Bioelectrochemistry 80 (2011) 114–120

Contents lists available at ScienceDirect

Bioelectrochemistry
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / b i o e l e c h e m

Amperometric glucose sensor based on glucose oxidase immobilized on

gelatin-multiwalled carbon nanotube modified glassy carbon electrode
Arun Prakash Periasamy, Yu-Jung Chang, Shen-Ming Chen ⁎
Department of Chemical Engineering and Biotechnology, National Taipei University of Technology, No.1, Section 3, Chung-Hsiao East Road, Taipei 106, Taiwan, ROC

a r t i c l e i n f o a b s t r a c t

Article history: We investigated the direct electrochemistry of glucose oxidase (GOx) at gelatin-multiwalled carbon
Received 22 April 2010 nanotube (GCNT) modified glassy carbon electrode (GCE). GOx was covalently immobilized onto GCNT
Received in revised form 16 June 2010 modified GCE through the well known glutaraldehyde (GAD) chemistry. The immobilized GOx showed a pair
Accepted 18 June 2010
of well-defined reversible redox peaks with a formal potential (E0′) of − 0.40 V and a peak to peak separation
Available online 26 June 2010
(ΔEp) of 47 mV. The surface coverage concentration (Г) of GOx in GCNT/GOx/GAD composite film modified
Keywords:
GCE was 3.88 × 10− 9 mol cm− 2 which indicates the high enzyme loading. The electron transfer rate constant
Direct electrochemistry (ks) of GOx immobilized onto GCNT was 1.08 s− 1 which validates a rapid electron transfer processes. The
Glucose oxidase composite film shows linear response towards 6.30 to 20.09 mM glucose. We observed a good sensitivity of
Gelatin 2.47 μA mM−1 cm− 2 for glucose at the composite film. The fabricated biosensor displayed two weeks
Multiwalled carbon nanotubes stability. Moreover, it shows no response to 0.5 mM of ascorbic acid (AA), uric acid (UA), acetaminophen
Electrocatalysis (AP), pyruvate (PA) and lactate (LA) which shows its potential application in the determination of glucose
from human serum samples. The composite film exhibits excellent recovery for glucose in human serum at
physiological pH with good practical applicability.
© 2010 Elsevier B.V. All rights reserved.

1. Introduction the high sensitivity and good stability of the immobilized enzyme [3].
In addition, MWCNT platform offers more compatible environment to
In recent years, rapid and accurate monitoring of blood glucose the immobilized GOx and thereby favors the rapid shuttling of
level is required for the control of diabetes. Among various electronic electrons with the electrode surface [12]. Though the MWCNT based
devices commercially available in the market, with low-cost of glucose sensors possess high sensitivity and good electrocatalytic
production, rapid response, high sensitivity and good selectivity, activity they have a major drawback that the prepared MWCNT
glucose oxidase (GOx) based enzyme sensors have been widely dispersions precipitate on long storage and are not quite stable.
employed to monitor the blood glucose level [1]. However, one of the Further, the dispersing agents may cause the electrode surface passive
most challenging aspects in the development of highly sensitive or cause any damage to the walls of MWCNT. Thus there is a need for
glucose sensor is the selection of a suitable matrix for GOx exploring a suitable dispersing agent for MWCNT.
immobilization. With high mechanical strength, excellent conductiv- Gelatin is the chief protein present in skin, bones and white
ity and antifouling properties, multiwalled carbon nanotubes connective tissues which is generally separated by the partial hydrolysis
(MWCNTs) have been immensely used in glucose sensors for GOx of collagen [13]. It is extremely heterogeneous; composed of polypep-
immobilization [2]. Some of the recently reported MWCNT matrices tides of various sizes and it possesses a molecular weight distribution in
for GOx immobilization are ionic liquid n-octylpyridinum hexafluor- the range of 15,000 to 250,000. Previously, Zheng and Zheng reported
ophosphate dispersed MWCNT [3], MWCNT/gold nanoparticles- that the immobilization of non-polar amino acid chain of gelatin at the
Teflon composites [4], Chitosan dispersed MWCNT [5], gold nano- side wall of MWCNT takes place through hydrophobic–hydrophobic
particles/thiol group functionalized MWCNT/Chitosan-ionic liquid [6], interactions and which leads to the formation of stable dispersion of
cobalt hexacyanoferrate nanoparticles/gold nanoparticles (seed)/ MWCNT [14]. In addition, other related reports found elsewhere in the
MWCNT composites [7], crystalline gold nanoparticles modified literature substantiate that the stability of the MWCNT/gelatin
MWCNT [8], platinum nanoparticle/MWCNT composites [9], nitrogen dispersion could be comprehensively augmented by increasing the
doped MWCNT [10] and polymerized ionic liquid-wrapped MWCNT amount of MWCNT [15,16]. Considering such advantageous properties,
[11]. The inclusion of MWCNT into the immobilization matrix leads to in the present study, we prepared a stable MWCNT dispersion using
gelatin as the dispersing agent. For the first time, we utilized the gelatin
dispersed MWCNT (GCNT) matrix to immobilize the model enzyme
⁎ Corresponding author. Tel.: + 886 2270 17147; fax: + 886 2270 25238. GOx. We noticed a rapid direct electron transfer processes between GOx
E-mail address: smchen78@ms15.hinet.net (S.-M. Chen). and the glassy carbon electrode (GCE). Further, a good linear range and

1567-5394/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.bioelechem.2010.06.009
 A.P. Periasamy et al. / Bioelectrochemistry 80 (2011) 114–120 115

reasonable sensitivity for glucose was observed at the composite film dimethylformamide (DMF) solution and then ultrasonicated for 1 h. The
with GCNT. The good performance of this glucose sensor could be DMF dispersed MWCNT (DCNT) was then used for the fabrication of
attributed to the stability of GCNT and this opens up new horizons in the DCNT/GOx/GAD film on GCE following the similar fabrication procedure
field of functionalized MWCNT based glucose sensor related researches. employed for GCNT/GOx/GAD composite film.

2. Experimental 3. Results and discussions

2.1. Reagents and apparatus 3.1. Direct electrochemistry of GOx

GOx, type X-S from Aspergillus niger and gelatin from porcine skin, Fig. 1 shows the cyclic voltammograms obtained at bare, GCNT,
type A were purchased from Sigma-Aldrich and used as received. DCNT, DCNT/GOx/GAD and GCNT/GOx/GAD film modified GCEs in
MWCNT with O.D. 10–15 nm, I.D. 2–6 nm, and length 0.1–10 μm was deoxygenated PBS at the scan rate of 50 mV s− 1. Cyclic voltammo-
obtained from Aldrich and used without further purification. 25 wt.% grams were recorded in the potential range of −0.8 to 0.2 V. Initially,
glutaraldehyde (GAD) solution in water was obtained from Sigma- the GCNT loading was optimized through CV studies for the efficient
Aldrich. D(+) glucose (Dextrose, anhydrous) and acetaminophen were immobilization of GOx. The effect of GCNT loading on cathodic peak
purchased from Wako pure chemical Industries, Ltd. Ascorbic acid, current (Ipc) at GCNT/GOx/GAD composite film modified GCE is
uric acid and L (+) lactic acid (98%) were obtained from Sigma- shown in Fig. 1 inset. It is obvious from the inset plot that Ipc increases
Aldrich. Sodium pyruvate solution, 100 mM, sterile-filtered, cell with increase in the amount of GCNT. Maximum Ipc is observed for
culture tested was also obtained from Sigma-Aldrich. The supporting 20 μg cm− 2 of GCNT i.e. 0.13 mL cm− 2 of GCNT. This optimized
electrolyte used for all experiments is pH 7 phosphate buffer solution amount of GCNT is used for all experiments. In Fig. 1, in the absence of
(PBS) prepared using 0.05 M Na2HPO4 and NaH2PO4 solutions. All the GOx, no significant redox peaks are noticed in this potential range for
reagents used were of analytical grade and doubly distilled water was bare, GCNT or DCNT modified GCEs. The GOx/GAD film too exhibits no
utilized for the preparation of all aqueous solutions. redox peaks. Interestingly, a pair of well-defined redox peaks with
Electrochemical studies were performed in a conventional three maximum peak current is observed at GCNT/GOx/GAD composite film
electrode cell. Prior to each experiment, all solutions were deoxygen- modified GCE in the same potential window. E0′ associated with this
ated by passing pre-purified N2 gas for 10 min. Cyclic voltammetry (CV) redox couple is −0.40 V which is close to the E0′ of other GOx
studies were carried out using CHI 1205a work station. BAS GCE with an modified electrodes reported previously [17–19]. The peak to peak
electrode surface area of 0.079 cm2 was used as working electrode and separation (ΔEp) is 47 mV. This redox couple could be thus attributed
Pt wire with 0.5 mm diameter was used as counter electrode. All the to the direct electron transfer of GOx at the GCNT modified GCE
potentials were referred with respect to standard Ag/AgCl reference surface. A similar redox couple with much lowered peak current is
electrode. IM6ex ZAHNER (Kroanch, Germany) was used for electro- observed at DCNT/GOx/GAD film on GCE. E0′ for this redox couple is
chemical impedance spectroscopy (EIS) studies. CHI-750 potentiostat −0.41 V and ΔEp is 33 mV. The higher redox peak current observed at
was used for amperometric i–t curve studies. Surface morphology of the GCNT/GOx/GAD composite film on GCE shows a faster electron
prepared GCNT was studied using JEM 2007 model transmission transfer process. Furthermore, the surface coverage concentration (Г)
electron microscope (TEM). Hitachi S-3000 H scanning electron
microscope (SEM) and Being nano-instruments CSPM 4000, atomic
force microscope (AFM) were employed for the surface morphological
characterization of GOx, GCNT/GOx/GAD and DCNT/GOx/GAD films.

2.2. Fabrication of GCNT/GOx/GAD composite film

GCNT used in this work was prepared according to the procedure

reported elsewhere in the literature [14]. Briefly, 10 mg of MWCNT was
dispersed into 1 mg mL− 1 of aqueous gelatin solution and the resulting
mixture was ultrasonicated for 1 h, until a homogeneous, black
dispersion containing 2 mg mL− 1 MWCNT was obtained. 0.25% GAD
and 10 mg mL− 1 GOx solutions were prepared in PBS and stored at 4 °C.
Prior to electrode modification, GCE surface was polished well on a clean
Buehler polishing cloth using 0.05 μm alumina slurry. The polished GCE
surface was washed several times with double distilled water,
ultrasonicated for 10 min and finally dried at room temperature. The
cleaned GCE was then used for the fabrication of GCNT/GOx/GAD
composite film. About 0.13 mL cm− 2 of GCNT was drop casted above
the pre-cleaned GCE surface and dried. Then 0.05 mL cm− 2 of GOx
solution was spread gently over the GCNT modified GCE and the enzyme
solution was left to dry for 10 min at 35 °C. Finally, 0.03 mL cm− 2 of
0.25% GAD solution was drop casted above the GOx modified GCE and
dried for 10 min. The free –NH2 group of GOx and –CHO group of GAD
react to form the covalent linkage at the GCNT matrix (see Fig S1
supplementary data). Thus GOx has been covalently immobilized at the
GCNT matrix through GAD cross-linking. The obtained GCNT/GOx/GAD
composite film modified GCE was then gently rinsed a few times with
double distilled water and used for further experiments. In order to
Fig. 1. Cyclic voltammograms obtained at bare GCE ( ), GOx/GAD ( ), DCNT/
investigate the importance of using GCNT as immobilization matrix, GOx/GAD ( ), GCNT ( ), DCNT ( ) and GCNT/GOx/GAD ( ) film modified GCEs
MWCNT without gelatin coating has also been prepared. For this in deoxygenated PBS at 50 mV s− 1 scan rate. Inset shows the influence of GCNT loading
purpose, about 10 mg of MWCNT was dispersed into 10 mL of N,N- on Ipc observed at GCNT/GOx/GAD composite film modified GCE.
116 A.P. Periasamy et al. / Bioelectrochemistry 80 (2011) 114–120

of GOx at both GCNT/GOx/GAD and DCNT/GOx/GAD modified GCEs solutions (pH 1 to 11). The redox peak current remains stable in the
has been calculated. The maximum Г value of 3.88 × 10− 9 mol cm− 2 is pH range of 1 to 11. The redox couple is also reproducible when the
observed at GCNT/GOx/GAD composite film on GCE which is 33.12% composite film modified GCE is transferred from one buffer solution to
higher than that of DCNT/GOx/GAD film (Г = 2.59 × 10− 9 mol cm− 2). another. The influence of pH over the anodic peak potential (Epa),
This shows higher enzyme loading at the GCNT/GOx/GAD composite cathodic peak potential (Epc) and E0′ of FAD/FADH2 redox couple at
film. GCNT/GOx/GAD composite film modified GCE is shown in Fig. 3 inset.
From the inset plot of E0′ vs. pH, it is apparent that both Epa, Epc and E0′
3.2. Different scan rate studies exhibit a linear dependence over different pHs. The correlation
coefficient is 0.987. The slope value is found to be −49 mV/pH which
Fig. 2 shows the cyclic voltammograms of GCNT/GOx/GAD is closer to −50.3 and −50.9 mV/pH reported for GOx immobilized at
composite film modified GCE in deoxygenated PBS at different scan bamboo shaped MWCNT and poly(diallydimethylammonium) chloride
rates. Both Ipa and Ipc increase linearly with increase in scan rates from (PDDA) wrapped MWCNT respectively [21,22]. It is also close to the
20 to 200 mV s− 1. This indicates that the electron transfer process theoretical value of Nernstian equation for equal number of proton and
occurring at GCNT/GOx/GAD composite film on GCE is a surface electron transfer process. Thus the FAD/FADH2 redox couple reported in
confined process. Fig. 2 inset shows the linear dependence of Ipa and this study is an equal number of proton and electron transfer processes.
Ipc on different scan rates from 20 to 200 mV s− 1. The electron
transfer rate constant (ks) for GOx at GCNT/GOx/GAD composite film 3.4. Surface morphological characterizations using TEM, SEM and AFM
modified GCE is calculated using Laviron equation [20]. studies

Log ks = α Log ð1−αÞ + ð1−αÞLog α−Log ðRT = nFνÞ− α ð1−αÞ nFΔEp = 2:3 RT: TEM study has been carried out to investigate the surface
ð1Þ morphology of the prepared GCNTs. Fig. S2 shows the TEM image of
GCNT at 10 nm magnification. It is clear from Fig. S2 that GCNTs are well
Where, R is the gas constant (8.314 J mol−1 K− 1), T is the room dispersed in aqueous gelatin solution and they retain their well-defined
temperature (298.15 K) and ΔEp is the peak separation of the FAD/ nanostructures. The outer wall of MWCNT is coated well with gelatin.
FADH2 redox couple. Here, α value is assumed as ≈0.5 and the The gelatin coating around the walls of MWCNT offers good biocom-
number of electrons transferred is considered as 2. The ks value at patibility to the immobilized GOx. As reported previously by Zheng and
GCNT/GOx/GAD composite film on GCE is calculated to be 1.08 s− 1 Zheng the good stability of MWCNTs in aqueous gelatin solution could
and it is higher than that reported by Yao and Shiu for GOx be attributed to the strong interactions between the hydrophobic amino
immobilized at MWCNT-gold colloid composites with poly (diallydi- acid chain of gelatin and hydrophobic side walls of MWCNTs.
methylammonium chloride) (PDDA) coatings (ks = 1.01 s− 1) [19]. Prior to AFM and SEM analyses, ITO surfaces were cleaned well,
This result shows that GCNT matrices efficiently facilitate electron sonicated in acetone–water mixture for 15 min and then dried. GOx,
transfer process between GOx and the electrode surface. DCNT/GOx/GAD and GCNT/GOx/GAD films were fabricated on clean
ITO surface. Then dried at 25 °C and used for AFM and SEM
3.3. Effect of pH characterizations. Fig. 4(a) and (a′) shows the AFM and SEM images
of only GOx film. Small spherical voids surrounded by small bead like
Fig. 3 shows the effect of pH on GOx (FAD/FADH2) redox couple at structures are seen on the film surface. Whereas the AFM and SEM
GCNT/GOx/GAD composite film modified GCE in various buffer images of DCNT/GOx/GAD film coated ITO show large bead like
structures without any spherical voids (See Fig. 4 (b) and (b′)). This
may be due to the filling up of the voids by the GAD. Similarly, Fig. 4(c)

Fig. 2. Cyclic voltammograms recorded at GCNT/GOx/GAD composite film modified GCE Fig. 3. Cyclic voltammograms obtained at GCNT/GOx/GAD composite film modified GCE
in deoxygenated PBS at different scan rates. The scan rates from inner to outer are: 20, in deoxygenated various buffer solutions (pH 1–11) at the scan rate of 50 mV s− 1. Inset
40, 60, 80, 100, 120, 140, 160, 180 and 200 mV s− 1. Inset shows the linear dependence shows the influence of pH on Epa ( ), Epc ( ) and E0′ ( ) of GCNT/GOx/GAD
of Ipa (●) and Ipc (○) on scan rate (20 to 200 mV s− 1). composite film.
 A.P. Periasamy et al. / Bioelectrochemistry 80 (2011) 114–120 117

and (c′) shows the AFM and SEM images of uniform GCNT/GOx/GAD GOx/GAD film modified GCEs in PBS containing 5 mM Fe(CN)3−/4− 6 .
composite film coated ITO. Here, closely packed larger bead like The inset shows the Randles equivalence circuit model. In the Nyquist
structures are seen. This validates higher GOx loading at GCNT matrix plot, a semicircle portion results from the parallel combination of
and it could be attributed to the larger surface area of GCNT. Thus SEM electron transfer resistance (Ret) and double layer capacitance (Cdl)
and AFM studies reveal the discriminate surface morphology between resulting from electrode impedance [23]. In the present study, EIS of
GOx, DCNT/GOx/GAD and DCNT/GOx/GAD films. GOx modified GCE exhibits a larger semicircle as shown in Fig. 5(a).
Whereas, depressed semicircle with smaller diameter is observed at
3.5. EIS studies at different film modified GCEs bare/GCE (see Fig. 5(b)). This indicates that greater hindrance to
electron transfer occurs at the GOx modified GCE rather than bare/
The GOx immobilization and its loading at various modified GCE. The increased electron transfer resistance observed at GOx
electrode surfaces have been investigated using EIS. Fig. 5(a) shows modified GCE could be due to the thick protein layer surrounding the
the real and imaginary part of the impedance spectra represented as FAD redox centre of GOx [24]. Interestingly, no significant semicircles
Nyquist plots (Zim vs. Zre) for only GOx, DCNT/GOx/GAD and GCNT/ are noticed at both DCNT/GOx/GAD and GCNT/GOx/GAD film

Fig. 4. AFM images of (a) GOx, (b) DCNT/GOx/GAD and (c) GCNT/GOx/GAD films. SEM images of (a′) GOx, (b′) DCNT/GOx/GAD and (c′) GCNT/GOx/GAD films.
118 A.P. Periasamy et al. / Bioelectrochemistry 80 (2011) 114–120

Fig. 6. Cyclic voltammograms recorded at GCNT/GOx/GAD composite film modified GCE

in (a) 0.05 and (b) 4.3 μM glucose containing oxygenated PBS. Scan rate: 50 mV s− 1.
Fig. 5. (a) EIS of GOx (●), DCNT/GOx/GAD (○) and GCNT/GOx/GAD (▼) film modified Similarly, cyclic voltammograms obtained at DCNT/GOx/GAD film modified GCE in (a′)
GCEs in PBS containing 5 mM Fe(CN)3− 4−
6 /Fe(CN)6 . (b) Shows the EIS of bare/GCE (■) in
0.05 and (b′) 4.3 μM glucose containing PBS under similar conditions as in (a) and (b).
PBS containing 5 mM Fe(CN)3− 4−
6 /Fe(CN)6 . Frequency: 0.1 Hz to 100 kHz. Inset shows
(c) Represents the electrocatalytic response of bare/GCE in highest glucose concentra-
the Randles circuit for above modified GCEs. tion (4.3 μM). Inset plot shows the linear dependence of peak current vs. [glucose]/μM
for GCNT/GOx/GAD (●) and DCNT/GOx/GAD (○) film modified GCEs.

modified GCEs in this frequency range (see Fig. 5(a)). This shows that
rapid electron transfer occurs at these modified GCEs which are
response towards glucose is higher than that observed at DCNT/GOx/
attributed to the excellent conductivity of GCNT and DCNT.
GAD film or bare/GCE (see Fig. 6(b′) and (c)). This shows the efficient
The Ret and Cdl values calculated for various modified GCEs are
electrocatalytic ability of the composite film towards glucose oxidation.
presented in Table S1. From Table S1 it is obvious that, lowest Ret value
Fig. 6 inset shows the linear dependence of peak currents on various
is observed at GCNT modified GCE i.e., Ret = 2.77 Ω. Comparatively,
glucose concentration additions at GCNT/GOx/GAD and DCNT/GOx/
DCNT modified GCE possesses about 17 times higher Ret value i.e.
GAD film modified GCEs. From the inset plot the correlation coefficients
48.88 Ω. The lower Ret value observed at GCNT modified GCE validates
of GCNT/GOx/GAD and DCNT/GOx/GAD film modified GCEs are
the higher conductance of GCNT than DCNT. In contrast, the Ret value
obtained as 0.9946 and 0.9666. Similarly, their sensitivity values are
noticed at GCNT/GOx/GAD modified GCE is found to be higher than
3.60 and 0.80 μA mM−1 cm− 2, respectively. The higher sensitivity
that of DCNT/GOx/GAD and any other modified GCEs discussed in
obtained at the composite film shows the good catalytic efficiency of
Table S1 (except GOx/GCE). The higher Ret value observed at GCNT/
GCNT than DCNTs. The proposed GCNT/GOx/GAD composite film
GOx/GAD composite film modified GCE could be due to the more
modified GCE exhibits good electrocatalytic response in the linear
amount of GOx present on the electrode surface which hinders the
concentration range from 1.04 to 3.76 mM of glucose.
electron transfer. Thus EIS studies show that more amount of GOx is
well immobilized at the GCNT/GOx/GAD composite film, which is in
good agreement with the surface coverage studies discussed in
3.7. Amperometric determination of glucose at GCNT/GOx/GAD
Section 3.1.
composite film modified GCE

3.6. Electrocatalysis of GCNT/GOx/GAD composite film towards oxygen Fig. 7 shows the amperometric i–t curve recorded at GCNT/GOx/GAD
reduction and glucose determination composite film modified rotating disc GCE in oxygenated PBS. The
electrode potential is kept constant at −0.44 V. Small shoulder peaks
The influence of dissolved oxygen at GCNT/GOx/GAD composite film appeared immediately for the successive addition of minimum glucose
modified GCE in deoxygenated and oxygenated PBS is shown in Fig. S3 concentrations (0.01 and 0.02 mM), as indicated by arrows in Fig. 7
(a) and (b) (see supplementary data). From the results it is obvious that lower inset. These shoulder peaks appeared as a result of oxygen
the reduction peak current observed at GCNT/GOx/GAD composite film reduction. With further increase in glucose addition, the oxygen
modified GCE in oxygenated PBS is higher than that observed in consumption increases and thus the catalytic current decreases
deoxygenated PBS. This result reveals that dissolved oxygen is reduced gradually and finally levels off for very high glucose concentration
effectively at GCNT/GOx/GAD composite film. Similar observations have (25.84 mM) (see Fig. 7 upper inset). The fabricated GCNT/GOx/GAD
been reported at GOx modified electrodes [17,25]. Further, the composite film modified GCE exhibits steady amperometric response
electrocatalytic response towards glucose has been comparatively towards glucose in the linear concentration range of 6.30 to 20.09 mM.
investigated at both GCNT/GOx/GAD and DCNT/GOx/GAD film modified The correlation coefficient is 0.9995. The sensitivity value at the
GCEs in oxygenated PBS. Fig. 6(a) shows that in the presence of low composite film is 2.47 μA mM−1 cm− 2 for glucose. In Table 1, the
glucose concentration (0.05 μM), a higher catalytic reduction current electroanalytical values observed at the GCNT/GOx/GAD composite film
corresponding to oxygen reduction is noticed at GCNT/GOx/GAD in this work are comparatively presented with other GOx modified
composite film than that of DCNT/GOx/GAD film on GCE (Fig. 6(a′)). electrodes reported previously [19,26–30]. From Table 1 it is clear that
This result shows that the composite film possesses higher electro- GCNT/GOx/GAD composite film possesses good linear range (6.30–
catalytic activity for glucose. Similarly, in the presence of high glucose 20.09 mM) for glucose which is significantly higher than the linear
concentration (4.3 μM) the catalytic reduction current observed at the range obtained at other GOx modified electrodes (see Table 1). The
composite film significantly decreases (see Fig. 6(b)). Thus glucose has other GOx modified electrodes used for comparison are GOx immobi-
been effectively oxidized at the composite film. This electrocatalytic lized at carbon nanotube-gold colloid composites with PDDA coatings
 A.P. Periasamy et al. / Bioelectrochemistry 80 (2011) 114–120 119

in the consecutive weeks. This result reveals that the composite film
exhibits a reasonable stability for two weeks. Here, the good stability
of GCNT/GOx/GAD composite film could be attributed to the stability
of GCNT.
The selectivity of the developed sensor is mandatory especially for
the determination of glucose from serum samples. In human serum,
glucose is present with other common interferences like ascorbic acid
(AA), uric acid (UA), acetaminophen (AP), pyruvate (PA) and lactate
(LA). The normal physiological levels of glucose, AA and AP in human
serum are about 5, 0.2 and 0.1 mM respectively [31,32]. Similarly the
normal level of lactate in the human serum is 0.52 ± 0.33 mM [33].
Since the normal level of most of the common interferences in human
serum is within 0.5 mM, we have used the same concentration for the
interference study. Similarly, the normal glucose level in human
serum (5 mM) has been used. The technique used for the interference
study is amperometric i–t curve. It is clear from Fig. S6 that the GCNT/
GOx/GAD composite film exhibits a sharp, rapid response towards
5 mM glucose injected into oxygenated PBS. However, no response is
Fig. 7. Amperometric i–t response at GCNT/GOx/GAD composite film modified rotating observed at the composite film when 0.5 mM of AA, UA, AP, PA and LA
disc GCE for the addition of 0.01 to 25.84 mM glucose into continuously stirred
oxygenated PBS. Applied potential: − 0.44 V; rotation rate: 900 RPM. The upper inset
is successively injected into the same oxygenated PBS (see Fig. S6).
shows the plot of linear dependence of peak current with [glucose]/mM. The lower This result shows that the GCNT/GOx/GAD composite film is highly
inset shows the enlarged view of amperometric i–t response noticed at GCNT/GOx/GAD selective towards glucose determination. This validates that the
composite film within 300 s for 0.01 and 0.02 mM glucose concentration additions into composite film is suitable for the selective glucose determination in
continuously stirred oxygenated PBS.
human serum under physiological conditions.

(GCE/CNT/Au/PDDA-GOx), gold nanoparticles-MWCNT multi layer

3.9. Real sample analysis
membrane (GNP/MWCNTs/GOx), poly pyrrole functionalized MWCNTs
(PPy-cMWCNTs-GOx), multilayer membrane with sodium dodecyl
The practical applicability of GCNT/GOx/GAD composite film has
sulfate functionalized MWCNT and PDDA on 3-mercapto-1-propane-
been corroborated through real sample analysis. The technique used
sulfonic acid, sodium salt self assembled gold/titanium deposited poly
is amperometric i–t curve. Serum samples were collected from a
(ethylene terephthalate) substrate [GOx/PDDA]3/[SDS-MWCNT/
healthy woman. During amperometric experiments, definite volume
PDDA]3-/MPS/Au/Ti/PET, poly(toluidine blue O) on CNT film (GCE/
of serum was injected into continuously stirred oxygenated PBS
CNT/PTBO-GOx) and chitosan-NdPO4 nanoparticles composite film
containing GCNT/GOx/GAD composite film modified rotating GCE.
(GOx/NdPO4 NPs/CHIT). The good linear range and the excellent
The working electrode potential is kept constant at −0.44 V. Fig. S5
electrocatalytic ability of the GCNT/GOx/GAD composite film reported
shows the amperometric i–t response obtained at GCNT/GOx/GAD
in the present study could be attributed to the good biocompatibility of
composite film modified GCE towards various glucose concentration
GCNT matrix for GOx.
additions. The composite film exhibits a rapid electrocatalytic
response to glucose in the linear range from 0.04 to 0.16 mM (see
3.8. Stability and interference studies Fig. S5 inset supplementary data). The correlation coefficient is
0.9781. Further, from the amount of added, found glucose concentra-
Fig. S4 shows the stability plot of GCNT/GOx/GAD composite film tions, recovery has been calculated and the results are given in Table
modified GCE stored in PBS at 4 °C for 4 consecutive weeks. About S2. From Table S2 it is clear that the GCNT/GOx/GAD composite film
96.37% peak current is retained at the composite film even after exhibits an excellent average recovery of 97.27% for glucose
2 weeks storage in PBS. However, the peak current drops significantly determination in human serum. The determination of glucose in
human serum with good linear range and excellent recovery shows
the good practical applicability of the composite film. Further, real
Table 1 sample analysis substantiates the unique ability of the GCNT matrices
Comparison of electroanalytical values for glucose oxidation at various GOx modified
in promoting the catalytic efficiency of GOx towards glucose.
electrodes.

Film type Epa (V) Sensitivity Correlation Linear Ref.

i–t coefficient concentration 4. Conclusions
range in mM

GCE/CNT/Au/ − 0.3 3.96 mA – 0.5–5 [19] We successfully investigated the direct electrochemistry of GOx at
PDDA-GOx M− 1 cm− 2 GCNT modified GCE. More stable aqueous dispersion of MWCNT was
GNP/MWCNTs/ − 0.2 7.3 mA 0.994 up to 9 [26] prepared by wrapping the hydrophobic walls of MWCNT with gelatin.
GOx mM− 1 cm− 2
The prepared GCNT was used as a novel matrix for GOx immobiliza-
PPy-cMWCNTs- + 0.9 95 nA mM− 1 – up to 4 [27]
GOx tion. GCNT present in the composite film greatly facilitates the
[GOx/PDDA]3/ + 0.6 5.6 μA 0.998 0.02–2.2 [28] electron transfer processes between GOx and GCE. The developed
[SDS-MWCNT/ mM− 1 cm− 2 glucose biosensor possesses significant characteristics like good
PDDA]3−/MPS/ biocompatibility, excellent selectivity, reasonable stability, rapid
Au/Ti/PET
GCE/CNT/PTBO- − 0.1 14.5 mA – 1–7 [29]
response, promising electrocatalytic activity towards glucose with
GOx mM− 1 cm− 2 good linear range. Moreover, it exhibits rapid electrocatalytic
GOx/NdPO4 + 0.4 1.92 μA 0.999 0.15–10 [30] response towards glucose in human serum samples without any
NPs/CHIT mM− 1 common interference at physiological pH. The excellent recovery
GCNT/GOx/GAD − 0.44 2.47 μA 0.9995 6.30–20.09 This
achieved at the composite film shows the robust performance of the
mM− 1 cm− 2 work
developed biosensor towards glucose determination.
120 A.P. Periasamy et al. / Bioelectrochemistry 80 (2011) 114–120

Acknowledgements [14] W. Zheng, Y.F. Zheng, Gelatin-functionalized carbon nanotubes for the bioelec-
trochemistry of hemoglobin, Electrochem. Commun. 9 (2007) 1619–1623.
[15] L. He, Q.W. Di, L.C. Hua, L.L. Bai, Z.G. Li, A novel gelatin-carbon nanotubes hybrid
This work was supported by the National Science Council and the hydrogel, Macromol. Biosci. 3 (2003) 720–724.
Ministry of Education of Taiwan (Republic of China). [16] S. Haider, S. Park, K. Saeed, B.L. Farmer, Swelling and electroresponsive
characteristics of gelatin immobilized onto multi-walled carbon nanotubes,
Sens. Actuators B 124 (2007) 517–528.
[17] C. Shan, H. Yang, J. Song, D. Han, A. Ivaska, L. Niu, Direct electrochemistry of
Appendix A. Supplementary data glucose oxidase and biosensing for glucose based on graphene, Anal. Chem. 81
(2009) 2378–2382.
Supplementary data associated with this article can be found, in [18] A. Salimi, E. Sharifi, A. Noorbakhsh, S. Soltanian, Immobilization of glucose oxidase
on electrodeposited nickel oxide nanoparticles: direct electron transfer and
the online version, at doi: 10.1016/j.bioelechem.2010.06.009. electrocatalytic activity, Biosens. Bioelectron. 22 (2007) 3146–3153.
[19] Y. Yao, K. Shiu, Direct electrochemistry of glucose oxidase at carbon nanotube-
gold colloid modified electrode with poly(diallyldimethylammonium chloride)
References coating, Electroanalysis 20 (2008) 1542–1548.
[20] E. Laviron, General expression of the linear potential sweep voltammogram in the
[1] M. Musameh, J. Wang, A. Merkoci, Y. Lin, Low-potential stable NADH detection at case of diffusionless electrochemical systems, J. Electroanal. Chem. 101 (1979)
carbon nanotube-modified glassy carbon electrodes, Electrochem. Commun. 4 19–28.
(2002) 743–746. [21] N. Jia, L. Liu, Q. Zhou, L. Wang, M. Yan, Z. Jiang, Bioelectrochemistry and enzymatic
[2] R.T. Kachoosangi, M.M. Musameh, I.A. Yousef, J.M. Yousef, S.M. Kanan, L. Xiao, S.G. activity of glucose oxidase immobilized onto the bamboo-shaped CNx nanotubes,
Davies, A. Russell, R.G. Compton, Carbon nanotube-ionic liquid composite sensors Electrochim. Acta 51 (2005) 611–618.
and biosensors, Anal. Chem. 81 (2009) 435–442. [22] D. Wen, Y. Liu, G. Yang, S. Dong, Electrochemistry of glucose oxidase immobilized
[3] B. Wu, G. Zhang, Y. Zhang, S. Shuang, M.M.F. Choi, Measurement of glucose on the carbon nanotube wrapped by polyelectrolyte, Electrochim. Acta 52 (2007)
concentrations in human plasma using a glucose biosensor, Anal. Biochem. 340 5312–5317.
(2005) 181–183. [23] S. Zong, Y. Cao, H. Jua, Direct electron transfer of hemoglobin immobilized in
[4] V. Serafin, L. Agui, P.Y. Sedeno, J.M. Pingarron, Glucosinolate amperometric multiwalled carbon nanotubes enhanced grafted collagen matrix for electro-
bienzyme biosensor based on carbon nanotubes-gold nanoparticles composite catalytic detection of hydrogen peroxide, Electroanalysis 19 (2007) 841–846.
electrodes, Electroanalysis 21 (2009) 1527–1532. [24] J. Wang, Electrochemical glucose biosensors, Chem. Rev. 108 (2008) 814–825.
[5] B. Wu, S. Hou, M. Yu, X. Qin, S. Li, Q. Chen, Layer-by-layer assemblies of chitosan/ [25] R. Garjonyte, A. Malinauskas, Amperometric glucose biosensor based on glucose
multi-wall carbon nanotubes and glucose oxidase for amperometric glucose oxidase immobilized in poly (o-phenylenediamine) layer, Sens. Actuators B 56
biosensor applications, Mater. Sci. Eng. C 29 (2009) 346–349. (1999) 85–92.
[6] D. Ragupathy, A.I. Gopalan, K. Lee, Synergistic contributions of multiwall carbon [26] Y. Liu, S. Wu, H. Ju, L. Xu, Amperometric glucose biosensing of gold nanoparticles
nanotubes and gold nanoparticles in a chitosan-ionic liquid matrix towards and carbon nanotube multilayer membranes, Electroanalysis 19 (2007) 986–992.
improved performance for a glucose sensor, Electrochem. Commun. 11 (2009) [27] Y. Tsai, S. Li, S. Liao, Electrodeposition of polypyrrole-multiwalled carbon
397–401. nanotube-glucose oxidase nanobiocomposite film for the detection of glucose,
[7] S. Wang, L. Lu, M. Yang, Y. Lei, G. Shen, R. Yu, A novel cobalt hexacyanoferrate Biosens. Bioelectron. 22 (2006) 495–500.
nanocomposite on CNT scaffold by seed medium and application for biosensor, [28] X.B. Yan, X.J. Chen, B.K. Tay, K.A. Khor, Transparent and flexible glucose biosensor
Anal. Chim. Acta 651 (2009) 220–226. via layer-by layer assembly of multi-wall carbon nanotubes and glucose oxidase,
[8] R.B. Rakhi, K. Sethupathi, S. Ramaprabhu, A glucose biosensor based on deposition Electrochem. Commun. 9 (2007) 1269–1275.
of glucose oxidase onto crystalline gold nanoparticle modified carbon nanotube [29] Y. Yao, K. Shiu, Low potential detection of glucose at carbon nanotube modified
electrode, J. Phys. Chem. B 113 (2009) 3190–3194. glassy carbon electrode with electropolymerized poly (toluidine blue O) film,
[9] D. Wen, X. Zou, Y. Liu, L. Shang, S. Dong, Nanocomposite based on depositing Electrochim. Acta 53 (2007) 278–284.
platinum nanostructure onto carbon nanotubes through a one-pot, facile [30] S. Qinglin, L. Kai, L. Lei, Z. Jianbin, Direct electrochemistry of glucose oxidase
synthesis method for amperometric sensing, Talanta 79 (2009) 1233–1237. immobilized on NdPO4 nanoparticles/chitosan composite film on glassy carbon
[10] S. Deng, G. Jian, J. Lei, Z. Hub, H. Ju, A glucose biosensor based on direct electrodes and its biosensing application, Bioelectrochemistry 74 (2009)
electrochemistry of glucose oxidase immobilized on nitrogen-doped carbon 246–253.
nanotubes, Biosens. Bioelectron. 25 (2009) 373–377. [31] E. Scavetta, B. Ballarin, D. Tonelli, A cheap amperometric and optical sensor for
[11] C. Xiao, X. Chu, B. Wu, H. Pang, X. Zhang, J. Chen, Polymerized ionic liquid- glucose determination, Electroanalysis 22 (2010) 427–432.
wrapped carbon nanotubes: the promising composites for direct electrochemistry [32] H. Tang, J. Chen, S. Yao, L. Nie, G. Deng, Y. Kuang, Amperometric glucose biosensor
and biosensing of redox protein, Talanta 80 (2010) 1719–1724. based on adsorption of glucose oxidase at platinum nanoparticle-modified carbon
[12] L. Agui, P.Y. Sedeno, J.M. Pingarron, Role of carbon nanotubes in electroanalytical nanotube electrode, Anal. Biochem. 331 (2004) 89–97.
chemistry, Anal. Chim. Acta 622 (2008) 11–47. [33] T. Nishijima, M. Nishina, K. Fujiwara, Measurement of lactate levels in serum and
[13] L. Yurong, L.M. Geever, J.E. Kennedy, C.L. Higginbotham, P.A. Cahill, G.B. bile using proton nuclear magnetic resonance in patients with hepatobiliary
McGuinness, Thermal behavior and mechanical properties of physically cross- diseases: its utility in detection of malignancies, Jpn. J. Clin. Oncol. 27 (1997)
linked PVA/gelatin hydrogels, J. Mech. Behav. Biomed. Mater. 3 (2010) 203–209. 13–17.