View Article Online / Journal Homepage / Table of Contents for this issue

Analyst Dynamic Article Links < C

Cite this: Analyst, 2012, 137, 5113

www.rsc.org/analyst PAPER
Development of amperometric lysine biosensors based on Au nanoparticles/
multiwalled carbon nanotubes/polymers modified Au electrodes†
Nidhi Chauhan,a Anamika Singh,b Jagriti Narang,a Swati Dahiyab and C. S. Pundir*a
Received 14th May 2012, Accepted 31st August 2012
Published on 03 September 2012 on http://pubs.rsc.org | doi:10.1039/C2AN35629E

DOI: 10.1039/c2an35629e

The construction of two amperometric L-lysine biosensors is described in this study. The construction
Downloaded by University of New Hampshire on 24 February 2013

comprises the covalent immobilization of lysine oxidase (LOx) onto nanocomposite composed of gold
nanoparticles (AuNPs) and carboxylated multiwalled carbon nanotubes (c-MWCNT), decorated on (i)
polyaniline (PANI) and (ii) poly 1,2 diaminobenzene (DAB), electrodeposited on Au electrodes. The
biosensors were characterized by scanning electron microscopy (SEM), Fourier transform infrared
(FTIR) and electrochemical impedance spectroscopy (EIS) studies. The optimum response (current)
was observed within 2 s at pH 7.0 and 25
C for LOx/AuNPs/c-MWCNT/PANI/Au, and 4 s at pH 7.0
and 30
C for LOx/AuNPs/c-MWCNT/DAB/Au electrodes. There was a linear relationship between
current and lysine concentration ranging from 5.0 to 600 mM for LOx/AuNPs/c-MWCNT/PANI/Au
with a detection limit of 5.0 mM, and 20 to 600 mM for the LOx/AuNPs/c-MWCNT/DAB/Au electrode
with a detection limit of 20 mM. The PANI modified electrode was in good agreement with the standard
HPLC method, with a better correlation (r ¼ 0.992) compared to the DAB modified electrode
(r ¼ 0.986). These observations revealed that the PANI modified Au electrode was better than the DAB
modified electrode, and hence it was employed for the determination of lysine in milk, pharmaceutical
tablets and sera. The PANI modified electrode showed a half life of 120 days, compared to that of
90 days for the DAB modified electrode, after their 100 uses, when stored at 4
C.

Introduction consuming and require tedious sample pretreatment, expensive

Lysine, one of nine essential amino acids, is popularly used as an
additive in the animal feed industry, as well as a supplement in
human nutrition.1 Therefore it is considered as an index of the
nutritional quality of food stuff and pharmaceutical formula-
tions.2–4 A low level of lysine in body fluids indicates certain
disease conditions, i.e. tiredness, inability to concentrate,
bloodshot eyes, retarded growth, hair loss, anemia and repro-
duction problems.5 People with a high level of lysine may have an
intellectual disability or behavioral problems. Due to the clinical,
nutritional and pharmaceutical significance of this amino acid,
rapid and sensitive methods are required for the detection of
lysine in diverse matrices.6,7
Various analytical methods are available for determining lysine,
such as chemical assays based on the ninhydrin reaction, capillary
electrophoresis, gas phase chromatography and HPLC.8,9
Although these methods provide fruitful results, they are time
a
Department of Biochemistry, M. D. University, Rohtak-124 001,
Haryana, India. E-mail: pundircs@rediffmail.com; Fax: +91-126274640;
Tel: +91 9416492413
b
Department of Biotechnology, UIET, Kurukshetra University,
Kurukshetra-136 119, Haryana, India
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c2an35629e
equipment and trained personnel to operate them.2,10 Biosensing
methods are comparatively more simple, rapid, sensitive, specific,
inexpensive, portable and reusable. Lysine biosensors employing a
nylon/collagen membrane mounted on different electrodes,11–15
aminosilylated glass,5 polycarbonate membranes,16 activated
alkylaminated controlled pore glass,17 poly(o-phenylenediamine)
membranes,18,3 polyurethane hydrogel,19 PTFE (polytetrafluoro-
ethylene) tube,6 imidodiacetic acid chelating beads on glassy carbon
foil20 and gold–mercaptopropionic acid self assembled mono-
layers21 have met with limited success, due to the leakage of enzyme
from the support during washing.
Recently, nanomaterials have been used as novel supports to
immobilize and modify biomolecules for the improvement of
biosensors. Carbon nanotubes (CNTs) have been employed in
biosensors as effective catalyst supports due to their large surface
areas, unique structural and electromechanical properties, good
biocompatibility, easy preparation and surface renewal.22 Gold
nanoparticles (AuNPs) are already established as an ideal
material for biosensors, especially for those using electro-
chemical detection, because the gold surface is suitable for the
binding of biomolecules, and the metal facilitates fast and direct
electron transfer.23
Although most of the lysine biosensors have been based on
non-conducting polymer matrices like diaminobenzene (DAB),17

This journal is ª The Royal Society of Chemistry 2012 Analyst, 2012, 137, 5113–5122 | 5113

the use of conducting polymer matrices like polyaniline (PANI)
is yet to be studied. PANI is a unique conducting polymer due to
its relatively facile synthesis, high conductivity and environ-
mental stability. To the best of our knowledge, no attempt has
been made to combine it with nanomaterials for the improve-
ment of lysine biosensors. We describe herein the construction of
two novel amperometric lysine biosensors based on the covalent
immobilization of lysine oxidase (LOx) onto AuNPs/carboxyl-
ated multiwalled carbon nanotubes (c-MWCNT)/PANI and
AuNPs/c-MWCNT/DAB electrodeposited on Au electrodes,
and their comparison and application.
Published on 03 September 2012 on http://pubs.rsc.org | doi:10.1039/C2AN35629E
Experimental
Chemicals and reagents
Downloaded by University of New Hampshire on 24 February 2013
Lysine oxidase (EC 1.4.3.14 from Trichoderma viride), 1,2-di-
aminobenzene, and aniline from Sigma Chemical Co., USA,
sodium citrate dehydrate and hydrogen tetrachloroaurate tri-
hydrate from Sisco Research Laboratory, Mumbai, India,
carboxylated multiwalled carbon nanotubes (c-MWCNT)
(Functionalized MWCNT, 12 walls, length 15–30 mm, purity
90%, metal content: nil) from Intelligent Materials Pvt. Ltd.,
Panchkula (Haryana), India, and Alamin M Forte (amino acids
with minerals capsules) manufactured by M/S Albert David
Limited, Ghaziabad, India, were used. Au wire (1.5  0.05 cm2,
23 carat) was purchased from a local market. All other chemicals
were of analytical reagent grade. Double distilled water (DW)
was used throughout the experiments.
Apparatus
Amperometric measurements were conducted with a potentio-
stat/galvanostat (model: Autolab AUT83785, Ecochemie, the
Netherlands). The morphology of the AuNPs was studied by
transmission electron microscopy (TEM) at Punjab University,
Chandigarh. Ultrasonication was performed on Misonix Ultra-
sonic Liquid Processors (Mode XL-2000 series). Fourier trans-
form infrared (FTIR) spectroscopy of the modified electrodes
was carried out with a FTIR spectrometer (model: iS10, Ther-
moelectron, USA). Scanning electron microscopy
(SEM) measurements of the electrodes were carried out at the
Department of Chemistry, M. D. University, Rohtak. Lysine
analysis of milk samples by HPLC was carried out commercially
at Arbro Pharmaceuticals Ltd. (Analytical Division), New Delhi,
India.
Preparation of gold nanoparticles (AuNPs)
AuNPs were prepared according to the method of McFarland
et al.,24 with modifications. 20 ml of 1.0 mM hydrogen tetra-
chloroaurate (HAuCl4) solution in a flask was kept on a hot plate
under constant stirring. To this boiling solution, 1% trisodium
citrate dehydrate (2 ml) was added dropwise. The gold sol was
formed gradually, as the citrate reduced the HAuCl4 solution to
gold. The change of colour of the solution from blackish to wine
red confirmed the aggregation of the nanoparticles. The colloidal
gold solution was then stored in dark bottles at 4
C after
cooling. The morphological characterization of the AuNPs was
carried out by TEM.
5114 | Analyst, 2012, 137, 5113–5122
View Article Online
Construction of the enzyme electrodes
Preparation of AuNPs/c-MWCNT/PANI modified Au elec-
trode. Aniline (50 ml) was added to 10.0 ml of 1 N HCl and elec-
trodeposited onto the Au electrode through a cyclic voltammetric
technique using a potentiostat/galvanostat. Prior to electrode-
position, the Au electrode was polished with 0.05 mM alumina
slurry and then immersed in piranha solution (a hot mixed solu-
tion of conc. H2SO4 and 30% H2O2 in a 3 : 1 ratio (v/v)) for 15 min,
followed by ultrasonic cleaning with DW. The electrochemical
polymerization of aniline onto the Au electrode (working elec-
trode) was achieved through cyclic voltammetry, by applying ten
polymerization cycles at 0.0 to +1.5 V.22 1 mg c-MWCNTs was
suspended in 1 ml of a mixture of concentrated H2SO4 and HNO3
in a 3 : 1 ratio (v/v) and ultrasonicated for 2 h to get a finely
dispersed black colored solution of MWCNTs. The PANI coated
Au electrode was dipped into this c-MWCNT solution for 24 h at
room temperature. The resulting c-MWCNT/PANI/Au electrode
was washed thoroughly with DW to remove unbound matter and
then immersed into the AuNPs solution for 12 h at 4
C. The
electrode was washed again with DW to remove any unbound
matter, and kept in a dry Petri plate at 4
C until use.
Preparation of AuNPs/c-MWCNT/DAB modified Au electrode.
Prior to use, the Au electrode (1.5  0.05 cm2) was polished with
alumina slurry, treated with piranha solution and washed in DW
as described above. Then the electrode was held at a fixed
potential of +1.2 V vs. Ag/AgCl in NaOH (1 M), for 5 min
and then scanned between 0.2 V and +1 V (at a scan rate of
50 mV s1) for 5 min in the same NaOH solution. The Au
electrode was pre-treated by potential cycling from 0.2 V to
+1.2 V in 0.5 M H2SO4 for 20 min. A DAB film was electro-
deposited onto the electrode surface by means of cyclic voltam-
metry at 0.0 to +0.8 V in a 0.05 M solution of the
1,2-diaminobenzene monomer in sodium phosphate buffer (pH
7.4). The electrode was scanned until the current reached a value
which remained constant after further cycling. The potential was
continuously cycled until a minimum value of current was
reached which remained constant after further cycling. This
indicated that the electrode surface was completely covered by
the polymer.18 The AuNPs/c-MWCNT/DAB modified Au elec-
trode was prepared from the DAB coated Au electrode as
described above.
Immobilization of LOx enzyme on the modified Au electrodes.
Both types of enzyme electrodes were prepared by dropping
50 ml of LOx solution (6 U) onto the surface of the modified
electrodes and keeping them at 4
C for 24 h. The enzyme
electrodes were rinsed with reaction buffer, dried and stored at
4
C until use.
The fabrication of the lysine biosensors based on the LOx/
AuNPs/c-MWCNT/PANI modified Au electrode is summarized
in Scheme 1A. Firstly, both PANI and DAB were electro-
deposited onto two separate Au electrodes with free –NH2
groups at their ends, which linked to the –COOH groups of
MWCNTs through amide bonds (–CONH–). The electrodepo-
sition method was selected to produce a polymer layer on elec-
trode surface, as this method was easy to carry out and the layer
thickness could be controlled. AuNPs were deposited on the
This journal is ª The Royal Society of Chemistry 2012

Published on 03 September 2012 on http://pubs.rsc.org | doi:10.1039/C2AN35629E
Downloaded by University of New Hampshire on 24 February 2013
Scheme 1 (A) Schematic representation of the fabrication of the lysine oxidase/AuNPs/c-MWCNT/PANI modified Au electrode. (B) Schematic
representation of the electron flow and current generation in the lysine biosensor employing LOx/AuNPs/c-MWCNT/PANI/Au as the working
electrode.
surface of PANI/c-MWCNT and DAB/c-MWCNT by the
physical adsorption method. Then LOx was immobilized cova-
lently, as the –NH2 group on surface of the enzyme forms an
amide bond (–CONH–) with the –COOH group of the
MWCNTs.
Scanning electron microscopy (SEM) of LOx electrodes during
modifications
To study SEM of both the Au electrodes modified by AuNPs, c-
MWCNT, PANI/DAB and LOx, they were cut into small pieces
(1 cm) and placed on a copper mount of 2 cm diameter using a
spray gun, and their micrographs were taken with a scanning
electron microscope.
Construction and response measurements of amperometric LOx
biosensors
The electrochemical response of the LOx/AuNPs/c-MWCNT/
PANI/Au and LOx/AuNPs/c-MWCNT/DAB/Au electrodes was
studied using cyclic voltammetry. During the catalytic cycle, the
LOx is first reduced and then regenerated by oxidation with
molecular oxygen in the sample solution. The electron is now
accepted by the AuNPs/c-MWCNT/PANI to form reduced
This journal is ª The Royal Society of Chemistry 2012
View Article Online
AuNPs/c-MWCNT/PANI. This reduced AuNPs/c-MWCNT/
PANI is oxidized by releasing an electron, which further migrates
to the Au electrode and forms the current. The operative mode of
a lysine biosensor based on the above sequence of events is the
electrochemical oxidation of H2O2 releasing e and completing
the cycle at the working electrode (Scheme 1B). The mechanism
of electron flow and current generation is same in both the
biosensors.
Amperometric biosensors for the measurement of lysine were
prepared by using the modified Au electrode (LOx/AuNPs/c-
MWCNT/PANI/Au or LOx/AuNPs/c-MWCNT/DAB/Au) as
the working electrode, Ag/AgCl as the reference and Pt wire as
the auxiliary electrode connected through a potentiostat/galva-
nostat. Cyclic voltammetric studies were carried out on the
electrodes at different stages of modification. The three electrode
system was dipped into 20 ml of 50 mM sodium phosphate
buffer, pH 7.5. The reaction was started by adding 0.1 ml of 0.1
mM lysine, and the current (mA) generated was recorded at
different voltages. Cyclic voltammograms of the modified elec-
trodes were recorded after applying a suitable voltage.
The maximum current was obtained at +0.4 V, and hence in
subsequent studies the electrodes were polarized at this voltage to
generate current.
Analyst, 2012, 137, 5113–5122 | 5115
 View Article Online

from the hospital of the local Pt BDS P. G. Institute of Medical

L-lysine þ O2 ƒƒƒƒƒƒƒƒƒ! a-keto-3-aminocaproate þ H2 O2
Lysine oxidase

þ NH3
þ0:4V
H2 O2 ƒƒƒƒƒƒƒ! 2Hþ þ O2 þ 2e-
To evaluate the activity of the LOx/AuNPs/c-MWCNT/
PANI/Au and LOx/AuNPs/c-MWCNT/DAB/Au electrodes, the
modified electrodes were characterized by a cyclic voltammo-
gram in the presence and absence of lysine. For this purpose,
cyclic voltammograms of the working electrodes in an unstirred
Published on 03 September 2012 on http://pubs.rsc.org | doi:10.1039/C2AN35629E
Science, Rohtak, and analyzed for lysine using the present elec-
trode. The procedure for the measurement of lysine in these
samples was the same as described for the response measurement
of the electrodes, except that lysine was replaced by the sample.
The content of lysine in these samples was determined from the
standard curve of lysine concentration vs. current in mA
(Fig. 1A) (prepared in buffer).
Storage stability and reusability of modified Au electrodes
To reuse the working electrodes, they were washed by dipping in

a test tube containing 2 ml of reaction buffer. The electrodes were

0.1 M KCl solution (20 ml) and 0.1 M sodium phosphate buffer,
stored dry in a refrigerator at 4
C when not in use.
pH 7.0 (5 ml) with and without 0.1 mM lysine solution, were
recorded in the potential range of 0 to +1.5 V at a scan rate of
50 mV s1. The well defined oxidation and reduction peaks of the
Downloaded by University of New Hampshire on 24 February 2013

electrode were increased after addition of 0.1 mM lysine, which

clearly indicates the catalytic properties of the modified elec-
trode. Hence, in the subsequent studies, the electrochemical
measurements of the electrodes obtained in the absence of lysine
were subtracted from those obtained in the presence of lysine.

Optimization of biosensors
Optimization of the lysine biosensors was accomplished by
measuring their response at different pH values (from 5.5 to 10.0)
using citrate buffer (pH 5.5), phosphate buffer (pH 6.0, 6.5, 7.0
and 7.5), Tris buffer (pH 8.0 and 8.5) and bicarbonate buffer (pH
9.0 and 10.0), each at 0.1 M final concentration containing 0.1 M
potassium chloride. To determine the optimum temperature, the
reaction mixture was incubated at temperatures ranging from
20
C to 55
C at intervals of 5
C. To study the effect of substrate
concentration, different lysine concentrations ranging from 5 to
600 mM were tested by cyclic voltammetry from 0.0 to +1.5 V at a
scan rate of 20 mV s1. The limit of detection (LOD) was
calculated from the standard deviation (SD) of the response and
the slope of the calibration curve (S) at levels approximating the
LOD, according to the formula, LOD ¼ 3.3(SD/S). The SD of
the response was determined from the standard deviation of the
y-intercepts of the regression line. The amperometric response
was also measured in the presence of potential interfering
substances, methionine, phenylalanine, proline, serine, uric acid,
ascorbic acid, histidine, cysteine, tyrosine, arginine, tryptophan,
glutamic acid and ornithine, each at 1 mM concentration.

Amperometric determination of lysine in milk, pharmaceutical

tablets and serum
As the PANI modified electrode showed better analytical prop-
erties than the DAB modified electrode, it was employed for the
detection of lysine in real samples, e.g. milk, pharmaceutical
tablets and serum. Protein samples from milk were digested in
6 N HCl for 15 min. Prior to the assay, the samples were brought
to pH 7.0 with 1 M KOH. The samples were then assayed for
lysine content. To measure the L-lysine content in pharmaceutical
tablets, namely Alamin M Forte (amino acids with minerals Fig. 1 (A) A calibration curve between the lysine concentration (mM)
capsule), a 1000 mg tablet was dissolved in the sodium phosphate and current response (mA) of the biosensor based on (a) LOx/AuNPs/c-
buffer (0.1 M, pH 7.0) after grinding with a pestle and mortar, MWCNT/PANI/Au and (b) LOx/AuNPs/c-MWCNT/DAB/Au. (B)
and then analysed for lysine. Fresh serum samples were collected Transmission electron microscope (TEM) image of the AuNPs.

5116 | Analyst, 2012, 137, 5113–5122 This journal is ª The Royal Society of Chemistry 2012

Results and discussion
The AuNPs synthesized in this study were essentially very fine
and monodisperse with a diameter of ca. 18–20 nm, as seen from
their TEM image (Fig. 1B).
Electrode surface characterization by SEM
The morphologies of bare Au, PANI/Au, c-MWCNT/PANI/Au,
AuNPs/c-MWCNT/PANI/Au, DAB/Au, c-MWCNT/DAB/Au,
AuNPs/c-MWCNT/DAB/Au, LOx/AuNPs/c-MWCNT/PANI/
Au and LOx/AuNPs/c-MWCNT/DAB/Au electrodes were char-
Published on 03 September 2012 on http://pubs.rsc.org | doi:10.1039/C2AN35629E
acterized by SEM studies. SEM of the bare Au electrode (ESI,
Fig. S1a†) showed a smooth and featureless morphology.
However, fine fiber-like structures indicate the presence of a
Downloaded by University of New Hampshire on 24 February 2013
polymer layer on the Au electrode surfaces (ESI, Fig. S1b and f†).
The tubular structures in ESI, Fig. S1c and g† and the granular
structures in ESI, Fig. S1d and h† indicate the presence of
c-MWCNT and AuNPs, respectively, on the surface. Globular
Fig. 2 (A) FTIR spectra of (i) PANI, (ii) c-MWCNT/PANI/Au electrode and (iii) LOx/AuNPs/c-MWCNT/PANI/Au electrode. (B) FTIR spectra of (i)
DAB, (ii) c-MWCNT/DAB/Au electrode and (iii) LOx/AuNPs/c-MWCNT/DAB/Au electrode.
This journal is ª The Royal Society of Chemistry 2012
View Article Online
shapes on the electrode surfaces (Fig. S1e and i†) show the pres-
ence of the enzyme layer after immobilization.
FTIR characterization of the modified electrodes
FTIR spectra of the electrodes show the chemical bonding
involved in the covalent immobilization of the enzyme (LOx)
onto the AuNPs/c-MWCNT/PANI and AuNPs/c-MWCNT/
DAB electrodes. Fig. 2A shows the FTIR spectra of the PANI/
Au electrode (curve i), c-MWCNT/PANI/Au electrode (curve ii)
and LOx/AuNPs/c-MWCNT/PANI/Au electrode (curve iii).
Curve i shows peaks of the quinoid and benzenoid structures of
PANI at 1577 and 1485 cm1, and bands at 1319 cm1 assigned
to the –C–N stretching of 2
amines. It shows that the aniline was
polymerized to form a PANI chain linked through –NH– bonds
at the Au electrode, with two free –NH2 groups at its ends. Curve
ii shows the peak of the C–O bond of free –COOH groups at
1294.99 and 1343.93 cm1, and the –CONH– bonds of MWCNT
immobilized on PANI at 1625 cm1. It reveals that the –COOH
Analyst, 2012, 137, 5113–5122 | 5117
 View Article Online

group at one end of the MWCNT is attached to the –NH2 groups

of PANI through –CONH– bonds. The disappearance of the
peak of free –COOH at 1294 cm1 in curve iii confirms that the
enzyme was immobilized covalently onto the free –COOH group
of MWCNT through –CONH– bonds. There was one more peak
at wavelength 578.00 cm1, which was due to the out of plane
C–H bending vibration.
Similarly, Fig. 2B shows the FTIR spectra for the DAB/Au
electrode (curve i), c-MWCNT/DAB/Au electrode (curve ii) and
LOx/c-MWCNT/DAB/Au electrode (curve iii). Curve i shows
the DAB spectrum with a broad band in the region between 3700
and 3000 cm1, generally assigned to primary and secondary
amino groups present in the polymer. The band at 1636 cm1 is
Published on 03 September 2012 on http://pubs.rsc.org | doi:10.1039/C2AN35629E

broad and can be assigned to C]C stretching vibrations of the

aromatic rings. Another band at 1400 cm1 may correspond to
the ring stretching vibrations of the phenazine type structures in
Downloaded by University of New Hampshire on 24 February 2013

the DAB backbone. Bands assigned to C–N stretching vibrations

of the quinoid and benzoid rings appear at 1319 and 1287 cm1,
respectively. The peaks at 850 cm1 are due to the out of plane
bending motion of the C–H bonds of the 1,2,4,5-tetrasubstituted
benzene nuclei of phenazine units. The presence of the bands at
884 cm1 due to the in-plane bending motion of the C–H bonds
of the 1,2,4-trisubstituted benzene rings indicates the presence of
open rings in the phenazine units. The signal at 614 cm1 can be
assigned to ring deformation. The peak at 578 cm1 is due to out
of plane C–H bending vibrations. Curve ii reveals new strong
peaks at 3200–3500, which can be assigned to the N–H, NH2 and
OH stretching modes. The peak at 1669 cm1 reveals the amide
–CONH– linkage formation. The other peak at around 1620
cm1 can be assigned to the NH2 scissoring mode, that overlaps
with the C]C mode. The peaks at around 1540–1590, 1400–
1480, 1200–1380 and 1100 cm1 correspond to C]C stretching
in the nanotubes, aromatic ring modes, C–N and C–O stretching
modes, respectively. Like the PANI modified electrode, the
disappearance of the peak of free –COOH at 1200 cm1 in curve
iii confirmed that the enzyme was immobilized covalently onto
the free –COOH group of the MWCNTs through –CONH–
bonds.
Impedance measurements
Electrochemical impedance spectroscopy (EIS) was carried out
to measure the charge transfer resistance (RCT) at the interface of
the modified electrodes. The value of the electron transfer resis-
tance (semicircle diameter) depends on the dielectric and insu-
lating features at the electrode–electrolyte interface. An increase
in RCT value indicates a resistance or hindrance of electron flow
due to the addition of a substance on the surface of the electrode,
which leads to an increase in RCT value. The impedance scans
were studied in the frequency range of 0.01 Hz to 10 kHz (Fig. 3A
and B). The RCT values of the AuNPs/c-MWCNT/PANI and
AuNPs/c-MWCNT/DAB modified Au electrodes (1000 and
1100 U respectively, curve b) are lower than those of the
c-MWCNT/PANI/Au and c-MWCNT/DAB/Au electrodes
(1500 and 1350 U respectively, curve a), as the AuNPs act as an
electron-transfer accelerator, and help to enhance the current
response of the enzyme electrode, thus decreasing the resistance
charge transfer. However, the RCT value increases after the
immobilization of LOx onto the AuNPs/c-MWCNT/PANI and
Fig. 3 (A) Nyquist plots of the sensing electrode response at different
stages in the electrode assembly process: c-MWCNT/PANI/Au (a);
AuNPs/c-MWCNT/PANI/Au electrode (b) and LOx/AuNPs/c-
MWCNT/PANI/Au electrode (c). (B) Nyquist plots of the sensing elec-
trode response at different stages in the electrode assembly process:
c-MWCNT/DAB/Au electrode (a); AuNPs/c-MWCNT/DAB/Au elec-
trode (b) and LOx/AuNPs/c-MWCNT/DAB/Au electrode (c). All spectra
were recorded in the presence of 10 mM [Fe(CN)6]3/4 in 0.1 M KCl as a
redox-active indicator. Frequency range: 0.01 Hz to 10 kHz.
AuNPs/c-MWCNT/DAB modified Au electrodes (1850 and
1490 U respectively, curve c). Further increase in RCT with the
immobilized LOx confirms the successful immobilization of
enzyme on the conducting polymer surface. This increase in RCT
is attributed to the fact that most biological molecules, including
enzymes, are poor electrical conductors at low frequencies
(<10 kHz) and hinder the electron transfer.
Cyclic voltammetric studies
The cyclic voltammetry (CV) responses for different electrodes in
50 mM pH 7.0 phosphate buffer solution containing 0.1 mM
lysine at a scan rate of 50 mV s1 were recorded. No redox
reaction occurred on the bare Au electrode. After the Au elec-
trode was modified by a layer of AuNPs/PANI, the CV curve
changed dramatically from an electrochemically capacitive shape
to a reversible shape. However, the electrodes coated with
MWCNT/PANI/Au electrode again exhibited reversible cyclic
voltammetry with large cathodic and anodic peaks, resembling
that of the bare Au electrode. This was due to the fact that
AuNPs/PANI and MWCNT/PANI possess better conductivity
than the bare Au electrode, and facilitated the electron transfer.
Compared with the bare Au electrode, the AuNPs/MWCNT/
PANI/Au electrode exhibited a bigger peak-to-peak separation,
indicating high electron transfer between the analyte and the
electrode surface. Therefore, we can conclude that the AuNPs on
the modified MWCNTs/PANI/Au electrode not only provide the

5118 | Analyst, 2012, 137, 5113–5122 This journal is ª The Royal Society of Chemistry 2012
 View Article Online

Table 1 Determination of lysine in H-milk (HCl hydrolyzed milk)

samples using the LOx/AuNPs/c-MWCNT/PANI/Au biosensor

Lysine conc. by biosensor

H-milk samples (mM), mean  SD (n ¼ 5) RSD (%)

Sterilized non fat dry milk 199.4  1.14 0.221

Sterilized whole milk 212.7  1.23 0.179
Clarified raw milk 234.3  1.71 0.245

Electrochemical characterization of LOx/AuNPs/c-MWCNT/

PANI/Au and LOx/AuNPs/c-MWCNT/DAB/Au electrodes by
cyclic voltammetry
Published on 03 September 2012 on http://pubs.rsc.org | doi:10.1039/C2AN35629E

In the case of the c-MWCNT/PANI and c-MWCNT/DAB

modified Au electrodes, the current responses of the Au electrode
Downloaded by University of New Hampshire on 24 February 2013

to the addition of 0.1 mM lysine at different stages of construction

Fig. 4 (A) Cyclic voltammograms of (a) bare Au electrode, (b) c-
MWCNT/PANI modified Au electrode, (c) AuNPs/c-MWCNT/PANI
modified Au electrode and (d) LOx/AuNPs/c-MWCNT/PANI/Au elec-
trode in 50 mM pH 7.0 phosphate buffer solution containing 0.1 mM
lysine at a scan rate of 50 mV s1. (B) Cyclic voltammograms of (a) bare
Au electrode, (b) c-MWCNT/DAB modified Au electrode, (c) AuNPs/c-
MWCNT/DAB modified Au electrode and (d) LOx/AuNPs/c-MWCNT/
DAB/Au electrode in 50 mM pH 7.0 phosphate buffer solution con-
taining 0.1 mM lysine at a scan rate of 50 mV s1.
necessary conduction pathways, but also act like nanoscale
electrodes, promoting the electron transfer between the analyte
and the electrode surface. On the other hand, when AuNPs or
MWCNTs were deposited onto the Au electrode, a remarkable
increase in current was seen. The maximum current was observed
when AuNPs were loaded on the MWCNTs and doped on the
electrode. This can be attributed to the synergistic effect of these
two nanomaterials (ESI, Fig. S2†).
were studied (Fig. 4A and B). The cyclic voltammograms
Fig. 4A(a–d) are different from the cyclic voltammograms
Fig. 4B(a–d) because of the different polymer used (PANI or
DAB). The bare electrode yielded a very small current response to
lysine (trace a), that was not detectable due to the H2O2 not being
electrooxidized at the bare electrode surface when a potential of
0.0 to +1.5 V was applied. The small current indicated that the
direct oxidation of the substrate at the Au electrode was ineffi-
cient. CV of c-MWCNT/PANI/Au revealed one anodic (oxida-
tion/upward peak) and one cathodic (reduction/downward peak)
peak due to the grouping of c-MWCNT/PANI. At the
c-MWCNT/PANI/Au electrode (Fig. 4A, trace b), a slight
increase in current response to H2O2 was observed compared with
the bare Au electrode, which may be ascribed to the fact that the
relative surface area of the electrode was increased after the
immobilization of c-MWCNT onto it. The introduction of
AuNPs onto the c-MWCNT/PANI/Au electrode dramatically
improved the current response (trace c). The significant increase in
the current response after the addition of AuNPs to the electrode
can be ascribed to the good catalytic activity of AuNPs for the
electrooxidation of H2O2. An anodic current was found to
increase from 0.05 to 0.075 mA with c-MWCNT/AuNPs. In the
case of the c-MWCNT/DAB/Au electrode, anodic current was
found to increase from 0.250 to 0.415 mA with c-MWCNT/
AuNPs. The increase in current response after the covalent
immobilization of LOx onto the modified AuNPs/c-MWCNT/
PANI/Au and AuNPs/c-MWCNT/DAB/Au electrodes indicated
the immobilization of LOx.

Table 2 Lysine levels in serum samples of apparently healthy adults, as

determined by the biosensor based on the LOx/AuNPs/c-MWCNT/
PANI/Au electrode

Serum sample Sex Lysine (mM), mean  SD (n ¼ 5)

S1 F 229.0  1.00
S2 F 244.0  0.70
Fig. 5 (A) The voltammograms of the LOx/AuNPs/c-MWCNT/PANI/ S3 M 216.8  0.44
Au electrode obtained in the presence of different concentrations of S4 F 278.8  0.44
lysine: (a) 5, (b) 50, (c) 100, (d) 200, (e) 300, (f) 400, (g) 500 and (h) 600 mM S5 M 220.0  0.70
in phosphate buffer (50 mM, pH 7.0) at a scan rate of 50 mV s1. (B) The S6 M 271.3  0.30
S7 M 274.5  0.44
voltammograms of the LOx/AuNPs/c-MWCNT/DAB/Au electrode S8 M 274.7  0.45
obtained in the presence of different concentrations of lysine: (a) 20, (b) S9 F 263.8  0.80
50, (c) 100, (d) 200, (e) 300, (f) 400, (g) 500 and (h) 600 mM in phosphate S10 F 233.8  0.90
buffer (50 mM, pH 7.0) at a scan rate of 50 mV s1.

This journal is ª The Royal Society of Chemistry 2012 Analyst, 2012, 137, 5113–5122 | 5119
 Downloaded by University of New Hampshire on 24 February 2013
Published on 03 September 2012 on http://pubs.rsc.org | doi:10.1039/C2AN35629E

Table 3 Comparison of performance characteristics of membrane based LOx biosensors with the present biosensors

Method of Type of Temp Response LOD Storage

Reference Support matrix immobilization transducer pH (
C) time Linearity (M) (M) stability Application
6 4 6
11 Nylon net/collagen Glutaraldehyde DO metric 7.5 45 5–8 s 6.7  10 to 6.7 10 6.7  10 6 months Wheat extracts
membrane clark type crosslinking from
electrode young plants
16 Immobilon/polycarbonate Glutaraldehyde Amperometric 7.0 25 2 min 5  106 to 5  104 5  107 10 days Animal feed
membrane on platinum crosslinking (batch)
electrode and 30 s
(flow)
12 Nylon membrane Glutaraldehyde Chemiluminescent 8.5 20 10 s 2.5  106 to 103 106 Up to 300 Flow injection
crosslinking injections analysis
18 DAB on platinum Passive adsorption Amperometric 7.5 NRa 12–16 s 1  105 to 1  103 2  107 NR NR

5120 | Analyst, 2012, 137, 5113–5122

electrode
13 Nylon membrane on Glutaraldehyde Amperometric 7.5 30 42 s Up to 1  104 8.2  107 25 days Pharmaceutical
graphite–methacrylate crosslinking samples
electrode
19 Polyurethane hydrogel Entrapment Amperometric 7.0 RT NR 1  105 to 2.5  104 1  105 6 months Fermentation
on platinum electrode samples from
microbial
cultures
3 Poly(o-phenylenediamine) Glutaraldehyde Potentiometric 9.0 37 NR 1  104 to 1  102 1  104 40 days NR
membrane on Si–Gold crosslinking
strip electrode
2 Ruthenium–rhodium coated BSA–glutaraldehyde Amperometric 7.0 30 NR 2  106 to 1.25  104 2  106 NR Milk, pasta
glassy carbon electrode coupling
20 Imidodiacetic acid Non-covalent Electrochemiluminescent 8.5 NR 180 s 1  106 to 0.5  103 1  106 NR NR
chelating beads onto
glassy carbon foil
14 Nylon membrane Covalent Amperometric 5.0 37 60 s 1.0  106 to 3.3  105 0.5  106 1 month Bran and pasta
(polyazetidine samples
as immobilizing agent)
17 Activated alkylaminated Glutaraldehyde Amperometric 6.0 NR 80 samples 1  105 to 1.6  104 5  106 Two weeks NR
controlled pore glass crosslinking Chemiluminescent 6.0 per hour 2.05  107 to 2.75  4  108
105
15 Nylon membrane onto Glutaraldehyde Potentiometric NR NR 135 s 5  104 to 1  102 1.05  105 NR NR
ammonium electrode crosslinking 1.25  105
6 Transparent PTFE tube Covalent coupling Chemiluminometric 9.8 RT 11 samples 8  105 to 4  104 4  108 37 days Human serum
to vinyl polymer per hour
4
21 Gold–mercaptopropionic EDC and NHS Amperometric 7.4 RT NR 1  104 to 8  103 1  10 NR NR
acid self assembled coupling
monolayer
5 Aminosilylated glass BHA/glutaraldehyde Chemiluminometric 7.4 NR 90 samples 1  105 to 1  103 1  105 1 month NR
co-crosslinking per hour
6
LOx/AuNPs/c- Au electrode Covalent Amperometric 7.0 25 2s 5  106 to 6  104 5  10 120 days Milk, serum
MWCNT/ and amino
PANI/Au acid
electrode tablets
LOx/AuNPs/c- Au electrode Covalent Amperometric 7.0 30 4s 25  106 to 6  104 25  106 90 days Milk, serum
MWCNT/ and amino
DAB/Au acid tablets
electrode
a
NR ¼ Not reported.

This journal is ª The Royal Society of Chemistry 2012

View Article Online

Our results show that the LOx/AuNPs/c-MWCNT/PANI
modified Au electrode demonstrated a greater synergistic effect
toward the oxidation of lysine compared with the LOx/AuNPs/c-
MWCNT/DAB modified Au electrode.
Optimization of lysine biosensors
The experimental conditions affecting the biosensor response
were studied in terms of the effects of pH, incubation tempera-
ture, time and substrate (lysine) concentration. The flow of
electrons (i.e., current) was measured in mA at +0.4 V (ESI,
Fig. S3a–c†). Both the modified Au electrodes showed a
maximum response at pH 7.0, which is more or less similar to
Published on 03 September 2012 on http://pubs.rsc.org | doi:10.1039/C2AN35629E
that of free enzyme. The optimum temperatures of the LOx/
AuNPs/c-MWCNT/PANI and LOx/AuNPs/c-MWCNT/DAB
biosensors were 25
C and 30
C, respectively, which are lower
Downloaded by University of New Hampshire on 24 February 2013
than that of the free enzyme (37
C). The decrease in the
optimum temperature of the enzyme after immobilization might
be due to the increase in enzyme rigidity upon immobilization
through covalent binding. The PANI and DAB based biosensors
showed optimum response within 2 s and 4 s, respectively.
Effect of lysine concentration on the response of enzyme
electrodes
The effect of lysine concentration on biosensor response was
measured by adding different concentrations (5–600 mM) of
lysine, which was oxidized, producing an electroactive H2O2.
Formation of H2O2 was detected by its oxidation generating
electrons (i.e., current at the electrode). Fig. 5A and B show the
cyclic voltammograms of the enzyme electrodes’ response at
different lysine concentrations. There was a linear increase in
oxidation current with the increase in lysine concentration in the
range 20–600 mM for the DAB electrode, and 5–600 mM for the
PANI modified electrode. The current response observed in case
of the PANI electrode was higher than that of the DAB modified
Au electrode.
Interference study
There was practically no interference in the response measure-
ment of either of the electrodes by methionine, phenylalanine,
proline, serine, uric acid, ascorbic acid, histidine, cysteine, tyro-
sine, arginine, tryptophan, glutamic acid or ornithine, each at a
final concentration of 1 mM (ESI, Table S1†).
Evaluation of biosensor
The LOD of the LOx/AuNPs/c-MWCNT/PANI modified Au
electrode was 5 mM, which is better than that of the LOx/AuNPs/
c-MWCNT/DAB/Au electrode (20 mM). The analytic recovery of
a known amount of lysine added into spiked milk samples was
determined by the present biosensor. The mean analytic recov-
eries of added 50 mM and 100 mM were 96.00  0.7% and 93.00 
0.8% for the PANI modified Au electrode, and 92.00  1.5% and
90.00  2.0% for the DAB modified Au electrode, respectively. To
test the reliability of the present biosensors, the lysine content in
six milk samples was estimated on a single day (within a batch),
and five times again after storage at 20
C for one week (between
batches). The results showed that determinations were consistent,
This journal is ª The Royal Society of Chemistry 2012
View Article Online
and within and between batch coefficients of variation were low
for each electrode. However, the coefficient of variation for the
PANI modified Au electrode was the lowest (within batch: 1.59%,
and between batch: 2.17%) compared to the DAB modified Au
electrode (within batch: 1.70%, and between batch: 4.40%). The
high precision indicated the reproducibility and consistency of the
developed biosensors. To study the correlation of the present
biosensors, the lysine levels in 15 H-milk (HCl hydrolysed milk)
samples determined by the present biosensor (y) were compared
with those determined by the standard HPLC method (x) (from
the same batch). The values obtained by the LOx/AuNPs/c-
MWCNT/PANI/Au and LOx/AuNPs/c-MWCNT/DAB/Au
electrodes were very close to those obtained by the standard
method. The correlation coefficients (r) were 0.992 and 0.986
respectively. Thus the LOx/AuNPs/c-MWCNT/PANI/Au
method showed better correlation than the DAB based sensor.
Application
The lysine contents of milk, amino acid tablets and serum, as
determined by the present PANI modified biosensor, are given in
the tables. The lysine content was in the range 199.4 to 234.3 mM
in the H-milk samples (Table 1), 24.6 mg in an amino acid tablet,
and 216.8 to 278.8 mM in sera of apparently healthy persons with
a mean of 248.0 mM (n ¼ 10) (Table 2), which is within the
normal established range (150–250 mM).
Storage stability and reusability
The half life of the LOx/AuNPs/c-MWCNT/PANI/Au biosensor
was 4 months, which is better than that of the LOx/AuNPs/c-
MWCNT/DAB biosensor (3 months) (ESI, Fig. S4†).
A comparison of various analytical properties of the present
biosensors with the earlier reported biosensor is summarized in
Table 3.
Conclusion
The use of PANI/DAB and nanoparticles (AuNPs/c-MWCNT)
in the construction of lysine biosensors has resulted in its much
improved analytical performance over earlier reported biosen-
sors in terms of response time (2 s and 4 s), limit of detection (5
and 20 mM), higher stability (up to 4 months) and no interference
from various serum substances. The PANI modified electrode
showed better performance than the DAB modified electrode.
References
1 S. Anastassiadis, Recent Pat. Biotechnol., 2007, 1, 11–24.
2 S. C. Kelly, P. J. O’Connell, C. K. O’Sullivan and G. G. Guilbault,
Anal. Chim. Acta, 2000, 412, 111–119.
3 I. D. Karalemas, A. G. Constantinos and D. S. Papastathopoulos,
Talanta, 2000, 53, 391–402.
4 N. Garcia-Villar, J. Saurina and S. Hernandez-Cassou, Anal. Chim.
Acta, 2003, 477, 315–324.
5 F. Preuschoff, U. Spohn and D. Janasek, Biosens. Bioelectron., 1994,
9, 543–549.
6 N. Kiba, T. Miwa, M. Tachibana, K. Tani and H. Koizumi, Anal.
Chem., 2002, 74, 1269–1274.
7 A. Guerrieri, T. R. I. Cataldi and R. Ciriello, Sens. Actuators, B, 2007,
126, 424–430.
8 B. M. Nair, A. Laser, A. Burvall and N. G. Asp, Food Chem., 1978, 3,
283–291.
Analyst, 2012, 137, 5113–5122 | 5121

9 H. Yoshida, Y. Nakano, K. Koiso, H. Nohta, J. Ishida and
M. Yamaguchi, Anal. Sci., 2001, 17, 107–112.
10 I. B. Z. Diaz and S. C. Ricke, Appl. Microbiol. Biotechnol., 2003, 62,
268–273.
11 E. Vrbova and M. Marek, Anal. Chim. Acta, 1992, 270, 131–136.
12 F. Preuschoff, U. Spohn and D. Janasek, Biosens. Bioelectron., 1994,
9, 543–549.
13 J. Saurina, S. Hernandez-Cassou, S. Alegret and E. Fabregas,
Biosens. Bioelectron., 1999, 14, 211–220.
14 F. Mazzei, F. Botre, G. Favero, E. Podesta and C. Botre, Microchem.
J., 2009, 91, 209–213.
15 N. Garcia-Villar, J. Saurina and S. Hernandez-Cassou, Anal. Chim.
Acta, 2003, 477, 315–324.
16 M. G. Lavagnini, D. Moxone and G. Pallexhi, Talanta, 1993, 40,
1301–1306.
Published on 03 September 2012 on http://pubs.rsc.org | doi:10.1039/C2AN35629E
Downloaded by University of New Hampshire on 24 February 2013
5122 | Analyst, 2012, 137, 5113–5122
View Article Online
17 A. M. Almuaibed and A. Townshend, Anal. Chim. Acta, 1997, 338,
149–154.
18 A. Curulli, S. Kelly, C. O’Sullivan, G. G. Guilbault and G. Palleschi,
Biosens. Bioelectron., 1998, 13, 1245–1250.
19 H. Olschewski, A. Erlenkotter, C. Zaborosch and G. C. Chemnitius,
Enzyme Microb. Technol., 2000, 26, 537–543.
20 C. A. Marquette, A. Degiuli and L. J. Blum, Biosens. Bioelectron.,
2003, 19, 433–439.
21 R. K. Shervedani, Z. Hemmatian and A. Hatefi-Mehrjardi,
Bioelectrochemistry, 2009, 75, 124–129.
22 N. Chauhan, J. Narang and C. S. Pundir, Analyst, 2011, 136, 1938–
1945.
23 S. Andreescu and L. A. Luck, Anal. Biochem., 2008, 375, 282–290.
24 A. D. McFarland, C. L. Haynes, C. A. Mirkin, R. P. Van Duyne and
H. A. Godwin, J. Chem. Educ., 2004, 81, 544A–544B.
This journal is ª The Royal Society of Chemistry 2012