Glucose biosensor based on electrodeposition of platinum nanoparticles onto carbon nanotubes and immobilizing enzyme with chitosan-SiO2 sol–gel

Available online at www.sciencedirect.com

Biosensors and Bioelectronics 23 (2008) 1010–1016

Glucose biosensor based on electrodeposition of platinum nanoparticles onto carbon nanotubes and immobilizing enzyme with chitosan-SiO2 sol–gel Yongjin Zou a,b , Cuili Xiang a,b , Li-Xian Sun a,∗ , Fen Xu a,∗∗ a

Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China b Graduate School of the Chinese Academy of Sciences, Beijing 100049, China

Received 6 July 2007; received in revised form 10 October 2007; accepted 12 October 2007 Available online 25 October 2007

Abstract A novel amperometric biosensor, based on electrodeposition of platinum nanoparticles onto multi-walled carbon nanotube (MWNTs) and immobilizing enzyme with chitosan-SiO2 sol–gel, is presented in this article. MWNTs were cast on the glass carbon (GC) substrate directly. An extra Nafion coating was used to eliminate common interferents such as acetaminophen and ascorbic acids. The morphologies and electrochemical performance of the modified electrodes have been investigated by scanning electron microscopy (SEM) and amperometric methods, respectively. The synergistic action of Pt and MWNTs and the biocompatibility of chitosan-SiO2 sol–gel made the biosensor have excellent electrocatalytic activity and high stability. The resulting biosensor exhibits good response performance to glucose with a wide linear range from 1 ␮M to 23 mM and a low detection limit 1 ␮M. The biosensor also shows a short response time (within 5 s), and a high sensitivity (58.9 ␮Am M−1 cm−2 ). In addition, effects of pH value, applied potential, rotating rate, electrode construction and electroactive interferents on the amperometric response of the sensor were investigated and discussed in detail. © 2007 Elsevier B.V. All rights reserved. Keywords: Pt nanoparticles; Chitosan-SiO2 sol–gel; Glucose biosensor; Multi-walled carbon nanotubes

1. Introduction Enzymatic biosensors based on immobilizing bioactive enzymes and related materials have been an attractive and popular field for scientific research (Sun et al., 1998; Yuan et al., 2005; Zou et al., 2007). Due to their rapid response, high sensitivity and intrinsic selectivity, biosensors have many potential applications in various fields, including medical diagnostics, pharmaceuticals, and environmental control. Electrochemical sensors, especially amperometric biosensors, hold a leading position among biosensors. In the case of electrochemical biosensor, the effective immobilization of an enzyme on an electrode surface with a high retention of its biological activity

∗

Corresponding author at: Materials & Thermochemistry Laboratory, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China. Tel.: +86 411 84379213; fax: +86 411 84379213. ∗∗ Corresponding author at: Materials & Thermochemistry Laboratory, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China. Fax: +86 411 84379213. E-mail addresses: [email protected] (L.-X. Sun), [email protected] (F. Xu). 0956-5663/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2007.10.009

is a crucial point for the commercial development of biosensors. During recent years, numerous approaches had been taken to immobilize enzymes onto solid substrates including adsorption, covalent coupling and tethering via an intermediate linker molecule (Yang et al., 2006a,b; Tang et al., 2004; Ricci et al., 2003). Among various approaches, the method of sol–gel processing is particularly advantageous to the immobilization of biomolecules (Wang et al., 2000; Lin et al., 2007). Braun et al. reported the attempt to encapsulate proteins inside SiO2 glasses in 1990 firstly (Braun et al., 1990). Since then, these new kinds of inorganic materials are particularly attractive to the development of electrochemical biosensor. They can be prepared under mild conditions and exhibit tunable porosity, high thermal stability, chemical inertness, ability to form films and negligible swelling in aqueous and non-aqueous solution (Yang et al., 2004). The enzymes and proteins can be immobilized within sol–gel matrices whilst maintaining their native properties and reactivates, which make a potential tool for the development of biosensors. Moreover, sufficient amount of trapped interstitial water in gels plays an important role in the retention of the tertiary structure and active reactivity of

Y. Zou et al. / Biosensors and Bioelectronics 23 (2008) 1010–1016

encapsulated biomolecules. The pore size in sol–gel can be controlled into an appropriate size for the diffusion of the analyte to the redox active sites, at the same time preventing enzyme leakage. This property is suitable of development of new biosensors. Unfortunately, it has also been observed that some gels lose their activities after the first few cycles, primarily ascribed to leaching of enzymes from the sol–gel matrix (Shi et al., 2007). Recently, the organic–inorganic sol–gels have received significant interest because the incorporation of organic polymers in the inorganic sol–gel can lead to new composite materials possessing the properties of each component that would be useful in particular applications. Incorporation of organic polymers, especially those with amino or amide groups, allows the formation of molecular hybrids often stabilized by strong hydrogen bonding. Polymers such as chitosan are able to form hybrids with SiO2 gel by using this method. It has been shown that chitosanSiO2 possesses a highly porous microstructure and demonstrates an excellent performance for enzyme immobilization (Yang et al., 2004). However, to the best of our knowledge, there is no report on fabrication of biosensor by immobilizing enzymes with chitosan-SiO2 gels. On the other hand, carbon nanotubes (CNTs) have emerged as new class nanomaterials that are receiving considerable interest because of their unique structure, high chemical stability and high surface-to-volume ratio. These properties make them extremely attractive for fabricating chemical sensors (Lim et al., 2005; Wang et al., 2003). Recently, composite materials based on integration of CNTs and some other materials to possess properties of the individual components with a synergistic effect have gained growing interest (Zhang and Gorski, 2005). Materials for such purposes include conducting polymers, redox mediators and metal nanoparticles. For example, coupling CNTs with Pt nanoparticles resulted in remarkable improvement of the electroactivity of the composite materials toward H2 O2 . With glucose oxidase (GOD) as an enzyme model, they constructed a glassy carbon or carbon fiber microelectrode-based glucose biosensor (Hrapovic et al., 2004). Lim et al. (2005) reported a glucose biosensor based on electrodeposition of palladium nanoparticles and GOD onto Nafion-solubilized carbon nanotube electrode. In the present paper, the Pt/MWNTs nanocomposite was fabricated in a simple and robust way. Platinum nanoparticles can be grown by electrodeposition onto MWNTs directly with the average diameter of the nanoparticles about 30–40 nm. The resulting Pt/MWNTs material brings new capabilities for electrochemical devices by using the synergistic action of Pt nanaoparticles and CNTs. The immobilization of glucose oxidase onto electrode surfaces was carried out by chitosan-SiO2 gel. After casing amount of Nafion solution on the gel, the biosensor was fabricated. The electrochemical behavior of the modified electrode has been investigated by amperometric method. The influence of pH value, applied potential, rotating rate, electrode construction and electroactive interferents on the sensor performance is also evaluated. The resulted biosensor exhibits high sensitivity, good reproducibility, long-term stability and freedom of interference from other co-existing electroactive species.

1011

2. Experimental 2.1. Reagents The MWNTs (95% 20–60 nm) were purchased from Shenzhen Nanotech. Port. Co., Ltd. (Shenzhen, China). The MWNTs were treated with concentrated nitric acid (70%) at 120 ◦ C for 24 h and then filtered, rinsed with deionized water and dried. H2 PtCl6 ·6H2 O was obtained from Beijing Chemical Reagent Co. (Beijing, China). A 10 mM H2 PtCl6 +0.1 M HCl stock solution was prepared for electrodeposition of Pt nanoparticles. Chitosan (MW 1.9–3.1 × 105 ; 92.5% deacetylation) was purchased from Nantong Shuanglin (China). It was refined twice by dissolving it in dilute acetic acid solution, filtered, precipitated with aqueous NaOH, and finally dried in vacuum at room temperature. Tetraethoxysilance (TEOS) was purchase from Shanghai reagent company (Shanghai China). GOD (from Aspergillus niger, 300,000 unit g−1 ) was purchased from Sanland (America). d-Glucose was used without further purification and glucose solutions were stored overnight at room temperature before use. Nafion 117 solutions (1 wt%) were prepared by dilution with alcohols of 5 wt% Nafion 117 solutions. All other chemicals from commercial source were of analytical grade and used as received. 0.1 M phosphate buffer solution (PBS), which was made from Na2 HPO4 and NaH2 PO4 , was employed as supporting electrolyte. Deionized water was used throughout the experiments. 2.2. Apparatus and measurements The electrodeposition of Pt nanoparticles and amperometric measurements were performed using an IM6e electrochemical workstation (Zahner-Elektrik, Kronach, Germany). All electrochemical experiments were carried out with a conventional three-electrode system. The working electrode was the Bioanalytical Systems (BAS) cavity GC electrode (3-mm diameter). Prior to each experiment, the GC electrode was polished successively with 1 ␮m, 0.3 ␮m and 0.05 ␮m ␣-alumina powder, rinsed thoroughly with double-distilled water between each polishing step, utrasonicated successively in 1:1 (v/v) nitric acid, acetone and double-distilled water and then allowed to dry at room temperature. The rotating disk electrode experiments were performed by BAS rotator system in conjunction with an IM6e. The rotating rate is 4000 rpm when detect H2 O2 and glucose unless stated otherwise. An Ag/AgCl (saturated with NaCl) reference electrode was used for all measurements and all the potentials were reported in this paper versus this reference electrode. A platinum wire was used as a counter electrode. Before all batch amperometric experiments, the potential of each electrode was held at the operating value, allowing the background current to decay to a steady-state value. Cyclic voltammetric (CV), electrochemical impedance spectroscopy (EIS) measurements were performed in the presence of 5 mM K3 [Fe(CN)6 ]/K4 [Fe(CN)6 ] (1:1) and 0.1 M KCl. For EIS tests, a frequency range of 100 KHz–0.1 Hz was utilized with potential amplitude of 5 mV. The electrolyte solution was purged with high-purity nitrogen for at least 15 min prior to

1012

Y. Zou et al. / Biosensors and Bioelectronics 23 (2008) 1010–1016

CV and EIS experiments and a nitrogen environment was then maintained for the solution in the cell during the measurement process. All experiments were performed at room temperature (20 ± 1 ◦ C). Scanning electron microscopy (SEM) images were obtained by using JSM-6360LV SEM (JEOL, Japan). 2.3. Preparation of chitosan-SiO2 sol–gel One percent chitosan solution was prepared by dissolving chitosan in 2% acetic acid solution with magnetic stirring for about 2 h, The chitosan-SiO2 sol–gel was prepared by mixing 2 ml of TEOS and 2 ml of above chitosan solution in a beaker under magnetic stirring at room temperature. The mixture initially composed of two phases was made uniform by stirring vigorously till the SiO2 -containing phase distributed evenly in the aqueous solution while the hydrolysis reaction took place. After 3 h, the opaque and white sol was formed. 2.4. Fabrication of the modified electrodes The fabrication of the Pt/MWNTs glassy carbon electrode was summarized as follows. Two milligrams of purified MWNTs was dispersed in 5 ml dimethylformamide (DMF) with the aid of ultrasonic agitation to give a 0.4 mg ml−1 black suspension. The GC electrode was treated by dropping a suspension (5 ␮l) of the MWNTs in DMF and then dried under an infrared lamp. The MWNT/GC obtained was thoroughly rinsed with water. Electrodeposition of platinum on MWNTs/GC electrode was carried out in an electroplating bath. The composition of the electroplating bath consisted of 10 mM H2 PtCl6 and 0.1 M HCl, The MWNTs/GC-modified electrode was immersed in the plating bath and a constant potential of −0.20 V was applied for 5 min. The amount of loaded Pt nanoparticles was evaluated from the charge consumed during electrodeposition, assuming that Pt4+ to Pt0 reduction is 100% efficient. For the preparation of the enzyme electrode, 10 mg of GOD was dissolved in 2 ml of 0.1 M phosphate buffer solution (PBS, pH 7) and 250 ␮l of the mixture was added to 250 ␮l of the chitosan-SiO2 sol–gel. The above mixture was hand-mixed thoroughly and a homogenous solution was obtained. Then 5 ␮l of the mixture was dropped on the surface of the Pt/MWNTs/GC electrode. After drying for 10 h at 4 ◦ C in refrigerator, a 1% (3 ␮l) Nafion solution was placed on the enzyme surface as protective film. Before the electrochemical measurements, the enzyme electrodes were washed with PBS (pH 7.0). 3. Results and discussion 3.1. Characterization of the modified GC Fig. 1 shows the SEM images of the modified glassy carbon electrodes at the optimized condition, (a) MWNTs (b) Pt/MWNTs (c) GOD-Chitosan-SiO2 /Pt/MWNTs. As shown in Fig. 1a, the porous MWNTs film has large surface area which provides an ideal matrix for the distribution of Pt. The SEM images also reveal that the MWNTs, with a diameter ranging

Fig. 1. SEM images of the modified electrodes. (a) MWNTs/GC, (b) Pt/MWNTs/GC and (c) GOD-Chitosan-SiO2 /Pt/MWNTs/GC.

from 30 nm to 60 nm, are well distributed on the surface and that most of the MWNTs are in the form of small bundles or single tubes. Such small bundles and single tubes assembled homogeneously on the substrate are believed to be very beneficial for the modified electrode. From Fig. 1b, it can be seen the Pt nanoparticles (389 ␮g cm−2 ) are dispersed well on the MWNTs substrate. The particle size is about 30–40 nm. The lattice planes of Pt are also clearly visible indicating the crystalline nature of catalytic

Y. Zou et al. / Biosensors and Bioelectronics 23 (2008) 1010–1016

Fig. 2. CVs at (a) GC electrode, (b) MWNTs/GC electrode and (c) Pt/MWNTs/GC electrode. Supporting electrolyte, 0.1 M KCl containing 5.0 mM 3−/4− Fe(CN)6 ; scan rate, 50 mV/s.

Pt. These nanostructured Pt functionalized MWNTs are good H2 O2 sensors with high sensitivity. To fully utilize the catalytic property of Pt nanoparticles and the metal–MWNT interfaces, it is very important to disperse them efficiently on the MWNTs. Since CNTs are chemically inert and hydrophobic in nature, activating their surfaces is essential and this has motivated numerous studies to improve metal dispersions on MWNTs. Surface functionalization of CNTs by application of extremely aggressive reagents, such as HNO3 or H2 SO4 or a mixture of two is an easy technique to achieve nanostructured metal dispersed MWNTs (Kumar and Ramaprabhu, 2006). Chemical functional groups, namely -COOH and -OH derived during the vigorous acid treatment (chemical oxidation) resulting in enhanced hydrophilic nature of the MWNTs, act as anchoring sites for metal nanoparticles. From Fig. 1c, it can be seen that the nanohybrid films are compact, homogeneous and densely packed on the electrode surface after casting the GOD-Chitosan-SiO2 sol–gel on the Pt/MWNTs-modified GC electrode. The composite system can provide a favorable microenvironment for the enzyme to retain its good bioactivity. CV and EIS were used to characterize the modification of 3−/4− the electrode in 5 mM Fe(CN)6 and 0.1 M KCl solution. Fig. 2 compares the CVs response at GC, MWNTs/GC and Pt/MWNTs/GC electrodes in the above solution, respectively. After modified with MWNTs, the anodic peak and cathodic peak is increased (curve b), indicating MWNTs can improve the surface area of electrode. When Pt nanoparticles are dispersed on the surface of MWNTs/GC electrode, the peak current (curve c) increases greatly. The peak current is much higher and the separation of the anodic and cathodic peak potentials (Ep ) is smaller than these at GC and MWNTs/GC electrodes, indicating that Pt nanoparticles play an important role in increasing the electroactive surface area. Fig. 3 displays the EIS of the electrodes, obtained in 5 mM 3−/4− Fe(CN)6 and 0.1 M KCl solution. As shown in Fig. 3, at the naked GC electrode, the electron transfer resistance (Ret ) can be estimated to be 750  (curve a). After coated with MWNTs

1013

Fig. 3. EIS of (a) GC electrode, (b) MWNTs/GC electrode and (c) 3−/4− and 0.1 M KCl. Pt/MWNTs/GC electrodes in 5.0 mM Fe(CN)6

and Pt/MWNTs films, the Ret decreases dramatically, nearly to zero at curves b and c, indicating that the MWNTs and Pt/MWNTs form high electron conduction pathways between the electrode and electrolyte, and obviously improve the diffusion of ferricyanide toward the electrode surface (Liu et al., 2005). Furthermore, the Ret of Pt/MWNTs-modified GC electrode is the lowest among the three electrodes, which demonstrates that the Pt/MWNTs-modified GC has excellent electrocatalytic activity. These results were consistent with the CV tests. 3.2. Electrochemical performance of the Pt/MWNTs/GC electrode In addition to high electrochemical behavior for Fe(CN)4− 6 , the combined Pt/MWNTs-modified GC electrode exhibited the highest electrocatalytic activity toward H2 O2 . Fig. 4 displays the response to increasing levels of H2 O2 in 10 ␮M steps at electrode at potentials of 0.6 V. The response reached a steadystate signal within 5 s. Fast responses may be due to the easy

Fig. 4. Current–time recording obtained at the Pt/MWNTs/GC-modified electrode upon increasing the concentration of H2 O2 in steps of 10 ␮M at 0.6 V.

1014

Y. Zou et al. / Biosensors and Bioelectronics 23 (2008) 1010–1016

diffusion of H2 O2 in the nanocomposite film. The detection limit is 200 nM on signal-to-noise ratio of 3 which is much better than that obtained using mesoporous Pt electrode (Evans et al., 2002). Repeated use of the electrode did not affect its performance and good reproducibility was obtained. For example, 20 ␮M H2 O2 was measured continuously 10 times, and a relative standard deviation (R.S.D.) of 4.1% was obtained. Excellent reproducibility again proved the stability of the assembled film. The Pt nanoparticles used in this work, dispersed well on the surface of CNTs, may provide a large available surface and enhance the electrocatalytic activity for H2 O2 electrooxidation. The excellent electrocatalytic activity of the electrode toward H2 O2 means that the electrode can represent a new electrochemical platform that provides operational access to a large number of oxidase-based enzymes for fabricating biosensors. 3.3. Effect of Pt electrodeposition time on the response current of the biosensor Pt nanoparticles were electrodeposited on the surface of CNTs by the potentiostatic method. The effect of the amount of deposited Pt nanoparticles on the response current of the biosensor was investigated. The response current of the Pt/MWNTs/GC to the addition of 5 mM glucose increased with the increase of the Pt deposition time from 1 min to 5 min. However, when the Pt deposition time was more than 5 min, the response current decreased slightly. This may be associated with the decrease of the real surface area of the electrode resulting from the deposition of a large amount of Pt nanoparticles on the electrode surface. Therefore, the deposition time of 5 min was selected in the following investigation. 3.4. Effect of pH value on the response current of the biosensor Investigation of the effect of the pH value on the performance of the biosensor is of great importance, because the activity of the immobilized GOD is pH dependent (Luo et al., 2004). The response current of the enzyme electrode increases in the pH range 5.0–7.0, the current response decreases and the background current increased at higher pH values. The current shows the maximum value at pH 7.0. 3.5. Effect of rotating rate on the response current of the biosensor To obtain a steady-state signal and to increase the rate of mass transfer at the electrode surface, the rotating disk electrode (RDE) method was employed so that the reaction is not transported limited. The effect of rotating rate for glucose detection was also investigated. The current at 0.6 V increases with increasing rotation speed and approaches a plateau at a rotation speed of 4000 rpm. The response current does not increase and keep steady at higher speed. Therefore, the rotating rate 4000 rpm was chosen for glucose detection.

Fig. 5. Effect of applied potential on the response current in pH 7, 0.1 M PBS rotating rate, 4000 rpm.

3.6. Effect of applied potential on the response current of the biosensor The applied potential was changed from 0.4 V to 0.9 V, and the corresponding response current to 2 mM glucose was measured (shown in Fig. 5). For the fabricated enzyme electrode, the oxidation of the enzymatically formed H2 O2 started at potential of 0.4 V. The response current increased rapidly with the increase of applied potential when the potential was less than 0.6 V. This indicated that the response of the enzyme electrode was controlled by the electrochemical oxidation of H2 O2 . When the potential was higher than 0.6 V, a current plateau appeared. The appearance of such a current plateau is attributed to the rate-limiting process of enzymatical kinetics, and the potential at which the current plateau appears is dependent upon the electrode nature. Similar result was also obtained in other work (Chu et al., 2007). A potential of 0.6 V was selected as the operational potential of the enzyme electrode. 3.7. Amperometric determination of glucose at the enzyme electrode Fig. 6 illustrates a typical current–time plot for the enzyme electrode upon the successive addition of 1 mM glucose at 0.6 V. The current response of the enzyme electrode increased along with glucose concentration. The enzyme-modified electrode reached 95% of the steady-state current within 5 s, which is much faster than the reported result (50 s) in the pure silica matrix and (20 s) in the Silica sol–gel composite film. The fast response can be attributed to two aspects. First, The MWNTs provide a high surface area for Pt nanoparticles loading. The Pt–MWNTs nanocomposites with improved electrocatalytic activity can electrooxidize the byproduct (H2 O2 ) of enzyme reaction rapidly. Second, the chitosan-SiO2 gel exhibited highly porous structure and the sol–gel film was very thin because of large amount of volume shrinkage during the drying process which was conducive to fast diffusion of the substrate from bulk solution to the enzyme membrane. Therefore, the enzyme elec-

Y. Zou et al. / Biosensors and Bioelectronics 23 (2008) 1010–1016

1015

tain the activity of the enzyme. The synergistic action of Pt and MWNTs made the biosensor have excellent electrocatalytic activity. The immobilized process might give rise to the microenvironment to change enzyme and affect the intrinsic properties of enzyme which could improve the affinity to glucose. 3.8. Interference tests

Fig. 6. Typical current–time plot for the biosensor upon the successive addition of 1 mM glucose at 0.6 V in 0.1 M pH 7.0 PBS, rotating rate, 4000 rpm. Inset: calibration curve of the biosensor as a function of glucose concentrations.

trode did not pose electrochemical and diffusion obstacles to the substrate. Fig. 6 (inset) shows the calibration curve of glucose at the enzyme electrode. The enzyme electrode gave a linear response to glucose in the range from 1 ␮M to 23 mM with a correlation coefficient of 0.999. The electrode has a sensitivity of 58.9 ␮Am M−1 cm−2 and a detection limit 1 ␮M at the signalto-noise ratio of 3. The sensitivity of the resulting biosensor is much higher than the reported 29.9 ␮Am M−1 cm−2 (Hrapovic et al., 2004) and 0.98 ␮Am M−1 cm−2 (Yang et al., 2006a,b). The detection is lower and the linear range is much wider than that of biosensor based on Pt microparticles dispersed in nanofibrous polyaniline (Zhou et al., 2005). The high sensitivity and wide linear range of the enzyme electrode can be attributed to the excellent electrocatalytic activity of Pt/MWNTs nanocomposites and the biocompatible microenvironment around the enzyme. Large amount of hydrogen bonds in the sol–gel hybrid material is favorable to maintain the active configuration of the enzyme molecule. When glucose concentration is high, a plateau current was observed, showing the characteristics of Michaelis–Menten kinetics. The apparent Michaelis–Menten constant (Km ) and the maximum current density (imax ) can be obtained by an amperometric method as suggested by Shu and Wilson (1976):   1 1 1 Km + = is imax Cg imax where is is the steady-state current, Cg the concentration of glucose, Km the apparent Michaelis–Menten constant and imax is the maximum current. From the curve of the i−1 versus s Cg−1 , based on the experimental data from Fig. 6, the apparent Michaelis–Menten constant Km and the maximum current density imax were estimated to be 14.4 mM, 122.85 ␮Am M−1 , respectively. The value of Km is much lower than the reported 22 ± 2 mM (Wang et al., 1998; Sampath and Lev, 1996) and 33 mM in solution phase (Swoboda and Massey, 1965). These results show that the biosensor possesses higher biological affinity to glucose. The sol–gel film is favorable to main-

The interferences from electroactive compounds commonly present in physiological samples of glucose such as ascorbic acid and acetaminophen used to cause problems in the accurate determination of glucose. The potential interference was examined in detail. At 0.6 V, the response of the glucose biosensor was affected by the addition f 0.1 mM ascorbic acid and acetaminophen, respectively, as these species also yielded response current the electrode surface. The response signals of acetaminophen were nearly eliminated at the potential of 0.6 V, while the addition of 0.1 mM ascorbic acid can produce oxidation current. However, high sensitivity of the biosensor towards glucose made the interference of ascorbic acid negligible. The good selectivity may attribute to the Nafion polymer, is as an effectively perselective barrier which can circumvent the entry of anionic biological interferences. 3.9. Reproducibility and stability of the enzyme electrode The reproducibility of enzyme electrode construction was estimated from the response to 5 mM glucose at five enzyme electrodes prepared under the same conditions. The results revealed that the biosensor has satisfied reproducibility with a relative standard deviation of 5.2%. The operational stability of the enzyme electrode was measured at 0.6 V in 0.10 M PBS containing 5 mM glucose. There is less than 2.4% relative deviation for five times continuous determinations of the same sample. Excellent reproducibility seems to mean that little amount enzyme was leaked out from the electrode surface which indicates the stability advantage of the sol–gel binder. Though Pt/MWNTs-modified GC electrode has good electrocatalytic activity, the biosensor is unstable if there is no effective way for enzyme immobilization. For comparison, several experiments have been carried out to obtain a stable and active enzymatic layer. Nafion/GOD/Pt/MWNTs, Nafion/ChitosanGOD/Pt/MWNTs-modified GC electrode were also fabricated with the same amount of enzyme. In all cases, no steady state signal was observed from the Nafion/GOD/Pt/MWNTsmodified GC electrode probably because of a not complete entrapment of the enzyme by the Nafion. In fact, no response to the substrate was obtained when electrodes were reused, which is in agreement with the reference (Ricci et al., 2003). A Nafion/Chitosan-GOD/Pt/MWNTs film was then tested to obtain a more stable layer. The resulting membranes were also unstable because chitosan gel swells unsteadily in water. The storage stability of the biosensor was also studied. The steady-state response current of 5 mM glucose was determined every 2 days. When not in use, the biosensor was stored dry

1016

Y. Zou et al. / Biosensors and Bioelectronics 23 (2008) 1010–1016

at 4 ◦ C. The results show that the steady-state response current only decreases by 10% after 50 days (25 times) measurements, which indicates that the enzyme electrode was considerably stable. The good stability of the biosensor can be attributed to two points: on one hand, as the enzyme is physically entrapped in the chitosan-SiO2 gel, which has large quantities of hydroxyl and amido groups that are favorable to maintain the activity of the enzyme, can form strong interaction between the enzyme and the hydrogel; on the other hand, the chitosan-SiO2 provide a good microenvironment for GOD immobilization. The bottleneck effect of the silica sol–gel prevents the enzyme from leaking. 4. Conclusions In the present paper, a novel route for fabricating biosensor was developed. Pt nanoparticles were electrodeposited on MWNTs matrix in a simple and robust way. The immobilization of glucose oxidase onto Pt/MWNTs electrode surfaces was carried out by chitosan-SiO2 gel. The resulting biosensor exhibited wide linear range from 1 ␮M to 23 mM with a high sensitivity of 58.9 ␮Am M−1 cm−2 . The enzyme electrode also demonstrated excellent stability and non-interferences. The good results may be attributed to the synergistic action of Pt and MWNTs and the good biocompatibility of chitosan-SiO2 sol–gel. The hybrid nanocomposites provide a new electrochemical platform for designing a variety of bioelectrochemical devices with high sensitivity and good stability. Acknowledgments The authors wish to express their gratitude and appreciation for the financial support from the National Natural Science Foundation of China (nos. 20473091 and 50671098).

References Braun, S., Rappoport, S., Zusman, R., Avnir, D., Ottolenghi, M., 1990. Mater. Lett. 10, 1–5. Chu, X., Duan, D., Shen, G., Yu, R., 2007. Talanta 71, 2040–2047. Evans, S.A.G., Elliott, J.M., Andrews, L.M., Bartlett, P.N., Doyle, P.J., Denuault, G., 2002. Anal. Chem. 74, 1322–1326. Hrapovic, S., Liu, Y., Male, K.B., John, H., Luong, T., 2004. Anal. Chem. 76, 1083–1088. Kumar, M.K., Ramaprabhu, S., 2006. J. Phys. Chem. B 110, 11291–11298. Lim, S.H., Wei, J., Lin, J., Li, Q., You, J.K., 2005. Biosens. Bioelectron. 20, 2341–2346. Lin, C.-L., Shih, C.-L., Chau, L.-K., 2007. Anal. Chem. 79, 3757–3763. Liu, Y., Wang, M., Zhao, F., Guo, Z., Chen, H., Dong, S., 2005. J. Electroanal. Chem. 581, 1–10. Luo, X., Xu, J., Zhao, W., Chen, H., 2004. Sens. Actuators B 97, 249–255. Ricci, F., Amine, A., Palleschi, G., Moscone, D., 2003. Biosens. Bioelectron. 18, 165–174. Shi, G., Sun, Z., Liu, M., Zhang, L., Liu, Y., Qu, Y., Jin, L., 2007. Anal. Chem. 79, 3581–3588. Sun, L.X., Xu, F., Tatsuhiro, O., 1998. Talanta 47, 1165–1174. Sampath, S., Lev, O., 1996. Anal. Chem. 68, 2015–2021. Shu, F.R., Wilson, G.S., 1976. Anal. Chem. 48, 1679–1686. Swoboda, B.E., Massey, V., 1965. J. Biol. Chem. 240, 2209–2215. Tang, H., Chen, J., Yao, S., Nie, L., Deng, G., Kuang, Y., 2004. Anal. Biochem. 331, 89–97. Wang, B., Li, B., Deng, Q., Dong, S., 1998. Anal. Chem. 70, 3170–3174. Wang, B., Zhang, J., Dong, S., 2000. Biosens. Bioelectron. 15, 397–402. Wang, S.G., Zhang, Q., Wang, R., Yoon, S.F., Ahn, J., Yang, D.J., Tian, J.Z., Li, J.Q., Zhou, Q., 2003. Electrochem. Commun. 5, 800–803. Yang, M., Yang, Y., Liu, Y., Shen, G., Yu, R., 2006a. Biosens. Bioelectron. 21, 1125–1131. Yang, M., Yang, Y., Yang, H., Shen, G., Yu, R., 2006b. Biomaterials 27, 246–255. Yang, Y.M., Wang, J.W., Tan, R.X., 2004. Enzyme Microb. Technol. 34, 126–131. Yuan, J., Wang, K., Xia, X., 2005. Adv. Funct. Mater. 15, 803–809. Zhang, M., Gorski, W., 2005. J. Am. Chem. Soc. 127, 2058–2059. Zhou, H., Chen, H., Luo, S., Chen, J., Wei, W., Kuang, Y., 2005. Biosens. Bioelectron. 20, 1305–1311. Zou, Y., Sun, L., Xu, F., 2007. Biosens. Bioelectron. 22, 2669–2674.