Electrosynthesis of polyaniline–mutilwalled carbon nanotube nanocomposite films in the presence of sodium dodecyl sulfate for glucose biosensing

MWCNTs (with >95% purity) were purchased from Shenzhen Nanotech Port (China). Glucose oxidase (GOx) EC 1.1.3.4 (from Aspergillus niger), 25% glutaraldehyde and SDS were used as received (Sigma Aldrich). Aniline (Merck) was distilled under reduced pressure and stored below 0 °C before use. β-d-(+)-glucose, H₂SO₄, KCl, KH₂PO₄, K₂HPO₄ and other reagents were of analytical grade and used without further purification. The glucose stock solution was allowed to mutarotate for 24 h in order to reach the equilibrium of α- and β- anomers at room temperature prior to use. Phosphate buffered saline (PBS, 20 mM KH₂PO₄ + 20 mM K₂HPO₄ + 0.1 M KCl, pH 6.8) was used as supporting electrolytes. All of the solutions were prepared with doubly distilled water.

Electrochemical measurements were performed on an Autolab PGSTAT12 Electrochemical Analyser (EcoChemie, Netherlands) under the control of GPES version 4.9. The experiments were carried out using a conventional three-electrode system with interdigitated planar platinum-film microelectrodes (IDμE) as the working electrode, a Pt wire as the auxiliary electrode, and a saturated calomel electrode (SCE) as the reference electrode. Field-emission scanning electron microscope (FE-SEM) images of the films were recorded with Hitachi S-4800 (Japan).

The interdigitated platinum-film planar microelectrodes (IDμE) were fabricated on a silicon wafer using a photolithographic metal lift-off process. Two pairs of the interdigitated electrodes were prepared by sputtering 50 nm Cr and 200 nm Pt on a layer of silicon dioxide (SiO₂) with a thickness of around 100 nm by means of dry thermal oxidation on top of the wafer. The microelectrode digit width was of 50 μm and the gap between them was 20 μm, and their length was around 1 mm (figure 1).

Zoom In Zoom Out Reset image size

Two milligrams of MWCNTs were dispersed in 0.1 mM of SDS aqueous solution and sonicated for 1 h [8,13]. Then aniline monomer solution (0.01 M of aniline in 0.1 M of H₂SO₄) was dissolved in SDS and MWCNTs solution under ultrasonic stirring for 15 min at room temperature. The PANi–MWCNTs modified IDμE was fabricated by the potentiostatic method at a constant potential of +0.9 V versus SCE. After polymerization the electrode was rinsed with distilled water and placed into a PBS solution. For comparison, a PANi/IDμE was prepared under the same conditions.

Fresh solutions of glucose oxidase (GOx) were prepared in PBS and stored at 4 °C. The immobilization of GOx on the surface of the PANi–MWCNTs film was achieved by cross-linking through the glutaraldehyde agent. Firstly, the PANi–MWCNTs electrodes were placed in saturated glutaraldehyde vapor for 30 min then dried in air for 15 min at room temperature. Next, 8 μl of the GOx solution (2 U μl⁻¹) was spread on the PANi–MWCNTs surface using a drop method. The product, the formed GOx/PANi–MWCNT/Pt electrode, was rinsed thoroughly with PBS solution and stored at 4 °C when not in use.

Cyclic voltammetry (CV) was used to characterize the redox properties of PANi–MWCNTs modified IDμE. The cyclic voltammograms obtained for both PANi and PANi–MWCNTs composite electrodes at a scan rate of 50 mV s⁻¹ in the potential range of −0.20 to +1.0 V are shown in figure 2.

Zoom In Zoom Out Reset image size

The current density of PANi–MWCNTs electrode is larger than that of the PANi electrode. Three identifiable anodic peaks with cathodic counterparts are observed for the PANi–MWCNTs electrode but not clearly for the PANi electrode. Here MWCNTs played an important role in increasing electroactivity and/or its specific surface area of the electrode. It significantly improves the overall biosensor performance.

Figures 3(A) and (C) show the low magnification scanning electron micrographs obtained for PANi–MWCNTs and GOx/PANi–MWCNTs films on IDμE, respectively. The SEM image of PANi–MWCNTs shows spongy and porous morphology of polyaniline. The high magnification SEM of the PANi–MWCNT sample (figure 3(B)) indicates that carbon nanotubes dispersed in the polymer matrix and formed the homogeneous nanocomposites. It can be helpful in the entrapment of enzyme molecules. When GOx was added and allowed to covalently attach to PANi–MWCNT surface, the films became thicker and constituted a gel configuration (figure 3(D)). These small globules indicate the presence of enzyme (GOx) onto PANi–MWCNT surface.

Zoom In Zoom Out Reset image size

Figure 4 shows the cyclic voltammograms (CV) of GOx/PANi–MWCNT/IDμE in the absence (curve a) and in the presence (curves b and c) of glucose in the PBS solution. The range of potential was from 0.0 to +0.8 V at a scan rate of 50 mV s⁻¹.

Zoom In Zoom Out Reset image size

The CV curves of the electrodes corresponding to the addition of different glucose concentrations show an increase in anodic current when the potential varied from +0.2 to +0.8 V in comparison to that without glucose addition. It is worth noting that there are no changes in shape and position of the cyclic voltammograms when glucose concentrations were added. Here GOx has been successfully immobilized onto the PANi–MWCNT surface. It is also suggests that the range of potential response was due to hydrogen peroxide generation during the enzymatic reaction [14].

As can be seen in figure 4, the large response current was obtained within the range of +0.2 to +0.8 V. However, at high potential there may be interferences from other oxidable species such as ascorbic acid (AA), uric acid (UA), and acetaminophen (AAP). They commonly present in physiological samples of glucose and cause problems in the accuracy of the glucose determination. In this study the effect of the applied potential on amperometric response for glucose sensor was investigated.

The amperometric responses of AA, UA, AAP, and glucose (1 mM) in the potential range of +0.2 to +0.8 V are shown in figure 5. Commonly, the susceptibility to interference species is positively correlated to the potential of the sensor. As a sequence, the lower potential results in the lower susceptibility and the increase of the sensor selectivity. When the applied potential is higher, the response current increases, however, it is strongly influenced by the interferences. When the applied potential increased from +0.6 to +0.7 V, the sensor currents of AA, UA, and AAP increased from 4.7 to 8.8, 3.6 to 8.89, and 1.9 to 4.8 μA, respectively. The errors in determination of 1 mM glucose at a potential of +0.6 V in solution caused by 1 mM AA, 1 mM UA, and 1 mM AAP are 22.38, 17.14 and 8.81%, respectively. Compared to the normal physiological levels that have been detected from blood samples, which are 0.13 mM for AA, 0.3 mM UA, and 0.125 mM for AAP [15], the biosensor exhibits essentially no response to such interfering species. Thus, the value of +0.6 V was chosen as applied potential of the GOx/PANi–MWCNT sensor.

Zoom In Zoom Out Reset image size

Figure 6 shows amperometric response at the GOx/PANi–MWCNT/IDμE in PBS at +0.6 V (versus SCE) for the successive addition 1 mM of glucose.

Zoom In Zoom Out Reset image size

The bioelectrode exhibits a rapid and sensitive current response to each injection of glucose concentration. The time required to reach 95% of the maximum steady-state current was less than 5 s. The plot of current versus glucose concentration (inset, figure 3(C)) depicted a linear relationship over the range of glucose (1–12 mM) until saturation current at a high glucose-concentration (>12 mM). The linear regression equation was I (μA) = 1.07 × C_glucose (mM) + 20.06 with the correlation coefficient of 0.99039. This linear range is clinically sufficient for monitoring of the glucose levels in normal human blood, 4.4 − 6.6 mM [16].