Electroless plated gold as a support for carbon nanotube electrodes

Article abstract of DOI:10.1016/j.electacta.2008.11.055

A monolayer of -NH₂ terminated 3-aminopropyltriethoxysilane (APS) was self-assembled onto a p-type silicon (1 0 0) substrate. This amine terminated silane monolayer provided an electrostatic point of attachment for citrate stabilised gold collo

Full text of DOI:10.1016/j.electacta.2008.11.055

Electrochimica Acta 54 (2009) 3191–3198
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Electroless plated gold as a support for carbon nanotube electrodes
∗
Benjamin S. Flavel, Jingxian Yu, Amanda V. Ellis, Joseph G. Shapter
School of Chemistry, Physics & Earth Sciences, Flinders University, Sturt Road, Bedford Park, Adelaide SA 5001, Australia
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 25 September 2008
Received in revised form
A monolayer of –NH2 terminated 3-aminopropyltriethoxysilane (APS) was self-assembled onto a p-type
silicon (1 0 0) substrate. This amine terminated silane monolayer provided an electrostatic point of attach-
ment for citrate stabilised gold colloid nanoparticles, which act as ‘seed’ particles for the electroless
deposition of gold, creating an electrolessly deposited gold layer on silicon. A –NH2 terminated cysteamine
monolayer was then deposited onto the gold layer and carbon nanotubes, with high carboxylic acid func-
tionality, were immobilised via a condensation reaction. A redox active molecule ferrocenemethanol
was then chemically attached to the immobilised carbon nanotubes. These nanostructures were used as
working electrodes in cyclic voltammetry to observe the oxidation and reduction of ferrocene. Important
electrochemical parameters such as electrode kinetics, electron transfer rate and surface concentration
of the redox active molecules were obtained, providing information on the ability of electroless plated
gold surfaces to act as supports for carbon nanotube-based electrodes. This information has also provided
insights into the behaviour of vertically aligned carbon nanotubes immobilised on nanoscale gold wires,
which have been previously fabricated using atomic force microscopy.
2
0 November 2008
Accepted 22 November 2008
Available online 3 December 2008
Keywords:
Silicon
Cyclic voltammetry
Electroless plating
Carbon nanotubes
Gold
Cysteamine
Nanoparticle
©
2008 Elsevier Ltd. All rights reserved.
1
. Introduction
glassy carbon electrode and held in place with a Nafion coating.
Similarly, Zhou and coworkers [14] fabricated a modified electrode
by drop coating a N,N-dimethylformamide (DMF) solution of single-
walled carbon nanotubes (SWCNTs) on a gold disk electrode.
However, more recently it is the chemically attached, vertically
aligned, carbon nanotube electrodes which are showing promise
for future device fabrication [15–18]. These systems provide direct
electrical communication between the underlying electrode and
redox active species with no need for redox mediators and show
potential to create reagent-less sensing devices [5]. The most com-
mon method to create chemically immobilised carbon nanotube
electrode arrays is by self-assembly [7–9,19–21]. Oxidative treat-
ment of SWCNTs results in the creation of carboxylic acid end
groups that can be converted to carbodiimide leaving groups using
dicyclohexylcarbodiimide, which in turn can be reacted with an
amine terminated self-assembled monolayer [4].
A substantial body of work exists utilising a self-assembled alka-
nethiol [7,19,21] or silane [6] monolayer to chemically attach carbon
nanotubes to gold electrodes. For example Lee and Lee [7] have
used 11-mercaptoundecanoic acid to attach carbon nanotubes to a
patterned gold array on polyethylene terephtalate to create flex-
ible field emitter arrays. Liu and coworkers [19] have prepared
electrochemically addressable carbon nanotube microelectrodes
using an insulating amino-n-undecylmercaptan monolayer. Zeng
and Huang [6] have created a multi-walled carbon nanotube/(3-
mercaptopropyl)trimethoxysilane bilayer modified gold electrode
for the electrochemical study and determination of fluphenazine.
Finally, Gooding et al. [20] have measured the electrochemical
Since their discovery [1–3] carbon nanotubes have received
considerable research attention due to their unique structural,
mechanical and electronic properties [4,5]. Carbon nanotubes
display metallic, semiconducting and superconducting electron
transport, store guest molecules within their hollow core and have
the largest elastic modulus of any known material [5]. It is expected
that these unique properties will see carbon nanotubes used in a
variety of different fields, such as nanoelectronic devices [4,6], elec-
tron field emission sources [7], batteries and ultra-sharp probes for
scanning probe microscopy [8,9]. One field that has shown a great
deal of interest in carbon nanotubes is electrochemistry, in which
nanostructuring of carbon nanotube electrodes helps to promote
electron transfer reactions [10]. Furthermore, due to the carbon
nanotubes’ superb chemical stability, low mass density, low resis-
tivity and large surface area they are the ideal base material for
electrochemical sensors or electrodes [5,10,11].
There are primarily two strategies employed to modify electrode
surfaces with carbon nanotubes [5]. These are through physisorp-
tion or chemisorption via chemical reaction or electrochemical
surface activation [5,12]. An example of a physisorption strategy
was reported by Cai and Chen [13] in which carbon nanotubes dis-
persed in cetyltrimethylammonium bromide were spread onto a
∗ Corresponding author. Tel.: +61 8 82012005.
E-mail address: joe.shapter@flinders.edu.au (J.G. Shapter).
0
013-4686/$ – see front matter © 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2008.11.055

3192
B.S. Flavel et al. / Electrochimica Acta 54 (2009) 3191–3198
characteristics of ferrocenemethylamine-modified single-walled
carbon nanotubes immobilised to a self-assembled monolayer of
mercaptoethylamine on a polished gold electrode.
2.2. Synthesis of gold colloid nanoparticles
Prior to use, all glassware was thoroughly cleaned with a 3:1
(v/v) solution of 36% hydrochloric acid (Univar) and 70% nitric
acid (Univar). Colloidal gold nanoparticles were prepared using the
method outlined by Chen and coworkers [29] and Dong and cowork-
ers [27]. Firstly, 1 mL of a 1% aqueous solution of 99.9% gold (III)
chloride (Sigma–Aldrich) was added to 100 mL of HPLC grade water
produced by a Milli-Q Plus system (Millipore), hereafter referred to
as Milli-Q water, under vigorous stirring for 1 min. One millilitre of
a 1% aqueous solution of 99% sodium citrate (Sigma–Aldrich) was
then added and further stirred for 1 min. Finally, 1 mL of a 0.075%
solution of 99% sodium borohydride (Sigma–Aldrich) in 1% aque-
ous sodium citrate was added and the solution stirred for 5 min. An
intense burgundy solution, characteristic of gold nanoparticles [32],
Although substantial work has been performed to create elec-
trodes on polycrystalline [21], magnetron sputtered [7], thermally
evaporated [22–24] and bulk [19,20,25] gold surfaces, other forms
of gold remain to be characterised as support substrates. Electro-
less gold deposition has recently received considerable research
interest as a result of its ability to create coatings on small selected
areas [26–29]. Electroless deposition is an autocatalytic redox pro-
cess, occurring only on catalytic surfaces [29], which is simple
and inexpensive to perform using solution chemistry [27]. In this
3
+
method Au is reduced to bulk metal resulting in surface con-
fined growth of the catalytic gold nanoparticles followed by an
eventual coalescence to form a complete layer [27,28]. Previously,
Dong and coworkers [27] have assessed the suitability of gold
films prepared by electroless plating for use as electrochemical
electrodes. Using the technique of cyclic voltammetry, electroless
plated gold was found to be highly stable with no observable change
in voltammogram shape after several hundred cycles, suggesting
it is ideal for use as an electrode material. However, the effect
of changes in surface topography, important to electrochemistry
◦
was then obtained and stored at 4 C until required. The prepared
nanoparticles are approximately 7 nm in diameter [30].
2.3. Immobilisation of gold nanoparticles to APS
The process to immobilise colloidal gold nanoparticles onto a
APS self-assembled monolayer on a highly boron doped p-type sil-
icon (1 0 0) substrate (0.5 mm thickness, 1 mꢀ cm resistivity and
polished on one side) from Virginia Semiconductor Inc., USA)
is shown schematically in Fig. 1, steps 1–4. Self-assembly of a
silane monolayer on silicon has been previously described in detail
[15,30,33,34]. Self-assembly of APS onto a hydroxylated silicon sur-
[
20] which occurs as a result of placing a self-assembled mono-
layer, or subsequently carbon nanotubes, on the surface is unclear.
To date, very little work has been directed towards the charac-
terisation of self-assembled monolayers on electroless gold. One
example by Evans and coworkers [26] used infrared spectroscopy
to measure self-assembled monolayer order. For a self-assembled
monolayer of octadecanethiol a high level of monolayer order-
ing was observed compared to a di(nonacosa-10,12-diyn) disulfide
layer which showed considerable levels of disorder. The differences
in the ordering of these layers would be expected to influence the
electrochemical behaviour if they were used as electrodes.
In our previous work [30] carbon nanotubes were immobilised
on gold nanostructures electrolessly plated on lithographically
defined regions of 3-aminopropyltriethoxysilane (APS) utilising
a scanning probe technique known as atomic force anodisation
lithography. Due to the inherent difficulty in creating an electrical
connection to a gold nanostructure for electrochemical analysis,
important information about the electron transfer kinetics could
not be obtained. This work attempts to address this issue by using
electroless deposition to create a complete gold layer on a silicon
wafer, which due the increased surface area, could easily be con-
nected. The chemical approach used remained exactly the same
as that used for the nanostructures. Important electrochemical
parameters such as electrode kinetics, electron transfer rate and
surface concentration of redox active molecules could therefore
be obtained. To our knowledge this work also provides the first
electrochemical analysis of a carbon nanotube electrode chemi-
cally attached to a self-assembled monolayer on electroless plated
gold.
face results in a surface terminated with a –NH functionality, Fig. 1,
2
step 3, which provided the point of attachment for citrate stabilised
gold colloid nanoparticles, step 4. The amine terminated silicon
substrate was subsequently immersed into the gold colloid solution
(pH 5) for a period of 2 h. The gold colloid assembly was driven by
the electrostatic attraction between the negative, citrate-capped,
gold nanoparticles and the partially protonated amine layer (pKa
7.5) [35–38].
2.4. Electroless plating of gold
Prior to use the immobilised gold nanoparticle substrates were
ultra-sonicated for 2 min, thoroughly rinsed with water and then
blown dry under a stream of ultra-pure nitrogen. This ensured
that only nanoparticles electrostatically attached to the surface
remained when placed into the gold plating solution. Fig. 1, step 5,
shows how the gold nanoparticle monolayer acts as ‘seed’ particles,
to initiate the electroless deposition of gold. The substrate was then
placed into a gold plating solution consisting of aqueous 0.4 mM
99% hydroxylamine hydrochloride (Sigma–Aldrich) and 0.1% gold
chloride trihydrate for a period of up to 1 min before being thor-
oughly washed with Milli-Q water. This resulted in the creation of
a gold layer on the surface of the silicon substrate.
2.5. Attachment of ferrocenemethanol-modified carbon
nanotubes to electroless plated gold nanostructures
2. Experimental
2
.1. Atomic force microscopy (AFM)
The preparation of carboxylated SWCNTs has been described in
detail previously [15,16,33]. Fig. 1, steps 6–8 shows the approach
used to immobilise ferrocenemethanol-modified carbon nanotubes
onto the electroless plated gold layer using a cysteamine monolayer
as a molecular anchor. The gold plated substrate was immersed
into a 1 mM 95% cysteamine (Sigma–Aldrich) ethanol solution
for a period of 40 min. The substrate was then removed, thor-
oughly washed with ethanol and dried under a stream of ultra-pure
nitrogen and then placed into a 5 mL 99.9% DMF (Southern Cross
Atomic force microscope images were taken in air with a multi-
mode head and Nanoscope IV controller (Digital Instruments,
Veeco, Santa Barbara) operating in tapping mode. Commercially
available silicon cantilevers (FESP-ESP series, Veeco probes, Santa
Barbara) with fundamental resonance frequency between 70 and
8
5 kHz were used. Topographic (height) images were obtained at
a scan rate of 1 Hz with the parameters set point, amplitude, scan
size, and feedback control optimised for each sample. All images
presented represent background-subtracted data (using the flatten
feature in the Digital Instruments or WSxM [31] software).
ꢀ
Scientific Pty. Ltd.) solution containing both 2.5 mg of 99.0% N,N -
dicyclohexylcarbodiimide (DCC) (Fluka Production GmbH) and
1 mg of carboxyl functionalised carbon nanotubes. Prior to placing

B.S. Flavel et al. / Electrochimica Acta 54 (2009) 3191–3198
3193
Fig. 1. Mechanism for the fabrication of ferrocenemethanol-modified carbon nanotube electrodes on electroless plated gold.
the gold plated substrate in for 24 h, the carbon nanotube/DCC/DMF
suspension was ultra-sonicated for 5 h to evenly disperse the car-
bon nanotubes. After removal from the nanotube suspension, the
gold plated silicon wafer with the immobilised carbon nanotubes
attached was rinsed in acetone and allowed to dry. Fig. 1, step
the attached carbon nanotube substrate was immersed into 3 mL
of 99.9% dimethyl sulfoxide (Sigma–Aldrich) containing 0.3 mg of
99.0% DCC and 1.5 mg of 97% ferrocenemethanol (Sigma–Aldrich)
for 48 h at 25 C. The substrate was then rinsed in acetone and dried
with nitrogen.
◦
7, shows the immobilised carbon nanotubes which still have car-
boxylic acid moieties available after attachment which can then
be used for further modification. In this case, the redox active
molecule, ferrocenemethanol, was attached to the free carboxyl
moieties using a condensation reaction [15,16,18]. To achieve this
2.6. Electrochemical techniques
Electrochemical measurements were performed with
BAS100B Electrochemical Analyser (Bioanalytical Systems Inc.,
a

3194
B.S. Flavel et al. / Electrochimica Acta 54 (2009) 3191–3198
Fig. 2. Schematic of the specially designed electrochemical cell and the ferrocenemethanol-modified carbon nanotube surface examined.
USA), operating in cyclic voltammetry mode using a specially
designed electrochemical cell, which has been described previ-
ously [15,16,18] and shown schematically in Fig. 2. Small step sizes
were used to ensure that there are no effects from digitisation of
the sweep voltages. Unlike previous work the working electrode
was connected to the electroless plated gold layer instead of the
silicon substrate. All data was collected using the BAS100W (Bio-
analytical Systems Inc., USA) software V2.3 with data presented
representing a Fourier transform of the raw signal. The background
capacitive current was subtracted using the UTLIS, Utilities for
Data Analysis software (Dirk Heering). An electrolyte solution
3. Results and discussion
Typically a self-assembled monolayer of cysteamine on flat gold
substrates is created by exposure to a cysteamine solution for a
period of 5–24 h ensuring a dense well-aligned layer [9,21,39,40].
However, in this work due to the porous nature [27,28,41] of the
electrolessly deposited gold, extended exposure to the cysteamine
ethanol solution was found to completely delaminate the gold. For
extended exposure times it was believed that the cysteamine was
able to penetrate under the gold, displacing it from the surface.
This is most likely due to the high affinity of gold for sulphur
containing moieties [42,43]. An exposure time of less than 1 h
was therefore performed and was found to leave the gold intact
whilst also providing a sufficient coating. Fig. 3 shows an AFM
image of a gold layer on a silicon substrate created by electroless
−1
containing 1 mmol L
TBAP) (Fluka) in acetonitrile (Scharlau Chemie) was used. The
potential was swept initially in the anodic direction from 100 to
of 98% tetrabutylammonium perchlorate
(
−1
1200 mV at scan rates of 100–3000 mV s
.
Fig. 3. AFM images of an (a) electroless plated gold layer on silicon before and (b) after attachment of ferrocenemethanol-modified carbon nanotubes.

B.S. Flavel et al. / Electrochimica Acta 54 (2009) 3191–3198
3195
Fig. 4. (a) Cyclic voltammograms showing the scan rate dependence of ferrocenemethanol-modified carbon nanotubes attached to electroless plated gold and (b) background
−1
subtracted 100 mV s voltammogram.
deposition for a period of 25 s both (a) before and (b) after attach-
ment of ferrocenemethanol-modified carbon nanotubes. Prior to
the attachment of carbon nanotubes, Fig. 3(a), an approximately
subtracted data to be at 717 and 680 mV, with peak currents of 64
and −70 nA, respectively. This represents a peak separation (ꢁEp)
of 37 mV and is a decrease of 59 mV compared to a bulk gold sur-
face with ferrocenemethanol in solution, as shown as an inset in
(a), where anodic and cathodic peak positions were found to be
741 and 645 mV, respectively. As discussed in our previous work
attaching carbon nanotubes directly to silicon [15,16], the reduc-
tion in peak separation upon attachment of ferrocenemethanol to
the carbon nanotube is attributed to the highly efficient electron
transfer down the length of the nanotube. However, in previ-
ous work peak separations of 136 and 56 mV were obtained
for ferrocenemethanol-modified carbon nanotubes attached to a
macro- and nanoscale silicon electrode [16]. This is an increase in
peak separation of up to 99 mV compared to that observed in the
present work implying that the electron transfer process in previous
work was substantially slower. This can be explained by considering
the substrate composition for the different cases. In this work the
substrate comprised of electroless plated gold. Previously a silicon
substrate with a thin insulating oxide layer was used. It is believed
that the presence of the oxide layer in the case of the silicon lead
to the reduction in electron transfer rate leading to larger peak
separations.
4
0 nm thick, rough and porous gold layer typical of electroless depo-
sition [27,28,41] can be seen topographically and 3-dimensionally.
The image root mean square (Img. RMS) is a measure of rough-
ness and a value of 1.41 nm was determined for a 6 ␮m × 6 ␮m
area. After attachment, Fig. 3(b), both vertically and horizontally
aligned carbon nanotubes are found to be immobilised on the
cysteamine-modified gold surface. The image root mean square
roughness was found to increase to 4.29 nm, indicating a substan-
tially rougher surface due to the presence of carbon nanotubes.
A ‘saw-tooth’ like cross-section, which is characteristic of verti-
cal alignment [15,18,33,44–46] can be seen with carbon nanotube
bundle heights of between 5 and 58 nm. This is in agreement with
previous work [33]. However, unlike previous work on silicon sub-
strates the shorter carbon nanotubes (5–10 nm) are not as obvious
in the three-dimensional AFM image due to the rough electroless
gold layer.
Fig. 4(a) shows the raw data obtained from cyclic voltammetry at
−1
scan rates of 100, 200, 400 and 550 mV s and (b) the background-
subtracted voltammogram for ferrocenemethanol-modified carbon
nanotubes. Distinct redox waves can clearly be seen with the
anodic and cathodic peak positions found from the background-
At slow scan rates, the peak separation observed is constant and
perhaps surprisingly is non-zero. For attached species in an ideal

3196
B.S. Flavel et al. / Electrochimica Acta 54 (2009) 3191–3198
Fig. 5. Cyclic voltammograms of (a) control electrodes and (b) cysteamine self-assembled on electroless plated gold at scan rates of 150–550 mV s 1
−
.
system, zero peak separation at low scan rates would be expected
but with complicated nanotube systems that will never be seen as
zero peak separation implies all the redox species are in the same
environment. Due to the various environments and complexity of
the systems being examined nanotube systems do not behave ide-
ally. This has been observed in previous work in this lab and others
and there is never a non-zero peak separation [18,47,48]. This is
particularly common when interfaces are fabricated in a stepwise
manner. At higher scan rates where the voltammetry is controlled
by the rate of electron transfer, the peak separation does change as
a function of scan rate.
from ferrocenemethanol-modified carbon nanotubes a variety of
control experiments were performed at a scan rate of 100 mV s
−
1
and are shown in Fig. 5(a). Firstly, a cysteamine monolayer was
assembled onto a gold layer created by electroless deposition. Sec-
ondly, carbon nanotubes were immobilised to the cysteamine. Then
finally carbon nanotubes were immobilised onto the gold layer and
exposed to a ferrocenemethanol solution in the absence of DCC
hence no attachment of ferrocenemethanol is expected. As antici-
pated, the peaks observed at 717 and 680 mV in Fig. 4, which were
assigned to ferrocenemethanol redox waves are not observed in any
of the control experiments. However, small oxidative oscillations
can be seen between 300 and 700 mV, especially in the case of the
cysteamine monolayer, and are attributed to the oxidative desorp-
tion of cysteamine. For substrates consisting of carbon nanotubes
and carbon nanotubes exposed to ferrocenemethanol solution in
the absence of DCC, although cysteamine is present as a molecular
anchor for the carbon nanotube, pronounced cysteamine desorp-
tion waves were not as obvious. It is believed that in these cases
the attached carbon nanotube acts as a barrier preventing signif-
icant desorption with desorption only possible from bare regions
containing no carbon nanotubes.
Along with the redox waves consistent with ferrocenemethanol
a small amount of oxidation is observed between 500 and 680 mV,
as shown in Fig. 4(a). This can be attributed to oxidative desorp-
tion of the underlying cysteamine monolayer [22,24,49]. As a result
of the anodic ferrocenemethanol wave overlaying the cysteamine
oxidation, the anodic peak current in the raw signal is observed
to be larger than the cathodic. For example, as shown in Fig. 4(a),
−
1
the raw peak currents for a scan rate of 550 mV s
were found
to be approximately 580 and −420 nA for the anodic and cathodic
waves, respectively. Typically for a completely reversible electro-
chemical process symmetry between anodic and cathodic peak
currents would exist [50]. To confirm that the increased oxida-
tion current was from cysteamine and to ensure that the redox
waves previously attributed to ferrocenemethanol were indeed
As this is the first time a self-assembled monolayer on electroless
plated gold has been used for electrochemistry further investiga-
tion into the cysteamine desorption was carried out. To highlight
the oxidative desorption, Fig. 5(b) shows the raw data obtained from

B.S. Flavel et al. / Electrochimica Acta 54 (2009) 3191–3198
3197
cyclic voltammetry at scan rates of 150, 300, 450 and 550 mV s−1 of
a cysteamine monolayer assembled on the electrolessly deposited
gold. Two oxidative waves at 526 and 680 mV can clearly be seen.
The oxidation of the cysteamine monolayer involves the oxidation
−
−
of the sulphur bond to yield either SO3 or SO2 moieties, caus-
ing the desorption of the self-assembled monolayer from the gold
surface [49]. Previously Porter and coworkers [22] have investi-
gated the desorption of alkanethiolate monolayers from a variety
of gold electrodes. Vacuum evaporation was used to deposit a gold
layer on mica, silicon and glass substrates to which butanethiol and
octanethiol were self-assembled. In the case of the Au/mica sur-
face an atomically flat Au (1 1 1) surface was obtained as opposed
to substantially rougher, stepped microstructures obtained for the
Au/silicon and Au/glass substrates. During electrochemical desorp-
tion a single wave was observed for the Au/mica substrate but two
waves were observed for the Au/silicon and Au/glass [22], which is
consistent with our work. This difference between the Au/mica and
the silicon or glass was attributed to the Au/mica substrate con-
sisting largely of a single binding site for the alkanethiol compared
to the Au/silicon and Au/glass substrates, which had a large num-
ber of non Au (1 1 1) binding sites due to their increased surface
roughness. Similarly, it is believed that the observed multi-wave
desorption of cysteamine from electroless gold in this work can
be attributed to the pronounced heterogeneity of the thiol bind-
ing sites due to the highly porous nature [27,28,41] of electrolessly
deposited gold. Furthermore, as a result of the short self-assembly
time for cysteamine the layer is likely to be highly disordered
[
9,21,39,40] on the gold surface. Previously, in work conducted by
Gooding et al. [4,21,25,51] to create electrodes by chemically attach-
ing carbon nanotubes to bulk gold, the oxidation of cysteamine
was not reported. This suggests that the use of electroless plated
gold as a support for self-assembled monolayer formation is not as
suitable as other forms of gold when used as an electrode for elec-
trochemistry, especially when precise knowledge of peak potentials
or currents is required.
Fig. 6. (a) Dependence of peak current on scan rate and (b) determination of the
dimensionless parameter m.
1
4
−2
−8
−2
(or 6.89 × 10 molecules cm
)
and 9.59 × 10 mol cm
or
16
−2
(5.77 × 10 molecules cm ) which have been obtained previ-
ously [16] for macro- and nanoscale ferrocenemethanol-modified
carbon nanotube electrodes on silicon, respectively. It is possible
that in this work the observed reduction in surface concentration
could be attributed to the short exposure time of cysteamine to
gold resulting in less chemically attached carbon nanotubes and
hence binding sites for ferrocenemethanol. However, the surface
concentration is comparable to recent work by Gooding et al. [20]
who have immobilised ferrocenemethylamine-modified carbon
nanotubes to a bulk gold surface and obtained a surface coverage
Information about the electrode kinetics for the fer-
rocenemethanol redox waves was then obtained using cyclic
−
1
voltammetry at scan rates of 100–3000 mV s
. Oxidation
and reduction peak currents were obtained from background-
subtracted curves and plotted against scan rate, as shown in
Fig. 6(a). Fig. 6(a) shows that a linear relationship exists between
peak current and scan rate indicating that the electrochemical
response measured is from surface bound electro-active species
[
16,18,52] confirming ferrocenemethanol attachment to the car-
−
10
−2
bon nanotube. Furthermore, as can be seen in Fig. 4(a) the peak
separation remains relatively constant with increasing scan rate,
as expected for a surface bound redox-active species. The surface
concentration of ferrocenemethanol molecules can thus be calcu-
lated by determining the area under the oxidation/reduction peaks
as described by Laviron’s theory [52]:
of 2.5 × 10
mol cm .
The apparent rate constant of electron transfer for the carbon
nanotube electrode can also be calculated using Laviron’s method
[52] for the condition of ꢁEp < 200/n mV. By considering the value
of the transfer coefficient, ˛, to be between 0.3 and 0.7, the average
electron transfer rate, ks, can be estimated according to the formula:
2
2
n F Aꢂ v
nFQv
4RT
mvnF
Ip =
=
(1)
ks =
(2)
4
RT
RT
where Ip is the peak current (A), ꢂ is the surface concentra-
where m is a dimensionless parameter related to the peak-to-
peak separation. By using the slope from a plot of ꢃ = f (m ), as
−2
2
−1
tion (mol cm ), A is the electro-active area (cm ), Q is the
peak area of the voltammogram (C), and n is the number of
electrons involved. The electro-active area is the geometric
area of the sample as the roughness of the nanotube surface
would be very difficult to take into account and hence making
determination of the actual area impossible. Fig. 4(b) shows
that the peak area for the anodic and cathodic waves at a
shown in Fig. 6(b), mꢃ was determined and hence ks found to be
−
1
36.95 ± 2.46 s . This represents an increase in electron transfer
rate by a factor of about two compared to previous work on sili-
−
1
con where ks was found to be 15.6 s [16]. This increase in transfer
rate is to be expected as the oxide layer on silicon reduces the
transfer rate compared to that observed for gold. Furthermore, it
is higher than most previously observed transfer rates for immo-
bilised carbon nanotube electrodes [15,16,18,21] although Gooding
−
1
−8
and
scan rate of 100 mV s
7
were calculated to be 7.22 × 10
.99 × 10 C, respectively. From these measurements the average
ferrocenemethanol surface concentration was determined to be
−8
−
1
et al. [20] have calculated a transfer rate of 459 ± 132 s
for
−
10
−2
13
−2
1.50 × 10
mol cm (or 9.06 × 10 molecules cm ). This surface
ferrocenemethylamine-modified carbon nanotubes. It is believed
that the observed increase in electron transfer by Gooding et al.
−
9
−2
concentration is ten times lower than the 1.14 × 10 mol cm

3
198
B.S. Flavel et al. / Electrochimica Acta 54 (2009) 3191–3198
[16] B.S. Flavel, J. Yu, J.G. Shapter, J.S. Quinton, Electrochim. Acta 53 (2008) 5653.
can be attributed to the use of a bulk gold electrode reducing elec-
tron percolation which is expected to exist for a porous [27,28,41]
electroless plated layer. Despite this, upon performing a complete
electrochemical analysis the attachment of vertically aligned car-
bon nanotubes to nanoscale lateral gold wires, as shown previously
[
17] J. Yu, S. Mathew, B.S. Flavel, M.R. Johnston, J.G. Shapter, J. Am. Chem. Soc. 130
2008) 8788.
(
[18] J. Yu, J.G. Shapter, M.R. Johnston, J.S. Quinton, J.J. Gooding, Electrochim. Acta 52
(2007) 6206.
[
19] P. Diao, Z. Liu, B. Wu, X. Nan, J. Zhang, Z. Wei, Chem. Phys. Phys. Chem. 10 (2002)
98.
8
[
30], looks promising as a future approach to fabricate vertically
[
20] J.J. Gooding, A. Chou, J. Liu, D. Losic, J.G. Shapter, D.B. Hibbert, Electrochem.
Commun. 9 (2007) 1677.
integrated carbon nanotube wires in electronic devices.
[21] J.J. Gooding, R. Wibowo, J. Liu, W. Yang, D. Losic, S. Orbons, F.J. Mearns, J.G.
Shapter, D.B. Hibbert, J. Am. Chem. Soc. 125 (2003) 9006.
4
. Conclusions
[
22] M.M. Walczak, C.A. Alves, B.D. Lamp, M.D. Porter, J. Eletroanal. Chem. 396 (1995)
03.
[23] M.D. Porter, T.B. Bright, D.L. Allara, C.E.D. Chidsey, J. Am. Chem. Soc. 109 (1987)
559.
24] C.A. Widrig, C. Chung, M.D. Porter, J. Electroanal. Chem. 310 (1991) 335.
1
This work has provided a detailed analysis of the electro-
3
chemical and kinetic parameters of ferrocenemethanol-modified
carbon nanotubes immobilised on an electrolessly deposited gold
layer. The chemical approach used to create the electrodes was
exactly the same as previously used to create nanostructure sys-
tems of carbon nanotubes on gold wires. Cyclic voltammetry of the
chemically modified carbon nanotube working electrodes demon-
strated that the electrochemical signal was from surface bound
redox species. The number of surface bound ferrocenemethanol
[
[25] J. Liu, A. Chou, R. Wibowo, M.N. Paddon-Row, J.J. Gooding, Electroanalysis 17
(2005) 38.
26] H. Menzel, M.D. Mowery, M. Cai, C.E. Evans, Adv. Mater. 11 (1999) 131.
27] Y. Jin, X. Kang, Y. Song, B. Zhang, G. Cheng, S. Dong, Anal. Chem. 73 (2001) 2843.
[
[
[
[
[
28] S. Hrapovic, Y. Liu, G. Enright, F. Bensebaa, J.H.T. Luong, Langmuir 19 (2003)
3958.
29] F. Guan, M. Chen, W. Yang, J. Wang, S. Yong, Q. Xue, Appl. Surf. Sci. 240 (2005)
24.
30] B.S. Flavel, J. Yu, A.V. Ellis, J.S. Quinton, J.G. Shapter, Nanotechnology 19 (2008)
445301.
[31] I. Horcas, R. Fernandez, J.M. Gomez-Rodriguez, J. Colchero, J. Gomez-Herrero,
A.M. Baro, Rev. Sci. Instrum. 78 (2007) 13705.
13
−2
molecules was calculated to be 9.06 × 10 molecules cm and the
−
1
electron transfer rate determined to be 36.95 ± 2.46 s . This rep-
resents a substantially higher electron transfer rate compared to
other carbon nanotube-based electrodes. Due to the roughness of
the porous gold layer two oxidative cysteamine desorption peaks
where also observed. It is believed that the development of electro-
less techniques to create a more heterogeneous, smooth gold layer
are required to reduce cysteamine monolayer desorption.
[
32] C.D. Keating, M.D. Musick, M.H. Keefe, M.J. Natan, J. Chem. Educ. 76 (1999)
949.
[
[
[
33] B.S. Flavel, J. Yu, J.G. Shapter, J.S. Quinton, Carbon 45 (2007) 2551.
34] B.S. Flavel, J. Yu, J.G. Shapter, J.S. Quinton, Soft Matter 5 (2009) 164.
35] A.N. Shipway, E. Katz, I. Willner, Chem. Phys. Phys. Chem. 1 (2000) 18.
[36] J. Park, H. Lee, Colloids Surf. A 257–258 (2005) 133.
[37] A.N. Shipway, M. Lahav, I. Willner, Adv. Mater. 12 (2000) 993.
38] Q. Li, J. Zheng, Z. Liu, Langmuir 19 (2003) 166.
39] R.M. Petoral, K. Uvdal, Colloids Surf. B 25 (2002) 335.
[40] S.K. Kim, J. Zhang, M. Lee, H.Y. Choi, H. Lee, Curr. Appl. Phys. 6S1 (2006) e48.
41] T.N. Vorobyova, O.N. Vrublevskaya, A.V. Vengura, Surf. Coat. Technol. 200 (2005)
[
[
References
[
[
[
[
[
[
[
[
[
[
1] P.G. Wiles, J. Abrahamson, Carbon 16 (1977) 341.
2] J. Abrahamson, P.G. Wiles, B.L. Rhoades, Carbon 37 (1999) 1873.
3] S. Iijima, Nature 354 (1991) 56.
4] J.J. Gooding, Electrochim. Acta 50 (2005) 3049.
5] J. Wang, Electroanalysis 17 (2005) 7.
2481.
[42] S. Xu, G. Tu, B. Peng, X. Han, Anal. Chim. Acta 570 (2006) 151.
[43] C.M. Pradier, M. Salmain, Z. Liu, C. Methivier, Surf. Interf. Anal. 34 (2002) 67.
[44] J. Yu, J.G. Shapter, J.S. Quinton, M.R. Johnston, D.A. Beattie, Phys. Chem. Chem.
Phys. 9 (2007) 510.
[45] J. Yu, B.S. Flavel, J.G. Shapter, Fullerenes, Nanotubes, Carbon Nanostruct. 16
(2008) 18.
[46] J. Yu, D. Losic, M. Marshall, T. Bocking, J.J. Gooding, J.G. Shapter, Soft Matter 2
(2006) 1081.
[47] J. Liu, M.N. Paddon-Row, J.J. Gooding, J. Phys. Chem. B 108 (2004) 8460.
[48] S.E. Creager, G.K. Rowe, Electroanal. Chem. 420 (1997) 291.
[49] M. Wirde, U. Gelius, Langmuir 15 (1999) 6370.
6] B. Zeng, F. Huang, Talanta 64 (2004) 380.
7] O. Lee, K. Lee, Appl. Phys. Lett. 82 (2003) 3770.
8] Z. Liu, Z. Shen, T. Zhu, S. Hou, L. Ying, ACS J. Surf. Colloids 16 (2000) 3569.
9] X. Nan, Z. Gu, Z. Liu, J. Colloid Interf. Sci. 245 (2002) 311.
[
[
[
10] P. Zhang, F. Wu, G. Zhao, X. Wei, Bioelectrochemistry 67 (2005) 109.
11] M. Guo, J. Chen, L. Nie, S. Yao, Electrochim. Acta 49 (2004) 2637.
12] G.G. Wildgoose, C.E. Banks, H.C. Leventis, R.G. Compton, Microchim. Acta 152
2006) 187.
13] C. Cai, J. Chen, Anal. Biochem. 325 (2004) 285.
14] L. Wang, J. Wang, F. Zhou, Electroanalysis 16 (2004) 627.
15] B.S. Flavel, J. Yu, J.G. Shapter, J.S. Quinton, J. Mater. Chem. 17 (2007) 4757.
(
[50] A.J. Bard, L.R. Faulkner, Electrochemical Methods: Fundamentals and Applica-
tions, John Wiley & Sons, New York, 2000, Ch. 6.
[51] A. Chou, T. Bocking, N.K. Singh, J.J. Gooding, Chem. Commun. (2005) 842.
[52] B. Ye, S. Qu, F. Wang, L. Li, J. Chin. Chem. Soc. 52 (2005) 1111.
[
[
[