Fabrication of a novel electrochemical sensing platform based on a core–shell nano-structured/molecularly imprinted polymer for sensitive and selective determination of ephedrine

Abstract

A simple methodology was used to develop a novel sensor based on a core–shell/molecularly imprinted polymer (MIP) for the determination of ephedrine. The Fe₃O₄@SiO₂@TiO₂-MIP nanocomposite with a well-defined core–shell structure was synthesized using a simple imprinting strategy. The morphology of the synthesized nanocomposite was observed by scanning electron microscopy, X-ray diffraction, Fourier transform infrared spectroscopy and the properties of the sensor were examined by cyclic voltammetry and electrochemical impedance spectroscopy. The fabricated electrochemical sensor based on the Fe₃O₄@SiO₂@TiO₂-MIP nanocomposite exhibits great features such as a remarkably low detection limit of 0.0036 μmol L⁻¹ (3S_b/m), superb selectivity in discriminating ephedrine from its structural analogues and good anti-interference ability towards several co-existing substances. Also, it was used to detect the concentration of ephedrine in the linear range of 0.0090–2.8 μmol L⁻¹. Moreover, the proposed method demonstrates excellent repeatability and stability, with a relative standard deviation (RSD) of less than 1.4% and 1.6%, respectively. Analysis of ephedrine in pharmaceutical dosage forms and biological fluids is successfully carried out without the assistance of complicated pretreatment. The Fe₃O₄@SiO₂@TiO₂-MIP nanocomposite presented here with admirable merits makes it a promising candidate for developing an electrochemical sensor device to perform routine analysis of ephedrine.

---

Introduction

Ephedrine (EP) is a sympathomimetic amine drug. It is an important drug which is used in the treatment of bronchial asthma, and in allergic conditions like utricaria.¹ EP is also used as appetite suppressant, decongestant and to treat hypotension. However, EP alkaloids have amphetamine-like properties at high doses and can cause nervousness, tachycardias, hypertension, seizures and psychosis.^1–3 Also, it has an improving effect on physical performance and is included in the doping list of forbidden pharmacological substances indicated by the medical commission of the international Olympic committee.^3,4 Recently, EP assessment in food products, pharmaceutical formulations, human fluids of athletes and detection of drug toxicity and abuse, has gained a growing interest. An athlete may be suspected as “positive” if the concentration of this compound is too high in his/her biological fluids. Furthermore, in the clinical and pharmacokinetic studies, the concentrations of EP in serum plasma and saliva are often mg L⁻¹ level or even lower.^2,3,5 Therefore, the sensitive monitoring of this species in biological samples is of great and continuous interest.⁶

Several studies have reported the determination of EP based on different analytical techniques such as high performance liquid chromatography (HPLC), liquid chromatography-mass spectrometry, gas chromatography-mass spectrometry, electrophoresis, spectrophotometry, spectrofluorimetry and electrochemical methods.^2,3 However, there may be several disadvantages for these methods. For example, some of them are time consuming and tedious, while others use large biological fluid volumes and need expensive instruments and toxic solvents, plus time consuming procedures.

Electrochemical methods, especially electrochemical biosensors/sensors based on MIP, have attracted more attention in recent years due to their recognition properties (high selectivity), high sensitivity, simplicity, good stability, low cost, fast response and real time detection.^7–9 MIP has been demonstrated as a powerful technique in designing and synthesizing some artificial receptor molecules which is based on the copolymerization of functional monomers and cross-linking agent in the presence of the template molecule. MIPs with artificially generated recognition sites, which are specific in shape and size to the target molecule, are capable of specifically binding a target molecule in preference to other closely related compounds.^10,11

Recently, nanotechnology and nanoparticles have introduced with MIP to improve their structural and functional properties such as low density of imprinted sites, slow binding time, heterogeneous distribution of imprinted sites, insulating polymer and low electrochemical signal in electrochemically application of MIPs. The imprinted nanomaterials can be prepared either mixing the polymers with nanoparticle or by synthesizing the MIPs in the form of nanoparticles.^12,13 The former way is more flexible, easy to use and more popular technique, known as surface imprinting in which nanoparticles were used as a supporting materials for MIP synthesis. Typically, surface imprinting has been adopted to avoid entrapment of template in cross-linked polymers. Surface imprinting over nanosized support materials with unique structure, high chemical stability, good conductivity and high surface-to-volume ratio is very common and popular too, which provides easy access of template molecule to the imprinting site in order to provide an alternative way can improve mass transfer and reduce permanent entrapment of templates.^12–15 Metal nanoparticles (Au, Ag, Pt, etc.),¹⁶ carbon nanotubes,¹⁷ graphene sheet,¹⁸ quantum dots,¹⁹ and magnetic nanoparticles²⁰ like nanomaterials/nanoparticles were used as platform for MIP synthesis. Among them, the magnetic nanoparticles (MNPs) of metal oxide (e.g., NiO, Fe₃O₄ and Co₃O₄) have been shown as most proficient substrate material for preparation of imprinted polymers. When these MNPs are coated with polymer shells, the resulting imprinted nanomaterial possesses the characteristics of both MNPs as well as MIP. The imprinted MNPs have been widely applied in the field of catalysis, environmental pollutants analysis, pharmaceuticals, food sector, chromatography and sensors.^12,20,21

However, pure magnetic particles are prone to form aggregates and their magnetic properties can be altered in complex environmental and biological systems. To solve the above problems, a suitable protective coating on a magnetic core is often used. Silica has been considered as one of the most ideal shell materials due to its reliable chemical stability, biocompatibility and versatility in surface modification. Because of the synergistic effects of the two distinct nanomaterials, the core–shell silica magnetic nanoparticles have high surface areas. Sensors/biosensors modified with core/shell MNPs have much potential in terms of sensitive detection of analytes.²² However, modified core/shell MNPs as a platform for MIP synthesis is rare. In order to enhance the electrode sensitivity and amplify the electrochemical response, Fe₃O₄@SiO₂@TiO₂ (FST) was introduced in insulating MIP matrix. Titanium dioxide is the most widely investigated material among all of the metal oxides because of its unique properties and promising applications in variety of fields, including photocatalysis,²³ supercapacitors,²⁴ lithium ion batteries,²⁵ solar cells,²⁶ sensors,^27,28 and biosensors.²⁹ TiO₂ is chemically stable in both acidic and alkaline solutions and has high catalytic activity for the reduction of several small organic molecules. The presence of core/shell nanostructure in the composition of electrode modifier enables fast electron transfer kinetics, increases the electroactive surface area, and reduces over-potential.

EP has been a popular template for molecular imprinting studies, being employed in fundamental studies, as a template and analyte for chromatography studies, capillary electrophoresis, solid-phase extraction, and sensor devices.³⁰ In this paper, by using methyl methacrylate (MMA) as the functional monomer, magnetic molecular imprinted polymer (MMIP) was prepared based on FST via a microwave heating method. It was incorporated as a modifier into carbon paste electrode (CPE) for enhancing in sensitivity and selectivity of the voltammetric determination of EP. FST, as a noble metal nanocomposite modifier, plays dual role in this work: it works as a solid support for imprinting; it also acts as a conductive material in the non-conductive polymer matrix. This motif in fabrication of modified CPE offered several advantages such as ease of preparation, lower limits of detection, a wide usable concentration range and high selectivity and stability. Thus, FST-MIP/CPE exhibit both predetermined selective molecular recognition properties and high electrical conductivity.

Experimental

Apparatus and chemicals

Electrochemical measurements were carried out by Behpajoh potentiostat/galvanostat system (model BHP-2065). A three electrode configuration consisted of CPE and modified CPEs (as working electrodes), Ag/AgCl (as a reference electrode) and a platinum wire (as an auxiliary electrode). X-ray powder diffraction (XRD, X'Pert Pro MPD, PANalytical, Netherland) was employed to analyze the chemical components of the composites. Fourier transform infrared (FTIR) spectra (KBr dispersed pellets) in the range of 400–4000 cm⁻¹ (model spectrum 100 FTIR of Perkin Elmer spectrometer). The morphology of synthesized materials was determined by a ZEISS-EM902 transmission electron microscope (TEM). The morphology images of composites were observed with a XL30 scanning electron microscope (SEM-EDX, Philips Netherland). Impedance measurements were performed using an IVIUM-STAT potentiostat (Ivium Technologies, The Netherlands). The validation of results was performed by HPLC (Agilent 1100, USA).

All the reagents were analytical graded and used without further purification. EP hydrochloride was prepared from Sigma-Aldrich. All required solutions were prepared using deionized distilled water (DDW). Britton–Robinson (B–R) universal buffer solution (0.04 M boric acid, 0.04 M acetic acid and 0.04 M phosphoric acid) as the supporting electrolyte was prepared in DDW that had been titrated to the desired pH with 0.2 M NaOH.

Synthesis of the FST-MIP

Magnetic Fe₃O₄ nanoparticles were prepared according to the literature.³¹ Also, the preparation process of FST followed by the literature³² with some modifications, the specific process was as follows: briefly, 6 mL tetraethyl orthosilicate (TEOS) was added into 40 mL anhydrous ethanol with mechanical agitation for 20 min at 40 °C, the mixed solution (contained 40 mL anhydrous ethanol, 3 mL DDW and 12 mL concentrated ammonia) was added into above solution drop by drop, the reaction was conducted under mechanical agitation for 30 min. Subsequently, 2.0 g magnetic Fe₃O₄ were added, and the reaction was continued for another 8 h. Finally, the sample was collected by magnet and washed with DDW and anhydrous ethanol for several times, vacuum dried at room temperature, Fe₃O₄@SiO₂ (FS) was obtained. In the following step, 8 mL tetrabutyltitanate (TBOT) was added into 40 mL anhydrous ethanol with mechanical agitation for 20 min at 40 °C, the mixed solution (contained 40 mL anhydrous ethanol, 3 mL DDW and 0.3 mL HCl) was added drop wise into above mixture. When the sol was formed, 2 g FS was added into above sol and kept stirring to the gel. Finally, the gel was vacuum dried at room temperature and calcinated for 4 hours at 500 °C to obtain FST.

FST-MIP was prepared under optimized synthesis method as follows: 2.5 g polyethylene glycol 4000 (PEG 4000) and 0.5 g FST were dissolved in 8 mL DDW by ultrasonication at room temperature, hereinafter called modified FST. Subsequently, 0.015 g EP, 0.04 mL MMA and 1 mL DMSO were put together into the microwave reactor. After magnetic stirring for 15 min, this mixed solution was placed for 12 h in the dark under a nitrogen atmosphere (the prepolymerization process). Afterwards, 0.25 mL trimethylolpropanetrimethacrylate, 0.005 g 2,2'-azobis(2-methylpropionitrile) as free radicals initiator and above modified FST were added into this microwave reactor, the polymerization was carried out in the microwave synthesizer at 500 W and 60 °C for 1 h under a nitrogen atmosphere. Subsequently, EP as the molecular template was removed by adding 150 mL DDW into above solution under the UV light irradiation for 5 h with the magnetic agitation under an air atmosphere. The solid sample was washed with DDW and anhydrous ethanol and vacuum dried at room temperature. The preparation of magnetic non-imprinted polymer (FST-NIP) was followed FST-MIP, but without adding EP.³³ The schematic of the molecular imprinting process for EP determination is shown in Scheme 1.

Scheme 1 The schematic of the molecular imprinting process for EP determination.

Fabrication of FST-MIP/CPE

Graphite powder and paraffin oil were mixed in 75 : 25 (w/w%) ratio in a mortar, followed by grinding for making a paste as unmodified CPE electrode. A portion of the prepared paste was packed firmly a piston-driven CPE holder (2.5 mm diameter). The Fe₃O₄/CPE was prepared by mixing 10% (w/w) Fe₃O₄, 65% (w/w) graphite powder and 25% (w/w) paraffin oil in a mortar and pestle. The mixture was homogenized and was used in the same way as the case of the unmodified electrode. The FST/CPE was prepared by mixing the unmodified mixture with 10% w/w FST and transferred into the CPE holder. The FST-MIP/CPE and FST-NIP/CPE were prepared by mixing the carbon pastes with FST-MIP (15% w/w) and FST-NIP (15% w/w), respectively, and then the resulted mixtures transferred into the syringes. The paste was carefully packed in to the syringe tip to avoid possible air gaps, which often enhance the electrode resistance. The external surface of the carbon paste was smoothed with soft paper. A new surface was produced by scraping out the old surface and replacing the new carbon paste.

Samples preparation

The serum samples were centrifuged and then after filtering, diluted with B–R buffer solution (pH = 10.5) without any further treatment. Urine samples were analyzed immediately after their collections. 10 mL of the sample was centrifuged for 15 min at 2000 rpm. The supernatant was filtered using a 0.45 mm filter and then diluted 5 times with B–R buffer solution (pH = 10.5). The solution was transferred into the voltammetric cell to be analyzed without any further pretreatment. The standard addition method was used for the determination of EP.³⁴

For pharmaceutical formulation of EP, each tablet contained a dose of 25 mg EP as a substantial and electrochemically active component of the tablet (without any other generally known additive drugs), as the manufacturer guarantees in drug information leaflet. Eight tablets were completely grinded and homogenized and 120 mg of powdered tablet was dissolved in DDW with intensive stirring of magnetic stirrer for 15 min to ensure that tablets were dissolved completely. The mixture was filtered through a filter paper to obtain a clear filtrate and then quantitatively transferred into 100 mL volumetric flask. To obtain final concentrations in the range of calibration curve, the sample solutions were suitably diluted with supporting electrolyte.

Results and discussion

Characterization

The FT-IR spectra of Fe₃O₄, FS, FST and FST-MIP are shown in Fig. 1a. In all spectra, the absorption band about 3400–3600 cm⁻¹ was assigned to the stretching vibration of hydroxyl group, the absorption band at 1600–1700 cm⁻¹ was assigned to the bending vibration of O–H.^33,35 As shown in Fig. 1a, for Fe₃O₄, the presence of Fe–O bond vibration at 620 cm⁻¹ are obviously observed. This pattern corresponds to the Fe–O bonds, which is reported to belong to bulk magnetite. The deposition of silica network on the magnetite surface by Fe–O–Si bonds through silanization was also confirmed by obtaining relevant FT-IR spectrum. The corresponding absorption band cannot be seen in the FT-IR spectrum because it appears at around 600 cm⁻¹ and inevitably overlaps with the Fe–O vibration of Fe₃O₄ (ref. 31). However, the strongly absorbing region of 1020–1110 cm⁻¹ in the spectrum of FS in Fig. 1a results from the vibration of Si–O–H and Si–O–Si groups. For FST, the absorption bands in the range from 500 to 900 cm⁻¹ relating to stretching vibration of Ti–O–Ti bond are clearly observed. Furthermore, compared with FST, a lot of additional absorption bands appeared in the spectrum of FST-MIP and indicated that the surface-imprinted layer has been successfully coated on the surface of FST.³³

Fig. 1 Comparison of (a) FTIR spectra (b) XRD patterns for Fe₃O₄, FS, FST, FST-MIP (c) SEM images for Fe₃O₄, FST, FST-MIP and (d) their TEM images.

The XRD patterns of the synthesized Fe₃O₄, FS, FST and FST-MIP were presented in Fig. 1b. The XRD pattern of Fe₃O₄ was in agreement with the JCPDS card no. 19-0629, presenting the characteristic peaks [at 2θ about: 30, 35, 43, 53, 57 and 63.0°] of cubic spinel structure. No impurity was observed. The characteristic diffraction peaks of Fe₃O₄ are weakened because of the silica coat and the mixed group modification, and the diffraction peak of amorphous silica can be observed (FS). In addition, FST was composed of anatase TiO₂ (JCPDS card no. 21-1272) along with the magnetic phase. The peaks about at 25° and 48° revealed the presence of anatase TiO₂ in the synthesized nanocomposites.³⁶ In addition, the pattern of FST-MIP was nearly the same with FST, indicating that the phase of anatase TiO₂ was not changed by coating the surface-imprinted layer.

SEM was performed to obtain an insight into the surface morphology of the different synthesized materials, as shown in Fig. 1c. Fig. 1c shows the SEM of Fe₃O₄ nanoparticles, FST core/shell and FST-MIP, respectively. The average diameter of Fe₃O₄ nanoparticles is 10–20 nm with a spherical shape in, the aggregation of the nanoparticles can be discerned clearly. In Fig. 1c, the FST core–shell structure can be observed. The dispersity of FST is also improved, and the average size is increased to about 25–45 nm. It can be seen that the surface of FST-MIPs exhibits a porous and rough structure. This porousness and roughness of the surface of MIP should be considered as a factor providing in the surface area. Also, its SEM image demonstrated the small cavities of template molecules. Created cavities in leached polymer can be related to the removal of EP from polymer particles after the leaching process.

Fig. 1d shows the TEM images of all the MNPs. It revealed that the Fe₃O₄ particles were of irregular shape. However, core–shell structure of FST was successfully prepared with regular spherical shape and relatively narrow size distribution. The SiO₂ and TiO₂ layers were clearly seen to be uniformly coated on Fe₃O₄ dark core. After imprinting process, an external polymer layer was clearly observed around FST particles, which suggested that MIPs layer had been successfully grafted on the surface of FST particles.

Electrochemical impedance spectroscopy studies

Electrochemical impedance spectroscopy (EIS) is a suitable technique for investigating of the electrode surface dependent charge transfer process (interfacial properties, i.e., resistance and capacitance). The curve of the EIS includes a semicircular part and a linear part. The semicircular part at higher frequencies corresponds to the electron-transfer-limited process and its diameter is equal to the electron transfer resistance, which controls the electron transfer kinetics of the redox probe at the electrode interface. The measurements are generally performed in faradaic mode in frequency range 0.1–100000 Hz and formal potential of 0.2 V, using redox-probe ferrocyanide/ferricyanide in order to focus on the variations of the charge transfer resistance (it is equal to the semicircle diameter) between the solution and the electrode surface. The equivalent circuit compatible with the Nyquist diagram recorded for fabricated sensor is depicted in Fig. 2. In this circuit, R_s, C_dl and R_ct represent solution resistance, a capacitance for the double-layer and charge transfer resistance, respectively. W is a finite-length Warburg short-circuit term coupled to R_ct, which accounts for the Nernstian diffusion. This equivalent circuit (Fig. 2 inset) was used to fit the impedance spectra and extract the values of C_dl and R_ct. After the modification of the electrode composition with various specific materials, the value of R_ct changes, due to their conduction properties. By monitoring the value of R_ct after each modification step in electrode composition, most sensitive sensing layer can be selected with remarkable accuracy. Fig. 2 illustrated the impedance spectra observed at each step in the presence of 1.0 mmol L⁻¹ Fe(CN)₆^3−/4− solution.

Fig. 2 Nyquist plots for various prepared electrodes in the 1.0 mmol L⁻¹ [Fe(CN)₆]^4−/3− and 0.1 mol L⁻¹ KCl. Inset: equivalent circuit for this system.

As shown in Fig. 2, when CPE was unmodified, the semicircle associated with R_ct increased obviously, suggesting that CPE has a low rate of electron transfer due to its poor conductivity (1180 Ω). However, after CPE was modified with Fe₃O₄, the R_ct value decreased distinctively to 810 Ω indicating that Fe₃O₄ could accelerate the electron transfer between the electrochemical probe [Fe(CN)₆]^3−/4− and the electrode surface.

It was observed that FST/CPE, displayed almost straight line in the Nyquist plot and the slope of lines increased dramatically, indicating that there was improved diffusion of ferricyanide toward the electrode surface. As shown in Fig. 2, after CPE modification with FST, the R_ct decreased to 130 Ω, suggesting faster electron transfer kinetics of [Fe(CN)₆]^3−/4− on the electrode surface. At FST-MIP/CPE, the R_ct increased to 390 Ω, indicating that the MIPs formed an additional barrier on the surface of electrode to block the electron exchange between the solution and the electrode. This is due to the fact that the polymeric modifier was not conductive and there were almost no channels for the active probe to access the electrode surface. After removing the template, the R_ct decreased to 240 Ω, indicating that the template was successfully removed and formed some cavities which facilitated the electron exchange between the redox and electrode surface. Also, the estimated C_dl values for CPE, Fe₃O₄/CPE, FST/CPE, FST-MIP/CPE before removal and FST-MIP/CPE after the removal were 2.73, 2.67, 2.56, 3.09 and 2.45 μF, respectively. The measured capacitance usually arises from the series combination of several elements, such as analyte binding to a sensing layer on an electrode.

Determination of electrochemical active surface area

In order to calculate the electrochemical active surface areas of bare CPE and various modified electrodes, the cyclic voltammetry (CV) of potassium ferrocyanide, as a redox probe, was performed by Randles–Sevcik equation:³⁷

I_p = (2.69 × 10⁵)n^3/2AC*D^1/2ν^1/2 (1)

where I_p refers to the anodic peak current, n is the total number of electron transferred (n = 1), A is the effective surface area of the electrode, D is the diffusion coefficient for K₄[Fe(CN)₆] = 7.6 × 10⁻⁶ cm² S⁻¹, C* is the concentration of K₄[Fe(CN)₆] and ν is the scan rate. The effective surface areas of various electrodes were calculated from the slope of the I_p versus ν^1/2 plot. The surface area of CPE (0.082 cm²) is less than surface areas of other electrodes. The calculated surface areas for Fe₃O₄/CPE, FST/CPE, FST-MIP/CPE (Fig. S1 in ESI†) and FST-NIP/CPE were 0.107, 0.177, 0.143 and 0.121 cm², respectively. Compared to the unmodified electrode, FST/CPE nanocomposite spread on the electrode surface increased the number of sites for electron transfer to electroactive species. The results showed that surface area of FST-MIP/CPE is about two fold greater than unmodified CPE. These increase of the electroactive surface area in modified electrodes exhibited the influence of used nanocomposite as an effective modifier that provide a large surface and facilitate the electron transfer between the electrode and the solution.

Electrochemical behaviour

Electrochemical performance of various prepared electrodes for the determination of 2.5 μmol L⁻¹ EP in the B–R buffer (pH = 10.5) have been investigated by CV and are presented in Fig. 3. For CPE, cyclic voltammogram did not exhibit any voltammetric peak. After modifying of the electrode with Fe₃O₄, a weak anodic peak appears at potential about 0.75 V. A small oxidation peak of EP was observed at the FST-NIP/CPE due to non-selective binding sites in the NIP. Also, this might be attributed to the weak conductivity of NIP, due to the electro-insulating characteristics of it which has fewer cavities than MIP. At FST/CPE, a significant increase in peak current of EP was observed. This indicates better electrochemical properties of FST/CPE due to its large surface area and high conductivity. Also, the apparent peak shape for EP at FST/CPE is improved against Fe₃O₄/CPE, because the presence of FST providing the excellent electrochemical activity and good antifouling property. Oxidation current increased dramatically when the MMIP nanocomposite was introduced to the CPE. This result may be attributed to the synergistic effect between the three components. The remarkable specific surface area, conductivities and high affinity of sensing layer may be the main contribution that can amplify the electrochemical signal.²⁷ In the absence of EP, the FST-MIP/CPE did not show any peak in the potential region 0.4 to 1.0 V.

Fig. 3 Cyclic voltammograms for 2.5 μmol L⁻¹ EP in B–R buffer solution (pH = 10.5) on the surface of various electrodes.

Effect of pH on the electrochemical oxidation of EP

The effect of the solution pH on the peak current of EP was evaluated when the accumulation and electrochemical oxidation perform in two separated solutions and also when accumulation and detection occur in the same solution. In order to investigate the effect of pH, the prepared electrode was inserted into the EP solution with various pH, and then was incubated for 6 min at a constant stirring rate. Then, the electrode was removed from the EP solution and after washing, voltammograms were recorded. The results of this experiment are shown in Fig. S2.† As it is seen the EP accumulation increased with the increase of the pH values to 9.5. However, after 9.5, peak current decreased. The isoelectric point (i_ep) is an important characteristic of adsorbents. Their surface charges are largely dependent on the pH of the solution, the pH_PZC caused by the amphoteric behavior of surface groups and the interaction between surface sites and the electrolyte species. The i_ep of FST-MIP/CPE was calculated about 5.4 using reported method.³¹ The surface charge is positive at pH values lower than pH_PZC, neutral at pH_PZC, and negative at pH values higher than it. The accumulation of EP increased by increasing the solution pH from 5.0 because FST-MIP has negative charge while EP is in protonated form. The electrostatic interaction between EP and the electrode surface at this pH range could be responsible for increasing in currents. According to the literature³ the pK_a value of EP is 9.6. When the pH of solution exceeded the pK_a of EP or increased further (pHs > 9.6), EP tended more to exist in natural form, which would gradually decrease to interact with electrode surface electrostatically and current values decline.

The effects of different pH values on the electrochemical oxidation current of accumulated EP at the surface of the FST-MIP/CPE were also investigated in the B–R buffer solutions (EP-free solution) in a pH range of 5.0–12.0 (Fig. S3†). The obtained results show that peak currents of EP increased with increasing the solution pH from 5.0 to 11.0, beyond which it decreased. In general, when the pH of the solution was much lower than the pK_a, owing to the strong protonation, the oxidation might become difficult and responses were lower. It was concluded that the anodic peak current increased as pH rises from 5.0 to 11.0. However, further increase in the pH of the buffer solution decreases the peak current. The decrease of the peak current in pH values higher than 11.0 might be resulted from the loss of activity of FST-MIP in alkaline medium.

Furthermore, the effect of pH on peak potential and peak current of EP was investigated over the pH range of 5.0–12.0 employing B–R buffers when accumulation and detection occur in the same solution. As shown in Fig. 4b, with increasing pH, the current increases and highest peak currents were obtained for pH between 10.0 and 11.0. It could be explained by the consequence of deprotonation involved in the oxidation process which was facilitated at higher pH values. At pHs below the EP pK_a, (pH < 9.6), where the EP species is initially protonated, the oxidation current increased markedly with pH. The variation of the surface properties of the electrodes and electrostatic interactions between EP and the electrode surface at different pHs also could be responsible for this phenomenon. After pH 11, oxidation current decreases with increasing pH. It may be related to the stability of sensing layer is decreased in strong alkaline media.

Fig. 4 (a) Differential pulse voltammograms for fabricated sensor in 1.0 μmol L⁻¹ of EP at various pHs (pH 5.0, 6.0, 7.0, 8.0, 8.5, 9.0, 9.5, 10.0, 10.5, 11.0, 12.0). (b) Effect of pH on the peak currents of EP (c) effects of pH on the peak potentials of EP and (d) cyclic voltammograms in different pHs.

Fig. 4c shows that the peak potential of EP is also pH dependent and the potentials shifted negatively when the pH of the solutions increases due to the participation of protons in the electrode reaction. This behaviour can be explained due to the deprotonation step involved in oxidation process that is facilitated at higher pH values. These dependences are linear over the whole studied pH range and two linear plots obtained; one from pH 5.0 to 9.5 and the second from pH 9.5 to 12.0. Each curve showed a deviation from linearity at pH = 9.5, which is virtually coincident with the pK_a value of EP (pK_a = 9.6).³ They can be described by the following equations (eqn (2) and (3)):

E = −0.087pH + 1.59 (R² = 0.997) (pH 5.0 to 9.5) (2)

E = −0.031pH + 1.06 (R² = 0.988) (pH 9.5 to 12.0) (3)

The cyclic voltammograms of 2.0 μmolL⁻¹ EP solution at FST-MIP/CPE in various pHs are shown in Fig. 4d. EP showed an irreversible anodic peak and no reduction peak was obtained in the reverse scans at any pHs. A small current was observed at pH 5.0, which increased up to pH 10–11 and then decreased gradually with the further increase of buffer pH. Fig. 4d demonstrates the electrode reaction of EP is an irreversible surface reaction. For the irreversible surface electrochemical reaction, the relationship between the peak potential E_p and the pH in the CV is expressed in the eqn (4):³⁸

and slope of plot E_p vs. pH is given by:where m is the number of protons involved in the oxidation, n′ is number of electrons transferred before the rate-determining step and βn′ + 1 is the transfer coefficient of the rate-determining step. Therefore, in addition to number of electrons transferred before the rate-determining step and number of protons, the transfer coefficient has also influence on slope of plot E_p vs. pH. At pH values in the range 5.0 to 9.6, the experimental measurement resulted in a 0.087 V pH⁻¹ shift that was almost similar to that shown in previous electrochemical investigations³⁹ and it is suggested that the number of electrons and protons are not equal in the oxidation of EP. When the pH of the solution was lower than the pK_a(<9.6), EP is in the protonated form (EPH⁺) and the number of hydrogen ions and electrons taking part in the electrode reaction was estimated as 3 and 2, respectively. But at pH values > 9.6 EP is in the natural form and slope was found the 0.031 V that indicate a change in number of electrons and protons, most likely involving potential determining role of the deprotonation of the EP after the abstraction of the electron.

The reaction mechanism for the oxidation of EP is as given in Scheme 2.^39,40

Scheme 2 Suggested oxidation mechanism for EP.

Effect of scan rate

The influences of scan rates on the peak current of FST-MIP/CPE towards oxidation of EP were studied by CV in the range of 25–150 mV s⁻¹ in order to investigate the electron-transfer process of the selected molecules on the electrode. As can be seen in Fig. 5, the anodic peak currents increased with the increasing of scan rate, and it presented a good linear relation between the anodic currents (I) and the scan rates (ν). The linear equation was shown as follows:

I (μA) = 0.092v + 0.22 (R² = 0.996) (for EP oxidation) (6)

Fig. 5 Cyclic voltammograms for FST-MIP/CPE in B–R buffer of pH 10.5 containing 2.5 μmol L⁻¹ of EP with scan rates ranging from 25, 50, 75, 100, 125 and 150 mV s⁻¹, and insets show the linear relationship of the anodic peak current versus scan rate.

suggested that the oxidation of EP was an adsorption-controlled irreversible process.

Influence of FST-MIP amount and incubation time

The amount of the FST-MIP used for modifying the CPE is a critical factor in the extraction of EP onto surface of fabricated sensor. Therefore, different quantities of FST-MIP content were embedded into the carbon paste matrix to fabricate the FST-MIP sensors for EP. The results showed, upon increasing the MIP content in the carbon paste, the oxidation peak current increased and reached to a maximum at 15% (w/w) of MIP in the electrode composition. It is believed that at high weight ratios of FST-MIP to graphite, the recognition sites on the electrode surface is increased and enhanced the electrode response. However, at higher weight ratios than 15%, the oxidation current response decreased; most probably due to increase in nonconductive MIP content and decrease the conductivity and the electron transfer capability at the electrode surface. Thus, the graphite : FST-MIP : binder weight ratio of 60 : 15 : 25 was chosen as the best composition for the developed MIP sensor performance.

The sensitivity of the proposed method was undoubtedly improved by the incubation time. The effects of incubation period on the EP extraction were evaluated and found that with increasing of incubation period up to 6 min, the voltammetric response increased and longer incubation times did not affect considerably the EP extraction. Therefore, this value was chosen for subsequent experiments.

Analytical performance

The FST-MIP/CPE was used for the assessment of the analytical utility of the proposed method as a means for the analysis of trace amounts of EP. The DPV method using the suggested sensor were applied as a method featuring high sensitivity, very low detection limit and good selectivity in the presence of other compounds in B–R buffer solution with pH 10.5. Quantitative evaluation is based on the linear correlation between the peak current and concentration. Linear calibration curves were obtained for EP in the range of 0.0090–2.8 μmol L⁻¹ with a regression equation of I_p (μA) = 8.967C (μmol L⁻¹) + 0.016 and r = 0.9993 (Fig. 6). The detection limits (based on 3S_b/m) of the procedures were found to be 0.0036 μmol L⁻¹.

Fig. 6 Differential pulse voltammograms of different concentrations of EP in B–R (pH = 10.5) and under optimum conditions. Inset shows the plot of peak current as a function of EP concentration in the ranges of 0.0090–2.8 μmol L⁻¹.

Interference study

In order to investigate the selectivity of the prepared electrode for the determination of EP, several species were checked as potential interfering species in their analysis. The potentially interfering substances were chosen from the group of substances commonly found with EP in the pharmaceuticals and/or in the biological fluids. The tolerance limit was taken as the maximum concentration of foreign species that caused a relative error of approximately ±5% for the determination of 0.3 μmol L⁻¹ each of EP plus the potential interfering substances at pH 10.5. The interference study was conducted by placing the modified carbon paste into a solution containing target analytes at optimum conditions. It was found that 600-fold of K⁺, Na⁺, NH⁴⁺, Mg²⁺, Ca²⁺, SCN⁻, ClO₄⁻, NO₃⁻, SCN⁻ have no influence on the signal of 0.3 μmol L⁻¹ of E_p. Glucose, lactose and sucrose showed no changes in the signals until to 500-fold excess was used. In addition, ascorbic acid, dopamine and alanine also did not interfere until a 400-fold excess was achieved. 18-fold excess of norephedrine as an analogous of EP has effect on EP signal. Like EP, this molecule poses a hydroxyl group and an amino group located at the same position in the aliphatic substituent as EP thus explaining its interference effect. These results suggested that the determination of EP in pharmaceutical formulations and biological samples at an FST-MIP/CPE was not significantly affected by the most common interfering species.

Determination of EP in real samples

In real samples analysis, FST-MIP/CPE was utilized to detect EP content in commercial pharmaceutical samples and also in the biological fluids. The results can be seen in Table 1 indicated that the fabricated sensor has good efficiency for the determination of EP in with satisfactory results. The recoveries were determined by spiking the samples with a certain amount of standard solutions of EP, and the results were found to be 97.5–102%. So this method could be applied to EP detection in biological samples and pharmaceutical samples with satisfactory results.

Table 1 Results for EP determination (μmol L⁻¹) in various real samples obtained by the proposed method under the optimum conditions

Sample	Added (μmol L^−1)	Found (μmol L^−1)	Recovery (%)	HPLC
Human serum
Sample 1	0.05	0.048	96.0	0.048
	1.00	0.990	99.0	1.01
	2.00	1.950	97.5	1.97
Sample 2	0.00	0.26	—	0.24
	0.4	0.67	102.5	0.67
	0.8	1.03	96.2	1.05
Urine
Sample 1	0.100	0.097	97.0	0.098
	1.50	1.53	102.0	1.46
	2.50	2.45	98.0	2.45
Sample 2	0.00	0.43	—	0.41
	0.30	0.72	96.6	0.73
	0.60	1.02	98.3	1.01
Tablet
	0.00	0.830	—	0.84
	0.06	0.892	103.3	0.888
	0.10	0.934	104.0	0.933
	1.00	1.84	101.0	1.81

---

Repeatability, reproducibility and stability

To evaluate the repeatability of the FST-MIP/CPE, the net response of the sensor before and after incubation in 0.10 and 0.25 μmol L⁻¹ EP solutions were measured with seven replicates. The relative standard deviations (RSDs) were 1.4% and 1.2% respectively for the seven successive assays. In the case of reproducibility of electrodes, six electrodes were prepared in a completely same manner. Then, the reproducibility was performed in the determination of 0.10 and 0.25 μmol L⁻¹ EP solutions, and the RSDs of six tests were less than 1.6%. Furthermore, the storage stability of the sensor was investigated. The results showed that the sensor lost only 6.0% of its initial response after it was stored in laboratory for 21 days. Therefore, the FST-MIP/CPE has good repeatability, reproducibility and stability.

Conclusion

We have developed a simple but effective MIP sensor for electrochemical determination of EP. Fe₃O₄@SiO₂@TiO₂ nanocomposite was introduced into the MIP to improve the electrochemical signal and recognition capacity of the sensor for the detection of EP. This simple and selective sensor exhibits not only good sensitivity to EP but also has excellent reproducibility and stability. The sensor also has a wider linear range and lower limit to EP than others (Table 2) and shows fast binding kinetics to EP due to its high ratio of surface imprinted sites and large surface-to-volume ratios. This sensor has been used successfully to detect EP in the biological fluids and pharmaceutical samples, indicating that the sensor will be as a practical tool in clinical and toxicological laboratories.

Table 2 Comparison of some characteristics of the proposed electrode with previously reported

Electrode	Method	Linear range (μmol L^−1)	Detection limit (μmol L^−1)	Ref.
MIPE/PPY-GCE	CV	500–3000	Not given	5
GCE	DPV	34.7–99.2	5	41
Poly(AHNSA)/GCE	SWV	8–1000	0.79	42
C18 bonded silica gel/CPE	ASV	0.37–9.9	0.46	43
MWCNT-GCE	DPV	1–100	0.75	44
CoPC/CPE	Amperometry	1–100	0.8	45
Carbon fiber	SWV	62–438	Not given	46
FST-MIP/CPE	DPV	0.009–2.8	0.0036	This work

---

Acknowledgements

The authors wish to thank the Researches and Technology Council, Baqiyatallah University of Medical Sciences, Tehran, for supports.

References

1. K. Parfitt, S. Sweetman, P. Blake and A. Parsons, Pharmaceutical, London, 32rd edn, 1999, p. 585.
2. N. T. Abdel-Ghani, M. S. Rizk and M. Mostafa, Spectrochim. Acta, Part A, 2013, 111, 131–141.
3. S. S. Hassan, A. H. Kamel and H. A. El-Naby, Talanta, 2013, 103, 330–336.
4. International Olympic Committee Medical Code, World Anti-Doping Agency, website:/http://www.wada-ama.org.
5. E. Mazzotta, R. Picca, C. Malitesta, S. Piletsky and E. Piletska, Biosens. Bioelectron., 2008, 23, 1152–1156.
6. D. L. Deng, J. Y. Zhang, C. Chen, X. L. Hou, Y. Y. Su and L. Wu, J. Chromatogr. A, 2012, 1219, 195–200.
7. H. Bagheri, H. Khoshsafar, S. Amidi and Y. Hosseinzadeh Ardakani, Anal. Methods, 2016, 8, 3383–3390.
8. L. Zhao, F. Zhao and B. Zeng, Sens. Actuators, B, 2013, 176, 818–824.
9. H. da Silva, J. Pacheco, J. Silva, S. Viswanathan and C. Delerue-Matos, Sens. Actuators, B, 2015, 219, 301–307.
10. H. Bagheri, A. Shirzadmehr and M. Rezaei, J. Mol. Liq., 2015, 212, 96–102.
11. B. Rezaei, O. Rahmanian and A. A. Ensafi, Sens. Actuators, B, 2014, 196, 539–545.
12. S. Patra, E. Roy, R. Das, P. Karfa, S. Kumar and R. Madhuri, Biosens. Bioelectron., 2015, 73, 234–244.
13. D. Y. Li, Y. P. Qin, H. Y. Li, X. W. He, W. Y. Li and Y. K. Zhang, Biosens. Bioelectron., 2015, 66, 224–230.
14. S. Patra, E. Roy, P. Karfa, S. Kumar, R. Madhuri and P. K. Sharma, ACS Appl. Mater. Interfaces, 2015, 7, 9235–9246.
15. Y. Lv, T. Tan and F. Svec, Biotechnol. Adv., 2013, 31, 1172–1186.
16. S. Patra, E. Roy, R. Madhuri and P. K. Sharma, Biosens. Bioelectron., 2015, 63, 301–310.
17. S. Patra, E. Roy, R. Madhuri and P. K. Sharma, Biosens. Bioelectron., 2015, 66, 1–10.
18. E. Roy, S. Patra, D. Kumar, R. Madhuri and P. K. Sharma, Biosens. Bioelectron., 2015, 68, 726–735.
19. X. Ren and L. Chen, Biosens. Bioelectron., 2015, 64, 182–188.
20. T. Madrakian, A. Afkhami, H. Mahmood-Kashani and M. Ahmadi, Talanta, 2013, 105, 255–261.
21. H. Bagheri, A. A. Asgharinezhad and H. Ebrahimzadeh, Food Analytical Methods, 2016, 9, 876–888.
22. A. Afkhami, A. Shirzadmehr, T. Madrakian and H. Bagheri, J. Mol. Liq., 2014, 199, 108–114.
23. N. Li, X. Lang, W. Ma, H. Ji, C. Chen and J. Zhao, Chem. Commun., 2013, 49, 5034–5036.
24. X. Lu, M. Yu, G. Wang, T. Zhai, S. Xie and Y. Ling, et al., Adv. Mater., 2013, 25, 267–272.
25. V. Etacheri, J. E. Yourey and B. M. Bartlett, ACS Nano, 2014, 8, 1491–1499.
26. A. K. Chandiran, M. Abdi-Jalebi, M. K. Nazeeruddin and M. Grätzel, ACS Nano, 2014, 8, 2261–2268.
27. L. Yang, H. Zhao, S. Fan, B. Li and C. P. Li, Anal. Chim. Acta, 2014, 852, 28–36.
28. Z. Chu, L. Shi, Y. Liu, W. Jin and N. Xu, Biosens. Bioelectron., 2013, 47, 329–334.
29. X. Liu, J. Zhang, S. Liu, Q. Zhang, X. Liu and D. K. Wong, Anal. Chem., 2013, 85, 4350–4356.
30. R. J. Ansell, J. K. Kuah, D. Wang, C. E. Jackson, K. D. Bartle and A. A. Clifford, J. Chromatogr. A, 2012, 1264, 117–123.
31. H. Bagheri, A. Afkhami, M. Saber-Tehrani and H. Khoshsafar, Talanta, 2012, 97, 87–95.
32. W. Zhang, J. Li, Z. Lu, Y. Luo, F. Chen and P. Huo, et al., Desalin. Water Treat., 2013, 52, 6598–6610.
33. Z. Lu, F. Chen, M. He, M. Song, Z. Ma and W. Shi, et al., Chem. Eng. J., 2014, 249, 15–26.
34. H. Bagheri, S. M. Arab, H. Khoshsafar and A. Afkhami, New J. Chem., 2015, 39, 3875–3881.
35. S. Xu, W. Shangguan, J. Yuan, M. Chen, J. Shi and Z. Jiang, Nanotechnology, 2008, 19, 095606.
36. S.-H. Wu, J.-L. Wu, S.-Y. Jia, Q.-W. Chang, H.-T. Ren and Y. Liu, Appl. Surf. Sci., 2013, 287, 389–396.
37. A. J. Bard and L. R. Faulkner, Electrochemical Methods, Fundamentals and Applications, John Wiley & Sons, Inc., 1980, pp. 213–248.
38. Q. Lin, Q. Li, C. Batchelor-McAuley and R. G. Compton, J. Phys. Chem. C, 2015, 119, 1489–1495.
39. A. Gazaliev, S. Fazylov, É. Sim and M. Z. Zhurinov, Chem. Nat. Compd., 1989, 25, 447–450.
40. M. Masui and S. Ozaki, Chemical and Pharmaceutical Bulletin, 1978, vol. 26, pp. 2153–2159.
41. M. Chicharro, A. Zapardiel, E. Bermejo, J. Perez and L. Hernandez, Anal. Chim. Acta, 1993, 273, 361–368.
42. M. Amare, W. Lakew and S. Admassie, Anal. Bioanal. Electrochem., 2011, 3, 365–378.
43. M. Chicharro, A. Zapardiel, l. E. Bermejo, J. Perez and L. Hernandez, Anal. Lett., 1994, 27, 1809–1831.
44. H. Ahmar and A. R. Fakhari, Anal. Methods, 2012, 4, 812–818.
45. E. G. Cookeas and C. E. Efstathiou, Analyst, 2000, 125, 1147–1150.
46. M. Platts, R. B. Smith, N. Mould and J. Davis, Electrochem. Commun., 2006, 8, 633–637.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra09488k

---