Carbon nanotube/polysulfone composite screen-printed electrochemical enzyme biosensors{

Transcription

1 PAPER The Analyst Carbon nanotube/polysulfone composite screen-printed electrochemical enzyme biosensors{ Samuel Sánchez, a Martin Pumera, b Enric Cabruja c and Esteve Fàbregas* a Received 27th June 2006, Accepted 26th October 2006 First published as an Advance Article on the web 9th November 2006 DOI: /b609137g The fabrication, evaluation and attractive performance of multiwall carbon nanotube(mwcnt)/ polysulfone biocomposite membrane modified thick-film screen-printed electrochemical biosensors are reported. The fabricated carbon nanotube/polysulfone (CNT/PS) strips combine the attractive advantages of carbon nanotube materials, polysulfone matrix and disposable screen-printed electrodes. Such thick-film carbon nanotubes/polysulfone sensors have a well defined performance, are mechanically stable, and exhibit high electrochemical activity. Furthermore, biocompatibility of CNT/PS composite allows easy incorporation of biological functional moiety of horseradish peroxidase by phase inversion technique. The comparison of graphite with MWCNT as conductor material is described in this paper. The proposed H 2 O 2 biosensor exhibited a linear range (applied potential, 20.2 V) from 0.02 to 0.5 mm and a K M app of 0.71 mm. 1. Introduction There is a strong interest in preparing new nanomaterials composed of carbon nanotubes (CNT) and organic binders for electrochemical and materials science applications. 1 The coupling of polymers with carbon nanotubes and biorecognition elements forming a composite is of increasing importance due to its simplicity of construction and its ability to incorporate conducting materials into porous polymers in order to form electrochemical biosensors. We describe and characterize in this paper horseradish peroxidase/cnt/polysulfone thickfilm screen-printed electrodes for electrochemical sensing. Since the discovery of multiwalled carbon nanotubes (MWCNT) in 1991 by Iijima 2 and their single walled (SWCNT) counterparts two years later 3 they have attracted a huge interest because of their unique chemical, mechanical and electronic properties. 4 The remarkable conductivity of MWCNT (regarded as metallic) and their interesting electrocatalytic and electrochemical properties have led to an explosion of research activity in the field of CNT based electrochemical (bio) sensors in recent years. 5 9 The majority of the CNT-modified electrodes have been prepared by casting CNT films on the surface of glassy carbon electrodes. While this CNT electrode preparation scheme is useful for the study of electrocatalytic properties of carbon nanotubes, new fabrication schemes and more rigid materials are needed to broaden the application of CNT-based electrochemical sensors. Carbon nanotube (bio) composites with Teflon, 16 epoxy, 17,18 chitosan 19,20 or polypyrrole 21 were prepared in order to improve robustness of CNT electrodes and to facilitate immobilization of biocomponents. Recently, it has been demonstrated that CNT-based inks are highly suitable for the microfabrication of thick-film electrochemical sensors. 22 Such screen-printed CNT sensors, based on thick-film fabrication, are mechanically stable with good resistance to mechanical abrasion and they offer large scale mass production of highly reproducible low-cost electrochemical biosensors. 22,23 The thick-film technology for fabrication of CNT screen-printed electrodes combine the attractive advantages of CNT materials and screen-printed electrodes and this opens the door for a wide range of sensing applications. The knowledge of new materials, such as polysulfone (PS), allows us to expand the ability of constructing amperometric and potentiometric sensors (Scheme 1) Polysulfone showed high resistance in extreme ph conditions as well as good thermal stability. 27,28 Moreover, polysulfone is soluble in dimethylformamide, and membranes are easily prepared by conventional phase inversion technique. 29 Its porosity allows it to be used in micro, ultrafiltration and reverse osmosis processes 30 as well as in the development of composite membranes to facilitate transport. 31 It is also possible to modify the chemical nature of the polysulfone matrix in order to optimize membrane composition. Furthermore, it is easy and fast to prepare electrodes based on PS membranes a Grup de Sensors i Biosensors, Universitat Autònoma de Barcelona, Bellaterra, Barcelona, Spain. esteve.fabregas@uab.es; Fax: ; Tel: b ICYS, National Institute for Material Science, 1-1 Namiki, Tsukuba, Japan c Centre Nacional de Microelectrònica (CNM) Bellaterra, Barcelona, Spain { Electronic supplementary information (ESI) available: Fig. S1 S3. See DOI: /b609137g Scheme 1 matrix. Structural representation of polysulfone used as polymeric 142 Analyst, 2007, 132, This journal is ß The Royal Society of Chemistry 2007

2 In this paper, we describe and characterize carbon nanotube/ polysulfone composite thick-film screen-printed electrodes for amperometric sensing and biosensing. The combination of MWCNT and polysulfone results in a novel composite material, consisting of an interconnected CNT polymer network, and possessing mechanical flexibility, high toughness, and high porosity, while retaining the attractive electrochemical behaviour of CNT electrodes and biocompatibility of polysulfone. In the next sections we will describe that such CNT/polysulfone functional modules have favourable electrochemical properties which lead to sensitive detection systems based on thick-film screen-printed electrodes. The soft composite CNT/PS electrodes also serve as a reservoir of horseradish peroxidase (HRP) enzyme. This new biosensor allows an easy and fast incorporation of biological reagents into polymeric matrices for biosensing applications. 2. Experimental 2.1. Apparatus Amperometric experiments were performed with a Bioanalytical system (BAS) LC-4C amperometric controller connected to a BAS X Y recorder. Cyclic voltammograms were recorded with the AUTOLAB PGSTAT10 Electrochemical Analyzer (Eco Chemie BV, Netherlands). All the experiments were performed in a three-electrode cell. The working electrodes were made by screen printing using a Dek248 semi-automatic system. The squeegees used were soft polymer-type and the pressure applied during the printing process was set to 7 kg cm 22. A double-sweep process was programmed at a speed of 20 mm s 21. The Ag/AgCl reference electrode (Orion ) filled with 0.1 M KCl as external reference solution and a platinum auxiliary electrode were used. A stirring bar and a magnetic stirrer provided the convective transport during amperometric measurements. A Scanning Electron Microscope (SEM, Hitachi S-570, Tokyo, Japan) was used to study the morphology of the membrane surface Reagents and solutions Polysulfone (PS) was obtained from BASF (BASF Ultrasons S 3010 natur), using as solvent N,N-dimethylformamide (DMF) from Panreac (Spain). The buffer used for the electrochemical reaction was phosphate buffer solution, PBS (0.1 M KCl, 0.1 M Na 2 HPO 4, HCl to ph 7.0). Potassium ferricyanide (Sigma) and hydrogen peroxide (Merck, Hohenbrunn, Germany) solutions used as analytes were prepared daily in double-distilled water. Horseradish peroxidase (HRP, 100 units mg 21 ) was obtained from Sigma (Spain). A conductive paste was formed mixing PS/DMF suspension with MWCNT or graphite in different concentrations and stirring it for 10 min and sonification. This PS suspension modified with either MWCNT or graphite is deposited onto graphite electrodes to form a soft composite and used as sensor or biosensor. Two kinds of multiwall carbon nanotubes (y95% pure) with different characteristics were obtained from Aldrich (Stenheim, Germany), for their characteristics see Table 1. Table 1 Name Further purification was accomplished by stirring the MWCNT in 2 mm nitric acid (Panreac, Spain) for 24 h and drying at 80 uc in a furnace. 17 Graphite powder (particle size less than 50 mm) was obtained from BDH, UK. The Acheson carbon ink (Electrodag 400 B), conductive silver ink (Electrodag 6037 SS) and insulating ink (Minico M 7000) were obtained from Acheson colloids Co. (USA) Methods Characteristics of MWCNT used in this work Outside diameter Inside diameter Length Purity MWCNT nm 5 15 nm mm y95% MWCNT nm 5 10 nm mm y95% Preparation of carbon/polysulfone screen-printed electrode. The amperometric sensors used in the present investigation consisted of a single working screen-printed electrode deposited onto polycarbonate (PC) substrate. Silver ink acting as a conductive layer was printed and cured in a furnace at 60 uc overnight. Carbon paste ink was printed and cured at the same temperature overnight. A nonconductive isolating ink was applied and cured at 60 uc overnight. The reaction area of the working electrode was 20 mm 2. Fig. 1 shows a schematic of the structure of MWCNT/PS screen-printed electrode. The carbon/polysulfone composites were prepared as follows. The MWCNTs or graphite suspension was mixed with the 7.5% wt 25 PS DMF solution for 10 min under continuous stirring. The mixed ratios (carbon/ps-dmf suspension) were 6.5, 9.6, 12.5, 15.0, 17.6 wt%. Serigraphy is applied to print the composite onto the reaction region of 12 working electrodes. The electrodes were then immersed into bidistilled water for the phase inversion during 5 min and rinsed for 1 min and then the electrodes were dried at room temperature Preparation of HRP/MWCNT/polysulfone by phase inversion method. A similar procedure as described in section Fig. 1 (A) The enzyme/mwcnt/polysulfone screen-printed thickfilm electrochemical detector, top view; (B) Cross section of the detection area of enzyme/mwcnt/polysulfone screen-printed detector; (C) schematic drawing showing structure of HRP/MWCNT/ PS composite; (a) polycarbonate substrate, (b) insulator layer, (c) HRP/MWCNT/polysulfone conducting composite, (d) silver contact for the working electrode, (e) carbon ink contact layer. This journal is ß The Royal Society of Chemistry 2007 Analyst, 2007, 132,

3 was carried out to print the PS-membranes onto bare electrodes, but in this case a solution of HRP is used in the phase inversion process. Concentrations of HRP of 5.34 Units ml 21 were tested. The carbon/ps membranes were dipped for 5 min into the solution and the biosensors were rinsed for 1 min with bidistilled water and dried at room temperature Electrochemical measurements Voltammetric measurements. Cyclic voltammetric experiments were performed to study the conducting composite at a scan rate of 50 mv s 21 adding 0.1 mm ferricyanide to the PBS buffer solution. Different MWCNT/PS and graphite/ PS loading ratios were analyzed in order to determine the highest cathodic peak Amperometric measurements. The current response of the MWCNT/PS and the graphite/ps sensors were recorded by adding aliquots of ferricyanide to the bulk solution. Hydrodynamic voltammograms were carried out for 4.6 mm hydrogen peroxide in PBS buffer solution using the graphite/ PS, MWCNT50/PS and MWCNT200/PS screen-printed electrodes. The carbon/polysulfone composition ratio used for CNT was 6.5 wt% and for graphite 17.6 wt%, if not stated otherwise. The stirring rate was y500 rpm. Hydrogen peroxide was added to the bulk solution for the characterization of HRP-biosensors using hydroquinone as mediator. Measurements were carried out in PBS buffer supporting electrolyte medium under stirring conditions. A potential of 2200 mv vs. SCE reference electrode was applied to the working electrode. Reproducibility of the construction and repeatability of the sensor response were measured in order to study the sensors properly. All measurements were performed at room temperature. 3. Results and discussion The attractive behavior of new MWCNT/PS composite screenprinted electrodes was demonstrated in connection with the detection of hydrogen peroxide and potassium ferricyanide, owing to the importance of these compounds in a wide range of sensing and biosensing applications. SEM microscopy was employed to gain insight into the nature and structure of the new polysulfone composites. Fig. 2 compares the SEM images of MWCNT200/PS and MWCNT50/PS with graphite/ps composites. Electrodes show very different morphologies since graphite/ps (A) shows a rigid composition where graphite granules are compactly distributed on the surface yielding a less porous and denser surface. In contrast, MWCNT50/PS (B) and MWCNT200/PS (C) composite electrodes show a porous and spongy morphology. This difference in porosity leads to a higher surface area for MWCNT/PS composite and the consequent easy access to the interior of the membrane leading to a higher electrolyte/ mass transport through it. The study of two MWCNT with different length and diameter, and graphite, was carried out in order to choose which one shows the higher sensitivity and usability for the production of screen-printed polysulfone composite electrodes. Fig. 2 SEM images of the surfaces of the (A) graphite/polysulfone, (B) MWCNT50/polysulfone and (C) MWCNT200/polysulfone modified screen-printed thick-film electrodes. Accelerating voltage, 15 kv; carbon/polysulfone composition ratio, 17.6 wt% (A) and 6.5 wt% (B, C). For this reason the composites are studied by varying the MWCNT or graphite loading into DMF PS suspension. Fig. 3 compares peak height obtained from cyclic voltammograms for ferricyanide, recorded at 50 mv s 21. For MWCNT50/PS composite (b) a decrease of the signal with loading into the suspension is observed. This can be due to the lack of homogeneity of this suspension with the increase of CNT loading. For MWCNT200/PS composite (a) an increase of the peak height with the loading is observed but the variations of both reproducibility of the membrane and the instability of the signal when MWCNT200 loading was higher than 9.6 wt% made us choose the 6.5 wt% loading. A content of 6.5 wt% of MWCNT50 and MWCNT200 is found to be enough to obtain the optimum response. In contrast, for graphite/ps composite (c) electrodes there is a significant increase of the peak high with loading up to 17.6 wt% (note that the MWCNT/PS electrode signal is in any case higher than graphite/ps electrode). A higher graphite loading in the composite makes the preparation of composite difficult. Thus, 6.5 wt% of MWCNT and 17.6 wt% of graphite in polysulfone matrix are taken as optimum loadings for the preparation of electrodes. Fig. 3 Influence of the carbon loading upon reduction peak current of 0.1 mm potassium ferricyanide at a potential of 20.2 V in cyclic voltammetry. (a) MWCNT200/polysulfone, (b) MWCNT50/polysulfone and (c) graphite/polysulfone. Scan rate, 50 mv s 21. Start potential, 1.2 V. Registration potential, 20.2 V. Supporting electrolyte, PBS buffer (0.1 M, ph 7.0). 144 Analyst, 2007, 132, This journal is ß The Royal Society of Chemistry 2007

4 Fig. 4 Hydrodynamic voltammograms for 4.6 mm hydrogen peroxide using the (a) graphite/polysulfone, (b) MWCNT50/polysulfone and (c) MWCNT200/polysulfone screen-printed electrodes. Supporting electrolyte, PBS buffer (0.1 M, ph 7.0); carbon/polysulfone composition ratio, 17.6 wt% (a) and 6.5 wt% (b, c); stirring rate, y500 rpm. From experiments related to Fig. 3 it is clear that an important increase in the sensitivity during electrochemical measurements can be achieved by adding MWCNT instead of graphite into carbon/polysulfone composite. Fig. 4 shows the comparison of hydrodynamic voltammograms (HDV) for 4.6 mm hydrogen peroxide at the graphite/ PS (a), MWCNT50/PS (b) and MWCNT200/PS (c) composite electrodes. No redox activity is observed for graphite/ps electrode for potentials lower than +0.5 V. In contrast, the MWCNT200/PS and MWCNT50/PS composites showed an amperometric response starting at +0.3 V and +0.4 V, respectively. For potentials higher than this value, a rapid increase in current is observed for both MWCNT/PS electrodes. In addition to the lowering of the detection potential, the carbon nanotube based polysulfone composites exhibit larger current signal. Thus, the coupling of MWCNT with PS polymer does not disturb the MWCNT electrocatalytic properties and the comparison with graphite/ps composite is used to demonstrate some advantages of the new CNT/PS sensors. Due to this electrocatalytic effect, H 2 O 2 can be detected at lower potentials with MWCNT/PS soft composite. The different carbon materials used for construction of carbon/ps screen-printed electrodes has a profound effect upon their electrochemical behavior. Fig. 5 compares calibration plots for potassium ferricyanide obtained at polysulfone electrodes with different carbon materials. All electrode compositions yield highly linear calibration plots over the entire concentration range. The sensitivity is the lowest for graphite/ps soft composite (a), despite high carbon loading in this composite (17.6% wt.). The sensitivity increases for MWCNT50/PS composite (b), while maximum sensitivity is reached for MWCNT200/PS composite (c) (carbon loading for both MWCNT/PS composites is 6.5 wt%). These results show that the MWCNTs maintain their conducting properties although they are immersed in a hydrophobic PS matrix. Each point of the calibration curve represents the average amperometric response of six different electrodes for the addition of potassium ferricyanide. It is clearly visible that the smallest RSD and the highest reproducibility of signal between Fig. 5 Calibration plots for potassium ferricyanide using polysulfone carbon screen-printed electrodes of different compositions. (a) Graphite/polysulfone, (b) MWCNT50/polysulfone, and (c) MWCNT200/polysulfone composite. Operating potential, 20.2 V; Supporting electrolyte, PBS buffer (0.1 M, ph 7.0); Error bars represent standard deviation for measurements on different electrodes (n = 6). Carbon/polysulfone composition ratio, 17.6 wt% (a) and 6.5 wt% (b,c). different electrodes were obtained with MWCNT200/PS composite. These results, in addition to others mentioned above, led us to use MWCNT200 as the best of three conducting materials into the polymeric matrix of PS composite. The deviation of amperometric response of MWCNT200/PS is attributed to the construction of the screen-printed electrode since the bare electrodes (without carbon/ps layer) showed the same deviation as MWCNT200/PS composite. For this reason, we believe that the fabrication and deposition of the soft MWCNT200/PS composite does not imply a lack of reproducibility, giving an easy, fast and cheap way of construction for soft polysulfone composite as electrochemical sensors. Long-term stability of all carbon/ps screen-printed electrodes was measured over 2 months for all carbon/ps configurations, as shown in Fig. 6. The MWCNT200 composite (c) electrodes showed a stable current response for two months with a slight signal decrease for potassium ferricyanide. The MWCNT50/PS composite (b) electrodes showed lower but also stable response. In contrast, graphite/ps composite (a) electrode shows a decrease of signal response after 40 days. This can be attributed to the enhanced mechanical properties of MWCNT vs. graphite into the polysulfone polymeric matrix. The electrode was also used for the preparation of enzyme biosensors by the incorporation of HRP into the composite matrix by phase inversion method. 25 It is well known that HRP catalyzes the reduction of hydrogen peroxide to water. Fig. 7 shows the amperometric response at 2200 mv for HRP/ MWCNT200/PS biocomposite towards the addition of H 2 O 2 (each point represents the average signal from three different biosensors). A linear relationship is shown for a concentration range from 0.02 to 0.5 mm of hydrogen peroxide, showing a detection limit of 25 mm. The apparent Michaelis-Menten constant K app M was calculated to be 0.71 mm from the Lineaweaver Burk plot (inset Fig. 7). This K app M indicates that the enzyme immobilized in the MWCNT200/PS biocomposite keeps its activity with a very low diffusion barrier. This This journal is ß The Royal Society of Chemistry 2007 Analyst, 2007, 132,

5 We demonstrated that the signal response is increased in comparison with graphite/polysulfone and electrocatalytic properties of MWCNT are not diminished by incorporating them in PS matrix. An easy-to-prepare and cheap MWCNT/ PS composite also allows the incorporation of biological functional moieties by phase inversion technique, which was demonstrated in this work by incorporation of HRP enzyme. HRP enzyme is very important as a label in electrochemical biosensing (immunoassay, DNA assay). Novel MWCNT/PS thick-film screen-printed disposable sensors and biosensors offer mass-production low-cost fabrication scheme and offer great promise for point-of-care disposable biosensing. Our preliminary results towards development of MWCNT/PS immunosensor show promising disposable electrochemical devices. Fig. 6 Long-term stability of the response to 0.1 mm potassium ferricyanide using (a) graphite/polysulfone, (b) MWCNT50/polysulfone and (c) MWCNT200/polysulfone screen-printed electrodes. Operating potential, 20.2 V; Supporting electrolyte, PBS buffer (0.1 M, ph 7.0); carbon loading, 17.6 wt% (a) and 6.5 wt% (b,c). K app M value was smaller than others for H 2 O 2 biosensors based on sol gel (4.6 mm) 32 and other composites (2.0 mm) 33 and based on siloxane homopolymer (2.5 mm). 34 This is a great advantage over other composites since the MWCNT200/PS membrane maintains the conducting properties of MWCNT and allows the easy and fast incorporation of the enzyme with a very low K app M and high sensitivity (0.12 ma mm 21 ). 4. Conclusions The experiments described above indicate that the carbon nanotube/polysulfone composite can be used for fabrication of screen-printed electrodes with attractive sensing performance. Fig. 7 Current dependance upon hydrogen peroxide concentration for HRP/MWCNT200/polysulfone biocomposite screen-printed electrode (6.5 wt% loading). Conditions: 5.3 U ml 21 HRP in the biocomposite, 1.8 mm hydroquinone used as mediator; buffer PBS ph 7.0. Error bars represent a standard deviation of 3 different electrodes. Acknowledgements S.S. and E.F. would like to thank the Spanish Ministry of Education and Science (MAT ) for its financial support. M.P. is grateful to the Japanese Ministry for Education, Culture, Sports, Science and Technology for funding through the ICYS program. References 1 D. Tasis, N. Tagmatarchis, A. Bianco and M. Prato, Chem. Rev., 2006, 106, S. Iijima, Nature, 1991, 354, S. Iijima and T. Ichihashi, Nature, 1993, 363, P. M. Ajayan, Chem. Rev., 1999, 99, J. Wang, Electroanalysis, 2005, 17, 7. 6 J. J. Gooding, Electrochim. Acta, 2005, 50, A. Merkoci, M. Pumera, X. Llopis, B. Perez, M. del Valle and S. Alegret, Trends Anal. Chem., 2005, 24, C. E. Banks and R. G. Compton, Analyst, 2006, 131, G. G. Wildgoose, C. E. Banks, H. C. Leventis and R. G. Compton, Microchim. Acta, 2006, 152, M. Musameh, J. Wang, A. Merkoci and Y. Lin, Electrochem. Commun., 2002, 4, R. R. Moore, C. E. Banks and R. G. Compton, Anal. Chem., 2004, 76, N. S. Lawrence, R. P. Deo and J. Wang, Electroanalysis, 2005, 17, C. E. Banks, R. R. Moore, T. J. Davies and R. G. Compton, Chem. Commun., 2004, 16, N. S. Lawrence, R. P. Deo and J. Wang, Electrochem. Commun., 2004, 6, M. Pumera, X. Llopis, A. Merkoci and S. Alegret, Microchim. Acta, 2006, 152, J. Wang and M. Musameh, Anal. Chem., 2003, 75, M. Pumera, A. Merkoci and S. Alegret, Sens. Actuators, B, 2006, 113, B. Perez, M. Pumera, A. Merkoci and S. Alegret, J. Nanosci. Nanotechnol., 2005, 5, M. Zhang, A. Smith and W. Gorski, Anal. Chem., 2004, 76, L. Jiang, R. Wang, X. Li, L. Jiang and G. Lu, Electrochem. Commun., 2005, 7, J. Wang and M. Musameh, Anal. Chim. Acta, 2005, 539, J. Wang and M. Musameh, Analyst, 2004, 129, J. Wang, B. Tian and K. R. Rogers, Anal. Chem., 1998, 70, A. González-Bellavista, J. Macanás, M. Muñoz and E. Fàbregas, Sens. Actuators, B, 2006, 115, B. Prieto-Simon and E. Fàbregas, Biosens. Bioelectron., 2006, 22, S. Sánchez and E. Fàbregas, Biosens. Bioelectron., 2006, DOI: /j.bios , in press. 27 W. J. Wrasildo, Asymmetric membranes, US Patent 4,629,563, December 1986, Brunswick. 146 Analyst, 2007, 132, This journal is ß The Royal Society of Chemistry 2007

6 28 M. Kraus, M. Heisler, I. Katsnelson and D. Velazques, Filtration membranes and method of making the same, US Patent 4,900,449, February 1990, Gelman. 29 Membrane Technology in the Chemical Industry, ed. S. P. Nunes and K. V. Peinemann, Weinheim, Germany, 2001, ch. 4, pp Membrane Handbook, ed. W. S. W. Ho and K. Sirkar, Van Nostrand Reinhold, New York, J. Macanás and M. Muñoz, Anal. Chim. Acta, 2005, 534, B. Wang, J. Zhang, G. Cheng and S. Dong, Anal. Chim. Acta, 2000, 407, V. S. Tripathi, V. B. Kandimalla and H. Ju, Biosens. Bioelectron., 2005, 21, M. P. G. Armada, J. Losada, I. Cuadrado, B. Alonso, B. González, C. M. Casado and J. B. Zhang, Sens. Actuators, B, 2004, 101, 143. This journal is ß The Royal Society of Chemistry 2007 Analyst, 2007, 132,