optimization of n2o decomposition rhox/ceria...

Optimization of N2O decomposition RhOx/ceria catalysts and design of a high N2-selective

deNOx system for diesel vehicles

Verónica Rico Pérez

www.ua.es

www.eltallerdigital.com

Optimization of N2O decomposition

RhOx/ceria catalysts and design of a high N2-selective deNOx system for diesel vehicles

PhD Thesis July 2013


UNIVERSIDAD DE ALICANTE

Departamento de Química Inorgánica Grupo de Materiales Carbonosos y Medio Ambiente (MCMA)

Optimization of N2O decomposition RhOx/ceria catalysts and design of a high N2-selective deNOx system for

diesel vehicles

Memoria presentada para aspirar al grado de Doctor


Directores del trabajo:

Agustín Bueno López Concepción Salinas Martínez de Lecea

Profesor Titular de Catedrática de Química Inorgánica Química Inorgánica

Alicante, Julio 2013

Index

i

SUMMARY OF CONTENTS 1

CHAPTER 1 7

Introduction

1.1 Environmental effects of N2O. 9

1.2 Sources of N2O. 10

1.3 Diesel versus Gasoline engines. 12

1.4 TWC for gasoline vehicles. 15

1.5 Gas pollution control in diesel vehicles. 16

1.5.1 Diesel Oxidation Catalysts (DOC). 17

1.5.2 Diesel Particulate Filters (DPF). 17

1.5.3 NOx emissions control. 21

1.5.3.1 Catalytic decomposition of NOx. 21

1.5.3.2 NOx Storage Reduction (NSR). 22

1.5.3.3 Selective Catalytic Reduction (SCR). 24

1.6 N2O abatement. 28

1.7 Background of the research group. 29

1.8 Objectives. 31

CHAPTER 2 45

Characterization techniques and catalytic tests

2.1 Characterization techniques. 47

2.1.1 N2 adsorption at -196 ºC. 47

2.1.2 X-Ray Diffraction (XRD). 50

2.1.3 Raman spectroscopy. 54

2.1.4 X-ray Photoelectron Spectroscopy (XPS). 57

2.1.5 Temperature Programmed Reduction with H2 (H2-TPR). 63

2.1.6 Transmission Electron Microscopy (TEM). 64

Index

ii

2.1.7 Scanning Electron Microscopy (SEM). 66

2.1.8 Thermobalance coupled to a Mass Spectrometer (TG-MS). 67

2.2 Catalytic tests at different scales. 68

2.2.1 Powder and small size honeycomb monolith catalysts. 69

2.2.2 Medium-size honeycomb monolith catalysts. 70

2.2.3 Full-size honeycomb monolith catalysts. 72

CHAPTER 3 79

Effect of the calcination conditions of the RhOx/CeO2

catalysts on N2O decomposition activity

3.1 Introduction. 81

3.2 Experimental. 82

3.2.1 Study of the metal precursors decomposition. 82

3.2.2 Catalysts preparation. 82

3.2.3 Catalysts characterization. 83

3.2.4 N2O decomposition tests. 84

3.2.5 CO oxidation tests. 84

3.3 Results and discussion. 84

3.3.1 Thermogravimetry - Mass Spectroscopy (TG-MS) study of

metal precursors decomposition. 84

3.3.2 XRD, Raman spectroscopy and N2 adsorption at -196 ºC

characterization. 86


3.3.4 Characterization by XPS of fresh catalysts and after “in

situ” pre-treatments with N2O at 225 ºC. 90

3.3.5 H2-TPR characterization. 93

3.3.6 TEM characterization. 95

3.3.7 CO oxidation tests. 96

3.4 Conclusions. 97

Index

iii

CHAPTER 4 105

Preparation of RhOx/CeyPr1-yO2 N2O decomposition

catalysts by rhodium nitrate impregnation with different

solvents







4.3.1 Catalysts temperature during rhodium nitrate thermal

decomposition. 109


4.3.3 Catalysts characterization by N2 adsorption at -196 ºC,

XRD and Raman spectroscopy. 114

4.3.4 Catalysts characterization by TEM, XPS and H2-TPR. 120

4.4 Conclusions. 129

CHAPTER 5 135

Preparation, characterization and N2O decomposition

activity of honeycomb monolith-supported

RhOx/Ce0.9Pr0.1O2 catalysts







5.3.1 SEM-EDS-chemical mapping characterization. 141

5.3.2 XRD and Raman spectroscopy characterization. 145

Index

iv

5.3.3 Characterization by N2 adsorption at -196 ºC. 147

5.3.4 H2-TPR characterization. 148

5.3.5 TEM characterization. 149



CHAPTER 6 161

NOx reduction to N2 with commercial fuel in a real diesel

engine exhaust using a dual bed which constists of

Pt/Beta zeolite and RhOx/ceria monolith catalysts


6.2 Experimental details. 165


6.2.1.1 Medium-size Pt/Beta zeolite/monolith catalysts

preparation. 165

6.2.1.2 Full-size Pt/Beta zeolite/monolith catalyst preparation. 168

6.2.1.3 Medium-size RhOx/Ce0.9Pr0.1O2/monolith catalyst

preparation. 170

6.2.2 SEM characterization. 171

6.2.3 Catalytic tests. 171


6.3.1 Study of the Beta zeolite suspensions viscosity. 171

6.3.2 Effect of the Beta zeolite suspension viscosity on medium-

size monoliths dip-coating. 174

6.3.3 Medium-size monolith coating with Beta zeolite in

consecutive dipping steps. 175

6.3.4 SEM characterization of medium-size monolith catalysts. 178

6.3.5 SCR experiments performed with commercial diesel fuel

and the medium-size Pt/Beta zeolite/monolith catalyst. 180

6.3.6 SCR experiments performed with commercial diesel fuel

and the full-size Pt/Beta zeolite/monolith catalyst. 182

Index

v

6.3.7 SCR experiments performed at 300 ºC with commercial

diesel fuel and a dual-bed which contains Pt/Beta zeolite

and RhOx/Ce0.9Pr0.1O2 medium-size monolith catalysts. 186


CHAPTER 7 195

General Conclusions

RESUMEN 201

Optimización de catalizadores RhOx/ceria para la

descomposición de N2O y diseño de un sistema deNOx

altamente selectivo a N2 para vehículos diésel

ABBREVIATIONS 227

CURRICULUM VITAE 233

Summary of contents

Summary

3

Summary of contents

This thesis gathers a full research process ranging from basic or

fundamental investigation to actual implementation. This work deals with

rhodium catalysts supported on ceria-based materials preparation,

characterization and catalytic test towards N2O decomposition for a better

understanding of the effect of their physico-chemical properties on the

catalytic behavior.

This thesis is structured in seven chapters. Chapter 1 describes

the environmental problems caused by N2O and its main sources, focusing

on diesel engines. A review of technology and catalysts used for diesel

pollutants abatement is presented, including previous work performed in

our group and the main objectives of this PhD thesis.

Chapter 2 describes the characterization techniques employed to

obtain information about the physico-chemical properties of the materials

and the experimental systems and conditions used to carry out the

catalytic tests.

Chapters 3 to 6 present the discussion of experimental results. In

Chapter 3 the calcination conditions for catalysts preparation were

modified in order to improve the distribution of rhodium and the RhOx-ceria

interface of RhOx/CeO2 catalysts. Improved catalytic activity for N2O

decomposition and CO oxidation was obtained by flash calcination, which

consisted of introducing the ceria support-impregnated rhodium precursor

in a furnace which was pre-heated at 250 ºC. The speed at which water

evaporates from the ceria-based support seemed very important for the

catalyst properties, because it affects the size of the RhOx particles on the

final catalyst. The following publication describes the results obtained in

part of this study: V. Rico-Pérez, M.A. Velasco-Beltrán, Q. He, Q. Wang,

C. Salinas-Martínez de Lecea, A. Bueno-López. Preparation of

ceria-supported rhodium oxide sub-nanoparticles with improved

catalytic activity for CO oxidation. Catalysis Communications 33 (2013)

47.

In Chapter 4 was studied the effect of the solvent (water, ethanol

or acetone) used to impregnate CeyPr1-yO2 (y = 1, 0.9 or 0.5) supports with

Summary

4

rhodium nitrate, in order to prepare N2O decomposition catalysts. The

activity of the catalysts studied was related with the RhOx-support

interaction, and both the nature of the ceria support and of the solvent

used for rhodium impregnation affected such interaction. Ceria doping with

10 % praseodymium had a positive effect on RhOx-support interaction, but

the benefit on the catalytic activity was only obtained for water

impregnation because the temperature peaks created during calcination of

ethanol and acetone-impregnated catalysts promoted Ce0.9Pr0.1O2 and

RhOx sintering. The interaction between RhOx and Ce0.5Pr0.5O2 was not as

good as that with Ce0.9Pr0.1O2. The best catalyst was obtained by

impregnating Ce0.9Pr0.1O2 with a water solution of rhodium. However, if

acetone or ethanol must be used for any reason the bare ceria support is

more suitable (under the calcination conditions of this study; 250 to 500 ºC

at 10 ºC/min) because is more stable towards sintering during solvents

combustion.

In Chapter 5 RhOx/Ce0.9Pr0.1O2 active phases were loaded by

sequential impregnation into small-size (1 cm diameter) cordierite

honeycomb monoliths following the procedure (i) cerium and

praseodymium nitrates impregnation + calcinations and ii) rhodium nitrate

impregnation + calcination. The supported catalysts were characterized by

XRD, Raman spectroscopy, SEM-EDS, TEM-EDS and H2-TPR, and tested

for N2O decomposition. Rhodium oxide particles were selectively attached

to the Ce0.9Pr0.1O2 and not to the cordierite. The optimum content of

rhodium was 0.2 wt.% (in total weight). The calcination procedure

significantly affected the supported catalyst features. The best catalyst was

prepared by flash calcination yielding to smaller RhOx particles and

lowering the temperature for surface Rh-Ce-Pr entities reduction in

comparison to ramp calcination, improving both the distribution of active

phases on the cordierite substrate and the catalytic activity for N2O

decomposition. These results have been published in: V. Rico-Pérez,

S. Parres-Esclapez, M.J. Illán-Gómez, C. Salinas-Martínez de Lecea,

A. Bueno-López. Preparation, characterization and N2O decomposition

activity of honeycomb monolith-supported Rh/Ce0.9Pr0.1O2 catalysts.

Applied Catalysis B: Environmental 107 (2011) 18.

In Chapter 6 medium-size (2.3 cm diameter) and full-size (14 cm

diameter) Pt/Beta zeolite monolith catalysts were successfully prepared

and tested in SCR technology for NOx abatement in a real diesel engines

Summary

5

exhaust using commercial diesel fuel as reducing agent. Also a

medium-size dual bed catalytic system consisting of a

Pt/Beta zeolite/monolith SCR catalyst located upstream a N2O

decomposition RhOx/Ce0.9Pr0.1O2/monolith catalyst were tested in the

same reaction with the objective to improve N2 selectivity. The dip-coating

method was optimized for the Pt/Beta zeolite/monolith preparation. The

catalyst used for N2O decomposition, RhOx/Ce0.9Pr0.1O2/monolith, was

prepared by nitrate precursor decomposition. The production of N2O as

undesired NOx reduction product, which is a drawback of platinum SCR

catalysts, has been solved by using the dual bed configuration, where both

monolith catalysts operated at the same temperature, and 100 % N2

selectivity has been obtained.

Finally, Chapter 7 summarizes the most relevant conclusions

obtained in this thesis.

Since this thesis applies for the International Doctorate mention,

this is written in English and a summary in Spanish is presented at the end

of the thesis to fulfil the requirements of this mention.

CHAPTER 1

Introduction

This chapter describes the environmental problems caused by

N2O and its main sources, focusing on diesel engines. A review of

technology and catalysts used for diesel pollutans abatement is

presented, including previous work performed in our group, and the

main objectives of this PhD thesis.

Introduction

9

1.1 Environmental effects of N2O.

Nitrous oxide (N2O) is naturally produced by biological processes

occurring in soil and water and by a variety of anthropogenic activities

related with agriculture, energy, industry and waste management activities

[1, 2].

From an environmental point of view, Kapteijn et al. [3] reported

some years ago that N2O was considered as a relatively harmless species

and suffered from a lack of interest by scientists, engineers and politicians

for long time, due to the underestimation and unawareness of the potential

contribution of this species to environmental problems. From the middle

80’s a growing concern was noticed since N2O was identified as a

contributor to the destruction of the ozone in the stratosphere. Nowadays,

it is well known that N2O is a dangerous environmental pollutant because it

contributes to the destruction of the stratospheric ozone, being at the same

time a greenhouse gas (GHG).

The greenhouse effect is a phenomenon caused by strong

absorbance of infrared radiation in the atmosphere, increasing the

temperature of the Earth. Without a natural greenhouse effect the average

Earth temperature would be 33 ºC below the actual one and this would

hinder life on the planet. The problem occurs when the amount of

greenhouse gases increases, enhancing the natural greenhouse effect,

and as consequence, the average global temperature and the related

climate changes also increase.

N2O is the major source of NOx (NO + NO2) in the stratosphere,

and therefore, is an important natural regulator of the stratospheric ozone.

The overall influence of N2O on the ozone layer is complex and very

different from that of substances covered by the Montreal protocol in

September 1987. At this time, there were no doubts about the negative

effect of N2O on the ozone depletion layer [1]. Indeed, N2O emissions are

the largest of all the ODSs (Ozone Depleting Substances) and are

projected to remain the largest for the rest of the 21st century [4].

The environmental impact of this gas is increasing because the

atmospheric N2O concentration is increasing mainly by human activities

[1, 5].

Chapter 1

10

Although N2O is not the major contributor to global warming

(∼6 %), it is much more potent than either the other two most common

anthropogenic greenhouse gases, CO2 and CH4. Due to its long lifetime of

approximately 150 years in the atmosphere, N2O has around 310 times the

Global Warming Potential (GWP)1 of CO2. For this reason, a relatively

limited emission (compared to other greenhouse gases) is equivalent to

about 10 % of the CO2 emission [1, 2]. At the Third Conference of the

Parties (COP-3) of the United Nations Framework on Climate Change

(UNFCC), in Kyoto (Japan) in December 1997, legal binding targets were

set for reducing emissions of six greenhouse gases (CO2, CH4, N2O, HFC,

PFC, SF6) to be achieved in the period 2008–2012 [6]. In Doha, Qatar, on

8 December 2012, the "Doha Amendment to the Kyoto Protocol" was

adopted. The composition of Parties changed and the amendment

included a second commitment period, from 1 January 2013 to 31

December 2020, a revised list of greenhouse gases (GHG) and some

updates of several articles pertaining to the first commitment period.

During the second commitment period, Parties committed to reduce GHG

emissions by at least 18 % below 1990 levels in the eight-year period from

2013 to 2020 [6].

1.2 Sources of N2O.

As mentioned before N2O was produced by both natural and

anthropogenic sources, the main sources of atmospheric N2O being the

microbial action in soils, the manufacture of nylon and adipic acid, and

fossil fuel combustion in stationary and mobil sources [1-3, 7]. It was also

reported that chemical processes associated with the production and use

of nitric acid and fluidized bed combustion were N2O sources, and their

contribution to the total N2O emissions amount were about 20 %.

A new inventory of U.S. Greenhouse gas emissions and sinks:

1990–2011 has been published indicating that although the direct

greenhouse gases CO2, CH4, and N2O occur naturally in the atmosphere,

1 The Global Warming Potential (GWP) of a greenhouse gas is defined as the

ratio of the time-integrated radioactive forcing from the instantaneous release of 1 kilogram (kg) of a trace substance relative to that of 1 kg of a reference gas (IPCC 2001) [1].

http://unfccc.int/kyoto_protocol/doha_amendment/items/7362.php

Introduction

11

human activities have changed their atmospheric concentrations. From the

pre-industrial era (i.e., ending about 1750) to 2010, concentrations of these

greenhouse gases have increased globally by 39, 158 and 18 %,

respectively [1]. Figure 1.1 illustrates the relative contribution of the direct

greenhouse gases to the total U.S. emissions in 2011 and the main

anthropogenic sources of N2O emissions. As observed in Figure 1.1 N2O

represents 5.6 % of global greenhouse gases emission. Agricultural soil

management, stationary fuel combustion, mobile source fuel combustion,

manure management and nitric acid production were the major sources.

From 1990 to 2011, a 4 % N2O emissions (14.5 Tg CO2 eq) increase has

been detected [1].

2011 Sources of N2O Emissions

2011 Greenhouse Gas Emissions by Gas

(percentages based on Tg CO2 Eq)

Figure 1.1. 2011 Greenhouse Gas Emissions (percentages based on Tg CO2-eq.) and

Sources of N2O Emissions [1].

Emissions of N2O that can be reduced in the short term are

associated with chemical production and energy industry (∼35 % in the

EU). This emission is concentrated in a limited number of large facilities,

which holds promise for an economic and efficient reduction strategy to

fulfill the Kyoto commitment. In the particular case of adipic acid and nitric

acid production, technologies are commercially available, but their

application (extrapolation or adaptation) to other sources, e.g. light-duty

Chapter 1

12

vehicles, is not feasible due to the dissimilar characteristics of the

exhausts. Diluted N2O streams, relatively low temperature

(typically < 500 ºC), and presence of catalyst inhibitors (O2, H2O, NOx and

SO2) are features of some light-duty vehicles [1-3, 8].

Historically, medium-duty vehicles were not part of Corporate

Average Fuel Economy (CAFE) regulations. However, manufacturers

provide GHG emission data for carbon dioxide (CO2), methane (CH4), and

nitrous oxide (N2O) to the Environmental Protection Agency (EPA) since

2011 [9, 10]. The proposed emission standards for both N2O and CH4 from

medium-duty vehicles certified on a chassis or engine dynamometer are2

50 mg/mile and 50 mg/bhp-h (milligrams per brake horsepower-hour) and

must be reported as CO2 equivalent (CO2-eq) in g/mile, taking into account

the global warming potential of the other gases. Thus emissions at the

proposed standards equate to 14.9 gCO2-eq/mile and 14.9 gCO2-eq/bhp-h

for N2O, and 1.25 g CO2-eq/mile and 1.25 gCO2-eq/bhp-h for CH4 [9, 11].

As above reported, one of the main N2O sources is mobile

combustion. This thesis is focused on N2O abatement in diesel engines.

For a better understanding of the problem faced, a brief explanation of

gasoline and diesel engines operation and their exhaust emissions will be

carried out in the next section.

1.3 Diesel versus Gasoline engines.

Currently, automobile prevails across the globe as the most

popular and necessary mode of transportation in our daily lives. About

50 million cars are produced every year, and over 700 million cars are

used worldwide [12]. The increasing number of cars is seen as the major

source of pollutants causing the decrease in air quality, health problems,

and recirculation of polluted air by frequent ambient temperature inversion

and formation of “photochemical smog” in the major cities [13, 14].

2 “miles” are used instead of kilometers in this document when talking about

US regulations, in order to keep the original terminology (1 mile = 1.609 Km).

Introduction

13

Due to environmental concerns and the increasing price of

traditional fossil fuels, alternative fuels are now receiving more attention

than before in many countries. Also, many developed countries are

currently encouraged to find out alternative approaches to promote fuel

economy and reduce the environmental impact from internal combustion

diesel engines [15].

Diesel and gasoline engines have different modes of operation. In

diesel engines, the fuel auto-ignites as it is sprayed into the combustion

chamber at a high pressure. A spark is not required to ignite the diesel

fuel, which is a heavier, less volatile mixture of hydrocarbons than gasoline

and chemically more susceptible to auto-ignition. Relative to gasoline

engines, diesel engines have higher compression ratios, faster

combustion, lower throttling losses, operate leaner (i.e., at a greater

air–fuel ratio) and work with O2 excess. In gasoline engines (also known as

Otto engines), air and fuel are mixed before introduction into the cylinder.

The air-to-fuel ratio is constant and is often chosen to be stoichiometric,

i.e. about 14.6 on an air-to-fuel weight basis [14]. As a result, diesel

engines have an inherently greater thermodynamic efficiency than gasoline

engines. Diesel fuel also has approximately 12 % greater volumetric

energy content than gasoline, and hence, diesel engines operate with a

higher volumetric fuel economy (lower fuel consumption) than gasoline

engines [9, 15], which tipically leads to lower CO2 emission.

The use of catalysts for purifying exhaust gases is absolutely

necessary and indispensable in every vehicle taking into account current

regulations [12, 14]. The major pollutants emmited by a gasoline engine

are HC (hydrocarbon), CO (carbon monoxide), NOx (NO with only traces

of NO2) and PM (particulate matter) in very low concentration. Standards

of exhausts emissions were tightened in subsequent years which

conducted to a successful development and introduction of the three-way

catalysts (TWC). Nowadays, all new gasoline cars are equipped with a

catalytic convertor, reducing CO, HC and NOx emissions.

Currently, CO and HC emissions from diesel engines are about the

same than catalytically equipped gasoline engines. However, NOx and

certainly PM emissions from diesel engines are much higher. TWC does

not match in diesel engine mainly for two different reasons:

Chapter 1

14

High O2 concentration that hinders direct reduction of NOx.

The need of a specific system for soot abatement based on

a filter which should be regenerated.

Even though diesel engines were considered clean in comparison

with gasoline exhaust gases (when TWC were introduced, PM

concentration was not legislated), NOx and PM emissions from diesel

engines are much higher. Since 1982 PM standards for diesel engines

have also been tightened over the years [14].

Typical diesel exhaust gases are subdivided in three groups:

harmless compounds, regulated harmful compounds (harmful compounds

subjected to regulation) and unregulated harmful compounds. The first

group (O2, CO2, H2O and N2) is harmless in the sense that the compounds

have no direct adverse effect on health, despite CO2 does contribute to the

greenhouse effect. The second group of regulated harmful compounds

includes CO, HC, NOx, PM and SOx, and the unregulated harmful

compounds include aldehydes, ammonia, cyanide, benzene, toluene and

PAHs (polycyclic aromatic hydrocarbon) [14]. Since 1993 an increasing

restriction in emission standards over the regulated harmful compounds

group has been done [13].

Emission targets of different contaminants have become more

restrictive over the years and several techniques are promising to reduce

emissions from diesel engines. Modified or alternative fuels have been

studied, trying to correlate emission levels with certain fuel specifications in

order to optimize fuel composition towards low emissions. Also, some

engine modifications have been very effective in reducing diesel engine

emission levels and have been implemented. Finally, after-treatment

techniques have been studied extensively for NOx and also for PM

removal from diesel exhaust gases [14]. The after treatment techniques

will be the alternative to reduce diesel emissions used in this work. In the

next section, the different configuration commercialized or proposed in the

literature for CO and HC oxidation, NOx abatement and finally N2O

decomposition will be presented.

Introduction

15

1.4 TWC for gasoline vehicles.

In gasoline engines, CO and HCs oxidation catalysts eventually

chosen in the early catalytic converters, were based on platinum group

metals (PGMs). The catalytic oxidation of CO and HCs (HmCn) would

follow the reactions (1) and (2).

2CO + O2 → 2CO2 (1)

4HmCn + (m+4n)O2 → 2mH2O + 4nCO2 (2)

NOx has also to be controlled, and Pt/Rh catalyst formulations

were used for the selective reduction of NO to N2, under rich fuel

conditions, according to reactions (3) and (4).

(8n+2m)NO + 4HmCn → (4n+m)N2 + 2mH2O + 4nCO2 (3)

2NO + 2CO → N2 + 2CO2 (4)

Early TWC used two catalysts. The engine ran slightly rich fuel to

enable reduction of NOx over a Pt/Rh catalyst, and air was introduced

before a second catalyst to oxidize the CO and HCs excess. After some

time, some European car manufacturers, notably Volkswagen and Volvo,

used Pt/Rh oxidation catalysts and when operated around the air/fuel

stoichiometric point, it was found they could provide good NOx control and

oxidize HC and CO at the same time. The use of Pt/Rh catalysts to control

HC, CO and NOx simultaneously became the preferred system because

all three major pollutants were controlled by one catalyst, and the concept

was christened as the “three way catalyst” (TWC) [13].

To use a catalyst in real gas streams it is necessary to load the

active phases (powder) into an appropriate inert support. Diverse

substrates can be used, such as pellets or inorganic oxide particles,

honeycomb ceramic monoliths, ceramic foams, etc [16-20]. In the 1960s,

Johnson Matthey was involved with catalytic control of gaseous pollutants

such as NOx from nitric acid plants. When a plan to reduce emissions from

US cars by 90 % was announced in 1970, the company was ready to

develop catalytic technology for automotive emission control. Two types of

Chapter 1

16

pellet catalysts were first introduced, but Johnson Matthey concentrated on

the “monolithic” type that is used in all auto-catalytic systems today [13].

Among the potential catalyst supports, honeycomb monoliths

present some attractive properties. They provide a good contact between

the active phases of the catalyst and the treated gases and they present a

high dust tolerance. The pressure drop is minimized because the gas flow

is not significantly impeded through the catalytic bed and also they have no

degradation problems typically occurring, for instance, in particles friction

[16-19].

By far, the most successful substrate materials are based on

compositions which when extruded and fired at high temperature form

aligned cordierite, 2MgO·2Al2O3·5SiO2, and about 85 % of the substrates

on cars today are made of this material [13].

The monoliths had no catalytic activity, they are inert supports, and

a process was developed for coating the monoliths with both high surface

area alumina and platinum group metals (PGMs). Today, TWCs are based

on combinations of platinium and/or palladium and rhodium, alumina and

ceria (also known as cerium oxide or CeO2), together with a variety of

support stabilizers, activity promoters, and selectivity improvers. Elements

used include iron, nickel, manganese, calcium, strontium, barium,

lanthanum, neodymium, praseodymium and zirconium. While most

commercial catalysts contain one, and usually more, of these minor

elements, all of them contain PGMs, alumina and ceria in different

proportions.

1.5 Gas pollution control in diesel vehicles.

The different pollutants emmited by diesel engines require specific

post-combustion strategies, opposite to that described for TWC in gasoline

vehicles. CO and HC are ususally oxidized in a diesel oxidation catalyst

(DOC), soot particles are collected in filters and oxidized and NOx

reduction must be accomplished by adding a reductant in the presence of

a suitable catalyst. All these strategies are discussed in this section.

Introduction

17

1.5.1 Diesel Oxidation Catalysts (DOC).

Catalytic emission control from diesel vehicles began with the

fitting of platinum monolithic oxidation catalysts to Volkswagen diesel cars

in 1989 [13], before the demanding European legislation was in place. The

catalyst removed the characteristic odor of diesel exhaust and controlled

HC and CO emissions. With the introduction of the legislation in 1993

fitting of oxidation catalysts to all new European diesel cars became

necessary. However, the temperature of a diesel car exhaust is low

compared to that of a gasoline counterpart and this was a major challenge

for catalyst design [13].

The oxidation of HCs and CO in diesel exhausts is not

straightforward as temperatures are low, due to the fuel efficient nature of

the diesel engine, so higher amounts of HCs can absorb on active sites

and block them. Also poisons, such as sulphur oxides, are strongly

adsorbed. Traditionally platinum-based oxidation catalysts are used. When

the engine is started, the catalyst is insufficiently warm to oxidize HCs

initially present, but incorporating zeolites into the catalyst significantly

improved the performance during the so-called “cold start” period (when

the temperature is not high enough for the catalyst to work). The zeolite

adsorbs HCs preventing inhibition of the platinum active sites. This

improves low temperature CO oxidation and HC removal. At higher

temperature the HCs are desorbed and oxidized on the platinum sites [13].

1.5.2 Diesel Particulate Filters (DPF).

Several types of ceramic and sintered metal diesel particulate

filters (DPFs) have been developed for soot (PM) removal and the most

successful and commonly used are porous ceramic wall-flow filters.

Porous refractory materials used to make them include cordierite, silicon

carbide (SiC) and aluminum titanate (Al2O3TiO2) among others. The

preferred materials to manufacture DPF filters are cordierite and SiC due

to their adequate properties under regeneration conditions. Melting

temperatures (∼1400 and ∼2700 ºC, for cordierite and SiC, respectively)

and expansion coefficients (2.0 × 10−6

and 4.3 × 10−6

ºC-1

from 25 to

800 ºC, respectively) are important differences in the physical properties

between cordierite and SiC. Due to these differences, SiC is able to

Chapter 1

18

support higher temperatures reached during filters regeneration than

cordierite, but is more prone to suffer damages due to thermal shock

[21, 22].

A DPF filter is illustrated schematically in Figure 1.2. Alternate

channels are plugged on one side and opened on the other of the piece

and vice versa, so the exhaust gas is forced through the channel walls but

the PM does not and it is trapped in the filter. As PM accumulate in the

filter, the backpressure across increases. Before backpressure is too high

and the engine stops working the filter must be regenerated, so PM is

removed by oxidizing to CO2 and H2O. The arrows indicate the gas flow

through the walls [13].

Figure 1.2. Schematic representation of a ceramic diesel wall-flow filter.

The DPFs regeneration can be reached by different ways:

-The PSA (Peugeot-Citroën Societé d’Automobiles) system: a

cerium-fuel additive leads to the formation of ceria particles well embedded

into the PM structure, which lower its ignition temperature. Once a high

pressure drop is detected by a sensor, fuel is injected and its combustion

produces an increase of the exhaust gas temperature that promotes PM

ignition. Ceria catalyzed PM combustion diminishes the amount of fuel

required for trap regeneration. Recently, iron based catalysts are also

being used [21].

-Passive regeneration: known as Continuously Regenerating Trap

(CRT®), where, at speeds around 100 km/h, the temperature can be

Introduction

19

sufficiently high for NO to be oxidized over an upstream platinum oxidation

catalyst, producing NO2 and also oxidazing CO and HCs:

2NO + O2 → 2NO2 (5)

NO2, which is more oxidizing than NO and O2, rapidly reacts with

PM in the filter [13, 21, 23]. Another alternative additionally includes a

catalyst in the filter to re-oxidize NO to NO2 therefore improving the overall

efficiency of PM removal. This alternative is called Catalyzed Continuously

Regenerating Trap (CCRT) [24].

-Active regeneration: it is employed on cars that periodically

increase the exhaust gas temperature to burn PM in the filter with oxygen

(typically starting at 550 – 600 ºC) every 400 – 2000 km, depending on

actual driving conditions. Here, additional fuel from the engine is oxidized

over upstream platinum or platinum/palladium oxidation catalyst to provide

the high temperature to initiate PM burning that is then carefully controlled

by restricting the O2 available by throttling the engine.

According to the different technologies, three filter systems (named

as Generation 1, 2 and 3) have been developed and used commercially for

cars that use periodic active regeneration in which catalysis has key roles,

and these are illustrated in Figure 1.3.

Generation 1. This system employs one or two platinum based

oxidation catalysts upstream of a filter to control HC and CO emissions,

and to convert NO to NO2 for passive PM combustion when conditions

permit this to take place. It requires a fuel additive that is converted to

oxide in the engine which is retained in the filter and can lower the

temperature for PM combustion [13].

Generation 2. This system has the advantage of not using a fuel

additive, so it does not require a fuel additive tank and the associated

pump, etc. This system has one or two separate oxidation catalysts

upstream of a filter. The filter has platinum or platinum/palladium catalyst in

its channel walls to promote PM combustion, and today many cars use this

configuration [13].

Chapter 1

20

Generation 3. It does not require a fuel additive nor an upstream

catalyst, and it combines in a single filter (typically SiC, Al2O3TiO2, or

cordierite) with the oxidation catalyst to oxidize HC and CO during normal

driving, and to periodically oxidize extra partially burnt fuel to raise the

temperature to combust PM with O2 during active regenerations. The

catalyst also oxidizes NO to NO2 to provide some passive PM removal

during high-speed driving. This system is thermally the most efficient, and

during active regenerations only the filter must be heated, which is

mounted on the engine turbocharger to minimize heat losses. The

oxidation reactions used to boost the temperature take place actually in the

filter in the same location as the retained PM. In contrast, earlier systems

with a separate upstream catalyst lose considerable heat during

regenerations to the environment via the pipe between the turbocharger

and the filter [13].

Figure 1.3.Three filter systems used on European diesel cars [13], where DOC =

Diesel Oxidation Catalyst, DPF = Diesel Particulate Filter and CSF = Catalyst Soot

Filter.

During these PM combustion processes, few NOx is reduced to

N2, typically less than 15 – 20 %, and an additional NOx removal system

must be located downstream the PM filter, where the temperature is

usually lower than 400 ºC [25].

Introduction

21

1.5.3 NOx emissions control.

In this section the technologies that are already commercially

available for NOx abatement in vehicles, and also some proposals for

possible implementation are described. Some of these technologies are

not only suitable to remove NOx from the exhaust gases of diesel and

lean-burn engines, but also to remove NOx from the flue gases of large

combustion plants used for heating or power generation.

Three different processes have been proposed for NOx

abatement:

i) Catalytic decomposition of NOx.

ii) NOx Storage Reduction (NSR), also known as Lean NOx

Trap (LNT), NOx trapping or NOx absorbing catalysts

(NACs).

iii) Selective Catalytic Reduction (SCR).

1.5.3.1 Catalytic decomposition of NOx.

Decomposition of NOx is thermodynamically favorable at

temperatures below 900 ºC, but the activation energy required for this

reaction is too high. Therefore, a catalyst is necessary to lower the

activation energy thus facilitating the reaction. This approach is the

simplest and the most desirable because no reducing agent is required.

However, the relatively high activation energy of a direct NOx

decomposition limits the practical use of this approach [25].

Over different catalysts studied, Ishihara et al. [26] proposed

La0.7Ba0.3Mn0.6Cu0.2In0.2O3 as active catalyst for NO decomposition even in

the coexistence of H2O, O2, and SO2. For this type of catalyst, the NOx

species are strongly adsorbed, and removal of these adsorbed species

and/or surface oxygen plays an important role in the sequence of reaction

steps. It seems likely that NO decomposition proceeds on the vacant

adsorption site, which is formed by removing adsorbed NOx and/or surface

oxygen, and N2O may be the intermediate species. The reaction

Chapter 1

22

temperature, which is too high, makes this catalyst difficult to apply under

actual exhaust conditions [27].

1.5.3.2 NOx Storage Reduction (NSR).

The first NSR catalyst was developed and launched by Toyota in

1994 [23]. Reactions (5 - 9) take place during lean and rich cyclic periods.

During lean operation (excess air) NO is oxidized to NO2 over a platinum

based component in the presence of excess O2, forming a stable nitrate of

an alkaline catalyst component. During the rich or stoichiometric operation,

the exhaust lowers the O2 partial pressure to a point where the nitrate

phase is not stable at the normal operating temperature and it

decomposes, that effectively is the reverse of the storage process, and the

NOx formed is reduced to N2 usually over rhodium based catalysts

[13, 24, 27-29].

2NO + O2 → 2NO2 (5)

Lean

NO2 + MCO3 → MNO3 + CO2 (6)

2MNO3 → MO + 2NO + O2 (7)

2MO + CO2 → MCO3 (8) Rich

2NO + 2CO → N2 + 2CO2 (9)

In the reactions (5) to (9), M typically represents an alkaline earth

metal cation. NOx-trapping catalysts commonly have two layers with the

bottom one containing platinum for oxidation of NO to NO2 together with

the storage component that can be mainly a barium or strontium

compound. Rhodium is incorporated into the top layer. During the periodic

enrichments of the exhaust gas the NOx released from the bottom has to

pass through the rhodium-containing layer, in the presence of a reductant,

being reduced to N2 as shown in Figure 1.4 [24, 29, 30].

During the lean phase, NOx is removed from the exhaust gas by

adsorption onto platinum sites where it is oxidized to NO2 and then

Introduction

23

converted into a solid nitrate phase. During rich phase, the NSR capacity is

recovered by releasing NOx that is reduced to N2 over rhodium catalyst

sites [31].

Figure 1.4. Schematic operation of a NOx Storage Reduction (NSR) catalyst: (a) lean

phase; b) rich phase.

Although a large number of investigations have been conducted,

some problems still remain unsolved. The main problem was the catalysts

deactivation by sulphur, as barium component forms very stable sulfates

from traces of SOx derived from fuel sulphur compounds [27, 32, 33].

Another problem is the thermal deterioration due to reaction of the NOx

storage material with compounds within the wash-coat and particle growth

of both precious metals and NOx storage material. Also contact between

the platinum particles and metal alkaline earth oxide (BaO) reduces the

platinum oxidation activity by an electronic effect. This causes a lowering

of the activity at low temperature (below 250 ºC), a critical problem

especially for emissions from light-duty diesel engines, since a large part of

the testing cycle is characterized by low-temperature emissions, typically in

the range 120–200 ºC [28, 33].

Chapter 1

24

Most NSR catalysts are based on the utilization of BaO as the NOx

storage component (about 15 wt.%), but new generation catalysts also

contain alkali metal oxides to improve high-temperature behavior, TiO2

nanoparticles to enhance regeneration and Rh/ZrO2 to promote in-situ

generation of H2 during periodic high-temperature (about 650 ºC)

regeneration [28].

1.5.3.3 Selective Catalytic Reduction (SCR).

The SCR is one of the technologies proposed for NOx emissions

control in diesel-engine exhausts to convert NOx to N2 in an O2-rich

environment [24, 27, 34, 35]. In the SCR process, NOx reduction

successfully competes with the reduction of O2, even though the latter is

present in a large excess. Mainly based on laboratory experiments, several

reductants have been proposed to accomplish the SCR of NOx in O2-rich

gas streams (lean fuel conditions), including H2, CO, different HCs,

ammonia (NH3), urea (CO(NH2)2) or even diesel fuel among others

[24, 27, 34, 35].

Some of the SCR technology for mobile sources is based on the

well-established technology used in power plants, where NH3 or urea is

injected in the post-combustion stream for NOx reduction to N2 [35]. Based

on that technology, it has been proved that with an appropriate catalyst

NH3 can be used as a very selective NOx reductant in NH3-SCR system

for mobile sources. Among all the catalysts tested, platinum catalysts can

work at relatively low temperatures, and vanadium-based catalysts are

commonly used at temperatures typical of heavy-duty diesel engine

exhaust gas. Traditionally, V2O5 supported on TiO2 has been used

because its resistance to sulphur poisoning. Also there has been much

work reported on metal exchanged zeolite systems; especially those

containing iron, cobalt, copper, and cerium [24]. Fe/zeolite catalysts have

been commercialized and used in several NH3-SCR applications,

achieving high NOx conversions, but oxidation of NH3 affords NO at high

temperatures so the apparent conversion of NO decreases as increasing

amounts of NO are formed from NH3 [24].

Transportation of NH3 (being a corrosive compound) hinders this

technology implementation, becoming a promising alternative the use of

https://en.wikipedia.org/wiki/Carbon


https://en.wikipedia.org/wiki/Nitrogen


Introduction

25

urea. NH3 is then derived from an aqueous urea solution (via urea

hydrolysis, reaction (10)) that is injected into the hot exhaust upstream of

the SCR catalyst [13].

(NH2)2CO + H2O → 2NH3 + CO2 (10)

Koebel et al. [35] reported that NOx in diesel exhaust is usually

composed of > 90 % NO and therefore, once NH3 is generated, the main

reaction of SCR with NH3 will be:

4NH3 + 4NO + O2 → 4N2 + 6H2O (11)

This reaction implies a 1:1 stoichiometry for NH3 and NO and the

consumption of some O2. The non O2 consuming reaction (12) is much

slower and is therefore not relevant in lean combustion gases:

4NH3 + 6NO → 5N2 + 6H2O (12)

On the other hand, the reaction rate with equimolar amounts of NO

and NO2 (reaction (13)) is much faster than that of the main reaction (11)

4NH3 + 2NO + 2NO2 → 4N2 + 6H2O (13)

At high temperatures (>400 ºC) the commonly used catalysts

based on TiO2–WO3–V2O5 tend to form N2O (reaction (14))

4NH3 + 4NO + 3O2 → 4N2O + 6H2O (14)

At still higher temperatures, NH3 may be oxidized to NO (reaction

(15), thus limiting the maximum NOx conversion:

4NH3 + 5O2 → 4NO + 6H2O (15)

In spite of many research efforts have been focussed on the

utilization of NH3 or urea as reductants, this technology has not been

successfully developed for light-duty vehicles, and one of the reasons is

the weight penalty of the additional urea/NH3 tank. In addition, a network of

urea/NH3 suppliers should be available to fill up the tanks.

Chapter 1

26

SCR of NOx with HC has been proposed as an alternative to the

urea/NH3 technology. The SCR of NOx with HC has been studied in detail

for HC of different nature (propene [36-39], octane [40], methanol [39],

ethanol [39], acetaldehyde [41] and decane [38]). However, as far as we

know, direct reduction of NOx with commercial diesel fuel has not been

reported, and one of the goals of this study is to do so. The main

advantage of diesel fuel with regard to some other reductants is that it is

already on board, and additional tanks and filling facilities are not required.

Using HC as a reductant, the reduction of NOx successfully

competes with the reduction of O2, even though the latter is present in a

large excess [24].

The reactivity of HCs in lean-NOx conversion depends on their

nature, the catalyst and the temperature [30]. At higher temperatures

competitive oxidation of HC becomes increasingly important, and then

most of the HC reductant is oxidized giving little opportunity for NOx

reduction. A feature of many lean-NOx reduction reactions is that there is

insufficient reduction capability on the surface to reduce NOx completely to

N2, and a significant amount of N2O is released as reaction by-product

[24].

The relative importance of N2O emission depends on the nature of

the catalyst surface concerned, the nature and concentration of reductant,

and the temperature as well as exhaust gas flow rates, etc [24].

The proposed catalysts for HC-SCR include PGMs [42-45], copper

[46], iridium [47], and silver [44] among others. Among noble metals,

platinum is very active for HC-SCR at low temperature (T < 300 ºC), and is

not significantly affected by the presence of H2O in the exhaust stream

[27, 48-50]. It has also been reported that platinum catalysts are resistant

to SO2 at sulphur levels similar to those in vehicle exhaust gas

(25 – 50 ppm) [51].

Bearing in mind the sensitivity of the HC-SCR reaction in noble

metals to the reaction conditions it is not surprising that different

mechanisms have been proposed [30]. In the case of platinum catalysts,

two alternative mechanistic schemes have been developed to explain the

Introduction

27

kinetics of the HC-SCR reaction. The first reaction scheme proposed was

the following, which is described for propene as model hydrocarbon [30]:

C3H6 → CxHy(ads) (16)

CxHy(ads) + O2→ I1(ads) (17)

I1(ads) + NO → CN(ads) (18)

CxHy(ads) + NO → CN(ads) (19)

I1(ads) + O2 → CO2 + H2O (20)

CN(ads) + NO/NO2 → N2O + N2 + CO2 (21)

CN(ads) + O2 → N2 + CO2 (22)

where CxHy (ads) represents adsorbed HCs and I1 represents some form

of an activated HC intermediate created after partial O2 oxidation.

The second mechanism for the HC-SCR reaction on platinum

catalysts involves the dissociation of NO and subsequent removal of

adsorbed oxygen by the reductant. This model was first presented by

Burch [30] and the main features of this mechanism are summarized

below:

NO ↔ NO(ads) (23)

NO(ads) ↔ N(ads) + O(ads) (24)

N(ads) + N(ads) ↔ N2(ads) (25)

NO(ads) + N(ads) ↔ N2O (26)

O2(g) ↔ 2O(ads) (27)

CxHy ↔ CxHy(ads) (28)

CxHy(ads) + (2x + y/2)O(ads) ↔ xCO2 + y/2H2O (29)

Chapter 1

28

In addition, at low temperatures the following reactions could

occur:

NO(ads) + N(ads) ↔ N2(ads) + O(ads) (30)

NO(ads) + NO(ads) ↔ N2O(ads) + O(ads) (31)

Zeolite-containing catalyst formulations can provide enhanced NOx

reduction due to their ability of maintaining a high concentration of HC in

the catalyst [30].

As described above N2O is one of the non-desired by-products

from SRC of NOx with HC. The main challenge of this thesis is to develop

and optimize a catalyst for N2O decomposition to be located after a SCR

catalyst used for NOx reduction in diesel exhausts, by using diesel fuel as

reductant.

1.6 N2O abatement.

The development of N2O abatement systems during nitric acid

production has been broadly studied by industry, research institutions and

universities. According to the process, different locations and catalysts

have been tested which operate at different temperatures. To apply that

knowledge at the end pipe exhaust treatment in diesel engines of light-duty

vehicles, the low temperature options will fit better.

From studies made for N2O abatement in nitric acid plants it can

be said that the low-temperature catalytic decomposition of N2O is

definitely more beneficial than SCR, because additional reducing agents

are not needed in this case. This reduces the cost of the reductant and

avoids the emissions involved (slip or undesired combustion products)

[2, 28]. However, the use of a catalyst in the low temperature zone is

connected with the requirement of high activity and resistance to inhibitors

always present in the exhaust gases, such as O2, H2O and residual NOx.

Catalytic decomposition of N2O was a subject of numerous studies

involving simple oxides [2, 52], perovskites [53], spinels [54-57], zeolites

[58], hydrotalcites [59, 60], mesoporous silicas [61] and supported

catalysts [62-64]. However, many of the catalysts proposed in the literature

do not show a good activity and stability in N2O conversion under realistic

Introduction

29

conditions of feed composition and space velocities. Laboratory results

often deviate from what is normally met within industrial practice and in

most of the cases severe stability test under realistic conditions have not

been taken into consideration.

Among all the catalyst reported in the literature working in the

range of low temperatures (200-600 ºC) in a batch reactor [3], including

pure oxides (CaO, Fe2O3, CuO, Rh2O3, IrO2, CoO, La2O3, pure and

doped CeO2), solid solutions (CoO in MgO), spinels (MCo2O4 where

M = Co, Cu, Ni, Zn, Ni+Cu), hydrotalcites (M-Al-CO3-HT where M=Co, Ni,

Cu), zeolites (Fe-ZSM-5), supported systems over alumina (Mn2O3, Rh),

and silica (Cr, Co, Ni, Fe)), we will focus on rhodium supported on pure or

doped ceria catalysts. Based on previous work, this system presents high

activity at laboratory and nitric acid pilot plant scale [65, 66] and also

shows resistance to H2O, NO and O2 atmospheres [67]. For all these

reasons, it was assumed that the RhOx/ceria formulations could be also

suitable catalysts for N2O decomposition in diesel exhausts, after the SCR

catalyst for NOx reduction with HCs.

1.7 Background of the research group.

This research work has been developed at the Carbon Materials

and Environment Research Group (MCMA) in the Inorganic Chemistry

Department that belongs to the University of Alicante. This group has wide

experience in environmental catalysis; it may be high-lighted several thesis

and numerous papers related with research on the pollutants abatement in

diesel engine exhausts, especially in NOx abatement and N2O

decomposition over different materials.

Regarding ceria and doped ceria, they were studied as catalysts

for soot combustion in two doctoral theses [68, 69], which have greatly

contributed to the knowledge and understanding of the physico-chemical

properties of ceria based supports used in the present thesis [70-80].

With respect to NOx abatement, another thesis [81] was developed

working with platinum, palladium and rhodium among other metals

supported over activated carbon and different zeolites for HC-SCR in NOx

abatement. Ptatinum-based catalysts were the most active at

low-temperature (200 - 250 ºC), showing thermal stability and resistance to

Chapter 1

30

poisons. The main drawback of ptatinum catalysts was the low selectivity

to N2 which was not improved with other metals [82, 83]. Later on it was

demonstrated that Pt/Beta zeolite achieved higher NOx conversions and

slightly higher selectivity to N2 due to the acidic properties of the support,

relatively open structure and thermostability [84, 85].

As a solution for the low selectivity to N2 of platinum-based

catalysts in C3H6-SCR, a system based on a double catalytic bed for

deNOx and deN2O was developed. The problem of that system was that,

under optimal conditions, the first bed operated at 200 ºC and the second

bed at 425 ºC [82]. Then, the gas heating would be required between both

beds and for a practical application it would result in unacceptable extra

fuel consumption. For this reason the desired conditions would be a dual

catalytic bed system that would operate in a similar temperature window.

To accomplish this requirement is another objective of the present thesis.

Focusing on N2O abatement, N2O decomposition was carried out

in our group with Fe-ZMS-5 catalysts using different reductants [86-88] and

also by direct N2O decomposition with rhodium supported on pure and

doped ceria [65]; proving the high activity for N2O decomposition related

with the redox properties of the support and concluding that the support

nature is crucial. A further work ended up with another thesis [66] focused

on the study of the N2O decomposition mechanism on rhodium/ceria

catalysts [89] and its application in nitric acid plants [90]. Rhodium,

palladium and platinum supported on alumina, bare and doped ceria with

lanthanum or praseodymium were also studied concluding that the best

catalytic composition was rhodium over praseodymium doped ceria [63]. In

addition the use of different gases (O2, H2O and NOx) in the gas stream

and their effect on N2O decomposition was also studied [90].

All these previous results allowed us to select RhOx/CeyPr1-yO2 as

a promising catalyst for N2O removal in diesel exhaust engines.

Introduction

31

1.8 Objectives.

Bearing in mind the above described background, the general

objective of the present thesis is to develop and optimize an effective

catalyst for N2O abatement in diesel engine exhausts, to be located after a

Pt/Beta zeolite SCR catalyst for NOx reduction with diesel fuel. To achive

this general goal, the specific objectives to be achieved are:

To study the effect of rhodium/ceria catalysts calcination conditions

in the catalyst physico-chemical properties and N2O decomposition

activity.

To study the effect of using different solvents (water, ethanol and

acetone) for rhodium loading over ceria-based supports in the

catalysts physico-chemical properties and N2O decomposition

activity.

To prepare, characterize and test for N2O decomposition at

laboratory honeycomb monolith-supported RhOx/Ce0.9Pr0.1O2

catalysts.

To prepare and test in a real diesel engine exhaust honeycomb

monolith-supported Pt/Beta zeolite catalysts for NOx reduction with

commercial fuel.

To test a dual bed consisting of a honeycomb monolith-supported

Pt/Beta zeolite catalyst followed by a honeycomb monolith-

supported RhOx/ceria catalyst in order to reduce NOx to N2 with

commercial diesel fuel. Diesel fuel will be introduced before the first

bed and both beds must operate at the same temperature.

Chapter 1

32

References.

[1] http://www.epa.gov/climatechange/ghgemissions/usinventoryreport

.html; Inventory of U.S. Greenhouse Gas Emissions and Sinks:

1990–2011-February 2013. Access in April 2013.

[2] J. Pérez-Ramírez, F. Kapteijn, K. Schöffel, J.A. Moulijn. Formation

and control of N2O in nitric acid production. Where do we stand

today?. Applied catalysis B: Environmental 44 (2003) 117.

[3] F. Kapteijn, J. Rodríguez-Mirasol, J.A. Moulijn. Heterogeneous

catalytic decomposition of nitrous oxide. Applied catalysis B:

Environmental 9 (1996) 25.

[4] A.R. Ravishankara, J.S. Daniel, R.W. Portmann. Nitrous Oxide

(N2O): The dominant ozone-depleting substance emitted in the

21st Century. Science 326 (2009) 123.

[5] H. Rodhe. A comparison of the contribution of various gases to the

greenhouse effect. Science 248 (1990) 1217.

[6] http://unfccc.int/kyoto_protocol/items/2830.php. Access May 2013.

[7] J. Haber, T. Machej, J. Janas, M. Nattich. Catalytic decomposition

of N2O. Catalysis Today 90 (2004) 15.

[8] G. Centi, S. Perathoner, F. Vazzana, M. Marella, M. Tomaselli,

M. Mantegazza. Novel catalysts and catalytic technologies for NO2

removal from industrial emissions containing O2, H2O and SO2.

Advances in Environmental Research 4 (2000) 325.

[9] T.J. Wallington, C.K.Lambert, W.C.Ruona. Diesel vehicles and

sustainable mobility in the U.S. Energy Policy 54 (2013) 47.

[10] Guidance for implementing section 141 of the energy

independence and security act of 2007. Federal Vehicle Fleets and

Low Greenhouse Gas-Emitting Vehicles. EPA-420-B-10-008.

February 2010.

Introduction

33

[11] Greenhouse gas emissions standards and fuel efficiency standards

for medium and heavy-duty engines and vehicles. Proposed Rule.

Docket No. EPA-HQ-OAR-2010-0162-0574,

http://federal.eregulations.us/rulemaking/document/EPA-HQ-OAR-

2010-0162-0002. Access March 2013.

[12] H. Shinjoh. Rare earth metals for automotive exhaust catalysts.

Journal of Alloys and Compounds 408–412 (2006) 1061.

[13] M.V. Twigg. Catalytic control of emissions from cars. Catalysis

Today 163 (2011) 33.

[14] J.P.A. Neeft, M. Makkee, J.A. Moulijn. Review article diesel

particulate emission control. Fuel Processing Technology 47

(1996) 1.

[15] H-K. Wang, C-Y Cheng, Y-C. Lin, K-S. Chen. Emission reductions

of air pollutants from a heavy-duty diesel engine mixed with various

amounts of H2/O2. Aerosol and Air Quality Research 12 (2012)

133.

[16] T.A. Nijhuis, A.E.W. Beers, T. Vergunst, I. Hoek, F. Kapteijn,

J.A. Moulijn. Preparation of monolithic catalysts. Catalysis

Reviews–Science and Engineering 43 (2001) 345.

[17] T. Vergunst, M.J.G. Linders, F. Kapteijn, J.A. Moulijn.

Carbon-based monolithic structures. Catalysis Reviews – Science

and Engineering 43 (2001) 291.

[18] C. Moreno-Castilla, A.F. Perez-Cadenas. Carbon-based

honeycomb monoliths for environmental gas-phase applications.

Materials 3 (2010) 1203.

[19] A.E.W. Beers, T.A. Nijhuis, N. Aalders, F. Kapteijn, J.A. Moulijn.

BEA coating of structured supports - performance in acylation.

Applied Catalysis A: General 243 (2003) 237.

http://federal.eregulations.us/rulemaking/document/EPA-HQ-OAR-2010-0162-0002

http://federal.eregulations.us/rulemaking/document/EPA-HQ-OAR-2010-0162-0002

Chapter 1

34

[20] K. Galejová, L. Obalová, K. Jirátová, K. Pacultová, F. Kovanda.

N2O catalytic decomposition – effect of pelleting pressure on

activity of Co-Mn-Al mixed oxide catalysts. Chemical Papers 63

(2009) 172.

[21] A. Bueno-López. Diesel soot combustion ceria catalysts. Applied

Catalysis B: Environmental (2013).

http://dx.doi.org/10.1016/j.apcatb.2013.02.033.

[22] M.V. Twigg, P.R. Phillips. Cleaning the air we breathe–controlling

diesel particulate emissions from passenger cars. Platinum Metals

Review 53 (2009) 27.

[23] S. Matsumoto. DeNOx catalyst for automotive lean-burn engine.

Catalysis Today 29 (1996) 43.

[24] M.V. Twigg, Progress and future challenges in controlling

automotive exhaust gas emissions. Applied Catalysis B:


[25] S. Roy, A. Baiker. NOx storage–reduction catalysts: from

mechanism and materials properties to storage–reduction

performance. Chemical Reviews 109 (2009) 4054.

[26] T. Ishihara, K. Anami, K. Takiishi, K. Yamada, H. Nishiguchi,

Y. Takita. Direct decomposition of NO on Cu-doped

La(Ba)Mn(In)O3 perovskite oxide under coexistence of O2, H2O

and SO2. Chemistry Letters 32 (2003) 1176.

[27] Z. Liu, S. I. Woo. Recent advances in catalytic deNOx science and

technology. Catalysis Reviews – Science and Engineering 48

(2006) 43.

[28] G. Centi, G.E. Arena, S. Perathoner. Nanostructured catalysts for

NOx storage–reduction and N2O decomposition. Journal of

Catalysis 216 (2003) 443.

http://dx.doi.org/10.1016/j.apcatb.2013.02.033

Introduction

35

[29] W. Bogner, M. kramer, B. Krutzsch, S. Pischinger, D. Voigtlander,

G. Wenninger, F. Wirbeleit, M.S. Brogan, R.J. Brisley,

D.E. Webster. Removal of nitrogen oxides from the exhaust of a

lean-tune gasoline engine. Applied Catalysis B: Environmental 7

(1995) 153.

[30] R. Burch. Knowledge and know‐how in emission control for mobile

applications. Catalysis reviews – Science and Engineering 46

(2004) 271.

[31] M.V. Twigg. H. Gandhi 1941-2010: Contributions to the

development and implementation of catalytic emissions control

systems. Platinum Metals Review 55 (2011) 43.

[32] W.S. Epling, L.E. Campbell, A. Yezerets, N.W. Currier, J.E. Parks.

Overview of the fundamental reactions and degradation

mechanisms of NOx storage/reduction catalysts. Catalysis reviews

– Science and Engineering 46 (2004) 163.

[33] S. Matsumoto, Y. Ikeda, H. Suzuki, M. Ogai, N. Miyoshi. NOx

storage-reduction catalyst for automotive exhaust with improved

tolerance against sulfur poisoning. Applied Catalysis B:


[34] A. Fritz, V. Pitchon. The current state of research on automotive

lean NO(x) catalysis. Applied Catalysis B: Environmental 13 (1997)

1.

[35] M. Koebel, M. Elsener, M. Kleemann. Urea-SCR: a promising

technique to reduce NOx emissions from automotive diesel

engines. Catalysis Today 59 (2000) 335.

[36] A. Bueno–López, M.J. Illán–Gómez, C. Salinas-Martínez de Lecea.

Effect of NOx and C3H6 partial pressures on the activity of

Pt–Beta–coated cordierite monoliths for deNOx C3H6–SCR.


Chapter 1

36

[37] M. Haneda, Y. Kintaichi, H. Shimada, H. Hamada. Selective

reduction of NO with propene over Ga2O3-Al2O3: effect of sol-gel

method on the catalytic performance. Journal of Catalysis 192

(2000) 137.

[38] V. Houel, D. James, P. Millington, S. Pollington, S. Poulston,

R. Rajaram, R. Torbati. A comparison of the activity and

deactivation of Ag/Al2O3 and Cu/ZSM-5 for HC-SCR under

simulated diesel exhaust emission conditions. Journal of Catalysis

230 (2005) 150.

[39] A. Obuchi, I. Kaneko, J. Oi, A. Ohi, A. Ogata, G.R. Bamwenda,

S. Kushiyama. A practical scale evaluation of catalysts for the

selective reduction of NO(x) with organic substances using a diesel

exhaust. Applied Catalysis B: Environmental 15 (1998) 37.

[40] K. Eränen, L.E. Lindfors, F. Klingstedt, D.Y. Murzin. Continuous

reduction of NO with octane over a silver/alumina catalyst in

oxygen-rich exhaust gases: combined heterogeneous and

surface-mediated homogeneous reactions. Journal of Catalysis

219 (2003) 25.

[41] S.J. Schmieg, B.K. Cho, S.H. Oh. Selective catalytic reduction of

nitric oxide with acetaldehyde over NaY zeolite catalyst in lean

exhaust feed. Applied Catalysis B: Environmental 49 (2004) 113.

[42] R. Burch, P.J. Millington. Selective reduction of nitrogen oxides by

hydrocarbons under lean-burn conditions using supported platinum

group metal catalysts. Catalysis Today 26 (1995) 185.

[43] J. Pérez-Ramírez, J.M. García-Cortés, M.J. Illán-Gómez,

F. Kapteijn, J.A. Moulijn, C. Salinas-Martínez de Lecea. Reduction

of NO by propene over Pt, Pd and Rh-based ZSM-5 under

lean-burn conditions. Reaction Kinetics and Catalysis Letters 69

(2000) 385.

Introduction

37

[44] R. Burch, J.P. Breen, F.C. Meunier. A review of the selective

reduction of NOx with hydrocarbons under lean-burn conditions

with non-zeolitic oxide and platinum group catalysts. Applied

Catalalysis B: Environmental 39 (2002) 283.

[45] C. Thomas, O. Gorce, F. Villain, G. Djega-Mariadassou. Influence

of the nature of the noble metal on the lean C3H6-assisted

decomposition of NO on Ce0.68Zr0.32O2-supported catalysts. Journal

of Molecular Catalysis A: Chemical 249 (2006) 71.

[46] U. Shibata, K. Shimizu, A. Satsuma, T. Hattori. Influence of

hydrocarbon structure on selective catalytic reduction on NO by

hydrocarbons over Cu-Al2O3. Applied Catalysis B: Environmental

37 (2002) 197.

[47] K.C. Taylor, J.C. Schlatter. Selective reduction of nitric oxide over

noble metals. Journal of Catalysis 63 (1980) 53.

[48] A.R. Vaccaro, J. Perez-Ramirez, J.M. García-Cortés, C. Salinas-

Martinez de Lecea, G.Mul, F. Kapteijn, J.A. Moulijn. Dual-bed

catalytic system for removal of NOx-N2O in lean-burn engine

exhausts. Environmental Challenges and Greenhous Gas Control

for Fossil Fuel Utilization in the 21st Century (2002) 229.

[49] M. Iwamoto, H. Yahiro, H.K. Shin, M. Watanabe, J. Guo, M. Konno,

T. Chikahisa, T.Murayama. Performance and durability of Pt-MFI

zeolite catalyst for selective reduction of nitrogen monoxide in

actual diesel engine exhaust. Applied Catalysis B: Environmental 5

(1994) L1.

[50] G. Garrigós-Pastor, S. Parres-Esclapez, A. Bueno-López,

M.J. Illán-Gómez, C. Salinas-Martínez de Lecea. Power-bench

demonstration of the Pt-catalyzed C3H6-SCR of NOx in a diesel

exhaust. Applied Catalysis A: General 354 (2009) 63.

[51] M.D. Amiridis, K.L. Roberts, C.J. Pereira. The selective catalytic

reduction of NO by propylene over Pt supported on dealuminated

Y zeolite. Applied Catalysis B: Environmental 14 (1997) 203.

Chapter 1

38

[52] P. Stelmachowski, F. Zasada, W. Piskorz, A. Kotarba, J.F. Paul,

Z. Sojka. Experimental and DFT studies of N2O decomposition

over bare and Co-doped magnesium oxide-insights into the role of

active sites topology in dry and wet conditions. Catalysis Today

137 (2008) 423.

[53] J.P. Dacquin, C. Lancelot, C. Dujardin, P. Da Costa, G. Djega-

Mariadassou, P. Beaunier, S. Kaliaguine, S. Vaudreuil, S. Royer,

P. Grange. Influence of preparation methods of LaCoO3 on the

catalytic performances in the decomposition of N2O. Applied

Catalysis B: Environmental 91(2009) 596.

[54] W. Piskorz, F. Zasada, P. Stelmachowski, A. Kotarba, Z. Sojka.

Decomposition of N2O over the surface of cobalt spinel: a DFT

account of reactivity experiments. Catalysis Today 137(2008) 418.

[55] F. Zasada, P. Stelmachowski, G. Maniak, J.F. Paul, A. Kotarba,

Z. Sojka. Potassium promotion of cobalt spinel catalyst for N2O

decomposition - accounted by work function measurements and

DFT modelling. Catalysis Letters 127 (2009) 126.

[56] P. Stelmachowski, G. Maniak, A. Kotarba, Z. Sojka. Strong

electronic promotion of Co3O4 towards N2O decomposition by

surface alkali dopants. Catalysis Communications 10 (2009) 1062.

[57] P. Stelmachowski, F. Zasada, G. Maniak, P. Granger, M. Inger,

M. Wilk, A. Kotarba, Z. Sojka. Optimization of multicomponent

cobalt spinel catalyst for N2O abatement from nitric acid plant tail

gases: laboratory and pilot plant studies. Catalysis Letters 130

(2009) 637.

[58] S. Sklenak, P.C.Andrikopoulos, B. Boekfa, B. Jansang,

J. Novakova, L. Benco, T. Bucko, J. Hafner, J. Dedecek,

Z. Sobalik. N2O decomposition over Fe-zeolites: structure of the

active sites and the origin of the distinct reactivity of Fe-ferrierite,

Fe-ZSM-5, and Fe-Beta. A combined periodic DFT and

multispectral study. Journal of Catalysis 272 (2010) 262.

http://www.sciencedirect.com/science/article/pii/S0920586107007808





http://www.sciencedirect.com/science/article/pii/S092058610800076X





http://www.chemweb.com/redirect?jid=SV10562&iid=0013000003&aid=14







Introduction

39

[59] L. Obalova, K. Karaskova, K. Jiratova, F. Kovanda. Effect of

potassium in calcined Co–Mn–Al layered double hydroxide on the

catalytic decomposition of N2O. Applied Catalysis B: Environmental

90 (2009) 132.

[60] K. Karaskova, L. Obalova, K. Jiratova, F. Kovanda. Effect of

promoters in Co–Mn–Al mixed oxide catalyst on N2O

decomposition. Chemical Engineering Journal 160 (2010) 480.

[61] L. Chmielarz, P. Kustrowski, M. Drozdek, M. Rutkowska,

R. Dziembaj, M. Michalik, P. Cool, E.F. Vansant. SBA-15

mesoporous silica modified with rhodium by MDD method and its

catalytic role for N2O decomposition reaction. Journal of Porous

Materials 18 (2011) 483.

[62] J.P. Dacquin, C. Dujardin, P. Granger. Surface reconstruction of

supported Pd on LaCoO3: consequences on the catalytic

properties in the decomposition of N2O. Journal of Catalysis 253

(2008) 37.

[63] S. Parres-Esclapez, M.J. Illán-Gómez, C. Salinas-Martínez de

Lecea, A. Bueno-López. On the importance of the catalyst redox

properties in the N2O decomposition over alumina and ceria

supported Rh, Pd and Pt. Applied Catalysis B: Environmental 96

(2010) 370.

[64] S. Parres-Esclapez, F.E. López-Suárez, A. Bueno-López,

M.J. Illán-Gómez, B. Ura, J. Trawczynski. Rh-Sr/Al2O3 catalyst for

N2O decomposition in the presence of O2. Topics in Catalysis 52

(2009) 1832.

[65] A. Bueno-López, I. Such-Basáñez, C. Salinas-Martínez de Lecea.

Stabilization of active Rh2O3 species for catalytic decomposition of

N2O on La-, Pr-doped CeO2. Journal of Catalysis 244 (2006) 102.

[66] S. Parres-Esclapez. N2O decomposition rhodium/ceria catalysts:

From principles to practical application. Tesis Doctoral. University

of Alicante (2011).













Chapter 1

40

[67] J. Oi, A. Obuchi, G.R. Bamwenda, A. Ogata, H. Yagita,

S. Kushiyama, K. Mizuno. Decomposition of nitrous oxide over

supported rhodium catalysts and dependency on feed gas

composition. Applied Catalysis B: Environmental 12 (1997) 277.

[68] I. Atribak. CeO2 y óxidos mixtos CeO2-ZrO2 como catalizadores de

la combustión de carbonilla en corrientes gaseosas con NOx y O2.

Tesis Doctoral. University of Alicante (2009).

[69] N. Guillén-Hurtado. Eliminación de carbonilla generada por

motores diésel mediante combustión catalizada por óxidos mixtos

basados en cerio. Tesis Doctoral. University of Alicante (2013).

[70] I. Atribak, A. Bueno-López, A. García-García. Combined removal

of diesel soot particulates and NOx over CeO2–ZrO2 mixed oxides.

Journal of Catalysis 259 (2008) 123.

[71] I. Atribak, A. Bueno-López, A. García-García. Further insights into

the key features of ceria–zirconia mixed oxides governing the

catalyzed soot combustion under NOx/O2. Topics in Catalysis 52

(2009) 2088.

[72] I. Atribak, A. Bueno-López, A. García-García, B. Azambre.

Contributions of surface and bulk heterogeneities to the NO

oxidationactivities of ceria–zirconia catalysts with composition

Ce0.76Zr0.24O2prepared by different methods. Physical Chemistry

Chemical Physics 12 (2010) 13770.

[73] M. Valencia, E. López, S. Andrade, M.L. Iris, N. Guillén-Hurtado,

V. Rico-Pérez, A. García-García, C. Salinas-Martínez de Lecea,

A. Bueno-López. Evidences of the cerium oxide-catalyzed DPF

regeneration in a real diesel engine exhaust. Topics in Catalysis 56

(2013) 452.

[74] N. Guillén-Hurtado, A. García-García, A. Bueno-López. Isotopic

study of ceria-catalyzed soot oxidation in the presence of NOx.


Introduction

41

[75] N. Guillén-Hurtado, V. Rico-Pérez, A. García-García, D. Lozano-

Castelló, A. Bueno-López. Three-Way Catalysts: past, present and

future. Dyna-Colombia 79 (2012) 114.

[76] N. Guillén-Hurtado, A. Bueno-López, A. García-García. Catalytic

performances of ceria and ceria-zirconia materials for the

combustion of diesel soot under NOx/O2 and O2. Importance of the

cerium precursor salt. Applied Catalysis A: General 437-438 (2012)

166.

[77] B. de Rivas, N. Guillén-Hurtado, R. López-Fonseca, F. Coloma-

Pascual, A. García-García, J.I. Gutiérrez-Ortiz, A. Bueno-López.

Activity, selectivity and stability of praseodymium-doped CeO2 for

chlorinated VOCs catalytic combustion. Applied Catalysis B:

Environmental 121-122 (2012) 162.

[78] N. Guillén-Hurtado, A. Bueno-López, A. García-García. Surface

and structural characterization of co-precipitated CexZr1-xO2

(0 < x < 1) mixed oxides. Journal of Materials Science 47 (2012)

3204.

[79] I. Atribak, N. Guillén-Hurtado, A. Bueno-López, A. García-García.

Influence of the physico-chemical properties of CeO2-ZrO2 mixed

oxides on the catalytic oxidation of NO to NO2. Applied Surface

Science 256 (2010) 7706.

[80] N. Guillén-Hurtado, I. Atribak, A. Bueno-López, A. García-García.

Influence of the cerium precursor on the physico-chemical features

and NO to NO2 oxidation activity of ceria and ceria-zirconia

catalysts. Journal of Molecular Catalysis A: Chemical 323 (2010)

52.

[81] J.M García-Cortés. Solución a la problemática contaminante de

motores diesel. Estudio fundamental de la reducción de óxidos de

nitrógeno con propeno sobre catalizadores de platino. Tesis

Doctoral. University of Alicante (2003).

Chapter 1

42

[82] J. Pérez-Ramirez, J.M. García-Cortés, F. Kapteijn, M.J. Illán-

Gómez, A. Ribera, C. Salinas-Martínez de Lecea, J.A. Moulijn.

Dual-bed catalytic system for NOx–N2O removal: a practical

application for lean-burn deNOx HC-SCR. Applied Catalysis B:


[83] J.M. García-Cortés, J. Pérez-Ramírez, M.J. Illán-Gómez,

F. Kapteijn, J.A. Moulijn, C. Salinas-Martínez de Lecea.

Comparative study of Pt-based catalysts on different supports in

the low-temperature de-NOx-SCR with propene. Applied Catalysis

B: Environmental 30 (2001) 399.

[84] J.M. García-Cortés, J. Pérez-Ramírez, J.N. Rouzaud,

A.R. Vaccaro, M.J. Illán-Gómez, and C. Salinas-Martínez de

Lecea. On the structure sensitivity of deNOx HC-SCR over Pt-Beta

catalysts. Journal of Catalysis 218 (2003) 111.

[85] A. Bueno-López, D. Lozano-Castelló, I. Such-Basáñez,

J.M. Garcia-Cortes, M.J. Illán-Gómez, C. Salinas-Martínez de

Lecea. Preparation of Beta-coated cordierite honeycomb monoliths

by in situ synthesis utilization as Pt support for NOx abatement in

diesel exhaust. Applied Catalysis B: Environmental 58 (2005) 1.

[86] M.N. Debbagh-Boutarbouch, J.M. García-Cortés, M.S. El Begrani,

C. Salinas-Martínez de Lecea, J. Pérez-Ramírez. Catalytic

conversion of N2O over FeZSM-5 zeolite in the presence of CO

and NO. Applied Catalysis B: Environmental 54 (2004) 115.

[87] M.N. Debbagh, C. Salinas-Martínez de Lecea, J. Pérez-Ramírez.

Catalytic reduction of N2O over steam-activated FeZSM-5 zeolite.

Comparison of CH4, CO, and their mixtures as reductants with or

without excess O2. Applied Catalysis B: Environmental 70 (2007)

335.

[88] M.N. Debbagh, A. Bueno-López, C. Salinas-Martínez de Lecea,

J. Pérez-Ramírez. Kinetics of the N2O + CO reaction over

steam-activated FeZSM-5. Applied Catalysis A: General 327

(2007) 66.

Introduction

43

[89] S. Parres-Esclapez, I. Such-Basáñez, M.J. Illán-Gómez,

C. Salinas-Martínez de Lecea, A. Bueno-López. Study by isotopic

gases and in situ spectroscopies (DRIFTS, XPS and Raman) of the

N2O decomposition mechanism on Rh/CeO2 and Rh/ɣ-Al2O3



Lecea, A. Bueno-López. Preparation and characterization of

ɣ-Al2O3 particles-supported Rh/Ce0.9Pr0.1O2 catalyst for N2O

decomposition in the presence of O2, H2O and NOx. International

Journal of Greenhouse Gas Control 11 (2012) 251.

CHAPTER 2

Characterization techniques and

catalytic tests

This chapter describes the characterization techniques used to

obtain information about the physico-chemical properties of the

catalysts. These techniques are N2 adsorption at -196 ºC, X-Ray

Diffraction (XRD), Raman spectroscopy, X-ray Photoelectron

Spectroscopy (XPS), Temperature Programmed Reduction with H2

(H2-TPR), Transmission Electron Microscopy (TEM), Scanning

Electron Microscopy (SEM), and Thermogravimetry-Mass

Spectroscopy (TG-MS). The experimental set-ups and reaction

conditions used in the catalytic tests are also described in this chapter.

Characterization and experimental techniques

47

This chapter describes the characterization techniques and

experimental methods used in this thesis to obtain information about the

physico-chemical properties of the catalysts together with the experimental

details about the set-ups used to perform catalytic tests. The details about

the preparation of the different catalysts studied throughout this thesis

have been described in the experimental section of each chapter devoted

to the presentation of experimental results, because most details of these

preparation procedures are specific for each particular chapter.

2.1 Characterization techniques.

Catalysts are defined by composition, structure, surface and

textural properties that are related with their catalytic activities. In order to

understand and determine their physico-chemical properties several

characterization techniques were used in this work.

N2 adsorption at -196 ºC was used to determine

(Brunauer-Emmett-Teller) BET surface areas. Crystalline properties were

studied by X-Ray Diffraction (XRD) and Raman spectroscopy. The atomic

surface composition and metal oxidation states were determined by X-ray

Photoelectron Spectroscopy (XPS). Temperature Programmed Reduction

with H2 (H2-TPR) was employed to study the reducibility of the samples.

The metal particle size was determined by Transmission Electron

Microscopy (TEM). Finally, on supported catalysts, the active phase

distribution on the solid substrate was observed by Scanning Electron

Microscopy (SEM). Thermogravimetry-Mass Spectroscopy (TG-MS) was

used to study the decomposition of the catalysts precursors.

A brief description of each technique, the experimental equipment

used and the operation conditions is presented below.

2.1.1 N2 adsorption at -196 ºC.

Many of the most popular methods for determining the surface

area of powders and porous materials depend on the measurement of gas

adsorption. The appearance of Langmuir's comprehensive review of the

nature of adsorption [1] stimulated several investigators to consider the

possibility of using gas adsorption for surface area determination.

Chapter 2

48

Nowadays the physical adsorption (or physisorption) of gases is one of the

most used techniques to study the porous texture of solids. The first

significant advances were made by Brunauer and Emmett [2, 3] and their

work prepared the way for the development of the Brunauer-Emmett-Teller

(BET) theory in 1938 [4].

An adsoption isotherm is a successive measure of the adsorbed

quantities of a certain gas (N2 in this case) by a solid depending on the

equilibrium pressure when the solid and gas are in contact in a closed

volume at a constant temperature. The amount of adsorbed gas by the

solid and the pressure at constant temperature is related to the porosity of

the solid, allowing the calculation of the solid surface area. The adsorbate

must meet certain properties being chemically inert, having a saturation

pressure relatively high at work temperature and presenting a spherical

shape to minimize uncertainty in calculating the cross section of the

molecule. Then the following molecules can be used for this approach: N2,

CO2, Ar, He, CH4 and H2O, but the most common absorbate which has

been also used in this work is N2, working at -196 ºC.

Over the past 50 years the BET method has become an extremely

popular method for determining the surface area of adsorbents, catalysts

and various other finely divided and porous materials.

Two stages are involved in the evaluation of the surface area by

the BET method from physisorption isotherm data. First, it is necessary to

construct the BET plot and from it to derive the value of the monolayer

capacity, nm. The second stage is the calculation of the specific surface

area, a(BET), from nm, and this requires a knowledge of the average area,

am, occupied by each molecule in the completed monolayer (i.e. the

molecular cross-sectional area) [5].

The BET equation is conveniently expressed in the linear form:

0mm

0

0

P

P

Cn

1C

Cn

1

P

P1 n

P

P

(1)


49

where P is the pressure, P0 is the saturation pressure, n is the number of

moles adsorbed, nm is the number of moles adsorbed in the monolayer

and C is a parameter related to the heat of adsorption.

In the original work of Brunauer, Emmett and Teller [4] and

subsequent studies [6, 7] it was found that Type II nitrogen isotherms on

various adsorbents gave linear BET plots over the approximate range

P/Po = 0.05-0.30.

The BET plot of

0

0

P

P1 n

P

P

versus 0P

P should be a straight line,

from which the values of C and nm can be calculated by resolving the

simultaneous equations of the slope, s:

mn

1Cs

(2)

and intercept, i:

Cn

1i

m

(3)

From the value of nm the specific area of the material is determined

by the following equation:

21amm 10NanS (4)

where S is the apparent surface area of the material (m2/g), am is the area

occupied by one adsorbate molecule (nm2/molecule; in the case of N2 at

-196 ºC is 0.162 nm2) and Na is the Avogadro's number

(6.0231023

molecules/mol).

In this thesis the surface areas of the catalysts studied have been

determinated applying the BET method (equation 1) to the data obtained

from N2 adsorption isotherms at -196 ºC, using a volumetric equipment

Chapter 2

50

Quantachrome (modele Autosorb-6B) which is shown in Figure 2.1 and is

available in the laboratories of the MCMA group at the University of

Alicante. This system includes two units, a degassing unit and an

adsorption unit. The degassing unit consists of a vacuum system and six

independent furnaces. Before the N2 adsorption isotherms were measured,

the samples were degassed for 4 hours at 150 ºC. The degassed samples

were moved to the physical adsorption unit, where all subsequent

experimental steps take place. The measurement process is controlled by

a computer. The steps include a vacuum step and a filling step and

different pressures of adsorbate (0 < P/P0 < 1 for N2 at -196 °C) are

supplied. The system is able to calculate the adsorbed gas volume at each

relative pressure.

Figure 2.1. Degassification (left) and adsorption (right, Quantachrome Autosorb-6B)

units used to obtain N2 adsorption isotherms at -196 ºC.

2.1.2 X-Ray Diffraction (XRD).

A first approximation of a crystal can be considered as an ordered

and periodic atoms aggrupation. X-rays are electromagnetic radiation with

wavelengths in the range 0.5 - 2.5 Å. Since this is of the same order of

magnitude as the interatomic distances in solids, X-rays are frequently

used to study the internal (crystalline) structure of materials [8]. When a

solid is reached by an X-ray beam, the X-rays are reflected, and in most


51

cases the reflected waves interfere one to each other being annulled.

However if the solid presents crystalline order, in certain directions the

waves are reinforced to form a new wave front. This constructive

interference is known as diffraction. These preferential directions are

directly related with the lattice geometry, while the intensity of the diffracted

beams depends on the type and form of atoms aggrupation in the crystal.

Therefore, the study of the geometry and intensity of the diffraction allow

the determination of the crystal structure [9].

XRD is based on the optical interferences produced when a

monochromatic radiation crosses a split of similar thickness to the

wavelength of the radiation. The X-ray wavelength is around few

Angstroms (Å), that is the same order than interatomic distances of the

crystal structures. When the solid to be analyzed is irradiated with X-rays,

these are diffracted with angles that depend of the interatomic distances

according to the Bragg’s law (see Figure 2.2):

send2n hkl (5)

where dhkl is the distance between planes, is the radiation wavelength

(nm) used to obtain the diffractogram, θ is the angle of the diffracted X-ray

beam with the surface normal (rad) and n is an integer number [10].

Figure 2.2. Diffraction (i.e. constructive interference of the scattered X-rays) will occur if

the Bragg’s equation (equation. 5) is fulfilled.

Chapter 2

52

The representation of the radiation intensity after interaction with

the sample as a function of incidence angle (usually 2θ) is called

diffractogram and is characteristic of each crystal. XRD provides both

qualitative and quantitative information about crystalline solids and

indicates the structure and position of atoms within the structure [9].

From the position of the reflections collected in the X-ray

diffractogram and Bragg’s equation (5), the interatomic distance dhkl can be

determined. The parameter dhkl depends on the crystalline system and it is

related with the lattice parameters through diverse equations. In the case

of ceria, which crystallizes in a cubic fluorite structure, the diffraction peak

corresponding to the plane (111) is the most intense reflection and is used

to calculate the lattice parameter. For this approach firstly the d111 is

calculated by the following equation:

111

111sen2

d

(6)

From d111 value the lattice parameter (a) can be determined by the

following crystallographic relation [9]:

2

222

2

hkl a

)lkh(

d

1 (7)

where (a) is the lattice parameter (nm) and h,k,l are the Miller’s indices of a

specific crystallographic plane ((111) in this case).

The crystal size can also be calculated with this technique from the

full-width at half maximum (FWHM) of the (111) peak. The Scherrer’s

equation (equation 8) can be used to determine the average size of the

crystals (D) for a crystalline solid [9]:

cos

KD (8)

where K is the Scherrer’s constant or shape factor (0.94 has been

proposed for ceria and praseodymium doped ceria [11]), is the radiation

wavelength used to obtain the diffractogram (in our particular case


53

= 0.15406 nm), β is the FWHM expressed in radians and θ is the

diffraction angle.

Nevertheless, Scherrer’s equation, despite being widespread, does

not take into account factors as crystalline perfection, crystal curvature,

tensions, etc. that also affects FWHM [9, 11]. When these factors affect

FWHM, Williamson-Hall’s method is used to obtain a more accurate crystal

size value using the following expression:

cosθd

sinθΔd4

cosθD

0.9βββ StrainSizeTotal

(9)

where βTotal is the FWHM, is the wavelength of the incident radiation, D is

the crystal size, θ is the diffraction angle, and ∆d is the difference of the d

spacing corresponding to a typical peak. The Williamson-Hall’s method

separates the contribution of the crystal size and strain (caused by atoms

that are not in their ideal positions in a non-uniform crystal) in the FWHM

[11]. When βTotalcosθ is plotted versus 4sinθ the crystal size can be

calculated from the intercept value and the tension value from the slope of

the straight line [12].

This technique has been used in this study to elucidate the

crystalline structure of both bare and doped ceria catalysts, to determine

their lattice parameter and to calculate the average size of the crystals,

using the above mentioned Scherrer’s and Williamsom-Hall’s equations,

(8) and (9) respectively.

The equipment used is a Bruker D-8 Advance with Göebel mirror

(for no flat samples), with a high temperature chamber (up to 900 ºC), and

an X-ray generator Kristalloflex K 760-80 F (Power: 3 KW, Tension:

20 - 60 KV and Current: 5 - 80 mA). All the measure conditions are

controlled by the computer system. This diffractometer is available in the

scientific instrumentation area of the University of Alicante [10].

X-ray diffractograms were recorded using CuKα radiation

( = 1.5406 Å), between 10º and 80º (2θ) with steps of 0.02º and a step

time of 3 sec. The working conditions of the diffractometer were a power of

1600 KW, a tension of 40 KV and a current of 40 mA.

Chapter 2

54

Figure 2.3. X-ray diffractometer Bruker D-8 Advance.

2.1.3 Raman spectroscopy.

This technique is a tool to study solid materials, especially with

crystalline structure. The information obtained by Raman spectroscopy

complements that obtained by X-ray diffraction and both techniques are

non-destructive for the material analyzed.

Raman spectroscopy analysis is carried out by the interaction of a

monochromatic light beam with a material. The oscillating electric field of

the incident radiation (with frequency ν0) causes an oscillation in the

electronic density of the molecule, originating the Rayleigh dispersion (with

the same frequency as the incident radiation, elastic photon-molecule

collision) and the Raman dispersion (with a frequency change, inelastic

photon-molecule collision).

Raman dispersion is a small fraction of all dispersed radiation

which provides information about the sample analyzed, where in the

Raman dispersion, the photon is dispersed with a different frequency than

the incident ν0 due to the energy transfer between the molecule and the

photon. Raman dispersion is the result of electron excitement momentarily

in a prohibited quantum levels, but never achieving an excited electronic

level, being the relaxation from the prohibited quantum level to permitted

quantum levels very fast (see Figure 2.4). The difference between the new


55

frequencies formed (Raman bands) and the original radiation frequency is

characteristic of the irradiated molecule and is numerically equal to some

of its vibration or rotation frequencies [13].

ν ν

ν0 - νk ν0 + νk

Figure 2.4. Energy diagram showing energetic transitions leading to IR absortion,

Rayleigh and Raman dispersion.

Raman dispersion can be Stokes or anti-Stokes. Stokes Raman

dispersion is produced when the dispersed photon has lower frecuency

than indicent one (ν0), the frequency of the dispersed photon being

(ν0 – νk). On the contrary, anti-Stokes Raman dispersion is produced when

the dispersed photon has higher energy than the incident frequency, being

(ν0 + νk). The new frequencies (–νk and +νk) are the Raman frequencies

and each material has a set of values νk characteristic of each structure.

Figure 2.4 shows an energetic diagram that compiles the energetic

transitions above commented.

In regular Raman spectroscopy, anti-Stokes lines are much

weaker than Stokes lines and usually they are not used, because both

types of lines give the same information. Stokes lines are usually placed in

the positive part of the Raman spectra plot, and the Reyleigh band will be

the beginning of the axis (see Figure 2.5).

Chapter 2

56

Figure 2.5. Position of Stokes lines in the Raman shift axis.

As mentioned, Raman dispersion is an extremely weak

phenomenon, so this technique is not adequate for trace analysis. Another

inconvenience is the potential fluorescence of the samples. This

phenomenon, even in weakly fluorescent samples, masks the Raman

signal and hinders the analysis of the Raman spectrum [13].

In this thesis Raman spectroscopy has been used to obtain

information about the structure of bare and doped ceria catalysts. Two

different equipments available in the scientific instrumentation area of

University of Alicante have been used.

In Chapters 3 and 4, Raman spectra were recorded in a Jobin

Yvon Horiba Raman dispersive spectrometer with a variable-power He-Ne

laser source (632.8 nm) and using a confocal microscope. The detector is

a CCD Peltier-cooled detector (Figure 2.6).


57

Figure 2.6. LabRam (Jobin-Ivon) Raman dispersive spectrometer, with a confocal

microscope.

On the other hand, a Bruker RFS 100/S Fourier transform Raman

spectrometer with a variable-power Nd:YAG laser source (1064 nm) was

used in Chapter 5 (Figure 2.7), which is particularly suitable for the

analysis of fluorescent samples. This device uses a liquid nitrogen cooled

Ge detector.

Figure 2.7. FT-Raman (Bruker RFS/100) spectrometer with coupled microscope.

2.1.4 X-ray Photoelectron Spectroscopy (XPS).

X-ray Photoelectron Spectroscopy is known as XPS and also as

Electron Spectroscopy for Chemical Analysis (ESCA). This technique is

Chapter 2

58

able to obtain the chemical composition of a solid surface, analyzing up to

few nanometers depth, and providing information about the oxidation state

and environment of the different elements. Most elements can be detected

by XPS, except hydrogen [14].

In the material analyzed each electron has unique identification

information depending of the surrounding environment. The goal of this

technique is to find out from which atom they are coming. To free the

electron from the nucleus attractive force it is necessary to excite them by

exposure to X-ray bombardment. Magnesium and aluminium are the

normal X-ray sources that emit at 1253.6 and 1486.6 eV respectively,

bringing enough energy to free the electron from the nucleus.This process

is known as photoemission principle, and is shown in Figure 2.8. Once an

electron is removed, another electron from a more external layer jump

down to cover the gap, releasing the energy corresponding to the

difference between the initial and final orbitals. This energy is characteristic

of each element.

Figure 2.8. Photoemission principle. Process of electron excitation by an X-ray beam

and the subsequent release of the photoelectron.

According to the principle of energy conservation the energy of the

incident photon must be the sum of the binding and kinetic energies of the

removed electron (equation 10):


59

Bindingkinetic E E hv (10)

The XPS analyses are carried out under high vacuum and the

electrons emmited are drived in these conditions to an electron analyzer to

determine the Binding Energy (BE). As mentioned before, the BE is

characteristic for an element and the number of photoelectrons emmited is

proportional to the concentration of each element in a surface. XPS

spectrum obtained by plotting Intensity (which depends of the amount of

electrons collected) versus their BE (in eV) provides information about the

atomic surface composition and concentration [14, 15].

In the current study, the rhodium, cerium and praseodymium

oxidation states have been estimated for the different catalysts, together

with the Ce/Pr surface ratios. The Rh 3d photoelectron spectra of the

catalysts show two peaks, corresponding to the 3d5/2 and 3d3/2 transitions

(around 309 and 313 eV, respectively). Both peaks provide similar

information about the oxidation state of rhodium. As reported in the

literature, the Rh 3d5/2 peak appears at 307.0–307.5 eV for Rh0, at about

308 eV for Rh+, and from 308.3 to 310.5 eV for Rh

3+ [12, 14]. The

percentage of each rhodium species was determinated by peak

deconvolution analysis (Figure 2.9).

300 305 310 315 320

Inte

nsity (

a.u

.)

Binding Energy (eV)

Rh0

Rh3+

Figure 2.9. Rh 3d5/2 spectrum of a representative sample RhOx/CeyPry-1O2

The Ce3+

percentage (with regard to total surface cerium) was

estimated following the method proposed elsewhere [12, 16, 17] and the

Chapter 2

60

Pr3+

percentage (with regard to total surface praseodymium) by using the

semi-quantitative method proposed by Borchert et al. [18].

875 885 895 905 915 925

Inte

nsity (

a.u

.)

Binding Energy (eV)

v0

v

v'

v''

v'''

u0

u

u'

u''

v'''

Ce 3d

Figure 2.10. Ce 3d spectrum of a representative sample RhOx/CeyPry-1O2

The proportion of Ce3+

cations with regard to the total cerium was

calculated as the ratio of the sum of the intensities of the u0, u’, v0, and v’

bands to the sum of the intensities of all the bands (see in Figure 2.10

spectra of Ce 3d) [16].

uv area

'uu'vv area100%

CeCe

Ce 0043

3

(11)

In comparison to cerium oxide, less literature is available on XPS

analysis of praseodymium oxide. Pr 3d spectra have a rather similar shape

like the Ce 3d spectra of Ce3+

/Ce4+

compounds (Figure 2.11) and also

allow studying redox processes. However, the precise analysis of the

oxidation states by deconvolution of the spectra into components has been

less elaborated until now. Three spin-orbit split doublets labeled as a/b,

a’/b’, and a’’/b’’ were found. The 3d3/2 sub-level presents an additional

feature labeled “t”, which can be explained by a multiplet effect.


61

920 930 940 950 960 970

Inte

nsity (

a.u

.)

Binding Energy (eV)

a

a'

a''

b

b'

tb''

Figure 2.11. Pr 3d spectrum from representative sample RhOx/CeyPr1-yO2

It has not been reported the direct possibility to determine the

amounts of praseodymium ions in the different oxidation states from the

relative peak areas. Doublet a’’/b’’ is absent in the spectra of clean Pr2O3

and can therefore be assigned to Pr4+

ions. However, doublets a/b and

a’/b’ are present in Pr2O as well as in PrO2 and cannot be assigned to a

specific oxidation state. Semi-quantitative estimation of the amount of Pr3+

ions was done using the following expression reported by Borchert et al.

[18].

'a area

''a area

28.0

11100%

PrPr

Pr43

3

(12)

XPS spectra were recorded with two different equipments available

in the scientific instrumentation area of University of Alicante. XPS spectra

of Chapter 3 were obtained with a VG‐Microtech Multilab 3000

spectrometer (Figure 2.12) equipped with a hemispherical electron

analyzer with 9 channeltrons (with energy of 2 ‐ 200 eV step) and Mg and

Al X‐ray radiation sources. In our case, Mg Kα radiation (1253.6 eV) was

used, with the detector in constant energy mode and a pass energy of

50 eV.

Chapter 2

62

The pressure in the analysis chamber was maintained at

5∙10‐10 mBar. The BE scales and Auger kinetic energy (KE) were

established using the C1s transition at 284.6 eV as reference [15].

A high temperature-pressure cell with controlled atmosphere is

available in this device for in situ pre-treatments. This cell has been used

to pre-treat selected catalysts with a 1000 ppm N2O/He flow at 225 ºC

(typical N2O decomposition conditions in laboratory experiments) before

the XPS analysis.

Figure 2.12. X-ray Photoelectron spectrometer VG-Microtech Multilab 3000.

A fully automated K-Alpha spectrometer from Thermo-Scientific

(Figure 2.13) was used in Chapter 4. It has a source of electrons and ions

for automated charge balancing and an argon ion source for high-precision

etching with a beam size of less than 300 μm. The high performance

colour optical system allows precise alignment of analysis position [15]. All

spectra were collected using Al-Kα radiation (1486.6 eV),

monochromatized by a twin crystal monochromator, yielding a focused

X-ray spot with a diameter of 400 μm, at 3 mA × 12 kV. The alpha

hemispherical analyzer was operated in the constant energy mode and

pass energy of 50 eV. Charge compensation was achieved with the

system flood gun that provides low energy electrons and low energy argon

ions from a single source.


63

Figure 2.13. Fully automated X-ray Photoelectron spectrometer K-Alpha from

Thermo-Scientific

2.1.5 Temperature Programmed Reduction with H2 (H2-TPR).

Temperature Programmed Reduction by H2 is a technique to study

the reduction of solids. To carry out the experiments a H2 flow was used

and the reduction was carried out in temperature programmed conditions,

i.e., increasing the temperature using a constant heating rate. The

equipment registers thermic conductivity changes of the gas stream that

has interacted with the sample. In this study, this technique allows to

determine the reducibility of the catalysts, which will be related to their

catalytic performance.

These tests have allowed the analysis of the catalysts prepared

(RhOx/CeyPr1-yO2) and to observe how praseodymium doping ceria, and

the synthesis conditions affect the redox properties of the ceria catalysts.

H2-TPR experiments were carried out in a Micromeritics Pulse ChemiSorb

2705 device (Figure 2.14), consisting of a tubular quartz reactor (inner

diameter 5 mm) coupled to a Thermal Conductivity Detector (TCD)

analyzer to monitor H2 consumption. A cold trap was placed before the

TCD, consisting of a mixture of isopropyl alcohol and liquid nitrogen

(temperature −89 ºC). This equipment was available in the MCMA group

laboratories at the University of Alicante.

Chapter 2

64

Most experiments were carried out with 20 mg of fresh catalyst,

which were pre-treated in situ at 500 ºC for 1 hour in a 50 mL/min flow of

5 vol.% O2 in He. Once cold, the flow gas was switched to 40 mL/min of

5 vol.% H2 in Ar and the temperature was increased at 10 °C/min up to

1050 °C. Taking into account that the active phases used in chapter 5 are

diluted by the cordierite support, 60 mg of fresh catalysts were used in this

case.

Figure 2.14. Micrometitics Pulse Chemisorb 2705 equipment.

2.1.6 Transmission Electron Microscopy (TEM).

In a TEM microscope a thin sample (maximum thickness 100 nm,

to be transparent to the electrons) is irradiated with a high energy electron

beam (200 keV in our case). As shown in Figure 2.15, part of these

electrons are transmitted, other are scattered and some of them result in

interactions that produce different phenomena such as light emission,

secondary and Auger electrons, X rays, etc. The electrons passing through

the sample may scatter due to interaction with the sample (elastic or

inelastic) or may not experience any trajectory change. Elastically

scattered electrons can be used to record diffraction images and no

scattered electrons project direct images of the dispersed material.


65

Inelastically scattered electrons are responsible for the background noise

in electron microscopy.

Figure 2.15. Scheme of the electron beam interaction with a sample.

The TEM microscope uses the transmission/scattering of the

electrons to form images, the diffraction of the electrons to obtain

information about the crystal structure and the emission of characteristic

X-rays to determine the elemental composition of the sample [19]. It is

possible to determine the morphology of the sample (size and position of

the microcrystals or the particles observed) and also the crystallography

(position of the crystal planes, study of the defects, and etc; as well as

chemical composition of the material).

TEM characterization was performed using a JEOL (JEM-2010)

microscope, equipped with a detector (Oxford, model INCA Energy TEM

100) for microanalysis Energy-Dispersive X-ray Spectroscopy (EDS). A

few droplets of an ultrasonically dispersed suspension of the catalyst in

ethanol were placed in a copper grid with lacey carbon film and dried at

ambient conditions. This microscope (Figure 2.16) is available in the

scientific instrumentation area of University of Alicante.

Chapter 2

66

Figure 2.16. JEOL (JEM-2010) microscope.

In this work, this technique was used to observe the catalysts at

nanometer scale and enable analyzing the crystal size and morphology of

the samples. The samples composition was measured by EDS.

2.1.7 Scanning Electron Microscopy (SEM).

SEM provides morphological and topographical information of the

solids surface. To obtain an image of the surface, this is scanned with a

very thin electron beam with high energy that gives several types of signals

as retrodispersed electrons, secondary electrons, etc [20].

Catalysts were characterized by SEM in a Hitachi S-3000N

microscope with a secondary electrons detector, a retrodispersed

electrons detector and X-ray detector (XFlash 3001 de Bruker) for

microanalysis (EDS) and chemical mapping, this equipment is available in

in the scientific instrumentation area of University of Alicante (Figure 2.17).


67

Figura 2.17. SEM Hitachi S-3000N microscope.

2.1.8 Thermobalance coupled to a Mass Spectrometer (TG-MS).

Thermogravimetry (TG) is a technique based in the weight change

(gain or loss) during a thermal treatment performed in a controlled

atmosphere [21]. In addition, the detection and analysis of emitted gases

provides complementary information about the process occurring during

the thermal treatment, and a mass spectrometer has been used in this

study for this purpose.

The TG-MS system (Figure 2.18) used in this study is available in

the scientific instrumentation area of University of Alicante and consists of

a thermobalance TG–DTA (Mettler Toledo model TGA/SDTA

851e/LF/1600) coupled to a mass spectrometer (Pfeiffer Vacuum model

Thermostar GSD301T). This combined thechnique has been used to study

the decomposition of 10 mg of either Ce(NO3)3·6H2O or Ceria-impregnated

Rh(NO3)3 under 100 mL/min flow of synthetic air, with a heating rate of

50 °C/min.

Figura 2.18. TG-MS system.

Chapter 2

68

2.2 Catalytic tests at different scales.

As it has been previously mentioned, the details about the

preparation of the different catalysts studied in this thesis have been

described in the experimental section of each chapter devoted to the

presentation of experimental results. These catalysts include both powder

and cordierite honeycomb monolith-supported catalysts. Honeycomb

monoliths of different size have been used and are shown in Figure 2.19,

including cylindrical substrates of 1, 2.3 and 14 cm of diameter, which are

referred to as small, medium and full-size monoliths, respectively. In order

to study the N2O decomposition over catalysts of such different shape and

size specific experimental set-ups have been required, and these set-ups

are described in this section.

Figure 2.19. Different size cordierite honeycomb monoliths used in this thesis.


69

2.2.1 Powder and small size honeycomb monolith catalysts.

N2O decomposition experiments were performed at laboratory

scale with powder and small size honeycomb monolith catalysts. These

experiments consisted of point-by-point isothermal reactions, increasing

the temperature with intervals of 25 ºC, which were extended until the

steady state was reached. The gas composition was analyzed by a

HP 6890 gas chromatograph equipped with a TCD and two columns

(Porapak Q, for N2O and CO2, and Molecular Sieve 13X, for O2 and N2 and

CO). In the particular case of Chapter 5, a chemiluminescence analyzer

(Signal 400VM) was additionally used for NO and NO2 monitoring. This

system is available in the laboratories of the MCMA group at the University

of Alicante (see Figure 2.20).

N2O decomposition tests of Chapters 3 and 4 were performed with

powder catalysts in a U-shaped fix-bed quartz reactor, located in a vertical

furnace at atmospheric pressure, with a 100 mL/min flow

(GHSV = 2000 h−1

) of 1000 ppm N2O in He, using 100 mg of catalyst

(Figure 2.20).

In Chapter 3 CO oxidation experiments were also performed in the

same set-up using a 100 mL/min flow of 1000 ppm CO/5 % O2/He and

100 mg of powder catalyst (GHSV = 42000 h−1

).

Most N2O decomposition tests of Chapter 5 were performed with

catalysts supported in small size honeycomb monoliths, in a cylindrical

reactor (cylindrical shape with 1 cm inner diameter) at atmospheric

pressure. The total flow rate was 500 mL/min (GHSV = 27000 h−1

). N2O

decomposition tests were carried out under different gas streams:

1000 ppm N2O/He or 1000 ppm N2O/1000 ppm NO/5 % O2/0.6 % H2O/He

(Figure 2.20).

Particular experiments performed in Chapter 5 with a powder

catalyst were carried out in the same reactor to that used to test the

monoliths, and the amount of powder catalyst introduced in the reactor is

similar to the amount of active phase present in a small-size monolith

catalyst. The powder catalyst was diluted with SiC for the experiments to

prepare a catalytic bed with the volume of a monolith. This facilitates the

comparison of results obtained with the powder active phase and with the

Chapter 2

70

supported catalysts, since N2O conversion results can be directly

compared.

Analysis Unit:

Cromatograph HP6890

Quimiluminiscence analyzer

Oven

Temperarture controller

Quartz reactor

Feed and mix gases unit

N2O

NO

He

H2OO2

Figura 2.20. Scheme and picture of the set-up used to perform the catalytic tests with

powder and small-size monolith catalysts.

2.2.2 Medium-size honeycomb monolith catalysts.

NOx reduction (SCR) and N2O decomposition experiments were

carried out with medium-size honeycomb monoliths catalysts in the

power-bench facilities belonging to the MCMA group of the University of

Alicante (see Figure 2.21).The system contains a Turbo Diesel 2.0 engine

running with commercially available diesel fuel at 880 rpm (idle conditions).


71

At this running conditions, the exhaust flow was 800 L/min, and the gas

composition was 17.2 % O2, 2.4 % CO2, 10 % H2O, 100 ppmV NO,

35 ppmV NO2, 120 ppmV CO and 10 ppmV THC (THC = total

hydrocarbons, expressed as CH4). The gas exhaust temperature ranged

from 90 to 50 ºC along the exhaust pipe.

Stream to

analyzers

Figure 2.21. Power bench for catalytic test of medium-size monolith catalysts

A gas stream of 10 L/min was continuously pumped out from the

main stream and was used to perform the tests. This gas flow passed

throughout two consecutive furnaces with independent control of

temperature, the first one containing a crucible with commercial diesel fuel

and the second one with the monolith catalysts.

Experiments in single or dual-bed configurations were performed.

Only the SCR catalyst (Pt/Beta zeolite/monolith) was placed in the second

furnace for single-bed experiments, and additionally, the deN2O catalyst

(RhOx/Ce0.9Pr0.1O2/monolith) was located in the same furnace,

downstream the SCR catalyst, for dual-bed experiments. The amount of

Chapter 2

72

diesel fuel evaporated in the first furnace, and used in the SCR process in

the second furnace, was controlled by fixing the temperature of the first

furnace. In these experiments, the space velocity (GHSV) in one monolith

was 19000 h-1

.

The gas composition was continuously monitored during the

experiments by specific gas analyzers from Signal Instruments for THC

(model 3000HM), CO, CO2 and N2O (model 7000FM) NO and NO2 (model

4000VM), and O2 (model 8000 M).

The following parameters were determined:

in

2in

out2

outin2

in

][NO[NO]

][NO-[NO]- ][NO [NO] 100· % removalNOx

(13)

out2

outin2

in

out2

2][NO-[NO]- ][NO [NO]

O][N 1/2 -1100· %y selectivit N (14)

in

outin

[THC]

[THC]- [THC] 100· % removal THC (15)

where the superscripts “in” and “out” refer to inlet and outlet

concentrations, respectively.

2.2.3 Full-size honeycomb monolith catalysts.

SCR experiments were also performed in a engine test bench

Horiba Titan S190 (Figure 2.22), with the full-size monolith loaded with

Pt/Beta zeolite and a 1.6 HDI diesel engine running with commercial diesel

fuel at 1100 rpm and different loading (torque between 45 and 83 N·m).

These experiments were performed in the facilities of the Galician

Automotive Technology Centre (Centro Tecnológico de Automoción de

Galicia, CTAG).


73

Diesel injector

gas to opacimeter

Gas flow Fuel injector

Gas to analyzers

Fuel pump

Pressure sensor

Pressure sensor

Thermocouple

Thermocouple

Gas to opacimeter

Figure 2.22. Photograph of the engine test bench and detail of the metal holder used

for the full-size SCR monolith catalyst, which is fitted in the exhaust pipe.

Depending on the engine loading, the temperature at the entrance

of the SCR catalyst and the gas composition ranged as indicated in Table

2.1

Table 2.1. Gas temperature and composition measured at the entrance of the monolith

catalyst (without fuel injection) for different engine loading.

Exhaust

gas flow

(kg/h)

Torque

(N·m)

GHSV

(h-1

)

Gas

temp

(ºC)

O2

(%)

CO2

(%)

CO

(ppm)

THC

(ppm)

NO

(ppm)

NO2

(ppm)

17 45 9460 220 8 11 534 310 91 6

17 60 10035 250 7 12 320 375 218 0

18 83 11640 300 5 13 455 143 411 0

Chapter 2

74

In these experiments the whole exhaust stream was treated.

Diesel fuel was used as reducing agent, being fed to the exhaust stream

by a fuel injector located at the entrance of the metal holder used to fit the

catalyst to the exhaust pipe (see Figure 2.22.). The catalyst holder was

designed and fabricated for this study. The holder was provided with five

connections at the inlet part of the monolith catalyst, in order to connect a

thermocouple, a pressure sensor, the fuel injection system and the two gas

conductions connected to the gas analyzers and gas opacimeter,

respectively. At the monolith catalyst exit, the holder was provided with

four connections for a thermocouple, a pressure sensor and also two gas

conductions connected to the gas analyzers and the gas opacimeter,

respectively. Horiba Mexa 7170D analyzers were used to follow gas

composition (NOx, CO, CO2, THC, and O2).

The fuel injection system consisted of a fuel pump, a

programmable touch screen, a programmable logic controller (PLC), and a

low-pressure injector. Both the frequency and amount of fuel pulsed could

be selected in order to ensure a precise control of the THC concentration

in the inlet gas stream.

SCR experiments were performed by feeding different amounts of

fuel in order to reach THC (total hydrocarbons) concentrations in the

exhaust (measured at the entrance of the catalyst holder) in the range

1000 - 5000 ppm. The SCR experiments were extended until all engine

parameters and gas compositions were at steady state.


75

References.

[1] I. Langmuir. The constitution and fundamental properties of solids

and liquids. Part I. Solids. Journal of the American Chemical

Society 38 (1916) 2221.

[2] S. Brunauer, P.H. Emmett. The use of Van der Waals adsorption

isotherms in determining the surface area of iron synthetic

ammonia catalysts. Journal of the American Chemical Society 57

(1935) 1754.

[3] S. Brunauer, P.H. Emmett. The use of low temperature van der

Waals adsorption isotherms in determining the surface areas of

various adsorbents. Journal of the American Chemical Society 59

(1937) 2682.

[4] S. Brunauer, P.H. Emmett, E. Teller. Adsorption of Gases in

Multimolecular Layers. Journal of the American Chemical Society.

60 (1938) 309.

[5] J. Rouquerol, F. Rouquerol, K.S.W. Sing. Adsorption by Powders

and Porous Solids: Principles, Methodology and Applications.

Academic Press (1998).

[6] F. Rouquerol, J. Rouquerol, B. Imelik. Etude de la texture des

solides poreux. 4. Validite de la loi BET dans le cas de ladsorption

dazote dargon et de butane des adsorbants poreux. Bulletin de la

Société chimique de France 3 (1964) 635.

[7] S.J. Gregg, K.S.W. Sing. Adsorption, surface area and porosity.

2nd

edition. Academic Press. London (1982).

[8] http://fys.kuleuven.be/iks/nvsf/experimental-facilities/x-ray-

diffraction-2013-bruker-d8-discover. Access May 2013.

[9] M. Rodríguez-Gallego. La difracción de los rayos X. Alhambra

Universidad. (1982).

[10] http://sstti.ua.es/en/scientific-instrumentation/x-ray-unit/x-ray-

difracction.html. Acces May 2013

http://books.google.es/ebooks?output=ws2&as_brr=5&q=inauthor%3A%22Jean%20Rouquerol%22

http://books.google.es/ebooks?output=ws2&as_brr=5&q=inauthor%3A%22Fran%C3%A7oise%20Rouquerol%22

http://books.google.es/ebooks?output=ws2&as_brr=5&q=inauthor%3A%22Kenneth%20S.W.%20Sing%22

http://books.google.es/books?id=hNCREhZ7Wt8C&hl=es&source=gbs_book_other_versions

http://books.google.es/books?id=hNCREhZ7Wt8C&hl=es&source=gbs_book_other_versions

http://fys.kuleuven.be/iks/nvsf/experimental-facilities/x-ray-%20diffraction-2013-bruker-d8-discover.%20Access%20May%202013

http://fys.kuleuven.be/iks/nvsf/experimental-facilities/x-ray-%20diffraction-2013-bruker-d8-discover.%20Access%20May%202013

http://sstti.ua.es/en/scientific-instrumentation/x-ray-unit/x-ray-difracction.html.%20Acces%20May%202013

http://sstti.ua.es/en/scientific-instrumentation/x-ray-unit/x-ray-difracction.html.%20Acces%20May%202013

Chapter 2

76

[11] F. Zhang, Q. Jin, S-W. Chan. Ceria nanoparticles: size, size

distribution, and shape. Journal of Applied Physics 95 (2004) 4319.




[13] http://sstti.ua.es/en/scientific-instrumentation/unit-single-crystal-x-

ray-and-vibrational-spectroscopy-and-optical/raman-

spectroscopy.html. Access May 2013

[14] http://www.lasurface.com/xps/index.php. Acces May 2013

[15] http://sstti.ua.es/en/scientific-instrumentation/x-ray-unit/x-ray-

photoelectron-spectroscopy.html. Acces May 2013

[16] N. Guillén-Hurtado, I. Atribak, A. Bueno-López, A. García-García,


and NO to NO2 oxidation activity of ceria and ceria–zirconia


52.






[18] H. Borchert, Y.V. Frolova, V.V. Kaichev, I.P. Prosvirin,

G.M. Alikina, A.I. Lukashevich, V.I. Zaikovskii, E.M. Moroz,

S.N. Trukhan, V.P. Ivanov, E.A. Paukshtis, V.I. Bukhtiyarov,

V.A. Sadykov. Electronic and chemical properties of

nanostructured cerium dioxide doped with praseodymium. Journal

of Physical Chemistry B 109 (2005) 5728.

[19] http://sstti.ua.es/en/scientific-instrumentation/microscopy-

unit/transmission-electron-microscopy.html. Access May 2013.

[20] R.F. Egerton. Physical principles of electron microscopy: an

introduction of TEM, SEM and AEM. Springer (2008).

http://sstti.ua.es/en/scientific-instrumentation/unit-single-crystal-x-ray-and-vibrational-spectroscopy-and-optical/raman-spectroscopy.html.%20Access%20May%202013



http://www.lasurface.com/xps/index.php.%20Acces%20May%202013

http://sstti.ua.es/en/scientific-instrumentation/x-ray-unit/x-ray-photoelectron-spectroscopy.html.%20Acces%20May%202013

http://sstti.ua.es/en/scientific-instrumentation/x-ray-unit/x-ray-photoelectron-spectroscopy.html.%20Acces%20May%202013

http://sstti.ua.es/en/scientific-instrumentation/microscopy-unit/transmission-electron-microscopy.html.%20Access%20May%202013

http://sstti.ua.es/en/scientific-instrumentation/microscopy-unit/transmission-electron-microscopy.html.%20Access%20May%202013


77

[21] http://sstti.ua.es/en/scientific-instrumentation/thermal-analysis-

unit/thermogravimetry-tg-atd-tg-mass-spectrometry.html. Access

May 2013.

http://sstti.ua.es/en/scientific-instrumentation/thermal-analysis-unit/thermogravimetry-tg-atd-tg-mass-spectrometry.html

http://sstti.ua.es/en/scientific-instrumentation/thermal-analysis-unit/thermogravimetry-tg-atd-tg-mass-spectrometry.html

CHAPTER 3

Effect of the calcination conditions of

the RhOx/CeO2 catalysts on N2O

decomposition activity

In this chapter the effect of the calcination conditions (ramp

and flash calcinations, starting calcinations at 25 or 250/350 ºC,

respectively) of 2.5%RhOx/CeO2 catalysts on the catalytic activity for

N2O decomposition has been studied. The calcination conditions of

cerium nitrate have neither effect on the physicochemical properties of

ceria, observed by XRD, Raman spectroscopy and N2 adsorption, nor

on the catalysts performance for N2O decomposition. On the contrary,

flash calcination of rhodium nitrate improved the catalytic activity for

N2O decomposition. This is attributed to the small size of RhOx

nanoparticles obtained which allows a higher rhodium oxide-ceria

interface, favouring the reducibility of the ceria surface and stabilizing

the RhOx species under reaction conditions.

Effect of the RhOx/CeO2 calcination conditions

81

3.1 Introduction.

In catalytic applications, ceria is used either as metal support or as

catalyst itself [1-3]. The use of ceria as noble metal support has attracted

intense interest due to their vast applications in heterogeneous catalysis.

Noble metal/ceria catalysts are widely used in important applications such

as TWC for gas pollution control in gasoline vehicles [4], fluid catalytic

cracking (FCC) [1], PM combustion [5], selective catalytic reduction of NOx

[6], electrocatalyst in cathodes for phosphoric acid fuel cells [7, 8], partial

oxidation of methane [9], combustion of volatile organic compounds [10]

and N2O decomposition [11] among others. Many factors, including the

size and distribution of the noble metal particles, the surface morphology

and defects on the oxides, affect the performance of noble metal/ceria

catalysts. Indeed, effectively controlling the size of noble metal particles is

crucial for maintaining high catalytic activity. The behavior of the noble

metal-ceria interface is of critical importance in this regard [1, 3] and new

ways for optimizing the noble metal-ceria interactions, in order to improve

the catalytic performance, are being investigated [12].

The promoting effect of ceria-based materials in the catalytic

activity of rhodium, and other platinum group metals, is very well known for

several chemical reactions [4, 12-22]. The best example is found in TWC

used in gasoline vehicles for the simultaneous removal of NOx, HC and

CO. Also, rhodium, supported on different materials (ZnO, ceria, ZSM-5),

was investigated for N2O decomposition, Rh/ceria being the most active

[23].

In a previous work performed in our group, noble metals (rhodium,

palladium and platinum) supported on -Al2O3, pure ceria and lanthanum

and praseodymium-doped ceria were studied as N2O decomposition

catalysts [11]. It was concluded that rhodium is the most active noble metal

among those tested, and that the redox properties of the ceria support

affect the catalytic activity of the supported noble metals. For rhodium

catalysts, those with a ceria-based support present higher activity than

Rh/-Al2O3 demonstrating the participation of ceria, either pure or doped, in

the N2O decomposition process. The ceria support interacts strongly with

rhodium partially stabilizing cationic species of the noble metal during N2O

decomposition [24]. For this reason, the formulation adopted for these

Chapter 3

82

catalysts is RhOx/CeO2, evidencing the cationic nature of the metal. The

low-temperature activity of RhOx/CeO2 for N2O decomposition was

attributed to electron excess sites at the micro-interfaces between the

dispersed RhOx particles and the ceria support [24, 25].

The RhOx particles size and their interaction with the ceria support

could be affected by the preparation variables, such as the drying and

calcination conditions used during the powder catalysts synthesis. Such

drying and calcination conditions are also known to affect the distribution of

active phases along the honeycomb monoliths channels [26], which are

the real configuration for many practical applications. The preparation of

honeycomb monolith-supported catalysts is the objective of further

chapters of this thesis, and knowing the effect of the drying and calcination

conditions on the RhOx/CeO2 active phase performance will be useful to

design a suitable preparation procedure for the honeycomb

monoliths-supported catalysts used in further chapters.

Bearing this in mind, the goal of the current chapter is to get further

insights into the effect of the drying and calcination conditions on the

properties of N2O decomposition RhOx/CeO2 catalysts

3.2 Experimental.

3.2.1 Study of the metal precursors decomposition.

Decomposition of 10 mg of either Ce(NO3)36H2O or

ceria-impregnated with rhodium nitrate (Rh(NO3)3/Ce25 sample, see

section 3.2.2 for nomenclature details) was performed in air in a TG-MS

set up in order to select the adequate starting heating temperatures for

calcination of catalysts precursors. The experimental details of these tests

were described in Chapter 2.

3.2.2 Catalysts preparation.

Different RhOx/CeO2 catalysts have been synthetised for this

study. Powder CeO2 was prepared by Ce(NO3)3·6H2O (Alfa-Aesar,

99.5 wt.%) calcination in two different ways:


83

Ramp calcination: consisted of heating the cerium precursor in a

muffle furnace (static air) from 25 to 600 ºC at 10 ºC/min,

maintaining the maximum temperature for 90 min. This ceria

support is referred to as Ce25.

Flash calcination: consisted of heating the empty muffle furnace at

250 ºC. Then, the cerium precursor was introduced and the

temperature was raised at 10 ºC/min up to 600 ºC, maintaining the

maximum temperature for 90 min. This ceria support is referred to

as Ce250.

Rhodium was loaded on both ceria powders by incipient wetness

impregnation with an aqueous solution of Rh(NO3)3·xH2O (Sigma-Aldrich,

~36 wt.% as Rh) of the appropriate concentration to obtain 2.5 wt.% of

rhodium on the catalysts. The impregnated supports were placed in the

furnace immediately after impregnation. The rhodium precursor

decomposition was also performed using different calcination conditions

(ramp and flash). The muffle furnace temperature was stabilized at 25 ºC

(ramp calcination) or at 250 or 350 ºC (flash calcinations) before the ceria-

impregnated with rhodium nitrate was introduced (denoted by Rh25,

Rh250 and Rh350, respectively). In all cases the heating rate was

10 ºC/min and the final temperature was 500 ºC, maintaining this

temperature for 30 min. Following the described procedures, four

RhOx/CeO2 catalysts have been prepared, which are referred to as

Rh25Ce25, Rh250Ce25, Rh350Ce25 and Rh350Ce250.

3.2.3 Catalysts characterization.

Catalysts crystalline properties were studied by XRD and Raman

spectroscopy. BET surface area was determined from the N2 adsorption

isotherms at -196 ºC. XPS was used to determine the oxidation states of

the elements. In addition, samples reducibility was examined by

experiments of H2-TPR. Finally, particle size distribution of rhodium oxide

was determined by TEM. These experimental techniques and the

conditions used for each measurement were previously described in detail

in Chapter 2.

Chapter 3

84

3.2.4 N2O decomposition tests.

N2O decomposition tests were performed at laboratory scale using

a 100 mL/min flow (GHSV = 42000 h−1

) of 1000 ppm N2O in He, and

100 mg of catalyst. Details of these experiments were described in

Chapter 2.

3.2.5 CO oxidation tests.

CO oxidation has been selected as an additional test reaction, in

order to analyze if the conclusions achieved about the effect of the thermal

treatment conditions on the catalytic decomposition of N2O can be

extended to some other reactions. CO oxidation experiments were

performed at atmospheric pressure in the same experimental set-up used

for N2O decomposition (see details in Chapter 2). In this case 100 mL/min

flow of 1000 ppm CO/5 % O2/He and 100 mg of catalyst

(GHSV = 42000 h−1

) were used.

3.3 Results and discussion.

3.3.1 Thermogravimetry - Mass Spectroscopy (TG-MS) study of metal

precursors decomposition.

The thermograms corresponding to Ce(NO3)3·6H2O and

Rh(NO3)3/Ce25 decomposition are plotted in Figure 3.1 together with

representative MS signals. In both experiments, most water molecules

(m/z 18) evolved below 250 ºC with the corresponding weight loss. During

Ce(NO3)3·6H2O decomposition (Figure 3.1.a), three well-defined m/z 18

peaks were observed assigned to coordination water release. In the

Rh(NO3)3/Ce25 decomposition experiment (Figure 3.1.b) a broad m/z 18

band with two maxima appeared. It must be outlined the difference in the

Y-axis scales due to the low Rh(NO3)3 percentage in the sample studied.


85

0

10

20

30

40

50

60

70

80

90

100

8.00E-11

1.80E-10

2.80E-10

3.80E-10

4.80E-10

5.80E-10

6.80E-10

0 100 200 300 400 500 600 700 800

Ma

ss (%

)

MS

sig

nal (a

.u.)

Temperature (ºC)

(a)

MS 18 (H2O)

MS 30 + MS46 (NOx)

94

95

96

97

98

99

100

0 100 200 300 400 500 600 700 800

Ma

ss (%

)

MS

sig

nal (a

.u.)

Temperature (ºC)

(b)

MS 18 (H2O) MS 30 + MS46 (NOx)

Figure 3.1. Metal precursors decomposition followed by TG-MS. (a) Ce(NO3)3·6H2O

and (b) Rh(NO3)3/Ce25.

Cerium nitrate (Figure 3.1.a) decomposed in a relatively narrow

range of temperature (240-320 °C) with NO (m/z 30), NO2 (m/z 46) and O2

(m/z 32, which is not shown for clarity) release. This result lead us to select

25 and 250 ºC for the preparation of the ceria supports in ramp (slow water

release and slow nitrate decomposition) and flash (rapid water release and

slow nitrate decomposition) calcination conditions, respectively.

Ceria impregnated with rhodium nitrate (Figure 3.1.b) decomposed

in the range 200-550 ºC, and decomposition in such wide range of

temperature is an evidence of the interaction of the noble metal species

Chapter 3

86

with the ceria support (otherwise a narrower range of decomposition

temperatures would be expected). Taking the results of Figure 3.1.b into

account, three temperatures were selected to start the thermal treatments

for catalysts preparation: 25, 250 and 350 ºC.

In the calcination from 25 ºC, water release will be slow and the

impregnated rhodium salt will be allowed to move on the ceria particles

surface forced by the concentration gradients created during drying. Once

dry, rhodium nitrate is expected to be distributed on the ceria surface more

heterogeneously than if water is released very rapidly (as in calcinations

starting at 250 or 350 ºC), and this heterogeneity is expected to affect the

final RhOx particle size, as it will be demonstrated afterwards. It is also

expected that this kind of slow treatment starting at 25 ºC will also lead to a

poor distribution of active phase into a cordierite monolith when the

catalyst is scaled up in Chapters 5 and 6.

On the contrary, in the calcination starting at 250 ºC most water

release will be very rapid but not rhodium nitrate decomposition, while

starting at 350 ºC both water release and rhodium nitrate decomposition

will be fast.

3.3.2 XRD, Raman spectroscopy and N2 adsorption at -196 ºC

characterization.

All catalysts were characterized by XRD, Raman spectroscopy and

N2 adsorption at -196 ºC. These three techniques have provided

information mainly about the ceria support properties. XRD and Raman

spectroscopy are complementary, and provide information about the

structure of the ceria supports. XRD is sensitive to the cations position in

the ceria lattice, while Raman spectra of cerium oxides are caused by

oxide anions vibration.

All XRD diffractograms (Figure 3.2) show fluorite structure

characteristic reflections, corresponding to the (111), (200), (220), (311),

(222) and (400) planes. No other peaks but those of fluorite were observed

in the diffractograms [27] indicating that rhodium species should be highly

dispersed and/or should present low crystallinity.


87

10 20 30 40 50 60 70 80

Inte

nsity (

a.u

.)

2θ (º)

Rh25Ce25

Rh250Ce25

Rh350Ce25

Rh350Ce250

(111)

(200)

(220)(311)

Figure 3.2. XRD patterns of fresh catalysts.

The lattice parameter values were determined from XRD patterns,

and the values obtained are presented in Table 3.1. The differences are

within the experimental error of the determination. The average crystallite

sizes of the ceria particles were determined with the Scherrer’s and

Williamson-Hall’s equations [28]. The values obtained are included in

Table 3.1 as well as the BET surface areas. Similar crystallite sizes and

BET surface areas have been found for all catalysts.

Table 3.1. XRD and N2 adsorption characterization results.

*The lattice parameter experimental error is estimated to be ± 0.0014 nm in our

experimental conditions.

Raman spectra included in Figure 3.3 are also similar for all

catalysts. All of them show the main band at about 465 cm-1

assigned to

the F2g mode of the fluorite-type structure of cerium oxides, based on the

Sample

Lattice

parameter

(nm)*

Crystal Size

by Scherrer

(nm)

Crystal Size by

Williamson-Hall

(nm)

BET

surface

area

(m2/g)

Rh25/Ce25 0.5401 12 14 64

Rh250/Ce25 0.5412 12 13 66

Rh350/Ce25 0.5401 12 13 66

Rh350/Ce250 0.5424 12 13 69

Chapter 3

88

face-centered cubic cell [29,30], and a small peak at ca. 230 cm-1

assigned

to RhOx species [31-33].

150 250 350 450 550 650 750

Inte

nsity (

a.u

.)

Raman Shift (cm-1)

Rh25Ce25

Rh250Ce25

Rh350Ce25

Rh350Ce250

Figure 3.3. Raman spectra of fresh catalysts.

In conclusion, the XRD, Raman spectroscopy and N2 adsorption

characterization reveal that all the ceria supports seems to be very similar,

that is, the heating conditions used in the calcinations steps do not

significantly affect ceria properties.


N2O decomposition experiments were performed with a 1000 ppm

N2O/He stream in the temperature range 175-275 ºC. The N2O

decomposition profiles in steady-state are plotted in Figure 3.4. As

expected, the N2O decomposition percentage increases with temperature,

reaching total conversion at 275 ºC for all the catalysts. According to these

results, and in agreement with previous XRD, Raman spectroscopy and N2

adsorption characterization, the decomposition conditions of cerium nitrate

has no significant effect on the catalytic performance (profiles for

Rh350Ce25 and Rh350Ce250 are almost equal) while rhodium nitrate

decomposition affects the activity. The catalyst where rhodium nitrate was

decomposed in ramp (starting heating at 25 ºC (Rh25Ce25)), which is the

most conventional calcination procedure, presents lower activity than

catalysts calcined by a flash procedure (Rh250Ce25 and Rh350Ce25


89

catalysts). This can be tentatively attributed to the improved interaction

between RhOx and ceria particles obtained by flash calcinations, and this

interaction is studied in detail in the coming sections.

0

20

40

60

80

100

175 200 225 250 275

N2O

decom

positio

n

(%)

Temperature (ºC)

Rh25Ce25

Rh250Ce25

Rh350Ce25

Rh350Ce250

Figure 3.4. N2O decomposition in steady state as a function of temperature.

Additional information about the catalyst performance is obtained

from the N2O decomposition profiles as a function of time for the different

temperatures, that is, from the behavior of the N2O decomposition profiles

before the steady-state is achieved. The steady state was reached after

20-30 minutes of reaction for all catalysts at temperatures below 225 ºC. At

225 ºC longer times were required to reach a constant N2O decomposition

level. On the other hand, at temperatures above 225 ºC the steady state

was again reached in 20-30 minutes. The curves corresponding to 225 ºC

are included in Figure 3.5 for selected catalysts (Rh25Ce25 and

Rh250Ce25; rhodium nitrate decomposed by ramp and flash calcinations,

respectively). From a qualitative point of view both profiles are similar, with

a first step of 35-40 minutes where the N2O decomposition level increased

until a pseudo steady-state was reached. The conversions then increased

again for 20-25 additional minutes before the real steady-state level was

stabilied. Despite both profiles on Figure 3.5 are qualitatively similar, the

sample Rh250Ce25 needed more time to reach the final steady-state than

Rh25Ce25, the former also being more active than the latter. The

transformations suffered by these two catalysts at 225 ºC before the

steady state was achieved were studied by XPS.

Chapter 3

90

35

40

45

50

55

60

0 20 40 60 80

N2O

decom

positio

n (

%)

Time (min)

Rh25Ce25

Rh250Ce25

Figure 3.5. N2O decomposition at 225 ºC as a function of time.

3.3.4 Characterization by XPS of fresh catalysts and after “in situ”

pre-treatments with N2O at 225 ºC.

XPS spectra were recorded with the fresh Rh250Ce25 and

Rh25Ce25 catalysts and with these catalysts pre-treated in situ with

1000 ppm N2O/He at 225 ºC for different periods of time. The profiles

obtained for the Rh 3d (see Figure 3.6 as an example) and Ce 3d

transitions were qualitatively similar to those previously obtained with some

other RhOx/ceria catalysts [24, 34, 35]. The percentages of Rh0 (with

regard to total rhodium) and Ce3+

(with regard to total cerium) were

calculated following the methods reported elsewhere [24, 34, 35] and

described in Chapter 2. Rh0

and Ce3+

percentages are compiled in Table

3.2.


91

303 308 313 318

Inte

nsity (

a.u

.)

Binding Energy (eV)

t=0 min

t=5 min

t=15 min

t=60 min

Figure 3.6. Rh 3d photoelectron spectra of the Rh25Ce25 catalyst as an example.

As described in Chapter 2, the peaks assigned to Rh0 appear at

307.0–307.5 eV and the one assigned to Rh3+

appear from 308.3 to

310.5 eV. Therefore, rhodium appeared fully oxidized on both fresh

catalysts (t = 0 min), which was expected since the catalysts were calcined

at 500 ºC, and the Ce3+

percentages (38-40 %) were in accordance with

values typically obtained for similar synthesis conditions [24, 34, 36].

As observed in Figure 3.6 and Table 3.2, rhodium was

progressively reduced and only Rh0 was identified in both catalysts after

the 15 min pretreatment, while after 60 minutes both Rh3+

and Rh0

were

observed again. The behavior of the cerium oxidation state (Table 3.2) was

qualitatively similar to that of rhodium, Ce4+

being first slightly reduced to

Ce3+

and reoxidized afterwards. It has to be mentioned that only the trend

of the metals oxidation state must be considered, but not the absolute

values, since the reducing environment of the XPS measurements (high

vacuum and an electron beam) could affect such absolute oxidation state

values.

These XPS results (Table 3.2), together with the N2O

decomposition profiles included in Figure 3.5, evidence that the

RhOx/CeO2 catalysts suffer an activation process at 225 ºC, and that the

reoxidation of the catalysts (both for rhodium and cerium) is stronger for

the most active catalyst (Rh250Ce25; see Table 3.2 for time = 60 min),

suggesting a deeper transformation during the activation period.

Chapter 3

92

Table 3.2. Rh0 and Ce

3+ percentages determined by XPS after in situ thermal

treatments with 1000 ppm N2O at 225 ºC for different times.

This activation process is consistent with the reaction mechanism

previously proposed for the RhOx/CeO2-catalyzed N2O decomposition

[24, 36]:

Rh–O + N2O → Rh* + N2 + O2 (1)

Rh* + N2O → Rh*–O + N2 (2)

2Rh*–O ↔ 2Rh* + O2 (3)

Rh* + Ce*–O → Rh*–O + Ce* (4)

Ce* + N2O → Ce*–O + N2 (5)

2Ce*–O ↔ 2Ce* + O2 (6)

Rh*–O + Ce*–O ↔ Rh* + Ce* + O2 (7)

Ce–OH + N2O → Ce* + N2 + H2O (8)

Ce* + Ce–O + H2O → 2Ce–OH (9)

Sample

Pre-treatment

Time

(min)

Ce3+

(%)

Rh0

(%)

Rh25Ce25 0 38 0

5 40 34

15 42 100

60 39 75

Rh250Ce25 0 40 0

5 43 28

15 44 100

60 38 60


93

where Rh*–O and Rh* represent oxidized and reduced rhodium sites,

respectively, and Ce*–O and Ce* are oxidized and reduced ceria sites.

Before the steady state is achieved, the Rh3+

and Ce4+

reduction

observed by XPS evidence that the reducing steps (1), (3), (6), (7) and (8)

prevail against those where N2O oxidizes catalysts sites (steps (2) and

(5)), which makes sense taking into account that the fresh catalysts is

highly oxidized. After some time, once rhodium is reduced, N2O and also

Ce4+

oxide partially re-oxidizes the rhodium sites via steps (2), (4) and (5),

reaching a net null balance between the oxidizing and reducing steps

rates, that is, achieving the real steady state.

Coming back to the fact that this behavior was neither observed

below 225 ºC nor above this temperature, it is postulated that the

activation process does not occurs at low temperatures because the steps

involving reduction and re-oxidation of cerium are not occurring in a great

extent. Secondly, since the experiments are performed sequentially, the

effect is not observed at 250 ºC and higher temperatures because the

catalysts have already been activated at 225 ºC.

3.3.5 H2-TPR characterization.

Additional information about the catalysts has been obtained by

H2-TPR. The TCD profiles obtained in these experiments are included in

Figure 3.7. Three peaks are observed for all catalysts. The

lowest-temperature peak at around 100 ºC is attributed to both RhOx

reduction and noble metal-catalyzed surface reduction of ceria due to the

hydrogen spillover on the surface [14, 35, 37-39]. The TCD signal

appearing between 200 and 400 ºC can be attributed to different events:

the reduction of surface cerium oxide which is not in close contact with

rhodium, the decomposition of surface carbonates (occluded within the

ceria structure) and/or the reduction of hydroxyl and peroxide/superoxide

surface groups [15,40-42]. The H2 consumed above 700 ºC is assigned to

bulk ceria reduction, that is, to the reduction of Ce4+

cations placed within

the bulk oxide particles.

A detailed analysis of the lowest temperature reduction peak

shows a slightly lower reducibility of the catalyst prepared by ramp

calcination (Rh25Ce25, see Figure 3.7) in comparison to the flash calcined

Chapter 3

94

catalysts. A larger ceria surface reducibility must be a consequence of a

better noble metal-support interaction since the process occurs through the

catalytic action of the noble metal.

0 150 300 450 600 750 900 1050

TD

C s

ignal (a

.u.)

Temperature (ºC)

Rh25Ce25

Rh250Ce25

Rh350Ce25

Rh350Ce250

25 50 75 100 125 150 175 200

TC

D s

ign

al (a

.u.)

Temperature (ºC)

Figure 3.7. H2-TPR profiles of fresh catalysts after an in situ pretreatment with

5 % O2/He at 500 ºC.

The catalysts with larger reducibility at low temperature also exhibit

the higher catalytic activity for N2O decomposition (Figure 3.4). This

observation is in agreement with the conclusions of previous studies,

where a relationship between ceria surface reducibility and N2O

decomposition capacity was obtained for a set of RhOx/CeO2 catalysts

prepared with different ceria carriers (either pure and doped with

lanthanum or praseodymium) [11]. Moreover, the rate limiting step of the

RhOx/CeO2 catalyzed N2O decomposition mechanism, once the steady

state is achieved, was reported to be the reduction of the catalyst sites by

N2O [24]. The redox properties of the support are decisive for RhOx

stabilization under reaction conditions, the larger the amount of reducible

CeO2 at the surface, the better the catalytic activity. These results suggest

that flash calcination of rhodium nitrate (starting heating either at 250 or

350 ºC) allows obtaining a much better noble metal–support interaction

than the conventional ramp calcinations (starting heating at 25 ºC)

[24, 35, 43], which was confirmed by TEM analysis.


95

3.3.6 TEM characterization.

Catalysts Rh25Ce25 and Rh250Ce25 were selected for TEM

characterization. Figure 3.8 shows, as an example, a representative

micrograph of each catalyst. The typical crystalline planes of the fluorite

structure of ceria are clearly observed in both micrographs as well as some

RhOx particles.

In the catalyst prepared by conventional ramp calcination

(Rh25Ce25), RhOx particles smaller than 3 nm are distinguished. In this

case, the number of RhOx particles which are clearly distinguished from

the ceria support is enough to make a RhOx particle size distribution (inset

in Figure 3.8.a). Most RhOx particles size is in the range 0.5-1 nm.

In the catalyst prepared by flash calcination of rhodium nitrate

(Rh250Ce25; Figure 3.8.b), RhOx particles have been hardly observed,

assuming that most of them are much smaller than 0.5 nm, and evidencing

a better dispersion of RhOx on this catalyst. This difference in RhOx size is

attributed to the different water release rate during drying. Water

evaporates much slower during conventional ramp calcination than during

a flash heating, creating concentration gradients along the ceria surface.

Consequently, rhodium nitrate is more heterogeneously distributed than if

water is released very rapidly (as in flash calcination) affecting the final

RhOx particle size.

As a summary, smaller RhOx particles are supported on ceria by

flash calcinations of impregnated rhodium nitrate in comparison to

conventional ramp calcinations. Catalysts prepared by this procedure are

more active than the counterpart prepared in ramp, since the RhOx-ceria

interface is improved.

Chapter 3

96

0.0 0.5 1.0 1.5 2.0 2.5 3.00

10

20

30

40

50

60

Per

centa

ge

(%)

Particle size (nm)

(a) (b)

Figure 3.8. TEM micrographs of catalysts (a) Rh25Ce25 and (b) Rh250Ce25.

3.3.7 CO oxidation tests.

As discussed, the flash calcination of ceria-impregnated rhodium

nitrate improves the catalytic activity for N2O decomposition with regard to

conventional ramp heating. In order to analise if this is a particular

conclusion that only applies to the N2O decomposition reaction, or on the

contrary, it can be extended to some other catalyzed reactions, CO

oxidation was selected as catalytic reaction test with practical relevance.

Figure 3.9 plots CO conversions obtained in steady state at

temperatures from 25 ºC to 90 ºC.

The activity of the catalyst prepared by flash calcination is

considerably higher than that calcined in ramp, showing significant activity

even at room temperature. For instance, the amount of CO oxidized at

45 ºC by Rh250Ce25 doubles that of the catalyst calcined by conventional

ramp. The CO oxidation rates achieved by this novel catalyst with RhOx

sub-nanoparticles are also much higher than those typically reported in

literature for rhodium catalyst [44].


97

0

20

40

60

80

100

0 25 50 75 100

CO

oxid

ation

(%

)

Temperature (ºC)

Rh25Ce25

Rh250Ce25

Figure 3.9. CO conversion in oxidation reactions as a function of temperature for

Rh250Ce25 and Rh25Ce25.

In accordance with the previous discussion about the N2O

decomposition results and about the catalysts characterization, the

enhanced activity for CO oxidation of the catalyst prepared by flash

calcination is attributed to the larger noble metal oxide-support interface

due to the smaller RhOx particles obtained by flash calcination. Highly

active catalytic sites are located at the RhOx-ceria interface, as reported in

previous studies for CO oxidation and also for some other catalytic

reactions. For instance, it has been reported that the oxidation of CO

occurs at the interface between the rhodium particles and the ceria support

[45], and also that the active sites for low temperature N2O dissociation are

located at the RhOx-CeO2 interface [24].

3.4 Conclusions.

The conclusions of this study can be summarized as follows:

The calcinations method (ramp or flash) of cerium nitrate to obtain

the ceria support has no effect neither on the ceria properties

(those observed by XRD, Raman spectroscopy and N2 adsorption)

Chapter 3

98

nor on the RhOx/CeO2 catalyst performance for N2O

decomposition.

Flash calcination of rhodium nitrate impregnated on ceria improves

the catalytic activity for N2O decomposition and CO oxidation of

RhOx/CeO2 catalysts in comparison to that of similar catalysts

calcined in ramp.

The improved catalytic activity for N2O decomposition and CO

oxidation of catalysts with rhodium nitrate decomposed by flash

calcinations is attributed to the smaller size of RhOx nanoparticles

(smaller than 0.5 nm), which allow a larger noble metal

oxide-support interface.

The improved noble metal oxide-support interface favors the

reducibility of the ceria surface and stabilizes the RhOx species

under N2O decomposition conditions.

Fresh RhOx/CeO2 catalysts suffer a transformation process at

225 ºC during the N2O decomposition reaction, during which both

rhodium and cerium are reduced and re-oxidized. This process is

consistent with the mechanism proposed for N2O decomposition on

RhOx/CeO2.


99

References.

[1] A. Trovarelli, C. De Leitenburg, M. Boaro, G. Dolcetti. The

utilization of ceria in industrial catalysis. Catalysis Today 50 (1999)

353.

[2] M.F. Luo, J.M. Ma, J.Q. Lu, Y.P. Song, Y.J. Wang. High-surface

area CuO–CeO2 catalysts prepared by a surfactant-templated

method for low-temperature CO oxidation. Journal of Catalysis 246

(2007) 52.

[3] A. Trovarelli. Catalytic properties of ceria and CeO2-containing

materials. Catalysis Reviews - Science and Engineering 38 (1996)

439.

[4] J. Kaspar, P. Fornasiero, M. Graziani. Use of CeO2-based oxides

in the three-way catalysis. Catalysis Today 50 (1999) 285.

[5] I. Atribak, I. Such-Basáñez, A. Bueno-López, A. García-García.

Comparison of the catalytic activity of MO2 (M=Ti, Zr, Ce) for soot

oxidation under NOx/O2. Journal of Catalysis 250 (2007) 75.

[6] M. Adamowska, S. Muller, P. Da Costa, A. Krzton, P. Burg.

Correlation between the surface properties and deNOx activity of

ceria-zirconia catalysts. Applied Catalysis B: Environmental 74

(2007) 278.

[7] Q. He, S. Mukerjee, R. Zeis, S. Parres-Esclapez, M.J. Illán-Gómez,

A. Bueno-López. Enhanced Pt stability in MO2 (M = Ce, Zr or

Ce0.9Zr0.1)-promoted Pt/C electrocatalysts for oxygen reduction

reaction in PAFCs. Applied Catalysis A: General 381 (2010) 54.

[8] Q. He, S. Mukerjee, S. Parres-Esclapez, A. Bueno-López. Effect of

praseodymium oxide and cerium-praseodymium mixed oxide on

the Pt electrocatalyst performance for oxygen reduction reaction in

PAFCs. Journal of Applied Electrochemmistry 41 (2011) 891.

[9] S. Eriksson, S. Rojas, M. Boutonnet, J.L.G. Fierro. Effect of

Ce-doping on Rh/ZrO2 catalysts for partial oxidation of methane.

Applied. Catalysis A: General 326 (2007) 8.



http://www.sciencedirect.com/science/article/pii/S0926860X10002292



http://apps.webofknowledge.com/full_record.do?product=UA&search_mode=Refine&qid=7&SID=V2NAcE@P89BcCEDC4B3&page=2&doc=17


Chapter 3

100

[10] J.I. Gutiérrez-Ortiz, B. de Rivas, R. López-Fonseca, J.R. González-

Velasco. Catalytic purification of waste gases containing VOC

mixtures with Ce/Zr solid solutions. Applied Catalysis B:






(2010) 370.

[12] M. Cargnello, P. Fornasiero, R.J. Gorte. Opportunities for tailoring

catalytic properties through metal-support interactions. Catalysis

Letters 142 (2012) 1043.

[13] S. Bernal, J.J. Calvino, M.A. Cauqui, J.M. Gatica, C. Larese,

J.A. Pérez-Omil, J.M. Pintado. Some recent results on

metal/support interaction effects in NM/CeO2 (NM: noble metal)

catalysts. Catalysis Today 50 (1999) 175.

[14] P. Fornasiero, J. Kaspar, V. Sergo, M. Graziani. Redox Behavior of

High-surface-area Rh-, Pt-, and Pd-Loaded Ce0.5Zr0.5O2 mixed

oxide. Journal of Catalysis 182 (1999) 56.

[15] P. Fornasiero, R. Dimonte, G.R. Rao, J. Kaspar, S. Meriani,

A. Trovarelli, M. Graziani. Rh-Loaded CeO2-ZrO2 solid-solutions as

highly efficient oxygen exchangers: dependence of the reduction

behavior and the oxygen storage capacity of the structural

properties. Journal of Catalysis 151 (1995) 168.

[16] G. Rao, B. Mishra. Structural, redox and catalytic chemistry of ceria

based materials. Bulletin of the Catalysis Society of India 2 (2003)

122.

[17] A.S. Ivanova. Physicochemical and catalytic properties of systems

based on CeO2. Kinetics and Catalysis 50 (2009) 797.



http://www.scopus.com/source/sourceInfo.url?sourceId=16377&origin=resultslist






101

[18] T. Miyazawa, K. Okumura, K. Kunimori, K. Tomishige. Promotion

of oxidation and reduction of Rh species by interaction of Rh and

CeO2 over Rh/CeO2/SiO2. Journal of Physical Chemistry 112

(2008) 2574.

[19] P. Vidmar, P. Fornasiero, J. Kaspar, G. Gubitosa, M. Graziani.

Effects of trivalent dopants on the redox properties of Ce0.6Zr0.4O2

mixed oxide. Journal of Catalysis 171 (1997) 160.

[20] P. Bera, A. Gayen, M.S. Hegde, N.P. Lalla, L. Spadaro, F. Frusteri,

F. Arena, Promoting effect of CeO2 in combustion synthesized

Pt/CeO2 catalyst for CO oxidation. Journal of Physical Chemistry. B

107(2003) 6122.

[21] Q. He, S. Mukerjee, B. Shyam, D. Ramaker, S. Parres-Esclapez,

M.J. Illán-Gómez, A. Bueno-López. Promoting effect of CeO2 in the

electrocatalytic activity of rhodium for ethanol electro-oxidation.

Journal of Power Sources 193 (2009) 408.

[22] L.O.O. Costa, S.M.R. Vasconcelos, A.L. Pinto, A.M. Silva,

L.V. Mattos, F.B. Noronha, L.E.P. Borges. Rh/CeO2 catalyst

preparation and characterization for hydrogen production from

ethanol partial oxidation. Journal of Materials Science 43 (2008)

440.


S. Kushiyama, K. Mizuno. Decomposition of nitrous oxide over



[24] S. Parres-Esclapez, I. Such-Basañez, M.J. Illán-Gómez,





http://www.springerlink.com/content/?Author=L.+O.+O.+Costa

http://www.springerlink.com/content/?Author=S.+M.+R.+Vasconcelos

http://www.springerlink.com/content/?Author=A.+L.+Pinto

http://www.springerlink.com/content/?Author=A.+M.+Silva

http://www.springerlink.com/content/?Author=L.+V.+Mattos

http://www.springerlink.com/content/?Author=F.+B.+Noronha

http://www.springerlink.com/content/?Author=L.+E.+P.+Borges

Chapter 3

102

[25] J. Cunningham, J.N. Hickey, R. Cataluna, J.C. Conesa, J. Soria,

A. Martínez-Arias. Interfacial RhOx/CeO2 sites as locations for low

temperature N2O dissociation. Studies in Surface Science

Catalysis 101 (1996) 681.



reviews – Science and Engineering 43 (2001) 345.

[27] D. Terribile, A. Trovarelli, J. Llorca, C. de Leitenburg, G. Dolcetti.

The preparation of high surface area CeO2-ZrO2 mixed oxides by a

surfactant-assisted approach. Catalysis Today 43 (1998) 79.

[28] C. Bueno-Ferrer, S. Parres-Esclapez, D. Lozano-Castelló,

A. Bueno-López. Relationship between surface area and crystal

size of pure and doped cerium oxides. Journal of Rare Earths 28

(2010) 647.

[29] A. Mineshige, T. Taji, Y. Muroi, M. Kobune, S. Fujii, N. Nishi,

M. Inaba, Z. Ogumi. Oxygen chemical potential variation in

ceria-based solid oxide fuel cells determined by Raman

spectroscopy. Solid State Ionics 135 (2000) 481.

[30] A. Bueno-Lopez, K. Krishna, M.Makkee, J.A. Moulijn. Enhanced

soot oxidation by lattice oxygen via La3+

-doped CeO2. Journal of

Catalysis 230 (2005) 237.

[31] S. Music, A. Saric, S. Popovic, M. Ivanda. Formation and

characterization of nanosize α-Rh2O3 particles. Journal of

Molecular Structure 924–926 (2009) 221.

[32] V.V. Pushkarev, V.I. Kovalchuk, J.L. d’Itri. Probing defect sites on

the CeO2 surface with dioxygen. Journal of Physical Chemistry B

108 (2004) 5341.

[33] A. Nakajima, A. Yoshihara, M. Ishigame. Defect-induced Raman

spectra in doped CeO2. Physical Review B 50 (1994) 13297.




103





52.









[37] J. Silvestre-Albero, F. Rodríguez-Reinoso, A. Sepúlveda-

Escribano. Improved metal-support interaction in Pt/CeO2–SiO2

catalysts after zinc addition. Journal of Catalysis 210 (2002) 127.

[38] S. Imamura, J-I. Tadani, Y. Saito, Y. Okamoto, H. Jindai, C. Kaito.

Decomposition of N2O on Rh-loaded Pr/Ce composite oxides.


[39] Y. Liu, K. Murata, M. Inaba, I. Takahara, K. Okabe. Synthesis of

ethanol from syngas over Rh/Ce1−xZrxO2 catalysts. Catalysis Today

164 (2011) 308.

[40] F.M.Z. Zotin, L. Tournayan, J. Varloud, V. Perrichon, R. Frety.

Temperature-programmed reduction: limitation of the technique for

determining the extent of reduction of either pure ceria or ceria

modified by additives. Applied Catalysis A: General 98 (1993) 99.

[41] A. Trovarelli. Catalysis by Ceria and Related Materials. Catalytic

Science Series. vol. 2, Imperial College Press (2002).

[42] S. Bernal, J.J. Calvino, G.A. Cifredo, J.M. Gatica, J.A. Pérez-Omil,

J.M. Pintado. Hydrogen chemisorption on ceria: influence of the

oxide surface area and degree of reduction. Journal of the

Chemical Society, Faraday Transactions 89 (1993) 3499.

Chapter 3

104

[43] V. Rico-Pérez, S. Parres-Esclapez, M.J. Illán-Gómez, C. Salinas-

Martínez de Lecea, A. Bueno-López. Preparation, characterization

and N2O decomposition activity of honeycomb monolith-supported

Rh/Ce0.9Pr0.1O2 catalysts. Applied Catalysis B: Environmental 107

(2011) 18.

[44] S. Kim, K. Qadi, S. Jin, A.S. Reddy, B. Seo, B.S. Mun, S.H. Joo,

J.Y. Park. Trend of catalytic activity of CO oxidation on Rh and Ru

nanoparticles: role of surface oxide. Catalysis Today 185 (2012)

131.

[45] I. Manuel, C. Thomas, C. Bourgeois, H. Colas, N. Matthess,

G. Djéga-Mariadassou. Comparison between turnover rates of CO

oxidation over Rh0 or Rh

x+ supported on model three-way

catalysts. Catalysis Letters 77 (2001) 193.

(

b)

http://www.scopus.com/record/display.url?eid=2-s2.0-0035780648&origin=resultslist&sort=plf-f&src=s&st1=I.+Manuel&nlo=&nlr=&nls=&sid=CD766D47B32FF9707B5C70693D7E7B52.aXczxbyuHHiXgaIW6Ho7g%3a230&sot=b&sdt=cl&cluster=scopubyr%2c%222001%22%2ct&sl=22&s=AUTHOR-NAME%28I.+Manuel%29&relpos=0&relpos=0&searchTerm=AUTHOR-NAME%28I.+Manuel%29+AND+%28+LIMIT-TO%28PUBYEAR%2C2001%29+%29+



CHAPTER 4

Preparation of RhOx/CeyPr1-yO2 N2O

decomposition catalysts by rhodium

nitrate impregnation with different

solvents

The effect of the solvent (water, ethanol or acetone) used to

impregnate CeyPr1-yO2 (y = 1, 0.9 or 0.5) supports with rhodium nitrate

in order to prepare N2O decomposition catalysts has been studied. The

activity for N2O decomposition of the catalysts studied was related with

the RhOx-support interaction, and both the nature of the ceria support

and of the solvent used for rhodium impregnation affected such

interaction. Ceria doping with 10 % praseodymium had a positive effect

on RhOx-support interaction, but the benefit on the catalytic activity

was only obtained for water impregnation. The interaction between

RhOx and Ce0.5Pr0.5O2 was not as good as that with Ce0.9Pr0.1O2. The

best catalyst was obtained by impregnating Ce0.9Pr0.1O2 with a water

solution of rhodium.

Effects of the solvents

107

4.1 Introduction.

In addition to the reasons mentioned in Chapters 1 and 3, ceria

based materials are of interest in catalysis because of their oxygen storage

capacity (OSC) and lattice oxygen mobility. These properties are

dependent of the crystal size and defects as well as could be varied by

ceria doping. Regarding ceria-based oxides used as a metal catalyst

support, they are materials affecting the catalytic behavior of the metals

loaded. In this sense, praseodymium-doped ceria has showed enhanced

performance with regard to pure ceria as rhodium oxide support for

catalytic decomposition of N2O [1]. Incorporation of praseodymium to ceria

modifies the lattice oxygen properties and catalytic performance, due to

the lower metal-oxygen binding energy in the mixed oxide, and the

increase in the number of defects (oxygen vacancies); improving the N2O

decomposition activity of rhodium oxide when combined with the

praseodymium doped ceria [2, 3].

In the previous chapter, the calcination conditions were modified in

order to improve the distribution of rhodium and to enlarge the

rhodium-ceria interface of RhOx/CeO2 catalysts. Improved catalytic activity

for N2O decomposition and CO oxidation was obtained by flash

calcination, which consisted of introducing the ceria support-impregnated

rhodium precursor in a furnace which was pre-heated at 250 ºC. The

solvent used for rhodium precursor impregnation was water, as usually,

and the speed at which water evaporates from the ceria-based support

seemed very important for the catalyst properties, because it affects the

size of the RhOx particles on the final catalyst [4, 5]. This lead us to look for

different procedures to accelerate the evaporation of the rhodium

precursor solvent, and one of the options was to change water by some

other more volatile solvent.

In addition, the features of the solvent used in the impregnation

step not only potentially affects the distribution of the impregnated metal

precursor, due to the evaporation rate, but could also had some other

important roles. The optimum conditions for a successful infiltration of the

support pores mainly depend on the surface polarity, the polarity of the

solvent, and the solubility of the precursor in the solvent. For a surface with

a high density of polar functions a polar solvent will be suitable, in order to

Chapter 4

108

obtain a high degree of wettability and convenient diffusion through the

pores [6]. The polarity and density of the solvent used to impregnate a

metal precursor are also important parameters to be taken into account in

the impregnation of metal precursors on coated substrates, such us

honeycomb monoliths previously coated with a mixture of oxides (alumina,

ceria-based oxides, etc), because the solution must enter into the

substrate channels [7]. Therefore, the conclusions of this study will be also

useful for the preparation of honeycomb monoliths-supported catalysts in

further chapters.

As far as we know, the effect of the solvent used to impregnate

praseodymium doped ceria supports (samples that will be referred as

doped cerias from now on) with a rhodium salt on the properties of the

obtained catalysts has not been reported, and the goal of the current study

is to compare three solvents for such purpose: water, ethanol and acetone.

4.2 Experimental.


Nine catalysts, labelled as RhOx(solvent)/CeyPr1-yO2, were

prepared, the solvent could be water, ethanol or acetone and y takes

values of 1, 0.9 or 0.5. Cerium and praseodymium nitrate precursors

(Ce(NO3)3·6H2O (Aldrich, 99.99 wt.%) and (Pr(NO3)3·6H2O (Aldrich,

99.9 wt.%)) were mixed in an agate mortar to obtain CeO2, Ce0.9Pr0.1O2

and Ce0.5Pr0.5O2 after calcination at 600 ºC for 90 min (heating rate

10 ºC/min). Rhodium was loaded on these oxides by incipient wetness

impregnation with the proper amount of Rh(NO3)3·xH2O (Sigma-Aldrich,

~36 wt.% as Rh) dissolved in water, ethanol or acetone in order to obtain

1 wt.% rhodium on the final catalysts. The catalysts were calcined in flash

conditions, that is, the impregnated supports were introduced in a muffle

furnace that was pre-heated at 250 ºC, and then the temperature was

increased at 10 ºC/min up to 500 ºC (the maximum temperature was

maintained for 30 min).

Additionally, three portions of bare ceria were impregnated with

rhodium nitrate solutions using water, ethanol and acetone, respectively,

and were placed in test tubes with a thermocouple located inside the


109

solids. The tubes were introduced in a vertical furnace that was previously

heated at 250 ºC, and the temperature was registered as a function of

time.


The techniques used for catalysts characterization have been XRD

and Raman spectroscopy to study the crystalline properties of the

catalysts, XPS to determine oxidation states and surface composition,

H2-TPR to know the catalysts reducibility, N2 adsorption isotherms to

determine the BET surface area and TEM to observe RhOx particle size.

The details of the experimental systems and procedures were described in

Chapter 2.


N2O decomposition tests were performed at laboratory scale using

a 100 mL/min flow (GHSV = 42000 h−1

) of 1000 ppm N2O in He, and

100 mg of catalyst. Details of these experiments were described in

Chapter 2.


4.3.1 Catalysts temperature during rhodium nitrate thermal

decomposition.

The effect of the solvent used for rhodium precursor impregnation

on the temperature profile during the further calcination step was studied

as described in section 4.2.1 of the present chapter. The same experiment

was performed with an empty test tube. The temperature profiles

registered are plotted on Figure 4.1.

Chapter 4

110

0

50

100

150

200

250

300

350

0 5 10 15 20 25 30

Tem

pera

ture

(ºC

)

Time (min)

empty test tube

RhOx(acetone)/CeO2

RhOx(H2O)/CeO2

RhOx(ethanol)/CeO2

Figure 4.1. Temperature profiles during the thermal treatment of ceria-impregnated

rhodium precursor (using water, ethanol or acetone as solvent) in test tubes placed in a

vertical furnace pre-heated at 250 ºC.

The temperature profile of the empty test tube shows a sharp

increase during approximately 5 minutes followed by a smooth increase

reaching a constant value around 220 ºC (slightly lower than the set-point

temperature = 250 ºC) after 10 minutes. The temperature profiles of the

impregnated ceria samples are different. In the case of the sample

impregnated with the water solution, the temperature increased until

100 ºC, and reached a plateau at this temperature that can be attributed to

water evaporation. A second increase of temperature occurs afterwards,

reaching the same temperature than the empty tube in 15 min. The

samples impregnated with the ethanol or acetone solutions exhibit a first

step of solvent evaporation (boiling temperature = 78 ºC and 56 ºC for

ethanol and acetone, respectively) but considerable shorter than in the

case of water, since a sharp increase in temperature is observed reaching

a temperature considerably higher than the furnace temperature

(as observed for the empty tube). This behavior is attributed to the

exothermal combustion of the solvent. As it will be appealed for several

times throughout the chapter, the temperature increase ocurred during the

thermal treatment affects the final features of the catalysts impregnated

with acetone or ethanol rhodium solutions.


111


N2O decomposition catalytic tests were performed with the nine

catalysts prepared, and the conversion curves obtained are compiled in

Figure 4.2.

The nature of the ceria-based support and the solvent used for

rhodium precursor impregnation affect the final activity of the catalysts. For

bare ceria, the type of solvent has no effect on the behavior of the

catalysts (Figure 4.2.a). The three RhOx(solvent)/CeO2 catalysts

decompose N2O from 200 ºC approximately and achieve total

decomposition at 375 ºC following the same decomposition profile. On the

contrary, for the catalysts prepared with doped ceria supports the solvent

used for rhodium precursor impregnation strongly modifies their behavior

(Figures 4.2.b and 4.2.c). Regardless the molar fraction of praseodymium

on doped-ceria, the best results were obtained with catalysts impregnated

with the water solution of rhodium nitrate. Both ethanol and acetone

impregnations lead to a significant decrease on the catalytic activity with

regard to the counterpart catalysts impregnated with water. The N2O

decomposition curves obtained for praseodymium-containing catalysts

impregnated with acetone or ethanol rhodium solution were delayed by

50-75 ºC with regard to curves of the catalysts impregnated with the water

solution of rhodium. The effect of the solvent on the physicochemical

properties of the catalysts, and at the end on their catalytic performance, is

analyzed in detail in the coming sections.

Chapter 4

112

0

20

40

60

80

100

200 225 250 275 300 325 350 375 400 425

N2O

de

co

mp

ositio

n(%

)

Temperature (ºC)

RhOx(H2O)/CeO2

RhOx(ethanol)/CeO2

RhOx(acetone)/CeO2

(a)

0

20

40

60

80

100

200 225 250 275 300 325 350 375 400 425

N2O

decom

positio

n(%

)

Temperature (ºC)

RhOx(H2O)/Ce0.9Pr0.1O2

RhOx(ethanol)/Ce0.9Pr0.1O2

RhOx(acetone)/Ce0.9Pr0.1O2

(b)

0

20

40

60

80

100

200 225 250 275 300 325 350 375 400 425

N2O

decom

positio

n(%

)

Temperature (ºC)


(c)



(c)

Figure 4.2. N2O decomposition as a function of temperature for catalysts supported on:

(a) CeO2, (b) Ce0.9Pr0.1O2 and (c) Ce0.5Pr0.5O2.


113

The temperatures required to decompose 50 % of N2O (T50) in

these catalytic tests have been compiled in Table 4.1.

Table 4.1. Temperature required to decompose 50 % of N2O (T50) in the catalytic tests.

Catalyst T50 (ºC)

RhOx(H2O)/CeO2 252

RhOx(ethanol)/CeO2 252

RhOx(acetone)/CeO2 252

RhOx(H2O)/Ce0.9Pr0.1O2 242

RhOx(ethanol)/Ce0.9Pr0.1O2 287

RhOx(acetone)/Ce0.9Pr0.1O2 301

RhOx(H2O)/Ce0.5Pr0.5O2 252

RhOx(ethanol)/Ce0.5Pr0.5O2 326

RhOx(acetone)/Ce0.5Pr0.5O2 319

Comparing the T50 values, it can be concluded that the

impregnation with ethanol or acetone rhodium solutions has a negative

effect on catalysts supported on doped ceria with regard to catalysts

supported on bare ceria. However, using water as the solvent, superior

performance of RhOx(H2O)/Ce0.9Pr0.1O2 is observed with regard to

catalysts with bare and 50 % praseodymium doped ceria supports. The

positive effect of 10 % ceria doping with praseodymium is in agreement

with previous publications [2, 8]. The N2O decomposition capacity of

catalysts prepared by water impregnation of rhodium follows the trend:

RhOx/Ce0.9Pr0.1O2 > RhOx/CeO2 ≈ RhOx/Ce0.5Pr0.5O2

According to this trend, ceria doping with 10 % praseodymium had

a positive effect on the catalytic activity, as already observed [2, 8], while

50 % praseodymium doping had no effect. As it will be discussed

afterwards, the amount of praseodymium not only affects ceria properties

but also the RhOx-ceria interaction and this can explain the observed

trend. For a future work, it will be desirable to perform an optimization

study of the praseodymium amount on the RhOx/CeyPr1-yO2 catalysts

Chapter 4

114

(by using water impregnation of rhodium precursor), but this is out of the

scope of the current study.

4.3.3 Catalysts characterization by N2 adsorption at -196 ºC, XRD and

Raman spectroscopy.

To understand the effect of the solvent used for rhodium precursor

impregnation on the final activity of the catalysts, N2 adsorption at -196 ºC,

XRD and Raman spectroscopy techniques were used to analyze the

physicochemical properties of the materials prepared. The methods used

were described in Chapter 2. These techniques provide (not only but

mainly) information about the properties of the ceria-based supports. The

characterization results obtained are presented in Table 4.2, including the

BET surface area of the supports and catalysts; the ceria supports crystal

size and the lattice parameters of the ceria-based supports determined by

XRD analysis.

Table 4.2. Results of the N2 adsorption and XRD characterization.

Sample

BET

surface

area (m2/g)

Crystal

size

(nm)

Lattice

parameter

(nm)

CeO2 61 - -

RhOx(H2O)/CeO2 60 14 0.5413

RhOx(ethanol)/CeO2 60 14 0.5412

RhOx(acetone)/CeO2 56 15 0.5412

Ce0.9Pr0.1O2 50 - -

RhOx(H2O)/Ce0.9Pr0.1O2 50 18 0.5417

RhOx(ethanol)/Ce0.9Pr0.1O2 31 21 0.5416

RhOx(acetone)/Ce0.9Pr0.1O2 33 21 0.5415

Ce0.5Pr0.5O2 18 - -

RhOx(H2O)/Ce0.5Pr0.5O2 20 24 0.5412




115

The BET surface areas of all catalysts prepared with the un-doped

ceria support are almost equal (56-60 m2/g), regardless the solvent used

for rhodium impregnation, and are also similar to that of the ceria support

(61 m2/g). In accordance with the BET values, the ceria crystal sizes and

ceria lattice parameters corresponding to these three catalysts are also

similar to each other. These results allow concluding that the nature of the

solvent used for rhodium impregnation do not affect the particle size/area

(both parameters are related to each other in this type of oxides) of the

bare ceria support, which is in agreement with the same catalytic activity

obtained with the three praseodymium-free catalysts (see Figure 4.2.a).

The BET surface area of the Ce0.9Pr0.1O2 support is 50 m2/g,

slightly lower than that of the pure ceria. The catalysts with Ce0.9Pr0.1O2

support impregnated with ethanol or acetone solutions present a

considerably lower BET area (31-33 m2/g) than the support, while the

catalyst impregnated with the water solution of rhodium

(RhOx(H2O)/Ce0.9Pr0.1O2) keeps the same BET area than the support

(50 m2/g). These results must be related to the N2O decomposition results

obtained with these three Ce0.9Pr0.1O2-supported catalysts (see

Figure 4.2.b), that is, the highest activity was obtained with the

water-impregnated catalyst (also with the highest BET surface area among

catalysts of this series) and the worse catalytic results were obtained with

ethanol/acetone-impregnated catalysts. As it was previously demonstrated

(see Figure 4.1), temperature gradients are created during the calcination

of catalysts that are impregnated with ethanol or acetone rhodium

solutions (while not with water). These gradients created due to the

exothermic combustion of the solvents should favor Ce0.9Pr0.1O2 sintering

and decrease the activity of the resulting catalysts. The BET surface area

of a Ce0.9Pr0.1O2 sample impregnated with acetone (but without rhodium)

and calcined under the same conditions than the catalysts was 50 m2/g

(the same than that of fresh Ce0.9Pr0.1O2). This is an evidence of the

rhodium role catalyzing the solvents combustion and support sintering.

Finally, the BET area is low and very similar for all

Ce0.5Pr0.5O2-containg catalysts (17-20 m2/g), being also similar to that of

the Ce0.5Pr0.5O2 support. This means that rhodium impregnation and

further calcination do not affects the area of this Ce0.5Pr0.5O2 support,

which is already much lower to that of CeO2 synthesized in equal

conditions (61 m2/g). In this case, a relationship between catalytic activity

Chapter 4

116

and catalyst sintering during calcination is not found, since the BET areas

of all Ce0.5Pr0.5O2-containg catalysts are similar while important differences

on activity were observed (see Figure 4.2.c). As it will be demonstrated by

XPS, TEM and H2-TPR characterization afterwards, the RhOx-CeyPr1-yO2

interaction also plays a key role on the activity of these catalysts, and the

nature of the solvent used for rhodium impregnation affects such

interaction.

As a summary, the BET surface area of CeO2 (61 m2/g) and

Ce0.5Pr0.5O2 (18 m2/g) do not change significantly upon rhodium

impregnation and calcination, regardless the solvent used for rhodium

impregnation, while the area of Ce0.9Pr0.1O2 (50 m2/g) drops (to

31-33 m2/g) upon rhodium impregnation with ethanol or acetone solutions

and further calcination. On the contrary, impregnation with the water

solution of rhodium precursor does not affect the surface area of

Ce0.9Pr0.1O2.

Additional information about the features of the ceria-based

supports was obtained from the XRD diffractograms (Figure 4.3) and

Raman spectra (Figure 4.4) of the catalysts. All X-ray diffractograms only

contain the main reflections typical of a fluorite-structured material with a

face centered cubic unit cell, corresponding to the (111), (200), (220),

(311), (222) and (400) planes. Evidences of segregated phases are not

obvious on Figure 4.3. However, the presence of segregated PrOx species

is difficult to be detected by XRD, because the XRD patterns of such PrOx

species are quite similar to that of ceria [9]. However, in some doped ceria

samples, asymmetric XRD peaks could suggest the presence of

segregated CeO2-rich and PrOx-rich separated phases, but this is not the

case of the diffractograms on Figure 4.3.


117

10 20 30 40 50 60 70 80

Inte

nsity (

a.u

.)

2θ(°)

(a)

(111)

(200)

(220)(311)

(222) (400)

3

1

2

27.5 28 28.5 29 29.5

Inte

nsity (

a.u

.)

2θ(°)

28.54⁰

10 20 30 40 50 60 70 80

Inte

nsity (

a.u

.)

2θ(°)

(b)(111)

(200)(220)

(311)

(222) (400)

6

4

5

27.5 28.5 29.5

Inte

nsity (

a.u

.)

2θ(°)

28.52⁰

10 20 30 40 50 60 70 80

Inte

nsity (

a.u

.)

2θ(°)

9

7

8

(c)(111)

(200)(220)

(311)

(222) (400)

27.5 28 28.5 29 29.5

Inte

nsity (

a.u

.)

2θ(°)

28.50⁰

28.54⁰

Figure 4.3. X-ray diffractograms of catalysts supported on (a) CeO2, (b) Ce0.9Pr0.1O2

and (c) Ce0.5Pr0.5O2. Diffractograms 1, 4, 7 correspond to samples impregnated with

ethanol; 2, 5, 8 to samples impregnated with acetone and 3, 6, 9 to samples

impregnate with water.

Chapter 4

118

The position and shape of the diffraction peaks is quite similar for

all catalysts. See, for instance, the zoom of the (111) peaks inset on

Figure 4.3. As a result, the lattice parameter of the ceria-based supports is

also quite similar for all catalysts (see data on Table 4.2). The expansion

and contraction of the crystal lattice is expected to occur due to ceria

doping with large or small cations, respectively [10]. However, the sizes of

the Ce3+/4+

cations (0.114 nm/0.097 nm) are quite similar to those of the

Pr3+/4+

cations (0.113 nm/0.096 nm), and therefore, the partial substitution

of cerium by praseodymium cations has a minor effect of the lattice

constant of doped ceria. The slightly higher lattice constant values

obtained with some doped ceria catalysts, with regard to values of

catalysts with bare ceria, must be attributed to the presence of more

+3 cations, which are larger than +4 cations, mainly Pr3+

because Pr4+

is

reduced more easily than Ce4+

[11].With regards to crystal size the data

are in accordance to the changes observed in BET surface area.

Raman spectroscopy characterization is consistent with XRD

conclusions. Evidences of praseodymium incorporation into the ceria

framework have been observed by this technique. As a general behavior,

four Raman bands are detected on the spectra included on Figure 4.4, but

these four bands are not clear in all spectra.

The band at 444-463 cm−1

is ascribed to the Raman active F2g

mode of fluorite ceria. This can be viewed as a symmetric breathing mode

of the oxygen atoms surrounding each cation. The intensity of this peak is

highest for catalysts with the bare ceria support (Figure 4.4.a). An slight

deformation of ceria structure can be elucidated because of the

introduction of praseodymium into the ceria structure (in agreement with

the lattice parameter determinated by XRD, Table 4.2), due to the

presence of Pr3+

cations that are bigger than Ce4+

, affecting the oxygen

breathing mode and therefore F2g signal intensity. In addition, fluorescence

produced by praseodymium also diminishes the intensity of the main peak.

Moreover, it is also important to highlight the position of F2g peak, which

undergoes a shift towards lower Raman Shifts when increasing the

praseodymium content, being an evidence of praseodymium introduction

within the fluorite lattice of ceria (Figure 4.4). These results are in

agreement with other reported work [9].


119

150 350 550 750 950 1150 1350

Inte

nsity (

a.u

.)

Raman Shift (cm-1)

RhOx(H2O)/CeO2

RhOx(ethanol)/CeO2

RhOx(acetone)/CeO2

(a) 463 cm-1

150 350 550 750 950 1150 1350

Inte

nsity (

a.u

.)

Raman Shift (cm-1)




(b)

458 cm-1

150 350 550 750 950 1150 1350

Inte

nsity (

a.u

.)

Raman Shift (cm-1)




(c)

444 cm-1

Figure 4.4. Raman spectra of catalysts supported on (a) CeO2, (b) Ce0.9Pr0.1O2 and (c)

Ce0.5Pr0.5O2 (c).

Chapter 4

120

The weak peak at 1170 cm-1

, which is not observed in all catalysts,

has been related to surface oxygen groups [13]. Several interpretations

have been proposed for peaks around 200 cm-1

and 570 cm-1

[2, 8, 10, 12-16 ]. Some authors have assigned peaks at ca. 195 and 570

cm-1

to RhOx species [8, 14-16] and others attributed these bands to the

formation of CeyPr1-yO2 solid solutions, because a physical mixture of the

pure cerium and praseodymium oxides did not show these features [13].

Going into more detail, these bands at 195 and 570 cm−1

have been

assigned to lattice defects, which results in the creation of oxygen

vacancies that is, to the asymmetric oxygen vibration caused by the

formation of vacancies [10, 13].

The creation of vacant sites on ceria would explain the important

increase of the relative intensity of the 570 cm-1

band; with regard to the

intensity of the main F2g mode at 444-463 cm−1

by increasing the

praseodymium content in our catalysts (see Figure 4.4.b and 4.4.c). The

creation of vacant sites on ceria by praseodymium doping is an evidence

of solid solution formation [13].

4.3.4 Catalysts characterization by TEM, XPS and H2-TPR.

The results obtained by TEM, XPS and H2-TPR, presented and

discussed in this section, provide (not only but mainly for the purposes of

the current study) information about the RhOx particles and their interaction

with the ceria-based supports.

TEM images of selected catalysts are included on Figure 4.5. All

the micrographs show the crystals corresponding to the ceria-based

supports. The ceria crystalline planes are even identified in some of them.

The size of the ceria-based crystals seems consistent with the BET areas

of the catalysts included on Table 4.2, that is, the size of the ceria-based

crystals observed by TEM for RhOx(H2O)/CeO2 and

RhOx(H2O)/Ce0.9Pr0.1O2 (60 an 50 m2/g, respectively) are smaller than

those of RhOx(H2O)/Ce0.5Pr0.5O2, RhOx(acetone)/Ce0.9Pr0.1O2 and

RhOx(acetone)/Ce0.5Pr0.5O2 (20, 33 and 18 m2/g, respectively). The ceria

particles size observed by TEM also correlates with the crystal sizes

obtained by XRD (see Table 4.2).


121

Small dark spots (marked with red circles) corresponding to RhOx

nanoparticles are observed in all TEM images as well. The size of the

RhOx particles observed in Figure 4.5 is smaller than 2 nm in most cases.

However, a deeper TEM analysis of these catalysts, and of some others of

similar composition previously studied [5], confirmed the presence of very

small RhOx particles (even smaller than 1 nm) that can be hardly observed

with the magnification used to take the images on Figure 4.5.

The EDS local analysis of the area of the TEM images and the

global analysis of the catalysts by FRX confirmed that the amount of

rhodium is similar in all catalysts, and therefore, the amount of RhOx

particles of ca. 2 nm observed in the TEM images of Figure 4.5 can be

related with RhOx dispersion. If only few RhOx particles are observed in a

certain TEM image is because most RhOx particles are much smaller than

the detection limit, that is, RhOx is highly dispersed. On the contrary, if

there are a lot of RhOx spots in a TEM image of Figure 4.5 is because

RhOx is less dispersed.

Comparing the three TEM images (Figure 4.5) of catalysts

prepared by water impregnation of rhodium nitrate (a) (RhOx(H2O)/CeO2,

(b) RhOx(H2O)/Ce0.9Pr0.1O2 and (c) RhOx(H2O)/Ce0.5Pr0.5O2) it is observed

that there are much more RhOx particles on the catalyst with the

Ce0.5Pr0.5O2 support than on those with Ce0.9Pr0.1O2 or CeO2, and this is an

evidence of the worst RhOx dispersion over Ce0.5Pr0.5O2. This lower RhOx

dispersion on RhOx(H2O)/Ce0.5Pr0.5O2 is coincident with the lowest BET

area of this catalyst (see data on Table 4.2). On the other hand, comparing

the TEM images of the catalysts RhOx(H2O)/Ce0.9Pr0.1O2 and

RhOx(acetone)/Ce0.9Pr0.1O2, more RhOx particles are observed on the

second catalysts, that could be related with the BET area values (60 and

33 m2/g, respectively).

Chapter 4

122

(a)

(b)

(c)(d)

(e)

Figure 4.5. TEM images of catalysts (a) RhOx(H2O)/CeO2, (b) RhOx(H2O)/Ce0.9Pr0.1O2,

(c) RhOx(H2O)/Ce0.5Pr0.5O2, (d) RhOx(acetone)/Ce0.9Pr0.1O2, (e)

RhOx(acetone)/Ce0.5Pr0.5O2. RhOx particles have been circled.


123

The conclusion of this TEM characterization is that both the nature

of the ceria-based support and of the solvent used for rhodium

impregnation affect RhOx dispersion. As a general trend, RhOx dispersion

on CeO2 and Ce0.9Pr0.1O2 is better than on Ce0.5Pr0.5O2, and RhOx

dispersion is better for water-impregnated catalysts than for ethanol or

acetone-impregnated catalysts. Some of these conclusions are supported

by the XPS characterization results.

Figure 4.6 shows the Rh 3d photoelectron spectra of all catalysts.

Two peaks are observed in all spectra, corresponding to the 3d5/2 and 3d3/2

transitions (around 309 and 313 eV, respectively). Both peaks provide

similar information about the oxidation state of rhodium. In all spectra, the

position of these peaks evidence that rhodium is as Rh3+

. The position of

the Rh 3d5/2 peaks is 309.0–310.6 eV for all catalysts tested. As reported in

the literature, the Rh 3d5/2 peak appears at 307.0–307.5 eV for Rh0, at

about 308 eV for Rh+, and from 308.3 to 310.5 eV for Rh

3+ [2, 17-19].

There are subtle differences in the position of the Rh 3d peaks on

Figure 4.6 that deserve a detailed analysis. As explained in Chapter 2, the

binding energy, and therefore the position of each element peaks, depend

on their surroundings. In RhOx-ceria catalysts, there is a negative charge

density transfer from the noble metal to the ceria support, and the extent of

such transfer affects the position of the Rh 3d peaks.

Chapter 4

124

300 305 310 315 320

Inte

nsity (

a.u

.)

Binding Energy (eV)

RhOx(H2O)/CeO2

RhOx(ethanol)/CeO2

RhOx(acetone)/CeO2

(a)

309.4

309.5

309.5

300 305 310 315 320

Inte

nsity (

a.u

.)

Binding Energy (eV)




(b)

309.9

309.8

310.0

300 305 310 315

Inte

nsity (

a.u

.)

Binding Energy (eV)




(c)

309.0

309.0

309.0

Figure 4.6. Rh 3d XPS spectra of catalysts supported on (a) CeO2, (b) Ce0.9Pr0.1O2 and

(c) Ce0.5Pr0.5O2.


125

Regardless the solvent used for rhodium impregnation, the position

of the 3d5/2 peak is 309.5 eV for all RhOx/CeO2 catalysts (Figure 4.6.a),

while it is delayed to slightly higher values for RhOx/Ce0.9Pr0.1O2 catalysts

(Figures 4.6.b) and to lower values for RhOx/Ce0.5Pr0.5O2 (Figures 4.6.c).

This means that the RhOx-support interaction strongly depends on the

support nature, and ceria doping with 10 % praseodymium favors the

RhOx-support interaction while 50 % praseodymium doping hinders the

interaction. This is consistent with the lowest RhOx dispersion (see TEM

images on Figure 4.5 and previous discussion) obtained with the

Ce0.5Pr0.5O2 support and with the lowest BET area of this support

(see Table 4.2).

Table 4.3. Ce3+

and Pr3+

percentages (with regard to total Ce and Pr surface contents,

respectively) and Ce/Pr atomic ratio determined by XPS.

Ce3+

(%) Pr3+

(%) Ce/Pr

RhOx(H2O)/CeO2 37 - -

RhOx(ethanol)/CeO2 35 - -

RhOx(acetone)/CeO2 34 - -

RhOx(H2O)/Ce0.9Pr0.1O2 28 58 4.0



RhOx(H2O)/Ce0.5Pr0.5O2 31 51 0.6



The cerium and praseodymium oxidation states have been

estimated by XPS, and the values obtained are compiled in Table 4.3

together with the Ce/Pr surface ratios. The Ce3+

percentage (with regard to

total surface cerium) was estimated following the method proposed

elsewhere [2, 20] and the Pr3+

percentage (with regard to total surface

praseodymium) by using the semi-quantitative method proposed by

Borchert et al. [21], this procedure was described in Chapter 2. The

percentage of Ce3+

is similar for all catalysts prepared with the bare ceria

support (34-37 %) and these percentages slightly decrease upon

Chapter 4

126

praseodymium doping, evidencing the insertion of the dopant into the ceria

lattice. The Pr3+

percentages are much higher than those of Ce3+

due to

the easier reducibility of Pr4+

with regard to Ce4+

. Moreover, the presence

of Pr3+

cations, in somehow, partially decreases the reduction of Ce4+

.

Regarding the Ce/Pr ratios, they are well below the expected

nominal ratios deduced from the stoichiometric formula of the mixed oxides

(9 for Ce0.9Pr0.1O2 and 1 for Ce0.5Pr0.5O2). Rodríguez-Luque et al. [22]

reported the rhodium nanocrystallites decoration by patches of support in

Rhodium/CeyPr1-yO2 catalysts. It was argued that, during rhodium

impregnation, the acid character of the rhodium solution promotes Pr3+

leaching, and after catalyst drying and calcination, such species are

accumulated on the particles surface. This would explain the preferential

accumulation of praseodymium on the surface of our catalysts.

As a summary, the XPS analysis leads to conclude that

praseodymium is partially inserted into the ceria lattice for catalysts with

doped ceria supports, but with a preferential enrichment of praseodymium

on the particles surface. Such praseodymium doping affects the

RhOx-support interaction, being improved for 10 % praseodymium doping

but hindered for 50 % doping.

The RhOx-support interaction is known to affect the RhOx/ceria

catalysts reducibility, which is closely related to the catalytic activity for

N2O decomposition. In order to study such reducibility, H2-TPR

experiments were performed and the profiles obtained are compiled in

Figure 4.7.

Three peaks are shown in most H2-reduction profiles, as expected

[1, 2, 23]. As already mentioned in Chapter 3, the lowest-temperature peak

can be attributed to the reduction of RhOx, and in some cases, also to the

rhodium-catalyzed ceria-based support surface reduction. The

intermediate-temperature peak is attributed by some authors to surface

ceria reduction as well, but not catalyzed by the noble metal, whereas

other authors relate this peak to surface and/or bulk carbonates

decomposition [24] and/or to surface hydroxyls, peroxides or superoxides

reduction. Finally, the peak at highest temperature is attributed to bulk

ceria-support reduction.


127

0 150 300 450 600 750 900 1050

Inte

nsity (

a.u

.)

Temperature (ºC)

RhOx(H2O)/CeO2

RhOx(ethanol)/CeO2

RhOx(acetone)/CeO2

(a)

0 150 300 450 600 750 900 1050

Inte

nsity (

a.u

.)

Temperature (ºC)




(b)

0 150 300 450 600 750 900 1050

Inte

nsity (

a.u

.)

Temperature (ºC)




(c)

Figure 4.7. H2-TPR profiles of catalysts supported on (a) CeO2, (b) Ce0.9Pr0.1O2 and

(c) Ce0.5Pr0.5O2.

Chapter 4

128

Special attention must be paid to the lowest temperature reduction

peak [2], mainly taking into account the symmetry of the peak. The

catalysts with high activity for N2O decomposition, those prepared with the

bare-ceria support and those impregnated with a water solution of rhodium

(see Figure 4.2), present single low-temperature H2 reduction peaks, while

two overlapped peaks are evident for catalysts with lower activity (those

impregnated with acetone or ethanol solutions in doped ceria supports)

[2, 4, 9, 25].

The presence of double-peaks or pronounced shoulders on the low

temperature H2 reduction peaks occur because Rh3+

, Pr4+

and Ce4+

are

reduced sequentially, while symmetric peaks are obtained if such

reductions occur simultaneously [4]. Therefore, the shape of this peak is

related with the RhOx-support interaction (and with the formation of doped

ceria solid solutions). The catalysts with good RhOx-support interaction

present high catalytic activity [1, 8]. It has been reported that the most

active sites for N2O decomposition are located at the RhOx-ceria interfaces

[26]. Also, a very effective N2O decomposition mechanism has been

demonstrated to occur on RhOx/ceria catalysts, where a synergy between

rhodium and ceria sites is proposed (see discussion on Chapter 3). This

effective mechanism needs a good RhOx-ceria interaction [12].

Regarding the peak at high temperature attributed to bulk

reduction, this band disappeared by increasing the amount of

praseodymium, evidencing improved reducibility with regard to pure ceria

due to increasing oxygen mobility into the catalyst [9].

In conclusion, the activity for N2O decomposition of the catalysts

studied is related with the RhOx-support interaction. Both the nature of the

ceria support and the solvent used for rhodium precursor impregnation

affect such interaction. Ceria doping with 10 % praseodymium has a

positive effect on such interaction (see XPS results; Figure 4.6) and on the

catalytic activity (see Figure 4.2 and Table 4.1), but the benefit on the

catalytic activity is only obtained using water for impregnation. On the

contrary, when Ce0.9Pr01O2 is impregnated with ethanol or acetone

solutions of rhodium, the temperature gradients created during calcination

(Figure 4.1) promote support (see BET areas on Table 4.2) and RhOx

sintering (see RhOx sizes on TEM images; Figure 4.5) hindering the

RhOx-Ce0.9Pr01O2 interaction. The interaction between RhOx and


129

Ce0.5Pr0.5O2 is not as good as that with Ce0.9Pr0.1O2 (see larger RhOx

particles in TEM images (Figure 4.5) and the XPS binding energies of Rh3+

(Figure 4.6), but it seems to be enough to keep a high catalytic activity if

the impregnation is carried out with a rhodium precursor water solution. In

fact, the H2-TPR peak at low temperature (Figure 4.7.c) demonstrates a

considerable RhOx-support interaction. However, acetone or ethanol

impregnation also leads to an important decrease in activity, and this can

be only attributed to RhOx sintering, since Ce0.5Pr0.5O2 with a considerably

lower BET surface area does not sinter with regard to water impregnation

(the same BET areas were obtained for all Ce0.5Pr0.5O2 catalysts; see

Table 4.2). Finally, pure CeO2 is more stable towards sintering than doped

ceria under the calcination conditions of this study (contrarily to the

phenomenon observed at high-temperature calcination), and does not

sinter regardless the solvent used for rhodium impregnation, and for this

reason all the RhOx/CeO2 catalysts kept the same activity.

According to this study, the best catalyst is obtained by

impregnating Ce0.9Pr01O2 with a water solution of rhodium precursor.

However, if acetone or ethanol must be used for any reason (to improve

wettability of a honeycomb monolith channels, for instance) the bare ceria

support is more suitable.

4.4 Conclusions.

The effect of the solvent (water, ethanol or acetone) used to

impregnate CeyPr1-yO2 (y = 1, 0.9 or 0.5) supports with rhodium nitrate, in

order to prepare N2O decomposition catalysts, have been studied and the

following main conclusions can be summarized:

Both the nature of the ceria support and the solvent used for

rhodium precursor impregnation affect RhOx-support interaction,

modifing the activity for N2O decomposition of the catalysts

studied.

The use of ethanol or acetone as solvent has a very negative effect

on Ce0.9Pr0.1O2 and Ce0.5Pr0.5O2-containing catalysts, due to the

sintering of both the support and RhOx particles. This negatively

affects the RhOx-support interaction which directly hinders the

Chapter 4

130

catalytic activity for N2O decomposition. This negative effect is due

to the solvent combustion catalyzed by rhodium.

Ceria doping with 10 % praseodymium has a positive effect on the

RhOx-support interaction, observed as a negative charge density

transfer from the noble metal to the ceria support (only obtained

using water for impregnation) that improves the catalytic activity.

The interaction between RhOx and Ce0.5Pr0.5O2 is not as good as

that with Ce0.9Pr0.1O2, but it seems to be enough to keep a high

catalytic activity if rhodium is impregnated with a water solution.

However, acetone or ethanol impregnation leads to an important

decrease in activity, and this must be attributed to RhOx sintering

because Ce0.5Pr0.5O2 does not sinter additionally with regard to

water impregnation.


131

References.

[1] S. Imamura, J. Tadani, Y. Saito, Y. Okamoto, H. Jindai, C. Kaito.

Decomposition of N2O on Rh-loaded Pr/Ce composite oxides.





[3] A. Trovarelli. Catalysis by Ceria and Related Materials. Catalytic

Science Series. Vol. 2, Imperial College Press (2002).




Rh/Ce0.9Pr0.1O2 catalysts. Applied Catalalysis B: Environmental

107 (2011) 18.

[5] V. Rico-Pérez, M.A. Velasco-Beltrán, Q. He, Q. Wang, C. Salinas-

Martínez de Lecea, A. Bueno-López. Preparation of


catalytic activity for CO oxidation. Catalysis Communications 33

(2013) 47.

[6] J. Roggenbuck, T. Waitz, M. Tiemann. Synthesis of mesoporous

metal oxides by structure replication: Strategies of impregnating

porous matrices with metal salts. Microporous and Mesoporous

Materials 113 (2008) 575.



Reviews - Science and Engineering 43 (2001) 345.





(2010) 370.

Chapter 4

132





Environmental 121–122 (2012) 162.

[10] J.R. McBride, K.C. Hass, B.D. Poindexter, W.H. Weber. Raman

and X-ray studies of Ce1-xRExO2-y where RE=La, Pr, Nd, Eu, Gd,

and Tb. Journal of Applied Physics 76 (1994) 2435.

[11] S. Bernal, G. Blanco, M.A. Cauqui, A. Martiın, J.M. Pintado,

A. Galtayries, R. Sporken. Oxygen buffering capacity (OBC) of

praseodymium-modified CeO2: influence of the Pr distribution in

the ceria host lattice. Surface and Interface analysis 30 (2000) 85.





catalysts. Journal of Catalysis. 276 (2010) 390.

[13] M-F. Luo, Z-L. Yan, L-Y. Jin. Structure and redox properties of

CexPr1−xO2−δ mixed oxides and their catalytic activities for CO,

CH3OH and CH4 combustion. Journal of Molecular Catalysis A:

Chemical 260 (2006) 157.

[14] S. Music, A. Saric, S. Popovic, M. Ivanda. Formation and

characterization of nanosize α-Rh2O3 particles. Journal of

Molecular Structure 924–926 (2009) 221.

[15] M. Kurnatowska, L. Kepinski. Structure and thermal stability of

nanocrystalline Ce1-xRhxO2-y in reducing and oxidizing atmosphere.

Materials Research Bulletin 48 (2013) 852.

[16] V.V. Pushkarev, V.I. Kovalchuk, J.L. d’Itri. Probing defect sites on

the CeO2 surface with dioxygen. Journal of Physical Chemistry B

108 (2004) 5341.

[17] http://www.lasurface.com. Access May 2013



http://www.lasurface.com/


133

[18] X.D. Wu, L.H. Xu, D. Weng. The thermal stability and catalytic

performance of Ce-Zr promoted Rh-Pd/ɣ-Al2O3 automotive

catalysts. Applied Surface Science 221 (2004) 375.

[19] J. Soria, A. Martínez-Arias, J.L.G. Fierro, J.C. Conesa. Effect of

outgassing treatments on the surface reactivity of Rh/CeO2,

catalysts: CO adsorption. Vacuum 46 (1995) 1201.




catalysts. Journal Molecular Catalysis A: Chemical 323 (2010) 52.

[21] H. Borchert, Y.V. Frolova, V.V. Kaichev, I.P. Prosvirin,

G.M. Alikina, A.I. Lukashevich, V.I. Zaikovskii, E.M. Moroz,

S.N. Trukhan, V.P. Ivanov, E.A. Paukshtis, V.I. Bukhtiyarov,

V.A. Sadykov. Electronic and chemical properties of

nanostructured cerium dioxide doped with praseodymium. Journal

of Physical Chemistry B 109 (2005) 5728.

[22] M.P. Rodríguez-Luque, J.C. Hernandez, M.P. Yeste, S. Bernal,

M.A. Cauqui, J.M. Pintado, J.A. Perez-Omil, O. Stephan,

J.J. Calvino, S. Trasobares. Preparation of Rhodium/CexPr1-xO2

catalysts: a nanostructural and nanoanalytical investigation of

surface modifications by transmission and scanning-transmission

electron microscopy. Journal of Physical Chemistry C 112 (2008)

5900.

[23] P. Fornasiero, P. Di Monte, G.R. Rao, J. Kaspar, S. Meriani,

A. Trovarelli, M. Graziani. Rh-loaded CeO2-ZrO2 solid solutions as


behavior and the oxygen storage capacity on the structural

properties. Journal of catalysis 151 (1995) 168.




modified by additive. Applied Catalysis A: General 98 (1993) 99.

Chapter 4

134

[25] K. Krishna, A. Bueno-López, M. Makkee, J.A. Moulijn. Potential

rare earth modified CeO2 catalysts for soot oxidation:

I. Characterization and catalytic activity with O2. Applied Catalysis




temperature N2O dissociation. Studies in Surface Science and

Catalysis 101 (1996) 681.




CHAPTER 5

Preparation, characterization and N2O

decomposition activity of honeycomb

monolith-supported RhOx/Ce0.9Pr0.1O2

catalysts

RhOx/Ce0.9Pr0.1O2 active phases have been loaded by

sequential impregnation on cordierite honeycomb monoliths following

the procedure: (i) cerium and praseodymium nitrates impregnation +

calcinations and (ii) rhodium nitrate impregnation + calcination. The

supported catalysts have been characterized by XRD, Raman

spectroscopy, SEM-EDS, TEM-EDS and H2-TPR, and tested for N2O

decomposition. Rhodium oxide particles are selectively attached to

Ce0.9Pr0.1O2 and not to cordierite.The optimum content of rhodium has

been 0.2 % in total weight. The calcination procedure significantly

affects the supported catalyst features. The best catalyst was prepared

by flash calcination, yielding better distribution of doped ceria in the

cordierite monolith, smaller RhOx particles and lowering the

temperature for surface Rh-Ce-Pr entities reduction in comparison to

ramp calcination. This improves both the distribution of active phases

on the cordierite substrate and the catalytic activity for N2O

decomposition.

Honeycomb monolith-supported RhOx/Ce0.9Pr0.1O2 catalysts

137

5.1 Introduction.

As described in the previous chapters, the RhOx/Ce0.9Pr0.1O2

catalyst is a promising candidate for low temperature N2O decomposition.

In Chapter 3, the calcination conditions were modified in order to improve

the distribution of rhodium and the RhOx-ceria interface of RhOx/CeO2

catalysts. Improved catalytic activity for N2O decomposition and CO

oxidation [1] was obtained by flash calcination, which consisted of

introducing the ceria support impregnated with the rhodium precursor in a

furnace which was pre-heated at 250 ºC. The speed at which water

evaporates from the ceria-based support seemed very important for the

catalyst properties, because it affects the size of the RhOx particles on the

final catalyst. This led us to look for a procedure to accelerate the

evaporation of the solvent used for rhodium impregnation and water was

compared with more volatile solvents (acetone and ethanol) in Chapter 4.

This study was carried out with RhOx catalysts with CeO2, Ce0.9Pr0.1O2 and

Ce0.5Pr0.5O2 supports. The main conclusion obtained was that the nature of

the solvent used for rhodium nitrate impregnation affects the activity of the

catalysts not only due to the evaporation rate but also due to the

exothermic combustion of acetone and ethanol during calcination. This had

a very negative effect on Ce0.9Pr0.1O2 and Ce0.5Pr0.5O2-containing catalysts

due to the induced sintering of these praseodymium-doped ceria and also

of the RhOx particles.

According to these previous studies, the best powder catalyst

prepared until now is RhOx/Ce0.9Pr0.1O2, with rhodium loaded with a water

solution of the nitrate precursor and with further calcination performed in

the so-called flash conditions.

One of the challenges in N2O emission control is to develop a

catalyst able to work in real gas streams. To use a catalyst in real gas

streams it is necessary to load the previously optimized active phase

(powder) into an appropriate inert support. Diverse substrates can be used

for this purpose, such as pellets or inorganic oxide particles, honeycomb

ceramic monoliths, ceramic foams, etc. [2-6]. Among those supports,

honeycomb monoliths present some attractive properties. They provide a

good contact between the active phases of the catalyst and the treated

gases, they present a high dust tolerance, the pressure drop is minimized

Chapter 5

138

because the gas flow is not significantly impeded through the catalytic bed

and also they have no degradation problems typically occurring, for

instance, by particles friction [2, 3, 4, 5].

The most common material used to manufacture honeycomb

monoliths is cordierite (2MgO:5SiO2:2Al2O3). Cordierite presents high

mechanical strength and low thermal expansion coefficient [2, 3, 4].

Incorporation of catalytically active phases into monoliths is not trivial, and

optimization studies of the loading process are required. Some studies of

the decomposition of N2O used a series of monolithic (ceria-alumina

washcoated cordierite) supported transition metal and noble metal oxide

catalysts [7]. Rh/ɣ-Al2O3–sepiolite monolithic catalysts have been studied

[8] and the importance of the drying process in the active phase

distribution into the monolith channels has been reported [2].

The goal of the current chapter is to give a step forward towards

scaling up the utilization of N2O decomposition RhOx/Ce0.9Pr0.1O2 catalysts

in real gas streams. Active phases with composition RhOx/Ce0.9Pr0.1O2

have been loaded into cordierite honeycomb monoliths following a simple

method, consisting of impregnation of the monoliths with the metal

precursors and calcination. The effect of the calcination conditions has

been studied, and the noble metal oxide amount has been also optimized.

The coated monoliths have been characterized and tested for N2O

decomposition in a N2O/He stream. The best supported catalyst has been

also evaluated for N2O decomposition in the presence of the main

inhibitors found in a diesel exhaust stream (O2 + NO + H2O).

5.2 Experimental.


Supported catalysts were prepared by impregnation of cordierite

honeycomb monoliths (by Corning; 1 cm diameter; 1.4 cm length, ~0.6 g

weight) with an aqueous solution of Ce(NO3)3·6H2O (Aldrich, 99.99 wt.%)

and Pr(NO3)3·6H2O (Aldrich, 99.9 wt.%), with the appropriate

concentrations to obtain a Ce:Pr molar ratio of 9:1. After calcination at

600 ºC, the monoliths contain Ce0.9Pr0.1O2 loading amounts of

12.5 ± 0.5 wt.%. Then they were impregnated with an aqueous solution of


139

Rh(NO3)3·xH2O (Sigma-Aldrich, ~36 wt.% as Rh) to obtain the desired

rhodium loading in total base. After that, the catalysts were calcined in

static air at 500 ºC.

Two critical variables have been evaluated in this study: the

rhodium loading and the calcination conditions.Six samples with different

rhodium loading were prepared, denoted by x%RhOx/Ce0.9Pr0.1O2/M

(x = wt.% of rhodium in total basis from 0.001 to 0.9 % and

M = honeycomb monolith). These samples were prepared following the

so-called ramp-calcination process. After impregnation of the cerium and

praseodymium salts, these monoliths were dried at 80 ºC for 120 min, and

then the temperature was raised up to 600 ºC at 10 ºC/min, maintaining

the maximum temperature for 90 min. The calcination of the Ce0.9Pr0.1O2-

containing monoliths after impregnation of the rhodium salt was carried out

by heating the monoliths from room temperature to 230 ºC at 10 ºC/min

and from 230 to 500 ºC at 1ºC/min.

A sample labeled as 0.2%RhOx/Ce0.9Pr0.1O2/M (flash) was

prepared following a similar procedure to that described in Chapter 3. In

this case, after impregnation of the monolith with the cerium and

praseodymium precursors, the calcination was carried out by introducing

the impregnated monolith in a previously heated furnace at 250 ºC,

keeping the monolith at this temperature for 60 min, and increasing the

temperature up to 600 ºC afterwards (heating rate, 10 ºC/min). The

maximum temperature was maintained for 90 min. After cooling, the

monolith was impregnated with a water solution of the rhodium precursor,

and the impregnated monolith was introduced in a furnace previously

heated at 250 ºC. After 60 min at 250 ºC, the temperature was increased

up to 500 ºC (heating rate, 1 ºC/min). This thermal treatment has been

referred to as “flash calcination”.

In order to ensure mechanical resistance of the coated phases in

the supported catalysts, the poorly bound solids were blown out with a

pressurized air stream after the last calcinations step.

A powder sample, denoted by RhOx/Ce0.9Pr0.1O2 (flash), was also

prepared. This sample consisted of a powder active phase with similar

composition to that on the 0.2%RhOx/Ce0.9Pr0.1O2/M (flash) monolith

catalyst. The Ce0.9Pr0.1O2 powder was prepared with the required amounts

Chapter 5

140

of the cerium and praseodymium precursors, which were mixed in a mortar

and calcined in static air. Rhodium was loaded afterwards by incipient

wetness impregnation. The calcination steps were carried out by following

the flash calcination procedure. Table 5.1 compiles general information

about all the catalysts prepared in this study.

Table 5.1. Nomenclature, properties and preparation conditions of the catalysts.

Nomenclature Rh

(%)a

Active phase and

conformation

Calcination

conditions

Ce0.9Pr0.1O2/M 0

Ce0.9Pr0.1O2 active

phase supported into

the monolith

ramp

calcination

0.001%RhOx/Ce0.9Pr0.1O2/M 0.001

RhOx/Ce0.9Pr0.1O2 active


the monolith

ramp

calcination

0.1%RhOx/Ce0.9Pr0.1O2/M 0.1



the monolith

ramp

calcination

0.2%RhOx/Ce0.9Pr0.1O2/M 0.2



the monolith

ramp

calcination

0.6%RhOx/Ce0.9Pr0.1O2/M 0.6



the monolith

ramp

calcination

0.9%RhOx/Ce0.9Pr0.1O2/M 0.9



the monolith

ramp

calcination

RhOx/Ce0.9Pr0.1O2 (flash) 1b

Unsupported


phase (powder)

flash

calcination

0.2%RhOx/Ce0.9Pr0.1O2 /M (flash) 0.2



the monolith

flash

calcination

a Rh content expressed in % of total weight.

b Powder sample; the ratio rhodium to Ce0.9Pr0.1O2 is equivalent to that present in the

supported samples 0.2%RhOx/Ce0.9Pr0.1O2/M (flash) and 0.2%RhOx/Ce0.9Pr0.1O2/M.


141


Catalysts were characterized by Raman spectroscopy and XRD to

determine crystalline structure, N2 adsorption at −196 ºC to calculate BET

surface area, TEM to mesure RhOx particle size, SEM to observe the

active phase distribution on the monolith and H2-TPR to determine the

reducibility of the samples.

The supported catalysts were milled and the powder obtained was

used in all characterization techniques, except in SEM-EDS-chemical

mapping characterization that was used to observe the distribution of the

active phase on the monolith. Detailed information about these

characterization techniques was reported in Chapter 2


N2O decomposition tests were performed in a cylindrical reactor at

atmospheric pressure. The total flow rate was 500 mL/min

(GHSV = 27000 h−1

). N2O decomposition tests were carried out under

different gas streams: 1000 ppm N2O/He or 1000 ppm N2O/1000 ppm

NO/5 % O2/0.6 % H2O/He. For more details see Chapter 2.


5.3.1 SEM-EDS-chemical mapping characterization.

Figure 5.1 shows the micrographs obtained for two monolithic

catalysts prepared by ramp calcination (Figure 5.1.a) and flash calcination

(Figure 5.1.b); samples denoted 0.2%RhOx/Ce0.9Pr0.1O2/M and

0.2%RhOx/Ce0.9Pr0.1O2/M (flash), respectively. In both pictures the active

phase is observed (light grey), coating the monolith walls (dark grey).

Chapter 5

142

(a)

(b)

Figure 5.1. SEM microphotographs of 0.2%RhOx/Ce0.9Pr0.1O2 loaded into honeycomb

monoliths (a) ramp calcination; (b) flash calcination.

As revealed by SEM micrographs, flash calcination procedure

(Figure 5.1.b) allows a more homogeneous active phase coating of the

monolith than ramp calcination. This is attributed to the migration of the

precursor salts solution along the monolith channels during ramp

calcinations, due to the temperature gradients generated. This effect is

minimized during flash calcination. This heterogeneous distribution of the

active phases upon conventional ramp calcination has been also reported

by some other authors for monoliths loaded with different compounds, as

for instance nickel salts [2].

In order to study in detail the effect of calcination conditions in the

coating homogeneity, chemical composition of the monolith surface of

samples 0.2%RhOx/Ce0.9Pr0.1O2/M and 0.2%RhOx/Ce0.9Pr0.1O2/M (flash)

has been determined by EDS analysis. Figure 5.2 shows cerium

concentration (%), which has been selected as representative element of


143

the coating for different channels. The analysis was done in different

positions along the channels and also in different channels of the monolith,

as indicated in the inset scheme of Figure 5. 2.

0

20

40

60

80

100

1 2 3 4 5

Ce (

%)

Channel

(a)Channel

1 8

Position (mm)0 7

0

20

40

60

80

100

1 2 3 4 5

Ce

(%

)

Channel

0 2 4 6 7

(b)

Position (mm)

Figure 5.2. EDS analysis of cerium distribution in the monoliths

0.2%RhOx/Ce0.9Pr0.1O2: (a) ramp calcination; (b) flash calcination.

The EDS analysis in Figure 5.2 also shows a more homogeneous

distribution of cerium using flash calcination (Figure 5.2.b) in comparison to

ramp calcination (Figure 5.2.a). In both samples the amount of cerium

along the central channels (channels 3 and 5) show not relevant

Chapter 5

144

differences in cerium concentration for the different positions. On the

contrary, outer channels of the monoliths (channel 1) have a higher

amount of active phase than inner (central) channels. Cerium

concentration in outer channels oscillates depending on the position

analyzed, and these changes are much more important in the sample

prepared by ramp calcination (Figure 5.2.a).

Chemical mapping was also used to analyze the distribution of the

different components of the coating into the monolith. Figure 5.3.a exhibits

a SEM micrograph of the catalyst 0.2%RhOx/Ce0.9Pr0.1O2 /M (flash) and

the distributions of rhodium (Figure 5.3.b; red), silicon (Figure 5.3.c; blue),

and cerium (Figure 5.3.d; green) were selected as elements representative

of RhOx particles, cordierite and Ce0.9Pr0.1O2, respectively. The active

phase is observed at the top of the pictures, forming a layer of several

micrometers over the cordierite substrate. This analysis points out that

rhodium oxide particles are preferably attached to Ce0.9Pr0.1O2, since the

areas where rhodium appears (Figure 5.3.b) preferably are those where

cerium is also detected (Figure 5.3.d). For a proper interpretation of these

analyses, it must be taken into account that the intensity of rhodium is

lower than that of the other elements, because the concentration of

rhodium is very low and the noise signal for this element is high. This type

of selective RhOx-ceria interaction, and the noble metals dispersion

capacity of ceria, is something very well known that has been deeply

studied in the context of the TWC used in gasoline vehicles for gas

pollution control in this type of vehicles [9].

Previous studies of N2O decomposition carried out with powder

RhOx/ceria catalysts [10-12] showed that rhodium catalytic activity

depends on the ceria promoter effect, and that is why it is important to

ensure a good contact between RhOx and the Ce0.9Pr0.1O2 mixed oxide in

supported catalysts.


145

(a) (b)

(c) (d)

Figure 5.3. SEM-chemical mapping analysis of the sample

0.2%RhOx/Ce0.9Pr0.1O2/M (flash): (a) SEM picture, (b) silicon analysis (blue), (c) cerium

analysis (green), and (d) rhodium analysis (red).

5.3.2 XRD and Raman spectroscopy characterization.

The structure of doped ceria was studied by XRD and Raman

spectroscopy, both techniques providing complementary information. XRD

is sensitive to the position of the cations in the lattice, while Raman spectra

of ceria are caused by oxide anions vibration.

XRD patterns of selected samples are compiled in Figure 5.4,

including the bare cordierite monolith (without active phase), the powder

active phase RhOx/Ce0.9Pr0.1O2 (flash) and the powders obtained upon

milling of the monoliths coated with the active phases

0.2%RhOx/Ce0.9Pr0.1O2 prepared by ramp or flash calcinations.

The powder active phase without cordierite (sample

RhOx/Ce0.9Pr0.1O2 (flash)) presents reflections of the fluorite structure of

Chapter 5

146

ceria, corresponding to the planes (111), (200), (220) and (311). No other

peaks but those of fluorite were observed in the diffractograms, and

evidences of the segregation of praseodymium phases were not obtained

[10-14].

10 20 30 40 50 60 70 80

Inte

nsity (

a.u

.)

2 θ (º)

bare cordierite

0.2%RhOx/Ce0.9Pr0.1O2/M

0.2%RhOx/Ce0.9Pr0.1O2/M (flash)

RhOx/Ce0.9Pr0.1O2 (flash)(111) (200) (220)(311)

Figure 5. 4. XRD patterns of selected samples.

The active phases supported on monoliths,

0.2%RhOx/Ce0.9Pr0.1O2/M and 0.2%RhOx/Ce0.9Pr0.1O2/M (flash), also

present the characteristic peaks of the fluorite structure of ceria, confirming

the incorporation of the active phase into the monolith. These

characteristic peaks of doped ceria appear together with cordierite peaks.

However, overlapping of many of fluorite and cordierite peaks hinders a

detailed analysis of the structure of the supported doped ceria, since the

fluorite peaks broadening cannot be accurately measured. Changes of

relative intensities of peaks at 2 28 º and 30 º in samples

0.2%RhOx/Ce0.9Pr0.1O2/M and 0.2%RhOx/Ce0.9Pr0.1O2/M (flash) with

respect to the bare cordierite correspond to the superposition of (111)

reflexion from doped ceria with that of cordierite at 28 º.

The selected samples have been also characterized by Raman

spectroscopy and the spectra obtained are included in Figure 5.5. The

fluorite structure of ceria, based on the face-centered cubic cell, shows a

characteristic Raman band at ca. 460 cm-1

assigned to the F2g mode [11].

This band is identified in all the Raman spectra included in Figure 5.5,

except in that of the bare monolith. The position of the F2g bands in


147

Figure 5.5 (at ca. 460 cm-1

) is slightly lower to the reference pure ceria

value (465 cm-1

), and this shift in the position is an evidence of the

introduction of praseodymium cations into the ceria lattice, as explained in

Chapter 4. It is worth noting that the position of the F2g peak is exactly the

same both for the pure active phase (RhOx/Ce0.9Pr0.1O2 (flash)) and for the

(0.2%RhOx/Ce0.9Pr0.1O2/M and 0.2%RhOx/Ce0.9Pr0.1O2/M (flash)).

supported samples. This suggests that the introduction of praseodymium

cations into the ceria lattice in coated monoliths is very similar to that

occurring when the active phase is prepared in powder.

200 250 300 350 400 450 500 550 600

Inte

nsity (

a.u

.)

Raman shift (cm-1)

bare monolithF2g mode of ceria (fluorite structure)



RhOx/Ce0.9Pr0.1O2 (flash)

Figure 5.5. Raman Spectra of selected samples.

5.3.3 Characterization by N2 adsorption at -196 ºC.

The BET surface areas of the catalysts are compiled in Table 5.2.

The bare cordierite has a very low surface area (1 m2/g) and the powder

catalyst presents an area of 57 m2/g. The BET area of the supported

catalysts ranges around 4-8 m2/g, and these were the expected values

considering the area of the powder catalyst and the active phase loading

(12.5 ± 0.5 wt.%). These results suggest that the size of the Ce0.9Pr0.1O2

particles (responsible of the BET area) [14] is similar in the powder and in

the supported active phases.

Chapter 5

148

Table 5.2. BET areas of the samples.

Sample BET surface area

(m2/g)

Bare monolith 1

RhOx/Ce0.9Pr0.1O2 (flash) 57

0%RhOx/Ce0.9Pr0.1O2/M 8

0.001%RhOx/Ce0.9Pr0.1O2/M 7

0.1%RhOx/Ce0.9Pr0.1O2/M 6




0.2%RhOx/Ce0.9Pr0.1O2/M (flash) 6

5.3.4 H2-TPR characterization.

Figure 5.6 includes the H2-TPR profiles of catalysts

0.2%RhOx/Ce0.9Pr0.1O2/M and 0.2%RhOx/Ce0.9Pr0.1O2/M (flash). The

species on the catalysts that can be reduced by H2 are the Rh3+

, Pr4+

and

Ce4+

. The phenomena occurring in H2-TPR were described in the previous

chapter. As a brief reminder, first peak at low temperature is attibuted to

RhOx reduction and to the noble metal-catalyzed surface reduction of the

doped ceria supports [2, 13]. Different events can contribute to the peaks

appearing between 200 and 600 ºC as reduction of surface doped ceria

which is not in close contact with RhOx [10], decomposition of surface

carbonates or carbonates occluded within the CeO2 structure [15, 16] and

the reduction of hydroxyls groups or some other surface oxygen groups on

ceria [17, 18].

As mentioned in previous chapters, N2O decomposition capacity of

the catalysts is related with the shape and position of the low temperature

H2 reduction peak [11]. The supported catalyst prepared by flash

calcination (0.2%RhOx/Ce0.9Pr0.1O2/M (flash)) shows a quite symmetric

peak, pointing out a simultaneous reduction of the different cations

(Rh3+

, Ce4+

and Pr4+

) due to a more homogeneity of this catalyst. On the

contrary, the sample prepared by ramp calcination

(0.2%RhOx/Ce0.9Pr0.1O2/M) presents an asymmetric peak with a shoulder


149

at low temperature revealing the successive reduction of the different

cations. It is reasonable to think that Rh3+

reduction takes place first,

followed by the Pr4+

and Ce4+

reduction [11, 19, 20]. These results suggest

that the flash calcination procedure allows obtaining a much better noble

metal-support interaction than the conventional ramp calcinations, which is

consistent with the conclusions of Chapter 3 performed with powder

RhOx/CeO2 catalysts.

0 200 400 600 800

TC

D S

ignal (a

.u.)

Temperature (ºC)



50 100 150 200

TC

D S

ignal (a

.u.)

Temperature (ºC)

Figure 5.6. H2-TPR profiles of selected samples.

5.3.5 TEM characterization.

The supported catalysts 0.2%RhOx/Ce0.9Pr0.1O2/M and

0.2%RhOx/Ce0.9Pr0.1O2/M (flash) were studied by TEM microscopy and

Figure 5.7 shows, as an example, a picture of each catalyst.

Chapter 5

150

5 nm

(a) (b)

5 nm

Figure 5.7. TEM photograph of catalysts (a) 0.2%RhOx/Ce0.9Pr0.1O2/M and

(b) 0.2%RhOx/Ce0.9Pr0.1O2 /M (flash). Red circles show RhOx particles.

In the catalyst prepared by ramp calcination (Figure 5.7.a) RhOx

particles of ca. 2 nm (see red circles) are clearly distinguished from those

of Ce0.9Pr0.1O2, which exhibit a well-defined crystalline structure that is

observed in some areas. On the contrary, in the catalyst prepared by flash

calcination (Figure 5.7.b) RhOx particles have been hardly observed,

suggesting a much smaller particle size, and therefore a better dispersion

of rhodium than for the catalyst prepared by ramp calcination. A similar

conclusion was achieved in Chapter 3, where smaller RhOx particles were

observed by TEM on powder RhOx/CeO2 catalysts prepared by flash

calcination with regard to those on a similar catalyst calcined in ramp.

In order to confirm that the absence of RhOx particles in Figure

5.7.b can be really attributed to the small size of the particles and to rule

out the possibility that RhOx is not observed because is heterogeneously

distributed on the sample EDS analyses of both samples have been

performed. The EDS measurements, included in Table 5.3, confirm the

presence of similar amounts of rhodium in both samples (2.1 and

1.7 wt.%).

The small particle size of RhOx in the sample prepared by flash

calcinations (0.2%RhOx/Ce0.9Pr0.1O2/M (flash)) is consistent with the

conclusions of the H2-TPR experiments (Figure 5.6), that is, the small size


151

of the RhOx particles allows a good contact with the support and improves

the reduction of the Ce4+

and Pr4+

surrounding the noble metal particles.

Table 5.3. EDS chemical analysis of photographs shown in Figure 5.7. The

percentages are expressed by weight.

Sample Rh (%) Ce (%) Pr (%) O (%)

0.2%RhOx/Ce0.9Pr0.1O2/M 2.1 71.5 7.1 19.3

0.2%RhOx/Ce0.9Pr0.1O2/M (flash) 1.7 72.4 8.9 17.0


The effect of the noble metal loading on the catalytic

decomposition of N2O has been studied with the set of catalysts prepared

by ramp calcination using a N2O/He stream. The N2O decomposition

profiles obtained in these experiments are compiled in Figure 5.8, where

N2O decomposition (%) is plotted versus the rhodium content of the

catalysts for different reaction temperatures. In agreement with previous

studies carried out with powder active phases [10, 11], the supported

sample without rhodium (Ce0.9Pr0.1O2/M) has no catalytic activity in the

range of temperature studied, while all the catalysts with rhodium are able

to decompose N2O, even the catalyst with very low noble metal loading

(0.001%RhOx/Ce0.9Pr0.1O2/M). As expected, N2O decomposition increases

with temperature, almost reaching the complete conversion at 350 ºC for

all the catalysts. According to these results, the best formulation is

0.2%RhOx/Ce0.9Pr0.1O2/M because the addition of higher amount of

rhodium does not improve the conversion attained at each temperature.

This means that the specific activity of rhodium decreases by increasing

the rhodium loading above 0.2 wt.%.

The best formulation has been selected to study the effect of the

calcination conditions in the catalytic performance of the coated monoliths.

In Figure 5.9 the profiles obtained with the supported catalysts

0.2%RhOx/Ce0.9Pr0.1O2/M and 0.2%RhOx/Ce0.9Pr0.1O2/M (flash), and also

the results obtained with the powder catalyst RhOx/Ce0.9Pr0.1O2 (flash) are

compiled.

Chapter 5

152

0

25

50

75

100

0.0 0.2 0.4 0.6 0.8 1.0

N2O

decom

positio

n (

%)

Rh (%)

350 ºC

325 ºC

300 ºC

275 ºC

250 ºC225 ºC

Figure 5.8. N2O decomposition (%) in a 1000 ppm N2O/He stream versus rhodium

content at different temperatures for x%RhOx/Ce0.9Pr0.1O2/M catalysts where

0 ≤ x ≤ 0.9.

Comparing the N2O decomposition profiles obtained with the two

catalysts (Figure 5.9) prepared by flash calcinations (powder and

supported); it is observed that the N2O decomposition onset temperature is

the same. This means that the procedure followed to load the active

phases into the monoliths is able to mimic the properties of the powder

active phase. However, after certain level of decomposition (ca. 15 %), the

supported catalyst (0.2%RhOx/Ce0.9Pr0.1O2 /M (flash)) outperforms the

conversions of the powder active phase (RhOx/Ce0.9Pr0.1O2 (flash)),

successfully achieving the first objective of this chapter.

The shape of the N2O decomposition curve obtained with the

powder sample evidences an important contribution of the diffusion

phenomena to the N2O decomposition rate, while this behavior is not

obvious for the supported catalysts. This is not surprising since monolith

catalysts are characterized by their good mass-transfer in comparison to

randomly packed beds [2]. This observation could have practical

relevance, and honeycomb monoliths could be more suitable supports to

be used in real applications than randomly packed beds, like those of

particles or pellets. These diffusion phenomena is observed in this chapter

but not in previous ones due to the much lower ratio between mass of

active phase and flow of gas used in the experiments (current chapter;


153

75 mg of catalyst active phase and 500 mL/min of gas; previous chapters:

100 mg of catalyst active phase and 100 mL/min of gas).

Regarding the calcination treatments, the supported catalyst

prepared by ramp calcination is much less effective for N2O decomposition

than the counterpart prepared by flash calcination, with 100 ºC of delay in

the N2O decomposition temperature at 50 % conversion. This can be

attributed to the improved interaction between RhOx and Ce0.9Pr0.1O2

particles obtained by flash calcinations, as deduced from H2-TPR and

TEM-EDS characterization. It is interesting that the flash calcination is not

only advantageous in preparing powder catalysts, as discussed in

Chapter 3, but also in preparing monolith supported catalysts.

0

25

50

75

100

200 250 300 350 400

N2O

deco

mp

ositio

n

(%)

Temperature (ºC)

0.2%RhOx/Ce0.9Pr0.1O2/M (flash)circles:1st runstars: 2nd run


RhOx/Ce0.9Pr0.1O2 (flash)

Figure 5.9. N2O decomposition (%) as a function of temperature (in a 1000 ppm

N2O/He stream) for catalysts with similar amount of rhodium and different calcination

procedure and shape.

In order to evaluate the stability of the supported catalysts two

consecutive cycles of N2O decomposition experiments were conducted

with the 0.2%RhOx/Ce0.9Pr0.1O2/M (flash) catalyst. The results of these

experiments, also included in Figure 5.9, were equal, proving that the

heating and cooling cycles do not modify the catalyst performance.

Finally, the best catalyst (0.2%RhOx/Ce0.9Pr0.1O2/M (flash)) among

all those prepared in this study has been tested using a complex gas

mixture which includes the main inhibitors typically present in a diesel

Chapter 5

154

exhaust stream (NOx, O2, and H2O). Null NOx removal was observed in

these experiments once the steady-state conditions were achieved, and

only N2O decomposition was detected. The N2O decomposition results

obtained are compiled in Figure 5.10. As expected, the presence of

inhibitors in the gas mixture shifts the N2O decomposition curve to higher

temperatures in comparison with the pure N2O/He stream. For the complex

gas mixture evaluated, the N2O decomposition started at 350 ºC and total

conversion was reached at 525 ºC. In spite of the important delay in the

N2O decomposition curve due to the presence of catalyst inhibitors

observed in Figure 5.10, the range of temperature obtained could be

suitable for N2O removal in a diesel engine exhaust with the N2O

decomposition catalyst located after a NOx reduction SCR catalyst. This

hypothesis will be confirmed in the next chapter.

0

25

50

75

100

100 150 200 250 300 350 400 450 500 550

N2O

decom

positio

n (

%)

Temperature (ºC)

1000 ppm N2O1000 ppm N2O1000 ppm NO5% O2

0.6% H2O

Figure 5.10. Effect of reactive gas atmosphere in the catalytic activity for N2O

decomposition. Tests conducted with 1000 ppm N2O/He or with 1000 ppm N2O/

1000 ppm NO/5 % O2/0.6 % H2O/He using the catalyst 0.2%RhOx/Ce0.9Pr0.1O2/M

(flash).

The inhibiting effect of NOx, O2 and H2O on the catalytic activity for

N2O decomposition of 0.2%RhOx/Ce0.9Pr0.1O2/M (flash) observed in

Figure 5.10 is consistent with a previous DRIFTS study performed with ɣ-

Al2O3 particles-supported Rh/Ce0.9Pr0.1O2 [21]. Among these three

inhibitors, the strongest effect was attributed to NOx while the weakest to

O2. The inhibiting effect of O2 was attributed to its reversible chemisorption

on catalyst sites, while the effect of H2O and NOx was mainly related with


155

the irreversible chemisorption on the catalysts sites. The inhibiting effect of

H2O is not as high as that of NOx because the product of H2O

chemisorption (Ce–OH surface groups) is suitable for N2O chemisorption

and decomposition, while the surface nitrogen species created upon NOx

chemisorption are not [21].

5.4 Conclusions.

RhOx/Ce0.9Pr0.1O2 active phases have been loaded into cordierite

honeycomb monoliths, and the effects of the rhodium amount and of the

calcination conditions have been studied. The characterization and

catalytic test for the N2O decomposition of the supported catalysts allow

the following conclusions:

The nature of Ce0.9Pr0.1O2 in powder and supported catalysts

seems to be similar.

In honeycomb monoliths-supported catalysts, RhOx is preferentially

attached to Ce0.9Pr0.1O2, and not to the cordierite substrate.

Monolith supported catalyst has similar N2O decomposition

capacity than powder active phases of the same composition at

low N2O conversions (below ~15 % for the conditions of the current

study), but they outperform the conversion of powder catalysts due

to the improved mass transfer of the supported catalysts.

For the experimental conditions used in this study, the optimum

content of rhodium in total weight base is 0.2 wt.%. Increasing the

amount above this loading does not increase the N2O

decomposition capacity.

The distribution of the supported active phases on the cordierite

substrate depends on the calcinations conditions. More

homogeneous coatings are obtained by flash calcination in

comparison to conventional ramp calcinations.

The calcination procedure also affects the Rh-Ce-Pr interactions.

Flash calcination yields smaller RhOx particles and improves the

low temperature reduction of surface Rh-Ce-Pr entities in

Chapter 5

156

comparison to ramp calcination, which in turn improves the

catalytic activity for N2O decomposition.

The catalyst denoted by 0.2%RhOx/Ce0.9Pr0.1O2/M (flash) has

proved to be able to decompose N2O in presence of NO, O2 and

H2O.

The main goal of this work has been suscesfully accomplished and

N2O decomposition RhOx/Ce0.9Pr0.1O2 catalysts have been

successfully supported on honeycomb monoliths and have been

tested in simulated gas streams in the presence of typical

inhibitors.


157

References.





(2013) 47.



Reviews – Science and Engineering 43 (2001) 345.


Carbon-based monolithic structures. Catalysis Reviews – Science


[4] C. Moreno-Castilla, A.F. Pérez-Cadenas. Carbon-based


Materials 3 (2010) 1203.

[5] A.E.W. Beers, T.A. Nijhuis, N. Aalders, F. Kapteijn, J.A. Moulijn,

BEA coating of structured supports-Performance in acylation.


[6] K. Galejová, L. Obalová, K. Jirátová, K. Pacultová, F. Kovanda.



(2009) 172.

[7] V. Boissel, S. Tahir, C.A. Koh. Catalytic decomposition of N2O over

monolithic supported noble metal-transition metal oxides. Applied

Catalysis B: Environmental 64 (2006) 234.

[8] S. Suárez, C. Saiz, M. Yates, J.A. Martin, P. Avila, J. Blanco.

Rh/γ-Al2O3–sepiolite monolithic catalysts for decomposition of N2O

traces. Applied Catalysis B: Environmental 55 (2005) 57.







Chapter 5

158




[11] S. Parres-Esclapez, M.J. Illán-Gómez, C.Salinas-Martínez de




(2010) 370.






[13] J. Silvestre-Albero, F. Rodríguez-Reinoso, A. Sepúlveda-

Escribano. Improved metal-support interaction in Pt/CeO2–SiO2

catalysts after zinc addition. Journal of Catalysis 210 (2002) 127.

[14] C. Bueno-Ferrer, S. Parres-Esclapez, D. Lozano-Castelló,

A. Bueno-López. Relationship between surface area and crystal

size of pure and doped cerium oxides. Journal of Rare Earths 28

(2010) 647.




modified by additives. Applied Catalysis A: General 98 (1993) 99.

[16] P.Fornasiero, P. Di Monte, G.R. Rao, J.Kaspar, S. Meriani,

A. Trovarelli, M. Graziani. Rh-loadead CeO2-ZrO2 solid solutions as


behavior and the oxygen storage capacity on the structural

properties. Journal of catalysis 151 (1995)168.

[17] A. Trovarelli. Catalysis by ceria and related materials. Catalytic






159

[18] S. Bernal, J.J. Calvino, G.A. Cifredo, J.M. Gatica, J.A. Pérez-Omil,

J.M. Pintado. Hydrogen chemisorption on ceria: influence of the

oxide surface area and degree of reduction. Journal of the

Chemical Society, Faraday Transactions 89 (1993) 3499.

[19] K. Krishna, A. Bueno-López, M. Makkee, J.A. Moulijn. Potential

rare earth modified CeO2 catalysts for soot oxidation:

I. Characterization and catalytic activity with O2. Applied Catalysis






Environmental 121–122 (2012) 162.









CHAPTER 6

NOx reduction to N2 with commercial

fuel in a real diesel engine exhaust

using a dual bed which constists of

Pt/Beta zeolite and RhOx/ceria

monolith catalysts

In this chapter medium-size and full-size Pt/Beta zeolite

monolith catalysts have been prepared and tested in SCR technology

for NOx abatement in real diesel engines exhausts. A medium-size

dual bed catalytic system consisting of a Pt/Beta zeolite/monolith SCR

catalyst located upstream a N2O decomposition

RhOx/Ce0.9Pr0.1O2/honeycomb monolith catalyst has been also

successfully prepared and tested for NOx reduction to N2. The

dip-coating method was optimized for the Pt/Beta zeolite/monolith

preparation. The catalyst RhOx/Ce0.9Pr0.1O2/monolith was prepared by

nitrate precursor decomposition. The production of N2O as undesired

NOx reduction product, which is a drawback of platinum SCR

catalysts, has been solved by using the dual bed configuration, where

both monolith catalysts operated at the same temperature, and 100 %

N2 selectivity has been obtained.

Medium and full-size catalysts for deNOx and deN2O

163

6.1 Introduction.

As reported in Chapter 1, NOx removal in diesel-engine exhausts

can be mainly accomplished by two technologies: SCR (Selective Catalytic

Reduction) and NOx Storage Reduction (NSR) 1-3.

The main drawback of NOx removal on diesel vehicles is the

oxygen excess in the exhaust. For this reason, in both technologies (SCR

and NSR) a reducing agent must be fed into the exhaust stream, which

must react with NOx in the presence of excess O2. Different reducing

agents have been studied, most of them at laboratory scale, including H2,

CO, different HCs, NH3, urea, etc. 1-9].

In the NSR process 1-3, the NOx removal is carried out in cycles

of storage and reduction steps. During the storage steps, which occur in

normal driving conditions, NOx is chemisorbed on a basic oxide present on

the catalyst. During the reduction steps, the reducing gas is fed and reacts

with the nitrogen compounds previously stored (and also with O2).

In the SCR process [2-9], the reducing agent is continuously fed

into the gas stream and a selective catalyst is used to promote the reaction

of the reducing gas with NOx with respect to the unproductive direct O2

combustion.

During the last years, many research efforts have been focussed

on the utilization of NH3 or urea as reductants, in somehow inspired by the

well-established technology used in power plants, and some commercial

systems are now available for stationary diesel engines and heavy-duty

diesel vehicles [2, 6-9]. However, this technology has not been

successfully developed for light-duty vehicles, and one of the reasons is

the weight penalty of the additional urea/NH3 tank. In addition, a network of

urea/NH3 suppliers should be available to fill up the tanks.

An alternative option that overcomes these limitations of the

urea/NH3 technology would be the utilization of diesel fuel itself as reducing

agent. Diesel fuel is already on board of diesel vehicles, and therefore,

additional tanks and filling facilities are not required. The SCR of NOx has

been studied in detail with several HCs of different nature (propene

[10-12], octane [13], methanol [12], ethanol [12], acetaldehyde [14] and

Chapter 6

164

decane [11]. Nevertheless, most of these studies were performed under

laboratory conditions, but for real diesel exhausts the number of scientific

articles describing NOx removal studies is quite limited 2, 15, 16].

Recently, Cho et al. [17] have successfully tested in a real diesel

exhaust a NOx reduction system consisting of a plasma reactor, a diesel

fuel reformer and a dual-bed catalytic reactor. In this system the diesel fuel

reformer produces highly reactive oxygenated HCs for NOx reduction.

However, as far as we know, direct reduction of NOx with commercial

diesel fuel has not been reported.

In previous studies performed in our group, small-size (1 cm

diameter) cordierite honeycomb monoliths were coated with Pt/Beta zeolite

active phase and this supported catalyst was successfully tested for SCR

of NOx with propene at laboratory conditions [18, 19]. The activity of such

coated monolith was quite promising (NOx conversions above 80 % were

achieved), and the main drawback was that N2O was the main NOx

reduction product (while N2 is the desired one). In the study reported in the

previous chapter, small-size cordierite honeycomb monoliths were coated

with RhOx/Ce0.9Pr0.1O2 and successfully tested for low temperature N2O

decomposition at laboratory as well [20]. These results lead us to think that

a dual bed of Pt/Beta zeolite/monolith and RhOx/Ce0.9Pr0.1O2/monolith

could be a suitable option for NOx conversion into N2, with both catalytic

beds operation at the same temperature and in a range of temperatures

suitable for a diesel engine exhaust (around 200-250 ºC for propene as

reductant, and at higher temperatures for heavier HCs, as it will be later

demonstrated in this chapter). The dual bed configuration was previously

tested but using two catalytic systems working at different temperatures

[21].

One of the goals of the current study is to coat larger cordierite

honeycomb monoliths (medium-size monoliths; diameter 2.3 cm) with

Pt/Beta zeolite and to test the SCR of NOx in a real diesel engine exhaust

by using commercial diesel fuel as reducing agent. This system will be

scaled up into a full-size Pt/Beta zeolite monolith catalyst in order to

demonstrate and report, for the first time, that the SCR of NOx can be

performed in a real diesel exhaust stream by commercial diesel fuel using

this catalyst.


165

Another goal is to prepare a medium-size (diameter 2.3 cm)

RhOx/Ce0.9Pr0.1O2/monolith catalyst and to test the proposed dual bed

configuration (Pt/Beta zeolite/monolith and RhOx/Ce0.9Pr0.1O2/monolith) in

a real diesel exhaust to ensure that N2O is not released as undesired NOx

reduction by-product.

6.2 Experimental details.


Cordierite honeycomb monoliths of different size have been used

in this study, which were shown in Figure 2.18 (Chapter 2).

The small-size monolith was used in Chapter 5 and previous

studies [18, 19, 20], and the medium and full-size monoliths have been

used in the current chapter.

The medium-size monoliths were supplied by CTI (Céramiques

Techniques et Industrielles; France) and the dimensions are 2.3 cm

diameter and 7.5 cm length, with 400 cpsi3. These monoliths were used

both as Pt/Beta zeolite support and as RhOx/Ce0.9Pr0.1O2 support.

The full-size monoliths were supplied by Corning and the

dimensions are 14.4 cm diameter and 14 cm length, with 400 cpsi. One of

these monoliths was used as Pt/Beta zeolite support.

6.2.1.1 Medium-size Pt/Beta zeolite/monolith catalysts preparation.

Beta zeolite was loaded into the medium-size honeycomb

monoliths by dip-coating [22-24], with a water suspension of commercial

powder ammonia Beta zeolite (provided by Zeolyst International),

surfactant (Teepol) and binder (a 40 wt.% suspension of colloidal silica in

water; Ludox AS-40). A detailed study of the dip-coating process was

performed and different variables were evaluated. Suspensions with

different proportions of zeolite, surfactant and binder were studied.

3 cpsi: cell pieces per square inch (1 inch = 2.54 cm).

Chapter 6

166

To obtain homogeneously dispersed mixtures, the slurries were

stirred with a high-shear mixer (UltraTurrax T50 from IKA Labortechnik) for

15 minutes at 8000 rpm. The effect of the zeolite, surfactant and binder

amounts on the slurry viscosity was studied with a viscometer SV10 from

A&D.

The monoliths were dipped into the slurries with a home-made

dip-coater (Figure 6.1) at a rate of 1.5 cm/min, both for introducing the

monoliths into the slurries and to take them out. Monoliths coating either in

a single step or in several consecutive dipping steps were studied.

Figure 6.1. Home-made dip-coater with accessories for different size of monoliths.

In all cases, the coated monoliths were first dried overnight at room

temperature while rotating in horizontal position on the device shown in

Figure 6.2, and afterwards, at 90 ºC for 1 hour in static horizontal position.

Finally, selected monoliths were heat-treated in air in a furnace by heating

from room temperature to 200 ºC at 1 ºC/min and from 200 to 500 ºC at

10 ºC/min, keeping the maximum temperature for 4 hours.


167

Figure 6.2. Home-made rotating dryer.

A Beta zeolite-coated monolith (with 3.3 wt.% of zeolite coated in

two steps, 0.04 gzeolite/cm3monolith, by using fresh Beta zeolite slurry of

composition: 9 wt.% Beta zeolite, 0.24 wt.% binder and 0.36 wt.%

surfactant; viscosity 1.3 mPas) was selected for platinum loading. This

monolith was prepared with the optimal conditions obtained in section

6.3.3 of the current chapter.

[Pt(NH3)4](NO3)2 (Alfa-Aesar, 99.99 wt.%) was used as platinum

precursor, and the nominal platinum loading was 1 wt.% on zeolite basis.

The amount of water used to dissolve the platinum precursor and to

impregnate the monolith was the minimum amount that ensured no water

dropping upon impregnation. This amount was determined experimentally,

being 4.5 gwater/gzeolite. 50 % of the platinum solution was homogeneously

dropped with a pipette to one of the end sides of the Beta coated

substrate, and afterwards, the remaining 50 % was dropped to the other

one. The impregnated monolith was dried at 90 ºC in vertical position for

1 hour, changing the position of the monolith up-side-down every

10 minutes. After this step the monolith was apparently dry and was kept

overnight at the same temperature in horizontal position. Finally, the

monolith was calcined in air at 500 ºC (heating rate 5 ºC/min) keeping the

maximum temperature for 2 hours. Figure 6.3 shows the medium-size

Pt/Beta zeolite/monolith tested in the SCR experiments with commercial

diesel fuel.

Chapter 6

168

Figure 6.3. Photograph of the medium-size Pt/Beta zeolite/monolith tested in the SCR

experiments with commercial diesel fuel.

6.2.1.2 Full-size Pt/Beta zeolite/monolith catalyst preparation.

Beta zeolite was also loaded into a full-size honeycomb monolith

by dip-coating 22, 23, 24 with a water suspension of commercial powder

ammonia Beta zeolite (provided by Zeolyst International), using Teepol as

surfactant and a 40 wt.% suspension of colloidal silica in water; Ludox

AS-40 as binder. The optimized suspension was obtained from the study

described in section 6.3.3.

To obtain a homogeneously dispersed mixture, the slurry was

stirred with the mixer UltraTurrax T50 for 15 minutes at 1500 rpm. After

5 minutes stabilization, the slurry viscosity was 1.3 mPa·s. Using this

viscosity, blowing the suspension excess with compressed air was not

necessary.

The monolith was dipped into the slurry with a home-made

dip-coater (Figure 6.1) at a rate of 1.5 cm/min, both for introduction of the

monolith into the slurry and to take it out. Four consecutive dipping steps

were performed.

The coated monolith was dried overnight at room temperature

while rotating in horizontal position. Afterwards, it was heat-treated in air in

a furnace by heating from room temperature to 200 ºC at 1 ºC/min and


169

from 200 to 500 ºC at 10 ºC/min, keeping the maximum temperature for

4 hours. The Beta zeolite loading was 10.5 wt.% after calcination

(0.05 gzeolite/cm3monolith).

Finally, the Beta zeolite-coated monolith was impregnated with a

water solution of [Pt(NH3)4](NO3)2 (Alfa-Aesar 99.99 wt.%) to obtain 1 wt.%

platinum loading on zeolite basis. The amount of water used to dissolve

the platinum precursor and to impregnate the monolith was also the

minimum amount that ensured no dropping upon impregnation

(4.5 gwater/gzeolite). 50 % of the platinum solution was homogeneously

dropped with a pipette to one of the end sides of the Beta zeolite-coated

substrate, and afterwards, the remaining 50 % was dropped to the other

one. The impregnated monolith was dried at 90 ºC in vertical position,

changing the position of the monolith up-side-down every 30 minutes for

4 hours. After this time the monolith was apparently dried and was kept

overnight at 90 ºC in horizontal position. Finally, the monolith was calcined

in air at 500 ºC (heating rate 5 ºC/min) keeping the maximum temperature

for 2 hours. Figure 6.4 shows a picture of the full-size honeycomb monolith

loaded with Pt/Beta zeolite.

Figure 6.4. Photograph of the full-size Pt/Beta zeolite/monolith.

Chapter 6

170

6.2.1.3 Medium-size RhOx/Ce0.9Pr0.1O2/monolith catalyst preparation.

The active phase RhOx/Ce0.9Pr0.1O2 was loaded on a medium-size

honeycomb monolith by using Ce(NO3)3·6H2O (Alfa-Aesar, 99.5 wt.%),

Pr(NO3)3·6H2O (Alfa-Aesar, 99.9 wt.%) and Rh(NO3)3·xH2O

(Sigma–Aldrich, ∼36 wt.% as rhodium). Figure 6.5 shows a picture of the

honeycomb monolith loaded with RhOx/Ce0.9Pr0.1O2.

Figure 6.5. Medium-size RhOx/Ce0.9Pr0.1O2/monolith.

The preparation conditions were similar to those described as flash

calcination for a small-size (1 cm diameter) monolith, in Chapter 5,

corresponding to the optimum catalyst. The impregnated monolith, with

cerium and praseodymium precursors water solution, was introduced in

the furnace at 250 ºC for 1 hour and then was flash calcined (at 600 ºC for

90 min, using a heating rate of 10 ºC/min).The concentration of the

solution used for the impregnation was fixed to obtain 3 wt.% of

Ce0.9Pr0.1O2 loading (0.04 gCe0.9Pr0.1O2/cm3monolith). This weight increase

was confirmed gravimetrically.

Afterwards, the Ce0.9Pr0.1O2-containing monolith was impregnated

with a water solution of the rhodium precursor, with the proper

concentration to obtain 1 wt.% rhodium loading on Ce0.9Pr0.1O2 basis,

following the procedure previously described for platinum impregnation on


171

Beta zeolite-containing monoliths. After rhodium impregnation the monolith

was introduced in a furnace which was previously heated at 250 ºC, and

after 1 hour at this temperature, the temperature was increased at

1 ºC/min until 500 ºC [25].

6.2.2 SEM characterization.

SEM pictures of selected coated monoliths were taken. This

characterization technique and the equipment used were described in

Chapter 2.

6.2.3 Catalytic tests.

SCR experiments were performed with the medium and full-size

Pt/Beta zeolite/monolith catalysts. Dual-bed experiments for simultaneous

deNOx and deN2O were also performed with the medium-size

RhOx/Ce0.9Pr0.1O2/monolith catalyst located downstream the medium-size

Pt/Beta zeolite/monolith catalysts. Experiments with medium-size monolith

catalysts were performed at the University of Alicante and SCR

experiments with the full-size monolith catalyst were carried out at the

facilities of the Galician Automotive Technology Centre (Centro

Tecnológico de Automoción de Galicia, CTAG) as described in Chapter 2.


6.3.1 Study of the Beta zeolite suspensions viscosity.

Figure 6.6 shows the change of viscosity with time for several Beta

zeolite water suspensions prepared with different Beta zeolite and

surfactant concentrations. As a general trend, viscosity decreased with

time until a constant value was achieved. Both the zeolite and surfactant

concentrations affected the change of viscosity with time and the stable

viscosity value achieved. The viscosity decrease with time was attributed

to zeolite sedimentation, which in most cases occurred during few minutes

after the mixer was stopped.

The surfactant concentration significantly affected viscosity

changes with time. As expected, this effect was more important for the

Chapter 6

172

zeolite suspension with the highest zeolite loading (18 and 24 wt.% are

compared in Figure 6.6). As a general trend, the higher the surfactant

concentration the faster the stable viscosity level was achieved.

0

20

40

60

80

0 10 20 30 40 50 60

Vis

cosity (

mP

a.s

)

Time (min)

24 wt. % Beta zeolita + 0.25 wt. % binder

18 wt. % Beta zeolita + 0.25 wt. % binder

Surfactant (%)

0 0.40.8 1.6

Surfactant (%)

0 0.40.8 1.6

Figure 6.6. Change of viscosity with time of several Beta zeolite water suspensions of

different composition. (Time = 0 min when the mixer was stopped).

The practical conclusion from results on Figure 6.6 was that, for

dip-coating of monoliths on this type of suspensions, a certain stabilization

time was necessary before the monolith can be dipped, and this time

ranged from few minutes (3 at least) until 1 hour, depending on the zeolite

and surfactant concentrations.

As deduced from Figure 6.6, the zeolite concentration had an

important effect on the suspension viscosity, the higher the zeolite

concentration, the higher the suspension viscosity. This important variable

was studied in more detail. Figure 6.7 compiles viscosity values for several

Beta zeolite water suspensions of different composition, with zeolite

concentrations between 18 and 27 wt.%. The viscosity data were

measured after the time required in each case to obtain stable values. The

effect of surfactant and binder concentrations was evaluated as well. The

results on Figure 6.7 allowed concluding that the zeolite amount had the

strongest effect on viscosity. For instance, viscosity values ranged from 13

to 110 mPa·s for zeolite concentrations between 18 and 27 wt.%

respectively. However, for a particular zeolite concentration, the change of

the surfactant concentration between 0 and 1.6 wt.% only induced a


173

viscosity change of ± 5 mPa·s and the change of binder concentration

between 0.24 and 0.46 wt.% only modified viscosity in ± 2 mPa·s.

0

20

40

60

80

100

120

0 0.5 1 1.5

Vis

co

sity (

mP

a.s

)

Surfactant (wt.%)

27 wt.% beta zeolite + 0.24 wt.% binder






Figure 6.7. Viscosity of Beta zeolite water suspensions with different compositions. (All

data were measured after the time required in each case to obtain stable viscosity

values).

The relationship between zeolite concentration and suspension

viscosity is plotted on Figure 6.8 for suspensions with similar surfactant

and binder concentrations. A wider range of zeolite concentrations to that

explored on Figure 6.7 was plotted on Figure 6.8, where an exponential

relationship is observed. The increase of viscosity was small for zeolite

concentrations between 5 and 13 wt.% (from 1 to 3 mPa·s respectively)

while viscosity increased drastically above 18 wt.% zeolite concentration.

Chapter 6

174

0

20

40

60

80

100

120

0 5 10 15 20 25 30

Vis

cosity (

mP

a.s

)

Beta zeolite (wt.%)

0.24 wt.% binder0.36 wt.% surfactant

Figure 6.8. Effect of the Beta zeolite concentration on suspension viscosity, (all data

were measured after the time required in each case to obtain stable viscosity values).

6.3.2 Effect of the Beta zeolite suspension viscosity on medium-size

monoliths dip-coating.

Once the effect of the Beta zeolite concentration on suspensions

viscosity was studied, honeycomb monoliths were dipped on Beta zeolite

suspensions of different composition and the monolith weight increase was

determined gravimetrically after drying the coated monoliths, calcination at

500 ºC and blowing the loosely bound zeolite with pressurised air. Figure

6.9 shows the effect of the Beta zeolite concentration (Figure 6.9.a) and

viscosity of the suspension (Figure 6.9.b) on the monolith weight increase

after a single dip-coating step.

A linear relationship between zeolite concentration and weight

increase was obtained for Beta zeolite suspensions up to 20 wt.%

concentration, corresponding to viscosities ≤ 23 mPa·s. Above this

viscosity the zeolite suspension was not able to penetrate into the monolith

channels and the weight increase dropped to zero. In a review article [26]

was reported that the length of honeycomb monoliths that can be

washcoated is mainly controlled by the viscosity of the washcoating

solution, being possible to washcoat monolith bodies up to 25 cm long with

solutions of viscosity below 30 mPa·s. However, our current results

evidence that this general rule do not apply to Beta zeolite suspensions,

since a threshold at 23 mPa·s was found for honeycomb monoliths of


175

7.5 cm length. In this conditions the maximum zeolite loading obtained was

1.5 wt.%.

0.0

0.3

0.6

0.9

1.2

1.5

1.8

0 5 10 15 20 25 30

Weig

ht

incre

ase (

%)

Beta zeolite (wt.%)

(a)0.24 wt.% binder0.36 wt.% surfactant

0.0

0.3

0.6

0.9

1.2

1.5

1.8

0 20 40 60 80 100 120

We

igh

t in

cre

ase

(%

)

Viscosity (mPa·s)

(b)


Figure 6.9. Effect of the Beta zeolita concentration (a) and viscosity of the suspension

(b) on the weight increase of the monolith after one dip-coating step (after drying,

calcinations at 500 ºC, and blowing the loosely bound zeolite with pressurised air).

6.3.3 Medium-size monolith coating with Beta zeolite in consecutive

dipping steps.

Since the maximum weight increase achieved in the dipping

conditions used for Figure 6.9 experiments was 1.5 wt.%, different

Chapter 6

176

strategies were explored in order to increase the amount of Beta zeolite

loaded on the monolith.

In the experiments reported in section 6.3.2 it was observed that

an important proportion of the Beta zeolite remained on the water

suspension after dipping the monolith. In order to analyze the potential

reutilization of the same Beta zeolite suspension, several consecutive

coating steps were carried out with the same slurry. Figure 6.10 compiles

the monolith weight increase achieved in consecutive coating steps

performed with the same monolith and with the same Beta zeolite

suspension (circles). These data were obtained after drying the monolith

and blowing the loosely bound zeolite with pressurised air (this procedure

was repeated after every dipping step).

0

1

2

3

4

5

6

7

8

9

0 1 2 3 4 5 6 7 8 9

Weig

ht

incre

ase (

%)

Number of dip-coating steps

9 wt.% Beta zeolite


Figure 6.10. Monoliths weight increase after several dip-coating steps. Circles: using

the same Beta zeolite suspension in all steps. The monolith was dried and the loosely

bound zeolite was blown with pressurised air after each dip-coating step; Squares:

Fresh Beta zeolite suspension was used in each dip-coating step. (Solid squares:

weight increase after dip-coating and drying; Open squares: weight after drying,

calcination at 500 ºC and air blowing).

The amount of zeolite loaded on the monolith increased

progressively with the number of dip-coating steps until a maximum

loading of 2.8 wt.% was achieved after eight steps. As observed, the

weight of the monolith did not increase linearly. This was attributed to the

progressive depletion of the zeolite on the water suspension. In

conclusion, the Beta zeolite suspension could be used in consecutive


177

dipping steps but the yield of the process would be lower to that obtained

by using fresh suspension on each step.

In order to increase the zeolite loading, four consecutive dipping

steps were performed with a new monolith, using fresh Beta zeolite

suspension on each step. The effect of calcination was also analyzed, and

the monolith was calcined at 500 ºC after the second and fourth steps. The

monolith weight increase achieved on these conditions is also included on

Figure 6.10 (squares).

The monolith weight increased progressively in consecutive

coating steps, and as expected, the zeolite loading was higher

(Figure 6.10; squares) to that obtained using the same suspension for

several times (Figure 6.10; circles). The maximum zeolite loading achieved

using fresh Beta zeolite suspension on each step was about 7.6 wt.%

(after four dipping steps), while only 2.8 wt.% was achieved with the same

zeolite suspension. Note that the zeolite loading increase almost linearly

during the first three dipping steps performed with fresh Beta zeolite

suspension (Figure 6.10, squares) but in the fourth one the loading

decrease from that tendency. This is because the monolith channels

become narrower upon zeolite accumulation, and this hinders further

zeolite loading. Actually, after the fourth dip-coating step part of the

monolith channels were closed, as it will be shown below.

Special attention must be paid to the weight increase in the first

step. The different weight increase observed in both series of Figure 6.10

must be attributed to the fact that, in one case (solid squares) the monolith

was not blown with pressurised air after drying, while it was in the other

series (circles). According to this comparison, a significant amount of

zeolite (the loading decreased from 2 to 0.6 %) was removed upon blowing

with pressurised air, and this amount of zeolite removed is wasted.

However, the proportion of zeolite removed by air blowing after calcination

at 500 ºC was much lower (see open squares), and this procedure is

therefore more convenient. These results evidenced that the blowing step

performed after drying is not really necessary if the monolith channels are

not blocked by zeolite, and blowing the loosely bound zeolite could be

done after calcination.

Chapter 6

178

Opening up the monolith channels with pressurised air is a habitual

procedure which is typically done after dip-coating and before calcination

[22-24, 26-29]. This is done in this way because the channels usually

become blocked after dip-coating, since high viscosity suspensions are

preferred in order to achieve high washcoat loading. Opening blocked

channels after calcination, once the binder has sintered, is almost

impossible for long channels and that is why air blowing is usually done

before calcination. According to our optimization study, an alternative

procedure consists of avoiding channels blockage by using low-viscosity

suspensions, and therefore the cleaning step can be done after

calcination, once the binder has sintered and the zeolite layer has been

stabilized. In this case, several consecutive dipping steps are required in

order to increase the active phase loading, and the goal of air blowing is

not to open closed channels but to remove loosely bound zeolite.

6.3.4 SEM characterization of medium-size monolith catalysts.

Figure 6.11 shows SEM images of selected coated monoliths.

Figures 6.11.a and 6.11.b show general views of several channels of the

Beta zeolite coated monoliths with 3.3 and 7.6 wt.% coating, respectively,

corresponding to data represented by open squares on Figure 6.10

(calcined samples). A detail of the few micrometers coating layer is also

shown in the inset of Figure 6.11.a. It is observed that the channels of the

monolith with lower coating content (3.3. wt.%; Figure 6.11.a) remain open

while those of the monolith with higher coating loading (7.6 wt.%; Figure

6.11.b) are partially or totally blocked.

Taking into account the optimization study, the dip-coating

conditions selected to prepare the medium and full-size

Pt/Beta zeolite/monolith catalysts were those corresponding to the sample

with 3.3 wt.% of zeolite (one of the open squares of Figure 6.10, coated in

two steps by using fresh Beta zeolite on each step).


179

Figure 6.11 SEM images of selected coated monoliths. (a) Beta zeolite coated

monolith with 3.3 wt.% coating, (b) Beta zeolite coated monolith with 7.6 wt.% coating

and (c) RhOx/Ce0.9Pr0.1O2/monolith catalyst. Figures 6.11.a and 6.11.b correspond to

open squares on Figure 6.10; Figures 6.11.a and 6.11.c are the samples used in the

catalytic tests sections 6.3.5 and 6.3.7.

Chapter 6

180

Figure 6.11.c shows pictures of the RhOx/Ce0.9Pr0.1O2/monolith

catalyst, where it is clearly observed that channels are also open in this

case. This is in accordance with our previous study performed to optimize

the preparation of this kind of supporter catalysts (Chapter 5 and reference

[20]).

6.3.5 SCR experiments performed with commercial diesel fuel and

the medium-size Pt/Beta zeolite/monolith catalyst.

Figure 6.12 shows results of the SCR experiments performed with

different diesel fuel concentrations and the medium-size Pt/Beta

zeolite/monolith catalyst at reaction temperatures between 300 and

400 ºC. NOx removal curves (Figure 6.12.a) present the typical

volcano-shape profile and the THC consumption (Figure 6.12.c) increases

with temperature [3, 4, 9]. A maximum in the curves of NOx removal level

versus temperature is typically reported in SCR NOx experiments

performed with model HCs. The temperature of the maximum and the

maximum NOx removal level achieved depend both on the nature of the

catalysts and the nature and concentration of the HC. Below the maximum

temperature the NOx-HC reaction takes place and the process is selective

in a certain extent, while above the maximum the selectivity decreases and

the HCs are consumed by undesired O2 combustion with less contribution

of the NOx-HC reaction pathway. The optimum temperature for platinum

catalysts is typically lower than that of some other transition metal catalysts

such us cupper or cobalt catalyst [4].

In our experiments, the optimum temperature for maximum NOx

removal was 350 ºC for all THC concentrations tested, and the maximum

NOx removal level achieved in these experimental conditions was 50 % for

3400-4500 ppm THC inlet concentration. The temperature for maximum

NOx removal is higher to that typically obtained with propene as reducing

agent and Pt/Beta zeolite catalysts tested at laboratory under simulated

gas streams (210-230 ºC) [18, 30]. This can be attributed to several

experimental differences between our previous laboratory experiments

[18, 30] and the current pilot plant tests: (i) catalyst active phase to gas

flow ratio, (ii) nature of the reducing agent and (iii) gas stream composition

(laboratory streams only included NOx and 5 % O2 in an inert carrier while

the real stream also includes H2O, CO2 and much more O2 (17.2 %)).


181

0

20

40

60

80

275 300 325 350 375 400 425

NO

x r

em

oval (%

)

Catalyst temperature (ºC)

45003400220013006001500

THC inlet concentration (ppm)

(a)

0

20

40

60

80

100

275 300 325 350 375 400 425

N2

sele

ctivity (

%)


4500

3400

2200

1300

600


(b)

0

20

40

60

80

100

275 300 325 350 375 400 425

TH

C r

em

oval (%

)


45003400220013006001500


(c)

Figure 6.12. SCR experiments performed with commercial diesel fuel and a

Pt/Beta zeolite/monolith catalyst (a) NOx removal, (b) N2 selectivity and (c) THC

removal.

Chapter 6

182

These results demonstrate that the SCR of NOx with diesel fuel is

feasible in a real exhaust by using a medium-size Pt/Beta zeolite/monolith

catalyst. After this proof of concept, next step is to scale up this catalyst

into a full-size catalyst, as described in section 6.2.1.2, and to test it.

One of the weak points of this technology is the reduction of part of

the NOx to N2O instead of to N2, which is the innocuous and desired

reaction product. This has been reported by several authors working at

laboratory [18, 30-33]. The selectivity results presented on Figure 6.12.b

show that, in our experimental conditions, N2 is always the main NOx

reduction product (N2 selectivity ≥ 67 %) but N2O is detected at 300 ºC for

all THC inlet concentrations tested and at 350 ºC for the low THC inlet

concentrations (600 and 1300 ppm THC inlet concentration) which are

desirable from the economic point of view and also to obtain the lower HC

emission as observed in Figure 6.12.c. The reason for the high N2

selectivity achieved in the current power bench experiments must be the

higher operation temperature compared to other laboratory experiments

where platinum-catalysts work below 250 ºC [18, 30]. However, in order to

avoid N2O emission dual-bed experiments have been performed and are

described afterwards.

6.3.6 SCR experiments performed with commercial diesel fuel and

the full-size Pt/Beta zeolite/monolith catalyst.

Figure 6.13 compiles the results obtained in SCR experiments

performed with the full-size Pt/Beta zeolite/monolith catalyst. THC and CO

removal increased with temperature (Figures 6.13.a and 6.13.b,

respectively) and NOx removal (Figure 6.13.c) followed a typical

volcano-shape profile, with maxima at 250 ºC for all THC inlet

concentrations studied. This behavior is similar to that typically observed in

laboratory experiments performed with model HCs 30-32.


183

0

20

40

60

80

100

200 225 250 275 300 325

TH

C r

em

ova

l (%

)

Temperature at catalyst inlet (ºC)

1000

2000

3000

4000

5000


(a)

0

20

40

60

80

100

200 225 250 275 300 325

CO

re

mo

va

l (%

)


1000

2000

3000

4000

5000


(b)

0

5

10

15

20

25

30

35

200 225 250 275 300 325

NO

x r

em

oval (%

)


1000

2000

3000

4000

5000


(c)

Figure 6.13. Catalytic results obtained at steady state with a full-size

Pt/Beta zeolite/monolith catalyst upon injection of different amounts of commercial

diesel fuel versus temperature at catalyst inlet.

Chapter 6

184

As expected, the profiles of NOx removal (volcano shape) in the

full size catalyst behaves in similar way than medium-size catalyst

(section 6.3.5). However, in the experiments performed with the full-size

monolith catalyst, it must be also taken into account that the inlet NOx

concentration increases with the exhaust temperature (see data on

Table 2.1), and this also affects the shape of the NOx removal profiles on

Figure 6.13.c.

The temperature for maximum NOx removal (250 ºC) was only

slightly higher than values previously obtained in laboratory experiments

with 1 cm diameter Pt/Beta zeolite/honeycomb monolith by using propene

as model hydrocarbon (210-230 ºC) 18, 19, confirming that the Pt/Beta

zeolite monolith catalyst is a good candidate for low temperature NOx

reduction with commercial diesel fuel. Maximum temperature for NOx

removal is lower in this section than in the previous one, although as

explained before this temperature depends on several factors.

0

5

10

15

20

25

30

35

0 1000 2000 3000 4000 5000 6000

NO

x r

em

ova

l (%

)


Temperature = 300 ºC411 ppm NOx

Temperature = 220 ºC97 ppm NOx

Temperature = 250 ºC 218 ppm NOx

at catalyst inlet :

Figure 6.14. Effect of the amount of fuel injected on NOx removal.

For a better observation of the effect of the amount of fuel injected

in the NOx removal, the NOx conversion has been represented versus

THC inlet concentration, for several inlet temperatures, in Figure 6.14. An

optimum concentration at 3000 ppm THC inlet concentration is observed.

Only considering the Le Chatelier's principle one would expect a

continuous increase of NOx removal with THC inlet concentration, but this


185

only occurred below 3000 ppm THC. Two arguments have been proposed

to explain the NOx removal decrease at high THC concentration 16. In

one hand, for the HC-NOx reactions to occur on a catalyst surface the

reactants must be chemisorbed on the catalyst, and a huge excess of one

of the reactants (THC in this case) hinders the chemisorption of the

remaining gases, inducing a certain poisoning effect. On the other hand,

the HC combustion reactions are highly exothermic, and the increase of

temperature within the catalyst promotes the HC-O2 reaction, decreasing

the selectivity of the process. This effect is evidenced in Figure 6.15,

where the temperature at the catalyst outlet is plotted versus the THC inlet

concentration. In most experimental conditions, the gas temperature

decreases within the catalytic bed, since most outlet temperatures are

below the corresponding inlet temperatures. However, an increasing trend

was obtained with the THC concentration for experiments performed at the

same inlet temperature due to the exothermic character of the THC

combustion reaction, therefore affecting the SCR selectivity.

200

250

300

1500 2000 2500 3000 3500 4000 4500

Te

mp

era

ture

at

ca

taly

st o

utle

t (º

C)


Temperature at catalyst inlet = 300 ºC



Figure 6.15. Effect of the amount of fuel injected on catalyst temperature.

The results obtained in this study confirm that the SCR of NOx can

be successfully achieved by commercial diesel fuel in real light-duty

vehicles with a Pt/Beta zeolite/monolith catalyst, therefore avoiding

additional tanks on board which are required for some other reductants like

urea or NH3. After this proof of concept, next step for further studies would

be to optimize the experimental variables in order to achieve NOx removal

levels as high as possible, for instance, optimization of the catalyst

Chapter 6

186

preparation, platinum loading, fuel injection, catalyst temperature control,

utilization of several consecutive small catalysts instead of in a single bed

with fuel injection before each catalyst, etc.

6.3.7 SCR experiments performed at 300 ºC with commercial diesel

fuel and a dual-bed which contains Pt/Beta zeolite and

RhOx/Ce0.9Pr0.1O2 medium-size monolith catalysts.

In spite of the good results presented in section 6.3.5, in order to

achieve 100 % N2 selectivity, dual bed experiments were performed with

the medium-size Pt/Beta zeolite/monolith catalyst (located upstream) and

the RhOx/Ce0.9Pr0.1O2/monolith catalyst (located downstream). Both

medium-size monoliths were placed on the same reactor and operate at

300 ºC. This temperature was selected because was the temperature

where more N2O was released in the single-bed SCR experiments (see

Figure 6.12.b).

Figure 6.16 compiles the NOx removal (Figure 6.16.a) and N2

selectivity (Figure 6.16.b) values obtained for different THC inlet

concentrations, where single and dual-bed configuration experiments are

compared. The N2 selectivity results obtained with the dual-bed

configuration (Figure 6.16.b) confirm that the rhodium catalyst avoids N2O

emission as undesired NOx reduction product, achieving 100 % N2

selectivity for all THC inlet concentrations tested. As far as we know, it is

the first time that this achievement is reported. A dual-bed configuration

was already proposed by Pérez-Ramírez et al. [21] in order to improve the

low N2 selectivity of SCR platinum catalysts, and several N2O

decomposition catalysts were evaluated in that study (Co-Rh and Co-Pd

mixed oxides derived from hydrotalcite-like compounds and ion-exchanged

Fe-ZSM-5 and Pd-ZSM-5 zeolites). That study was performed under

laboratory conditions by using propene as model HC. In that case the

optimum temperatures for the first and second beds were 200 and 425 ºC

respectively, and this ruled out the practical utilization of the mentioned

N2O decomposition catalysts in vehicles because the gas stream should

be heated up between the deNOx and deN2O catalysts. In that study [21],

it was noted that a much more active catalysts was required for practical

application of the dual-bed system to a realistic car exhaust situation, and

the results reported in the current study suggest that the


187

RhOx/Ce0.9Pr0.1O2/monolith catalysts is a suitable candidate to accomplish

this task.

0

10

20

30

40

50

0 1000 2000 3000 4000

NO

x r

em

oval (%

)


Pt/beta/honeycomb monolith +RhOx/Ce0.9Pr0.1O2/honeycomb monolith

Pt/beta/honeycomb monolith

Pt/beta zeolite/monolith + RhOx/Ce0.9Pr0.1O2/monolith

Pt/beta zeolite/monolith

(a)

0

20

40

60

80

100

0 1000 2000 3000 4000

N2

se

lectivity (

%)


Pt/beta+Rh/CePr

Pt/beta

Pt/beta zeolite/monolith + RhOx/Ce0.9Pr0.1O2/monolith

Pt/beta zeolite/monolith

(b)

Figure 6.16. SCR experiments performed at 300 ºC with commercial diesel fuel and a

dual bed of medium-size Pt/Beta zeolite/monolith and RhOx/Ce0.9Pr0.1O2/monolith

catalysts (a) NOx removal and (b) N2 selectivity.

The dual-bed configuration studied in the current study has a

certain benefit on NOx removal. For low THC inlet concentrations (THC

inlet concentration ≤ 2200 ppm) the NOx removal levels achieved with the

single and dual-bed configuration are quite similar, increasing with the

THC inlet concentration. This means that the RhOx/Ce0.9Pr0.1O2/monolith

catalyst do not contributes to NOx removal in this range of THC

concentrations. On the contrary, the NOx removal level remained more or

Chapter 6

188

less constant in the experiments performed with the single bed for THC

inlet concentration > 2200 ppm, because the THC excess poisons the

platinum catalyst, while increases monotonically in the dual bed

configuration experiments. Therefore, for high THC excess the

RhOx/Ce0.9Pr0.1O2/monolith catalyst contributes to NOx removal, which is

an additional benefit of the dual bed configuration. These results are in

agreement with the conclusions reported by Kotsifa et al. [34], who

compared the SCR of NO by propene over supported platinum and

rhodium catalysts. It was reported [34] that rhodium catalysts are active in

NOx SCR and selective for N2 production at low oxygen partial pressures,

but under severe oxidizing environments they lose their activity while

maintain N2 selectivity. This explains why our rhodium catalyst only

showed SCR activity at high THC inlet concentration, but improved N2

selectivity regardless the THC inlet concentration. On the contrary, Kotsifa

et al. [34] concluded that platinum catalysts are not able to catalyse NO

reduction at low oxygen partial pressures due to the accumulation of

hydrocarbonaceous intermediates on the metal surface, which hinder

adsorption and dissociation of NO. The reaction is promoted by higher O2

concentrations, where activation of HC enables efficient regeneration of

the catalytically active sites, in agreement with the proposed mechanism

[35]. The decrease of N2 selectivity of platinum catalysts with increasing O2

concentration was related with the higher coverage of adsorbed atomic

oxygen.

6.4 Conclusions.

The main conclusions of the current chapter can be summarized

as follows:

The preparation of the Pt/Beta zeolite/monolith has been optimized

by adjusting the viscosity of surfactant and binder-containing

Beta zeolite suspensions (viscosities ≤ 23 mPa·s). The effect of

zeolite concentration on slurry viscosity was much more relevant

than those of surfactant and binder.

A certain stabilization time was necessary before the monolith can

be dipped on Beta zeolite slurries, and this time ranged from few


189

minutes (3 at least) until 1 hour, depending on the zeolite and

surfactant concentrations.

The blowing step performed after drying is not really necessary if

the monolith channels are not blocked by zeolite, and blowing the

loosely bound zeolite could be done after calcination, but several

consecutive dipping steps are required in order to obtain the Beta

zeolite loading suitable for a practical application (3.3 wt.% in the

current study).

SCR experiments performed in real diesel exhausts with medium

and full-size Pt/Beta zeolite/monolith catalysts demonstrated that

NOx reduction is feasible with commercial diesel fuel.

The SCR behaviors observed in the real exhausts with commercial

diesel fuel were similar to that typically reported in laboratory

experiments performed with model HC.

The N2 selectivity of the medium-size Pt/Beta zeolite/monolith

catalyst in the studied conditions is significantly higher than

previous laboratory results, but the 100 % N2 selectivity is only

obtained by using RhOx/Ce0.9Pr0.1O2/monolith catalyst located

downstream the Pt/Beta zeolite/monolith catalyst, with both

monolith catalysts (deNOx and deN2O) operating at the same

temperature.

Chapter 6

190

References.

[1] W.S. Epling, L.E. Campbell, A. Yezerets, N.W. Currier, J.E. Parks.

II. Overview of the fundamental reactions and degradation

mechanisms of NOx storage/reduction catalysts. Catalysis



lean NO(χ) catalysis. Applied Catalysis B: Environmental 13 (1997)

1.

[3] R. Burch. Knowledge and know-how in emission control for mobile

applications. Catalysis Reviews – Science and Engineering 46

(2004) 271.

[4] Y. Traa, B. Burger, J. Weitkamp. Zeolite-based materials for the

selective catalytic reduction of NOx with hydrocarbons.

Microporous and Mesoporous Materials 30 (1999) 3.

[5] P.L.T. Gabrielsson. Urea-SCR in automotive applications. Topics in

Catalysis 28 (2004) 177.

[6] T. Johnson. Diesel engine emissions and their control: an

overview. Platinum Metals Review 52 (2008) 23.




[8] M.V. Twigg. Progress and future challenges in controlling



[9] Z. Liu, S.I. Woo. Recent advances in catalytic deNOx science and


(2006) 43.


191




(2000) 137.



deactivation of Ag/Al2O3 and Cu/ZSM-5 for HC- SCR under


230 (2005) 150.







oxygen-rich exhaust gases: combined heterogeneous and surface-

mediated homogeneous reactions. Journal of Catalysis 219 (2003)

25.




[15] M. Makkee, H.C. Krijnsen, S.S. Bertin, H.P.A. Calis, C.M. Van Den

Bleek, J.A. Moulijn. Bench-scale demonstration of an integrated

deSoot-deNOx system. Catalysis Today 75 (2002) 459.



demonstration of the Pt-catalyzed C3H6-SCR of NOx in a diesel


[17] B.K. Cho, J.H. Lee, C.C. Crellin, K.L. Olson, D.L. Hilden,

M.K. Kim, P.S. Kim, I. Heo, S.H. Oh, I.S. Nam. Selective catalytic

reduction of NOx by diesel fuel: plasma-assisted HC/SCR system.




Chapter 6

192


J.M. García-Cortés, M.J. Illán-Gómez, C. Salinas-Martínez de


by in situ synthesis. Utilization as Pt support for NOx abatement in


[19] A. Bueno-López, M.J. Illán-Gómez, C. Salinas-Martínez de Lecea.


Pt-Beta-coated cordierite monoliths for deNOx C3H6-SCR. Applied

Catalysis A: General 302 (2006) 244.




Rh/Ce0.9Pr0.1O2 catalysts. Applied Catalalysis B: Environmental

107 (2011) 18.

[21] J. Pérez-Ramírez, J.M. García-Cortés, F. Kapteijn, M.J. Illán-

Gómez, A. Ribera, C. Salinas-Martínez de Lecea, J.A. Moulijn.

Dual-bed catalytic system for NOx–N2O removal: a practical

application for lean-burn deNOx HC-SCR. Applied Catalysis B:


[22] A.E.W. Beers, T.A. Nijhuis, F. Kapteijn, J.A. Moulijn. Zeolite coated

structures for the acylation of aromatics. Microporous and

Mesoporous Materials. 48 (2001) 279.

[23] A.E.W. Beers, R.A. Spruijt, T.A. Nijhuis, F. Kapteijn, J.A. Moulijn.

Esterification in a structured catalytic reactor with counter-current

water removal. Catalysis Today 66 (2001) 175.


BEA coating of structured supports-performance in acylation.



193

[25] V. Rico-Pérez, M.A. Velasco-Beltrán, H. Qinggang, W. Qi,

C. Salinas-Martínez de Lecea, A Bueno-López. Preparation of



(2013) 47.

[26] T.A. Nijhuis, A. E. W. Beers, T. Vergunst, I. Hoek, F. Kapteijn,



[27] A.F. Pérez-Cadenas, F. Kapteijn, J.A. Moulijn. Tuning the

morphology of monolith coatings. Applied Catalysis A: General 319

(2007) 267.

[28] J. Tsou, J. Cameiro, A.F. Perez-Cadenas, S. Eijsbouts,

M. Makkee, J.A. Moulijn. Preparation of monolithic catalysts for

hydrodesulfurization. Studies in Surface Science and Catysis 162

(2006) 143.

[29] A. F. Pérez-Cadenas, M. M. P. Zieverink, F. Kapteijn, J. A. Moulijn.

High performance monolithic catalysts for hydrogenatin reactions.


[30] J.M. García-Cortés, J. Pérez-Ramírez, J.N. Rouzaud,

A.R. Vaccaro, M.J. Illán-Gómez, C. Salinas-Martínez de Lecea. On

the structure sensitivity of deNOx HC-SCR over Pt-Beta catalysts.


[31] R. Burch, J.P. Breen, F.C. Meunier. A review of the selective

reduction of NOx with hydrocarbons under lean-burn conditions

with non-zeolitic oxide and platinum group metal catalysts. Applied

Catalysis B: Environmental 39 (2002) 283.

[32] J.M. García-Cortés, J. Pérez-Ramírez, M.J. Illán-Gómez,

F. Kapteijn, J.A. Moulijn, C. Salinas-Martínez de Lecea.

Comparative study of Pt-based catalysts on different supports in

the low-temperature de-NOx-SCR with propene. Applied Catalysis


Chapter 6

194

[33] M.D. Amiridisa, K.L. Roberts, C.J. Pereira. The selective catalytic

reduction of NO by propylene over Pt supported on dealuminated Y

zeolite. Applied Catalysis B: Environmental 14 (1997) 203.

[34] A. Kotsifa, D.I. Kondarides, X.E. Verykios. A comparative study of

the selective catalytic reduction of NO by propylene over supported

Pt and Rh catalysts. Applied Catalysis B 80 (2008) 260.

[35] J.M. García-Cortés, M.J. Illán-Gómez, C. Salinas-Martínez de

Lecea. The selective reduction of NOx with propene on Pt-Beta

catalyst: a transient study. Applied Catalysis B: Environmental 74

(2007) 313.




CHAPTER 7

General Conclusions

General conclusions

197

In this PhD thesis the optimization of the preparation conditions for

RhOx/CeyPr1-yO2 catalysts, in powder and supported in several cordierite

honeycomb monolith sizes, has been developed for further scaling up the

catalytic N2O decomposition process. The optimized conditions include

calcination process, use of different solvents (water, ethanol and acetone)

for rhodium loading, and amounts of rhodium, cerium and praseodymium

in the catalyst. A major understanding of the physico-chemical properties

and catalytic role of these materials in N2O decomposition process has

been facilitated by several characterization techniques and catalytic tests

carried out at different scales.

The main conclusions that are drawn from this research work are

as follows:

The calcination methods (ramp or flash) of cerium nitrate to obtain

the ceria support, has neither effect on the ceria properties (those

observed by XRD, Raman spectroscopy and N2 adsorption) nor on

the RhOx/CeO2 catalyst performance for N2O decomposition. On

the contrary, flash calcination of rhodium nitrate impregnated on

ceria support improves the catalytic activity for N2O decomposition

and CO oxidation of RhOx/CeO2 catalysts in comparison to that of

similar catalysts calcined in ramp.

The improved catalytic activity for N2O decomposition and CO

oxidation of catalysts with rhodium nitrate decomposed by flash

calcinations is attributed to the smaller size of RhOx nanoparticles

(smaller than 0.5 nm), allowing a larger noble metal oxide-support

interface favoring the reducibility of the ceria surface and stabilizing

the RhOx species under reaction conditions.

Fresh RhOx/CeO2 catalysts suffer a transformation process at

225 ºC during the N2O decomposition reaction, during which both

rhodium and cerium species are reduced and re-oxidized. This

process is consistent with the mechanism proposed for N2O

decomposition on RhOx/CeO2.

Both the nature of pure or praseodymium-doped ceria support and

the solvent used for rhodium precursor impregnation (with water,

Chapter 7

198

ethanol or acetone) affect RhOx-support interaction that determines

the activity for N2O decomposition of the catalysts studied.

The use of ethanol or acetone as solvent has a very negative effect

on Ce0.9Pr0.1O2 and Ce0.5Pr0.5O2-containing catalysts, regarding the

sintering of both supports and RhOx particles, affecting negatively

the RhOx-support interaction which directly hinders the catalytic

activity for N2O decomposition. This negative effect is due to the

solvent combustion catalyzed by the rhodium salt.

Ceria doping with 10 % praseodymium has a positive effect on the

RhOx-support interaction observed as a negative charge density

transfer from the noble metal to the ceria support (only obtained

using water for impregnation) that improves the catalytic activity.

The interaction between RhOx-Ce0.5Pr0.5O2 is not as good as that

with Ce0.9Pr0.1O2, but it seems to be enough to keep a high

catalytic activity if rhodium is impregnated with a water solution.

However, acetone or ethanol impregnation also leads to an

important decrease in activity, which is attributed to RhOx sintering.

RhOx/Ce0.9Pr0.1O2 active phases have been successfully loaded

into cordierite honeycomb monoliths. The nature of Ce0.9Pr0.1O2 in

powder and supported catalysts seems to be similar.

In honeycomb monoliths-supported catalysts, RhOx is preferentially

attached to Ce0.9Pr0.1O2, and not to the cordierite substrate.

For the experimental conditions studied in Chapter 5, the optimum

content of rhodium, in total weight-base of the monolith is 0.2 wt.%.

Increasing the amount above this loading does not increase the

N2O decomposition capacity.

The distribution of the supported active phases on the cordierite

substrate depends on the calcinations conditions. More

homogeneous coatings are obtained by flash calcination in

comparison to conventional ramp calcinations, affecting the

Rh-Ce-Pr interactions.

General conclusions

199

Supported catalysts have superior N2O decomposition capacity

than similar powder active phases due to its better mass-transfer in

comparison to randomly packed beds, avoiding gas diffusion

limitations in N2O decomposition rate.

The catalyst denoted by 0.2%RhOx/Ce0.9Pr0.1O2/M (flash) has

proved to be able to decompose N2O in presence of NO, O2 and

H2O accomplishing one of the main goals of this work which was scaling up RhOx/Ce0.9Pr0.1O2 catalysts for N2O decomposition in

real gas streams.

Diesel-SCR catalysts for NOx removal where also optimized and

prepared in order to build an efficient dual-bed system for

simultaneous deNOx and deN2O abatement.

The preparation of the Pt/Beta zeolite/monolith has been optimized

by adjusting the viscosity of surfactant and binder-containing Beta

zeolite suspensions (viscosities ≤ 23 mPa·s). The effect of zeolite

concentration on slurry viscosity was much more relevant than

those of surfactant and binder.

A certain stabilization time was necessary before the monolith can

be dipped on Beta zeolite slurries, and this time ranged from few

minutes (3 at least) until 1 hour, depending on the zeolite and

surfactant concentrations.

Several consecutive dipping steps are required in order to obtain

the Beta zeolite loading suitable for a practical application

(3.3 wt.% in the current study). The blowing step performed after

drying is not really necessary if the monolith channels are not

blocked by zeolite, and blowing the loosely bound zeolite could be

done after calcination.

SCR experiments performed in real diesel exhausts with medium

and full-size Pt/Beta zeolite/monolith catalysts demonstrated that

NOx reduction is feasible with commercial diesel fuel.

Chapter 7

200

The SCR behaviors observed in the real exhausts with commercial

diesel fuel were similar to that typically reported in laboratory

experiments performed with model HC.

The N2 selectivity of the medium-size Pt/Beta zeolite/monolith

catalyst in the studied conditions is significantly higher than

previous laboratory results, but the 100 % N2 selectivity is only

obtained by using RhOx/Ce0.9Pr0.1O2/monolith catalyst located

downstream the Pt/Beta zeolite/monolith catalyst, with both

monolith catalysts (deNOx and deN2O) operating at the same

temperature.

RESUMEN

Optimización de catalizadores

RhOx/ceria para la descomposición de

N2O y diseño de un sistema deNOx

altamente selectivo a N2 para

vehículos diésel

Catalizadores RhOx/ceria para la descomposición de N2O

203

1. Introducción General.

1.1. Efectos medioambientales del N2O.

El óxido nitroso se produce de forma natural en procesos

biológicos que se dan en suelos y aguas y también son emitidos por varias

fuentes antropogénicas relacionadas con la agricultura, energía, industria

y gestión de residuos [1, 2]

Desde el punto de vista medioambiental, en 1996 Kaptein y col. [3]

señalaron que con anterioridad el N2O había sido considerado un gas

inofensivo y no había presentado gran interés para los científicos,

ingenieros y políticos durante largo tiempo debido a la infravaloración y

desconocimiento de su potencial en la contribución a los problemas

medioambientales. Desde mitad de los años 80 se notó una creciente

preocupación al descubrir que el N2O contribuye a la destrucción de la

capa de ozono. Actualmente se sabe que el N2O es un contaminante muy

peligroso por lo comentado anteriormente y por ser un gas de efecto

invernadero.

El impacto medioambiental del N2O se ve incrementado como

consecuencia del aumento de las concentraciones atmosféricas a

consecuencia de las actividades humanas mencionadas anteriormente

[1, 4]. Aunque no es el gas que más contribuye al calentamiento global,

contribuyendo más el CO2 y el CH4. Su potencial de calentamiento es 310

veces superior al del CO2. Esto hace que emisiones relativamente

pequeñas (comparadas con otros gases de efecto invernadero) sean

equivalentes a un 10 % de las emisiones de CO2 [1, 2].

Por estos motivos, en la 3ª conferencia de la Convención Marco

de las Naciones Unidas sobre el Cambio Climático (UNFCC) en Kyoto en

1997, se legislaron límites de emisiones de seis gases de efecto

invernadero (CO2, CH4, N2O, HFC, PFC and SF6) para llevarse a cabo en

el periodo comprendido entre 2008-2012 [5]. En Doha, Qatar, a finales del

2012 se adoptaron nuevas enmiendas a incluir en el Protocolo de Kyoto,

incluyendo nuevos compromisos con una mayor ambición para reducir las

emisiones de gases de efecto invernadero, en un 18 % respecto al nivel

de emisiones de 1990 en el periodo comprendido entre 2013 y 2020 [5].

Resumen de la tesis doctoral

204

1.2. Fuentes de emisión de N2O.

Como se ha mencionado, las principales fuentes antropogénicas

de N2O son las derivadas de actividades agrarias y de las industrias de

producción de ácido adípico, ácido nítrico y nylon, así como la combustión

de combustibles fósiles tanto en fuentes estacionarias como en fuentes

móviles [1, 2].

Las emisiones de N2O que se pueden reducir a corto plazo están

asociadas a la producción de compuestos químicos e industria energética.

En concreto para la producción tanto de ácido adípico como ácido nítrico

ya existen tecnologías comercialmente disponibles para la reducción de

emisiones de N2O. Sin embargo su aplicación (extrapolación o

adaptación) a otras fuentes de emisión, como por ejemplo, en vehículos

ligeros no es factible debido a las diferentes características en la

composición de los gases de escape. En los vehículos ligeros, el N2O sale

a concentraciones y temperaturas relativamente bajas (T < 500 ºC)

acompañado de gases (O2, H2O, NOx, and SO2) que inhiben la actividad

de los catalizadores [2, 3, 6].

Como se ha mencionado anteriormente, una de las principales

fuentes de emisión de N2O son los vehículos y esta tesis se centra en la

eliminación de N2O procedente de motores diésel.

1.3. Motores diésel frente a motores gasolina.

Actualmente, los automóviles son el modo de transporte más

popular y necesario en nuestra vida cotidiana y en todo el mundo el

parque automovilístico asciende a más de 700 millones de vehículos [7].

En las últimas décadas, el incremento en el número de automóviles se

considera una de las principales fuente de contaminantes que disminuye

la calidad del aire, produciendo problemas de salud y dando lugar al

“smog fotoquímico” en las ciudades grandes [8].

Los motores diésel y gasolina funcionan de modo diferente. En los

motores diésel, el combustible en forma de spray se auto-inflama por la

alta presión alcanzada en la cámara de combustión, no siendo necesaria

la chispa inicie la combustión. El diésel es más pesado que la gasolina

debido a la mezcla de HC menos volátiles, siendo más susceptible a la


205

auto-ignición. En comparación con los motores gasolina, los motores

diésel trabajan con mayores ratios de compresión, combustionan más

rápido, operan en condiciones pobres de combustible (mayor relación

aire-combustible) y trabajan en exceso de O2. En los motores de gasolina

(también conocidos como motores Otto), el aire y el combustible se

mezclan antes de introducirlos en el cilindro, manteniendo constante el

ratio aire-combustible, normalmente estequiométrico, que es 14.6 en peso

[9]. Como resultado, los motores diésel tienen mayor eficiencia

termodinámica que los motores gasolina. Con combustible diésel se

consigue alrededor de un 12 % de energía en base volumétrica más que

la gasolina, favoreciendo que los motores diésel operen con una mayor

economía en combustible (menor consumo) que los motores gasolina

[10, 11].

El uso de catalizadores para purificar los gases de escape es

absolutamente necesario e indispensable en todos los vehículos [9, 12].

Los principales contaminantes procedentes de motores gasolina son HC,

CO, NOx (mezcla de NO y NO2) y material particulado (PM), este en muy

bajas concentraciones. La legislación relativa a estos contaminantes se ha

ido restringiendo a lo largo de los años, favoreciendo así el desarrollo e

introducción de los catalizadores de tres vías (TWC). Actualmente, en

occidente, prácticamente todos los coches nuevos con motor de gasolina

están equipados con estos catalizadores para reducir las emisiones de

CO, HC y NOx.

En los automóviles diésel, las emisiones de CO e HC son del

mismo orden de magnitud que las emitidas por vehículos a gasolina

equipados con TWC. Sin embargo las emisiones de NOx y PM en los

motores diésel son muy superiores. Los catalizadores TWC no se pueden

usar en motores diésel principalmente por dos razones:

La alta concentración de O2, pues dificulta la reducción de los

NOx.

La necesidad de un sistema específico para la eliminación de PM

basado en un filtro que retiene el PM y que periódicamente tiene

que ser regenerado.


206

Al igual que ha pasado con las emisiones procedentes de los

motores de gasolina, la legislación para las emisiones en motores diésel

también se ha vuelto más restrictiva con los años y se han desarrollado

varias técnicas para reducir dichas emisiones. De entre todas, las técnicas

de post-tratamiento se han estudiado extensamente para la eliminación de

NOx y PM procedentes de motores diésel [9]. En este trabajo se aborda el

estudio de la eliminación de NOx procedentes de los motores diésel.

1.4. Control de contaminantes procedentes de vehículos diésel.

Los distintos contaminantes emitidos por motores diésel requieren

estrategias específicas de post-combustión, contrariamente a lo que pasa

en los TWC para vehículos de gasolina. El CO y los HC son normalmente

oxidados en un catalizador de oxidación (DOC), el PM es recogido en un

filtro de partículas (DPF) y la reducción de NOx se realiza añadiendo un

reductor en presencia de un catalizador adecuado. Se han propuesto tres

procesos diferentes para la eliminación de NOx:

Descomposición catalítica de NOx.

Almacenamiento y reducción de NOx (NSR).

Reducción catalítica selectiva (SCR).

De estos tres procesos propuestos este trabajo está centrado en

la estrategia SCR para convertir NOx en N2 en una atmosfera rica en O2

[13-16]. En el proceso SCR, la reducción de NOx compite con éxito con la

reducción de O2, aunque el último esté en gran exceso. Estudios basados

principalmente en experimentos en el laboratorio proponen varios

reductores incluyendo H2, CO, distintos HCs, NH3 y CO(NH2)2 (urea), o

incluso el mismo combustible diésel [14-17].

Basándose en la tecnología bien implantada en plantas de

energía, donde el NH3 o la urea se inyectan a la corriente de

post-combustión para la reducción de NOx a N2 [14], con un catalizador

apropiado; el NH3 se está utilizando como reductor selectivo de NOx en

fuentes móviles. Los catalizadores basados en pentóxido de vanadio se

utilizan en vehículos diésel pesados (autobuses, camiones, etc). Por otro

lado, catalizadores basados en Fe/zeolita se han comercializado y usado






207

en varias aplicaciones SCR con NH3 como reductor, alcanzando

conversiones de NOx altas [13].

Sin embargo, el transporte de NH3 (que es un compuesto

corrosivo) dificulta la implementación de esta tecnología, convirtiendo la

urea en una alternativa, que por hidrolisis a alta temperatura da lugar al

NH3. Sin embargo, esta tecnología no ha sido implementada para

vehículos ligeros porque el tanque de urea/NH3 tiene el inconveniente de

añadir peso al automóvil, además de que sería necesario instaurar una

red de suministradores de estos reductores para llenar los tanques.

Los sistemas SCR de NOx con hidrocarburos (HC) han sido

propuestos como una alternativa a la tecnología con urea/NH3. Se han

probado HC de distinta naturaleza (propeno [18-21], octano [22], metanol

[21], etanol [21], acetaldehído [23] y decano [20]). Sin embargo, hasta

donde sabemos, la reducción directa de NOx con diésel comercial no ha

sido publicada, y uno de los objetivos de este trabajo es hacerlo. La

principal ventaja del combustible diésel con respecto a otros reductores es

que ya se encuentra en el vehículo y, por tanto, no son necesarios

tanques ni suministros adicionales.

La reactividad de los HCs en la conversión de NOx depende de su

naturaleza, del catalizador utilizado y de la temperatura, y por tanto

distintos HCs se comportan de forma diferente [24]. A altas temperaturas

la reacción de HC con O2 toma cada vez más relevancia y la mayoría de

los reductores (HC) se oxidan dificultando así la reducción de NOx. Una

característica de la reducción de NOx en exceso de O2 es la insuficiente

capacidad de reducción completa de los NOx a N2, dando lugar a una

considerable cantidad de N2O emitido [13]. La generación de N2O

depende de la naturaleza del catalizador, la naturaleza y concentración del

agente reductor, la temperatura y el flujo de salida de los gases de

escape, entre otros factores [13].

De entre todos los catalizadores propuestos para HC-SCR parece

que el platino es muy activo a bajas temperaturas (T < 300 ºC), y no se ve

significativamente afectado por la presencia de H2O en la corriente de

escape [14, 25-27]. Además, este es resistente al SO2 en los niveles de

azufre presentes en el escape del vehículo (25-50 ppm) [28].


208

1.5. Eliminación de N2O.

La descomposición catalítica de N2O ha sido estudiada con

catalizadores de distinta naturaleza. Las principales propiedades que debe

tener un catalizador para esta aplicación en escapes de vehículos es que

sea altamente activo a bajas temperaturas y resistente a los inhibidores

presentes en los gases de escape tales como O2, H2O y NOx residual.

Entre los ejemplos de catalizadores encontrados en la bibliografía se

pueden señalar los óxidos simples [2], perovskitas [29], spinelas [30-33],

zeolitas [34], hidrotalcitas [35], silicas mesoporosas [36] y catalizadores

soportados [37-39]. Sin embargo, muchos de los catalizadores propuestos

en la literatura no muestran una buena actividad ni estabilidad para la

conversión de N2O en condiciones reales.

Basándose en la experiencia de estudios previos en el grupo, de

entre todos los catalizadores propuestos, este trabajo se centra en

catalizadores de rodio soportado sobre ceria pura (CeO2) o dopada con

praseodimio (CeyPr1-yO2). Este tipo de catalizadores presenta gran

actividad en los experimentos realizados tanto en el laboratorio como en

una planta piloto de producción de ácido nítrico [40, 41] presentando gran

resistencia a H2O, NO y O2 [42]. Por ello, la hipótesis de partida es que los

catalizadores RhOx/ceria pueden ser adecuados para la descomposición

de N2O en vehículos diésel, situándolo detrás del catalizador SCR para la

reducción de NOx.

1.6. Objetivos.

Teniendo en cuenta lo descrito anteriormente, el objetivo general

de esta tesis es desarrollar y optimizar un catalizador efectivo para la

eliminación de N2O en las corrientes de escape de los vehículos diésel.

Éste irá situado después del catalizador SCR cuya composición fue

optimizada en trabajos anteriores y corresponde a Pt/zeolita Beta

soportado en un monolito de cordierita. Para alcanzar este objetivo

general, se deben lograr unos objetivos específicos, que se describen a

continuación:

Estudiar el efecto del tratamiento de calcinación de los

catalizadores rodio/ceria en las propiedades físico-químicas de los

mismos y en la actividad para la descomposición de N2O.


209

Estudiar el uso de distintos disolventes (agua, etanol y acetona) al

impregnar los soportes basados en ceria con los precursores de

rodio y ver cómo afectan a las propiedades físico-químicas de los

catalizadores y su actividad para la descomposición de N2O.

Preparar, caracterizar y probar en el laboratorio catalizadores

RhOx/Ce0.9Pr0.1O2 soportados en monolitos tipo panal de abeja de

cordierita.

Preparar un catalizador Pt/zeolita Beta soportado en un monolito

tipo panal de abeja de cordierita de tamaño real y probarlo en una

corriente real de escape de un motor diésel para la reducción de

NOx con diésel comercial.

Probar un doble lecho catalítico basado en los catalizadores

Pt/zeolita Beta y RhOx/Ce0.9Pr0.1O2 soportados en monolitos de

cordierita que operen a la misma temperatura para la reducción de

NOx a N2 con diésel comercial, que se inyecta antes del primer

catalizador.

2. Técnicas de caracterización.

Los catalizadores se definen por su composición, estructura, y por

sus propiedades estructurales y texturales que están relacionadas con su

actividad catalítica. Para entender y determinar las propiedades

físico-químicas de los catalizadores en este trabajo se han utilizado

diversas técnicas de caracterización.

La termogavimetría acoplada a un espectrómetro de masas

(TG-MS) se utilizó para seleccionar la temperatura adecuada a la que se

empieza a descomponer los precursores de los catalizadores. La

adsorción de N2 a -196 ºC se utilizó para determinar el área superficial

BET. Las propiedades estructurales fueron estudiadas mediante difracción

de rayos X (DRX) y espectroscopia Raman. La composición atómica

superficial y los estados de oxidación de los metales fueron determinados

mediante espectroscopia fotoelectrónica de rayos X (XPS). La reducción a

temperatura programada con H2 se empleó para estudiar la reducibilidad

de las muestras. El tamaño de partícula se determinó mediante DRX y


210

microscopia electrónica de transmisión (TEM). Finalmente, la distribución

de la fase activa fue observada mediante microscopía electrónica de

barrido (SEM).

3. Efecto de las condiciones de calcinación de los catalizadores

RhOx/CeO2 en la descomposición de N2O.

En aplicaciones catalíticas, la ceria se usa como soporte de otras

especies o como catalizador en si misma [43-45]. Este tipo de catalizador

se usa principalmente para el control de contaminantes procedente de los

vehículos de gasolina (TWC) [46], combustión de PM [47], reducción

selectiva catalítica de NOx [48], oxidación parcial de metano [49],

combustión de compuestos orgánicos volátiles [50] y descomposición de

N2O [38] entre otros. En esta última aplicación, el tamaño y la distribución

de partículas de metal noble, así como la morfología superficial y los

defectos de los óxidos afectan el funcionamiento del catalizador Rh/ceria,

siendo crucial el comportamiento de la interfase metal-soporte con

relación a la actividad catalítica [40, 51-53]. En este sentido en este

trabajo se investiga la optimización de estas interacciones para mejorar el

funcionamiento del catalizador.

Esta parte del trabajo corresponde al estudio del efecto de las

condiciones de calcinación en catalizadores cuya composición es

2.5%RhOx/CeO2. Se estudian dos formas de calcinación. Cuando la

calcinación empieza a 25 ºC se denomina calcinación en rampa mientras

que si empieza a 250 o 350 ºC se denomina calcinación flash. Este

estudio se realiza a escala de laboratorio en un sistema compuesto por

una unidad de alimentación y mezcla de gases, un reactor, un horno y un

controlador de temperatura y por último la unidad de análisis compuesta

por un cromatógrafo de gases para analizar N2O, O2 y N2.

Los ensayos catalíticos de descomposición de N2O se realizaron

en un reactor de cuarzo en forma de U que se introduce en un horno

vertical a presión atmosférica usando un flujo de 100mL/min

(GHSV = 42000 h−1

) de 1000 ppm N2O en He, utilizando 100 mg de

catalizador.


211

Los ensayos de oxidación de CO se realizaron en el mismo

sistema que los ensayos catalíticos anteriores usando 100 mg de

catalizador y un flujo de 100 mL/min (GHSV=42000 h−1

) de

1000 ppm CO/5 % O2/He y 100 mg de catalizador.

Este estudio deja patente que el método de calcinación (rampa o

flash) del nitrato de cerio utilizado para obtener el soporte de ceria no

afecta ni las propiedades de la ceria (observado mediante isotermas de

adsorción de N2, DRX y espectroscopia Raman) ni la actividad del

catalizador RhOx/CeO2. Por el contrario, la calcinación flash del nitrato de

rodio usado para impregnar la ceria mejora la actividad catalítica tanto

para la descomposición de N2O como para la oxidación de CO (reacción

de gran relevancia que se usó para confirmar si las mejoras del

catalizador sólo eran válidas para la descomposición de N2O o se podía

extrapolar a otras reacciones).

Estas mejoras en la actividad catalítica se atribuyen a un menor

tamaño de partícula RhOx que permite mejorar la interfase metal-soporte

cuando se hace una calcinación flash en lugar de calcinación en rampa,

favoreciendo a su vez la reducibilidad de la superficie de la ceria y

estabilizando las especies RhOx en condiciones de reacción.

4. Preparación de catalizadores RhOx/CeyPr1-yO2 para la

descomposición de N2O con distintos disolventes.

Además de las razones mencionadas anteriormente, los

materiales basados en ceria son de interés por su capacidad de

almacenamiento de oxígeno (OSC) y la movilidad de éste en la estructura.

Estas propiedades dependen del tamaño de cristal, de los defectos y de la

presencia de dopantes (en este estudio el praseodimio). La incorporación

del praseodimio en la estructura de la ceria modifica la movilidad del

oxígeno en la celdilla unidad y su comportamiento catalítico, ya que

disminuye la (BE) energía del enlace M-O en la ceria dopada, aumentando

el número de defectos en la red y por tanto las vacantes de oxígeno.

Como consecuencia se mejora la eficiencia en la descomposición de N2O

[40, 54].


212

En el apartado anterior se concluyó que la velocidad de

evaporación del disolvente que impregna la ceria es un factor

determinante para las propiedades del material, afectando al tamaño de

las partículas de RhOx en el catalizador final [55, 56]. Por ello en esta

segunda parte se presenta el estudio realizado sobre el efecto del

disolvente (agua, etanol o acetona) usado en la impregnación de los

soportes CeyPr1-yO2 (donde y es 1, 0,9 o 0,5) con nitrato de rodio para la

preparación de los catalizadores. Este estudio se realiza a escala de

laboratorio en el mismo sistema y condiciones descritos en el apartado

anterior para los ensayos de descomposición de N2O.

Las conclusiones obtenidas fueron que tanto la naturaleza del

soporte de ceria como del disolvente usado afectan la interacción

RhOx-soporte, la cual determina la actividad catalítica de los catalizadores

estudiados.

Tanto el etanol como la acetona afectan muy negativamente a los

catalizadores que contienen Ce0.9Pr0.1O2 y Ce0.5Pr0.5O2 en lo referente a la

sinterización del soporte y de las partículas de RhOx. Este hecho está

relacionado con la propia combustión de los disolventes catalizada por el

nitrato de rodio, afectando negativamente la interacción RhOx-soporte lo

que conduce a catalizadores menos efectivos para la descomposición de

N2O.

Mediante XPS se observó que el dopado de la ceria con un 10 %

de praseodimio tiene un efecto positivo en la interacción RhOx-soporte

observada como una transferencia de carga negativa desde el metal noble

al soporte (sólo obtenida con el uso de agua para la impregnación) y esto

mejora la eficiencia del catalizador.

En cuanto a la interacción RhOx-Ce0.5Pr0.5O2 no es tan buena

como con Ce0.9Pr0.1O2 pero parece ser suficiente para mantener una alta

actividad catalítica, si la impregnación con rodio se hace utilizando agua.

Pues cuando se utiliza acetona o etanol se produce un considerable

descenso en la actividad y esto solo puede atribuirse a la sinterización de

las partículas de RhOx puesto que Ce0.5Pr0.5O2 con considerable baja área

superficial BET no sinteriza con respecto a la impregnación con agua.


213

5. Preparación, caracterización y actividad catalítica para la

descomposición de N2O de catalizadores monolíticos

RhOx/Ce0.9Pr0.1O2.

De acuerdo con los resultados anteriores, el mejor catalizador en

polvo preparado hasta el momento es RhOx/Ce0.9Pr0.1O2,donde el rodio ha

sido soportado sobre Ce0.9Pr0.1O2 con un disolución acuosa del nitrato

precursor y una posterior calcinación en condiciones flash.

Uno de los retos en el control de las emisiones de N2O es

desarrollar un catalizador capaz de trabajar en corrientes de gases reales.

Para emplear un catalizador en estas condiciones es necesario incorporar

la fase activa (RhOx/Ce0.9Pr0.1O2) previamente optimizada en polvo, a un

soporte inerte apropiado. Se pueden diversos sustratos con este

propósito, por ejemplo, pellets o partículas de óxidos inorgánicos,

monolitos cerámicos tipo panal de abeja, espumas cerámicas, etc [57-61].

De entre los soportes mencionados, los monolitos cerámicos tipo panal de

abeja presentan algunas propiedades atractivas, proporcionando un buen

contacto entre las fases activas del catalizador y los gases a tratar.

También tienen alta tolerancia al polvo y la caída de presión se minimiza

porque el flujo de gas no se ve obstaculizado de manera significativa a

través del lecho catalítico. Además no padecen problemas de degradación

como los que ocurren típicamente por ejemplo por la fricción de partículas

[57-60]

La cordierita (2MgO:5SiO2:2Al2O3) es el material más

comúnmente utilizado en la fabricación de monolitos tipo panal de abeja.

La cordierita presenta alta resistencia mecánica y bajo coeficiente de

expansión térmica [57-60]. La incorporación de una fase catalíticamente

activa en el monolito no es trivial, requiriéndose estudios de optimización

del proceso de carga de las fases activas.

El objetivo de esta tercera etapa del estudio fue preparar un

catalizador en forma de monolito (panel de abejas) y utilizarlo para la

descomposición de N2O en corrientes de gases reales. La fase activa de

composición RhOx/Ce0.9Pr0.1O2 se incorpora al interior del monolito

mediante un método sencillo consistente en la impregnación de los

monolitos con una disolución acuosa que contiene los precursores

metálicos y una posterior calcinación. En este caso, nuevamente, se ha


214

estudiado el efecto de las condiciones de calcinación. Además se ha

optimizado la cantidad de rodio que debe contener el catalizador. Los

monolitos recubiertos en su interior por la fase activa RhOx/Ce0.9Pr0.1O2

han sido caracterizados y probados para la descomposición de N2O en

una corriente gaseosa de N2O/He. El mejor catalizador soportado ha sido

evaluado en presencia de los principales inhibidores (O2 + NOx + H2O)

encontrados en el escape de un vehículos diésel.

Los ensayos catalíticos se realizaron a escala de laboratorio en el

sistema experimental antes mencionado incluyendo un analizador de

quimiluminiscencia para monitorizar NO y NO2. En esta ocasión se utilizó

un reactor cilíndrico de cuarzo y un horno horizontal. El flujo utilizado fue

de 500 mL/min (GHSV = 27000 h−1

) utilizando distintas atmosferas

1000 ppm N2O/He o 1000 ppm N2O/1000 ppm NO/5 % O2/0.6 % H2O/He y

un catalizador soportado. Para el catalizador en polvo se usó una cantidad

similar a la fase activa contenida en el soporte monolítico y se diluyo con

SiC has obtener un volumen de lecho similar al ocupado por el monolito.

Las resultados obtenidos en este apartado llevaron a la conclusión

de que la composición y estructura del Ce0.9Pr0.1O2 en polvo y soportado

en la cordierita son similares a bajas temperaturas pero a partir de una

cierta temperatura el catalizador soportado supera la conversión

conseguida por el catalizador en polvo. Esto se atribuye a la buena

transferencia de masa de los catalizadores soportados.

Para las condiciones experimentales aplicadas en este estudio el

contenido óptimo en rodio es de 0.2 wt.% en base a peso total.

Aumentando este valor no se consigue mejorar la capacidad de

descomposición de los catalizadores.

En los catalizadores soportados en el monolito se observa que las

partículas de RhOx se adhieren preferentemente al Ce0.9Pr0.1O2 y no al

sustrato de cordierita. Además, la distribución de la fase activa soportada

en el sustrato monolítico depende de las condiciones de calcinación,

obteniéndose un recubrimiento más homogéneo por calcinación flash en

comparación con la calcinación en rampa.

También se observó que el proceso de calcinación afecta a las

interacciones Rh-Ce-Pr. La calcinación flash da lugar a menores tamaños


215

de partículas RhOx mejorando la reducibilidad de las entidades

superficiales Rh-Ce-Pr del catalizador a bajas temperaturas en

comparación con la calcinación en rampa. De esta forma se mejora la

actividad catalítica.

Finalmente, el catalizador denominado 0.2%RhOx/Ce0.9Pr0.1O2/M

(flash) ha demostrado ser capaz de descomponer N2O en presencia de

NOx, O2 y H2O. Por tanto, el objetivo principal de este apartado, consiste

en la preparación de un monolito con la fase activa RhOx/Ce0.9Pr0.1O2 para

su utilización en corrientes de gases reales, ha sido completado con éxito.

6. Reducción de NOx a N2 con combustible comercial en la corriente

de escape de un motor diésel utilizando un doble lecho de los

catalizadores monolíticos Pt/zeolita Beta y RhOx/Ce0.9Pr0.1O2.

La reducción de NOx (en exceso de O2) procedente del escape de

los motores diésel se puede conseguir principalmente mediante dos

tecnologías: Reducción catalítica selectiva (SCR) y el almacenamiento y

posterior reducción de NOx (NSR) [62-64] y en ambos casos es necesario

añadir un reductor.

Se han propuesto varios reductores incluyendo H2, CO, diferentes

HCs, NH3, urea, etc [62-70] en la tecnología SCR. La utilización comercial

de NH3 o urea como reductor para el sistema SCR ya está disponible para

camiones y autobuses, pero en el caso de coches, esta tecnología no se

ha implementado debido al inconveniente del sobrepeso que supone un

depósito adicional de NH3/urea. Como nueva alternativa proponemos el

combustible diésel como agente reductor y un catalizador Pt/zeolita Beta,

basándonos en la experiencia de trabajos previos [71, 72].

Esta última fase del estudio consiste en la preparación de

catalizadores monolíticos de tamaño medio y real con Pt/zeolita Beta

como fase activa para utilizarlos en la tecnología SCR para la eliminación

de NOx en el escape de un motor diésel real utilizando combustible

comercial como agente reductor, tanto en un lecho simple SCR (a escala

media y escala real) como en un doble lecho catalítico de tamaño medio el

cual consiste en el catalizador SCR Pt/zeolita Beta/monolito localizado

delante del catalizador de descomposición de N2O


216

RhOx/Ce0.9Pr0.1O2/monolito. Los ensayos catalíticos a escala media fueron

realizados en la Universidad de Alicante y a escala real en el Centro

Tecnológico de Automoción de Galicia (CTAG).

Los ensayos con catalizadores de tamaño medio se realizaron con

un motor Turbo Diesel 2.0 a 880 rpm (en ralentí). En estas condiciones de

funcionamiento, el flujo de escape era de 800 L/min, y la composición fue

17.2 % O2, 2.4 % CO2, 10 % H2O, 100 ppmV NO, 35 ppmV NO2,

120 ppmV CO y 10 ppmV THC (THC = hidrocarburos totales, expresados

como CH4). La temperatura del gas a lo largo del tubo de escape estaba

comprendida entre 90 y 50 ºC. Se extrajo una corriente de gas de

10 L/min continuamente de la corriente principal y se usó para realizar los

ensayos. Este flujo pasó a través de dos hornos consecutivos con un

control independiente de temperatura, el primero conteniendo un crisol

con diésel comercial y el segundo los catalizadores monolíticos.

Los ensayos catalíticos de reducción de NOx con catalizadores a

escala real se llevó a cabo en un banco de pruebas Horiba Titan S190 en

un motor diésel 1.6 HDI a 1100 rpm y cargas distintas (torque entre 45 y

83 N.m) usando diésel como reductor. La temperatura y la composición

del gas a la entrada del catalizador SCR dependen de la carga del motor.

En cuanto a la preparación de los catalizadores, se ha optimizado

el método de recubrimiento por inmersión (conocido como dip-coating)

para la preparación del catalizador Pt/zeolita Beta/monolito. El catalizador

(RhOx/Ce0.9Pr0.1O2/monolito) de descomposición de N2O, se ha preparado

por descomposición de nitratos en forma similar al método óptimo descrito

en [55]. La producción de N2O como producto no deseado generado en la

reducción de NOx, es el mayor inconveniente de los catalizadores de

platino para SCR, este problema se ha resuelto con la configuración de

doble lecho catalítico donde los dos catalizadores monolíticos operan a la

misma temperatura consiguiendo un 100 % de selectividad hacia el N2. El

objetivo de escalar el catalizador SCR a escala real o completa es

demostrar por primera vez que el sistema SCR puede ser implementado

en el escape de un vehículo diésel usando diésel comercial como

reductor.

De este estudio se concluye que la preparación del catalizador

Pt/zeolita Beta/monolito ha sido optimizada ajustando la viscosidad de la


217

suspensión de zeolita Beta con surfactante y aglomerante a valores

≤ 23 mPa.s. El efecto de la concentración de zeolita es mucho más

relevante que el del surfactante o aglomerante en la viscosidad de la

suspensión.

Antes de que el monolito sea sumergido en la suspensión de

zeolita es necesario un cierto tiempo de estabilización de la suspensión.

Este tiempo de estabilización tiene un intervalo de entre unos pocos

minutos (3 por lo menos) y una hora, dependiendo de las concentraciones

de zeolita y surfactante.

La etapa de soplado después del secado no es realmente

necesaria en el proceso de preparación, si los canales del monolito no

están bloqueados por la zeolita. Por tanto la eliminación por soplado de la

zeolita débilmente unida al monolito podría hacerse después de la

calcinación. Sin embargo, son necesarias varias etapas de inmersión del

monolito en la suspensión de zeolita para incrementar la carga necesaria

para una aplicación práctica.

Los experimentos SCR realizados en el escape de un vehículo

diésel a escala media y real con catalizadores Pt/ zeolita Beta/monolito

han demostrado que la reducción de NOx es posible con diésel comercial

como reductor. El comportamiento de estos catalizadores, en cuanto a la

reducción de NOx se refiere, es similar a los realizados previamente a

escala de laboratorio con HC modelo.

La selectividad hacia N2 del catalizador Pt/zeolita Beta/monolito en

las condiciones estudiadas es significativamente mayor que en resultados

previos de laboratorio, pero para alcanzar un 100 % de selectividad es

necesario utilizar el doble lecho catalítico situando primero el catalizador

Pt/zeolita Beta/monolito y luego el RhOx/Ce0.9Pr0.1O2/monolito, ambos

catalizadores operando a la misma temperatura.


218

Referencias.

[1] http://www.epa.gov/climatechange/ghgemissions/usinventoryreport

.html; Inventory of U.S. Greenhouse gas emissions and sinks:

1990–2011-February 2013. Access in April 2013.

[2] J. Pérez-Ramírez, F. Kapteijn, K. Schöffel, J.A. Moulijn. Formation

and control of N2O in nitric acid production. Where do we stand

today?. Applied catalysis B: Environmental 44 (2003) 117.

[3] F. Kapteijn, J. Rodríguez-Mirasol, J.A. Moulijn. Heterogeneous

catalytic decomposition of nitrous oxide. Applied catalysis B:


[4] H. Rodhe. A Comparison of the contribution of various gases to the

greenhouse effect. Science 248 (1990) 1217.

[5] http://unfccc.int/kyoto_protocol/items/2830.php. Access May 2013

[6] G. Centi, S. Perathoner, F. Vazzana, M. Marella, M. Tomaselli,

M. Mantegazza. Novel catalysts and catalytic technologies for NO2

removal from industrial emissions containing O2, H2O and SO2.

Advances in Environmental Research 4 (2000) 325.



[8] M.V. Twigg. Catalytic control of emissions from cars. Catalysis

Today 163 (2011) 33.

[9] J.P.A. Neeft, M. Makkee, J.A. Moulijn. Review article diesel

particulate emission control. Fuel Processing Technology 47

(1996) 1.

[10] T.J. Wallington, C.K. Lambert, W.C. Ruona. Diesel vehicles and

sustainable mobility in the U.S. Energy Policy 54 (2013) 47.

http://unfccc.int/kyoto_protocol/items/2830.php.%20Access%20May%202013


219

[11] H-K. Wang, C-Y Cheng, Y-C. Lin, K-S. Chen. Emission reductions

of air pollutants from a heavy-duty diesel engine mixed with various

amounts of H2/O2. Aerosol and Air Quality Research 12 (2012)

133.



[13] M.V. Twigg, Progress and future challenges in controlling





(2006) 43.



1.




[17] M.V. Twigg, P.R. Phillips. Cleaning the air we breathe – controlling

diesel particulate emissions from passenger cars. Platinum Metals

Review 53 (2009) 27.

[18] A. Bueno–López, M.J. Illán–Gómez y C. Salinas-Martínez de

Lecea. Effect of NOx and C3H6 partial pressures on the activity of

Pt–Beta–coated cordierite monoliths for deNOx C3H6–SCR.





(2000) 137.


220



deactivation of Ag/Al2O3 and Cu/ZSM-5 for HC-SCR under


230 (2005) 150.







oxygen-rich exhaust gases: combined heterogeneous and surface-

mediated homogeneous reactions. Journal of Catalysis 219 (2003)

25.




[24] R. Burch. Knowledge and know‐how in emission control for mobile

applications. Catalysis reviews – Science and Engineering 46

(2004) 271.

[25] A.R. Vaccaro, J. Pérez-Ramírez, J.M. Garcia-Cortés, C. Salinas-

Martínez de Lecea, G.Mul, F. Kapteijn, J.A. Moulijn. Dual-bed

catalytic system for removal of NOx-N2O in lean-burn engine

exhausts. Environmental Challenges and Greenhous Gas Control

for Fossil Fuel Utilization in the 21st Century (2002) 229.

[26] M. Iwamoto, H. Yahiro, H.K. Shin, M. Watanabe, J. Guo, M. Konno,

T. Chikahisa, T.Murayama. Performance and durability of Pt-MFI

zeolite catalyst for selective reduction of nitrogen monoxide in

actual diesel engine exhaust. Applied Catalysis B: Environmental 5

(1994) L1.


221



demonstration of the Pt-catalysed C3H6-SCR of NOx in a diesel


[28] M.D. Amiridis, K.L. Roberts, C.J. Pereira. The selective catalytic

reduction of NO by propylene over Pt supported on dealuminated Y

zeolite. Applied Catalysis B: Environmental 14 (1997) 203.

[29] J.P. Dacquin, C. Lancelot, C. Dujardin, P. Da Costa, G. Djega-

Mariadassou, P. Beaunier, S. Kaliaguine, S. Vaudreuil, S. Royer,

P. Grange. Influence of preparation methods of LaCoO3 on the

catalytic performances in the decomposition of N2O. Applied

Catalysis B: Environmental 91(2009) 596

[30] W. Piskorz, F. Zasada, P. Stelmachowski, A. Kotarba, Z. Sojka.

Decomposition of N2O over the surface of cobalt spinel: a DFT

account of reactivity experiments. Catalysis Today 137 (2008) 418.

[31] F. Zasada, P. Stelmachowski, G. Maniak, J.F. Paul, A. Kotarba,

Z. Sojka. Potassium promotion of cobalt spinel catalyst for N2O

decomposition - accounted by work function measurements and

DFT modelling. Catalysis Letters 127 (2009) 126.

[32] P. Stelmachowski, G. Maniak, A. Kotarba, Z. Sojka. Strong

electronic promotion of Co3O4 towards N2O decomposition by

surface alkali dopants. Catalysis Communications 10 (2009) 1062.

[33] P. Stelmachowski, F. Zasada, G. Maniak, P. Granger, M. Inger,

M. Wilk, A. Kotarba, Z. Sojka. Optimization of multicomponent

cobalt spinel catalyst for N2O abatement from nitric acid plant tail

gases: laboratory and pilot plant studies. Catalysis Letters 130

(2009) 637.

http://www.scopus.com/record/display.url?eid=2-s2.0-68749113955&origin=resultslist&sort=plf-f&src=s&st1=dacquin&nlo=&nlr=&nls=&sid=9BD716503253C359EB402FCF1830BCCF.euC1gMODexYlPkQec4u1Q%3a80&sot=b&sdt=cl&cluster=scopubyr%2c%222009%22%2ct&sl=20&s=AUTHOR-NAME%28dacquin%29&relpos=1&relpos=1&searchTerm=AUTHOR-NAME%28dacquin%29+AND+%28+LIMIT-TO%28PUBYEAR%2C2009%29+%29+

http://www.scopus.com/record/display.url?eid=2-s2.0-68749113955&origin=resultslist&sort=plf-f&src=s&st1=dacquin&nlo=&nlr=&nls=&sid=9BD716503253C359EB402FCF1830BCCF.euC1gMODexYlPkQec4u1Q%3a80&sot=b&sdt=cl&cluster=scopubyr%2c%222009%22%2ct&sl=20&s=AUTHOR-NAME%28dacquin%29&relpos=1&relpos=1&searchTerm=AUTHOR-NAME%28dacquin%29+AND+%28+LIMIT-TO%28PUBYEAR%2C2009%29+%29+










222

[34] S. Sklenak, P.C.Andrikopoulos, B. Boekfa, B. Jansang,

J. Novakova, L. Benco, T. Bucko, J. Hafner, J. Dedecek,

Z. Sobalik. N2O decomposition over Fe-zeolites: Structure of the

active sites and the origin of the distinct reactivity of Fe-ferrierite,

Fe-ZSM-5, and Fe-Beta. A combined periodic DFT and

multispectral study. Journal of Catalysis 272 (2010) 262.

[35] K. Karaskova, L. Obalova, K. Jiratova, F. Kovanda. Effect of

promoters in Co-Mn-Al mixed oxide catalyst on N2O

decomposition. Chemical Engineering Journal 160 (2010) 480.

[36] L. Chmielarz, P. Kustrowski, M. Drozdek, M. Rutkowska,

R. Dziembaj, M. Michalik, P. Cool, E.F. Vansant. SBA-15

mesoporous silica modified with rhodium by MDD method and its

catalytic role for N2O decomposition reaction. Journal Porous

Materials 18 (2011) 483.

[37] J.P. Dacquin, C. Dujardin, P. Granger. Surface reconstruction of

supported Pd on LaCoO3: Consequences on the catalytic

properties in the decomposition of N2O. Journal of Catalysis 253

(2008) 37.





(2010) 370.

[39] S. Parres-Esclapez, F.E. López-Suárez, A. Bueno-López,

M.J. Illán-Gómez, B. Ura, J. Trawczynski. Rh-Sr/Al2O3 catalyst for

N2O decomposition in the presence of O2.Topics in Catalysis 52

(2009) 1832.











http://www.scopus.com/record/display.url?eid=2-s2.0-71749105186&origin=resultslist&sort=plf-f&src=s&st1=parres+esclapez&sid=CD766D47B32FF9707B5C70693D7E7B52.aXczxbyuHHiXgaIW6Ho7g%3a120&sot=b&sdt=b&sl=28&s=AUTHOR-NAME%28parres+esclapez%29&relpos=9&relpos=9&searchTerm=AUTHOR-NAME%28parres+esclapez%29

http://www.scopus.com/record/display.url?eid=2-s2.0-71749105186&origin=resultslist&sort=plf-f&src=s&st1=parres+esclapez&sid=CD766D47B32FF9707B5C70693D7E7B52.aXczxbyuHHiXgaIW6Ho7g%3a120&sot=b&sdt=b&sl=28&s=AUTHOR-NAME%28parres+esclapez%29&relpos=9&relpos=9&searchTerm=AUTHOR-NAME%28parres+esclapez%29


223

[41] S. Parres-Esclapez. N2O decomposition rhodium/ceria catalysts:

From principles to practical application. Tesis Doctoral, University

of Alicante (2011).


S. Kushiyama, K.Mizuno. Decomposition of nitrous oxide over



[43] A. Trovarelli, C. De Leitenburg, M. Boaro, G. Dolcetti. The

utilization of ceria in industrial catalysis. Catalysis Today 50 (1999)

353.

[44] M.F. Luo, J.M. Ma, J.Q. Lu, Y.P. Song, Y.J. Wang. High-surface

area CuO–CeO2 catalysts prepared by a surfactant-templated

method for low-temperature CO oxidation. Journal of Catalysis 246

(2007) 52.

[45] A. Trovarelli. Catalytic Properties of Ceria and CeO2-Containing

Materials. Catalysis Reviews - Science and Engineering 38 (1996)

439.



[47] I. Atribak, I. Such-Basáñez, A. Bueno-López, A. García-García.

Comparison of the catalytic activity of MO2 (M=Ti, Zr, Ce) for soot

oxidation under NOx/O2. Journal of Catalysis 250 (2007) 75.

[48] M. Adamowska, S. Muller, P. Da Costa, A. Krzton, P. Burg.

Correlation between the surface properties and deNOx activity of

ceria-zirconia catalysts. Applied Catalysis B: Environmental 74

(2007) 278.

[49] S. Eriksson, S. Rojas, M. Boutonnet, J.L.G. Fierro. Effect of

Ce-doping on Rh/ZrO2 catalysts for partial oxidation of methane.







224

[50] J.I. Gutiérrez-Ortiz, B. de Rivas, R. López-Fonseca, J.R. González-

Velasco. Catalytic purification of waste gases containing VOC

mixtures with Ce/Zr solid solutions. Applied Catalysis B:


[51] S. Imamura, R. Hamada, Y. Saito, K. Hashimoto, H. Jindai.

Decomposition of N2O on Rh/CeO2/ZrO2 composite Catalyst.

Journal of Molecular Catalysis A: Chemical 139 (1999) 55.



temperature N2O dissociation. Studies in Surface Science

Catalysis 101 (1996) 681.

[53] M. Cargnello, P. Fornasiero, R.J. Gorte, Opportunities for tailoring

catalytic properties through metal-support interactions. Catalysis

Letters 142 (2012) 1043.

[54] A. Trovarelli. Catalysis by ceria and related materials. Catalytic





Rh/Ce0.9Pr0.1O2 catalysts. Applied catalysis B: Environmental 107

(2011) 18.





(2013) 47.

[57] T.A. Nijhuis, A.E.W. Beers, T. Vergunst, I. Hoek, F. Kapteijn, J.A.

Moulijn. Preparation of monolithic catalysts. Catalysis Reviews -

Science and Engineering 43 (2001) 345.


Carbon-Based Monolithic Structures. Catalysis Reviews – Science





225

[59] C. Moreno-Castilla, A.F. Pérez-Cadenas. Carbon-based


Materials 3 (2010) 1203.


BEA coating of structured supports - Performance in acylation.


[61] K. Galejová, L. Obalová, K. Jirátová, K. Pacultová, F Kovanda.



(2009) 172.

[62] W.S. Epling, L.E. Campbell, A. Yezerets, N.W. Currier, J.E. Parks

II. Overview of the fundamental reactions and degradation

mechanisms of NOx storage/reduction catalysts. Catalysis




1.

[64] R. Burch. Knowledge and know-how in emission control for mobile

applications. Catalysis Reviews – Science and Engineering 46

(2004) 271.

[65] Y. Traa, B. Burger, J. Weitkamp. Zeolite-based materials for the

selective catalytic reduction of NOx with hydrocarbons.

Microporous and Mesoporous Materials 30 (1999) 3.

[66] P.L.T. Gabrielsson. Urea-SCR in automotive applications. Topics in

Catalysis 28 (2004) 177.

[67] T. Johnson. Diesel engine emissions and their control: an

overview. Platinum Metals Review 52 (2008) 23.





226

[69] M.V. Twigg. Progress and future challenges in controlling





(2006) 43.


J.M. García-Cortés, M.J. Illán-Gómez, C. Salinas-Martínez de


by in situ synthesis. Utilization as Pt support for NOx abatement in


[72] A. Bueno-López, M.J. Illán-Gómez, C. Salinas-Martínez de Lecea.


Pt-Beta-coated cordierite monoliths for deNOx C3H6-SCR. Applied

Catalysis A: General 302 (2006) 244.

Abbreviations

Abbreviations

229

Abbreviations

BE Binding Energy

BET Brunauer-Emmett-Teller

ca. Latin word “circa” which means “around”

CAFE Corporate Average Fuel Economy

CCRT Catalyzed Continuously Regenerating Trap

CeO2 Ceria, cerium oxide

CeyPr1-yO2 Doped ceria, cerium praseodymium mixed oxide

CRT® Continuously Regenerating Trap

deNOx NOx abatement

deN2O N2O abatement

DPFs Diesel Particulate Filters

EDS Energy-Dispersive X-ray Spectroscopy

EPA U.S. Environmental Protection Agency

ESCA Electron Spectroscopy for Chemical Analysis

FCC Fluid Catalytic Cracking

FWHM Full-Width at Half Maximum

GHG Greenhouse Gas

GWP Global Warming Potential

H2-TPR Temperature Programmed Reduction with H2

HC Hidrocarbons

Abbreviations

230

i.e. Latin phrase "id est", meaning "that is"

IPCC Intergovernmental Panel on Climate Change

LNT Lean NOx Trap

MCMA Carbon Materials and Environment Research Group

from Spanish “Materiales Carbonosos y Medio

Ambiente”

mg/bhp-h Milligrams per brake horsepower-hour

NACs NOx absorbing catalysts

NOx (NO+NO2) Mononitrogen oxides (nitric oxide and nitrogen

dioxide).

NSR NOx Storage Reduction

ODSs Ozone Depleting Substances

OSC Oxygen Storage Capacity

PAHs Polycyclic Aromatic Hydrocarbon

PGMs Platinum Group Metals

PM Particulate Matter or soot

SEM Scanning Electron Microscopy

SCR Selective Catalytic Reduction

TCD Thermic Conductivity Detector

TEM Transmission Electron Microscopy

Tg CO2-eq. Teragrams (or million metric tons) of CO2 equivalent

TG-MS Thermogravimetry-Mass Spectroscopy

Abbreviations

231

TWC Three Way Catalysts

UNFCCC United Nations Framework Convention on Climate

Change

XRD X-Ray Diffraction

XPS X-ray Photoelectron Spectroscopy

Curriculum Vitae


235

Curriculum Vitae

Verónica Rico Pérez was born on the 11th of January of 1981 in

Alicante, Spain. She earned her Chemistry degree from University of

Valencia in 2008. After that, she spent one year in Germany doing

internship in the company Sartorious Biotech GmbH in Goettingen. In 2009

she started a master in Materials Science at the University of Alicante and

followed her PhD studies in the Inorganic Chemistry Department of this

University. This doctoral thesis describes the most important findings of the

research she performed in N2O decomposition and diesel-SCR for NOx

abatement from 2009 to 2013.

List of publications

1. G. Maniak, P. Stelmachowski, A. Kotarba, Z. Sojka, V. Rico-Pérez, A. Bueno-López. Rationales for the selection of the best precursor for potassium doping of cobalt spinel based deN2O catalyst. Applied Catalysis B: Environmental 136–137 (2013) 302.

2. M. Valencia, E. López, S. Andrade, Iris M.L., N. Guillén-Hurtado, V. Rico-Pérez, A. García-García, C. Salinas-Martínez de Lecea, A. Bueno-López. Evidences of the cerium oxide-catalyzed DPF regeneration in a real diesel engine exhaust. Topics in Catalysis 56 (2013) 452.

3. V. Rico-Pérez, M.A. Velasco-Beltrán, H. Qinggang, W. Qi, C. Salinas-Martínez de Lecea, A Bueno-López. Preparation of ceria-supported rhodium oxide sub-nanoparticles with improved catalytic activity for CO oxidation. Catalysis Communications 33 (2013) 47.

4. N. Guillén-Hurtado, V. Rico-Pérez, A. García-García, D. Lozano-Castelló, A. Bueno-López. Three-way catalysts: past, present and future. Dyna, año 79, Edición Especial, pp. 114. Medellín, Octubre 2012. ISSN 0012-73533.

5. V. Rico-Pérez, S. Parres-Esclapez, M.J. Illán-Gómez, C. Salinas- Martínez de Lecea, A. Bueno-López, Preparation, characterization and N2O decomposition activity of honeycomb monolith-supported Rh/Ce0.9Pr0.1O2 catalysts. Applied Catalysis B: Environmental 107 (2011) 18.

http://link.springer.com/search?facet-author=%22M.+Valencia%22

http://link.springer.com/search?facet-author=%22E.+L%C3%B3pez%22

http://link.springer.com/search?facet-author=%22S.+Andrade%22

http://link.springer.com/search?facet-author=%22Iris+M.L.%22

http://link.springer.com/search?facet-author=%22N.+Guill%C3%A9n+Hurtado%22

http://link.springer.com/search?facet-author=%22V.+Rico+P%C3%A9rez%22

http://link.springer.com/search?facet-author=%22A.+Garc%C3%ADa+Garc%C3%ADa%22

http://link.springer.com/search?facet-author=%22C.+Salinas+Mart%C3%ADnez+de+Lecea%22

http://link.springer.com/search?facet-author=%22A.+Bueno+L%C3%B3pez%22

Curriculum Vitae

236

List of contributions to international conferences

1. V. Rico-Pérez, C. Salinas-Martínez de Lecea, A. Bueno-López.

Effect of calcination method in the N2O decomposition RhOx/CeO2

catalysts. 7th International Conference on Environmental Catalysis

(ICEC). Lyon, France. 2012.-Poster

2. M. Valencia, E. López, S. Andrade, M.L. Iris, N. Guillén-Hurtado,

V. Rico-Pérez, A. García-García, C. Salinas-Martínez de Lecea,

A. Bueno-López. Power-bench demonstration of the Ceria-

Praseodymia-catalyzed DPF regeneration. 7th International

Conference on Environmental Catalysis (ICEC). Lyon, France.

2012.-Poster

3. V. Rico-Pérez, C. Salinas-Martínez de Lecea, A. Bueno-López. Efecto del método de calcinación en la interacción metal-soporte en catalizadores RhOx/CeO2. XII Congreso Nacional de Materiales/ XII Congreso Iberoamericano de Materiales. pp. 40. Alicante, Spain. 2012.-Oral. ISBN 978-84-695-3316-1.

4. M. Valencia, E. López, S. Andrade, Iris M.L., N. Guillén-Hurtado, V. Rico-Pérez, A. García-García,C. Salinas-Martínez de Lecea, A. Bueno López. Evidences of the cerium oxide-catalyzed DPF regeneration in a real diesel engine exhaust. The ninth international Congress on Catalysis and Automotive Pollution Control (CAPoC9). Brussels, Belgium. 2012.-Oral.

5. G. Maniak, P. Stelmachowski, V. Rico-Pérez, A. Bueno-López A. Kotarba, Z. Sojka. Supported Co3O4 catalyst for deN2O reaction: preparation, characterization and activity. XLIV Polish annual conference on catalysis. Cracow, Poland. 2012.-Poster

6. V. Rico-Pérez, S. Parres-Esclapez, M.J. Illán-Gómez, C. Salinas-Martínez de Lecea, A. Bueno-López. Rh/Ce0.9Pr0.1O2/monolith preparation, characterization and catalytic performance for N2O decomposition. EUROPACAT X. Glasgow, United Kingdom. 2011.-Poster.




















237

7. V. Rico-Pérez, S. Parres-Esclapez, A. Bueno-López, M.J. Illán-Gómez, C. Salinas-Martínez de Lecea. Preparation, characterization and catalytic performance for N2O decomposition of Rh/Ce0.9Pr0.1O2/monolith. 2

nd International Symposium on Air

Pollution Abatement Catalysis. APAC. Cracow, Poland. 2010.- Poster. ISBN 978-83-926523-3-5

List of contributions to national conferences

1. V. Rico, S. Parres, M.J. Illán, C. Salinas, A. Bueno. Catalizadores Rh/Ce0,9Pr0,1O2 soportados en monolitos de cordierita. Preparación, caracterización y ensayos catalíticos de descomposición de N2O. La catálisis ante la crisis energética y ambiental (Reunión de la Sociedad Española de Catálisis SECAT). pp. 169. Zaragoza, Spain. 2011.-Oral. ISBN 978-84-939090-0-0.

2. M. Valencia, E. López, S. Andrade, M.L. Iris, N. Guillén, V. Rico, A. García, C. Salinas, A. Bueno. Demostración en condiciones reales de la regeneración catalítica de filtros de partículas diesel utilizando óxidos de cerio. La catálisis ante la crisis energética y ambiental (Reunión de la Sociedad Española de Catálisis SECAT). pp.199. Zaragoza, Spain. 2011.-Poster. ISBN 978-84-939090-0-0.

3. V. Rico-Pérez, S. Parres-Esclapez, M.J. Illán-Gómez, C. Salinas-Martínez de Lecea, A. Bueno-López. Catalizadores Rh/Ce0.9Pr0.1O2 soportados en monolitos de cordierita. Preparación, caracterización y ensayos catalíticos de descomposición de N2O. XIV Reunión Científica Plenaria de Química Inorgánica y VIII Reunión Científica. Plenaria de Química de Estado Sólido (QIES). Cartagena, Spain. 2010.-Oral. ISBN 978-84-693-3704-2.


Carbon Materials and Environment Research Group

Inorganic Chemistry Department, University of

Alicante,

Apto. 99, San Vicente del Raspeig

E-03080 Alicante, Spain

E-mail: [email protected]

This thesis gathers a full research process ranging from basic or

fundamental investigation to actual implementation. The

environmental problems caused by N2O and its main sources,

focusing on diesel engines, are described. This work deals with

the preparation conditions for RhOx/CeyPr1-yO2 catalysts, in

powder and supported in several cordierite honeycomb monolith

sizes, for further scaling up the catalytic N2O decomposition

process. The optimized conditions include calcination process,

use of different solvents (water, ethanol and acetone) for

rhodium loading, and amounts of rhodium, cerium and

praseodymium in the catalyst. A major understanding of the

physico-chemical properties and catalytic role of these materials

towards N2O decomposition process has been facilitated by

several characterization techniques and catalytic tests carried

out at different scales, including the monolithic materials

currently employed by the automotive industry.

The production of N2O as undesired NOx reduction product in

diesel engine emissions, which is a drawback of platinum SCR

catalysts, has been solved by using the dual bed configuration,

where both monolith catalysts operated at the same

temperature, and 100 % N2 selectivity has been obtained.

optimization of n2o decomposition rhox/ceria...

Documents