QUEENSLAND UNIVERSITY OF TECHNOLOGY
SCHOOL OF PHYSICAL AND CHEMICAL SCIENCES
THE MORPHOLOGY AND STRUCTURE OF INTERCALATED AND PILLARED CLAYS
Submitted by Loc Van Duong to the school of Physical and Chemical Sciences, Queensland University of Technology, in partial fulfilment of the requirements of the degree of Doctor of Philosophy.
i
This thesis is dedicated to my wife Bach-Yen Thi Nguyen and children Tien-Nam Nguyen Duong, Thuy-Tien Nguyen Duong and Dang-Khoa Nguyen Duong
ii Acknowledgements
I would like to take the opportunity to express my sincere thanks to people who have contributed greatly to my research:
My principal supervisor, Dr. Theo Kloprogge, for his support, advice and guidance throughout my program of study. My co-supervisors Professor Ray Frost and Dr. Thor
Bostrom for their invaluable advice, encouragement and discussions.
Other people who involved in helping me with techniques and applications include:
• Dr Barry Wood, University of Queensland for his advice and assistance on the
XPS equipment and techniques.
• Dr Gregory Watson, Griffith University for assistance and comments on the
Scanning Probe microscope work.
• Dr Marek Zbik, Ian Wak Research Centre, South Australia for discussion on the
application of Atomic Force Microscope for clay minerals.
• Dr Wayde Martens for discussion and comments on vibrational spectroscopy
work on clay minerals.
• Professor H. Zhu and Professor HongPing He for the comments on organoclay
and pillared clay application.
Gratefully acknowledgements go to the X- Ray Analytical Facility, Mr Tony Raftery and Mr Lambert Bekessy
The Queensland University of Technology (QUT) Faculty of Science is acknowledged for providing time off and funding for me to complete my research.
iii Finally I would like to thank my wife Bach-Yen and my children Tien-Nam, Thuy-
Tien and Dang- Khoa for their understanding, patience, assistance and supporting me during the time of study.
iv
The work contained in this thesis has not been previously submitted for a degree or diploma at any other tertiary educational institution. To the best of my knowledge and belief, the thesis contains no material previously published or written by another person except where due reference is made.
______Signed
______Date
v Abstract
This thesis is submitted in a format of published papers by the candidate. Advanced methods of electron microscopy and X-ray spectroscopy have been used to study the relationship between the pillars and the silicate structure ranging from Al13 and Ga13 complexes to the final products Al- and Ga-pillared clays. The Al13 and Ga13 pillared montmorillonites have been prepared by conventional and ultrasonic methods. The ultrasonic method has been proven to be effective and showed very good catalytically activity. Transmission electron microscopy combined with elemental mapping by EDS showed the distribution of the Ga and Al pillars in the clay structure. The use of gallium allowed the independent observation of the Ga pillar distribution from the Al distribution in the clay structure.
XPS spectra of the Ga13 pillared montmorillonites showed that the pillars has been changed from the original Keggin structure with a 7+ charge to something more stable with a lower charge upon intercalation. No direct evidence of the inverted silicon tetrahedron structure bonding to the pillar structure, as suggested by Plee in his original thesis, was observed. For comparative reasons the major aluminium hydroxide minerals in bauxite (gibbsite, bayerite and (pseudo-) boehmite) were studied.
Detailed information about the Al13 structure was obtained by studying the basic sulphate and nitrate salts with XPS. The XPS results of a set of starting clays in comparison to the pillared clays indicated that small changes in the binding energy could explain the changes in the pillar structure and the formation of chemical bonds to the clay tetrahedral sheets during the calcination leading to the final products.
vi LIST OF PUBLICATIONS DERIVED FROM THIS RESEARCH
1. Duong, L.V., Wood, B.J. and Kloprogge, J.T. (2005). XPS study of basic aluminium sulphate and basic aluminium nitrate. Materials Letters, 59 (14/15), 1932-1936. 2. Duong, L., Bostrom, T., Kloprogge, T. and Frost, R., (2005). The distribution of Ga in Ga- pillared montmorillonites: a transmission electron microscopy and microanalysis study. Microporous and Mesoporous Materials, 82, 165-172. 3. Duong, L.V., Kloprogge, J.T., Frost, R.L. and van Veen, J.A.R. (2007) An improved route
for the synthesis of Al13-pillared montmorillonite catalysts. Journal of Porous Materials 14(1), 71-79 4. Kloprogge, J.T., Duong, L.V., Wood, B.J. and Frost, R.L. (2005) X-ray photoelectron spectroscopic study of the major minerals in bauxite: gibbsite, bayerite and (pseudo-) boehmite. Journal of Colloid and Interface Science, 296, 572-576. 5. He, H., Zhou, Q., Frost, R.L., Wood, B.J., Duong, L.V., and Kloprogge, J.T. (2007) A X- ray photoelectron spectroscopy study of HDTMAB distribution in organoclays. Spectrochimica Acta A, 66, 1180-1188. 6. Kloprogge, J.T., Duong, L.V. and Frost, R.L. (2005). A review of the synthesis and characterization of pillared clays and related porous materials for cracking of vegetable oil to produce biofuels. Environmental Geology, 47(7), 967-981 7. Kloprogge, J.T., Xi, Y., Duong, L.V. and Frost, R.L. (2005) High-resolution X-ray photoelectron spectroscopy of alkyl ammonium intercalated montmorillonites. 13th International Clay Conference, Tokyo, Japan, August 21-27, 113. 8. Kloprogge, J.T., Duong, L.V. and Frost, R.L. (2005). High-resolution XPS of boehmite: What is the difference between boehmite and pseudoboehmite? 13th International Clay Conference, Tokyo, Japan, August 21-27, 93.
9. Duong, L.V., Kloprogge, J.T., Frost, R.L. and Wood, B.J. (2005) The structure of Al13 and th Ga13 pillars in pillared montmorillonites. 13 International Clay Conference, Tokyo, Japan, August 21-27, 77.
vii LIST OF PUBLICATIONS RELATED TO THIS RESEARCH
1. Frost, R. L., W. N. Martens, Duong, L., Kloprogge, J. T. (2002). Evidence for molecular assembly in hydrotalcites. Journal of Materials Science Letters 21(16): 1237-1239. 2. He, H., R. L. Frost, Bostrom, T.,Yuan P., Duong L.,Yang .D., Xi Y. Kloprogge T. (2006). Changes in the morphology of organoclays with HDTMA+ surfactant loading. Applied Clay Science 31(3-4): 262-271. 3. Kloprogge, J.T., Duong, L.V., Frost, R.L. and Wood, B.J. (2008) Baseline studies of the Clay Minerals Society Source Clays: X-ray photoelectron spectroscopy. Clays and Clay Minerals, submitted. 4. Kloprogge, J.T., Duong, L.V., Martens, W.N., van der Eerden, A.J.M. and Frost, R.L. (2007) Mid- and Near-Infrared transmittance spectroscopy of hydrothermally synthesised
2:1 phyllosilicates in the system Na2O-Al2O3-SiO2-H2O. In Progress in Solid State Chemistry Research (ed. Buckley R.W.), Chapter 7, 285-300. 5. Kloprogge, T., Duong L., Frost, R., Bostrom, T. (2005). Environmental SEM: Application Of Low Voltage Sem, Heating Stage Sem And Variable Relative Humidity Sem To Study Uncoated Minerals. Microscopy and Analysis(July):17-17. 17. 6. Kloprogge, J.T., Broekmans, M., Duong, L.V., Martens, W.N., Hickey, L. and Frost, R.L.
(2006) Low temperature synthesis and characterisation of lecontite, (NH4)Na(SO4).2H2O. Journal of Materials Science, 41, 3535- 3529. 7. Ruan, H. D., R. L. Frost, Kloprogge, J. T., Duong, L., Schulze, D. G. (2001). Photo acoustic spectroscopy of kaolinite and gibbsite surfaces. 12th ICC, 22-28July 2001, Bahia Blanca. Argentina. 171. 8. Ruan, H. D., R. L. Frost, Kloprogge, J. T., Schulze, D. G., Duong, L. (2001). FT-Raman spectroscopy and SEM of gibbsite, bayerite, boehmite and diaspore in relation to the characterization of bauxite. The 12th International Clay Conference, 22-28 July 2001, Bahia Blanca, Argentina. 184. 9. Ruan, H. D., R. L. Frost, Kloprogge, J. T., Duong, L. (2002). Infrared spectroscopy of goethite dehydroxylation: III. FT-IR microscopy of in situ study of the thermal transformation of goethite to hematite. Spectrochimica Acta, Part A:Molecular and Biomolecular Spectroscopy 58(5): 967-981. 10. Ruan, H. D., R. L. Frost, Kloprogge, J. T., Duong L. (2002). Far Infared Spectroscopy of Alumina phases. Spectrochim. Acta, Part A 58A (1): 265-272.
viii 11. Ruan, H. D., R. L. Frost, Kloprogge,T., Duong, L (2002). Infrared spectroscopy of goethite dehydroxylation. II Effect of aluminium substitution on the behaviour of hydroxyl units. Spectrochimica Acta 58: 479-491.
ix TABLE OF CONTENTS
Acknowledgments ………………………………………...……………………………..iv
Abstract ………………………..………………………………………...…….…………vi
List of publication derived from this research …………………….…………...... vii
List of publication related to this research ……………………………………...………viii
Table of contents ……………………………………………….………………………....x
List of tables ………………………………………………………………..…………....xv
List of figures ……………………………………………………………...……...... ….xvii
CHAPTER 1 ...... 1
LITERATURE REVIEW ON CLAY MINERALOGY, PILLARED CLAY AND
TECHNIQUES USED FOR STUDYING PILLARED CLAYS
1.1 Introduction ...... 1
1.2 Clay Mineralogy ...... 2
1.2.1 The basic structure of clay minerals ...... 2
1.2.1.1 The tetrahedral sheet ……………………....……………. .3
1.2.1.2 The octahedral sheet ……………………....………….…..3
1.2.2 Classification of the clay minerals ...... 8
1.2.2.1 The 1:1 structure ………………….…………………..…..8
1.2.2.2 The 2:1 structure ………………….…………………..…..9
1.3 Intercalated and pillared clays ...... 14
1.3.1 Introduction ...... …14
x 1.3.2 Intercalated clays ...... 14
1.3.3 Pillared clays ...... 16
1.4 Analytical techniques for studying clays and their modifications ...... 30
1.4.1 X-Ray Diffraction ...... 30
1.4.2 Electron Microscopy and Microanalysis ...... 34
1.4.3 Atomic Force Microscopy ...... 41
1.4.4 Vibrational Spectroscopy and Solid State NMR ...... 44
1.4.5 X-Ray Photoelectron Microscopy ...... 45
1.5 General discussion ...... 50
1.6 Aims and objectives ...... 52
1.7 References ...... 54
CHAPTER 2 ...... …...72
THE DISTRIBUTION OF Ga IN Ga-PILLARED MONTMORILLONITES: A
TRANSMISSION ELECTRON MICROSCOPY AND MICROANALYSIS STUDY
2.1 Abstract ...... 73
2.2 Introduction ...... 73
2.3 Materials and methods ...... 76
2.3.1 Starting materials ...... 76
2.3.2 X-ray diffraction ...... 77
2.3.3 Sample preparation for Transmission Electron Microscopy ...... 77
2.3.4 Transmission Electron Microscopy ...... 80
2.3.5 Nitrogen adsorption-desorption ...... 80
xi 2.4 Results and discussion ...... 82
2.5 Conclusions ...... 92
2.6 Acknowledgement ...... 94
2.7 References ...... 95
CHAPTER 3...... 98
AN IMPROVED ROUTE FOR THE SYNTHESIS OF AL13 PILLARED
MONTMORILLONITE CATALYSTS
3.1 Abstract ...... 99
3.2 Introduction ...... 100
3.3 Experimental ...... 102
3.3.1 Starting materials ...... 102
3.3.2 Analytical techniques ...... 102
3.3.2.1 ...... X-Ray diffraction (XRD) ………………….…………… ..102
3.3.2.2 Scanning Electron Microscopy (SEM) ...… ….… 103
3.3.2.3 Atomic Force Microscopy (AFM) ……………...… …. 103
3.3.3 N2 adsorption /desorption ...... 108
3.3.4 Catalytic testing ...... 108
3.4 Results and discussion ...... 109
3.5 Conclusions ...... 119
3.6 Acknowledgement ...... 119
3.7 References ...... 120
xii CHAPTER 4 ...... 122
HIGH RESOLUTION XPS STUDY OF THE INTERNAL STRUCTURE OF AL-
AND Ga-PILLARS IN PILLARED CLAY CATALYSTS
4.1 Abstract ...... 123
4.2 Introduction ...... 124
4.3 Experimental ...... 126
4.3.1 Sample preparation ...... 126
4.3.2 X-ray diffraction (XRD) ...... 126
4.3.3 N2 adsorption /desorption ...... 127
4.3.4 X-ray Photo-electron Spectroscopy (XPS) ...... 127
4.4 Results and discussion ...... 128
4.5 Conclusions ...... 140
4.6 Acknowledgement ...... 140
4.7 References ...... 141
CHAPTER 5 ...... 144
XPS STUDY OF BASIC ALUMINUM SULPHATE AND BASIC ALUMINIUM
NITRATE
5.1 Abstract ...... 145
5.2 Introduction ...... 146
5.3 Experimental ...... 147
5.3.1 Basic aluminium sulphate and nitrate ...... 147
5.3.2 XPS analysis ...... 148
xiii 5.4 Results and discussion ...... 149
5.5 Acknowledgement ...... 157
5.6 References ...... 158
CHAPTER 6 ...... 160
SUMMARY AND SUGGESTIONS FOR FUTURE WORK
6.1 SUMMARY ...... 160
6.2 SUGGESTIONS FOR FUTURE WORK ...... 163
6.3 REFERENCES ...... 165
CHAPTER 7 ...... 166
ADDITIONAL SUPPORT PAPERS
7.1 REVIEW OF THE SYNTHESIS AND CHARACTERIZATION OF
PILLARED CLAYS AND RELATED POROUS MATERIALS FOR
CRACKING OF VEGETABLE OIL TO PRODUCE BIOFUELS...... 167
7.2 X- RAY PHOTOELECTRON SPECTROSCOPIC STUDY OF THE
MAJOR MINERALS IN BAUXITE: GIBBSITE, BAYERITE AND
(PSEUDO-) BOEHMITE...... 204
7.3 A X-RAY PHOTOELECTRON SPECTROSCOPY STUDY OF HDTMAB
DISTRIBUTION WITHIN ORGANOCLAYS...... 212
xiv LIST OF TABLES:
Table 1.1 Classification of common phyllosilicates. ………………………………12
Table 1.2 Atomic and X-ray notations ……………………………………..………49
Table 2.1 Characteristic concentrations (single analysis) in weight % from X-ray
microanalyses in the TEM of the starting and the Ga13-, Al12Ga- and Al13-
pillared montmorillonites, and the formula calculation based on 22
oxygens ………………………………………………………….………83
Table 3.1 BET surface area and pore volume and average pore diameter of Al-
pillared montmorillonites …………………………...…………………103
Table 3.2 Temperatures for 40% n-heptane conversion on Pd-loaded pillared clays
(calcined 450°C, Pd loading 0.4 wt%) and two standard catalysts (ASA
and Pd/Al2O3:Si) …………………………...…………………………109
Table 4.1 XRD and N2 adsorption/desorption results for the starting montmorillonite
SWy-2, the Al13- and Ga13-intercalated montmorillonites and the calcined
Al13- and Ga13-pillared montmorillonites ……………………………...119
Table 4.2 Chemical analysis (at %) of the starting clay SWy-2 and the Al and Ga
pillared equivalents …………………………………………………….134
Table 5.1 Chemical composition (in atom%) from the XPS analyses of the basic
aluminium sulphate at room temperature and after calcination at 200 and
400°C and basic aluminium nitrate ………………………………….....137
Table 5 2 Binding energies (in eV) of the basic sulphate at room temperature and
after calcination at 200 and 400°C …………………………………….143
xv LIST OF FIGURES
Fig. 1.1 Diagram showing the single tetrahedron unit and the arrangement of the
tetrahedral sheet …………………………………………………...... …….4
Fig. 1.2 Diagram showing the single octahedral unit and the arrangement of the
octahedral sheet ………………………………………………...…...………6
Fig. 1.3 Diagram showing the arrangement of the tetrahedral and octahedral sheet to
form an ideal structure of clay minerals with no substitution .………...... ….7
Fig. 1.4 The structure of typical montmorillonite minerals ……..….……….……..13
Fig. 1.5 Representation of the pillaring process showing d spacing changing from the
beginning and the end product after calcination ………………...………...19
Fig. 1.6 XRD patterns of untreated and treated montmorillonites from Miles,
Queensland, Australia ………………...……………………...………...….33
Fig. 1.7 Image formation in optical microscope, transmission microscope and
scanning electron microscope. In the optical microscope the source is light
bulb whereas in the SEM and TEM the source is an electron beam ………35
Fig. 1.8 SEM image of pillared montmorillonite (Top) and TEM image of particle of
Montmorillonite (bottom) ……………………...…….……………………39
Fig. 1.9 SEM images of Kaolinite in sandstone shows typical book structure ….....40
Fig. 1.10 Typical diagram of an atomic force microscope showing a tip mounted on a
cantilever (C). The tip scans over the surface of the sample. The laser light
from the back of the cantilever reflects to a two element photosensitive
xvi diode (PSD) and the output signal can be used in a feedback loop to control
the vertical position of the sample ………………...………………………42
Fig. 1.11 Diagram showing Photoelectron and Auger electron processes …...…...…48
Fig. 2.1 Preparation of cross sections of a clay sample for TEM ………………….79
Fig. 2.2 XRD patterns of a) starting Wyoming SWy-2 montmorillonite, b)
montmorillonite exchanged with Ga13 showing an interlayer spacing of 19.9
Å, and c) Ga13 pillared montmorillonite with a spacing of 17.9Å ………...81
Fig. 2. 3 TEM image of a) a grain of Al12Ga pillared montmorillonite; and images
and electron diffraction patterns from sectioned material: b) Ga13 pillared
montmorillonite; c) Al12Ga pillared montmorillonite and d) Al13 pillared
montmorillonite …………………………………………………...………84
Fig. 2.4 Elemental X-ray maps for Ga, Si and Al from a cross section of a single
grain of Ga13 pillared montmorillonite ……… ……87
Fig. 2.5 EDX spectra from analyses in the TEM of small grains of (a) Ga13 pillared,
(b) Al12Ga pillared, and (c) Al13 pillared montmorillonites. The spectra are
shown overlaid with a spectrum from the starting material. The C and Cu
peaks derive from the resin or thin carbon coating and the TEM grid
material respectively ………………………………………………………88
Fig. 2. 6 Pore size distribution of Ga13 pillared montmorillonite ………………….91
Fig. 3.1a Non-calcined Al13- montmorillonite SAz-1 ……………………...………104
Fig. 3.1b Non-calcined Na-exchanged Miles montmorillonite intercalated with Al13
……………………...... ………104
xvii Fig. 3.1c Non-calcined Miles montmorillonite intercalated with Al13……....……105
Fig. 3. 2a Al-pillared montmorillonite SAz-1 calcined at 450°C ...…………...….107
Fig. 3. 2b Al-pillared Miles montmorillonite calcined at 450°C ……………….....107
Fig. 3. 2c Al-pillared Na-exchanged Miles montmorillonite calcined at 450°C .…108
Fig. 3.3a BET adsorption and desorption isotherms for Al-pillared montmorillonite
SAz-1 after ultrasonic treatment for 5 minutes and calcined at 450°C …..111
Fig. 3.3b BET adsorption and desorption isotherms for Al-pillared Miles
montmorillonite after ultrasonic treatment for 5 minutes and calcined at
450°C .……………………………………………………………………112
Fig. 3. 4a Al-pillared montmorillonite SAz-1 (ultrasonic treatment 5 minutes, calcined
at 450°C) ...…………...…………………………………………………..113
Fig. 3. 4b Al-pillared montmorillonite SAz-1 (ultrasonic treatment 10 minutes,
calcined at 450°C) ...…………...……………………………………..…..113
Fig. 3. 4c Al-pillared montmorillonite SAz-1 (ultrasonic treatment 20 minutes,
calcined at 450°C) ...…………...…………………………………..……..114
Fig. 3. 5a AFM raw image (left) and FTIR processed image (right) of Miles
montmorillonite ...…………...……………………………………..……..117
Fig. 3. 5b AFM raw image (left) and FTIR processed image (right) of Miles
montmorillonite after 10 minutes ultrasonic treatment ……..……..……..117
Fig. 3. 5c AFM raw image (left) and FTIR processed image (right) of Al-pillared
Miles montmorillonite after 20 minutes ultrasonic treatment ……..……..118
xviii Fig. 4.1 XPS survey scan of the Wyoming montmorillonite starting material …...131
Fig. 4.2 High resolution XPS spectra of Si 2p, Al 2p and O 1s of the starting
montmorillonite SWy-2 ………………………………………………....136
Fig. 4.3 High resolution XPS spectra of Si 2p, Al 2p, Ga 2p and O 1s of the Ga-
pillared montmorillonite SWy-2 ……………………………………..…..137
Fig. 4.4 High resolution XPS spectra of Al 2p of Al13 sulfate after calcination at
400°C (top ), gibbsite with only AlVI (middle), and corundum with AlVI
(bottom ) ……………………………………………………………....….138
Fig. 4.5 High resolution XPS spectra of Si 2p, Al 2p and O 1s of the Al-pillared
montmorillonite SWy-2 …………………………………………...….….139
Fig. 5.1 SEM images of a basic aluminium sulphate crystal at room temperature and
after calcination at 400°C …………………………………….………….141
Fig. 5.2 XPS survey scan of basic aluminum sulphate …………………….……..142
Fig. 5.3a Al 2p high resolution spectrum of basic aluminium sulphate ……………145
Fig. 5.3b O 1s high resolution spectrum of basic aluminium sulphate ……………146
Fig. 5.3c S 2p high resolution spectrum of basic aluminium sulphate …………….146
xix CHAPTER 1
1.0 LITERATURE REVIEW ON CLAY MINERALOGY, PILLARED CLAY AND TECHNIQUES USED FOR STUDYING PILLARED CLAYS
1.1 Introduction
The term “clay” has been used to either indicate fine particles, with a grain size less than 2 µm or minerals belonging to the clay minerals group (Velde, 1992). In this thesis clay refers to a member of the clay minerals group, as part of the phyllosilicates. Clay is present almost everywhere, flying up in the air as dust particles, covering the surface of the earth as part of the soils, and below the surface as part of sedimentary rocks. Clay is mainly formed through the process of weathering of primary silicate minerals such as feldspars. The characteristics of clay deposits are depended on the source rocks, the weathering processes, transportation and the environmental conditions (Velde, 1992).
The first documented use of clay under the term Kauling Earth was in the middle of the Ming Dinasty, China, in AD 1604 (Liu and Bai, 1982). The two most popular clay minerals are kaolinite and montmorillonite. Kaolinite is the main component of the kaolin deposits, which has been used for making ceramics in China since the seventeenth
1 century (Chen et al., 1997). The name “kaolin” comes from the Chinese word kauling, which is the name of a village near Jauchou Fu, central China (Chen et al., 1997).
In Europe Von Liebig used acid treated clay as catalyst (Liebig, 1865). Professor
Heinrich Ries of Cornell University was probably the first American geologist to study clay minerals in 1920 (Grim, 1988). Montmorillonite, named after the town
Montmorillon, France, is the main component of bentonite deposits, and has the capacity to absorb water molecules between the layers around the exchangeable cations. Bentonite is used in many applications including drilling mud for oil drilling, as absorbents for waste water or for stopping leakage in soils, rocks and dams (Grim, 1988)
The structure of clay minerals has only been understood since the discovery of X-rays and the development of X-ray diffraction methods, with the first analyses of clay particles published in 1923 (Hadding, 1923). The development of advanced analytical techniques such as vibrational spectroscopy, electron microscopy and surface analysis have helped clay mineralogists to study in detail the morphology, structure and chemistry of these minerals down to the atomic scale.
1.2 Clay Mineralogy
1.2.1 The basic structure of clay minerals
The clay minerals belong to the phyllosilicates or layer silicates. Members of this group have a platy habit and perfect (001) cleavage. They are generally soft, the hardness ranges from 2 to 3 on the Mohs scale and the colour varies from yellow to pure white depending mainly on the substitution of Fe into the structure. Most of the clay minerals
2 belong to the triclinic or monoclinic system. The four important clay mineral groups are: kandites, illites, smectites and vermiculites. Kaolinite from the Kaolin group and montmorillonite from the smectite group are the most common clay minerals found in nature.
The basic structure of clay minerals can be obtained though the stacking of two sheets: tetrahedral sheets and octahedral sheets sometime separated by an interlayer. Different clay minerals are formed by (1) different combination of these two units and the interlayer and (2) changes in the composition of the sheets.
1.2.1.1 The tetrahedral sheet
The basic unit of this layer is a tetrahedron, which contains normally one Si4+ in the centre with four O2- at the corners. The tetrahedra are linked to neighbouring tetrahedra by sharing three oxygen atoms each to form a hexagonal mesh pattern. All the unshared corners with the apical oxygen atoms point in the same direction to form part of the adjacent octahedral sheet (Figure 1.1).
1.2.1.2 The octahedral sheet
The basic unit of this layer is an octahedron, which contains mainly Al3+ or Mg2+ surrounded by six oxygen atoms or hydroxyl groups. When the cations are trivalent, the sheet contains two cations per half unit cell and one vacancy. This is a structure similar to gibbsite, Al(OH)3, and is known as a dioctahedral structure. When the cations are divalent, the sheet contains three cations per half unit cell and no vacancy. This is a structure similar to brucite, Mg(OH)2, and is known as a trioctahedral structure.
Octahedral sheets can contain other cations including Li+, Fe2+, and Fe3+ (Figure 1.2).
3
Single tetrahedron unit Corner linked tetrahedral sheet (top view)
Apical Oxygen
Silicon
Tetrahedral sheet formed by joining of tetrahedral unit (side view)
Fig. 1.1 Diagram showing the single tetrahedron unit and the arrangement of the tetrahedral sheet
4 The octahedral and tetrahedral sheets are covalently linked through sharing of the apical oxygen atoms of the tetrahedral sheet with the octahedral sheet creating the metal cation (octahedral) - O- Si (tetrahedral) link as seen in Figure 1.3. The octahedral sheet is commonly slightly smaller than the tetrahedral sheet (Triolo et al., 1988). To compensate for this difference, the tetrahedra in the tetrahedral sheet rotate resulting in a distortion sheet.
The cations in the tetrahedral and/or octahedral sheet can be replaced by other cations to cause a negative charge in the structure. The charge can be located in the tetrahedral and/or octahedral sheet and is very useful to identify different species of clay minerals.
The structure of the clay minerals is quite complex. The combination of the basic clay units and the interlayer with substitution of different cations in tetrahedral or octahedral layers can result in different species. These different minerals have quite distinct properties and can be grouped together by similar layer type, layer charges, or the type of interlayer species. Not all clay minerals are suitable to make pillared clays; the groups with layer charges ranging from 0.4 to 1.8 with hydrated exchangeable cations in the interlayer are the most common starting material. In the structure of the clay mineral, the charges can be unbalanced, resulting in minerals containing a permanent charge (negative or positive) or no charge (on the surface) as in the case of kaolinite. A second source of charge is found along the edges of clay minerals, which depended on the pH of the solution. The classification of common phyllosilicates is listed in Table 1.1. The two most popular minerals with the 1:1 and 1:2 structure, kaolinite and montmorillonite, are discussed below.
5
Edge linked octahedral sheet (top Single octahedral unit view)
Octahedral sheet formed by joining of octahedral unit (side view) Al, Mg
O, OH
Fig. 1.2 Diagram showing the single octahedral unit and the arrangement of the octahedral sheet
6 O
T
(a)
T
O
T (b)
Octahedral
Tetrahedra
(c)
Fig. 1.3 Diagram showing the arrangement of the tetrahedral and octahedral sheet to form an ideal structure of clay minerals with no substitution: (a) and (b) showed the 1:1 and 2:1clay mineral structure (side view), (c) The link between the tetrahedral and octahedral layers is the result of sharing apical oxygen from the tetrahedral layers and the unshared ions normal to the octahedral sheet 7 1.2.2 Classification of the clay minerals
1.2.2.1 The 1:1 structure
The so called 1:1 clay structure contains one tetrahedral and one octahedral sheet as the repeat unit with no layer charge. The interlayer is occupied by hydroxyl groups and oxygen atoms from the octahedral and tetrahedral sheets connected by weak hydrogen bonds. Members of this group can be dioctahedral such as kaolinite, containing Al and Si with no substitutions, or trioctahedral such as chrysotite containing Mg and Si ( see Table
1.1)
Kaolinite belongs to a group of 1:1 clay minerals known as the Kaolin group. In this group dickite and nacrite are the polymorphs of kaolinite and distinguished by the stacking of layers along the c axis. Halloysite has similar structure as kaolinite but has more water in the interlayer resulting in the d (001) spacing increasing from 7.15 Å to
10Å. The structure of kaolinite consists of a repetition of layers existing of one Al
(O,OH) octahedral sheet coupled to a single SiO4 tetrahedral sheet.
Within this structure a differentiation can be made between four hydroxyl groups.
These hydroxyl groups are generally known as the inner and outer hydroxyl groups.
Three of these groups stick out of the layers and form the hydrogen bonds to the next layer. The outer hydroxyl groups (OuOH) are located in the outer, upper and unshared plane whereas the inner hydroxyl groups are located in the lower shared plane of the octahedral sheet. The location of the hydrogen atoms is very difficult to study by XRD due to X-ray scattering by hydrogen atoms but they do have a large neutron incoherent scattering (Bailey, 1988). Vibrational spectroscopic techniques have been found very
8 useful for studying the location of hydroxyl groups in kaolinite (Ledoux and White,
1966), hence a large amount of literature may be found on this subject. The IR OH- stretching modes of the hydroxyl groups from of the kaolinite can be observed in the region between 3600 and 3700cm-1, (Bailey, 1988; Frost, 1997). The source of interlayer bonding in kaolinite structure comes from the hydrogen bonds between the octahedral layer and the oxygen of the tetrahedral layer.
The two sheets of the kaolinite are bonded together by Van der Waals forces and hydrogen bonding. No interlayer cations or layer charge are present in the kaolinite structure. The layers are connected by Si-O-Al bonds. Kaolinite has quite often been qualified as clay that is not capable of expansion. However specific organic molecules can be inserted into the kaolinite structure resulting in the expansion of kaolinite along the c axis (Kristof et al., 1998; Sidheswaran et al., 1987; Frost et al., 2000b; Frost et al.,
1999c).
1.2.2.2 The 2:1 structure
The 2:1 clay layer type is characterized by an octahedral sheet sandwiched between two tetrahedral sheets. Depending on the layer charge the 2:1 layers groups can be divided into two subgroups:
• Subgroup 2:1 with no layer charge and no exchangeable cations
The dioctahedral mineral of this group is pyrophyllite, Al4Si8O20(OH)4, and the trioctahedral equivalent is talc, Mg6Si8O20(OH)4. The layer thickness of the octahedral and two tetrahedral sheets together is about 9.5 Ǻ.
9 • Subgroup 2:1 with layer charge and exchangeable cations
The layer charge of the smectite group range from 0.4- 1.2 and the species of this group correspond to the most commercial clays known as bentonite. Members of this group have the ability to exchange interlayer cations and water with the surrounding environment.
Smectite is a name for a group of 2:1 layer minerals, either dioctahedral or trioctahedral in nature, which can expand when water or organic molecules are introduced into the interlayer space. Montmorillonite and beidellite are dioctahedral smectites with mainly Al in the octahedral sheet, which have been formed by alteration of volcanic ash or tuffs (Hewitt, 1917). The trioctahedral smectites, such as saponite, hectorite and stevensite, contain mainly Mg+ (with or without Li+ or vacancies) in the octahedral sheet. The charge can be located in the tetrahedral sheet (saponite) or in octahedral sheet (hectorite).
The substitution of cations, which have different valencies, can lead to charge unbalances within a sheet and can be totally or partially balanced by the adjacent sheet.
The remaining net charge or layer charge is negative for the 2:1 layer. This charge is balanced by large hydrated cations, such as Na+, K+, Ca2+ and Mg2+, which coordinate to the basal surfaces of the tetrahedral sheets from the adjacent layers. These charged cations are referred to as “interlayer cations” (see Figure 1.4). The layer charge arises primarily in the octahedral sheet for montmorillonite, hectorite and in the tetrahedral sheet for beidellite, saponite and nontronite. Smectites are often referred to as “swelling clays” due to the ability of expanding when contact with water. The distance between two
10 clay layers can vary from 12Ǻ to 16Ǻ depending on the size of the exchangeable ions and
the amount of water. The charge imbalance is much higher compared to illites and these
electrical property results in the high cation adsorption capacity of the smectite structure.
Layer Interlayer Layer Octahedral Formula Mineral type species charge sheet type name
The smectites may be Fe3+-rich (nontronites) or Al-rich (montmorillonite-beidellites)
although complete compositional substitutions are possible. Volkhonskoite is a Cr rich
(>15% Cr2O3) dioctahedral smectite with Cr contains mainly in octahedral sheet.
Sauconite is a trioctahedral smectite contaning Zn in the octahedral sheet. Other smectites
containing vanadium, nickel and copper have also been reported (Guven and Hover,
1979, Brindley and Maksimovic, 1974). The ability to absorb water can result in an
increased distance along the c axis of the clay layers. The free and rapid exchange of
cations, complexes and water in the smectite interlayer depends largely on the physico-
chemistry of the environment (Weaver and Pollard, 1973; Velde, 1985; Newman, 1987)
and can be used to produce a large variety of intercalated and pillared clays.
11 1:1 none or H2O ~0 di Al4Si4O10(OH)8 kaolinite only tri Mg6Si4O10(OH)8 chrysotile
2:1 None ~0 dil Al4Si8O20(OH)4 pyrophyllite
tri Mg6Si8O20(OH)4 talc
n+ . hydrated 0.4-1.2 di Mgx/n [Al4-xMgx][Si8]O20(OH)4 nH2O montmorillonite exchangeable n+ . di Mgx/n [Al4] [Si8-xAlx]O20(OH)4 nH2O beidelite
cations n+ . di Mgx/n [Fe4][Si8-xAlx]O20(OH)4 nH2O nontronite
n+ . tri Mgx/n [Mg6][Si8-xAlx]O20(OH)4 nH2O saponite
n+ . tri Mgx/n [Mg6-xLix][Si8]O20(OH.F)4 nH2O (F-)hectorite
n+ . tri Mgx/n [Mg6-xVacancyx][Si8]O20(OH)4 nH2O stevensite
2+ . 1.2-1.8 intermediat [Mg,Ca]x/2 [Al4-xMgx][Si8]O20(OH)4 8H2O vermiculite e 2+ . [Mg,Ca]x/2 [Mg6][Si8-xAlx]O20(OH)4 nH2O vermiculite tri
non-hydrated 1.8-2.0 di Na2[Al4][Si6Al2]O20(OH)4 paragonite cations tri K2[Mg.Fe]6[Si6Al2]O20(OH,F)4 phlogopite
. 2:2 hydroxyl sheet 0.4-2.0 tri [(AlxMg6-x)(OH)6][Mg6][Si8-xAlx]O20(OH)4 clinochlore
3+ 2+ 2+ tri [Fe xFe 6-x)(OH)6x][Fe 6-yMgy][Si8-xAlx]O20(OH)4 thuringite
12
12-16Ǻ
Fig. 1.4 The structure of typical smectite minerals
13 1.3 Intercalated and pillared clays
1.3.1 Introduction
Clay minerals with an ability to exchange interlayer cations have attracted a lot of interest for use in industrial applications. To modify the clay structure, different compounds have been used to intercalate between the layers with the hope of increasing the pore size. Barrer and McLeod were the first to introduce the process of intercalation of organic compounds into the structure of clay minerals (Barrer and MacLeod, 1955).
These products are quite suitable for applications such as absorbent, fillers and thickeners
(Schoonheydt et al., 1993), but were found to decompose at relatively low temperatures making them unsuitable for most catalytic applications (Barrer and MacLeod, 1955).
The next development stage of pillared clay coincided with a crisis in the oil industry.
The search for new type of materials with relatively large pore-sizes to deal with larger molecules in the crude oil, and good thermal and hydrothermal stability has resulted in the development of inorganic polyoxocations as pillaring agents that could be intercalated between the clay layers and, when calcined, produced fixed metal oxide pillars, providing new types of acidic, highly porous, thermally stable materials with a high specific surface area (200 to 500 m2g-1).
1.3.2 Intercalated clays
Intercalation is the process of inserting organic or inorganic molecules into the interlayer of the clay structure The principle of the intercalation process in kaolinite was pointed out by Lagaly in 1984 (Lagaly, 1984). The new organic molecules, which were introduced into the kaolinite layers were found to have broken the hydrogen bond
14 between the hydroxyl group of the octahedral sheet and the oxygen atoms of the tetrahedral sheet to form new bonds with the more hydrophobic siloxane layer or with the more hydrophilic hydroxyl groups.(Weiss et al., 1966; Lagaly, 1984; Costanzo and Giese,
1990).
The intercalation of organic molecules such as hydrazine, urea, formamide, acetamide,
DMSO and acetate has been discussed in a review by Frost et al. (2000a). The ideas of intercalation kaolinite with organic molecules to make it expanded has given rise to a new field of research and since then more molecules have been tested for dimensions and bonding properties with the kaolinite structure. The most common molecules, which have been used for kaolinite intercalation are hydrazine (NH2-NH2), urea (NH2-C=O-
NH2) and formamide (HC=O-NH2).
In general, the organic molecules that intercalated into kaolinites can be divided into three groups. Group I contains molecules that can form strong hydrogen bonds with the siloxane layer. These molecules include hydrazine (NH2-NH2), urea (NH2-C=O-NH2) and formamide (HC=O-NH2). Kaolinite intercalated with molecules from this group gives expansion of the basal spacing from 7.2Å to 10.4Å. (Ledoux and White, 1966); (Johnston and Stone, 1990); (Frost et al., 1999b), (Cruz and Franco, 2000).
Group II contains molecules with strong dipole interactions. These molecules can interact with the siloxane layer of the kaolinite. Members of this group include dimethylsulfoxide ([CH3]2SO), and dimethylselenoxide. ([CH3]2SeO). Kaolinite intercalated with molecules from this group results in a large expansion of the basal spacing, from 7.2Å to approximately 11.2Å (Frost et al., 1998b). The intercalation of
15 kaolinite with these molecules can be used as a precursor for other inorganic alkaline salts.(Lahav, 1990; Lapides et al., 1997).
Group III contains alkali salts of short chain fatty acids such as acetic and propionic acid. (Wada, 1961;Ledoux and White, 1966; Weiss et al., 1966). The surface hydroxyl groups on the kaolinite can form hydrogen bonds with fatty acid salt. Kaolinite intercalated with molecules from this group gives a basal expansion from 7.2Å to approximately 14.2Å.
Organoclays have been used for treating soil to remove contaminants, e.g., polychlorinated biphenyls and polyaromatic hydrocarbons. Organoclays have been modified with a range of pillaring agents such as Al(OH)3, SiO2, and TiO2 and been found to have organophilic properties (McLeod, 1997). In a study of HDTMA+ / montmorillonite , the morphology of organoclay has been shown strongly depending on the surfactant packing density within the montmorillonite interlayer space. Thermal treatment has an important effect on the stability of organoclays, reflected by significant changes in the basal spacing (He et al., 2006).
1.3.3 Pillared clays
The use of inorganic compounds as pillaring agents has provided an alternation to organic compounds as pillaring agents which suffer the low thermostability. Most of the recent research on pillared clays has been concentrated on Al13 pillared clays which use
7+ the Al13 polyoxocation [Al13O4(OH)24(H2O)12] , as the pillaring agent of which was first studied by Johansson in the ‘60s (Johansson, 1960).
16 The Al13 complex consists of a central tetrahedral aluminium cation, which is surrounded by twelve edge-linked octahedrally coordinated aluminium cations
(Johansson, 1962).
In general the pillaring process can be described according to the diagram in Figure
1.5. The smectite used as a starting material is ion exchanged with Na or Li in order to obtain maximum swelling prior to the exchange with the large inorganic pillaring complex. To obtain for example Al pillared clays, a solution containing the Al13 polyoxocation is mixed with the Na or Li exchanged clay (smectite group) suspension.
After the cation exchange reaction between the clay and the polyoxocation has taken place, the suspension is washed; the expanded clay is separated, dried and calcined at around 400 to 500°C to convert the complex to intercalated fixed metal oxide pillars in the clay interlayers. The calcination process of Al pillared smectites can liberate protons from the pillar, which can then diffuse into the clay sheet, lowering the thermal stability.
In pillared beidellite, the formation of Si-OH groups was observed during dehydration in the IR emission spectra. This indicated that at the same time protons are liberated, which can interact with the tetrahedral Si-O-Al bonds upon calcination and forming new
Si-OH-Al bonds. A reaction between the Al13 pillar and the protonated Si-OH-Al linkage will yield Altetrahedral-O-Alpillar linkages (Chevallier et al., 1994; Kloprogge and Frost,
1999; Kloprogge et al., 1999b; Kurschner et al., 1998). No clear reaction has been observed so far between pillars and montmorillonite upon calcination although a stable bond is formed, whereas in pillared beidellite a structural transformation links the pillar to some sort of inverted tetrahedra of the tetrahedral sheet (Kloprogge, 1998).
17 The first reports on Al13 pillared montmorillonite were published around 1977-1981 by several authors.(Brindley and Sempels, 1977; Lahav et al., 1978; Vaughan and Lussier,
1980; Vaughan et al., 1981). The Al13 complex has been used to prepare pillared clays with the d(001) around 17-18 Å and thermal stability up to 500ºC (Brindley and Sempels,
1977; Miehe et al., 1997)
The stability of Al13 pillared montmorillonite has been suggested to result from the formation of Al metal oxides through dehydration and dehydroxylation of the Al13 cations at high temperature as shown by the equation below. (Pinnavaia, 1983; Schoonheydt et al., 1994)
+ (3−n) + + ⎯−⎯→H⎯2O 6.5 + (3 − n) Al13O4 (OH)28+n Al2O3 H
The number and strength of the acid sites occurring in the clay minerals are important factors for catalytic applications. Two types of acid sites have been discussed in the literature: the Brönsted and Lewis acid sites. Both types of acid sites have been reported in pillared clays (Occelli and Tindwa, 1983; Pinnavaia, 1983). In general, a Brönsted acid is defined as a substance which can supply a proton and a Brönsted base is a substance which can accept a proton. A Lewis acid is defined as a substance that can receive a pair of electrons and the Lewis base can donate a pair of electrons.
18
Fig. 1.5 Representation of the pillaring process showing d spacing changing from the beginning and the end product after calcinations Modified after Gil et al. (Gil et al., 2000b)
19 The acidity of the pillared clays can be studied by calorimetric adsorption of pyridine and monitoring of pyridine desorption by IR spectroscopy (Occelli, 1985; Figueras et al.,
1990; Gonzalez Luz et al., 1990). Lewis acid sites have been found to be correlated with the number and nature of the pillars, whereas the Brönsted acid sites are located within the clay sheets (Bagshaw and Cooney, 1993). For example, Cr pillared clays show stronger Lewis acid sites than other pillared clays (Jiang et al., 1991; Auer and Hofmann,
1993). Acid sites can also be studied by IR combined with temperature programmed desorption of ammonium (Bodorado et al., 1994; Vogels et al., 2004; Zhao et al., 1994).
Bagshaw and Cooney (Bagshaw and Cooney, 1993) found that the characteristics of surface Lewis acid sites were associated with the pillar species while those of the
Brönsted acid sites are determined by both the pillar and the starting clays. The pillared clays have been found to have more Lewis acid sites than Brönsted acid sites (Kurian and
Sugunan, 2005). The surface acidity varies with the pillaring agents and stems from their pillars. During calcination the pillars are transformed to oxides. In this process protons are released and can produce the Lewis and Brönsted sites (Barrer and MacLeod, 1955;
Plee et al., 1985b; Xie et al., 1994).
When Na- exchanged saponite was used as a starting material, the Brönsted acid sites were not present but were later formed in the Al-pillared product, based on the observation of a pyridinium band at 1549 cm-1 in the infrared spectrum after the absorption of pyridine (Bergaoui et al., 1995). The Brönsted acid sites are thought to be related to OH groups formed by the acidic attack of Si-O-Al linkages, which occurs on the clay sheets during the pillaring process (Chevallier et al., 1994; Kurschner et al.,
20 1998). These hydroxyl groups were observed in the infrared spectrum at 3740-3720 cm-1 and 3597-3594 cm-1.
Changes in the local environment, such as short range order and coordination, in the interlayer, the octahedral or tetrahedral sheets in smectites or their pillared equivalents can be studied by Magic-Angle Spinning Nuclear Magnetic Resonance spectroscopy
(MAS-NMR).
The movement of water and interlayer cations in synthetic beidellite (Na0.7Al4.7Si7.
3O20(OH)4.nH2O) has been studied by Kloprogge et al. (Kloprogge, 1992) using solid- state 23Na and 27Al magic-angle spinning (MAS) NMR. The 23Na NMR chemical shift of
0.2 ppm indicated that the hydrated sodium cations in the interlayer resemble Na+ in solution. From 25°C to 85°C four water molecules per Na+ were found to be removed, which caused the basal spacing to decrease from 12.54 Å to 9.98 Å and the remaining water molecules with the Na+ was relocated much closer to the tetrahedral sheet, as indicated by the chemical shift of 1.5 ppm after the first dehydration step. The second dehydration occurs at a temperature around 400ºC but the removal of the remaining 2 water molecules did not cause any further decrease of the basal spacing (Kloprogge et al.,
1992).
Plee et al (1985a) used 27Al and 29Si to study pillared beidellite and suggested that the linking of Al13 and tetrahedral layer as a result of inversion of tetrahedral unit and the formation of Si-OH-Al group. In Plee’s original PhD thesis there is also evidence of a separate Al(IV) peak for the pillar next to the tetrahedral Al peak of the clay which shows
21 changes with calcinations. However, no one has been able to reproduce these results. In all cases a severe overlap of the two signals has been observed.
The study of saponite from Ballarat pillared with Al polycations by 29Si and 27 Al solid state NMR suggests that the pillaring mechanism involved H+ attack of the clay
27 tetrahedral sheet followed by AlO4 tetrahedra inversion. Al NMR revealed that the pillars are not modified upon calcination at 500°C although they undergo reversible dehydration reactions; at 750°C and above, a strong pillar reorganisation occurs prior to collapse of the global structure (Bergaoui et al., 1993).
The work of Kloprogge et al in 1999 showed that the structure of the pillared clays changed during the pillaring process. During calcination up to 500°C the pillars were not completely converted to the oxide and remained as hydroxyl groups (Kloprogge and
Frost, 1999; Kloprogge et al., 1999a; Kloprogge et al., 1999b). This fits in nicely with the work by Balek et al (Balek et al., 1998) using emanation thermal analysis, which showed that after heating to 500 C° followed by cooling to room temperature and a second heating cycle that there were still some volatiles present that were driven off during this second heating cycle.
In a study of montmorillonite pillared with single and mixed Keggin polycations and calcination to 900°C using 27Al and 29Si MAS-NMR, the result showed that the tetrahedral layers (SiO4) were not significantly affected by the pillaring process whereas the OH regions changed during the pillaring and calcination (Espinosa et al., 1994). The
OH vibration in the corresponding infrared spectra, which occurred at 3543cm-1, disappeared around 400°C owing to the formation of pillar-layer bonds. The new pillar
22 OH vibration occurring at 3695cm-1 and the Si-OH at 3750cm-1 disappeared after the samples were heated to 700°C (Espinosa et al., 1994).
Due to its similar electronic configuration, gallium (Ga) has chemical properties comparable to aluminum (Al) and can be used to improve the catalytic activity of the pillared clays (Bellaloui et al., 1990; Bradley and Kydd, 1993a). Bradley and coworkers have done an extensive research of the hydrolysis of gallium (Bradley, 1991; Bradley and
Kydd, 1993 a, b, c, d, e; Bradley et al., 1990). They found that Ga could form a similar
Keggin-type polyoxocation and may be used as pillaring agents in a similar method to
Al13. In Bradley’s study, the Al at the tetrahedral site in the Al13 structure was replaced by
Ga resulting in a better fit in the Keggin structure (Bradley, 1991). The small size of the
Al cation causes a slight distortion in the tetrahedral site in the Al13 structure. This distortion is minimized by replacing the Al cation by the larger Ga cation, which results in an increase of the stability of the GaAl12 Keggin structure (Bradley, 1991).
The structure of Ga pillared clays and the relationship between the physico-chemical properties and the local pillar structure was studied by Montarges et al., 1997; Montarges et al., 1999 and Bradley and coworkers (Bradley and Kydd, 1993a; b; d; e). They showed that the pillared products are structurally analogous to materials synthesized using Al13 polycations and possess similar physicochemical properties. Furthermore, the properties of Ga13- polyethyleneoxide (PEO) modified montmorillonites exhibit similar properties to
Al13-PEO-modified samples. (Montarges et al., 1999).
The use of Ga13 as a pillaring agent can be advantageous as it allows independent observation of the structural changes in both the pillar and the clay during calcination.
23 The basal spacing of Ga-pillared montmorillonite was found in some cases to collapse to
9.5 Å at 350°C due to the Ga polymer decomposing to Ga3+ cations (Kloprogge, 1994).
The reason for this collapse is still unclear.
Mixed Ga/Al pillared clays prepared by Coelho and Poncelet (1990) with partly hydrolyzed Ga/Al salt solutions with various Ga/Al ratios showed basal spacings between
18.7- 20.1 Å after intercalation, which decreased to 17.5Å after calcination .
Bradley and coworkers have contributed largely to the understanding of the hydrolysis of aqueous Ga and mixed Ga/Al solutions (Bradley, 1991; Bradley and Kydd, 1993c).
The hydrolysis of a mixed Ga/Al solution resulted in the presence of a sharp Ga NMR signal at 137.8 ppm that belongs to tetrahedral Ga indicating that the GaAl12 Keggin structure was formed. The GaO4 formed the central tetrahedron and was surrounded by
12 Al octahedra (Bradley and Kydd, 1993c). The basal spacings and surface areas of the
Ga/Al pillared clays were found to be very similar to those found by Coelho and Poncelet
(1991) and Gonzalez et al. (1991; 1992). The decrease in tetrahedral distortion by substituting GaIV for AlIV in the Keggin structure was found to be reflected in the increase of thermal stability in the following order: GaAl12- PILC > Al 13-PILC> Ga13-PILC
(Bradley and Kydd, 1993c). The Lewis acid strength decreases in the order Ga13 >Al13 >
GaAl12 whereas the Brönsted acid strength decreases in the opposite order (Bradley,
1991).
Bentonites pillared with Ga13, GaAl12 and Al13 were found to be thermally stable up to
700°C with catalytic data indicating the following order for activity for Ga13 >GaAl12>
24 Al13. The acidity studies showed more acid sites were present in all pillared clays, with both Lewis and Brönsted sites being detected (Berkeliev et al., 1992).
The methods of preparation and characterization of Ga12Al pillared clays have been reported by (Bellaloui et al., 1990; Holmgren, 1996, Kloprogge, 1994, Bagshaw and
Cooney, 1995, Caballero et al., 1995, Gil et al., 2000b). The current methods of making pillared clays may be divided into six types which are (1) Traditional method using Al13,
(2) Surface analysis /polymerization used for making chromia pillared clay (3) Via the sol gel.method (4) Direct intercalation of nano particles. (5) Using amines and surfactant. (6)
Using platelets themselves as pillars (De Stefanis and Tomlinson, 2006).
To improve the hydrothermal stability and catalytic properties of the pillared clays, mixed Al/metals pillared clays have been studied. Apart from mixed Ga/Al pillared clays, the most important combinations with Al are Fe, Si, and Zr. (Torii et al., 1992; Bergaya and Barrault, 1990; Bergaya et al., 1990; Doff et al., 1988).
The FeAl pillared clays can be prepared by using solutions of AlCl3, FeCl3 and NaOH.
Fe cations have been found to replace a few octahedral Al in the Al13 pillars (Bergaya and
Barrault, 1990; Carrado et al., 1988). There are two different methods of preparing mixed
Fe/Al complexes in solution. In the first method, Fe salts (nitrate or chloride) were added to a commercially available Al13 solution known as Chlorhydrol (ACH, commercially available from Reheis Chemical Company). A basal spacing of 18.8Å was observed after calcination at 480oC (Carrado et al., 1986a; Skoularikis et al., 1988). The second method comprises the co-hydrolysis of FeCl3 and AlCl3 by NaOH (Doff et al., 1988, Bergaya et al., 1990) or Na2CO3 before cation exchange with the clays; (Bergaya and Barrault,
25 1990). Further preparation and characterization of mixed Fe/Al pillared clays have been reported by several authors, e.g. (Kiricsi et al., 1997; Zhao et al., 1993; Huang et al.,
1991; Carrado et al., 1986a) and reviewed by Kloprogge (1998a).
Mixed Si/Al pillared clays were prepared by Gaaf et al. (1983) by using a solution of aluminum chlorhydrol and sodium silicate. The final products showed a weak XRD signal of an expanded structure together with the original non-expanded clay basal reflection indicating only partial intercalation. Another successful route was explored by
Occelli (1986) using sub-micron positively charged colloidal particles of alumina coated silica from. The basal spacing of the final products after treatment at 400°C was 18.8Å and they were stable up to 600°C in air. Other authors who have done extensive studies on hydroxy-silicoaluminum (HAS) include Tichit et al. (1988); Sterte and Shabtai
(1987b); Zhao et al. (1992). The basal spacings of the products were found to range from
17 Å to 26.5Å (Sterte and Shabtai, 1987a; Huang et al., 1991).
Occelli and his group has prepared mixed Zr/Al pillared clays using the REZAL67 and
RETZEL 36G (commercially available from Reheis Chemical Company) (Occelli, 1987).
6+ The formula of the complex is [Al8(OH)20ZrO] and the Al/Zr ratio is about 6.7. The cracking activity was intermediate between the cracking activities of Si/Al and Al pillared clays (Occelli, 1987).
Other mixed metal/Al pillared clays used for cracking that have been prepared include
UO2/Al, B/Zr/Al, and B/Si/Al ; (Wenyang et al., 1991a; Wenyang et al., 1991b, Carrado et al., 1986a, Carrado et al., 1986b). To produce the mixed UO2/Al pillared clay, a mixture of UO2(NO3)2 H2O and Al13 was used to react with Ca bentonite. The
26 incorporation of the uranyl has been found not to affect the cracking properties of the pillared clay (Carrado et al., 1986a). Wenzang et al. (1991) found that adding boron (B) to Si/Al or Zr/Al mixed metal pillared clays could improve the cracking ability of the pillared clays (Wenyang et al., 1991a; Wenyang et al., 1991b).
Although Al – pillared clay is the most popular material for industrial applications, other stable pillared clays have been prepared. The preparation, characterization and application of pillared clays and their catalytic properties have been discussed in a number of reviews including a special issue of Catalysis today in 1988 (Burch, 1988); and more recently by Kloprogge and coworkers (Ding et al., 2001; Kloprogge, 1998;
Kloprogge et al., 2005) and others (Baiker, 1996; Ballantine, 1986; De Stefanis and
Tomlinson, 2006; Gil et al., 2000a; Vaccari, 1998).
The most stable pillared clays that have been produced are Ti and Zr- pillared clays.
Ti pillared clays have been prepared and studied by a number of researchers: (Kurek,
1992; Bergaya et al., 1995; Chae et al., 1999, Liu et al., 2006, Binitha and Sugunan,
2006). Titanium has been found to form polymeric species in solution (Nabivanets and
Kudritskaya, 1967; Einaga, 1979). Various methods of preparing Ti complexes for pillaring processs have been investigated. The first method in which Ti complexes were formed comprised adding TiCl4 to 5 or 6 M HCl solution followed by dilution with distilled water and ageing from 3 hours up to as long as 20 days (Baksh et al., 1992;
Sterte, 1989; Bernier et al., 1991). This method produced highly acidic conditions which could cause leaching of a small amount of Al and Si from the clay structure (Bernier et al., 1991; Baksh et al., 1992).
27 The second method is based on the hydrolysis of various Ti alkoxides under milder acidic conditions (mostly 1 M HCl) (Farfan-Torres et al., 1992; Malla et al., 1989;
Yoneyama et al., 1989; Sychev et al., 1992; Del Castillo and Grange, 1993; Choudary et al., 1990). The d-spacing has been increased to around 24 – 25 Å for both methods. The experimental conditions of the Ti complexes-clay suspension were found to be critical to the morphology and texture of the pillared clays (Bernier et al., 1991; Sterte, 1989). The formation of an anatase phase outside the clay structure upon calcination has been reported based on XRD and IR spectral evidence (Bernier et al., 1991; Bagshaw and
Cooney, 1993).
The acidity of Ti pillared clays has been found to be increased during the synthesis process. Using temperature programmed desorption (TPD) of ammonia and pyridine adsorpsition/desorption experiments Bagshaw and Cooney (1993) and Bernier et al., suggested that the Lewis acid site must be located at the interface between pillar and the siloxane surface. (Bagshaw and Cooney, 1993; Bernier et al., 1991). The dealumination of the Ti-pillared clays during the synthesis process is the main reason for the increased acidity (Sychev et al., 1992). The formation of Brönsted acid sites has been reported from the pillaring of rectorite (Bagshaw and Cooney, 1993). Organic titanium pillared clays have been prepared by mixing a TiCl4 ethanol solution with a solution of glycerin and water. The final products showed basal spacings from 17.4 Å, to 21.3 Å (Jong et al.,
1994). In 1999, Ooka studied Ti pillared clays prepared by hydrothermal treatment. The study indicated that the relationship between the size of the crystallized TiO2 pillars and the average pore diameter was in good agreement, and could be controlled by changing the treatment conditions. The catalytic activity of the TiO2 pillars was found to be
28 enhanced by quantum-size effects: as the TiO2 pillar size decreased, the pillared montmorillonites exhibited higher catalytic activity and showed larger blue-shifts in their
UV absorption spectra. (Ooka et al., 1999). The photochemical and photocatalytic properties of the microcrystalline TiO2, which formed in the interlayer space after calcination of the pillared clays is very important for catalytic applications ; (Sychev et al., 1992; Yoneyama et al., 1989).
Other pillared clays using Zr, Cr, Fe, Si, V as pillars has been reviewed by Kloprogge
(Kloprogge, 1998), and Kloprogge and coworkers (Ding et al., 2001; Kloprogge et al.,
2005). In a more recent review in 2006, the method of making pillared clays for catalytic applications was discussed by De Stefanis and Tomlinson. In this review, small angle neutron scattering was mentioned as a new tool for studying interpillar distances (De
Stefanis and Tomlinson, 2006).
Boehmite AlOOH pillared montmorillonite has been studied by Sivakumar and coworkers (Sivakumar et al., 1997a; 1997b; 1994).
Imogolite and boehmite pillared clays have been prepared for some applications.
Imogolite Si2Al4O6(OH)8 is a semi-ordered structure which has been used successfully to intercalate montmorillonite and beidelite (Pinnavaia, 1992). The imogolite intercalated montmorillonite showed regular microporosity of adsorbates molecules with kinetic diameters from 8.6-10Ǻ (Johnson et al., 1988).
Recently, it has become possible to look at the morphology and the internal structure of clay minerals and pillared structures on an atomic scale with the development of
Atomic Force Microscopy (AFM) and high resolution Transmission Electron Microscopy
29 (HRTEM), vibrational spectroscopy and X ray photoelectron spectroscopy (XPS). There are still many questions about the relationship between the morphology and structure of clay minerals. The understanding of this structure will produce tailored clays for industrial applications.
1.4 Analytical techniques for studying clays and their modifications
1.4.1 X- Ray Diffraction
The discovery of X-rays by Röntgen in 1895 has given clay scientist a new method to study the fine structure of clay minerals. In 1923, Harding in Sweden published the fist
XRD pattern of the finest clay particles and found that the smallest clay particles were crystalline and had the same composition as other bigger particles in other samples
(Harding, 1923). The principles and application of the XRD method can be found in most textbooks covering geology, chemistry or material science (Moore and Reynolds, 1997;
Birks, 1969).
Since the first XRD pattern of clay was published, the technique has developed into an important tool for clay scientists for the characterization of crystalline materials. XRD can not only be used to study the structure of pure clays, mixed layer clays, but also to study changes in structure when expanding and pillaring clays with organic and/or inorganic molecules. Figure 1.6 shows the changes in XRD patterns of montmorillonite, intercalated montmorillonite and pillared montmorillonite with Ga pillars.
Mixed-layered clay structures, such as illite/smectite, can be determined by XRD, (Pons et al., 1995).
30 With the development of computing technology, there are a number of programs for the interpretation of XRD patterns. For example Krumm in 1999 discussed how to apply
WinStruct software to simulate XRD patterns of oriented clay minerals (Krumm, 1999).
Another program called “an expert system” has been used by Plancon in 2000 to interpret the mixed layer illite/ smectite patterns (Plancon and Drits, 2000). Other programs that have been used for clay structure are NEWMOD and mudmaster. (Denis Eberl et al 2000;
Plancon et al 1990).
The ratio of illite/ smectite has been found to be very stable in sedimentary bedding and could be used as an indicator to study basin history and correlation. Changes in structure of the clay particles, in particular the stacking, can be seen from the variety of d spacings which allows scientists to identify the different species with ordered or disordered structures.
Heating stage XRD is another way of studying clay minerals. The temperature can be controlled and XRD patterns can be collected at certain temperatures allowing the direct observation in situ of any structural changes taking place.
Aceman et al. (1997) studied Al-pillared clays preparing from five different smectites using heating stage X-ray diffraction. These clays, dioctahedral beidellite and montmorillonite and trioctahedral saponite, hectorite and Laponite, differ in the origin of isomorphic substitution and represent a series of decreasing basicity along the siloxane plane. They concluded that the Keggin ion lost its structural water around 200°C and dehydroxylated above 350°C. After heating to 500°C to 600°C, this polymeric cation,
31 which is thought to form the Al2O3 oxide, did not rehydrate. (Aceman et al., 1997)
(Aceman et al., 2000).
Recent developments in diffraction techniques include techniques such as small angle neutron scattering which can provide information about the inter pillar distances and permits the modelling of the pore structure (De Stefanis and Tomlinson, 2006).
32
17.9 Å
Ga13 calcined
19.9 Å
Ga13 exchanged
Counts
14 Å
Starting montmorillonite
3 8 13 18 23 degrees 2θ Cukα
Fig. 1.6 XRD patterns of clay and pillared clays showing structural changes from starting
montmorillonite to Ga exchanged montmorillonite and after calcination process
33 1.4.2 Electron Microscopy and X- Ray Microanalysis
In electron microscopy, a focused electron beam is used to illuminate a specimen.
Electrons are produced from a tungsten, LaB6 or field emission cathode, accelerated along the column and are focused by electromagnetic lenses. At the same time X-ray are generated from the interaction between the electrons and the specimen which can be used for elemental microanalysis of the specimen. Scanning electron microscopy can provide higher resolution information of clay materials than optical microscopy, down to a few nanometers in practice, but only of surfaces of the particles. It does provide a very good depth of field for three-dimensional imaging of clay particles. However, transmission electron microscopy with electron diffraction and microanalysis provides very high resolution images of clay particles and their structures. The first TEM was introduced in
1931 by the work of Knoll and Ruska (Marton, 1968) and since then has rapidly expanded for studying clay structures. The formation of an image in an optical microscope, transmission electron microscope and scanning electron microscope can be seen in Figure 1.7.
34
Fig. 1.7 Image formations in optical microscope, transmission microscope and scanning
electron microscope In the optical microscope the source is light bulb whereas in the
SEM and TEM the source is an electron beam (Buseck, P.R., 1992)
35 In an early study of clay minerals using electron microscopy techniques, TEM was mentioned in the Symposium held at the 7th International Clay Conference, Bologna,
Italy as one of the important advanced techniques for clay mineral analysis (Fripiat,
1982). The principle of transmission electron microscopy and application to clay structures including electron diffraction of a small particle, have been discussed by
Goldstein et al (1981), Buseck (1992). Occelli et al. (1986) used TEM to analyse the macro structure of pillared and delaminated hectorite catalysts. Two types of structure were reported: the “house of card” and the “face to face” structures. (Lipsicas et al.,
1984; Occelli, 1987). In kaolinite, high resolution transmission microscopy (HRTEM) revealed three types of surface layers crystals and defects present in the structure. Type 1 has a 7Å kaolinite surface layer, type 2 has a 10Å pyrophyllite like layers and type 3 has one or several 10Å collapsed smectite like layers (Ma and Eggleton, 1999).
The electron diffraction pattern of the kaolinite structure was used to determine the hydrogen atom position in the crystal structure (Raupach et al., 1987), and to study mix- layer illite/smectite (Veblen et al., 1990). Results from high-resolution transmission electron microscopy and lattice-energy calculations of mixed layering of illite-smectite showed the unit layers are O0.5TITO0.5 with O, T and I are octahedral , tetrahedral and interlayer sheets (Olives et al., 2000; Klimentidis and Mackinnon, 1986; Kohyama and
Fukami, 1982).
Crozier et al.,(1999) used the combination of TEM techniques such as bright field, dark field imaging and energy dispersive X- ray spectroscopy (EDX) to determine the location of the pillars in the [001] direction (or the basal plane projection) of the Zr- pillared montmorillonite. The results showed that zirconia pillars have an irregular shape
36 and the sizes were <50 Å, which indicated that zirconia dispersion was not ideally distributed throughout the whole sample.(Crozier et al., 1999).
In geological applications, TEM and high resolution TEM have been used to study the mixed layer of illite-smectite, to understand the diagenetic changes in clay minerals and to reconstruct the temperature history of the basin. Few SEM studies of clay minerals have been undertaken, except for some studies mainly focusing on kaolinites (Zbik and
Smart, 1998) and have not compared the morphology of the pillared clays with starting materials.
The exposure of the clays to the electron beam can lead to some structural damage.
The loss of some alkali elements (K, Na, Mg), low atomic number elements (Al), and high atomic number elements ( Fe, Ti ) during the process of microanalysis of silicates
(kaolinite, smectite, biotite, muscovite and pyrophyllite) at about 300 kV, were also reported (Ma et al., 1999).
The ability of smectites to expand when in contact with water causes complication with sample preparation for SEM and particularly TEM. Several attempts have been made to overcome this problem. Resin embedding techniques seem to be a popular for clay samples preparation (Bateman, 1995). Two popular types of resins, which are used by clay mineralogists, are LR White and Spurr (Spurr, 1969; Bateman, 1995). Araldite has previously been used to embed mixed layer illite-smectite and this was cut with diamond knife to obtain thin sample for TEM (Olives et al., 2000). The method of preparing oriented clay particles for TEM has been discussed by Duong et al. (2005). Ga
37 and Al pillared montmorillonite were embedded in Spurr resin, and sectioned in an ultramicrotome using a diamond knife (Duong et al., 2005).
Scanning electron microscopy and X-ray microanalysis permit an evaluation of surface morphology, chemical composition and the overall structure of clay particles (Figure 1.8,
Figure 1.9). This allows a comparison to be made between the final products and the starting materials which enables the results to be related to other observations made by
AFM and other techniques. The introduction of a heating stage, which could be fitted with an SEM, TEM, or field emission SEM and TEM and environmental SEM, would help clay mineralogists to study in much more detail the changing structure of pillared clays with increasing pressure and temperature. Kloprogge and coworkers used ESEM and low voltage SEM to study mineralogy of a bauxite sample (Kloprogge et al., 2006) and characteristics of uncoated minerals have also been studied (Lin et al., 2005). This work may open up the new areas in future, which can overcome the problem of sample preparation due to the expansion of montmorillonites when in contact with water.
Lee and Peacor (1986) noted that laurylamine hydrochloride destroys the original texture of the samples and should be used very carefully when characterize mixed layer illite/smectite clays.
38
Fig. 1.8 SEM image of pillared montmorillonite (Top) and TEM image of particle of Montmorillonite (bottom) 39
Fig. 1.9 SEM images of Kaolinite in sandstone shows typical book structure
40 1.4.3 Atomic Force microscopy
The atomic force microscope (AFM) was invented in the 80s by Binnig et al., 1986. It gives topographic images by using highly local interactions between a very fine tip and the surface of a sample. The tip and cantilever are made as one unit (Figure 1.10), and are made of silicon nitride (Si3N4) or silicon (Si). The shape of the tip can be pyramidal or conical depending on applications and ultra sharp tips are also available. The tip is scanned across the surface and its movement is used to form a high-resolution image.
Because the interaction between the tip and a sample interaction is a very small force, the
AFM does not require conductive samples and is therefore of interest for the study of mineral surfaces. The clay minerals with layered structure usually have a good cleavage so they are potentially very good samples for AFM study. On the other hand, the scanning tunneling microscope (STM) relies on a tunneling current between the tip and surface hence samples of this method need to be at least semi conductive.
AFM has been used to study the surface of clay minerals on a nanoscale (Zbik and
Smart, 1998; Hartman et al., 1990) but very few studies have been found on AFM directly of pillared clays
41
Fig. 1.10 Typical diagram of an atomic force microscope showing a tip mounted on a cantilever (C). The tip scans over the surface of the sample. The laser light from the back of the cantilever reflects to a two element
photosensitive diode (PSD) and the output signal can be used in a feedback loop to control the vertical position of the sample (Binnig et
al., 1986)
42 The AFM images of illite and montmorillonite collected in air and under humidities near 80% showed an ideal hexagonal array of oxygen ions in the siloxane surface of a 2:1 clay mineral. The distance from the center of the ring to another across the hexagonal is about 11Å (Hartman et al., 1990). The hexagonal pattern of bright spots has also been seen on the surface of Al- pillared rectorite as the collection of basal oxygens in the tetrahedral sheet of the pillared clay (Occelli, 1994). The AFM has been used for characterization of small kaolinite particles and compared with TEM results. AFM images of kaolinite from Weipa deposit, northern of Queensland, Australia showed steps on the basal plane of the kaolinite crystals in agreement with TEM images (Zbik and
Smart, 1998). The ratios of edge surface area (ESA) to total surface area (TSA) of kaolinite particles has been studied and compared with the BET method. The edge surface area is 18.2- 30.0% of the total surface area depending on the kaolinite standard.
(Bickmore et al., 2002; Zbik and Smart, 1998). Wu et a. (2001) imaged untreated montmorillonite and pillared montmorillonite with the Digital Nanoscope III AFM in contact mode and pointed out that the surface of untreated montmorillonite appeared more intergrated whereas pillared montmorillonite showed more layer steps, which could be the pillaring effect of the Keggin ions.
The effect of the pillaring process can be seen from the AFM images of Al13 pillared montmorillonite. A powder of pillared clays was formed by pressing the sample at 34
000kPa, gluing it on to a steel disk with epoxy resin and then imaging with a Nanoscope
III Digital Instrument AFM. The images showed that the surface of the Al13 pillared clays is free of Al species. The dimensions of the white spots on the surface are in good
43 agreement with the unit cell of other pillared montmorillonite but larger than the parent montmorillonite (Campos et al., 1998).
The AFM non-contact mode, which uses the weak attractive force between the tip and the sample to get the true atomic resolution has been reported recently (Lantz et al., 2001)
In the experiments performed by a low temperature AFM, operating with an ultrahigh vacuum environment with a commercial silicon cantilever, the short range chemical bonding forces between the apex of the silicon tip and the specific atomic sites on a silicon sample have been measured (Lantz et al., 2001).
Sayed Hassan et al. (2006) studied the distribution of edge and basal surface of phyllosilicate particles using low pressure argon adsorption and AFM analysis. The results show very good agreement between the two methods .
1.4.4 Vibrational Spectroscopy and solid state NMR
Vibrational spectroscopic techniques involve the interaction of electromagnetic waves
(light) and the vibration modes of molecules. Infrared and Raman spectroscopic techniques are the most common techniques, which have been used for studying minerals, organic and inorganic materials. Further reading about those techniques can be found from the publications of Banwell (1983), Farmer (1974), Gadsden (1975). Griffith
(1987), McMillan and Hofmeister (1988), Ross (1972). Raman spectroscopy is a non- destructive technique, which is based on the experiment by Sir C.V. Raman in 1928
(Raman, C.V and Krishnan, K.S., 1928). The difference between Raman and infrared is that Raman spectroscopy looks at the change in polarization of molecules where infrared looks at the changes in dipole moment of molecules. Raman microscopy and Fourier
44 transform Raman have been used to determine the location of the OH of the kaolinite structure (Frost and van der Gaast, 1997; Frost et al., 1998a; Frost et al., 1999a) and for the identification of minerals.
Nuclear Magnetic Resonance (NMR) is a domain of absorption spectroscopy. This method is based on the principle that certain nuclei possess a magnetic moment which interacts with an applied magnetic field. The NMR studies of a number of clay minerals and pillared clays have indicated that chemical shift in Al13 pillars and the relation between OH/Al molar ratio and the Al 13 concentration (Kloprogge, 1992).
NMR has been used to recognise the tetrahedral Al and octahedral Al in pillared smectite structure. Pillared beidellite showed a deep structural transition that can be described as the growth of a pillar network grafted on the structure of the clay. There is no reaction in pillared smectites with no substitution in the tetrahedral (hectorite and
Laponite) during the calcination process (Plee et al., 1985a; 1987).
The combination of vibrational and NMR techniques can give more detail understanding of the arrangement of atoms and their environment. However as is indicated in the literature, the position of the hydroxyls and the oxidation state of the clay minerals are very difficult to determine by these two techniques.
1.4.5 X-Ray Photoelectron spectroscopy
X-ray photoelectron spectroscopy (XPS) was first introduced in 1981 through the work of Siegbahn and coworkers in 1967 (Siegbahn et al., 1967) and since has been applied as a surface analysis technique for studying surface materials and thin films
45 including minerals. Clay minerals, however, due to the small particle size and special layer structure, are very good candidates for this technique. XPS has been regarded by clay mineralogists as one of the advanced techniques for studying of clay minerals
(Fripiat, 1982).
X-ray photoelectron spectroscopy uses photons from an X-ray source of known energy to eject core electrons from atoms at the surface of solids and the instrument is used to measure the kinetic energy of that electron. Table 1.2 shows the atomic and corresponding X-ray notations used in XPS. The binding energy of the free electron is determined by the relationship:
hυ = Ekinetic + Ebinding
where hυ represents the energy of the photon, Ekinetic the kinetic energy of the free electron and Ebinding the original (negative) binding energy. Hence this technique measures, at a given hυ, the kinetic energy of the emitted electrons and can therefore be used to determine the binding energy of each element present in the solid near the surface. Changes in the binding energy reflect changes in the local environment (nearest neighbours and next nearest neighbours) around a particular atom. At the same time as the core electrons are removed, it is also possible to remove Auger electrons. Figure 1.11 shows the photoelectric effect, which produces both a photoelectron and an Auger electron as a result.
It will be clear that only those electrons can be detected with a binding energy lower than the energy of the photons used. Most common photon sources are monochromatic
Al-Kα and Mg-Kα X-rays resulting in photons with energies of 1486.6 and 1253.6 eV,
46 respectively, while older instruments often use non-monochromatic Mg X-rays. In these cases the technique is known as X-ray Photo-electron Spectroscopy (XPS) or Electron
Spectroscopy for Chemical Analysis (ESCA). The second name points to the fact that by analysing the energy of the photoelectrons, their energy levels and thus the chemical identity can be determined. Especially for surfaces of solid materials XPS is a good qualitative and quantitative analytical technique.
47
Fig. 1.11 Diagram showing Photoelectron and Auger electron processes (Wagner, 1979 a)
48
Table 1.2 Atomic and X-ray notations (Wagner, 1979a)
49 Photoelectron spectroscopy as an analytical tool has only received limited interest in the field of clay mineral science. Photoelectron spectroscopy, together with Auger electron spectroscopy, gives information about the positions of the energy levels in atoms or molecules. XPS has been widely used to study for thin films and for other surface analysis. The application of this technique to study clay minerals will result in information about the band structure of these materials and the local atomic environment
(Duong et al., 2005; Duong et al., 2006).
1.5 General discussion
Clay minerals are naturally produced by weathering processes and have structures that belong to the group of phyllosilicates. Due to their special structures and properties, they can be mined in large quantities at low cost, have a wide variety of industrial applications, e.g. in the brick and tile industry. In addition a large amount of research has been focused on modifying clay to obtain a suitable structure for special industrial purposes. A good example of this is the delamination of kaolinite used in the production of paper. Intercalated and pillared clays have been studied and produced for catalytic purposes in the oil industry.
Due to the extremely small particle size and the complex sheet structure, many techniques have been used to study clay minerals. The morphology can be directly studied by techniques such as AFM and SEM and TEM. X-ray interactions with the crystal structure such as these used in techniques such as XPS or XRD can be used to obtain information not only on the crystal structure but also on its chemical composition.
Vibrational spectroscopy, including techniques such as Mid-infrared and Raman
50 spectroscopy, allow the investigation of the molecular structure of clay minerals based on measurement of the vibrations between atoms.
Many types of pillared clays have been produced to suit industrial applications.
Techniques used for producing pillared clays vary from simple hydrolysis followed by calcination in either an oven or a hydrothermal vessel to microwave treatment and with the pillars ranging from small relatively simple complexes to large metal complexes. The starting clays are mainly from the smectite group with montmorillonite and beidelite as the most popular species. Kaolinite has been used for intercalating of organic molecules with only limited industrial applications, while smectite clays intercalated with specific organic molecules to produce organoclays have a wide range of uses in environmental applications such as treatment of wastewater to remove unwanted organic species.
From the literature, a large variety of pillared clays have been produced based on different metal complexes and with different methods depending on the purposes of application. Although many papers have been published on the general structure of pillared clays, the detailed structure of the pillars and the type of bonding of the pillars to the silicate layers is still not fully understood. The literature review given in this chapter shows a large number of good research papers focusing on the general structure and the applications of pillared clays but hardly any research concentrating on high-resolution techniques such as AFM or TEM to study the pillar structure and its bonding to different types of clay layers. Current literature does not show any publication on the use of XPS as a technique to learn more about pillared clays.
51 This type of research based on the use of advanced techniques for studying clay products will provide a better understanding of the structure of pillared clays, further the relationship between the pillar structure and bonding mechanism to the silicate structure and the final catalytic properties of the pillared clays will be investigated.
1.6 Aims and objectives
The aims of this study are: 1) to develop a more efficient, less waste producing and less time consuming method for the preparation of pillared clays, 2) to gain a better understanding of the structure of pillared clays, and 3) to obtain a better understanding of the bonding mechanisms between the pillars and the silicate structure. As the pillared clay structure only varies along the c axis and in the interlayer region, this study will be focused mainly on the cross sections of the clays. For this reason, the similarity of the structures of Ga and Al pillars will be taken into account when preparing pillared clays for studies with TEM and XPS. A new method of preparing TEM samples in cross section will be developed for use with X-ray mapping and microanalysis.
Characterisation of the morphology of pillared montmorillonite with electron microscopy including scanning electron microscopy, microanalysis, X-ray mapping and transmission electron microscopy will be compared with the results from X-ray photoelectron spectroscopy.
The objectives of the study are to obtain a better understanding of the pillar structure and the bonding of the pillars to the silicate structure. This will be achieved by direct methods, such as TEM, and indirect methods such as XPS and XRD. The study will also look at the binding energies of various atoms in different types of clays, in particular Si,
52 Al and O and how these binding energies change upon the introduction of the Ga and Al pillars. This will help to better understand the results from electron microscopy to which will be useful to determine the effects of pillars and clay type on the final structure with emphasis on Al13 and Ga13 pillared montmorillonites.
53 1.7 References
Aceman, S., Lahav, N. and Yariv, S., 1997. XRD study of the dehydration and rehydration behavior of Al-pillared smectites differing in source of charge. J. Therm. Anal., 50(1-2): 241-256. Aceman, S., Lahav, N. and Yariv, S., 2000. A thermo-XRD study of Al-pillared smectites differing in source of charge, obtained in dialyzed, non-dialyzed and washed systems. Appl. Clay Sci., 17(3-4): 99-126. Auer, H. and Hofmann, H., 1993. Pillared clays: characterization of acidity and catalytic properties and comparison with some zeolites. Appl. Catal., A, 97(1): 23-38. Bagshaw, S.A. and Cooney, R.P., 1993. FTIR spectroscopic studies of surface-site probe species on metal oxide pillared clay surfaces. Proc. SPIE-Int. Soc. Opt. Eng., 2089 (9th International conference on fourier transform spectroscopy, 1993): 162-3. Bagshaw, S.A. and Cooney, R.P., 1995. Preparation and Characterization of a Highly Stable Pillared Clay: GaAl12-Pillared Rectorite. Chem. Mater., 7(7): 1384-9. Baiker, A., 1996. Heterogeneous catalysis. From fundamentals to reaction engineering. Chimia, 50(3): 65-73. Bailey, S.W., 1988. Chlorites: structures and crystal chemistry. In: Hydrous phyllosilicates (exclusive of micas). Rev. Miner., 19, 1 pp. Baksh, M.S., Kikkinides, E.S. and Yang, R.T., 1992. Characterization by physisorption of a new class of microporous adsorbents: pillared clays. Ind. Eng. Chem. Res., 31(9): 2181-9. Balek, V., Malek, Z. and Klosova, E., 1998. Emanation thermal analysis of intercalated montmorillonitic clay. J. Therm. Anal. Calorim., 53(2): 625-631. Ballantine, J.A., 1986. The reactions in clays and pillared clays. Nato Asi Ser., 165 (Chem. React. Org. Inorg. Constrained Syst.): 197-212. Banwell, C.N., 1983. Raman Spectroscopy, Fundamentals of Molecular Spectroscopy, pp. 124-154. Barrer, R.M. and MacLeod, D.M., 1955. Activation of montmorillonite by ion exchange and sorption complexes of tetraalkylammonium montmorillonites. Trans. Farady Soc., 51: 1290-1300.
54 Bateman, J.A., 1995. Mineralogical and geochemical traits of the Egret Member oil source rock (Kimmeridgian), Jeanne d'Arc Basin, offshore Newfoundland, Canada.Dalhousie University, Halifax, Canada, Master thesis 200p. Bellaloui, A., Plee, D. and Meriaudeau, P., 1990. Gallium containing pillared interlayer clays: preparation, characterization and catalytic properties. Appl. Catal., 63(2): L7-L10. Bergaoui, L. et al., 1993. Pillared clays - aging and stability in the presence of heavy metals. J. Soc. Alger. Chim., 3S(2): 51-66. Bergaoui, L., Lambert, J.F., Suquet, H. and Che, M., 1995. Study of the adsorptive properties of a pillared clay towards Cu2+ and Cd2+ as a function of pH. J. Chim. Phys. Phys.-Chim. Biol., 92(7-8): 1486-505. Bergaya, F. and Barrault, J., 1990. Mixed aluminum-iron pillared laponites: preparation, characterization and their catalytic properties in syngas conversion. Pillared Layered Struct.: Curr. Trends Appl., [Proc. Workshop]: 167-84. Bergaya, F., Gallot, P., Setton, R. and Damme, H.V., 1995. Pillaring of montmorillonite by titanium oxide. Clays Controlling Environ., Proc. Int. Clay Conf., 10th: 151-6. Bergaya, F., Hassoun, N., Gatineau, L. and Barrault, J., 1990. Mixed Al-Fe pillared laponites: preparation, characterization and catalytic. Scient. Bases Prep. Heter. Catal. 5th Int. Symp.: Louvain-la Neuve, Belgium. Berkeliev, T.K., Pautov, L.A. and Ignatenko, K.I., 1992. Fraipontite from Turkmenstan. Mineral. Zh., 14(2): 89-91. Bernier, A., Admaiai, L.F. and Grange, P., 1991. Synthesis and characterization of titanium pillared clays . Influence of the temperate of preparation. Appl. Catal., 77: 269-281. Bickmore, B.R., Nagy, K.L., Sandlin, P.E. and Crater, T.S., 2002. Quantifying surface areas of clays by atomic force microscopy. Amer. Min., 87: 780- 783. Binitha, N.N. and Sugunan, S., 2006. Preparation, characterization and catalytic activity of titania pillared montmorillonite clays. Micropor. Mesopor. Mater., 93(1-3): 82- 89. Binnig, G., Quate, C.F. and Gerber, C., 1986. Atomic Force Microscope. Phys. Rev. Letters, 56: 930- 933.
55 Birks, L.S., 1969. X-Ray Spectrochemical Analysis. Interscience publishers, New York Bodorado, S., Figueras, F. and Garrone, E., 1994. IR study of Broensted acidity of Al-pillared montmorillonite. J. Catal., 147(1): 223-30.
Bradley, M. and Kydd, A., 1993a. Gallium (Ga13), aluminum (Al13), gallium-aluminum
(GaAl12), and chromium-pillared montmorillonites: acidity and reactivity for cumene conversion. J. Catal., 141(1): 239-49. Bradley, S.M., 1991. Synthesis and characterization of heteropoly and isopoly oxometal cations and their use in the formation of pillared clay mineral catalysts, University Calgary, Canada 326 pp.
Bradley, S.M. and Kydd, R.A., 1993b. A comparison of the catalytic activities of Ga13-,
Al13-, GaAl12-, and chromium-pillar interlayered clay minerals and Ga-H-ZSM-5 zeolite in the dehydrocyclodimerization of propane. J. Catal., 142(2): 448-54. Bradley, S.M. and Kydd, R.A., 1993c. Comparison of the species formed upon base hydrolyses of gallium(III) and iron(III) aqueous solutions: the possibility of + existence of an [FeO4Fe12(OH)24(H2O)12]7 polyoxocation. J. Chem. Soc., Dalton Trans.(15): 2407-13.
Bradley, S.M. and Kydd, R.A., 1993d. GA13, Al13, GaAl12, and Chromium-Pillared Montmorillonites: Acidity and Reactivity for Cumene Convertion. J. Catal., 141: 239-249.
Bradley, S.M. and Kydd, R.A., 1993e. Gallium (Ga13), aluminum (Al13), gallium-
aluminum (GaAl12), and chromium-pillared montmorillonites: acidity and reactivity for cumene conversion. J. Catal., 141(1): 239-49. Bradley, S.M., Kydd, R.A. and Yamdagni, R., 1990. Study of the hydrolysis of combined aluminum(3+) and gallium(3+) aqueous solutions: formation of an extremely + stable dodecaaluminiogallium (GaO4Al12(OH)24(H2O)127 ) polyoxycation. Magn. Reson. Chem., 28(9): 746-50. Brindley, G.W. and Maksimovic, Z., 1974. The nature and nomenclature of hydrous nickel-containing silicates. Clay Miner., 10: 271-277. Brindley, G.W. and Sempels, R.E., 1977. Clay Miner., 12: 229. Burch, R.E., 1988. Pillared clays. Catal. Today, 2: 185.
56 Buseck, P.R., 1992. Principles of transmission electron microscopy. In: Buseck and Peter (Editors), Minerals and reactions at the atomic scale; transmission electron microscopy. Rev.n Miner., Mineralogical Society of America, Washington, DC, United States, pp. 1-36. Caballero, L., Dominguez, J.M. and Lorenzo, J., 1995. On the structural stability of smectites pillared with RE-Al and Al-Ga polyoxocations. Prepr. - Am. Chem. Soc., Div. Pet. Chem., 40(2): 283-5. Campos, D., Eic, M. and Occelli, M.L., 1998. Adsorption of n-decane and perfluorotributylamine in pillared and expanded clays. Fundam. Adsorpt., [Conf.], 6th: 129-134. Carrado, K.A., Kostapapas, A., Suib, S.L. and Coughlin, R.W., 1986a. Physical and chemical stabilities of pillared clays containing transition metal ions. Solid State Ionics, 22(1): 117-25. Carrado, K.A., Suib, S.L., Skoularikis, N.D. and Coughlin, R.W., 1986b. Chromium(III)- doped pillared clays (PILC's). Inorg. Chem., 25(23): 4217-21. Carrado, K.A., Thompson, A.R., Winans, R.E. and Botto, R.E., 1988. Structural studies of pillared clays and modified pillared clays. Mater. Res. Soc. Symp. Proc., 111(Microstruct. Prop. Catal.): 277-82.
Chae, H.J. et al., 1999. Characteristics of V2O5/Ti-PILC catalyst for the reduction of NO
by NH3. Stud. Surf. Sci. Catal., 125 (Porous Materials in Environmentally Friendly Processes): 595-602. Chen, P.Y., Lin, M.L. and Zheng, Z., 1997. On the origin of the name kaolin and the kaolin deposits of the Kauling and Dazhou areas, Kiangsi, China. Appl. Clay Sci., 12(1-2): 1-25. Chevallier, S., Franck, R., Suquet, H., Lambert, J.F. and Barthomeuf, D., 1994. Al- Pillared saponites. Part 1. IR studies. J. Chem. Soc., Faraday Trans., 90(4): 667- 74. Choudary, B.M., Valli, V.L.K. and Durga Prasad, A., 1990. An Improved Asymmetric Epoxidation of Allyl Alcohols using Titanium-pillared. J. Chem. Soc. Chem. Commun.: 1186-1187.
57 Coelho, A.V. and Poncelet, G., 1991. Gallium, aluminium and mixed gallium-aluminium pillared montmorillonite : Preparation and characterization. Appl. Catal., 77(2): 303-314. Costanzo, P.M. and Giese, R.F., 1990. Ordered and disordered organic intercalates of 8.4Å synthetically hydrated kaolinite. Clays Clay Miner., 38: 160. Crozier, P.A., Pan, M., Bateman, C., Alcaraz, J.J. and Holmgren, J.S., 1999. Direct imaging of zirconia pillars in montmorillonite by analytical electron microscopy. Clays Clay Miner., 47(6): 683-687. Cruz, M.D.R. and Franco, F., 2000. Thermal behavior of the kaolinite-hydrazine intercalation complex. Clays Clay Miner., 48(1): 63-67. Dennis D. Eberl, V. Drits, J. Srodon, and R. Nüesch 2000. Mudmaster. A program for calculating crystallite size distributions and strain from the shapes of X-ray diffraction peaks. US Geological Survey,53 p De Stefanis, A. and Tomlinson, A.A.G., 2006. Towards designing pillared clays for catalysis. Catal. Today. Nano-structured materials for catalysis, 114(2-3): 126- 141. Del Castillo, H.L. and Grange, P., 1993. Preparation and catalytic activity of titanium pillared montmorillonite. Appl. Catal.A, 103: 23-34. Ding, Z., Kloprogge, J.T., Frost, R.L., Lu, G.Q. and Zhu, H.Y., 2001. Porous clays and pillared clays-based catalysts. Part 2: A review of the catalytic and molecular sieve applications. J. Por.s Mater., 8(4):273-293. Doff, D.H., Gangas, N.H.J., Allan, J.E.M. and Coey, J.M.D., 1988. Preparation and characterization of iron oxide pillared montmorillonite. Clay Miner., 23(4): 367- 77. Duong, L., Bostrom, T., Kloprogge, T. and Frost, R., 2005. The distribution of Ga in Ga- pillared montmorillonites: A transmission electron microscopy and microanalysis study. Micropor. Mesopor. Mater., 82(1-2): 165-172. Duong, L., Kloprogge, T., Frost, R. and Wood, B., 2006. High resolution XPS study of the internal structure of Al- and Ga-pillars in pillared clay catalysts (Submitted). Appl. Clay Sci.
58 Einaga, H., 1979. Hydrolysis of Titanium(IV) in Aqueous (Na,H)Cl Solution. J. Chem. Soc.: Dalton Trans., 1917-1919. Espinosa, J., Gomez, S. and Fuentes, G.A., 1994. Synthesis and MAS-NMR analysis of highly stable pillared clays. Stud. Surf. Sci. Catal., 84(Zeolites and Related Microporous Materials, Pt. A): 283-90. Farfan-Torres, E.M., Sham, E. and Grange, P., 1992. Pillared clays: preparation and characterization of zirconium pillared montmorillonite. Catal. Today, 15(3-4): 515-26. Farmer, V., 1974. The Infrared Spectra of Minerals. Mineralogical Society, England. Figueras, F.Klapyta, Z., Massiani, P., Mountassir, Z., Tichit, D., Fajula, F., Gueguen, C., Bousquet, J., and Auroux, A., 1990. Use of competitive ion exchange for intercalation of montmorillonite with hydroxy-aluminum species. Clays Clay Miner., 38(3): 257-64. Fripiat, J.J., 1982. Advanced Techniques for clay mineral analysis (Invited contributions from the Symposium held at the 7th International Clay Conference, September 6-12,1981, Bologna , Italy). Elservier Scientific Publishing Company., Printed in The Netherlands. Frost, R.L., 1997. The structure of the kaolinite minerals - a ft-raman study. Clay Miner., 32(1): 65-77. Frost, R.L., Kloprogge, J.T., Kristof, J. and Horvath, E., 1998a. Raman spectroscopy of hydrazine intercalated kaolinites, 3rd Australian Conference on Vibrational Spectroscopy, University of Melbourne, Australia, pp. 180-181. Frost, R.L., Kloprogge, J.T., Russell, S.C. and Szetu, J.L., 1999a. Vibrational spectroscopy and dehydroxylation of aluminum (oxo)hydroxides. Appl. Spectrosc., 53(4): 423-433. Frost, R.L., Kristof, J., Horvath, E. and Kloprogge, J.T., 1998b. Molecular structure of dimethylsulphoxide intercalated kaolinites, 16th Australian Clay Minerals Society Conference, CIDC, QUT. Frost, R.L., Kristof, J., Horvath, E. and Kloprogge, J.T., 2000a. Kaolinite hydroxyls in dimethylsulphoxide intercalated kaolinites at 77K - a Raman spectroscopic study. Clay Miner., 35: 443-454.
59 Frost, R.L., Kristof, J., Paroz, G.N. and Kloprogge, J.T., 1999b. Intercalation of kaolinite with acetamide. Phys. Chem. Miner., 26: 257-263. Frost, R.L., Kristof, J., Rintoul, L. and Kloprogge, J.T., 2000b. Raman spectroscopy of urea and urea-intercalated kaolinites at 77 K. Spectrochim. Acta A, 56: 1681- 1691. Frost, R.L., Lack, D.A., Paroz, G.N. and Tran, T.H.T., 1999c. New techniques for studying the intercalcation of kaolinites from Georgia with formamide. Clays and Clay Miner., 47(3): 297-303. Frost, R.L. and van der .Gaast, .S.J., 1997. Kaolinite hydroxyls; a Raman microscopy study. Clay Miner., 32(3): 471-484. Gaaf, J.,Van Santen, R.,Knoester, A,.Van Wingerden, B.,1983. Synthesis and catalytic properties of pillared nickel substituted mica montmorillonite clays. J. Chem. Soc., Chem. Commun. (120, 655- 657 Gadsden, J.A., 1975. Infrared Spectra of Minerals and Related Inorganic Compounds. Butterworths, England, 75 pp. Gil, A., Gandia, L.M. and Vicente, M., 2000a. Recent Advances in the Synthesis and Catalytic Applications of Pillared Clays. Catal. Rev., 42: 145-212. Gil, A., Vicente, M.A. and Gandia, L.M., 2000b. Main factors controlling the texture of zirconia and alumina pillared clays. Micropor. Mesopor. Mater., 34(1): 115-125. Goldstein, J.J., Newbury, D., Echlin, P., Joy, D., Fiori,.C., and Lifshin, E., 1981. Scanning Electron Microscopy and X-ray Microanalysis. Plenum Press, New York, 673 pp Gonzalez, F., Pesquera, C., Benito, I. and Mendioroz, S., 1991. Aluminium -Gallium Montmorillonite with high thermal stability. J. Chem. Soc. Chem. Commu.: 587- 588. Gonzalez, F., Pesquera, C., Blanco, C., Benito, I. and Mendioroz, S., 1992. Synthesis and Characterization of Al- Ga Pillared clays with high thermal and hydrothermal stability. Inorg. Chem., 31: 727-731. Gonzalez Luz, W., Meyer Jaraba, A., Machado, F. and Rosa-Brussin, M., 1990. Alkylation of toluene over aluminosilicates. Rev. Soc. Venez. Catal., 3(1): 85-9.
60 Griffith, W.P., 1987. Spectroscopy of Inorganic-based Materials. J. Wiley & Sons Ltd, 119 pp. Grim, R., 1988. The History of the development of clay mineralogy. Clay Clay Miner., 36(2): 97-101. Guven, N. and Hover, W.F., 1979. A Vanadium smectite. Clay Miner., 14: 241-245. Harding, A., 1923. Eine rontgenographische Methode kristalline und kryptokristalline Substanzen zu Identifizieren. Z. Kristallogr, 58: 108-112. Hartman, H. et al., 1990. Molecular -scale imaging of clay mineral surfaces with the atomic force microscope. Clays Clay Miner., 38(4): 337-342. He, H. et al., 2006. Changes in the morphology of organoclays with HDTMA+ surfactant loading. Appl. Clay Sci., 31(3-4): 262-271. Hewitt, D.F., 1917. The origin of bentonite. J. Wash. Acad. Sci., 7: 196-198. Holmgren, J.S., 1996. Hydrothermally stable metal oxide solid solutions as carriers for catalytic composites, U.S. (UOP Inc., USA). Us, pp. 6 pp. Cont.-in-part of U.S. 5, 393, 722. Huang, Z. et al., 1991. Selection of zeolitic acid components for the mild hydrocracking catalysts in processing poor quality VGO. Proc. Int. Conf. Pet. Refin. Petrochem. Process.: 258-65. Jiang, D. et al., 1991. Structure and catalytic behavior of saponite and its cross-linking product. Chem. Res. Chin. Univ., 7(3): 156-61. Johansson, G., 1960. On the crystal structures of some basic aluminium salts. Acta Chem. Scand., 14: 771. Johansson, G., 1962. Acta Chem. Scand., 16: 403. Johnson, I.D., Werpy, T.A. and Pinnavaia, T.J., 1988. Tubular silicate-layered silicate intercalation compounds: a new family of pillared clays. J. Am. Chem. Soc., 110(25): 8545-7. Johnston, C.T. and Stone, D.A., 1990. Influence of hydrazine on the vibrational modes of kaolinite. Clays Clay Miner., 38: 121-128. Jong, S.-J., Lin, J.-T. and Cheng, S., 1994. Synthesis of titanium pillared clay using organic medium. Stud. Surf. Sci. Catal., 83 (Zeolites and microporous crystals): 33-40.
61 Kiricsi, I., Molnar, A., Palinko, I., Fudala, A. and Nagy, J.B., 1997. Nanoscale redox catalysts: Cr- and Cr,Al-pillared layer clays: characterization and catalytic activity. Solid State Ionics, 101-103(Pt. 2): 793-797. Klimentidis, R.E. and Mackinnon, I.D.R., 1986. High-resolution imaging of ordered mixed-layer clays. Clays Clay Miner., 34: 155-164. Kloprogge, J.T., 1992. Pillared clays . Preparation and characterisation of clay minerals and Aluminium-based pillaring agents,. PhD Thesis, Universtiy of Utrecht, The Netherlands. Kloprogge, J.T., 1994. Synthesis of aluminum expanded clay a new type of form- selective sieve in heterogeneous catalyst. Tijdschr. Klei, Glas Keram., 15(9): 218- 21. Kloprogge, J.T., 1998. Synthesis of smectites and porous pillared clay catalysts: a review. J. Por. Mater., 5(1): 5-41. Kloprogge, J.T. and Frost, R.L., 1999. Infrared emission spectroscopy of Al-pillared beidellite. Appl. Clay Sci., 15: 431-445. Kloprogge, J.T., Frost, R.L. and Fry, R., 1999a. Infrared emission study of the thermal transformation mechanism of Al-13-pillared clay. Analyst, 124(3): 381-384. Kloprogge, J.T., Fry, R. and Frost, R.L., 1999b. An infrared emission spectroscopic study of the thermal transformation mechanisms in Al-13-pillared clay catalysts with and without tetrahedral substitutions. J. Catal., 184(1): 157-171. Kloprogge, T., Duong, L. and Frost, R., 2005. A review of the synthesis and characterization of pillared clays and related porous materials for cracking of vegetable oil to produce biofuels . Env. Geol., 47: 967-981. Kloprogge, T., Duong, L., Wood, B. and Frost, R., 2006. XPS study of the major minerals in bauxite: Gibsite, bayerite and (pseudo-) boehmite. J. Coll. Interf. Sci., 296: 572-576. Kohyama, N.F.K. and Fukami, A., 1982. Interlayer hydrates and complexes of clay minerals observed by electron microscopy using an environmental cell. Internat. Clay Conf.. 1981 (van Olphen,H. and Veniale F. Eds): Developments in sedimentology 35.
62 Kristof, J., Frost, R.L., Kloprogge, J.T., Gabor, M. and Toth, M., 1998. Thermoanalytical investigation on intercalated kaolinites, 7th Eur. Symp. Thermal anal. Cal., Balatonfured, Hungary, pp. 319. Krumm, S., 1999. Simulation of XRD patterns from oriented clay minerals by WinStruct. In: Butler and John (Editors), Freeware and shareware in the geosciences. Computers & Geosciences. Pergamon, New York-Oxford-Toronto, International, pp. 501-509. Kurek, P.R., 1992. Acylation of aromatic compounds by acid anhydrides using solid acid catalysts, U.S. (UOP Inc., USA). Us, pp. 6 pp. Kurian, M. and Sugunan, S., 2005. Characterisation of the acid-base properties of pillared montmorillonites. Micropor. Mesopor. Mater., 83(1-3): 25-34. Kurschner, U., Seefeld, V., Parlitz, B., Gessner, W. and Lieske, H., 1998. Catalytic activity and acidity of Al pillared clays and zeolites in different hydrocarbon reactions. React. Kinet. Catal. Lett., 65(1): 17-23. Lagaly, G., 1984. Clay organic interaction. Phil. Trans. R. Soc. Lond., A311: 315-332. Lahav, N., 1990. Preparation of stable suspensions of delaminated kaolinite by combined dimethylsulfoxide-ammonium fluoride treatment. Clays Clay Miner., 38: 219. Lahav, N., Shani, U. and Shabtai, J., 1978. Cross-linked smectites I: Preparation and properties of some hydroxy-aluminum montmorillonites. Clays Clay Miner., 26: 107-115. Lantz, M. Hug, H. J.Hoffmann,R.Schendel, P.Kappenberger, P.Martin, S. Baratoff, A.Guntherodt, H., 2001. Quantitative measurement of short range chemical bonding forces. Science, 291(5513): 2580-2583. Lapides, I., Lahav, N., K, H., Michaelian and Yariv, S., 1997. X-ray study of the thermal intercalation of alkali halides into kaolinite. J. Therm. Anal., 49: 1423-1432. Lee, J.H. and Peacor, D.R., 1986. Expansion of smectite by laurylamine hydrochloride: ambiguities in Transmission Electron Microscope observations. Clays Clay Miner., 34: 69-73. Ledoux, R.L. and White, J.L., 1966. Infrared studies of hydrogen bonding interaction between kaolinite surfaces and intercalated potassium acetate, hydrazine, formamide and urea. J. Coll. Interf. Sci., 21: 127-152.
63 Liebig, V., 1865. Chemische Briefe. Winter'sche Verlagshandlung, Leipzig, 96 pp. Lin, Y.-H., Adebajo, M.O., Frost, R.L. and Kloprogge, J.T., 2005. Thermogravimetric analysis of hydrotalcites based on the takovite formula
NixZn6xAl2(OH)16(CO3).4H2O. J. Therm. Anal. Cal., 81(1): 83-89. Lipsicas, M. Raythatha, R.H. Pinnavaia, T.J. Johnson. I.D. Giese, R.F. Costanzo, P.M. Roberts, J.L., 1984. Nature, 309: 604. Liu, S., Yang, J.-H. and Choy, J.-H., 2006. Microporous SiO2-TiO2 nanosols pillared montmorillonite for photocatalytic decomposition of methyl orange. J. Photochem. Photobiol. A: Chem., 179(1-2): 75-80. Liu, S.Y. and Bai, K., 1982. Investigation on the historical records of kaolin. Chin. Ceram: 141–170. Ma, C. and Eggleton, R.A., 1999. Surface layer types of kaolinite; a high-resolution transmission electron microscope study. Clays Clay Miner., 47(2): 181-191. Malla, P., Yamanaka, S. and Komarneni, S., 1989. Unusual water vapor adsorption behavior of montmorillonite pillared with ceramic oxides. Solid State Ionics, 32- 33(Pt. I): 354-62. Marton, L., 1968. Early history of the Electron Microscope. San Francissco Press . San Francisco, 56 pp pp. McLeod, N.A., 1997. Pillared organoclays, and method for treating soils and materials contaminated with hazardous organic compounds, Brit. UK Pat. Appl. (Envirotreat Limited, UK). Gb, pp. 44 pp. McMillan, P.F. and Hofmeister, A.M., 1988. Infrared and Raman spectroscopy, In Spectroscopic methods in mineralogy and geology. In: F.C. Hawthorne (Editor), Rev. Miner., pp. 99-159. Miehe, J., Le Dred, R., Baron, J. and Saehr, D., 1997. Improvement of the pillaring process for synthetic and natural beidellite. Chem. Ind. (Dekker), 69 (Synthesis of Porous Materials): 523-536. Montarges, E., Michot, L.J. and Ildefonse, P., 1997. Structural studies of hydroxy-Ga pillared clays by Ga K-edge X-ray absorption spectroscopy. Clays Our Future, Proc. Int. Clay Conf., 11th: 451-456.
64 Montarges, E., Michot, L.J. and Ildefonse, P., 1999. Ga13-polyethyleneoxide-modified smectites: relationship between physico-chemical properties and local pillar structure. Micropor. Mesopor. Mater., 28(1): 83-101. Moore, D.M. and Reynolds, R.C., Jr., 1997. X- Ray diffraction and Analysis of Clay Minerals. Oxford University Press, Inc, New York, 378 pp. Nabivanets, B.I.a. and Kudritskaya, L.N., 1967. Polymerization of titanium(IV) compounds in hydrochloric acid solutions. Russ. J. Inorg. Chem., 12: 616. Newman, A.C.D., 1987. Chemistry of clays and clay minerals. Longman Scientific & Technical , England, 480 pp. Occelli, M.L., 1985. Surface and catalytic properties of some pillared clays. In: H.v.O. L.G. Schulz, and FA. Mumpton, eds, (Editor), Proc. Int. Clay Conf. The Clay Mineral Society , Bloomington , Indiana., Denver, pp. 319-323 (1987). Occelli, M., Senders, J.V. and Lynch, J., 1986. TEM Analysis of pillared and delaminated hectorites catalysts. J. Catal., 107: 557-565. Occelli, M.L., 1986. New routes to the preparation of pillared montmorillonite catalysts. J. Mol. Catal., 35(3): 377-89. Occelli, M.L., 1987. Process for the catalytic cracking of nitrogen-containing feedstocks, Eur. Pat. Appl. (Union Oil Co. of California, USA). Ep, patent number 4708786, pp. 15 pp. Occelli, M.L., 1994. Hydrocracking process employing catalyst containing zeolite beta and a pillared clay, U.S. (Union Oil Co. of California, USA). Us, patent number 5614453, pp. 10 pp. Occelli, M.L., Dominguez, J.M. and Eckert, H., 1993. Vanadium effects on pillared rectorite catalysts. J. Catal., 141(2): 510-23. Occelli, M.L. and Tindwa, R.M., 1983. Physicochemical properties of montmorillonite interlayered with cationic oxyaluminum pillars. Clays Clay Miner., 31(1): 22-8. Olives, J., Amouric, M. and Perbost, R., 2000. Mixed layering of illite-smectite: Results from high-resolution transmission electron microscopy and lattice-energy calculations. Clays Clay Miner., 48(2): 282-289.
65 Ooka, C. et al., 1999. Crystallization of hydrothermally treated TiO2 pillars in pillared montmorillonite for improvement of the photocatalytic activity. J. Mater. Chem., 9(11): 2943-2952. Pinnavaia, T., 1992. Expanded clays and other Microporous solids, ed M.L.Occelli and H.E. Robson, Van Nostrand Reinhold, New York, 1992, p1 Pinnavaia, T.J., 1983. Intercalated clay catalysts. Science, 220(4595): 365-371. Pinnavaia, T.J., Tzou, M.S. and Landau, S.D., 1985. New chromia pillared clay catalysts. J. Am. Chem. Soc., 107(16): 4783-5. Pinnavaia, T.J. Morland, M.M. and Endo, T., 1983. Silica-clay complexes. U.S. Patent, patent number 4367163 Plancon, A. and Zacharie, C., 1990. An expert system for the structural characterisation of kaolinites. Clay Miner., 25: 249-260. Plancon, A. and Drits, V.A., 2000. Phase analysis of clays using an expert system and calculation programs for X-ray diffraction by two- and three-component mixed- layer minerals. Clays Clay Miner., 48(1): 57-62. Plee, D., Borg, F., Gatineau, L. and Fripiat, J.J., 1985a. High-resolution solid-state aluminum-27 and silicon-29 nuclear magnetic resonance study of pillared clays. J. Am. Chem. Soc., 107(8): 2362-2369. Plee, D., Gatineau, L. and Fripiat, J.J., 1987. Pillaring processes of smectites with and without tetrahedral substitution. Clays Clay Miner., 35(2): 81-8. Plee, D., Schutz, A., Poncelet, G. and Fripiat, J.J., 1985b. Acid properties of a bidimensional zeolite. In: B. Imelik, C. Naccache, G. Coudurier, J. Ben Taarit and J.C. Vedrine (Editors), Catalysis by Acids and Bases. Stud. Surf. Sci. Catal. Elsevier, Amsterdam, pp. 343-50. Pons, C.H., De la Calle, C. and Martin de Vidales, J.L., 1995. Quantification curves for XRD analysis of mixed-layer 14Aa/ 10Aa clay minerals. Clays and Clay Minerals, 43(2): 246-254. Raman, C.V., Krishnan K.S., 1928. A new type of radiation. Nature 121, 501 Raupach, M., Barron, P.F. and Thompson, J.G., 1987. Nuclear magnetic resonance, infrared, and X-ray diffraction study of dimethylsulfoxide and dimethylselenoxide intercalates with kaolintie. Clays Clay Miner., 35: 208-219.
66 Ross, S.D., 1972. Inorganic Infrared and Raman Spectra. McGraw-Hill Book Company, London, 140 pp. Roulia, M., 2005. Synthesis and characterization of novel chromium pillared clays. Mater. Chem. Phys., 91(2-3): 281-288. Sayed Hassan, M., Villieras, F., Gaboriaud, F. and Razafitianamaharavo, A., 2006. AFM and low-pressure argon adsorption analysis of geometrical properties of phyllosilicates. J. Col. Interf. Sci., 296(2): 614-623. Schoonheydt, R.A., Leeman, H., Scorpion, A., Lenotte, I. and Grobet, P., 1994. The Al + pillaring of clays. Part II. Pillaring with [Al13O4(OH)24(H2O)12]7 . Clays Clay Miner., 42(5): 518-25. Schoonheydt, R.A. et al., 1993. The aluminum pillaring of clays. Part I. Pillaring with dilute and concentrated aluminum solutions. Clays Clay Miner., 41(5): 598-607. Sidheswaran, P., Ganguli, P. and Bhat, A.N., 1987. Thermal behavior of intercalated kaolinite. Thermochim. Acta, 118: 295-303. Siegbahn, K. et al., 1967. Electron Spectroscopy for Chemical Analysis. Atomic, Molecular and Solid State Structure Studies by Means of Electron Spectroscopy. Almquist and Wiksells, Uppsala. Sivakumar, S., Ghosh, S.K., Damodaran, A.D. and Warrier, K.G.K., 1997a. Dependence of concentration of montmorillonite and boehmite sol on the properties of al-pilc. Polyhedron, 16(16): 2785-2788. Sivakumar, S., Ghosh, S.K., Damodaran, A.D. and Warrier, K.G.K., 1997b. Influence of medium on the properties of boehmite intercalated montmorillonite. Mater. Lett., 31(1,2): 113-118. Sivakumar, S., Mukundan, P., Pillai, P.K., Damodaran, A.D. and Warrier, K.G.K., 1994. Thermal stability of montmorillonite intercalated and alumina pillared by boehmite sol. Ceram. Int., 20(5): 349-352. Skoularikis, N.D., Coughlin, R.W., Kostapapas, A., Carrado, K. and Suib, S.L., 1988. Catalytic performance of iron(III) and chromium(III) exchanged pillared clays. Appl. Catal., 39(1-2): 61-76. Spurr, A.R., 1969. A low -Viscocity Epoxy Resin Embedding Medium for Electron Microscopy. Ultrastruct. Res., 26: 31-43.
67 Sterte, J., 1989. Hydrothermal stability and catalytic cracking performance of some pillared clays. Prepr. - Am. Chem. Soc., Div. Pet. Chem., 34(3): 489-96. Sterte, J. and Shabtai, J., 1987a. Cross-linked smectites. V. Synthesis and properties of. Clays Clay Miner., 35: 429-439. Sterte, J. and Shabtai, J., 1987b. Cross-linked smectites. V. Synthesis and properties of hydroxy-siticoaluminum montmorillonites and fluorohectorites. Clays Clay Miner., 35: 429-439. Sychev, M., Kostoglod, N., Kosorukov, A., Goncharuk, V. and Kashkovski, V., 1992. Pillared clays: preparation and investigation of physicochemical and catalytic properties. Proc. - Pol.-Ger. Zeolite Colloq.: 63-78. Tichit, D. et al., 1988. Sintering of montmorillonites pillared by hydroxy-aluminum species. Clays Clay Miner., 36(4): 369-75. Torii, K., Iwasaki, T., Onodera, Y. and Nishiyama, Y., 1992. Mesoporous materials produced from novel trioctahedral nickel-rich smectite. Nippon Seramikkusu Kyokai Gakujutsu Ronbunshi, 100(April): 368-72. Triolo, I., Mottana, A. and Giampaolo, C., 1988. Synthesis and thermal behaviour of syngenite. Terra Cognita, 8(1): 79. Vaccari, A., 1998. Preparation and catalytic properties of cationic and anionic clays. Catal. Today, 41(1-3): 53-71. Vaughan, D.E.W. and Lussier, R.J., 1980. Preparation of molecular sieves based on pillared interlayered clays (PILC). Proc. Int. Conf. Zeolites, 5th: 94-101. Vaughan, D.E.W., Lussier, R.J. and Magee, J.S., Jr., 1981. Pillared interlayered clay products, U.S. (Grace, W. R., and Co., USA). Us, patent number 4271043pp. 5 pp. Cont.-in-part of U.S. Ser. No. 71,957. Veblen, D.R., Guthrie, G.D., Jr., Livi, K.J.T. and Reynolds, R.C., Jr., 1990. High- resolution transmission electron microscopy and electron diffraction of mixed- layer illite/ smectite; experimental results. Clays Clay Miner., 38(1): 1-13. Velde, B., 1985. Clay minerals. A physical chemical Explanation of their occurrence. Elsevier, Amsterdam, 427 pp. Velde, B., 1992. Introduction to clay minerals. Chapman & Hall, London, 197 pp.
68 Vieira Coelho, A. and Poncelet, G., 1990. Preparation, characterization and some catalytic properties of different smectites pillared with hydroxyaluminum and hydroxy-gallium solutions. Preliminary study. Pillared Layered Struct.: Curr. Trends Appl., [Proc. Workshop]: 185-93. Vogels, R.J.M.J., Kloprogge, J.T. and Geus, J.W., 2004. Brønsted and Lewis acidity of synthetic saponites and stevensites: effects of substitution, cation exchange and hydration state. Appl. Catal. A Wada, K., 1961. Lattice expansion of the kaolin minerals. Amer. Miner., 46: 79-91. Wagner, C.D., Rigg, W.M.,Davis, L.E.,Moulder, J.F. and Muilenberg, G.F. 1979a. Hand book of X-ray photoelectron spectroscopy. Eden Prairie, Minnesota, Physical Electronics Division Weaver, C.F. and Pollard, L.D., 1973. The chemistry of clay minerals. Elsevier, Amsterdam, 213 pp. Weiss, A., Thielepape, W. and Orth, H., 1966. Intercalation into kaolinite minerals. Proc. Int. Clay. Conf. Jerusalem, 1: 277-293. Wenyang, X., Yizhao, Y., Xianmei, X., Shizheng, L. and Jinping, L., 1991a. Preparation and properties of Al-B-Si composite pillared bentonite. React. Kinet. Catal. Lett, 44: p 115-120. Wenyang, X., Yizhao, Y., Xianmei, X., Shizheng, L. and Taoying, Z., 1991b. Catalytic cracking properties of Al-Zr-B composite pillared clays. Appl. Catal., 75: p 33-40. Wu, P. X., Z. W. Liao, Zhang, H.F. , Guo, J.G. (2001). Adsorption of phenol on inorganic- organic pillared montmorillonite in polluted water. Env. In., 26: 401- 407. Xie, X., Yao, Y., Cheng, G. and Cao, J., 1994. Preparation and characterization of Al- Mg-PILC. Fenzi Cuihua, 8(1): 63-9. Yoneyama, H., Haga, S. and Yamanaka, S., 1989. Photocatalytic activities of microcrystalline titania incorporated in sheet silicates of clay. J. Phys. Chem., 93(12): 4833-7. Zbik, M. and Smart, R.S.C., 1998. Nanomorphology of kaolinites: Comparative SEM and AFM studies. Clays Clay Miner., 46: 153-160.
69 Zhao, D., Wang, G., Yang, Y. and Guo, X., 1992. Synthesis and characterization of hydroxyaluminosilicate crosslinked montmorillonites with high basal spacings. Gaodeng Xuexiao Huaxue Xuebao, 13(3): 295-9. Zhao, D. et al., 1993. Preparation and characterization of hydroxy-iron-aluminum pillared clays. Clays Clay Miner., 41(3): 317-27. Zhao, D., Yang, Y., Xin, Q., Zhang, H. and Guo, X., 1994. Studies on the IR spectra of pillared clay molecular sieve. Guangpuxue Yu Guangpu Fenxi, 14(4): 27-30, 18.
70
Statement of Contribution
The authors listed below have certified* that:
1. they meet the criteria for authorship in that they have participated in the conception, execution, or interpretation, of at least that part of the publication in their field of expertise; 2. they take public responsibility for their part of the publication, except for the responsible author who accepts overall responsibility for the publication; 3. there are no other authors of the publication according to these criteria; 4. potential conflicts of interest have been disclosed to (a) granting bodies, (b) the editor or publisher of journals or other publications, and (c) the head of the responsible academic unit, and
5. they agree to the use of the publication in the student’s thesis and its publication on the Australasian Digital Thesis database consistent with any limitations set by publisher requirements. In the case of this chapter:
THE DISTRIBUTION OF Ga IN Ga-PILLARED MONTMORILLONITES: A TRANSMISSION ELECTRON MICROSCOPY AND MICROANALYSIS STUDY
Loc Duong, Thor Bostrom, Theo Kloprogge, and Ray Frost Published in Microporous and Mesoporous Materials 2005, 82, 165-172
Contributor Statement of contribution*
Loc Duong wrote the manuscript, designed experiments, conducted experiments, data analysis, data interpretation
Date
Thor Bostrom aided with data analysis and interpretation
Theo Kloprogge aided data analysis, editing
Ray Frost Aid editing
Principal Supervisor Confirmation
I have sighted email or other correspondence from all Co-authors confirming their certifying authorship and. ______Name Signature Date CHAPTER 2
2.0 THE DISTRIBUTION OF Ga IN Ga-PILLARED MONTMORILLONITES: A TRANSMISSION ELECTRON MICROSCOPY AND MICROANALYSIS STUDY
Loc Duong1,2*, Thor Bostrom1,2, Theo Kloprogge1, and Ray Frost1
Published in. Microporous and Mesoporous Materials 2005, 82, 165-172
Loc Duong, Inorganic Materials Research Program, School of Physical and Chemical
Sciences, Queensland University of Technology, GPO Box 2434, Brisbane, Qld 4001,
Australia
Thor Bostrom, Analytical Electron Microscopy Facility, Queensland University of
Technology, GPO Box 2434, Brisbane, Qld 4001, Australia
Theo Kloprogge, Inorganic Materials Research Program, School of Physical and
Chemical Sciences, Queensland University of Technology, GPO Box 2434, Brisbane,
Qld 4001, Australia
Ray Frost, Inorganic Materials Research Program, School of Physical and Chemical
Sciences, Queensland University of Technology, GPO Box 2434, Brisbane, Qld 4001,
Australia
72 2.1 Abstract
The distribution of Ga in the interlayer of montmorillonite pillared with a Ga13 polyoxocation complex has been studied by transmission electron microscopy, energy- dispersive X-ray microanalysis (EDX), X-ray mapping and powder X-ray diffraction in combination with N2 adsorption-desorption. To view the clay layers by TEM, the pillared clay was embedded in Spurrs resin in a preferred orientation, and sectioned with an ultramicrotome perpendicular to the layers. Montmorillonites pillared with Al13 and
Al12Ga complexes were also prepared for microanalysis in the TEM. The Ga X-ray peaks could be easily distinguished in the EDX spectra, allowing concentrations relative to other elements to be determined. Elemental X-ray maps for Ga, Si and Al in the Ga13 pillared clay cross-sections demonstrated that the Ga was homogeneously distributed throughout the crystal thickness. Comparison of the analytical data with that from the
Al13 and Al12Ga pillared clays and the starting material suggested that an approximately constant amount of the intercalated species per amount of Si in the clay became incorporated into the structure in each case. Calculation of the formula for the Ga- pillared montmorillonite showed that 0.89 Ga is present per formula unit containing 8
(Si+Al), which is equivalent to 20 silicate rings, each consisting of 6 tetrahedra, for every
Ga13 pillar. The actual dimension of the pillar, based results from the elemental analyses and XRD is 8.7Å and the mean distance between the pillars is 44.3Å, which is in good agreement with the average pore size of 39Å obtained by N2 adsorption-desorption measurements. This study shows a new approach for obtaining more detailed information on the pillars in pillared clay by combining analytical data from X-ray microanalysis with measurements by XRD and N2 adsorption-desorption.
73
Keywords: Ga13, Montmorillonite, Pillared clay, Textural properties, Transmission Electron Microscopy
2.2 Introduction
Pillared montmorillonites are microporous materials, obtained by ion exchange of montmorillonites with highly charged metallic species followed by a calcination process, that in the last 25 years have been developed as a new class of catalysts [1-5]. Producing suitable pillared montmorillonites for catalytic applications requires detailed understanding of the structure of the starting clay, the pillaring agent, and the size, shape and location of the oxidic pillars in the final products. Although extensive research has been undertaken in the field of pillared clay structures [1, 6-8], the complete structure of pillared clays especially the location, structure and bonding to the clay layers of the oxide pillars is still unknown.
Montmorillonites pillared with different polyoxocations have been prepared and studied for many years [1]. Al13 montmorillonites are the most common pillared clays [9], mainly because the Al polyoxocation gives a large basal d-spacing and, together with the very similar Al12Ga pillared montmorillonites, are quite stable at high temperatures, up to
o about 750 C. The structures of Al13 and Ga13 are very similar with both having a Keggin
IV IV VI structure in which a central Al O4 or Ga O4 is surrounded by twelve Al octahedra with water and hydroxyl groups [10-15] resulting in similar basal d-spacings in the final pillared clays.
74 Transmission electron microscopy (TEM) has provided the necessary spatial resolution for studying the microstructure of clays down to the lattice level, but has required specific sample preparation procedures to achieve this. A preparation method for studying smectite layers by TEM has been discussed by Kim et al. [16]. TEM together with energy-dispersive X-ray spectrometry (EDX) has allowed chemical analysis of small areas down to about 20 nm and X-ray mapping of regions of a few hundred nanometers.
This combination of structural and chemical analysis in the TEM is a powerful tool for studying complex, fine-grained clay minerals, and in particular pillared clays. However the achievable analytical spatial resolution may be limited by the electron beam sensitivity of some specimens. Ma et al. [17] used EDX in a TEM to study the loss of certain elements from a clay structure. The diffusion of alkali elements and higher atomic number elements (Fe, Ti) from clay minerals was found to be associated with crystal habit and instrumentation conditions. More specimen damage due to the electron beam was seen when the crystals were viewed in the plane perpendicular to the (001) direction than in the plane parallel to (001). The loss of elements may be reduced by using a lower beam current and a larger analysis area [17]. Crozier et al. [18] used analytical electron microscopy to confirm the location of Zr pillars in Zr-pillared montmorillonite. The Zr pillars were found to have an irregular shape and distribution in the silicate layers [18].
Montmorillonite pillared with Ga13 has some advantages over that pillared with Al13 for analytical TEM studies. As Ga has a higher atomic number (31) than Al, the intercalated Ga may potentially improve contrast in TEM images. Further, the element can be clearly detected by X-ray microanalysis without confusion with the Al in the clay structure. For comparative studies Al13 and Al12Ga pillared clays were also prepared. The
75 pillared montmorillonites were examined and analysed in the TEM mainly in cross- section, in other words with the c-axis oriented perpendicular to the electron beam. In this paper the results from transmission electron microscopy, EDX microanalysis, powder X- ray diffraction and N2 adsorption-desorption measurements have been used to deduce the distribution of Ga in the structure of Ga13 pillared montmorillonite. It is shown that an approximately constant proportion of the intercalated species is incorporated into the clay structure in pillaring, and it is possible to calculate an average size and distribution of the pillars.
2.3 Materials and methods
2.3.1 Starting materials
The starting materials used for this study were ≤ 2 μm fractions of Wyoming montmorillonite SWy-2, and Miles montmorillonite from Queensland, Australia. The
Miles montmorillonite has a significantly higher CEC than Swy-2. All samples were saturated with sodium through exchange with 1 M NaCl for 8 hours. The clays were washed five times with deionised water in order to remove residues of NaCl. A detailed description of the Miles material and the Al13-pillaring procedure have been provided by
Kloprogge et al. [19]. Miles montmorillonite was used to prepare the Al13 and Al12Ga pillared clays and SWy-2 montmorillonite was used to prepare the Ga13 pillared clay. The preparation of the Al12Ga and Ga13-pillared clays was analogues to that of the Al13- pillared clay. A 0.1M solution of NaOH was added to Ga(NO3)3 at a rate of 0.01 ml/min using a peristaltic pump under vigorous stirring at room temperature. The OH/Ga ratio was 2:1. The Al13, Al12Ga and Ga13 solutions were added to the aqueous clay suspensions
76 under continuous stirring for a period of four hours. The suspensions were then allowed to stand for several days. The pillared clays were washed 5 times with deionised water using a centrifuge. The samples were allowed to dry in air at ambient temperature.
Finally, the samples were heated at 2°C/min and calcined at 450°C for 8 hours.
2.3.2 X-ray diffraction
X-ray diffraction (XRD) was used to check for the intercalation process of Ga into the clay structure before and after calcination by observing the changes in the (001) spacings.
For powder XRD, the sample was ground and mixed with ethanol, deposited on a low background plate and dried at room temperature. Preferential orientation of the clay platelets is very common under these conditions, thus enhancing the (00l) reflections relative to other reflections. XRD patterns were collected using a PANalytical X’Pert Pro diffractometer with a rotating anode source and a diffracted beam curved graphite monochromator and CuKα radiation. Scans were made using a 0.05o step size at
0.5sec/step.
2.3.3 Sample preparation for Transmission Electron Microscopy
Smectites, due to their ability to absorb and desorb water and exchange cations, are very sensitive to the method used to prepare them for examination in the TEM and therefore care must be taken not to introduce preparation artefacts. For viewing the clay particles mainly parallel to the c-axis, a dilute suspension of montmorillonite in 70% alcohol was briefly ultrasonicated and a small drop of the suspension was placed on a thin carbon film on a TEM copper grid, allowed to dry and coated with a thin carbon layer to improve stability under the beam.
77 In order to observe the clay in a direction perpendicular to the c-axis so that the layers could be viewed in cross-section, the clay was embedded in Spurrs resin [20] for ultramicrotoming. The morphology, swelling properties, water absorption, and reaction with the resin have to be taken into account when preparing smectite samples for TEM.
Because of its low viscosity, Spurrs has long been used as an embedding agent for biological and material samples for electron microscopy, and was used here to obtain good penetration into the clay material. However the resin is intolerant of moisture in the sample, and therefore the samples had to be adequately dried before embedding. Figure
2.1 shows the steps used in the preparation of cross-sections of the pillared clays. Pillared montmorillonite was diluted with water and allowed to settle for about one week. The water was removed leaving the clay particles preferentially aligned with the bottom of the container. The sample was dried at 60-100oC for two weeks before addition of the resin.
After polymerisation overnight at 60oC, the small resin plug was removed from the base of the container, rotated 90° and sectioned with a diamond knife using a Reichert ultramicrotome to produce cross-sections of the embedded particles. The sections were
60-80 nm in thickness. The starting montmorillonites were also prepared for microanalysis, but as these materials had not been calcined they was first dehydrated with alcohol and acetone to remove any residual water before impregnation with Spurrs resin.
78
Clay in suspension, clay particles deposit slowly at the bottom of the container with flat grains mainly parallel to the STAGE 1 bottom
STAGE 2 Oriented clay embedded in Spurrs resin
Cross sections cut using an STAGE ultramicrotome 3
Fig. 2.1 Preparation of cross sections of a clay sample
for TEM
79 2.3.4 Transmission Electron Microscopy
Specimens were examined in a Philips CM200 transmission electron microscope fitted with a LaB6 cathode and operated at 200kV. For measurements of lattice spacings and electron diffraction patterns, TEM negatives were scanned at 600 or 1200dpi and measurements carried out on the digital images using image analysis software. Energy- dispersive X-ray microanalysis and X-ray mapping was carried out mainly in scanning transmission (STEM) mode using a Link thin-window X-ray detector and Link ISIS 300 microanalysis system (Oxford Instruments, UK). Quantitative calculations of element concentrations and atomic ratios were carried out using a thin-film matrix correction procedure, in which the total concentrations are normalised to 100%. For these calculations, the density of the material was taken as 3.0 g.cm-3 and the specimen thickness for each analysis was estimated from STEM images and the Si X-ray intensity.
2.3.5 Nitrogen adsorption-desorption
The surface areas of the starting montmorillonite and Ga13 pillared montmorillonite were calculated by the BET method using N2 adsorption-desorption on a Micromeritics
ASAP 2010 at a partial pressure range of 0.06 to 0.30. The pore size distribution and pore volume were calculated using the Tristar software.
80 (c) 17.9 Å
Ga13 calcined
19.9 Å
Ga13 exchanged Counts (b)
14 Å
Starting montmorillonite (a)
3 8 13 18 23 degrees 2θ Cukα
Fig. 2.2 XRD patterns of:
(a) starting Wyoming Swy-2 montmorillonite
(b) montmorillonite exchanged with Ga13 showing an interlayer spacing of 19.9 Å
(c) Ga13 pillared montmorillonite with a spacing oft 17.9 Å
81 2.4 Results and discussion
X-ray powder diffraction (XRD) patterns of the starting montmorillonite and Ga13 pillared montmorillonite demonstrated that the Ga had been pillared successfully. As the
Ga is intercalated into the structure the layers of the montmorillonite are propped apart, and the basal spacing (the d-spacing along the c-axis, or 001 reflection) increases from
o about 14Å to 19.9Å (Figure 2.2). After calcination at 450 C the d001 spacing reduced to
17.9Å. The clean pattern of the intercalated clay indicated that the intercalation process had completed and that no non-pillared montmorillonite remained in the sample.
Figure 2.3a shows the laminar structure of a single grain of Al12Ga pillared montmorillonite, prepared by simple deposition from suspension onto a TEM grid.
Figures 2.3b-d are high magnification TEM images of resin embedded sections of the
Ga13, Al12Ga and Al13 pillared montmorillonites, together with electron diffraction patterns showing the (001) diffraction spots. The images also show detailed views of lattice fringes and corresponding spacings in specific areas. The observed lattice fringe spacings from TEM of the pillared clays ranged from 12.9 to 18.2 Å, with a mean value of about 15.2 Å. However, it is clear from the micrographs that the fringe spacings can vary quite considerably, particularly where there is pronounced curvature of the crystals
(Figure 2.3c, arrow). Two different fringe spacings can be observed in the detailed view in Figure 3d. This variability should be reflected in some broadening of the (001) peak in the XRD pattern and this is what is observed (Figure 2.1). There is some suggestion that the layers are more clearly defined in the Al12Ga pillared montmorillonite, and this may be due to the fact that the central GaIV atom fits much better in the Keggin-type complex, thereby not only increasing the thermal stability [9], but also the stability of the pillared
82 clay under the electron beam in the TEM. We did observe a noticeable loss of structure in these materials due to beam damage after a short period of TEM viewing, especially at high magnifications.
Measurements of the (001) spacing from electron diffraction patterns of cross-sections of the pillared clays gave a mean value of 15.6 Å (range 14.4 – 17.4 Å), which is consistent with the spacings observed in the micrographs. The patterns also showed fine arcs at 4.45 Å, corresponding to (100) reflections, and in one case spots at 2.29 Å, probably (113) reflections from an adjacent crystal. Overall, the electron diffraction patterns were most consistent with the Montmorillonite-15A structure for a Wyoming montmorillonite (ICDD powder diffraction database #29-1498), for which d001, d100 and d113 are 15.542, 4.473 and 2.311 Å respectively. However both the observed lattice spacings in the micrographs and the measured (001) spacings from the electron diffraction patterns are on average lower than the basal spacing of 17.9 Å measured by
XRD for the Ga13 pillared clay. This difference may arise from the different preparation methods used. The XRD measurements were made on powdered material in ambient air, while the TEM measurements were made from thin sections of resin-embedded clay in a high vacuum. The TEM preparation required extensive drying of the clay samples prior to resin embedding and polymerisation and this may explain the differences observed.
83
Fig. 2.3 TEM image of:
(a) a grain of Al12Ga pillared montmorillonite; and images and electron diffraction patterns from sectioned material
(b) Ga13 pillared montmorillonite
(c) Al12Ga pillared montmorillonite and
(d) Al13 pillared montmorillonite
84 Figure 2.4 shows an X-ray intensity map for Al, Si and Ga in a thin cross-section of a single grain of the Ga13 pillared clay. The map demonstrates that the distribution of Ga is closely related to that of Al and Si, and that Ga is present throughout the grain thickness.
The estimated electron probe diameter in STEM mode under the conditions used was about 4 nm, consequently there is insufficient spatial resolution to distinguish the layers directly. To measure the amount of Ga in the Ga13 pillared clay, eight EDX analyses were taken from individual small grains or very small clusters of grains. Ga was present in all these spectra in roughly equivalent amounts. The other pillared clays were analysed in a similar manner. EDX spectra from the three pillared clays, compared in each case with a spectrum from the starting montmorillonite, are shown in Figure 2.5. The Ga K and L X- ray lines are clearly discernible even in the Al12Ga pillared material, and excess Al is evident in the clays pillared with Al13 and Al12Ga. Both atomic ratios to silicon and concentrations of the elements were calculated from the spectra. The ratio to Si was used since Si is a relatively constant element within the structure and is not affected by the pillaring process. The mean concentrations from EDX analyses of the three pillared clays as well as from the starting materials are listed in Table 2.1, and do show some variation between the different pillared and starting clays analysed.
From the spectra, the mean atomic ratio of Ga to Si was 0.024 ± 0.001 (SD, n = 10) for the Al12Ga pillared clay, and 0.235 ± 0.015 (SD, n = 8) for the Ga13 pillared clay. The ratio of total Al/Si was 0.646 in the Al13 pillared material, as compared to a mean of
0.373 in the Miles starting clay, therefore the overall Al13/Si ratio was 0.273. For the
Al12Ga pillared material, the total (Al+Ga)/Si ratio was 0.608, which gives an overall
85 Al12Ga/Si ratio of 0.235 after allowing for the Al in the starting material. Thus the ratios of pillared element to Si were 0.273, 0.235 and 0.235 respectively in the three pillared clays, suggesting that a roughly constant proportion of the pillared element is incorporated into the structure.
A more reliable correction for the Al content in the tetrahedral and octahedral layers is obtained by calculating the stoichiometric formula of the clay. This has been done in
Table 2.1 for the analysed clays using the mean element concentration data from the X- ray microanalyses. The analyses were calculated into a chemical formula based on a net negative charge of 44 (20 oxygen atoms + 4 hydroxyls). The calculated formula reflects the chemical analysis of SWy-2 montmorillonite well and is very close to the composition described for this CMS source clay [21]. For the Ga13 pillared clay, the formula indicates an average of 0.89 Ga atoms per structural unit.
86
Fig. 2.4 Elemental X-ray maps for Ga, Si and Al from a cross section of a single grain of Ga13 pillared montmorillonite
87 Fig. 2.5 EDX spectra from analyses in the TEM of small grains of:
(a) Ga13 pillared
(b) Al12Ga pillared
(c) Al13 pillared montmorillonites The spectra are shown overlaid with a spectrum from the starting material. The C and Cu peaks derive from the resin or thin carbon coating and the TEM grid material respectively 88
Element Starting Starting Ga13 Pilc Al13-Pilc Al12Ga- Pilc Mont. Miles Mont. Swy-2 Miles Miles (Swy-2) Na 1.23 2.15 0 0 0.20 Mg 1.49 1.71 0.3 1.21 1.07 Al 10.68 9.86 10.00 13.87 13.82 Si 27.12 26.66 25.15 25.85 29.41 K 0.06 0.22 0.02 0 0.03 Ca 0.09 0.64 0.15 0 0.09 Ti 0.10 0.05 0 0.25 0.18 Fe 2.78 2.39 2.55 2.26 2.88 Ga 0 0 2.85 0 0.72
Calculated formula based on 22 oxygens
Starting (Na K Ca )(Mg Fe3+ Ti Al )(Si Al )O (OH) .nH O montmorillonite 0.14 0.08 0.04 0.49 0.40 0.04 3.09 7.87 0.13 20 4 2 Swy-2 Ga pillared 3+ 13 Ga (Mg Fe Ti Al )(Si Al )O (OH) .nH O montmorillonite 0.89 0.60 0.21 0.02 3.09 7.87 0.13 20 4 2 Starting Miles (Na K Ca )(Mg Fe3+ Ti Al )(Si Al )O (OH) .nH O montmorillonite 0.95 0.02 0.084 0.60 0.12 0.005 3.11 7.95 0.05 20 4 2 Al13 pillared Al (Mg Fe3+ Ti Al )(Si Al )O (OH) .nH O montmorillonite 1.46 0.65 0.18 0.010 3.11 7.95 0.05 20 4 2 Al12Ga pillared Al Ga (Mg Fe3+ Ti Al )(Si Al )O (OH) .nH O montmorillonite 2.28 0.19 0.40 0.78 0.017 0.91 7.95 0.05 20 4 2
Table 2. 1: Characteristic concentrations (single analysis) in weight % from X-ray
microanalyses in the TEM of the starting and the Ga13-, Al12Ga- and
Al13-pillared montmorillonites (top), and the formula calculation based on 22 oxygens (bottom)
89 From the information above obtained by EDX analysis and the structure of the montmorillonite, it is possible to calculate the size of the Ga pillars in the Ga13 pillared clay. From the literature the thickness of a single clay layer (consisting of one octahedral sheet sandwiched between two tetrahedral sheets) is about 9.8 Å. A basal spacing for the
Ga pillared clay of 18.5 Å then results in a height of the Ga pillars of around 8.7 Å, which is close to the value of 9.8 Å for the hydrated complex in solution taking into account the decrease in size during the calcination in which the complex looses all its water and hydroxyl groups [22, 23]. The formula calculated from the EDX analyses shows 0.89 Ga per unit formula, which contains about 8 (Si+Al). There are 6 tetrahedra needed to form a silicate ring, so a total of 19.45 rings are needed to accommodate one Ga13 pillar. If we consider a total number of 20 silicate rings per pillar this will result in 10 rings of the top layer and another 10 rings of the bottom layer. From the literature the dimension of the silicate ring is about 5.3 Å [24] so 10 rings correspond to 53 Å. The Ga13 shape being a
Keggin structure can be considered to be equi dimensional so it is possible to use the height of Ga13 as indicated from XRD measurements of the basal spacing for the size of the pillar. Then the actual dimension of the pillars is 8.7 Å and the distance between the pillars is 53Å – 8.7Å = 44.3Å. The N2 adsorption-desorption measurements show a quite homogeneous pore size distribution with an average pore size of 39 Å (Figure 6), and the value of 44.3 Å calculated above is in very good agreement with this mean pore size
90
0.009
/g) 0.008 3 0.007 0.006 0.005 0.004 0.003 0.002 0.001 Pore volume (cm volume Pore 0 20 40 60 80 100 Pore diameter (Å)
Fig. 2.6 Pore size distribution of Ga13 pillared montmorillonite
91 This calculation is based on average dimensions and it is clear from the TEM images and electron diffraction measurements that the basal spacing does vary within the structure, so it would be expected that the actual distribution of the pillars within the interlayer space may not be homogeneous, but may be analogous to that found for Zr pillars by Crozier et al. [18].
2.5 Conclusions
Wyoming SWy-2 montmorillonite has been used to produce Ga13 pillared clay with a mean basal spacing of about 17.9Å, as determined by XRD. This material was prepared for analysis by TEM by orienting the clay particles, embedding them in Spurrs resin, and then ultramicrotoming to produce cross-sections of the clay grains. For comparison Al13 and Al12Ga pillared montmorillonites were prepared in a similar manner. Ga13 pillared montmorillonite appears to have a similar structure to the Al13 and Al12Ga pillared materials, but has the advantage that the Ga incorporated in the clay can be easily analysed by energy-dispersive X-ray microanalysis without confusion with the Al in the structure. Detailed X-ray maps of Ga, Si and Al in cross-sections of the clay grains showed a good correlation between the three elements and also indicated that the Ga was present throughout the structure. Direct measurements of basal layer spacings from TEM micrographs, as well as determinations of spacings from electron diffraction patterns, gave layer spacings that were about 10% lower than the values expected from XRD measurements. However this difference may have been mainly due to the drying, resin embedding and polymerisation used for processing the specimens for TEM.
92 By allowing for the average amount of Al in the structure of the starting clay material, it is estimated from the EDX microanalyses of the three clays that an approximately constant amount of the intercalated elements is incorporated into the montmorillonite structure. The estimated atomic fractions of the total intercalated species to silicon were
0.273, 0.235 and 0.235 for the Al13, Al12Ga and Ga13 pillared clays respectively. By detailed calculation of the clay stoichiometries from the EDX data, it was shown that 0.89
Ga atoms are present per formula unit, which indicates that there are 20 silicate rings consisting of 6 tetrahedral each per Ga13 pillar. Thus the average distance between the pillars has been calculated to be 44 Å. This value is close to the average pore size of 39 Å that was determined from N2 adsorption-desorption measurements of the pillared clay. As the ratio of Al12Ga to Si was similar to that for Ga13/Si, we expect that the distribution of
Al12Ga pillars is similar to that of Ga13. The ratio Al13/Si was somewhat higher than for
Al12Ga and Ga13, hence the average distance between the Al13 pillars should be somewhat less than the 44 Å estimated for the Ga13.
The information from microanalyses in the TEM, in combination with data from XRD and N2 adsorption-desorption measurements, has allowed us to determine some fundamental information about the distribution of the pillars in the pillared clay. The finding that nearly one pillared atom is incorporated per structural unit in the clay suggests that a specific bonding site may be involved in the pillaring process within the interlayer. Further details of the pillaring mechanism are currently being investigated using other techniques.
93 2.6 Acknowledgements
We wish to thank Mr Tony Raftery for his expert assistance with the XRD measurements, and members of the clay group in the Inorganic Materials Research
Program for helpful advice and discussion. We acknowledge financial support from the
Inorganic Materials Research Program, Faculty of Science, Queensland University of
Technology.
94 2.7 References
1. J. T. Kloprogge, J. Porous Mater. 5 (1998) 5. 2. S. M. Bradley, R. A. Kydd and K. K. Brandt, Stud. Surf. Sci. Catal. 73 (1992) 287. 3. S. M. Bradley and R. A. Kydd, J. Catal. 142 (1993) 448. 4. R. T. Yang, J. P. Chen, E. S. Kikkinides, L. S. Cheng and J. E. Cichanowicz, Ind. Eng. Chem. Res. 31 (1992) 1440. 5. A. Vaccari, Catal. Today 41 (1998) 53. 6. T. J. Pinnavaia, NATO ASI Ser., Ser. C 165 (1986) 151. 7. F. Figueras, Catal. Rev. - Sci. Eng. 30 (1988) 457. 8. W. Jones, Catal. Today 2 (1988) 357. 9. S. M. Bradley, R. A. Kydd, R. Yamdagni and C. A. Fyfe, Synth. Microporous Mater. (1992) 13. 10. S. M. Bradley, R. A. Kydd and R. Yamdagni, Magn. Reson. Chem. 28 (1990) 746. 11. S. M. Bradley, R. A. Kydd and R. Yamdagni, J. Chem. Soc., Dalton Trans. (1990) 2653. 12. A. Bellaloui, D. Plee and P. Meriaudeau, Appl. Catal. 63 (1990) L7. 13. A. V. Coelho and P. G., Appl. Catalysis 77 (1991) 303. 14. F. Gonzalez, P. C., B. I. and M. S., J. Chem. Soc. (1991) Chem. Commun. 15. E. Montarges, L. J. Michot and P. Ildefonse, Microporous and Mesoporous Mater. 28 (1999) 83. 16. J.-W. Kim, D. R. Peacor, D. Tessier and F. Elsass, Clays Clay Miner. 43 (1995) 51. 17. C. Ma, D. Fitzgerald, R. A. Eggleton and D. J. Llewellyn, Clays Clay Miner. 46 (1998) 301. 18. P. A. Crozier, M. Pan, C. Bateman, J. J. Alcaraz and J. S. Holmgren, Clays Clay Miner. 47 (1999) 683. 19. J. T. Kloprogge, R. Evans, L. Hickey and R. L. Frost, Appl. Clay Sci. 20 (2002) 157.
95 20. A. R. Spurr, Journal of Ultrastructural Research 26 (1969) 31–43. 21. A. R. Mermut and A. F. Cano, Clays Clay Miner. 49 (2001) 381. 22. M. L. Occelli, A. Auroux and G. J. Ray, Microporous Mesoporous Mater. 39 (2000) 43. 23. M. L. Occelli, J. A. Bertrand, S. A. C. Gould and J. M. Dominguez, Microporous Mesoporous Mater. 34 (2000) 195. 24. W. A. Deer, R. A. Howie and J. Zussman, An introduction to the Rock-Forming Minerals, 2nd ed., Addison Wesley Longman Ltd., Harlow, 1996.
96
Statement of Contribution
The authors listed below have certified* that:
1. they meet the criteria for authorship in that they have participated in the conception, execution, or interpretation, of at least that part of the publication in their field of expertise; 2. they take public responsibility for their part of the publication, except for the responsible author who accepts overall responsibility for the publication; 3. there are no other authors of the publication according to these criteria; 4. potential conflicts of interest have been disclosed to (a) granting bodies, (b) the editor or publisher of journals or other publications, and (c) the head of the responsible academic unit, and
5. they agree to the use of the publication in the student’s thesis and its publication on the Australasian Digital Thesis database consistent with any limitations set by publisher requirements. In the case of this chapter:
AN IMPROVED ROUTE FOR THE SYNTHESIS OF AL13PILLARED MONTMORILLONITE CATALYSTS
Loc V. Duong, Jacob T. Kloprogge, Ray L. Frost, and Job A. Veen Published in Journal of microporous and mesoporous materials 2007, 14, 71-79
Contributor Statement of contribution*
Loc Duong wrote the manuscript, designed experiments, conducted experiments, data analysis, data interpretation
Date
aided with data analysis and interpretation Jacob Kloprogge
Ray Frost editing
Job Veen Aid with catalytic testing
Principal Supervisor Confirmation
I have sighted email or other correspondence from all Co-authors confirming their certifying authorship and. ______Name Signature Date CHAPTER 3
3.0 AN IMPROVED ROUTE FOR THE
SYNTHESIS OF AL13PILLARED
MONTMORILLONITE CATALYSTS
Loc V. Duong*1, Jacob T. Kloprogge1, Ray L. Frost1, and Job A. Veen2
Published in Journal of microporous and mesoporous materials 2007, 14, 71-79
Loc Duong, Inorganic Materials Research Program, School of Physical and Chemical Sciences, Queensland University of Technology, GPO Box 2434, Brisbane, Qld 4001, Australia
Theo Kloprogge, Inorganic Materials Research Program, School of Physical and Chemical Sciences, Queensland University of Technology, GPO Box 2434, Brisbane, Qld 4001, Australia
Ray Frost, Inorganic Materials Research Program, School of Physical and Chemical Sciences, Queensland University of Technology, GPO Box 2434, Brisbane, Qld 4001, Australia
Rob van Veen, Shell International Chemicals, B.V., Shell Research and Technology Centre Amsterdam, Badhuisweg 3, 1031 CM Amsterdam, The Netherlands
98 3.1 Abstract
The distribution of Al13 pillars and the process of intercalation in montmorillonite can be enhanced through the application of an ultrasonic treatment. This paper describes the results of ultrasonic treatment in the preparation of Al-pillared montmorillonite with and without prior exchange with Na+. The resulting materials have been charactersed by X-ray diffraction, N2 adsorption/desorption, Scanning
Electron Microscopy and Atomic Force Microscopy. The catalytic activity was tested with the n-heptane hydroconversion test. Optimum results were obtained after ultrasonic treatment up to 20 minutes without prior Na-exchange before the Al13 intercalation. Longer ultrasonic treatment resulted in partial destruction of the pillared structure. The pore size diameter also increased with increasing ultrasonic treatment up to 20 minutes with values in the range of 4 nm. This behaviour can be explained by the loss of the typical house of cards structure after prolonged ultrasonic treatment. AFM showed that the pillars in the interlayer of the montmorillonite resulted in a distortion of the tetrahedral sheets of the clay. At atomic scale resolution it was clear that the pillar distribution is not homogenous, confirming earlier results using high resolution TEM. The effects of ultrasonic treatment on the catalytic activity is rather limited, although the pillared clays prepared with short ultrasonic treatments of 5 and 10 minutes performed slightly better.
Key words: AFM, Al13, pillared clay, montmorillonite, n-heptane conversion, N2 adsorption/desorption, SEM, ultrasonic treatment.
99 3.2 Introduction
Acid treated montmorillonite clays were the catalysts commonly initially used for cracking reactions of hydrocarbons in the 1930’s [1]. These acid-treated smectites catalysts were replaced after World War II with a more stable synthetic silica-alumina type which also gave better product distribution [1]. The emergence of zeolites in the
1960's revolutionised the process mainly because of their high activity, selectivity and resistance to collapse when treated at high temperatures [2, p30]. Nowadayds, ZSM-5 and Y-zeolite are among the most popular heterogeneous catalysts in the petrochemical industry. The interest now is in producing a catalyst with a larger pore size compared to zeolite (∼8 Å) so as to handle the cracking of heavier crude oil. The use of pillared clays has received considerable attention [3]because of their ability to achieve large pore sizes, but factors such as the large volumes of water and chemicals involved in the preparation, the thermal stability and coking properties still need to be overcome.
Altering the preparation of a PILC can have dramatic effects on properties such as thermal stability and acidity [3-5]. This area also has received considerable attention with many authors who are looking at ways to economise the process for commercial viability. Current problems in preparation are time and energy costs, water usage and preparation of the expanded clay suspensions.
The preparation of a pillared clay normally starts with an cation exchange step where the hydrated interlayer cation of the smectite clay is exchanged for sodium.
This way an increase in swellability is achieved, making it easier to incorporate those large metal polyoxocomplexes. This cation exchange involves the use of large
100 amounts of water and sodium salts. It would be a significant improvement if this step could be altered into a less waste producing step.
The use of ultrasonics for the cation exchange step has been reported [6, 7]. A Ca- montmorillonite was intercalated with the Al13 complex using ultrasonic treatment over a number of time periods. The most intense and sharpest peaks in the XRD patterns were observed for the calcined sample that had been left in the ultrasonic bath for 20 minutes. The same authors, in a later study [7], described how the exchangeable cations present in the smectite affected the ultrasonic treatment. They converted the Ca-montmorillonite to Na+ and La3+ forms by ion exchange. This gave exchangeable cations with valencies of +1, +2 and +3. The optimum times for ultrasonic treatment were found to be 5 minutes for the Na-exchanged form, 20 minutes for the Ca-exchanged form and 80 minutes for the La-exchanged form. The increase in time was ascribed to the higher charge ions being more tightly bound to the clay layers. This method of intercalation has a number of advantages that help to make large-scale production of pillared clays more viable. Firstly, it reduces the time needed from several hours to less than 30 minutes. It also requires no heat for the process, thus saving in costs and reducing the safety risks, although some safety issues arise with ultrasonics that would need to be addressed. Finally, the clay suspension required don’t have to be cation exchanged and can be more concentrated compared to conventional methods, thus using less water, sodium salts and space.
This paper describes a detailed study on the use of ultrasonic treatment of a number of smectite clays for the intercalation with Al13 Keggin complexes.
101 3.3 Experimental
3.3.1 Starting materials
The starting materials used for this study were ≤ 2 μm fractions of Cheto montmorillonite SAz-1, and Miles montmorillonite from Queensland, Australia. The
Miles montmorillonite has a significantly higher CEC than SAz-1 (see special issue nr. 5 of Clays and Clay Minerals, volume 49, 2001 for a detailed characterization of
SAz-1). A detailed description of the Miles material and the conventional Al13- pillaring procedure have been provided by Kloprogge et al. [8]. Pillaring with Al13 of non-exchanged and Na-exchanged montmorillonites was executed in an ultrasonic bath with increasing time intervals from 0 to 30 minutes at room temperature. For all experiments a clay suspension of 30 % (w/w) in distilled water was prepared under stirring for 30 minutes prior to the intercalation with Al13. After washing and drying at room temperature for 24 hours the expanded clays were calcined at 450°C for 2 hours
(heating rate 5°C/min.).
3.3.2 Analytical techniques
3.3.2.1 X-ray diffraction (XRD)
The nature of the resulting material was checked by X-ray powder diffraction
(XRD). The XRD analyses were carried out on a Philips wide angle PW 1050/25 vertical goniometer equipped with a graphite diffracted beam monochromator (Fig.
3.1). The d-values and intensity measurements were improved by application of an in- house developed computer aided divergence slit system enabling constant sampling area irradiation (20 mm long) at any angle of incidence. The goniometer radius was enlarged to 204 mm. The radiation applied was CoKα from a long fine focus Co tube
102 operating at 35 kV and 40 mA. The samples were measured at 50 % relative humidity in stepscan mode with steps of 0.02° 2θ and a counting time of 2s.
3.3.2.2 Scanning Electron Microscopy (SEM)
Scanning electron microscope (SEM) images were obtained on a FEI Quanta 200
Environmental Scanning Electron Microscope (FEI Company, USA) operated at an accelerating voltage of 15 kV.
3.3.2.3 Atomic Force Microscopy (AFM)
Sample preparation involved a SiO2 surface for attachment of the clay sheets. An industry-standard n-type Si wafer with RMS roughness of 0.2 nm covered by a native oxide layer constituted the surface. The Si surface was exposed to ultrasonic cleaning with iso-propyl-alcohol, followed by rinses in doubly distilled and deionized water
(DDDW). Specimens of 1 cm2 area were sectioned and attached to standard AFM mounts. The clay was diluted with DDDW (filtered) and allowed to dry on the Si surface at room temperature (~23 °C).
103
20 minutes Counts
10 minutes
5 minutes
2.5 7.5 12.5 17.5 22.5 Degrees 2 θ
Fig. 3.1a Non-calcined Al13- montmorillonite SAz-1
Na-exchanged - 20min Usound
Na-exchanged - 10min Usound
Na-exchanged - 5min Usound Counts
Non-exchanged - 20min Usound
Non-exchanged - 10min Usound
Non-exchanged - 5min Usound
2.5 7.5 12.5 17.5 22.5 Degrees 2θ
Fig. 3.1b Non-calcined Na-exchanged Miles montmorillonite intercalated with Al13 104
20 minutes Counts
10 minutes
5 minutes
2.5 7.5 12.5 17.5 22.5 Degrees 2 q
Fig. 3.1c Non-calcined Miles montmorillonite intercalated with Al13
105 The work was carried out on a JEOL JSPM-4200 system with a 25 μm tube scanner, with a z-range of ca. 3 μm. The system is based on the detection of the tip-to- surface forces through monitoring optical deflection of a laser beam incident on a force-sensing/imposing lever. The analyses were carried out under air-ambient conditions (temperature of 23°C and 65% relative humidity). The probes were of the beam-shape variety in order to ensure that only the simple lowest-order bending modes contributed to the response. Probes were obtained from Ultrasharp NT-MDT.
The characteristics of probes employed in the present study are listed below.
Surface Designation R A k (Nm-1) Tip r Chem. N A <10 <20° Si-oxide 0.03
RTip = Manufacturers radius of curvature at tip apex; Ar = quoted full tip cone angle; kN= the force constant of the lever along the z-axis (i.e., normal to the surface plane).
3.3.2.3.1 Operational modes
Contact mode: Frictional images were carried out in constant height mode with a lever-imposed loading in the range 5-20 nN. The scanning rate in the fast-scan direction was ca. 3 Hz, and a typical image was composed of 256x256 pixels.
3.3.2.3.2 Data processing
Frictional images were processed by subtraction of background and adjustment of brightness and contrast. Some images were enhanced through Fast Fourier Transform processing.
106 Counts 30 minutes
20 minutes
10 minutes
Sonicated 5 minutes
2.5 4.5 6.5 8.5 10.5 12.5 14.5 16.5 18.5 Degrees 2-Theta
Fig. 3.2a Al-pillared montmorillonite SAz-1 calcined at 450°C
30 minutes Counts
20 minutes
10 minutes
Sonicated 5 minutes
2.5 7.5 12.5 17.5 22.5 Degrees 2-Theta
Fig. 3.2b Al-pillared Miles montmorillonite calcined at 450°C
107 30 minutes Counts
20 minutes
10 minutes
Sonicated 5 minutes
2.5 4.5 6.5 8.5 10.5 12.5 14.5 16.5 18.5 Degrees 2-Theta
Fig. 3.2c Al-pillared Na-exchanged Miles montmorillonite calcined at 450°C
3.3.3 N2 adsorption/desorption
The surface areas of the starting montmorillonite and Al13 pillared
montmorillonites were calculated by the BET method using N2 adsorption-desorption
on a Micromeritics ASAP 2010 at a partial pressure range of 0.06 to 0.30. The pore
size distribution and pore volume were calculated using the Tristar software.
3.3.4 Catalytic testing
The pillared clays were prepared for hydroconversion of n-heptane by loading the
pillared clays with 0.4 wt% Pd by impregnation with tetramine. Catalysts were dried
at 120°C for 16 hours and subsequently reduced in a hydrogen flow (H2 flow rate 2.24
Nml/min, total pressure 30 bar) at 400°C for 2 hours. The conversions were carried
out in a conventional fixed bed reactor under various reaction temperatures to obtain a
constant conversion of 40%. Reaction conditions were: n-heptane/H2 molar ratio of
108 0.25, total pressure 30 bar, and GHSV = 1020 Nml/(g.h). Effluents were analysed by on-line gas chromatography. For comparative reasons two commercial catalysts were treated in a similar fashion as the pillared clays.
3.4 Results and discussion
Conventionally, the starting clay is pre-exchanged with sodium to assist the ion exchange process that occurs during the intercalation step. Then the Na-exchanged clay suspension is mixed with the pillaring solution over a time period with heat. This process can take up to 6 hours to complete as the pillaring solution is added drop-wise and the mixture is stirred for at least 2 hours at around 80oC. Although this method has had success in laboratory synthesis of pillared clays, it is not an ideal preparation technique for large-scale production as large amounts of water and heat are necessary.
Intercalation of Miles montmorillonite, its Na-exchanged form and SAz-1 with
Al13 in an ultrasonic bath resulted in PILCs with interlayer spaces of around 8.5 Å, taking into account a thickness of 9.8 Å for the 2:1 silicate layers, which compares well to that of a conventionally prepared PILC (Fig. 3.1). However, experiments at 0 minutes ultrasonic treatment did not give an expanded montmorillonite, the XRD pattern of a normal montmorillonite was obtained instead. The results showed that the
Na-exchanged PILC retained its XRD peak up to 20 minutes ultrasound treatment. It also showed no difference between the Na-exchanged form and the non-exchanged form thus doing away with the need for this preliminary step. The XRD patterns of the intercalated clay however, did show the formation of another peak at 30 minutes ultrasound duration, suggesting some loss of regularity in the layer structure. In general the best results were obtained with the shortest ultrasonic treatments,.i.e. 5 or
10 minutes (Fig. 3.2).
109 The N2 adsorption/desorption isotherms of the pillared clays display a typical hysteresis loop (Fig. 3.3), consistent with slit-like pores in the pillared clays. The BET surface areas are the highest for the pillared clays prepared with 5 or 10 minutes ultrasonic treatment (Table 3.1). Longer treatment does not result in a further increase and in most cases a significant decrease in surface area, supporting the XRD results.
The change in BET surface area however is not accompanied by any significant changes in the total pore volume (<144 nm pores), which is on average 0.13 cm3/g for all pillared clays.
The SEM images show that with increasing the ultrasonic treatment time the house of cards type structure, where most of the clay particles are in an edge-to-edge or edge-to-face type orientation is lost and is being replaced with an increasing amount of face-to-face orientations (Fig. 3.4). This difference in particle orientation can explain the loss in BET surface area, as a significant part of the external surface is lost upon face-to-face orientation of the clay particles. This also means that the changes observed in the XRD are mainly due to the changes in the interlayers of the clay where the regularity in the pillar distances is the main contributor to the width of the basal reflections.
110
Fig. 3.3a BET adsorption and desorption isotherms for Al-pillared montmorillonite SAz-1 after ultrasonic treatment for 5 minutes and calcined at 450°C
111
Fig. 3.3b BET adsorption and desorption isotherms for Al-pillared Miles montmorillonite after ultrasonic treatment for 5 minutes and calcined at 450°C
Clay SAz-1 Na- Na- Na- Miles Miles Miles exchanged exchanged exchanged Miles Miles Miles Treatment 5 min. 5 min 10 min 20 min 5 min. 20 min 30 min Ultrasonic, ultrasonic, ultrasonic, ultrasonic, Ultrasonic, ultrasonic, ultrasonic, calcined calcined calcined calcined calcined calcined calcined 450°C 450°C 450°C 450°C 450°C 450°C 450°C BET surface area 174 121 129 119 172 154 156 (m2/g) Total pore 0.13 0.13 0.12 0.12 0.14 0.14 0.14 volume of pores < 144 nm (cm3/g) Average pore 2.98 4.25 3.83 4.01 3.34 3.52 3.62 diameter (nm)*
Table 3.1 BET surface area and pore volume and average pore diameter of Al- pillared montmorillonites
112
Fig 3.4a Al-pillared montmorillonite SAz-1 (ultrasonic treatment 5 minutes, calcined at 450°C)
Fig 3.4b Al-pillared montmorillonite SAz-1 (ultrasonic treatment 10 minutes, calcined at 450°C) 113
Fig 3.4c Al-pillared montmorillonite SAz-1 (ultrasonic treatment 20 minutes, calcined at 450°C)
114 In order to obtain a better understanding of the pillar distribution and how this is related to the changes in the pore size high resolution AFM has been used. Fig. 3.5a gives an example of the starting montmorillonite with its layers oriented parallel to the mica surface. It clearly shows the hexagonal pattern associated with the six-membered rings of SiO4 tetrahedra in the tetrahedral sheet of the montmorillonite. The introduction of the Al-pillars results in a rather broad distortion of this pattern. Figures
3.5 show examples of the effects of Na-exchange prior to the pillaring and the prolonged ultrasonic treatment. The introduction of the Al-pillars with increasing ultrasonic treatment resulted in an increase in the distortion of the tetrahedral sheet.
The images also show that these distortions are not present at a regular interval, which may indicate that the pillars are not present in the interlayers at a regular distance.
There is no evidence for the presence of any pillars on the outside surface of the clay particles. This supports earlier work where high resolution TEM showed similar irregularities in the pillar distances between layers but also within layers [9].
The catalytical activity of the pillared clays was tested for the n-heptane conversion after loading the pillared clays with 0.4 wt% Pd. For comparison two commercially used catalysts have been tested (Table 3.2). In general the starting clay and Al-pillared clay prepared via the standard method without ultrasonic treatment did not perform well with the temperature required for 40% conversion around 425°C, which is comparable to Pd loaded on Al2O3 modified with TEOS (resulting in a very weak solid acid). In contrast the commercially used catalyst ASA (amorphous silica/alumina from American Cyanamid loaded in a similar fashion, and previously described in [10]) gives a temperature of 348°C. All the pillared clays gave temperatures in the range between 355 and 368°C. Minor differences are observed with respect to the preparation methods used. The use of Na-exchange prior to the
115 pillaring resulted in the worst performing catalysts. This reaffirms the previous XRD observation that the best material was obtained without the Na-exchange. The effect of the time used for the ultrasonic treatment shows no significant differences although the pillared clays ultrasonically treated for 5 or 10 minutes performed slightly better than those treated for 20 minutes. Overall the selectivity for the isomers is high and there is almost no cracking up to about 60% conversion. Exceptions are the starting montmorillonite and the pillared clay prepared via the standard method without ultrasonic treatment, as thermal cracking is easily achieved at temperatures above
400°C.
116
Fig 3.5a AFM raw image (left) and FTIR processed image (right) of Miles montmorillonite
Fig 3.5b AFM raw image (left) and FTIR processed image (right) of Miles montmorillonite after 10 minutes ultrasonic treatment
117
Fig 3.5c AFM raw image (left) and FTIR processed image (right) of Al-pillared Miles montmorillonite after 20 minutes ultrasonic treatment
Sample T(°C) required for 40% conversion Pd/Al2O3:Si 425 ASA 348 Cheto montmorillonite 424 Al-pillared Miles conventional method 426 Na-exchanged Al-pillard Miles 5 min US 368 Na-exchanged Al-pillard Miles 10 min US 362 Na-exchanged Al-pillard Miles 20 min US 368 Al-pillared Miles 5 min US 360 Al-pillared Miles 10 min US 363 Al-pillared Miles 20 min US 359 Al-pillared Cheto 5 min US 366 Al-pillared Cheto 5 min US 355 Al-pillared Cheto 5 min US 363
Table 3.2 Temperatures for 40% n-heptane conversion on Pd-loaded pillared clays (calcined 450°C, Pd loading 0.4 wt%) and two standard catalysts (ASA and Pd/Al2O3:Si)
118 3.5 Conclusions
This work shows that the use of ultrasonic treatment results in pillared clays without the need for Na-exchange prior to the pillaring process. The pillar distribution is clearly affected by the ultrasonic treatment. Overall best results are obtained with short ultrasonic treatment times, while prolonged ultrasonic treatment results in a decrease in surface area and pore diameter associated with a loss of regularity in the interpillar distances. The decrease in pore size and pillar distribution has a minor but observable effect on the catalytic activity in the n-heptane conversion test, where the pillared clays with the longest ultrasonic treatment performed worst. This means that the preparation of pillared clays on a commercial scale can be easier achieved since the extensive washing and sodium-exchange step can be left out and the intercalation of the pillaring complexes can be achieved in a much smaller timeframe without loosing catalytic activity.
3.6 Acknowledgements
The Inorganic Materials Research Program is thanked for the financial Support.
The authors wish to thank Chris Hildebrandt for his help in the experimental work.
We would also like to thank Shell Research and Technology Centre Amsterdam for their help with the catalytic testing and Greg Watson of the Scanning Probe
Microscopy Laboratory of Griffith University, Brisbane, for his help with the AFM work.
119 3.7 References
1. J. M. Thomas and W. J. Thomas, Principles and practice of heterogeneous catalysis, VCH Publishers Inc., New York, 1997. 2. P. B. Venuto and J. E. Thomas Habib, Fluid catalytic cracking with zeolite catalysts, Vol. 1, Marcel Dekker, Inc New York, 1979. 3. J. T. Kloprogge, Journal of Porous Materials 5, 5 (1998). 4. A. Vaccari, Applied Clay Science 14, 161 (1999). 5. R. Burch, Catalysis Today 2, 185 (1988). 6. S. P. Katdare, V. Ramaswamy and A. V. Ramaswamy, Catalysis Today 49, 313 (1999). 7. S. P. Katdare, V. Ramaswamy and A. V. Ramaswamy, Microporous and Mesoporous materials 37, 329 (2000). 8. J. T. Kloprogge, R. Evans, L. Hickey and R. L. Frost, Applied Clay Science 20, 157 (2002). 9. L. V. Duong, T. E. Bostrom, J. T. Kloprogge and R. Frost, L., Microporous and Mesoporous Materials 82, 165 (2005). 10 E. Booij, J.T. Kloprogge and van Veen, J.A.R., Clays and Clay Minerals 44, 774 (1996)
120
Statement of Contribution
The authors listed below have certified* that:
1. they meet the criteria for authorship in that they have participated in the conception, execution, or interpretation, of at least that part of the publication in their field of expertise; 2. they take public responsibility for their part of the publication, except for the responsible author who accepts overall responsibility for the publication; 3. there are no other authors of the publication according to these criteria; 4. potential conflicts of interest have been disclosed to (a) granting bodies, (b) the editor or publisher of journals or other publications, and (c) the head of the responsible academic unit, and
5. they agree to the use of the publication in the student’s thesis and its publication on the Australasian Digital Thesis database consistent with any limitations set by publisher requirements. In the case of this chapter:
HIGH RESOLUTION XPS STUDY OF THE INTERNAL STRUCTURE OF Al- AND Ga- PILLARS IN PILLARED CLAY CATALYSTS
Loc V. Duong, .Theo Kloprogge, R. Frost and Barry J. Wood Submited December 2007
Contributor Statement of contribution*
Loc Duong wrote the manuscript, designed experiments, conducted experiments, data analysis, data interpretation
Date
Supervision, aided with data analysis and interpretation Theo Kloprogge
Ray Frost editing
Barry Wood Aid with XPS techniques
Principal Supervisor Confirmation
I have sighted email or other correspondence from all Co-authors confirming their certifying authorship and.
______Name Signature Date CHAPTER 4
4.0 HIGH RESOLUTION XPS STUDY OF THE INTERNAL STRUCTURE OF AL- AND GA- PILLARS IN PILLARED MONMORILLONITE
Loc V. Duong, J.Theo Kloprogge, R. Frost and Barry J. Wood
Submitted
Loc Duong, Inorganic Materials Research Program, School of Physical and Chemical
Sciences, Queensland University of Technology, GPO Box 2434, Brisbane, Qld
4001, Australia
Theo Kloprogge, Inorganic Materials Research Program, School of Physical and
Chemical Sciences, Queensland University of Technology, GPO Box 2434,
Brisbane, Qld 4001, Australia
Ray Frost, Inorganic Materials Research Program, School of Physical and Chemical
Sciences, Queensland University of Technology, GPO Box 2434, Brisbane, Qld
4001, Australia
Barry J. Wood, Brisbane Surface Analysis Facility, University of Queensland,
Brisbane, Qld 4072, Australia
122 4.1 Abstract
Ga and Al-pillared clays were prepared using montmorillonite (SWy-2) as the starting material. The complexing agents Al13 and Ga13 were prepared through hydrolysis of
Al and Ga nitrate solutions with NaOH up to a Al/OH and Ga/OH molar ratio of 2/1.
After ion exchange the resulting intercalated montmorillonites were calcined at 450° for 8 hours. XRD shows an increase in the basal spacing from 14Ǻ to about 19Ǻ. XPS high resolution scans of Si 2p confirms the presence of only 1 type of Si in both the starting montmorillonite and the pillared samples. The Al 2p scan of the starting montmorillonite shows two bands at 75.90 eV associated with octahedral Al and at
76.23 eV associated with tetrahedral Al in the clay layers. The corresponding O 1s is dominated by a single strong band at 532.62 with a minor water band at 535.44 eV.
The Ga-pillared montmorillonite shows two bands in the Ga 2p high resolution scan at
1118.99 eV and 1120.77 eV associated with GaVI and GaIV , respectively. In the O 1s spectrum an extra band at 531.93 eV is visible, associated with the oxygen in the pillar structure in addition to the earlier observed bands associated with the clay layers. A similar result is obtained for the O 1s spectrum in the case of the Al-pillared clay. Chemical analysis of the montmorillonite and its corresponding pillared samples shows that 2.01 Ga and 2.76 Al are present in the clay interlayer per formula unit of
Al4(Si7.38Al0.62)O20(OH)4. Based on the montmorillonite layer charge and the amount of pillars present in the interlayer it is calculated that progressive hydrolysis has reduced the effective charge of 7+ present on the Al13 or Ga13 complex in solution to around 3 to 4+ in the pillars present in the clay interlayers.
Key words: Al13 Keggin structure, pillared clay, montmorillonite, XPS
123 4.2 Introduction
Smectite, the 2:1 dioctahedral phyllosilicates with the surface charge per unit cell from 0.4-1.2, has been commonly used as starting material for pillaring with different inorganic species for mainly catalytic purposes. Pillaring is a process in which the clay layers are propped apart by large inorganic species or props followed by a calcination step during which the prop is converted to a covalently bonded structure known as the pillar (Kloprogge, 1998, Kloprogge, 2005,(Bergaya, 2006 *). The most common inorganic complex which has been used is Al13. The Al13 was first reported by Johansson in 1960 (Johansson, 1960)and has a Keggin type cage structure. The final Al pillared montmorillonite would provide the microporous structure with the d spacing of about 18Å. Most of research has been done on the structure of the Al13 pilllared montmorillonite using XRD, SSA, NMR and TEM. Vibrational spectroscopy is a technique which helps to uncover the changes in the structure of the pillars after calcination
A major problem in this type of characterization is that it is not easy to distinguish between the Al from the pillars and the Al from the silicate structure. Ga13 provides a similar Keggin structure as Al13 but is in general more stable and fit in much better than Al in the Keggin structure. The similarity in the structure between Ga and Al pillared montmorillonite has been discussed by (Duong et al., 2005a) using TEM and microanalysis study. The use of gallium cation however allows one to now clearly distinguish between what is happening in the pillar structure and in the clay layers during the calcination step.
X ray photoelectron spectroscopy, introduced in 1967, is a surface technique which using the concept of binding energy different from one element to other and varies from local environment which reflects the bonding details of the element
124 involved in the structure (Siegbahn et al., 1967). The penetration depth (escape depth of the photoelectron) is about 0.5 - 10nm (50-100 Ǻ) which equals to about 5 atomic clay layers and is depended on the matrix and the X-ray source. Due to the very small size of smectite particles of less than 5 micron in diameter and up to about 20 nm in thickness, XPS can be used for the study of clay minerals as a bulk analytical technique instead of a surface technique. Earlier work in our laboratory has used XPS to study a series of Source Clays from The Clay Minerals Society Repository
(Kloprogge et al., 2008)
This paper reports the first experiments done on Al13- and Ga13-pillared montmorilonites using XPS with the objective to study in detail the changing in the pillar structure of the pillared clays. The structure of Al13 pillars has been compared to the Keggin structure in a basic aluminium sulphate and Ga13 pillars in the Ga13 pillared montmorillonite in order to gain a better understanding of the changes in the pillar structure upon calcination and the bonding structure of the pillared clay before and after calcination
125 4.3 Experimental
4.3.1 Sample preparation
The starting materials used for this study were ≤ 2 μm fractions of Wyoming montmorillonite SWy-2. All samples were saturated with sodium through exchange with 1 M NaCl for 8 hours. The clays were washed five times with deionised water in order to remove residues of NaCl. A detailed description of the Al13-pillaring procedure have been provided by Kloprogge et al. (2002). SWy-2 montmorillonite was also used to prepare the Ga13 pillared clay. The preparation of the Ga13-pillared clays was similar to that of the Al13-pillared clay. A 0.1M solution of NaOH was added to Ga(NO3)3 at a rate of 0.01 ml/min using a peristaltic pump under vigorous stirring at room temperature. The OH/Ga ratio was 2:1. The Al13 and Ga13 solutions were added to the aqueous clay suspensions under continuous stirring during four hours. The suspensions were then allowed to stand for several days. The pillared clays were washed 5 times with deionised water using a centrifuge. The samples were allowed to dry in air at ambient temperature. Finally, the samples were heated at
2°C/min and calcined at 450°C for 8 hours.
4.3.2 X-ray diffraction (XRD)
The nature of the resulting material was checked by X-ray powder diffraction.
The XRD analyses were carried out on a Philips wide angle PW 1050/25 vertical goniometer equipped with a graphite diffracted beam monochromator. The d-values and intensity measurements were improved by application of an in-house developed computer aided divergence slit system enabling constant sampling area irradiation (20 mm long) at any angle of incidence. The goniometer radius was enlarged to 204 mm.
The radiation applied was CoKα from a long fine focus Co tube operating at 35 kV
126 and 40 mV. The samples were measured at 50 % relative humidity in stepscan mode with steps of 0.02° 2θ and a counting time of 2s.
4.3.3 N2 adsorption/desorption
The surface areas of the starting montmorillonite and Al13 pillared montmorillonites were calculated by the BET method using N2 adsorption-desorption on a Micromeritics ASAP 2010 at a partial pressure range of 0.06 to 0.30.
4.3.4 X-ray Photo-electron Spectroscopy
The samples were analyzed in freshly powdered form in order to prevent surface oxidation changes. Prior to the analysis the samples were outgassed under vacuum for 72 hours. The XPS analyses were performed on a Kratos AXIS Ultra with a monochromatic Al X-ray source at 150 W. Each analysis started with a survey scan from 0 to 1200 eV with a dwell time of 100 milliseconds, pass energy of 160 eV at steps of 1 eV with 1 sweep. For the high resolution analysis the number of sweeps was increased, the pass energy was lowered to 20 eV at steps of 100 meV and the dwell time was changed to 250 milliseconds.
Band component analysis was undertaken using the Jandel ‘Peakfit’ software package, which enabled the type of fitting function to be selected and allows specific parameters to be fixed or varied accordingly. Band fitting was done using a Lorentz-
Gauss cross-product function with the minimum number of component bands used for the fitting process. The Gaussian-Lorentzian ratio was maintained at values greater than 0.7 and fitting was undertaken until reproducible results were obtained with correlations of r2 greater than 0.995.
127 4.4 Results and discussion
X-ray powder diffraction patterns of the starting montmorillonites and Ga13 and Al13 pillared montmorillonites demonstrated that the Ga and Al had been pillared successfully. The exchange of montmorillonite Swy-2 with large Keggin type molecules such as Al13 and Ga13 results in an expansion of the basal spacing from 14
Ǻ to about 19-20Ǻ (Table 4.1). Upon calcination these rather bulky complexes convert to a stable oxide bonded to the tetrahedral sheets. During this conversion there is a slight decrease in basal spacing. In addition pillaring with both the Ga and Al complexes strongly increased the BET surface area (Table 4.1).
Earlier work of Duong et al. (2005a) has shown that the pillars are relatively evenly distributed between the clay layers at crystal size resolution but that at high resolution there are distinct differences in the distribution. The use of high resolution
TEM did not provide any detailed information about the exact structure of the pillar after calcination. Kloprogge et al. (1999) used infrared emission spectroscopy to determine the changes that take place upon calcination of Al13 exchanged clays.
Exchange of montmorillonite with Al13 gave Al–OH and Al–H2O OH-stretching
−1 modes of Al13 at 3682 and 3538 cm . Calcination resulted in removal of the Al–OH
−1 mode >400°C. The Al–H2O band was replaced by bands at 3574 and 3505 cm , due to structural rearrangement within Al13. The intensities diminished, but are still observed at 800°C, suggesting that the pillar structure incompletely converted to
"Al2O3." A similar observation was made by Balek et al. (1998) using Emanation
−1 Thermal Analysis. Below 1750 cm the Al13–montmorillonite displays bands at 642,
1008, 1321, 1402, and 1512 cm−1. The 1512 cm−1 band disappeared at 500°C, followed by the other bands above 600°C. The 642 cm−1 band intensity diminished but
128 is still observed at 800°C. At 700°C a new band is observed at 722 cm−1due to Al–O bond formation. However no details about the exact pillar structure were obtained.
Basal spacing BET surface area Total Pore volume (001) (Ǻ) (m2/g) (<144nm) (cm3/g) SWy-2 14.0 22.3 SWy-2
Al13 exchanged 18.3 na Al13 exchanged
Ga13 exchanged 19.9 na Ga13 exchanged
GaAl12 pillared 19.1 122 GaAl12 pillared
Al13 pillared 17.3 120 Al13 pillared
Ga13 pillared 17.9 118 Ga13 pillared
Table 4.1 XRD and N2 adsorption/desorption results for the starting montmorillonite SWy-2, the Al13- and Ga13-intercalated montmorillonites and the calcined Al13- and Ga13-pillared montmorillonites
Plee et al.(1985) have used 27Al and 29Si solid state nuclear magnetic resonance (MAS-NMR) techniques to study the thermal transformation of Al pillars and their linkage with the clay sheets. Their 27Al MAS NMR spectra of beidellite revealed two separate AlIV resonances due to the tetrahedral Al present in the clay structure and in the centre of the Al13 pillar. In all other publications an overlap of both signals is reported instead of separate signals making it impossible to draw any definitive conclusions on the changes in the pillar structure upon calcination
(Kloprogge et al., 1994; Malla and Komarneni, 1993).
X-ray photoelectron spectroscopy is not only sensitive to the chemical composition but is also sensitive to the local environment of atoms in a crystal structure, which is reflected in changes in the binding energy and the occurrence of
129 multiple bands associated with different chemical environments. A good example of this was recently published on bauxite minerals, where boehmite (AlOOH) showed two distinct oxygen bands associated with an oxygen atom linked to the aluminium atom and the hydroxyl group. Similar observations for pillared clays might shed more light on the pillar structure after calcination.
XPS is in general considered to be a surface analysis technique. However, for powders with very small particle sizes, such as commonly observed for clay minerals, the analysis can be assumed to be close to a bulk analytical technique since the penetration depth of the X-rays is in the order of up to 100 Å. Figure 4.1 shows the
XPS survey scan of the starting montmorillonite. The major advantages of XPS as an analytical tool are the absence of sample preparation, rapid analysis and multi- elemental analysis including elements such as F, Cl or Li. Fig. 4.2 exhibits the high resolution scans for the O 1s, Si 2p and Al 2p of the starting clay. Since all clay minerals belong to the phyllosilicate group where silicon is only present as SiO4 tetrahedra linked together through their three basal oxygen atoms, a single band will be observed. This is indeed the case; a single Si 2p band is observed with a binding energy of 103.42 eV. Aluminium in montmorillonite can however be present in two different positions. First of all aluminium in six-coordination can be present in the octahedral sheet, which is sandwiched between two silicon tetrahedral sheets and secondly aluminium can be present in four-coordination as substitution of silicon in the tetrahedral sheet. A similar split has been observed in a series of clay minerals provided by the Clay Minerals Society Source Clay Repository (Kloprogge et al.,
2008). The O 1s is dominated by a single strong band at 532.62 eV with a minor band at 535.44 eV.
130 20000 Starting montmorillonite Swy-2 O 1s
15000 ) Al 2s
CPS Mg 2p (
y 10000 O Auger Mg Al 2p Na 1s Si 2s Auger Intensit Fe 2p Si 2p 5000 Na C 1s Auger
0 1200 1100 1000 900 800 700 600 500 400 300 200 100 0 Binding Energy (eV)
Fig. 4.1 XPS survey scan of the Wyoming montmorillonite starting material
The main band is due to the oxygen in the clay layers while the minor band is associated with a small amount of strongly absorbed water. Interestingly no distinction can be made between oxygen atoms and hydroxyl groups in the clay layers in contrast to some of the oxohydroxides. Exchange with Ga13 followed by calcination is thought to result in the transformation of the Ga13 into an oxidic material covalently bonded to the tetrahedral sheets of the montmorillonite. Ga13 is a Keggin complex similar to Al13 and thus exists of a central four-coordinated Ga surrounded by 12 six- coordinated Ga. Upon calcination of the Ga13-exchanged clay the resulting high resolution Ga 2p spectrum show two bands that, similar to Al13 in basic aluminium sulfate (Duong et al., 2005b), reflect the presence of both tetrahedral and octahedral
131 Ga in the pillar even after calcination (Fig. 4.3). The major band observed at 1118.99 eV is associated with the six-coordinated Ga while the small band at 1120.77 eV is associated with the four-coordinated Ga. The observed ratio of 1:11 is close to the theoretical ratio of 1:12. The Si 2p (103.53 eV) and Al 2p (75.42 and 76.31 eV) spectra do not change upon pillaring with Gallium. The O 1s spectrum is also rather similar to that of the original starting clay with a major band at 532.81 eV and a minor band associated with strongly bonded water at 534.37 eV. However a new band can be observed at 531.93 eV. Based on the above observations there is direct evidence for the formation of a different type of oxygen in the pillar structure. It appears that the local environment of the Si and Al in the tetrahedral sheets as reflected in the high resolution Si 2p and Al 2p spectra is not significantly influenced through the formation of covalent bonds with the Ga-pillar.
The observation of changes in the pillar structure and the bonding to the clay layers of aluminium pillars has always been hampered by the fact that a significant amount of aluminium is present in the clay structure. This makes independent observation of what happens in the pillar structure upon calcination very difficult.
Earlier work has shown that basic aluminium sulfate with the Al13 Keggin structure still intact exhibits two bands in the high resolution Al 2p spectrum (Duong et al.,
2005b). These two bands at 74.6 and 73.8 eV have been identified as the octahedral and tetrahedral Al respectively (Fig. 4.4). The interpretation of the 74.6 eV band as the octahedral Al is supported by similar binding energies observed for gibbsite and corundum. Upon exchange of Al13 and subsequent calcination of the intercalated montmorillonite a similar high resolution Al 2p spectrum is obtained with a major band at 74.37 eV and a minor band at 73.89 eV in both the exchanged and calcined samples (Fig. 4.5). Similar to the Gallium pillars the presence of tetrahedral
132 coordinated Aluminium is preserved in the pillared clay. Due to the excess amount of aluminium present in the pillared clay it is impossible to observe the bands associated with the aluminium in the clay layers. However, the high resolution Si 2p spectrum shows that there is no change in the binding energy of the silicon in the tetrahedral sheets, in agreement with the observations for the Ga-pillared montmorillonite.
Similarly the high resolution O 1s spectrum is almost identical to that observed for the
Ga-pillared montmorillonite with the O 1s of the clay layer at 532.75 eV, strongly bonded water at 534.38 eV and that of the pillar at 532.02 eV.
The chemical analyses based on the XPS spectra of the starting clay and its
Al- and Ga-pillared equivalents are represented in Table 4.2. The analysis of the montmorillonite SWy-2 and the Ga-pillared clay are in close agreement with earlier published chemical analyses based on single crystal energy dispersive X-ray analysis within the Transmission Electron Microscope (Duong et al., 2005a), taking into account that no analysis was included of the Fe and Mg content in the starting material. Comparison between the starting montmorillonite and the Ga- and Al- pillared clay shows the presence of 2.01 Ga and 2.76 Al per formula unit of
Al4(Si7.38Al0.62)O20(OH)4. The amount of Al bound in the interlayer per unit cell has been shown to vary only within a small range (2.78–3.07), equivalent to approximately one Al13 per 4.2 to 4.6 unit cells, and shows no correlation with the charge of the layer. The absence of a correlation suggests a more or less uniform monolayer of hydrated Al or Ga polyoxocations to be present in the interlayer
(Kloprogge, 1998),. Bergaoui et al., (1995) indicated that the amount of Al never exceeds one Al13 per 6 unit cells, due to steric constraints at the solid-liquid interface.
133 Al13 GaAl13- Swy2 exchanged Al-pilc pilc Ga-pilc Si 19.21 13.44 16.21 15.73 15.90 Al 11.81 13.68 16.21 14.43 9.78 Na 0.77 - - - - Ca 0.73 - - - - O 67.47 72.88 67.57 69.45 72.37 Ga - - - 0.39 1.97
Si/Al 1.6263 0.9825 1.0000 1.0901 1.6263 Si/O 0.2847 0.1844 0.2397 0.2265 0.2197 Si/Ga - - - 40.33 8.0711
Si 7.38 7.38 7.38 7.38 7.38 Al clay 4.62 4.62 4.62 4.62 4.62 Al pillar - 2.89 2.76 2.15 - Ga - - - 0.18 2.01 Na 0.21 - - - - Ca 0.42 - - - - O clay layer 24.00 24.00 24.00 24.00 24.00 O pillar - 8.90 5.25 6.26 6.85
H2O 1.92 7.11 1.53 2.74
O 1s clay 92% 78% 72% O 1s pillar - 17% 20%
O 1s H2O 8% 5% 8%
Table 4.2 Chemical analysis (at%) of the starting clay SWy-2 and the Al and Ga pillared equivalents
The montmorillonite layer charge of 0.62 and the amount of Ga and Al present in the pillars indicates that the charge on the Ga13 pillar upon intercalation and
134 washing has changed due to progressive hydrolysis from 7+ to about 4+ and for Al13 from 7+ to about 3+ in agreement with earlier work on Al13-pillared clays. Jones and
Purnell (1993) have shown that the Al-pillars exhibit an ionic charge slightly above
3+ instead of 7+ as assumed for Al13, which explains why more Al is introduced in the clay interlayer than a 7+ charged Al13 would allow (Purnell, 1990).
The Ga/O ratio in the pillar is about 0.29 indicating that the overall chemical composition of the pillar is rather different from that of the standard oxide Ga2O3 with a ratio of 0.67 as would be expected. Similarly the Al/O ratio in the Al pillars is about
0.53 instead of 0.67 as expected for Al2O3. A possible explanation for the difference may be the presence of residual hydroxyl groups on the outside of the pillars as a result of incomplete conversion of the Al13 complex upon calcination. Infrared emission spectroscopy and Emanation Thermal Analysis have shown that surface hydroxyls are present on Al13 pillared clays after calcination at 450°C (Kloprogge et al., 1999; Balek et al., 1998).
135 1000 Si 2p montmorilonite 900
800
700 600 103.42 eV 500 400 Intensity (CPS) 300
200
100 0 107 106 105 104 103 102 101 100 Binding energy (eV)
600 Al 2p montmorillonite
500 VI Al 75.90 eV 400
300
Intensity (CPS) IV 200 Al 76.23 eV
100
0 80 79 78 77 76 75 74 73 72 71 Binding energy (eV)
3000
O1s montmorillonite
2500 O 1s 532.62 eV
2000
1500 O 1s H2O 535.44 eV
(CPS) Intensity 1000
500
0 538 537 536 535 534 533 532 531 530 529 528 Binding energy (eV)
Fig. 4.2 High resolution XPS spectra of Si 2p, Al 2p and O 1s of the starting montmorillonite Swy-2
136 1000 600 Si 2p montmorilonite Al 2p montmorillonite 900 500 800 AlVI 75.90 eV 700 ) 400
600 )
CPS 103.42 eV ( CPS
( y 500 y 300
400 Intensit
Intensit IV 300 200 Al 76.23 eV
200 100 100
0 0 107 106 105 104 103 102 101 100 80 79 78 77 76 75 74 73 72 71 Binding energy (eV) Binding energy (eV)
700 1600 Ga 2p Ga-pillared montmorillonite O 1s Ga-pillared montmorillonite 1400 600 O 1s clay 532.81 e V GaVI 1118.99 eV 1200 500 ) ) 1000 CPS CPS 400 (
(
y y 800 300 GaIV 1120.77 eV 600 Intensit O 1s pillar 531.93 eV Intensit O 1s H2O 534.37 eV 200 400
100 200
0 0 1124 1123 1122 1121 1120 1119 1118 1117 1116 1115 538 537 536 535 534 533 532 531 530 529 528 Binding energy (eV) Binding energy (eV)
Fig. 4.3 High resolution XPS spectra of Si 2p, Al 2p, Ga 2p and O 1s of the Ga-pillared montmorillonite Swy-2
137 400 Al 2p Al13 sulfate 350 calcined 400οC
300 AlVI 74.6 eV 250
200 150 AlIV 73.9 eV Intensity (CPS)
100
50
0 80 79 78 77 76 75 74 73 72 71 70 Binding energy (eV)
300 Al 2p Al-pillared montmorillonite
250
VI 200 Al 74.37 eV
150 AlIV 73.89 eV
Intensity (CPS) 100
50
0 80 79 78 77 76 75 74 73 72 71 Binding energty (eV)
3000 Al 2p corundum
2500
2000 AlVI 74.2 eV
1500
Intensity (CPS) Intensity 1000
500
0 80 79 78 77 76 75 74 73 72 71 70 Binding energy (eV)
Fig. 4. 4 High resolution XPS spectra of Al 2p of Al13 sulfate after calcination at 400°C (top ), gibbsite with only AlVI (middle), and corundum with AlVI (bottom )
138 250 Si 2p Al-pillared montmorillonite 200
150
102.66 eV 100
Intensity (CPS)
50
0 107 106 105 104 103 102 101 100 Binding energy (eV)
300 Al 2p Al-pillared montmorillonite 250
VI 200 Al 74.37 eV
150
AlIV 73.89 eV
Intensity (CPS) Intensity 100
50
0 80 79 78 77 76 75 74 73 72 71 Binding energty (eV)
5000 O 1s Al-pillared montmorillonite 4500 4000 O 1s clay 532.75 eV 3500
3000 2500
2000 O 1s pillar 532.02 eV Intensity (CPS) 1500 O 1s H2O 534.38 eV 1000
500
0 538 537 536 535 534 533 532 531 530 529 528 Binding energy (eV)
Fig. 4. 5 High resolution XPS spectra of Si 2p, Al 2p and O 1s of the Al- pillared montmorillonite Swy-2
139 4.5 Conclusions
This paper successfully describes the use of high resolution XPS for the study of pillared clays. Detailed XPS analysis has shown to provide detailed information not only about the chemical composition of the clay and its pillared equivalents, but also provided insight in the chemical composition of the pillars itself. The change in ionic charge due to progressive hydrolysis during the intercalation and washing steps can be calculated to be from 7+ to values around 3 to 4+. The results of the Ga and Al-pillared clay are in close agreement with earlier published work applying different analytical techniques.
4.6 Acknowledgements
We wish to thank Wayde Martens for his expert assistance with preparation of the pillared clays, and members of the clay group in the Inorganic Materials Research
Program for helpful advice and discussion. We acknowledge financial support from the
Inorganic Materials Research Program, Faculty of Science, and Queensland University of
Technology.
140 4.7 References
Balek, V., Malek, Z. and Klosova, E., 1998. Emanation thermal analysis of intercalated montmorillonitic clay. J. Therm. Anal. Calorim., 53(2): 625-631. Bergaoui, L., Lambert, J.-F., Franck, R., Suquet, H. and Robert, J.-L., 1995. Al-pillared saponites. Part 3. Effect of parent clay layer charge on the intercalation-pillaring mechanism and structural properties. J. Chem. Soc., Faraday Trans., 91(14): 2229-39. Bergaya, F., 2006 *. Handbook of Clay Science. Duong, L., Bostrom, T., Kloprogge, T. and Frost, R., 2005a. The distribution of Ga in Ga-pillared montmorillonites: A transmission electron microscopy and microanalysis study. Microporous and Mesoporous Materials, 82(1-2): 165-172. Duong, L., Wood, B. and Kloprogge, T., 2005b. XPS study of basic aluminium sulphate and basic aluminium nitrate. Materials Letters . 59(14/15): 1932-1936. Johansson, G., 1960. On the crystal structures of some basic aluminium salts. Acta Chem. Scand., 14: 771. Jones, J.R. and Purnell, J.H., 1993. Synthesis and characterisation of alumina pillared Texas montmorillonite and determination of the effective Keggin ion charge. Catalysis Letters, 18: 137-140. Kloprogge, J.T., 1998. Synthesis of smectites and porous pillared clay catalysts: a review. J. Porous Mater., 5(1): 5-41. Kloprogge, J.T., Booy, E., Jansen, J.B.H. and Geus, J.W., 1994. The effect of thermal treatment on the properties of hydroxy-Al and hydroxy-Ga pillared montmorillonite and beidellite. Clay Minerals., 29(2): 153-67. Kloprogge, J.T., Duong, L.V., Frost, R.L. and Wood, B.J., 2008. Baseline studies of the Clay Minerals Society Source Clays: X-ray photoelectron spectroscopy. Clays and clay minerals. Kloprogge, J.T., Evans, R., Hickey, L. and Frost, R., 2002. Characterisation and Al- pillaring of smectites from Miles, Queensland (Australia). Applied Clay Sciencce, 20: 157- 163.
141 Kloprogge, J.T., Fry, R. and Frost, R.L., 1999. An infrared emission spectroscopic study of the thermal transformation mechanisms in Al-13-pillared clay catalysts with and without tetrahedral substitutions. Journal of Catalysis, 184(1): 157-171. Malla, P.B. and Komarneni, S., 1993. Properties and characterization of alumina and silica-titania pillared saponite. Clays Clay Miner., 41(4): 472-83. Plee, D., Borg, F., Gatineau, L. and Fripiat, J.J., 1985. High-resolution solid-state aluminum-27 and silicon-29 nuclear magnetic resonance study of pillared clays. J. Am. Chem. Soc., 107(8): 2362-9. Purnell, J.H., 1990. Pillared clays-retrospect, prospect and action. Pillared Layered Struct.: Curr. Trends Appl., [Proc. Workshop]: 107-13. Siegbahn, K. et al., 1967. Electron Spectroscopy for Chemical Analysis. Atomic, Molecular and Solid State Structure Studies by Means of Electron Spectroscopy. Almquist and Wiksells, Uppsala.
142
Statement of Contribution
The authors listed below have certified* that:
1. they meet the criteria for authorship in that they have participated in the conception, execution, or interpretation, of at least that part of the publication in their field of expertise; 2. they take public responsibility for their part of the publication, except for the responsible author who accepts overall responsibility for the publication; 3. there are no other authors of the publication according to these criteria; 4. potential conflicts of interest have been disclosed to (a) granting bodies, (b) the editor or publisher of journals or other publications, and (c) the head of the responsible academic unit, and
5. they agree to the use of the publication in the student’s thesis and its publication on the Australasian Digital Thesis database consistent with any limitations set by publisher requirements. In the case of this chapter: XPS STUDY OF BASIC ALUMINUM SULPHATE AND BASIC ALUMINIUM NITRATE
Loc V. Duong, Barry J. Wood and J.Theo Kloprogge* Published in Materials Letters. 2005, 59, 1932-1936
Contributor Statement of contribution*
Loc Duong wrote the manuscript, designed experiments, conducted experiments, data analysis, data interpretation
Date
aided with data analysis and interpretation Barry Wood
Theo Kloprogge Supervision, editing
Principal Supervisor Confirmation
I have sighted email or other correspondence from all Co-authors confirming their certifying authorship and. ______Name Signature Date CHAPTER 5
5.0 XPS STUDY OF BASIC ALUMINUM SULPHATE AND BASIC ALUMINIUM NITRATE
Loc V. Duong, Barry J. Wood and J.Theo Kloprogge*
Published in Materials Letters. 2005, 59, 1932-1936
Loc Duong, Inorganic Materials Research Program, School of Physical and Chemical
Sciences, Queensland University of Technology, GPO Box 2434,
Brisbane, Qld 4001, Australia
Barry J. Wood, Brisbane Surface Analysis Facility, University of Queensland, Brisbane,
Qld 4072, Australia
Theo Kloprogge, Inorganic Materials Research Program, School of Physical and
Chemical Sciences, Queensland University of Technology, GPO Box
2434, Brisbane, Qld 4001, Australia
144
5.1 Abstract
Basic aluminium sulphate and nitrate crystals were prepared by forced hydrolysis of aluminium salt solution followed by precipitation with a sulphate solution or by evaporation for the basic aluminium nitrate. X-ray Photoelectron Spectroscopy (XPS) confirms the chemical composition determined by ICP-AES in earlier work. High resolution XPS scans of the individual elements allows the identification of both the
IV central AlO4 group and the twelve aluminium octahedra in the
IV VI [ AlO4Al (OH)24(H2O)12] building unit by two Al 2p transitions with binding energies of
73.7 and 74.2 eV in both the basic aluminium sulphate and nitrate. Four different types of oxygen atoms were identified in the basic aluminium sulphate associated with the central
AlO4, OH, H2O and SO4 groups in the crystal structure with transitions at 529.4, 530.1,
530.7 and 531.8 eV, respectively.
Keywords: Al13, basic aluminium nitrate, basic aluminium sulphate, Characterisation methods, Keggin structure, X-ray techniques
145 5.2 Introduction
One of the most important aluminium complexes in solution is the so-called Al13, a
IV Keggin-type cage structure in which a central Al O4 is surrounded by 12
VI 3+ Al (OH)24(H2O)12. Forced hydrolysis of Al solutions by the addition of a base like sodium carbonate or sodium hydroxide or homogeneous hydrolysis by the decomposition of urea is known to result in the formation of this large aluminium (oxo)hydroxide complex. The structure of this complex was first studied by X-ray diffraction after precipitation in the form of two different basic aluminium sulphates in which the Al13 structure is retained [1-4].
Johansson and coworkers [1-4] described the precipitation of basic aluminium sulphates containing the Al13 building unit linked by hydrogen bonding to the oxygen atoms of the sulphate groups. The sodium-containing aluminium sulphate crystallised in the cubic system whereas the sodium-free sulphate crystallised in the monoclinic system.
In earlier work we [5, 6] reported the precipitation of monoclinic basic aluminium sulphate with a small amount of sodium and a trace of nitrate and of basic aluminium nitrate. Based on the chemical analyses by ICP-AES a chemical composition per unit cell
27 of Na0.1[AlO4Al12(OH)24(H2O)12](SO4)3.55.9H2O was reported for the sulphate. Al
Solid-state Magic-Angle Spinning Nuclear Magnetic Resonance spectroscopy showed that the Al13 units were still present in the crystal structure.
Kloprogge et al. [5, 7] and Teagarden et al. [8] have described the infrared spectrum and near infrared spectrum [9] of the monoclinic basic aluminium sulphate. The spectrum of basic aluminium sulphate is dominated by two strong water bands at 3247 and 1640
146 cm-1, a strong Al-OH stretching band at 3440 cm-1. The infrared spectrum was further
-1 dominated by the ν1 and ν3 bands at 981 and 1051 cm of the sulphate group in the Al13 sulphate structure. Furthermore the band at 724 cm-1 is assigned to an Al-O mode of the polymerised Al-O-Al bonds in the Al13 Keggin structure [10]. The Raman spectrum
2- -1 -1 showed the ν2 and ν4 SO4 triplets at 446, 459 and 496 cm and 572, 614 and 630 cm .
-1 The ν1 was observed as a single band at 990 cm , partly overlapped by the ν3 triplet at
979, 1009 and 1053 cm-1 [11, 12].
To date there are no publications available on the X-ray Photoelectron Spectroscopy
(XPS) of the basic aluminium sulphate and nitrate complexes. The objective of this report is to describe in detail the high-resolution XPS spectra of basic aluminium sulphate and compare those with the spectra of basic aluminium nitrate order to get more insight in the structure of this complex aluminium salt. As such this paper forms a continuation of our earlier work on infrared, infrared emission and Raman spectroscopy of these basic aluminium salts.
5.3 Experimental
5.3.1 Basic aluminium sulphate and nitrate
The synthesis and characterisation of the monoclinic basic aluminium sulphate used in this study has been extensively described by Kloprogge et al. [5-7, 10, 13]. The tridecameric aluminium polymer was obtained by forced hydrolysis of a 0.5 M aluminium nitrate solution with a 0.5 M sodium or potassium hydroxide solution until an
OH/Al molar ratio of 2.2 was reached. Next, the basic aluminium sulphate was precipitated by the addition of the appropriate amount of 0.5 M sodium sulphate and aged
147 for 42 days before removal from the solution. Crystals collected from the wall of the container were shown by XRD and SEM to be phase pure (Fig. 5.1). The basic aluminium nitrate was prepared from the same hydrolysed aluminium solution followed by very slow evaporation of the excess water at room temperature. This sample was shown to have an impurity in the form of KOH.
5.3.2 XPS analysis
The basic aluminium nitrate and sulphate samples were analyzed in freshly powdered form in order to prevent surface oxidation changes. Prior to the analysis the samples were out gassed under vacuum for 72 hours. The XPS analyses were performed on a Kratos
AXIS Ultra with a monochromatic Al X-ray source at 150 W. Each analysis started with a survey scan from 0 to 1200 eV with a dwell time of 100 milliseconds, pass energy of
160 eV at steps of 1 eV with 1 sweep. For the high resolution analysis the number of sweeps was increased, the pass energy was lowered to 20 eV at steps of 100 meV and the dwell time was changed to 250 milliseconds.
Band component analysis was undertaken using the Jandel ‘Peakfit’ software package, which enabled the type of fitting function to be selected and allows specific parameters to be fixed or varied accordingly. Band fitting was done using a Lorentz-Gauss cross- product function with the minimum number of component bands used for the fitting process. The Gaussian-Lorentzian ratio was maintained at values greater than 0.7 and fitting was undertaken until reproducible results were obtained with squared correlations of r2 greater than 0.995. Band width and shape of the O and Al transitions were set prior
148 to the band component analysis, based on the analysis of standard aluminium compounds corundum, α-Al2O3 and boehmite, AlOOH.
5.4 Results and discussion
The basic aluminium sulphate crystals exhibit a clear tetrahedral morphology. Some surface cracks are visible probably due to partial dehydration (Fig.5.1). Calcination of the basic aluminium sulphate does not significantly alter the morphology, although some minor powder formation can be observed on the surface. Energy dispersive X-ray analyses showed no change in the overall composition.
Fig. 5.2 shows the XPS survey scan of the basic aluminium sulphate, clearly showing the presence of sodium, aluminium, oxygen and sulphur. In addition there is a minor amount of N present as well. The chemical composition based on the survey scans of the basic aluminium sulphates at room temperature and after calcination at 200 and 400°C do not show any significant differences (Table 5.1).
25°C 200°C 400°C Al13 nitrate O 1s 72.2 71.2 68.7 64.4 Al 2p 18.3 18.9 21.9 14.3 S 2p 4.9 7.2 3.4 - N 1s 2.3 - - 11.2 Na 1s 2.4 2.7 6.1 -
Table 5.1. Chemical composition (in atom%) from the XPS analyses of the basic aluminium sulphate at room temperature and after calcination at 200 and 400°C and basic aluminium nitrate
149 The composition is very close to the composition reported earlier based on ICP-AES analysis of this compound, although the sodium content is somewhat higher [5]. This can probably be explained by the fact that ICP-AES of the redissolved basic aluminium sulphate crystals is less sensitive for sodium than XPS of the solid crystals. The basic aluminium nitrate shows a similar composition to that of the basic aluminium sulphate but is characterised by slightly lower oxygen content, due to the fact that the SO4 groups have been replaced by NO3 groups in the crystal lattice and the presence of the KOH impurity.
The high resolution scans of the different elements present in the basic aluminium sulphate crystals at room temperature and after calcination at 200 and 400°C are similar in both intensities and binding energies within the experimental error of the instrument
(Table 5.2), confirming the general observations discussed above. For sodium a single 1s transition is observed with a binding energy of 1073 eV, indicating a single position in the basic aluminium sulphate crystal, which is in accordance with the crystal structure described by Johansson [1-4]. Similarly the sulphur is only present as a single type of sulphate in the structure as shown by the presence of only one S 2p ½ and one S 2p 3/2 transitions (Fig 5.2).
150
Fig 5.1 SEM images of a basic aluminium sulphate crystal at room temperature and after calcination at 400°C
151
25°C 200°C 400°C
5000
4500 O 1s
4000
3500
3000
2500
2000 C 1s
Intensity (CPS) Na 1s S 2p 1500 N 1s Al 2p 1000
500
0 1200 1000 800 600 400 200 0 Binding energy (eV)
Fig. 5.2 XPS survey scan of basic aluminum sulphate
152 Al 2p VIAl 74.2 74.2 74.6 Al 2p IVAl 73.7 73.5 73.8 S 2p 1/2 171.0 171.3 171.7 S 2p 3/2 169.9 170.1 170.5 O 1s SO4 531.8 531.7 531.9 O 1s H2O 530.7 531.0 531.1 O 1s OH 530.1 530.3 530.4 O 1s O 529.4 529.5 529.7 Na 1s 1072.9 1073.1 1073.3
Table 5.2. Binding energies (in eV) of the basic sulphate at room temperature and
after calcination at 200 and 400°C
Much more informative though are the high resolution scans of aluminium and oxygen
(Fig 5.2). The aluminium high resolution scans show two overlapping bands associated
with two different Al 2p transitions with binding energies of 74.2 and 73.7 eV at room
temperature. The ratio of the two types of aluminium is in the order of 11:1 which is very
IV close to the 1:12 ratio observed in the Al13 complex where a central Al is surrounded by
twelve VIAl. Therefore the 74.2 eV transition is interpreted as being due to the twelve
aluminium octahedra in the Keggin structure, while the 73.7 eV transition is associated
with the central AlO4 tetrahedron. Similar values are observed for the Al 2p transitions in
the basic aluminium nitrate.
The oxygen high resolution scans are rather complex and contain a number of
overlapping transitions. Band component analysis indicate the presence of four transitions
at 531.6, 530.8, 530.1 and 529.5 eV associated with roughly 8, 40, 36 and 15 percent of
the total amount of oxygen present in the crystal structure. Related work in our laboratory
153 on aluminium (oxo-)hydroxides and oxides such as gibbsite, boehmite and corundum have shown that in general there is a clear distinction between oxygen atoms, hydroxyl groups and water with a shift in the binding energy towards higher values. Oxygen in these minerals is generally observed around 530.7 eV, which is close to the value of
529.4 eV in the basic aluminium sulphate. Similarly, hydroxyl groups in gibbsite and boehmite are identified by an oxygen 1s transition around 531.8 eV and water around 533 eV. Following an analogues interpretation of the oxygen transitions in the basic aluminium sulphate the four transitions are interpreted as being due to the central O around the IVAl, hydroxyl group around the VIAl, water in both around the VIAl and as crystal water and finally as oxygen in the sulphate groups. This last transition with a transition at 531.6 eV is similar to the value of 532.4 for oxygen in Al2(SO4)3 [14], 532 eV for CaSO4 [15] or 531.7 eV for CdSO4 [16]. Further evidence for this interpretation stems from the relative ratios of these transitions. Based on the chemical composition of the basic aluminium sulphate one would expect an atom concentration of about 7% for oxygen, 38% for the oxygen atoms in the hydroxyl groups, 34% for oxygen atoms in water molecules and 22% for oxygen atoms in the sulphate groups. These values are close to the observed values. No details can be obtained from the O 1s high resolution spectrum of the basic aluminium nitrate due to interference of the KOH impurity.
In summary this work has shown that XPS is not only a strong tool for obtaining chemical information of materials, but high resolution XPS will also give detailed information on the local environment of the different atoms in the structure. As such this technique forms an important additional analytical tool to the more common techniques
154 such as X-ray Diffraction, Infrared spectroscopy, Raman spectroscopy and solid-state
Nuclear Magnetic Resonance Spectroscopy.
400
350
300 VIAl 250
200 IVAl 150 Intensity (CPS)
100
50
0 78 77 76 75 74 73 72 71 70 Binding energy (eV)
Fig 5.3a Al 2p high resolution spectrum of basic aluminium sulphate
155 3000
2500 OH
2000 H2O
1500
SO4
Intensity (CPS) 1000 O
500
0 536 535 534 533 532 531 530 529 Binding energy (eV)
Fig.5.3b O 1s high resolution spectrum of basic aluminium sulphate
350
300
250 S 2p 3/2 S 2p 1/2 200
150 Intensity (CPS) 100
50
0 173 172 171 170 169 168 167 166 Binding energy (eV)
Fig 5. 3c S 2p high resolution spectrum of basic aluminium sulphate
156 5.5 Acknowledgements
The authors wish to thank the Inorganic Materials Research Program, Queensland
University of Technology, Brisbane, for the financial support
157 5.6 References
1. G. Johansson, Ark. Kemi. 20 (1963) 321.
2. G. Johansson, Acta Chem. Scand. 14 (1960) 771.
3. G. Johansson, G. Lundgren, L. G. Sillén and R. Söderquist, Acta Chem. Scand. 14
(1960) 769.
4. G. Johansson, Acta Chem. Scand. 16 (1962) 403.
5. J. T. Kloprogge, J. W. Geus, J. B. H. Jansen and D. Seykens, Thermochim. Acta
209 (1992) 265.
6. J. T. Kloprogge, P. J. Dirkin, J. B. H. Jansen and J. W. Geus, J. Non-Cryst. Solids
181 (1994) 151.
7. J. T. Kloprogge and R. L. Frost, Thermochim. Acta 320 (1998) 245.
8. D. L. Teagarden, J. F. Kozlowski and J. L. White, J. Pharm. Sci. 70 (1981) 758.
9. J. T. Kloprogge, H. Ruan and R. L. Frost, J. Mater. Sci. 36 (2001) 603.
10. J. T. Kloprogge, R. L. Frost and R. W. P. Fry, in 16th International Conference on
Raman Spectroscopy, Cape Town, South Africa, 1998, pp. 700.
11. J. T. Kloprogge and R. L. Frost, J. Mater. Sci. 34 (1999) 4367.
12. J. T. Kloprogge and R. L. Frost, J. Mater. Sci. 34 (1999) 4199.
13. J. T. Kloprogge, D. Seykens, J. W. Geus and J. B. H. Jansen, J. Non-Cryst. Solids
142 (1992) 94.
14. K. Arata and M. Hino, Appl. Catal. 59 (1990) 197.
15. A. B. Christie, J. Lee, I. Sutherland and J. M. Walls, Appl. Surf. Sci. 15 (1983)
224.
158 16. J. Riga, J. J. Verbist, P. Josseaux and A. K. Mesmaeker, Surf. Interface Anal. 7
(1985) 163.
159 CHAPTER 6
6.0 SUMMARY AND SUGGESTIONS FOR FUTURE WORK
6.1 Summary
This chapter summarises the results of this study presented in chapter two to chapter five. A more complete picture of Al- and Ga-pillared clays has been obtained by studying the preparation of the Al13 and Ga13 complexes and their intercalation in montmorillonite, including a new method applying ultrasonic treatment. The relationship of the Al and Ga pillars with the silicate structure of the pillared montmorillonite has been studied in detail using electron microscopy, X-ray microanalysis, and X-ray photoelectron spectroscopy
(XPS).
Pillared clays have been developed since 1955 with the hope to produce a new type of catalyst with relatively large pore sizes (in comparison to zeolites) to use for hydrocarbon cracking. The Al13 Keggin or cage complexes have since the early 1970s been the most popular pillars to use in producing pillared clays. Johansson and coworkers described the precipitation of basic aluminium sulphates containing the Al13 building unit linked by hydrogen bonding to the oxygen atoms of the sulphate groups (Johansson, 1960). The basic sulphate and aluminum nitrate from these complexes have been studied by NMR by a number of authors (see under section review of pillared clays).
160 The relationship between the Al and Ga pillars and the silicate structure was discussed in chapter 2. Detailed structures of the pillared Al and Ga montmorillonite (SWy-2 and
Miles) have been studied by TEM. Ga has similar chemical properties as Al but gives excellent stable pillared clays under the electron beam. The author has developed a method for the preparation of cross sections of clay samples perpendicular to the layers for TEM studies, which provided very good results. (Duong et al., 2005a). This method uses Spurr resin (Spurr, 1969) for embedding clay samples, which are then cut in the correct orientation with a diamond knife. These sections are very stable under the electron beam in TEM. Formulae of the Ga13 and Al13 pillared montmorillonites calculated from EDS results show that the atomic fractions of the total intercalated species to silicon were 0.273, 0.235 and 0.235 for the Al13, Al12Ga and Ga13 pillared clays respectively. It was shown that 0.89 Ga atoms are present per formula unit, which indicates that there are 20 silicate rings consisting of 6 tetrahedral each per Ga13 pillar.
The average distance between the pillars has been calculated to be 44 Å in agreement with pore size measurements.
The relationships between the Ga, Al and GaAl pillars and the silicate layers in cross section indicated that all the pillars intercalated very homogeneously at relatively low magnification as evidenced by an X-ray map and EDS results. However, at high magnification differences in pillar density and basal spacing was observed.
An improved route for the preparation of pillared montmorillonite by using ultrasonic treatment was described in chapter 3. In comparison to the commonly applied exchange times in the order of hours, short ultrasonic treatment of 5-10 minutes produced very good results, even without prior Na exchange. Atomic force microscopy showed no
161 evidence of any pillars on the outside of the clay particles. Furthermore, there were indications that the distribution of the Al 13 pillars may not be as homogeneous as is often thought, in agreement with the TEM results of the previous chapter. The new method of preparing intercalated clays by ultrasonic treatment has proved to be successful in providing pillared clays with similar properties to the conventional method without loosing catalytic activity. The short ultrasonic treatment time without the process of extensive washing and sodium exchange make this method very attractive for up-scaling from the laboratory to large commercial scale production.
The internal structure of pillared clay has been further examined in chapter 4 by XPS.
Chemical analysis of Wyoming montmorillonite SWy-2 and its pillared equivalents shows that 2.01 Ga and 2.76 Al is present in the clay interlayer per formula unit of
Al4(Si7.38Al0.62)O20(OH)4 (Duong et al., 2005b). The calculation from starting montmorillonite layer charge showed that the charge of the Ga13 complex has been reduced from 7+ in solution to around 3 to 4+ for the complex intercalated in the montmorillonite. The excess water in pillared montmorillonite represented as a band around 534.38 eV present in the sample after calcinations process. The amount of Ga observed in the XPS is significantly higher than in the TEM analyses. This is a reflection of the inhomogeneous distribution of the Ga in the clay sample. In the TEM single particles were analysed while in XPS a significantly larger volume of sample was irradiated with X-rays. The higher Ga content can either be due to a higher but inhomogeneous pillar density in the pillared clay or by the presence of gallium oxide particles in between the clay particles. Based on the TEM observations of differences in
162 basal spacings and the absence of clear oxide particles the first hypothesis seems to be the most acceptable.
In chapter 5 of this study the binding energies of Al, O and S in these complexes was examined for the first time by XPS. The chemical composition of the Keggin structure from the work of Kloprogge using ICP-AES has been confirmed. The high resolution Al
2p scans showed both Al IV at 73.7 eV and AlVI at 74.2 eV present in the structure with the ratios of 1:12 which is identical to the ideal structure of the Al13 with one central Al tetrahedron surrounded by 12 Al octahedra. The high resolution scans of the O 1s identified 4 types of oxygen: AlO4 associated with the four oxygen atoms in the central tetrahedron, OH and H2O in the 12 octahedra and SO4. Calcination of the basic aluminum at 200 and 400 0C showed no change in the high resolution scans.
6.2 Suggestion for future work
Although this study has added some useful information about the relationship between the Al13 pillars and the montmorillonite structure, more work is needed to determine the exact bonding between these pillars and the montmorillonite tetrahedral sheets. In addition, further study is necessary with respect to different starting materials, in particular the effects of different layer charges and the origin of the layer charge
(octahedral versus tetrahedral) on the pillaring mechanisms. Also the exact structure of the pillar itself after calcinations is still unknown, although some information has been obtained from vibrational spectroscopy, solid state NMR and XPS. Heating stage XPS will be most useful to study the calcination processes and the behavior of the hydroxyl groups in the pillars in situ, in particular when this is combined with other thermal
163 techniques such as infrared emission spectroscopy (IES) and thermogravimetric and differential scanning calorimetry (TG and DSC).
Finally, since most of the work so far has focused on Al, other metal pillared clays deserve further attention. In future studies the role of Atomic force microscope, XPS, neutron diffraction and Synchrotron X-ray radiation techniques can play an important part in providing further information about the pillared clay structures.
Synchrotron techniques are currently being developed in Australia. This will allow further study using Neutron diffraction to further elucidate the pillar structure, while
EXAFS can provide detailed chemical and structural information about a specific absorbing element whether it is a major component of a solid phase or trace component of the bulk phase, a soluble species. The EXAFS can be used to identify the charges of the local element in the silicate structure.
164 6.3 References
Duong, L., Bostrom, T., Kloprogge, T. and Frost, R., 2005a. The distribution of Ga in Ga-pillared montmorillonites: A transmission electron microscopy and microanalysis study. Microporous and Mesoporous Materials, 82(1-2): 165-172. Duong, L., Kloprogge, T., Frost, R. and Wood, B., 2005b. The structure of Al13 and Ga13 pillars in pillared montmorillonites., The 13th International Clay Conference. Oral presentation, Waseda University Tokyo Japan. Johansson, G., 1960. On the crystal structures of some basic aluminium salts. Acta Chem. Scand., 14: 771. Spurr, A.R., 1969. Journal of Ultrastructural Research, 26: 31-43.
165 CHAPTER 7
7.0 ADDITIONAL SUPPORT PAPERS
7.1 Review of the synthesis and Characterization of
pillared clays and related porous materials for
cracking of vegetable oil to produce biofuels
Kloprogge, J.T., Duong, L.V. and Frost, R.L. (2005) Published in. Environmental Geology, 47(7), 967-981.
7.2 X- ray photoelectron spectroscopic study of the
major minerals in Bauxite: Gibbsite, Bayerite and
(pseudo-) boehmite
J. Theo Kloprogge1, Loc V. Duong1, Barry J. Wood2, and Ray L. Frost1 Published in Journal of Colloid and Interface Science 296 (2) : pp 572- 576
7.3 A X-ray photoelectron spectroscopy study of
HDTMAB distribution within organoclays
Frost, Ray.L and He, Hongping and Zhou, Qin and Kloprogge, J.Theo, Duong, Loc and Wood, Barry. (2007) Published in Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy
166 A review of the synthesis and characterisation of pillared clays and related porous materials for cracking of
vegetable oils to produce biofuels
J.Theo Kloprogge*, Loc V. Duong and Ray L. Frost
Inorganic Materials Research Program, School of Physical and Chemical Sciences, Queensland University of Technology, GPO Box 2434, Brisbane, Qld 4001, Australia
Corresponding author: phone +61 7 3864 2184, fax +61 7 3864 1804, E- mail [email protected]
167
Abstract This paper presents an overview of the modification of clay minerals by propping apart the clay layers with an inorganic complex. This expanded material is converted into a permanent two-dimensional structure, known as pillared clay or shortly PILC, by thermal treatment. The resulting material exhibits a two-dimensional porous structure with acidic properties comparable to that of zeolites. Synthetic as well as natural smectites serve as precursors for the synthesis of Al, Zr, Ti, Fe, Cr, Ga, V, Si and other pillared clays as well as mixed Fe/Al, Ga/Al, Si/Al, Zr/Al and other mixed metal pillared clays. Biofuels form an interesting renewable energy source, where these porous catalytically active materials can play an important role in the conversion of vegetable oils, such as canola oil, into biodiesel. Transesterification of vegetable oil is currently the method of choice for conversion to biofuel. The second part of this review focuses on the catalysts and cracking reaction conditions used for the production of biofuel. A distinction has been made in three different vegetable oils as starting materials: canola oil, palm oil and sunflower oil.
Key words: Canola oil, Cracking reaction, Pillared clay, PILC
168
Introduction
Escalating crude oil prices and environmental awareness have increased interest in the use of renewable fuel sources. One area of attention is the upgrading of vegetable oils for use as a fuel or fuel additive. Besides being a renewable source, the use of vegetable oils has benefits economically and environmentally. Such oils are CO2 neutral and contain little, if any sulfur, nitrogen and metals, which are major pollutants in current fuel emissions (Katikaneni and others 1998). The possibility also exists for the reuse of current vegetable oil wastes such as wastes from fast food restaurants. As such, oils come from plants that can be easily grown; its production can be localized and adjusted according to demand. The conversion of the oil to fuel can therefore bring benefits to the community economically as well as making them no longer reliant on outside sources. Over the years, vegetable oils have been substituted for diesel for use in engines but this has led to problems such as carbon deposits, oil ring sticking and gelling of the lubricating oil (Ma and Hanna 1999). Because of such problems, research in this area has been centered on the conversion of these oils to a form that is similar to current fuels. One such fuel, which is currently gaining much attention, is biodiesel. This is a variety of ester-based oxygenated fuels made from vegetable oils or animal fats. There are several methods for the conversion of vegetable oils to biodiesel of which the most common is the transesterification process, in which an alcohol is reacted with the oil to form esters and glycerol (Ma and Hanna 1999). The esters are separated and commonly used as a mixture with petroleum diesel (20:80) to minimize engine modification requirements. Altin and others (2001), showed that vegetable oil methyl esters gave performance and emission characteristics close to petroleum diesel. The main problems associated with the increased use of this fuel are the costs of the oil and its processing. Also, the marketing of this product is limited to diesel engine applications. Another method for the conversion of vegetable oils to a useable fuel product is by catalytic cracking reactions. This is currently used in the petroleum and
169 petrochemical industry to convert high molecular weight oil components to lower molecular weight ones which can be used directly or blended for use as fuel (Thomas and Thomas 1997). The reaction process usually involves the mixing of catalysts with the oil feed at high temperature in a fluid catalytic cracker (FCC) unit. Here the hydrocarbon product is collected and the spent catalyst is directed to a regenerator
which oxidizes the coke that has collected on it, to CO, CO2 and H2O, to then be reused (Venuto and Habib 1979). Acid-treated clay of the montmorillonite type was the catalyst commonly used for initial cracking reactions in the 1930s (Thomas and Thomas 1997). Such catalysts were replaced after World War II with a more stable synthetic silica-alumina type which also gave better product distribution (Thomas and Thomas 1997). The emergence of zeolites in the 1960s revolutionized the process mainly because of their high activity, selectivity and resistance to collapse when treated at high temperatures (Venuto and Habib 1979). Their use is commonplace now with ZSM-5 and Y types being some of the most popular catalysts. The interest now is in producing a catalyst with a larger pore size compared to zeolite (∼8 Å) to handle the cracking of heavier crude oil. The use of pillared clays has received considerable attention because of their ability to achieve large pore sizes, but factors such as thermal stability and coking properties still need to be overcome. It is common knowledge that vegetable oils can be cracked into lighter fuel fractions by the use of such catalysts. There are, however, a number of problems associated with this process with cost being one of the major ones. Altin and others (2001) noted that at present vegetable oils are more expensive than diesel fuels. However he suggests that with an increase in consumption should come an increase in production and this would lead to more mechanised farming methods which would probably translate into a decrease in cost. Another problem with the use of vegetable oils for conversion to fuels is that the composition of such oils varies drastically between types. This means that a particular set of reaction conditions and catalyst type will give different products according to the starting oil.
170 Smectites
Clays are phyllosilicates or layer silicates with a layer lattice structure in which two- dimensional oxoanions are separated by layers of hydrated cations. The oxygen atoms define upper and lower sheets enclosing tetrahedral sites and a central sheet having the brucite or gibbsite structure enclosing octahedral sites. Smectites have two tetrahedral sheets around the central octahedral sheet in each layer, hence the name 2:1 phyllosilicate. These layers have a positive charge deficiency resulting from isomorphous substitutions (e.g Si4+ by Al3+ at tetrahedral sites or Al3+ by Mg2+ at octahedral sites). These negative layer charges are balanced by exchangeable hydrated interlayer cations such as Na+, K+ or Ca2+. The charge deficiency and the origin of this deficiency (octahedral vs tetrahedral) result in different physical and chemical properties, such as, thermal stability and swelling behaviour. Layer charges related to tetrahedral substitutions lead to a localised charge distribution, while layer charges related to octahedral substations are more distributed over the complete oxygen framework.
Pillared Interlayered Clays (PILCs)
As a consequence of increasing oil prices, PILCs were improved in the mid-1970s to optimise the catalytic cracking of crude oil. To increase the yield of lighter fractions from heavy crude oil, catalysts were required that had larger pore size and good thermal and hydrothermal stability (Ding and others 2001; Frost and others 1998; Kloprogge 1998). This was research focussed on the use of inorganic hydrated polyoxocations as pillaring agents. Such pillaring agents, when calcined, dehydrate and dehydroxylate to form a fixed metal oxide pillar with a high thermal stability and high surface area. The use of the Al13 polyoxocation was favoured as it had been extensively researched and reported previously and was easily prepared (Gil and Gandía 2000). The first step in the pillaring process is to prepare a pillaring agent. In the case of the
Al13 polyoxocation, two methods are commonly used: 1) mixing of aqueous AlCl3
171 with Al to form a chlorohydrate which is also commercially available, and 2) addition 3+ of a base to AlCl3 or Al(NO3)3 solutions with OH/Al ratios up to 2.5. The polyoxocation complex produced has been analysed and is thought to be the 7+ tridecamer [AlO4Al12(OH)24(H2O)12] , also referred to as the Keggin ion (Kloprogge 1998; Kloprogge and others 1992). The next step is the mixing of a clay suspension with this polyoxocation solution. This allows the interlayer cations in the clay to exchange with the polyoxocation in solution through cation exchange reaction or intercalation (Gil and Gandía 2000). After the intercalation process is complete, the clay is separated, washed and then calcined. The property of the stable pillared structure obtained is greatly affected by factors such as clay used, mixing and drying conditions and polycation/s used.
Vegetable oil
Vegetable oils are predominantly made up of triacylglycerols with a small amount of minor compounds (2-5%) (Cert and others 2000). Triacylglycerols are made up of one glycerol molecule joined to 3 fatty acids by an ester link. As shown in Figure 1, the type and concentration of fatty acid varies considerably from one vegetable oil to another. Hence, it is important to be aware of the composition of the vegetable oil used for the choice of the catalyst, as it will determine the type of reactions that are probable. The four major vegetable oils produced today are palm oil, soybean oil, sunflower oil and rapeseed oil. The production of palm oil has increased at a great rate over the past 5 years. One report estimates that the world trading of palm oil has grown 32% since 1997-98 (Anonymous 2001). This has led to a decrease in prices with the Malaysian average palm oil prices in 1999-00 falling to $314 per metric tonne a decrease of around 35% from 1998-99 (Guzman 2001). This reduction in price and high availability makes it ideal as a fuel source. Oil is extracted from the fruit of the oil palm which is usually grown in areas within 10o of the equator (Gunstone and Society of Chemical Industry (Great Britain) 1987). It produces two types of oil, one from the flesh (palm oil) and another from the kernel (palm kernel oil) of the fruit. These oils can then be separated into a high-melting
172 fraction (stearin) and a low-melting fraction (olein) (Gunstone 1996). The yield is about 4-5 tonnes/hectare of palm oil and 0.5 tonnes/hectare of palm kernel oil. The main use of soybean is the protein rich meal obtained after extraction of the oil. This is used for animal feed and makes up between 50-70% of its value (Gunstone 1996). Soybean oil prices are currently at the lowest they have been since 1986-87 (Guzman 2001). It is mainly produced in USA, China, Brazil and Argentina (Gunstone 1996). Soybean oil is highly unsaturated with linoleic and linolenic acid comprising over 60%. The remainder is predominantly oleic and palmitic acid. The sunflower plant is mainly grown in the former Soviet Union, the European Union, Argentina, China, USA, and Eastern Europe (Gunstone 1996). It too is at its lowest price since 1986-87 (Guzman 2001). The seed oil is also highly unsaturated, containing mostly linoleic and oleic acid. Rapeseed oil source has received the most attention with considerable research performed on the altering of the plants by breeding techniques and genetic modification. Such changes have been made to give the plant specific qualities such as tolerance to broad-spectrum herbicides (Friedt and Lühs 1998). The project is of interest for the possibility of modifying the fatty acid composition of the oil by using these techniques. Not only can the length of the fatty acid be altered, but also properties such as the degree of unsaturation, stereochemistry and position of double bonds (Friedt and Lühs 1998). It may be possible therefore to modify the plant to grow in a particular climate and produce oil that has the properties necessary for optimum cracking reactions. According to Friedt and Lühs (1998), rapeseed oil is ideal for non-food applications because of qualities such as relatively homogeneous composition, high degree of refinement, freedom from contaminants and also biodegradability. This is evidenced by the relatively large number of studies that have used canola oil, a modified rapeseed oil, to investigate cracking reactions. Unmodified rapeseed oil is high in erucic, oleic and linoleic acid. The composition varies greatly among plant variety with some oil products being high in the saturated lauric acid. It is mainly produced in northern Europe, China, India and Canada.
173 Australia has recently increased its production with estimates of about 300,000 tonne now achieved annually (Department of Natural Resources and Environment 2001). Although coconut oil is not produced on as large a scale as the oils mentioned previously, it would be of interest to study the cracking reactions associated with this oil as it is quite high in lauric acid. This makes it highly saturated oil and it would be expected that any major differences in reaction products could be largely attributed to this fact.
Pillared clays as cracking catalysts
As noted previously, a major problem associated with the use of pillared clays as catalysts has been their lack of thermal and hydrothermal stability above roughly 600 to 700°C depending on the pillar and clay used. Because of this, many studies have focussed on methods of improving such stability while also retaining, or increasing, its catalytic properties. Current research has examined how the introduction of various pillaring species alters the PILC properties (see e.g. review Kloprogge 1998). Another way of changing the property of PILCs has been to alter preparation techniques. This is usually examined after the pillaring species has been optimised so as to maximise the desired properties of the PILC. Each method has had various degrees of success and will be further discussed.
Various pillaring species This usually involves intercalation with one easily pillared cation, such as Al, and then the addition of another cation to give the PILC a specific property. Some cations that are used are Ce, Cr, Ga, La, Si, Ti, and Zr. Because there are a large number of cations available, extensive study has been carried out in this area and a few will be discussed here for an overview.
Gallium This cation has a number of chemical properties (eg. ionic radius) that are similar to Al3+, making it ideal for pillaring. Bradley and others (1990a) have shown that the
Ga13 and GaAl12 pillaring agents that can be formed by hydrolysis, are similar in
174 structure to the Al13 Keggin-ion species. The d001 spacing of the Ga13 species was
around 5.6% larger than Al13. This correlated well with their estimate of 5.7% for a Keggin-like structure. The pillaring solutions prepared were tested for thermal stability by allowing each to reflux until a precipitate formed. It was found that the
GaAl12 solution could be refluxed for over 3 weeks, over 5 times longer than the Al13 solution. Using NMR techniques, such stability was attributed to the overall increase in symmetry of the pillar owing to the better fit of the Ga3+ ion in the central position of the modified Keggin structure (Bradley and others 1990b). Further MAS NMR, IR and XRD studies confirmed that the GaAl12 structure is structurally analogous to the
Al13 species (Bradley and others 1992, 1993). Bradley and Kydd (1991) were also able to demonstrate how the thermal stability of
the GaAl12 PILC was markedly better than Al13 and Ga13. They found that the surface 2 area of the GaAl12 pillared clay dropped from 277 to 196 m /g over a temperature range of 200 to 700oC whereas the other two PILCs dropped to less than 115 m2/g. This correlated with the stability of the ions in solution. In a further article they examined the Brönsted acidic character of these pillared clays and concluded that the
GaAl12 pillared clay had the highest abundance (Bradley and Kydd 1993). González and others (1992) also produced thermally stable GaAl PILCs that retained 70% of its surface area when heated to 700oC. In a later study (Hernando and others 1996), they examined how modifying this PILC with cerium effected its thermal stability and catalytic properties. They determined that although the presence of Ce decreased the surface area of the PILC, it also increased the Brönsted acid site density making it more selective toward cracking reactions. González and others (1999), also studied the catalytic properties of GaAl PILCs with respect to the cracking of heavy oils. They produced thermally stable PILCs that retained around 85% of its surface area and micropore volume when heated to 700oC. As shown in Table 1, the GaAl PILC was the only one to exhibit a basal spacing by XRD at 700oC (17.3 Å). The results from their fixed-bed reactor at 482oC indicated that, although the GaAl PILC had the highest activity, it formed more gaseous product and had a higher coke formation than the Al PILCs. The gasoline product formed by the GaAl PILC did however have a higher octane number, which
175 somewhat compensates for lack of quantity. This higher octane number was attributed to the dehydrogenating effect of the Ga, which resulted in the production of more alkenes. Lenarda and others (1999) also produced PILCs using a solution containing an Al/Ga
ratio of 12/1. They further treated the PILC with NH3 vapours and repeated the pillaring process. The re-pillared clays had a similar XRD pattern to the standard PILC but had a decrease in surface area from 383 to 323 m2/g and an increase in pillar density of 30%. This was due to the re-pillaring process in which more pillars diffuse into the interlayer space and/or unpillared or partially pillared areas are completed (Fig. 2). They too found that the GaAl12 species retained its structure up to 700oC. Domínguez and others (1998) examined the hydrolysis of various mixed solutions
and concluded that for the Al-Ga system, the formation of GaAl12 species was more
probable in the pH range of 4 - 5, after which the Al13 species was more likely to form.
Lanthanum and Cerium Sterte (1991) assessed lanthanum as a rare earth cation that, "most readily formed a complex with aluminium suitable for pillaring". This author was able to produce LaAl pillared montmorillonite with basal spacings of around 26Å and surface areas of 300-500 m2/g. The pillaring solution was prepared by either refluxing mixtures of aluminium chlorohydrate and lanthanum chloride for to up to 120 hours or by treating the solution in an autoclave at 120o-160oC for 12-96 hours. He determined that a La/Al ratio of at least 1/5 was needed to produce large pore structures. From XRD analysis it was shown that between 72 and 96 hours of refluxing was required for maximum basal spacings while only 12 hours was needed for solutions autoclaved at 160oC. Booij and others (1996) also attained, by hydrothermal treatment, large pore LaAl and CeAl PILCs with basal spacings around 25Å and surface areas of about 430m2/g. They concluded that the La/Al or Ce/Al molar ratio can be as low as 1/30 as the formation of the initial polyoxocation is favoured by high Al concentrations. They
176 attributed the stability of the pillared structure to the higher Altetrahedral/Aloctahedral ratio compared to the normal Keggin structure. Valverde and others (2000) synthesized PILCs by mixing 0.05M solutions of
La(NO3)3 or Ce(NO3)3 with a bentonite clay suspension that had already been allowed to react with an Al polyoxocation solution. Basal spacings of 16 to 20Å were recorded with surface areas from 239 to 347 m2/g. Although these pillared clays were similar in structure and acid properties to Al PILCs, they showed an increase in thermal and hydrothermal stability. They concluded that the cations were not part of the pillar and that they delayed the dehydroxylation of the PILC with Ce being more effective than La. They suggested that this delaying effect was due to the cation either blocking the hexagonal cavities in the tetrahedral layer or somehow strengthening the bonding of the Al pillars. Various montmorillonites from different localities were pillared with Ce and Al by Pires and others (1998). They found that by changing the total metal concentration of the pillaring solution they could regulate the interlayer pore size. Basal spacings up to 22.2Å were recorded at 500oC with specific surface areas being as high as 300 m2/g at 700oC. They observed that the extent of octahedral substitution in the parent clay had a large influence on the thermal stability of the PILC produced.
Silicon By using two industrially produced silanes, 3-aminopropyltrimethoxysilane (APTMS) and 2-(2-trichlorosilylethyl)pyridine (TCSEP) as pillaring solutions, Fetter and others (1994) where able to produce pillared clays having good thermal stability. The PILC intercalated with TCSEP gave a number of interlayer spacings (up to 10Å), which were attributed to the range of polymeric species present in the solution. It had however, poor thermal stability with its structure almost collapsing at 700oC. The use of APTMS gave a more homogenous interlayer spacing of around 7Å and was stable up to 700oC. This PILC also maintained a large microporosity and some acidity at 700oC. A further study (Fetter and others 1995) used competitive ion exchange of Al- polyoxocations and TCSEP to produce a pillared clay exhibiting good thermal stability with an acidity comparable to that of HY zeolites. The SiAl PILC produced had a basal spacing of 17.4Å and a surface area of 278 m2/g when calcined at 600oC
177 (the temperature necessary to burn off the organic moiety). With further calcination at 700oC, the surface area only decreased to 244 m2/g with a basal spacing of 17.2Å. XRD patterns showed that the structure collapses at around 800oC. Sterte and Shabtai (1987) produced hydroxy-SiAl pillaring solutions by two methods:
(1) by mixing orthosilicic acid with AlCl3 and then treating with aqueous NaOH and
allowing to age and (2) by ageing an Al13 solution and then combining with orthosilicic acid. The SiAl pillared montmorillonite produced by the first method gave basal spacings around 19Å (table 3) and were not affected by the Si/Al ratio in the pillaring solution. The surface area did drop substantially as the Si/Al ratio increased. The same trend occurred for the PILCs prepared by the second method except the basal spacings were lower (∼17Å). The decrease in surface area was
attributed to the increase in substitution of -OH groups by the bulkier -OSi(OH)3 groups in the pillaring species. They also examined the thermal stability of the PILCs and noted a rapid decrease in surface area for Si/Al ratio of 0.53 as it was heated to 600oC whereas it was slower for Si/Al ratios of 1.04 and 2.08. Zhao and others (1992) also prepared SiAl PILCs by methods similar to Sterte and Shabtai (1987) and they found that the structure of the PILCs produced by both methods were similar but the SiO2 content of the PILC produced by method (2) was higher than that from method (1). This led them to likewise conclude that Si was incorporated into the Al pillars and 27Al-NMR was used to confirm that the pillar structure was similar to the Keggin structure in Al PILCs. Hence, they suggested two reactions occurring in solution:
7+ [Al13O4(OH)23(H2O)12] -OH + HO-[Si(OH)2]n-OH [1] 7+ [Al13O4(OH)23(H2O)12] -O-[Si(OH)2]n-OH + H2O
- 13AlCl3 + 28OH + 9H2O-[Si(OH)2]n-OH [2] 7+ [Al13O4(OH)23(H2O)12] -O-[Si(OH)2]n-OH
Thus a structure was proposed similar to that of Sterte and Shabtai (1987) and is depicted in Figure 3. Sterte and Shabtai (1987) also found that the SiAl PILCs had
178 more Brönsted and Lewis acid sites compared to Al-pillared species and a lower Lewis/Brönsted ratio. This was attributed to the presence of acidic silanol groups in the pillars. The cumene cracking ability of the SiAl PILC was examined and they found an increase in activity due to the incorporating of silica into the aluminium pillars.
Zirconium Farfan-Torres and others (1992) describe two methods of preparing Zr PILCs. Both
methods involve the mixing of zirconyl chloride (ZrOCl2) solution to a clay suspension with one having an extra step of refluxing at 100oC to force polymerization of the Zr complex. The non-refluxed solution produces a square planar complex that can then polymerize to give Zr8 and Zr12 units and upon calcination at 500oC, give PILCs with basal spacings of 16Å. This method has the disadvantage of being time consuming whereas the refluxed solution is quicker but produces PILCs with disordered layer structure and basal spacings around 15Å. This led them to investigate how parameters such as time of reaction, temperature of reaction and concentration all affect the degree of polymerization. They found that
although heating of the ZrOCl2 solution and higher contact time assisted the polymerisation process, it also causes the pH to drop which led to degradation of the clay structure. The pH drop is due to the zirconyl ion, present as the tetramer 8+ 2+ + [(Zr(OH)2.4H2O)4] , hydrolysing to form the tetramer [Zr4(OH)14.10H2O] and H . Although the authors do not examine the stability of the pillared clay above 500oC, they do show that the introduction of Zr enhances the acidic properties of the solid.
Ohtsuka (1993) intercalated sodium fluoride tetrasilic mica with ZrOCl2 solutions of various concentrations at room temperature and at an elevated temperature. He produced intercalated clays having interlayer spaces of 7, 12 and 14 Å according to the degree of polymerization of the zirconium tetramer. The dimension of this tetramer was given as 8.98 Å wide and 5.82 Å thick. This was given as the major species responsible for producing the 7 Å interlayer distances with two and three two- dimensional layers being responsible for the 12 and 14 Å species respectively. Halogens in solution (Cl-, Br-) were shown to form part of the zirconium tetramer and, if present in high concentrations, had an effect on the extent of polymerisation.
179 The PILC formed from the 14Å species has an homogenous pore structure and exhibited good thermal stability with the interlayer spacing falling to only 10Å upon heating to 700oC and the surface area reaching a maximum at 600oC. Gandía et al. (1999) pillared saponite and montmorillonite using, what they term, a non-aggressive method in which a commercial solution of zirconium in acetic acid was used. The intercalation step was performed with pH=3.3 at room temperature while the calcination was performed at 500oC. This method produced basal spacings of 14-24 Å with surface areas up to 300 m2/g. In a further study (Gill and others 2000), they found that upon calcination the Zr PILC actually increased in surface area. This increase was attributed in part to the decomposition of acetate ligands thus giving access to the porous network of the PILC. A comparison of the properties of the Zr PILC to the ZrAl species was made by
Cañizares and others (1999). They prepared the Zr-PILC using a ZrOCl2 solution mixed with bentonite. They tried three methods to produce the ZrAl PILC: (1) impregnating an Al-PILC with Zr, (2) mixing the ZrOCl2 solution to previously intercalated Al clay slurry and (3) mixing of Al and Zr pillaring solutions, addition of a basic solution to give OH/Al ratio of 2 and then ageing for 16 hours. The first method gave products which could not be calcined higher than 200oC before structural collapsed. The products of the second method could be calcined successfully but resulted in low basal spacings. The third method gave acceptable PILCs and was used for further analysis. They only analysed the PILCs up to 500oC but were able to conclude that, “the surface acidity, methane adsorption and thermal stability were increased by incorporating aluminium into the single oxide pillars”. It was also shown that the structure of the pillar varied with Al concentration with the Keggin structure predominating at higher Al/Zr ratios.
Chromium
A solution of CrCl3 and AlCl3 in Na2CO3 was used by Zhao and others (1995) to produce a CrAl PILC. Basal spacings of around 18 Å were achieved when calcined at 500oC with surface areas around 230 m2/g. The thermal stability was found to be greater than that of Cr PILC as well as possessing more acid sites. Toranzo and others 3+ 5+ (1997) stated that Cr can form the trimer [Cr3(OH)4(H2O)9] , the dimer
180 4+ 3+ 3+ [Cr2(OH)2(H2O)8] or no polymerization occurs depending on the Al /Cr ratio in the solution. They also found that CrAl PILCs were more stable thermally than Cr PILCs.
Titanium
Del Castillo and Grange (1993) determined that Ti(OEt)4 (titanium tetraethoxide) gave a polycationic precursor that gave a PILC of regular structure and was stable at 600oC. Regulating the pillar size and distribution was difficult as it depended greatly on solution pH and reaction temperature. The acidity of the Ti PILCs produced were determined to be mainly of the Lewis kind. Swarnakar and others (1996) prepared a TiAl pillaring solution by hydrolysis of
AlCl3 and Ti(OEt)4 for pillaring of beidellite and montmorillonite. From XRD analysis it was determined that the pillared beidellite was thermally more stable than the pillared montmorillonite with a peak seen at 700oC.
Tantalum The intercalation of montmorillonite using niobium and tantalum was performed early on by Christiano and others (1985). They produced PILCs with basal spacings around 18 Å but they were only stable to 400oC. Guiu and Grange (1994, 1997) examined ways to produce more stable Ta PILCs. They prepared a pillaring solution
by controlling the hydrolysis of Ta(OC2H5)5 in an ethanolic acidic solution. The PILC had a basal spacing of 26 Å and was stable to 600oC. A pillar precursor structure was
proposed as [Ta8O10(OR)20], R= H, C2H5. The pillaring of Ta was shown to produce stronger Lewis sites and new Brönsted sites.
Preparation techniques
Altering the preparation of a PILC can have dramatic effects on properties such as thermal stability and acidity. This area also has received considerable attention with many authors who are looking at ways to economise the process for commercial viability. Current problems in preparation are time and energy costs, water usage and mixing of clay solutions. Some of these issues, along with property changes, have been addressed and will now be discussed briefly.
181 Intercalation step
Mixing of the clay suspension with the pillaring solution over a time period with heat is the standard method for intercalation of clays. This process can take up to 6 hours to complete as the pillaring solution is added drop-wise and the mixture is stirred for at least 2 hours with heat supplied. Although this method has had success in laboratory synthesis of pillared clays, it is not an ideal preparation technique for large scale production because large amounts of water and heat are necessary. The use of ultrasonics in this step was reported by Katdare and others (2000). They intercalated a Ca-montmorillonite using ultrasonic treatment over a number of time periods. The most intense and sharpest peaks on XRD patterns were for the calcined sample that had been left in the ultrasonic bath for 20 minutes. They than varied the pillaring solution and determined the optimum Al3+/clay ratio as being 20 meq/g. This PILC (PILCUS) had a basal spacing of 19.2 Å and a BET surface area of 281 m2/g. To test the thermal and hydrothermal stability of the PILCUS, they heated it to 900oC in 200o steps as well as to 750oC with 100% steam for 8 hours. A similar treatment was given to a PILC (PILCONV) intercalated by the conventional method. The results showed that the structure of PILCONV collapsed at around 700oC while the PILCUS still had some structure at 900oC. The hydrothermal results were the same with the PILCUS still having a basal spacing of 18.1 Å and a surface area of 189 m2/g. The stability was attributed to the uniform pillaring obtained by the use of ultrasonics. They also reported how the use of ultrasonics did not alter properties such as acidity and catalytic activity. The same authors, in a later study (Katdare and others 2000), looked at how the exchangeable ions present in the starting clay affected the ultrasonic treatment. They converted the Ca-montmorillonite to Na+ and La3+ forms by ion exchange. This gave exchangeable cations with valencies of +1, +2 and +3. They found that the optimum times for ultrasonic treatment were 5 minutes for the Na form, 20 minutes for the Ca form and 80 minutes for the La form. The increase in time was due to the higher charge ions being more tightly bound to the clay layers. They also concluded that the 7+ role of ultrasound is to accelerate the [Al13] diffusion within the clay layers.
182 This method of intercalation has a number of advantages that help to make large- scale production of pillared clays more viable. First, it reduces the time needed from several hours to less than 30 minutes. It also requires no heat for the process, thus saving in costs and reducing the safety risks, although some safety issues arise with ultrasonics that would need to be addressed. Finally, the clay suspension required can be more concentrated compared to conventional methods, thus using less water and space. Fetter and others (1996) also looked at a way to speed up the intercalation step by using microwave irradiation. They made up a 10 wt% clay solution and added aluminium chlorohydrate to it making an Al/clay ratio of 5mmol/g. The sample was sealed and subjected to microwave irradiation for various time periods and then pillared by conventional methods. The samples prepared using microwaves gave surface areas some 20-30% higher than samples prepared by the conventional method (ie. mixing for 18 hours). They also found that the irradiation time had little effect on the surface area with a maximum of 347 m2/g being attained after only 5 minutes. In a later study, Fetter and others (1997) were able to similarly prepare a pillared clay using microwave irradiation for 7 minutes but with a more concentrated starting clay slurry of 50 wt%. They achieved a surface area of 331 m2/g, some 87 m2/g higher than the sample they prepared by the conventional method.
Separating and Washing
This step serves to remove any excess ions that are present in preparation for the calcining of the intercalated clay. The separation can be done by filtration or, for a faster result, centrifugation. The washing procedure is more time consuming, as it usually requires the re-suspension of the separated clay in deionised water with stirring for a time period. This process is done a number of times, usually until the filtrate is free of chloride ions as determined by the AgNO3 test. All of this is time consuming and would be hard to scale-up as vast amounts of water would be required. Thomas and Occelli (2000) examined the effect that washing had on an Al intercalated montmorillonite. They examined samples of the intercalated clay after
183 each washing up to 4 times with 400 cm3 of deionised water as well as a sample without any washing. The XRD results showed a broad weak peak for the unwashed sample, which then shifted to 18.9 Å for the first washing, and this reflection became sharper with subsequent washing. They conclude that the initial cations present in the interlayer space were not Keggin ions but that they were formed, "in situ by base hydrolysis of the different oligomers present". Hence, the washing procedure is necessary to form stable PILCs as their stability is related to the formation of these Keggin ions. Aceman and others (2000) examined how allowing the clay to age in the pillaring solution and how the use of dialysis compared to conventional washing methods. They also determined that initially aluminium is adsorbed into the interlayer space either in a monomeric state or as small oligomers. Several days were needed for these species to undergo hydrolytic oligomerization to form Keggin-like ions. This could only occur however, if the excess Al, Na and Cl ions had been removed by washing or by dialysis. For dialysis they put the intercalated sample in a Visking dialysis bag dipped in double-distilled water for one week at room temperature. For laponite and hectorite samples, the dialyzed and non-dialysed samples showed similar poor thermal stability. For montmorillonite, the dialyzed sample was comparable to the sample washed four times. For beidellite and saponite, the dialyzed samples gave the best thermal stability results with XRD peaks being more intense and sharper than those of the washed sample.
Introduction of organic molecules
The use of organic molecules, such as surfactants, in the pillaring solution to act as a swelling has also been examined. This helps to reduce significantly the amount of water required while also making the clay easier to filter. It involves the introduction of an organic molecule into the interlayer along with the pillaring species. The organic molecule, of known shape and size, can later be removed by heat to leave a more regular pore structure. By investigating how such organic molecules affect the pillaring process, specific pore sizes can be designed.
184 Michot and Pinnavaia (1992) incorporated a nonionic surfactant of general formula
C12-14H25-29O(CH2CH2O)5 with the usual Al-pillaring solution. The resultant PILC had a more uniform micropore distribution and a sharper, more symmetrical 001 reflection compared to one prepared without surfactant. The basal spacing was however smaller than normal (15.3 Å) but this was attributed to the surfactant limiting the condensation of Al13 units within the interlayer space. The surface area (305 m2/g) was also slightly larger than the conventional PILC (279 m2/g) and contained more mesopores. An important feature of this method is that the intercalated product is easily filtered and could be washed free of excess ions using two thirds less water than that required by the conventional method. Galarneau and others (1995) converted a Li-fluorohectorite to a quaternary- ammonium exchanged form and then mixed with a solution of neutral amine and tetraethylorthosilicate (TEOS). The interlayer space was swollen which allowed the TEOS to enter and hydrolyse. The surfactants were then removed by calcining at 600oC leaving a rigid silica framework between the layers. The basal spacings of the PILC ranged from 14.9 to 23.4 Å, depending on the chain length of the neutral amine used. Suzuki and others (1988) were able to use polyvinyl alcohol as a pre-swelling agent
for pillaring of montmorillonite with Al13. Hectorite was similarly (Suzuki and others 1991) and they showed how the presence of polyvinyl alcohol provided favourable conditions to allow smaller Al cations to hydrolyse into larger ones. As they increased the concentration of aluminium chlorohydroxide in the pillaring solution, the basal spacings increased, as did the pillar concentration. This was not the case with the PILC prepared without polyvinyl alcohol.
Cracking of vegetable oil
As mentioned earlier, transesterification of vegetable oil is currently the method of choice for conversion to a useable fuel. Many studies have however been done on the upgrading of such oils using various catalysts and reaction conditions. These studies have examined the products obtained by the cracking processes and have tried to correlate the effect that these variables have on the final producs. The problem with
185 any research done in this field is that because the starting material (vegetable oil) is a complex mixture that changes drastically between types, it is difficult to extrapolate or replicate the results. Hence, the reporting of any results in this review must be used as a guide only and can not be expected to extend totally to the anticipated outcome of this project.
Canola oil
Katikaneni and others (1995a) used a number of catalysts, including Al- PILC, to convert canola oil to fuel using a fixed bed reactor. They examined how each catalyst performed with respect to organic liquid product (OLP) yield, selectivity and coke formation. They found that HZSM-5 gave the highest yield of OLP of 63 mass % with the pillared clay being third giving a 55 mass % yield. The OLP of the PILC contained more aliphatic hydrocarbons than the other catalysts and the least amount of aromatic hydrocarbons. Their results showed that as the pore size of the catalyst was increased, the conversion of canola oil, the coke formation and the selectivity for aliphatics increased while the yield of hydrocarbons and the selectivity for aromatics decreased. This led them to the conclusion that medium pore catalysts enhanced the initial cracking and deoxygenation reactions needed for an optimum fuel yield. They were also able to propose a reaction pathway for conversion by looking at the products formed (Fig. 4). After the initial cracking occurs (step 1) further reaction steps were proposed for the heavy hydrocarbons and oxygenates formed. Both are thought to undergo secondary cracking (steps 2 and 5) to form gas products. The heavy oxygenates can also be deoxygenated (step 4) to form CO, CO2, methanol and
acetone while the heavy hydrocarbons undergo aromatization (step 3) to form C9+ aromatic hydrocarbons. The authors suggest that the initial cracking occurs in the inter-planar space and then diffuses into the pores of the clay sheets where further reactions proceed. One of these reactions is polymerization, which leads to coke formation and a clogging of the inter-planar space. A further study by these authors (Katikaneni and others 1995b) showed how co- feeding with steam during the reaction helps to increase olefin formation as well as increase the catalyst life by decreasing coke formation.
186 Vonghia and others (1995) suggested that the initial cracking reaction occurred via two mechanisms: β-elimination and γ-hydrogen transfer (Fig. 5). Both are initiated by the bonding of a carbonyl oxygen to a Lewis acid site on the catalyst. It is possible for both reactions to occur on a triacylglyceride molecule but β-elimination can only happen once. The effect that acidity, basicity and shape selectivity of a catalyst had on the conversion of canola oil was examined by Idem and others (1997). The various catalysts and their properties are listed in table 3. An empty reactor run was performed to evaluate the contribution of each catalyst, and the products of this run led them to the conclusion that initial decomposition of the oil to heavy hydrocarbons and heavy oxygenates was independent of catalyst properties. The effect shape selectivity had on the product was evaluated by comparing HZSM-5 and silicalite against the empty reactor run. These catalysts gave a higher OLP yield along with a lower gas yield. This led them to the conclusion that only limited secondary cracking was allowed because of the long molecules diffusing through the pore structure with minimal C-C bond scission. The OLP from the two catalytic runs
had a higher fraction of C7-C9 aromatic hydrocarbons. This was attributed to an increase in cyclization and aromatisation reactions (eg. Diels-Alder and dehydrogenation) that are allowed to occur within the pores of the catalyst. HZSM-5, silica-alumina and γ-alumina were compared to evaluate the role of catalyst acidity in conversion products. Product distribution for runs using silica-alumina and γ-alumina were very similar to the empty reactor run with yields being slightly higher. As both of these catalysts contain Brönsted and Lewis acid sites, it was suggested that the acidity of the catalyst did not determine the product selectivity. The increase in OLP yield and total aromatic hydrocarbons obtained by using HZSM- 5 was attributed mainly to its shape selective properties, not acidity. To examine how basicity affected the product yield, they used calcium oxide and magnesium oxide as catalysts. The results showed that the presence of basic centres inhibited secondary cracking and produced large amounts of residual oil. Katinkaneni and others (1998), in a later study, looked at the conversion products using HZSM, HS-mix and silica-alumina catalysts but they carried out the reactions
187 in a fluidised bed reactor. This involved the continuous flow of argon through the catalyst bed, which assists in catalyst regeneration. They summarized the reaction sequence (below) proposed from previous studies and related the catalyst properties to the conversion products by means of this sequence.
Canola oil ⇒ long-chain CxHy + long-chain [1]
oxygenated CxHy (thermal)
long-chain oxygenated CxHy ⇒ long-chain [2]
CxHy + H2O + CO2 + CO (thermal + catalytic)
long-chain CxHy ⇒ paraffins + olefins [3] (short and long-chain) (thermal + catalytic)
short-chain olefins ⇒ C2-C10 olefins (catalytic) [4]
C2-C10 olefins ⇔ aliphatic CxHy + [5]
aromatic CxHy (catalytic)
canola oil ⇒ coke (thermal) [6]
n(aromatic CxHy) ⇒ coke (catalytic) [7]
The Brönsted acidity of the catalyst can enhance the reactions that occur in steps 2-4 and this was confirmed by their results which showed that canola oil conversion increased as the reaction temperature and catalyst acid site density was increased. Such conversions were also enhanced by a decrease in fluidising gas velocity, as there was a greater contact time with the catalyst. The selectivity for OLP in a fluidised bed reactor was lower than that achieved in a fixed bed reactor. This too was
188 related to the shorter contact time in the fluidised-bed reactor, which did not allow the
formation of additional OLP from C2-C5 olefins in steps 4 and 5.
Palm oil
Leng and others (1999) used a fixed bed reactor to crack palm oil over HZSM-5 catalyst. The maximum formation of gasoline range hydrocarbons was achieved at 400oC with a low space velocity. The conversion of palm oil was low (40-70%) compared to those using canola oil (Idem and others 1997) where conversions up to 100% were achieved. This was attributed to the fact that palm oil contains more saturated fatty acids (palmitic acids) than canola oil and these have a greater stability than unsaturated fatty acids. A similar reaction pathway (Fig. 6) to that shown for canola oil was proposed for palm oil conversion over HZSM-5, with deoxygenation and primary cracking being the initial reactions. Likewise, secondary reactions were controlled by catalyst properties (eg. acidity and pore structure) as well as the reaction conditions (eg. temperature and flow rate). Twaiq and others (1999) also looked at palm oil conversion using HZSM-5 as well as zeolite β and ultrastable Y (USY) zeolites. They were able to achieve conversions of up to 99 wt % with gasoline yields of 28 wt%. They concluded that HZSM-5 was the best catalyst for conversion, gasoline yield, selectivity for aromatic and lower coke formation.
Sunflower oil
A study by Dandik and others (1998) examined the products of the conversion of used sunflower oil with HZSM-5 using a special fractionating pyrolysis reactor. A conversion of 96.6% was achieved at 420oC with an OLP yield of 33%. The length of the fractionating column on the reactor had an effect on the OLP content with an increase in length giving a significant increase in n-alkene content. This is given as a variable that can be adjusted to optimise the fuel product.
Acknowledgements
189 The authors wish to thank the Inorganic Materials Research Program, School of Physical and Chemical Sciences, Queensland University of Technology, for the infrastructural and financial support.
References
Anonymous (2001) India: Palm oil to lead global vegoil trade growth. Business line, January 18 Aceman S, Lahav N, Yariv S (2000) A thermo-XRD study of Al-pillared smectites differing in source of charge, obtained in dialyzed, non-dialyzed and washed systems. Applied Clay Science 17: 99-126 Altin R, Çetinkaya S, Yücesu, HS (2001) The potential of using vegetable oil fuels as fuel for diesel engines. Energy Conversion and Management 42: 529-538 Booij E, Kloprogge JT, Van Veen, JAR (1996) Preparation, structural characteristics and catalytic properties of large-pore rare earth element (Ce, La)/Al-pillared smectites. Clays and Clay Minerals 44: 774-782 Bradley SM, Howe RF, Kydd RA (1993) Correlation between aluminum-27 and gallium-71 NMR chemical shifts. Magn. Reson. Chem. 31: 883-886 Bradley SM, Kydd RA (1991) A comparison of the thermal stabilities Ga13, GaAl12 and Al13 pillared clay minerals. Catalysis Letters 8: 185-192 Bradley SM, Kydd RA (1993) Ga13, Al13, GaAl12 and chromium-pillared motmorillonites: Acidity and reactivity for cumene conversion. Journal of Catalysis 141: 239-249 Bradley SM, Kydd RA, Fyfe CA (1992) Characterization of the galloaluminate 7+ GaO4Al12(OH)24(H2O)12 polyoxocation by MAS NMR and infrared spectroscopies and powder x-ray diffraction. Inorg. Chem. 31: 1181-5 Bradley SM, Kydd RA, Yamdagni R (1990a) A comparison of the properties of the Ga13, GaAl12 and Al13 polyoxycations in solution and of their corresponding pillared clays. Mat. Res. Soc. Extended Abstract EA-24: 69-72 Bradley SM, Kydd RA, Yamdagni R (1990b) Study of the hydrolysis of combined Al3+ and Ga3+ aqueous solutions: formation of an extremely stable 7+ dodecaaluminiogallium (GaO4Al12(OH)24(H2O)12 ) polyoxycation. Magn. Reson. Chem. 28: 746-50 Cañizares P, Valverde JL, Kou MRS, Molina CB (1999) Synthesis and characterization of PILCs with single and mixedpillars from two different bentonites. A comparative study. Microporous and Mesoporous materials 29: 267-281 Cert A, Moreda W, Pérez-Camino MC (2000) Chromatographic analysis of minor constituents in vegetable oils. Journal of Chromatography 881: 131-148 Christiano SP, Wang J, Pinnavaia TJ (1985) Intercalation of niobium and tantalum M6Cl12n+ cluster cations in montmorillonite: a new route to pillared clays. Inorg. Chem. 24: 1222-7 Dandik L, Aksoy HA, Erdem-Senatalar A (1998) Catalytic conversion of used oil to hydrocarbon fuels in a fractionating pyrolysis reactor. Energy & Fuels 12: 1148-1152
190 Del Castillo HL, Grange P (1993) Preparation and catalytic activity of titanium pillared montmorillonite. Applied Catalysis A. 103: 23-34 Department of Natural Resources and Environment, V, Australia (2001) Oilseed Program Ding Z, Kloprogge JT, Frost RL, Lu GQ, Zhu HY (2001) Porous clays and pillared clay-based catalysts. Part 2: A review of teh catalytic and molecular sieve applications. Journal of Porous Materials 8: 273-293 Domínguez JM, Botello-Pozos JC, López-Ortega A, Ramírez MT, Sandoval-Flores G, Rojas-Hernández A (1998) Study of pillar precursors [Ga(III)-Al(III), Ln(III)-Al(III), Zr(IV)] for hydrothermally stable pillared clays. Catalysis Today 43: 69-77 Farfan-Torres EM, Sham E, Grange P (1992) Pillared clays: Preparation and characterization of zirconium pillared montmorillonite. Catalysis Today 15: 515-526 Fetter G, Heredia G, Maubert AM, Bosch P (1996) Synthesis of Al-intercalated monmorillonites using microwave irradiation. J. Mater. Chem. 6: 1857-1858 Fetter G, Heredia G, Velázquez LA, Maubert AM, Bosch P (1997) Synthesis of aluminium-pillared montmolrillonites using highly concentrated clay suspensions. Applied Catalysis A 162: 41-45 Fetter G, Tichit D, de Menorval LC, Figueras F (1995) Synthesis and characterization of pillared clays containing both Si and Al pillars. Applied Catalysis A 126: 165-176 Fetter G, Tichit D, Massiani P, Dutartre R, Figueras F (1994) Preparation and characterization of montmorillonites pillared by cationic silicon species. Clays and Clay Minerals 42: 161-169 Friedt W, Lühs W (1998) Recent developments and perspectives of industrial rapeseed breeding. Lipid 100: 219-226 Frost, R.L., Kristof, J., Paroz, G.N., and Kloprogge, J.T. (1998) Raman spectroscopy of intercalated kaolinites. 16th International Conference on Raman Spectroscopy, Cape Town, South Africa: 686-887 Galarneau A, Barodawalla A, Pinnavaia TJ (1995) Porous clay heterostructures formed by gallery-templated synthesis. Nature 374: 529-531 Gandía LM, Toranzo RVicente MA, Gil A (1999) Non-aggressive pillaring of clays with zirconium acetate. Comparison with alumina pillared clays. Applied Catalysis A 183: 23-33 Gil A, Gandía LM (2000) Recent advances in the synthesis and catalytic applications of pillared clays. Catal. Rev - Sci. Eng. 42: 145-212 Gill A, Vicente MA, Gandía LM (2000) Main factors controlling the texture of zirconia and alumina pillared clays. Microporous and Mesoporous materials 34: 115-125 González F, Pesquera C, Benito I, Herrero E, Poncelet G, Casuscelli S (1999) Pillared clays: catalytic evaluation in heavy oil cracking using a microactivity test. Applied Catalysis A 181: 71-76 González F, Pesquera C, Blanco C, Benito I, Mendorioz S (1992) Synthesis and characterisation of Al-Ga pillared clays with high thermal and hydrothermal stability. Inorg. Chem 31: 727-731
191 Guiu G, Gill A, Montes M, Grange P (1997) Tantalum-pillared montmorillonite: I. Synthesis and physicochemical characterization. Journal of Catalysis 168: 450-462 Guiu G, Grange P (1994) Synthesis and characterization of tantalum pillared montmorillonite. J. Chem. Soc., Chem. Commun.: 1729-30 Guiu G, Grange P (1997) Tantalum-pillared montmorillonite: II Acidic and catalytic properties. Journal of Catalysis 168: 463-470 Gunstone FD (1996) Fatty acid and lipid chemistry. Blackie Academic & Professional, London Gunstone FD, Society of Chemical Industry (Great Britain). (1987) Palm oil. Published for the Society of Chemical Industry by Wiley, Chichester [West Sussex] New York Guzman DD (2001) Oilseed and animal fats look for recovery this year. Chemical Market Reporter, 259, 7 Hernando MJ, Pesquera C, Blanco C, Benito I, Gonzalez F (1996) Effect of Ce on catalytic properties of pillared montmorillonite with Al- and GaAl- polyoxycations. Appl. Catal., A 141: 175-183 Idem RO, Katikaneni SPR, Bakhshi NN (1997) Catalytic conversion of canola oil to fuels and chemicals: roles of catayst acidity, basicity and shape selectivity on product distribution. Fuel Processing Technology 51: 101-125 Katdare SP, Ramaswamy V, Ramaswamy AV (2000) Factors affecting the preparation of alumina pillared montmorillonite employing ultrasonics. Microporous and Mesoporous materials 37: 329-336 Katikaneni SPR, Adjaye, JD, Bakhshi NN (1995a) Catalytic conversion of canola oil to fuels and chemicals over various cracking catalysts. The Canadian Journal of Chemical Engineering 73: 484-497 Katikaneni SPR, Adjaye, JD, Bakhshi NN (1995b) Studies on the catalytic conversion of canola oil to hydrocarbons: Influence of hybrid catalysts and steam. Energy & Fuels 9: 599-609 Katikaneni SPR, Adjaye JD, Idem RO, Bakhshi NN (1998) Performance studies of various cracking catalysts in the conversion of canola oil to fuels and chemicals in a fluidized-bed reactor. JAOCS 75: 381-391 Kloprogge JT (1998) Synthesis of smectites and porous pillared clay catalysts: Review. J. Porous Materials 5: 5-42 Kloprogge JT, Frost RL (1999) Raman microscopy study of basic aluminum sulfate. Part II. Raman microscopy at 77 K. J. Mater. Sci. 34: 4367-4374 Kloprogge JT, Seykens D, Geus JW, Jansen JBH (1992) A 27Al nuclear magnetic resonance study on the optimalization of the development of the Al13 polymer. J. Non-Cryst. Solids 142: 94-102 Lenarda M, Storaro L, Pellegrini G, Piovesan L, Ganzerla R (1999) Solid acid catalysts from clays. Part 3: benzene alkylation with ethylene catalyzed by aluminum and aluminum gallium pillared bentonites. Journal of Molecular Catalysis A: Chemical 145: 237-244 Leng TY, Mohamed AR, Bhatia S (1999) Catalytic conversion of palm oil to fuels and chemicals. The Canadian Journal of Chemical Engineering 77: 156-162 Ma F, Hanna MA (1999) Biodiesel production: a review. Bioresource Technology 70: 1-15
192 Michot LJ, Pinnavaia TJ (1992) Improved synthesis of alumina-pillared montmorillonite by surfactant modification. Chem. Mater. 4: 1433-1437 Ohtsuka K (1993) Microporous ZrO2-pillared clays derived from three kinds of Zr polynuclear ionic species. Chem. Mater. 5: 1823-1829 Pires J, Machado M, de Carvalho MB (1998) Porosity and thermal stability of PILCs prepared with clays from different origins and different metal- polyhydroxycationic species of Al and Al/Ce. J. Mater. Chem. 8: 1465-1469 Sterte J. (1991) Preparation and properties of large-pore La-Al-pillared montmorillonite. Clays and Clay Minerals 39: 167-173 Sterte J, Shabtai J (1987) Crosslinked smectites. V. Synthesis and properties of hydroxy-silicoaluminum montmorillonites and fluorhectorites. Clays Clay Miner. 35: 429-439 Suzuki K, Horio M, Masuda H, Mori T (1991) Control of the distance between the silicate layers of hectorite by pillaring with alumina in the presence of polyvinyl alcohol. J. Chem. Soc., Chem. Commun. 1991: 873-874 Suzuki K, Mori T, Kawase K, Sakami H, Iida S (1988) Unique pore structure formed in montmorillonite in the presence of polyvinyl alcohol and aluminium chlorhydroxide. J. Chem. Soc., Chem. Commun. 1988: 122-123 Swarnakar R, Brandt KB, Kydd RA (1996) Catalytic activity of Ti- and Al-pillared montmorillonite and beidellite for cumene cracking and hydrocracking. Applied Catalysis A 142: 61-71 Thomas JM, Thomas, WJ (1997) Principles and practice of heterogeneous catalysis. VCH Publishers Inc., New York Thomas SM, Occelli ML (2000) Effects of synthesis conditions on the thermal 7+ stability of a Texas montmorillonite expanded with [Al13O4(OH)24(H2O)12] cations. Clays and Clay Minerals 48: 304-308 Toranzo R, Vicente MA, Banares-Munoz MA (1997) Pillaring of a saponite with aluminum-chromium oligomers. Characterization of the solids obtained. Chem. Mater. 9: 1829-1836 Twaiq FA, Zabidi NAM, Bhatia S (1999) Catalytic conversion of palm oil to hydrocarbons: Performance of various zeolite catalysts. Ind. Eng. Chem. Res. 38: 3230-3237 Valverde JL, Cañizares P, Kou MRS, Molina CB (2000) Enhanced thermal stability of Al-pillared smectites modified with Ce and La. Clays and Clay Minerals 48: 424-432 Vassallo AM, Cole-Clarke PA, Pang LSK, Palmisane A (1992) Infrared emission spectroscopy of coal minerals and their thermal transformation. J. Applied Spectroscopy 46: 73-78 Venuto PB, Habib JET (1979) Fluid catalytic cracking with zeolite catalysts. 156 Marcel Dekker, Inc New York Vonghia E, Boocock DGB, Konar SK, Leung A (1995) Pathways for the deoxygenation of triglycerides to aliphatic hydrocarbons over activated alumina. Energy & Fuels 9: 1090-1096 Zhao D, Yang Y, Guo X (1992) Preparation and characterization of hydrosilicoaluminium pillared clays. Inorganic Chemistry 31: 4727-4732 Zhao D, Yang Y, Guo X (1995) Synthesis and characterization of hydroxy-CrAl pillared clays. Zeolites 15: 58-66
193
Table 1 - Basal spacings, (Å) at different temperatures (González and others 1992)
194 Table 2- Properties of SiAl PILCs produced by 2 methods (Sterte and Shabtai 1987) Method of Si/Al ratio in hydroxy- Surface area (m2/g) d(001)(Å) preparation SiAl solutions 1 0.0 458 19.2 1 0.53 369 19.3 1 1.04 324 19.5 1 2.18 278 19.0 2 0.0 499 17.6 2 0.5 498 17.7 2 1.0 460 17.2 2 2.0 343 17.0
195 Table 3- Characteristics of catalysts used by Idem and others (1997). Strength Acidic, Pore Surface Catalyst Type of Si/Al of acid Pore Shape basic or size area type acidity ratio or basic structure selectivity neutral (nm) (m-1g-1) sites HZSM-5 Acidic Mostly B 56 Strong Crystalline 0.54 329 V. high Silicalite Neutral None No Al N/A Crystalline 0.54 401 V. high Silica Neutral None No Al N/A Amorphous 11.46 211 None γ-alumina Acidic B and L 0 Moderate Amorphous 14.93 241 None Silica- Acidic B and L 0.79 Moderate Amorphous 3.15 321 None alumina Calcium Basic None N/A Strong Amorphous 11.86 7 None oxide Magnesium Basic None N/A Weak Amorphous 15.22 27 None oxide Empty neutral none N/A N/A N/A N/A None None reactor
196 Figure captions
Fig. 1 Compositions of the most common vegetable oils.
Fig. 2 Schematic representation of the repillaring process
Fig. 3 Schematic model of the SiAl pillaring complex (modified after Sterte and
Shabtai, 1987)
Fig. 4 Reaction pathway for canola oil conversion using pillared clay catalysts as proposed by Katikaneni and others (1995a,b).
Fig. 5 Two possible pathways for cracking reactions via β-elimination and γ- hydrogen transfer.
Fig. 6 Proposed reaction pathway for cracking of palm oil over HZSM-5 (Leng and others 1999).
197 Myristic
1% Oleic Palmitic 38% 47%
10%
Linoleic Palm Oil
Linolenic 8% Palmitic Linoleic 54% 11%
Oleic 22%
Soybean Oil
Palmitic
7%
Linoleic Oleic 75% 14%
Sunflower Oil Myristic
10%
Oleic 17% 46% Erucic
Linoleic 13% 4% 5%
Rapeseed Oil Palmitic Linolenic
Myristic 17% 48% Lauric
5% 3% 8% Oleic
Linoleic Palmitic Coconut Oil
Fig. 1 198
Fig. 2
199
Fig 3
200
Canola Oil
7 Deoxygenation and Cracking
Coke 1
Heavy Hydrocarbons + Heavy Oxygenates Polymerization 6
3 Cracking Deoxygenation 5 Aromatization 2 4
Aromatic Hydrocarbons CO + CO2 C1-C6 Hydrocarbon + Methanol Gases + Acetone
Fig. 4
201
Fig. 5 β-elimination
CH2 O C R1
O H O
C O C R1
O H CH R2
CH2 O C CH2
CH2
γ-hydrogen transfer
CH2 O O C O C R1
H O C R1 CH2 CH - R2 OH + +
CH2 O C CH2
202
Palm Oil
Deoxygenation and cracking
Coke Heavy hydrocarbons + Oxygenates
Secondary cracking
+ deoxygenation
Light olefins + Light paraffins (gasoline) Polymerisation + CO2 + alcohol + CO + H2O
Oligomerisation Aromatisation, Alkylation Isomerisation
Olefins + paraffin Aromatic hydrocarbons (gasoline, diesel & kerosine)
Gases (light olefins, paraffins, CO, CO2, H2O)
Fig. 6
203 XPS STUDY OF THE MAJOR MINERALS IN BAUXITE:
GIBBSITE, BAYERITE AND (PSEUDO-) BOEHMITE
J. Theo Kloprogge1, Loc V. Duong1, Barry J. Wood2, and Ray L. Frost1
1 Inorganic Materials Research Program, Queensland University of Technology, 2
George Street, GPO Box 2434, Brisbane, Q 4001, Australia
E-mail: [email protected]
2 Brisbane Surface Analysis Facility. The University of Queensland, Brisbane, Qld
4072, Australia
204
Abstract
Synthetic corundum (Al2O3), gibbsite (Al(OH)3), bayerite (Al(OH)3), boehmite
(AlOOH) and pseudoboehmite (AlOOH) have been studied by high resolution XPS.
The chemical composition based on the XPS survey scans were in good agreement with the expected composition. High resolution Al 2p scans showed no significant changes in binding energy, with all values between 73.9 and 74.4 eV. Only bayerite showed two transitions, associated with the presence of amorphous material in the sample. More information about the chemical and crystallographic environment was obtained from the O 1s high resolution spectra. Here a clear distinction could be made between oxygen in the crystal structure, hydroxyl groups and adsorbed water. Oxygen in the crystal structure was characterised by a binding energy of about 530.6 eV in all minerals. Hydroxyl groups, either present in the crystal structure or on the surface exhibited binding energies around 531.9 eV, while water on the surface showed binding energies around 533.0 eV. A distinction could be made between boehmite and pseudoboehmite based on the slightly lower ratio of oxygen to hydroxyl groups and water in pseudoboehmite.
Keywords: Bauxite, Bayerite, Boehmite, Corundum, Gibbsite, XPS
Introduction
Bauxite forms a major resource of aluminium in Australia and especially in
Queensland. For that reason research on the mineralogy of these bauxites are of importance to the mining industry. The major aluminum phases recognised in
205 bauxites and laterites are gibbsite also known as hydrargillite (γ-Al(OH)3), and boehmite (γ-AlOOH).
Gibbsite is the main mineral in bauxites formed in areas characterized by a tropical climate with alternating rainy and dry periods (monsoon). Bauxites with primarily boehmite appear to be more constrained to the subtropical areas
(Mediterranean type bauxite). Thermal action or low-grade metamorphism mostly favors diaspore formation. Furthermore, diaspore is formed as a minor constituent in many types of bauxite in addition to gibbsite and boehmite [1-3]. For comparative reasons bayerite (β-Al(OH)3)and corundum (Al2O3) have been incorporated in this study. The thermal behaviour and spectroscopy of these bauxite minerals has been reported in earlier work by our group [4-7].
X-ray Photoelectron Spectroscopy (XPS) is widely used for determining the surface composition of solid materials, including aluminium. Although XPS has become a powerful tool to identify different phases, it has been so far less successful in determining subtle changes in aluminum oxide/hydroxide minerals. Although the binding energies of the core lines (i.e. Al 2p, Al 2s, O 1s, O 2s) are easily measured, the differences in binding energy of Al among the aluminium oxides, hydroxides and oxyhydroxides are very small, generally in the order of 0 to 0.5 eV, which is in the same order of magnitude as the experimental precision of XPS [10-13]. Some limited work has been done on the use valence band XPS to distinguish these minerals
[14,15]. The oxygen core lines may however be more sensitive to changes in the crystal chemistry. This paper therefore reports on the possible use of high resolution
XPS for the identification of the major aluminium oxide/hydroxide/oxyhydroxide minerals.
206 Structure of the aluminum (oxo)hydroxides
Gibbsite Al(OH)3
Gibbsite is monoclinic (P21/n, a = 8.684 Å, b = 5.078 Å, c = 9.736 Å, β =
94.54°) with mostly a tabular pseudohexagonal habit. The structure can be visualized as sheets of hcp layers with open packing between successive sheets. In the lateral extension of the hexagonal closed packed sheets each Al cation is octahedrally coordinated by 6 OH groups and each hydroxyl group is coordinated by two Al cations with one octahedral site vacant [16,17]. This can also be visualized as double layers of OH groups with Al cations occupying two thirds of the interstices within the layers. Each double layer is positioned in such a way that the upper and lower neighboring layers have their hydroxyl groups directly opposite to each other and not in the position of the closest packing. This type of layer structure explains the perfect cleavage of gibbsite parallel to the basal plane (001).
Bayerite Al(OH)3
Bayerite is monoclinic (P21/a, a = 5.0626 Å, b = 8.6719 Å, c = 9.4254 Å, β =
90.26°) forming mostly very fine fibers in radiating hemispherical aggregates and sometimes flaky to tabular crystals to about 0.1 mm. The crystal lattice of bayerite is composed of layers of hydroxyl groups similar to those in gibbsite. These layers, however, are arranged in an AB-AB-AB sequence; in other words the hydroxyl groups of the third layer lie in the depressions between the hydroxyl positions of the second layer.
207
Boehmite AlOOH
Boehmite has the same structure as lepidocrocite (γ-FeO(OH)). The structure of boehmite consists of double layers of oxygen octahedra partially filled with Al cations [18]. Boehmite is orthorhombic with space group Amam (a = 3.6936 Å, b =
12.214 Å, c = 2.8679 Å) [17,19]. The stacking arrangement of the three oxygen layers is such that the double octahedral layer is in cubic closed packing. Within the double layer one can discriminate between two different types of oxygen. Each oxygen atom in the middle of the double layer is shared by four other octahedra, while the oxygen atoms on the outside are only shared by two octahedra. These outer oxygen atoms are hydrogen-bonded to two other similarly coordinated oxygen atoms in the neighboring double layers above and below. The stacking of the layers is such that the hydroxyl groups of one layer are located over the depression between the hydroxyl groups in the adjacent layer.
Experimental
Mineral samples
The aluminium phases used in this study are synthetic gibbsite produced in our laboratory, synthetic gibbsite (γ-alumina) produced by Baikowski International
Corporation (Charlotte, NC), pseudoboehmite synthesised by P. Buining [20], boehmite synthesized by Ray Frost, synthetic bayerite synthesized by Comelco. The samples were analysed for phase purity by X-ray diffraction prior to the XPS analysis.
208 X-ray diffraction has shown that the gibbsites, bayerite and the boehmites are pure.
For comparative reasons synthetic corundum produced by Baikowski International
Corparation (Charlott, NC) was used.
XPS analysis
The minerals were analyzed in freshly powdered form in order to prevent surface oxidation changes. Prior to the analysis the samples were out gassed under vacuum for 72 hours. The XPS analyses were performed on a Kratos AXIS Ultra with a monochromatic Al X-ray source at 150 W. Each analysis started with a survey scan from 0 to 1200 eV with a dwell time of 100 milliseconds, pass energy of 160 eV at steps of 1 eV with 1 sweep. For the high resolution analysis the number of sweeps was increased, the pass energy was lowered to 20 eV at steps of 100 meV and the dwell time was changed to 250 milliseconds.
Band component analysis was undertaken using the Jandel ‘Peakfit’ software package, which enabled the type of fitting function to be selected and allows specific parameters to be fixed or varied accordingly. Band fitting was done using a Lorentz-
Gauss cross-product function with the minimum number of component bands used for the fitting process. The Gaussian-Lorentzian ratio was maintained at values greater than 0.7 and fitting was undertaken until reproducible results were obtained with squared correlations of r2 greater than 0.995.
209 Results and discussion
For all minerals XPS survey spectra and high resolution core line spectra (O 1s and Al 2p) were obtained. Table 1 gives an overview of the chemical composition of the minerals analysed based on the XPS survey scans. In addition to the elements belonging to the mineral one always observes the presence of carbon, the so called advantageous or rubbish carbon. The C 1s transition of this carbon is used to correct for charging, which results in a shift of all other transitions. The amount of carbon is variable but can be as high as 22 atom %. The ratio of oxygen to aluminium for corundum is slightly higher than expected based on the composition of Al2O3. The reason for this becomes apparent from Fig. 1 where the O1 s spectrum shows the presence of three different oxygen species. In addition to the bulk oxygen with a binding energy of 530.7 eV from the crystal structure the surface of corundum contains hydroxyl groups with an O 1s binding energy of 532.1 eV and a minor amount of adsorbed water with an O 1s binding energy of 532.9 eV. All three signals have similar FWHM of about 1.4 eV (Table 2). Only one Al 2p transition is observed at 74.1 eV (Fig. 2).
For gibbsite and bayerite the ratio O to Al should be equal to 3 to 1 based on a compositon of Al(OH)3, which is exactly what is observed for gibbsite. For the bayerite however the ratio is slightly lower than expected but within experimental error. For gibbsite and bayerite only two O 1s transitions are observed at 531.8 and
533.2 eV and at 531.9 and 533.4 eV, respectively, associated with the hydroxyl groups in the crystal structure and absorbed water on the surface. Gibbsite and bayerite both show one major Al 2p transition at 74.3 and 74.4 eV respectively. In addition, bayerite shows a second transition at 75 eV. XRD showed that this bayerite
210 sample had a very low crystallinity and possibly some amorphous content. It might well be that this second transition is associated with this amorphous phase.
The difference between boehmite and pseudoboehmite has been a matter of discussion for a long time. In general there is some consensus that the unit cell of pseudoboehmite is slightly larger than that of boehmite. It has been indicated in the literature that this would be due to the incorporation of water in the crystal structure.
In this study two boehmite samples were analysed, one of which was thought to be pseudoboehmite based on the slightly different XRD pattern. The chemical analyses clearly show a difference in composition. Boehmite in its purest form has a chemical formula of AlOOH and therefore an O to Al ratio of 2 to 1 has to be expected. The boehmite sample has significantly less oxygen than expected whereas the pseudoboehmite has more oxygen than expected plus a trace amount of chlorine. The
Al 2p transitions are slightly different, although still within the experimental error.
The same is the case for the O 1s transitions, but again the values for the pseudoboehmite are slightly higher than for boehmite. The high resolution O 1s spectrum of boehmite shows a nearly 1:1 ratio of oxygen and hydroxyl groups as expected in boehmite. The amount of water in this sample is minimal. In the pseudoboehmite the amount of hydroxyl groups and water are both slightly higher than in boehmite. This may explain the slightly larger unit cell, where a small amount of the oxygen atoms has been replaced by hydroxyl groups and maybe even water molecules.
It is well known, and this study confirms this, that it is very difficult to unambiguously determine any chemical shifts in the Al 2p binding energies among the oxides, hydroxides and oxohydroxides, as these shifts are generally in the order of
0.2 to 0.5 eV, which is not much more than the typical precision of the XPS
211 instrument [10-13]. In general a chemical shift is caused by changes in the electrostatic potential field experienced by the core electrons. Oxidation number, ligand type and coordination (e.g tetrahedral vs. octahedral) can all change the chemical shift of the Al 2p line. This work shows that the differences in these parameters are very small for aluminium in the bauxite minerals.
The O 1s peaks show a slightly larger chemical shifts (up to 0.6 eV) than the
Al 2p peaks. In addition the O 1s peak allows one to distinguish between oxygen, hydroxyl groups and water in the crystal structure and can therefore be used as a technique to identify within the different bauxite minerals the difference between gibbsite and bayerite on one hand and boehmite and diaspore on the other hand.
However, due to the very small chemical shifts no distinction can be made between minerals with the same chemical composition such gibbsite and bayerite.
Refererences
[1] Schoen, R., Roberson, C. E., American Mineralogist 55 (1970) 43-77.
[2] Newman, A. C. D. Chemistry of clay and clay minerals, Longman Scientific &
Technical, Harlow, UK (1987) 480.
[3] van der Marel, H. W., Beutelspacher, H. Atlas of infrared spectroscopy of clay
minerals and their admixtures, Elsevier: Amsterdam (1974) 396.
[4] Ruan, H. D., Frost, R. L., Kloprogge, J. T., Journal of Raman Spectroscopy 32
(2001) 745-750.
[5] Ruan, H. D., Frost, R. L., Kloprogge, J. T., Applied Spectroscopy 55 (2001)
190-196.
212 [6] Ruan, H. D., Frost, R. L., Kloprogge, J. T., Duong, L., Spectrochim. Acta, Part
A 58 (2002) 265-272.
[7] Kloprogge, J. T., Ruan, H. D., Frost, R. L., Journal of Materials Science 37
(2002) 1121-1129.
[8] Frost, R. L., Kloprogge, J. T., Russell, S. C., Szetu, J. L., Applied
Spectroscopy 53 (1999) 423-434.
[9] Frost, R. L., Kloprogge, J. T., Russell, S. C., Szetu, J. L., Applied
Spectroscopy 53 (1999) 423-433.
[10] Strahlin, A., Hjertberg, T., Applied Surface Science 74 (1994) 263-275.
[11] Tsuchida, T., Takahashi, T., J. Mater. Res. 9 (1994) 29219-2224.
[12] Nylund, A., Olefjord, I., Surface and Interface Analysis 21 (1994) 283-289.
[13] Strohmeier, B. R., Surface and Interface Analysis 15 (1990) 51-56.
[14] Thomas, S., Sherwood, P. M. A., J. Chem. Soc., Faraday Trans. 89 (1993)
263-266.
[15] Thomas, S., Sherwood, P. M. A., Anal. Chem. 64 (1992) 2488-2495.
[16] Megaw, H. D., Zeitschrift für Kristallographie 87A (1934) 185-204.
[17] Ramos-Gallardo, A., Vegas, A., Zeitschrift für Kristallographie 211 (1996)
299-303.
[18] Milligan, W. O., McAtee, J. L., Journal of Physical Chemistry 60 (1956) 273-
277.
[19] Christoph, G. G., Corbato, C. E., Hofmann, A., Tettenhorst, R. T., Clays and
Clay Minerals 27 (1979) 81-86.
[20] Buining, P. A., Pathmamanoharan, C., Jansen, J. B. H., Lekkerkerker, H. N.
W., Journal of the American Ceramic Society 74 (1991) 1303-1307.
213 Table 1. Chemical compositions (atom %) of the alumina phases based on the
XPS analyses
corundum gibbsite bayerite boehmite pseudoboehmite O 48.56 62.50 60.19 59.44 62.54 Al 30.25 20.93 20.95 35.58 27.29 Na* bd 2.32 4.23 bd bd N* bd bd 1.56 bd bd Cl* bd bd 0.58 bd 1.35 C** 21.18 14.26 12.48 4.98 8.82 * Impurities ** Advantageous carbon bd - below detection limit
214 Table 2 Binding energies (in eV) of the alumina phases (FWHM in parenthesis).
Al 2p Al 1 Al 2 Corundum 74.1 - Gibbsite 74.4 - Bayerite 74.3 75.0 Boehmite 73.9 - Pseudoboehmite 74.3 -
O 1s O OH H2O Corundum 530.7 (1.4) 532.1 (1.4) 532.9(1.4) Gibbsite - 531.8 (1.5) 533.2 (1.5) Bayerite - 531.9 (1.7) 533.4 (1.7) Boehmite 530.5 (1.5) 531.8 (1.5) 533.0 (1.5) Pseudoboehmite 530.8 (1.6) 532.2 (1.6) 533.5 (1.6)
215
8000
O 530.7 eV 6000 OH 532.1 eV
4000 Intensity (CPS) H2O 532.9 eV 2000
0 537 536 535 534 533 532 531 530 529 528 527 Binding energy (eV)
Fig. 1a O 1s corundum
2000
OH 531.8 eV 1500
1000 Intensity (CPS) H2O 533.2 eV
500
0 537 536 535 534 533 532 531 530 529 528 527 Binding energy (eV)
Fig 1b O 1s gibbsite
216 1600
1400
1200 OH 531.9 eV
1000
800
Intensity (CPS) 600 H2O 533.4 eV 400
200
0 537 536 535 534 533 532 531 530 529 528 527 Binding energy (eV)
Fig 1c O 1s bayerite
700
600 OH 531.8 eV 500 O 530.5 eV
400
300 H2O 533.0 eV Intensity (CPS) 200
100
0 537 536 535 534 533 532 531 530 529 528 527 Binding energy (eV)
Fig. 1d O 1s boehmite
217 900
800
700 O 530.8 eV OH 532.2 eV 600
500
400
H2O 533.5 eV Intensity (CPS) 300
200
100
0 537 536 535 534 533 532 531 530 529 528 527 Binding energy (eV)
Fig. 1e O 1s pseudoboehmite
218 3000
2500
2000
1500
Intensity (CPS) 1000 74.1 eV 500
0 80 79 78 77 76 75 74 73 72 71 70 Binding energy (eV)
Fig. 2a Al 2p corundum
450
400
350
300
250
200
Intensity (CPS) 150
100 74.4 eV
50
0 80 79 78 77 76 75 74 73 72 71 70 Binding energy (eV)
Fig 2b Al 2p gibbsite
219 250
200 75.0 eV
150
Counts 100 74.3 eV
50
0 80 79 78 77 76 75 74 73 72 71 70 Binding energy (eV)
Fig. 2c Al 2p bayerite
250
200
150
100 Intensity (CPS) 73.9 eV
50
0 80 79 78 77 76 75 74 73 72 71 70 Binding energy (eV)
Fig. 2d Al 2p boehmite
220 500
400
300
200 Intensity (CPS)
74.3 eV 100
0 80 79 78 77 76 75 74 73 72 71 70 Binding energy (eV)
Fig. 2e Al 2p pseudoboehmite
221 A X-ray photoelectron spectroscopy study of HDTMAB distribution
within organoclays
Hongping He a,b, Qin Zhou a,c, Ray L. Frost b, Barry J. Wood d, Loc V. Duong b, J. Theo Kloprogge b,*
a Guangzhou Institute of Geochemistry, Chinese Academy of Sciences, Guangzhou 510640, China b Inorganic Materials Research Program, School of Physical and Chemical Sciences, Queensland University of Technology, GPO Box 2434, Brisbane, QLD 4001, Australia c Graduate University of the Chinese Academy of Sciences, Beijing 100039, China d Brisbane Surface Analysis Facility, University of Queensland, Brisbane, QLD 4072, Australia
Corresponding author: [email protected]
222 Abstract X-ray photoelectron spectroscopy (XPS) in combination with X-ray diffraction (XRD) and high-resolution thermogravimetric analysis (HRTG) has been used to investigate the surfactant distribution within the organoclays prepared at different surfactant concentrations. This study demonstrates that the surfactant distribution within the organoclays depends strongly on the surfactant loadings. In the organoclays prepared at relative low surfactant concentrations, the surfactant cations mainly locate in the clay interlayer whereas the surfactants occupy both the clay interlayer space and the interparticle pores in the organoclays prepared at high surfactant concentrations. The former adopts a lateral arrangement for the intercalated surfactants within the interlayer while the latter has a paraffin arrangement. This can well explain the dramatic surface area and pore volume decrease of organoclays compared to those of starting clays. XPS survey scans show that, at low surfactant concentration (< 1.0 CEC), the ion exchange between Na+ and HDTMA+ is dominant whereas both cations and ion pairs occur in the organoclays prepared at high concentrations (> 1.0CEC). High-resolution XPS spectra show that the modification of clay with surfactants has prominent influences on the binding energies of the atoms in both clays and surfactants, and nitrogen is the most sensitive to the surfactant distribution within the resultant organoclays.
Keywords: X-ray photoelectron spectroscopy; X-ray diffraction; Organoclay; Surfactant distribution; Surfactant loading
223 1. Introduction Organoclays represent a family of materials with hydrophobic surfaces, synthesized by modifying swelling clays with various surfactants. During the last 50 years, organoclays have attracted great interest in a number of applications, such as adsorbents for organic pollutants [1-4], rheological control agents [5], reinforcing fillers for plastics [6], clay-based nanocomposites7, and precursors for preparing mesoporous materials [8, 9]. Up to now, a large variety of organoclays have been synthesized using different surfactants [1-4,10-12] and their structures have been characterized using various techniques, including X-ray diffraction (XRD) [11,12], Fourier transform infrared spectroscopy (FTIR) [13-15], Raman spectroscopy [16], thermogravimetric measurement (TG) [17-21], magic-angle-spinning nuclear-magnetic-resonance (MAS NMR) [22,23] and transmission electron microscopy (TEM) [24-26]. In these cases, the detailed information about the interlayer structure, the conformation of the intercalated surfactant and thermal stability of the resultant organoclays rather than the surface characteristics was obtained. However, in various applications of organoclays, the surface characteristics of the resultant organoclays is of high importance since the affinity between the organoclays and the matrix depends strongly on the surface characteristics of the organoclays. Unfortunately, the aforementioned techniques provide little information about the surface characteristics of the organoclays. X-ray photoelectron spectroscopy (XPS) has been demonstrated to be a powerful technique to investigate the surface characteristics of various materials, including clay minerals and related products [27-30]. XPS can provide elemental analysis for essentially the entire periodic table. Because the electrons whose energies are analyzed arise from a depth of no greater than about 2 – 5 nm, the technique is surface-sensitive and suitable to investigate the surface characteristics of clays and the resultant organoclays. With the increase of applications in various fields, the study of organoclay surface characteristics will attract great interest. Zhu and coworkers [2] proposed that the various sorption mechanisms of organoclays for pollutants might result from the different distributions of surfactant within the organoclays. Recently, our study demonstrated that washing the organoclays with solvents resulted in the change of surface energy of the resultant organoclays, resulting from the removal of physically adsorbed surfactant [31]. Both of the abovementioned cases suggest that the distribution of surfactant have a significant effect on the surface property of the organoclays and a consequent influence on their applications. Unfortunately, to date, there is no publication available on XPS of organoclays, which can provide convincing evidence about the distribution of surfactant within organoclays. The objective of this report is to determine the surfactant distribution within the organoclays using XPS, in conjunction with X-ray diffraction (XRD) and high-resolution thermogravimetric analysis (HRTG). This study demonstrates that the distribution of surfactant (in the interlayer space and outside clay layer) depends strongly on the surfactant loading within organoclays and N 1s spectra are most
224 sensitive to the surfactant distribution. This is of high importance to well understand the microstructure of organoclays and for their applications.
2. Experimental 2.1 Materials Ca-montmorillonite (Ca-Mt) was obtained from Hebei, China. The sample was purified by sedimentation and the <2 µm fraction was collected and dried at 90 °C. The sample was ground through a 200 mesh sieve and sealed in a glass tube for use. + Its cation exchange capacity (CEC) is 90.8 meq/100g, determined by NH4 method as described in the literature [32]. Its chemical formula can be expressed as
Ca0.19Mg0.06Na0.01(Si3.96Al0.04)(Al1.44Fe0.09Mg0.47)O10(OH)2·nH2O, calculated from the chemical analysis result. The surfactant used in this study is hexadecyltrimethylammonium bromide (HDTMAB) with a purity of 99%, provided by YuanJu Chem. Co. Ltd., China.
2.2 Preparation of organoclays Before synthesis of HDTMA+ intercalated montmorillonites, sodium montmorillonite (Na-Mt) was prepared from Ca-Mt as follows: 10 g of the mixture of
Ca-Mt and Na2CO3 in the ratio of 94:6 was added into 100 ml of deionized water and stirred at 80 °C for 3 h. During the stirring, several drops of HCl were added into the 2- suspension to dissolve the CO3 . Na-Mt was collected by centrifugation and washed with deionized water until the solution was free of chloride (titration with AgNO3). The Na-Mt was dried at 105 °C, ground through a 200 mesh sieve and kept in a sealed bottle. The syntheses of HDTMA+ intercalated montmorillonites were performed by the following procedure: 2.5 g of Na-montmorillonite was first dispersed in 300 ml of deionized water and then a desired amount of HDTMAB was slowly added. The concentrations of HDTMA+ varied from 0.5 CEC to 2.5 CEC of montmorillonite. The reaction mixtures were stirred in a water bath for 9 h at 80 oC. All products were o washed free of bromide anions (titration with AgNO3), dried at 60 C and ground in an agate mortar to pass through a 200 mesh sieve. The HDTMA+ modified montmorillonite prepared at the concentration of 0.5 CEC was denoted as 0.5CEC-Mt and the others were marked in the same way.
2.3 Characterization X-ray diffraction (XRD) patterns of the samples were recorded between 1.5 and 20° (2θ) at a scanning speed of 2°/min, using Rigaku D/max-1200 diffractometer with Cu Kα radiation (30 mA and 40 kV). High-resolution thermogravimetric analysis (HRTG) was performed on a TA Instruments Inc. Q500 thermobalance. Samples were heated from room temperature o to 1000 °C at a heating rate of 10 °C/min with a resolution of 6 C under N2 atmosphere (80 cm3/min). Approximately 30 mg of finely ground sample was heated in an open platinum crucible.
N2 adsorption-desorption isotherms were gained at liquid nitrogen temperature
225 with a Micromeritics ASAP 2010 gas sorption analyzer (Quantachrome Co., USA). o Before measurement, the samples were pre-heated at 80 C under N2 for ca. 24 h. The specific surface area was calculated by the BET equation and the total pore volumes were evaluated from nitrogen uptake at relative pressure of ca. 0.99. The X-ray photoelectron spectroscopy (XPS) analyses were performed on a Kratos AXIS Ultra with a monochromatic Al X-ray source at 150 W. Each analysis started with a survey scan from 0 to 1200 eV with a dwell time of 100 ms, pass energy of 160 eV at steps of 1 eV with 1 sweep. For the high-resolution analysis, the number of sweeps was increased, the pass energy was lowered to 20 eV at steps of 100 meV and the dwell time was changed to 250 ms. Band component analyses were undertaken using using the Jandel ‘Peakfit’ software package, which enabled the type of fitting function to be selected and allows specific parameters to be fixed or varied accordingly. Band fitting was done using a Lorenz-Gauss cross-product function with the minimum number of component bands used for the fitting process [33].
3. Results and discussion 3.1 X-ray diffraction (XRD) Figure 1 shows the XRD patterns of montmorillonites and the resultant organoclays. The basal spacing of Na-Mt is 1.24 nm, which is a characteristic d value for Na-montmorillonite. However, after modification with surfactant, the interlayer height of montmorillonite is obviously increased. With an increase of surfactant concentration in the preparation solution, the basal spacings of the resultant organoclays increase in the following order: 1.24 nm (Na-Mt) → 1.48 nm (0.5CEC-Mt) → 1.78 nm (0.7CEC-Mt) → 1.95 nm (1.0CEC-Mt) → 2.23 nm (1.5CEC-Mt) → 3.61 nm (2.0CEC-Mt) → 3.84 nm (2.5CEC-Mt). Here, it can be understood that there are five different HDTMA+ arrangements adopted within the montmorillonite interlayer space, i.e., lateral monolayer in 0.5CEC-Mt, lateral bilayer in 0.7CEC-Mt, pseudotrilayer in 1.0CEC-Mt, paraffin monolayer in 1.5CEC-Mt and paraffin bilayer in 2.0CEC-Mt and 2.5CEC-Mt, respectively, in agreement with previous experimental and molecular modeling reports [11,12,34-36]. Figure 2(I) is the schematics of the organoclays with different surfactant arrangement models.
3.2 XPS characterization Figure 3 displays the XPS survey scans of HDTMAB, Na-Mt and the representative organoclays (0.7CEC-Mt and 2.5CEC-Mt). The XPS results clearly show that the presences of carbon, nitrogen and bromine in HDTMAB and sodium, aluminum, silicon, oxygen, magnesium and iron in Na-Mt. The XPS result is in an excellent agreement with our chemical analysis result of montmorillonite. In addition, there is a minor amount of oxygen in HDTMAB and carbon in Na-Mt, resulting from adsorbed CO2 [37]. The XPS survey scans show the presence of calcium in Ca-Mt (not shown) whereas in Na-Mt only sodium was observed (Fig. 3), indicating that the preparation of sodium montmorillonite from calcium montmorillonite in this study was successful. The ratios of the elemental atomic concentrations in montmorillonite and the
226 resultant organoclays are summarized in Table 1, deduced from the corresponding XPS analyses. This calculation shows that the content of surfactant within the organoclays increases in the order of 0.5CEC-Mt → 0.7CEC-Mt → 1.0CEC-Mt → 1.5CEC-Mt → 2.0CEC-Mt → 2.5CEC-Mt. This is in accordance with our previous studies using other techniques and the reports in the literature [11-25]. Meanwhile, the Al/Si ratio deduced from the XPS analysis (0.36) is in good agreement with the chemical analysis (0.37). However, the Al/Si ratio in the resultant organoclays decreases with the intercalation of surfactant as shown in Table 1. This results from the increase of the interlayer distance with the intercalation of surfactant, which leads to a decreasing possibility for detecting the Al-O(OH) octahedral sheets sandwiched between the two Si-O tetrahedral sheets of the montmorillonite. In the XPS survey scans of the resultant organoclays, prominent peaks corresponding to magnesium and a trace to iron are always recorded whereas that of sodium disappears. This reflects that both magnesium and iron are in the montmorillonite structure rather than in the interlayer. This is in agreement with our formula calculation. Meanwhile, the disappearance of the peak corresponding to sodium results from the exchange of sodium ions by surfactant cations. There is no peak corresponding to bromine recorded in the XPS survey scans of 0.5CEC-Mt, 0.7CEC-Mt and 1.0CEC-Mt whereas it is recorded in the XPS scans of 1.5CEC-Mt, 2.0CEC-Mt and 2.5CEC-Mt. The surfactant contents (in term of CEC) in the resultant organoclays, deduced from the thermogravimetric measurements (not shown), are shown in Table 1. Our calculation indicates that there is more than one CEC of surfactant in 1.5CEC-Mt, 2.0CEC-Mt and 2.5CEC-Mt. Here, it can be understood that, at relative low surfactant concentration (≤ 1 CEC in the present case), the intercalation is dominant and the surfactants enter into the clay interlayer space as cations. On the other hand, the surfactants exist within the organoclays in both formats of cations and molecules when the loaded surfactants are more than 1 CEC [38]. This concept is further supported by the high-resolution XPS scans.
3.3 High-resolution XPS 3.3.1 C 1s spectra The C 1s spectrum of HDTMAB is characterized by two transitions centered at 284.7 and 285.7 eV, corresponding to the C-C bond in the long chain and C-N, respectively (Fig. 4). The C 1s spectra of organoclays show a significant broadening with slight changes in binding energy, indicating more than one type of surfactant-clay interaction. The change trends of binding energy for C-C and C-N as function of CEC are different as shown in Figure 5. For the spectra corresponding to C-C, there is a significant binding energy decrease from HDTMAB to 0.5CEC-Mt and the binding energies for 0.7CEC-Mt and 1.0CEC-Mt are similar to that of 0.5CEC-Mt. However, there is a significant increase from 1.0CEC-Mt to 1.5CEC-Mt, then to 2.0CEC-Mt and finally to 2.5CEC-Mt. The C 1s binding energies of 2.0CEC-Mt and 2.5CEC-Mt are higher than that of HDTMAB. However, the C 1s binding energies of the resultant organoclays, corresponding to the C-N bond, are similar. Here, the C 1s spectra of organoclays show that the local molecular
227 environment of the surfactant has a prominent effect on the binding energy. Molecular modeling [34-36] demonstrates that in all arrangements of surfactant within the clay interlayer, the headgroups (nitrogen) of the alkyl chains will be close to the clay surface due to the strong electrostatic interaction between the negative clay surface and the positive headgroups of the alkyl chains. This can well explain the similar binding energy of C 1s in C-N for the organoclays with different surfactant arrangements. Previous reports [13-16,34-36] have shown that, in the organoclays with lower surfactant packing density, the alkyl chains within the interlayer space are parallel within the interlayer space and are individually separated. In this case, the repulsive interaction between the hydrocarbon chain ─ silicate surface is dominant whereas the interaction among the hydrocarbon chains is very weak. The local environment of the intercalated surfactant is absolutely different from those in bulk state. With the increase of surfactant packing density, the interchain interaction among the surfactants becomes the dominant force and the orientation of the hydrocarbon tail changes from parallel to the silicate surface within the interlayer space to parallel but at an angle to the silicate surface as shown by XRD and FTIR results [11-15]. The interaction among alkyl chains will increase with the increase of the surfactant packing density [11,12,21] and this will result in the ordered packing of the alkyl chains as indicated by FTIR and Raman spectroscopy [13-16]. The local environment of the surfactant within the resultant organoclays strongly depends on their loaded amounts, resulting in a variation of the C 1s binding energy associated with the C-C bond in the alkyl chains.
3.3.2 N 1s spectra The high resolution scans of nitrogen in HDTMAB and the representative organoclays are displayed in Figure 4. For HDTMAB, a single 1s transition is observed with a binding energy of 401.9 eV, which is similar to that reported in a previous study [39]. The high resolution scans of nitrogen in 0.5CEC-Mt, 0.7CEC-Mt and 1.0CEC-Mt show a single 1s transition with a slight increase of the binding energy (ca. 0.5 eV) and full-width-at-half-maximum (FWHM), indicating the local environment of the intercalated surfactant is different from that in bulk state. This is in accordance with the conclusion deduced from the C 1s spectra. However, the nitrogen high resolution scans of organoclays with a surfactant loading more than one CEC show two overlapping bands related to two different N 1s transitions. The N 1s spectrum of 2.5CEC-Mt is shown in Figure 4, in which two bands at ca. 403.6 and 402.6 eV, respectively, were recorded, reflecting two different local environments for the surfactant within the organoclay. The band with a bonding energy (402.6 eV) similar to that in 0.5CEC-Mt, 0.7CEC-Mt and 1.0CEC-Mt, should be attributed to the intercalated surfactant while the other one (ca. 403.6 eV) should be attributed to the surfactant outside the clay layers. This assumption is supported by the pore volume analyses of these samples. The nitrogen adsorption-desorption isotherms of Na-Mt and the resultant organoclays show that there are “ink-bottle” like pores in these clays, which could be
228 described as a “house of cards” structure [40,41]. Meanwhile, the pore volumes of these samples (Table 1) show that there is a dramatic decrease of the BET-N2 surface area from Na-Mt to 0.5CEC-Mt, followed by a smooth decrease till 1.0CEC-Mt, then a more pronounced decrease till 2.0CEC-Mt. The dramatic pore volume decrease of the organoclays with less than one CEC can be explained as the intercalation of the HDTMA+ cations into the clay interlayer space. And the significant pore volume decrease from 1.0CEC-Mt to the organoclays with a surfactant loading more than one CEC is resulted from the occupation of surfactant in the interparticle pores. This also has been elucidated by thermal analysis of organoclays [21].
3.3.3 Br 3d spectra As indicated by the survey scans of the resultant organoclays (Fig. 3), a very weak Br 3d transition begins to occur in the XPS spectrum of 1.5CEC-Mt. The intensity of the Br 3d transition obviously increases in the spectra of 2.0CEC-Mt and 2.5CEC-Mt. Figure 6 displays the high resolution Br 3d scans of HDTMAB, 2.0CEC-Mt and 2.5CEC-Mt. The Br 3d spectrum of HDTMAB displays two well-resolved transitions centered at 67.1 and 68.2 eV, corresponding to Br 3d5 and 3d3, respectively. In comparison to the XPS spectrum of HDTMAB, both 2.0CEC-Mt and 2.5CEC-Mt show a broad peak with low intensity and poor resolution for the two transitions. This reflects that the content of bromine in the organoclays is limited and disordered, and ion exchange between HDTMA+ and interlayer cations (Na+) is dominant [38].
3.3.4 O 1s and Si 2p spectra The high-resolution O 1s scan of Na-Mt and the simulated curves (Fig. 7a) show that it is difficult to distinguish O and OH in montmorillonite. This is different from the previous study about basic aluminum sulphate and basic aluminum nitrate, in which O and OH in the corresponding materials were clearly identified [33]. There is a small amount of water (ca. 2.39%) remaining in Na-Mt after exposure to ultra high vacuum (10-9 – 10-10 Torr), as shown by the simulated curves (Fig. 7a). The oxygen in motmorillonite structural sheets corresponds to a binding energy of 532.1 eV while that in water is ca. 535.0 eV. These values are in accord with those reported in the literature [33,42]. However, the high-resolution O 1s scans of the resultant organoclays (Fig. 7) do not show any transition corresponding to H2O, resulting from the hydrophobicity of the organoclays and high vacuum (10-9 Torr) in the detection chamber. Compared to Na-Mt, it can be seen that there is a slight decrease (ca. 1 eV) of the O 1s binding energy in the organoclays. The binding energy change of Si 2p transition from Na-Mt to the resultant organoclays is similar to that of O 1s (Fig. 7 d-f). The binding energy of Si 2p in Na-Mt is 103.0 eV while that for the resultant organoclays increases to 101.9 eV, with a decrease of 1.1 eV. Both the decreases of O 1s and Si 2p binding energies result from the change of the interlayer environment. This is in agreement with the proposal deduced from MAS NMR study of organoclays, which indicates that modifying clays with surfactant results in a measurable shielding of 29Si nuclei in clays [43].
229 During the intercalation, the prominent change for montmorillonite is that the exchangeable interlayer hydrated cations are replaced by the surfactant cations. As shown by the O 1s XPS spectrum (Fig. 7a), there is a minor amount of H2O in Na-Mt, corresponding to the strongly bound water of the interlayer cations rather than the surface adsorbed water [44,45]. This part of water links to the oxygen on the clay surface (Si-O tetrahedral sheet) through hydrogen bond [46,47]. After the interlayer hydrated cations are replaced by the intercalated surfactant, the main interaction between the clay and surfactant includes both the electrostatic attract between the positively charged headgroups (nitrogen) of the alkyl chains and the negatively charged clay surfaces, and a repulsive force between alkyl chain and clay surface as demonstrated by molecular modeling [34-36]. On the basis of abovementioned experimental results, the schematics for the structural evolution from Na-Mt to the resultant organoclays are built as shown in Figure 7(II). Obviously, two basic organoclay types are formed when modifying clay with surfactant due to the different surfactant distributions: 1) the surfactant mainly occupies the clay interlayer and 2) both the clay interlayer space and external surface are modified by surfactant. Our recent sorption experiments indicate that both surface sorption and partition are involved in the sorption mechanisms for 0.5CEC-Mt, 0.7CEC-Mt and 1.0CEC-Mt to p-nitrophenol whereas partition is dominant for 1.5CEC-Mt, 2.0CEC-Mt and 2.5CEC-Mt. This is a convincing evidence supporting our assumption of the surfactant distribution within organoclays.
4. Conclusions In this study, a series of organoclays with different surfactant arrangements within the clay interlayer were prepared. The surfactant distribution within the resultant organoclays was investigated by XPS in combination with XRD and HRTG. In the organoclays prepared at relative low surfactant concentrations (< 1.0CEC), the surfactant cations mainly occupy the clay interlayer with lateral arrangements (lateral monolayer, lateral bilayer and pseudotrilayer). However, when the surfactant concentrations are higher than 1.0CEC, the surfactants occupy both the clay interlayer space and the interparticle pores and paraffin type arrangements of surfactants (paraffin monolayer and paraffin bilayer) are adopted in the clay interlayer spaces. This gives excellent explanations about the dramatic surface area and pore volume decrease of organoclays and different sorption mechanisms (surface sorption and partition) involved in organoclay sorption experiments as reported in the literature. XPS survey scans show that the peaks corresponding to magnesium and iron are identical in all samples, reflecting these atoms in montmorillonite structure rather in the interlayer or impurities. The peaks corresponding to bromine only appear in the organoclays prepared at high surfactant concentrations (> 1.0CEC). This suggests that both surfactant cations and ion pairs occur in these organoclays, corresponding different interactions between surfactants and clays. The former relates with ion exchange and the latter with sorption. Generally, modifying clays with surfactants results in a decrease of binding energy of atoms in both clays and surfactants and broadening of the peaks. However,
230 with the increase of surfactant loadings, the FWHM of the related peaks decreases, suggesting the organoclay structure becomes more ordered. This study shows that nitrogen is the most sensitive to the surfactant distribution within the resultant organoclays and the C 1s binding energy of C-C bond in alkyl chain is sensitive to the local environment of surfactants in organoclays with different arrangements.
Acknowledgments The financial and infra-structural support from the National Natural Science Foundation of China (Grant No. 40372029 and International Cooperation Research Program), and the Inorganic Materials Research Program, Queensland University of Technology are gratefully acknowledged.
231 References (1) M.R. Stackmeyer, Appl. Clay Sci. 6 (1991) 39. (2) L.P. Meier, R. Nueesch, F.T. Madsen, J. Colloid Interface Sci. 238 (2001) 24. (3) L.Z. Zhu, B.L. Chen, Environ. Sci. Technol. 34 (2000) 2997. (4) L.Z. Zhu, B.L. Chen, X.Y. Shen, Environ. Sci. Technol. 34 (2000) 468. (5) E. Manias, G. Hadziioannou, G. Brinke, Langmuir 12 (1996) 4587. (6) Z. Wang, T.J. Pinnavaia, Chem. Mater. 10 (1998) 3769. (7) S.S. Ray, M. Okamoto, Prog. Polym. Sci. 28 (2003) 1539. (8) M. Nakatsuji, R. Ishii, Z.M. Wang, K. Ooi, J. Colloid Interface Sci. 272 (2004) 158. (9) R. Ishii, M. Nakatsuji, K. Ooi, Micropor. Mesopor. Mater. 79 (2005) 111. (10) F. Bergaya, G. Lagaly, Appl. Clay Sci. 19 (2001) 1. (11) G. Lagaly, Clay Miner. 16 (1981) 1. (12) J.X. Zhu, H.P. He, J.G. Guo, D. Yang, X.D. Xie, Chinese Sci. Bull. 48 (2003) 368. (13) R.A. Vaia, R.K. Teukolsky, E.P. Giannelis, Chem. Mater. 6 (1994) 1017. (14) Y.Q. Li, H. Ishida, Langmuir 19 (2003) 2479. (15) H.P. He, R.L. Frost, J.X. Zhu, Spectrochim Acta A 60 (2004) 2853. (16) H.P. He, R.L. Frost, Y.F. Xi, J.X. Zhu, J. Raman Spectros. 35 (2004) 316. (17) Y.Q. Li, H. Ishida, Chem. Mater. 14 (2002) 1398. (18) W. Xie, Z.M. Gao, W.P. Pan, D. Hunter, A. Singh, R.A. Vaia, Chem. Mater. 13 (2001) 2979. (19) W. Xie, R.C. Xie, W.P. Pan, D. Hunter, B. Koene, L.S. Tan, R.A. Vaia, Chem. Mater. 14 (2002) 4837. (20) S. Yariv, Appl. Clay Sci. 24 (2004) 225. (21) H.P. He, Z. Ding, J.X. Zhu, P. Yuan, Y.F. Xi, D. Yang, R.L. Frost, Clays Clay Miner. 53 (2005) 287. (22) H.P. He, R.L. Frost, F. Deng, J.X. Zhu, X.Y. Weng, P. Yuan, Clays Clay Miner. 52 (2004) 350. (23) J.X. Zhu, H.P. He, L.Z. Zhu, X.Y. Weng, F. Deng, J. Colloid Interface Sci. 286 (2005) 239. (24) S.Y. Lee, S.J. Kim, Colloid Surface A 211 (2002) 19. (25) S.Y. Lee, S.J. Kim, Clays Clay Miner. 50 (2002) 435. (26) Y.M. Dana, C. Yona, N. Shlomo, P. Rene, Environ. Sci. Technol.29 (2005) 1231. (27) T.N. Wittberg, P.S. Wang, Surf. Interface Anal. 27 (1999) 936. (28) G. Parthasarathy, B.M. Choudary, B. Sreedhar, A.C. Kunwar, R. Srinivasa, Am. Mineral. 88 (2003) 1983. (29) J.X. Du, J. Zhu, C.A. Wilkie, J.Q. Wang, Polym. Degrad. Stabil. 77 (2002) 377. (30) M. Bueno, A. Astruc, J. Lambert, M. Astruc, P. Behra, Environ. Sci. Technol. 35 (2001) 1411. (31) H.P. He, J. Duchet, J. Galy, J.F. Gérard, J. Colloid Interface Sci. 295 (2006) 202. (32) H.P. He, J.G. Guo, X.D. Xie, J.L. Pen, Environ. Int. 26 (2001) 347. (33) L.V. Duong, B.J. Wood, J.T. Kloprogge, Mater. Lett. 59 (2005) 1932. (34) P. Čapková, J.V. Burda, Z. Weiss, H. Schenk, J. Mol. Model. 5 (1999) 8. (35) Q.H. Zeng, A.B. Yu, G.Q. Lu, R.K. Standish, J. Phys. Chem. B 108 (2004) 10025. (36) H.P. He, J. Galy, J.F. Gerard, J. Phys. Chem. B. 109 (2005) 13301. (37) L. Bugyi, A. Oszkó, F. Solymosi, Surf. Sci. 461 (2000) 177. (38) Z. Klapyta, T. Fujita, N. Iyi, Appl. Clay Sci. 19 (2001) 5. (39) A.N. Buckley, M.D. Kelly, P.F. Nelson, K.W. Riley, Fuel Process. Technol. 43 (1995) 47. (40) G. Lagaly, S. Ziesmer, Adv. Colloid Interf. 100 (2003) 105. (41) D. Burgentzlé, J. Duchet, J.F. Gérard, A. Jupin, B. Fillon, J. Colloid Interface Sci. 278 (2004)
232 26. (42) T.L. Barr, M.A. Lishka, J. Am. Chem. Soc. 108 (1986) 3178. (43) I.K. Breakwell, J. Homer, M.A.M. Lawrence, W.R. Mcwhinnie, Polyhedron, 14 (1995) 2511. (44) A. F. Koster van Groos, S. Guggenheim, Am. Mineral. 72 (1987) 292. (45) Y.F. Xi, R.L. Frost, H.P. He, J.T. Kloprogge, T. Bostrom, Langmuir 21 (2005) 8675. (46) L. Yan, C.B. Roth, P.F. Low, Langmuir 12 (1996) 4421. (47) L. Yan, P.F. Low, C.B. Roth, Clay Clay Miner. 44 (1996) 749.
233
Table 1 Surfactant loadings and the ratios of atomic concentrations in montmorillonite and the resultant organoclays based on TG and XPS analyses.
Sample Na-Mt 0.5CEC-Mt 0.7CEC-Mt 1.0CEC-Mt 1.5CEC-Mt 2.0CEC-Mt 2.5CEC-Mt a SL (%) - 9.73 16.73 22.13 28.19 38.73 44.17 b SL (vs CEC) - 0.33 0.61 0.86 1.19 1.9 2.4 c C/Si - 2.08 2.28 2.69 3.41 5.79 6.58 d Al/Si 0.36 0.34 0.32 0.33 0.33 0.28 0.30 3 VP (cm /g) 0.107 0.061 0.060 0.056 0.037 0.011 0.007 a: surfactant loading within the corresponding organoclay, evaluated from high-resolution thermogravimetric analysis. b: surfactant loading expressed in CEC of montmorillonite (100 g). c: the ratios of carbon and silicon atomic concentrations in the organoclays. d: the ratios of aluminun and silicon atomic concentrations in Na-montmorillonite and the organoclays..
VP: pore volume determined by BJH method from N2 desorption isotherm.
234
Figure 1 XRD patterns of montmorillonite and the resultant organoclays.
235
Figure 2 The schematics of organoclays with different arrangements (I), and the schematics of Na-Mt and the resultant organoclays (II). A: lateral monolayer; B: lateral bilayer; C: pseudotrilayer; D: paraffin monolayer; E: paraffin bilayer.
236
Figure 3 XPS survey scans of HDTMAB, Na-Mt and the representative organoclays.
237
Figure 4 C 1s and N 1s high resolution XPS spectra of HDTMAB and the representative organoclays. thin solid line: experimental curve; thick solid line: deconvolution curve; dash line: fitting curve.
238
Figure 5 The C 1s binding energy change of C-C and C-N in the resultant organoclays.
239
Figure 6 Br 3d high resolution XPS spectra of HDTMAB, 2.0CEC-Mt and 2.5CEC-Mt. thin solid line: experimental curve; thick solid line: deconvolution curve; dash line: fitting curve.
240
Figure 7 O 1s and Si 2p high resolution XPS spectra of Na-Mt and the representative organoclays. thin solid line: experimental curve; thick solid line: deconvolution curve; dash line: fitting curve.
241