3:
To...

, ”W.

.1 .I

. :3.
. .suiyfiw...
. .
‘ a!

.
fin

v

u
“fill“
‘ \ .
0 1:.
.3...- _.1 1

s i... 1.

V ,1

I
is!)
.3-

up.“
JP; 3 can
, ms mama

I
i
‘3
.
5': .
fih’gfi"

”E

i.
.. an. . :

Eb,

:w\9wrr:aq‘ ! I
m “whims...“ |<. .él
.Pnuwkfiu . ‘ . ‘ . ,L. 3...:2 a. is > .
x45)? Hi}.:iiihh... . . . ‘ we. ‘nfiv.d. 4.4V \ 14.9..“

fiSrafiFJvu-fi. v . . ; Lu“... .b...:m..m.. «tamiflu

. . . . £1... . 19$»

. . X!

it... in .1... -

“firetfl 2:333
23.5.: kl»;

bu.

:

:1)?! I
\vil xlv
t; .
. , x :1
. n t

.it. ~

.I .21

.110

”:10!

 

 

 

 

iii i531“;

I/iiiiiiflWii/‘ii‘l/iil

This is to certify that the

dissertation entitled

THE INFLUENCE OF SURFACE STRUCTURE ON THE
CATALYTIC ACTIVITY OF ALUMINA SUPPORTED
TRANSITION METAL OXIDES: A STUDY OF CARBON
MONOXIDE AND METHANE OXIDATION

presented by

Paul Worn Park

has been accepted towards fulfillment
of the requirements for

PhD. degree in _£hemisin¥

v/mgfl]gww4

 

or professor

Date WI?) M7!

M5 U is an Affirmative Action/Equal Opportunity Institution 0-12771

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to roman this chockout from your record.
To AVOID FINES return on or baton date duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

7 i -_.____—

 

THE INFLUENCE OF SURFACE STRUCTURE ON THE CATALYTIC ACTIVITY
OF ALUMINA SUPPORTED TRANSITION METAL OXIDES: A STUDY OF
CARBON MONOXIDE AND METHANE OXIDATION

By

PAUL WORN PARK

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1996

ABSTRACT

THE INFLUENCE OF SURFACE STRUCTURE ON THE CATALYTIC ACTIVITY
OF ALUMINA SUPPORTED TRANSITION METAL OXIDES: A STUDY OF
CARBON MONOXIDE AND METHANE OXIDATION
By

PAUL WORN PARK

The projects described in this thesis examine the effect of rare earth and transition
metal oxide promoters on the structure and reactivity of base metal oxide/alumina
catalysts for total oxidation reactions. The research involves the preparation and testing
of alumina supported Single and mixed metal oxide catalysts designed for total oxidation
reactions. The catalysts were prepared with various active phase loadings, promoter
contents, and precursors to control the chemical state and structure of supported phases.
The identification and quantification of various supported phases have been determined
using surface (XPS), bulk (XRD, ESR, FTIR. TGA, BET), and microprobe (TEM)
techniques. The information derived from these techniques was correlated with CO and
CH4 oxidation activities to identify active sites and to examine the effect of promoter on
the structure and reactivity of the catalysts.

The effect of crystallinity on the photoreduction of Ce(IV) species during XPS
analysis of CeOz samples and Ce/A1203 catalysts has been determined. The crystallinity of
the samples was controlled by varying the calcination temperature or by using difierent Ce
precursors (ammonium Ce(IV) nitrate and Ce(IV) methoxyethoxide). XPS, XRD, and

TEM results indicated that the amorphous cerium oxide was reduced more extensively

than the crystalline material during XPS analysis.

XRD, XPS, and ESR results indicated that an isolated c0pper surface phase, an
interacting copper surface phase, and large CuO crystallites are formed on Cu/A1203
catalysts, depending on Cu content. The crystalline CuO and isolated copper surface
phase has been assigned to the active site for CO and CH4 oxidation, respectively.
Cu/Ale3 catalyst prepared using Cu(II) ethoxide showed higher Cu dispersion, less
crystalline CuO, and lower oxidation activity for CO and CH4 than the catalyst derived
from Cu(II) nitrate. For Cu/Ce/A1203 catalysts, Ce had little effect on the dispersion and
crystallinity of the copper species. However, Cu addition decreased the Ce dispersion and
increased the amount of crystalline CeOz. Cerium addition dramatically increased the CO
oxidation activity, but it had little effect on CH4 oxidation. This indicated that cerium
strongly interacted with crystalline CuO, but not with copper surface phase.

The characterization of Cr/A1203 catalysts indicated that Cr was present as a highly
dispersed Cr6+ surface phase, Cr3+ clusters, and large Cr203 crystallites, depending on Cr
content. The active phase for CH4 oxidation has been assigned to a Cr(III)-Cr(VI)
interaction species. For the Cu/Cr/Ale3 catalysts, Cu addition decreased the dispersion
of the chrOmium phase by reacting selectively with a dispersed Cr3+ species to form
CuCr204. The Cu dispersion also decreased. with increasing Cr loading due to the
formation of CuO and CuCrZOI. These phases contributed to the catalytic activity of CO
oxidation with increasing Cr content up to Cr/Al = 0.054. For further Cr addition, the
catalytic activity decreased due to the decrease in catalyst dispersion or to encapsulation

of the active site by excess Cr species.

ACKNOWLEDGMENTS

I would like to thank Dr. Jeff S. Ledford, my research adviser, for lead me to the
fields of surface science and catalysis which are my all time-favorite subjects. I greatly
appreciate for his advice, correction, and encouragement whenever I needed. I also
appreciate professor Thomas J. Pinnavaia, my official mentor, who understands and
supports me constantly throughout my graduate school years at Michigan State
University.

There were many technical staffs in the department who deserve to get credits for
my research. Dr. Atkinson: He have always been available when I needed to fix and to
upgrade my computer which becomes the best PC in the chemistry building. The Glass
Shop: Their glassblowing technique must be one of the best in the country. I cannot
imagine without them how I could do my experiments with my easy-breaking hand. The
Machine Shop: They designed lot of tools and saved me a lot of money. The Electronic
Shop & The Electric Designer: Their magic fingers created and fixed my sophisticated
instrument and saved me from the disaster.

I have several friends whose friendship have certainly made my graduate school life
enjoyable. I will remember the friends, Tom Curtis, Ed Thounsand, Jeff Rasimas, Mike
Thelen, Mark Waner, Radu Craciun, and Dana Spence. Especially, I thank our final five
orphans Kathy Severin, Greg Noonan, Lee Hoffman, Per Askeland for our discussion and
encouragement each other. I wish you my best regards and to keep our friendship forever.

This work could never be completed without the support of my Mothers and

Fathers who are taking care of me and my family with endless and devoted love. Finally,

this thesis will be dedicated to my lovable wife, Jongmi who sacrifices herself to support

me and to take care of our kids. I will promise you to be a better husband and a father.

TABLE OF CONTENTS

List of Tables ..................................................................................... ix

List of Figures .................................................................................... xi

Chapter 1. Air Pollution and Pollution Control Catalysts

1.1.

1.2.

1.3.

Air Pollution ....................................................................... 1
1.1.1. Introduction .............................................................. 1
1.1.2. Origin and Effect of Air pollutants ..................................... 2
1.1.3. Pollution Control. Techniques ........................................... 4
1.1.3.1. Recovery Techniques ......................................... 5
1.1.3.2. Combustion Techniques ...................................... 6
Pollution Control Catalysts ...................................................... 7
1.2.]. Introduction .............................................................. 7
1.2.1.1. Principle of Catalytic Incineration ........................... 7
1.2.1.2. Advantages of Catalytic Pollution Control ................. 8
1.2.1.3. Disadvantages of Catalytic Pollution Control .............. 8
1.2.2. Noble Metal Catalysts ................................................... 9
1.2.2.1. Advantages of Noble Metal Catalysts ....................... 9
1.2.2.2. Disadvantages of Noble Metal Catalysts ................... 10
1.2.3. Transition Metal Oxide Catalysts ..................................... 11
1.2.3.1. Advantages of Transition Metal Oxide Catalysts ......... 11
1.2.3.2. Principles of Transition Metal Oxide Catalysts ............ 12
1.2.4. Industrial Waste Treatment Catalysts ................................. 14
1.2.4.1. Volatile Organic Compounds (VOC) Emission ........... 14
1.2.4.2. NOx Emission ................................................. 16
1.2.5. Automobile Emission Control Catalysts .............................. 21
References ........................................................................ 28

Chapter 2. Catalysts Preparation and Characterization Techniques

2.1.

Preparation of Catalysts ......................................................... 35
2.1.1. ‘Y-A1203 ................................................................... 36
2.1.1.1. Preparation of'y-A1203 ....................................... 36

vi

2.2.

2.3.

2.4.
2.5.

Chapter 3.

3.1.
3.2.
3.3.
3.4.
3.5.

Chapter 4.

4.1.
4.2.
4.3.
4.4.
4.5.

2.1.1.2. Structure ofy-A1203 .......................................... 36

2.1.1.3. Advantages of y-A1203 support ............................. 38
2.1.2. Catalysts Preparation Techniques ..................................... 40
2.1.2.1. Incipient wetness impregnation ............................. 40
2.1.2.2. Ion exchange .................................................. 41
2.1.2.3. Grafting ........................................................ 41
2.1.2.4. Precipitation ................................................... 42
Bulk Characterization ........................................................... 43
2.2.1. Surface Area Measurement (BET) .................................... 43
2.2.2. Electron Spin Resonance (ESR) ....................................... 48
2.2.3. X-ray Diffraction (XRD) ............................................... 51
2.2.3.1. Qualitative analysis of XRD ................................. 51
2.2.3.2. Quantitative analysis of XRD ............................... 54
2.2.3.3. Particle size determination ................................... 55
Surface Characterization ........................................................ 56
2.3.1. Ultrahigh Vacuum Chamber ........................................... 56
2.3.2. Transmission Electron Spectroscopy (TEM) ........................ 57
2.3.2.1. Instrumentation for TEM .................................... 60
2.3.2.2. Sample preparation for TEM ................................ 62
2.3.3. X-ray Photoelectron Spectroscopy (XPS) ........................... 63
2.3.3.1. Qualitative analysis of XPS .................................. 63
2.3.3.2. Quantitative analysis ofXPS ................................ 65
2.3.3.3. Particle size determination ................................... 66
2.3.3.4. XPS shake up satellite peaks ................................ 69
Catalytic Activity Measurement ................................................ 71
References ........................................................................ 74

The Effect of Crystallinity on the Photoreduction of Cerium Oxide: A
Study of CeO; and CC/Ale3 Catalysts

Abstract ........................................................................... 77
Introduction ...................................................................... 78
Experimental ..................................................................... 80
Results and Discussion .......................................................... 84
References ...................................................................... 100

The Influence of Surface Structure on the Catalytic Activity of Alumina
Supported Copper Oxide Catalysts: Oxidation of Methane and Carbon
Monoxide

Abstract ......................................................................... 1 03
Introduction .................................................................... 104
Experimental ................................................................... 106
Results .......................................................................... l 1 1
Discussion ...................................................................... l 18

vii

4.6.
4.7.

Conclusions ..................................................................... 122
References ...................................................................... 1 24

Chapter 5. Characterization and CH4 Oxidation Activity of Cr/Ale3 Catalysts

5.1.
5.2.
5.3.
5.4.
5.5.
5.6.

Abstract ......................................................................... 1 27
Introduction .................................................................... 128
Experimental ................................................................... 1 3 0
Results and Discussion ........................................................ 134
Conclusions ..................................................................... 144
References ...................................................................... 145

Chapter 6. Characterization and CO Oxidation Activity of Cu/Cr/A1203 Catalysts

6.1.
6.2.
6.3.
6.4.
6.5.
6.6.

Abstract ......................................................................... 149
Introduction .................................................................... 1 50
Experimental ................................................................... 1 52
Results and Discussion ........................................................ 157
Conclusions ..................................................................... 171
References ...................................................................... 1 72

Chapter 7. Influence of Cerium Promotion and Catalyst Precursors on the

7.1.
7.2.
7.3.
7.4.
7.5.
7.6.

Structure and Total Oxidation Activity of Alumina Supported Copper

Oxide Catalysts
Abstract ......................................................................... 175
Introduction .................................................................... 176
Experimental ................................................................... 178
Results and Discussion ........................................................ 183
Conclusions ..................................................................... 194
References ...................................................................... 196

viii

LIST OF TABLES

Table ........................................................................................... Page

Table 3.1. Particle Sizes of Cerium Oxide Species in Ce/A1203 Catalysts
Determined from XRD Line Broadening and XPS Intensity Ratios
Measurements ...................................................................... 97

Table 4.1. Concentration of Crystalline Phases in Cuy Catalysts Calculated from
Quantitative XRD Data ......................................................... 113

Table 4.2. Particle Size of Copper Phases Determined from XPS Cu/Al Intensity
Ratios and XRD Line Broadening Calculations .............................. 115

Table 4.3. Turn Over Numbers and Activation Energies of Cu/A1203 Catalysts
for CO and CH4 Oxidations ..................................................... 117

Table 5.1. Concentration of Crystalline Phases in Cry Catalysts Calculated from
Quantitative XRD Data ......................................................... 137

Table 5.2. XPS Cr 2p3a/A1 2p intensity ratios of Cr(VI) and Cr(III) Species in
Cry Catalysts Determined Using Non-Linear Least-Squares Curve
Fitting (NLLSCF) ................................................................ 139

Table 5.3. Particle Size and Dispersion of Chromium Phases Determined from
XRD Line Broadening Calculations and XPS Cr/Al Intensity Ratios ...... 142

Table 5.4. Turn Over Numbers and Activation Energies of Cry Catalysts for
CH4 Oxidation. ................................................................... 144

Table 6.1. XPS Cu 2pm Binding Energies and shakeup/main Peak Intensity
Ratios Measured for CuCry Catalysts and Standard Compounds .......... 158

ix

Table 6.2.

Table 6.3.

Table 6.4.

Table 6.5.

Table 6.6.

Table 7.1.

Table 7.2.

Table 7.3.

Distribution of Cr Oxidation States Determined from XPS Cr 2pm

Spectra of CuCry Catalysts Measured by non-linear least-squares
curve fitting (NLLSCF) ......................................................... 162

Concentration of Crystalline Phases in CuCry Catalysts Calculated
from Quantitative XRD Data ................................................... 162

Particle Sizes of Chromium and Copper Species Determined from
XPS Intensity Ratios ............................................................ 166

Particle Sizes of CuO, CT303, and CuCrZO. Phases Determined from
XRD Line Broadening Calculations ........................................... 166

Turn Over Numbers and Activation Energies of CuCry Catalysts for
CO Oxidation ..................................................................... 169

Concentration of Crystalline Phases in Cu/Ce/A1203 Catalysts
Calculated from Quantitative XRD Data ...................................... 186

Particle Size of Cerium and Copper Phases Determined from XRD
Line Broadening Calculations and XPS Intensity Ratios .................... 187

Turn Over Numbers and Activation Energies of Cu/Ce/Ale3
Catalysts for CO and CH4 Oxidation ........................................... 192

Figure .....

Figure 1.1.

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.

LIST OF FIGURES

..................................................................................... Page
Simultaneous conversion of hydrocarbon (HC), CO, and NOx for
TWC as a function of air/fuel ratio. (From Heck, R. M.; Farrauto, R.
J. Catalytic Air Pollution Control: Commercial Technology; Van
Nostrand Reinhold: New York, 1995) ........................................... 24
Schematic diagram of the formation of various alumina phases .............. 37
Lattice struture of spinel .......................................................... 39
The five types of adsorption isotherms (From Satterfield, C. N.
Heterogeneous Catalysis in Practice; McGraw-Hill: New York,
1980) ................................................................................ 44
Energy level splitting of the electron produced by an applied
magnetic field ....................................................................... 49
Constructive interference of scattered X-rays from crystalline Lattice ....... 53
Schematic drawings of XPS chamber ............................................ 58
Diagram of electron beam interaction with a solid sample .................... 59
TEM column (From Flegler, S. L.; Heckman, Jr. J. W.; Klomparens,
K. L. Scanning and Transmission Electron Microscopy: An
Introduction; W. H. Freeman and Company: New York, 1993 .............. 61
Diagram of XPS process .......................................................... 64

xi

Figure 2.10.

Figure 2.11.

Figure 3.1.

Figure 3.2.

Figure 3.3.

Figure 3.4.

Figure 3.5.

Figure 3.6.

Figure 3.7.

Figure 3.8.

Figure 4.1.

Figure 4.2.

Figure 4.3.

Figure 5.1.

Kerkhof and Moulijn's model of the supported catalysts ...................... 67
Schematic of the catalytic reactor system ....................................... 72

Ce 3d XPS spectra measured for (a) CeOz and (b) Ce(III)
acetylacetonate hydrate ........................................................... 85

TEM micrographs and SAED patterns obtained for (a) Ce200 and (b)
Ce750 samples ..................................................................... 87

XRD patterns measured for (a) Ce200 and (b) Ce750 samples ............... 89

Ce 3d XPS spectra measured for the Ce200 sample after irradiation
times of (a) 15 min., (b) 30 min, (c) 60 min., (d) 120 min, (e) 180

min., and (f) 300 min .............................................................. 91

Ce 3d XPS spectra measured for the Ce750 sample after irradiation
times of (a) 15 min., (b) 30 min, (c) 60 min., (d) 120 min., (e) 180
min., and (f) 300 min .. ............................................................ 93

Fraction of CeOz reduced as a function of irradiation time for the
Ce200 (O) and Ce750 (0) samples .............................................. 94

XRD patterns measured for (a) alumina support and (b) CeA and (c)
CeN catalysts ....................................................................... 96

XPS Ce 3d spectra measured for (a) CeN and (b) CeA catalysts ............. 98

XRD patterns measured for (a) A1203, (b) Cu7.7, (c) Cu10, and (d)
Cu13 ............................................................................... 112

Cu 2p33/A1 2p XPS intensity ratios measured for Cuy catalysts(.)
plotted versus Cu/Al atomic ratio. Cu2p3/2/A12p intensity ratios
calculated for monolayer dispersion (line) ..................................... 114

Normalized ESR intensity (1 mg CuO) of Cuy catalysts relative to the
internal standard .................................................................. 1 l6

XRD patterns measured for (a) A1203, (b) Cr] 1, and (c) Crl3 ............. 136

xii

Figure 5.2. Cr 2p XPS spectra obtained for (a) Cr1.3, (b) Cr2.7, (c) 054, (d)
Cr8.0, (e) Cr] 1, and (f) Cr13 catalysts ........................................ 138

Figure 5.3. Cr 2p/Al 2p XPS intensity ratios measured for Cry catalysts (O)
plotted versus Cr/Al atomic ratio. Cr 2p / A] 2p intensity ratios
calculated for monolayer dispersion (line) ..................................... 141

Figure 6.1. XRD patterns measured for CuCry catalysts ( O alumina, I CuO, 0
cubic CuCrzO4, V tetragonal CuCrzO4, ‘il’ Cr203). (a) alumina (b)
CuCrO, (c) CuCr1.3, (d) CuCr2.7, (e) CuCr5.4, (f) CuCr8.0, (g)
CuCr10.7, and (h) CuCr13.4 ................................................... 159

Figure 6.2. Cr 2p3/2 XPS spectra measured for CuCry catalysts. (a) CuCr1.3, (b)
CuCr2.7, (c) CuCr5.4, (d) CuCr8.0, (e) CuCr10.7, and (f) CuCr13.4 ..... 160

Figure 6.3. Cu 2p3,g/Al 2p XPS intensity ratios of CuCry catalysts plotted versus
Cr/Al atomic ratio. Monolayer dispersion value = 2.28 ..................... 165

Figure 6.4. Cr 2p/Al 2p XPS intensity ratios of Cry (0) and CuCry(O) catalysts
plotted versus Cr/Al atomic ratio. Cr 2p/Al 2p intensity ratios
calculated for monolayer dispersion (line) ..................................... 167

Figure 7.1. XRD patterns measured for (a) alumina support, (b) CeA, (c) CeN,
(d) CuA, and (e) C uN catalysts ................................................ 185

Figure 7.2. XRD patterns measured for (a) CuACeA (b) CuNCeA, (c)
CuACeN, and (d) CuNCeN catalysts .......................................... 189

xiii

Chapter 1

Air Pollution and Pollution Control Catalysts

1.1. Air Pollution

1.1.1. Introduction

In recent years, there has been great concern about environmental problems which
have resulted from the development of technologies. A tremendous amount of
governmental and scientific effort has been exerted to improve the environment all over
the world. While we still want to continue our economic growth which creates more
products, more jobs, and a more comfortable and pleasurable lifestyle, at the same time,
we also want to protect ourselves from an environmental crisis. One of the answers to
this apparent paradox (economic growth and a clean environment) can be addressed by
developments in catalytic science and technology [I].

The definition of air pollution was given by Seinfeld [2] as follows, ‘any
atmospheric condition in which substances are present at concentrations above their
normal ambient levels to produce a measurable effect (i.e., undesirable effect) on human
beings, animals, plants, or materials’. By this definition, the substances can be any natural

or artificial airborne chemicals which may exist in the atmosphere in gas, liquid, or solid

2
forms. Since it is impossible and unreasonable to eliminate all man-made emissions, it is
more desirable to focus on the reduction of pollutants to the level of no harmful effects.
The air pollution problems cannot be solved by the efforts of a few scientists. The study
of emission sources, behavior of emissions in the atmosphere, and the effect of emissions
on animals and plants requires a knowledge of the chemistry of combustion, engineering
aspects of the equipment design, meteorology, mechanics, atmospheric chemistry, aerosol
physics, physiology, medicine, plant pathology and so on. Since air pollution is also a
global problem which requires international cooperation, not only science and technology
fields but also social, economic, and political factors may play a major role in the

abatement of air pollution problems.

1.1.2. Origin and Effect of Air Pollutants

The origin of air pollution is an emission source. Major emission sources are
transportation, electric power generation, refuse burning, industrial and domestic fuel
burning, and industrial process. These emission sources produce air pollutants such as
carbon monoxide, carbon dioxide, sulfur compounds, nitrogen compounds, alcohol,
aldehyde, hydrocarbons, halogen compounds, particulate matter and radioactive
compounds. These air pollutants cause many undesirable effects such as respiratory
illness, visibility diminution, damage to animals and plants, and possibly catastrophic

effects on the global scale [2-4].

3

Carbon Monoxide and Dioxide(C0 and CO2): With the exception of C02, C0 is
the most abundant air pollutant in the lower atmosphere especially in urban areas. The
major source of C0 is the incomplete combustion of fossil fuels from automobile exhaust.
Carbon monoxide is quite stable in the air and the rate of conversion to C02 is very low.
Carbon monoxide is a poisonous inhalant because it has a strong affinity (250 times
greater than oxygen) for hemoglobin and replaces oxygen. C02 is not normally regarded
as an air pollutant according to the definition of air pollution and is an inevitable final
product of the combustion process. However, although C02 effects may have been small
to date, it may cause a significant effect in the fiIture when its concentration increases in
the atmosphere. C02 absorbs infrared terrestrial radiation and can increase surface and
atmospheric temperatures (green-house effect).

Sulfur Oxides (SO/J5 The sources of atmospheric sulfur oxides (802, 803) are
combustion of fossil fiJels (typically coal), decomposition and combustion of organic
matter from electric power generation and industrial sources. SOx also appears in auto
exhaust gases depending on the sulfur content of the fuel. S02 is highly soluble and
absorbed in the upper respiratory system causing constriction of the airways. It also
contributes to acid rain. .

Nitrogen oxides (NOJ: The primary sources of nitrogen oxides (predominantly
N0 and less than 5% N02) are high temperature combustion processes (particularly above
1500 °C) and the oxidation of fiJel containing nitrogen compounds. The combustion
process fixes atmospheric nitrogen to produce first nitrogen monoxide (N0), which will

be converted into nitrogen dioxide (N02). Under the action of solar radiation N02

4
dissociates into N0 and atomic oxygen which gives rise to the formation of ozone 03. The
ozone produces a narrowing of the airways in the lung and accelerates an aging of lung
tissue by oxidation. Organic molecules react with ozone to form free radicals which in
turn act as a catalyst for the oxidation of N0 and hydrocarbons. Nitrogen oxides are
known to cause bronchitis, pneumonia, susceptibility to viral infection and damage to the
immune system. They also contributes to acid rain, urban smog, and ozone.

Hydrocarbons: The combustion of gasoline is a major source of hydrocarbon
emissions as a result of either simple evaporation before combustion or incomplete
combustion. Benzene, its analogs, and aromatic polycyclic hydrocarbons are responsible
for mutagenic or carcinogenic action. Hydrocarbons along with nitrogen oxides are a very
important contributors to photochemical smog.

Particles: Fine particles (mostly fly ash), which consist of inorganic and organic
compounds of high molecular weight, are emitted from combustion systems (especially
coal combustion). The particles, due to their small diameter, penetrate deep into the lungs
and are retained for indefinite periods of time. Motor vehicles emit a number of metals
such as Pb, Cr, Mn, Ba, V, Fe, A1, Cd, and Ni from engine exhaust, panel body alterations,
tires, brakes systems, etc. Many of these metals are toxic, in particular Cd, Ni, Cr are

carcinogenic, while Mn affects to the nervous system.

1.1.3. Pollution Control Techniques

Control techniques for limiting pollutant formation and emission at the source can

be classified into three groups: modification of the basic process, substitution of cleaner-

burning fuels, and cleaning of the effluent gases before release to the atmosphere. The
particular technique is selected depending on pollutants involved, the process responsible
for pollutant formation, and the required degree of control [2]. Cleaning of effluent gases
has received the most attention and is the most widely used among the three control
techniques.

The most frequently used technique to clean effluent gases from emission sources
is the application of add-on devices. These devices are usually divided into two types:
recovery and incineration devices. The suitable control device is selected on the basis of

stream-specific characteristics and desired control efficiency.

1.1.3. 1. Recovery techniques

One of the recovery techniques is the gas absorption method which uses a liquid
solution capable of dissolving the pollutant gas molecules. Absorption is a diffusion
process that involves the transfer of molecules from the gas state into the liquid state
because of the contaminant concentration gradient between two phases. For best
operation, the contaminants should be readily soluble in the liquid phase and undergo an
irreversible reaction with a scrubbing reagent. An alternative recovery technique is to
adsorb pollutants on a solid surface which can retain gas molecules by physisorption or
chemisorption. Adsorption is a semi-continuous, diffusion-limited, surface area dependent
process that involves percolation of a gas through a solid bed. Adsorption depends on
selectivity and high capture efficiency which usually requires a significant mass-transfer

gradient from the gas phase to the surface [2,5].

1.1.3.2. Combustion Techniques

The combustion method involves chemical alteration of the pollutant to non-
pollutant molecules through thermal or catalytic incineration of the waste gases. Thermal
incineration is currently the most widely used method to control emissions from industrial
volatile organic compound (VOC) sources because it can provide high (> 99%) VOC
destruction efficiencies when operated above their auto-ignition temperatures (usually 500
- 1000 °C) [6.7]. In addition, thermal incinerators are relatively simple and economical to
install and operate. Thermal incinerators are applied principally when low concentrations
of combustible materials are present in a waste stream or when a reliable supply of waste
gas is available to use as a fuel for preheat burners. Thermal incinerators are generally
used when organic aerosols, significant amounts of non-combustible, high boiling point
compounds, inorganic materials, dusts, catalyst poisons or inhibitors are present in the
waste gas stream. However, thermal incineration requires a high fuel cost since the waste
gases need to be heated up to high operating temperatures for effective operation. The
high operating temperature of thermal incinerators may produce undesirable secondary
emissions such as N0x (via fixation of atmospheric nitrogen) and dioxins (via destruction

of chlorinated hydrocarbons) and create difficulties in mechanical design [5-7].

7

1.2. Pollution Control Catalysts

1.2.1. Introduction

1.2.1. 1. Principle of Catalytic Incineration

The objective of introducing a catalyst in a combustor is to carry out
heterogeneous oxidation on its surface. By choice of a suitable catalyst, one can ensure
that an activation energy for a heterogeneous reaction is much lower than that for a purely
homogeneous reaction. For instance, the activation energies for the homogeneous
oxidation of propane are approximately 25-50 kcal/mol, while those for heterogeneous
oxidation reactions are approximately 10—20 kcal/mol. Consequently, appreciable
heterogeneous oxidation rates can be achieved for temperatures and fuel concentrations
much lower than those required for the homogeneous reactions to proceed [8].
Therefore, catalytic oxidation is an efficient and effective method for burning low heating
value gaseous fuels and for the destruction of trace organic contaminants in air [9]. To
obtain optimal performance from a combustor, the catalyst system should possess several
properties. First, the catalyst should be able to ignite a fuel/air mixture at the lowest
possible temperature (low light off temperature). Second, the catalyst activity Should be
sufficiently high to maintain complete combustion at the lowest temperature and the
highest mass throughput. And finally, the support should have a large surface area, low

pressure drop, and high thermal and shock resistance [8].

1.2.1.2. Advantages of Catalytic Pollution Control

Primary advantages of using catalysts for emission control are that pollutants can
be completely oxidized at relatively low temperatures and consequently the formation of
N0,, is greatly suppressed compared to thermal incineration [10]. For the thermal process,
high temperature is normally maintained by burning extra find which represents up to 40 %
of total Operating cost. Therefore, operating cost is lower than for thermal incineration.
The decreased temperature also allows the use of corrosion resistance alloys for reactors
rather than replaceable ceramic linings which corrode rapidly in the thermal incinerator
[11]. The introduction of a heterogeneous catalyst also allows better control of oxidation
over wider air/fuel ratios. If a fuel concentration is too low to maintain catalyst
temperature, a catalyst bed can be heated effectively to the temperature at which oxidation
is initiated. In addition, the use of a catalyst improves energy recovery by minimizing
energy loss as visible light (i.e. non-flammable combustion) [12]. These advantages have
led to widespread application of catalytic incineration for the control of waste gas emitted

from industrial processes and automobile converters.

1.2.1.3. Disadvantages of Catalytic Pollution Control

The primary disadvantage of using catalysts for emission control is that all
catalysts are eventually deactivated physically and chemically. Physical deactivation
results from mechanical attrition and sintering which lead to loss in active components and

surface area, respectively. Chemical deactivation results from presence of impurities

 

9
which can lead to poisoning and accumulation of deposits which can lead to fouling.
Thus, many catalysts require frequent and costly replacement of a catalyst bed and
interruption of process operation [5,13]. Further disadvantages of catalytic combustors
are that the volumetric heat release rates and mass transfer coefficients are very low
compared to those of the thermal combuster [14]. In addition, destruction efficiency for a
given compound may vary depending on the composition of a mixture in an emission

stream [15].

1.2.2. Noble Metal Catalysts

1.2.2. 1. Advantages of Noble Metal Catalysts

Most commercial catalysts for total oxidation reactions utilize noble metals such as
Pt and Pd. Although, in principle, any noble metal (such as Ag, Au, Ru, Os, Ir) may be
used as a total oxidation catalyst, their application is limited due to sintering, volatility
loss, and irreversible oxidation resulting from excess oxygen present at high temperature.
Limited supply and high cost Of these metals also minimize their use [8]. Compared to
base metal oxides, noble metals have many advantages such as high intrinsic activity for
oxidation reactions, high thermal stability, and sulfur poisoning resistance [16]. In
addition, Pt and Pd are readily prepared in a highly dispersed form on a number of support
materials. For these reasons, only small amounts (0.1-0.5 wt.%) are typically necessary
for the production of good combustion catalysts. The high intrinsic activity of these
metals is related to their ability to activate H2, 02, C-H, and O-H bonds [8]. On noble

metals, oxygen adsorption is relatively weak. This implies that the mechanism may

10
proceed through direct electron transfer between reactant and oxygen on metal surface.
The general mechanism of oxidation over noble metals is suggested to involve the

dissociative adsorption of oxygen [17].

02 +1] -+ [02] —> ZIOI (H)

where [] represents a surface active site.

1.2.2.2. Disadvantages of Noble Metal Catalysts

While noble metal catalysts have many advantages, there are several problems in
the operation of the catalysts. This is especially true for chlorinated hydrocarbon
oxidation because such systems are susceptible to C12 and HCl poisoning, byproducts of
the reaction [58,59]. Noble metals are deactivated by sintering at temperatures of
between 500 ~ 900 °C [8]. Sintering reduces the active surface area and catalyst activity.
It has been found that in oxidizing atmosphere, noble metals can disperse as oxides on
A1203 surface at temperatures below the decomposition temperatures of the oxides (Pt02,
585 °C; PdO, 790 °C) [18]. Pt02 interacts weakly with y-AI2O3 surface so that the Pt
dispersion decreases and large metallic crystallites form at temperatures greater than 600
°C under oxidizing conditions [16]. The prolonged exposure of Pd to oxygen has been
observed to cause structural changes in Pd metal that result in loss of catalytic activity for
methane oxidation at temperatures above 450 °C [19]. In addition, catalytic combustors

using noble metal catalysts convert very high levels of fuel bound nitrogen to N0x (50~90

11
%) under lean condition [8]. The high price and limited supply are also major

disadvantages for using noble metals.

1.2.3. Transition Metal Oxide Catalysts

1.2.3.1. Advantages of Transition Metal Oxide Catalysts

Transition metal oxide catalysts have been considered as suitable substitutes for
noble metal catalysts because of their natural abundance and low cost. Current issues are
the development of contaminant resistant and thermally stable base metal oxide catalysts
for automobile catalytic converters and industrial catalytic incinerators [20]. Transition
metal oxide catalysts have greater resistance to certain poisons such as halogens, As, Pb,
and P [21] and yield a lower level of fiIel bound nitrogen conversion to N0x (10~20 %),
even in lean fuel Operation, than noble metal catalysts [8]. Thus, the use of oxide based
catalysts such as Cr203-C03O4 [22] and NiO [23] appears to be very promising for N01,
control from both thermal and fuel bound nitrogen. Metal oxide catalysts typically have
lower catalytic activity and higher light-off temperatures than noble metal catalysts.
However, for typical catalytic combustion, catalysts are operated under a diffusion control
region where their activities are comparable to those of the noble metals and under such

operation, the catalytic activity (kinetic control) plays a secondary role in determining the

reaction rate [8].

12
1.2. 3. 2. Principles of Transition Metal Oxide Catalysts

Blazowski et al. [24] indicated that C0304, Cr203, CuO, NiO, Mn02 and V205 are
potential catalysts for hydrocarbon oxidation. MnO2 and CuO are promising candidates
for olefin oxidation [25,26] and C0304 possesses the highest activity of base metal oxides
for paraffin or large molecular weight olefin oxidation reactions [27]. On the basis of the
activity of catalyst per unit surface area, C0304 and CuO catalysts showed high CO
oxidation activity [92].

The activity of the transition metal oxide catalysts for oxidation reactions may
depend on redox behavior of metal ions [28], metal-oxygen bond strength [29], and types
of oxygen involved in the reaction [30]. In addition, there should be an optimum level of
interaction between active sites and reactants. If the interaction is too strong, the catalyst
will be deactivated by irreversible chemisorption. If the interaction is too weak, only a
small fraction of the surface is covered and the catalytic activity is very low [31]. Low
temperature ESR studies have shown that the interaction of oxygen gas with an oxide

surface proceeds the following steps [32].

02 (gas) + e' —> 02' (adsorbed) (1.2)
02' + e' —> 20' (adsorbed) (1.3)
0' + e' —> 02' (lattice) (1.4)

The mobility of oxygen species is also important for oxidation catalysts because

highly mobile oxygen should result in a highly active and non-selective catalyst [9]. Both

13

adsorbed and lattice oxygen participate in a catalytic oxidation process. Haber [33]
postulated that surface adsorbed oxygen may generally lead to complete oxidation,
whereas lattice oxygen is used for partially oxidized products. This may be because
surface adsorbed oxygen is more mobile than lattice oxygen. Though, in principle, either
02' or 0' may promote total oxidation, 0' may be more reactive than 02' due to its higher
mobility. In addition, 0' is known to oxidize CO and H2 even at -196 °C [30]. Lattice
oxygen becomes important when the oxidation processes are at high temperatures (> 500
°C).

Mixtures of these oxides often exhibit greater stability and combustion activity
than the single oxides [34-37]. Considerable improvement in thermal stability can be
achieved by utilization of complex oxides such as rare earth perovskites [38] and mixed
transition metal oxides [39]. Indeed, a binary transition metal oxide catalyst comprised of
Cr203-CO304 supported on alumina has been successfully used for the catalytic
combustion of lean propane-air mixtures [40]. Catalysts comprised of La203-Cr203 also
appear to be very promising for oxidation at high temperature [41]. CuO-Mn02
(hopcalite) [42] and CuO-Cr203 [43] catalysts are well known as some of the most active
oxidation catalysts among base metal oxide catalysts. These catalysts were also studied
for the removal of air pollutants such as CO and N0 from exhaust gas [44,45]. The
synergistic effect of mixed oxide catalysts compared to their individual oxide components
may be due to the readily available multiple energy levels of the metals and their associated

oxygen anions, which makes the organic reactant more accessible to the active oxygen

14
anions. This also may result in higher surface mobility of oxygen and/or the activated

complex as well as electron transport through the lattice,[31].

1.2.4. Industrial Waste Treatment Catalysts

1.2.4.1. Volatile Organic Compounds (VOC) Emission

The control of organic emissions from industrial processes was first accomplished
by catalytic incineration in the late 1940’s [46]. However, after thermal incineration was
developed during 1960’s [13], the first catalytic system was installed for commercial use in
1975. Catalytic oxidation is widely used for trace contaminant removal in Europe due to
the high cost of energy and relatively strict environmental regulations. The Clean Air Act
1990 in the US raised the number of toxic chemicals to be controlled by the year 2000
from 7 to 189 (70 % of these are VOCS) [47]. Both catalytic and thermal combustion are
used for controlling VOCs. Some areas where catalysts are used beyond thermal
combustion are in can, paper, coating, organic chemical and plywood manufacture, tire
production, asphalt blowing, odor control, waste water plants, automotive exhaust, and to
remove air contaminants in submarines [48]. For these processes, a highly active and non-
selective catalyst is required. Pt shows high oxidation activity for saturated hydrocarbons
and high molecular weight species [49]. Pd is preferred for CO, olefins, methane and low
molecular weight olefin oxidation [50]. Base metal oxide catalysts are used less frequently
for waste gas which is relatively free of contaminants or for fluidized bed processes in

which the catalytic surface is refreshed by abrasion [51]. The preferred support for

organic abatement catalysts is either ceramic (cordierite) or metallic (aluminum, stainless

15
steel) monolith or honeycomb. These support materials can minimize pressure drOp with
parallel channels (200-400 channels per square inch). Most commonly 0.1-0.5 wt.% Pt on
high surface area y-Al203 is used for a bead or washcoat on a monolith. In limited cases, a
small amount of Pd or Rh is added to promote a particular reaction where the Pt may be
deficient. If the feed contains a large amount of sulfur ( > 50 ppm S), less reactive carriers
such as Ti02 or Ot-Al203 are used for Pt [10] to avoid formation of sulfates.

An air stripper is used to remove VOCS from contaminated ground water. The
emission from air strippers contains a wide range of compounds and concentrations of
VOCS including chlorinated organics. Chlorinated organics cause byproduct formation in
thermal oxidation systems. For instance, thermal destruction of polychlorinated biphenyls
(PCBS) requires temperatures of at least 1200 °C and residence times of over 1 second.
Standard municipal incinerators merely vaporize and release PCBs into the atmosphere
[52] or produce even more toxic polychlorinated dibenzofiirans (PCDFs) and
polychlorinated dibenzo-p-dioxins (PCDDS) [53-55]. Hence, there is a compelling need
for development of economically viable, lower temperature processes capable of
effectively treating incinerator flue gas or other effluents that may be contaminated with
PCB vapors [56].

Catalytic combustion systems may be the answer for air stripper emission control
due to its low temperature Operation, relatively humidity insensitive, and its compatibility

with low gas concentrations [57]. Transition metal catalysts are suitable candidates for
VOC control because noble metals (Pt, Pd) are susceptible to C12 and HCl poisoning from

the destruction of chlorinated hydrocarbons [58-62]. Cr203/Al203 is considered to be a

16
suitable catalyst for chlorinated hydrocarbon oxidation because it produces low levels of
C12 which is a more toxic compound than HCl. Weldon and Senkan [63] have determined
the C12 selectivity for CH3CI oxidation to be in the 3-7 % range over Cr203/A1203,

whereas a 19-42 % range of C12 selectivity was reported for copper based catalysts [64].

1.2.4.2. N0, Emission

N0,IL consists primarily of N0 and N02, which are produced in all combustion
processes by the oxidation of N2 and fuel bound nitrogen. At temperatures higher than
1500 °C, the reaction proceeds at appreciable rates through the Zeldovich mechanism

[65].

N2+O—>NO+N (1.5)
N+02—)N0+0 (1.6)
N+OH—>NO+H (1.7)

The rate of the above reactions is controlled by reaction (1.5), which has a high activation
energy and, as a result, thermal NO. formation is highly temperature dependent. If a
compound in the combustion process has bound nitrogen (e.g., pyridine), NOx is readily
formed at much lower temperature through an oxidation process [66]. Recently, the
emissions of N20 from fiber production plants has received attention due to its potential to

contribution to global warming (strong infrared absorptivity).

17
The major contribution to N0x generation is stationary sources. NOx emission
from power plants and lean burn engines cannot be controlled by three-way catalysts due
to the uncontrolled stoichiometric air/fuel ratio. The most effective NOx control strategies

are catalytic decomposition of nitrogen oxides and selective catalytic reduction [67].

Catalytic NO, decomposition: The decomposition of N03‘ to nitrogen and oxygen is one
of the most challenging approaches to remove N0x fiom exhaust gases.
Thermodynamically, NO should be decomposed to its constituents elements at low

temperature.

N0 (g) —> 1/2N2 + 1/202 AG" = -207 kcal/mol (1.8)

Accordingly, a catalytic N03 decomposition method could be the simplest and cheapest
technique Since no added reductant is necessary. Silver promoted C0304 [68], perovskite
type oxides [69,70], Pt/Al203 [71], Rh/Al203 [72], Cu-ZSM-S [73] have been reported to
be active catalysts for N0,X decomposition. Cu-ZSM-S catalyst is considered to be the
most active system [67]. However, at present these catalysts are not active enough under
commercial conditions [74]. Furthermore, the Cu-ZSM-S catalyst is poisoned by 502 [75]

and by oxygen in the feed gas or produced from N0 decomposition [76].

Catalytic reduction of N0,r by ammonia: Selective catalytic reduction (SCR) of NOx
using NH3 was first discovered with Pt catalysts in excess amounts of oxygen [77]. The

technology of selective catalytic reduction with ammonia was well developed in Japan and

18
Western EurOpe to control of NOx emissions from electricity generating power plants. The

Clean Air Act 1990 will prompt widespread use of SCR technologies in USA.

The desired reactions are:

4NH3 + 4N0 'l' 02 -> 4N2 + 61120 (1.9)

4N1‘13 ‘1' 2N02 '1‘ 02 —) 3N2 + 61120 (1.10)

The undesired reactions are:

4N1'13 ‘1' 302 ---> 2N3 + 61120 (1.11)

4N1'13 + 502 —-)4NO + 61120 (1.12)

Pt catalysts were not applicable because they have poor selectivity for N03‘
reduction at temperatures above 250 °C. Above 250 °C, a V205/Al203 catalyst was used
first, but was restricted to sulfur free exhaust gases due to its high sensitivity to poisoning.
802 is converted to $03 which reacts to form sulfates (Al2(SO4)3) which plug catalyst
pores and cause deactivation. Further, as the catalyst deteriorates, the activity at a fixed
NH3/N03 ratio decreases and slippage of NH3 increases, with resultant discharge of NH3
into the atmosphere [67]. Therefore, an alternative method of NOx control has been
developed using vanadia/titania catalysts. These catalysts facilitate reaction between NOx
and NH3 or hydrocarbons in the presence of oxygen. Si02, W03, M03 can be used as

promoters. A key to the success of the foi catalyst is that it is not poisoned by S02

19

because the T102 carrier is non-sulfating under operating conditions [78-80]. V205 based
catalysts operate best in the temperature range between 260-450 °C. The maximum
exposure temperature for this catalyst must not exceed 450 °C, because the active anatase
phase of Ti02, with a surface area of 80-120 mZ/g, irreversibly converts to rutile, with a
surface area of less than 10 mZ/g. Recently, Cu/Zeolite catalysts have been developed that
function at higher temperatures over a broader range of temperatures than Pt and V205
based catalysts [81]. These catalysts have much lower tendency to oxidize ammonia to
NOx and the maximum operating temperature is about 600 °C. Above 600 °C, the zeolite
is deactivated by a de-alumination process whereby the Al ions migrate out of the Si02-
A1203 structure [10].

The SCR technique using ammonia has received much attention because ammonia
is a very selective reducing agent to NOx under oxidative conditions. The other reactants
such as CO, H2, CH4 and hydrocarbons react readily with oxygen in the gas stream. There
are several disadvantages of using ammonia as a reductant. NH3 is more expensive than
hydrocarbons especially methane, it requires special handling and storage, and a

sophisticated metering system is needed to avoid NH3 leakage [10].

Catalytic reduction of NO; by hydrocarbons: Reduction of NOx by natural gas or other
hydrocarbon has recently attracted much attention as a possible new de-nitrification
process for exhaust gases from power plants, diesel and lean-bum gasoline engines. The
motivation to develop N0x reduction in lean environments using hydrocarbons is to

improve fiIel economy. Operating an internal combustion engine under lean conditions

20
improves the combustion efficiency, power output, and decreases the emission of the

greenhouse gas CO2. Additionally, in many new power plants, natural gas is commonly

used as a fuel and is readily available [10,67,82].

CH4 + 2N0 + 02 —) N2 + 2H2O '1' C02 (113)

Catalysts comprised of copper ion exchanged zeolite (Cu-ZSM-S) were the first
discovered to be active and selective for N02, reduction with propane at 500 °C under
oxidative conditions. Cu-ZSM-S catalyzed N02 reduction by C2H4, C3H6, C4113, C3Hg in
the presence of 02 [83,84]. H-ZSM-S was a particularly good catalyst for the reduction
of nitric oxide by C3113 in the presence of 02 [85]. Its activity is actually promoted by the
presence of 02, but it is ineffective at lower temperatures and inhibited by H20. In
contrast, the use of ethylene or propene reduces NOx at lower temperatures (160 -200 °C).
An Fe ion exchanged mordenite catalyst was recently reported to be most active for N02
reduction by C2H4 in the presence of 02 at 200 °C [86]. Since CH4 is a major hydrocarbon
emission from a gas-cogeneration system, such as a electric power plant, it would be
advantageous to develop a catalyst which is active for reduction of NOx by CH4 in lean
conditions [87,88]. Li and Armor [89] have found that a Co-ZSM-S catalyst showed
potential for using natural gas as a reductant. However, the reaction rate was slow and
was inhibited strongly by H2O. Noble metal catalysts are also regarded as promising.

Alumina supported Pt catalysts show higher activity and durability than Cu-ZSM-S

21
catalysts under real diesel exhaust conditions [90]. However, at present, no commercial

process has been developed for catalytic reduction ofNOx under oxidative conditions.

1.2.5. Automobile Emission Control Catalysts

The development of catalytic reactors for cleaning exhaust gases of automobiles
has been stimulated successfully by legislation motivated by environmental and political
issues. In 1959 and 1960, laws on motor vehicle emission standards were first enacted
and these standards could be met with the modification of engine parameters. The Clean
Air Act of 1970 announced further tightening of the standards that went well beyond
existing technology and catalytic research increased considerably. The catalytic converter
has been considered as the only technology available for meeting the stringent automobile
exhaust standards [91].

Automobile catalytic converters have been used in the US since 1974 (1975 model
year vehicles) in order to meet emission standards. The initial objective was to reduce the
emission of CO and hydrocarbons. The engine was operated in the fiiel rich side and the
concentrations of C0 and hydrocarbons are relatively high, whereas the NO"
concentration is low. The nitrogen oxide standard could be met by using exhaust gas
recirculation which leads to the formation of lower levels of nitrogen oxides by reducing
the combustion flame temperature in the engine [91]. The oxidation catalysts used until
1979 were a combination of Pt and Pd and operated in the temperature range of 250 - 600
°C with space velocities from 10,000 - 100,000 l/hr depending on the engine size and

driving mode. Typical catalyst compositions were Pt and Pd in a 2.5:] or 5:1 ratio,

22
ranging from 1.6-3.1 g/car [10]. Many base metal oxide candidates were also investigated
in this period, such as Cu, Cr, Ni, Mn [92,93].

Since 1979, the NOx standards tightened and exhaust gas recirculation alone was
no longer sufficient to control nitrogen oxides emission. In the beginning, the dual bed
system with separate reduction and oxidation catalysts was considered due to difficulties
in controlling stoichiometric air/fiJel ratio. The engine was operated under fiIel rich
conditions for the first bed so that a catalyst could reduce N03 with H2, C0 and
hydrocarbons. Remaining exhaust gases would be oxidized in the second bed with air
injection [94]. The activity of alumina supported noble metal catalysts for NO reduction
by C0-H2 mixture was reported in the order Ru>Rh>Pd>Pt [95]. However, Ru, Pd, and
Pt were not proper catalysts for N03, reduction beds. Ru was found to be volatile by
forming Ru03 or Ru04 when the engine exhaust was oxidizing and the temperature
exceeded about 700 °C [96]. In addition, Ru04 is a highly toxic compound. Pt and Pd
catalysts are less useful as NO reduction catalysts due to large production of NH3 which
would be reoxidized to N0, in second oxidation bed. Rh was suitable for the N01,
reduction bed because it has a high activity for N0, reduction [95], is less inhibited by C0
and sulfur [97], and forms less NH; than Pt or Pd. It also has an extended N03
conversion window [98].

The next generation of catalysts named three-way catalyst (TWC) could catalyze
all reactions simultaneously: hydrocarbons and carbon monoxide are oxidized to C02 and
H20, while nitric oxides are reduced to N2. The noble metal Rh combined with Pt is able
to do both sets of reactions if the engine exhaust is operated close to stoichiometric

air/fuel ratio. A typical example of the influence of air/fuel ratio to the conversions is

23
given in Figure 1.1. On the rich side (air<fuel), the NOx reduction occurs readily, while
the C0 and hydrocarbon oxidation reactions are prevented by insufficient 02. As the
air/fuel ratio approaches the stoichiometric point (air—fuel), simultaneous catalytic
conversion of all three occurs in a narrow window. On the lean side (air>fuel), the CO
and hydrocarbon conversions are high, but at the sacrifice of the NOx reduction. A special
control system to maintain the stoichiometric air/fire] ratio at all times required an advance
in this technology. This was made possible by the development Of the 02 sensor, which
was positioned before the catalyst bed in the exhaust manifold [99]. The catalysts in use
contain Pt, Pd, and Rh as major constituents supported on La203 (and/or BaO) stabilized
(1-2 wt.%) alumina. The alumina is usually introduced as a washcoat (about 15 wt.% and

30-50 um thick) to a ceramic honeycomb monolith made of cordierite

(2Mg0-2Al203'58i02) [100] containing 200-400 square channels per square inch. A

converter typically contains 1-3 g Pt and 01-05 g Rh and 0-3 g Pd. Pt is an effective
oxidation catalyst for CO and the complete oxidation of hydrocarbons. Pd is more active
than Pt to promote the oxidation of CO and hydrocarbons but is more sensitive to
poisoning and sintering than Pt in the exhaust environment. Both Pt and Pd promote the
reduction of nitric oxide, but they are less effective than Rh. Rh also contributes to C0
oxidation, but is mostly required for N0, reduction. In addition to the noble metals, three
way catalysts contain Ce02 and possibly other additives such as La, Ni, Fe. These base

metal additives are believed to improve catalyst performance by extending conversion

Figure 1.1.

24

Removal Efficiency (x)

 

100

40

 

Ctoionoinotric
l Alr-Fuai Ratio

 

 

 

 

I I I I I
14.3 14.4 14.5 14.8 14.7 14.8 14.9

 

Air-Fuel Ratio (wot/wot.)

Simultaneous conversion of hydrocarbon (HC), CO, and NOx for
TWC as a fiinction of air/fiael ratio. (From Heck, R. M.; Farrauto,
R. J. Catalytic Air Pollution Control: Commercial Technology;
Van Nostrand Reinhold: New York, 1995).

25
during the rapid air/fuel ratio perturbations and help to stabilize the alumina support
against thermal degradation [91].

Ceria may be added at a loading of 10-30 wt.% to the washcoat of three-way
catalysts to store oxygen, promote the water-gas shift reaction, stabilize the noble metals
against thermal damage, and to alter CO oxidation kinetics [101,102]. Ce02 enhances
NOx conversion to N2 by its ability of oxygen storage derived from N0x decomposition
during lean condition. Stored oxygen is then available for reaction with C0 and
hydrocarbons during subsequent rich condition [103]. This oxygen storage behavior
broadens the operating air/fuel ratio window of the catalyst [104]. Ceria has been shown
to enhance the decomposition of NO by extending the time before the noble metal catalyst
is deactivated by the accumulation of surface oxygen derived from N0 decomposition.
For instance, Rh/CeO2 is deactivated more slowly than is Rh/Al203 during N0
decomposition, probably due to oxygen spillover from the noble metals to the reduced
ceria [105]. In addition, ceria addition to an alumina supported Rh catalysts was shown to
enhance NO reduction activity at low temperatures by decreasing the apparent activation
energy for the reaction of CO with N0 and by shifting to positive order the dependence of

the rate on NO partial pressure [106].

The reactions are indicated below:

Rich condition: 2Ce02 + CO -—> Ce203 + C02 (1.14)

2C602 + CxHy -) C8203 + mC02 + [11120 (1.15)

26

Lean condition: Ce203 + %02 (or NO)-—> 2Ce02 (1.16)

Another benefit of Ce02 is that it is a good steam-reforming (or water gas Shift) catalyst
and catalyzes the reaction of C0 and hydrocarbons with H2O to produce H2 in the rich

condition. Then the H2 reduces a portion of the NO, to N2:

C0 + H2O —) 1'12 '1' C02 (1.17)
CxHy + Z1120 -) 11112 + mC02 (1.18)
2N0 + 2112 -) N2 + 21120 (1.19)

Ce is the component mainly responsible for the release of H25 gas by reducing Ce(SOI)2
which is formed with 803 and C602 in the lean mode. Nickel has been found to be an
effective scavenger of H25 [107]. Lanthanum is added to the alumina washcoat for
thermal stabilization [108,109].

Recently, the replacement of Pt and/or Rh with Pd has been studied for TWC
because Pd is relatively less expensive than Pt and Rh [1 10]. First, the Pd/Rh catalyst was
developed for an automotive application [111,112] and then the Pt/Pd catalyst was
developed to further reduce TWC cost [1 1 1].

Although three way catalysts have been developed which successfully control
automobile emissions, they still need to be improved. A major challenge in this field is to

develop a catalytic technology which removes N0,i under lean-bum conditions with the

27
objective of increasing fuel efficiency and environmental protection with a lower level of

CO2 production [67].

10.

ll.

12.

13.

14.

15.

28
1.3. References

Cusumano, J. A. In Science and Technology in Catalysis 1994 ; Izumi, Y., Arai,
H.; Iwamoto, M., Ed.; Elsevier: Amsterdam, 1995; p 3.

Seinfeld, J. H. Air Pollution Physical and Chemical Fundamentals ;
McGraw-Hill: New York, 1975.

Perkins, H. C. Air Pollution; McGraw-Hill: New York, 1974.

Chiron, M. In Catalysis and Automotive Pollution Control; Crucq, A.
Frennet, A. Ed.; Elsevier: Amsterdam, 1987; p1.

Bethea, R. M. Air Pollution Control Technology: An Engineering Analysis
point of View; Litton Educational Publishing: New York, 1978.

“Control of Volatile Organic Emissions from Existing Stationary Sources,
Volume 1: Control Methods for Surface Coating Operations”; (Prepared for
US. Environmental Protection Agency.) Research Triangle Park, NC. EPA
Publication No. 450/2-76-028. November 1976.

Rolke, R.W. “Afterburner Systems Study” (Prepared for US. Environmental
Protection Agency.) Research Triangle Park, NC. Publication No.
EPAR272062. August 1972.

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

Satterfield, C. N. Heterogeneous Catalysis in Practice ; McGraw-Hill:
New York, 1980.

Heck, R. M.; Farrauto, R. J. Catalytic Air Pollution Control: Commercial
T echnology; Van Nostrand Reinhold: New York, 1995.

Manning, M. P. Hazardous Waste 1984, l, 41.

Trimm, D. L. Appl. Catal. 1983, 7, 249.

Hardison, L. C .; Dowd, E. J. Chemical Engineering Progress 1977, 73, 31
Pfefferle, L. D.; Pfefferle, W. C. Catal. Rev. -Sci. Eng. 1987, 29, 219.

US. EPA. "Parametric Evaluation of VOC/HAP Destruction Via Catalytic
Incineration." EPA-600/2-85-041 (NTIS PBSS-l91187). April 1985.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25

26.

27.

28.

29.

30.

31.

32.

33.

29

Kummer, J. T. Prog. Energy Combust. Sci. 1979, 6, 177.

Golodets, G. I. Heterogeneous Catalytic Reactions Involving Molecular
Oxygen; Elsevier: Amsterdam, New York, 1983.

Yao, H. C.; Sieg, M.; Plummer, H. K. Jr. J. Catal. 1979, 59, 365.
Cullis, C. F.; Willatt, B. M. J. Catal. 1983, 83, 267.
Yang, B. L., Chan, S. F., Chang, W. S., and Chen, Y. Z. J. Catal. 1991, 130, 52.

Cimino, A.; De Angelis, B. A.; Luchetti, A.; Minelli, G. J. Catal. 1976, 45, 316.

Prasad, R.; Kennedy, L. A.; Ruckenstein, E. Combust. Sci. Technol. 1981, 27,
45.

B. A. Folsom et al., “Fuel nitrogen conversion-The impact of catalyst type”,
Energy & Environmental Research Corporation, EPA-600/9-80-035 (1980).

Blazowski, W. S. and Walsh, D. E. Combustion Science and Technology, 1975,
10, 233.

Morovka, Y.; Ozaki, A. J. Catal. 1966, 5, 116.
Morovka, Y.; Morikawa, Y.; Ozaki, A. J. Catal. 1967, 7, 23.

Boreskov, G. K. Proc. 5th Int. Congress on Catalysis, Amsterdam, paper 71,
1972. '

Labinger, J. A; Ott, K. C. Catal. Lett. 1990, 4, 245.

Bond, G. C. Heterogeneous Catalysis: Principles and Applications; Clarendon
Press: Oxford, UK, 1974.

Margolis, L. Y. Catal. Rev. 1973, 8, 241.

Spivey, J. J. In Catalysis, vol 8 The Royal Society of Chemistry, Cambridge,
1989; p 157.

Kon, M. Ya; Schvets, V, A,; Kazanskii, V. B. Kinet. Kat. 1972, 13, 635.

Haber, J. Int]. Chem. Eng. 1975, 15, 21.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48

49.

50.

3O

Dwyer, F. G., Catal. Rev. -Sci. Eng 1972, 6, 261.
Prasad, R.; Kennedy, L. A.; Ruckenstein, E. Catal. Rev-Sci. Eng. 1984, 26, 1.
Weldon, J. and Senkan, S. M. Combust. Sci. and Tech. 1986, 4 7, 229.

Dadyburjor, D. B., Jewur, S. S., and Ruckenstein, E. Catal. Rev.- Sci. Eng.
1979,19, 293.

Coutures, J.-P.; et al. High Temp. Sci. 1980, 13, 331.
Prasad, R.; et al. Comb. Sci. Techn. 1980, 26( 1&2), 51.

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

Tong, H.; Snow, G. O; Chu, E. K.; Chang, R. C. S.; Angwin, M. J.; Pessagno,
S. L. NASA CR-l65396, Sep. 1981.

Yoon, C.; Cocke, D. L. J. Catal. 1988, 113, 267.
Severino, F.; Brito, J.; Carias, 0.; Laine, J. J. Catal. 1986, 102, 172.

Dennott, J. M. C. in “Pollution Control Review”, vol 2, Noyes Data Corp.,
Park Ridge, IL, 1971.

Stegenga, S.; van Soest, R.; Kapteijn, F.; Moulijn, J. A. Appl. Catal. B 1993,
2, 257.

Hardison, L. C., “A Summary of the Use of Catalysts for Stationary Emissions
Sources Control”, Fisrt National Symposium on the Use of Heterogeneous
Catalysts for Control of Air Pollution, Philadelphia, Pa. Nov. 21-22, 1968.
Chemical Engineering, 1990, 24.

Armor, J. N. Appl. Catal. B, 1992, l, 221.

Yao, Y.-F. l & EC Prod. Res. Dev. 1980, 19, 293.

Stein, K.; Feenan, J .; Hofer, L.; Anderson, R. Bureau of Mins Bulletin, 1962,
608.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

31

Hardison, L.; Dowd, E. Chemical Engineering Progress, 1977, 75, 31.
Farrell, J. B.; Salotto, B. V. Effect of Incineration on Metals, Pesticides and
PCBs in Sewage Sludge; National Symposium on the Ultimate Disposal of
Wastewaters, and Their Residuals; Raleigh: NC, 1973.

Buser, H.R.; Bosshardt, H. P.; Rappe, C. Chemosphere 1978, 1, 109.

Buser, H.R.; Bosshardt, H. P.; Rappe, C. Chemosphere 1978, 2, 165.

Olie, K.; Verrnuelen, P. L.; Hutzinger, O. Chemosphere 1977, 8, 455.
Subbanna, P.; Greene, H.; Desal, F. Environ. Sci. Techno]. 1988, 22, 557.

Agarwal, S. K.; Spivey, J. J. Environmental Progress 1993, 12, 182.

Heyes, C. J.; Irwin, J. G.; Johnson, H. A.; Moss, R. L. J. Chem. Tech.
Biotechnol. 1982, 32, 1034.

Cullis, C. F.; Keene, D. E.; Trimm, D. L. J. Catal. 1970, 19, 378.

Wang, Y.; Shaw, H.; Farrauto, R. ACS Symposium Series 45, 1992, 125.
Yu, T.; Shaw, H.; Farrauto, R. ACS Symposium Series 45, 1992, 141.
Simone, D.; Kennelly, T.; Brungard, N.; Farrauto, R. App]. Catal. 1991, 70, 87.
Weldon, J.; Senkan, S. M. C ombust. Sci. and Tech. 1986, 47, 229.

Laidig, G., Hoenicke, D.; Griesbaum, K. Erdol. Kohle. Erdgas Petrochem.
1981, 34, 329.

Zeldovich, J. Acta Physiochim, 1946, 2], 577.
Fenimore, C. Combustion and Flame 1972, 19, 289.
Tamaru, K.; Mills, G. A. Catal. Today 1994, 22, 349.

Hamada, H., Kuwahara, Y., Kindaichi, Y., Ito, T. Nat. Meeting Chem. Soc.
Jpn., 1Va39, 1988.

Shimada, H., Miyama, 8., Kuroda, H. Chem. Lett. 1988, 1797.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

32

Viswanathan, B. Catal. Rev. -Sci. Eng. 1992, 34, 337.
Amimazmi, A.; Benson, J. E.; Boudart, M.; J. Catal. 1973, 30, 55.
Hardee, J. R.; Hightower, J. W. J. Catal. 1984, 86, 137.

Iwamoto, M. In Future Opportunities in Catalytic and Separation Technology;
Misono, M.; Moro-oka, Y.; Kimura, 8., Ed; Elsevier: Amsterdam, 1990; p121.

Catalytica, “Prospects for Catalytic Nitric Oxide Decomposition in Air Pollution
Control: Multi-Client Study”, Catalytica, Mountain View, CA. 1993.

Iwamoto, M.; Yahiro, H.; Shundo, S.; Yu-u, Y.; Mizuno, N. App]. C atal. 1991,
69, 5.

Li, Y.; Hall, W. K. .1. Catal. 1991, 129, 202.
Cohn, 6.; Steele, 0.; Andersen, H.; US. Patent 2,975,025 1961
Bosch, H.; Janssen, F. Catal. Today, 1988, 2, 369.

Farrauto, R. J .; Heck, R. M.; Speronello, B. 1992, Chemical and Engineering
News, 1992, 70, 34.

Heck, R. M.; Chen, J. M.; Speronello, B. K.; Morris, L. in Environmeta]
Catalysis; Armor, J. Ed.; ACS Symposium Series 552; American Chemical
Society: Washington, DC, 1994; p 215.

Heck, R. M.; Chen, J. M.; Speronello, B. K. Environmental Progress, 1994,
13, 221.

Gopalakrishnam,R.; Stafford, P. R.; Davidson, J. E.; Hecker, W. C .;
Bartholomew, C. H. Appl. Catal. B, 1993, 2, 165.

Held, W.; Koenig, T. R.; Puppe, L. SAE Pasper 900496, 1990.
Iwamoto, M.; Hamada, H. ('atal. Today, 1991, 10, 57.

Hamada, H.; Kintaichi, Y.; Sasaki, M.; Ito, T.; Tabata, M. Appl. Catal. 1991,
64, L1.

Sato, S.; Hirabayashi, H; Yahiro, H.; Mizuno, N.; Iwamoto, M. Catal. Lett.
1992,12, 193.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

33

Armor, J. N.; Li, Y. Prepr. Div. Pet. Chem. Am. Chem. Soc. 1994, 39, 104.

Zhang, X.; Vannice, M. A.; Walters, A. B. Prepr. Div. Pet. Chem. Am. Chem.
Soc. 1994, 39, 141.

Li, Y.; Armor, J. Appl. Catal. B, 1992, 1, L31.

Obuchi, A.; Ohi, A.; Nakamura, M.; Ogata, A.; Mizuno, K.; Ohuchi, H. Appl.
Catal. B, 1993, 2, 71.

Taylor, K. C. In Catalysis andAutomotive Pollution Control; Crucq, A.;
Frennet, A. Ed.; Elsevier: Amsterdam, 1987; p 97.

Yu-Yao, Y.-F. J. Catal. 1975, 39, 104.

Kummer, J. In Catalyst for the Control of Automotive Pollutants; McEvoy, J.
Ed.; ACS Series 143; American Chemical Society: Washington, DC, 1975; p 178.

Truex, T. J.; Searles, R. A.; Sun, D. C. Platinum Metals Rev. 1992, 36, 2.
Kobylinski, T.; Taylor, B. J. Catal. 1974, 33, 376.

Shelef, M.; Gandhi, H. Platinum Met. Rev. 1974, 18, 2.

Summers, J.; Baron, K. J. Catal. 1979, 57, 380.

Taylor, K. C.; Schlatter, J. C. .l. Catal. 1980, 63, 53.

Wang, T.; Soltis, R.; Logothetis, E.; Cook, J.; Hamburg, D. “Static
Characteristics onrO2 Exhaust Gas Oxygen Sensors”, SAE 930352. Warrendale,
Pa.: SAE International. 1993.

Gulati, S. T. Society of Automotive Engineers Publ. No. 85013 and 900268.
Taylor, K. C. CHEMTECH, 1990 Sept, 551.

Nunan, J. G.; Robota, H. J.; Cohn, M. 1; Bradley, S. A. In Catalysis and
Automotive Pollution Control I]; Crucq, A. Ed.; Elsevier: Amsterdam, 1991; p
221.

Gandhi, H. S.; Piken, A. G.; Shelef, M.; Delosh, R. G. Society of Automotive
Engineers Publ. No. 760201.

104.

105.

106.

107.

108.

109.

110.

111.

112.

34

Harrison, 8.; Cooper, B. J .; Wilkins, A. J. J. Plat. Metals Rev, 1981, 25, 14.
Loof, P.; Kasemo, B.; Anderson, S.; Frestad, A. J. Catal. 1991, 130, 181.
Oh, S. H. J. Catal. 1990, 124, 477.

Golunski, S.; Roth, S. Catal. Today, 1991, 9, 105.

Schaper, H., DoeSburg, E. B. M., and Van Reijen, L. L. Appl. Catal. 1983, 7,
211.

Oudet, F., Courtine, P., and Vejux, A. J. Catal. 1988, 114, 112.
Muraki, 111.; Yokota, K.; Fujitani, Y. App]. Catal. 1989, 48, 93.

Lui, Y.; Dettling, J. “Evolution of a Pd/Rh TWC Catalyst Thechnology” SAE
930249, Warrendale, Pa: SAE International. 1993

Summers, J.; Williamson, W.; Henk, M. “Use of Palladium in Automotive
Applications” SAE 920094, Warrendale, Pa: SAE International. 1988.

Chapter 2

Catalysts Preparation and Characterization Techniques

2.1. Preparation of Catalysts

Heterogeneous catalysts used in our work consist of two or three components: the
active component, promoter, and support. The active component is responsible for a
catalytic reaction. The promoter is catalytically inactive for a reaction, however, it
enhances the catalytic performance through interaction with the active component. For
Cu/Cr/Al203 and Cu/Ce/Al203 catalysts of interest in our study, Cu is the main active
component for CO and CH4 oxidation and Cr and Ce are considered as promoters. The
support plays a role of carrier for the active phase or promoter. The support provides a
large surface area and disperses an active phase on it in order to maximize the surface area
of the active phase. The y-alumina used in our study provides about 200 mZ/g surface area
and 0.6 mL/g pore volume. Alumina is quite inactive for C0 and CH4 oxidation in our
study. In other cases, alumina can be used as a catalyst by itself for alcohol dehydration,

alkene isomerization, and deuterium exchange reaction [1].

35

36

2.1.1. y-AI2O3

2. 1. 1. 1. Preparation of y-A I203

Aluminas exist as a various amorphous and micro-crystalline forms which can be
distinguished by X-ray diffraction. y-AI2O3 is the most important catalyst support due to
its thermally stability, controllable surface acidity, and high surface area [2]. Aluminum
metal can be dissolved in both acidic and basic solvents because of its amphoteric
character. Dissolved aluminum species will exist as solvated A131 ions in acidic solutions
(< pH 2) and aluminate ions (A1045) in basic solutions (> pH 12). From using an aqueous
acidic Al3+ solution (e.g. aluminum sulfate solution), aqueous basic A1045' solution (e.g.
sodium aluminate solution), or mixture of the basic and the acidic solution, precipitation
(aluminum hydroxide) can be obtained if the pH of the solution is properly controlled by
addition of extra acid or base. The various structures of aluminas are generally prepared
by subsequent dehydration (aging or heating) of the precipitated aluminum hydroxide.
Since the structure transformation is determined by kinetic factors rather than
thermodynamics, a mixture of different alumina phases are frequently obtained [3]. The

transformation of aluminas observed upon aging and heating is Shown in Figure 2.1.

2.1.1.2. Structure of y-A I203

The oxygen arrangement in y-Al203 is cubic close packing, with occupation of the
cations (Al3') located partially in tetrahedral and octahedral sites. This arrangement is
similar to the spinel structure (e. g. MgAl2O4) which contains Mgzi ions in tetrahedral and

Al3+ in octahedral interstices. The unit cell of a cubic close packed structure

7
3

83:; «£52m 25:9» Co SEES.“ 65 mo EEwEv usannom ._.~ 8sz

cowacw MO?
339 38:3.

 

 

 

.o._<-o Al 6.25 l 5.2-» l 2:538 05:5ch Al Esme... _<
0.8.x 98:58 9.22-me 953-com + wvzavo 228.38.“?
:3
4 Al 6.2-x Alli 6.2-x All 0:236 9.3 meme
O. 8 _ .A O. o: .58 O. 825m
m m
15
O 2 In FREE mime :vmq w Invfl
n V
AON_<-o Al .O~_<-: Al 59? All .5328 ...O_<
’ A10 OmZA 8:995 oE—Eabu 2mm; mzoozc<
. O. 8:52 O. 832 28
v.9. mama
«vi.
5902 3:?“ 8:28 ...._<

 

 

”Os—e. mag—923‘ Al Em BEsoom o==_Smbooco_§ 41 228 302:5.

Us Om _ _A Us coo omen

38
is shown in Figure 2.2. A face centered cube (fcc) contains four oxygen atoms, four
octahedral and eight tetrahedral sites. The atomic ratio of metal to oxygen of a true spinel
structure is 3:4, as in Mn304, Fe304 or C0304, while the atomic ratio of alumina is 2:3.
Therefore, there are more cation vacant Sites in the A1203 structure (so called defect
spinel) than in a true spinel [2]. The unit cell consists of 32 oxygen and 21-1/3 aluminum
atoms, therefore 2-2/3 cation vacant sites per unit cell occur in the lattice [1]. These
vacant sites sometimes cause formation of solid solutions with metal cations during
catalyst preparation or aging at high temperature. The formula of the y-alumina is given
formally as Ho,5Alo,5[Al2]O4 in which two out of four octahedral sites and one out of the
eight tetrahedral sites are occupied by hydrogen and aluminum. However, hydrogen
atoms are not located in tetrahedral or octahedral positions. Instead, they are at the

surface as OH groups. One OH group is present for every two fcc units [3].

2.1.1.3. Advantages of y-A ]_203 support

Alumina support has been used the most widely in industry among variety of
support materials due to its several advantages for practical use [2,4]. Alumina is cheap
since its raw materials (Gibbsite or aluminum) are available in large amounts at low cost.
Alumina is inert for many chemical reactions, has good mechanical strength (such as
attrition resistance, hardness, compressive strength) and maintains high surface area at
high temperature. Alumina is amphoteric so that its surface can be electrically charged
either positively or negatively according to the pH of solution and can adsorb ions

selectively. Active metal (Pt) is dispersed as small clusters on alumina due to the very low

39

 

 

 

 

. Octahedral site
Oxygen
. Tetrahedral site

Figure 2.2. Lattice structure of spinel.

4O
diffusion of the metal. In addition, alumina can be shaped with an accurate control of its

porosity.

2.1.2. Catalyst Preparation Techniques

2.1.2.1. Incipient wetness impregnation

Incipient wetness impregnation (or dry impregnation) is widely used in industry
due to its simplicity. Accurate control of the amount of support and active ingredient can
be achieved by this procedure. The support is contacted with a solution of appropriate
active ingredient, corresponding in quantity to the total known pore volume of the
support. Capillary force ensures intimate contact of the solution with the interior support
surface (meso and micro-pores). The time required for incipient wetness impregnation can

be calculated from the following equation, if the contact angle is assumed to be zero

_ 411x
yd

 

(1)

where r] is the liquid viscosity, x is the penetration distance into a capillary, y is the surface
tension, and d is the pore diameter. A disadvantage of this method is the limited maximum
loading in a single impregnation due to the solubility of the reagent [2]. In our work, all
catalysts were prepared by incipient wetness technique with aqueous metal nitrate or
alcoholic metal alkoxide solution. The binary component catalysts, such as CulCr/Al203

or CulCe/Al203, were prepared by step-wise incipient impregnation methods. The Cr or

41
Ce modified alumina catalysts were prepared first and then Cu was impregnated
subsequently on the modified alumina. This Step-wise impregnation method allows us to

study supported binary oxide catalysts in a simple, systematic way.

2. 1. 2. 2. Ion exchange

Ion exchange is very suitable as a method for preparing atomically dispersed metal
[5] or metal oxide [6] catalysts. The principle of this method is based on the fact that the
support surface is polarized at the pH of the solution and counter ions are adsorbed by
electrostatic attraction. Adsorption phenomena involving ion complexes are determined
by the isoelectric point of the support, the pH of the aqueous solution, and the nature of
the ion complex. The isoelectric point (or zero point of charge) of y-alumina is in the
range of 7-9 (so called amphoteric point). Therefore its surface charge can be controlled
by solution pH in order to adsorb cations or anions selectively. Several disadvantages of
the ion exchange method must be considered, such as the adsorption capacity limit of the
support, metal complex stability, and the solubility of the support material in acidic or

basic solution [7].

2. 1.2. 3. Grafting

Grafting methods have been developed recently for preparing monolayer catalyst
dispersions [8,9]. The method is based on condensation reaction between the metal
alkoxide precursor and the hydroxyl group of the support. This chemical reaction yields

the corresponding alcohol and a strong metal-support bond. Since alkoxide groups react

42
with water to be polymerized, nonaqueous solution such as alcohol derivative is
commonly used. Several disadvantages of grafting methods are such as availability,
solubility, moisture sensitivity of metal alkoxide, and reactivity with hydroxyl group of

support.

2. 1. 2. 4. Precipitation

Precipitation is one of the commonly used methods for making catalysts. The
procedure involves precipitation from a mixture of more than two solutions or suspensions
of material. The principle of precipitation is that the reaction between aqueous metal salt
and basic solution (such as ammonium hydroxide, or ammonium carbonate) produces
insoluble metal hydroxide or carbonate. Several advantages of precipitation method are:
producing more homogeneous mixing of catalyst ingredients, homogeneous distribution of
active phase throughout bulk and surface of catalyst, suitable for preparing high loading
catalyst, and controllable pore size and distribution. However, there are disadvantages
when two or more metal compounds are precipitated in different rate or sequence rather
than simultaneously and some of the active phase will be unavailable to the surface due to

being encapsulated by other material [2].

43

2.2. Bulk Characterization

2.2.1. Surface Area Measurement (BET)

The adsorption of a particular molecular species from a gas or liquid phase onto
the surface of a solid is the principal method for measuring the total surface area of porous
structures. If one can control the conditions of monolayer coverage and calculate the
cross section of the molecule, then the surface area of a sample can be obtained from the
volume of the adsorbed molecules. The molecules used for measuring surface area should
be inert, small, spherical and easy to handle at the required temperature. Though not
spherical, N2 is usually used since it is cheap and readily available in high purity. For
measuring the surface area of solid materials, nonspecific physical adsorption is required.
Physical adsorption isotherms have been classified into five groups by Brunauer, Deming,
Deming, and Teller (called BDDT classification) [10] (Figure 2.3). Type I (Langmuir
type) is commonly found for activated carbons, silica gels, and zeolites that contain
micropores (< 10 A). Type II (sigmoid) is found for nonporous structures. Complete
monolayer coverage is achieved at point B. Type III is found for systems having weak
adsorption forces (e. g. water vapor on graphite). Type IV is a typically found for porous
materials. The adsorption isotherm increases significantly at higher P/Po ratio due to pore
condensation. Type V is similar to Type III with pore condensation at high P/Po ratio.
Type III and V are relatively rare and difficult to use for surface area determination due to

multilayer formation before completing monolayer coverage.

44

.83. 15> 262 “Eitsemvo—z 66.55% E hibesb Sentences»:
.Z .0 63:03am EBB Eco—LEE .8398? go womb o>c 2:.

omi .8388.“ 2523.
o o. _

 

 

 

 

_

 

AGE .2335 9523.
o. _

 

_

 

>—

 

V
m
0
n
N
c
p

m
q
m.

MN Sawm—

o

 

o
_
s
—

 

#— paqsospc tunomv

45

The most common method for measuring the surface areas of catalyst is that
developed by Brunauer, Emmett, and Teller (called the BET method) [11]. They
extended the Langmuir mechanism [12] to multimolecular layers. Langmuir assumed that
an adsorption site on the solid surface accommodates one adsorbed molecule and the
adsorption sites are uniform regardless of surface coverage, so that the adsorption
probability (heat of adsorption) is the same at all sites. He also proposed that the rate of
evaporation equals the rate of condensation for the first layer under equilibrium condition.
For multimolecular layers, Brunauer, Emmett, and Teller assumed that the adsorption rate
was proportional to the vacant sites of the lower layer and that the desorption rate was
proportional to the adsorbed molecules in that layer. They also assumed that the heat of
adsorption for all layers except the first one is equal to the heat of condensation of the
molecule. The summation over an infinite number of adsorbed layers gives the following

equation;

P __ 1 + (C -1)P
V(Po ‘- P) VmC VmCPo

 

(2)

where V is the volume of gas adsorbed at pressure P, V... is the volume of gas adsorbed in
monolayer, P0 is the saturation pressure of adsorbate gas at the experimental temperature,
C is a constant related exponentially to the heats of adsorption and condensation of the
gas. The BET equation yields a straight line when P/V(P,,-P) is plotted versus P/Po. The

BET plot is usually found to be linear in the range P/Po = 005-035, and this range is

46

usually used for surface area measurements. The deviation from linearity increases at

higher P/P0 values (> 0.35) due to multilayer adsorption and/or pore condensation. The

experimental error increases at lower P/P0 (< 0.05) due to the small amount of molecules

adsorbed. The slope (S) is expressed as

 

S = (C — 1) ,
WC
and the intercept (I) is expressed as
I = ]
VmC

Solving the equations (3) and (4) for Vm gives

v...=—‘—
S+I

 

(3)

(4)

(5)

The surface area of the catalyst can be calculated from Vm if the cross sectional area of an

adsorbed molecule is known. Practically, the BET measurement can be performed with

the value of the weight adsorbed on the sample (W & Wm) instead of volume (V & Va)

[13]. From the PVT relationship the amount of adsorbate adsorbed on the sample can be

calculated from calibrated integrator counts caused by thermal desorption of the

adsorbate.

47

— _ Veal — (6)

where W is the mass of adsorbate adsorbed on the sample, A is the sample integrator
counts, A... is the calibration integrator counts, Veal is the calibration volume (cm3), P is
the ambient pressure, M is the adsorbate molecular weight (g), R is the gas constant (82.1

mL atm deg'l mol’l). Therefore equation (2) and (5) can be rewritten as

 

 

 

P = 1 + (C — 1)P (7)
W(Po - P) WmC WmCPo
W = .___l._ (3)
m S + I
The total surface area of the sample (8.) is determined from following equation;
WmNAcs
51: M (9)

where Wm is the weight of adsorbate adsorbed at a coverage of one monolayer, N is
Avogadro’s number (602 x 1073), Acs is the cross sectional area of adsorbate molecule

(N2 = 0.162 nmz), M is the adsorbate molecular weight (g). The specific surface area S is

given by following equation

lib-D

pL

, .27

48

s= . & no
weight of sample (g)

 

2.2.2. Electron Spin Resonance (ESR)

Electron spin resonance (electron paramagnetic resonance or electron magnetic
resonance) can be applied only to paramagnetic species such as supported metal ions,
surface defects, adsorbed molecules, and ions on heterogeneous catalysts. In ESR,

different energy states arise from the interaction of the unpaired electron spin moment

with the applied magnetic field (Zeeman effect). The Zeeman Hamiltonian (if) for the

interaction of an electron with the magnetic field is given by following equation [14];

%=gBHsz on

where g is the gyromagnetic value for a free electron (2.0023193), B is the electron Bohr

magneton (eh/2mcc), SZ is the Spin operator, and If is the applied field Strength. In the

absence of an applied magnetic field, the electrons are oriented at random. However,
under applied magnetic field, the electron spin axes are oriented either in the same
direction (+) or in the opposite direction (-) with respect to H. Therefore, there are two
spin populations: electron spin of m, = -1/2 and m, = 1/2 has lower and higher energy state

respectively (Figure 2.4). The transition energy difference is given as [14]

49

4+
E E = +1/2gl3H
m8 = +1/2

/

hv = gBH

 

V

 

rns = +1/2 0

m,=-1/2 \

m5 = -1/2
E = -1/2gBH

 

 

Figure 2.4. Energy level splitting of the electron produced by
an applied magnetic field.

50

AE=hv=gB% an

The EPR experiment usually is performed by sweeping the magnetic field at fixed
frequency. When the energy difference of two states (AE) matches with the hv, electrons
spin flip from the antiparallel to parallel. This point is called the resonance condition at
which energy absorption occurs. Since absorption curves are weak, EPR spectra are
usually represented as a derivative curve. The microwave frequency region such as X-
band (~ 9.5 GHz) and Q-band (~35 GHz) frequency ranges are commonly used in ESR.
Water has to be avoid for ESR measurement due to Strong absorption of microwave.
Quartz sample tube is preferred to Pyrex one because Pyrex absorb more microwave and
also exhibit an IEPR signal.

In this study, ESR has been used for semi-quantitative analysis for copper ions in
Cu/Al203 catalysts. ESR selectively detects isolated Cu2+ in Cu/Al203 catalysts. Although
clustered and crystalline CuO have paramagnetic Cuzi ions, their resonance absorption is
too broad to be detected due to the interaction with neighbor copper ions. The number of
spins can be obtained by the ratio of catalyst sample to the standard signal based on

following equation [15];

= (Au/ASXSS/gu )2“s 55(55 +1)
u PL Su(su +1)

 

U”

51
where n“ and n5 are the number of spins per gram, A,' and A, are the areas of resonance
absorption curves of sample and standard, gu and g. are the g values, Su and S. are total
spins of sample and standard respectively, p is the mass per unit length of sample in the
sample tube, and L is experimental factor. Therefore, the intensity ratio of sample vs.
standard is proportional to the absolute number of spins per gram sample.

Crystalline CuSO4.5H20 or phosphorous doped silicon embedded in polyethylene
are usually used for intensity standards. In our work, a Ruby crystal standard was glued at
the center of the inside cavity for improved accuracy, because each loading of the cavity
will slightly alter the standing wave of the microwave field. However, the greatest source
of error usually comes from double integration of the experimental derivative curve due to
the difficulty in determinating the base line. Measuring peak to peak distance was adapted

in this work rather than double integration since both methods showed very similar trends.

2.2.3. X-Ray Diffraction (XRD)

2. 2. 3. l . Qualitative analysis of X RD

X-rays are sufficiently energetic to penetrate solid, and are well suited to probe
their internal structure by diffraction. XRD is used to identify bulk crystalline phases, to
quantify the amount of the crystalline phase, and to estimate particle sizes that are large
enough to be detected. When an X-ray source target (e.g. Cu) is bombarded with high
energy electron, the Cu Ku line (8.04 keV, 0.154 nm) is irradiated because a primary

electron creates a hole in the K shell which is filled by an electron from the L shell with

52
emission of an X-ray photon. The KB line is also observed due to the electron transition
from the M-shell to the K-shell. X-ray diffraction is the elastic scattering of X-ray
photons by atoms in a periodic lattice. When X-rays are scattered by the ordered
environment in a crystal, interference (both constructive and destructive) takes place
among the scattered rays because the distances between the scattering centers are of the
same order of magnitude as the wavelength of the radiation. Figure 2.5 shows how
diffraction of X-rays by crystal planes allows one to derive lattice spacing by using the

Bragg equation [16]

nl=2dsin(O) (14)

where n is the integer order of the reflection, A is the wavelength of the X-ray, d is the
distance between two lattice planes, O is the angle between the incoming X-rays and the
normal to the reflecting lattice plane.

The XRD pattern is obtained with a stationary X-ray source and a movable
detector. XRD peaks appear to be reflected from the crystal only if the angle of incidence
for the X-ray satisfies the condition sinO = nMZd. At all other angles, destructive
interference occurs. If one measures peak positions as the angle 28 (which is angle
between the incoming and the diffracted beams) from a XRD pattern, the Bragg equation
gives the corresponding lattice spacing, which are unique characteristic for a crystal
compound. For powder sample such as catalyst sample, which randomly oriented
crystalline particle, only a small portion of powder sample will be oriented by chance to

provide a right angle with the incident beam for constructive interference. The limitation

53

in coming X-ray diffracted X-ray

 

 

l
e o d
9 pdsinO I

 

Figure 2.5. Constructive interference of scattered X-rays from
crystalline lattice.

54
of XRD is that diffraction peaks are only observed when the sample contains large enough
crystalline particles. If the sample contains amorphous phase or very small crystalline
particle size (less than 3 nm), no diffraction peaks will be observed due to destructive

interference in scattering directions where the X-rays are out of phase [17].

2. 2. 3. 2. Quantitative analysis of XRD

In our work, quantitative X-ray diffraction data were obtained by comparing metal

oxide (e.g. CuO<Ill>, Cr203<104>, and CuCr204<311>) vs. support (Al2O3<400>)
intensity ratios measured for catalyst samples with intensity ratios measured for the
physical mixtures of pure CuO, Cr203, or CuCr2O4 and y-Al203. For the Ce/Al203
catalyst, the Ce02 <11 1> vs. AI2O3<440> intensity ratio was used because the
A1203<440> peak is distorted severely by the Ce02 <220> peak at high loading. Pure
CuO, Cr203, CeO2 or CuCr204 were finely grounded before measuring XRD intensities
for preparation of calibration curves (intensity ratio vs. weight percent of standard
compounds based on alumina). For maximum accuracy, the sample and the standard
compound peak widths (full width at half maximum) should be comparable. This method
assumes that metal oxides do not disrupt the y-Al203 spinel and do not affect the intensity
of the A1203 XRD line. The error in this method is estimated to be i 20 %.

Another quantitative XRD method compares the metal oxide (e.g. CuO) intensity
relative to the metal oxide intensity for physical mixtures of pure CuO and y-Al203. This
method does not use the A1203 peak as an internal standard. It assumes that the XRD

intensity is proportional to the mass of CuO. This method may be more accurate when

55
supported metal oxide absorbs X-ray significantly, such as at high loading of transition

metal or lanthanide element.

2. 2. 3. 3. Particle size determination

The width of diffraction peaks carries information on the dimension of the
reflecting planes. Perfect crystals produce very narrow diffraction lines. However, line
broadening occurs for small size crystallites (< 100 nm) due to incomplete constructive or
destructive interference. The mean crystallite Sizes (II) of the crystal particles can be

detemtined from XRD line broadening measurements using the Scherrer equation [18]:

d=KAchosO (15)

where it is the X-ray wavelength (Cu K“ = 0.154 nm), K is the particle shape factor, taken
as 0.9. The dimension d is defined as the effective thickness of the crystallite in a
direction perpendicular to the reflecting planes. 0n the basis of this definition there is only
a small dependence of K upon the crystallite shape. [3 is the full width at half maximum in
radians. O is the angle between the beam and normal on the reflecting plane (half of 26).
The detection limit of XRD technique depends on the X-ray source (target
material). According to the Bragg (14) and Scherrer equations (15), if one use X-ray with
smaller wavelength, the 3. and O terms become smaller. Therefore, smaller particles (d)
can be detected. For example, XRD patterns from smaller particles can be obtained using

a Ag (0.056 nm) or Mo (0.071 nm) target compared to Cu (0.154 am). Another way to

56
improve the XRD technique is to use synchrotron radiation source. The advantages are
high intensity of radiation, better signal to noise, shorter collection time and varying
wavelength of X-ray. The best resolved XRD patterns can be obtained by avoiding
incident radiation energy close to absorption edge of element, because the scattering
efficiency of an element decreases close to an absorption edge [17].

It must be noted that while X-ray line broadening measurements are useful as a
relative measure of average crystallite size, the Scherrer formula should not be used for
absolute determinations due to its failure to account for the effects of microstrain, the
consequences of particle size distribution, and the uncertainty in correcting for
instrumental broadening [19]. Hence, the values are meant to be used for relative

comparisons only.

2.3. Surface Characterization

2.3.1. Ultrahigh Vacuum Chamber

Electron spectrometers must operate under a vacuum of 10" Torr or lower. At
pressures higher than 10'6 Torr, the electrons would be scattered in the path from the
sample to the detector and the surface to be analyzed could be contaminated by gaseous
impurities. A reasonable criterion for proper working pressure would be that no more
than a few percent of an atomic layer of gaseous atoms should attach to the surface from

the gas phase during measurement.

57
From the simple kinetics theory of gases, the rate of arrival can be expressed as

follows [20],

 

 

}/ 22
2
[ RT J =351x10 P (16)

211M JTM

where r is particle flux (molecules cm'zs"), N is the number of gas molecules per cm3, R is
the gas constant, P is a pressure (Torr), T is a temperature (K), and M is a molecular
weight. If we assume that a monolayer consists of approximately 10ls atoms cm'2 on a
solid surface and the gas molecule sticking coefficient is 1, at least 109 Torr is necessary
to keep the surface clean for an hour.

The ultrahigh vacuum (UHV) chamber used in our work (Figure 2.6) is pumped by
a rough pump (Varian SD-90, 1.5 L/s pumping speed), a turbo molecular pump (Varian
V-80, pumping speed 80 L/s), and an ion pump (Perkin-Elmer 207-0230, pumping speed

220 US) in combination with a titanium sublimation pump [20]. The typical base pressure

of the chamber is normally approximately 1x103 Torr.

2.3.2. Transmission Electron Spectroscopy (TEM)

The TEM instrument is very similar to an optical microscope, if one replaces the

optical lenses by electromagnetic ones. TEM is able to observe the Size and shape of

58

sample manipulator

 

top view
X-ray source /
\ t
\x’;
analyzer x p . ,
prep chamber V 1”
\\ l;
‘1
side view
pumping port————> e
6
ion gauge ——>
o l t
. 1.1

 

 

 

 

 

4

ion pump and
tungsten sublimation pump

Figure 2.6. Schematic drawings of XPS chamber.

59

Primary electron beam

Secondary electrons

Auger electrons

X-rays Backscattered electrons

Photons

 

Diffracted electrons

 

i

Transmitted electrons

Figure 2.7. Diagram of electron beam interaction with a solid sample.

60
particles directly. In TEM, a primary electron beam of high energy and intensity passes
through condenser lenses to control the size of the beam that impinges on the sample,
which is subsequently magnified by the Objective lenses to produce a image. The
interaction of a high energy electron beam with a solid sample produces various electrons
and X-rays (Figure 2.7). The intensity of transmitted electrons (unscattered electrons),
which depends on the density and thickness of the sample, produces a two-dimensional
image of the atomic mass. Diffracted electrons formed by the elastic scattering of the
electron beam by the atoms in the specimen enable one to identify crystallographic phases
of the sample as well as dark field images. If an area of a sample is crystalline, scattering
occurs regularly, giving a pattern of sharp spots. Polycrystalline samples produce sharp
lines consisting of spots. If an area of a sample is amorphous, the pattern will contain
several diffuse rings. Backscattered electrons are formed from colliding electron beams
with atoms in the sample and scattering the electrons backward. Secondary electrons are
formed from several consecutive inelastic collisions. X-rays and Auger electrons are
formed from the relaxation of core ionized atoms and reveal information on the sample
composition. The emissions of photons ranging from UV to IR are mainly caused by
recombination of electron-hole pairs in the sample. Typical operating conditions of a
TEM instrument are 100-200 keV electrons, 106 Torr, 0.5 nm resolution and a

magnification of 105-10" [17].

2. 3. 2. 1. Instrumentationfm T EM
Figure 2.8 shows the TEM column used in this study. The electron gun chamber

commonly consists of a tungsten wire filament, shield, and anode.

61

\ High-voltage

cable

F Gun chamber

Condenser-lens

_ / aperture assembly

Specimen
/ exchange port

 

Condenser lens 1

 

Condenser lens 2

 

 

 

 

Cl 1 ) Ill

Objective lens an] Objective lens
aperture assembly

 

DIfffaCth“ lens Diffraction-lens

aperture assembly

 

 

Intermediate lens ,
Specrmen

Projector lens / traverse rods

 

 

 

 

0 O
Specimen viewing

f \ chamber

——i r—w

{1‘ —-—Plate camera port

 

 

 

 

Figure 2.8. TEM column (From Flegler, S. L.; Heckman, Jr. J. W.; Klomparens, K.
L. Scanning and Transmission Electron Microscopy: An Introduction;

W. H. Freeman and Company: New York, 1993.

62
The electrons produced by the heated tungsten filament are accelerated by the potential
difference between the filament and anode, which is normally 20 - 100 kV. In this study,
TEM images and selected area diffraction patterns were obtained with a JEOL 100CX2
using 120 kV primary voltage. The condenser-lens system is used to control electron
illumination on the sample and viewing screen. The first condenser lens can condense the
50 um electron beam, which is the diameter of electron beam at crossover in electron gun
chamber, to 1 um. The second condenser lens is used to control the beam brightness for
image viewing and photography. The objective lens, the diffraction lens, the intermediate
lens, and the projector lens all contribute to image magnification. The objective lens is the
first magnifying lens and the sample is inserted into it. The diffraction lens is used for
imaging diffraction patterns. The intermediate lens actually magnifies image by changing
intermediate lens current. The projector lens projects the final magnified the image onto
the viewing screen. Apertures are a piece of metal with a small hole, typically 30 - 1000
um, depending on their function. The various apertures are used to control the electron

beam diameter or to remove stray or widely scattered electrons [21].

2. 3. 2. 2. Sample Preparationfin' TIE/14

Generally, a thin copper grid (3 mm diameter, 50 pm thick) is used as a standard
sample support disk for TEM measurements. The powder form of a catalyst sample is
dispersed in water or organic solvent by an ultrasonic vibrator. In this work, methanol
was used as a liquid media to prepare samples. A drop of the suspension was deposited

on a holey carbon film coated copper grid and allowed to adsorb for about a minute. The

63
bulk of the solution was blotted off the edge of the drop with filter paper. The sample was

ready for viewing after drying excess solution in the air.

2.3.3. X-Ray Photoelectron Spectroscopy (XPS)

X-ray Photoelectron Spectroscopy (XPS or ESCA) is a powerfirl surface sensitive
techniques which provides chemical or physical information about surface atoms within
about 50-100 A thickness. From XPS measurements of heterogeneous catalysts, we can
determine the oxidation state of surface atoms, dispersion of catalyst on the support, and

quantitative analysis of all elements except H and He [22].

2. 3. 3. 1. Qualitative Analysis by XPS
The primary process of XPS is an ionization phenomenon brought about by
photons, the so-called photoelectric effect (Figure 2.9). A photon with energy (hv)

penetrates the surface and is absorbed by atoms. An electron with binding energy (Eb)

below the Fermi level escapes from the solid surface with a discrete kinetic energy (Ek).

Ek=hv-Eb-¢ (17)

where (b is the work fiinction of the spectrometer. From the binding energy, which is

characteristic for each element, the elements in a sample can be detected by XPS [23].

64

2|: _._O_Q_O_

23 L!
//

electrons

hv

hole

18 LOW

 

 

A+hv——a A++e'

Figure 2.9. Diagram of XPS Process.

65
2. 3.3.2. Quantitative analysis of XPS
XPS can be used as a quantitative tool for the determination of the chemical
composition of the surface region of a solid. We can determine the concentration of
impurities in a host material (e.g. supported catalyst) of known chemical composition.

The signal intensity (1,) observed in XPS can be expressed as follows [24],

11 =10111017~T(€I)D(81) (18)

where I0 is the X-ray flux, r], is the density of atoms of i, O, is the photo excitation
probability (photo ionization cross section), lfiei) is the mean free path (escape depth) of
an excited electron with kinetic energy (Si) in the host, and D(ei) is the fraction of

electrons detected by the analyzer (analyzer-detection efficiency). From the above

equation, we can find that the relative impurity vs. host material density as follows [25],

£111.): lieutenants.»

19
T1h I11917~T(81)D(€1) ( )

where n, and D11 are the density of impurity and host atoms. The relative energy
dependence of D(8h)/D(8i) can be determined (usually D(e) is proportional to 1/8) and

photoionization cross section (0’) have been reported by Scofield [26]. A method to

estimate escape depths in elemental solids and compounds has been reported by Penn [25].

66
A calculation of M48) for free-electron like materials (also valid for transition, noble, and

rare earth metals) is as follows,

Me) =[ 8 (20)

aane+bfl

where a and b depend on the electron concentration of the host material. However, in

order to obtain III/71h for eqn.(18), we need only the relative energy dependence

map/meg. Penn [25] has suggested a simple equation which is independent of a and b.

Met): 810“: '23) (21)
A1611) 82(ln81- 2'3)

 

 

This eqn.assumes (with i 5 "/0 error) that the mean free path ratio shows little or no
dependence on the density parameters a and b, but is dependent on eh and Si (for eh , a, 2

200 eV). Therefore, we can determine the relative impurity concentration.

2.3.3.3. Particle size determination

It has been shown by Defosse et al. [27] that one may calculate the theoretical
intensity ratio I°p/I°, expected for a supported phase (p) atomically dispersed on a carrier
(5). An extension of the Defosse model proposed by Kerkhof and Moulijn [28] has been

used in the present investigation. They proposed that catalysts consisted of sheets of

.1

I
I e

//////////////////////Z//// 2

WW I

I
I5
I

%////////////////////////////A 1+1 ..

Figure 2.10.

E. # of layers

It.

\\

 

Kerkhof and Moulijn's model of the supported catalysts.

68
support (dimension t) with cubic crystallites (dimension c) in between the layer structure
(Figure 2.10). The photoelectron cross sections and the mean escape depths of the
photoelectrons used in these calculations were taken from Scofield [26] and Penn [25]

respectively. In the case of a crystalline sample, the intensity ratio can be expressed as

i— s D(e, )o,2a(1 — e'B’ )

[1,] =(B-j ”(e,).,s,a_.-«)a..-B=) <22)
5 cryst b
where II, is the intensity of the electrons from element p in the catalyst, Is is the intensity of

the electrons from element 5 in the support, 0,, and OS are photoelectron cross sections,
(p/s)b is the bulk atomic ratio of the promoter and support, [31 and B2 are the dimensionless

support thickness, and D are the detector efficiencies which may be a function of the
kinetic energy (e) of the electrons. on is dimensionless crystallite size parameter which can

be calculated from following equation,
01 = c / A”, (23)

where A", is escape depth of electron from the element p in the catalyst traveling through
the element p.
In the case of a monolayer catalyst, or is small. Therefore the intensity ratio can be

expressed as follows,

69

I_, _(£) D(e,)o,(3,(1+e’9=) (24)
I mono- S 1’ D(£3)O,2(I'CTBZ)

B1 and [32 can be calculated from the following equations,
Bl=t/1,,, [La/1,, (25)

where A“ is the escape depth of electron from support traveling through the support. A,
is the escape depth of electron from the promoter traveling through support.

We can calculate the crystalline size of supported materials with the following eqn.
and eqn. (24), if it", is known. The following eqn. is obtained from eqn. (22) and eqn.

(23).

U

1‘ I
5 CM, (1— e‘“)

= (26)

s mono

where (Ip/I‘)m Is a experimental data, (IP/Is)mono is a theoretical monolayer Intensrty ratio

 

F—II ~
'3

—1

calculated from eqn. (24).

70
2.3.3.4. HS shake up satellite peaks

Shake-up processes arise when the photoelectron excites an outer electron to a
bound excited state in the atom or molecule. The photoelectron loses an amount of
kinetic energy corresponding to the excitation and that appears at a higher binding energy
in the spectrum. The origin of the satellite peaks is considered to be the promotion of 3d
electrons to 45 and/or 4p levels [29-32] or alternatively to be a charge transfer of ligand
electrons to unfilled 3d orbitals (e.g. 0 2p —> Cu 3d in case of CuO) [33-38]. Such
transitions are not seen for Cull or Cu0 compounds. It is well known [39] that transition
metal ions with unfilled 3d orbitals and completely empty 3d orbitals show satellite peaks
due to shake-up, that the satellite peaks are not observed when the 3d orbitals are
completely filled (Zn2+ and CuT), and that the transition is a monopole excitation,
conserving molecular symmetry and parity. Taking into consideration the above facts, the
explanation based on the ligand charge transfer seems more reasonable than that based on
the 3d —> 4s, 4p transitions. In this study, we discussed the shake up satellites associated
with the Cu(2p3/2) level on the basis of the monopole charge transfer mechanism. Thus,
the satellite Structure provides information about the bonding nature of Cuzi ion in the
catalyst.

In order to estimate the nature of the distorted copper oxide structure, a
correlation between the satellite intensity of the Cu(2p3/2) level and the symmetry of the
Cu2+ ion is required. The detailed structure of the satellite peaks depended on the catalyst
composition. Frost et al. [29] reported that Cu2+ ion in a distorted octahedral symmetry

Cu(acac.)H2O, Cu(NO3)23H20, Cu(OAc)2, CuC03 and Cu(OH)2 show generally stronger

71
and well resolved satellite peaks compared with those in a distorted square planar
symmetry (CuO, CuCl2, CuBr2). Strohmeier et al. [40] reported that the Cu 2p3/2 satellite
structure of bulk CuAl204 (Cu ions in 60 % T4 and 40 % 02, sites) was more intense than
that of CuO. The surface Cu2+ ions in CulAl203 catalysts at low loading (distributed 90 %
in a tetragonally distorted octahedral environment and 10 % in tetrahedral environment)
exhibit satellite peaks similar to CuO rather than CuAl204 [40,41,42]. It has been
suggested [40] that this did not necessarily mean that well dispersed or amorphous CuO
was present on the catalysts, but simply that the copper oxide surface spinel species is

chemically different from CuO and CuAl204.

2.4. Catalytic Activity Measurement

The catalytic reactor system built for this study is shown in Figure 2.11.
Measurement of C0 and CH4 oxidation activity was performed in a flow microreactor.
Approximately 30 and 100 mg of catalysts was supported on a glass frit (70 - 100 pm) for
C0 and CH4 oxidation, respectively. The temperature was measured with a K-type
thermocouple located just above the catalyst bed. The reactor was heated by a tube
fumace (Lindberg) with temperature being controlled within 1 °C by an Omega CN 1200
temperature controller. Reactant gas flow rates were held constant with Brooks 5850
mass flow controllers. Product gases were analyzed with a Varian 920 gas chromatograph
equipped with a TCD and interfaced to a Hewlett-Packard 3394A integrator. Samples

(0.5 and 5 cc) were collected by Valco sample loops and electric actuator for C0 and CH4

72

 

 

 

 

Zeolite

 

 

 

 

 

Mass Flow
Controllers

He

02
C0/02/He
CHJOJHe

 

 

 

 

 

 

 

 

 

 

 

\>

 

 

 

Bellows Valves

 

 

Controller

 

 

Temperature 1 Union Tee

 

 

 

 

Furnace :

 

Thermocouple
Glass Fritz

 

 

 

Vent

 

 

 

 

 

 

 

 

 

 

 

 

3-way Valve

 

 

 

 

 

Sample
Loop

Water Trap

 

Inte grater

Figure 2.11. Schematic of the catalytic reactor system.

73
oxidation, respectively. Reaction products were separated on a 6 ft 60/80 mesh
Carbosieve S-II column. Reaction temperature control, sample injection into GC,
integration of GC response were controlled automatically by a controller constructed in
house. Prior to the first activity measurement, the catalyst was pretreated with a mixture
of 5% O2/He (99.5 % purity for 02, 99.995% purity for He, AGA Gas Co.) stream (143
cc/min) at 350°C for 1 hr to remove any impurities adsorbed on the surface during catalyst
preparation and storage. CO oxidation reactions were performed with a 80 cm3/min flow
of 4.8% CO/9.8% 02/85.4% He gas mixture (AGA, purity > 99.99%) at 100-310°C.
Methane oxidation reactions were performed with a 15 cm3/min flow of 0.98%
CHI/5.25% 02/93.77% He gas mixture (AGA, purity > 99.99%) at 310-450 °C. All
activity measurements were obtained under steady-state conditions at conversions less
than 15%. Water produced during methane oxidation was frozen downstream from the
reactor in a trap maintained below -40 °C with a mixture of ethanol and dry ice. The TCD
response factor for each molecule was obtained using a standard gas mixture (Scotty II
Analyzed Gases 1% CH4, C0, C02, H2, 02 in N2, Supelco). The relative thermal response
factors for C0 and CH4 based on CO2 are 1.14 and 1.33 respectively, which agree well

with the reported values [43].

IO.

11.

12.

l3.

14.

15.

74

2.5. References

Knozinger, H.; Ratnasamy, P. Catal. Rev. Sci. Eng. 1978, 17, 31.

Satterfield, C. N. Heterogeneous Catalysis in Practice; McGraw-Hill: New York,
1980.

Doesburg, E. B. M. and van Hooff, J. H. C. In Catalysis: An Integrated Approach
to Homogeneous, Heterogeneous and Industrial Catalysis; Mouli j n, J. A.; van

Leeuwen, P.W.N.M.; van Santen, RA. Ed.; Elsevier: Amsterdam, 1993; pp 309.

Nortier, P.; Soustelle, M. in “Catalysis and Automotive Pollution Control”,
Crucq, A; Frennet, A. Eds. Elsevier, Amsterdam 1987 pp275.

Benesi, H. A.; Curtis, R. M.; Studer, H. P. J. Catal. 1968, 10, 328.

Scierka, S. J .; Houalla, M.; Proctor, A.; Hercules, D. M. J. Phys. Chem. 1995, 99,
1537.

Brunelle, J. P. in “Preparation of Catalysts II” Delmon, B.; Grange, P.; Jacobs, P.,
Poncelet, G., Ed. Elsevier, Amsterdam pp 21 1, 1979.

Kijenski, J.; Baiker, A.; Glinski, M.; Dollenmeier, P.; Wokaun, A. J. Catal. 1986,
101, 1.

Okabe, K.; Sayama, K.; Matsubayashi, N.; Shimomura, K; Arakawa, H. Bull.
Chem. Soc. Jpn. 1992, 65, 2520.

Brunauer, S.; Deming, L. S.; Deming, W. S.; Teller,E. J. Am. Chem. Soc, 1940,
62, 1723.

Brunauer, S.; Emmett, P.H.; Teller,E. J. Am. Chem. Soc, 1938, 60, 309.
Langmuir, I. J. Amer. Chem. Soc. 1916, 38, 2221.
Manual onuantasorb Jr. Quantachrome 1985.

Drago, R. S. “Physical methods for Chemists” 2nd edition, Saunders College
Publishing, New York 1992.

Delgass, W. N.; Haller, G. L.; Kellerman, R.; Lunsford, J. H. “Spectroscopy in
Heterogeneous Catalysis” Academic Press, New York 1979.

16.
1’7.

18.

19.

20.

21.

22.

23.

24.

25.
26.
27.
2.8-
29_
30-
31.
32.

33.

75

Bragg, W. L. Proc. Camb. Phil. Soc. 1912, 17, 43.
Niemantsverdriet, J. W. “Spectroscopy in Catalysis” VCH, Weinheim, 1993.

Klug, H. P.; Alexander, L. E., X -Ray Diffraction Procedures for Polycrystalline
and Amorphous Materials, lst.Ed. Wiley, New York, 1954.

Cohen, J. B. Ultramicroscopy 1990, 34, 41.

O’Hanlon, J. F. “A User’s Guide to Vacuum Technology” 2nd. edition, Wiley,
New York, 1989.

Flegler, S. L.; Heckman, Jr. J. W.; Klomparens, K. L. Scanning and Transmission
Electron Microscopy: An Introduction; W. H. Freeman and Company: New York,
1993.

Hercules, D. M.; Hercules, S. H. J. Chem. Educ. 1984, 61, 483.

Briggs, D. and Seah, M. P. Practical Surface Analysis by Auger and X-Ray
Photoelectron Spectroscopy, 2nd ed.; Wiley: New York, 1990.

Fadley, C. S., Baird, R. J., Siekhaus, W., Novakov, T., and Bergstrom, S. A. L. J.
Electron. Spectrosc, 1974, 4, 93.

Penn, D. R. J. Electron. Spectrosc. Relat. Phenom. 1976, 9, 29.

Scofield, J. H. J. Electron. Spectrosc. Relat. Phenom. 1979, 8, 129.

C. Deffosse, D. Canesson, P. G. Rouxhet and B. Delmon, J. Catal. 51(1978)269.
Kerkhof, F. P. J. M. and Moulijin, J. A. J. Phys. Chem. 1979, 83, 1612.

Frost, D. C.; Ishitani, A.; McDowell, C. A. Mo]. Phys. 1972, 24,861.
Rosencwaig, A. J. Phys. Rev.Lett. 1971, 27, 479.

Yin, L. Chem. Phys. Lett. 1974, 24, 81.

Fiermans,L. Surf Sci. 1975,47, l.

Hufner, S. Phys. Rev.B. 1973, 7, 5086.

34.

35

36.

3'7.

38.

39.

40.

41,

42.

76

Kim, K. S. J. Electron Specsc. Relat. Phenom. 1974, 3, 217.
Kim, K. S. J. Electron Specsc. Relat. Phenom. 1972 1, 253.
Wallbank, B. J. Electron Specsc. Relat. Phenom. 1974, 5, 259.
Wallbank, B. J. Phys. C. 1973, 6, 493.

Carlson, T.A. J. Electron Specsc. Relat. Phenom. 1974, 5,247.
Carlson “Photoelectron and Auger Spectroscopy” 1975.

Strohmeier, B. R.; Leyden, D. E.; Field, R. S.; Hercules, D. M.; Petrakis, L. J.
Catal. 1985, 94, 514.

Friedman, R. M.; Freeman, J. J.; Lytle, F. W. J. Catal. 1978, 55, 10.

Tikhov, S. F.; Sadykov, V. A.; Kryukova, G. N.; Paukshtis, E. A.; Popovskii, V.
V.; Starostina, T. G.; Kharlamov, G. V.; Anufrienko, V. F.; Poluboyarov, V. F.;
Razdobarov, V. A.; Bulgakov, N. N; Kalinkin, A. V. J. Catal. I992, 134, 506.

McNair, H. M.; Bonelli, E. J. “Basic Gas Chromatography” Varian Aerograph,
1969.

Chapter 3

The Effect of Crystallinity on the Photoreduction of Cerium Oxide:
A Study of CeOz and Ce/AIZO3 Catalysts

3.1. Abstract

The effect of crystallinity on the photoreduction of cerium oxide (Ce4+ to Cesi)

d uring X-ray photoelectron spectroscopy (XPS) analysis of CeOz and Ce/A1203 catalysts
has been determined. Cerium oxide samples were prepared by calcining cerium (IV)
methoxyethoxide in air at different temperatures. The structure of the resulting metal
Oxides was determined using X-ray diffraction (XRD) and transmission electron
microscopy (TEM). The oxide obtained from calcination at 200 °C consisted primarily of
amorphous cerium oxide, while the material produced at 750 °C contained large CeOz
cI3/stallites. XPS Ce 3d spectra obtained for the samples showed that the amorphous
cerium oxide sample was reduced more extensively than the crystalline material during
XPS analysis. Ce/A1203 catalysts (5.1 wt.% CeOz) were prepared by incipient wetness
impregnation using ammonium Ce(IV) nitrate and Ce(IV) methoxyethoxide precursors.
){RD and XPS analyses of the Ce/A1203 catalyst derived from the nitrate precursor
(designated “CeN”) indicated that most of the cerium was present as poorly dispersed

CeO; crystallites. The catalyst prepared using the alkoxide precursor (designated “CeA”)

77

78

also contained poorly dispersed cerium oxide; however, XRD data suggested that very
little of the cerium oxide phase was crystalline. The XPS Ce 3d spectrum obtained for the
CeN catalyst was typical of Ce(IV) oxide while the spectrum collected for the CeA
catalyst was similar to a Ce(III) compound. These results have been attributed to

enhanced photoreduction of the amorphous cerium oxide present in the CeA catalyst.

3.2. Introduction

Cerium oxide is an important component of superconductors [1,2], thin film
0 ptical devices [3,4], ceramics [5], gas sensors [6], and heterogeneous catalysts [7-13]. In
In any applications, knowledge of the cerium oxidation state is necessary for understanding
‘2 he properties of the material. For example, Tranquada et al. [1] have reported that the
d etermination of cerium valence in cerium doped copper oxide superconductors is
i m portant for understanding the details of electron-pairing in these systems. The oxygen
Storage capabilities of CeOz/A1203 catalysts are commonly attributed to reversible changes
i r) the cerium oxidation state [7-9, 12].
X-ray photoelectron spectroscopy (XPS) is frequently used to determine the
Oxidation state of surface species. However, it is well known that many types of
Compounds are not stable under the conditions used for XPS measurements [14,15].
Several factors such as X-ray flux [16], integrated X-ray dose [17], secondary electrons
from the X-ray source [18], sample charging [19], temperature [20], and high vacuum

[21] are known to be responsible for sample damage. Cerium oxide photoreduction

- 79
during XPS measurements has been reported by several researchers [19,22-25].

Paparazzo et al. [24,25] attributed cerium oxide photoreduction primarily to intense
heating of the sample surface. In contrast, Laachir et al. [26] found that CeOz samples
vvith both low and high surface area (5 mZ/g and 115 m2/g, respectively) were stable for up
to four hours during XPS analysis. Dausher et al. [22] reported that a Ti02-Ce02 mixed
oxide prepared using the sol-gel method was more reduced during XPS measurement than
t he physical mixture of the two oxides. These seemingly inconsistent results may be due
t O a difference in the structures of cerium oxides analyzed.
We are interested in the surface chemistry of cerium oxide materials that can be
u sed as chemical sensors [27] and heterogeneous catalysts [28]. XPS is widely utilized to
d etermine the effect of preparation method and sample pretreatment on the distribution of
c erium oxidation states in such materials [10,13,29,30]. 1E] Fallah et a1. [19] demonstrated
1: hat ceria photoreduction could be limited by controlling X-ray beam power, irradiation
t i me, and sample charging. However, extended analysis times are sometimes necessary to
obtain reasonable quality spectra for materials with low cerium contents. This may cause
a Change in cerium oxidation state and invalidate the results. Therefore, it is important to
kn ow which types of materials are susceptible to photoreduction during XPS analysis. In
this study, we have examined the effect of crystallinity on the photoreduction of cerium
Oxide. Amorphous and crystalline CeOz materials were prepared by calcining cerium(IV)
methoxyethoxide in air at 200 °C. and 750 °C, respectively. X-ray diffraction (XRD) and
transmission electron microscopy (TEM) have been used to determine the structure of

these oxide materials. In addition, Ce/A1203 catalysts were prepared using ammonium

8O
Ce(IV) nitrate and Ce(IV) methoxyethoxide precursors. XRD and XPS have been used to

determine the crystallinity, dispersion, and oxidation state of the alumina supported cerium
oxides. The structural information obtained for CeOz samples and Ce/A1203 catalysts has
been correlated with the extent of cerium photoreduction observed during XPS

measurements.

3.3. Experimental

Standard Materials. Cerium(III) acetylacetonate hydrate was obtained from
Aldrich Chemical Company, Inc. CeOz was prepared by calcining ammonium cerium(IV)
nitrate (Analytical Reagent, Mallianrodt, Inc.) in air at 800 °C for 16 h. The XRD
pattern of the CeOz standard compound matched the appropriate Powder Diffraction File
E: 3 l].

Cerium Oxide Sample Preparation. Cerium oxide samples were prepared by
allowing cerium(IV) methoxyethoxide solution (18-20 % cerium alkoxide in

methoxyethanol; Gelest, Inc.) to dry on quartz slide substrates at room temperature for 5

days in a nitrogen purged glove bag. These samples were subsequently calcined in air at

200 °C and 750 °C for 36 h and 16 h, respectively. TGA and FTIR analyses indicated that

Cerium methoxyethoxide decomposition was complete afier calcination steps. Samples
prepared at 200 0C and 750 °C are designated as Ce200 and Ce750, respectively.

Catalyst Preparation. Catalysts were prepared by pore volume impregnation of y-

alumina (Cyanamid, y-A1203, surface area = 203 mz/g, pore volume = 0.6 mL/g). The

81
alumina support was finely ground (< 230 mesh) and calcined in air at 500 °C for 24 h

prior to impregnation. Catalysts were prepared using a deionized water solution of
ammonium cerium(IV) nitrate (Mallinckrodt Inc., Analytical Reagent) or an ethanol
solution of cerium(IV) methoxyethoxide (Gelest Inc., 18-20 "/0 cerium alkoxide in
methoxyethanol). The catalyst derived from the cerium nitrate precursor (designated
“(IeN”) was dried in air at 120 °C for 24 h and calcined in air at 500 °C for 16 h. The
catalyst prepared using the cerium alkoxide precursor (designated “CeA”) was

impregnated and subsequently dried at room temperature for 48 h in a N2 purged glove

1) ag prior to firrther drying in air at 120 °C for 24 h and calcination in air at 500 °C for 16
h - The Ce loading was held constant at 5.1 wt.% CeOz.

BET Surface Area. Catalyst surface areas were determined using a QuantaChrome
Quantasorb Jr. Sorption System. Approximately 0.1 g of a catalyst sample was outgassed
i n a Ny/He mixture (5% N2) at 350 °C for 1 h prior to adsorption measurements. The
measurements were made using relative pressures of N; to He of 0.05, 0.08, and 0.15 (N2
811 rface area = 0.162 nmz) at 77 K.

Electron Microscopy. Transmission electron microscopy (TEM) images and
Selected area electron diffraction (SAED) patterns were obtained with a JEOL 100CX2
using 120 kV primary voltage. Cerium oxide samples were ultrasonically dispersed in

methanol and deposited on a holey carbon film supported on a conventional copper grid
(3 00 mesh, 3 mm).
X-Ray Diffraction. X-ray powder diffraction patterns were obtained with a

Rigaku XRD diffractometer employing Cu Kat radiation (2» = 1.5418 A). The X-ray was

82
operated at 45 kV and 100 mA. Diffraction patterns were obtained using a scan rate of

0.25 deg/min. with divergence slit and scatter slit widths of 1°. Cerium oxide samples
were run as powders mounted in the cavity of a circular silicon sample holder (C-12, Dow

Corning). Catalysts were run as powders packed into a glass sample holder having a 20 x

1 6 x 0.5-mm cavity. The mean crystalline sizes ((3 ) of the CeOz particles were determined

from XRD line broadening measurements using the Scherrer equation [32]:

d=IOt/Bc059 (1)

Where is. is the X-ray wavelength, K is the particle shape factor, taken as 0.9, and B is the
fill width at half maximum (fwhm), in radians, of CeOz <1 11> and <220> lines measured
for Ce/A1203 and CeOz samples, respectively.

Quantitative X-ray diffraction data for Ce/A1203 catalysts were obtained by

comparing Ce02 <1 1 1>/A1203<440> intensity ratios measured for catalyst samples with

peak ratios measured for physical mixtures of CeOz standard and y-A1203 support. This

method assumes that cerium addition does not disrupt the y-A1203 spinel and subsequently

afiect the intensity of the A1203 peak. The error in this method was estimated to be i 20
0/0 .

XPS Analysis. XPS data were obtained using a Perkin-Elmer Surface Science

instrument equipped with a magnesium anode (1253.6 eV) operated at 300 W (15 kV, 20

mA) and a 10-360 hemispherical analyzer operated with a pass energy of 50 eV. The

3:?

V]

(In

I“)

83
instrument typically operated at pressures near 1 x 10'8 Torr in the analysis chamber. The

take-off angle between the sample and analyzer was approximately 45° for all
measurements. The distance between the X—ray source and sample was maintained at
approximately 3/4 inch for all analyses. Spectra were collected using a PC137 board
interfaced to a Zeos 386SX computer. CeOz samples were analyzed as prepared while
catalyst samples were analyzed as powders dusted onto double-sided sticky tape or spray
coated onto a quartz slide using a methanol suspension of the catalyst. XPS binding
energies were measured with a precision of :t 0.2 eV, or better. Binding energies for the
st andard samples were referenced to the C ls line (284.6 eV) of the carbon overlayer.
S i mce the charging observed for the Ce200 and Ce750 samples changed during our
experiments, we were concerned that the C Is charge reference may be less accurate than
a Ce 3d derived charge correction. Thus, we have charge corrected the Ce 3d spectra
Obtained for the cerium oxide samples using the Ce 3d u'" peak (916.7 eV). Using the Ce
3 d u'" reference energy, the binding energy of the C 15 peak measured for the CeOz
Standard was 284.6 eV. The Ce 3d u'" peak has been commonly used to compensate for
Sample charging in the ceria photoreduction studies [19,24,25]. Binding energies for the
catalyst samples were referenced to the Al 2p peak (74.5 eV) of the alumina support. The
Ce 3d spectra of Ce200 and Ce750 samples were obtained at 15 min. intervals up to 300
min. The Ce 3d region of CeN and CeA catalysts were measured for 600 min. to obtain
reasonable quality spectra. The extent of cerium reduction was determined using

composite spectra [33] derived from the Ce 3d spectra measured for CeOz and Ce(III)

84

acetylacetonate hydrate standard compounds. We estimate the error in this method is less
than i 5 %.

Quantitative XPS Analysis. It has been shown by Defosse et al. [34] that one may
calculate the theoretical intensity ratio I°p/1°. expected for a supported phase (p) atomically
dispersed on a carrier (5). An extension of the Defosse model proposed by Kerkhof and
Moulijn [35] has been used in the present investigation. The photoelectron cross sections
and the mean escape depths of the photoelectrons used in these calculations were taken
from Scofield [36] and Penn [37], respectively. For a phase (p) present as discrete

particles, the experimental intensity ratio (Ip/Is) is given by the following expression:

Ip/r, = 1°,,/1°.[1-exp(-d/x,)]/ cm.,, (2)

where (I°,/I°,) is the theoretical monolayer intensity ratio, d is the length of the edge of the
cubic crystallites of the deposited phase and 3.? is the mean escape depth of the

photoelectrons in the deposited phase.

3.4. Results and Discussion

XPS Analysis of Standard ( 'omponnds. The XPS Ce 3d spectra of cerium
compounds are known to be complicated due to hybridization of the Ce 4f with ligand

orbitals and fractional occupancy of the valence 4f orbitals [38-44]. Figure 3.1 shows the

85

 

 

“ w " \

(b)

 

 

 

 

 

 

 

 

 

l l l

 

920 900 880
Binding Energy (eV)

Figure 3.1. Ce 3d XPS spectra measured for (a) CeO2 and (b) Ce(III)
acetylacetonate hydrate.

 

86
Ce 3d spectra measured for CeOz and Ce(III) acetylacetonate hydrate standard

compounds. The peak shapes are similar to those reported for Ce(IV) and Ce(III)
compounds [10,29,45,46]. The Ce 3d spectrum measured for CeOz contains three main
3d5/2 features at 882.6 eV (v), 889.3 eV (v"), and 898.6 eV (v"') and three main 3d3,2
features at 900.8 eV (u), 907.8 eV (u"), and 916.7 eV (u'"). The high binding energy
doublet v'"(u"') has been assigned to a final state with primarily 4f“ configuration. The v"
and v (u" and u) doublets are generally assigned to final states with strong mixing of 40
and 4f2 configurations. These states arise from the core hole potential in the final state and
4f hybridization in the initial state [38-43]. The Ce 3d spectrum measured for Ce(III)
acetylacetonate hydrate contains two main 3d5,2 peaks at 881.4 eV (v) and 885.2 eV (v')
and two main 3d3a features at 899.5 eV (u) and 903.6 eV (u'). These doublets have been
assigned to final states where 4fl and 4f2 configurations are strongly mixed [44].

Structure of Ce200 and Ce750. TEM images and SAED patterns of Ce200 and
Ce750 samples are shown in Figure 3.2. The micrograph obtained for the Ce200 sample
consists of thick agglomerates with no preferred shape or orientation; however, small
grains (~ 3 nm diameter) are observed at the edges of agglomerates. The SAED pattern
of the Ce200 sample shows seVeral diffuse rings indicative of an amorphous phase. The
TEM image obtained for the Ce750 sample shows faceted particles. The SAED pattern
shows several sets of sharp spots characteristic of a crystalline phase.

Figure 3.3 shows the XRD patterns obtained for Ce200 and Ce750 samples. The
most intense CeOz peak (CeOz <1 1 1>) was excluded from this figure due to overlap with

a broad silica peak from the sample substrate. The XRD pattern obtained for the Ce200

87

 

(a)

88

 

(b)

 

Figure 3.2. TEM micrographs and SAED patterns obtained for (a) Ce200 and (b)
Ce750 samples.

89

 

 

<220>
<311>

1000

 

 

<222>

JL J (b)

 

 

 

 

 

 

100

 

 

l l l

 

4O 50 60
Degrees (20)

Figure 3.3. XRD patterns measured for (a) Ce200 and (b) Ce750 samples. ~

gmdes
thank
pads l

C6505

REE

m

Oi

Im

90

sample shows broad, weak peaks characteristic of poorly crystallized CeOz. The
diffraction pattern measured for the Ce750 sample shows more intense, sharper CeOz
peaks. In addition, the CeOz <222> peak can be observed in the diffraction pattern of the
Ce750 sample. CeOz particle sizes determined using XRD line broadening calculations are
significantly larger for the Ce750 sample (34.8 nm) than for the Ce200 sample (2.2 nm).
For the Ce200 sample, the XRD determined CeOz particle size is similar to the grain size
observed in TEM.

In summary, TEM and XRD results indicate that calcination of cerium alkoxide
leads to the formation of cerium oxide. The extent of crystallinity is affected by the
calcination temperature. The Ce200 sample consists primarily of large, amorphous cerium
oxide particles that contain small CeOz crystallites. The Ce750 sample consists primarily
of large CeOz crystallites. No evidence for the presence of an amorphous cerium oxide
phase has been observed for the Ce750 sample.

Photoreduction of Ce200 and C e 750. Figure 3.4 shows the Ce 3d spectra
measured for the Ce200 sample as a function of irradiation time. The spectrum obtained
in the first 15 min. of analysis has features similar to those observed for the CeOz standard;
the Ce 3d5/2 binding energy (882.7 eV) is identical, within experimental error, to the value
measured for Ce02 (882.6 eV). As the analysis time increases, new features characteristic
of Ce3‘ appear in the Ce 3d spectra. For intermediate irradiation times (30 and 60 min),
this is most evident as a loss of resolution between the v and v" peaks of Ce4+ due to the
growth of the v' peak characteristic of C e3 ’. We believe that the new features appearing in
the Ce 3d spectrum are due to the photoreduction of Ce4+ to Ce‘ii during the XPS analysis.

This is consistent with numerous reports in the literature [19,22-25]. We also note that

91

 

 

(f)

(d)

(C)

(b)

(a)

l l l

 

920 900 880
Binding Energy (eV)
Figure 3.4. Ce 3d XPS spectra measured for the Ce200 sample after

irradiation times of (a) 15 min, (b) 30 min, (c) 60 min,
(d) 120 min., (e) 180 min, and (f) 300 min.

 

92
increasing analysis time broadens the Ce 3d5/2 region on the low binding energy side.

Paparazzo [24] has reported that such broadening can be attributed to the superposition of
“Ce203” components and CeOz features. The spectrum measured afier 300 min. clearly
shows features characteristic of Ce3+ and Ce“. The Ce 3d5,r2 binding measured for the
Ce3+ v' component (885.4 eV) in the irradiated sample is close to values measured for
Ce(III) acetylacetonate hydrate (885.2 eV) and reported for Ce203 (885.3 eV [25] and
885.8 eV [47]) compound. In addition, Le Normand et al. [48] reported that a v' peak
(885.5 eV) appeared in the XPS Ce 3d spectra measured for C602 reduced in H2 at 900
°C.

The XPS Ce 3d spectra measured for the Ce750 sample are shown as a fimction of
irradiation time in Figure 3.5. One can clearly see that these spectra also change with
time; however, the growth of Ce?" features is less dramatic than those observed for Ce200
sample. The XPS Ce 3d binding energies measured for the Ce750 sample are identical,
within experimental error, to the values obtained for the Ce200 sample. For the Ce750
sample, well resolved v and v" peaks are observed for the duration of the experiment.

Figure 3.6 shows the extent of cerium reduction with respect to irradiation time for
Ce200 and Ce750 samples. As anticipated from Figures 3.4 and 3.5, the extent of
reduction increases rapidly with time up to approximately 180 min. For longer analysis
times, the change in fraction reduced is less pronounced. This indicates that the rate of
reduction decreases with irradiation time. Similar results have been reported by Wallbank
et al. [49] in the case of Cqu photoreduction (Cu2+ to Cu"). Figure 3.6 also shows that

the cerium oxide calcination temperature affects the extent of reduction. - As noted above,

93

 

 

(f)

(6)

(d)

(C)

(b)

(a)

l l l

 

920 900 880
Binding Energy (eV)

Figure 3.5. Ce 3d XPS spectra measured for the Ce750 sample after
irradiation times of (a) 15 min, (b) 30 min., (c) 60 min.,
(d) 120 min, (e) 180 min, and (f) 300 min.

 

94

35

 

30- O O

25- O

20- O

 

15- 0

Fraction Reduced (%)
O
O
O

10- 0

54a

 

 

 

O-‘I I I 1 1 l I I I r I I I I I I I
O 50 100 150 200 250 300 350 400 450

Irradiation Time (min)

Figure 3.6. Fraction of CeOz reduced as a function of irradiation time
for the Ce200 (O) and Ce750 (0) samples.

95
TEM and XRD analyses have shown that the Ce200 sample consists of small CeOz

crystallites imbedded in large amorphous cerium oxide particles, while the Ce750 sample
contains only large CeOz crystallites. Thus, we believe the photoreduction results indicate
that amorphous cerium oxide is more readily reduced than crystalline CeOz during the
XPS measurement.

Structure of Cerium/Alumina Catalysts. The surface areas of CeN and CeA
catalysts (195 mZ/g) are identical and slightly lower than the value obtained for the A1203
support (203 mz/g). Figure 3.7 shows the )GKD patterns obtained for CeN and CeA
catalysts. The XRD pattern of the CeN catalyst shows intense peaks characteristic of
CeOz; however, only weak CeOz lines are observed in the diffraction pattern of the CeA
catalyst. Quantitative XRD measurements indicate that the CeN catalyst contains 4.6
wt.% crystalline CeO; phase, while the CeA catalyst contains only 0.3 wt.% crystalline
C602.

Variation in the particle size of the cerium oxides determined from XRD line
broadening calculations and XPS Ce/Al intensity ratios are shown in Table 3.1 for both
CeN and CeA catalysts. The CeOz particle size determined for the CeN catalyst using
XRD is slightly larger than the value obtained for the CeA catalyst. However, the C602
particle size calculated from XRD analysis of the CeA catalyst must be considered as
approximate due to the low intensity of the CeOz <1 1 1> peak. The cerium oxide particle
sizes determined for CeN and CeA catalysts using XPS intensity ratios are identical and

smaller than the values obtained from XRD.

96

 

 

 

f r r F
' CeO2
C602 <200>
(c) w <111>
(b)
A1203
<311>
A1203
(a) <220> A1203
<222>
r r r I
25 30 35 40

Degrees (20)

Figure 3.7. XRD patterns measured for (a) alumina support and (b)
CeA and (c) CeN catalysts.

 

97

In summary, XRD and XPS analyses indicate that the cerium oxide dispersion is
similar for catalysts derived from nitrate (CeN) and alkoxide (CeA) cerium precursors.
However, the CeN catalyst consists primarily of CeOz crystallites while most of the cerium

oxide present in the CeA catalyst is amorphous [28].

Table 3.1. Particle Sizes of Cerium Oxide Species in Ce/A1203
Catalysts Determined from m Line Broadening
and XPS Intensity Ratios Measurements.

 

 

mm
Samples l XRD XPS
CeN 5.1 1.6
CeA 4.5 1.6

 

 

Photoreduction of Cerium/A lununa Catalysts. Figure 3.8 shows the Ce 3d spectra
obtained for CeN and CeA catalysts. The XPS Ce 3d spectrum measured for the CeN
catalyst shows Ce“ and Ce‘i’ features similar to those observed for the photoreduced
Ce750 sample. For the CeA catalyst, the Ce 3d spectrum has v' and u' lines similar to
those observed for the Ce(IIl) acetylacetonate hydrate standard compound as well as the
u'" peak associated with Ce" species.

The presence of Ce3‘ in calcined CC/Al203 catalysts is ofien attributed to the
formation of bulk CeAlO; or a Ce-A1203 surface phase. We do not believe that the CeN
and CeA catalysts examined in this study contain either of these species since they were

prepared using Ce“ precursors and were calcined at a relatively low temperature. Shyu et

98

 

(b)

 

 

r 1 1
920 900 880

 

Binding Energy (eV)

Figure 3.8. Ce 3d XPS spectra measured for (a) CeN and (b) CeA catalysts.

99
al. [29] have reported that high temperature (1000 °C), long reaction times (80 h), and a

reducing atmosphere were required to form bulk CeAlO3 in Ce/A1203 catalysts.
Formation of a Ce3i—A1203 surface phase would seem to require a Ce“ precursor.
Further, the relatively poor Ce dispersion noted for CeN and CeA catalysts is not
consistent with the formation of a Ce-A1203 surface phase since surface phases are
normally considered to be highly dispersed (particle size < 1 nm). Thus we believe that
the Ce3+ species observed by XPS arise from cerium oxide photoreduction. The greater
extent of photoreduction shown by the CeA catalyst can be attributed to the predominance

of amorphous cerium oxide in the catalyst.

10.

ll.

12.

13.

I4.

15.

16.

17.

100

3.5. References

Tranquada, J. M.; Heald, S. M.; Moodenbaugh, A. R.; Liang, G.; Croft, M. Nature
1989, 337, 720.

Merchant, P.; Jacowitz, R. D.; Tibbs, K; Taber, R. C.; Landerrnan, S. S. Appl.
Phys. Lett. 1992, 60, 763.

Al-Robaee, M. S.; Rao, K. N.; Mohan, S. J. Appl. Phys. 1992, 7!, 2380.

Al-than, Z. T.; Hogarth, C. A.; Riddleston, N. Phys. Stat. Sol. (b) 1988, 145,
145.

Kleebe, H.-J.; Cinibulk, M. K. J. Mater. Sci. Lett. 1993, 12, 70.

Butta, N.; Cinquegrani, L.; Mugno, E.; Tagliente, A.; Pizzini, S. Sensors and
Actuators B, 1992, 6, 253.

Su, E. C.; Montreuil, C. N.; Rothschild, W. G. Appl. Catal. 1985, 17, 75.
Herz, R. K. Ind. Eng. Chem. Prod. Res. Dev. 1981, 20, 451.
Yao, H. C.; Yu-Yao, Y. F. J. Catal. 1984, 86, 254.

Le Normand, F.; Hilaire, L.; Kili, K.; Krill, G.; Maire, G. J. Phys. Chem. 1988, 92,
2561.

Jin, T.; Okuhara, T.; Mains, G. J.; White, J. M. J. Phys. Chem. 1987, 91, 3310.
Cho, B. K.; Shanks, B. H.; Bailey, J. E. J. Catal. 1989, 115, 486.

Bak, K.; Hilaire, L. Appl. Surf Sci. 1993, 70/71, 191.

Copperthwaite, R. G. Surf. Interface Anal. 1980, 2, l7.

Storp, S. Spectrochim. Acta, B, 1985, 40, 745.

Copperthwaite, R. G.; Lloyd, J. .1. Electron Spectrosc. Relat. Phenom. 1978, 14,
159.

Batista-Leal, M.; Lester, J. E.; Lucchesi, C. A. J. Electron Spectrosc. Relat.
Phenom. 1977, 11, 333.

18

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

101

Johnson, C. E. In Electronic States of Inorganic Compounds: New Experimental
Techniques, Nato Advanced Study Institutes, University of Oxford, Sept. 1974;
Day, P., Ed.; D. Reidel Publishing Company, Dordrecht-Holland, 1975; p 409-
424.

E1 Fallah, J.; Hilaire, L.; Romeo, M.; Le Normand, F. J. Electron Spec. Rel.
Phenom. 1995, 73, 89.

Hirokawa, K.; Honda, F .; Oku, M. J. Electron Spectrosc. Relat. Phenom. 1975, 6,
333.

Bumess, J. H.; Dillard, J. G.; Taylor, L. T. J. Amer. Chem. Soc. 1975, 97, 6080.

Dauscher, A.; Hilaire, L.; Le Normand, F.; Muller, W.; Maire, G.; Vasquez, A.
Surf Interface Anal. 1990, I6, 341.

Al-than, Z. T.; Hashemi, T.; Hogarth, C. A. Spectrochimica Acta. 1989, 441),
205.

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

Paparazzo, E.; Ingo. G. M.; Zacchetti, N. J. Vac. Sci. Techrrol. A. 1991, 9, 1416.
Laachir, A.; Perrichon, V.; Badri, A.; Lamotte, J.; Catherine, E.; Lavalley, C.; El
Fallah, J.; Hilaire, L.; Le Normand, F .:, Quéméré, E.; Sauvion, G. N.; Touret, O. J.
Chem. Soc. Faraday Trans. 1991, 87, 1601.

Askeland, P. A.; Ledford, J. S. to be published.

Park, P. W.; Ledford, J. S. to be published.

Shyu, J. 2.; Weber, W. 1-1.; Gandhi, H. S. J. Phys. Chem. 1988, 92, 4964.

Graham, G. W.; Schmitz, P. J.; Usmen, R. K.; McCabe, R. W. Catal. Lett. 1993,
17,175.

Powder Diffraction File, Inorganic Phases, JCPDS, 1983.

Klug, H. P.; Alexander, L. E., X -Ray Diffraction Procedures for Polycrystalline
and Amorphous Materials, lst.ed.; Wiley: New York, 1954.

XPS data manipulation performed using sofiware provided by Dr. Andrew
Proctor, University of Pittsburgh, PA.

 

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

102

Deffosse, C.; Canesson, D.; Rouxhet, P. G.; Delmon, B. J. Catal. 1978, 51, 269.
Kerkhof, F. P. I. M.; Moulijn, J. A. J. Phys. Chem. 1979, 83, 1612.

Scofield, J. H. J. Electron. Spectrosc. Relat. Phenom. 1979, 8, 129.

Penn, D. R. J. Electron. Spectrosc. Relat. Phenom. 1976, 9, 29.

Fujimori, A. Phys. Rev. B 1983, 28, 2282.

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

Kotani, A.; Mizuta, H.; Jo, T.; Parlebas, J . C. Solid State Commun. 1985, 53, 805.

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

Wuilloud, E; Delley, B.; Schneider, W.-D.; Baer, Y. Phys. Rev. Lett. 1984, 53,
202.

JO, 1.; Kotani, A. Phys. Scrp. 1987, 35, 570.
Nakano, T.; Kotani, A.; Parlebas, J. C. J. Phys. Soc. Jpn. 1987, 2201.

Creaser, D. A.; Harrison, P. G.; Morris, M. A.; Wolfindale, B. A. Catal. Lett.
1994, 23, 13.

Burroughs, P.; Hamnett, A.; Orchard, A. F; Thornton, G. J. Chem. Soc., Dalton
Trans. 1976, 1686.

Praline, G.; Koel, B. E.; Hance, R. L.; Lee, H.-I.; White, J. M J. Electron.
Spectrosc. Relat. Phenom. 1980, 21, 17.

Le Normand, F.; El Fallah, J.; Hilaire, L.; Legare, P.; Kotani, A.; Parlebas, J. C.
Solide State Communications 1989, 7], 885.

Wallbank, B.; Johnson, C. E.; Main, 1. G. J. Electron Spectrosc. Relat. Phenom.
1974, 4, 263.

Chapter 4

The Influence of Surface Structure on the Catalytic Activity of Alumina
Supported Copper Oxide Catalysts: Oxidation of Carbon Monoxide and
Methane

4.1. Abstract

X-ray photoelectron spectroscopy (XPS), X-ray diffraction (XRD) and electron
spin resonance (ESR) have been used to characterize a series of CU/A1203 catalysts. The
information obtained from surface and bulk characterization has been correlated with CO
and CH4 oxidation activity of the catalysts. For catalysts with Cu/Al atomic ratios 3
0.051, XPS data indicated that most of the Cu was present as a dispersed surface phase.
ESR results showed that the ratio of isolated/interacting copper surface phase decreased
with increasing Cu content. For catalysts with C u/Al atomic ratio 2 0.077, large CuO
crystallites were detected by XRD. The turn over number for CO oxidation increased
with increasing Cu content. This has been attributed to an increase in the amount of
crystalline CuO present in the catalysts. The specific activity for CH4 oxidation

decreased with increasing Cu content up to Cu/Al = 0.051. XPS results indicate that Cu
dispersion decreased with increasing Cu content. Since the turn over number has been

normalized by Cu dispersion obtained from XPS measurement, the decrease in CH4

103

Lfiiéial

f0! 1h:

104

oxidation activity could not be attributed to a decrease in Cu dispersion. Instead, we
propose that the isolated Cu surface phase is more active for CH4 oxidation than the
interacting copper surface phase or crystalline CuO. CH4 oxidation activities were similar

for the catalysts with Cu/Al atomic ratio 2 0.077.

4.2. Introduction

Copper oxide is one of the most active transition metal oxide catalysts for emission
control reactions [1-9]. Catalysts based on copper oxide are useful for the total oxidation
of CO [5,6,10-12], hydrocarbons [2,5,6,13], chlorinated hydrocarbons [l4], and alcohols
[15,16]. NO,i [7,17-19] and SO; reduction [19] reactions are also catalyzed by copper
oxide. Kummer [2] has reported that copper oxide catalyst exhibits a CO oxidation
activity per unit surface area similar to that of noble metal catalysts. Thus, supported
copper oxide catalysts have been considered as suitable substitutes for noble metal-based
emission control catalysts.

The structure of Cu/A1303 catalysts depends on both the metal loading and
calcination temperature [20-32]. For low copper loadings (< 8~10 wt.% Cu on 200m2/g
y-Ale3 support), copper is present as a well-dispersed surface phase. Magnetic
susceptibility and ESR measurements have been used to distinguish between two species

in the copper surface phase: isolated and interacting copper species [20-25]. The isolated
CUPric ions contribute to both magnetic susceptibility and resonance absorption, whereas

the interacting copper species (or ‘clustered copper ions’) do not give a detectable ESR

105
signal. The c0pper surface spinel is believed to result either from the diffusion of copper

ions into an alumina lattice [26] or from the interaction of a Cu2+ ion complex with
alumina hydroxyl groups [28]. Strohmeier et al. [26] have reported that the chemical state
of the spinel is different from both bulk CuO and CuA1204. Indeed, diffuse reflectance
spectroscopy results have shown that cupric ions in the surface spinel predominantly
occupy tetragonally distorted octahedral sites (> 90%) with only a small fraction located in
tetrahedral sites [21,29,32]. *Bulk CuAle4 (60 % tetrahedral Cuzi and 40 % octahedral
Cu”) can be formed only afier high temperature calcination (> 700 °C) [20,21,26]. For
high copper loadings (> 10 wt.% Cu), crystalline CuO is formed on the alumina support
[21,26].

The CO and CH4 oxidation activities of CU/N203 catalysts also depend on copper
content. Mooi and Selwood [33] have reported that the CO oxidation activity is the
highest for a 7 wt.% CuO/AIZO3 catalyst (supported on 200 mz/g y-AIZO3). Severino et al.
[10] reported that a 5 wt.% CuO/A1303 catalyst (supported on 100 mZ/g y-Ale3) shows

,. -

the maximum CO oxidation activity. The authors attributed the activity of the low loading
catalysts to copper ions dispersed on the alumina surface. However, Pierron et al. [34]
have reported that CO oxidation occurs on crystalline copper oxide phases. Tsikoza et al.
[35] and Marion et al. [363 7] reported that the CH4 oxidation activity per unit mole or
gram of copper decreases with increasing copper loading. Tsikoza et al. [35] suggested
that CH4 oxidation proceeds primarily on the well-dispersed copper surface phase.
Marion et al. [36,37] also suggested that the catalytic activity for methane oxidation

depends on the amount of well-dispersed copper ions. In addition, these authors

a?

tutu

106

attributed the decrease in CH4 oxidation activity with increasing Cu loading primarily to a
decrease in copper oxide dispersion.

Several previous studies of Cu/A1203 catalysts have focused on the structure or
activity of the catalyst. Little effort has been devoted to investigating systematically the
relationship between the surface structure and CO and CH4 oxidation activity of Cu/Ale3
catalysts. The present work is part of a broad study to investigate structure-reactivity
correlations for transition metal oxide based emission control catalysts. In this paper, X-
ray photoelectron spectroscopy (XPS), X-ray diffraction (XRD), and electron spin
resonance (ESR) have been used to determine the effect of Cu loading on the chemical
state and dispersion of copper species supported on y-alumina. The information derived
from these techniques is correlated with the CO and CH4 oxidation activity in order to

develop a more complete understanding of Cu/Al203 catalysts.

4.3. Experimental

Catalyst Preparation. Catalysts were prepared by pore volume impregnation of y-
alumina (Cyanamid, surface area = 203 mz/g, pore volume = 0.6 mL/g) using solutions of
copper(II) nitrate (Columbus Chemical Industries Inc, ACS Grade). The alumina was

finely ground (<230 mesh) and calcined in air at 500 °C for 24 h prior to impregnation.
The impregnated samples were dried in air at 120 °C for 24 h and calcined in air at 500 °C
for 16 h. The copper content of the CU/Aleg, series was varied from a Cu/Al atomic

1'3 ti o of 0 to 0.128 (0 to 20 wt.% CuO). Cu/A1203 catalysts will be designated by “Cuy”,

respec

match

Quam
outga
meaSL
05,

COME

Opera
0ft»:

35p0

107
where y is the Cu/Al atomic ratio (x102) of the catalysts. The catalyst colors change with

increasing copper content from pale blue (Cu0.3 and Cu0.6), to pale green (Cu1.3 and
Cu2.6), to green (Cu5.1), to dark green (Cu7.7), to gray (Cu10), to dark gray (Cul3).

Standard Materials. CuO was prepared by calcining copper(II) nitrate in air at
500 °C for 16 h. CuA1204 was prepared by calcining stoichiometric amounts of the
respective nitrates in air at 1000 °C for 24 h. XRD patterns of the standard compounds
matched the appropriate Powder Diffraction File [3 8].

BET Surface Area. Surface area measurements were performed using a
QuantaChrome Quantasorb Jr. Sorption System. Approximately 0.1 g of catalyst were
outgassed in a Nz/He mixture (5% N2) at 350 °C for 1 h prior to adsorption
measurements. The measurements were made using relative pressures of N; to He of
0.05, 0.08, and 0.15 (N; surface area = 0.162 nmz) at 77 K. Increasing the copper
content of the catalysts decreases the BET surface area by a maximum of 20 %.

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

Rigaku XRD diffractometer employing Cu K, radiation (7x = 1.5418 A). The X-ray was
operated at 45 kV and 100 mA. Diffraction patterns were obtained using a scanning rate
of 0.5 deg/min (in 29) with divergence slit and scatter slit widths of 1°. Samples were run
as powders packed into a glass sample holder having a 20 x 16 x 0.5-mm cavity.

The mean crystallite sizes (3) of the CuO particles were determined from XRD

line broadening measurements using the Scherrer equation [39]:

Z=Kmpcose (1)

there

iii wi

Cu0<
measr
app:

the A

1mm
[FLA ]
instn

Spec

108

where it is the X-ray wavelength, K is the particle shape factor, taken as 0.9, and B is the

full width at half maximum (fwhm), in radians, ofthe Cuo<il 1> line.

Quantitative X-ray diffraction data were obtained by comparing

CuO<ill>/A1203<400> intensity ratios measured for catalyst samples with intensity ratios
measured for physical mixtures of pure CuO and y—A1203. This method assumed that
copper addition did not disrupt the y-A1203 spinel and subsequently affect the intensity of
the A1203<400> line. The error in this method was estimated to be 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 300 W (15 kV, 20
mA) and a 10-360 hemispherical analyzer operated with a pass energy of 50 eV. The
instrument typically operates at pressures near 1 x 10’8 Torr in the analysis chamber.
Spectra were collected using a PC 137 board interfaced to a Zeos 386SX computer. XPS
spectra were also obtained using a VG Microtech spectrometer to confirm Cu 2pm
binding energies. The VG spectrometer is equipped with a magnesium anode (1253.6 eV)
operated at 300 W (15 kV, 20 mA) and a Clam2 hemispherical analyzer operated with a
pass energy of 50 eV. 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 or spray coated onto a quartz slide using a methanol suspension of the catalyst.

Binding energies for the catalyst samples and standard compounds which contained Al
Were referenced to the Al 2p peak (74.5 eV). The binding energy for CuO was referenced

to the C Is line (284.6 eV) of the carbon overlayer. Detector efficiency was corrected

 

109
using Cu 2p and Cu 3p peak intensities measured for copper metal foil which was cleaned

with Ar ion bombardment. Since the Cu 3p line overlaps the Al 2p line for the catalyst
samples, estimated Cu 3p intensities (from the ratio of Cu 3p to Cu 2p) were subtracted
from Al 2p intensities prior to calculating the Cu/Al intensity ratio. XPS binding energies
were measured with a precision of i 0.2 eV, or better.

Reduction of copper species during XPS experiments has been reported [26,40,41]
and attributed to several factors such as X-ray flux, X-ray dose, temperature, and
pressure. In order to minimize the effect of photoreduction on the results, all samples
were analyzed using the same distance between the X-ray source and sample (~ 2 cm) and
minimum data acquisition time (5 min). No significant photoreduction was observed using
these experimental conditions.

Quantitative XPS Analysis. It has been shown by Defosse et al. [42] that one may
calculate the theoretical intensity ratio I°,,/I°s expected for a supported phase (p)
atomically dispersed on a carrier (5). An extension of the Defosse model proposed by
Kerkhof and Moulijn [43] has been used in the present investigation. The photoelectron
cross sections and the mean escape depths of the photoelectrons used in these calculations
were taken from Scofteld [44] and Penn [45], respectively. For a phase (p) present as

discrete particles, the experimental intensity ratio l,,/1s is given by the following expression:

U], = 1°p/I°,[l-exp(-d/?tp)]/ cm, (2)

110

where I°P/I°. is the theoretical monolayer intensity ratio, d is the length of the edge of the
cubic crystallites of the deposited phase and A? is the mean escape depth of the
photoelectrons in the deposited phase. The Cu species particle sizes were determined by
solving the Eq.(2) for d. The relative Cu dispersion to the monolayer was estimated from
the ratio of VI. to 1°./1°,

ESR Study. ESR spectra were obtained using a Varian E-4 spectrometer with a
rectangular TE 10; cavity operated at a microwave frequency of 9.06 GHz and a power of
20 mW. Ruby crystal mounted inside the cavity at fixed orientation was used as an
intensity standard for determination of relative ESR intensities. The relative intensities per
unit mg CuO were obtained from the ratio of the peak intensity of catalyst sample to the
value of Ruby crystal. Powder samples were dried at 120 °C for 24 hr prior and run at 98
K in a quartz ESR tube.

C 0 Oxidation Activity. Measurement of CO oxidation activity was performed in a
flow microreactor. Approximately 0.03 g of catalyst were supported on a glass frit (70 -
100 um) and the temperature was measured with a K-type thermocouple located just
above the catalyst bed. The reactor was heated by a tube furnace (Lindberg) with
temperature being controlled within 1 °C by an Omega CN 1200 temperature controller.
Reactant gas flow rates were held constant with Brooks 5850 mass flow controllers.
Product gases were analyzed with a Varian 920 gas chromatograph equipped with a TCD

and interfaced to a Hewlett-Packard 3394A integrator. Reaction products were separated
on a 6 fi 60/80 mesh Carbosieve S-Il column. Prior to the first activity measurement,

catalysts were pretreated with a mixture of 5% Oz/He (99.5 % purity for 02, 99.995 %

pumy
impuri

reactic

coast:
punt}
were
late
trap r
(C0

than

mole

XPS

111m

loadi

111
purity for He, AGA Gas Co.) stream (143 cc/min) at 350 °C for 1 hr to remove any

impurities adsorbed on the surface during catalyst preparation and storage. CO oxidation

reactions were performed with a constant flow rate (80 cm3/min) of 4.8% CO/9.8°/o

07/85.4% He gas mixture (AGA, purity > 99.99 %) in the temperature range of 150-400
oC_

CH4 Oxidation Activity. Methane oxidation reactions were performed with a
constant flow (15 cm3/min) of 0.98% CH4/5.25% O;/93.77% He gas mixture (AGA,
purity > 99.99 %) in the temperature range of 360-430 °C. Approximately 0.1 g catalyst
were charged into the same type of microreactor used for CO oxidation measurements.
Water produced during methane oxidation was frozen downstream from the reactor in a
trap maintained < -40 °C with a mixture of ethanol and dry ice. All activity measurements
(CO and CH4 oxidation) were obtained under steady-state conditions at conversions less
than 15 %. The turn over number (TON) was expressed as mole CO or CH; oxidized per
mole of surface-bound c0pper. The surface bound copper content was determined by

XPS from the relative intensity ratios of Cu 2pm and Al 2p peaks.

4.4. Results

State and Dispersion of Copper. XRD patterns obtained for Cuy catalysts with
Clat/Al atomic ratios 3 0.051 (s 8 wt.% CuO) showed only lines characteristic of the
alumina carrier (Figure 4.1a). XRD patterns obtained for catalysts with higher Cu

10adings (Cu/Al atomic ratios 2 0.077) contain peaks characteristic of CuO (Figure 4.1b-

112

 

CuO
<i11> Cuo

<111>

CuO C110

<110> <262>
(d) w

 

 

 

 

 

(b)
AIZO,
< >
3 l 1 A120:
A120, . ‘ <222>
<220>
(a)
A120,
<400>
L l l l l l

 

 

25 30 35 40 45 50
Degrees (20)

l:igure 4.1. XRD patterns measured for (a) A1203, (b) Cu7.7, (c) CulO, and (d) Cu13.

113

Table 4.1. Concentration of Crystalline‘ Phases in Cuy
Catalysts Calculated from Quantitative XRD Data.

 

 

Catalysts Crystalline Phase (wt.%)
As prepared Measured
Cu 0.3 0.5 0
Cu 0.6 1 0
Cu 1.3 2 0
Cu 2.6 4 0
Cu 5.1 8 0
Cu 7.7 12 5.7
Cu 10 16 10.8
Cu 13 20 14.6

 

 

 

a- Valid for crystalline phases with particle sizes > 3.0 nm.

d)- The intensities of the CuO XRD peaks increase with increasing Cu content.
Quantitative XRD measurements (Table 4.1) show that the amount of crystalline CuO
increases from 5.7 to 14.6 wt.% as the Cu/Al atomic ratio increases from 0.077 to 0.13.
The XPS Cu 2133/2 binding energies measured for the Guy catalysts (935.1 i 0.2 eV) were
independent of Cu content and similar to the Cu 2pm binding energy measured for
CuAhOa (935.0 eV) and higher than the value measured for CuO (933.9 eV). The
SEltellite/main peak ratio measured for the Cuy catalysts (0.47 i 0.1) is similar to that
measured for CuO (0.45) and lower than the value for CuAlea (0.71).

Figure 4.2 shows the variation of the Cu 2p3;2/Al 2p intensity ratio measured for
the Cuy catalysts as a fiJnction of Cu/Al atomic ratio. The theoretical line calculated for
monolayer dispersion [43] is shown for comparison. The Cu/Al intensity ratio increases
with increasing Cu/Al atomic ratio up to 0.051. For higher Cu loadings, the Cu/Al

Intensity ratio measured for the catalysts levels off. Variation in the particle size of the

XPS Intensity Ratio (Cu2p3/2/A12p)

I“igure 4.2.

0

114

 

 

 

 

     

 

monolayer line
0 catalysts
.. O r 1
J
W I I l U I I I I l U r I

 

0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
Atomic Ratio (Cu/A1)

Cu 2P3/2/Al 2p XPS intensity ratios measured for Cuy catalysts(0) plotted
versus Cu/Al atomic ratio. Cu 2p3n/Al 2p intensity ratios calculated for
monolayer dispersion (line).

115

Table 4.2. Particle Size of Copper Phases Determined from
Cu/Al XPS Intensity Ratios and XRD Line
Broadening Calculations.

 

 

Catalysts Particle size (nm)
XPS XRD
Cu 0.3 0.4 -‘
Cu 0.6 0.6 -'
Cu 1.3 0.8 -‘
Cu 2.6 0.8 -'
Cu 5.] 0.9 -"
Cu 7.7 1.7 30
Cu 10 2.3 30
Cu 13 3.0 29

 

 

 

a- No CuO XRD peaks detected for these catalysts

copper species determined from Cu/Al XPS intensity ratios using Eq. (2) is given as a
function of the Cu loadings in Table 4.2. The mean particle size of the copper species
Calculated from XPS intensity ratios increases from 0.4 to 3.0 nm with increasing Cu/Al
atomic ratio. Particle sizes determined from XRD line broadening calculations using Eq.
( 1 ) are also shown in Table 4.2. For catalysts with Cu/Al atomic ratios 2 0.077, XRD line
broadening calculations show that the CuO crystallite sizes are independent of Cu loading
and significantly larger than the values determined using XPS.
Figure 4.3 shows the variation of the relative ESR signal intensity measured for the
Guy catalysts as a function of Cu/Al atomic ratio. For catalysts with Cu/Al atomic ratios 3
0- 051, relative ESR signal intensity decreases significantly with increasing Cu content.
I“Iowever, for catalysts with higher Cu/Al atomic ratios, the relative ESR signal intensities

are independent of Cu content.

 

 

 

 

 

 

 

 

 

 

 

 

 

116
18
16 _. A
'e
u- _fi/

14 /
.~ L
m
c: A
3 12
.5
G4
m "__i
LU 10 lllllll
B m 16% 230a“? 4000 48m
-521
v-g 8
0
Z 6-

4-

2-

U U
0.1.1.1.111oi.
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14

Atomic Ratio (Cu/A1)

I:igure 4.3. Normalized ESR intensity (1 mg CuO) of Cuy catalysts relative to the
internal standard.

117

Table 4.3. Turn Over Numbers‘ and Activation Energies” of
Cu/A1203 Catalysts for CO and CHa Oxidations.

 

 

CO Oxidation CH4 Oxidation
Catalysts Activation Energy TON (x104) Activation Energy TON (x104)
Cu 0.3 25.8 3.5 21.8 11.8
Cu 0.6 25.7 6.8 22.1 8.3
Cu 1.3 19.2 33.6 22.6 6.5
Cu 2.6 19.4 76.0 23.4 4.7
Cu 5.1 19.2 242 21.9 2.8
Cu 7.7 13.1 390 22.3 3.2
Cu 10 13.7 694 23.6 2.9
Cu 13 13.2 1010 21.4 2.9

 

 

 

 

a- Estimated from Arrhenius plots (kcal/mol) within 15 "/0 conversion.
b- Calculated at 260 and 390 0C for CO and CH4 oxidation respectively (5")
using exposed Cu species determined from XPS.

CO Oxidation Activity. Table 4.3 shows the specific turn over number (TON) for
CO oxidation over Cuy catalysts calculated at 260 °C using XPS estimates of the Cu
dispersion. The TON increases by nearly a factor of 300 as the Cu/Al atomic ratio
increases from 0.003 to 0.13. The activation energies calculated for CO oxidation are also
SFlown as a function of Cu content in Table 4.3. The activation energy decreases from
25-8 to 13.1 kcal/mol as the Cu/Al atomic ratio increases from 0.003 to 0.077. For
cEltrdlysts with Cu/Al atomic ratios 2 0.077, the activation energies are identical within
experimental error (13.3 i 0.3 kcal/mol).

CH4 Oxidation Activity. Table 4.3 shows the TON for CH4 oxidation over Cuy
canalysts calculated at 390 °C using XPS estimates of the Cu dispersion. The TON
decreases by nearly a factor of 4 as the Cu/Al atomic ratio increases from 0.003 to 0.051.

I:llt‘ther addition of Cu has little effect on the TON for CH4 oxidation. Table 4.3 also

118
shows that the activation energies calculated for CH4 oxidation are similar within

experimental error (22.4 :t 1.0 kcal/mol).
For catalysts with Cu/Al atomic ratio 3 0.006, the CO; selectivity measured at 390
°C increased from 63 % to 80 % with increasing copper content. No partial oxidation

product was observed for catalysts with Cu/Al atomic ratio 2 0.013.

4.5. Discussion

Structure of Cooper/Alumina Catalysts. For catalysts with Cu/Al atomic ratio 3
O - 051, the absence of XRD peaks characteristic of discrete CuO and the small particle size
calculated from Cu/Al XPS intensity ratios suggest that copper exists as a well-dispersed
COpper surface phase. These results are consistent with previous work by Friedman et a1.
[2 l ], which showed that the support capacity of several y-aluminas for copper surface
phase is approximately 4 wt.%Cu/lOOmZ/g support. Strohmeier et al. [26] also reported
tl'lat 9-10 wt.% copper (for 195 mZ/g y-alumina) is the maximum copper loading that can
be accommodated as a surface phase.

The Cu 2pm XPS spectra measured for the Cuy catalysts show Cu 2pm binding
erlergies characteristic of CuAlea and satellite/main peak intensity ratios close to the
Value obtained for CuO. It is well known [21,26] that Cu/A1203 catalysts contain Cu2+
SL1 rface phase and discrete CuO particles. In principle, photoelectrons from both discrete

CUO particles and Cuzi surface phase will contribute to the XPS signal. However, due to

119
the high dispersion of the surface phase relative to the CuO crystallites, the qualitative

features of the Cu 2pm XPS spectra are expected to be more representative of the surface
phase. Consequently, the observation of satellite/main peak intensity ratios typical of CuO
is surprising. However, this apparent discrepancy has been reported by previous
researchers [26,30]. Strohmeier et al. [26] suggested that this did not necessarily mean
that well-dispersed or amorphous CuO was present on the catalysts, but simply that the
copper surface species is chemically different from bulk CuO and CuA1204.

The significant decrease in the relative ESR intensity as a function of Cu content is

consistent with the result from previous work [21,22]. The ESR result indicates that
isolated/interacting copper surface phase ratio decreases as Cu loading is increased in the
range of 0 < Cu/Al 3 0.051. The former contributes the resonance absorption, which can
be assigned as tetragonally distorted, octahedrally-coordinated cupric ions [23]. The latter
contributes less ESR signal due to interaction with neighboring cupric ions [22]. The
decrease in Cu dispersion for catalysts with Cu/Al g 0.051 can be attributed to the
formation of interacting copper surface phase or copper oxide clusters.

For Cuy catalysts with Cu/Al atomic ratio 2 0.077, the increase in XPS determined
COpper oxide particle size with increasing Cu loading is consistent with the appearance of
X~ray diffraction lines characteristic of CuO. These results are similar to previous work
[2 1,26] which indicated that at a critical metal loading, the support becomes saturated
with copper surface phase and further addition of the metal results in formation of CuO
crystallites. The difference in CuO particle sizes determined using XPS and XRD may be

attributed to the limitations of X-ray diffraction. It is well known that XRD determined

120

particle sizes are skewed to large values since XRD does not detect highly dispersed
species. For the Cu rich catalysts, a significant fraction of the copper is present as a
surface phase or small CuO particles (d < 3.0 nm) that are detected by XPS, but not XRD.
Thus, smaller average particle sizes are expected from XPS calculations.

Effect of Catalyst Structure on C0 Oxidation Activity. The significant increase in
CO oxidation TON as a function of Cu content observed for the Cuy catalysts suggests
that CuO is the active phase for CO oxidation. In addition, when one considers that the
crystalline CuO is poorly dispersed compared to the copper surface phase, it is clear that
CuO is significantly more reactive than the copper surface phase. The increase in CO
oxidation activity with increasing Cu loading observed in this study is not consistent with
previous studies of CO oxidation over alumina supported copper oxide catalysts. The
optimum copper contents have been reported as approximately 5 wt.% Cu/95 m2/g y-
A1203 [10] and 7 wt.% Cu on 200 mZ/g y-Ale3 [33].

For the CO oxidation over alumina supported copper oxide catalysts, it is generally
considered that CO oxidation involves redox phenomena on the surface of copper oxide.
Jernigan and Somorjai [46] have recently proposed that CO oxidation proceeds by a redox
mechanism involving CuO and CuzO and that the reduction of CuO by CO is the rate
determining step. Since the oxidation reaction may require adsorption of CO at oxygen
vacancies, the availability of oxygen vacancies is important for copper oxide based CO
oxidation catalysts [11]. On the basis of CO TPR results, Severino et a1. [10] have
reported that the CuO crystalline phase is reduced at a lower temperature than the copper

surface phase. Therefore, the high CO oxidation activity of the CuO phase can be

121

attributed to more facile generation of oxygen vacancies during the CO oxidation reaction
compared to the copper surface phase.

Effect of Catalyst Structure on CH4 Oxidation Activity. The decrease in CH4
oxidation activity with increasing Cu loading has been reported in previous studies
[35,36]. Tsikoza et al. [35] suggested that CH4 oxidation occurred primarily on the
copper surface phase. In addition, Garbowski and Primet [28] have proposed that
octahedrally coordinated surface copper ions are accessible to reactants involved in
catalytic combustion reactions. Marion et al. [36] have reported that CH4 oxidation
activity for CuO/Ale3 catalysts is proportional to the number of surface copper species
and that the number of active centers decreased with increasing copper loading due to the
decrease in Cu dispersion. In this study, the decrease in CH4 oxidation activity for Cuy
catalysts cannot be attributed to a change in Cu dispersion, since the turn over number
was obtained by normalization with Cu dispersion. Instead, we propose that the isolated
Cu2+ species are more reactive for CH4 oxidation than the interacting copper surface
species. For the high loading catalysts (Cu/Al 2 0.077), the CH4 oxidation activity is
independent of Cu content, which can be ascribed to the formation of poorly dispersed
CuO crystallites.

The mechanism of CH4 oxidation differs in some respects from C0 oxidation.
First, the rate determining step of CH4 oxidation involves C-H bond breakage [47].
Second, oxidation of CH4 produces water which may poison catalysts [11]. Therefore
CH4 oxidation requires a higher reaction temperature, and thus the influence of oxygen

vacancies will be less important due to high mobility of lattice oxygen [13]. Liu and

122
Flytzani-Stephanopoulos [48] have proposed that acidic and basic sites on the catalyst

surface play an important role for C-H bond breakage and formation of intermediate
species. Isolated copper species may provide acidic/basic sites with more suitable strength
and geometry for CH; oxidation than interacting copper species or crystalline CuO.

CH4 TPR studies have been performed with CuO/A1203 catalysts. Kartheuser et
al. [49] reported a broad, featureless peak in the range of 500-700 °C for a 5 wt.%
CuO/325m2/g A1203 catalyst. They assigned this feature to the reduction of CuO or Cu-
A1203. H2 TPR results for Cu/A1203 catalysts showed that the reduction of copper surface
phase occurs at a lower temperature than for CuO [37,50]. These H2 TPR results are
usefiil for understanding the high CH4 oxidation activity of isolated copper species since

both H2 and CH4 utilize hydrogen to reduce copper species and produce water.

4.6. Conclusions

The combined use of several techniques to investigate the effect of Cu loading on
the structure and catalytic activity of Cu/A1203 catalysts for CO and CH4 oxidation leads
to the following conclusions.

1. For low Cu loadings (Cu/Al 5 0.051), copper exists as a dispersed copper
surface phase. ESR results indicate that the fraction of interacting Cu phase increases
with increasing Cu content. For Cu rich catalysts (Cu/A1 2 0.077), large CuO crystallites

are formed.

123

2. The increase in CO oxidation activity observed with increasing Cu content can
be attributed to an increase in concentration of crystalline CuO. The copper surface phase
shows very low CO oxidation activity. This has been explained with the ease of CuO
redox behavior compared to copper surface phase during the CO oxidation reaction.

3. The CH4 oxidation activity decreases with increasing Cu loading. We propose
that the isolated copper surface phase is more active for CH; oxidation than the interacting
copper surface phase or crystalline CuO. For higher Cu loadings, the CH4 oxidation
activity is independent of Cu loading, which can be attributed to the formation of poorly

dispersed CuO crystallites.

10.

11.

12.

13.

14.

15.

16.

17.

18.

124
4.7. References

Prasad, R., Kennedy, L. A. and Ruckenstein, E. Catal. Rev. -Sci. Eng. 1984, 26, l.
Kummer J. T., Prog. Energy Combust. Sci. 1980, 6, 177.
Dwyer, F. G. Catal. Rev-Sci. Eng. 1972, 6, 261.

Zwinkels, M. F. M.; Jarés, S. G.; Menon, P. G. Catal. Rev. -Sci. Eng. 1993, 35
319.

Yao, Y. F. Yu; Kummer, J. T. J. Catal. 1977, 46, 388.
Yao, Y. F. Yu J. Catal. 1975, 39, 104.

Kapteijn, F.; Stegenga, S.; Dekker, N. J. J.; Bijsterbosch, J. W.; Moulijn, J. A.
Catal. Today 1993, I6, 273.

Klimisch, R. L. General Motors Res. Publ. GMR-842 (1968).

Heyes, C. J .; Irwin, J. G.; Johnson, H. A.; Moss, R. L. J. Chem. Tech. Biotechnol.
1982, 32, 1025.

Severino, F.; Brito, J.; Cari’as, O. Laine, J. J. Catal. 1986, 102, 172.
Boon, A. O. M.; van Looij, Fr, Geus, J. W. J. Mol. Catal. 1992, 75, 277.

Lopez Agudo, A.; Palacios, J. M.; Fierro, J.L.G.; Laine, J.; Severino, F. Appl.
Catal. 1992, 91,43 .

Boon, A. O. M.; Huisman, H. M.; Geus, J. W. J. Mol. Catal. 1992, 75, 293.
Subbanna, P.; Greene, H.; Desal, F. Environ. Sci. Technol. 1988, 22, 557.

Ozkan, U. S.; Kueller, R. F.; Moctezuma, E. Ind. Eng. Chem. Res. 1990, 29,
1136.

Rajesh, H.; Ozkan, U. S. Ind. Eng. Chem. Res. 1993, 32, 1622.
London, J. W.; Bell, A. T. J Catal. 1973, 31, 96.

Huang, T.-J.; Yu, T.-C. Appl. Catal. 1991, 7], 275.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

125

Goetz, V. N.; Sood, A.; Kittrell, J. R. Ind. Eng. Chem. Prod. Res. Develop. 1974,
I3, 110.

Wolberg, A.; Roth, J. F. J. Catal. 1969, 15, 250.

Friedman, R. M.; Freeman, J. J.; Lytle, F. W. J. Catal. 1978, 55, 10.
Matsunaga, Y. Bull. Chem. Soc. Jpn. 1961, 34, 1291.

Berger, P. A.; Roth, J. F. J. Phys. Chem. 1967, 71, 4307.

Lumbeck, H.; Voitlander, J. J. Catal. 1969, 13, 117.

Centi, G.; Perathoner, S.; Biglino, D.; and Giamello, E. J. Catal. 1995, 151, 75.

Strohmeier, B. R.; Leyden, D. E; Field, R. S.; Hercules, D. M.; Petrakis, L. J.
Catal. 1985, 94, 514.

Ertl, G.; Hierl, R.; Knozinger, H.; Thiele, N.; Urbach, H.-P., Appl. Surf Sci. 1980,
5, 49.

Garbowski, E.; Primet, M. J. Chem. Soc. Chem. Commun. 1991, 11.

Lo Jacono, M.; Cimino, A.; Inversi, M. J. Catal. 1982, 76, 320.

Tikhov, S. F.‘, Sadykov, V. A.; Kryukova, G. N.; Paukshtis, E. A.; Popovskii, V.
V.; Starostina, T. G.; Kharlamov, G. V.; Anufrienko, V. F.; Poluboyarov, V. F.;
Razdobarov, V. A.; Bulgakov, N. N.; Kalinkin, A. V. J. Catal. 1992, I34, 506.
Summers, J. C. and Klimisch, R. L. Proc. Int. Cong): Catal. 51h 1972 vol 2, 293.
Freeman, J. J .; Friedman, R. M. J. Chem. Soc. Faraday Trans. 1. 1978, 758.

Mooi, J.; Selwood, P. W. J. Am. Chem. Soc. 1952, 74, 2461.

Pierron, E. D.; Rashkin, J. A.; Roth, J. F. J. Catal. 1967, 9 38.

Tsikoza, L. T.; Torasova, D. V.; Ketchik, S. V.; Maksimov, N. G.; Popovskii, V.
V. Kinet. Katal, 1981, 22, 1022.

Marion, M. C; Garbowski, E.; Primet, M. J. Chem. Soc, Faraday Trans. 1990,
86, 3027.

37.

38.

39.

40.

41.

42.

43.

44.

45.‘

46.

47

48.

49.

50.

126

Marion, M. C.; Garbowski, E.; Primet, M. J. Chem. Soc, Faraday Trans. 1991,
87, 1795.

Powder Diffraction File, Inorganic Phases, JCPDS, 1983.

Klug, H. P.; Alexander, L. E., X -Ray Diffraction Procedures for Polycrystalline
and Amorphous Materials, lst.Ed. Wiley, New York, 1954.

Rosencwaig, A. and Wertheim, G. K. J. Electron. Spectrosc. Relat. Phenom.
1972, l, 493.

Wallbank, B., Johnson, C. E., and Main, 1. G. J. Electron. Spectrosc. Relat.
Phenom. 1974, 4, 263.

Deffosse, C.; Canesson, D.; Rouxhet, P. G.; Delmon, B. J. Catal. 1978, 51, 269.
Kerkhof, F. P. J. M.; Moulijn, J. A. J. Phys. Chem. 1979, 83, 1612.

Scofield, J. H. J. Electron. Spectrosc. Relat. Phenom. 1979, 8, 129.

Penn, D. R. J. Electron. Spectrosc. Relat. Phenom. 1976, 9, 29.

Jernigan, G. G.; Somorjai, G. A. J. Catal. 1994, 147, 567.

Otto, K. Langmuir 1989, 5, 1364.

Liu W. and Flytzani-Stephanopoulos, M. J. Catal. 1995, 153, 304.

Kartheuser, B.; Hodnett, B. K.; Riva, A.; Centi, G.; Matralis, H.; Ruwet, M.;
Grange, P.; Passarini, N. IndEng. Chem. Res. 1991, 30, 2105.

Dumas, J. M.; Geron, C.; Kribii, A.; Barbier, J. Appl. Catal. 1989, 47, L9.

Chapter 5

Characterization and CH4 Oxidation Activity of Cr/A1203 Catalysts

5.1. Abstract

X-ray photoelectron spectroscopy (XPS) and X-ray diffraction (XRD) have been
used to characterize a series of Cr/Ale3 catalysts (designated “Cry”, y = Cr/AJ atomic
ratio). The information obtained from surface and bulk characterization has been
correlated with CH4 oxidation activity of the Cry catalysts. For catalysts with a Cr/Al
atomic ratio y = 0.013, XPS results indicated that most of the Cr was present as a highly
dispersed Cr6+ surface phase. Catalysts with intermediate Cr loadings (y = 0.027 ~ 0.080)
showed no XRD peak characteristic of chromium oxide. However, XPS data indicated
that the Cr dispersion decreased and the concentration of Cr3+ species increased with
increasingCr content. For catalysts with high Cr loadings (Cr/Al atomic ratio 2 0.107),
large Cr203 crystallites were detected by XRD. The specific activity for CH4 oxidation
increased with increasing Cr content up to Cr/Al = 0.107. This has been attributed to an
increase in the amount of Cr(IIl)-Cr(VI) cluster present in the catalysts. A decrease in
CH.. oxidation activity observed for a Cr rich catalyst (Cr/Al = 0.13) has been ascribed to

the formation of poorly dispersed Cr203.

127

128

5.2. Introduction

Alumina supported chromium oxides have been studied extensively due to their
potential application in emission control catalysis [1-3]. Desirable chemical and physical
properties, such as low temperature efficiency for hydrocarbon [4-6] and chlorinated
hydrocarbon oxidation [7-10], low selectivity for C12 formation [8,11], high resistance to
HCl poisoning [12] during chlorinated hydrocarbon oxidation, NO reduction with CO [13]
and NH3 [14], high thermal stability [6,15] and mechanical durability, [12] make Cr/A1203
catalysts suitable candidates for automotive emission control and industrial waste
treatment. Chromium oxide is also an effective promoter for copper oxide catalysts used
for CO and hydrocarbon oxidation and NO reduction [13,16,17].

The surface structure of alumina supported chromium oxide catalyst depends on
the metal loading, calcination temperature and hydration/dehydration conditions [18-26].
For low chromium loadings (s 10 wt.% Cr203/Al203 200 mZ/g), most of the chromium is
present as a highly dispersed Cr(Vl) surface species. XPS studies indicate that the
proportion of Cr(VI) on the alumina support decreases with increasing Cr content
[18,19,21] or with increasing calcination temperature [21,27]. This has been attributed to
the formation of Cr(IIl) species on the support surface [19,21,27]. Ellison and coworkers
[28-32] studied Cr/Al2O3 catalysts using magnetic susceptibility, DRS, SIMS and ESR

techniques. They proposed that a mixed oxidation state cluster (-Cr6i-02'-Cr3+-Oz'-Cr6+-;

y-phase) forms on the alumina support even at 1 wt.% Cr on 157 mz/g A1203. Wachs and

129

coworkers [22,33] reported the monolayer coverage of chromium oxide (12 wt.%
CrO3/Al2O3 180 mZ/g) based on Raman, XRD, and FTIR data. This is in good agreement
with the theoretical monolayer coverage calculated from the cross section of K2Cr2O7 [34]
and ISS results [35]. For high chromium loadings (> 10 wt.% Cr203/Al2O3 200 mZ/g),
crystalline 020:. is observed using Raman spectroscopy and X-ray diffraction. Bulk
Cr(III)-Al2O3 forms after high temperature calcination (2 800 °C) of the Cr/Al2O3
containing crystalline Cr2O3 [22].

Despite the importance of Cr/Al203. catalysts in emission control applications, the
study of CH; oxidation over Cr/Al2O3 catalysts has been relatively limited [36,37]. Little
effort has been devoted to investigating systematically the relationship between the surface
structure and the CH4 oxidation activity of Cr/AJ2O3 catalysts. Anderson et a1. [36] have
studied CH4 oxidation over various alumina supported transition metal oxide catalysts and
reported that chromium oxide shows the highest methane oxidation activity. Kuznetsova
et al. [37] have reported that CH4 oxidation activity increases and levels off with
increasing Cr content. They proposed that Cr(VI) is the main active component for the
CH4 oxidation reaction.

The present work is part of a broad study to investigate structure-reactivity
correlation for transition metal oxide based emission control catalysts. In this paper, X-
ray photoelectron spectroscopy (XPS) and X-ray diffraction (XRD) have been used to

determine the effect of Cr loading on the chemical state and dispersion of chromium oxide

Phases supported on y-alumina. The information derived from these techniques is

130

correlated with CH. oxidation activity to develop a more complete understanding of

Cr/Al2O3 catalysts.

5.3. Experimental

Catalyst Preparation. The Cr-modified alumina carriers were prepared by pore
volume impregnation of y-alumina (Cyanamid, surface area = 203 mZ/g, pore volume = 0.6
mL/g) using solutions of chromium (III) nitrate (Mallinckrodt, Analytical Reagent). The
alumina was finely ground (<230 mesh) and calcined in air at 500 °C for 24 h prior to
impregnation. The impregnated samples were dried at 120 °C for 24 h and calcined at 500
°C for 16 h. The chromium content of the Cr/Al2O3 series was varied from a Cr/Al atomic
ratio of O to 0.134 (0 to 20 wt.% Cr2O3). Cr/A1203 catalysts will be designated by "Cry",
where y is the Cr/Al atomic ratio (x 102) of the catalysts. The catalyst colors change with
increasing chromium content from pale yellow (Cr1.3), to yellow (Cr2.7), to dark yellow
(Cr5.4), to brown (Cr8.0), to dark brown (Crl 1 and Crl3).
Standard Materials. Cr203 was prepared by calcining chromium (III) nitrate at
500 °C for 16 h in air. CrO_~. (99.9 %) was obtained from Aldrich Inc. XRD patterns of
the standard compounds matched the appropriate Powder Diffraction File [3 8].
BET Surface Area. Surface area measurements were performed using a
QuantaChrome Quantasorb Jr. Sorption System. Approximately 0.1 grams of catalyst
Were outgassed in a N2/He mixture (5% N2) at 350 °C for l h prior to adsorption

measurements. The measurements were made using relative pressures of N2 to He of

131
0.05, 0.08, and 0.15 (N2 surface area = 0.162 nmz) at 77 K. Increasing the chromium

content of the catalysts decreases the BET surface area by a maximum of 10%.

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

Rigaku XRD diffractometer employing Cu K.ll radiation (it = 1.5418 A). The X-ray was
operated at 45 kV and 100 mA. Diffraction patterns were obtained using a scanning rate
of 0.5 deg/min (in 20) with divergence slit and scatter slit widths of 1°. Samples were run
as powders packed into a glass sample holder having a 20 x 16 x 0.5-mm cavity.

The mean crystallite sizes (El) of the Cr2O3 particles were determined from XRD

line broadening measurements using the Scherrer equation [39]:
d = K)» / [3 cos 9 (l)

where A 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 Cr2O3<104> line.

Quantitative X-ray diffraction data were obtained by comparing
Cr2O3<104>/AJ2O3<400> intensity ratios measured for catalyst samples with intensity
ratios measured for physical mixtures of pure Cr2O3, and y-Al2O3. This method assumed

that chromium addition did not disrupt the y-Al2O3 spinel and subsequently affect the
intensity of the A1203<400> line. The error in this method was estimated to be i 20%.
XPS Analysis. XPS data were obtained using a Perkin-Elmer Surface Science
il’lStrument equipped with a magnesium anode (1253.6 eV) operated at 300 W (15 kV, 20
mA) and a 10-360 hemispherical analyzer operated with a pass energy of 50 eV. The

Insttument typically operates at pressures near 1 x 10'8 torr in the analysis chamber.

rep
pre
WEI
min

usir

Sign;
p051

obta

La)
9)

lists
film

53131}

132

Spectra were collected using a PC137 board interfaced to a Zeos 386SX computer.
Samples were analyzed as powders dusted onto double-sided sticky tape or spray coated
onto a quartz slide using a methanol suspension of the catalyst. Binding energies for the
catalyst samples were referenced to the A1 2p peak (74.5 eV). The binding energies for
standard compounds that did not contain Al were referenced to the C ls line (284.6 eV)
of the carbon overlayer. XPS binding energies were measured with a precision of i 0.2
eV, or better.

Reduction of chromium species [18,27,40] during XPS experiments has been
reported and attributed to several factors such as X-ray flux, X-ray dose, temperature, and
pressure. In order to minimize the effect of photoreduction on the results, all samples
were analyzed using the same distance between X-ray source and sample (~3/4 inch) and
minimum data acquisition time (< 20 min.) No significant photoreduction was observed
using these experimental conditions.

The oxidation state distribution of chromium was determined by non-linear least-
squares curve fitting (NLLSCF) using the Cr2p3/2 envelope [41]. Chromium peak

positions were allowed to float slightly (i 0.1 eV) in order to fit the Cr2p3/2 envelope

obtained for each sample. The full width at half maximum (fwhm) of Cr(III) was fixed at

3 -3 eV (determined using Cr203 as a reference compound). Since the fwhm of Cr(V1)
measured for Cry catalysts did not agree with the value obtained for the 003 standard
Compound (2.] eV) and increased with decreasing Cr content, the fwhm of Cr(VI) was
allowed to float near 3.0 eV (i 0.3 eV). The fwhm of standard compounds and Cry

catalysts agree well with the values reported [42]. Scierka et al. [42] have examined the

ta.

dis

an;
fret

par

ilou
100 ll

abO't'e

133

Cr2p envelope of Cr/Al2O3 catalysts with principal component analysis and target testing.
The authors concluded that only Cr61 and Cr3+ components are observed on the Cr/Al203
catalyst.

Quantitative )G’S Analysis. It has been shown by Defosse et al. [43] that one may
calculate the theoretical intensity ratio I°,/I°, expected for a supported phase (p) atomically
dispersed on a carrier (5). An extension of the Defosse model proposed by Kerkhof and
Moulijn [44] has been used in the present investigation. The photoelectron cross sections
and the mean escape depths of the photoelectrons used in these calculations were taken
from Scofield [45] and Penn [46], respectively. For a phase (p) present as discrete

particles, the experimental intensity ratio Ip/I. is given by the following expression:

m. = 1°./1°.r1-expt-d/xm/ cm». (2)

where I°,,/1°s is the theoretical monolayer intensity ratio, d is the length of the edge of the
cubic crystallites of the deposited phase and XP is the mean escape depth of the
photoelectrons in the deposited phase. The Cr species particle sizes were determined by
solving the equation for d. The relative Cr dispersion to the monolayer was estimated
from the ratio of Ip/Is to 1°P/I°,.
CH4 Oxidation Activity. Measurement of CH; oxidation activity was performed in
a flow microreactor. Approximately 0.1 g of catalyst were supported on a glass frit (70 -
1 00 pm) and the temperature was measured with a K-type thermocouple located just

above the catalyst bed. The reactor was heated by a tube fumace (Lindberg) with

134

temperature being controlled within 1°C by an Omega CN 1200 temperature controller.
Reactant gas flow rates were held constant with Brooks 5850 mass flow controllers.
Product gases were analyzed with a Varian 920 gas chromatograph equipped with a TCD
and interfaced to a Hewlett-Packard 3394A integrator. Reaction products were separated
on a 6 ft 60/80 mesh Carbosieve S-II column. Prior to the first activity measurement, the
catalyst was pretreated with a mixture of 5% 02/He (99.5 % purity for 02, 99.995%
purity for He, AGA Gas Co.) stream (143 cc/min) at 350 °C for 1 hr to remove any
impurities adsorbed on the surface during catalyst preparation and storage. CH4 oxidation
reactions were performed with a constant flow rate (15 cm3/min) of 0.98 % CHa/5.25 %
O2/93.77 % He gas mixture (AGA, purity > 99.99%) in the temperature range of 3 10-450
°C. Water produced during methane oxidation was frozen downstream from the reactor
in a trap maintained < -40 °C with a mixture of ethanol and dry ice. All activity
measurements were obtained under steady-state conditions at conversions less than 15 %.
CO was observed as a partial oxidation product. The selectivity to CO2 was in the range
of 73-83 % over Cry catalysts at 10 % conversion. Turn over number (TON) was
calculated with CH; oxidation rate at 390 °C normalized by surface chromium atoms

determined from Cr content and dispersion.

5.4. Results and Discussion

Chemical State of Cr/A lumina Catalysts. XRD patterns obtained for Cry catalysts

with Cr/Al atomic ratios 3 0.080 showed only lines characteristic of the alumina carrier

135

(Figure 5.1). For higher Cr loadings (Cr/Al atomic ratios 2 0.107), peaks characteristic of
Cr2O3 were observed. The intensities of the 020; XRD peaks increase with increasing Cr
content. Quantitative XRD measurements (Table 5.1) show that the amount of crystalline
Cr2O3 increases from 0.9 to 5.5 wt.% as the Cr/Al atomic ratio increases from 0.107 to
0.134 (16 to 20 wt.% Cr2O3). Previous researchers have suggested that the support
becomes saturated with surface chromium oxide at a critical metal loading and further Cr
addition leads to the formation of crystalline Cr2O3 [22,47]. XRD results for the Cry
catalysts indicate that the critical chromium loading is approximately 15 wt.% Cr2O3. Zaki
et a1. [47] have reported that crystalline Cr2O3 forms at 10.2 wt.% Cr2O3 on 135 m2/g
A1203 support. This corresponds to 15.3 wt.% Cr2O3 after normalization for the surface
area of alumina used in our study.

Figure 5.2 shows the Cr 2p XPS spectra obtained for the Cry catalysts. The
average Cr 2p3/2 binding energies measured for the catalysts (579.7 3: 0.1 eV) are close to
the value determined for 00;. (579.4 eV) and higher than the values obtained for Cr2O3
(576.4 eV). The Cr 2p3,2 spectra also contain a shoulder at lower binding energy (577.0 1‘
0.2 eV) which is close to the value measured for Cr2O3. The intensity of the shoulder
increases with increasing Cr loading. The binding energies measured for the standard

compounds and catalysts agree well with values reported previously [l8,21,27,35,48].
Comparison of binding energies measured for Cry catalysts and standard compounds
indicates that the major peak and shoulder can be assigned to Cr(Vl) and Cr(III),
reSpectively. This is consistent with previous XPS studies of Cr/Al2O3 catalysts [18-21]
which showed that Cr(VI) species was present predominantly at low Cr concentration (< 5

wt - °/o Cr2O3 on 200 mZ/g A1203) and the concentration of Cr(IIl) species increased with

136

 

 

C1303
CrZO, <110>
<104>
Cr,03
<012>
C1303
_ <113>
‘ (C)
, " A120, (b)
<3 l> ' . .
A120,
A120, . 2)
<220> r
A120, (a)
<400>

l l l l L l

 

Figure 5.1.

25 30 35 4o 45 50
Degrees (20)

 

XRD patterns measured for (a) A1203, (b) Cr] 1, and (c) Cr13.

137

Table 5.]. Concentration of Crystalline' Phases in Cry Catalysts
Calculated from Quantitative XRD Data.

 

Crystalline Phase (wt%)

 

Catalysts As prepared Measured
Cr 1.3 2 3’
Cr 2.7 4 -b
Cr 5.4 8 3’
Cr 8.0 12 3’
Cr 11 16 0.9
Cr 13 20 5.5

 

 

 

a - Valid for crystalline phases with particle sizes > 3.0 nm.
b - No 0203 peaks detected forthese catalysts.

increasing Cr content. The variation of the Cr 2p3/2/Al 2p XPS intensity ratio of Cr(VI)
and Cr(III) species for the Cry catalysts determined using non-linear least-squares curve
fitting are shown in Table 5.2. The Cr(V1)/Al intensity ratio increases with increasing
Cr/Al atomic ratio up to 0.054. For higher Cr loadings, the Cr(VI)/Al intensity ratios
measured for the catalysts are independent of Cr content. This indicates that the alumina
support becomes saturated with Cr(VI) species at the catalyst with Cr/Al atomic ratio =
0.054 (8 wt.% Cr203) and further addition of chromium only leads to the formation of
Cr(III) species. Wachs and coworkers [22,33] have used the absence of Raman and XRD
peaks of 0203, the suppression of FTIR peaks of 0H, and C02 chemisorption to show
that monolayer coverage was reached at 12 wt.% Cr03 (for 180 mz/g A1203). This
corresponds to 10 wt.% Cr203 after normalization for chromium oxide molecular formula
and surface area of alumina used in our study. The discrepancy between the XRD and
3338 data for saturation point (15 wt.% and 8 wt.% Cr203/Al203, respectively) is due to

the different detection limits (selectivity) of the techniques. The saturation point

138

 

(f)

(d)

(C)

(b)

(a)

 

 

 

595 590 585 580 575

Binding Energy (eV)

Figure 5.2. Cr 2p XPS spectra obtained for (a) Cr1.3, (b) Cr2.7, (c) Cr5.4, (d) Cr8.0,
(e) Cr] 1, and (f) Crl3 catalysts.

139

Table 5.2. Cr 2p3/2/Al 2p XPS Intensity Ratios of Cr(VI) and
Cr(III) Species in Cry Catalysts Determined Using
Non-Linear Least-Squares Curve Fitting (NLLSCF).

 

Intensity Ratio

 

Catalysts Cr(VI)/Al Cr(III)/Al
Cr 1.3 0.21 0.05
Cr 2.7 0.31 0.13
Cr 5.4 0.57 0.25
Cr 8.0 0.54 0.45
Cr 11 0.52 0.64
Cr 13 0.51 0.75

 

d at ermined by XRD is based on large Cr203 crystallites (d > 3.0 nm), while well dispersed
Q r(VI) species are used for the determination derived from XPS. The Cr(III) species are
Ob served from the Cr1.3 catalyst and the Cr(III)/Al )G’S intensity ratio increases with
i h creasing Cr content (Table 5.2). The Cr(III) species on the low Cr content catalysts
(below monolayer coverage) has been detected by ESR [28-32] and XPS [18-20,42]
t echniques. The presence of Cr(IIl) species has been attributed to cluster formation [30]
61‘ calcination induced reduction [21]. The relative distributions of Cr(VI) and Cr(III)

Sli>ecies for the low Cr loading catalysts (e.g. 19 % Cr(III) for Crl.3) are similar to the
Values reported previously [19,42]. Scierka et al. [42] have found 17 % Cr(III) species
for the 2.5 wt.% Cr/Al203 (170 mz/g) catalyst which corresponds to 4.4 wt.% Cr203 for

alumina support used in this study. The firrther increase in Cr(III) content for Cr rich

e21talysts(Cr11 and Crl3) can be attributed to the formation of crystalline Cr203.

140

Dispersion of Cr/Alumina Catalysts. Figure 5.3 shows the variation of the Cr
2 p / A1 2p intensity ratio measured for the Cry catalysts as a function of the Cr/Al atomic
rat i o. The theoretical line calculated for monolayer dispersion [44] is shown for
co m parison. The Cr/Al intensity ratio measured for the Crl.3 catalyst is identical, within
exp erimental error, to the value predicted for monolayer dispersion. For higher Cr
1<>a.c1ings, the Cr/Al intensity ratios measured for the catalysts are lower than the
monolayer values. The increasing deviation of the Cr/Al intensity ratio from the
monolayer line with increasing Cr loading indicates that the Cr dispersion decreases
e0tttinuously with increasing Cr content [44]. We ascribed the decrease in Cr dispersion
t9 the formation of chromium oxide cluster or crystalline on the alumina support as a
fixt‘rction of Cr content.
Variation in the particle size of the chromium phase determined from Cr/Al XPS
i “tensity ratios using Eq. (2) is given as a function of the Cr loading in Table 5.3. The
mean particle size of the chromium phase calculated from XPS intensity ratios increases
fl‘om 0.5 to 1.9 nm with increasing Cr content. Particle sizes determined from XRD line
broadening calculations using Eq. (1) are also shown in Table 5.3. For catalysts with
Cr/Al atomic ratios 2 0.107, XRD line broadening calculations show that the 0203
particle sizes are significantly larger than the values determined using XPS. The difference
i h Cr203 particle sizes determined using XPS and XRD may be attributed to the limitations
of X-ray diffraction. It is well known that XRD determined particle sizes are skewed to
l arge values since XRD does not detect highly dispersed species. For the Cr rich catalysts,

a significant fraction of the chromium is present as a surface phase or small Cr203 particles

141

 

A
l

monolayer —’

l

u.)
l

N
j I
O
O
C

XPS Intensity Ratio (Cr2p/A12p)

u—d
l

 

 

 

O 1 1 U l r I I l l j I j 1
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14

Cr/Al Atomic Ratio

I:igure 5.3. Cr 2p/Al 2p XPS intensity ratios measured for Cry catalysts (0) plotted
versus Cr/Al atomic ratio. Cr 2p/Al 2p intensity ratios calculated for
monolayer dispersion (line).

0nd

load
mos
Cr(l

inCTl

142

 

 

Table 5.3. Particle Size and DiSpersion of Chromium Phases

Determined from XRD Line Broadening
Calculations and Cr/Al XPS Intensity Ratios.

Particle size (nm)
Catalysts XRD XPS Dispersionc (%)

Cr 1.3 J 3’ 100
Cr 2.7 -' 0.5 75.8
Cr 5.4 -' 0.8 67.6
Cr 8.0 -‘ 1.2 55.7
Cr 11 46 1.5 50.8
Cr 13 29 1.9 42.4

 

 

 

a - No Cr203 peaks detected for these catalysts.
b - Value corresponds to monolayer dispersion.
c - Calculated from XPS data.

((1 < 3.0 nm) that are detected by XPS but not XRD. Thus, smaller average chromium
QDrizide particle sizes are expected from XPS calculations.

Based on the XRD and XPS results presented above, the effect of chromium

1‘ Qading on Cry catalyst structure can be summarized as follows: (1) for low Cr loadings,

l"lfiost of the Cr is present as a highly dispersed Cr(VI) surface phase, (2) the fraction of

Cr(III) species present in the catalysts increases and the Cr dispersion decreases with
i l’Icreasing Cr content, and (3) Cr rich catalysts contain large Cr203 particles.

Effect of Catalyst Structure on CH4 Oxidation Activity. Table 5.4 shows the

SIZJeCific turnover number (TON) for CH4 oxidation over Cry catalysts calculated at 390 °C

using XPS estimates of the Cr dispersion. The TON increases approximately by a factor

of 3 as the Cr/Al atomic ratio increases from 0.013 to 0.11. Further addition of Cr

decreases the TON for CH4 oxidation. The activation energies for CH4 oxidation (Table

54) cal

kcaltnc

upon
moteasi
Colll)
loading
Ctlll)
enhanc:
several
PlOPOSI
ttachcn
additio:
for Ht
decreas
ElIItOUn
irate f
Nllanr
X0 rec

the (hp:

143

5 4 ) calculated for the Cry catalysts are similar within experimental error (20.2 i 0.9

1.: c alimol).
The variation in TON for CH4 oxidation as a function of Cr loading can be
ex p l ained in terms of the surface structure of the Cry catalysts. The increase in TON with
increasing Cr/Al ratio over the range 0.013 to 0.107 parallels a significant increase in the
Cr< III) content of the catalysts (Table 5.2). The inorease in TON as a function of Cr
10 ading up to Crll catalyst can be attributed to an increase in the formation of Cr(III)-
C—‘—'t‘(\/I) clusters that may be more active for CH4 oxidation than Cr(VI) species. The
e‘1'111anced activity of an alumina supported Cr(III)-Cr(VI) species has been reported for
S"F—E‘weral different reactions [28-32,47-52]. Ellison and coworkers [28,31,32] have
I) roposed that the mixed oxidation state cluster (y-phase) has high catalytic activity for
t‘e-Ezlctions that require both oxidation and reduction process (Zener mechanism). In
zi-Cldition, Zaki and coworkers [47,49,50] have reported that the y-phase was the active site
er H202 and propanol decomposition over chromium catalysts. For the Crl3 catalyst, the
Cl'Ecrease in TON relative to the Crll catalyst may be correlated to an increase in the
a~1'Tlount of large Cr203 particles present in the Cr rich catalyst. Large 020:, particles will
l”)ave fewer Cr(III)-Cr(VI) interfacial interactions compared to small Cr203 particles.
1\Tiiyama et al. [51,52] have reported that the y-phase was more active than bulk Cr203 for

NO reduction with NH3. They proposed that the metal - surface oxygen bond strength of

t he y-phase was weaker than that of bulk Cr203.

The c

firUCtl

Surfac
decree

mstal

Cilmm
Conn

Poorly

144

Table 5.4. Turn Over Numbers1 and Activation Energiesb of
Cry Catalysts for C114 Oxidation.

 

 

 

Catalysts TON (s’lxlOJ') Activation Energy
Cr 1.3 2.8 21.7
Cr 2.7 3.4 20.4
Cr 5.4 4.5 20.3
Cr 8.0 5.7 19.8
Cr 11 7.3 20.2
Cr 13 5.6 18.8
a - Calculated at 390 °C using exposed Cr species determined
from XPS.

b - Estimated from Arrhenius plots.

5.5. Conclusions

The combined use of several techniques to investigate the effect of Cr addition on the
S‘tnicture and CH4 oxidation activity of Cry catalysts leads to the following conclusions.
1. XPS data indicate that most of the Cr is present as a highly dispersed Cr(VI)
Starface phase in Crl.3 catalyst. The fraction of Cr(III) increases and the Cr dispersion
decreases with increasing Cr loading. Cr-rich catalysts (Crll and Crl3) contain large
Qtystallites of Cr203.
2. The TON for CH4 oxidation increases by a factor of 3 with increasing
c}u'omium loading up to the Crl 1 catalyst. This has been attributed to the formation of a
CI‘(III)-Cr(VI) cluster. Further addition of Cr decreases the TON due to the formation of

poorly dispersed Cr203 crystallites.

11.

12-

13-
14-

15-

145

5.6. References

Kummer, J. T. Prog. Energy Combust. Sci. 1980, 6, 177.
Armor, J. N. Appl. Catal. B 1992, 1, 221.

Zwinkels, M. F. M.; Jaras, S. G.; Menon, P. G., Griffin, T. A. Catal. Rev. -Sci.
Eng. 1993, 35, 319.

Dmuchovsky, B.; Freerks, M. C.; Zienty, F. B. J. Catal. 1965, 4, 577.
Kang, Y.-M.; Wan, B.-Z. Appl. Catal. A 1994, 114, 35.

Prasad, R.; Kennedy, L. A.; Ruckenstein, E. Combust. Sci. and Tech. 1980, 22,
271.

Manning, M. P. Hazardous Waste 1984, l, 41.
Weldon, J .; Senkan, S. M. Combust. Sci. and Tech. 1986, 47, 229.

Kolaczkowski, S. T.; Beltran, F.; Crittenden, B. D.; Jefferies, T. M. Process
Safety and Environmental Protection 1990, 68, 49.

Drago, R. S.; Petrosius, S. C.; Grunewald, G. C., and Brendley, Jr., W. H. In
Environmental Catalysis; Armor, J. N. Ed.; ACS Symposium Series 552; American
Chemical Society: Washington, DC, 1994; Chapter 28.

Agarwal, S. K; Spivey, J. 1.; Butt. J. B. Appl. Catal. 1992, 82, 259.

Johnson, A. J.; Meyer, F. G.; Hunter, D. I., Lombardi, E. F. Incineration of
polychlorinated biphenyls using a fluidised bed incinerator, RFP-3271, (Rockwell
Int.) 1981.

Shelef, M., Otto, K., Gandhi, H. J. Catal. 1968, 12, 361.

Wong, W. C.; Nobe, K. Ind. Eng. Chem. Prod. Res. Dev. 1986, 25, 179.

Simon, S.; van der Pol, A.; Reijerse, E. J.; Kentgens, A. P. M.; van Moorsel, G.-J.
M. P.; de Boer, E. J. Chem. Soc, Faraday Trans. 1995, 91, 1519.

Dwyer, F. G. Catal. Rev. 1972, 6, 261.

17-

ll 1E3: ..

TIL SEQ!

2C)

146

Kapteijn, F., Stegenga, S., Dekker, N. J. J., Bijsterbosch, J. W., Moulijn, J. A.
Catal. Today 1993, 16, 273.

Okamoto, Y.; Fujii, M.; Imanaka, T.; Teranishi, S. Bull. Chem. Soc. Jpn. 1976, 49,
859.

Jagannathan, K.; Srinivasan, A.; Rao, C. N. R. J. Catal. 1981, 69, 418.

Gorriz, O. F.; Corberén, V. C.; Fierro, J. L. G. Ind. Eng. Chem. Res. 1992, 3],
2670.

Rahman, A.; Mohamed, M. H.; Ahmed, M.; Aitani, A. M. Appl. Catal. 1995, 121,
203.

Vuurman, M. A.; Hardcastle F. D.; Wachs, I. E. J. Mol. Catal. 1993, 84, 193.
Hardcastle, F. D.; Wachs, I. E. J. Mol. Catal. 1988, 46, 173.

Vuurman, M. A.; Wachs, 1. E; Stufltens, D. J .; Oskam, A. J. Mol. Catal. 1993,
80, 209.

Vuurman, M. A.; Wachs, I. E. J. Phys. Chem. 1992, 96, 5008.

Weckhuysen, B. M.; Schoonheydt, R. A.; Jehng, J.-M.; Wachs, I. E.; Cho, S. J.;
Ryoo, R.; Kilistra, S.; Poels, E. J. Chem. Soc. Faraday Trans. 1995, 91, 3245.

Cimino, A.; De Angelis, B. A.; Luchetti, A.; Minelli, G. J. Catal. 1976, 45, 316.

Ellison, A.; Oubridge, J. O. V.; Sing, K. S. W. Trans. Faraday Soc. 1970, 66,
1004.

Ellison, A.; Sing, K. S. W. J. Chem. Soc., Faraday Trans. 1978, 74, 2017.
Ellison, A. J. Chem. Soc, Faraday Trans. 1, 1984, 80, 2567.

Ellison, A.; Sing, K. S. W. J. Chem. Soc. Faraday Trans. 1978, 74, 2807.
Ellison, A J. Chem. Soc, Faraday Trans. 1, 1984, 80, 2581.

Turek, A. M.; Wachs, I. E.; DeCanio, E. J. Phys. Chem. 1992, 96, 5000.

Vuurman, M. A.; Stuflcens, D. J.; Oskam, A.; Moulijn, J. A.; Kapteijn, F. J. Mol.
Catal. 1990, 60, 83.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

147

Scierka, S. J .; Houalla, M.; Proctor, A.; Hercules, D. M. J. Phys. Chem. 1995, 99,
1537.

Anderson, R. B.; Stein, K. C.; Feenan, J. J .; Hofer, L. J. E. Ind. Eng. Chem. 1961,
53, 809.

Kuznetsova, L. L.; Paukshtis, E. A.; Shkurina, G. P.; Shkrabina, R. A.;
Koryabkina, N. A.; Arendarskii, D. A.; Barannik, G. B.; Ismagilov, Z. R. Catal.
Today 1993, 17, 209.

Powder Diffraction File: Inorganic Phases, Joint Committee on Powder
Diffraction Standards, PA, 1983.

Klug, H. P.; Alexander, L. E., X-Ray Diffraction Procedures for Polycrystalline
and Amorphous Materials. lst.Ed. Wiley, New York, 1954.

De Angelis, B. A. J. Electron Spectrosc. Relat. Phenom. 1976, 9, 81.

Software provided by Dr. Andrew Proctor, University of Pittsburgh, Pittsburgh,
PA.

Scierka, S. J.; Proctor, A.; Houalla, M.; Fiedor, J. N.; Hercules, D. M. Surf
Interface Anal. 1993, 20, 901.

Deffosse, C .; Canesson, D.; Rouxhet, P. G.; Delmon, B. J. Catal. 1978, 51, 269.
Kerkhof, F. P. J. M.; Moulijn, J. A. J. Phys. Chem. 1979, 83, 1612.

Scofield, J. H. J. Electron. Spectrosc. Relat. Phenom. 1979, 8, 129.

Penn, D. R. J. Electron. Spectrosc. Relat. Phenom. 1976, 9, 29.

Zaki, M. 11.; Fouad, N. E.; Leyrer, J.; Knézinger, H. Appl.Catal. 1986, 21, 359.

Griinert, W.; Shpiro, E. S.; Feldhaus, R.; Anders, K. Antoshin, G. V.; Minachev,
KH. M. J. Catal. I986, 100, 138.

Fahim, R. B.; Zaki, M. I.; Gabr, R. M. Surf Technol. 1981, 12, 317.

Fahim, R. B.; Zaki, M. 1.; Gabr, R. M. Appl. Catal. 1982, 4, 189.

148

51. Niiyama, H.; Murata, K.; Ebitani, A.; Echigoya, E. J. Catal. 1977, 48, 194.

52. Niiyama, H.; Murata, K.; Echigoya, E. J. Catal. I977, 48, 201.

Chapter 6

Characterization and CO Oxidation Activity of Cu/Cr/Al2O3 Catalysts

6.1. Abstract

X-ray photoelectron spectroscopy (XPS) and X-ray diffraction (XRD) have been
used to characterize a series of Cu/Cr/Al203 catalysts. The catalysts were prepared by
stepwise incipient wetness impregnation of chromium first, followed by copper. The
copper loading was held constant at 8 wt.% CuO and chromium loadings were varied
from Cr/Al atomic ratios from 0 to 0.0134 (0 to 20 wt.% Cr203). The information
obtained from surface and bulk characterization has been correlated with CO oxidation
activity of the Cu/Cr/Al203 catalysts. The copper addition decreased the dispersion of
chromium phase by reacting selectively with a dispersed Cr3i species to form large
crystalline CuCr2Oa. The Cu dispersion also decreased with increasing Cr loading. For
low Cr loading catalysts (Cr/Al g 0.027), CO oxidation activity increased with the addition
of Cr due to the formation of crystalline CuO on the chromium modified alumina. This
has been attributed to the prevention of Cu ion diffusion to alumina lattice vacancies by
highly dispersed chromium species. The Cu/Cr/Al203 catalyst of Cr/Al = 0.054 showed

the highest C0 oxidation activity due to the formation of an active CuCr204 phase. For

further addition of Cr (Cr/Al 2 0.080), the catalytic activity for C0 oxidation decreased.
149

150
This might be due to the decrease in active phase dispersion and encapsulation of the

active site with excess Cr species.

6.2. Introduction

The combination of copper and chromium oxide was recognized as a promising
active component for emission control reactions in the early years of pollution control
catalysis research [1-5]. The Cu-Cr catalyst system has been studied extensively for
various reactions, including CO oxidation [6-21], hydrocarbon oxidation [6-10,22-23],
N0 reduction [19-20,24], alcohol and aldehyde oxidation [25-26], and sulfumated [27]
and chlorinated [28] hydrocarbon oxidation. Shelef et al. [4] and Kapteijn et al. [20] have
performed a screening study of transition metal oxide catalysts for C0 oxidation and N0
reduction. They have reported that Cu-Cr catalysts showed consistently higher catalytic
activity than single oxide catalysts based on Cu, Ni, Co, Fe, Mn, Cr, or V. Barnes [6]
studied automotive pollution control catalysts for exhaust treatment such as CO and
hydrocarbon oxidations. He reported that the alumina supported 8 wt.% Cu-7 wt.% Cr
catalyst showed similar activity to 0.3 wt.% Pt catalyst. In addition, Stegenga et al. [29]
have found that monolith supported 10 wt.% Cu-Cr/Al203 catalyst showed a three-way
catalytic behavior (CO and hydrocarbon oxidation, and N0 reduction, simultaneously)
comparable to noble metal catalysts Operated under the same conditions.

In the copper chromite catalyst system, copper oxide has been proposed as an

active component for C0 oxidation reaction, while chromium oxide has been considered

151
as a promoter [13, 19,23,30]. The roles of Cr promoter to the Cu species are believed to

limit catalyst reduction [12-14,24,31], prevent catalyst poisoning [10,12,15], inhibit bulk
copper aluminate formation [30], improve thermal stability [10], and increase catalyst
dispersion [17,30]. Furthermore the mixed oxide catalyst shows greater activity and
stability than the single oxide catalysts [32,33]. The synergistic effect of Cu-Cr catalysts
has been attributed to the electronic interaction between copper and chromium species
[13,17,23].

Despite a considerable amount of effort devoted to the study of Cu-Cr catalyst, the
understanding of Cu-Cr catalyst system is still not clear. Stegenga et al. [29] reported that
the optimum metal composition of Cu-Cr/Al203 catalyst was Cu/Cr = 2 for CO oxidation.
However, Severino et al. [12] found that Cu/Cr = 1 was a optimal metal composition of
alumina supported catalyst for C0 oxidation. Also, uncertainty has arisen with regard to
the pretreatment for the most efficient catalyst. Severino et a1. [12] reported that catalyst
prereduction enhanced catalytic activity efficiently for CO oxidation. However, Chien et
al. [23] showed that oxidative pretreatment of Cu-Cr catalyst was more beneficial than
reductive pretreatment for CO oxidation.

Much of the previous research on the Cu-Cr catalysts have focused on the
understanding of catalytic properties such as activity or stability. Little effort has been
devoted to the relationship between surface structure of Cu/Cr/Al203 and its catalytic
activity. The present work is a part of a broad study to investigate the effect of base metal
oxide promoters on the structure and reactivity of copper oxide based emission control

catalysts. In this paper, X-ray photoelectron spectroscopy (XPS) and X-ray diffraction

152
(XRD) were used to determine the effect of Cr loading on the chemical state and

dispersion of chromium and c0pper oxide phases supported on y-alumina. The
information derived from these techniques has been correlated with C0 oxidation activity

to develop a more complete understanding of the Cu/Cr/Al203 catalyst.

6.3. Experimental

Standard Materials. Cr03 (99.9 %) was obtained from Aldrich Chemical
Company, Inc. CuO and Cr203 were prepared by calcining metal nitrates in air at 500 °C
for 16 h. CuCr2Oa and CuAl204 were prepared by calcining stoichiometric mixtures of the
respective nitrates in air at 1000 °C. for 24 h. XRD patterns of the standard compounds
matched with the appropriate Powder Diffraction File [34].

Catalyst Preparation. Catalysts were prepared by pore volume impregnation of y-
alumina (Cyanamid, y-A1203, surface area = 203 mZ/g, pore volume = 0.6 mL/g). The
alumina support was finely ground (< 230 mesh) and calcined in air at 500 °C for 24 h
prior to impregnation. Catalysts were prepared using deionized water solutions of
chromium(III) nitrate (Mallinckrodt, Analytical Reagent) and copper(II) nitrate
(Columbus Chemical Industries Inc, ACS Grade). The impregnated catalysts were dried
in air at 120 °C for 24 h and subsequently calcined in air at 500 °C for 16 h. Mixed metal
Oxide catalysts (Cu/Cr/Al203) were prepared by the step wise impregnation of chromium

first, followed by Cu. Chromium impregnated alumina (Cr/A1203) was dried and calcined

153
prior to the introduction of copper. Catalysts with added copper were dried and calcined

again under the same conditions. The copper content was held constant at 8 wt.% (as
CuO) of the alumina support. The chromium content was varied from an Cr/Al atomic
ratio of 0 to 0.134 (0 to 20 wt.% Cr203). Catalyst samples will be designated by "Cry"
and "CuCry", where y is the Cr/Al atomic ratio (x 102).

BET Surface Area. Catalysts surface area were determined using a QuantaChrome
Quantasorb Jr. Sorption System. Approximately 0.1 g of catalyst was outgassed in a
N2/He mixture (5% N2) at 350 0C for 1 h prior to adsorption measurement. The
measurement was 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 BET surface areas of catalyst samples decreased
with increasing catalyst loading by a maximum of 20 %.

X-Ray Diffraction. X-ray powder diffraction patterns were obtained with a
Rigaku XRD diffractometer employing Cu Ka radiation (A = 1.5418 A). The
diffractometor was operated at 45 kV and 100 mA. Diffraction patterns were obtained
using a scan rate of 0.5 deg/min. with divergence slit and scatter slit widths of 1°. Samples

were run as powders packed into a glass sample holder having a 20 x 16 x 0.5-mm cavity.

The mean crystalline sizes (d) of the CuO, Cr203. and CuCr204 particles were determined

from XRD line broadening measurement using the Scherrer equation [3 5]:

d=Kiu/Bc059 (1)

154
where A is the X-ray wavelength, K is the particle shape factor, taken as 0.9, and B is the

fill] width at half maximum (fwhm), in radians, of the CuO<T11>, Cr203<104>, or

CuCr204<31l> lines. Quantitative X-ray diffraction data were obtained by comparing

CuO<T11>, Cr203<104>, or CuCr2Oa<3 1 1> vs A1203<400> intensity ratios measured for
catalyst samples with intensity ratios measured for the physical mixtures of pure CuO,
Cr203, or CuCr204 and y-Al203. This method assumes that copper and chromium addition
does not disrupt the y-Al203 spinel and subsequently affect the intensity of the
AJ2O3<400> line. The error in this method was estimated to be 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 300 W (15 kV, 20
mA) and a 10-360 hemispherical analyzer operated with a pass energy of 50 eV. The
instrument typically was operated at pressures near 1 x 10'8 torr in the analysis chamber.
Spectra were collected using a PC137 board interfaced to a Zeos 386SX computer.
Samples were analyzed as powders dusted onto double-sided sticky tape or spray coated
onto a quartz slide using a methanol suspension of the catalyst. Binding energies for the
catalyst samples and CuAl2Oa were referenced to the Al 2p peak (74.5 eV). The binding
energies for standard compounds with no A1 were referenced to the C Is line (284.6 eV)
of the carbon overlayer. XPS binding energies were measured with a precision of i 0.2
eV, or better.

Reduction of copper [36-38] and chromium [39-41] species during the XPS
experiment has been reported and attributed to several factors such as X-ray flux, X-ray

dose, temperature, and pressure. In order to minimize the effect of photoreduction on the

155

results, all samples were analyzed using the same X-ray flux, a constant distance between
the X-ray source and the sample, and a minimum data acquisition time.

Quantitative XPS Analysis. It has been shown by Defosse et al. [42] that one may
calculate the theoretical intensity ratio I°,/I°. expected for a supported phase (p)
atomically dispersed on a carrier (5). An extension of the Defosse model proposed by
Kerkhof and Moulijn [43] has been used in the present investigation. The photoelectron
cross sections and the mean escape depths of the photoelectrons used in these calculations
were taken from Scofield [44] and Penn [45], respectively. For a phase (p) present as

discrete particles, the experimental intensity ratio IP/I, is given by the following expression:

1px], = I°,,/1°,[1-exp(-d/r.,)]/ do.p (2)

where I°,,/I°, is the theoretical monolayer intensity ratio, d is the length of the edge of the
cubic crystallites of the deposited phase and 2,, is the mean escape depth of the
photoelectrons in the deposited phase.

Curve Fitting. XPS data manipulation performed using ‘GOOGLY’ PC software
[46]. The Cr 2p3/2 range was used to measure Cr binding energies and a relative oxidation
state distribution by non-linear least-squares curve fitting (NLLSCF). The Cr peak
positions were allowed to float slightly (i 0.1eV) in order to fit into the Cr 2p3/2 envelope
of each sample. The full width at half maximum (FWHM) of Cr(IIl) was fixed as 3.3 eV,
a value that was determined from the line width for Cr203. The Cr(VI) FWHM of

Cr/Al203 catalysts did not agree with the value for 003 (2.1 eV). Also, the width

156
increased with decreasing Cr content. The FWHM of Cr(VI) was allowed to float from

3.0 :l: 0.3 eV. Those FWHM of standard compounds and Cr/Al203 catalysts are well
agreed with the values reported by Scierka et a1 [41].

Catalytic Activity Measurement The measurement of CO oxidation activity was
performed in a flow microreactor (1/2 inch glass tube). Approximately 0.03 g of catalyst
was supported on a glass frit (70 - 100 um) and the temperature was measured with a K-
type thermocouple located just above the catalyst bed. The reactor was heated by a tube
furnace (Lindberg) with temperature being controlled within 1 °C by an Omega CN 1200
temperature controller. Reactant gas flow rates were held constant with Brooks 5850
mass flow controllers. Product gases were analyzed with a Varian 920 gas
chromatography equipped with a TCD and interfaced to a Hewlett-Packard 3394A
integrator. Reaction products were separated on a 6 ft 60/80 mesh Carbosieve S—II
column. Prior to the first activity measurement, the catalysts were pretreated with a
mixture of 5% 02/He (99.5 % purity for 02, 99.995% purity for He, AGA Gas Co.)
stream (143 cc/min) at 350 °C for 1 hr to remove any impurities adsorbed on the surface
during catalyst preparation and storage. CO oxidation reactions were performed with a 80
cm3/min flow of 4.8% C0/9.8% 02/85.4% He gas mixture (AGA, purity > 99.99%). The
reaction rate was calculated from the TCD sensitivity corrected CO and CO2
chromatographic response of the reactor output. All activity measurements were obtained
under steady-state conditions at conversions less than 15 %. The specific turn over

number (TON) was calculated from the C0 oxidation rate at 240 °C which was

157

normalized with respect to Cu dispersion as estimated by XPS data. The activation energy

was determined from an Arrhenius plot for CO oxidation.

6.4. Results and Discussion

Chemical Composition of Cu/Cr/AI2O3 Catalysts

Standard Compounds. The XPS Cu 2p3/2 binding energies and satellite/main peak
intensity ratios for standard compounds are shown in Table 6.1. CuA1204 and CuCr204
have higher binding energies and satellite/main peak intensity ratios than the values
obtained for CuO. Those binding energies and satellite/main peak intensity ratios of the
standard compounds are consistent with previously reported values [3 8,47,48].

Ctr/A1203 Catalyst. The XPS Cu 2pm binding energy measured for the CuCrO
catalyst is higher than the value obtained for CuAlgOa and CuO, but closer to the value of
CuAl204 (Table 6.1). Figure 6.1 shows the XRD patterns obtained for alumina support
and CuCry catalysts. The XRD pattern of the CuCrO catalyst shows only lines
characteristic of alumina. The XPS and XRD results indicate the predominant presence of
dispersed copper ‘surface spinel’ phase on the alumina support [38,48]. However, the
satellite/main peak ratio measured for the CuCrO catalyst is closer to the value of CuO
than to that of CuAl2Oa. This result conflicts with XRD data which indicates the absence
of crystalline CuO for the CuCrO catalyst. Strohmeier et al. [38] have explained this

discrepancy that well dispersed or amorphous CuO was not necessarily present on the

158

Table 6.1. XPS Cu 2p3/2 Binding Energies and Shakeup/Main
Peak Intensity Ratios Measured for CuCry Catalysts
and Standard Compounds.

 

Cr/Al Atomic Ratio (x103) BE of Cu 2pm shakeup/main peak ratio

 

0.0 935.4 0.42
1.3 935.3 0.48
2.7 935.5 0.56
5.4 ‘ 935.5 0.62
8.0 ' 935.5 0.63
10.7 935.5 0.69
13.4 935.5 0.72
CuO 933.9 0.45
CuAlgOa 935.0 0.71
CuCrgOa 934.5 0.70

 

catalyst, but the surface spinel copper species might be chemically different from both bulk
CuO and CuAl2Oa.

Cu/Cr/AlgO3 Catalysts- The average XPS Cu 2p3,2 binding energy measured for
CuCry catalysts (935.5 : 0.1. eV) is identical, within experimental error, to the value
obtained for CuCrO catalyst (Table 6.1). This indicates that the addition of chromium had
little effect on the XPS Cu 2p32 binding energies measured for the catalysts.

Figure 6.2 shows the Cr 2p XPS spectra of CuCry catalysts. The Cr 2p3/2 binding
energies of a main peak (580.0 : 0.1 eV) and a shoulder (577.4 i 0.1 eV) measured for
the catalysts has been assigned to the Cr(VI) and Cr(III) species, respectively [49]. The
relative distribution of Cr(VI) and Cr(III) oxidation states in CuCry catalysts indicates that

Cr(III) species increase with increasing Cr content (Table 6.2). The XPS binding energies

159

 

 

 

 

+
+
+
+
(h)
(g)
(f)
(e)
' . (d)

 

 

Figure 6.1. '

25 30 35 40 45 50

Degrees (20)

XRD patterns measured for CuCry catalysts ( O alumina, l CuO, 0 cubic
CuCr204, V tetragonal CuCr2Oa, ‘f‘r’ Cr203). (a) alumina (b) CuCrO, (c)
CuCr1.3, (d) CuCr2.7, (e) CuCr5.4, (f) CuCr8.0, (g) CuCr10.7, and (h)
CuCr13.4.

160

 

(f)

(6)

(d)

(0)

(b)

 

 

 

595 590 585 580 575

Binding Energy (eV)

Figure 6.2. Cr 2pm XPS spectra measured for CuCry catalysts. (a) CuCr1.3, (b)
CuCr2.7, (c) CuCr5.4, (d) CuCr8.0, (e) CuCr10.7, and (f) CuCr13.4.

161
and relative oxidation state distribution of Cr species for the CuCry catalysts are consistent

with the results obtained for Cry catalysts within experimental error [49]. This indicates
that the post-addition of Cu does not affect Cr oxidation states in CuCry catalysts.

Our previous study [49] indicated that the chromium oxide structure of Cry
catalysts depended on the Cr content. Three levels of Cr loading will be discussed the
structure of CuCry catalysts: low Cr loading (0.013 S Cr/Al atomic ratios 3 0.027),
intermediate Cr loading (0.054 s Cr/Al g 0.080) and high Cr loading (Cr/Al 2 0.107).

For CuCry catalysts with low Cr loading, the XRD patterns show peaks
characteristic of CuO (Figure 6.1). Table 6.3 shows the amount of crystalline CuO,
tetragonal CuCr2Oa, and Cr203 present in the CuCry catalysts as a function of Cr/Al
atomic ratio. The amount of crystalline CuO increases from 0.4 to 1.9 wt.% as the Cr/Al
atomic ratio increases from 0.013 to 0.027. The formation of CuO phase can be explained
in terms of a decrease in interaction between copper species and alumina support [50]. A
significant number of the lattice vacant sites normally accessible to the Cu2+ ions during
catalyst preparation steps are blocked by pre-loaded chromium oxide species (well
dispersed Cr6’ species). Thus, Cui‘ ions are not able to enter the alumina lattice to form
the copper surface spinel phase. Upon calcination, these Cu} ions agglomerate to form
large particles of CuO. The formation. of a CuO phase on the Cr-modified alumina
indicates that the copper species does not readily interact with the Cr°+ species to form
CuCr204 phase. This is consistent with the result of the relative distribution of Cr(VI) and
Cr(III) oxidation states in CuCry catalysts which were similar to the values of Cry

catalysts [49]. For CuCry catalysts with intermediate Cr loading, the XRD patterns show

162

Table 6.2. Distribution of Cr Oxidation States Determined from
Cr 2p322 XPS Spectra of CuCry Catalysts Measured
by Non-Linear Least-Squares Curve Fitting

 

 

 

(NLLSCF).
Cr/Al Atomic Ratio Relative concentration (%)
(x103) Cr(VI) Cr(III)
1.3 79 21
2.7 77 23
5.4 73 27
8.0 57 43
10.7 54 46
13.4 47 53
Table 6.3. Concentration of Crystalline“ Phases in CuCry

Catalysts Calculated from Quantitative XRD Data.

 

 

Cr/Al Atomic Weight Percent of Crystalline Phase

Ratio (x103) Cr203 CuO CUCT304
0.0 b 0.0 b
1.3 0.0 0.4 0.0
2.7 0.0 1.9 0.0
5.4 0.0 0.0 0.3
8.0 0.0 0.0 3.4
10.7 1.3 0.0 3.4
13.4 8.1 0.0 3.1

 

a - Valid for crystalline phases with particle sizes > 3.0 nm.
b - Catalyst does not contain chromium.

163
peaks characteristic of tetragonal CuCr204 (Figure 6.1). In addition, a peak observed at

359° in the diffraction pattern of the CuCr5.4 catalyst can be assigned to a cubic CuCr204
<311>. The crystalline tetragonal CuCr204 content of the catalysts increases from 0.3 to
3.4 wt.% as the Cr/Al atomic ratio increases from 0.054 to 0.080 (Table 6.3). For
CuCr5.4 and CuCr8.0 catalysts, the XRD patterns indicate that CuCr204 is the major
crystalline phase. This is consistent with the increase of the Cu 2p3/2 XPS satellite/main
peak ratio up to 0.63 (Table 6.1).

For CuCry catalysts with high Cr loading, the peaks characteristic of crystalline
Cr203. are observed in the XRD patterns (Figure 6.1). The crystalline Cr203 content of the
catalysts increases from 1.3 to 8.1 wt.%, however, the crystalline CuCr204 content is
independent of the Cr addition as the Cr/Al atomic ratio increases from 0.107 to 0.134
(Table 6.3). The Cu 2p3,2 XPS satellite/main peak intensity ratios for the high Cr loading
catalysts (Table 6.1) are close to the value obtained for CuCr204. This result indicates
that most of the Cu reacts with Cr species to form CuCr204 instead of interacting with the
alumina support. The comparable Cr203 content obtained for the CuCry and the
analogous Cry catalysts [49] suggests that little CUCT304 phase is formed by the
interaction of Cu with large Cr203 crystallites. This is readily understood when one
considers that large Cr2O3 crystallites have relatively less surface area available to react

with Cu species than small Cr203 crystallites.

164
Dispersion of Cu/Cr/Al203 Catalysts.

Copper species. Figure 6.3 shows the variation of the Cu ZpyMAl 2p XPS
intensity ratios measured for the CuCry catalysts as a function of Cr/Al atomic ratio. The
decrease in the XPS intensity ratios with increasing Cr content indicates a decrease of Cu
species dispersion in CuCry catalysts as a function of Cr loading [43]. However, it should
be noted relatively high Cu 2p322/Al 2p XPS intensity ratio observed for CuCr5.4 catalysts.
This is probably due to the formation of CuCr204 phase from well dispersed Cr oxide
species on the alumina support. The variation in copper particle size as a function of Cr
content determined from Cu/Al XPS intensity ratios and XRD line broadening is shown in
Table 6.4 and 6.5, respectively. The particle size of Cu species in CuCr0 catalyst (1.0 nm)
calculated fi'om XPS data indicates that Cu is well dispersed over alumina carrier. This is
consistent with XRD result which does not show any crystalline phase. The Cu phase
particle sizes calculated from XPS data increase from 1.0 to 2.1 nm with increasing Cr
loadings. This result agrees with XRD data that the formation of large CuO and CuCr204
crystallites is observed in the CuCry catalysts with increasing Cr content. CuO and
CuCr204 particle sizes determined using XRD line broadening calculations are significantly
larger than the particles sizes determined from XPS data. The difference in particle size
determined from XPS and XRD results may be understood in terms of the detection limit
of XRD. Since XRD detects large particles (> 3 nm) exclusively, the particle sizes
calculated from XRD line broadening result have tendency to be large values.

Chromium species. Figure 6.4 shows the variation of the Cr 2p/Al 2p XPS
intensity ratios measured for CuCry catalysts (open circles) as a firnction of Cr/Al atomic

ratio. Cr 2p/Al 2p XPS intensity ratios obtained for the analogous Cry catalysts (solid

165

 

 

 

 

1.1
1.0 C
2.5-
0.9 -
3
Q 0
a: It
'2
g 0 8 - .
.9. 0
35? 0.7 -
U)
fi
8
E. ‘ 0
0.6 -
' o
0.5 U T U l I I I l I ' U I I
0.00 0.02 0. 04 0.06 0.08 0. 10 0.12 0.14
Cr/Al Atomic Ratio

Figure 6.3. Cu 2p3,2/Al 2p XPS intensity ratios of CuCry catalysts plotted versus Cr/Al
atomic ratio. Monolayer dispersion value = 2.28.

166

Table 6.4. Particle Sizes of Chromium and Copper Species
Determined from XPS Intensity Ratios.

 

 

Cr/Al Atomic Particle Size (nm)

Ratio (x103) Chromium Copper
0.0 a 1.0
1.3 b 1.3
2.7 0.7 1.5
5.4 1.2 1.4
8.0 2.0 1.8
10.7 2.0 2.1
13.4 2.5 2.1

 

a - Catalyst does not contain chromium.
b - Catalyst dispersion is close to monolayer.

Table 6.5. Particle Sizes of CuO, Cr203. and CuCrgO. Phases
Determined from XRD Line Broadening

 

 

Calculations.
Cr/Al Atomic Particle Size(nm)
Ratio (x103) CuO Cr203 CUCl'gos
0.0 a b b
1.3 17 a a
2.7 16 a a
5.4 a a 31
8.0 a a 29
10.7 a 44 30
13.4 a 33 28

 

a - No diffraction lines detected.
b - Catalyst does not contain chromium.

167

 

 

 

 

5

4-
a . monolayer —’
N
3‘
o.
o 3-
C.)
v
.2 3
Ei
Dd
€2- o
r: C
2
E. . O O 0
g o

1- 0 O

0
0 I I I l I l I l I I I I fl

0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14

Cr/Al Atomic Ratio

Figure 6.4. Cr 2p/Al 2p XPS intensity ratios of Cry (0) and CuCry(O) catalysts
plotted versus Cr/Al atomic ratio. Cr 2p/Al 2p intensity ratios calculated
for monolayer dispersion (line).

168
circles) [49] and the theoretical line calculated from the Kerkhof and Moulijn model [43]

for monolayer dispersion are also shown for comparison. The Cr/Al intensity ratios
measured for the CuCry catalysts with Cr/Al atomic ratio 3 0.027 are similar to the values
measured for Cry catalysts. The absence of XRD peaks characteristic of discrete Cr
phases and the similarity of the Cr/Al intensity ratios measured for the catalysts with the
values calculated for monolayer dispersion indicate that Cr is highly dispersed over the
alumina carrier. For the CuCry catalysts with Cr/Al atomic ratio 2 0.054, the Cr/Al
intensity ratios are lower than the values obtained for the analogous Cry catalysts. This
indicates that Cu addition decreases the dispersion of the Cr species. This is consistent
with XRD results which showed large crystalline CUCT304 and G203 phases for the
catalysts with Cr/Al atomic ratio 2 0.054. The variation in Cr phase particle size as a
function of Cr loading determined from Cr/Al XPS intensity ratios and XRD line
broadening is shown in Tables 6.4 and 6.5, respectively. Cr particle sizes calculated from
XPS data generally increase with increasing Cr loading. As for the copper species, the
Cr2O3 particle sizes determined for CuCry catalysts using XRD line broadening

calculations are significantly larger than the values determined using XPS.

Effect of Catalyst Structure on C0 Oxidation Activity.

Table 6.6 shows the specific turn over numbers (TON) and activation energies of
CuCry catalysts for C0 oxidation as a function of Cr/Al atomic ratio. The specific TON
increases by a factor of 8 with increasing the Cr/Al atomic ratio from 0.0 to 0.054.

However, further addition of Cr decreases the specific TON for CO oxidation. The

169

Table 6.6. Turn Over Numbers' and Activation Energies” of
CuCry Catalysts for CO Oxidation.

 

 

 

Cr/Al Atomic TON Activation Energy
Ratio (x102) (s'lx102) (kcal/mol)

0.0 4 17.9

1.3 15.3 16.5

2.7 24.3 14.4

5.4 31.5 15.8

8.0 16.2 16.2

10.7 19.7 17.5

13.4 15.2 16.3

a— Determined from reaction rate at 240 °C normalized by Cu

dispersion.

b- Estimated from Arrhenius plots within 15 % conversion.

activation energies of CuCry catalysts are independent of Cr content (16.4 i 1.1
kcal/mol). The C0 oxidation activity can be attributed predominantly to the Cu species in
CuCry catalysts, since Cr species show a little CO oxidation activity compared to Cu
catalysts [6,13,19,23]. In this study, Cry catalysts showed negligible CO oxidation activity
compared to CuCr0 catalyst at the same reaction condition. Addition of Cr up to a Cr/Al
atomic ratio of 0.027 increases the TON by a factor of 6 compared to the CuCrO catalyst.
We attributed this to the formation of an active CuO phase which was observed in XRD
patterns for CuCrl.3 and CuCr2.7 catalysts. In our previous work [51], we concluded
that crystalline CuO was more active for C0 oxidation reaction than Cu surface phase.
This has been attributed to that redox ability of CuO/Al2O3 was more favorable than

dispersed Cu surface phase during CO oxidation reaction [12].

170
It should be noted that CuCr5.4 catalyst shows the highest specific TON among

the CuCry catalysts, although Cu and Cr dispersion decreases continuously as a function
of Cr content. This has been correlated with the formation of CuCr204 phase on the
CuCr5.4 catalyst which was detected by XRD. This is consistent with that satellite/main
peak ratio of CuCr5.4 catalyst increases to 0.62 which is close to the value of pure
CuCr204 (0.70) (Table 1). Therefore we attributed the activity improvement of CuCr5.4
catalyst to the formation of active phase (CuCr204). CuCr204 has been known more
active catalyst than CuO for CO oxidation [4,13,19]. Severino and Laine [13] have
studied alumina supported and unsupported Cu-Cr catalysts for CO oxidation. They
suggested that chromium might enhance the catalytic activity by providing structure
oxygen (02') to copper species and electron transfer from copper to chromium. Kapteijn
et a1. [19] have also proposed recently that the Cu-Cr oxide catalyst was more active than
a single oxide for CO oxidation due to more available of surface oxygen on the mixed
oxide catalyst.

The further addition of Cr (Cr/Al atomic ratio 2 0.080) decreases C0 oxidation
activity for CuCry catalysts. XPS and XRD data indicate that the dispersions of both Cu
and Cr are decreased significantly and a large CuCr204 crystalline phase is formed on the
catalyst. In addition, excess Cr203 phase, which is not active for CO oxidation reaction, is
present on the catalyst surface. Thus, low dispersion and covering active sites by excess

Cr phase may decrease catalytic activity for high Cr content CuCry catalysts.

171
6.5. Conclusions
The combined use of several techniques to investigate the effect of Cr addition on the
structure and C0 oxidation activity of Cu/Cr/Al203 catalysts leads to the following
conclusions.

1. Cu addition decreased Cr dispersion in CuCry catalysts (Cr/Al 2 0.054)
compared to the Cr dispersion in analogue Cry catalysts. The dispersion of Cu species in
CuCry catalysts also decreased with increasing Cr content. The decrease in dispersion of
both Cu and Cr species has been attributed to the formation of large crystalline CuO,
CuCr2Oa, and Cr203 on the alumina support.

2. For CuCry catalysts with low Cr contents (Cr/Al g 0.027), the presence of
highly dispersed Cr6+ phase inhibited the formation of copper surface phase and enhanced
the formation of crystalline CuO over the alumina support. The CO oxidation activity
increased in these catalysts (CuCrl.3 and CuCr2.7) has been attributed to an increase in
the amount of crystalline CuO.

3. The CuCr5.4 catalyst showed the highest CO oxidation activity in this study.
This is due to the formation of active CuCr204 phase. Post loaded copper species
interacted with well dispersed Cr?" species to form CuCr204 rather than with highly
dispersed Créi species or large Cr203 crystallites.

4. The CuCry catalysts with high Cr contents (Cr/Al 2 0.08) showed the decreased
CO oxidation activity. This has been attributed to a decrease in the dispersion of Cu and

Cr species and encapsulation of the active site with excess Cr203.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

172

6.6. References

Frazer, J. C. W. US. Pat. 2031475, 1936.

Roth, J. F.; Doerr, R. C. Ind. Eng. Chem. 1961, 53, 293.

Hofer, L. J. E; Gussey, R; Anderson, R. B. J. Catal. 1964, 3, 451.

Shelef, M.; Otto, K.; Gandhi, H. J. Catal. 1968, 12, 361.

Dwyer, F. G. Catal. Rev.- Sci. Eng. 1972, 6, 261.

Hertl, W.; Farrauto, R. J. J. Catal. 1973, 29, 352.

Yu Yao, Y. F.; Kummer, J. T. J. Catal. 1977, 46, 388.

Barnes, G. J. Adv. Chem. Series 1975, 143, 72.

Farrauto, R. J.; Hoekstra, K. E; Shoup, R. D. U. S. Patent 3,870,658; 1975.
Yu Yao, Y. F. J. Catal. 1975, 39, 104.

Kummer, J. T. Adv. Chem. Series 1975, 143, 178.

Severino, F.; Brito, J .; Carias 0.; Laine,J. J. C atal. 1986, 102, 172.

Severino, F .; Laine, J. Ind. Eng. Chem. Prod. Res. Dev. 1983, 22, 396.

Laine, J .; Albomoz, A.; Brito, J .; Carias, 0.; Castro, G.; Severino, F.; Valera, D.
In Catalysis and Automotive Pollution Control; Crucq, A. and Frennet, A. Ed.;
Elsevier, Amsterdam, 1987; pp 387-393.

Laine, J.; Severino, F. Appl. Catal. 1990, 65, 253.

L6pez Agudo,A., Palacios, J. M., Fierro, J .L.G., Laine, J ., and Severino, F. Appl.
Catal. 1992, 91, 43.

Bijsterbosch, J. W.; Kapteijn, F .; Moulijn, J. A. J. Mol. Catal. 1992, 74, 193.

Dekker, F. H. M.; Dekker, M. C .; Bliek, A.; Kapteijn, F.; Moulijn, J. A. Catal.
Today 1994, 20, 409.

Stegenga, S.; van Soest, R.; Kapteijn, F.; Moulijn, J. A. Appl. Catal. B 1993, 2,
257.

20.

21.

22.

23.

24.

25.

26.

27

28.

29.

30.

31.

32.

33.

34.

35.

36.

173

Dekker, N. J. J.; Hoom, J. A. A.; Stegenga, S.; Kapteijn, F.; Moulijn, J. A. AIChE
J. 1992, 38, 385.

Kapteijn, F.; Stegenga, S.; Dekker, N. J. J .; Bijsterbosch, J. W.; Moulijn, J. A.
Catalysis Today 1993, 16, 273.

Rastogi, R. P.; Singh, G.; Dubey, B. L.; Shukla, C. S. J. Catal. 1980, 65, 25.
Chien, C.-C.; Chuang, W.—P.; Huang, T.-J. Appl. Catal. A. 1995, I31, 73.

Tarasov, A. L.; Osmanov, M. 0.; Shvets, V. A.; Kazanskii, V. B. Kinet. Catal.
1990, 31, 565.

McCabe, R. W.; Mitchell, J. Ind. Eng. Chem. Prod. Res. Dev. 1983, 22, 212.
Rajesh, H.; Ozkan, U. S. Ind. Eng. Chem. Res. 1993, 32, 1622.

Heyes, C. J .; Irwin, J. G.; Johnson, H. A.; Moss, R. L. J. Chem. Tech. Biotechnol.
1982, 32, 1034.

Subbanna, P.; Greene, H.; Desal, F. Environ. Sci. T echnol. 1988, 22, 557.
Stegenga, S.; Dekker, N.; Bijsterbosch, J .; Kapteijn, F.; Moulijn, J .; Belot,
G.;Roche, R. In Catalysis and Automotive pollution Control I]; Crucq, A. Ed.;
Elsevier, Amsterdam, 1991; pp 353-369.

Stegenga, S. Ph. D. Thesis, University of Amsterdam, The Netherlands, 1991.

Fattakhova, Z. T.; Ukharskii, A. A.; Shiryaev, P. A.; Berrnan, A. D. Kinet. Katal.
1986, 27, 884.

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

Dadyburjor, D. B; Jewur, S. S.; Ruckenstein, A. E. Catal. Rev.- Sci. Eng. 1979,
19, 293.

Powder Diffraction File, Inorganic Phases, JCPDS, 1983.

Klug, H. P.; Alexander, L. E., X -Ray Diffraction Procedures for Polycrystalline
and Amorphous Materials, lst.Ed. Wiley, New York, 1954.

Rosencwaig, A. and Wertheim, G. K. J. Electron. Spectrosc. Relat. Phenom.
1972, 1, 493.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

174

Wallbank, B., Johnson, C. E, and Main, 1. G. J. Electron. Spectrosc. Relat.
Phenom. 1974, 4, 263.

Strohmeier, B. R.; Leyden, D. E; Field, R. S.; Hercules, D. M. J. Catal. 1985,
94, 514.

Okamoto, Y.; Fujii, M.; Imanaka, T.; Teranishi, S. Bull. Chem. Soc. Jpn. 1976,
49, 859. '

De Angelis, B. A. J. Electron Spectrosc. Relat. Phenom. 1976, 9, 81.

Scierka, S. J.; Proctor, A.; Houalla, M.; Fiedor, J. N.; Hercules, D. M. Surf
Interface Anal. 1993, 20, 901. ’

Deffosse, C.; Canesson, D.; Rouxhet, P. G.; Delmon, B. J. Catal. 1978, 51, 269.
Kerkhof, F. P. J. M.; Moulijn, J. A. J. Phys. Chem. 1979, 83, 1612.

Scofield, J. H. J. Electron. Spectrosc. Relat. Phenom. 1979, 8, 129.

Penn, D. R. J. Electron. Spectrosc. Relat. Phenom. 1976, 9, 29.

‘GOOGLY’ written by Dr. Andrew Proctor, University of Pittsburgh, PA.

Capece, F. M.; Di Castro, V.; Furlani, C.; Mattogno, G.; Fragale, C.; Gargano,
M.; Rossi, M. J. Electron Spec. Rel. Phenom. 1982, 27, 1 l9.

Friedman, R. M.; Freeman, J. 1.; Lytle, F. W. J. Catal. 1978, 55, 10.
Park, P. W.; Ledford, J. S. submitted to J. Phys. Chem.

Lo Jacono, M.; Schiavello, M. in “Preparation of Catalysis” Delmon, E; Jacobs,
P. A.; Poncelet, G. Eds. pp 474-487. Elsevier, Amsterdam, 1976.

Park, P. W.; Ledford, J. S. to be published.

Chapter 7

The Influence of Surface Structure on the Catalytic Activity of Cerium
Promoted Copper Oxide Catalysts on Alumina: Oxidation of Carbon
Monoxide and Methane

7.1. Abstract

X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) have been
used to characterize a series of Cu/Ce/Al203 catalysts. Catalysts were prepared by
incipient wetness impregnation using metal nitrates and alkoxides precursors. Catalyst
loadings were held constant at 12 wt.% CuO and 5.1 wt.% CeO2. Mixed oxide catalysts
were prepared by impregnation of cerium first, followed by copper. The information
obtained from surface and bulk characterization has been correlated with CO and CH4
oxidation activity of the catalysts. Cu/Al203 catalysts prepared using Cu(II) nitrate (CuN)
and Cu(II) ethoxide (CuA) precursors consist of a mixture of copper surface phase and
crystalline CuO. The CuA catalyst shows higher dispersion, less crystalline CuO phase,
and lower oxidation activity for CO and CH4 than the CuN catalyst. For Cu/Ce/Al203
catalysts, Ce has little effect on the dispersion and crystallinity of the copper species.
However, Cu impregnation decreases the Ce dispersion and increases the amount of

crystalline CeO2 present in the catalysts, particularly in catalysts prepared using the CeA

175

176

support. Cerium addition dramatically increases the C0 oxidation activity, however, it has

little effect on C114 oxidation.

7.2. Introduction

Copper oxide is a well known component of catalysts for CO [1-6], hydrocarbon
[1-3,7], chlorinated hydrocarbon [8] and alcohol oxidation [9,10] as well as for NO" [11-
13] and $02 reduction [13]. These air purification reactions using heterogeneous catalysts
have been important topics due to their application to automobile exhaust converters and
industrial waste incinerators. Copper oxide based catalysts have been considered as
suitable substitutes for noble metal catalysts in emission control applications due to their
high catalytic activity for the purification reactions, tolerant to 802 [3] and refractory to
high temperature [14]. Several researchers reported that copper oxide based catalysts
show similar activity to noble metal catalysts for CO oxidation [2], butanal and mercaptan
oxidation [15], and reduction ofNO by CO [13].

Rare earth oxide additives have been widely applied as textural and structural
promoters for supported noble metal catalysts. Cerium addition prevents y-alumina from
sintering at high temperature and improves the dispersion and thermal stability of noble
metals [16-20]. In addition, cerium oxide has been added to noble metal catalysts for
treatment of automobile exhaust to promote the water-gas shift reaction [21,22], to
suppress the CO inhibition effect [23-25], and to remove C0, hydrocarbons, and N0x

simultaneously using its oxygen storage ability [17,18,26-29].

177
Previous studies of rare earth promoters have focused primarily on their effects on

the structure and activity of noble metal-based catalysts. The study of the promoters effect
on supported transition metal oxide catalysts has been relatively limited. Recently cerium
oxide has been studied both as a promoter [30] and as a support [31] for oxidation
reactions over copper oxide based catalysts. Agarwal et a1. [30] found that a ceria
promoted hopcalite catalyst has a higher initial activity for hydrocarbon oxidation than a
Pt catalyst. Bedford and LaBarge [3 1] showed that copper-chromium oxide impregnated
on high surface area ceria has three-way catalyst behavior and retains catalyst activity after
aging at high temperature. Flytzani-Stephanopoulos and coworkers [32-34] reported
recently that a Cu-Ce-O composite catalyst shows high activity for CO and CI-Ia oxidation
and high resistance to water poisoning. They attributed the enhanced catalytic activity and
stability to the strong interaction of copper and cerium oxides. The effect of a CeO2
promoter on alumina supported copper oxide catalysts has been reported similar to the
effect it has on noble metal catalysts [35,36]. The authors reported that the addition of
Ce02 enhances redox behavior of copper ion, increases dispersion of copper oxides,
provides surface oxygen, and improves oxygen storage capacity of catalyst.

While, cerium oxide promoted copper oxide based catalysts have recently drawn
an attention of many researchers, the effect of cerium oxide on the surface structure and
reactivity of copper catalysts has not been studied extensively. The present work is part of
a broad study to investigate structure-reactivity correlation for rare earth oxide promoted
transition metal oxide based emission control catalysts. In this paper, X-ray photoelectron

spectroscopy (XPS) and X-ray diffraction (XRD) have been used to determine the effect

178

of catalyst precursors and cerium promotion on the chemical state and dispersion of
copper oxide phases in a series of Cu/Ce/Al203 catalysts. The information derived from
these techniques is correlated with CO and CH4 oxidation activity to develop a more

complete understanding of Cu/Ce/A12O3 catalysts.

7.3. Experimental

Catalyst Preparation. Catalysts were prepared by pore volume impregnation of y-
alumina (Cyanamid, surface area = 203 mz/g, pore volume = 0.6 mL/g). The alumina was
finely ground (< 230 mesh) and calcined in air at 500 °C for 24 h prior to impregnation.
The Ce/Al2O3 catalysts were prepared using a deionized water solution of ammonium
cerium(IV) nitrate (Mallinckrodt Inc, Analytical Reagent) or an ethanol solution of
cerium(IV) methoxyethoxide (Gelest Inc., l8-20% cerium methoxyethoxide in
methoxyethanol). The impregnated sample derived from the aqueous cerium nitrate
solution (designated “CeN”) was dried in air at 120 °C for 24 h and calcined in air at 500
°C for 16 h. The sample prepared using the ethanol cerium methoxyethoxide solution
(designated “CeA”) was impregnated and subsequently dried at room temperature for 48 h
in a N2 purged glove bag prior to further drying in air at 120 °C for 24 h and calcination in
air at 500 °C for 16 h. The Ce loading was held constant at 5.1 wt.% CeO2. The
Cu/Al203 catalysts were prepared with a deionized water solution of copper(II) nitrate
(Columbus Chemical Industries Inc, ACS Grade) or an aqueous diethylenetriamine

solution (DETA/water ratio = 1/3 by volume) of copper(II) ethoxide. DETA was

179
necessary due to the limited solubility of c0pper(II) ethoxide in water and common

organic solvents. Since an aqueous DETA solution was used to dissolve copper(II)
ethoxide, a Cu-DETA complex should be formed in the impregnating solution [37]. The
impregnated samples derived from the aqueous copper nitrate solution (designated
“CuN”) and catalysts prepared with the aqueous DETA copper ethoxide solution
(designated “CuA”) were dried and calcined under the same conditions as used for CeN
and CeA samples respectively. The Cu loading was held constant at 12 wt.% CuO.
Mixed metal oxide catalysts were prepared by step-wise impregnation of Ce first, followed
by Cu. The cerium modified alumina carrier was dried and calcined prior to introduction
of copper. After the addition of copper, the catalysts were dried and calcined again at the
same conditions described above depending on the choice of precursors. Catalysts
prepared using nitrate and alkoxide precursors will be symbolized using "N" and "A",
respectively.

Standard Materials. CuO and CeO2 were prepared by calcining copper(II) nitrate
and ammonium cerium(IV) nitrate in air at 500 °C for 16 h. CuAl2Oa was prepared by
calcining stoichiometric amounts of the respective nitrates in air at 1000 °C for 24 h.
CeAlO3 was prepared by reduction of a dried CeA sample at 800 °C in H2 (AGA, 99.99%)
for 8 h. XRD patterns of the standard compounds matched the appropriate Powder
Diffraction File [3 8].

BET Surface Area. Surface area measurements were performed using a
QuantaChrome Quantasorb Jr. Sorption System. Approximately 0.1 g of catalyst were

outgassed in a N2/He mixture (5% N2) at 350 0C for l h prior to adsorption

 

180

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 addition of copper
and/or cerium to the alumina carrier decreased the BET surface area by a maximum of 20
%.

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

Rigaku XRD diffractometer employing Cu Ka radiation (2» = 1.5418 A). The X-ray was
operated at 45 kV and 100 mA. Diffraction patterns were obtained using a scanning rate
of 0.5 deg/min (in 20) with divergence slit and scatter slit widths of 1°. Samples were run

as powders packed into a glass sample holder having a 20 x 16 x 0.5-mm cavity.

The mean crystallite sizes (d) of the CuO and CeO2 particles were determined

from XRD line broadening measurements using the Scherrer equation [39]:

d: KK/BcosG (1)

where it is the X-ray wavelength, K is the particle shape factor, taken as 0.9, and B is the

fill] width at half maximum (fwhm), in radians, ofthe CuO <111> or CeO2<l 1 l> line.

Quantitative X-ray diffraction data were obtained by comparing

Cu0<111>/Al203<400> or Ce02 <1 1 l>/A1203<440> intensity ratios measured for
catalyst samples with intensity ratios measured for physical mixtures of pure CuO or CeO2

and y-Al203. This method assumed that copper or cerium addition did not disrupt the 7-

181
A1203 spinel and subsequently affect the intensity of the A1203 lines. The error in this

method was estimated to be 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 300 W (15 kV, 20
mA) and a 10-360 hemispherical analyzer Operated with a pass energy of 50 eV. The
instrument typically operates at pressures near 1 x 10‘8 torr in the analysis chamber.
Spectra were collected using a PC137 board interfaced to a Zeos 386SX computer.
Samples were analyzed as powders dusted onto double-sided sticky tape or spray coated
onto a quartz slide using a methanol suspension of the catalyst. Binding energies for the
catalyst samples and standard compounds which contained Al were referenced to the Al
2p peak (74.5 eV). The binding energies for standard compounds that did not contain Al
were referenced to the C ls line (284.6 eV) of the carbon overlayer. XPS binding
energies were measured with a precision of i 0.2 eV, or better.

Reduction of copper [40-42] and cerium [43-46] species during XPS experiments
has been reported and attributed to several factors such as X-ray flux, X-ray dose,
temperature, and pressure. In order to minimize the effect of photoreduction on the
results, all‘samples were analyzed using the same distance between X-ray source and
sample (~3/4 inch) and minimum data acquisition time (5 min. scan for Cu 2p3r2 spectra).
No significant photoreduction of copper species was observed using these experimental
conditions. However, Ce 3d region had to be scanned for extended time (~ 600 min.) in
order to obtain reasonable quality spectra. Therefore Ce XPS data were used only for

quantitative analysis to measure Ce dispersion on alumina.

182
Quantitative XPS Analysis. It has been shown by Defosse et al. [47] that one may

calculate the theoretical intensity ratio 1°./1°, expected for a supported phase (p)
atomically dispersed on a carrier (5). An extension of the Defosse model proposed by
Kerkhof and Moulijn [48] has been used in the present investigation. The photoelectron
cross sections and the mean escape depths of the photoelectrons used in these calculation
were taken from Scofield [49] and Penn [50], respectively. For a phase (p) present as

discrete particles, the experimental intensity ratio I,/I. is given by the following expression:

141. = 1°./1°. rI-exp (-d/7~p)] / do. (2)

where 1°P/I°, is the theoretical monolayer intensity ratio, d is the length of the edge of the
cubic crystallites of the deposited phase and it? is the mean escape depth of the
photoelectrons in the deposited phase.

CO Oxidation Activity. Measurement of CO oxidation activity was performed in a
flow microreactor. Approximately 0.03 g of catalyst were supported on a glass hit (70 -
100 um) and the temperature was measured with a K-type thermocouple located just
above the catalyst bed. The reactor was heated by a tube fumace (Lindberg) with
temperature being controlled within 1°C by an Omega CN 1200 temperature controller.
Reactant gas flow rates were held constant with Brooks 5850 mass flow controllers.
Product gases were analyzed with a Varian 920 gas chromatograph equipped with a TCD
and interfaced to a Hewlett-Packard 3394A integrator. Reaction products were separated

on a 6 ft 60/80 mesh Carbosieve S-II column. Prior to the first activity measurement, the

183
catalyst was pretreated with a mixture of 5% O2/He (99.5 % purity for 02, 99.995%

purity for He, AGA Gas Co.) stream (143 cc/min.) at 350 °C for 1 hr to remove any
impurities absorbed on the surface during catalyst preparation and storage. CO oxidation
reactions were performed with a constant flow rate (80 cm3/min) of 4.8% CO/9.8%
02/85.4% He gas mixture (AGA, purity > 99.99%) in the temperature range of 100-220
°C. A11 activity measurements were obtained under steady-state conditions at conversions
less than 15%.

CH; Oxidation Activity. Methane oxidation reactions were performed with a
constant flow (15 cm3/min) of 0.98% CHa/5.25% 02/93.77% He gas mixture (AGA,
purity > 99.99%) in the temperature range of 360-420 °C. Approximately 0.1 g catalyst
were charged into the same type of microreactor used for C0 oxidation measurements.
Water produced during methane oxidation was frozen downstream from the reactor in a
trap maintained < -40 °C with a mixture of ethanol and dry ice. No partial oxidation
product was observed. All activity measurements were obtained under steady-state

conditions at conversions less than 15%.

7.4. Results and Discussions

Chemical State and Dispersion of Cerium/A lumina Catalysts. The cerium oxide
chemical state and dispersion of CeN and CeA catalysts have been reported previously
[46]. In summary, the Ce dispersion for the CeN and CeA was found to be poor and

Similar each other (Table 7.2). For the CeA, only 0.3 wt.% cerium oxide was detectable

184
by XRD (Figure 7.1b), indicating that most of the cerium oxide is present as an

amorphous phase. A large amount of crystalline Ce02 was observed for CeN (Figure
7.1c), although a low loading ceria and high surface area alumina were used in this study.
This result does not agree with the results of other researchers [20,51]. For cerium
catalysts with < 2.6 umol ceria/m2 alumina, which is corresponding to < 9 wt.% Ce0j200
m2 A1203, Ce species were present as CeA103 dispersed surface phase and/or as small
CeO2 crystallites which were not detectable by XRD. This discrepancy may be readily
understood when one considers that the cerium precursor used in this study (Ce‘f) was
different from the one used in other studies (Ce3i). The limited interaction between the
Ce4+ precursor and alumina support may produce less Ce3+ ion occupancy in cation
vacancies in alumina lattice due to charge conflict. This leads to the formation of CeO2
during sample preparation.

Chemical State of Copper/Alumina Catalysts. The XRD patterns obtained for
CuA and CuN catalysts are shown in Figure 7.1d and 7.1e. The XRD pattern of the CuN
catalyst shows intense peaks characteristic of CuO, however, weak CuO lines are
observed in the diffraction pattern of the CuA catalyst. Quantitative XRD measurements
indicate that the CuN catalyst contains 5.5 wt.% crystalline CuO phase, while the CuA
catalyst contains only 0.8 wt.% crystalline CuO (Table 7.1). For alumina supports with
surface area comparable to that used in this study, previous results [40,52] suggest that up
to 8 wt.% Cu can be incorporated into the alumina lattice as a Cu-Al203 surface phase

when copper nitrate precursor is used to prepare the catalyst. Thus, our finding that the

185

 

 

CuO
<1 1 1) CuO

<lll>

 

CuO

<no> hv*hd

 

 

(d)

(C)

#22. :fi‘: ‘2‘ ; W)

 

 

Figure 7.1.

A120, <311> 15.1203
<220> <222>

l l l l
25 30 35 40

Degrees (20)

CuA, and (e) CuN catalysts.

 

XRD patterns measured for (a) alumina support, (b) CeA, (c) CeN, (d)

186

Table 7.1. Concentration of Crystalline‘ Phases in Cu/Ce/A12O3
Catalysts Calculated from Quantitative XRD Data.

 

Crystalline Phase (wt.%)

 

Catalysts Ce02 CuO
CeN 4.6 b
CeA 0.3 b
CuN b 5.5
CuA b 0.8

CuNCeN 4.7 6.0

CuACeN 4.1 0.6

CuNCeA 3 .4 5.5

CuACeA 1.5 0.8

 

 

 

a - Valid for crystalline phases with particle sizes > 3.0 nm.
b - Catalysts do not contain the component.

presence of 6.5 wt.% of CuO as a dispersed phase for CuN catalyst is in good agreement
with previous work. In addition, the XPS Cu 2p3/2 binding energies measured for CuN
and CuA catalysts (935.1 eV) are consistent with the value measured for CuAl204 (935.0
eV) and higher than the value measured for CuO (933.9 eV). Thus, the XRD and XPS
results indicate that copper is present predominantly as a dispersed copper surface phase
on the alumina support.

Dispersion of Copper/Alumina Catalysts. Particle sizes determined from XRD
line broadening calculations and from Cu/Al XPS intensity ratios for CuN and CuA
catalysts are shown in Table 7.2. CuO particle sizes determined using XRD line
broadening calculations are identical for CuN and CuA catalysts. While, the copper

species particle size determined for the CuA catalyst using XPS is smaller than the value

187

Table 7.2. Particle Size of Cerium and Copper Phases
Determined from XRD Line Broadening
Calculations and XPS Intensity Ratios.

 

Particle Size (nm)

 

Catalysts Cerium Copper
XRD XPS XRD XPS
CeN 5.1 1.6 -' -‘
CeA 4.5 1.6 -' -‘
CuN -“ -‘ 30 1.9
CuA -“ -" 30 1 . 1
CuNCeN 5.1 3.0 32 2.1
CuACeN 5.4 3.6 32 1.1
CuNCeA 4.1 4.1 32 2.2
CuACeA 4.6 4.7 30 1.1

 

 

 

a - Catalysts do not contain the component.

obtained for the CuN catalyst. XPS and XRD analyses of the CuA catalyst indicate the
formation of less crystalline and more dispersed copper species on the alumina support
compared to CuN catalyst. From quantitative XRD data, approximately 11.2 wt.% CuO
in CuA which is not detected by XRD can be assigned as a dispersed copper species on
the alumina surface. However, the relatively poor Cu dispersion compared to monolayer
noted for CuA catalyst indicates that it is not entirely due to a Cu-Al203 surface phase,
since copper surface phase is normally considered to be highly dispersed. It can be
assumed that a poorly dispersed amorphous copper oxide phase is partially formed on
CuA catalyst, which is not detectable by XRD. Uchikawa and Mackenzie [37] have

proposed that a polymer structure of gel containing Cu-DETA unit was obtained afier a

188
solution of copper ethoxide-aqueous DETA was heated at 90 °C. Therefore, the polymer

structure is possibly obtained during the drying of aqueous copper ethoxide-DETA
solution and then an amorphous CuO phase may be formed subsequently in calcination
step. It is also possible that Cu-DETA complexes aggregate less during catalyst
preparation than the copper nitrate precursor.

The CuO particle sizes of all copper catalysts determined with XRD are
significantly larger than the values obtained from XPS (Table 7.2). This difference may be
attributed to the limitations of X-ray diffraction. It is well known that XRD determined
particle sizes are skewed to large values since XRD does not detect highly dispersed
species. For the copper catalysts, a significant fraction of the Cu is present as a copper
surface phase or small CuO particles (d < 3.0 nm) that are detected by XPS but not by
XRD. Thus, smaller average particle sizes are expected from XPS calculations.

Chemical State of Copper/Cerium’Alumina Catalysts. Figure 7.2 shows the XRD
patterns obtained for Cu/Ce/Al203 catalysts. The XRD patterns show only peaks
characteristic of CuO and CeO2. This is in agreement with studies have found that copper
oxide is immiscible with cerium oxide [33]. For the Cu/Ce/Al203 catalysts prepared using
the CeN (i.e. CuNCeN and CuACeN), Cu addition has little effect on the intensity of
Ce02 XRD peaks. This can be attributed to that a large amount of poorly dispersed
crystalline CeO2 (4.1 - 4.7 wt.%) is already present in CeN. For the Cu/Ce/A12O3 catalysts
derived from the CeA (i.e. CuNCeA or CuACeA), Cu addition increases the intensity of

CeO2 XRD peaks. The Cu addition effect on the Ce02 peak intensity is observed more

Figure

189

 

 

 

 

 

 

 

 

 

 

CuO
<111> CuO
<11 l>
CeO2
<111>
O
<110> ‘
(d)
CeO2
Q00>
(C)
l
(b)
(a)
1 1 r 1
25 30 3 5 40

Degrees (20)

Figure 7.2. XRD patterns measured for (a) CuACeA, (b) CuNCeA, (c) CuACeN, and
(d) CuNCeN catalysts.

190
strongly in the diffraction pattern of CuNCeA catalyst than that of CuACeA catalyst.

Quantitative XRD measurements (Table 7.1) show that Cu addition increases the amount
of crystalline Ce02 from 0.3 wt.% (CeA) to 1.5 wt.% (CuACeA) and 3.4 wt.%
(CuNCeA). Since cerium oxide in CeA initially was a amorphous phase [46], post loaded
copper species may act as a seed to form a crystalline Ce02 during sample preparation.
We attribute the more pronounced increase in crystalline Ce02 content observed for the
CuNCeA catalyst to the dissolution of cerium oxide in an acidic copper nitrate solution
(pH 2.2) followed by agglomeration during subsequent drying and calcination steps of the
catalyst preparation. Cerium oxide is insoluble in basic impregnating solution (pH 12.4)
used for CuA derived catalysts.

For the Cu/Ce/Al203 catalysts, quantitative XRD measurements show that the
CuO crystallinity is identical within experimental error to the value measured for
analogous CuN and CuA catalysts. In addition, Ce has little effect on the XPS Cu 2p3/2
binding energies (935.0 i 0.1 eV) which were identical to the value measured for CuAl2Oa
and higher than the value of CuO. This indicates that chemical state of copper species is
independent of Ce promoter in this study.

Dispersion of Copper/Cerium/A lumina Catalysts. The particle sizes of Cu and Ce
species calculated using XRD line broadening and XPS intensity ratios are shown in Table
7.2. XRD line broadening calculations show that CeO2 particle sizes of Cu/Ce/A12O3
catalysts are not changed by Cu addition and are similar to those of analogous CeN and
CeA supports. However, the cerium particle size calculated from Ce/Al XPS intensity

ratios increases significantly after Cu addition. This indicates that the addition of copper

191
Species dramatically decreases the dispersion of cerium species. Since the addition of Cu

increases the crystallinity of cerium oxide, we can attribute the decrease of dispersion to
the agglomeration of cerium species during catalyst preparation step. In addition, we can
not exclude an encapsulation phenomena involving the deposition of copper species on the
surface of the cerium oxide phase, which would lead to a decrease in the Ce/Al XPS
intensity ratio and dispersion subsequently. Such encapsulation effect would be more
prominent for CuA due to its higher dispersion than CuN.

The particle sizes of copper oxide calculated from XRD line broadening
calculations are identical regardless of cerium promotion and copper precursors. The
particle sizes of copper species calculated from Cu/Al XPS intensity ratio depend only on
the precursors used for catalyst preparation. The absence of a significant effect of Ce
addition on the CuO crystallinity and dispersion can be understood in terms of the
Ce/A12O3 surface structure. Since the CeN and CeA contain a low concentration of
poorly dispersed cerium oxide, most of the y-Al203 surface is exposed to the Cu
introduction. This exposed alumina dominates the alteration of the copper structure in the
catalysts.

Effect of Catalyst Structure on CO Oxidation Activity. Table 7.3 shows the
specific turnover numbers (TON) and activation energies for CO oxidation over Cu/Al203
and Cu/Ce/Al203 catalysts calculated at 160 °C using XPS estimates of the Cu dispersion.
The results of CeN and CeA catalysts were not shown here due to their negligible CO
oxidation activity compared to the copper catalysts. The CuN catalyst shows higher TON

and lower activation energy than the values of the CuA catalyst. Comparison of the CO

192

Table 7.3. Turn Over Numbers‘ and Activation Energies” of
Cu/Ce/A1203 Catalysts for CO and CH4 Oxidation.

 

 

CO Oxidation CHa Oxidation

Catalysts TON Activation Energy TON Activation EnergL

CuN 1.1 12.0 3.7 22.7

CuA 0.3 15.4 1.2 21.6
CuNCeN 8.7 10.9 3.7 22.7
CuACeN 3.0 12.1 1.3 23.8
CuNCeA 13.7 12.6 4.5 21.9
CuACeA 1.5 12.8 1.4 22.6

 

 

 

a - Calculated at 160 °C and 390 °C for CO and CH4 oxidation respectively
using exposed Cu species determined from XPS (s'1 x 102).
b - Estmiated from Arrhenius Plots (kcal/mol) within 15 % conversion.

oxidation activities measured for the CuN and CuA catalysts indicates that the catalyst
containing more crystalline CuO (i.e. CuN) is more active than catalysts containing less
CuO phase (CuA). This is consistent with our previous study [53], which showed that
crystalline CuO was more active for C0 oxidation than a dispersed copper surface phase.
The Ce promoted catalysts show increased TONs by a factor of 5 to 12 compared
to unpromoted copper catalysts. The activation energies for CO oxidation calculated for
cerium promoted catalysts are similar within a experimental error (12.1 i 0.9 kcanol).
The higher CO oxidation TON for cerium promoted copper catalysts indicates that cerium
oxide is very effective promoter when one considers that only a small amount of cerium
oxide (5.1 wt.%) was used in this study. Changing crystallinity and dispersion of cerium
oxide species after Cu addition indicates that copper oxide interacts strongly with cerium

oxide during catalyst preparation. Therefore, cerium species included in the CuO particles

193
may lead to increased CO oxidation activity. Liu and Flytzani-Stephanopoulos [33,34]

reported that Cu-Ce-O composite catalyst shows high CO oxidation activity, comparable
to that of the Pt/alumina catalyst. They attributed the enhanced catalytic activity to the
strong interaction of copper and cerium oxides. It also has been reported recently [3 5,36]
that ceria promoter enhances the redox behavior of copper ions and produces surface
oxygen ions over CuO/CeO2/y-alumina catalysts. The Cu/Ce/Al203 catalysts derived from
CuN show higher CO oxidation activity than the catalyst prepared with CuA. This is
consistent with the result of CuN and CuA catalysts, which shows higher CO oxidation
activity for the catalyst containing more crystalline CuO.

Effect of Catalyst Structure on CH4 Oxidation Activity. Table 7.3 shows the
specific turnover numbers (TON) and activation energies for CH4 oxidation calculated at
390 °C. The results of CeN and CeA catalysts were not shown here due to their negligible
CH4 oxidation activity compared to the copper catalysts. The TONs of the catalysts
derived from CuN is about 3 times higher than the catalysts prepared with CuA. The Ce
addition has little effect on the CH4 oxidation TON for copper catalysts. The activation
energies are identical within experimental error (22.6 i- 0.8 kcal/mol).

Since the active site of copper oxide catalyst for CH4 oxidation reaction has been
assigned to the isolated copper surface phase [53], the catalyst containing more isolated
copper surface phase should be more effective. In our study, CuA catalyst may contain
less isolated copper surface phase than CuN catalyst, although Cu dispersion of CuA is
higher than that of CuN. The formation of less isolated copper surface phase can be

attributed to the formation of Cu-DETA polymer structure during catalyst preparation

194
step [37]. We propose that this polymer type Cu-DETA complex leads to the formation

of an interacting copper surface phase (i.e. clustered c0pper species) and/or amorphous
copper oxide which may be a less active phase than isolated c0pper species for CH4
oxidation. The no cerium promoter effect on CH4 oxidation reaction for Cu/Ce/Al203
catalysts can be explained by little interaction between poorly dispersed cerium oxide and

isolated copper surface phase.

7.5. Conclusions

The combined use of several techniques to investigate the effect of cerium
promoter and metal precursors on the structure of Cu/Ce/A12O3 catalysts leads to the

following conclusions.

1. The CuA catalyst shows enhanced Cu dispersion compared to CuN catalyst.
However, CuA catalyst is less active for CO and CH4 oxidation than CuN catalyst. This
has been attributed to that CuA catalyst contain less active phases than the CuN catalyst
for CO and CH4 oxidation which are crystalline CuO and isolated copper surface species,
respectively.

2. Ce has little effect on the dispersion and crystallinity of copper species. Whereas
Cu addition decreases the Ce dispersion and increases the amount of crystalline Ce02

present in the Cu/Ce/Al203 catalysts particularly in catalysts prepared using the CeA.

195
3. Ce addition increases the CO oxidation activity of the copper catalysts. This has

been attributed to that cerium oxide interacts strongly with CuO crystalline which is active
phase for CO oxidation. While, the Ce may not promote the isolated copper surface

species which is active site for CH.; oxidation.

10.

11.

12.

13.

14.

15.

16.

17.

18.

196
7.6. References

Yao, Y. F. Yu; Kummer, J. T. J. Catal. 1977, 46, 388.

Kummer J. T., Prog. Energy Combust. Sci. 1980, 6, 177.

Yao, Y. F. YuJ. Catal. 1975, 39, 104.

Severino, F.; Brito, J.; Cari’as, O. Laine, J. J. Catal. 1986, [02, 172.
Boon, A. Q. M.; van Looij, F.; Geus, J. W. J. Mol. Catal. 1992, 75, 277.

Lopez Agudo, A.; Palacios, J. M.; Fierro, J .L.G.; Laine, J.; Severino, F. Appl.
Catal. 1992, 91, 43.

Boon, A. Q. M.; Huisman, H. M.; Geus, J. W. J. Mol. Catal. 1992, 75, 293.
Subbanna, P.; Greene, H.; Desal, F. Environ. Sci. T echnol. 1988, 22, 557.

Ozkan, U. S.; Kueller, R. F.; Moctezuma, E. Ind. Eng. Chem. Res. 1990, 29,
1136.

Rajesh, H.; Ozkan, U. S. Ind. Eng. Chem. Res. 1993, 32, 1622.

Kapteijn, F.; Stegenga, S.; Dekker, N. J. J.; Bijsterbosch, J. W.; Moulijn, J. A.
Catal. Today 1993, 16, 273.

Huang, T.-J.; Yu, T.-C. Appl. Catal. 1991, 71, 275.

Goetz, V. N.; Sood, A.; Kittrell, J. R. Ind. Eng. Chem. Prod. Res. Develop. 1974,
I3, 110.

R. Prasad, L. A. Kennedy and E. Ruckenstein, Catal. Rev-Sci. Eng, 1984, 26, l.

Heyes, C. J.; Irwin, J. G.; Johnson, H. A.; Moss, R. L. J. Chem. Tech. Biotechnol.
1982, 32,1025.

Yao, Y. F. Yu; Kummer, J. T. J. Catal. 1987, 106, 307.
Su, E. C.; Montreuil, C. N.; Rothschild, W. G. Appl. Catal. 1985, 17, 75.

Gandhi, H. S.; Piken, A. G.; Shelef, M.; Delosh, R. G. SAE paper No. 760201,
1976.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

197

Dictor, R.; Roberts, S. J. Phys. Chem. 1989, 93, 5846.

Le Normand, F .; Hilaire, L.; Kili, K.; Krill, G.; Maire, G. J. Phys. Chem. 1988, 92,
2561.

Kim, G. Ind. Eng. Chem. Prod. Res. Dev. 1982, 21, 267.

Herz, R. K.; Sell, J. A. J. Catal. 1985, 94, 166.

Oh, S. H. J. Catal. 1990, 124, 477.

Yao, Y. F. YuJ. Catal. 1984, 87, 152.

Oh, S. H.; Eickel, C. C. J. Catal. 1988, 112, 543.

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

Herz, R. K. Am. Chem. Soc. Symp. Ser. 1982, 178, 59.

Jin, T.; Okuhara, T.; Mains, G. J.; White, J. M. J. Phys. Chem. 1987, 91, 3310.
Cho, B. K.; Shanks, B. H.; Bailey, J. E. .I. Catal. I989, 115, 486.

Agarwal, S. K.; Spivey, J. J.; Butt, J. B. Appl. Catal. 1992, 81, 239.

Bedford, R. E; LaBarge, W. J. US. Patent 5,063,193 Nov. 5, 1991.

Liu, W.; Sarofim, A. F; FIytzani-Stephanopoulos Appl. Catal. B. 1994, 4, 167.
Liu, W.; Flytzani-Stephanopoulos, M. J. Catal. 1995, 153, 304.

Liu, W.; Flytzani-Stephanopoulos, M. J. C atal. 1995, 153, 317.

Lu, G.; Wang, R. Cuihua Xuebao 1991, I2, 314.

Tian, Y.; Fu, Y.; Lin, P. Cuihua Xuebao 1994, I5, 189.

Uchikawa, F .; Mackenzie, J. D. J. Mater. Res. 1989, 4, 787.

Powder Diffraction File, Inorganic Phases, JCPDS, 1983.

Klug, H. P.; Alexander, L. E, X-Ray Diffraction Procedures for Polycrystalline
and Amorphous Materials, lst.Ed. Wiley, New York, 1954.

40.

.41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

198

Strohmeier, B. R.; Leyden, D. E; Field, R. S.; Hercules, D. M. J. Catal. 1985,
94, 514.

Wallbank, B., Johnson, C. E, and Main, 1. G. J. Electron. Spectrosc. Relat.
Phenom. 1974, 4, 263.

Rosencr-s' g, A. and Wertheim, G. K. J. Electron. Spectrosc. Relat. Phenom.
1972, 1. -, . 3.

Al-than, Z. T.; Hashemi, T.; Hogarth, C. A. Spectrochimica Acta. 1989, 44b,
205.

Paparazzo, E. Surf Sci. 1990, 234, L253.
Paparazzo, E; Ingo, G. M.; Zacchetti, N. J. Vac. Sci. T echrrol. A. 1991, 9, 1416.
Park P.W.; Ledford J. S. accepted to be published in Langmuir

Deffosse, C.; Canesson, D.; Rouxhet, P. G.; Delmon, B. J. Catal. 1978, 51, 269.
Kerkhof, F. P. J. M.; Moulijn, J. A. J. Phys. Chem. 1979, 83, 1612.

Scofield, J. H. J. Electron. Spectrosc. Relat. Phenom. 1979, 8, 129.

Penn, D. R. J. Electron. Spectrosc. Relat. Phenom. 1976, 9, 29.

Shyu, J. 2.; Weber, W. H.; Gandhi, H. S. J. Phys. Chem. 1988, 92, 4964.
Friedman, R. M.; Freeman, J. J.; Lytle, F. W. J. Catal. 1978, 55, 10.

Park, P.W.; Ledford. J. S. to be published

      
     

HIGH ‘1

c N STATE UNIV. LIBRARIES
llllllllllllllllllllllIlllllllllllllllllllllllllllllllll
1293014217313

  

I

3