Download Encyclopedia of Nanoscience and Nanotechnology

Biosensors Based on Carbon Nanotubes
Yuehe Lin
Wassana Yantasee
Fang Lu
Pacific Northwest National Laboratory, Richland, Washington, U.S.A.
Joseph Wang
Mustafa Musameh
New Mexico State University, Las Cruces, New Mexico, U.S.A.
Yi Tu
Zhifeng Ren
Boston College, Chestnut Hill, Massachusetts, U.S.A.
INTRODUCTION
Carbon nanotube (CNT) is an attractive material for the
development of biosensors because of its capability to
provide strong electrocatalytic activity and minimize
surface fouling of the sensors. This article reviews the
recent successful development of biosensors based on
CNT materials. Specifically, biosensors from two fabrication regimes have been investigated: 1) the coimmobilization of CNTs and enzymes on electrode surfaces and
2) the growth of controlled-density aligned CNTs for the
fabrication of nanoelectrode arrays. In the first regime, the
CNTs are either dispersed in solvents [e.g., sulfuric acid,
dimethylformamide (DMF)], dissolved in Nafion solution
for electrode coating, or mixed with Teflon as an electrode
material for reagentless biosensors. In the second regime,
the nanoelectrode arrays consisting of millions of vertically aligned CNTs, each acting as an individual electrode, have been fabricated through a nonlithographic
method. We also demonstrate the capability of CNTs to
promote the oxidation/reduction (redox) reactions of
hydrogen peroxide and nicotinamide adenine dinucleotide
(NADH), which are involved in a wide range of amperometric biosensors associated with oxidase and dehydrogenase enzymes, respectively. With these electrocatalytic
properties of CNTs, the applications of CNT-based biosensors examined in our laboratories include the lowpotential detections of glucoses, organophosphorous compounds, and alcohol.
were first introduced in 1991.[22] CNTs are distinguished
according to their structural properties:[23] a single-wall
CNT (SWCNT) consists of a single graphitic sheet rolled
into a cylinder (with 1 to 2 nm o.d. and several microns in
length), and multiwall CNT (MWCNT) consists of
graphitic sheets rolled into closed concentric tubes (with
50 nm o.d. and microns in length), each separated by van
der Waals forces to have a gap of 3.4 Å. Carbon nanotubes
have been known to promote electron-transfer reactions
of cytochrome c,[2] NADH,[1,6] catecholamine neurotransmitters,[3] and ascorbic acid.[4] This is attributed to
their electronic structure, high electrical conductivity, and
redox active sites. Carbon is also a versatile electrode material that can undergo various chemical and electrochemical
modifications to produce suitable surfaces for high electrode responses. Carbon electrodes have a wide useful
potential range, especially in the positive direction,
because of the slow kinetics of carbon oxidation. These
excellent properties of CNTs have been successfully exploited in our laboratories to develop either amperometric
biosensors based on the immobilization of CNTs,[1,5]
coimmobilization of CNTs and enzymes,[6,8] or the growth
of controlled-density aligned CNTs into the nanoelectrode
arrays[7,9,10] for the detection of glucoses,[5,6,10] organophosphorous compounds,[8] and alcohol,[6] as summarized
in the following sections.
BIOSENSORS BASED ON
IMMOBILIZATION OF CNTs
BACKGROUND
Solubilization and Immobilization of CNTs
There has been enormous interest in exploiting CNTs in
electrochemical and biological sensors[1–21] since they
CNTs have shown potential for applications in chemical/
biological sensors and nanoscale electronic devices. A
Dekker Encyclopedia of Nanoscience and Nanotechnology
DOI: 10.1081/E-ENN 120025031
Copyright D 2004 by Marcel Dekker, Inc. All rights reserved.
Copyright © 2004 by Marcel Dekker, Inc. All rights reserved.
B
major barrier for developing such CNT-based devices is
the insolubility of CNTs in most solvents. This challenge
has been addressed through covalent modification[24,25]
or noncovalent functionalization[26,27] of the CNTs. A
‘‘wrapping’’ of CNT in polymeric chains (i.e., poly(pphenyl-enevinylene)[26] or poly{(m-phenylenevinylene)co-[2,5-dioctyloxy-(p-phenylene)-vinylene]}[27]) has improved the solubility of CNTs without impairing their
physical properties.[28] In our work,[5] a well-known
perfluorosulfonated polymer, Nafion, has been used to
solubilize single-wall and multiwall CNTs. Because of
their unique ion-exchange, discriminative, and biocompatibility properties, Nafion films have been used extensively to modify electrode surfaces and to construct
amperometric biosensors.[29,30] Similar to other polymers
used to wrap and solubilize CNTs, Nafion bears a polar
side chain. We have found that CNTs can be suspended
in solutions of Nafion in phosphate buffer or alcohol.
Increasing the Nafion content from 0.1 to 5 weight percent (wt.%) results in a dramatic enhancement of the
solubility of both single-wall and multiwall CNTs, which
can be observed by the naked eye. A homogeneous
solution of the Nafion/CNT complex is observed in Nafion solution, but no such solubilization is observed in
ethanol or phosphate-buffer solutions containing no Nafion. The CNT/Nafion association does not impair the
electrocatalytic properties of CNTs with respect to the
redox reaction of hydrogen peroxide. The Nafion-induced solubilization of CNT thus permits a variety of
applications, including the modification of electrode surfaces for preparing amperometric biosensors.
Because CNTs are insoluble in most solvents, previously reported CNT-modified electrodes have relied on
casting a CNT/sulfuric acid solution onto a surface of
electrodes,[1,31] a procedure that is not compatible with the
immobilization of biocomponents. Therefore we have
developed a new and simple method for preparing effective CNT-based biosensors from CNT/Teflon composite material.[6] Carbon composites, based on the dispersion
of graphite powder within an insulator, offer convenient
bulk modification for the preparation of reagentless and
renewable biosensors. Teflon has been used as a binder for
graphite particles for various electrochemical-sensing applications.[32,33] Our approach relies on CNTs as the sole
conductive component rather than as the modifier cast on
other electrode surfaces. The bulk of CNT/Teflon composites hence serve as a reservoir for the enzymes, in the
same manner as their graphite-based counterparts. The
preparation is very simple: a certain amount of CNTs is
hand-mixed in the dry-state with granular Teflon to obtain
a desired composition of CNT/Teflon. The CNT content
of the new composites has a large effect upon their electrochemical behavior. Fig. 1a compares the calibration
plots for potassium ferricyanide obtained at electrodes
Copyright © 2004 by Marcel Dekker, Inc. All rights reserved.
containing different MWCNT loadings. All electrode
compositions yield highly linear calibration plots over the
entire concentration range. The sensitivity increases with
the CNT loading between 10 and 60 wt.% and decreases
thereafter. The influence of the CNT content was also
examined using cyclic voltammetry (CV) experiments.
The CV ferricyanide cathodic and anodic currents increased linearly with the CNT content between 30 and
70 wt.% (Fig. 1b). Composites containing more than 70
wt.% CNTs are too dry and porous and have poor
mechanical stability. Electrode resistance leading to a
poorly defined and negligible CV is observed at the electrodes containing low CNT content. Fig. 1b shows the
sensitivity-loading (B) and resistance-loading (A) profiles.
Too high Teflon (< 30 wt.% CNTs) leads to high resistance to a nearly insulating matrix and low sensitivity,
while too high CNT (> 75 wt.% CNT) leads to an
operation beyond the mass-limiting plateau associated
Fig. 1 (a) Calibration plots for potassium ferricyanide using
electrodes with varying MWCNT/Teflon ratio (by weight) of
10:90 (A), 30:70 (B), 40:60 (C), 60:40 (D), and 80:20 (E).
Operating conditions: potential, + 0.1 V; supporting electrolyte,
phosphate buffer (0.05 M, pH 7.4); stirring rate, 400 rpm. (b)
Influence of the CNT loading upon the electrode resistance (A)
and amperometric response to 1 mM ferricyanide (B) measured
with the 60:40 wt.% MWCNT/Teflon electrode. Other conditions are as in (a). (From Ref. 6.)
with the shift of the voltammetric signal. Therefore a CNT
content of 40 to 60 wt.% is suggested.
The CNT/Teflon composites have the combined
advantages of CNTs and bulk composite electrodes that
permit a wide range of applications without the need for a
graphite surface. Certain amounts of enzymes [e.g., glucose oxidase (GOx) and alcohol dehydrogenase (ADH)]
and cofactor (e.g., NAD+) can be mixed with the CNT/
Teflon composite and used as electrode materials,
depending upon specific needs. The CNT/Teflon coating
was later investigated in our laboratory and displayed a
marked electrocatalytic action toward hydrogen peroxide
and NADH and hence is promising for the development of
biosensors for glucose (in connection with oxidase
enzymes) and ethanol (in connection with dehydrogenase
enzymes), respectively.
B
Fig. 3 Hydrodynamic voltammograms for 1 mM hydrogen
peroxide at the 60:40 wt.% graphite/Teflon electrode (a) and the
60:40 wt.% MWCNT/Teflon electrode (b). Other conditions are
as in Fig. 1a. (From Ref. 6.)
Electrocatalytic Activity of CNTs to Redox
Reactions of Hydrogen Peroxide
Hydrogen peroxide is involved in a wide range of biosensing applications associated with oxidase enzymes. In
our laboratories, the enhancement of the redox activity of
hydrogen peroxide by CNTs is investigated using the
previously described CNT/Nafion[5] and CNT/Teflon[6]
electrode materials.
To study the electrocatalytic activity of CNTs for
hydrogen peroxide oxidation/reduction, the CNT/Nafion
(from a 0.5 wt.% Nafion solution containing 2 mg/mL
of CNTs) is coated on a glassy carbon (GC) electrode
Fig. 2 Cyclic voltammograms of 5 10 3 M hydrogen peroxide at unmodified (A) and MWCNT/Nafion-modified (B) GC
electrodes. Operating conditions: scan rate, 50 mV/sec;
electrolyte, phosphate buffer (0.05 M, pH 7.4). Upper inset
shows the amperometric response after increasing hydrogen
peroxide concentration for 1 10 3 M incrementally while
potential is held at 0.0 V. (From Ref. 5.)
Copyright © 2004 by Marcel Dekker, Inc. All rights reserved.
surface. Fig. 2 displays the cyclic voltammograms for
5 10 3 M hydrogen peroxide recorded at a bare GC
electrode (A) and the CNT/Nafion-modified GC electrode (B). Significant oxidation and reduction currents
starting around + 0.20 V are observed on the CNT/
Nafion-coated electrode, while none is observed at the
bare GC electrode. The CNT/Nafion-coated electrode
offers a marked decrease in the overvoltage for hydrogen peroxide reaction and hence allows low-potential amperometric detection. The inset of Fig. 2 shows
the amperometric response at 0.0 V to successive additions of hydrogen peroxide. While the modified
electrode (B) responds very rapidly and favorably to
the changes in hydrogen peroxide concentration, no
response is observed at the bare GC electrode (A). A
similar decrease in hydrogen peroxide overvoltage is
observed at other CNT-modified electrodes (not
shown), indicating that Nafion does not impair the
electrocatalytic properties of CNT. The CNT/Nafioncoated electrode is reliable and not affected by
regenerating the surface; six successive measurements
of hydrogen peroxide, each recorded on a freshly
polished surface, show reproducible results with %RSD
of less than 4.
Similar substantial lowering of the detection potential
and significantly improved current signals for hydrogen
peroxide by CNTs have also been found at CNT/Teflon
electrodes, obtained by packing 60/40 wt.% of CNT/
Teflon into the electrode cavity of a glass sleeve with
a copper wire as the electrical contact. Control experiments, using graphite-based Teflon composites, were
performed in parallel. Fig. 3 compares the hydrodynamic voltammograms (HDVs) for 1 mM hydrogen
Fig. 4 (a) Cyclic voltammograms for 5 10 3 M NADH at unmodified (A), MWCNT-modified (B), and SWCNT-modified (C) GC
electrodes. Operating conditions: scan rate, 50 mV/sec; electrolyte, phosphate buffer (0.05 M, pH 7.4). Dotted lines represent the
background response. (b) Hydrodynamic voltammograms for 1 10 4 M NADH at the unmodified (A) and the MWCNT-modified (B)
GC electrodes. Operating conditions: stirring rate, 500 rpm; electrolyte, phosphate buffer (0.05 M, pH 7.4). (c) Current–time recordings
obtained after increasing the NADH concentration of 1 10 4 M (each step) at unmodified (A) and MWCNT-modified (B) GC
electrodes. Inset shows the corresponding calibration curve. Operating conditions: potential, + 0.3 V; others as in (b). (From Ref. 1.)
Copyright © 2004 by Marcel Dekker, Inc. All rights reserved.
peroxide at the graphite/Teflon (a) and CNT/Teflon (b)
electrodes. Compared to the graphite/Teflon electrode,
the CNT/Teflon electrode responds more favorably to
hydrogen peroxide over the entire potential range (0.0
to 1.0 V) with significant response starting at + 0.20 V.
The Teflon binder is proven to not impair the electrocatalytic properties of CNTs.
B
Electrocatalytic Activity of CNTs
to Redox Reactions of NADH
b-Nicotinamide adenine dinucleotide (NADH) is a cofactor in several hundred enzymatic reactions of NAD+/
NADH-dependent dehydrogenases. The electrochemical
oxidation of NADH has thus been the subject of numerous
studies related to the development of amperometric
biosensors.[34] Problems inherent to such anodic detection
are the large overvoltage encountered for NADH oxidation at ordinary electrodes[35] and surface fouling associated with the accumulation of reaction products.[36] CNTs
have thus been used in our recent work[1,6] as the new
electrode material to reduce the overpotential for NADH
oxidation and to alleviate surface fouling problems.
In the previous study,[1] single-wall and multiwall
CNTs were dispersed in concentrated sulfuric acid, and
each was subsequently cast on a glassy carbon electrode.
Fig. 4a shows the cyclic voltammograms of NADH measured at unmodified (A), MWCNT-modified (B), and
SWCNT-modified (C) glassy carbon electrodes. Both
modified electrodes yield an approximately twofold larger
NADH peak, compared to the unmodified electrode. The
oxygen-rich groups on the CNT surface, introduced during the acid dispersion, are perhaps responsible for such
electrocatalytic behavior for the oxidation of NADH.
Fig. 4b shows a HDV of 1 10 4 M NADH, which
reflects the electrocatalytic behavior of the CNT coating
with varying potentials. The MWCNT-coated electrode
(B) responds to NADH over the entire 0.0- to 1.0-V range,
while the bare electrode (A) responds only at potentials
higher than + 0.6 V. Fig. 4c shows that successive additions of 1 10 4 M NADH result in increasing response
detected at the CNT-modified electrode (B) but no response
at the unmodified electrode (A) when the detection potential was kept low (i.e., 0.3 V). Evidently, the electrocatalytic action of CNT enables the fast response (i.e.,
10 sec to reach the steady state) to the change of NADH
concentrations at the low-detection potential. The amperometric response of 5 10 3 M NADH appears to be
very stable; the decay of the signal is less than 10% and
25% after a 60-min period at the MWCNT-modified and
SWCNT-modified electrodes, compared with 75% and
53% at the graphite-coated and acid-treated electrodes,
respectively. This shows the capability of CNTs in resist-
Copyright © 2004 by Marcel Dekker, Inc. All rights reserved.
Fig. 5 Hydrodynamic voltammograms for 1 mM NADH at
the 60:40 wt.% graphite/Teflon electrode (a) and the 60:
40 wt.% MWCNT/Teflon electrode (b). Other conditions are as
in Fig. 1a. (From Ref. 6.)
ing the fouling effects and in preventing the diminishing of
signals in successive cyclic voltammetric detections. The
resistance to fouling of CNT-based electrodes has yet to
be understood.
In our more recent study,[6] the electrocatalytic effect
of CNTs to facilitate low-potential amperometric measurements of NADH has been investigated using CNT/
Teflon electrodes. Fig. 5 compares the amperometric
response (at +0.40 V) of the graphite/Teflon (a) and
CNT/Teflon (b) electrodes to successive additions of
0.1 mM NADH. Only the CNT/Teflon electrode
responds very rapidly to the changes in the level of
NADH, producing steady-state signals within 8 to 10
sec. The favorable signals are accompanied by a low
noise level.
The CNT coating offers remarkably decreased overvoltage for the NADH oxidation as well as reduced
surface fouling effects of the electrodes. These characteristics indicate the great promise of CNTs for developing
highly sensitive, low-potential, and stable amperometric
biosensors based on dehydrogenase enzymes.
Applications of CNT-Immobilized Biosensors
Our laboratories have exploited the capability of CNTs to
promote redox activity of hydrogen peroxide to develop
oxidase-based amperometric biosensors, including those
for detecting glucose[5,6,10] and organophosphorous compounds.[8] Similarly, the electrocatalytic properties of
CNTs in reducing the overpotential for the redox reaction
of NADH suggest their potential use in dehydrogenasebased amperometric biosensors such as those for alcohol
detection.[6] These applications of CNT-based biosensors
are summarized as follows.
Glucose detection
The CNT/Nafion/GOx-modified GC electrode was used in
a flow-injection system to measure glucoses.[5] Fig. 6
compares the amperometric responses for relevant physiological levels of glucose, ascorbic acid, acetaminophen,
and uric acid at the CNT/Nafion/GOx-modified GC
electrode (B) and Nafion/GOx-modified GC electrode
(A). In Fig. 6, the accelerated electron-transfer reaction of
hydrogen peroxide at the CNT/Nafion/GOx-modified GC
electrode allows for glucose measurements at very low
potentials (i.e., 0.05 V) where interfering reactions are
minimized. As a result, a well-defined glucose signal (d)
is observed, while the signals of acetaminophen (a), uric
acid (b), and ascorbic acid (c) are negligible. No such
discrimination is obtained at the Nafion/GOx biosensor
(without the CNT) (A) held at +0.80 V, where large
oxidation peaks are observed for all interferences, indicating that the permselective (charge-exclusion) properties of Nafion are not adequate to fully eliminate anionic
interferences. In short, the coupling of the permselective
properties of Nafion with the electrocatalytic action of
CNT allows for glucose detection with effective discrimination against both neutral and anionic redox constituents. Similarly, the CNT/Nafion-coated electrodes have
also been demonstrated to dramatically improve the signal
of dopamine in the presence of the common ascorbic
acid interference.
Fig. 6 Flow-injection signals for 2 10 4 M acetaminophen
(a), 2 10 4 M ascorbic acid (b), 2 10 4 M uric acid (c), and
1 10 2 M glucose (d), at the Nafion/GOx-modified GC
electrode (A) at + 0.8 V, and the MWCNT/Nafion/GOx-modified
GC electrode (B) at 0.05 V, and flow rate of 1.25 mL/min.
(From Ref. 5.)
Copyright © 2004 by Marcel Dekker, Inc. All rights reserved.
Fig. 7 Current–time recordings for successive 2-mM additions
of glucose at the graphite/Teflon/GOx (a) and the MWCNT/
Teflon/GOx (b) electrodes measured at + 0.6 (A) and + 0.1 V (B).
Electrode composition, 30:69:1 wt.% carbon/Teflon/GOx. Other
conditions are as in Fig. 1a. (From Ref. 1.)
The detection of glucose has also been performed on
CNT/Teflon-based electrodes, which are immobilized
with GOx enzyme.[6] Fig. 7 compares the amperometric
response to successive additions of 2 mM glucose at the
graphite/Teflon/GOx (a) and the MWCNT/Teflon/GOx (b)
electrodes using operating potentials of + 0.6 V (A) and
+ 0.1 V (B). The CNT-based bioelectrode offers substantially larger signals, especially at low potential, reflecting
the electrocatalytic activity of CNT. Such low-potential
operation of the CNT-based biosensor results in a highly
linear response (over the entire 2- to 20-mM range) and a
slower response time ( 1 min vs. 25 sec at +0.6 V). The
glucose biocomposite based on single-wall CNTs results
in a more sensitive but slower response than that based on
multiwall CNTs. The low-potential detection also leads to
high selectivity (i.e., effective discrimination against
coexisting electroactive species). Despite the absence of
external (permselective) coating, the glucose response
at + 0.1 V was not affected by adding the common
acetaminophen and uric acid interferences at 0.2 mM. A
similar addition of ascorbic acid resulted in a large
interference, reflecting the accelerated oxidation of this
compound at the CNT surface.[4] Reproducible activity
was observed after 2 weeks of dry storage at 4°C, which is
in good agreement with the high stability of enzymes in
Teflon-based carbon composites.[33]
Organophosphorus compound detection
Organophosphorous (OP) compounds are very toxic and
are thus widely used as pesticides and chemical-warfare
agents (CWAs). Recently, we have successfully used
MWCNTs in developing an amperometric biosensor for
OP compounds.[8] Specifically, the MWCNT is used to
modify screen-printed carbon electrodes, which are subsequently coimmobilized with acetylcholinesterase (ACHE)
and choline oxidase (CHO) enzymes. The MWCNTmodified electrode has demonstrated a significant catalytic effect for the redox reaction of hydrogen peroxide,
leading to the development of a novel biosensor for the
assay of OP compounds with enhanced sensitivity.
The ACHE enzyme is known to play an important role
in cholinergic transmission as a catalyst for the rapid
hydrolysis of the neurotransmitter acetylcholine to acetate
and choline as follows:
ACHE
Acetylcholine þ H2 O ! Acetate þ Choline
was used to oxidize CNTs to create a carboxylic acid
group. Both enzymes, ACHE and CHO, were then coimmobilized on the CNTs via carbodiimide chemistry by
forming amide linkages between their amine residues and
carboxylic acid groups on the CNT surface.[10,43] The
optimum biosensor was found to contain a loading of 0.8
mU ACHE and 1.5 U CHO on the electrode, with a 2-mM
ACH as the substrate.
An amperometric method was employed to study the
sensor response time of the ACHE inhibition after the
spike of methyl parathion, as a representative OP compound. In Fig. 8a, the amperometric response was rapid
(i.e., within 30 sec) after the ACH addition, reflecting the
fast diffusion of enzyme substrates and the products (as
the CNTs are membrane-free porous). When methyl
parathion was successively added to the test area of the
ðiÞ
However, in the presence of OP compounds, the rate of
choline production is reduced. The capability of OP
compounds to inhibit ACHE activity is well known[37–41]
and thus is being exploited in developing biosensors for
OP compound detection. In our work, the amperometric
biosensor for OP compounds is based on coimmobilization of ACHE and CHO on a printed CNT electrode. In
the biosensor based on ACHE/CHO enzymes, choline that
is produced in reaction (i) serves as a substrate for the
CHO enzyme in the presence of oxygen to produce
hydrogen peroxide as follows:
CHO
Choline þ O2 ! Betaine Aldehyde þ H2 O2
ðiiÞ
The hydrogen peroxide that is produced in reaction (ii)
can be detected amperometrically, and the amperometric
response is negatively proportional to the amount of an
OP compound that is introduced into the system.
The inhibition of ACHE activity by OP compounds is
an irreversible process; once exposed to the OP compounds, the enzyme is inactivated, and the sensor can be
reused only after an appropriate enzyme reactivation.[42]
Therefore we have attempted to develop biosensors that
are low-cost and disposable using screen-printed carbon
electrodes.[8] To make the sensor, the suspension of
MWCNTs in N,N-DMF was cast on the surface of screenprinted carbon electrodes to form a thin-film of CNTs.
The screen-printed carbon only serves as a conducting
base for the CNT electrode. The electrochemical method
Copyright © 2004 by Marcel Dekker, Inc. All rights reserved.
Fig. 8 (a) Amperometric response of methyl parathion at the
CNT/ACHE/CHO-immobilized screen-printed biosensor. Operating conditions: 0.1 M phosphate buffer/0.1 M NaCl (pH 7.4),
potential of + 0.50 V. (b) Inhibition effects by three organophosphates, chlorpyrifos (A), fenitrothion (B), and methyl parathion
(C) on the enzyme activity, measured with the CNT/ACHE/
CHO-immobilized screen-printed biosensor. Other conditions
are as in (a). (From Ref. 8.)
B
biosensor, the response decreased significantly and rapidly in the first 10 min and more slowly thereafter. The
significant inhibition effect of methyl parathion to the
catalytic activity of ACHE reduces the production of
hydrogen peroxide, leading to low signals.
In addition to methyl parathion, the inhibition effects of the other two OP compounds were investigated
using the CNT-modified, ACHE/CHO-immobilized electrode. Fig. 8b compares the enzyme activities with a
function of time after the spike of each OP compound,
including (A) chlorpyrifos, (B) fenitrothion, and (C)
methyl parathion.
The enzyme activity is the ratio of Ii (a steady-state
current obtained in the presence of a given OP compound)
to Io (that obtained in the absence of the OP compound).
After spiking with OP compounds, the enzyme activity
decreases with time. The high inhibition effect of OP
compounds can be correlated to low enzyme activity of
ACHE. Their inhibition effects are in the following ascending order: methyl parathion > fenitrothion >chlorpyrifos.
The relative inhibition of methyl parathion at the CNTmodified, ACHE/CHO-immobilized biosensor as a function of methyl parathion concentration was investigated
using a preincubation method in which the biosensor was
exposed to the incubation solution containing methyl
parathion for 10 min before the change in enzyme activity
was measured. Successive incubation measurements were
performed with varied methyl parathion concentrations.
The CNT/ACHE/CHO biosensor has good analytical
characteristics for methyl parathion, including a broad
dynamic linear range (up to 200 mM, r 2 = 0.96), high
sensitivity (0.48% inhibition/mM), and low detection limit
(LDL =0.05 mM). These improved characteristics reflect
the catalytic activity of CNTs that promotes the redox
reaction of hydrogen peroxide produced during ACHE/
CHO enzymatic reactions with their substrate, as well as
the large surface area of CNT materials. The hand-held
electrochemical detector (i.e., CHI1232 from CHI Instrument, Inc.) coupled with the disposable biosensor developed in this work will potentially facilitate the field
screening of OP pesticides and nerve agents with fast speed,
high efficiency, low cost, and small sample size needed.
Alcohol detection
The attractive low-potential detection of NADH, along
with minimal surface fouling, makes CNT extremely
attractive for amperometric biosensors of ethanol through
the incorporation of ADH/NAD+ within the three-dimensional electrode matrix. Specifically, the reagentless biocomposite was prepared by mixing the desired amounts of
the ADH enzyme and the NAD+ cofactor with the CNT/
Teflon composite to obtain the final composition of
28.5:65:1.5:5 wt.% CNT/Teflon/ADH/NAD+. The mix-
Copyright © 2004 by Marcel Dekker, Inc. All rights reserved.
Fig. 9 Current–time recordings for successive 1-mM additions
of ethanol at the graphite/Teflon/ADH/NAD+ (a) and the
MWCNT/Teflon/ADH/NAD+ (b). Operating potential, + 0.2 V;
electrode composition, 28.5:65:1.5:5 wt.% carbon/Teflon/ADH/
NAD+. Other conditions are as in Fig. 1a. (From Ref. 6.)
ture was then packed firmly into the electrode cavity of a
glass sleeve with a copper wire as the electrical contact.
Fig. 9 compares the performance at a low detection
potential (i.e., + 0.20 V) of the CNT/Teflon-based electrode (b) to that of graphite/Teflon-based electrode (a).
Only the CNT/Teflon-based electrode (b) responds favorably to successive additions of 1 mM ethanol. The
response is relatively fast ( 60 sec to reach steady state)
and nonlinear. The greatly enhanced biosensing of ethanol
at the CNT-based electrode suggests the accelerated
oxidation of NADH at low-potential detection. In addition
to being reagentless, the biosensor does not require a
redox mediator (to shuttle the electrons from the NADH
product to the surface), which is commonly used for such
low-potential detection of ethanol.
BIOSENSORS BASED ON
CONTROLLED-DENSITY ALIGNED CNTs
Fabrication
Although vertically aligned CNTs have good material
properties[44] (e.g., good electrical conductivity, ability to
promote electron transfer reactions) and are of the right
size (20 to 200 nm) for nanoelectrode arrays (NEAs), they
lack the right spacing. To make each nanotube work as an
individual nanoelectrode, the spacing needs to be sufficiently larger than the diameter of the nanotubes to
prevent the diffusion layer overlap from the neighboring
electrodes.[45]
Recently, a nonlithography method that allows the
fabrication of low-site-density aligned CNT arrays with an
interspacing of more than several micrometers has been
developed.[46] From these low site density CNTs, the
NEAs consisting of millions of nanoelectrodes with each
electrode being less than 100 nm in diameter were
successfully fabricated.[7,9,10] As the total current of the
loosely packed electrode arrays is proportional to the total
number of individual electrodes, having the number of the
electrodes up to millions is highly desirable. The size
reduction of each individual electrode and the increased
total number of the electrodes result in improved signalto-noise ratio (S/N) and detection limits.[47,48]
In growing the low-site-density aligned CNT arrays, Ni
nanoparticles were randomly deposited on a 1-cm2 Crcoated silicon substrate (Fig. 10a) by applying a pulse
current to the substrate in NiSO4 electrolyte solution. The
size and the site density of the Ni nanoparticles were
controlled by the amplitude and the duration of the pulse
current. On these Ni particles, the CNTs were grown [Fig.
10b] in the plasma-enhanced chemical vapor deposition
(PECVD) system at 650°C for 8 min with 160 sccm NH3
and 40 sccm C2H2 gases with a total pressure of 15 Torr and
a plasma intensity of 170 W. The aligned CNT arrays had a
site density of 1 106–3 106/cm2, a length of 10 to 12 mm,
and a diameter of 50 to 80 nm. In the early days of
fabricating the electrode arrays,[7,46] a thin layer of SiO2
was coated on the surface by magnetron sputtering to
insulate the Cr layer. This was followed by applying MBond 610 coating (epoxy-phenolic adhesive from Vishay
Intertechnology, Inc., Shelton, CT) to further insulate the
Cr and provide mechanical support to the CNTs. This
insulation is sometimes not sufficient to prevent current
leakage, which results in the distortion of the cyclic
voltammetry curve. Recently,[9,10] we improved the fabrication method using better insulating materials and packing procedures that solve the leakage problem, coupled
with a pretreatment of the electrode to reduce the peak
separation. We replaced the SiO2 and M-bond coatings
used as the passivation layer in our previous attempts with
Epon epoxy resin 828 (Miller-Stephenson Chemical Co.,
Inc., Sylmar, CA) and m-phenylenediamine (MPDA) as a
hardener. After these steps, the CNTs were half-embedded
in the polymer resin, and the protruding part of the CNTs
beyond the polymer resin was mechanically removed
by polishing with a lens, followed by ultrasonication in
water. Then the electronic connection was made on the
CNT-Si substrate to make the CNT nanoelectrode arrays
(Fig. 10c). Finally, the electrode arrays were pretreated by
electrochemical etching in 1.0 M NaOH at 1.5 V for 90 sec
prior to the electrochemical characterizations. The lifetime
of the NEAs employing new epoxy is significantly
improved: there is no degradation for several weeks
because of the excellent stability of the epoxy layer.
Applications of Controlled-Density
Aligned CNT Electrodes
Fig. 10 Fabrication scheme of a low-site-density aligned CNT
nanoelectrode array. (From Ref. 7.)
Copyright © 2004 by Marcel Dekker, Inc. All rights reserved.
The nanoelectrode array, consisting of millions of CNT
electrodes, yields an excellent sigmoidal voltammogram
for K3Fe(CN)6 and a low S/N.[9] The sigmoidal voltammogram is a characteristic of microelectrodes having
radial diffusion. The steady-state current arises because
the rate of electrolysis approximates the rate of diffusion
of an analyte to the electrode surface.[49] The scan-rateindependent limiting current behavior was observed up
B
Fig. 11 (a–b) Amperometric responses for 5 mM glucose (G), 0.5 mM ascorbic acid (AA), 0.5 mM acetaminophen (AC), and 0.5 mM
uric acid (UA) at the GOx-modified, CNT-nanoelectrode array and the potentials of + 0.4V (a) and 0.2V (b). Electrolyte: 0.1 M
phosphate buffer/0.1 M NaCl (pH 7.4). (c) Amperometric response at the GOx-modified, CNT-nanoelectrode array for each successive
addition of 2 10 3 M glucose. Inset shows the corresponding calibration curve. Potential: 0.2 V, other conditions are as in (a–b).
(From Ref. 10.)
to 500 mV/sec at the nanoelectrode array. This indicates
that there is no diffusion layer overlapping between the
electrodes because most of the CNTs are separated from
their nearest neighbors by at least 5 mm, much larger than
the diameter of each nanotube (50 to 80 nm).
The CNT-NEAs have a wide range of applications.
They have been investigated in our laboratory as electrochemical sensors without further modification for the
detection of drugs, such as 4-acetamidophenol, and metal
ions, such as lead.[9]
For use as a glucose biosensor,[10] the GOx enzyme
molecules were attached to the broken tips of the CNTs
via carbodiimide chemistry by forming amide linkages
between their amine residues and carboxylic acid groups
on the CNT tips. Fig. 11a and b compares the amperometric responses for 5 mM glucose (G), 0.5 mM ascorbic
Copyright © 2004 by Marcel Dekker, Inc. All rights reserved.
acid (AA), 0.5 mM acetaminophen (AC), and 0.5 mM uric
acid (UA) at the GOx-modified NEA and the potentials of
+ 0.4 V (a) and 0.2V (b). Well-defined cathodic and
anodic glucose responses are obtained at the CNT/GOxbased biosensor at both potentials. However, the glucose
detection at lower operating potential ( 0.2 V) is
significantly less influenced by interferences, indicating
high selectivity toward the glucose substrate. Such a
highly selective response to glucose is obtained at the
CNT/GOx-based biosensor without the use of mediators
and permselective membranes. The amperometric response at the CNT/GOx-based biosensor for each successive addition of 2 10 3 M glucose is presented in
Fig. 11c with the corresponding calibration curve in the
inset. The linear response to glucose is up to 30 mM, and
the steady state is reached within 20 to 30 sec.
For many years, numerous researches have emphasized
the use of anti-interference layers or artificial electron
mediators for improving the selectivity of amperometric
biosensors. CNTs eliminate potential interferences through
the preferential detection of hydrogen peroxide at the CNTbased electrodes. Such development of interference-free
transducers will significantly simplify the design and
fabrication of biosensors. The biosensors based on lowsite-density aligned CNTs are also suitable for the highly
selective detection of glucose in a variety of biological
fluids (e.g., saliva, sweat, urine, and serum).
CONCLUSION
The electrocatalytic properties of CNTs in promoting the
redox reaction of hydrogen peroxide and NADH are being
exploited in our work in developing biosensors based on
oxidase and dehydrogenase enzymes, respectively. With
our CNT-based biosensors, low-potential detections of
alcohol, organophosphorous compounds, and glucose are
possible with many advantages over conventional devices.
The capability of CNTs to reduce the overvoltage for the
oxidation of hydrogen peroxide and NADH allows the
detection of these species at low potential. At such low
potential, most interfering species in the test samples do
not undergo oxidation, thus eliminating potential interference. CNTs also minimize the surface fouling of biosensors, thus imparting higher stability onto these devices.
While the CNT/Nafion-based biosensors use the electrocatalytic activity of CNTs and the permselectivity of
Nafion to detect glucose with effective discrimination
against most neutral and anionic redox constituents, the
CNT/Teflon composite allows the reagentless approach to
fabricate biosensors with flexible coimmobilization of
enzymes and cofactors for specific biosensing needs.
Owing to the CNTs’ capability to promote electron-transfer activity, the artificial mediators that shuttle electrons
between the enzymes and the electrodes are not required
at the CNT-based biosensors, thereby eliminating the
dependence upon the electroactive species and enhancing
the reproducibility.
For future work, we will investigate the capability of
CNTs to promote electron-transfer reactions of other
biologically and environmentally important compounds.
The biosensor fabrication technology demonstrated in this
work holds a great future for developing routine, onsite
amperometric biosensors based on both oxidase and
dehydrogenase enzymes, such as those for cholesterol,
alcohol, lactate, acetylcholine, choline, hypoxanthine, and
xanthine. Although this article focuses on biosensors
based on CNTs, other oriented conducting nanowires, e.g.,
oriented conducting polymer nanowires recently synthe-
Copyright © 2004 by Marcel Dekker, Inc. All rights reserved.
sized in our laboratory,[50,51] should also provide an
alternative ideal platform for biosensing applications.
ACKNOWLEDGMENTS
The work performed at Pacific Northwest National
Laboratory (PNNL) was supported by the Laboratory
Directed Research and Development program. The work
performed at Boston College was supported by the U.S.
Department of Energy (DOE) (DE-FG02-00ER45805)
and by the National Science Foundation (NSF) (CMS0219836). Work performed at New Mexico State University was supported by NSF. The research described in this
paper was performed in part at the Environmental
Molecular Sciences Laboratory, a national scientific
user facility sponsored by the DOE’s Office of Biological
and Environmental Research and located at PNNL. PNNL
is operated for DOE by Battelle under Contract DEAC06-76RL01830.
REFERENCES
1.
Musameh, M.; Wang, J.; Merkoci, A.; Lin, Y. Lowpotential stable NADH detection at carbon-nanotube-modified glassy carbon electrodes. Electrochem. Commun. 2002, 4, 743 – 746.
2. Wang, J.; Li, M.; Shi, Z.; Li, N.; Gu, Z. Direct
electrochemistry of cytochrome c at a glassy carbon
electrode modified with single-wall carbon nanotubes. Anal. Chem. 2002, 74 (9), 1993 – 1997.
3. Wang, J.; Li, M.; Shi, Z.; Li, N.; Gu, Z. Electrocatalytic oxidation of norepinephrine at a glassy
carbon electrode modified with single wall carbon
nanotubes. Electroanalysis 2002, 14 (3), 225 –230.
4. Wang, Z.; Liu, J.; Liang, Q.; Wang, Y.; Luo, G.
Carbon nanotube-modified electrodes for the simultaneous determination of dopamine and ascorbic
acid. Analyst 2002, 127 (5), 653 –658.
5. Wang, J.; Musameh, M.; Lin, Y. Solubilization of
carbon nanotubes by Nafion toward the preparation of
amperometric biosensors. J. Am. Chem. Soc. 2003,
125 (9), 2408 – 2409.
6. Wang, J.; Musameh, M. Carbon nanotube/teflon
composite electrochemical sensors and biosensors.
Anal. Chem. 2003, 75, 2075 – 2079.
7. Tu, Y.; Lin, Y.; Ren, Z.F. Nanoelectrode arrays
based on low site density aligned carbon nanotubes.
Nano Lett. 2003, 3 (1), 107 –109.
8. Lin, Y.; Lu, F.; Wang, J. Disposable carbon
nanotube modified screen-printed biosensor for
amperometric detection of organophosphorus pesticides and nerve agents. Electroanalysis, in press.
B
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
Tu, Y.; Lin, Y.; Yantasee, W.; Huang, Z.P.;
Carnahan, D.L.; Ren, Z.F. Carbon nanotubes based
nanoelectrode array: Fabrication, evaluation and the
application in voltammetric detection. Analyst,
submitted.
Lin, Y.; Lu, F.; Tu, Y.; Ren, Z.F. Glucose biosensor
based on low site density aligned carbon nanotubes.
Nano Lett., submitted.
Britto, P.J.; Santhanam, K.S.V.; Ajayan, P.M.
Carbon nanotube electrode for oxidation of dopamine. Bioelectrochem. Bioenerg. 1996, 41, 121 –
125.
Britto, P.J.; Santhanam, K.S.V.; Rubio, A.; Alonso,
A.J.; Ajayan, P.M. Improved charge transfer at
carbon nanotube electrodes. Adv. Mater. 1999, 11,
154– 157.
Campbell, J.K.; Sun, L.; Crooks, R.M. Electrochemistry using single carbon nanotubes. J. Am.
Chem. Soc. 1999, 121, 3779 – 3780.
Azamian, B.R.; Davis, J.J.; Coleman, K.S.; Bagshaw, C.B.; Green, M.L.H. Bioelectrochemical
single-walled carbon nanotubes. J. Am. Chem.
Soc. 2002, 124, 12664 – 12665.
Nguyen, C.V.; Delzeit, L.; Cassell, A.M.; Li, J.;
Han, J.; Meyyappan, M. Preparation of nucleic
acid functionalized carbon nanotube arrays. Nano
Lett. 2002, 2, 1079 – 1081.
Li, J.; Ng, H.T.; Cassell, A.; Fan, W.; Chen, H.; Ye,
Q.; Koehne, J.; Han, J.; Meyyappan, M. Carbon
nanotube nanoelectrode array for ultrasensitive
DNA detection. Nano Lett. 2003, 3, 597 –602.
Xu, J.Z.; Zhu, J.J.; Wu, Q.; Hu, Z.; Chen, H.Y. An
amperometric biosensor based on the coimmobilization of horseradish peroxidase and methylene blue
on a carbon nanotubes modified electrode. Electroanalysis 2003, 15, 219– 224.
Zhao, Q.; Gan, Z.H.; Zhuang, Q.K. Electrochemical
sensors based on carbon nanotubes. Electroanalysis
2002, 14, 1609 – 1613.
Yu, X.; Chattopadhyay, D.; Galeska, I.; Papadimitrakopoulos, F.; Rusling, J.F. Peroxidase activity of
enzymes bound to the ends of single-wall carbon
nanotube forest electrodes. Electrochem. Commum.
2003, 5, 408 – 411.
Sotiropoulou, S.; Chaniotakis, N.A. Carbon nanotube array-based biosensor. Anal. Bioanal. Chem.
2003, 375, 103 – 105.
Shim, M.; Kam, N.W.S.; Chen, R.J.; Li, Y.M.; Dai,
H.J. Functionalization of carbon nanotubes for
biocompatibility and biomolecular recognition.
Nano Lett. 2002, 2, 285 –288.
Iijima, S. Helical microtubules of graphitic carbon.
Nature 1991, 354 (6348), 56– 58.
Saito, R.; Dresselhaus, G.; Dresselhaus, M.S.
Copyright © 2004 by Marcel Dekker, Inc. All rights reserved.
Physical Properties of Carbon Nanotubes; Imperial
College Press: London, 1998.
24. Chen, J.; Hamon, M.A.; Hu, H.; Chen, Y.S.; Rao,
A.M.; Eklund, P.C.; Haddon, R.C. Solution properties of single-walled carbon nanotubes. Science 1998,
282 (5386), 95– 98.
25. Boul, P.J.; Liu, J.; Mickelson, E.T.; Huffman, C.B.;
Ericson, L.M.; Chiang, I.W.; Smith, K.A.; Colbert,
D.T.; Hauge, R.H.; Margrave, J.L.; Smalley, R.E.
Reversible sidewall functionalization of buckytubes.
Chem. Phys. Lett. 1999, 310 (3–4), 367 –372.
26. Star, A.; Stoddart, J.F.; Steuerman, D.; Diehl, M.;
Boukai, A.; Wong, E.W.; Yang, X.; Chung, S.W.;
Choi, H.; Heath, J.R. Preparation and properties of
polymer-wrapped single-walled carbon nanotubes.
Angew. Chem., Int. Ed. 2001, 40 (9), 1721 –1725.
27. Riggs, J.E.; Guo, Z.X.; Carroll, D.L.; Sun, Y.P.
Strong luminescence of solubilized carbon nanotubes. J. Am. Chem. Soc. 2000, 122 (24), 5879 –
5880.
28. O’Connell, M.J.; Boul, P.; Ericson, L.M.; Huffman,
C.; Wang, Y.H.; Haroz, E.; Kuper, C.; Tour, J.;
Ausman, K.D.; Smalley, R.E. Reversible watersolubilization of single-walled carbon nanotubes by
polymer wrapping. Chem. Phys. Lett. 2001, 342 (3–
4), 265– 271.
29. Fan, Z.H.; Harrison, D.J. Permeability of glucose
and other neutral species through recast perflourosulfonated ionomer films. Anal. Chem. 1992, 64
(11), 1304 – 1311.
30. Fortier, G.; Vaillancourt, M.; Belanger, D. Evaluation of Nafion as media for glucose-oxidase
immobilization for the development of an amperometric glucose biosensor. Electroanalysis 1992, 4 (3),
275– 283.
31. Zhao, Q.; Gan, Z.H.; Zhuang, Q.K. Electrochemical
sensors based on carbon nanotubes. Electroanalysis
2002, 14 (23), 1609 – 1613.
32. Wang, J.; Reviejo, A.J.; Angnes, L. Graphite–
teflon enzyme electrode. Electroanalysis 1993, 5
(7), 575– 579.
33. Del Cerro, M.A.; Cayuela, G.; Reviejo, A.J.;
Pingarron, J.M.; Wang, J. Graphite–teflon–peroxidase composite electrodes. Application to the direct
determination of glucose in musts and wines.
Electroanalysis 1997, 9 (14), 1113 –1119.
34. Lobo, M.J.; Miranda, A.J.; Tunon, P. Amperometric
biosensors based on NAD(P)-dependent dehydrogenase enzymes. Electroanalysis 1997, 9 (3), 191 –202.
35. Blaedel, W.J.; Jenkins, R.A. Study of the electrochemical oxidation of NADH. Anal. Chem. 1975,
47 (8), 1337 –1343.
36. Wang, J.; Angnes, L.; Martinez, T. Scanning
tunneling microscopic probing of surface fouling
during the oxidation of nicotinamide coenzymes.
Bioelectrochem. Bioenerg. 1992, 29 (2), 215 –221.
37. Rosenberry, T.L. Advances in Enzymology and
Related Areas of Molecular Biology; John Wiley
& Sons, Inc.: New York, 1975; 43, 103.
38. Zhang, S.; Zhao, H.; John, R. Development of a
quantitative relationship between inhibition percentage and both incubation time and inhibitor concentration for inhibition biosensors—Theoretical and
practical considerations. Biosens. Bioelectron.
2001, 16 (9–12), 1119 –1126.
39. Fennouh, S.; Casimiri, V.; Burstein, C. Increased
paraoxon detection with solvents using acetylcholinesterase inactivation measured with a choline
oxidase biosensor. Biosens. Bioelectron. 1997, 12
(2), 97 –104.
40. Cremisini, C.; Disario, S.; Mela, J.; Pilloton, R.;
Palleschi, G. Evaluation of the use of free and immobilized acetylcholinesterase for paraoxon detection with an amperometric chlorine oxidase based
biosensors. Anal. Chim. Acta 1995, 311 (3), 273 –280.
41. Guerrieri, A.; Monaci, L.; Quinto, M.; Palmisano, F.
A disposable amperometric biosensor for rapid
screening of anticholinesterase activity in soil
extracts. Analyst 2002, 127 (1), 5 –7.
42. Dzydevich, S.V.; Shul’ga, A.A.; Soldatkin, A.P.;
Hendji, A.M.N.; Jaffrezic-Renault, N.; Martelet, C.
Conductometric biosensors based on cholinesterases
for sensitive detection of pesticides. Electroanalysis
1994, 6, 752– 758.
43. Edwards, R. Immunodiagnostics—A Practical Approach; Oxford University Press: Oxford, 1999; 67.
Copyright © 2004 by Marcel Dekker, Inc. All rights reserved.
44.
Ren, Z.F.; Huang, Z.P.; Xu, J.W.; Wang, J.H.; Bush,
P.; Siegal, M.P.; Provencio, P.N. Synthesis of large
arrays of well-aligned carbon nanotubes on glass.
Science 1998, 282 (5391), 1105 –1107.
45. Morf, W.E.; De Rooij, N.F. Performance of
amperometric sensors based on multiple microelectrode arrays. Sens. Actuators. B, Chem. 1997, 44
(1–3), 538– 541.
46. Tu, Y.; Huang, Z.P.; Wang, D.Z.; Wen, J.G.;
Ren, Z.F. Growth of aligned carbon nanotubes
with controlled site density. Appl. Phys. Lett.
2002, 80 (21), 4018 – 4020.
47. Menon, V.P.; Martin, C.R. Fabrication and evaluation of nanoelectrode ensembles. Anal. Chem. 1995,
67 (13), 1920 – 1928.
48. Weber, S.G. Single-to-noise ratio in microelectrodearray-based electrochemical detectors. Anal. Chem.
1989, 61 (4), 295 –302.
49. Wehmeyer, K.R.; Wightman, R.M. Scan rate
dependence of the apparent capacitance at microvoltammetric electrodes. J. Electroanal. Chem.
1985, 196 (2), 417 –421.
50. Liang, L.; Liu, J.; Windisch, C.F.; Exarhos, G.J.;
Lin, Y. Direct assembly of large arrays of oriented
conducting polymer nanowires. Angew. Chem. Int.
Ed. 2002, 41 (19), 3665 –3668.
51. Liu, J.; Lin, Y.; Liang, L.; Voigt, J.A.; Huber, D.L.;
Tian, Z.; Coker, E.; McKenzie, B.; McDermott, M.J.
Templateless assembly of molecularly aligned
conductive polymers nanowires: A new approach
for oriented nanostructures. Chem. Eur. J. 2003, 9
(3), 604 –611.
B