-. .31!
2. 51.3 b u I. ,
3‘ . . ,

 

 

i i...

._ gsssiis ississas sis , 1

7] v... Ev

 

'HESIS

(147/97)

ESR SITY LIBRARIES

Will “I l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IllllllllUlllllllllllllll

3 12930

This is to certify that the

thesis entitled

New Synthetic Methods for
Preparation of Cerium Promoted
y-ALzOBCatalysts

presented by

Radu Craciun

has been accepted towards fulfillment
of the requirements for

M . s. degree in ChCM;ST}}/

O (,ng

Major progs'or
Date 9 August (cl GIL-

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

LIBRARY _

Michigan State
University

 

 

 

\

PLACE ll RETURN BOX to roman this Mom from your ncord.
To AVOID FINES Mum on or baton dd. duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_—ll—_T_T

MSU I. An Affirmative Action/Equal Opportunity Intuition
W

 

m1

 

NEW SYNTHETIC METHODS FOR PREPARATION OF CERIUM
PROMOTED y-ALzO3 CATALYSTS

By

Radu Craciun

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Chemistry

1996

ABSTRACT

NEW SYNTHETIC METHODS FOR PREPARATION OF CERIUM
PROMOTED y-ALZO3 CATALYSTS

By

Radu Craciun

In this study, the structure of Cer/y-AIZO3 catalysts have been investigated using
diffuse reflectance infrared spectroscopy, X-ray diffraction, and X-ray photoelectron
spectroscopy. Fourier transform infrared and 1H nuclear magnetic resonance
spectroscopy have provided information about the structures of various precursors used
for impregnation. CeOz/AIZO3 catalysts were prepared by incipient wetness using Ce4+
amonium nitrate and by wet impregnation using cerium methoxyethoxide precursor.
Different cerium loadings on A1203 supports have been considered in this study.
Impregnation with nitrate solutions occurs with formation of Ce(OH)x which interacts
with the support through hydrogen bonds, whereas the wet impregnation occurs through a
grafting process which implies a covalent bond formation between the cerium alkoxide
and the surface hydroxyl groups on alumina. After catalyst calcination, the ratio between
the crystalline and amorphous CeOz on the A1203 is a function of loading, thermal
treatment, and precursor. Ce02 particle size have increased with the cerium loading and
calcination temperature. Surface analysis have indicated that cerium is present as Ce4+ on

the catalysts surface.

to my father

iii

ACKNOWLEDGMENTS

I would like to thank Dr. Jeffrey Ledford and Dr. James Jackson for their support
and guidance during the course of this work and for all the knowledge I obtained in this
period of time. I would like to thank Liliana for the understanding me during these years.
I would also like to thank the Ledford group members, especially to my friends Paul and

Mike, for their help and assistance during the work in the lab, for the past two years.

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................ vi

LIST OF FIGURES ........................................................................................................... vii

INTRODUCTION ............................................................................................................... 1

CHAPTER 1

INTRODUCTION TO POLLUTION CONTROL CATALYSIS ...................................... 2
1.1. Overview of Emission Control Reactions ....................................................... 3
1.2. Rare earth promoters used in different oxidation reactions ........................... 10
1.3. References ...................................................................................................... 16

CHAPTER 2

NEY SYNTHETIC METHODS FOR PREPARATION OF CERIUM PROMOTED

y-AL203 CATALYSTS ...................................................................................................... 20
2.1. Introduction .................................................................................................... 20
2.2. Experimental ................................................................................................. 22
2.3. Results and discussion ................................................................................... 26
2.4. Conclusions .................................................................................................... 52
2.5. References ...................................................................................................... 53

CHAPTER 3

THE EFFECT OF DIOL COMPLEXATION AND CALCINATION ATMOSPHERE

ON THE STRUCTURE OF CEOzly-ALZO3 CATALYSTS ............................................. 56
3.1. Introduction .................................................................................................... 56
3.2. Experimental .................................................................................................. 58
3.3. Results and discussion ................................................................................... 60
3.4. Conclusions .................................................................................................... 73
3.5. References ...................................................................................................... 74

CHAPTER 4

FUTURE WORK ............................................................................................................... 76
4.1. Preparation of transition metal oxide catalysts .............................................. 76
4.2. Characterization of mixed metal oxide catalysts ........................................... 77
4.3. Mechanism and kinetics of emission control reactions ................................. 79
4.4. References ...................................................................................................... 81

Table 2.1

Table 2.2

Table 2.3.

Table 3.1.

Table 3.2.

LIST OF TABLES

Semiquantitative XRD data and Ce02 particle size determined from line
broadening calculations for the Ce8-catalysts ........................................... 39

The effect of cerium content on surface area and particle size evaluated
from XRD line broadening of CeyAl catalysts calcined at 500°C ............. 42

Concentration of crystalline Ce02 in CeyN, CeyA and CeyHG catalysts
calculated from quantitative XRD data ...................................................... 47

1H NMR data of mesa-2,4-pentanediol for modified cerium alkoxide
precursors ................................................................................................... 62

Ce02 article size and crystallinity of DquA catalysts calcined at 500°C
determined from XRD line broadening calculations ................................. 66

vi

Figure 2.1

Figure 2.2

Figure 2.3

Figure 2.4

Figure 2.5

Figure 2.6

Figure 2.7

Figure 2.8

Figure 2.9

Figure 2.10

Figure 2.11

Figure 2.12

LIST OF FIGURES

DRIFTS spectra of standard compounds: a) y-Ale3 support; b) Ce02
obtained from pure cerium nitrate calcination; c) Ce(OH)4 ....................... 27

DRIFTS spectra acquired during preparation of the Ce8N catalysts:
a) pure y-AIZO3; b) dried at 120°C; c) calcined at 500°C .......................... 29

DRIFTS spectra acquired during preparation of the Ce8NR catalysts:
a) pure y—AIZO3; b) dried at 120°C; c) calcined at 500°C .......................... 30

DRIF TS spectra acquired after thermal treatment of the Ce8A catalysts:
a) pure y-A1203; b) fresh; c) dried at 120°C; d) dried at 200°C;
e) calcined at 300°C; 0 calcined at 500° C ................................................ 32

DRIFTS spectra acquired afler thermal treatment of the Ce8HG catalysts:
a) pure y-A1203; b) fresh; c) dried at 120°C; (1) dried at 200°C; e) calcined
at 300°C; f) calcined at 500°C ................................................................... 33

XRD patterns for Ce8N catalysts treated at different temperatures:
a) 25°C; b) 120°C; c) 500°C; (1) 800°C ..................................................... 35

XRD patterns for Ce8A catalysts treated at different temperatures:
a) 25°C; b) 120°C; c) 500°C; d) 800°C ..................................................... 37

XRD patterns of the Ce8-series of catalysts calcined at 500°C:
a) pure y-AIZO3; b) Ce8N; c) Ce8NR; d) Ce8A; e) Ce8HG ....................... 38

Comparison of XRD patterns of the Ce8HG and Ce8A catalysts calcined
at different temperature: a) Ce8A calcined at 500°C; b) Ce8A calcined at
800°C; c) Ce8HG calcined at 500°C; (1) Ce8HG calcined at 800°C ......... 41

X-ray diffraction patterns measured for CeyN catalysts: a) A1203;
b) CelN; c) Ce2N d) Ce4N; e) Ce6N ........................................................ 43

X-ray diffraction patterns measured for CeyA: a) A1203; b) CelA;
c) Ce2A; d) Ce4A; e) Ce6A ....................................................................... 45

X-ray diffraction patterns measured for CeyHG catalysts: a) A1203;
b) CelHG; c) Ce2HG; d) Ce4HG; e) Ce6HG ............................................ 46

vii

Figure 2.13 Ce XPS spectra for standard cerium compounds: a) Ce3+AlO3;
b) Ce4+02 .................................................................................................... 50

Figure 2.14 Ce XPS photoreduction spectra for Ce8A catalyst, calcined at 800°C after:
a) 5 min.; b) 15 min.; c) 30 min.; (I) 1 h; e) 4 h, of scanning ..................... 51

Figure 3.1 1H NMR spectra of the precursor solutions: a) DOCeAp; b) DlCeAp
c) D2CeAp; d) D3CeAp ............................................................................. 63

Figure 3.2 FTIR spectra of the precursor solutions: a) DOCeAp; b) DlCeAp;
c) D2CeAp; d) D3CeAp ............................................................................. 65

Figure 3.3 XRD patterns for the DquA catalysts calcined at 500°C: a) A1203;
b) DOCeA; c) DlCeA; d) D2CeA; e) D3CeA catalysts ............................. 67

Figure 3.4 XRD patterns for the DquA catalysts calcined at 800°C: a) DOCeA;
b) DlCeA; c) D2CeA; d) D3CeA catalysts ............................................... 69

Figure 3.5 XRD patterns for the DOCeA catalysts calcined under different
atmosphere: a) pure y-AIZO3; b) DOCeA calcined under hydrogen;
c) DOCeA calcined under air ...................................................................... 70

Figure 3.6 XPS spectra of the DquA catalysts calcined at 500°C: a) DOCeA;
b) DlCeA; c) D2CeA; d) D3CeA catalysts ............................................... 72

viii

INTRODUCTION

A number of previous studies have examined the effects of preparation method,
cerium content, and catalyst pretreatment on the structure of Ce/Ale3 catalysts.
However, little effort has been devoted to the use of different cerium precursors to control
the chemical state and dispersion of the cerium supported on y-Ale3 catalysts. In this
study, the effects of cerium precursor, impregnation methods, and conditioning
treatments on the structure of Ce/A1203 catalysts has been examined using diffuse
reflectance infrared Fourier spectroscopy (DRIF TS), X-ray diffraction (XRD) and X-ray
photoelectron spectroscopy (XPS). Combined information from these techniques will be
used to gain information about the structure of the modified precursors, and the
modifications observed in the structure of Ce/AIZO3 catalysts. The work focuses on
developing new synthetic methods that offer control over the bulk and surface structure

of Ce02 deposited on a y-alumina support.

CHAPTER 1

INTRODUCTION TO POLLUTION CONTROL CATALYSIS

1.1. Overview of Emission Control Reactions

Even though some catalytic reactions were previoudly known and used for
production of different chemicals, e.g. ether from alcohol dehydration, the term
“catalysis” was introduced into the scientific chemical language, only in 1835 by
Berzelius. He was trying to name and explain some of his observations regarding the
influence of different additives to a reaction which led to an increase in its rate.
Doberreiner revealed a classic example of the catalysis process in an attempt to produce
water from hydrogen and oxygen. He directed a jet of hydrogen at a powdered platinum
catalyst. The fine particles instantly became hot and the hydrogen caught fire. This was
the scientific principle for the so called “feuerzeug” meaning lighter or tinder box [1,2].
So, total oxidation reactions were among the first chemical transformations performed
under heterogeneous catalysis conditions, when platimun played the role of the solid
catalyst for the gas phase reaction between H2 and 02. From that time, the demand for
different chemicals required by the industry and the great technical developments in new
analysis techniques and apparatus allowed the introduction of a wide variety of catalysts
for total oxidation or combustion reactions.

The great interest in the last years in environmental protection and air pollution
control, has necessitated new technology and instruments development. Air pollution
causes respiratory illness and other diseases, damage to plants and animals, and other

effects on a global scale. Air pollution is defined as any atmospheric condition in which

substances are present at concentration high enough above their normal ambient levels to
produce a measurable effect on man, animals, vegetation, or materials. The study of
emission sources, such as the internal combustion engine, burning of coal and oil in
industrial furnaces, requires a knowledge of the chemistry of combustion as well as of
engineering aspects. The secondary products from a chemical process or the refitse from
different technologies, can be an important factor in pollution. The high cost of burning
these materials can play a key role for large investments, which may lead to the
development of new low-cost technology for environmental protection. One of the best
ways to control air pollution is to prevent contamination from entering the atmosphere.
In fact there are three strategies available for reducing the impact of chemicals on the
environment: waste minimization, emission abatement, and remediation. Waste
minimization calls for the design and development of products and processes that are
inherently low polluting or non-polluting. The abatement of emission, considered mostly
for air pollution minimization, can often be achieved by trapping harmful effluents or
converting them to harmless substances. Remediation is the name given to the use of
catalytic processes to clean up contaminated soil or water. This process is potentially the
most cost-effective approach, either by using various solid materials or enzymes as
catalysts, for contaminant decomposition into harmless compounds.

Unpolluted air is a concept implying a composition of air which would not
adversaly affect people’s health. The substances usually considered air pollutants are
chemicals such as sulfur (SOX), halogen (C3HxCly), or nitrogen (NOX) based compounds

and radioactive compounds (Rn). Concentrations of air pollutants are measured in ppm

or ug/m3. Recent studies have shown that high concentrations of chlorinated
hydrocarbons influence the ozone concentration in the stratosphere. Protection of the
ozone layer by controlling emissions from chemical industry has been a significant
scientific concern. Particular attention has focused on the mechanisms involved in
atmospheric ozone formation and consumption due to UV radiation and various other
chemicals present in the air [3].

Combustion of waste materials is another way to avoid environmental pollution.
Catalytic combustion or incineration as an economical alternative to simple combustion is
an important part of modern emission control technology. Interest has therefore
developed in the study of the catalytic total oxidation reactions. The main advantages
offered by catalytic incineration are the wide range of fuel concentration used in the
process and the low temperature at which the reactions can occur compared with the
uncatalyzed ones.

Catalytic exhaust gas treatments, especially for automobiles, operate in a
temperature range of 600-800°C. The use of total oxidation catalysts for high-
temperature processes (>1000°C) became more difficult because catalysts such as
supported noble metals or transition metal oxides are not resistant to such conditions.
Therefore, it has become necessary to develop new materials which may be used as
catalysts or supports for these types of processes. The development of advanced
heterogeneous catalysts involves several important activities. These include the

development of sophisticated preparation procedures that can be used to design specific

surface structures, guided by examination of the active sites as well as promoters/supports
and promoters/transition metal oxides interactions in the active catalytic phases.

At present, the catalyst systems used in these processes are based on noble metals
which are expensive and susceptible to poisoning. Transition metal oxide catalysts have
been suggested as suitable substitutes for noble metals even though they are not as active
and are also subject to poisoning. The advantages of transition metal oxides are their low
cost and better resistance to chlorine poisoning. By improving their stability, selectivity
and activity using suitable promoters, transition metal oxides may become good
replacements for noble metal-based catalysts.

In thermal combustion (1500-2000°C), gas-phase radicals are initiated thermally
after which oxidation reactions occur rapidly. At high temperature nitrogen oxides are
formed from nitrogen and oxygen in the air. The presence of these nitrates oxides in
exhaust gases is undesirable from a pollution control point of view. Control of the
reaction at these high temperatures is very difficult due to the limited flexibility in the
fuel concentration. These difficulties challenge science to develop new catalytic
processes which can avoid the inconvenience of thermal incineration.

The mechanisms involved in catalytic incineration are complex and a better
understanding of the processes which take place on the catalyst surface would be
desirable. Prasad et a1. [3] summarized some of the properties for suitable catalysts used
in catalytic incineration processes. Among these can be mentioned a low ignition
temperature of the fuel/air mixture, high catalytic activity and durability over time,

resistance to high temperature and mechanical shocks, and suitable supports with high

surface area. There are a few classes of materials which approach all these requirements,
but none that fulfilled all.

Noble Metals. Catalytic incineration over noble metal catalysts has been widely
used for decades. The advantages of these catalysts are given by the high catalytic
activity and selectivity. Platinum, palladium and rhodium are the most commonly used
active phases for catalytic combustion applications [4-6]. A desirable characteristic of the
noble metal catalysts is their high resistance to sulfur poisoning compared with the
transition metal oxide catalysts [7]. Highly dispersed Pt and Pd catalysts are easily
prepared using a number of support materials like silica or alumina.

Oxidation over noble metals is considered to be a structure sensitive reaction
which means that the surface configuration of the catalyst will strongly influences the
activity, selectivity, reaction pathway and the rate [8-12]. Briot et al. [8-9] have
attributed the activity of supported platinum and palladium catalysts to the high
specificity of surface oxygen on large crystallites. For these type of catalysts, an activity
enhancement was observed at low temperature, close to the catalyst ignition temperature.
Catalysts with high noble metal dispersion (small crystallites) exhibit high catalytic
activity [10]. No oxygen pressure dependence on the reaction rate was found. In the case
of methane oxidation, the reaction order was found to be around one in methane with
little influence of oxygen concentration on the reaction rate. The dissociation of the
methane or oxygen-methane complexes at the catalyst surface was proposed to be the
rate-limiting step in the mechanism [12].

Transition Metal Oxides. Oxidation over solid metal oxide catalysts has been

widely studied during the last years. The main advantage over noble metal catalysts is

the low cost of the raw material. By choosing a well defined composition of the metal
oxide in the catalyst, an oxidation activity close to that of the noble metal catalysts can be
obtained. It was found that the formation of nitrogen oxides can be avoided using metal
oxide catalysts [13]. Oxygen can be activated by interaction with the catalyst surface.
Coordination or incorporation into the oxide lattice may be an important step in
catalyst activation. Lattice incorporation leads to a more strongly bound oxygen than in
the case of simple physical adsorption. The lattice oxygen was found to be more active in
the case of selective oxidation processes whereas the surface oxygen was more important
in complete oxidation reactions [14]. Participation of lattice oxygen in complete
oxidation reactions increases with temperature. Wise et al. [15] observed that the ratio of
the rate constants corresponding to the reactions of lattice oxygen and adsorbed oxygen
increases with temperature from 0 at 500°C to 0.3 at 800°C for methane oxidation on a
LaFeO3 catalyst. At higher temperatures lattice oxygen interactions dominate the
oxidation process. Many studies have been focused on the suitability of different types of
metal oxides for catalytic incineration, trying to rationalize their catalytic activity.
Comparison studies of the catalytic activities in terms of conversion for a variety of
oxides such as Cr203, C0203 [16], Mn203 [17] V205 [18], or CuO [19] led to the
conclusion that many of these oxides are potentially good catalysts for total oxidations.
Mixed oxides with perovskite structures XYO3, (where X and Y are different
transition metals), revealed high activity in total oxidation reaction [20]. The use of
perovskite-type oxides as catalysts was first reported in 1970, by Meadowcrofi [21].

Their well defined and stable structure make them suitable for oxidation catalysts.

Perovskite electronic structures are comparable with those of the transition metals.
Seiyama [20] summarized in a review article the important aspects of total catalytic
oxidation reaction on various perovskite oxides, like sorption and desorption of oxygen,
H2 oxidation and total oxidation of hydrocarbons. Many of the pervoskite structures
incorporated lanthanides (La, Ce, Pr) and/or alkaline earth metals (Ba, Sr, Ca) in
combination with transition or noble metals. Structures like Lal_xerBO3, La1,xerMnO3,
La1_xerCoO3 [22-24] are some of the most recent perovskite-type catalysts prepared and
used successfully in total oxidation reactions.

From the material presented above we can conclude that a great number of high-
temperature-stable complex oxide materials are promising catalysts for catalytic
incineration.

Supports. Another important element in the structure and composition of catalyst
is the support. The chemical nature of these may play an important role in the catalyst
activity. Supports are mostly used to improve thermal and chemical stability while
decreasing the amount of the expensive active component required to make a catalyst.
The surface area and chemical nature of the support play major roles in the behavior of
the catalysts. The most often used supports are based on silicates and alurninates, but
often less resistant, low surface area materials with low surface area such as MgO or
TiOz, are used to improve the selectivity of a catalyst or as a catalyst site itself. The
geometry and the dimension of the particle used as support are very important factors in
the catalyst activity. The characteristics needed for a support to be used in high-
temperature reactions such as combustion can be summarized as follows: i) high thermal

stability and mechanical strength; ii) no chemical reaction between support and the

catalytic component; iii) high volume throughput [25]. It is hard to find an ideal material
that meets all the requirements for an excellent support. Many times, additives or
promoters are used to improve the properties of the supports used for catalyst preparation.
Silica is one of the most commonly used support materials; at high temperature and in the
presence of oxygen, silica exhibits sintering phenomena. y-AIZO3 is another support that
shows resistance to the sintering process compared to silica. However, at very high
temperature (1000°C) the y-A1203 phase is transformed into a-Ale3 which has a lower
surface area [26]. Many studies on alumina based catalysts show loss in surface area due
to sintering of particles by solid-state diffusion. Consequently, the thermal treatment
used for the support material also affects a catalyst’s properties [27].

As was previously mentioned, addition of promoter may stabilize the support
inhibiting the sintering process. Promoters can be divided in two different classes:
structural and textural. Structural promoters will lead to a transformation of the catalysts
surface configuration which will affect the catalyst activity. Textural promoters will have
less influence on the structure of the catalyst but will improve the textural properties of
the catalyst like surface area, pore volume size, and mechanical resistance of the catalyst
surface. Alkali earth metals (Ba, Ca, Sr) are well known as additives with consequence
on the textural properties as well as mechanical resistance of the catalysts. Lanthanide
oxides are known to be used as both textural and structural promoters, influencing the
thermal resistance and the surface configuration of the catalyst. Another way to
promoter a catalyst is to introduce lattice flaws into the structure. These kind of

promoters were called electronic promoters, substances with semiconductor character.

10

These materials will accept or lose electrons from conduction band. Impurities in the

lattice of the catalyst will determine an increase its catalytic activity.

1.2. Rare Earth Promoters used in Different Oxidation Reaction

As was described before, catalysis has become an important field of study for
environmental pollution control processes. Dispersion of the active component, loading
and surface structure are important parameters in designing a proper catalyst. The main
chemical effect of a catalyst in total oxidation reactions is to force them to completion,
yielding water, carbon dioxide and nitrogen as final products (assuming only C, O, H, N
are in the feed). Another important effect is a decrease in the incineration temperature.

Metal oxide based catalysts are of particular interest for these types of processes.
Procedures such as temperature treatment, or additives such as promoters or activators
may strongly influence the surface structure of the catalyst and its catalytic activity.
Manipulation of these factors during catalyst preparation offers control over activity,
selectivity and resistance of the catalyst to various poisons or inhibitors. Tailoring a
catalyst surface by controlling the dispersion and/or particle size of the active phase,
crystalline structure and oxidation state of the metals on the surface, ultimately gives
control over the catalytic process.

Promoter effect. Rare earth additives have been employed extensively as textural
and structural promoters in supported metal catalysts [28-30]. Cerium oxide in particular
has been widely used for these purposes [31-39]; its promoter effect is given by the
nonstoichiometrical crystal structure [40-41], oxygen storage capabilities [42-45], redox

properties [46] and very good thermal resistance [47-53]. Cerium promotion enhances

II

the catalytic activity of transition metals and/or transition metal oxides supported
catalysts for CO oxidation and NOx reduction because of its ability to form oxygen-
deficient oxides under reducing conditions [49,40]. Due to their importance in a variety
of catalytic applications, the structures of supported cerium catalysts have been the focus
of many investigations [29-59]. Miki et al. [42] have used XRD to examine the structure
of a 20 wt.% CeOz/Ale3 catalyst oxidized and reduced at high temperature (900°C).
High temperature reduction of the catalysts led to the CeAlO3. Reoxidation of the
reduced catalysts had no effect on LaAlO3 whereas CeAlO3 was reoxydized to CeOz.
Shyu et a1. [57] reported XPS, Raman spectroscopy, and TPR results for a series of
Ce/AIZO3 catalysts. They concluded that Ce existed as a CeAlO3 precursor, small Ce02
crystallites, and large Ce02 particles on Ce/Al203 catalysts. Recently, Graham et al. [3 5]
reported that lanthanum promotion increases the dispersion and the range of reversible
reducibility of alumina supported CeOz.

Spectroscopic measurements combined with ab initio calculations have shown
that in the case of supported rare earth oxides on alumina, modification of the electronic
structure of the lantlranide cation occurs affecting the catalytic properties of the catalyst.
Charge redistribution over an alumina surface implies a modification of the acid-base
characteristic of the active sites [60]. Lanthanide cations are known to have strong
interaction with the support which results in an electronic density redistribution. Metal-
support interactions influence both the electronic state of the metal and the surface
morphology. The A1203 support provides a weak ligand field, favoring the high-spin

state of the metal and a change in the oxidation state [61].

12

The crystallographic and solid-state properties of the lanthanide oxides can be
largely limited to the so-called sesquioxide stoichiometry (M203), the only stable
composition ordinarily observed for these materials. Exceptions from this structure have
been observed in the cases of Equ and SmOx. Dioxides having a face-centered cubic
structure (fluorite type structure), are known in the case of Ce, Pm and Th but only the
cerium stoichiometry is more stable than the corresponding sesquioxide structure, mostly
due to the stability of the Ce“ oxidation state. Ce203 can be prepared only with
difficulty, by extended reduction of Ce02 at high temperature or by vacuum
decomposition of certain salts. Ce203 oxidizes easily in air to the dioxide. In addition,
the oxide system of Ce and Tb contain well-defined composition of LnOx (1.5 < x < 2.0).
In all three systems mentioned above as characteristic for lanthanides oxides, the stable
phases shows variable Ln3+/Ln4+ ratios together with the appropriate number of 02' ions
needed to achieve electrical neutrality. Rare earth sesquioxides exhibit characteristic
polymorphism and can exist in one or more distinct crystalline modifications, depending
on the radius of the tetravalent metal ion and the temperature for oxide preparation and
quenching. Pertinent parameters of the rare earth cations and stable rare earth oXide
phases are available in literature [62]. X-ray diffraction patterns of most of these
materials are also available for comparative purposes [63]. Various oxide modifications
may be further distinguished by bond differences that occur in the 200-700 cm'l, in the
IR-spectral region. All anhydrous rare earth oxides are z75°/o ionic having a good basic

character (important property in catalytic purpose).

l3

Lanthanide oxides undergo dehydration/rehydration processes. This cyclic
process was used to prepare Ln203 having highly reproducible catalytic activities for
alcohol dehydrogenation/dehydration and for double bond isomerizations of olefins [64].
For some reactions, catalytic activities are independent of electronic or magnetic
properties of the oxides and is determined by relative acid-base properties or by the
surface structure of the catalysts whereas for other types of reactions paramagnetic nature,
lattice oxygen mobility and cation variable valence play important roles, governing the
catalytic behavior. For instance, the paramagnetic rare earth oxides have been shown in
several studies to be effective catalysts for ortho-para hydrogen conversion [65]. The rate
is strongly influenced by the temperature which affects the paramagnetic properties of the
catalyst materials. Diamagnetic oxides such as Y203 or Lu203 can be activated by
thermal treatment which apparently generates the paramagnetic ions Y2+ and Lu2+ [66].

Preparation methods. The precursors and/or methods used for preparation can
strongly influence the surface structure of a catalyst and consequently its catalytic activity
and selectivity. The most widely used precursor for cerium impregnation utilized often
for preparation of three way catalysts for automotive exhaust systems, is the Ce3+-nitrate
solution [67]. After calcination, a well dispersed cerium phase is observed on the A1203
support, at low cerium loading [67-68]. Miki et al. [42] prepared a cerium catalyst by
incepient wetness impregnation and compared its surface structure with a cerium catalyst
prepared by the sol-gel technique, which uses Ce3+'nitrate dissolved in ethylene glycol
and aluminum tri-isopropoxide. Different Cer surface structures were obtained which

led to modified catalytic behavior.

l4

Ion exchange is an alternative method used for preparation of catalysts [69, 70].
The method is based on the electrostatic adsorption of the ions at a pH lower than the
isoelectric point (IEP) characteristic for each support when the surface is positively
charged and anion adsorption can occur. Positive and negative charged support surfaces
can adsorb positive or negative charged ions from the solution. The exact chemistry
depends on support, pH and species in solution. The pH value of the solution will act as
a surface charge selection switch, favoring the deposition of a given ion. Using this
method, molybdenum was successfully well dispersed on TiOz [71] and A1203 [72]
supports. Difficulties in controlling the equilibrium of ionic species present in the
precursor solutions at various pH values preclude the general utilization of this technique.

For metal impregnation, Kummer et al. [73] proposed a method which involves
thermal transport from bulk phase of a metal foil to a dispersed phase on the support
surface. The metal concentration on such catalysts is low but highly dispersed. In this
way, cerium was deposited on the support by incipient wetness impregnation using Cey-
nitrate solution, before the transition metal impregnation. Such catalysts used CeOz as an
additive favoring the metal transport (in this case, Pt) and dispersion, leading to a very
complex CeOz-metal surface structure.

Dufour et a1. [74] used organometallic compounds for rhodium impregnation of a
silica support. The interaction mode of the Rh(n3-C3H5) with the oxide support varies
according to the nature and the degree of hydroxylation of the support. Different grafted
rhodium species were found on the support surface after impregnation. Didillon et al.

[75] extended the use of organometallics for other transition metals (group VIII-metals)

15

and found that highly dispersed metal phases can be successfully anchored onto low-
surface area A1203. Usmen et al. [76] used a La3+ pretreated alumina as a support to form
a well dispersed. cerium phase during impregnation. They found that well dispersed La”-
phases interact with CeOz leading to a better cerium and platinum dispersion. Higher
cerium dispersion leads to the retention of more reducible oxides such as CeAlO3 with
consequences on the oxygen storage capacity of the catalyst [5 7].

Van Hengsturn et al. [77] and later Baiker et al. [18] have developed a synthetic
method for catalyst preparation that involves alkoxide grafting using a metal-organic
compound. Generally, an alkoxide grafting reaction may be written as follows:

Support-OH + M - (OR)x —> Support - O - M (OR)x,1 + R-OH (1-1)

The chemistry of this method involves the formation of a covalent bond between
the metal center and the support surface, followed by the elimination of an alkoxide
ligand from the metal center (M) as an alcohol. Excellent dispersion was obtained for
vanadium impregnation on different supports (SiOz, A1203 or TiOz), when vanadium tert-
isobutyl solution was used as a precursor.

For example, silica-supported ceria catalysts of various content (less than 11%
Ce) were prepared by anchoring cerium acetylacetonate in organic media and studied
their surface strucuture. The resulting Ce02 particles were 1-3 nm diameter after
calcination in oxygen at 673 K. The highly dispersed ceria obtained by using the cerium
acetylacetonate precursor shows a shift to higher XPS binding energies for Ce3d and 01 3,
increased CO stretching in CO-Ce4+ terminal groups, and a blue shift of the band gap
measured by difiuse reflectance in comparison with catalysts prepared using the classical

preparation by incipient wetness using cerium nitrate [7 8].

16

The use of a “hot grafting” technique refers to a grafting process using a metal-
alkoxide precursor under reflux condition and elevated temperature. This technique can
be an alternative way to improve the interaction between the precursor and support
surface [79]. The high cost as well as the sensitivity to air and water of these metal-
organic compounds [80], confer to the grafting method a disadvantage for large scale
catalysts impregnation. It was found that in some cases, the presence of ethanol in the
system as an exchange ligand influences the reactivity of cerium alkoxide toward grafting
with the hydroxyl groups from the support surface [81].

Recently, lanthanide alkoxide catalysts were successfully used in some organic
reactions like synthesis of vitamins and steroids [82, 83], being promising catalysts for
heterogeneous catalysis processes applied in organic synthesis. If these alkoxides are
prepared from chiral alcohols, the catalysts obtained may be used to perform asymmetric
synthesis, largely applied in the drug industry [84]. A monolayer dispersion of the
impregnated lanthanide-alkoxide, fixed on a solid support, may confer on the catalysts

increased selectivity and activity in these types of organic synthesis.

1.3 References

1. Collins, P. Ambix, 1976, 23, 96.

2. Thomas, J. M. Angew. Chem. Int. Ed. Engl. 1994, 33, 913-937.

3. Prasad, R.; Kennedy, L. A.; Ruckenstein, E. Catal. Rev. - Sci. Eng. 1984, 26(1),
1-58.

4. Hicks, R. F. Chem. Eng. Sci. 1990, 45(3), 2647-2651.

5. Worley, S. D.; Rice, C. A.; Mattson, G. A.'; Curtis, C. W.; Guin, J. A. J. Chem.
Phys. 1982, 76, 20-24.

6. Taylor, K. C. in Catalysis and Automotive Pollution Control by Crucq, A.;
Fressnet, A. 1987, Elsevier, Amsterdam.

7. Johnson, D. W. Jr. J. Catal. 1977, 48, 87-97.

8. Briot, P. Appl. Catal. 1990, 59, 141-152.

10.
ll.
12.
13.
14.
15.
16.

17.
18.

19.
20.
21.
22.
23.
24.
25.

26. '
27.

28.

29.

30.
31.

32.

33.

34.

35.

36.

37.
38.

39

40.

I7

Briot, P.; Primet, M. Appl. Catal. 1991, 68, 301-314.

Hick, R. F. .1. Catal. 1990, 122, 295-306.

Baldwin, T. R.; Burch, R. Appl. Catal. 1990, 66, 359-381.

Otto, K. Langmuir, 1989, 5, 1364-1369.

Ismagilov, Z. R.; Kerzhentsev, M. A. Catal. Rev. -Sci. Eng. 1990, 32(1,2), 51-103.
Arakawa, T.; Tsuchi, Y. S.; Shiokawa, J. J. Catal. 1982, 74, 317-322.

Wise, H.; McCarty, J. G. Catal. Today, 1990, 8, 231-248.

Prasad, R.; Kennedy, L. A.; Ruckenstein, E. Combust. Sci. Technol. 1980, 22,
271-275.

Dwyer, F. G. Catal. Rev. 1972, 6(2), 261.

Kifenski, J .; Baiker, A.; Glinski, M.; Dollenmeier, P.; Wokaun, A. J. Catal. 1986,
101, 1-11.

Lu, G.; Wang, R. Cuilua Xuabao, 1991, 12(4), 261-267 (Ch).

Seiyama, T. Catal. Rev. - Sci. Eng. 1992, 34(4), 281-300.

Meadowcroft, D. B. Nature, 1970, 226, 847.

Jaecnicke, S.; Chuah, G. K. Ber. Bunsenges. Phys. Chem. 1992, 96(1), 1-9.
Arai, H. Appl. Catal. 1986, 26, 265-276.

Zhang, H. M.; Teraoka, Y.; Yamazoe, N. Appl. Catal. 1989, 6(1,2), 155-162.
Zwinkels, M. F. M.; Jaras, S. G.; Menon, P. G.; Griffin, T. A Catal. Rev. - Sci.
Eng. 1993, 35(3), 319-358

Beguin, B.; Garbowski, E.; Primet, M. J. Catal. 1991, 127, 595-604.

Tijburg, I. I. M.; Geus, J. W.; Zandbergen, H. W. J. Mater. Sci. 1991, 26,
6479-6486.

Alvero, R; Bemal, A.; Carrizosa, 1.; Odriozola, J. A. Inorg. Chim. Acta, 1987,
140, 45-48.

Xie, Y. C.; Qian, M. X.; Tang, Y. Q. China-Japan-US. Symposium on
Heterogenous Catalysis Related to Energy Problem, 1982, B. 10C,China.

Yang, J. Y.; Swartz, W. E. Spectrosc. Lett. 1984, I 7, 331-332.

Baiker, B. G.; Clark, N. J. in Studies in Surface Science and Catalysis, Delmon,
B. Grange, P. Jacobs, P. A. Poncelet, G. Elsevier, Amsterdam, 1987, 31, 455-62.
Kieffer, R.; Kiennemann, A.; Rodriquez, M.; Bemal, S.; Rodriquez-Izquierdo, J.
M. Appl. Catal. 1988, 42, 77-86.

Shyu, J. Z.; Otto, K.; Watkins, W. L. H.; Graham, G. W.; Belitz, R. K.; Gandhi,
H. S. J. Catal., 1988, 114, 23-28.

Hay, C. M.; Jennings, J. R.; Lambert, R. M.; Nix, R. M.; Owen, G.; Rayment, T.
Appl. Catal. 1988, 37, 291-304.

Graham, G. W.; Schmitz, P. J .; Usmen, R. K.; McCabe, R. W. Cat. Lett. 1993, 1 7,
175-184.

Masuda, K.; Kawai, M.; Kuna, K.; Kachi, N.; Mizukami, F. in Preparation of
Catalysts VEd. Poncelet, G.; Jacobs, P. A.; Grange, P.; Delmon, B.; Elsevier,
Amsterdam, 1991, 229-238.

Sugiyama, S.; Matsumura, Y.; Moffat, J. B. J. Catal. 1993, 139, 338-350.
Ledford, J. S. Ph. D. Thesis, University of Pittsburgh, 1989.

Dulamita, N. Rev. Rom. Chim. 1985, 36(11), 1050.

Steinberg, M. Anal. Appl. Rare Earth Mater. Nato Adv. Study Inst. 1972, 263-270.

41.

42.

43.
44.
45.
46.

47.
48.

49.
50.
51.
52.
53.
54.
55.
56.
57.
58.

59.
60.

61.
62.
63.

65.
66.
67.

68.

69.

70.
71.
72.

73.
74.

18

Perrichon, V.; Laachir, A.; Bergeret, G.; F rety, R.; Tournayan, L.; Touret, O.; J.
Chem. Soc. FaradayTrans. 1994, 90(5), 773-781.

Miki, T.; Ogawa, T.; Haneda, M.; Kakuta, N.; Ueno, A.; Tat'eishi, S.; Matsuura,
S.; Sato, M. J. Chem. Phys. 1990, 94, 6464-6467.

Yao, H. C.; Yao yu, Y. F. J. Catal. 1984, 86, 254-265.

Su, E. C.; Montreuil, C. N.; Rothschild, W. G. Appl. Catal. 1985, 1 7, 75-86.
Loof, P.; Kasemo, B.; Keck, K. E. .1. Catal. 1989, 118, 339-348.

Haneda, M.; Mizushima, T.; Kakuta, N.; Ueno, A.; Sato, Y.; Matsuura, S.;
Kasahara, K.; Sato, M. Bull. Chem. Soc. Jpn. 1993, 66, 1279-1288.

Schaper, H.; Doesburg, E. B. M.; VanReijen, L. L. Appl. Catal. 1983, 7, 211-220.
Pembo, J.M.; Lenzi, M.; Lenzi, J.; Lebugle, A. Surf Inter. Anal. 1990, 15(11),
663.

Yu-Yao, Y.F.; Kummer, J.T. J. Catal. 1987, 106, 307.

Chojnacki, T.; Krause, K.; Schmidt, L.D. J. Catal. 1991, 128(1), 161.

Oudet, F.; Courtine, P.; Vejux, A]. J. Catal. 1988, 114, 112.

Yang, J .K.; Swartz, W.E. Spectrosc. Lett. 1984, 1 7(6& 7), 331.

Gaugin, R.; Graulier, M.; Papee, D. Adv. Chem. Ser. 1975, 143, 147.
Koberstein, E. Unst. St. Proc. Catal., Matros, Y.S., Ed., 1990, 643.

Yamaguchi, T.; Ikeda, N.; Hattony, H.; Tanabe, K. J. Catal. 1981, 67, 324.
Raen, S.; Braten, N.A.; Grepstad, J.K.; Qui, S.L. Phys. Scr. 1990, 41(6), 1001.
Shyu, J. Z.; Weber, W. H.; Gandhi, H. S. J. Phys. Chem. 1988, 92, 4964.
Ledford, J. S.; Houalla, M.; Proctor, A.; Hercules, D. M. J. Phys. Chem. 1989, 93,
6770-6777.

Dulamita, N.; Pop, A.; Corbeanu, V. Studio, 1984, 29, 3-11.

Capitan, M. J .; Centeno, M. A.; Malet, P.; Carrizosa, I.; Odriozola, J. A. J. Phys.
Chem. 1995, 99, 4655-4660.

Tauster, S. J.; Fung, S. C.; Garten, R. L. J. Am. Chem. Soc. 1978, 100, 170.
Komer, R.; Ricken, M.; Nolting, J. J. Solid State Chem. 1989, 78, 136-147.
Powder Diflraction File, Inorganic Phases, JCPDS, 1983.

Rosynek, M. P.; Cat. Rev. - Sci. Eng. 1977, 16(1), 111-154.

Baron, K.; Selwood, P. W. J. Catal. 1973, 28, 422-426.

Aries, J. A.; Selwood, P. W. J. Catal. 1974, 33, 284-288.

Kubsh, J. E.; Rieck, J. S.; Spencer, N. D. in Catalysis and Automotive Pollution
Control 11, ed. Crucq, A. Elsevier, Amsterdam, 1991, 125-138.

Diwell, A. F .; Rajaram, R. R.; Shaw, H. A.; Truex, T. J. in Catalysis and
Automotive Pollution Control 11, ed. Crucq, A. Elsevier, Amsterdam, 1991,
139-152.

Burch, R.; Flambard, A. R. in Preparation of Catalysts III by Poncelet, G.;
Grange, P.; Jacobs, P. A. Els. Sci. Publ. Amsterdam, 1983, 311-320.

Brunelle, J. P. Pure Appl. Chem, 1978, 50(9,10), 1211-1229.

Luthra, N. P.; Cheng, W. C. J. Catal. 1987, 107, 154-160.

VanVeen, J. A. R.; deWitt, H.; Emeis, C. A.; Hendriks, P. A. J. M. J. Catal. 1987,
107, 579-582.

Kummer, J. T.; Yu-Yao, Y. F. J. Catal. 1987, 106, 307-312.

Dufour, P.; Houtman, C.; Santini, C. C.; Basset, J. M.; Hsu, L. Y.; Shore, S. G.

75.

76.
77.
78.
79.
80.
81.
82.

83.
84.

19

J. Am. Chem. Soc. 1992, 114, 4248-4257.

Didillon, B.; El Mansour, A.; Candy, J. P.; Basset, J. M.; Le Peltier, F.;
Bournonville, J. P. in Preparation of Catalysts Ved. Poncelet, G.; Jacobs, P. A.;
Grange, P.; Delmon, B. Elsevier, Amsterdam, 1991, 717-723.

Usmen, R. K.; Graham, G. W.; Watkins, W. L. H.; McCabe, R. W. Catal. Lett.
1995, 30, 53-63.

Van Hengstum, A. J.; Van Ommen, J. G.; Bosch, H.; Gellings, P. J. Appl. Catal.
1983, 5, 207-217.

Bensalem, A.; Bozon-Verduraz, F.; Delamar, M.; Bugli, G. Appl. Catal. 1995,
121, 81-93.

Smith, P. D.; Klendworth, D. D.; McDaniel, M. P. .1. Catal. 1987, 105, 187-193.
Namy, J. L.; Souppe, J.; Kagan, H. B. J. Org. Chem. 1984, 49, 2045-2049.
Mehrotra, R. C.; Kapoor, P. N.; Botsoara, J. M. Coord. Chem. Rev. 1980, 31,
67-91.

Kirk, D. N.; Slade, C. J. J. Chem. Soc. Perkin Trans, 1980, 1, 2591-2569.
Evans, D. A.; Hoveyda, A. H, J. Am. Chem. Soc. 1990, 112, 6447-6449.
LeBrun, A.; Namy, J. L.; Kagan, H. B. Tetrahedron Letters, 1991, 32(21),
2355-2358.

CHAPTER 2

NEW SYNTHETIC METHODS FOR PREPARATION OF CERIUM
PROMOTED y-ALZO3 CATALYSTS

2.1 Introduction

Cerium oxide has been employed extensively as a textural and structural promoter
for supported metal catalysts. The structural promotion effect was attributed to the
formation of new active sites (crystallites or structural defects) formed by Cer species
on the catalyst surface [1-5]. The textural promotion effect is given by the excellent
thermal and mechanical resistance conferred on the promoted catalyst. Detailed studies
were done to investigate the correlation between catalyst structure and activity. Due to its
many catalytic applications, the structure of supported cerium oxide has been the focus of
many investigations [6-18]. CeOz/y-A1203 catalysts are commonly used in automotive
catalytic converters and for selective or total oxidation reactions. Cerium promotion
enhances the catalytic activity for CO oxidation mostly due to its ability to form oxygen-
deficient oxides under reducing conditions [8]. In discussing the effects of cerium
promoters in three-way catalysts, Nuna et al. [17] concluded that a decrease in the CeOz
crystallites and increase in cerium content led to an increase in catalytic activity and
thermal durability. The improvement in catalyst activity have been attributed to the
ability of CeOz to fix platinum (or Pd) on the catalyst’s surface [19-27].

The structure of a Cer/A1203 system is strongly dependent on: i) preparation
method (impregnation techniques), ii) the chemical nature of the precursors used for

preparation, and iii) temperature treatment during catalyst preparation and iiii) cerium

20

21

loading [IO-22]. Le Normand et al. [20] examined a series of CeOz/AIZO3 and Pd-
CeOz/AIZO3 catalysts supported on y-A1203 using X-ray photoelectron spectroscopy
(XPS) and X-ray absorption spectroscopy (XAS). They found a clear correlation between
the cerium loading and its surface structure. For cerium loading below 1.5 wt%, these
authors reported that cerium exists as Ce3+ cations in octahedral coordinated sites on the
A1203 surface. For higher cerium loading, XPS and XAS data indicated the formation of
three-dimensional cerium oxides. Cerium loading has an effect on the stability and
dispersion of the Pd deposited on the alumina support, as well. Shyu et al. [21] reported
XPS, Raman spectroscopy, and temperature-programmed reduction (TPR) results for a
series of Ce/AIZO3 catalysts. They concluded that Ce existed as a CeAlO3 precursor,
small Ce02 crystallites, and large Ce02 particles on Ce/AIZO3 catalysts. Miki et al. [22]
used XRD to examine the structure of a 20 wt% CeOz/A1203 catalyst oxidized and
reduced at high temperature (900°C). High temperature reduction of the catalysts lead to
the formation of CeAIO3. Reoxidation of the reduced catalysts had no effect on the
LaAlO3 whereas CeAlO3 was reoxidized to CeOz.

Previous literature data enphasized that a key factor to consider in the study of
CeOz/A1203 catalyst system is cerium dispersion. Several approaches can be considered
in order to improve cerium dispersion. One is the use of efficient impregnation
techniques that optimize promoter dispersion on the high surface area support. Pre-
irnpregnation of the support with elements which may favor the cerium dispersion
further improve the cerium oxide dispersion. Another approach is to change the chemical

nature of the precursor solution used for support impregnation to optimize the interaction

22

with the support surface and to avoid as much as possible side reactions (e. g. precipitation
or self polymerization) which may affect the promoter dispersion.

This study is focused on the investigation of chemical interactions and structural
transformations of Cer species supported on a y-A1203 support when various
impregnation techniques and thermal treatments were used for catalyst preparation. The
interaction between the support and cerium promoter will be investigated using diffuse
reflectance infrared Fourier transform spectroscopy (DRIFTS). The effect of cerium
loading and thermal treatment on the dispersion of Ce02 in a series of y-A1203 promoted
catalysts have been examined using X-ray diffraction (XRD). The oxidation state of
supported CeO,‘ has been investigated using XPS. Information derived from the surface
and bulk characterization techniques has been used to establish the effects of cerium

precursor and preparation method on the structure of the CeOz/y-A1203 catalysts.

2.2 Experimental

Catalyst Preparation. Catalysts were prepared by impregnation of y-alumina
(American Cyanamid, y-A1203, surface area = ZOOmZ/g, pore = 0.6 mL/g) using various
cerium precursors. The alumina support was finely ground (<230 mesh) and calcined in
air at 500°C for 24 h prior to impregnation. Catalysts derived from cerium nitrate used
water solution of Ce4+-ammonium nitrate precursor (Mallinckrodt Co.) and Ce3+-nitrate
99.9% hexahydrate (Aldrich Co.). Catalysts obtained from cerium alkoxide used Ce“-
methoxyethoxide 18-20% in methoxyethanol (Gelest Co.) as precursor. The cerium

content was varied from yx10'2= 0.0x10'2 to 8.0x10'2, where y is the Ce/Al atomic ratio

23

(or from 0 to 22 wt% CeOz). The catalysts for this study were obtained i) by incipient
wetness impregnation, using Ce3+-nitrate (designated “CeyNR”) and Ce4+-ammonium
nitrate (designated “CeyN”), ii) by wet grafting impregnation using the Ce“-
methoxyethoxide solution. Grafting impregnation was performed under inert atmosphere
and continuous mixing, for 48 h. Two different grafting approaches were used: a)
“simple grafting” at room temperature without any additives (designated “CeyA”) and b)
“hot grafting” when the impregnation was carried out at 80-85°C under reflux (designated
“CeyHG”), in the presence of ethanol. All samples were dried at 120°C for 16 h and
calcined at 500°C and 800°C, respectively. In order to follow the transformations that
take place during the thermal treatment, the Ce8-samples were calcined at 200°C and
300°C and analyzed. The actual cerium concentrations in the catalysts were determined
by inductively coupled plasma (ICP) analysis, the results showing an estimated error of i
10% in comparison with those theoretically calculated and added during preparation.

Standard Materials. Ce02 was prepared by calcining Ce4+-ammonium nitrate
(Aldrich Chemical Company) at 500°C for 16 h. XRD patterns of the standard
compounds matched the appropriate Powder Diffraction File [28].

BET Surface Area. Surface area measurements were performed using a Quanta-
Chrome Quantasorb Jr. Sorption System. Approximately 0.1 grams of catalyst were
outgassed in N2 at 165°C for 12h prior to adsorption measurements. The measurements
were made using relative pressures of N2 to He of 0.05, 0.08, and 0.15 (N2 surface area =

0.162 nmz) at 77 K. The estimated error for the surface area measurement is i 5%.

1

24

Difiitse Reflectance. Diffuse reflectance spectra were acquired using a Mattson-
Galaxy F TIR-3020 instrument with a diffuse reflectance attachment from Spectra-Tech
Inc. The catalyst powders (20—25 mg) were put into the sample holder, introduced into
the DRIFTS attachment. The IR beam is focused using an ellipsoidal mirror from the
attachment, diffuses into the powder, where it undergoes reflection, refraction, diffusion
and absorption before reemerging at the surface. After the beam is collected from the
sample using the same ellipsoidal mirrors, is focused onto the IR detector and analyzed.
The spectra collected are in Kubelka-Munk units versus wavenumber [29]. IR spectra
were acquired with resolution of 2.0 cm'1 over a 400-4000 cm'1 wavenumber range.

X-Ray Dzfiraction. X-ray powder diffraction patterns were obtained with a

Rigaku XRD diffractometer employing Cu Ka radiation (7» = 1.541838 A). The X-ray
was operated at 45 kV and 100 mA. Diffraction patterns were obtained using a scan rate
of 0.5 deg/min with slits = 1mm. Powdered samples were mounted on glass slides by

pressing the powder into an indentation on one side of the slide.

The mean crystallite size (d) of the CeOz particles was determined from XRD
line broadening measurements using the Scherrer equation [30]:

d = kit/Boost) (1)
where it is the X-ray wavelength, k is the particle shape factor, taken as 0.9 for cubic
particles, and [3 is the full width at half maximum (fwhm), in radians, of the Ce02 <111>
line and 0 is the diffraction angle. Semi-quantitative X-ray diffraction data were obtained
by comparing Ce02<111>/A1203<440> intensity ratios measured for catalyst samples

with peak ratios measured for physical mixtures of Ce02 and y-A1203. A calibration

25

curve was made using a physical mixture of Ce02 and y-A1203 with different
concentration (gCeOZ/g mm). This method assumes that cerium addition does not disrupt
the y-A1203 spinel and consequently does not affect the intensity of the A1203 <440> line.
The error in this method was estimated to be approximately i 20%.

XPS Analysis. XPS data were obtained using a Perkin-Elmer Surface Science
instrument equipped with a magnesium anode (1253.6 eV) operated at 300W (15 kV, 20
mA) and a 10-360 hemispherical analyzer operated with a pass energy of 50 eV. Spectra
were collected using a PC137 board interfaced to a Zeos 386SX computer. The
instrument typically operates at pressures below 1 x 10‘8 Torr in the analysis chamber.
Samples were analyzed as powders dusted onto double-sided sticky tape. Binding
energies for the catalyst samples were referenced to the A12p peak (74.5 eV). XPS
binding energies were measured with a precision of i 0.2 eV, or better.

Binding energies for the standard samples were referenced to the C15 line (284.6
eV) of the carbon overlayer. Since charging was observed for the standard Ce3+ and Ce4+
samples studied, the C1, charge reference may be less accurate that a Ce3d derived charge
correction. Thus, charge correction was done using the Ce3d u’” peak (916.7 eV) from
cerium oxide sample. Using the Ce3d u’” reference energy, the binding energy of the Cl,
peak measured for the CeOz standard was 284.6 eV. The Ce3d u’” peak has been
commonly used to compensate for sample charging in ceria photoreduction studies [31].
Pure CeOz (Aldrich) was used for the Ce“ XPS spectrum and CeAlO3 (just prepared) for
Ce3+ XPS spectrum were used as standard compounds. CeAlO3 was prepared by

reduction of cerium oxide at 900°C under H2 for 16 hours.

26

2.3 Results and Discussion

Precursor-Support Interaction. During this study, it was observed that Ce8N and
Ce8NR catalysts show a similar behavior to the parameters varied during the preparation
and thermal treatment. As a result, the data obtained from these samples will be treated
and interpreted together. Figure 2.1 presents DRIFTS spectra of y-Ale3, CeOz and
Ce(OH)4. Characteristic for the alumina spectrum (a) is the presence of the 3690 cm'1
band, which is attributed to the free hydroxyl groups present on the surface. The broad
band at around 3500 cm”, indicates that many of the hydroxyl groups are hydrogen
bonded [32]. The other bands observed in the spectrum are due to the residual C02 and
H20 (1638 cm”) adsorbed on the alumina surface, hard to remove by simple nitrogen
purging, and the bands around 1200-100 cm'1 are attributed to Al-O stretch and bend
from alumina structure [32-34].

Due to the sensitivity of the DRIFTS analysis, several well defined bands for
CeOz were observed which are otherwise hard to see using simple FTIR analysis. Figure
2.1. (b) shows the DRIFTS spectrum for neat CeOz. As in the case of the alumina
support (a), the free and coupled hydroxyl groups from the surface, can be easily
distinguished. The bands observed at 1556 cm”, 1473 cm”, 1299 cm", and 1034 cm'1
were attributed in a recent study made by Verduraz et al. [33], to the carbonate and
carboxylate species present on the Ce02 surface. The band at 1556 cm'1 corresponds to a
carboxylate species, 1473 cm'1 and 1299 cm'1 to a bidentate carbonate and 1034 cm'1 to a

monodentate carbonate surface species. The last type of carbonate species can be found

27

1043
1614 .
1762
2010
‘1
\
1556 '2?9
‘ \ 1473
1034
(C) \ /’ '

(b)

 

.. e1»... - 1-- L_- J.— - .- t-
4000.0 3000.0 2000.0 1000.0

Wavenumber [cm '1]

Figure 2.1 DRIFTS spectra of standard compounds: a) y-AIZO3 support; b) CeOz
obtained from pure cerium nitrate calcination; c) Ce(OH)4

28

even after calcination at high temperature (800°C). The Ce(OH)4 spectrum shown in
Figure 2.1. (0), includes a broad band characteristic of the -OH stretch (3500-3000 cm”)
involved in intramolecular hydrogen bonds, more intense bands characteristic of C02 and
H20, bands characteristic of carboxyl or carbonyl surface species and bands characteristic
of deformation of O-H bond (the broad band between 1500 - 1200 cm") [34].

Figures 2.2 and 2.3 show the DRIFTS spectra of the Ce8N and Ce8NR catalysts,
respectively, dried at 120°C (b) and calcined at 500°C (c) for comparison with the
spectrum of pure y-AIZO3 (a). In both cases, the spectra reveal the formation of Ce(OH)x
species on y-AIZO3 after drying the catalyst at 120°C. The process may be favored by the
weak acidity of the alumina surface, which favors the formation of hydrated particles
from Ce3+/Ce4+-nitrate solution. Simple water solvation and drying of cerium nitrate
compounds led to identical DRIFTS spectra as the starting compounds. Together with
the formation of the Ce(OH)4 on the support, the alumina band at 3690 cm"]
corresponding to the free hydroxyl groups disappears and a strong increase in the
intensity of the broad band around 3500-3000 cm'I is observed, corresponding to
hydrogen bonded hydroxyl groups. This fact suggests that the interaction between the
cerium precursor and alumina support has only a physical nature, basically dipole-dipole
interactions and hydrogen bonds. After calcination at 500°C [Figures 2.2 (c), and 2.3
(c)], the IR spectra of the catalysts became similar with the spectrum of the pure y-A1203

support. The IR band at 3690 cm'1 corresponding to free surface hydroxyl groups from

29

 

 

(b)

 

(3)
LL —l-~—- YL_ _4_ ,_ -im ‘___--z -
4000.0 3000.0 2000.0 1000.0

Wavenumber [cm '1]

Figure 2.2 DRIFTS spectra acquired during preparation of the Ce8N catalysts:
a) pure y-A1203; b) dried at 120°C; c) calcined at 500°C

30

 

 

(b)

 

L 11L ,LtL

4000.0 3000.0 2000.0 1000.0

Wavenumber [cm ‘1]

Figure 2.3 DRIFTS spectra acquired during preparation of the Ce8NR catalysts:
a) pure y-A1203; b) dried at 120°C; c) calcined at 500°C

31

alumina is regenerated in both cases, by removing the hydroxyl groups from the cerium
species as CeOZ is forming during calcination. Some remaining bands from bending -OH
groups are still observable at 1520 cm'1 and 1100 cm".

Figure 2.4, shows the DRIFTS spectra for the Ce8A catalysts acquired during
impregnation and thermal treatment. As was mentioned before, the grafting process
involves a chemical interaction between the support and precursor. The interaction of the
cerium alkoxide with the support may take place either through a reaction with the
hydroxyl groups involved in intramolecular hydrogen bonds or reaction with the free
hydroxyl groups from the surface leading to the formation of Al-O-Ce-Alkoxide bonds
[35]. If the Ce-OH phase is formed in the solution, the precursor-support interaction
could be an intermolecular interaction through hydrogen bonds. During “simple grafting”
DRIF TS spectra from Figure 2.4 (c) show the disappearance of the absorption band at
3690 cm'I characteristic of the free hydroxyl groups from the y-AIZO3 surface. The
presence of the absorption bands at 2837 cm’1 characteristic of hydrocarbon C-H
stretching, are observed for the pure cerium alkoxide precursor (not shown), and even for
catalysts calcined at 200°C when all the solvent is removed from the catalyst supports.
This observation supports the idea of a covalent bond interaction between the cerium
alkoxide and alumina surface. In addition, no Ce(OH)4 phase formation was observed
during “simple grafting “ as DRIFTS spectra of the dried catalyst indicates. After
calcination, the DRIFTS spectra of the catalyst indicate the regeneration of the band

correSponding to the free hydroxyl group from the support surface and the band at 1523

32

 

“is“ —~-1 —~1eee 1 e- - 1 1—+ + «+1
4000.0 3000.0 3200.0 2300.0 2000.0 1000.0 1200.0 800.0

Wavenumber [cm'l]

Figure 2.4 DRIFTS spectra acquired after thermal treatment of the Ce8A catalysts:
a) pure y-AIZO3; b) fresh; 0) dried at 120°C; (1) dried at 200°C;
e) calcined at 300°C; f) calcined at 500° C

33

 

(a)

14ss_+--r1s -+ +41 we.i-w-s1s.-mt.
4000.0 3600.0 3200.0 2800.0 2000.0 1600.0 1200.0 800.0

Wavenumber [cm-l]

Figure 2.5 DRIF TS spectra acquired after thermal treatment of the Ce8HG catalysts:
a) pure y-A1203; b) fresh; c) dried at 120°C; d) dried at 200°C; e) calcined
at 300°C; f) calcined at 500°C

34

cm'1 which can be assigned to a surface carboxylate species formed on Ce02 [33].

Figure 2.5 shows DRIFTS spectra of the pure alumina support (a) for comparison
with the spectra corresponding to the Ce8HG catalysts dried and calcined at different
temperature (b-f). It can be observed from the spectrum (b) that the band corresponding
to the free hydroxyl groups from the surface of pure alumina disappeared following
impregnation with the cerium alkoxide solution. The solvent was removed from the
catalyst by distillation under vacuum. Alkoxides exhibit characteristic C-O stretch bands
in the 950-1200 cm'l domain. Carboxylates presumably from oxidation of alkoxide
exhibit a C-O stretching bands at 1580 cm”. Ceriurn-oxygen bands, which appear in the
500-250 cm'1 region, are hard to be observed due to the strong absoption bands of the
support in this domain [36]. Figure 2.5 shows bands characteristic of the alkoxide even
for catalysts calcined at 200° - 300°C (c-d), which may indicate that the cerium alkoxide
species survived during impregnation and was dispersed on the support through a grafting
reaction [see reaction (1)]. The intense band from 1580 cm'1 indicates also the formation
of some metal (Al or Ce)-carboxylate species on the surface, during the “hot grafting”
process. Comparison of the DRIFTS spectra of Ce8A in Figure 2.4 with those for
Ce8HG in Figure 2.5 reveal similaries in absorption bands. This suggests that the
grafting process occurs in the same way in both cases, but that the higher temperature and
reflux conditions of the “hot grafting” may carry out the reaction more efficiently.

Bulk Characterization. Figure 2.6 presents XRD patterns for cerium species

present on the alumina support during thermal treatment. Figure 2.6 a shows peaks that

35

C802
<111>

 

 

(C)

(b)

(a)
L [L L LL l LL .,
20.0 30.0 40.0 50.0

Degrees 2 9

Figure 2.6 XRD patterns for Ce8N catalysts treated at different temperatures:
a) 25°C; b) 120°C; c) 500°C; (1) 800°C

36

correspond to the Ce(NO3)6(NH4)2 species (marked on the spectrum from 1 to 7). A
DRIF TS spectrum of Ce8N catalyst dried at room temperature indicates bands similar
with those observed for cerium ammonium nitrate (not shown). After drying above
125°C and removing the water from the system, the XRD patterns observed in Spectra
2.6 b-d, correspond mostly to the cerium oxide crystalline phase. DRIFTS spectra are
consistent with this observation; the peaks observed by probing the catalysts surface
correspond to the cerium hydroxide phase (see Figures 2.1 c, 2.2 b,c, and 2.3 b,c). As the
temperature of the thermal treatment increases, the Ce02 <111> peak becomes more
intense and narrower. This fact is attributed to the growth of the CeOz crystalline phase
with temperature. Figure 2.7 shows the XRD patterns of the Ce8A catalysts collected at
different treatment temperatures. XRD patterns for Ce8A catalysts just dried at room
temperature (25°C) after impregnation (a) and dried at 125°C in the oven, where only the
XRD pattern corresponding to poorly crystalline CeOz species (very broad and less
intense peaks) can be observed. Only after calcination at 500°C or higher can a broad
XRD pattern for CeOz <111>, corresponding to the crystalline Ce02 phase [Figure 2.7 (c-
d)] be clearly observed. So, during impregnation and thermal treatment of the Ce8A
catalyst, less or smaller crystalline Ce02 particles are formed than in the case of Ce8N
catalyst. In Figure 2.8 are presented the XRD 'pattems for y-A1203 support (a) for
comparison with the Ce8-catalysts prepared by using various precursors (b-e) after
calcination at 500°C in air. Table 2.1 presents the Ce02 particle size evaluated from the
XRD pattern presented in Figure 2.8, using the Scherrer equation (1). The XRD peak at

20 = 28.5°, corresponding to the main CeOz <111> peak, is utilized for our calculations.

37

 

 

 

 

CeO 2
<440>
((1)
C60 2
<200>
C602
<111>
(C)
(b)
(a)
4 m” L 1— - A. - , —+-—
20.0 30.0 40.0 50.0
Degrees 2 0

Figure 2.7 XRD patterns for Ce8A catalysts treated at different temperatures:
a) 25°C; b) 120°C; c) 500°C; d) 800°C

38

()

(d)

 

(C) C602
<11 1>
CeO2
<200>
(b)
A1203 A] O
A1203 <31 l> 2 3
<220> <222>
(a)
i — rel-w *“W—~— f - * -- —+—1
25.0 30.0 35.0 40.0
Degrees 20

Figure 2.8 XRD patterns of the Ce8-series of catalysts calcined at 500°C:

a) pure y-A1203; b) Ce8N; c) Ce8NR; d) Ce8A; e) Ce8HG

39

Table 2.1 Semiquantitative XRD data and CeOz particle size determined from line
broadening calculations for the Ce8-catalysts

Ce8A Ce8HG Ce8NR Ce8N

- after calcination at 500°C
d [nm] 4.1 3.6 7.0 8.0
Ce02 % 38 88 100 100
crystalline

- after calcination at 800°C
(1 [nm] 13.9 11.6 9.7 11.7
CeOz % 68 100 95 98.5
crystalline

The particle size for Ce8N and CeSNR catalysts are comparable (7-8 nm). The
percentage of the Ce02 crystalline phase present in the catalysts after calcination at
800°C, are 100% from” the amount of cerium species added to the system during
preparation, for both catalysts. This mean that almost all the cerium precursor used for
impregnation was transformed in large crystallites. Among the four catalysts in Figure
2.8, the C602 <111> peak is the lowest in intensity for Ce8A. The crystalline phase
formed from the total amount of cerium used for impregnation reaches a value of only
38%. The rest of the cerium is present as amorphous Cer or very small C602
crystallites (< 20A). XRD analysis indicates that the Ce8HG catalyst calcined at 500°C
has a much higher proportion of Ce02 crystallites (88%) compared with the Ce8A
catalyst obtained by the “simple grafting” technique (38%). On the other hand, the
particle size for the Ce8A catalyst calculated from .XRD line broadening (4.1 nm) is
slightly larger than that for the Ce8HG catalyst (3.6 nm). The differences observed in

Ce02 particle size of the Ce8-catalysts can be attributed to the impregnation technique.

40

Based on previous data, transition metal alkoxide precursor led to a well dispersed oxide
on the support following calcination at temperature above 500°C. In the case of cerium
alkoxide, either “simple grafiing” or “hot grafting” method led to the formation of Ce02
crystallites larger than 3 nm, following calcination (far from a monolayer dispersion).
This fact can be attributed to the polymerization of cerium alkoxide during the
impregnation, which favor the formation of large Ce02 particles, after calcination at
500°C and 800°C. The small difference in CeOz particle size observed between Ce8A
and Ce8HG can be explained considering two opposite effects during catalyst
preparation: i) the reflux and mixing conditions used in the case of “hot grafting”
impregnation favored a better cerium dispersion; ii) the higher temperature during
impregnation determined a higher percentage of Ce02 crystalline phase for Ce8HG in
comparison with Ce8A catalyst. However, the CeOz particle size is much lower than in
the case of the Ce8N and Ce8NR catalysts, which corresponds to a better cerium
dispersion. The small difference in C602 particle size observed between Ce8A and
Ce8HG can be explained by considering the reflux and mixing conditions used in the case
of “hot grafting” impregnation which give the better dispersion. The higher temperature
during impregnation apparently determines a greater crystalline percentage of Ce02 phase
for Ce8HG compared with Ce8A catalyst.

Figure 2.9 shows XRD patterns for the Ce8A and Ce8HG calcined at 500°C in
comparison with those obtained after calcination at 800°C. For Ce8A catalysts (Figures

2.9 a and 2.9 b), calcination at higher temperature led to formation of larger CeOz

41

C802
<111>

 

 

 

 

C302
<200>
(d)
(C)
(b)
(a)
h _ .L l--—— -1 w—L—uk
25.0 30.0 35.0 40.0

Degrees 2 0

Figure 2.9 Comparison of XRD patterns of the Ce8HG and Ce8A catalysts calcined
at different temperature: a) Ce8A calcined at 500°C; b) Ce8A calcined at
800°C; c) Ce8HG calcined at 500°C; (1) Ce8HG calcined at 800°C

42

particles (13.9 nm) and a sharp increase in the crystalline Ce02 phase (68%), as
semiquantitative XRD analysis revealed. Figures 2.9 c and 2.9 d show the modification
in the XRD pattern corresponding to CeOz <111> for the Ce8HG catalyst, calcined at
500°C and 800°C. Sample calcination at 800°C increases the particle size from 4.1 to
11.6 nm but does not significantly increase the crystallinity of the cerium phase (100%).
This fact corresponds to a decrease in the cerium dispersion and suggests that the small
crystallites are mobile enogh to get together and form large crystalline particles.

Cerium loading eflect. Table 2.2 shows the variation in the BET surface areas of
the Ce/A1203 catalysts as a fiinction of Ce content. As can be observed, there are no real
differences in the surface area of the catalysts following cerium impregnation. Since
CeOz has essentially no porosity, a simple weight correction to the measurement values

show that the surface area does not change significantly.

Table 2.2. The effect of cerium content on surface area and particle size evaluated
from XRD line broadening of CeyAl catalysts calcined at 500°C

Ce/Al Atomic CeyN CeyA CeyHG

Ratio 828* (1 § 11* d Sza“ d
(x102) 1m /g1 [mm] [m an [um] 1m /g1 [mm]

0 200 - 200 - 200 —

1.0 194 4.1 188 a 191 a
2.0 195 4.9 175 2.8b 190 3.0
4.0 200 5.9 183 3.7 190 3.5
6.0 201 5.6 191 4.0 193 3.9
8.0 210 7.0 203 4.1 192 3.6

a - no Ce02 XRD peaks detected; b - the value must be considered approximate due to
the low intensity of the Ce02 peak; * based on weight of alumina in the catalyst

(e)

 

J ,,, ,, .l 7 W I 771 .
30.0 35.0 40.0

Degrees 20

Figure 2.10 X-ray diffraction patterns measured for CeyN catalysts: a) A1203;
b) CelN; c) Ce2N d) Ce4N; e) Ce6N

44

XRD analysis was used to examine the effect of cerium loading on the Ce02
dispersion on y-A1203 support when different impregnation techniques were used for
catalyst preparation. XRD patterns for the CeyN catalysts are shown in Figure 2.10, for
CeyA catalysts in Figure 2.11, and for CeyHG in Figure 2.12. From these figures, the
characteristic patterns corresponding to the CeOz crystalline phase can be observed
clearly. The Ce02 <111> peak has been used to evaluate the CeOz particle size using line
broadening calculations. In each case, the relative intensities of the CeOz features and the
particle sizes have increased with increasing Ce content. Table 2.2 shows the calculated
values for CeOz particle size in the case of the CeyN, CeyA, and CeyHG.

In the case of CeyN catalysts (Figure 2.10), Ce02 particle sizes increased with
increasing cerium loading from 4.1 nm in the case of the CelN catalysts to 7.0 nm for the
Ce8N catalyst. The diffraction pattern of the CelA catalyst (Figure 2.11) showed only
lines characteristic for the y-A1203 carrier. For CeyA catalysts with Ce/Al atomic ratios 2
0.02, peaks characteristic of Ce02 are observed. The relative intensities of the Ce02 lines
increase with Ce loading. The CeOz particle size calculated for the CeyA catalysts with y
S 4.0 are very close to each other (3.7 nm compared with 4.1 nm), but significantly
smaller than the values obtained for CeyN catalysts of the same composition. A similar
trend is observed for the CeyHG catalysts as is seen in Figure 2.12. For the CelHG
catalyst, no pattern specific to Ce02 crystalline phase is observed, but as the cerium
loading is increased, y 2 2.0, the line characteristic of Ce02 <111> is well defined. The
relatively small particle size for CeyHG catalysts may indicate a better dispersion of the

Ce02 crystalline phase or amorphous phase formation.

45

 

<220>

 

25.0 30.0 35.0 40.0

Degrees 29

Figure 2.11 X-ray diffraction patterns measured for CeyA: a) A1203; b) CelA;
c) Ce2A; d) Ce4A; e) Ce6A

46

(e)
CeOZ
<1 I I) Ce02
<200>
(d)
(C)

 

 

1. * + -|- + ~+~ -—

25.0 30.0 35.0 40.0
Degrees 29

Figure 2.12 X-ray diffraction patterns measured for CeyHG catalysts: a) A1203;
b) CelHG; c) Ce2HG;d) Ce4HG; e) Ce6HG

47

Table 2.3 Concentration of crystalline Ce02 in CeyN, CeyA and CeyHG catalysts
calculated from quantitative XRDa data

Ce/Al Atomic Weight % Weight % Ce02 crystalline
Ratio CeOz CeOz in Catalysts from the total cerium added
(x10'2) CeyN CeyA CeyHG
1.0 3.4 74 b b
2.0 6.8 93 3 43
4.0 13.5 100 38 64
6.0 20.3 100 42 67
8.0 26.9 100 38 88

a - valid for CeOz particle size > 3.0 nm.

b - no C602 XRD peaks detected for this catalyst.

Table 2.3 shows the concentration of crystalline Ce02 phase in the CeyN , CeyA
and CeyHG catalysts determined from semiquantitative XRD analysis. In all cases, the
amount of crystalline CeOZ increases with increasing Ce loading.

For CeyN catalyst, the amount of CeOz crystalline phase is close to the weight
percent of CeOz used for catalyst preparation (Table 2.3). This implies that all the cerium
nitrate used for the catalyst impregnation was transformed into crystalline cerium oxide
phase. The formation of crystalline CeOz phase was attributed to the dehydration of the
Ce(OH)4 species formed on the alumina surface afier impregnation with the aqueous
nitrate solution [27, 37]. These led after calcination at 500°C to large CeOz crystallites
due to dehydration of the Ce(OH)4 formed after impregnation and drying. The
differences between the experimental and initial values for cerium oxide concentration
present in the catalysts (13.5% versus 14.7% in the case of the Ce4N catalyst) are in the
range of the 20% error considered acceptable for the semiquantitative XRD calculations.

The amount of crystalline CeOz present in the CeyA series (Table 2.3) increases with

48

increasing Ce content; however, the concentrations are significantly lower than the
amount of cerium oxide present in the catalysts. This fact can be attributed to the
formation of cerium amorphous phase or small crystallites (< 2 nm). Table 2.2 shows the
variation in the Ce02 particle size as a function of Ce content for the CeyA catalysts
determined from XRD line broadening calculations.

A similar analysis of the CeyHG catalysts shows that the amounts of Ce02
crystalline phase in these catalysts are lower than the amounts of cerium used for catalyst
preparation but higher than for the CeyA series. As in the case of the CeyA catalysts, this
finding can be attributed either to an amorphous phase or to small CeOz crystallites. If
small Ce02 crystallites are formed on the alumina support, it would suggest that a well
dispersed cerium phase is formed on the alumina surface. However, previous data
presented in the first part of this work has indicated that cerium alkoxide solution
agglomerates on the alumina surface during impregnation due to polymerization of the
alkoxide, which competes with the grafting reaction. As data for CeyA and CeyHG series
have shown, grafting impregnation led to formation of more amorphous cerium oxide
The size of crystalline CeOz range from 3.0 to 3.9 nm, close to the detection limit of the
XRD analysis. Incepient wetness impregnation using cerium nitrate (CeyN series) led to
the formation of large CeOz particles with a high degree of crystallinity.

As a conclusion, the chemical instability, low concentration of the cerium
alkoxide precursor used for impregnation, and the incomplete grafiing reaction of the
alkoxide with the support which favored cerium alkoxide polymerization led after

calcination at 500C mostly to amorphous phase and large CeOz.

49

Cerium Oxidation State. Figure 2.13 show the Ce XPS spectra of the standard
Ce3 +A103 (a) and Ce4+02 (b) compounds. The XPS Ce3d spectra of standard compounds
are complicated due to hybridization of the Ce 4f with ligand orbitals and fractional
occupancy of the valence 4f orbital [38-40]. The Ce3d spectrum obtained for CeAlO3
[Figure 2.13 (a)] is relatively a simple one, containing two major 3d5,2 peaks at 883.1 eV
(v) and 886.5 eV (v’) and the main Ce3d3,2 features which appear at 900.1 eV (u) and
905.5 eV (u’). These peaks have been assigned to cerium with mixed 4fl and 4f2 orbital
configurations. The Ce3d spectrum measured for Ce02 [Figure 2.13. (b)] contains three
main 3d5,2 features at 883.2 eV (v), 889.2 eV (v”), and 899.4 eV (v”’). The three main
Ce3d3,2 features appear at 901.1 eV (u), 907.7 eV (u”), and 917.3 eV (u”’). The high

’9’

binding energy doublet v”’(u ) has been assigned to cerium with 4f) orbital
configuration. The v” and v (u” and u) doublets are assigned to cerium with mixed 4fi
and 4f2 orbital configurations. These states appear due to the core hole potential in the
final state and the 4f hybridization in the initial state [39-44]. As can be observed, cerium
compounds which contain only Ce3+ do not have u”’(v”’) peaks characteristic of the 4f)
orbital configuration. This observation implies that the u’” peak can be used to indicate
the presence of Ce4+ on the catalysts. However, the relative intensity of the u’” feature
cannot be used to determine the amount of Ce4+ present in the catalysts unless the

dispersions of the Ce4+ and Ce3+-phases are known and it is assured that cerium

photoreduction does not occur [41, 42].

50

 

 

V
u
u"
V
v"
VI
u! ‘ (b)
(a)
up - .e l e. 1
920.0 900.0 880.0
Binding Energy [eV]

Figure 2.13 Ce X128 spectra for standard cerium compounds: 3) Ce3+AIO3;
b) Ce *0,

51

 

920.0 900.0 880.0
Binding Energy (eV)

Figure 2.14 Ce XPS photoreduction spectra for Ce8A catalyst, calcined at 800°C after:
a) 5 min.; b) 15 min.; c) 30 min.; (1) 1 h; e) 4 h, of scanning

52

Figure 2.14 shows the Ce XPS spectra for Ce8A catalyst, calcined at 800°C,
collected at different X-ray exposure times. For short scanning time (Figure 2.14 a) the
surface cerium phase is present as Ce“. As the scanning time increases [Figure 2.14 (b-
e)], the v’-peak intensity increases corresponding to the appearance of surface Ce3+
species. Previous studies have indicated that Ce02 undergoes photoreduction during XPS
analysis in high vacuum due to intense heating of the sample surface, the presence of free
electrons into the chamber and the crystallinity of the exposed particle [43-46]. Park and
Ledford [47] have shown that the photoreduction of cerium is also a function of Ce02
crystallinity. The amorphous phase is reduced more than the crystalline one. The results
in Figure 2.14 indicate that cerium supported on y-A1203 is photoreduced even in the case
of the catalyst calcined at 800°C. The photoreduction and the relatively low cerium
dispersion found for the catalyst in this study preclude XPS quantitative analysis of
Ce3+/Ce4+ supported species. For Ce8N, Ce8NR and Ce8HG catalysts, XPS spectra
indicate the presence of cerium as Ce4+ species, being present mostly as CeOz crystalline
phase. In this case, X-ray photoreduction during XPS analysis is significantly decreased
due to the higher crystallinity of the cerium species.

2.4. Conclusions

On the basis of the results obtained from bulk and surface techniques used to
characterize the various cerium modified y-A1203 catalysts obtained by using different

3+I4+
precursors, we can conclude that: 1) Ce

-nitrate precursors, impregnated via the
incipient wetness method, led to large CeOz crystallites; 2) the grafting technique, using a

cerium alkoxide precursor, is an alternative preparation method for y-A1203 impregnation;

53

the grafting process takes place through a reaction between surface hydroxyl groups (free
or hydrogen bonded) from the support and cerium precursor; during this process no
evidence of Ce(OH)4 formation is observed; 3) Simple grafting led to only a slight
improvement in cerium dispersion compared with the traditional incipient wetness
approach using cerium nitrate; specific to the grafting method is the formation of Ce“-
amorphous phase on the support surface after calcination at 500°C; 4) XRD data show
variations in crystallinity and cerium dispersion on alumina surface, when different
precursors and Ce-loadings were used for catalyst preparation; 5) XPS data indicates that

cerium is present on the catalyst surface as Ce“ unless it is photoreduced during analysis.

2.5 References

1. Ledford, J. S.; Houalla, M.; Proctor, A.; Hercules, D. M. J. Phys. Chem. 1989, 93,
6770-6777.

2. Barrault, J .; Guilleminot, A.; Achard, J. C.; Paul-Boncour, V.; Percheron-Guegan,
A.; Appl. Catal. 1986, 21, 307-312.

3. Fleisch, T. H.; Hicks, R. F .; Bell, A. T. J. Catal. 1984, 87, 398-413.

4. Reick, J. S.; Bell, A. T. J. Catal. 1985, 96, 88-105

5. Diwel, A. F.; Rajaram, R. R.; Truex, T. J. Catalysis and Automotive Pollution
Control 11 Edit by Crucq, A. 1991, Elsevier, Amsterdam, 139-152.

6. Pembo, J. M.; Lenzi, M.; Lenzi, J.; Lebugle, A. Surf Inter. Anal. 1990, 15(11),
663-668.

7. Yu-Yao, Y.F.; Kummer, J .T. J. Catal. 1987, 106, 307-312.

8. Chojnacki, T.; Krause, K.; Schmidt, L.D. J. Catal. 1991, 128(1), 161-165.

9. Schaper, H.; Doesburg, E. B. M.; VanReijen, L. L. Appl. Catal. 1983, 7, 211-220.

10. Oudet, F.; Courtine, P.; Vejux, A. J. J. Catal. 1988, 114, 112-118.

11. Yao, H. C.; Yu-Yao, Y. F. J. Catal. 1984, 86, 254-265.

12. Yang, J. K.; Swartz, W. E. Spectrosc. Lett. 1984, 17(6& 7), 331-337.

13. Gaugin, R.;Grau1ier, M.; Papee, D. Adv. Chem. Ser. 1975, 143, 147-149.

14. Koberstein, E. Unst. St. Proc. Catal., Matros, Y.S., Ed., 1990, 643-648.

15. Yamaguchi, T.; Ikeda, N.; Hattony, H.; Tanabe, K. J. Catal. 1981, 6 7, 324—330.

16. Raen, S.; Braten, N. A.; Grepstad, J. K.; Qui, S. L. Phys. Scr. 1990, 41 (6),
1001-1002.

17. Nunan, J.; G.; Robota, H. J.; Cohn, M. J.; Bradley, S. A. in Catalysis and

18.

19.

20.

21.

22.
23.

24.

25.

26.

27.

28.
29.

30.
31.
32.
33.
34.
35.
36.
37.
38.

39.
40.

41.
42.

54

Automotive Pollution Control 11, ed Crucq, A. Elsevier, Amstredam, 1991,
21-230.

Rodriguez, R. I.; Guerrero, R. A.; Fierro, J. L. G. Appl. Surf Sci. 1989, 40(3),
239-241.

Silver, R. G.; Summers, J. C.; Williamson, W. B. in Catalysis and Automotive
Pollution Control 11, ed. Crucq, A. Elsevier, Amsterdam, 1991, 167-180.
LeNormand, F .; Hilaire, L.; Kili, K.; Krill, G.; Maire, G. J. Phys. Chem. 1988, 92,
2561-2568.

Shyu, J. Z.; Otto, K.; Watkins, W. L. H.; Graham, G. W.; Belitz, R. K.; Gandhi,
H. S. J. Catal., 1988, 114, 23-33.

Miki, T.; Ogawa, T.; Ueno, A.; Matsuura, S.; Sato, M. Chem. Lett. 1988, 565-568.
Haneda, M.; Mizushima, T.; Kakuta, N.; Ueno, A.; Sato, Y.; Matsuura, S.;
Kasahara, K.; Sato, M. Bull. Chem. Soc. Jpn. 1993, 66, 1279-1288.

Mizukami, F .; Maeda, K.; Watanabe, M.; Masuda, K.; Sano, T.; Kuno, K. in
Catalysis and Automotive Pollution Control 11, Edit by Crucq, A. 1991, Elsevier,
Amsterdam, 557-568.

Leclercq, G.; Danthy, C.; Mabilon, G.; Leclercq, L. in Catalysis and Automotive
Pollution Control 11, Edit by Crucq, A. 1991, Elsevier, Amsterdam, 181-194.
Weibel, M.; Garin, F .; Bernhardt, P.; Maire, G.; Prigent, M. in Catalysis and

Automotive Pollution Control 11, Edit by Crucq, A. 1991, Elsevier, Amsterdam,
95-205.

Munuera, G.; Fernandez, A.; Gonzalez-Elipe, A. R. in Catalysis and Automotive
Pollution Control 11, Edit by Crucq, A. 1991, Elsevier, Amsterdam, 207-219.
Powder Dififaction File, Inorganic Phases, JCPDS, 1983.

Bell, A. T. in Vibrational Spectroscopy of Molecules on Surfaces Ed. Yates, J. T.
1987, New York, 105-134.

Klug, H.P.; Alexander, L.E. in X-ray Diflraction Procedures for Polycrystalline
and Amorphous Materials, l-st Ed. Wiley, New York, 1954.

Paparazzo, E.; Ingo, G. M.; Zacchetti, N. J. Vac. Sci. Technol. A, 1991, 9, 1416.
Peri, J. B. J. Phys. Chem. 1965, 69(1), 211-219.

Verduraz, F. B.; Bensalem, A. J. Chem. Soc. Faraday Trans. 1994, 90(4),
653-657.

Nakamoto, K. in Infiared and Raman Spectra of Inorganic and Coordination
Compounds, 4-th Ed. Wiley, New York, 1986, 231, 384.

Kifenski, J .; Baiker, A.; Glinski, M.; Dollenmeier, P.; Wokaun, A. J. Catal. 1986,
101, 1-11.

Fujimori, A. Phys. Rev. B, 1983, 28(4), 2281-2283.

Dulamita, N.; Pop, A.; Corbeanu, V. Studio, 1984, 29, 3-11.

Wuilloud, E.; Delley, B.; Schneider, W. D.; Baer, Y. Phys. Rev. Lett. 1984,
202-207.

Fujimori, A. Phys. Rev. Lett. 1984, 53, 2518.

Wuilloud, E.; Delley, B.; Schnieder, W. D.; Baer, Y. Phys. Rev. Lett. 1984, 53,
2519.

Jo, J.; Kotani, A. Phys. Scrp. 1987, 35, 570-575.

Praline, G.; Koel, B. E.; Hance, R. L.; Lee, H. 1.; White, J. M. J. Electron Spec.

43.
44.
45.

46.
47.

55

Relat. Phenom. 1980, 21 , 1746.

El Fallah, J .; Hilaire, L.; Romeo, M.; Le Normand, F.; J. Electron Spectrosc. Rel.
Phenom. 1995, 73, 89.

Daucher, A.; Hilaire, L.; Le Normand, F.; Muller, W.; Maire, G.; Vasquez, A.;
Surf Interface Anal. 1990, 16, 341.

Paparazzo, E. Surf Sci. 1990, 234, L253.

Paparazzo, E.; Ingo, G. M.; Zacchetti, N. J. J. Vac. Sci. Technol. A, 1991, 9, 1416.
Park, W. P.; Ledford, L. S. Langmuir, 1996, 12, 1794-1799.

CHAPTER 3

THE EFFECT OF DIOL COMPLEXATION AND CALCINATION ATMOSPHERE
ON THE STRUCTURE OF CEOz/y—ALZO3 CATALYSTS

3.1. Introduction

Anchoring cerium alkoxide species in an organic medium followed by calcination
has recently proved to be a reliable preparation technique for supported Ce02 catalysts
[1]. In our previous study of cerium promoted catalysts (chapter 2), we have used simple
and hot grafting techniques to prepare a high loadings of small CeOz particles supported
on a y-AIZO3 catalyst. The method is based on the chemical interaction between a cerium
alkoxide compound and the hydroxyl groups from the support surface [2]. Grafting
impregnation led, after calcination, to poorly dispersed Ce02 on alumina. This was
attributed partly to the agglomeration of the cerium alkoxide molecules on the support
surface through a polymerization process, during impregnation, drying, and calcination.
A significant amount of the cerium phase deposited on the alumina surface was found to
be in an amorphous state. Increasing the calcination temperature to 800°C enlarged the
CeOz crystallites deposited on the y-AIZO3 support. Other previous studies have
emphasized the strong correlation between cerium surface structure and its promoter
and/or catalytic properties [3-10]. This work is a part of our study which investigates
new synthetic methods for preparation of catalysts.

In solution, cerium can form complexes with one, two, three, or four bidentate
organic ligands. Different discrete structures are obtained by using variable cerium-

ligand molar ratios. Sanchez et al. [11] have studied the sol-gel chemistry of cerium

56

57

isopropoxide using 2,4-pentanedione (acac) as a ligand for the gelation process. Using
EXAFS analysis combined with F TIR and Raman spectroscopy they found that cerium
isopropoxide is present in the gel as a monomer or dimer alkoxide. Cerium as a metal
center can coordinate one, two or three (acac) molecules forming complexes with well
defined structure. Based on this idea, it should be possible to make a new complex
alkoxide precursor, stable enough to avoid polymerization but reactive to be used for
grafting impregnation of a high surface area support, like A1203, TiOz or SiOz. Diols are
organic compounds with two hydroxyl groups in the molecules. These hydroxyl groups
can be geminal, or bonded to the same carbon atom (very unstable); vicinal or bonded to
two neighboring carbon atoms; or attached to carbon atoms farther apart from each other
[12]. Diols are generally viscous liquids which can function as moderate bidentate
ligands in complexation reactions due to the nonbonded electrons on the oxygen atoms in
their hydroxyl groups [13]. Livage et al. [14] explained the chemical modification of
metal alkoxides [(Ti-OR)4] with “acac” in a similar manner, using FTIR and lH-NMR,
respectively. The structural differences are related to the ability of the alkoxide groups to
behave as exchange ligands. Another factor which might be considered is the humidity
during the preparation of the diol modified cerium alkoxide. It is well known that
atmospheric water vapor plays an important role in the gelation process. To avoid
gelation before and during impregnation, a dry atmosphere is needed. Extreme
conditions like high vacuum or high temperature easily decompose the complexes.

The impregnation process in this case, is similar with that described previously

(see chapter 2). The interaction between the precursor and support may imply a reaction

58

of the cerium alkoxide with the surface hydroxyl groups from the support. This chapter
explores the effects of two variables: i) 2,4-pentarrediol complexation in the precursor
solution used for grafting impregnation; ii) the calcination atmosphere during the thermal
treatments on the surface structure of Ce02 deposited on y-Ale3 support. Fourier
transform infrared spectroscopy (FTIR) and nuclear magnetic resonance (N MR) analysis
are used to characterize the molecular structure of the diol modified precursors. X-ray
diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) provide information
about the bulk and surface structure of the calcined Ce-promoted/AIZO3 catalysts

prepared using the modified precursors.

3.2 Experimental

Catalyst Preparation. Catalysts were prepared by grafting impregnation of y-
alumina (American Cyanamid, y-A1203, surface area = 200m2/g, pore = 0.6 mL/g) using
solutions of modified cerium alkoxide (Cerium Methoxyethoxide solution 18-20% in
methoxyethanol, Gelest Inc.) with a diol (2,4-pentanediol 99%, mixture of isomers,
Aldrich) at various molar ratios, as precursor. Samples where assigned as DquAp for
precursors and DquA for catalysts (where q = diolzcerium alkoxide molar ratio). The
alumina support was finely ground (<230 mesh) and calcined in air at 500°C for 24 h
prior to impregnation. Catalysts derived from pure alkoxide used Ce4+-methoxyethoxide
(18-20% alkoxide in methoxyethanol) obtained from Gelest Inc. The modified cerium
methoxyethoxide precursors were prepared using 2,4-pentanediol obtained from Aldrich

Chemical Company. The cerium content chosen for our catalysts was Ce/Al = 6.0 x 10'2

59

atomic ratio (or 16.5% Ce02 by weight), a value close to the monolayer coverage for a y-
A1203 support. Different diolzcerium alkoxide molar ratios in the precursor solutions
were used to prepare the DquA catalysts, where q = 0, 1, 2, and 3. The catalysts were
prepared by wet impregnation of the y-A1203 support with the alkoxide and the modified
precursor solutions, at room temperature, under inert atmosphere and continuous mixing
for 24 h. After impregnation, the catalysts were dried at 120°C for 12 h and calcined 24
hours in air at 500°C and 800°C, respectively.

FT IR Spectroscopy. An FTIR Mattson Inst. Galaxy-3020 was used for infrared
analysis of the precursor solutions. Thin precursor films were prepared on KBr and
analyzed. IR spectra were acquired with resolution of 2.0 cm'1 on a 400-4000cmil
wavenumber range. The spectra are presented in absorption units.

NMR analysis. 1H-NMR spectra for the modified cerium precursors were
collected with a Varian Gemini 300 MHz instrument and computer interface. D-
Chloroforrn (singlet at 7.23 ppm) at room temperature was used as reference and solvent.
The spectrum of pure 2,4-pentanediol in D-chloroforrn was acquired as a chemical shift
reference in the study of the modified precursors.

X-Ray Difii'action. X-ray powder diffraction patterns were obtained with a

Rigaku XRD diffractometer employing Cu Ka radiation (3. = 1.541838 A). The X-ray
was operated at 45 kV and 100 mA. Diffraction patterns were obtained using a scan rate
of 0.5 deg/min with DS and SS slits = 1. Powdered samples were mounted on glass

slides by pressing the powder into an indentation on one side of the slide. The mean

60

crystallite size (3) of the Ce02 particles was determined from XRD line broadening
measurements using the Scherrer equation [15]:
d=klecosG (1)

where it is the X-ray wavelength, k is the particle shape factor, taken as 0.9, and [3 is the
full width at half maximum (fwhm), in radians, of the Ce02 <111> line and 0 is the
diffraction angle. Semiquantitative X-ray diffraction data were obtained by comparing
Ce02 <111> / A1203 <440> intensity ratios measured for catalyst samples with peak
ratios measured for physical mixtures of Ce02 and y-A1203.

XPS Analysis. XPS data were obtained using a Perkin-Elmer Surface Science
instrument equipped with a magnesium anode (1253.6 eV) operated at 300W (15 kV, 20
mA) and a 10-360 hemispherical analyzer operated with a pass energy of SOeV. Spectra
were collected using a PC137 board interfaced to a Zeos 386SX computer. The
instrument typically operates at pressures below 1 x 10'8 torr in the analysis chamber.
Samples were analyzed as powders dusted onto double-sided sticky tape. Binding
energies for the catalyst samples were referenced to the Al 2p peak (74.5 eV). XPS
binding energies were measured with a precision of i 0.2 eV, or better. As reference

spectra for cerium oxidation state analysis was used data from Chapter 2, Figure 2.13.

3.3 Results and Discussion
Modified precursors. In solution, different discrete structures can be obtained by
using different ceriurn-ligand molar ratios [2]. Previous studies on cerium alkoxide

coordination chemistry used 2,4-pentanedione (acac) at different metal to ligand atomic

61

ration, for metal stabilization by complexation [11,12]. Due to the high reactivity of
cerium methoxyethoxide, addition of acac solution led instantly to a precipitate. The
formation of the precipitate precluded the utilization of acac modified cerium alkoxide
solution for catalyst preparation because grafting impregnation requires the use of a
homogeneous solution as precursor. A moderate modifier, 2,4-pentanediol (diol), was
chosen for this purpose. Adding the diol at different diol:alkoxide molar ratios,
homogeneous cerium solutions were obtained. In time, the solutions changed color and
formed gels. Aging is an important factor in gels preparation. To avoid the gelation
process, the fresh modified solution was used immediately for impregnation. The
strongest gelation effect was observed in the case of the diol modified cerium alkoxide at
a molar ratio of 3:1. Figure 3.1. shows the lN-NMR spectrum of the cerium
methoxyethoxide solution (a) in comparison with the diol modified alkoxide solutions.
From spectrum (a), it can be observed that the resonances corresponding to the -O-CH3 (8
= 3.28 ppm), -CH2- (8 = 3.43 ppm) and -O-CH2- (8 = 3.66 ppm) protons are broad instead
of a narrow singlets for (-CH3) or multiplets (-CH2-) peaks. This was attributed in the
literature [11] to averaging resonances obtained from dynamical chemical exchange of
the methoxyethoxide ligand and different ligand groups (terminal or solvated molecules),
with the methoxyethanol solvent.

Figure 3.1. (b-d) shows the 1H-NMR for the diol modified cerium alkoxide
solutions at different molar ratios [(a)-1:l, (b)-2:l, (c)-3:l]. In these cases, the spectra

show characteristic resonances for the protons corresponding to the methoxyethanol from

62

the cerium solution: a singlet for the -CH3 and a multiplet for the -CH2- and -O-CH2-

groups.

Table 3.1 1H NMR data of meso-2,4-pentanediol for modified cerium alkoxide

precursors
Chemical Shift 8 [ppm]
CH3‘ 'CH2‘ -CH-
Diol 1.116(d, J=6 Hz, 3H) 1.464(m, 2H) 3.949(m, 1H)
1.144(d, J=6 Hz, 3H) 4.045(m, 1H)
DlCeAp 1.096(d, J=6Hz, 3H) 1.443(m,2H) 3.416(m,1H)
1.124(d, J=6Hz, 3H) 3.629(m, 1H)
D2CeAp 1.123(d, J=6Hz, 3H) 1.476(m, 2H) 3.441(m, 1H)
1.159(d, J=6 Hz, 3H) 3.458(m, 1H)
D3CeAp 1.044(d, J=6Hz, 3H) 1.388(m, 2H) 3.368(m, 1H)
1.080(d, J=6Hz, 3H) 3.584(m, 1H)

Table 3.1 shows the chemical shifts observed for the 2,4-pentanediol protons in
the modified solution (DquAp) in comparison with the pure diol solution [16]. Due to
the two asymmetric carbons in positions 2 and 4, of the meso-2,4-pentanediol molecules,
resonances corresponding to the -CH- protons in the molecules appear at two positions, 61
= 3.949 ppm (m) and 52 = 4.045 ppm (m) [17]. Spectra of the diol in the modified
precursors exhibits shifting of the free diol resonances due to interactions with cerium
ions. For instance considering the peaks characteristic to -CH3 group, 5on3 is 1.096 ppm

(1.124 ppm) for the DlCeAp, 1.123 ppm (1.159 ppm) for the D2CeAp and 1.044 ppm

63

 

UR - m «9

 

 

 

 

ULLA - (b)

 

(a)

 

 

 

5 [ppm]

Figure 3.1 1H NMR spectra of the precursor solutions: a) DOCeAp; b) DlCeAp
c) D2CeAp; d) D3CeAp precursors

64

(1.080 ppm) for D3CeAp compared with the 5013 = 1.116 ppm (1.144 ppm) for the pure
2,4-pentanediol solution. These new resonance values for the chemical shifts of the diol
protons can be assigned to the bonded diol ligands that are coordinated to the cerium
atom. The presence of diol in the system may have an effect as moderator, changing the
viscosity of the solution by initiating new chemical processes in the system and
interfering in the dynamic equilibrium established between the ligands and solvent
molecules attached to the cerium atom [11].

Additional information about the complexation and gelation process can be
obtained from FTIR analysis. IR spectra of the precursor solutions (DyCeAp) are
presented in Figure 3.2. Metal alkoxide exhibits characteristic bands for -C-O- vibrations
at ca. 1000 cm'1 and M-0- vibration at 600-300 cm"1 wavenumber [l8]. Cerium
isopropoxides exhibit a strong and characteristic absorption band, corresponding to the -
C-O- vibration from a -C-O-Ce group, at 980 cm'l [19]. Pure cerium methoxyethoxide
IR spectrum shows the analogous band at 910 cm". As the diol is added to the alkoxide,
a new band appears at 937 cm'1 which can be attributed to the new diol -C-O-Ce bend.
Other new bands appeared in the IR spectra of the diol modified solution at 1085 cm",
1330 cm'l and 1710 cm". Absorptions in the region of 1710 cm'1 usually indicate
carboxylate groups in the system. It is possible that during the complexation and gelation
process, carboxylate groups have been formed by ligand decomposition in the cerium
alkoxide. With cerium methoxyethoxide molecules 2,4-pentanediol can form a weak

complex which inhibits self polymerization of the alkoxide, initially present as a

65

1330 1935
\.

\ I 9,37

1710 . l

(d) + 4

(C)

(b)

 

_L P - |-~ _ _|_. -_-
1600.0 1200.0 800.0 400.0

Wavenumber [crrr1 ]

Figure 3.2 FTIR spectra of the precursor solutions: a) DOCeAp; b) DlCeAp;
c) D2CeAp; d) D3CeAp precursors

66

monomer. The diol modified precursor was used for grafting impregnation of the y-
A1203 support. The impregnation process in this case, is similar with that described
previously. The interaction between the precursor and support implies a reaction of the
cerium alkoxide with the surface hydroxyl groups from the support. The complex
formation determined a change in the precursor structure and the viscosity of the solution
used for impregnation modifying the cerium chemical configuration deposited and
anchored on the alumina support

Bulk Characterization. Figure 3.3 (b-e) shows the XRD patterns of the DquA
catalysts calcined at 500°C under air, in comparison with the y-A1203 support (Figure 3.3
a). The presence of CeOz crystalline phase was observed for all catalyst samples. The
particle sizes for the crystallites formed on the catalysts’ surfaces can be evaluated from

line broadening calculations. The results for DquA catalysts are presented in Table 3.2.

Table 3.2 Ce02 particle size and crystallinity of DquA catalysts calcined at 500°C
determined from XRD line broadening calculations

Catalyst Ce02 Crystalline Phase [%] Particle Size [nm]
500°C 800°C 500°C 800°C
DOCeA 52 100 4.0 9.0
D1CeA 64 100 3.2 3.5
D2CeA 64 77 4.1 1 1.9
D3CeA 63 96 4.2 15.9

The Ce02 particle sizes are around 4.0 nm with the exception of the catalyst with

 

- t- _, I , ~+~ ~
20.0 30.0 40.0

Degrees 2 0

Figure 3.3 XRD patterns for the DquA catalysts calcined at 500°C: a) A1203;
b) DOCeA; c) D1CeA; d) D2CeA; e) D3CeA catalysts

68

q = 1, where the particle size is about 3.2 nm. Smaller cerium oxide particles usually
correspond to a better dispersion on the high surface area support. This observation can
be related with a good interaction of the cerium promoters with the support during
grafting impregnation. The C602 crystalline phase present in the DquA catalysts
calcined at 500°C are around 50% of the total amount of cerium loading added to the
catalyst.

The presence of Ce02 increases the thermal resistance of catalysts used in
exothermic catalytic processes [20-22]. To reveal the structural transformation of cerium,
the DquA catalysts were calcined at high temperature and their structural changes were
followed after the thermal treatment. Figure 3.4. (a-d) shows the XRD pattern for the
DquA catalysts calcined at 800°C. The observed XRD pattern from the figure
corresponding to 20 = 28.5°, was assigned to CeOz <111> phase. The peak at 20 = 32.5°
is a sample holder artifact. The intense CeOz <111> pattern suggests that the cerium in
these catalysts is mostly crystalline. The semiquantitative XRD data presented in Table
3.2 indicate a significant amount of amorphous or small crystallites even after calcination
at 800°C, in the case of the D2CeA (77%) in comparison with the other catalysts where
cerium is present only as CeOz crystalline phase. At this moment, there are no clear
explanation of this observation for this particular catalyst. However, it is clear that higher
temperatures favor crystalline phase formation.

Figure 3.5 (a) presents the XRD pattern for y-A1203 support for comparison with

69

 

 

 

 

 

 

 

 

 

v ‘1. l , - - H—L—m — 4————
20.0 30.0 40.0 50.0
Degrees 2 0
Figure 3.4 XRD patterns for the DquA catalysts calcined at 800°C: a) DOCeA;

b) D1CeA; c) D2CeA; d) D3CeA catalysts

70

(C)
Al 2 03
<400>
CeO
2
CeO2 <200>
<1 1 1>
ll
0))
Al 20 3
<31 l>
. L
Al 2 03 I
<220>
(a)
l | I
30.0 40.0 50.0
Degrees 2 0
Figure 3.5 XRD patterns for the DOCeA catalysts calcined under different

atmosphere: a) pure y-AIZO3; b) DOCeA calcined under hydrogen;
c) DOCeA calcined under air

71

the pattern observed for DOCeA catalysts [spectrum (b)] calcined under a reducing
atmosphere (hydrogen flow) and calcined under air [spectrum (c)]. As can be seen in
Figure 3.5 (b), the CeOz <111> peak almost disappeared when the calcination was
performed under a hydrogen flow, at 500°C. This fact can be attributed to the formation
of amorphous cerium phases which are not revealed by XRD analysis. It is well known
in the literature that surface Ce3+-species are easily oxidized to Ce“. However, H2
reduction of supported Ce02 may disrupt the crystal structure such that room temperature
oxidation does not reform the crystal structure. It is unlikely that Ce3+AlO3 is formed at
this temperature, these kind of species being observed only after calcination under
hydrogen at much higher temperature [23].

Surface Characterization. The XPS spectra for the standard Ce3+AlO3 (a) and
Ce4+02 (b) compounds used as reference are presented in Figure 2.13 (Chapter 2). We
have shown that cerium is photoreduced during XPS analysis (Figure 2.14) even in the
case of a sample which was previously calcined at very high temperature. The high
inhomogeneity of the cerium dispersion on the alumina support, combined with the
cerium photoreduction effect observed for amorphous phase precludes quantitative
analysis of Ce3+/Ce4+ using XPS measurements. Park and Ledford [24] have shown that
the amorphous cerium phase is more strongly reduced than crystalline CeOz and the
fraction of cerium photoreduced decreases with an increase in the Ce02 crystalline phase.
Figure 3.6 (a-d) shows XPS spectra for the DquA catalysts calcined at 500°C in air,

acquired after the same scanning time. This implies that all the catalysts were exposed to

72

917.3 905-5 /901'1

 

 

886 5 883.2
\\ /
\
I I
I I
‘ I
I I I I f
I I
I I l
I I
* I I I (d)
I I I I
I I '
I
I ‘ ~ 02)
I
I I
I I ' .
I . I I-
I I I Q
I I
I ‘ I
I I
(a)
4 *"f—v—— #— ———--~I_—--- ‘# — ——~ ——-’+*
920.0 900.0 880.0
Binding Energy [eV]

Figure 3.6 XPS spectra of the DquA catalysts calcined at 500°C: a) DOCeA;
b) D1CeA; c) D2CeA; d) D3CeA catalysts

73

the X-ray for a same period of time. Figure 3.6 (a) shows that the intensity of the Ce v’
(u’) feature characteristic of Ce3+ is dominant in the Ce3d XPS spectrum even if cerium
was initially present as Ce4+-phase. The intensity of the Ce v’ (u’) feature is lower in the
case of the D2CeA and D3CeA. In these cases, the Ce4+-features (u”, u and v”, v)
dominates the Ce3d XPS spectrum. The XPS spectrum observed for the D1CeA catalyst
is a mixed Ce4+-Ce3+-like shape. Our previous studies (Chapter 2) have indicated that
cerium is present as an amorphous phase when the simple grafting method was used for
alumina impregnation. Under these conditions, long X-ray exposures should cause
cerium to appear strongly reduced in the XPS spectrum. The decreased photoreduction
effect observed in the case of the DquA catalysts with q 2 1, obtained by impregnation
with the diol modified precursors, can be related to the amount of Ce02 crystalline phase
formed. As Table 3.2. shows, the Ce02 crystallinity increases in the order: DOCeA <
D1CeA < D2CeA < D3CeA. This analysis has provided some additional information
about the crystallinity of the cerium phase deposited on the alumina support for the

DquA series of catalysts.

3.4 Conclusions

FTIR and lH-NMR analysis were used to analyze the molecular structure of the
modified cerium alkoxide precursors. Cerium alkoxide can form a weak complex with
2,4-pentanediol, which shows, after aging, gelation behavior. The diol modified cerium
alkoxide solution used for grafting impregnation a y-AIZO3 support led to a different

cerium structure on the support than in the case of the unmodified cerium alkoxide

74

precursor. XPS and XRD analysis were used to characterize the cerium oxide on y-
alumina catalysts prepared from modified precursors with different diolzcerium alkoxide
molar ratio solutions. The findings for DquA catalysts indicate that after calcination,
different CeOz structures were obtained when the diol modified cerium alkoxide
precursors was used for alumina impregnation. The diol chelation effect influences the
impregnation process with consequences on the surface configuration of the promoter.
X-ray photoreduction is an important factor in evaluating the oxidation state of cerium on
a catalyst surface; it also offers qualitative information about the presence of amorphous
Cer. Cerium alkoxide impregnated support calcined under an H2 atmosphere leads to

formation of more amorphous Cer phase.

3.5 References

l. Bensalem, A.; Bozon-Verduraz, F .; Delamar, M.; Bugli, G. Appl. Catal. A, 1995,

121, 81-93.

2. Kifenski, J .; Baiker, A.; Glinski, M.; Dollenmeier, P.; Wokaun, A. J. Catal. 1986,
101, 1-11.

3. Schaper, H.; Doesburg, E.B.M.; Van Reijen, L.L. Appl. Catal. 1983, 7, 211-220.

4. Yang, J. K.; Swartz, W. E. Spectrosc. Lett. 1984, 1 7(6, 7), 331-334.

5. Yu-Yao, Y. F.; Kummer, J. T. J. Catal. 1987, 106, 307-312.

6. Oudet, F .; Courtine, P.; Vejux, A. J. J. Catal. 1988, 114, 112-118.

7. Pembo, J. M.; Lenzi, M.; Lenzi, J.; Lebugle, A. Surf Inter. Anal. 1990, 15(11),
663-668.

8. Koberstein, E. Unst. St. Proc. Catal., Matros, Y. S., Ed., 1990, 643-647.

9 Chojnacki, T.; Krause, K.; Schmidt, L.D. J. Catal. 1991, 128(1), 161-165.

10. Ribot, F.; Sanchez, C.; Livage, J. in Chemical Processing of Advanced Materials,
1992, Wiley, NewYork, 267-275.

11. Sanchez, C; Toledano, P.; Ribot, F. Mat. Res. Soc. Symp. Proc. 1990, 180, 47-59.

12. Solomons, G. T. W. in Organic Chemistry - 5-th Edition, 1992, Wiley & Sons
Inc. New York, p125, p694.

13. Kemp, W. in Organic Spectroscopy, Wiley & Sons, New York, 1975.

14. Leaustic, A.; Babonneau, F.; Livage, J. Chem. Mot. 1989, 1(2), 240-247.

15. Klug, H. P.; Alexander, L. E. X-ray Dififaction Procedures for Polycrystalline
and Amorphous Materials, 1-sted.; 1954, Wiley: New York.

16.

17.

18.

19.
20.

21.
22.

23.
24.

75

Itoh, 0.; Ichikawa, Y.; Katano, H.; Ichikawa, K. Bull. Chem. Soc. Jpn. 1976,
49(5), 1353-1358.

Drago, R. S. in Physical Methods for Chemists 2-nd Ed. 1992, Saunders C.P.
Orlando, 211-273.

Nakamoto, K. Infiared and Ramon Spectra of Inorganic and Coordinative
Compounds, 3-rd Ed., 1976, Wiley, Ney York, 230-231.

Brown, L. M. Mazdiyasni, K. S. Inorg. Chem. 1970, 9(12), 2783-2786.
Haneda, M.; Mizushima, T.; Kakuta, N.; Ueno, A.; Sato, Y.; Matsuura, S.;
Kasahara, K.; Sato, M. Bull. Chem. Soc. Jpn. 1993, 66, 1279-1288.

Graham, G. W.; Schmitz, P. J .; Usmen, R. K.; McCabe, R. W. Cat. Lett. 1993,
1 7, 175-184.

Shyu, J. Z.; Weber, W. H.; Gandhi, H. S. J. Phys. Chem. 1988, 92, 4964-4970.
Powder Dififaction File Search Manual, 1. C. D. D, Swarthmore, 1987, 57,534.
Park, W. P.; Ledford, L. S. Langmuir, 1996, 12, 1794-1799.

CHAPTER 4

FUTURE WORK

4.1. Preparation of Transition Metal Oxide (Mn, Cr) Catalysts

One of the goals in this study has been to develop new synthetic methods which
offer us control over the surface structure of catalysts. This objective is important
because it would allow intentional design of catalytic properties into a particular catalys.
Differences in promoter and/or supported metal oxide bulk and surface structure would
be reflected in the catalysts’ stability, activity and selectivity. Preliminary results have
indicated that from various preparation techniques, like grafting and pore volume
impregnation, Ce/Al-catalysts with different crystallinity and Ce02 particle size were
obtained. We have characterized the structure of the cerium immobilized on the y-
alumina surface. Different activity and selectivity in oxidation reactions are expected due
to the various surface structures of the catalysts. In the future work, it is intended to use
similar methods of preparation to control the dispersion of multi-component catalysts
with transition (Cr, Mn) metal oxide surface active phase and rare earth oxide promoters.
Incipient wetness and alkoxide grafting will be used as preparation methods for catalyst
impregnation using different Mn (Cr) precursor solutions. Based on previous studies on
Mn (Cr) oxide based catalysts [1-6] as well as on the metal alkoxide chemistry of these
elements [7,8,9], a new generation of metal oxide supported catalysts could be developed

in this work.

76

77

4.2 Characterization of mixed metal catalysts

To understand the effects of promoter structure and composition, preparation
conditions, active surface phases and catalytic reactions using the new class of chromium
and manganese oxide cerium promoted y-alumina catalysts, several experimental probes
will be used. To obtain molecular level information about the preparation mechanism
(characterization of the grafting reaction) and the nature of the precursors used for
impregnation a detailed FTIR study will be performed. The goal of these IR studies is to
look for the existance and possible modes of formation of the chelated structures and
consequently to learn about other possible structures of the cerium methoxyethoxide in
the precursor solution with different chelating agents (like B-diketonates). With these
studies in hand, it would be possible to explain in more detail the nature of the
impregnation and understand the “modified grafting” mechanism. It is desirable to obtain
information about the relation between the catalysts structure and the reaction mechanism
and kinetics using in-situ DRIFTS measurements or to use EPR spectroscopy in the case
of a radical mechanism. Due to the paramagnetic properties of the cerium and manganese
oxides, EPR may be a proper way to analyze their influence on a reaction mechanism.

An important aspect of this research is the characterization of the crystallinity and
the dispersion of the active phases (promoter or active component). In order to accurately
determine the distribution of the metal oxide phase, the catalysts must be rigorously
characterized using bulk and surface analytical techniques. XRD will be used to

determine the chemical and physical nature of the species on catalyst surfaces, identifying

78

the crystalline phases in the catalysts by means of lattice structural parameters and
obtaining data regarding the particle size.

Additional information about the structure morphology and crystallography of the
supported species can be provided by electron microscopy. The method can reveal
information about the composition and internal structure of the particle. Transmission
electron microscopy (TEM) will be used in this work to provide a second measure of the
crystalline phase(s) and to characterize the amorphous or the highly dispersed surface
phases that can not be detected by XRD. These methods can provide more information
about the dispersion of the different species from the catalysts surface.

XPS can be used to obtain information about chemical state of supported phases
in the top 3 nm from the catalysts surface (approximately 10 atomic layers). However, in
many cases one is interested in the structure of the first atomic layer of the catalyst
surface. For example, several of the grafted catalysts to be prepared for this study involve
the formation of phases that are expected to be only 0.5-1.0 mm thick [8]. XPS can follow
chemical changes in the chemical nature of these surface phases but changes in particle
geometry may be difficult to discern. Ion scattering spectroscopy (ISS) can be used to
determine the composition and morphology of these monolayer systemsl. DRIFTS
analysis may complete the system characterization by giving information about the
catalytic properties of the new catalysts. It offers the advantage that it is an excellent

method for analyzing powder materials avoiding the tedious preparation of wafers used in

 

1The 188 method uses ions as input “particles”, with energies of 0.5-20 keV; it measures the energy of the
scattered ions as output “particles” after binary collisions. The method provides information on the first
layer from the surface monolayer (adsorbates on single-crystal substrate) [10]

79

common IR analysis. The analysis is very sensitive being able to detect adsorbed
molecules at minimum 10-5 of the monolayer coverage. Quantitative possibilities and in-
situ capabilities of DRIFTS makes the method very attractive for the purpose of this

project (kinetics and mechanistic measurements) [11].

4.3 Mechanism and kinetics of emission control reactions

In this part of the research, a transition metal oxide (Cr and Mn) base on a Ce(La)
promoted y-alumina catalyst will be designed and the system’s catalytic activity,
selectivity and stability will be probed in total oxidation reaction for CO, CH, and CHCI3
streams as well as in ethylbenzene oxydehydrogenation reaction. The effect of the
catalysts structure on catalytic properties will be carefully studied. Activity data (rate
constant determination and Arrhenius plots for activation energy) will be collected at low
conversion to limit heat and mass transfer effects on the rate measurements. This will
allow the determination of structure-activity relationships that could lead to more
advanced catalysts. In addition, the best catalysts will be studied at high conversion to
simulate actual industrial operating conditions and to observe the reaction mechanisms,
lifetimes of the catalysts and deactivation processes. With its excellent qualitative and
quantitative possibilities, DRIFTS analysis will be used for evaluation of the mechanism
and kinetics for the studied reaction. IR analysis will be performed in a cell capable of
obtaining temperature between 100K and 773K in a variety of reactive atmospheres [12].

This will allow measurements of adsorbed species to be performed in-situ. DRIFTS can

80

readily detect changes in the surface concentrations of CO or hydrocarbons fragments
adsorbed on a catalyst surface close to 10'5 monolayer coverage [11, 13].

It is intended to use the power of surface analytical techniques to determine the
nature of active sites and the mechanism of a probe process which can simulate a real
reaction used in emission control processes. An XPS instrument is equipped with an in-
situ reactor that can perform reactions at temperature up to 773K at atmospheric pressure
and transfer the reacted sample to the ultrahigh vacuum chamber without exposure to the
air. This will allow us to perform XPS analysis of the activated catalyst surface and
reveal the changes of the surface configuration after reactant adsorption. Chromium and
manganese are elements very labile to oxidation state changes. In-situ XPS
measurements may be crucial for identifying the active sites from these types of catalysts.
Oxidation-reduction studies using these cells may give us additional information about
the flexibility of oxidation state changes of the catalyst surface and the consequences on
the surface configuration. The great interest for the study of a Cr(Mn) oxide based
catalysts indicates their importance in industry and recently in emission control
processes. Despite the considerable amount of effort devoted to the study of Cr(Mn)
catalyst, the active sites of these catalysts, the role of promoters, the mixed effect of
single oxides, and the nature of any surface phases is still little known. More detailed
work, based on new catalyst preparation methods and advanced surface studies using the
new and sophisticated surface analysis techniques, is necessary for a better understanding
of these processes. These studies can be applied for industrial environmental protection

technologies and selective oxydehydrogenation reactions.

>19

10.
ll.

12.
13.

81

4.4 References

Ivanova, A. S.; Dzisco, V. A.; Moroz, E. M.; Nosckova, S. P. Kinet. Katal. 1986,
27(2), 428-434.

Robinsin, E. Rev. Port. Quin. 1977, 19(1-4), 103-106.

Johnson, A. J.; Meyer, F. G.; Hunter, D. 1.; Lombardi, E. F. Incineration of
polychlorinated Biphenyl using a F luidized Bed Incineration, Rockwell
International, 18-th Sep., REP-3271, 1981.

Bedford, R. E.; LaBarge, J. Base Metal Automotive Exhaust Catalysts with
Improved Activity and Stability and Method of making the Catalysts, US 5063193
A 5 Nov. 1991, 1-9.

Prasad, R.; Kennedy, L. A.; Ruckenstein, E. Combust. Sci. T echnoI. 1980, 22,
271-275.

Agarwal, S. K.; Spivey, J. J.; Butt, J. B. Appl. Colo], 1992, 81, 259-263.

Weldon, J.; Senkan, S. M. Combust. Sci. Technol. 1986, 47, 229-234.

Ribot, F.; Sanchez, C.; Livage, J. Chemical Processing of Advanced Materials,
1992, Wiley, NewYork, 267.

Sanchez, C; Toledano, P.; Ribot, F. Mat. Res. Soc. Symp. Proc. 1990, I80, 47.
Hercules, D. M. J. Chem. Ed. 1984, 61 592.

Craciun, R.; Dulamita, N.; Miller, D. J .; Kackson, J. E. submitted and accepted to
Progress in Catalysis, 1996.

Harnadeh, I. M.; King, D.; Griffiths, P. R. J. Catal. 1984, 88, 264-72.

Fraser, D. J. J.; Griffiths, P. R. Appl. Spectrosc. 1990, 44, 193-199.

 

 

"IIIIIIIIIIIIIIIIIIIIIIIII