Tyrosinase is a copper-containing enzyme that converts phenolic compounds to the corresponding quinones in the presence of oxygen [25]. In order to explore the influence of the immobilization matrix in the performance of the biosensors, Langmuir-Blodgett (LB) and carbon paste electrodes (CPE) were constructed and immersed in 4 × 10⁻⁴ M phosphate buffer solutions (0.01 M, pH 7) containing one monophenol (phenol), one diphenol (catechol) or one polyphenol (catechin). The response of the biosensors was registered in the range from −0.5 V to +0.5 V at a scan rate of 0.050 V × s⁻¹.

In previous works, it has been demonstrated that the presence of lutetium bisphthalocyanine (LuPc₂) electron mediator improves the sensitivity of the Langmuir-Blodgett biosensors towards phenols [31,32,33]. A similar effect has been observed here for CPE biosensors. This is illustrated in Figure 1 where the electrochemical response of LB (Figure 1a,c) and CPE electrodes (Figure 1b,d) towards catechol with and without electron mediator is shown.

In absence of electron mediator, the enzyme oxidizes the phenol to the quinoid form, whose back electrochemical reduction can be detected at −0.2 V versus Ag/AgCl/KCl 3 M. (Figure 1a,b). In the presence of the lutetium bisphthalocyanine, the intensity of the cathodic peak at −0.2 V increases both in LB and CPE sensors, but in the case of LB films the increase is of almost one order of magnitude. Simultaneously a shift to lower potentials is observed in the presence of LuPc₂. The shift is more marked in LB films where it appears at positive potentials (+0.1 V) (Figure 1.a) whereas in CPE electrodes the peak is observed at ca. −0.1 V (Figure 1.b). This result seems to indicate that the biomimetic environment provided by the LB films improves the interaction between the phthalocyanine, the enzyme and the phenol, facilitating the electron transfer and hence the biocatalytic activity.

The presence of LuPc₂ is also involved in two other interesting observations. First, a redox pair associated to the reversible one electron reduction of the phthalocyanine ring should be observed at E_1/2 of ca. −0.24 V [37,38]. Surprisingly, this peak is clearly seen in biomimetic LB films, but it is almost absent in the CPE electrodes. Secondly, in the presence of the electron mediator, an anodic peak at +0.38 V associated to the electrochemical oxidation (non enzymatically mediated) of catechol is observed in LB film based electrodes. As reported in previous studies, the oxidation peak of catechol (using carbon electrodes) appears at 0.59 V (at pH 3.6) [16]. According to this result, it can be concluded that the presence of LuPc₂ facilitates the oxidation of catechol, which occurs at lower potentials than in carbon electrodes, confirming the electrocatalytic effect of the LuPc₂. It is important to remark that this anodic peak is clearly observed in LB films but cannot be observed in CPE sensors. Because the electrocatalytic effect of the LuPc₂ is not so efficient in CPE electrodes, the non enzymatic oxidation peak appears at higher potentials and cannot be observed in the voltage range studied.

Apart from the differences in positions and relative intensities of the peaks from LB to CPE sensors, it is also remarkable that the intensity of the voltammograms registered using biomimetic electrodes is clearly higher (ca. one order of magnitude) than the intensity measured in non biomimetic electrodes.

The response of the biosensors towards other phenolic compounds including catechin (Figure 1e,f) and phenol (Figure 1g,h) has also been studied. As observed in the figure, tyrosinase is able to use mono, di- and polyhydroxyphenols as substrates, but the affinity depends on the nature of the phenolic species.

The response towards catechin, displays the same general behavior than the response towards catechol, but some particularities are observed. For both phenols, the cyclic voltammograms exhibit (a) a cathodic peak at ca. −0.1 V related to the reduction of the previously oxidized o-quinone, (b) an anodic peak at ca. 0.4 V associated to the electrochemical oxidation of the phenol and (c) the redox pair associated to the phthalocyanine. The affinity of the enzyme towards catechin is lower than the affinity to catechol (a diphenol). For this reason, the intensity of the peaks related to catechin (both anodic electrochemical oxidation of phenol and cathodic reduction of the o-quinone) is lower than that of the phthalocyanine. The main difference is that the intensity of the reduction peak observed when catechin solution was analyzed appear at +0.1 V for LB biosensor and −0.08 V for the CPE biosensor. The electrochemical oxidation of catechin appears only in the case of LB biosensor (at +0.3 V) and cannot be clearly observed in the CPE.

In previous studies, it has been established that the oxidation peak of catechin appears at 0.455 V (at pH 3) and the formal potential increases by 0.066 V per pH unit [16]. Taking into account these preceding data, it can be concluded that the presence of LuPc₂ facilitates the oxidation of catechin, which occurs at lower potentials than in carbon electrodes; in addition, the electron mediator increases drastically the intensity of the signals of both the enzymatic and the electrochemical process with respect to the signals obtained in absence of the electron mediator thus demonstrating an electrocatalytic effect of the system.

Finally, the LB biosensor immersed in phenol solution shows a reduction peak at +0.01 V. In the case of the CPE biosensor the reduction peak appear at −0.07 V. The presence of the reduction peak indicates that the enzyme shows certain biological activity in both solid substrates. However, the anodic peak corresponding to the oxidation of the phenol is not observed nor in LB or in CPE electrodes.

Therefore, the value of the potential that must be applied to monitor reduction or oxidation of the species at the electrode surface was in −0.08 to +0.1 V range, potential frame that allows to reach a minimum of possible electrochemical interferences.

Influence of scan rate in biosensors response was carried out by registering the cyclic voltammograms of the biosensors at different sweep rates, from 0.01 to 0.21 V × s⁻¹. The effect of the scan rate is illustrated in Figure 2 for catechol (Not all CVs are shown).

In both classes of electrodes, the intensity of the cathodic peak (at ca. −0.1 V) resulting from the reduction of the o-quinone enzymatically formed, were proportional to the sweep rate, according to the Laviron equation (Equation 1) (Figure 2c,d): I_c = n²F² νAΓ/4 RT (1) where I_c is the cathodic peak current (Ampere), n is the number of electrons involved in the redox process, F is the Faraday constant (F = 96,485.3365 C/mol), ν is the potential scan rate (V × s⁻¹), A is the electrode area (cm²), Γ is the surface coverage of the redox species (mol × cm⁻²), R is the ideal gas constant (8.3144621 J × K⁻¹ × mol⁻¹) and T is the temperature (K) [39]. This linear relationship between I and the scan rate indicates that there is no diffusion limitation and that redox processes are controlled by the electron transfer of the adsorbed species at the electrode surface [40,41]. From the slope of the I vs. ν graph the total surface coverage Γ could be calculated. The principal parameters of linear regression equation of the plots I_c vs. ν are presented in Table 1.

As observed in the Table, the Ty/LuPc₂-LB biosensors showed higher surface coverage values (Γ) than CPEs, indicating that the number of active sites is superior in the nanostructured LB thin films than in the carbon-based matrix sensors. From this data it can be deduced that nanostructured sensors show an improvement of the enzymatic activity which is related to the higher surface area provided by the nanostructured environment and to the preservation of the enzymatic activity in the biomimetic environment.

The influence of the scan rate in the intensity of the peaks was also studied using the peak related to the electrochemical oxidation of catechol and catechin (anodic peak at ca. 0.4 V) (this peak was absent in phenol). In the case of the LB films, the intensity of this peak increased linearly with the square root of the sweep rate (Figure 2c,d) indicating a diffusion controlled processes in accordance with the Randles-Sevcik equation (Equation 2) [37]: I_pa = 2.687 × 10⁵n^3/2ν ^1/2D^1/2AC (2) where I_pa is the anodic peak current (Ampere), n is the number of electrons involved in the redox process, ν is the potential scan rate (V × s⁻¹), D is the diffusion coefficient (cm² × s⁻¹), A is the electrode surface area (cm²), and C is the concentration (mM). From the slope of I_a vs. v^1/2 plot, the diffusion coefficient D could be calculated (Table 2).

In the case of CPEs, and in voltammograms registered at low scan rates, the anodic peak at 0.4 V was very weak. Nevertheless, when increasing the scan rate, the intensity of the peak increased progressively and the peak could be clearly observed at high scan rates. The intensity of the anodic peak shown by CPE sensors was also linearly dependent with the square root of the scan rate following the Randles-Sevcik equation, where the diffusion of species (from the bulk to the electrode surface and inside the sensitive layer) is controlling the process.

From the results displayed in Table 2, it is evidenced that the Ty/LuPc₂-LB sensors showed diffusion coefficients two orders of magnitude higher than CPE sensors pointing that the electrochemical processes are faster in the case of nanostructured thin film.

The response to increasing amounts of the phenolic compounds was studied using cyclic voltammetry and amperometry. The effect of the concentration in the Voltammetric responses is illustrated in Figure 3 for an LB sensor immersed in catechin (from 10 μM to 40 mM) at a scan rate of 0.05 V × s⁻¹. As observed in the figure, the intensity of both the enzymatic and the electrochemical peaks increased with the concentration of the phenolic compound (not all CVs were shown).

Figure 4 shows the amperometric signals for the Ty/LuPc₂-CPE biosensor at −0.07 V (Figure 4a) and for the Ty/LuPc₂-LB biosensor at +0.01 V (Figure 4b) upon the addition of successive aliquots of phenol to the 0.01 M PBS (pH 7.0), under constant stirring. Well-definite reduction currents proportional to the concentration of phenol were observed, which result from the electrochemical reduction of o-quinone enzymatically formed at the biosensor surface.

The Ty/LuPc₂-LB biosensor reached 95% of steady-state current in less than 4 s. The response rate was much slower in the case of the Ty/LuPc₂-CPE biosensor that reached the steady state in 10 s. The faster response observed in the nanostructured films could be attributed to the favorable enzymatic environment provided by the fatty acid layer.

Figure 5 shows the correlation between the response current of the biosensors and the concentration of phenol in PBS (pH 7.0) at +0.01 V for Ty/LuPc₂-LB biosensor and −0.07 V for Ty/LuPc₂-CPE biosensor.

The response current of the LB nanostructured biosensor is linear with the phenol concentration in the range from 10 to 120 μM, indicating that the catalytic reaction of Ty is a first-order reaction. Concentrations higher than 120 μM cause a lost of the linearity: the current increases slowly, and the enzyme reaction shows a change from first to zero-order. The corresponding detection limits were calculated according to the 3s_b/m criterion, where m is the slope of the calibration graph, and s_b is estimated as the standard deviation (n = 7) of the amperometric signals obtained for the lowest concentration of the calibration plot. The detection limit calculated for the LB biosensor was 5.4 μM. Using the same method, the detection limit calculated for the CPE biosensor was 8.57 μM.

Similar results were obtained for catechol and catechin. In all cases the detection limit was lower for the nanostructured biomimetic LB sensor and the lowest detection limit was obtained for catechol.

From the calibration data, the Hill coefficient (h) was calculated by representing the log[I/(I_max − I)] vs. log [S] (the logarithm of substrate concentration). A Hill coefficient of 1.09 ± 0.03 (R² = 0.952) was obtained for the reduction process of o-quinone formed from the enzymatic reaction on the electrode surface for Ty/LuPc₂-LB biosensor. In the case of Ty/LuPc₂-CPE biosensor a Hill coefficient of 0.94 ± 0.03 was calculated.

Both types of sensors showed an h parameter close to unity demonstrating that the kinetics of the enzymatic reaction fitted into a Michaelis–Menten type. The value slightly higher than 1 obtained for the LB biosensor (h = 1.09 ± 0.03) indicates a positive cooperative effect between the occupied active sites. A negative cooperative effect between the occupied active sites takes place in the case of Ty/LuPc₂-CPE biosensor (h = 0.94 ± 0.03).

The apparent Michaelis–Menten constant () was calculated for the immobilized Ty by using the linearization of Lineweaver-Burk expressed by Equation 3.

1/I = 1/I_max + /(I_max × [S]) (3) where I is the cathodic current, I_max is the steady-state current, is the apparent Michaelis-Menten constant and [S] is the concentration of substrate.

The maximum current response and apparent Michaelis–Menten constant were calculated from the intercept and slope respectively [42]. The values obtained for both biosensors immersed in phenolic compounds solutions are presented in Table 3.

The small value of Michaelis–Menten constant found in the biomimetic LB sensor indicates that the affinity between the tyrosinase and the substrate is stronger in LB than in CPE sensors. In turn, the superior I_max found in LB sensors is a reflection of the higher sensitivity of the nanostructured biosensor.

The results confirm that the affinity of the tyrosinase towards the catechol (a diphenol) is better than towards other types of phenols such as catechin or towards a monophenol (the phenol) that showed the highest and lowest I_max.

The stability of the biosensors was studied at room temperature as well as in cooled conditions (at 4 °C). The stability of the enzyme inside of the solid matrices (LB and CPE) was monitored during one month by measuring the changes in the intensity of the cathodic peak when the electrodes were immersed in 50 μM catechol. In both types of electrodes, a very slow decrease in the response current was observed. When stored at 4 °C, biosensors based on LB films maintained the 85% of their initial enzymatic activity for 1 month, whereas in CPE an 80% of their activity was maintained after one month. This result could be explained by taking into account the improved compatibility of the enzyme with the lipidic biomimetic environment provided by the LB film. These long-term stabilities of LB and CPE biosensors are better than those reported in literature (<15 days) [43].

Lifetime data indicate that both the LB and CPE biosensors can be cycled up to 50 times. The decrease in signal intensity expressed as percentage of decay was lower than 4% over 50 continuous scan cycles in the case of LB biosensor and 6% in the case of CPE biosensor.

The reproducibility of the biosensor signals was investigated at a catechol concentration of 2.0 × 10⁻⁴ M. In the case of LB biosensor, the observed mean steady-state current was 32.52 μA with a relative standard deviation (S.D.) of 2.6% for seven determinations. In the case of the CPE biosensor, the mean steady-state current was 5.44 μA with a relative standard deviation (S.D.) of 4.3% for seven determinations.

All reagents were of high purity and used without further purification (Sigma–Aldrich). All solutions were prepared with deionised water (18.3 MΩ × cm resistivity, Milli-Q, Millipore). Mushroom tyrosinase (from mushroom EC 232-653-4), with a noted activity of 5370 U/mg of solid (product T3824-250KU), was purchased from Sigma Chemical Co. (USA). A 67 μg × μL⁻¹ solution of tyrosinase in buffer phosphate 0.01 M (pH = 7) was used for the enzyme immobilization. The buffer phosphate solution (PBS) 0.01 M (pH 7) was prepared from potassium monobasic and dibasic phosphate salts. The lutetium (III) bisphthalocyaninate (LuPc₂) was synthesized and purified in their neutral radical state following earlier published method [44].

Carbon paste electrodes containing tyrosinase and LuPc₂ (Ty/LuPc₂-CPE) were prepared as previously reported by mixing graphite powder (Ultracarbon^®, Ultra F purity), lutetium bisphthalocyanine (15%, w/w) and Nujol (Fluka) as the binder of the composite mixture [45]. Pastes were packed into the body of a 1mL PVC (polyvinyl chloride) syringe and compressed. A metallic copper wire was used as an electrical contact. The enzyme, tyrosinase (Ty), was immobilized on the carbon paste surface by a dip coating technique.. For this purpose, 10 μL of 0.01 M phosphate buffer (pH 7.0) containing 67 μg × μL⁻¹ of the enzyme, was dip coated onto the surface of the carbon paste electrode.

LB thin films were prepared in a KSV 5000 System 3 Langmuir–Blodgett trough equipped with a Wilhelmy plate to evaluate the surface pressure. LB films containing tyrosinase, LuPc₂ and arachidic acid (Ty/LuPc₂-LB) were prepared by spreading a chloroform solution (10⁻⁵ M) of arachidic acid and LuPc₂ onto a water subphase (NaCl 0.1M, phosphate buffer 0.01 M of pH 7 in ultrapure water–Millipore MilliQ; 20 °C). After the evaporation of the solvent, 20 μL of a 67 μg × μL⁻¹ solution of tyrosinase in 0.01 M phosphate buffer (pH 7) was injected drop by drop underneath the air/water interface. The floating film was compressed using a symmetrical two barrier compression system. At a surface pressure of 30 mN × m⁻¹, 30 monolayers were deposited onto an ITO (indium tin oxide) glass. The substrate speed used was 2 mm × min⁻¹. LB thin films were prepared by Y-type deposition with a transfer ratio close to 1. The biosensor was rinsed using phosphate buffer solution and stored at 4 °C.

Electrochemical experiments were performed using a three electrode system. The potentiostat used was an EG & G PARC, Model 2273 potentiostat/galvanostat (Princeton Applied Research Corp.). All measurements were carried out in 0.01 M phosphate buffer adjusted to pH 7.0. Temperature control at 25 °C was achieved using a circulating bath (Neslab).