Post on 19-Aug-2020
ZAHRA SARSHAR
HYDROCARBON TRAPADSORBENTS FOR REDUCING COLD-START EMISSIONS OF
AUTOMOTIVE EXHAUST
Mémoire présenté à la Faculté des études supérieures de l'Université Laval
dans le cadre du programme de maîtrise en génie chimique pour l'obtention du grade de Maître ès sciences (M. Sc.)
DEPARTEMENT DE GÉNIE CHIMIQUE FACULTÉ DES SCIENCES ET DE GÉNIE
UNIVERSITÉ LA V AL QUÉBEC
2008
© ZAHRA SARSHAR, 2008
11
Résumé
Avec l'entrée en vigueur des nouvelles réglementations sur les émissions polluantes
des moteurs à essence, de plus en plus d 'attention est portée sur la phase « de démarrage à
froid» pendant laquelle environ 80% des hydrocarbures (HC) sont émis. Notre étude a porté
sur le concept de diffusion "single-file" (dans les zéolithes 1-D) utilisé comme une
approche possible pour diminuer les émissions de HC durant le démarrage à froid. Ce
mécanisme est envisagé sur une variété de zéolithes aux structures MTW et IFR et pour
l' étude de leur capacité de piégeage des molécules plus légères émises souvent avant que le
catalyseur trois-voies atteigne sa température d' allumage.
Dans ce projet, une série de zéolithes telles que AI- et Fe-ZSM-12 (structure MTW)
avec différents cations (Na+, H+, Ag+ et Mg+) ainsi que B-SSZ-42 (structure IFR) sous la
forme Na + et H+ ont été synthétisées. La caractérisation par désorption à température
programmée (DTP) d'éthylène et de toluène, utilisés comme adsorbats,. a été examinée
pour établir le potentiel de ces zéolithes comme pièges pour les HC. Nous avons aussi
démontré la dépendance entre les performances des zéolithes en fonction du type de métal
substitué (Al, Fe) dans le réseau, l' ampleur de cette substitution, le type de cation échangé,
la structure des canaux, le diamètre des pores, et ainsi la présence d'autres types de
molécules adsorbées comme la vapeur d'eau et/ou le gaz CO2.
Dans la série des échantillons AI-ZSM -12, celui qui a été sous forme protonique (H
ZSM-12), montre des capacités adsorption plus élevées pour les deux adsorbats, AI-ZSM-
12 (Ag) montre des capacités élevées et stables pour piéger les deux adsorbats dans tous les
mélanges étudiés, montra1).t ainsi son insensibilité par rapport au CO2 et/ou H 20. Il a été
démontré que cet échantillon avait des températures de désorption plus élevées, de l'ordre
de 300 oC, pour les deux adsorbats. La substitution isomorphe d'Al par Fe dans les
structures MTW n'a pas donné de grands changements au niveau de la capacité
d' adsorption ni dans la température de désorption de ces deux adsorbats.
-- --~-- -------------------------------------~------------------------------------------
111
B-SSZ-42 montre une capacité remarquable d'adsorption sous les deux formes Na+ et
H+ pour l' éthylène et le toluène .. En effet, la capacité d'adsorption des deux adsorbats est
restée approximativement inchangée en présence des molécules de H20 et CO2 co
adsorbées.
Les propriétés structurelles et acid~s de tous les échantillons synthétisés ont été aussi
examinées par différentes techniques telles que: DRX, BET, MEB, MET, RMN,
spectroscopie d' absorption atomique, FTIR de pyridine adsorbée, désorption des
hydrocarbures à la température programmée et la thermo-désorption d' ammoniac.
IV
Absfract As regulations for emissions from gasoline engines become stricter, attention has
been focused . on the start-up phase (cold-start), when about 80% of total hydrocarbons
(HC) emissions of an engine' are released. In this thesis, the concept of single-file diffusion
(in 1-D zeolites) was employed as a feasible approach to control automotive HC emissions
during cold-start. This mechanism was investigated over a variety of zeolites with MTW
and IFR structures for studying their trapping ability for lighter HC molecules, which are
often desorbed before the three-way catalyst reaches its light-off temperature.
In this work, a series of zeolites such as: AI- and Fe-ZSM-12 (MTW structure) with
different cations (Na+, H+, Ag+ and Mg2+) and B-SSZ-42 (IFR structure) in Na+ and H+
forms were synthesized. Temperature-programmed desorption (TPD) of ethylene and
toluene, as probe adsorbates, was employed to screen these synthesized zeolites as potential
HC traps. The performance of the synthesized samples as HC traps was found to be
affected by the type of framework substituted metal, extent of metal substitution, the type
of exchangeable cation, channel structure and pore diameter of the samples and also the
presence of other constituents like water vapour and/or CO2 in the adsorbed phase.
AI-ZSM-12 samples, in particular 'a proton exchang~d AI-ZSM-12 among the ZSM-
12 samples, were found to have higher adsorption capacity for both adsorbates. A silver
exchanged MTW zeolite, AI-ZSM-12 (Ag) exhibited high and stable trapping capacities for
both probe molecules in aIl mixtures investigated in this study, thus showing essentiaIly no
sensitivity to CO2 and/or H20. It was also found that 'AI-ZSM-12 samples showed higher
desorption temperature of about 300 oC for both probe molecules. Isomorphous substitution
of Al by Fein MTW structures did not lead to drastic çhanges in adsorption capacities and
desorption temperatures of the two adsorbates.
B-SSZ-42 in both Na+ and H+ forms demonstrated remarkably high adsorption
capacity for both ethylene and toluene. Indeed, the adsorption capacity of both sorbates
remained approximately unchanged in the presence ofH20 and CO2 molecules.
- - --- - --- ----- - - - - - -------
v
The structural and acidic properties of aIl synthesized samples were further
investigated by a variety of characterisation techniques such as XRD, BET, SEM, TEM,
lIB MAS NMR, atomic absorption spectroscopy, FTIR of adsorbed pyridine, temperature
programmed desorption of hydrocarbons, thermodesorption of ammonia.
VI
Foreword This dissertation was composed in the form of scientific articles, which at the time of
the thesis submission were accepted for publication. The first author for the two articles is
the submitter of this M.Sc thesis.
In the first chapter, a general introduction, problem definition and bibliographie
review are presented. The objectives of this thesis are also introduced in this chapter. The
second chapter is devoted to the description of experimental details of the synthesis
procedures and characterisation techniques. The first part of this chapter is focused on
different approaches for AI-ZSM-12 and Fe-ZSM-12 samples and compares the various
methods of synthesis. The second part is devoted to the description of the variety of
characterisation methods used in this study.
The third chapter reports the results of the first paper focusing on ZSM -12 zeolite,
which was selected as a hydrocarbon trap adsorbent for cold-start emissions. This paper
was accepted in the scientific journal Applied Catalysis B: Environmental in August 2008
(Ms. Ref. No.: APCATB-D-08-00494R1).
The fourth chapter contains the second paper concentrating on SSZ-42 zeolite as a
hydrocarbon trap adsorbent. In this paper, several characterisation techniques were carried
out to characterize the structural and acidic properties of SSZ-42 zeolite for the first time.
This paper was accepted in the scientific journal Microporous and Mesoporous Materials
in September 2008(Ms. Ref. No.: MICMAT-D-08-00687).
Finally, in chapter 5, the general conclusions and sorne recommendations for future
work complete the thesis.
VIl
Acknowledgements
1 wish to thank my advisor Dr. Serge 'Kaliaguine for introducing me to the exciting
world of zeolites and the opportunity to perform ' research in the area of zeolites. He gave
me the freedom to pursue the research directions and projects that 1 was interested in and
fully supported my work by providing all the necessary resources; 1 greatly appreciate it.
Many people in our research group have helped me greatly. Dr. Mohammad-Hassan
Zahedi-Niaki and Mf. Gilles Lemay trained me in the use of aIl the equipment and
analytical instruments in our labs. 1 also appreciate the cooperation with Dr. Mladen Éic
and Qinglin Huang, both from the Chemical Engineering Department of University of New
Brunswick. Thanks are due to the Faculty, Staff and other graduate students of the
Chemical Engineering Department, Université Laval for their help and friendship.
It is difficult to describe my gratitude to my family in a few words. They have made a
lot of sacrifices during my education and always encouraged me to pur sue my goals. 1 wish
to thank them especially my mother, Vajiheh, a constant source of encouragement and
support.
The financial support provided by Natural Sciences and Engineering Research
Council of Canada (NSERC) throughout this study is highly appreciated.
Finally, and with deep sense of affection, 1 offer my sincere thanks to my husband,
Hossein, for his encouragement and support throughout my M.Sc. program.
VIn
To my lovely mother: Vajiheh
IX
Table of contents
Résumé ............................................................................................................................................. ii
Abstract ........................................................................................................................................... iv
Foreword ......................................................................................................................................... vi
Acknowledgements ........................................................................................................................ vii
Table of contents .............................................................................................................................. ix
List of Tables ................................................................................................................................ xiii
List of Figures ............................................................................................................................... xiv
1. Introduction ............................................................................. .................................................... 2
1.1 General Background ...................................................................................................... ~ ....... 2
1.2 Zeolite Molecular Sieves ....................................................................................................... 3
1.2.1 Introduction .................................... ~ ................................................................................ 3
1.2.2 Definition and Structure .......................................................................... ~ ....................... 4
1.2.3 Categories and properties ................................................................................................ 5
1.2.4 Synthesis ......................................................................................................................... 6
1.2.4.1 Introduction .... '.' ........................................................................................................ 6
1.2.4.2 Hydrothermal Synthesis ........................................................................................... 7
1.2.4.3 Mechanism of Crystallization .................................................................................. 7
1.2.4.4 Reactants and Gel Compositions in Hydrothermal synthesis .................................. 8
1.2.5 Application of Zeolites .................................................................................................. 10
1.3 The Most Common Strategy Proposed to Solve the Cold-Start Problem ............................ 13
1.4 Objectives ............................................................................................................................. 14
2. Experim·ental ............................................................................................................................. 16
2.1 Synthesis ................................................................................... ~ ............................... ~ .......... 16
x
2.1.1 Synthesis of AI-ZSM-12 ............................................................................................... 16
2.1.2 Ion exchange of ZSM-12 .............................................................................................. 20 ,
2.1.3 Synthesis ofFe-ZSM-12 ............................................................................................... 20
2.1.4 Synthesis ofSSZ-42 ...................................................................................................... 23
2.1.5 Ion ex change of SSZ-42 ................................................................ ~ ............................... 24
2.2 Characterisation ................................................................................................................... 24
2.2.1 Elemental Analysis .......................................................... ~ ............................................ 24
2.2.2 Nitrogen Adsorption ........................................... ~ ......................................................... 25
2.2.3 X-ray Diffraction (XRD) .............................................................................................. 26
2.2.4 Scanning electron Microscopy (SEM) .......................................................................... 26
2.2.5 Transmission Electron Microscopy (TEM) .................................................................. 26
2.2.6 Nuclear Magnetic Resonance Spectroscopy (NMR) .................................................... 27
2.2.7 Fourrier Transform Infrared Spectroscopy (FTIR) ....................................................... 27
2.2.8 Thermodesorption of Ammonia (TPD of Ammonia) ................................................... 28
2.2.9 Temperature-Programmed Desorption ......................................................................... 29
2.3 References ............................................................................................................................ 31
3. MTW Zeolites for Reducing Cold-Start Emissions of Automotive Exbaust ...................... 37
Résumé ................................................................................... ~ ....................................................... 37
Abstract ..................................................................................................................................... : .... 38
3.1 Introduction .......................................................................................................................... 39
3.2 Experimental Details ............................................................................................................ 40
3.2.1 AI-ZSM-12 synthesis .................................................................................................... 40
3.2.2 Ion exchange ofZSM-12 .............................................................................................. 40
3.2.3 Fe-ZSM-12 Synthesis ................................................................................................... 40
Xl
3.3 Structural Characterisation .... · ............................................................................................... 42
3.3.1 XRD .................................................................................................................. ~ ............ 42
3.3.2 SEM .......... ~ ................................................................................................................... 42
3.3.3 Elemental Analysis ....................................................................................................... 42
3.3.4 Nitrogen Adsorption ..................................................................................................... 42
3.3.5 Fourrier Transform Infrared Spectroscopy (FTIR) ....................................................... 43
3.3.6 Ammonia Thermodesorption (TPD ofadsorbed ammonia) .......................................... 43
3.3.7 Temperature-Programmed ~esorption ......................................................................... 43
3.4. Results and Discussions ....................................................................................................... 44
3.4.1 XRD .............................................................................................................................. 44
3.4.2 SEM .............................................................................................................................. 44
3.4.3 Elemental Analysis ....................................................................................................... 44
3.4.4 Nitrogen Adsorption ............................................... ; ................................................ , ..... 47
3.4.5 Fourrier Transform.lnfrared Spectroscopy (FTIR) ....................................................... 48
3.4.6 Ammonia Thermodesorption (TPD of adsorbed ammonia) ......................................... 51
3.4.7 Temperature-Programmed Desorption ................................................ .......................... 53
3.4.7.1 Binary Toluene-Ethylene mixture ............................................................. · ............. 53
3.4.7.2 Temary Toluene-Ethylene-C02 mixture ................................................................ 59
3.4.7.3 Ternary ~oluene-Ethylene-H20 mixture ............................................................... 60
3.4.7.4 Quatemary Toluene-Ethylene-C02-H20 mixture .................................................. 60
3.5. Conclusion ...................................................................................................................... · .... 63 Acknowledgements ........................................................................................................................ 64
3.6 References ................................................................................................................................. 65
4. Synthesis, Structural and Acidity Characterisations of the Large-Pore Zeolite SSZ-42 for
Controlling Cold-Start Emissions ............................................................................................... 68
Xll
Résumé ........................................................................................................................................... 68
Abstract .......................................................................................................................................... 69
4.1 Introduction .......................................................................................................................... 70
4.2 Experimental Details ............................................................................................................ 71
4.3 Results and .Discussions ............................................................................................ .. .......... 73
4.3.1 XRD .............................................................................................................................. 73
4.3.2 SEM and TEM .............................................................................................................. 74
4.3.3 Nitrogen Adsorption ..................................................................................................... 75
4.3.4 FTIR ............................................... ............................................................................... 77
4.3.4.1 FTIR spectroscopy of framework vibrations ......................................................... 77
4.3.4.2 FTIR spectroscopy of adsorbed pyridine ............. , ................................................. 79
4.3.5 Ammonia Thermodesorption (TPD ofadsorbed ammonia) ......................................... 81
4.3.6 lIB MAS NMR .............................................................................................................. 83
4.3.7 Temperature-Programmed Desorption ......................................................................... 84
4.3.7.1 Binary Toluene-Ethylene mixtures ........................................................................ 84
4.3.7.2 Ternary and Quaternary Toluene-Ethylene- CO2 /H20 mixtures .......................... 88
4.4 Conclusion ........................................................................................................................... 90 Acknowledgements ........................................................................................................................ 90
4.5 References ............................................................................................................................ 91
5. General Conclusions and Recommendations ........................................................................ 93
5.1 General Conclusions ............................................................................................................ 93
5.2 Recommendations ................................................................................................................ 95
XllI
List of Tables
Table 3.1: Gel composition of synthesized samples ...................................................................... 41
Table 3.2: Results ofbulk chemical composition of calcined samples by AAS ........................... 47
Table 3.3: TexturaI properties from nitrogen adsorption isotherrns at 77 K .................................. 48
Table 3.4: FTIR results of chemisorbed pyridine (desorption at 423 K) ....................................... 49
Table 3.5: N·H3-TPD results .......................................................................................................... 52
Table 3.6: Toluene-ethylene binary TPD results ........................................................................... 55 .
Table 4.1: TexturaI properties from nitrogen adsorption isotherrns af77 K .................................. 77
Table 4.2: FTIR reSUltS of chemisorbed pyridine (desorption at 150 OC) ..................................... 79
Table 4.3: NH3-TPD results ...................................... ~ ................................................................... 81
Table 4.4: Toluene-ethylene binary TPD results ........................................................................... 84
XIV
List of Figures
Fig. 1.1: T04 tetrahedra where a is the 0-T -0 bond angle and ~ is the T -0-T bond angle ............ 5
Fig. 1.2: the shape of para-xylene that can diffuse freely in the channels of silicalite .................. 11
Fig. 1.3: Sodium zeolite A, used as a water softener in detergent powder .................................... 13
Fig. 2.1: XRD patterns showing the progress of ZSM-12 crystallization with time from a gel of composition Na20: 0.016Ab03: 1.93Si02 : MTEACI: 1893 H20 at 433 K .... ... ... ........ ... ............. 18
Fig. 2.2: Crystallization kinetics ofZSM-12 in the system with Si/Al = 60, OH-/Si02 = 0.3 , Na20/ (MTEA)20= 1.5 and H20/OH- = 125 ....................................................... .......................... 19
Fig. 3. 1: XRD of selected as-synthesized samples. (A) Al-ZSM-12 (Na); (B) Al-ZSM-12 (H); (C) Fe-ZSM-12 (40); (D) Al-Fe-ZSM-12 (40) .............................................................................. 45
Fig. 3. 2: SEM pictures of selected as-synthesized samples; (A) Na-ZSM-12; (B) Fe-ZSM-12 (40); (C) Fe-ZSM-12 (70) .............................................................................................................. 46
Fig. 3.3: FTIR spectra of (A) Na-ZSM-12; (B) Ag-ZSM-12, before and after pyridine chemisorption at different temperatures (B and L denote Bronstead .......................... .................. 50
Fig. 3.4: Ammonia thermodesorption profile of Ag-ZSM-12 ............................................... ........ 53
Fig. 3.5: Toluene/ethylene binary mixture TPD profiles of selected as-synthesized samples. (A) Na-ZSM-12; (B) H-ZSM-12; (C) Ag-ZSM-12; (D) Fe-ZSM-12 (70); ......................................... 57
Fig. 3.6: Volume of the binary (ethylene + toluene) adsorbed phase as a function ofmicropore volume .............................................................................................................. .............................. 58
Fig. 3.7: Comparison of trapping capacities of ethylene in gas mixtures under different conditions: binary: dry; ternary with CO2: CO2; ternary with H20: wet; quaternary: wet-C02 .... 61
Fig. 3.8: Comparison oftrapping capacities oftoluene in gas mixtures under different conditions: binary: dry; ternary with CO2: CO2; ternary with H20: wet; quaternary: wet-C02 ....................... 62
Fig. 4.1: XRD of selected samples. (A) reference (siliceous ITQ-4); (B) H-SSZ-42 .................... 74
Fig. 4.2: SEM pictures of as-synthesized samples; (A, B) Na-SSZ-42; (C, D) H-SSZ-42 ........... 75
Fig. 4.3: TEM images and diffraction patterns ofH-SSZ-42 .................................................. ...... 76
xv
Fig. 4.4: IR spectra of the samples: (a) as-synthesized Na-SSZ-42; (b) calcined .......................... 78 Fig. 4.5: FTIR spectra of (a) Na-SSZ-42; (b) H-SSZ-42, before and after pyridine chemisorption at different temperatures (B and L denote Bronsted and Lewis sites, respectively) ................................................................................................................................... 80
Fig. 4.6: Ammonia thermodesorption profile of (a) Na-SSZ-42; (b) H-SSZ-42 ........................... 82
Fig. 4.7: 1 lB MAS NMR spectra of calcined samples: (a) Na-SSZ-42; (b) H-SSZ-42 .................. 83
. Fig. 4.8: Toluene/ethylene binary mixture TPD profiles of selected samples. (A) Na-SSZ-42; (B) H-SSZ-42 ....................................................................................................................................... 85
Fig. 4.9: Volume of the binary (ethylene + toluene) adsorbed phase as a function of micropore volume .......................................................................................................................... 87
Fig. 4.10: Comparison of trapping capacities of ethylene/toluene in gas mixtures under different conditions: binary (dry); temary with CO2(C02); temary with H20 (wet) and quatemary( wet-CO2) ..................•.........................•.................................................................................................. 89
1
Chapter 1
Introduction & Literature Review
2
1. Introduction
1~1 General Background Automobile emissions containing unbumed hydrocarbons, carbon monoxide, and nitrogen
oxides constitute one of the most significant impacts on the environment and human society and )
have led to strict environmentallegislations and restrictive emission standards for vehicles [1 , 2].
In order to meet the emission standards, three-way catalysts (TWC) have been developed and
used on cars with gasoline-buming engines [3]. Three-way catalyst (TWC) contains individual or
a combination of the platinum group metals (PGM) palladium, platinum and rhodium (Pd, Pt and
Rh, respectively) to oxidize the hydrocarbons (HCs) and carbon monoxide (CO) to carbon
dioxide (C02) and water (H20) and to reduce the nitrogen oxides (NOx) to nitrogen. Since the
first introduction of emission control devices in automobiles in the 1970s, much effort has been
made to continuously improve automotive catalysis in order to satisfy the stricter legislation
regarding engine emission control [4]. It has long been known that the largest fraction of about
800/0 of the total HC emission is due to a cold-start, which lasts typically during the first 1-2
minutes from starting a vehicle. Tailpipe HC emissions during the cold-start are high because the '
catalyst is not at its light-off temperature, usually above 170 oC for fresh catalyst [5] , to
efficiently oxidize the HCs [6].
Different approaches have been attempted to solve this problem. First,. there is a series of
ideas which are based upon methods of quickly bringing the catalyst to working temperature.
Close-coupled or manifold mounted catalysts are placed in positions very near the engine, thus
reducing the time necessary for the heat of the engine exhaust to increase the catalyst temperature
. [7-9]. In the electrically heated catalyst (EHC), electrical power is provided to heat the catalyst at
start-up [10-13]. Exhaust-gas ignition (EGI) works by deliberately running the engine under very
rich conditions (air-to-fuel (AIF) ratio ~9) so that large quantities ofhydrogen are produced in the
exhaust, which is th en ignited by a glow-plug upstream of the catalyst [14, 15]. Another proposaI
is the combustion heated catalyst (CHC) in which hydrogen and oxygen are fed to the catalyst
prior to starting the engine, and the exothermic combustion reaction heats the catalyst [16]. A
variety of heat storage devices have also been suggested, aIl of which work on the principle of
3
retaining heat from the time the car was last shutdown until the following cold-start. However, aIl
of the solutions discussed above have inherent disadvantages. Close-coupled catalysts are located
in the valuable space near the engine compartment (which is inconvenient for engine design and
can result in a loss of power) and also must be robust to very high temperature exposure. EHCs
are bulky and require large amounts of power at start-up, often requiring the use of a second
battery. EGI systems require the complexity of implementing a glow-plug into the exhaust
stream. Feeding hydrogen to the catalyst prior to starting the engine is inconvenient, and requires
a ready source of hydrogen on-board. Heat storage devices are bulky, involve the use of
expensive materials, and may be difficult to fabricate.
The most common strategy regarding a potential solution to the cold-start problem is to
first trap hydrocarbons by adsorption in a porous material upstream of the catalytic converter to
be released when the catalyst has reached its working temperature. For this modification, various
solutions involving hydrocarbon adsorbents such as activated carbons or zeolites have been
proposed [17-27]. Zeolites are found to be the preferred adsorbents due to their stability under
elevated temperature and their thermodynamic affinity to HC' s. In the following section, a brief
literature review on structure and properties of zeolite molecular sieves is reported.
1.2 Zeolite Molecular Sieves
1.2.1 Introduction
In 1756, ' the Swedish mineralogist Axel Fredrick Cronstedt discovered that stilbite, a
natural mineraI , visibly lost water when heated and he named the class of materials zeolites from
the classical Greek words meaning 'boiling stone'. In fact, zeolites are classified in two types:
natural and synthetic. N atural zeolites were first discovered in cavities and vugs of basalts. At the
end of the 19th century, they were also found in sedimentary rocks [28]. Up to now, over 40 types
of natural zeolites have been found, but fewer than 30 of them have had their structures solved.
At present, natural zeolites are widely used in the fields of drying and separation of gases and
liquids, softening of hard water, treatment of sewage, and melioration of soils. Synthesis of
zeolites was first conducted at the end of the 19th century through mimicking of the geothermal
conditions for natural zeolites. In the 1940s, low-silica zeolites were first synthesized. In the
4
1950s, zeolites were mainly used in drying, separation, and purification of gases. Since the
1960s, zeolites have been widely used as catalysts and catalyst supports in petroleum refining. At
present, zeolites have become the most important adsorbents and catalysts in the petroleum
industry. Although, conipared with natural zeolites, synthesized zeolites have many advantages
such as: high purity, uniform pore size, and better ion-exchange abilities, natural zeolites are
more applicable when there are huge demands and lower quality requirements [29].
1.2.2 Definition and Structure
Zeolites are three-dimensional, mlcroporous, crystalline solids with well-defined pore
structures whose regular frameworks are formed by corner linked T04 tetrahedra (T = Si, Al) and
each oxygen being shared between two T elements. Figure 1.1 shows the T04 tetrahedra sharing
a common oxygen vertex. The presence of the trivalent metal cation Ae+ in the framework of the
zeolites creates a negative charge in the lattice which can be compensated by a cation. Cations
and water are located in the pores. The T04 tetrahedra (primary building units) are assembled to
form secondary building units (SBU). These SBU assemble into large polyhedral building blocks
which form thus the framework of zeolites. Zeolites have void space (cavities or channels) that
can host cations, water, or other molecules. The empirical formula of an aluminosilicate zeolite
can be expressed as:
where, the charge-balancing nonframework cation M has valence il, x is 2.0 or more, and y is the
number ofmoles ofwater in voids [30].
The Si and Al atoms in the framework can be substituted by other elements, su ch as: Ge
(for Si) and B, Fe, Ti, Zr, Zn, Co, Mn (for Al), which is namely isomorphic substitution. In the
related aluminophosphates (AIP04) , each negatively charged AI04 tetrahedron is balanced by a
,positively charged P04 tetrahedron,. and nonframework cations are not needed. Still other variants
include the silicoaluminophosphate (SAPO) structures in which the Si substitutes sorne P in the
AIP04 framework, each added Si needs a nonframework cation to balance the charge on the
framework.
5
Fig. 1.1: T04 tetrahedra where a is the 0-T -0 bond angle and ~ is the T -0-T bond angle
1.2.3 Categories and properties
Zeolites are divided into three categories based on the ~ Sil Al ratio: low (Sil Al = 1.0", 1.5),
medium (Sil Al = 2.0",5.0) and high (Sil Al = 10", 1 00) Sil Al ratio zeolites. These solids were
extensively explored and the emergence . of these zeolites facilitated the study of both the
structure and property of molecular sieves and porous compounds [31]. In order to increase the
thermal stability and acidity of zeolites, Breck et al. synthesized zeolite Y (Si/AI = 1.5",3.0) ~
which played an extremely important role in the catalysis of hydrocarbon conversion. From then
on, a variety of zeolites with a Sil Al ratio of 2",5 (intermediate silica zeolites) which include
mordenite, zeolite L, erionite, chabazite, clinoptilolite, zeolite 0 , etc, have been synthesized [32].
At the beginning of the 1960s, scientists at Mobil Corporation started to use organic amines and
quatemary alkylammonium cations as templates in the hydrothermal synthesis of high-silica
zeolites, and this is considered a milestone in the progress of zeolite synthesis. In 1972, Argauer
and Landolt synthesized the first important member of the pentasil family, ZSM-5, using Pr4NCI
or Pr4NOH as the" template at 120 oC [33] , whereas in 1973, Chu synthesized ZSM-11 using
BU4N+ as the template [34]. In 1974, Rosinski and Rubin prepared ZSM-12 using Et4N+ as the
template [35]. The rapid progress in synthesis of high-silica zeolites· facilitated the study of the
secondary synthesis of zeolites. Sorne high-silica zeolites such as zeolite Y (Sil Al >3), which
were difficult to synthesize directly, could be prepared from zeolites with medium Sil Al ratios
through steam treatment or de-alumination in framework by reaction with Si.
6
Zeolites possess special properties that rnake thern unIque when cornpared with other
inorganic oxide materials. These properties include: the microporous character of the uniform
pore dimensions, the ion exchange properties, the separation and catalysis properties, the ability
to develop internaI acidity, the high thermal stability and the high internaI surface area. These
fascinating properties are essentially determined by their unique structural characters, such as: the
size of the pore window, the accessible void space, the dimensionality of the channel system and
the numbers and locations of cations, etc.
Many zeolites are thermally stable to over SOO°C. Sorne are stable in an alkaline
environment,· and sorne are stable in acidic media. They are stable to ionizing radiation and can
thus be used to adsorb radioactive cations. Zeolites can separate molecules based on size, shape,
polarity and degree of unsaturation among others. Zeolites can be regenerated using relatively
easy methods such as: heating to remove adsorbed materials, ion exchanging with sodium to
remove cations or pressure swing to remove adsorbed gases.
Zeolites modification implies an irreversible change, unlike ion exchange or adsorption.
There are a number of different ways that zeolites can be modified. The framework of the zeolite
can be modified by synthesizing zeolites with metal cations other than aluminium and silicon in
the framework, which is called isomorphous substitution. The framework of the zeolites can be
modified by dealumination to increase the silica . and increase the hydrophobic nature of the
zeolites. Another method is ion exchange of cation (usually Na +) by other cations (Li+, K+, Rb +,
Cs +, Be2+, Mg2+, Ca2+, etc.). There are appropriate methods to amend zeolites and their properties
that impart unique characteristics to them.
1.2.4 Synthesis
1.2.4.1 Introduction
Zeolites are generally synthesized by hydrothermal crystallization of a heterogeneous gel,
which is composed of a liquid and a solid phase. The reaction mixture contains the sources of the
elements which will form the final solid (T = Si, Al, P, Ti, Fe etc.), the mineralizer which can be
7
OH- or P-, mineraI cations which are normally alkali or alkaline-earth, in many cases organic
compounds acting as structure directing agents and usually water as solvent. The synthesis of
zeolites by the hydrothermal method is genera}ly in alkaline aqueous media, under autogeneous
pressure and at temperatures between 60 and L'QO oc depending on the type of zeolite that one
wishes to obtain (Outh and Kessler, [36]). In addition to hydrothermal method, the formation of
zeolitic microporous crystals can be also achieved by dry gel conversion (DOC) technique. In
1990, Xu et al. [37] have reported for the first time that a dry aluminosilicate gel can be
transformed to ZSM-5 zeolite when contacted with water vapour and a volatile amine (as
structure directing agent). To avoid the contact with the dry gel, which is placed in the middle of
the autoclave, the water and amine are initially placed in their liquid form at the bottom of the
autoclave. Upon heating, the vaporization ofwater and amine takes place. The vapours reach dry
gel and bring about crystallization. This method is beyond our work, thus its details are not
reported in this thesis however, the related details, important reactants and their roles in zeolite
hydrothermal synthesis are discussed.
1.2.4.2 Hydrothermal Synthesis
Hydrothermal synthesis refers to the synthetic reactions conducted at appropriate
temperature (100-1000°C) and pressure (1-100 MPa) in aqueous or organic solvents within a
specially sealed container or high-pressure autoclave under subcritical or supercritical conditions.
Under hydrothermal conditions, the physical and chemical properties of reactants can be
significantly changed. Hydrothermal syntheses and related reactions have become increasingly
important routes for the preparation of microporous materials.
1.2.4.3 Mechanism of Crystallization
The hydrothermal synthesis process of a zeolite basically consis-ts of two stages: the initial
formation of the hydrated aluminosilicate gel and the following crystallization process of the gel.
In fact, the crystallization process of the .hydrated aluminosilicate gel is very complicated. No
decisive conclusions have been reached for this crystallization process so far. Trong On and
Kaliaguine, in chapter 3 of the book Nanoporous M~terials [38] , explained two mechahisms of
zeolite germination. In the first mechanism, it was considered that nucleation is occurring within
the hydrogel and the crystals are formed by solid-solid transformation. In the second approach, it
8
was supposed that nucleation is taking place directly from liquid phase and, once the nuclei reach
a critical size to from stable nanoparticles, they grow into crystal by the progressive incorporation
of dissolved species. However regardless of the liquid- or solid-phase transformation mechanism,
it is commonly accepted that the crystallization process consists of four steps:
(1) Condensation ofpolysilicate and aluminate anions
(2) Nucleation of zeolites
(3) Growth ofnuclei
(4) Crystal growth of zeolites which sometimes results in secondary nucleation
1.2.1.4 Reactants and Gel Compositions in Hydrothermal synthesis'
The basic reactants used in the synthesis of zeolites include a silicon source, an aluminium
source, metal ions, a base, a mineraliser and water. Sorne additives, su ch as organic template or
inorganic salts, could be critical for the successful crystallization of a specific zeolite. Among
them, th.e silicon and aluminium sources are the two most important reactants. Frequently used
silicon and aluminium source reactants are listed below [29]:
Silicon sources:
Water glass: (Na20.xSi02), where x is modulus; sodium silicate: Na2Si03.9H20 ; silica gel:
Ludox-AS-40 colloidal sol: Si02 40 wt% , NH4 + (counter ion); Ludox-HS-40 colloidal sol: Si02
40 wt%, Na+ (counter ion); Fumed silica: Aerosil~200, Cab-O-Sil M-5; tetra-ethyl orthosilicate
(TEOS): Si(OC2Hs)4; tetra-methyl orthosilicate (TMOS): Si(OCH3)4.
Aluminium sources:
Sodium aluminate: NaAI02; Boehmite (pseudo-boehmite): AIOOH, Ab03 70%, H20 300/0;
aluminium hydroxide (Gibbsite), AI(OH)3; aluminium isopropoxide: AI(O-iC3H7)3; aluminium
nitrate: Al(N03)3.9H20 ; metallic aluminium.
The type of zeolite that can be crystallized from a synthetic system depends on the various
parameters which are listed below.
Sil Al ratio:
9
The Sil Al ratio in the parent mixture plays an important role in determining the structure and
composition of the final product. However, there is no quantitative correlation between the Sil Al
ratio in the product and that in the gel composition. Usually, the Si/AI ratio in the precurs<?r gel is
higher than that in the crystallized product. The excess of silicon is left in the mother liquid. For
sorne zeolites structures, the Si/AI ratio in the crystallized product can be tuned without loss of
structure. For instance, sorne zeolites like ZSM-5 can have both high- and low- silica forms. The
Sil Al ratio can enormously affect the zeolite properties and functions by causing sorne changes in
structure, rnorphology and physicochemical properties of zeolites.
Alkalinity
Zeolite synthesis is usually performed under basic or strongly basic conditions. Many zeolites can
be crystallized from the basic Na20-Ab03-Si02-H20 system. For this specific system, the
alkalinity is defined as the OH-;Si ratio or the concentration of base (H20lNa20). Basically,
increasing the ratio of OH-/Si leads to a higher solubility of silicon and aluminium sources, which
will alter the polymerization state. Moreover, a higher alkalinity can decrease the polymerization
degree of the silicate anions, speed up the polyrnerization of the polysilicate and aluminate and
accelerate the crystallization of zeolites [39].
Another effect of alkalinity on the crystallization of zeolites is that increasing the alkalinity of the
system favours the formation of aluminium-rich zeolites. The most remarkable example is the
zeolite faujasite. High-silica Y zeolite can only be synthesized from the precursor gel with a low
alkalinity (OH-/Si), whereas low-silica X zeolite can be crystallized frorn that with high alkalinity
[29].
10
Inorganic cations:
It is found that inorganic cations also play an important role in the crystallization of zeolites. For
instance, the zeolites analcime (ANA), cancrinite (CAN), chabazite (CHA), gmelinite (GME),
faujasite (FAU), A (LTA) and phillipsite (PHI) could be crystallized from the alumino-silicate
crystallization system (Ah03-Si02-H20) in the presence of sodium-containing species, whereas
zeolites KE, KF, KZ, KG, KR, KL, KM, KQ and KW could be promoted from the same
precursor gel system in the presence of potassium-containing species [32]. Moreover, zeolite L
(LTL), offretite (OFF) and erionite (ERI) can be crystallized from this synthetic system in the
presence of both sodium- al).d potassium-containing species.
Aluminosilicate zeolites are normally synthesized under basic conditions. The introduction of
OH- ions to the synthetic system will necessarily lead to the introduction of correlated cations.
These positively charged cations play an important role in the polymerization of polysilicates and
aluminates by affecting the polymeric state and their distribution and have an important effect on
the colloidal chemistry of aluminosilicate as weIl. Moreover, cations present in the synthetic
system also have significant effects on the formation of the framework structure of zeolites [40].
Structure Directing Agents:
In the synthesis of high-silica zeolites and phosphate-based molecular sieves, sorne orgarnc
amines are commonly introduced into the ' reaction system to act as templates or structure
directing agents (SDAs). They are located in the channels or cages of zeolites, playing the
following roles in the formation of specific channels and cages: space filling, structure directing
and true templating [41]. The role o.f this material in formation of a given structure is really
significant since using different structure directing agents for a specific structure leads to
formation of zeolites with different morphology and structural properties. 1
1.2.5 Application of Zeolites _ Since 1950s, there have been three tr~ditional fields of application for molecular sieves and
porous materials:
(1) Separation, purification, drying and environment process
(2) Petroleum refining, petrochemical, coal and fine chemical industries
Il
(3) Ion exchange, detergent industry, radioactive waste storage, soil rernediation and treatrnent of
liquid waste.
The shape-selective properties of zeolites are the basis for their use in molecular adsorption.
The ability to adsorb preferentially certain molecules, while excluding others, has opened up a
wide range of molecular sieving applications. Sometimes, it is sirnply a matter of the size and
shape of pores controlling access into the zeolite. In other cases, different types of molecules
enter the zeolite, but sorne diffuse through the channels more quickly, leaving others stuck
behind, as in the purification of para-xylene by silicalite (Figure 1.2).
Fig. 1.2: the shape of para-xylene that can diffuse freely in the channels of silicalite
Cation-containing zeolites are extensively used as desiccants due to their high affinity for
water, and also find application in gas separation, where rnolecules are differentiated on the basis
of their electrostatic interactions with the metal ions. Conversely, hydrophobic high silica zeolites
preferentially adsorb organic sol vents. Zeolites can thus separate rnolecules based on differences
of size, shape and polarity.
Zeolites contribute to a cleaner and safer environment in a great number of ways. In fact
nearly every application of zeolites has been driven by environmental concems or plays a
significant role in reducing toxic waste and energy consumption.
In powder detergents, zeolites replaced harmful phosphate builders, now banned in many
parts of the world because of water pollution risks. Catalysts, by definition, make a chemical
12
process more efficient, thus saving energy and indirectly reducing pollution. Moreover, processes
can be carried out in fewer steps, minimizing unnecessary waste and by-products. As solid acids,
zeolites reduce the need for corrosive liquid acids, and as redox catalysts and sorbents, they can
rempve atmospheric pollutants, such as engine exhaust gases and ozone-depleting CFCs. Zeolites
can also be used to separate harmful organics from water, and also in removing from it heavy
metal ions, including those produced by nuclear fission.
Zeolites have the ability to act as catalysts for chemical reactions, which take place within
the internaI cavities. An important class of reactions is that catalyzed by hydrogen-exchanged
zeolites, whose framework-bound protons give rise to very high acidity. This is exploited in
many organic reactions, including crude oil cracking, isomerisation and fuel synthesis. Zeolites
can also serve as oxidation or reduction catalysts, often after metals have been introduced into the
framework. Examples are the use of titanium ZSM-5 in the production of caprolactam, and
copper zeolites in NOx decomposition. Underpinning aIl these types of reaction is the unique
microporous nature of zeolites, where the shape and size of a particular pore system exerts a
steric influence on the reaction, controlling the access of reactants and products. Thus zeolites are
often said to act as shape-sélective catalysts. Increasingly, attention has focused on fine-tuning
the properties of zeolite catalysts in order to carry out very specific syntheses of high-value
chemicals e.g. pharmaceuticals and cosmetics.
The loosely-bound nature of extra-framework metal Ions, su ch as Na+ in zeolite NaA
(Figure 1.3), means that they are often readily exchanged for other types of metal when in
aqueous solution. This is exploited in a major way in water softening, where alkali metals such as
sodium or potassium prefer to exchange out of the zeolite, being replaced by the "hard" calcium
and magnesium ions from the water. Many commercial washing powders thus contain substantial
amounts of zeolite. Commercial waste water containing heavy metals, and nuclear effluents
containing radioactive isotopes can also be cleaned up using such zeolites.
13
Fig. 1.3: Sodium zeolite A, used as a water softener in detergent powder
In addition to the traditional application fields, zeolites and related porous materials may
also find application in new areas such as microelectronics and molecular device manufacture.
1.3 The Most Common Strategy Proposed to Solve the Cold,;.Start Problem
As mentioned in previous section, the most common strategy toward solving the cold-start
problem is applying a hydrocarbon adsorbent to trap HCs and avoid releasing them until the
three-way catalyst reaches its light-off temperature. This adsorbent is preferably a zeolite. A
series of zeolites BEA, MFI, MOR and X [19, 42, 43] and silicoaluminophosphate molecular
sieves [44] were .investigated regarding their hydrocarbon adsorption capacities under a variety of
conditions [45]. Burke et al. [46] reported that beta zeolite (BEA) is a promising material for this
application, while Elanogovan et al. [47] found that SSZ-33 is superior to BEA based on their
adsorption studies of a series of medium and large pore silicoaluminate zeolites. On the other
hand, Czaplewski et al. [48] observed an interesting phenomenon of trappirig molecules, within
the one-dimensional channel of EUO and MOR zeolites, as compared to more typical three
dimensional zeolites. This phenomenon was designated as the single-file diffusion mechanism,
which prevents the passage of molecules by one another inside the micropores of the one
dimensional zeolites. AIl these HC trapping solid adsorbents have been generaIly found to
adequately trap heavier components in the exhaust, e.g. , aromatics, but the light HC ' s, e.g. , C2-C4
fractions desorb from the trap much before the catalyst has reached a high enough temperature
for efficient combustion to occur.
14
1.4 Objectives
The objective of this work is to develop novel hydrocarbon trap adsorbents with high
adsorption capacity, hydrothermal stability and good selectivity. To achieve this goal, two series
of zeolites with MTW (ZSM-12) and IFR (SSZ-42) structures were selected to be synthesized
and characterized by a variety of techniques in order to investigate their capabilities as
hydrocarbon trap adsorbents for the cold-start problem. These two materials were selected based
on our previous works. Iliyas et al. [49, 50] synthesized a variety of zeolites including 10 and 12
oxygen ring aperture, 10R and 12R respectively, and their capabilities were tested as hydrocarbon
adsorbents for solving cold-start problem. They showed that the 10R zeolites had very low
micropore filling compared t6 the 12R zeolites and only the latter ones allowed fast enough
adsorption of hydrocarbons. According to these results, we chose two 12R zeolites with different
structures, ZSM-12 and SSZ-42. The literature showed that these solids have hydrothermal
stability and high adsorption capacity for trapping hydrocarbons [50, 79].
ZSM-12 is a high-silica large pore zeolite with 12-oxygen ring and non-intersecting one
dimensional straight channels. The channel dimension is 5.7 x6.1 0 A. SSZ-42 is also a large pore
one-dimensional microporous material with undulating, 12-MR channel and its channel
dimension is 6.4 x7.5 0 A.
In next chapter, the synthesis of these two materials and the characterisation techniques will
be discussed.
Chapter 2
Experimental
15
2. Experimental
2.1 Synthesis
2.1.1 Synthesis of AI-ZSM-12
16
ZSM -12 (MTW) is a high silica zeolite first synthesized by Rosiriski and Rubin in 1974
[51]. ZSM-12 has a mono dimensional 12-membered ring channel system with pore openings of
5.6 x 6.0 A [52]. The acidic form of ZSM-12 has been claimed to be a very useful catalyst in
cracking, hydrocracking and various other petroleum refining processes [51]. The aluminium rich
form of ZSM-12 would be particularly useful in the above processes due to its higher acid site
density and hydrogen transfer capability. Another interesting property of ZSM-12 is its
remarkable time stability for several hydrocarbon conversion reactions [53].
A variety of organlc cations such as quaternary tetraalkylammonium, mixed
alkyllarylammonium compounds and diquaternary ammonium cations have been claimed to act
as templates in the synthesis of ZSM-12 [54]. Compounds containing the ~etraethylammonium
(TEA) cation such as tetraethylammonium bromide (TEABr) ànd tetraethylammonium hydroxide
(TEAOH) or methyltriethylammonium bromide (MTEABr) have been commonly used as the
templates. Jacobs and Martens [54] provided an excellent overview of the literature on synthesis
of ZSM-12 and other members of its familY. They concluded that with monomeric quatemary
ions, the lower Si/Allimit for ZSM-12 that one can achieve was around 40. It is also weIl known
that, at lower Si/Al ratios, the gel crystallizes into ~-zeolite [51]. Low Si/Al ZSM-12 seems to
have been crystallized successfully only when N-containing polymers were used as templates
[54]. Other zeolites like TPZ-3 and CZH-5, which are considered to be members of the ZSM-12
family, have been synthesized from aluminium rich mixtures. By comparing the XRD patterns pf
these materials with that of ZSM-12, Jacobs and Martens [54] concluded that these claimed
materials are probably ZSM-12-1ike materials but have somewhat different XRD patterns due to
the presence of impurities like kenyaite and cristobalite. Therefore, these materials have the same
ZSM -12 structure, but differ from ZSM -12 in their degree of crystallographic purity.
17
Ernst et al. [55] carried out a systematic study on the synthesis of ZSM-12 using MTEABr
as the organic templ,~.te. Their results suggested that the sodium cation played a very important
role in the nucleation and crystal growth process, and MTEA acts only as a pore filling agent.
Upon increasing the aluminium content of the reaction mixture, crystallization rates decreased
and also resulted in poorly formed crystals, which led them to concliIde that incorporation of
aluminium into the framework of ZSM -12 is a difficult disruptive process. From this study the
all;thors were able to '-arrive at optimum conditions for synthesis of ZSM-12 using MTEA. The
factors which affect the cryst~llization of ZSM -12 in t~e presence of MTEA and TEA were also
studied by Katovic and Giordano [56]. They found that MTEA expanded the formation domain
of ZSM-12 and favored aluminium incorporation into the MTW framework, which led them to
conclude that MTEA was a more specific template for the synthesis of ZSM-12.
In zeolite synthesis, the template is one of the most cost-bearing factors. ZSM-12 synthesis
using MTEABr would be relatively expensive, while use of TEAOH or preferably TEABr as the
template would prove much more economical. The synthesis of ZSM-12 using TEA cation has
not been studied systematically and earlier attempts to produce ZSM-12 with Si/Al < 50 using
TEA did not produce pure ZSM-12. Ernst et al. [55] report that with TEA and low Si/Al ratios
« 60), ZSM-5 was synthesized together with ZSM-12. The original patent [1] also states that
ZSM-12 is synthesized preferentially at high Si/Al ratios (> 42.5). Katovic and Giordano [56]
report the lower Si/Allimit for successful synthesis of ZSM-12 using TEA to be around 50. The
crystallization field of ZSM -12 has been identified as a function of key parameters such as Si/Al,
OH-/Si02 and TEA/ Si02 ratios.
The synthesis procedure used in this work is the one p:roposed by Ernst et al. [55]. In a
typical preparation, 80.9 g -of a sodium silicate solution (Merck, 28.5 wt% Si02, 8.8 wt% Na20 ,
62.7 wt% H20) was mixed with 80 g distilled water. To this, a solution of 30 g
methyltriethylammoniumchloride (Fluka) in 100 g water was added, under stirring, followed by
adding a solution containing 2.4 g Al(N03)3.9H20 (Merck) in 30g H20. Afterwards, 4.7 g H2S04
(Merck, 98%) was added under vigorous stirring. The-obtained gel was th en transferred into a
teflon-lined stainless steel autoclave and the crystallization was conducted in a fumace at 160°C
without agitation for 7 days. The final products were separated by centrifugation and then
18
repeatedly washed with distilled water, followed by drying in air at 70°C ovemight. Finally,
calcination at 550°C was carried out ovemight. The molar composition of the resulting gel was
calculated as Si02/Ah03 = 120, OH-/Si02 = 0.3, Na20/ (MTEA)2 0= 1.5 and H20/OH- = 125.
As mentioned, the required crystallization time in this method is 7 days. However, it can
be seen from Figure 2.1 that crystallization of the zeolite takes place rapidly after about 41 hours
and increases to about 90% crystallinity in 48 hours. With more time there is a slow increase in
the crystallinity according to Figure 2.2.
138 h
118 h
65 h
48 h
41 h
24 h
17 h
5 10 15 20 25 30 35
2 theta
Fig. 2.1: XRD patterns showing the progress of ZSM-12 crystallization with time from a gel of composition Na20: 0.016Ah03: 1.93Si02: MTEACI: 1893 H20 at 433 K.
-- - ---~
19
100 --. .--. / .
80
~ ~
60
~ S fil
40 ~ U
20 • / •
0 0 30 60 90 120 150
Time (hours)
Fig. 2.2: Crystallization kinetic~ ofZSM-12 in the system with Si/Al = 60, OH-/Si02-= 0.3, Na20/ (MTEA)20= 1.5 and H20/OH- = 125.
In the TEAOH synthesis, crystallization in presence of NaOH was very similar to the
TEACI synthesis; however, crystallization was mu ch slower in the case of gels containing KOH,
with the crystallinity approaching 100% only after around six days. When sodium was eliminated
from the gel or when its amount was greatly decreased, crystallization did not occur suggesting
that it is playing a role in nucleation. Jacobs and Martens state that larger cations like potassium
have a structure breaking effect since they are less strongly hydrated, which explains the sluggish
'crystallization in the KOH containing gels [54].
Temperature is one of the important factors in zeolite synthesis. In ZSM-12 synthesis,
lowering the temperature as weIl as the pH, as suggested by Ernst et al. [55] prevented the
formation of crlstobalite. By performing synthesis experiments at temperatures ranging from 140
to 175°C, a temperature of around 160°C was found to be ideal for the crystallization of ZSM-12.
Higher temperatures resulted in formation of impurity phases ' while lower temperatures were
20
more favorable for crystallization of zeolite beta [55]. A similar observation was made by Eapen
et al. [57] who studied crystallization of zeolite beta.
2.1.2 Ion exchange of ZSM-12
The calcined zeolites were three-fold ion-exchanged at room temperature with O.l M
ammonium nitrate solution. Ag incorporation was carried out by ion-exchange in the liquid phase
using silver nitrate over H-form of zeolite sainples. For this, the solution was stirred at 80°C for 2
hours. Thereafter, the samples were centrifuged and washed with distilled water and th en dried
ovemight at 100°C.
2.1.3 Synthesis of Fe-ZSM-12
One of the emerglng trends in the area of zeolite modification is the isomorphous
substitution of Si or Al in the framework by other elements like: Ge, Ga, B, Fe, Ti, etc.
Isomorphous substitution can affect and amend the zeolite properties such as: the acidity, redox
properties, catalytic activity and other functions of the parent zeolite. Isomorphously substituted
zeolites are finding applications not only in petrochemical industry but also in organic synthesis
reactions [58-62]. Barrer [28] and Tielen et al. [63] have reviewed possibilities of isomorphous·
substitution of elements other than Al and Si in the zeolite framework. Szostak and Thomas [64]
reported the synthesis of sodalite, a small-pore zeolite structure (eight-ring), with significant
quantities of iron in the framework (Si02/Fe203 =6-30). Several aliicles describe the synthesis and
physicochemical properties of ferrisilicates belonging to the medium-pore (ten-ring) zeolites [65-
70]. An excellent review article was published by Ratnasamy and Kumar [71] describing the
ferrisilicate analogs of zeolites. However, very little is reported about the substitution of iron in
large-pore zeolites (twelve-ring), such as Beta [72] and HY [73] and ZSM-12 [71] zeolites, in the
literature. In this research work, the complete and partial isomorphous substitutions of AI-ZSM-
12 with Fe atoms were investigated.
Three factors appear to be critical in the consistent and successful preparation of
ferrisilicate molecular sieves over a wide range of Si02/Fe203 ratios: the avoidance of iron
21
hydroxide precipitation, the necessity of using low-molecular-weight silica sources and the need
to suppress the formation of iron complexes with the organic amine crystal-directing agents. Iron
. tends to precipitate as a rust-red colloidal ferric hydroxide at pH > 4 [74]. Once formed, ferric
hydroxide is almost completely insoluble, thereby limiting availability of FeO-2 species for
incorporation into the silicate units during crystal growth. This problem is not encountered in
zeolite synthesis, since any colloidal aluminium hydroxide formed redissolves more readily at the
pH necessary for zeolite crystallization [28]. The ability of iron (III) to precipitate silica species
through complex f?rmation was reported in early studies of silicate and soil chemistry [75-77].
Complexes occur between the two species at pH values of 3 and 4. Therefore, initial formation of
a ferrisilicate complex at low pH avoids precipitation of rust-red iron hydroxide. Once the iron is
complexed in this manner, the formation of ferric hydroxide at elevated pH appears to be
suppressed. The pH is then adjusted with OH- (as NaOH) to the basic conditions necessary to
promote molecular sieve crystallization. Upon addition of the silica-containing solution to the
acidic iron (III) solution, distinct ~olor changes are observed with changing pH. Below pH 3, the
solution is bright yellow, turning turbid and peach color at around pH 4. The solution rapidly
increases in opacity with disappearance of color when above pH 6. The final pH used in this
synthesis is between 8 and Il. Within this range the opaque gels are either white or pale yellow,
depending on the amount of iron used in the formation of the gel. The yellow gel disappears upon
crystallization, resulting in highly crystalline white materials. For instance, the ZSM-5 structure
crystallizes over a range of Si02/Fe203 between 15 and 200 [78].
The amount of silica needed to react with aIl the iron is dependent on the extent of
polymerization within that silica source. Sodium metasilicate (Fisher Scientific) and N-brand
silicate (PQ Corp.), both consisting of low-molecular-weight silicaspecies, produce good-quality
white to pale yellow colored gels with Si02/Fe203 as low as 5 to 8. Ferrisilicate analogues with
the sodalite structure have been crystallized in the lower Si02/Fe203 range [62]. On the other
hand, Ludox (DuPont Co.) and Cab-O-Sil (Cabot Corp.) containing 200A particles of high
molecular-weight polymeric silica beads do not react appreciably with the soluble iron (III). In
these latter cases, iron (III) can complex to the surface of the beads only. After the available sites
on the bead surface' have reacted, the iron begins to react with itself, resulting in a rust -colored
gel. Limited incorporation of iron (III) into the final crystalline material is then observed. The
22
uncoordinated iron oxide phase produces a pink- or brown-colored molecular sieve, showing the
presence of both tetrahedral (framework) and octahedral (non-framework) iron. Prolonged
digestion in NaOH, which breaks the silica polymers into smaller units, increases iron
incorporation.
The presence of a complexing template or crystal-directing agent, which binds strongly to
the iron (III) species also prevents it from forming the ferrisilicate gel. A strongl y coordinating
neutral amine template, such as pyridine, is found to limit incorporation of the iron into the gel
and the final crystalline molecular sieve structure. N oncoordinating tetrapropylammonium
bromide, free of the neutral amine impurity, does not appear to impede incorporation and is the
template of choice in crystallizing the ZSM-5 structure. However, in order to avoid any possible
complex formation between the iron and any free amine impurity present, the TP A +, added as the
bromide salt, is introduced after the ferrisilicate gel is formed [78]. Thus, the order of adding the
reactants in gel preparation of ferrisilicate is a very important factor for controlling the pH of the
gel.
The presence of the tetrahedrally ·coordinated iron in the ferrisilicate lattices can be
confirmed by a variety of techniques such as: IR spectroscopy of the symmetric (Si-O-Fe)
vibration in the framework vibration region of the crystalline ferrisilicate, expansion of the unit
cell dimensions (Iron is approximately 1.5 times larger than the silicon or aluminium ion), EPR
and M5ssbauer spectroscopy.
In this work, Fe-ZSM-12 gels were prepared using the method proposed by Ratnasamy and
Kumar [72]. In a.typical preparation of Fe-ZSM-12 [73] , a required amount of sodium silicate
(27.8 wt.% Si02, 8.8 wt.% Na20 and 63.4 wt.% H20) in 40 g H20 was added slowly under
stirring to another solution comprising specified amount of Fe(N03)3.9H20, 40 g H20 and 3.0-
4.0 g H2S04 (96-98 wt.%). To the above mixture, adequate amount of
methyltriethylammoniumchloride (MTEA-CI, Fluka) in 40 g H20 was added under vigorous
stirring. The pale lemon-coloured gel , so obtained, was stirred for about half an hour before
transferring into a teflon-lined stainless steel autoclave. The typical molar gel composition was:
Si02 / Fe20 3 = 140, Si02 / MTEA+ = 2.5 , Si02 / Na+ = 1.67, OH- / Si02 = 0.214 and H20 / Si02
= 47. The crystallization was carried out at 160°C for 5 days. The color of the crystalline product
was white. The final products were separated by centrifugation and then repeatedly washed with
23
distilled water, followed by drying in air at 70°C ovemight. Thereafter, calcination was carried
out at 600°C ovemight. Isomorphous substitution of Al with Fe atoms were performed either
completely or partially. Three gels with complete substitution of Al atoms with Si/Fe ratios of 40,
50 and 70 and two gels with partial substitution with Si/Fe ratios of 40 and 60 were prepared and
their structural and acidity· properties were studied by different characterisation techniques which
will be explained in what follows.
2.1.4 Synthesis of SSZ-42
SSZ-42 lS a high-silica large pore zeolite with an undulating, one dimensional 12-
membered T-atom ring (12R) channel system. The pore diameter at the narrowest point in the xz
projection is 6.4 A, while in the perpendicular projection (yz) it is 6.7 A. The cage at the widest
point is ca. 10 A. The chann~l exhibits side pockets, however, there are no intersecting channels
in this molecular sieve [79]. SSZ-42 was firstly synthesized by Zones and Rainis by using N
benzyl-l , 4-diazabicyclo [2.2.2] octane cation as a structure directing agent [80].. In basic media,
SSZ-42 can be synthesized only as a borosilicate, while the purely siliceous ·counterpart, ITQ-4,
can be synthesized in the presence of fluoride ions by using the same organic template [81]. SSZ-
42 can be made aluminium free using essentially aluminium 'free silicon sources. This zeolite can
also be prepared directly as a boralite or as an alumino(boro) silicate by first synthesizing a
boralite, then substituting boron with aluminium for a portion of the boron [82].
The procedure used in this work for synthesizing B-SSZ-42 was based on the method
proposed by S. Zones [82]. In a typical preparation, 3 millimoles of template N-benzyl-l ,4-
diazabicyclo octane cation was used to dissolve 0.06 g sodium borate decahydrate (Fluka) in 5.5
ml deionised water .and then 0.6 g Cabosil M5 silica was slurried into the resulting mixture. The
reaction mixture was heated in a Teflon-lined stainless steel autoclave at 150°C fof 17 days
without agitation. Afterwards, the final products were separated by centrifugation then repeatedly
washed with deionised water, followed by drying in air at 70°C ovemight. Finally, calcination at
873K was carried out ovemight. Because of the long crystallization time of the synthesis, the
reactions were seeded for larger batches. In one preparation, the reaction was seeded with SSZ-42
in the as-synthesized form to the extent of about 1 % of the silica and in this case the needed time
24
for synthesis was reduced to 4 days. The typical molar gel composition was: 0.150 R20: 0.018
Na20: 0.037 B20 3: Si02: 43.3 H20 (R signifies the template).
2.1.5 Ion exchange of SSZ-42
Ion exchange of the calcined sample was carried out using a 1/1/20 mass ratio of SSZ-42 /
ammonium acetate/water, this mixture being heated at 95°C for two hours. After cooling, the
exchanged zeolites were filtered and washed with water. Elemental analysis showed that about
84% of the Na+ cations were exchanged with H+ cations.
2.2 Characterisation
General characterisation techniques, as well as particular techniques, were utilized to
characterize AI-, Fe-ZSM-12 and B-SSZ-42 samples in comparison to reference samples of
zeolites. These include elemental analyses, nitrogen adsorption, XRD, SEM, TEM, lIB MAS
NMR, FTIR of adsorbed pyridine after desorption at different temperature (373 , 423 ,473 , 573
and 673 K), TPD of ammonia, TPD of ethylene and toluene (as probe molecules) in four different
mixture conditions: binary (toluene-ethylene), ternary (toluene-ethylene-C02), temary (toluene
ethylene-H20) and quatemary (toluene-ethylene-C02-H20). These techniques will be briefly
described below.
2.2.1 Elemental Analysis
The bulk chemical composition of the samples was determined by elemental analysis. The
normalized chemical composition can be generally represented as (MxSiy) O2, where M is the
element such as Al or Fe, whereas x and y stand for the atomic fractions of the corresponding
elements. The bulk composition of the calcined samples (practically Si/M atomic ratio = y/x) was
determined by the so called wet analysis method. A complete acid digestion prior to analysis by
inductively coupled plasma-atomic emission spectroscopy (ICP-AES) or flame atomic adsorption
spectroscopy (AAS) was required. Typically, for acid digestion a weighed amount of about 100
mg of solid sample was placed in a 100 ml polyethylene digestion bottle with a binary acid
mixture (2 ml of concentrated HF (48-49%)and. 25 ml of diluted Hel (1 M)) and 50 ml of
25
demineralized water. The bottle was then placed in an oven at 60°C and shaken overnight to
completely dissolv~ the sample. The digested sample was then transferred to an . acid-cleaned
polyethylene volumetric flask and the total volume adjusted to 100 ml by adding mineralized
water. The solution was ready for analysis for elements of interest using the appropriate method.
The AAS analysis was carried out using a Perkin-Elmer 1100B atomic adsorption spectrometer.
The ICP analyses were made using a P40 from Perkin-Elmer. AlI the solutions were analyzed by
reference to aqueous calibration standards.
2.2.2 Nitrogen Adsorption
Nitrogen adsorption is a standard tool to investigate porous materials. The sorption
isotherms of nitrogen measured at its condensation temperature (-196°C) reflect the texturaI
characteristics of the materials. In this work, the nitrogen adsorption measurements were made to
characterize the texturaI properties of the calcined samples including the total surface area BET
and the micropore volumes.
The adsorption and/or desorption isotherms of nitrogen at 77 K were· obtained using an
Omnisorp-lOO automatic analyzer after degassing the calciried samples at 300°C for at least 4 h
under vacuum 0.013-0.0013 Pa. The linear part of the Brunauer-Emmett-Teller (BET) equation
(P/po = 0.06-0.10) was used to calculate specific surface area. The t-plot method was applied to
quantitatively determirte the micropore volumes of zeolites. t was calculated using the formula
for the statistical thickness of the adsorbed nitrogen multilayer (Equation 2.1), which was
developed by Harkins and Jura in 1944 [83]. By plotting the volume of N2 adsorbed, Va, as a
function of the thickness, t, at-plot can be made. The calculation were made for a thickness in the
range 0.45< t < 1.0 nm.
13.99 t=~==========~
P (0.034 -log(- ))
. . Po
(2.1)
where P is the vapour pressure of the gas and Po is the equilibrium vapour pressure of the liquid.
Usually, the presence of micropores in the solid is indicated by a positive intercept at zero
pressure and the rnicropore volume is calculated from this intercept.
26
2.2.3 X-ray Diffraction (XRD)
For identifying and describing the structure of a particular zeolite, X-ray diffraction is the
most frequently used technique. The unique structures of the different zeolites, in terms of atom
positions and unit cell dimensions, are reflected in characteristic positions and relative intensities
of peaks observed. The XRD patterns of known structure types of zeolites and other microporous
molecular sieves are compiled in the "Collection of Simulated XRD Powder Patterns for
Zeolites" (Treacy and Higgins, 2001) [84] as weIl as in the online databases of the IZA
(www.iza-structure.org/databases). Powder X-ray ' diffraction patterns (XRD) of the as
synthesized and calcined samples were recorded using a Siemens D5000 powder diffractometer
with Cu KÀ radiation (À = 1.54184 Â). The scanning par~meters were as follows: the .28 values
from 5 to 40°, step size of 0.02° and step time of 1 s.
2.2.4 Scanning electron Microscopy (SEM)
Scanning electron micrographs were recorded to determine the crystallite Slze and
characterize the morphology of the materials, using a JEOL JSM-840A SEM operated at 15-20
kV. The powder sample was directly pasted on a metallic support with epoxy adhesive then
coated with a very thin film of gold using a high vacuum chemical vapour depositor. This was
done because zeolites are insulator materials and electron bombardment causes high
concentrations of electrical charges accumulated on the zeolite surface. This surface charging
gives a very , glowing image of the zeolite particles. The gold thin film can improve the image
clarity by conducting the electrical charge accumulated on the zeolite surfaces.
2.2.5 Transmission Electron Microscopy (TEM)
The TEM can be used to examine internaI structure and ' composition of thin, thinned, or
sectioned specimens. Convergent beam, electron diffraction provides information on crystal
structure and crystallography. The TEM images were recorded using a JEOL 2011 STEM.
27
2.2.6 Nuclear Magnetic Resonance Spectroscopy (NMR)
Solid state NMR is an effective tool for characterizing complex solids such as rnicroporous
and mesoporous materials. This technique can provide very precise information on the chemical
environment of a particular atom including bond length and angles as well as the coordination
symmetry. Solid state NMR is a complementary technique to XRD, because both crystalline and
amorphous materials can be investigated. However, while XRD provides information about the
long range ordering, NMR allows investigation on the short range ordering (local environment)
and structure.
In this regard, -correlations between lIB chemical shifts and the zeolite structure have been
well established in several works [85-88]. Il .
B MAS NMR spectra wer~ recorded at room temperature at a resonance frequency of
96.25 MHz using a Brucker ASX 300 spectrometer. lIB MAS NMR spectra were obtained with a
90° -pulse duration of 2.5 JlS, repetition time of 2s, and spinning rate of 3.5 kHz. Boron
trifluoride diethyl etherate was used as extemal reference for Il B MAS NMR analysis.
2.2.7 Fourrier Transform Infrared Spectroscopy (FTIR)
Infrared spectroscopy undoubtedly represents an important t601 in catalysis research, being
most widely used, and usually most effective, spectroscopic method for direct characterisation of
the surface structure of heterogeneous catalysts. Different FTIR techniques have been applied in
order to determine the framework vibrations by IR spectroscopy, verify the functional groups
(surface hydroxyl groups) by IR spectroscopy, probe the acidic character using IR spectroscopy
of adsorbed probe molecules (pyridine) and probe the accessibility of functional surface groups
using IR spectroscopy of adsorbed probe molecules (2,6-ditertbutylpyridine, DTBPy).
In this work, we used FTIR of adsorbed pyridine to characterize acid sites of all ZSM-12
and SSZ-42 samples as weIl as the IR spectroscopy of framework vibrations of SSZ-42 samples.
AlI FTIR spectra were recorded on a Biorad FTS-60 spectrometer connected to a personal
computer through the conventional RS-232 interface for data acquisition. Generally, the spectra
- - ------_._--------------- ----------
28
were collected in the absorbance mode at a spectral resolution of 2 cm-l. The spectra were
processed using the Galactic GRAMS/32 AI software (version 6.0).
FTIR coupled with adsorption of basic probe molecules, like pyridine, yields information
conceming the nature of acid sited (Bronsted and Lewis) while the FTIR spectra after adsorption
at different temperatures offers information about the acid strength. About 8 mg wafers of
calcined samples were first evacuated (0.013 Pa) at 300°C for 24 h. The spectra were then
recorded at room temperature. After that, the samples were exposed to 2373.1 Pa of pyridine
vapour at 24°C for 10 min. In each step of pyridine desorption, the samples weré evacuated
(0.013 Pa) at different temperatures (100, 150, 200, 300 and 400°C) for 12h before the spectra
were recorded at room temperature in the evacuated cell.
. The relative 'concentrations of Bronsted and Lewis acid sites can be determined directly
from the integrated intensities of the corresponding bands in the IR spectra after adsorption
followed by desorption of pyridine at "150°C. The molar extinction coefficients for pyridine
adsorbed on Bronsted and Lewis acid sites are 1.88 and 1.42 cm/Ilmol, respectively. In this work,
the acidity was calculated in terms of relative concentration of acid sites and expressed as
Bronsted/ Lewis ratio (Equation 2.2),
C Bronsted
C l ewis
1.88MB r2
W 1.88MB ----x " W 1.42ML r
2 1.42ML
(2.2)
where CBronsted lS the concentration of Bronsted acid sites (mmol/g catalyst), C Lewis is the
concentration of Lewis acid sites (mmol/g catalyst), M Band M L are the integrated intensities of
the IR bands (cm-1) at 1548 cm- l (Bronsted acid sites) and 1450 cm-1 (Lewis acid sites), r is the
diameter (cm) and w the weight of the wafer (mg).
2.2.8 Thermodesorption of Ammonia (TPD of Ammonia) Temperature-programmed desorption (TPD) of ammonia is a widely used method to assess
the acidity of zeolites in terms of the number of overall acid sites (Bronsted and Lewis) as weIl as
their strength. This method is based on the temperature programmed desorption in order to
monitor the changes of the surface coverage of ammonia, which is pre-adsorbed on the surface of
29
the zeolite with increasing the temperature. The rate of adsorption can be thus followed as a
function of temperature by measuring the concentration of ammonia in the gas phase over the
sample. Generally, in the TPD profile, the presence of sites of high acid strength is reflected in a
maximum of the corresponding desorption peak at a high temperature of desorption. Moreover,
the number of acid sites can be determined by measuring the integral area under the peaks.
The NH3 TPD profiles were obtained using a RXM 100 apparatus (Advanced Scientific
designs Inc.) equipped with 'an on-line mass spectrometer (MS, UTI 100) and a
thermoconductivity detector (TCD). About 0.1 g of calcined sample was placed in a U-shaped
quartz reactor between two plugs of quart~ wool. The ,sample was then pretreated at 550°C under
a flow of 20% oxygen in helium and cooled down to room temperature under the same flowing
gas. The reactor was then heated to 1000 e and pure ammonia was admitted to the reactor (1 atm)
for 15 min. Physisorbed NH3 was removed by purging with helium. Heating was then carried out
at a rate of 5°C/min until 550°C, under the same gas flow. Quantification of the response peaks
was obtained after calibration of the TC)) response curves.
2.2.9 Temperature-Programmed Desorption
Temperature-programmed desorption (TPD) was carried out in a custom-made flow device
with helium as a carrier gas. Toluene and ethylene were used as probe molecules for heavy and
light components of automobile exhaust gases, respectively. For each experiment, 10 mg of the
sample was placed between two sintered discs in a 0.14 in LD. stainless steel tube. Subsequently,
acti~ation of the sample was carried out in a flow of helium at 300°C for 2 h and the sample was
then cooled down to a temperature of 30°C. After this procedure, adsorption of the probe
molecules was performed by introducing the sorbates diluted in helium at 30°C (50 cm3STP/
min) for a period of 15 min. The partial pressures of toluene and ethylene during the adsorption
step were kept constant at 0.25 kPa. The gas phase and the weakly adsorbed sorbates on the
surface of crystals were purged with pure helium at 30°C .for 15 min.
Desorption was then performed by heating the sample in the flow ofhelium from 30 to
350°C at a linear heating rate of 20°C/min. After the final temperature was reached, desorption
was continued for a further period of 10 min to desorb any residual sorbate. Helium flow rate was
30
maintained at 50 cm3 (STP)/min. The effluent gas stream during TPD was continuously
monitored using a quadrupole mass spectrometer (Dycor Dymaxion Quadrupole MS).
For temperature-programmed desorption tests of temary and quatemary mixtures, the
ethylene (0.25%) and toluene (0.25%) binary mixture was co-adsorbed with 6% water vapour or
8% C02 in their respective temary mixtures, as well as with 14% combined water-C02, thus
forrning the quatemary mixture. Also, to ensure that the desired concentration of water vapour
was attained, the transfer line was heated to avoid partial condensation.
2.3 References
[1] R. Burch, Catal. Rev. 46 (2004) 271-333.
[2] T. Kanazawa, Catal. Today 96 (2004) 171-177.
[3] K.C. Taylor, CHEMTECH (1990) 551.
[4] F. Haab, H. Fuess, Adv. Eng. Mater. 7(2005) 899.
[5] M.E. Davis, T. Okubo, US Patent, (2005) 0166581.
[6] F. Zhao, Technologies for Near-Zero-Einission Gasoline-Powered Vehicles, SAE
International, Pennsylvania (2007).
[7] N.R. Collins, G.R. Chandler, R.J. Brisley, P.J. Andersen, P.J. Shady, S.A. Roth, SAE Paper
960799 (1996).
[8] B. Pfalzgraf, M. Rieger, G. Ottowitz, SAE Paper 960261(1996).
[9] R.J. Farrauto, R.M. Heck, Catal. Today 55 (2000) 179-187.
[10] J.E. Kubsh, SAE Paper 941996 (1994).
[11] K.P. Reddy, S.T. Gulati, D.W. Lambert, P.S. Schmidt, D.S. Weiss, SAE Paper 940782
(1994).
[12] H. Mizuno, F. Abe, S. Hashimoto, T. Kondo, SAE Paper 940466 (1994).
[13] P.F. Kuper, W. Maus, H. Swars, R. Bruck, F.-W. Kaiser, SAE Paper 940465 (1994).
[14] T. Ma, N. Collings, T. Hands, SAE Paper 920400 (1992).
[15] T. Tsoi-Hei Ma, UK patent application GB 2280128 A (1995).
[16] T. Kirchner, A. Donnerstag, A. Koénig, G. Eigenberger, in: N. Cruse, A. Frennet, J.M.
Bastin (Eds.), 4th International Congress on Catalysis and Automotive Pollution Control
(1997) p. 39.
[17] R.G. Silver, D. Dou, C.W. Kirby, R.P. Richmond, J.D.S. Balland, SAE Technical Paper
Series 972843(1997).
[18] D.S. Lafyatis, T.H. Ballinger, G. Lammey, J.C. Frost, SAE Technical Paper Series 980421
(1998).
[19] D.S. Lafyatis, G.P. Ansell, S.C. Bennette, J.C. Frost, P.J. Millington, R.R. Rajaram, A.P.
Walker, T.H. Ballinger, Appl. Catal. B: Environ.18(1998)123.
[20] T.H. Ballinger, W.A. Manning, D.S. Lafyatis, SAE Technical Paper Series 970741(1997).
[21] J. Abthoff, R. Kemmler, H. Klevin, M. Matt, H. Robota, W. Wolsing, J. Wiehl, S. Dunne,
SAE Technical Paper S.eries 980422 (1998).
31
32
[22] N.R. Burke, D.L. Trimm, R. Howe, N.W. Cant, Zeolites as exhaust e,mission Traps, in: I.E.
Aust (Ed.), Proceedings of the CHEMECA 98 Conference, port Douglas, Paper No.
87(1998).
[23] M.D. Patil, K.E. Morse, SAE Technical paper Series 980419 (1998).
[24] K. Otto, C.N. Montreuil, O. Todor, R.W. McCabs, H.S. Gandhi, Ind. Eng. Chem. Res. 30
(1991) 2333.
[25] T. Kobayashi, T. Yamada, K. Kayano, SAE Technical Paper Series 970745 (1997).
[26] E.W. KaÎser, W.O. Siegl, Y.I. Henig, R.W. Anderson, F.H. Trinker, Environ. Sci. Technol.
25 (1991) 2005.
[27] E.W. Kaiser, J.M. Andino, W.O. Siegel, R.H. Hammerle, J.W. Butler, J. A Waste Manage.
Assoc. 41(1991) 195.
[28] R.M. Barrer, Hydrothermal Chemistry o/Zeolites, Academic Press, London (1982).
[29] R. Xu, W. Pang, J. Yu, Q. Huo, J. Chen, Chemistry o/Zeolites and Related Porous
Materials, John Wiley and Sons, Singapore (2007).
[30] J.D. Sherman, Proc. Natl. Acad. Aci. 96 (1999) 3471-3478.
[31 ]E.M. Flanigen, Molecular Sieve Zeolite Technology- The First Twenty-five Years.
Proceedings o/the Fifth International Conference on Zeolites, ed. L.V.C.Rees, Heyden,
London (1980) 760-780.
[32] D.W. Breck, Zeolite Molecular Sieves, John Wiley and Sons, New York (1974).
[33] R.J. Argauer, G.R. Landolt, US Patent, 320288? (1972).
[34] Y.F. Chu, US Patent, 4847055 E1989).
[35] E.J. Rosinski, M.K. Rubin, US Patent, 3832449 (1974).
[36] J.L. Guth, H. Kessler, R. Wey, Stud. Surf. Sci. Catal. 28 (1986) 121.
[37] W. Xu, J. Dong, J.Li, W.Li, F. Wu, J.Chern. Soc. Chern. Commun. (1990) 755.
[38] D. Trong On, S. Kaliaguine, Zeolite/Mesoporous Molecular Sieve Composite Materials
(chapter 3), Nanoporous Materials (Science & Engineering) edited by G. Q. Lu, X.S. Zhao,
Published by Imperial College Press, vol. 4, 2004.
[39] D. Domine, J.Quobex, Molecular Sieves 78, Society ofChemistry Industry, London, 1968.
[40] M. Mertens, J.A. Martens, P.J. Orobet, P.A. Jacobs, Guidlines/or Mastering the Properties
0/ Molecular Sieves- Relationship between the Physicochemical Properties o/Zeolite
Systems and their Low Dimonsionality. NATO ASI, Ser. B, ed. D. Barthomeuf, E.G.
Derouane, W. Holdrich, Plenum Press, New York, London, 1 (1990) 221.
[41] M.E. Davis, R.f. Lobo,Chem. Mater. 4 (1992) 756-768.
[42] R.M. Heck, R.J. Farrauto, Appl. Catal. A-Gen. 221 (2001) 443-457.
[43] B. Bigot, V.H. Peuch, J. Phys. Chem. B, 102 (1998) 8696-8703.
[44] S.P. Elangovan, M. Ogura, Y. Zhang, N. Chino, T. Okubo, Appl. Catal. B-Environ,
57 (2005) 31-36.
[45] N.R. Burke, D.L. Trimm, R. Howe, N.W. Cant, Zeolites as exhaust emission Traps, in:
I.E. Aust (Ed.), Proceedings of the CHEMECA 98 Conference, Port Douglas, Paper
No. 87, 1998.
[46] N.R. Burke, D.L. Trimm, R.F. Howe, Appl. Catal. B-Environ, 46 (2003) 97-104.
[47] S.P. Elangovan, M. Ogura, M.E. Davis, T. Okubo, J. Phys. Cheille B, 108 (2004)
13059.
[48] K.F. Czaplewski, T.L. Reitz, YJ. Kim, R.Q. Snurr, Micropor. Mesopor. Mater. 56
(2002) 55-64.
33
[49] A. Iliyas, M.H. Zahedi-Niaki, M. Eic, S. Kaliaguine, Micropor. Mesopor. Mater. 102 (2007) 171-177.
[50] A. Iliyas, Z. Sarshar, M.H. Zahedi-Niaki, M. Eic' , S. Kaliaguine, Submitted.
[51] E.J. Rosinski and M.K. Rubin, US Patent 3,832,449 (1974).
[52] C. H. Baerlocher, W. M. Meier, D. H. OIson, Atlas of zeolite framework types, Elsevier,
Amsterdam (2001).
[53] W. Zhang and P.G. Smimiotis, Catal. Lett. 60(1999) 223- 228.
[54] P.A. Jacobs, J.A. Martens, Synthesis of High-Silica Aluminosilicate Zeolites, Elsevier,
Amsterdam (1987) 297-319.
[54] S. Ernst, P. A. Jacobs, J. A. Martens, and J. Weitkamp, Zeolites, 7 (1987) 458-462.
[56] A.Katovic and G. Giordano, Factors affecting the MTW zeolite crystallization
processin: ML. Occelli, H Kessler (Eds.) , Synthesis of Porous Materials: Zeolites,
Clays and Nanostructures, Marcel Dekker, New York (1997) 127-137.
[57] M.J. Eapen, K.S.N. Reddy and V.P. Shiralkar, Zeolites, 14 (1994) 295-302.
[58] W.F. Holderich, in P.A. Jacobs and R.A. Van Santen (Eds.), Proceedings of the 8th
International Zeolite Conference, Amsterdam, July 10-14, 1989 (Stud. Surf Sci. Catal.,
Vol. 49), Elsevier, Amsterdam (1989) p. 69.
[59] W.F. Holderich, M. Hesse and F. Naumann, Angew. Chem. Int. Ed. Engl. , 27 (1988)
226.
[60] D.R.C. Huybrechts, L de Bruycker and P.À. Jacobs, Nature, 345 (1990) 240.
[61] R.A.T. Deshmukh, 1. Reddy, B.M. Bhawal, V.P. Shiralkar and S. Rajappa, J. Chem.
Soc. Perkin Trans. l, (1990) 1217.
[62] B. Notari, in T. Inui, S. Namba and T. Tatsumi (Eds.), Proceedings of an International
Symposium on Chemistry of Microporous Crystals, Tokyo, June 1990 (Stud. SurfSci.
Catal. , Vol. 60), Kodansha, Tokyo, Elsevier, Amsterdam (1991) p. 343.
[63] M. Tielen, M. Geelen and P.A. Jacobs, Acta. Phys. Chem., 31 (1985) 1.
[64] R. Szostak and T.L Thomas, J. Chera. Soc. Chem. Commun. , 2 (1986) 113.,
[65] P. Ratnasamy, R.B. Borade, S. Sivasankar, V.P. Shiralkar, S.G. Hegde, Acta. Phys.
Chera., 31 (1985) 137.
[66] R.B. Borade, Zeolites, 7 (1987) 398.
[67] R. Szostak and T.L. Thomas, J. Catal. , 100 (1986) 555.
[68] C.T.W. Chu and C.D. Chang, 1. Phys. Chem., 89 (1985) 1569 . .
[69] R. Kumar and P. Ratnasamy, J. Catal. , 121 (1999) 89.
[70] K.G. Ione, N.D. Tien, N.B. Klyueva and LA. Vostdkova, Catal. Sci. 'Technol. , (1985)
281.
[71] P. Ratnasamy and R. Kumar, Catal. Today, 9 (1991) 238.
[72] R. Kumar, A. Thangaraj , R.N. Bhat and P. Ratnasamy, Zeolites, 10 (1990) 85.
[73] A.J. Chandwadkar, R.N. Bhat and P. Ratnasamy, Zeolites, Il (1991)42.
[74] R. Szostak and T. L. Thomas, J. Catal., submitted.
[75] B. Wichterlova and P. Jiru, React. Kinet. Catal. Lett. , 1980, 13, 197.
[76] A. Gutsze, H. Winkler and M. Zawack, Magnetic Resonance in Colloid and Interface
Science, ed. J. P. Fraissard and H. A. Resing (D. Reidel, Dordrecht, 1980), p. 467.
[77] D. K. Simmons, Master' s Thesis (Georgia Institute of Technology, 1986).
[78] R. Szostak, T. L. Thomas, J. Catal. , 100 (1986) 555-557.
[79] C.Y. Chen, L.W. Finger, R.C. Medrud, P.A. Crozier, I.Y. Chan, T.V. Harris, S.I.
Zones, Chem. Commun. 18 (1997) 1775-1776.
[80] S.I. Zones, A. Rainis, WO Patent 95/908793 (1995).
[81] M.A. Camblor, A. Corma, L.A. Villaescùsa, Chem. Commun. (1997) 749.
[82] S.I. Zones, US Patent, (1998) 5770175.
34
35
[83] W.D. Harkins, G. Jura, J. Am. Chem. Soc., 66 (1944) 1362-66.
[84] M. M. J. Treacy, J. B. Higgins, Collection of Simulated XRD Powder Patterns for Zeolites,
Fifth revised edition, Elsevier, Oxford (2007).
[85] D. Trong On, M.P. Kapoor, L. Bonneviot, S. Kaliaguine, Z. Gabelica, J. Chem. Soc. Faraday Trans. 92 (1996) 1031-1038.
[86] G. Coudurier, A. Auroux, J. C. Vedrine, R. D. Farlee, L. Abrams, R.D. Shannon, J. Card. , 108 (1987) 1.
[87] K. F. M. G. J. ScholIe, W. S. Veeman, Zeolites, 5 (1985) 118.
[88] G. L. Turner, K. A. Smith, R. J. Kirkpatrick and E. Oldfield, J .Magn. Reson. , 67 (1986), 544.
Chapter 3
MTW Zeolites for Reducing Cold-Start
Emissions of Automotive Exhaust
Z. Sarshar 1, M. H. Zahedi-Niaki ], Q. Huang 2, M. Eié 2 and S. Kaliaguine 1*
1 Department of Chemical Engineering, Université Laval, Québec, Canada, G 1 K 7P4,
2 Department of Chemical Engineering, University of New Brunswick, Fredericton, N.B.,
Canada, E3B 5A3.
Article in Press in scientific journal of Applied Catalysis OB: Environmental
(doi: 1 0.1 016/j.apcatb.2008.08.025).
36
3. MTW Zeolites for Reducing Cold-Start Emissions of Automotive Exbaust z. Sarshar l , M. H. Zahedi-Niaki l , Q. Huang 2, M. Eié 2 and S. Kaliaguine 1*
1 Department of Chemical Engineering, Université Laval, Québec, Canada, G 1 K 7P4,
2 Department of Chemical Engineering, University of New Brunswick, Fredericton, N.B. ,
Canada, E3B SA3.
Résumé
37
L' entrée en vigueur des nouvelles législations environnementales sévères conduit au besoin
l'un meilleur contrôle des émissions au démarrage à froid des moteurs à essence. Dans ce travail
de recherche, une série de tamis moléculaires présentant des canaux unidimensionnels avec des
ouvertures de pores à 12 oxygènes (12R), ayant la structure MTW (MTW est une dénomination
pour ZSM-12 dans la' nomenclature de structures des zéolithes de l ' IZA) a été synthétisée et
caractérisée par différentes techniques telles que, DRX, MEB, BET, l ' analyse élémentaire par la
spectroscopie d 'absorption atomique, FTIR de pyridine adsorbée et la désorption à température
programmée (DTP). Les échantillons synthétisés ont été examinés comme adsorbants
d'hydrocarbures (HC' s) en utilisant l'éthylène et le toluène comme molécules sondes
respectivement légère ( oléfine) et lourde ( aromatique) qui sont présentes dans les gaz
d'échappement des moteurs au démarrage à froid. Les analyses de DTP ont été effectuées après
l 'adsorption sous quatre conditions de mélanges différents: le binaire (éthylène- toluène), le
ternaire (éthylène-toluène-C02) , le ternaire (éthylène-toluène-H20) et le quaternaire (éthylène
toluène-C02-H20). Les résultats ont démontré que l ' échantillon ZSM-12 échangé à l ' argent (Ag
ZSM -12), possède des capacités de piégeage élevées et stables pour les deux sorbats dans tous les
mélanges étudiés, montrant ainsi son insensibilité par rapport au CO2 et/ou H20. De plus, une
haute température de désorption, en particulier pour le toluène, a été associée à un nombre élevé
de sites acides de Lewis et Bronsted forts. La substitution isomorphe d'Al par Fe dans les
structures MTW n'a pas produit de grands changements au niveau de la capacité d'adsorption ni
dans la température de désorption de ces deux adsorbats.
Mots-Clés: Piège à Hydrocarbures, ZSM-12, échange d'Ions, substitution isomorphe,
émissions au démarrage à froid, désorption à température programmée, diffusion "single file" ,
adsorption.
- - - - - --- - - - - - --- - -------"
----------------------------------------------- ----
38
Abstract Increasingly strict environmental legislations have led to the need for better control of
vehicle cold-start emissions. In this research work, a series of one-dimensional channel molecular
sieves with 12 oxygen ring apertures (12R) having MTW structure (MTW is the designation for
ZSM-12 in the IZA nomenclature of zeolite structures), have been synthesized and characterized
by different techniques such as: XRD, SEM, BET surface area, elemental analysis by atomic
absorption spectroscopy, FTIR of adsorbed pyridine and temperature-prograrnrned desorption
(TPD). The synthesized samples were tested as hydrocarbon (HC' s) trap adsorbents using toluene
and ethylene as heavy (aromatic) and light (olefin) probe molecules present in the exhaust stream
at engine cold-start. TPD tests were perforrned after adsorption under four different mixture
conditions: binary (toluene-ethylene), ternary (toluene-ethylene-C02), ternary (toluene-ethylene
H20) and quaternary (toluene-ethylene-C02-H20). The results demonstrated that a silver
exchanged MTW zeolite, AI-ZSM-12 (Ag) exhibited high and stable trapping capacities for both
probe molecules in aU mixtures investigated in this study, thus showing essentiaUy no sensitivity
to CO2 and/or H20. Moreover, a high desorption temperature, particularly for toluene, was
associated with the large amount of strong Lewis and Bronsted acid sites. Isomorphous
substitution of Al by Fein MTW structures did not lead to drastic changes in adsorption
capacities and desorption temperatures of the two sorbates.
Keywords: Hydrocarbon trap, ZSM-12, Ion-exchange, Isomorphous substitution, Cold-start emissions, Temperature programmed desorption, Single-file diffusion, Adsorption.
39
3.1 Introduction
Over the recent years, the low emission standards have forced automobile and catalyst
manufacturers to focus on reducing the cold-start hydrocarbon ("HC' s") emissions. Cold-start is
referring to the short time period of 1-2 minutes, after engine ignition, before the three-way
catalyst reaches its light-off temperature. During this period, about 70-80% of the total
hydrocarbons are released without .converting due to sluggish catalyst activity at low temperature.
Different approaches have been attempted to solve this problem including combustion heated
catalyst (CHC) [1] , eiectrically heated catalyst [2-5] , close-coupled catalyst [6, 7] , exhaust-gas
ignition (EGI) [8, 9] and heat storage devices but aIl these solutions encountered difficulties such
as: cost, complexity in design and fabrication and durability. It has been found that the effective
solution to the cold-start problem is to design a system employing an adsorbent which traps the
hydrocarbons temporarily, followed by their graduaI desorption from the porous structure upon
increasing exhaust temperature. For this solution, zeolites have been found preferred adsorbent
materials due to their thermal stability and their thermodynamic affinity to HC' s. The critical
factors for any emission trap are the adsorption capacity and desorption temperature which must
be higher than the catalyst light-offtemperature [10].
A series of zeolites BEA, MFI, MOR and X [11-13] and silicoaluminophosphate molecular
sieves [14] were investigated regarding their hydrocarbon adsorption capacities, under a variety
of conditions [15]. Burke et al. [16] reported that beta zeolite (BEA) is a promising material for
this application, while Elanogovan et al. [17] found that StandardOil Synthetic Zeolite-thirty
three (SSZ-33) is superior to BEA based on their adsorption studies of a series of medium and
large pore silicoaluminate zeolites. On the other hand, Czaplewski et al. [18] observed an
interesting phenomenon of trapping molecules within the one-dimensional channel of EUO and
MOR zeolites, as compared to more typical three- dimensional zeolites. This phenomenon was
. designated as the single-file diffusion mechanism, which prevents the passage of molecules by
one another inside the micropores of the one-dimensional zeolites.
In a previous work, we have studied a number of one-dimensional zeolites which could
potentially be employed as adsoJ;bents for HC traps during cold-start. ZSM-12 was selected as a
promising candidate from these preliminary screening tests, based on its high trapping capacity
and desorption temperature [19, 22]. In particular ZSM -12 ~as found to be hydrothermally stable
up to 973-1073 K in the presence of 10% water vapour in helium [22]. In the present study, the
40
influence of isomorphous metal substitutions, cation exchange and acid properties of synthesized
ZSM -12 (MTW) samples on trapping capacities and desorption temperatures of both probe
~olecules were investigated.
3.2 Experimental Details
3.2.1 AI-ZSM-12 synthesis AI-ZSM -12 was synthesized using a method proposed by Ernst et al. [20]. In a typical
preparation, 80.9 g of a sodium silicate solution (Merck, 28.5 wt% Si02, 8.8 wt% Na20 , 62.7
wt% H20 ) was mixed with 80 g of distilled water. To this, a solution of 30 g
methyltriethylammoniumchloride (Fluka) in 100 g of water was added, under stirring, followed
by adding a solution containing 2.4 g AI(N03)3.9H20 (Merck) in 30g H20. Afterwards, 4.7 g
H2S04 (Merck, 98%) was added under vigorous stirring. In the next step, the gel was transferred
into a teflon-lined stainless steel autoclave and the crystallization was conducted in a furnace at
1600 e without agitation for 7 days. The final products were separated by centrifugation and then
repeatedly washed with distilled water, followed by drying in air at 700 e overnight. Finally,
calcination at 5500 e was carried out overnight. The molar composition of the resulting gel was
calculated as Si02/Ah03 = 120, OH-/Si02 = 0.3 , Na20/ (MTEA)20= 1.5 and H20/OH- = 125.
3.2.2 Ion exchange of ZSM-12 The calcined zeolites were three-fold ion-exchanged at room temperature with 0.1 M
ammonium nitrate solution. Ag incorporation was carried out by ion-exchange in the liquid phase
using silver nitrate over H-form of zeolite samples. For this, the solution was stirred at 800 e for 2
hours. Thereafter, the samples were centrifuged and washed with distilled water and then dried
overnight at 100oe.
3.2.3 Fe-ZSM-12 Synthesis The hydrothermal synthesis of Fe-ZSM-12 was carried out by modifying the procedure
reported by Ernst et al. [21] for the synthesis ofZSM-12. In a typical preparation ofFe-ZSM-12,
a required amount of sodium silicate (27.8 wt.% Si02, 8.8 wt.% Na20 and 63.4 wt.% H20) in 40 g
H20 was added slowly under stirring to another solution comprising specified amount of
Fe(N03)3.9H20, 40 g H20 and 3.0-4.0 g H2S04 (96-98 wt.%). To the above mixture, adequate
41
amount of methyltriethylammoniumchloride (MTEA-CI, Fluka) in 40 g H20 was added under
vigorous stirring. The pale lemon-coloured gel, so obtained, was stirred for about half an hour
before transferring into a teflon-lined stainless steel autoclave. The typical molar gel composition
was: Si02 1 Fe203 = 140, Si02 1 MTEA + = 2.5 , Si02 1 Na + = 1.67, OH- 1 Si02 = 0.214 and H20 1
Si02 = 47. The crystallization was' carried out at 433 K for 5 days. The color of the crystalline
product was white. The final products were separated by centrifugation · and then repeatedly
washed with distilled water, followed by drying in air at 70°C ovemight. Thereafter, calcination
was carried out at 600°C ovemight. The gel compositions of the different samples are presented
in Table 3.1.
Table 3.1: Gel composition of synthesized samples
Sample . Acronym Si lAI Si IFe Gel composition *
r x y z m
ZSl-29 AI-ZSM-12 60 1.0 0.016 0.000 1.89 226.13
ZSl-26 Fe-ZSM-12 40 1.0 0.000 0.030 2.39 122.10
ZSl-36 Fe-ZSM-12 50 1.0 0.000 0.024 2.39 122.1 0
ZSI-24 Fe-ZSM-12 70 1.0 0.000 0.017 2.39 122.10
ZSl-28 AI-Fe-ZSM-12 40 40 1.0 0.014 0.014 2.39 122.1 0
ZSl-32 AI-Fe-ZSM-12 60 60 1.0 0.010 0.010 2.39 122.10
* : r R: x Ab03: y Fe203: z Si02: m H20 ; R = MTEACI
42
3.3 Structural Characterisation
3.3.1 XRD In X-ray diffraction patterns the unique structures of the different zeolites, in terms of atom
positions and unit cell dimensions, are reflected in characteristic positions and relative intensities
of observed peaks. Powder X-ray diffraction patterns (XRD) of the as-synthesized and calcined
samples were recorded using a Siemens D5000 powder diffractometer with Cu KÀ radiation (À =
1.54184 A).
3.3.2 SEM Scanning electron micrographs were recorded to determine the crystallite Slze and
characterize the morphology of the materials, using a JEOL JSM-840A SEM operated at 15-20
kV.
3.3.3 Elemental Analysis The bulk chemical compositions of the samples were determined by elemental analysis.
The bulk composition of the calcined samples (practically Si/M atomic ratio = y/x) was
determined by the so called wet analysis method. A complete acid digestion prior to analysis by
inductively coupled plasma-atomic emission spe.ctroscopy (ICP-AES) or flame atomic absorption
spectroscopy (AAS) was required. The AAS analysis was carried out using a Perkin-Elmer
1100B atomic absorption spectrometer. The l CP analyses were carried out using a P40
spectrometer also from Perkin-Elmer.
3.3.4 Nitrogen Adsorption The sorption isotherms of nitrogen measured at its condensation temperature (-196°C)
reflect the texturaI characteristics of the materials. In this work, the nitrogen adsorption
measurements were performed to characterize the texturaI properties of the calcined samples,
including the total BET surface area and the micropore volume.
The adsorption and/or desorption isotherms of nitrogen at 77 K were obtained uSlng an
Omnisorp-100 automatic analyzer after degassing the calcined samples at 300°C for at least 4 h
under vacuum 0.013-0.0013 Pa. The linear part of the Brunauer-Emmett-Teller (BET) equation
(P/po = 0.06-0.1.0) was used to calculate specific surface area. The t-plot method was applied to
quantitatively determine the micropore volumes of zeolites.
43
3.3.5 Fourrier Transform Infrared Spectroscopy (FTIR) FTIR coupled with adsorption of basic probe molecules like pyridine yields information
conceming the nature of acid sited (Bronsted and Lewis), while the FTIR spectra after adsorption
at different temperatures provide information about the acid strength. FTIR of adsorbed pyridine
was performed using a Biorad FTS-60 spectrometer. About 8 mg wafers of calcined samples
were first evacuated (0.013 Pa) at 300°C for 24 h. The spectra were then recorded at room
temperature. After that, the samples were exposed to 2373.1 Pa of pyridine vapour at 24 oC for 10
min. In each step of pyridine desorption, the samples were ev~cuated (0.013 Pa) at different
temperatures (100, 150, 200, 300 and 400°C) for 12h before the spectra were recorded at room
temperature in the evacuated cell.
3.3.6 Ammonia Thermodesorption (TPD of adsorbed ammonia) Ammonia thermodesorption profiles were obtained using a RXM-100 catalyst
characterisation system (ASDI) equipped with an online mass spectrometer (UTI 100) and a
thermal conductivity detector (TCD). About 0.1 g of calcined sample was placed in a U-shaped
quartz reactor between two plugs of quartz wool. The sample was then pretreated at 550°C under
a flow of 200/0 oxygen in helium and cooled down to room temperature under the same flowing
gas. The reactor was then heated to 100°C and pure ammonia was admitted to the reactor (1 atm)
for 15 min. Physisorbed NH3 was removed by.purging with helium. Heating was then carried out
at a rate of 5°C/min until 550°C, under the same gas flow. Quantification of the response peaks
was obtained after calibration of the TCD response curves.
3.3.7 Temperature-Programmed Desorption Temperature-programmed desorption ('"fPD) was carried out in a custom-made. flow device
with helium as a carrier gas. Toluene and ethylene were used as probe molecules for heavy and
light components of automobile exhaust gases, respectively. For each experiment, 10 mg of the
sample was placed between two sinter discs in a 0.14 in LD. stainless steel tube. Subsequently,
activation of the sample was carried out in a flow of helium at 300°C for 2 h, and then cooled
down to a temperature of 30°C. After this procedure, adsorption of the probe molecules was
performed by introducing the sorbates diluted with helium at 30°C (50 cm3STP/ min) for a period
of 15 min. The partial pressures of toluene and ethylene during the adsorption step were kept
- - - - - ------------
44
constant at 0.25 kPa. The gas phase and the weakly adsorbed sorbates on the surface of crystals
were purged with pure helium at 30°C for 15 min.
Desorption was then performed by heating the sample in the flow of helium from 30 to
350°C at a linear heating rate of 20°C/min. After the final temperature was reached, desorption
was continued for a further period of 10 min to desorb any residual sorbate. Helium flow rate was
maintained at 50 cm3 (STP)/min. The effluent gas stream during TPD was continuously
monitored using a quadrupole mass spectrometer (Dycor Dymaxion Quadrupole MS).
3.4. Results and Discussions
3.4.1 XRD Selected XRD patterns of calcined samples are shown in Fig. 3.1. These patterns confirm
that the samples are pure with the MTW crystal structure. The high intensity of the XRD lines
indicates that the samples are weIl crystalline materials.
3.4.2 SEM The SEM images of sorne selected samples are displayed in Fig. 3.2. The individual
crystals ofNa-ZSM-12 (Fig. 3.2A) appear to be rice-shaped with average crystal diameter of 2.7
~m, while Iliyas et al. [22] reported the spherical shape crystallite because of using
tetraethylammonium hydroxide (TEAOH) as the structure directing agent. Isomorphous
substitution of Al by Fe results in changing the morphology \ to the spherical shape or hank
shaped crystallites (Fig. 3.2B), as weIl as appearance of sorne plate-shaped crystallites (Fig.
3.2C). Figures 2 A and B show that the particles are agglomerates of very small crystals (a
fraction of micron in size). On the other hand, Na-ZSM-12 particles (Fig. 3.2A) look much more
homogeneous.
3.4.3 Elemental Analysis The results of bulk composition analyses are summarized in Table 3.2. From this table, it
can be found that comparable bulk phase chemical compositions observed for the solid products
and the corresponding gel phases.
45
Moreover, the degree of ion ex change of calcined Na-ZSM-1 2 for H+, Ag+ and Mg2+
cations in samples ZSl-29-1 , 29-2 and 29-3 were found to be 0.78, 0.96 and 0.65, respectively.
:i .!!. ~ 'u) c Cl) .... .E ëü c C')
,Ci)
2000
:i -!!. 1500
b ëi) c S .E 1000 ëü c en en
2000
1800
1600
1400
1200
1000
800
600
400
200
500
(8)
(A)
10 15 20 25 30 35 40
2theta
(0)
(C)
10 15 20 25 30 35 40
2theta
Fig. 3.1: XRD ofselected as-synthesized samples. (A) AI-ZSM-12 (Na); (B) AI-ZSM-1 2 (H); (C)
Fe-ZSM-1 2 (40)~ (D) AI-Fe-ZSM-12 (40)
,Fig. 3.2: SEM pictures of selected as-synthesized samples; (A) Na-ZSM-12; (B) Fe-ZSM-12 (40); (C) Fe-ZSM-12 (70)
46
47
Table 3.2: Results of bulk chemical composition of calcined samples by AAS
Sample Trivalent element (M) Si/M a Si/M b
ZSI-29 Al 79 60
ZSI-28 Al, Fe 40c 40 c
ZSI-28 Al, Fe 47d 40 d
ZSI-32 Al, Fe 62c 60 c
ZSl-32 Al, Fe 55d 60 d
ZSI-24 Fe 68 70
ZSI-36 Fe 49 50
ZSl-26 Fe 45 40
a: ratio in solid product, b: ratio in the gel; c: M= Al, d: M= Fe
3.4.4 Nitrogen Adsorption The measured values of specific surface area and micropore volume of the samples are
shown in Table 3.2. They reveal that AI-ZSM -12 samples in aU cationic forms (except exchanged
with Mg2+) possess high surface area (close to 300 mf/g) and micropore volume (above 0.12
cm3/g), which are desirable properties related to sorbates adsorption capacity. Substituting A13+
with Fe3+ leads to lower surface area and/or mict:opore volume probably due to extra--framework
oxide clusters that can create pore clogging in zeolites and reduce the pore volume available for
nitrogen adsorption. The higher the content of extra-framework F e3+, the lower the surface area
and micropore volume.
48
Table 3.3: TexturaI properties from nitrogeIi adsorption isotherms at -196°C
Sample* Trivalent Si /Me Ionie form Surface Micropore element Area Volume
(m2/g) (ml/g)
ZSl-29 Al 60 Na+ 342 0.131
ZSl-29-1 Al 60 H+ 324 0.123
ZSl-29-2 Al 60 Ag+ 293 0.122
ZSl-29-3 Al 60 Mg+ 257 0.099
ZSl~26 Fe 40 Na+ 283 0.120
ZSl-36 Fe 50 Na+ 290 0.122
ZSl-24 Fe 70 Na+ 311 0.123
ZSl-28 Al, Fe 40 Na+ 339 0.128
ZSl-32 Al, Fe 60 Na+ 235 0.101
* AlI samples have MTW structure.
Specifie surface area values are ~ 5m2/g whereas microp~re volumes are ~ O.OOlml/g.
3.4.5 Fourrier Transform Infrared Spectrosc.opy (FTIR) In this characterisation technique, Bronsted acid centres are detected by a band around
1548 cm-J , which originates from pyridinium ions (CsHsNH+) created via protonation of pyridine
molecules by surface acidic hydroxylgroups [23]. On the other hand, Lewis acidic centres are
characterized by bands around 1450 cm-l, originating from pyridine coordinatively linked to
Lewis acidic sites. However, bands of physisorbed pyridine are expected in the range close to
Lewis acidic centres, i.e. , around 1443 cm-1• These additional bands may be removed at higher
desorption temperature, e.g. , above 150°C. The measured amounts of both acid sites decreased as
. pyridine desorption temperature increased and finally no pyridine desorbing from acidic sites was
observed at 400°C for aIl sampI es.
The concentration ratios of Bronsted to Lewis acid sites for all samples are presented in
Table 3.4. H-ZSM-12 has the highest measured CBrbnsted/CLewis and this ratio decreases in the
order of: H+ > Na+ > Mg2+ >Ag+. Ag-ZSM-12 has the largest amount of Lewis acid sites (the
T
49
lowest ratio) in comparison to the other samples. The F.e-ZSM-12 sample with Sil Fe = 40 also
has a high Lewis acid site content. The 1448 cm- l peak area, related to Lewis sites, decreases
with increasing Si/Fe ratio in the Fe substituted samples, whereas CBrônsted/CLewis ratio increases.
Consequently, the higher the Sil Fe ratio in Fe-ZSM-12 samples, the higher the ratio
CBrônsted/CLewis. This ratio is increased by partial substitution' of Fe by Al into the framework. For
instance in the case of AI-Fe-ZSM-12 (40), this ratio is 2.9, whereas for Fe-ZSM-12 (40) it is 1.9.
Thus, the Si/M ratio (M is a trivalent element), the nature of the trivalent element and the type of
counter-cation are the most important factors that influence the nature and quantity of the acid
sites which in tum control the adsorption/desorption behaviour of the zeolites. Typical
chemisorbed pyridine FTIR spectra of Na-ZSM-12 and Ag-ZSM-12 at different desorption
temperatures are shown in Fig. 3.3.
Table 3.4: FTIR results of chemisorbed pyridine (desorption at I ~ O°C)
S~mple
Na-ZSM-12
H-ZSM-12
Ag-ZSM-12
Mg-ZSM-12
Fe-ZSM-12(40)
Fe-ZSM-12(50)
Fe-ZSM-12(70)
AI-Fe-ZSM-12 (40)
Al-Fe-ZSM-12 (60)
Peak area (a.u.) 1448 cm- l
0.010
0.012
0.471
0.120
0.272
0.213
0.078
0.228
0.107
*calculated as A1550 x 1.88 1 Al448x 1.~2
Peak area (a.u.) 1550 cm- l
0.347
0.732
0.183
0.220
· 0.398
0.521
0.335
0.498
0.403
C Brônsted/CLewis *
45.9
80.8
0.5
2.4
1.9
3.2
5.7
2.9
5.0
0,7 L+B L Physisorbed
B
/ pyridine
0,6 100 C
150 C
Q) 0,5
0 c 200 C Q) .c 0 0,4 CI) .c 300 C <x:
0,3
0,2
1600 1550 1500 1450 1400
-1 Wavenumber (cm )
0,7
~ L
0,6
Q) 0,5
0 c Q) .c 0 0,4 CI) .c <x:
0,3
before pyridine adsorption
0,2
1600 1550 1500 1450 1400 1 -1
Wavenumber (cm )
Fig. 3.3: FTIR spectra of (A) Na-ZSM-12; (B) Ag-ZSM-12, before and after pyridine chemisorption at different temperatures (B and L denote Br6nstead
and Lewis sites, respectively)
50
51
3.4.6 Ammonia Thermodesorption (TPD of adsorbed ammonia) The TPD re~ults of adsorbed ammonia of aIl ZSM-12 samples are shown in Table 3.5,
while a typical NH3-TPD profile of Ag-ZSM-12 sample is shown in Fig. 3.4. Three distinct
regions are observed in the TPD profiles." These peaks can be designated as 1, 2 and 3 and
correspond to weak, medium and strong acid centers in zeolites, respectively.
1t is also seen from this table that the amount of adsorbed ammonia in the low temperature
region of the desorption profile, which can be assigned to Lewis acid sites, increases in the order:
Ag+> H+> Mg2+> Na+ for AI-ZSM-12 and Si/Fe =40> 50 >70 for Fe-ZSM-12 samples. In fact,
Ag-ZSM-12 and Fe-ZSM-12 (40) have the largest ~mounts of Lewis acid sites among the Al and -
Fe samples, respectively. The amount of adsorbed ammonia in the medium temperature region of
desorption profile can be assigned either to Lewis or Bronsted acid sites, depending on the
samples. Finally, the amount of adsorbed ammonia in the high temperature region of the profile,
which may be attributed to strong Bronsted acid sites, increases in the order: H+ > Na + > Ag + >
Mg2+. It can also be seen that by increasing the Sil Fe ratio for Fe-ZSM-12 samples, the amount
of these strong acid sites increases.
Comparing these peak temperatures ln the three distinct reglons can lead to a clear
assessment regarding the acid strength of 'the sites. G:enerally, it can be seen that the ammonia
desorption temperatures (from NH3-TPD profiles) are higher for AI-ZSM-12 samples than for Fe
ZSM -12 samples. This corresponds to the fact that Al as a trivalent element in the framework of
zeolites can generate stronger acid sites than Fe. Among AI-ZSM-12 samples, the one having Ag
counter-ions shows the highest desorption temperature in the low temperature region (200 OC)
and thus stronger Lewis acid sites compared to Na + cations. The strongest Bronsted acid sites
correspond to NH3 desorbing from proton sites in Na-ZSM-12 and H-ZSM-12 at 395 and 380 oC,
respectively.
AlI the above conclusions are confirmed by the FTIR of chemisorbed pyridine. Comparing
the NH3-TPD for Fe-ZSM-12 samples indicates a systematic decrease in the amounts of NH3
desorbed at the lowest temperature (T peak]) as the Si/Fe ratio increases. At the same time, the
amounts of ammonia desorbed as the other two peaks (peaks 2 and 3) both increase. This would
be coherent with peak 1 being ascribed to Lewis acid sites (presumably Na+ cations), whereas
peaks 2 and 3 could be related to Bronsted acid sites of different · acid strengths. Thedensity of
both Bronsted sites would thus ·increase by increasing Si/Fe ratio. This suggests that the Na+ ~H+
52
ex change equilibrium is displaced toward a higher surface H+ density with the change in acid
strength associated with the decreased density of lattice iron. Again aH the above conclusions are
coherent with the increase in C Bronsted/CLewis ratios reported in Table 3.4.
The observed desorption temperatures and desorbed NH3 amounts ln Al-Fe ZSM-12
samples are found to be intermediate values of AI-ZSM-12 and Fe-ZSM-12 results.
Table 3.5: NH3-TPD results
Sample Tpeak 1 NH3 NH3 Tpeak2 NH3 NH3 Tpeak3 NH3 NH3
(OC) moles/gcat moles/m2 (OC) moles/gcat moles/m2 (OC) moles/gcat moles/m2
* 1 05 * 1 07 *105 *107 *105 *107
Na-ZSM-12 160 2.9 0.8 212 27.4 8.0 395 190.9 55.8
H-ZSM-12 143 24.3 7.5 310 70.2 21.7 380 218.1 67.3
Ag-ZSM-12 200 34.5 11.8 263 9.5 32.4 375 70.3 24.0
Mg-ZSM-12 118 10.2 4.0 141 21.9 8.5 193 . 30.l 11.7
Fe-ZSM-12 117 37.7 18.5 175 8.3 4.2 230 11.9 4.2 (40)
Fe-ZSM-12 133 10.8 3.7 175 13.8 4.7 242 15.7 5.4 (50)
Fe-ZSM-12 145 5.1 1.6 175 15.l 4.9 245 17.3 5.6 (70)
Ai-Fe- 143 20.6 6.l 205 47.8 14.1 320 34.2 1.0 ZSM-12
(40)
AI-Fe- 173 19.2 8.2 220 30.1 12.8 320 54.6 23.3 ZSM-12
(60)
0.6~-------------------"T----__ --.
0.5
0.4
0.3
0.2
/
/ /
ulli ' ,4 . •
. 1 ~ 2 ~ 3 :
Tem perature
- - - experimental - - - - --Peak 1 -- --- -- Peak 2 - -- -- Peak 3
Sum of eaks
~ . - .-.-: ... .. ------_ .- .. ------- -- - -- ---------------
100 150 200 250 300 350 400 450 500 550
Temperature (C)
Fig. 3.4: Ammonia thermodesorption profile of Ag-ZSM-12
3.4.7 Temperature-Programmed Desorption
3.4.7.1 Binary Toluene-Ethylene mixture
53
Representative TPD profiles obtained over selected samples are presented in Fig. 3.5 and
the results are summarized in Table 3.6. Several factors, such as mode of metal substitution, acid
properties, framework or extra-framework ~pecies in the pore lattice, extent of metal su~stituti6n
and/or distribution have to be considered to understand the observed trapping capacities and
profiles [19]. It can be observed from these results that the desorption profiles of ethylene and
toluene are homothetic, compatible with the single file diffusion mechanism, since the smaIler,
less-strongly adsorbed molecule (ethylene) is desorbed simultaneously with the larger, more
strongly adsorbed one (toluene) [19, 22]. For aIl the toluene profiles, there are two desorption
54
peaks, as shown in Fig. 5. The high temperature peak is more important in the case ofNa-ZSM-
12 (Fig. 3.5A) than in the other samples, which could he attributed to stronger interactions of
toluene with the strong Bronsted acid sites present in this sample. A dominant desorption peak is
obse~ed for ethylene in the temperature range 113 to 161°C in aIl samples, except in the cases
of Na and H-ZSM-12. These materials exhibit two desorption peaks, the second ones being
observed at 282 and 215°C, respectively. The main desorption peak of ethylene in the
temperature range of 113 to 161°C is similar to the first desorption peak of toluene in the range of
138-160°C, particularly in the case of Fe samples (Fig. 3.5D). The second high temperature
desorption peak of toluene, over the temperature range of 183-311 oC, can be attributed to
specific strong interaction of tolueJ?e molecules with stronger acid sites.
In the case of Ag-ZSM-12, the high desorption temperature of the second peak reaches up
to 300°C due to strong interactions between toluené and Ag trapping centres. Thus, this
hydrocarbon can be retained up to temperatures weIl above the light -off temperature of the
catalytic muffler.
As noted above the second toluene peak is most pronounced for Na-ZSM-12 sample
(Fig. 3.5 A), around 282 oC, which is also weIl above the light-off temperature of the muffler
catalyst.
It can be noticed that a clear relation exists between desorption temperature and the degree
of substitution of F e3+ in iron substituted samples. By decreasing the amount of F e3
+ in these
samples, the desorption temperatures of both ethy lene and toluene increase due to appearance of
progressively stronger Bronsted acid sites (Table 3.6). For instance, the higher desorption peak of
toluene in the case of Si/F e = 70 reaches 284°C.
The total numbers of moles of ethylene and toluene desorbed were converted into adsorbed
volumes assuming molar volumes of 0.105 and 0.075 ml /mmol for toluene and ethylene,
respectively. The sums of these values are estimates of total sorbate volume. They are plotted
against micropore volume in Fig. 3.6 in a same manner as in ref. [19]. The curv.e shown in
Fig. 3.6 was established from our previous studies of about 30 different zeolites and AIPO' s. It is
reported here for sake of a global comparison with these previous data.
AlI the samples have micropore volume> 0.10 ml/go The micropore volume values on this
graph are close to each other since aIl samples have the same MTW structure. It is clear that H
ZSM -12 has the highest adsorption capacity among these samples and this capacity decreases in
55
the order: H+>Na+>Ag+>Mg2+ for Al based ZSM-12. It is obvious from Fig. 3.6 that by
increasing the Sil Fe ratio in the samples, the adsorption capacity is also increased. This is due to
creation of stronger acid sites, as established by FTIR results of adsorbed pyridine and
thermodesorption of ammonia, and as discussed in the previous section.
Table 3.6: Toluene-ethylene binary TPD results
Sample* Desorbed Molar Desorption (trivalent amounts ratio temperature element) (mmol/g) Toluene/ (OC)
ethylene
Ethylene Toluene Ethylene Toluene
Na-ZSM- 0.104 0.778 7.5 138; 282 138; 282 12 (Al)
H-ZSM-12 0.045 1.001 22.2 130; 215 142; 183 (AI)
Ag-ZSM- 0.108 0.626 5.8 125 150; 311 12 (AI)
Mg-ZSM- 0.069 0.340 4.9 148 153 ; 248 12 (Al)
Na-ZSM-12 0.037 0.388 10.5 113 147;262 (Fe, 40)
Na-ZSM-12 0.042 0.440 10.5 147 149;270 (Fe, 50)
Na-ZSM-12 0.047 0.559 11.9 . 161 160;284 (Fe, 70)
Na-ZSM-12 0.045 0.551 12.2 122 140; 272 (Al, Fe, 40)
Na-ZSM-12 0.029 0.339 11.7 152 157;276 (Al , Fe, 60)
*: The samples have MTW structure with 12R aperture, pore diameter 5.6x6.0 A and non-intersecting 1 D pore lattice.
~ "Ci) c: .s .: '0 Cl)
.~ ni E 0 Z
~ "Ci) c: .s .s '0 Cl)
.~ ni E 0 Z
30~------------------------------------------~
20
10
30
20
10
30
20
10
Œ2J f\ t \
1 ~
l \ 1
,/ 1
(a)
200 250
Temperature (C)
~
(b)
f'\ (a)
150 200 250
Temperature (C)
~
Temperature (C)
300 350 400
300 350 400
350 400
56
57
80
~ 70
~ ~ 60 IJIÎ
'u; 1 ~ c: .s 50 .: , \ "C
40 i Cl)
,~ 1 \ cu , • (b) E 30 1 '\ 0
z , 20
10
100 150 200 250 300 350 400
Temperature (C)
80
~. 70
~ 60 'u; c: ~ 50 ,S "C
40 Cl)
,~ ~ E 30 ~
0 Z
20
10
100 150 200 250 300 350 400
Temperature (C)
Fig. 3.5: Toluenelethylene binary mixture TPD profiles of selected as-synthesized samples. (A) Na-ZSM-12; (B) H-ZSM-12; (C) Ag-ZSM-12; (D) Fe-ZSM-12 (70);
(E) AI-Fe-ZSM-12 (40).
0/14 -,-----------------'---------,-------, ,
0,12
~ ]. 0,1 "'C al .c ~
~ "'C ~ 0,08 c al
>-.c .... l.iJ t 0,06 c ~ ::J
~ Ô 0,04 al E ~
"0 >
0,02 ./ , ,
./
, 1 ,
./
, , , , ./
, ./
./ ./
1 , ,
./
./ , , ,
, 1
, ./
./ ./
./ ./ ,
./ , ./
" 1
Fe-ZSM-12(40)
Mg-ZSM-12 Â . $1ft)
AI-F e-ZS~1--
12(60)
./ ./ ,
./ ,
./ , , , ./
1
• Na-ZSM-12
o ~----~--~---=~----~----~----~----~--~-----!
o 0,02 0 ,04 0,06 0,08 0,1 0 ,12 0,,14 0,16 0,18
MicroporevoJume {mile)
Fig. 3.6: Volume of the binary (ethylene + toluene) adsorbed phase as a function ofmicropore volume
58
The two samples AI-Fe-ZSM-12 (60) and AI-ZSM-12 (Mg) have low adsorption capacities
probably due to extra-framework oxidic material that causes pore blocking and reduces the
trapping capacity of the sorbates. In fact, the various sorbates used in this work may not access
the same pores, so that the micropore volume determined from nitrogen physisorption may not be
entirely accessible to the hydrocarbon probes used in this study.
The molar ratio of adsorbed toluene/ethylene is reported in Table 3.6. The AI-ZSM-12
samples, except for the one in H+ form, display low values of this ratio in the range of 5-7.5. For
instance, in the case of Ag-ZSM7" 12, only one ethylene molecule is trapped by 6 toluene
59
molecules in its micropores. The small relative content of ~thylene compared to toluene is indeed
surprising since ethylene should diffuse faster within the micropore structure. This result is likely
associated with the pore aperture diameter of ZSM-12 (5.6 x 6.0 A) which might be too large to be
completely obstructed by the toluene molecule. Thus the penetration of toluene might force the
ethylene molecules out of the pore lattice, making the single-file diffusion process less efficient.
For iron substituted samples, this ratio varies in the range of 10.5-12.2 indicating that in these
cases the amount of ethylene trapped per molecule of toluene is even lower. This can be
explained by the effect of isomorphous substitution of F e3+ into the framework of zeolites, which
leads to modification in morphology and pore expa!lsion. Indeed introducing a F e-O bond
(1.84 A) in place of a AI-O bond (1.73 A) should slightly enlarge the pore aperture and therefore
make the single-file diffusion ev en less efficient.
Besides toluene and ethylene, which are representative of aromatic and olefin components
ln automobile exhaust, other components such as H20 and CO2 are present in high
concentrations. It is thus reasonable to collect, investigate and compare the TPD profiles of the
samples loaded in different mixture conditions: binary toluene/ethylene mixture, ternary
toluene/ethylene/C02 mixture, ternary toluene/ethylene/H20 mixture and quatemary
toluene/ethylene/C021H20 mixture. The TPD profiles and the results of desorption temperature
and adsorption capacities of the samples in the case of binary adsorption are presented in Fig. 3.5
and Table 3.6. For the other cases, the results are summarized in Fig. 3.7 and 3.8 discussed
below.
3.4.7.2 Ternary Toluene-Ethylene-C02 mixture From the TPD profiles of all samples in the presence of CO2, it has been found ' that the
adsorption capacities of both ethylene and toluene decrease. This reduction in trapping capacities
is explained by the competitive adsorption of CO2 molecules, which can occupy a portion of
hydrocarbon trapping sites. This relative decline in adsorption capacities of both probe molecules
is remarkably lower in the case of Ag-ZSM-12.
These differences in sensitivity to CO2 suggest that adsorption sites of different nature
and/or different energetics of adsorption are implemented in these materials. CO2 is believed to
interact with bases. In zeolites, the negatively charged oxygen neighbouring the cation is
considered a basic site [24]. A stronger Lewis acid, such as the ones observed by NH3 TPD in
60
Ag-ZSM-12, corresponds to a weaker base, .thus explaining the minor variation in adsorption
capacity upon adsorption of CO2 observed for toluene and ethylene in this solid.
In the case of Fe-ZSM-12, it is seen from Fig. 3.7 and 3.8 that in binary ethylene/toluene
adsorption, both ethylene and toluene adsorption capacities increase at increasing SilFe ratio.
Thus, the comparison with NH3 TPD data (Table 3.5) suggests that in this case the two
hydrocarbons are predominantly adsorbed on Bronsted acid sites. The competitive adsorption
effect of CO2 seems however to decrease at increasing Si/Fe. This in tum suggests competitive
adsorption of CO2 on the ensemble of the Lewis acid and conjugated lattice base. Thus, at least in
this case, both Bronsted and Lewis acids constitute adsorption sites for the two hydrocarbons.
3.4.7.3 Ternary Toluene-Ethylene-H20 mixture The same trend of reduction in adsorption capacities of both sorbates has been obtained for
aIl samples in the presence of H20. This reduction is ev en larger than in the presence of CO2.
Here again, a smaller reduction in toluene and ethylene capacity was found for Ag-ZSM-12
compared to the other zeolites. Very drastic reduction in both toluene and ethylene capacity was
systematically obtained on aIl Fe-ZSM-12 samples.
3.4.7.4 Quaternary Toluene-Ethylene-COr H20 mixture This experiment is a close simulation of the conditions prevailing at the exhaust of an
automobile engine. From the TPD profiles, it was found that the trapping capacities of aIl
samples in the presence of both CO2 and H20 are reduced, but the degree of reduction is again
not similar for aIl. samples. The decline in adsorption capacities of both toluene and ethylene in
the case of Ag-ZSM-12 is not very significant contrary to H-ZSM-12 and almost aIl the Fe-ZSM-
12 samples.
Indeed, the significant competitive adsorption of water results from the fact . that both
Bronsted and Lewis acid sites can adsorb water. The strong interaction corresponding to
hydration of the proton implies a strong water adsorption on Bronsted acid sites. This may be an
explanation for the lower sensitivity of hydrocarbon adsorption on Ag-ZSM-12 as discussed
above. Indeed, the dominant hydrocarbon adsorption site would be the Lewis acid in this solid.
0,12
0,1
0 ,08 UI dry
0,06 il C02
0,04 =wet
1"" wet-C02 0,02
o
Samples
0,12
€ 0,1 u (\S Cl. (\S u 0,08
III dry 0)-c:J2' '50 (5
0/06 Cl.E ~E ...,-
=wet
ltt C02
1· a> 0,04 c: + wet-C0 2 a> ~ ..c: 0,02 ..., W
0
~~ <o~ ~~~ :1,'=
~ ~' 1t-C3 ~ «rd «r<f
Samples
Fig. 3.7: Comparison oftrapping capacities of ethylene in gas mixtures under different conditions: binary: dry; temary with CO2: CO2; temary with H20: wet; quatemary: wet-C02
61
1
0,9
0,5
0,4
0,3
0,2
0,1
o ooF--== Na*ZSM-12 H-ZSM-12 Ag-ZSM-12 Mg-ZSM-12
Samples
0 ,7 //'------------------------
0 ,6 f-'".-------------=~--"---
l:) :~.·~lË ~1;,~-i2 Samples
62
III dry
- C02
=wet
-+ wet-C02
'''dry
mC 02
E wet
I wet-C 0 2
Fig. 3.8: Comparison oftrapping capacities oftoluene in gas mixtures under different conditions: binary: dry; ternary with CO2 : ·C02; ternary with H20: wet; quaternary: wet-C02
63
3.5. Conclusion A systematic study of 12R microporous zeolites with 1 D channel lattice as hydrocarbon
trap adsorbents for ethylene/toluene under the four conditions of binary, temary (C02 or H20 )
and quatemary mixture in presence of both H20 and CO2 has been performed. These one
dimensional zeolites have the same MTW structure but they are differing in sorne parameters
such as: trivalent elernent (M = Al or/and Fe), Si/M ratio in their lattice, and the nature of cations
employed for synthesis and ion exchange (Na, H, Ag and Mg). The role of these factors in
modifying the properties of microporous zeolites acting as adsorbents for reducing the cold-start
emissions of automobiles has been investigated.
The single file diffusion mechanism operated more efficiently in the microporous channels
of the AI-ZSM-12 samples, exc'ept for AI-ZS~-12 (H) compared to Fe-ZSM-12 samples.
It was moreover found that Fe-ZSM-12 samples were not effective adsorbents for trappirig
ethylene/toluene owing to their low rate of hydrocarbon adsorption. Their ethylene and toluene
adsorption capacities and desorption temperature were not found to be high enough for trapping
these hydrocarbons efficiently during the few minutes of the heating time of the exhaust gases
during automotive cold-start. Moreover, their adsorption capacity decreases considerably in the
presence ofH20 and,or/ CO2 .
It was shown that AI-ZSM-12 samples possess higher adsorption capacities and desorption
temperatures for both ethylene and toluene. In this case, Na-ZSM-12 demonstrates a desorption
temperature around 282°C for both ethylene and toluene due to its strong Bronsted acid sites that
create effective interactions with the probe molecules. In the presence of CO2 and/or H20 , the
adsorption capacity of AI-ZSM-12 samples is recluced but not to such large extent as in the case
ofFe-ZSM-12.
On the other hand, Ag-ZSM-12 sample showed an unexpected behaviour under aIl
conditions of TPD experiments. The adsorption capacities of both ethylene and toluene were not
significantly changed and showed to be relatively insensitive to the presenc~ of CO2 and H20. In
this case, desorption temperatures of around 311°C and 400°C for toluene were indeed reached in
TPD tests for binary mixture of hydrocarbons and quatemary mixture of hydrocarbons,
respectively. These results were explained by the dominating role of strong Lewis acid sites
present in this adsorbent. Consequently, this solid could be considered as a promising adsorbent
for controlling cold-start emissions.
64
Acknowledgements The authors gratefully acknowledge N aturai Sciences and Engineering Research Councii of
Canada (NSERC) for providing financiai support for this study.
3.6 References [1] T. Kirchner, A. Donnerstag, A. Koenig, G. Eigenberger, in: N. Cruse, A. Frennet, J.M.
Bastin (Eds.), 4th International Congress on Catalysis and Automotive Pollution
Control, 1997: p. 39.
[2] J.E. Kubsh, SAE Paper, 1994.941996.
[3] K.P. Reddy, S.T. Gu1ati, D.W. Lambert, P.S. Schmidt, D.S. Weiss, SAE Paper, 1994.
940782.
[4] H. MizlillO, F. Abe, S. Hashimoto, T. Kondo, SAE Paper, 1994.940466.
[5] P.F. Kuper, W. Maus, H. Swars, R. Bruck, F.W. Kaiser, SAE Paper, 1994.940465.
[6] N.R. Collins, G.R. Chandler, R.J. Brisley, P.J. Andersen, P.J. Shady, S.A. Roth, SAE
Paper, 1996. 960799.
[7] B. Pfalzgraf, M. Rieger, G. Ottowitz, SAE Paper, 1996.960261.
[8] T. Ma, N.Collings, T. Hands, SAE Paper, 1992.920400.
[9] T. Tsoi-Hei Ma, UK patent application GB 2280128 A, 1995.
[10] S.P. Elangovan, M. Ogura, S. Ernst, M. Hartmann, S. Tontisirin, M.E. Davis, T. Okubo, Micropor. Mesopor. Mater. 96 (2006) 210-215.
[11] D.S. Lafyatis, G.P. Ansell, S.C. Bennett, J.C. Frost, P.l. Millington, R.R. Rajaram, A.P. Walker, T.H. Ballinger, Appl. Catal. B-Environ, 18 (1998) 123-135.
[12] R.M. Heck, R.J. Farrauto, Appl. Catal. A-Gen. 221 (2001) 443-457.
[13] B. Bigot, V.H .. Peuch, J. Phys. Chem. B, 102 (1998) 8696-8703.
[14] S.P. Elangovan, M. Ogura, Y. Zhang, N. Chino, T. Okubo, Appl. Catal. B-Environ, 57 (2005) 31-36.
[15] N.R. Burke, D.L. Trimm, R. Howe, N.W. Cant, Zeolites as exhaust emission Traps, in:
I.E. Aust (Ed.), Proceedings of the CHEMECA 98 Conference, Port Douglas, Paper
No. 87, 1998.
[16] N.R. Bruke, D.L. Trimm, R.F. Howe, Appl. Catal. B-Environ, 46 (2003) 97-104.
[17] S.P. Elangovan, M. Ogura, M.E. Davis, T. Okubo, J. Phys. Chem. B, 108 (2004)
13059.
[18] K.F. Czaplewski, T .. L. Reitz, Y.J. Kim, R.Q. Snurr, Micropor. Mesopor. Mater. 56
(2002) 55-64.
65
[19] A. Iliyas, Z. Sarshar, M.H. Zahedi-Niaki, M. Eic' , S. Kaliaguine, Submitted.
[20] S. Ernst, P.A. Jacobs, J.A. Martens, J. Weitkamp, Zeolites, 7 (1997) 458-462.
[21] P. Ratnasamy, R. Kumar, Catal. Today, 9 (1991) 329-416.
66
[22] A. Iliyas, M.H. Zahedi-Niaki, M. Eic, S. Kaliaguine, Micropor. Mesopor. Mater. 102 (2007) 171-177.
[23] J. Pérez-Ramirez, J.C. Groen, A. Brückner, M.S. Kumar, U. Bentrup, M.N. Debbagh, L.A. Villaescusa, J. Catal. 232 (2005) 318-334.
[24] M. Huang, A. Adnot, S. Kaliaguine, J. Catal. 137 (1992) 322-332.
Chapter 4
Synthesis, Structural and Acidity Characterisations of the Large-Pore Zeolite SSZ-42 for Controlling Cold-Start Emissions
z. Sarshar 1, M. H. Zahedi-Niaki 1, Q. Huang 2 and, S. Kaliaguine 1*
67
IDepartment ofChemical Engineering, Université Laval, Québec, Canada, G1K 7P4,
2Department of Chemical Engineering, University of New Brunswick, Fredericton, N.B. , Canada,
E3B SA3.
Article in press in scientific journal Microporous and Mesoporous Materials
(doi: 1 0.1 016/j.micromeso.2008.09.023).
68
4. Synthesis, Structural and Acidity Characterisations of the Large-Pore Zeolite SSZ-42 for Controlling Cold-Start Emissions z. Sarshar l, M. H. Zahedi-Niaki l, Q. Huang 2 and S. Kaliaguine 1 *
IDepartment of Chemical Engineering, Université Laval, Québec, Canada, G 1K 7P4,
2Department of Chemical Engineering, University of New Brunswick, Fredericton, "N.B., Canada,
E3B 5A3.
Résumé Na-SSZ-42 (Si/B = 14) a été synthétisé et l'ion Na+ a été échangé par H+ pour obtenir H-SSZ-42,
et les deux échantillons ont été examinés en tant qu'adsorbants pour piège à hydrocarbures pour
la réduction des émissions produites lors du démarrage à froid. Cette investigation a été effectuée
par la technique de désorption à température programmée utilisant l'éthylène et le toluène
respectivement comme molécules sondes légères (oléfine) et lourdes (aromatiques) présentes "
dans les gaz d'échappement des moteurs au démarrage à froid. Ces matériaux ont été aussi
caractérisés par une variété de techniques dont FTIR de pyridine adsorbée ainsi que la
thermodésorption de l'ammoniac pour investiguer leur acidités ainsi que par DRX, MEB, lIB
MAS RMN et BET pour étudier leur propriétés structurelles. La thermodésorption de
l'ammoniac combinée avec FTIR de pyridine a indiqué des sites acides faibles dans ces deux
matériaux en raison de la présence d'atomes de bore comme élément trivalent dans le réseau. La
lIB MAS RMN a confirmé que la plupart des atomes de bore sont en coordination tétraédriques
(B04) dans le réseau. Ces deux matériaux ont aussi démontré une capacité d'adsorption
extrêmement élevée pour les deux sorbats, ainsi qu'un comportement très particulier en présence
d'eau et/ou CO2. En effet, la capacité d'adsorption de ces deux adsorbats n'est pas été
significativement affectée par la présence de molécules de H20 et de CO2 adsorbées.
Mots-Clés: Zéolithes, SSZ-42, démarrage à Froid, adsorption, désorption à température programmée, FTIR de Pyridine chimisorbée, poches latérales, diffusion" single file".
69
Abstract Na-SSZ-42 (a boralite with Si/B =14) was synthesized and ion-exchanged to obtain H-SSZ-
42. They were both tested as hydrocarbon traps for reducing cold-start emissions. This
investigation was carried out by the temperature-programmed desorption technique using toluene
as a heavy (aromatic) and ethylene as a light (olefin) probe molecule present in the exhaust
stream at engine cold-start. These materials were also characterized by a variety of techniques
including FTIR of adsorbed pyridine and thermodesorption of ammonia to investigate its acidity
and XRD, SEM, lIB NMR and BET to study its structural properties. The thermodesorption of
ammonia along with FTIR of adsorbed pyridine indicated acid sites with low strength in these
two materials due to presence of boron atoms as trivalent element in their lattice. lIB NMR
confirmed that most of the boron was incorporated as tetrahedral B04 units in the framework.
These two rnaterials showed remarkably high adsorption capacity for both ethylene and toluene,
with very special behaviour in presence of H20 and/or CO2 which exist in exhaust gases. Indeed,
the adsorption capacity of both sorbates was not significantly affected by the presence of H20
and CO2 molecules.
Keywords: Zeolite, SSZ-42, Cold-Start, Adsorption, Temperature Programmed Desorption, FTIR of Cbemisorbed Pyridine, Side Pockets, Single-file Diffusion
-- ---- ----------------------------------~
70
4.1 Introduction SSZ-42 is a high-silica large pore zeolite with an undulating, one dimensional 12-
membered T-atom ring (12R) channel system. The pore diameter at the narrowest point in the xz
projection is 6.4 À, while in the perpendicular projection (yz) it is 6.7 À. The cage at the widest
point is ca. 10 À. The channel exhibits side pockets however there are no intersecting channels in
this molecular sieve [1]. Moreover, this zeolite was shown to be thermally and hydrothermally
stable up to at least 800 oC [2].
SSZ-42 was firstly synthesized by Zones and Rainis by using N-benzyl-1 , 4-diazabicyclo
[2.2.2] octane cation as a structure directing agent [3]. In basic media, SSZ-42 can be synthesized
only as a borosilicate, while the purely siliceous counterpart, ITQ-4, can be synthesized · in the
presence of fluoride ions by using the same organic template [4]. SSZ-42 can be ma~e aluminium
free using essentially aluminium free silicon sources. This zeolite can also be prepared directly as
a boralite or as an alumino(boro) silicate by first synthesizing a boralite and then substituting
boron with aluminium for a portion of the boron [5].
Cold-start emissions of vehicles equipped with three-way catalysts is still remaining as an
environmental problem that compel automobile and catalyst manufacturers to focus on
controlling the cold-start HC emissions. The cold-start phase is referring to the first 1-2 minutes
after engine ignition in which about 70-80% of the total hydrocarbons are released. A three-way
catalyst must reach a relatively high temperature (light-off temperature), usually above 170°C
. (for fresh catalyst), before its efficiency rises to a level needed to meet emissions standards [6].
Among aIl the suggested solutions for cold-start problem, the hydrocarbon traps appear as an
effective method to control the emissions of automobiles. Hydrocarbon trapping aims at
adsorbing hydrocarbons from the exhaust stream when the temperature of the catalytic system is
low and releasing them when the temperature is raised to the light-off point. For this solution,
zeolites have been proposed as adsorbent materials due to their stability under elevated
temperatures and their" thermodynamic affinity to HC's. A series of zeolites including BEA, MFI,
MOR, X, SSZ-33, ZSM-12 and silicoaluminophosphate molecular sieves have been investigated
for their hydrocarbon adsorption capacity under a variety of conditions [7-14]. In a previous
work, it was found that Ag-ZSM-12 was a promising adsorbent for this application due to its low
sensitivity to H20 and CO2 which are essentially present in the exhaust stream of automobiles
[15]. However, it is needed to find an adsorbent that possesses a higher adsorption capacity. In
~---- ._----
71
this work, SSZ-42 was selected as a candidate for investigation of its adsorption capacity and
desorption temperature of both probe molecules (toluene · and ethylene) as hydrocarbon trap
during cold-start. The structural and acid properties of this material in both Na and H forms were
thus investigated.
4.2 Experimental Details The procedure for synthesizing B-SSZ-42 was based on the method proposed by S. Zones
[5]. In a typical preparation, 3 millimoles of template N-benzyl-l ,4-diazabicyclo octane cation
was used to dissolve 0.06 g sodium borate decahydrate (Fluka) in 5.5 ml deionis~d water and
then 0.6 g Cabosil M5 silica was slurried into the resulting mixture. The reaction mixture was
heated in a Teflon-lined stainless steel autoclave at 150°C for 17 days without agitation.
Afterwards, the final products were separated by centrifugation and then repeatedly washed with
deionised water, followed by drying in air at 70°C overnight. Finally, calcination at 600°C was
carried out overnight. Because of the long crystallization time of the synthesis, the reactions wère
seeded for larger batches. In one preparation, the reaction was seeded with SSZ-42 in the as
synthesized form to the extent of about 1 % of the silica and in this case the needed time for
synthesis was reduced to 4 days. The typical molar gel composition was 0.150 R20: 0.018 Na20:
0.037 B20 3 : Si02: 43.3 H20 (R signifies the template).
Ion exchange of the calcined sample was carried out using a 1/1/20 mass ratio of SSZ-42 /
ammonium acetate/water, this mixture being heated at 95°C for two hours. After cooling, the
exchanged zeolites was filtered and washed with water. Elemental analysis showed that about
84 % of the Na + cations were exchanged with H+ cations.
Powder X-ray diffraction patterns (XRD) of the as-synthesized and calcined samples were
recorded using a Siemens D5000 powder diffractometer with Cu KÀ radiation (À = 1.54184 Â).
Scanning electron micrographs were recorded to determine the crystallite size and
characterize the motphology of the materials, using a JEOL JSM-840A SEM operated at
15-20kV.
The TEM can be used to examine internaI structure and composition of thin, thinned, or
sectioned specimens. Convergent beam electron diffraction provides information on crystal
structure and crystallography. The TEM images were recorded using a JEOL 2011. STEM.
72
The nitrogen adsorption measurements were performed to characterize the texturaI
properties of the calcined samples, including the totalBET surface area and the micropore
volume. The adsorption and/or desorption isotherms of nitrogen at -196 oC were obtained using
an Omnisorp-1 00 automatic analyzer after degassing the calcined samples at 300°C for at least
4h under vacuum (0.013-0.0013 Pa). The linear part of the Brunauer-Emmett-TelIer (BET)
equation (PlPo = 0.06-0.10) was used to calculate specific surface area. The t-plot method was
applied to determine the micropore volume of zeolites.
St~ctural information of the synthesized samples using the FTIR spectroscopy of zeolite
framework vibrations in the mid-IR region of 400-1400 cm- l was colIected. FTIR, coupled with
adsorptionldesorption of basic probe molecules like pyridine yielding information conceming the
nature/strength of acid sited (Bronsted and Lewis), was also carried out. For this technique, about
8 mg wafers of calcined samples were first evacuated (0.013 Pa) at 300 oC for 24 h. The spectra
were then recorded at room temperature. The samples were then exposed to 17.8 torr of pyridine
vapour at 24 oC for 10 min. In each step of pyridine desorption, the samples were evacuated
(0.013 Pa) at different temperatures (100, 150, 200, 300 and 400 OC) for 12h before the spectra
were recorded at room temperature in the evacuated celle AlI these FTIR characterisations were
performed using a Biorad FTS-60 spectrometer.
Ammonia thermodesorption profiles were · obtained uSlng a RXM-100 catalyst
characterisation system (ASDI) equipped with an online mass spectrometer (UTI 100) and a
thermal conductivity detector (TCD). About 0.1 g of calcined sample was placed in a U-shaped
quartz reactor between two plugs of quartz wool. The sample was then pretreated at 550°C under
a flow of 200/0 oxygen in helium and cooled down to room temperature under the same flowing
gas. The rèactor was then heated to 100 oC and pure ammonia was admitted to the reactor (1 atm)
for 15 min. Physisorbed NH3 was removed by purging with helium. Heating was then carried out
at a rate of 5°C/min until 823 K, under the sarne gas flow. Quantification of the response peaks
was obtained after calibration of the TCD response curves.
]]B MAS NMR spectra were recorded at room temperature at a resonance frequency of
96.25 MHz using a Brucker ASX 300 spectrometer. I]B MAS NMR spectra were obtained with a
90° -pulse duration of 2.5 ~s , repetition time of 2s, and spinning rate of 3.5 kHz. Boron
trifluoride diethyl etherate was used as external reference for]] B MAS NMR analysis.
73
Temperature-programmed desorption (TPD) was carried out in a custom-made flow device
with helium as a carrier gas. Toluene and ethylene were used as probe molecules for heavy and
light components of automobile exhaust gases, respectively. For each experiment, 10 mg of the
sample was placed between two sinter discs in a 0.14 in LD. stainless steel tube. Subsequently,
activation of the sample was carried out in a flow of heliurri at 300 oC for 2 h, and then cooled
down to a temperature of 30 oC. After this procedure, adsorption of the probe molecules was
perforrned by introducing the sorbates diluted with helium at 30 oC (50 cm3STP/ min) for a
period of 2 hours. The partial pressures of toluene and ethylene during the adsorption step were
kept constant at 0.25 kPa. The gas phase and the .weakly adsorbed sorbates on the surface of
crystals were purged with pure helium at 30 oC for 15 min.
Desorption was then perforrn~d by heating the sample in the flow ofhelium from 30 to
350°C at a linear heating rate of 20 OC/min. After ·the final temperature was reached, desorption
was continued for a further period of 10 min to desorb any residual sorbate. Helium flow rate was
maintained at 5Q cm3 (STP)/min. The effluent gas stream during TPD was continuously
monitored using a quadrupole mass spectrometer (Dycor. Dymaxion Quadrupole MS).
For temperature-programmed desorption tests of ternary and quaternary mixtures, the
ethylene (0.25%) and toluene (0.250/0) binary mixture was co-adsorbed with 6% water vapour or
8% C02 in their respective ternary mixtures, as weIl as with 14% combined water-C02, thus
forming the quaternary mixture. AIso, to ensure that the desired concentration of water vapour
was attained, the transfer line was heated to avoid a partial condensation.
4.3 Results and Discussions
4.3.1 XRD The XRD patterns confirrn that the samples are pure with the IFR crystal structure. XRD
pattern of the calcined H-SSZ-42 sample is shown in Fig. 4.1 along with the reference (siliceous
ITQ-4) one related to IFR structure. The high intensity of the XRD lines indicates that the
samples are weIl crystalline materials. The unit cell dimensions were calculated using the \.
CR YSFlRE 2000 software. It was shown that this zeolite has monoclinic structure with C2/m
symmetry. The calculated a, band c were 18.5623, 13.4092 and 7.5970 A respectively while the
unit cell parameters reported for the ITQ-4 reference mentioned were 18.6524, 13.4958 and
74
7.631Â. The calculated angles (p,a ,y) were 101.461°, 90° and 90° respectively whereas they
were 101.978°, 90° and 90° for the reference mentioned. Thus, the unit cell dimension was
smaller for B-SSZ-42 compared to purely siliceous ITQ-4. This can be due to B-O distance
which is 1.39 Â while for Si-O bond length is 1.61 Â.
2500~-------------------------------------------
)( 0 .3
2000
::l 1500 ~ ~ -Ci) c: Cl) 1000 +'" c:
500
o
5 10 15 20 25 30 35 40
2 theta
Fig. 4.1: XRD of selected samples. (A) reference (siliceous ITQ-4); (B) H-SSZ-42.
4.3.2 SEM and TEM The SEM images of calcined samples are displayed in Fig. 4.2. The morphology of this
zeolite particles comprised of cylindrical (elongated) crystallites with average crystal diameter of
about 1 0 ~m. These are analogous in shape to those reported in the literature [16] which however
were only 1-2 ~m long. The tiny particles on the surface of the principal crystals, as can be seen
from Fig (4.2-b), were thought likely to be sm aIl crystals of SSZ-42 and confirmed by TEM
images of H-SSZ-42 (Fig. 4.3).
75
4.3.3 Nitrogen Adsorption . The .results of BET surface area and micropore volume of the samples are shown in Table
4.1. As can be observed from this Table, the surface area and micropore volume of these two
samples are about 500 m2/g and 0.19 ml/g respectively. These results indicate that SSZ-42 could
be an appropriate adsorbent due to its high micropore volume which can lead to high adsorption
capacity.
Fig. 4.2: SEM pictures of as-synthesized samples; (A, B) Na-SSZ-42; (C, D) H-SSZ-42
76
Fig. 4.3: TEM images ~nd diffraction patterns ofH-SSZ-42
77
Table 4.1: TexturaI properties from nitrogen adsorption isotherms at 77 K
Sample* Trivalent Si /Me Ionic form Surface Micropore element Are a Volume
(m2/g) (ml/g)
ZS 1-37-1 B 14 Na+ 495 0.193
ZS 1-37-2 B 14 H+ 502 0.187
* The samples have IFR structure.
4.3.4 FTIR
4.3.4.1 FTIR spectroscopy offramework vibrations IR spectra in the mid-IR region of 400-1400 cm-lof the .as-synthesized and calcined
samples are reported in Fig. 4.4. The weIl defined IR bands at 800 and 527cm-1 and the saturated
one at nearly 1012 cm-1 (in the 1000-1300 cm-l region) are characteristic ofSi04 tetrahedron units
[17]. The intensity of these bands decreases in the case of H-SSZ-42 compared to Na-SSZ-42.
The asymmetric B-O stretching band which is usually around 1100 cm-l , shifted to lower
frequencies upon increasing the degree of isomorphous substitution. The presence of B is
revealed by the appearance of bands at 920, 715 and 670 cm-1• The 1380 cm-1 band observed in
the case of as-synthesized Na-SSZ-42 was assigned to tricoordinated framework boron and the
intensity of this band decreases after calcination of the as-synthesized sam pIe which is consistent
with the literature [20]. The band at 920 cm-l can be attributed to the presence of the
tetracoordinated framework boron [18-20]. This band decreases in the case of calcined samples
indicating that by 'Calcination sorne of the boron atoms were removed from tetracoordinated
framework position. From Fig. (4.4a) and (4.4b), it can be seen that in the case ofNa-SSZ-42, the
bands of the calcined sample were generally shifted to higher frequencies in comparison to as
synthesized sample.
c o Ou; tn
"e tn c cu
1400 1300 1200 1100 1000 900 800 700 600 500 400
-1 Wavenumber (cm )
~ 1380
(d)
(c)
1400 1300 1200 1100 1000 900 800 700 600 500 400
-1 Wavenumber (cm )
Fig. 4.4: IR spectra of the samples: (a) as-synthesized Na-SSZ-42; (b) calcined Na-SSZ-42; (c) as-synthesized H~SSZ-42; (d) calcined H-SSZ-42.
78
79
4.3.4.2 FTIR spectroscopy of adsorhed pyridine The FTIR spectra of the samples after adsorption/desorption of pyridine are shown in
Fig. 4.5. Bronsted acid centres are detected by a band around 1548 cm-l , which originates from
pyridinium ions (CsHsNH+) created via protonation of pyridine molecules by surface acidic
hydroxyl groups [21]. On the other hand, Lewis acid centres are characterized by bands around
1455 cm-l , originating from pyridine coordinatively linked to Lewis acidic sites. Physisorbed
pyridine was observed around 1444 cm-1 in the case ofNa-SSZ-42. This band is then removed by
increasing the temperature to 150°C, however no characteristic band contributed to physisorbed
pyridine was observed in the case of H-SSZ-42. The measured amounts of both acid sites
decreased as pyridine desorption temperature increased and finally no pyridine desorbing from
acidic sites was observed at 200°C for both samples.
The concentration ratios of Bronsted to Lewis acid sites for aH samples are presented in
Table 4.2. It can be observed from this table and Fig. 4.5 that the amounts of Lewis acid sites in
both Na-SSZ-42 and H-SSZ-42 are very minor. As expected, the concentration of Bronsted acid
sites in H-SSZ-42 is significantly higher than in Na-SSZ-42 sample.
Table 4.2: FTIR results of chemisorbed pyridine (desorption at 150 OC)
Sample
Na-SSZ-42
H-SSZ-42
Surface area (a. u.) 1455 cm;-I
0.020
0.018
*calcuhtted as A1SSOx 1.88 / A1448x 1.42
Surface area (a. u.) 1548 cm-I
0.291
0.603
C Bronsted/CLewis *
19.3
44.4
(1) (.) c: (1) .c '-0 t/) .c c:x:
Cl) U t: Cl)
.c ~
o t/) .c oC(
0,0
-0,2
-0,4
-0,6
0,2.----------------------"'T"""'T""T'I
0,0
-0,2 200 oC
300°C
400°C
-0,4 before pyridine adsorption
1700 1650
~ 100°C
150°C
200°C
300°C
400°C
before pyridine adsorption
physisorbed pyridine ___
B
1600 1550
-1 Wavenumber (cm )
B+L
1500 1450
1700 1650 1600 1550 1500 1450
-1 Wavenumber (cm )
Fig. 4.5: FTIR spectra of (a) Na-SSZ-42; (h) H-SSZ-42, hefore and after pyridine chemisorption at different temperatures (B and L denote
Bronstead and Lewis sites, respectively)
80
81
4.3.5 Ammonia Thermodesorption (TPD of adsorbed ammonia) The TPD results of adsorbed ammonia of the samples are shown in Table 4.3 , while NH3-
TPD profiles of both samples a!e shown in Fig. 4.6. Three regions are observed in TPD profiles
and the peak in each region are designated as 1, 2 and 3 which correspond to acid sites present in
the samples with different strength.
The amount of adsorbed ammonia in the low temperature region of desorption profiles
which can be assigned to Lewis acid sites is higher in the case ofNa-SSZ-42 in comparison to H
SSZ-42, whereas the amount of adsorbed ammonia in the medium temperature region of the
desorption profile, which can be assigned to Bronsted acid sites, is more significant in the case of
H-SSZ-42. The high temperature peak in third region can be associated to desorption of ammonia
linked to strong acid sites on extraframework boron species. The amount of adsorbed ammonia in
this region is slightly larger in the case of H-SSZ-42, which is probably due to more
extraframework boron in the lattice being created upon ion-ex change of Na + with H+ cations.
Comparing these peak temperatures in these three regions can lead to a clear assessment
regarding the acid strength of the sites. From Table 4.3 , it can be observed that in the case of Na
SSZ-42, the temperature peaks in the three regions are higher in comparison to H-SSZ-42. This
indicates that both Lewis and Bronsted acid sites, as weIl as the strong acid sites on
extraframework boron species, are stronger in the case ofNa-SSZ-42.
AlI the above conclusions are in agreement with the FTIR of chemisorbed pyridine results.
As discussed in literature for other high-silica zeolites [17-19] , trivalent boron in isomorphous
substitution in SSZ-42 generate acid sites of much lower strength than the other trivalent
elements such as Al, Fe, Ga.
Table 4.3: NH3- TPD results
Sample T peak ] NH} NH} T peak2 NH} NH} Tpeak3 NH} NH3
moles/m2 moles/m2 (OC) moles/gcat moles/m2 (OC) moles/gcat (OC) moles/gcat
X ] 05 x107 X] 05
X ] 07 X 105
X 107
Na-SSZ- 151 28.7 5.8 176 ]3.8 2.8 226 14.0 2.8 42
H-SSZ- 124 11.0 2.2 149 32.0 6.4 178 ] 6.5 3.3 42
0,8~------------------------------------------------________________________________________ ~
0,7 - - - - Experimental . ..... Peak 1 _.- ._. Peak 2 - .. - .. - Peak 3
-Sum
0,4
/ -,/ ' . - - _'-'-.f: :_" - " . ... ">'~'_ ' _'_ ' _' _'_''':'~---=-=~---'''~~-'''''';;'''---f
0,3+-~-'~-'----~~~~~--~~~ __ ~~----~~~ __ -4
50 100 150 200 250 300 350 400 450 500 550
Temperature (C)
0,8 ,----------------------------------------------------_________________________ ---.,
0,7
0,6
0,5
0,4
- - 1
- - - - Experimental .. .... Peak 1 _._. _. Peak 2 _ .. _ .. - Peak 3
-----Sum
. . . .. . ,_. -. -. - _.=- -......;::..---=-::- -=-_"":"' _____ --_-_ -__ ~_:__"":"' _____ -::_~""""""-=-..;;.....---------__1
0,3 +--.--~~-~~--.----r-~-.----__r__~_.---.-_~~___y_--__.-.....---l 50 100 150 200 250 300 350 400 450 500 550
Temperature (C)
Fig. 4.6: Ammonia thermodesorption profile of (a) Na-SSZ-42; (h) H-SSZ-42.
82
83
4.3.6 lIB MAS NMR The lIB MAS NMR spectra of the calcined samples are shown in Fig. 4.7. A single peak
observed at ,..., -3.0 ppm for Na-SSZ-42 sample confirms that most of the boron has been
incorporated as tetrahedral B04 units in the framework. Two broad lines of mu ch weaker
intensity around -2 ppm correspond to framework-linked trigonal boron. In the case of H-SSZ-
42, the intensity of the -3 ppm line representing the tetrahedral framework boron is decreased and
it is slightly shifted to -2.9 ppm, while the trigonal framework boron lines are relatively
increased, with an extra hump appearing around 2 ppm (Fig. 4. 7b). The làtter can be attributed to
extraframework trigonal boron. The small shift from -3.0 to -2.9 ppm for the tetrahedral boron
signal after ion-exchange with H+ cations c.an be explained by the change in boron chemical
environment [17-20].
8 6 4
+2
2 o o
-2
-3.0
(b)
(a)
-4 -6 -8
Fig.4.7: 11 B MAS NMR spectra ofcalcined samples: (a) Na-SSZ-42; (b) H-SSZ-42.
84 ·
4.3.7 Temperature-Programmed Desorption
4.3.7.1 Binary Toluene-Ethylene mixtures TPD profiles obtained over both Na-SSZ-42 and H-SSZ-42 samples are presented in
Fig. 4.8 and the results are summarized in Table 4.4. It can be observed from these results that thè
single-file diffusion mechanism is operating in these samples since the smaIler, less-strongly
adsorbed molecule (ethylene) is desorbed simultaneously with the larger and presumably and
more-strongly adsorbed one (toluene) [14, 22]. It is seen from Fig. 4.8 that aIl these profiles
(ethylene or· toluene) comprise only one peak. In the case of Na-SSZ-42, toluene and ethylene
desorb simultaneously at 157°C while the desorption temperature in the case of H-SSZ-42 is
slightly lower (153°C) for both probe molecules. The relatively low desorption temperature of
both sorbates is indeed associated to the weak acid sites created by boron atoms in the zeolite
structure as, confirmed by thermodesorption of ammonia and FTIR of adsorbed pyridine. For
example we recently conducted a study on toluene-ethylene thermodesorption in Al and Fe
substituted ZSM-12 zeolite (MTW) showing much stronger acid sites than the B substituted SSZ- .
42 [14]..
Table 4.4: Toluene-ethylene binary TPD results
Sample Desorbed Molar Desorption (trivalent amounts ratio temperature element) (mmol/g) Toluenel (OC)
ethylene
Ethylene Toluene Ethylene Toluene
Na-SSZ-42 0.081 0.748 9.2 157 157
H-SSZ-42 0.164 1.147 7.0 154 153
85
250
f\(bl 200
~ : \ 0ëi) III il 150 III •
c: " , Q) • :Ë
, III .. Z • • " .
" Q) . " " o~ 100 . il
co . ; : E . • "- : • 0 : z .
l1li il : . 50 • . \
(a)
0 50 100 150 200 250 300 350 400
Temperature (C)
250
(\l 200
~ Ou; c 150 Q) ... / \ .E
"'C Cl)
o~ ca 100 1 \ E • • "- • • • 0 • Z • • •
50 : : · • •
0 50 100 150 200 250 300 350 400
Temperature (C)
Fig. 4.8: Toluene/ethylene binary mixture TPD profiles of selected samples. (A) Na-SSZ-42; (B)
H-SSZ-42.
86
It can be observed from Table 4.4 that the adsorption capacity for both ethylene and toluene
is considerably high, especially in the case of H-SSZ-42. In fact, this large adsorption capacity of
SSZ-42 is due to large volume of the straight channels in the structure of this zeolite [1]. The
higher adsorption capacity of H-SSZ-42 (1.15 and 0.16 mmol/g for toluene and ethylene
respectively) may be due to the smaller size of cation H+ in comparison to Na+, which thus can
provides a larger available pore volume for" adsorption of toluene and ethylene.
The total numbers of moles of ethylene and toluene desorbed were converted into adsorbed
volumes assuming molar volumes of -0.105 and 0.075 ml Immol for toluene and ethylene,
respectively. The sums of these values are estimates of total sorbate volume. They are plotted
against micropore volume in Fig. 4.9 in a same manner as in ref. [14, 15]. The S-shaped curve
shown in Fig. 4.9 was obtained from a systematic study of above 30 different 10R and 12R
straight channels zeolites, in exactly the same adsorptionlTPD conditions as the SSZ-42 data
reported here [14, 15]. In addition ta the SSZ-42 data, we report also the data for mordenite [14]
and FAPO-36 [15]. These two solids show micropore volumes close to the one of SSZ-42 and
they are the only ones to show significant deviations for the S-shâped curve. This particular
behaviour was associated to the presence of side pockets in the straight channels ofthese zeolites.
These pockets with 6R aperture are accessible to nitrogen (used in the estimation of the
micropore volume) but not accessible to toluene. The horizontal distance between the data point
and S-shaped curve is thus considered an estimate of the sicle pocket specifie volume. From
Fig. 4.9, these estimates are 0.03, 0.06, 0.07 and 0.085cm3/g for F APO-36, H-SSZ-42, Na-SSZ-
42 and mordenite, respectively.
The molar ratio of adsorbed toluene/ethylene is reported in Table 4.4. In the case of H-SSZ-
42, only one ethylene molecule is trapped by 7 toluene molecules in its micropores. The small
relative content of ethylene compared to toluene is indeed surprising since ethylene should
diffuse faster within the micropore structure. This result is likely associated with the pore
aperture diameter of ZSM-12 (6.2 x 7.2Â) which might be too large ta be completely obstructed
by the toluene molecule. Thus, the penetration of toluene might force the ethylene molecules out
of the pore lattice, making the single-file diffusio~ process less efficient.
0,18
0,16
0,14
:ê ::êo,12 "'C CI)
.c o VI
"'C ro
~ 0,1 CI)
>. .s::. W + Q)
~,08 :J
~ -o CI)
E ~.06 >
0.04
0.02
o o 0,05 0,1
i ~ 1 ~ : 1 1 -- 1 Il
1 _-; -- -- -- ---4.- .- .
0,15
Pore volume mllg
FAPO-36
H-SSZ-42
Na-SSZ-42
Mordenite
0.2 0 25
Fig. 4.9: Volume of the binary (ethylene + toluene) adsorbed phase as a function
of micropore volume
87
88
4.3.7.2 Ternary and Quaternary Toluene-Ethylene- C02 1H20 mixtures The effect of co-adsorbed H20 and/or CO2, which are present in high concentration in
automobile exhaust, along with both ethylene and toluene on the adsorption capacity -of Na and
H-SSZ-42 were investigated. The TPD profiles of the samples loaded in different mixture
conditions: binary toluene/ethylene mixture, temary toluene/ethylene/C02 mixture, temary
toluene/ethylenelH20 mixture and quatemary toluene/ethylene/C02/H20 mixture were
established.
From the TPD profiles of the samples in the presence of co-adsorbed CO2 or H20 , it has
been -found that the adsorption capaciti~s of both ethylene and toluene remained almost
unchanged. Such a property would be a significant advantage of these materials as trap
adsorbents in the cold-start control. Contrary to almost aIl zeolites investigated in our laboratory
for reducing cold-start emissions, which had their adsorption capacity for hydrocarbons strongly
influenced by co-adsorption of CO2 or H20 , SSZ-42 in both forms (Na+ and H+) exhibited a very
stable behaviour. This behaviour is related with the very weak Bronsted and Lewis acid sites
associated with the isomorphous substitution ofboron in the zeolite lattice. Fig 4.10 compares the
adsorption capacity of both samples in the four cases of binary, temary and quatemary
adsorption.
1,4
.~ 1,2
0 (0
1 c.. (0 0 Oc; 0,8 r:::::: .- 0 &:E CO E 0,.6 ~
+0# -(J) r:: Cl) 0,4 ::::s 0 1-
0,2
0
0,18
0)6
~ 0, 14 (.) ('0 a. CO 0!12 (J
0) ....... c ~ 0/1 '0. (5 0.. E ('0
E 0,08 ~ .... -<U c 0,06 (1)
~ ..s:: 0,04 LU
0102
0
dry C02 w et
Samples
dry C02 wet
Samples
w et-C02
wet .. C0 2
ENa-SSZ-42
- H-SSZ-42
:: Na .. SSZ-42
• H-SSZ-42
89
Fig. 4.10: Comparison of trapping capacities of ethylene/toluene in gas mixtures under
different conditions: binary (dry); ternary with CO2 (C02) ; ternary with H20 (wet) and quaternary( wet -C02) .
90
4.4 C,onclusion This work investigated the synthesis, properties, physicochemical and acidïc
characterisation of the high-silica large-pore zeolite SSZ-42 which has an undulating, one
dimensional 12-membered T-atom ring (12R) channel system. Ion-exchanged was carried out on
Na-SSZ-42 to obtain H-SSZ-42. These materials showed high BET surface area (",500 m2/g ) and
micropore volume ("" 0.2 ml/g ). llB MAS NMR confirmed that the boron was incorporated as
tetrahedral B04 units in the framework of zeolites. The results obtained by FTIR of adsorbed
pyridine and thermodesorption of ammonia indicated that these materials comprise weak acid
sites due to the presence of B as the trivalent element in their frameworks. H-SSZ-42 showed a
higher content of Bronsted acid sites while the acidic sites of Na-SSZ-42 exhibited a slightly
higher strength in comparison to H-SSZ-42.
The capability of the synthesized SSZ-42 samples as adsorbents for the application of
reducing cold-start emissions was also examined through TPD technique of ethylene and toluene
as probe molecules. TPD tests of both samples revealed that Na and H-SSZ-42 possess
considerably high adsorption capacities of the two sorbates.
It was also shown that these materials have extremely stable adsorption behaviour in
presence of H20 and/or CO2 as their trapping capacities for hydrocarbons remained unchanged
when the TPD tests were performed in the presence of co-adsorbed H20 and CO2. The desorption
temperature of both probe molecules were however relatively low (around 160 OC) owing to the
low acid strength of their acid sites.
Acknowledgements The , authors gratefully acknowledge Natural Sciences and Engineering Research Council of
Canada (NSERC) for providing financial support for this study. We thank Dr. Stacey 1. Zones
for his cooperation in providing the template material for synthesis.
· 4.5 References [1] C.Y. Chen, L.W. Finger, R.C. Medrud, P.A. Crozier, LY. Chan, T.V. Harris, S.L Zones,
Chem. Commun. 18 (1997) 1775-1776.
[2] C.Y. Chen, L.W. Finger, R.C. Medrud, C.L. Kibby, P.A. Crozier, LY. Chan, T.V. Harris, L.W. Beck, S.L Zones, Chem. Eur. 1. 4 (1998) 1312-1323.
[3].S.L Zones, A. Rainis, WO Patent 95/908793 (1995).
[4] M.A. Camblor, A. Corma, L.A. Villaescusa, Chem. Commun. (1997) 749.
[5] S.L Zones, US Patent, (1998) 5770175.
[6] M.E. Duvis, T. Okubo, US Patent, (2005) 0166581.
[7] D.S. Lafyatis, G.P. Ansell, S.C. Bennett, J.C. Frost, P.J. Millington, R.R. Rajaram, A.P. Walker, T.H. Ballinger, Appl. Catal. B-Environ, 18 (1998) 123-135.
[8] R.M. Heck, R.J. Farrau~o , Appl. Catal. A-Gen. 221 (2001) 443-457.
[9] B. Bigot, V.H. Peuch, J. Phys. Chem. B, 102 (1998) 8696-8703. .
[10] S.P. Elangovan, M. Ogura, Y. Zhang, N. Chino, T. Okubo, Appl. Catal. B-Environ, 57 (2005) 31-36.
[11] N.R. Bruke, D.L. Trimm,. R.F. Howe, Appl. Catal. B-Environ, 46 (2003) 97-104.
[12] S.P. Elangovan, M. Ogura, M.E. Davis, T. Okubo, J. Phys. Chem. B, 108 (2004) 13059.
[13] K.F. Czaplewski, T.L. Reitz, Y.J. Kim, R.Q. Snurr, Micropor. Mesopor. Mater. 56 (2002) 55-64.
[14] A. Iliyas, Z. Sarshar, M.H. Zahedi-Niaki, M. Eic ', S. Kaliaguine, Submitted.
[15] Z. Sarshar, M. H. Zahedi-Niaki, Q. Huang, M. Eié and S. Kaliaguine, Submitted.
[16] C.Y. Chen, L.W. Finger, R.C. Medrud, P.A. Crozier, LY. Chan, T.V. Harris, S.L Zones, Chem. Commun. 4 (1998) 1312-1323.
[17] E. Dumitriu, D. Trong On, S. Kaliaguine, J. Catal. 170 (1997) 150-160.
[18] D. Trong On, S. Kaliaguine, L .. Bonneviot, J. Catal. 157 (1995) 235-243.
91
[19] D. Trong On, M.P. Kapoor, L. Bonneviot, S. Kaliaguine, Z. Gabelica, J . Chem. Soc. Faraday Trans. 92 (1996) 1031-1038.
[20] R. de Ruiter, A.P.M. Kentgens, J. Grootendorst, J.C. Jansen, H. van Bekkum, Zeolites, 13 (1993) 128-138.
[21] J. Pérez-Ramîrez, lC. Groen, A. Brückner, M.S. Kumar, U. Bentrup, M.N. Debbagh, L.A. Villaescusa, J. Catal. 232 (2005) 318-334.
92
[22] A. Iliyas, M.R. Zahedi-Niaki, M. Eic, S. Kaliaguine, Micropor. Mesopor. Mater. 102 (2007) 171-177.
93
5. General Conclusions and Recommendations
5.1 General Conclusions Aluminosilicate and ferrisilicate crystalIine ZSM -12 zeolites having MTW structure and
one-dimensional, non-intersecting channel s, with 12R pore apertures were synthesized. AlI the
ZSM-12 zeolites were prepared in Na+ form, then the aluminosilicate samples were ion
exchanged to obtain H+, Ag + and Mg2+ forms. In the second phase of this work, boralite SSZ-42
zeolite was prepared in sodium form and ion-exchanged ta yield the H form ofB-SSZ-42.
Characterisation of aIl samples was first accompli shed to assess their chemical, texturaI and
structural properties. The techniques, which were applied, encompassed elemental analysis,
nitrogen adsorption, XRD, FTIR of framework vibrations and SEM. B-SSZ-42 samples were also
chara~terized by TEM and lIB MAS NMR. The acidic properties of the zeolites were
characterized using FTIR of adsorbed pyridine and thermodesorption of ammonia. AlI these
characterisation methods were applied to direct us to potential zeolites that can be further
investigated for the main goal of this work, which is the development of a hydrocarbon (HC' s)
trap adsorbent for cold-start emission control. Afterwards, the capability of the zeolites as HC' s
trap adsorbents were examined using temperature-programmed desorption of two probe
molecules, ethylene and toluene. 'The TPD tests were carried out under four different mixture
conditions: binary (toluene-ethylene), temary (toluene-ethylene-C02) , temary (toluene-ethylene
H20) and quatemary (toluene-ethylene-C02-H20). The general ou~comes are as follows:
Nitrogen adsorption of the samples revealed that Al-ZSM-12 posseses higher specifie surface
area and micropore volume than Fe-ZSM-12 samples. B-SSZ-42 showed considerably high
values about 500 m2/g and 0.19'ml/g for surface area and micropore volume, respectively.
Elemental analysis confirmed that the Si/M (M= Al or Fe) of the samples was comparable
in both the solid products and their corresponding gel phases.
94
FTIR of adsorbed pyridine and TPD of ammonia indicated the existence of both Bronsted
and Lewis acid sites in the samples. In TPD profiles of ammonia, three distinct regions were
detected which relate to the different types of acid centers. Among AI-ZSM-12 samples, Ag
ZSM-12 owns the highest and strongest amount of Lewis acid sites. Among Fe-ZSM-12 samples,
the amount of Lewis acid sites decreased by increasing the Si/Fe ratio, whereas the amount of
Bronsted acid sites increased. It was also found that the strength of both Bronsted and Lewis acid
sites increased as the Si/Fe ratio increased.
lIB MAS NMR analyses of SSZ-42 samples indicated that most of the boron had been
incorporated as tetrahedral B04 units in the framework.
TPD of ethylene and toluene over the samples revealed that the single file diffusion
mechanism operated more efficiently in the microporous channels of the AI-ZSM-12 samples,
except for AI-ZSM-12 (H) compared to Fe-ZSM-12 samples. It was moreover found that Fe
ZSM -12 samples were not effective adsorbents for trapping ethylene/toluene owing to . their low
rate of hydrocarbon adsorption. Moreover, their adsorption capacity decreases considerably in the
presence of H20 and or/ CO2 .
It was shown that AI-ZSM-12 sa.mples possess higher adsorption capacities and desorption
temperatures for both ethylene and toluene than Fe-ZSM-12 samples. In this case, Na-ZSM-12
demonstrates a maximum desorption temperature of around 282°C for both ethylene and toluene
due to its strong Bronsted acid sites that create effective interactions with the probe molecules. In
the case of Ag-ZSM-12, adsorption capacity of ethylene and toluene were not significantly
changed in the presence of H20 and/or CO2. In this case, desorption temperatures of around
311°C and 400°C for toluene were indeed reached in TPD tests for binary mixture of
hydrocarbons and quaternary mixtures, respectively. These results were explained by the
dominating role of strong Lewis acid sites present in this adsorbent.
Indeed, TPD of hydrocarbons over B-SSZ-42 samples showed considerably high
adsorption capacity for both probe molecules. Moreover, It was also shown that these materials
have extremely stable 'adsorption behaviour in presence of H20 and/or CO2 as their trapping
95
capacities for hydrocarbons rernained essentially unchanged when the TPD tests were performed
in the presence of co-adsorbed H20 and C02. The desorption ternperatu~e of both probe
rnolecules were however relatively low (around 160 OC) owing to the low acid strength of their
acid sites.
5.2 Recommendations The following are sorne recomrnendations for future works:
1. Synthesis of B-SSZ-42 (Ag) and performing the TPD tests due to good results of
AI-ZSM-12 (Ag).
2. Synthesis of AI-SSZ-42 in order to achieve stronger acid sites in SSZ-42 that can possess
higher desorption ternperature of ethylene and toluene.
3. Study the feasibility of zeolitic/rnesoporous rnaterials as hydrocarbon trap adsorbents
for cold-start emission control due to their larger pore filling and adsorption capacity and
the possibility of yielding higher hydrocarbon desorption ternperatures.