Anal. Chem. 2004, 76, 5547-5551
Amperometric Glucose Biosensor Based on Assisted Ion Transfer through Gel-Supported Microinterfaces Carlos M. Pereira, Joaquim M. Oliveira,† Ricardo M. Silva, and Fernando Silva*
Departamento de Quı´mica, CIQ-L4, Faculdade de Cieˆ ncias da Universidade do Porto, R. do Campo Alegre, 687, 4169-007 Porto, Portugal
A novel amperometric glucose sensor was developed based on the facilitated proton transfer across microinterfaces between two immiscible electrolyte solutions. The combination of a 1,3:2,4-dibenzylidene sorbitol/2-nitrophenyl octyl ether gel membrane and 3-(2-pyridyl)-5,6diphenyl-1,2,4-triazine as the ionophore allows the transfer of protons from water to the gellified organic phase; the gel membrane is supported on arrays of microholes drilled on a polyester film. The protons are generated as the result of the dissociation of gluconic acid produced during the enzymatic degradation of glucose by glucose oxidase. The characteristics of the glucose sensor were investigated using several experimental conditions, namely, the concentration of ligand and enzyme. The electrochemical response is typical of an enzymatic electrode and displays a linear behavior in the range 0.2-3 mM glucose. The effect of the experimental parameters of the voltammetric technique was also optimized with the aimof improving sensor sensitivity. In the past decade, there has been a growing interest to use interfaces between two immiscible electrolyte solutions (ITIES) to study enzymatic reactions since ITIES are recognized as a simple model of a biological membrane, and therefore, the use of such electrochemical systems to build biosensor devices is a natural development of the research carried out in this field of electrochemistry. Furthermore, ion-transfer kinetics across the ITIES are considered to be fast1 even when amphiphilic molecules are adsorbed at the liquid-liquid interface,2-4 which could be an advantage over other methods where the electron transfer can be impaired by the adsorption of bulky molecules. The first work describing a biosensor application of an ITIES was reported by Osakai et al.5 The paper describes the develop* To whom correspondence should be addressed. E-mail:
[email protected]. Fax: +351 226082959. Phone: +351 226082913. † Present address: Laborato ´ rio 3B’s, Departamento de Engenharia de Polı´meros da Universidade do Minho, Campus de Gualtar, 4710-057 Braga, Portugal. (1) Murtoma¨ki, L.; Kontturi, K. J. Electroanal. Chem. 1998, 449, 225-229. (2) Allen, R. M.; Williams, D. E. Faraday Discuss. 1996, 104, 281-293. (3) Grandell, D.; Murtoma¨ki, L.; Kontturi, K.; Sundholm, G. J. Electroanal. Chem. 1999, 463, 242-247. (4) Kontturi, A.-K.; Kontturi, K.; Murtoma¨ki, L.; Quinn, B.; Cunnane, V. J. J. Electroanal. Chem. 1997, 424, 69-74. (5) Osakai, T.; Kakutani, T.; Senda M. Anal. Sci. 1988, 4, 529-530. 10.1021/ac0498765 CCC: $27.50 Published on Web 08/06/2004
© 2004 American Chemical Society
ment of a biosensor based on the ammonium produced during the urease decomposition of urea. The facilitated transfer of ammonium ion by dibenzo-18-crown-6 through the ITIES gives rise to an amperometric signal, which is used as the detection method. These authors have shown that the peak current obtained was proportional to the urea concentration in the test solution, and a linear response was reported in the range between 1 and 5000 µM. Later Osborne and Girault6 described a micro-ITIES (µ-ITIES) transducter for creatinine based on the same principle introduced by Osakai et al. The creatinine device showed a linear amperometric response for the range 20 µM-1 mM. These authors used microhole membranes to support the liquid-liquid interfaces. The use of microinterfaces aimed to enhance the rate of the mass transfer, in this way improving the electrode response. Microinterfaces, and in this particular case, microinterfaces supported at poly(ethylene terephthalate) films, present other advantages, namely, increasing the stability of the interface and reducing the effect of ohmic drop that could be a major problem when highly resistive organic solvents are used. Although a diversity of methods can be found in the literature,7-12 the applications of ITIES in the construction of sensing devices,13,14 and particularly biosensing5,6 devices, have proved to be an important development for electrochemical detection. Due to their relatively low applied potentials, ITIES-based biosensors could avoid some interferences from readily oxidizable species, as described by Zhang and Wang,15 thus, creating an alternative method to the electrochemical detection of peroxide produced during the enzymatic cycle.16-18 (6) Osborne, M. D.; Girault, H. H. Mikrochim. Acta 1995, 117, 175-185. (7) Clark, L. C.; Lyons, C. Ann. N. Y. Acad. Sci. 1962, 102, 29-45. (8) Senda, M. Ann. N. Y. Acad. Sci. 1990, 613, 79-94. (9) 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. Anal. Chem. 1984, 56, 667-671. (10) Albery, W. J.; Bartlett, P. N.; Craston, D. H. J. Electroanal. Chem. 1985, 194, 223-235. (11) Foulds, N. C.; Lowe, C. R. Anal. Chem. 1988, 60, 2473-2478. (12) Mehrvar, M.; Bis, C.; Scharer, J. M.; Moo-Young, M.; Luong, J. H. Anal. Sci. 2000, 16, 677-692. (13) Osakai, T.; Kakutani, T.; Senda, M. Anal. Sci. 1987, 3, 521-526. (14) Lee, H. J.; Pereira, C. M.; Silva, F.; Girault, H. H. Anal. Chem. 2000, 72, 5562-5566. (15) Zhang, C.; Wang, K. Anal. Lett. 2002, 35, 869-880. (16) Trojanowicz, M.; Miernik, A. Electrochim. Acta 2001, 46, 1053-1061. (17) Rodrı´guez, M. C.; Rivas, G. A. Anal. Lett. 2000, 33, 2373-2389. (18) Sugawara, K.; Fukushi, H.; Hoshi, S.; Akatsuka, K. Anal. Sci. 2000, 16, 1139-1143.
Analytical Chemistry, Vol. 76, No. 18, September 15, 2004 5547
Figure 1. Schematic representation work principle of the glucose biosensor based on proton transfer through an aqueous-organic gel microinterface: (a) organic phase; (b) microhole membrane; (c) aqueous phase.
The use of ITIES to analyze glucose could at a first glance seem to be an impossible task since glucose is a neutral molecule and a detection system based on ITIES relies on the transfer of charged ions through the liquid-liquid interface. However, glucose oxidase (GOx) catalyzes the oxidation of β-D-glucose by oxygen to D-gluconolactone,19 which subsequently hydrolyzes spontaneously to gluconic acid; simultaneously, the reduction of oxygen to hydrogen peroxide occurs, as exemplified in Figure 1. This model is consistent with the “ping-pong” mechanism described in the literature, where the β-D-glucose oxidation induces enzyme reduction, while the natural acceptor O2 acts as a twoelectron oxidant,20 This is followed by the proton transfer through the liquid-gel interface in a process similar to that described by Osakai et al.5 and more recently by Pereira et al.,21 who reported a Zn(II) amperometric sensor based on the assisted ion transfer through gellified µITIES. In this paper, an amperometric glucose sensor, based in a gellified µITIES is described. We will demonstrate that the gluconic acid, generated during enzymatic cycle can be followed by cyclic voltammetry. The voltammetric signal is a consequence of ionophore-assisted transfer of protons across the aqueous 2-nitrophenyl octyl ether (o-NPOE) gellified interface. This work will open the way to new applications for the ITIES-based biosensor. EXPERIMENTAL SECTION Chemicals. The aqueous- and organic-phase solvents were respectively ultrapure water (Milli-Q, Millipore purification system) (r )18 MΩ‚cm) and o-NPOE (Fluka, Selectophore). The organic solvent was used without any further purification; however, the bottle was always kept in the refrigerator. All solvent manipulations were carried out in a glovebox under a nitrogen atmosphere. The supporting electrolytes used were 10 mM NaCl (Merck, p.a. grade) in the aqueous phase and 10 mM tetraoctylammonium (19) Su, Q.; Klinmann, J. P. Biochemistry 1999, 38, 8572-8581. (20) Pierce, D. T.; Unwin, P. R.; Bard, A. J. Anal. Chem. 1992, 64, 1795-1804. (21) Pereira, C. M.; Tirilly, N.; Martins, M. C.; Silva F. Fresenius J. Anal. Chem. 2001, 369, 609-612.
5548
Analytical Chemistry, Vol. 76, No. 18, September 15, 2004
tetraphenylborate (TOATPB) in the organic phase; the latter was prepared as described elsewhere.22 The organic gellifying agent was 1,3:2,4-dibenzylidene sorbitol (Milliken) with and agar (DIFCO) as aqueous gellifying agent. The complexing agent used in this work was 3-(2-pyridyl)-5,6diphenyl-1,2,4-triazine (PDT; Aldrich). D-(+)-Glucose monohydrate, GOx (EC 1.1.3.4; 23 and 200 units/mg from Aspergillus niger) were supplied by Fluka. GOx solutions were prepared by dissolving analytical-grade reagent in ultrapure water. All glassware was cleaned by rinsing first with acetone (Riedel-de Hae¨n and Pronalab) and ultrapure water thoroughly. Microarrays. The microhole array was prepared by UV excimer laser photoablation of 12-mm-thick Melinex (S grade) film at the Laboratoire d’Electrochimie (EPFL, Lausanne, France). The detailed procedure was given by Seddon et al.23 In this work, the array is made of 66 holes each of 10-µm diameter. The film was fixed either to a glass cylinder or to a Teflon cylinder using a solvent-resistant fluorosilicone RTV 730 sealant (Dow Corning Corp.). Gel Preparation. The details of preparation and properties of the gellified organic solution have been given by Silva et al.24 The gellifying agent (1.8% w/w) and the macrocyclic ligand were added to a 10 mM solution of TOATPB in o-NPOE. The mixture was heated at 150 °C, with stirring until the solution became clear. The hot gel was then applied on the microhole array film and was left overnight to set. The aqueous reference gel for the organic phase was prepared by adding agar-agar (2% w/w) to a 10 mM NaTPB aqueous solution and 10 mM NaCl and heating the mixture at 90 °C with stirring. Electrochemical Measurements. The transfer of H+ across the o-NPOE/water (W) interface assisted by PDT present in the o-NPOE phase was followed by cyclic voltammetry. Electrochemical measurements were carried using a SI 1287 electrochemical interface (Solatron Instruments) and a EG&G potentiostat model 273 in a two-electrode mode. The schematic representation of the electrochemical cell used throughout this work can be described as follows:
More detailed information on the cell geometry can be found in the literature.24 The polarized o-NPOE/W interface is indentified by the asterisk *. Cyclic voltammograms (CVs) were recorded with a scan rate of 0.050 V/s unless otherwise specified. After each addition of glucose solution, the cell was homogenized with a magnetic stirrer for 30 s. The peak current values were taken after stabilization of the CVs, which usually required five cycles (∼180 s). The peak current is then corrected for the baseline and plotted as a function of glucose concentration. (22) Pereira, C. M.; Schmickler, W.; Silva, F.; Sousa, M. D. J. Electroanal. Chem. 1997, 436, 9-15. (23) Seddon, B. J.; Shao, Y.; Fost, J.; Girault, H. H. Electrochim. Acta 1994, 39, 783-791. (24) Silva, F.; Sousa, M. D.; Pereira, C. M. Electrochim. Acta 1997, 42, 30953103.
Figure 3. Sensor response for the addition of D-(+)-glucose: (a) at different GOx concentrations and 5 mM PDT; (b) at different PDT concentrations and 10 nM GOx.
Figure 2. Representation of typical voltammograms for microinterfaces recorded (a) in the absence of PDT and with GOx 10 nM, (b) in the absence of GOx and with 20 mM PDT, and (c) with 10 nM GOx in the aqueous phase and 20 mM PDT in the organic gel.
A Crison-MicropH 2000 was used for the pH measurements. Temperature control of the electrochemical cell was achieved using a thermostated water circulating bath, type SU5, from Grant Instruments. RESULTS AND DISCUSSION As mentioned before, two major factors influence the response of a sensor, the biocatalyst and the transducer. To establish the use of an ITIES as a transducer for the glucose biosensor, the effect of the different components over the sensor response was tested. Figure 2a) shows the voltammograms recorded when the ionophore (PDT) is absent in the organic gel; although GOx is present in the aqueous solution, the increment of glucose concentration does not change the voltammetric response despite
the small changes in the solution viscosity.25 The voltammograms shown in Figure 2b demonstrate that in the absence of GOx in the aqueous phase successive additions of glucose to the solution do not cause any noticeable change on the voltammetric response. Only when both PDT and GOx are simultaneously present is an increase in current with glucose concentration observed at more positive potentials (Figure 2c). This increase in current corresponds to the assisted transfer of H+ by PDT from the aqueous to the gellified organic phase. The peak current (measured at -0.050 V as represented in Figure 2c by the dotted line), after subtraction from the baseline current, was used for the evaluation of the amperometric behavior of the sensor. The same feature of current increase is observed when a hydrochloric acid solution is added do the aqueous phase and PDT is present in the organic phase. Further characterization of the ITIES-based glucose biosensor was carried out by assessing the effect of the concentration of GOx and ligand as well as the effect of the temperature. Panels a and b of Figure 3 display the influence, respectively, of the GOx and PDT concentrations on the amperometric signal of the glucose sensor. Depending on the concentration of the enzyme, two contrasting behaviors are observed: (1) For lower concentration of enzyme (1, 10 nM), there is an increase of the amperometric signal with the amount of GOx that (25) Shao, Y.; Girault, H. H. J. Electroanal. Chem. 1990, 282, 59-72.
Analytical Chemistry, Vol. 76, No. 18, September 15, 2004
5549
Figure 5. Temperature effect on current intensity as a function of D-(+)-glucose concentration. Figure 4. Calibration plots of GOx behavior for different activities. Measurements carried out at -0.010 V, and I represents the difference between the peak current and the baseline (recorded at aqueous phase 10 mM NaCl and 10 nM GOx).
can be related to an increase in the amount of gluconic acid produced. (2) For higher concentrations of GOx (0.5, 10 µM), the sensor performance is reduced with a reduction, and loss, of linearity. This effect could be related, possibly, with the formation of micelles or with the formation of enzyme bylayers. This suggestion is supported by the work of Georganopoulou et al.,26 who reported evidence of the formation of bilayers at the liquid-liquid interface when the enzyme concentration is greater than 1 µM. To work with a well-defined interface, the remaining experiments were carried out with 10 nM GOx solution. The results in Figure 3b show that there is no marked effect of the ligand concentration on the voltammetric behavior of the ITIES-based glucose biosensor. Although the changes are negligible, we decided to use a 5 mM solution of PDT since under these conditions better defined voltammograms are obtained. The experimental results in Figure 4 compare the behavior observed for two commercial enzymes showing that the use of an enzyme with higher activity improves the linearity of the amperometric response for the lower ranges of glucose concentration. However, the maximum current obtained is not affected by the nature of the enzyme because the changes observed are within the experimental error. An interesting result displayed in Figure 4 concerns the fact that the standard deviation of the experimental results is higher for higher concentrations of glucose in solution. In this situation, the standard deviation of four experiments is ∼12%. An interesting conclusion that can be drawn from the analysis of the data in Figure 4 is that, despite lower proton concentration generated from low concentrations of glucose in solution, the standard deviation for those results is smaller than 5% and much smaller than the standard deviation measured for higher glucose concentration levels. (26) Georganopoulou, D. G.; Williams, D. E.; Pereira, C. M.; Silva, F.; Su, T.; Lu, J. R. Langmuir 2003, 19, 4977-4984.
5550 Analytical Chemistry, Vol. 76, No. 18, September 15, 2004
Table 1. K app M Values for Different Temperatures Obtained by Nonlinear Curve Fit of the Hill-Type Form of the Michaelis-Menten Equation temperature/°C
Imax/nA
KM/mM
10 15 25 35 40
14.1 ( 0.8 17.9 ( 1.4 25 ( 2 33 ( 1 46 ( 4
2.5 ( 0.4 2.1 ( 0.5 3.0 ( 0.6 2.6 ( 0.2 4.2 ( 0.9
The possible influence of slow adsorption of GOx on the amperometric response was assessed by performing measurements after allowing different times of contact. The measurements taken after 30 min, 1, 2, and 17 h of adsorption time with a 10 nM GOx aqueous solution do not show any relevant change in the amperometric response, and therefore, the adsorption time of the GOx prior to use with the sensor was established to be 30 min. The effect of temperature has been long recognized as being a significant factor in the performance of enzyme-based devices. In particular, it may lead to enzyme denaturation, damaging it in an irreversible way, as reported by Georganopoulou et al.26 To evaluate the effect of temperature on the performance of the ITIESbased biosensor, several experiments were carried on in the temperature range between 10 and 40 °C. The results show that an increase in temperature does not cause any alteration in the shape of the voltammograms; however, an increase in the peak current is observed as shown in Figure 5. The data from Figure 5 may also be analyzed in order to extract the apparent Michaelis-Menten constants for the enzymatic process.27 Table 1 shows the KM and Imax values at different temperatures. These values are significantly different from the reported value of KM for the free enzyme solution (6.2 mM)28 and from values found in the literature.18,27,29 The differences between the measured KM and the value of the free enzyme could be explained by the effect of the immobilization process on the sensor (27) Lowry, J. P.; McAteer, K.; El Atrash, S. S.; Duff, A.; O’Neill, R. D. Anal. Chem. 1994, 66, 1754-1761. (28) Cosnier, S.; Gondran, C.; Pellec, A. L.; Senillou, A. Anal. Lett. 2001, 34, 61-70. (29) Gregg, B. A.; Heller, A. Anal. Chem. 1990, 62, 258-263.
Figure 6. Comparison between the potentiometric and the amperometric response.
performance as reported by Pierce et al.,20 the characteristics of the adsorbed layer,30 or the effect of the pH on KM. Another aspect that should be retained is the increase of the maximum current with the temperature, which demonstrates the large temperature range stability of the sensor.9 Figure 6 compares the data obtained using an amperometric and a potentiometric sensor as a function of glucose concentration. The amperometric measurements were carried out after the potentiometric measurements, in the same cell, just allowing the time to remove the glass electrode in order to reduce the noise during the voltammetric measurements. As can be observed, the steady-state current, has a linear region for lower concentrations of glucose (
