Biosensors and Bioelectronics 22 (2007) 1317–1322
Enzyme electrodes based on sono-gel containing ferrocenyl compounds Barbara Ballarin ∗ , Maria Cristina Cassani, Rita Mazzoni, Erika Scavetta, Domenica Tonelli Dipartimento di Chimica Fisica ed Inorganica, Universit`a di Bologna, V.le Risorgimento 4, 40136 Bologna, Italy Received 9 February 2006; received in revised form 25 May 2006; accepted 31 May 2006 Available online 18 July 2006
Abstract An amperometric-mediated glucose sensor has been developed by employing a silica sono-gel carbon composite electrode (SCC). The chosen mediators, ferrocene (Fc) and 1,2-diferrocenylethane (1), have been immobilized in the sono-gel composite matrix. The complex 1 has been employed for the first time as an electron transfer mediator for signal transduction from the active centre of the enzyme to the electrode conductive surface. After the optimisation of the construction procedure the best operative conditions for the analytical performance of the biosensor have been investigated in terms of pH, temperature and applied potential. Cyclic voltammetric and amperometric measurements have been used to study the response of both the glucose sensors, which exhibit a fast response and good reproducibility. The sensitivity to glucose is quite similar (6.7 ± 0.1 A/mM versus 5.3 ± 0.1 A/mM) when either Fc or 1 are used as mediators as are the detection limit ca. 1.0 mM (S/N = 3) and the range of linear response (up to 13.0 mM). However, the dynamic range for glucose determination results wider when using 1 (up to 25.0 mM). The apparent Michaelis–Menten constants, calculated from the reciprocal plot under steady state conditions, are 27.7 and 31.6 mM for SCC-Fc/GOx and SCC-1/GOx electrodes, respectively, in agreement with a slightly higher electrocatalytic efficiency for the mediator 1. © 2006 Elsevier B.V. All rights reserved. Keywords: Ferrocene; 1,2-Diferrocenylethane; Sono-gel; Composite electrode; Biosensor; Glucose oxidase
1. Introduction The use of composite materials based on conductive phases dispersed within polymeric matrices has led to important advances in the development of sensor devices. In such systems two or more materials are combined, each of them retaining its original nature, while the composite shows distinctive chemical, mechanical and physical behavior (Ruschau et al., 1989). Among the large number of matrices proposed in the literature, sol–gel materials offer an alternative route for processing materials at low temperature and into a large variety of structures (Binker and Schrerer, 1989). Recently, we have developed a new class of sono-gel carbon composite (SCC) electrodes obtained by sonocatalysis in which high-energy ultrasounds (HEU) were directly applied to precursors to promote hydrolysis in acid medium without the addition of any solvent (Blanco et al., 1999; Cordero-Rando et al., 2002; Ballarin et al., 2003). These SCCs exhibit higher density, reduced reticulation, more homogeneous structure and
∗
Tel.: +39 051 209 3704; fax: +39 051 2093690. E-mail address:
[email protected] (B. Ballarin).
0956-5663/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2006.05.034
minimized volume contraction in comparison with the classic sol–gel composites (Hidalgo-Hidalgo-de-Cisneros et al., 2001). All these properties are fundamental to provide high electrical conductivity and reduced capacitive currents in the electrochemical materials. Moreover, the sono-gel composites are suitable to be modified with the addition of redox-active systems to produce bioelectrochemical sensors, e.g. glucose enzyme electrodes that function with glucose oxidase (GOx). Among the different mediators described in the literature (Mahenc and Aussaresses, 1979; Ikeda et al., 1984; Crumbliss et al., 1986; Zhang et al., 2000), ferrocene (Fc) and its derivatives, first reported by Cass et al. (1984), have proved to be the most efficient electron transfers for the GOx enzymatic reaction. In fact, they are generally small enough to come into proximity of the active site of GOx (Alvarez-Icaza et al., 1995; Forrow et al., 2002; Forrow and Walters, 2004) providing an excellent overlap of the -orbitals of the cycloplentadienyl ring with those of the enzyme prosthetic group. Considering the foregoing, we wondered whether waterinsoluble linked ferrocenes (immobilized in SCC as stabilizing matrix) could yield improved performances as electron transfer mediators to oxidoreductase enzymes, starting with GOx. With this objective in mind, our choice fell on ferrocene
1318
B. Ballarin et al. / Biosensors and Bioelectronics 22 (2007) 1317–1322
and on one of the simplest biferrocene systems, namely 1,2diferrocenylethane (1), which has been never studied before as enzymatic mediator. Note that the presence of a bridging –(CH2 )2 – group in 1 avoids metal–metal interactions. In fact, it has been shown that for molecules containing n identical non-interacting centres (sufficiently spaced) the current–potential responses should have the same shape as those obtained for the corresponding molecule with just one centre. Only, the magnitude of the current response will be enhanced accordingly (Flanagan et al., 1978). The biferrocene 1 is characterised by low steric hindrance on the cyclopentadienyl ring and by a reduction potential lower than Fc in CH2 Cl2 (+0.35 V instead of +0.44 V versus SCE) (Ferguson et al., 1996). Moreover, although the redox behavior of 1 in organic solvents such as CH2 Cl2 , CH3 CN and DMF has been described in detail (Morrison et al., 1973; Bildstein et al., 1995; Ferguson et al., 1996; Nowicka et al., 2004) no data are available in aqueous media. In this paper we report on the electrochemical and catalytical behavior of SCC-1 electrodes compared to the SCC-Fc ones.
HCl (catalyst) solution were mixed in a glass vessel and submitted to high-energy ultrasound (HEU) for 10 s until hydrolysis started (Blanco et al., 1999; Cordero-Rando et al., 2002; Ballarin et al., 2003). MTMOS was used to obtain hydrophobic electrodes (Gun et al., 1994) so that only the outer section of the electrode was active. In particular, 1 g of graphite powder or 1 g of a powder consisting of graphite and equal percentages of 1 (20%, w/w; 0.50 mmol) or ferrocene (20%, w/w; 1.07 mmol) was added to the sono-gel mixture, so that in both electrodes the same molar amount of Fe is present. In all cases the resulting paste was homogenized for 1 min. The sticky black pastes obtained (named: SCC, SCC-1, and SCC-Fc, respectively) were used to fill in glass tubes (70 mm length, 5 mm diameter) up to a height of about 3 mm and finally allowed to dry overnight at ambient conditions (25 ◦ C). The electrical contact was assured by a 0.5 mm diameter copper wire. The electrodes were first polished with 1200 P grit emery paper (silicon carbide), then gently with weighing paper and finally were washed with Milli-Q water. The unmodified SCC electrode, obtained by adding only graphite, was used as a blank.
2. Experimental
2.3. Preparation of the biosensors
2.1. Chemicals
Forty milligrams of bovine serum albumin (BSA) and 10 mg GOx were dissolved in 1 mL of KPB, pH 7.0 (solution A). A 2.5% aqueous glutaraldehyde (GA) solution was prepared (solution B). Fifteen microliters of solution A and 5 L of solution B were mixed to obtain solution C. Then, 15 L of solution C were dipped on the electrode surface of the appropriate SCC-modified electrode and allowed to dry at room temperature for 90 min. The films obtained were homogeneous and well adherent to the electrode surface. After the preparation, the electrodes were stored at 4 ◦ C, soaked into the KPBS, pH 7.0.
Glucose oxidase (E.C. 1.1.3.4) from Aspergillus niger, (50,000 U/mg), ␣-d(+)glucose, glutaraldehyde (GA) 25% (v/v), and bovine serum albumin (BSA) were obtained from SigmaAldrich. Ferrocene was purchased from Aldrich, and K2 HPO4 , and KH2 PO4 , from Fluka. All the reagents were of analytical grade and were used without further purification. The biferrocene 1,2-diferrocenylethane was prepared as described in the literature (Rinehart et al., 1959); methyl trimethoxysilane (hereinafter abbreviated MTMOS, Merck) and graphite powder (spectroscopic grade RBW, SGL Carbon Ringsdorff, Germany) were used as purchased. All potassium–phosphate buffers (KPB) were prepared with water obtained from a Milli-Q system (Millipore). The derivative 1 was characterised by elemental analysis (ThermoQuest Flash 1112 Series EA Instrument) and spectroscopic methods. The NMR spectra were recorded using Varian MercuryPlus VX 400 (1 H, 399.9; 13 C, 100.6 MHz) instrument, referenced internally to residual solvent resonances, and recorded at 298 K for characterisation purposes. ESI-MS analysis were performed by direct injection of methanol solutions of the metal complex using a WATERS ZQ 4000 mass spectrometer. 1 H NMR (CDCl3 ) of 1: δ = 4.270 (s, 4H; CH2 ), 4.226 (t, 3J 3 H,H = 1.6 Hz, 4H; C5 H4 ) 4.141 (t, JH,H = 1.6 Hz, 4H; C5 H4 ), 13 1 4.117 (s, 10H; C5 H5 ); C-{ H} NMR (CDCl3 ): δ = 83.7 (Cq ), 69.3 (CH; C5 H4 ), 68.4 (CH; C5 H5 ), 68.3 (CH; C5 H4 ), 67.9 (CH2 ). ESI-MS: m/z (%): 199 (100) [(C10 H9 FeCH2 )2 ]2+ . 2.2. Preparation of ferrocenes SCC electrodes The procedure used for the preparation of the SCC matrix was the following: 0.5 mL of MTMOS and 0.1 mL of 0.2 M
2.4. Electrochemical measurements A three-electrode cell configuration was used for all the electrochemical measurements. A Pt auxiliary electrode and a saturated calomel electrode (SCE) were employed; all the potentials presented are quoted with respect to SCE. Unless otherwise stated, electrochemical measurements were carried out in a deoxygenated 0.1 M KPBS with a multimode electrochemical system (Autolab PGSTAT 20, EcoChemie, Utrecht, The Netherlands) controlled by a personal computer, via GPES software, at room temperature. Oxygen removal from the working solutions was achieved by purging with nitrogen. Chronoamperometric measurements were carried out in a stirring solution, at a constant rate of 300 rpm. The stirring rate of 300 rpm is fast enough for a reasonable equilibration time but not so high to generate air bubbles or noise problems (Niu and Lee, 2002). After about 100 s, a time sufficient to obtain a stable background current, different aliquots (25, 50 or 100 L) of 0.1 M glucose solution were successively added to 5 mL of KPBS (0.1 M at pH 5.5 or 7.0). The steady state current response was measured.
B. Ballarin et al. / Biosensors and Bioelectronics 22 (2007) 1317–1322
1319
Fig. 1. Optical microscopy image of SCC-1 electrode surface before the deposition of the enzymatic film. Magnification 100×.
3. Results and discussion 3.1. Surface morphology Surface morphology was evaluated by optical microscopy (Reichert-Jung MeF3A with a JVC-TK-C1381 color camera) before and after the deposition of the enzymatic film. Fig. 1 reports, as example, the image of a SCC-1 surface, from which the three different components of the matrix (silica cluster, graphite and 1,2-diferrocenylethane particles) are well evident. After the deposition of the enzymatic film (figure not reported) a relatively homogeneous coating is observable with the absence of cracks on the whole surface. From the microscopic image, the thickness of the enzymatic film was estimated to be about 2 m. 3.2. Electrochemistry of SCC-Fc and SCC-1 The electrochemical properties of SCC-modified electrodes were studied by cyclic voltammetry (CV) in KPB solutions, pH 5.5 and 7.0. The CV of SCC-Fc with different scan rates at pH 5.5 is reported in Fig. 2A. At low scan rate (5 mV s−1 ) the CV exhibits an anodic peak at 234 mV in the forward scan of potential, related to the oxidation of Fc to Fc+ , whereas in the reverse scan of potential, a cathodic peak appears at 142 mV related to the reduction of Fc+ (Ep = 92 mV; E◦ = 188 mV, calculated as the half-sum of the cathodic and anodic peak potentials). A stable signal is obtained after three successive scans. The oxidation and reduction of Fc/Fc+ redox couple presents a quasireversible behavior because the peak separation potential is greater than the 59/n mV value, expected for a reversible system. Analogously, the SCC-1 electrodes present an anodic and cathodic well distinct peaks system at 456 and 269 mV, respectively (Ep = 187 mV; E◦ = 362.5 mV) as shown in Fig. 2B. Also in this case a stable CV response was registered after three or four continuous cycles. The higher Ep values observed in high and low scan rates compared to those reported by Morrison et al. (1973) in acetonitrile solution, could be imputed to a slower electron transfer process that occurs inside the composite matrix.
Fig. 2. Cyclic voltammograms in 0.1 M KPBS, pH 5.5 at different scan rates: a, b, c and d (5, 10, 20 and 50 mV s−1 , respectively) of: (2A) SCC-Fc and (2B) SCC-1.
At a scan rate greater than 20 mV s−1 only one oxidation wave was detected, corresponding to a two-electron process, with SCC-1 electrode, i.e. the iron atoms are both oxidized at the same potential; this data are in good agreement with that observed by Morrison et al. (1973). The oxidation of one iron metal center does not affect the process involving the other center, even if the compound is present in a composite system. At lower scan rate (i.e. 5 mV s−1 ), in the anodic region, a shoulder became visible at lower potential (ca. 384 mV); in any case two different processes are not discernible. The cyclic voltammograms of blank SCC electrodes (not reported) recorded in the same solution, did not show any redox reaction peak. The plots of cathodic and anodic peak currents as a function of potential scan rates (not reported) indicating that both SCC-Fc than -1 exhibit electrochemical responses, which are characteristic of redox species confined on the electrode surfaces. At higher scan rates (>25 mV s−1 ) the peak currents versus scan rate plots deviate from linearity and the peak current became proportional to the square root of potential scan rate, which indicated that the peak current is diffusion controlled (Salimi et al., 2004). The mediator surface concentration in SCC-1 and SCC-Fc was evaluated using the following equation: =
Q nfA
where Q is the charge obtained by integrating the anodic peak at low voltage scan rate and the other symbols have their usual meaning (n = 1 for Fc and n = 2 for 1). The calculate values are Fc = 9.0 × 10−9 mol cm−2 and 1 = 1.0 × 10−7 mol cm−2 .
1320
B. Ballarin et al. / Biosensors and Bioelectronics 22 (2007) 1317–1322
Fig. 3. Differential pulse voltammograms (DPV) of SCC-Fc and SCC-1 in 0.1 M KPBS at pH 7.0; V = 50 mV.
The differential pulse voltammetry, obtained on SCC-1 (see as example Fig. 3 (pH 7.0)) show one larger oxidation wave (w1/2 = 238 mV) if compared with that observed on SCC-Fc (w1/2 = 123 mV) but two waves are not discernible; moreover, it seems that pH variations (from 7.0 to 5.5) do not affect the redox reaction potential. 3.3. Electrochemical behavior of SCC-1 and SCC-Fc enzyme electrodes The addition of BSA in the enzymatic solution plays an important role in stabilizing the polymeric film, as reported by Yokoyama et al. (1996), and in enhancing the activity of the immobilized enzyme (Chen et al., 2002). Therefore, such a solution was employed to obtain SCC-Fc/GA–GOx–BSA (SCCFc/GOx) and SCC-1/GA–GOx–BSA (SCC-1/GOx) biosensors. The CVs obtained for the SCC-1/GOx and SCC-Fc/GOx electrodes in KPBS (pH 7.0), at 5 mV s−1 scan rate (not reported) showed stable and reproducible signals after 5 scans. Both electrochemical responses were characterised by an increased peak separation with respect to the signals recorded with the corresponding electrode when GOx film was absent. The effect of the potential scan rate on the electrocatalytic current of SCC-1/GOx towards glucose was studied by cyclic voltammetry. Different voltammograms were registered at various scan rates (5–100 mV s−1 ) in a KPBS, pH 7.0, containing 3.4 mM glucose. The catalytic effect of the mediator became evident only at low scan rate, i.e. 5 mV s−1 , when the increase of the oxidation current is not associated to an increase of the reduction current. An analogous behavior was obtained with SCC-Fc/GOx electrode. This behavior is indicative of the enzyme-dependent catalytic reduction of the oxidized form of 1 or Fc and supports a kinetic limitation in the reaction between the redox-active site of GOx and mediator. 3.4. Chronoamperometric studies and stability of the sensor For the calibration of enzyme-modified electrodes, amperometric measurements were carried out in a degassed phosphate buffers (pH 7.0) stirred at the constant rate of 300 rpm,
Fig. 4. Current–time curve for a SCC-1/GOx sensor upon sequential addition of 2.5 mM glucose in pH 7.0 KPBS. The applied potential was +0.30 V; (inset) calibration plots relative to data 5A (), and to data obtained by a SCC-Fc/GOx sensor () in the same experimental condition.
by adding successive amounts of glucose while the electrode potential was kept at an opportunely fixed value (i.e.+0.30 V versus SCE). The chronoamperometric response curve registered at SCC/GA–GOx–BSA electrode, i.e. a blank response, at +0.30 V and pH 7.0 showed a very low increase of the current with the addition of glucose in solution (1.5 mM each addition) and a saturation current was obtained after the third addition (4.5 mM). A typical current versus time curve in a concentration range from 1.0 to 20.0 mM of glucose for a SCC-1/GOx electrode is reported in Fig. 4. The response time defined as the time necessary to reach 95% of the steady state current was typically less than 10 s. The i versus glucose concentration plot, reported in inset of Fig. 4, showed a linear range extended from 1.9 to about 13.0 mM (R = 0.995). Beyond this concentration, the calibration curve becomes non-linear; however, still at a concentration of 25.0 mM no saturation is evident (dynamic range). The detection limit (LOD) calculated as ylod = yb + 3σ b (using the value of the calculated intercept, as an estimate of the blank signal (yb ) and Sy/x , standard deviation of the regression, in place of σ b , blank standard deviation) (Miller and Miller, 1988), was 1.1 mM. The sensitivity, expressed as the slope of the linear region of the calibration curves resulted 6.7 ± 0.1 A/mM. Analogously, in the inset of Fig. 4 is also reported the calibration curve for a SCC-Fc/GOx. A linear plot is observable in the range 1.9 to about 13.0 mM (R = 0.998), with a detection limit and a sensitivity of 1.3 mM and 5.3 ± 0.1 A/mM, respectively. In this case, the dynamic range result up to 15.0 mM.
B. Ballarin et al. / Biosensors and Bioelectronics 22 (2007) 1317–1322
The repeatability of the sensor current response was investigated testing four glucose concentrations (1.9, 2.9, 3.9 and 4.7 mM) with a SCC-1/GOx electrode during a 18 h period and the relative standard deviation (R.S.D.) calculated on the slope of 10 repeated calibration plots resulted 5.1%. The fabrication reproducibility of the glucose sensors was evaluated by the current response of five SCC-1/GOx electrodes made independently from different batches and modified with enzyme as described in Section 2. Four glucose concentrations (1.9–4.7 mM) were tested for each electrode and the RDS, calculated on the slope of the calibration plots, resulted 6.6%. Different values were obtained for SCC-Fc/GOx electrode; the R.S.D. for repeatability and reproducibility measurements were calculated as 17.8% and 32.8%, respectively. The stability of both the glucose sensors was tested by amperometric measurements over a period of 1 month using 4 mM glucose in KPBS, pH 7.0. In-between the measurements, the enzyme sensors were stored in 0.1 M KPBS, pH 7.0, at 4 ◦ C to preserve the GOx activity (Pandey et al., 1999). The response for both the electrodes showed a loss of 20% of the initial value during the first 10 days and decreased to about 50% of its value in a period of 1 month. This result is very similar to that observed by Lev and others (Lev and Sampath, 1996). With the aim to optimise the stability, studies are still in progress. 3.5. Effect of pH, temperature and applied potential The enzymatic response of SCC-1/GOx and SCC-Fc/GOx sensors were studied in a range of pH 5.5–7.0 using phosphate buffers, pH values out of this range gives an irreversible denaturation of the enzyme. Both the electrodes exhibited the optimum response in terms of range of linearity and sensitivity at pH 7.0. The effect of temperature on the two sensors has been investigated between 23 and 80 ◦ C, at pH 7.0 phosphate buffers in the presence of 4 mM glucose. The results shows that the steady state current response increases with temperature, reaching a maximum at about 60 ◦ C. Thereafter, the response declines rapidly as a result of denaturation of the enzyme. Five different applied potentials (0.20, 0.25, 0.30, 0.35 and 0.40 V versus SCE) were investigated in the chronoamperometric studies in order to evaluate the value allowing the highest electrocatalytic response for SCC-1/GOx electrode. The operative conditions should also minimise the electrooxidation of possible interferents, i.e. moving the potential at lower positive values. All the tested potentials were chosen at a value in which the oxidation of FeII sites was involved but not still complete (before the anodic current maximum peak). The best calibration curves in terms of sensitivity and linearity was obtained at potential range of 0.30–0.35 V; a little change in slope was observable at a potential of 0.40 V (curve not reported). At 0.25 V a slight decrease in sensitivity and linearity range was visible. A dramatic change was evident with an applied potential of 0.20 V: at this potential, indeed, the redox process involving the FeII /FeIII couple only partially proceeds and there are not enough sites for an efficient catalytic reaction. As regards SCC-Fc/GOx electrode again a slight decrease in sensitivity and
1321
Fig. 5. Variation of the amperometric current at +0.30 V as a function of glucose concentration; (inset) Lineweaver–Burke type plot for the determination of app KM obtained from the data of figure. () SCC-Fc/GOx and () SCC-1/GOx electrode, respectively.
linearity range was obtained by changing the potential from 0.30 to 0.25 V. app
3.6. Measuring of catalytic efficiency, KM and Imax It is known that for a biosensor the ratio of the catalytic current ik (kinetically controlled current), to the diffusion controlled current id , (ik /id ) can be taken as measure of catalytic efficiency (Niu and Lee, 2002; Kane et al., 1998 and references cited therein). The ik /id evaluated in this study were 4.6 and 3.7 for SCC-Fc/GOx and SCC-1/GOx, respectively, suggesting that the electrocatalytic efficiency of the two mediators was almost the same, although slightly better for SCC-Fc/GOx. app The apparent Michaelis–Menten constants KM were calculated from the data of Fig. 5 ( = SCC-1/GOx; = SCCFc/GOx) and the Lineweaver–Burke plot (inset of Fig. 5) (Kamin and Wilson, 1980). It can be seen that fairly good linear relationships were obtained for both curves. From the interapp cepts on the abscissa of these plot the values of KM = 27.7 and 31.6 mM were estimated for SCC-Fc/GOx and SCC-1/GOx app electrodes, respectively. In this case, the lower KM value obtained for SCC-Fc/GOx shows a slightly higher catalytic efficiency of this sensor, in respect to SCC-1/GOx, in agreement with the results obtained for the ik /id ratios. 4. Conclusions In this work a new class of rigid conducting composites, based on a silica sono-gel–carbon matrix, has been used to incorporate
1322
B. Ballarin et al. / Biosensors and Bioelectronics 22 (2007) 1317–1322
the electron transfer mediators responsible for the re-oxidation of the GOx enzyme active centre. This system can be employed in the assembling of a new glucose biosensor. Moreover, we have proved the capability of 1 to shuttle electrons from biological redox compounds to the electrode surface, as an alternative to other ferrocene derivatives. The sensors’ lifetime is comparable to that of other glucose sensors reported in literature (Li et al., 1997 and references cited therein). Studies are still in progress to increase the sensor stability and to investigate the effect of the amount of mediator on the glucose response. The results obtained prompt us to further investigate the usefulness of 1 or analogues biferrocenes for the electrocatalytic determination of other analytes, with or without the presence of an enzymatic substrate. Acknowledgements The authors wish to thank the University of Bologna and the Ministero della Istruzione, Universit`a e Ricerca (MIUR), for financial support. References Alvarez-Icaza, M., Kalisz, H.M., Hecht, H.J., Aumann, K.-D., Schomburg, D., Schmid, R.D., 1995. Biosens. Bioelectron. 10, 735–742. Ballarin, B., Cordero-Rando, M.M., Blanco, E., Hidalgo-Hidalgo de Cisneros, J.L., Seeber, R., Tonelli, D., 2003. Collect. Czech. Chem. Commun. 68, 1420–1436. Bildstein, B., Denifl, P., Wurst, K., Andr`e, M., Baumgarten, M., Friedrich, J., Ellmerer-M¨uller, E., 1995. Organometallics 14, 4334–4342. Binker, C.J., Schrerer, G.W., 1989. Sol–Gel Science: The Physics and Chemistry of Sol–Gel Processing. Academic Press, San Diego, CA (Chapters 1 and 2). Blanco, E., Esquivias, L., Litr`an, R., Pinero, M., Ramirez-del-Solar, M., de laRosa Fox, N., 1999. Appl. Organomet. Chem. 13, 399–418. 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. Anal. Chem. 56, 667–671.
Chen, X., Hu, Y., Wilson, G.S., 2002. Biosens. Bioelectron. 17, 1005–1013. Cordero-Rando, M.M., Hidalgo-Hidalgo de Cisneros, J.L., Blanco, E., NaranjoRodriguez, I., 2002. Anal. Chem. 74, 2423–2427. Crumbliss, A.L., Hill, H.A.O., Page, D., 1986. J. Electroanal. Chem. 206 (1–2), 327–331. Ferguson, G., Glidewell, Ch., Opromolla, G., Zakaria, Ch.M., Zanello, P., 1996. J. Organomet. Chem. 517, 183–190. Flanagan, J.B., Margel, S., Bard, A.J., Anson, F.C., 1978. J. Am. Chem. Soc. 100, 4248–4253. Forrow, N.J., Sanghera, G.S., Walters, S.J., 2002. J. Chem. Soc., Dalton Trans., 3187–3194. Forrow, N.J., Walters, S.J., 2004. Biosens. Bioelectron. 19, 763–770. Gun, G., Tsionsky, M., Lev, O., 1994. Anal. Chim. Acta 294 (3), 261–270. Hidalgo-Hidalgo-de-Cisneros, J.L., Cordero-Rando, M.M., Naranjo-Rodriguez, I., Blanco-Ollero, E., Esquivias-Fedriani, L., 2001. ES Patent P200100556. Ikeda, T., Katasho, I., Kamei, M., Senda, M., 1984. Agric. Biol. Chem. 48, 1969–1976. Kane, S.A., Iwuoha, E.I., Smyth, M.R., 1998. Analyst 123, 2001–2006. Kamin, R.A., Wilson, G.S., 1980. Anal. Chem. 50, 1198–1205. Lev, O., Sampath, S., 1996. Anal. Chem. 68, 2015–2021. Li, J., Chia, L.S., Goh, N.K., Tan, S.N., Ge, H., 1997. Sens. Actuators B 40, 135–141. Mahenc, J., Aussaresses, H., 1979. C. R. Acad. Sci. Paris 289 (14), 357– 359. Miller, J.C., Miller, J.N., 1988. Statistics for Analytical Chemistry, 2nd ed. Ellis Horwood Limited, Publishers, Chichester, England (Chapter 5). Morrison, W.H., Krogsrud S.Jr., Hendrickson, D.N., 1973. Inorg. Chem. 12, 1998–2004. Nowicka, A.M., Stojek, Z., Ciszkowska, M., 2004. Anal. Lett. 37 (5), 979–994. Niu, J., Lee, J.Y., 2002. Sens. Actuators B 82, 250–258. Pandey, P.C., Upadhyay, S., Pathak, H.C., 1999. Sens. Actuators B 60, 83–89. Rinehart Jr., K.L., Michejda, C.J., Kittle, P.A., 1959. J. Am. Chem. Soc. 81, 3162–3163. Ruschau, G.R., Newnham, R.E., Runt, J., Smith, B.E., 1989. Sens. Actuators 20, 269–275. Salimi, A., Alizadeh, V., Hadadzadeh, H., 2004. Electroanalysis 16, 1984–1991. Yokoyama, K., Shibasaki, T., Murakami, Y., 1996. Denki Kagaku 64 (12), 1221–1227. Zhang, C., Haruyama, T., Kobatake, E., Aizawa, M., 2000. Anal. Chim. Acta 408, 225–232.
