Analytica Chimica Acta 809 (2014) 141–147

Contents lists available at ScienceDirect

Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca

Thin-film electrochemical sensor for diphenylamine detection using molecularly imprinted polymers Vera L.V. Granado a , Manuel Gutiérrez-Capitán b , César Fernández-Sánchez b , M. Teresa S.R. Gomes a , Alisa Rudnitskaya a , Cecilia Jimenez-Jorquera b,∗ a b

CESAM and Chemistry Department, University of Aveiro, 3810-193 Aveiro, Portugal Instituto de Microelectrónica de Barcelona (IMB-CNM), CSIC, 08193 Bellaterra, Barcelona, Spain

h i g h l i g h t s

g r a p h i c a l

a b s t r a c t

• New

voltammetric sensor for diphenylamine detection. • Thin-film miniaturized electrode modified with a molecularly imprinted polymer. • Molecularly imprinted particles anchored into an electropolymerised poly(3,4-ethylenedioxythiophene) membrane. • Analysis of diphenylamine in spiked apple juice samples.

a r t i c l e

i n f o

Article history: Received 28 July 2013 Received in revised form 14 November 2013 Accepted 16 November 2013 Available online 23 November 2013 Keywords: Molecularly imprinted polymers Voltammetric detection Thin-film electrode Diphenylamine

a b s t r a c t This work reports on the development of a new voltammetric sensor for diphenylamine based on the use of a miniaturized gold electrode modified with a molecularly imprinted polymer recognition element. Molecularly imprinted particles were synthesized ex situ and further entrapped into a poly(3,4ethylenedioxythiophene) polymer membrane, which was electropolymerized on the surface of the gold electrode. The thickness of the polymer layer was optimized in order to get an adequate diffusion of the target analyte and in turn to achieve an adequate charge transfer at the electrode surface. The resulting modified electrodes showed a selective response to diphenylamine and a high sensitivity compared with the bare gold electrode and the electrode modified with poly(3,4-ethylenedioxythiophene) and nonimprinted polymer particles. The sensor showed a linear range from 4.95 to 115 ␮M diphenylamine, a limit of detection of 3.9 ␮M and a good selectivity in the presence of other structurally related molecules. This sensor was successfully applied to the quantification of diphenylamine in spiked apple juice samples. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Molecular imprinting has gained popularity during the last decades as a technique of synthesizing polymer materials with

∗ Corresponding author at: Instituto de Microelectrónica de Barcelona (IMBCNM), CSIC, Campus de la UAB, 08193 Bellaterra, Barcelona, Spain. Tel.: +34 93 594 7700; fax: +34 93 580 1496. E-mail address: [email protected] (C. Jimenez-Jorquera). 0003-2670/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.aca.2013.11.038

chemically selective recognition sites [1–3]. Molecular imprinting consists in the polymerization of the monomer mixture in the presence of the target molecule or template in an inert solvent. After polymerization, the template is removed from the polymer matrix, thus leaving cavities or specific binding sites in the resulting material that selectively interact with the template molecule and that give rise to molecularly imprinted polymer matrices (MIPs). The high specificity and stability of MIPs as well as the possibility to synthesize polymers for practically any analyte render them attractive artificial ligands or receptors for various analytical applications including chemical sensing [4].

142

V.L.V. Granado et al. / Analytica Chimica Acta 809 (2014) 141–147

The use of the MIPs as stationary phases for liquid chromatography [5], sorbents for solid-phase extraction [6,7] and ligand binding assays [8] have been reported. Although less numerous, MIPs have also been applied as recognition elements in various types of sensors, including mass [9], impedimetric [10], amperometric [11] and potentiometric [12–14] sensors, among others. An important aspect when using a MIP as a selective receptor in chemical sensing is the selection of the appropriate transducer for transforming the corresponding MIP binding event into an analytical signal. Different methods for the integration of MIPs with the transducer have been reported in the literature [15] and have been applied to the fabrication of electrochemical sensors [16,17]. MIPs have been prepared in situ by electropolymerization in the presence of the template [18,19]. This is an easy and fast technique producing stable membranes with controlled thickness and good adherence to the electrode surface. However, these sensors often show poor selectivity. Drop-casting, spin-coating or spray coating of a pre-polymerization mixture onto the transducer surface [20] are other easy and rapid reported methods. The resulting membranes have proved to be stable in both aqueous and organic solutions. All these techniques are limited to particular polymer formulations and often involve laborious synthesis processes. Entrapping the MIP particles into gels [21] or using different types of composites such as inks containing conductive carbon materials [22] or composites containing MIP particles, graphite and a solid binding matrix [23] are alternative methods for the fabrication of MIP-based electrochemical sensors. These types of membranes are susceptible to non-specific adsorption, leakage and the preparation procedure is also laborious. The advantage of this approach is the possibility to use established MIP formulations. Another approach for the preparation of MIP-based sensors combines the methods explained above and consists of the electrosynthesis of a polymeric matrix for the entrapment of the MIP particles onto the transducer surface [15,18]. With this approach, a conducting polymer is synthesized electrochemically in the required monomer solution also containing the MIP suspension. Thus, MIP microparticles are entrapped within the generated polymer matrix and in turn directly immobilized onto the transducers surface. One of the most appealing aspects of this approach is the possibility of decoupling MIP synthesis and immobilization, thus enabling better optimization of each step separately. At the same time it takes advantage of the versatility of the electrochemical polymerization processes that enables a strict control over the membrane thickness and the surface active area of the synthesized polymer layer. Diphenylamine (DPA) is an agrochemical product with high antioxidant properties widely employed to control storage scald on apples and pears [24]. Considering its low solubility in water, it is not completely removed from the fruit and therefore, DPA and its derivatives can appear in processed products such as apple juice. This product is included in the third European Union (EU) list of priority pollutants [25] and actually its use is no longer authorized within the EU [26]. Therefore, as a reference the maximum allowed concentration for apples and pears established in the EU directive 91/414/EEC was between 5 and 10 mg kg−1 . The analytical techniques for the determination of diphenylamine are based on gas chromatography, high-performance liquid chromatographic combined with different detection techniques, mass spectrometric methods and spectrophotometry. Although these methods are quite sensitive, they require treatment of the sample, are usually time consuming and use expensive instrumentation. Regarding sensors for DPA detection, no any electrochemical sensor has been reported in the literature. In the present work, the fabrication and characterization of a miniaturized thin-film voltammetric sensor for the detection of DPA is presented. MIP particles were synthesized following

a protocol previously described [27] and deposited on the surface of microfabricated gold electrodes [28] by entrapment in an electropolymerised poly(3,4-ethylenedioxythiophene) (PEDOT) membrane. The use of miniaturized thin-film electrodes has multiple advantages for the preparation of MIP based sensors. Firstly, the miniaturized size of these sensors permits the use of small amounts of all the required reagents and solutions and secondly, the use of deposition techniques scalable to mass production is feasible. MIP based sensors have been optimized according to the electropolymerization time and have been characterized for DPA detection. The good results obtained with this sensor have permitted to test their feasibility for DPA measurements in spiked apple juice samples. 2. Materials and methods 2.1. Reagents and solutions Acetonitrile (99%), magnesium perchlorate hydrate (99%), 3,4ethylenedioxythiophene (97%), 1-naphtylamine (98%), nitric acid (69%), methanol (99%), potassium nitrate (99%), potassium hydroxide (85%), potassium ferricyanide (99%), catechol, bisphenol A and gallic acid were all purchased from Sigma–Aldrich and used as received. Acetic acid glacial (99.7%) was purchased from Panreac and used as received. All solutions were prepared using de-ionized water. Real samples of concentrated apple juice were purchased from supermarket. 2.2. Apparatus Gold thin-film electrodes were fabricated at the Instituto de Microelectrónica de Barcelona (IMB-CNM) according to standard photolithographic technology. The fabrication process is explained elsewhere [29]. Working area of the electrode was 4.39 mm2 . Three-electrode amperometric cell was used for batch measurements, which included a gold electrode as working electrode, an external Pt counter electrode (XE100, Radiometer Analytical, Villeurbanne CEDEX, Lyon, France) and a Ag/AgCl (KCl 10%, w/v) reference electrode (0726.100 Metrohm, Herisau, Switzerland). A ␮-Autolab potentiostat/galvanostat (EcoChemie B.V., Utrech, The Netherlands), using GPES 4.7 software package (General Purpose Electrochemical System) was used for batch voltammetric measurements. SEM images were recorded on the gold coated polymeric films using a scanning electron microscope (SEM, Auriga from Zeiss) operated at 5–10 kV. 2.3. Electrode preparation Polymers imprinted with DPA were prepared according to the procedure reported elsewhere [27]. Briefly, methacrylic acid (MAA) was used as a monomer, trimethylolpropane trimethacrylate (TRIM) as a cross-linker, 2,2 -azobis(2-methylpropionitrile) as a catalyst and acetonitrile as a solvent. Synthesis was run under nitrogen atmosphere at 60 ◦ C. Imprinted polymers were prepared using monomer/template ratio 7.5 and polymerization time 10 h. The respective non-imprinted polymer (NIP) was selected for the electrode membrane preparation according to the previous results [27]. Prior to the membrane deposition, gold electrodes were chemically cleaned and electrochemically activated. First, the electrode surface was rinsed using a brush successively wetted in ethanol, deionized water, sulphuric acid (6 M) and deionized water, and then dried under nitrogen flow. Electrochemical activation was

V.L.V. Granado et al. / Analytica Chimica Acta 809 (2014) 141–147

143

made by cycling the potential 20 times between 0.8 and −2.2 V in a 0.1 M KNO3 solution. Electrodes with electropolymerized membranes containing the imprinted polymer (MIP), the respective non-imprinted polymer (NIP) and no particles (blank) have been prepared. Electrosynthesis solution was prepared by dissolving 1 mM of EDOT monomer in the 0.1 M magnesium perchlorate solution in acetonitrile. Then, for the preparation of the membranes with polymeric particles 20 mg of MIP or NIP particles were added to the solution and well dispersed by ultrasonication for 10 min. Three different methodologies were used in order to deposit the PEDOT/MIP film on the gold electrode: by performing consecutive cyclic scans between −0.6 and 1.2 V, by fixing the potential at 1.2 V (potentiostatic) and by fixing the current through the electrode (galvanostatic). The best results in terms of DPA permeability were obtained under galvanostatic conditions by applying a current of 20 ␮A in the above solution, previously deareated using a nitrogen flow. With the aim of optimizing membrane thickness, different electropolymerization times were considered: 1, 2, 4 and 8 min. Modified electrodes were rinsed with acetonitrile, dried and kept in the air while not in use. The electrochemical behaviour of the modified electrodes was tested in a 0.1 M KNO3 solution containing 1.0 mM K3 Fe(CN)6 at 50 mV s−1 . 2.4. Voltammetric measurements The electrochemical response of the bare gold electrode to DPA was firstly investigated using cyclic voltammetry (CV). Two cycles in the potential range from 0.0 to 1.0 V were recorded in a deaerated 0.1 M HNO3 pH 2 solution containing 83.3 ␮M DPA at a scan rate of 0.1 V s−1 . For further experiments, detection of DPA was made by differential pulse voltammetry (DPV) in the potential range from 0.6 to 0.85 V. This technique provides a high sensitivity given that the effect of the charging current is minimized. The measurement parameters were as follows: 0.1 V s−1 scan rate; −0.2 V standby potential; 0.025 V modulation amplitude; 0.05 s modulation time; 10 s equilibration time. The area of the recorded peak was used as the analytical signal. All experiments were performed at room temperature. Electrochemical performance of PEDOT/MIP modified electrodes was evaluated also by DPV with DPA solutions in 0.1 M HNO3 whose pH was adjusted to 2, 4 and 6 by addition of the required volume of 1 M NaOH solution. The calibration curves were obtained by addition of increasing amounts of a 0.5 mM DPA stock solution. Calibration measurements were made in the concentration range 4.95–115 ␮M of DPA. After each measurement the thin-film electrodes were washed with solutions of methanol and acetic acid (1:9). Three replicates for each concentration and electrode were made. Selectivity of the modified electrodes was evaluated for catechol, gallic acid, aniline, bisphenol A and 1-naphthylamine. Measurements were performed in a 0.1 M HNO3 solution pH 2 containing an equimolar concentration of DPA and the interfering compound. Three replicates were also made. Voltammetric selecsel , were determined using the following tivity coefficients, Kvoltamm equation [30]: sel Kvoltamm =

A

t

Ai



−1 ×

Ci Cj

(1)

where At and Ai are the area under the peak recorded in the mixed solution (DPA and interfering specie) and in the solution containing only DPA, respectively, and Ci and Cj are the concentrations of DPA and interfering species, respectively. Stability of PEDOT/MIP modified electrodes was evaluated by regularly making calibration measurements in DPA solutions during an 18 day-period. Three calibrations per sensor were made on each occasion and three sensors were used each time. After

Fig. 1. (A) Cyclic voltammograms of the gold electrode in a solution without DPA and containing 83.3 ␮M DPA and (B) DPV response of the gold electrode to changes in DPA concentration. In both cases, the background solution is 0.1 M HNO3 (pH 2) and scan rate of 0.1 V s−1 .

measurements, sensors were washed with the methanol and acetic acid (1:9 v:v) mixture, dried and kept on the air. Sensors with PEDOT/MIP membranes were used for the detection of DPA in apple juice. Three different brands of 100% apple juice were purchased at a local retailer. Juice samples were filtrated, diluted two-fold with 0.2 M HNO3 solution and spiked with DPA at the concentration level of 7.5 ␮M (approx. 1.3 mg kg−1 ). This value corresponds to typical levels of DPA detected in apple juice [31]. Samples pH was adjusted to pH 2 with 1 M NaOH prior to measurements. DPA concentration in juice was determined by the standard addition method. Measurements were made in triplicate for each juice sample. 3. Results and discussion 3.1. Voltammetric response of the bare gold electrode to DPA Prior to membrane deposition, electrochemical behaviour of DPA was studied by cyclic voltammetry (CV) on a bare gold electrode. CV scans in a 0.1 M HNO3 pH 2 solution without DPA (Blank) and containing 83.3 ␮M DPA are shown in Fig. 1(A). A high anodic peak at 0.73 V is observed on the forward scan, which can be ascribed to the oxidation of DPA to DPAH+• radical cation and its further dimerization to give phenylbenzidine [32]. On the reverse scan two peaks appear at 0.52 V and 0.35 V, which correspond to the reduction of DPAH2+ cations and/or DPAH+• radicals and the reversible redox process of the DPA dimer or polymer, respectively. Cathodic peak potentials are shifted to the more negative values compared to the reference [32], which is presumably due

144

V.L.V. Granado et al. / Analytica Chimica Acta 809 (2014) 141–147

Fig. 2. SEM images of PEDOT/MIP membranes deposited on the gold electrode: membrane electrodeposited during 1 min (A), 2 min (B), 4 min (C) and 8 min (D). Also a SEM image of the PEDOT without MIP synthesized during 1 min is shown (E).

to the difference in the background electrolyte used (0.1 M HNO3 vs. 1 M H2 SO4 ). Therefore, the recording of the oxidation peak of DPA at 0.73 V by scanning the potential from 0.60 V to 0.85 V was selected for the detection of DPA by differential pulse voltammetry (DPV). Fig. 1(B) illustrates how the oxidation peak increases with the addition of increased amounts of DPA. It is well known that DPA oxidation involves protonation and thus depends on the pH of the solution. A significant decrease of both anodic and cathodic peak intensity with increase of pH has been reported in the literature [32]. In order to test the effect of pH on the sensor response, calibrations were performed at pH 2, 4 and 6. The slope (sensitivity) values of the corresponding calibration curves were 1.7 × 10−3 ␮C/␮M, 1.6 × 10−3 ␮C/␮M, and 1.4 × 10−3 ␮C/␮M, respectively. The decrease of sensitivity as the pH increases is likely to be related to the coupled chemical reactions that take place in solution during the DPA electrochemical process. DPA shows a pKa around 1. At pHs higher than 3, DPA is deprotonated in solution. It has been already reported that the redox species generated during the electrochemical anodic process are more prone to dimerization and polymerization. At pH 2, these negative effects are avoided

due to DPA being mostly protonated. Therefore, further measurements were carried out at pH 2. Under these conditions, the area under the anodic peak at +0.73 V was directly related to the concentration of DPA in solution. The results showed a sensitivity of 0.32 × 10−3 ␮C/␮M in a linear concentration range of 4.95–115 ␮M. 3.2. Optimization of the membrane electropolymerization process It was expected that MIP particles would promote DPA preconcentration on the electrode surface and its transport through the membrane thus enhancing sensitivity and selectivity of the modified sensor compared to the bare gold electrode. First experiments were addressed to optimize the method of immobilization of MIP particles. Immobilization of MIP particles on the surface of the electrode was carried out by synthesizing electrochemically a conducting polymer PEDOT membrane using a solution containing the EDOT monomer and the MIP suspension. As expected, the gold electrodes were well coated with the electropolymerized PEDOT membranes containing the particles imprinted with DPA. The thickness of

V.L.V. Granado et al. / Analytica Chimica Acta 809 (2014) 141–147

145

Table 1 Analytical characteristics of modified electrodes and thickness of PEDOT/MIP membranes electropolymerized during different periods of time. Standard deviation is shown within parentheses. Parameters of unmodified gold electrode are shown for comparison. Electrode

Membrane thickness (nm)

Slope × 103 (␮C/␮M) (n = 5)

Detection limit (␮M)

Bare gold 1 min 2 min 4 min 8 min

0 30–50 90–100 140–190 700–800

0.32 (0.05) 1.74 (0.01) 1.30 (0.10) 1.50 (0.07) 0.37 (0.05)

5.4 (0.5) 18 (6.0) 14 (1.0) 78 (10.0)

the PEDOT polymeric membrane was firstly optimized taking into account that it could significantly influence the analytical performance of the resulting sensor. Too thin membranes could result in low sensor selectivity by allowing other species to diffuse to the gold surface. By contrast, thicker membranes may lead to the decrease of the sensitivity due to the negative effect on the analyte transport through the polymeric layer. Therefore, electropolymerized membranes of different thickness were generated by varying the polymerization time from 1 min to 8 min. The thickness of the obtained PEDOT/MIP membranes varied about 20 times: from 30–50 nm for 1 min to 700–800 nm for 8 min (Table 1). SEM images of the transversal membrane cuts were used for the estimation of the membrane thickness (not shown). Fig. 2 shows topographical SEM images of PEDOT based membranes containing MIP particles. It can be observed that, regardless the electropolymerization time, agglomerates of MIP particles are formed on the surface of the PEDOT membrane. For shorter polymerization times and thinner membranes (Fig. 2a), MIP particles seem to be attached to the surface with PEDOT polymer network serving as a kind of a hook. For longer polymerization times (4–8 min) and consequently thicker membranes, a higher density of MIP particles can be observed, which appears to be anchored and coated by the polymer (Fig. 2b–d). A SEM image of a PEDOT membrane without MIP particles is also shown in Fig. 2e for comparative purposes. In Fig. 3 it is shown in more detail how the PEDOT layer grafts the MIP particles (Fig. 3a) and the cross-section of one MIP particle (Fig. 3b), carried out by cutting the particle with a focused ion beam. It is also indicated the diameter of this particle and the apparent thickness of the metal layer and the PEDOT layer. Selection of the optimal membrane thickness was carried out by comparing the analytical characteristics (sensitivity and detection limit) of the resulting sensors. Modification of the gold electrode with a PEDOT/MIP membrane resulted in a 5 times increase of the sensitivity compared with the bare electrode, as expected (Table 1). In Fig. 4 it is shown a DPV voltammogram of the PEDOT/MIP. On the other hand, increase of the membrane thickness up to 90 nm led to a significant decrease in sensitivity and increase of the detection limit for DPA (Table 1). Moreover, the electrode response became non-linear at higher DPA concentrations for membranes with polymerization times longer than 4 min. Considering these results, membrane polymerized during 1 min was chosen as the optimal for the sensor fabrication and used for all further experiments.

Fig. 3. SEM images of PEDOT/MIP membranes (1 min electrodeposition). (A) Image of MIP particles grafted by PEDOT film. (B) Cross-section of the sensor surface. The diameter of a MIP particle and the apparent thickness of the metal layer and the PEDOT membrane are indicated.

3.3. MIP sensor response to DPA In order to study the features of the MIP membrane, studies of the sensor response with the PEDOT membrane and the respective non-imprinted polymer (PEDOT/NIP) were carried out. As shown in Fig. 5 calibrations of sensors with DPA in a range of 4.95–115 ␮M revealed that electrodes modified with MIP particles displayed higher sensitivity compared to the electrodes with PEDOT or PEDOT/NIP membranes indicating that DPA is capable

Fig. 4. Differential pulse voltammetry response of the PEDOT/MIP modified electrode to changes in DPA concentration using a 0.1 M HNO3 (pH 2) background solution and scan rate of 0.1 V s−1 .

146

V.L.V. Granado et al. / Analytica Chimica Acta 809 (2014) 141–147 Table 2 Selectivity coefficients for several compounds calculated through Eq. (1).

PEDOT NIP MIP

0.20

Area / µC

0.15

Compound

sel Kvoltamm

Gallic acid Aniline Catechol 1-Naphtylamine

0.025 0.216 0.128 0.659

0.10 Table 3 Results of DPA determination in spiked apple juices using sensors modified with PEDOT/MIP membranes. Standard deviations are shown within the parentheses.

0.05

0.00

0

20

40

60

80

100

120

[DPA] / µM Fig. 5. Calibration plots for the different sensors: with PEDOT film, with PEDOT/NIP membrane and with PEDOT/MIP membranes. Concentration range of 4.95–115 ␮M DPA.

of diffusing through the PEDOT film to the gold electrode surface where it is oxidized. Therefore NIP does not have any influence on this process. The enhanced response of the PEDOT/MIP modified electrodes can be attributed to the existence of cavities with complementary shape and size. The stabilization of the interaction is based on weak hydrogen-bonds between the DPA and the functional groups of the MIP, which enabled pre-concentration of the DPA and facilitated its diffusion to the electrode surface thus giving rise to the signal enhancement. The response characteristics of MIP modified sensor were a sensitivity of 1.74 × 10−3 ␮C/␮M in a linear range of 4.95–115 ␮M and a limit of detection of 5.4 ␮M DPA (LOD = 3 ␤ /S, where  ␤ and S are the standard deviation of the intercept and the slope of the calibration plot respectively). This LOD (around 1 mg kg−1 ) is lower than the maximum level established by EU directive. Regarding the stability study, the response of three electrochemical sensors containing PEDOT/MIP membranes in DPA solutions were measured during 18 days (Fig. 6). Significant decrease of the slope was observed between the first and second measurement carried out 6 days later, which could be attributed to the changes in the membrane occurred during its first contact with DPA aqueous solution. After that, the sensor sensitivity was quite stable decreasing only 15% between day 6 and day 18. The detection limit decreased slightly after the first measurement, remaining stable at values of about 5 ␮M during the next 12 days.

Apple juice sample

DPA (␮M)

Recovery

1 2 3 4 5 6

7.7 (0.6) 7.7 (0.3) 7.3 (0.2) 7.5 (0.3) 7.5 (0.4) 7.8 (0.1)

103% 103% 97% 100% 100% 104%

Selectivity of the sensors modified with PEDOT/MIP membranes was evaluated towards compounds with structures similar to DPA: aniline, catechol, gallic acid and 1-naphtylamine. Aniline, catechol and gallic acid are not electroactive in the potential window used for the DPA detection. According to the literature data, their oxidation peaks are situated at +0.55 V for the aniline [33], +1.12, +1.36 and +1.5 V for gallic acid [34] and +0.4 V for catechol [35] (potentials vs. saturated calomel reference electrodes). Oxidation peak at +0.8 V in acidic media has been reported for 1-naphthylamine [36]. Therefore, it was expected that interference of these compounds on the sensor performance would result from the competition for the sorption sites in the MIP particles, while in case of 1-naphthylamine the overlapping of its anodic peak could also contribute to the interference for DPA response. High selectivity of the MIP modified electrodes was observed in the presence of all compounds except 1-naphthylamine (Table 2). From the results shown in Table 2, it can be assumed that the main interference mechanism is based on the competition for the binding sites of the MIP particles, except for 1-naphtylamine. For this molecule a combination of voltammetric signal overlapping and MIP-binding site competition appears to take place. 3.4. Determination of DPA in real samples Determination of DPA residues in fruits can be rather cumbersome due to the matrix complexity and low concentrations of DPA, which requires sample treatment at various stages of extraction. The analytical assessment of the developed voltammetric sensors for the quantification of DPA in vegetal matrices was evaluated in apple juices spiked with DPA. Results of determination of DPA concentration by standard addition method are presented in Table 3. As can be seen, the change of matrix, from water to apple juice, did not appear to interfere in the DPA quantification. Values from 97% to 104% recovery for the spiked samples were found. It is worth mentioning that the detected concentration is lower than the maximum allowed limits for DPA residue in fruits [31]. Therefore, the developed MIP modified electrodes could be applied to the control of DPA residues in fruit juices. 4. Conclusions

Fig. 6. Evolution of the calibration slope (sensitivity) and detection limit of a PEDOT/MIP modified electrode measured during 18 days (mean values and standard deviations for 3 electrodes are shown).

A voltammetric sensor for diphenylamine based on the immobilization of MIP particles on the surface of a miniaturized gold electrode using an electrosynthesized PEDOT matrix has been developed. First optimization results demonstrated that the increase of the PEDOT membrane thickness resulted in the

V.L.V. Granado et al. / Analytica Chimica Acta 809 (2014) 141–147

deterioration of sensitivity and detection limit. The optimum membrane thickness of about 50 nm was obtained with a polymerization time of 1 min. High selectivity of modified electrodes to phenolic compounds structurally similar to DPA such as gallic acid, aniline and catechol was found while 1-naphtylamine interfered with sensor response. These MIP-based voltammetric sensors displayed significantly lower detection limits (5.4 ␮M vs. 290–350 ␮M) and consequently wider working range compared to the potentiometric sensors previously developed [27]. Therefore, the voltammetric sensors described in this work appear to be more suitable for the detection of DPA at low levels present in foodstuffs and environment. This premise has been demonstrated with the sensor application to determination of DPA in spiked apple juice samples obtaining from 97% to 104% recovery for levels typically found in fruits. It is worth mentioning that this is the first time that a DPA voltammetric sensor is described in the literature. This sensor allows the detection of DPA with a high sensitivity and a limit of detection able to be applied for food analysis, thus offering an alternative to current methods for DPA detection based on conventional analytical instrumentation. Acknowledgements Financial support for the stage of Vera Granado at the IMB-CNM by REDSENSA, CYTED, project 510AC0408 is acknowledged. This work was supported by MINECO (Madrid) under grant TEC201129045-C04-01. References [1] C. Alexander, H.S. Andersson, L.I. Andersson, R.J. Ansell, N. Kirsch, I.A. Nicholls, J. O’Mahony, M.J. Whitcombe, J. Mol. Recognit. 19 (2006) 106–180. [2] L. Chen, S. Xu, J. Li, Chem. Soc. Rev. 40 (2011) 2922–2942. [3] E.L. Holthoff, F.V. Bright, Anal. Chim. Acta 594 (2007) 147–161. [4] S.A. Piletsky, A.P.F. Turner, Electroanalysis 14 (2002) 317–323.

[5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36]

147

E. Turiel, A. Martin-Esteban, Anal. Bioanal. Chem. 378 (2004) 1876–1886. D. Jiang, X. Sun, Y. Zhang, Anal. Methods 4 (2012) 3760–3766. J. Yin, Z. Meng, Y. Zhu, M. Song, H. Wang, Anal. Methods 3 (2011) 173–180. R.J. Ansell, D. Kriz, K. Mosbach, Curr. Opin. Biotechnol. 7 (1996) 89–94. L. Cao, X.C. Zhou, S.F.Y. Li, Analyst 126 (2001) 184–188. Z.L. Cheng, E.K. Wang, X.R. Yang, Biosens. Bioelectron. 16 (2001) 179–185. C. Pellicer, A. Gomez-Caballero, N. Unceta, M. Aranzazu Goicolea, R.J. Barrio, Anal. Methods 2 (2010) 1280–1285. R. Liang, R. Zhang, W. Qin, Sens. Actuators B: Chem. 141 (2009) 544–550. S.P. Pogorelova, A.B. Kharitonov, I. Willner, C.N. Sukenik, H. Pizem, T. Bayer, Anal. Chim. Acta 504 (2004) 113–122. R.N. Liang, D.A. Song, R.M. Zhang, W. Qin, Angew. Chem. Int. Ed. 49 (2010) 2556–2559. C. Malitesta, E. Mazzotta, R.A. Picca, A. Poma, I. Chianella, S.A. Piletsky, Anal. Bioanal. Chem. 402 (2012) 1827–1846. M.C. Blanco-Lopez, S. Gutierrez-Fernandez, M.J. Lobo-Castanon, A.J. MirandaOrdieres, P. Tunon-Blanco, Anal. Bioanal. Chem. 378 (2004) 1922–1928. V. Suryanarayanan, C.T. Wu, K.C. Ho, Electroanalysis 22 (2010) 1795–1811. K.C. Ho, W.M. Yeh, T.S. Tung, J.Y. Liao, Anal. Chim. Acta 542 (2005) 90–96. E. Mazzotta, R.A. Picca, C. Malitesta, S.A. Piletsky, E.V. Piletska, Biosens. Bioelectron. 23 (2008) 1152–1156. M.C. Blanco-Lopez, M.J. Lobo-Castanon, A.J. Miranda-Ordieres, P. Tunon-Blanco, Biosens. Bioelectron. 18 (2003) 353–362. Z.H. Zhang, Y.F. Hu, H.B. Zhang, L.J. Luo, S.Z. Yao, J. Electroanal. Chem. 644 (2010) 7–12. T. Yamazaki, Z. Meng, K. Mosbach, K. Sode, Electrochemistry 69 (2001) 969–972. T. Alizadeh, L.A. Fard, Anal. Chim. Acta 769 (2013) 100–107. B.D. Whitaker, Postharvest Biol. Technol. 18 (2000) 91–97. O. Drzyzga, Chemosphere 53 (2003) 809–818. European Food Safety Authority, EFSA J. 9 (2011) 2336. V.L.V. Granado, A. Rudnitskaya, J.A.B.P. Oliveira, M.T.S.R. Gomes, Talanta 94 (2012) 133–139. R. Olive-Monllau, J. Orozco, C. Fernandez-Sanchez, M. Baeza, J. Bartroli, C. Jimenez-Jorquera, F. Cespedes, Talanta 77 (2009) 1739–1744. J. Orozco, G. Suárez, C. Fernández-Sánchez, C. McNeil, C. Jiménez-Jorquera, Electrochim. Acta 53 (2007) 729–736. C. Macca, J. Wang, Anal. Chim. Acta 303 (1995) 265–274. Environmental Protection Agency, vol. 64, 1998, p. 8273. G. Inzelt, J. Solid State Electrochem. 6 (2002) 265–271. D. Bejan, A. Duca, Croat. Chem. Acta 71 (1998) 745–756. M.K. Carter, J. Mol. Struct. 831 (2007) 26–36. D. Nematollahi, M. Rafiee, J. Electroanal. Chem. 566 (2004) 31–37. W.F. Smyth, M.R. Smyth, Pure Applied Chem. 59 (1987) 245–256.