- Email: [email protected]
Electron transfer mediator micro-biosensor fabrication by organic plasma process
Biosensors and Bioelectronics 21 (2005) 957–964
Electron transfer mediator micro-biosensor fabrication by organic plasma process Atsunori Hiratsuka a , Ken-ichi Kojima b , Hitoshi Muguruma c,∗ , Kyong-Hoon Lee d , Hiroaki Suzuki b , Isao Karube a,e a
c
Laboratory of Advanced Bioelectronics, National Institute of Advanced Industrial Science and Technology, 1-1-1 Higashi, Tsukuba-shi, Ibaraki 305-8562, Japan b Institute of Materials Science, Tsukuba University, 1-1-1 Tennodai, Tsukuba-shi, Ibaraki 305-8573, Japan Department of Electronic Engineering, Shibaura Institute of Technology, 3-9-14 Shibaura, Minato-ku, Tokyo 108-8548, Japan d Department of Mechanical Engineering, Northwestern University, 2145 Sheridan Road, Evanston, Illinois 60208, USA e Faculty of Bionics, Tokyo University of Technology, Katakura 1404-1, Hachioji-shi, Tokyo 192-0982, Japan Received 27 January 2005; received in revised form 25 February 2005; accepted 1 March 2005 Available online 5 April 2005
Abstract We propose a new strategy for constructing a mediator-type biosensor as a Bio-MicroElectroMechanical Systems (BioMEMS) application. A vinylferrocene plasma-polymerized film (PPF) was deposited directly onto the surface of an electrode under dry conditions. The resulting redox film was extremely thin, adhered well onto a substrate (electrode), and had a highly crosslinked network structure. This technique, capable of polymeric deposition of any kind of monomer, can also serve the purpose of anti-fouling coating, or layer-to-layer interface creation. With a subsequent plasma process, additional polymeric layer of hydrophilic acetonitrile was superimposed onto the existing vinylferrocene-PPF surface to offer crucial features such that the wettability could be adjusted for a better electron transfer, and amino functional groups could be attached to immobilize a large amount of enzyme. Based upon this scheme, the device fabrication could be designed in a manner that the whole procedure was made up of dry wafer-handling processes, which is compatible with mass production. A prototype device was fabricated to have an array of needle-shaped amperometric micro-biosensors. The resultant thin polymer layer carried a large number of the mediator molecules, accomplishing a lower overpotential (+410 mV) and a rapid response time (<5 s). Stressing the advantages of the plasma polymerization process together with some additional features accomplished in our device fabrication, we would discuss new possibilities in the field of BioMEMS. © 2005 Elsevier B.V. All rights reserved. Keywords: Plasma-polymerized film; Electron transfer mediator; Ferrocene; Glucose biosensor; Enzyme
1. Introduction Amperometric enzyme biosensors for biochemical and medical uses have been widely and seriously developed for the last two decades. This type of biosensor, which is a combination between a biocatalyst and an electronic device, converts electron transfers involved in an enzymatic action into an electrochemical response, which provides quantitative information on a target analyte. For constructing a glucose ∗
Corresponding author. Tel.: +81 3 5427 7813; fax: +81 3 5476 3165. E-mail address: [email protected] (H. Muguruma).
0956-5663/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2005.03.002
biosensor, which has been one of the most intensively and successfully commercialized, the electron transfer mediator is a widely accepted technique to enhance the efficiency of the sensor signal acquisition. The mediator inter-relays the electron transfer between the catalytic center of the enzyme and the electrode surface. Among so many kinds of mediators, ferrocene has been undisputedly regarded as one of the most representative, since the first introduction of its concept for the mediator-based enzymatic biosensor (Cass et al., 1984). Usually, the enzyme and mediators are embedded or coimmobilized in a polymer cast on an electrode surface. Due
958
A. Hiratsuka et al. / Biosensors and Bioelectronics 21 (2005) 957–964
to their size, the enzyme molecules are well fixed within the polymer, but a possible leak of the much smaller mediator molecules causes problems such as sensor deactivation and contamination of sample. Moreover, when it comes to the sensor design, the wettability balance needs to be tuned between the hydrophobic ferrocene and the hydrophilic enzyme. The final design considering these aspects significantly determines the sensor characteristics. As such, many strategies for the electron transfer system have been proposed (Padeste et al., 2004; Kurita et al., 2004; Alonso et al., 2004; Lawrence et al., 2004). The field of micro electromechanical systems (MEMS) or micro total analysis systems (TAS) has played a significant role in the development of fast chemical analysis, which is usually related with running small-volume samples on a microsensor array or a microfluidic device (Auroux et al., 2002). The microscale approach is especially important to bioanalytical systems (so-called “BioMEMS”) due to the inherent nature of bioreactions in terms of the sample availability and the slow reaction speed. In case of enzyme biosensor-integrated MEMS systems, one of the key fabrication issues may be how to immobilize biosensing molecules contained in a liquid onto a given substrate, in line with dry processes. However, most of immobilization methods available are unsuitable for a manufacture-oriented design of biosensor and/or BioMEMS because “wet processes” are hardly unavoidable. Furthermore, they are conventionally spin- or dip-coated, which makes it difficult to achieve thin, homogeneous, reproducible, and strongly adhesive film properties. Taking all these into consideration, we propose a new strategy for the fabrication of an electrochemical enzymatic biosensor with an electron mediator for BioMEMS application, whose immobilization step relies upon organic plasma polymerization (Yasuda, 1985). Plasma-polymerized films (PPFs) of vinylferrocene were reported by the Murray’s group (Nowak et al., 1980; Daum and Murray, 1979), but these were mainly about the electrochemical characterizations instead of sensor applications. The amperometric enzyme biosensors based on plasma polymerizaion were early reported by some researchers (Kampfrath and Hintsche, 1989; Yoshimura and Hozumi, 1996; Mutlu et al., 1994, 1998). Likewise, we have been exploiting organic plasma processes for the construction of biosensors and/or BioMEMS devices (Muguruma and Karube, 1999; Muguruma et al., 2000; Hiratsuka et al., 2001, 2004). In this article, the possibility of such concept is expanded to a biosensor involving the “electron transfer mediator”, as a precursor to mass production-compatible bioMEMS devices.
2. Materials and methods 2.1. Materials Acetonitrile, potassium dihydrogenphosphate, disodium hydrogenphosphate, d-glucose, potassium chloride, glutar-
Fig. 1. Equipment and geometry used for plasma polymerization.
aldehyde, hydrogen chloride and hydrogen peroxide were purchased from Wako (Osaka, Japan). The enzyme glucose oxidase (GOD) from Aspergillus niger (EC 1.1.3.4, type VIIS, 181600 units/g) was purchased from Sigma (St. Louis, MO, USA). Hexamethyldisiloxane (HMDS), tetraethylammonium perchlorate (Et4 NClO4 ), and vinylferrocene purchased from Aldrich (Milwaukee, WI, USA). All reagents used in this work were prepared without further purifications. All solutions were made with deionized water (18 M cm resistivity). 2.2. Plasma polymerization The apparatus used for the plasma polymerization (Model BP-1, Samco International Laboratories, Inc., Kyoto, Japan) is shown in Fig. 1. Two electrodes were placed horizontally 15 cm above the sample stage (external electrode reactors). An RF generator (Model RFG-300, Samco), coupled to an automatching network to minimize the reflective power, was employed. The working frequency of the power supply was 13.56 MHz. The vinylferrocene monomer was directly placed on the sample stage, which was heated to ca. 60 ◦ C in order to provide a high monomer pressure. The acetonitrile monomer was placed in a bottle reservoir, which was connected to an inlet line through the mass flow controller. With plasma-free exposure of the electrode to the vinylferrocene vapor, we could not find any electrochemical- or other experimental-evidences that ferrocene was immobilized. 2.3. Device fabrication The device, which is an electrochemical micro-biosensor with the three electrodes configuration, was fabricated on the basis of semiconductor layer-by-layer processes as shown in Fig. 2. It was formed on a 150-m thick glass substrate, whose planar dimension was approximately 50 mm × 50 mm. All metal layers were sputter-deposited and patterned by a lift-off process. Glass slides used to make thin
A. Hiratsuka et al. / Biosensors and Bioelectronics 21 (2005) 957–964
Fig. 2. Decomposed structure of the needle-type biosensor: (1) Pt for counter electrode (200 nm); (2) Cr for adhesion layer (40 nm); (3) glass substrate; (4) Cr for adhesion layer (40 nm); (5) Pt for working electrode (200 nm); (6) Ag/AgCl for reference electrode (300 nm); (7) polyimide (7 m); (8) vinylferrocene plasma-polymerized film (150 nm); (9) acetonitrile plasmapolymerized film, (20 nm) introducing amino groups onto the surface; (10) glutaraldehyde reacted with the amino groups on 9 layer; (11) immobilized glucose oxidase via glutaraldehyde. Inset is the cross-section of the working electrode.
film electrodes were cleaned in a 50% nitric acid for an hour and then rinsed with water and acetone. The working, reference, and counter electrodes were formed with 200-nm thick platinum (Pt) backbone patterns. The Pt thin films were sputtered under normal condition, with an apparatus manufactured by Shibaura Engineering Works Co., Ltd. (Model CFS-4ES-231, Tokyo, Japan). The Pt film thickness (200 nm) was determined from a surface profiler (Dektak ST, Veeco Instruments Inc., Tokyo, Japan). A 40-nm thick chromium (Cr) intermediate layer was used to promote the adhesion of the Pt layer. A 300-nm thick silver layer was formed only on the reference electrode area. To delineate the active area for each electrode a 7m thick polyimide layer was formed. In curing the polyimide, the devices were subjected to bakings for 15 min at 150 ◦ C, 15 min at 200 ◦ C, and 30 min at 300 ◦ C. The dimension of the openings for the working electrode was 0.5 mm × 1.0 mm. A novel thin-film Ag/AgCl anode structure was used in this device (Suzuki et al., 1998). As reported in the previous study, the lifetime of a usual thin-film made of Ag/AgCl is very restricted and its element starts to deteriorate once the film is exposed to an electrolyte solution. In order to solve
959
the problem, the entire silver layer was protected with the polyimide film and the AgCl layer was grown from one edge of the silver pattern. The protecting layer was effective in suppressing the dissolution of AgCl. A vinylferrocene PPF was deposited onto a platinum electrode. The deposition parameters were as follows: power; 80W, pressure; 1.0 Pa, plasma exposure time; 2 min (150 nm of film thickness). The vinylferrocene monomer was directly placed on the electrode (see Fig. 1) under glow discharge according to the Murray’s method (Nowak et al., 1980). Subsequently, an acetonitrile PPF was superimposed onto the vinylferrocene layer above the platinum electrode. The deposition parameters were as follows: power; 80-W, flow rate; 15 mL min−1 , pressure; 2.1 Pa, plasma exposure time; 1 min (20 nm of film thickness). The surface of the acetonitrile PPF had a multitude of amino groups to which enzyme molecules could be immobilized. At the last stage, the immobilization of the enzyme was achieved by applying 2.5% aqueous glutaraldehyde (GA) solution to the surface of the plasma-polymerized film, washing with water, and then dropping 10 mg/mL GOD (Sigma) solution in a phosphate buffer (20 mM, pH 7) onto the film. After 10 minutes, the device was washed with water and stored at 4 ◦ C until use. Twentyeight sensors were batch-fabricated on the glass substrate shown in Fig. 3. Each and every sensor in a needle shape (0.5 mm × 1.0 mm) was obtained by cutting the substrate with a dicing saw (Model DAD321, Disco, Tokyo, Japan). 2.4. Measurement Electrochemical studies were performed with a BAS50W potentiostat (Bioanalytical Systems, West Lafayette, IN, USA). The three-electrode configuration equipped with the fabricated devices was used. Electrochemical measurements were carried out in a 5-mL vessel at laboratory ambient temperature (20 ± 1 ◦ C). The supporting electrolyte was a phosphate buffer (20 mM, pH 7.4).
3. Results and discussion 3.1. Characterization of vinylferrocene plasma-polymerized film on sputtered Pt thin film A plasma-polymerized vinylferrocene film on an electrode was studied by the Murray’s group (Nowak et al., 1980; Daum and Murray, 1979), but there were still good reasons enough for us to revisit this topic for our own purpose: first, the structure and properties of any PPF strongly depend on the reactor geometry, the discharge power, and the chamber pressure of a given plasma generator. Second, we used sputtered thin film electrodes, on which a plasma-polymerized film would present different characteristics from that of a “bulk” electrode (Nowak et al., 1980). A scanning electron microscopic (SEM) image and a cyclic voltammogram of a Pt thin film electrode covered with a vinylferrocene PPF are shown in
960
A. Hiratsuka et al. / Biosensors and Bioelectronics 21 (2005) 957–964
Fig. 3. Picture of arrayed needle biosensors. Twenty-eight of biosensors array are cut into each device.
Fig. 4. The distinct cathodic and anodic peaks due to the ferrocene/ferrocenium couple were observed from the film prepared by the 80-W plasma power. The difference between the cathodic- and the anodic-peak potentials (Ep ) was 80 mV. The formal potential (E0 ) versus Ag/AgCl was +450 mV. The ferrocenes in the plasma-polymerized film seemed substitutionally similar to those in a polyvinylferrocene film (Nowak et al., 1980). The surface coverage of the ferrocene redox sites was estimated to 4.5 × 10−8 mol/cm2 . The redox peak of a vinylferrocene PPF deposited with the 20-W discharge power was smaller than that with the 80-W. This can be explained by our observation (Fig. 4a) that the 20-W discharge power was too low to form a homogeneous structure with an ample number of the redox ferrocenes.
The peak of the redox vinylferrocene PPF formed with the 120-W power was smaller and more widened with a positive potential shift than that of the 80-W. This may also be explained by another observation of ours (Fig. 4c) that the high discharge power of 120-W gave a homogeneous film structure but destructively transformed a greater part of the redox ferrocenes into electrochemically inactive Fe(III) species (Nowak et al., 1980). Consequently, we used this value (80-W) as the optimal plasma discharge power in this work. The dependence of cyclic voltammograms on their scan rates was investigated using one of the fabricated electrodes (Fig. 5). With the increasing scan rate, the cathodic peak potential shifted toward in the more negative direction, and the anodic peak potential in the more positive direction,
A. Hiratsuka et al. / Biosensors and Bioelectronics 21 (2005) 957–964
961
Fig. 4. Cyclic voltammograms and surface image (scanning electron spectroscopy) of electrode modified by vinylferrocene plasma-polymerized film as a function of the power of polymerization: (a) 20-W; (b) 80-W; (c) 120-W. Solvent was 0.1 M Et4 NClO4 /CH3 CN. Scan rate was 10 mV s−1 .
as generally observed, probably due to the hysteresis. This observation was more obvious than in the case of a bare electrode (data not shown), which may be related with the surface diffusion. On the other hand, the magnitude of a peak current was linearly proportional to the square root of the relevant scan rate. This can be explained by the Randles– Sevcik equation (Bard and Faulkner, 1980), applicable to the case of a semi-infinite volume of a diffusing reactant in contact with the electrode. 3.2. Device fabrication Generally, the lifetime of a conventional thin-film Ag/AgCl electrode is shortened and the electrode deteriorates in a highly concentrated chloride solution. In order to fix this problem, we employed a new type of Ag/AgCl electrode with high durability (Suzuki et al., 1998), whose design was intended for integration into a miniaturized device. The
Ag/AgCl layer was covered with a polyimide protective layer and 20 holes (0.1 mm × 0.1 mm) were opened on the layer by photolithographic patterning. In our work, we were highly conscious of developing a fabrication scheme compatible with mass-production. Most of all, it is noteworthy that our process could leave out the use of a solvent from the corresponding conventional steps. For example, the matrix for immobilizing the enzyme and the mediator could be created only after several minutes’ exposure of a substrate to plasma generated under vacuum. This enabled us to continue to run dry wafer processes up to the last stage, at which a wet chemistry could be introduced for the enzyme immobilization without meaningful interruption of the overall procedure. Here, one might imagine even advanced protocols more appropriate for the mass production, in which the sputtering of thin films and their surface modifications with plasma polymerization are carried out sequentially at a batch.
962
A. Hiratsuka et al. / Biosensors and Bioelectronics 21 (2005) 957–964
the discriminative peak was observed at +450 mV, indicating that the hydrogen peroxide was oxidatively decomposed by the ferrocenium ions (Fc+ ). The mechanism can be expressed as follows. First, glucose is oxidized by the immobilized enzyme. -d-glucose + O2 → (GOD) ␦-gluconolactone + H2 O2 (1) The hydrogen peroxide produced diffuses through the acetonitrile PPF layer and reaches the vinlyferrocene PPF layer. Subsequenly, the hydrogen peroxide is oxidized by ferricenium ions (Zhou et al., 1997). 2Fc+ + H2 O2 → 2Fc + 2H+ + O2
(2)
Meanwhile, ferrocenes (Fc) reductively created from ferriceniums are oxidized back electrochemically, resulting in electron transfers to the electrode. Fc → Fc+ + e−
Fig. 5. Cyclic voltammogram of plasma-polymerized vinylferrocene deposited in platinum electrode thin film in 0.1 M Et4 NClO4 /CH3 CN. (a–e) 5, 10, 20, 50, 100, and 200 mV s−1 , respectively. Insets: dependence of peak current (Ip ) on square root of potential sweep rate.
3.3. Sensor characteristics Fig. 6 shows a cyclic voltammogram, which was obtained in the absence and presence of glucose. The catalytic current resulted from the added glucose, which diffuses into the film and reacts oxygen in the presence of glucose oxidase. When hydrogen peroxide produced by the enzymatic action was directly oxidized with the surface of the electrode, large and broad current increments were observed in a potential range from +400 to +800 mV (Hiratsuka et al., 2001). In this work,
Fig. 6. Cyclic voltammograms obtained using the device in 20 mM phosphate buffer (pH 7.4): (a) without glucose and (b) with 10 mM of glucose. The scan rate was 10 mV s−1 .
(3)
To our knowledge, this work is probably the first report on a vinylferrocene-mediated biocatalytic electric current observed in a generic aqueous medium. In general, observation of such a current should be accompanied by the use of a specific supporting electrolyte such as LiClO4 and Et4 NClO4 (Daum and Murray, 1979). We believe that this can be ascribed to the hydrophilic acetonitrile PPF layer, without which the peak current coming from the enzymatic reaction as shown in Fig. 6 was “not” observed. Only with the hydrophobic surface of the vinylferrocene PPF, first, the mass transport would be seriously retarded across the heterogeneous interface between an aqueous medium and the ferrocene membrane, and, second, it would be difficult to immobilize the enzyme thereon. The previous solution to this problem was introduction of a hydrophilic polymer matrix (Saito and Watanabe, 1998, 1999). Compared with these, our method is capable of controlling the film properties in a simpler, more reproducible and robust manner, which provides a more manufacture-oriented protocol. The structure of the acetonitrile PPF consists of highlybranched- and incompletely-crosslinked aliphatic hydrocarbon backbone chains. The film contains a lot of nitrogen atoms in various forms of functional groups, among which 32% were primary amines (Hiratsuka et al., 2000). This provides not only a good matrix for the immobilization of biomaterials, but also a decent diffusion barrier due to positive charges at a physiological pH. The mesh size of the acetonirile PPF is small enough to impede molecules larger than hydrogen peroxide. (Muguruma et al., 2000; Hiratsuka et al., 2000). This size discrimination effect comes from a highly branched network structure formed upon the random nature of the plasma polymerization. Our group reported (Hiratsuka et al., 2000) that the diffusion coefficient of hydrogen peroxide through the acetonitrile PPF was larger than those of other clinical interferants such as ascorbic acid, acetaminophen, and uric acid.
A. Hiratsuka et al. / Biosensors and Bioelectronics 21 (2005) 957–964
Fig. 7. Calibration curve obtained using a device in 20 mM phosphate buffer (pH 7.4). The oxidation current at +410 mV vs. Ag/AgCl upon cyclic voltammetry was used. Insets: upper left, time response of the device when 5 mM of glucose added. Response time was 4 s; lower light, an enlargement of the low glucose concentration region. Each point represents the average and the vertical bars designate the standard deviation (n = 3).
Fig. 7 shows a calibration curve obtained using a device in phosphate buffer (20 mM, pH 7.4). The current response increased linearly from 1 mM to 20 mM. The electrochemical responses were well differentiated enough to cover the clinically significant concentration range of glucose within the dynamic range of the sensor. The apparent Michealis conapp stant (KM ) and the maximum limiting current (Imax ), calculated from the Michealis–Menten analysis, were 17.7 mM and 8.9 A/cm2 , respectively. A rapid response time (<5 s) was another benefit of our thin film structure. When it comes to the dynamic range and the response time, however, our results were almost identical to those of our previous PPFbased glucose biosensors without an electron transfer mediator (Muguruma et al., 2000; Hiratsuka et al., 2001). The primary benefit of the ferrocene film layer may involve the lower constant potential (410 mV), than the conventional ones (600–700 mV), for amperometric oxidation of hydrogen peroxide. This, in turn, is energy-saving and less prone to high background signal due to the interfering ascorbic acid. The operational stability of the devices was tested. Measurements up to 20 mM glucose were repeated 20 times, for which the observed current dropped only to more than 90% of the first response.
4. Conclusions We successfully demonstrated a microfabricated array of needle-shaped glucose biosensors as a BioMEMS application. The electron transfer mediator, vinylferrocene, was massively deposited as an intermediate layer by the organic plasma polymerization method. The hydrophobicity of the mediator layer could be overcome by superimposing addi-
963
tional hydrophilic plasma polymerized layer on it. As a vital result, the efficiency of the electron transport at the heterogeneous interface was remarkably enhanced without the use of special electrolytes required for vinylferrocene. As a simple and reproducible dry process aiming at mass production, our methodology introduced here provides multi-faceted advantages in devising any kind of bioelectronic devices for the sake of immobilization of a biological component; anti-fouling coating, and wettability control. There are some advantages of micro-biosensors, which are related with their small feature sizes. First, they are capable of running a small-volume, suitable for handling expensive or rarely available biological samples. Second, it can also achieve high sensitivity and fast response time. Third, the design of a massive biosensor array can be implemented for high throughput detection of multiple anlaytes. Last, in case of in vivo applications, high spatial resolution can be achieved.
References Alonso, B., Armada, P.G., Losada, J., Cuadrado, I., Gonz´alez, B., Casado, C.M., 2004. Amperometric enzyme electrodes for aerobic and anaerobic glucose monitoring prepared by glucose oxidase immobilized in mixed ferrocene-cobaltcenium dendrimers. Biosen. Bioelectron. 19, 1617–1625. Auroux, P.-A., Iossifidis, D., Reyees, D.R., Manz, A., 2002. Micro total analysis systems. 2. Analytical standard operations and applications. Anal. Chem. 74, 2637–2652. Bard, A.J., Faulkner, L.R., 1980. Electrochemical Methods Fundamental and Applications. John Wiley & Sons, New York. Cass, A.E.G., Davis, G., Francis, G.D., Hill, H.A.O., Aston, W.J., Higgins, I.J., Plotkin, E.V., Scott, L.D.L., Turner, A.P.F., 1984. Ferrocenemediated enzyme electrode for amperometric determination of glucose. Anal. Chem. 56, 667–671. Daum, P., Murray, W.R., 1979. Chemically modified electrodes. Part XXII. Solvent effects on the electrochemistry of thin films of plasma polymerized vinylferrocene. J. Electroanal. Chem. 103, 289– 294. Hiratsuka, A., Muguruma, H., Nagata, R., Nakamura, R., Sato, K., Uchiyama, S., Karube, I., 2000. Mass transport behavior of electrochemical species through plasma-polymerized thin film on platinum electrode. J. Membr. Sci. 175, 25–34. Hiratsuka, A., Kojima, K., Suzuki, H., Muguruma, H., Ikebukuro, K., Karube, I., 2001. Integration of microfabricated needle-type glucose sensor devices with a novel thin-film Ag/AgCl electrode and plasmapolymerized thin film: mass production techniques. Analyst 126, 658–663. Hiratsuka, A., Muguruma, H., Lee, K.-H., Karube, I., 2004. Organic plasma process for simple and substrate-independent surface modification of polymeric BioMEMS devices. Biosen. Bioelectron. 19, 1667–1672. Kampfrath, G., Hintsche, R., 1989. Plasma-polymerized thin films for enzyme immobilization in biosensors. Anal. Lett. 22, 2423–2431. Kurita, R., Tabei, H., Iwasaki, Y., Hayashi, K., Sunagawa, K., Niwa, O., 2004. Biocompatible glucose sensor prepared by modifying protein and vinylferrocene monomer composite membrane. Biosens. Bioelectron. 20, 518–523. Lawrence, N.S., Deo, R.P., Wang, J., 2004. Biocatalytic carbon paste sensors based on a mediator pasting liquid. Anal. Chem. 76, 3735– 3739. Muguruma, H., Karube, I., 1999. Plasma-polymerized films for Biosensors. TrAC, Trends Anal. Chem. 18, 62–68.
964
A. Hiratsuka et al. / Biosensors and Bioelectronics 21 (2005) 957–964
Muguruma, H., Hiratsuka, A., Karube, I., 2000. Thin-film glucose biosensor based on plasma-polymerized film: simple design for mass production. Anal. Chem. 72, 2671–2675. Mutlu, M., Mutlu, S., Boyacii, H., Alp, B., Piskin, E., 1998. Highlinearity glucose enzyme electrodes for food industries: preparation by a plasma polymerization technique. ACS Symp. Ser. 690, 57–65. Mutlu, S., Mutlu, M., Piskin, E., 1994. Sandwich-type amperometric enzyme electrodes for determination of glucose. ACS Symp. Ser. 556, 71–83. Nowak, R.J., Schultz, F.A., Uma˜na, M., Lam, R., Murray, R.W., 1980. Chemically modified electrodes. Radiofrequency plasma polymerization of vinylferrocene on glassy carbon and platinum electrode. Anal. Chem. 52, 315–321. Padeste, C., Steiger, B., Grubelnik, A., Tiefenauer, L., 2004. Molecular assembly of redox-conductive ferrocene-streptavidin conjugates— towards bio-electrochemical devices. Biosen. Bioelectron. 20, 545– 552.
Saito, T., Watanabe, M., 1998. Characterization of poly(vinylferroceneco-2-hydroxyethyl methacrylate) for use as electron transfer mediator in enzyme glucose sensor. React. Funct. Polym. 37, 263–269. Saito, T., Watanabe, M., 1999. Electron transfer reaction from glucose oxidase to an electrode via redox copolymers. Polym. J. 31, 1149– 1154. Suzuki, H., Hiratsuka, A., Sasaki, S., Karube, I., 1998. Problems associated with the thin-film Ag/AgCl reference electrode and a novel structure with improved durability. Sens. Actuators, B B46, 104–113. Yasuda, H., 1985. Plasma Polymerization. Academic Press, New York. Yoshimura, K., Hozumi, K., 1996. Response characteristics of a glucose electrode with a sensing membrane prepared by plasma polymerization. Microchem. J. 53, 404–412. Zhou, D.-M., Ju, H.-X., Chen, H.-Y., 1997. A miniaturized glucose biosensor based on the coimmobilization of glucose oxidase and ferrocene perchlorate in nafion at a microdisk platinum electrode. Sens. Actuators, B 40, 89–94.
Electron transfer mediator micro-biosensor fabrication by organic plasma process Atsunori Hiratsuka a , Ken-ichi Kojima b , Hitoshi Muguruma c,∗ , Kyong-Hoon Lee d , Hiroaki Suzuki b , Isao Karube a,e a
c
Laboratory of Advanced Bioelectronics, National Institute of Advanced Industrial Science and Technology, 1-1-1 Higashi, Tsukuba-shi, Ibaraki 305-8562, Japan b Institute of Materials Science, Tsukuba University, 1-1-1 Tennodai, Tsukuba-shi, Ibaraki 305-8573, Japan Department of Electronic Engineering, Shibaura Institute of Technology, 3-9-14 Shibaura, Minato-ku, Tokyo 108-8548, Japan d Department of Mechanical Engineering, Northwestern University, 2145 Sheridan Road, Evanston, Illinois 60208, USA e Faculty of Bionics, Tokyo University of Technology, Katakura 1404-1, Hachioji-shi, Tokyo 192-0982, Japan Received 27 January 2005; received in revised form 25 February 2005; accepted 1 March 2005 Available online 5 April 2005
Abstract We propose a new strategy for constructing a mediator-type biosensor as a Bio-MicroElectroMechanical Systems (BioMEMS) application. A vinylferrocene plasma-polymerized film (PPF) was deposited directly onto the surface of an electrode under dry conditions. The resulting redox film was extremely thin, adhered well onto a substrate (electrode), and had a highly crosslinked network structure. This technique, capable of polymeric deposition of any kind of monomer, can also serve the purpose of anti-fouling coating, or layer-to-layer interface creation. With a subsequent plasma process, additional polymeric layer of hydrophilic acetonitrile was superimposed onto the existing vinylferrocene-PPF surface to offer crucial features such that the wettability could be adjusted for a better electron transfer, and amino functional groups could be attached to immobilize a large amount of enzyme. Based upon this scheme, the device fabrication could be designed in a manner that the whole procedure was made up of dry wafer-handling processes, which is compatible with mass production. A prototype device was fabricated to have an array of needle-shaped amperometric micro-biosensors. The resultant thin polymer layer carried a large number of the mediator molecules, accomplishing a lower overpotential (+410 mV) and a rapid response time (<5 s). Stressing the advantages of the plasma polymerization process together with some additional features accomplished in our device fabrication, we would discuss new possibilities in the field of BioMEMS. © 2005 Elsevier B.V. All rights reserved. Keywords: Plasma-polymerized film; Electron transfer mediator; Ferrocene; Glucose biosensor; Enzyme
1. Introduction Amperometric enzyme biosensors for biochemical and medical uses have been widely and seriously developed for the last two decades. This type of biosensor, which is a combination between a biocatalyst and an electronic device, converts electron transfers involved in an enzymatic action into an electrochemical response, which provides quantitative information on a target analyte. For constructing a glucose ∗
Corresponding author. Tel.: +81 3 5427 7813; fax: +81 3 5476 3165. E-mail address: [email protected] (H. Muguruma).
0956-5663/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2005.03.002
biosensor, which has been one of the most intensively and successfully commercialized, the electron transfer mediator is a widely accepted technique to enhance the efficiency of the sensor signal acquisition. The mediator inter-relays the electron transfer between the catalytic center of the enzyme and the electrode surface. Among so many kinds of mediators, ferrocene has been undisputedly regarded as one of the most representative, since the first introduction of its concept for the mediator-based enzymatic biosensor (Cass et al., 1984). Usually, the enzyme and mediators are embedded or coimmobilized in a polymer cast on an electrode surface. Due
958
A. Hiratsuka et al. / Biosensors and Bioelectronics 21 (2005) 957–964
to their size, the enzyme molecules are well fixed within the polymer, but a possible leak of the much smaller mediator molecules causes problems such as sensor deactivation and contamination of sample. Moreover, when it comes to the sensor design, the wettability balance needs to be tuned between the hydrophobic ferrocene and the hydrophilic enzyme. The final design considering these aspects significantly determines the sensor characteristics. As such, many strategies for the electron transfer system have been proposed (Padeste et al., 2004; Kurita et al., 2004; Alonso et al., 2004; Lawrence et al., 2004). The field of micro electromechanical systems (MEMS) or micro total analysis systems (TAS) has played a significant role in the development of fast chemical analysis, which is usually related with running small-volume samples on a microsensor array or a microfluidic device (Auroux et al., 2002). The microscale approach is especially important to bioanalytical systems (so-called “BioMEMS”) due to the inherent nature of bioreactions in terms of the sample availability and the slow reaction speed. In case of enzyme biosensor-integrated MEMS systems, one of the key fabrication issues may be how to immobilize biosensing molecules contained in a liquid onto a given substrate, in line with dry processes. However, most of immobilization methods available are unsuitable for a manufacture-oriented design of biosensor and/or BioMEMS because “wet processes” are hardly unavoidable. Furthermore, they are conventionally spin- or dip-coated, which makes it difficult to achieve thin, homogeneous, reproducible, and strongly adhesive film properties. Taking all these into consideration, we propose a new strategy for the fabrication of an electrochemical enzymatic biosensor with an electron mediator for BioMEMS application, whose immobilization step relies upon organic plasma polymerization (Yasuda, 1985). Plasma-polymerized films (PPFs) of vinylferrocene were reported by the Murray’s group (Nowak et al., 1980; Daum and Murray, 1979), but these were mainly about the electrochemical characterizations instead of sensor applications. The amperometric enzyme biosensors based on plasma polymerizaion were early reported by some researchers (Kampfrath and Hintsche, 1989; Yoshimura and Hozumi, 1996; Mutlu et al., 1994, 1998). Likewise, we have been exploiting organic plasma processes for the construction of biosensors and/or BioMEMS devices (Muguruma and Karube, 1999; Muguruma et al., 2000; Hiratsuka et al., 2001, 2004). In this article, the possibility of such concept is expanded to a biosensor involving the “electron transfer mediator”, as a precursor to mass production-compatible bioMEMS devices.
2. Materials and methods 2.1. Materials Acetonitrile, potassium dihydrogenphosphate, disodium hydrogenphosphate, d-glucose, potassium chloride, glutar-
Fig. 1. Equipment and geometry used for plasma polymerization.
aldehyde, hydrogen chloride and hydrogen peroxide were purchased from Wako (Osaka, Japan). The enzyme glucose oxidase (GOD) from Aspergillus niger (EC 1.1.3.4, type VIIS, 181600 units/g) was purchased from Sigma (St. Louis, MO, USA). Hexamethyldisiloxane (HMDS), tetraethylammonium perchlorate (Et4 NClO4 ), and vinylferrocene purchased from Aldrich (Milwaukee, WI, USA). All reagents used in this work were prepared without further purifications. All solutions were made with deionized water (18 M cm resistivity). 2.2. Plasma polymerization The apparatus used for the plasma polymerization (Model BP-1, Samco International Laboratories, Inc., Kyoto, Japan) is shown in Fig. 1. Two electrodes were placed horizontally 15 cm above the sample stage (external electrode reactors). An RF generator (Model RFG-300, Samco), coupled to an automatching network to minimize the reflective power, was employed. The working frequency of the power supply was 13.56 MHz. The vinylferrocene monomer was directly placed on the sample stage, which was heated to ca. 60 ◦ C in order to provide a high monomer pressure. The acetonitrile monomer was placed in a bottle reservoir, which was connected to an inlet line through the mass flow controller. With plasma-free exposure of the electrode to the vinylferrocene vapor, we could not find any electrochemical- or other experimental-evidences that ferrocene was immobilized. 2.3. Device fabrication The device, which is an electrochemical micro-biosensor with the three electrodes configuration, was fabricated on the basis of semiconductor layer-by-layer processes as shown in Fig. 2. It was formed on a 150-m thick glass substrate, whose planar dimension was approximately 50 mm × 50 mm. All metal layers were sputter-deposited and patterned by a lift-off process. Glass slides used to make thin
A. Hiratsuka et al. / Biosensors and Bioelectronics 21 (2005) 957–964
Fig. 2. Decomposed structure of the needle-type biosensor: (1) Pt for counter electrode (200 nm); (2) Cr for adhesion layer (40 nm); (3) glass substrate; (4) Cr for adhesion layer (40 nm); (5) Pt for working electrode (200 nm); (6) Ag/AgCl for reference electrode (300 nm); (7) polyimide (7 m); (8) vinylferrocene plasma-polymerized film (150 nm); (9) acetonitrile plasmapolymerized film, (20 nm) introducing amino groups onto the surface; (10) glutaraldehyde reacted with the amino groups on 9 layer; (11) immobilized glucose oxidase via glutaraldehyde. Inset is the cross-section of the working electrode.
film electrodes were cleaned in a 50% nitric acid for an hour and then rinsed with water and acetone. The working, reference, and counter electrodes were formed with 200-nm thick platinum (Pt) backbone patterns. The Pt thin films were sputtered under normal condition, with an apparatus manufactured by Shibaura Engineering Works Co., Ltd. (Model CFS-4ES-231, Tokyo, Japan). The Pt film thickness (200 nm) was determined from a surface profiler (Dektak ST, Veeco Instruments Inc., Tokyo, Japan). A 40-nm thick chromium (Cr) intermediate layer was used to promote the adhesion of the Pt layer. A 300-nm thick silver layer was formed only on the reference electrode area. To delineate the active area for each electrode a 7m thick polyimide layer was formed. In curing the polyimide, the devices were subjected to bakings for 15 min at 150 ◦ C, 15 min at 200 ◦ C, and 30 min at 300 ◦ C. The dimension of the openings for the working electrode was 0.5 mm × 1.0 mm. A novel thin-film Ag/AgCl anode structure was used in this device (Suzuki et al., 1998). As reported in the previous study, the lifetime of a usual thin-film made of Ag/AgCl is very restricted and its element starts to deteriorate once the film is exposed to an electrolyte solution. In order to solve
959
the problem, the entire silver layer was protected with the polyimide film and the AgCl layer was grown from one edge of the silver pattern. The protecting layer was effective in suppressing the dissolution of AgCl. A vinylferrocene PPF was deposited onto a platinum electrode. The deposition parameters were as follows: power; 80W, pressure; 1.0 Pa, plasma exposure time; 2 min (150 nm of film thickness). The vinylferrocene monomer was directly placed on the electrode (see Fig. 1) under glow discharge according to the Murray’s method (Nowak et al., 1980). Subsequently, an acetonitrile PPF was superimposed onto the vinylferrocene layer above the platinum electrode. The deposition parameters were as follows: power; 80-W, flow rate; 15 mL min−1 , pressure; 2.1 Pa, plasma exposure time; 1 min (20 nm of film thickness). The surface of the acetonitrile PPF had a multitude of amino groups to which enzyme molecules could be immobilized. At the last stage, the immobilization of the enzyme was achieved by applying 2.5% aqueous glutaraldehyde (GA) solution to the surface of the plasma-polymerized film, washing with water, and then dropping 10 mg/mL GOD (Sigma) solution in a phosphate buffer (20 mM, pH 7) onto the film. After 10 minutes, the device was washed with water and stored at 4 ◦ C until use. Twentyeight sensors were batch-fabricated on the glass substrate shown in Fig. 3. Each and every sensor in a needle shape (0.5 mm × 1.0 mm) was obtained by cutting the substrate with a dicing saw (Model DAD321, Disco, Tokyo, Japan). 2.4. Measurement Electrochemical studies were performed with a BAS50W potentiostat (Bioanalytical Systems, West Lafayette, IN, USA). The three-electrode configuration equipped with the fabricated devices was used. Electrochemical measurements were carried out in a 5-mL vessel at laboratory ambient temperature (20 ± 1 ◦ C). The supporting electrolyte was a phosphate buffer (20 mM, pH 7.4).
3. Results and discussion 3.1. Characterization of vinylferrocene plasma-polymerized film on sputtered Pt thin film A plasma-polymerized vinylferrocene film on an electrode was studied by the Murray’s group (Nowak et al., 1980; Daum and Murray, 1979), but there were still good reasons enough for us to revisit this topic for our own purpose: first, the structure and properties of any PPF strongly depend on the reactor geometry, the discharge power, and the chamber pressure of a given plasma generator. Second, we used sputtered thin film electrodes, on which a plasma-polymerized film would present different characteristics from that of a “bulk” electrode (Nowak et al., 1980). A scanning electron microscopic (SEM) image and a cyclic voltammogram of a Pt thin film electrode covered with a vinylferrocene PPF are shown in
960
A. Hiratsuka et al. / Biosensors and Bioelectronics 21 (2005) 957–964
Fig. 3. Picture of arrayed needle biosensors. Twenty-eight of biosensors array are cut into each device.
Fig. 4. The distinct cathodic and anodic peaks due to the ferrocene/ferrocenium couple were observed from the film prepared by the 80-W plasma power. The difference between the cathodic- and the anodic-peak potentials (Ep ) was 80 mV. The formal potential (E0 ) versus Ag/AgCl was +450 mV. The ferrocenes in the plasma-polymerized film seemed substitutionally similar to those in a polyvinylferrocene film (Nowak et al., 1980). The surface coverage of the ferrocene redox sites was estimated to 4.5 × 10−8 mol/cm2 . The redox peak of a vinylferrocene PPF deposited with the 20-W discharge power was smaller than that with the 80-W. This can be explained by our observation (Fig. 4a) that the 20-W discharge power was too low to form a homogeneous structure with an ample number of the redox ferrocenes.
The peak of the redox vinylferrocene PPF formed with the 120-W power was smaller and more widened with a positive potential shift than that of the 80-W. This may also be explained by another observation of ours (Fig. 4c) that the high discharge power of 120-W gave a homogeneous film structure but destructively transformed a greater part of the redox ferrocenes into electrochemically inactive Fe(III) species (Nowak et al., 1980). Consequently, we used this value (80-W) as the optimal plasma discharge power in this work. The dependence of cyclic voltammograms on their scan rates was investigated using one of the fabricated electrodes (Fig. 5). With the increasing scan rate, the cathodic peak potential shifted toward in the more negative direction, and the anodic peak potential in the more positive direction,
A. Hiratsuka et al. / Biosensors and Bioelectronics 21 (2005) 957–964
961
Fig. 4. Cyclic voltammograms and surface image (scanning electron spectroscopy) of electrode modified by vinylferrocene plasma-polymerized film as a function of the power of polymerization: (a) 20-W; (b) 80-W; (c) 120-W. Solvent was 0.1 M Et4 NClO4 /CH3 CN. Scan rate was 10 mV s−1 .
as generally observed, probably due to the hysteresis. This observation was more obvious than in the case of a bare electrode (data not shown), which may be related with the surface diffusion. On the other hand, the magnitude of a peak current was linearly proportional to the square root of the relevant scan rate. This can be explained by the Randles– Sevcik equation (Bard and Faulkner, 1980), applicable to the case of a semi-infinite volume of a diffusing reactant in contact with the electrode. 3.2. Device fabrication Generally, the lifetime of a conventional thin-film Ag/AgCl electrode is shortened and the electrode deteriorates in a highly concentrated chloride solution. In order to fix this problem, we employed a new type of Ag/AgCl electrode with high durability (Suzuki et al., 1998), whose design was intended for integration into a miniaturized device. The
Ag/AgCl layer was covered with a polyimide protective layer and 20 holes (0.1 mm × 0.1 mm) were opened on the layer by photolithographic patterning. In our work, we were highly conscious of developing a fabrication scheme compatible with mass-production. Most of all, it is noteworthy that our process could leave out the use of a solvent from the corresponding conventional steps. For example, the matrix for immobilizing the enzyme and the mediator could be created only after several minutes’ exposure of a substrate to plasma generated under vacuum. This enabled us to continue to run dry wafer processes up to the last stage, at which a wet chemistry could be introduced for the enzyme immobilization without meaningful interruption of the overall procedure. Here, one might imagine even advanced protocols more appropriate for the mass production, in which the sputtering of thin films and their surface modifications with plasma polymerization are carried out sequentially at a batch.
962
A. Hiratsuka et al. / Biosensors and Bioelectronics 21 (2005) 957–964
the discriminative peak was observed at +450 mV, indicating that the hydrogen peroxide was oxidatively decomposed by the ferrocenium ions (Fc+ ). The mechanism can be expressed as follows. First, glucose is oxidized by the immobilized enzyme. -d-glucose + O2 → (GOD) ␦-gluconolactone + H2 O2 (1) The hydrogen peroxide produced diffuses through the acetonitrile PPF layer and reaches the vinlyferrocene PPF layer. Subsequenly, the hydrogen peroxide is oxidized by ferricenium ions (Zhou et al., 1997). 2Fc+ + H2 O2 → 2Fc + 2H+ + O2
(2)
Meanwhile, ferrocenes (Fc) reductively created from ferriceniums are oxidized back electrochemically, resulting in electron transfers to the electrode. Fc → Fc+ + e−
Fig. 5. Cyclic voltammogram of plasma-polymerized vinylferrocene deposited in platinum electrode thin film in 0.1 M Et4 NClO4 /CH3 CN. (a–e) 5, 10, 20, 50, 100, and 200 mV s−1 , respectively. Insets: dependence of peak current (Ip ) on square root of potential sweep rate.
3.3. Sensor characteristics Fig. 6 shows a cyclic voltammogram, which was obtained in the absence and presence of glucose. The catalytic current resulted from the added glucose, which diffuses into the film and reacts oxygen in the presence of glucose oxidase. When hydrogen peroxide produced by the enzymatic action was directly oxidized with the surface of the electrode, large and broad current increments were observed in a potential range from +400 to +800 mV (Hiratsuka et al., 2001). In this work,
Fig. 6. Cyclic voltammograms obtained using the device in 20 mM phosphate buffer (pH 7.4): (a) without glucose and (b) with 10 mM of glucose. The scan rate was 10 mV s−1 .
(3)
To our knowledge, this work is probably the first report on a vinylferrocene-mediated biocatalytic electric current observed in a generic aqueous medium. In general, observation of such a current should be accompanied by the use of a specific supporting electrolyte such as LiClO4 and Et4 NClO4 (Daum and Murray, 1979). We believe that this can be ascribed to the hydrophilic acetonitrile PPF layer, without which the peak current coming from the enzymatic reaction as shown in Fig. 6 was “not” observed. Only with the hydrophobic surface of the vinylferrocene PPF, first, the mass transport would be seriously retarded across the heterogeneous interface between an aqueous medium and the ferrocene membrane, and, second, it would be difficult to immobilize the enzyme thereon. The previous solution to this problem was introduction of a hydrophilic polymer matrix (Saito and Watanabe, 1998, 1999). Compared with these, our method is capable of controlling the film properties in a simpler, more reproducible and robust manner, which provides a more manufacture-oriented protocol. The structure of the acetonitrile PPF consists of highlybranched- and incompletely-crosslinked aliphatic hydrocarbon backbone chains. The film contains a lot of nitrogen atoms in various forms of functional groups, among which 32% were primary amines (Hiratsuka et al., 2000). This provides not only a good matrix for the immobilization of biomaterials, but also a decent diffusion barrier due to positive charges at a physiological pH. The mesh size of the acetonirile PPF is small enough to impede molecules larger than hydrogen peroxide. (Muguruma et al., 2000; Hiratsuka et al., 2000). This size discrimination effect comes from a highly branched network structure formed upon the random nature of the plasma polymerization. Our group reported (Hiratsuka et al., 2000) that the diffusion coefficient of hydrogen peroxide through the acetonitrile PPF was larger than those of other clinical interferants such as ascorbic acid, acetaminophen, and uric acid.
A. Hiratsuka et al. / Biosensors and Bioelectronics 21 (2005) 957–964
Fig. 7. Calibration curve obtained using a device in 20 mM phosphate buffer (pH 7.4). The oxidation current at +410 mV vs. Ag/AgCl upon cyclic voltammetry was used. Insets: upper left, time response of the device when 5 mM of glucose added. Response time was 4 s; lower light, an enlargement of the low glucose concentration region. Each point represents the average and the vertical bars designate the standard deviation (n = 3).
Fig. 7 shows a calibration curve obtained using a device in phosphate buffer (20 mM, pH 7.4). The current response increased linearly from 1 mM to 20 mM. The electrochemical responses were well differentiated enough to cover the clinically significant concentration range of glucose within the dynamic range of the sensor. The apparent Michealis conapp stant (KM ) and the maximum limiting current (Imax ), calculated from the Michealis–Menten analysis, were 17.7 mM and 8.9 A/cm2 , respectively. A rapid response time (<5 s) was another benefit of our thin film structure. When it comes to the dynamic range and the response time, however, our results were almost identical to those of our previous PPFbased glucose biosensors without an electron transfer mediator (Muguruma et al., 2000; Hiratsuka et al., 2001). The primary benefit of the ferrocene film layer may involve the lower constant potential (410 mV), than the conventional ones (600–700 mV), for amperometric oxidation of hydrogen peroxide. This, in turn, is energy-saving and less prone to high background signal due to the interfering ascorbic acid. The operational stability of the devices was tested. Measurements up to 20 mM glucose were repeated 20 times, for which the observed current dropped only to more than 90% of the first response.
4. Conclusions We successfully demonstrated a microfabricated array of needle-shaped glucose biosensors as a BioMEMS application. The electron transfer mediator, vinylferrocene, was massively deposited as an intermediate layer by the organic plasma polymerization method. The hydrophobicity of the mediator layer could be overcome by superimposing addi-
963
tional hydrophilic plasma polymerized layer on it. As a vital result, the efficiency of the electron transport at the heterogeneous interface was remarkably enhanced without the use of special electrolytes required for vinylferrocene. As a simple and reproducible dry process aiming at mass production, our methodology introduced here provides multi-faceted advantages in devising any kind of bioelectronic devices for the sake of immobilization of a biological component; anti-fouling coating, and wettability control. There are some advantages of micro-biosensors, which are related with their small feature sizes. First, they are capable of running a small-volume, suitable for handling expensive or rarely available biological samples. Second, it can also achieve high sensitivity and fast response time. Third, the design of a massive biosensor array can be implemented for high throughput detection of multiple anlaytes. Last, in case of in vivo applications, high spatial resolution can be achieved.
References Alonso, B., Armada, P.G., Losada, J., Cuadrado, I., Gonz´alez, B., Casado, C.M., 2004. Amperometric enzyme electrodes for aerobic and anaerobic glucose monitoring prepared by glucose oxidase immobilized in mixed ferrocene-cobaltcenium dendrimers. Biosen. Bioelectron. 19, 1617–1625. Auroux, P.-A., Iossifidis, D., Reyees, D.R., Manz, A., 2002. Micro total analysis systems. 2. Analytical standard operations and applications. Anal. Chem. 74, 2637–2652. Bard, A.J., Faulkner, L.R., 1980. Electrochemical Methods Fundamental and Applications. John Wiley & Sons, New York. Cass, A.E.G., Davis, G., Francis, G.D., Hill, H.A.O., Aston, W.J., Higgins, I.J., Plotkin, E.V., Scott, L.D.L., Turner, A.P.F., 1984. Ferrocenemediated enzyme electrode for amperometric determination of glucose. Anal. Chem. 56, 667–671. Daum, P., Murray, W.R., 1979. Chemically modified electrodes. Part XXII. Solvent effects on the electrochemistry of thin films of plasma polymerized vinylferrocene. J. Electroanal. Chem. 103, 289– 294. Hiratsuka, A., Muguruma, H., Nagata, R., Nakamura, R., Sato, K., Uchiyama, S., Karube, I., 2000. Mass transport behavior of electrochemical species through plasma-polymerized thin film on platinum electrode. J. Membr. Sci. 175, 25–34. Hiratsuka, A., Kojima, K., Suzuki, H., Muguruma, H., Ikebukuro, K., Karube, I., 2001. Integration of microfabricated needle-type glucose sensor devices with a novel thin-film Ag/AgCl electrode and plasmapolymerized thin film: mass production techniques. Analyst 126, 658–663. Hiratsuka, A., Muguruma, H., Lee, K.-H., Karube, I., 2004. Organic plasma process for simple and substrate-independent surface modification of polymeric BioMEMS devices. Biosen. Bioelectron. 19, 1667–1672. Kampfrath, G., Hintsche, R., 1989. Plasma-polymerized thin films for enzyme immobilization in biosensors. Anal. Lett. 22, 2423–2431. Kurita, R., Tabei, H., Iwasaki, Y., Hayashi, K., Sunagawa, K., Niwa, O., 2004. Biocompatible glucose sensor prepared by modifying protein and vinylferrocene monomer composite membrane. Biosens. Bioelectron. 20, 518–523. Lawrence, N.S., Deo, R.P., Wang, J., 2004. Biocatalytic carbon paste sensors based on a mediator pasting liquid. Anal. Chem. 76, 3735– 3739. Muguruma, H., Karube, I., 1999. Plasma-polymerized films for Biosensors. TrAC, Trends Anal. Chem. 18, 62–68.
964
A. Hiratsuka et al. / Biosensors and Bioelectronics 21 (2005) 957–964
Muguruma, H., Hiratsuka, A., Karube, I., 2000. Thin-film glucose biosensor based on plasma-polymerized film: simple design for mass production. Anal. Chem. 72, 2671–2675. Mutlu, M., Mutlu, S., Boyacii, H., Alp, B., Piskin, E., 1998. Highlinearity glucose enzyme electrodes for food industries: preparation by a plasma polymerization technique. ACS Symp. Ser. 690, 57–65. Mutlu, S., Mutlu, M., Piskin, E., 1994. Sandwich-type amperometric enzyme electrodes for determination of glucose. ACS Symp. Ser. 556, 71–83. Nowak, R.J., Schultz, F.A., Uma˜na, M., Lam, R., Murray, R.W., 1980. Chemically modified electrodes. Radiofrequency plasma polymerization of vinylferrocene on glassy carbon and platinum electrode. Anal. Chem. 52, 315–321. Padeste, C., Steiger, B., Grubelnik, A., Tiefenauer, L., 2004. Molecular assembly of redox-conductive ferrocene-streptavidin conjugates— towards bio-electrochemical devices. Biosen. Bioelectron. 20, 545– 552.
Saito, T., Watanabe, M., 1998. Characterization of poly(vinylferroceneco-2-hydroxyethyl methacrylate) for use as electron transfer mediator in enzyme glucose sensor. React. Funct. Polym. 37, 263–269. Saito, T., Watanabe, M., 1999. Electron transfer reaction from glucose oxidase to an electrode via redox copolymers. Polym. J. 31, 1149– 1154. Suzuki, H., Hiratsuka, A., Sasaki, S., Karube, I., 1998. Problems associated with the thin-film Ag/AgCl reference electrode and a novel structure with improved durability. Sens. Actuators, B B46, 104–113. Yasuda, H., 1985. Plasma Polymerization. Academic Press, New York. Yoshimura, K., Hozumi, K., 1996. Response characteristics of a glucose electrode with a sensing membrane prepared by plasma polymerization. Microchem. J. 53, 404–412. Zhou, D.-M., Ju, H.-X., Chen, H.-Y., 1997. A miniaturized glucose biosensor based on the coimmobilization of glucose oxidase and ferrocene perchlorate in nafion at a microdisk platinum electrode. Sens. Actuators, B 40, 89–94.