Langmuir 2006, 22, 5837-5842

5837

Synthetic Nanocrystalline Diamond as a Third-Generation Biosensor Support Jorge Rubio-Retama,*,† Jorge Hernando,‡ Beatriz Lo´pez-Ruiz,§ Andreas Ha¨rtl,‡ Doris Steinmu¨ller,| Martin Stutzmann,‡ Enrique Lo´pez-Cabarcos,† and Jose´ Antonio Garrido‡ Departamento de Quı´mica-Fı´sica II, Facultad de Farmacia, UCM, Spain, Walter Schottky Institut, Technische UniVersita¨t Mu¨nchen, Germany, Departamento de Quı´mica Analı´tica, Facultad de Farmacia, UniVersidad Complutense de Madrid, Spain, and F-BeSt Coating GmbH, Austria ReceiVed January 17, 2006. In Final Form: April 14, 2006 Horseradish peroxidase (HRP) has been immobilized on the surface of functionalized nanocrystalline diamond (NCD) thin films. The structure of the modified NCD surface as well as the electrochemical behavior of the whole system was characterized by impedance spectroscopy and cyclic voltammetry. The proximity of HRP heme groups to the NCD surface allowed direct electron transfer between them, resulting in two separated one-electron-transfer peaks at 0.05 V and 0.29 V vs Ag/AgCl, corresponding to the cathodic and anodic process, respectively. The heterogeneous electron-transfer constant for both processes was calculated to be 0.066 s-1, the charge-transfer coefficient R ) 0.49, and the immobilized enzymatic layer about 2‚10-10 mol/cm2. The modified NCD electrode was used as a thirdgeneration biosensor for hydrogen peroxide determination showing a linear response in the 0.1-45 mM H2O2 range, at +0.05 V vs Ag/AgCl.

1. Introduction Diamond is an interesting material due to its physicochemical stability, large electrochemical potential window, and chemical sensitivity.1 These properties make diamond an excellent candidate for electrochemical applications, although its high cost has limited the use of this material. However, nowadays, the development of diamond growth by chemical vapor deposition (CVD) has enabled the preparation of large-area synthetic diamond wafers on different substrates at reasonable cost. Depending on growth parameters such as gas mixture, temperature, and substrate seeding, CVD growth of diamond can produce different kinds of films, which are generally classified according to the crystal grain size as polycrystalline (with a grain size about 1 µm), nanocrystalline (grain size about 100 nm), and ultrananocrystalline (UNCD) (grain size below 10 nm) diamond films. The electronic properties of UNCD and NCD films strongly depend on the presence of sp2-hybridized carbon atoms. Both UNCD2 and NCD3 films have been characterized as electrodes for electrochemical applications, proving that these electrodes exhibit a large electrochemical potential window and a low background current, even with a relatively high sp2 content in the films. These features, together with its inherent biocompatibility,4 make NCD an interesting candidate to prepare highly active electrodes for biosensors that could be used in “in vivo” applications. In this work, we have investigated the functionalization of NCD thin films with horseradish peroxidase. HRP reacts with * Corresponding author. E-mail: [email protected]. † Departamento de Quı´mica-Fı´sica II, Facultad de Farmacia, UCM. ‡ Walter Schottky Institut, Technische Universita ¨ t Mu¨nchen. § Departamento de Quı´mica Analı´tica, Facultad de Farmacia, Universidad Complutense de Madrid. | F-BeSt Coating GmbH. (1) Prado, C.; Flechsig, G. H.; Grundler, P.; Foord, J. S.; Marken, F.; Compton, R. G. Analyst 2002, 127, 329-332. (2) Yoshiyuki, S.; Małgorzata, A.; Witek, P. S.; Greg, M. S. Chem. Mater. 2003, 15, 879-888. (3) Chen, Q.; Gruen, D. M.; Krauss, A. R.; Corrigan, T. D.; Witek, M.; Swain, G. M. J. Electrochem. Soc. 2001, 148, E44-E51. (4) Carlisle, J. A. Nature Mater. 2004, 3, 668-669.

H2O2, producing two equivalents of an oxidized agent called compound I. This form contains a porphyrin radical and an oxyferryl heme group (Fe4+dO). The porphyrin radical can abstract an electron from the substrate, forming a new intermediate known as compound II, which generates the native enzyme after its oxyferryl heme reduction.5,6 When HRP is immobilized on an electrode surface, it can extract the electron directly from the electrode, regenerating the enzyme without any mediator. The electron transfer can be described as a tunneling process between the enzyme catalytic center and the electrode surface.7-9 The electron-transfer rate constant, kET, is given by eq 1,

kET ) k0 exp(-βd)

(1)

where d is the distance over which tunneling occurs, k0 is the electron-transfer rate constant when d ) 0, and β is a parameter that depends on the energy barrier and the nature of the medium. This phenomenon has been used to develop third-generation amperometric biosensors,8,9 which are more selective, by working at low potentials and faster because they do not need the diffusion of mediators to transform the enzymatic chemical signal into an electrical current. These features open the possibility to enhance the hydrogen peroxide determination that could be used to optimize a wide range of amperometric biosensors or clinical techniques such as ELISA,37 in which HRP plays an important roll. In this study, HRP (E.C.1.11.17) has been used as a model enzyme to be immobilized covalently onto a NCD electrode to prepare a biosensor.5,10 The NCD surface modification was (5) Hayashi, K.; Horiuchi, T.; Kurita, R.; Torimitsu, K.; Niwa, O. Biosens. Bioelectron. 2000, 15, 523-529. (6) Li, J.; Dong, S. J. Electroanal. Chem. 1997, 431, 19-22. (7) Bard, J. A.; Faulkner, L. R. Electrochemical Methods: Fundamentals and Applications, 2nd ed.; John Wiley & Sons: New York, 2001. (8) Li, J.; Guangjin, C.; Dong, S. J. Electroanal Chem. 1996, 416, 97-104. (9) Kuznetsov, B. A.; Shumakovich, G. P.; Koroleva, O. V.; Yaropolov, A. I. Biosens. Bioelectron. 2001, 16, 73-84. (10) Long, D. D.; Marx, K. A.; Zhou, T. J. Electroanal. Chem. 2001, 501, 107-113.

10.1021/la060167r CCC: $33.50 © 2006 American Chemical Society Published on Web 05/17/2006

5838 Langmuir, Vol. 22, No. 13, 2006

Rubio-Retama et al.

investigated by atomic force microscopy (AFM), X-ray photoelectron spectroscopy (XPS), and impedance spectroscopy. The enzymatic activity as well as the biosensor functionality was studied by using amperometric methods. 2. Experimental Section Nanocrystalline Diamond Films. Nanocrystalline diamond films were grown by F-BeSt Coating (Austria) on silicon substrates by hot-wire chemical vapor deposition from a CH4 (5-10%)/H2 (95-90%) mixture. The thickness of the diamond film was about 2 µm, with an average grain size of about 20 nm. The conductivity of the NCD films, resulting from the combined effect of impurities, such as nitrogen, and grain boundary conduction, was below 0.1 Ω-1 cm-1. On the basis of X-ray photoelectron spectroscopy (XPS), we estimate the contribution from sp2-hybridized orbitals of carbon atoms to less than 5%, keeping the rest sp3 hybridized. Instrumentation. AFM images were recorded using a Digital Instrument Multi-Mmode AFM operating in tapping mode with standard silicon tips. XPS spectra were recorded on a SPECS EA 10 spectrometer. The X-ray source was a SPECS RQ 20/38 with a twin Al/Mg anode. All electrochemical experiments were performed with a Parstat 2263 Perkin-Elmer potentiostat with a conventional three-electrode electrochemical cell by using hydrogenated nanocrystalline diamond (∼1 mm2) as the working electrode, a twisted platinum wire as the counter electrode, and Ag/AgCl saturated with KCl as the reference electrode. Electrochemical measurements were performed in 10 mM phosphate buffer (PBS) with 90 mM KCl at pH 6.2 and 25 °C. Impedance measurements were performed at a bias potential of 0.05V vs Ag/AgCl in the frequency range from 0.01 to 100 kHz. Chemicals. The cross-linker agent 2,2,2-trifluorine-N-9′-decenil acetamide (TFA) was synthesized by the Organic Chemistry and Biochemistry Department (Technische Universita¨t Mu¨nchen). HRP (E.C.1.11.17), succinic anhydride, 1-ethyl-3-(3dimethylaminopropyl)-carbodiimide (EDC), and N-hydroxy-sulfo-succinamide (NHS) were purchased from Sigma. The PBS solution was prepared by mixing 2 mM K2HPO4, 90 mM KCl, and 8 mM KH2PO4.

Figure 1. (A) AFM tapping mode image of an unmodified NCD electrode. (B) Shape analysis from a small section of NCD surface (red triangles).

3. Results and Discussion Surface Morphology of NCD. As illustrated in Figure 1, the surface of the diamond electrode as studied by AFM shows a very rough structure in which large clusters of grains are visible, as previously reported by Ha¨rtl et al.11 Figure 1b shows the surface roughness along the line drawn in Figure 1a: while within one cluster the roughness is below 2 nm, the superstructure formed by the clusters has a much larger roughness, on the order of 100 nm. Due to this characteristic surface morphology of NCD, the active area of the electrode is increased, which is recognized as a very important factor in enzyme-related kinetic phenomena12 NCD Biosensor Preparation. With the aim of preparing a NCD biosensor, HRP was anchored to the diamond surface by using TFA as linker. TFA attachment was performed via a photochemical method, which has been previously discussed in detail.11,13 The process was carried out by covering a small NCD piece (around 1 mm2) with a solution containing TFA and illuminating during 15 h with UV light at 254 nm under nitrogen atmosphere. As it has been recently reported,14 UV light (11) Haertl, A.; Schmich, E.; Garrido, J. A.; Hernando, J.; Catharino, S. C. R.; Walter, S.; Feulner, P.; Kromka, A.; Steinmu¨ller, D.; Stutzmann, M. Nature Mater. 2004, 3, 736-742. (12) Lau, C. H.; Grehan, K. J.; Compton, R. G.; Foord, F. S.; Marden, F. Diamond Relat. Mater. 2003, 12, 590-595. (13) Yang, W.; Auciello, O.; Butler, J. E.; Cai, W.; Carlisle, J. A.; Gerbi, J. E.; Gruen, D. M.; Knickerbocker, T.; Lasseter, T. L.; Russell, J. N.; Smith, L. M.; Hamers, R. J. Nature Mater. 2002, 1, 253-257. (14) Nichols, B. M.; Butler, J. E.; Russell, J. N.; Hamers, R. J. J. Phys. Chem. B 2005, 109, 8523-8532.

Figure 2. XPS spectra for the NCD modified surface showing the F1s peak as a function of illumination time.

illumination leads to the activation of the diamond surface, inducing an electrophilic attack to the TFA double bond, similar to the previously reported process on silicon by Cai et al.15 Subsequently, the NCD was thoroughly cleaned by sonication for 15 min in CCl4, followed by another 15 min in methanol. The attachment of TFA to the NCD surface was investigated by XPS. (15) Cai, W. J. Phys. Chem. B 2002, 206, 2656.

Nanocrystalline Diamond as a Biosensor Support

Langmuir, Vol. 22, No. 13, 2006 5839

Figure 3. Enzyme immobilization scheme on the NCD surface. (1) TFA attachment, (2) TFA-deprotection, (3) carboxyl group insertion, and (4) enzyme immobilization.

The trifluorine protective group of the TFA molecule can be detected by evaluating the XPS spectra in the fluorine F(1s) region. As can be seen in Figure 2, the intensity of the F1(s) peak at 688.5 eV increases with the illumination time, confirming the attachment of TFA to the NCD surface. No F(1s) signal was detected if the UV light was switched off during the incubation process. Before immobilizing the enzyme, the TFA monolayer was deprotected by immersing the NCD film in 25% tetramethylammonium hydroxide methanol solution for 1 h and subsequent rinsing with deionized water. This process releases the trifluorine acetate group, yielding an amine-terminated alkyl monolayer. With the aim of inserting a carboxyl group where the enzyme will be attached via an amide bond, 5 mL of succinic anhydride in acetone (0.2 g/mL) was added to a phosphate buffer solution at pH 6.2 where the NCD sample was immersed. During this latter process, the pH was kept around 6.0 by adding small amounts of KOH solution. The amine terminal group reacts with the succinic anhydride, yielding a terminal carboxyl group. After

Figure 4. Randle’s equivalent circuit.

this treatment, the sample surface was covered for 15 min with 400 µL of 200 mM EDC and 50 mM NHS. Both compounds produce the carboxyl group activation required to link the enzyme. Then, the sample was cleaned with water, dried with nitrogen gas, and covered with a HRP solution at 2 mg/mL for 12 h at 4 °C. Finally, the NCD electrode was rinsed with buffer to remove the noncovalently bonded enzyme and subsequently immersed

5840 Langmuir, Vol. 22, No. 13, 2006

Rubio-Retama et al.

Figure 5. Nyquist plot of: (A) unmodified NCD electrode, (B) NCD electrode covered by a TFA layer. Open and solid symbols correspond to experimental points and simulated values, respectively. The impedance measurements were carried out at 0.05 V vs Ag/AgCl in the frequency range 0.01-100 kHz.

in 1 M ethanolamine to saturate the remaining active carboxyl groups. Figure 3 schematically illustrates the modification of the NCD surface. The electrical contact to the NCD film was made by a gold wire contacting a gold metal pad on the NCD surface, which was evaporated prior to any chemical modification. The metal pad and the gold wire were covered with silicone glue in such a way that only the surface covered with HRP was exposed to the electrolyte when it was used as a working electrode. Characterization by Impedance Spectroscopy. Impedance spectroscopy is a powerful tool for studying the interface properties of surface-modified electrodes. An AC-modulated small signal voltage is applied to the electrochemical cell, and the AC current response is recorded as a function of the modulation frequency. The resulting complex impedance of the electrochemical cell can be represented by the simple equivalent circuit shown in Figure 4 The different elements of the circuit correspond to the resistance of the solution (RS), a capacitive element (CD), which in some cases can be described as a double layer capacitance, and two other elements related to the charge-transfer processes. RCT, the electron-transfer resistance, dominates in the case of kinetic limited processes. The Warburg impedance (ZW), a frequency dependent term, represents the resistance to the mass transport of electroactive species, dominating in the case of diffusionlimited kinetics. The Nyquist diagram of the complex impedance represents the imaginary versus the real part of the impedance. In general, for the equivalent circuit of Figure 4, the Nyquist plot will show a semicircle and a linear region. The semicircle at higher frequencies corresponds to an electron-transfer-limited process, and the linear portion at lower frequencies corresponds to the diffusion-limited process. We have measured the nonfaradaic impedance of the NCD electrode when no redox species was present in the electrochemical cell. In this situation, any contribution from the Warburg impedance can be neglected. The corresponding plot is shown in Figure 5A. The experimental data (open symbols in Figure 5A) have been fitted with a model (simulation data as solid symbols in Figure 5) based on the equivalent circuit of Figure 4, assuming no contribution from the Warburg impedance. The value of the electron-transfer resistance for the case of the unmodified NCD electrode is RCT ≈ 5500 Ω. After modification of the electrode surface with the TFA molecules, RCT shows a conspicuous increase (RCT ≈ 59 000 Ω) as can be seen in Figure 5B. The resistance increment indicates that the organic layer acts as an insulating barrier that hinders the current flow to the electrode.

Figure 6. Nyquist plot of HRP-modified NCD biosensor. Open and solid symbols correspond to experimental points and simulated values, respectively. AC amplitude of 0.05 V; frequency range 0.01-100 kHz.

Figure 6 shows the Nyquist plot of the protein-covered electrode. In the high-frequency region, a new contribution to the total impedance is visible. In the case of the protein-modified electrode, the simple equivalent circuit of Figure 4 cannot describe the complex situation, which is schematically shown in Figure 7. The presence of the enzyme will modify the properties of the molecules beneath, adding new capacitive terms. Remarkable in Figure 6 is that the protein modification introduces a component with very low resistance to electron-transfer: the RCT of the high-frequency component is about 400 Ω. As discussed below, this low resistive path results from the direct charge transfer between the enzyme’s redox group and the electrode surface. Characterization by Cyclic Voltammetry. Cyclic voltammetry was applied to study the electron transfer between the enzyme heme group and the modified electrode surface. The voltammogram depicted in Figure 8A shows no redox process for the unmodified NCD electrode. In contrast, two redox waves can be seen on the HRP-NCD electrode (Figure 8B), with oxidation and reduction potentials at 0.29 and 0.05 V vs Ag/AgCl, respectively. The redox peaks positions were calculated by convolutional and semidifferential electroanalysis.33,34 Similar redox peaks, reported for HRP immobilized with mercaptopropionic acid on gold electrodes6 and, more recently, on platinum electrodes,16 have been attributed (16) Yan, L.; Rou, Y.; Yaqin, C.; Dianping, T.; Jianyuan, D.; Xia, Z. Sens. Actuators, B 2005, in press.

Nanocrystalline Diamond as a Biosensor Support

Langmuir, Vol. 22, No. 13, 2006 5841

Figure 9. Cyclic voltammograms of HRP-modified NCD electrode in 50 mM pH 6.2 PBS at different scan rates. The inset shows the variation of the current intensity with the scan rates.

Figure 7. Schematic model of the two-layer structure of the modified NCD surface, consisting of an enzymatic layer anchored to an organic layer.

Figure 10. Variation of peak potentials versus the logarithm of the scan rates.

Figure 8. Cyclic voltammogram of (a) the unmodified electrode and (b) a modified electrode with immobilized HRP in 50 mM PBS buffer at pH 6.2. Scan rate 20 mV/s.

to reversible oxidation and reduction (Fe2+ S Fe3+) of the HRP heme group. Figure 9 illustrates the cyclic voltammetry of the enzymemodified NCD electrode at different scan rates from 5 to 60 mV/s. The anodic and cathodic peaks show a shift in the peakto-peak separation when the scan rate increases. Furthermore, as shown in the inset of Figure 9, the absolute value of the current intensity increases linearly with the scan rate for the anodic and cathodic peaks, indicating a surface-controlled electrode process. From the area of the cathodic peak, the total amount of charge passing through the electrode has been calculated to be Q ≈ 8‚10-7 C. From Faraday’s law, we have

Q ) nFAΓ

(2)

where n is the number of electrons, A is the electrode area (∼0.01 cm2), F is Faraday’s constant, and Γ the average covered surface (mol/cm2). From the calculated charge Q, a value of Γ ≈ 2‚10-10 mol/cm2 of HRP immobilized on the NCD electrode surface is derived. This value is slightly higher than the theoretical value

obtained for a flat surface homogeneously covered with the hydrated enzyme (volume 35 × 60 × 75 Å3).17 Values of the same order of magnitude were obtained when HRP was immobilized on SnO2.18 We attribute this result to the active area of the NCD surface used as immobilization support, as discussed above (see Figure 1), which is capable of accommodating a larger amount of enzyme than a flat surface. One of the most important parameters to describe the electrochemical behavior of the immobilized HRP is the electron-transfer constant (kET), which quantifies the direct electron-transfer efficiency. We have used the Laviron’s expression for a diffusionaless electrochemical system to determinate kET,19 which relates kET to the scan rate and the anodic and cathodic peak positions. Figure 10 shows the linear dependence of the anodic and cathodic potential peaks on the scan rate υ (V/s) on a logarithmic scale. According to Laviron, the plot yields two straight lines with slopes of -2.3 RT/RnF for the cathodic peak and 2.3 RT/nF(1 - R) for the anodic peak, respectively. From these slopes, a value of R ) 0.49 ( 0.02 was calculated. The separation between the anodic and cathodic peaks (Epa - Epc) is larger than the theoretical value (∆Ep ) 0) for a (17) Gajhede, M.; Schuller, D. L.; Henriksen, A.; Smith, A. T.; Poulos, T. L. Nat. Struct. Biol. 1997, 4, 1032-1038. (18) Jia, N.; Zhou, Q.; Liu, L.; Yan, M.; Jiang, Z. J. Electroanal. Chem. 2005, 580, 213-221 (19) Laviron, E. J. Electroanal, Chem. 1979, 101, 19-28, 101.

5842 Langmuir, Vol. 22, No. 13, 2006

Figure 11. Calibration curve of H2O2 obtained from the HRPNCD biosensor working at 0.05 V vs Ag/AgCl. The inset represents the current-time response of the biosensor.

surface-controlled process, which has been interpreted as the result of conformational effects,6 for instance, due to the different orientation of the enzyme catalytic group respect to the electrode surface. The electron-transfer constant was calculated from the equation19

log(kET) ) R log(1 - R) + (1 - R) log(R) R(1 - R)nF∆Ep RT log (3) nFυ 2.3RT We obtained kET ≈ 0.066 ( 0.003 s-1. This value is lower than the one obtained when resonant π-system molecules are used instead of TFA,20-34 and could be related to the isolation (20) Young-Tae, K.; Mannan, B.; Yoo-Bo, S. Biosens. Bioelectron. 2003, 19, 227-232. (21) Chen, J.; Yan, F.; Dai, Z.; Ju, H. Biosens. Bioelectron. 2005, 21, 330336. (22) Modern Aspects of Electrochemistry; Bockris, J. O’M., Conway, B. E., White, R. E., Eds,; Plenum Press: NY, 1992; Vol. 22. (23) Ferapontova, E. E. Electrochim. Acta 2004, 49, 1751-1759. (24) Ferna´ndez-Sa´nchez, C.; McNeil, C. J.; Rawson, K. Trends Anal. Chem. 2005, 24, 37-48. (25) Ferri, T.; Poscia, A.; Santucci, R. Bioelectrochem. Bioenerg. 1998, 45, 221-226. (26) Gruen, D. M. Annu. ReV. Mater. Sci. 1999, 29, 211-259. (27) Kamin, R. A.; Wilson, G. S. Anal. Chem. 1980, 52, 1198-1205. (28) Liu, H. H.; Tian, Q.; Lu, Z. X.; Zhang, Z. L.; Zhang, M.; Pang, D. W. Biosens. Bioelectron. 2004, 20, 294-304. (29) Pei, S.; Kian, P.; Wei, C.; Heng, Z. Diamond Relat. Mater. 2005, 14, 426-431. (30) Rahman, M. A.; Won, M. S.; Shim, Y. B. Biosen. Bioelectron. 2005, 21, 257-265. (31) Trushina, E.; Oda, R.; Landers, J.; McMurray, C. Electrophoresis 1997, 18, 1890-1898. (32) Yi, X.; Xian, J. X.; Yuan, C. H. Anal. Biochem. 2000, 278, 22-28. (33) Matsumiya, M.; Terazono, M.; Tokuraku, K. Electrochim. Acta 2005, in press. (34) Toman, J. J.; Corn, R. M.; Brown, S. D. Anal. Chim. Acta 1981, 1231, 87-19.

Rubio-Retama et al.

produced by the aliphatic monolayer acting as a barrier in the direct electron transfer from the heme group to the electrode surface. Biosensor Application. The activity of the biosensor was assayed according to the following standard procedure: the HRPNCD electrode was dipped into a vigorously stirred solution of 50 mM PBS, pH 6.2, and overpolarized at 0.05 V vs Ag/AgCl. Then, the current was recorded while H2O2 was added to the solution. The inset in Figure 11 shows the amperometric response of the HRP-NCD-modified electrode for successive additions of H2O2. The response time (below 5 s) shown in the inset of Figure 11 indicates a fast redox process. The calibration curve shown in Figure 11 gives an indication of the large linear range of this device, which covers from 0.1 to 45 mM (r ) 0.996). This result, together with the low reading potential (close to 0 V vs Ag/AgCl), allows the detection of H2O2, minimizing interferences due to ascorbic acid or uric acid present in biological samples.35-37 Biosensor Stability. The stability study carried out for 15 days on the HRP-NCD biosensor showed that the enzymatic activity was constant over this period on a device stored at room temperature, which is a clear indication of its robustness.

Conclusions Horseradish peroxidase has been effectively immobilized on nanocrystalline diamond thin films. The proximity between the prosthetic group and the electrode surface enables direct electron transfer. In addition, the HRP-modified NCD electrode was sensitive to the presence of H2O2 in a large linear range at 0.05 V vs (Ag/AgCl) in the four biosensors investigated. Thus, we can conclude that the immobilized enzyme maintains its activity while showing a good stability. Our results confirm that enzyme immobilization on NCD can open new possibilities for biosensor applications based on diamond thin films that could be used for in vivo applications or for optimizing clinical methodologies that are based on the measurement of hydrogen peroxide produced during the analytical process. Acknowledgment. Part of this work was funded via the NaDiNe Nano Diamond Network project, the Deutsche Forschungsgemeinschaft DFG (SFB 563/B15), the Spanish Science and Educational Ministry (MAT2003-03051-C03-03), the Comunidad de Madrid (6R/MAT/0501/2004), and the UCM (PR45/05-14177). LA060167R (35) Rubio-Retama, B. J.; Lo´pez Ruiz, B.; Lo´pez Cabarcos, E. Talanta 2005, 68, 99-107. (36) Rubio-Retama, B. J.; Lo´pez Ruiz, B.; Lo´pez Cabarcos, E.; Mecerreyes, D. Biosens. Bioelectron. 2004, 20, 1111-1117. (37) Darain, F.; Park, S. P.; Shim, Y. B. Biosens. Bioelectron. 2003, 18, 773780.