Anal. Chem. 2000, 72, 2671-2675
Thin-Film Glucose Biosensor Based on Plasma-Polymerized Film: Simple Design for Mass Production Hitoshi Muguruma,*,† Atsunori Hiratsuka,‡ and Isao Karube‡
Department of Environmental Systems Engineering, Kochi University of Technology, Tosayamada, Kochi 782-8502, and Research Center for Advanced Science and Technology, The University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo 153-8904, Japan
We propose a simple thin-film glucose biosensor based on a plasma-polymerized film. The film is deposited directly onto the substrate under dry conditions. The resulting films are extreme thin, adhere well onto the substrate (electrode), and have a highly cross-linked network structure and functional groups, such as amino groups, which enable a large amount of enzyme to be immobilized. Since this design allows fabrication through a dry process, with the exception of the enzyme immobilization, which is the last stage of the process, the chip fabrication can be designed as a full-wafer process to achieve mass production compatibility. The resulting sensors produced using this film are more reproducible, exhibit lower noise, and reduce the effect of interference to a greater degree than sensors made using conventional immobilization methods, e.g., via 3-(aminopropyl)triethoxysilane. The obtained film is a good interfacial design between enzyme and electrode; enzyme two-dimensionally locates very close to the electrode in a manner that is quite reproducible. Therefore, a wide dynamic range (up to 60 mM) and rapid response time (11.5 ( 0.8 s) were obtained. Because of its highly cross-linking network structure, the amperometric response due to interferences such as ascorbic acid and acetaminophen was reduced by size discrimination of plasma-polymerized films. The development of miniaturized biosensors for in vivo or in vitro monitoring of physiologically important chemicals is an important tool in the field of medicine.1-5 Miniaturized biosensors are typically fabricated using micromachine and/or silicon technology processes, such as vacuum evaporation, plasma etching, * To whom correspondence should be addressed. E-mail:
[email protected]. † Kochi University of Technology. ‡ University of Tokyo. (1) Willins E.; Atanasov, P. Med. Eng. Phys. 1996, 18, 273-288. (2) Hu, Y.; Wilson, G. S. J. Neurochem. 1997, 68, 1745-1752. (3) Ishikawa, M.; Schmidtke, D. W.; Raskin, P.; Quinn, C. A. P. J. Diab. Comput. 1998, 12, 295-301. (4) Jobst, G.; Moser, I.; Varahram, M.; Svasek, P.; Aschauer, E.; Trajanoski, Z.; Wach, P.; Kotanko, P.; Skrabal, F.; Urban, G. Anal. Chem. 1996, 68, 3173-3179. (5) Trajanoski, Z.; Wach, P.; Gfrerer, R.; Jobst, G.; Urban, G.; Kotanko, P.; Skrabal, F. Biosens. Bioelectron. 1996, 11, 479-487. 10.1021/ac000014n CCC: $19.00 Published on Web 05/04/2000
© 2000 American Chemical Society
photolithography, and lift-off.6-10 Miniaturized biosensors may offer several advantages over conventional medical techniques. Miniaturized biosensors allow small features to be measured, providing extremely high spatial resolution. In addition, small measurement volumes are possible while retaining relatively high concentrations of the original sample. The small size of the device is also advantageous for the design of biosensor arrays for multiple detection.4,11-13 A further advantage of miniature biosensors is that these sensors can be produced at low cost.14-16 However, there is one problem with respect to how well the biological components (i.e., wet process) are combined to the surface of the transducers (i.e., dry process). The performance of the biosensors is dominated by the combination technique of these two components. This technique is often called “interfacial design”. Thin-film technology for interfacial design of miniaturized biosensors in order to maintain the designed high-performance sensor characteristics, such as response time, sensitivity, stability, dynamic range, reproducibility, and reusability, is thought to be a key technology. Thus, the conventional technique for interfacial design is the use of silane coupling,9,17-19 sol-gel matrix,20,21 cellulose acetate,22-24 Nafion,23-25 poly(L-lysine),26 electropolymerized film,27-29 and self(6) Kovacs, G. T.; Peterson, K.; Albin, M. Anal. Chem. 1996, 68, 407A-412A. (7) Bratten, C. D. T.; Cobbold, P. H.; Cooper, J. M. Anal. Chem. 1997, 69, 253-258. (8) Sreenivas, G.; Ang, S. S.; Fritsch, I.; Brown, W. D.; Gerhardt, G. A.; Woodward, D. J. Anal. Chem. 1996, 68, 1858-1864. (9) Murakami, Y.; Takeuchi, T.; Yokoyama, K.; Tamiya, E.; Karube, I.; Suda, M. Anal. Chem. 1993, 65, 2731-2735. (10) Steinkuhl, R.; Dumschat, C.; Sundermeier, C.; Hinkers, H.; Renneberg, R.; Cammann, K.; Knoll, M. Biosens. Bioelectron. 1996, 11, 187-190. (11) Ma˜da˜ras¸ , M. B.; Popescu, I. C.; Ufer, S.; Buck, R. P. Anal. Chim. Acta 1996, 319, 335-345. (12) Xie, B.; Danielsson, B. Anal. Lett. 1996, 29, 1921-1932. (13) Urban, G.; Jobst, G.; Keplinger, F.; Aschauer, E.; Tilado, O.; Fasching, R.; Kohl, F. Biosens. Bioelectron. 1992, 7, 733-739. (14) Perdomo, J.; Sundermeier, C.; Hinkers, H.; Morell, O. M.; Seifert, W.; Knoll, M. Biosens. Bioelectron. 1999, 14, 27-32. (15) Puig-Lleixa`, C.; Jime´nez, C.; Alonso, J.; Bartrolı´, J. Anal. Chim. Acta 1999, 389, 179-188. (16) Steinkuhl, R.; Sundermeier, C.; Hinkers, H.; Dumschat, C.; Cammann, K.; Knoll, M. Sens. Acuators B 1996, 33, 19-24. (17) Jung, S.-K.; Wilson, G. S. Anal. Chem. 1996, 68, 591-596. (18) Murray, R. W. Acc. Chem. Res. 1980, 13, 135-141 and references therein. (19) Turner, R. F. B.; Harrison, D. J.; Rajotte, R V.; Baltes, H. P. Sens. Actuators 1990, B1, 561-564. (20) Wang, B.; Li, B.; Wang, X.; Xu, G.; Wang, Q.; Dong, S. Anal. Chem. 1999, 71, 1935-1939.
Analytical Chemistry, Vol. 72, No. 11, June 1, 2000 2671
assembling monolayers.17,30 However, such methods are unsuitable for mass production of biosensors because they involve “wet processes”. Furthermore, Nafion and the other conventional polymers on electrode surfaces are difficult to control so as to produce thin, homogeneous, reproducible, and strongly adhesive films, because the polymers are fabricated by the spin- and dipcoating methods. Gooding et al.30,31 reported that most classical geometry enzyme electrodes, in which the enzyme is immobilized in a three-dimensional reaction matrix placed over a planer electrode, show the response of the sensor to be very sensitive to the thickness of the reaction matrix. Self-assembled monolayers (SAMs) of alkanethiols are available for gold electrodes because gold does not have a stable oxide. SAMs on platinum, which are the best catalytic electrodes for hydrogen peroxide caused by oxidative enzymatic reaction,30,32 are very weak and fragile.30,33 In addition, electron mediators such as ferrocene,34-36 osmium complex,37 tetrathiafulvalene,38 and hydroquione39 are widely used in most oxidaze-catalyzed biosensors. The mediators play a role as a medium of electron transfer between electrodes and enzymatic product located at the reaction center. The benefits are (i) the widening of the dynamic range of glucose detection influenced by oxygen depletion and (ii) a reduction in the effect of interfering agents such as ascorbic acid, since the mediator lowers the polarized potential for amperometric detection.37 However, some practical limitations exist. The sensor design for efficient electron transfer is complicated. In addition, most of the mediator is too toxic for use in in vivo monitoring. With these limitations, plasma-polymerized films (PPFs) are thought to offer a new alternative.40,41 The PPF is achieved in glow discharge or plasma in vapor phase. The properties of the film include (i) the ability to be made extremely thin (0.98); 0-60 mM, sensitivity; 0.56 µA mM-1 cm-2, error; (2.8% (5-30 mM). Device 2: linear range (>0.98); 0-20 mM, sensitivity; 1.34 µA mM-1 cm-2, error; (10.3% (5-30 mM).
Figure 2. Time base measurement of glucose in 20 mM phosphate buffer (pH 7.4). Applied potential, +700 mV vs Ag/AgCl. Device based on (a) plasma-polymerized film (device 1) and (b) 3-(aminopropyl)triethoxysilane (device 2). Sampling time is 0.5 s.
high-quality devices and a mass production system may be obtained. The PPF-modified Pt electrode and in situ process, used together, prevent the oxidation of the electrode surface by exposure to ambient air when the sputtered electrode is removed from the chamber. Since the process is a dry process, the coatings form on complicated patterns and three-dimensional shapes as thin, homogeneous, strongly adhesive films that have reproducible features. Therefore, the plasma process is verified to be a simple process for electrode fabrication. Furthermore, PPFs tend to be more biocompatible than other conventional materials. In fact, PPFs have been used as biomaterials, e.g., contact lenses.48 Sensor Characteristics. The sensor response is due to the following enzymatic reaction: GOx
β-D-glucose + O2 98 δ-gluconolactone + H2O2
(1)
The hydrogen peroxide generated in this reaction can be electrochemically oxidized at a Pt anode electrode. The electrode is polarized at +700 mV compared to the Ag/AgCl reference electrode.
H2O2 f 2H+ + O2 + 2e-
(2)
The catalytic ability of the Pt electrode is sensitive to the surface (48) Ho., C.-P.; Yasuda, H. J. Biomed. Mater. Res. 1988, 22, 919-937.
2674 Analytical Chemistry, Vol. 72, No. 11, June 1, 2000
roughness.30 The sputtered Pt electrodes used in the present case were obtained from the same batch in order to obtain reproducibility of sensor response. Figure 2 shows the steady-state amperometric response. The films of both devices were so thin that the distance between the electrode and the reaction center of enzyme was very close. As a result, a short response time (11.5 ( 0.8 s; glucose concentration, 5-20 mM)49 was obtained. Although a similar result for PPF-based devices was preliminarily reported by Kampfrath and Hintsche,50 our data first demonstrated that the noise level of baseline of device 1 (PPF-base) is 2-5 times smaller (glucose concentration, 5-20 mM) than that of device 2 (APTES-base). The detection limit (S/N ratio, 3) of device 1 and device 2 are 0.12 and 0.25 mM, respectively. The significant reasons are presented below. Due to the numerous pinhole (or pore) defects of APTES layer, the sensor response of device 2 is easily affected by artifacts, as compared to device 1; whereas PPFs are flawless and have a very highly cross-linking network.18 The immobilized layer of device 1 appears to presenting a diffusion barrier to the solution species, causing a partitioning between electrode and the bulk solution. The PPF also prevents interfering agents from adsorbing onto the surface and diminishing the convection (discussed below). Although based on Figure 2a, the response does not seems to be linear between 0 and 20 mM, the linearity (R) of this range is more than 0.98, which is an acceptable range of error (see Figure 3 and discussion below). This nonlinearity is because the kinetics of the enzymatic reaction obeys the Michaelis-Menten equation. For further improvement, the external diffusion control layer14,16,23 for glucose is needed. Figure 3 shows the calibration plot for glucose based on the CV current response at +700 mV compared to that of Ag/AgCl. On the basis of the data shown in Figure 3, the Eadie-Hofstee form51 of the Michaelis-Menten equation was employed in order to obtain the apparent Michaelis constants and (49) The response time is defined as the time needed to reach 95% of the maximum response. (50) Kampfrath, G.; Hinsche, R. Anal. Lett. 1989, 22, 2423-2431. (51) Hoshi, T.; Anzai, J.; Osa, T. Anal. Chem. 1995, 67, 770-774.
Table 1. Enzyme-Substrate Kinetic Parameters Estimated from Eadie-Hofstee Plot of Figure 3 KMapp/mM
devices 1 2
3.91 4.11
Table 2. Errors in Glucose Caused by Interfering Agentsa
Imax/µA cm-2
errorb (%) interference
81.2 81.9
0.2 mM ascorbic acid 0.3 mM ascorbic acid 0.1 mM acetoaminophen 0.2 mM acetoaminophen 0.2 mM urea 0.2 mM glycine 0.3 mM glycine
the maximum current,
I ) Imax - KMapp(I/C)
device
