823

Full Paper

Carbon Nanotubes Paste Electrodes. A New Alternative for the

Development of Electrochemical Sensors
G. A. Rivas,* M. D. Rubianes, M. L. Pedano, N. F. Ferreyra, G. L. Luque, M. C. Rodrguez, S. A. Miscoria
INFIQC, Departamento de Fsico Qumica, Facultad de Ciencias Qumicas, Universidad Nacional de Córdoba, Ciudad Universitaria,
5000 Córdoba, Argentina
*e-mail: [email protected]

Received: October 2, 2006

Accepted: November 12, 2006

Abstract
In this work we summarize the recent activities of our group regarding the analytical performance of a new composite
material, the so-called carbon nanotubes paste electrode (CNTPE) obtained by dispersion of multiwall carbon
nanotubes in mineral oil. The electrocatalytic properties towards different redox systems, especially those involved in
important enzymatic reactions are discussed. Significant shifting in the overpotentials for the oxidation and/or
reduction of hydrogen peroxide, NADH, phenol, catechol, dopamine, ascorbic acid, uric acid and hydroquinone are
obtained at CNTPE in comparison with the analogous graphite paste electrode (CPE). The usefulness of the electrode
as a matrix for immobilizing enzymes is also demonstrated. Highly sensitive and selective glucose quantification is
accomplished even without using permselective films or redox mediators. Enzymatic biosensors obtained by
incorporation of lactate oxidase, polyphenol oxidase and alcohol dehydrogenase/NADþ within the composite material
have allowed the successful quantification of lactate, phenol, dopamine, catechin and ethanol. The sensitive
quantification of traces of oligonucleotides and double stranded calf thymus DNA by adsorptive stripping is reported.
The confined DNA layer demonstrated to be stable either in air, acetate or phosphate buffer. The advantages of
incorporating copper particles for the quantification of amino acids and albumin is also discussed.

Keywords: Multiwall carbon nanotubes, Glucose oxidase, Lactate oxidase, Polyphenol oxidase, Alcohol dehydro-
genase, NADH, Catechin, Iridium, Copper, Nucleic acids, Enzymatic biosensors, DNA, Amino acids, Albumin

DOI: 10.1002/elan.200603778

1. Introduction their diameter and chirality [2, 5, 8]. According to the

structural parameters, SWCNTs can be either a metal,
Since 1991, carbon nanotubes (CNTs) have been the center semiconductor or small-gap semiconductor [2, 5, 6 – 8].
of numerous investigations due to their outstanding struc- Since the topological defects in nanotubes result in local
tural, electronic and mechanical properties that make them perturbations of their electronic structure, the pentagonal
a very unique material [1 – 6]. Carbon nanotubes are built defect of the caps make them more metallic than the
from sp2 carbon units and present a seamless structure with cylinders. These defects enhance the chemical reactivity of
hexagonal honeycomb lattices, being several nanometers in the ends giving the possibility to functionalize them, to open
diameter and many microns in length [1, 4]. Each end of the the tubes and fill them with foreign substances [1, 9, 10].
nanotube is capped with half of a fullerene molecule. The Due to their properties, CNTs have received enormous
curvature in the planar hexagonal graphite lattice is attention for the preparation of electrochemical sensors, as
obtained from topological defects like pentagons [4, 5]. it was widely reviewed [3, 11, 12]. Since they are not readily
There are two groups of carbon nanotubes, multiwall soluble in usual media, different strategies for immobilizing
(MWCNTs) and singlewall (SWCNTs) carbon nanotubes them on the surface of electrodes have been proposed.
[4, 5, 7]. MWCNTs can be visualized as concentric and Dispersion of CNTs in acidic solutions [13, 14], N,N’-
closed graphite tubules with multiple layers of graphite dimethylformamide [15], Nafion [16, 17] and chitosan [18]
sheet defining a hole typically from 2 to 25 nm and separated have been successfully used. The incorporation of CNTs in
by approximately 0.34 nm [4, 5, 7]. SWCNTs consist of a composite matrices using different binders like Teflon
single graphite sheet rolled seamlessly, defining a cylinder of [19], bromoform [20], mineral oil [21 – 24] and inks [25]
1 – 2 nm diameter. MWCNTs can be considered as a has been also proposed. The resulting electrodes modified
mesoscale graphite system while SWCNTs are real single with CNTs have been employed for the detection of several
large molecules [8]. bioanalytes like glucose [19, 21], DNA [13, 22], homocystein
The combination of size, structure and topology gives [26], neurotransmitters and related compounds [20, 21, 24,
nanotubes important mechanical and surface properties. 27], uric acid [28] and ethanol [19, 23], amitrol [29] among
The electrical properties of CNTs depend sensitively on others.

Electroanalysis 19, 2007, No. 7-8, 823 – 831 E 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
824 G. A. Rivas et al.

In this work we present a review of the current activities of and Ag/AgCl, 3 M NaCl (BAS, Model RE-5B) were used as
our group on the analytical applications of an electrode counter and reference electrode, respectively. All potentials
proposed by us some years ago, the carbon nanotubes paste are referred to the latter. A magnetic stirrer provided the
electrode (CNTPE), a composite material obtained by convective transport during the amperometric and DNA
dispersion of MWCNTs in mineral oil. The electrochemical accumulation experiments.
behavior of different redox systems at CNTPE as well as the Constant-current chronopotentiometric experiments
performance of enzymatic electrodes developed by incor- were performed with a TraceLab Potentiometric Stripping
poration of enzymes within the CNTPE matrix is reported in Unit PSU 22 (Radiometer, France) connected to a PC.
the following sections. According to the Trace Lab protocol, the potentials were
sampled at a frequency of 30 kHz and the derivative signals
(dt/dE) versus potential (E) were recorded following base-
line fitting.
2. Experimental
The working electrodes were graphite paste electrode
(CPE), carbon nanotubes paste electrode (CNTPE) or
2.1. Reagents
glassy carbon paste electrode (GCPE). The CNTPE was
Hydrogen peroxide (30% V/V aqueous solution) was prepared by mixing in an agata mortar MWCNTs powder
purchased from Baker. Ascorbic acid was obtained from and mineral oil (Aldrich) in a ratio 60.0% w/w nanotubes
Fluka. Dopamine, 3,4-dihydroxyphenylacetic acid (dopac), powder and 40.0% w/w mineral oil. CPE and GCPE were
glucose oxidase (GOx) (Type X-S, Aspergillus niger, prepared in a similar way by mixing graphite powder or
(EC 1.1.3.4), 157 500 Units per gram of solid, Catalog glassy carbon microparticles with mineral oil. CPE and
number G-7141), lactate oxidase (LOx) (from Pediococcus CNTPE containing enzymes were prepared in analogous
species, 36 units/mg solid), alcohol dehydrogenase (ADH) way by including the desired amount of enzyme. CNTPE
(EC 1.1.1.1. from bakers yeast, 428 units/mg protein, 393 containing copper (CNTPE-Cu) was prepared in a similar
units/mg solid), NADþ, NADH, lactate, polyphenol oxidase way, mixing first the copper microparticles with the mineral
(PPO) (EC 1.14.18.1 from mushroom, 2870 units/mg solid), oil for 1 min, followed by the incorporation of the carbon
bovin seric albumin (A-4503) and all amino acids were nanotubes powder and additional mixing for 30 min. A
purchased from Sigma. Uric acid, glucose, hydroquinone, portion of the resulting paste was packed firmly into the
and phenol, l-amino acids were obtained from Merck. cavity (3.0 mm diameter) of a Teflon tube. The electric
Catechol was from Mallinckrodt. Ethanol, methanol and contact was established via a stainless steel screw. A new
isopropanol were from Baker. Other chemicals were surface was obtained by smoothing the electrode onto a
reagent grade and used without further purification. Copper weighing paper. In case of DNA experiments, the mineral oil
(99% purity) microparticles ( 325 mesh, 10% max þ 325 was DNAse, RNAse, protease free (Aldrich)
mesh, 99%) were acquired from Alfa Aesar. Scanning Electronic Microscopy (SEM) pictures were
MWCNTs powder (diameter 20 – 50 nm and lengths of obtained with a Hitachi S3000N Microscope equipped with
1 – 5 microns (short CNTs) and 5 – 20 microns (long CNTs), secondary and backscattered electron detectors, as well as
95% purity, were obtained from NanoLab, (U.S.A). Glassy with a X-ray detector.
carbon spherical powder 0.4 – 12 mm, type 2 was purchased
from Alfa AEsar. Graphite powder (grade # 38) was from
Fisher.
2.3. Procedure
OligoX (5’-ATG TGG AAA ATC TCTAGC AGT-3’) was
purchased from Life Technologies (Grand Island, New The amperometric and cyclic voltammetric experiments
York, USA) as its ammonium salt. Double stranded calf were carried out in a phosphate buffer solution (0.050 M,
thymus DNA (dsDNA) (activated and lyophilized, catalog pH 7.40). The amperometric ones were performed by
number D-4522) was purchased from SIGMA (St. Louis, applying the desired potential and allowing the transient
MO). All other reagents were of analytical grade. DNA current to decay to a steady-state value prior to the addition
stock solutions (nominally 1000 mg/L) were prepared with of the analyte and the subsequent current monitoring.
TE buffer (1  concentrate, 10 mM Tris-HCl, 1 mM EDTA,
pH 8.0).
2.3.1. Nucleic Acids Detection
Ultrapure water (1 ¼ 18 MW) from a Millipore-MilliQ
system was used for preparing all the solutions. Nucleic acids detection consisted of the following steps:
– Pretreatment: By applying 1.300 V for 20 s in a 0.200 M
acetate buffer solution pH 5.00.
2.2. Apparatus – Nucleic acid immobilization: performed at 0.200 V for a
given time by immersion of the pretreated CNTPE in a
The amperometric and voltammetric measurements were
stirred 0.200 M acetate buffer pH 5.00 containing DNA
performed with EPSILON (BAS) and TEQ-02 potentio-
– Washing of the DNA modified electrode with a 0.200 M
stats. The electrodes were inserted into the cell (BAS, Model
acetate buffer solution pH 5.00 for 5 s.
MF-1084) through holes in its Teflon cover. A platinum wire

Electroanalysis 19, 2007, No. 7-8, 823 – 831 www.electroanalysis.wiley-vch.de E 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Carbon Nanotubes Paste Electrodes 825

– Chronopotentiometric transduction: it was performed

after medium exchange in a 0.200 M acetate buffer
solution pH 5.00 applying a constant current of 8.0 mA
with an initial potential of 0.500 V. The anodic signal at
around 1.0 V, corresponding to the guanine oxidation,
was used as the analytical signal.

Repetitive measurements were carried out by polishing the

surface of the electrode on a weighing paper and repeating
the above assay. All experiments were performed at room
temperature.

2.3.2. Amperometric Determination of Aminoacids

Measurements were conducted in a stirred 0.050 M phos-
phate buffer solution pH 7.40 by applying a potential of
0.000 V. Fig. 1. SEM picture of CNTPE (60.0% w/w CNT and 40.0% w/w
oil). Magnification 10 000  .

2.3.3. Albumin Determination

A square-wave voltammogram between  0.200 V and
1.000 V was obtained after holding the CNTPE-Cu at
 0.100 V for 10 min in a stirred solution of albumin. The
voltammetric parameters are the following: frequency of
25 Hz, pulse amplitude of 25 mV, staircase step 10 mV and
scan rate of 0.250 Vs1

3. Results and Discussion

3.1. Electrochemical Behavior of Different Analytes at

CNTPE
Our group proposed for the first time a new composite
material based on the dispersion of MWCNTs within
mineral oil, the called carbon nanotubes paste electrode
(CNTPE) [21]. The results demonstrated that the presence Fig. 2. Cyclic voltammograms for 1.0  103 M ascorbic acid (A),
of the nonconductive phase did not impair the excellent uric acid (B), dopamine (C) and hydroquinone (D) at different
properties of CNTs for the electron transfer. A homoge- electrodes: CPE (dotted line) and CNTPE (solid line). Supporting
neous distribution of CNTs within the mineral oil after electrolyte: 0.050 M phosphate buffer solution pH 7.40. Scan rate:
0.100 V/s. CNTPE (60.0% w/w CNT and 40.0% w/w oil. Adapted
incorporation into the electrode can be observed in the SEM
from Figure 2, Electrochemistry Communications 5 (2003) 689 –
pictures (Fig. 1). 694.
A potential window comparable to that of CPE and better
than that of the analogous glassy carbon composite was
obtained in different supporting electrolyte solutions [30]. reversibility was found at the CNTPE for dopamine and
Figure 2 displays cyclic voltammograms obtained at hydroquinone, compounds that display a quasireversible
0.100 V/s for 1.0  103 M ascorbic acid (A), uric acid (B), behavior at CPE, In fact, the DEp for dopamine and
dopamine (C) and hydroquinone (D) at CPE (dotted line) hydroquinone decreases 141 mV and 181 mV, respectively,
and at CNTPE (solid line). An important decrease in the while the anodic-to-cathodic peak currents ratio decreases
overvoltages for the oxidation of these redox systems was from 5.21 to 3.00 for dopamine and from 1.79 to 1.12 for
observed at CNTPE, indicating that CNTs effectively hydroquinone. These results demonstrate that the improved
improve the electron-transfer kinetics. It is important to electrocatalytic activity of CNTs due to the nanotubes
remark that the excellent electrocatalytic properties of the dimensions, electronic structure and topological defects
carbon nanotubes composite material were observed even present on their surface [31, 32], are retained even in the
with electrodes containing just 10.0% w/w CNTs (not presence of the pasting liquid.
shown). A decrease in the oxidation peak potential of 230 Hydrogen peroxide is an important molecule involved in
and 160 mV was obtained at CNTPE for ascorbic acid and numerous enzymatic processes. Therefore, the development
uric acid, respectively. A significant improvement in the of new methodologies for the sensitive and selective hydro-

Electroanalysis 19, 2007, No. 7-8, 823 – 831 www.electroanalysis.wiley-vch.de E 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
826 G. A. Rivas et al.

nanotubes (1 – 5 mm) demonstrated to be more effective for

the oxidation of ascorbic acid, dopamine and dopac than the
long ones (5 – 20 mm) (not shown). The effect of the content
of mineral oil is another important aspect to consider when
studying these materials. For instance, no significant differ-
ences were observed in the voltammetric parameters of
ascorbic acid, dopac and dopamine when using composites
containing 40.0 and 45.0% w/w mineral oil. Pastes with
50.0% w/w oil were difficult to handle and a shifting of the
peak potentials in the positive direction was observed,
suggesting a slower charge transfer. When long CNTs were
used to prepare the composite, the situation was different.
Pastes containing 40.0% w/w oil were difficult to handle due
Fig. 3. Cyclic voltammograms for 2.0  102 M hydrogen perox-
to the inherent characteristics of the long CNTs that require
ide obtained at graphite composite electrodes containing different more oil to obtain a homogeneous dispersion. Pastes with
amounts of carbon nanotubes: 0.0 (þ), 10.0 (&), 50.0 (*) and 60.0 45.0 and 50.0% w/w long CNTs gave a similar response,
(—)% w/w. In all cases the percentage of mineral oil was 40.0% w/ demonstrating that with these CNTs is possible to obtain
w. The 100% w/w was completed, when necessary, with graphite composites with more oil content.
powder. Scan rate: 0.100 V/s. Supporting electrolyte: 0.050 M
phosphate buffer solution pH 7.40.

3.3. Effect of Electrochemical Pretreatments

gen peroxide determination is receiving enormous atten-
tion. Figure 3 shows the cyclic voltammetric response of Different schemes for CNTs activation have been proposed.
2.0  102 M hydrogen peroxide at carbon composites Many of them have been based on the chemical oxidation. In
electrodes containing different amounts of MWCNTs. general, as a consequence of the pretreatment, there is an
Even for composites containing 10.0% w/w CNTs, the increase in the density of oxygenated functional groups on
catalytic effect is clear. The overvoltages for the oxidation the surface of CNTs, mainly carboxylic ones [5]. Depending
and reduction of hydrogen peroxide largely decrease in the on the pretreatment conditions (nature of the acid, temper-
presence of CNTs and the oxidation and reduction currents ature, time and sonication), the tubules can be opened or
present a noticeable increase. For instance, at  0.100 V the even shortened [34, 35]. Electrochemical pretreatments
reduction currents for hydrogen peroxide were 0.090, 0.360, have been also used, although, in general, they were used as
0.790 and 1.360 mA, while the oxidation currents at 0.700 V a complement of a previous chemical oxidation scheme.
were 0.250, 0.950, 2.880 and 3.260 mA at carbon composites The state of the surface of CNTPE plays an important role
containing 0.0; 10.0; 50.0 and 60.0% w/w CNTs, respectively. on the electrode kinetics. For instance, the electrochemical
The significant decrease in the overvoltage for the reduction behavior of dopamine was highly dependent on the surface
of hydrogen peroxide has allowed working at potentials pretreatment. Various activation procedures were evaluat-
where the interference of easily oxidizable compounds can ed, being the one performed by cycling the potential
be eliminated, opening the doors to many analytical between  1.00 and 1.50 V (18 cycles) at 1.0 V/s in a
applications connected to enzymatic electrodes based on 0.050 M phosphate buffer solution pH 7.40 the one that
the use of hydrogen peroxide as analytical signal. has allowed us to obtain the best response. As it is shown in
The effect of CNTs on the electrochemical behavior of Figure 4, after this pretreatment the DEp for 1.0  103 M
NADH is also interesting. It is widely known that the dopamine decreases 30 mV while the anodic oxidation peak
oxidation of this compound requires elevated overvoltages current increases from 34.3 to 64.9 mA and the currents peak
at carbon materials and that the electrode is passivated after ratio decreases from 3.00 to 2.80.
its oxidation [33]. In the case of CNTPE the oxidation starts The influence of different potentiostatic and potentiody-
at  0.100 V, that is, at potentials 400 mV less positive than at namic pretreatments on the adsorption and electrooxida-
CPE, evidencing once more the unique electrocatalytic tion of DNA was also evaluated. Potentiodynamic pretreat-
properties of CNTs (not shown) [23]. The amperometric ments performed either in a 0.200 M acetate buffer solution
response of 1.0  105 M NADH solution at CNTPE at pH 5.00 or in a 0.050 M phosphate buffer solution pH 7.40
0.400 V decreased only 20.0% after 15 min, at variance with by cycling the potential between  1.000 and 1.500 V at
the behavior observed at CPE, where a decrease of 80.0% 1.0 V/s were not effective because they produced a large
was obtained after the same period (not shown). increase in the background signal that makes the chrono-
potentiometric detection of guanine electrooxidation very
difficult. On the contrary, potentiostatic pretreatments
demonstrated to be effective for the adsorption and electro-
3.2. Effect of the Length of CNTs and Paste Composition
oxidation of oligonucleotides, being the selection of the
The influence of the nature of CNTs on the electrochemical applied potential and time a critical point. The pretreat-
response of the electrodes was also evaluated. Short carbon ments were performed by applying different potentials in a

Electroanalysis 19, 2007, No. 7-8, 823 – 831 www.electroanalysis.wiley-vch.de E 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Carbon Nanotubes Paste Electrodes 827

Fig. 4. Cyclic voltammograms for 1.0  103 M dopamine at

untreated CNTPE (dotted line) and at electrochemically activated
CNTPE (solid line). Electrochemical pretreatment: 18 cycles
between  1.00 and 1.50 V at 1.0 Vs1 in supporting electrolyte.
Supporting electrolyte: 0.050 M phosphate buffer solution
pH 7.40. Scan rate: 0.100 V/s.

0.200 M acetate buffer solution pH 5.00 for 20 s. The

guanine oxidation signal rose with the increase in the Fig. 5. A) Amperometric recordings obtained at CNTPE-GOx
activation potential, mainly due to the increase in the (10.0% w/w) for successive additions of 2.0 mM glucose. B)
density of oxygenated groups on CNTs. However, potentials Average calibration plot obtained from (A). Working potential:
more positive than 1.300 V produced a large background  0.100 V. Supporting electrolyte: 0.050 M phosphate buffer
solution pH 7.40. Adapted from Figure 4, Electrochemistry Com-
current as a consequence of the catalytic effect of CNTs on munications 5 (2003) 689 – 694.
the solvent oxidation. This increase in the background
current made the resolution of the guanine oxidation peak
very complicated. A pretreatment of 20 s at 1.300 V was selection of the working potential is very important to
enough to improve the guanine oxidation signal. Under obtain the selective determination of the hydrogen peroxide
those conditions, a signal almost 13 times higher than that at enzymatically generated. Figure 5A shows the amperomet-
CPE pretreated was obtained. Experiments performed at ric response at  0.100 V obtained at CNTPE-GOx (10.0%
1.300 V for different times showed that times longer than w/w) for successive additions of 2.0 mM glucose. A well
20 s produce a larger increase in the background signal (not defined, fast and very sensitive response was obtained at the
shown). CNTPE-GOx. Figure 5B shows the corresponding calibra-
tion plot obtained as an average of four independent
experiments. The response was linear even up to 25 mM
glucose (4.5 g/L), covering, thus, not only the physiological
3.4. Enzymatic Biosensors Based on the Use of CNTPE
range, but also pathological values. The corresponding
The dramatic improvement in the electrochemical behavior sensitivity and detection limit were (1.13  0.01)  104 nA
of NADH, hydroquinone and hydrogen peroxide, repre- M1 (r ¼ 0.9994) and 0.6 mM (0.11 g/L), respectively. There-
sents a very interesting starting point for the development of fore, the sensitivity obtained with CNTPE-GOx was 43
enzymatic biosensors that involve the detection of these larger than that with the analogous CPE-GOx.
compounds, like oxidases and dehydrogenases-NADþ de- Figure 6 shows the effect of common interferents on the
pendent. The performance of the resulting biosensors amperometric response of 5.0  103 M glucose at  0.100 V
involving CNTs is discussed below. using CNTPE after additions of 1.0  104 M ascorbic acid
(AA), 5.0  105 M acetaminophen and 2.5  104 M uric
acid (UA). No interference of these easily oxidizable
3.4.1. CNTPE Modified with Glucose Oxidase and Lactate
compounds was found even at these levels, indicating that
Oxidase
under these conditions it is possible to obtain not only a
Based on the excellent electrocatalytic properties of CNTs sensitive but also a highly selective glucose determination
towards the oxidation and reduction of hydrogen peroxide, without adding any membrane or redox mediator.
the CNTPE was used for the preparation of glucose The usefulness of a CNTPE containing lactate oxidase
biosensors by the incorporation of GOx into the composite (LOx) for the detection of lactate was also studied [23]. As
matrix. As it is widely known, GOx catalyzes the oxidation expected, due to the noticeable improvement in the electron
of glucose to gluconolactone while oxygen, its natural transfer of hydrogen peroxide at CNTPE, a well defined and
mediator, is converted into hydrogen peroxide [36]. The fast response to lactate was reported. A linear range was

Electroanalysis 19, 2007, No. 7-8, 823 – 831 www.electroanalysis.wiley-vch.de E 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
828 G. A. Rivas et al.

Fig. 6. Current-time profiles performed at CNTPE-GOx (10.0%

w/w) for one addition of 1.0  104 M ascorbic acid (AA), 5.0 
105 M acetaminophen, 2.5  104 M uric acid (UA) and 5.0 
103 M glucose (GLU). Other conditions as in Figure 5.

observed up to 7.0  103 M lactate, with a sensitivity of

(14.4  0.4) mA M1 and a detection limit of 3.0  104 M.

3.4.2. CNTPE Modified with Polyphenol Oxidase

Polyphenol oxidase is an enzyme that catalyzes the oxida-
tion of phenols and catechols to the corresponding quinones
in the presence of oxygen [23]. Figure 7A shows a calibra-
tion plot obtained from amperometric experiments at
 0.050 V after successive additions of 1.0  106 M dopa-
mine. The sensitivity was (3.36  0.03)  103 mA M1, that is,
12 times higher than that obtained with CPE-PPO, while the
detection limit was 1.0  106 M.
The CNTPE-PPO was also used for the quantification of
polyphenols, compounds that are receiving important
attention in the pharmacological and food industry due to
their interesting antioxidant properties. The bioelectrode
demonstrated to be very useful for the detection of catechin,
a polyphenol present in some products like tea and wines. Fig. 7. A) Calibration plot for dopamine obtained from ampero-
Figure 7B shows the amperometric recordings obtained at metric experiments at PPO-CNTPE. B) Amperometric recording
 0.050 V for successive additions of 1.0  105 M catechin. for successive additions of 1.0  105 M catechine at PPO-CNTPE.
The corresponding calibration plot is shown in Figure 7C. C) Calibration plot for catechin obtained from (B). Electrode
composition: 1.5% w/w PPO, 60.0% w/w carbon nanotubes, 38.5%
The sensitivity and detection limits were (2.03  0.09)  103
w/w mineral oil. Working potential:  0.050 V.
mA M1 (r ¼ 0.996) and 1.0  106 M, respectively. The
electrode demonstrated to be highly stable, since after 60
days the sensitivity was the same as the original one.
3.5. Direct Quantification of Nucleic Acids at CNTPE
3.4.3. CNTPE Modified with Alcohol Dehydrogenase
Figure 8 shows the calibration plot for calf thymus double
A fast response was obtained at 0.400 V after additions of stranded DNA after 5 min accumulation at 0.200 V. There is
5.0  103 M ethanol at CNTPE modified with ADH (12.0% a linear relationship up to 15.0 mg L1 and a sensitivity of
w/w) and NADþ (12.0% w/w) with a sensitivity of (44  (12.3  0.9) ms L mg1, (r ¼ 0.995). Detections limits of 2 and
3) mA M1 (not shown) [23]. No response was observed for 170 mg L1 (calculated as 3 times the standard deviation of
methanol even for high concentrations. In the presence of the blank over the sensitivity), were obtained for oligox and
isopropanol the sensitivity was one third of that obtained for dsDNA, respectively.
ethanol. The inset of Figure 8 depicts chronopotentiograms for
The electrode demonstrated to be highly stable since after different concentrations of oligoX: 0.30, 0.50, 0.80 and
2 months at 4 8C the response was still the same as the first 1.00 mg L1. The calibration plot for this oligonucleotide
day, while after 6 months it decreased just 24%. The gave a sensitivity of (163  5) ms L mg1, (r ¼ 0.995), and a
bioelectrode was used to determine the content of ethanol in linear range up to 1.00 mg L1 after 1 min accumulation at
different alcoholic beverages. 0.200 V (not shown).

Electroanalysis 19, 2007, No. 7-8, 823 – 831 www.electroanalysis.wiley-vch.de E 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Carbon Nanotubes Paste Electrodes 829

Fig. 8. Calibration plot for dsDNA at pretreated CNTPE. The

inset shows chronopotentiograms for oligo (dX)21 . Experimental
conditions: Adsorption at 0.200 V for 1 min (A) or 5 min (B) in
0.200 M acetate buffer solution pH: 5.00, containing different
concentrations of the corresponding nucleic acid. Stripping: in
0.200 M acetate buffer solution pH: 5.00; i ¼ 8.0 mA, initial
potential: 0.500 V. Adapted from Figure 5, Electrochemistry
Communications 6 (2004) 10 – 16.

The layer of DNA confined to the electrode demonstrated

to be stable in air, in 0.200 M acetate buffer pH 5.00 and in
0.020 M phosphate buffer pH 7.40 þ 0.50 M NaCl. The
stability of the adsorbed nucleic-acid layer upon transfer to
the blank solutions results promising for developing differ-
ent biosensors based on the use of CNTPE and nucleic acids
as biorecognition element.

3.6. Effect of the Presence of Copper Microparticles

within the CNTPE Fig. 9. Amperometric recordings for successive additions of 5 
104 M l-amino acids at CNTPE 60/40 (a) and CNTPE þ Cu 18%
3.6.1. Glucose Biosensor (b) for A) l-histidine, B) l-lisine, C) l-proline.

Since copper is an excellent catalyst for the reduction of

hydrogen peroxide [37] it was incorporated into the CNTPE amino acids are not electroactive, no response is observed at
to improve the sensitivity of glucose determinations. The CNTPE. On the contrary, at CNTPE/Cu, a fast and sensitive
careful selection of the amount of copper has allowed an response is obtained after successive additions of the amino
important improvement in sensitivity without compromis- acids due to the facilitated oxidation of copper. It is
ing the selectivity. A value of 0.77% w/w was selected as the important to remark that the sensitivity is highly dependent
optimum to achieve the best compromise between sensitiv- on the nature of the amino acid due to the strength of the
ity and selectivity [38]. complex formation between Cu (II) and each amino acid
The sensitivity of CNTPE-Cu-GOx was almost three [40]. Therefore, the incorporation of copper within CNTPE
times larger than in the absence of copper, although the have allowed us the detection of amino acids, electroactive
linear range was more restricted. The electrode demon- or not, at very low potentials and at physiological pHs.
strated to be highly stable since after 60 days at 4 8C the Based on these properties, the electrode was also used for
response was almost the same as the first day. the square-wave voltammetric-detection of albumin. Fig-
ure 10 shows square-wave voltammograms obtained for
10.0 mg/mL albumin at CNTPE (dotted line) and CNTPE-
3.6.2. Aminoacids and Albumin Sensor
Cu (solid line). At the composite without metal, there is only
The incorporation of copper into the carbon nanotubes the direct oxidation of tyrosine and triptophan residues at
composite was also used to design amino acids and albumin around 0.7 V. In the presence of copper, besides the direct
sensors based on the facilitated copper oxidation due to the amino acids oxidation, there is a huge peak at  0.100 V due
complex formation with the amino acid [39]. to the facilitated oxidation of copper in the presence of the
Figure 9 shows amperometric recordings at  0.100 V for amino acid residues. A linear relationship between peak
successive additions of 5.0  104 M l-histidine, l-lisine and current and concentration was obtained up to 25.0 mg/L
l-proline at CNTPE (a) and CNTPE-Cu (b). Since the albumin. The methodology was validated using the classical

Electroanalysis 19, 2007, No. 7-8, 823 – 831 www.electroanalysis.wiley-vch.de E 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
830 G. A. Rivas et al.

(CONICET), Secretara de Ciencia y Tecnologa de la

Universidad Nacional de Córdoba (SECyT) and Agencia
Nacional de Promoción Cientfica y Tecnológica for the
financial support. M. D. R. and M. L. P. thank CONICET;
and G. L. L. thanks ANPCyT for the fellowships received.
The authors also thank Prof. Manuel Chicharro, Esperanza
Bermejo, Alberto Sánchez Arribas and Antonio Zapardiel
for their helpful discussions and for the SEM experiments.

6. References
[1] S. Iijima, Nature 1991, 354, 56.
Fig. 10. Square-wave voltammograms for 10.0 mg/mL albumin
[2] H. Dai, Acc. Chem. Res. 2002, 35, 1035.
at CNTPE (dotted line) and at CNTPE-Cu (18.0%, w/w) (solid
[3] C. E. Banks, R. G. Compton, Analyst 2006, 131, 15.
line). Conditions: protein accumulation at  0.100 V for 10 min
[4] M. L. Cohen, Mater. Sci. Eng. C 2001, 15, 1.
followed by square-wave voltammogram between  0.200 and
[5] P. M. Ajayan, Chem. Rev. 1999, 99, 1787.
1.000 V. Frequency of 25 Hz, pulse amplitude of 25 mV, staircase
[6] P. M. Ajayan, T. W. Ebbesen, Rep. Prog. Phys. 1997, 60, 1025.
step 10 mV and scan rate of 0.250 V s1.
[7] H. Dai, Surface Science 2002, 500, 218.
[8] P. M. Ajayan, O. Z. Zhou, Applications of Carbon nanotubes
in Carbon nanotubes (Eds: G. Dresselhaus, Ph. Avouris)
spectrophotometric determination, with excellent correla-
Springer, Heidelberg, 2001, pp. 391 – 425.
tion. [9] P. M. Ajayan, T. W. Ebbesen, T. Ichihashi, S. Iijima, K.
Tanigaki, H. Hiura, Nature 1993, 362, 522.
[10] S. C. Tsang, Y. K. Chen, P. J. F. Harris, M. L. H. Green,
4. Conclusions Nature 1994, 372, 159.
[11] J. Wang, Electroanalysis 2005, 17, 7.
CNTPE combines the advantages of composite materials [12] M. Valcarcel, B. M. Simonet, S. Cardenas, E. B. Suarez, Anal.
Bioanal. Chem. 2005, 382, 1783.
with the unique electrocatalytic properties of carbon nano- [13] M. Musameh, J. Wang, A. Merkoci, Y. Lin, Electrochem.
tubes. The increased reactivity of carbon nanotubes has Commun. 2002, 4, 743.
allowed the fast and highly sensitive detection of different [14] J. Wang, A.-N. Kawde, M. Musameh, Analyst 2003, 128, 912.
analytes like hydrogen peroxide, ascorbic acid, uric acid, [15] X. X. Yan, D.-W. Pang, Z.-X. Lu, J.-Q. Li, H. Tong, J.
dopamine, dopac, NADH and hydroquinone. Trace levels of Electroanal. Chem. 2004, 1, 47.
nucleic acids could be also detected by adsorptive stripping [16] J. Wang, M. Musameh, Y. Lin, J. Am. Chem. Soc. 2003, 125,
2408.
at CNTPE.
[17] G. A. Rivas, S. A. Miscoria, J. Desbrieres, G. Barrera, Talanta
The feasibility to incorporate enzymes and cofactors into 2007, 71, 270.
the composite matrix has paved the way for the construction [18] M. Zhang, A. Smith, W. Gorski, Anal. Chem. 2004, 76, 5045.
of different enzymatic electrochemical biosensors. For in- [19] J. Wang, M. Musameh, Anal. Chem. 2003, 75, 2075.
stance, the important decrease in the reduction overpotential [20] P. J. Britto, K. S. V. Santhanam, P. M. Ajayan, Bioelectro-
of hydrogen peroxide (400 mV) has allowed the highly chem. Bioenerg. 1996, 41, 121.
[21] M. D. Rubianes, G. A. Rivas, Electrochem. Commun. 2003, 5,
sensitive and selective glucose quantification based on the
689.
incorporation of GOx into the composite matrix without [22] M. L. Pedano, G. A. Rivas, Electrochem. Commun. 2004, 6,
need of redox mediators or anti-interferents membranes. The 10.
CNTPE has also offered an attractive chronopotentiometric [23] M. D. Rubianes, G. A. Rivas, Electroanalysis 2005, 17, 73.
detection of metals, like copper, associated with very short [24] F. Valentini, A. Amine, S. Orlanducci, M. L. Terranova, G.
accumulation times [21], expanding, in this way, the analytical Palleschi, Anal. Chem. 2003, 75, 5413.
application of this composite electrode. [25] J. Wang, M. Musameh, Analyst 2004, 129, 1.
[26] N. S. Lawrence, R. P. Deo, J. Wang, Talanta 2004, 63, 443.
The incorporation of copper into the CNTPE has allowed [27] M. Chicharro, A. Sánchez, E. Bermejo, A. Zapardiel, M. D.
the highly sensitive detection of electroactive and non Rubianes, G. A. Rivas, Anal. Chim. Acta 2005, 543, 84.
electroactive amino acids and albumin at very low potentials [28] Z. Wang, Y. Wang, G. A. Luo, Analyst 2002, 127, 1353.
and at physiological pH. [29] M. Chicharro, E. Bermejo, M. Moreno, A. Sánchez, A.
In summary, this new composite material represents a Zapardiel, G. A. Rivas, Electroanalysis 2005, 17, 476.
very good alternative for designing novel electrochemical [30] M. C. Rodrguez, G. A. Rivas, Anal. Chim. Acta 2002, 459,
43.
sensors with a large number of analytical applications.
[31] P. J. Britto, K. S. V. Santhanam, V. Alonso, A. Rubio, P. M.
Ajayan, Adv. Mater. 1999, 11, 154.
[32] T. W. Odom, J.-L. Huang, P. Kim, C. M. Lieber, J. Phys.
5. Acknowledgements Chem. B 2000, 104, 2794.
[33] C. Hua, S. Walsh, M. Smyth, I. Švancara, K. Vytřas,
The authors thank Fundación Antorchas, Consejo Nacional Electroanalysis 1992, 4, 107.
de Investigaciones Cientficas y Técnicas de Argentina

Electroanalysis 19, 2007, No. 7-8, 823 – 831 www.electroanalysis.wiley-vch.de E 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Carbon Nanotubes Paste Electrodes 831

[34] O. Zhou, H. Shimoda, B. Gao, S. Oh, L. Fleming, G. Yue, [38] G. L. Luque, N. F. Ferreyra, G. A. Rivas, Microchim. Acta
Acc. Chem. Res. 2002, 35, 1045. 2006, 152, 277.
[35] C. G. Hu, W. L. Wang, S. X. Wang, W. Zhu, Y. Li., Diamond [39] J.-M. Zen, C.-T. Hsu, A. S. Kumar, H.-J. Lyu, K.-Y. Lin,
Rel. Mater. 2003, 12, 1295. Analyst 2004, 129, 841.
[36] Methods in Enzymology, Vol. XLIV (Ed: K. Mosbach), [40] G. L. Luque, N. F. Ferreyra, G. A. Rivas, Talanta 2007, 71,
Academic Press, London 1976. 1282.
[37] M. C. Rodrguez, G. A. Rivas, Electroanalysis 2001, 13, 1179.

Electroanalysis 19, 2007, No. 7-8, 823 – 831 www.electroanalysis.wiley-vch.de E 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim