You are on page 1of 7

中国科技论文在线 http://www.paper.edu.

cn
Talanta 71 (2007) 1096–1102

Magnetic loading of carbon nanotube/nano-Fe3O4


composite for electrochemical sensing
Song Qu a , Joseph Wang b , Jilie Kong a ,
Pengyuan Yang a , Gang Chen a,∗
a Department of Chemistry, Fudan University, Shanghai 200433, China
b Departments of Chemical & Materials Engineering and Chemistry, Arizona State University,
Tempe, AZ 85287-5001, USA
Received 19 April 2006; received in revised form 31 May 2006; accepted 1 June 2006
Available online 10 July 2006

Abstract
An electrochemical sensing platform was developed based on the magnetic loading of carbon nanotube (CNT)/nano-Fe3 O4 composite on
electrodes. To demonstrate the concept, nano-Fe3 O4 was deposited by the chemical coprecipitation of Fe2+ and Fe3+ in the presence of CNTs in an
alkaline solution. The resulting magnetic nanocomposite brings new capabilities for electrochemical devices by combining the advantages of CNT
and nano-Fe3 O4 and provides an alternative way for loading CNT on electrodes. The fabrication and the performances of the magnetic nanocomposite
modified electrodes have been described. Cyclic voltammetry (CV) and constant potential measurement indicated that the incorporated CNT
exhibited higher electrocatalytic activity toward the redox processes of hydrogen peroxide. In addition, chitosan (CTS) has also been introduced
into the bulk of the CNT/nano-Fe3 O4 composite by coprecipitation to immobilize glucose oxidase (GOx) for sensing glucose. The marked
electrocatalytic activity toward hydrogen peroxide permits effective low-potential amperometric biosensing of glucose, in connection with the
incorporation of GOx into CNT/Fe3 O4 /CTS composite. The accelerated electron transfer is coupled with surface renewability. TEM images and
XRDs offer insights into the nature of the magnetic composites. The concept of the magnetic loading of CNT nanocomposites indicates great
promise for creating CNT-based biosensing devices and expands the scope of CNT-based electrochemical devices.
© 2006 Elsevier B.V. All rights reserved.

Keywords: Carbon nanotube; Magnetic loading; Electrochemical sensing; Composite; Magnetite; Hydrogen peroxide; Glucose

1. Introduction powder microelectrode [12], CNT/Teflon composite electrode


[13], CNT screen-printed electrode [14], CNT/epoxy compos-
Since Iijima reported the existence of carbon nanotubes in ite electrode [15], CNT/copper composite electrode [16], etc.
1991 [1], they have attracted much attention because of their To modify CNT on the surface of electrodes, CNTs were usu-
high electrical conductivity, mechanical strength and chemi- ally dispersed in solvents [17,18] or polymer solutions (such as
cal stability [2,3]. The unique properties of carbon nanotube Nafion [19] and chitosan (CTS) [20] solutions) that were cast on
(CNT) make them extremely attractive for electrochemical sen- the electrodes because CNTs were insoluble in most solvents.
sors and biosensors [4–6]. Recent investigations demonstrated Recently, methods based on electrochemical deposition [21] and
that CNT shows strong electrocatalytic activity and the mini- electrochemical polymerization [22,23] have been employed for
mization of surface fouling when it was employed to improve loading CNT on electrodes. CNT-based electrodes have been
the electrochemical response of some important bioactive sub- demonstrated to reduce the overpotential of hydrogen peroxide
stances [7–10]. A variety of carbon nanotube electrodes have significantly [4,21]. The ability of carbon nanotubes to promote
been fabricated for sensing different redox compounds, includ- the electron-transfer reactions of hydrogen peroxide suggests
ing CNT modified screen-printed carbon electrode [11], CNT great promise for oxidase-based amperometric biosensors [4].
However, the selectivity and stability of the electrochemical
determination of hydrogen peroxide were usually inadequate
∗ Corresponding author. Tel.: +86 21 6466 1130; fax: +86 21 6564 1740. at conventional electrodes because there is a large overpoten-
E-mail address: gangchen@fudan.edu.cn (G. Chen). tial for the direct oxidation of hydrogen peroxide. A traditional

0039-9140/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.talanta.2006.06.003

转载
中国科技论文在线 http://www.paper.edu.cn
S. Qu et al. / Talanta 71 (2007) 1096–1102 1097

approach to lowering the oxidation potential of hydrogen per- acetic acid while GOx solution was made by dissolving 50 mg
oxide has relied on the use of redox mediators with low formal GOx in 10 mL of 50 mM phosphate buffer (pH 7.4). 5 mg/mL of
potentials [24]. glutaraldehyde solution was prepared by the appropriate dilution
Magnetic loading provides an alternative approach to load of 25 wt.% solution of glutaraldehyde using 50 mM phosphate
various materials on electrodes. Willner et al. have reported the buffer (pH 7.4).
magnetic control of enzymatic reactions using magnetic parti-
cles functionalized with redox relay units [25,26]. Ferrocene- 2.2. Apparatus
functionalized magnetic particles were thus used for electri-
cally contacting glucose oxidase (GOx) and the electrode. The Cyclic voltammetry (CV) and constant potential measure-
use of magnetic fields to align CNT [27] and to self-assemble ments were performed with a CHI 830B electrochemical ana-
CNT devices [28] has been reported. Recently, Musameh and lyzer (Shanghai Chen-Hua Instruments Co., Shanghai, China)
Wang have demonstrated that the magnetic and catalytic prop- in a three-electrode system consisted of a 2.2 mm diameter
erties of CNT can be exploited for the magnetoswitchable con- graphite-epoxy composite electrode with small magnets accom-
trol of electron-transfer reactions [29]. Herein, we report on modated inside, a platinum auxiliary electrode, and a saturated
a novel electrochemical-sensing platform for the detection of calomel electrode (SCE) as the reference electrode. XRD mea-
hydrogen peroxide based on the magnetic loading of carbon surements were carried out using a Rigaku D/max-rB diffrac-
nanotube/nano-Fe3 O4 (CNT/Fe3 O4 ) composite on electrodes. tometer (Rigaku, Tokyo, Japan) with Cu K␣1 radiation (40 kV,
The magnetic nanocomposite dispersed in water could be loaded 60 mA). The size and morphology of magnetic nanoparticles
onto the surface of magnetic electrode with the aid of the Fe3 O4 were observed by TEM using a JEOL model JEM-2011(HR)
nanoparticles. Although CNT/Fe3 O4 nanocomposites have been at 200 kV. The sample for TEM analysis was obtained by dip-
synthesized to adsorb heavy metal ions [30] and to minimize the ping a holy grid to nanoparticles-dispersed ethanol solution and
sedimentation of Fe3 O4 particles [31], there is no earlier report evaporated in air at room temperature.
on the magnetic loading of CNT/Fe3 O4 composite on electrodes
for electrochemical sensing. 2.3. Preparation of CNT/Nano-Fe3 O4 composites
In the present work, it has been demonstrated that mag-
netic CNT/nano-Fe3 O4 composite prepared by coprecipita- Further purification of the multiwalled carbon nanotubes was
tion exhibited excellent electrocatalytic ability in the reduction accomplished by stirring the CNT in concentrated nitric acid at
and oxidation of hydrogen peroxide. In addition, CNT/nano- 60 ◦ C for 12 h. Subsequently, the obtained CNT was dried at
Fe3 O4 /chitosan composite has also been prepared by copre- 110 ◦ C for 4 h. The CNT/Fe3 O4 nanoparticle composite was
cipitation to immobilize glucose oxidase covalently for sensing prepared by the in situ chemical coprecipitation of Fe2+ and
glucose. The attractive analytical performances of the magnetic Fe3+ in alkaline solution in the presence of CNTs. The molar
nanocomposite modified electrodes, along with the character- ratio of Fe2+ :Fe3+ was 1:2. The magnetic composite was pre-
ization of the nanocomposites, are illustrated in the following pared by suspending 1.0 g purified CNT in 200 mL of solution
sections. containing 1.7 g (4.33 mmol) (NH4 )2 Fe(SO4 )·6H2 O and 2.51 g
(8.66 mmol) NH4 Fe(SO4 )2 ·12H2 O at 50 ◦ C under N2 atmo-
2. Experimental sphere. After the solution was sonicated (30 W, 40 kHz) for
10 min, 10 mL of 8 M NH4 OH aqueous solution was added
2.1. Reagent and solutions dropwise to precipitate the iron oxides while the mixture solu-
tion was sonicated. The pH of the final mixture should be in
␤-d-Glucose and glucose oxidase (EC 1.1.3.4, from the range of 11–12. To promote the complete growth of the
Aspergillus niger, 158 units/mg) were obtained from Sigma (St. nanoparticle crystals, the reaction was carried out at 50 ◦ C for
Louis, MO, USA). Glutaraldehyde (25 wt.% solution in water) 30 min under constant mechanical stirring. The precipitate was
was the product of Merck (Germany). Chitosan (CTS, 75–85% isolated in the magnetic field, and the supernatant was separated
deacetaylation) was purchased from Aldrich (Milwaukee, WI, from the precipitate by decantation. Impurities (such as sulfate
USA). Multiwall carbon nanotubes (MWCNT, 30–60 nm diam- and ammonia) in the CNT/Fe3 O4 samples were removed by
eter, 5–15 ␮m long), with a purity of 95%, were provided by washing with copious amounts of doubly distilled water and the
Shenzhen Nanoport Company (Shenzhen, China). Ammonium precipitate was isolated by a permanent magnet. The obtained
ferrous sulfate and ammonium ferric sulfate were both pur- CNT/Fe3 O4 nanocomposite was then washed with 150 mL of
chased from SinoPharm (Shanghai, China). Other chemicals absolute alcohol for three times. Subsequently, the composite
were all analytical grade. was dried under vacuum. The content of CNT in the compos-
The supporting electrolyte for the electrochemical investi- ite was approximately 50 wt.%. If no CNT was added to the
gations was 50 mM phosphate buffer (pH 7.4). Stock solutions coprecipitation solution of Fe2+ :Fe3+ , pure Fe3 O4 nanoparticles
of 1 M glucose were allowed to mutarotate for 24 h before use could be obtained for the control experiments. The chemical
and subsequently kept in a 4 ◦ C refrigerator, while solutions of reaction of the Fe3 O4 precipitation occurring in the container is
hydrogen peroxide (1 M) were prepared by diluting a 30 wt.% as follows:
a stocking solution with doubly distilled water. CTS solution
(0.2 wt.%) was prepared by dissolving 0.4 g of CTS in 10 mg/mL 2Fe3+ + Fe2+ + 8NH3 ·H2 O = Fe3 O4 + 4H2 O + 8NH4 +
中国科技论文在线 http://www.paper.edu.cn
1098 S. Qu et al. / Talanta 71 (2007) 1096–1102

15 cm long, 50 ␮m diameter) was inserted into the glass tube


with the aid of the stainless-steel rod until the twisted end of
the copper wire was buried into the graphite-epoxy composite
inside the glass tube. Subsequently, the composite was allowed
to cure at room temperature for at least 12 h. Note that the cop-
per wire should be immobilized just beside the inner wall of the
glass tube so that NdFeB permanent magnets (Fig. 1(c), 1.5 mm
diameter, 4 mm height, Shanghai Heng-qiang Magnetic Mate-
rial Co. Ltd., Shanghai, China) could be taken out from the tube.
Prior to use, the magnetic microdisc electrode was successively
polished with emery paper and alumina powder, sonicated in
doubly distilled water. Fig. 1B is the photograph of the magnetic
electrode.

2.5. Magnetic loading

Before loading, 15 pieces of NdFeB permanent magnets men-


Fig. 1. Schematic diagram (A) and photograph (B) of the magnetic electrode and
tioned above were inserted into the glass tube of the electrode
(C) schematic of the magnetic loading process on the CNT–Fe3 O4 composite. until the magnet touched the inner surface of the graphite-epoxy
(a) Copper wire, (b) glass tube, (c) NdFeB magnet, (d) graphite-epoxy compos- composite electrode. 20 mg of the magnetic composites was dis-
ite electrode, (e) water drop with CNT/Fe3 O4 composite dispersed inside, (f) persed in 10 mL of water with the aid of sonication and will be
Plexiglas plate, (g) CNT/Fe3 O4 composite loaded on the electrode and (h) water referred to as the loading solution. And then, 10 ␮L aliquots
drop.
of this loading solution containing 20 ␮g CNT were taken out
with a pipette under sonication and were dropped on the surface
When the CNT, ammonium ferrous sulfate and ammonium fer- of a piece of clean Plexiglas plate. Glass plate could not work
ric sulfate involved in the above procedures were dissolved in well because it is hydrophilic. The amount of the composite
200 mL CTS solution (0.1 wt.%) instead of water, the obtained loaded on the magnetic electrode can be changed by controlling
precipitate was CNT/Fe3 O4 /CTS composite. The primary amino the volume of the loading solution on the Plexiglas plate. As
groups of CTS in the composite can facilitate the immobiliza- shown in Fig. 1C, the carbon disc part of the magnetic electrode
tion of enzymes such as GOx. A 20 mg of the CNT/Fe3 O4 /CTS was allowed to touch the top of the water drop (Fig. 1(e)). The
composite was dispersed in 2 mL of 5 mg/ml GOx solution by CNT/nano-Fe3 O4 composite would move toward the surface of
1 min sonication and the mixture was kept in a 4 ◦ C refriger- the electrode and aggregated there to form a modification layer
ator for 2 h. And then, the supernatant was removed from the (Fig. 1(g)). The magnets inside the electrode tube can prevent
precipitate by decanting with the aid of a magnet. The precip- the magnetic particles from escaping. When the magnets inside
itate was dispersed in 2 mL of 5 mg/mL of glutaraldehyde for tube were taken out, the magnetic modification layer could be
15 min to cross-link the primary amino groups of chitosan and removed by flushing with water or sonicating in water. Pure mag-
GOx. The magnetic enzyme material was washed with water and netite nanoparticles and CNT/Fe3 O4 /CTS/GOx could be loaded
immersed in 50 mM phosphate buffer (pH 7.4) at 4 ◦ C overnight on the surface of the electrode in the same procedures. Based on
to remove the reagents and the excess nonimmobilized GOx the CVs before and after loading CNT/nano-Fe3 O4 composite
from the composite. The obtained magnetic biosensing compos- (not shown), the modification layer can be removed completely
ite, CNT/Fe3 O4 /CTS/GOx, was dispersed in 10 mL of 50 mM by sonication when the magnets were taken out.
phosphate buffer (pH 7.4) and was kept in a 4 ◦ C refrigerator for
future use. 2.6. Procedures

2.4. Magnetic electrode fabrication The cyclic voltammetry measurements were carried out in the
desired potential range at a scan rate of 100 mV/s and no solu-
Fig. 1 illustrated schematic diagrams of the magnetic elec- tion agitation was necessary during the run. Measurements of the
trode. Epoxy resin and hardener (Zhejian Cixi Tiandong Adhe- amperometric response to the sample solution were performed
sive Co. Ltd., Ningbo, China) were mixed thoroughly at a weight at a desired potential while the solution was being magnetically
ratio of 2:1. Graphite powder (Aldrich, Wilwaukee, WI, USA) stirred at 300 rpm continuously. Because the supporting elec-
and epoxy resin/hardener were hand-mixed at a ratio of 1:1 trolytes were magnetically stirred, the magnetic electrode should
(w/w). The graphite-epoxy composite was subsequently packed be inserted into the electrolyte through a hole inside the rub-
into a glass tube (Fig. 1(b), 2.2 mm i.d., 5 mm o.d., 60 mm long) ber plug on a cylinder electrochemical cell (20 mm i.d., 50 mm
by pressing the end into a sample of the composite to a depth of height) to prevent it from vibrating. The surface of the magnetic
∼3 mm. A stainless-steel rod (1.5 mm diameter) was employed electrode was downwards and should be immersed into the solu-
to press the graphite-epoxy composite (Fig. 1(d)) inside the tube tion for no more than 5 mm so that the influence of the magnetic
to acquire a flat inner surface. A piece of copper wire (Fig. 1(a), stirrer could be minimized.
中国科技论文在线 http://www.paper.edu.cn
S. Qu et al. / Talanta 71 (2007) 1096–1102 1099

3. Results and discussion that the CNTs in the composite were coated with nano-Fe3 O4
that facilitated the magnetic loading of CNT on the surface of
In order to load the magnetic nanocomposite on the electrode, the electrodes.
a specially designed magnetic electrode has been prepared by Fig. 2C displays the X-ray diffraction (XRD) patterns of CNT
using NdFeB permanent magnets (Fig. 1). The magnets facili- and the CNT/nano-Fe3 O4 composite. Diffraction peaks assigned
tate the magnetic loading of the nanocomposite and can prevent to CNTs at 2θ = 26.5◦ [32] can be clearly seen for both pure
the composite from falling off in the solution. The magnetic CNT and CNT/nano-Fe3 O4 composite, indicating that the CNTs
electrode can carry a 0.5 g iron screw on its surface. When the structure was not destroyed after the chemical coprecipitation of
magnets were taken away or put back, the magnetic compos- Fe3 O4 . As shown in Fig. 2C, the other two weak peak of CNT
ite loaded on the electrode would fall off or could be reloaded, merged with that of Fe3 O4 . Six characteristic peaks for Fe3 O4
allowing a simple and fast method for renewing the modifica- (2θ = 30.2◦ , 35.6◦ , 43.3◦ , 53.7◦ , 57.3◦ and 62.8◦ ), marked by
tion layer on the electrodes. Transmission electron microscopy their indices ((2 2 0), (3 1 1), (4 0 0), (4 2 2), (5 1 1) and (4 4 0))
(TEM) was performed to observe the microstructure of the mag- [33], were observed for the CNT/nano-Fe3 O4 composite, indi-
netic composite powders. Fig. 2A illustrates the TEM image cating that the resultant Fe3 O4 particles in the composite were
of the CNT/nano-Fe3 O4 composite. The diameter of Fe3 O4 pure Fe3 O4 with a spinel structure.
nanoparticles in Fig. 2A is in the range of 10–30 nm. A higher The magnetic lodging of magnetic CNT composite permits
resolution TEM image of the CNT/nano-Fe3 O4 composite as a variety of manipulations, including the modification of elec-
shown in Fig. 2B illustrates the nature of the composite mate- trode surfaces and the fabrication of biosensors. The CNT/Fe3 O4
rial. Although the composites had been sonicated in ethanol association does not impair the electrocatalytic properties of
before TEM measurements, a great amount of Fe3 O4 nanopar- CNT. Such behavior has been demonstrated for the redox pro-
ticles were found on the surface of CNTs, indicating the strong cesses of hydrogen peroxide that are of considerable interest
interaction between CNT and nano-Fe3 O4 . It could be concluded for the operation of oxidase-based amperometric biosensors.

Fig. 2. (A and B) Typical transmission electron micrographs (TEMs) of the MWCNT/nano-Fe3 O4 composite and (C) XRD pattern of CNTs and the CNT/Fe3 O4
composites. (a) MWCNT and (b) Fe3 O4 nanoparticles.
中国科技论文在线 http://www.paper.edu.cn
1100 S. Qu et al. / Talanta 71 (2007) 1096–1102

Fig. 3 compares the cyclic voltammograms (CVs) of 0.05 M


phosphate buffer solution (pH 7.4) containing (a, b, e) 0, (f)
1, (g) 2, (c, d, h) 3, (i) 4 and (j) 5 mM hydrogen peroxide at
bare (a, c), Fe3 O4 nanoparticle modified (b, d) and CNT/nano-
Fe3 O4 composite modified (e–j) electrodes at a scan rate of
100 mV/s. The electrode loaded with CNT/nano-Fe3 O4 com-
posite exhibits significantly enhanced currents for the oxidation
and reduction of hydrogen peroxide starting around +0.45 and
−0.1 V (versus SCE), respectively. Thus, the CNT/Fe3 O4 mod-
ified electrode not only improves the redox currents but also
decreases the overvoltage potential for the oxidation and reduc-
tion of hydrogen peroxide, allowing convenient low-potential
amperometric detection. The catalytic activity can be demon-
strated by the well-defined peaks at +0.7 V (A) and −0.36 V
(B). Fig. 3A(e–j) illustrates that the addition of hydrogen per-
oxide to the supporting solution in the steps of 1 mM results in
a nearly linear increase in the peak currents for the oxidation of
hydrogen peroxide. In contrast, poor electrochemical response
was obtained at the surface of both the bare (c) and the Fe3 O4
(d) modified electrodes over the commonly used potential range.
The oxidation current of hydrogen peroxide starts to increase at
potentials that are higher than +0.85 V while the reduction cur-
rents increase at potentials that are lower than −0.3 V at both
CNT-free electrodes. It can be concluded that Fe3 O4 loaded on
the magnetic electrodes does not exhibit electrocatalytic activity
toward the redox processes of hydrogen peroxide and the elctro-
catalytic activity of CNT/Fe3 O4 loaded on the electrode mainly
comes from CNT. The equations for the electrooxidation and
electroreduction of H2 O2 over CNT can be illustrated as follow:

H2 O2 + 2H2 O − 2e → 2H3 O+ + O2
Fig. 3. Cyclic voltammograms of (a, b and e) 0, (f) 1, (g) 2, (c, d and h) 3, (i) 4,
(j) 5 mM hydrogen peroxide in 0.05 M phosphate buffer solution (pH 7.4) at the
bare (a and c), Fe3 O4 nanopartile modified (b and d) and CNT/Fe3 O4 composite H2 O2 + 2H3 O+ + 2e → 4H2 O
modified (e–j) electrodes in the potential range of 0 to 1.0 (A) and 0.3 to −0.5 V
(B). The amount of the loaded magnetic materials, 20 ␮g; scan rate, 100 mV/s. The amount of CNT/Fe3 O4 plays a crucial role in the electrocat-
alytic reduction of hydrogen peroxide. Fig. 4A shows the effect
of the amount of CNT/nano-Fe3 O4 composite on the amper-

Fig. 4. (A) Effect of the amount of CNT/nano-Fe3 O4 composite on the current response of 5 mM hydrogen peroxiode at potential of −0.1 V (vs. SCE) and (B)
hydrodynamic voltammograms (HDVs) for 1 mM hydrogen peroxide at a bare (a) and a CNT/nano-Fe3 O4 composite (b) electrode. The amount of the loaded
composite, 20 ␮g; electrolyte, 50 mM phosphate buffer (pH 7.4); stirring rate, 300 rpm.
中国科技论文在线 http://www.paper.edu.cn
S. Qu et al. / Talanta 71 (2007) 1096–1102 1101

Fig. 5. Amperometric response of the bare (a) and the CNT/Fe3 O4 composite modified (b) electrodes to the successive addition of 1 mM hydrogen peroxide at the
potential of (A) −0.1 and (B) +0.4 V (vs. SCE). (C) Amperometric response of the bare (a) and the magnetic electrode loaded with Fe3 O4 /CNT/CTS–GOx composite
(b) to the successive addition of 1 mM glucose at the potential of −0.1 V (vs. SCE). Also shown (in the insets) are the plots of amperometric currents versus the
concentrations of hydrogen peroxide (A and B) and glucose (C) at the electrodes modified with magnetic CNT composite. Other conditions, as in Fig. 4.

ometric current response of 5 mM hydrogen peroxide at the concentration in the steps of 1 mM at the bare (a) and the
constant potential of −0.1 V (versus SCE). The current increases CNT/Fe3 O4 coated (b) electrodes at the detection potential of
to a value of 16.8 ␮A, upon raising the nanocomposite amount +0.4 V. The greatly enhanced amperometric response and the
between 10 and 20 ␮g, and decreases gradually to 15.9 ␮A upon short response time (2–3 s) at the CNT-based device reflects
increasing the amount to 30 ␮g. The dropping of the current the accelerated oxidation of hydrogen peroxide, in accordance
implies that the excess composite may disturb the mass trans- with the HDV data illustrated in Fig. 4B. Although the oxida-
formation on the surface of the electrode. Fig. 4B compares the tion potential has been significantly reduced to 0.3–0.5 V by
hydrodynamic voltammograms (HDVs) for 1 mM hydrogen per- the CNT/nano-Fe3 O4 composite, some coexistent substances
oxide at the bare electrode (a) and the CNT/Fe3 O4 composite (such as dopamine, ascorbic acid, uric acid, etc.) in the body
modified electrode (b). It is obvious that the current response fluids at physiological levels may interfere with the determi-
for the oxidation and the reduction of hydrogen peroxide at nation because they are electroactive at the potentials above
the bare electrode was poor. A small response was observed 0.2 V. So the glucose biosensor based on the magnetic loading of
at more positive or more negative potentials for the oxidation the CNT/Fe3 O4 /CTS/GOx composite on the base electrode was
or reduction of hydrogen peroxide, respectively. In contrast, the allowed to run at potential of −0.1 V in reduction mode so that
CNT/Fe3 O4 composite modified electrode exhibited enhanced the interferences of dopamine, ascorbic acid and uric acid were
current response with significant decrease in the overvoltages. minimized. Fig. 5C illustrates the amperometric response of the
Such electrocatalytic activities facilitate the low-potential magnetic electrode loaded with Fe3 O4 /CNT/CTS/GOx compos-
amperometric measurements of hydrogen peroxide. Fig. 5A ite to the successive addition of 1 mM glucose. Well-defined
compares the amperometric response (at −0.1 V) of the bare (a) current responses for glucose were obtained at the magnetic
and the CNT/Fe3 O4 modified (b) electrodes to the successive biosensor. The reaction occurring at the biosensor is very fast to
additions of 1 mM hydrogen peroxide, along with the resulting reach a dynamic equilibrium upon each addition of the sample
calibration plots (insets). As expected from the voltammetric solution, generating a steady-state current signal within 5–6 s.
data, the bare and the Fe3 O4 modified electrodes are not respon- The high sensitivity is accompanied by a low noise level (despite
sive to these concentration changes using this low detection the large voltammetric background current). In order to test the
potential. However, well-defined and fast amperometric signals selectivity of the magnetic biosensor, amperometric responses
are observed for these 1 mM-H2 O2 additions at the CNT/Fe3 O4 were obtained by a batch addition of interfering species (0.2 mM
modified electrode, generating steady-state signals within 2–3 s. ascorbic acid, 0.2 mM uric acid and 0.2 mM dopamine) after the
The CNT-based magnetic electrode offers substantially larger addition of 5 mM glucose. The interferences of dopamine, ascor-
signals, reflecting the electrocatalytic activity of CNT (b). In bic acid and uric acid, have not been found because they are not
addition, the hydrogen peroxide response was not affected by electroactive at the detection potential of −0.1 V. Because chi-
regenerating the surface of the electrode. Such renewability tosan dissolved in the acidic solution can be precipitated in alka-
was indicated from a series of seven successive measurements line solution, CTS was coprecipapted with Fe3+ and Fe2+ in the
each recorded on a freshly loaded surface, yielding satisfactory presence of CNT to form magnetic CNT/Fe3 O4 /CTS composite,
reproducible signals with relative standard derivation (R.S.D.) a versatile biosensing material. The CNTs inside the composite
of 5.3% (not shown). form a conductive network that can accommodate Fe3 O4 , CTS
The CNT/nano-Fe3 O4 coated electrodes exhibit highly elec- and other functional components. The primary amino groups in
trocatalytic activity toward not only the reduction but also the CTS facilitate the further immobilization of enzyme, DNA or
oxidation of hydrogen peroxide. Fig. 5B illustrates the current- proteins, while the magnetic moiety of the composite provides
time recordings obtained by increasing the hydrogen peroxide conveniences for magnetic loading and magnetic isolation.
中国科技论文在线 http://www.paper.edu.cn
1102 S. Qu et al. / Talanta 71 (2007) 1096–1102

4. Conclusions [5] R.S. Chen, W.H. Huang, H. Tong, Z.L. Wang, J.K. Cheng, Anal. Chem. 75
(2003) 6341–6345.
[6] J. Wang, M. Musameh, Anal. Chem. 75 (2003) 2075–2079.
It has been demonstrated that CNT/nano-Fe3 O4 composite
[7] Z.H. Wang, J. Liu, Q.L. Liang, Y.M. Wang, G. Luo, Analyst 127 (2002)
can be magnetically loaded on electrodes for electrochemical 653–658.
sensing, providing an alternative rapid way for creating CNT- [8] G. Zhao, S.Q. Zang, K.Z. Liu, S. Lin, J. Liang, X.Y. Guo, Z.J. Zhang, Anal.
based devices for electrochemical sensing by combining the Lett. 35 (2002) 2233–2244.
advantages of CNT and nano-Fe3 O4 . Various chemical and bio- [9] J. Wang, M. Musameh, Analyst 128 (2003) 1382–1385.
[10] M. Musameh, J. Wang, A. Merkoci, Y.H. Lin, Electrochem. Commun. 4
logical components could be readily incorporated into the bulk
(2002) 743–746.
of the magnetic nanocomposite by coprecipitation and cross- [11] J. Wang, G. Chen, M.P. Chatrathi, M. Musameh, Anal. Chem. 76 (2004)
linking. It has been demonstrated that the CNT/nano-Fe3 O4 298–302.
composite loaded on the electrode with the aid of magnets [12] Y.D. Zhao, W.D. Zhang, H. Chen, Q.M. Luo, Talanta 58 (2002) 529–
offered an electrocatalytic platform for electrochemical sensing. 534.
[13] J. Wang, M. Musameh, Anal. Chem. 75 (2003) 2075–2079.
The magetic nanocomposite can be removed upon taking away
[14] J. Wang, M. Musameh, Analyst 129 (2004) 1–2.
the magnets inside the electrodes, providing a useful avenue for [15] G. Chen, L.Y. Zhang, W. Wang, Talanta 64 (2004) 1018–1023.
preparing renewable CNT-based biosensors and CNT modified [16] J. Wang, G. Chen, M. Wang, M.P. Chatrathi, Analyst 129 (2004) 512–515.
electrodes. The biosensor fabrication technology demonstrated [17] H. Luo, Z. Shi, N. Li, Z. Gu, Q. Zhuang, Anal. Chem. 73 (2001) 915–920.
in this work is readily applicable to the fabrication of other [18] F.H. Wu, G.C. Zhao, X.W. Wei, Electrochem. Commun. 4 (2002) 690–694.
[19] J. Wang, M. Musameh, Y. Lin, J. Am. Chem. Soc. 125 (2003) 2408–2409.
biosensors based on oxidases, such as biosensors for cholesterol,
[20] M. Zhang, A. Smith, W. Gorski, Anal. Chem. 76 (2004) 5045–5050.
alcohol, lactate, acetylcholine, choline, hypoxanthine, xanthine, [21] X.L. Luo, J.J. Xu, J.L. Wang, H.Y. Chen, Chem. Commun. 16 (2005)
etc. In addition, other carbon materials such as graphite powder, 2169–2171.
activated carbon or carbon black can also be coated with mag- [22] M.L. Guo, J.H. Chen, J. Li, B. Tao, S.Z. Yao, Anal. Chim. Acta 532 (2005)
netic nanoparticles by the procedures developed in this work for 71–77.
[23] M. Gao, S. Huang, L. Dai, G. Wallace, R. Gao, Z. Wang, Angew. Chem.
electrochemical sensing, solid-phase microextraction, etc.
Int. Ed. 39 (2000) 3664–3667.
[24] J. Wang, Electroanalysis 13 (2001) 983–988.
Acknowledgments [25] R. Hirsch, E. Katz, J. Wasserman, I. Willner, J. Am. Chem. Soc. 122 (2000)
12053–12054.
[26] E. Katz, I. Willner, Electrochem. Commun. 4 (2001) 201–204.
This work was supported by Shanghai Science Committee [27] M. Fujiwara, E. Oki, M. Hamada, Y. Tanimoto, I. Mukouda, Y. Shimomura,
(051107089 and 2004ZR140150212), NSFC (20405002), State J. Phys. Chem. A 105 (2001) 4383–4386.
Education Ministry of China and the 863 program of China. [28] D. Long, J. Lazorcik, R. Shahidhar, Adv. Mater. 16 (2004) 814–819.
[29] M. Musameh, J. Wang, Langmuir 21 (2005) 8565–8568.
[30] X.J. Peng, Z.K. Luan, Z.C. Di, Z.G. Zhang, C.L. Zhu, Carbon 43 (2005)
References 855–894.
[31] H.T. Pu, F.J. Jiang, Nanotechnology 16 (2005) 1486–1489.
[1] S. Iijima, Nature 354 (1991) 56–58. [32] W.Z. Zhu, D.E. Miser, W.G. Chan, M.R. Hajaligol, Mater. Chem. Phys. 82
[2] R.H. Baughman, A. Zakhidov, W.A. de Heer, Science 297 (2002) 787–792. (2003) 638–647.
[3] P.M. Ajayan, Chem. Rev. 99 (1999) 1787–1799. [33] D.H. Chen, M.H. Liao, J. Mol. Catal. B Enzyme 16 (2002) 283–
[4] J. Wang, Electroanalysis 17 (2005) 7–14. 291.

You might also like