Microchemical Journal 119 (2015) 6674

Contents lists available at ScienceDirect

Microchemical Journal
journal homepage: www.elsevier.com/locate/microc

Improvement of the detection limit for biosensors: Advances on the

optimization of biocomposite composition
R. Montes, J. Bartrol, M. Baeza , F. Cspedes
Grup de Sensors i Biosensors, Departament de Qumica, Facultat de Cincies, Edici C-Nord, Universitat Autnoma de Barcelona, 08193 Cerdanyola del Valls (Bellaterra), Spain

a r t i c l e

i n f o

Article history:
Received 28 July 2014
Received in revised form 18 September 2014
Accepted 11 November 2014
Available online 14 November 2014
Keywords:
Amperometric biosensor
EIS
CV
Graphiteepoxy
GOD
(Bio)composite

a b s t r a c t
In this work the application of advanced characterization techniques in the development of amperometric biosensors based on biocomposites is described. The optimization of the conductive particle distribution and the
amount of the biological material inside the biomaterial have allowed an improvement of the electrochemical
properties, regarding the electroanalytical properties such as signal stability and limit of detection. The high
signal-to-noise ratio obtained in the electrochemical transduction has allowed enhancing the limit of detection
of the biosensor. In the present study, it has been demonstrated the feasibility of electrochemical impedance
spectroscopy (EIS) and cyclic voltammetry (CV) for the characterization and optimization of biosensors based
on graphiteepoxyenzyme, using an enzyme model. The optimum biocomposite proportion based on graphiteepoxy which incorporates the enzyme glucose oxidase (GOD) on the matrix ranges between 16% and 17%
of graphite using 1% and 2% of enzyme. This range provides the optimal electroanalytical properties. Low limit
of detection and good sensitivity have been achieved. Furthermore, confocal laser scanning microscopy was
used to visualize the enzyme distribution onto the surface electrode.
2014 Elsevier B.V. All rights reserved.

1. Introduction
The need of detecting small amounts of different kinds of compounds, which are usually present in a complex matrices, has led to
the development of new sensitive, economical and robust (bio)sensor
devices based on composite materials that allow to perform in situ
and real analysis [1,2]. Moreover, the use of conductive materials
as transducers or conductive phase based on carbon materials
(i.e. graphite, carbon nanotubes, etc.) and dispersed on a polymeric
matrix (i.e. epoxy, methacrylate, Teon, etc.) has been opened a new
generation of rigid conducting composites that have been applied in
the realization of (bio)sensors. Electrodes obtained by using a mixture
of particulate conductive carbon phase and an insulating matrix represent an attractive approach for the fabrication of electrochemical (bio)
sensors, whose surfaces can be renewed by polishing [36]. Biocompatibility and capability to incorporate chemical species without their loss
in operating medium are of utmost importance [7]. A biosensor based
on a biocomposite is dened as a rigid material made by combining
two or more materials of different nature (phases) where at least one
of them has a biological origin [8]. An important aspect of the development of biosensors is the method of immobilization of the enzyme [9].
The biological component of biosensors has traditionally been placed
on the surface of the transducer, either by direct adsorption [10],

Corresponding author. Tel.: +34 935814927.

E-mail address: mariadelmar.baeza@uab.cat (M. Baeza).

http://dx.doi.org/10.1016/j.microc.2014.11.004
0026-265X/ 2014 Elsevier B.V. All rights reserved.

cross-linking [11,12] or covalent attachment [13,14] or immobilized

previously on a membrane [15]. Another alternative strategy proposed
in the development of biocomposites is the immobilization of the
biological component inside the matrix of the composite by physical
entrapment [1620]. This immobilization that allows forming a rigid
and renewable sensing surface showed excellent performance for enzymatic determinations.
An important feature of composite electrodes is that their overall
analytical performance is strongly inuenced by the carbon loading
within polymeric matrix. It is due to carbon loading that inuences
directly on the electrochemical surface and inner structure (bulk
resistance) of the composite electrode [21,22]. Both parameters strongly affect on the overall electroanalytical performance of such composite
electrodes [23]. The characterization and optimization of composites
based graphiteepoxy have been widely studied using different strategies based on several techniques as well as percolation theory
[2426], atomic force microscopy (AFM) [3,25] or chronoamperometry
[3,24]. Up to now the principle applied to the optimization of the composite proportions has been done using the percolation theory, under
the criteria of maximizing the conductive particle loading, without
losing its physical and mechanical stability, but without taking care if
the composite provides the best electroanalytical characteristics of
response [24,2729]. Recently, it has established new alternative
strategies of characterization which demonstrates that if the composite
proportions are optimized the response of the electrode is improved
[30] in terms of the signal-to-noise ratio which has a direct relationship
with the limit of detection.

R. Montes et al. / Microchemical Journal 119 (2015) 6674

These techniques are electrochemical impedance spectroscopy

(EIS) and cyclic voltammetry (CV). EIS measurements provide, in
an easy way, information about the electron-transfer rate, double-layer
capacitance, contact resistance and resistance of the solution [31,32].
The electrochemical properties required by a transducer are high
electron-transfer rate, the lowest double-layer capacitance and ohmic resistance in order to guarantee the optimal electroanalytical characteristics
of the electrode response as high sensitivity, a high signal/noise ratio, and
low detection limits. Consequently, by EIS technique it is possible to determine the optimal composite composition that exhibits these improved
electroanalytical properties. These results can also be complemented with
voltammetric measurements. This is the rst time that these strategies
are applied in the characterization and optimization of biocomposite electrodes in order to develop more efcient biosensors for determining low
analyte concentrations in a nal analytical application.
The study of the electrochemical properties of the biocomposites as
a function of the conductive particle and biological charge material present on the transducer matrix has been performed. In the development
of amperometric biosensors based on biocomposites, the incorporation
of a biological compound produces a modication in the spatial separation and inner distribution of the conductive particles. The electrochemical response of a biosensor based on a biocomposite depends on
the electron-transfer rate and also the active area of the electrode.
Moreover, the electronic-transfer on the surface electrode depends
on the conductive particle loading and hence on their distribution
[33,34]. The improvement of the electrochemical properties of the biosensor is due to an appropriate distribution of the conductive particles
inside the biocomposite and, in consequence, in the biosensor surface.
In the literature some references have been found which realize a
study of the inuence of the enzyme load. On one side, Prez et al. [35]
reported a study of L-lactate biosensors based on polysulfone/carbon
nanotubes membranes where the criterion of optimization of the
amount of enzyme presented on the biosensor is the amount that provides the wider linear range according to the analytical requirements.
On the other side, Wang et al. [36] reported a study based on enzyme
dispersed on carbon nanotubes for monitoring glucose where the criterion of optimizing the enzyme loading followed is the amount that provides the best electroanalytical signal. Moreover, under the same
criterion, Caro-Jara et al. [37] optimize the GOD/HRP ratio in a
bienzymatic amperometric biosensor. However, in any case the optimization of the transducer has been considered.
The main goal of this study is the use of alternative strategies
of characterization [30], in order to characterize and optimize the
biocomposite composition based on graphiteepoxy which incorporates different amounts of an enzyme model on the polymeric matrix.
Enzymatic amperometric glucose biosensors have been widely studied
in the last four decades [38] because of the relatively high durability of
the enzyme and low cost, so glucose oxidase (GOD) has been chosen as
an enzyme model for this study.
Firstly, we have constructed two series of graphiteepoxyGOD with
different graphite loadings between the percolation threshold zone and
amounts of GOD. We have applied the electrochemical strategies of
characterization in order to obtain the optimized biocomposite composition. The analytical response of the optimized biocomposite has been
evaluated with synthetic samples of glucose. Besides the electroanalytical characterization, the surface of the biosensors has been also characterized by optical techniques such as uorescence microscopy.

67

(Crison instruments, Alella, Barcelona, Spain), an AgCl covered silver

wire and the constructed graphite biocomposite electrodes were used
as a counter, reference, and working electrodes, respectively.
Amperometric measurements were done using an amperimeter LC4C (Bio analytical Systems Inc., West Lafayette, IN, USA), connected to a
personal computer by data acquisition card ADC-42 Pico Technology
(St. Neots, Cambridgeshire, UK) for data registering and visualization.
Electroanalytical experiments were carried out in 20 mL glass cell, at
room temperature (25 C), using three-electrode conguration. A single
junction reference electrode Ag/AgCl Orion 900100 (Thermo Electron
Corporation, Beverly, MA, USA) and platinum-based electrode were
used as reference and auxiliary, respectively. The graphite biocomposite
electrodes were used as working electrode. A magnetic stirrer provided
the convective transport during the amperometric measurements.
Confocal laser scanning microscopy microphotographs were taken
with a LEICA TCS SP2 microscope.
2.2. Chemical reagents
Graphite powder (particle size 50 m) was received from Merck
(Merck Millipore, Darmstadt, Germany). Epoxy resin Epotek H77A and
hardener Epotek H77B were obtained from Epoxy Technology (Epoxy
Technology, Billerica, MA, USA). Potassium ferricyanide/ferrocyanide
(99.8%), potassium chloride (99.5%), potassium phosphate monobasic (99.5%), nitric acid (65%), potassium dibasic-anhydrous (98%),
D -(+)-glucose (99.5%) and glucose oxidase type VII from Aspergillus
niger (174,400 units/g) were supplied from Sigma-Aldrich (St. Louis,
MO, USA) and used without further purication. EZ-Link Sulfo-NHSLC-Biotin was purchased from Thermo Scientic (Rockford, USA). All
the dissolutions were prepared using deionized water from Milli-Q
system (Millipore, Billerica, MA, USA).
2.3. Fabrication of the electrodes
2.3.1. Working electrodes
Handmade graphiteepoxy composites were prepared following the
conventional methodology previously established in the research group
[39]. A resin Epotek H77 and their corresponding hardener compound
were mixed in the ratio 20:3 (w/w). The graphite composite was prepared by loading different amounts of graphite (13, 14, 15, 16, 17, 18,
19 and 20% (w/w)) into the epoxy resin before hardening. The composite was homogenized for 30 min. After the homogenization time, the
glucose oxidase amount (1% and 2% for each series, respectively) was introduced to the composite paste and homogenized for 15 min more. The
nal biocomposite paste electrode was allowed to harden during 5 days
at 40 C and when not in use it was stored at 4 C [16]. Finally the electrode surface was polished with different sandpapers of decreasing
grain size. The nal electrode dimensions were 28 mm2 and 3 mm for
its geometric area and thickness, respectively.
2.3.2. GraphiteepoxyGOD electrodes for uorescence measurements
In order to perform the uorescence measurements, graphiteepoxy
GOD biocomposite was prepared as discussed in Section 2.3.1, but it was
introduced into a special support. Hand-made biocomposite was placed
into a PVC disk with an external diameter of 35 mm, internal diameter
of 15 mm, and thickness of 3 mm. The process of hardening was 5 days
at 40 C and when not in use it was stored at 4 C. The electrode surface
was polished using different sandpapers of decreasing grain size.

2. Experimental
2.4. Procedure
2.1. Apparatus
Electrochemical impedance spectroscopy and voltammetric measurements were performed using a computer controlled Autolab PGSTAT12
potentiostat/galvanostat (Eco Chemie, Utrecht, The Netherlands) with
a three-electrode conguration. A platinum-based electrode 53-671

2.4.1. Electrochemical characterization

EIS and voltammetric measurements were made in a 0.1 M potassium
chloride solution containing 0.01 M potassium ferricyanide/ferrocyanide
under quiescent condition. Amperometric detection was made under
force convection by stirring the solution with magnetic stirrer.

68

R. Montes et al. / Microchemical Journal 119 (2015) 6674

2.4.2. Microscopic characterization of biosensor surface

The distribution of the glucose oxidase enzyme on the electrode surface was studied by confocal laser scanning uorescence microscopy.
In this instance a uorescent conjugate was used for the labeling of
the graphiteepoxyGOD electrode surface by covalent bond between
the amino groups of GOD and the carboxyl groups of biotin. The background control was performed using non-biotinylated uorescein surface of graphiteepoxyGOD electrode.
The biocomposite surface was labeled in two steps. An incubation
step with 100 L of biotin/streptavidinCy5 solution (10 mM of biotin
and 0.02 mg/mL of streptavidinCy5 uorescent dye) in PBS (0.1 M
and pH = 7.6) at 25 C during 60 min was performed. After that,
three washing steps with PBS buffer (0.1 M and pH = 7.6) at 25 C
during 5 min were realized.
3. Results and discussion
3.1. Electrochemical impedance spectroscopy measurements
By means of electrochemical impedance spectroscopy we could obtain general trends in the electrochemical parameters of the graphite
epoxyenzyme biocomposites such as the solution resistance or the
ohmic resistance (R) (which comprises the resistance of the solution
(Rs) plus any contact resistance (Rc)), the charge-transfer resistance
(Rct) and the double-layer capacitance (Cdl) for composites with different graphite loadings and amount of GOD. These parameters were obtained by tting the impedance spectra to an equivalent circuit. This
circuit was sufciently suitable to interpret the R, Rct and Cdl values
in terms of interfacial phenomena that occur at the electrochemical
cell [30]. Such electrochemical characterization was performed in the
4
presence of benchmark redox species such as Fe(CN)3
6 /Fe(CN)6
which is very sensitive to the electrode surface characteristics [40].
The impedance spectra were recorded in the frequency range 0.1 Hz
to 100 kHz at the redox equilibrium potential. The signal amplitude to
perturb the system was 10 mV.
Two series of eight different compositions of graphiteepoxy composite electrodes were constructed by varying the graphite loading
from 13% to 20% and the amount of glucose oxidase was 1% and 2%,
respectively. The interval of graphite composition was selected based
on the percolation curve for the graphiteepoxy composite previously
established in an earlier work [25]. This interval of composition corresponds to the low resistivity region, where it is important to highlight
that the graphite loading has to be high enough to guarantee a
conducting pathway and sufciently low bulk resistivity. For each
graphiteepoxyGOD composition, three equal electrodes were fabricated and evaluated. The corresponding parameters of EIS, the ohmic
resistance (R), charge-transfer resistance (Rct) and the double-layer
capacitance (Cdl) were obtained by tting the impedance spectra to a
simple equivalent circuit proposed (see inset Fig. 1). Using such approach, we attempted to obtain a general trend of these physical parameters as a function of the graphite loading, using constant amount of
glucose oxidase for each series.
The impedance response of biocomposites with 1% and 2% of GOD is
depicted in Fig. 1A and 1B, respectively. As we can see, the impedance
behavior is dominated by small diameter semicircle representing
kinetic-controlled electrode process, though in some cases the diffusion
control starts to be discerned at low frequencies, when the amount of
graphite is high corresponding to the low resistivity area (from 20% to
16% of graphite loading). On the other hand, the impedance plot for
composites with low graphite loading (from 15% to 13% of graphite
loading) appears to be dominated by a big diameter semicircle and
only the kinetic-controlled electrode process is present in the recorded
frequency range. We could observe that when the amount of GOD increases from 1% (Fig. 1A) to 2% (Fig. 1B) the diameters of the impedance
response for each composition are higher. So, an increase of the GOD
amount produces an increase of the kinetics-controlled electrode

process and the electron-transfer is more limited. This can be associated

with the separation of the conductive particles due to the increase of
enzyme load.
Fig. 2A shows the variations of the ohmic resistance as a function of
the graphite composition for both GOD amounts. Ionic concentration,
type of ion, electrode area and dry resistance of the composite are factors that inuence on this resistance. At low graphite loads, the ohmic
resistance is dominated by composite resistance whereas at higher
graphite loads the ohmic resistance decreases to low values and it is
dominated by the solution resistance [25,26]. In general trends, the
ohmic resistance values decrease when the graphite loading increases
for both GOD concentrations. The increase of the ohmic resistance becomes more signicant when the enzyme load increases. Therefore,
the modication of the matrix by a third element has an inuence on
the active area of the electrode because the conductive material is separated by the enzyme particles, decreasing the active area of the electrode and increasing the value of the ohmic resistance. Consequently,
in order to assure the sensitivity and low response time it is important
that ohmic resistance is low.
The quantitative values of Rct are depicted in Fig. 2B. This parameter
is inversely proportional to the heterogeneous charge-transfer rate and
also affects the sensitivity and response time of the electrode. A decrease of the Rct is observed with the increase of the graphite load
which indicates the strong relation between electrochemical reactivity
and the surface characteristics of the conducting material. As carbon
load increases, the probability of having more electroactive sites increases and hence the electrode kinetics. Therefore low values in the
charge-transfer resistance allow increasing the electroanalytical applications of the electrode. As well as in ohmic resistance, the values of
Rct are higher when the GOD amount increases. The normalization of
this parameter by the electrochemical electroactive area should be constant in metal electrode surface (see Table 1(A) and 1(B)). Graphite as a
carbon material exhibits electrochemical anisotropy as well as it was reported for CNTs [30]. Low values on the charge-transfer resistance of the
biocomposites will guarantee fast electron-transfer and their appropriate use in electroanalytical applications. According to the results obtained biocomposites between 20% and 16% of graphite loading and both
amounts of GOD presented lower charge-transfer resistance values.
Finally, it is important to consider the double-layer capacitance
value which is directly related to the charging or background current.
Fig. 2C depicts the decrease of the double-layer capacitance values
with the decrease of the graphite loadings with constant GOD amount.
An increase of the values of Cdl is observed when the amount of GOD
increases on biocomposite electrodes with the same load of graphite.
High proportions of graphite increase remarkably the background current because the proportion of conducting material exposed on the surface area becomes higher and that not enhances the signal-to-noise
ratio and consequently increased the detection limit of the analyte.
Therefore, an increase on the amount of GOD increases the Cdl value
due to the presence of the enzyme on the matrix that reduces the
graphite portion on the surface.
Based on the impedance results obtained for the biosensors and
taking into account the properties required by an electrode for electroanalytical purposes, such as rapid response time, low limit of detection
and high sensitivity, the interval between 16% and 17% of graphite loading seems to fulll these requirements, for both GOD concentrations
evaluated. The fabricated composites in such interval of composition
depicted similar electrochemical behavior, an increase of the electrochemical reproducibility and analytical properties improved.
3.2. Cyclic voltammetry characterization
In order to complement the results obtained by EIS technique we
have also performed cyclic voltammetry (CV) measurements. Cyclic
voltammograms were taken for the different biocomposite composition
4
electrodes in the presence of the benchmark Fe(CN)3
6 /Fe(CN)6 redox

R. Montes et al. / Microchemical Journal 119 (2015) 6674

(A)

2000
1800
1600
1400
1200

Z (Imag)

69

13% graphite
14% graphite
15% graphite
16% graphite
17% graphite
18% graphite
19% graphite
20% graphite

Cdl
Rs

Rc
Rct

1000
800
600
400
200
0
-500

500

1000

1500

2000

2500

3000

3500

2000

2500

3000

3500

Z (Real)

(B)

2000
1800
1600
1400

Z (Imag)

1200

13% graphite
14% graphite
15% graphite
16% graphite
17% graphite
18% graphite
19% graphite
20% graphite

1000
800
600
400
200
0
-500

500

1000

1500

Z (Real)
4
Fig. 1. Nyquist plots for different graphite loading electrodes with (A) 1% and (B) 2% of GOD in the presence of Fe(CN)3
6 /Fe(CN)6 . The equivalent circuit used for the impedance spectra
tting is shown in (A).

couple under the same experimental condition (10 mV/s of scan). For
both GOD concentrations, it is important to highlight that none of the
compositions exhibit sigmoidal shaped voltagramms so there is no electrochemical behavior as a microelectrode array where radial diffusion
dominates mass transport; at least in the interval of graphite studied.
Biocomposite electrodes exhibit the typical peak-shaped [4,6,41] prole
corresponding to more massive electrodes with planar diffusion where
linear diffusion controls mass transport (see Fig. 3A and 3B). Different parameters were extracted from the cyclic voltagramms such as the peak
separation potential (E) and peak current (Ip) as shown in Table 1 (A
and B) for each series of graphiteepoxyGOD biocomposite. As we can
see in Table 1, there is an increase of the peak current with the graphite
loading increases associated with an increase of the electroactive area,
together with a decrease of peak separation related to an enhancement

of the electron-transfer rate. This behavior is observed for both GOD concentrations. The relative electroactive area was estimated from the peakshaped voltammograms by quantifying the peak current with the use of
this relationship, Ip = 3.01 105 n3/2 ( Dred )1/2 A Cred [31], which is
appropriate for electron-transfer controlled process. In this equation
corresponds to the transfer coefcient which was considered to be approximately 0.5, Dred = 6.32 106 cm2s1 corresponds to the diffusion
coefcient of the reduced species, = 0.01 Vs1 represents the scan rate,
A is the electroactive area and Cred = 0.01 M is the bulk concentration of
the electroactive species. We also evaluate the exchange current (io) from
Tafel plots (log current vs. potential), a parameter which provides information about the reversibility of the process. From the value of the exchange current we can also evaluate the charge-transfer resistance
through the relation io = RT/nFRct. The charge-transfer resistance values

Fig. 2. Values of (A) ohmic resistance, (B) charge transfer resistance and (C) double-layer capacitances, with their corresponding standard deviation (n = 3) for the different graphite
4
loading electrodes with 1% and 2% of GOD, using the redox probe Fe(CN)3
6 /Fe(CN)6 .

70

R. Montes et al. / Microchemical Journal 119 (2015) 6674

(A) 10000
1% GOD
2% GOD

R (ohms)

8000
6000
4000
2000
0
12

13

14

15

16

17

18

19

20

21

% graphite

(B)

5000
1% GOD
2% GOD

Rct (ohms)

4000
3000
2000
1000
0
12

13

14

15

16

17

18

19

20

21

18

19

20

21

% graphite

(C) 1.80e-4
1.60e-4

1% GOD
2% GOD

1.40e-4

Cdl (F)

1.20e-4
1.00e-4
8.00e-5
6.00e-5
4.00e-5
2.00e-5
0.00

12

13

14

15

16

17

% graphite

R. Montes et al. / Microchemical Journal 119 (2015) 6674

obtained by EIS are following the same trend like the results obtained
by CV technique. We have normalized the Rct with respect to the
electroactive area and it is observed that there is a decrease on the
value when the active area increases showing the evident inuence of
the electrochemical anisotropy of the graphite which can be more noticeable as the graphite loading is increased. From the comparison of the normalized Rct values extracted from cyclic voltammetry with the ones
obtained by electrochemical impedance spectroscopy measurements it
can be observed that the values agree quite well being more similar in
the optimal range of composition.

3.3. Electroanalytical performance

Glucose was used as analyte for evaluating the electroanalytical
properties. When GOD and glucose interact, hydrogen peroxide is produced at the biosensor surface. In this case, hydrogen peroxide produced by the biocatalytic reaction can be amperometrically measured
by direct oxidation on the surface of the biosensor. Measurements
were carried out at 1050 mV polarization potential xed. The response
of biosensor to changes in glucose concentration was evaluated by
hydrodynamic amperometric measurements for all biocomposite compositions studied.
The analytical parameters such as detection limit, sensitivity and
linear range were evaluated for 16% and 17% of graphite composite
with 1% and 2% of GOD and compared to those obtained with 20% of
graphite loading with the same amount of GOD. For each composition
three electrodes were evaluated.
As we can see from Table 2, the sensitivity increases when the graphite loading increases and the limit of detection increases remarkably,
using the same amount of glucose oxidase. The experimental results
show that in the optimal composite composition range (1617% of
graphite loading with 12% of GOD), there are no signicant differences
regarding the sensitivity and limit of detection for both amounts of GOD.
However, we can observe that the biocomposite with 16% of graphite
loading and 1% of GOD presents a wider linear range. When the amount
of GOD decreases from 2% to 1%, the sensitivity is slightly inferior but the
detection limit remains constant. Nevertheless, when it is compared
with the 20% of graphite loading, in spite of the sensitivity and the linear
range increased, the limit of detection also increases in one order of
magnitude. These results show that the biocomposites in the optimal
composition range provide quite good sensitivity, low limit of detection
and a wide linear range.

71

In the literature it has been reported that non-optimized electrodes

based on graphiteepoxyGOD (18:80:2 (w/w)) which compared to
the results obtained with the optimized biocomposite electrodes had
lower sensitivity and higher limit of detection [18]. Hence, this fact
reafrms that complementary strategies to optimize the electrochemical properties of biosensors, both conducting material and enzyme material load, improve the electroanalytical properties of the biosensors
developed.
3.4. Microscopic characterization of biosensor surface
Confocal laser scanning uorescence microscopy was used to evaluate the distribution of the biological compound in the working electrode
surface. This study was performed with the biotinylation and marking
with a uorescein dye of the surface electrode (see Experimental
section 2.4.2). The compositions evaluated in this study are 16% and
20% of graphite with both amounts of GOD. The surface of different
biocomposite electrodes with different amounts of graphite load (16%
and 20%) and glucose oxidase (1% and 2%) is shown in Fig. 4. The intense
uorescence area corresponds to the enzyme biding with the uorescent
marker. After this study uorescence microscopy conrms that the glucose oxidase is distributed homogenously in the electrode surface.
4. Conclusion
The proposed electrochemical techniques, such as electrochemical
impedance spectroscopy and cyclic voltammetry, as well as hydrodynamic amperometry are suitable for the characterization and optimization of biocomposites based on graphiteepoxyenzyme. By these
alternative techniques, and especially EIS, provide a versatile tool to optimize the biocomposite composition. Some relevant parameters can be
extracted from EIS, as ohmic resistance which is related to the percolation theory, charge-transfer resistance which can be related to the heterogeneous electron-transfer and which depends on the surface
electrochemical reactivity (active area) and nally the electrode capacitance which can be correlated to the background current and consequently with the signal/noise ratio. As a result, all these parameters
are relevant to the biosensor analytical response and the evaluation of
these parameters allows choosing the biocomposite composition
which fullls the electroanalytical requirements of high sensitivity,
fast response and low limits of detection. In the present work the optimized composition ranges between 16% and 17% of graphite loading

Table 1
Cyclic voltammetry parameters for the different composite compositions with (A) 2% of GOD and (B) 1% of GOD. Io corresponds to the exchange current, Rct to the charge transfer resistance, Ip to peak current, A to active area and E to the peak separation potencial. Rct A and REIS
ct A correspond to the Rct obtained by voltammetric and EIS measurements, respectively,
and normalized with respect to the active area.
Electrodes

io (A)

(A)
13%
14%
15%
16%
17%
18%
19%
20%

1.93
3.03
5.93
1.51
1.79
2.20
5.27
8.46

106
106
106
105
105
105
105
105

(B)
13%
14%
15%
16%
17%
18%
19%
20%

2.60
5.97
1.08
1.85
2.00
2.34
6.01
7.76

106
106
105
105
105
105
105
105

Rct ()

Ip (A)

12873.2
8327.2
4254.9
1671.0
1409.6
1146.9
478.8
298.2

7.80
8.21
1.01
1.23
1.37
1.43
1.73
2.03

8.26
1.00
1.23
1.35
1.35
1.46
1.82
2.06

9704.7
4226.36
2336.2
1363.9
1261.6
1078.3
419.8
325.2

A (cm2)

E (V)

Rct A ( cm2)

2
REIS
)
ct A ( cm

105
105
104
104
104
104
104
104

0.15
0.15
0.19
0.23
0.26
0.27
0.32
0.38

1.5666
1.3245
0.9720
0.6622
0.5732
0.5091
0.2563
0.1780

1931
1249
789
384
366
310
153
113

469
400
441
276
264
250
97
77

105
104
104
104
104
104
104
104

0.15
0.19
0.23
0.25
0.25
0.27
0.34
0.38

1.5595
1.0895
0.7868
0.5803
0.4949
0.4733
0.2314
0.1923

1456
803
537
341
315
291
143
124

456
337
348
193
175
173
44
60

72

R. Montes et al. / Microchemical Journal 119 (2015) 6674

(A)

3,00e-4

13% graphite
14% graphite
15% graphite
16% graphite
17% graphite
18% graphite
19% graphite
20% graphite

2,00e-4

Intensity (A)

1,00e-4

0,00

18000
16000

Rct (ohms)

14000

-1,00e-4

12000
10000
8000
6000
4000

-2,00e-4

2000
0
12

-3,00e-4
-1,5

13

14

15

16

17

18

19

20

21

% graphite

-1,0

-0,5

0,0

0,5

1,0

1,5

2,0

2,5

Potential (V)

(B)

3,00e-4
13 % grafito
14 % grafito
15 % grafito
16 % grafito
17 % grafito
18 % grafito
19 % grafito
20 % grafito

2,00e-4

Intensity(A)

1,00e-4

0,00

18000
16000

Rct (ohms)

14000

-1,00e-4

12000
10000
8000
6000
4000

-2,00e-4

2000
0
12

13

14

15

16

17

18

19

20

21

% graphite

-3,00e-4
-1,5

-1,0

-0,5

0,0

0,5

1,0

1,5

2,0

Potential (V)
Fig. 3. Cyclic voltammogram for biocomposites with (A) 1% of GOD and (B) 2% of GOD using 0.01 M ferricyanide/ferrocyanide and 0.1 M KCl. Scan rate 10 mV/s. The inset gure shows the
trend of charge transfer resistance for different biocomposite compositions.

for graphiteepoxyGOD model electrodes for both GOD proportions.

Small variations in compositions around this optimal range do not produce high changes in the electrochemical behavior. To conrm these
Table 2
The calibration parameters for 20%, 17% and 16% of graphite biocomposite electrode with
1% and 2% of GOD using amperometric measurements with glucose as analyte and PBS
0.1 M at pH = 7.0 as background electrolyte.
% Graphite

% GOD

Sensitivitya
(ALmmol1)
(% RSD)

LOD
(mmolL1)

Linear range
(mmolL1)

16

1
2
1
2
1
2

0.185
0.21
0.214
0.231
0.74
1.35

0.037
0.046
0.036
0.035
0.265
0.24

0.0371.28
0.0460.70
0.0360.70
0.0350.70
0.2650.81
0.241.80

17
20
a

n = 3, 95% condence level.

0.003 (2%)
0.01 (7%)
0.005 (3%)
0.009 (4%)
0.01 (2%)
0.05 (4%)

0.001
0.008
0.007
0.009
0.006
0.01

predictions, we showed the electroanalytical detection of glucose

which demonstrates that although the sensitivity decreases, compared
with 20% graphite; we have obtained a better limit of detection (one
order of magnitude lower) and wider linear range. It is important to
highlight that these results showed that to obtain low limit of detection
the amount of graphite loading it is more critical than the enzyme loading in biocomposites. Moreover, there are no signicant differences in
the electroanalytical parameters between 1% and 2% of GOD, when it
used the same graphite loading. Finally, in the present work it has
been demonstrated the applicability of these techniques in the characterization and optimization of biosensors based on biocomposites
which will allow us to apply them to other biological compounds for
their application in the determination of pesticides, DNA, immunoassay,
food safety, etc., where it is very important to achieved low analyte
concentrations. In addition, the homogeneous and stable immobilization of the enzyme in the biocomposite surface by entrapment has
been demonstrated by confocal laser microscopy.

R. Montes et al. / Microchemical Journal 119 (2015) 6674

73

A1

A2

B1

B2

C1

C2

D1

D2

Fig. 4. Confocal laser scanning uorescence microphotograph for biocomposites with 16% with (A1) 1% and (C1) 2% of GOD and 20% of graphite load and (B1) 1% and (D2) 2% of GOD submitted to biotinylated uorescein. Microphotographs (A2), (B2), (C2) and (D2) correspond to non-biotinylated uorescein. Laser excitation was 620 nm. Voltage was 352 V.

74

R. Montes et al. / Microchemical Journal 119 (2015) 6674

Acknowledgments
R. Montes thanks Universitat Autnoma de Barcelona (UAB) for the
award of PIF studentship.

[20]

[21]

References
[22]
[1] R. Oliv-Monllau, A. Pereira, J. Bartrol, M. Baeza, F. Cspedes, Highly sensitive CNT
composite amperometric sensors integrated in an automated ow system for the
determination of free chlorine in waters, Talanta 81 (2010) 15931598, http://dx.
doi.org/10.1016/j.talanta.2010.03.008.
[2] A.B. Ibez, A. Guts, M. Baeza, F. Cspedes, Electronic tongue applied to phenolic
compounds analysis, Anal. Lett. 41 (2008) 14191429.
[3] D.O. Hare, J.V. Macpherson, A. Willows, On the microelectrode behaviour of
graphiteepoxy composite electrodes, Electrochem. Commun. 4 (2002) 245250.
[4] S. Ramrez-Garca, S. Alegret, F. Cespedes, R.J. Forster, Carbon composite electrodes:
surface and electrochemical properties, Analyst 127 (2002) 15121519, http://dx.
doi.org/10.1039/b206201a.
[5] L. Rassaei, M. Sillanp, M.J. Bonn, F. Marken, Carbon nanoberpolystyrene
composite electrodes for electroanalytical processes, Electroanalysis 19 (2007)
14611466, http://dx.doi.org/10.1002/elan.200703887.
[6] J. Wang, Practical considerations, Anal. Electrochem.Wiley-VCH, New York, 2000,
pp. 100139.
[7] D. Bellido-Milla, L.M. Cubillana-Aguilera, M. El Kaoutit, M.P. Hernndez-Artiga, J.L.
Hidalgo-Hidalgo de Cisneros, I. Naranjo-Rodrguez, et al., Recent advances in
graphite powder-based electrodes, Anal. Bioanal. Chem. 405 (2013) 35253539,
http://dx.doi.org/10.1007/s00216-013-6816-2.
[8] F. Cespedes, S. Alegret, New materials for electrochemical sensing II. Rigid carbon
polymer biocomposites, Trends Anal. Chem. 19 (2000) 276284.
[9] A. Numnuam, P. Thavarungkul, An amperometric uric acid biosensor based on
chitosancarbon nanotubes electrospun nanober on silver nanoparticles, Anal.
Bioanal. Chem. 406 (2014) 37633772, http://dx.doi.org/10.1007/s00216-0147770-3.
[10] A. Parra, E. Casero, L. Vzquez, F. Pariente, E. Lorenzo, Design and characterization of
a lactate biosensor based on immobilized lactate oxidase onto gold surfaces, Anal.
Chim. Acta. 555 (2006) 308315, http://dx.doi.org/10.1016/j.aca.2005.09.025.
[11] S.A. Marzouk, V.V. Cosofret, R.P. Buck, H. Yang, W.E. Cascio, S.S. Hassen, A conducting
salt-based amperometric biosensor for measurement of extracellular lactate accumulation in ischemic myocardium, Anal. Chem. 69 (1997) 26462652 http://
www.ncbi.nlm.nih.gov/pubmed/9230678.
[12] M.M. Barsan, C.M.A. Brett, A new modied conducting carbon composite electrode
as sensor for ascorbate and biosensor for glucose. Bioelectrochemistry 76 (2009)
135140, http://dx.doi.org/10.1016/j.bioelechem.2009.03.004.
[13] F. Palmisano, R. Rizzi, D. Centonze, P.G. Zambonin, Simultaneous monitoring of
glucose and lactate by an interference and cross-talk free dual electrode amperometric biosensor based on electropolymerized thin lms, Biosens. Bioelectron. 15
(2000) 531539 http://www.ncbi.nlm.nih.gov/pubmed/11419650.
[14] J. Haccoun, B. Piro, L.D. Tran, L.A. Dang, M.C. Pham, Reagentless amperometric detection of l-lactate on an enzyme-modied conducting copolymer poly(5-hydroxy-1,4naphthoquinone-co-5-hydroxy-3-thioacetic acid-1,4-naphthoquinone), Biosens.
Bioelectron. 19 (2004) 13251329, http://dx.doi.org/10.1016/j.bios.2003.11.006.
[15] S. Snchez, M. Pumera, E. Fbregas, Carbon nanotube/polysulfone screen-printed
electrochemical immunosensor, Biosens. Bioelectron. 23 (2007) 332340, http://
dx.doi.org/10.1016/j.bios.2007.04.021.
[16] F. Cspedes, E. Martnez-Fbrega, J. Bartrol, S. Alegret, Amperometric enzymatic
glucose electrode based on an epoxy graphite composite, Anal. Chim. Acta. 273
(1993) 409417.
[17] J. Wang, L. Fang, D. Lopez, H. Tobias, Highly selective and sensitive amperometric
biosensing of glucose at ruthenium-dispersed carbon-paste enzyme electrodes,
Anal. Lett. 26 (1993) 18191830.
[18] B. Perez, M. Pumera, M. del Valle, A. Merkoci, S. Alegret, Glucose biosensor based on
carbon nanotube epoxy composites, J. Nanosci. Nanotechnol. 5 (2005) 16941698.
[19] E. Zacco, M.I. Pividori, X. Llopis, M. Del Valle, S. Alegret, Renewable Protein A
modied graphiteepoxy composite for electrochemical immunosensing, J.

[23]

[24]

[25]

[26]

[27]

[28]

[29]
[30]

[31]

[32]

[33]
[34]
[35]

[36]

[37]

[38]

[39]

[40]
[41]

Imunol. Methods 286 (2004) 3546, http://dx.doi.org/10.1016/j.jim.2003.11.

014.
M.D. Rubianes, G. Rivas, Enzymatic biosensors based on carbon nanotubes paste
electrodes, Electroanalysis 17 (2005) 7378, http://dx.doi.org/10.1002/elan.
200403121.
M. Baeza, R. Oliv-Monllau, M.J. Esplandiu, Advances on rigid conducting composites for electroanalytical applications, in: Deborah B. Son (Ed.), Resin Compos.
Prop. Prod. Appl.Nova Science Publishers, Inc., 2011, pp. 153211.
M.J. Esplandiu, M. Baeza, R. Oliv-Monllau, F. Cspedes, Chapter 5: Development of
Tunable Nanocomposites Made from Carbon Nanotubes for Electrochemical
Applications, 2010.
H. Zhao, D.O. Hare, Characterisation and modeling of conducting composite electrodes, J. Phys. Chem. C 112 (2008) 93519357, http://dx.doi.org/10.1021/
jp711366u.
S. Ramrez-Garca, S. Alegret, F. Cspedes, R.J. Forster, Carbon composite microelectrodes: charge percolation and electroanalytical performance, Anal. Chem. 76
(2004) 503512, http://dx.doi.org/10.1021/ac034536p.
R. Oliv-Monllau, M. Baeza, J. Bartrol, F. Cspedes, Novel amperometric sensor
based on rigid near-percolation composite, Electroanalysis 21 (2009) 931938,
http://dx.doi.org/10.1002/elan.200804494.
J. Trijueque, J.J. Garca-Jareo, J. Navarro-Laboulais, A. Sanmatas, F. Vicente, Ohmic
drop of Prussian-blue/graphite + epoxy electrodes, Electrochim. Acta 45 (1999)
789795, http://dx.doi.org/10.1016/S0013-4686(99)00257-1.
M. Pumera, A. Merkoi, S. Alegret, Carbon nanotubeepoxy composites for electrochemical sensing, Sensors Actuators B Chem. 113 (2006) 617622, http://dx.doi.
org/10.1016/j.snb.2005.07.010.
S. Ramrez-Garca, F. Cespedes, S. Alegret, Development of conducting composite
materials for electrochemical sensing in organic media, Electroanalysis 13 (2001)
529535.
F. Cespedes, E. Martnez-Fbregas, S. Alegret, New materials for electrochemical
sensing I. Rigid conducting composites, Trends Anal. Chem. 15 (1996) 296304.
R. Oliv-Monllau, M.J. Esplandiu, J. Bartrol, M. Baeza, F. Cspedes, Strategies for the
optimization of carbon nanotube/polymer ratio in composite materials: applications
as voltammetric sensors, Sensors Actuators B Chem. 146 (2010) 353360, http://dx.
doi.org/10.1016/j.snb.2010.02.017.
M. Pacios, M. del Valle, J. Bartroli, M.J. Esplandiu, Electrochemical behavior of rigid
carbon nanotube composite electrodes, J. Electroanal. Chem. 619620 (2008)
117124, http://dx.doi.org/10.1016/j.jelechem.2008.03.019.
M.J. Esplandiu, M. Pacios, L. Cyganek, J. Bartroli, M. del Valle, Enhancing the electrochemical response of myoglobin with carbon nanotube electrodes, Nanotechnology
20 (2009) 355502, http://dx.doi.org/10.1088/0957-4484/20/35/355502.
N. Bojorge, E. Alhadeff, Graphite-composites alternatives for electrochemical biosensor,
in: J. Cuppoletti (Ed.), Met. Ceram. Polym. Compos. Var. UsesInTech, 2011, pp. 597620.
A. Sadana, N. Sadana, Handbook of Biosensors and Biosensor Kinetics, Elsevier, B.V.,
2011
S. Prez, S. Snchez, E. Fbregas, Enzymatic strategies to construct L-lactate biosensors based on polysulfone/carbon nanotubes membranes, Electroanalysis 24
(2012) 967974, http://dx.doi.org/10.1002/elan.201100628.
J. Wang, M. Musameh, Enzyme-dispersed carbon-nanotube electrodes: a needle microsensor for monitoring glucose, Analyst 128 (2003) 1382, http://dx.doi.org/10.
1039/b309928h.
N. Caro-Jara, R. Mundaca-Uribe, C. Zaror-Zaror, J. Carpinelli-Pavisic, M. ArandaBustos, C. Pea-Farfal, Development of a bienzymatic amperometric glucose
biosensor using mesoporous silica (MCM-41) for enzyme immobilization and its
application on liquid pharmaceutical formulations, Electroanalysis 25 (2013)
308315, http://dx.doi.org/10.1002/elan.201200391.
J. Wang, Glucose biosensors: 40 years of advances and challenges, Electroanalysis 13
(2001) 983988, http://dx.doi.org/10.1002/1521-4109(200108)13:12b983::AIDELAN983N3.0.CO;2-#.
A. Morales, F. Cespedes, J. Muoz, S. Alegret, E. Martinez-Fbregas, Hydrogen peroxide amperometric biosensor based on a biocomposite, Anal. Chim. Acta. 332 (1996)
131136.
R.L. McCreery, Advanced carbon electrode materials for molecular electrochemistry,
Chem. Rev. 108 (2008) 26462687, http://dx.doi.org/10.1021/cr068076m.
K. Stulik, C. Amatore, K. Holub, V. Marecek, W. Kutner, Microelectrodes. Denitions,
characterization and applications, Pure Appl. Chem. 72 (2000) 14831492.