Upload
others
View
17
Download
0
Embed Size (px)
Citation preview
Structure and Properties of Nanomaterials: From Inorganic Boron Nitride Nanotubes to the
Calcareous Biomineralized Tubes of H. dianthus
by
Adrienne Elizabeth Tanur
A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy
Department of Chemistry University of Toronto
© Copyright by Adrienne Elizabeth Tanur 2012
ii
Structure and Properties of Nanomaterials: From Inorganic Boron Nitride Nanotubes to the
Calcareous Biomineralized Tubes of H. dianthus
Adrienne Elizabeth Tanur
Doctor of Philosophy
Department of Chemistry
University of Toronto
2012
Abstract
Several nanomaterials systems, both inorganic and organic in nature, have been extensively
investigated by a number of characterization techniques including atomic force microscopy
(AFM), electron microscopy, Fourier transform infrared spectroscopy (FTIR), and energy
dispersive x-ray spectroscopy (EDX). The first system consists of boron nitride nanotubes
(BNNTs) synthesized via two different methods. The first method, silica-assisted catalytic
chemical vapour deposition (SA-CVD), produced boron nitride nanotubes with different
morphologies depending on the synthesis temperature. The second method, growth vapour
trapping chemical vapour deposition (GVT-CVD), produced multiwall boron nitride nanotubes
(MWBNNTs). The bending modulus of individual MWBNNTs was determined using an AFM
three-point bending technique, and was found to be diameter-dependent due to the presence of
shear effects. The second type of nanomaterial investigated is the biomineralized calcareous
iii
shell of the serpulid Hydroides dianthus. This material was found to be an inorganic-organic
composite material composed of two different morphologies of CaCO3, collagen, and
carboxylated and sulphated polysaccharides. The organic components were demonstrated to
mediate the mineralization of CaCO3 in vitro. The final system studied is the proteinaceous
cement of the barnacle Amphibalanus amphitrite. The secondary structure of the protein
components was investigated via FTIR, revealing the presence of β-sheet conformation, and
nanoscale rod-shaped structures within the cement were identified as β-sheet containing amyloid
fibrils via chemical staining. These rod-shaped structures exhibited a stiffer nature compared
with other structures in the adhesive, as measured by AFM nanoindentation.
iv
Acknowledgments
I've heard it said
That people come into our lives for a reason
Bringing something we must learn
And we are led
To those who help us most to grow
If we let them
And we help them in return
Well, I don't know if I believe that's true
But I know I'm who I am today
Because I knew you
- Glinda the Good Witch (Wicked, the musical)
I am indebted to my supervisor, Professor Gilbert Walker, for his support and guidance over the
years. Always perceptive, attentive, encouraging, and ready with a helping hand, I would not be
in the position I am now without him. His infectious enthusiasm for science and team-building
activities outside of the lab won’t be forgotten, and I am proud to be a member of his science
family.
I would also like to thank my supervisory committee members, Professors Al-Amin Dhirani and
Zhenghong Lu for their helpful input and advice. I am very appreciative of the time and
consideration they have given me over the course of my graduate studies.
To the members of Walker Labs, past and present – it is astonishing how well we all get along!
I cannot imagine what lab life would be like without you. To the postdocs, Doctors Shan Zou,
Nikhil Gunari, Weiqing Shi, Zahra Fakhraai, and Leela Reddy: thank you for the knowledge and
experience you shared with me, as both mentors and friends. To my fellow graduate students,
Doctors Shell Ip, James Li, Ruby Sullan, and Isaac Li; Melissa Paulite, Claudia Grozea,
Christina MacLaughlin, Dan Lamont, Alex Kumachev, Colin Zamecnik, Duncan Smith-
Halverson, and Alex Stewart: it has been a pleasure working with such curious, industrious, and
humorous people. In particular, I would like to thank Shell, James, Isaac, and Melissa for our
numerous insightful discussions which helped me to solve many problems and come up with
v
new ideas. To my co-authors Nikhil, Ruby, and Melissa: I have learned so much from working
with you – not only about science; but about perseverance, determination, and integrity.
Certain measurements have helped me to re-focus time and time again, and for this I am grateful
to the S4S team (you know who you are). I am also grateful to my best girls Aliza Kassam and
Courtney Smyth for always being there, sharing my highs and lows. Thanks as well to my
EngSci friends, who saw me through undergrad and beyond.
I also appreciate the support of my extended family, who sat patiently through a “what is nano?”
presentation when I tried to explain what I do, and who always made sure I was supplied with
leftovers from family parties. To my second family, the Joes: thank you for making me a part of
your family, and for your support.
Thank you Mom and Dad, for all of the sacrifices you have made for me and for your constant
encouragement and support. I appreciate that you both (quietly) expect great things from me
because you believe that I can achieve them. To my big brother Luke: you probably set me on
this path by passing on your love of reading and science fiction. Thank you for listening to my
presentations and putting up with my mess while we lived together, and for always looking out
for me. To my little sister Cheryl: you are the true chemist and I loved having you around in the
department. We must be more alike than we would care to admit, because everyone recognized
that we are sisters. Having you around gives me the confidence to rise to any occasion, because I
have to act my part as the big sister.
Finally, to my fiancé Chris Joe: Thank you for your patience with me over the past 10 years, for
understanding how much this means to me, and for your constant love.
Adrienne Elizabeth Tanur
August 7th
, 2012
vi
Table of Contents
Acknowledgments.......................................................................................................................... iv
Table of Contents ........................................................................................................................... vi
List of Symbols ............................................................................................................................. xii
List of Abbreviations ................................................................................................................... xiii
List of Tables ............................................................................................................................... xiv
List of Figures ................................................................................................................................xv
1 Introduction .................................................................................................................................1
1.1 Nanomaterials: Materials Revolution, Natural Evolution ...................................................1
1.2 Nanoscale Characterization Methods ..................................................................................2
1.3 Hexagonal Boron Nitride Nanomaterials .............................................................................2
1.4 Marine Fouling Organisms: Adhesive Nanomaterials .........................................................4
1.5 Summary of Thesis ..............................................................................................................5
1.6 References ............................................................................................................................5
2 Atomic Force Microscopy and Spectroscopy ...........................................................................10
2.1 Introduction ........................................................................................................................10
2.2 Contact Mode Imaging ......................................................................................................11
2.3 Intermittent Contact (Tapping) Mode Imaging..................................................................12
2.4 Force Spectroscopy ............................................................................................................13
2.5 References ..........................................................................................................................14
3 Synthesis of Boron Nitride Nanotubes ......................................................................................16
3.1 Permissions ........................................................................................................................16
vii
3.2 Abstract ..............................................................................................................................16
3.3 Introduction ........................................................................................................................16
3.3.1 Arc Discharge ........................................................................................................16
3.3.2 Laser Heating/Ablation ..........................................................................................17
3.3.3 Templated Synthesis ..............................................................................................19
3.3.4 Chemical Vapour Deposition .................................................................................21
3.3.5 Mechano-Thermal (Ball milling and Annealing) ..................................................23
3.3.6 Chemical Synthesis ................................................................................................24
3.3.7 Comparison of Methods .........................................................................................24
3.4 Experimental Methods .......................................................................................................25
3.4.1 Method 1: Silica-Assisted Catalytic Chemical Vapour Deposition .......................25
3.4.2 Method 2: Growth Vapour Trapping Chemical Vapour Deposition .....................26
3.5 Results ................................................................................................................................27
3.5.1 Macroscopic Description of Products ....................................................................27
3.5.2 Nanotube Morphology ...........................................................................................29
3.6 Discussion ..........................................................................................................................30
3.6.1 BNNT Growth Mechanisms in CVD Synthesis ....................................................30
3.6.2 Qualitative Comparison of Methods ......................................................................34
3.7 Conclusions ........................................................................................................................34
3.8 Contributions......................................................................................................................35
3.9 References ..........................................................................................................................35
4 Structural and Chemical Characterization of Boron Nitride Nanotubes ...................................39
4.1 Abstract ..............................................................................................................................39
4.2 Introduction ........................................................................................................................39
4.3 Experimental Methods .......................................................................................................40
viii
4.3.1 Synthesis ................................................................................................................40
4.3.2 Electron Microscopy ..............................................................................................40
4.3.3 Energy Dispersive X-ray........................................................................................41
4.3.4 Fourier Transform Infrared Spectroscopy .............................................................41
4.4 Results ................................................................................................................................41
4.4.1 Scanning Transmission Electron Microscopy .......................................................41
4.4.2 EDX Characterization ............................................................................................43
4.4.3 FTIR Characterization ...........................................................................................46
4.5 Discussion ..........................................................................................................................47
4.5.1 Nanotube Morphology and Structure.....................................................................47
4.5.2 Defect Characterization .........................................................................................48
4.6 Conclusions ........................................................................................................................49
4.7 References ..........................................................................................................................49
5 Diameter-Dependent Bending Modulus of Individual Multiwall Boron Nitride Nanotubes ...50
5.1 Abstract ..............................................................................................................................50
5.2 Introduction ........................................................................................................................50
5.3 Experimental Methods .......................................................................................................52
5.3.1 MWBNNT Synthesis .............................................................................................52
5.3.2 Chemical and Structural Characterization .............................................................52
5.3.3 Sample Preparation and AFM Measurements .......................................................53
5.4 Results and Discussion ......................................................................................................54
5.4.1 Characterization of MWBNNTs ............................................................................54
5.4.2 AFM Three-Point Bending ....................................................................................54
5.4.3 Elastic Properties of MWBNNTs ..........................................................................60
5.5 Conclusion .........................................................................................................................66
ix
5.6 Contribution .......................................................................................................................67
5.7 References ..........................................................................................................................67
6 Insights into the composition, morphology, and formation of the calcareous shell of the
serpulid Hydroides dianthus .....................................................................................................72
6.1 Permissions ........................................................................................................................72
6.2 Abstract ..............................................................................................................................72
6.3 Introduction ........................................................................................................................72
6.4 Experimental Methods .......................................................................................................74
6.4.1 Tubeworm Collection and Preservation.................................................................74
6.4.2 X-Ray Diffraction (XRD) ......................................................................................75
6.4.3 Fourier Transform Infrared Spectroscopy (FTIR) .................................................75
6.4.4 Inductively Coupled Plasma Atomic Emission Spectrometry (ICPAES)
Analysis..................................................................................................................76
6.4.5 Electron Probe Microanalysis (EPMA) .................................................................76
6.4.6 Separation of the Organic Tube Lining and the Soluble Organic Matrix
(SOM) ....................................................................................................................76
6.4.7 Scanning Electron Microscopy (SEM) and Energy Dispersive X-Ray (EDX) .....76
6.4.8 Atomic Force Microscopy (AFM) Imaging and Nanoindentation ........................77
6.4.9 Light Microscopy and Chemical Staining .............................................................78
6.4.10 Amino Acid Analysis of the Soluble Organic Matrix (SOM) ...............................78
6.4.11 In vitro CaCO3 Mineralization Experiments ..........................................................79
6.5 Results ................................................................................................................................80
6.5.1 Bulk Composition of the Tube Shell and Adhesive Material ................................80
6.5.2 Tube Shell Ultrastructure and Spatial Composition ..............................................82
6.5.3 Adhesive Material Structure and Composition ......................................................84
6.5.4 Mechanical Properties of the Adhesive Material ...................................................86
x
6.5.5 Mechanical Properties of the Tube Shell ...............................................................88
6.5.6 Characterization of the Organic Tube Lining ........................................................88
6.5.7 Characterization of the SOM .................................................................................93
6.5.8 Characterization of the Remineralized Organic Tube Lining ................................96
6.5.9 Characterization of the SOM Mineralization Precipitates .....................................98
6.6 Discussion ........................................................................................................................101
6.6.1 Tube Layering and Mechanical Properties ..........................................................101
6.6.2 CaCO3 Polymorphs and Morphologies ................................................................104
6.6.3 Insights into the Formation and Attachment of the Adhesive Material to the
Substrate ...............................................................................................................105
6.6.4 SOM Composition ...............................................................................................106
6.6.5 IOM Composition ................................................................................................107
6.6.6 Summary of the Structure and Composition of the Tube Shell and Adhesive
Material ................................................................................................................109
6.6.7 Role of the Organic Tube Lining in Tube Formation ..........................................109
6.7 Conclusions ......................................................................................................................112
6.8 Contributions....................................................................................................................113
6.9 References ........................................................................................................................113
7 Nanoscale Structures and Properties of the Proteinaceous Cement of the Barnacle
Amphibalanus amphitrite ........................................................................................................119
7.1 Permissions ......................................................................................................................119
7.2 Abstract ............................................................................................................................119
7.3 Introduction ......................................................................................................................119
7.3.1 FTIR Characterization of Protein Secondary Structure .......................................120
7.3.2 Barnacle Cement: Proteinaceous Glue .................................................................121
7.4 Experimental Methods .....................................................................................................121
xi
7.4.1 Barnacle Rearing ..................................................................................................121
7.4.2 Fourier Transform Infrared (FTIR) Spectroscopy ...............................................122
7.4.3 Atomic Force Microscopy (AFM) Imaging and Indentation ...............................123
7.4.4 Scanning Electron Microscopy (SEM) and Energy Dispersive X-ray (EDX) ....123
7.4.5 Chemical Staining ................................................................................................124
7.5 Results ..............................................................................................................................124
7.5.1 FTIR Spectra ........................................................................................................124
7.5.2 AFM Images, Force Curves, and Moduli Histograms .........................................125
7.5.3 SEM and EDX .....................................................................................................129
7.5.4 Amyloid-Selective Staining .................................................................................130
7.6 Discussion ........................................................................................................................130
7.6.1 Significance of β-sheet Conformation in Barnacle Cement ................................130
7.7 Conclusions ......................................................................................................................132
7.8 Contributions....................................................................................................................132
7.9 References ........................................................................................................................132
8 Summary and Outlook ............................................................................................................135
8.1 Summary of Thesis ..........................................................................................................135
8.2 Outlook ............................................................................................................................136
8.2.1 Boron Nitride Nanotubes .....................................................................................136
8.2.2 Nanomaterials in Nature ......................................................................................136
8.3 References ........................................................................................................................137
xii
List of Symbols
a suspended length to the left of applied force
b suspended length to the right of applied force
D diameter
EB bending modulus
EY Young’s modulus
F loading force
G shear modulus
I second moment of area
k spring constant
keff effective spring constant
L suspended length
R radius of curvature
δ deflection (Chapter 5)
δ indentation (Chapter 6, 7)
ν Poisson’s ratio
xiii
List of Abbreviations
AFM atomic force microscopy/microscope
BN boron nitride
BNNT boron nitride nanotube
c-BN cubic boron nitride
CNT carbon nanotube
CVD chemical vapour deposition
DCBM double clamped beam model
EDX energy dispersive x-ray spectroscopy
FTIR Fourier transform infrared spectroscopy
GVT-CVD growth vapour trapping chemical vapour deposition
h-BN hexagonal boron nitride
HR-TEM high resolution transmission electron microscopy/microscope
IR infrared
LO longitudinal optical
MSBM mixed support beam model
MWBNNT multiwall boron nitride nanotube
MWCNT multiwall carbon nanotube
SA-CVD silica-assisted chemical vapour deposition
SE standard error
SEM scanning electron microscopy/microscope
SSBM simply supported beam model
STEM scanning transmission electron microscopy/microscope
SWBNNT single wall boron nitride nanotube
SWCNT single wall carbon nanotube
TEM transmission electron microscopy/microscope
TO transverse optical
XRD x-ray diffraction
xiv
List of Tables
Table 3.1 Comparison of BNNT Synthesis Methods ................................................................... 25
Table 4.1 Infrared modes of h-BN and c-BN ............................................................................... 40
Table 4.2 FTIR Peak Positions for BN Nanomaterials. ............................................................... 48
Table 6.1 Chemical composition of seawater vs. the artificial seawater used in all CaCO3
precipitation experiments .............................................................................................................. 79
Table 6.2 Summary of IR bands for the tube shell sample FTIR spectrum. ................................ 81
Table 6.3 Summary of IR bands for the organic tube lining and the SOM FTIR spectra. .......... 94
Table 6.4 Amino acid composition for the SOM. ........................................................................ 95
Table 7.1 IR peaks and the corresponding fraction of the observed secondary structures found in
gummy barnacle cement sample ................................................................................................. 125
xv
List of Figures
Figure 1.1 Structures of h-BN and graphite. .................................................................................. 3
Figure 1.2 Structure of a carbon nanotube and a boron nitride nanotube. ..................................... 3
Figure 2.1 Schematic of atomic force microscope. ...................................................................... 11
Figure 3.1 Position of alumina boat within quartz test tube. ....................................................... 27
Figure 3.2 GVT-CVD set-up. ...................................................................................................... 27
Figure 3.3 Quartz boat after Method 1, SA-CVD (1150 oC, 1 h). ............................................... 28
Figure 3.4 Alumina boat after Method 2, GVT-CVD (1150 oC, 2 h). ......................................... 28
Figure 3.5 Electron micrographs of the products of SA-CVD.. .................................................. 29
Figure 3.6 SEM image of MWBNNTs synthesized by GVT-CVD ............................................ 30
Figure 3.7 Schematic of BNNT growth. ...................................................................................... 32
Figure 4.1 Scanning transmission electron microscopy images of BNNTs produced by SA-CVD
....................................................................................................................................................... 42
Figure 4.2 Transmission electron microscopy images of BNNTs produced by GVT-CVD. ...... 42
Figure 4.3 EDX spectrum of SA-CVD BNNTs on Si substrate prior to HF purification ........... 43
Figure 4.4 EDX spectrum of SA-CVD Bamboo BNNTs (after HF purification) on Si substrate.
....................................................................................................................................................... 44
Figure 4.5 SEM image of GVT-CVD BNNTs on lacy C TEM grid and corresponding EDX
elemental maps.............................................................................................................................. 45
Figure 4.6 EDX spectrum of GVT-CVD BNNTs on lacy C TEM grid. ..................................... 46
xvi
Figure 4.7 Fourier transform infrared spectra of BNNTs produced by SA-CVD and GVT-CVD.
....................................................................................................................................................... 47
Figure 5.1 SEM and TEM images of MWBNNTs ...................................................................... 55
Figure 5.2 FTIR spectrum of MWBNNTs. .................................................................................. 56
Figure 5.3 Beam schematics describing beam bending boundary conditions. ............................ 57
Figure 5.4 SEM and AFM images and maps of MWBNNTs on patterned Si substrate ............. 59
Figure 5.5 Representative AFM force curves. ............................................................................. 59
Figure 5.6 Tube effective stiffness (keff) vs. position along suspended tube (a/L) ....................... 60
Figure 5.7 Bending modulus vs. tube outer diameter .................................................................. 62
Figure 5.8 Determination of the Young’s modulus and shear modulus via a fit to plot of 1/EB vs
(D/L)2 ............................................................................................................................................ 65
Figure 6.1 Overview of serpulid tube structure (Hydroides dianthus) ........................................ 73
Figure 6.2 XRD and FTIR spectra for powdered sample of entire tube ...................................... 80
Figure 6.3 SEM images of tube shell transverse cross section .................................................... 84
Figure 6.4 SEM images of adhesive material, transverse cross section. ..................................... 85
Figure 6.5 SEM images of the variety of crystal morphologies observed for Mg-calcite and
aragonite. ....................................................................................................................................... 86
Figure 6.6 SEM images of the adhesive material surface, substrate side, showing the presence
and incorporation of various biofilm components ........................................................................ 87
Figure 6.7 AFM height images and elastic moduli histograms of adhesive material surface ..... 88
xvii
Figure 6.8 SEM images illustrating the different crystal morphologies and orientations present at
the adhesive material surface, adjacent to the substrate ............................................................... 89
Figure 6.9 Optical images of organic tube lining......................................................................... 89
Figure 6.10 SEM images of the adhesive material (lumen-side) and the EDTA-treated tube
lining ............................................................................................................................................. 90
Figure 6.11 EDX spectrum for the “smooth layer” (possibly an organic sheet). ........................ 91
Figure 6.12 Optical image of the insoluble organic matrix after Masson’s trichrome staining .. 91
Figure 6.13 AFM amplitude images and linescans of fibres from the insoluble organic matrix.
....................................................................................................................................................... 92
Figure 6.14 FTIR spectra of the tube shell organic matrices ....................................................... 93
Figure 6.15 SEM images of the IOM, treated overnight with 0.5 M EDTA ............................... 96
Figure 6.16 SEM images and EDX spectra of crystals formed on the EDTA-demineralized
organic tube lining sample, after 24 hours .................................................................................... 97
Figure 6.17 FTIR spectrum of the particles formed in the SOM in vitro crystallization
experiment, for the control sample ............................................................................................... 98
Figure 6.18 A) FTIR spectrum of the particles formed in the SOM in vitro crystallization
experiment, for the 48 hour SOM sample ..................................................................................... 99
Figure 6.19 SEM images and EDX spectra of the crystal products of the in vitro SOM
crystallization experiments ......................................................................................................... 100
Figure 6.20 Force profile obtained from nanoindentation on the adhesive material surface,
performed in artificial seawater .................................................................................................. 102
Figure 6.21 Summary of tube structure and composition .......................................................... 110
xviii
Figure 7.1 Typical FTIR spectrum of barnacle glue on CaF2 substrate. .................................... 120
Figure 7.2 FTIR spectra of the bulk cement from A. amphitrite. .............................................. 125
Figure 7.3 AFM topographic images of the barnacle cement. ................................................... 126
Figure 7.4 Nanoscale morphology of the bulk barnacle cement................................................ 127
Figure 7.5 Elastic modulus distribution of individual nanostructures/components observed by
AFM ............................................................................................................................................ 128
Figure 7.6 SEM images and EDX spectra of the barnacle cement resettled on aluminum foil . 129
Figure 7.7 Chemical staining images of the barnacle cement with amyloid-selective dyes ...... 130
1
1 Introduction
1.1 Nanomaterials: Materials Revolution, Natural Evolution
“Nano” has become a familiar and often used buzzword in the arenas of science and engineering.
Typically, a nanomaterial possesses features on a 1 – 100 nm length scale which may result in
properties that differ from the bulk or macroscopic form of the material. Properties of
nanomaterials are often tunable based on the size of the relevant feature. A popular example of
this is quantum dots, which are semiconducting nanoparticles small enough that quantum
confinement effects dominate the electronic properties of the particles. Whereas the band gap of
bulk semiconductors is determined by the material’s crystal structure and chemical composition,
the bandgap of quantum dots is related to their diameter. As a consequence, nanoparticles with
the same composition but different sizes can emit different wavelengths of light, with smaller
particles emitting shorter wavelengths, and larger particles emitting longer wavelengths.1
Nanomaterials such as quantum dots, fullerenes, and carbon nanotubes are relatively new
discoveries (circa mid 1980 to early 1990’s),2-4
and their unique structures and properties have
inspired a materials revolution in which researchers are designing and exploiting materials in a
fundamentally different way. On the other hand, there are many examples of remarkable
nanomaterials in the natural world that have been around for more than a billion years.
Organisms have evolved a variety of functional nanomaterials, examples which include the
superhydrophic self-cleaning surfaces of lotus leaves, exceptionally strong spider silk, and
adhesive spatulae structures on gecko toes.5-7
Discovery and study of such natural nanomaterials
has led to the design of biomimetic materials. For instance, gecko-inspired dry adhesives have
been fabricated out of polymer nanopillars and carbon nanotubes.8-10
Nanoscale systems must be studied with a different set of considerations than macroscale
systems. The properties of the surface become increasingly dominant as the surface-to-volume
ratio increases with decreasing particle size. In solids, surface atoms exist in a different
environment than bulk atoms. The symmetry of the bulk crystal is broken, and surface atoms
can have dangling bonds, functional groups, and interactions with adsorbed species. As a result,
when a significant proportion of the atoms in a particle are surface atoms, the properties of the
2
particle will deviate from bulk material properties. Some typical changes include a decrease in
melting temperature and an increase in reactivity.11
1.2 Nanoscale Characterization Methods
The development of instrumentation capable of characterizing the structure and properties of
materials at the nanoscale is responsible for the explosion of nanotechnology research over the
past few decades. Atomic force microscopy (AFM) is an important and versatile tool for the
study of surfaces at the nanoscale, in ideal cases demonstrating a lateral resolution of < 1 nm,
and a vertical resolution of < 1 Å.12
AFM force spectroscopy and nanoindentation techniques
can measure the mechanical properties of materials at the nanoscale. A particular advantage of
AFM characterization is that it is largely non-destructive, and little to no sample preparation is
required. In addition, it is capable of operating in a variety of conditions, such as in vacuum,
ambient conditions in air, and in liquids.
Electron microscopy, including scanning electron microscopy (SEM) and transmission electron
microscopy (TEM) remain standard tools for the characterization of nanomaterials, especially
when combined with integrated techniques such as energy dispersive x-ray spectroscopy, and
electron diffraction.
1.3 Hexagonal Boron Nitride Nanomaterials
The two main forms of boron nitride are cubic boron nitride (c-BN) and hexagonal boron nitride
(h-BN). These allotropes have the same basic structure as the well-known carbon allotropes
diamond (cubic form) and graphite (hexagonal form). As a result, the mechanical properties of
C and BN materials are similar – c-BN is super hard, almost as hard as diamond, and h-BN is
soft and lubricious, like graphite.13
The crystal structure of h-BN is shown in Figure 1.1 below,
along with the crystal structure of graphite as a comparison. It can be seen that the basal plane
stacking differs between h-BN and graphite, with B atoms in one layer stacking on top of N
atoms in the layer below. Stacking differences aside, both materials are anisotropic with strong
covalent bonding within an atomic plane, and weak van der Waals bonding between planes. A
major difference between h-BN and graphite is their electrical and optical properties. Whereas
3
graphite is a semimetal, h-BN is an electrical insulator and is white in colour. This is a result of
the localization of π-electrons on the nitrogen atoms, due to their higher electronegativity.
Figure 1.1 Structures of h-BN and graphite. From Souche et al.14
Boron nitride materials do not occur naturally, unlike their carbon counterparts. Hexagonal
boron nitride was first synthesized in 1842 by Balmain.15
It is a non-oxide ceramic, and
possesses a variety of useful material properties including high lubricity, good thermal
conductivity, high oxidation resistance and chemical stability, and non-wetting by molten glass
and metals. As a result, h-BN in micro-powder form is an important industrial material and is
used as a solid lubricant, as an additive in cosmetics, and in mould release applications.13, 16, 17
Figure 1.2 Structure of a carbon nanotube and a boron nitride nanotube. From Golberg et al.18
Following the discovery of carbon nanotubes (CNTs),4 boron nitride nanotubes (BNNTs) were
predicted in 1994 by Blase and Rubio, and synthesized in 1995 by Chopra et al.19-21
BNNTs can
be synthesized in various forms, such as single-wall nanotubes,22-24
double-wall nanotubes,25
multiwall nanotubes,26, 27
and bamboo morphology nanotubes.28, 29
Typically, these BNNTs have
diameters ranging from 1 – 200 nm, and lengths greater than 10 μm; recently, synthesis of
4
BNNTs with lengths greater than 1 mm has been reported.30
Like its parent material, h-BN,
BNNTs exhibit excellent oxidation resistance and thermal conductivity.31, 32
The electronic
properties of BNNTs are largely independent of chirality and diameter,19
which makes them
ready-to-use for many applications since the as-synthesized material does not have to be sorted
as is the case for carbon nanotubes. Taking advantage of their stable wide band gap, polymer
encapsulants loaded with BNNTs have been developed for the electrical insulation of
microelectronics components and optoelectronic devices.33, 34
BNNTs also show promise as
deep-UV emitters.35
Multiwall boron nitride nanotubes (MWBNNTs) have been shown to possess exceptional elastic
properties, similar to those of carbon nanotubes. The Young’s modulus of MWBNNTs has been
experimentally determined to range from several hundred GPa to over 1 TPa.36-39
With their
one-dimensional structure and high Young’s modulus, MWBNNTs are ideal reinforcement
components for plastic, glass, and ceramic composite materials.40-44
1.4 Marine Fouling Organisms: Adhesive Nanomaterials
Biofouling is a major problem for seafaring vessels, particularly those in the shipping industry.
A wide variety of marine organisms including barnacles, oysters, tubeworms, and algae will
attach to and grow on ship hulls. Accumulation of these organisms occurs while the ships are
stationary (in port) and the increased drag of the fouling layer results in significant extra fuel
consumption or slower speeds by the ships once they are underway.45
In addition, biofouling
leads to the introduction of foreign and potentially invasive species to non-native ecosystems, as
the fouling organisms are transported from port to distant port. Toxic tributyl tin-based biocidal
paints which prevent organism settlement are no longer viable anti-fouling solutions due to
environmental concerns.46
Researchers are attempting to address the biofouling problem by
developing novel non-toxic coatings that can be applied to ship hulls to repel the settlement of
fouling organisms outright, or enable the release of the organisms under the hydrodynamic forces
generated when vessels get up to cruising speed. In order to design these novel coatings, it is
necessary to have an understanding of the adhesion mechanisms that fouling organisms employ
to firmly attach themselves to surfaces.
5
Bivalves and barnacles use proteinaceous adhesives, and in some species the proteins responsible
for attachment strength have been identified.47
Oysters and serpulid tubeworms, on the other
hand, employ a cement-like layer comprised largely of calcium carbonate.48
It is becoming
apparent that the mechanical properties of these adhesive materials are often a result of
nanostructured components, by natural design. The study of these adhesive materials at the
nanoscale is therefore an active area of research, which may lead to new solutions to the
biofouling problem.
1.5 Summary of Thesis
Chapter 2 details the imaging and nanomechanical characterization capabilities of the atomic
force microscope, and the model used for analysis of indentation measurements. Chapter 3
reviews synthesis methods for boron nitride nanotubes. Two chemical vapour deposition based
methods which were utilized during the course of this thesis are described in detail. The
experimental methods are presented, together with the morphology of the synthesized nanotubes
as determined via electron microscopy. Possible growth mechanisms for the formation of boron
nitride nanotubes are discussed. In Chapter 4, the characterization of the resulting products via
FTIR, SEM, TEM, EDX is presented. In Chapter 5, the mechanical properties of individual
boron nitride nanotubes are investigated via AFM bending experiments. Euler and Timoshenko
beam models are used to calculate the bending modulus of individual tubes. A comprehensive
study of the calcareous shell of the tubeworm Hydroides dianthus in a biomineralization context
is presented in Chapter 6. Lastly, in Chapter 7, the nanoscale structures in barnacle cement are
investigated via FTIR, EDX, SEM, and AFM.
1.6 References
1. Alivisatos, A. P., Semiconductor clusters, nanocrystals, and quantum dots. Science 1996,
271, 933–937 .
2. Brus, L., Quantum crystallites and nonlinear optics. Applied Physics A: Materials Science
& Processing 1991, 53, 465–474.
3. Kroto, H. W.; Heath, J. R.; O'Brien, S. C.; Curl, R. F.; Smalley, R. E., C60:
Buckminsterfullerene. Nature 1985, 318, 162–163.
4. Iijima, S., Helical microtubules of graphitic carbon. Nature 1991, 354, 56–58.
6
5. Barthlott, W.; Neinhuis, C., Purity of the sacred lotus, or escape from contamination in
biological surfaces. Planta 1997, 202, 1–8.
6. Giesa, T.; Arslan, M.; Pugno, N. M.; Buehler, M. J., Nanoconfinement of spider silk
fibrils begets superior strength, extensibility, and toughness. Nano Lett. 2011, 11, 5038–
5046.
7. Autumn, K.; Liang, Y. A.; Hsieh, S. T.; Zesch, W.; Chan, W. P.; Kenny, T. W.; Fearing,
R.; Full, R. J., Adhesive force of a single gecko foot-hair. Nature 2000, 405, 681–685.
8. Lee, H.; Lee, B. P.; Messersmith, P. B., A reversible wet/dry adhesive inspired by
mussels and geckos. Nature 2007, 448, 338–341.
9. Yurdumakan, B.; Raravikar, N. R.; Ajayan, P. M.; Dhinojwala, A., Synthetic gecko foot-
hairs from multiwalled carbon nanotubes. Chem. Commun. 2005, 3799–3801.
10. Sethi, S.; Ge, L.; Ci, L.; Ajayan, P. M.; Dhinojwala, A., Gecko-inspired carbon nanotube-
based self-cleaning adhesives. Nano. Lett. 2008, 8, 822–825.
11. Shu, Q.; Yang, Y.; Zhai, Y.; Sun, D.; Xiang, H.; Gong, X.-g., Size-dependent melting
behavior of iron nanoparticles by replica exchange molecular dynamics. Nanoscale 2012.
12. Bhushan, B.; Marti, O., Atomic Force Microscope. In Springer Handbook of
Nanotechnology, Bhushan, B., Ed. Springer - Verlag: Berlin, 2004; pp 331–346.
13. Haubner, R.; Wilhelm, M.; Weissenbacher, R.; Lux, B., Boron Nitrides - Properties,
Synthesis and Applications. In High Performance Non-Oxide Ceramics II, Jansen, M.,
Ed. Springer Verlag: Berlin, Heidelberg, 2002; pp 1–45.
14. Souche, C.; Jouffrey, B.; Hug, G.; Nelhiebel, M., Orientation sensitive EELS-analysis of
boron nitride nanometric hollow spheres. Micron 1998, 29, 419–424.
15. Balmain, W. H., Bemerkungen uber die bildung von verbindungen des bors und siliciums
mit stickstoff und gewissen metallen. J. Prakt. Chem. 1842, 27, 422–430.
16. Greim, J.; Schwetz, K. A., Boron Carbide, Boron Nitride, and Metal Borides. In
Ullmann's Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag GmbH & Co.
KGaA: 2000; pp 219–234.
17. Lipp, A.; Schwetz, K. A.; Hunold, K., Hexagonal boron nitride: Fabrication, properties
and applications. J. Eur. Ceram. Soc. 1989, 5, 3–9.
18. Golberg, D.; Bando, Y.; Tang, C. C.; Zhi, C. Y., Boron nitride nanotubes. Adv. Mater.
2007, 19, 2413–2432.
19. Blase, X.; Rubio, A.; Louie, S. G.; Cohen, M. L., Stability and band gap constancy of
boron nitride nanotubes. Europhys. Lett. 1994, 28, 335–340.
7
20. Rubio, A.; Corkill, J. L.; Cohen, M. L., Theory of graphitic boron-nitride nanotubes.
Phys. Rev. B 1994, 49, 5081–5084.
21. Chopra, N. G.; Luyken, R. J.; Cherrey, K.; Crespi, V. H.; Cohen, M. L.; Louie, S. G.;
Zettl, A., Boron-nitride nanotubes. Science 1995, 269, 966–967.
22. Lee, R. S.; Gavillet, J.; Chapelle, M. L. d. l.; Loiseau, A.; Cochon, J. L.; Pigache, D.;
Thibault, J.; Willaime, F., Catalyst-free synthesis of boron nitride single-wall nanotubes
with a preferred zig-zag configuration. Phys. Rev. B 2001, 64, 121405.
23. Smith, M. W.; Jordan, K. C.; Park, C.; Kim, J.-W.; Lillehei, P. T.; Crooks, R.; Harrison,
J. S., Very long single- and few-walled boron nitride nanotubes via the pressurized
vapor/condenser method. Nanotechnology 2009, 20, 505604.
24. Loiseau, A.; Willaime, F.; Demoncy, N.; Hug, G.; Pascard, H., Boron nitride nanotubes
with reduced numbers of layers synthesized by arc discharge. Phys. Rev. Lett. 1996, 76,
4737–4740.
25. Cumings, J.; Zettl, A., Mass-production of boron nitride double-wall nanotubes and
nanococoons. Chem. Phys. Lett. 2000, 316, 211–216.
26. Zhi, C.; Bando, Y.; Tan, C.; Golberg, D., Effective precursor for high yield synthesis of
pure BN nanotubes. Solid State Commun. 2005, 135, 67–70.
27. Lee, C. H.; Wang, J. S.; Kayatsha, V. K.; Huang, J. Y.; Yap, Y. K., Effective growth of
boron nitride nanotubes by thermal chemical vapor deposition. Nanotechnology 2008, 19,
455605–5.
28. Ma, R.; Bando, Y.; Sato, T., Bamboo‐like boron nitride nanotubes. J. Electron Microsc.
2002, 51, S259–S263.
29. Zhang, L.; Wang, J.; Gu, Y.; Zhao, G.; Qian, Q.; Li, J.; Pan, X.; Zhang, Z., Catalytic
growth of bamboo-like boron nitride nanotubes using self-propagation high temperature
synthesized porous precursor. Mater. Lett. 2012, 67, 17–20.
30. Chen, H.; Chen, Y.; Liu, Y.; Fu, L.; Huang, C.; Llewellyn, D., Over 1.0 mm-long boron
nitride nanotubes. Chem. Phys. Lett. 2008, 463, 130–133.
31. Chen, Y.; Zou, J.; Campbell, S. J.; Caer, G. L., Boron nitride nanotubes: Pronounced
resistance to oxidation. Appl. Phys. Lett. 2004, 84, 2430–2432.
32. Bando, Y.; Golberg, D.; Tang, C.; Zhang, J.; Ding, X.; Fan, S.; Liu, C., Thermal
conductivity of nanostructured boron nitride materials. J. Phys. Chem. B 2006, 110,
10354-10357.
33. Zhi, C.; Bando, Y.; Terao, T.; Tang, C.; Kuwahara, H.; Golberg, D., Towards
Thermoconductive, Electrically Insulating Polymeric Composites with Boron Nitride
Nanotubes as Fillers. Adv. Funct. Mater. 2009, 19, 1857–1862.
8
34. Ravichandran, J.; Manoj, A. G.; Liu, J.; Manna, I.; Carroll, D. L., A novel polymer
nanotube composite for photovoltaic packaging applications. Nanotechnology 2008, 19,
085712–5.
35. Li, L. H.; Chen, Y.; Lin, M. Y.; Glushenkov, A. M.; Cheng, B. M.; Yu, J., Single deep
ultraviolet light emission from boron nitride nanotube film. Appl. Phys. Lett. 2010, 97,
141104–3.
36. Suryavanshi, A. P.; Yu, M. F.; Wen, J. G.; Tang, C. C.; Bando, Y., Elastic modulus and
resonance behavior of boron nitride nanotubes. Appl. Phys. Lett. 2004, 84, 2527–2529.
37. Golberg, D.; Costa, P. M. F. J.; Lourie, O.; Mitome, M.; Bai, X.; Kurashima, K.; Zhi, C.;
Tang, C.; Bando, Y., Direct force measurements and kinking under elastic deformation of
individual multiwalled boron nitride nanotubes. Nano Lett. 2007, 7, 2146–2151.
38. Chopra, N. G.; Zettl, A., Measurement of the elastic modulus of a multi-wall boron
nitride nanotube. Solid State Commun. 1998, 105, 297–300.
39. Ghassemi, H. M.; Lee, C. H.; Yap, Y. K.; Yassar, R. S., Real-time fracture detection of
individual boron nitride nanotubes in severe cyclic deformation processes. J. Appl. Phys.
2010, 108, 024314–4.
40. Zhi, C. Y.; Bando, Y.; Wang, W. L. L.; Tang, C. C. C.; Kuwahara, H.; Golberg, D.,
mechanical and thermal properties of polymethyl methacrylate-BN nanotube composites.
J Nanomater. 2008, 642036–5.
41. Lahiri, D.; Rouzaud, F.; Richard, T.; Keshri, A. K.; Bakshi, S. R.; Kos, L.; Agarwal, A.,
Boron nitride nanotube reinforced polylactide-polycaprolactone copolymer composite:
Mechanical properties and cytocompatibility with osteoblasts and macrophages in vitro.
Acta Biomater. 2010, 6, 3524–3533.
42. Bando, Y.; Golberg, D.; Tang, C.; Terao, T.; Zhi, C. Y., Dielectric and thermal properties
of epoxy/boron nitride nanotube composites. Pure Appl. Chem. 2010, 82, 2175.
43. Choi, S. R.; Bansal, N. P.; Garg, A., Mechanical and microstructural characterization of
boron nitride nanotubes-reinforced SOFC seal glass composite. Materials Mater. Sci.
Eng., A 2007, 460–461, 509–515.
44. Qing, H.; Yoshio, B.; Xin, X.; Toshiyuki, N.; Chunyi, Z.; Chengchun, T.; Fangfang, X.;
Lian, G.; Dmitri, G., Enhancing superplasticity of engineering ceramics by introducing
BN nanotubes. Nanotechnology 2007, 18, 485706.
45. Schultz, M. P., Effects of coating roughness and biofouling on ship resistance and
powering. Biofouling 2007, 23, 331–341.
46. Champ, M. A., Economic and environmental impacts on ports and harbors from the
convention to ban harmful marine anti-fouling systems. Mar. Pollut. Bull. 2003, 46, 935–
940.
9
47. Kamino, K., Underwater adhesive of marine organisms as the vital link between
biological science and material science. Mar. Biotechnol. 2008, 10, 111–121.
48. Yamaguchi, K., Shell structure and behaviour related to cementation in oysters. Mar.
Biol. 1994, 118, 89–100.
Figure 2. sdf
Table 2. sdf
Table 3. srdfg
10
2 Atomic Force Microscopy and Spectroscopy
2.1 Introduction
The scanning tunneling microscope (STM) is credited with being the instrument that brought
about the nano-revolution because it was the first instrument capable of imaging molecules and
atoms. It was invented by Binnig and Rohrer, with the first results published in 1982.1-3
The
impact and potential of the instrument on diverse fields of science earned the inventors the Nobel
Prize just a few years later, in 1986. STM is a scanned probe technique in which a sharp metallic
probe is brought into close proximity with a conductive sample surface, and a bias is applied
between the probe and the sample. Electron tunneling occurs between the tip and the sample,
producing a small tunneling current. The topography of the sample can be traced by using the
tunneling current as a feedback parameter. As the tip raster scans across the surface (moving in
the x and y directions), the height of the tip can be raised or lowered (in the z direction) such that
the tunneling current is kept constant, thus forming a topographic image of the sample surface.
The minute translations of the tip in the x, y, and z directions are achieved with piezoelectric
transducers. Following the invention of the STM, Binnig, together with Quate and Gerber, went
on to develop another scanned probe instrument with an applicability that was not restricted to
conductive samples, the atomic force microscope (AFM).4
In AFM, a variety of forces can be probed, including van der Waals, electrostatic, interatomic,
capillary, adhesion, and mechanical contact forces. Forces (as small as 10-18
N) experienced by a
sharp probe mounted on a cantilever cause the cantilever to deflect (by as little as 10-4
Å). This
deflection is used as the feedback parameter, in contrast to the tunneling current in STM. In the
first prototype of the instrument, an STM was used to monitor the minute deflections of the
cantilever.4 The first atomic resolution image obtained was of boron nitride, an insulator.
5
Subsequently, in 1988, Meyer and Amer developed an optical method for detecting the
deflection of the cantilever. In this method, a low powered laser is reflected from the back of the
cantilever and onto a position sensitive detector.6 Commercial AFMs employ this set-up, which
is shown schematically in Figure 2.1. A photodiode with two or four segments is used as the
position sensitive detector, and voltage differences generated by the position of the laser spot on
the various diode segments are measured to monitor the cantilever deflection.
11
Figure 2.1 Schematic of atomic force microscope.7
The AFM is a very versatile instrument, capable of imaging in vacuum, ambient conditions in
air, or in liquid environments. In this chapter, the imaging modes of the AFM will be described.
The AFM force spectroscopy technique and its application to the nanomechanical
characterization of materials will also be presented.
2.2 Contact Mode Imaging
Contact mode imaging involves the raster scanning of the AFM tip while the tip is in contact
with the sample surface. The attractive and repulsive forces experienced by the tip cause the
cantilever beam to deflect, and this deflection is measured via the optical method described
above. The deflection is used as a feedback parameter in order to control the height of the tip (z
direction) such that a constant force is maintained between the tip and sample. In this manner,
the topography of the surface can be traced. The force is calculated from Hooke’s Law, F = k
∙Δx, in which F is the force, k is the spring constant of the cantilever, and Δx is the deflection of
the cantilever.
12
Low spring constant cantilevers are often used in contact mode imaging in order to increase the
sensitivity, maximizing the deflection of the cantilever. In addition to mapping surface
topography, the lateral force experienced by the cantilever can be used to map frictional forces
when a scan direction perpendicular to the cantilever axis is used.
When imaging samples in air, there is a thin water layer coating the sample surface. This layer
can influence contact mode imaging due to capillary forces. One strategy to minimize this effect
is to image the sample in water (or other liquid). However, depending on the sample of interest,
this is not always appropriate. Another problem encountered with contact mode imaging is tip
contamination and sample damage, which commonly occur when imaging soft samples such as
polymers, Langmuir Blodgett films, and lipid bilayers.
2.3 Intermittent Contact (Tapping) Mode Imaging
In order to address the issue of sample damage in contact mode imaging, an intermittent contact
mode (TappingMode, a trademark of Digital Instruments) was developed in which a high spring-
constant cantilever is driven to oscillate near its resonance frequency (~300 kHz), with an
oscillation amplitude of 20 – 100 nm.8 The tip probes the surface with each oscillation and, as a
result, the contact and lateral forces between the tip and sample is minimized. Interaction
between the tip and sample causes changes to the oscillation amplitude, phase, and frequency of
the cantilever. In tapping mode, the height of the tip is adjusted via feedback such that the
oscillation amplitude remains constant, in order to obtain a topographic image of the surface.
The phase lag between the oscillation driving the cantilever oscillation and the actual cantilever
oscillation can also be monitored to generate a phase image. Phase changes occur as a result of
energy dissipation, and are material dependent. This allows for contrast between materials with
different adhesion, elastic, and viscoelastic properties. Phase imaging is particularly useful when
there is little height variation between two distinct sample materials, as in the case of lipid
bilayers and block copolymer films, for example. In contrast to the cantilevers used in contact
mode imaging, stiffer cantilevers (~40 N/m) are used for tapping mode.9 In Chapter 5,
topographic images of boron nitride nanotubes on a patterned Si substrate are acquired using
tapping mode.
13
2.4 Force Spectroscopy
Force curves, which plot the force versus the distance between the tip and sample, can be
obtained from a given location on a sample. The AFM can be used to stretch single molecules,
such as polymer chains or proteins, between the sample and the tip. In this case, the force is
plotted against the tip-sample separation distance as the tip is withdrawn from the surface.
Phenomena such as single chain polymer elongation and hydrophobic hydration, and protein
domain unfolding can be studied with this technique.10-13
The AFM tip can also be used to apply a load to a point on the sample, in which the tip is
brought down to the sample surface and pressed a small distance into it. The force is plotted
against the tip-sample separation distance as the tip is brought down to, and into, the surface.
This nanoindentation technique can be used to determine the Young’s modulus of a sample.14, 15
Force curves are analyzed with the Sneddon-Hertz model, which describes loading vs.
indentation depth when a paraboloidal object (the AFM tip) contacts an elastic planar film of
infinite thickness (the sample). Non-adhesive contact between the two materials is assumed. In
the expression given below in Equation 1.1, F is the loading force [N], E is Young’s modulus
[Pa], R is the radius of curvature of the tip [m], δ is the indentation [m], and ν is the Poisson’s
ratio.16, 17
Equation 1.1
2/3
2 )1(3
4
v
REF
AFM nanoindentation is used in Chapters 6 and 7 to study the mechanical properties of
tubeworm and barnacle adhesive materials.15, 18
A force indentation curve is collected from a single point on the sample. In order to correlate
mechanical properties with topographic features, a topographic image may be divided into
pixels, and a force curve can be collected from each pixel. This process is known as force
mapping. In Chapter 5, force mapping is employed in order to correlate the location of the
applied force on a boron nitride nanotube suspended across a trench. This allows for the
boundary conditions of the nanotube beam to be determined, allowing for a more accurate
determination of the bending modulus using beam mechanics analysis.
14
2.5 References
1. Binnig, G.; Rohrer, H.; Gerber, C.; Weibel, E., Tunneling through a controllable vacuum
gap. Appl. Phys. Lett. 1982, 40, 178–180.
2. Binnig, G.; Rohrer, H.; Gerber, C.; Weibel, E., Surface studies by scanning tunneling
microscopy. Phys. Rev. Lett. 1982, 49, 57–61.
3. Binnig, G.; Rohrer, H., Scanning tunneling microscopy. Surf. Sci. 1983, 126, 236–244.
4. Binnig, G.; Quate, C. F.; Gerber, C., Atomic force microscope. Phys. Rev. Lett. 1986, 56,
930–933.
5. Albrecht, T. R.; Quate, C. F., Atomic resolution imaging of a nonconductor by atomic
force microscopy. J. Appl. Phys. 1987, 62, 2599–2602.
6. Meyer, G.; Amer, N. M., Novel optical approach to atomic force microscopy. Appl. Phys.
Lett. 1988, 53, 1045–1047.
7. OverlordQ Atomic force microscope block diagram.
http://en.wikipedia.org/wiki/File:Atomic_force_microscope_block_diagram.svg
(accessed August 8th, 2012).
8. Zhong, Q.; Inniss, D.; Kjoller, K.; Elings, V. B., Fractured polymer/silica fiber surface
studied by tapping mode atomic force microscopy. Surf. Sci. Lett.1993, 290, L688–L692.
9. arc a, R.; Pérez, R., Dynamic atomic force microscopy methods. Surf. Sci. Rep. 2002,
47, 197–301.
10. Bemis, J. E.; Akhremitchev, B. B.; Walker, G. C., Single polymer chain elongation by
atomic force microscopy. Langmuir 1999, 15, 2799–2805.
11. Li, I. T. S.; Walker, G. C., Effect Of temperature on the mechanical properties of
fibronectin. Biophys. J. 2009, 96, 641a.
12. Meadows, P. Y.; Bemis, J. E.; Walker, G. C., Single-molecule force spectroscopy of
isolated and aggregated fibronectin proteins on negatively charged surfaces in aqueous
liquids. Langmuir 2003, 19, 9566–9572.
13. Shi, W.; Walker, G., Mechanical desorption of single fibronectin type III module from
hydrophilic and hydrophobic surfaces. Biophys. J. 2011, 100, 480a.
14. Sun, Y.; Guo, S.; Walker, G. C.; Kavanagh, C. J.; Swain, G. W., Surface elastic modulus
of barnacle adhesive and release characteristics from silicone surfaces. Biofouling 2004,
20, 279–289.
15. Sullan, R. M. A.; Gunari, N.; Tanur, A. E.; Chan, Y.; Dickinson, G. H.; Orihuela, B.;
Rittschof, D.; Walker, G. C., Nanoscale structures and mechanics of barnacle cement.
Biofouling 2009, 25, 263–275.
16. Hertz, H., Über die berührung fester elastischer körper (On the contact of elastic solids).
J. Reine Angew. Mathematik. 1881, 92, 156–171.
17. Sneddon, I. N., The relation between load and penetration in the axisymmetric boussinesq
problem for a punch of arbitrary profile. Int. J. Eng. Sci.1965, 3, 47–57.
15
18. Tanur, A. E.; Gunari, N.; Sullan, R. M. A.; Kavanagh, C. J.; Walker, G. C., Insights into
the composition, morphology, and formation of the calcareous shell of the serpulid
Hydroides dianthus. J. Struct. Biol. 2010, 169, 145–160.
Figure 3. sadf
Equation 2 rtr
Equation 3 sdf
16
3 Synthesis of Boron Nitride Nanotubes
3.1 Permissions
The experimental material in this chapter is presented with permission from Reddy, A. L. M.;
Tanur, A. E.; Walker, G. C. Synthesis and hydrogen storage properties of different types of
boron nitride nanostructures. Int. J. Hydrogen Energy 2010, 35, 4138-4143.
3.2 Abstract
The major synthesis techniques for boron nitride nanotubes are reviewed in this chapter. Two
methods utilized over the course of this thesis are presented in detail for the synthesis of
multiwall boron nitride nanotubes (MWBNNTs), based on chemical vapour deposition (CVD).
The first method involves the mechanochemical activation of precursor powders prior to
annealing at temperatures of 1050 – 1200 oC in an NH3 atmosphere. The second method
involves an apparatus to trap the growth vapours of the precursor powders during annealing at
1200 oC in an NH3 atmosphere. Possible growth mechanisms of BNNTs are discussed.
3.3 Introduction
Despite the interest in the unique properties of boron nitride nanotubes and their similarity to
carbon nanotubes, there is much less literature published on BNNTs. A large part of this is due
to the difficulty in synthesizing high-quality BNNTs in sufficiently large quantities. As a
refractory ceramic, the commercial production of h-BN requires extremely high temperatures,
from 1000 – 5500 oC, depending on the method used.
1 Given this, it is no wonder that a major
focus of research is to achieve synthesis of BNNTs at reasonably low temperatures (i.e. < 1500
oC). The main categories of BNNT synthesis methods are summarized below.
3.3.1 Arc Discharge
Arc discharge was the first reported method for the synthesis of carbon nanotubes (CNTs).2 In
this method, two conducting electrodes are placed close together, with a gap of 1-2 mm
separating them. A DC current is applied, resulting in a potential difference between the
electrodes. At a certain threshold current, electrical arcing occurs between the two electrodes,
producing a plasma discharge. Material is vaporized from the end of the anode. The electrode
17
material is chosen based on the desired synthesis product. For carbon nanotubes, graphite
electrodes are used. After arcing, product can be found deposited as soot on the cathode, as well
as on the walls of the arc discharge chamber (different product species may be found in each
location). The quality and quantity of product synthesized is dependent on a number of
parameters, including the type and pressure (sub-atmospheric) of inert gas in the discharge
chamber, the current and voltage, the plasma temperature, the composition of the electrodes, and
the geometry of the apparatus.3
Due to its electrically insulating nature, pure h-BN cannot be used on its own as the electrodes in
arc discharge synthesis. Boron nitride nanotubes were first synthesized by Chopra et al.4 in 1995
by an arc discharge method. Currents from 50 – 140 A were applied to maintain a voltage of 30
V between the anode, an h-BN filled W rod, and the cathode, a cooled Cu electrode. The arcing
took place within a He gas atmosphere, and the temperature at the anode was in excess of the
melting point of tungsten (> 3400 oC). The method produces MWBNNTs with diameters < 10
nm, and lengths > 200 nm. Loiseau et al. reported the synthesis of few to single-wall BNNTs
using a similar arc discharge method, using HfB2 electrodes in a N2 atmosphere.5 Cumings and
Zettl produced bulk quantities of nearly monodisperse double-wall BNNTs by using electrodes
with a low metal content. Elemental B powder was mixed together with 1 at% of Ni and Co.
The powder was melted to form ingots, which were used as the electrodes. Arcing was
performed in a N2 environment.6
3.3.2 Laser Heating/Ablation
In laser heating or ablation methods, a laser is used to heat a target material to very high
temperatures (> 4000 oC). Continuous lasers heat the target and the synthesis product is
collected from the target surface. Pulsed lasers will ablate the target, and synthesis product is
transported away from the target area by a carrier gas to a collecting substrate downstream.7
In 1996, Golberg et al. used a continuous high powered CO2 laser (up to 240 W) to irradiate a
target of single crystal c-BN powder within a diamond anvil cell in a N2 environment. The area
of the target on which the laser was focused either melts as c-BN, or undergoes a phase change
to h-BN before melting, depending on the N2 pressure, which varied from 5 – 15 GPa. Short (<
30 nm) MWBNNTs with circular and polygonal cross sections 3 – 15 nm in diameter were
18
observed on both melted c-BN flakes as well as recrystallized h-BN.8 The synthesis of single to
few-wall BNNTs was achieved by Yu et al. in 1998, using an oven-laser ablation method. A
pulsed excimer laser was used to irradiate a hot pressed target composed of h-BN powder mixed
with 1 at% each of Ni and Co nanopowders. Prior to irradiation, the target was heated within a
tube furnace to 1200 oC under a flow of various carrier gases (e.g., Ar, N2, He). The type of
carrier gas was found to affect the number of walls of the BNNTs –– the use of Ar and He
resulted in predominantly SWBNNTs, and the use of N2 resulted in mainly double-wall BNNTs.
The diameters of the tubes ranged from 1.5 nm to 8 nm, and the lengths were < 100 nm.9
Following the first report by Golberg et al., continuous CO2 laser heating was used to synthesize
long ropes of BNNTs (~ 40 μm). Laude et al. used a hot pressed h-BN powder target, and heated
it with a 70 W CO2 laser for 3 min in a N2 atmosphere. BNNTs with two to four walls were
produced, with up to several tens of tubes bundled into ropes.7
Gram quantities of BNNTs were synthesized by Lee et al. using a laser ablation set-up that could
be operated continuously.10
This method employed a 1 kW CO2 laser and a N2 atmosphere. A
catalyst-free BN target was moved over the course of the synthesis to expose new areas for
continuous ablation. The yield of BNNTs was estimated to be 0.6 g/h, and consisted of mostly
SWBNNTs. HRTEM characterization revealed that the chirality of the BNNTs was
predominantly zig-zag (85%).10
A pressurized vapour/condenser (PVC) method was developed by Smith et al.11
In a pressurized
N2 environment (2 -20 atm), boron vapour is produced through the laser heating of a B-
containing target, and a condenser wire is placed within the plume of B vapour in order to
generate B droplets in its wake. The droplets encounter N2 near the shear layer of the B plume,
and BNNTs form and assemble into fibrils. The process is versatile, and was found to be
effective with a variety of targets (hot and cold pressed BN, amorphous B powder, cast B) and
condenser materials (BN, B, stainless steel, Cu, Nb, W). The BNNT fibrils produced have
diameters of ~1 mm, and are composed of individual tubes > 100 μm long and < 10 nm in
diameter (SWBNNTs, DWBNNTs, and few-wall BNNTs). In a 200 mg run, a fibril mass 15 cm
long and approximately 1.5 cm wide was produced, with an appearance similar to cotton balls.
The raw material could be processed into yarn by twisting.
19
Typically, laser heating and ablation methods are high temperature methods. However, using a
plasma enhanced pulsed laser deposition (PLD) technique, Wang et al.12
were able to synthesize
BNNTs directly on a substrate at a temperature of 600 oC. Oxidized Si substrates were coated
with a thin Fe film, and installed on a heater within the deposition chamber. The chamber was
evacuated to a high vacuum, and then backfilled with N2. After heating the substrate to 600 oC,
an RF generator generated plasma on the substrate surface, which induced a negative DC voltage
on the substrate such that positive ions were accelerated towards the substrate. After 10 min of
plasma treatment at 600 oC, the thin Fe film broke down into nano-sized particles. A pulsed
Nd:YAG laser was then used to irradiate a rotating h-BN target situated above the substrate, and
the ablated vapour was propelled towards the substrate. At substrate biases of -360 to -450 V,
BNNTs were synthesized, and observed to grow from the Fe nanoparticles on the substrate. The
BNNTs were multiwalled and had diameters of 20 nm and less. Multiple tubes from closely
spaced Fe particles were found to form vertical bundles.12
3.3.3 Templated Synthesis
BNNTs can also be produced by using other nanostructures as templates. Han et al.13
used
CNTs to synthesize bulk quantities of BNNTs via a substitution reaction (Equation 3.1) in which
the C atoms were substituted by B and N atoms.
Equation 3.1 B2O3 + 3C (nanotubes) + N2 2 BN (nanotubes) + 3CO
A crucible containing a layer of MWCNTs (~10 nm diameter) over B2O3 powder was heated to
1500 oC under a flow of N2. The resulting BNNTs had similar dimensions to the starting CNTs,
but displayed a higher crystallinity.
Another type of one-dimensional nanostructure that has been used to template BNNT growth
consists of SiC nanowires. Zhong et al.14
demonstrated that the surface of SiC nanowires can be
coated by a liquid B-N containing polymer, produced from the decomposition of ammonia
borane at 1450 oC in an N2 environment. The decomposition products of ammonia borane
include H2, BH2NH2 (monomeric aminoborane), (BHNH)3 (borazine), and B2H6 (diborane). As
the products coating the SiC nanowires polymerizes into BN, H2 is evolved and trapped beneath
the film. At 1450 oC, H2 etches SiC, and by the end of the reactions the SiC nanowire is
20
completely etched away, leaving a hollow BN tube behind. These BNNTs had slightly larger
diameters compared with the SiC nanowires (150 nm vs. 100 nm), and possessed a unique
beaded structure in which sections of straight tubular walls are joined together by spheric shells.
The authors hypothesize that surface tension causes the liquid film to bead on the nanowire
surface, resulting in the final beaded morphology.14
Nanoporous materials such as mesoporous silica and anodic aluminum oxide (AAO) have been
used to template BNNT growth by deposition on the inner walls of the pores. The BNNTs can be
subsequently freed from the template using template-specific etchants. One particular advantage
that this membrane-assisted template process has over other techniques is the ability to form
monodisperse vertically aligned nanotubes, with control over the tube diameter, spacing, and
length. This is desirable for many applications, because BNNTs have anisotropic properties such
as thermoconductivity and Young’s modulus, both of which are greater in the direction of the
tube axis.15, 16
Therefore, to exploit these properties to their fullest, alignment of the nanotubes
in devices and composite materials is necessary. Li et al.17
used a chemical vapour deposition
(CVD) technique by reacting BCl3 and NH3 gases which formed BN via the reaction in Equation
3.2:
Equation 3.2 BCl3 (g) + NH3 (g) BN (s) + HCl (g)
The silica template was dissolved by treatment with 5% HF. The pore size of the mesoporous
silica used was ~7 nm, and TEM analysis of the freed BNNTs revealed that they had similar
diameters.
Bechelany et al.18
synthesized vertically aligned supported MWBNNT arrays via the infiltration
of AAO with liquid polymeric borazine. After infiltration, the impregnanted AAO samples were
heated to 200 oC to create thermoset polymer films coating the walls of the AAO channels. The
samples were then annealed in N2 at 1200 oC, and the polymer films underwent solid state
thermolysis to produce BN coatings on the channel walls. A thin film of BN covered the
surface of the template, such that the nanotube array was supported by it once freed from the
template by 48% HF. To improve the crystallinity of the BNNTs, samples after HF treatment
were annealed in N2 at temperatures between 1450 – 1800 oC. The AAO membrane used had a
200 nm pore size and pore lengths of ~60 μm, and the MWBNNTs were found to have similar
21
dimensions. Other AAO membranes with different pore diameters and channel length could be
used to control the dimensions of the BNNTs.
3.3.4 Chemical Vapour Deposition
MWBNNTs can be produced with chemical vapour deposition (CVD) based techniques. These
techniques involve the introduction of gaseous precursors into a reaction chamber containing a
heated substrate. Gas-phase reactions occur, forming intermediate species; these species can
form solids in the gas phase which are deposited onto the substrate, where they can become
crystal nucleation centers. Alternatively, gaseous intermediate species can adsorb directly onto
the substrate and heterogeneous reactions can occur at the gas-solid interface, forming the solid
species which become nucleation centers.19
The majority of CVD methods developed for BNNT synthesis involve the use of a metal-based
catalyst, and relatively mild reaction temperatures < 1300 oC. Compared with other techniques
such as arc discharge or laser ablation, the BNNTs produced by CVD are predominantly
multiwalled and larger in diameter (~10 – 100 nm). The first report of CVD synthesis of
BNNTs, by Lourie et al.,20
used borazine gas (B3N3H6) and Co, Ni, NiB, and Ni2B catalyst
particles. The catalyst particles were suspended in ethanol and deposited onto oxidized Si
substrates. The reaction was carried out at 1000 – 1100 oC under N2 flow. It was found that the
NiB and Ni2B catalysts produced the highest yield and quality of MWBNNTs, compared with
Co and Ni. The MWBNNTs grew directly from these catalyst particles, were 10 – 50 nm in
diameter, and possessed lengths up to 5 μm.
In other works, the most commonly used gas precursors for the source of N are N2 and NH3. For
the source of B, amorphous B powders, B thin films, and alloy particles such as Fe-B can be
used.21-24
Tang et al.25
combined B and MgO powders and heated the mixture to 1300 oC in an
induction furnace. At this temperature, the following reaction takes place:
Equation 3.3 2B (s) + 2MgO (s) B2O2 (g) + 2Mg (g)
The B2O2 vapour was then transported via an Ar carrier gas flow into a cooler (1100 oC) reaction
chamber, where NH3 gas was introduced. BNNTs were produced via Equation 3.4.
22
Equation 3.4 B2O2 (g) + 2NH3 (g) BN (s) + 2H2O (g) + H2 (g)
Following the work of Tang et al., a high-yield process for MWBNNTs was developed by Zhi et
al.,26
using a mix of FeO, MgO, and B powders at the precursor material. When heated in an
induction furnace under NH3 flow at temperatures between 1100 – 1700 oC, the MgO precursor
was found to effectively produce B2O2 vapour, while the FeO precursor was reduced to Fe and
served as a catalyst for the growth of BNNTs. These two processes resulted in a high product
yield (~200 mg/h), and the yield was found to increase with increasing temperature. This
method is known as boron oxide chemical vapour deposition (BOCVD).
Transition metals such as Fe, Co, and Ni are typical catalysts, and can be used in nanoparticle
form, or as thin films coated on substrates (which de-wet and form nanoparticles of a
characteristic size, depending on the film thickness, when heated).24, 27, 28
There are several CVD methods which are notable due to their unique precursors, products,
and/or equipment. Kim et al.29
used a floating nickelocene catalyst (solid nickelocene was
heated at 80 oC to form a vapour) and borazine gas. These two gases were reacted in the
presence of N2 and NH3 at 1200 oC, and at low borazine pressures, highly crystalline doublewall
BNNTs were formed (~2 nm diameter). The yield of BNNTs was estimated to be ~20 mg/h,
with 70% being comprised of doublewall BNNTs, with SWBNNTs and few-wall BNNTs
making up the other fraction. Ma et al.30
showed that a metal-based catalyst is not necessary for
the synthesis of BNNTs. They used a melamine diborate precursor (B4N3O2H) powder, which
was heated to 1700 oC in an induction furnace under N2 flow. MWBNNT product was found
downstream, in a zone of the furnace where the temperature was ~1200 oC. A common feature
of the MWBNNTs was the presence of bulbous tips on the nanotube ends. HRTEM revealed
that these tips were BN cages filled with amorphous material. In another study without metal
catalysts, Gan et al.22
grew BNNT films on oxidized Si substrates. The precursors were B films
deposited on the substrates, and NH3 gas. Synthesis was carried out at 1175 oC. It was observed
that growth was preferential on SiO2 as opposed to plain Si, and that SiO2 played a catalytic role
in the growth of BNNTs.22
Films up to 1 mm thick were synthesized, comprised of tubes with
diameters < 100 nm.
23
Assisted CVD methods have been developed in order to achieve low temperature synthesis of
BNNTs (< 1000 oC). A microwave plasma-enhanced CVD method was used by Guo et al.
27, 28
to synthesize MWBNNTs on Ni and Co coated oxidized Si substrates at a temperature of 800 oC.
The gas precursors used were diborane (B2H6, diluted in H2 to 5 vol%), NH3, and H2. Under
optimum process conditions, MWBNNTs several microns long with diameters of 5 – 20 nm
were produced on the substrates. Su et al.31
synthesized MWBNNTs at 900 oC using a home-
built plasma-assisted CVD system. In addition to the plasma, finely dispersed Fe catalyst
particles were prepared on high surface area SiO2/Al2O3 supports, in order to promote nanotube
growth at the lower temperature. B2H6, NH3, O2, and Ar were used as process gases.
3.3.5 Mechano-Thermal (Ball milling and Annealing)
A synthesis method with promise for industrial scale-up is the mechano-thermal method, in
which precursor powders are subjected to ball-milling for extended periods in order to break
them down into nano-sized particles with highly activated surfaces prior to thermal annealing.
High-energy ball milling involves a stainless steel cell into which the powder is loaded along
with several hard balls (usually made out of stainless steel or tungsten carbide). The cell is then
mechanically shaken, such that the powder gets mixed and broken down by a variety of
processes including grinding, fracturing, and plastic deformation. Due to these processes the
surfaces of the particles become very reactive with many dangling bonds, and the particles have
a high density of structural defects such as dislocations, grain boundaries, etc. The disordered
surfaces and defects allow for non-equilibrium chemical reactions to proceed at low
temperatures.
Chen et al. introduced the technique for BNNTs in 1999,32
as an alternative to arc discharge and
laser heating. The milling cell was loaded with B powder and NH3 gas and milled for 150 h.
The gas pressure within the cell was monitored, and was observed to decrease during the first 40
h of milling, and then increase until it stabilized at 100 h. The pressure changes were a signature
of a nitriding reaction between the B and NH3, in which NH3 was first adsorbed onto the B
particles, followed by the reaction shown in Equation 3.5:
Equation 3.5 B + NH3 BN + 3/2 H2
24
The milled sample is then annealed at 1000 oC and higher under N2 flow for 6 h, allowing the
nitriding reaction to proceed to completion. XRD analysis of the products confirms the presence
of h-BN, and also indicates that a Fe-B phase is present. Small amounts (5 wt%) of Fe were
introduced from the stainless steel cell and milling balls, and served as catalysts for the BNNT
synthesis. MWBNNTs with diameters ranging from 20 – 100 nm were produced with this
method, and higher temperatures were found to increase the tube diameters.
Other variations on the mechano-thermal method include the milling and annealing of h-BN
powder in NH3 and N2,33, 34
of B powder in N2,35
and of B2O3 powder in N2 and NH3.36
Milling
times were typically 100 h and longer, and annealing was carried out at temperatures of 1000 –
1300 oC for periods of 6 h and longer. A mixture of BNNT types was produced, consisting of
bamboo and cylindrical MWBNNTs. A yield of 600 mg BNNTs per 1 g milled h-BN powder
was demonstrated by Li et al.36
As further evidence for the catalyst hypothesis, these studies
reported the incorporation of impurities into the milled powders; namely Fe, Cr, and Ni (or W in
the case of WC milling balls and cell), and traces of Si and Al.
3.3.6 Chemical Synthesis
Another method which is promising for industrial scale-up is the self-propagation high-
temperature synthesis (SHS) and annealing process, developed by Wang et al.37
The key
element of this method is the synthesis of a porous precursor via SHS. B2O3, Mg, and CaB6
powders are mixed in a blender, then heated to 750 oC in an Ar atmosphere. These precursors
undergo the reaction described in Equation 3.6 to form the porous precursor, B18Ca2(MgO)9.
Equation 3.6 3B2O3 + 9Mg + 2 CaB6 B18Ca2(MgO)9
The precursor is then annealed in a horizontal tube furnace at 1150 oC under a NH3 flow for 6 h.
Gram quantities of MWBNNTs (20 – 300 nm in diameter) were obtained after synthesis, with a
product yield above 80 wt%. The process was further developed to yield kilogram quantities of
BNNTs with various morphologies (MWBNNTs, bamboo BNNTs), using a B31Fe17(MgO)27
porous precursor synthesized by SHS.38
3.3.7 Comparison of Methods
A summary of the synthesis methods discussed above is presented in Table 3.1.
25
Table 3.1 Comparison of BNNT Synthesis Methods
Synthesis Method Temperature
(oC)
Morphology Tube Diameter
(nm) Yield
Arc Discharge > 3400 SW, DW,
few-wall BNNT < 10
Low4, 5
Moderate6
Laser
Heating/Ablation > 4000
SW, DW,
few-wall BNNT < 10
Low8, 9
Moderate7
High10, 11
CVD/Chemical
Synthesis 800 – 1700
MWBNNT,
bamboo BNNT38
10 – 200
Moderate
High26, 37, 38
Mechano-Thermal 1000 – 1300 MWBNNT,
bamboo BNNT 50 – 200
Moderate
High36
3.4 Experimental Methods
3.4.1 Method 1: Silica-Assisted Catalytic Chemical Vapour Deposition
The first method employed in this thesis and presented here is based on the work of Reddy et al.
(2010).39
Silica-assisted chemical vapour deposition (SA-CVD) is a bulk synthesis method
combining the use of a Fe2O3-SiO2/Al2O3 supported catalyst, ball milling, and CVD.
3.4.1.1 Preparation of the Precursor Powder
In a typical preparation, 3.5 g of Fe(NO3)3∙9H2O, 5 g of fumed Al2O3, and 80 mL of methanol
are combined. The mixture is covered and stirred overnight, following which the solvent is
evaporated, leaving behind a powder. Following this, 0.65 g of the Fe2O3-Al2O3 catalyst, 1 g of
amorphous B powder, and 1 g of fumed SiO2 are combined with 80 mL of methanol. Again, the
mixture is covered and stirred overnight. After solvent evaporation, the precursor powder is
ready for ball milling treatment. The powder is loaded into a stainless steel cell with several 1
cm stainless steel balls, and milled for 20 h.
26
3.4.1.2 CVD Process
The ball milled precursor, if used some time after the milling process, is ground with a mortar
and pestle, then sprinkled evenly into a quartz boat. The boat is then placed in the centre of a
tube furnace, and heated to 1000 – 1150 oC under 80 sccm Ar gas flow. Upon reaching the set
temperature, 50 sccm of NH3 is flowed into the process tube, and the Ar flow is stopped. The
reaction is allowed to proceed under these conditions for 2 h, then the NH3 flow is stopped and
the Ar flow restarted, to protect the products from oxidation while the system cools down to
room temperature.
3.4.1.3 Purification
After synthesis, the quartz boat is filled with a grey powder. In order to remove the SiO2 and
Al2O3 components, the product is washed with 1 M HF, and filtered several times. HCl can be
used to dissolve Fe impurities, while HNO3 can be used to dissolve any B.40
3.4.2 Method 2: Growth Vapour Trapping Chemical Vapour Deposition
The second method employed in this thesis follows the method of Lee et al (2008).24
Growth
vapour trapping chemical vapour deposition (GVT-CVD) is a simple thermal CVD process using
a conventional tube furnace set-up, with the critical components being a vacuum pump and a
quartz test tube which is used to trap the growth vapours evolved during the reactions close to
the nucleation sites.
3.4.2.1 Preparation of the Precursor Powder
Amorphous B, MgO, and Fe2O3 nanopowders (Sigma Aldrich, MO) are combined in a mortar
and pestle in a 4:1:1 molar ratio.
3.4.2.2 CVD Process
In a typical run, 100-150 mg of precursor powder is measured into a 3 mL alumina boat. Si
substrates can be placed over top of the boat to partially cover it. The boat is then placed at the
closed end of a quartz test tube, as shown in Figure 3.1 below.
27
Figure 3.1 Position of alumina boat within quartz test tube.
The quartz test tube is then placed within the quartz process tube of the tube furnace
(ThermoScientific), such that the closed end of the test tube and alumina boat are located within
the middle of the heating element region. The process tube is then sealed, and evacuated to ~25
kPa via a mechanical pump. The vacuum pump is kept running for the duration of the process.
A flow of NH3 gas is started at 200 sccm and kept flowing for the duration of the temperature
ramp-up (~10 oC/min) and dwell time at 1200
oC (1-2 h). At the end of the dwell time, the NH3
flow is shut off, and the system is allowed to cool to room temperature under vacuum. A
schematic of the set-up is shown in Figure 3.2.
Figure 3.2 GVT-CVD set-up.
After the synthesis, a white powder can be found in the alumina boat, covering the remains of the
precursor powder (the reaction does not go to completion in 1-2 h).
3.4.2.3 Purification
Product collected from Si substrates can be used without purification. For the product in the
boat, HCl can be used to dissolve Fe impurities, while HNO3 can be used to dissolve any B.40
3.5 Results
3.5.1 Macroscopic Description of Products
After a typical SA-CVD run, the quartz boat and its contents are as shown in Figure 3.3. A white
powder covers the transparent colourless quartz boat, and a grey powder covers the remnants of
the precursor powders, which were rust-coloured prior to the synthesis and black afterward.
BNNT products were found within the grey powder, and not the white powder.
28
Figure 3.3 Quartz boat after Method 1, silica-assisted CVD (1150 oC, 1 h).
After a typical GVT-CVD run, the alumina boat and its contents are as shown in Figure 3.4. A
white powder covers the remnants of the precursor powders, which were rust-coloured prior to
the synthesis and black afterward. The walls of the white alumina boat remained largely clean of
material. BNNTs were found to be present in the white powder (see Chapter 4 for
characterization details).
Figure 3.4 Alumina boat after Method 2, GVT-CVD (1150 oC, 2 h).
29
3.5.2 Nanotube Morphology
The BN products synthesized via SA-CVD are shown in Figure 3.5 below. Different synthesis
temperatures resulted in different BN nanostructure morphologies.
Figure 3.5 Electron micrographs of the products of SA-CVD. a) TEM image of BN flower structure, synthesized at 1000
oC. b) TEM image of short bamboo BNNTs, synthesized at 1050
oC. c) TEM image of long bamboo BNNTs,
synthesized at 1100 oC. d) SEM image of long bamboo BNNTs. e) TEM image of straight MWBNNT, synthesized at
1150 oC. f) SEM image of MWBNNTs.
30
The BN products synthesized via GVT-CVD are shown in Figure 3.6 below.
Figure 3.6 a) SEM image of MWBNNTs synthesized by GVT-CVD. b) TEM image of MWBNNTs.
3.6 Discussion
3.6.1 BNNT Growth Mechanisms in CVD Synthesis
3.6.1.1 Redox Reactions – Generation of Growth Vapours
At temperatures above 600 oC, a number of redox reactions occur amongst the solid precursor
materials and the NH3 process gas. In SA-CVD and GVT-CVD, B reacts with Fe2O3 to form
B2O2 vapour and metallic Fe, as shown in Equation 3.7.21
B2O2 vapour is also produced from a
reaction of B and SiO2, in the SA-CVD method (Equation 3.8).22
In GVT-CVD, MgO reacts
with B to produce B2O2 vapour, as shown in Equation 3.9.25
Equation 3.7 6B (s) + 2Fe2O3 (s) 3B2O2 (g) + 4Fe (s)
Equation 3.8 2B (s) + SiO2 (s) B2O2 (g) + Si (s)
Equation 3.9 2B (s) + 2MgO (s) B2O2 (g) + 2Mg (g)
BNNTs can be formed from the reaction between B2O2 and NH3 gases, as described by Equation
3.10:
Equation 3.10 B2O2 (g) + 2NH3 (g) 2BN (s) + 2H2O (g) + H2 (g)
31
However, the dominant growth mechanism of BNNTs in the CVD synthesis methods explored in
this work involves the Fe nanoparticles which are formed after reduction of the Fe2O3 precursor
powder. As in the case of CNTs, in which transition metals including Fe have been
demonstrated to catalyze nanotube growth in CVD methods,41
the Fe nanoparticles catalyze
BNNT growth by serving as a vessel for the concentration of B and N and controlling the
precipitation of BN via the nanoparticle morphology.
3.6.1.2 Formation of Low Melting Point Eutectic Nanoparticles
The catalytic ability of the Fe nanoparticles relies in part on the particles being in a liquid or
semi-molten state, in which the particles are quasi-spherical. The melting point of Fe is 1538
oC,
42 which is a higher temperature than the ones employed in Methods 1 and 2 (~1200
oC). Fe
nanoparticles can exhibit melting point depression, with smaller particles melting at a lower
temperature than the bulk material. However, for nanoparticles 25 nm and larger, which is the
estimated size of the Fe nanoparticles in Methods 1 and 2, the melting point would only be
depressed by ~150 oC.
43 Therefore, pure Fe nanoparticles would remain solid at the synthesis
temperature of ~1200 oC. When B and N diffuse into the Fe nanoparticles, alloys can form,
namely Fe2B and Fe4N. Fe4N does not possess catalytic properties. It can also be reduced back
to Fe by B at high temperatures (> 700 oC). On the other hand, Fe2B is a low melting point
eutectic, with a melting point of ~1200 oC.
42 It is more thermodynamically stable than Fe4N, and
is therefore the dominant alloy formed under the reaction conditions. Fe2B alloy particles can
form via solid state diffusion between adjacent B particles and Fe particles, as well as via
decomposition of B2O2 on the surface of Fe particles, and the subsequent diffusion of B into the
Fe particles, as depicted in Figure 3.7a.40
In the SA-CVD method, Si can also alloy with Fe to
produce low melting point eutectics (FeSix), with melting points ~1200 oC.
44,45
3.6.1.3 Vapour-Liquid-Solid Hypothesis
Many studies of CVD methods propose that BNNTs are formed via a vapour-liquid-solid (VLS)
process. The Fe2B nanoparticles constitute the liquid phase, while the gaseous B and N sources
(B2O2, NH3) make up the vapour phase. The process gases or growth vapours (N2 or NH3; B2O2)
decompose onto the metal nanoparticle surfaces, allowing for the diffusion of N and B atoms
into the nanoparticles. When the concentration of B and N within the nanoparticle reaches
32
supersaturation, B and N precipitate out on the surface of the metal nanoparticles in a layer-by-
layer fashion to form a cap of BN (the solid phase), as shown in Figure 3.7b.23
It is also a
possibility that the B atoms which diffused into the Fe2B particle eventually segregate to form a
shell on the outside of the nanoparticle, and then react directly with N from the decomposed NH3
vapour to form the BN cap.31, 40
As BN growth continues, the liquid nanoparticle undergoes a shape change from spherical to
elongated, due to capillary forces from the BN cap/growing BNNT (Figure 3.7c). This shape
change allows for the growth of cylindrical BN walls, forming a BNNT.46
Figure 3.7 Schematic of BNNT growth, showing (a) the Fe-B alloy catalyst particle and its uptake of B and N from gaseous sources (and also B via solid state diffusion), (b) the precipitation of a BN cap, (c) BNNT growth from an
elongated Fe-B alloy particle, and (d) contraction of Fe-B alloy particle and formation of a new BN cap. The growth process repeats to form the bamboo BNNT structure.
Bamboo BNNTs are typically formed at lower temperatures than straight-walled cylindrical
MWBNNTs. The VLS hypothesis can explain the formation of the bamboo structure by a
stuttered MWBNNT growth process in which the shape of the metal nanoparticle is dictated by
the balance of surface tension and capillary forces. In this process, a BN cap is formed and
growth of a short segment of cylindrical walls occurs with the metal nanoparticle elongating due
to capillary forces. The elongated shape of the nanoparticle is not stable below a certain
temperature, and surface tension causes the metal nanoparticle to contract, reforming a sphere.
The process then repeats, with the formation of a new BN cap and the growth of another short
segment of cylindrical tube wall. This process is illustrated in Figure 3.7d above.
33
The colour change of the precursor powders from rust-coloured to black indicates that the iron
oxide component changes from Fe2O3 to Fe to Fe3O4 over the course of the reaction.
3.6.1.4 Achieving BNNT Growth at Low Temperatures
In Method 1, the design and synthesis of the supported catalyst material and subsequent ball
milling result in the successful growth of BNNTs at relatively low temperatures (1100 – 1150
oC). The choice of catalyst support has been found to influence nanotube growth in studies of
CNT synthesis. SiO2 and Al2O3 are two support materials that appear to promote growth. The
high surface area of the fumed Al2O3 enables a high dispersion of Fe2O3 particles with minimal
aggregation. Furthermore, there is a strong interaction between the two oxides. The Al2O3
stabilizes the Fe2O3 particles to high temperatures by inhibiting their reducibility.47
Therefore,
the formation of the catalytic Fe nanoparticles occurs at temperatures closer to the optimal
synthesis temperature, which gives the particles less time to sinter and aggregate together. This
ensures that the particle sizes of the catalyst remain small and amenable to nanotube growth.48
SiO2 has a similar interaction with Fe2O3, although not as strong as Al2O3. The introduction of
SiO2 into the Fe2O3/Al2O3 support promotes efficient BNNT growth by creating larger pore
volumes and increasing the surface area of the supported catalyst exposed to growth vapours
during synthesis.31
The relatively short (20 h) ball milling treatment of the precursor powders
breaks down large aggregates and thoroughly mixes together all components, resulting in a fine
powder with a uniform distribution of precursors and high surface area. The redox reaction
between SiO2 and B also increases the concentration of B2O2 vapour, which leads to a higher
nucleation rate.
Three strategies are employed to increase the nucleation rate for BNNTs at 1200 oC in the GVT-
CVD method. The first strategy is the efficient generation of B2O2 vapour in the early stages of
the synthesis via the redox reaction between MgO and B, following the boron oxide chemical
vapour deposition (BOCVD) method developed by Zhi et al.26
The second strategy is to trap the
growth vapours evolved from the redox reactions between the precursors, such that there is a
high concentration of B source in the vicinity of the Fe catalyst particles within the alumina boat.
The trapping is accomplished by means of a quartz test tube, with the alumina boat located
within the test tube at the closed end. The closed end faces the flow of incoming NH3 gas, and
therefore protects the growth vapours from being transported downstream away from the boat.
34
The open end is 60 cm away from the closed end, and allows the NH3 gas to diffuse into the test
tube and reach the alumina boat. Lastly, the third strategy for promoting nucleation is to carry
out the synthesis under vacuum conditions. The mild vacuum (~0.025 MPa) created by the
mechanical pump increases the vapour pressure of the growth species, which in turn results in a
higher probability of nuclei formation. The probability of nuclei formation is described in
Equation 3.11, in which A is a constant, σ is the surface energy of the catalyst particles, k is the
Boltzmann constant, T is the temperature in Kelvin, and α = p/po (p = vapour pressure of the
growth species, po = equilibrium vapour pressure of the condensed phase).
Equation 3.11
For constant T and σ, the probability of nuclei formation can be increased by increasing the
vapour pressure of the growth species, and therefore α.24
3.6.2 Qualitative Comparison of Methods
Methods 1 and 2 have different strengths and weaknesses as synthesis methods for BNNTs. SA-
CVD possesses a higher growth rate/yield than the GVT-CVD method, and is more amenable to
industrial scale-up. The control over BN nanostructure morphology via synthesis temperature is
useful for the study of each type of nanostructure, and also gives insight into the growth
mechanisms involved. On a laboratory scale, the preparation of the precursor powder is a time
intensive process, and the many components make purification of the final product more
difficult. The GVT-CVD method has the advantage of very quick and simple precursor powder
preparation, and BNNTs collected from substrates covering the reaction boat can be used for
most applications without further purification. However, the growth rate and yield of GVT-CVD
are lower than those of SA-CVD, and as a result make GVT-CVD not as suitable to industrial
scale-up.
3.7 Conclusions
The main methods for synthesizing BNNTs were summarized. Two methods utilized over the
course of this thesis were presented in detail, namely a silica-assisted chemical vapour deposition
technique, and a growth vapour trapping chemical vapour deposition technique. The resulting
35
products are described qualitatively (to be presented in further detail in Chapter 4), and possible
growth mechanisms are discussed.
3.8 Contributions
The author set up the synthesis equipment and process for the methods of Reddy et al. and Lee et
al. at the University of Toronto. BNNTs were synthesized according to the protocols,
characterized (see next chapter), and used in various experiments.
3.9 References
1. Haubner, R.; Wilhelm, M.; Weissenbacher, R.; Lux, B., Boron Nitrides - Properties,
Synthesis and Applications. In High Performance Non-Oxide Ceramics II, Jansen, M.,
Ed. Springer Verlag: Berlin, Heidelberg, 2002; pp 1–45.
2. Iijima, S., Helical microtubules of graphitic carbon. Nature 1991, 354, 56–58.
3. Prasek, J.; Drbohlavova, J.; Chomoucka, J.; Hubalek, J.; Jasek, O.; Adam, V.; Kizek, R.,
Methods for carbon nanotubes synthesis - review. J. Mater. Chem. 2011, 21, 15872–
15884.
4. Chopra, N. G.; Luyken, R. J.; Cherrey, K.; Crespi, V. H.; Cohen, M. L.; Louie, S. G.;
Zettl, A., Boron-nitride nanotubes. Science 1995, 269, 966–967.
5. Loiseau, A.; Willaime, F.; Demoncy, N.; Hug, G.; Pascard, H., Boron nitride nanotubes
with reduced numbers of layers synthesized by arc discharge. Phys. Rev. Lett. 1996, 76,
4737–4740.
6. Cumings, J.; Zettl, A., Mass-production of boron nitride double-wall nanotubes and
nanococoons. Chem. Phys. Lett. 2000, 316, 211–216.
7. Laude, T.; Matsui, Y.; Marraud, A.; Jouffrey, B., Long ropes of boron nitride nanotubes
grown by a continuous laser heating. Appl. Phys. Lett. 2000, 76, 3239–3241.
8. Golberg, D.; Bando, Y.; Eremets, M.; Takemura, K.; Kurashima, K.; Yusa, H.,
Nanotubes in boron nitride laser heated at high pressure. Appl. Phys. Lett. 1996, 69,
2045–2047.
9. Yu, D. P.; Sun, X. S.; Lee, C. S.; Bello, I.; Lee, S. T.; Gu, H. D.; Leung, K. M.; Zhou, G.
W.; Dong, Z. F.; Zhang, Z., Synthesis of boron nitride nanotubes by means of excimer
laser ablation at high temperature. Appl. Phys. Lett. 1998, 72, 1966–1968.
10. Lee, R. S.; Gavillet, J.; Chapelle, M. L. D. L.; Loiseau, A.; Cochon, J. L.; Pigache, D.;
Thibault, J.; Willaime, F., Catalyst-free synthesis of boron nitride single-wall nanotubes
with a preferred zig-zag configuration. Phys. Rev. B 2001, 64, 121405.
11. Smith, M. W.; Jordan, K. C.; Park, C.; Kim, J.-W.; Lillehei, P. T.; Crooks, R.; Harrison,
J. S., Very long single- and few-walled boron nitride nanotubes via the pressurized
vapor/condenser method. Nanotechnology 2009, 20, 505604.
36
12. Wang, J.; Kayastha, V. K.; Yap, Y. K.; Fan, Z.; Lu, J. G.; Pan, Z.; Ivanov, I. N.;
Puretzky, A. A.; Geohegan, D. B., Low temperature growth of boron nitride nanotubes on
substrates. Nano Lett. 2005, 5, 2528–2532.
13. Han, W.; Bando, Y.; Kurashima, K.; Sato, T., Synthesis of boron nitride nanotubes from
carbon nanotubes by a substitution reaction. Appl. Phys. Lett. 1998, 73, 3085–3087.
14. Zhong, B.; Song, L.; Huang, X. X.; Wen, G. W.; Xia, L., Synthesis of boron nitride
nanotubes with SiC nanowire as template. Mater. Res. Bull. 2011, 46, 1521–1523.
15. Bando, Y.; Golberg, D.; Tang, C.; Zhang, J.; Ding, X.; Fan, S.; Liu, C., Thermal
conductivity of nanostructured boron nitride materials. J. Phys. Chem. B 2006, 110,
10354–10357.
16. Chopra, N. G.; Zettl, A., Measurement of the elastic modulus of a multi-wall boron
nitride nanotube. Solid State Commun. 1998, 105, 297–300.
17. Li, N.; Li, X.; Geng, W.; Zhao, L.; Zhu, G.; Wang, R.; Qiu, S., Template synthesis of
boron nitride nanotubes in mesoporous silica SBA-15. Mater. Lett. 2005, 59, 925–928.
18. Bechelany, M.; Bernard, S.; Brioude, A.; Cornu, D.; Stadelmann, P.; Charcosset, C.;
Fiaty, K.; Miele, P., Synthesis of boron nitride nanotubes by a template-assisted polymer
thermolysis process. J. Phys. Chem. C 2007, 111, 13378–13384.
19. Choy, K. L., Chemical vapour deposition of coatings. Prog. Mater Sci. 2003, 48, 57–170.
20. Lourie, O. R.; Jones, C. R.; Bartlett, B. M.; Gibbons, P. C.; Ruoff, R. S.; Buhro, W. E.,
CVD growth of boron nitride nanotubes. Chem. Mater. 2000, 12, 1808–1810.
21. Li, J.; Lin, H.; Chen, Y.; Su, Q.; Huang, Q., The effect of iron oxide on the formation of
boron nitride nanotubes. Chem. Eng. J. 2011, 174, 687–692.
22. Gan, Z. W.; Ding, X. X.; Huang, Z. X.; Huang, X. T.; Cheng, C.; Tang, C.; Qi, S. R.,
Growth of boron nitride nanotube film in situ. Appl. Phys. A. 2005, 81, 527–529.
23. Fu, J. J.; Lu, Y. N.; Xu, H.; Huo, K. F.; Wang, X. Z.; Li, L.; Hu, Z.; Chen, Y., The
synthesis of boron nitride nanotubes by an extended vapour–liquid–solid method.
Nanotechnology 2004, 15, 727.
24. Lee, C. H.; Wang, J. S.; Kayatsha, V. K.; Huang, J. Y.; Yap, Y. K., Effective growth of
boron nitride nanotubes by thermal chemical vapor deposition. Nanotechnology 2008, 19,
455605–5.
25. Tang, C.; Bando, Y.; Sato, T.; Kurashima, K., A novel precursor for synthesis of pure
boron nitride nanotubes. Chem. Commun. 2002, 1290–1291.
26. Zhi, C.; Bando, Y.; Tan, C.; Golberg, D., Effective precursor for high yield synthesis of
pure BN nanotubes. Solid State Commun 2005, 135, 67–70.
27. Guo, L.; Singh, R. N., Selective growth of boron nitride nanotubes by plasma-enhanced
chemical vapor deposition at low substrate temperature. Nanotechnology 2008, 19,
065601.
28. Guo, L.; Singh, R. N., Catalytic growth of boron nitride nanotubes using gas precursors.
Physica E 2009, 41, 448–453.
37
29. Kim, M. J.; Chatterjee, S.; Kim, S. M.; Stach, E. A.; Bradley, M. G.; Pender, M. J.;
Sneddon, L. G.; Maruyama, B., Double-walled boron nitride nanotubes grown by floating
catalyst chemical vapor deposition. Nano Lett. 2008, 8, 3298–3302.
30. Ma, R.; Bando, Y.; Sato, T., CVD synthesis of boron nitride nanotubes without metal
catalysts. Chem. Phys. Lett. 2001, 337, 61–64.
31. Su, C.-Y.; Chu, W.-Y.; Juang, Z.-Y.; Chen, K.-F.; Cheng, B.-M.; Chen, F.-R.; Leou, K.-
C.; Tsai, C.-H., Large-scale synthesis of boron nitride nanotubes with iron-supported
catalysts. The Journal of Physical Chemistry C 2009, 113, 14732–14738.
32. Chen, Y.; Fitz Gerald, J.; Williams, J. S.; Bulcock, S., Synthesis of boron nitride
nanotubes at low temperatures using reactive ball milling. Chem. Phys. Lett. 1999, 299,
260–264.
33. Chen, Y.; Chadderton, L. T.; Gerald, J. F.; Williams, J. S., A solid-state process for
formation of boron nitride nanotubes. Appl. Phys. Lett. 1999, 74, 2960–2962.
34. Singhal, S.; Srivastava, A.; Pant, R.; Halder, S.; Singh, B.; Gupta, A., Synthesis of boron
nitride nanotubes employing mechanothermal process and its characterization. J. Mater.
Sci. 2008, 43, 5243–5250.
35. Kim, J.; Lee, S.; Uhm, Y. R.; Jun, J.; Rhee, C. K.; Kim, G. M., Synthesis and growth of
boron nitride nanotubes by a ball milling–annealing process. Acta Mater. 2011, 59, 2807–
2813.
36. Li, Y.; Zhou, J. e.; Zhao, K.; Tung, S.; Schneider, E., Synthesis of boron nitride
nanotubes from boron oxide by ball milling and annealing process. Mater. Lett. 2009, 63,
1733–1736.
37. Wang, J.; Gu, Y.; Zhang, L.; Zhao, G.; Zhang, Z., Synthesis of boron nitride nanotubes
by self-propagation high-temperature synthesis and annealing method. J. Nanomater.
2010, 2010, 540456–6.
38. Wang, J.; Zhang, L.; Zhao, G.; Gu, Y.; Zhang, Z.; Zhang, F.; Wang, W., Selective
synthesis of boron nitride nanotubes by self-propagation high-temperature synthesis and
annealing process. J. Solid State Chem. 2011, 184, 2478–2484.
39. Reddy, A. L. M.; Tanur, A. E.; Walker, G. C., Synthesis and hydrogen storage properties
of different types of boron nitride nanostructures. Int. J. Hydrogen Energy 2010, 35,
4138–4143.
40. Koi, N.; Oku, T.; Inoue, M.; Suganuma, K., Structures and purification of boron nitride
nanotubes synthesized from boron-based powders with iron particles. J. Mater. Sci. 2008,
43, 2955–2961.
41. Öncel, Ç.; Yürüm, Y., Carbon nanotube dynthesis via the vatalytic CVD method: A
review on the effect of reaction parameters. Fullerenes Nanotubes Carbon Nanostruct.
2006, 14, 17–37.
42. Okamoto, H., B-Fe (Boron-Iron). J. Phase Equilib. Diffus. 2004, 25, 297–298.
38
43. Shu, Q.; Yang, Y.; Zhai, Y.; Sun, D.; Xiang, H.; Gong, X.-G., Size-dependent melting
behavior of iron nanoparticles by replica exchange molecular dynamics. Nanoscale 2012
(accepted manuscript).
44. Tang, C. C.; Ding, X. X.; Huang, X. T.; Gan, Z. W.; Qi, S. R.; Liu, W.; Fan, S. S.,
Effective growth of boron nitride nanotubes. Chem. Phys. Lett. 2002, 356, 254–258.
45. Schlesinger, M. E., Thermodynamics of solid transition-metal silicides. Chem. Rev. 1990,
90, 607–628.
46. Loh, K. P.; Lin, M.; Yeadon, M.; Boothroyd, C.; Hu, Z., Growth of boron nitride
nanotubes and iron nanowires from the liquid flow of FeB nanoparticles. Chem. Phys.
Lett. 2004, 387, 40–46.
47. Berry, F. J.; Lin, L.; Liang, D.; Wang, C.; Tang, R.; Zhang, S., An investigation of metal-
support interactions in some iron, ruthenium, and iron-ruthenium catalysts by in situ iron-
57 Mossbauer spectroscopy. Appl.Catal. 1986, 27, 195–205.
48. Esmaieli, M.; Khodadadi, A.; Mortazavi, Y., Catalyst support and pretreatment effects on
carbon nanotubes synthesis by chemical vapor deposition of methane on iron over SiO2,
Al2O3 or MgO. Int. J. Chem. Reactor Eng. 2009, 7, A36–17.
Equation 4 f
Figure 4. t
Table 4.
39
4 Structural and Chemical Characterization of Boron Nitride Nanotubes
4.1 Abstract
In this chapter, the structural and chemical characterization of boron nitride nanotubes via
electron microscopy, FTIR, and EDX is presented.
4.2 Introduction
Electron microscopy and its related techniques are invaluable characterization tools for the study
of nanomaterials. In general, scanning electron microscopy (SEM) has the capability to image at
magnifications ranging from x 10 – 100 000, while high resolution transmission electron
microscopy (HR-TEM) can achieve magnifications up to x 10 000 000. TEM can measure the
spacing between planes of atoms (lattice imaging), yielding crystallographic information.1, 2
In
addition to imaging, electron microscopy is often performed together with integrated techniques
such as energy dispersive x-ray (EDX) analysis. EDX provides chemical analysis through the
detection of characteristic x-ray lines emitted by the sample due to interaction with the electron
beam. Elements with Z ≥ 5 can be detected qualitatively, with quantitative analysis possible for
Z ≥ 11. The combination of SEM-EDX allows for chemical analysis on the microscale, and
correlation between micro-nano structures and elemental composition.3
For bulk characterization of chemical composition, Fourier transform infrared (FTIR)
spectroscopy is a commonly used technique.4 FTIR can differentiate between hexagonal and
cubic boron nitride phases. The sp2 bonding in h-BN gives rise to two characteristic lattice
vibrational modes which are infrared-active: E1u, corresponding to in-plane stretching between B
and N atoms, and A2u, corresponding to out-of-plane stretching between B and N atoms. In c-
BN, the sp3 bonding results in one infrared-active mode, F2. The frequencies of the modes for h-
BN and c-BN are summarized in Table 4.1 below.5
40
Table 4.1 Infrared modes of h-BN and c-BN
Material Infrared Mode Wavenumber (cm-1
)
h-BN E1u (TO) 1367
E1u (LO) 1510
A2u (TO) 767
A2u (LO) 783
c-BN F2 1045
4.3 Experimental Methods
4.3.1 Synthesis
Boron nitride nanotubes were synthesized by two methods, as described in Chapter 3. Briefly, a
silica-assisted chemical vapour deposition technique (SA-CVD) was used to produce two types
of BNNTs. Bamboo morphology tubes were grown at a temperature of 1100 oC, while multiwall
BNNTs (MWBNNTs) were grown at 1150 oC. Following synthesis, the products were purified
by HNO3 and HF treatment. MWBNNTs were also synthesized using a growth-vapour-trapping
chemical vapour deposition technique (GVT-CVD).
4.3.2 Electron Microscopy
BNNT samples were dispersed in DI H2O or ethanol via sonication for ~30 min. Drops of the
BNNT suspension were placed on Si substrates or lacy C-coated TEM grids and allowed to dry.
SEM images were acquired with accelerating voltages of 10 – 15 kV, and currents of 20 – 30 μA
(S-5200, Hitachi, Japan). STEM images were obtained at 200 kV and ~40 μA emission current
(HD-2000, Hitachi, Japan).
41
4.3.3 Energy Dispersive X-ray
EDX spectra and element maps were obtained with accelerating voltages of 10-15 kV (Inca
System, Oxford Instruments, United Kingdom).
4.3.4 Fourier Transform Infrared Spectroscopy
SA-CVD BNNT samples were dispersed in KBr powder and pellets were made. Transmission
FTIR was performed on the pellets (Spectrum BX, Perkin Elmer, MA). GVT-CVD BNNTs and
h-BN nanoparticles (< 100 nm diameter, MKNano, Canada) were dispersed in ethanol via
sonication, and deposited onto a Ge attenuated total internal reflection (ATR) crystal. The
solvent was allowed to dry, and then spectra were obtained in the ATR configuration (Spectrum
BX, Perkin Elmer, MA).
4.4 Results
4.4.1 Scanning Transmission Electron Microscopy
STEM images obtained in Z-contrast mode are shown in Figure 4.1, showing two nanotubes
produced by Method 1. Figure 4.1a shows a ~60 nm diameter straight-wall BNNT several
microns long. A magnified image of the root of the nanoparticle, attached to a round catalyst
particle, is shown in Figure 4.1b. The tip of the nanotube in Figure 4.1a in shown in greater
detail in Figure 4.1c. Two bamboo segments can be seen to be attached to the straight-wall
portion of the nanotube. Figure 4.1d shows a ~1 μm long fragment of a bamboo BNNT with a
diameter of ~100 nm.
A low resolution TEM image of BNNTs produced by GVT-CVD is shown in Figure 4.2a. The
nanotubes are hollow, with diameters ranging from 30 – 60 nm. Periodic dark spots are visible
on the nanotube walls. A high resolution TEM image of one such dark spot is shown in Figure
4.2b. This particular tube has ~23 walls. The corresponding fast Fourier transform is given in
Figure 4.2c. The two main diffraction spots correspond to [0002] reflections in h-BN, and
indicate an inter-wall spacing of 0.34 nm.6
42
Figure 4.1 Scanning transmission electron microscopy images of BNNTs produced by SA-CVD. (a) Nanotube with catalyst particle attached at root. (b) Higher magnification image of catalyst particle in (a). (c) Higher magnification
image of nanotube tip in (a). (d) Broken bamboo morphology nanotube.
Figure 4.2 Transmission electron microscopy images of BNNTs produced by GVT-CVD. (a) Low-resolution TEM image showing hollow nanotubes with periodic dark spots within their walls. (b) High resolution TEM image of a dark
spot from (a). (c) Fast Fourier transform of nanotube wall from (b).
43
4.4.2 EDX Characterization
The EDX spectrum for SA-CVD MWBNNTs prior to HF purification is given in Figure 4.3a.
The dominant peaks are O, Al, and Si, with Si having the highest number of counts due to the Si
substrate upon which the MWBNNTs are deposited. These EDX signals indicate the presence of
Al2O3 and SiO2, precursor materials which can be removed by HF treatment. Figure 4.3b shows
an expanded view of the spectrum below 1 keV, in which B, N, and O signals are present. The
sample region from which the EDX spectrum was collected is indicated by the pink box in
Figure 4.3c.
Figure 4.3 (a) EDX spectrum of SA-CVD BNNTs on Si substrate prior to HF purification. (b) Low energy region of spectrum from (a). (b) SEM image indicating the area over which the EDX spectrum was collected (pink box).
44
A bamboo BNNT sample produced via SA-CVD was analyzed with EDX following HF
purification. The spectrum shown in Figure 4.4a shows B, N, O, and Si signals (again, the Si
signal is largely due to the substrate). No Al is apparent, and the O signal is lower than the N
signal, indicating that the majority of the oxide precursors have been eliminated via HF
treatment. An expanded view of the B, N, and O peaks is given in Figure 4.4b, and the sample
area from which the measurement was taken is indicated by the pink box in Figure 4.4c.
Figure 4.4 EDX spectrum of SA-CVD Bamboo BNNTs (after HF purification) on Si substrate.
In order to obtain spatially correlated elemental composition, EDX element maps were acquired.
Results for unpurified MWBNNTs produced via GVT-CVD, are shown in Figure 4.5 below,
45
with an SEM image of the region of interest presented at the top of the figure, followed by
individual EDX element maps. These MWBNNTs were deposited on a lacy C coated TEM grid.
The nanotube features are clearly visible in the B and N maps. Some detector overlap between
the B and C x-ray lines appears to cause some nanotube features to be visible in the C map. The
O, Mg, and Fe maps show that the material in the upper left corner of the SEM image is
comprised of residual precursors, namely MgO and Fe (likely in Fe3O4 form, see previous
chapter for details the of synthetic redox reactions)
Figure 4.5 SEM image of GVT-CVD BNNTs on lacy C TEM grid and corresponding EDX elemental maps.
46
The corresponding EDX spectrum acquired over the area shown above in Figure 4.5 is given in
Figure 4.6a below. The dominant peaks in the spectrum are of C, O, Mg, Fe, and Cu. The high
C signal is due to the lacy C coating on the TEM grid, and the Cu signal is from the TEM grid
itself. An expanded view from 0 to 2.5 keV is shown in Figure 4.6b. No B peak can be seen due
to overlap with the C peak. The N signal is apparent, and small Al and Si peaks are also present.
Al and Si may have been introduced into the BNNT sample from the Al2O3 boat and the Si
substrate which partially covered the boat during the synthesis. The small peak at ~2.1 keV could
not be assigned to any characteristic x-ray lines.
Figure 4.6 EDX spectrum of GVT-CVD BNNTs on lacy C TEM grid.
4.4.3 FTIR Characterization
The FTIR spectra for SA-CVD bamboo BNNTs, SA-CVD MWBNNTs, GVT-CVD
MWBNNTs, and h-BN nanoparticles are given in Figure 4.7. All materials exhibit two major
bands at ~800 cm-1
and ~1375 cm-1
which is characteristic of h-BN (A2u, E1u TO). The GVT-
CVD MWBNNT sample exhibits another peak at ~1500 cm-1
, which corresponds to the E1u LO
mode. The other samples do not exhibit two distinct TO and LO modes, but a shoulder which
47
likely corresponds to the LO mode is apparent in the h-BN nanoparticle spectrum. The broad
asymmetric lineshapes of the E1u TO peaks in the SA-CVD sample spectra indicate that some
contribution from the LO mode is present.
Figure 4.7 Fourier transform infrared spectra of BNNTs produced by SA-CVD and GVT-CVD. The spectrum for <100 nm h-BN nanoparticles is shown for comparison.
4.5 Discussion
4.5.1 Nanotube Morphology and Structure
The observation of the SA-CVD MWBNNT with a catalyst particle attached at its root, and
bamboo segments at its tip (Figure 4.1a-c) supports the growth mechanism hypothesis described
in Chapter (see Figure 3.7). In the initial stage of growth, the catalyst particle could not maintain
48
a stable morphology for the growth of a straight-walled tube, and as a result, bamboo segments
were formed. The catalyst particle was eventually able to maintain an elongated shape, perhaps
due to an increase in temperature, such that the remainder of the nanotube formed a MWBNNT
morphology. At the end of the growth process, as the temperature decreased, the catalyst particle
formed a sphere once again. Note that the diameter of the catalyst particle is larger than the
diameter of the nanotube. When elongated, the particle’s diameter decreases (in the direction
perpendicular to the nanotube axis) which templates the diameter of the nanotube.
4.5.2 Defect Characterization
The peak position of the E1u (TO) mode is sensitive to defects within the h-BN lattice. As lattice
defects increase, the peak will shift to higher wavenumber. The peak positions of the BNNTs
synthesized by SA-CVD and GVT-CVD are summarized in Table 4.2 together with the peak
positions for the h-BN nanoparticles. For comparison, the peak positions for a thin film of single
crystalline h-BN are also listed.7
Table 4.2 FTIR Peak Positions for BN Nanomaterials.
BN Material/Infrared Mode E1u, TO (cm-1
) E1u, LO (cm-1
) A2u (cm-1
)
Bamboo BNNTs (SA-CVD) 1384 - 797
MWBNNTs (SA-CVD) 1381 - 798
MWBNNTs (GVT-CVD) 1367 1513 805/819
h-BN nanoparticles 1377 ~1480 shoulder 783
Single crystal h-BN thin film
(from Shi et al.7)
1370 - 823
Based on the E1u, TO peak position, the bamboo BNNTs have the most defective structure,
followed by the SA-CVD MWBNNTs and the h-BN nanoparticles. The GVT-CVD MWBNNTs
appear to be highly crystalline, with a peak position of 1367 cm-1
. The narrow band width and
the distinctly separate LO peak at 1510 cm-1
are also strong indications of the high crystallinity
of the GVT-CVD MWBNNT sample.
49
4.6 Conclusions
The morphology of the BNNTs produced by SA-CVD and GVT-CVD were investigated with
SEM. EDX is a useful tool for detecting elements of interest within the material, and correlating
the elemental composition with nanostructures. The crystal structure of GVT-CVD BNNTs was
investigated with HR-TEM, and confirmed the tubular h-BN nature of the nanotubes. FTIR was
employed in order to provide bulk characterization of the chemical composition of the BNNTs.
4.7 References
1. Verhoeven, J. D., Scanning Electron Microscopy. In ASM Handbook, Volume 10 -
Materials Characterization, Whan, R. E., Ed. ASM International: 1986; pp 490–515.
2. Romig, A. D. J., Analytical Transmission Electron Microscopy. In ASM Handbook,
Volume 10 - Materials Characterization, Whan, R. E., Ed. ASM International: 1986; pp
429–489.
3. Heinrich, K. F. J.; Newbury, D. E., Electron Probe X-Ray Microanalysis. In ASM
Handbook, Volume 10 - Materials Characterization, Whan, R. E., Ed. ASM
International: 1986; pp 516–535.
4. Marcott, C., Infrared Spectroscopy. In ASM Handbook, Volume 10 - Materials
Characterization, Whan, R. E., Ed. ASM International: 1986; pp 109–125.
5. Geick, R.; Perry, C. H.; Rupprech.G, Normal modes in hexagonal boron nitride. Phys.
Rev. 1966, 146, 543–547.
6. Terauchi, M.; Tanaka, M.; Matsuda, H.; Takeda, M.; Kimura, K., Helical nanotubes of
hexagonal boron nitride. J. Electron Microsc.1997, 46, 75–78.
7. Shi, Y.; Hamsen, C.; Jia, X.; Kim, K. K.; Reina, A.; Hofmann, M.; Hsu, A. L.; Zhang, K.;
Li, H.; Juang, Z.-Y.; Dresselhaus, M. S.; Li, L.-J.; Kong, J., Synthesis of few-layer
hexagonal boron nitride thin film by chemical vapor deposition. Nano. Lett. 2010, 10,
4134–4139.
Figure 5. sds
50
5 Diameter-Dependent Bending Modulus of Individual Multiwall Boron Nitride Nanotubes
5.1 Abstract
The mechanical properties of individual multiwall boron nitride nanotubes (MWBNNTs)
synthesized by a growth-vapor-trapping chemical vapor deposition method are investigated by a
three-point bending technique via atomic force microscopy. Multiple locations on suspended
tubes are probed in order to determine the boundary conditions of the supported tube ends. The
bending moduli (EB) calculated for 20 tubes with diameters ranging from 18 to 55 nm confirm
the exceptional mechanical properties of MWBNNTs, with an average EB of 760 ±30 GPa. For
the first time, the bending moduli of MWBNNTs are observed to increase with decreasing
diameter, ranging from 100 ± 20 GPa to as high as 1800 ± 300 GPa. This diameter dependence is
evaluated by Timoshenko beam theory. The Young’s modulus and shear modulus were
determined to be 1800 ± 300 GPa and 7 ± 1 GPa, respectively for a trimmed data set of 16 tubes.
The low shear modulus of MWBNNTs is the reason for the detected diameter-dependent
bending modulus and is likely due to the presence of inter-wall shearing between the crystalline
and faceted helical nanotube structures of MWBNNTs.
5.2 Introduction
Boron nitride nanotubes (BNNTs), first predicted in 19941,2
and synthesized in 1995,3 have
attracted increasing attention in recent years due to their unusual properties. Although
structurally similar to carbon nanotubes (CNTs), BNNTs have significantly different optical and
electronic properties. BNNTs are much more insulating than CNTs, with a band gap of 5 – 6 eV
which is largely independent of tube chirality or diameter.2 Theoretical studies have indicated
that the axial Young’s modulus of single wall BNNTs (SWBNNTs) is of the same order as that
of carbon nanotubes (~1 TPa).4,5
BNNTs’ mechanical properties, together with their high
aspect ratio, high thermal conductivity,6 optical transparency, electrically insulating character,
and high resistance to oxidation (up to 1100oC)
7 make them ideal fillers for technologically
relevant composite materials such as seals and encapsulants,8-11
and biomaterials.12
In addition,
BNNTs show promise for a diverse range of other applications, including hydrogen storage,13,14
targeted drug delivery,15
and optoelectronic devices such as lasers and light emitting diodes.16,17
51
For BNNTs to be successfully employed in the aforementioned applications, a better
understanding of their mechanical properties is required. This is particularly important for
applications which rely on the mechanical properties of individual tubes, such as resonators and
sensors,18
and microtubule mimics.19
In contrast to CNTs, only a handful of experimental
studies have been conducted on BNNTs to determine their Young’s modulus. Chopra and Zettl20
used the resonance technique of Treacy et al.21
to determine that an arc-discharge multiwall
BNNT (MWBNNT) (3.5 nm outer diameter) had a modulus of ~1.22 TPa. Electric-field-
induced resonance experiments by Suryavanshi et al.22
yielded moduli of 505 – 1031 GPa for a
set of 18 tubes, with outer diameters ranging from 34 – 94 nm. Golberg et al.23
determined
moduli of 0.5 – 0.6 TPa (40 nm and 100 nm outer diameter tubes) via in-situ bending
experiments using an atomic force microscope (AFM) set-up within a transmission electron
microscope (TEM). Using a similar set-up, Ghassemi et al.24
measured 5 MWBNNTs with outer
diameters of 38 – 51 nm, and found that the average modulus was ~0.5 TPa. Depending on the
choice of shell thickness, the Young’s modulus of a 1.9 nm diameter SWBNNT was found to
range from 0.87 to 1.11 TPa.25
The wide range of moduli observed indicates a need for further
study in order to elucidate the influence of factors such as synthesis technique, nanotube
structure, and morphology on the Young’s modulus.
Three-point bending tests conducted with AFM have been used to characterize the modulus of a
variety of high aspect ratio structures, including CNTs,26-28
nanowires,29,30
and electrospun
polymers.31
Typically, the nanotubes or wires are deposited onto a stiff substrate with a
topographical pattern, such as polished porous Al2O3 membranes or Si gratings patterned with
trenches. The tubes occasionally lie over pores or trenches, and the midpoint of the suspended
portion is subjected to a downward force applied by the AFM tip. Force-displacement curves are
obtained, and the bending modulus can be calculated directly from the slope of the force curve
together with the geometrical parameters of the tube’s diameter and suspended length. In most
studies, the supported beam ends are assumed to have clamped boundary conditions due to the
adhesion between the nanomaterial and the substrate. However, this assumption can be
unfounded and can be a source of systematic error in the determination of the bending modulus.
Other beam end boundary conditions include simply supported and mixed support in which one
end is clamped, and the other end is simply supported. Depending on the support conditions, the
solution of the Euler-Bernoulli beam equation takes on different forms, yielding different
52
expressions for the bending modulus. The appropriate boundary conditions for an individual
tube can be determined if multiple locations along the length of the suspended tube are probed.
This allows for a more accurate determination of the modulus value, as demonstrated by
Shanmugham et al.,32
Chen et al.,33
Kluge et al.,34
and Gangadean et al.35
In this study, we use AFM to measure the bending modulus Eb of MWBNNTs synthesized by a
growth-vapor-trapping chemical vapor deposition (GVT-CVD) technique.36
A force mapping
technique is used in order to collect force curves from various locations along the length of the
suspended tube. We show that for our sample the majority of tubes possess simply supported
ends instead of clamped ends. Based on these boundary conditions, we calculate the bending
moduli for tubes of various diameters, and we present a discussion about the diameter
dependence that is observed.
5.3 Experimental Methods
5.3.1 MWBNNT Synthesis
MWBNNTs were synthesized via the growth vapor trapping chemical vapor deposition
technique previously described by Lee and coworkers.36
The MWBNNTs were collected on Si
substrates and sonicated in ethanol to form a MWBNNT suspension.
5.3.2 Chemical and Structural Characterization
MWBNNTs were synthesized via the growth vapor trapping chemical vapor deposition
technique previously described by Lee and coworkers.36
The MWBNNTs were collected on Si
substrates and sonicated in ethanol to form a MWBNNT suspension.
The MWBNNTs were characterized with scanning electron microscopy (SEM), low and high
resolution transmission electron microscopy (TEM, HR-TEM), and Fourier transform infrared
spectroscopy (FTIR). The morphology of the as-synthesized MWBNNTs was characterized with
SEM (S-4700, Hitachi, Japan). For the TEM measurements, the MWBNNT suspension was
dropped onto a holey carbon TEM grid and allowed to dry. Bright-field low resolution TEM
images were acquired at 30 kV, 17.5 μA emission current (S-5200, Hitachi, Japan). Bright-field
HR-TEM images were acquired at 200kV, 39 μA emission current (HD-2000, Hitachi, Japan).
FTIR spectra were taken using an attenuated total internal reflection (ATR) set-up. The
53
MWBNNT suspension was dropped onto a ZnSe ATR crystal and allowed to dry. For
comparison purposes, hexagonal boron nitride (h-BN) nanoparticles (MK-hBN-N70, MK Impex
Canada, Mississauga, Ontario) were also characterized by FTIR. Spectra were recorded on an
FTIR spectrometer (Spectrum BX, Perkin-Elmer, Waltham, MA) at a resolution of 1 cm-1
.
5.3.3 Sample Preparation and AFM Measurements
For the AFM sample preparation, the MWBNNT suspension was dropped onto clean Si
substrates patterned with trenches 400 nm wide and 200 nm deep (LightSmyth Technologies,
Eugene OR) and was allowed to dry. AFM height images of the tubes on the patterned substrate
were acquired in air under ambient conditions using AC (intermittent contact) mode (MFP-3D,
Asylum Research, Santa Barbara CA). Si probes (NCH, Nano World, Neuchâtel Switzerland)
with tip radii of ~20 nm were used. The optical lever sensitivity of the cantilevers was calibrated
by acquiring force curves in contact mode on a clean Si substrate. The spring constant of each
cantilever used was determined by the thermal method and found to range from 33 to 46 N/m.37
A discussion of the applicability of the thermal method for high spring constant cantilevers is
presented in the Supporting Information. AFM force maps (typically 2 μm x 0.5 μm with 32 x
16 points) were obtained of MWBNNTs spanning trenches. A force curve (applied force F
versus tip-sample separation) was collected at each point on the map. The force curves
corresponding to the points along the suspended portion of the tube (as determined from the
height map and the force curves themselves) were analyzed to extract the effective tube stiffness,
keff, by a linear fit to the slope of the force curve. A more detailed description of the force
mapping method is given in the Supporting Information.
The suspended length L for a given tube was determined from the AFM height map as well as
from higher-resolution AFM height images acquired in tapping mode. The lateral dimension of
the pixels making up the force map was used to estimate the errors associated with the values of
position (a, b) and suspended length. The tube diameter was determined by the height of the
tube on the substrate from the tapping mode height images.
Equation 5
54
5.4 Results and Discussion
5.4.1 Characterization of MWBNNTs
Electron microscopy images of the MWBNNTs produced by the GVT-CVD method are shown
in Figure 5.1. The scanning electron microscope (SEM) image in Figure 1a shows straight fibers
with diameters ranging from ~15 to 60 nm. Figure 5.1b depicts a low resolution bright field
TEM image of the as-synthesized MWBNNTs, and confirms the hollow tubular nature of the
fibers. A high resolution TEM image of a dark region in a tube wall is shown in Figure 5.1c.
The layers appear crystalline with an interlayer spacing of ~0.34 nm, as determined from the
(002) diffraction spots in the fast Fourier transform for this region (not shown). This spacing is
consistent with the crystal structure of hexagonal boron nitride and BNNTs.24,36
The FTIR spectrum of the MWBNNTs is shown in Figure 5.2. For comparison, the spectrum for
commercially available hexagonal BN (h-BN) nanoparticles is also given. The MWBNNT
spectrum exhibits peaks at ~1368 cm-1
and ~1510 cm-1
, which correspond to the in-plane
transverse optical (TO) and longitudinal optical (LO) E1u modes of h-BN. The TO E1u mode is a
stretching mode along the tube axis, while the LO E1u mode is a stretching mode along the tube
circumference. A weak feature at around 800 cm-1
is shown enlarged in the inset of Figure 5.2.
A shoulder is visible at ~819 cm-1
, and a peak at ~806 cm-1
. These spectral features correspond
to the out-of-plane TO and LO A2u modes of h-BN. 24,38
5.4.2 AFM Three-Point Bending
In three-point bending experiments, slender wires can be modeled as an elastic string (pure
stretching), a stiff beam (pure bending), or a combination of the two. Heidelberg et al.39
presented a generalized approximation for these behaviors, in which a force F is applied to the
mid-point of the suspended wire and the wire ends are assumed to be clamped:
Equation 5.1 ) 24I
A+(1
L
I192E=F 2
centercenter3
Bcenter
In the above expression, δ is the deflection of the wire, EB is the bending modulus, A is the cross
sectional area of the wire, and I is the second moment of area. At small displacements, the wire
undergoes pure bending which is described by the first linear term. At large displacements,
55
Figure 5.1 (a) SEM image of MWBNNTs. Inset: Higher magnification SEM image showing straight, slender fibres 15 – 60 nm in diameter. (b) Low resolution TEM image of MWBNNTs. (c) High resolution TEM image of a MWBNNT
wall near a typical dark spot shown in (b).
56
Figure 5.2 FTIR spectrum (blue line) of MWBNNTs. FTIR spectrum of hBN nanopowders (red line) is also shown for comparison.
axial tensile stresses are induced as the wire stretches which are described by the cubic second
term (F δ3). In this study, only pure bending is considered because the experiments are
conducted within the small deflection regime, in which the maximum deflection does not exceed
the radius of the wire.
To model the MWBNNTs in this study as stiff beams undergoing pure bending, Euler-Bernoulli
beam theory was employed. It should be noted that this theory assumes a homogeneous isotropic
material, which is not the case for multiwall nanotubes. Nevertheless, simulations indicate that
this approximation offers an adequate description of nanotube bending mechanics prior to
buckling.40
As a result, this approach has been widely used in AFM bending experiments on
nanowires and nanotubes.26-29,31
Unique solutions to the beam equation depend on the boundary
conditions of the beam ends, which can be considered to be clamped (no deflection or slope at
beam end), or simply supported (no deflection or bending moment at beam end). Beam
schematics are presented in Figure 5.3, which summarize the three models considered in this
work: simply supported beam model (a, SSBM), double clamped beam model (b, DCBM), and
mixed support beam model (c, MSBM).41
57
Figure 5.3 Beam schematics describing beam bending boundary conditions.
The corresponding equations are as follows:
Equation 5.2
22
3
ba
ILEF B
(SSBM)
Equation 5.3
33
33
ba
IELF B
(DCBM)
Equation 5.4
aLba
IELF B
3
1232
3
(MSBM)
In these equations, L is the suspended length of the beam, and a and b are the suspended lengths
on both sides of the applied force F, where a + b = L. I, the second moment of area, is taken to
be I = πD4/64 which is defined for a solid cylindrical wire with a circular cross section, where D
is the diameter. Hence, in this approximation of the multiwall nanotube beam, only the outer
diameter is taken into account and not the inner diameter. In order to determine the appropriate
boundary conditions for each tube, the AFM tip is used to apply a force at different positions
along the suspended tube, not just at the midpoint. Therefore, AFM force curves (plots of F
versus the tip-sample separation (= beam deflection, δ)) are collected at multiple locations along
the tube. The linear slope of a force curve directly yields the effective tube stiffness, keff = F/ δ.
The boundary conditions for the tube are determined by plotting keff versus the position along the
58
tube (a/L) and performing fits to the various beam models (Equations 2-4). The bending
modulus EB is then determined using the appropriate beam model.
Figure 5.4 shows SEM (a) and AFM images (b-d) of suspended MWBNNTs on patterned Si
trenches (400 nm wide and 200 nm deep). An AFM height image of a typical MWBNNT
spanning a trench is shown in Figure 5.4b. The height image is subsequently divided up into
pixels (typically 32 x 16 or 64 x 32) by the AFM software, and force curves are collected at each
point (pixel) during the force mapping procedure. The corresponding AFM height map image
illustrating the spatial location (x,y) of each of the force measurements is shown in Figure 5.4c.
The height in each pixel is determined from the Z range distance at which the tip first engages
the sample during the extend portion of the force curve. Figure 5.4d shows the AFM height
image of the MWBNNT after the force map was performed, and its similarity to Figure 5.4b
indicates that the tube did not shift or deform as a result of the force measurements. Typical
force curves collected from different locations on a MWBNNT are shown in Figure 5.5. The red
dotted line corresponds to a force curve obtained from a location where the tube is supported by
the Si substrate (red ‘x’ in Figure 5.4c, illustrative purpose only), while the blue solid line
corresponds to a force curve obtained from a position where the tube is suspended over a trench
(blue ‘x’ in Figure 5.4c, illustrative purpose only). The slope of the blue solid line in Figure 5.5
is equivalent to keff.
Plots of the effective tube stiffness (keff) versus position along the suspended tube (a/L) fitted
with Equations (1) to (3) corresponding to SSBM, DCBM, and MSBM are given in Figure 5.6.
The values of keff located near the ends (a/L < 0.2, a/L > 0.8) of the suspended tubes were not
included in the fits, because of the large error associated with fitting force curves with large
slopes (theoretically, at the ends, keff approaches infinity). Figure 5.6a shows that SSBM fits the
data better than DCBM signifying that the MWBNNT is an example of a simply supported tube.
On the other hand, data in Figure 5.6b fits the MSBM model well and is therefore an example of
a tube which is fixed on one end (its left side) and simply supported on the opposite end (its right
side). Despite the fact that all of the tubes examined were on the same sample, various support
conditions were observed which demonstrates the importance of determining the boundary
conditions for each individual tube.
59
Figure 5.4 (a) SEM image of MWBNNTs on patterned Si substrate. (b) AFM height image before force mapping was performed. (c) Height map image corresponding to a force map acquired at a deflection trigger of 1 nm. (d) AFM
height image after force map was acquired.
Figure 5.5 (Red dotted line) Force curve obtained from a point on a tube supported by substrate. (Blue solid line) Force curve obtained from a point on a tube suspended over a trench.
60
Figure 5.6 Tube effective stiffness (keff) vs. position along suspended tube (a/L). (a) A simply supported tube. (b) A
mixed support tube, with left side fixed and right side simply supported.
5.4.3 Elastic Properties of MWBNNTs
5.4.3.1 Bending Modulus
The bending moduli EB determined for 20 tubes with diameters ranging from 18 – 55 nm are
shown in Figure 5.7. EB ranged from 100 ± 20 to 1800 ± 300 GPa, with an average of 760 ± 30
GPa. The error in EB was determined via error propagation, using an error of 10% for the tube
61
outer diameter D, half the lateral pixel width in the force map for the tube length L and lengths
on each side of the loading position (a, b), and an error of 20% for effective spring constant of
the tube keff.
It is worth noting that there is the possibility that the calculated bending moduli may be
underestimated due to inaccurate assumptions about the cross sectional geometry. The
nanotubes were approximated to be solid wires, with a solid circular cross section. This model
was chosen because it was not possible for us to determine the inner diameter of the nanotubes
we probed with AFM, given that we deposited the tubes on a substrate that is not amenable to
TEM analysis. A more accurate model would be a hollow cylinder, with an inner diameter Di
and an outer diameter Do, which results in a second moment of area expression of I = π/64(Do4 –
Di4). Modeling the tube as a solid wire as opposed to a hollow cylinder underestimates the
modulus; however, in the most extreme case (i.e. very large diameter tube with only a few walls)
the underestimation is on the order of 30%. Based on the range of Di/Do ratios observed in
TEM data for 24 tubes in the same MWBNNT production batch as the tubes used in the bending
experiments (presented in the Supporting Information), the underestimation is closer to 10% for
our particular sample.
From the plot of EB versus tube outer diameter shown in Figure 5.7, it is evident that there is a
decreasing trend for the bending modulus with increasing tube diameter. For both CNTs and
BNNTs, the Young’s modulus should theoretically approach an upper limit defined by the in-
plane elastic constant of graphite and h-BN, respectively.4,42
For h-BN single crystals, this
constant was measured to be c11 = 811 GPa,43
while for single crystal graphite, c11 = 1109
GPa.4,42
These measurements correlate well with various theoretical calculations.44-46
This limit
is expected to apply to MWBNNTs as well, because the modulus depends mainly on intra-wall
bonds. Simulations suggest that the Young’s modulus of a MWCNT is slightly higher than that
of a SWCNT, for the same outer diameter, due to the effect of inter-wall van der Waals forces in
MWCNTs.47
The average EB of the MWBNNTs studied in the present investigation matches the
c11 elastic constant of h-BN quite closely. However, the origin of the wide range of bending
moduli and the diameter dependence requires further analysis.
62
Figure 5.7 Bending modulus vs. tube outer diameter. The beam model used for calculating EB is denoted by black
(SSBM), blue (MSBM), and red (DCBM).
5.4.3.2 Investigation of Diameter Dependence: Shear Effects
In experimental studies of multiwall nanotubes a wide range of modulus values has been
measured. For MWCNTs, Treacy et al.21
were first to show that CNTs have Young’s moduli in
the TPa range, using a thermal excitation method. They found that arc-discharge MWCNTs with
outer diameters ranging from 5.4 nm to 24.8 nm had Young’s moduli of 0.4 to 4.15 TPa. A
number of other studies have also produced Young’s moduli in the TPa range, for arc-discharge
MWCNTs.26,27,48,49
Within these studies, despite the focus on the ~1 TPa measurements as
validation of the superior mechanical properties of CNTs, there are many instances of tubes with
lower moduli, on the order of tens to hundreds of GPa. In the work of Salvetat et al.27
catalytic
CVD MWCNTs were also studied and found to have an average modulus of 27 GPa, which is
dramatically lower than the average modulus of 810 GPa measured for arc-discharge MWCNTs.
63
Additional studies also observe lower moduli for catalytic CVD and pyrolytic MWCNTs, in
certain cases as low as tens of GPa.28,50-52
Typically, catalytic CVD and pyrolysis synthesis
methods produce tubes with defective structures compared with the highly crystalline tubes
synthesized by arc-discharge. While point defects do not affect the modulus by more than a few
percent,53
extended defects can cause the modulus to drop by as much as two orders of
magnitude.27,51
In some studies, within sample sets of nanotubes produced under the same conditions, the
modulus is observed to drop with increasing tube diameter. This diameter dependence can be
attributed to three possibilities; namely, the probing of an elastic rippling mode in bending
experiments,49
the presence of defects,28,52
or shear effects. Due to the linearity of the force
curves obtained in the present study, it is unlikely that rippling modes are the cause of the low
moduli measurements observed for larger tubes. Although it was not rare to acquire force curves
which exhibited kinks, potentially due to tube buckling or tip slipping events, fits were only
made to the initial linear portion of the force curves (for deflections less than 10 nm) after
contact. In terms of defects, the low resolution TEM image (Figure 5.1b) shows long, straight
nanotubes with uniform diameters. The high resolution TEM image (Figure 5.1c) shows that the
dark spots present in the tube walls in Figure 1b are crystalline. The MWBNNTs do not appear
to exhibit the type of pronounced structural defects that were found to affect the modulus of
catalytic or pyrolytic MWCNTs, as discussed above. In beam bending experiments, shear must
always be considered for short, stocky beams – those which have a length-to-diameter ratio L/D
< 10. The length-to-diameter ratio L/D was measured to be greater than 10 for all tubes in this
study, which indicates that if shear effects are present, they are not a result of the experimental
geometry. Rather, they can be an indication of a material’s anisotropy.28,54
If shear effects are present, then the bending modulus is not equivalent to the Young’s modulus.
In order to determine whether the Young’s modulus of the MWBNNTs is diameter dependent,
the contribution of shear deflection to the total deflection in the bending experiment must be
quantified. This approach follows Salvetat and coworkers’ bending and shear analysis of single-
wall CNT ropes.55
64
The bending modulus is related to the Young’s modulus EY and the shear modulus G using the
following relationship, determined by Timoshenko beam theory:28,55-57
Equation 5.5
2
211
L
D
G
f
EE
s
YB
In this expression, fs is a shape factor which has a value of 10/9 for a cylindrical beam, and γ is a
shear term coefficient with values of 3, 1.715, and 0.75 for DCBM, MSBM, and SSBM
respectively. The Timoshenko beam theory converges to the Euler-Bernoulli beam theory when
the beam is rigid in shear (G ∞). In this case, the bending modulus is equal to the Young
modulus and is not diameter dependent (which is not the case here). EY and G in our case can be
estimated by plotting 1/EB against (D2/L
2), as shown in Figure 5.8. A linear fit weighted by the
error in 1/EB was obtained for a trimmed data set of 16 tubes. The shear coefficient was taken as
γ = 1.152, determined by the number of tubes exhibiting each type of boundary condition (16
tubes total = 12 simply supported tubes + 2 mixed support tubes + 2 doubly clamped tubes). EY
and G were determined to be 1800 ± 300 GPa and 7 ± 1 GPa, respectively. The expected shear
modulus for a MWBNNT should be on the order of several hundred GPa, based on the
calculations for MWCNTs47,58
which find that GMWCNT ~ 500 GPa. This value is on the order of
the intralayer shear modulus. However, the value of G that we determined for MWBNNTs is
much lower than this, and is close to the value of the c44 elastic constant of h-BN, c44 = 7.7 ± 5,
measured by Bosak et al.43
This elastic constant is equivalent to the interlayer shear modulus of
h-BN, and describes the shear between basal planes. In the case of the MWBNNT structure this
corresponds to shearing between tube walls, which can only occur if there are discontinuities due
to the presence of extended defects within the tube walls.
Although no extensive defects are apparent from the TEM characterization of the MWBNNTs,
as discussed above, the dark spots within the tube walls in the low resolution TEM image (Figure
5.1b) and their somewhat regular pattern within a given tube warrant additional consideration. A
detailed electron diffraction study by Celik-Aktas et al.59
determined that the dark spots can be
attributed to a helical nanotube structure in which the tube is comprised of two or more helices
(each comprised of multiple walls) which wrap to form the entire nanotube. In this structure, the
dark spots correspond to a strongly diffracting helix, which is locally highly crystalline. The
highly crystalline regions are joined together by line defects which result in a faceted helix.
65
Figure 5.8 Determination of the Young’s modulus and shear modulus via a fit to plot of 1/EB vs (D/L)2. The beam
model used for calculating 1/EB is denoted by black (SSBM), blue (MSBM), and red (DCBM).
The lighter regions of the tube wall form the other helix, which possesses the conventional
nested coaxial cylindrical structure expected for multiwall nanotubes. Based on this multi-helix
nanotube structure, it is conceivable that the line defects within the faceted helix as well as the
interface between faceted and cylindrical helices make inter-wall shearing a possibility.
Therefore, as our analysis of the bending data suggests, shear cannot be ignored in the
calculation of the elastic modulus, and shear effects arise from nanotube anisotropy (G << EY)
and the presence of defects within the nanotubes, and not from the experimental geometry.
Our finding that the shear modulus of MWBNNTs is orders of magnitude smaller than the
Young’s modulus indicates that the existing theoretical models are not sufficient in predicting
the mechanical properties of such extremely anisotropic structures, particularly when structural
66
defects are present.47
Experiments performed on MWCNTs support this assertion. Guhados et
al.60
determined that EY = 350 ± 110 GPa and G = 1.4 ± 0.3 GPa for 13 MWCNTs grown by a
CVD method, while Wei et al.61
found that EY ranged from 300 – 900 GPa while G ranged from
30 to 800 MPa, for a sample of 8 tubes. Both studies attribute the low shear modulus to defects
in the structure of the nanotubes. There are several possible benefits of having a low shear
modulus: (1) Taking advantage of its high melting temperature, the shear modulus of
MWBNNTs cast within metals or ceramics would enable damping of vibrations. This could
result in quieter, more durable materials.62-64
(2) Local distortions allowed due to the low shear
modulus could enable MWBNNTs to adapt to local structure variations while maintaining
rigidity on long length scales (longitudinal distortions), imparting toughness to otherwise brittle
composite materials.48
(3) The mutual compensation of shear modulus and Young’s modulus,
where by tubes of different diameters have similar bending stiffness could allow for lower purity
BNNT materials in BNNT coated interfaces for release applications.65,66
(4) With a shear
modulus on the order of the value for h-BN, MWBNNTs can be used as a high-temperature solid
lubricant additive in industrially relevant composites.67
The nanotube structures would have the
added advantages of enabling more efficient heat transport on longitudinal length scales,6 and
increasing the wear resistance of the composite due to reinforcement of the matrix.68
5.5 Conclusion
The bending modulus of individual multiwall boron nitride nanotubes (MWBNNTs) was
measured via AFM bending experiments. Boundary conditions for the beam bending model
were determined by using a force mapping technique. MWBNNTs were found to have excellent
mechanical properties, with an average bending modulus of 760 ± 30 GPa, which is consistent
with the theoretically predicted value for BNNTs. Shear effects were found to be non-negligible,
and the Young’s modulus and shear modulus were determined to be 1800 ± 300 GPa and 7 ± 1
GPa, respectively. The experimental geometry and the dimensions of the nanotubes were not
major contributors to the shear effects; rather, it is likely that inter-wall shearing occurred
between crystalline and faceted cylindrical helices in these MWBNNTs.
67
5.6 Contribution
The author performed the majority of the work described above. A.E.T. carried out the
characterization and AFM measurements and undertook the data analysis to determine the elastic
properties of the MWBNNTs. A.E.T. was the first author on a paper based on the above
submitted to a refereed journal for publication. The MWBNNTs were provided by coauthors J.
Wang and Y.K. Yap, and the preliminary AFM experiments upon which the above work was
based were performed by coauthors A.L.M. Reddy and D. N. Lamont.
5.7 References
1. Rubio, A.; Corkill, J. L.; Cohen, M. L., Theory of graphitic boron-nitride nanotubes.
Phys. Rev. B 1994, 49, 5081-5084.
2. Blase, X.; Rubio, A.; Louie, S. G.; Cohen, M. L., Stability and band-gap constancy of
boron-nitride nanotubes. Europhys. Lett 1994, 28, 335-340.
3. Chopra, N. G.; Luyken, R. J.; Cherrey, K.; Crespi, V. H.; Cohen, M. L.; Louie, S. G.;
Zettl, A., Boron-nitride nanotubes. Science 1995, 269, 966-967.
4. Hernandez, E.; Goze, C.; Bernier, P.; Rubio, A., Elastic properties of single-wall
nanotubes. Appl. Phys., A 1999, 68, 287-292.
5. Verma, V.; Jindal, V. K.; Dharamvir, K., Elastic moduli of a boron nitride nanotube.
Nanotechnology 2007, 18, 435711-6.
6. Bando, Y.; Golberg, D.; Tang, C.; Zhang, J.; Ding, X.; Fan, S.; Liu, C., Thermal
conductivity of nanostructured boron nitride materials. J. Phys. Chem., B 2006, 110,
10354-10357.
7. Chen, Y.; Zou, J.; Campbell, S. J.; Caer, G. L., Boron nitride nanotubes: Pronounced
resistance to oxidation. Appl. Phys. Lett. 2004, 84, 2430-2432.
8. Zhi, C. Y.; Bando, Y.; Wang, W. L. L.; Tang, C. C. C.; Kuwahara, H.; Golberg, D.,
Mechanical and thermal properties of polymethyl methacrylate-BN nanotube composites.
J. Nanomater. 2008, 642036-5.
9. Ravichandran, J.; Manoj, A. G.; Liu, J.; Manna, I.; Carroll, D. L., A novel polymer
nanotube composite for photovoltaic packaging applications. Nanotechnology 2008, 19,
085712-5.
10. Bando, Y.; Golberg, D.; Tang, C.; Terao, T.; Zhi, C. Y., Dielectric and thermal properties
of epoxy/boron nitride nanotube composites. Pure Appl.Chem. 2010, 82, 2175.
11. Choi, S. R.; Bansal, N. P.; Garg, A., Mechanical and microstructural characterization of
boron nitride nanotubes-reinforced SOFC seal glass composite. Mater. Sci. Eng., A 2007,
460–461, 509-515.
12. Lahiri, D.; Singh, V.; Benaduce, A. P.; Seal, S.; Kos, L.; Agarwal, A., Boron nitride
nanotube reinforced hydroxyapatite composite: Mechanical and tribological performance
68
and in-vitro biocompatibility to osteoblasts. J. Mech. Behav. Biomed. Mater. 2011, 4, 44-
56.
13. Ma, R. Z.; Bando, Y.; Zhu, H. W.; Sato, T.; Xu, C. L.; Wu, D. H., Hydrogen uptake in
boron nitride nanotubes at room temperature. J. Am. Chem. Soc. 2002, 124, 7672-7673.
14. Reddy, A. L. M.; Tanur, A. E.; Walker, G. C., Synthesis and hydrogen storage properties
of different types of boron nitride nanostructures. Int. J. Hydrogen Energy 2010, 35,
4138-4143.
15. Ciofani, G.; Raffa, V.; Yu, J.; Chen, Y.; Obata, Y.; Takeoka, S.; Menciassi, A.;
Cuschieri, A., Boron nitride nanotubes: A novel vector for targeted magnetic drug
delivery. Curr. Nanosci. 2009, 5, 33-38.
16. Jaffrennou, P.; Barjon, J.; Lauret, J. S.; Maguer, A.; Golberg, D.; Attal-Tretout, B.;
Ducastelle, F.; Loiseau, A., Optical properties of multiwall boron nitride nanotubes. Phys.
Status Solidi B 2007, 244, 4147-4151.
17. Li, L. H.; Chen, Y.; Lin, M. Y.; Glushenkov, A. M.; Cheng, B. M.; Yu, J., Single deep
ultraviolet light emission from boron nitride nanotube film. Appl. Phys. Lett. 2010, 97,
141104-3.
18. Chowdhury, R.; Adhikari, S., Boron-nitride nanotubes as zeptogram-scale
bionanosensors: Theoretical investigations. IEEE Trans. Nanotechnol. 2011, 10, 659-667.
19. Pampaloni, F.; Florin, E.-L., Microtubule architecture: inspiration for novel carbon
nanotube-based biomimetic materials. Trends Biotechnol. 2008, 26, 302-310.
20. Chopra, N. G.; Zettl, A., Measurement of the elastic modulus of a multi-wall boron
nitride nanotube. Solid State Commun. 1998, 105, 297-300.
21. Treacy, M. M. J.; Ebbesen, T. W.; Gibson, J. M., Exceptionally high Young's modulus
observed for individual carbon nanotubes. Nature 1996, 381, 678-680.
22. Suryavanshi, A. P.; Yu, M. F.; Wen, J. G.; Tang, C. C.; Bando, Y., Elastic modulus and
resonance behavior of boron nitride nanotubes. Appl. Phys. Lett. 2004, 84, 2527-2529.
23. Golberg, D.; Costa, P. M. F. J.; Lourie, O.; Mitome, M.; Bai, X.; Kurashima, K.; Zhi, C.;
Tang, C.; Bando, Y., Direct force measurements and kinking under elastic deformation of
individual multiwalled boron nitride nanotubes. Nano Lett. 2007, 7, 2146-2151.
24. Ghassemi, H. M.; Lee, C. H.; Yap, Y. K.; Yassar, R. S., Real-time fracture detection of
individual boron nitride nanotubes in severe cyclic deformation processes. J. Appl. Phys.
2010, 108, 024314-4.
25. Arenal, R.; Golberg, D.; Wang, M.-S.; Xu, Z.; Loiseau, A., Young modulus, mechanical
and electrical properties of isolated individual and bundled single-walled boron nitride
nanotubes. Nanotechnology 2011, 22, 265704-6.
26. Salvetat, J. P.; Bonard, J. M.; Thomson, N. H.; Kulik, A. J.; Forro, L.; Benoit, W.;
Zuppiroli, L., Mechanical properties of carbon nanotubes. Appl. Phys. A 1999, 69, 255-
260.
69
27. Salvetat, J. P.; Kulik, A. J.; Bonard, J. M.; Briggs, G. A. D.; Stockli, T.; Metenier, K.;
Bonnamy, S.; Beguin, F.; Burnham, N. A.; Forro, L., Elastic modulus of ordered and
disordered multiwalled carbon nanotubes. Adv. Mater. 1999, 11, 161-165.
28. Lee, K.; Lukić, B.; Magrez, A.; Seo, J. W.; Briggs, . A. D.; Kulik, A. J.; Forró, L.,
Diameter-dependent elastic modulus supports the metastable-catalyst growth of carbon
nanotubes. Nano Lett. 2007, 7, 1598-1602.
29. Ni, H.; Li, X. D.; Gao, H. S., Elastic modulus of amorphous SiO2 nanowires. Appl. Phys.
Lett. 2006, 88, 043108-3.
30. Wen, B. M.; Sader, J. E.; Boland, J. J., Mechanical Properties of ZnO Nanowires. Phys.
Rev. Lett. 2008, 101, 175502-4.
31. Shin, M. K.; Kim, S. I.; Kim, S. J.; Kim, S. K.; Lee, H.; Spinks, G. M., Size-dependent
elastic modulus of single electroactive polymer nanofibers. Appl. Phys. Lett. 2006, 89,
231929-3.
32. Shanmugham, S.; Jeong, J.; Alkhateeb, A.; Aston, D. E., Polymer nanowire elastic
moduli measured with digital pulsed force mode AFM†. Langmuir 2005, 21, 10214-
10218.
33. Chen, Y.; Dorgan, B. L.; McIlroy, D. N.; Aston, D. E., On the importance of boundary
conditions on nanomechanical bending behavior and elastic modulus determination of
silver nanowires. J. Appl. Phys. 2006, 100, 104301-7.
34. Kluge, D.; Abraham, F.; Schmidt, S.; Schmidt, H.-W.; Fery, A., Nanomechanical
properties of supramolecular self-assembled whiskers determined by AFM force
mapping. Langmuir 2010, 26, 3020-3023.
35. Gangadean, D.; David, N. M.; Brian, E. F.; Aston, D. E., Winkler boundary conditions
for three-point bending tests on 1D nanomaterials. Nanotechnology 2010, 21, 225704.
36. Lee, C. H.; Wang, J. S.; Kayatsha, V. K.; Huang, J. Y.; Yap, Y. K., Effective growth of
boron nitride nanotubes by thermal chemical vapor deposition. Nanotechnology 2008, 19,
455605-5.
37. Hutter, J. L.; Bechhoefer, J., Calibration of atomic-force microscope tips. Rev Sci Instrum
1993, 64, 1868-1873.
38. Wirtz, L.; Rubio, A., Vibrational properties of boron-nitride nanotubes: Effects of finite
length and bundling. IEEE Trans. Nanotechnol. 2003, 2, 341-348.
39. Heidelberg, A.; Ngo, L. T.; Wu, B.; Phillips, M. A.; Sharma, S.; Kamins, T. I.; Sader, J.
E.; Boland, J. J., A Generalized Description of the Elastic Properties of Nanowires. Nano
Lett 2006, 6, 1101-1106.
40. Iijima, S.; Brabec, C.; Maiti, A.; Bernholc, J., Structural flexibility of carbon nanotubes.
The Journal of Chemical Physics 1996, 104, 2089-2092.
41. Hsu, T. H., Structural Engineering and Applied Mechanics Data Handbook. Gulf
Publishing Company: Houston, 1988; Vol. 1: Beams.
42. Santosh, M.; Maiti, P. K.; Sood, A. K., Elastic properties of boron nitride nanotubes and
their comparison with carbon nanotubes. J. Nanosci. Nanotechno. 2009, 9, 5425-5430.
70
43. Bosak, A.; Serrano, J.; Krisch, M.; Watanabe, K.; Taniguchi, T.; Kanda, H., Elasticity of
hexagonal boron nitride: Inelastic x-ray scattering measurements. Phys. Rev. B 2006, 73,
041402-4.
44. Green, J. F.; Bolland, T. K.; Bolland, J. W., Theoretical elastic behavior for hexagonal
boron-nitride. J. Chem. Phys. 1976, 64, 656-662.
45. Boettger, J. C., All-electron full-potential calculation of the electronic band structure,
elastic constants, and equation of state for graphite. Phys. Rev. B 1997, 55, 11202-11211.
46. Ohba, N.; Miwa, K.; Nagasako, N.; Fukumoto, A., First-principles study on structural,
dielectric, and dynamical properties for three BN polytypes. Phys. Rev. B 2001, 63,
115207-9.
47. Lu, J. P., Elastic properties of carbon nanotubes and nanoropes. Phys. Rev. Lett. 1997, 79,
1297-1300.
48. Wong, E. W.; Sheehan, P. E.; Lieber, C. M., Nanobeam mechanics: Elasticity, strength,
and toughness of nanorods and nanotubes. Science 1997, 277, 1971-1975.
49. Poncharal, P.; Wang, Z. L.; Ugarte, D.; de Heer, W. A., Electrostatic deflections and
electromechanical resonances of carbon nanotubes. Science 1999, 283, 1513-1516.
50. Gaillard, J.; Skove, M.; Rao, A. M., Mechanical properties of chemical vapor deposition-
grown multiwalled carbon nanotubes. Appl. Phys. Lett. 2005, 86, 233109-3.
51. Gao, R. P.; Wang, Z. L.; Bai, Z. G.; de Heer, W. A.; Dai, L. M.; Gao, M.,
Nanomechanics of individual carbon nanotubes from pyrolytically grown arrays. Phys.
Rev. Lett. 2000, 85, 622-625.
52. Lukic, B.; Forro, L.; Seo, J. W.; Bacsa, R. R.; Delpeux, S.; Beguin, F.; Bister, G.;
Fonseca, A.; Nagy, J. B.; Kis, A.; Jeney, S.; Kulik, A. J., Catalytically grown carbon
nanotubes of small diameter have a high Young's modulus. Nano Lett. 2005, 5, 2074-
2077.
53. Sammalkorpi, M.; Krasheninnikov, A.; Kuronen, A.; Nordlund, K.; Kaski, K.,
Mechanical properties of carbon nanotubes with vacancies and related defects. Phys. Rev.
B 2004, 70, 245416-8.
54. Kaplan-Ashiri, I.; Cohen, S. R.; Apter, N.; Wang, Y.; Seifert, G.; Wagner, H. D.; Tenne,
R., Microscopic Investigation of Shear in Multiwalled Nanotube Deformation. The
Journal of Physical Chemistry C 2007, 111, 8432-8436.
55. Salvetat, J. P.; Briggs, G. A. D.; Bonard, J. M.; Bacsa, R. R.; Kulik, A. J.; Stockli, T.;
Burnham, N. A.; Forro, L., Elastic and shear moduli of single-walled carbon nanotube
ropes. Phys. Rev. Lett. 1999, 82, 944-947.
56. Kis, A.; Mihailovic, D.; Remskar, M.; Mrzel, A.; Jesih, A.; Piwonski, I.; Kulik, A. J.;
Benoit, W.; Forro, L., Shear and Young's moduli of MoS2 nanotube ropes. Adv. Mater.
2003, 15, 733-736.
57. Timoshenko, S., Strength of materials. 3d ed.; Van Nostrand: New York, 1955; p 2 v.
58. Shen, L.; Li, J., Transversely isotropic elastic properties of multiwalled carbon
nanotubes. Phys Rev B 2005, 71, 035412.
71
59. Celik-Aktas, A.; Zuo, J. M.; Stubbins, J. F.; Tang, C. C.; Bando, Y., Double-helix
structure in multiwall boron nitride nanotubes. Acta Crystallogr A 2005, 61, 533-541.
60. Guhados, G.; Wan, W.; Sun, X.; Hutter, J. L., Simultaneous measurement of Young's and
shear moduli of multiwalled carbon nanotubes using atomic force microscopy. J. Appl.
Phys. 2007, 101, 033514-5.
61. Wei, X.-L.; Liu, Y.; Chen, Q.; Wang, M.-S.; Peng, L.-M., The very-low shear modulus of
multi-walled carbon nanotubes determined simultaneously with the axial Young's
modulus via in situ experiments. Adv. Funct. Mater. 2008, 18, 1555-1562.
62. Sueyoshi, H.; Rochman, N. T.; Kawano, S., Damping capacity and mechanical property
of hexagonal boron nitride-dispersed composite steel. J. Alloy Compd. 2003, 355, 120-
125.
63 Deng, C. F.; Wang, D. Z.; Zhang, X. X.; Ma, Y. X., Damping characteristics of carbon
nanotube reinforced aluminum composite. Mater. Lett. 2007, 61, 3229-3231.
64. Balani, K.; Agarwal, A., Damping behavior of carbon nanotube reinforced aluminum
oxide coatings by nanomechanical dynamic modulus mapping. J. Appl. Phys. 2008, 104,
063517-6.
65. Lipp, A.; Schwetz, K. A.; Hunold, K., Hexagonal boron nitride: Fabrication, properties
and applications. J. Eur. Ceram. Soc. 1989, 5, 3-9.
66. Haubner, R.; Wilhelm, M.; Weissenbacher, R.; Lux, B., Boron Nitrides - Properties,
Synthesis and Applications. In High Performance Non-Oxide Ceramics II, Jansen, M.,
Ed. Springer Verlag: Berlin, Heidelberg, 2002; pp 1-45.
67. Chen, W.; Gao, Y.; Chen, C.; Xing, J., Tribological characteristics of Si3N4–hBN ceramic
materials sliding against stainless steel without lubrication. Wear 2010, 269, 241-248.
68. Zhou, S.-M.; Zhang, X.-B.; Ding, Z.-P.; Min, C.-Y.; Xu, G.-L.; Zhu, W.-M., Fabrication
and tribological properties of carbon nanotubes reinforced Al composites prepared by
pressureless infiltration technique. Composites Part A 2007, 38, 301-306.
Equation 6 fg
Table 5. sdf
Table 6. sd
72
6 Insights into the composition, morphology, and formation of the calcareous shell of the serpulid Hydroides dianthus
6.1 Permissions
The material in this chapter is presented with permission from Tanur, A. E.; Gunari, N.; Sullan,
R. M. A.; Kavanagh, C. J.; Walker, G. C. J. Struct. Biol. 2010 169, 145-160. Copyright 2009
Elsevier Inc.
6.2 Abstract
To date, the calcareous tubes of serpulid marine worms have not been studied extensively in a
biomineralization context. The structure and composition of the tube shell and adhesive cement
of the marine tubeworm Hydroides dianthus were studied using a variety of characterization
techniques, including powder XRD, FTIR, SEM, EDX, and AFM. The tube and cement were
determined to be inorganic-organic composite materials, consisting of inorganic aragonite
(CaCO3) and Mg-calcite ((Ca0.8,Mg0.2)CO3) crystals, and both soluble and insoluble organic
matrices (SOM and IOM). SEM imaging revealed a variety of crystal morphologies. AFM
nanoindentation of the inorganic components yielded Young’s moduli of ~20 Pa in the wet
state, and ~50 GPa in the dry state. Amino acid analysis of the SOM indicated substantial
amounts of acidic and non-polar neutral amino acids. Part of the insoluble organic tube lining
was identified as being composed of collagen-containing fibres aligned in a criss-crossed
structure. The SOM and organic tube lining were found to contain carboxylated and sulphated
polysaccharides. In an artificial seawater solution, the SOM and the organic tube lining
mediated CaCO3 mineralization in vitro.
6.3 Introduction
Recently, much interest has been paid to the structure and formation of inorganic-organic
composite materials by various organisms. Biomineralization processes in nature produce
inorganic crystalline materials with exceptionally fine control over nucleation and growth
processes, and overall material organization. All of this is achieved under ambient conditions
73
largely through the use of organic macromolecules, which are also incorporated into the
crystalline structures to form the inorganic-organic composite material.1-3
Calcium carbonate, in both calcitic and meta-stable aragonitic forms, is one of the most common
and abundant biominerals and is widely used by many marine species, such as in certain
echinoderms, brachiopods, scleractinian corals, and molluscs as a structural material.4
Numerous studies have been conducted on the material structure and properties of nacre, sea
urchin spines, oyster shell folia, and coral, and models for the biomineralization processes in
each of these organisms have been proposed and developed.5-11
Unlike the aforementioned
marine invertebrates, the calcareous tube-building marine worms in the serpulid family
(Annelida, Polychaeta) have not been studied extensively with regards to CaCO3
biomineralization. Serpulids are sessile animals that spend their adult lives within tubes they
have built, which are permanently attached to a substrate (Figure 6.1). The thin material which
is in contact with the substrate is referred to in this paper as the adhesive material, and the thicker
cylindrical portion is referred to as the tube shell. The tubes are composed of calcite, aragonite,
or a combination of these two CaCO3 polymorphs, although typically one polymorph is
dominant. Organic sheets have also been occasionally observed within the tube shell walls and
lining the inside of the tubes.12, 13
Figure 6.1 Overview of serpulid tube structure (Hydroides dianthus). A) Cartoon of tubeworm, transverse cross section. Thickness of adhesive material is exaggerated circa 20X, relative to the shell wall. B) Bottom view of
tubeworm, with white coloured tube walls, and light brown (grey in Figure) adhesive materials. The wider end of the tube is the head section (most recent growth), and the thinner end is the tail section (oldest part of the tube).
Ultrastructural studies of tube shell cross-sections from a number of recent and fossil serpulid
species by Vinn et al.12
indicate that roughly 75% of serpulid tubes consist of a single layer, with
74
the remaining 25% being composed of two to four layers. Layers are classified by the overall
orientation and organization of the inorganic crystals making up the tube. The current model for
serpulid tube formation involves the secretion of calcareous granules and organic components
from the calcium-secreting glands in the anterior portion of the worm, located on the
peristomium near the base of the collar. The glandular openings are within a region known as
the ventral shield, whose epithelial mucocytes secrete additional organic components. New
material is secreted and added to the anterior end of the existing tube, and moulded by the
worm’s collar.14-18
Although the calcium-secreting glands play a pivotal role in the tube
formation process, it is unclear how certain serpulids achieve layered structures within the tube,
some with very intricate organization, through this mechanism alone. It has been hypothesized
that extracellular mediation may be involved, such as in the case of molluscs, but to date the
macromolecules and their roles in the biomineralization process have not been studied in
detail.12, 19
In this paper, we identify several organic components of the calcareous tube of the serpulid
Hydroides dianthus, and provide evidence for their participation in an extracellular
biomineralization mechanism. In addition, we present a description of both the tube shell and
the adhesive material in terms of their ultrastructure, crystal polymorph and morphology, and
mechanical properties.
6.4 Experimental Methods
6.4.1 Tubeworm Collection and Preservation
Tubeworms were collected near the Sebastian inlet in the Indian River lagoon system in Florida,
USA (27 oN, 80
oW). Rectangular 10.16 x 20.32 cm (4 x 8 in) glass panels were coated with a
polydimethylsiloxane-based anti-fouling material (Sylgard 184, Dow Corning, Midland, MI) in
thicknesses ranging from 110 – 380 μm. The panels were fastened to PVC frames and
suspended from a floating raft at a depth of 1 m. The panels were submerged until they were
encrusted with a variety of marine biofoulers, including H. dianthus.
Upon removal from the lagoon, the encrusted panels were kept moist and immediately shipped
overnight to the University of Toronto on September 9th
, 2008. The panels were placed into a 45
L aquarium filled with 35 ppt artificial seawater (Instant Ocean®
Sea Salt, Marineland,
75
Cincinnati, OH) and kept at room temperature. The tubeworms were fed daily with rotifers (Bio-
Pure, Hikari, Hayward, CA). For all experiments, tubeworms were pushed off of the substrate in
shear, and the worms removed. For experiments in which the samples were required to be in a
dry state, tube shells were rinsed with deionized water and then dried in an oven set at 80 oC for
1 hour, under 30 mmHg vacuum.
6.4.2 X-Ray Diffraction (XRD)
Dry tube shells were crushed into a fine powder using a mortar and pestle. Two types of samples
were prepared, one consisting of both the shell and adhesive material from a single tube, and the
other consisting of only adhesive material. For the latter sample type, the adhesive material was
collected from 3 different worms on the same panel, and ground up separately from the rest of
the tube material. Samples were subjected to powder XRD analysis, using an automated
diffractometer (AXS D5000, Siemens/Bruker, Madison, WI) system and Cu-Kα radiation (50
kV, 35 mA). A solid-state Si/Li Kevex detector was used for the removal of Kβ lines.
Diffraction patterns were collected on a θ/2θ Bragg-Brentano reflection geometry, with fixed
slits. A step scan mode was used for data acquisition, with a step size of 0.02o 2θ, and a counting
time of 2.5 s per step. Qualitative identification of the crystalline components was performed
using the Search/MatchTM
routine, part of the data processing software EvaTM
v.8.0
(Siemens/Bruker, Madison, WI). Rietveld refinement was carried out with profile fitting
software (TopasTM
, Siemens/Bruker, Madison, WI).
6.4.3 Fourier Transform Infrared Spectroscopy (FTIR)
Powdered tube shell samples (from the same samples that were used for XRD analysis) were
dispersed in KBr pellets. Spectra were obtained with a Fourier transform infrared spectrometer
(FTIR) (Spectrum BX, Perkin-Elmer, Waltham, MA), using a resolution of 2 cm-1
. Spectra of
the organic tube lining were taken using an attenuated total reflection set-up. The soluble
organic matrix (SOM) was deposited onto a CaF2 window and allowed to dry under ambient
conditions, and its spectrum was taken in transmission mode. For the in vitro mineralization
experiments, the Au substrates were analyzed via FTIR using a reflectance set-up.
76
6.4.4 Inductively Coupled Plasma Atomic Emission Spectrometry (ICPAES) Analysis
25 mg of powdered tube shell was dissolved in 5 mL of concentrated HCl and diluted with
deionized water to 100 mL in a volumetric flask. Solutions were assayed by Inductively
Coupled Plasma Atomic Emission Spectrometry (ICP AES), on a spectrometer (Optima
3000DV, Perkin-Elmer, Waltham, MA).
6.4.5 Electron Probe Microanalysis (EPMA)
Polished cross-sectioned samples were carbon coated and subjected to electron probe
microanalysis (EPMA), using an electron microprobe instrument (SX50, Cameca, France). Ca
and Mg molar ratios were calculated and averaged from measurements at six different locations
within the tube shell cross-section.
6.4.6 Separation of the Organic Tube Lining and the Soluble Organic Matrix (SOM)
After removal of the worm from the tube, the inner organic lining of the tube was removed with
tweezers. The lining constitutes part of the insoluble organic matrix (IOM) of the tube shell
along with other organic sheets within the tube which were not extracted. After separation from
the tube shell, the organic tube lining was subjected to treatment with 0.5 M EDTA overnight.
For the extraction of the SOM, 1 mL of 0.5 M EDTA (adjusted to a pH of 8) was added to 110
mg of powdered tube shell in an eppendorf tube. The sample was vortexed for 1 min every 30
min for 6 hrs, and subsequently refrigerated at 4 oC overnight. Afterwards, the eppendorf tube
was centrifuged for 10 min at 13.5 g. The supernatant was removed and diluted with deionized
water to form a 0.05 M EDTA solution, which was then used for the AA analysis and the in vitro
mineralization experiments.
6.4.7 Scanning Electron Microscopy (SEM) and Energy Dispersive X-Ray (EDX)
For cross-sectioned samples, tube shells were embedded in a low-viscosity embedding media
(Spurr, Canemco, Quebec, Canada), sectioned, and polished. Samples were etched for 1-2 min
in 1% acetic acid. For measurements on the tube adhesive material, tube shells were fixed using
77
a 0.1M glutaraldehyde-paraformaldehyde phosphate buffer (pH 7.2) solution, serially dehydrated
using 50%, 70%, 90%, and 100% ethanol solutions, then critical point dried with CO2 using a
critical point dryer (Autosamdri 810, Tousimis Research Corporation, Rockville, MD). All
samples were mounted directly onto aluminum stubs using carbon tape and/or carbon paint, and
carbon coated. Measurements were obtained with a field emission scanning electron
microscope (SEM) (S-5200, Hitachi, Japan) equipped with an energy dispersive x-ray (EDX)
system (Inca, Oxford Instruments, United Kingdom). EDX measurements were performed at 10
kV, with a probe current of 20 μA.
6.4.8 Atomic Force Microscopy (AFM) Imaging and Nanoindentation
Tapping mode AFM images and nanoindentation measurements were obtained in 35 ppt artificial
seawater employing atomic force microscopy (AFM) (MFP 3D, Asylum Research, Santa
Barbara, CA). AFM images were obtained using V-shaped silicon nitride cantilevers (MLCT-
AUNM, Veeco Inc., Santa Barbara, CA) with nominal spring constants of ~ 0.5 N/m.
The mechanical properties of the tubeworm shell and the adhesive region were measured by
nanoindentation experiments using rectangular shaped silicon tips (NCH_W, Nano World,
Switzerland). Prior to use, individual cantilever spring constants were determined using the
thermal noise spectrum method.20
The spring constant was determined to be 39 N/m. The
resulting force-indentation curves were analyzed with custom-programmed analysis software
(Igor Pro, Wavemetrics, Portland, OR), as described previously.21, 22
Indentation measurements on the adhesive section were carried out by first obtaining several 20
x 20 μm images in a ~8000 μm2 region in artificial seawater using a contact-mode tip (0.01 N/m)
and then replacing the tip with a rectangular silicon cantilever (39 N/m) for the mechanical tests.
In addition to the measurements performed on the adhesive material, unetched polished cross-
sectioned samples were also examined via nanoindentation, in order to minimize the error in the
determination of the modulus due to topographic variations. Indentation measurements were
also performed on the cross-sectioned samples in the dry state as well as the wet state. The
polished samples were immersed in artificial sea water for 8 days prior to the measurements
performed in the wet state.
78
Tapping mode AFM imaging in air was performed on samples of the organic tube lining. For
these samples, the IOM was sonicated in DI water for 2 min (Bransonic 2510, Branson,
Danbury, CT) in order to break it up into fragments. The fragments were then deposited onto
freshly cleaved mica substrates, and allowed to dry before imaging.
6.4.9 Light Microscopy and Chemical Staining
A light microscope (CME, Leica Microsystems Inc., Bannockburn, IL) was used for optical
imaging of the organic tube lining. Thin sections of the lining were glued to a piranha cleaned
glass slide using epoxy glue. The glue was allowed to cure for 2 hours at room temperature in a
laminar flow hood. The glass slide was then washed with ethanol and then rinsed in distilled
water. The tube lining was then subjected to Masson’s trichrome staining (Trichrome stain AB
solution, Sigma-Aldrich, St. Louis, MO), per the Sigma-Aldrich protocol.
6.4.10 Amino Acid Analysis of the Soluble Organic Matrix (SOM)
Amino acid analysis of the EDTA extract was performed on an UltraPerformance Liquid
Chromatograph (UPLC) (Acquity, Waters Corporation, Milford, MA). Dried protein sample
was hydrolyzed by a vapour phase or liquid phase reaction, using 6 M HCl with 1% phenol at
110 oC for 24 hrs. Glycoprotein sample was hydrolyzed using 6 M HCl with 1% phenol at 110
oC for 28 hrs (to quantify amino sugars and obtain accurate serine measurements). A work
station (PICO-TAG, Waters Corporation, Milford, MA) was used for the drying and hydrolysis
procedures. After hydrolysis, the excess HCl was removed from the hydrolysis tube under
vacuum, and the sample was treated with a re-drying solution consisting of
methanol:water:triethylamine (2:2:1) and dried under vacuum. The sample was then derivatized
for 20 min at room temperature using a derivatizing solution of
methanol:water:triethylamine:phenylisothiocyanate (PITC, 7:1:1:1). The solution was removed
under vacuum, and the sample was treated once again with the re-drying solution and dried to
remove any traces of PITC.
The derivatized sample was dissolved in a given amount of sample diluents (pH 7.4) and an
aliquot was injected into the UPLC BEH C18 column (1.7 μm, 2.1 mm x 100 mm) running on a
gradient for UPLC. The column temperature was 48 oC. The PITC-amino acids were detected at
254 nm. Data acquisition and collection was obtained using a TUV detector. The whole system
79
was controlled using chromatography software (Empower 2.0, Waters Corporation, Milford,
MA).
6.4.11 In vitro CaCO3 Mineralization Experiments
For the organic tube lining remineralization experiment decalcified lining material was rinsed in
distilled water and then placed in an uncovered crystallization dish containing 4 mL of 35 ppt
artificial seawater solution. For the SOM mineralization experiment, 50 μL of the SOM extract
was deposited onto a flame-annealed Au substrate, and allowed to adsorb for 10 min.
Afterwards, the substrate was placed into an uncovered crystallization dish to which 4 mL of 35
ppt artificial sea water solution was added. Flame-annealed Au substrates (no SOM deposition)
were placed into dishes with 4 mL of 35 ppt artificial seawater solution as well, as controls.
Crystallization was performed using the ammonium carbonate diffusion method.23
The dishes
were placed in a chamber containing powdered (NH4)2CO3 and sealed using parafilm.
Precipitates were collected after various time intervals. For the organic tube lining experiment,
the IOM was removed after 24 hr, and for the SOM experiments, the Au substrates were
removed after 90 min and after 48 hr. Samples were rinsed in distilled water and allowed to dry
at ambient conditions.
For reference, the chemical compositions of sea water and artificial seawater are given in Table
6.1.
Table 6.1 Chemical composition of seawater vs. the artificial seawater used in all CaCO3 precipitation experiments. (Atkinson, M.J., Bingman, C., 1997. Elemental composition of commercial seasalts. J. Aquaricult. Aquat. Sci. 8, 39-
43.)
Salinity Na Mg Ca K Sr Cl- SO4- BO3
HCO3-,
CO32-
Molar Mass
(g/mol)
Seawater 35 470 53 10.3 10.2 0.09 550 28 0.42 1.90 22.9898
Instant
Ocean 29.65 462 52 9 9.4 0.19 521 23 0.44 1.90 24.305
80
6.5 Results
6.5.1 Bulk Composition of the Tube Shell and Adhesive Material
The XRD spectra for the tube shell and the tube adhesive both showed the same peaks,
indicating that the crystal structure and distribution is similar for the bulk tube shell material and
for the adhesive material. The XRD spectrum for the bulk tube shell is shown in Figure 6.2A.
The peaks correspond to a combination of two polymorphs of CaCO3; namely, aragonite and
magnesium calcite (Mg-calcite). After Rietveld refinement of the spectrum, the relative
proportions of the two polymorphs were found to be 40 ± 1 % aragonite, and 60 ± 1 % Mg-
calcite. The unit cell parameters of the Mg-calcite component were found to be a = 4.921 ±
0.002 Å and c = 16.714 ± 0.007 Å (rhombohedral unit cell, space group R c), with a
corresponding unit cell volume of 350.5 ± 0.2 Å. Based on the site occupancies of Ca and Mg in
the 6b position, the empirical formula of the Mg-calcite was determined to be (Ca0.8Mg0.2)CO3.
Figure 6.2 A) XRD spectrum for powdered sample of entire tube. Aragonite and Mg-calcite peaks are labeled A and C respectively. B) FTIR spectrum for powdered sample of entire tube. Inset: Features of ν4 bands.
81
The FTIR spectra for the tube shell and the tube adhesive exhibited the same features, and thus
only the tube shell spectrum is shown in Figure 6.2B. As shown in Table 6.2, the peak positions
of the carbonate ion modes confirm the presence of both aragonite and calcite. Comparison of
the observed calcite peaks with those of an abiogenic calcite sample indicates that the biogenic
sample contains Mg-calcite. Shifts in the peak positions and line width broadening of the ν2 and
ν4 modes, as compared to those of pure calcite, are characteristic of the presence of magnesium
in the calcite structure. In particular, the ν4 mode is ideal for estimation of the Mg content in Mg-
calcite because it is both a function of composition and independent of formation conditions.24, 25
The ν2 and ν4 modes in pure calcite are 876 cm-1
and 712 cm-1
, respectively, and these modes are
shifted to 873 cm-1
, and 713cm-1
, with a shoulder at 723cm-1
(see inset of Figure 6.2B). Given
that the ν4 mode of aragonite is at 713 cm-1
, the shoulder at 723 cm-1
is attributed to Mg-calcite.
Using the relationship developed by Bottcher et al.,25
ν4 [cm-1
] = 39.40XMg + 712.20, our FTIR
results indicate a Mg content of ~27 mol%. However, the relationship is only valid for XMg <
0.23; hence, qualitatively the FTIR data confirms the XRD identification of a high Mg-calcite
present in the tubeworm shell.
Table 6.2 Summary of IR bands for the tube shell sample FTIR spectrum and comparison with geological calcite and aragonite samples, in units of wavenumbers (cm
-1). Abbreviations: vvs = very very strong, vs = very strong, s =
strong, m = medium, vw = very weak, sh = shoulder, br = broad.
Carbonate ion mode H. dianthus Geological Calcite Geological Aragonite
ν1 + ν4 ~1796 (vw,br) 1800 (m) 1788 (m)
ν3 1482 (vvs), 1429 (sh) 1422 (vvs) 1477 (vvs)
ν1 1082 (m) - 1083 (s)
ν2
873 (vs) 876 (vs) -
858 (s) - 855 (vs)
844 (sh) 848 (sh) 843 (sh)
ν4
723 (sh,br), 713 (m) 712 (s) 713 (s)
700 (m), ~693 (sh,br) 696 (sh) 700 (m)
82
The presence of Mg was also confirmed by electron probe microanalysis and ICP analysis,
which yielded estimates of 16 ± 2 mol% and ~11 mol%, respectively. The ICP result is low in
comparison to the other results because the aragonite phase could not be separated from the
calcite phase for the analysis.
6.5.2 Tube Shell Ultrastructure and Spatial Composition
An overview of a transverse tube cross-section is shown in Figure 6.3A. The adhesive material
is only a few tens of microns thick; in contrast, the tube shell is on the order of a hundred to
several hundred microns thick. Concentric layers are evident within the tube shell, and the
layers terminate roughly perpendicular to the adhesive material at the base of the shell. Figure
6.3B shows the cross-section of one of the tube shell walls in greater detail. The direction of the
exterior (ext) and interior (lumen) of the tube shell is indicated by the arrow label. As previously
described by Vinn et al.,26
the shell of H. dianthus consists of three distinct layers, each with a
specific crystal organization. In the following description, we use the terminology established by
Vinn et al.12
The outermost layer, in the top left of Figure 6.3B, possesses a spherulitic prismatic
structure (SPHP). Below this is a layer with an irregularly oriented prismatic structure (IOP).
The last layer, terminating in the inner wall of the tube shell (shown as the bottom two-thirds of
Figure 6.3B), has a lamello-fibrillar (LF) structure, in which the orientation of crystals changes
with successive growth increments.
Figure 6.3C-F show the structural details of the three layers and the transitions between layers.
The local chemical composition of or within the layers is indicated, as determined through EDX.
First, in Figure 6.3C, the SPHP layer was found to be composed of aragonite (bottom left in
image), while the IOP layer was a mixture of aragonite (adjacent to the SPHP layer) and Mg-
calcite (top left in image). Next, Figure 6.3D shows the interface between the IOP layer (bottom
left of image) and the LF layer (top right of image). Despite the difference in crystal
morphology and organization, both regions were found to consist of Mg-calcite. Inside the LF
layer, Figure 6.3E shows that both aragonite and Mg-calcite can be present in this particular
structure. Lastly, Figure 6.3F shows the LF layer terminating into the inner wall of the shell
(upper right), on which there is an organic layer (the tube lining). In the lower left of the image,
among the Mg-calcite crystals, another organic layer is present. Both organic layers appear
smooth and unaffected by the acetic acid etching process.
83
84
Figure 6.3 SEM images of tube shell transverse cross section, showing a layered structure with each layer consisting of a different crystal morphology and organization, and/or polymorph. The head of the arrows indicate the direction of the tube interior (lumen). In Figs. C-F, the local chemical composition is indicated by A – aragonite, MC – Mg-calcite,
O = organic. A) Large scale view of cross section, with fragments of adhesive material also visible. The arrow indicates the region over which Figs. C-F were taken. B) Tube shell wall, showing the three distinct layers (SPHP =
spherulitic prismatic structure, IOP = irregularly oriented prismatic structure, LF = lamello-fibrillar structure). C) Details of outer tube wall, showing the SPHP and IOP layers. D) Details of the middle region of the tube wall,
showing the transition from the IOP layer to the LF layer. E) Details of the LF layer, showing the presence of both aragonite and Mg-calcite within the layer. F) Details of the inner tube wall, showing several organic membranes. G) Representative EDX spectrum for Mg-calcite crystals. H) Representative EDX spectrum for aragonite crystals. I)
Representative EDX spectrum for organic material.
Representative EDX spectra for Mg-calcite, aragonite, and organic material are shown in Figure
6.3G-I. Any Cd signal present in the spectra is the result of contamination within the SEM
chamber. Both CaCO3 polymorphs exhibit strong Ca peaks as well as C and O peaks, but can be
distinguished by the presence of a Mg peak in the case of Mg-calcite, and a Sr peak (with no Mg
peak visible) in the case of aragonite. Sr is able to substitute into the aragonite lattice because
the Ca-O distances are larger than they are in the calcite lattice.27
The organic membranes are
distinguished by high C, O, and S signals.
6.5.3 Adhesive Material Structure and Composition
The adhesive material has an apparently different structural organization than the tube shell.
Although its thickness is comparable to that of the SPHP layer (on the order of 10-20 μm), it
appears to be made up of multiple thinner layers, each several microns thick, as shown in Figure
6.4. The adhesive layers appear crystalline, like the tube wall layers, but a greater amount of
organic material (sheet and fibre features) is associated with the crystals, as seen in Figure 6.4B-
D. The transition between the bulk tube wall crystals and the adhesive material is shown in
Figure 6.4A, with an organic sheet separating the two regions. The adhesive material at the
outermost edge of the tube wall has a layer consistent with the outer SPHP structure (Figure
6.4B). Cross sections of the adhesive fragments ‘a’ and ‘b’ (labelled in Figure 6.3A) are shown
in Figure 6.4C and 3D respectively.
A wide variety of different crystal morphologies was observed within the adhesive material, with
several distinct forms observed for both Mg-calcite and aragonite polymorphs. Figure 6.5A-C
show SEM images of Mg-calcite crystals and Figure 6.5D-F are images of aragonite crystals, as
determined by EDX. Besides the rhombohedral crystal form shown in Figure 6.5C, Mg-calcite
can also take the form of cauliflower-like aggregates made up of triangular layers (Figure 6.5A),
and dumbbells, shown in various stages of growth in Figure 6.5B.
85
Figure 6.4 SEM images of adhesive material, transverse cross section. Layers are evident, some with apparently different structures than those observed in the bulk tube. Sample subjected to a 2 min etching with 1% acetic acid.
A) Base of tube wall, showing transition between adhesive region and the crystals of the bulk wall. B) Cross section of adhesive material extending out from the outer tube wall. C) Cross section of adhesive fragment ‘a’ (labeled in
Fig. 6.3A). D) Cross section of adhesive fragment ‘b’ (from Fig. 6.3A).
In fact, the latter two forms were observed much more often than the rhombohedral form. Most
aragonite crystals observed had an acicular habit, whether in a randomly oriented arrangement as
in Figure 6.5D, or in bundles with a common orientation, as in Figure 6.5E. However, globular
aggregates of aragonite were also observed, as shown in Figure 6.5F. Crystals observed on the
substrate side of the adhesive material were often found to be associated with thin fibre and sheet
structures (possibly organic), as shown in Figure 6.5C and Figure 6.5E-F. Bacteria and diatoms
were also observed on or embedded within the adhesive material, as shown in Figure 6.6.
86
Figure 6.5 SEM images of the variety of crystal morphologies observed for A-C) Mg-calcite and D-F) aragonite.
6.5.4 Mechanical Properties of the Adhesive Material
The mechanical properties of the tubeworm adhesive material were measured using AFM
indentation measurements. The Young’s modulus was determined by considering load-
indentation dependence for a paraboloidal tip shape given by Equation 6.1:
Equation 6.1 2/3
2 )1(3
4
v
REF Y
Here, F is the loading force in nN, EY is Young’s modulus in Pa, R is the radius of curvature of
the tip in nm, is the indentation in nm, and is Poisson’s ratio. In order to estimate the
Young’s modulus, a value of 0.2 was used for Poisson’s ratio, derived from averaging the bulk
Poisson’s ratios for calcite and aragonite (0.32 and 0.16, respectively).28, 29
AFM topography images collected in seawater of defects in the adhesive material surface which
was in contact with the substrate are shown in Figure 6.7A and Figure 6.7B. The defects
illustrate the different crystal morphologies and orientations present in the material. SEM
images of the adhesive material surface, showing similar features to those observed in the AFM
images, are shown in Figure 6.8.
87
Figure 6.6 SEM images of the adhesive material surface, substrate side, showing the presence and incorporation of various biofilm components. A-B) Bacteria among crystals and on surface. C-D) Diatoms embedded in adhesive
material. E-F) Fibrous networks on the surface and within the adhesive material, some showing signs of mineralization.
88
Figure 6.7 A) 5 x 5 μm and B) 3 x 3 μm AFM height images of adhesive material surface (substrate side), taken in solution (sea water). C) 20 x 20 μm AFM height image of adhesive material area, representative of area over which indentation measurements in solution were taken (adhesive material). D) “Wet state, adhesive”: Histogram of elastic moduli derived from indentation measurements in solution (adhesive material, substrate side). E) “Wet state, shell”: Histogram of elastic moduli derived from indentation measurements in solution (transverse tube shell cross-section,
polished). F) “Dry state, shell”: Histogram of elastic moduli derived from indentation measurements in air (transverse tube shell cross-section, polished).
Figure 6.7D shows a histogram of elastic moduli obtained via nanoindentation measurements
made over the area shown in the representative AFM topographic image Figure 6.7C. A
modulus of 3 ± 1 GPa was obtained from these measurements.
6.5.5 Mechanical Properties of the Tube Shell
Nanoindentation on polished cross-sectioned samples of the tube shell in artificial seawater
solution and in air was performed, yielding Young’s modulus values of 22 ± 3 Pa and 51 ± 8
GPa respectively, as shown in the histograms in Figure 6.7E-F.
6.5.6 Characterization of the Organic Tube Lining
The insoluble organic tube lining was characterized via light microscopy, SEM, chemical
staining, FTIR, and AFM. Figure 6.9A shows a partially broken tube, revealing the lining which
also has a tubular form. Figure 6.9B-C show the insoluble organic tube after removal from the
calcareous tube shell, at 10X magnification (Figure 6.9B) and 40X magnification (Figure 6.9C).
89
Figure 6.8 SEM images illustrating the different crystal morphologies and orientations present at the adhesive material surface, adjacent to the substrate. All images were obtained away from the tube walls. Note the smooth
regions where the crystal growth was terminated by the substrate. A) Columnar-type crystals, roughly perpendicular to the substrate. B) Spherulitic needles, terminating at the substrate in various angles. C) Aggregates with a layered
triangular habit. D) Bundles of acicular crystals, amongst smaller crystals with a triangular habit.
Figure 6.9 Optical images of organic tube lining. A) Partially broken tube, showing exposed organic matrix, 10X magnification. B) Tube lining after removal from the tube, 10X magnification. C) IOM, 40X magnification.
90
SEM observations of the lumen-side of the adhesive material fragments reveal that criss-crossed
layers of fibres are located on the inner wall of the tube, and that the fibres are covered by a
smooth thin layer, possibly an organic sheet, as shown in Figure 6.10A. The smooth layer is
shown in detail in the inset of Figure 6.10A, and EDX measurements indicate the presence of C,
O, Ca, Mg, Na, P, and S (Figure 6.11). A region in which a layer of fibres is in intimate contact
with calcareous crystals is shown in Figure 6.10B. The structure of the fibres is more clearly
seen in Figure 6.10C. Each ~3 μm fibre is made up of many smaller fibres, some in what
appears to be a twisted conformation. EDX measurements of the fibres yielded results similar
those obtained for the smooth layer, with varying levels of S. Cross-sectional details of the
organic layer lining the inside of the tube are shown in Figure 6.10D. Layers of fibre and sheet
structures are apparent.
Figure 6.10 SEM images of the adhesive material, lumen-side (A-C) and of the EDTA-treated tube lining (D). Fibre structures similar to those observed via optical microscopy are evident. A) Fibres covered by a smooth sheet-like structure. Inset: Details of sheet surface. B) Fibres associated with crystalline aggregates. C) Details of fibres,
showing them to be composed of smaller fibres. D) Cross-section of the EDTA-treated organic tube lining, showing layers of fibres and sheets.
91
Figure 6.11 EDX spectrum for the “smooth layer” (possibly an organic sheet) shown in the inset of Fig. 6.10A, showing the presence of C, O, Ca, Mg, Na, P, and S. The Cd signal is due to contamination within the SEM
chamber.
Masson’s trichrome staining was performed in order to identify the organic tube lining. Some
areas were observed to be stained blue, which indicates a positive result for the presence of
collagen, as shown in Figure 6.12. Sonicated fragments of the lining were deposited onto mica
substrates and examined via AFM, in order to characterize the diameter and structure of the
smaller fibres. Two types of fibres were distinguished via AFM, as shown in Figure 6.13. The
first type consisted of fibres with a diameter ~ 150 nm, which exhibited banding with a 67 nm
spacing, which is similar to vertebrate collagen type I.30
The second type was unstriated, with a
diameter of ~100 nm.
Figure 6.12 Optical image of the insoluble organic matrix after Masson’s trichrome staining, 40X magnification. The blue colour indicates a positive result for the presence of collagen.
92
Figure 6.13 1 x 1 μm AFM amplitude images and linescans of fibres from the insoluble organic matrix. A) Striated fibre. B) Smooth fibre.
The FTIR spectrum of the tube lining is given in Figure 6.14, and the observed band frequencies
are summarized in Table 6.3. The absence of calcite bands at 1416 cm-1
, 870 cm-1
, and 719 cm-1
(ν3, ν2, and ν4 carbonate modes) indicate that the lining is fully de-mineralized. The dominant
features of the spectrum appear to be signatures of carbohydrates. The broad peak at 3284 cm-1
is due to bonded –OH, while the components in the range of 2850 – 2950 cm-1
are attributed to
CH stretching. In the 1000 – 1100 cm-1
region, the band at 1096 cm-1
arises from C-O and C-C
stretching in pyranose rings, while the peaks at 1030 and 1007 cm-1
are assigned to C-O-C and
C-C stretches, somewhat similar to modes observed in α and β-glucan.31, 32
Peaks at 797 and 777
cm-1
could be due to α and β-galactose skeletal bending.33
The strong bands at 1594 and 1400
cm-1
are attributed to –COO- (carboxylic acid salt), while the bands at 1323 and 1231 cm
-1
correspond to sulphate ester vibrations.34
93
Figure 6.14 FTIR spectra of the tube shell organic matrices. A) Organic tube lining spectrum. B) SOM spectrum.
6.5.7 Characterization of the SOM
The FTIR spectrum of the SOM is given in Figure 6.14B, and a summary of the bands is
presented in Table 6.3 (below). The broad bands at 3377 and 3260 cm-1
are due to bonded –OH,
and the bands around 2850 – 2950 cm-1
are assigned as above in the organic tube lining FTIR
spectrum. Carbohydrate signatures arising from C-C, C-O-C, and C-O stretches are present in
the region between 980 – 1050 cm-1
. The lower frequency components at 925 and 851 cm-1
are
assigned to ring vibration and α-anomeric CH deformation, respectively. The band at 851 cm-1
may also be assigned to a C-O-S bending mode of sulphate esters, and additional sulphate ester
modes are observed at 1332, 1258, 1181, and 1114 cm-1
. Carboxylic acid salt (-COO-) bands are
observed at 1601 and 1407 cm-1
, and the bands at 1739 and 963 cm-1
are assigned to COOH
modes.
94
Table 6.3 Summary of IR bands for the organic tube lining and the SOM FTIR spectra.
IR Band (cm-1
) Assignment IR Band (cm-1
) Assignment
Organic
Tube
Lining SOM
Organic
Tube
Lining SOM
- 3377 Bonded –OH, -NH, -
NH2 1231 1258
Sulphate ester (S-O
stretch)
3284 3260 Bonded –OH, -NH, -
NH2 - 1181
Phenol, sulphonic
acid
2962 2955
-CH2, -CH3, bonded
OH in COOH
- 1114 SO42: S=O stretch
2921 2881 1096 - C-O, C-C stretch
2849 2838 - 1050 C-C bond in
alcohols, C-O stretch
- 2731 -CHO, bonded –OH
in COOH 1030 1029
C-C bond in
alcohols, C-O-C and
C-O stretch
- 2651 Bonded –OH in
COOH 1007 1003 C-C stretch
2350 - CO2 (background) - 984 C-O stretch, -
CH=CH2
- 1739 -CHO, -COOH - 963 -COOH
1594 1601 -COO
- asymm.
stretch 915 925 ring vibration
- 1489 Aromatic - 851
Sulphate ester (C-O-
S bending), α-
anomeric CH
deformation, C1-OH
1400 1407 -COO- - 832 Aromatic
1323 1332 SO2 asymm. stretch 797 - α-galactose skeletal
bending
- 1282 H-C-C, C-C 777 - β-galactose skeletal
bending
- 710 Aromatic
95
In addition to the FTIR measurements, the SOM extract was subjected to amino acid analysis.
The relative proportions of each type of residue are shown in 0. The majority of the amino acids
present (~61 %) are nonpolar and neutral, including significant amounts of glycine and proline.
Aspartic and glutamic acid, both polar and negatively charged, make up ~19 % of the
composition.
Table 6.4 Amino acid composition for the SOM.
Amino Acid Composition (mol%)
Asp 8.42
Glu 10.20
Ser 6.36
Gly 14.47
Arg 2.76
Thr 6.54
Ala 7.75
Pro 21.42
Tyr 3.06
Val 4.89
Met 0.87
Ile 3.69
Leu 5.34
Phe 2.21
Lys 1.94
96
6.5.8 Characterization of the Remineralized Organic Tube Lining
Prior to the remineralization experiment, the demineralised organic tube lining was subjected to
SEM and EDX analysis. The fibre and sheet features were found to be smooth, and C, O, Na,
and S were detected via EDX, as shown in Figure 6.15.
Figure 6.15 SEM images of the IOM, treated overnight with 0.5 M EDTA. A-B) Fibre structures, showing the smooth fibre surface. C) EDX spectrum of IOM (associated image not shown).
The absence of Ca indicates that the lining was fully decalcified following the overnight EDTA
treatment, which is consistent with the FTIR spectrum of the lining. SEM examination of the
97
remineralized tube lining revealed that the sheet-like structures were covered in crystals
exhibiting dumbbell morphology and a cauliflower-like texture, all in the same size range of ~5
μm in length and diameter. The cauliflower-like texture of the crystals is shown in detail in the
bottom left of Figure 6.16A, and a crystal in an earlier stage of growth before the formation of
the dumbbell shape is shown in the upper right of Figure 6.16A. Although the density of
coverage varied slightly from site to site, the morphology and size of the crystals was generally
constant along the surface of the lining. The EDX spectrum of the dumbbell-shaped crystals is
shown below the SEM image in Figure 6.16A. Strong C, O, and Ca signals indicate that they are
composed of CaCO3, but both Mg and Sr signals are present. The higher Sr signal indicates the
composition is predominantly aragonite. Crystals with a different morphology were observed
near fibre structures, as shown in Figure 6.16B. Ellipsoidal leaf-like features are evident. The
EDX spectrum of these crystals is shown in the bottom half of Figure 6.16B and the presence of
a strong Mg signal in addition to the C, O, and Ca signals suggest that the crystals are composed
of Mg-calcite. Details of the fibre structures are shown in Figure 6.16C, showing granular
features along the fibres. In addition to the strong S signal, there are small Ca and Mg signals,
which suggest that the fibres are slightly mineralized.
Figure 6.16 SEM images and EDX spectra of crystals formed on the EDTA-demineralized organic tube lining sample, after 24 hours. A) Crystals with cauliflower-like texture and dumb-bell morphology, with a high Sr content. B)
Crystal with ellipsoidal leaf-like features, with a high Mg content. C) Organic tube lining fibres, beginning to show signs of mineralization (Ca, Mg signals).
98
6.5.9 Characterization of the SOM Mineralization Precipitates
The precipitates collected after 90 min and 48 hr in the crystallization chamber were analyzed
via FTIR, SEM, and EDX. The FTIR spectrum for the 90 min control and SOM samples
identified the precipitates as being composed of amorphous calcium carbonate (ACC). The
control sample appeared to contain more precipitate than the SOM sample, from its higher
absorbance. The FTIR spectrum for the control sample is given in Figure 6.17. Broad peaks at
862 and 1070 cm-1
were observed, as well as the splitting of the ν3 carbonate ion mode into two
bands at 1410 and 1470 cm-1
. The band at 1654 cm-1
is assigned to water, indicating that the
ACC is partially hydrated. These features are all consistent with the spectrum for ACC.35, 36
For
the 48 hr precipitates, the FTIR spectrum for the SOM sample (Figure 6.18) indicated the
presence of aragonite. The control sample could not be identified via FTIR due to low surface
coverage of the Au substrate.
2000 1800 1600 1400 1200 1000 800 6000.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Absorb
ance / A
Wavenumber / cm-1
Figure 6.17 FTIR spectrum of the particles formed in the SOM in vitro crystallization experiment, for the control sample. The broad peaks at 862 and 1070 cm
-1 as well as the splitting of the ν3 carbonate ion mode into two bands
at 1410 and 1470 cm-1
correspond to amorphous calcium carbonate (ACC).
99
Figure 6.18 A) FTIR spectrum of the particles formed in the SOM in vitro crystallization experiment, for the 48 hour SOM sample. The peaks correspond to aragonite. B) Details of ν2 and ν4 carbonate ion modes.
The morphology of the ACC particles formed on the control sample differed from that of the
SOM sample, as shown in Figure 6.19AI-II and Figure 6.19BI-II. The control ACC particles
were spherical (0.5 – 1 μm in diameter) and smooth, and there was a high coverage density of the
particles on the substrate. Ca, C, and O signals in the EDX spectrum (Figure 6.19AIII) confirm
B
A
100
the CaCO3 composition of the particles. A high Mg signal and a low Sr signal were also
observed. In contrast, the SOM ACC particles were irregularly shaped and formed micron-sized
aggregates with sub-particles sizes of ~100 nm. C, O, Ca, and Mg were detected via EDX, as
shown in Figure 6.19BIII. The surface coverage of the substrate was sparser for the SOM ACC
particles than for the control ACC particles.
Figure 6.19 SEM images and EDX spectra of the crystal products of the in vitro SOM crystallization experiments. AI-
III) Control sample after 90 min. Spherical ACC particles with a high Mg content. BI-III) SOM sample after 90 min. Irregularly shaped ACC particles. CI-III) Control sample after 48 hr. Flower-like crystals with dumb-bell morphology,
and rhombic “petals”. Higher Mg signal than Sr. DI-III) Crystals with a bundle-of-needles morphology. Higher Sr signal than Mg.
101
The morphology of the 48 hr precipitates also differed between the control and SOM samples, as
shown in Figure 6.19CI-II and Figure 6.19DI-II. The control crystals were larger, about 10-15
μm in diameter, and possessed a flower-like morphology, with the “petals” exhibiting the
rhombic symmetry typical of calcite. From the EDX spectrum given in Figure 6.19CIII, C, O,
and Ca signals indicate a CaCO3 composition, and Mg and Sr were also detected, with the Mg
signal being higher than the Sr one. The surface coverage for the control particles was quite low,
which explains the difficulty with the FTIR spectrum. The SEM and EDX data suggest a Mg-
calcite composition. In comparison, the surface coverage for the SOM particles was quite high,
and the crystals had a needle-aggregate morphology. The particle size was consistently ~5 μm
on the long axis of the crystals. The EDX spectrum given in Figure 6.19CIII shows relatively
high Mg and Sr signals, in addition to the C, O, and Ca signals. The Sr signal is slightly higher
than that for Mg. Despite the presence of both Mg and Sr, the FTIR spectrum only exhibited
peaks for aragonite, and not calcite.
6.6 Discussion
6.6.1 Tube Layering and Mechanical Properties
The relatively low modulus of 3 ± 1 GPa obtained via nanoindentation on the adhesive material
surface (substrate side) could have two different origins. First, it could indicate the presence of a
thin organic film adsorbed onto the adhesive material, originating from the biofilm coating the
substrate upon which the worm settled, or from organic secretions from the worm. Second, the
low Young’s modulus could be due to hydration of the adhesive material, in particular, by the
outermost few nanometers. The modulus may be observed in the force profile shown in Figure
6.20 that shows two regions. The first region (i) of indentation from 0 nm to 7 nm exhibits a
good fit to the Sneddon model, giving a modulus of 6.3 ± 0.3 GPa (fit shown as the grey solid
line). The second region (ii) of indentation from 7 nm to 8 nm shows a sharp increase in force.
Using 6 nm as an estimate for the contact point, this second region was fit (shown as the red
solid line), yielding a Young’s modulus of ~36 Pa. This value is consistent with literature
values for biogenic calcite and aragonite in the wet state,37, 38
whereas the modulus obtained from
region (i) cannot be attributed to calcite or aragonite because it is too low. Therefore, the
Young’s modulus of 3 ± 1 Pa (which was obtained from analyses of region (i) of the force
profiles) arises from the presence of a thin organic film on the adhesive material surface.
102
Figure 6.20 Force profile obtained from nanoindentation on the adhesive material surface, performed in artificial seawater. The solid grey line denotes the fit for region (i) of indentation from 0 nm to 7 nm, yielding a Young’s
modulus of 6.3 +/- 0.3 GPa. The solid red line denotes the fit for region (ii) between 7 nm and 8 nm, calculated using a contact point set at 6 nm. The Young’s modulus for this region is ~36 GPa.
The Young’s modulus range of 40 – 65 GPa (FWHM) obtained via nanoindentation
measurements on the polished tube cross-section (in air) is consistent with literature values for
inorganic calcite and aragonite, for directions other than the c-axis of the crystals.39, 40
The range
is also consistent with biogenic calcite and aragonite measurements, such as those made on
gastropod nacre (aragonite) and calcitic fibre structures in brachiopods.38, 41
Both calcite and
aragonite are anisotropic materials, and therefore the measured Young’s modulus will depend on
the orientation of the crystals. Although the Young’s modulus of aragonite is typically higher
than that of calcite, due to the various unknown crystal orientations within the shell, regions of
calcite and aragonite could not be distinguished via the nanoindentation measurements. In order
to estimate the Young’s modulus, the bulk Poisson’s ratios for each polymorph were averaged.
Nanoindentation performed on the same sample in artificial seawater solution yielded lower
Young’s modulus values, with a range of 18 – 24 GPa (FWHM). This is consistent with studies
of nacre, in which the Young’s modulus is lower in the wet state compared with the dry state.38,
42 It is the organic phase of the nacre composite that becomes softer upon hydration; for the
tubeworm composite, the same effect is likely. For nacre, the wet Young’s modulus value is
typically ~10 Pa lower than the dry value. The tube shell exhibits a larger decrease in Young’s
modulus due to hydration compared with nacre (~20 GPa), which is likely due to its different
structure and composition.
103
Due to crystal anisotropy, the crystal organization in each ultrastructural layer plays a role in the
overall mechanical properties of the tube. This has been observed in studies of two calcitic
brachiopods, one with a semi-nacre structure, and the other with a calcitic fibre structure. The
semi-nacre structure was found to have a higher Young’s modulus as well as a greater hardness,
which correlates with the observation that the dominant orientation of crystals in the semi-nacre
structure consisted of the c-axis of the crystals being perpendicular to the surface.41
The spatial distribution of aragonite and Mg-calcite is moderately correlated with the three
ultrastructural layers, as shown in Figure 6.3C-F. The outer SPHP layer is composed of
aragonite, and the innermost layer in the LF region is composed of Mg-calcite. The middle
LF/IOP region of the tube wall is predominantly composed of Mg-calcite, although aragonite
was also occasionally detected. This polymorph distribution could offer some benefits for the
mechanical stability of the tube. Aragonite is denser, harder and less brittle than calcite,43
properties which make it beneficial for the outermost layer of the tube. The Mg content of the
calcite, at 20 mol%, is high in comparison to other biogenic forms of Mg-calcite. The elastic
modulus and hardness of calcite is known to increase with increasing Mg content.44-46
The
incorporation of such a large amount of Mg could have evolved as a means of developing a
stronger tube.
In general, laminar structures have been found to exhibit superior fracture resistance and
toughness. Propagating cracks will tend to arrest or deflect at the interface between layers due to
the differences in the structure or composition of each layer.47, 48
Serpulids that create multi-
layered tubes, such as H. dianthus, would therefore likely have an advantage over serpulids with
single-layered tubes, since the laminar organization of the shell increases its fracture resistance to
loads normal to the tube surface. Robust tubes would provide greater protection from predators,
such as sea urchins, clam worms, and fish.49, 50
It is interesting to consider the fact that the teeth
of such predators are also biomineralized structures, which have evolved to be exceptionally hard
and tough materials; in particular, Mg-calcite sea urchin teeth have hardnesses in the range of 3.5
– 4 GPa,51-53
which is on par with the hardness of aragonite. From an evolutionary perspective,
serpulid phylogeny is divided into two major clades, and multi-layered serpulid tubes with
complex oriented structures are found to occur in only one of the clades. It is thought that the
104
clade which is comprised of organisms with simpler structures and single-layered tubes is the
plesiomorphic condition for Serpulidae.12
6.6.2 CaCO3 Polymorphs and Morphologies
The presence of both aragonite and Mg-calcite, in nearly equal proportions, is unusual because
most biomineralized marine materials consist of one dominant polymorph, such as in the cases of
sea urchin spines (Mg-calcite), mollusc shells (aragonite), and coral (aragonite).10, 54, 55
Although
serpulid tubes are commonly bimineralic,12, 56, 57
in most species, either calcite or aragonite
comprises 75% or more of the tube material. To our knowledge, only one other serpulid species,
Pyrgopolon ctenactis, has a tube which consists of equal proportions of calcite and aragonite
(50.9 and 49.1 %, respectively).12
In his study of two serpulid worms, H. brachyacantha and E.
dianthus (= H. dianthus), Neff observed that tubes from each species contained large amounts of
both high Mg-calcite and aragonite, but did not specify their relative proportions.58
Similarly,
Vinn et al.12
observed that H. dianthus was bimineralic, but their XRD data did not include
results for H. dianthus.
The variety of crystal morphologies observed in the adhesive material for each polymorph is also
significant because it implies that there are a number of different macromolecular interactions or
biomineralization pathways. Most of the morphologies, with perhaps the exception of the
rhombohedral crystals shown in Figure 6.5C-D, are markedly different from the abiogenic
crystal forms for calcite and aragonite. For abiogenic crystals, under equilibrium conditions the
final morphology is closely related to the symmetry of crystal structure of the material, due to the
different surface energies of various crystal planes. High surface energy planes grow quickly
and are thermodynamically driven to minimize their surface area, while low surface energy
planes grow more slowly and therefore form the faces of the crystal with the largest surface
areas.3, 59
In biogenic systems, macromolecules such as aspartic acid-rich proteins and sulphated
glycosaminoglycans (GAGs, formerly known as mucopolysaccharides) can interact with certain
crystal planes, altering growth rates and therefore the final crystal morphology.3, 60
Ions such as
Mg2+
and Sr2+
can also influence the polymorph and morphology of the crystals by the same
thermodynamic principles.27
105
In the adhesive material, the acicular aragonite forms observed in H. dianthus are consistent with
the typical habit of aragonite formed by cyanobacteria and corals.61, 62
The dumbbell Mg-calcite
forms are similar to calcite precipitates obtained through in vitro mineralization by marine
bacteria and to Mg-calcite precipitates formed in the presence of the SOM extracted from the
giant seastar P. giganteus.63, 64
The crystal morphologies of the precipitates formed in the in vitro mineralization experiments in
the present study were different than those observed in the tube shell and adhesive material.
However, the precipitates produced in the control experiments (without the presence of organic
components from the tubeworm) did exhibit morphologies or features consistent with non-
biogenic forms of ACC and Mg-calcite, while those formed in the SOM and IOM experiments
possessed biogenic characteristics. Furthermore, the morphology and polymorph of the
precipitates formed in the SOM in vitro mineralization experiment after 48 hrs were different
between the control sample and the SOM sample. Lastly, the different relative abundances of
precipitate in the control versus the SOM sample for each time interval indicate that there are
different rates of formation for each calcium carbonate polymorph, due to the absence or
presence of the SOM. Therefore, the in vitro mineralization experiments show that both the
SOM and IOM components mediate the calcification process.
6.6.3 Insights into the Formation and Attachment of the Adhesive Material to the Substrate
In considering the biomineralization mechanisms contributing to the formation of the adhesive
material, one cannot rule out the possible role of the biofilm present on the substrate. Biofilms
tend to form on hard substrates exposed to aqueous solution, such as seawater, and are composed
of organic molecules and microorganisms such as bacteria, and diatoms. It is known that
biofilms influence marine invertebrate larvae settlement, and in the case of H. elegans, a serpulid
species closely related to H. dianthus, the larvae settle preferentially on substrates covered in a
biofilm.65-67
Some components of the biofilm, such as the extracellular polysaccharides
produced by bacteria, are capable of inducing CaCO3 mineralization.68, 69
It is possible that the
adhesive material forms in an environment containing macromolecules produced by the
inhabitants of the biofilm as well as the tubeworm, which could explain the variety of crystal
morphologies observed within the adhesive material.
106
The calcareous tube of H. elegans exhibited increased adhesion strength for individuals settled
on biofilms versus clean substrates.70
Nanoindentation measurements in this study on H.
dianthus revealed a thin soft layer on the surface of the adhesive material, suggesting the
presence of a biofilm or organic secretion. In addition, SEM images show bacteria and diatoms
incorporated within the adhesive material (Figure 6.6), and fibre and sheet structures, likely
organic, are observed to be associated with crystals (Figure 6.5C, Figure 6.5E-F). These
observations of the intimate contact and merging of the adhesive material with the underlying
biofilm could explain the increased adhesion strength of tubes attached to biofilm-coated
substrates.
6.6.4 SOM Composition
No amide bands are immediately visible in the FTIR spectrum, although it is possible that they
are obscured due to band overlap with carboxylic acid salt and sulphate ester resonances. Due to
these dominant bands and the carbohydrate signatures, carboxylated and sulphated
polysaccharides constitute the majority of the SOM, and proteins form a minority component.
The most abundant amino acids in the SOM of H. dianthus were found to be Asp (8.4 mol%),
Glu (10.2 mol%), Gly (14.5 mol%), and Pro (21.4 mol%). This is largely consistent with the
SOM amino acid composition averaged from data for three serpulid species,71
in which Asp,
Glu, and Gly are major components, although the Asp content is higher (~20 mol%) and the Pro
content is much lower (~5 mol%) than observed for H. dianthus. Pro plays a role in the
stabilization of collagen conformation,72
and therefore the high Pro content could be due to
collagen from the tube lining, some of which could have been solubilised.
The presence of Asp, Glu, and Gly is also consistent with studies of various calcareous
biomineralizing organisms such as molluscs and coral. Analyses of these organisms typically
reveal that acidic amino acids (aspartic acid and glutamic acid) and neutral amino acids (such as
serine and glycine) are major components of the SOM, with the acidic amino acids making up 30
– 50% of the total.60, 73-75
In those studies, Asp in particular was prominent, making up 20% or
more of the total amino acid composition. Sequences of aspartic acid separated by neutral amino
acids, as well as poly-Asp molecules have been shown to be able to bind with Ca2+
; in addition,
poly-Asp takes on a beta-sheet structure upon binding with Ca2+
. This structure exposes the
107
negatively charged amino acids in such a way that it can interact with specific crystal faces,
matching the distances between Ca2+
ions in calcite or aragonite lattices.76
In H. dianthus, the
acidic amino acid content is low in comparison with the aforementioned biomineralizing
organisms, including other serpulids. However, the Asp and Glu content in H. dianthus is
similar to that observed in the SOM of P. giganteus ossicles (Echinodermata), as reported by
Gayathri et al.64
In their work, the seastar’s SOM was found to accelerate the conversion of
amorphous calcium carbonate into crystalline Mg-calcite, despite the relatively low acidic amino
acid content (~18 mol%). Therefore, although the general observation among CaCO3
biomineralizing species is that high amounts of negatively charged amino acids correlates with
control over crystal growth and morphology, there are other factors to consider. As mentioned
above, macromolecules such as acidic or sulphated polysaccharides can also play a role in Ca2+
binding and mineralization, and indeed the SOM contains significant amounts of carboxylated
and sulphated polysaccharides as determined via FTIR.
6.6.5 IOM Composition
The relatively high S signals observed in the EDX spectra for the organic tube lining and other
organic membranes within the tube of H. dianthus could indicate the presence of sulphated
GAGs, which are negatively charged linear polysaccharide chains covalently bound to a core
protein, with the resulting complex known as a proteoglycan.77
It is unlikely that such a high
signal would arise from proteins alone (disulphide bonds), unless there were a substantial amount
of cysteine residues; furthermore, previous studies have noted the presence of high amounts of
sulphated polysaccharides associated with annelid collagen.15, 78-80
The polysaccharide and
sulphate ester signatures in the FTIR spectrum of the organic tube lining (Figure 6.14) appear to
provide further evidence for the presence of sulphated GAGs. However, no protein signatures
are immediately apparent in the FTIR spectrum, despite the fact that collagen was also detected
within the lining. It is possible that the amide bands are obscured by the absorptions
corresponding to carboxylated polysaccharides, particularly by the carboxylic acid salt bands at
1594 and 1400 cm-1
. Upon close inspection of the 1594 cm-1
band, which is not symmetric, there
appear to be higher and lower frequency components contributing to the shape of the broad band,
with the higher frequency component possibly corresponding to the amide I band. Regardless,
the intensity of the polysaccharide bands and the lack of obvious amide bands suggest that
108
proteins are a relatively minor component of the lining, and that the polysaccharides are mainly
composed of sugars which lack amino groups and hence are not GAGs. Overall, the spectrum
bears a strong resemblance to that of cationic salts (Na+, Ca
2+) of alginic acid, an acidic
polysaccharide found in seaweed,31, 81
with additional peaks indicating the presence of sulphated
polysaccharides, such as sulphated fucans.82
Cations can cross-link individual polysaccharide chains, resulting in the formation of gels.
Acidic polysaccharides and alginates are known to self-assemble into a variety of structures,
including lamellar, sponge-like, and sheet structures,83
in part through this cation cross-linking
process. The organic sheet structures observed via SEM could therefore be composed of
polysaccharides. It is also possible that the fibre structures are partially composed of
polysaccharides, in addition to the collagen fibrils.
Based on the above, in conjunction with the SEM data, the organic tube lining is comprised of
layers of collagen-containing fibres and organic sheets, which are composed of sulphated and
carboxylated polysaccharides. This is consistent with a description of a “coating membrane”
given by Muzii,84
for Eupomatus (=Hydroides) dianthus, which tentatively identifies the
membrane as being composed of protein fibres and dispersed sulphomucins. In the present
study, however, although the presence of sulphated polysaccharides is confirmed, the presence of
GAGs is not. The organic membranes within the tube shell may be similar, based on the S EDX
signal, although the presence of collagen or fibre features has not been verified in these
membranes.
Both polysaccharides and collagen have been found to influence biomineralization. Sulphated
and carboxylated polysaccharides have the capability to bind cations, thereby playing roles in the
nucleation process.23, 76, 85
Studies by Cuif et al.86
have identified varying distributions of
sulphated proteoglycans spatially correlated with varying CaCO3 polymorph distribution.
Collagen gels as well as solubilised collagen have been used for in vitro mineralization
experiments, and found to affect the polymorph and morphology of the resulting CaCO3
crystals.87-89
In the case of the collagen gels, the gel was found to have two effects on the crystal
morphology. Firstly, the gel creates constricted volumes within its fibre network, which
provides supersaturated microenvironments for nucleation and growth, as well as physical
109
boundaries for constraining crystal size or orientation. Secondly, deformation of the gel can also
change the alignment of the collagen fibres, which could then interact with the crystal faces
themselves to alter the morphology. Solubilised collagen influences calcite morphology by
adsorbing to different crystal planes, resulting in a variety of crystal habits depending on the
collagen concentration.
6.6.6 Summary of the Structure and Composition of the Tube Shell and Adhesive Material
The preceding observations are summarized in Figure 6.21, which shows cartoons of the
transverse cross-sections of both the tube shell and the adhesive material. In Figure 6.21A, the
three distinct ultrastructural layers, SPHP, IOP, and LF, are shown to be composed
predominantly of aragonite, Mg-calcite, and Mg-calcite, respectively. Organic sheets are shown
within the mineralized layers of the shell, and the organic tube lining is shown on the inner wall
of the tube shell. In Figure 6.21B, although no distinct ultrastructural layers could be easily
distinguished, the observed spatial distribution of aragonite and Mg-calcite is shown, with
aragonite closer to the substrate side and Mg-calcite closer to the organic tube lining and the
lumen side. The abundance of organic sheets is also depicted within the mineral layer. Both the
organic sheets and the organic tube lining contain sulphated polysaccharide components.
6.6.7 Role of the Organic Tube Lining in Tube Formation
The structure and composition of the organic tube lining of H. dianthus is reminiscent of
descriptions of polychaete cuticles, with the exception of the observation of banding in
individual collagen fibrils. In many polychaete worms, including serpulids, the cuticle consists
of unstriated collagen fibrils in an amorphous or fine filamentous matrix, with the fibrils
frequently present in a criss-cross layered structure, located in the basal layer of the cuticle. The
collagenous cuticle is produced by secretory cells in the epidermis, and is adherent to the
epidermis in most polychaetes. Striated collagen has been observed in polychaetes, but only for
non-cuticular interstitial forms, located under the epidermal cells in the extracellular
matrix.80,90,91
Because the organic tube lining was found to be present and intimately associated
with the mineralized tube wall after worm removal, it is not cuticular material. However, the
110
similarities in composition and structure suggest that the organic tube lining is also produced by
secretory cells in the epidermis.
Figure 6.21 Summary of tube structure and composition. A) Cartoon of tube shell, transverse cross-section. LF = lamello-fibrillar structure, IOP = irregularly oriented prismatic structure, SPHP = spherulitic prismatic structure. B)
Cartoon of adhesive material, transverse cross-section.
Most observations of the tubes of various serpulid species have found the inner tube walls to be
smooth, with no mention of criss-crossed fibre structures.12, 13
However, organic sheets lining
the inner tube walls were observed in the serpulids M. cavatica and S. giganteus, with a mesh-
like structure noted for the latter species.12
Furthermore, the inner tube walls of one polychaete
species, Chaetopterus variopedatus, were found to consist of aligned protein fibres with a criss-
crossed layered structure, in an amorphous matrix. The worm itself was found to be lacking a
cuticle entirely, and it was hypothesized that C. variopedatus sloughs off its cuticle and uses it as
a template for tube construction.92
It has not been verified whether the fibre layers are cuticular,
although they appeared to be formed by the epidermis. Although the tube of C. variopedatus is
111
not calcareous, being composed of proteins and GAGs, C. variopedatus demonstrates an
example of a tube formation strategy utilizing a framework of aligned protein fibres.
H. dianthus may employ a similar strategy, using its organic tube lining as a scaffold for the
biomineralization of CaCO3. A further similarity between H. dianthus and C. variopedatus is the
observation that the protein fibrils in the tube lining exhibit striation, which is unusual since the
vast majority of annelid cuticular collagen has been found to be smooth. A banding of 64.1 nm
was observed for C. variopedatus via SEM, which is close to the 67 nm banding we observed in
H. dianthus via AFM. These values are consistent with the values for vertebrate collagen type I,
which has a spacing of 64 – 67 nm.30
The association of the tube lining’s collagen-containing fibres with CaCO3 crystals at the tube
wall interface (Figure 6.10), and the detection of elements consistent with Mg-calcite in the
fibres and on the smooth sheet at the lumen interface further support the idea that H. dianthus
employs the organic tube lining as a scaffold for biomineralization. A previous study showed
that the decalcified IOM of H. dianthus was capable of becoming recalcified in an inorganic
solution mimicking the ion concentrations found in molluscan extrapallial fluid.93
The IOM was
described as a tube-shaped structure, and presumably consisted of both the tube lining and the
organic sheets within the tube walls. Whether mineralization is induced by collagen
components, the polysaccharide components or a combination of the two remains to be
determined conclusively; however, the results of the in vitro recalcification of the organic tube
lining in artificial seawater (Figure 6.19) confirm that the lining as a whole can induce CaCO3
mineralization. The majority of the Sr-rich crystals observed on the surface of the IOM smooth
sheet structures had the same morphology and were similar in size and chemical composition.
This indicates that the nucleation and growth conditions for these CaCO3 crystals were very
similar, which suggests that the macromolecules associated with or comprising the sheet
structures play a role in the biomineralization process. The IOM fibre structures also appear to
influence the form of the CaCO3 precipitates, based on observations of Mg-rich crystals near the
fibres with a different morphology than the Sr-rich crystals. The fibres themselves exhibited
granular features which were not observed on EDTA-treated IOM fibres, and the detection of
Mg and Ca signals implies that the fibres were becoming mineralized, possibly via a Mg2+
selective mechanism.
112
As a last point to consider, from the serpulid studies of Hedley and Vovelle,15-17
it was observed
that the mucus-secreting cells of the ventral and lateral epithelium contained a high amount of
calcium, with only the cells of the calcium-secreting glands containing more. As mentioned
earlier, it is possible that the organic tube lining is also produced by epidermal secretory cells.
Therefore, the worm could create a high local concentration of Ca2+
ions adjacent to its body
through epidermal mucus secretions, which could then bind to the highly negatively charged
sulphated and carboxylated polysaccharides in the organic tube lining, initializing the
biomineralization process on the organic tube lining.
6.7 Conclusions
The calcareous tube shell and adhesive material of the serpulid H. dianthus were studied from a
biomineralization perspective. The tube shell consists of three layers: an outer layer composed
of aragonite crystals with a spherulitic prismatic (SPHP) structure, a middle irregularly oriented
prismatic (IOP) Mg-calcite layer, and an inner lamello-fibrillar (LF) Mg-calcite layer. The
Young’s moduli of the inorganic components, as measured via AFM nanoindentation on
polished cross-sectioned samples, ranged from ~22 GPa in the wet state, to ~51 GPa in the dry
state. Both soluble and insoluble organic matrices were extracted from the tube shells.
H. dianthus appears to utilize its organic tube lining, which is composed of striated and smooth
collagen and carboxylated and sulphated polysaccharides, as an extracellular scaffold for the
biomineralization of Mg-calcite. This is consistent with a speculation by Neff,58
who
hypothesized that the calcium-secreting glands of H. dianthus formed the aragonitic outer wall
(SPHP layer) of the tube, while the ventral epithelium played a role in the formation of the Mg-
calcitic inner layers (IOP and LF layers). In vitro mineralization experiments in artificial
seawater confirmed the ability of the organic tube lining and the SOM to mediate calcification,
with some degree of control over the final morphology and polymorph of the resulting crystals.
Further characterization of the soluble and insoluble organic matrices is required in order to
better understand their roles in the biomineralization process. Purification of the matrices in
order to separate the various proteins and polysaccharides would allow for the determination of
the specific roles of each component in the biomineralization process.
113
6.8 Contributions
The author performed the SEM, EDX, and FTIR measurements and analysis and wrote the
manuscript. A.E.T. also undertook an extensive literature survey in order to present and interpret
all of the experimental results in the context of calcium carbonate biomineralization.
6.9 References
1. Weiner, S.; Addadi, L., Design strategies in mineralized biological materials. J. Mater.
Chem. 1997, 7, 689–702.
2. Veis, A., Crystals and Life: An Introduction. In Biomineralization, John Wiley & Sons,
Ltd: 2010; pp 1–35.
3. Meldrum, F. C.; C lfen, H., Controlling mineral morphologies and structures in
biological and synthetic systems. Chem Rev 2008, 108, 4332–4432.
4. Mann, S., Biomineralization: Principles and Concepts in Bioinorganic Materials
Chemistry. Oxford University Press: Oxford, 2001.
5. Checa, A. .; Rodr guez-Navarro, A. B., Self-organisation of nacre in the shells of
Pterioida (Bivalvia: Mollusca). Biomaterials 2005, 26, 1071–1079.
6. Addadi, L.; Joester, D.; Nudelman, F.; Weiner, S., Mollusk shell formation: A source of
new concepts for understanding biomineralization processes. Chem. Eur. J. 2006, 12,
980–987.
7. Rousseau, M.; Meibom, A.; Gèze, M.; Bourrat, X.; Angellier, M.; Lopez, E., Dynamics
of sheet nacre formation in bivalves. J. Struct. Biol. 2009, 165, 190–195.
8. Lee, S. W.; Kim, Y. M.; Kim, R. H.; Choi, C. S., Nano-structured biogenic calcite: A
thermal and chemical approach to folia in oyster shell. Micron 2008, 39, 380–386.
9. Mount, A. S.; Wheeler, A. P.; Paradkar, R. P.; Snider, D., Hemocyte-mediated shell
mineralization in the eastern oyster. Science 2004, 304, 297–300.
10. Allemand, D.; Ferrier-Pagès, C.; Furla, P.; Houlbrèque, F.; Puverel, S.; Reynaud, S.;
Tambutté, É.; Tambutté, S.; Zoccola, D., Biomineralisation in reef-building corals: from
molecular mechanisms to environmental control. Comptes Rendus Palevol. 2004, 3, 453–
467.
11. Cuif, J.-P.; Dauphin, Y., The two-step mode of growth in the scleractinian coral skeletons
from the micrometre to the overall scale. J. Struct. Biol. 2005, 150, 319–331.
12. Vinn, O.; Ten Hove, H. A.; Mutvei, H.; KirsimÄE, K., Ultrastructure and mineral
composition of serpulid tubes (Polychaeta, Annelida). J. Linn. Soc. London, Zool 2008,
154, 633–650.
13. Nishi, E., On the internal structure of calcified tube walls in Serpulidae and Spirorbidae
(Annelida, Polychaeta). Mar. Fouling 1993, 10, 17–20.
114
14. Swan, E. F., The calcareous tube secreting glands of the serpulid polychaetes. J.
Morphol. 1950, 86, 285–314.
15. Hedley, R. H., Studies of serpulid tube formation: I. The secretion of the calcareous and
organic components of the tube by Pomatoceros Triqueter. Q. J. Microsc. Sci. 1956, 97,
411–419.
16. Hedley, R. H., Studies of serpulid tube formation. Q. J. Microsc. Sci. 1956, 97, 421–427.
17. Vovelle, J., Processus glandulaires impliques dans la reconstitution du tube chez
Pomatoceros triqueter. Bull. Lab. Mar. Dinard. 1956, 42, 10–32.
18. Neff, J. M., Ultrastructural studies of the secretion of calcium carbonate by the serpulid
polychaete worm, Pomatoceros caeruleus. Z. Zellforsch. 1971, 120, 160–186.
19. Vinn, O.; KirsimÄE, K.; Ten Hove, H. A., Tube ultrastructure of Pomatoceros
americanus (Polychaeta, Serpulidae): Implications for the tube formation of serpulids.
Est. J. Earth Sci. 2009, 58, 148–152.
20. Hutter, J. L.; Bechhoefer, J., Calibration of atomic-force microscope tips. Rev. Sci.
Instrum. 1993, 64, 1868–1873.
21. Sun, Y.; Guo, S.; Walker, G. C.; Kavanagh, C. J.; Swain, G. W., Surface elastic modulus
of barnacle adhesive and release characteristics from silicone surfaces. Biofouling 2004,
20, 279–289.
22. Sullan, R. M. A.; Gunari, N.; Tanur, A. E.; Chan, Y.; Dickinson, G. H.; Orihuela, B.;
Rittschof, D.; Walker, G. C., Nanoscale structures and mechanics of barnacle cement.
Biofouling 2009, 25, 263–275.
23. Addadi, L.; Moradian, J.; Shay, E.; Maroudas, N. G.; Weiner, S., A chemical model for
the cooperation of sulfates and carboxylates in calcite crystal nucleation: Relevance to
biomineralization. Proc. Natl. Acad. Sci. 1987, 84, 2732–2736.
24. White, W. B., Infrared Spectra of Minerals. Mineralogical Society: London, 1974.
25. Böttcher, M. E.; Gehlken, P.-L.; Steele, D. F., Characterization of inorganic and biogenic
magnesian calcites by Fourier Transform infrared spectroscopy. Solid State Ionics 1997,
101–103, Part 2, 1379–1385.
26. Vinn, O.; Ten Hove, H. A.; Mutvei, H., On the tube ultrastructure and origin of
calcification in sabellids (Annelida, Polychaeta). Palaeontology 2008, 51, 295–301.
27. Meldrum, F. C., Calcium carbonate in biomineralisation and biomimetic chemistry. Int.
Mater. Rev. 2003, 48, 187–224.
28. Mavko, G.; Mukerji, T.; Dvorkin, J., The Rock Physics Handbook: Tools for Seismic
Analysis in Porous Media. Cambridge University Press: Cambridge, 1998.
29. Barthelat, F.; Li, C.-M.; Comi, C.; Espinosa, H. D., Mechanical properties of nacre
constituents and their impact on mechanical performance. J.Mater. Res. 2006, 21, 1977–
1986.
30. Bonucci, E., Role of collagen fibrils in calcification. CRC Press: Boca Raton, 1992.
115
31. Leal, D.; Matsuhiro, B.; Rossi, M.; Caruso, F., FT-IR spectra of alginic acid block
fractions in three species of brown seaweeds. Carbohydr. Res. 2008, 343, 308–316.
32. Charles, A. L.; Huang, T. C.; Chang, Y. H., Structural analysis and characterization of a
mucopolysaccharide isolated from roots of cassava (Manihot esculenta Crantz L.). Food
Hydrocolloids 2008, 22, 184–191.
33. Mathlouthi, M.; Koenig, J. L., Vibrational Spectra of Carbohydrates. In Advances in
Carbohydrate Chemistry and Biochemistry, Tipson, R. S.; Derek, H., Eds. Academic
Press: 1987; Vol. 44, pp 7–89.
34. Matsuhiro, B., Vibrational spectroscopy of seaweed galactans. Hydrobiologia 1996, 326-
327, 481–489.
35. Addadi, L.; Raz, S.; Weiner, S., Taking advantage of disorder: Amorphous calcium
carbonate and its roles in biomineralization. Adv. Mater. 2003, 15, 959–970.
36. Loste, E.; Wilson, R. M.; Seshadri, R.; Meldrum, F. C., The role of magnesium in
stabilising amorphous calcium carbonate and controlling calcite morphologies. J. Cryst.
Growth 2003, 254, 206–218.
37. Burkhardt, A.; Hansmann, W.; Märkel, K.; Niemann, H.-J., Mechanical design in spines
of diadematoid echinoids (Echinodermata, Echinoidea). Zoomorphology 1983, 102, 189–
203.
38. Jackson, A. P.; Vincent, J. F. V.; Turner, R. M., Comparison of nacre with other ceramic
composites. J. Mater. Sci. 1990, 25, 3173-3178.
39. Chen, C.-C.; Lin, C.-C.; Liu, L.-G.; Sinogeikin, S. V.; Bass, J. D., Elasticity of single-
crystal calcite and rhodochrosite by Brillouin spectroscopy. Am. Mineral. 2001, 86,
1525–1529.
40. Liu, L.-G.; Chen, C.-C.; Lin, C.-C.; Yang, Y.-J., Elasticity of single-crystal aragonite by
Brillouin spectroscopy. Phys. Chem.Miner. 2005, 32, 97–102.
41. Pérez-Huerta, A.; Cusack, M.; Zhu, W., Assessment of crystallographic influence on
material properties of calcite brachiopods. Mineral. Mag. 2008, 72, 563–568.
42. Bruet, B. J. F.; Qi, H. J.; Boyce, M. C.; Panas, R.; Tai, K.; Frick, L.; Ortiz, C., Nanoscale
morphology and indentation of individual nacre tablets from the gastropod mollusc
Trochus niloticus. J. Mater. Res. 2005, 20, 2400–2419.
43. Roberts, W. L.; Rapp, G. R. J.; Weber, J., Encyclopedia of Minerals. Van Nostrand
Reinhold: New York, 1974.
44. Hozman, P., Das unterschiedliche Harteverhalten biogener und anorganischer
Calcitkristalle. Bochum: 1983; Vol. 10.
45. Zhang, J. Z.; Reeder, R. J., Comparative compressibilities of calcite-structure carbonates:
Deviations from empirical relations. Am. Mineral. 1999, 84, 861–870.
46. Xu, L.; Renner, J.; Herwegh, M.; Evans, B., The effect of dissolved magnesium on creep
of calcite II: transition from diffusion creep to dislocation creep. Contrib.Mineral. Petrol.
2009, 157, 339–358.
116
47. He, M. Y.; Hutchinson, J. W., Crack deflection at an interface between dissimilar elastic
materials. Int. J. Solids Struct. 1989, 25, 1053–1067.
48. Parmigiani, J. P.; Thouless, M. D., The roles of toughness and cohesive strength on crack
deflection at interfaces. J. Mech. Phys. Solids 2006, 54, 266–287.
49. Schochet, J., Opercular regulation in the polychaete Hydroides dianthus (verrill, 1873) II
Control of opercular regulation. J. Exp. Zool. 1973, 184, 259–279.
50. Haines, J. L.; Maurer, D., Quantitative faunal associates of the serpulid polychaete
hydroides dianthus. Mar.Biol. 1980, 56, 43–47.
51. Markel, K.; Gorny, P., On the functional anatomy of sea urchin teeth (Echinodermata,
Echinoidea). Z. Morphol. Tiere. 1973, 75, 103–114.
52. Wang, R. Z.; Addadi, L.; Weiner, S., Design strategies of sea urchin teeth: structure,
composition and micromechanical relations to function. Philos. Trans. R. Soc. London,
Ser. B 1997, 352, 469–480.
53. Ma, Y.; Cohen, S. R.; Addadi, L.; Weiner, S., Sea urchin tooth design: An “all-calcite”
polycrystalline reinforced fiber composite for grinding rocks. Adv. Mater. 2008, 20,
1555–1559.
54. Magdans, U.; Gies, H., Single crystal structure analysis of sea urchin spine calcites. Eur.
J. Mineral. 2004, 16, 261–268.
55. Compere, E. L. J., Determination of calcite:aragonite ratios in mollusc shells by infrared
spectra. Limnol. Oceanogr. 1973, 18, 326–331.
56. Lowenstam, H. A., Environmental relations of modification compositions of certain
carbonate secreting marine invertebrates. Proc. Nat. Acad. Sci. 1954, 40, 39–48.
57. Bornhold, B. D.; Milliman, J. D., Generic and environmental control of carbonate
mineralogy in serpulid (polychaete) tubes. J. Geol. 1973, 81, 363–373.
58. Neff, J. M., Mineral regeneration by serpulid polychaete worms. Biol. Bull. 1969, 136,
76–90.
59. Wulff, G., On the question of the rate of growth and dissolution of crystal surfaces. Z.
Kristallogr. Mineral. 1901, 34, 449–530.
60. Weiner, S., Aspartic acid-rich proteins: Major components of the soluble organic matrix
of mollusk shells. Calcified Tissue Int. 1979, 29, 163–167.
61. Macintyre, I. G.; Reid, R. P., Comment on the origin of aragonite needle mud; a picture is
worth a thousand words. J. Sediment. Res. 1992, 62, 1095–1097.
62. Constantz, B. R., Coral skeleton construction: A physiochemically dominated process.
Palaios 1986, 1, 152–157.
63. Buczynski, C.; Chafetz, H. S., Habit of bacterially induced precipitates of calcium
carbonate and the influence of medium viscosity on mineralogy. J. Sediment. Res. 1991,
61, 226–233.
Equation 7
117
64. Gayathri, S.; Lakshminarayanan, R.; Weaver, J. C.; Morse, D. E.; Kini, R. M.;
Valiyaveettil, S., In vitro study of magnesium-calcite biomineralization in the skeletal
materials of the seastar Pisaster giganteus. Chem. Eur. J. 2007, 13, 3262–3268.
65. Hadfield, M. G.; Unabia, C.; Smith, C. M.; Michael, T. M., Settlement preferences of the
ubiquitous fouler Hydroides elegans. In Recent Developments in Biofouling Control,
Fingerman, M.; Nagabhushanam, R.; Sarojini, R., Eds. Oxford & IBH: New Delhi, 1994;
pp 65–74.
66. Unabia, C. R. C.; Hadfield, M. G., Role of bacteria in larval settlement and
metamorphosis of the polychaete Hydroides elegans. Mar. Biol. 1999, 133, 55–64.
67. Huang, S.; Hadfield, M. G., Composition and density of bacterial biofilms determine
larval settlement of the polychaete Hydroides elegans. Mar. Ecol. Prog. Ser. 2003, 260,
161–172.
68. Wada, N.; Okazaki, M.; Tachikawa, S., Effects of calcium-binding polysaccharides from
calcareous algae on calcium carbonate polymorphs under conditions of double diffusion.
J. Cryst. Growth 1993, 132, 115–121.
69. Kawaguchi, T.; Decho, A. W., A laboratory investigation of cyanobacterial extracellular
polymeric secretions (EPS) in influencing CaCO3 polymorphism. J. Cryst. Growth 2002,
240, 230–235.
70. Zardus, J. D.; Nedved, B. T.; Huang, Y.; Tran, C.; Hadfield, M. G., Microbial biofilms
facilitate adhesion in biofouling invertebrates. Biol. Bull. 2008, 214, 91–98.
71. Gaill, F.; Hunt, S., IV. Tubes. In The Ultrastructure of Polychaeta, Westheide, W.;
Hermans, C. O., Eds. Gustav Fischer Verlag: Stuttgard, New York, 1988; pp 61–70.
72. Wainwright, S. A., Collagen. In Mechanical Design in Organisms, Princeton University
Press: Princeton, 1982; pp 81–93.
73. Wheeler, A. P., Phosphoproteins of oyster (Crassostrea virginica) shell organic matrix. In
Hard Tissue Mineralization and Demineralization, Suga, S.; Watabe, N., Eds. Springer-
Verlag: Tokyo, 1992; pp 171–187.
74. Sarashina, I.; Endo, K., Primary structure of a soluble matrix protein of scallop shell;
implications for calcium carbonate biomineralization. Am. Mineral. 1998, 83, 1510–
1515.
75. Puverel, S.; Tambutté, E.; Pereira-Mouriès, L.; Zoccola, D.; Allemand, D.; Tambutté, S.,
Soluble organic matrix of two Scleractinian corals: Partial and comparative analysis.
Comp. Biochem. Physiol. B: Biochem. Mol. Biol. 2005, 141, 480–487.
76. Weiner, S.; Hood, L., Soluble protein of the organic matrix of mollusk shells: a potential
template for shell formation. Science 1975, 190, 987–989.
77. Arias, J. L.; Fern ndez, M. a. S., Polysaccharides and proteoglycans in calcium
carbonate-based biomineralization. Chem. Rev. 2008, 108, 4475–4482.
78. Rudall, K. M., Comparative biology and biochemistry of collagen. In Treatise on
Collagen 2: Biology of Collagen Part A, Academic Press: New York, 1968; pp 83–137.
118
79. Cássaro, C. M.; Dietrich, C. P., Distribution of sulfated mucopolysaccharides in
invertebrates. J. Biol. Chem. 1977, 252, 2254–61.
80. Hausen, H., Comparative structure of the epidermis in polychaetes (Annelida).
Hydrobiologia 2005, 535-536, 25–35.
81. Lawrie, G.; Keen, I.; Drew, B.; Chandler-Temple, A.; Rintoul, L.; Fredericks, P.;
Grøndahl, L., Interactions between alginate and chitosan biopolymers characterized using
FTIR and XPS. Biomacromolecules 2007, 8, 2533–2541.
82. Silva, T. M. A.; Alves, L. G.; Queiroz, K. C. S. d.; Santos, M. G. L.; Marques, C. T.;
Chavante, S. F.; Rocha, H. A. O.; Leite, E. L., Partial characterization and anticoagulant
activity of a heterofucan from the brown seaweed Padina gymnospora. Braz. J. Med.
Biol. Res. 2005, 38, 523–533.
83. Andrade, L. R.; Salgado, L. T.; Farina, M.; Pereira, M. S.; Mourão, P. A. S.; Amado
Filho, G. M., Ultrastructure of acidic polysaccharides from the cell walls of brown algae.
J. Struct. Biol. 2004, 145, 216–225.
84. Muzii, E. O., A coating membrane in the calcareous tubes of serpulids. Publ. Staz. Zool.
Napoli 1968, 36, 135–137.
85. Addadi, L.; Weiner, S., Control and design principles in biological mineralization.
Angew. Chem. Int. Ed. 1992, 31, 153–169.
86. Cuif, J. P.; Dauphin, Y.; Farre, B.; Nehrke, G.; Nouet, J.; Salomé, M., Distribution of
sulphated polysaccharides within calcareous biominerals suggests a widely shared two-
step crystallization process for the microstructural growth units. Mineral. Mag. 2008, 72,
233–237.
87. Falini, G., Crystallization of calcium carbonates in biologically inspired collagenous
matrices. Int. J. Inorg. Mater. 2000, 2, 455–461.
88. Shen, F. H.; Feng, Q. L.; Wang, C. M., The modulation of collagen on crystal
morphology of calcium carbonate. J. Cryst. Growth 2002, 242, 239–244.
89. Jiao, Y.; Feng, Q.; Li, X., The co-effect of collagen and magnesium ions on calcium
carbonate biomineralization. Mater. Sci. Eng., C 2006, 26, 648–652.
90. Storch, V.; Welsch, U., Uber die feinstruktur der polychaeten-epidermis (Annelida).
Morph. Tiere 1970, 66, 310–322.
91. Lepescheux, L., Spatial organization of collagen in annelid cuticle: order and defects.
Biol.Cell 1988, 62, 17–31.
92. Brown, S. C.; McGee-Russell, S., Chaetopterus tubes: Ultrastructural architecture. Tissue
and Cell 1971, 3, 65–70.
93. Bernhardt, A. M.; Manyak, D. M.; Wilbur, K. M., In vitro recalcification of organic
matrix of scallop shell and serpulid tubes. J.Mollusc. Stud. 1985, 51, 284–289.
Figure 7. sfdtg
Table 7. swrdtg
119
7 Nanoscale Structures and Properties of the Proteinaceous Cement of the Barnacle Amphibalanus amphitrite
7.1 Permissions
The material in this chapter is presented with permission from Sullan, R. M. A.; Gunari, N.;
Tanur, A. E.; Chan, Y.; Dickinson, G. H.; Orihuela, B.; Rittschof, D.; Walker, G.C. Nanoscale
structures and mechanics of barnacle cement. Biofouling 2009, 25, 263-275.
7.2 Abstract
In this work, we investigate the proteins comprising barnacle cement, which is an adhesive
nanomaterial for firmly attaching the organism to a substrate. The presence of β-sheet
components within the cement was established via FTIR spectroscopy. Rod shaped structures
were identified by AFM and SEM imaging, and their organic nature was determined with EDX.
Chemical staining revealed that the rods contained beta-sheet structure components. AFM
indentation on the rods indicated a stiffer nature of these structures compared with other
structures within the cement.
7.3 Introduction
Proteins are nature’s designer nanomaterial, formed by directed self-assembly of amino acids
into a polypeptide chain (via transcription and translation from the blueprints encoded within
genes). An immense variety of polypeptides is possible, due to the combinatorial possibilities
afforded by the 20 types of amino acids. Adding to this richness of diversity, polypeptides fold
into specific structures, dictated by the interactions between their constituent amino acids and
with their surrounding solvent. Furthermore, individual proteins can associate to form larger
structures. This so-called hierarchical design is an important concept and goal in
nanotechnology, in which nanoscale components come together to form functional materials. It
is a protein’s structure which dictates its function, and as a result, an active area of research
concerns the determination of protein structure in order to better understand how it functions in
its particular physiological roles.
120
The sequence of amino acids within a polypeptide is known as its primary structure. When a
protein folds, local segments form repeating structures, stabilized by hydrogen bonds. These
structures are known as secondary structures, of which the most common are α-helices, random
coils, and β-sheets. The overall three-dimensional structure of a folded protein, consisting of a
number of secondary structures, is called the tertiary structure. Proteins with a globular tertiary
structure are generally water soluble and many function as enzymes. Proteins with a fibrous
tertiary structure, such as collagen and keratin, tend to function as structural materials.1
7.3.1 FTIR Characterization of Protein Secondary Structure
Characterization of proteins by FTIR can yield information about the secondary structure by
probing the vibrational modes of functional groups and atomic configurations. A resulting
spectrum contains several well-known features denoted as amide bands, as shown in Figure 7.1.
Figure 7.1 Typical FTIR spectrum of barnacle glue on CaF2 substrate.
The band used most often for the determination of secondary structure is the amide I Band,
which spans the 1600 – 1700 cm-1
spectral region. This band arises primarily from C=O
stretching modes, with minor contributions from out-of-phase CN stretching, CCN deformation,
and NH in-plane bending. Because of these modes, the amide I Band is sensitive to protein
backbone structure, and is not affected by side chains. The amide I Band is centered at around
1650 cm-1, and for proteins with β-structures, band splitting occurs due to a transition dipole
coupling (TDC) mechanism. TDC is essentially a resonance interaction between oscillating
121
dipoles of neighbouring groups, in this case amide groups, and the coupling depends on their
relative orientation and distance.
In terms of interpreting FTIR spectra of proteins, two techniques are commonly used: the second
derivative method,2 and the deconvolution method.
3 Taking the second derivative of a spectrum
enhances subtle changes in the line shapes of bands, thereby revealing distinct secondary
structure band positions within the broad amide I Band. The band center positions obtained from
the second derivative method can be further used as input parameters for the deconvolution
method. This method assumes a Lorentzian line shape for the original unresolved components,
and Gaussian band shapes for deconvolved components.
7.3.2 Barnacle Cement: Proteinaceous Glue
Sessile marine fouling organisms such as mussles and barnacles employ proteinaceous glues in
order to anchor themselves to substrates. In some species, the proteins responsible for the
attachment strength have been identified.4 In this study, the link between adhesive
composition/morphology and mechanical toughness is explored.
7.4 Experimental Methods
7.4.1 Barnacle Rearing
Adult A. amphitrite (= B. amphitrite) on silicone (Silastic® T2 (Dow Corning) or Veridian®
(International Paint)) panels were prepared at Duke Marine Laboratory (Rittschof et al. 1984;
Holm et al. 2005), and shipped overnight to the University of Toronto. The panels were placed in
a 45 L fish tank filled with 35 ppt artificial seawater (Marineland, Instant Ocean Mix, Ohio) at
room temperature under a constant 12 h light/dark cycle. Aliquots (1.5–2 ml) of nauplius larvae
of the brine shrimp Artemia sp. were fed to the barnacles daily. Before performing experiments,
the barnacles were removed from the silicone panels in shear using a mechanical force gauge
(Imada, Northbrook, Illinois), following the procedure in ASTM D5618–94,5 to ensure future
reference capabilities.
122
7.4.2 Fourier Transform Infrared (FTIR) Spectroscopy
Barnacles with opaque glue were detached from Veridian panels and reattached onto CaF2
windows for two days and kept in 35 ppt seawater at room temperature. Reattached barnacles
were forced off the CaF2 windows before IR spectroscopy was performed on the remaining
gummy cement. Fresh glue secreted by the adhesive ducts of barnacles was also collected and
smeared on CaF2 windows.
IR spectra at 2 cm-1
resolution on the cement were collected using an FTIR spectrometer
equipped with a liquid nitrogen cooled MCT detector and dry air purger (Nicolet, Nexus 470,
Minnesota). The region from 1600 to 1700 cm-1
(corresponding to the amide I band) was
examined in detail. A clean CaF2 window with no cement sample was scanned to provide a
background spectrum.
The resulting spectra were processed via Fourier self-deconvolution, followed by peak-fitting, as
per the well-established method of Byler and Susi.3 First, the second derivative of the original
FTIR spectrum was taken in order to identify the peak positions of the multiple secondary
structure components within the Amide I band. Next, Fourier self-deconvolution was performed
on the original FTIR spectrum in order to sharpen the spectral features (secondary structure
components) present in the Amide I band. This process was performed with the FTIR analysis
software (GRAMS, Salem, NH), which uses the Griffiths-Pariente method. This method utilizes
two filters, an exponential filter of the form e2πγx
, and a low pass smoothing filter. The
exponential filter performs the deconvolution; but in doing so, it also increases the noise in the
spectrum. Therefore, the smoothing filter is used to compensate for the increased noise. The
software accepts two input parameters: “γ” (= full width at half height of the widest resolvable
peak), and “Smoothing %”. Values of γ = 6.5 cm-1
(following the parameters chosen by Byler
and Susi3), and Smoothing % = 75 were used. The frequencies of the resulting resolved bands
were checked against those identified in the second derivative spectrum. The deconvolved
spectrum was then subjected to peak fitting in order to determine the integrated areas of the
resolved bands. The Levenberg–Marquadt method was used, assuming a Lorentzian line shape
for the deconvolved spectrum and Gaussian peak shapes for the fit. The integrated areas were
then used to estimate the percentage of each major type of secondary structure (eg. α-helix,
random coil, β-sheet, turns) found in the barnacle protein. Uncertainties when using this
123
approach are typically 10%.6 Bands were assigned to their respective secondary structures based
on the summary of amide I band IR spectral features in H2O by Barth.7
7.4.3 Atomic Force Microscopy (AFM) Imaging and Indentation
Barnacles with opaque glue were detached from Veridian panels and reattached onto glass
coverslips for two days and kept in 35 ppt seawater at room temperature. Reattached barnacles
were forced off the glass coverslips in shear, and AFM experiments were carried out on the
remaining cement in artificial seawater.
AFM topographic images of the cement were obtained using tapping (non-contact) mode
(Asylum Research, MFP 3D, Santa Barbara, CA). A Si3N4 probe (Veeco, DNP, California) was
used, with a tip radius of ~25 nm. The spring constant was determined to be 0.58 N/m by the
thermal noise method.8 Indentation measurements were obtained in contact mode on various
morphologies identified in the topographic images. Force versus indentation curves were
obtained from the deflection of the AFM cantilever as a function of the piezo vertical
displacement plots. The data were processed using Igor Pro software (Wavemetrics, Portland,
OR) and fit with the Sneddon-Hertz model (Equation 7.1) in order to extract the Young’s
modulus.
Equation 7.1 2/3
2 )1(3
4
v
REF Y
In the expression above, F is the loading force [N], EY is Young’s modulus [Pa], R is the radius
of curvature of the tip [m], δ is the indentation [m], and ν is the Poisson’s ratio, which is taken
here as 0.5.9, 10
7.4.4 Scanning Electron Microscopy (SEM) and Energy Dispersive X-ray (EDX)
Barnacles with opaque glue were removed from the silicone panels and reattached onto
aluminum foil for longer than one week during which time they were fed with nauplius larvae of
the brine shrimp Artemia sp. daily. The barnacles were detached and the remaining cement on
the foil was coated with carbon for imaging in the SEM (S-570, Hitachi, Japan). EDX spectra
(Oxford Instruments, Inca Systems, United Kingdom), were then obtained on the fibrillar
124
features observed under the SEM. In all SEM-EDX measurements, a 5 keV beam energy was
used, resulting in an interaction volume of ~1 μm3.
7.4.5 Chemical Staining
Thioflavin T (ThT) and Congo red were used as purchased (Sigma-Aldrich, Madison). ThT
binding was carried out at room temperature as described by Mostaert et al.11
Barnacles with
opaque glue were detached from Veridian panels and reattached onto CaF2 windows for a
maximum of two days and kept in artificial seawater at room temperature. The cement remaining
on a CaF2 window after the barnacle was removed was stained with 10 mM ThT for 5 min. It was
then rinsed with excess 18 MΩ deionized water (Millipore, Milli-Q), and air-dried. Images were
taken with a laser confocal microscope (Leica, TCS SP2, Germany). For Congo red staining, the
barnacles were reattached to glass coverslips for two days and then removed; glass was used
instead of CaF2 because of the birefringence of CaF2. The cement remaining on the coverslip was
stained with 0.5% Congo red in 50% ethanol in water for 5 min. It was then rinsed with excess
18 MΩ water and blow dried with N2 gas prior to examination using the same microscope as
before, now equipped with crossed polarizers.
7.5 Results
7.5.1 FTIR Spectra
The transmission FTIR spectrum of barnacle cement on a CaF2 substrate gave signatures of β-
sheet and a large fraction of random coil conformations. Figure 7.2 shows the deconvoluted
FTIR spectrum for cement of A. amphitrite, fit with four bands (SE: 0.0003, correlation value R2
= 0.999). To check the validity of this fit, multiple fits were performed on the same deconvoluted
FTIR spectrum, from two to 11 bands. The SE did not undergo significant improvement for fits
with more than four bands. Assignment of the four bands to their corresponding secondary
structure resulted in the identification of a large random coil component at 1655 cm-1
(79%), and
smaller low and high frequency β-sheet components, at 1623, 1637 and 1692 cm-1
. The
distribution of secondary structures obtained from the FTIR spectra by the peak fitting algorithm
is summarized in Table 7.1. Note that this distribution applies to the barnacle cement as a whole,
which is made up of a number of different proteins in unknown proportions.
125
Figure 7.2 FTIR spectra of the bulk cement from A. amphitrite. Top spectrum, the original data before processing, offset for clarity; Bottom spectrum, the deconvoluted data, with a four-band peak fit showing peaks corresponding to random coil structure (1655 cm
-1) and low (1623, 1637 cm
-1) and high (1692 cm
-1) frequency β-sheet components.
Table 7.1 IR peaks and the corresponding fraction of the observed secondary structures found in gummy barnacle cement sample
Secondary structure band positions in H2O (cm-1
) Relative peak area fraction (%)
Low-frequency β-sheet components (1623, 1637) 17
Random coil (1655) 79
High-frequency β-sheet component (1692) 4
7.5.2 AFM Images, Force Curves, and Moduli Histograms
After reattachment to glass substrates, barnacles typically to produced ~1 mm wide glue deposits
below the periphery of their baseplate, consistent with the location of the glue ducts. After
pushing these barnacles off the substratum, the residual glue on the substratum was imaged with
AFM. Typical topography images are shown in Figure 7.3. Mesh structures such as those
reported earlier by Weigemann and Watermann were observed.12
AFM images of the barnacle
126
cement in both 35 ppt artificial seawater (Figure 7.3A) and in air (Figure 7.3B) exhibit this mesh
morphology. The mesh structure in Figure 7.3 is the general picture of the adhesive when
scanned in a larger area.
Figure 7.3 AFM topographic image of the barnacle cement in (A) 35 ppt sea water and (B) air. Gray-scale provides
height reference (right hand of each image). Images are 15 x 15 m2.
Upon zooming in, the mesh in Figure 7.3A was observed to be composed of a mixture of
different structures such as those shown in Figure 7.4A–C. Figure 7.4A shows clusters of
globular structures with diameters ranging from 60 to 100 nm. Pearl-like arrangements of the
globular aggregates are clearly seen. Smaller globular structures (10–30 nm) and a small, rod-
like structure of 11 nm in diameter and ~300 nm length were also observed (Figure 7.4B).
Figure 7.4C shows a larger, more regular rod-shaped structure. An unstructured aggregate
(Figure 7.4D) is also seen in the matrix (Figure 7.4E). Thus, the bulk cement is composed of the
mesh comprised of the structures in Figure 7.4A–C and of the matrix (Figure 7.4E) with some
127
unstructured aggregates (Figure 7.4D). There is no specific pattern or localization of those
structures. The relative fraction of those occurring is variable from one barnacle sample to
another, even from within the same species.
Figure 7.4 The bulk barnacle cement is composed of both the mesh structure and the matrix. A: Globular aggregates. B: Smaller rod-like and smaller globular features. (C) A larger, more regular rod-shaped structure
comprising the mess. (D) An unstructured aggregate in the matrix. (E) Bulk glue. The black dots in figures (C) to (E) indicate the point of indentation.
AFM nanoindentation was performed on the rod-shaped structure (Figure 7.5A), the unstructured
aggregate (Figure 7.5B), and the matrix (Figure 7.5C), to determine and compare their elastic
moduli. The black dot in each figure indicates the point of indentation. Individual fits for the
larger, more regular rod-shaped structure, the unstructured aggregate, and the matrix and the
corresponding histogram of elastic modulus are shown below the images of each structure in
Figure 7.5A–C.
128
Figure 7.5 In situ determination of the elastic modulus of the barnacle cement. Elastic modulus distribution of individual nanostructures/components observed by AFM. (A) Fit to the force-indentation plot with the corresponding histogram of the elastic modulus for larger, more regular rod-shaped structure. (B) The unstructured aggregate. (C)
The matrix. Insets in (B) and (C) are linescans (taken across the feature, as shown by the line drawn across it) of the larger, more regular rod-shaped structure and unstructured aggregate, respectively. For both rod-shaped and
unstructured aggregate structures, the height of the feature is ~250 nm.
The corresponding force-indentation plots were fit by Sneddon mechanics.9 The solid line
represents the fit to indentation data by the paraboloidal tip model (Equation 7.1). The larger,
more regular rod-shaped structure (Figure 7.5A) was found to have a Young’s modulus in the
range of 20–90 MPa; the unstructured aggregate, 0.20–2 MPa; and the matrix, 1–10 MPa. Insets
in Figure 7.5A and B are the linescans (taken across the feature as shown by the line drawn
across it) of the larger, more regular rod-shaped structure and unstructured aggregate,
respectively. For both rod-shaped and unstructured aggregate structures, the height of the
indented feature is ~250 nm and the rod had a diameter of ~600 nm. Indentations were <20% of
the sample thickness so finite thickness effects that can affect the predicted modulus are
minimal.13
129
7.5.3 SEM and EDX
To determine the chemical composition of the rod-shaped features observed in the AFM, SEM
images coupled to EDX were obtained. The rod-shaped features observed under SEM (Figure
7.6A) showed high counts of carbon, nitrogen and oxygen in the EDX line scan spectrum (Figure
7.6B) suggestive of the organic nature of the components. Figure 7.6C and D shows an EDX
area mapping, indicating the spatial location of the elements in a given region. Consistent with
the EDX line scan in Figure 7.6A and B, the rod-shaped features in Figure 7.6C showed signals
of C, N and O (Figure 7.6D) coming from the rod-shaped structures, again indicative of the
organic nature of the components. The high counts of Al for both the EDX line scan and area
mapping are attributed to the background aluminum foil where the cement was attached. The Mg
background was similar to Al. The F elemental map is shown as a control. The elemental X-ray
lines are indicated for each line scan and map (a = α).
Figure 7.6 SEM images and EDX spectra of the barnacle cement resettled on aluminum foil. (A) SEM image showing the rod-shaped structures and the corresponding (B) EDX line scan spectra showing high counts of elemental
carbon, nitrogen and oxygen coming from the rod-shaped structures. (C) SEM image of the rod-shaped structures and the corresponding (D) elemental maps of the elements. Fair signals of C, N and O coming from the rod-shaped structures are obtained. The elemental X-ray lines from which the EDX signals were derived are indicated for each
line scan and map (a = α)
130
7.5.4 Amyloid-Selective Staining
To test for the presence of amyloids within the cement, chemical staining with amyloid-selective
dyes was performed. ThT staining resulted in fluorescence of the rod-shaped structures (Figure
7.7B), and Congo-red staining resulted in apple-green birefringence under cross-polarized light
(Figure 7.7D). For comparison, an image of the barnacle cement in the absence of ThT is shown
in Figure 7.7A, and a bright field image of the cement stained with Congo red is shown in Figure
7.7C.
Figure 7.7 Chemical staining images of the barnacle cement with amyloid-selective dyes. (A) Confocal images of the barnacle cement on a CaF2 window without ThT. (B) Stained with ThT. (C) Bright field image of the barnacle cement stained with Congo red. (D) Polarized light image of Congo red sample. Fluorescence and apple-green birefringence
of the rod-shaped structures are observed when the barnacle cement is stained with ThT and Congo red, respectively.
7.6 Discussion
7.6.1 Significance of β-sheet Conformation in Barnacle Cement
AFM indentation on some of the morphologies seen in Figure 7.4A–E gave a wide range of
elastic modulus values as shown in Figure 7.5A–C, indicating that the elasticity of the cement is
heterogeneous. From the force indentation plot in Figure 7.5A, the depth of indentation changes
only slightly as the force is increased for the rod-shaped structure. In the case of the unstructured
aggregate and of the matrix, the depth of indentation increases as the force is increased. This is
131
reflective of the more compliant nature of these morphologies. These results indicate the stiffer
nature of the larger, more regular rod-shaped structure than both the unstructured aggregate and
the matrix. This is also evident from the magnitude of their elastic moduli. The rod-shaped
structure has an elastic modulus in the range of 20 – 90 MPa, which is similar in range to what
has been observed for the amyloid fibrils in insulin (5–50 MPa).14
Materials with higher elastic modulus typically require more stress to fracture.15
The higher
elastic modulus value for the rod-shaped structure suggests that it has a stiffer nature with higher
cohesive forces. This causes the load applied to it at any point to be distributed across its length,
making a composite more resistant to fracture. A simple way to understand how the rod-shaped
structures could contribute to the cement strength is to consider that they distribute load across a
relatively wide region of the bulk cement. The results do not allow determination of how
important this effect is for barnacle cement toughness, but their presence needs to be considered
in a detailed model for fracture. Kamino et al.16
previously suggested that the 100 kDa protein
present in the barnacle cement of M. rosa may be similar to the proteins involved in the
formation of amyloid fibrils. This was based on the abundance of β-sheet structures in the 100
kDa barnacle protein, its alternating hydrophobic and hydrophilic profile, and its very insoluble
nature. It was reported that the pattern of alternating polar and non-polar residues in a cross-β
structure is essential in the formation of the insoluble amyloid fibrils.17
Recently, amyloid-like sequences have been found in the primary structure of the protein in the
bulk cement of the M. rosa.4 Moreover, the 100 kDa cement protein of M. rosa is particularly
rich in isoleucine (Ile), valine (Val) and threonine (Thr) residues, which are the three amino acid
residues reported to have the highest propensity to form the β-sheet structure.18
Chemical staining of the barnacle cement with amyloid-selective dyes was done to check for the
possible presence of amyloids in the cement proteins of A. amphitrite. The apple-green
birefringence under polarized light when stained with Congo red (Figure 7.7C), and fluorescence
of ThT (Figure 7.7B) indicate the presence of an amyloid fibril structure in the barnacle cement.
Both ThT and Congo red dyes are known to be amyloid-selective.19-22
The fluorescence observed
in the presence of ThT and the apple-green birefringence with Congo red are related to the cross-
132
β core structure, both of which are indications of the binding of the dyes to an amyloid fibril
structure in the barnacle cement.
A variety of rod-shaped structures (or fibre-like features) were observed under the optical
microscope, but only a small portion gave an apple-green birefringence and fluorescence when
stained with Congo red and ThT, respectively. This indicates that the amyloid fibrils in the
barnacle cement only comprise a small fraction of the bulk cement. Such a small fraction is still
significant because, in most fiber reinforced composite materials, fibers need to be present in
only a minute fraction to lead to a considerable increase in the toughness of the material.23, 24
Non-amyloid fibers could also play a similar mechanical role.
7.7 Conclusions
Different nanoscale structures were observed in the bulk cement of the barnacle A. amphitrite
through AFM, viz. a mesh which is composed of rod-shaped, threadlike and globular structures
and a matrix with unstructured aggregates. Indentation data suggested a stiffer nature of the rod-
shaped structure over the other components of the bulk cement. The EDX spectrum of the rod-
shaped structures was suggestive of its organic nature. The FTIR spectra supported the presence
of a β-sheet conformation, whereas the results of chemical staining with both ThT and Congo red
confirmed the presence of a small fraction of amyloid fibrils in the bulk cement.
7.8 Contributions
The author performed the FTIR, SEM, and EDX measurements and analysis. The material in
this chapter, based on Sullan, R. M. A.; Gunari, N.; Tanur, A. E.; Chan, Y.; Dickinson, G. H.;
Orihuela, B.; Rittschof, D.; Walker, G.C. Nanoscale structures and mechanics of barnacle
cement. Biofouling 2009, 25, 263-275, has been presented with an emphasis on A. E. T.’s
contributions.
7.9 References
1. Matthews, H. R.; Freedland, R. A.; Miesfeld, R. L., Proteins. In Biochemistry - A Short
Course, John Wiley & Sons: New York, 1997; pp 25–43.
2. Susi, H.; Michael Byler, D., Protein structure by Fourier transform infrared spectroscopy:
Second derivative spectra. Biochem. Bioph. Res. Co. 1983, 115, 391–397.
133
3. Byler, D. M.; Susi, H., Examination of the secondary structure of proteins by
deconvolved FTIR spectra. Biopolymers 1986, 25, 469–487.
4. Kamino, K., Underwater adhesive of marine organisms as the vital link between
biological science and material science. Mar. Biotechnol. 2008, 10, 111–121.
5. International, A., Standard test method for measurement of barnacle adhesion strength in
shear. 1997, 2005; Vol. Designation D5618–94.
6. Dousseau, F.; Pezolet, M., Determination of the secondary structure content of proteins in
aqueous solutions from their amide I and amide II infrared bands. Comparison between
classical and partial least-squares methods. Biochemistry 1990, 29, 8771–8779.
7. Barth, A., Infrared spectroscopy of proteins. Biochimica et Biophysica Acta (BBA) -
Bioenergetics 2007, 1767, 1073–1101.
8. Hutter, J. L.; Bechhoefer, J., Calibration of atomic-force microscope tips. Rev. Sci.
Instrum. 1993, 64, 1868–1873.
9. Sneddon, I. N., The relation between load and penetration in the axisymmetric boussinesq
problem for a punch of arbitrary profile. Int. J. Eng. Sci. 1965, 3, 47–57.
10. Sun, Y.; Guo, S.; Walker, G. C.; Kavanagh, C. J.; Swain, G. W., Surface elastic modulus
of barnacle adhesive and release characteristics from silicone surfaces. Biofouling 2004,
20, 279–289.
11. Anika, S. M.; Suzanne, P. J., Beneficial characteristics of mechanically functional
amyloid fibrils evolutionarily preserved in natural adhesives. Nanotechnology 2007, 18,
044010.
12. Wiegemann, M.; Watermann, B., Peculiarities of barnacle adhesive cured on non-stick
surfaces. J. Adhes. Sci. Technol. 2003, 17, 1957–1977.
13. Akhremitchev, B. B.; Walker, G. C., Finite sample thickness effects on elasticity
determination using atomic force microscopy. Langmuir 1999, 15, 5630–5634.
14. Guo, S.; Akhremitchev, B. B., Packing density and structural heterogeneity of insulin
amyloid fibrils measured by AFM nanoindentation. Biomacromolecules 2006, 7, 1630–
1636.
15. Griffith, A. A., The phenomena of rupture and flow in solids. Philos. Trans. R. Soc.
London, Ser. A 1921, 221, 163–198.
16. Kamino, K.; Inoue, K.; Maruyama, T.; Takamatsu, N.; Harayama, S.; Shizuri, Y.,
Barnacle cement proteins. J. Biol. Chem. 2000, 275, 27360–27365.
17. West, M. W.; Wang, W.; Patterson, J.; Mancias, J. D.; Beasley, J. R.; Hecht, M. H., De
novo amyloid proteins from designed combinatorial libraries. Proc. Natl. Acad. Sci.
1999, 96, 11211–11216.
18. Xiong, H.; Buckwalter, B. L.; Shieh, H. M.; Hecht, M. H., Periodicity of polar and
nonpolar amino acids is the major determinant of secondary structure in self-assembling
oligomeric peptides. Proc. Natl. Acad. Sci. U.S.A.1995, 92, 6349–6353.
19. Missmahl, H. P.; Hartwig, M., Polarisationsoptische untersuchungen an der
amyloidsubstanz. Virchows Arch. Path. Anat. 1953, 324, 489–508.
134
20. Janigan, D. T., Experimental amyloidosis. Structural relationships of amyloid and
reticulin in tissue sections and isolated preparations. Am. J. Pathol. 1966, 49, 657–678.
21. Stopa, B.; Piekarska, B.; Konieczny, L.; Rybarska, J.; Spolnik, P.; Zemanek, G.;
Roterman, I.; Krol, M., The structure and protein binding of amyloid-specific dye
reagents. Acta Biochim. Polonica 2003, 50, 1213–1227.
22. Khurana, R.; Coleman, C.; Ionescu-Zanetti, C.; Carter, S. A.; Krishna, V.; Grover, R. K.;
Roy, R.; Singh, S., Mechanism of thioflavin T binding to amyloid fibrils. J. Struct. Biol.
2005, 151, 229–238.
23. Melanitis, N.; Galiotis, C.; Tetlow, P. L.; Davies, C. K. L., Interfacial shear stress
distribution in model composites part 2: Fragmentation studies on carbon fibre/epoxy
systems. J. Compos. Mater.1992, 26, 574–610.
24. Savastano Jr, H.; Warden, P. G.; Coutts, R. S. P., Potential of alternative fibre cements as
building materials for developing areas. Cem. Concr. Compos. 2003, 25, 585–592.
135
8 Summary and Outlook
8.1 Summary of Thesis
The study of nanomaterials requires a multi-faceted approach in order to determine the
morphology, chemical composition, and mechanical properties of the nanoscale features. Each
technique provides complementary information, helping to build a fuller understanding of the
links between material structure, properties, and function. For the study of boron nitride
nanotubes (Chapters 3-5), scanning and transmission electron microscopy (SEM and TEM) were
essential techniques for the characterization of nanotube morphology and structure. Energy
dispersive x-ray (EDX) spectroscopy and mapping, combined with SEM, identified the presence
of elements of interest that were expected to be present in the synthesized nanotubes. B and N
signals were observed from the nanotubes, and O, Fe, Mg, and Al were also detected within the
synthesized product prior to purification as remnants of the precursor and catalyst powders.
SEM and EDX were also indispensable tools in the identification of aragonite and Mg-calcite
phases within the calcareous tube of H. dianthus (Chapter 6). It was possible to choose specific
locations to probe with EDX based on the SEM images, such that the different CaCO3
polymorphs could be identified and correlated with layers within the tube based on the Sr signal
for aragonite, and the Mg signal for Mg-calcite. In Chapter 7, EDX was used to determine that
the nanoscale rods observed within the barnacle cement were organic and proteinaceous in
nature, through the observation of C, N, and O signals.
Fourier transform infrared spectroscopy (FTIR) was useful for characterizing the bulk chemical
composition of boron nitride nanotubes (BNNTs) in Chapters 4 and 5. Shifts in the peak
position of the E1u TO mode for h-BN were qualitative indications of the crystallinity of the
nanotubes. It was also able to provide signatures of nanoscale structures, namely the β-sheet
conformation within proteins in barnacle cement (Chapter 7). In addition, in Chapter 6, FTIR
determined that the soluble and insoluble organic matrices within the calcareous shell of H.
dianthus contained substantial amounts of carboxylated and sulphated polysaccharides,
molecules which have been demonstrated to mediate CaCO3 mineralization through the
interaction between their functional groups and ions within seawater.
136
Atomic force microscopy (AFM) was used to determine the bending modulus of individual
BNNTs in Chapter 5. The force mapping technique was employed in order to determine the
boundary conditions of each nanotube, and revealed that the assumption made in many studies
that such systems possessed only clamped ends was false. As a result, the bending modulus was
determined using the appropriate beam models for each type of boundary condition. AFM
nanoindentation was used to measure the Young’s modulus of biological materials in Chapters 6
and 7, in the investigation of tubeworm and barnacle adhesives both in air, and in an artificial sea
water solution. The versatility of the AFM is showcased through the above studies.
8.2 Outlook
8.2.1 Boron Nitride Nanotubes
The potential of BNNTs as an advanced functional material remains to be fully realized. Most
synthesis methods require further development in order to produce industrial scale yields. The
unique set of properties that BNNTs possess makes it a worthwhile endeavour, however. Several
exciting possible applications for BNNTs include their incorporation as an emitting layer in
organic light emitting diode architectures, outputting UV light; thermally conductive electrically
insulating composite polymer encapsulants for electronics; additives to ceramics and metals to
enhance damping characteristics; and targeted boron neutron capture therapy agents for the
treatment of cancerous tumors.1-4
8.2.2 Nanomaterials in Nature
The study of the materials design principles employed by organisms such as tubeworms and
barnacle continues to inspire the design and synthesis of novel biomimetic materials. In
particular, biomineralization offers lessons in the growth of nano and microcrystals in mild
aqueous conditions, with exceptional control over chemical composition and morphology. The
organic molecules which mediate biomineralization, including polysaccharides and proteins,
give materials scientists new ideas and new tools with which to create new materials.5-7
137
8.3 References
1. Li, L. H.; Chen, Y.; Lin, M. Y.; Glushenkov, A. M.; Cheng, B. M.; Yu, J., Single deep
ultraviolet light emission from boron nitride nanotube film. Appl. Phys. Lett. 2010, 97,
141104–3.
2. Zhi, C.; Bando, Y.; Terao, T.; Tang, C.; Kuwahara, H.; Golberg, D., Towards
thermoconductive, electrically insulating polymeric composites with boron nitride
nanotubes as fillers. Adv. Funct. Mater. 2009, 19, 1857–1862.
3. Sueyoshi, H.; Rochman, N. T.; Kawano, S., Damping capacity and mechanical property
of hexagonal boron nitride-dispersed composite steel. J. Alloy Compd. 2003, 355, 120–
125.
4. Ciofani, G.; Raffa, V.; Yu, J.; Chen, Y.; Obata, Y.; Takeoka, S.; Menciassi, A.;
Cuschieri, A., Boron nitride nanotubes: A novel vector for targeted magnetic drug
delivery. Curr. Nanosci. 2009, 5, 33–38.
5. Barnard, A. S.; Russo, S. P., Modeling the environmental stability of FeS2 nanorods,
using lessons from biomineralization. Nanotechnology 2009, 20, 115702.
6. Oaki, Y.; Adachi, R.; Imai, H., Self-organization of hollow-cone carbonate crystals
through molecular control with an acid organic polymer. Polym. J. 2012, 44, 612–619.
7. Nudelman, F.; Sommerdijk, N. A. J. M., Biomineralization as an inspiration for materials
chemistry. Angew. Chem. Int. Ed. 2012, 51, 6582–6596.