
was Ar. The uptake of CO was measured for the two steps. The chemisorbed CO was then calculated
by subtracting the physisorbed CO and the reactor system contribution, measured in the second CO
pulse. The number of surface Pt atoms was inferred assuming that CO adsorbed on the surface with
the ratio CO:Pt= 0.8 according to the results of Foger et al [63] and Olsson et al. [33] (Figure 9). The
expression for the dispersion was then:
Figure 9: Schema of CO adsorption on Pt particle surface for measuring the dispersion.

Dispersion and average particle size are correlated and different relations have been proposed
assuming different particle shapes. Assuming spherical particles (or hemispherical with
the flat surface in contact with the support) Guo et al. proposed the relation
where dPt is
the average particle size and DPt is the platinum dispersion [64]. The CO:Pt ratio was kept constant
for all the analyzed data, assuming that the stoichiometry of CO chemisorption does not depend on
particle size. This simplification is not fundamentally correct since the stoichiometry of CO adsorption
on Pt varies between 0.6 and 1 [65].
5.2.3. Transmission electron microscopy
Transmission electron microscopy was used in Paper I to characterize the fresh and SO2 aged
catalysts and to support the dispersion results calculated from CO chemisorption. The catalysts
powders were ground in an agate mortar and subsequently put on holey carbon films on TEM Cu
grids. The particles were imaged using an FEI Titan 80-300 TEM with a probe Cs (spherical aberration)
corrector operated at 300 kV. The images were recorded using a high angle annular dark field
(HAADF) detector in the scanning TEM imaging mode (STEM), providing Z number contrast. The
electron probe size was about 0.2 nm for these studies. Figure 10 shows a STEM micrograph of the
fresh catalyst. The synthesized catalysts are well-dispersed with a low average particle size (1nm).
21
Figure 10: STEM micrograph of the fresh 1wt% Pt/Al2O3 catalyst.

5.2.4. Reactor setup
All experiments (aging, dispersion measurement and activity test) were performed in similar flow
reactors composed of a quartz tube heated by a coiled electric resistance (DOC studies) or a furnace
(SCR study) in which the monolithic catalyst was placed. The reactors were equipped with a set of
mass flow controllers that supplied gases with the appropriate volumetric flow. The outlet of the
reactors was connected to analysis instruments in order to measure and record the effluent gas
chemical composition. In paper I, specific detectors were mounted on the reactor to detect CO, CO2,
NO and NO2. The reactor used to perform the other DOC studies was equipped with a mass
spectrometer and a gas FTIR analyzer that could simultaneously detect quantitatively various gases.
For the intra-catalyst study of NH3-SCR, one MS was used.
Figure 11: Schematic representation of the reactor setup used in Paper I.
22

5.2.5. Temperature control
In all experiments, thermocouples were inserted from the back of the reactor. The number of
thermocouples used varied with the reactor employed. In general, one thermocouple was placed in
front of the catalyst and controlled the heating system to make the temperature equal to the set
temperature at that position. A second thermocouple was placed in a channel of the monolith in
order to measure the catalyst temperature. Efforts were made in order to reproduce the probed
position (chosen channel and depth into this channel). An additional thermocouple and an empty
monolith were placed downstream in Paper I, to limit and check the temperature gradient.
5.2.6. NO oxidation
NO oxidation activity was measured on degreened and aged catalysts using the following procedure.
A flow of 3000ml/min containing 500 ppm NO and 8 % O2 with or without water vapor (5 %) was sent
over the catalyst for 30 min at room temperature in order to saturate the surface with NO. Then the
temperature was ramped up to 500C at the rate of 5C/min. The NO and NO2 concentrations of the
outlet gas was quantified by a chemiluminescence NOx analyzer and recorded at the frequency of
1Hz, in Paper I. In Paper II and III, multigas FTIR was employed to record quantitatively the NOx outlet
concentrations.
After the ramping, the temperature was held constant for 5 min at 500C to make sure that steady
state was established. The same gas flow was maintained during the cooling. In Paper II and III, the
cooling was operated at the same controlled rate as the heating in order to measure the activity
during extinction and characterize the hysteresis behavior of the reaction.
5.2.7. C3H6 oxidation
C3H6 oxidation activity was evaluated in flow reactor by similar temperature-programmed reaction in
presence of 5 % H2O. The temperature was increased and then decreased at a rate of 1C/min
between 130C and 250C. Propene outlet concentration was measured by the FTIR instrument.
5.2.8. Aging under reactive atmosphere
The aging procedure was the same for each catalyst and all aging temperatures, and was performed
in the flow reactor used for the activity test. Initially an Ar flow of 1000 mL/min was sent over the
catalyst for 10min at room temperature before starting to heat the reactor up to the aging
temperature. As soon as the reactor temperature stabilized at the target temperature the catalyst
was exposed to its reactive atmosphere for 2h. After this period, the aging atmosphere was replaced
by a flow of pure Ar and the reactor was cooled. The total flow remained constant during the entire
process (1000 ml/min) and only the gas composition varied from pure Ar to aging atmosphere. Like
the heating, the cooling was as short as possible. To accelerate the cooling step, the insulation wool
around the reactor was immediately removed after the 2h aging step.
5.2.9. Advanced analytic technique: SpaciMS
In automotive catalysis, monolithic converters are used in order to achieve a high gas-catalyst
interface and a low pressure drop. Ideally, a monolith behaves like a plug-flow reactor and
concentrations of reactants and products evolve along the catalyst. These intra-catalyst evolutions
cannot be detected with the usual flow reactor experiments and analytical techniques. It is of
importance to know what occurs inside the catalyst in order to assess the reaction mechanisms. For
23

example, reaction intermediates can be formed at the front of the catalyst and subsequently react.
Intra-catalyst measurement is a large advantage in order to investigate these features.
For this purpose the Spatially Resolved Capillary Inlet technique has been developed and was first
adapted to mass spectrometer by the Oak Ridge National Laboratory and Cummins Inc. [4, 66-69]
and was later also used together with FTIR spectrometer [50]. A flexible capillary is connected to the
instrument inlet and the other end is inserted into a selected channel of the monolith (Figure 12).
The capillary is typically made of fused-silica and coated with polyimide; its outer diameter is ca. 150200 m. The capillary tip can be translated throughout the catalyst in order to determine the gas
composition at any position. Such data are valuable for the development of kinetic models since it
allows one to collect data at realistic flow rate and temperature leading to high, if not full, conversion
at the catalyst outlet.
Figure 12: Schematic representation of SpaciMS system.

One difficulty with intra-catalyst measurement is the flow disturbances caused by the capillary. This
effect depends on the volume of the capillary actually present in the catalyst. Thus, the outer
diameter of the capillary is an important parameter. It should be much less than the channel
diameter to limit the impact on the gas flow. The capillary probe pumps reaction gases, which
modifies locally the concentrations of the reactants, the pressure and the flow velocity. A low
sampling rate of ca. 10 L/min limits this effect. S et al. [70] have investigated the intra-catalyst
oscillations of the concentrations and temperature for the oxidation of CO. Using measurements and
CFD simulations, they demonstrated that the intrusion of a capillary probe in a channel had a
negligible effect on the catalyst behavior. Another drawback of using a capillary inlet is the
interactions between gas molecules and the inner surface of the capillary. They become nonnegligible as the inner diameter of the capillary decreases and when the flow is decreasing.
Interactions are greater for sticky molecules, such as NH3 and NO2. This can cause that the
instrument response to a pulse is a more or less broad peak which complicates the time resolved
analysis. The capillary was heated and maintained at 200C to weaken molecular interactions with
the capillary walls and reduce adsorption inside the capillary. In this way, the time resolution was
improved.
24

SpaciMS has been successfully used at ORNL for investigating lean NOx-trap catalyst behavior during
storage and regeneration [66-68] and fuel cells [69]. Other groups have performed intra-catalyst
measurements on various reactions to get an insight into the spatial consumption of reactants [42,
50, 71, 72].
For the second part of this thesis, the SpaciMS technique has been adapted to study the reduction of
NO by NH3 according to the NH3-SCR reaction. These experiments were conducted at Oak Ridge
National Laboratory in USA.
5.2.10. Reactor setup
The micro reactor used for the NH3-SCR study is shown in Figure 13. It was composed of two
cylindrical furnaces. The larger one was the pre-heater, in which feed gases circulated into coiled
pipes to be heated up before subsequent mixing. The furnace with smaller diameter contained the
quartz-tube reactor into which the catalyst was placed. Quartz rods were placed in the tube
upstream and downstream from the catalyst to create gas mixing and to lower the axial temperature
gradients inside the reactor.
The capillary probes were inserted into the tube from the inlet (right side of the reactor on Figure 13)
and K-type thermocouples were inserted from the outlet. The gas flows were metered by mass flow
controllers and sent to two three-way switching valves, which were computer controlled. These
electrovalves enable a fast turnaround time when switching gas stream to the reactor. The gas lines
between the pre-heater and the reactor, at the outlet of the reactor as well as the capillaries were
heated and maintained at a temperature above 200C to avoid adsorption in the pipes and the
formation of ammonium nitrate. Two capillary probes, the tips of which were separated by
1.25 inches, were actually used during all experiments. One capillary was placed into the catalyst,
while the second was always located in front of the catalyst. Both capillaries were connected to a
multiport valve, connected to the MS inlet. The gas flow from either of the capillaries could be
selected for analysis and switching from one capillary to the other was made during experiment with
a remote control (Figure 12). The inlet gas composition was determined at the end of each step, by
switching to the inlet capillary, to get a better contrast between inlet and intra-catalyst
concentrations.
Figure 13: Picture of the reactor setup for the spatio-temporal study of NH3-SCR at Oak Ridge
National Laboratory.
25
26
Chapter 6
6. Challenges specific to automotive catalysts: aging and sulfur
poisoning
Catalysts often lose their activity with operating time and need to be replaced or regenerated. In
automotive emission control catalysis, this deactivation is mainly caused by durable catalyst
poisoning and sintering. The sintering is the loss of catalytic surface due to collapse of the pore
structure and metal agglomeration. These phenomena take place at elevated temperature. The
poisoning is the alteration of the catalyst caused by a chemical component that reacts with or blocks
the active sites.
6.1. Aging of catalyst
6.1.1. Thermal aging
When a supported catalyst is subjected to high temperature, structural changes occur. The porous
support can lose surface area by pore collapse leading to the formation of larger pores. Phase
transition can also affect the surface area of the support material. Thus, loose packed -Al2O3 is
converted to close packed -Al2O3, the stable form of alumina, at high temperature. The temperature
increases also the mobility of the metal nanoparticles on the surface, which consequently start to
migrate. This phenomenon results in metal agglomeration in larger particles, more
thermodynamically stable due to a lower fraction of low-coordinated atoms. This process is called
sintering and generally causes a loss of activity correlated to the loss of exposed surface sites. Figure
14 illustrates the decrease of Pt dispersion owing to aging at high temperature (600, 700, 800 and
900C) in an inert gas flow. A linear dispersion decrease can be noted until 700C, followed by a
plateau at higher temperature. The mechanism and the kinetic of sintering have been studied to
develop a sintering model. Two models were proposed: the particle migration model [73] and the
atomic migration model [74, 75]. In the former, Ruckenstein and Pulvermacher described the
formation of large particles by migration of small particles on the surface and their incorporation to
the larger particles structure by collision. This model implies the decrease of the total number of
particles. Flynn and Wanke [74, 75] suggested that sintering was initiated by extraction of metal
atoms or molecules from small particles followed by the rapid transport of these free atoms to large
particles. This mechanism, also known as Oswald ripening, predicts simultaneous shrinkage of small
particles and growth of larger particles. It can also explain the redispersion of sintered catalysts by a
rapid cooling, early in the sintering process, when atoms accumulate on the surface or the presence
of trapping sites on the support that are able to anchor the free atoms during their migration. The
driving force of Flynn and Wankes model is the particle size difference which implies no sintering of
unisized catalyst and faster sintering for a catalyst presenting a broad particle size distribution.
Experiments have showed that a mixture of fresh and presintered catalyst, i.e. containing large
particles, sinters more rapidly [76].
27
Chapter 6: Challenges specific to automotive catalysis: aging and sulfur poisoning ________________
In Paper I, aging in argon showed to decrease the platinum dispersion linearly until 700C (Figure 14).
After further aging at higher temperature, only minor dispersion loss was observed and the
dispersion leveled off.
70
Pt dispersion (%)
60
50
40
30
20
10
0
400
500
600
700
800
Catalyst temperature during the aging (C)
900
Figure 14: Pt dispersion decrease by thermal aging (2h at 600, 700, 800 and 900C) in argon flow.
The activity resulting from the dispersion changes are presented in Figure 15, where the activity
improvement by each aging step is evidenced. These results show that large particles are more active
for NO oxidation, in agreement with several studies in literature [6, 12, 15, 17].
80
NO conversion (%)
70
Aging temperature (C)
60
900
800
700
600
500
50
40
30
20
10
0
0
100
200
300
400
500
600
Figure 15: NO oxidation activity of Pt/Al2O3 degreened (labeled as 500) and aged at 600, 700, 800 and
900C. Gas composition: 500ppm NO + 8% O2 in Ar (3L/min); ramping rate 5C/min.
28
_________________________________________________________________ 6.1. Aging of catalyst

6.1.2. Effect of reactive atmosphere
The objective in Paper I was to investigate the role of different aging atmosphere on the dispersion
evolution. Indeed, the atmosphere has a significant role in the sintering rate since it can influence the
support-metal interaction and yields the formation of intermediate species, such as PtOx in the case
of aging in oxygen, that enhance the Pt atom mobility. Figure 16 reports the dispersion of catalysts
aged in different atmospheres after each aging step.
Effect of O2
In presence of oxygen, the dispersion decreases linearly with aging temperature, as for argon aging.
However, the minimum dispersion attained after aging at 900C was lower in presence of O2. Oxygen
can form platinum oxides molecules that migrate from the particles according to the atomic
migration mechanism [77]. This process is able to explain the redispersion observed in presence of
oxygen [78] at temperature below 600C. Above this temperature, Pt oxides decompose to metallic
platinum and oxygen. Lee and Kim [78] suggested that Pt atoms formed on the surface were likely
very mobile and would be captured by large particles. At high temperature, Pt can be transported in
vapor phase as PtO2(g) and deposited on the surface where the oxide decomposes. Matam et al. [20]
noted the more pronounced decrease of dispersion, measured by CO chemisorption, of a model DOC
aged in oxidative atmosphere compared to lean diesel exhausts atmosphere, containing also CO and
HC. They also observed that aging in air led to heterogeneous particle size and the formation of very
large particles.
Effect of H2
In H2, the dispersion decreased only moderately at 600C indicating a slower sintering process. After
aging at 700C, the catalyst aged in H2 still had the highest dispersion but presented, after the last
aging step, similar dispersion as the argon catalyst.
29
100
90
Ar
Ar
H2
H2
O2
O2
SO2
SO2
O2/SO2
O2/SO2
Pt dispersion (%)
80
70
60
50
40
30
20
10
0
400
500
600
700
800
900
1000
Figure 16: Evolution of Pt dispersion with aging temperature of catalysts aged in different
atmospheres (Ar, 10% O2, 1% H2, 30ppm SO2 and 30ppm SO2 + 10% O2).
Effect of SO2
Figure 16 shows that the presence of SO2 during aging led to significant dispersion drop already at
600C. However, the apparent Pt dispersion increased after aging at 800C in 30 ppm SO2 which
could be the result of either Pt redispersion, promoted by SO2, or underestimation of dispersion by
CO chemisorption after aging at 600 and 700C. The reduction treatment prior to dispersion
measurement can reduce SO2 species to atomic S, which covers the platinum particles and decreases
the available platinum surface. The redispersion feature has been reproduced on a second sample
aged in SO2, but attempt to redisperse aged catalysts by SO2 treatment at 800C failed. It is possible
that the dispersion after SO2 aging at temperature lower than 800C was underestimated. The
important sintering at the end of the procedure was however confirmed by microscopy analysis.
Figure 17 evidenced the sintering with the presence of large particles homogeneously dispersed on
the surface (Figure 17a) having similar size (15-20nm) and faceted morphology (Figure 17b). Chang et
al. [79] investigated sulfur poisoning of Pt/Al2O3 catalysts and observed by EXAFS an important Pt
sintering caused by sulfur. Olsson and Karlsson [80] observed a significant decrease of Pt dispersion
upon SO2 exposure, accompanied with a continuous NO conversion increase during SO2 exposure.
However, Krkkinen et al. [81] did not find any sintering evidence by SO2 treatment of a Pt/Al2O3
catalyst.
30
_________________________________________________________________ 6.1. Aging of catalyst
(a)
(b)
Figure 17: STEM images of Pt/Al2O3 aged in SO2 + O2 at the end of the aging procedure (last aging
step at 900C).
100
Peak NO conversion (%)
90
80
70
60
Ar
Ar
H2
H2
O2
O2
SO2
SO2
O2/SO2
O2/SO2
50
40
30
20
10
0
400
500
600
700
800
900
1000
Figure 18: Evolution of catalytic activity, represented by the maximum of NO conversion reached, of
Pt/Al2O3 catalysts aged in different atmospheres (Ar, 10% O2, 1% H2, 30ppm SO2 and 30ppm SO2 +
10% O2).
Figure 18 represents the highest NO conversion of the catalysts after every aging steps. In general,
NO oxidation activity was improved by thermal aging, as large particles demonstrated higher activity
than small ones. Differences arose from aging atmosphere though. The activity of the two samples
aged in presence of oxygen increased dramatically after the first aging (>70 % conversion) whereas
31
the three other samples were moderately improved (<50 % conversion). The second particularity of
the oxygen aged catalysts is the activity decline after aging at high temperature, which takes place at
800C in O2 and at 900C in O2+SO2. The catalysts aged in Ar, H2 and SO2 were improved by each aging
step.
The relation between activity and Pt dispersion is not straightforward indicating atmosphere-induced
effects on the Pt particles. Cabi et al. [82] investigated with environmental TEM the morphology
changes of Pt particles induced by the nature of the gas surrounding the nanoparticles. They
evidenced the facet transition from (111), more stable under H2 pressure, to (001) in presence of O2.
From our aging results, it could be noticed that the catalysts aged in SO2 and SO2+O2 showed the
greatest activity (>80 % NO conversion), which suggested a beneficial effect of SO2 aging and directed
the continued research work in that direction.
6.2. Interaction of sulfur compounds with oxidation catalysts
SO2 enters into the vehicle exhausts due to the presence of sulfur in the fuel as well as in the lube oil.
Sulfur dioxide has a great ability to adsorb and store as sulfite or sulfate on a metal oxide material
such as alumina. Alkaline sulfates, such as barium sulfates, are easily formed which results in that SO2
is a difficult poison for Lean NOx-trap Ba-based catalysts. This NOx-removal technology aims at
trapping NOx by formation of barium nitrate prior their reduction into N2. The storage of sulfur on
barium sites reduces dramatically the NOx-storage capacity.
However, SO2 has shown interesting properties in promotion of diverse reactions. The promotion of
propane oxidation [36, 61, 83, 84] and methane oxidation [85] by SO2 has been reported.
6.2.1. Adsorption and storage of sulfur species
SO2 adsorption and storage on catalysts has been studied by infra-red absorption and temperatureprogrammed reduction (TPR). Yao et al. [36] identified the formation of sulfites formed on alumina,
which were subsequently oxidized into sulfates upon SO2 exposure. SO2 oxidation and consequently
SO2 storage is facilitated by the presence of Pt [23, 36, 86]. In absence of O2, Corro et al. [61]
observed the adsorption of SO2 below 250C, which was accompanied by a second SO2 uptake at
higher temperature in presence of O2 forming sulfate species. This second uptake was greater and
took place at lower temperature on Pt/Al2O3 than on Al2O3, indicating the promotion of Pt on SO2
storage. SO2 adsorption is also promoted by the presence of NO, both in pre-adsorbed and gas form
[87]. Two main species form on alumina: sulfate, favored at high exposure temperature and long
exposure time, and weakly chemisorbed SO2. The SO2 storage decreases at higher deposition
temperature but the sulfate amount is not affected. The exposure temperature has therefore a
major effect on the adsorbed species nature and stability. Wakita et al. [88] observed by XPS the
simultaneous presence of sulfur as both SO42- and S2- upon exposure of catalysts to SO2 or H2S. They
concluded that sulfate was stored on alumina and S2- covered Pt. Krcher et al. [23] suggested that
sulfation in presence of water processes via oxidation of SO2 into SO3 and formation of liquid sulfuric
acid on the surface. In a slower step, subsequent sulfation of the washcoat by stable species occurred
by sulfur migration, independently of SO2 concentration.
The support plays a crucial role in SO2 storage since it has been observed that SiO2, unlike Al2O3, was
unable to store SO2 [89]. SO2 storage is also very low on TiO2 support [34]. In contrary, the presence
32
________________________________ 6.2. Interaction of sulfur compounds with oxidation catalysts

of absorbent elements such as barium boosts the sulfur storage [90]. According to the results by
Streber et al. [91], on Pt (111) single crystal surface, SO2 adsorbs molecularly at low temperature
(120 K). A part of SO2 is converted to SO3 on oxygen precovered surface. On the clean Pt surface,
slow heating creates SO2 transition from up-standing to lying formation before all SO2 is desorbed at
360 K. A small amount of atomic S produced by SO2 disproportionation remains at that temperature.
On O-precovered surface, heating causes the increasing SO3 coverage at the expense of SO2. SO3 is
subsequently oxidized in SO4. The SOx species starts desorbing at 230 K until the surface becomes
sulfur-free at 520 K. Adsorption and thermal evolution of SO2 on Pt(110) [92] also showed the SO2
molecular adsorption at low temperature and the conversion to SO3 and atomic S and subsequently
SO4. After heating to 650 K, SO4 has desorbed and S has almost completely left the surface. This
temperature suggests that in our experiments the platinum is sulfur-free after the reduction at 450C
preceding the dispersion measurement.
6.2.2. Stability of sulfur adsorbates
The IR absorption bands attributed to sulfates by Yao et al. [36] disappeared at 800C in vacuum or
700C in H2 flow, indicating the high stability of these adsorbates. It has been reported that SO2
adsorbs weakly on the basic sites and strongly on the acidic sites of alumina. The adsorption of
weakly adsorbed SO2 is increased by the presence of O2, likely by creating new basic sites and their
release takes place below 600C [87]. However, the presence of both NO and O2 converts weakly
adsorbed species into strongly adsorbed SO2, likely Al-O-SO3 or SO2. Nam and Gavalas [93] have
4
identified two kinds of SO2 adsorbed species on -Al2O3 that differ in stability and reduction products
distribution. The more weakly adsorbed type desorbs without reduction as SO2 while the most stable
species desorbs only during reduction as H2S. This finding is in agreement with our results showing a
category of species that decomposes into SO2 and H2S upon reduction at 500C (Figure 19) and a
second type of more stable species, reduced at higher temperature and producing solely H2S (Figure
20). Figure 19 displays the SO2 and H2S release observed during reduction at 500C of SO2-aged
Pt/Al2O3. Before the introduction of H2 thermal desorption of SO2 weakly adsorbed can be noticed.
Introduction of H2 immediately provokes an intense SO2 release followed by the H2S desorption, as
observed by Nam and Gavalas [93].
Luo et al. [94] showed that sulfur release off a DOC was the most effective in presence of H2,
compared to CO, CO2, H2O and inert He. In contrary, oxidizing conditions inhibited sulfur release. The
relative H2S/SO2 ratio during release of sulfur was found to be dependent on the initial sulfur loading
with a higher SO2 proportion released at high sulfur loading [94]. In reductive atmosphere, the DOC
desulfation was achieved at 700C. In the same study, Luo et al. proposed that desulfation takes
place on Pt and is therefore controlled by surface sulfates migration to Pt particles.
33
900
Introduction of 2% H2
Outlet concentration (ppm) /

Temperature (C)
800
H
H2S
2S
SO
SO2
2
feed
TT feed
700
600
500
400
300
200
100
0
-100
0
2000
4000
Time (s)
6000
8000
140
120
100
800
700
M-Pt-M23
H
2S
TTcatalyst
cat
600
80
500
60
400
40
300
20
200
100
-20
2000
4000
6000
8000
Time(s)
10000
Temperature (C)
Outlet concentration (ppm)
Figure 19: Species released by reduction (2% H2 ) at 500C of Pt/Al2O3 after SO2 exposure.
0
12000
Figure 20: H2S release during TPR of a sulfated Pt/Al2O3 (2% H2, 5C/min).
6.2.3. Interaction of surface sulfur species with a DOC in operating conditions
Many studies have revealed a promotional effect of SO2 in combustion of saturated HC such as
propane [36, 61, 83, 84]. In contrary, SO2 reduces the activity of Pt/Al2O3 catalysts for CO oxidation as
well as propylene oxidation [36, 43]. Li et al. [95] demonstrated that freshly deposited sulfur has a
detrimental effect on NO and C3H6 oxidation due to its location close to the active sites, supported by
a change in activation energy. On the other hand, they showed that vehicle-aged high sulfur loading
catalysts presented similar activity before and after desulfation. Further experiment consisting in
sulfation followed by a long step at 400C showed that sulfur migrated from active sites to the
support resulting in a minor impact on the subsequent activity, as sulfur deposited on vehicle-aged
34

catalysts. Pazmio et al. [96] evidenced by EXAFS the migration of sulfur from Pt particles to the
support during NO oxidation conditions and concluded that the formation of nitrates could displace
sulfur from Pt. Our results are in line with the sulfur migration from Pt sites to the support. They also
support the conclusion of Li et al. [95] since we evidenced the major detrimental impact of the less
stable sulfur species, presumably located on the Pt or the Pt-support interface, while the more stable
species caused a minor activity loss for NO and C3H6 oxidation.
70
Degreened
NO conversion (%)
60
(a) Heating
Sulfated-reduced
Sulfated
50
40
30
20
10
0
0
100
200
300
400
500
600
70
Degreened
NO conversion (%)
60
(b) Cooling
Sulfated-reduced
50
Sulfated
40
30
20
10
0
0
100
200
300
400
500
600
Figure 21: NO conversion measured by temperature-programmed reaction (500ppm NO + 8% O2 +

5% H2O) over Pt/Al2O3 degreened, sulfated (22h; 250C; 30ppm SO2 + 8% O2 + 500ppm NO + 5% H2O)
and reduced after sulfation (1h; 500C; 2% H2). (a) Heating 5C/min. (b) Cooling -5C/min.
35
In Paper III, SO2 exposure of a model DOC was conducted for 22h at 250C in presence of water,
oxygen and nitric oxide. As shown in Figure 19 and Figure 20, this treatment leads to catalyst
sulfation by creation of two types of sulfur species.
Figure 21 shows the impact of sulfur species that can be released under reducing condition at 500C
on NO oxidation activity. The comparison between the activity of the degreened and the sulfated
catalyst during heating shown in Figure 21a demonstrates the negative effect of this type of species.
This is confirmed by the significant improvement observed after reduction and release of these sulfur
adsorbates. During cooling (Figure 21b), the sulfated catalyst gained activity to display the same
profile as the reduced catalyst. This observation combined with the absence of sulfur desorption
during the activity tests suggests that the sulfur species have been displaced from active sites to the
support and therefore had a minor effect during the cooling.
The SO2 exposure effect is worth to mention since a tremendous activity increase was noted after
sulfation and reduction. However, if the samples were not reduced after the sulfation, it had a
negative effect on C3H6 combustion, characterized by an increase of 50 % conversion temperature
(T50) of 12C. The same effect of DOC sulfation was reported by Russell et al. [43]. Since C3H6
oxidation test directly followed the NO oxidation step in our protocol, the lower C3H6 oxidation
performance, during heating (Figure 22a), due to sulfation indicates that sulfur migrated back to the
active sites after having been displaced by NO at high temperature as observed by Pazmio et al.
[96]. However, this sulfur migration did not occur in the presence of propylene as suggested by the
absence of activity recovery during cooling (Figure 22b). The reason can be the lower temperature
applied in the C3H6 oxidation experiment or the weaker competition of C3H6 for adsorption sites. The
reduction of sulfated sample, however, revealed the beneficial effect of SO2 exposure on C3H6
combustion since higher activity was measured after the treatment for both heating and cooling.
36
C3H6 conversion (%)
120
(a) Heating
100
Degreened
Sulfated-reduced
Sulfated
80
60
40
20
0
120
140
160
180
Feed temperature (C)
200
220
200
220
120
C3H6 conversion (%)
(b) Cooling
100
Degreened
Sulfated-reduced
Sulfated
80
60
40
20
0
120
140
160
180
Figure 22: C3H6 conversion measured by temperature-programmed reaction (500ppm C3H6 + 8% O2 +

5% H2O) over Pt/Al2O3 degreened, sulfated (22h; 250C; 30ppm SO2 + 8% O2 + 500ppm NO + 5% H2O)
and reduced after sulfation (1h; 500C; 2% H2). (a) Heating 5C/min. (b) Cooling -5C/min.
The sulfation also influenced the conversion hysteresis for both NO and C3H6 oxidation since it
became larger indicating either a strong promotion during heating or a limiting effect during cooling.
The activity measured after desulfation gave information to clarify this point. Indeed, after release of
the most stable fraction of sulfur species (Figure 20), the NO oxidation activity during heating was
not modified (Figure 23a), whereas the activity during cooling was improved by the absence of these
surface species (Figure 23b).
37
70
70
(a) Heating
50
40
30
20
Sulfated-reduced
Desulfated
10
(b) Cooling
60
NO conversion (%)
NO conversion (%)
60
50
40
30
20
Sulfated-reduced
Desulfated
10
0
0
200
400
600
200
400
Figure 23: Effect of stable sulfur species created by SO2 exposure on NO oxidation. The reduced
catalyst contains the most stable species and the desulfated was sulfur-free. (a) Heating 5C/min.
(b) Cooling -5C/min.
This result implies that the stable sulfur species, likely sulfates stored on alumina support, are able to
interact during the cooling and decrease the activity. The migration of sulfates from the support to
the active sites in operating condition is a possible way to explain this finding. After desulfation,
similar conclusion can be drawn concerning propene oxidation, with an unchanged ignition activity
and an improved extinction activity.
38
600
Chapter 7
7. Effect of the support and modification of the acido-basic properties
In diesel oxidation catalysts, like in most of heterogeneous catalysts, the active phase is dispersed on
a support material, which provides high surface area, mechanical strength and thermal stability. The
interaction between metal particles and the support is an important parameter to anchor and retain
the metal and avoid deactivation due to particle growth. Metal oxides such as Al2O3, TiO2, SiO2 and
ZrO2 are widely used as catalytic supports. The support can be inert for the reaction or present
reactive surface sites. Thus, Brnsted and Lewis acid sites are accessible on alumina (Al2O3) surface.
These sites can for instance store molecules in reaction conditions and participate in the reaction.
Many Pt-based catalysts with various metal oxide supports have been screened for NO oxidation
application. Schmitz et al. [16] evaluated the impact of various parameters such as Pt precursors, Pt
loading and support on the NO oxidation activity. It was concluded that the support had the most
important effect and SiO2-supported Pt catalysts show higher rate than Al2O3 catalysts. The
classification established by Benard et al. [15] regarding activity ranks the catalysts as follow: Pt/SiO2
> Pt/Al2O3 > Pt/CeO2 > Pt/TiO2 > Pt/ZrO2 = Pt/YSZ = Pt/Al2O3-SiO2 which is in agreement with other
studies [13, 16]. It is worth to note that the acid strength of Al2O3, SiO2 and ZrO2, evaluated by the
color change of Hammetts indicators, follows the same trend since on a pKa scale: SiO2 < Al2O3 <
ZrO2 [62]. There seems to be a correlation between support acidity and NO oxidation activity of
supported-Pt catalysts. Yazawa et al. [62, 97-99] have described the effect of acidic metal oxide
supports on the oxidation state of platinum. They demonstrated that acidic supports create an
electron deficiency on platinum, which becomes therefore more resistant to oxidation.
7.1. Support modification by alkaline additives
To control the nature of the support, a mixed support material can be used and dopant can be
added. Alkali and alkaline earths are additives that provide higher electron density to the catalysts
and are able to improve the activity of Pt/Al2O3 towards lean NO reduction with C3H6 [31, 100].
The promotion by alkali and alkaline earth (Na, Ba, K, Rb, Li, Cs) has been described by Yentekakis et
al. [31, 101] for the lean reduction of NOx with various reductants over Pt and Pd-based catalysts. Napromotion was particularly investigated and improved activity and N2-selectivity at low temperature
has been reported for NO reduction by propene [31]. Propene oxidation was also promoted by Na
whereas NO oxidation was inhibited by the Na doping. High sodium loading, however, led to a
complete suppression of NO reduction by propene.
The alkali promotion originates in the strengthening of the metal-NO bond along with the weakening
of the N-O bond. In general, the increased electron density induced by the alkali promoter
strengthens the metal-electron acceptor adsorbate bonds and reduces the strength of the metalelectron donor adsorbate bonds. The NOx reduction by propene process suffers from the relative
adsorption of NO and propene. Indeed, there is a competition between the two reactants to adsorb
on the catalyst. An equal coverage of NO and C3H6 is desired to optimize the reaction rate but, in
reality, propene adsorbs strongly on Pt/Al2O3 catalysts and covers the surface, which becomes
39
Chapter 7: Effect of the support and modification of the acido-basic properties __________________
inaccessible to NO. Na promotion enables to change the adsorption equilibrium by decreasing the
propene adsorption ability and increasing NO adsorption. It was demonstrated that the Pt-O bond
strength, known to play an important role in NO oxidation, was strengthened by alkali promotion.
Strong Pt-O bond and platinum oxidation inhibits NO oxidation over Pt catalysts, which led us to
examine the possibility of lowering the Pt-O bond strength.
7.2. Enhanced acidity by addition of chlorine
Chlorine belongs to the halogen group of elements and is therefore electronegative, i.e. it has a
strong ability to drag electrons. The presence of chlorine in the catalyst composition can therefore
decrease the electron density of the active metal and affect negatively the catalyst performance.
However, the electronegativity of chlorine enhances the acidic nature of the catalyst [102], which is
the wanted feature to carry out some catalytic reactions. Chlorinated alumina catalysts are employed
in paraffins chemistry to perform acido-basic reactions such as isomerization, cyclisation, cracking
[103-105]. During preparation, chlorine anions replace the hydroxyl groups of the alumina surface.
7.3. Sulfation to increase the catalyst acidity
Electro-attractor groups also influence the acido-basic nature of supported-metal catalysts. Thus
catalysts can be modified by adsorption and storage of molecules such as SO2. Extensive work has
been performed on the effects of catalyst sulfation, which has been described as one way of
preparing solid superacids, i.e. having an acidity higher than 100 % sulfuric acid [106]. The formation
of adsorbed sulfur oxide complexes boosts the strength of acid sites. In order to identify the
adsorbed species and the acido-basic centers, IR spectroscopy was widely used. Sulfate species
absorb IR and the IR absorption of an appropriate probe molecule, adsorbed on the catalyst, enables
to quantify and evaluate the strength of acido-basic sites. In particular, the complex structures,
presented in Figure 24 generate strong acidity due to the S=O bonds that can readily accommodate
an excess of electrons [106]. Yamaguchi et al. [106, 107] have identified the bidentate sulfate (A)
species as the compound that is mainly responsible for the enhanced acidity, whereas Saur et al.
[108] ruled out the formation sulfate (A) on Al2O3 and TiO2. Instead Saur et al. inferred that the
sulfate species responsible for acidity contained only one S=O bond and proposed the adsorption of
species (B) (Figure 24) that can be converted in contact with water into Brnsted acid sites.
O
O
O
M
O
M
O
M
O
M
(A)
O
M
(B)
Figure 24: Configuration of sulfates adsorbed on metal oxide surface identified as acidity promoter.
40
______________________________________________ 7.3. Sulfation to increase the catalyst acidity

The creation of Brnsted acid sites, absent on pure alumina, has been evidenced on sulfated alumina
by the protonation of pyridine, a rather weak base [109, 110]. This new acidic centers conferred a
better oxidation activity and a significant ability to cracking. Pure alumina however was more active
for reduction than sulfated alumina. Przystajko et al. [109] postulated that sulfate groups
destabilized adjacent hydroxyl groups and favored their protonation, generating the Brnsted sites.
Sulfated zirconia showed very high acidity and good propane combustion activity has been achieved
by Pt deposited onto this material [62]. The sulfate addition modified tremendously the nature of the
catalyst since unmodified Pt/ZrO2 catalyst was not active at all for the combustion of propane. The
strength of weak Lewis acid sites of ZrO2 and TiO2 was increased by sulfation [110]. Fe2O3 is also
greatly promoted by sulfation. The presence of sulfur in S6+ form generates indeed high acidity and
high activity for cyclopropane isomerization [107].
7.4. Oxidation ability of an acidity-enhanced DOC
Two acidity-enhanced Pt/Al2O3 were prepared by chlorination and sulfation in acidic solution to
evaluate the effect of support acidity towards NO and C3H6 oxidation. The first consequence of the
acidic treatment was the increase of platinum dispersion compared to the unmodified Pt/Al2O3
catalyst. The temperature-programmed technique used to evaluate the effect of surface sulfur
species was again employed.
NO oxidation
The degreened sulfated catalyst (M-S) and chlorinated catalyst (M-Cl) presented a lower oxidation
activity than the standard degreened Pt/Al2O3 (M-Pt) during heating (Figure 25a). This is probably
attributable to the poisoning of active sites by the excess of sulfate and chlorine rather than the
enhanced acidity. However during cooling, unlike unmodified Pt catalyst, M-S and M-Cl showed a
better NO oxidation activity than during heating (Figure 26), suggesting that the active sites were
operational. Thus, during cooling, the sulfated sample turned out to be the most active, while M-Cl
and M-Pt displayed very similar activity profiles (Figure 25b).
60
(a) Heating
50
NO conversion (%)
NO conversion (%)
60
M-S
M-Cl
M-Pt
40
30
20
10
(b) Cooling
50
M-S
40
M-Cl
30
M-Pt
20
10
0
0
200
400
600
200
400
Figure 25: NO conversion comparison (500ppm NO + 8% O2 + 5% H2O) between chlorinated Pt/Al2O3

(M-Cl), sulfated Pt/Al2O3 (M-S) and unmodified Pt/Al2O3 (M-Pt) during (a) heating (5C/min) and
(b) cooling (-5C/min).
41
600
50
45
40
35
30
25
20
15
10
5
0
(a) chlorinated
NO conversion (%)
NO conversion (%)
Heating
Cooling
100
200
300
400
500
600
50
45
40
35
30
25
20
15
10
5
0
(b) sulfated
Heating
Cooling
100
200
300
400
500
Figure 26: NO conversion profiles (500ppm NO + 8% O2 + 5% H2O) during heating (5C/min) and
cooling (-5C/min) of (a) chlorinated Pt/Al2O3 (M-Cl) and (b) sulfated Pt/Al2O3 (M-S).
The conclusion that can be drawn from the results reported above is that enhanced acidity limits the
deactivation of the catalyst due to platinum oxide formation usually observed during cooling. The
electroattraction of sulfates and chlorine groups decrease the electron density of platinum particles
and thus decreases platinum ability to bind strongly oxygen atoms. Sulfation proved to be more
efficient than chlorination for this purpose.
C3H6 oxidation
After NO oxidation, propylene oxidation was performed over each of the three catalysts (M-Pt, M-S,
M-Cl). The three activity profiles are comparable during heating (Figure 27 a), which indicates that
the species that caused the initial low NO oxidation performance were removed. It also means that
propene combustion is affected only to a small extent by the enhanced support acidity. However,
during cooling, there is a trend showing the slight negative effect of enhanced acidity on propylene
combustion (Figure 27 b). This result is in line with the improved propene oxidation obtained by
alkaline doping by Yentekakis et al. [31] since, in our case, the electron density of Pt is expected to be
lowered.
42
600
__________________________________________ 7.4. Oxidation ability of an acidity-enhanced DOC
120
(a) Heating
100
C3H6 conversion (%)
C3H6 conversion (%)
120
80
60
M-S
M-Cl
M-Pt
40
80
60
20
0
150
170
190
210
230
Feed temperature ( C)
M-S
40
20
130
(b) Cooling
100
M-Cl
M-Pt
130
250
150
170
190
210
230
250
Figure 27: C3H6 conversion comparison (500ppm C3H6 + 8% O2 + 5% H2O) between chlorinated
Pt/Al2O3 (M-Cl), sulfated Pt/Al2O3 (M-S) and unmodified Pt/Al2O3 (M-Pt) during (a) heating (1C/min)
and (b) cooling (-1C/min).
7.5. SO2 aging of chlorinated and sulfated catalysts
The acidity-enhanced catalysts were aged for 22 h at 250C in the following gas mixture: 500 ppm NO
+ 8 % O2 + 30 ppm SO2 + 5 % H2O diluted in argon (3 L/min). They were then reduced for one hour at
500C in 2% H2. The aging and reduction procedure increased the oxidation ability of M-Cl and M-S as
it does for the unmodified Pt/Al2O3 catalyst. However, it can be seen that both catalysts were less
active during cooling for NO oxidation (Figure 28) and therefore followed the usual trend observed
for Pt/Al2O3 catalyst. That suggests the loss of their acidic properties during aging that conferred a
protection against Pt oxides formation.
80
60
50
40
30
20
Heating
Cooling
10
0
0
(b) sulfated
70
NO conversion (%)
70
NO conversion (%)
80
(a) chlorinated
200
400
60
50
40
30
20
Heating
Cooling
10
0
600
200
400
Figure 28: NO conversion profiles (500ppm NO + 8% O2 + 5% H2O) during heating (5C/min) and
cooling (-5C/min) of (a) chlorinated Pt/Al2O3 (M-Cl) and (b) sulfated Pt/Al2O3 (M-S) after aging with
500ppm NO + 8% O2 + 30ppm SO2 + 5% H2O and reduction.
43
600
7.6. SO2 storage of chlorinated and presulfated Pt/Al2O3
Similar sulfation treatment as described in Paper I and III was carried out to assess the SO2 storage
capacity of chlorinated and presulfated Pt/Al2O3. Subsequently, one reduction step at 500C was
performed in order to remove and quantify the less stable species and one TPR up to 800C was
conducted to remove the most stable species. The results of the sulfur release are presented in
Figure 29 and 30. In general, the sulfur storage is lower on acidity-enhanced catalyst than on
unmodified Pt/Al2O3 (Figure 29). Silica, which is more acidic than alumina, is known to be insensitive
to SO2, which is line with our results. The distribution between less and more stable sulfur species is
interesting since it differs greatly between samples. While on Pt/Al2O3 the two types of species were
released in equivalent amount, important difference exists in their respective proportion on M-S and
M-Cl, as shown in Figure 29. M-S released a high amount of low-stability species that corresponds to
the maximum observed among the three catalysts. However a small amount of sulfur, compared to
M-Pt and M-Cl (Figure 30), was released during TPR, indicating a low amount of stable species. A
fraction of released sulfur likely originates from the sulfates deposited in the catalyst preparation.
That might explain the high amount of less stable species. On M-S, the presence of a third type, more
stable, of sulfur species, which is not released during TPR, cannot be ruled out. The chlorinated
catalyst, M-Cl, was the only catalyst that released more sulfur during the TPR than during the
reduction at 500C. The amount of less stable species was particularly low, indicating that chlorine
prevented the formation of this kind of species. The release during TPR was interestingly very similar
to the release recorded for M-Pt (Figure 30). This feature suggests that chlorine influenced neither
the formation nor the stability of such sulfur species.
Sulfur release (mol/gcat)
reduction at 500 C
TPR up to 800 C
Total
1400
1200
1000
800
600
400
200
0
M-Pt
M-S
M-Cl
Figure 29: Stepwise sulfur release showing the distribution and stability of species formed during
sulfation of chlorinated (M-Cl), pre-sulfated (M-S) and unmodified Pt/Al2O3 (M-Pt).
44
__________________________________________ 7.4. Oxidation ability of an acidity-enhanced DOC
120
100
80
900
M-Pt
800
M-Cl
700
M-S
600
T cat
500
60
400
40
300
20
200
100
-20
2000
4000
6000
8000
Time (s)
10000
H2S outlet concentration (ppm)
140
0
12000
Figure 30: H2S release during Temperature-Programmed Reduction up to 800C (TPR) in 2% H2 of MCl, M-S and M-Pt.
45
46
Chapter 8
8. Intra-catalyst insight of SCR with NH3 on Cu-BEA
8.1. Experimental measurements

NH3-SCR was carried out on a Cu-Beta zeolite prepared at Chalmers containing 4.3 wt.% copper. The
catalyst was characterized following a 4-step protocol developed by Cummins to study several
reactions occurring in operating conditions [111]:
1. NO oxidation
2. NO reduction by NH3
3. NH3 oxidation
2 NO + O2 2 NO2
4 NO + 4 NH3 + O2 4 N2 + 6 H2O
4 NH3 + 3 O2 2 N2 + 6 H2O
NH3 storage in presence and absence of NO was also studied and the storage capacities of the
catalyst were determined. The SpaciMS analysis gave the spatial distribution for all of these features.
The 4-step protocol consisted in changing the gas composition of the feed by turning NH 3 and/or NO
supply on and off in order to investigate NH3 storage and the reactions mentioned above. The
description of the protocol is given in Table 1.
Table 1: Feed composition during each step of the 4-step protocol (total flow 510 sccm).
In order to simulate real conditions, the gas feed contained 10 % O2, 5 % H2O throughout the whole
protocol. Krypton was added for MS signal analysis and correction. The water vapor was generated
by evaporation of liquid water heated at a fixed temperature in a hot bath. The vapor was then
carried by a flow of Ar. In the first step, 205 ppm NO was added to the base stream to clean the
surface of adsorbed NH3 as well as studying the NO oxidation. In step 2, NO was kept in the flow and
183 ppm NH3 was added in order to perform the SCR reaction described in chapter 4 as well as
studying the dynamic of NH3 storage in operating conditions. When the SCR reaction reached steady
state, NO was turned off, marking the start of step 3. In this step, the storage of NH3 in the absence
of NO was studied and the reaction of NH3 oxidation was observed. Finally in step 4, NH3 was
switched off and NO was simultaneously started in order to observe the reaction occurring between
the flowing NO and NH3 stored on the surface. The total storage capacity of the catalyst could be
determined by the consumption of NO and the release of NH3 in this last step. After the fourth step,
NO and NH3 were turned off in order to get a baseline of concentrations for subsequent calculations.
The capillary was then translated to the next measurement position and the 4-step protocol could be
47
Chapter 8: Intra-catalyst insight of SCR with NH3 on Cu-BEA __________________________________

run at this new location. Before moving on to the next position, the protocol was repeated twice and
the results from the second cycle were analyzed.
[NO]:
[NH3]:
Concentration (ppm)
200
Clean
Standard SCR
NH3 Saturate
Clean
205 ppm
0
205 ppm
183 ppm
0
183 ppm
205 ppm
0
Step 1
Step 2
Step 3
NO Oxidation
NOx-free NH3 Oxidation
NO
N2
150
100
DC
50
NO2
Step 4
NO Oxidation
UC
TC
Possible parasitic NH3 Oxidation
NH3
Figure 31: Schematic experimental concentration profiles during the 4-step protocol and
representation of storage capacity integration. DC is the dynamic storage capacity, observed during
SCR, UC refers to the unused capacity, measured during NH3 oxidation, and TC stands for total
capacity.
8.1.1. NO and NH3 oxidation
Cu-zeolite catalysts are active for oxidation of NO and NH3, which was verified in step 1 and 3 of our
protocol (Figure 31), respectively. Ammonia oxidation reaction can interfere during SCR by
consuming ammonia and therefore needs to be controlled in order to adjust the injection of urea
(precursor of NH3) necessary to reduce NOx, in practical application. NO oxidation is important, since
it produces NO2 and therefore changes the rate and mechanism of SCR reaction from Standard SCR
to Fast SCR or NO2-SCR. It has been proposed that NO oxidation was in fact the rate determining
step (RDS) of the SCR reaction [50]. However, other studies have suggested that this is not the case.
For example H2O has a detrimental effect on NO oxidation, but only minor impact on the SCR rate
[112], which suggests that NO oxidation is not the RDS. With the use of spatially resolved
measurements, the extent of reaction along the monolithic catalyst was evaluated (Figure 32).
48
_______________________________________________________ 8.1. Experimental measurements

25
120
15
NH3 conversion (%)
20
325C
325C
10
400C
400C
5
0
(b)
100
80
325C
325C
60
400C
400C
40
20
0
1/8 1/4 3/8 1/2 5/8 3/4 7/8 1

Axial position (Fraction of catalyst length)
1/8 1/4 3/8 1/2 5/8 3/4 7/8 1

Figure 32: Performance of the Cu-BEA SCR catalyst at 325C and 400C for (a) NO oxidation and
(b) NH3 oxidation, presented as the conversion throughout the catalyst (0=inlet, 1=outlet).
The catalyst is more active at higher temperature for both oxidation reactions and did not show any
activity at 200C. NO oxidation increased in the front half of the catalyst (Figure 32 a) and then
stabilized due to the inhibition effect and the dissociation of NO2. The NO conversion reached at
400C was 23 %. Higher activity was noted for NH3 oxidation as seen in Figure 32 b. At 325C, NH3
conversion increased almost linearly from the inlet to the outlet where it exhibited a maximum at
45 %. At 400C a sharp conversion increase was observed in the front part leading to the total NH3
conversion by the half catalyst position.
8.1.2. Selective catalytic reduction of NO by NH3
Contrary to NH3 and NO oxidation, the SCR occurred at all three temperatures. SCR was studied
during step 2 and was evidenced at the start of step 4. The amount of catalyst used to reach 100 %
conversion, the NH3:NO stoichiometry and the N2O production were especially considered.
Steady state concentrations

SCR (ppm)
NO conversion (%)
(a)
200
180
160
140
120
100
80
60
40
20
0
-20
325C
NH
NH3
3
NO
NO
N
N2
2
N
N2O
2O
1/8 1/4 3/8 1/2 5/8 3/4 7/8

Figure 33: Steady state evolution of NO, N2, NH3 and N2O concentrations during SCR at 325C
(205ppm NO + 183ppm NH3 + 10% O2 + 5% H2O)
49

Since the feed contained less NH3 than NO, the conversion of NO was normalized to the inlet NH3
concentration as:
(Eq. 8.1)
This normalized NO conversion can be directly compared to NH3 conversion to assess the
stoichiometry of the reaction. In Figure 34, it can be seen that the NH3 conversion and the
normalized NO conversion profiles are very close, which means that as much ammonia as nitric oxide
disappears and, consequently the stoichiometry of the SCR reaction over our Cu-BEA catalyst is
NH3:NO=1. Even at 400C (Figure 34 c), where NH3 oxidation is fast, over-consumption of ammonia
was not observed. This is likely due to that the SCR rate was significantly higher than NH3 oxidation
rate at this temperature. In these conditions, 100 % conversion was reached at the outlet at all three
temperatures and effluent-based analysis would not be able to display the rate differences.
However, the SpaciMS technique revealed important information about the kinetics of SCR reaction.
Indeed, Figure 34 shows that the SCR zone, defined as the length of catalyst necessary to convert
100 % of NO, becomes shorter and condensed to the front as the operating temperature increases
from 200C to 400C. These kinetics data, acquired in realistic conditions, are valuable for the
development of a kinetic model.
50
Normalized NO conversion (%)
_______________________________________________________ 8.1. Experimental measurements

120
90%
100
SCR zone
80
60
40
NH3
NH3
20
NO
NO
0
0
1/8
1/4
3/8
1/2
5/8
3/4
7/8
120
90%
(b) 325C
100
80
60
40
20
NH3
NH3
SCR zone
NO
NO
0
0
(a) 200C
120
1/8
1/4
3/8
1/2
5/8
3/4
7/8
90%
(c) 400C
100
80
60
40
NH3
NH3
20
NO
NO
SCR zone
0
0
1/8
1/4
3/8
1/2
5/8
3/4
7/8
Figure 34: NO and NH3 conversion during SCR (step 2) with highlight on the SCR zone at (a) 200C,
(b) 325C and (c) 400C.
51

The major products of SCR are N2, which is wanted and N2O, which must be as low as possible due to
its high global warming potential (~300 times higher than CO2). The production of N2O has been
found non-monotonic with respect to the temperature. Indeed, the outlet concentration of N2O was
higher at 200C and 400C than at 325C (Figure 35). This result suggests the possibility of multiple
mechanisms to form N2O in order to describe N2O at both high and low temperature. At low
temperature, decomposition of ammonium nitrate has been proposed to be the main source of N2O
[46, 51, 113]. At high temperature, on Fe-zeolite, N2O formation has been correlated to NO2 [46, 50,
114]. On Cu-zeolite, copper dimer sites have been found responsible for N2O production at high
temperature [113].
N2O concetration (ppm)
20
15
10
400C
400C
200C
200
C
325C
325
C
0
0
1/8
1/4
3/8
1/2
5/8
3/4
7/8
Figure 35: Steady state N2O concentration profiles during SCR step 2 at 200C, 325C and 400C.
8.1.3. NH3 storage capacity
The NH3 storage capacity of the catalyst was studied in presence and absence of NO. In step 2, NO,
NH3 and oxygen were flowing and SCR occurred. In parallel, NH3 was stored on the catalyst, which
defined the dynamic storage capacity (DC) of the catalyst at a particular position. The unused
capacity (UC) was obtained by integrating the NH3 storage during step 3, without NO. The total
storage capacity (TC) could be determined in two ways: by addition of the dynamic and the unused
capacity DC+UC or by direct measurement in step 4, when the stored NH3 was either released or
converted into N2 and N2O.
The integrated area representing the storage capacities are shown in Figure 31. The total storage
capacity increases with decreasing the temperature. At 400C, no storage could have been
measured. The storage distribution between dynamic and unused capacity depends, however, on the
axial position. TC was inferred from step 4 where stored NH3 was quickly released or quickly reacted
under the action of NO to form N2. The consumption of NO and the NH3 release in step 4 were used
to determine the total storage capacity. DC, as indicated in Figure 31, was calculated in step 2 by
integration of NO and NH3 signals, relative to steady state level. The storage (DC and TC) at 200C and
52
_______________________________________________________ 8.2. Kinetic modeling simulations

325C is represented in Figure 36. It can be noticed that in the front of the catalyst, DC and TC are
very similar, which reveals that all storage sites were filled during SCR. It also means that the SCR
reaction does not affect the ammonia storage and both processes occur simultaneously, which is an
important information regarding the dynamic behavior of the SCR catalyst. The two curves separate
at a location that corresponds to the 90 % NO conversion zone, when DC remains almost constant
while TC keeps increasing. Indeed, downstream of the SCR zone, no ammonia remains in the gas
phase and no ammonia storage can be detected. This is in agreement with our experimental results
displaying constant DC after the SCR zone.
20
70
(a)
200C
90%
90%
50
NH3 storage (mol)
NH3 storage (mol)
60
40
30
20
DC
10
0
0
TC
15
10
DC
TC
5
0
0
1/8 1/4 3/8 1/2 5/8 3/4 7/8 1

Axial position (fraction of catalyst length)
(b) 325C
1/8 1/4 3/8 1/2 5/8 3/4 7/8 1

Axial position (fraction of catalyst length)
Figure 36: Ammonia dynamic storage capacity measured during SCR step (DC) and total storage
capacity measured in step 4 (TC). (a) at 200C and (b) at 325C. The respective position of 90 % NO
conversion is displayed on the figures.
8.2. Kinetic modeling simulations
Kinetic modeling applies kinetic theory and chemical reaction engineering to evaluate the rate of a
reaction in desired operating conditions. Kinetic modeling can be used to find empirically reaction
kinetic parameters; it can provide information about reaction mechanism and can help to determine
the rate determining step of a complex reaction, which is especially applicable to catalytic reactions.
Many kinetic models assuming different mechanisms have been developed to describe NH3-SCR
[115-121]. Experimental data used to fit the kinetic parameters are usually based on effluent gas
composition over a wide range of initial conditions and temperature. We developed a kinetic model
based on steady state intra-monolith concentration profiles. The kinetic parameters were tuned in
order to simulate axial concentration evolutions that match the experimental measurements.
The simulations were carried out with the commercial software AVL BOOST Aftertreatment in
connection with a user-defined Fortran script, in which reaction rate equations and parameters were
defined. The system was reduced to one channel, assuming all channels identical, and discretized in a
series of 15 elements, in which the mass balance (Eq. 8.1) of each species was solved
(Eq. 8.2)
53

where wk,g is the mass fraction of species k, Deff is an effective diffusion coefficient, ri(ckL, Ts)
represents the molar reaction rate of the surface reaction i, i,k is the stoichiometric coefficient of the
species k in the reaction i and g is the volume fraction of the gas phase in the total volume.
Considering the low working temperature range, homogeneous gas phase reactions were negligible
and not taken into account in the model. The transport from the gas phase to the washcoat was
described by the film transport model. The Arrhenius equation (Eq. 8.3) was applied to express the
temperature-dependence of the rate constant of the considered reactions.
(Eq. 8.3)
The kinetic parameters A (pre-exponential factor) and EA (activation energy) were tuned to obtain
the best fit to experimental intra-catalyst concentration profiles based on steady state values. Before
fitting parameters related to NH3-SCR (step 2 in the protocol), NO oxidation, NH3
adsorption/desorption and NH3 oxidation were independently studied. NO and NO2 concentrations
measured in step 1 were used to determine adequate parameters for NO oxidation. NO oxidation
was simulated in the model as a global gas phase reaction (r3 in Table 2), which do not imply an
actual gas phase mechanism but is a simplification in the model. In the same way, NH3 and N2 steady
state profiles were used to fit ammonia oxidation parameters. This reaction was represented in the
model by a reaction between O2 in the gas phase and adsorbed NH3 (r2 in Table 2), which again is
only a simplification in the global kinetic model. The adsorption and desorption process (r1 in Table
2) was studied in an independent experiment consisting in adsorption of NH3 at 40C followed by a
temperature ramping up to 600C to release the stored ammonia. The site density was tuned to fit
the ammonia release during the ramping phase. The coverage-dependence of the activation energy
of ammonia desorption was taken into account. It was thus expressed using a Temkin isotherm,
which considers the adsorbate - adsorbate interactions: (Eq. 8.4):
ENH3,des = E0NH3,des (1- NH3)
(Eq. 8.4)
where is a constant, NH3 the coverage and E0NH3,des the activation energy for 0 coverage. This
expression was employed in many kinetic models [118-120] and agrees with Wilken et al. [122]
micro-calorimetry results. The coverage dependent heat of adsorption was determined in microcalorimetry experiments (112.75*(1-0.39*NH3-S1)). The SCR parameters as well as the N2O formation
parameters were simultaneously tuned to fit the observed steady state concentrations at the end of
step 2. Two mechanisms for N2O production were included in the model to describe the higher yields
obtained at 200C and 400C compared to 325C, which cannot be described by one single Arrhenius
reaction. The low temperature route proposed is a two-step process starting with the formation of
an intermediate NH3-NO adsorbed species that decomposes in a subsequent step into N2O, N2 and
H2O according to reaction 6 and 7 in Table 2.
54

Table 2: Reaction steps and rate expressions included in the kinetic model.
Nb
Reactions
Rate expression
NH3 + S1 NH3-S1
r1 = k1f * v * [NH3] - k1b * S1-NH3
4NH3-S1 + 3O2 2N2 + 6H2O + 4 S1
r2 = k2 * S1-NH3 * [O2] ^ 1
NO + 0.5 O2 NO2
r3 = k3f * [NO] * [O2] ^ 0.5 k3b * [NO2]
4NH3-S1 + 4NO + O2 4N2 + 6H2O + 4S1
r4 = k4 * S1-NH3 * [NO] * [O2] ^ 2
2NH3-S1 + 2NO + O2 N2O + N2 + 3H2O + 2S1
r5 = k5 * S1-NH3 * [NO] * [O2]
NH3-S1 + NO (S1)-NH3_NO
r6 = k6f * S1-NH3* [NO] * k6b * S1-NH3_NO
2 S1-NH3_NO + O2 N2O + N2 + 3 H2O + 2S1
r7 = k7 * S1-NH3_NO * [O2] * v
The reaction order of oxygen 1 and 2 were determined in experiments with varying oxygen
concentrations to 0.6 and 0.5, respectively.
Local concentration (ppm)
250
250
(a) 325C
(b) 400C
200
200
150
150
NO model
model
NO22 model
model
NO
NO exp
exp
NO22 exp
NO
100
50
NO model
model
NO
NO22 model
model
NO exp
exp
NO
NO22 exp
100
50
0
0
1/4
1/2
3/4
1
Axial position (fraction of total length)
1/4
1/2
3/4
1
Axial position (fraction of total length)
Figure 37: Experimental and simulated steady state NO and NO2 concentration profiles during NO
oxidation (step 1). (a) at 325C and (b) at 400C.
To reduce the number of parameters to be fitted, some kinetic parameters were set to literature
values. That was the case for the activation energy of NO oxidation reaction, which was set to
48 kJ/mol according to Olsson et al. [117]. Only the pre-exponential factor was therefore tuned for
that reaction, which led to an adequate curve fit both at 325C and 400C, as seen in Figure 37.
However, the activation energy for ammonia oxidation reported by Sjvall et al. [118] (162.5 kJ/mol)
was too high to fit with a good agreement our spatial experimental data. Table 3 summarizes the
reaction steps and their respective parameters used in the model. NH3 adsorption as well as NH3-NO
complex formation are non-activated reversible steps. The model predicted the experimentally
observed SCR stoichiometry, in other words, the absence of ammonia over-consumption. NH3 overconsumption was observed over Fe-zeolite catalysts [50, 116, 123] and is a crucial parameter in
dosing NH3 injection.
55
Table 3: Summary of kinetic parameters used in the kinetic model.
Pre-exponential factor a
Reaction
Activation energy (kJ/mol)
NH3 adsorption
9279
NH3 desorption
6.76e8
112.75
NH3 oxidation
8.00e8
130.0
NO oxidation
1.10e5
48.0
SCR
2.50e8
60.0
N2O formation high temperature
1.90e16
160.0
65
8.48e10
65.0
7.33e18
95.0
N2O formation low temperature

complex formation
complex dissociation
complex decomposition
a
The unit of the reaction rate is kmol/(m3s) and the gas phase concentration are dimensionless.
The model validation was based on the transient response to gas composition changes when the
whole 4-step experiment was simulated. The model describes the experimental features adequately.
One example of full experiment simulation, at 200C and 1/4 of catalyst is shown below (Figure 38).
70
NH3
exp
NH
3 exp
NH3
sim
NH
3 sim
200
(a)
N2O concentration (ppm)
NH3 concentration (ppm)
250
150
100
50
0
0
-50
1000
2000
3000
4000
5000
60
N2O exp
exp
N2O
50
N2O sim
sim
N2O
40
30
20
10
0
-10 0
-20
Time (s)
56
(b)
1000
2000
3000
Time (s)
4000
5000

250
(c)
200
NO exp
NO sim
N2 concentration (ppm)
NO concentration (ppm)
250
150
100
50
0
0
-50
1000
2000
3000
4000
(d)
200
N2
N2sim
sim
150
100
50
0
5000
0
-50
Time (s)
N2
N2exp
exp
1000
2000
3000
4000
Time (s)
Figure 38: Comparison of simulated and experimental concentration profiles for (a) NH3 , (b) N2O ,
(c) NO and (d) N2 at 200C and 1/4 monolith length from the inlet.
At this position, the steady state concentration of all components and in all the four steps is in line
with the experimental data. The transient evolution in step 2 is captured well by the model for all
components. The ammonia storage indicated in step 3 by the time to reach ammonia signal steady
state is accurately modeled. However, the model predicts too high NO consumption and N2 and N2O
production in step 4. These conclusions differ depending on the considered temperature and
location but the simulated results agree well with the experimental data, in general.
57
5000
58
Chapter 9
9. Concluding remarks
In this thesis, two major aspects of diesel emission control were addressed. A typical diesel oxidation
catalyst and a lean NOx reduction catalyst were studied during flow reactor experiments. The activity
response of the model DOC to aging treatment and sulfur poisoning was specifically investigated. The
performance of the SCR catalyst was evaluated with intra-monolith spatially resolved measurement
and was used to develop a kinetic model.
Aging an alumina-supported Pt catalyst at high temperature in various atmospheres revealed the
faster sintering of Pt particles in presence of SO2. In contrast, H2 seemed to prevent sintering up to
700C. In oxidative and inert flow, the dispersion decreases with aging temperature with similar
linear trend. Long aging at lower temperature (250C) in presence of water, oxygen and SO2, followed
by reduction in H2, led to significant dispersion loss and simultaneous activity improvement for both
NO and C3H6 oxidation. The role of surface sulfur stored during the aging step was investigated in
order to attempt to explain the resulting catalytic activity. Two types of sulfur species characterized
by different stability and reducibility were identified on the catalyst after aging. The less stable
species could be removed by reduction treatment at 500C but had a major detrimental effect on
oxidation performance. Nevertheless, our results indicate that these species are mobile and can be
replaced on the active sites by NO provided that the operating temperature is high enough. The
second type of sulfur species, more stable, was decomposed at higher temperature to form H2S and
could likely be assigned to sulfates stored on the alumina support. The effect of the stable sulfur
adsorbates is minor and affects only the activity during extinction suggesting their migration from the
support to the active sites at elevated temperature. Since the presence of adsorbed sulfur species
causes an activity loss, our results suggest that the beneficial effect of SO2 aging is due to sulfurassisted platinum sintering. It would be interesting to reproduce similar SO2 aging study on Pd and
bimetallic Pd-Pt-based catalysts since commercial DOCs contain also palladium. Bimetallic Pd-Pt
catalysts are known to be more resistant to sintering and it would be interesting to see if SO2
treatment is able to promote the particle agglomeration on this type of catalysts. Finally, the SO2
sensitivity of Pd catalysts could lead to differences in SO2 storage and release as well as in
performance.
The deactivation of DOC towards NO oxidation has been evidenced by the decrease of activity with
time on stream and the lower activity during extinction yielding an inverse hysteresis. This
phenomenon has been attributed to the reversible oxidation of platinum. The tendency of platinum
to get oxidized varies with the nature of the support material. Platinum supported on acidic support
are more resistant to oxidation, which motivated the study of acidity-enhanced Pt catalyst for DOC
application.
The deposition of sulfate and chlorine in acidic solution during the preparation of Pt/Al2O3 was the
employed method to increase the catalyst support acidity. The initial consequence of such treatment
is a lower NO oxidation activity during reaction ignition but the catalysts do not exhibit deactivation
during extinction. As unmodified Pt/Al2O3, chlorinated and sulfated catalysts were subjected to SO2
59
Chapter 9: Concluding remarks _________________________________________________________

poisoning to study the sulfur storage, the platinum sintering and the subsequent oxidation activity.
Lower sulfur release was noted upon reduction of acidity-enhanced catalysts, suggesting that SO2
storage is reduced on acidic supports. After the long SO2 step, deactivation appeared during the
cooling phase of NO oxidation. This indicates the loss of the acidic character of the catalysts, likely
due to the loss of chlorine and sulfate, respectively. Further characterization of the fresh and used
catalysts would be valuable. The determination of the initial and final chlorine/sulfate content as well
as the trapping of released species during the flow reactor tests would enable one to correlate the
chlorine/sulfate content to catalyst acidity and oxidation performance. It would also be good to
evaluate Pt catalysts supported on various materials with acidic nature such as silica-alumina.
The characterization of a Cu-BEA catalyst for NH3-SCR, following a four-step protocol, with intramonolith spatial measurements disclosed the utilization of the catalyst for reaction and storage. NO
oxidation and NH3 oxidation were noticed at 325 and 400C. In particular 100 % ammonia was
oxidized at 400C at the catalyst exit. However, during SCR, equal amount of ammonia and nitric
oxide was consumed at any axial position in the catalyst, indicating no ammonia over-consumption
and SCR stoichiometry of NO:NH3= 1. Selective reduction of NO by NH3 in lean condition is fast and
uses only a limited zone in the front of the catalyst to be complete. The required zone to reach full
conversion is condensed to the inlet as the operating temperature increases from 200 to 400C. Thus
at 400C ammonia is already consumed by SCR reaction before direct ammonia oxidation becomes
significant. The analysis of ammonia storage revealed that all sites within the SCR zone are filled
during SCR conditions (NO, NH3 and O2). A kinetic model was developed based on spatially resolved
steady state concentrations. The advantage of SpaciMS measurement is the possibility to collect
interpretable data in realistic gas composition and temperature conditions. Indeed, conversion
differences inside the monolith are observed in different conditions, even when outlet conversion is
100 %. A global kinetic model was developed and parameters were tuned to fit axial concentration
profiles obtained by spatially-resolved measurement. The model used a global kinetic expression for
NO oxidation and described the SCR kinetics as a reaction between adsorbed ammonia and gas
phase NO. The model includes the formation of N2O as a side product of SCR. Two mechanisms
accounting for low and high temperature N2O formation were proposed. The model agreed well with
the experiment as it reproduced well the transient response to gas composition changes during the
experimental protocol. In the future, the study of the Fast SCR reaction to develop further the kinetic
model could be carried out. This implies overcoming the experimental difficulties of working with
SpaciMS at high NO2 concentration, due to the stickiness of this gas in the capillary probe.
60
List of abbreviations
DOC
Diesel Oxidation Catalyst
BET
Brunauer Emmett Teller
SCR
Selective Catalytic Reduction
HC
Hydrocarbon
TWC
Three-Way Catalyst
LNT
Lean NO x-Trap
ASC
Ammonia Slip Catalyst
DPF
Diesel Particulate Filter
PM
Particulate Matter
TEM
Transmission Electron Microscopy
SpaciMS
Spatially-resolved Capillary Inlet Mass Spectrometer
XPS
X-ray Photoelectron Spectroscopy
TOR
Turn Over Rate
EXAFS
Extended X-ray Absorption Fine Structure
DRIFT
Diffuse Reflectance Infrared Fourier Transform
TPD
Temperature Programmed Desorption
NSR
NOx Storage Reduction
HAADF
High Angle Annular Dark Field
STEM
Scanning Transmission Electron Microscopy
MS
Mass Spectrometer
FTIR
Fourier Transform Infra Red
CFD
Computational Fluid Dynamics
ORNL
Oak Ridge National Laboratory
TPR
Temperature-Programmed Reduction
IR
Infra Red
DC
Dynamic Capacity
UC
Unused Capacity
TC
Total Capacity
RDS
Rate Determining Step
61
62
Acknowledgments
The Swedish Foundation for Strategic Research is gratefully acknowledged for funding this project.
This work was carried out at the Competence Centre for Catalysis (KCK), Chalmers University of
Technology. KCK is financially supported by Chalmers University of Technology, the Swedish Energy
Agency and the member companies: AB Volvo, Volvo Car Corporation, Scania CV AB, Haldor Topsoe
A/S and ECAPS AB.
I would like to thank Professor Bengt Andersson and Professor Louise Olsson for giving me the
opportunity to do my PhD at Chalmers.
I would like to thank again Louise, as my supervisor, for her guidance, her support and positive
attitude, for giving me the chance to visit ORNL and giving me freedom and independence in my
research.
Bill Partridge and Jae-Soon Choi, my co-workers and friends from ORNL, for making my visit in
Tennessee so enjoyable, for the good work we achieved and all the things I learned with you.
I would also like to thank all the colleagues who helped me through the years. I want to thank
especially my friends/colleagues Slavi and Azis for their support and assistance and for all the
discussions and good times.
Finally, I thank my family and friends who did not see me much in the past five years but kept
supporting me.
63
64
Bibliographic references
[1]
S.G. Masters, D. Chadwick, Applied Catalysis B: Environmental 23 (1999) 235-246.
[2]
F.C. Meunier, J.P. Breen, V. Zuzaniuk, M. Olsson, J.R.H. Ross, Journal of Catalysis 187
(1999) 493-505.
[3]
R.M. Heck, Catalysis Today 53 (1999) 519-523.
[4]
J.S. Choi, W.P. Partridge, C.S. Daw, Applied Catalysis B: Environmental 77 (2007) 145156.
[5]
J. Desprs, M. Elsener, M. Koebel, O. Krcher, B. Schnyder, A. Wokaun, Applied
Catalysis B: Environmental 50 (2004) 73-82.
[6]
S.S. Mulla, N. Chen, L. Cumaranatunge, G.E. Blau, D.Y. Zemlyanov, W.N. Delgass, W.S.
Epling, F.H. Ribeiro, Journal of Catalysis 241 (2006) 389-399.
[7]
S.S. Mulla, N. Chen, W.N. Delgass, W.S. Epling, F.H. Ribeiro, Catalysis Letters 100
(2005) 267-270.
[8]
D. Bhatia, R.W. McCabe, M.P. Harold, V. Balakotaiah, Journal of Catalysis 266 (2009)
106-119.
[9]
L. Olsson, H. Persson, E. Fridell, M. Skoglundh, B. Andersson, Journal of Physical
Chemistry B 105 (2001) 6895-6906.
[10]
L. Olsson, B. Westerberg, H. Persson, E. Fridell, M. Skoglundh, B. Andersson, Journal of
Physical Chemistry B 103 (1999) 10433-10439.
[11]
M. Crocoll, S. Kureti, W. Weisweiler, Journal of Catalysis 229 (2005) 480-489.
[12]
J.H. Lee, H.H. Kung, Catalysis Letters 51 (1998) 1-4.
[13]
E. Xue, K. Seshan, J.R.H. Ross, Applied Catalysis B: Environmental 11 (1996) 65-79.
[14]
L. Olsson, E. Fridell, Journal of Catalysis 210 (2002) 340-353.
[15]
S. Benard, L. Retailleau, F. Gaillard, P. Vernoux, A. Giroir-Fendler, Applied Catalysis B:
Environmental 55 (2005) 11-21.
[16]
P.J. Schmitz, R.J. Kudla, A.R. Drews, A.E. Chen, C.K. Lowe-Ma, R.W. McCabe, W.F.
Schneider, C.T. Goralski Jr, Applied Catalysis B: Environmental 67 (2006) 246-256.
[17]
P. Denton, A. Giroir-Fendler, H. Praliaud, M. Primet, Journal of Catalysis 189 (2000)
410-420.
[18]
A. Boubnov, S. Dahl, E. Johnson, A.P. Molina, S.B. Simonsen, F.M. Cano, S. Helveg, L.J.
Lemus-Yegres, J.-D. Grunwaldt, Applied Catalysis B: Environmental 126 (2012) 315-325.
[19]
H.F. Wang, Y.L. Guo, G. Lu, P. Hu, Journal of Physical Chemistry C 113 (2009) 1874618752.
[20]
S.K. Matam, E.V. Kondratenko, M.H. Aguirre, P. Hug, D. Rentsch, A. Winkler, A.
Weidenkaff, D. Ferri, Applied Catalysis B: Environmental 129 (2013) 214-224.
[21]
K. Hauff, U. Tuttlies, G. Eigenberger, U. Nieken, Applied Catalysis B: Environmental
123-124 (2012) 107-116.
[22]
W. Hauptmann, M. Votsmeier, J. Gieshoff, A. Drochner, H. Vogel, Applied Catalysis B:
[23]
O. Krcher, M. Widmer, M. Elsener, D. Rothe, Industrial and Engineering Chemistry
Research 48 (2009) 9847-9857.
[24]
K. Mudiyanselage, Y. Cheol-Woo, J. Szanyi, Journal of Physical Chemistry C 113 (2009)
5766-5776.
[25]
S. Ovesson, B.I. Lundqvist, W.F. Schneider, A. Bogicevic, Physical Review B (Condensed
Matter and Materials Physics) 71 (2005) 115406-115401.
[26]
C.B. Wang, C.T. Yeh, Journal of Catalysis 178 (1998) 450-456.
[27]
C.B. Wang, C.T. Yeh, Applied Catalysis A: General 209 (2001) 1-9.
[28]
P. Briot, A. Auroux, D. Jones, M. Primet, Applied Catalysis 59 (1990) 141-152.
[29]
R.W. McCabe, C. Wong, H.S. Woo, Journal of Catalysis 114 (1988) 354-367.
[30]
H. Oh, J. Luo, W.S. Epling, Catalysis Letters 141 (2011) 1746-1751.
65
[31]
I.V. Yentekakis, V. Tellou, G. Botzolaki, I.A. Rapakousios, Applied Catalysis B:
[32]
K. Irani, W.S. Epling, R. Blint, Applied Catalysis B: Environmental 92 (2009) 422-428.
[33]
L. Olsson, M. Abul-Milh, H. Karlsson, E. Jobson, P. Thormhlen, A. Hinz, Topics in
Catalysis 30-31 (2004) 85-90.
[34]
M.F. Irfan, J.H. Goo, S.D. Kim, Environmentalist 31 (2011) 4-10.
[35]
M.F. Irfan, J.H. Goo, S.D. Kim, S.C. Hong, Chemosphere 66 (2007) 54-59.
[36]
H.C. Yao, H.K. Stepien, H.S. Gandhi, Journal of Catalysis 67 (1981) 231-236.
[37]
A. Baylet, C. Capdeillayre, L. Retailleau, J.L. Valverde, P. Vernoux, A. Giroir-Fendler,
Applied Catalysis B: Environmental 102 (2011) 180-189.
[38]
S. Benard, A. Baylet, P. Vernoux, J.L. Valverde, A. Giroir-Fendler, Catalysis
Communications 36 (2013) 63-66.
[39]
L.M. Carballo, E.E. Wolf, Journal of Catalysis 53 (1978) 366-373.
[40]
M. Haneda, T. Watanabe, N. Kamiuchi, M. Ozawa, Applied Catalysis B: Environmental
142143 (2013) 8-14.
[41]
P. Marcot, A. Fakche, B. Kellali, G. Mabilon, P. Prigent, J. Barbier, Applied Catalysis B:
[42]
O. Shakir, A. Yezerets, N.W. Currier, W.S. Epling, Applied Catalysis A: General 365
(2009) 301-308.
[43]
A. Russell, C. Henry, N.W. Currier, A. Yezerets, W.S. Epling, Applied Catalysis A: General
397 (2011) 272-284.
[44]
J.-H. Lee, H. Kung, Catalysis Letters 51 (1998) 1-4.
[45]
A. Abedi, R. Hayes, M. Votsmeier, W.S. Epling, Catalysis Letters 142 (2012) 930-935.
[46]
M. Devadas, O. Krcher, M. Elsener, A. Wokaun, N. Sger, M. Pfeifer, Y. Demel, L.
Mussmann, Applied Catalysis B: Environmental 67 (2006) 187-196.
[47]
M. Koebel, G. Madia, M. Elsener, Catalysis Today 73 (2002) 239-247.
[48]
I. Nova, C. Ciardelli, E. Tronconi, D. Chatterjee, B. Bandl-Konrad, Catalysis Today 114
(2006) 3-12.
[49]
M. Koebel, M. Elsener, M. Kleemann, Catalysis Today 59 (2000) 335-345.
[50]
J.-Y. Luo, X. Hou, P. Wijayakoon, S.J. Schmieg, W. Li, W.S. Epling, Applied Catalysis B:
Environmental 102 (2011) 110-119.
[51]
A. Grossale, I. Nova, E. Tronconi, D. Chatterjee, M. Weibel, Journal of Catalysis 256
(2008) 312-322.
[52]
T. Komatsu, M. Nunokawa, I.S. Moon, T. Takahara, S. Namba, T. Yashima, Journal of
Catalysis 148 (1994) 427-437.
[53]
K. Rahkamaa-Tolonen, T. Maunula, M. Lomma, M. Huuhtanen, R.L. Keiski, Catalysis
Today 100 (2005) 217-222.
[54]
S.J. Schmieg, S.H. Oh, C.H. Kim, D.B. Brown, J.H. Lee, C.H.F. Peden, D.H. Kim, Catalysis
Today 184 (2012) 252-261.
[55]
J.H. Kwak, R.G. Tonkyn, D.H. Kim, J. Szanyi, C.H.F. Peden, Journal of Catalysis 275
(2010) 187-190.
[56]
J.H. Kwak, D. Tran, S.D. Burton, J. Szanyi, J.H. Lee, C.H.F. Peden, Journal of Catalysis
287 (2012) 203-209.
[57]
H. Sjvall, L. Olsson, E. Fridell, R.J. Blint, Applied Catalysis B: Environmental 64 (2006)
180-188.
[58]
F. Cabello Galisteo, R. Mariscal, M. Lpez Granados, M.D. Zafra Poves, J.L.G. Fierro, V.
Krger, R.L. Keiski, Applied Catalysis B: Environmental 72 (2007) 272-281.
[59]
J. Pazmio, S. Mulla, W.N. Delgass, F.H. Ribeiro, J.T. Miller, Kinetic studies on the
stability of Pt for NO oxidation: Effect of sulfur and long term aging, 2009.
[60]
C.R. Apesteguia, T.F. Garetto, C.E. Brema, J.M. Parera, Applied Catalysis 10 (1984) 291295.
[61]
G. Corro, R. Montiel, C. Vzquez L, Catalysis Communications 3 (2002) 533-539.
66
[62]
Y. Yazawa, N. Takagi, H. Yoshida, S.-i. Komai, A. Satsuma, T. Tanaka, S. Yoshida, T.
Hattori, Applied Catalysis A: General 233 (2002) 103-112.
[63]
K. Foger, J.R. Anderson, Applications of Surface Science 2 (1979) 335-351.
[64]
I. Guo, T.T. Yu, S.E. Wanke, Changes in Pt crystallite sizes as a result of treating
Pt/Al2O3 catalysts in different atmospheres, 1988, pp. 21-32.
[65]
K. Foger, Catalysis : Science and technology, Springer, Verlag Berlin Heidelberg New
York Tokyo, 1984.
[66]
J.S. Choi, W.P. Partridge, C.S. Daw, Applied Catalysis A: General 293 (2005) 24-40.
[67]
J.-S. Choi, W.P. Partridge, W.S. Epling, N.W. Currier, T.M. Yonushonis, Catalysis Today
114 (2006) 102-111.
[68]
W.P. Partridge, J.S. Choi, Applied Catalysis B: Environmental 91 (2009) 144-151.
[69]
W.P. Partridge, T.J. Toops, J.B. Green, T.R. Armstrong, Journal of Power Sources 160
(2006) 454-461.
[70]
J. S, D.L.A. Fernandes, F. Aiouache, A. Goguet, C. Hardacre, D. Lundie, W. Naeem,
W.P. Partridge, C. Stere, Analyst 135 (2010) 2260-2272.
[71]
K. Irani, W.S. Epling, R. Blint, Topics in Catalysis 52 (2009) 1856-1859.
[72]
A. Russell, W.S. Epling, H. Hess, H.Y. Chen, C. Henry, N. Currier, A. Yezerets, Industrial
and Engineering Chemistry Research 49 (2010) 10311-10322.
[73]
E. Ruckenstein, B. Pulvermacher, Journal of Catalysis 29 (1973) 224-245.
[74]
P.C. Flynn, S.E. Wanke, Journal of Catalysis 34 (1974) 390-399.
[75]
[76]
[77]
R.M.J. Fiedorow, S.E. Wanke, Journal of Catalysis 43 (1976) 34-42.
[78]
T.J. Lee, Y.G. Kim, Journal of Catalysis 90 (1984) 279-291.
[79]
J.R. Chang, S.L. Chang, T.B. Lin, Journal of Catalysis 169 (1997) 338-346.
[80]
L. Olsson, H. Karlsson, Catalysis Today 147 (2009) S290-S294.
[81]
M. Krkkinen, M. Honkanen, V. Viitanen, T. Kolli, A. Valtanen, M. Huuhtanen, K.
Kallinen, M. Vippola, T. Lepist, J. Lahtinen, R.L. Keiski, Topics in Catalysis 56 (2013) 672-678.
[82]
M. Cabi, S. Giorgio, C.R. Henry, M.R. Axet, K. Philippot, B. Chaudret, The Journal of
Physical Chemistry C 114 (2010) 2160-2163.
[83]
A.F. Lee, K. Wilson, R.M. Lambert, C.P. Hubbard, R.G. Hurley, R.W. McCabe, H.S.
Gandhi, Journal of Catalysis 184 (1999) 491-498.
[84]
A. Hinz, M. Skoglundh, E. Fridell, A. Andersson, Journal of Catalysis 201 (2001) 247257.
[85]
D. Bounechada, S. Fouladvand, L. Kylhammar, T. Pingel, E. Olsson, M. Skoglundh, J.
Gustafson, M. Di Michiel, M.A. Newton, P.A. Carlsson, Physical Chemistry Chemical Physics 15 (2013)
8716-8721.
[86]
J.C. Summers, Environmental Science and Technology 13 (1979) 321-325.
[87]
Y. Xie, Y. Chen, Y. Ma, Z. Jin, Journal of Hazardous Materials 195 (2011) 223-229.
[88]
H. Wakita, Y. Kani, K. Ukai, T. Tomizawa, T. Takeguchi, W. Ueda, Applied Catalysis A:
General 283 (2005) 53-61.
[89]
J. Dawody, M. Skoglundh, L. Olsson, E. Fridell, Journal of Catalysis 234 (2005) 206-218.
[90]
L. Limousy, H. Mahzoul, J.F. Brilhac, P. Gilot, F. Garin, G. Maire, Applied Catalysis B:
[91]
R. Streber, C. Papp, M.P.A. Lorenz, O. Hfert, E. Darlatt, A. Bayer, R. Denecke, H.P.
Steinrck, Chemical Physics Letters 494 (2010) 188-192.
[92]
P. Zebisch, M. Stichler, P. Trischberger, M. Weinelt, H.P. Steinrck, Surface Science 371
(1997) 235-244.
[93]
S.W. Nam, G.R. Gavalas, Applied Catalysis 74 (1991) 53-64.
[94]
J.-Y. Luo, D. Kisinger, A. Abedi, W.S. Epling, Applied Catalysis A: General 383 (2010)
182-191.
[95]
J. Li, A. Kumar, X. Chen, N. Currier, A. Yezerets, SAE Technical Papers 2 (2013).
67
[96]
J.H. Pazmio, J.T. Miller, S.S. Mulla, W. Nicholas Delgass, F.H. Ribeiro, Journal of
Catalysis 282 (2011) 13-24.
[97]
Y. Yazawa, N. Kagi, S.I. Komai, A. Satsuma, Y. Murakami, T. Hattori, Catalysis Letters 72
(2001) 157-160.
[98]
Y. Yazawa, H. Yoshida, T. Hattori, Journal of Synchrotron Radiation 8 (2001) 560-562.
[99]
Y. Yazawa, H. Yoshida, N. Takagi, N. Kagi, S. Komai, A. Satsuma, Y. Murakami, T.
Hattori, in: F.V.M.S.M. Avelino Corma, G.F. Jos Luis (Eds.), Studies in Surface Science and Catalysis,
Elsevier, 2000, pp. 2189-2194.
[100]
I.V. Yentekakis, R.M. Lambert, M. Konsolakis, N. Kallithrakas-Kontos, Catalysis Letters
81 (2002) 181-185.
[101]
M. Konsolakis, I.V. Yentekakis, Applied Catalysis B: Environmental 29 (2001) 103-113.
[102]
A. Melchor, E. Garbowski, M.V. Mathieu, M. Primet, Journal of the Chemical Society,
Faraday Transactions 1: Physical Chemistry in Condensed Phases 82 (1986) 3667-3679.
[103]
J. Basset, F. Figueras, M.V. Mathieu, M. Prettre, Journal of Catalysis 16 (1970) 53-61.
[104]
J.P. Giannetti, R.T. Sebulsky, Product R&D 8 (1969) 356-361.
[105]
T. Cai, S. Liu, J. Q, S. Wong, Z. Song, M. He, Applied Catalysis A: General 97 (1993)
113-122.
[106]
T. Yamaguchi, Applied Catalysis 61 (1990) 1-25.
[107]
T. Yamaguchi, T. Jin, K. Tanabe, Journal of Physical Chemistry 90 (1986) 3148-3152.
[108]
O. Saur, M. Bensitel, A.B.M. Saad, J.C. Lavalley, C.P. Tripp, B.A. Morrow, Journal of
Catalysis 99 (1986) 104-110.
[109]
W. Przystajko, R. Fiedorow, I.G. Dalla Lana, Applied Catalysis 15 (1985) 265-275.
[110]
M. Waqif, J. Bachelier, O. Saur, J.C. Lavalley, Journal of Molecular Catalysis 72 (1992)
127-138.
[111]
K. Kamasamudram, N.W. Currier, X. Chen, A. Yezerets, Catalysis Today 151 (2010) 212222.
[112]
R. Nedyalkova, K. Kamasamudram, N.W. Currier, J. Li, A. Yezerets, L. Olsson, Journal of
Catalysis 299 (2013) 101-108.
[113]
G. Delahay, B. Coq, S. Kieger, B. Neveu, Catalysis Today 54 (1999) 431-438.
[114]
A. Grossale, I. Nova, E. Tronconi, Catalysis Today 136 (2008) 18-27.
[115]
M. Colombo, I. Nova, E. Tronconi, Catalysis Today 197 (2012) 243-255.
[116]
P.S. Metkar, M.P. Harold, V. Balakotaiah, Chemical Engineering Science 87 (2013) 5166.
[117]
L. Olsson, H. Sjvall, R.J. Blint, Applied Catalysis B: Environmental 81 (2008) 203-217.
[118]
H. Sjovall, R.J. Blint, L. Olsson, Journal of Physical Chemistry C 113 (2009) 1393-1405.
[119]
H. Sjvall, R.J. Blint, A. Gopinath, L. Olsson, Industrial and Engineering Chemistry
Research 49 (2010) 39-52.
[120]
H. Sjvall, R.J. Blint, L. Olsson, Applied Catalysis B: Environmental 92 (2009) 138-153.
[121]
E. Tronconi, I. Nova, C. Ciardelli, D. Chatterjee, B. Bandl-Konrad, T. Burkhardt, Catalysis
Today 105 (2005) 529-536.
[122]
N. Wilken, K. Kamasamudram, N.W. Currier, J. Li, A. Yezerets, L. Olsson, Catalysis
Today 151 (2010) 237-243.
[123]
S. Shwan, J. Jansson, J. Korsgren, L. Olsson, M. Skoglundh, Catalysis Today 197 (2012)
24-37.
68

Fundamental Studies of Catalytic Systems For Diesel Emission Control

Uploaded by

Copyright:

Available Formats

Fundamental Studies of Catalytic Systems For Diesel Emission Control

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fundamental Studies of Catalytic Systems For Diesel Emission Control

Uploaded by

Copyright:

Available Formats

T HESIS FOR THE D EGREE OF D OCTOR O F P HILOSOPHY

Fundamental studies of catalytic

Department of Chemical and Biological Engineering

Fundamental studies of catalytic systems for diesel emission control

XAVIER AUVRAY, 2013

Doktorsavhandlingar vid Chalmers tekniska hgskola

Department of Chemical and Biological Engineering

Fundamental studies of catalytic systems for diesel emission control

Contribution to the papers:

6.1.1. Thermal aging .............................................................................................................. 27

Figure 1: Combination of catalysts to reduce emissions of diesel vehicles.

_________________________________________________ Introduction emission control catalysis

Chapter 3: Diesel oxidation catalyst _____________________________________________________

_____________________________________________________________ 3.1. NO oxidation over Pt

Figure 3: Hysteresis phenomenon during NO oxidation over Pt/Al2O3 (500ppm NO + 8% O2 + 5% H2O).

Chapter 3: Diesel oxidation catalyst _____________________________________________________

Catalyst temperature (C)

_____________________________________________________________ 3.1. NO oxidation over Pt

Chapter 3: Diesel oxidation catalyst _____________________________________________________

___________________________________________________________ 3.2 Hydrocarbon oxidation

C3H6 concentration (ppm)

Chapter 3: Diesel oxidation catalyst _____________________________________________________

4 NH3 + 3 NO2 7/2 N2 + 6 H2O

Chapter 4: Selective catalytic reduction of NOx by ammonia __________________________________

__________________________________________________________________ 4.3. Side reactions

Chapter 5: Catalyst synthesis and flow reactor experiments __________________________________

Chlorinated Al2O3, higher

Figure 8: Mechanism of alumina chlorination by substitution of hydroxyle surface groups.

__________________________________________________________ 5.2. Catalyst characterization

Figure 9: Schema of CO adsorption on Pt particle surface for measuring the dispersion.

Chapter 5: Catalyst synthesis and flow reactor experiments __________________________________

Figure 10: STEM micrograph of the fresh 1wt% Pt/Al2O3 catalyst.

Figure 11: Schematic representation of the reactor setup used in Paper I.

__________________________________________________________ 5.2. Catalyst characterization

Chapter 5: Catalyst synthesis and flow reactor experiments __________________________________

Figure 12: Schematic representation of SpaciMS system.

__________________________________________________________ 5.2. Catalyst characterization

_________________________________________________________________ 6.1. Aging of catalyst

_________________________________________________________________ 6.1. Aging of catalyst

________________________________ 6.2. Interaction of sulfur compounds with oxidation catalysts

Outlet concentration (ppm) /

Outlet concentration (ppm)

________________________________ 6.2. Interaction of sulfur compounds with oxidation catalysts

Figure 21: NO conversion measured by temperature-programmed reaction (500ppm NO + 8% O2 +

________________________________ 6.2. Interaction of sulfur compounds with oxidation catalysts

C3H6 conversion (%)

Figure 22: C3H6 conversion measured by temperature-programmed reaction (500ppm C3H6 + 8% O2 +

______________________________________________ 7.3. Sulfation to increase the catalyst acidity

Figure 25: NO conversion comparison (500ppm NO + 8% O2 + 5% H2O) between chlorinated Pt/Al2O3

__________________________________________ 7.4. Oxidation ability of an acidity-enhanced DOC

C3H6 conversion (%)

C3H6 conversion (%)

Feed temperature (C)

Sulfur release (mol/gcat)

__________________________________________ 7.4. Oxidation ability of an acidity-enhanced DOC

Catalyst temperature (C)

H2S outlet concentration (ppm)

8.1. Experimental measurements

Chapter 8: Intra-catalyst insight of SCR with NH3 on Cu-BEA __________________________________