Construction of novel sensitive electrochemical sensor for electro-oxidation and determination of citalopram based on zinc oxide nanoparticles and multi-walled carbon nanotubes

Materials Science and Engineering C 59 (2016) 847–854

Contents lists available at ScienceDirect

Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec

Construction of novel sensitive electrochemical sensor for electro-oxidation and determination of citalopram based on zinc oxide nanoparticles and multi-walled carbon nanotubes Hamed Ghaedi a,⁎, Abbas Afkhami b, Tayyebeh Madrakian b, Farzaneh Soltani-Felehgari b a b

Faculty of Engineering, Bushehr Branch, Islamic Azad University, Bushehr, Iran Faculty of Chemistry, Bu-Ali Sina University, Hamedan, Iran

a r t i c l e

i n f o

Article history: Received 15 April 2015 Received in revised form 20 September 2015 Accepted 29 October 2015 Available online 30 October 2015 Keywords: Chemically modified electrode Citalopram determination Electrochemical sensor Square wave voltammetry Human serum Pharmaceutical preparations

a b s t r a c t A new chemically modified carbon paste electrode (CMCPE) was applied to the simple, rapid, highly selective and sensitive determination of citalopram in human serum and pharmaceutical preparations using adsorptive square wave voltammetry (ASWV). The ZnO nanoparticles and multi-walled carbon nanotubes modified CPE (ZnO– MWCNT/CPE) electrode was prepared by incorporation of the ZnO nanoparticles and multi-walled carbon nanotubes (MWCNT) in carbon paste electrode. The limit of detection and the linear range were found to be 0.005 and 0.012 to 1.54 μmol L−1 of citalopram, respectively. The effects of potentially interfering substances on the determination of this compound were investigated and found that the electrode is highly selective. The proposed CMCPE was used to the determination of citalopram in human serum, urine and pharmaceutical samples. This reveals that ZnO–MWCNT/CPE shows excellent analytical performance for the determination of citalopram in terms of very low detection limit, high sensitivity, very good repeatability and reproducibility over other methods reported in the literature. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Analysis of drugs in pharmaceutical products and biological samples is growing in importance, both in the development of more selective and effective drugs and in understanding their therapeutic and toxic effects [1–4]. Knowledge of drug levels in body fluids, such as serum, allows the optimization of pharmacotherapy and provides a basis for studies of patient compliance, bioavailability, pharmacokinetics, and the influence of co-medications. The treatment of depressive disorders is usually accomplished with antidepressants [5,6]. Therapeutic drug monitoring is underutilized in the field of psychiatry because the therapeutic ranges of antidepressants seem quite broad, leading to the generally accepted notion of low toxicity [7]. Citalopram (CIT, Scheme 1) is an antidepressant drug, which selectively potentiates serotonin neurotransmission by inhibiting serotonin reuptake. It offers similar therapeutic efficacy and a more favorable tolerability profile than the tricyclic antidepressants [8]. Citalopram is a racemic mixture, its pharmacological effect resides mainly in the S-(1) enantiomer and, to a lesser degree, in the S-(1)-desmethyl citalopram [9]. Nevertheless, for bioequivalence studies the recent guidance [10] recommends measurement of the race-mate using an achiral assay. ⁎ Corresponding author. E-mail address: [email protected] (H. Ghaedi).

http://dx.doi.org/10.1016/j.msec.2015.10.088 0928-4931/© 2016 Elsevier B.V. All rights reserved.

The advantages of the therapeutic profile of one important antidepressant, citalopram [11–13], have led to its increasing use in the treatment of depression. However, even if this compound has fewer undesirable side effects, it can lead to major intoxications [11]. Therefore, the development of a rapid and specific method allowing the screening and the determination of this new antidepressant drug in biological fluids could be of great interest either in therapeutic drug monitoring use or in toxicological screening in the case of a suicide involving this compound [14]. Analytical methods for the detection of anti-depressants are not only of interest in the field of clinical toxicology, but also in forensics because they are often involved in intoxications [15–19]. The methods described for the determination of citalopram in biological samples involve gas chromatography and high performance liquid chromatography [20,21]. One of the important steps in an analytical method is the extraction of the compounds of interest from the sample matrix. Liquid–liquid extraction (LLE) and solid-phase extraction (SPE) are the most common techniques for isolation and enrichment of citalopram prior to chromatographic analysis [22–29]. These methods have many disadvantages, as they are tedious, laborintensive and time-consuming. Formation of emulsion is a major drawback during LLE process. This method also requires the use of large amount of highly purified solvents, which are often hazardous and lead to the production of toxic laboratory waste. However SPE needs a large volume of samples to obtain high enrichment factors. This could

848

H. Ghaedi et al. / Materials Science and Engineering C 59 (2016) 847–854

2. Experimental 2.1. Apparatus and chemicals

Scheme 1. Structure of CIT.

be a major limitation especially when biological samples are being analyzed. Therefore development of simple, fast and sensitive methods was necessary. Over the past five decades, carbon paste, i.e. a mixture of carbon (graphite) powder and a binder (pasting liquid), has become one of the most popular electrode materials used for the laboratory preparation of various electrodes, sensors, and detectors [30–32]. In recent years, to improve the sensitivity, selectivity, detection limit and other features of CPE, chemically modified carbon paste electrodes (CMCPEs) have been used. For example, within the electrode structure various materials such as chelating agents, ion exchangers or functionalized nanoparticles have been used as modifiers [33–37]. The operation mechanism of such CMCPEs depends on the properties of the modifier materials used to import selectivity and sensitivity towards the target species. Initially, non-conductive reagents, such as mineral oil or paraffin oil were used as binders [30]. Nowadays, due to the unique properties of carbon nanotubes (CNTs) such as ordered structure with high aspect ratio, ultra-light weight, high mechanical strength, high electrical conductivity, high thermal conductivity, metallic or semi-metallic behavior and high surface area, they have been widely used for the development of chemically modified electrodes [38–40]. The combination of these characteristics makes CNTs unique materials with the capability to promote electron transfer reaction and improve sensitivity in electrochemistry, and thus they are widely used to prepare modified electrodes. Moreover, extensive efforts have been devoted to design novel CNTs modified electrodes to improve the voltammetric determinations of organic [40–42] and inorganic compounds [38,43]. In comparison to the conventional CPEs, the carbon nanotube paste electrodes have shown a considerable enhancement in electrochemical signals leading to improvement of the detection limit in the voltammetric measurements. In the present work, we investigated the performance of a ZnO and multi-walled carbon nanotube modified CPE using a room temperature ionic liquid (RTIL) as the binder (ZnO–MWCNT/CPEIL) as the working electrode for the determination of trace amounts of CIT by adsorptive square wave voltammetry (ASWV). The created selectivity in this method makes the electrode very suitable for the detection of trace amounts of CIT in various real samples. To the best of our knowledge, this electrode represents the better limit of detection and selectivity rather than other reported methods for electro-determination of CIT [44,45]. Also, in this paper, harmful substances, such as mercury, were not used to construct the electrode for the termination of this compound. The prepared modified electrode was successfully applied to the voltammetric determination of CIT in pharmaceutical and clinical samples. To the best of our knowledge this is the first report on the investigation of the electrochemical behavior of CIT at the modified CPE surface. The combination of the above-mentioned characteristics makes modified CPEs as a unique sensor with the potential for the diverse applications.

All electrochemical experiments including cyclic voltammetry (CV), square wave voltammetry (SWV) and other methods were performed using an Autolab Potentiostate/Galvanostate, Model 302 N. A conventional three-electrode system was used with a carbon paste working electrode (unmodified or modified), a saturated Ag/AgCl reference electrode and a Pt wire as the counter electrode. A magnetic stirrer (PAR305) with a Teflon-coated magnet was used to provide the convective transport during the preconcentration step. The whole measurements were automated and controlled through the programming capacity of the apparatus. A pH-meter, Model 713, with a glass electrode (Metrohm, Swiss), was used to determine pH values of the solutions. The pharmaceutical samples (Citalopram hexal tablets from Hexal, Germany) were obtained from local drug stores. All the chemicals were of analytical grade, or better purchased, and used as received. Unless otherwise stated, all the solutions were prepared with doubly distilled water (DDW). MWCNT (purity more than 95%) with outer diameter between 5 and 20 nm, inner diameter between 2 and 6 nm and tube length from 1 to 10 μm was purchased from Plasmachem GmbH (Germany). Paraffin oil and graphite powder (mesh size b 100 μm) were obtained from Merck Company and used as received and n-octylpyridinum hexa-fluoro phosphate (OPFP) was purchased and used from Hangzhou Kemer Chemical Limited Company. Britton–Robinson (B–R) buffer was prepared in DDW and was used as supporting electrolyte. pH adjustments were performed with 0.01–1.0 mol L−1 HCl or NaOH solutions. 2.2. Pretreatment of multi-walled carbon nanotube materials A pre-treatment of the CNTs is usually necessary to eliminate graphitic nanoparticles, amorphous carbon, metallic impurities, and/or to improve the electron transfer properties and/or to allow further functionalization [46]. The pre-treatment consists of exposing the MCNTs to an acidic solution of sulfuric, nitric or hydrochloric acid, or mixture of these acids at room temperature, under refluxing or under sonication for different times [46,47]. Following one of the purification methodologies, 500 mg of MWCNT was heated at 400 °C using an air flow of 12 mL min−1 (quartz tubular reactor of 14 mm diameter), for 1 h. To eliminate metal oxide catalysts, the heated processed amount of MWCNTs was dispersed in 60 mL of 6.0 mol L−1HCl for 4 h under ultrasonic agitation; filtered on a Whatman No. 42 filter paper and washed until the pH of the solution was neutral; and finally, dried. 2.3. Preparation of modified carbon paste electrode The modified carbon paste electrode was prepared by mixing 75% (w/w) ZnO/MWCNT/graphite powder (8:14:53) with 25% (w/w) n-octyl-pyridinum hexa-fluoro phosphate (OPFP, as the binder) in a mortar and pestle. The mixture amount of 0.20 g was homogenized in a mortar for 30 min and the resulting composite was dispersed in dichloromethane (for more homogeneity of the electrode composite components leading to an increase in the reproducibility after each electrode surface polishing). The homogenized composite was stirred by a magnetic stirrer till the solvent evaporated completely. Then, the prepared modified composite was air dried for 24 h. Finally the homogenized paste was then inserted into a plastic needle-type capillary tube with a 1.5 mm diameter and a 5 cm length, using a 0.5 mm diameter copper wire connected to the measurement system. 2.4. Analytical procedure The analysis of CIT using ASWV was carried out in a 25.0 mL aliquot (pH 5.0) using the following steps after purging with nitrogen for at

H. Ghaedi et al. / Materials Science and Engineering C 59 (2016) 847–854

849

least 5 min: (a) pre-conditioning step was performed by applying a potential of 1.00 V vs. Ag/AgCl for 30 s before each measurement to ensure dissolution of the remaining deposits on the surface of the modified electrode; (b) the preconcentration step proceeded at 0.800 V vs. Ag/AgCl for 120 s; at the end of the preconcentration time, stirring was stopped and a 10 s rest period was allowed for the solution to become quiescent; (c) the adsorptive square wave voltammograms were recorded when the potential was swept from 0.500 to 1.450 V vs. Ag/AgCl. In the pre-conditioning and preconcentration processes, the detection solutions were stirred with a magnetic stirrer. The peak currents at about 0.900 V vs. Ag/AgCl for CIT were measured. All measurements were carried out at room temperature (25.0 ± 0.1 °C). Calibration graphs were prepared for the peak current against CIT concentrations. Calibration graphs were prepared by plotting the net anodic peak currents vs. CIT concentration in solution. 2.5. Analysis of serum and urine samples Drug free human blood and/or urine samples, obtained from healthy volunteers were centrifuged at 3000 rpm for 30 min at room temperature and separated serum samples were stored frozen until assay. Acetonitrile removes serum proteins more effectively, the addition of 1–1.5 vol of serum or urine is sufficient to remove the proteins. After vortexing for 30 s, the mixture was then centrifuged for 10 min at 3000 rpm to separate urine and serum protein residues and the supernatant was taken carefully. Appropriate volumes of this supernatant were transferred into the 50.0 mL volumetric flask and then, 1.0 mL of stock solution of CIT was added and diluted up to the volume with B–R buffer at pH 5.0. Appropriate volumes from this liquor were spiked to electrochemical cell containing B–R solution (the final volume of the solution in electrochemical cell was 25 mL). For pharmaceutical samples, a tablet was powdered and homogenized and then about 10.0 mg of homogenized powder was accurately weighed and dissolved in 50 mL of distilled water. Finally, appropriate volumes from these solutions were transferred to electrochemical cell containing B–R solution (the final volume of the solution in electrochemical cell was 25 mL) and then the electrochemical signal was determined. 3. Results and discussion 3.1. Electrochemical characterization of the ZnO–MWCNT/CPEIL To clarify the differences among the electrochemical performance of the CPE, carbon paste electrode using RTIL as the binder (CPE IL ), MWCNTs modified CPE IL (MWCNT/CPE IL ) and ZnO and MWCNTs modified CPE IL (ZnO–MWCNT/CPE IL) , electrochemical impedance spectroscopy (EIS) was used as a procedure for the characterization of each electrode surface. As such, the Nyquist plots for 1.0 × 10−3 mol L−1K3[Fe(CN)6] in 1.0 mol L−1 KNO3 showed a significant difference in responses for the electrodes (Fig. 1). The semicircular elements correspond to the charge transfer resistances (Rct) at the electrode surface with a large diameter was observed for the bare CPE in the frequency range 10−1–106 Hz. However, the diameter of the semicircle diminished when CPEIL and MWCNT/CPEIL were employed. Additionally, the charge transfer resistance (Rct) values obtained from this observation implied that the charge transfer resistance of the electrode surface decreased and the charge transfer rate increased upon using ZnO–MWCNT/CPEIL. A Warburg at 45° was also observed for all the electrodes of interest. The Rct value for the ZnO–MWCNT/CPEIL electrode was less than that for CPE, CPEIL and MWCNT/CPEIL. Comparison of the Rct obtained for different electrodes indicates that the modification of electrodes with MWCNTs and ZnO and using RTIL as the binder instead of paraffin caused easier and faster charge transfer at the electrode surface. ZnO nanostructures have great advantages, such as high surface area, nontoxicity, good environmental

Fig. 1. Impedance plots of the CPE (a), CPEIL (b), MWCNT/CPEIL (c) and ZnO–MWCNT/ CPEIL (d) at optimum composition, for 1.0 × 10−3 mol L−1 K3[Fe(CN)6] in 1.0 mol L−1 KNO3.

acceptability, inexpensive, electrochemical activity and high electron communication features [48,49]. At the same time RTIL provides a biocompatible interface with high electron transfer rate and inherent catalytic ability [48,49]. Also RTIL had exhibited the advantages such as resistivity towards electrode fouling, high rates of electron transfer and the inherent catalytic activity [50]. Therefore the ZnO–MWCNT/ CPEIL was chosen as the optimum composition of the modified electrode for further studies. As it is shown in SEM image for modified electrode (Fig. 2), at the surface of modified CPE the layer of irregular flakes of graphite powder, MWCNTs and ZnO nanoparticles are observed. The area of the electrode was obtained by the cyclic voltammetric method using 1.0 mmol L−1 K3Fe(CN)6 at different scan rates. For a reversible process, the following Randles–Sevcik formula can be used [51].   Ipa ¼ 2:69  105 n3=2 AC0 D1=2 ν1=2

where, Ipa refers to the anodic peak current, n is the number of electrons transferred, A is the surface area of the electrode, D is the diffusion coefficient, υ is the scan rate and C0 is the concentration of K3Fe(CN)6. For 1.0 mmol L−1 K3Fe(CN)6 in 0.1 mol L−1 KCl electrolyte, n = 1 and D = 7.6 × 10−6 cm2 s−1. Then from the slope of the plot of Ipa versus υ1/2 relation, the electroactive area was calculated. In this case, the slope was obtained as 1.3 × 10−5 and 1.9 × 10−5 μA (Vs−1)1/2 and the area of electrode was obtained to be 1.75 × 10− 2 cm2 and 2.56 × 10−2 cm2 for MWCNT/CPE and ZnO–MWCNT/CPE, respectively which had shown that the presence of ZnO nanoparticles and MWCNTs

Fig. 2. SEM image for modified ZnO–MWCNT/CPE electrode.

850

H. Ghaedi et al. / Materials Science and Engineering C 59 (2016) 847–854

3.3. Optimization of variables

Fig. 3. Cyclic voltammograms for 0.61 μmol L−1 CIT in B–R buffer solution of pH 5.0 obtained in different electrodes. Conditions: pH 5.0; scan rate 80 mV s−1, at (a) CPEIL, (b) MWCNT/CPEIL, (c) ZnO–MWCNT/CPEIL surface.

was increasing the surface area of the modified electrode much more than MWCNT.

3.2. Preliminary investigations The electrooxidation of CIT in B–R solution of pH 5.0 was studied at the surface of CPEIL, MWCNT-CPEIL and ZnO–MWCNT/CPEIL using the cyclic voltammetric technique in the potential range 0.300 to 1.550 V vs. Ag/AgCl. The CVs for 0.61 μmol L−1 CIT solution were recorded at pH 5.0 after a 120 s at 0.800 V vs. Ag/AgCl accumulation step using different electrodes (scan rate 80 mV s−1). The resulting voltammograms (Fig. 3) show an electrochemically irreversible system. The CVs CPEIL in 0.61 μmol L−1 CIT showed a weak S-shape anodic peak in the positive going scan at about 1.100 V vs. Ag/AgCl. The CV for modified MWCNT/ CPEIL in the presence of CIT showed an anodic peak at about 1.030 V vs. Ag/AgCl. But at ZnO–MWCNT/CPEIL a sharp anodic peak at about 0.870 V vs Ag/AgCl was observed. For all the electrodes, the cathodic counterpart of the anodic peak was not observed. The enhancement in peak current and improving in the shape of ZnO–MWCNT-CPE in comparison to other tested electrodes indicates the role of nano-materials (ZnO and MWCNT) as the suitable elements in the construction of a sensor for determination of the CIT.

3.3.1. Carbon paste electrode composition Some important features of the carbon paste electrodes, such as the properties of the binder, the binder/graphite powder ratio and especially, the nature and amount of the nano-materials used have significant influence on the sensitivity and selectivity of the electrodes. Preliminary studies using cyclic and square wave voltammetry, described above, showed the effect of ZnO nanoparticles and MWCNTs in improving the catalytic activity of the electrode composition in the electrochemical oxidation of CIT. As described above, by introducing ZnO nanoparticles and MWCNTs to the carbon paste electrode, the surface area and the electron transfer rate increased and consequently the electrochemical response of CIT was enhanced. Effect of the amount of ZnO nanoparticles and MWCNTs on the voltammetric response was also studied (Fig. 4). The peak current increased by increasing the amount of MWCNTs and ZnO nanoparticles up to 14 and 8%, respectively. The increase in the anodic peak current, by increasing modifiers, can be due to the larger microscopic area of the modified electrode and also enhanced accumulation efficiency. On the other hand, enhancement of the nano-materials (ZnO and MWCNTs) amount more than the optimum value, leads to a level off in the electrode response. 3.3.2. Influence of scan rate To understand the reaction mechanism, the effect of scan rate on the peak currents of CIT at different scan rates from 5 to 300 mV s−1 (Fig. 5) was investigated. As shown in Fig. 5, the oxidation peak current of CIT was linearly proportional with the scan rate in the range 5–300 mV s−1 in the B–R buffer solution of pH 5.0, with the equation of Ipa (μA) = 0.482v (mV s−1) + 0.6923 (r2 = 0.9969). This indicates that the electrode process was controlled by adsorption rather than diffusion. The Ep of the oxidation peak was dependent on the scan rate. The plot of Ep versus log υ was linear having a correlation coefficient of and the relation between peak potential and log υ can be communicated by the equation, Ep = 0.1207log υ + 0.8388 (0.986), where Ep and υ are peak potential in V and scan rate in V s− 1, respectively. For an adsorption-controlled and irreversible electrode process, according to Laviron [52], Ep is defined by the following equation, 0

Ep ¼ E0 þ

0

2:303RT RTk 2:303RT log logυ þ αnF αnF αnF

where α is the transfer coefficient, k0 is the standard heterogeneous rate constant of the reaction, n is the number of electrons transferred, υ is

Fig. 4. Effect of the amount of modifiers (A) ZnO and (B) MWCNT on the stripping peak current for 0.30 μmol L−1 of CIT solution. Condition: pH 5.0; deposition potential, 0.80 V vs. Ag/AgCl; deposition time, 120 s; resting time, 10 s; SW frequency, 60 Hz; pulse amplitude, 55.0 mV.

H. Ghaedi et al. / Materials Science and Engineering C 59 (2016) 847–854

851

Fig. 5. Dependence of the peak current on scan rate, Condition: pH 5.0; scan rate between 5 and 300 mV s−1.

the scan rate and Eo′ is the formal redox potential. Other symbols have their common meanings. Thus, the value of αn can be easily calculated from the slope of Ep versus log υ. In this system, the slope was found to be 0.1207, taking T = 298 K and substituting the values of R and F, αn was calculated to be 0.49. The transfer coefficient for an irreversible process can be calculated from following equation [53,54]: α¼

47:7 mV Ep‐Ep=2

Where Ep/2 is the potential where the current is at half the peak value. From this, we obtained the value of α to be 0.247. So, the number of electron (n) transferred in the electro oxidation of CIT was calculated to be 1.98 ≈ 2. 3.3.3. Influences of pH The influence of the pH on the oxidation peak current of 0.15 μmol L−1, CIT, was studied using B–R buffer. It was observed that as pH of the medium was gradually increased, the potential kept on shifting towards less positive values, suggesting the involvement of proton in the oxidation reaction. Below pH 3.0, CIT does not give any oxidation peak. Over the pH range 3.0–8.0, the peak potential (Ep) of CIT is a linear function of pH. From the plot of Ep vs. pH, slope of −95.90 mV pH−1 was observed for CIT. The Ep has a linear relationship with pH of the buffer solution regarding the following equation:  2  Ep ðVÞ ¼ −0:0959pH þ 1:427 R 2 ¼ 0:999 This behavior is in agreement with a two electrons–three proton. It was observed that the peak current for CIT was maximum at pH 5.00

Fig. 7. Effect of deposition potential and time on the stripping peak current for 0.30 μmol L−1 of CIT solution. Condition: pH 5.0; deposition potential, 0.800 V vs. Ag/ AgCl; deposition time, 120 s; resting time, 10 s; SW frequency, 60 Hz; pulse amplitude, 55.0 mV.

(Fig. 6). Thus, this pH was employed for further studies. The oxidation reaction of CIT including proton elimination and therefore, in the strong acidic media the elimination of the H+ is very hard and therefore the extent in which the reaction proceeds decreases. At pHs higher than 5.0 the CIT may start to decompose which caused the a decrease in accumulation efficiency and the concentration of the CIT at the electrode surfaced, therefore the electrochemical signal decreased at pH higher than 5.0.

3.3.4. Effect of accumulation time and potential It was significant to fix the accumulation potential and accumulation time when adsorption studies were intended. Both conditions could affect the amount of adsorption of CIT at the electrode. Bearing this in mind, the effect of accumulation potential and time on ASWV signal was studied (Fig. 7). The concentration of CIT was 0.30 μmol L−1. When accumulation potential was varied from 0.000 to 0.800 V vs. Ag/AgCl, the peak current increased because the applied potential is near the oxidation potential of CIT and therefore the CIT molecules tend to acomulated at the electrode surface. But at potentials higher than 0.800 V vs. Ag/AgCl the peak current dramatically decreased due to the fact that by applying the potentials higher or about the oxidation potential of CIT, the drug molecoules do not have any time to accumulate at the surface. CIT molecules oxidized immediately by arriving to the electrode surface at this potential (the potential higher or about 0.900 V). Therefore the accumulation efficiency was decreased and the electrochemical signals were decreasing consequently. Hence, a potential of 0.800 V vs Ag/AgCl was applied as the accumulation potential. Also the influence of accumulation time ranging from 0 to 200 s on the oxidation of CIT at MIP-MWCNT/CPEIL was investigated (Fig. 7). The peak current increased gradually as the accumulation time increased from 0 to 120 s. However, with increase in accumulation time higher than 120 s, the peak current was not shown any significant change and tended to be almost stable. Therefore, the optimal accumulation time of 120 s was chosen for further investigations.

Table 1 Optimum values for the studied parameters.

Fig. 6. Effect of pH on the Ep and voltammetric response for 0.015 μmol L−1 of CIT. Condition: pH 5.0; deposition potential, 0.80 V vs. Ag/AgCl; deposition time, 120 s; resting time, 10 s; SW frequency, 60 Hz; pulse amplitude, 55.0 mV.

Parameter

Range studied

Accumulation potential (V vs. Ag/AgCl) Accumulation time (s) Pulse amplitude (mV) Voltage step (mV) Frequency (Hz)

0.00–1.10 0–200 10–150 1–10 10–100

Optimum value 0.80 120 55 4 60

852

H. Ghaedi et al. / Materials Science and Engineering C 59 (2016) 847–854

Fig. 8. Net square wave voltammograms of ZnO–MWCNT/CPEIL solutions with different concentrations of CIT. Conditions: pH 5.00; deposition potential, 0.800 V vs Ag/AgCl; deposition time, 120 s; resting time, 10 s; SW frequency, 60 Hz; pulse amplitude, 55.0 mV.

3.3.5. Other optimum conditions The optimum conditions for the ASWV signal of CIT were also recognized by measuring the current dependence on some instrumental parameters including, accumulation potential, accumulation time, modulation amplitude, modulation time and voltage step. These parameters were optimized for obtaining maximum signal-to-noise ratio. Optimum values for the studied parameters are given in Table 1. As can be seen in Table 1, deposition potential 0.80 V vs. Ag/AgCl; deposition time 120 s; resting time 10 s; SW frequency 60 Hz and pulse amplitude, 55.0 mV were chosen as the optimum conditions. 3.4. Analytical characterization Calibration graphs were constructed under the optimum conditions described above using ZnO–MWCNT/CPEIL. Fig. 8 shows the net square wave voltammograms for different concentrations of CIT obtained under the optimum conditions. As seen in Fig. 8a linear range of 0.012–1.54 μmol L−1 was observed with line equation of Ip (μA) = 35.70 CCIT(μmol L−1) + 1.19, (R2 = 0.9995). The limit of detection defined as LOD = 3Sb/m, where LOD, Sb and m are the limit of detection, standard deviation of the blank and the slope of the calibration graph, respectively, was found to be 0.005 μmol L−1. Sb was estimated by six replicate determinations of blank signals. The repeatability of the electrode in the determination of CIT was evaluated by performing six determinations with the same standard solutions of CIT. The relative standard deviation (RSD) for the response of electrode towards a 0.06 μmol L− 1 of CIT solution was 2.80%. The reproducibility of the response of the electrode was also studied. Six electrodes were prepared from the same batch and were evaluated by performing the determination of 0.06 μmol L−1 of CIT solution. The RSD for the responses between electrodes was 3.12%. The results show that the repeatability and reproducibility of the sensor for the determination of CIT are acceptable. The electrode was stable for 7 weeks. After that the response of the electrode decreased and the noise in the responses increased.

less than ± 5.0% at a concentration of 0.15 μmol L− 1 of CIT. Ascorbic acid, citric acid, uric acid, dopamine, glucose, caffeine, K+, Ca2+, NO− 3 , 2+ NH+ have no effect on the Ip of CIT up to 150-fold excess 4 and Mg (150:1, interferent:analyte mass ratio (50 ng mL−1 of CIT). This suggests that the determination of CIT in the pharmaceutical and biological samples at ZnO–MWCNT/CPEIL is not affected significantly by the common interfering species present along with the molecules of interest.

3.6. Analytical applications The modified sensor was successfully applied to the determination of CIT in human urine and serum and in tablet samples. CIT was also determined after the addition to the samples and recovery values were calculated. According to Table 2, the recovery values are between 96.20 and 102.80% which are acceptable. The satisfactory recovery for the human urine and serum and the good agreements between the obtained value and the tablet manufacturer's value indicate that the proposed method had great potential in the practical sample analysis (manufacturer's value was given as 20.00 mg tablet−1). Thus the sensor provides a good alternative for the determination of CIT in real samples. Also, comparison between the analytical characteristics of the present method and some previous reports for the determination of CIT is shown in Table 3. This reveals that ZnO–MWCNT/CPEIL shows excellent analytical performance for determination of CIT in terms of very low detection limit, high sensitivity, very good repeatability and reproducibility over other methods reported in the literature [44,45].

Table 2 Determination of CIT in human urine samples by the proposed method (N = 5). Sample

Spiked value (μmol mL−1)

Found (μmol mL−1)

Recovery (%)

Human serum

– 0.154 0.460 – 0.154 0.460 –

0.00 0.148 ± 0.002 0.450 ± 0.010 0.00 0.158 ± 0.006 0.471 ± 0.01 19.25 ± 0.17⁎

– 96.10 97.80 – 102.60 102.0.4 –

3.5. Interferences In order to evaluate the selectivity of the method for the determination of CIT, the influence of potentially interfering substances on the determination of this compound was investigated. The tolerance limit for interfering species was considered as the maximum concentration that gave a relative error

Human urine

Citalopram Hexal tablet (20 mg tablet−1) ⁎ In mg.

H. Ghaedi et al. / Materials Science and Engineering C 59 (2016) 847–854 Table 3 Comparison of some characteristics of the different electrodes for the determination of CIT. Electrodes⁎

Linear range (μmol L−1)

LOD (μmol L−1)

Ref

DME GCE ZnO–MWCNT/CPE

0.10–20 0.096–28.0 0.012–1.54

0.047 0.0078 0.005

[44] [45] This work

⁎ GCE; Glassy carbon electrode, DME; Dropping mercury electrode.

4. Conclusions Using new modified CPE to the fabrication of a novel sensor for the determination of CIT using adsorptive stripping square wave voltammetry was investigated. The sensitivity was enhanced considerably by preconcentration of CIT on the modified electrode surface and by increase in the electroactive surface area due to the presence of nanomatreials. Moreover, the proposed method is very sensitive having sub-micromolar detection limits for the determination of CIT. Finally, the sensor was applied to the determination of CIT in biological and pharmaceutical samples and the results were satisfactory, which suggests that ZnO–MWCNT/CPEIL can act as a powerful sensor for the determination of CIT in biological and pharmaceutical samples. Acknowledgments The authors acknowledge the Bu-Ali Sina University Research Council and Center of Excellence in Development of Environmentally Friendly Methods for Chemical Synthesis (CEDEFMCS) for providing support to this work. References [1] K.K. Akerman, J. Jolkkonen, H. Huttunen, I. Penttila, High-performance liquid chromatography method for analysing citalopram and desmethylcitalopram from human serum, Ther. Drug Monit. 20 (1998) 25–29. [2] G. Tournel, N. Houdret, V. Hedouin, M. Deveaux, D. Gosset, M. Lhermitte, High-performance liquid chromatographic method to screen and quantitate seven selective serotonin reuptake inhibitors in human serum, J. Chromatogr. B 761 (2002) 147–158. [3] O.V. Olesen, K. Linnet, Simplified high-performance liquid chromatographic method for the determination of citalopram and desmethylcitalopram in serum without interference from commonly used psychotropic drugs and their metabolites, J. Chromatog. B 675 (1996) 83–88. [4] H.Q. Meng, D. Gauthier, Simultaneous analysis of citalopram and desmethylcitalopram by liquid chromatography with fluorescence detection after solid-phase extraction, Clin. Biochem. 38 (2005) 282–285. [5] H. Bagheri, F. Khalilian, E. Babanezhad, A. Eshaghi, A.R. Rouini, Modified solvent microextraction with back extraction combined with liquid chromatography-fluorescence detection for the determination of citalopram in human plasma, Aanal. Chim. Acta 610 (2008) 211–216. [6] H. Juan, Z. Zhiling, L. Huande, Simultaneous determination of fluoxetine, citalopram, paroxetine, venlafaxine in plasma by high performance liquid chromatography– electrospray ionization mass spectrometry (HPLC–MS/ESI), J. Chromatogr. B 820 (2005) 33–39. [7] P. Baumann, F. Larsen, The pharmacokinetics of citalopram, Rev. Contemp. Pharmaco. 6 (1995) 287–295. [8] R.J. Milne, K.L. Goa, Citalopram. A review of its pharmacodynamic and pharmacokinetic properties, and therapeutic potential in depressive illness, Drugs 41 (1991) 450–477. [9] M. Kosel, C.B. Eap, M. Amey, P. Baumann, Analysis of the enantiomers of citalopram and its demethylated metabolites using chiral liquid chromatography, J. Chromatogr. B 719 (1998) 234–238. [10] Guidance for Industry: Bioavailability and Bioequivalence Studies for Orally Administered Drug Products — General Considerations, US Department of Health and Human Services, Food and Drug AdministrationCenter for Drug Evaluation and Research, Rockville, MD, 2000. [11] S.E. Mgller, F. Larsen, R.P.E. Pitsiq, Effect of citalopram on plasma levels of oral theophylline, Clin. Therap. 22 (2000) 1494–1501. [12] L. Dalgaard, C. Larsen, Metabolism and excretion of citalopram in man: identification of O-acyl-and N-glucuronides, Xenobiotica 29 (1999) 1033–1041. [13] M. Shipkova, E. Wieland, Glucuronidation in therapeutic drug monitoring, Clin. Chim. Acta 358 (2005) 2–23. [14] K.E. Goeringer, L. Raymon, G.D. Christian, B.K. Logan, Postmortem forensic toxicology of selective serotonin reuptake inhibitors: a review of pharmacology and report of 168 cases, J. Forensic Sci. 45 (2000) 633–648.

853

[15] A. de Meester, G. Carbutti, L. Gabriel, J.M. Jacques, Fatal overdose with trazodone: case report and literature review, Acta Clin. Belg. 56 (2001) 258–261. [16] T. Azaz-Livshits, A. Hershko, E. Ben-Chetrit, Paroxetine associated hepatotoxicity: a report of 3 cases and a review of the literature, Pharmacopsychiatry 35 (2002) 112–115. [17] K.E. Goeringer, I.M. McIntyre, O.H. Drummer, Postmortem tissue concentrations of venlafaxine, Forensic Sci. Int. 121 (2001) 70–75. [18] C.A. Kelly, N. Dhaum, W.J. Laing, F.E. Strachan, A.M. Good, D.N. Bateman, Comparative toxicity of citalopram and the newer antidepressants after overdose, J. Toxicol. Clin. Toxicol. 42 (2004) 67–71. [19] D.E. Adson, S. Erickson-Birkedahl, M. Kotlyar, An unusual presentation of sertraline and trazodone overdose, Ann. Pharmacother. 35 (2001) 1375–1377. [20] H.H. Maurer, J. Bickeboeller-Friedrich, Screening procedure for detection of antidepressants of the selective serotonin reuptake inhibitor type and their metabolites in urine as part of a modified systematic toxicological analysis procedure using gas chromatography–mass spectrometry, J. Anal. Toxicol. 24 (2000) 340–347. [21] E. Lacassie, J.M. Gaulier, P. Marquet, J.F. Rabatel, G. Lachatre, Methods for the determination of seven selective serotonin reuptake inhibitors and three active metabolites in human serum using high-performance liquid chromatography and gas chromatography, J. Chromatogr. B 742 (2000) 229–238. [22] J. Macek, P. Ptacek, J. Kl yma, Rapid determination of citalopram in human plasma by high-performance liquid chromatography, J. Chromatogr. B 755 (2001) 279–285. [23] C. Pistos, I. Panderi, J. Atta-Politou, Liquid chromatography–positive ion electrospray mass spectrometry method for the quantification of citalopram in human plasma, J. Chromatogr. B 810 (2004) 235–244. [24] L. Kristoffersen, A. Bugge, E. Lundanes, L. Slørdal, Simultaneous determination of citalopram, fluoxetine, paroxetine and their metabolites in plasma and whole blood by high-performance liquid chromatography with ultraviolet and fluorescence detection, J. Chromatogr. B 734 (1999) 229–246. [25] P. Molander, A. Thomassen, L. Kristoffersen, T. Greibrokk, E. Lundanes, Simultaneous determination of citalopram, fluoxetine, paroxetine and their metabolites in plasma by temperature-programmed packed capillary liquid chromatography with on-column focusing of large injection volumes, J. Chromatogr. B 766 (2001) 77–87. [26] H. Juan, Z. Zhiling, L. Huande, Simultaneous determination of fluoxetine, citalopram, paroxetine, venlafaxine in plasma by high performance liquid chromatography– electrospray ionization mass spectrometry (HPLC–MS/ESI), J. Chromatogr. B 820 (2005) 33–39. [27] Q.H. Meng, D. Gauthier, Simultaneous analysis of citalopram and desmethylcitalopram by HPLC following solid-phase extraction, Clin. Biochem. 38 (2005) 282–285. [28] S.M.R. Wille, K.E. Maudens, C.H. Van Peteghem, W.E.E. Lambert, Development of a solid phase extraction for 13 ‘new’ generation antidepressants and their active metabolites for gas chromatographic–mass spectrometric analysis, J. Chromatogr. A 1098 (2005) 19–29. [29] J.J. Berzas Nevado, M.J. Villasenor Llerena, C.G. Cabanillas, V.R. Robledo, Screening of citalopram, fluoxetine and their metabolites in human urine samples by gas chromatography–mass spectrometry: a global robustness/ruggedness study, J. Chromatogr. A 1123 (2006) 130–133. [30] M. Mzloum-Ardakani, H. Beitollah, M.K. Amini, F. Mirkhalaf, M.A. Alibeik, New strategy for simultaneous and selective voltammetric determination of norepinephrine, acetaminophen and folic acid using ZrO2 nanoparticles-modified carbon paste electrode, Sensor. Actuat. B 151 (2010) 243. [31] J. Tashkhourian, S.F. Nami-Ana, A sensitive electrochemical sensor for determination of gallic acid based on SiO2 nanoparticle modified carbon paste electrode, Mater. Sci. Eng. C 52 (2015) 103–110. [32] M.B. Gholivand, M. Torkashvand, E. Yavari, Electrooxidation behavior of warfarin in Fe3O4 nanoparticles modified carbon paste electrode and its determination in real samples, Mater. Sci. Eng. C 48 (2015) 235–242. [33] A. Afkhami, H. Ghaedi, Multiwalled carbon nanotube paste electrode as an easy, inexpensive and highly selective sensor for voltammetric determination of Risperidone, Anal. Methods 4 (2012) 1415–1420. [34] A. Afkhami, T. Madrakian, S.J. Sabounchei, M. Rezaei, S. Samiee, M. Pourshahbaz, Construction of a modified carbon paste electrode for the highly selective simultaneous electrochemical determination of trace amounts of mercury(II) and cadmium(II), Sensor. Actuat. B 161 (2012) 542–548. [35] A. Afkhami, T. Madrakian, H. Ghaedi, Construction of a chemically modified electrode for the selective determination of nitrite and nitrate ions based on a new nanocomposite, Electrochim. Acta 66 (2012) 255–264. [36] A.A. Ensafi, H. Karimi-Maleh, S. Mallakpour, Electroanalysis, N-(3,4dihydroxyphenethyl)-3,5-dinitrobenzamide-modified multi-wall carbon nanotubes paste electrode as a novel sensor for simultaneous determination of penicillamine, uric acid and tryptophan, Electroanal. 23 (2011) 1478–1487. [37] A. Afraz, A.A. Rafati, M. Najafi, Optimization of modified carbon paste electrode with multiwalled carbon nanotube/ionic liquid/cauliflower-like gold nanostructures for simultaneous determination of ascorbic acid, dopamine and uric acid, Mater. Sci. Eng. C 44 (2014) 58–68″. [38] M.R. Ganjali, N. Motakef-Kazami, F. Faridbod, S. Khoee, P. Norouzi, Determination of Pb2+ ions by a modified carbon paste electrode based on multi-walled carbon nanotubes (MWCNTs) and nanosilica, J. Hazard. Mater. 173 (2010) 415–419. [39] A. Nojeh, G.W. Lakatos, S. Peng, K. Cho, R.F.W. Pease, A carbon nanotube cross structure as a nanoscale quantum device, Nano Lett. 3 (2003) 1187–1190. [40] H. Beitollahi, H. Karimi-Maleh, H. Khabazzadeh, Nanomolar and selective determination of epinephrine in the presence of norepinephrine using carbon paste electrode modified with carbon nanotubes and novel 2-(4-Oxo-3-phenyl-3,4dihydro-quinazolinyl)-N′-phenyl-hydrazinecarbothioamide, Anal. Chem. 80 (2008) 9848–9851.

854

H. Ghaedi et al. / Materials Science and Engineering C 59 (2016) 847–854

[41] R.H. Ouyang, Z.Q. Zhu, C.E. Tatum, J.Q. Chambers, Z.-L. Xue, Simultaneous stripping detection of Zn(II), Cd(II) and Pb(II) using a bimetallic Hg–Bi/single-walled carbon nanotubes composite electrode, J. Electroanal. Chem. 656 (2011) 78–84. [42] S. Shahrokhian, Z. Kamalzadeh, A. Bezaatpour, D.M. Boghaei, Differential pulse voltammetric determination of N-acetylcysteine by the electrocatalytic oxidation at the surface of carbon nanotube-paste electrode modified with cobalt salophen complexes, Sensor. Actuat. B 133 (2008) 599–606. [43] Y.H. Li, X.Y. Liu, X.D. Zeng, Y. Liu, X.T. Liu, W.Z. Wei, S.L. Luo, Simultaneous determination of ultra-trace lead and cadmium at a hydroxyapatite-modified carbon ionic liquid electrode by square-wave stripping voltammetry, Sensor. Actuat. B 139 (2009) 604–610. [44] H. Nouws, C. Delerue-Matos, A. Barros, Electrochemical determination of citalopram by adsorptive stripping voltammetry-determination in pharmaceutical products, Anal. Lett. 39 (2006) 1907–1915. [45] C. Annamalai, R. Kunjithapatham, S.K. Rathinavelu, Electrochemical behavior and voltammetric determination of an antidepressant drug in pharmaceuticals and biological fluids samples, J. Biosci. Res. 2 (2011) 142–159. [46] F. Valentini, A. Amine, S. Orlandocci, M.L. Terranova, G. Palleschi, Carbon nanotubes purification: preparation and characterization of carbon nanotube paste electrodes, Anal. Chem. 75 (2003) 5413–5421.

[47] T. Madrakian, A. Afkhami, M. Ahmadi, H. Bagheri, Removal of some cationic dyes from aqueous solutions using magnetic-modified multi-walled carbon nanotubes, J. Hazard. Mater. 196 (2011) 109–114. [48] W. Sun, Z.Q. Zhai, D.D. Wang, S.F. Liu, K. Jiao, Electrochemistry of hemoglobin entrapped in a Nafion/nano-ZnO film on carbon ionic liquid electrode, Bioelectrochem. 74 (2009) 295–300. [49] M.G. Zhang, A. Smith, W. Gorski, Carbon nanotube–chitosan system for electrochemical sensing based on dehydrogenase enzymes, Anal. Chem. 76 (2004) 5045–5050. [50] D. Wei, A. Ivaska, Applications of ionic liquids in electrochemical sensors, Anal. Chim. Acta 607 (2008) 126–135. [51] A. Sevcik, Collect. Czechoslov. Chem. Commun. 13 (1948) 349–377. [52] E. Laviron, General expression of the linear potential sweep voltammogram in the case of diffusionless electrochemical systems, J. Electroanal. Chem. 101 (1979) 19–28. [53] A.J. Bard, L.R. Faulkner, Electrochemical Methods Fundamentals and Application, second ed. Wiley, 2004 236. [54] G. Alberti, R. Palombari, F. Pierri, Use of NiO, anodically doped with Ni(III), as reference electrode for gas sensors based on proton conductors, J. Solid State Ionics 97 (1997) 359–364.