evaluación de propiedades superficiales y catalíticas del ... · pdf...

Evaluación de propiedades superficiales y catalíticas del

proceso de adsorción de etileno y oxígeno sobre catalizadores de oro soportados en óxidos mixtos CeZr

Lucía Pérez Ramírez

Universidad Nacional de Colombia

Facultad de Ingeniería, Departamento de Ingeniería Química y Ambiental

Bogotá, Colombia

2017

Evaluación de propiedades superficiales y catalíticas del

proceso de adsorción de etileno y oxígeno sobre catalizadores de oro soportados en óxidos mixtos CeZr


Tesis de grado presentada como requisito parcial para optar al título de:

Magister en Ingeniería – Ingeniería Química

Director:

Julio César Vargas Sáenz

Línea de Investigación:

Catálisis Heterogénea

Grupo de Investigación:

Procesos Químicos y Bioquímicos


Facultad de Ingeniería, Departamento de Ingeniería Química y Ambiental

Bogotá, Colombia

2017

Evaluation of the surface and catalytic properties of the ethylene and oxygen adsorption process on gold catalysts supported on mixed

oxides CeZr


A dissertation submitted in partial fulfilment of the requirements for the degree of:

Master of Engineering – Chemical Engineering

Advisor:

Julio César Vargas Sáenz

Research field:

Heterogeneous Catalysis

Research group:

Chemical and Biochemical Processes


School of Engineering, Department of Chemical and Environmental Engineering

Bogotá, Colombia

2017

He aquí que el temor del Señor es la sabiduría,

Y el apartarse del mal, la inteligencia.

Job 28:28

Acknowledgements

First, I praise God for its faithfulness. He filled me with strength and courage during this

challenging journey.

Many thanks to my advisor, Professor Julio César Vargas, for his direction, patience and

persistent help in all circumstances, and to Professor Sébastien Thomas, who provided me

with the opportunity to conduct part of my research at the Institut de Chimie et Procédés

pour l’Energie, l’Environnement et la Santé (ICPEES) facilities at Université de Strasbourg

under a CNRS collaboration partnership. I also acknowledge the economic support

provided by the ICPEES-CNRS during this research stay, as well as the funding received

by the Department of Chemical Engineering and the “División de Investigación y Extensión

Sede Bogotá” through the “Call for International Mobility 2016-2018”.

Thanks to Professor Anne-Cécile Roger for having me at her team at the ICPEES. To

Thierry Dintzer, Thierry Romero, Sécou, Christophe Mélart and Corinne Ulhaq for their

precious help on the characterisation of the catalysts.

I am especially grateful to have been received in the Laboratoire Catalyse et Spectrochimie

(LCS), Caen, and to have learned so much of infrared from Philippe Bazin.

The funding received during my master programme from Universidad Nacional de

Colombia by the scholarship “Asistente Docente”, is gratefully acknowledged, as well as

the funding received by Colciencias through the call for “Jovenes Investigadores”.

Thanks to my friends at Universidad Nacional de Colombia, Université de Strasbourg and

LCS Caen, specially Valentin, Kasia and Ibrahim, for making the path even more enjoyable.

Last but not least, I thank my family for being my unconditional support and motivation.

Resumen y Abstract XI

Abstract

Surface and catalytic properties of mixed oxides CeO2, Au/CeO2, CeZr and Au/CeZr, during

the total oxidation of ethylene is presented in this work. Four types of supports were used:

synthesised and commercial CeO2 and CeZr mixed oxides. Synthesised supports were

prepared by pseudo sol gel method while commercial supports were provided by Solvay-

Rhodia. The mass CeO2/ZrO2 ratio was 0.8/0.2 for the synthesised support and 0.6/0.4 for

the commercial one. Gold was subsequently deposited via direct anion exchange. XRD

and Raman spectroscopy showed the effective formation of the fluorite type structure on

each catalyst. Textural properties were determined via N2 adsorption, demonstrating the

low influence of gold deposited as active phase. Temperature programmed desorption

(TPD) and IR operando spectroscopy were used to stablish the fluid-surface interaction,

elucidate the reaction mechanism during the oxidation reaction and determine the role of

gold as active phase. Results show that there is no adsorption of ethylene or oxygen on

any of the catalysts at room temperature, despite increasing the amount of gold deposited

from 1 to 2%. In fact, only catalysts with gold are active for the oxidation reaction and only

at temperatures above 100 °C. The formation of formate and carbonate species as

intermediary precursors for the subsequent CO2 desorption, are seen as responsible for

the oxidation of ethylene following the Mars and Van Krevelen model.

Keywords: mixed oxides, IR operando spectroscopy, ceria-based catalysts, total

oxidation, ethylene

XII Evaluation of the surface and catalytic properties of the ethylene and oxygen

adsorption process on gold catalysts supported on mixed oxides CeZr

Resumen

Las propiedades superficiales y catalíticas de los óxidos mixtos CeO2, Au/CeO2, CeZr y

Au/CeZr, durante la oxidación total del etileno se presentan en este trabajo. Se utilizaron

cuatro tipos de soportes: óxidos CeO2 y CeZr sintetizados y comerciales. Los soportes

sintetizados se prepararon mediante el método de pseudo sol gel, mientras que los

soportes comerciales fueron proporcionados por Solvay-Rhodia. La proporción másica de

CeO2/ ZrO2 fue de 0.8/0.2 para el soporte sintetizado y de 0.6/0.4 para el soporte comercial.

El oro como fase active se depositó posteriormente a través del intercambio aniónico

directo. La espectroscopia XRD y Raman mostró la formación efectiva de la estructura del

tipo fluorita en cada catalizador mediante difracción de rayos X (XRD) y espectroscopía

Raman. Las propiedades texturales se determinaron por adsorción de N2, demostrando la

baja influencia del oro depositado como fase active sobre estas. Se utilizaron las técnicas

de desorción a temperatura programada (TPD) y espectroscopia infrarrojo (IR) operando

para establecer la interacción fluido-superficie, elucidar el mecanismo de reacción durante

la oxidación total de etileno y determinar el papel de la fase activa. Los resultados muestran

que no hay adsorción de etileno ni de oxígeno en ninguno de los catalizadores a

temperatura ambiente, a pesar de aumentar la carga de oro de 1 a 2%. De hecho, sólo los

catalizadores con oro son activos para la reacción de oxidación y sólo a temperaturas por

encima de 100 ºC. Se observa la formación de especies de formiato y carbonato como

precursores intermedios para la posterior formación y desorción de CO2, y responsables

de la oxidación del etileno siguiendo el modelo propuesto por Mars y Van Krevelen.

Palabras clave: óxidos mixtos, espectroscopía IR operando, catalizadores a base de

ceria, oxidación total, etileno

Contents XIII

Contents

Pág.

Abstract.......................................................................................................................... XI

List of figures................................................................................................................ XV

List of tables .............................................................................................................. XVIII

Introduction ..................................................................................................................... 1

1. Catalytic oxidation of ethylene ................................................................................ 3 1.1 Ethylene........................................................................................................... 3

1.1.1 Basic properties and reactivity ............................................................... 4 1.1.2 Ethylene as aging hormone ................................................................... 5 1.1.3 Technologies for ethylene elimination ................................................... 6

1.2 Catalyst selection ........................................................................................... 10 1.2.1 Cerium oxides ..................................................................................... 14 1.2.2 Cerium-Zirconium oxides .................................................................... 15 1.2.3 Gold as active phase ........................................................................... 17

1.3 Studies on the fluid-surface interaction .......................................................... 18 1.3.1 Temperature programmed desorption (TPD) ....................................... 18 1.3.2 Infrared spectroscopy .......................................................................... 21

1.4 Scope ............................................................................................................ 31

2. Cerium oxides as oxidation catalysts for ethylene elimination .......................... 33 2.1 Catalyst synthesis .......................................................................................... 33 2.2 Characterisation............................................................................................. 34

2.2.1 Surface area – N2 isothermal adsorption-desorption ........................... 34 2.2.2 Crystalline structure – XRD ................................................................. 35 2.2.3 Solid phase formation – Raman spectrocopy ...................................... 38

2.3 Evaluation of the fluid-surface interactions ..................................................... 39 2.3.1 Study of the adsorption-desorption mechanisms of probe molecules C2H4 and O2 via TPD ......................................................................................... 39 2.3.2 Study of the intermediary species during ethylene total oxidation via operando IR spectroscopy ................................................................................. 42

3. Cerium-zirconium oxides as oxidation catalysts for ethylene elimination ........ 55 3.1 Catalyst synthesis .......................................................................................... 55 3.2 Characterisation............................................................................................. 56

3.2.1 Surface area – N2 isothermal adsorption-desorption ........................... 56 3.2.2 Crystalline structure – XRD ................................................................. 57 3.2.3 Solid phase formation – Raman spectroscopy .................................... 59

XIV Evaluation of the surface and catalytic properties of the ethylene and oxygen


3.3 Evaluation of the fluid-surface interactions .................................................... 61

3.3.1 Study of the adsorption-desorption mechanisms of probe molecules C2H4 and O2 via TPD ........................................................................................ 61 3.3.2 Study of the intermediary species during ethylene total oxidation via operando IR spectroscopy ................................................................................ 61

4. Adsorbent (zeolite) + catalyst: a promissing alternative ..................................... 65

5. Conclusions and recommendations ..................................................................... 73

References .................................................................................................................... 75

Contents XV

List of figures

Pág.

Figure 1-1: Synthesis of chemical intermediates and products from ethylene ...................4

Figure 1-2: Generalized patterns of growth, respiration and ethylene during development,

maturation and senescence of climacteric and non-climacteric fruits .................................6

Figure 1-3: Schematic representation of the possible reaction paths and products of the

C2H4 + O2 reaction ...........................................................................................................11

Figure 1-4: Mars-van Krevelen mechanism of oxidation reactions on oxide based

catalysts ..........................................................................................................................13

Figure 1-5: CeO2 crystal structure ..................................................................................14

Figure 1-6: Cubic Close Packed (FCC) Anion Arrangement, a) circles labelled T

represent the centres of the tetrahedral interstices, b) circles labelled O represent the

centres of the octahedral interstices ................................................................................15

Figure 1-7: Oxygen (a) storage (b) release mechanism in Ce-Zr oxides .........................16

Figure 1-8: Phase diagram of the CeO2-ZrO2 binary system ...........................................17

Figure 1-9: O2-TPD profiles of (a) Pt(0.2%)/CeO2 submitted to CO2 adsorption, (b)

Pt(0.2%)/CeO2 without CO2 adsorption, and (c) CeO2 samples .......................................20

Figure 1-10: Scheme of a classic dispersive IR spectrometer .........................................22

Figure 1-11: Scheme of a typical FTIR spectrometer ......................................................24

Figure 1-12: Energy profile of a non-catalysed reaction (---) and a catalysed reaction (---)

........................................................................................................................................25

Figure 1-13: A one-dimesional potential energy schematic showing qualitative aspects of

oxygen adsorption ...........................................................................................................27

Figure 1-14: Potential adsorbed-ethylene species ..........................................................28

Figure 1-15: Chatt-Dewar-Duncanson model for ethylene adsorption .............................28

Figure 1-16: Experimental IR frequency bands of different types of formate species over

metal oxides [comb1 stands for νas(COO) + δ(CH) and comb2 stands for νs(COO) + δ(CH)]

........................................................................................................................................30

Figure 1-17: Experimental IR frequency bands of different types of surface carbonate

species over metal oxides ...............................................................................................31

Figure 2-1: XRD patterns of ceria samples with synthesised and commercial (*) supports

........................................................................................................................................36

Figure 2-2: Raman spectra for synthesised and commercial (*) ceria .............................38

Figure 2-3: Scheme of the TPD setup used ....................................................................40

Figure 2-4: Possible configurations of the valve system in the TPD setup ......................40

XVI Evaluation of the surface and catalytic properties of the ethylene and oxygen


Figure 2-5: Evolution of mass spectra signal intensities with time on stream during (a)

ethylene desorption (b) oxygen desorption over Au(2%)/CeO2 sample ........................... 41

Figure 2-6: Scheme of the operando system .................................................................. 43

Figure 2-7: General and detailed views of the “sandwich” reactor-cell. 1=Spectrometer

base-plate, 2=IR cell support, 3=adjusting nut for air tightness, 4=gas outlet, 5=air cooling

outlet, 6=gas inlet, 7=air cooling inlet, 8=external shell, 9=thermocouple location, 10=IR

beam, 11=stainless steel ring, 12=Kalrez O-ring, 13=KBr windows, 14=wafer holder,

15=oven location ............................................................................................................. 43

Figure 2-8: Evolution of temperature and mass spectra signal (m/z= 28 for ethylene, 18

for water, 44 for CO2) intensities with time on stream [total flow 25 cc·min-1: 400 ppm C2H4

+ 20% O2 in Ar] when the reaction is performed over CeO2 ............................................ 45

Figure 2-9: IR spectra of CeO2 before (a) and after (b) the reaction ............................... 45



+ 20% O2 in Ar] when the reaction is performed over Au(1%)/CeO2 ................................ 46

Figure 2-11: IR spectra of Au(1%)/CeO2 before (a) and after (b) the reaction ................ 47

Figure 2-12: IR stacked spectra (800-1150 cm-1) of Au(1%)/CeO2 (a) just before the

reaction; (b) 0 min, 25 °C; (c) 20 min, 35 °C; (d) 40 min, 80 °C; (e) 1 h, 126 °C; (f) 1 h 25

min, 171 °C; (g) 1 h 45 min, 214 °C; (h) 2 h 10 min, 259 °C; (i) 2 h 30 min, 280 °C; [total

flow 25 cc·min-1: 400 ppm C2H4 + 20% O2 in Ar] ............................................................. 48


reaction; (b) 0 min, 25 °C; (c) 20 min, 35 °C; (d) 40 min, 80 °C; (e) 1 h, 126 °C; (f) 1 h 25

min, 171 °C; (g) 1 h 45 min, 214 °C; (h) 2 h 10 min, 259 °C; (i) 2 h 30 min, 280 °C; [total

flow 25 cc·min-1: 400 ppm C2H4 + 20% O2 in Ar] ............................................................. 49


reaction; (e) 1 h, 126 °C; (f) 1 h 25 min, 171 °C; (g) 1 h 45 min, 214 °C; (h) 2 h 10 min,

259 °C; (i) 2 h 30 min, 280 °C; [total flow 25 cc·min-1: 400 ppm C2H4 + 20% O2 in Ar] ..... 49

Figure 2-15: Proposed reaction mechanism for ethylene total oxidation over

Au(1%)/CeO2 (part a) ...................................................................................................... 51


Au(1%)/CeO2 (part b) ...................................................................................................... 52

Figure 3-1: Global Au/CexZr1-xO2 catalyst synthesis procedure ...................................... 56

Figure 3-2: XRD patterns of ceria-zirconia samples with synthesised and commercial (*)

supports. Composition (wt%) Ce/Zr for CZ=0.8/0.2, for CZ*=0.6/045. ............................. 58

Figure 3-3: Raman spectra for synthesised and commercial (*) ceria-zirconia (CZ) ....... 60



+ 20% O2 in Ar] when the reaction is performed over Au(2%)/CZ* .................................. 62

Figure 3-5: IR spectrum of Au(2%)/CZ* before (a) and after (b) the reaction .................. 63

Figure 3-6: IR stacked spectra (1200-1700 cm-1) of Au(2%)/CZ* (a) just before the

reaction (b) 0 min, 25°C (c) 20 min, 35 °C (d) 40 min, 80 °C (e) 1 h, 123 °C (f) 1 h 25 min,

167 °C (g) 1 h 45 min, 211 °C (h) 2 h 5 min, 252 °C (i) 2 h 25 min, 280 °C [total flow 25

cc·min-1: 400 ppm C2H4 + 20% O2 in Ar] .......................................................................... 64

Contents XVII

Figure 4-1: Schematic of different zeolite structures .......................................................66

Figure 4-2: Evolution of mass spectra signal (m/z= 27 for ethylene, 44 for CO2) intensities

with time on stream during ethylene desorption under air over Zeolite A .........................67


with time on stream during ethylene desorption (a) under air (b) under argon over Zeolite

A + Au(1%)/CeO2 system ................................................................................................69



A + Au(2%)/CZ* system...................................................................................................70

Contents XVIII

List of tables

Pág.

Table 1-1: Observed wavenumbers (cm-1) and assignments for ethylene adsorbed on

metal oxides .................................................................................................................... 29

Table 2-1: Specifications of the precursor used for the Ce-based catalyst preparation ... 33

Table 2-2: Textural characteristics of samples with synthesised and commercial (*) ceria

supports .......................................................................................................................... 34

Table 2-3: Structural characteristics of CeO2, Au(1%)/CeO2, CeO2* and Au(2%)/CeO2*

oxides. ............................................................................................................................ 37

Table 3-1: Specifications of the precursors used for the Ce-Zr-based catalyst preparation

....................................................................................................................................... 56

Table 3-2: Textural characteristics of samples with synthesised and commercial (*) ceria-

zirconia supports ............................................................................................................. 57

Table 3-3: Structural characteristics of CZ, Au(1%)/CZ, CZ* and Au(2%)/CZ mixed

oxides. ............................................................................................................................ 59

Table 4-1: Summary of the results of the TPD experiences for zeolite A under air, zeolite

A +Au/CeO2 (under air and argon), and zeolite A +Au/CZ* (under air and argon) ........... 68

Introduction

Ethylene is a plant hormone that regulates the processes of maturation and senescence of

various horticultural products and is one of the compounds responsible for post-harvest

losses, which can be as high as 40% [1]. This problem is particularly important in tropical

regions like Colombia, since most horticultural products are climacteric and, therefore, emit

ethylene and are sensitive to concentrations above 0.1 μL L-1 [2].

Many technologies have been developed to control ethylene emissions. The most widely

used are cold storage and hormone control, which have presented drawbacks due to their

extensive energy consumption, to the toxicity of the compounds used and to the rapidity

with which some products regenerate their ethylene receptors [3]. Taking this into account,

the use of non-invasive and efficient technologies for the removal of ethylene is a must in

order to develop industrial solutions.

The actual trend in research for ethylene elimination has been the hormonal control of

ethylene by the use of 1-methylcyclopropene (1-MCP). Regarding physicochemical

methods, some authors have reported the use of substances to absorb ethylene such as

silica, zeolites or activated carbon, having the problem of rapidly saturating with water due

to the high relative humidity (80% - 100%) in the environment where the horticultural

products are stored. It has also been reported the use of strongly oxidizing agents such as

potassium permanganate, although it is a toxic substance. As for the catalytic oxidation,

metals such as palladium, platinum and gold have shown an important activity for oxidation

reactions at relatively low temperatures (0°C - 200°C). Advanced oxidation through the use

of titanium dioxide and ultraviolet light, has also been reported. It is noteworthy that these

researches have shown successful results in horticultural products in temperate latitudes,

finding problems in tropical products such as bananas, mango and avocado [4].

The present work aims to evaluate the surface and catalytic properties of mixed oxides of

cerium-zirconium for the elimination of ethylene by catalytic oxidation as a strategy to

2 Introduction

increase the post-harvest useful life of fruits and vegetables. This thesis complements the

doctoral work of Rolando Mendoza [5], in which it is proposed to develop a catalyst for the

total oxidation of ethylene based on gold nanoparticles supported on mixed oxides of Ce-

Zr-Co. The influence of the presence of gold and the nature of the support on the adsorption

mechanism of ethylene and oxygen, are studied through the use of spectroscopic

techniques.

A catalyst composed of a mixed oxide of cerium/zirconium as the support/carrier and gold

nanoparticles active site, is used. The selection of gold as active principle is due to its high

catalytic activity at low temperatures, when it has a particle size smaller than 4 nm. The

selection of a mixed oxide cerium-zirconium is due to the high dispersion of gold in ceria

(cerium oxide) and the possibility of having a good redox cycle due to the oxygen vacancies

that can be generated by zirconium and cerium oxide.

This document has been divided into five main chapters. The first one is a brief introduction

to the state of the art and some generalities concerning current ethylene elimination

methods for horticultural applications, including total catalytic oxidation as an alternative. It

describes the catalyst used, some structural and physicochemical properties, and the

reason for selecting a ceria-based oxide for this specific purpose. It also introduces the

main techniques used to elucidate the fluid-surface interactions that take place during the

oxidation reaction.

Chapter two and three are dedicated to the study of ceria and ceria-zirconia based

catalysts, respectively, on the total oxidation of ethylene. The influence of gold as active

phase is elucidated. Synthesis, characterisation and evaluation techniques used are

described in detail among with the results obtained.

Chapter four explores a novel strategy for ethylene elimination through a system composed

of an adsorbent and a catalyst. This, in view of the fact that the catalyst is only active at

temperatures above 100 °C. The system allows to use a regenerable adsorbent (in this

case, a zeolite) to trap a certain quantity of ethylene at room temperature. Ethylene is

subsequently desorbed and reacts with the catalyst in the presence of hot air.

Chapter five summarises some conclusions and recommendations for future work.

1. Catalytic oxidation of ethylene

In this chapter, basic properties of ethylene are described, as well as its role as ripening

hormone in plants and fruits. Current technologies for ethylene elimination are presented,

introducing total catalytic oxidation as an alternative to what is conventionally used. The

use of a proper catalyst is key in the development of this strategy at the industrial scale,

thus, the selection of ceria-zirconia based catalyst is explained in detail. Finally, the main

techniques used to study the fluid-surface interaction are introduced.

1.1 Ethylene

Ethylene (C2H4), also known as ethene, is a colourless, gaseous organic compound. It is

the simplest of the alkenes (structures that contain a carbon-carbon double bond) and is

the most important organic compound manufactured nowadays, with a worldwide

production of over 150 million tons in 2016, exceeding that of any other organic compound

[6]. It is the building block for a vast range of chemicals from polymers to solvents.

It is mostly used to produce ethylene oxide, ethylene dichloride, ethylbenzene and

polyethylene, which are precursors to many other compounds (Figure 1-1) [7]. Ethylene is

usually obtained from petroleum and natural gas, by steam cracking, which is the thermal

breakdown of long chain saturated hydrocarbons into smaller, often unsaturated ones. In

the US, Canada and the Middle East, it is primarily obtained by cracking of ethane and

propane. In Europe and Asia, it is mainly obtained from cracking naphtha, gasoil and

condensates with the coproduction of propylene, C4 olefins and aromatics [8]. It can also

be obtained by advanced catalytic cracking of olefins.

4 Evaluation of the surface and catalytic properties of the ethylene and oxygen


Figure 1-1: Synthesis of chemical intermediates and products from ethylene [7]

1.1.1 Basic properties and reactivity

Ethylene has a poor solubility in water 3.5 mg·100 mL-1 but a good solubility in diethyl ether

due to its nonpolar nature. Its gas density is 1.178 kg·m-3 at 15 °C, its boiling point −103.7

°C (1 atm) and its melting point −169.2 °C (1 atm) [9].

A closer look into the molecular structure of ethylene allows us to understand the reasons

for its multiple applications, functionality and reactivity. In order to have a deeper

understanding, we must remember that a double carbon-carbon bond is composed by σ-

bond and a π-bond. A σ-bond is a covalent bond resulting from the formation of a molecular

orbital by the axial or end-to-end overlap of atomic orbitals, and a π-bond, results from the

formation of a molecular orbital by side-to-side overlap of atomic orbitals along a plane

perpendicular to a line connecting the nuclei of the atoms. π-bonding engages overlapping

of p (in the ethylene case) or d atomic orbitals, which result in the formation of π (bonding)

and π* (antibonding) molecular orbitals. Bonding orbitals stabilise the molecu le and have

lower energies because they are closer to the nuclei of the atoms. Antibonding orbitals have

higher energy levels and less electron density between the nuclei [10].

Usually the π-bond is weaker than the σ-bond, thus, it is not necessary a large amount of

energy to break it. It dominates the chemistry of ethene and is responsible for its high

reactivity. Being a region of high electron density, the π-bond is very vulnerable to attack

by electrophiles. Many reactions of ethylene are catalysed by transition metals, which bind

transiently to the ethylene using both the π and π* orbitals.

Catalytic oxidation of ethylene 5

1.1.2 Ethylene as aging hormone

In addition to having a wide range of practical and commercial applications, ethylene is also

known as the aging hormone in plants. It is synthesised in most tissues of plants and has

several effects on their physiology. It plays a key role in the processes of growth,

development, maturation, and senescence of all climacteric plants or fruits. Ethylene is

biologically active at very low concentrations in the ppm and ppb range, in fact, it can induce

quality loss of a wide variety of fruits and vegetables when present at concentrations above

0.1 μL·L-1 [11].

Fruits and vegetables may be classified depending on whether or not they exhibit a peak

in respiration and ethylene production during ripening (Figure 1-2). Climacteric species are

characterised by an increase in respiration ethylene production as they ripen [12], thus,

they are subject to massive deterioration during the post-harvest period, characterised by

excessive softening and changes in taste, aroma and suface colour [13]. Non-climacteric

species, such as grape, orange, pineapple and leafy vegetables, are characterised by the

lack of ethylene-associated respiratory peak, thus, they do not continue to ripen after

harvest. They may eventually soften and rot, but this is due to moisture loss, decay and

tissue degeneration.

Deterioration of climacteric fruits is accentuated during the commercialisation stage since

these fruits are transported for great distances and stored during long periods of time.

Taking into account the importance of the product perception by the final consumer, and

that this can be associated to quality parameters such as fruit firmness, percentage of

acidity and soluble solids which at the same time are indirectly related to sensory quality

characteristics (such as colour, taste, texture and aroma) and nutritional characteristics

(content of vitamins, minerals, antioxidants, etc.), a deterioration in these properties results

in large economic losses [14]. In fact, according to the Food and Agriculture Organization,

post-harvest losses are estimated to be as high as 40% [1]. This is why there has been a

broad interest in the search for alternatives and strategies to prolong the conditions of the

fruit during the post-harvest, consumption and commercialisation stage, by removing

ethylene and, therefore, suppressing its effects [15].



Figure 1-2: Generalized patterns of growth, respiration and ethylene during development,

maturation and senescence of climacteric and non-climacteric fruits [16]

1.1.3 Technologies for ethylene elimination

Techniques to limit biosynthesis of ethylene mostly focus on increasing or decreasing

temperature and modifying the atmosphere in which fruits are preserved, which implies

economic investments in transport and cold storage [17]. One of the disadvantages of using

low temperatures is cold damage, when temperatures below the critical temperature of

each product are used [3], which in tropical fruits can be as high as 10-12 °C. As for the

use of controlled and modified atmospheres, it is sought to delay the maturation of the

product by varying the contents of O2 and CO2 in the surrounding environment, affecting

the processes of respiration and gas emission. However, industrial application of this

alternative has not been very successful due to the lack of economic incentives to extend

its use. In addition, the product is generally exposed to varying temperatures during


handling, transportation, storage and marketing. This creates a problem in the design of

modified atmospheres because it is difficult to maintain optimum gas concentrations when

the surrounding temperature is not constant [18].

The complete elimination of ethylene can be achieved through various chemical or physical

methods, which exploit the ability of certain substances to oxidise, decompose or adsorb

ethylene through different mechanisms (degradation/oxidation, adsorption, inhibition of the

receptor and the synthesis, catalysis and photocatalysis).

1.1.3.1 Chemical/physical methods

Ethylene can undergo various reactions for degradation. In fact, its double bond makes it

highly reactive and prone to different reaction mechanisms. Ethylene can adsorb UV

radiation at 161, 166 and 175 nm [19]. When it reacts with ozone, it produces water, carbon

dioxide (CO2), carbon monoxide (CO), and formaldehyde [20]. Even in low ozone

concentrations, effective removal of ethylene can also be achieved as ultraviolet light

interacts with oxygen (O2) in the air to form ozone, which breaks the ethylene molecule,

and at the same time, UV light interacts directly with ethylene to degrade it [21]. However,

this reaction is inefficient in atmospheres of very low ethylene concentration, such as those

found in the storage and transport processes of fresh fruit, so the commercial potential of

ozone as an ethylene scrubber is limited [22].

One of the first chemical agents used for the oxidation of ethylene was potassium

permanganate (KMnO4), with the ability to decompose ethylene to acetaldehyde and,

subsequently, acetic acid, which in turn can be oxidised to CO2 and H2O [23]. KMnO4 is

often supported on celite, vermiculite, silica gel, zeolite or alumina pellets in order to

increase the surface area [24]. Currently, commercial formulas containing KMnO4 are the

most commonly used in the industry to reduce post-harvest losses. However, direct contact

of potassium permanganate with fruits is not allowed due to its high toxicity, so films or

filters are used in combination with inert substrates such as silica gel, activated carbon,

zeolite or alumina in order to avoid this [25].

Ethylene may be subjected to hydrogenation in the presence of metal catalysts. It can also

react with halogens (mainly chlorine and bromine), in halogenation and hydrohalogenation



reactions to form dihaloalkanes. In this way, ethylene can be removed by passing it over

brominated activated carbon to form dibromoethane [26]. These types of filters are relatively

efficient in the removal of ethylene, however, the bromine can also react with water to form

HBr and Br2. These compounds are harmful to plant tissues and corrosive to stainless steel.

In addition, the brominated activated carbon is hygroscopic, so it is possible that the filter

adsorbs a lot of moisture, which is inconvenient [22]. Alternatively, the ethylene may react

with hydrogen halides to form ethyl halides. It can also react with concentrated sulfuric acid

to form ethyl hydrogen sulphate or with water in acid medium to form ethanol [27].

It is clear that ethylene is a highly reactive compound and can be degraded by different

reaction routes. However, typical reactions generally require high ethylene concentrations

and/or high pressure and temperature conditions.

In addition to the chemical degradation methods, ethylene can be adsorbed by different

substances including activated carbon, crystalline aluminosilicate molecular sieves,

Kieselguhr, bentonite, Fuller's earth, silica gel, aluminium oxide and zeolites [28]. The

activated carbon may be granular, powder or fibre. According to Martínez-Romero et al.

[29], the best results in terms of ethylene adsorption rates were obtained with granular

carbon (80%), followed by powder (70%) and fibre (40%).

The use of some regenerable adsorbents, such as propylene glycol, hexylene glycol,

squalene, phenylmethylsilicone, polyethylene and polystyrene has also been investigated

[30]. These have the advantage of being reusable after being purged. However, the main

problem is that adsorption techniques only transfer ethylene to the adsorbent solid matrix,

rather than destroying it.

At the commercial scale, only some of the technologies described above are used in

product packaging. Substances designed to remove ethylene during storage are generally

designed to go into the package (in the form of sachets) or integrated with the packaging

material. Adsorbents based on KMnO4, for example, come in the form of sachets, blankets,

or tubes. The sachets are used in individual fruit and vegetable boxes while the others are

used in vehicles where they are transported [31]. KMnO4 is not impregnated directly into

the package due to its toxicity. Typically, these products contain about 4-6% KMnO4


supported on an inert substrate [32]. It has been proposed to use other materials such as

activated carbon, bentonite, Kieselguhr, zeolites and aluminosilicates. Metal catalysts on

activated carbon also effectively remove ethylene. Commercially, sachets containing

activated carbon impregnated with palladium are used [33]. A variety of packages

impregnated with clays or 'activated earth' have also been commercialised. Usually these

are bags of polyethylene impregnated with zeolites. Although these minerals have some

capacity for adsorption of ethylene, there is no information to prove its effectiveness. In fact,

in plastic films, the ethylene would have to diffuse through the polymeric matrix of the

package before coming into contact with the dispersed mineral, making this an excessively

slow and inefficient process [31].

1.1.3.2 Biological methods

Alternatively, ethylene synthesis inhibitors such as aminoethoxyvinylglycine (AVG) and

aminooxyacetic acid (AOA) can also be employed, which effectively retard the senescence

of climacteric products by inhibiting the action of 1-aminocyclopropane-1-carboxylate

synthase (ACCS), the enzyme that catalyses the synthesis of 1-aminocyclopropane-1-

carboxylic acid (ACC), a precursor of ethylene [34]. Ethylene synthesis inhibitors are

applied pre-harvest. Another strategy commonly used involves the use of the ethylene

action inhibitor, 1-methylcyclopropene (MCP-1), which block ethylene by binding to its

receptors. It is applied post-harvest [35], [36].

1.1.3.3 Catalytic oxidation of ethylene

During the last years, a recent development has been seen in the removal of biogenic

ethylene by total catalytic oxidation. Among the drawbacks of the catalysts which have been

reported above is that they are limited by their low efficiency and high reaction temperature.

Considering that the storage or transport temperature of the fruit should be lower than 10°C

[37], the presence of an active catalyst at relatively low temperatures is essential for the

development and implementation of this process at the commercial level. Other desirable

characteristics for this catalyst are: a high ethylene adsorption capacity and the capacity to

promote the breakdown of the carbon-carbon bond (C=C).



Several catalysts with transition metals as active phase (Pd, Pt, Rh) [38] supported on metal

oxides (TiO2, Al2O3, zeolites) [39], [40] or black carbon have been evaluated in the catalytic

oxidation of ethylene. But metals like palladium and platinum are only active at

temperatures above ambient. The use of Cu-Mn porous oxides has also been studied,

however, these materials only exhibit high conversion in the ethylene oxidation reaction at

temperatures around 200 °C [41]. The use of photocatalytic materials that are irradiated

with ultraviolet light to remove ethylene through a light-promoted oxidation mechanism on

TiO2 nanoparticles has also been investigated [42] and also using TiO2/SiO2 mixed phase

nanocomposites [23]. However, the main limitation is the low efficiency of photocatalysis in

TiO2.

A very attractive option is the use of Au/Co3O4 and Au-Ag/Al2O3 catalysts [43]. Ahn et al.

[44] observed that at 0 °C, the total oxidation reaction of ethylene can achieve a conversion

of up to 93.7% on an Au/Co3O4 catalyst. It is interesting that although the use of the Co3O4

support exhibits no activity, the addition of gold as the active phase greatly increases the

activity of the catalyst. In fact gold has a great catalytic activity when it has a crystal size

smaller than 4 nm. But for the oxidation reaction it is necessary, at the same time, a

promoter or support that allows the formation of oxygen vacancies, for which mixed oxides

are generally used to generate a good redox cycle.

1.2 Catalyst selection

Depending on the type of the catalyst and reaction conditions, one could either favour the

selective or partial oxidation of ethylene leading to the formation of ethylene oxide

(epoxide), or the total oxidation or combustion reaction leading to the formation of CO2 and

water, as seen in Figure 1-3. Usually, total oxidation reactions are thermodynamically much

more favourable, although it is common that products of partial oxidation are intermediates

in the combustion process that forms carbon oxides.


Figure 1-3: Schematic representation of the possible reaction paths and products of the

C2H4 + O2 reaction [38]

In presence of water, ethylene oxidation may even lead to the formation of acetic acid via

acetaldehyde. It can also just hydrate and form ethanol following another catalytic path [45].

Overall in a catalytic reaction, activity and selectivity depend pretty much on the chemical

nature of the catalyst. This includes the composition and distribution of the components,

the textural properties (surface area, pore volume, pore size), and other process variables

[46]. Among some of the desirable characteristics of the catalyst in order to favour the total

oxidation of ethylene to CO2, the following phenomena is expected:

(i) High ethylene adsorption capacity

(ii) The preferential cleavage of C-C, O-H, and C-H bonds

(iii) No cleavage of C-O bonds

(iv) No hydrogenation of CO or CO2 to form alkanes

(v) High oxygen mobility

(vi) High activity at low temperatures

The C-H bond breaking of alkenes is enhanced by strongly basic surface lattice oxygen

and cations that are soft acid and undergo redox readily. Due to their good redox

(oxidoreductive) properties, multicomponent oxides are often used for oxidation reactions.

These are based on a few number of transition-metal oxides as basic components. For



example, for selective oxidation reactions, these basic components include oxides of

molybdenum, vanadium, copper, iron, antimony, tin and uranium. For total oxidation

reaction the basic components include oxides of cerium, cobalt, iridium, and zirconium [47],

[48].

Often in these cases, a second element is also added, acting as a promoter which main

purpose is to enhance catalyst activity and selectivity, by enhancing the rate of the rate-

determining steps. Improved catalyst activity can be due to many reasons: the formation of

new compounds, increase of acidity or basicity, or generation of defect sites for

hydrocarbon or oxygen activation. Promoters can be transition or post-transition metals.

Generally, presence of small quantities of promoter can optimise the performance of the

principal active elements [47].

Even though a mixed oxide support favours oxidation reactions due to their high oxygen

exchange, it is still necessary an active principle which acts as a strong dehydrogenating

agent. The most used materials as active phase are noble metals like palladium or platinum,

due to their high dehydrogenating capacity. In addition, since the early studies by Haruta

and co-workers, there has been a growth of papers and reports about the unusual activity

of nanosised gold-based materials in oxidation reactions [49].

Within a mixed oxide based catalyst, total combustion reactions follow a Mars-Van Krevelen

mechanism, where the catalyst surface oxidises the hydrocarbon and then is re-oxidised

by molecular oxygen from the gas phase, following a redox cycle. This is shown in a more

detailed way in Figure 1-4. The reactant molecule interacts strongly with the catalyst at site

containing M1n+. Lattice oxygen is consumed to form products, and the M1 cation is reduced.

The active site and cation M1 are re-oxidised by the migration of lattice oxygen from M2,

which in turn is re-oxidised by the gas phase oxygen. M2 may or may not be distinct from

M1 [50]. Although Figure 1-4 shows a Mars-Van Krevelen cycle for a bimetallic oxide, a

catalyst composed of a transition metal and a metal oxide might follow the same type of

mechanism as well.


Figure 1-4: Mars-van Krevelen mechanism of oxidation reactions on oxide based catalysts

[50]

There are some additional features to optimise total combustion reactions. This can be

achieved by increasing the residence time of the surface intermediates, strengthening the

adsorption of the desired products and optimising the amount of weakly adsorbed oxygen

or the density of combustion sites. The support also plays an important role on this, since

reaction products also depend on its surface area. A not so large surface area limits the

rate of total combustion products, since it does not favour the subsequent reactions to CO2

of partially oxidised products. While a large surface area enhances the dispersion of active

elements and favours the formation of total combustion products, which are dependent on

their ability to diffuse from the active centre into the gas phase [51].

Fluorite-type mixed oxides are characterised by a high oxygen mobility. Cerium based

catalysts are widely known for their redox properties and high performance due to the

generation of oxygen vacancies, which make them ideal in oxidation reactions. It has also

been reported that the introduction of zirconium improves the redox properties of cerium

and thermal stability, enhancing the oxygen storage capacity and, therefore, its catalytic

properties [52]. In addition, gold as active phase has shown a great catalytic activity when

it has a particle size smaller than 4 nm. Gold has a strong interaction with ceria-based



supports due to their defect-type structure, it shows a good dispersion and it helps

enhancing the reducibility of the material [53], [54].

1.2.1 Cerium oxides

Cerium is the second element in the lanthanide series. It has a ground state electron

configuration [Xe] 4f15d16s2 and it can exhibit both +3 and +4 oxidation states, although it

acquires a more stable structure on the tretavalent form Ce4+. Cerium oxide crystallises in

a fluorite-type structure. It has a face-centred cubic unit cell (fcc) with a space group Fm3m.

Each cerium cation (Ce4+) is coordinated by eight equivalent nearest-neighbour oxygen

anions (O2-) at the corner of a cube, each anion being tetrahedrally coordinated by four

cations [55], as seen in Figure 1-5.

Figure 1-5: CeO2 crystal structure [56]

If we take a closer look to the common ccp (cubic close) anion packing, it can have

tetrahedral as well as octahedral interstices. Figure 1-6 helps to illustrate both the centres

of the tetrahedral (8 sites) and octahedral (4 sites) holes in the ccp arrangement of an fcc

unit cell. Comparing to the ceria structure, it is easy to see that all the tetrahedral sites in

ceria are occupied by the oxide ions whereas the octahedral sites remain empty. This is

why oxygen mobility is favoured, since the oxygen anions tend to move into the octahedral

holes [57].


Figure 1-6: Cubic Close Packed (FCC) Anion Arrangement, a) circles labelled T represent

the centres of the tetrahedral interstices, b) circles labelled O represent the centres of the octahedral interstices [58]

The outstanding role of ceria has been recognised in three-way catalysis, water-gas shift

(WGS) reactions, wet oxidation, combustion reactions and solid oxide fuel cells [59]-[61].

This is due to its high OSC (oxygen storage capacity) attributed to the reversible

transformation of oxidation states between Ce4+ and Ce3+ ions [62].

1.2.2 Cerium-Zirconium oxides

One well-known drawback in ceria catalysts is their low stability in high temperature ranges.

Textural properties of ceria can be improved by the introduction of doping agents in the

fluorite lattice. One of the most efficient dopants studied has been zirconium [60], [63]. The

incorporation of elements such as Zr4+ in the ceria network enhances the catalyst redox

properties by increasing its oxygen storage capacity, favours the reduction of Ce4+ to Ce3+,

preserves the oxygen vacancies and prevents catalyst sintering by enhancing the thermal

resistance [52].

The catalytic effects of zirconium introduction into the ceria structure can be mainly

explained by the distortion that it causes. This distortion is introduced primarily by the

difference in the ionic sizes between Ce4+ (0.87 Å) and Zr4+ (0.72 Å) when in a hexahedral

position. Taking into account that during the oxygen storage and release process, the

volume of the cerium cell increases due to the change in the cerium oxidation state, from

Ce4+ (0.87 Å) to Ce3+ (1.01 Å), the stress energy generated from this volume increase can

be diminished by Zr introduction, facilitating the valence change process [64], [65]. The



oxygen storage-release mechanism of a Ce-Zr mixed oxide can be well appreciated in

Figure 1-7.

Figure 1-7: Oxygen (a) storage (b) release mechanism in Ce-Zr oxides [66]

The Ce-Zr solid solution can adopt a different crystal structure depending on the Ce/Zr ratio

and temperature. According to the CeO2-ZrO2 phase diagram (Figure 1-8), below 1000 °C

(1273 K), ceria-zirconia exhibits a cubic (c) structure when Zr content is less than 20%. This

structure is common to both pure ceria and cubic zirconia. However, at high Zr contents

(>80%), the compound exhibits a monoclinic (m) symmetry. In the intermediate region, the

true nature of the Ce-Zr structure is still matter of debate due to the presence of a number

of stable and metastable phases of tetragonal symmetry. In fact, three different tetragonal

phases can be distinguished t, t’ and t’’. The t phase has been reported to be the most

stable tetragonal phases. The t’’ phase is an intermediate between t’ and c; in fact, it shows

no tetragonality and is generally referred to as cubic phase, since its XRD pattern is indexed

in the cubic Fm3m space group [67], [68].

Although according to the phase diagram the limit between the tetragonal and the cubic

phase may be found at CeO2 molar compositions of around 65%, previous reports have

found that a Ce0.5Zr0.5O2 is still compatible with the cubic symmetry, being on the limit of the

phase transition. This composition can either take the tetragonal or cubic form by simply

changing the preparation conditions [69].


Figure 1-8: Phase diagram of the CeO2-ZrO2 binary system [70]

1.2.3 Gold as active phase

Among with group VIII metals, Cu and Ag from group IB have been the most widely used

in catalytic reactions due to their activity. The catalytic excellence of group VIII metals can

be attributed to the presence of unpaired electrons in the d-band, which is why they are

classified as paramagnetic. While, in the case of Cu, Ag and Au from group IB, they have

fully occupied d-electron bands (diamagnetic), however, they have a relatively low

ionization potential, being able to readily lose electrons to yield d-band vacancies. The

presence of unpaired electrons in the d-band is essential to hydrogen chemisorption, as

they stabilise an intermediate weakly-held form without which dissociation cannot occur

[71].

Gold, despite belonging to the group IB of the periodic table, has often been regarded to

have a poor catalytic activity. In bulk, it is chemically inert. However, recent discoveries

have shown that when it is small enough (particle size < 4nm) it turns out to be surprisingly

active for many reactions. In fact, there has been a significant growth on papers about the

high activity of nanosised gold since the early studies by Haruta and co-workers, who

realised its unusual catalytic activity at sub-ambient temperature [72]. It is thought that the



exceptional ability of gold is ascribed to the presence of Au atoms in low coordination states

on the surface of the nanoparticles [73], [74].

As just mentioned, gold’s particle size is critical when in catalytic reactions. The particle

size of an active metal depends highly on the preparation method used [75]. However, the

general catalytic performance is not defined only by Au particle diameter but also by other

major factors like support selection and contact structure. In fact, in this case, gold’s

remarkable ability also depends on the mutual interaction of the nanoparticles and the

support, which not only depends on the synthesis method, but also the support used [74].

Taking this into account, a ceria-based support is also an excellent alternative because of

gold high dispersion on it. Gold also promotes the reducibility of ceria at low temperatures,

(although this degree of reduction depends on the percentage of incorporated zirconia) [76],

[53]. In some situations, the active site of the correspondent reaction is the interface gold-

support, as it is the case of CO oxidation [77]. But for WGS, for example, active sites are

small gold particles in a close contact with oxygen vacancies of ceria [78].

1.3 Studies on the fluid-surface interaction

Since the presence of an active catalyst at relatively low temperatures is a key for the

development and application of the catalytic oxidation of ethylene at the commercial scale,

the identification and understanding of the adsorption and reaction mechanisms that take

place during this process are fundamental for the proper design of a ceria-based material.

Adsorption-desorption of probe molecules can be studied via temperature programmed

desorption (TPD). This technique helps to quantify the amount of gas that adsorbs on the

catalyst surface at room temperature. Identification of molecular intermediary species that

adsorb on the surface during the reaction can be performed by infrared (IR) operando

spectroscopy.

1.3.1 Temperature programmed desorption (TPD)

The quantification of chemical species that dissociate or desorb the surface can be

performed by temperature programmed desorption (TPD), also referred as thermal

desorption spectroscopy (TDS). Through TPD techniques it is also possible to determine


the number, type and strength of available active sites on the catalyst surface from

measurements of the amounts of desorbed gas at several temperatures [79].

Adsorption takes place when an attractive interaction between a particle and a surface is

strong enough to overcome the disordering effect of thermal motion. This can occur by

mechanisms of physisorption or chemisorption. Physisorptive bonds occur when the

interaction is essentially the result of van-der-Waals forces, they are characterised by

dissociation energies of approximately 50 kJ·mol-1. Chemisorption, as stated by its name,

involves the formation of a chemical bond which occurs by the overlap between the

molecular orbitals of the adsorbed molecule and the surface atoms. Chemisorptive bonds

often require that an activation barrier is overcome, this is why chemisorption usually leads

to dissociation of the molecule. These bonds are characterised by dissociation energies

exceeding 50 kJ·mol-1 [80].

A typical TPD system is mainly composed of a gas flow set-up, reactor or sample tube that

goes into a furnace, and detector. The procedure is simple, once the sample has been

degasified, reduced or pre-treated in some way, a constant stream of the analyte gas is

flowed through the sample (a fixed bed of catalyst) so that it adsorbs onto the active sites

of the catalyst. The programmed desorption begins by increasing the temperature linearly

with time while a continuous stream of inert carrier gas flows through the sample. At a

certain temperature, the thermal energy will exceed the heat of adsorption of the molecule,

thus, the bond between the adsorbate and the adsorbent will break, and the adsorbed

species will be released from the surface and swept along with the carrier gas. If there is a

presence of different active metals, the chemical bond between the adsorbed molecule and

each type of metal will probably be of different energy. Consequently, molecules adsorbed

on each active metal will require a different level of thermal energy to break the bond and

desorb, resulting in different peaks in the final TPD spectrum.

The detector is generally a Quadrupole Mass Spectrometer (QMS), although it is also usual

to attach a differential thermal conductivity (TCD) detector. Either the mass over charge

(m/Z) of TCD measured by the detector at any time is proportional to the instantaneous

molecular concentration of desorbed molecules. The volume of desorbed species obtained

from the integration of the peak, combined with a factor of stoichiometry, allows to obtain



the number of active sites. As said before, presence of multiple peaks indicate the energy

differences in the active sites [81].

One of the important considerations in order to propose the reaction mechanism of the total

oxidation of ethylene on ceria-based samples is to study the adsorption-desorption

mechanisms of the reactants, in this case, C2H4 and O2, via TPD. Literature concerning

C2H4 or O2 desorption profiles on ceria or ceria-zirconia samples is very scarce. In fact,

there are no reported C2H4- or O2-TPD profiles of gold-containing ceria-based samples.

However, there are some reported O2-TPD profiles of other transition metals supported on

ceria. Figure 1-9 shows O2-TPD profiles of high surface area platinum-containing ceria

sample [82]. In this case, Gaillard wanted to stablish whether there was a competitive

adsorption between CO2 and O2, and the role of platinum as active phase during the

activation and adsorption of oxygen. First, it is noticeable that no oxygen desorption is

observed arising directly from the ceria support alone (c). This is probably due to the lack

of Pt atoms which may act as portholes for oxygen adsorption, as generally described in

the literature [83]. This means, the activation of molecular oxygen is done on the transition

metal atoms. Also, the MS signal (m/e = 32) exhibits two desorption peaks at about 450

and 620 °C, consistent with two kinds of adsorbed oxygen species. The fact that the shape

and area of the peaks are not affected by CO2 adsorption indicates the lack of competition

between the CO2 probe molecule and adsorbed oxygen.

Figure 1-9: O2-TPD profiles of (a) Pt(0.2%)/CeO2 submitted to CO2 adsorption, (b)

Pt(0.2%)/CeO2 without CO2 adsorption, and (c) CeO2 samples [82]


1.3.2 Infrared spectroscopy

Spectroscopy techniques take advantage of the interaction between light and matter. The

interaction between radiation, which is a flow of energy photons, and a molecule in a given

state can result in different phenomena such as absorption, diffusion and emission. The

exact nature of this interaction depends on the energy of the radiation [84].

In the case of the infrared radiation, it happens to correspond to most of the molecular

vibration frequencies. Thus, if a molecule is irradiated by an electromagnetic wave in the

infrared domain, the incident wave will be absorbed whenever the frequency of the incident

wave is equal to one of the vibration frequencies νvib of the molecule. This will only happen

if the transition is allowed, that is, if the dipole moment of the molecule changes as a result

of the excitation in the vibrational states [85], [86].

The typical units in infrared spectroscopy are wavenumbers, �̃�, which is just the reciprocal

of the wavelength, λ, according to equation (1.1). The energy of the irradiated photon can

be calculated as the product of the Planck’s constant, h, and the frequency, 𝜈, according to

equation (1.2).

�̃� =1

𝜆=𝜈

𝑐 (1.1)

𝐸 = ℎ ∙ 𝜈 = ℎ ∙ �̃� ∙ 𝑐 (1.2)

The total portion of the infrared radiation in the electromagnetic spectrum extends from

14000 to 20 cm-1, although the mid-infrared range, which approximately goes from 4000-

400 cm-1, is the most useful. It gives information about the strength of the bonds between

atoms and the nature of molecular vibrations. The contributions of different groups within a

specific molecule appear in different frequencies. Thus, IR spectroscopy is used for

qualitative analysis to determine specific functional groups and molecular structure. Since

the intensity of the bands observed may indicate the concentration of the species, it is

possible to make a quantitative analysis in some cases [87].

IR spectroscopy is capable to analyse gas, liquid and solid samples. In heterogeneous

catalysis, it is often used to observe adsorbed species on a solid surface. The adsorption



of a molecule onto a surface often comes with a loss of symmetry, activating vibrational

modes that were not permitted before, when in the gas phase. In conclusion, interaction

between adsorbed species and their environment (the surface or other molecules) influence

the observed bands on the spectrum.

1.3.2.1 Dispersive IR spectrometers

Dispersive IR spectrometers are often referred to as scanning spectrometers. Their basic

components are a radiation source, grating-slit system (monochromator), and detector, as

shown in Figure 1-10. They generally have a double-beam design with two equivalent

independent beams from the same source which pass through both the sample and a

reference species for analytical comparison. These reference and sample beams are

alternately focused on the detector by making use of an optical chopper, such as, a sector

mirror. After the source energy is directed along both the sample and the reference, the

emitted wavefront of radiation is dispersed by a monochromator into its component

frequencies. The monochromator is composed by a grating, which disperses the light into

its component frequencies (spatially separated wavelengths), and a narrow slit, which

chooses which frequencies go onto the detector. Finally, the detector converts the analogue

spectral output into an electrical signal, which is further processed by the computer using

mathematical algorithm to arrive at the final spectrum [88].

Figure 1-10: Scheme of a classic dispersive IR spectrometer [89]

The common IR radiation sources are inert solids that are electrically heated to a

temperature up to 1300-1700 °C to promote thermal emission of radiation in the infrared


region of the electromagnetic spectrum. The most used ones are the Nernst glower, which

is a cylinder of rare earth oxides with platinum wires sealed to the ends, the globar source,

which is a silicon carbide rod, the incandescent wire source, which is a tightly wound coil of

nichrome wire, and the carbon dioxide laser [90].

As for the detectors, they can be classified in three categories: thermal, pyroelectric and

photoconducting. Examples of thermal detectors are a thermocouple or a bolometer. They

can be used over a wide range of wavelengths and they operate at room temperature. Their

main disadvantages are slow response time and lower sensitivity relative to other types of

detectors. Pyroelectric detectors consist of a pyroelectric material, triglycerine sulphate is

the most commonly used. They have a fast response time which make them ideal in most

Fourier transform IR instruments. Photoconducting detectors rely on the interaction

between photons and a semiconductor. They consist of a semiconductor materials, mercury

cadmium telluride (MCT) is widely used for mid-IR region. It is the most sensitive detector

with better response characteristics than the others [91].

Dispersive IR instruments operate in the frequency domain. The wavelengths are measured

one at a time, with the gratingslit combination selecting the wavelengths being measured.

The slit controls the spectral bandwidth, which is how wide a range of frequencies are

permitted to strike the detector. Narrower slits give better resolution by distinguishing more

closely spaced frequencies of radiation. Wider slits provide better sensitivity by allowing

more light to reach the detector [86].

1.3.2.2 Fourier transform infrared (FTIR) spectrometers

FTIR spectrometer belongs to the third generation of IR spectrometer, and it is the most

widely used since the mid-1980s. Unlike the dispersive IR spectrometer, it operates in the

time domain and collects all wavelengths simultaneously. It is able to do that by first

collecting an interferogram of a sample signal using an interferometer, and then performing

a Fourier Transform (FT) on the interferogram to obtain the spectrum. Thus, the FT is simply

a mathematical function to convert data from the time domain to the frequency domain [92].

As show in Figure 1-11, an FTIR is typically based on The Michelson Interferometer

Experimental Setup. It consists of a source, interferometer and detector. The interferometer



consists of a beam splitter and two mirrors, a fixed one and a movable one that translates

back and forth, very precisely thanks to a laser frequency mechanism. Radiation from the

source strikes the beam splitter and separates into two beams roughly of the same intensity.

One beam is reflected off the beam splitter to the fixed mirror; the second is transmitted

through the beam splitter to the movable mirror, thus, the path length of this beam is

variable. The fixed and moving mirrors reflect the radiation back to the beam splitter, where

it is recombined forming an interference pattern. Interference will be constructive if the

difference in path lengths is an exact multiple of the wavelength. If the difference in path

lengths is half the wavelength then destructive interference will result [93], [94].

Figure 1-11: Scheme of a typical FTIR spectrometer [95]

Then, the IR beam passes through the sample, which can absorb some portion of the

energy. The transmitted portion reaches the detector, which records the total intensity and

yields an interferogram. The signal goes then through an amplifier, and is subsequently

converted by an analogue-to-digital converter. Finally, it is transferred to a computer in

which Fourier transform is carried out.

The final spectrum can be presented as transmittance (%T) or absorbance. It is important

to collect a reference or “background” single beam which is then automatically subtracted

from the sample spectrum.


FTIR has three major advantages over dispersive IR spectroscopy. Since all wavelengths

being transmitted at once, many scans can be completed and averaged on an FTIR in a

shorter time than one scan on most dispersive instruments. In addition, it does not limit the

amount of light reaching the detector using a slit and it uses fewer mirrors than the

dispersive IR instrument, reducing the occurrence of reflecting losses. This means more

energy reaches the sample and hence the detector, which results in an improved signal-to-

noise ratio. Moreover, it does not require external calibration standards. The laser that

controls the velocity of the movable mirror provides reference for both precision and

accuracy of the spectrometer. Therefore, spectra can be compared with confidence, leading

to a higher precision [96], [97].

1.3.2.3 Operando Infrared Spectroscopy

Characterisation of the active sites within a catalyst is a key to its optimisation and rational

design. It is on these sites where reaction takes place going through the stages of reactant

adsorption, intermediate species formation and generated products desorption. In fact, the

formation of intermediate species is a key when crossing the energy barrier compared to a

non-catalysed reaction, as represented in Figure 1-12. Active centres are linked to

intermediates but can also be poisoned by other species, which affect the overall catalytic

performance. It is important to consider that they are not static entities, but they undergo

modifications depending on the reaction conditions and surface restructuration

phenomena. This is why, the study and characterization of these active centres should be,

as well, under actual reaction conditions.

Figure 1-12: Energy profile of a non-catalysed reaction (---) and a catalysed reaction (---)

[98]



The term operando is borrowed from Latin and means "working" or "operating". Operando

IR spectroscopy allows studying a catalyst under flow of reactants using real operating

conditions very close to the application (industrial conditions). Therefore, this method is

able to provide significant insights on the relevant catalytic steps, active sites and reaction

intermediates during a given process, in order to elucidate the reaction mechanism on a

given catalyst. It can even serve to differentiate an active site from a bare adsorption centre.

Note that IR is a very sensitive technique. Thus, as it serves to identify important reaction

surface species it can also detect surface impurities such as water, carbonates and other

residual species after synthesis. Presence of these species may inhibit probe adsorption,

or simply disturb the correct spectral interpretation by band overlapping. Therefore, the

necessity of their elimination (catalyst pre-treatment). Thermal stability of these species

strongly depends on acid–base properties of the material: on relatively acidic compounds,

like ceria or zirconia, they can be removed at mild temperatures. As a matter of fact, it is

important that the pre-treatment temperature never exceeds the calcination temperature of

the catalyst. If that would be the case, it would lead to surface reconstruction and sintering

phenomena [99], [100].

1.3.2.4 Oxygen activation

The interaction of molecular oxygen with transition-metal surfaces has been largely studied

due to its importance as a fundamental step in many catalytic reactions. Despite extensive

research, the elementary steps of oxygen chemisorption on transition metals remain difficult

to characterize accurately, although many authors agree on a chemisorbed-precursor

mechanism that involves an initial physically and/or chemically intact adsorbed species

acting as a precursor to dissociative chemisorption (Figure 1-13) [101]-[104].


Figure 1-13: A one-dimesional potential energy schematic showing qualitative aspects of

oxygen adsorption [101]

On many transition metal surfaces, including Pt, Pd, Ni, Ag and Cu, two chemisorbed

molecular states of oxygen are frequently observed in vibrational spectroscopy: a peroxo-

like form (O22-), with an O-O stretching frequency of 610-650 cm-1, and a superoxo-like form

(O2-), with a stretching frequency of 810-870 cm-1 [105]. The O-O bond would then stretch

until it breaks, to form atomic adsorbed oxygen species.

1.3.2.5 Ethylene activation and oxidation

The interaction of olefins, mainly ethylene, with catalyst surfaces has been the object of

many investigations by vibrational techniques. Although there is no specific literature on

ethylene adsorption onto ceria-based catalysts, there have been studies concerning the

adsorption of ethylene onto some noble and transition metals, metal oxides and some types

of zeolites.

Ethylene has different adsorption mechanisms depending on the noble or transition metal

used (Figure 1-14), which include the (a) π-bonded, (b) di-σ bonded, (c) ethylidyne, (d)



ethylidene, (e) vinylidene and (f) vinyl [106]. However, in the majority of cases it adsorbs

following the two first cases. The di-σ mode, features a strong interaction where two of the

ethylene carbon atoms attach to two metal surface atoms (the atomic orbitals around the C

atom hybridise as in ethane). The π-bonded mode easily explains the weakening of the

ethylene C=C bond following the Chatt-Dewar-Duncanson model (Figure 1-15). In this

model, π-electrons are donated to an empty orbital of a metal atom on the catalyst surface.

This is accompanied by back bonding from the filled d orbitals on the catalyst to the π*

orbitals of the alkene [107].

Figure 1-14: Potential adsorbed-ethylene species [106]

Figure 1-15: Chatt-Dewar-Duncanson model for ethylene adsorption [108]

Depending on the ethylene adsorption mechanism, it is common to find frequency bans in

the C=C or C-C bond stretching region on the corresponding IR spectra. The weaker

interaction of π -bonded ethylene compared with the di-σ mode implies that the π -bonded

case has a smaller energy cost. [109] For example, in the case of small gold clusters, the


π-bonded mode dominates over the di-σ mode. [110] However, its detection is not easy

since adsorbed ethylene is very unstable on metal surfaces. In fact, for its detection to be

possible, it is necessary a low-temperature measure (below room temperature) [111], [112].

Ethylene characteristic vibrational frequencies can be shifted when adsorbed onto a

specific material. The shift with respect to the original position of the bands is due to the

metal- π -bond and highly depends on its strength. In fact, these bands can be even masked

by the strong skeletal vibrations of the catalysts [113].

For example, two main characteristic frequencies of free gaseous ethylene are 1623 and

1342 cm-1, which correspond to C=C stretching and CH2 in-plane bending (scissoring),

respectively. One of the early studies of π -adsorbed ethylene species on Pt, Rh and Pd

based-catalysts states that these bands can be shifted down to 1495 and 1199 cm-1,

respectively, because of the gas-metal interaction [114]-[116]. As for the out-of-plane

deformation modes, such as CH2 wagging, these are generally shifted upwards in the

majority metal oxides. Some of the observed wavenumbers (cm-1) and assignments for

ethylene adsorbed on metal and metal oxides are shown in Table 1-1.

Table 1-1: Observed wavenumbers (cm-1) and assignments for ethylene adsorbed on

metal oxides [113]-[116]

ν1 (sC=C) ν2 (δCH2) ν7+ ν8 (2ωCH2)

Free 1625 1343 1888

Pt 1495 1199 ---

Rh 1499 1214 ---

Pd 1510 1241 ---

SiO2 1618 1340 1922

HY 1612 1338 1940

MgO --- 1338 1898

CaO --- 1342 1898

NiO --- 1338 1908

Fe2O3 1618 1340 1960

ZnO 1599 1340 1960

TiO2 1612 1341 1960

ZrO2 1610 1341 1975

In addition to the strong shift upwards of the out-of-plane deformation modes and the shift

downwards of the C=C stretching frequencies as well as of all CH stretching, the



perturbation caused by the surface-gas interaction can cause the activation on IR of the

modes that are Raman active in the CH stretching region.

Nevertheless, as said before, ethylene is very unstable when on a catalyst surface. On the

pathway in oxidation reactions, it is not easy to distinguish adsorbed ethylene as

intermediary species. The observation of carboxylate species that end up in the formation

of carbon dioxide, is more common. These species are mainly formates, acetates and

carbonates.

There are three main types of conformations of carboxylate groups that can be adsorbed

on the catalyst surface. Single-bonded species are usually denominated as ‘monodentate’

while double-bonded are called ‘bidentate’. Bidentate species can either be chelating, when

both oxygen atoms bind to one surface site, or bridged if the oxygen atoms bind to different

sites. When all three oxygen atoms are bound to the surface, the conformation is called

polydentate, or bulk, as it can be the case for carbonate species. The main carboxylate

species formed during the production of CO2 as well as their assigned IR bands are shown

in Figure 1-16 and Figure 1-17. Usually, bridged bidentate species are the most stable

conformations.

Figure 1-16: Experimental IR frequency bands of different types of formate species over metal oxides [comb1 stands for νas(COO) + δ(CH) and comb2 stands for νs(COO) + δ(CH)]

[117]


Figure 1-17: Experimental IR frequency bands of different types of surface carbonate

species over metal oxides [118]

1.4 Scope

Total catalytic oxidation of ethylene is proposed as an alternative for its elimination within

the storage systems of horticultural products. However, the presence of an active catalyst

at low temperatures is a must for the development of the process at industrial scale. For

this purpose, ceria-based mixed oxides with gold as active phase are proposed as active

materials.

This work aims to contribute with the characterisation of ceria and ceria-zirconia-based

catalysts as active materials for the total oxidation of ethylene, and the elucidation of the

fluid-surface interactions that take place during the adsorption of the reactants in order to

propose a possible reaction pathway.

2. Cerium oxides as oxidation catalysts for ethylene elimination

Ceria-based catalysts were tested on the reaction of total oxidation of ethylene. Two kinds

of ceria supports were used, one was synthesised at the laboratory following a pseudo sol-

gel method, the other one was provided by Solvay-Rhodia. Gold was then deposited in both

supports at different compositions: 1% for synthesised support and 2% for the commercial

one. The catalysts composed with commercial supports will be referred in the text with this

symbol: *.

2.1 Catalyst synthesis

The different catalysts were synthesised according to the pseudo sol-gel method described

by Vargas et al. [119]. The method is based on the thermal decomposition of propionate

precursors. For the case of cerium oxide, the precursor salt was cerium (III) acetate hydrate.

The specifications of the precursor used are shown in Table 2-1. The necessary amount of

the organometallic salt was dissolved in propionic acid, which was used as modifier agent

with a cation concentration of 0.12 mol·L-1. The solution was kept under total reflux at 100

°C for 1 h. This procedure ensured the effective formation of the polymeric precursor in

solution. Then, the solvent was removed by controlled evaporation at atmospheric pressure

and an approximate temperature of 116-120 °C to obtain a yellow resin. The solvent

evaporation took around 3 h. Finally, the resin was calcined in air with a temperature ramp

of 2 °C·min-1 until reaching 500 °C and maintained 6 h at this temperature.

Table 2-1: Specifications of the precursor used for the Ce-based catalyst preparation

Reactive Supplier CAS Formula LOT

Cerium(III) acetate

sesquihydrate

Alfa Aesar 537-00-8 Ce(C2H3O2)3•1.5H2O E03W017



After obtaining the support, gold was subsequently deposited on the oxide by the method

of direct ion exchange (DAE) between the support and a solution of chloroauric acid

(HAuCl4) with a concentration of 10-4 mol·L-1. The quantity of the solution depended on the

concentration in weight of gold to be deposited. The solution was kept at 70 °C for 1.5 h.

Then, the solid was separated by filtration and washed several times with ammonia and de-

ionised water to remove chlorine, which improves acidity and prevents sintering, as

reported by Ivanova et al. [120]. Finally, the catalyst was calcined in air with a temperature

ramp of 5 °C·min-1 until reaching 300 °C and maintained 4 h at this temperature.

2.2 Characterisation

The ceria-based oxides were characterised by several techniques. The textural

characteristics were determined by N2 adsorption at 760 mmHg and -196 °C. The structural

properties were characterised using X-ray Diffraction (XRD) and Raman spectroscopy,

which were also useful to verify the effective formation of the fluorite-type structure.

2.2.1 Surface area – N2 isothermal adsorption-desorption

The surface area and the pore volume of the mixed oxides catalysts were determined by

N2 adsorption at -196 °C using a Micromeritics ASAP 2420 SurfaceArea and Porosity

Analyser. The samples were previously degasified during 1 night by heating up to 200 °C

in order to eliminate humidity. Total surface area was derived from the BET model, the total

pore volume (Vpore) was measured at P/P0 = 0.95 as a single point and the pore diameter

(Dpore) was calculated according to the formula: Dpore = 4Vpore/SBET, where SBET is the

BET surface area. The textural properties of CeO2, Au(1%)/CeO2, CeO2* and

Au(2%)/CeO2* are shown in Table 2-2.

Table 2-2: Textural characteristics of samples with synthesised and commercial (*) ceria

supports

SBET (m²·g-1) Vpore at P/P° = 0.95 (cm3·g-1) Dpore (nm)

Ceria 79.2 0.19 9.5

Au(1%)/Ceria 78.2 0.19 9.9

Ceria* 199.8 0.14 2.8

Au(2%)/Ceria* 203.1 0.15 2.9

Cerium oxides as oxidation catalysts for ethylene elimination 35

The introduction of gold as active phase does not have a significant effect on the textural

properties of the catalysts, this can be remarked on the fact that the variations on the BET

surface area when comparing the sample with or without gold, fall within the range of the

experimental error of the equipment used. However, it is clear that the ceria commercial

support has a smaller pore diameter than the synthesised one and, consequently, a higher

surface area. The differences in this property can be ascribed to the fact that the synthesis

method for both supports was different, taking into account that the so-called synthesised

supports were obtained via pseudo sol-gel method and the commercial ones were provided

by Rhodia. More specifically, it is very likely that calcination temperatures for both kind of

supports were different. It has been reported that, in general, an increase in the calcination

temperature of a catalyst leads to a decrease in the surface area [121].

2.2.2 Crystalline structure – XRD

The crystalline structure of the mixed oxides catalysts was determined by X-ray diffraction

(XRD) in a Brücker AXS-D8 Advanced equipment, furnished with a LYNXEYE detector with

a nickel filter in order to filtrate Cu-Kα radiation (λ= 1.5404 Å). The 2θ range scatter used

was 10º to 80º with a 0.05º step size at a scan rate of 5º·min-1. The crystallite size was

determined using the Debye-Scherrer equation (eq. 2.1).

𝐷 =𝐾∙𝜆

𝛽 ∙cos𝜃 (2.1)

Where K corresponds to the Scherrer constant (K= 0.9); λ corresponds to the wavelength

of the X-ray used (λ= 1.5404 Å); β is the line broadening at half the maximum intensity

(FWHM), after subtracting the instrumental line broadening, in radians, and θ is the Bragg

reflection angle in radians.

The cubic lattice parameter of the cubic phase-centred fluorite was calculated from the four

most intense diffraction peaks (111), (200), (220) and (311) according to equation 2.2,

considering a FCC system where h, k and l are the Miller indexes and d is the interplanar

distance.

𝑎 = 𝑑ℎ𝑘𝑙√ℎ2 + 𝑘2 + 𝑙2 (2.2)



The effective formation of the fluorite structure was verified. The XRD patterns for the ceria-

based catalysts are presented in Figure 2-1.

Figure 2-1: XRD patterns of ceria samples with synthesised and commercial (*) supports

For all the samples, diffraction peaks at 28.6°, 33.1°, 47.5°, 56.5°, 59.2° and 69.5° were

observed, which relate to the planes (111), (200), (220), (311), (222) and (400),

respectively. These peaks correspond to a fluorite cubic structure CeO2 (JCPDS 03-065-

5923). No diffraction peaks associated to gold as the active phase were detectable. They

would have corresponded to reflections at 38.5 and 44.8°, relating to (111) and (200) planes

of metallic gold (JCPDS 04-0784). The fact that Au0 is not detected in the samples suggests

that the metallic particle size obtained with the deposition technique used may be smaller

than the dimension required for detection (ca. 5 nm) [53]. The peaks of the commercial

ceria-based supports were broader than the synthesised ones, indicating a lower crystallite

size. Lower crystallite size tends to have a higher active metal dispersion which may lead

to better catalyst activity [122]-[124].


Results for the cubic lattice parameter of the cubic phase-centred fluorite and the average

crystallite size are reported in Table 2-3. Since no significant differences on XRD patterns

from samples with and without gold were observed, it was suggested that the presence of

gold does not induce any modification of the support phase. However, the values for the

lattice parameter of samples with gold are somewhat higher than the ones observed for

ceria bare support. Although the content of gold is very low, a plausible explanation to this

could be due to the fact that deposited metallic gold (Au0) forms an interface with the ceria

support. In this thin interface, the existence of Au3+ is likely, due to the presence of oxygen

atoms on the ceria lattice. These Au3+ ions may partially attract oxygen atoms to the

surface, leaving some vacancies in the oxide bulk and resulting in a reduction of cerium

atoms [125]. Remember that the reduction process results in an increase on the ionic size

of cerium from Ce4+ (0.87 Å) to Ce3+ (1.01 Å). This explains the fact that patterns of the

samples with gold are slightly shifted to lower Bragg angles. It is also noticeable that the

crystallite size of commercial samples is smaller than for synthesised ones.

Table 2-3: Structural characteristics of CeO2, Au(1%)/CeO2, CeO2* and Au(2%)/CeO2*

oxides.

Catalyst Cubic lattice “a” (Å) Crystallite size (nm)

CeO2 5.39 9.0

Au(1%)/CeO2 5.40 9.0

CeO2* 5.42 4.7

Au(2%)/CeO2* 5.43 4.7

The above behaviour was already seen by Fu et al. [126], although they prepared gold-

ceria samples by a different synthesis method (deposition precipitation and coprecipitation),

they also noticed and increase in the lattice constant when gold was present. Also, they did

not observe any peaks corresponding to metallic gold for samples with gold content less

than 3%.

Results from XRD analysis allows to corroborate the effectiveness of the synthesis method

in the formation of the solid cubic phase of ceria.



2.2.3 Solid phase formation – Raman spectrocopy

The solid phase formation was studied by Raman spectroscopy. The Raman spectra were

recorded in a dispersive Horiva Jobin Yvon LabRam HR800 Microscope, with a He-Ne

green laser (532.14 nm) working at 5 mW, and with grating of 600 g.mm-1. The microscope

used a 50x objective and a confocal pinhole of 1000 μm.

The Raman spectra of CeO2 and CeO2* are presented in Figure 2-2. Raman spectra from

gold-ceria samples are not shown a very poor signal to noise (S/N) ratio. This could be

possible due to a high fluorescence interference from gold. Even weak fluorescence from

the sample is often much stronger than the Raman scattering, resulting in a large

background. Fluorescence can be reduced by selecting lasers in the UV or NIR region

[127], [128].

Figure 2-2: Raman spectra for synthesised and commercial (*) ceria

Bands at 260, 470 and around 600 cm-1 were noticed. The main band at 470 cm-1

correspond to a first order F2g Raman active mode of fluorite type lattice, which originates

from Ce–8O stretching vibrations. This mode in CeO2 shifts and broadens with decreased

particle size, as seen for the commercial support. The band at 260 cm-1 is present only in

ceria lattice with high defect concentration and it usually indicates the presence oxygen


containing species on the surface. A weak and broad peak at about 600 cm-1 can be

assigned to the intrinsic oxygen vacancies [122], [129], [130].

The results obtained from the Raman spectra are found to be consistent with the results

obtained from the XRD patterns of CeO2 samples. The characterisation analysis allowed to

verify the effective formation of the fluorite-type structure of the ceria samples, leading to

the conclusion that the pseudo sol-gel synthesis method was effective.

2.3 Evaluation of the fluid-surface interactions

Evaluation of the catalyst was divided in two stages. First stage comprehended the study

of the adsorption-desorption mechanisms of C2H4 and O2 separately on the catalyst

surface, at room temperature. For this purpose, a TPD system was proposed, in order to

quantify the amount of gas adsorbed. Second stage was the identification of molecular

species that adsorbed on the surface during the reaction. This was done via operando IR

spectroscopy.

2.3.1 Study of the adsorption-desorption mechanisms of probe molecules C2H4 and O2 via TPD

The TPD system was composed of the gas flow set-up, where activation and reaction flows

could be prepared independently, reactor, furnace, and detector (Figure 2-3). In this case,

the detector was a Hiden Analytical QGA (quantitative gas analysis) mass spectrometer

system configured for continuous analysis of gases and vapours at pressures near

atmosphere in standard form.

Note that instead of having a 6-way valve joining the reactor, as in the majority of setups,

two 4-way valves were installed. This allowed to independently investigate either the carrier

flow of the analyte flow via MS, via bypass or reactor. Four configurations were possible,

as shown in Figure 2-4, (a) analysis of analyte gas when flowing via bypass, (b) analysis of

carrier gas when flowing through the reactor, (c) analysis of analyte gas when flowing

through the reactor, (d) analysis of carrier gas when flowing via bypass. This configuration

helped to create baselines in an easier way and to detect possible leakages in the system

during the reaction.



Figure 2-3: Scheme of the TPD setup used

Figure 2-4: Possible configurations of the valve system in the TPD setup

The procedure used was: during the activation, the sample was heated up to 300 °C with a

flow of 50 cc·min-1 of dry air and a ramp of 5 °C·min-1. Sample was kept at 300 °C for 1 h,

30 min under dry air and 30 min under argon. Then, the sample was cooled down under

argon, while the analyte gas (either pure O2 or ethylene with a concentration of 1000 ppm

in N2) was being sent via bypass. Once the reactor temperature reached 25 °C, the valve

was switched to send the analyte gas through the reactor with a flow of 25 cc·min-1 to


perform the adsorption at room temperature. Once the analyte gas m/Z signal was stable,

meaning that adsorption sites were full or that adsorption did not take place at all, the carrier

gas was sent through the sample (either dry air or argon). After stabilisation of the signal,

the temperature was raised progressively up to 300 °C (8 °C·min-1) for analyte desorption.

The quantity of catalyst loaded was 50 mg.

Contrary to what was expected, according to the results by Ahn et. al. using a ceria-based

catalyst [44], no adsorption of ethylene or oxygen took place on ceria-based catalysts (with

or without gold) at room temperature (Figure 2-5).

Figure 2-5: Evolution of mass spectra signal intensities with time on stream during (a) ethylene desorption (b) oxygen desorption over Au(2%)/CeO2 sample



This could either mean that ethylene does not effectively adsorb on this kind of surfaces at

room temperature, or that its partial pressure was not high enough for the adsorption to

take place, taking into account that the concentration used was 1000 ppm. Remember that

according to Langmuir model [131], the dynamic process of adsorption can be considered

as (eq. 2.3):

𝐴(𝑔) + 𝑙 𝑘𝑎𝑑𝑠↔ 𝐴 − 𝑙 (2.3)

Where A(g) is the adsorbent, 𝑙 is a free surface site, kads is the velocity constant for the

adsorption process and kdes is the constant for the desorption process. According to this,

the global constant would be K= kads/kdes. If A(g) has a low partial pressure, then there is a

very low chance that the surface is covered by adsorbed species, since their permanence

highly depend on the reactant pressure.

2.3.2 Study of the intermediary species during ethylene total oxidation via operando IR spectroscopy

The operando system used for this study is provided in Figure 2-6; it is composed of four

main parts: the infrared spectrometer, the IR reactor-cell, the gas flow set-up and the

exhaust gas analysers. The reactor-cell is connected to the gas-system including mass flow

controllers for the introduction of gases into the lines. The two gas mixtures, so called the

activation and the reaction flow, can be prepared and sent independently to the reactor cell.

The system allows investigating the exhaust gases (reactive and/or reaction products) by

a Quadrupole Mass Spectrometer (Pfeiffer Omnistar GSD 301), while complementary

information on the gas phase can be gained by FTIR spectroscopy through a gas micro-

cell of 0.088 cm3 volume. The IR operando system is equipped with one single IR source,

thus, if complementary information on the gas phase is required, the IR beam is

alternatively sent to the catalyst or to the gas phase analysis through a mobile mirror [132].

Regarding the IR data, 32 scans were collected per spectrum at a time resolution of 1

spectrum each two minutes with a Thermo Scientific Nicolet 6700 spectrometer, equipped

with a MCT detector. Spectra were processed by the Nicolet OMNIC software.

𝑘𝑑𝑒𝑠


Figure 2-6: Scheme of the operando system [132]

Figure 2-7: General and detailed views of the “sandwich” reactor-cell. 1=Spectrometer

base-plate, 2=IR cell support, 3=adjusting nut for air tightness, 4=gas outlet, 5=air cooling outlet, 6=gas inlet, 7=air cooling inlet, 8=external shell, 9=thermocouple location, 10=IR beam, 11=stainless steel ring, 12=Kalrez O-ring, 13=KBr windows, 14=wafer holder, 15=oven location [133]



For each experiment, the catalyst was pressed into a self-supported wafer with a diameter

of 16 mm and a mass of 20 mg approximately. The sample was first activated through an

oxidative procedure, rising the temperature to 287 °C with a ramp of 5 °C·min-1, and

maintaining at this temperature for 1 h. Then, the sample was cooled down to 25 °C in 40

min. During the cooling, the reaction mixture was prepared independently and sent via

bypass. Once the sample arrived 25 °C, the 6-way valve was switched to direct the reaction

flow through the cell. Stabilisation was reached after 15 min. The sample was afterwards

heated up to 287 °C with a ramp of 2 °C.min-1, to observe the reaction temperature. The

flow conditions were as follows: with a total flow rate of 25 cc·min-1, the activation flow was

composed of 20% of oxygen diluted in Argon and the reaction flow was composed by a

mixture 400 ppm C2H4 and of 20% oxygen diluted in Argon. The reaction was studied within

the temperature range from 25 to 287 °C.

Two ceria-based samples were tested. The first one, was a provided ceria support. The

second one, was the same support with 1% gold deposited as active phase. The exhaust

gas analysis showed that there is no production of CO2 when testing the support alone

(Figure 2-8). Common products from partial oxidation reactions were not detected either,

thus, oxidation does not take place. As for the IR data, the spectra before through the

reaction remained almost invariable, thus, no intermediary species were detected (Figure

2-9). A decrease in some peaks around 1000, 1380 and 1500-1600 cm-1 region is observed

but this could be due to a continuous subsequent desorption of trapped bulk carbonate

species. This kind of species could have been ‘trapped’ during the catalyst synthesis, more

specifically, during the evaporation of propionic acid producing bulk CO2 species inside the

catalyst matrix.


Figure 2-8: Evolution of temperature and mass spectra signal (m/z= 28 for ethylene, 18 for water, 44 for CO2) intensities with time on stream [total flow 25 cc·min-1: 400 ppm C2H4

+ 20% O2 in Ar] when the reaction is performed over CeO2

Figure 2-9: IR spectra of CeO2 before (a) and after (b) the reaction

On the contrary, the catalyst with gold exhibited activity on the total oxidation of ethylene,

but only at high temperatures, starting to convert ethylene to CO2 at around 136 °C, as

shown in Figure 2-10. Results for operando IR spectroscopy during the oxidation reaction

on Au(1%)/CeO2 catalyst, are reported in Figures 2-11 to 2-14.

800130018002300280033003800

Ab

sorb

ance

(a.

u.)

Wavenumber (cm-1)




+ 20% O2 in Ar] when the reaction is performed over Au(1%)/CeO2

The first remark that can be made from Figure 2-10 is that periodic phenomena were

observed during the reaction (small oscillations on the m/Z signals). This situation was

already observed by Bazin et al. [132], who stated that, since the oscillation period

correlated with the infrared beam scanning duration, they were clearly due to the periodic

infrared beam heating of catalyst surface, equivalent to a thermal effect of +3.5 °C. This is

due to gold nanoparticles sensitivity; as a result, a slight change in the temperature may

influence the reactivity.

The complete IR spectrum (800-4000 cm-1) of the catalyst surface before and after the

reaction is shown in Figure 2-11. Initial bands around 850, 1400-1350 and 1480-1420 cm-1

are detected. These bands can be attributed to the presence of bulk polydentate carbonate

species present within the catalyst framework. The existence of carbonate species before

the reaction can be due to the fact that CO2 could be trapped in the catalyst framework

during the synthesis or calcination procedures. The broad band around 3600 cm-1 is due to

the presence of surface hydroxyl species, which is common in the most of oxide catalysts

[135].


Figure 2-11: IR spectra of Au(1%)/CeO2 before (a) and after (b) the reaction

Figures 2-12 to 2-14 show the stacked IR spectra of the catalyst surface during the

operando experience. Looking into close detail to the region between 800-1150 cm-1

(Figure 2-12), the formation of weak bands around 1105 and 890 cm-1 can be observed.

These bands can be related to the adsorption of dioxygen species onto the surface. They

are associated to the formation of superoxide and peroxide species, respectively. These

bands only appeared at temperatures above 80 °C, which indicates the absence of surface

oxygen species at room temperature. It is clear how these bands attenuated and

disappeared as the sample was progressively heated. This could be due to dissociation of

oxygen molecular species into atomic species (O, O-) and later migration into the ceria

lattice [136]. This confirms what was already seen in Section 2.3.1, in which there was no

evidence of oxygen adsorption at room temperature. It is common within reactions

performed with mixed oxides catalyst that lattice oxygen is consumed for the formation of

products while this is constantly regenerated by molecular oxygen coming from the

reactants. The rise of a strong band at 1070 cm-1 which can be associated to the formation

of surface carbonate species, is also observed.

Higher bands are seen at 2933, 2842, 1545, 1379, 1356 cm-1 (Figure 2-13 and Figure 2-

14). These can be readily attributed to surface formate species [137]. These bands grew

significantly when the sample was heated on the presence of C2H4 and O2, reaching their

maximum at around 194 °C, and then followed by a steeped decrement. It is possible that

these surface formate species are partly oxidised to surface carbonate species at

temperatures above 200 °C.

800130018002300280033003800

Ab

sorb

ance

(a.

u.)

Wavenumber (cm-1)




reaction; (b) 0 min, 25 °C; (c) 20 min, 35 °C; (d) 40 min, 80 °C; (e) 1 h, 126 °C; (f) 1 h 25 min, 171 °C; (g) 1 h 45 min, 214 °C; (h) 2 h 10 min, 259 °C; (i) 2 h 30 min, 280 °C; [total flow 25 cc·min-1: 400 ppm C2H4 + 20% O2 in Ar]

Finally, the bands at 1430 and 1570 cm-1 correspond to the symmetric and asymmetric

stretching of the CO2- bridging mode, respectively, which can be associated to the formation

of bidentate carbonate species. The ‘negative’ peaks around 1140 and 1180 cm-1 are

related to the consumption of surface OH groups [138].



reaction; (b) 0 min, 25 °C; (c) 20 min, 35 °C; (d) 40 min, 80 °C; (e) 1 h, 126 °C; (f) 1 h 25 min, 171 °C; (g) 1 h 45 min, 214 °C; (h) 2 h 10 min, 259 °C; (i) 2 h 30 min, 280 °C; [total flow 25 cc·min-1: 400 ppm C2H4 + 20% O2 in Ar]


reaction; (e) 1 h, 126 °C; (f) 1 h 25 min, 171 °C; (g) 1 h 45 min, 214 °C; (h) 2 h 10 min, 259 °C; (i) 2 h 30 min, 280 °C; [total flow 25 cc·min-1: 400 ppm C2H4 + 20% O2 in Ar]



Taking into account the above considerations, a basic reaction pathway according to the

IR results can be proposed. Although it is not clear how or where ethylene adsorption takes

place, it is clear that it breaks to form two kinds of surface species: formates and

carbonates. These two intermediary species are the precursor for the subsequent formation

of CO2. Lattice oxygen is being constantly regenerated by molecular oxygen that comes

from the gas stream. Water is easily formed due to the initial presence of hydroxyl species

on the catalyst surface. These hydroxyl species are constantly being regenerated, as well,

by the dehydrogenation of ethylene.

If a detailed description with elementary steps is desired, two types of mechanisms are

possible. The first one is a non-competitive mechanism where ethylene and oxygen adsorb

on different active sites. The active sites are represented by 𝑙1 and 𝑙2, where 𝑙1=ceria and

𝑙1=Au, or vice versa.

1) 3[𝑂2 + 2𝑙1 → 2𝑂 − 𝑙1]

2) 𝐶2𝐻4 + 6𝑙2 → 2𝐶 − 𝑙2 + 4𝐻 − 𝑙2

3) 2𝐶 − 𝑙2 + 4𝑂 − 𝑙1 → 2𝐶𝑂2 + 4𝑙1 + 2𝑙2

4) 4𝐻 − 𝑙2 + 2𝑂 − 𝑙1 → 2𝐻2𝑂 + 4𝑙2 + 2𝑙1

The second one, is a competitive mechanism where both ethylene and oxygen adsorb onto

the same active site.

1) 3[𝑂2 + 2𝑙1 → 2𝑂 − 𝑙1]

2) 𝐶2𝐻4 + 6𝑙1 → 2𝐶 − 𝑙1 + 4𝐻 − 𝑙1

3) 2𝐶 − 𝑙1 + 4𝑂 − 𝑙1 → 2𝐶𝑂2 + 6𝑙1

4) 4𝐻 − 𝑙1 + 2𝑂 − 𝑙1 + 6𝑙2 → 4𝐻 − 𝑙2 + 2𝑂 − 𝑙2 + 6𝑙1

5) 4𝐻 − 𝑙2 + 2𝑂 − 𝑙2 → 2𝐻2𝑂 + 6𝑙2

In both competitive and non-competitive cases, it is possible that the adsorbed species

migrate to the ceria surface by spillover phenomena.

The most feasible detailed reaction pathway is represented in Figures 2-15 and 2-16.

Ethylene is adsorbed and activated on the metallic gold surface. Dehydrogenation takes

place and the given species migrate to the ceria surface by spillover mechanisms. Oxygen

adsorption takes place directly on the ceria surface.



Au(1%)/CeO2 (part a)




Au(1%)/CeO2 (part b)

The proposed reaction mechanism can be described in detail by the following steps:

1) Ethylene adsorption on gold (π-bonded)

2) Ethylene dehydrogenation and π-bond weakening

3) Hydrogen migration to the ceria surface by spillover mechanisms to form water by

associating with surface Ce-OH species (consumption of OH groups), and ethylene

C=C bond cleavage to form two C-H adsorbed species

4) Oxygen rearrangement on the ceria lattice

5) C-H group migration to the ceria surface by spillover mechanisms and formation of

formate species

6) Partial oxidation of formates to produce adsorbed carbonate species

7) CO2 formation and desorption


8) Regeneration of ceria lattice oxygen by adsorption and dissociation of molecular

oxygen coming from the reactive atmosphere, following a Mars van Krevelen

mechanism

In conclusion, gold has a crucial role on the performance of the catalyst, since the support

alone does not show any activity on the reaction of ethylene oxidation. Moreover, it is likely

that the dissociative adsorption of oxygen takes place on the gold metallic sites and

regeneration of ceria lattice oxygens takes place by migration of adsorbed oxygens by

spillover mechanisms. Besides gold, three main active sites on the ceria surface get

involved during the oxidation reaction. These include surface hydroxyl species, surface

oxygen species and generated vacancies. The presence of these species is also vital for

the reaction to take place since hydroxyl species favour the later formation of water and

vacancies are necessary in order for the Mars van Krevelen oxidation-reduction cycle to

take place.

3. Cerium-zirconium oxides as oxidation catalysts for ethylene elimination

Ceria-zirconia mixed oxides were tested on the reaction of total oxidation of ethylene. Two

kinds of ceria-zirconia supports were used, one was synthesised at the laboratory following

a pseudo sol-gel method, the other one was provided by Solvay-Rhodia. The mass

CeO2/ZrO2 ratio was 0.8/0.2 for the synthesised support and 0.6/0.4 for the commercial

one. Gold was then deposited in both supports at different compositions: 1% for

synthesised support and 2% for the commercial one. The catalysts composed with

commercial supports will be referred in the text with this symbol: *.

3.1 Catalyst synthesis

Cerium-zirconium mixed oxide catalysts were synthesised following the same procedure

mentioned in section 2.1, according to the pseudo sol-gel like method described by Vargas

et al. [119]. In this case, the salts used were cerium (III) acetate hydrate and zirconium (IV)

acetylacetonate. The specifications of the precursors used are shown in Table 3-1. Note

that the amount of the organometallic salt of zirconium depends on the final molar

composition desired. The salts were dissolved separately in propionic acid, kept under total

reflux at 100 °C for 1 h, and subsequently mixed. The solvent was removed by controlled

evaporation at atmospheric pressure and an approximate temperature of 116-120 °C to

obtain a yellow resin, which was finally calcined at 500 °C. Procedure for gold deposition

was the same as described in section 2.1. The global procedure is represented in Figure

3-1.



Table 3-1: Specifications of the precursors used for the Ce-Zr-based catalyst preparation

Reactive Supplier CAS Formula LOT

Cerium(III) acetate

sesquihydrate

Alfa Aesar 537-00-8 Ce(C2H3O2)3•1.5H2O E03W017

Zirconium(IV) 2,4-

pentanedionate

Alfa Aesar 17501-44-9 C20H28O8Zr G31Y032

Figure 3-1: Global Au/CexZr1-xO2 catalyst synthesis procedure

3.2 Characterisation

The ceria-zirconia mixed oxides were characterised by several techniques. The textural

characteristics were determined by N2 adsorption at 760 mmHg and -196 °C. The structural

properties were characterised using X-ray Diffraction (XRD) and Raman spectroscopy,

which were also useful to verify the effective formation of the fluorite-type structure.

For details concerning the characterisation analysis and calculations performed, please

refer to Section 2.2.

3.2.1 Surface area – N2 isothermal adsorption-desorption

The textural properties of CZ, Au(1%)/CZ, CZ* and Au(2%)/CZ* mixed oxides are shown in

Table 3-2.

Cerium-zirconium oxides as oxidation catalysts for ethylene elimination 57

Table 3-2: Textural characteristics of samples with synthesised and commercial (*) ceria-

zirconia supports

SBET (m²/g) Vpore at P/P° = 0.95 (cm3/g) Dpore (nm)

CZ 67.1 0.07 4.1

Au(1%)/CZ 62 0.06 4.1

CZ* 52.8 0.15 11.1

Au(2%)/CZ* 59.7 0.15 10.4

Contrary to what was seen for the ceria-based catalysts, the introduction of gold has a

remarkable effect on the BET surface area of the CZ catalysts, more specifically, a

reduction of the surface area for synthesised samples, and an increase for commercial

ones. A decrease on this property can be explained by the fact that the material was

subjected again to heat treatment after gold deposition. It is not clear why the opposite

happens with the commercial support, although it may be related to the fact that both

supports have different contents of gold. It seems then that for the latter case, the

introduction of gold has a favourable effect on the textural properties of the catalyst, even

after the heat treatment (calcination after gold deposition).

For the four samples it is clear that the surface area is low when compared to ceria-based

catalysts from Table 2-3. This behaviour was already observed by Gómez-García [139]

when synthesising the samples via co-precipitation method. There is no evident explanation

for this phenomena and it would be necessary to prepare more sample varying the Ce/Zr

ratio in order to make more general conclusions.

In this case, it seems to be that the introduction of zirconium into the fluorite matrix as

“doping agent” does not have a favourable effect on the textural properties of the catalyst,

since it induces a loss in the surface area.

3.2.2 Crystalline structure – XRD

The formation of the fluorite structure was verified by XRD. The XRD patterns for the ceria-

zirconia mixed oxides catalysts are presented in Figure 3-2.



Figure 3-2: XRD patterns of ceria-zirconia samples with synthesised and commercial (*)

supports. Composition (wt%) Ce/Zr for CZ=0.8/0.2, for CZ*=0.6/045.

For the CZ and Au(1%)/CZ samples, diffraction peaks at 28.8°, 33.5°, 48.0°, 57.0°, 59.9°

and 69.5° were observed, which relate to the planes (111), (200), (220), (311), (222) and

(400) of the FCC structure, respectively. Although in the case of CZ* and Au(2%)/CZ*

samples these peaks are shifted to higher Bragg angles, all samples exhibit XRD patterns

similar to cubic CeO2 (JCPDS 034-0394), indicating the integration of zirconium into the

CeO2 cubic lattice forming a solid solution. Also, no diffraction peaks associated to the

presence of tetragonal or monoclinic zirconia were observed.

No diffraction peaks associated to gold as the active phase were detectable. They would

have corresponded to reflections at 38.5 and 44.8°, relating to (111) and (200) planes of

metallic gold (JCPDS 04-0784). The fact that Au0 is not detected in the samples suggests

that the metallic particle size obtained with the deposition technique used may be smaller

than the dimension required for detection (ca. 5 nm) [53]. The peaks of the synthesised

ceria-zirconia based supports were broader than the commercial ones, indicating a lower

crystallite size.


Results for the cubic lattice parameter of the cubic phase-centred fluorite and the average

crystallite size are reported in Table 3-3. For comparison, the structure of all the catalysts

was assimilated to a cubic structure and the pseudo-cubic lattice parameter has been

calculated. No significant differences on XRD patterns and on lattice parameters from the

samples with and without gold were observed, which suggested that the introduction of gold

does not induce any modification of the support phase. From Figure 3-2 it can be seen that

the diffraction peaks were shifted systematically to higher Bragg angles with increasing

ZrO2 content. This observation was attributed to the shrinkage of the lattice due to the

replacement of Ce4+ (0.87 Å) with smaller cation radius Zr4+ (0.72 Å), and can be confirmed

when comparing the lattice parameters of both supports with different Zr content.

Table 3-3: Structural characteristics of CZ, Au(1%)/CZ, CZ* and Au(2%)/CZ mixed oxides.

Catalyst Cubic lattice “a” (Å) Crystallite size (nm)

Ce0.8Zr0.2O2 5.36 5.4

Au(1%)/Ce0.8Zr0.2O2 5.36 5.1

Ce0.5Zr0.5O2* 5.29 8.0

Au(2%)/ Ce0.5Zr0.5O2* 5.29 8.1

The above behaviour was already remarked by Araque [52] and Gómez-García [139] who

observed that the cubic lattice parameter diminished with an increase of the the zirconium

content in the sample.

Differences on the crystallite size of synthesised and commercial samples might be due to

the difference on the synthesis methods.

3.2.3 Solid phase formation – Raman spectroscopy

The presence of a tetragonal phase in the mixed oxides synthesised was studied by Raman

spectroscopy. The Raman Spectra of CZ and CZ* are presented in Figure 3-3.



Figure 3-3: Raman spectra for synthesised and commercial (*) ceria-zirconia (CZ)

Bands at 185, 303, 470 and 629 cm-1 were noticed. The main band at 470 cm-1 correspond

to a first order F2g Raman active mode of fluorite type lattice, which originates from Ce–8O

stretching vibrations. This mode in CeO2 shifts and broadens with decreased particle size,

as seen for the commercial support. The band at 303 cm-1 has been attributed to the

tetragonal substitution of oxygen atoms from the ideal fluorite lattice, while the band at 630

cm-1 has been attributed to the presence of oxygen vacancies. The high incorporation of Zr

atoms in the ceria network generated a tetragonal phase-like distortion, which is observed

in the intensity and location of these bands, although this distortion does not mean a

modification of the cubic solid phase.

The results obtained from the Raman spectra can be compared with those obtained by

Araque [52], who observed the same peaks for CZ samples with different mass ratios. They

are also found to be consistent with the results obtained from the XRD patterns of CZ

samples. The characterisation analysis allowed to verify the effective formation of the

fluorite-type structure of the ceria-zirconia samples, leading to the conclusion that the

pseudo sol-gel synthesis method was effective.


3.3 Evaluation of the fluid-surface interactions

The study of the adsorption-desorption mechanisms of C2H4 and O2 separately on the

catalyst surface was made via TPD. The identification of molecular species that adsorbed

on the surface during the reaction was done via operando IR spectroscopy.

For details concerning the TPD and operando IR systems, please refer to Section 2.3.

3.3.1 Study of the adsorption-desorption mechanisms of probe molecules C2H4 and O2 via TPD

Similar to what was seen for ceria-based catalyst, no adsorption of ethylene or oxygen took

place on ceria-zirconia based catalysts (with or without gold) at room temperature. As for

the ceria case, this could either mean that ethylene does not effectively adsorb on this kind

of surfaces at room temperature, or that its partial pressure was not high enough for the

adsorption to take place, taking into account that the concentration used is 1000 ppm.

Further tests can be made by increasing the partial pressure of ethylene using a higher

concentration in order to increase the adsorption constant.

3.3.2 Study of the intermediary species during ethylene total oxidation via operando IR spectroscopy

Two ceria-zirconia based samples were tested. The first one, was the commercial CZ*

support. The second one, was the same support with 2% gold deposited as active phase.

The exhaust gas analysis showed that there is no production of CO2 when testing the

support alone, thus, oxidation does not take place. On the contrary, the catalyst with gold

exhibited a good performance on the total oxidation of ethylene, but only at high

temperatures, starting to convert to CO2 at around 106 °C, as shown in Figure 3-4. Results

for operando IR spectroscopy during the reaction on Au(2%)/CZ* catalyst, are reported.




+ 20% O2 in Ar] when the reaction is performed over Au(2%)/CZ*

The complete IR spectrum (800-4000 cm-1) of the catalyst surface before the reaction is

shown in Figure 3-5. The initial spectrum has carbonate characteristic IR signatures: with

doublet bands at about 1540 and 1450 cm-1 (derived from the asymmetric stretching

vibration, ν3, of the unconstrained CO32– ion) and a singlet band at 880 cm–1 (the out-of-

plane bending vibration, ν2) [140]. The existence of carbonate species before the reaction

can be due to the fact that CO2 could be trapped in the catalyst framework during the

synthesis or calcination procedures. The band centred at 1720 cm-1 can be ascribed to a

bridged CO site, in which the carbonyl oxygen atom interacts with a cerium cation of the

support [141]. This can be due to adsorption of CO present in the air [142]. Two bands at

around 2920 and 2850 cm-1 are assigned to the asymmetrical and symmetrical stretching

vibrations of CH2 groups The broad band around 3600 cm-1 is due to the presence of

surface hydroxyl species [135].


Figure 3-5: IR spectrum of Au(2%)/CZ* before (a) and after (b) the reaction

Figure 3-6 show the stacked IR spectra of the catalyst surface during the operando

experience on the region 1200-1700 cm-1, where changes on the IR bands are more

evident. First thing to remark is the disappearing of the band around 1525 cm-1, which can

be associated to the desorption of carbonate species in the catalyst present due to the

synthesis method. Meanwhile, the formation of weak bands around 1546, 1370, 1358 cm-1

can be observed. These can be related to the formation of formate species and correspond

to the vibrational modes of νas(CO2-), δ(CH) and νsy(CO2

-), respectively [137]. Nevertheless,

the band at 1358 cm-1 can also be assigned to νsy(CO32-) of carbonate species, along with

the band near 1430 cm-1 and 1070 cm-1.

800130018002300280033003800

Ab

sorb

an

ce (

a.u

.)

Wavenumber (cm-1)



Figure 3-6: IR stacked spectra (1200-1700 cm-1) of Au(2%)/CZ* (a) just before the

reaction (b) 0 min, 25°C (c) 20 min, 35 °C (d) 40 min, 80 °C (e) 1 h, 123 °C (f) 1 h 25 min, 167 °C (g) 1 h 45 min, 211 °C (h) 2 h 5 min, 252 °C (i) 2 h 25 min, 280 °C [total flow 25 cc·min-1: 400 ppm C2H4 + 20% O2 in Ar]

Although formate and carbonate bands are difficult to distinguish individually due to

possible overlapping, one can conclude that the ceria-zirconia based catalyst has a similar

reaction mechanism to the ceria-based one, making a special emphasis on the formation

of intermediary formates and carbonates for the later transformation and desorption into

CO2, as previously shown in Section 2.3.2, Figures 2 15 and 2-16.

In conclusion, as seen for the ceria-based samples, gold has a crucial role on the

performance of the ceria-zirconia based catalysts, since the CZ support alone does not

show any activity on the reaction of ethylene oxidation. Ceria-zirconia based oxides also

follow a Mars van Krevelen oxidation-reduction mechanism during ethylene total oxidation

reaction.

4. Adsorbent (zeolite) + catalyst: a promissing alternative

Zeolites are nanoporous crystalline aluminosilicates that present a specific framework

structure with channels on its inside. These channels allow a variety of applications in many

fields such as catalysis and ion exchange. There are many types of zeolites, around 40 on

its mineral form and more than 150 synthetic ones, which vary depending on its framework

structure. The most important mineral type commercially are chabazite, faujasite and

mordenite, and the most popular synthetic ones are types A and X, synthetic mordenite and

their ion-exchanged varieties.

The structure of a zeolite is basically composed of silicon and aluminum atoms bonded via

oxygen. Given that an O atom has a charge of -2, while a Si atom has a charge of +4, a

pure free-defect SiO2 structure would have a neutral charge. However, an atom of Al has a

charge of +3, thus, aluminosilicate structures like zeolites have an unbalanced charge

distribution. This is compensated by positive ions, generally alkali metals Na+, K+, alkaline

metals, Ca2+, Ba2+, and protons H+, which adsorb to the surface. Zeolites are highly

hygroscopic, so besides metal cations they might also contain hydroxyl groups in their

pores or surface [143], [144].

The ‘sodalite cage’ is a common framework structure constructed of silicon, aluminium and

oxygen atoms. Figure 4-1 shows how zeolites are comprised from this sodalite cages to

form different structures of pores and rings. Each line represents a bridging oxygen while

the intersections locate the tetrahedral atoms [145].



Figure 4-1: Schematic of different zeolite structures [143]

Zeolites have been widely used as adsorbents in separation and purification processes in

the past decades due to their high cation-exchange ability as well as to the molecular sieve

properties [146], [147]. This is because their pores are uniform in size and they are in the

same size range as small molecules, so they can discriminate molecules on the basis of

size [144]. In addition to changes to the cationic structure, the Si/Al ratio can be varied.

Thus, zeolites may be tailored with widely different adsorptive properties, by the appropriate

choice of framework structure, cationic form, and silica-to-alumina ratio in order to achieve

the selectivity required for a given separation. As said before, they have a high affinity by

polar molecules, like water, but they can be hydrophobic when the Si/Al ratio is increased.

Hydrophobic zeolites are generally used in the removal of volatile organic compounds from

air [148].

As mentioned in chapter 1, a lot of C2H4-adsorbing substances such as potassium

permanganate (KMnO4), activated carbon and minerals (zeolites and clays) have been

characterised and used, though the search has been focused considerable attention on

zeolites [149].

Adsorbent (zeolite) + catalyst: a promissing alternative 67

Taking into account the fact that the catalyst proposed in this work is not active at room

temperature, it is proposed an enhanced strategy in which a regenerable adsorbent (a

zeolite) is used before the catalyst to ‘trap’ or adsorb ethylene at room temperature. Once

the zeolite is saturated with ethylene, it is heated under flow of hot gas for the ethylene to

desorb and the reaction to take place at the adequate temperature on the catalyst surface.

The zeolite used on the experiments was a Zeolite type A synthesised from a mining

residue, which characterisation is reported by Medina [150].

The procedure was the same as described at Section 2.3.1 with the only difference that 50

mg of zeolite were loaded before the catalyst bed.

All catalysts (ceria-based and ceria-zirconia based) were tested, but only the two with the

highest conversion upon the use of this system are shown, which are Au(1%)/CeO2 and

Au(2%)/CZ*. Zeolite A was also separately tested in order to quantify the amount of

ethylene that it can trap. TPD profile for zeolite A is reported in Figure 4-2. TPD profiles for

both catalysts are reported in Figure 4-3 and Figure 4-4. Two tests were performed for each

catalyst, the difference was the carrier gas under which desorption took place (air or argon).


with time on stream during ethylene desorption under air over Zeolite A



The TPD profile for zeolite A shows that the MS signal intensity for CO2 remains constant

during the whole process, meaning there is no formation of this production during the

heating and desorption process. This indicates that zeolite is not active for ethylene

oxidation, it only serves as ‘trap’ to adsorb it, as mentioned before.

On the opposite, the TPD profiles for both catalysts show that there is effective formation

of CO2 no matter the carrier gas utilised. Moreover, although the form of the peaks are

different when the carrier gas is different (and this is due to slight differences on the flow

conditions), the area is the same, thus, the quantity of desorbed unreacted ethylene and

desorbed CO2 is the same. This proves that the oxygen consumed during the oxidation

reaction is not the one that comes from the air but the one coming from the catalyst lattice.

These lattice oxygens have to be subsequently regenerated by the molecular oxygen that

comes from the air, either way, the reaction won’t keep taking place.

It can be seen from Figures 4-3 and 4-4 that, for both catalysts, CO2 signal is composed of

two peaks. This can be explained by the fact that CO2 formation can be either via two

intermediary species: formate or carbonate.

Table 4-1 shows a summary of the results of all TPD experiences done. It can be seen from

the results that conversions up to 58% can be achieved with this enhanced strategy.

Table 4-1: Summary of the results of the TPD experiences for zeolite A under air, zeolite

A +Au/CeO2 (under air and argon), and zeolite A +Au/CZ* (under air and argon)

mL adsorbed C2H4 /mg zeolite (x10-4)

mL desorbed C2H4 /mg zeolite (x10-4)

mL produced CO2 /mg cat (x10-3)

Ethylene conversion (%)

ZA (air) 11.5 ± 1.7 11.5 ± 1.7 ---- ----

ZA +Au/CeO2 (air) 9.03 ± 0.14 3.90 ± 0.58 1.03 ± 0.15 57%

ZA +Au/CeO2 (argon) 8.82 ± 0.13 2.87 ± 0.26 1.05 ± 0.10 57%

ZA +Au/CZ* (air) 12.3 ± 0.18 5.17 ± 0.47 1.43 ± 0.21 58%

ZA +Au/CZ* (argon) 12.4 ± 0.12 5.30 ± 0.50 1.37 ± 0.13 56%




A + Au(1%)/CeO2 system




with time on stream during ethylene desorption (a) under air (b) under argon over Zeolite A + Au(2%)/CZ* system


Ethylene conversion (X) was calculated by assuming that the initial moles of ethylene

(adsorbed) were equal to the sum of the desorbed moles of ethylene and the ones

consumed during the reaction. This was done due to the fact that it was not possible to

correctly calculate the peak of adsorbed ethylene with the technique, it was only possible

to quantify the desorption peak. Consumed moles of ethylene were calculated based on

stoichiometric calculations with respect to desorbed CO2. It is interesting to see that

although the quantity of adsorbed ethylene should be the same for all the systems (taking

into account that any of the catalyst adsorbs ethylene at room temperature), it appears to

be higher for systems with a zirconia. There is no evident explanation for this phenomena,

although it could be due to slight differences on the time during which the carrier gas was

flown through the system before heating.

According to the results, it appears to be that the enhanced adsorbent+catalyst strategy

proved to be effective for ethylene removal and oxidation in environments of low ethylene

concentration, although it does no achieve total conversion. Further recommendations to

enhance the strategy would include changing the type of zeolite used, to increase the

quantity of trapped ethylene, or increasing the carrier gas temperature very fast so that the

desorption of ethylene from the zeolite takes place at the same time as the catalyst

becomes active. This could help to avoid losses from ethylene that desorbs but does not

react, since the temperature is still not high enough on the catalyst surface. Adsorption

enhancement can be also achieved by the co-incorporation of Au and metal oxide into the

zeolite matrix [151].

5. Conclusions and recommendations

The so-called enhanced adsorbent+catalyst strategy in order to remove ethylene from

environments at low concentrations, such as the ones found in climacteric fruits storage,

proved to be effective for the purpose described. Depending on the catalyst used, the

system can achieve conversions of up to 58%. Other types of zeolites could be used to

enhance the quantity of ethylene that can be adsorbed. Further work should be made in

order to improve the technique and make it possible to quantify the adsorption peak of

ethylene.

The enhanced strategy was proposed due to the low interaction of the catalyst with ethylene

at room temperature. No effective adsorption of ethylene took place on the catalyst surface.

This could be due to the partial pressure of the ethylene in the flow stream, taking into

account the low concentration used (1000 ppm), which could not be enough for the

adsorption to take place, according to Langmuir mechanism.

Gold as active phase is fundamental for the activation of the ethylene and further oxidation

reaction. According to the IR spectra of the catalyst surface, formate and carbonate species

are primarily formed during the reaction. Both species are responsible for subsequent CO2

formation and desorption. No species associated to ethylene adsorption were detected.

This might be due to the low stability of adsorbed ethylene species, fast cleavage of the

double bond and transformation into formate and carbonate.

Ethylene oxidation follows a Mars van Krevelen mechanism, were oxygen in the products

(CO2 and H2O) comes from ceria structural oxygen. This is confirmed by the fact that the

reaction also takes place in the absence of atmospheric oxygen.



The presence of polydentate carbonate species in the catalyst before the reaction might be

due to CO2 trapped during the synthesis process. To avoid their presence, a heating

treatment is necessary before the use of the catalyst. It is possible that CO2 species get

constantly ‘trapped’ within the bulk of the catalyst when the ethylene oxidation takes places.

Therefore, a regeneration system (heating treatment after a specific time of use of the

catalyst) might be needed in order to avoid its accumulation.

The formation of the fluorite type structure was effectively verified for all types of oxides by

XRD and Raman spectroscopy. The insertion of gold as active phase does not have an

effect on the structural properties of the catalysts. The insertion of zirconia does not change

the original fluorite cubic-type structure of ceria although it has a clear effect on the lattice

parameter. The shrinkage of the lattice parameter is due to the differences in the Ce4+ and

Zr4+ ionic sizes. Tetragonal phase-like distortions may be seen for high contents of zirconia

(Ce/Zr>0.5).

Recommendations to improve ethylene adsorption on the ceria-based catalysts include the

increase of the ethylene concentration used. As for the adsorbent+catalyst strategy

proposed changing the type of zeolite used in order to increase the quantity of trapped

ethylene, would be an option. Adsorption enhancement can be also achieved by the co-

incorporation of Au and metal oxide into the zeolite matrix.

References

[1] Food and Agriculture Organization of the United Nations, “Global food losses and food

waster”, Rome, 2011.

[2] D. Valero and M. Serrano, Postharvest biology and technology for preserving fruit quality,

Boca Raton: CRC Press, 2010.

[3] S. Lurie and C. H. Crisosto, “Chilling injury in peach and nectarine,” Postharvest Biology

and Technology, vol. 37, no. 3, pp. 195-208, 2005.

[4] D. Zagory, “Ethylene-removing packaging”, in Active Food Packaging, New York,

Springer US, 1995, pp. 38-54.

[5] R. Mendoza, “Desarrollo de un catalizador para la oxidación del etileno a bajas

temperaturas como una estrategia de remoción de etileno de cámaras de

almacenamiento de productos hortofrutícolas”, Tesis de doctorado en ejecución,

Universidad Nacional de Colombia, Bogotá, 2015.

[6] Research and Markets, “The ethylene technology report 2016”, New York, 2016.

[7] OECD, “Screening information dataset (SIDS) ethylene”, IPCS Inchem, 1998.

[8] ICIS, “Ethylene production and manufacturing process”, 2007.

[9] National Center for Biotechnology Information, “PubChem compound database;

CID=6325”, 2017. [Online]. Available:

https://pubchem.ncbi.nlm.nih.gov/compound/Ethene#section=Top. [Accessed 2017].

[10] G. Rayner-Canham, Descriptive inorganic chemistry, New York: Pearson, 2000.



[11] M. E. Saltveit, “Effect of ethylene on quality of fresh fruits and vegetables”, Postharvest

Biology & Techology, vol. 15, no. 3, pp. 279-292, 1999.

[12] M. Vendrell, E. Domínguez-Puigjaner and I. Llop-Tous, “Climacteric versus non-

climacteric physiology”, IV International Conference on Postharvest Science, vol. 553, pp.

345-349, 2001.

[13] E. J. Corpas Iguarán and O. A. Tapasco Alzate, “Hallazgos de la biosíntesis del etileno

en frutas climatéricas y de los factores que afectan la ruta metabólica”, Alimentos Hoy,

vol. 22, no. 31, pp. 46-63, 2014.

[14] C. Chang and A. B. Bleecker, “Ethylene biology. More than a gas”, Plant Phisiology, vol.

136, no. 2, pp. 2895-2899, 2004.

[15] Comisión del Codex Alimentarius, “Justificación respecto al uso de etileno para la

maduración de fruta”, Organización Mundial de la Salud, Quebec, 2011.

[16] R. Wills, B. McGlasson, D. Graham and D. Joyce, Postharvest: an introduction to the

physiology & handling of fruit, vegetables & ornamentals, 5 ed., Wallingford: CAB

international, 1998, p. 29.

[17] J. R. Gorny and A. A. Kader, “Regulation of ethylene biosynthesis in climacteric apple

fruit by elevated CO2 and reduced O2 atmospheres”, Postharvest Biology and

Technology, vol. 9, no. 3, pp. 311-323, 1996.

[18] J. Brecht, K. Chau, S. Fonseca, F. Oliveira, F. Silva, M. Nunes and R. Bender,

“Maintaining optimal atmosphere conditions for fruits and vegetables throughout the

postharvest handling chain”, Postharvest Biology and Technology, vol. 27, no. 1, pp. 87-

101, 2003.

[19] J. D. Roberts and M. C. Caserio, Modern organic chemistry, New York: W.A. Benjamin,

Inc, 1967.

[20] W. Scott, E. Stephens, P. Hanst and R. Doerr, “Further developments in the chemistry of

the atmosphere”, Proceedings of the American Petroleum Institute, vol. 37, no. 111, pp.

171-183, 1957.

[21] A. Shorter and K. Scott, “Removal of ethylene from air and low oxygen atmospheres with

ultraviolet radiation”, Lebensmittel Wissenschaft und Technologie, vol. 19, no. 2, pp. 176-

179, 1986.

[22] M. Rooney, Active food packaging, New South Wales: Springer, 1995.

References 77

[23] M. de Chiara, S. Pal, A. Licciulli, M. Amodio and G. Colelli, “Photocatalytic degradation of

ethylene on mesoporous TiO2/SiO2 nanocomposites: Effects on the ripening of mature

green tomatoes”, Biosystems Engineering, vol. 132, pp. 61-70, 2015.

[24] A. Salcedo Echavarría, “Evaluación de absorbedores de etileno compuestos por mezclas

de permanganato de potasio y vermiculita como retardantes de madurez de frutos de

tomate (Solanum lycopersicum L.)”, Tesis de maestría, Universidad Nacional de

Colombia, Bogotá, 2014.

[25] R. N. Arteca, Introduction to horticultural science, Boston: Cengage Learning, 2006.

[26] Z. Talib, Ethylene in the storage of fresh produce. Developments in Food Preservation,

New York: Applied Science Publishers, 1983.

[27] R. T. Morrison and R. N. Boyd, Organic chemistry, New York: Prentice Hall, 1992.

[28] S. Kays and R. Beaudry, “Techniques for inducing ethylene effects”, Acta Horticulturae,

vol. XXII IHC 201, pp. 77-116, 1987.

[29] D. Martínez-Romero, G. Bailén, M. Serrano, F. Guillén, J. M. Valverde, P. Zapata, S.

Castillo and D. Valero, “Tools to maintain postharvest fruit and vegetable quality through

the inhibition of ethylene action: a review”, Critical Reviews in Food Science and

Nutrition, vol. 47, no. 6, pp. 543-560, 2007.

[30] R. Ahvenainen, Novel food packaging techniques, Cambridge: Woodhead Publishing

Series in Food Science, Technology and Nutrition, 2003.

[31] A. L. Brody, E. R. Strupinsky and L. R. Kline, Active packaging for food applications, New

York: CRC Press, 2001.

[32] F. B. Abeles, P. W. Morgan and M. E. Saltveit Jr., Ethylene in plant biology, Academic

Press, 1992.

[33] K. Abe and A. E. Watada, “Ethylene absorbent to maintain quality of lightly processed

fruits and vegetables”, Journal of Food Science, pp. 1589-1592, 1991.

[34] Q. Huai, Y. Xia, Y. Chen, B. Callahan, N. Li and H. Ke, “Crystal structures of 1-

aminocyclopropane-1-carboxylate (ACC) synthase in complex with

aminoethoxyvinylglycine and pyridoxal-5'-phosphate provide new insight into catalytic

mechanisms”, Journal of Biological Chemistry, vol. 276, no. 41, pp. 38210-38216, 2001.



[35] C. Watkins, “The use of 1-methylcyclopropene (1-MCP) on fruits and vegetables”,

Biotechnology Advances, vol. 24, no. 4, pp. 389-409, 2006.

[36] C. Espinosa Lagunas, “Inhibición de la síntesis y de la acción de etileno en

comportamiento en postcosecha de chirimoya (Annona cherimola Mill.) Var. Concha

Lisa.”, Universidad de Chile, Santiago de Chile, 2005.

[37] L. Elsgaard, “Ethylene removal at low temperatures under biofilter and batch conditions”,

Applied and Environmental Microbiology, vol. 66, no. 9, pp. 3878-3882, 2000.

[38] Ş. Bobaru, “Ethylene oxidation over Ag(111) and Pt(111)”, in High-pressure STM studies

of oxidation catalysis, Leiden, Doctoral thesis, Leiden University, 2006, pp. 113-122.

[39] D.-R. Park, J. Zhang, K. Ikeue, H. Yamashita and M. Anpo, “Photocatalytic oxidation of

ethylene to CO2 and H2O on ultrafine powdered TiO2 photocatalysts in the presence of

O2 and H2O”, Journal of Catalysis, vol. 185, no. 1, pp. 114-119, 1999.

[40] F. Diehl, J. Barbier Jr, D. Duprez, I. Guibard and G. Mabilon, “Catalytic oxidation of heavy

hydrocarbons over Pt/Al2O3. Influence of the structure of the molecule on its reactivity”,

Applied Catalysis B: Environmental, vol. 95, no. 3, pp. 217-227, 2010.

[41] E. C. Njagi, H. C. Genuino, C. K. King’ondu, S. Dharmarathna and S. L. Suib, “Catalytic

oxidation of ethylene at low temperatures using porous copper manganese oxides”,

Applied Catalysis A: General, vol. 421, pp. 154-160, 2012.

[42] M. Hussain, S. Bensaid, F. Geobaldo, G. Saracco and N. Russo, “Photocatalytic

degradation of ethylene emitted by fruits with TiO2 nanoparticles”, Industrial &

Engineering Chemistry Research, vol. 50, no. 5, p. 2536–2543, 2011.

[43] W. J. Xue, Y. F. Wang, P. Li, Z.-T. Liu, Z. P. Hao and C. Y. Ma, “Morphology effects of

Co3O4 on the catalytic activity of Au/Co3O4 catalysts for complete oxidation of trace

ethylene”, Catalysis Communications, vol. 12, no. 13, pp. 1265-1268, 2011.

[44] H.-G. Ahn, B.-M. Choi and L. Do-Jin, “Complete oxidation of ethylene over supported

gold nanoparticle catalysts”, Journal of Nanoscience and Nanotechnology, pp. 599-603,

2006.

[45] W. Chu, Y. Ooka, Y. Kamiya and T. Okuhara, “Reaction path for oxidation of ethylene to

acetic acid over Pd/WO3-ZrO2 in the presence of water”, Catalysis Letters, vol. 101, no.

3-4, pp. 225-228, 2005.

[46] D. L. Trimm, Design of industrial catalysts, Amsterdam: Elsevier, 1980.

References 79

[47] H. H. Kung, “Desirable catalyst properties in selective oxidation reactions”, Industrial &

Engineering Chemistry, Product Research and Development, vol. 25, no. 2, pp. 171-178,

1986.

[48] S. Bernal, J. J. Calvino, J. M. Gatica, C. López Cartes and J. M. Pintado, “Chemical and

Nanostructural Aspects of the Preparation and Characterisation of Ceria and Ceria-Based

Mixed Oxide-Supported Metal Catalysts”, in Catalysis by Ceria and Related Materials,

London, Imperial College Press, 2002, pp. 85-168.

[49] Z. Ma, S. H. Overbury and S. Dai, “Gold nanoparticles as chemical catalysts”, in

Encyclopedia of inorganic chemistry, New York, John Wiley & Sons, 2009.

[50] C. Mistry, R. Mewada, V. Srivastava and R. Jasra, “Characteristics of oxidation and

oxidative dehydrogenation catalysts for gas phase reactions: a review”, Nirma University,

Ahmedabad, 2011.

[51] F. Cavani and G. Centi, “Oxidation processes”, in Encyclopaedia of hydrocarbons, Rome,

ENI, 2005, pp. 617-635.

[52] M. C. Araque, “Valorización del glicerol por el reformado con vapor para la producción de

hidrógeno”, Tesis de doctorado, Universidad Nacional de Colombia, Bogotá, 2011.

[53] M. Vicario, J. Llorca, M. Boaro, C. de Leitenburg and A. Trovarelli, “Redox behavior of

gold supported on ceria and ceria-zirconia based catalysts”, Journal of Rare Earths, vol.

27, no. 2, pp. 196-203, 2009.

[54] T. Tabakova, L. Ilieva, I. Ivanov, R. Zanella and J. Sobczak, “Influence of the preparation

method and dopants nature on the WGS activity of gold catalysts supported on doped by

transition metals ceria”, Applied Catalysis B: Environmental, vol. 136, pp. 70-80, 2013.

[55] A. Trovarelli, “Structural Properties and Nonstoichiometric Behavior of CeO2”, in

Catalysis by Ceria and Related Materials, London, Imperial College Press, 2002, pp. 15-

50.

[56] A. Younis, D. Chu and S. Li, “Cerium oxide nanostructures and their applications”, in

Functionalized nanomaterials, Rijeka, InTech, 2016.

[57] A. Trovarelli, Catalysis by ceria and related materials, London: World Scientific, 2002.

[58] A. Pietrovito and P. Davies, “Ionic Crystal Structures”, University of Pennsylvania, 2007.

[Online]. Available: https://www.seas.upenn.edu/~chem101/sschem/ionicsolids.html.

[Accessed 26 Feb 2017].



[59] G. R. Rao and B. G. Mishra, “Structural, redox and catalytic chemistry of ceria based

materials”, Bulletin of the Catalysis Society of India 2, vol. 2, pp. 122-134, 2003.

[60] A. Trovarelli, C. de Leitenburg, M. Boaro and G. Dolcetti, “The utilization of ceria in

industrial catalysis”, Catalysis Today, vol. 50, no. 2, pp. 353-367, 1999.

[61] A. Kopp Alves, F. Amorim Berutti and F. A. Lucca Sánchez, “Nanomaterials and

catalysis”, in Nanostructured materials for engineering applications, Berlin, Springer,

2011.

[62] C. Zhang, X.-D. Wen, B.-T. Teng, Y. Zhao and M. Fan, “Catalytic effects of Zr doping ion

on ceria-based catalyst”, Fuel Processing Technology, vol. 131, pp. 1-6, 2015.

[63] M. Boaro, C. de Leitenburg, G. Dolcetti and A. Trovarelli, “The dynamics of oxygen

storage in ceria–zirconia model catalysts measured by CO oxidation under stationary and

cycling feedstream compositions”, Journal of Catalysis, vol. 193, no. 2, pp. 338-347,

2000.

[64] R. D. Shannon, “Revised effective ionic radii and systematic studies of interatomic

distances in halides and chalcogenides”, Acta Crystallographica, vol. 32, no. 5, pp. 751-

767, 1976.

[65] H. Sobukawa, “Development of ceria-zirconia solid solutions and future trends”, Toyota

Technical Journal R&D Review, vol. 37, no. 4, 2002.

[66] P. E. V. de Miranda, “Charge transfer in ceramic anode for solid oxide fuel cell”, Matéria

(Rio J.), vol. 21, no. 1, 2016.

[67] P. Fornasiero, G. Balducci, R. Di Monte, J. Kašpar, V. Sergo, G. Gubitosa, A. Ferrero and

M. Graziani, “Modification of the redox behaviour of CeO2 induced by structural doping

with ZrO2”, Journal of Catalysis, vol. 164, no. 1, pp. 173-183, 1996.

[68] S. Meriani, “Features of the ceria-zirconia system”, Materials Science and Engineering,

vol. A109, pp. 121-130, 1989.

[69] Y. Masamoto, K. Morimoto, N. Ishizawa and M. Yoshimura, “Zirconia–ceria solid solution

synthesis and the temperature–time–transformation diagram for the 1: 1 composition”,

Journal of the American Ceramic Society, vol. 76, no. 7, pp. 1745-1750, 1993.

[70] J. Kaspar and P. Fornasiero, “Structural Properties and Thermal Stability of Ceria-

Zirconia and Related Materials”, in Catalysis by Ceria and Related Materials, London,

Imperial College Press, 2002, pp. 217-242.

References 81

[71] G. Bond, “Periodic variations in the catalytic properties of metals”, Platinum Metals

Review, vol. 12, no. 3, pp. 100-105, 1968.

[72] M. Haruta, T. Kobayashi, H. Sano and N. Yamada, “Novel gold catalysts for the oxidation

of carbon monoxide at a temperature far below 0 °C”, Chemistry Letters, vol. 16, no. 2,

pp. 405-408, 1987.

[73] M. M. Schubert, S. Hackenberg, A. C. van Veen, M. Muhler and V. Plzak, “CO oxidation

over supported gold catalysts—“Inert” and “active” support materials and their role for the

oxygen supply during reaction”, Journal of Catalysis, vol. 197, no. 1, pp. 113-122, 2001.

[74] B. Acosta, E. Smolentseva, S. Beloshapkin, R. Rangel, S. Fuentes, A. Simakov and M.

Estrada, “Gold supported on ceria nanoparticles and nanotubes”, Applied Catalysis A:

General, vol. 449, pp. 96-104, 2012.

[75] M. Haruta, N. Yamada, T. Kobayashi and S. Iijima, “Gold catalysts prepared by

coprecipitation for low-temperature oxidation of hydrogen and of carbon monoxide”,

Journal of Catalysis, vol. 115, no. 2, pp. 301-309, 1989.

[76] J. C. Vargas, S. Ivanova, S. Thomas, A.-C. Roger and V. Pitchon, “Influence of gold on

Ce-Zr-Co fluorite-type mixed oxide catalysts for ethanol steam reforming”, Catalysts, vol.

2, no. 1, pp. 121-138, 2012.

[77] Z. Zhou, S. Kooi, M. Flytzani-Stephanopoulos and H. Saltsburg, “The role of the interface

in CO oxidation on Au/CeO2 multilayer nanotowers”, Advanced Functional Materials, vol.

18, no. 18, pp. 2801-2807, 2008.

[78] T. Tabakova, F. Boccuzzi, M. Manzoli and D. Andreeva, “FTIR study of low-temperature

water-gas shift reaction on gold/ceria catalyst”, Applied Catalysis A: General, vol. 252, no.

2, pp. 385-397, 2003.

[79] M. Fadoni and L. Lucarelli, “Temperature programmed desorption, reduction, oxidation

and flow chemisorption for the characterisation of heterogeneous catalysts. Theoretical

aspects, instrumentation and applications”, in Studies in surface science and catalysis,

Amsterdam, Elsevier, 1999, pp. 177-225.

[80] S. L. Schroeder and M. Gottfried , “Temperature-programmed desorption (TPD) Thermal

desorption spectroscopy (TDS)”, FU Berlin, Berlin, 2002.

[81] A. Auroux, Calorimetry and thermal methods in catalysis, Berlin: Springer, 2013.



[82] F. Gaillard, “Characterisation of Pt/ceria catalysts by one-pass TPD analysis”, Catalysis

Letters, vol. 95, no. 1-2, pp. 23-29, 2004.

[83] D. Martin and D. Duprez, “Mobility of Surface Species on Oxides. 1. Isotopic Exchange of

18O2 with 16O of SiO2, Al2O3, ZrO2, MgO, CeO2, and CeO2-Al2O3. Activation by

Noble Metals. Correlation with Oxide Basicity”, The Journal of Physical Chemistry, vol.

100, no. 2, p. 9429–9438, 1996.

[84] D. Ball, Field guide to spectroscopy, Bellingham: SPIE Publication, 2006.

[85] B. Stuart, Modern infrared spectroscopy, New York: Wiley, 1996.

[86] B. Stuart, Infrared spectroscopy: fundamentals and applications, New York: Wiley, 2004.

[87] F. Daniels, J. Williams, P. Bender, R. Alberty, C. Cornwell and J. Harriman, Experimental

physical chemistry, 7 ed., New York: McGraw-Hill, 1970.

[88] D. Skoog and J. Leary, Principles of instrumental analysis, 4 ed., Philadelphia: Harcourt

Brace Jovanovich, 1992.

[89] T. Theophanides, “Introduction to infrared spectroscopy”, in Infrared spectroscopy -

materials science, engineering and technology, Rijeka, InTech, 2012.

[90] W. D. Perkins, “Fourier transform-infrared spectroscopy: Part l. Instrumentation”, Journal

of Chemical Education, vol. 63, no. 1, pp. A5-A10, 1986.

[91] A. L. Smith, Applied infrared spectroscopy: fundamentals, techniques, and analytical

problem-solving, New York: Wiley, 1979.

[92] B. C. Smith, Fundamentals of fourier transform infrared spectroscopy, Boca Raton: CRC

Press, 1996.

[93] P. R. Griffiths, Chemical infrared fourier transform spectroscopy, New York: Wiley, 1975.

[94] P. R. Griffiths and J. A. de Haseth, Fourier transform infrared spectrometry, New York:

Wiley, 1986.

[95] Thermo Scientific, “Advantages of a fourier transform infrared spectrometer”, Technical

Note 50674, 2015.

[96] R. M. Silverstein, F. X. Webster and D. J. Kiemle, Spectrometric identification of organic

compounds, New York: Wiley, 2005.

References 83

[97] V. Saptari, Fourier-transform spectroscopy instrumentation engineering, Bellingham:

SPIE Publication, 2003.

[98] A. Vimont, F. Thibault-Starzyk and M. Daturi, “Analysing and understanding the active

site by IR spectroscopy”, Chemical Society Reviews, vol. 39, p. 4928–4950, 2010.

[99] M. Daturi, C. Binet, J.-C. Lavalley, A. Galtayries and R. Sporken, “Surface investigation

on CexZr1-xO2 compounds”, Physical Chemistry Chemical Physics, vol. 1, no. 24, pp.

5717-5724, 1999.

[100] S. Rousseau, O. Marie, P. Bazin and M. Daturi, “Investigation of methanol oxidation over

Au/catalysts using operando IR spectroscopy: determination of the active sites,

intermediate/spectator species, and reaction mechanism”, Journal of the American

Chemical Society, vol. 132, no. 31, pp. 10832-10841, 2010.

[101] P. D. Nolan, M. C. Wheeler, J. E. Davis and C. B. Mullins, “Mechanisms of Initial

Dissociative Chemisorption of Oxygen on Transition-Metal Surfaces”, Accounts of

Chemical Research, vol. 31, no. 12, p. 798–804, 1998.

[102] P. Hofmann, K. Horn, A. M. Bradshaw and K. Jacobi, “The adsorption and condensation

of oxygen on aluminium at low temperaturf”, Surface Science, vol. 82, no. 2, pp. L610-

L614, 1979.

[103] C. Kohrt and R. Gomer, “Adsorption of Oxygen on the (110) Plane of Tungsten”, The

Journal of Chemical Physics, vol. 52, no. 6, pp. 3283-3294, 1970.

[104] C. Wang and R. Gomer, “Sticking Coefficients of CO, O2 and Xe on the (110) and (100)

Planes of Tungsten”, Surface Science, vol. 84, no. 2, pp. 329-354, 1979.

[105] M. Mavrikakis and Y. Xu, “O2 dissociation on transition metals: Insights from first-

principles calculations”, Preprints of Papers- American Chemical Society, Division of Fuel

Chemistry, vol. 47, pp. 223-225, 2002.

[106] A. W. J. Smith, S. Poulston and L. Rowsell, “A New Palladium-Based Ethylene

Scavenger to Control Ethylene-Induced Ripening of Climacteric Fruit”, Platinum Metals

Review, vol. 53, no. 3, pp. 112-122, 2009.

[107] R. L. Augustine, Heterogeneous Catalysis for the Synthetic Chemist, New York: CRC

Press, 1996, pp. 11-13.

[108] R. A. Van Santen, P. W. N. M. van Leeuwen, J. A. Moulijn and B. A. Averill, Catalysis: An

Integrated Approach, London: Elsevier, 2000, pp. 154-156.



[109] R. A. Van Santen, Modern Heterogeneous Catalysis: An Introduction, London: Wiley,

2017, p. 309.

[110] A. Lyalin and T. Taketsugu, “Adsorption of Ethylene on Neutral, Anionic, and Cationic

Gold Clusters”, The Journal of Physical Chemistry C, vol. 114, no. 6, p. 2484–2493,

2010.

[111] G. Busca, V. Lorenzelli, G. Ramis and V. Sanchez Escribano, “Chemistry of olefins at

metal oxide surfaces: a tool for surface science investigation of oxide catalysts”, Materials

Chemistry and Physics, vol. 29, no. 1-4, pp. 175-189, 1991.

[112] Y. Soma, “Low temperature infrared spectra of ethylene adsorbed on alumina-supported

platinum and palladium”, Journal of the Chemical Society, Chemical Communications,

vol. 24, pp. 1004-1005, 1976.

[113] G. Busca, V. Lorenzelli, G. Ramis, J. Saussey and J. Lavalley, “FT-IR spectra of ethylene

molecularly adsorbed on metal oxides”, Journal of Molecular Structure, vol. 267, pp. 315-

329, 1992.

[114] J. Hiraishi, “The vibrational spectra of several platinum-ethylene complexes:

K[PtCl3(C2H4)]·H2O (Zeise's salt), K[PtCl3(C2D4)] · H2O and [PtCl2(C2H4)]2”,

Spectrochimica Acta Part A: Molecular Spectroscopy, vol. 25, no. 4, pp. 749-760, 1969.

[115] Y. Soma, “Infrared spectra of ethylene adsorbed on transition metals at low temperatures

and hydrogenation of the adsorbed species”, Journal of Catalysis, vol. 59, no. 2, pp. 239-

247, 1979.

[116] Y. Soma, “Infrared study of ethylene hydrogenation on alumina-supported platinum at low

temperature”, Journal of Catalysis, vol. 75, no. 2, pp. 267-274, 1982.

[117] M. Calatayud, S. E. Collins, M. A. Baltanás and A. L. Bonivardi, “Stability of formate

species on β-Ga2O3”, Physical Chemistry Chemical Physics, vol. 9, pp. 1397-1405,

2009.

[118] H. Li, X. Jiao, L. Li, N. Zhao, F. Xiao, W. Wei, Y. Sun and B. Zhang, “Synthesis of glycerol

carbonate by direct carbonylation of glycerol with CO2 over solid catalysts derived from

Zn/Al/La and Zn/Al/La/M (M = Li, Mg and Zr) hydrotalcites”, Catalysis Science &

Technology, vol. 5, no. 2, pp. 989-1005, 2015.

[119] J. C. Vargas, S. Libs, A.-C. Roger and A. Kiennemann, “Study of Ce-Zr-Co fluorite-type

oxide as catalysts for hydrogen production by steam reforming of bioethanol”, Catalysis

Today, vol. 107, pp. 417-425, 2005.

References 85

[120] S. Ivanova, C. Petit and V. Pitchon, “A new preparation method for the formation of gold

nanoparticles on an oxide support”, Applied Catalysis A: General, vol. 267, no. 1, pp.

191-201, 2004.

[121] A. Bumajdad, M. I. Zaki, J. Eastoe and L. Pasupulety, “Microemulsion-based synthesis of

CeO2 powders with high surface area and high-temperature stabilities”, Langmuir, vol.

20, no. 25, pp. 11223-11233, 2004.

[122] R. Jain, A. S. Poyraz, D. P. Gamliel, J. Valla, S. L. Suib and R. Maric, “Comparative study

for low temperature water-gas shift reaction on Pt/ceria catalysts: Role of different ceria

supports”, Applied Catalysis A: General, vol. 507, pp. 1-13, 2015.

[123] A. Trovarelli, C. De Leitenburg, G. Dolcetti and J. Lorca, “CO2 methanation under

transient and steady-state conditions over Rh/CeO2 and CeO2-promoted Rh/SiO2: The

role of surface and bulk ceria”, Journal of Catalysis, vol. 151, no. 1, pp. 111-124, 1995.

[124] W. Y. Hernández, F. Romero-Sarria, M. A. Centeno and J. A. Odriozola, “In situ

characterization of the dynamic gold−support interaction over ceria modified Eu3+.

Influence of the oxygen vacancies on the CO oxidation reaction”, The Journal of Physical

Chemistry C, vol. 114, no. 24, pp. 10857-10865, 2010.

[125] S. Ivanova, “Formation de nanoparticules d’or supportées : De la préparation à la

réactivité catalytique”, Thèse de doctorat, Université Louis Pasteur de Strasbourg,

Strasbourg, 2004.

[126] Q. Fu, W. Deng, H. Saltsburg and M. Flytzani-Stephanopoulos, “Activity and stability of

low-content gold–cerium oxide catalysts for the water–gas shift reaction”, Applied

Catalysis B: Environmental, vol. 56, no. 1, pp. 57-68, 2005.

[127] M. R. Kagan and R. L. McCreery, “Reduction of fluorescence interference in Raman

spectroscopy via analyte adsorption on graphitic carbon”, Analytical Chemistry, vol. 66,

no. 23, pp. 4159-4165, 1994.

[128] J. R. Ferraro, Introductory raman spectroscopy, Boston: Academic Press, 2002.

[129] Z. Popović, Z. Dohčević-Mitrović, M. Ŝćepanović, M. Grujić-Brojčin and S. Aŝkrabić,

“Raman scattering on nanomaterials and nanostructures”, Annalen der Physik (Leipzig),

vol. 523, no. 1-2, pp. 62-74, 2011.

[130] W. Weber, K. Hass and J. McBride, “Raman study of CeO2: Second-order scattering,

lattice dynamics, and particle-size effects”, Physical Review B, vol. 48, no. 1, pp. 178-

185, 1993.



[131] I. Langmuir, “The adsorption of gases on plane surfaces of glass, mica and platinum”,

Journal of the American Chemical Society, vol. 40, pp. 1361-1403, 1918.

[132] P. Bazin, S. Thomas, O. Marie and M. Daturi, “New insights into the methanol oxidation

mechanism over Au/CeO2 catalyst through complementary kinetic and FTIR operando

SSITKA approaches”, Catalysis Today, vol. 182, no. 1, p. 3–11, 2012.

[133] T. Lesage, C. Verrier, P. Bazin, J. Saussey and M. Daturi, “Studying the NO x-trap

mechanism over a Pt-Rh/Ba/Al2O3 catalyst by operando FT-IR spectroscopy”, Physical

Chemistry Chemical Physics, vol. 5, no. 20, pp. 4435-4440, 2003.

[134] I. Malpartida, E. Ivanova, M. Mihaylov, K. Hadjiivanov, V. Blasin-Aubé, O. Marie and M.

Daturi, “CO and NO adsorption for the IR characterization of Fe 2+ cations in ferrierite: an

efficient catalyst for NOx SCR with NH3 as studied by operando IR spectroscopy”,

Catalysis Today, vol. 149, no. 3, pp. 295-303, 2010.

[135] F. Bozon-Verduraz and A. Bensalem, “IR studies of cerium dioxide: influence of

impurities and defects”, Journal of the Chemical Society, Faraday Transactions, vol. 90,

no. 4, pp. 653-657, 1994.

[136] K. Fujimoto, T. Uchijima, M. Masai and T. Inui, New aspects of spillover effect in

catalysis: for development of highly active catalysts, Japan: Elsevier, 1993.

[137] C. Li and Q. Xin, “FT-IR spectroscopic investigation of methane adsorption on cerium

oxide”, The Journal of Physical Chemistry, vol. 96, no. 19, pp. 7714-7718, 1992.

[138] Z. Lu, A. Karakoti, L. Velarde, W. Wang, P. Yang, S. Thevuthasan and H.-f. Wang,

“Dissociative binding of carboxylic acid ligand on nanoceria surface in aqueous solution:

a joint in situ spectroscopic characterization and first-principles study”, The Journal of

Physical Chemistry, vol. 117, no. 46, pp. 24329-24338, 2013.

[139] M. A. Gómez García, “Absorption - réduction des NOx provenant d’installations fixes sur

systèmes catalytiques HPW-métal supportés”, Université Louis Pasteur de Strasbourg,

Strasbourg, 2004.

[140] M. E. Fleet, X. Liu and P. L. King, “Accommodation of the carbonate ion in apatite: An

FTIR and X-ray structure study of crystals synthesized at 2–4 GPa”, American

Mineralogist, vol. 89, no. 10, p. 1422–1432, 2004.

[141] A. B. Kroner, M. A. Newton, M. Tromp, O. M. Roscioni, A. E. Russel, A. J. Dent, C.

Prestipino and J. Evans, “Time-resolved, in situ DRIFTS/EDE/MS studies on alumina-

References 87

supported rhodium catalysts; effects of ceriation and zirconiation on rhodium-CO

interactions”, ChemPhysChem, vol. 15, no. 14, pp. 3049-3059, 2014.

[142] M. E. Swanson, “Interaction of carbon monoxide with partially reduced ceria-zirconia

supported catalysts”, Doctoral thesis, University of Pittsburgh, Pittsburgh, 2008.

[143] M. E. Davis, “Zeolites and molecular sieves: not just ordinary catalysts”, Industrial &

Engineering Chemistry Research, vol. 30, no. 8, p. 1675–1683, 1991.

[144] S. Wang and Y. Peng, “Natural zeolites as effective adsorbents in water and wastewater

treatment”, Chemical Engineering Journal, vol. 156, no. 1, p. 11–24, 2010.

[145] M. Mabilia, R. A. Pearlstein and A. J. Hopfinger, “Molecular modeling of zeolite structure.

1. Properties of the sodalite cage”, Journal of the American Chemical Society, vol. 109,

no. 26, p. 7960–7968, 1987.

[146] D. Barthomeuf, “Zeolites as adsorbents and catalysts. The interactive system encaged

molecule/zeolite framework”, Studies in Surface Science and Catalysis, vol. 105, pp.

1677-1706, 1997.

[147] B. Cekova, D. Kocev, E. Kolcakovska and D. Stojanova, “Zeolites as alcohol adsorbents

from aqueous solutions”, Acta Periodica Technologica, vol. 37, pp. 83-87, 2006.

[148] M. Chandak and Y. Lin, “Hydrophobic zeolites as adsorbents for removal of volatile

organic compounds from air”, Environmental technology, vol. 19, no. 9, pp. 941-948,

1998.

[149] B. Erdoğan, M. Sakızcı and E. Yörükoğulları, “Characterization and ethylene adsorption

of natural and modified clinoptilolites”, Applied Surface Science, vol. 254, no. 8, p. 2450–

2457, 2008.

[150] A. F. Medina Villamil, “Valoración de un residuo minero rico en óxidos de silicio mediante

su transformación hacia zeolitas a través de un proceso hidrotérmico”, Tesis de maestría,

Universidad Nacional de Colombia, Bogotá, 2015.

[151] Q. H. Trinh, S. B. Lee and Y. S. Mo, “Removal of Ethylene from Air Stream by Adsorption

and Plasma-Catalytic Oxidation Using Silver-based Bimetallic Catalysts Supported on

Zeolite”, Journal of Hazardous Materials, vol. 285, pp. 525-534, 2015.

evaluación de propiedades superficiales y catalíticas del ... · pdf...

Documents