Number 45 | 2009

Top science …free institutional access

Number 1 | 2009

Chemical Science
www.rsc.org/chemicalscience Volume 1 | Number 1 | 2010
Chemical Science

Chemical Communications
www.rsc.org/chemcomm Number 45  |  7 December 2009  |  Pages 6865–7052

ChemComm
ISSN 2041-6520
Pages 1–140

New for 2010

Chemical Science - a new journal presenting findings of exceptional significance from across the chemical
sciences. www.rsc.org/chemicalscience

MedChemComm - focusing on medicinal chemistry research, including new studies related to

biologically-active chemical or biochemical entities that can act as pharmacological agents with therapeutic
potential or relevance. www.rsc.org/medchemcomm

Polymer Chemistry - publishing advances in polymer chemistry covering all aspects of synthetic and
biological macromolecules, and related emerging areas. www.rsc.org/polymers

New for 2009

Analytical Methods - highlights new and improved methods for the practical application of analytical
science. This monthly journal will communicate research in the advancement of analytical techniques for use
by the wider scientific community. www.rsc.org/methods

Integrative Biology - focusing on quantitative multi-scale biology using enabling technologies and tools to
exploit the convergence of biology with physics, chemistry, engineering, imaging and informatics.
www.rsc.org/ibiology

Metallomics - covering the research fields related to metals in biological, environmental and clinical systems.
www.rsc.org/metallomics

Nanoscale - publishing experimental and theoretical work across the breadth of nanoscience and
nanotechnology. www.rsc.org/nanoscale

Free institutional access, managed by IP address, is available on all these titles.

For more details, and to register, visit www.rsc.org/free_access_registration
Pages 6865–7052

www.rsc.org/journals
ISSN 1359-7345

Registered Charity Number 207890

FEATURE ARTICLE FEATURE ARTICLE
Julie V. Macpherson, Patrick R. Unwin et al. Zhongfan Liu et al.
Electrochemistry at carbon nanotubes: Raman spectroscopy of strained single-
1359-7345(2009)45;1-T
perspective and issues walled carbon nanotubes
FEATURE ARTICLE www.rsc.org/chemcomm | ChemComm

Electrochemistry at carbon nanotubes: perspective and issues

Ioana Dumitrescu, Patrick R. Unwin* and Julie V. Macpherson*
Received (in Cambridge, UK) 18th May 2009, Accepted 17th August 2009
First published as an Advance Article on the web 23rd September 2009
DOI: 10.1039/b909734a

Electrochemistry at carbon nanotubes (CNTs) is a large and growing field, but one in which there is
still uncertainty about the fundamental activity of CNTs as electrode materials. On the one hand, there
are many reports which focus on the favourable electrochemical properties of CNT electrodes, such as
enhanced detection sensitivity, electrocatalytic effects and reduced fouling. On the other hand, other
studies suggest that CNTs may be no more electroactive than graphitic powder. Furthermore, it has
been proposed that the catalytic nanoparticles from which CNTs are formed may dominate the
electrochemical characteristics in some instances. A considerable body of the literature presumes that
the CNT sidewall is inert and that edge-plane-graphite-like open ends and defect sites are responsible
for the electron transfer activity observed. In contrast, studies of well characterised single-walled
nanotube (SWNT) electrodes, either as individual tubes or as two-dimensional networks, suggest
sidewall activity. This review highlights how the various discrepancies in CNT electrochemistry
may have arisen, by taking a historical view of the field and identifying crucial issues that still need
to be solved. When assessing the behaviour of CNT electrodes, it is vitally important that careful
consideration is given to the type of CNT used (SWNT or multi-walled), the quality of the material
(presence of impurities), the effect of chemical processing steps in the fabrication of electrodes and the
experimental arrangements adopted. Understanding these key features is an essential requirement
to develop a fundamental understanding of CNT electrochemistry, to allow a wide range of
electroanalytical applications, and to move the field forward rationally. As part of this process,
high resolution electrochemical and electrical imaging techniques are expected to play a significant
role in the future, as well as theoretical developments which examine the fundamentals of electron
transfer at different types of CNTs and their characteristic surface sites.

1. Introduction aspect ratio, high strength and excellent current carrying

Carbon nanotubes (CNTs) possess a remarkable array of
structural, mechanical and electronic properties, such as high
Department of Chemistry, University of Warwick, Coventry, UK CV4 7AL.
E-mail: p.r.unwin@warwick.ac.uk, j.macpherson@warwick.ac.uk
capability.1 These properties have fuelled an exponential
increase in the number of publications and patents on CNTs
over the last decade2 and led to CNTs being considered as a
‘‘hot topic’’ in many areas.3 Much research has concentrated
on the application of CNTs in areas such as composite
materials, scanning probe microscopy, field emission,

Ioana obtained her BSc in Patrick was born in West

Ioana Dumitrescu
analytical chemistry from Yorkshire in 1964 and
the University of Bucharest, educated at the Universities
Romania, in 2004. She then of Liverpool and Oxford.
obtained a Chevening Scholar- Following fellowships at
ship to study for an MSc at the Balliol College Oxford and at
University of Warwick, UK. the University of Texas at
In 2005 she started her PhD Austin, Patrick was appointed
in the Electrochemistry and to a lectureship in Physical
Interfaces group of Prof. Julie Chemistry at the University
Macpherson and Prof. Patrick of Warwick in 1991. In 1998
Unwin at the University he was promoted to Professor
of Warwick. Her research of Chemistry and is currently
interests lie in the field of the Director of the Warwick
electrochemistry at nanoscale Patrick R. Unwin Centre for Analytical Science.
interfaces. His research interests lie in the
development of new experimental techniques and the theoretical
treatment of coupled interfacial/mass transport problems.
Unknown to many, Patrick is an accomplished song writer and
guitarist.

6886 | Chem. Commun., 2009, 6886–6901 This journal is

c The Royal Society of Chemistry 2009
nanoelectronics and solar cell technology.4,5 CNTs have also
found extensive use as analytical tools, in the construction of
devices and sensors exploiting their small size and fascinating
electrical properties.6
In the electrochemical arena, the combination of high aspect
ratio, nanometre-sized dimensions, good electrical conductivity7
and low capacitance in the pristine state (arising from the low
density of states of SWNTs)6,8,9 dictates that CNTs have the
capability to make excellent electrodes.6,10 Carbon electrode
materials have been used extensively in electroanalytical
chemistry since the early 1960s.11,12 Attributes include chemical
stability, wide electrochemical potential window in aqueous
solution, biocompatibility and ready availability. In addition,
the complex surface chemistry of carbon can, in some cases,
aid electrocatalysis.6 There has been much excitement about
the prospect of CNTs as new carbon electrode materials,
but, as discussed herein, questions remain regarding the
fundamental electrochemical properties of CNTs, that need
to be resolved for the field to advance in a rational manner.
This review aims to highlight unresolved issues and unanswered
questions by tracing the history of CNTs in electroanalytical
chemistry, pointing out inconsistencies that have developed in
the literature and suggesting ways to move the field forward.
1.1 Structure and electronic properties of carbon nanotubes
It is important to consider the electronic properties of CNTs,
compared with macroscopic carbon structures, as these
may influence the electron transfer (ET) characteristics of
CNT-based electrodes. CNTs are part of the sp2 carbon
family, along with graphene (a two-dimensional honeycomb
structure) and highly oriented pyrolytic graphite (HOPG:
comprising layers of graphene sheets). Single-walled CNTs
(SWNTs)13 can be viewed as one rolled-up sheet of graphene,
while multi-walled CNTs (MWNTs)14 are concentrical tubes
of rolled-up graphene, separated by the interlayer distance of
graphite (0.34 nm), as shown in Fig. 1. MWNTs can also exist
in herringbone and bamboo configurations.15 SWNTs are
typically 0.7–3 nm in diameter, while a diameter range of
2–30 nm is characteristic for MWNTs. Both SWNTs and
MWNTs are usually microns to many tens of microns in
length, although it is possible to grow individual SWNTs up
to centimetres long.16
Julie was born in Sutton Cold-
field in 1972 and was educated
at the University of Warwick.
She obtained a University
Royal Society Fellowship in
1999 and upon completion of
her fellowship in 2007 was
promoted to a professorship
at the University of Warwick.
Her research interests lie in
the development of high reso-
lution electrochemical imaging
techniques and carbon-based
electrodes, fundamentals and
Julie V. Macpherson applications. She is also a
published photographer.
This journal is
c The Royal Society of Chemistry 2009
SWNTs have interesting electrical properties, deriving from
the sp2 hybridisation of carbon bonds, quasi-one-dimensional
nature, and cylindrical symmetry.1 The electronic structure of
SWNTs has been studied extensively and is generally described
using band structure representations, taking graphene as a
starting point and then wrapping to form a tube. The band
structure of graphene is quite unusual;17 it is a zero-band gap
semiconductor, and has both metallic and semiconducting
properties in different directions. Depending on the chirality
of the wrap, the resulting SWNTs can be either metallic
(mSWNT) or semiconducting (sSWNT). In a random sample
of SWNTs, it is predicted that 1/3rd will be metallic, whilst
the remaining 2/3rd are semiconducting (with a band gap
proportional to 1/d, where d is the diameter of the tube and
of the order of B0.5 eV).1 With MWNTs, only one of
the concentric tubes needs to have metallic character for the
overall electronic properties to be essentially metallic. The
band structure of graphene and the density of states (DOS) of
two SWNTs are shown in Fig. 1(a). Pristine SWNTs, graphene
and HOPG all have low DOS around the Fermi level,6,9
compared to a typical metal. This implies low quantum
capacitance values. For the specific case of SWNTs, theoretical
modelling by Heller et al. showed that the ET rate will be
different for SWNTs of different chirality,9 although this has
not yet been addressed experimentally.
MWNTs and HOPG are three-dimensional structures and
thus display terraces and steps (edges), along with point and
other surface and sub-surface defects.15 In contrast, SWNTs
and graphene comprise only surface atoms (Fig. 1(b)) and
defects and on sufficiently small length-scales can exhibit
ballistic conduction.18,19 Common defects of the six-membered-
ring SWNT structure are the inclusion of five- or seven-
membered rings, sp3 hybridised sites and open ends, the
latter typically from chemical processing (often terminated in
oxygen functionalities).20,21
A large body of work, hitherto, suggests that CNTs
(both SWNTs and MWNTs) possess ‘‘special’’ characteristics
(vide infra) for voltammetric analysis, compared with other
carbon electrode materials such as glassy carbon and HOPG.
These characteristics include electrocatalytic properties and
high sensitivity for amperometric detection, but the reasons
have not yet been fully elucidated.22 There is also significant
debate about the fundamental electroactivity of CNTs, which
has been likened to that of HOPG.15 For example, MWNTs
have been described as HOPG-like, with a sidewall which
equates to the basal plane of graphite and open ends
(after chemical processing) and locations on the tube axis
where graphitic sheets terminate, which are both likened to
edge-plane-steps on the HOPG surface.15 Conventional
voltammetric data (averaged over the CNTs and support
electrode) have been interpreted to suggest that the basal plane
of HOPG and the sidewalls of MWNTs are essentially inert
and that step edges (in HOPG) or edge-plane-like
defects, including open ends (in MWNTs), dominate the
electrochemistry.23,24 In SWNTs edge-plane-like defects, such
as oxygenated ends, are also considered responsible for the
electroactivity.25,26 In contrast, other work employing isolated
SWNTs has demonstrated reversible ET, suggesting that the
sidewalls are active for ET.27,28
Chem. Commun., 2009, 6886–6901 | 6887
Fig. 1 (a) Structure and electronic properties of graphene and SWNTs. SWNTs can be visualised as a graphene sheet wrapped along the chiral
vector. Energy dispersion for graphene: the valence and conduction bands meet at singular points in reciprocal space. DOS for a (6,4) sSWNT and
a (5,5) mSWNT (data taken from http://www.photon.t.u-tokyo.ac.jp/Bmaruyama/kataura/1D_DOS.html); (b) the structural connection between
HOPG, graphene, SWNTs and MWNTs. HOPG and MWNTs are 3D structures, while graphene and SWNTs are only surface atoms.

Currently, there are two fundamental questions in CNT
electrochemistry that still need to be addressed in full. (1) Do
different surface features and defects have different activity and
is this dependent on the type of CNT employed? (2) What role
do catalyst nanoparticles, used to grow CNTs, play in defining
electrochemical activity? In addition, there is also interest
in understanding how the metallic/semiconducting nature of
individual SWNTs influences their intrinsic electroactivity.
1.2 Growth, purification and characterisation of
CNT materials
The way CNTs are grown is vitally important as, for the
majority of growth schemes, significant impurity levels can
result, unless stringent post-growth purification procedures are
used. Given the interfacial nature of electrochemical processes,
the CNT surface state is very important; moreover, it must be
recognised that a SWNT consists entirely of surface atoms!1
The presence of surface impurities, mostly amorphous carbon
and metal nanoparticles (NPs), together with charged surface
groups from post-growth purification, e.g. acidic functionalities,
when strong acid cleaning treatments are employed,21 is likely
to affect the behaviour of CNT-based electrodes. Careful
characterisation of CNTs (vide infra) is thus essential in
order to understand electrochemical processes. As highlighted
herein, certainly in early work with CNT electrodes, such
characterisation was rarely undertaken.
The main methods for growing CNTs are electric arc
discharge, laser ablation, high-pressure carbon monoxide
(HiPCO), and catalysed chemical vapour deposition
(cCVD).29 The first two employ solid-state carbon sources
and involve carbon vaporisation at very high temperatures.
HiPCO produces CNTs from gas-phase reactions of iron
carbonyl in carbon monoxide at high pressures. Although
high-quality CNT structures are usually obtained from the
first three processes, there are large amounts of by-products,
such as graphitic debris, metal NPs and fullerenes
(Fig. 2(a,b,c)).30–32 For example, for arc discharge, the content
of catalyst NPs is often higher than 2 : 1 in catalyst NPs : carbon
weight ratio.31 Thus, CNTs prepared via these routes nearly
always require a clean-up procedure after growth.
In contrast to the preceding methods, cCVD typically
involves the catalytic decomposition of a carbon containing
gas at catalytic metallic NPs, although it has recently been
shown that non-metals can be employed.33,34 When cCVD
conditions are carefully optimised CNTs of high crystallinity
and low defect density can be obtained,29 especially for SWNT
growth. Amorphous carbon content is minimal and metal NP
content is limited typically to a carbide encapsulated particle
at the end of a SWNT (Fig. 2(d)).35
When amorphous carbon and metal NPs content is high, as
for the first three growth methods detailed, it is essential that
the impurities are removed if one wishes to work with a carbon
material and not a composite. Not surprisingly, there has been

6888 | Chem. Commun., 2009, 6886–6901 This journal is

c The Royal Society of Chemistry 2009

Fig. 2 Electron micrograph images of CNTs grown using different
methods, before purification (a) arc-discharge;30 (b) laser-ablation;31
(c) HiPCO;32 and (d) cCVD.35 (a) From ref. 30, with kind permission
of Springer Science and Business Media. (b) Copyright Wiley-VCH
Verlag GmbH & Co. KGaA. Reproduced with permission from
ref. 31. (c) Copyright (2001), American Institute of Physics. Reproduced
with permission from ref. 32. (d) Reprinted with permission from
ref. 35. Copyright 2000 Elsevier Ltd.
extensive interest into methods of removal of impurities from
CNTs. Several highly effective purification techniques have
been established, including flocculation, microfiltration,
chromatography and centrifugation,36 although these are used
rarely. Much more common has been the use of oxidative
treatments, commonly employing strong oxidising acids. Acid
purification typically involves refluxing CNTs in HNO3,
H2SO4, HCl or mixtures of these acids. Mixtures of concentrated
HNO3 and H2SO4 are most effective in removing amorphous
carbon,37 while the HCl treatment, in conjunction with sonication,
is considered to assist the removal of catalytic NPs.38,39
Different groups advocate slightly different acidic treatment
conditions based on variations in duration, concentration, use
of sonication and repeated cycles.
An important consideration is the secondary effect of the
acid purification process, which typically results in CNT
‘‘cutting’’, i.e. the opening of CNT ends and oxidative damage
of the sidewall. This causes cut ends and defects to form in the
CNT sp2 structure which are functionalised with oxygen-
containing groups, such as carboxylic acids, ketones and
alcohols.40–43 While oxygen-containing surface groups can
aid certain electrochemical reactions,44 the presence of sidewall
defects causes significant changes in the CNT electronic states
near the Fermi level, typically resulting in an increase in
resistivity of the CNT at these points.45–47 Thus, the effect of
the purification procedure on the resulting structure and
electronic properties of the CNT must be considered carefully
when fabricating devices or electrode materials using purified
CNTs.44
This journal is
c The Royal Society of Chemistry 2009
2. Development of carbon nanotube electrodes
2.1 Early investigations of CNT electrochemistry
The first report of an electrode based on CNTs was by Britto
et al.48 in 1996. To make the electrode, as-grown arc-produced
(unpurified) CNTs were mixed with bromoform and then
packed inside a glass tube. Little work was carried out to
characterise the chemical composition of the CNT material
employed. Voltammetric investigation of dopamine oxidation
revealed characteristic diffusion controlled peaks on a significant
background current, with a peak-to-peak separation (DEp)
indicative of reversible two ET (Fig. 3(a)). The data were very
interesting, as reversible dopamine oxidation had not been
achieved previously on a carbon electrode, without prior
activation of the surface.6
Further work using a similar unpurified CNT paste electrode
demonstrated a clear response for the electrochemical detection
of dissolved oxygen (Fig. 3(b)).49 Supporting density functional
theory calculations by the same group indicated that for the
specific case of oxygen reduction, pentagon and heptagon sites
(possible defect sites on a CNT) could act as favourable sites
for ET.49 By default, in these calculations, the CNTs were
assumed to be free from amorphous carbon and metal NPs,
which is not the case for an unpurified sample (as discussed in
Section 1.2).
In an attempt to simplify the experimental configuration
and aid the interpretation of the electrochemical response,
Campbell et al.50 were the first to perform experiments with a
single CNT (MWNT of 100 nm diameter), which was
electrically attached to an etched wire and dipped into
an aqueous electrolyte solution. Voltammetric measurements
of Ru(NH3)63+/2+ gave steady-state sigmoidal cyclic
voltammograms (CVs), with a limiting current that scaled
with immersion depth. However, transmission electron micro-
scopy (TEM) analysis showed that the sidewall of the CNT
was coated with amorphous carbon, so that essentially no
information was obtained on ET at the CNT surface.
A similar electrode configuration was reported later by
Downs et al.,51 but whiskers of loosely packed arc-discharge,
unpurified MWNTs were used as electrodes for the electro-
chemical oxidation of aniline. In these studies it was shown
that the polyaniline films on the CNT whiskers were oxidised
more readily than on conventional electrodes such as Pt.
Nugent et al.52 compared the response of uninsulated
bundles of unpurified MWNTs (grown by arc discharge),
shown in Fig. 3(c), to that of basal plane graphite (BPG)
and edge plane graphite (EPG), for the one-electron oxidation
of ferrocyanide to ferricyanide. Ferro/ferricyanide has been
used extensively in the electrochemical literature for ET
studies; however, it is becoming increasingly recognised that
this redox system is non-ideal due to issues with adsorption
and decomposition, especially on carbon surfaces.6,53 The
response at the MWNT bundles was Nernstian (Fig. 3(c)),
with an ideal peak-to-peak separation (DEp) of 59 mV,
which was significantly faster than that observed on BPG
(DEp E 300 mV) and EPG (DEp E 70 mV). The authors
found the reversible response of the MWNT electrode
puzzling, especially when compared to that seen at EPG. They
Chem. Commun., 2009, 6886–6901 | 6889
Fig. 3 Early developments in CNT electrochemistry. (a) CV of dopamine oxidation at a CNT paste electrode showing reversible behaviour;48
(b) CV of oxygen reduction at a CNT electrode compared to a carbon paste electrode. No images of the electrode assemblies are provided for these
studies;49 (c) SEM image of MWNT bundles electrode and CVs showing Nernstian response for ferrocyanide oxidation.52 (a) Reprinted with
permission from ref. 48. Copyright 1996 Elsevier Ltd. (b) Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission from
ref. 49. (c) Reprinted with permission from ref. 52. Copyright 2001 American Chemical Society.

assigned the increased ET rate to several possible factors
including CNT helicity, low dimensionality and the presence
of topological defects on the sidewalls of the CNTs. Interestingly,
no mention was made of possible contributions from impurities
in the CNT sample.
In summary, all of the studies above employed unpurified
MWNTs grown using arc discharge. Consequently, although
in many cases the results showed more reversible (fast) ET
characteristics than alternative carbon electrodes, they do not
provide a fundamental understanding of the intrinsic properties
of CNTs as an electrode material.
2.2 CNT modified electrodes: expansion of the field
CNT film electrodes were first introduced by Liu et al.,54 who
prepared thin-film electrodes by drop casting purified SWNTs
suspended in organic solvents onto underlying Pt and Au
electrodes. Using an essentially identical fabrication procedure,
Luo et al. subsequently reported the first use of purified
SWNTs drop cast onto a glassy carbon (GC) electrode, for
voltammetric analysis.55 Scanning electron microscopy (SEM)
showed the resultant SWNT-GC electrode to have an uneven
distribution, as shown in Fig. 4(a), with many SWNTs
aggregating in large bundles and SEM evidence of amorphous
carbon. However, dopamine, epinephrine, and ascorbic acid
showed a favourable voltammetric response at the SWNT-
modified electrode, giving higher currents and improved peak-
to-peak potentials than the bare GC electrode (Fig. 4(a)).
Following this report, the field expanded greatly with
hundreds of papers on the electrocatalytic properties of
CNT-modified electrodes. With variations on a theme, the
basic electrode configuration of randomly dispersed, purified
CNTs on the surface of a conducting support macroelectrode,
usually GC or BPG, was used extensively for the electrochemical
detection of a multitude of redox-active molecules in solution.
Noteworthy are early reports of the low-potential detection of
NADH,56 norepinephrine,57 cytochrome c58 and the simultaneous
detection of dopamine and ascorbic acid,59 as shown in
Fig. 4(b). Variations of this method included dispersing the
CNTs with different binders (N,N-dimethylformamide,60
dihexadecyl hydrogen phosphate,61 Nafion62), forming a
CNT composite,63,64 screen printing,65,66 drop coating without
binders,59,67 employing a nanotube paper as the electrode
(using filtration techniques)68 and abrasion attachment.69,70
Much of this work is described in reviews by Wang,10 Gooding71
and Banks and Compton.15 The generally-claimed benefits of

6890 | Chem. Commun., 2009, 6886–6901 This journal is

c The Royal Society of Chemistry 2009
Fig. 4 (a) SEM image of the first SWNT-modified electrode prepared by drop-casting a suspension of SWNTs onto GC and CVs showing
favourable response of the SWNT-modified electrode for dopamine oxidation, compared to the bare GC support;55 (b) SEM images of electrode
surfaces and CVs of norepinephrine,57 cytochrome c58 and dopamine with ascorbic acid59 at CNT-modified electrodes showing enhanced responses
when compared to the bare support. (a) Reprinted with permission from ref. 55. Copyright 2001 American Chemical Society. (b) Norepinephrine:
Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission from ref. 57. Cytochrome c: reprinted with permission from
ref. 58. Copyright 2002 American Chemical Society. Ascorbic acid: From ref. 59 (Copyright 2002). Reproduced by permission of The Royal Society
of Chemistry.

these CNT-modified electrodes were reduced overpotentials
and increased peak currents (more facile ET kinetics) providing
higher sensitivity and selectivity compared to the bare support
electrode. In many cases, the electrode material was little
characterised, thus the fundamental reasons for the apparent
improvements in electrode performance still remained elusive.
It is worth noting that a recent paper has suggested that for the
case of electrodes composed of multiple layers of MWNTs, the
observed electrochemical responses could be due to thin layer
diffusion within the porous layer and not electrocatalytic
effects.72
Other studies reported that for the electrochemical detection
of NADH, epinephrine, and norepinephrine, graphite powder-
modified BPG electrodes (prepared by abrasively attaching
graphite or pipetting a graphite solution onto the BPG
electrode) displayed similar enhanced currents and reduced
peak-to-peak separations, to MWNT electrodes (used as
received at 495% purity) prepared in the same way.69 These
authors consequently advised caution in assigning any unique
electrocatalytic properties to CNTs alone.
In cases where CNTs have been purified or chemically
treated, it is also important to consider any possible effects
of the chemical processing procedure on the resulting material,
when preparing CNT-modified electrodes. For example, acid
purification cuts, shortens and acid functionalises CNTs,44,73
whilst the use of a casting solvent, to prevent clumping of
CNTs on the support electrode surface, involves solubilisation
adjuvants, such as polymers, surfactants, proteins or harsh
sonication treatments.74 These solubilisation methods can
result in coating of the CNT surface which can be difficult to

This journal is

c The Royal Society of Chemistry 2009 Chem. Commun., 2009, 6886–6901 | 6891
remove, or in further attack of the CNT structure (in the
case of sonication). Combined, these factors can affect
the performance of composite CNT electrodes. Understanding
the influence of these treatments is clearly important when
assessing the electrochemical properties of CNTs.
3. Fundamental electrochemical properties of
CNTs
The previous section highlighted early work and the considerable
promise of CNT electrodes, particularly for electroanalysis.
However, this work also raised questions as to the intrinsic
activity of CNTs. This section discusses some of the key issues,
progress made to date, as well as some of the questions that
remain and need to be addressed.
3.1 Role of nanoparticles
The overwhelming majority of CNT studies have utilised
commercially sourced material, grown using HiPCO, arc
discharge or laser ablation. This material can be obtained
with different grades of purity, or can be purified by the user
prior to application. Several groups have shown that even
after the extensive use of established purification procedures,
CNTs may still contain significant amounts of metallic
impurities,75–77 encapsulated within graphite sheets or located
outside the tubes (Fig. 5(a,b)). This has led to the conclusion
that the electrochemical response of particular redox systems
could be greatly enhanced by the presence of residual metallic
NPs77–81 and that analysis of the CNTs82,83 before and after
purification is essential.76
It has been suggested that residual iron NPs in MWNTs are
responsible for the ‘‘electrocatalytic’’ oxidation of hydrazine
(Fig. 5(c)) and reduction of hydrogen peroxide on MWNT-
modified electrodes.78,80 In these studies, the MWNTs were
postulated to act merely as electrical ‘‘wires’’ (showing no
intrinsic electrochemical activity), connecting the active iron
oxide NPs.78,80 Copper impurities, contained as traces in the
iron catalyst NPs, have also been associated with the observed
electrochemical response of MWNTs toward the reduction of
halothane.79 Electrocatalysis of hydrazine oxidation and
hydrogen peroxide reduction has also been observed at
‘‘purified HiPCO’’ SWNT-modified electrodes.84 TEM images
of the as-purchased purified SWNTs revealed the presence
of a high density of metal NPs, trapped in graphite layers,

Fig. 5 (a) SEM image of as-purchased ‘‘purified HiPCO’’ SWNTs showing a high amount of catalyst NPs present. XPS data indicate that even
super-washing (reflux in HNO3 for 36 hours) does not eliminate all iron present;76 (b) TEM images of catalyst NPs sheathed by several graphene
sheets even after acid cleaning;77 (c) CVs of hydrazine oxidation at BPPG, EPPG, Fe-modified BPPG and MWNT-modified BPPG electrodes;78
(d) CVs for O2 reduction at N-CNTs before and after passivation of iron NPs.85 (a) From ref. 76 (Copyright 2007). Reproduced by permission of
The Royal Society of Chemistry. (b) Reprinted with permission from ref. 77. Copyright 2007 American Chemical Society. (c) Copyright
Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission from ref. 78. (d) Reprinted with permission from ref. 85. Copyright 2007
American Chemical Society.

6892 | Chem. Commun., 2009, 6886–6901 This journal is

c The Royal Society of Chemistry 2009
distributed amongst bundles or located on the outside of
the SWNTs. X-Ray photoelectron spectroscopy (XPS) data
indicated that a significant amount of iron remained even after
further acid wash (20 hours or 36 hours in 2 M HNO3), as
shown in Fig. 5(a).
Lyon et al. demonstrated that iron NPs encapsulated within
graphene layers of CNTs, grown using a floating vapour
deposition process, display potential-dependent redox activity
that varies with supporting electrolyte composition.85 In
neutral solutions of dibasic sodium phosphate, acetate and
citrate, Fe2+ dissolution and Fe2+/3+ redox activity were
observed, but no dissolution was seen in potassium nitrate,
due to the formation of oxide layers. In addition, Fe2+/3+
electrochemistry could be completely suppressed by passivating
the iron NPs through repeated cycling in supporting electrolyte.
The passivation process did not affect oxygen reduction at
nitrogen doped CNTs (N-CNTs), prepared by drop casting
CNTs onto a GC electrode, suggesting that iron impurities are
not required for enhanced oxygen reduction at N-CNTs
(Fig. 5(d)).85
It has recently been shown that the issue of catalyst NPs is
more complex when multi-component metallic impurities are
present.81,86 Only iron-based impurities were found to be
responsible for the reduction of hydrogen peroxide at acid-
purified double-walled CNTs containing residual cobalt
catalyst NPs, which themselves contained iron impurities.86
On the other hand, all metallic components of catalyst NPs
were found to participate in the electrochemical oxidation of
hydrazine.81 While these studies evidence contributions
from metal nanoparticles in the voltammetric response of
CNTs, they do not allow absolute assessment of the relative
contributions of NPs and the CNTs themselves in the electro-
chemical process. Furthermore, chemical treatments to
remove metal NPs may impact the response of the CNTs
and this is difficult to quantify from gross CV measurements.
3.2 Role of defects: electrochemical cyclic voltammetric
studies
Much of the work on understanding CNT electroactivity
draws on the premise that the sidewall of the CNT is largely
inactive and that edge-plane-like graphitic defect sites (i.e. open
ends and locations on the tube axis where graphitic sheets
terminate) are responsible for ET.23–25 This analysis largely
originates from analogies drawn between MWNTs and
HOPG.24,69 The question then arises as to whether assumptions
made about the electrochemical behaviour of MWNTs can be
extrapolated to SWNTs? Some researchers believe this to be
the case,22,25,26,87 although others have been careful not to
make direct comparisons between SWNTs and MWNTs. This
is an important point because studies on cCVD grown SWNTs
suggest that the sidewall may have much more electrochemical
activity27,28 than previously considered22,25 and that both
the experimental configuration and the chemical state of the
SWNT, in relation to the redox couple investigated, must be
taken into account. A further point of interest is the nature of
defects in CNTs. For example, are edge-plane-like defects the
only important defect sites when considering electrochemical
activity, or do other sites (including the pristine sidewall) also
This journal is
c The Royal Society of Chemistry 2009
contribute to the overall response? This aspect is discussed
further in Section 3.4.
Li et al. examined the ferro/ferricyanide couple at electrodes
comprising (i) SWNT paper, prepared by laser ablation, acid
purified and then collected by filtration; and (ii) cCVD grown
MWNT towers consisting of randomly curved CNTs growing
upwards from the substrate, with and without a post-growth
thermal anneal.68 In CV measurements, the SWNT paper
yielded a DEp value of 96 mV, smaller than observed at
MWNT towers (170 mV). Heat treating the towers prior to
use resulted in a DEp value of 230 mV. The authors suggested
that heat treatment resulted in the elimination of any amorphous
carbon present on the sidewall of the MWNTs. This, they
postulated, led to slower ET kinetics, as the basal plane-like
sidewalls were revealed, although no direct microscopic
evidence was presented for the removal of amorphous carbon.
It was also suggested that an abundance of graphitic edge-
plane-like defects on the SWNT paper electrodes, created by
prolonged acid purification, was responsible for the fastest ET
kinetics on the SWNT papers.
Banks et al. compared the voltammetric response of the
ferri/ferrocyanide couple at EPG and BPG, as well as at
MWNT-modified BPG (495% purity), as shown in Fig. 6(a).24
BPG showed slow ET kinetics, with DEp E 350 mV, which
was interpreted as ET occurring exclusively at a sparse density
of steps on the BPG with the basal plane itself essentially
inert. In contrast, EPG gave DEp of 78 mV and the
MWNT-modified BPG electrode showed ideal reversible
behaviour (Fig. 6(a)). Making the assumption that the sidewall
of the MWNT is similar to the basal plane of graphite, the
authors concluded that the electrocatalytic nature of the
MWNTs must reside solely in ET from the presence of
edge-plane-like sites on CNTs. Pumera et al.88 found that
extensive electrochemical oxidation of multi-layer MWNT
electrodes increased both the density of edge plane defects
(TEM analysis) and oxygen-containing functionalities (high
resolution XPS). The resulting rate constant of ET for the
ferrocyanide/ferricyanide couple was also found to increase,
by over two orders of magnitude, compared to that at
as-received MWNTs.
Liu et al. found that ET for ferrocyanide oxidation was
more facile at vertically aligned SWNTs (DEp 59 mV)
compared to randomly dispersed 2D (termed bed-spread),
acid-purified SWNTs (DEp 99 mV) on a gold surface, as shown
in Fig. 6(b).87 The vertically aligned SWNTs were prepared by
acid treating HiPCO prepared SWNTs, to purify, shorten (to
B120 nm) and form oxygen-containing moieties at the ends
of the tubes, due to acid cutting and sonication.87 The
carboxylate groups at the nanotube ends were then chemically
bound to a cysteamine-modified gold electrode at B16%
surface coverage (Fig. 6(b)). Working with the assumption
that ET on SWNTs is governed by open end sites, the
favourable electrochemical properties of the vertically aligned
CNTs were attributed to the increased surface concentration
of end groups on the electrode, compared to the bed-spread case.
It is evident that the majority of reports infer the contribution
of ‘‘defects’’ in the electrochemistry of CNT electrodes largely
from CV analysis. While defects may play an important role in
the electrochemistry of certain species, especially those that go
Chem. Commun., 2009, 6886–6901 | 6893
Fig. 6 (a) CVs of ferricyanide reduction at BPPG, EPPG and MWNT-modified BPPG electrodes24 and SEM image of a BPPG electrode
abrasively modified with MWNTs;69 (b) CVs of ferrocyanide oxidation at vertically aligned and horizontally-spread SWNT-modified gold
electrodes. AFM topography image and scan profile of the vertically aligned SWNT electrode.87 (a) BPPG electrodes: From ref. 24
(Copyright 2004). Reproduced by permission of The Royal Society of Chemistry. MWNTs: Reprinted with permission from ref. 69. Copyright
2004 American Chemical Society. (b) Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission from ref. 87.

via surface-mediated ‘‘inner-sphere’’ ET,6 there is still no direct
high resolution electrochemical evidence that a nanotube
sidewall is completely inactive, as has been claimed.24–26,69,89
The resolution of this issue is likely to require experimental
techniques that can visualise electrochemical activity at a local
scale and relate this to the underlying structural and electrical
properties of CNTs. Theory is also expected to play a major
role in addressing this problem.
3.3 Role of defects versus sidewalls in ET activity: studies
without the influence of a conducting support
To explore fully the interplay between defect and sidewall ET
activity, there is an urgent need for experimental methodologies
and techniques that go beyond CV analysis of CNT-modified
electrodes. The results of these approaches must also be
correlated with growth procedures, fabrication techniques
and characterisation methods, so that the composition of the
CNT electrode is very well defined.
A move in this direction has been an increasing number of
studies utilising cCVD-grown SWNTs on Si/SiO2 substrates.
There are multifold advantages to this approach. SWNTs
grown by this method are clean (with little or no amorphous
carbon), provided the correct growth conditions are employed,29
and this can be verified using high resolution structural
characterisation techniques such as atomic force microscopy
(AFM), scanning tunnelling microscopy (STM), TEM and
Raman analysis.90 The inherent defect density in SWNTs
grown in this way is low, typically one defect for every
100 nm to 4 mm length of SWNT.91,92 The incorporation of
residual catalytic metal impurities into the SWNTs is unavoid-
able, as with other CNT growth methods. However, the metal
content in cCVD-grown CNTs is significantly less than with
other growth methods, because each SWNT usually contains
only one NP which is typically encapsulated at one end of the
SWNT. Crucially, and in contrast to all of the electrodes
described above, the substrate on which the SWNTs are grown
is inert (insulating), so that any measured electrochemical
response comes solely from the SWNTs. Finally, the system
is amenable to many device configurations, including transistor
gate measurements which can provide key information on the
conductance of the SWNTs.93
The most compelling evidence that the SWNT sidewall is
electrochemically active comes from studies of individual
cCVD-grown SWNTs as nanoelectrodes for electrochemistry.27

6894 | Chem. Commun., 2009, 6886–6901 This journal is

c The Royal Society of Chemistry 2009
Fig. 7 (a) CV of the one electron FcTMA+ oxidation at an individual SWNT nano-electrode in the absence of background electrolyte; inset
represents an AFM image of the single SWNT electrode device;27 (b) CVs of dopamine oxidation before and after oxygen plasma treatment; insets
represent SCM images (3  3 mm) of a SWNT network before and after plasma treatment, showing a decrease in the intrinsic conductivity of
SWNTs manifested by a decrease in phase contrast;44 (c) schematic representation of the procedure for preparing the CNT electrodes with only the
nanotube tip (CNT-T) or sidewall (CNT-S) accessible to electrolyte and CVs for H2O2 and ascorbic acid oxidation at CNT-S and CNT-T
electrodes (with or without oxygen surface functionalities).94 (a) Reprinted with permission from ref. 27. Copyright 2005 American Chemical
Society. (b) Reprinted with permission from ref. 44. Copyright 2007 American Chemical Society. (c) Copyright Wiley-VCH Verlag GmbH & Co.
KGaA. Reproduced with permission from ref. 94.

An AFM image of the single SWNT electrode device is
depicted in Fig. 7(a) inset. With only the SWNT sidewall
(and no ends or catalyst NPs) exposed to solution, a steady-
state voltammetric current proportional to the exposed SWNT
length was observed for the oxidation of FcTMA+, as shown
in Fig. 7(a). This proves that tube ends or catalyst particles
do not need to be present for ET to occur for this redox
couple.
Dumitrescu et al. investigated the effect of chemical treatments
on the electrochemical and electrical properties of 2D connected
networks of cCVD-grown SWNTs on inert, insulating surfaces.
Exposing CNTs to particular chemical treatments results in
the creation of sidewall defects and open ends which become
chemically functionalised with oxygen-containing groups.
Wet-gate conductivity and high resolution scanning conductance
microscopy (SCM) measurements showed that the main effect
of exposing the SWNTs to acid reflux or an oxygen plasma
was a significant decrease in network conductivity (Fig. 7(b)
inset).44 Acid reflux selectively removed mSWNTs from a
sample and broke the connectivity of the SWNT network by
cutting and etching the SWNTs, while plasma functionalisation
increased the intrinsic resistivity, by disruption of the sp2
network due to sidewall addition of oxygen-containing
groups. Working under conditions where the current passed
was small enough not to be offset by the increase in electrode
resistance, the CV behaviour for the outer-sphere ET mediator,
Ru(NH3)63+, was no different whether the network had been
chemically treated or not. However, functionalisation through
defect creation made ET more facile for the oxidation of
dopamine (Fig. 7(b)),44 which is thought to undergo inner-
sphere ET and is considered to be promoted by adsorption at
oxygenated-surface functionalities.6

This journal is

c The Royal Society of Chemistry 2009 Chem. Commun., 2009, 6886–6901 | 6895
 Gong et al. also found that the electrochemistry of different
redox couples at CNT electrodes was also not necessarily
facilitated by oxygen-functionalised tube ends or sidewall
defects.94 Instead, the electrochemical response depended
strongly on the redox process under investigation. In these
studies, the electrochemistry of CNT tips and sidewalls was
probed separately using vertically aligned forests of super-long
(5 mm length) CNTs, grown by water-assisted cCVD. Oxygen-
containing groups on sidewalls and tip ends were introduced
by electrochemical treatments in acidic solution. Masking with
a non-conducting coating ensured that only the CNT sidewalls
or tips were exposed to solution (Fig. 7(c)). Various redox
processes were investigated; for example, H2O2 oxidation was
observable on sidewalls, whether the surface was oxygenated
or not, but not at tips. In contrast, ascorbic acid oxidation was
promoted by oxygen-containing functionalities, with little
difference between the tip and the sidewall. (Fig. 7(c)).
3.4 Significance of defects
Whichever synthesis method is employed, completely defect-
free CNTs, over the length-scales typically employed in
electrochemical experiments, are almost impossible to produce.
Although advances in synthesis methods claim defect-free
materials,29 evidence is usually based on macroscopic
characterisation methods, such as Raman spectroscopy, which
lacks the sensitivity to identify a single defect in an otherwise
pristine material.95
Inherent defects formed during the growth of SWNTs are
local impurities, Stone–Wales defects (rotated bonds), vacancies
(point defects) and sp3 hybridised defects (terminated by
chemical groups pertinent to the growth process, e.g. –H).21
Whilst similar types of defects may exist on the sidewalls of
MWNTs, it is the edge-plane-like graphitic sites, either at the
open ends of MWNTs or at locations along the tube axis
where graphite sheets terminate at the surface (of particular
importance in bamboo and herringbone-like CNTs), that are
typically referred to as active sites for ET in the literature.15
In the majority of cases, sidewall defects are thought
to reduce the conductance of a SWNT due to resonant
back-scattering.46 The presence of sp3-type carbon causes
drastic changes in the electronic states near the Fermi level
and the presence of many chemical groups (e.g. –COOH,
–NH2, –OH, –CH3, –H, –F) increases the resistivity of
SWNTs.45 Typical acid purification treatments also result in
defect creation. Interestingly, theoretical calculations show
that, in contrast to sidewall functionalised –COOH sites,
vacancy decorated –COOH sites can cause the electrical
conductivity of a sSWNT to increase.96
Identification of individual defects in CNTs requires high
resolution scanned probe techniques such as STM or scanning
gate microscopy (SGM).97 Using such techniques, Fan et al.
found that by applying a carefully-defined sequence of
electrochemical potentials to an individual SWNT grown by
cCVD and connected in an electrical device, it was possible to
nucleate selectively metal at defects (sites of highest local
energy), as shown in Fig. 8(a).91 Electrodeposition of nickel
on SWNTs at small overpotentials showed that only a few
metal dots nucleated and grew on SWNT sidewalls, with
6896 | Chem. Commun., 2009, 6886–6901
repeated cycles of stripping and deposition decorating
the same positions.91 This was assigned to the presence of
defects on the SWNT sidewalls and was confirmed with SGM
measurements, which identified regions of localised high
resistance which correlated precisely with the metal deposition
sites. The mean defect density for in situ cCVD-grown CNTs
was determined as one per 4 mm length of SWNT, from over
two hundred measurements. However, the large standard
deviation indicated that many SWNTs showed micron lengths
with no defects whilst others showed many defects clustered
together over a smaller length-scale. CNTs grown ex situ by
arc vaporisation, suspended in solvent and deposited onto a
wafer showed much higher defect densities, of about one per
100 nm.91
It is interesting to note that an overpotential of just 100 mV
was required to achieve uniform coverage of the entire SWNT
sidewall by nickel.91 By changing the applied potential, it was
possible to control systematically the density of NPs nucleated
on the SWNT surface, up to the limiting case of almost
contiguous wire formation. Contiguous electrochemical
decoration of SWNT sidewalls with silver, platinum and gold
has also been reported by Day et al.98 (Fig. 8(b)) and Quinn
et al.99 These studies again point to the sidewall of a SWNT
being able to support ET.
Dumitrescu et al. described a lithographic procedure for
producing ultramicroelectrodes (UMEs) from 2D networks of
pristine SWNTs, with o1% surface coverage (Fig. 8(c)). This
format is of considerable interest as it provides the possibility
of generating high mass-transfer rates and studying ET under
steady-state conditions.28 The steady-state behaviour at SWNT
network UMEs was as expected for conventional solid metal
UMEs of the same size due to diffusional overlap between
neighbouring SWNTs. Simple outer sphere redox couples
(FcTMA+/2+ and Ru(NH3)63+/2+) showed reversible behaviour
at the SWNT UMEs (Fig. 8(c)). Considering the high mass
transport rates to the SWNT network, a minimum value for the
standard ET rate of 3 cm s 1 was calculated for FcTMA+/2+,
assuming that the entire length of SWNT was active to ET.
3.5 Metallic versus semiconducting SWNTs
Electrochemical studies of individual cCVD-grown SWNTs,
in a format where measurement of the intrinsic electrical
properties was also possible, showed that both mSWNTs
and sSWNTs were similar for the oxidation of FcTMA+/2+
couple (Fig. 9(a)).27 However, this is not surprising given that,
in this potential range, sSWNTs are considerably p-doped
under electrolyte-gated conditions.8
Interestingly, studies of SWNTs grown using the same
approach but in a 2D network format indicated that the
electrochemical response of the entire network was sensitive
to the potential of the redox couple.100 The magnitude of the
current signal for the Fe(phen)32+/3+ couple was significantly
greater than that expected based on the current measured
for the Ru(NH3)63+/2+ couple (Fig. 9(b)). Simultaneously
recorded wet-gate measurements showed that the reduction
of Ru(NH3)63+ took place in the same potential region where
the network conductivity was smallest, likely due to the
depletion of charge carriers in the sSWNTs.100
This journal is
c The Royal Society of Chemistry 2009
Fig. 8 (a) SEM images showing selective nickel electrodeposition onto SWNTs at very low overpotentials;91 (b) SEM image of a SWNT network
electrodeposited with silver (top) and platinum (bottom), showing contiguous NP deposition on the nanotube sidewall;98 (c) SEM images of a
100 mm diameter SWNT UME device (top) and the o1% surface coverage SWNT network (bottom) and CVs of 0.025 mM to 1 mM FcTMA+
oxidation at the SWNT network UME.28 Note due to charging, the lateral dimensions of the SWNTs in these SEM images will always appear
significantly larger than they actually are, resulting in a false impression for the surface coverage. (a) Copyright 2005 Nature Publishing Group.
Reproduced with permission from ref. 91. (b) Reprinted with permission from ref. 98. Copyright 2005 American Chemical Society. (c) Reprinted
with permission from ref. 28. Copyright 2008 American Chemical Society.

Several authors have used metal electrodeposition to
elucidate differences in the electronic structure of SWNTs.
Work by Day et al. showed that only under defined potential
conditions could differences in metal nucleation patterns be
observed on 2D networks of SWNTs.101 Using small over-
potentials for the electrodeposition of palladium, preferential
nucleation sites became apparent from SEM analysis
(Fig. 9(c)); this was attributed to favoured deposition onto
mSWNTs, although it may also be indicative of SWNTs
with favourable defect densities. Huang et al. used cCVD
to grow super-long (up to cm length) parallel arrays of
SWNTs on an insulating surface and subjected them to silver
electrodeposition.102 Using Raman microscopy it was possible
to correlate the nucleation characteristics (observed using
SEM) with the electronic structure (Fig. 9(d)). It was found
that B32% of SWNTs showed no silver deposition and weak
Raman signal. They postulated that the defect density on these
SWNTs was high enough to make the SWNTs too resistive to
show appropriate nucleation behaviour. These SWNTs were
termed ‘‘quasi-insulators’’. Of the remaining SWNTs, only
B5% were identified by Raman as mSWNTs, with the
remaining B62% identified as sSWNTs. Under the deposition
potential conditions applied, no significant difference in the
nucleation characteristics between mSWNT and sSWNTs was
observed.102 Work by Zhang et al. also showed that no
difference could be observed for copper electrodeposition on
individual cCVD-grown sSWNTs and mSWNTs.103 However,
work by Qian et al. also using ultralong cCVD-grown SWNTs
(3–5 mm) showed that mSWNTs and sSWNTs (identified
electrically) did show different silver deposition behaviour,

This journal is
c The Royal Society of Chemistry 2009 Chem. Commun., 2009, 6886–6901 | 6897
Fig. 9 (a) CVs of FcTMA+ oxidation at three individual mSWNT and three individual sSWNT nano-electrodes for three different exposed
lengths of SWNT;27 (b) CVs of Ru(NH3)63+ reduction and Fe(phen)32+ oxidation at a gold and a SWNT network band electrode and
simultaneous conductivity measurements as a function of electrode potential;100 (c) SEM image of palladium electrodeposition onto a SWNT
network at low overpotentials, showing preferential deposition;101 (d) SEM image of Ag deposition onto two SWNTs and corresponding Raman
spectra from isolated individual SWNTs, taken at the indicated locations.102 (a) Reprinted with permission from ref. 27. Copyright 2005 American
Chemical Society. (b) Reprinted with permission from ref. 100. Copyright 2004 American Chemical Society. (c) Reprinted with permission from
ref. 101. Copyright 2007 American Chemical Society. (d) Reprinted with permission from ref. 102. Copyright 2009 American Chemical Society.

albeit subtle.104 In particular, mSWNTs showed a slightly
higher particle density than sSWNTs attributed to the lower
resistivity of the mSWNT. It was argued that such effects
should only be observable with longer SWNTs, where
potential drop effects along the length of the SWNT would
be exacerbated.
4. Recent developments and future perspectives
4.1 Electrochemical sensing of biomolecules
Carbon electrodes are a key component of electrochemical
biosensors;6 however, conventional biosensors based on screen
printed carbon or GC electrodes have several shortcomings,
including poor sensitivity and stability, large response times
(resulting from the large capacitance) and slow ET rates.
CNTs, in principle, could overcome many of these issues,
not withstanding the concern over practical applications of
CNTs due to possible toxicity issues.105–108 Indeed, as part of
ongoing work recent studies show that systemic administration
of CNTs does not have any deleterious effect on mammalian
health.109
Thus far, in vitro electrochemical detection of biomolecules
has proved extremely successful, with CNTs being used to
detect NAD+, nitric oxide, cytochrome c, morphine, cysteine,
glucose, indole acetic acid, cholesterol110,111 and DNA.112 In
the latter example, arrays of micro-disk electrodes, comprising
bundles of vertically aligned cCVD-grown MWNTs embedded
in SiO2, with the ends selectively functionalised with oligo-
nucleotides, yielded a detection limit of a few attomoles.112
The array format resulted in the minimisation of background
and non-faradaic processes and a much reduced electrode
response time. Recently, Liu et al. refined the fabrication
technique using an epoxy insulation.113 Similarly, Bertoncello
et al. showed that cCVD-grown SWNT network electrodes of
low surface coverage (o1%) on an insulating surface gave
unprecedented low background currents, enabling trace level
CV detection of FcTMA+ at levels below 25 nM and dopamine
below 100 nM (Fig. 10(a)). The latter represents the highest
CV detection sensitivity for dopamine at an untreated,

6898 | Chem. Commun., 2009, 6886–6901 This journal is

c The Royal Society of Chemistry 2009
Fig. 10 (a) SEM image of a o1% surface coverage SWNT network and CVs of FcTMA+ and dopamine oxidation at trace level
concentrations;114 (b) SEM image of a B100% surface coverage SWNT network, repeated CVs for the oxidation of dopamine in PBS, showing
high electrode stability and CVs for the oxidation of dopamine in 4% (w/w) albumin, in a concentration range of 1 mm to 1 mM.116 (a) Reprinted
with permission from ref. 114. Copyright 2007 American Chemical Society. (b) Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced
with permission from ref. 116.

unmodified carbon-based electrode material.114 This was due
to the inherently low capacitance values associated with
pristine SWNTs and the low surface coverage.
Gabay et al. used thick (10–100 mm) cCVD-grown CNT
mats as conducting adhesion layers for outer layer neuronal
cells. The cells spontaneously migrated and adhered to the
electrode surface and showed low-noise extracellular signal
recordings due to the improved cell–electrode coupling.115
However, the 3D nature of the CNTs gave these materials
very large specific capacitance values (up to 10 F cm 3),115
making them highly unsuitable for low concentration electro-
chemical detection. Recently, Dumitrescu et al. showed that
by carefully tuning cCVD conditions, it was possible to
produce ultrathin CNT mats of near complete surface coverage
that still retained remarkably low capacitance, but were not
susceptible to electrode fouling in in vivo mimics.116 These
electrodes were used successfully for the electrochemical
detection of dopamine at mM concentrations, in albumin-
containing solutions, with no deterioration of the signal from
electrode fouling even after extensive use, as shown in
Fig. 10(b). The electrodes were shown to outperform other
carbon materials currently available.116
An exciting new area of application is the use of CNT
coated electrodes for studying brain function. Keefer et al.
showed that conventional tungsten and stainless steel wire
electrodes can be coated with CNTs to enhance their
performance.117 The CNT coating improved both recording
and electrical stimulation of neurons in culture and in vivo, by
decreasing the electrode impedance and increasing charge
transfer. These hybrid electrodes could be used in ongoing
projects in the field of brain–machine interfaces, such as the
restoration of movement in tetraplegia.118
5. Conclusions and suggestions
The field of CNT electrochemistry is large and growing, with
many favourable electrochemical properties reported for CNT
electrodes, such as enhanced detection sensitivity, electrocatalytic
effects and reduced fouling. However, as pointed out in this
review, it is clear that uncertainties still remain regarding the
origin of the electrochemical signal. For example, much of the
CNT electrochemistry literature is based on the assumption
that the sidewalls of a CNT are inert and that only the open
ends of CNTs and edge-plane-like defect sites on the CNT
sidewall are responsible for the ET activity observed.23,24 This
body of work is largely based on MWNT electrodes
although similar claims have been for SWNT electrodes.22,25
The latter assertions appear to conflict with studies of metal
electrodeposition91,98,99 and the reversible electrochemistry
observed at SWNT UMEs.27,28
Concerns have also been raised about the purity of CNTs
employed in electrochemical studies, with evidence presented
that catalytic metal NPs (required for CNT growth) are
responsible for some observed electrochemical responses.77,78,80,84
In order to elucidate the intrinsic electroactivity of CNTs, it is
thus essential that measurements are made using CNTs which
have been thoroughly characterised, particularly for impurity
content. Additionally, the effect of any chemical processing
steps on the CNT structure and properties must also be
accounted for.
Recent studies have demonstrated that for individual
pristine SWNTs, open ends and catalytic NPs are not required
for fast ET of simple outer sphere redox couples.27 To
determine the relative contributions of different types of surface
sites, at the range of CNTs available, will require the use of

This journal is

c The Royal Society of Chemistry 2009 Chem. Commun., 2009, 6886–6901 | 6899
ultra-high resolution electrochemical imaging, to interrogate
the electrochemical activity of single defect sites or indeed
individual catalytic NPs on a well-defined system. 31
To summarise, the literature on CNT electrochemistry is
complex in several aspects. In part, this situation has arisen 32
due to the variable source and quality of the CNTs, the
33
chemical processing steps employed, incomplete characterisation
and the nature of the experimental arrangements adopted. 34
This review advocates a fresh look at CNT electrochemistry
35
and a thorough appraisal of all of the evidence in the literature.
As highlighted herein, progress towards a fundamental under-
standing of CNT electrochemistry is likely to be realised 36
through the use of ultra-high resolution electrochemical
37
imaging techniques in parallel with theoretical developments
of ET at sidewalls and the different types of defects found in
the various types of CNTs. 38
39
Notes and references
40
1 R. Saito, G. Dresselhaus and M. S. Dresselhaus, Physical Properties
of Carbon Nanotubes, Imperial College Press, London, 1998. 41
2 ISI Web of KnowledgeSM (www.isiknowledge.com) results for the
term ‘‘carbon nanotubes’’ in the publication title.
3 J. Giles, Nature, 2006, 441, 265. 42
4 C. Dekker, Phys. Today, 1999, 52, 22–28.
5 R. H. Baughman, A. A. Zakhidov and W. A. de Heer, Science, 43
2002, 297, 787–792.
6 R. L. McCreery, Chem. Rev., 2008, 108, 2646–2687.
7 R. Saito, M. Fujita, G. Dresselhaus and M. S. Dresselhaus, Appl. 44
Phys. Lett., 1992, 60, 2204–2206.
8 S. Rosenblatt, Y. Yaish, J. Park, J. Gore, V. Sazonova and 45
P. L. McEuen, Nano Lett., 2002, 2, 869–872.
9 I. Heller, J. Kong, K. A. Williams, C. Dekker and S. G. Lemay, 46
J. Am. Chem. Soc., 2006, 128, 7353–7359.
10 J. Wang, Electroanalysis (N. Y.), 2005, 17, 7–14. 47
11 R. N. Adams, Anal. Chem., 1958, 30, 1576–1576. 48
12 M. D. Hawley, S. V. Tatawawadi, S. Piekarski and R. N. Adams,
J. Am. Chem. Soc., 1967, 89, 447–450. 49
13 S. Iijima and T. Ichihashi, Nature, 1993, 363, 603–605.
14 S. Iijima, Nature, 1991, 354, 56–58. 50
15 C. E. Banks and R. G. Compton, Analyst, 2006, 131, 15–21.
16 L. X. Zheng, M. J. O’Connell, S. K. Doorn, X. Z. Liao, 51
Y. H. Zhao, E. A. Akhadov, M. A. Hoffbauer, B. J. Roop,
Q. X. Jia, R. C. Dye, D. E. Peterson, S. M. Huang, J. Liu and 52
Y. T. Zhu, Nat. Mater., 2004, 3, 673–676.
17 A. K. Geim and K. S. Novoselov, Nat. Mater., 2007, 6, 183–191. 53
18 A. Javey, J. Guo, Q. Wang, M. Lundstrom and H. Dai, Nature,
2003, 424, 654–657. 54
19 C. T. White and T. N. Todorov, Nature, 1998, 393, 240–242.
20 T. W. Odom, J.-L. Huang, P. Kim and C. M. Lieber, J. Phys. 55
Chem. B, 2000, 104, 2794–2809.
21 A. Hirsch, Angew. Chem., Int. Ed., 2002, 41, 1853–1859. 56
22 M. Pumera, Chem.–Eur. J., 2009, 15, 4970–4978.
23 C. E. Banks, T. J. Davies, G. G. Wildgoose and R. G. Compton, 57
Chem. Commun., 2005, 829–841.
24 C. E. Banks, R. R. Moore, T. J. Davies and R. G. Compton, 58
Chem. Commun., 2004, 1804–1805.
25 A. Chou, T. Bocking, N. K. Singh and J. J. Gooding, Chem. 59
Commun., 2005, 842–844.
26 A. F. Holloway, K. Toghill, G. G. Wildgoose, R. G. Compton, 60
M. A. H. Ward, G. Tobias, S. A. Llewellyn, B. N. Ballesteros,
M. L. H. Green and A. Crossley, J. Phys. Chem. C, 2008, 112, 61
10389–10397.
27 I. Heller, J. Kong, H. A. Heering, K. A. Williams, S. G. Lemay 62
and C. Dekker, Nano Lett., 2005, 5, 137–142.
28 I. Dumitrescu, P. R. Unwin, N. R. Wilson and J. V. Macpherson, 63
Anal. Chem., 2008, 80, 3598–3605. 64
29 H. Dai, Surf. Sci., 2002, 500, 218–241.
30 A. G. Rinzler, J. Liu, H. Dai, P. Nikolaev, C. B. Huffman, 65
F. J. Rodriguez-Macias, P. J. Boul, A. H. Lu, D. Heymann,
6900 | Chem. Commun., 2009, 6886–6901
D. T. Colbert, R. S. Lee, J. E. Fischer, A. M. Rao, P. C. Eklund
and R. E. Smalley, Appl. Phys. A: Mater. Sci. Process., 1998, 67,
29–37.
H. Hiura, T. W. Ebbesen and K. Tanigaki, Adv. Mater., 1995, 7,
275–276.
M. J. Bronikowski, P. A. Willis, D. T. Colbert, K. A. Smith and
R. E. Smalley, J. Vac. Sci. Technol., A, 2001, 19, 1800–1805.
B. Liu, W. Ren, L. Gao, S. Li, S. Pei, C. Liu, C. Jiang and
H.-M. Cheng, J. Am. Chem. Soc., 2009, 131, 2082–2083.
S. Huang, Q. Cai, J. Chen, Y. Qian and L. Zhang, J. Am. Chem.
Soc., 2009, 131, 2094–2095.
J.-F. Colomer, C. Stephan, S. Lefrant, G. Van Tendeloo,
I. Willems, Z. Kónya, A. Fonseca, C. Laurent and J. B. Nagy,
Chem. Phys. Lett., 2000, 317, 83–89.
T. W. Ebbesen, P. M. Ajayan, H. Hiura and K. Tanigaki, Nature,
1994, 367, 519.
Y. Li, X. Zhang, J. Luo, W. Huang, J. Cheng, Z. Luo, T. Li,
F. Liu, G. Xu, X. Ke, L. Li and H. J. Geise, Nanotechnology,
2004, 15, 1645–1649.
J.-M. Moon, K. H. An, Y. H. Lee, Y. S. Park, D. J. Bae and
G.-S. Park, J. Phys. Chem. B, 2001, 105, 5677–5681.
K. R. Moonoosawmy and P. Kruse, J. Am. Chem. Soc., 2008,
130, 13417–13424.
H. Hu, P. Bhowmik, B. Zhao, M. A. Hamon, M. E. Itkis and
R. C. Haddon, Chem. Phys. Lett., 2001, 345, 25–28.
M. A. Hamon, H. Hu, P. Bhowmik, S. Niyogi, B. Zhao,
M. E. Itkis and R. C. Haddon, Chem. Phys. Lett., 2001, 347,
8–12.
D. B. Mawhinney, V. Naumenko, A. Kuznetsova, J. T. Yates,
J. Liu and R. E. Smalley, Chem. Phys. Lett., 2000, 324, 213–216.
A. Kuznetsova, I. Popova, John T. Yates, M. J. Bronikowski,
C. B. Huffman, J. Liu, R. E. Smalley, H. H. Hwu and J. G. Chen,
J. Am. Chem. Soc., 2001, 123, 10699–10704.
I. Dumitrescu, N. R. Wilson and J. V. Macpherson,
J. Phys.Chem. C, 2007, 111, 12944–12953.
J. Zhao, H. Park, J. Han and J. P. Lu, J. Phys. Chem. B, 2004,
108, 4227–4230.
H. J. Choi, J. Ihm, S. G. Louie and M. L. Cohen, Phys. Rev. Lett.,
2000, 84, 2917–2920.
H. Park, J. Zhao and J. P. Lu, Nano Lett., 2006, 6, 916–919.
P. J. Britto, K. S. V. Santhanam and P. M. Ajayan, Bioelectrochem.
Bioenerg., 1996, 41, 121–125.
P. J. Britto, K. S. V. Santhanam, A. Rubio, J. A. Alonso and
P. M. Ajayan, Adv. Mater., 1999, 11, 154–157.
J. K. Campbell, L. Sun and R. M. Crooks, J. Am. Chem. Soc.,
1999, 121, 3779–3780.
C. Downs, J. Nugent, P. M. Ajayan, D. J. Duquette and
K. S. V. Santhanam, Adv. Mater., 1999, 11, 1028–1031.
J. M. Nugent, K. S. V. Santhanam, A. Rubio and P. M. Ajayan,
Nano Lett., 2001, 1, 87–91.
A. Heras, A. Colina, V. Ruiz and J. López-Palacios, Electro-
analysis (N. Y.), 2003, 15, 702–708.
C.-Y. Liu, A. J. Bard, F. Wudl, I. Weitz and J. R. Heath,
Electrochem. Solid-State Lett., 1999, 2, 577–578.
H. Luo, Z. Shi, N. Li, Z. Gu and Q. Zhuang, Anal. Chem., 2001,
73, 915–920.
M. Musameh, J. Wang, A. Merkoci and Y. Lin, Electrochem.
Commun., 2002, 4, 743–746.
J. Wang, M. Li, Z. Shi, N. Li and Z. Gu, Electroanalysis (N. Y.),
2002, 14, 225–230.
J. Wang, M. Li, Z. Shi, N. Li and Z. Gu, Anal. Chem., 2002, 74,
1993–1997.
Z. Wang, J. Liu, Q. Liang, Y. Wang and G. Luo, Analyst, 2002,
127, 653–658.
J. Wang, M. Li, Z. Shi, N. Li and Z. Gu, Microchem. J., 2002, 73,
325–333.
K. Wu, S. Hu, J. Fei and W. Bai, Anal. Chim. Acta, 2003, 489,
215–221.
J. Wang, M. Musameh and Y. Lin, J. Am. Chem. Soc., 2003, 125,
2408–2409.
J. Wang and M. Musameh, Anal. Chem., 2003, 75, 2075–2079.
F. Valentini, A. Amine, S. Orlanducci, M. L. Terranova and
G. Palleschi, Anal. Chem., 2003, 75, 5413–5421.
P. Fanjul-Bolado, P. Queipo, P. J. Lamas-Ardisana and
A. Costa-Garcı́a, Talanta, 2007, 74, 427–433.
This journal is
c The Royal Society of Chemistry 2009
 66 Y. Ye and H. Ju, Biosens. Bioelectron., 2005, 21, 735–741.
67 A. Guiseppi-Elie, C. Lei and R. H. Baughman, Nanotechnology,
2002, 13, 559–564.
68 J. Li, A. Cassell, L. Delzeit, J. Han and M. Meyyappan, J. Phys.
Chem. B, 2002, 106, 9299–9305.
69 R. R. Moore, C. E. Banks and R. G. Compton, Anal. Chem.,
2004, 76, 2677–2682.
70 C. E. Banks and R. G. Compton, Analyst, 2005, 130, 1232–1239.
71 J. J. Gooding, Electrochim. Acta, 2005, 50, 3049–3060.
72 I. Streeter, G. G. Wildgoose, L. Shao and R. G. Compton, Sens.
Actuators, B, 2008, 133, 462–466.
73 J. Liu, A. G. Rinzler, H. Dai, J. H. Hafner, R. K. Bradley,
P. J. Boul, A. Lu, T. Iverson, K. Shelimov, C. B. Huffman,
F. Rodriguez-Macias, Y.-S. Shon, T. R. Lee, D. T. Colbert and
R. E. Smalley, Science, 1998, 280, 1253–1256.
74 J. Chen, A. M. Rao, S. Lyuksyutov, M. E. Itkis, M. A. Hamon,
H. Hu, R. W. Cohn, P. C. Eklund, D. T. Colbert, R. E. Smalley
and R. C. Haddon, J. Phys. Chem. B, 2001, 105, 2525–2528.
75 N. S. Lawrence, R. P. Deo and J. Wang, Electroanalysis (N. Y.),
2005, 17, 65–72.
76 K. Jurkschat, X. Ji, A. Crossley, R. G. Compton and C. E. Banks,
Analyst, 2007, 132, 21–23.
77 M. Pumera, Langmuir, 2007, 23, 6453–6458.
78 C. E. Banks, A. Crossley, C. Salter, S. J. Wilkins and
R. G. Compton, Angew. Chem., Int. Ed., 2006, 45, 2533–2537.
79 X. Dai, G. G. Wildgoose and R. G. Compton, Analyst, 2006, 131,
901–906.
80 B. Sljukic, C. E. Banks and R. G. Compton, Nano Lett., 2006, 6,
1556–1558.
81 M. Pumera and H. Iwai, J. Phys. Chem. C, 2009, 113,
4401–4405.
82 T. Kolodiazhnyi and M. Pumera, Small, 2008, 4, 1476–1484.
83 C. Ge, F. Lao, W. Li, Y. Li, C. Chen, Y. Qiu, X. Mao, B. Li,
Z. Chai and Y. Zhao, Anal. Chem., 2008, 80, 9426–9434.
84 J. Kruusma, N. Mould, K. Jurkschat, A. Crossley and
C. E. Banks, Electrochem. Commun., 2007, 9, 2330–2333.
85 J. L. Lyon and K. J. Stevenson, Langmuir, 2007, 23, 11311–11318.
86 M. Pumera and H. Iwai, Chem. Asian J., 2009, 4, 554–560.
87 J. Liu, A. Chou, W. Rahmat, M. N. Paddon-Row and
J. J. Gooding, Electroanalysis (N. Y.), 2005, 17, 38–46.
88 M. Pumera, T. Sasaki and H. Iwai, Chem. Asian J., 2008, 3,
2046–2055.
89 J. J. Gooding, A. Chou, J. Liu, D. Losic, J. G. Shapter and
D. B. Hibbert, Electrochem. Commun., 2007, 9, 1677–1683.
90 H. Dai, Acc. Chem. Res., 2002, 35, 1035–1044.
91 Y. Fan, B. R. Goldsmith and P. G. Collins, Nat. Mater., 2005, 4,
906–911.
92 N. R. Wilson and D. H. Cobden, Nano Lett., 2008, 8, 2161–2165.
93 P. L. McEuen, Nature, 1998, 393, 15–17.
94 K. Gong, S. Chakrabarti and L. Dai, Angew. Chem., Int. Ed.,
2008, 47, 5446–5450.
This journal is
c The Royal Society of Chemistry 2009
95 S. K. Doorn, L. Zheng, M. J. O’Connell, Y. Zhu, S. Huang and
J. Liu, J. Phys. Chem. B, 2005, 109, 3751–3758.
96 C. Wang, G. Zhou, J. Wu, B.-L. Gu and W. Duan, Appl. Phys.
Lett., 2006, 89, 173130.
97 P. G. Collins, Defects and disorder in carbon nanotubes, in Oxford
Handbook of Nanoscience and Technology: Frontiers and Advances,
ed. A. V. Narlikar and Y. Y. Fu, Oxford University Press,
Oxford, 2009.
98 T. M. Day, P. R. Unwin, N. R. Wilson and J. V. Macpherson,
J. Am. Chem. Soc., 2005, 127, 10639–10647.
99 B. M. Quinn, C. Dekker and S. G. Lemay, J. Am. Chem. Soc.,
2005, 127, 6146–6147.
100 T. M. Day, N. R. Wilson and J. V. Macpherson, J. Am. Chem.
Soc., 2004, 126, 16724–16725.
101 T. M. Day, P. R. Unwin and J. V. Macpherson, Nano Lett., 2007,
7, 51–57.
102 Y. Qian, S. Huang, F. Gao, Q. Cai, L. Zhang and W. Hu, J. Phys.
Chem. C, 2009, 113, 6983–6988.
103 M. Zhang, Z. Zhou, X. Yang and X. Ye, Electrochem. Commun.,
2008, 10, 1559–1562.
104 P. Qian, Z. Wu, P. Diao, G. Zhang, J. Zhang and Z. Liu, J. Phys.
Chem. C, 2008, 112, 13346–13348.
105 K. Donaldson, R. Aitken, L. Tran, V. Stone, R. Duffin,
G. Forrest and A. Alexander, Toxicol. Sci., 2006, 92, 5–22.
106 J. M. Worle-Knirsch, K. Pulskamp and H. F. Krug, Nano Lett.,
2006, 6, 1261–1268.
107 A. Shvedova, V. Castranova, E. Kisin, D. Schwegler-Berry,
A. Murray, V. Gandelsman, A. Maynard and P. Baron,
J. Toxicol. Environ. Health, 2003, 66, 1909–1926.
108 S. K. Manna, S. Sarkar, J. Barr, K. Wise, E. V. Barrera,
O. Jejelowo, A. C. Rice-Ficht and G. T. Ramesh, Nano Lett.,
2005, 5, 1676–1684.
109 Z. Liu, C. Davis, W. Cai, L. He, X. Chen and H. Dai, Proc. Natl.
Acad. Sci. U. S. A., 2008, 105, 1410–1415.
110 K. Balasubramanian and M. Burghard, Anal. Bioanal. Chem.,
2006, 385, 452–468.
111 L. Agüı́, P. Yáñez-Sedeño and J. M. Pingarrón, Anal. Chim. Acta,
2008, 622, 11–47.
112 J. Li, H. T. Ng, A. Cassell, W. Fan, H. Chen, Q. Ye, J. Koehne,
J. Han and M. Meyyappan, Nano Lett., 2003, 3, 597–602.
113 X. Liu, K. H. R. Baronian and A. J. Downard, Anal. Chem.,
2008, 80, 8835–8839.
114 P. Bertoncello, J. P. Edgeworth, J. V. Macpherson and
P. R. Unwin, J. Am. Chem. Soc., 2007, 129, 10982–10983.
115 T. Gabay, M. Ben-David, I. Kalifa, R. Sorkin, Z. R. Abrams,
E. Ben-Jacob and Y. Hanein, Nanotechnology, 2007, 18, 035201.
116 I. Dumitrescu, J. P. Edgeworth, P. R. Unwin and
J. V. Macpherson, Adv. Mater., 2009, 21, 3105–3109.
117 E. W. Keefer, B. R. Botterman, M. I. Romero, A. F. Rossi and
G. W. Gross, Nat. Nanotechnol., 2008, 3, 434–439.
118 V. Parpura, Nat. Nanotechnol., 2008, 3, 384–385.
Chem. Commun., 2009, 6886–6901 | 6901