SYNTHESIS AND CHARACTERISATION OF MULTI-WALLED CARBON
NANOTUBES ON SUPPORTED CATALYSTS VIA CATALYTIC
CHEMICAL VAPOUR DEPOSITION
TEE JIA CHEE
UNIVERSITI TEKNOLOGI MALAYSIA
iii
Beloved mum and dad,
just for you.
iv
ACKNOWLEDGEMENT
In this column of acknowledgements, I would particularly like to express my
sincere appreciation and gratitude to both of my supervisors, Assoc. Prof. Dr. Nor
Aziah Buang and Prof. Dr. Ahmad Fauzi Ismail for providing me noble guidance and
valuable advices throughout the period of this master programme.
Besides that, I am also grateful to Mr. Azmi, Mr. Jaafar, Madam Noor
Khaidawati (Faculty of Science) and Madam Suhaila Mohd. Sanip, Mr. Sohaimi, Mr.
Ng Be Cheer (Membrane Research Unit, Faculty of Chemical Engineering and
Natural Resources Engineering) as well as Mr. Jefri, Mr. Ayub and Mr. Zainal
Abidin (Faculty of Mechanical Engineering). They had put considerable time and
effort in assisting me doing my laboratory works. Without them, I might not be able
to complete all the tests so successfully.
Appreciation is also acknowledged to my friends Norhuda, Zatur and Yusran
for their moral support and concerns and to everyone who has contributing to the
success of this study, either directly or indirectly.
Last but foremost, I wish to show my gratitude to my parents and my brother
for their care and support. Without my parents’ love and encouragement throughout
my life I will not able to be as present. Finally, I wish to thank my dear Jerry, for all
the love and happiness he brought to me. I cannot remember how many times he
came to be with me and sit by my side while I was doing my works. I thank God for
bringing him into my life.
v
ABSTRACT
The main hindrance to employ carbon nanotubes (CNTs) commercially is the
inability to control the growth of the nanotubes and to grow bulk amounts of carbon
nanotubes. However, recently the Chemical Vapour Deposition (CVD) has been
modified by applying various supported metals catalysts in the production of CNTs.
Therefore, in this research we focus on the effects of supported catalysts in the
synthesis of CNTs via Catalytic Chemical Vapour Deposition (CCVD) method. The
CCVD method was used to synthesize high quality CNTs in high yield and
economical cost with controlling of the CNTs characteristics and morphologies. A
practical and high performance CCVD system has been designed and built. The fixed
bed flow reactor in the CCVD system is specifically fabricated to carry out the
pyrolysis of hydrocarbon to produce CNTs. The supported catalysts of cobalt (Co),
iron (Fe) and mixture of these metals (Co/Fe) were prepared by using the alumina
(Al2O3), molecular sieves (MS) and anodic aluminium oxide (AAO) template as
supports. All supported catalysts were prepared by impregnation method. The asprepared supported catalysts were subjected to calcination at 450 °C. The catalysts
were characterised using X-Ray Diffraction (XRD) technique. Acetylene (C2H2) was
selected as the carbon precursor and the reaction was performed at 700 °C for 30
minutes. The yields of the reaction collected as black depositions on the catalysts.
The characterisations of the yield were carried out using Scanning Electron
Microscopy (SEM), Field-Emission Scanning Electron Microscopy (FE-SEM) and
Transmission Electron Microscopy (TEM) as well as Energy Dispersive X-Ray
Analysis (EDAX) techniques. Catalysts prepared were active in the production of
CNTs. The most active catalysts were identified as Al-Co/Fe(3.0)Cal,
AAO-Co/Fe(1.0)Cal and MS-Co/Fe(3.0)Cal as they generated high carbon contents
of 72.00, 64.03 and 48.50 wt.% respectively. The as-grown CNTs over various
catalysts showed high degree of graphitisation, purity and density with
configurations of bundles, arrays and coils. The CNTs yields were classified as
multi-walled carbon nanotubes (MWNTs). The best MWNT consists of 11 layers of
turbostratic graphene wall with inner diameter of 3.57 nm and outer diameter of
11.43 nm as well as distance between layers of 0.33 nm. The CNTs grown over
Al2O3 supported catalysts followed the tip growth mechanism whereas the CNTs
grown over MS supported catalysts followed the base growth mechanism.
vi
ABSTRAK
Halangan utama dalam menggunakan nanotiub karbon (CNTs) secara
komersial adalah ketidakupayaan untuk mengawal pertumbuhan nanotiub karbon dan
tidak dapat menghasilkan nanotiub karbon dalam jumlah yang banyak. Walau
bagaimanapun, kebelakangan ini kaedah Pemendapan Wap Kimia (CVD) telah
diubahsuai dengan menggunakan pelbagai jenis mangkin logam berpenyokong
dalam penghasilan CNTs. Oleh itu, kajian ini menfokuskan kepada kesan mangkin
berpenyokong ke atas sintesis CNTs melalui teknik Pemendapan Wap Kimia
Bermangkin (CCVD). Teknik CCVD ini digunakan untuk mensintesis CNTs
berkualiti tinggi dengan hasil yang tinggi, kos yang ekonomik dan dapat mangawal
ciri-ciri dan morfologi CNTs terhasil. Sistem CCVD yang praktikal dan
berkeupayaan tinggi telah direka bentuk dan dibina. Reaktor pengaliran dasar tetap
dalam CCVD sistem ini dibina khas untuk proses pirolisis hidrokarbon untuk
menghasilkan CNTs. Mangkin berpenyokong jenis kobalt (Co), ferum (Fe) and
campuran kedua-dua logam ini (Co/Fe) telah disediakan dengan menggunakan
alumina (Al2O3), penapis molekul (MS) dan templat anodik aluminium oksida
(AAO) sebagai penyokong. Kesemua mangkin berpenyokong telah disediakan
dengan kaedah pengisitepuan. Mangkin berpenyokong tersedia telah dikalsinkan
pada suhu 450 °C. Semua mangkin berpenyokong telah dicirikan dengan teknik
Pembelauan Sinar-X (XRD). Asetilena (C2H2) telah dipilih sebagai bahan asas
karbon. Tindak balas telah dijalankan pada suhu 700 °C selama 30 minit dan hasil
daripada tindak balas dikumpul sebagai serbuk hitam yang termendap atas mangkin.
Pencirian hasil ini telah dilakukan dengan menggunakan teknik Mikroskopi Imbasan
Elektron (SEM), Mikroskopi Imbasan Elektron Pemancaran Medan (FE-SEM) dan
Mikroskopi Transmisi Elektron (TEM) serta Analisis Penyerakan Tenaga Sinar-X
(EDAX). Mangkin berpenyokong yang disediakan adalah aktif dalam penghasilan
CNTs. Mangkin yang teraktif dikenalpasti sebagai Al-Co/Fe(3.0)Cal,
AAO-Co/Fe(1.0)Cal dan MS-Co/Fe(3.0)Cal kerana ia menghasilkan kandungan
karbon yang tinggi, iaitu masing-masing 72.00, 64.03 and 48.50 % berat. CNTs
tersedia atas pelbagai jenis mangkin menunjukan darjah grafitasi, ketulinan dan
kepadatan yang tinggi dengan konfigurasi jenis gumpalan, teratur and lingkaran.
Hasil CNTs itu telah diklasifikasikan sebagai nanotiub karbon dinding berganda
(MWNTs). MWNTs yang terbaik terdiri daripada 11 lapis dinding grafin turbostratik
dengan diameter dalaman 3.57 nm dan diameter luaran 11.43 nm serta jarak antara
lapisan 0.33 nm. Pertumbuhan CNTs bagi mangkin berpenyokong Al2O3 adalah
melalui mekanisma pertumbuhan hujung, manakala mangkin berpenyokong MS
melalui mekanisma pertumbuhan pangkal.
vii
TABLE OF CONTENTS
CHAPTER
TITLE
PAGE
TITLE PAGE
i
DECLARATION
ii
DEDICATION
iii
ACKNOWLEDGEMENT
iv
ABSTRACT
v
ABSTRAK
vi
TABLE OF CONTENTS
vii
LIST OF TABLES
xiii
LIST OF FIGURES
xv
LIST OF ABBREVIATIONS
xxi
LIST OF APPENDICES
xxii
INTRODUCTION
1
1.1
Research Background
1
1.2
Problem Statement
2
1.3
Research Objectives
3
1.4
Scopes of Study
4
CARBON AND CARBON NANOTUBES
5
Carbon and Carbon Structures
5
1
2
2.1
2.1.1 Carbon and Bonding between Carbon
Atoms
2.1.2 Graphite
5
7
viii
2.2
3
2.1.3 Diamond
8
2.1.4 Fullerenes
10
2.1.5 Carbon Fibers
12
Carbon Nanotubes
13
2.2.1 Structure of Carbon Nanotubes
13
2.2.2 Properties of Carbon Nanotubes
16
2.2.3 Common Synthesis Methods
18
2.2.4 Growth Mechanism of Carbon Nanotubes
22
2.2.5 Applications of Carbon Nanotubes
23
CATALYTIC CHEMICAL VAPOUR
DEPOSITION (CCVD)
3.1
26
An Overview of Catalytic Chemical Vapour
Deposition (CCVD) Method
26
3.2
Metal Oxides Supported Catalysts
27
3.3
Anodic Aluminium Oxide (AAO) Template
Supported Catalysts
37
3.4
Types of Carbon Precursor
40
3.5
Reaction Temperature
40
3.6
Modified Catalytic Chemical Vapour Deposition
4
4.1
(CCVD)
42
RESEARCH METHODOLOGY
44
Preparation of Supported Catalysts
44
4.1.1 Preparation of Alumina and Molecular
Supported Catalysts
44
4.1.2 Preparation of Anodic Aluminium Oxide
(AAO) Template Supported Catalysts
4.2
Synthesis of Carbon Nanotubes
45
47
4.2.1 Design and Fabrication of Custom Built
Catalytic Chemical Vapour Deposition
(CCVD) System
47
ix
4.3
4.2.2 Production of Carbon Nanotubes
49
Characterisation Techniques
50
4.3.1 X-Ray Diffraction (XRD)
50
4.3.1.1 Particle Size Measurement
4.3.2 Electron Microscopy
51
53
4.3.2.1 Scanning Electron Microscopy
(SEM)
53
4.3.2.2 Field-Emission Scanning Electron
Microscopy (FE-SEM)
55
4.3.2.3 Transmission Electron Microscopy
(TEM)
55
4.3.3 Energy Dispersive X-Ray Analysis (EDAX) 57
5
PRODUCTION OF CARBON NANOTUBES
(CNTs) OVER ALUMINA AND ANODIC
ALUMINIUM OXIDE (AAO) SUPPORTED
5.1
CATALYSTS
60
Alumina Supported Catalysts
60
5.1.1 Physical Appearance of Alumina Supported
Catalysts
5.1.2 X-Ray Diffraction (XRD) Analysis
61
62
5.1.2.1 XRD Analysis of Alumina
Supported Cobalt Catalysts
62
5.1.2.2 XRD Analysis of Alumina
Supported Ferrum Catalysts
67
5.1.2.3 XRD Analysis of Alumina
Supported Cobalt/Ferrum Catalysts 71
5.1.2.4 Particle Size of Alumina Supported
Catalysts
5.1.2.5 Conclusion of XRD Analysis
73
75
5.1.3 As-Grown Carbon Deposition Yield and
Physical Appearance
75
x
5.1.4 Scanning Electron Microscopy (SEM)
Analysis
78
5.1.4.1 SEM Analysis of Alumina
Supported Cobalt Catalysts and the
CNTs Yield
78
5.1.4.2 SEM Analysis of Alumina
Supported Ferrum Catalysts and the
CNTs Yield
81
5.1.4.3 SEM Analysis of Alumina
Supported Cobalt/Ferrum Catalysts
and the CNTs Yield
5.1.4.4 Conclusion of SEM Analysis
84
88
5.1.5 Field Emission-Scanning Electron
Microscopy (FE-SEM) Analysis
89
5.1.6 Transmission Electron Microscopy (TEM)
Analysis
81
5.1.7 Energy Dispersive X-Ray Analysis (EDX) 95
5.2
Anodic Aluminium Oxide (AAO) Supported
Catalysts
98
5.2.1 Physical Appearance of AAO Supported
Catalysts
5.2.2 X-Ray Diffraction (XRD) Analysis
98
99
5.2.3 As-Grown Carbon Deposition Yield and
Physical Appearance
99
5.2.4 Scanning Electron Microscopy (SEM)
Analysis
101
5.2.4.1 SEM Analysis of AAO Supported
Cobalt Catalysts and the CNTs
Yield
101
5.2.4.2 SEM Analysis of AAO Supported
Ferrum Catalysts and the CNTs
Yield
105
xi
5.2.4.3 SEM Analysis of AAO Supported
Cobalt/Ferrum Catalysts and the
CNTs Yield
5.2.4.4 Conclusion of SEM Analysis
106
109
5.2.5 Energy Dispersive X-Ray Analysis (EDX) 110
6
PRODUCTION OF CARBON NANOTUBES
(CNTs) OVER MOLECULAR SIEVES
SUPPORTED CATALYSTS
113
6.1
Molecular Sieves Supported Catalysts
113
6.2
Physical Appearance of Molecular Sieves
6.3
Supported Catalysts
114
X-Ray Diffraction (XRD) Analysis
115
6.3.1 XRD Analysis of Molecular Sieves
Supported Cobalt Catalysts
115
6.3.2 XRD Analysis of Molecular Sieves
Supported Ferrum Catalysts
120
6.3.3 XRD Analysis of Molecular Sieves
Supported Cobalt/Ferrum Catalysts
125
6.3.4 Particle Size of Alumina Supported
Catalysts
6.3.5 Conclusion of XRD Analysis
6.4
6.5
126
128
As-Grown Carbon Deposition Yield and
Physical Appearance
129
Scanning Electron Microscopy (SEM) Analysis
131
6.5.1 SEM Analysis of Molecular Sieves
Supported Cobalt Catalysts and the CNTs
Yield
131
6.5.2 SEM Analysis of Molecular Sieves
Supported Ferrum Catalysts and the CNTs
Yield
134
xii
6.5.3 SEM Analysis of Molecular Sieves
Supported Cobalt/Ferrum Catalysts and the
CNTs Yield
6.5.4 Conclusion of SEM Analysis
6.6
140
Field Emission-Scanning Electron Microscopy
(FE-SEM) Analysis
6.7
137
141
Transmission Electron Microscopy (TEM)
Analysis
142
Energy Dispersive X-Ray Analysis (EDAX)
146
CONCLUSION
148
7.1
Overall Conclusion
148
7.2
Future Works
152
REFERENCES
153
APPENDICES
165
6.8
7
xiii
LIST OF TABLES
TABLE NO.
TITLE
PAGE
2.1
Shape of cross section for three types of carbon nanotubes. 15
2.2
A summery of the major production methods and their
efficiency.
4.1
21
Denotation for the alumina, molecular sieves (MS) and
Anodic Aluminium Oxide (AAO) template supported
metal catalysts.
5.1
46
Phases, peak position (2θ) and d values (Å) in the X-ray
diffractogram patterns of the Al2O3 support, Al-Co(2.5)
and Al-Co(2.5)Cal catalysts.
5.2
64
Phase, peak position (2θ) and d values (Å) in the X-ray
diffractogram patterns of Al-Co(40.0)Cal catalyst calcined
at 450 °C / 17 h and 750 °C / 17 h.
5.3
66
Phases, peak position (2θ) and d values (Å) in the X-ray
diffractogram patterns of the Al2O3 support, Al-Fe(2.5)
and Al-Fe(2.5)Cal catalysts.
5.4
69
Phase, peak position (2θ) and d values (Å) in the X-ray
diffractogram patterns of Al-Fe(30.0)Cal catalyst calcined
at 450 °C / 17 h and 750 °C / 17 h.
5.5
70
Phases, peak position (2θ) and d values (Å) in the X-ray
diffractogram patterns of the Al2O3 support, Al-Co/Fe
and Al-Co/Fe(2.5)Cal catalysts.
5.6
73
The particle size (d) of Co3O4 face-centred cubic and
Fe2O3 rhombohedral particles in Al2O3 supported Co/Fe
catalysts.
74
xiv
5.7
The Al2O3 supported catalysts and the yield (%), the
physical appearance and density of the as-synthesized
carbon deposit in CCVD of C2H2.
5.8
77
EDX analysis data of surface compositions of various
Al2O3 supported catalysts before and after decomposition
of C2H2.
5.9
96
The AAO template supported catalysts and the yield (%),
the physical appearance and density of the
as-synthesized carbon deposit in CCVD of C2H2.
5.10
98
EDX analysis data of surface compositions of various
AAO supported catalysts before and after decomposition
of C2H2.
6.1
111
Phases, peak position (2θ) and d values (Å) in the X-ray
diffractogram patterns of the MS support, MS-Co(2.5)
and MS-Co(2.5)Cal catalysts.
6.2
118
Phase, peak position (2θ) and d values (Å) in the X-ray
diffractogram patterns of MS-Co(40.0)Cal catalyst calcined
at 450 °C / 17 h and 750 °C / 17 h.
6.3
120
Phases, peak position (2θ) and d values (Å) in the X-ray
diffractogram patterns of the MS support, MS-Fe(2.5)
and MS-Fe(2.5)Cal catalysts.
6.4
122
Phase, peak position (2θ) and d values (Å) in the X-ray
diffractogram patterns of MS-Fe(30.0)Cal catalyst
calcined at 750 and 1000 °C for 5 hours.
6.5
124
Phases, peak position (2θ) and d values (Å) in the X-ray
diffractograms of the MS support, MS-Co/Fe(2.5)
and MS-Co/Fe(2.5)Cal catalysts.
6.6
127
The particle size (d) of Co.333Na.333(AlSiO4)(H2O)2.92
cubic and Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic phase in
MS supported bimetallic catalysts.
6.7
128
The MS supported catalysts and the carbon deposition
yield (%), the physical appearance and density of the
as-synthesized carbon deposit.
130
xv
6.8
EDAX analysis data of surface compositions of various
MS supported catalysts before and after decomposition
of C2H2.
7.1
147
Comparison of purity, density and configurations of
as-grown CNTs and CNFs over various supported
catalysts.
150
xvi
LIST OF FIGURES
FIGURE NO.
2.1
TITLE
PAGE
The graphite hexagonal single crystal structure
(Dresselhaus et al., 1995). The planes of carbon (graphene
layers) are stacked in an ABAB sequence. The A and B
carbon sites are denoted by open circles and the A’ and
B’ sites by black circles. The in-plane lattice constant is
denoted by a0 and the vectors of the unit cell in the
directions a1, a2 and c are indicated.
2.2
Diamond crystal structure consisting of the FCC lattice
positions and the tetrahedral atom positions.
2.3
9
The C60 molecule showing single bond (a5) and double
bond (a6).
2.4
7
10
Carbon fiber morphologies for (a) as-deposited at 1100 °C
and (b) after heat treatment at 3000 °C (Dresselhaus et al.,
1995).
2.5
Classification of carbon nanotubes: (a) armchair, (b)
zig-zag, and (c) chiral nanotubes (Saito et al., 1998).
2.6
14
The unfolded nanotube on a 2-dimensional graphene
lattice (Kong, 2002).
2.7
12
16
Schematic experimental set-ups for carbon nanotubes
growth methods; (a) arc-discharge, (b) laser ablation and
(c) chemical vapour deposition (CVD) (Dresselhaus and
Avouris, 2001).
2.8
20
Visualization of a possible carbon nanotube growth
mechanism (Daenen et al., 2003).
22
xvii
3.1
The schematic diagram of an anodic aluminium oxide
(AAO) template.
4.1
38
Schematic diagram of the Catalytic Chemical Vapour
Deposition (CCVD) system (fixed bed flow reactor).
48
4.2
Photograph of the fixed bed flow reactor.
48
4.3
Photograph of the gas mixing component in the CCVD
system.
49
4.4
The diffraction of X-rays by crystal lattices.
51
4.5
The components of a Scanning Electron Microscope
(SEM) (Reimer, 1984).
4.6
54
Schematic ray path for a Transmission Electron
Microscope (TEM).
56
4.7
The interaction of an electron beam with an atom.
58
5.1
XRD patterns of Al2O3 support, Al–Co(2.5),
Al–Co(2.5)Cal, Al–Co(3.0) and Al–Co(3.0)Cal catalysts.
: Al2O3 tetragonal and
5.2
: Co3O4 face-centred cubic.
XRD patterns of Al-Co(40.0)Cal catalyst calcined at
450 °C / 17 h and 750 °C / 17 h.
: Co3O4 face-centred
cubic.
5.3
63
66
X-ray diffractogram patterns of various Al-Fe catalysts
together with Al2O3 support. : Al2O3 tetragonal and
S: Fe2O3 rhombohedral.
5.4
68
XRD patterns of Al-Fe(30.0)Cal catalyst calcined at
450 °C / 17 h and 750 °C / 17 h. S: Fe2O3
rhombohedral.
5.5
70
XRD diffractogram patterns of Al-Co/Fe catalysts and
Al2O3 support. : Al2O3 tetragonal,
: Co3O4
face-centred cubic and S: Fe2O3 rhombohedral.
72
5.6
SEM micrograph of dried Al2O3 support (500×).
78
5.7
SEM micrographs of Al2O3 supported Co catalysts:
(a) Al-Co(3.0) (1000×) and (b) Al-Co(3.0)Cal (1000×).
79
xviii
5.8
SEM micrographs of as-grown CNTs over various
Al2O3 supported Co catalysts, (a) Al-Co(2.5) (10000×),
(b) Al-Co(2.5)Cal (10000×), (c) and (d) Al-Co(3.0)
(5000× and 10000×) and (e) Al-Co(3.0)Cal (10000×).
5.9
80
SEM micrographs of Al2O3 supported Fe catalysts: (a)
Al-Fe(2.5)Cal catalyst (1000×) and (b) Al-Fe(3.0)
catalyst (1000×).
5.10
82
SEM micrographs of as-grown CNTs over different
Al2O3-Fe catalyst, (a) Al-Fe(2.5) (10000×), (b)
Al-Fe(2.5)Cal (10000×), (c) Al-Fe(3.0) (10000×), (d)
and (e) Al-Fe(3.0)Cal (5000× and 10000×).
5.11
83
SEM micrographs of various Al2O3-Co/Fe catalysts:
(a) Al-Co/Fe(2.5) (1000×), Al-Co/Fe(2.5)Cal (1000×)
and (c) Al-Co/Fe(3.0) (1000×).
5.12
85
SEM micrographs of CNTs grown over various of
Al-Co/Fe catalysts: (a) Al-Co/Fe(2.5) (10000×), (b)
Al-Co/Fe(2.5)Cal (10000×), (c) and (d) Al-Co/Fe(3.0)
(10000×), (e) to (h) Al-Co/Fe(3.0)Cal (2500×, 5000×
and 10000×).
5.13
87
FE-SEM micrographs of as-grown CNTs over
Al-Co/Fe(3.0)Cal catalyst, (a) 60000× magnification and
(b) 250000× magnification.
5.14
90
TEM micrographs of as-grown MWNTs produced by
Al-Co(3.0) catalyst, (a) scale bar: 100 nm and (b) scale
bar: 20 nm.
5.15
TEM micrographs of MWNTs formed by Al-Fe(3.0)
catalyst, (a) scale bar: 100 nm and (b) scale bar: 10 nm.
5.16
91
92
TEM micrographs of as-synthesized MWNTs over
Al-Co/Fe catalysts: (a) and (b) Al-Co/Fe(2.5)Cal (scale
bar: 50 nm and 10 nm) as well as (c) and (d) Al-Co/Fe(3.0)
(scale bar: 100 nm and 10 nm).
5.17
94
TEM micrographs of MWNTs grown over
Al-Co/Fe(3.0)Cal catalyst: (a) scale bar: 100 nm, (b)
scale bar: 5 nm and (c) scale bar: 10 nm.
95
xix
5.18
X-ray diffractogram patterns of AAO template and
AAO-Co(0.5) catalyst.
5.19
SEM micrograph of anodic aluminium oxide (AAO)
template (20000 ×).
5.20
99
101
SEM micrographs of AAO template supported Co
catalysts: (a) AAO-Co(0.5)Cal (20000 ×) and (b)
AAO-Co(1.0)Cal (20000 ×).
5.21
102
SEM micrographs of carbon deposit over various Co
catalysts: (a) and (b) AAO-Co(0.5) (10000× and 2000×),
(c) and (d) AAO-Co(0.5)Cal (10000×), (e) and (f)
AAO-Co(1.0) (20000× and 10000×), (g) and (h)
AAO-Co(1.0)Cal (10000×).
5.22
SEM micrographs of different AAO-Fe catalysts:
(a) AAO-Fe(0.5)Cal catalyst (20000×) and (b)
5.23
103
105
SEM micrographs of as-grown CNTs over various
AAO-Fe catalysts: (a) AAO-Fe(0.5) (10000×), (b)
AAO-Fe(0.5)Cal (10000×), (c) AAO-Fe(1.0) (10000×),
(d) and (e) AAO-Fe(1.0)Cal (10000×).
5.24
106
Micrographs from SEM analysis of AAO supported
bimetallic catalysts: (a) AAO-Co/Fe(0.5) (10000×) and
(b) AAO-Co/Fe(1.0) (5000×).
5.25
107
SEM micrographs of CNTs grown over: (a)
AAO-Co/Fe(0.5) (10000×), (b) and (c) AAO-Co/Fe(0.5)Cal
(10000× and 5000×), (d) and (e) AAO-Co/Fe(1.0) (10000×
and 5000×), (f) and (g) AAO-Co/Fe(1.0)Cal (5000× and
10000×).
6.1
108
XRD diffractogram patterns of MS support, MS-Co(2.5),
MS-Co(2.5)Cal, MS-Co(3.0) and MS-Co(3.0)Cal catalysts.
¡: NaSiAlO4 face-centred cubic and T:
Co.333Na.333(AlSiO4)(H2O)2.92 cubic phase.
117
xx
6.2
XRD patterns of MS-Co(40.0)Cal catalyst calcined at
450 °C / 17 h and 750 °C / 17 h.T: Co3O4 face-centred
cubic.
6.3
119
XRD patterns of various MS-Fe catalysts together with
MS support. ¡: NaSiAlO4 face-centred cubic and
Ø : Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic phase.
6.4
121
XRD patterns of MS-Fe(30.0)Cal catalyst calcined at
750 °C and 1000 °C for 5 hours. Ø: Fe2O3 rhombohedral. 124
6.5
X-ray diffractogram patterns of MS support and series
of MS-Co/Fe catalysts. ¡ : NaSiAlO4 face-centred cubic,
T : Co.333Na.333(AlSiO4)(H2O)2.92 cubic and Ø :
Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic phase.
6.6
SEM micrograph of dried molecular sieves (MS) support
(1000×).
6.7
131
SEM micrographs of MS supported Co catalysts: (a)
MS-Co(2.5)Cal (1000×) and (b) MS-Co(3.0) (1000×).
6.8
126
132
SEM micrographs of as-grown CNTs over various MS
supported Co catalysts: (a) and (b) MS-Co(2.5) (5000×
and 10000×), (c) MS-Co(2.5)Cal (5000×), (d) MS-Co(3.0)
(10000×), (e) and (f) MS-Co(3.0)Cal (10000×).
6.9
SEM micrographs of MS supported Fe catalysts: (a)
MS-Fe(3.0) (1000×) and (b) MS-Fe(3.0)Cal (1000×).
6.10
133
135
SEM micrographs of as-grown CNTs over series of MS-Fe
catalysts: (a) and (b) MS-Fe(2.5) (10000×), (c) and (d)
MS-Fe(2.5)Cal (10000×), (e) and (f) MS-Fe(3.0) (10000×),
(g) and (h) MS-Fe(3.0)Cal (10000×).
6.11
SEM micrographs of MS supported Fe catalysts: (a)
MS-Fe(3.0) (1000×) and (b) MS-Fe(3.0)Cal (1000×).
6.12
136
137
SEM micrographs of CNTs grown over the bimetallic
catalysts: (a) and (b) MS-Co/Fe(2.5) (10000×), (c) and
(d) MS-Co/Fe(2.5)Cal (5000× and 2000×), (e) and (f)
MS-Co/Fe(3.0) (10000× and 300×), (g) and (h)
MS-Co/Fe(3.0)Cal (5000×).
139
xxi
6.13
FE-SEM micrographs of as-grown CNTs over
MS-Co/Fe(3.0)Cal catalyst, (a) 60000× magnification
and (b) 250000× magnification.
6.14
TEM micrographs of the as-grown MWNTs over
MS-Co(3.0)Cal catalyst, (a) and (b) scale bar: 10 nm.
6.15
141
142
TEM micrographs of as-grown MWNTs from: (a) and (b)
MS-Fe(2.5)Cal catalyst, (c) and (d) MS-Fe(3.0) catalyst
(for all images scale bar: 10nm).
6.16
143
TEM micrographs of as-grown MWNTs from
MS-Co/Fe(2.5) catalyst, (a) scale bar: 100 nm, (b) scale
bar: 10 nm and (c) scale bar: 10 nm.
6.17
144
TEM micrographs of as-synthesized MWNT over
MS-Co/Fe(3.0)Cal catalyst, both of the micrographs have
scale bar of 5 nm.
145
xxii
LIST OF ABBREVIATIONS
AAO
-
Anodic Aluminium Oxide template
Al2O3
-
Alumina support
C
-
Carbon
Cal
-
Calcination
CVD
-
Chemical Vapour Deposition
CCVD
-
Catalytic Chemical Vapour Deposition
C2H2
-
Acetylene
CNFs
-
Carbon Nanofibers
CNTs
-
Carbon Nanotubes
Co
-
Cobalt
Cu
-
Copper
EDX
-
Energy Dispersive X-Ray Analysis
FE-SEM
-
Field Emission-Scanning Electron Microscopy
Fe
-
Iron
MgO
-
Magnesium Oxide
Mo
-
Molibdenum
MS
-
Molecular Sieves support
MWNTs
-
Multi-Walled Carbon Nanotubes
Ni
-
Nickel
sccm
-
Standard cubic centimetres per minute
SEM
-
Scanning Electron Microscopy
SiO2
-
Silica
SMSI
-
Strong Metal-Support Interaction
SWNTs
-
Single-Walled Carbon Nanotubes
TEM
-
Transmission Electron Microscopy
V
-
Vanadium
XRD
-
X-Ray Diffraction
xxiii
LIST OF APPENDICES
APPENDIX
A
TITLE
Mathematical Expressions for Stoichiometric Amount
of Metal Salt Used in Wet Impregnation
B
E
168
Mathematical Expressions for Crystallite Size of Co3O4
Cubic Crystal Using Scherrer Equation
D
165
Mathematical Expressions for Stoichiometric Amount
of Metal Salt Used in Dip Coating
C
PAGE
171
Mathematical Expressions for Particle Size, Length and
Diameter from the SEM Micrograph
173
Publications
174
CHAPTER 1
INTRODUCTION
1.1
Research Background
The past decade has witnessed tremendous effort and progress in the field of
carbon nanotubes. Ever since the discovery of carbon nanotubes by Iijima (1991), it
has captured the attention of researchers worldwide. Understanding their unique
properties and exploring their potential applications have been a main driving force
for this area.
Throughout history, the allotropes of carbon have played a number of
important roles in technology. In ancient times, diamond was celebrated for its
hardness and beauty, and carbon black was used as a colorant. The industrial age
brought greater interest in graphite and related carbon materials as a source of carbon
vapour in arc-lamps and as clean-burning fuels. Graphite-like carbon materials are
now widely used for their unique mechanical, electrical and thermal properties.
Very small diameter (less than 10nm) carbon filaments were prepared in the
1970’s and 1980’s by the decomposition of hydrocarbons at high temperatures
(Dresselhaus and Avouris, 2001). Direct stimulus to study carbon filaments of very
small diameters more systematically came from the discovery of fullerenes by Kroto
and Smalley.
2
In December 1990 at a carbon-carbon composites workshop, papers were
given on the status of fullerene research by Smalley, the discovery of a new synthesis
method for the efficient production of fullerenes and a review of carbon fibers
research by M.S. Dresselhaus. Discussions at the workshop stimulated Smalley to
speculate about the existence of carbon nanotubes of dimensions comparable to C60.
These conjectures were later followed up in August by an oral presentation at a
fullerene workshop by Dresselhaus on the symmetry proposed for carbon nanotubes
capped at either end by fullerene hemispheres (Saito et al., 1998).
However, the real breakthrough on carbon nanotube research came with the
experimental observation of carbon nanotubes in 1991 by Iijima of the NEC
Laboratory in Tsukuba, Japan using High-Resolution Transmission Electron
Microscopy (HRTEM) (Iijima, 1991). It was this work which bridged the gap
between experimental observation and the theoretical framework of carbon
nanotubes in relation to fullerenes and as theoretical examples of 1D system. Since
the pioneering work by Iijima, the study of carbon nanotubes has progressed rapidly.
1.2
Problem Statement
The main hindrance to employing carbon nanotubes (CNTs) in real world is
the inability to control the growth of the nanotubes and to grow bulk amounts of
carbon nanotubes. There are three main techniques to grow carbon nanotubes: arcdischarge, laser ablation and chemical vapour deposition (CVD). The first two
methods are high temperature processes that produce high quality CNTs, but they
cannot grow mass quantities of nanotubes within a reasonable amount of time. The
CVD technique is able to grow bulk amounts of nanotubes and arrays of multiwalled carbon nanotubes (MWNTs). However, these nanotubes contain a vast
amount of defects along the length of the tubes due to the relatively low synthesis
temperature of 600 – 1200 °C.
3
Nevertheless, some progress has been recently obtained, the chemical vapour
deposition (CVD) has been modified by applying various supported metals catalysts
in the production of CNTs. The catalytic chemical vapour deposition (CVD) method
supplies CNTs in high yield and low costs, but also at controlling the CNTs
characteristics and morphologies. Being a catalytic process, the combinations of
transition metals and support can be changed depending on the characteristics
required, such as the alignment and diameter of the nanotubes. The CCVD synthesis
of CNTs can be carried out at low temperature and ambient pressure.
1.3
Research Objectives
This research is intended to synthesize carbon nanotubes (CNTs) of high
yield and purity at economical cost. Therefore, this research is conducted to achieve
the following primary objectives:
1.
To produce carbon nanotubes (CNTs) using Catalytic Chemical Vapour
Deposition (CCVD) method.
2.
To study the effects of supported catalysts in the synthesis of carbon
nanotubes (CNTs).
3.
To characterise the supported catalysts and carbon nanotube yields
chemically and physically.
4.
To identify the best catalyst-support combination to catalyze the carbon
nanotubes growth.
4
1.4
Scopes of Study
In order to achieve the objectives, this research is focusing on the following
scopes:
(i)
Designing and fabricating an effective Catalytic Chemical Vapour Deposition
(CCVD) system contains a fixed bed flow reactor which facilitates the
production of CNTs.
(ii) Preparing supported catalysts by using alumina (Al2O3) beads, molecular
sieves (MS) beads and anodic aluminium oxide (AAO) template as catalyst
supports and cobalt (Co) or ferrum (Fe) as metal catalysts with impregnation
or dip coat techniques. Applying calcination treatment to enhance the activity
of the catalysts.
(iii) Producing CNTs through catalyze pyrolysis of acetylene (C2H2) at optimum
temperature, reaction time and gas flow rate.
(iv) Characterising the supported catalysts and CNT yields using X-ray
Diffraction (XRD) technique to reveal the active catalyst sites and effects of
calcination.
(v)
Investigating the morphologies and topologies of the supported catalysts and
the CNT yields through Scanning Electron Microscopy (SEM), FieldEmission Scanning Electron Microscopy (FE-SEM) and Transmission
Electron Microscopy (TEM) techniques.
(vi) Determining the surface composition of a sample (metal catalyst and carbon
content) using Energy Dispersive X-Ray Analysis (EDAX) technique.
(vii) Comparing the performance of supported catalysts in the production of CNTs
to figure out the best combination of support, catalyst, loading and treatment
for the supported catalysts.
CHAPTER 2
CARBON AND CARBON NANOTUBES
2.1
Carbon and Carbon Structures
2.1.1
Carbon and Bonding between Carbon Atoms
Carbon-based materials, clusters and molecules are unique in many ways.
Carbon is a Group 4A element on the periodic table with 6 protons, 6 neutrons, 2
core electrons and 4 valence electrons. The electron configuration of the carbon atom
is:
1s22s22p2
The 1s2 orbital contains two strongly bonded core electrons. Four more
weakly bonded electrons occupy the 2s22p2 valence orbitals. In the crystalline phase,
the valence electrons give rise to 2s, 2px, 2py and 2pz orbitals which are important in
forming covalent bonds in carbon materials. Further, carbon has an atomic mass unit
(amu) of 12.01 and an atomic radius of 0.77 Å.
The unique properties of carbon are consistent with the ability of the carbon
atom to hybridize. Since the energy difference between the upper 2p energy levels
and the lower 2s level in carbon is small compared with the binding energy of the
chemical bonds, the electronic wave functions for these four electrons can readily
mix with each other, thereby changing the occupation of the 2s and three 2p atomic
6
orbitals so as to enhance the binding energy of the C atom with its neighbouring
atoms.
The general mixing of 2s and 2p atomic orbitals is called hybridization,
whereas the mixing of a single 2s electron with one, two or three 2p electrons is
called spn hybridization with n = 1, 2 and 3 (Dresselhaus and Avouris, 2001). The
other Group 4A elements (Si, Ge, etc.) do not possess this ability since they only
hybridize into sp3 structure. This unique ability of carbon is primarily due to the
absence of spherical inner electron orbitals 1s, allowing for the mixing of only the
valence s and p orbitals (Ward, 2002). The various bonding states are connected with
certain structural arrangements, so that sp bonding gives rise to chain structures, sp2
bonding to planar structures and sp3 bonding to tetrahedral structures, these have lead
to 3 main structures associated with carbon (graphite, diamond and fullerenes).
Lately, much attention has focussed on small carbon clusters, since the
discovery of fullerenes and carbon nanotubes. The physical reason why these
nanostructures form is that a graphene layer (defined as a single 2-dimensional (2D)
layer of 3D graphite) of finite size has many edge atoms with dangling bonds, and
these dangling bonds correspond to high energy states. Therefore the total energy of
a small number of carbon atoms (30 – 100) is reduced by eliminating dangling
bonds, even at the expense of increasing the strain energy, thereby promoting the
formation of closed cage clusters such as fullerenes and carbon nanotubes.
The accompanying sub-sections briefly discuss the properties of each carbon
allotrope (structure) and also carbon fibers. Graphite and diamond are two main
allotropic forms of bulk carbon. Fullerenes and carbon nanotubes constitute other
allotropic forms.
7
2.1.2
Graphite
Under ambient conditions and in bulk form, the graphite phase with strong inplane trigonal bonding is the stable phase. Graphite assumes the sp2 hybridized form
of carbon, meaning it can bond with three other carbon atoms that lie with the plane
at a bond angle of 120°. The crystal structure of graphite containing 4 carbon atoms
per unit cell, resembles a hexagonal structure with 6 covalently bonded carbon atoms
in-plane, making a honeycomb structure and a weak Van der Waals attraction out-ofplane holding the graphite layers together (Figure 2.1).
Figure 2.1: The graphite hexagonal single crystal structure (Dresselhaus et al.,
1995). The planes of carbon (graphene layers) are stacked in an ABAB sequence.
The A and B carbon sites are denoted by open circles and the A’ and B’ sites by
black circles. The in-plane lattice constant is denoted by a0 and the vectors of the unit
cell in the directions a1, a2 and c are indicated.
As demonstrated in Figure 2.1, the ideal crystal structure of graphite consists
of layers in which the carbon atoms are arranged in an open honeycomb network,
such that the A and B atoms on consecutive layers are on top of one another, but the
A’ atoms in one plane are over the unoccupied centres of the adjacent layers, and
similarly for the B’ atoms. This give rise to an ABAB planar stacking arrangement.
The in-plane nearest-neighbour distance (aC-C) of carbon atoms are ~1.421 Å, with a
8
lattice constant of a0 = 2.462 Å, a c-axis lattice constant c0 of 6.708 Å and an
interplanar distance of c0 / 2 = 3.3539 Å (Dresselhaus et al., 1995).
The unique bonding characteristics of graphite allows for it to have some
interesting thermal, mechanical and electrical properties. Graphite has a density of
2.27 g/cm3 and a very high melting point of ≥ 3550 °C. Mechanically, graphite is one
of the softest materials around with a Mhos hardness of 0.5 – 1.0 (Mhos hardness is
measured on a 0 (softest) – 10 (hardest) scale). In comparison steel has a mhos
hardness of 5 – 8.5. Electrically, the graphite structure is highly anisotropic,
exhibiting semi-metallic behavior, with a small band overlap of ~ 0.04 eV, the
valence band and the conduction band touch and are degenerate at certain points
along the energy bands.
2.1.3
Diamond
The second most widely recognized allotrope of carbon is diamond. Diamond
is different with graphite in that it hybridizes into the sp3 state. The hybridization
scheme of diamond causes the carbon atoms to form a tetrahedral lattice with the
strong covalent carbon-carbon (C-C) bonds at an angle of 109.5°. Diamond
crystallizes into a face-centered cubic (FCC) structure with half of the tetrahedral
sites filled (known as the diamond cubic crystal structure). In the ideal diamond
structure (Figure 2.2), every carbon atom is surrounded by four other carbon atoms at
the corners of a regular tetrahedron with a cube edge length of a0 = 3.567 Å. The
nearest-neighbour C-C distance is 1.544 Å, nearly 10 % larger than in graphite, yet
the atomic density of diamond (1.77 × 1023 cm-3) is 56 % higher than in graphite, due
to the high anisotropy of the graphite structure. The diamond crystal is highly
symmetric.
Diamond has a density of 3.51 g/cm3. Diamond is not the thermodynamically
stable form of carbon at standard temperature and pressure. The stable form of
carbon is graphite. Diamond is a metastable carbon allotrope. Diamond forms by
9
placing graphite under extremely high pressures and temperatures, causing the
carbon bond to rearrange into the denser, more stable form of the diamond cubic
structure. Diamond does not rearrange back into graphite structure upon cooling due
to the very high activation energy for this reaction to occur. Diamond is one of the
strongest materials known with a Mhos hardness of 10, melts at the extremely high
temperature of 4000 °C and is insulating with a band gap of ~ 5.4 eV.
Figure 2.2: Diamond crystal structure consisting of the FCC lattice positions and the
tetrahedral atom positions.
The extreme hardness of diamond has allowed it to be utilized for cutting and
drilling tools for mining and other industries. It is also finding applications for
scratch resistant coatings with the developing technology of diamond thin films.
Diamond films are also finding prominence in the electronics field, since they
become semiconducting upon doping. This is of interest since diamond can
withstand temperatures ≥ 1000 °C and able to handle high power (Ward, 2002).
10
2.1.4
Fullerenes
Fullerenes were discovered in 1985 by Kroto, Smalley and Curl at Rice
University. The discovery of C60 directly led to the discovery of carbon nanotubes in
1991 by Iijima. C60 or the buckminsterfullerene or buckyball involves 60 carbon
atoms in a ‘soccer ball’ shaped structure. A model of a fullerene is given in Figure
2.3 (Dresselhaus et al., 1995).
The 60 carbon atoms in C60 are known to be located at the vertices of a
truncated icosahedron where all carbon sites are equivalent (see Figure 2.3). To a
first approximation, we can assume that C60 as a ‘rolled-up’ graphene sheet (a single
layer of crystalline graphite), these are because: (1) the average nearest neighbour
carbon-carbon (C-C) distance aC-C in C60 (1.44 Å) is almost identical to that in
graphite (1.42 Å), (2) each carbon atom in graphite and in C60 is trigonally bonded to
three other atoms in an sp2-derived bonding configuration, and (3) most of the faces
on the regular truncated icosahedron are hexagons (Rao et al., 1995).
Figure 2.3: The C60 molecule showing single bond (a5) and double bond (a6).
A regular truncated icosahedron has 90 edges of equal length, 60 equivalent
vertices, 20 hexagonal faces and 12 additional pentagonal faces to form a close shell,
consistent with Euler’s theorem (Dresselhaus et al., 1995). Two single C-C bonds are
11
located along a pentagonal edge at the fusion of a hexagon and a pentagon. In C60,
the bond lengths for the single bonds a5 are 1.46 Å (Figure 2.2). The third bond is
located at the fusion between two hexagons and is a double bond with a bond length
a6 which is measured to be 1.40 Å.
It should be noted that the C60 is not the only form of closed shell spherical
carbon hollow structure. C70, C82 etc. are other forms of fullerenes, however, Euler’s
theorem must still be satisfied. Since the bonding requirements of all the valence
electrons in C60 are satisfied, it is expected that C60 has filled molecular levels.
Because of the closed-shell properties of C60 (and also other fellerenes), the nominal
sp2 bonding between adjacent carbon atoms occurs on a curved surface, in contrast to
the case of graphite where the sp2 trigonal bonds are truly planar (Rao et al., 1995).
This curvature of the trigonal bonds in C60 leads to some admixture of sp3 bonding,
characteristic of tetrahedrally bonded diamond, but absent in graphite.
Inspection of the C60 molecular structure shows that every pentagon of C60 is
surrounded by five hexagons; the pentagon, together with its five neighbouring
hexagons, has the form of the corannulene molecule. Another molecular subunit on
the C60 molecule is the pyraclene (also called pyracylene) subunit, which consists of
two pentagons and two hexagons in the arrangement. Adherence to the ‘isolated
pentagon rule’, whereby the distance between pentagons on the fullerene shell is
maximized, reduces local curvature and stress and gives added stability to C60
(Dresselhaus et al., 1995).
At the present, there are no practical uses for fullerenes, yet, some research
has been conducted on attaching molecules to the fullerenes, possibly allowing for
the utilization of fullerenes for medical purposes, such as inhibiting HIV by attaching
to the virus and preventing the replication of the virus.
12
2.1.5
Carbon Fibers
Carbon fibers are an important class of graphite-related materials and predate
the discovery of carbon nanotubes. Carbon fibers consist of layers of graphite planes
deposited parallel (or normal) to the fiber axis as seen in Figure 2.4. This stacking of
graphite planes gives the fibers good mechanical properties, such as high strength
and modulus, along the direction of the axis. However, the fibers are very brittle and
fracture easily under a compressive stress. The properties of carbon fibers are closely
related to multi-walled carbon nanotubes (MWNTs) and very frequently MWNTs are
present as the precursor in the centre of a vapour grown carbon fiber.
Figure 2.4: Carbon fiber morphologies for (a) as-deposited at 1100 °C and (b) after
heat treatment at 3000 °C (Dresselhaus et al., 1995).
The heat treatments at high temperatures improve the properties and
graphitisation of the carbon fibers, which causes the fiber to resemble a carbon
nanotube’s structure. One of the important attributes of carbon fibers is that the
Chemical Vapour Deposition (CVD) growth of the fibers closely parallels the growth
of carbon nanotubes by CVD. CVD grown fibers can be prepared over a wide range
of diameters (from ~ 10 nm to more than 100 µm) and these fibers have hollow cores
(Dresselhaus et al., 1995). Carbon fibers have several uses such as for mechanical
strengthening of light weight polymer composite materials.
13
2.2
Carbon Nanotubes
2.2.1
Structure of Carbon Nanotubes
Carbon nanotubes (CNTs) are unique tubular structure of nanometer
diameter and large length/diameter ratio (~ 1000). CNTs can be considered
conceptually as a prototype one-dimensional (1D) quantum wire (Dresselhaus and
Avouris, 2001). The fundamental building block of carbon nanotubes is the very long
all-carbon cylindrical single-walled carbon nanotubes (SWNTs), one atom in wall
thickness and tens of atoms around the circumference (typical diameter ~ 1.4 nm).
A single-walled carbon nanotube can be described as a graphene sheet rolled
into a seamless hollow cylindrical shape with a high degree of molecular perfection.
A multi-walled carbon nanotube consists of concentric cylinders with an interlayer
spacing of 0.34 nm and a diameter of typically 10 – 20 nm, and these were the first
type of nanotubes observed experimentally. The lengths of two types of nanotubes
can be up to several hundreds of microns or even centimetres (Kong, 2002). They
can either occur in individual tubes or bundle up into ‘ropes’ due to the Van der
Waals interaction between their side walls.
Nanotube structure is one dimensional with axial symmetry, and in general
exhibiting a spiral conformation, called chirality. The chirality is given by a single
vector called chiral vector, Ch. The carbon bonds found on the sidewalls of nanotubes
are of predominantly sp2 character (Williams, 2001), though with decreasing tube
radius, the curvature-induced strain will give rise to intermediate sp3 character.
Carbon nanotubes are stronger than diamond (Yu et al., 2000) but have a bulk
density comparable to graphite.
An interesting and essential fact about the structure of a CNT is the
orientation of the six-membered carbon ring (hexagon) in the honeycomb lattice
relative to the axis of the nanotube. Three examples of single-wall carbon nanotubes
(SWNT) are shown in Figure 2.5. From this figure, the direction of the six-
14
membered ring in the honeycomb lattice can be taken almost arbitrarily, without any
distortion of the hexagons. This fact provides many possible structures for carbon
nanotubes (Saito et al., 1998).
Figure 2.5: Classification of carbon nanotubes: (a) armchair, (b) zig-zag, and (c)
chiral nanotubes (Saito et al., 1998).
The primary symmetry classification of a carbon nanotube is as either being
achiral (symmorphic) or chiral (non-symmorphic). An achiral carbon nanotube is
defined by a carbon nanotube whose mirror image has an identical structure to the
original one. There are only two cases of achiral nanotubes; armchair and zigzag
nanotubes, as shown in Figure 2.5 (a) and (b), respectively. Chiral nanotubes (Figure
2.5 (c)) exhibit a spiral symmetry whose mirror image is not super-imposable on to
the original one (Saito et al., 1998). In the chemical nomenclature, such structures
are called axially chiral. Axial chirality is commonly discussed in connection with
optical activity. The terminations of each of the three nanotubes are cis-type
(armchair), trans-type (zig-zag) and mixture of cis and trans (chiral) as shown on
Table 2.1.
A SWNT consist of two separate regions with different physical and chemical
properties. The first is the sidewall of the tube and the second is the termination of
the tube. The terminations are often called caps or end caps and its structure is
15
similar to or derived from a smaller fullerene, such as C60. C-atoms placed in
hexagons and pentagons form the end cap structures. From Euler’s theorem that
twelve pentagons are needed in order to obtain a closed cage structure, which
consists of only pentagons and hexagons (Daenen et al., 2003). The combination of a
pentagon and five surrounding hexagons results in the desired curvature of the
surface to enclose a volume. The isolated pentagon rule also states that the distance
between pentagons on the fullerene shell is maximized in order to obtain a minimal
local curvature and surface stress, resulting in a more stable structure.
Table 2.1: Shape of cross section for three types of carbon nanotubes.
Type
(a) Armchair
(b) Zig-zag
(c) Chiral
Shape of cross section
cis-type
trans-type
Mixture of cis and trans
The other structure of which a SWNT is composed is a cylinder. It is
generated when a graphene sheet of a certain size that is wrapped in a certain
direction. The structure of a cylinder is conveniently explained in terms of its onedimensional unit cell, defined by the vectors Ch and T in Figure 2.6. T, called the
translation vector, defines the direction of carbon nanotubes axis. Ch called chiral
vector, represents the circumference of a carbon nanotube. It connects two
crystallographically equivalent sites on the graphene sheet: Ch = nâ1 + mâ2, where â1
and â2 are the unit vectors of 2-dimentional graphene sheet. By folding a cylinder so
that the beginning and end of a (n, m) lattice vector in the graphene plane are joined
together, one obtain a (n, m) nanotube (Kong, 2002). The (n, m) indices determine
the diameter of the nanotube and also the chirality. The (n, n) tubes are armchair
nanotubes (Figure 2.5 (a)) since the atoms around the circumference are in armchair
pattern. The (n, 0) nanotubes are termed ‘zig-zag’ in view of the atomic
configuration along the circumference (Figure 2.5 (b)). The other types of nanotubes
are chiral, with the rows of hexagons spiraling along the nanotube axis (Figure 2.5
(c)).
16
Figure 2.6: The unfolded nanotube on a 2-dimensional graphene lattice (Kong,
2002).
SWNTs with different chiral vectors have dissimilar properties such as
optical activity, mechanical strength and electrical conductivity. Thus, different
diameter, chirality and cap structures give a variety of geometries in carbon
nanotubes.
2.2.2
Properties of Carbon Nanotubes
Electronic, molecular and structural properties of carbon nanotubes are
determined to a large extent by their nearly one dimensional structure. The most
important properties of CNTs and their molecular background are discussed below.
Electrical conductivity: Extensive theoretical studies have shown that a
single-walled nanotube (SWNT) can behave as a well-defined metallic,
semiconducting or semi-metallic wire depending on its chirality and diameter
(Hamada et al., 1992 and Kane and Mele, 1997). For (n, m) nanotubes with n = m
(arm-chair nanotubes), they are always metallic; for (n, m) nanotubes with
n – m ≠ 3 × integer, the nanotubes are semiconductors with moderate band gaps. The
17
energy gap scale with the tube diameter as 1/d and is on the order of 0.5 eV for a
SWNT with a typical diameter d = 1.4 nm. For n – m = 3 × integer, the tubes would
be semimetals, but become small-gap semiconductors (band gap scales with 1/d2 ~
10 meV for d ~ 1.4 nm) due to a curvature induced orbital rehybridization effect
(Kane and Mele, 1997). The resistance to conduction is determined by quantum
mechanical aspects and is independent of the nanotube length (Louie, 2001). The
extreme sensitivity of the electronic properties on the structural parameters is a
unique feature for carbon nanotubes. This uniqueness leads to rich physical
phenomena in nanotube systems, but also poses a significant challenge to chemical
synthesis in terms of controlling the nanotube diameter and chirality.
Mechanical Properties: Carbon nanotubes have excellent mechanical
characteristics due to the nature of their strong carbon-carbon bonding and their
seamless structure. Nanotubes are among the strongest and most resilient materials.
Carbon nanotubes are predicted to have far superior mechanical properties in
comparison to carbon fibers, and also being highly flexible (Yakobson et al., 2001).
A nanotube has Young’s modulus of 1.2 TPa and a tensile strength about a hundred
times higher than steel and can tolerate large strains before mechanical failure.
Chemical reactivity: The chemical reactivity of a CNT is enhanced as a
direct result of the curvature of the CNT surface. CNT reactivity is directly related to
the π-orbital mismatch caused by an increased curvature (Daenen et al., 2003). A
smaller nanotube diameter results in increased reactivity.
Thermal Properties: The thermal properties of carbon nanotubes are closely
related to the thermal properties of graphite, but contain many characteristics of 1D
structure. The specific heat, C of carbon nanotubes is mainly dependent on the
phonon contribution, similar to 3D graphite or a graphene sheet . The thermal
conductivity of carbon nanotubes is relatively high at moderate temperatures.
Optical activity: Theoretical studies have discovered that the optical activity
of chiral nanotubes disappears if the nanotubes become larger (Daenen et al., 2003).
18
2.2.3
Common Synthesis Methods
Three relatively efficient methods to synthesize carbon nanotubes have been
identified: laser ablation, carbon arc synthesis and chemical vapour deposition
(CVD). Three of these methods depend on the use of catalysts. In each of these
methods carbon atoms self-assemble to form nanometer size tubes of carbon.
Arc-discharge and laser ablation methods for the growth of nanotubes have
been actively pursued in the past ten years. Both methods involved the condensation
of carbon atoms generated from evaporation of solid carbon sources. The
temperatures involved in these methods are close to the melting temperature of
graphite, 3000 – 4000 ºC (Dresselhaus and Avouris, 2001). However, due to the high
temperatures characteristic for arc-discharge and laser ablation methods, it is very
difficult to ensure adequate stability and reproducibility of the products (Krivoruchko
et al., 2000).
The original arc-discharge experiments of 1991 produced multi-walled tubes
that possessed well-defined graphitic layers. This technique involves placing two
graphite electrodes in close proximity for arcing to occur under the application of an
electric field (Figure 2.7 (a)). Under these conditions, carbon is consumed from the
anode via evaporation (Papadopoulos, 2000). Carbon atoms are evaporated by
plasma of helium gas ignited by high currents passed through opposing carbon anode
and cathode. This arc-evaporated material then re-condenses on the cathode and the
subsequent deposit contains CNTs.
Arc-discharge has been developed into an excellent method for producing
both high quality multi-walled nanotubes and single-walled nanotubes (Dresselhaus
and Avouris, 2001). Multi-walled carbon nanotubes (MWNTs) can be obtained by
controlling the growth conditions such as the pressure of inert gas in the discharge
chamber and the arcing current. The nanotubes are typically bound together by
strong Van der Waals interactions and form tight bundles. MWNTs produced by arcdischarge are very straight, indicative of their high crystallinity. For the growth of
19
single-walled tubes, a metal catalyst is needed in the arc-discharge system. In 1993,
Bethune and co-workers used a carbon anode containing a small percentage of cobalt
catalyst in the discharge experiment, and found abundant SWNTs generated in the
soot material (Popov, 2004).
The growth of high quality SWNTs at the 1-10 g scales was achieved by
Smalley and co-workers using a laser ablation (laser oven) method (Figure 2.7 (b))
(Dresselhaus and Avouris, 2001). The method utilized intense laser pulses to ablate a
carbon target containing 0.5 atomic percent of nickel and cobalt. The target was
placed in a tube-furnace heated to 1200 ºC. During laser ablation, a flow of inert gas
was passed through the growth chamber to carry the grown nanotubes downstream to
be collected on a cold finger. The produced SWNTs are mostly in the form of ropes
consisting crystals via Van der Waals interactions.
A schematic experimental set-up for CVD growth is depicted in Figure 2.7
(c). The growth process involves heating a catalyst material to high temperatures in a
tube furnace and flowing through a hydrocarbon gas into the tube reactor for a period
of time. Materials grown over the catalyst are collected upon cooling the system to
room temperature. Both multi- and single-walled tubes have been grown by this
technique (Li et al., 1996 and Dai et al., 1996). The main factors in growing CNTs
via CVD are the hydrocarbons, catalysts and growth temperatures (Dresselhaus and
Avouris, 2001).
The active catalyst species in CVD method are typically transition metal
nanoparticles formed on a support material such as alumina. The general nanotube
growth mechanism in a CVD process involves the dissociation of hydrocarbon
molecules catalyzed by the transition metal, and dissolution and saturation of carbon
atoms in the metal nanoparticles. For MWNTs growth, most of the CVD methods
employ ethylene or acetylene as the carbon feedstock and the growth temperature is
typically in the range of 550 – 750 ºC. Iron, nickel or cobalt nanoparticles are often
used as catalyst. CVD is seen as providing a potentially cost-effective solution for
the large-scale production of nanotubes and can allow the controlled growth of CNTs
at specific sites for potential device applications (Papadopoulos, 2000).
20
(a)
He
Arc-discharge
I
(b)
Laser Ablation
Oven temperature ~ 1200 °C
(c)
CVD
CnHm
Catalyst
Oven temperature 500-1000 °C
Figure 2.7: Schematic experimental set ups for carbon nanotubes growth methods:
(a) arc-discharge, (b) laser ablation and (c) chemical vapour deposition (CVD)
(Dresselhaus and Avouris, 2001).
In 1993, Yacaman et al. (1993) invented a CVD modified method called
catalytic chemical vapour deposition (CCVD). Henceforward, the CCVD method is
widely used to synthesis CNTs. The CCVD method will be discussed in details in
Chapter 3: Catalytic Chemical Vapour Deposition (CCVD).
Iron, cobalt and nickel are also the favoured catalytic metals used in laser
ablation and arc-discharge. This simple fact may hint that the laser, discharge and
21
CVD growth methods may share a common nanotube growth mechanism, although
very different approaches are used to provide carbon feedstock (Dresselhaus and
Avouris, 2001). Table 2.2 is a short summary of the three most common techniques
to synthesize CNTs.
Table 2.2: A summery of the major production methods and their efficiency.
Method
Who
How
Arc Discharge
Ebbesen and Ajayan,
NEC, Japan, 1992
Connect two graphite
rods to a power supply,
place a few mm apart,
and throw the switch.
At 100 A, carbon
vaporizes and forms a
hot plasma.
Chemical Vapour
Deposition
Endo,
Shinshu
University, Japan.
Place substrate in oven,
heat to 600 °C and
slowly add a carbonbearing gas. As gas
decomposes it frees up
carbon atoms, which
combine in the form of
CNTs.
Typical yield
SWNT
30 – 90 %
20 – 100 %
Short
tubes
with Long
tubes
with
diameters of 0.6 – 1.4 diameters ranging from
nm.
0.6 – 4.0 nm.
MWNT
Short tubes with inner Long
tubes
with
diameter of 1 – 3 nm diameters in the range
and outer diameter of ~ of 10 – 240 nm.
10 nm.
Advantages
Can easily produce
SWNT and MWNT.
SWNTs
have
few
structural
defects;
MWNTs
without
catalyst,
not
too
expensive, open air
synthesis promising.
Disadvantages Tubes tend to be short
with random sizes and
directions; often needs
purification.
Easiest to scale up to
industrial production;
long length, simple
process,
SWNT
diameter controllable,
quite pure.
Laser Ablation
Smalley,
Rice
University, 1995
Blast graphite with
intense laser pulses; use
the laser pulses rather
than
electricity
to
generate carbon gas
from which the CNTs
form;
try
various
conditions until hit on
one
that
produces
significant amounts of
SWNTs.
Up to 70 %
Long bundles of tubes
(5 – 20 µm), with
individual diameter 1 –
2 nm.
Not very much interest
in this technique, as it
is too expensive, but
MWNT synthesis is
possible.
Primarily SWNTs, with
good diameter control
and few defects. The
reaction product is
relatively pure.
CNTs usually MWNTs Costly technique, it
and often riddled with requires expensive laser
defects.
and
high
power
requirement, but is
improving.
22
2.2.4
Growth Mechanism of Carbon Nanotubes
The way in which nanotubes are formed is not exactly known. The growth
mechanism is still a subject of controversy, and more than one mechanism might be
operative during the formation of CNTs. One of the mechanisms consists out of three
steps. First a precursor to the formation of nanotubes and fullerenes, C2, is formed on
the surface of the metal catalyst particle. From this metastable carbide particle, a rodlike carbon is formed rapidly. Secondly there is a slow graphitization of its wall, as
demonstrated in Figure 2.8 (Daenen et al., 2003). This mechanism is based on in-situ
TEM observation. The exact atmospheric conditions depend on the technique used,
as they are specific for a technique. The actual growth of the nanotube seems to be
the same for all techniques.
Figure 2.8: Visualization of a possible carbon nanotube growth mechanism (Daenen
et al., 2003).
There are several theories on the exact growth mechanism for nanotubes. One
theory postulates that metal catalyst particles are floating or are supported on the
substrate. It presumes that the catalyst particles are spherical or pear-shaped, in
which case the deposition will take place on only one half of the surface (this is the
lower curvature side for the pear shaped particles). The carbon diffuses along the
23
concentration gradient and precipitates on the opposite half, around and below the
bisecting diameter. However, it does not precipitate from the apex of the hemisphere,
which accounts for the hollow core that is characteristic of these CNTs.
For supported metals, CNTs can form either by ‘extrusion’ (also known as
‘root growth’ and ‘base growth’) (Figure 2.8) in which the nanotube grows upwards
from the metal particles that remain attached to the substrate, or the particles detach
and move at the head of the growing nanotube, labeled ‘tip-growth’ as shown in
Figure 2.8. Depending on the size of the catalyst particles, SWNT or MWNT are
grown.
Fonseca et al. (1998) explained the growth of tubules during the catalytic
decomposition of acetylene in three steps: the decomposition of acetylene, the
initiation reaction and the propagation reaction. First, dehydrogenative bonding of
acetylene to the catalyst surface will free hydrogen and produce C2 moieties bonded
to the catalyst coordination sites. These C2 units are the building blocks of the
nanotubes. Second, at the initial stage, the first layer of C2 units diffusing out of the
catalyst remains at a Van der Waals distance from the C2 layer coordinated to the
catalyst surface. Third, the C2 units are inserted between the catalyst coordination
sites and the growing nanotube. From the catalyst surface, a new C2 unit will again
displace the previous one that becomes part of the growing tube and so on. The
length of the nanotubes is directly proportional to the reaction time and it is possible
to interrupt the reaction and continue it without affecting the quality of the produced
nanotubes. Whereas, the turbostratic structure of nanotube depends on the reaction
temperature.
According to Kibria et al. (2001 and 2002), by observing the SEM images, if
each nanotube ended by showing a white spot, this indicates that the CNTs were
grown by tip growth mode. These white spots originated from the metal clusters. The
carbon feed stock from acetylene dissolved in metallic Co to form metastable cobalt
carbide. This cobalt carbide continues dissolving more carbon and became
oversaturated which then precipitated graphitic carbon in the forms of nanotubes by
keeping the metallic Co cluster at the top end of the CNTs.
24
Catalytically decomposed carbon species of the hydrocarbon are assumed to
dissolve in the metal nanoparticles and after reaching supersaturation, precipitate out
in the form of fullerene dome extending into a carbon cylinder with no dangling
bonds and hence, minimum energy. When the substrate-catalyst interaction is strong,
a CNT grows up with the catalyst particle rooted at its base (the base growth model).
When the substrate-catalyst interaction is weak, the catalyst particles are lifted up by
the growing CNT and continue to promote CNT growth at its tip (the tip growth
model). Formation of SWNT or MWNT is governed by the size of the catalyst
particle. Broadly speaking, when the particle size is a few nanometers, SWNT form,
whereas particles a few tens of nanometers wide favour MWNT formation (Ando et
al., 2004).
Yen et al. (2004) had prepared CNTs by thermal chemical vapour deposition
using acetylene on Ni coated silicon substrate. They postulated that the growth of
CNTs involves three main steps: (i) decomposition of hydrocarbon gas at the surface
of catalyst nanoparticles; (ii) diffusion of the resultant carbon atom in the
nanoparticles to form the nucleation seed; and (iii) precipitation of carbon atoms at
the nanoparticles interface to form nanotubes. The size and chemical composition of
the metal nanoparticles determines the diameter and structural perfection (degree of
graphitization) of the CNTs. The existence of catalyst particle at the tip of the
nanotubes confirms the fact that catalyst plays an important role in determining the
CNTs structure.
2.2.5
Applications of Carbon Nanotubes
Along with the improvement of the production techniques for nanotubes,
progress is also being made in their application. Recently, enormous interests have
been aroused by the discovery (Dillon et al., 1997) and reproduction (Chambers et
al., 1999 and Chen et al., 1999) of high hydrogen absorption capacity in carbon
nanotubes, although CNTs have relatively small surface area and pore volume. Lee
et al. (2000) have reported results of calculation for hydrogen storage behaviour in
25
SWNTs by density functional calculations. If these encouraging experimental results
can be reproduced easily and the large scale production of carbon nanotubes made
available in the near future, it will be possible to produce fuel cells using CNTs as
hydrogen storage medium.
The estimated high Young’s modulus and tensile strength as well as the
lightweight of CNTs (Andrews et al., 1999) of the nanotubes has lead to speculations
for their possible use in composite materials with improved mechanical properties
and even for bullet proof vests. In addition, their exceptional flexibility and stiffness,
coupled with their high aspect ratio has generated great interest for use as scanning
probe microscopy (SPM) tips. Several firms around the world are developing
nanotube field emitter arrays for use in flat panel displays and this may be the first
large-scale commercial application of nanotubes (Kong, 2002). Another area
receiving great attention is using CNTs as high-density ion storage material for
lithium and hydrogen batteries (Papadopoulos, 2000). One possible future field of
application for CNTs is in molecular electronics. Their small size, ability to be either
metallic or semiconducting and the low heat dissipation of metallic SWNTs are very
desirable for nanoscale electronic devices and their interconnections.
CHAPTER 3
CATALYTIC CHEMICAL VAPOUR DEPOSITION (CCVD)
3.1
An Overview of Catalytic Chemical Vapour Deposition (CCVD) Method
Recently many researchers have turned their attention onto the development
of methods for the production of CNTs in large-scale to realize their speculated
applications. Among these, catalytic chemical vapour deposition (CCVD) seems to
be one of the most promising methods for large-scale production of CNTs, stemming
from the relatively easy opportunity to upscale both the preparation and the
purification methods.
Compared to other synthesis methods, CCVD is a simple and economic
technique for synthesizing CNTs at low temperature and ambient pressure. It is
versatile in that it harnesses a variety of hydrocarbon in any state (gas, liquid or
solid), enables the use of various substrates and allows CNT growth in a variety of
forms, such as powder, thin or thick films, aligned or entangled, straight or coiled, or
even a desired architecture of nanotubes at predefined sites on a patterned substrate
(Ando et al., 2004). It also offers better control over growth parameters.
In fact, CCVD has been used for producing carbon filaments and fibre since
1959. Using the same technique, soon after the discovery of CNTs by Iijima, Endo et
al. reported CNTs grown from pyrolysis of benzene at 1100 °C, while Yacaman et
al. formed clear helical MWNTs at 700 °C from acetylene. After CNTs synthesis via
27
CCVD carried out by Yacaman et al. (1993) and Ivanov et al. (1994), modified or
new procedures were carried out.
3.2
Metal Oxides Supported Catalysts
In the field of heterogeneous catalysis, a number of oxides, clays and zeolites
have been used as catalyst support to disperse and stabilize metallic particles.
Catalytic properties of these solids are known to depend upon the interaction
between the support and the metal particles, which in turn depends on their method
of preparation. For the catalyst preparation, various methods are utilized, included
wet or dry impregnation, ion-exchange and mechanical grinding of support and metal
precursor. The catalytic activity also influenced by the concentration of the metal
solution. Iron (Fe), cobalt (Co) and nickel (Ni) supported catalysts were found to be
most active in growing CNTs. The most frequently used hydrocarbons are acetylene,
ethylene and benzene, due to their high carbon content.
Transition metals (Fe, Co and Ni) are the most commonly used catalysts for
CNT growth, since the phase diagram of carbon and these metals suggest finite
solubility of carbon in these transition metals at high temperatures (Ando et al.,
2004). The catalyst particle size has been found to dictate the tube diameter. Hence,
metal nanoparticles of controlled size can be used to grow CNTs of desired diameter.
The material, morphology and textural properties of the substrate greatly
affect the yield and quality of the resulting CNTs. Zeolite supports with catalysts in
their nanopores have resulted in significantly higher yields of CNTs with a narrow
diameter distribution (Ando et al., 2004). Alumina materials are reported to be better
catalyst supports than silica owing to their strong metal-support interaction, which
allows high metal dispersion and thus a high density of catalytic sites. Such
interactions prevent metal species from aggregating and forming unwanted large
clusters that lead to graphite particles or defective MWNTs. The key to obtaining
28
high yields of pure CNT is achieving hydrocarbon decomposition on catalyst sites
alone and avoiding spontaneous pyrolysis.
According to Konya and Kiricsi (2004), Si-, Mg- and Al- containing
materials (silica, alumina, zeolites, magnesia) prove to be applicable as supports, the
different forms of carbon (graphite and activated carbon) are not suitable in CNT
production. It was also shown that increasing reaction time results in the formation of
larger amounts of amorphous carbon due to the deactivation of the catalytic sites.
Ivanov et al. (1995) and Luo et al. (2002) had reasoned that the catalytic
chemical vapour deposition (CCVD) has more advantages over arc-discharge
method. The results showed that CNTs grown over Fe and Co supported catalysts
possessed well-crystalline graphite layers, compared to the as-grown CNTs over Cu
and Ni catalysts. The observation of encapsulated metal particles in the CNTs can be
the outcome of a weak metal-support interaction. The SiO2 support inhibits carbon
condensation and stabilizes the metal dispersion. The very fine metal clusters can be
localized in the solid-state exchange zeolites in the small cages, supercages or
intercrystalline spaces. However, because of the steric limitations, only the metal
particles at the outer surface and in supercages could be available for CNTs growth.
The catalytic production of CNTs was investigated by Hernadi et al. (1996a)
using various Fe catalysts. For the catalyst preparation, Fe acetate was used in order
to obtain the oxide without any anionic contamination after calcinations. However,
Fe acetate is not completely soluble in any solvent, so small crystals of the salt would
adhere to the outer surface of the support, though their amount is not that high. It
seems that the zeolite supported catalysts prepared by impregnation method shows
higher activity compared to samples prepared by ion exchange method. In this study,
most of the Fe catalysts produced CNTs which are not regularly bent and no helices
are present. They also revealed that by using fixed bed flow reactor, deposition of
soot begins on the catalyst surface and outer surface of the nanotubes at reaction time
more than 1 hour.
29
Fonseca et al. (1996) work on supported Co and Fe catalysts had effectively
produced CNTs in large amounts by catalytic decomposition of acetylene. The
results showed that the hydrogenated catalysts yield less carbon deposit compared to
the untreated catalysts. During the hydrogenation treatment, the catalyst was exposed
to high temperature. As a consequence Co particles had a chance to migrate from the
inner pores to the outer surface and assembled. Thus, reducing dispersion of the Co
particles and giving larger particle size resulting in higher amount of amorphous
carbon and thick nanotubes.
Silica and zeolite supported transition metal (Co, Cu and Fe) catalysts have
been prepared and utilized for the decomposition of acetylene at 700 °C by
Fonseca et al. (1998). In the absence of metal particles on the support, only the
homolytic decomposition of carbon occurs, generating amorphous carbon and fibers.
It can be concluded that catalysts calcined at 450 °C performed better in forming
CNTs.
In the production of CNTs over Ni/Cu mixture catalyst via CCVD process,
the Cu metal was found to have no effect on the growth of CNTs as discussed by
Jong et al. (2002). The Cu atoms would cluster rather than uniformly disperse on the
surface of the substrate. As Cu atoms are not the active catalytic sites, there is no
carbon deposited on the area covered by Cu clusters. These clustered areas would
limit the reactive areas of adsorption and precipitation of carbon and lead to
formation of filament instead of tubular structure. Therefore, Cu metal is not suitable
to be used as the catalyst in the synthesis of CNTs.
Similar to the synthesis of MWNTs, Colomer et al. (1999) have succeeded in
synthesizing SWNTs over different types of supported metal catalysts by
decomposition of ethylene. The Co and Fe catalysts seem to be more active than Ni
catalyst. For the catalysts prepared from a mixture of metals, the best activity is
found for Co/Fe and Fe/Co/Ni mixtures supported on alumina. Possibly, the
preparation of catalysts with mixtures of metals induces a better dispersion of the
metals on the surface of support and results the formation of alloy followed by
segregation. They also proposed that the support has no strong influence on the
30
growth mechanism, but plays an important role in the dispersion of the metal on the
surface and control the catalytic cracking of the hydrocarbon.
Colomer et al. (2000) have shown that the large-scale production of SWNTs
is possible by CCVD using the decomposition of methane over supported Co, Fe, Ni
or mixture catalysts at 1000 °C. In this work, metal particles larger than 20 nm are
found inactive in the synthesis of SWNTs. Each SWNTs bundle is hypothesized to
grow from a single metal particle. It was determined that the Co-Fe mixture catalyst
gives the best activity.
The reason for different quality and quantity of CNTs generated on supported
catalysts is unclear. Kukovecz et al. (2000) suggested several factors, such as the
solubility of carbon in the metal phase, the formation of metal particles with different
particle size distribution and the hindered mobility of metal atoms toward the cluster
formation. The varying acidity of the support could influence the reducibility of the
metal species. In the Al2O3-SiO2 mixture system, the support is known to have
maximum acidity at around 20 wt.% Al content, beneficially increase the rate of
acetylene decomposition. The supported Fe catalysts are efficient in producing
MWNTs of somewhat lower quality, as the Co catalysts produced good quality
nanotubes, however, the Co/Fe bimetallic catalyst produced the best quality and
quantity of CNTs.
The influence of Co concentration, reaction conditions, catalyst preparation
method and the nature of the support on the formation of CNTs were investigated by
Piedigrosso et al. (2000). The results showed that with Co loading higher than
5 wt.%, the activity of the catalyst is decreased. When the reaction time was longer
than 1 hour, amorphous carbon was formed and deposited not only on the surface of
the catalyst but also on the external walls of the tubes. This effect may be due to the
increasing ‘graphite’ surface area of the external surface of the nanotubes, which
favourable for homolytic decomposition of the reactant. The flow rate of acetylene
(reactant) and nitrogen (carrier gas) need to be coordinated in order to achieve
optimum residence time of acetylene on the catalyst active sites.
31
For the silica supported catalysts prepared by porous impregnation, the metal
particles are in the inner pores of the support and they must first migrate from the
pores to the outer surface to be active for the formation of CNTs. On the other hand,
for silica supported catalysts prepared by ion-adsorption precipitation, most of the
metal precipitates are on the catalyst surface and hence are directly active. Larger
numbers of helical tubes were form over Co-silica catalysts prepared by this method.
Piedigrosso et al. (2000) also revealed that the diameter of the catalyst particles
should be close to the inner diameter of the nanotubes, while the number of
turbostratic layers of each nanotube might depend on the flow rate of acetylene per
active site. The outer layers of the MWNTs seem to be formed simultaneously with
the inner layer.
In Willems et al. (2000) as well as Mukhopadhyay and Mathur (2003) work,
Co/Mo, Co/V and Co/Fe mixtures supported either on zeolite or corundum alumina
were used as catalysts to produced MWNTs. Depending on the catalyst used, it is
possible to control the carbon deposit and the outer diameter distribution of
nanotubes. Mo and V alone are totally inactive in the production of nanotubes.
Co/Mo catalysts are capable to produce turbostratic nanotubes containing no
amorphous carbon on their outer surface. The thickness of the tubes can be directed
through the dispersion of catalytic particles on the surface of the support and the flow
rate of acetylene. Optimized dispersion of the catalytic particles reduced the quantity
of acetylene coming into contact with the active sites, resulting thin nanotubes with
only a few concentric walls. It was also reported that it was much easier to separate
CNTs from the zeolite than from the alumina.
It is well establish and documented that the support plays an important role in
determining the catalytic activity of a metal present in it. The state of the metal on a
support depends on the kind of metal-support interaction, which in turn depends on
the nature of the support and preparation method. According to Nagaraju et al.
(2002), the diameter of CNTs growing on a metal particle mainly depends on the
dispersion of the metal particles on the support.
32
As reported by Nagaraju et al. (2002), hydrated alumina is basic due to the
presence of surface hydroxyl groups. When this alumina mixed with salt solutions of
Fe/Co, there is a possibility of better mixing and homogeneous distribution of metal
particles than that in the case of commercial Al2O3 which is neutral. Hence Fe and
Co alloy supported on hydrated alumina resulting in the formation of higher
quantities of MWNTs. Metal(s) when supported on SiO2 resulted in only a moderate
generation of carbon deposit. Lower activity of SiO2 compared to Al2O3 supported
metals indicate that the size of the metal particles produced in the former is
somewhat larger than the average nanometer scale clusters required for producing
CNTs.
Nhut et al. (2002) have carried out a comparison study between commercial
purified CNTs and their homemade CNTs. The commercial CNTs were produced by
high temperature arc-discharge method, while the homemade CNTs were
synthesized by the catalytic decomposition of acetylene over silica supported Fe
catalyst (20 wt.% of Fe) at 750 °C. An amorphous carbon layer was presented on the
outer surface of the commercial CNTs which could be due to carbon formation by
pyrolysis of the reactant mixture during the synthesis. Additionally, some residual
metallic iron catalyst was encapsulated inside the graphene layers. In other case, the
graphene layers of the homemade CNTs were not so straight and parallel. The
disorder between carbon planes was assigned to the relatively low temperature
catalytic growth process. However, it was noted that almost no amorphous carbon
was found on the outer surface of the homemade CNTs.
The MgO supported catalysts were also active towards formation of MWNTs
as discussed by Willems et al. (2002). Again, the high activity of mixed metals in
producing nanotubes was proven. Metals are differently dispersed on various
supports, owing to differences in interactions between the metal and support surface.
The MWNTs obtained appear as aggregates and many irregularities are observed in
the nanotubes walls. Willems et al. (2002) make the hypothesis that the presence of
aggregates rather than bundles of nanotubes are due to a poor graphitization process.
The decomposition of the acetylene and the nanotubes growth are not well
33
coordinated. The catalyst is so active in the decomposition that the arrangement of
the nanotube building units is not efficient.
Perez-Cabero and co-workers (2004) described the synthesis of carbon
nanotubes by catalytic decomposition of acetylene at 700 °C over several Fe-silica
catalysts. It is found that the CNT products highly dependent on the metal content,
the Fe precursor employed and the preparation conditions of the silica supported
catalyst. Meanwhile, the iron-nitrates produced the highest activity in acetylene
decomposition, but the iron-phthalocyanine produced the best CNTs in quality, size
homogeneity and amorphous carbon free. The impregnation method allows more
homogeneous metallic dispersion on the silica support than the depositionprecipitation method.
In the supported catalyst method of metal catalyzed CNT synthesis, the
support can interact both physically and chemically with the catalyst particle (Tauster
et al., 1978 and Yudanov et al., 1997). Physically, the solid support can control the
size of the metal catalyst particle. After solvent evaporation, the metal salt particle
size is determined by the solution concentration and by the liquid volume that had
been contained within a given pore. The support can also act to disperse the metal
catalyst particles formed upon decomposition of the metal salt precursors by
physically holding the particle in place.
In addition, the support can also interact chemically with the catalyst particle
and dramatically affect its catalytic activity. Acquisition of negative charge by the
metal catalyst from the substrate can enhance its catalytic activity by strengthening
back donation of electron density into antibonding orbitals of the adsorbate
(hydrocarbon molecules). This electron sharing between catalyst and adsorbate
sufficiently weakens the bonding within the adsorbate, resulting in its dissociation
(Davis and Klabunde, 1982). Dissociation of a hydrocarbon on the metal catalyst is
the first step in a widely accepted mechanism for metal catalyzed carbon nanotube
synthesis.
34
The metal catalyst particle can acquire negative charge through interaction
with either the metal cation or oxygen anion sites in the support (Yudanov, et al.,
1997 and Davis and Klabunde, 1982). A strong interaction between the catalyst
nanoparticle and the cationic sites on the support is known as strong metal-support
interaction (SMSI) (Vander Wal et al., 2001). SMSI arises from partial reduction of
the oxide which enables the metal cation of the support to donate partial negative
charge to the supported metal nanoparticle. H-atoms generated by the supported
catalyst initiate the partial reduction of the oxide and are thus integral to the SMSI
process.
For support oxides that are not readily reduced or catalysts not capable of H2
dissociation (enabling reduction of the support), another potential interaction
involves the catalyst particle and the exposed oxygen anion sites on the support
(Tauster et al., 2001 and Yudanov, et al., 1997). The anions, with their large negative
charge, can act as Lewis base sites donating negative charge to the metal
nanoparticle (weak metal-support interaction). In a similar manner, the partial
negative charge accepted from the oxide can enhance the catalytic activity of the
metal nanoparticle.
Vander Wal et al. (2001) reasoned out that the structure of the catalyst
support (include the lattice steps, defects and crystalline edge sites) are apparently
not critical to the interactions with the supported metal catalyst and would not affect
the nanotube growth. The pore density and pore uniformity of the support is more
crucial to form a narrow size range of the metal catalyst nanoparticles and
subsequently produce narrow size distribution of the nanotubes. Calcination supports
the forming of a highly dispersed catalyst oxide. A larger amount of negative charge
is transferred to the highly dispersed catalyst particles from the partial reduced
support via either SMSI or Lewis base sites.
Kathyayini et al. (2004) studied the role of Ca and Mg oxides, hydroxides
and carbonates as supports for Fe, Co and a mixture of Fe/Co catalysts in the
production of CNT by CCVD method. This work emphasize that the nature of the
support, the catalyst and the type of interaction between the metal and the support
35
had a great influence on the yield of carbon deposit. On a given type of Ca and Mg
salt supports, the catalytic activity of the metal ions towards carbon deposit was
found to vary in the same order, i.e. Fe/Co > Co > Fe. Further, Fe/Co supported on
CaCO3 or MgO are the best catalysts.
Methane decomposition into filamentous carbons was carried out over
Fe2O3/Al2O3 and Fe2O3/SiO2 as discussed by Takenaka et al. (2004). The catalytic
activity of Fe2O3/Al2O3 was higher than that of Fe2O3/SiO2. The difference of the
catalytic performance can be explained by the particle size of the catalytically active
species. Fe2O3 in the fresh catalyst would be reduced stepwise with methane during
the reaction, i.e. Fe2O3 → Fe3O4 → FeO → α-Fe metal and Fe3C (cementite). α-Fe
metal as well as Fe3C are the active species for the decomposition. Baker et al.
(1987) proposed that interaction of α-Fe with silica supports brings about a decrease
in carbon solubility and diffusion rate of carbon atoms in the metal. The Fe2O3/Al2O3
produced MWNTs and chain-like carbon fibers, though Fe2O3/SiO2 formed many
spherical carbon units, in addition to chain-like carbon fibers.
In the work by Shajahan et al. (2004), CNTs were grown over impregnated
5 – 40 wt.% Co/Mo-MgO catalysts by decomposition of C2H2 at 800 °C. It was
observed that the production of SWNTs and MWNTs was highly dependent on the
amount of metal loading. The Mo-MgO catalyst was inactive in growing CNTs.
However, in the Co/Mo-MgO catalyst, the inactive Mo generated CoMoO4 species,
which was that main precursor for the formation of active Co cluster to grow CNTs.
Small cluster of Co was generated from small particle of CoMoO4 species and grew
SWNTs over 5 and 10 wt.% Co/Mo-MgO catalysts. The particle size of CoMoO4
increased with increasing metal loading and thus the size of released Co cluster
increased.
Liu, B.C. et al. (2004) had studied the production of SWNTs by catalytic
decomposition of C2H2 over Al2O3 and MgO supported Fe/Mo bimetallic catalyst.
The results indicated that Al2O3 and MgO support materials are useful to a large-
36
scale synthesis (mass production) of high quality SWNTs and also demonstrated that
C2H2 can be very ideal feedstock to synthesize SWNTs over Fe/Mo-MgO catalyst.
Carbon nanofibers (CNFs) similar to MWNTs were synthesized from a
Fe2O3-Al2O3 composite ceramic catalyst through CCVD by Maruyama et al. (2004).
The Fe2O3-Al2O3 system has three phases: Fe2O3 solid solution, Al2O3 solid solution
and high temperature phase FeAlO3. It was revealed that different CNF construction
and structures were obtained depending on the catalyst phase. In the case that the
catalyst used is FeAlO3, aligned CNF which similar to MWNTs was produced. This
is because Fe was consumed in formation of catalyst particles for CNF growth,
FeAlO3 concentration became poor in Fe. Subsequently, Al2O3 precipitation has been
induced as a secondary phase on FeAlO3 and these precipitation planes became the
cleavage plane. On these cleavage planes, Fe catalyst particles formed at the same
time and condition, as a result, the catalyst particles are homogeneous and capable to
grow CNF in one direction forming aligned CNF.
Liu, D.Y. et al. (2004) had grown CNTs using a simple Fe catalyst prepared
by electrochemical deposition method. Although the electrochemical deposition is a
cheap and simple technique to deposit Fe particles on the substrate, but the
effectiveness of the catalysts to grow CNTs is relatively low. Only the Fe catalyst
particles with size less than 100 nm are effective in growing CNTs. The growth of
CNTs is affected by the morphology of the catalyst.
Recently Lu, A.H. et al. (2005) have discovered the formation of amorphous
CNTs by a CCVD procedure using mesoporous silica SBA-15 as matrix with Fe2O3
catalyst. The amorphous CNTs exhibit very special geometry with the carbon atom
layers packed vertical to the tube axis. The layers of carbon in these CNTs are
randomly displaced with respect to each other showing low degree of graphitization.
The authors proposed that the ‘tip growth’ mechanism took place as the Fe
nanoparticles was observed encapsulated in the tube walls of the amorphous CNTs.
An analogous phenomenon of catalyst migration in synthesis of CNTs has also been
observed by Hernadi et al. (1996b) and Piedigrosso et al. (2000). The external
37
diameter of the CNTs is larger than that of the SBA-15 pores because of the
additional thickness of the tube walls.
Flahaut et al. (2005) had suggested that catalyst with the higher specific
surface area possess better homogeneity and leading to a more homogeneous
population of smaller catalytic metal particles. Thus, this catalyst is capable in
producing CNTs with fewer walls and lower diameters as well as a narrow diameter
distribution.
Coiled CNTs were successfully prepared by Lu, M. et al. (2004a and 2005)
via CCVD over Co particles supported on silica gel under low gas flow rates. When
CCVD process is scaled up for the industrial production of CNTs, it may cause
environmental and safety concerns due to the consumption of hydrocarbons, such as
acetylene. In addition, the yield of the CNTs do not require high gas flow rate for the
deposition, consequently most of the gases in the CCVD process are discharged into
the atmosphere without participation in the chemical reaction. As a result, the flow
rate of C2H2 can be reduced to 5 – 20 sccm.
3.3
Anodic Aluminium Oxide (AAO) Template Supported Catalysts
Porous
anodic
aluminium
oxide
(AAO)
templates
were
prepared
electrochemically from aluminium metal. These membranes are unique materials,
with a high density of straight, cylindrical pores (up to 1011 pores/cm2) of a very
uniform size, and can be obtained with different pore sizes ranging from 5 nm to
more than 200 nm (Fuertes, 2002). The parameters of the AAO template can be well
dominated by controlling of oxidized time, temperature, anodization voltage and pH
value choice of acids (Li et al., 2003). The AAO template also exhibits good
chemical inertia and physical stability. These nanopores constitute an ideal host for
carrying out the synthesis of nanomaterials because of their special characteristics.
The schematic diagram of anodic aluminium oxide (AAO) template is shown in
Figure 3.1.
38
Several authors have explored template synthesis within the straight
nanochannels of different kinds of nanomaterials based on carbon, metal, polymer or
mixed precursors, with various configurations, i.e. nanotubes, nanofibrils or coaxial
rods. However, AAO template has low thermal stability, AAO membranes has a
tendency to warp during heat treatment (Jeong and Lee, 2003). In addition, AAO
membranes are brittle and easily cracked.
Anodic Aluminium Oxide
Figure 3.1: The schematic diagram of an anodic aluminium oxide (AAO) template.
Over recent years, ordered CNTs arrays have generally been prepared using
the template synthesis method. The most relevant feature of this method is that it
enables high-density arrays of well-aligned carbon nanotubes, of a similar diameter,
wall thickness and length to be prepared (Fuertes, 2002). The CNTs arrays perfectly
copy the three-dimensional structures of the nanochannel in the template, where the
resulting outer diameter of the CNTs matches the inner diameter of the channels.
Template synthesis method can also improve the selectivity and uniformity of CNTs.
The CNTs growth in AAO template is one of the more promising approaches owing
to the possibility of tailoring the size and density of the AAO template pores.
39
Two different procedures have been reported for the synthesis of carbon
nanotubes inside the nanochannels of an AAO template: (a) by liquid phase
impregnation of the AAO template with a polymer (e.g. polyacrylonitrile,
polyfurfuryl alcohol) followed by thermal carbonization and (b) by thermal
decomposition of a hydrocarbon within the nanochannels at high temperatures
(CCVD) (Fuertes, 2002). The first method is difficult to control the deposited carbon
accurately. On the other hand, CCVD is simpler and allows a more precise control of
the growing of carbon nanotubes.
In the conventional CNTs arrays produced in AAO template, alumina itself is
a catalyst in the hydrocarbon decomposition process due to the Lewis acid nature of
its surface sites (Scokart and Rouxhet, 1982). Although alumina is far less effective
than metal, its presence complicates the mechanism of CNTs growth. The CNTs are
formed by the joint catalytic activities of metal nanoparticles and alumina walls (Jing
et al., 1998). The metal catalyst might merely initiate the growth of the tube through
the initial catalytic decomposition of hydrocarbon. Subsequently, carbon as carbide
transport along the surface of the metal nanoparticles to the alumina channel wall,
where tube growth proceeds. In this case, the inside of the channel wall covered by
carbon flakes that form a tube-like structure.
Works by Lu et al. (2004b) have shown that high density CNTs can be
fabricated in silicon coated AAO template by CCVD. With this approach, the silicon
coating on the channel wall may contribute to the formation of CNTs by eliminating
the catalytic behaviour of the alumina, in which the CNTs are only catalyzed by the
small metal nanoparticle independently throughout the entire growth process. The
nanochannel with silicon coating is akin to a nanometer-sized test tube or
nanoreactor for the reaction between the metal nanoparticles and the carbon atoms.
The base growth mechanism mainly responsible for the CNTs growth in the silicon
coated nanochannel. Moreover, the pressure and flow rate of the hydrocarbon inside
the channels is different than on the outside. For this reason, the alumina channels
with larger diameter and thinner silicon coating may be favoured for its function as
nanoreactors.
40
3.4
Types of Carbon Precursor
Shajahan et al. (2004) explained that equilibrium is essential between the
acetylene (C2H2) feed stock and the surface of catalyst particles during the reaction.
From their results, the equilibrium appeared at 10 – 15 sccm (standard cubic
centimetres per minute) of C2H2 flow rate. Excessive carbon deactivates catalyst and
deposits high amorphous carbon.
A new type of precursor, coal gas was utilised by Qiu et al. (2004) in the
preparation of SWNTs by CCVD method. The SWNTs have been successfully
synthesized from coal gas with ferrocene as catalyst. The main components of the
coal gas are H2, N2, CH4, CO, CO2 and C2 – C3 hydrocarbons. Undoubtedly, the CO
and CH4 species would be the main carbon source for forming CNTs, while the H2
and N2 species act as carrier gas.
3.5
Reaction Temperature
Generally, low temperature CCVD (600 – 900 °C) yields MWNTs, whereas a
higher temperature (900 – 1200 °C) reaction favour SWNTs growth, indicating that
SWNTs have a higher formation energy. This presumably due to SWNT has small
diameter and this result in high curvature and high strain energy (Ando et al., 2004).
This could explain why MWNTs are easier to grow from most hydrocarbons than
SWNTs. Common efficient precursors of MWNTs, such as acetylene and benzene
are unstable at higher temperature and lead to deposition of certain amounts of
carbonaceous compounds.
An attempt has been made by Kukovecz et al. (2000) to synthesize MWNTs
using Co, Fe and Ni supported on different types of silica-alumina to investigate the
rules governing their nanotube producing activity. Acetylene was used as the source
of carbon. The effect of reaction temperature in the range of 497 – 697 °C (770 –
41
970 K) and the flow rate of the hydrocarbon has been investigated. The amount of
deposited carbon increased with increasing reaction temperature and the flow rate of
acetylene. At 697 °C, each catalyst showed high production of carbon nanotubes and
the amount of helical tubes is also more pronounced. The yield of tube formation was
very low at 597 °C while the catalysts were inactive at 497 °C. However, higher
temperatures can induce the formation of amorphous carbon.
A comparison study had been carried out by Nagaraju et al. (2002) on the
catalytic activity of Fe and Co supported on metal oxides in the production of CNTs.
It is observed that at 500°C, none of the catalysts was active in the formation of
CNTs. Generation of CNTs was observed only in the reactions carried out above
500 °C. A best yield of MWNTs is resulted at 700 °C on hydrated alumina and
containing a mixture of Fe and Co on it.
Andrews et al. (2003) described the synthesis of MWNTs by the
decomposition of a hydrocarbon vapour over a dispersed Fe catalyst. The data
showed that in the temperature range of 700 – 775 °C, MWNTs production rates is
the highest. Outside this range, a concomitant increase in the conversion of carbons
into amorphous carbon was detected.
The synthesis of MWNTs using Fe catalyst deposited in situ on quartz
substrates (Jacques et al., 2003) showed higher reaction temperature leading to the
growth of larger catalyst particle that control the outer tube diameters. They
concluded that the nanotube outer diameter is directly correlated to the catalyst
particle size.
Beta zeolite supporting Co particles also permits the growth of MWNTs by
ethylene CCVD as reported by Ciambelli et al. (2005). The nanotubes are sinuous
and entangled, with an inner diameter determined by metal particle dimensions.
Carbon deposit yield increases with reaction time and temperature, while reaction
yield increases with reaction time. Selectivity of nanotubes at about 70 % was
obtained after 60 min at 700 °C. The increase of reaction time leads to a greater wall
42
thickness, resulting from the increase of wall number. This increase is likely due to a
progressive ordering of amorphous carbon formed on the external surface of CNTs.
The carbon samples have surface areas increasing with reaction time and
temperature, as a combined effect of increased selectivity to nanotubes and porosity
of amorphous carbon.
3.6
Latest Catalytic Chemical Vapour Deposition (CCVD)
Valiente
et
al.
(2000)
have
applied
two
different
techniques,
thermogravimetric analysis and UV-Raman, to analyze carbon deposition by
catalytic decomposition of acetylene over an iron-based catalyst. The UV-Raman
result shows the chemical conversion of the adsorbed carbon species to olefinic
species, polymerised olefinic species, aromatic species and finally to carbon
nanotubes. Valiente et al. (2000) concluded that carbon deposition has an induction
period of about 5 minutes and the reaction temperature enhances the formation of the
nanotubes. The XRD pattern of the raw iron-based catalyst shows characteristic of
Fe2O3.
Zou et al. (2004) had synthesized aligned carbon nanotube arrays on
Fe2O3/SiO2/Si substrates using simple thermal chemical vapour deposition (STCVD)
apparatus at atmospheric pressure, which was made possible by substituting complex
apparatus for the growth of CNTs. The alignment of the CNTs may be controlled by
Van der Waals interactions among the CNTs and the overcrowding effect.
Moshkalyov et al. (2004) had presented a study of MWNTs growth using two
different catalytic chemical vapour deposition (CCVD) techniques, i.e. low pressure
plasma-enhanced and atmospheric pressure thermal CVD. Thin films of Ni were
used as catalyst. In the case of low pressure plasma-enhanced CVD, it was found that
it is impossible to achieve adequate growth of MWNTs under low temperature of
500 – 550 °C. Meanwhile, the atmospheric pressure thermal CVD was effective in
growing more regular and straighter nanotubes at temperature higher than 700 °C.
43
This technique gave the maximum growth rate as high as ~ 1 – 2 µm/min with tubes
length up to 30 – 40 µm. In addition, the results also shown that the nanotubes
growth density depends strongly on the catalyst treatment.
The latest alcohol catalytic chemical vapour deposition technique was used
by Inoue et al. (2005) to investigate the relationships among catalyst species, alcohol
and synthesis temperature (600 – 1000 °C). For ethanol, the Mo/Co catalyst
functions best among the three catalysts, i.e. Fe/Co, Mo/Co and Rh/Pd. In the case of
methanol, Fe/Co work more efficiently than Mo/Co.
However, these latest CCVD methods required complicated apparatus and
power supply and the yield of CNTs are not promising. Therefore, in this work the
typical CCVD was applied but with new approach in the preparation of supported
catalysts in order to enhance the CNTs growth.
CHAPTER 4
RESEARCH METHODOLOGY
4.1
Preparation of Supported Catalysts
4.1.1
Preparation of Alumina and Molecular Sieves Supported Catalysts
Two types of catalyst support were used, i.e. alumina beads (aluminium oxide
beads with diameters 2 – 4 mm, supplied by Sigma-Aldrich) and molecular sieve
beads (type 3A, supplied by Fisher). Molecular sieves (MS) are a mixture of metal
oxide, mainly alumina (Al2O3) and silica (SiO2). Both of the supports were obtained
from commercial source.
The supported catalysts were prepared by soaking technique of wet
impregnation method. The supports in beads form were dried in oven for 24 hours at
120 °C to eliminate water vapour and impurities. After cooling, all the supports were
impregnated with either Co or Fe acetate salt aqueous solutions. An amount of 1.00 g
of the dried support beads (either alumina or MS) was weighed. A stoichiometric
amount of hydrated cobalt acetate salt, Co(CH3COO)2.4H2O (Fluka) of 2.5 wt.% of
Co relative to the weight of the support (Appendix A.1) was diluted with minimum
amount of distilled water. A pinkish coloured solution was obtained. The support
was immersed in the solution until saturation, and then dried in oven for 30 minutes
at 90 °C. This procedure was repeated until all of the solution was adsorbed by the
support. The sample was then dried for 24 hours at the same temperature. As for the
non-calcined sample, the resulting supported catalyst was ready to be used in the
45
synthesis of carbon nanotubes (CNTs). For calcination treatment, the catalyst was
placed in a furnace and heated to 450 °C for 17 hours. For the preparation of
supported 3.0 wt.% Co catalyst, an amount of hydrated cobalt acetate salt containing
3.0 wt.% of Co were used.
The preparation of Fe supported catalysts was similar to the procedures above
except for the use of ferrum acetate salt, Fe(CH3COO)2 (Aldrich) as a replacement to
cobalt acetate salt. The calculation was shown in Appendix A.2. In the preparation of
Co and Fe mixture supported catalysts, equal amount of Co and Fe acetate salt were
used. For example, 1.25 wt.% of each acetate salt were mixed to obtain 2.5 wt.% of
metal catalyst loaded on the support (Appendix A.3).
4.1.2
Preparation of Anodic Aluminium Oxide (AAO) Template Supported
Catalysts
In this method, anodic aluminium oxide (AAO) template (Whatman Anodisc
47, 200 nm diameter pores and 50 µm thickness) was used as the catalyst support.
The anodic aluminium oxide (AAO) template is thin and poor in absorbing aqueous
solution, therefore, dip coating technique (impregnation method) has to be applied.
In order to get 0.5 M of cobalt acetate aqueous solution, a stoichiometric amount of
hydrated cobalt acetate salt, Co(CH3COO)2.4H2O (Fluka) (Appendix B.1) was
dissolved with distilled water in a 25 mL volumetric flask. The solution was shaken
well to make sure the salt dissolved completely. The pinkish solution was poured
into a petri dish and 2 pieces of AAO templates were immersed in the solution for
30 minutes. Then the wet AAO templates were dried in an oven at 90 °C for
30 minutes. The immersion and drying procedure were repeated 3 times. The
templates were then dried for 24 hours at the same temperature. Part of the sample
was then calcined at 450 °C for 17 hours. For the preparation of higher quantity of
Co catalyst, cobalt acetate aqueous solution of 1.0 M was used to dip coated the
templates.
46
The Fe supported catalysts were prepared using the same procedures except
for the use of ferrum acetate salt, Fe(CH3COO)2 (Aldrich) replacing cobalt acetate
salt. The calculation was showed in Appendix B.2. In the preparation of Co and Fe
mixture solution, same amount of Co and Fe acetate salt were used. To obtain 0.5 M
mixture solution, 0.25 M of Co solution was combined with 0.25 M of Fe solution
(Appendix B.3). The notations for all of the supported catalysts prepared in this work
are shown in Table 4.1. These notations will be used in the following discussion.
Table 4.1: Notations for the alumina, molecular sieves (MS) and Anodic Aluminium
Support
Oxide (AAO) template supported metal catalysts.
Loadings
of metal
catalyst
Al2O3
2.5 wt.%
3.0 wt.%
MS
2.5 wt.%
3.0 wt.%
AAO
0.5 M
1.0 M
Types of metal catalyst
Calcination*
Co
Fe
Co/Fe
8
Al–Co(2.5)
Al–Fe(2.5)
Al–Co/Fe(2.5)
9
Al–Co(2.5)Cal
Al–Fe(2.5)Cal
Al–Co/Fe(2.5)Cal
8
Al–Co(3.0)
Al–Fe(3.0)
Al–Co/Fe(3.0)
9
Al–Co(3.0)Cal
Al–Fe(3.0)Cal
Al–Co/Fe(3.0)Cal
8
MS–Co(2.5)
MS–Fe(2.5)
MS–Co/Fe(2.5)
9
MS–Co(2.5)Cal
MS–Fe(2.5)Cal
MS–Co/Fe(2.5)Cal
8
MS–Co(3.0)
MS–Fe(3.0)
MS–Co/Fe(3.0)
9
MS–Co(3.0)Cal
MS–Fe(3.0)Cal
MS–Co/Fe(3.0)Cal
8
AAO–Co(0.5)
AAO–Fe(0.5)
AAO–Co/Fe(0.5)
9
AAO–Co(0.5)Cal
AAO–Fe(0.5)Cal
AAO–Co/Fe(0.5)Cal
8
AAO–Co(1.0)
AAO–Fe(1.0)
AAO–Co/Fe(1.0)
9
AAO–Co(1.0)Cal
AAO–Fe(1.0)Cal
AAO–Co/Fe(1.0)Cal
* 8: without calcination; 9: treated with calcination.
47
4.2
Synthesis of Carbon Nanotubes (CNTs)
4.2.1
Design and Fabrication of Custom Built Catalytic Chemical Vapour
Deposition (CCVD) System
A simple thermal Catalytic Chemical Vapour Deposition (CCVD) system has
been designed and fabricated for synthesize of CNTs. The system is simple and
includes a cost effective fixed bed flow reactor. It is easy to handle and does not
require expensive power supply or high pressure reaction chamber.
The system can be divided into 3 major components: the gas sources, the gas
mixing component and the tube furnace. There are 2 types of gases used, acetylene
(C2H2) and nitrogen (N2). Each gas tank was connected to a regulator. The amount of
the gas was measured with a flow meter. Both of the gas streams were passed
through a moisture trap to make sure the gas was dry. The moisture trap contained
silica gel. The gas streams were controlled by an on-off valve before mixing up. The
function of the gas mixing chamber is to mix the gas uniformly prior to the reaction.
The gas mixture was measured again before it flown into the furnace.
The custom built reactor is a fixed bed flow reactor where the solid form
catalyst has been placed stationary in the high temperature zone within the furnace to
react with the gas source. The carbon deposition yield will deposit on the active
catalyst and the CNTs can be collected upon cooling of the system after the reaction.
A Carbolite CTF 12/65/550 model tube furnace was used. In the furnace, the
acetylene was heated up and reacted to the supported catalyst to synthesis CNTs. The
schematic diagram of the system is shown in Figure 4.1. Meanwhile Figure 4.2 and
Figure 4.3 showed the photographs of the CCVD system.
48
On-off Moisture Flow meter 1
valve 1 Trap 1
Flow meter 3
Gas mixing
chamber
Co/Fe supported
catalysts
Flow meter 2
Regulator 1
Gas in
On-off
valve 2
Gas out
Tube furnace
Vycor tube
(to fume
cupboard)
Ceramic boat
Moisture
Trap 2
Regulator 2
N2
C2H2
Tube furnace
Gas mixing component
Gas sources
Figure 4.1: Schematic diagram of the Catalytic Chemical Vapour Deposition
(CCVD) system (fixed bed flow reactor).
Nitrogen
(N2)
Tube
furnace
and
controller
Acetylene
(C2H2)
Gas mixing
component
Figure 4.2: Photograph of the fixed bed flow reactor.
49
Gas
mixing
chamber
Flow
meter
On-off
valve
Moisture
trap
Figure 4.3: Photograph of the gas mixing component in the CCVD system.
4.2.2
Production of Carbon Nanotubes
The experiment was carried out in a horizontal tube furnace at atmospheric
pressure. The prepared and calcined supported catalyst is spread on a ceramic boat,
which is placed inside a Vycor tube. Acetylene (C2H2) with purity of 99.9999 % was
used as the source of carbon for the production of carbon nanotubes (CNTs) and
nitrogen (N2) gas (99.9 %) acts as the carrier gas. After purging with nitrogen at 120
standard cubic centimetres per minute (sccm) for 15 minutes, acetylene was allowed
to flow at 15 sccm. The decomposition of C2H2 was performed at 700 °C (5 °C/min)
with reaction time of 30 minutes. The percentage of carbon deposited due to the
catalytic decomposition of C2H2 was obtained from Equation 4.1 (Nagaraju et al.,
2002):
Carbon deposit (%) = 100
( mtot − m cat )
m cat
(Equation 4.1)
where mcat and mtot are the mass of the catalyst before and after the reaction,
respectively.
50
4.3
Characterisation Techniques
In this research, combinations of various characterisation tools were adopted
in elucidating the bulk and surface complexity of the active catalysts and carbon
nanotubes (CNTs) produced. The combinations of several characterisation
techniques undoubtedly provide valuable interpretable data concerning the structure
and morphological properties of the supported catalysts and the CNTs synthesized.
Thus,
catalyst
characterisation
provides
more
fundamental
and
practical
understanding of the CNT growth in relating the type, composition and particle size
of catalyst with catalytic performance. There are five main techniques that will be
discussed: X-ray Diffraction Analysis (XRD), Scanning Electron Microscopy
(SEM), Energy Dispersive X-ray Analysis (EDX), Field Emission-Scanning Electron
Microscopy (FE-SEM) and Transmission Electron Microscopy (TEM). The SEM,
FE-SEM and TEM techniques are generally grouped as Electron Microscopy (EM).
4.3.1
X-ray Diffraction (XRD)
In heterogeneous catalysis, even though the active sites are located on the
surface of the catalysts, the bulk structure plays a significant role too. This is because
the catalysts characteristics depend strongly on the bulk structure. XRD is the main
technique for the identification of bulk crystallographic phases present in the catalyst
shown by the diffraction pattern. In addition, the possible presence of amorphous
phase can also be detected by XRD.
The X-ray diffraction phenomenon occurs when the X-ray photon is
elastically scattered by atoms, which are arranged in a periodic lattice. Figure 4.4
illustrates the diffraction of X-rays by crystal lattices.
51
X-rays
Crystal
lattices
θ
d
Figure 4.4: The diffraction of X-rays by crystal lattices.
By using the Bragg’s law, the lattice spacing that depends on the angular
position, θ, can be derived (Equation 4.2). The corresponding lattice spacing is
characteristic for each particular compound that enables the phase identification of
the compound.
nλ = 2d sin θ
(Equation 4.2)
where the n is reflection order (1, 2, 3…), λ is x-ray wavelength and d is lattice
spacing as well as the θ is the angle between the incoming X-rays and the normal to
the reflecting lattice plane.
The analysis of catalyst materials by XRD was conducted using the
Diffractometer D500 Siemens Crystalloflex with CuKα (λ = 1.54060 Å) as the
radiation source. Scans were performed in a step mode of 0.05°/step and
1 second/step. The data of Al2O3 and AAO supported catalysts were collected over a
2θ range of 20 to 70°. Whereas, 2θ range of 5 to 70° were used for collecting data of
MS supported catalysts.
4.3.1.1 Particle Size Measurement
Phase identification using X-ray diffraction relies mainly on the positions of
the peaks in a diffraction profile and to some extent on the relative intensities of
these peaks. The shapes of the peaks, however, contain additional and often valuable
52
information. The shape, particularly the width, of the peak is a measure of the
amplitude of thermal oscillations of the atoms at their regular lattice sites (Krill and
Birringer, 2005).
The particle size of the supported catalysts generally is too fine to be
measured by electron microscopy. Laser scattering methods are complicated and
time consuming. Measurement of particle size by XRD solves these shortcomings.
Particles are coherent (free of defect) crystal units that diffract in phase. The particle
size is a measurement of adjacent, repeating crystalline units. Particle size is
performed by measuring the broadening of a particular peak in a diffraction pattern
associated with a particular planar reflection from within the crystal unit cell
(Rikagu, 2005). The well known Scherrer Equation (Equation 4.3) explains peak
broadening in terms of incident beam divergence which makes it possible to satisfy
the Bragg condition for non-adjacent diffraction planes. Once instrument effects have
been excluded, the particle size is easily calculated as a function of peak width
(specified as the full width at half maximum peak intensity (FWHM)), peak position
and wavelength.
d=
Kλ
FWHM cos θ
(Equation 4.3)
where d is particle size (Å), K is unit cell geometry dependent constant (0.94), λ
equals to wavelength of x-rays (Å) and FWHM is full width at half maximum peak
intensity (radians) as well as θ representing the Bragg angle (°).
The particle size is inversely related to the FWHM of an individual peak, the
narrower the peak, the larger the particle size. This is due to the periodicity of the
individual particle domains, in phase, reinforcing the diffraction of the X-ray beam,
resulting in a tall narrow peak. If the crystals are defect free and periodically
arranged, the X-ray beam is diffracted to the same angle even through with multiple
layers of the specimen. If the crystals are randomly arranged or have low degrees of
periodicity, the result is a broader peak (Rikagu, 2005).
53
4.3.2
Electron Microscopy
Just as ‘light microscope’ is a generic term covering a range of instruments
for producing magnified images using glass lenses and visible or ultraviolet light, so
the name ‘electron microscope’ does not refer to a specific design of instrument but
to a family of instruments which produce magnified images by the use of
electrostatic or electromagnetic lenses and fast-moving electrons. They all share the
ability to give images of high or very high resolution over a very useful depth of field
(Watt, 1985).
Within the electron microscope family there are two well defined ranges of
microscope, corresponding to transmission and reflection (metallurgical) light
microscopes, which look directly at the internal structure of translucent specimens
and the outside features of bulk material, respectively.
4.3.2.1 Scanning Electron Microscopy (SEM)
In the context of catalyst and CNT characterisation, Scanning Electron
Microscopy (SEM) is the most important instrument for the investigation of topology
and morphology of the specimens. The major components of the SEM are illustrated
in Figure 4.5. These components require careful alignment and proper set up.
A very fine probe of electrons originated from a thermionic electron with
energies up to 30 or 40 keV is focused at the surface of the specimen in the
microscope and scanned across it in a raster or pattern of parallel lines. As the
electron beam impinges on the specimen surface, the emission of secondary electrons
with energies of a few tens of eV and re-emission or reflection of high-energy
backscattered electrons from the primary beam occurs. The intensity of emission of
both secondary and backscattered electrons is very sensitive to the angle at which the
electron beam strikes the surface. The emitted electron current is collected and
amplified; variations in the resulting signal strength as the electron probe is scanned
54
across the specimen are used to vary the brightness of the trace of a cathode ray tube
being scanned in synchronism with the probe. There is thus a direct positional
correspondence between the electron beam scanning across the specimen and the
fluorescent image on the cathode ray tube (Watt, 1985).
Figure 4.5: The components of a Scanning Electron Microscope (SEM) (Reimer,
1984).
In handling a SEM, voltage of the emitted beam and the beam current are
important factors. At higher voltage, more electrons are emitted; however, more
charging is associated with the image. For CNTs, charging is not significant as the
nanotubes and associated deposited graphite layers are conducting. Whereas higher
beam current provide higher energy to the electrons, therefore more surface emitted
electrons resulting better resolution (Ward, 2002).
Special specimen preparation techniques are rarely needed in scanning
electron microscopy. However, charging effects can be avoided, by coating the
specimen with a thin conductive film for example gold (Au) coating (Reimer, 1984).
In this work, a Philips XL 40 SEM was utilized to investigate all of the
supported catalysts and as-grown CNTs with 30.0 keV probe under working distance
around 10 mm.
55
4.3.2.2 Field Emission-Scanning Electron Microscopy (FE-SEM)
The Field Emission-Scanning Electron Microscope (FE-SEM) provides a
view of the surface or topology of the sample. The FE-SEM operates by emitting a
beam of electrons from a tungsten filament and then scanning (rastering) the probe
beam of electrons across the sample surface, ejecting secondary electrons,
backscattered electrons and Auger electrons. The FE-SEM is distinct from a normal
SEM by how the electrons are ejecting from the tungsten filament.
In an FE-SEM, electrons are emitted by applying a very high voltage
(external field), bending the potential barrier that binds the electrons onto the atom.
After a sufficient voltage is applied, the electrons are allowed to escape the atom by
tunnelling into the vacuum (Ward, 2002). The electrons are then accelerated by the
high potential field. This electron beam generation allows for better resolution and
less charging in the FE-SEM since more electrons are emitted from the filament at
lower energies. The sample preparation for FE-SEM is relatively simple. It is
assumed that CNTs are conducting that no Au coating is required to view the CNTs.
The carbon nanotubes produced in this work were examined by a JOEL JSM-6700F
model FE-SEM with magnification of 60000× and 25000×.
4.3.2.3 Transmission Electron Microscopy (TEM)
In a conventional Transmission Electron Microscope (TEM), a thin specimen
is irradiated with an electron beam of uniform current density with energy in the
range of 60 – 150 keV. Electrons are emitted in the electron gun by thermionic
emission from tungsten hairpin cathodes or LaB6 rods or by field emission from
pointed tungsten filaments (Reimer, 1984). The latter are used when high gun
brightness is needed. Unlike SEM, the TEM cannot view surface morphologies, but
sample cross-sections, lattice planes, atomic defects, crystal diffraction, etc. can all
be determined with a TEM. For CNTs, this allows for the determination of the type
of carbon nanotubes (SWNT or MWNT), diameter of the tube, amount of
56
graphitization and amorphous deposits and other structural properties of the tubes
such as helicity (using diffraction), bending and flattening of the tubes (Ward, 2002).
Figure 4.6 shows the schematic ray path for a Transmission Electron Microscope
(TEM).
Figure 4.6: Schematic ray path for a Transmission Electron Microscope (TEM).
The lens aberrations of the objective lens are so great that it is necessary to
work with very small objective apertures to achieve resolution of the order of 0.2 –
0.5 nm. Bright field contrast is produced either by absorption of the electrons
scattered through angles larger than the objective aperture (scattering contrast) or by
interference between the scattered wave and the incident wave at the image point
(phase contrast). The phase of the electron waves behind the specimen is modified by
the wave aberration of the objective lens. This aberration and the energy spread of
the electron gun, which is the order of 1 – 2 eV, limits the contrast transfer of high
special frequencies.
Electrons interact strongly with atoms by elastic and inelastic scattering. The
specimen must therefore be very thin, typically of the order of 5 nm – 5 µm for
57
100 keV electrons, depending on the density and elemental composition of the object
and the resolution desired (Reimer, 1984).
TEM can provide high resolution because elastic scattering is an interaction
process that is highly localized to the region occupied by the screened Coulomb
potential of an atomic nucleus, whereas inelastic scattering is more diffuse; it spreads
out over about a nanometer.
For viewing of the CNTs with a TEM, the substrates are placed into a vile
with ethanol. The vile is then placed into an ultrasonic bath and sonicated for
~10 minutes. This allows for the detachment of the nanotubes from the surface of the
substrates. After sonication, a pipette is used to extract some of the nanotubes in
solution. The solution is then slowly deposited onto a Cu grid with a holey carbon
mask. The ethanol is then allowed to evaporate and the sample can then be placed
into the TEM for viewing. In this work, a JOEL JEM-2010 model TEM was used to
analyze the as-grown CNTs.
4.3.3
Energy Dispersive X-ray Analysis (EDX)
An Energy Dispersive X-ray Analyzer (EDX) is a common accessory which
gives the Scanning Electron Microscope (SEM) a very valuable capability for
elemental analysis. The electron beam in an SEM has energy typically between 5 to
20 keV.
When an incident electron beam hits atoms of the sample, secondary and
backscattered electrons are emitted from the sample surface. However, these are not
the only signals emitted from the sample. When the incident beam bounces through
the sample creating secondary electrons, it leaves thousands of the sample atoms
with holes in the electron shells where the secondary electrons used to be. If these
holes are in inner shells, the atoms are not in a stable state. To stabilize the atoms,
electrons from outer shells will drop into the inner shells, however, because the outer
58
shells are at a higher energy state, to do this the atom must lose some energy. It does
this in the form of X-rays. Figure 4.7 illustrates the interaction of an electron beam
with an atom.
Secondary
electron
emission
Incident beam electron
Outer electron
jumps to inner
shell, releasing
X-ray emission
Figure 4.7: The interaction of an electron beam with an atom.
The detector used in EDX is the Lithium drifted Silicon detector. This
detector must be operated at liquid nitrogen temperatures. When an X-ray strikes the
detector, it will generate a photoelectron within the body of the Si. As this
photoelectron travels through the Si, it generates electron-hole pairs. The electrons
and holes are attracted to opposite ends of the detector with the aid of a strong
electric field. The size of the current pulse thus generated depends on the number of
electron-hole pairs created, which in turn depends on the energy of the incoming Xray. Thus, an X-ray spectrum can be acquired giving information on the elemental
composition of the material under examination.
The X-rays emitted from the sample atoms are characteristic in energy and
wavelength to not only the element of the parent atom, but also which shells lost
electrons and which shells replaced them. For instance, if the innermost shell (the K
shell) electron of an atom is replaced by an L shell electron, a K alpha X-ray is
59
emitted from the sample. If an M shell electron of the atom replaced the K shell
electron, a K beta X-ray is emitted. So, an EDX spectrum of an element would have
a few peaks. Since lower atomic number elements have fewer filled shells, they have
fewer X-ray peaks. While some of the high atomic numbered X-rays can be over 50
keV, a spectral range of 0 – 20 keV can detect all the elements from boron to
uranium.
EDX analysis can also quantify the elements it detects. A quantitative
analysis can be performed by a standardless or standards analysis. A standardless
analysis quantifies the elements by calculating the area under the peak of each
identified element and after taking account for the accelerating voltage of the beam
to produce the spectrum, performs calculations to create sensitivity factors that will
convert the area under the peak into weight or atomic percent. One of the most
popular algorithms is ZAF where Z stands for the atomic number of the element, A
and F are the absorbance and fluorescence values to compensate for the X-ray peak
interaction. From this, the atomic and weight percent are calculated.
Standard quantifications are performed in a similar fashion except the areas
under the elemental peaks are compared to standards files which are spectra of the
elements to be quantified acquired under the exact conditions of the unknown
spectrum. Since it requires these additional spectra, this form of analysis is more time
consuming and generally is less accurate than a standardless quantification unless the
X-ray peaks of the elements to be quantified overlap or the elements are in trace
quantities.
In this work, the EDX analyses of supported catalysts and products of
decomposition of C2H2 were performed using the EDX Microanalysis System
Software with standardless method. The most common X-rays detected in the
samples experimented are the Al Kα, Si Kα, Fe Kα, and Lα, Co Kα, and Lα, O Kα and
C Kα. Each sample was examinated in several points and the average data were
calculated.
CHAPTER 5
PRODUCTION OF CARBON NANOTUBES (CNTs) OVER ALUMINA AND
ANODIC ALUMINIUM OXIDE (AAO) TEMPLATE SUPPORTED
CATALYSTS
5.1
Alumina Supported Catalysts
Aluminas are a family of aluminum oxides (Al2O3) that can take many forms.
As a general term, alumina refers to the lower temperature transition alumina forms.
The most common use for alumina is in the manufacturing of aluminium, a metal
commonly used in beverage cans, automobiles, and many other applications where a
lightweight, durable metal is required. Alumina is also manufactured for use as
adsorbents and catalysts, as well as numerous other applications.
The alumina used in this work is high surface area beads that are activated by
a proprietary process to achieve specific surface chemistry and reactivity. It is
commonly used as adsorbents, desiccants and catalysts. The chemical properties, size
and structure of this alumina is tailored to specific applications.
Group VIIIB elements are known for their catalytic activity, in particular for
their ability to catalyze hydrocarbon decomposition. The metal catalysts selected in
this work are cobalt (Co), ferrum (Fe) and mixture of cobalt and ferrum (Co/Fe). The
acetate salt of metals are utilized as the precursor of the metal oxide in the supported
catalysts. Colomer et al. (1999) results showed that the Co and Fe catalysts are more
active than Ni catalyst and the Co/Fe catalysts gave best activity. The Mo
61
and V are utilised as catalysts in Willems et al. (2000) work, but the Mo and V alone
are totally inactive in the production of CNTs.
The metal catalysts (Co or Fe) loadings of supported catalysts were set in the
range of 2.5 – 3.0 wt %. This is because the particle size of the metal oxide increases
with increasing metal loading and the size of metal cluster also increases (Shajahan
et al., 2004). Larger catalyst clusters have higher tendency to grow multi-walled
carbon nanotubes (MWNTs) with greater diameter which is low in quality. For that
reason, the metal catalysts loading used in this work are low but satisfactory to grow
good quality CNTs.
As described by Perez-Cabero et al. (2004), the impregnation method allows
more homogeneous metallic dispersion on the support. Therefore, we have choosen
the wet impregnation method to incorporate the Co or Fe catalyst onto various
supports.
5.1.1
Physical Appearance of Alumina Supported Catalysts
The alumina supported catalysts obtained after drying at 90 °C remained in
beads form and not hygroscopic. The alumina support is originally white in colour.
After associated with either Co or Fe or Co/Fe metal catalyst, the support exhibited
different colours. The Al–Co catalysts showed purple pinkish colour, while the Al-Fe
catalysts showed brownish orange colour. The colour of the Al-Co/Fe catalysts is
light orange.
In all cases, the catalysts containing 3.0 wt.% metal showed more intense
colour compared to the catalysts with 2.5 wt.% metal. However, the colour of the
catalysts changed after calcination. After calcination at 450 °C, the Al-Co catalysts
turn to dirty green, whereas the Al-Fe catalysts turned into dark orange and Al-Co/Fe
catalysts turned into olive green. This indicates that the incorporation of the metal
catalysts onto the support gave different physical appearance to various catalysts.
62
5.1.2
X-ray Diffraction (XRD) Analysis
The alumina support and series of alumina supported catalysts were analyzed
using XRD mainly for the determination of active phases that would contribute to a
good catalytic activity in the synthesis of CNTs. Apart from that, the effect of varied
preparation parameters such as types of metal catalysts, amount of metal catalyst
loading and effect of calcination could be observed too. Thus, a contributing factor
that resulted in the variation of catalysts activities could be reasoned out. The
structural and phase of the catalysts were identified upon comparing the 2θ and d
values of the studied catalysts with that of reference materials from the powder
diffraction file (PDF 2001).
5.1.2.1 XRD Analysis of Alumina Supported Cobalt Catalysts
X-ray diffractogram patterns of dried Al2O3 support, Al–Co(2.5),
Al–Co(2.5)Cal, Al–Co(3.0) and Al–Co(3.0)Cal catalysts were shown in Figure 5.1.
The data were tabulated in Table 5.1. The Al2O3 support was observed as Al2O3
tetragonal and found to be amorphous. The main peaks for Al2O3 tetragonal phase
was observed at 2θ values of 67.382, 66.973 and 32.822° or d values of 1.3886,
1.3961 and 2.7264 Å (PDF 2001 d values (Å): 1.3996, 1.3959 and 2.7280).
In the Al-Co(2.5) catalyst, the face-centred cubic Co3O4 were detected at 2θ
values of 36.793, 31.325 and 65.358° with d values of 2.4407, 2.8532 and 1.4266 Å
(PDF 2001 d values (Å): 2.4370, 2.8580 and 1.4290). The Al2O3 support remained in
the same phase, but the structure has been suppressed as shown in Figure 5.1. This
probably due to the presence of Co3O4 face-centred cubic phase (Tee et al., 2004a).
In the case of higher loading of Co (3.0 wt.%), the intensity of Co3O4 face-centred
cubic peak at 2θ = 36.7° increases. The 2θ and d values of the Al-Co(2.5) and AlCo(2.5)Cal catalysts are shown in Table 5.1.
63
Al-Co(3.0)Cal
Intensity (%)
Al-Co(3.0)
Al-Co(2.5)Cal
Al-Co(2.5)
Al2O3
20
30
40
50
60
70
2θ (°)
Figure 5.1: XRD patterns of Al2O3 support, Al–Co(2.5), Al–Co(2.5)Cal, Al–Co(3.0)
and Al–Co(3.0)Cal catalysts. : Al2O3 tetragonal and
: Co3O4 face-centred cubic.
64
Table 5.1: Phases, peak position (2θ) and d values (Å) in the X-ray diffractogram
patterns of the Al2O3 support, Al-Co(2.5) and Al-Co(2.5)Cal catalysts.
Samples
Phase Structure
2θ (°)
67.382
66.973
32.822
67.402
Al2O3 tetragonal
67.044
32.756
Al–Co(2.5)
36.793
Co3O4 face31.325
centred cubic
65.358
67.198
Al2O3 tetragonal
66.942
32.807
Al–Co(2.5)Cal
36.742
Co3O4 face31.325
centred cubic
64.409
* dobs : d spacing values obtained from XRD analyses.
dref : d spacing values obtained from the reference.
Al2O3
Al2O3 tetragonal
dobs* (Å)
1.3886
1.3961
2.7264
1.3882
1.3948
2.7318
2.4408
2.8532
1.4266
1.3920
1.3967
2.7276
2.4441
2.8532
1.4256
dref * (Å)
(PDF 2001)
1.3996
1.3959
2.7280
1.3996
1.3959
2.7280
2.4370
2.8580
1.4290
1.3996
1.3959
2.7280
2.4370
2.8580
1.4290
In spite of calcination on both of the Co catalysts at 450 °C, there was no
phase or structural transformation observed. The Al2O3 tetragonal and Co3O4 facecentred cubic phases were preserved (Figure 5.1). However, the intensity of Co3O4
face-centred cubic phase had increased, which suggested that the degree of
crystallinity of both Al-Co(2.5) and Al-Co(3.0) catalysts had increased with
calcination. Comparing the diffractograms of the catalysts to that of the support,
there are no difference in the peak positions, except for very low intensity peaks and
peak broadening.
In the diffractograms of Al2O3 supported Co catalysts, only the broad peaks
of metallic Co appeared. In fact, Shajahan et al. (2004) had proven that the formation
of an ideal Co3O4-Al2O3 phase in the catalysts is responsible for the XRD reflections
at the identical positions of the support Al2O3. Further, Shajahan et al. (2004) also
explained that the Co crystallite was small in size and dispersed well over the
65
support, so in the XRD patterns, the catalysts showed broad peaks, their intensities
decreased and concurrently no clear peak for the presence of Co.
In order to determine the active catalyst phases more accurately, an Al2O3
supported catalyst of 40 wt.% Co loading were prepared for the purpose of XRD
analyses. The high loading of Co catalyst (40.0 wt.%) is expected to show high
intensity peaks in the XRD patterns. The Al-Co(40.0) catalyst were calcined at two
temperatures, i.e. 450 °C for 17 hours and 750 °C for 5 hours. The reason to carry
out the calcination at 750 °C is to verify the active catalyst phase appeared during the
CNTs synthesis at 700 °C. The XRD patterns of Al-Co(40.0)Cal were shown in
Figure 5.2 and the data were tabulated in Table 5.2. In the Al-Co(40.0) catalyst
calcined at 450 °C / 5 h, the face-centred cubic Co3O4 were detected at 2θ values of
36.850, 31.258 and 65.219° with d values of 2.4371, 2.8591 and 1.4293 Å (PDF
2001 d values (Å): 2.4370, 2.8580 and 1.4290). The Al2O3 support cannot be
identified in the diffractograms, it is due to the high loading of Co catalyst (40 wt%)
was over saturated the Al2O3 support.
From the XRD pattern of Al-Co(40.0)Cal catalyst calcined at 750 °C for 17
hours, there was no phase transformation occurred in the XRD pattern (Figure 5.2).
The main peaks of Co3O4 face-centred cubic phase appeared at 2θ values of 36.860,
31.258 and 65.213° with d values of 2.4365, 2.8592 and 1.4295 Å, as shown in Table
5.2. Therefore, through the XRD analyses of Al-Co(40.0)Cal catalyst, we proved that
the Co catalyst existed as Co3O4 face-centred cubic.
66
Intensity (%)
Calcination: 750 °C / 5 h
Calcination: 450 °C / 17 h
20
30
40
50
60
70
2θ (°)
Figure 5.2: XRD patterns of Al-Co(40.0)Cal catalyst calcined at 450 °C / 17 h and
750 °C / 17 h. : Co3O4 face-centred cubic.
Table 5.2: Phases, peak position (2θ) and d values (Å) in the X-ray diffractogram
patterns of Al-Co(40.0)Cal catalyst calcined at 450 °C / 17 h and 750 °C / 17 h.
Sample
Calcination
Phase
Structure
2θ (°)
36.850
31.258
450 °C / 17 h
65.219
Co(40.0)
36.860
Co3O4 face31.258
750 °C / 5 h
centred cubic
65.213
* dobs : d spacing values obtained from XRD analyses.
dref : d spacing values obtained from the reference.
Co3O4 facecentred cubic
dobs* (Å)
2.4371
2.8591
1.4293
2.4365
2.8592
1.4295
dref * (Å)
(PDF
2001)
2.4370
2.8580
1.4290
2.4370
2.8580
1.4290
67
5.1.2.2 XRD Analysis of Alumina Supported Ferrum Catalysts
The XRD analyses of various Al-Fe catalysts as well as the Al2O3 support are
shown in Figure 5.3. The data are tabulated in Table 5.3. From the XRD pattern, the
Al-Fe catalysts were generally found to be in amorphous state with the presence of
Fe2O3 rhombohedral phase. The amorphous XRD patterns of Al-Fe catalysts were in
agreement with Kukovecz et al. (2000) work. The high intensity peak of Al-Fe(2.5)
at 33.273° or d value of 2.7010 Å (PDF 2001 d value: 2.6900 Å) was suggestively
assigned to Fe2O3 rhombohedral phase. The main peak of Fe2O3 rhombohedral phase
existed at 33.273° overlapped with the main peak of Al2O3 tetragonal at 32.858°, as a
result a broad peak appear around 33°. The Al2O3 support was observed to remain in
the same phase after the incorporation of Fe metal, but the structure of the Al2O3
tetragonal phase was found to be suppressed showed by the presence of lower
intensity peaks.
The calcination treatment has resulted in the increase of the peak intensities
and the Fe2O3 rhombohedral phase in the Al-Fe(2.5)Cal catalyst became more
dominant. The peaks at 2θ values of 33.263, 35.731 and 54.239° with d values of
2.7022, 2.5240 and 1.6879 Å (PDF 2001 d values (Å): 2.6900, 2.5100 and 1.6900)
are assigned to Fe2O3 rhombohedral phase. The increment of Fe loading to 3.0 wt.%
gave sharper and higher intensity of the Fe2O3 rhombohedral at 2θ values of 33.268,
35.750 and 54.231° and d values of 2.7067, 2.5013 and 1.6900 Å. From the
diffractogram of Al-Fe(3.0)Cal catalyst (Figure 5.3), it is observed that the peaks of
Fe2O3 rhombohedral became more dominant with higher peak intensities.
This observation was found by Valiente et al. (2000) and Takenaka et al.
(2004) in their previous work. As reported by Valiente et al. (2000), the catalyst
prepared by Fe aqueous solution showed peaks of Fe2O3 phase in the XRD
diffractograms. Takenaka et al. (2004) found that Fe catalysts supported on Al2O3
prepared by impregnation were also exist as Fe2O3 crystallite, the calcination in air at
500 °C did not change the Fe2O3 structure.
68
Al-Fe(3.0)Cal
Intensity (%)
Al-Fe(3.0)
Al-Fe(2.5)Cal
Al-Fe(2.5)
Al2O3
20
30
40
50
60
70
2θ (°)
Figure 5.3: X-ray diffractogram patterns of various Al-Fe catalysts together with
Al2O3 support. : Al2O3 tetragonal and S: Fe2O3 rhombohedral.
69
Table 5.3: Phases, peak position (2θ) and d values (Å) in the X-ray diffractogram
patterns of the Al2O3 support, Al-Fe(2.5) and Al-Fe(2.5)Cal catalysts.
Samples
Phase Structure
2θ (°)
67.382
66.973
32.822
67.198
Al2O3 tetragonal
66.636
32.858
Al–Fe(2.5)
33.273
Fe2O3
35.738
rhombohedral
54.224
67.351
Al2O3 tetragonal
66.789
32.705
Al–Fe(2.5)Cal
33.263
Fe2O3
35.731
rhombohedral
54.239
* dobs : d spacing values obtained from XRD analyses.
dref : d spacing values obtained from the reference.
Al2O3
Al2O3 tetragonal
dobs* (Å)
1.3886
1.3961
2.7264
1.3920
1.4023
2.7235
2.7010
2.5231
1.7002
1.3892
1.3995
2.7359
2.7022
2.5240
1.6879
dref * (Å)
(PDF 2001)
1.3996
1.3959
2.7280
1.3996
1.3959
2.7280
2.6900
2.5100
1.6900
1.3996
1.3959
2.7280
2.6900
2.5100
1.6900
The XRD patterns of Al-Fe catalysts showed low intensity peaks of Fe2O3
rhombohedral, it is due to the low loading of the Fe catalyst. To overcome this, an
Al2O3 supported Fe catalyst with 30 wt.% loading was prepared, the catalyst was
undergone calcinations at 450 °C for 17 hours and 750 °C for 5 hours. The
diffractograms of the Al-Fe(30.0)Cal catalyst were demonstrated in Figure 5.4,
together with the 2θ and d values in Table 5.4. For the Al-Fe(30.0)Cal catalyst
calcined at 450 °C, the XRD pattern clearly depicted that the Fe catalyst appeared as
Fe2O3 rhombohedral phase. The peaks were assigned at 2θ of 33.191, 35.730 and
54.271° with d values of 2.6970, 2.5109 and 1.6889 Å (PDF 2001 d values (Å):
2.6900, 2.5100 and 1.6900) as shown in Table 5.3. The calcination at 750 °C did not
resulted any changes to the diffractogram pattern of the Al-Fe(30.0)Cal catalyst. The
catalyst was observed to remain as the Fe2O3 rhombohedral phase. Thus through the
XRD analyses of Al-Fe(30.0)Cal catalyst, it is verified that the Fe catalyst existed on
the Al2O3 support as Fe2O3 rhombohedral phase.
70
Intensity (%)
Calcination: 750 °C / 5 h
Calcination: 450 °C / 17 h
20
30
40
50
60
70
2θ (°)
Figure 5.4: XRD patterns of Al-Fe(30.0)Cal catalyst calcined at 450 °C / 17 h and
750 °C / 17 h. S: Fe2O3 rhombohedral.
Table 5.4: Phases, peak position (2θ) and d values (Å) in the X-ray diffractogram
patterns of Al-Fe(30.0)Cal catalyst calcined at 450 °C / 17 h and 750 °C / 17 h.
Sample
Calcination
Phase
Structure
2θ (°)
33.191
35.730
450 °C / 17 h
54.271
Fe(30.0)
33.185
Fe2O3
35.791
750 °C / 5 h
rhombohedral
54.171
* dobs : d spacing values obtained from XRD analyses.
dref : d spacing values obtained from the reference.
Fe2O3
rhombohedral
dobs* (Å)
2.6970
2.5109
1.6889
2.6974
2.5068
1.6917
dref * (Å)
(PDF
2001)
2.6900
2.5100
1.6900
2.6900
2.5100
1.6900
71
5.1.2.3 XRD Analysis of Alumina Supported Cobalt/Ferrum Catalysts
X-ray diffractogram patterns of a series of Al-Co/Fe catalysts together with
the Al2O3 support are shown in Figure 5.5. The 2θ and d values are tabulated in
Table 5.5. From the diffractograms, the catalysts are observed to be amorphous with
broad peaks as described by Kukovecz et al. (2000). The Al2O3 support structure did
not collapse after the impregnation of Co and Fe.
Interestingly, the XRD pattern of Al-Co/Fe catalysts showed the existence of
both Co and Fe in their individual phase, i.e. Co3O4 face-centred cubic and Fe2O3
rhombohedral phase. For Al-Co/Fe(2.5) catalyst, the 2θ values at 36.974, 31.297 and
65.342° and d values of 2.4293, 2.8557 and 1.4269 Å (PDF 2001 d values (Å):
2.4370, 2.8580 and 1.4290) corresponding to the existence of Co3O4 face-centred
cubic phase. The peaks at 2θ values of approximately 32.729, 64.548 and 29.968°
and d values of 2.7340, 1.3856 and 2.9793 Å (PDF 2001 d values (Å): 2.7280,
1.3830 and 2.9810) are due to the existence of Fe2O3 rhombohedral phase. This
observation indicates that there is no interaction between the Co3O4 face-centred
cubic phase and Fe2O3 rhombohedral phase. In the Al-Co/Fe calcined catalysts, the
Co3O4 face-centred cubic and Fe2O3 rhombohedral phase remained unchanged.
However, Fe2O3 rhombohedral phase is more dominant in the form of higher
intensity peaks (Tee et al., 2004b).
On the other hand, the structure of the Al-Co/Fe(3.0) and Al-Co/Fe(3.0)Cal
catalysts demonstrate better crystallinity in the diffractograms, as both of the
catalysts illustrated higher peak intensities (Figure 5.5). This may be due to the
higher concentration of the Co and Fe catalyst. However, the calcination did not
change the XRD patterns of the Al-Co/Fe(3.0) and Al-Co/Fe(3.0)Cal catalyst.
72
Al-Co/Fe(3.0)Cal
Intensity (%)
Al-Co/Fe(3.0)
Al-Co/Fe(2.5)Cal
Al-Co/Fe(2.5)
Al2O3
20
30
40
50
60
70
2θ (°)
Figure 5.5: XRD patterns of Al-Co/Fe catalysts and Al2O3 support. : Al2O3
tetragonal,
: Co3O4 face-centred cubic and S: Fe2O3 rhombohedral.
73
Table 5.5: Phases, peak position (2θ) and d values (Å) in the X-ray diffractogram
patterns of the Al2O3 support, Al-Co/F(2.5) and Al-Co/Fe(2.5)Cal catalysts.
Samples
Phase Structure
2θ (°)
67.382
Al2O3
Al2O3 tetragonal
66.973
32.822
67.248
Al2O3 tetragonal
66.941
32.729
36.974
Al–Co/Fe(2.5)
Co3O4 face31.297
centred cubic
65.342
33.278
Fe2O3
35.740
rhombohedral
54.218
67.291
Al2O3 tetragonal
66.729
32.776
36.918
Al–Co/Fe(2.5)Cal
Co3O4 face31.294
centred cubic
65.144
33.269
Fe2O3
35.738
rhombohedral
54.229
* dobs : d spacing values obtained from XRD analyses.
dref : d spacing values obtained from the reference.
dobs* (Å)
1.3886
1.3961
2.7264
1.3911
1.3967
2.7339
2.4293
2.8557
1.4269
2.6910
2.5132
1.6988
1.3903
1.4006
2.7301
2.4328
2.8560
1.4308
2.7011
2.5136
1.6915
dref * (Å)
(PDF 2001)
1.3996
1.3959
2.7280
1.3996
1.3959
2.7280
2.4370
2.8580
1.4290
2.6900
2.5100
1.6900
1.3996
1.3959
2.7280
2.4370
2.8580
1.4290
2.6900
2.5100
1.6900
5.1.2.4 Particle Size of Alumina Supported Catalysts
It is known that the particle size of the catalyst in the supported catalysts
system is important in determining the diameter of the as-grown CNTs. The particle
size of the Al2O3 supported catalysts were measured by Scherrer Equation (Equation
4.3) using data from the XRD analyses. Only the particle size of the Al-Co/Fe
catalysts are discussed here as these catalysts showed the best XRD patterns and
expected to give higher activity towards the synthesis of CNTs. The particle size of
the Co3O4 face-centred cubic and Fe2O3 rhombohedral particles were tabulated in
Table 5.6, while the example of calculation was shown in Appendix C.
74
It is observed that the particle sizes for the Co3O4 face-centred cubic particle
are in the range of 28 – 39 nm. Meanwhile, the Fe2O3 rhombohedral phase exhibited
smaller particle sizes of 22 – 30 nm. There is no exact trend of the particle size of the
Co3O4 face-centred cubic phase. This is because the particle size is not affected by
either the increment of Co loading or the calcination. On the contrary, the particle
size of Fe2O3 rhombohedral phase can be controlled through Fe catalyst loading and
calcination. The particle size increases when the Fe catalyst loading increases. Also,
the calcined catalysts tend to have larger Fe2O3 rhombohedral particles.
Table 5.6: The particle size (d) of Co3O4 face-centred cubic and Fe2O3 rhombohedral
particles in Al2O3 supported Co/Fe catalysts.
Catalysts
Particle size, d (nm)
Co3O4 face-centred cubic
Fe2O3 rhombohedral
Al-Co/Fe(2.5)
33.73
23.54
Al-Co/Fe(2.5)Cal
28.26
29.54
Al-Co/Fe(3.0)
31.08
22.35
Al-Co/Fe(3.0)Cal
38.64
29.89
The particle sizes in the Al-Co/Fe catalysts are in agreement with the
previous work. The Co and Fe catalyst particles form on the Al2O3 support are less
than 40 nm, which is concurring to the average nanometer scale required for
producing CNTs. Nagaraju et al. (2002) emphasized that nanometer scale clusters are
required for producing CNTs. On the other hand, Takenaka et al. (2004) disclosed
that an average crystallite size of Fe2O3 in fresh Fe2O3/Al2O3 catalysts are almost
constant (ca. 25 nm) irrespective of Fe2O3 loadings. The results showed that the size
of Fe2O3 catalyst particles obtained in this work is comparable to 25 nm.
75
5.1.2.5 Conclusion of XRD Analysis
The XRD deduction of the Al2O3 supported catalysts showed that the
incorporation of the Co, Fe or mixture of Co/Fe onto the Al2O3 support was
accomplished. All of the metal catalyst phases were identified through the XRD
analyses and the XRD results were verified through from the analysis of high loading
catalysts. Basically, the Al2O3 support was discovered as Al2O3 tetragonal, while the
Co catalysts were identified as Co3O4 face-centred cubic phase and the Fe catalysts
were assigned as the Fe2O3 rhombohedral phase. In the Co/Fe catalysts, the metal
catalyst showed the same phases as if they are isolated in the system.
Generally, the structure of Al2O3 support was suppressed by the metal
catalyst after the impregnation. The increase in metal loading of the catalysts resulted
in higher degree of crystallinity in the XRD diffractograms. The calcination at
450 °C had improved the structure of the catalysts but did not change the phase of
the catalysts, as shown by the sharp and highly intense peaks in XRD diffractograms.
According to Kukovecz et al. (2000), all the alumina supported Co and Fe samples
were amorphous and even after calcination no new phase was detected in the
catalysts. This explained the XRD pattern of the Al2O3 support and the Co or Fe
catalysts prepared in this work. The bimetallic catalyst of Al2O3 showed better
characteristic towards the synthesis of CNTs, compared to the monometallic catalyst
system.
5.1.3
The As-Grown Carbon Deposition Yield and Physical Appearance
It is known that C2H2 is more reactive than CH4 at the same reaction
temperature, resulting in high conversion into carbon product. Therefore, C2H2 is
more beneficial carbon source to obtain high-yeild CNTs compared with CH4. In this
work, C2H2 gas was used as the carbon precursor with flow rate of 15 sccm.
According to Shajahan et al. (2004), excessive carbon deactivates catalyst and
deposits high amorphous carbon. Temperature higher than 700 °C can induce the
76
formation of amorphous and unwanted carbonaceous product as reported by
Kukovecz et al. (2000), therefore a moderate reaction temperature of 700 °C was
applied. The use of N2 gas as the carrier gas can improve the mobility of the C2H2
gas and achieve optimum residence time of acetylene on the catalyst active sites
(Piedigrosso et al., 2000). A tarry viscous liquid was deposited at exist end of the
reactor when C2H2 gas was used.
The as-synthesized carbon deposits were black in colour in all cases but the
texture was distinctly different. From the difference in texture of the carbon deposit
obtained from the syntheses, at the first instance one can get an idea if a particular
catalyst was active in producing CNTs. In general, when the carbon deposit appeared
spongy with deposition of black fine powder, the catalyst was good for CNT
production. On the other hand, when the deposit remained in beads form but only
turned black without powder deposition, the activity was low. The percent of carbon
deposition yield of the reaction by Al2O3 supported catalysts and the physical
appearance as well as the density of the as-synthesized carbon deposit were tabulated
in Table 5.5.
From the observation on the physical appearance of the carbon deposit
obtained from different catalysts (Table 5.5), several inferences can be drawn. A
mixture of Co and Fe was more active than the individual metals, which produced
carbon deposit with spongy texture and powder deposition. The catalysts with higher
loading of metal (3.0 wt.%) exhibited higher activity compared to the 2.5 wt.%
loading of metal catalysts. Al-Co/Fe(3.0)Cal was found to be the best catalyst for the
production of high quantity of carbon deposit.
The carbon deposition yield is defined as the percentage of carbonaceous
products deposited on the supported catalyst due to the catalytic decomposition of
C2H2, which was obtained from Equation 3.1. The weight percentages of carbon
deposit obtained from various Al2O3 supported catalysts are given in Table 5.5. At
first it is observed that all of the Al2O3 supported catalysts were active in
transforming the C2H2 precursor into carbonaceous product, the carbon deposit yield
77
were varied from 21.7 % to 53.8 %. This range of carbon deposition yield is found to
be competitive to Nagaraju et al. (2002) work.
Table 5.7: The Al2O3 supported catalysts and the yield (%), the physical appearance
and density of the as-synthesized carbon deposit in CCVD of C2H2.
Carbon
deposition
Physical appearance
Density*
yield (%)
Al-Co(2.5)
23.66
Not spongy
+
Al-Co(2.5)Cal
28.49
Not spongy
+
Al-Co(3.0)
41.10
Slightly spongy, fine powder
++
Al-Co(3.0)Cal
47.65
Spongy, lots of fine powder
+++
Al-Fe(2.5)
21.73
Not spongy
+
Al-Fe(2.5)Cal
21.83
Not spongy
+
Al-Fe(3.0)
24.56
Slightly spongy
++
Al-Fe(3.0)Cal
26.55
Spongy, lots of fine powder
+++
Al-Co/Fe(2.5)
23.58
Spongy, fine powder
++
Al-Co/Fe(2.5)Cal
26.40
Spongy, fine powder
++
Al-Co/Fe(3.0)
40.49
Very spongy, plenty of fine powder
+++
Al-Co/Fe(3.0)Cal
53.82
Very spongy, plenty of fine powder
+++
* Density of carbon deposition yield from observations; +: low density, ++: moderate
density, +++: high density.
Catalysts
The bimetallic catalysts produced greater quantity of carbon deposition
compared to Co and Fe individually. Interestingly, in the case of Al-Co and Al-Co/Fe
catalysts, the carbon deposition yield increased dramatically when the metal catalyst
loading was increased to 3.0 wt.%, this might due to the increase of active catalyst
sites and consequently increase the rate of acetylene decomposition to produce
higher amount of yield. However, this observation was not obvious in the Al-Fe
catalysts. The Al-Fe catalysts showed relatively lower deposition yield. This
observation was similar to Kathyayini et al. (2004) research, where the carbon
deposit were vary in the order of Co/Fe > Co > Fe catalysts. The carbon deposition
yields together with the physical appearance of the carbon deposit contribute a very
useful indication of the activity of the catalysts.
78
5.1.4
Scanning Electron Microscopy (SEM) Analysis
5.1.4.1 SEM Analysis of Alumina Supported Cobalt Catalysts and the CNTs
Yield
Scanning Electron Microscopy (SEM) analysis of dried Al2O3 support
(Figure 5.6) shows that the support surface was covered with particles of average
size of 6.4 µm. The example of particle size calculation is shown in Appendix D.1.
Figure 5.6: SEM micrograph of dried Al2O3 support (1000×).
After the impregnation of Co catalyst, the Al2O3 particles were covered by
the Co catalyst, resulting in Co3O4-Al2O3 composite material. As depicted in Figure
5.7 (a), the Co3O4-Al2O3 composite was observed as irregular flat shape particles on
the surface of Al-Co(3.0) catalyst. The Al-Co(2.5) catalyst has similar morphology.
Through this observation, the suppression of Al2O3 tetragonal phase in the
diffractograms of Al-Co catalysts can be explained. The Al2O3 structure showed
lower peak intensity in the XRD diffractograms of catalysts because it was
overwhelmed by the metal catalyst particles.
It is noticeable that the calcination treatment had produced well dispersed
Co3O4 crystals with sharp edges, as seen in micrograph of Al-Co(3.0)Cal catalyst
(Figure 5.7 (b)). This indicates that the calcined Al-Co catalysts have higher
crystallinity, which is in agreement with XRD analyses in the previous section.
79
(a)
(b)
Figure 5.7: SEM micrographs of Al2O3 supported Co catalysts: (a) Al-Co(3.0)
(1000×) and (b) Al-Co(3.0)Cal (1000×).
It should be noted from the SEM images, the carbon nanotubes can be
observed as are highly graphitized tubules but the carbon nanofibers is amorphous
fibrous layers (Buang et al., 2005). Therefore, through the SEM analysis, we can
differentiate the growth of CNTs from the growth of CNFs on the surface of
supported catalysts. The SEM micrographs demonstrate images of the CNT growth
density and the nanotubes texture.
SEM micrographs of as-synthesized carbon deposits over different Al-Co
catalysts at 700 °C under the flow of C2H2/N2 mixture for 30 minutes are shown in
Figure 5.8 (a) – (e). SEM analyses revealed that the density and the nature of CNTs
produced were unique in each case. In Figure 5.8 (a), it can be seen that there is no
tube-like structure appeared on the surface of the Al-Co(2.5) catalyst, the catalyst
particles were covered by some carbonaceous materials. Hence, the Al-Co(2.5)
catalyst is only active to decompose C2H2 into carbonaceous products but inactive in
growing the CNTs.
In Al-Co(2.5)Cal catalyst (Figure 5.8 (b)), the growth of CNTs can be clearly
observed. The entangled and sinuous nanotubes are grown from the catalyst particles.
The nanotubes have narrow diameter and length up to a few micrometers. The
average diameter of the as-grown CNTs on the catalyst is 89.3 nm. The calculation
of length and diameter of the CNTs are shown in Appendix D.2.
80
(a)
(b)
(c)
(d)
(e)
Figure 5.8: SEM micrographs of as-grown CNTs over various Al2O3 supported Co
catalysts, (a) Al-Co(2.5) (10000×), (b) Al-Co(2.5)Cal (10000×), (c) and (d)
Al-Co(3.0) (5000× and 10000×) and (e) Al-Co(3.0)Cal (10000×).
Interestingly when the Al-Co(3.0) catalyst was used, the CNTs growth was
enhanced significantly. The density and quality of the CNTs grown over these
catalysts are greatly improved (Ismail et al., 2004a). From the image of the as-grown
CNTs over Al-Co(3.0) catalyst in Figure 5.8 (c), the Co3O4 particles were not visible
as being covered up by the highly dense CNTs. At high magnification (Figure 5.8
(d)), the CNTs can be seen as bundles of tubes with fluffy and spongy texture. The
81
CNTs possess high quality and density with diameter of approximately 53 nm and
length less than 10 microns. Obviously, the as-grown nanotube bundles are pure
without purification, as no amorphous carbon and carbonaceous particles deposited
on the catalyst. Here, the amorphous carbon was defined as carbonaceous deposit
which is not in the form of ordered nanotubes and graphite.
Further, the Al-Co(3.0)Cal catalyst also showed capability in producing
CNTs bundles, as depicted in Figure 5.8 (e). However, the as-grown CNTs over AlCo(3.0)Cal catalyst have lower density, where the nanotubes are shorter and less
fluffy. For both of 3.0 wt. % Co catalysts, the yield and the physical appearance of
the carbon deposit indicated higher quantity of the yield and better activity of these
catalysts.
Fonseca et al. (1998) disclosed that the calcination process at 450 °C
thermally decomposes Co acetate to Co oxide, resulting well-dispersed Co oxide
particles in the inner pores of the silica support. The hydrogenation steps of the
catalyst preparation are not necessary and even have unfavourable effect on the
catalyst performance. Since acetylene is able to reduce Co oxide under reaction
conditions to generate the active sites in situ during the induction period of the
reaction, it can be concluded that calcination at 450 °C is sufficient in forming good
quality CNTs. In conclusion, all of the Co catalysts are active in the formation of
CNTs with different efficiency, except for the Al-Co(2.5) catalyst.
5.1.4.2 SEM Analysis of Alumina Supported Ferrum Catalysts and the
CNTs Yield
From the SEM micrograph of Al-Fe(2.5)Cal (Figure 5.9 (a)), we can see that
the surface morphology of the catalyst is rough and the Fe2O3 crystal particles are
tightly bonded to the Al2O3 support. As a result, the suppression of Al2O3 tetragonal
phase was observed in the XRD diffractograms of various Al-Fe catalysts, this is
caused by the wrapping of support structure by the Fe2O3 catalyst particles. The
82
Al-Fe(2.5) catalyst has similar morphology as the calcined catalyst, so the
micrograph is not discussed here.
Refering to the micrograph of Al-Fe(3.0) catalyst (Figure 5.9 (b)), the Fe2O3
crystal particles became more dense and fully covered the Al2O3 surface. But,
calcination at 450 °C did not change the morphology of the Al-Fe(3.0)Cal catalyst.
This observation is in concurrence with the XRD data, where the catalysts contain
3.0 wt.% Fe showed higher crystallinity and more ordered structure.
(a)
(b)
Figure 5.9: SEM micrographs of Al2O3 supported Fe catalysts: (a)Al-Fe(2.5)Cal
catalyst (1000×) and (b) Al-Fe(3.0) catalyst (1000×).
The SEM micrographs of as-grown CNTs over Al-Fe catalysts have clearly
attested the ability of these catalysts to grow CNTs. Figure 5.10 (a) shows the high
magnification image of as-grown CNTs over Al-Fe(2.5) catalyst. The bundles of
CNTs shows uniform diameter (ca. 71 nm) with length of a few microns. On the
other hand, the Al-Fe(2.5)Cal catalyst seems to have comparable activity with the
Al-Fe(2.5) catalyst. As seen on Figure 5.10 (b), the as-grown CNTs do not show
much difference with the previous case.
83
(a)
(b)
(c)
(d)
(e)
Figure 5.10: SEM micrographs of as-grown CNTs over different Al2O3-Fe catalyst,
(a) Al-Fe(2.5) (10000×), (b) Al-Fe(2.5)Cal (10000×), (c) Al-Fe(3.0) (10000×), (d)
and (e) Al-Fe(3.0)Cal (5000× and 10000×).
However, the Al-Fe(3.0) catalyst also showing lower activity in growing
CNTs. As illustrated in Figure 5.10 (c), the surface of the catalyst was not completely
covered by CNTs. Although the CNTs grown in greater length, but they are less
dense and possess diameter range of 89 nm – 100 nm. In contrast, the Al-Fe(3.0)Cal
catalyst shows very encouraging CNTs growth. The lower magnification image
84
(Figure 5.10 (d)) is used to convey the contiguous mass of nanotubes with high
uniformity, whereas the higher magnification micrograph (Figure 5.10 (e)) conveys
the curved nanotubes morphology and smooth surface. The nanotubes grow
extremely long with low diameter distribution, the estimated diameter is around
70 nm. The texture of the nanotubes can be seen as fluffy and in bundle form.
At this stage, it was identified that the Al-Fe(3.0)Cal catalyst had shown the
best catalytic activity amongst the Al2O3 supported Fe catalysts. Nevertheless, all of
the Fe catalysts are active towards CNTs growth with different effectiveness. In fact,
the Fe catalysts are performing better than the Co catalysts.
5.1.4.3 SEM Analysis of Alumina Supported Cobalt/Ferrum Catalysts and the
CNTs Yield
A representative SEM image of Al-Co/Fe(2.5) catalyst is shown in
Figure 5.11 (a). This catalyst surface has rough texture and full of flaky particles
with some undefined shape particles on top of it. The morphology gives us an idea
that the Fe2O3 rhombohedral phase is more dominant than the Co3O4 face-centred
cubic phase, since the morphology of this catalyst is more analogous to the Fe
catalyst. Albeit in the XRD analysis, the Fe2O3 rhombohedral also showing higher
intensity peak. However, subsequent to the calcination, the Al-Co/Fe(2.5)Cal catalyst
surface is covered by random size particles and the flaky structure disappear, as
depicted in Figure 5.11 (b). This indicates that the Co3O4 phase is started to become
visible on the catalyst.
For the Al-Co/Fe(3.0) catalyst (Figure 5.11 (c)), the particles appear on the
catalyst showed wide diameter range. These particles showed morphology similar to
the Co3O4 phase. As described in previous section, the XRD diffractogram pattern of
this catalyst gave higher intensity peaks. The increased in the particle size is due to
the increase of metal catalysts concentration. Thus, the XRD analysis of the Al-
85
Co/Fe(3.0) catalyst also show higher crystallinity. The calcination did not resulted in
significant physical changes to the Al-Co/Fe(3.0)Cal catalyst.
(a)
(b)
(c)
Figure 5.11: SEM micrographs of various Al2O3-Co/Fe catalysts: (a) Al-Co/Fe(2.5)
(1000×), Al-Co/Fe(2.5)Cal (1000×) and (c) Al-Co/Fe(3.0) (1000×).
With respect to the good XRD patterns and surface morphology of various
Al2O3 supported Co/Fe catalysts, these catalysts are expected to show excellent
activity towards the growth of CNTs. The presence of CNTs on Al-Co/Fe(2.5)
catalyst are observed in Figure 5.12 (a). The nanotubes are observed to have grown
from the catalyst particles. The nanotubes exhibit length of less than 10 microns, but
some of the nanotubes can even grow longer than 10 µm. The textures of the tubes
are smooth without any particles adhere on it along with the average diameter of
about 71 nm. But, the density of the CNTs is low, the catalyst particles are not
completely covered by the CNTs.
86
After calcination, the capability of the Al-Co/Fe(2.5)Cal catalyst to grow
CNTs is significantly improved. Figure 5.12 (b) depicts that the CNTs grew with
high density and the entire catalyst surfaces are fully grown by the CNT bundles. The
nanotubes are twisted and grown in bundles. Uniform diameter distribution of the
nanotubes also being observed in this case. The growth direction of the CNTs is
found to be random and unordered.
Surprisingly, when the Al-Co/Fe(3.0) catalyst was used, aligned CNTs were
observed (Figure 5.12 (c)). The arrays of CNTs grown orderly with length up to
10 µm. In addition, the as-grown nanotubes possess relatively small diameter of
36 nm with narrow diameter distribution. In Figure 5.12 (d), we can see that most of
the as-grown CNTs over this catalyst are in alignment. The texture of the nanotubes
is fluffy. It can be deducted that the purity and graphitization of the CNTs are good
because there is no amorphous carbon and unwanted carbonaceous materials
deposited on the sample.
The Al-Co/Fe(3.0)Cal catalyst showed excellent activity in growing the
aligned CNTs. As depicted in Figure 5.12 (e) and (f), the nanotubes grown in the
same direction and arranged in arrays. The density of the CNTs is extremely high,
the surface of the catalyst were highly saturated by the CNTs arrays. The quasialignment of the CNTs is clearly demonstrated in Figure 5.12 (g) and (h). The
diameter of the nanotubes is down to 25 nm. The nanotubes are parallel to each other
to form an array of CNTs. Bunch of nanotubes grew from a cluster of catalyst
particles, as pointed by white circles in Figures 5.12 (f) and 5.12 (h). The purity and
quality of the as-grown CNTs is undoubtedly superior compared to other Al2O3
supported catalysts. As a conclusion, all of the supported bimetallic catalysts are
suitable to grow CNTs with good quality and purity.
87
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
Figure 5.12: SEM micrographs of CNTs grown over various of Al-Co/Fe catalysts:
(a) Al-Co/Fe(2.5) (10000×), (b) Al-Co/Fe(2.5)Cal (10000×), (c) and (d)
Al-Co/Fe(3.0) (10000×), (e) to (h) Al-Co/Fe(3.0)Cal (2500×, 5000× and 10000×).
88
The excellent performance of the Al-Co/Fe(3.0) and Al-Co/Fe(3.0)Cal
catalyst in growing quasi-aligned CNTs can be attributed to their high crystallinity
shown in the XRD patterns. Although calcination did not bring any significant
changes to the physical properties of Al-Co/Fe(3.0)Cal catalyst, but it developed a
better catalytic performance. In addition, Kukovecz et al. (2000) reported that the
bimetallic catalyst Co/Fe gave the best quality and quantity of CNTs. Willems et al.
(2000) have accounted that mixture of metals greatly improve the quantity of carbon
deposition with the 1:1 ratio of metals mixture giving the best yield. Yet again, the
high activity of mixed metals in producing nanotubes has been proved by Willems et
al. (2002). Moreover, the presence of aligned CNTs in the Co/Fe catalysts is in
agreement with Mukhopadhyay and Mathur (2003) result, which revealed that the
present of both Co and Fe metal catalysts created the alignment of the CNTs.
5.1.4.4 Conclusion of SEM Analysis
Generally, all of the Al2O3 supported catalysts are active in growing CNTs
with a variety of efficiency, except Al-Co(2.5) catalyst. The as-grown CNTs over
Al2O3 supported catalysts showed satisfactory purity and density.
Amongst the Co catalysts, the Al-Co(3.0) catalyst is the best by forming high
purity CNTs in fluffy texture. Compared to the Fe catalysts, the Co catalysts
performed better in forming higher density of CNTs. The Al-Fe(3.0)Cal catalyst was
identified as the best Fe catalyst as it grew mass of nanotubes with high uniformity.
Nagaraju et al. (2002) revealed that the alumina support is basic, the exposed
oxygen anions sites on the alumina support can interact with the catalyst particles.
The oxygen anions sites act as Lewis base sites donating negative charge to the metal
nanoparticles. This interaction is a weak metal-support interaction. The partial
negative charge accepted from alumina can enhance the catalytic activity of the
metal nanoparticles as described by Davis and Klabunde (1982).
89
Kukovecz et al. (2000) described that the supported Fe catalysts are efficient
in producing CNTs of somewhat lower quality, as the Co catalysts produced good
quality nanotubes. However, the bimetallic catalyst Co/Fe gave the best quality and
quantity of CNTs.
In this work, the Al2O3 supported bimetallic catalysts showed outstanding
performance in terms of density and quality of as-grown CNTs. The bimetallic
catalysts enable the growth of quasi-aligned CNTs. The alignment of CNTs was
further improved by calcination as occurred to the CNTs grown from the AlCo/Fe(3.0)Cal catalyst. Therefore, the ability of Al2O3 supported catalysts to grow
CNTs is established with bimetallic catalysts performing the best activity.
The results suggest that calcination is essential for achieving the best
dispersion, activity and uniformity of the catalyst particles (Vander Wal et al., 2001).
After calcination, a non-volatile, non-wetting and highly dispersed catalyst oxide is
formed. The partial reduced alumina support transfers a larger amount of negative
charge to the highly dispersed catalyst particles via Lewis base sites. Therefore
through calcination, a better catalyst-support system is achieved.
5.1.5
Field Emission-Scanning Electron Microscopy (FE-SEM) Analysis
In this work, only the CNTs grown over Al-Co/Fe(3.0)Cal catalyst was
examined by field emission-scanning electron microscope. The FE-SEM
micrographs provide a view of the surface or topology of the sample.
The FE-SEM micrographs of the CNTs grown over Al-Co/Fe(3.0)Cal
catalyst are shown in Figure 5.13 (a) and (b). At low resolution (Figure 5.13 (a)), the
as-grown CNTs are observed in bundles with uniform diameter. It is clearly shown
that the surface of catalyst particles was fully covered by CNT bundles, this proved
that the Al-Co/Fe(3.0)Cal catalyst gave high yield of CNTs. The alignment of CNTs
90
was not observed here as it was destroyed during the sample preparation for FE-SEM
analysis.
(a)
(b)
Figure 5.13: FE-SEM micrographs of as-grown CNTs over Al-Co/Fe(3.0)Cal
catalyst, (a) 60000× magnification and (b) 250000× magnification.
High resolution FE-SEM micrographs (Figure 5.13 (b)) had proven that the
Al-Co/Fe(3.0)Cal catalyst is able to grow CNTs instead of carbon nanofibres. The
smooth surface and the hollow structure of CNTs can be observed in the
micrographs. The outer diameter of the CNT is estimated as 7.14 nm and length up to
400 nm. The growth direction of the CNTs is random and the nanotubes are
entangled and twisted. It is observed that each nanotube ended by showing a white
spot (as pointed by the arrows in Figure 5.13 (b)). These white spots originated from
the metal catalyst particles, which indicates that the CNTs were grown by tip growth
mechanism. This observation is similar to the work of Shajahan et al. (2004). The
substrate-catalyst interaction via Lewis base sites is weak, hence the catalyst particles
are lifted up by the growing CNT and continue to promote CNT growth at its tip (tip
growth mechanism), as described by Ando et al. (2004).
91
5.1.6
Transmission Electron Microscopy (TEM) Analysis
The high resolution images of transmission electron microscopy (TEM)
analysis revealed the density and the nature of CNTs. From the transparent images,
we are able to observe the hollow structure and define a material as the CNT. We
also can calssified it as a SWNT or a MWNT depend on the number of wall layers.
In addition, the degree of graphitization and turbostratic arrangement of the CNT
wall structure also can be observed.
The TEM micrographs of as-grown CNTs obtained from Al-Co(3.0) catalyst
are presented in Figure 5.14 (a) and (b). As depicted in the low resolution image
(Figure 5.14 (a)), the CNTs grown in random direction from the Co catalyst base and
extended with an open end. The CNTs appeared as entangled bundles. The CNTs
possessed uniform diameter. The CNTs produced by Al-Co(3.0) catalyst are
MWNTs as observed from Figure 5.14 (b). The nanotube walls are turbostratic but
not straight, two sides of the MWNT has different layers of wall. The inner diameter
was estimated as 12.28 nm whereas the outer diameter was 30.64 nm. Some
amorphous carbon was attached to the outer layer of the MWNT. However, the
MWNT shows high degree of graphitization with low defect observed on the wall
structure.
(a)
(b)
Open end
of a CNT
Amorphous
carbon
Figure 5.14: TEM micrographs of as-grown MWNTs produced by Al-Co(3.0)
catalyst, (a) scale bar: 100 nm and (b) scale bar: 20 nm.
92
Figure 5.15 (a) and (b) represent the TEM micrographs of CNTs grown over
Al-Fe(3.0) catalyst. As observed from Figure 5.15 (a), the CNTs are short and
twisted and presented as aggregates, only a few nanotubes appeared as long and
straight tubes. The presence of aggregates rather than bundles of nanotubes is due to
a poor graphitization process, as concluded by Willems et al. (2002). The walls of
nanotube are not straight. The dark spots in the micrographs were caused by overlapping of carbon layers. From Figure 5.15 (b), this CNTs product can be classified
as MWNT, but the graphite layers are not parallel. The MWNT was curved and
covered by amorphous carbon at the outer layer. This MWNT consist of 7 layers of
the graphene walls with inner and outer diameter of 8.39 and 14.84 nm respectively.
(a)
(b)
Aggregates
of CNT
Amorphous
carbon
Figure 5.15: TEM micrographs of MWNTs formed by Al-Fe(3.0) catalyst, (a) scale
bar: 100 nm and (b) scale bar: 10 nm.
Many irregularities were observed in the nanotubes walls of the Al-Fe(3.0)
catalyst grown MWNT, as depicted in Figure 5.15 (a). The decomposition process of
the hydrocarbon and the CNTs growth are not well coordinated. The Fe catalyst is so
active in the decomposition that the arrangement of the carbon building units is poor.
The same observation was published by Willems et al. (2002).
The degree of graphitization and quality of the MWNTs from Al-Co(3.0)
catalyst are better compared to Al-Fe(3.0) catalyst and it is in agreement with the
SEM analysis. In Colomer et al. (1999) work, the micrographs similar to Figure 5.15
93
(a) were obtained from CNTs grown over Al2O3-Co/Ni/Fe catalyst. Therefore, we
assume that the Ni catalyst did not facilitate the graphitization of CNTs.
For the synthesis using bimetallic catalysts, the as-grown CNTs are expected
to possess higher quality and degree of graphitization, as revealed in the SEM and
FE-SEM analysis. As explained previously, the Co catalyst help to increase
graphitization of nanotubes, whereas Fe catalyst is effective in decompose C2H2 for
nanotubes growth. This is in agreement with Colomer et al. (1999) work, which
explained that the catalysts prepared from a mixture of metals are more efficient for
CNT production than those prepared with a single metal.
In Figure 5.16 (a), again the CNTs grown over Al-Co/Fe(2.5)Cal catalyst
were in the form of aggregates. The CNTs produced are MWNTs as indicated by
Figure 5.16 (b). The carbon layers of the MWNT are crooked and the degree of
graphitization is low. There is less amorphous carbon attached on the outer surface of
MWNT compared to CNTs form by Fe catalyst. This proved that the Co catalyst help
to reduce the disarrangement of carbon building units.
When the loading of Co/Fe catalyst was increased to 3.0 wt.%, the quality of
CNTs product was improved (Figure 5.16 (c)), the bundles of CNTs observed instead
of nanotubes aggregates. The CNTs grown are longer and the numbers of crooked
CNTs were less. The well-graphitized and low defect CNTs were seen in Figure 5.16
(d), the wall of this MWNT were made up of 7 graphene layers. The inner was
calculated as 5.00 nm and the outer diameter was 14.29 nm.
94
(a)
(b)
Aggregates
of CNT
Amorphous
carbon
(d)
(c)
Turbostratic
MWNT
Figure 5.16: TEM micrographs of as-synthesized MWNTs over Al-Co/Fe catalysts:
(a) and (b) Al-Co/Fe(2.5)Cal (scale bar: 50 nm and 10 nm) as well as (c) and (d) AlCo/Fe(3.0) (scale bar: 100 nm and 10 nm).
The capability of Al-Co/Fe(3.0)Cal catalyst to grow good quality CNTs is
superior. As illustrated in Figure 5.17 (a), the CNTs were very long and wellgraphitized with uniform diameter. There is a catalyst particle trapped at the tip of
the MWNT, as pointed in the micrograph. In Figure 5.17 (b), we observed a MWNT
with very thin walls, it only consists of 6 layers of graphene walls. The MWNT
possessed inner diameter of 4.21 nm and outer diameter of 8.16 nm. From Figure
5.17 (c), the MWNT is free from amorphous carbon deposition, its layered wall are
turbostratic and free from defect. Thus, the TEM analysis is again assured the high
performance of the bimetallic catalysts in growing high quality MWNTs.
95
(a)
(b)
Catalyst
particle
(c)
Figure 5.17: TEM micrographs of MWNTs grown over Al-Co/Fe(3.0)Cal catalyst:
(a) scale bar: 100 nm, (b) scale bar: 5 nm and (c) scale bar: 10 nm.
5.1.7
Energy Dispersive X-ray (EDX) Analysis
The energy dispersive X-ray (EDX) analysis was carried out to measure the
surface compositions of the catalyst. This information is about the fraction of the
active metals dispersed over the surface of catalyst. The analytical data for various
catalysts is tabulated in Table 5.8. It can be seen that oxygen (O) and the aluminium
(Al) are the main elements in the Al2O3 support and the silicon (Si) is only present in
trace amount. For all of the Co catalysts, the impregnation of cobalt (Co) catalyst
was successful to load the Co on the surface of Al2O3 support. The amount of Co
96
detected is higher than the Co concentration used in impregnation, this is due to the
adsorption of the Co solution at surface of the support without penetration into the
whole support. The Co loading used (2.5 and 3.0 wt.%) are calculated based on the
weight of the whole support (Appendix A), so it is reasonable to detect higher
amount of Co catalyst on the surface. This observation also occurred in the Fe
catalysts and bimetallic catalysts.
Table 5.8: EDX analysis data of surface compositions of various Al2O3 supported
catalysts before and after decomposition of C2H2.
Surface Elements (wt.%)
O
Al
Si
Co
Fe
8
Al2O3
40.92 57.27 1.81
8
43.55 52.80
3.65
Al–Co(2.5)
9
28.18 42.97
3.85
8
38.66 56.68
4.66
Al–Co(2.5)Cal
9
18.09 32.25
5.05
8
41.84 49.20
8.96
Al–Co(3.0)
9
15.01 36.69
6.01
8
24.81 59.27 4.28 11.64
Al–Co(3.0)Cal
9
17.22 33.73
3.31
8
42.22 44.59
13.19
Al–Fe(2.5)
9
16.70 31.38
6.07
8
41.98 49.51
8.51
Al–Fe(2.5)Cal
9
10.22 19.27
16.17
8
23.32 59.21
17.47
Al–Fe(3.0)
9
13.18 39.83
7.74
8
27.67 65.26
7.07
Al–Fe(3.0)Cal
9
14.58 38.65
7.37
8
31.71 48.86
13.34 6.09
Al–Co/Fe(2.5)
9
5.24 23.65
7.62
9.99
8
34.34 57.80
3.71
4.15
Al–Co/Fe(2.5)Cal
9
11.27 35.80
4.56
3.27
8
35.98 55.51
5.21
3.30
Al–Co/Fe(3.0)
9
5.09 15.60
3.85
1.70
8
36.52 56.91
3.58
2.99
Al–Co/Fe(3.0)Cal
9
5.91 19.06
1.84
1.19
* Reaction: decomposition of C2H2 at 700 °C for synthesis of CNTs.
8: before decomposition; 9: after decomposition.
Samples
Reaction*
C
25.00
44.61
42.29
45.74
45.85
54.34
39.25
39.40
53.50
45.10
73.76
72.00
Without any exception, the amount of Co element increased with increasing
of catalyst loading from 2.5 wt.% to 3.0 wt.%, as Al-Co(2.5) catalyst contained
3.65 wt.% of surface Co and Al-Co(3.0) catalyst hold 8.96 wt.% of Co. The Fe
97
catalysts also found to behave in the same way. Conversely, the bimetallic catalysts
loading decreased with increasing of Co/Fe loading. It is proposed that when the
amount of Co and Fe used increased, the catalyst hindered each other to adsorb on
the support, resulting lower metal composition.
For the Co catalysts (2.5 and 3.0 wt.%), the Co element increase after the
calcination. This indicates that the calcination help in rearrange the oxide phase of
the catalyst resulting in better dispersion of the catalyst particles on the support. This
assumption is supported by XRD patterns of calcined Co catalysts which showed
higher intensity of Co3O4 face-centred cubic phase. However, for both of the Fe
catalysts, the Fe element decreased after calcination. The Fe particles are believe to
aggregate and form bigger particles after calcination, but bigger particles have lower
surface area as a result lower Fe composition was detected in the EDAX analysis
(Table 5.8).
In the bimetallic catalysts, although the amount of Co and Fe loadings are the
same but the Fe metal showed higher composition. From the XRD diffractograms,
the Fe2O3 rhombohedral phase is more dominant than the Co3O4 face-centred cubic
phase, most probably, the Fe2O3 rhombohedral layer lies on top of the Co3O4 facecentred cubic species.
All of the supported catalysts are able to decomposed C2H2 into carbonaceous
materials which detected as surface carbon elements. The Al-Co(2.5) catalysts gave
the lowest carbon content, as it is unable to generate CNTs. Other Co catalyst have
carbon content in the range of 42.29 to 45.74 wt.%, apparently these C component
are CNTs. The carbon contents of supported Fe catalysts are in the range of 39.25 to
54.34 wt.%. The high efficiency of bimetallic catalysts to produce CNTs is proven
by the high carbon content from EDX analyses. The Al-Co/Fe(3.0) and AlCo/Fe(3.0)Cal catalysts showed extremely high carbon content of about 70 wt%.
These observations were supported by the high density of CNTs yield as observed
from the SEM micrographs. These EDX data are concurrent with the carbon
deposition yield (%) which discussed previously in section 4.1.3.
98
5.2
Anodic Aluminium Oxide (AAO) Supported Catalysts
In order to produce ordered CNTs arrays, the template synthesis method was
chosen. In this work, a new and simple template synthesis method was proposed. It is
done by dip coating of the anodic aluminium oxide (AAO) template with a metal
acetate solution followed by thermal decomposition of C2H2 into the nanochannels of
the AAO template. The metal catalysts used are Co, Fe and mixture of Co and Fe.
The acetate salt of these metals is utilised as the source of the metal in the AAO
supported catalysts. The concentrations of Co or Fe catalyst solutions were either 0.5
or 1.0 M.
5.2.1
Physical Appearance of AAO Supported Catalysts
The AAO template is initially white in colour and it is not hygroscopic so it
can be used in the dip coat process directly. Following the dip coating process, the
AAO templates were incorporated with Co, Fe or Co/Fe catalyst, shown by the
colour changes. The AAO supported catalysts preserve in the form of template after
drying at 90 °C. All of the AAO-Co catalysts showed purple pinkish colour, whereas
the AAO-Fe catalysts showed brownish colour and the bimetallic catalysts with light
orange colour. The catalysts of 3.0 wt.% metal loading gave more intense colour
compared to catalysts of 2.5 wt%. After calcination, the AAO-Co catalysts changed
to dirty green and the AAO-Fe catalyst turned into dark orange as well as the
AAO-Co/Fe catalysts became olive green.
99
5.2.2
X-ray Diffraction (XRD) Analysis
The X-ray diffraction (XRD) patterns showed that all of the AAO supported
catalysts exist as amorphous materials, as depicted in Figure 5.16. It is found that no
peak can be assigned as the Co or Fe species present in the samples, the XRD
patterns were full of interference. Therefore, the XRD analysis is not a suitable
characterisation method for the AAO supported catalysts.
Intensity (%)
AAO-Co(0.5)
20
AAO template
30
50
40
60
70
2θ (°)
Figure 5.18: X-ray diffractogram patterns of AAO template and AAO-Co(0.5)
catalyst.
5.2.3
The As-Grown Carbon Deposition Yield and Physical Appearance
The AAO template is brittle and has low thermal stability, thus after the
decomposition at high temperature, the product was cracked into pieces. In all
reaction, the carbon deposit on the AAO template catalyst was black in colour with
different thickness and density. The difference of thickness and density provide
general information about the activity of the catalysts. When the carbon deposit
appeared thick and bulky with deposition of black powder, the catalyst is active in
100
forming CNT. If the carbon deposition on the template catalyst is thin and no powder
deposition observed, indicates that the catalyst activity was low. The weight percent
of carbon deposition yield, the physical appearance and the density of the as-grown
CNT were shown in Table 5.7. The carbon deposition yield was calculated from
Equation 4.1.
Table 5.7: The AAO template supported catalysts and the yield (%), the physical
appearance and density of the as-synthesized carbon deposit in CCVD of C2H2.
Carbon
Catalysts
deposition
yield (%)
AAO-Co(0.5)
-14.07
AAO-Co(0.5)Cal
18.07
AAO-Co(1.0)
-12.89
AAO-Co(1.0)Cal
19.14
AAO-Fe(0.5)
-11.41
AAO-Fe(0.5)Cal
4.68
AAO-Fe(1.0)
-2.88
AAO-Fe(1.0)Cal
4.85
AAO-Co/Fe(0.5)
-4.52
AAO-Co/Fe(0.5)Cal
46.15
Physical appearance
Density
++
Thin carbon deposition, fine powder
+
Thin deposition, fine powder
+++
Thick deposition, lots of fine powder
+
Thin deposition, no powder
Thin deposition, no powder
+
Thin deposition, no powder
+
Thin deposition, fine powder
++
Thin deposition, no powder
+
++
Thick deposition, fine powder
Very thick deposition, lots of fine
+++
powder
AAO-Co/Fe(1.0)
-5.59
Very thick deposition, lots of fine
+++
powder
AAO-Co/Fe(1.0)Cal
21.57
++
Thick deposition, fine powder
* Density of carbon deposition yield from observations; +: low density, ++: moderate
density, +++: high density.
From the physical appearance of the carbon deposit, we found that all of the
AAO supported catalysts were capable of converting C2H2 into carbonaceous
product. It is noticed that some of the carbon deposition yield (%) showed negative
values. According to Kathyayini et al. (2004), the metal oxides support lost about 18
to 55 % of their initial weight under the flow of nitrogen at 700 °C, i.e. the
temperature chosen for CNT production. For that reason, if the carbon deposit
generated is less than the weight loss of the support, the deposition yield will show
negative value. The catalysts showed carbon deposition yield (%) of positive values
generated higher amount of deposition.
101
The density of the products demonstrated that the bimetallic catalysts
produced greater quantity of carbon deposition compared to individual Co and Fe.
The AAO-Fe catalysts produced lowest deposition yield. The ability of the catalysts
to produce carbon deposit vary in the order of Co/Fe > Co > Fe catalysts.
5.2.4
Scanning Electron Microscopy (SEM) Analysis
5.2.4.1 SEM Analysis of AAO Supported Cobalt Catalysts and the CNTs Yield
The micrograph of anodic aluminium oxide (AAO) template (Figure 5.19)
showed that the template is porous with pore diameters in the range of 190 to
220 nm.
Figure 5.19: SEM micrograph of anodic aluminium oxide (AAO) template
(20000 ×).
Subsequent to the impregnation of Co catalyst, the pores became wider, as
illustrated in micrograph of AAO-Co(0.5)Cal catalyst (Figure 5.20 (a)). Some of the
walls of the pore collapsed and two pores combined to become a larger pore. This
phenomenon occurred because the cobalt acetate solution used is slightly acidic and
corrosive, it react with the AAO template to widen the pores. The non-calcined
AAO-Co(0.5) catalyst has similar morphology. For the AAO-Co(1.0) catalyst, the
wall thickness increased compared to the AAO-Co(0.5) catalyst, as seen in Figure
102
5.20 (b). It can be explained by the higher loading of Co catalyst (1.0 M) resulting in
the higher deposition of Co on the template walls.
(b)
(a)
Figure 5.20: SEM micrographs of AAO template supported Co catalysts: (a) AAOCo(0.5)Cal (20000 ×) and (b) AAO-Co(1.0)Cal (20000 ×).
SEM micrographs of as-synthesized carbon deposits on different AAO-Co
catalysts at 700 °C under the flow of C2H2/N2 mixture are shown in Figure 5.21 (a) –
(h). The micrographs revealed that each sample produced CNTs with different
density and configuration. The cross section view of carbon deposit on AAO-Co(0.5)
catalyst are shown in Figure 5.21 (a). It was observed that CNTs were grown on the
wall of the template. The CNTs are short and winding with uniform diameter. The
diameter of the CNTs is much smaller than the nanopore diameter. Interestingly,
there are carbon nanofibers (CNFs) growth over the AAO-Co(0.5) catalyst as
depicted in Figure 5.19 (b). This product is classified as CNF because it has an
average diameter of 0.56 µm. The straight CNFs are seen grown out from the
nanochannels in the form of arrays.
When the AAO-Co(0.5)Cal catalyst was used, the CNTs growth was
improved (Figure 5.21 (c) and (d)). From the cross section view of Figure 5.21 (c),
small quantity CNTs grown from the wall of the nanochannels are found to be
twisted and short. In contrast, the density of CNTs grown on the surface of the
catalyst template is high, as observed in Figure 5.21 (d). The CNTs bundles are fluffy
and having regular diameter. There is no CNFs growth observed in this sample.
103
(a)
(b)
CNFs
(c)
(d)
CNTs
(e)
(f)
CNTs
CNFs
(g)
(h)
Coiled CNTs
Figure 5.21: SEM micrographs of carbon deposit over various Co catalysts: (a) and
(b) AAO-Co(0.5) (10000× and 2000×), (c) and (d) AAO-Co(0.5)Cal (10000×), (e)
and (f) AAO-Co(1.0) (20000× and 10000×), (g) and (h) AAO-Co(1.0)Cal (10000×).
104
Figure 5.21 (e) revealed that the AAO-Co(1.0) catalyst showed higher
activity in decomposition of C2H2, as the plenty of carbon deposition was obtained
after the reaction. The CNTs fully covered up the catalyst surface, these CNTs are
crooked and massive due to the amorphous carbon deposition. The cross section
view of this sample is similar with Figure 5.19 (c), except for the length of the CNTs
is higher. Figure 5.19 (f) showed that the AAO-Co(1.0) catalyst also active in
forming CNFs arrays with average diameter of 0.86 µm. In between the CNFs, there
were a few nanotubes observed, as pointed in the micrographs. These CNTs are long
and winding, growing on the CNF surface.
The CNTs were grown actively on the AAO-Co(1.0)Cal catalyst as illustrated
in Figure 5.19 (g). The CNTs are high in density and purity, it is sinuous but possess
uniform diameter. The length of CNTs also greatly improved. In Figure 5.19 (h), the
forest of CNTs was detected on the surface of the catalyst without any side product, a
few coiled CNTs were also observed. As reported by Lu et al. (2004b), the Co
catalyst is used to grow coiled CNTs by catalytic decomposition of C2H2.
The growth of CNFs is related to the big catalyst particles obtained from the
dip coating of AAO template. As discussed by Liu, D.Y. et al. (2004), the big
catalyst particles favour the growth of CNF instead of CNTs. In addition, the carbon
deposition yield (%) of the AAO-Co(0.5)Cal and AAO-Co(1.0)Cal catalysts showed
negative values. This indicates that the transformation of the carbon precursor to
CNTs is less effective, therefore producing CNFs instead of CNTs. In conclusion, all
of the AAO supported Co catalysts are active in forming CNTs with different
effectiveness.
105
5.2.4.2 SEM Analysis of AAO Supported Ferrum Catalysts and the CNTs Yield
The AAO-Fe catalysts have similar morphologies with the AAO-Co
catalysts. The widening of nanopores and collapse of walls are also occurred with the
AAO-Fe catalysts after the dip coating process, as observed in Figure 5.22 (a).
Exceptionally, the AAO-Fe(1.0)Cal catalyst showed the deposition of Fe catalyst
particles on the surface of AAO template. Figure 5.22 (b) showed the Fe particle
clusters have undefined shape, it deposited on the surface of the template.
(a)
(b)
Figure 5.22: SEM micrographs of different AAO-Fe catalysts: (a) AAO-Fe(0.5)Cal
catalyst (20000×) and (b) AAO-Fe(1.0)Cal catalyst (5000×).
The SEM micrographs of the AAO-Fe catalysts after decomposition of C2H2
have clearly attested the ability of these catalysts in producing CNTs. Figure 5.23 (a)
is the image of as-grown CNTs over AAO-Co(0.5) catalyst, the density is high and
the CNTs reached satisfactory length with no impurities observed. Analogously, the
AAO-Fe(0.5)Cal catalyst also produced high density CNTs (Figure 5.23 (b)), the
nanotubes possess uniform diameter of 59.65 nm. In addition, the AAO-Fe(1.0)
catalyst seemed to have equal activity with the 0.5 M loading catalysts. In Figure
5.23 (c), the as-grown CNTs do not have significant different with the previous case.
Conversely, some interesting images were captured from the yield of
AAO-Fe(1.0)Cal catalyst. The nanotubes are seen to have grown from the Fe catalyst
layer on the surface of AAO template, but it did not grow inside the nanochannel.
This indicated that the dip coating of catalyst onto the walls of nanochannel is not
106
successful to direct the growth of CNTs within the nanochannels. Apparently, the
catalytic activity of AAO-Fe catalysts is better than the AAO-Co catalysts as the
AAO-Fe catalysts produced CNTs of high density without growth of CNF.
(a)
(b)
(c)
(d) Fe catalyst layer
Figure 5.23: SEM micrographs of as-grown CNTs over various AAO-Fe catalysts:
(a) AAO-Fe(0.5) (10000×), (b) AAO-Fe(0.5)Cal (10000×), (c) AAO-Fe(1.0)
(10000×) and (d) AAO-Fe(1.0)Cal (10000×).
5.2.4.3 SEM Analysis of AAO Supported Cobalt/Ferrum Catalysts and the
CNTs Yield
The AAO supported Co/Fe catalysts share the same morphology regardless of
the catalyst loading and calcination. A representative SEM image of AAO-Co/Fe
catalyst is shown in Figure 5.24 (a), where it was noticed that the catalyst particles
were deposited inside the nanochannels. The coupling of Co and Fe catalyst is
successful to deposit the catalyst into the nanochannels. In the case of bimetallic
107
catalyst, the formation of catalyst particles on the surface of the template became
more dominant (Figure 5.24 (b)). These catalyst particles were observed on all of the
bimetallic catalysts, except the AAO-Co/Fe(0.5) catalyst.
(a)
(b)
Figure 5.24: Micrographs from SEM analysis of AAO supported bimetallic
catalysts: (a) AAO-Co/Fe(0.5) (10000×) and (b) AAO-Co/Fe(1.0) (5000×).
In relation to the high carbon deposition yield (%) and good physical
appearance, the AAO supported bimetallic catalysts are predicted to perform better
than the individual Co or Fe catalysts. Figure 5.23 (a) depicts that the
AAO-Co/Fe(0.5) catalyst grown CNTs with high density. The catalyst surface was
wrapped by a thick layer of CNTs, the CNTs grown in random direction with
uniform diameter. The purity of the CNTs is high, no amorphous carbon was
observed.
The pairing of Co and Fe catalyst has found to promote the growing of long
coiled CNTs with length more than 10 µm, as observed in Figure 5.23 (a). This is a
new discovery as no researcher ever used bimetallic catalyst to produce coiled CNTs.
According to Lu et al. (2004b), the inhomogeneous growth activity through the
larger and irregular particle favours the formation of coiled nanotubes.
108
(a)
(b)
Long coiled
CNTs
(d)
(c)
CNT
arrays
AAO
catalyst
(f)
(e)
(g)
Figure 5.25: SEM micrographs of CNTs grown over: (a) AAO-Co/Fe(0.5) (10000×),
(b) and (c) AAO-Co/Fe(0.5)Cal (10000× and 5000×), (d) and (e) AAO-Co/Fe(1.0)
(10000× and 5000×), (f) and (g) AAO-Co/Fe(1.0)Cal (5000× and 10000×).
109
After calcination, the CNTs yield of AAO-Co/Fe(0.5)Cal was greatly
enhanced as shown in Figure 5.25 (b). Two different configurations of CNTs were
obtained, i.e. the bundles and arrays. However, the graphitization degree of the CNTs
is low which showed by the rough surface of the CNTs. The arrays of CNT are found
to have grown out from the nanochannels of the template catalyst, as seen in Figure
5.25 (c). This is due to the deposition of the catalyst particles inside the nanochannels
as depicted in Figure 5.24 (a).
Interestingly, the AAO-Co/Fe(1.0) and AAO-Co/Fe(1.0)Cal catalysts
behaved in the same way with the 0.5 M catalyst concentration. The
AAO-Co/Fe(1.0) catalyst generated bundles of CNTs, as in Figure 5.25 (d) and (e).
The nanotubes are high in density with acceptable purity. Similarly, the
AAO-Co/Fe(1.0)Cal catalyst grew both the bundles (Figure 5.25 (f)) and arrays of
CNTs (Figure 5.25 (g)), the arrays of CNTs were grown from the nanochannels. The
density of the CNTs grown over AAO-Co/Fe(1.0)Cal catalyst is the highest.
There are two types of growth mechanism observed from the bimetallic
catalysts; (1), the CNTs bundles grown from the catalysts particles on the template
surface; (2), the arrays of CNT grown over the catalyst inside the nanochannels of
the template catalyst. Thus, the bimetallic catalysts produce higher amount of CNTs.
5.2.4.4 Conclusion of SEM Analysis
In general, all of the AAO supported catalysts are active in producing CNTs.
The AAO-Co(0.5)Cal and AAO-Co(1.0)Cal catalysts showed less efficiency in
converting precursor into CNTs and hence producing CNFs. The AAO supported Fe
catalysts produced purely CNTs but the density and purity of the product is not good
enough. The bimetallic catalysts not only showed high activity in producing CNTs
but also coiled CNTs. Apparently, the bimetallic catalysts are the best catalysts
compared to the individual Co and Fe catalysts. The attempt to synthesize aligned
110
CNTs which copy the shape of the nanochannels was partly achieved by the
bimetallic catalysts.
According to Scokart and Rouxhet, (1982), the alumina in AAO template
itself is a catalyst in the hydrocarbon decomposition process due to the Lewis acid
nature of its surface sites. Jing et al. (1998) reported that the CNTs are formed by the
joint catalytic activities of metal particles and alumina walls. Subsequent to the
decomposition of C2H2 initiated by the metal particles, the carbide transport along
the surface of the metal nanoparticles to the alumina channel wall, where tube
growth proceeds. In this study, this phenomenon was demonstrated in Figure 5.19 (c)
and (g), 5.21 (a) as well as 5.23 (c) and (g), which the nanochannel walls was
covered by carbon flakes that form a tube-like structure.
5.2.5
Energy Dispersive X-ray (EDX) Analysis
The Energy Dispersive X-ray (EDAX) analysis was carried out to measure
the surface compositions of the AAO supported catalyst. This information is about
the fraction of the active metals dispersed over the surface of the catalyst. The
analytical data for various catalysts is tabulated in Table 5.10.
Initially, the oxygen (O) and aluminium (Al) were found as the main
components in the AAO template. The Co and Fe elements do not exist in the AAO
template. Subsequent to the dip coating of Co onto the AAO template, the Co
catalyst was detected in all of the AAO-Co catalysts. The AAO-Co(0.5) catalyst
contained 5.14 wt.% Co on the surface, the calcination increased the Co contents to
14.98 wt.%. However, the 1.0 M loading Co catalysts show dissimilar trend. The
AAO-Co(1.0) catalyst has 12.11 wt.% of Co whereas the AAO-Co(1.0)Cal catalyst
contained 7.45 wt.% of Co. The amount of Co increased after the decomposition of
C2H2 suggest that the CNTs grown in the tip growth mode. The Co catalyst particles
located at the tips when the CNTs grew and it appeared on the surface and resulting
higher composition.
111
After the decomposition of C2H2, all of the AAO-Co catalysts were
effectively converted the precursor into carbon materials, the C compositions of the
yield were varied from 40.90 to 73.08 wt.%. This can be related to the physical
appearance of the Co catalysts deposition yield, which are thick and powdery. The
high content of the C is also attributed to the existence of CNFs.
Table 5.10: EDX analysis data of surface compositions of various AAO supported
catalysts before and after decomposition of C2H2.
Surface Elements (wt.%)
O
Al
Co
Fe
8
AAO
32.99
67.01
8
29.01
65.85
5.14
AAO–Co(0.5)
9
7.93
12.84
11.69
8
23.51
61.51
14.98
AAO–Co(0.5)Cal
9
5.91
8.70
16.90
8
30.31
57.58
12.11
AAO–Co(1.0)
9
5.74
9.33
11.85
8
27.49
65.06
7.45
AAO–Co(1.0)Cal
9
6.72
39.22
13.16
8
32.38
63.21
4.41
AAO–Fe(0.5)
9
21.54
52.37
3.77
8
25.24
66.50
8.26
AAO–Fe(0.5)Cal
9
29.48
48.20
1.77
8
31.31
54.88
13.81
AAO–Fe(1.0)
9
11.94
48.42
5.47
8
23.64
42.49
33.87
AAO–Fe(1.0)Cal
9
20.47
56.21
4.59
8
25.38
64.82
6.89
2.91
AAO–Co/Fe(0.5)
9
8.28
36.40
4.89
2.14
8
25.39
64.54
5.16
4.91
AAO–Co/Fe(0.5)Cal
9
2.95
26.67
4.73
1.86
8
26.60
54.54
13.24
5.62
AAO–Co/Fe(1.0)
9
10.15
0.86
1.51
0.97
8
24.49
41.01
16.73
17.77
AAO–Co/Fe(1.0)Cal
9
1.97
28.68
3.61
1.71
* Reaction: decomposition of C2H2 at 700 °C for synthesis of CNTs.
8: before decomposition; 9: after decomposition.
Samples
Reaction*
C
67.54
68.49
73.08
40.90
22.32
20.55
34.17
18.73
48.29
63.79
86.51
64.03
In the case of Fe catalysts, the amount of Fe increases regards of the
increment of Fe loading, i.e. from 4.41 to 13.81 wt.%. The calcination also helps in
raising the Fe contents, higher Fe compositions were detected in the calcined AAO-
112
Fe catalysts, as shown in Table 5.10. The formation of Fe catalyst particles as
observed in SEM analyses, explained the high surface composition of Fe.
However, the C contents of the yield of AAO supported Fe catalysts are
relatively low, which range from 18.73 to 34.17 wt.%. The AAO-Fe catalysts
produced CNTs without CNFs, therefore the C contents were not high. From the
physical appearance observation, the carbon deposition yields of all Fe catalyst are
thin and solid without powder.
Referring to Table 5.10, the Co contents of the AAO-Co/Fe(0.5) catalyst
were in the range of 5.16 to16.73 wt.%, whereas the Fe catalysts contained Fe
composition of 2.91 to 17.77 wt.%. The amount of Co and Fe catalyst is almost equal
and balance. By using these bimetallic catalysts, the decomposition of C2H2 yielded
carbon contents of 48.29 to 86.51 wt.%. The high carbon contents were due to the
high density CNTs produced by the bimetallic catalysts contributed by the two types
of growth mechanism (as discussed in Section 5.2.4.3).
CHAPTER 6
PRODUCTION OF CARBON NANOTUBES (CNTs) OVER MOLECULAR
SIEVES SUPPORTED CATALYSTS
6.1
Molecular Sieves Supported Catalysts
Molecular sieves (MS) are crystalline metal alumino-silicates having a three
dimensional interconnecting network of silica and alumina tetrahedral. The structure
of MS are alike a sponge on a molecular scale. Natural water of hydration is removed
from this network by heating to produce uniform cavities.
MS have a solid framework, defined by large internal cavities where
molecules can be adsorbed. These cavities are interconnected by pore openings
which molecules can pass through, and because of their crystalline nature, the pores
and cavities are all precisely the same size. Depending on the size of the openings,
they can adsorb molecules readily, slowly, or not at all, thus functioning as
‘molecular sieves’ by adsorbing molecules of certain sizes while rejecting larger
ones. Synthetic MS possess the unique ability to selectively adsorb molecules by
size.
The MS utilised in this study is type 3A with 4 – 8 mesh particle size. The
Co, Fe and mixture of Co and Fe are selected as metal catalysts and the acetate salt
of these metals is utilized as the precursor of the metal oxide in the supported
catalysts. Jong et al. (2002) discovered that Cu metal is not suitable to be used as the
114
catalyst and the Ni/Cu mixture catalyst was found showing less effectiveness towards
the growth of CNTs
The metal catalysts (Co or Fe) loadings of supported catalysts were set in the
range of 2.5 – 3.0 wt %. As the particle size of the metal cluster increases with
increase of metal loading (Shajahan et al., 2004). For that reason, the metal catalysts
loading used in this work is low so that the size of the metal clusters is small enough
to be adsorb into the cavities of MS. Fonseca et al. (1996) described that catalyst
particles inside the pores of support can react slowly with the C2H2 molecules during
the decomposition, producing thin nanotubes with lower amount of amorphous
carbon.
The impregnation method was applied in order to integrate the metal catalyst
onto the MS support because Perez-Cabero et al. (2004) found that the impregnation
method allows more homogeneous metallic dispersion on the support compared to
the deposition-precipitation method.
6.2
Physical Appearance of Molecular Sieves Supported Catalysts
Subsequent to the drying at 90 °C, the impregnated MS catalysts remained in
beads form and not hygroscopic. The MS support is originally light beige colour.
After incorporated with either individual Co or Fe or Co/Fe metal catalyst, the
support exhibited different colours. The MS supported Co catalysts showed blue
purplish colour, while the MS supported Fe catalysts showed brownish colour. The
colour of the MS-Co/Fe catalysts is light brown.
Commonly, the MS supported catalysts containing 3.0 wt.% metal showed
more intense colour compared to the catalysts with 2.5 wt.% metal loading.
However, the colour of the catalysts changed after the calcination at 450 °C. The
calcined MS-Co catalysts turn into olive green, whereas the MS-Fe catalysts turned
115
into dark brown colour and MS-Co/Fe catalysts changed to dark dirty green. This
indicates that the incorporation of the metal catalysts onto the MS support showed
different colours.
6.3
X-ray Diffraction (XRD) Analyses
The X-ray diffraction analyses of the molecular sieves support and series of
molecular sieves supported catalysts were carried out in order to determine the active
phases that contribute to a high catalytic property in the synthesis of CNTs. The
effect of preparation parameters such as types of metal catalysts, amount of metal
catalyst loading and effect of calcination have been studied too. Thus, a contributing
factor that resulted in the variation of catalysts activities could be reasoned out. The
structural and phase of the catalysts were identified by comparing the 2θ and d
values of the studied catalysts with that of reference materials from the powder
diffraction file (PDF 2001).
6.3.1
XRD Analysis of Molecular Sieves Supported Cobalt Catalysts
The XRD diffractograms of dried MS support and series of MS supported Co
catalysts are shown in Figure 6.1, the 2θ and d values are tabulated in Table 6.1. The
MS support was observed as NaSiAlO4 face-centred cubic. The NaSiAlO4 facecentred cubic phase was observed at 2θ values of 7.100, 10.127 and 12.431° or d
values of 12.4406, 8.7276 and 7.1144 Å (PDF 2001 d values (Å): 12.3113, 8.7054
and 7.1079).
In the MS-Co(2.5) catalyst, the existence of Co.333Na.333(AlSiO4)(H2O)2.92
cubic as well as the NaSiAlO4 face-centred cubic phase have been detected. The
Co.333Na.333(AlSiO4)(H2O)2.92 cubic phase was observed at 2θ values of 7.146,
10.150 and 23.939° or d values of 12.3594, 8.7077 and 3.7142 Å (PDF 2001 d
116
values (Å): 12.2670, 8.6741 and 3.6986). Meanwhile, the MS support remained in
the same phase. As shown in Figure 6.1, the Co.333Na.333(AlSiO4)(H2O)2.92 cubic
phase overlapping with the NaSiAlO4 face-centred cubic phase, as a result the
intensity of the peaks increased. Calcination at 450 °C did not result any changes to
the Co.333Na.333(AlSiO4)(H2O)2.92 cubic phase except decrement of peaks intensity.
The 2θ and d values of MS-Co(2.5)Cal catalyst are summarised in Table 6.1. The
diffractograms depicted the MS support and MS-Co catalyst to be of crystalline
structure.
When the Co loading increased to 3.0 wt.%, there was no major changes to
the XRD diffractograms. The peaks at 2θ values of 7.124, 10.148 and 23.981° or d
values of 12.3976, 8.7097 and 3.7078 Å (PDF 2001 d values (Å): 12.2670, 8.6741
and 3.6986) was distinctive of Co.333Na.333(AlSiO4)(H2O)2.92 cubic phase. It was also
observed that the main peaks of NaSiAlO4 face-centred cubic phase combined with
the peaks of Co.333Na.333(AlSiO4)(H2O)2.92 cubic, excluding the peak at 2θ = 12.426°
with d value of 7.1176 Å (dref. : 7.1079 Å).
Despite calcination, there was no structural transformation observed in the
MS-Co(3.0)Cal catalysts. The Co.333Na.333(AlSiO4)(H2O)2.92 cubic and NaSiAlO4
face-centred
cubic
phase
were
preserved
but
the
intensity
of
Co.333Na.333(AlSiO4)(H2O)2.92 cubic phase had decreased, as depicted in Figure 5.1.
By comparing the XRD diffraction pattern of the MS support to that of the
Co catalysts, there are similarity in the peak positions. It is elucidated that Co cations
from cobalt acetate had substituted the Na ions in the NaSiAlO4 structure, resulting
the Co.333Na.333(AlSiO4)(H2O)2.92 phase. Through this interaction, the MS support
plays an important role in dispersion of the Co catalyst and control the catalytic
cracking of the hydrocarbon.
117
MS-Co(3.0)Cal
Intensity (%)
MS-Co(3.0)
MS-Co(2.5)Cal
MS-Co(2.5)
MS
5
Figure
6.1:
10
X-ray
40
2θ (°)
diffractogram patterns
20
30
60
50
of
MS
70
support,
MS-Co(2.5),
MS-Co(2.5)Cal, MS-Co(3.0) and MS-Co(3.0)Cal catalysts. ¡: NaSiAlO4 facecentred cubic and T: Co.333Na.333(AlSiO4)(H2O)2.92 cubic phase.
118
Table 6.1: Phases, peak position (2θ) and d values (Å) in the X-ray diffractogram
patterns of the MS support, MS-Co(2.5) and MS-Co(2.5)Cal catalysts.
Samples
Phase Structure
2θ (°)
7.100
MS
10.127
12.431
7.146
NaSiAlO4
10.150
face-centred cubic
12.446
MS-Co(2.5)
7.146
Co.333Na.333(AlSiO4)(H2O)2.92
10.150
cubic
23.939
7.116
NaSiAlO4
10.147
face-centred cubic
12.400
MS-Co(2.5)Cal
7.116
Co.333Na.333(AlSiO4)(H2O)2.92
10.147
cubic
23.954
* dobs : d spacing values obtained from XRD analyses.
dref : d spacing values obtained from the reference.
NaSiAlO4
face-centred cubic
dobs* (Å)
12.4406
8.7276
7.1144
12.3594
8.7077
7.1063
12.3594
8.7077
3.7142
12.4118
8.7100
7.1323
12.4118
8.7101
3.7119
dref * (Å)
(PDF
2001)
12.3113
8.7054
7.1079
12.3113
8.7054
7.1079
12.2670
8.6741
3.6986
12.3113
8.7054
7.1079
12.2670
8.6741
3.6986
With the purpose to determine the active catalyst phases individually, a MS
supported catalyst with 40 wt.% of Co loading were prepared for the XRD analyses.
The MS-Co(40.0) catalyst was calcined at two temperatures, i.e. 450 °C for 17 hours
and 750 °C for 5 hours. The reason to carry out the calcination at 750 °C is to verify
the active catalyst phase appeared during the CNTs synthesis at 700 °C. The high
loading of Co catalyst (40.0 wt.%) showed high intensity peaks in the XRD patterns.
The XRD patterns of MS-Co(40.0)Cal were shown in Figure 6.2 and the data were
tabulated in Table 6.2.
From XRD pattern of MS-Co(40.0)Cal catalyst (calcined at 450 °C), peaks at
2θ values of 36.882, 31.267 and 65.356° with d values of 2.4350, 2.8584 and
1.4267 Å were assigned as Co3O4 face-centred cubic phase (PDF 2001 d values (Å):
1.3996, 1.3959 and 2.7280). Similarly, the XRD pattern of MS-Co(40.0)Cal catalyst
(calcined at 750 °C), the main peaks of Co3O4 face-centred cubic phase appeared at
119
2θ values of 36.883, 31.264 and 65.362° with d values of 2.4350, 2.8587 and 1.4266
Å, as in Table 6.2. The MS support cannot be identified in the diffractograsms,
because the high loading of Co catalyst (40 wt.%) resulting high concentration of Co
metal and subsequently over saturated the MS support. However, this result did not
represent the actual catalysts used in the synthesis of CNTs, because there are no
interaction occurred between the MS support and the Co metal.
Intensity (%)
Calcination: 750 °C / 5 h
Calcination: 450 °C / 17 h
5
10
20
30
40
2θ (°)
50
60
70
Figure 6.2: XRD patterns of MS-Co(40.0)Cal catalyst calcined at 450 °C / 17 h and
750 °C / 17 h. T: Co3O4 face-centred cubic.
120
Table 6.2: Phase, peak position (2θ) and d values (Å) in the X-ray diffractogram
patterns of MS-Co(40.0)Cal catalyst calcined at 450 °C / 17 h and 750 °C / 17 h.
Sample
Calcination
Phase
Structure
2θ (°)
36.882
31.267
450 °C / 17 h
65.356
Co(40.0)
36.883
Co3O4 face31.264
750 °C / 5 h
centred cubic
65.362
* dobs : d spacing values obtained from XRD analyses.
dref : d spacing values obtained from the reference.
Co3O4 facecentred cubic
6.3.2
dobs* (Å)
2.4350
2.8584
1.4267
2.4350
2.8587
1.4266
dref * (Å)
(PDF
2001)
2.4370
2.8580
1.4290
2.4370
2.8580
1.4290
XRD Analysis of Molecular Sieves Supported Ferrum Catalysts
The XRD analyses of the MS-Fe catalysts of 2.5 wt.% and 3.0wt.% loadings
and MS support are shown in Figure 6.3. The data are tabulated in Table 6.3. The
MS support and MS-Fe catalysts are found to be highly crystalline. The MS-Fe(2.5)
catalysts were detected as Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic phase. The peaks
were at 2θ values of 7.146, 10.127 and 23.992° or d values of 12.3595, 8.7273 and
3.7061 Å (PDF 2001 d values (Å): 12.2380, 8.6536 and 3.6899). The two main peaks
of the Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic phase showed similar position with the
NaSiAlO4 face-centred cubic phase of MS support, as observed in the diffractograms
(Figure 6.3). The integration of Fe catalyst to the MS support increased the
intensities of peaks at 2θ of 7.146 and 10.127°.
However, calcination at 450 °C had deteriorated the intensity of peak at 2θ of
23.960° even though the Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic phase was
maintained. The Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic phase in MS-Fe(2.5)Cal
catalyst were assigned at 2θ values of 7.116, 10.105 and 23.960° with d values of
12.4123, 8.7460 and 3.7110 Å.
121
MS-Fe(3.0)Cal
Intensity (%)
MS-Fe(3.0)
MS-Fe(2.5)Cal
MS-Fe(2.5)
MS
5 10
60
50
40
70
2θ (°)
Figure 6.3: XRD patterns of various MS-Fe catalysts together with MS support. ¡:
20
30
NaSiAlO4 face-centred cubic and Ø: Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic phase.
122
Table 6.3: Phases, peak position (2θ) and d values (Å) in the X-ray diffractogram
patterns of the MS support, MS-Fe(2.5) and MS-Fe(2.5)Cal catalysts.
Samples
Phase Structure
2θ (°)
7.100
MS
10.127
12.431
7.146
NaSiAlO4
10.127
face-centred cubic
12.394
MS–Fe(2.5)
7.146
Fe2.7Na2.0(Si12Al12O48)(H2O)14.8
10.127
cubic
23.992
7.116
NaSiAlO4
10.105
face-centred cubic
MS–
12.434
Fe(2.5)Cal
7.116
Fe2.7Na2.0(Si12Al12O48)(H2O)14.8
10.105
cubic
23.960
* dobs : d spacing values obtained from XRD analyses.
dref : d spacing values obtained from the reference.
NaSiAlO4
face-centred cubic
dobs* (Å)
12.4406
8.7276
7.1144
12.3595
8.7273
7.1356
12.3595
8.7273
3.7061
12.4123
8.7460
7.1126
12.4123
8.7460
3.7110
dref * (Å)
(PDF
2001)
12.3113
8.7054
7.1079
12.3113
8.7054
7.1079
12.2380
8.6536
3.6899
12.3113
8.7054
7.1079
12.2380
8.6536
3.6899
As observed in Figure 6.3, the increment of Fe catalyst loading did not differ
the
X-ray
diffraction
pattern
of
the
MS-Fe(3.0)
catalyst.
The
Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic and NaSiAlO4 face-centred cubic phase were
remained the same. The 2θ values of 7.159, 10.131 and 23.913° with d values of
12.3381, 8.7243 and 3.7181 Å corresponded to the Fe2.7Na2.0(Si12Al12O48)(H2O)14.8
cubic phase. The NaSiAlO4 face-centred cubic phase is sharing the same peaks as
Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic phase at 2θ positions of 7.159 and 10.131°. As
a result of the calcination, the MS-Co(3.0)Cal catalyst had slightly increased the
intensities of Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic peaks at 2θ of 7.097, 10.136 and
23.909° or d values of 12.4446, 8.7200 and 3.7188 Å (Tee et al., 2004a).
It is discovered that the Fe cations from ferrum acetate are replacing the Na
ions in the NaSiAlO4 structure resulting the Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic
phase without changing the structure of the alumino-silicate cages in MS. Therefore,
123
the peaks of these two phases appeared at the same position in the XRD patterns.
This interaction holds the Fe nanoparticles in place and increases the dispersion of
the catalyst particles.
For the purpose of XRD analyses, a MS supported Fe catalyst with Fe
loading of 30.0 wt.% was prepared and calcined at 450 °C / 17 h, 750 °C / 5 h and
1000 °C / 5 h. From the X-ray diffractograms, the MS-Fe(30.0)Cal which calcined at
450 °C and 750 °C were found to be amorphous and no phase can be detected,
therefore the calcination temperature was further increased to 1000 °C. The XRD
patterns of MS-Fe(30.0)Cal calcined at 750 and 1000 °C were shown in Figure 6.4.
The XRD pattern of MS-Fe(30.0)Cal catalyst (1000 °C) was highly crystalline and
Fe2O3 rhombohedral phase was detected at 2θ values of 33.193, 35.848 and 54.218°
or d values of 2.6968, 2.5029 and 1.6904 Å (PDF 2001 d values (Å): 2.6900, 2.5100
and 1.6900) as tabulated in Table 6.4. Thus, it is verified that the Fe species is Fe2O3
rhombohedral without interaction with the support.
124
Intensity (%)
Calcination: 1000 °C / 5 h
Calcination: 750 °C / 5 h
5
10
20
30
40
2θ (°)
50
60
70
Figure 6.4: XRD patterns of MS-Fe(30.0)Cal catalyst calcined at 750 °C and
1000
°C for 5 hours. Ø: Fe2O3 rhombohedral.
Table 6.4: Phase, peak position (2θ) and d values (Å) in the X-ray diffractogram
patterns of MS-Fe(30.0)Cal catalyst calcined at 750 and 1000 °C for 5 hours.
Sample
Fe(30.0)Cal
Calcination
Phase
Structure
2θ (°)
dobs* (Å)
1000 °C / 5 h
Fe2O3
rhombohedral
33.193
35.848
54.218
2.6968
2.5029
1.6904
dref * (Å)
(PDF
2001)
2.6900
2.5100
1.6900
125
6.3.3 XRD Analysis of Molecular Sieves Supported Cobalt/Ferrum Catalysts
The X-ray diffraction patterns of series of MS supported bimetallic catalysts
are shown in Figure 6.5. Table 6.5 showed the 2θ and d values of these
diffractograms. The incorporation of Co and Fe onto the MS support was achieved
because the peaks in the diffractograms of MS-Co/Fe catalysts increased
dramatically. All of the bimetallic catalysts possessed high degree of crystallinity by
showing sharp peaks in the diffractograms. In all cases, the MS support was still
detected as the NaSiAlO4 face-centred cubic phase.
In MS-Co/Fe(2.5) catalyst, the 2θ values at 7.114, 10.094 and 23.957° or d
values
of
12.4154,
8.7563
and
3.7114
Å
are
attributed
to
the
Co.333Na.333(AlSiO4)(H2O)2.92 cubic and Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic phase.
The PDF 2001 d values are shown in Table 6.3. The calcination at 450 °C did not
give any changes to the diffractogram of MS-Co/Fe(2.5)Cal catalyst.
The detection of the MS-Co/Fe(3.0) catalyst begun with the intense peaks at
2θ values of approximately 7.134, 10.067 and 23.976° or d values of 12.3816, 8.7794
and 3.7085 Å that were spotted as the Co.333Na.333(AlSiO4)(H2O)2.92 cubic and the
Fe2.7Na2.0(Si12Al12O48)(H2O)14.8
cubic
phase.
The
diffractograms
of
MS-Co/Fe(3.0)Cal was identical to the non-calcined sample, with no significant
changes observed.
The Co.333Na.333(AlSiO4)(H2O)2.92 cubic and Fe2.7Na2.0(Si12Al12O48)(H2O)14.8
cubic phase have analogous structure. For that reason they shared the same
diffraction pattern by showing intense peaks at the same positions. These two metal
interact with the NaSiAlO4 face-centred cubic phase in the same manner, in which
their ions substituted the Na ions in the support structure.
126
MS-Co/Fe(3.0)Cal
Intensity (%)
MS-Co/Fe(3.0)
MS-Co/Fe(2.5)Cal
MS-Co/Fe(2.5)
MS
5 10
20
30
40
2θ (°)
50
60
70
Figure 6.5: X-ray diffractogram patterns of MS support and series of MS-Co/Fe
catalysts. ¡: NaSiAlO4 face-centred cubic, T: Co.333Na.333(AlSiO4)(H2O)2.92 cubic
and Ø: Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic phase.
127
Table 6.5: Phases, peak position (2θ) and d values (Å) in the X-ray diffractogram
patterns of the MS support, MS-Co/Fe(2.5) and MS-Co/Fe(2.5)Cal catalysts.
Samples
Phase Structure
2θ (°)
7.100
MS
10.127
12.431
7.114
NaSiAlO4
10.094
face-centred cubic
12.433
7.114
MS–
Co.333Na.333(AlSiO4)(H2O)2.92
Co/Fe(2.5)
10.094
cubic
23.957
7.114
Fe2.7Na2.0(Si12Al12O48)(H2O)14.8
10.094
cubic
23.957
7.149
NaSiAlO4
10.143
face-centred cubic
12.424
7.149
MS–
Co.333Na.333(AlSiO4)(H2O)2.92
10.143
Co/Fe(2.5)Cal
cubic
23.995
7.149
Fe2.7Na2.0(Si12Al12O48)(H2O)14.8
10.143
cubic
23.995
* dobs : d spacing values obtained from XRD analyses.
dref : d spacing values obtained from the reference.
NaSiAlO4
face-centred cubic
6.3.4
dobs* (Å)
12.4406
8.7276
7.1144
12.4154
8.7563
7.1135
12.4154
8.7563
3.7114
12.4154
8.7563
3.7114
12.3548
8.7139
7.1184
12.3548
8.7139
3.7056
12.3548
8.7139
3.7056
dref * (Å)
(PDF
2001)
12.3113
8.7054
7.1079
12.3113
8.7054
7.1079
12.2670
8.6741
3.6986
12.2380
8.6536
3.6899
12.3113
8.7054
7.1079
12.2670
8.6741
3.6986
12.2380
8.6536
3.6899
Particle Size of Molecular Sieves Supported Catalysts
The particle size of the MS supported catalysts were measured by Scherrer
Equation (Equation 4.3) using full width at half maximum peak intensity (FWHM)
obtained from the XRD analyses. There is only a particle size (d) value obtained for
each catalyst due to the diffraction patterns of Co.333Na.333(AlSiO4)(H2O)2.92 cubic
and Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic phase in the bimetallic catalysts were
detected at identical peak position. The particle sizes (d) of MS supported bimetallic
catalysts were tabulated in Table 6.6, while the example of calculation was shown in
Appendix C.
128
Table 6.6: The particle size (d) of Co.333Na.333(AlSiO4)(H2O)2.92 cubic and
Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic phase in MS supported bimetallic catalysts.
Catalysts
Particle size, d (nm)
MS-Co/Fe(2.5)
75.74
MS-Co/Fe(2.5)Cal
74.18
MS-Co/Fe(3.0)
71.76
MS-Co/Fe(3.0)Cal
74.56
The particle sizes of the MS-Co/Fe catalysts are in the range of 71 – 76 nm.
The particle sizes are almost the same and are not affected by either the increment of
metal catalyst loading or the calcination process (Tee et al., 2004b). The Co and Fe
ions were clarified as the replacement ions of the Na ions in the NaSiAlO4 structure,
thus the particle sizes of the Co or Fe supported catalysts are in a narrow range.
The similar particle sizes of the bimetallic catalysts signify that the catalyst
particles are evenly dispersed on the MS support. The diameter of CNTs growing on
a metal particle mainly depends on the dispersion of the metal particles on the
support (Nagaraju et al., 2002).
6.3.5
Conclusion of XRD Analysis
The XRD patterns of the MS supported catalysts showed that the Co, Fe and
mixture of Co and Fe were successfully incorporated onto the MS support. The
diffractograms of MS support and all MS support catalyst showed sharp peaks
indicated high crystallinity properties of these materials. Essentially, the MS support
was detected as NaSiAlO4 face-centred cubic phase, meanwhile the Co catalysts
were identified as Co.333Na.333(AlSiO4)(H2O)2.92 cubic phase and the Fe catalysts
were assigned as Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic phase. In the bimetallic
catalysts, the Co and Fe catalysts showed the same phase as the individual Co or Fe
catalysts.
129
The X-ray diffractogram patterns of the MS supported catalysts have
comparable patterns independent of the catalyst loadings and the calcination at 450
°C.
It
is
because
the
Co.333Na.333(AlSiO4)(H2O)2.92
cubic
and
the
Fe2.7Na2.0(Si12Al12O48)(H2O)14.8 cubic phase have analogous structure with the
NaSiAlO4 face-centred cubic phase, which attributed to the Na ions exchanged by
the Co and Fe ions. The interaction between the MS support and metal catalysts are
strong as the metal ions were doped into the alumino-silicate structure of MS.
The particle sizes of the bimetallic catalysts are approximately 70 nm since
the entire bimetallic catalysts have same molecular structure. In all MS supported
catalysts, the dispersion of the Co and Fe catalysts on the MS supports is high and
contributes to better activity towards the synthesis of CNTs.
6.4
As-Grown Carbon Deposition Yield and Physical Appearance
The catalytic decomposition of C2H2 was carried out with C2H2 flow rate of
15 sccm at 700 °C for 30 min. The MS supported catalysts generated black colour
carbon deposits with distinctly different texture. The yields were deposited on the
MS supported catalyst beads and some of the catalysts were able to produce
carbonaceous powder within the container of the catalyst beads. From the textures of
the carbon deposits, we can get an idea regarding activity of a particular catalyst in
producing CNTs. In general, when the carbon deposit appeared spongy with
deposition of black fluffy powder, the catalyst was superior in CNT production. In
contrast, if the yield remained in beads form with only black settlement on its surface
without powder deposition, the activity was relatively low. The percent of carbon
deposition yield of the reaction using MS supported catalysts and the physical
appearances as well as the density of the as-synthesis carbon deposit were tabulated
in Table 6.7.
From Table 6.7, it is noticed that some of the carbon deposition yield (%)
showed negative values. According to Kathyayini et al. (2004), the metal oxides
130
support lost about 18 to 55 % of their initial weight under the flow of nitrogen at
700 °C. Thus, if the carbon deposit generated is less than the weight loss of the
support, the deposition yield will show negative value. The catalysts that have shown
positive carbon deposition yield (%) would generate higher amount of deposition.
Table 6.7: The MS supported catalysts and the carbon deposition yield (%), the
physical appearance and density of the as-synthesized carbon deposit.
Carbon
deposition
Physical appearances
Density
yield (%)
MS-Co(2.5)
-3.64
Not spongy
+
MS-Co(2.5)Cal
-23.44
Not spongy
+
MS-Co(3.0)
-5.61
Not spongy
+
MS-Co(3.0)Cal
-2.98
Not spongy, small amount of powder
+
MS-Fe(2.5)
4.59
Spongy, lots of powder
++
MS-Fe(2.5)Cal
9.09
Spongy, lots of powder
++
Slightly spongy, small amount powder
MS-Fe(3.0)
-7.63
++
MS-Fe(3.0)Cal
-6.14
Not spongy
+
MS-Co/Fe(2.5)
26.63
Very spongy, plenty of powder
+++
MS-Co/Fe(2.5)Cal
35.30
Extremely spongy, plenty of powder
+++
MS-Co/Fe(3.0)
27.15
Very spongy, plenty of powder
+++
MS-Co/Fe(3.0)Cal
46.37
Extremely spongy, plenty of powder
+++
* Density of carbon deposition yield from observations; +: low density, ++: moderate
density, +++: high density.
Catalysts
The MS-Co catalysts showed low density of carbon deposition, the texture of
the yield was not fluffy. The carbon deposition yields (%) of these catalysts are in
between -23.44 and -2.98 %, suggested the conversion of C2H2 to carbonaceous
materials was ineffective.
When the MS-Fe catalysts were used, the textures and density of carbon
depositions have improved with the carbon deposition yields varying from -7.63 to
9.09 %. The MS-Fe catalysts produced modest spongy yield and small amount of
carbonaceous powder.
The density and physical appearance of the products indicate that the
bimetallic catalysts are the best catalysts compared to the MS-Co and MS-Fe
131
catalysts. The carbon deposits grown over MS-Co/Fe catalysts were highly dense
with plenty of carbon powder obtained. The carbon deposition yield of
MS-Co/Fe(3.0)Cal catalyst are found to be as high as 46.37 %, whereas other Co/Fe
catalysts manage to generate around 30 % yield. The ability to produce carbon
deposit vary in the order of catalyst used as Co/Fe > Co > Fe.
6.5
Scanning Electron Microscopy (SEM) Analysis
6.5.1
SEM Analysis of Molecular Sieves Supported Cobalt Catalysts and the
CNTs Yield
Scanning Electron Microscopy (SEM) images of dried MS support is shown
in Figure 6.6. The support surface was flat and powdery with rough texture. After the
impregnation of cobalt acetate solution, the MS-Co(2.5)Cal catalyst surface was
completely covered by the irregular shaped Co particles, as seen in Figure 6.7 (a).
The average size of the Co particles was estimated as 2.5 µm. The example of
particle size calculation is shown in Appendix D.1.
Figure 6.6: SEM micrograph of dried molecular sieves (MS) support (1000×).
The MS-Co(2.5) catalyst has similar morphology because the calcination at
450 °C did not change the morphology of the non-calcined and calcined samples.
From Figure 6.7 (b), it is observed that MS-Co(3.0) catalyst showed similar
132
morphology as the 2.5 wt. % catalysts except for the higher density of the Co
particles. The higher loading of Co (3.0 wt.%) contributes to the high density of
catalyst particles on the catalyst surface. The Co particles possessed average size of
2.3 µm. The existance of Co catalyst particles on the support reveals that the support
had physically interacted with the Co particles by dispersing and holding the particle
on the surface (Tauster et al., 1978 and Yudanov et al., 1997).
(b)
(a)
Figure 6.7: SEM micrographs of MS supported Co catalysts: (a) MS-Co(2.5)Cal
(1000×) and (b) MS-Co(3.0) (1000×).
SEM micrographs of as-grown CNTs over different MS-Co catalysts are
shown in Figure 6.8 (a) – (f). The SEM analyses discovered that all of the Co
catalysts were active in producing CNTs with different density. In Figure 6.8 (a), it is
noticeable that the MS-Co(2.5) catalyst grown CNTs with adequate density. The asgrown CNTs are winding and some assembled to form bundles of CNTs. The CNTs
are seen to have grown from the catalyst particles. From the high magnification
micrograph (Figure 6.8 (b)), the average diameter of the as-grown CNTs over MSCo(2.5) catalyst is nearly 100 nm. The calculation of CNTs diameter of is shown in
Appendix D.2. The surface of the nanotubes is smooth without deposition of
amorphous carbon. A few coiled CNTs were also observed in the micrograph.
In Figure 6.8 (c), the growth of CNTs over MS-Co(2.5)Cal catalyst is
obvious. The density of the nanotubes was higher compared to the previous case; the
surface of the catalyst was covered by a thick layer of CNTs. The as-grown CNTs
have uniform diameter of approximately 71.4 nm. The CNTs grown in unsystematic
133
direction and assemble in bundles. In contrast, the MS-Co(3.0) catalyst only grown
CNTs with fair density as depicted in Figure 6.8 (d). The surface of the catalyst did
not fully covered by the nanotubes. The length of the nanotubes is extended to 6 µm.
(a)
(b)
Coiled
CNTs
(d)
(c)
(e)
Coiled
CNTs
(f)
Figure 6.8: SEM micrographs of as-grown CNTs over various MS supported Co
catalysts: (a) and (b) MS-Co(2.5) (5000× and 10000×), (c) MS-Co(2.5)Cal (5000×),
(d) MS-Co(3.0) (10000×), (e) and (f) MS-Co(3.0)Cal (10000×).
Interestingly, the growth of CNTs greatly improved when the MS-Co(3.0)Cal
catalyst was used, as illustrated in Figure 6.8 (e) and (f). The CNTs can be seen as
134
bundles with fluffy and spongy texture. The CNTs possess high quality as no
amorphous carbon or carbonaceous particle was observed in Figure 6.8 (f). The
average diameter was calculated as 71.4 nm. This catalyst was able to form very long
coiled CNTs as spotted by the circle in Figure 6.8 (e). According to Lu et al. (2004b),
the supported Co particles able to grow coiled CNTs through CCVD of C2H2.
Piedigrosso et al. (2000) also reported the growth of coiled CNTs over a Co-silica
catalyst. Here, we can draw a conclusion that the Co catalysts were accountable to
the growth of coiled CNT regardless to the types of support.
In general, the MS supported Co catalysts are active in synthesizing CNTs
with different effectiveness. The SEM analyses were parallel with the carbon
deposition yield (%) discussed in Section 6.3, which the MS-Co(3.0)Cal catalyst
showed the highest yield amongst the Co catalysts.
6.5.2
SEM Analysis of Molecular Sieves Supported Ferrum Catalysts and the
CNTs Yield
From the SEM micrographs of MS-Fe(3.0) (Figure 6.7 (a)), the Fe particles
with cubic shape can be observed and the particle size is about 1.8 µm. The
morphology show that the MS-Fe(3.0) catalyst is highly crystalline. The Fe particle
morphology agreed with the results of Liu, D.Y. et al. (2004). Further, Hernadi et al.
(1996a) reported that the Fe particles are the acetate salt crystal. However, the cubic
Fe particles collapsed after the calcination at 450 °C shown in Figure 6.9 (b). The
cubic Fe particles had shrinked and became unclear.
From the micrographs of carbonaceous powder produced by MS-Fe(2.5)
catalyst (Figure 6.10 (a)), the nanotubes grown in bundles with fluffy texture. The
density of as-grown CNTs is high. However, the CNTs grown on the surface of
catalyst beads showed different morphology, as illustrated in Figure 6.10 (b). The
nanotubes are seen grown from the catalyst particles discreetly (Ismail et al., 2004b).
The average diameter of the as-grown CNTs is 72.7 nm. The MS-Fe(2.5) catalyst
135
grown a lots of CNTs and part of the nanotubes fall off from the surface of the
catalyst forming the powder yield.
(b)
(a)
Figure 6.9: SEM micrographs of MS supported Fe catalysts: (a) MS-Fe(3.0) (1000×)
and (b) MS-Fe(3.0)Cal (1000×).
The ability of producing CNTs was improved by calcination, as observed in
the micrographs of CNTs over MS-Fe(2.5)Cal catalyst, i.e. Figure 6.10 (c) and (d).
The winding CNTs possessed uniform diameter of 65.5 nm. The purity of the CNTs
is high as no amorphous carbon was observed. The powder yield over the
MS-Fe(2.5)cal catalyst showed wavy configuration and assembled in bundles with
feathery texture as depicted in Figure 6.10 (d).
When the loading of catalyst was increased to 3.0 wt.%, the activity of the
MS-Fe(3.0) catalyst was also increased. The high density CNT bundles grown over
MS-Fe(3.0) catalyst were shown in Figure 6.10 (e). The nanotubes are twisted and
high in purity; the average diameter is approximately 70.8 nm. Interestingly, in the
micrograph of powder deposition (Figure 6.8 (f)), a newly discovered carbon
nanomaterial was observed grown in between the bundles of nanotubes, it is called
carbon nanowalls (CNWs). The morphology of CNTs grown over MS-Fe(3.0)Cal
catalyst was similar to the MS-Fe(3.0) catalyst as seen in Figure 6.10 (g) and (h).
This indicates that calcination does not improve the activity of the catalyst. The
CNTs grown over this catalyst are rather thick with diameter of about 90.9 nm.
136
(a)
(b)
(c)
(d)
Wavy
CNTs
(e)
(f)
Carbon
nanowalls
(CNWs)
(g)
(h)
Figure 6.10: SEM micrographs of as-grown CNTs over series of MS-Fe catalysts:
(a) and (b) MS-Fe(2.5) (10000×), (c) and (d) MS-Fe(2.5)Cal (10000×), (e) and (f)
MS-Fe(3.0) (10000×), (g) and (h) MS-Fe(3.0)Cal (5000× and 10000×).
137
The morphologies of the as-grown CNTs are in agreement with the carbon
deposition yield and physical appearance discussed previously. The SEM
micrographs reflected that MS supported Fe catalysts have higher efficiency in
producing CNTs compared to the Co catalysts, where the MS-Fe(2.5)Cal catalyst
showed highest activity amongst the Fe catalysts. The MS-Fe catalysts do not
produce coiled CNTs which in agreement with Hernadi et al. (1996a) results.
6.5.3
SEM Analysis of Molecular Sieves Supported Cobalt/Ferrum Catalysts
and the CNTs Yield
The surface of MS-Co/Fe(3.0) catalyst was covered by a layer of rough
particles as observed in Figure 6.11 (a). There was also a few discreet particles
dispersed on the catalyst surface. The loading of Co and Fe in this catalyst is
1.5 wt.% respectively, which is lower than the individual Co or Fe catalysts loadings
of 2.5 or 3.0 wt.%. Therefore, lower concentration of catalyst impregnated on the
surface of support resulting lesser catalyst particles. The morphology of the MSCo/Fe(3.0)Cal catalyst was flatter compared to the non-calcined sample, due to the
shrinkage of catalyst particles.
(a)
(b)
Figure 6.11: SEM micrographs of MS supported Fe catalysts: (a) MS-Fe(3.0)
(1000×) and (b) MS-Fe(3.0)Cal (1000×).
138
According to the high carbon deposition yield (26.63 – 46.37 wt.%) and good
physical appearance, the bimetallic catalysts are likely to produce the best yield of
CNTs. CNTs grown on the MS-Co/Fe(2.5) catalyst showed high density as observed
in Figure 6.12 (a). The snaky CNTs achieved satisfactory length of about 5 µm with
its surface free from amorphous carbon. Interestingly, the nanotubes are observed to
be attached to the catalyst particles as in Figure 6.12 (b), which a coiled CNT is seen
to be bridging a CNT bundle and another catalyst particle. The coiled CNT has
different coil diameter and coil pitch, the tip was a loop wire of nanotube and the
bottom showed wavy configuration. We proposed that the Fe catalyst enhanced the
yield of CNTs and the Co catalyst responsible for the present of the coiled CNT.
As a result of calcination, the activity of MS-Co/Fe(2.5)Cal catalyst had
improved. This catalyst not only grew CNTs bundles but also arrays of CNTs as
illustrated in Figure 6.12 (c). The quasi-aligned CNTs was curved and showed ‘S’
shape. This image was identical to the micrographs of zeolite supported Co/Fe
catalyst in Mukhopadhyay and Mathur (2003) work. The nanotubes are fine and
having uniform diameter. An interesting occurrence was observed in Figure 6.12 (d),
where the CNTs grown in between two catalyst particles. The two catalyst particles
were close to each other and the growing nanotubes was linked and forming straight
CNTs.
The activity of the MS-Co/Fe(3.0) catalyst was equivalent with the MSCo/Fe(2.5) catalyst as it produced bundles of CNTs as seen in Figure 6.12 (e). The
diameter of the nanotubes was consistent with average of 61.9 nm. From the
micrograph of the powdery yield (Figure 6.12 (f)), it is discovered that this catalyst
also produce small amount of CNFs. The CNFs were aligned but showed
heterogeneous diameter.
139
(a)
(b)
Loop wire
Wavy
nanotubes
(c)
(d)
CNT bridges
CNT
bundles
(e)
(f)
CNFs
(g)
(h)
Figure 6.12: SEM micrographs of CNTs grown over the bimetallic catalysts: (a) and
(b) MS-Co/Fe(2.5) (10000×), (c) and (d) MS-Co/Fe(2.5)Cal (5000× and 2000×), (e)
and (f) MS-Co/Fe(3.0) (10000× and 300×), (g) and (h) MS-Co/Fe(3.0)Cal (5000×).
140
The MS-Co/Fe(3.0)Cal catalyst grew very high density of CNTs as observed
in Figure 6.12 (g). The CNT yield was thick and they twisted to form bundles. While,
Figure 6.12 (h) showed the packing of the CNTs with furry texture and high density.
The micrographs of CNTs over this catalyst also showed bridging CNTs as in Figure
6.12 (d), but the micrographs was not shown here. The morphology of the CNTs
yield showed that the MS-Co/Fe(3.0)Cal catalyst has highest activity among the
bimetallic catalysts.
6.5.4
Conclusion of SEM Analysis
Collectively, the activity of MS supported catalysts in synthesizing CNTs are
measurable and the MS supported bimetallic catalysts showed the best performance
amongst the MS supported catalysts. A variety of CNT configurations were observed
from the micrographs of as-grown CNTs over MS supported catalysts, i.e. bundles,
arrays, bridge, coil etc. The MS-Co(3.0)Cal, MS-Fe(2.5)Cal and MS-Co/Fe(3.0)Cal
are amongst the best catalysts.
According to the results from XRD and SEM analyses, the interaction
between the MS support and the catalyst nanoparticles are classified as strong metalsupport interaction (SMSI) as discussed by Vander Wal et al. (2001). This strong
interaction arise from the metal cations of the support donating the partial negative
charge to the supported metal particles. This strong interaction can enhance the
catalytic activity by strengthening back donation of electron density into the
antibonding orbitals of the hydrocarbon molecules and subsequently weakens the
bonding within the molecules results in the dissociation of the hydrocarbon
molecules (Davis and Klabunde, 1982). According to Ando et al. (2004), when the
metal-support interaction is strong, the CNTs tend to grow up with the catalyst
particle rooted at its base (base growth model).
141
6.6
Field-Emission Scanning Electron Microscopy (FE-SEM) Analysis
The as-grown CNTs over MS-Co/Fe(3.0)Cal was examined by FieldEmission Scanning Electron Microscope, the micrographs were shown in Figure 6.13
(a) and (b). At low resolution (Figure 6.13 (a)), the CNTs were seen assembling in
bundles, some of the nanotubes have length up to 6µm. The nanotubes were
entangled and sinuous. There is no by product deposited on the CNTs, designated
that as-grown CNTs is pure without any purification.
Further, from the high resolution micrograph, the growth of CNTs is verified.
The hollow structure of the CNTs can be clearly observed, the wall of the nanotubes
is smooth which proved the high degree of graphitization. The outer diameter of the
CNTs is calculated as 14.29 nm. There is no catalyst particle found on the tip of
tubes, proved that the growth mechanism followed the base growth model.
(a)
(b)
Figure 6.13: FE-SEM micrographs of as-grown CNTs over MS-Co/Fe(3.0)Cal
catalyst, (a) 60000× magnification and (b) 250000× magnification.
142
6.7
Transmission Electron Microscopy (TEM) Analysis
The transmission electron microscopy (TEM) analysis is used to convey the
graphitization quality, defection, purity and wall structure of CNT yield. The TEM
micrographs facilitate the classification of SWNT and MWNT. Figure 6.14 (a) and
(b) presented the TEM micrographs of CNTs grown over MS-Co(3.0)Cal catalyst. In
figure 6.14 (a), the CNT has straight and turbostratic walls, it consists of 18 graphene
layers. No defect was observed in the wall structure, except the most external layer
was jagged and curved; reveal a high degree of graphitization. The outer diameter of
the nanotube was 33.55 nm, while the inner diameter was approximately 13.44 nm.
Further, another micrograph (Figure 6.14 (b)) showed that some by products of
decomposition attached on the surface of the CNT, the substance was amorphous as
the layers of the atoms are randomly displaced and it is kind of amorphous carbon.
(a)
(b)
Multi-layers
graphene wall
Amorphous
carbon
Turbostratic
nanotubes
Figure 6.14: TEM micrographs of the as-grown MWNTs over MS-Co(3.0)Cal
catalyst, (a) and (b) scale bar: 10 nm.
The TEM micrographs of CNTs grown over MS supported Fe catalysts were
presented in Figure 6.15 (a) - (d). As observed from Figure 6.15 (a), the MWNT has
lower quality because the walls were wavy and not parallel. The walls of the CNTs
were not continuous and contain some defects. A layer of amorphous carbon was
wrapping the MWNT. However, in Figure 6.15 (b) a better quality MWNT was
spotted. This CNT was constructed by 21 graphene layers of turbostratic
143
arrangement which were straight, parallel and free from defect. Adjacent to this
nanotube, there were a big mass of amorphous carbon.
(a)
(b)
Defects
Turbostratic
graphene
layers
Amorphous
carbon
Amorphous
carbon
. (c)
(d)
End cap of
MWNT
Figure 6.15: TEM micrographs of as-grown MWNTs from: (a) and (b)
MS-Fe(2.5)Cal catalyst, (c) and (d) MS-Fe(3.0) catalyst (for all images scale bar:
10nm).
From Figure 6.15 (c), the CNTs grown over MS-Fe(3.0) catalyst was
observed as a crooked MWNT, its walls were curved and not parallel. This MWNT
has similar quality as the CNT observed in Figure 6.15 (a) but with fewer layers of
walls. There was an internal compartment entrapped inside the longer tube, the end
cap of the short nanotubes can be clearly seen. In a close up image (Figure 6.15(d)),
144
the walls of the MWNT are uneven, suggested that the degree of graphitization is
low.
The TEM images of CNTs grown over MS-Co/Fe(2.5) catalyst were
presented in Figure 6.16 (a) – (c). From the low resolution image (Figure 6.16 (a)),
apparently the as-grown CNTs were high in quality and density. The CNTs were
observed elongated out from a catalyst particle and grew up to several microns. A
few open end of the nanotubes were detected, implied that the MS-Co/Fe(2.5)
catalyst grew CNTs in base growth mechanism.
(b)
(a)
Amorphous
carbon
Open ends of
nanotubes
Catalyst particle
(c)
Cross section
view
Figure 6.16: TEM micrographs of as-grown MWNTs from MS-Co/Fe(2.5) catalyst,
(a) scale bar: 100 nm, (b) scale bar: 10 nm and (c) scale bar: 10 nm.
145
In Figure 6.16 (b), the yield from the MS-Co/Fe(2.5) catalyst were assured as
the MWNT. The MWNT have more than 15 layers of graphene walls and same
layers in each side of the hollow structure. The walls are parallel and well-arranged,
no defect was identified. The inner and outer diameters of the MWNT measured
from the micrographs were 5.31 nm and 25.63 nm respectively. The cross section
view of the MWNT can be seen in Figure 6.16 (c), the layers of the nanotube wall
are concentric and well-arranged.
The TEM micrographs of MS-Co/Fe(3.0)Cal catalyst grown MWNTs are
shown in Figure 6.17 (a) and (b). Figure 6.17 (a) attested that the as-grown MWNT
has superb characteristic. The purity and the graphitization of the MWNT are very
high. The carbon building units arranged in well order to produce extremely
turbostratic and defect-free graphene layers. The inner diameter is 3.57 nm whereas
the outer diameter is 11.43 nm. The distance between layers is measured as 0.33 nm,
which is in agreement with inter-layer distance of graphite. In Figure 6.17 (b), it is
noticeable that the MWNT were wrapped by a layer of amorphous carbon, but the
MWNT still possessing good quality. The TEM micrographs revealed that the MS
supported catalysts are capable in producing MWNT with different quality and
quantity.
(a)
(b)
Amorphous
carbon
Figure 6.17: TEM micrographs of as-synthesized MWNT over MS-Co/Fe(3.0)Cal
catalyst, both of the micrographs have scale bar of 5 nm.
146
6.8
Energy Dispersive X-ray (EDX) Analysis
The energy dispersive X-ray (EDX) analysis was utilised to quantify the
surface compositions of the catalysts. The EDX analysis data were shown in Table
6.8. It is known that the molecular sieves (MS) are metal alumino-silicates, so it is
reasonable that the main component in the MS support are oxygen (O), aluminium
(Al) and silicon (Si). In addition to the main elements, there are traces amount of
natrium (Na), magnesium (Mg), chlorine (Cl) and kalium (K). These elements were
collectively group as E in Table 6.6, because their existences are not significant. The
MS support is free from cobalt (Co) or ferrum (Fe) before the impregnation step.
The impregnation of Co onto the MS support was achieved as the Co element
was detected in all MS-Co catalysts in the range of 4.81 – 14.44 wt.%. It is observed
that the calcination at 450 °C reduced the amount of Co elements in both of the
MS-Co/Fe(2.5)Cal and MS-Co/Fe(3.0)Cal catalysts. This is owing to the high
temperature during the calcination facilitating the surface Co to be incorporated into
the structure of MS, thus reducing the surface Co. The incorporation of Fe onto the
MS support were demonstrated by the detection of Fe element which range from
6.80 – 18.32 wt.%. The amount of Fe element did not show an exact trend. In the
case of bimetallic catalysts, the amount of Fe element is higher than the Co element.
The composition of Co and Fe were reduced by calcination process.
The MS supported catalysts were effective in converting C2H2 into CNTs as
the carbon (C) content of the catalysts varied from 24.84 to 61.84 wt.%. The MS-Co
catalysts generated modest quantity of carbon as a reflection of the lower activity in
the synthesis of CNTs. Meanwhile the Fe catalysts produced higher carbon content,
as in Table 6.8. This is concurring to the carbon deposition yield (%) and SEM
micrographs.
Throughout the characterisations, the MS supported bimetallic catalysts were
recognized as the best catalyst as these catalysts produced high graphitization and
purity. The MS-Co/Fe(2.5) and MS-Co/Fe(2.5)Cal catalysts gave the highest carbon
147
content. It might due to the higher amount of amorphous carbon in the samples as
shown in TEM micrographs.
From the C contents, we can draw a conclusion that the activity of the
bimetallic catalyst is the highest, followed by MS-Fe catalysts and the MS-Co
catalysts. The carbon contents are in agreement with the physical appearance
observation (Section 6.4).
Table 6.8: EDX analysis data of surface compositions of various MS supported
catalysts before and after decomposition of C2H2.
Surface Elements (wt.%)
O
Al
Si
Co
Fe
E#
8 36.50 16.57 23.64
MS
23.29
8 33.90 14.01 19.01 12.47
20.61
MS–Co(2.5)
9 15.53 12.95 20.17 9.12
17.18
8 29.00 17.92 27.37 4.81
20.90
MS–Co(2.5)Cal
9 22.73 11.63 16.24 5.97
15.38
8 34.82 13.08 18.08 14.44
19.58
MS–Co(3.0)
9 14.84 13.73 21.33 7.29
17.97
8 25.98 15.46 24.15 12.44
21.97
MS–Co(3.0)Cal
9 14.68 12.17 17.77 8.76
14.27
8 33.18 12.52 18.55
15.35 20.40
MS–Fe(2.5)
9 16.14 11.79 18.65
7.09 15.90
8 23.05 14.52 25.63
18.32 18.48
MS–Fe(2.5)Cal
9
9.15
9.36 14.37
8.07 11.08
8 24.54 16.65 27.19
12.89 18.73
MS–Fe(3.0)
9 13.03 10.51 17.05
8.54 18.18
8 27.56 16.75 26.48
6.80 22.41
MS–Fe(3.0)Cal
9
8.63
8.75 13.54
12.41 11.57
8 24.88 14.06 22.20 4.81 13.21 20.84
MS–Co/Fe(2.5)
9
6.18
6.83 10.41 2.66
4.76
7.32
8 22.98 17.20 27.70 5.11
6.14 20.87
MS–
Co/Fe(2.5)Cal
9 10.36 8.43 12.89 2.36
1.57
9.60
8 24.72 16.23 26.22 5.42
7.64 19.77
MS–Co/Fe(3.0)
9 14.64 10.12 13.99 2.14
0.84 10.60
8 21.52 16.74 27.73 4.20
6.46 23.35
MS–
Co/Fe(3.0)Cal
9 12.83 10.21 14.29 2.08
1.23 10.86
* Reaction (R): decomposition of C2H2 at 700 °C for synthesis of CNTs.
#
E: Elements included Na, Mg, Cl and K.
8: Before decomposition; 9: after decomposition.
Samples
R*
C
25.05
28.05
24.84
32.35
30.43
47.97
32.69
45.10
61.84
54.79
47.67
48.50
CHAPTER 7
CONCLUSION
7.1
Overall Conclusion
In this research, a high performance Catalytic Chemical Vapour Deposition
(CCVD) system was designed and built. The CCVD system consists of high
temperature tube furnace, fixed bed flow reactor and gas sources. The system is
useful and suitable for the synthesis of carbon nanotubes (CNTs).
All supported catalysts were successfully prepared by impregnation method,
which the alumina (Al2O3) and MS supported catalysts were prepared by the soaking
technique, whereas the anodic aluminium oxide (AAO) template supported catalysts
were prepared by dip coating technique. The cobalt (Co) and ferrum (Fe) metal were
effectively incorporated into the supports to form single or bimetallic supported
catalysts as indicated by the colour changes of the supported catalysts.
The X-Ray Diffraction (XRD) characterization has successfully identified the
crystallographic phases present in the Al2O3 and MS supported catalysts. However,
the XRD analysis was inappropriate for AAO supported catalysts as the materials are
amorphous. From the XRD pattern, the Al2O3 support was observed as Al2O3
tetragonal phase. In the Al2O3 supported catalysts, the presence of Co3O4 facecentred cubic phase and Fe2O3 rhombohedral phase were detected in the Co catalysts
and Fe catalysts respectively, while the bimetallic catalysts consist both of the active
149
catalyst phases. Meanwhile, the MS support was observed as NaSiAlO4 face-centred
cubic phase from the X-ray diffractogram. Whereas the MS-Co catalysts were
detected as Co.333Na.333(AlSiO4)(H2O)2.92 cubic phase and the MS-Fe catalysts
existed in Fe2.7Na2.0(Si12Al12O48) (H2O)14.8 cubic phase. The XRD patterns showed
that the bimetallic catalysts supported on Al2O3 and MS showed best diffraction
patterns.
In Al-Co/Fe catalysts, the particle sizes of Co3O4 face-centred cubic phase
were in the range of 28.26 – 38.64 nm and the particle sizes of Fe2O3 monoclinic
phase were varied form 22.35 to 29.89 nm. Nevertheless, The MS-Co/Fe catalysts
have only one particle size value for both of the active catalyst phases, i.e.
Co.333Na.333(AlSiO4)(H2O)2.92 cubic phase and Fe2.7Na2.0(Si12Al12O48) (H2O)14.8 cubic
phase, which are approximately 70 nm.
The carbon precursor chosen are pure acetylene (C2H2) and the catalytic
decomposition of C2H2 was carried out at 700 °C (ramping rate 5 °C/min) for 30 min
under the flow of C2H2/N2 at 15/120 sscm. The combination of precursor, catalysts
and parameters in this research was superb and producing excellent CNTs yield.
The Scanning Electron Microscopy (SEM), Field Emission-Scanning
Electron Microscopy (FE-SEM) and Transmission Electron Microscopy (TEM)
analyses gave comprehensible images of the as-grown CNTs over series of catalysts.
The physical appearance of the catalysts and CNTs produced provided brief
information about the samples.
From the electron microscopy analyses, the Al-Co catalysts are active in the
formation of CNTs with different efficiency except the Al-Co(2.5) catalyst. Amongst
the Al2O3 catalysts, the Al-Co(3.0), Al-Fe(3.0)Cal and Al-Co/Fe(3.0)Cal are the best
catalysts in terms of growing high quality and purity CNTs. The Al2O3 supported
catalysts produced MWNTs but did not favour the growth of coiled CNTs. The
quality and quantity of the as-grown CNTs were summarized in Table 7.1. All of the
AAO supported catalysts are able to grow CNTs in different configuration and
150
density. The AAO-Co(1.0)Cal, AAO-Fe(1.0) as well as AAO-Co/Fe(1.0)Cal are
more effective catalysts in synthesing CNTs. The details of the CNTs yield are
tabulated in Table 7.1.
Table 7.1: Comparison of purity, density and configurations of as-grown CNTs and
CNFs over various supported catalysts.
Configurations of CNTs #
Bundles Alignment Coil
Al-Co(2.5)
9
Al-Co(2.5)Cal
1
1
9
Al-Co(3.0)
4
4
9
Al-Co(3.0)Cal
2
2
9
Al-Fe(2.5)
3
3
9
Al-Fe(2.5)Cal
2
3
9
Al-Fe(3.0)
2
1
9
Al-Fe(3.0)Cal
4
4
9
Al-Co/Fe(2.5)
3
2
9
Al-Co/Fe(2.5)Cal
3
3
9
9
Al-Co/Fe(3.0)
4
4
9
9
Al-Co/Fe(3.0)Cal
5
5
9
AAO-Co(0.5)
1
2
9
AAO-Co(0.5)Cal
2
3
9
9
AAO-Co(1.0)
3
3
9
9
AAO-Co(1.0)Cal
4
4
9
AAO-Fe(0.5)
3
3
9
AAO-Fe(0.5)Cal
4
4
9
AAO-Fe(1.0)
4
4
9
AAO-Fe(1.0)Cal
3
2
9
9
AAO-Co/Fe(0.5)
4
4
9
9
AAO-Co/Fe(0.5)Cal
4
4
9
AAO-Co/Fe(1.0)
3
4
9
9
AAO-Co/Fe(1.0)Cal
5
5
9
9
MS-Co(2.5)
3
4
9
9
MS-Co(2.5)Cal
2
3
9
9
MS-Co(3.0)
3
2
9
9
MS-Co(3.0)Cal
5
5
9
MS-Fe(2.5)
3
4
9
MS-Fe(2.5)Cal
5
5
9
MS-Fe(3.0)
4
4
9
MS-Fe(3.0)Cal
4
4
9
9
MS-Co/Fe(2.5)
4
4
9
9
MS-Co/Fe(2.5)Cal
4
5
9
9
9
MS-Co/Fe(3.0)
3
4
9
9
MS-Co/Fe(3.0)Cal
5
5
* Relative rating of purity and density: 1 – lowest and 5 – highest.
#
9 - Existance of CNT bundles, alignment and coils.
Catalysts
Purity*
Density*
CNFs
-
9
9
9
151
From the quality and density of as-grown CNTs, all of the MS supported
catalysts showed very outstanding performance, in which the MS-Co(3.0)Cal,
MS-Fe(2.5)Cal and MS-Co/Fe(3.0)Cal were identified as the best catalysts. The
descriptions of the MWNTs are tabulated in Table 7.1. Referring to Table 7.1,
despite to the types of support, the ability to produce CNTs was vary in the order of
Co/Fe > Co > Fe catalysts. The Co catalysts are effective in forming coiled CNTs
whereas the bimetallic catalysts are accountable for the alignment of CNTs. In
general, the Co catalysts produced good quality nanotubes, while the Fe catalysts are
efficient of C2H2 decomposition in giving higher carbon deposition yield, thus, the
bimetallic catalysts showed the best performance in producing high quality and
quantity CNTs.
The Energy Dispersive X-Ray (EDX) Analysis provided useful information
about the composition of the supported catalysts and carbon deposits. The Co, Fe and
Co/Fe metals were successfully incorporated into the supports with sufficient amount
to initiate the growth of CNTs. The Al2O3 supported catalysts produced carbon
contents of 25.00 – 73.76 wt.%, the AAO supported catalysts produced 18.73 – 86.51
wt.% carbon as well as the MS supported catalysts generated 25.05 – 61.84 wt.% of
carbon.
Collectively, the performance of a catalyst in growing CNTs mainly depends
on the types of metal catalysts and quantity of the catalysts. From the experimental
data, we found that the bimetallic catalysts produced great quality and quantity CNTs
regardless to the types of support. The higher loading catalysts also enhanced the
growth of CNTs. The activity of CNTs production seems not strongly affected by the
types of support, it is noticed that the supports merely play a role in providing a
platform for the dispersion of catalyst particles and growth of CNTs. In addition, the
calcination treatment was found to be less significant in improving the activity of the
catalysts towards the synthesis of CNTs.
Through the observation of micrographs, the growth mechanism of the CNTs
can be suggested. The Al2O3 support interacts with the metal catalysts through Lewis
base sites (weak metal-support interaction), whereas the MS supported catalysts have
152
strong metal-support interaction (SMSI). Therefore, it can be concluded that the
Al2O3 supported catalysts grow CNTs through the tip growth mechanism while the
MS supported catalysts follow the base growth mechanism.
7.2
Future Works
The research on the synthesis of CNTs is still new in Malaysia, more works
need to be carried out in order to improve and control the synthesis method for the
production of CNTs with required structures. Some of the recommendations are:
1. The Catalytic Chemical Vapour Deposition (CCVD) system can be modified to
increase the production quantity. Replace the fixed bed flow reactor with the
floating catalyst reactor and investigate the CNTs synthesis.
2. Diversify the types of supports in preparation of supported catalysts as the
support can affect the metal-support interaction in the system. Supports such as
MgO and CaCO3 are potential catalyst supports.
3. Introducing transition metals such as Ni, V and Mo as catalysts in the preparation
of supported catalysts. The bimetallic and trimetallic combinations of these
metals need to be studied. The mixture of Co or Fe with these metals are an
interesting study to be investigated.
4. Utilise other characterization methods to study the supported catalysts and carbon
deposition yield comprehensively. The active catalyst sites in the supported
catalysts can be identified accurately using X-Ray Photoelectron Spectroscopy
(XPS) and the as-grown MWNTs can be verified through Raman Scattering
analysis.
5. Other compounds such as methane, benzene, alcohol and carbon dioxide are
potential carbon precursors for the synthesis of CNTs. The parameters of
153
catalytic decomposition should be customized according to the properties of the
precursor.
REFERENCES
Ando, Y., Zhao, X., Sugai, T. and Kumar, M. (2004). Growing Carbon Nanotubes.
Materials Today. Review Feature: 22 – 29.
Andrews, R., Jacques, D., Rao, A.M., Rantell, T., Derbyshire, F. Chen, Y., Chen, J.
and Haddon, R.C. (1999). Nanotube composite carbon fibers. Applied Physics
Letters. 75: 1329.
Andrews, R., Jacques, D., Dickey, E.C. and Qian, D. (2003). Synthesis and Growth
Mechanisms of Multiwalled Carbon Nanotubes. Kentucky/University of
Kentucky. unpublished. (http://www.caer.uky.edu).
Baker, R. T. K., Chludzinski Jr., J. J. and Lund, C. R. F. (1987). Further Studies of
the Formation of Filamentous Carbon from the Interaction of Supported Iron
Particles with Acetylene. Carbon. 25: 295 – 303.
Buang, N.A., Aziz, M., Sanip, S., Tee, J.C., Abidin, Z. and Ismail, A.F. (2005).
Effect of Pretreatment of Synthetic and Natural Carbons as Starting Materials for
Carbon Nanotubes. Journal of Metastable and Nanocrystalline Materials. 23:
285 – 288.
Chambers, A., Park, C. Baker, R.T.K. and Rodriguez, N.M. (1998). Hydrogen
Storage in Graphite Nanofibers. The Journal of Physical Chemistry B. 102: 4253
– 4256.
Chen, P., Wu, X. Lin, J. and Tan, K.L. (1999). High Hydrogen Uptake by AlkaliDoped Carbon Nanotubes under Ambient Pressure and Moderate Temperatures.
Science. 285: 91 – 93.
155
Ciambelli, P., Sannino, D., Sarno, M., Fonseca, A. and Nagy, J.B. (2005). Selective
Formation of Carbon Nanotubes over Co-modified Beta Zeolite by CCVD.
Carbon. 43: 631 – 640.
Colomer, J.F., Bister, G., Willems, I., Konya, Z., Fonseca, A., Van Tendeloo, G. and
Nagy, J.B. (1999). Synthesis of Single-Wall Carbon Nanotubes by Catalytic
Decomposition of Hydrocarbons. Chemical Communications. 14: 1343 – 1344.
Colomer, J.F., Stephan, C., Lefrant, S., Van Tendeloo, G., Willems, I., Konya, Z.,
Fonseca, A., Laurent, Ch. And Nagy, J.B. (2000). Large-Scale Synthesis of
Single-Wall Carbon Nanotubes by Catalytic Chemical Vapor Deposition
(CCVD) Method. Chemical Physics Letters. 317: 83 – 89.
Daenen, M., de Fouw, R.D., Hamers, B., Janssen, P.G.A., Schouteden, K. and Veld,
M.A.J. (2003). The Wondrous World of Carbon Nanotubes. Denmark/Eindhoven
University of Technology. unpublished. (http://students.chem.tue.nl/ifp03)
Dai, H., Rinzler, A.G., Nikolaev, P., Thess, A., Colbert, D.T. and Smalley, R.E.
(1996). Single-wall nanotubes produced by metal-catalyzed disproportionation
of carbon monoxide. Chemical Physics Letters. 260: 471.
Davis, S.C. and Klabunde K.J. (1982). Unsupported Small Metal Particles:
Preparation, Reactivity, and Characterization. Chemical Reviews. 82(2): 153 –
208.
Dillon, A.C., Jones, K.M., Bekkedahl, T.A., Kiang, C.H., Bethune, D.S. and Heben,
M.J. (1997). Storage of Hydrogen in Single-Walled Carbon Nanotubes. Nature.
386: 377 – 379.
Dresselhaus, M.S., Dresselhaus, G. and Eklund, P.C. (1995). Science of Fullerenes
and Carbon Nanotubes. California: Academic Press Inc.
Dresselhaus, M.S. and Avouris, P. (2001). Introduction to Carbon Materials
Research. In: Dresselhaus, M.S., Dresselhaus, G. and Avouris, P. Carbon
Nanotubes: Synthesis, Structure, Properties and Applications. Berlin, Germany:
Springer-Verlag. 1 – 9.
156
Flahaut, E., Laurent, Ch. and Peigney, A. (2005). Catalytic CVD Synthesis of
Double and Triple-Walled Carbon Nanotubes by the Control of the Catalyst
Preparation. Carbon. 43: 375 – 383.
Fonseca, A., Hernadi, K., Nagy, J.B., Bernaerts, D. and Lucas A.A. (1996).
Optimization of Catalytic Production and Purification of Buckytubes. Journal of
Molecular Catalysis A: Chemical. 107: 159 – 168.
Fonseca, A., Hernadi, K., Piedigrosso, P., Colomer, J.F., Mukhopadhyay, K.,
Doome, R., Lazarescu, S., Biro, L.P., Lambin, Ph., Thiry, P.A., Bernaerts, D.
and Nagy, J.B. (1998). Synthesis of Single- and Multi-Wall Carbon Nanotubes
Over Supported Catalysts. Applied Physics A: Materials Science and
Processing. 67: 11 – 12.
Fuertes, A.B. (2002). Template Synthesis of Carbon Nanotubules by Vapour
Deposition Polymerization. Letters to the edito /Carbon. 40: 1600 – 1602.
Hamada, N., Sawada, S. I. and Oshiyama, A. (1992). New One-Dimensional
Conductors: Graphitic Microtubules. Physical Review Letters. 68(10): 1579 –
1581.
Hernadi, K., Fonseca, A., Nagy, J.B., Bernaerts, D., Fudala, A. and Lucas, A.A.
(1996a). Fe-Catalyzed Carbon Nanotube Formation. Carbon. 34(10): 1249 –
1257.
Hernadi, K., Fonseca, A., Nagy, J.B., Bernaerts, D., Fudala, A. and Lucas, A.A.
(1996b). Catalytic Synthesis of Carbon Nanotubes Using Zeolite Support.
Zeolites. 17(5 – 6): 416 – 412.
Iijima, S. (1991). Helical Microtubules of Grafitic Carbon. Nature. 354(6348): 56 –
58.
Inoue, S., Nakajima, T. and Kikuchi, Y. (2005). Synthesis of Single-Wall Carbon
Nanotubes from Alcohol Using Fe/Co, Mo/Co, Rh/Pd Catalysts. Chemical
Physics Letters. 406: 184 – 187.
157
Ismail, A.F., Tee, J.C., Buang, N.A., and Sanip, S.M. (2004a). Effect of Catalyst
Supports on the Growth of CNTs. 3rd Annual Conference of the Asia Pacific
Nanotechnology Forum and 2nd Shanghai International Nanotechnology
Cooperation Synposium. 9 – 12 December 2004. Sydney, Australia: APNF, Inc,
150 – 155.
Ismail, A.F., Tee, J.C., Buang, N.A., and Sanip, S.M. (2004b). Scanning Electron
Microscope (SEM) Analysis of As-Grown CNTs over Supported Catalysts. 3rd
Annual Conference of the Asia Pacific Nanotechnology Forum and 2nd Shanghai
International Nanotechnology Cooperation Synposium. 9 – 12 December 2004.
Sydney, Australia: APNF, Inc, 48 – 52.
Ivanov, V., Nagy, J.B., Lambin, P., Lucas, A., Zhang, X.B. and Zhang, X.F. (1994).
The Study of Carbon Nanotubules Produced by Catalytic Method. Chemical
Physics Letters. 223: 329 – 335.
Ivanov, V., Fonseca, A., Nagy, J.B., Lucas, A., Lambin, P., Bernearts, D. and Zhang,
X.B. (1995). Catalytic Production and Purification of Nanotubules Having
Fullerene-Scale Diameters. Carbon. 33(12): 1727 – 1738.
Jacques, D., Villain, S., Rao, A.M., Andrews, R., Derbyshire, F., Dickey, E.C. and
Qian,
D.
(2003).
Synthesis
of
Multiwalled
Carbon
Nanotubes.
Kentucky/University of Kentucky. unpublished. (http://www.caer.uky.edu).
Jeong, S.H. and Lee, K.H. (2003). Fabrication of the Aligned and Patterned Carbon
Canotube Field Emitters Using the Anodic Aluminium Oxide Nano-Template on
a Si Wafer. Synthetic Metals. 139: 385 – 390.
Jing, L., Martin, M. and Tom, H.L. (1998). Nanoscale Electroless Metal Deposition
in Aligned Carbon Nanotubes. Chemistry of Materials. 10: 1963 – 1967.
Jong, W.J., Lai, S.H., Hong, K.H., Lin H.N. and Shih, H.C. (2002). The Effect of
Catalysis on the Formation of One-Dimensional Carbon Structured Materials.
Diamond and Related Materials. 11: 1019 – 1025.
158
Kane, C.L. and Mele, E.J. (1997). Size, Shape and Low Energy Electronic Structure
of Carbon Nanotubes. Physical Review Letters. 78(10): 1932 – 1935.
Kathyayini, H., Nagaraju, N., Fonseca, A. and Nagy, J.B. (2004). Catalytic Activity
of Fe, Co and Fe/Co Supported on Ca and Mg Oxides, Hydroxides and
Carbonates in the Synthesis of Carbon Nanotubes. Journal of Molecular
Catalysis A: Chemical. 223: 129 – 136.
Kibria, A.K.M.F., Mo, Y.H. and Nahm, K.S. (2001). Synthesis of Carbon Nanotubes
over Nickel-Iron catalysts Supported on Alumina under Controlled Conditions.
Catalysis Letters. 71(3 – 4): 229 – 236.
Kibria, A.K.M.F., Mo, Y.H., Nahm, K.S. and Kim, M.J. (2002). Synthesis of
Narrow-Diameter Carbon Nanotubes from Acetylene Decomposition over an
Iron-Nickel Catalyst Supported on Alumina. Carbon. 40: 1241 – 1247.
Kong, J. (2002). Carbon Nanotubes: Synthesis, Integration and Properties. Stanford
University: Ph.D Thesis.
Konya,
Z.
and
Kiricsi,
I.
(2004).
Catalytic
Production,
Purification,
Characterization and Application of Single- and Multiwall Carbon Nanotubes.
Hungary/University
of
Szeged.
unpublished.
(http://www.sci.u-szeged.hu
/appchem/nanotube.htm)
Krill, C.E. and Birringer, R. (2005). Crystallite Size Measurement Using X-ray
Diffraction.
Davis/University
of
California.
unpublished.
(www.matsci.ucdavis.edu/MatSciLT/EMS-162L/Files/XRD-CSize1.pdf).
Krivoruchko, O.P., Maksimova, N.I., Zaikovskii, V.I. and Salanov, A.N. (2000).
Study of Multiwalled Graphite Nanotubes and Filaments Formation from
Carbonised Products of Polyvinyl Alcohol Catalytic Graphitization at 600 –
800°C in Nitrogen Atmosphere. Carbon. 38: 1075 – 1082.
159
Kukovecz, A., Konya, Z., Nagaraju, N., Willems, I., Tamasi, A., Fonseca, A., Nagy,
J.B. and Kiricsi, I. (2000). Catalytic Synthesis of Carbon Nanotubes over Co, Fe
and Ni Containing Conventional and Sol-Gel Silica-Aluminas. Physical
Chemistry Chemical Physics. 2: 3071 – 3076.
Lee, S.M. and Lee, Y.H. (2000). Hydrogen Storage in Single-Walled Carbon
Nanotubes. Applied Physics Letters. 76: 2877 – 2879.
Li, W.Z., Xie, S.S., Qian. L.X., Qian B.H., Chang, B.H., Zou, B.S., Zhou, R.A. and
Wang, G. (1996). Large-scale synthesis of aligned carbon nanotubes. Science.
274: 1701.
Li, M.K., Lu, M., Kong, L.B., Guo, X.Y. and Li, H.L. (2003). The Synthesis of
MWNTs/SWNTs Multiple Phase Nanowire Arrays in Porous Anodic Aluminum
Oxide Templates. Materials Science and Engineering A. 354: 92 – 96.
Liu, B.C., Lyu, S.C., Jung, S.I., Kang, H.K., Yang, C.-W., Park, J.W., Park, C.Y. and
Lee, C.J. (2004). Single-Walled Carbon Nanotubes Produced by Catalytic
Chemical Vapor Deposition of Acetylene over Fe-Mo/MgO Catalyst. Chemical
Physics Letters. 383: 104 – 108.
Liu, D.Y., Chen, J.H., Deng, W., Zhou, H.H. and Kuang, Y.F. (2004). Simple
Catalyst for the Direct Growth of Carbon Nanotubes onto Substrate by
Chemical Vapour Deposition. Materials Letters. 58: 2764 – 2767.
Louie, S.G. (2001). Electronic Properties, Junctions and Defects of Carbon
Nanotubes. In: Dresselhaus, M.S., Dresselhaus, G. and Avouris, P. Carbon
Nanotubes: Synthesis, Structure, Properties and Applications. Berlin, Germany:
Springer-Verlag. 113 – 121.
Lu, A.H., Schmidt, W., Tatar, S.D., Spliethoff, B., Popp, J., Kiefer, W. and Schuth,
F. (2005). Formation of Amorphous Carbon Nanotubes on Ordered Mesoporous
Silica Support. Carbon. 43: 1811 – 1814.
160
Lu, M., Liu, W.-M., Guo, X.-Y. and Li, H.-L. (2004a). Coiled Carbon Nanotubes
Growth via Reduced-Pressure Catalytic Chemical Vapour Deposition. Carbon.
42: 805 – 811.
Lu, M., Wang, Z., Li, H.-L., Guo, X.-Y. and Lau, K.-T. (2004b). Formation of
Carbon Nanotubes in Silicon-Coated Alumina Nanoreactor. Letters to Editor,
Carbon. 42: 1846 – 1849.
Lu. M., Lau, K.-T., Xu, J.-C., Li and H.-L. (2005). Coiled Carbon Nanotubes Growth
and DSC Study in Epoxy-Based Composites. Colloids and Surfaces A:
Physicochemical Engineering Aspects. 257 – 258: 339 – 343.
Luo, G., Li, Z., Wei, F., Xiang, L., Deng, X. and Jin, Y. (2002). Catalysts Effect on
Morphology of Carbon Nanotubes Prepared by Catalytic Chemical Vapor
Deposition in a Nano-Agglomerate Bed. Physica B. 323: 314 – 317.
Maruyama, M., Fukasawa, T., Suenaga, S. and Goto, Y. (2004). Vapor-Grown
Carbon Nanofibers Synthesized from a Fe2O3-Al2O3 Composite Catalyst.
Journal of the European Ceramic Society. 24: 463 – 468.
Moshkalyov, S.S., Moreau, A.L.D., Guittierrez, H.R., Cotta, M.A. and Swart, J.W.
(2004). Carbon Nanotubes Growth by Chemical Vapor Deposition Using Thin
Film Nickel Catalyst. Materials Science and Engineering B. 112: 147 – 153.
Mukhopadhyay, K. and Mathur, G.N. (2003). Synthesis of 2D Quasi-Aligned
Multiwalled Carbon Nanotubes by Catalytic Chemical Vapor Deposition
Method. Journal of Nanoscience. 2(3): 153 – 164.
Nagaraju, N., Fonseca, A., Konya, Z. and Nagy, J.B. (2002). Alumina and Silica
Supported Metal Catalysts for the Production of Carbon Nanotubes. Journal of
Molecular Catalysis A: Chemical. 181: 57 – 62.
Nhut, J.M., Vieira, R., Pesant, L., Tessonnier, J.-P., Keller, N., Ehret, G, Cuong, P.H.
and Ledoux, J. (2002). Synthesis and Catalytic Uses of Carbon and Silicon
Carbide Nanostructures. Catalysis Today. 76: 11 – 32.
161
Papadopoulos, C. (2000). Nanotube Engineering and Science: Synthesis and
Properties of Highly Ordered Carbon Nanotube Arrays and Y-junction Carbon
Nanotubes. University of Toronto: MSc. Thesis.
Perez-Cabero, M., Monzon, A., Rodriguez-Ramos, I. and Guerrero-Ruiz, A. (2004).
Syntheses of CNTs over Several Iron-Supported Catalysts: Influence of the
Metallic Precursors. Catalysis Today. 93 – 95: 681 – 687.
Piedigrosso, P., Konya, Z., Colomer, J.F., Fonseca, A., Van Tendeloo, G. and Nagy,
J.B. (2000). Production of Differently Shaped Multi-wall Carbon Nanotubes
Using Various Cobalt Supported Catalysts. Physical Chemistry Chemical
Physics. 2: 163 – 170.
Popov, V. N. (2004). Carbon Nanotubes: Properties and Application. Materials
Science and Engineering R. 43: 61 – 102.
Qiu, J., An, Y., Zhao, Zhao, Z., Li, Y. and Zhou, Y. (2004). Catalytic Synthesis of
Single-Walled Carbon Nanotubes from Coal Gas by Chemical Vapor Deposition
Method. Fuel Processing Technology. 85: 913 - 920.
Rao, C.N.R., Seshadri, R., Govindaraj, A. and Sen, R. (1995). Fullerenes, Nanotubes,
Onions and Related Carbon Structures. Materials Science and Engineering.
15(95): 209 – 262.
Reimer, L. (1984). Tranmission Electron Microscopy: Physics of Image Formation
and Microanalysis. Berlin, Germany: Springer-Verlag.
Rikagu (2005). Crystallite Size and Microstrain Analysis of Thin Films. AMIA
Laboratories. unpublished. (http://www.rigakumsc.com/contract/amia_res_TNC01.htm).
Saito, R., Dresselhaus, G. and Dresselhaus, M.S. (1998). Physical Properties of
Carbon Nanotubes. London: Imperial College Press.
Scokart, P.O. and Rouxhet, P.G. (2002). Comparison of the Acid-Basen Properties of
Various of Oxides and Chemically Treated Oxides. Journal of Colloid and
Interface Science. 86(1): 96 – 104.
162
Shajahan, Md., Mo, Y.H., Kibria, A.K.M.F., Kim, M.J. and Nahm, K.S. (2004). High
Growth of SWNTs and MWNTs from C2H2 Decomposition over Co-Mo / MgO
Catalysts. Carbon. 42: 2245 – 2253.
Takenaka, S., Serizawa, M. and Otsuka, K. (2004). Formation of Filamentous
Carbons over Supported Fe Catalysts through Methane Decomposition. Journal
of Catalysis. 222: 520 – 531.
Tauster, S.J., Fung, S.C. and Garten, R.L. (1978). Strong Metal-Support Interactions.
Group 8 Noble Metals Supported on Titanium Dioxide. Journal of the American
Chemical Society. 100(1): 170 – 175.
Tee, J.C., Buang, N.A., Sanip, S.M. and Ismail, A.F. (2004a). Characterization on
Series of Supported Metal Oxides Catalysts for the Synthesis of Carbon
Nanotubes (CNTs). Journal of Solid State Science and Technology Letters
(Supplement). 11(2): 34.
Tee, J.C., Buang, N.A., Sanip, S.M. and Ismail, A.F. (2004b). Synthesis of Carbon
Nanotubes (CNTs) Using Supported Catalysts. Advances in Fual Cell Research
and Development in Malaysia. 3 – 5 December 2004. Kuala Lumpur: BATC,
UTM, 170 – 179.
Valiente, A.M., Lopez, P.N., Ramos, I.R., Ruiz, A.G., Li, C. and Xin, Q. (2000). In
Situ Study of Carbon Nanotube formation by C2H2 decomposition on an IronBased Catalyst. Carbon. 38: 2003 – 2006.
Vander Wal, R.L., Ticich, T.M. and Curtis, V.E. (2001). Substrate-Support
Interactions in Metal-Catalyzed Carbon Nanofiber Growth. Carbon. 39: 2277 –
2289.
Ward, J.W. (2002). Thin Film Catalyst and Substrate Interactions on the CCVD
(Catalytic Chemical Vapor Deposition) Growth of Carbon Nanotubes.
Rensselaer Polytechnic Institute: Ph.D. Thesis.
Watt, I.M. (1985) The Principles and Practice of Electron Microscopy. Cambridge,
Great Britain: Cambridge University Press.
163
Willems, I., Konya, Z., Colomer, J.F., Van Tendeloo, G., Nagaraju, N., Fonseca, A.
and Nagy, J.B. (2000). Control of the Outer Diameter of Thin Carbon Nanotubes
Synthesized by Catalytic Decomposition of Hydrocarbons. Chemical Physics
Letters. 317: 71 – 76.
Willems, I., Konya, Z., Fonseca, A. and Nagy J.B. (2002). Heterogeneous Catalytic
Production and Mechanical Resistance of Nanotubes Prepared on Magnesium
Oxide supported Co-based Catalysts. Applied Catalysis A: General. 229: 229 –
233.
Williams, K.A. (2001). Synthesis and Characterization of Single-Walled Carbon
Nanoutubes and Investigation of their Molecular Adsorption Properties. The
Pennsylvania State University: Ph.D Thesis.
Yacaman, M.J., Yoshida, M.M., Rendon, L. and Santiesteban, J.G. (1993). Catalytic
Growth of Carbon Microtubules with Fullerene Structure. Applied Physics
Letters. 62(2): 202 – 204.
Yakobson, B.I. and Avouris, P. (2001). Mechanical Properties of Crabon Nanotubes.
In: Dresselhaus, M.S., Dresselhaus, G. and Avouris, P. Carbon Nanotubes:
Synthesis, Structure, Properties and Applications. Berlin, Germany: SpringerVerlag. 287 – 298.
Yen, J.H., Leu, I.C., Lin, C.C. and Hon, M.H. (2004). Effect of Catalyst Pretreatment
on the Growth of Carbon Nanotubes. Diamond and Related Materials. 13: 1237
– 1241.
Yu, M.F., Files, S.B., Arepalli, S. and Ruoff, R.S. (2000). Tensile Loading of Ropes
of Single Wall Carbon Nanotubes and Their Mechanical Properties. Physical
Review Letters. 84(24): 5552 – 5555.
Yudanov, I., Pacchioni, G., Neyman, K. and Rosch, N. (1997). Systematic Density
Functional Study of the Adsorption of Transition Metal Atoms on the MgO(001)
Surface. The Journal of Physical Chemistry B. 101(15): 2786 – 2792.
164
Zou, X.P., Abe, H., Shimizu, T., Ando, A., Nakayama, Y., Tokumoto, H., Zhu, S.M.
and Zhou, H.S. (2004). Simple Thermal Chemical Vapour Deposition Synthesis
and Electrical Property of Multi-Walled Carbon Nanotubes. Physica E. 24: 14 –
18.
APPENDIX A
MATHEMATICAL EXPRESSIONS FOR STOICHIOMETRIC AMOUNT OF
METAL SALT USED IN WET IMPREGNATION
A.1
2.5 wt. % of Co Relative to 1.00 g of Support
= 58.93 g mole-1
Atomic weight of Co
Molecular weight of Co(CH3COO)2.4H2O = 248.93 g mole-1
Weight of support used = 1.00 g
Support : Co (wt. %) =
97.5
:
2.5
= 97.5 / 97.5 : 2.5 / 97.5
=
1
: 0.02564
Weight of Co on 1.00 g of support = 1.00 g × 0.02564
= 0.02564 g
Mole of Co = 0.02564 g / 58.93 g mole-1
≈ 0.0004351 mol
1
mole of Co ≡
1
mole of Co(CH3COO)2.4H2O
0.0004351 mole of Co ≡ 0.0004351 mole of Co(CH3COO)2.4H2O
Stoichiometric amount of Co(CH3COO)2.4H2O = 0.0004351 mole × 248.93 g mole-1
= 0.1083 g ≈ 0.11 g
166
A.2
2.5 wt. % of Fe Relative to 1.00 g of Support
= 55.85 g mole-1
Atomic weight of Fe
Molecular weight of Fe(CH3COO)2 = 173.94 g mole-1
Weight of support used = 1.00 g
Support : Fe (wt. %) =
97.5
:
2.5
= 97.5 / 97.5 : 2.5 / 97.5
=
1
: 0.02564
Weight of Fe on 1.00 g of support = 1.00 g × 0.02564
= 0.02564 g
Mole of Fe = 0.02564 g / 55.85 g mole-1
≈ 0.0004591 mol
1
mole of Fe ≡
1
mole of Fe(CH3COO)2
0.0004591 mole of Fe ≡ 0.0004591 mole of Fe(CH3COO)2
Stoichiometric amount of Fe(CH3COO)2 = 0.0004591 mole × 173.94 g mole-1
= 0.07986 g
≈ 0.08 g
A.3
2.5 wt. % of Co/Fe Relative to 1.00 g of Support
Weight of support used = 1.00 g
Support : Co : Fe(wt. %) =
97.5
:
1.25
:
1.25
= 97.5 / 97.5 : 1.25 / 97.5 : 1.25 / 97.5
=
1
:
0.01282 : 0.01282
167
Weight of Co on 1.00 g of support = 1.00 g × 0.01282
= 0.01282 g
Mole of Co = 0.01282 g / 58.93 g mole-1
≈ 0.0002175 mol
1
mole of Co ≡
1
mole of Co(CH3COO)2.4H2O
0.0002175 mole of Co ≡ 0.0002175 mole of Co(CH3COO)2.4H2O
Stoichiometric amount of Co(CH3COO)2.4H2O = 0.0002175 mole × 248.93 g mole-1
= 0.05414 g
≈ 0.05 g
Weight of Fe on 1.00 g of support = 1.00 g × 0.01282
= 0.01282 g
Mole of Fe = 0.01282 g / 55.85 g mole-1
≈ 0.0002295 mol
1
mole of Fe ≡
1
mole of Fe(CH3COO)2
0.0002295 mole of Fe ≡ 0.0002295 mole of Fe(CH3COO)2
Stoichiometric amount of Fe(CH3COO)2 = 0.0002295 mole × 173.94 g mole-1
= 0.03992 g
≈ 0.04 g
APPENDIX B
MATHEMATICAL EXPRESSIONS FOR STOICHIOMETRIC AMOUNT OF
METAL SALT USED IN DIP COATING
B.1
0.5 M of Cobalt Acetate Aqueous Solution
Molecular weight of Co(CH3COO)2.4H2O = 248.93 g mole-1
Concentration of Co aqueous solution = 0.5 M
Volume of solution = 25 mL = 0.025 L
0.5 mole of Co ≡ 1 L of solution
X mole of Co ≡ 0.025 L of solution
X = 0.025 × 0.5
= 0.0125 mole
Stoichiometric amount of Co(CH3COO)2.4H2O = 0.0125 mole × 248.93 g mole-1
= 3.1116 g
≈ 3.11 g
169
B.2
0.5 M of Ferrum Acetate Aqueous Solution
Molecular weight of Fe(CH3COO)2 = 173.94 g mole-1
Concentration of Fe aqueous solution = 0.5 M
Volume of solution = 25 mL = 0.025 L
0.5 mole of Fe ≡ 1 L of solution
Y mole of Fe ≡ 0.025 L of solution
Y = 0.025 × 0.5
= 0.0125 mole
Stoichiometric amount of Fe(CH3COO)2 = 0.0125 mole × 173.94 g mole-1
= 2.1743 g
≈ 2.17 g
B.3
0.5 M of Co/Fe Acetate Aqueous Solution
Concentration of Co aqueous solution = 0.25 M
Volume of solution = 25 mL = 0.025 L
0.25 mole of Co ≡ 1 L of solution
X mole of Co ≡ 0.025 L of solution
X = 0.025 × 0.25
= 0.00625 mole
Stoichiometric amount of Co(CH3COO)2.4H2O = 0.00625 mole × 248.93 g mole-1
= 1.5558 g
≈ 1.56 g
Concentration of Fe aqueous solution = 0.25 M
Volume of solution = 25 mL = 0.025 L
170
0.25 mole of Fe ≡ 1 L of solution
Y mole of Fe ≡ 0.025 L of solution
Y = 0.025 × 0.25
= 0.00625 mole
Stoichiometric amount of Fe(CH3COO)2 = 0.00625 mole × 173.94 g mole-1
= 1.0871 g
≈ 1.09 g
APPENDIX C
MATHEMATICAL EXPRESSIONS FOR CRYSTALLITE SIZE OF Co3O4
CUBIC CRYSTAL USING SCHERRER EQUATION
C.1
Co3O4 Cubic Crystal in Al-Co/Fe(2.5) Catalyst
d=
where d
Kλ
FWHM cos θ
(Equation 4.3)
= Crystallite size (Å)
K
= Unit cell geometry dependent constant (0.94)
λ
= Wavelength of X-rays (Å)
FWHM = Full width at half maximum peak intensity (radians)
θ
= Bragg angle (°)
Peak 1:
d =
1 . 4482
0 . 003665 cos (15 . 6485 )
= 410 . 37 Å
Peak 2:
d =
1 . 4482
0 . 004294 cos (18 . 4870 )
= 355 . 61 Å
172
Peak 3:
d =
1 . 4482
0 . 006999 cos ( 32 . 6170 )
= 245 . 80 Å
410.37 + 355.61 + 245.80
Å
3
= 337.26 Å
d average =
= 33.73 nm
APPENDIX D
MATHEMATICAL EXPRESSIONS FOR PARTICLE SIZE, LENGTH AND
DIAMETER FROM THE SEM MICROGRAPHS
D.1
Calculation of Particle Size of Catalyst
measurement of particle diameter from micrograph
× scale bar unit
measurement of scale bar
14 mm
=
× 20 µm
44 mm
= 6.36 µm
Particle size =
≈ 6.4 µm
D.2
Calculation of CNTs Length
Length =
measuremen t of CNTs length from micrograph
measuremen t of scale bar
2.5 mm
=
× 2 µm
56 mm
= 0.0893 µm
≈ 89.3 nm
× scale bar unit
APPENDIX E
PUBLICATIONS
Buang, N.A., Aziz, M., Sanip, S., Tee, J.C., Abidin, Z. and Ismail, A.F. (2005).
Effect of Pretreatment of Synthetic and Natural Carbons as Starting Materials for
Carbon Nanotubes. Journal of Metastable and Nanocrystalline Materials. 23:
285 – 288. (http://www.scientific.net).
Ismail, A.F., Tee, J.C., Buang, N.A., and Sanip, S.M. (2004). Effect of Catalyst
Supports on the Growth of CNTs. 3rd Annual Conference of the Asia Pacific
Nanotechnology Forum and 2nd Shanghai International Nanotechnology
Cooperation Synposium. 9 – 12 December 2004. Sydney, Australia: APNF, Inc,
150 – 155.
Ismail, A.F., Tee, J.C., Buang, N.A., and Sanip, S.M. (2004). Scanning Electron
Microscope (SEM) Analysis of As-Grown CNTs over Supported Catalysts. 3rd
Annual Conference of the Asia Pacific Nanotechnology Forum and 2nd Shanghai
International Nanotechnology Cooperation Synposium. 9 – 12 December 2004.
Sydney, Australia: APNF, Inc, 48 – 52.
Tee, J.C., Buang, N.A., Sanip, S.M. and Ismail, A.F. (2004). Characterization on
Series of Supported Metal Oxides Catalysts for the Synthesis of Carbon
Nanotubes (CNTs). Journal of Solid State Science and Technology Letters
(Supplement). 11(2): 34.
175
Tee, J.C., Buang, N.A. and Sanip, S.M. (2003), Heat Treatment Studies on Activated
Carbon and Charcoal for Carbon Nanotube Synthesis. The Proceedings of
Annual Fundamental Science Seminar (AFSS), 2003. 20 – 21 May 2003. UTM
Skudai, Malaysia: Ibnu Sina Institute, 134 – 141.
Tee, J.C., Buang, N.A., Aziz. M., Sanip, S.M. and Ismail, A.F. (2003). Heat
Treatment Studies on Activated Carbon and Charcoal for Carbon Nanotubes
Synthesis. Advances in Malaysian Fual Cell Research and Development. 27 – 30
June 2003., Bangi, Malaysia: Faculty of Engineering, UKM, 187 – 196.
Ismail, A.F., Aziz, M., Buang, N.A., Husin, R., Saidin, M.K., Arshad, K.A., Tee J.C.,
Sanip, S.M., Abidin, Z.R.Z. and Razak, N.A. (2003). Morphological Studies of
Pretreated Carbon Materials for Carbon Nanotubes Synthesis. National
Symposium on Science and Technology (NSST). 28 – 30 July 2003. Kuala
Lumpur, Malaysia: MOSTI.
Tee, J.C., Buang, N.A. and Ismail, A.F. (2004). Synthesis and Characterization of
Carbon Nanotubes Using Supported Catalyst Methods. Colloquium of Fuel Cell
- Carbon Nanotubes. 18 May 2004. UTM Skudai: CNT Reaserch Group, 10 –
15.
Tee, J.C., Buang, N.A., Sanip, S.M. and Ismail, A.F. (2004). Synthesis of Carbon
Nanotubes (CNTs) Using Supported Catalysts. Advances in Fual Cell Research
and Development in Malaysia. 3 – 5 December 2004. Kuala Lumpur, Malaysia:
BATC, UTM, 170 – 179.
Tee, J.C., Buang, N.A., Ismail, A.F., Majid, Z.A., Aziz, M. and Sanip, S.M. (2005).
Synthesis of Carbon Nanotubes (CNTs) via Custom Built Catalytic Chemical
Vapour Deposition (CCVD) System. Annual Fundamental Science Seminar
(AFSS), 2005. 4 – 6 July 2005. UTM Skudai, Malaysia: Ibnu Sina Institute.
Tee Jia Chee. Kajian Awal terhadap Struktur Fizikal dan Komposisi Bahan
Berasaskan Karbon bagi Penghasilan Karbon Nanotiub untuk Sistem
Penyimpanan Hidrogen. B.Sc. Thesis. Universiti Teknologi Malaysia; 2003.