- Email: [email protected]
Ultra-sensitive voltammetric sensor for trace analysis of carbamazepine
Analytica Chimica Acta 674 (2010) 182–189
Contents lists available at ScienceDirect
Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca
Ultra-sensitive voltammetric sensor for trace analysis of carbamazepine Alfredina Veiga a , Ana Dordio a , A.J. Palace Carvalho a,c , Dora Martins Teixeira a,b , Jorge Ginja Teixeira a,c,∗ Chemistry Department, Évora University, CLAV Rua Romão Ramalho n.◦ 59, 7000-671 Évora, Portugal Institute of Mediterranean Agricultural and Environmental Sciences (ICAAM), Évora, Portugal c Chemistry Center of Évora (CQE), Évora, Portugal a
b
a r t i c l e
i n f o
Article history: Received 26 April 2010 Received in revised form 18 June 2010 Accepted 23 June 2010 Available online 1 July 2010 Keywords: Carbamazepine Multi-walled carbon nanotubes (MWCNTs) Voltammetric determination Environmental samples Pharmaceutical formulations
a b s t r a c t A multi-walled carbon nanotubes (MWCNTs) film-coated glassy carbon electrode (GCE) was used for the voltammetric determination of carbamazepine (CBZ). The results showed that this simple modified electrode exhibited excellent electrocatalytic activity towards the oxidation of CBZ. The voltammetric response of CBZ at this film-modified electrode increased significantly when compared with that at a bare glassy carbon electrode and the sensor response was reproducible. The proposed method was applied to the quantification of CBZ in wastewater samples, collected in a municipal wastewater treatment plant, and in pharmaceutical formulations. The developed methodology yields results in accord with those obtained by chromatographic techniques commonly used in the quantification of pharmaceutical compounds in real samples. Good recoveries have been obtained and the limits of detection and quantification (40 and 140 nM, respectively) are among the lowest that have been reported to date for this pharmaceutical compound using voltammetric techniques. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Carbamazepine (Fig. 1), chemically known as 5Hdibenzo[b,f]azepine-5-carboxamide, is an anticonvulsant, antiepileptic and mood stabilizing drug used primarily in the treatment of epilepsy and bipolar disorder. It is also used to treat other affective disorders such as resistant schizophrenia, ethanol withdrawal, restless leg syndrome, psychotic behavior associated with dementia and post-traumatic stress disorders [1,2]. Carbamazepine (CBZ) is among the most widely prescribed drugs, in a variety of dosages and pharmaceutical forms [3]. Due to its high consumption rates in modern society [3] it became important to develop and establish new, fast and accurate methodologies for the determination of this drug in applications like quality control analysis of pharmaceutical formulations, clinical control of CBZ consumers, and especially in the field of analytical environmental chemistry, since CBZ is currently considered one of the emergent pollutants to be frequently monitored in ground and surface waters [3–5]. For this reason, CBZ has also been proposed as an anthropogenic marker in water bodies [6]. The introduction of CBZ in the aquatic ecosystem occurs mainly through human excreta of the unabsorbed fraction of CBZ, and dis-
∗ Corresponding author at: Chemistry Department, Évora University, CLAV Rua Romão Ramalho n.◦ 59, 7000-671 Évora, Portugal. Tel.: +351 266 745304; fax: +351 266 745303. E-mail address: [email protected] (J.G. Teixeira). 0003-2670/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2010.06.031
posal of unused or expired drug [3,4,7]. In wastewater treatment plants (WWTP) CBZ is recalcitrant to the conventional treatment processes used in WWTPs and, as it is released in the environment, it proves to be highly resistant to biodegradation [8–10], which helps explain why it is one of the most frequently detected pharmaceutical residues in water bodies around the world. If the pharmaceutical residues are not effectively removed from water by drinking water treatment plants, they will be unintentionally consumed by humans [3]. During wastewater treatment, CBZ was found to be removed by less than 10% [11–13]. As a result, CBZ has been detected at concentrations up to 1.075 g L−1 (4.5 nM) in surface water samples and up to 0.610 g L−1 (2.6 nM) in groundwater samples [3,12,13]. It has also been found in drinking and sea water, but at lower concentration levels [3,14]. In spite of the trace levels detected (ng L−1 to g L−1 ) it is important to take into account the effects of long-term exposure to pharmaceuticals like CBZ on public health and aquatic ecology. Additionally, the unknown synergistic effects of the pharmaceutical mixtures must also be considered another potential problem [14]. Therefore, the occurrence and fate of pharmaceuticals in the aquatic environment has been recognized as one of the emerging issues in analytical and environmental chemistry in the last decade. Several analytical methods have been published concerning the determination of CBZ in pharmaceutical formulations and clinical and environmental samples. The most commonly used are chromatographic techniques, like high performance liquid chromatography (HPLC) with UV [15,16] or diode-array detection [17],
A. Veiga et al. / Analytica Chimica Acta 674 (2010) 182–189
Fig. 1. Molecular structure of carbamazepine (CBZ).
liquid chromatography coupled with mass spectrometry (LC–MS) [18], LC–MS tandem [14,19,20] and gas chromatography coupled with mass spectrometry (GC–MS) [21,22], because of the high sensitivity, selectivity and lower quantification limits that these techniques can achieve (ng–g L−1 range) [23]. Other chromatographic techniques, like micellar electrokinetic chromatography (MEKC) [24], have also been used. Studies using spectrophotometric techniques, such as UV/vis spectrophotometry [5,25,26], and electroanalytical techniques, like voltammetry [5,27–31], are comparatively very scarce and rarely adopted. The few voltammetric methods known rely on cathodic behavior of CBZ at mercury electrodes and on its anodic behavior at carbon electrodes. Among these, the limits of detection vary between 938 and 1800 nM. Despite its less frequent use on the detection and determination of CBZ, it is well known that the voltammetric methods can overcome some of the problems associated to the chromatographic methods. While the latter methods always require the use of expensive and complex equipment, and some of them involve laborious and time-consuming multiple stage pre-concentration steps, the voltammetric methods can be very rapid, simple and sensitive, depending of the electrode materials and techniques used. In addition, they require low cost running time operations and equipments. These equipments can be also portable, which is particularly suitable for on-site monitoring of pharmaceuticals in field analysis of contaminated samples. The aim of this work was to develop a new, simple and ultrasensitive electrochemical methodology, for the quantification of CBZ in environmental samples and pharmaceutical formulations, using voltammetric techniques. The electrochemical sensor was a glassy carbon electrode (GCE) simply modified with multi-walled carbon nanotubes (MWCNTs) [32–35]. Several recent studies have been published dealing with the quantification of pharmaceutical compounds [36–41] using this simple and distinctive sensor. Attending to this and to the fact that very little improvements have been reported on voltammetric determinations of CBZ [5,27–31], a MWCNTs film-coated GCE [35,38,40–43] is proposed in this work, for the determination of this pharmaceutical substance. To the best of our knowledge this was never tried or reported. The proposed method was successfully applied to the quantification of CBZ in commercially available medicinal tablets and in wastewater samples collected from a municipal WWTP in the region of Évora – Portugal. Additionally, the voltammetric methodology yields results that are in accord with the results obtained by HPLC–UV and LC–MS techniques.
183
Clofibric acid (97% purity), atenolol (≥98% purity), diclofenac sodium salt (≥98% purity) and ibuprofen (99.8% purity) were all obtained from Sigma–Aldrich. MWCNTs with >95% purity (7–15 nm o.d. × 3–6 nm i.d. × 0.5–200 m length) were purchased from Sigma–Aldrich (Ref. 694185). These were used as received, without any chemical oxidation or washing step. Dihexadecyl hydrogen phosphate (DHP) was purchased from Sigma–Aldrich (Ref. D2631) and stored at −18 ◦ C. All other chemical reagents used (e.g., Na2 HPO4 , KH2 PO4 , HCl and NaOH) for solutions preparation were reagent grade from Merck and used as received. All solutions were prepared with ultra-pure water (resistivity ≥ 18 M cm) from a Millipore MilliQ system (Simplicity® UV, Millipore Corp., France), and methanol (HPLC gradient grade, Merck). The Proaxen® solution was prepared in a mixture of MeOH:H2 O (4:96, v/v), by dissolving a tablet (300 mg oxcarbazepine) in 20 mL of methanol, followed by its dilution with ultra-pure water on a volumetric flask of 500 mL. The final concentration was 600 mg L−1 oxcarbazepine (or 2.38 × 10−3 M). All other stock solutions of pharmaceutical substances (1.0 mM) were prepared in a mixture of MeOH:H2 O (40:60, v/v). All solutions were sonicated in an ultra-sonic bath for 15 min. Working solutions (0.2 mM) of pharmaceutical substances, including the oxcarbazepine solution, were prepared daily by suitable dilution (with ultra-pure water) of the respective stock solution before the experiments. A phosphate buffer stock solution (Na2 HPO4 /KH2 PO4 , with pH 7.00 and 0.1 M) was used to prepare the supporting electrolyte solutions. Whenever necessary the pH was adjusted either with 4.7 M HCl or 5 M NaOH to the studied values. The wastewater samples were collected after a secondary treatment stage in a wastewater treatment plant (WWTP), serving a small rural community population of ca. 400 inhabitants of Alentejo – Portugal. These samples were collected in amber glass bottles and transported to the laboratory in ice. Prior to analysis, the samples were filtered through a 0.45 m PTFE filter and stored in a refrigerator (at 4 ◦ C). The wastewater samples were characterized by the determination of the following wastewater quality parameters, according to the APHA-AWWA-WPCF methods [44]: total suspended solids (TSS), pH and total and soluble chemical oxygen demand (CODt and CODs ) of the filtered samples. 2.2. Pharmaceutical preparations and spiked wastewater samples The Tegretol® solutions were prepared in a mixture of MeOH:H2 O (20:80, v/v), by dissolving a tablet (200 mg CBZ) in 200 mL of methanol, followed by its dilution with ultra-pure water on a volumetric flask of 1 L. The final concentration was 200 g mL−1 CBZ (or 0.85 × 10−3 M). These solutions were sonicated in an ultra-sonic bath for 15 min, before use. Appropriate aliquots of wastewater samples, unspiked and spiked with 0.500 g mL−1 of CBZ, were subjected to a solid phase extraction (SPE) pre treatment. The samples were percolated through a 3 mL LiChrolut® C18 column, which was previously conditioned with 7.5 mL of methanol and 7.5 mL of ultra-pure water.
2. Experimental
2.3. Electrochemical apparatus and measurements
2.1. Reagents and solutions
Cyclic, linear sweep and square wave voltammetric experiments were carried out using an Autolab PGSTAT 20 (Eco Chemie, Utrecht, Netherlands). The instrument was computer-controlled using GPES (General Purpose Electrochemical System) software, version 4.9. All the measurements were carried out in a three-electrode measuring cell. The working electrodes were a bare and a MWCNT modified glassy carbon disc electrode of 2.0 mm diameter, the auxiliary electrode was a platinum rod, and the reference electrode was Ag/AgCl/3 M KCl. These two electrodes and the glassy car-
CBZ (>99% purity) was obtained from Sigma–Aldrich and the pharmaceutical preparation containing this active compound (Tegretol® tablets with 200 mg CBZ, from Novartis® ) was acquired on a local pharmacy. Oxcarbazepine was used from the pharmaceutical preparation Proaxen® (tablets with 300 mg oxcarbazepine, from Pentafarma/Grünenthal S.A.® ). This product was also acquired on a local pharmacy.
184
A. Veiga et al. / Analytica Chimica Acta 674 (2010) 182–189
bon electrode were purchased from Metrohm. All the potentials in the text are quoted vs. the reference electrode. The used glassy carbon disc electrodes were hand-polished using 0.3 m alumina powder on a polishing cloth followed by successively rinsing with acetone and deionised water, before each experiment or modification. Cyclic voltammetric (CV) measurements were carried out at scan rates of 25, 50, 75, 100, 150, 200, 250 and 300 mV s−1 using a step potential of 2.44 mV. Linear sweep voltammetric (LSV) measurements were carried out at scan rates of 100 mV s−1 . The square wave voltammetric (SWV) measurements were carried out at a frequency of 100 Hz, step potential of 1.2 mV and amplitude of 20 mV. All voltammetric measurements were performed between −0.700 V and some potential value between 1.300 and 1.500 V. All voltammograms were recorded at a constant temperature of 22 ◦ C. Solutions in the cell were purged with oxygen-free nitrogen for 5 min prior to the measurement. A nitrogen blanket was maintained above solutions surface during the measurements. The pH was controlled by a Metrohm pH/mV meter (Model 632) with a glass pH electrode. To homogenize the stock solutions and dispersions, an ultra-sonic bath (Transsonic T660/H Elma® ) was used. 2.4. Preparation of glassy carbon electrode modified with MWCNTs film (MWCNTs–GCE) In the preparation of MWCNTs–GCE, 1 mg MWCNTs was dispersed in 1 mL ultra-pure water containing 1 mg DHP, with the aid of ultra-sonic agitation for about 30 min, to produce a 1 mg mL−1 MWCNTs dispersion. The DHP was used as a dispersing agent to obtain a homogeneous dispersion of multi-walled carbon nanotubes in water and a uniform and stable MWCNTs film over the GCE surface [35]. As previously stated, prior to modification the GCE surface was treated and cleaned. After this, 2–12 L of the MWCNTs dispersion was cast on the GCE surface and dried in a vacuum desiccator. To study the effect of DHP, a DHP–GCE was prepared in a similar way, by casting the GCE surface with a fixed amount of a prepared solution of 1 mg DHP/1 mL ultra-pure water. 2.5. HPLC–UV and LC–ESI-MS/MS analysis High performance liquid chromatography (HPLC) with UV/vis spectrometry detection, using an Elite LaChrom HPLC system equipment (Hitachi, Japan), was used to quantify CBZ in commercial tablets and wastewater samples. The separation was performed in isocratic mode, with a mobile phase composed by 75:25 (%, v/v) acetonitrile:water (acidified with phosphoric acid 0.1%, v/v), at a flow rate of 1.0 mL min−1 , and using a reversed phase analytical column Zorbax Eclipse XDB-C18 with 5 m particle size. The UV detector wavelength was set at 210 nm. LC–ESI-MS/MS analyses of the samples were carried out in a LCQ Advantage ThermoFinnigan mass spectrometer equipped with an electrospray ionisation source and using an ion trap mass analyzer. The conditions of analysis were: capillary temperature of 275 ◦ C, source voltage of 5.0 kV, source current of 100.0 A, and capillary voltage of 15.0 V in positive ion mode. The mass spectrometer equipment was coupled to an HPLC system with autosampler (Surveyor ThermoFinnigan). The analytical column was a reversed phase Thermo Hypersil gold (C18, particle size 5 m, 150 mm × 2.1 mm). The quantification of CBZ was performed with an isocratic program using methanol as eluent A and water acidified with 0.1% (v/v) formic acid as eluent B. The mobile phase was composed by 75% eluent A:25% eluent B (v/v) at a flow rate of 0.3 mL min−1 . The LC–ESI-MS/MS analyses were performed in the SRM (selected reaction monitoring) mode, using the most intense transition described for CBZ (237.0 → 193.9) [19]. Five replicate injections of 20 L were made for each sample previously filtered through a 0.45 m filter. Calibration curves were
Fig. 2. Cyclic voltammetric responses of 6.0 × 10−6 M CBZ in 0.1 M phosphate buffer solution (pH 6.89) at (a) MWCNTs film-coated GCE (8 L of dispersion) and (c) bare GCE. The dotted lines (b) and (d) represent the corresponding blank responses. Scan rate: 100 mV s−1 .
constructed using a set of CBZ standard solutions, prepared with methanol/water (20:80, v/v) or wastewater, with concentrations ranging from 0.25 to 5.0 mg L−1 (1.06–21.16 × 10−6 M). 3. Results and discussion 3.1. Cyclic voltammetric behavior of carbamazepine The electrochemical behavior of CBZ in 0.1 M phosphate buffer with pH 6.89 was investigated by cyclic voltammetry (CV). Fig. 2 depicts the cyclic voltammograms for 6.0 × 10−6 M CBZ at (a) MWCNTs–GCE (8 L of dispersion) and (c) bare GCE. The dotted lines (b) and (d) represent the corresponding blank responses. Within the potential window studied it can be seen that the voltammetric response of CBZ at the MWCNTs–GCE was strongly enhanced, when compared with that of the bare GCE electrode. At this unmodified electrode (Fig. 2c), carbamazepine showed an irreversible behavior with a broad and very weak oxidation peak, around 1.170 V. Above 1.300 V no other CBZ oxidation peak was observed (results not shown), in contrast with previously reported results by Kalanur and Seetharamappa [31]. This was confirmed with further measurements up to concentrations of 15.0 × 10−6 M CBZ. Above this concentration the second oxidation peak, as reported by those authors, begins to be distinguished. Meanwhile, as can be seen from Fig. 2a carbamazepine exhibited at the MWCNTs–GCE a well-defined and very intense irreversible oxidation peak, P+ , around 1.090 V, with no additional post-peak in the forward scan or counterpart peak in the reverse scan. Above this anodic region and at higher CBZ concentrations no further peaks were observable in the voltammograms. The shift of the peak potential to lower anodic potentials and the peak current improvement obtained for CBZ oxidation at the MWCNTs–GCE, in comparison to the GCE response, reflects the decrease of the overpotential for this anodic reaction and the enhancement of the electron transfer between the CBZ molecule and the modified electrode. As expected for these electrodes, these can be ascribed to the strong adsorptive ability and electrocatalytic surface properties of the MWCNTs [35]. On the basis of the few reports about the anodic behavior of the CBZ molecule, the electrode process under discussion can be related with the oxidation of the nitrogen atom in the central ring, which results in the formation of cation radicals (Scheme 1) [31]. Due to the large active surface area, the high electronic reactivity
A. Veiga et al. / Analytica Chimica Acta 674 (2010) 182–189
185
Scheme 1. Probable reaction mechanism for electrooxidation “P+ ” of CBZ at the MWCNTs/GCE (adapted from Ref. [31]).
of the MWCNTs film and the complexity of the oxidation process we believe that the two steps leading to the loss of electrons, as illustrated in Scheme 1, occur simultaneously and cannot be distinguished in the voltammogram. On consecutive cyclic voltammetric sweeps, the electrochemical signals of CBZ at the modified electrode become consecutively poorer (Fig. 3), suggesting that the electrode surface is blocked by the adsorption of the reaction products of CBZ, which reduces the effective reaction sites of the modified electrode surface. As it is
well known, this is a factor that can compromise the voltammetric determination of analytes in routine analysis. However, it was found in this study that the voltammetric response was maintained (RSD = 0.67% for n = 5) if the electrode was kept at an open circuit potential (ocp) for at least 3 min and stirring with a rotation speed of 1000 revolutions min−1 between each run. Similar requirements were deemed necessary for reproducible voltammetric analysis of other tricyclic antidepressant drugs using other electrodes [45]. Following this procedure, the voltammetric determination can be performed with the same modified electrode, without any need for cleaning or discarding it. Probably during the stirring at an open circuit potential, the adsorbed products are released from the electrode surface, and a reproducible quantity of CBZ molecules is accumulated at reactive sites of the MWCNT electrocatalytic layer. Therefore, the voltammetric study of CBZ towards its electrochemical characterization and the analytical quantification was performed with the data obtained from the first voltammetric scan (cyclic voltammetry, CV, or linear sweep voltammetry, LSV). 3.2. Optimization of the experimental variables
Fig. 3. Successive cyclic voltammograms of 6.0 × 10−6 M CBZ at MWCNTs filmcoated GCE. Other conditions are as in Fig. 2.
3.2.1. Influence of the equilibration potential and time In electrochemical studies, where the interaction between the electroactive analyte and the electrode surface can have a strong influence on the quality of the signal to be measured, it is important to fix the potential and time conditions that precede the measurement. In addition to the ocp conditions used, the equilibration potential and time can affect the amount of CBZ that accumulates at the electrode. Bearing this in mind, the effect of both factors on peak current response was also studied. When equilibration poten-
186
A. Veiga et al. / Analytica Chimica Acta 674 (2010) 182–189
Fig. 4. Influence of solution pH on peak potential and peak current of peak P+ . Other conditions are as in Fig. 2.
tial was varied from −0.700 to 0.700 V, it was found that the peak current decreased between −0.500 and 0.700 V (Fig. S1, supplementary data). Hence, an equilibration potential of −0.700 V was adopted. The peak current increased very little when the equilibration time varied from 1 to 300 s (Fig. S1, supplementary data). This indicates that this period of time has little influence on the oxidation peak, when compared with the previous 3 min undertaken at the ocp. As already suggested, it is possible that during this stage, and for a given CBZ solution, the electrode surface becomes saturated with CBZ molecules. To avoid overextending the time of analysis, an equilibration time of 1 s was chosen. 3.2.2. Influence of pH The effect of pH on the anodic response (P+ ) of CBZ at the modified electrode was tested over the pH range 2.0–11.8 using the same supporting electrolyte solution. Within this pH range, the peak potential Ep,a is almost pH independent between pH 2 and 5, and above this value is shifted to less positive values as the pH of the solution increases (Fig. 4). The greater variations occur between pH 5.14 and 7.25 and for pH > 11. In the pH range 5.1–7.3, the Ep,a varies linearly with pH, with a negative slope of 27.9 mV pH−1 . This indicates that the number of protons released from the molecule, in this anodic process and at this pH range, seems to be half the number of electrons transferred to the electrode [45,46]. This is in agreement with the proposed mechanism by Kalanur and Seetharamappa [31] (Scheme 1), since four electrons and two protons are involved, when one dicationic radical dimer is produced. As can be also observed, the peak current is very sensitive to pH changes between pH 5.1 and 8.0, achieving a maximum value around 6.9 (the pH of the phosphate buffer used as supporting electrolyte). Outside this pH range the peak current decreases significantly when pH ≥ 10. Therefore, further studies were performed at pH 6.89. 3.2.3. Influence of the scan rate on peak P+ The dependence of peak potential, Ep , as well as peak current, ip , on the scan rate () were studied in the range 25–300 mV s−1 . The Ep is a function of the scan rate, shifting to a positive direction, and the plot of Ep vs. log (Fig. 5) was linear. It was also found that the peak current, ip , is proportional to the square root of the scan rate (ip (A) = 28.0611/2 (V s−1 )1/2 − 1.6687; R2 = 0.9909; the relation ip vs. was not linear, presenting a R2 = 0.9765). These diagnostic tests prove that the electrode reaction under study is irreversible [47]. The linear relationship found between ip and 1/2 suggests that the electrooxidation is a diffusion-controlled process. However, a lin-
Fig. 5. Cyclic voltammograms of [CBZ] = 6.0 × 10−6 M on MWCNTs-modified GCE at different scan rates (V s−1 ): (a) 0.025; (b) 0.050; (c) 0.075; (d) 0.100; (e) 0.150; (f) 0.200; (g) 0.250; (h) 0.300. Other conditions are as in Fig. 2. The inset shows the plot of anodic peak potential (Ep ) vs. log . The error bars of Ep values are too short to be represented.
ear relationship with a slope close to 0.62, observed between log ip and log (Fig. 6, with ip and expressed in A and V s−1 , respectively) seems to be indicative of some adsorption contributions to the electrode reaction. It can be said that this reaction at the modified GCE, is a “mixed” diffusion–adsorption process [47–49]. This is also in agreement with the proposed mechanism by Kalanur and Seetharamappa [31]. According to these authors, one of the two oxidation reactions is a diffusion-controlled process and the other is an adsorption-limited process. As already noted, at the MWCNTs–GCE, the electrode reaction is a “mixed” diffusion–adsorption process because the two irreversible oxidation processes seem to occur together in a very short time. This, in fact, can be confirmed by square wave voltammetry (SWV) measurements (Fig. 7) as the forward and backward components of the current present only one peak with same height and sign, resulting in an overall voltammogram showing no peaks [50]. 3.2.4. Influence of the amount of MWCNTs/DHP dispersion Fig. 8 illustrates the relationship between the volume of 1 mg mL−1 dispersion used to modify the GCE surface and the anodic peak current of CBZ. It was found that the peak current increased in general with an increasing volume of the dispersion.
Fig. 6. Plot of log ip vs. log . Same conditions as reported in Fig. 5.
A. Veiga et al. / Analytica Chimica Acta 674 (2010) 182–189
Fig. 7. Square wave voltammetric response of 2.7 × 10−6 M CBZ in 0.1 M phosphate buffer solution (pH 6.89) at MWCNTs film-coated GCE (10 L of dispersion). it – total current; if – forward current; and ib – backward current. Effective scan rate: 120 mV s−1 .
Above 10 L the peak current changed only slightly. Since the signal was not improved above this quantity, and because the dispersion spreads beyond the limits of the conductive surface of GCE electrode, the volume of the dispersion chosen in the analytical determination of CBZ was 10 L. 3.3. Calibration curve
Fig. 9. Calibration curve of carbamazepine established by LSV, under the optimized experimental conditions. Between each measurement the modified electrode was maintained at ocp during 3 min and constant rotation of 1000 rpm. Table 1 Comparison of detection limits for CBZ at different electrodes. Electrodes
Detection limits/nM
Refs.
HMDEa HMDEa AgNPs/SPCEb GCEc MWCNTs/GCEd
1500 1000 938 1800 40
[5] [28] [30] [31] Present work
a b
The calibration curve for CBZ (Fig. 9) was established by linear sweep voltammetry (LSV), under the optimized experimental conditions described above. A scan rate of 100 mV s−1 was used. At these conditions, the plot of peak current, ip , vs. CBZ concentration was found to be linear in the range 0.13–1.60 M (30.7–378 g L−1 ). The limits of detection (LOD) and quantification (LOQ) of CBZ were calculated using the typical equations LOD = 3 s m−1 and LOQ = 10 s m−1 (where s is the standard deviation of the peak currents of the blank (five runs) and m is the slope of the calibration curve). The LOD and LOQ obtained were 40 and 140 nM (9.5 and 33.1 g L−1 ), respectively. To the best of our knowledge, this LOD is the lowest that has been reported for carbamazepine using electrochemical techniques (Table 1). At higher scan rates, the linearity between the peak current, ip , and CBZ concentration became unsatisfactory.
187
c d
Hanging mercury drop electrode. Silver nanoparticle-modified carbon screen-printed electrode. Glassy carbon electrode. Multi-walled carbon nanotubes-modified glassy carbon electrode.
3.4. Interferences Table 2 lists the influence of other pharmaceutical substances, which possibly can occur in wastewater [10], on the oxidation signal of CBZ. Under the optimum experimental conditions, it was found that diclofenac and clofibric acid are the strongest interferents. Both pharmaceuticals enhanced the oxidation signal of CBZ, without distorting its shape. In the other three cases, the effect in the peak current response of CBZ was substantially more reduced. For all pharmaceuticals tested, the interferences remained at the same level when the concentration of the compounds was two times the CBZ concentration. Among the tested substances, diclofenac was the only pharmaceutical that produced an additional voltammetric peak (at 0.630 V approx.) at the modified GCE, as noted by Yang et al. [41]. Attending to these effects, it is advisable to perform the analytical determination of CBZ at real samples using the standard addition method, instead of the calibration curve method. 3.5. Analytical determination of CBZ The MWCNTs–GCE was used to determine CBZ in more realistic samples such as tablets and in spiked and unspiked wastewater Table 2 Effect of potential interferents on the voltammetric response of 0.150 M CBZ. Potential interferents in wastewaters (0.05 M)
Fig. 8. Peak current of P+ vs. amount of MWCNT/DHP dispersion on the GCE surface. [CBZ] = 6.0 × 10−6 M in 0.1 mol L−1 phosphate buffer solution (pH 6.89). Same voltammetric conditions as in Fig. 2. The plot includes the equivalent anodic response of CBZ at bare GCE.
Clofibric acid Ibuprofen Atenolol Diclofenac sodium Oxcarbazepine
Signal change (%) ±SD +6.5 +1.6 −3.2 +13.8 −2.8
± ± ± ± ±
1.7 0.6 1.0 2.3 1.2
188
A. Veiga et al. / Analytica Chimica Acta 674 (2010) 182–189
Fig. 10. Standard additions calibration curves for the determination of CBZ in () Tegretol® tablet, spiked wastewater (sww) sample (䊉) and unspiked wastewater (ww) sample (). Each curve shows the analytical signal (peak height) obtained with the respective sample solution at the modified GCE (10 L of MWCNTs dispersion), before and after successive standard additions of 47.3 g mL−1 CBZ working solution. Same voltammetric conditions as in Fig. 9. The inset figure shows the linear sweep voltammograms obtained with the spiked wastewater (sww) sample (without and with the six standard additions (solid lines)). The dashed line is the linear sweep voltammetric signal of the unspiked wastewater sample. Table 3 Wastewater quality parameters of samples used in this work, according to the APHAAWWA-WPCF methods [44]. Chemical characteristics
Values
pH TSS (mg L−1 )a CODt (mg L−1 )b CODs (mg L−1 )c
8.29 57 133 82
a b c
± ± ± ±
0.05 3 2 2
Total suspended solids. Total chemical oxygen demand. Soluble chemical oxygen demand.
samples, by LSV. In addition to carbamazepine as the active ingredient, composition of tablets used in the determination of CBZ includes other inactive ingredients, namely cellulose compounds, dextrates, iron oxides, magnesium stearate, mannitol, polyethylene glycol, sodium lauryl sulfate and titanium dioxide. The wastewater samples were also characterized in terms of some of the most common parameters of wastewater quality (Table 3). In general, organic matter load and suspended solids were somewhat high but still within the Portuguese legal limits for discharge into water bodies. The content of CBZ in all samples was obtained using the standard addition method, at the optimized experimental conditions. In the tablet assays, a suitable aliquot (15 L) of Tegretol® stock solution was diluted in 15 mL of 0.1 M phosphate buffer (pH 6.89). The standard additions were made with the working solution (200 M = 47.3 g mL−1 ) of CBZ. In the wastewater assays, the same procedure was repeated. In this case, an appropriate aliquot
(1.5 mL) of the unspiked or spiked wastewater sample (treated by SPE) was diluted in 13.5 mL of supporting electrolyte solution. Fig. 10 shows the standard addition curves obtained for the Tegretol® solution and the wastewater samples. In Table 4 the results obtained in the voltammetric analysis of five replicates from each sample are summarized. In addition, HPLC–UV and LC–ESIMS/MS were also used in the quantification of CBZ to show the importance and accuracy of this voltammetric method. The results obtained with the developed sensor were in good agreement with the content of CBZ advertised in the label of the tested commercial formulation, and with the known content of the spiked wastewater samples. The good recoveries obtained indicate that the proposed modified electrode and the voltammetric method has a great potential for the practical sample analysis of CBZ. The complex chemical composition of both the tablet ingredients and the wastewater components (after SPE pretreatment of the wastewater samples) do not significantly affect the determination of CBZ. In the analysis of the Tegretol® solution, better recoveries were obtained using this method, when compared with the other chromatographic methods. In the case of the spiked wastewater samples, the concentrations of CBZ obtained using HPLC and LC–MS were substantially higher than the known CBZ content. This fact points out one drawback of these techniques, which is the occurrence of matrix effects. These effects are caused by the high susceptibility of the ESI source and UV detector to other components present in the matrix, which may result in the suppression or enhancement of the signal. As referred above, in order to minimize the influence of interferents in the analytical determination of CBZ at real samples, the standard
Table 4 Determination of carbamazepine in Tegretol® , unspiked and spiked wastewater samples. Sample
Tegretol® 200 Spiked wastewater Unspiked wastewater a b c
RSD = 2.19%. RSD = 5.71%. n.d. – not detected.
CBZ content
200 mg tablet−1 0.500 g mL−1 Unknown
Experimental value ± SD found by
Recovery in this method (%)
This method
HPLC
LC–MS
201.1 ± 4.4a 0.479 ± 0.027b n.d.c
185.2 ± 6.6 0.588 ± 0.036 n.d.c
182.1 ± 4.9 0.558 ± 0.030 n.d.c
100.6 95.8 –
A. Veiga et al. / Analytica Chimica Acta 674 (2010) 182–189
addition method is a suitable alternative. However, that technique is more easily performed and less time consuming using voltammetric methods than in chromatographic methods.
[11] [12] [13] [14]
4. Conclusions
[15] [16] [17]
The MWCNTs–GCE and the voltammetric method presented in this work provide a good technique for the determination of CBZ with good repeatability, reproducibility, and recovery along with a very low detection limit. This electroanalytical technique is a promising alternative to the often reported chromatographic methods owing to its simplicity, rapidity, reliability and low cost of analysis. The multi-walled carbon nanotubes-modified GCE electrode exhibited a markedly electrocatalytic activity towards the oxidation of CBZ, leading to a considerable improvement in the signal response with the effect of significantly lowering the detection and quantification limits in comparison with other voltammetric techniques. The method has been applied with success to the determination of CBZ in real pharmaceutical dosage forms as well as spiked and unspiked wastewaters. For that reason, the authors believe that this can be a valuable analytical tool in quality control for the pharmaceutical industry as well in environmental applications. In the future, it is planned to apply the proposed methodology in the determination of carbamazepine in biological fluids. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.aca.2010.06.031. References [1] J.B. Leikin, F.P. Paloucek, Poisoning and Toxicology Handbook, 4th ed., Informa Health Care, New York, 2008. [2] L.J. Doods (Ed.), Drugs in Use: Clinical Case Studies for Pharmacists, 4th ed., Pharmaceutical Press, London, 2010, p. 357. [3] Y. Zhang, S.U. Geißen, C. Gal, Chemosphere 73 (2008) 1151–1161. [4] K. Kümmerer, A. Schuster, in: K. Kümmerer (Ed.), Pharmaceuticals in the Environment: Sources, Fate, Effects and Risks, 3rd ed., Springer, 2008, p. 46. [5] L. Campanella, A. Ambrosi, F. Bellanti, M. Tomassetti, Curr. Anal. Chem. 2 (2006) 229–241. [6] M. Clara, N. Kreuzinger, B. Strenn, Water Res. 38 (2004) 947–954. [7] K. Fent, A. Weston, D. Caminada, Aquat. Toxicol. 76 (2006) 122–159. [8] K. Stamatelatou, V. Vavilin, G. Lyberatos, Bioresour. Technol. 88 (2003) 131–136. [9] M. Clara, N. Kreuzinger, B. Strenn, O. Gans, H. Kroiss, Water Res. 39 (2005) 97–106. [10] V. Matamoros, C. Arias, H. Brix, J.M. Bayona, Water Res. 43 (2009) 55–66.
[18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44]
[45] [46] [47] [48] [49] [50]
189
T.A. Ternes, Water Res. 32 (1998) 3245–3260. H. Heberer, Toxicol. Lett. 131 (2002) 5–17. T. Heberer, J. Hydrol. 266 (2002) 175–189. L. Hu, H.M. Martin, O. Yarce-Bulted, M.N. Sugihara, K.A. Keating, T.J. Strathmann, Environ. Sci. Technol. 43 (2009) 509–515. Y. Kouno, C. Ishikura, M. Homma, K. Oka, J. Chromatogr. B 695 (1997) 349–353. K.M. Patil, S.L. Bodhanka, J. Pharma. Biomed. Anal. 39 (2005) 181–186. L. Hua, M. Delgado, S.T. Iannaccone, L.J. Forman, C.M. Eggers, Ther. Drug Monit. 15 (1993) 317–327. J.L. Maggs, M. Pirmohamed, N.R. Kitteringham, B.K. Park, Drug Metab. Dispos. 25 (1997) 275–280. N.M. Vieno, T. Tuhkanen, L. Kronberg, J. Chromatogr. A 1134 (2006) 101–111. G.F. van Rooyen, D. Badenhorst, K.J. Swart, H.K. Hundt, T. Scanes, A.F. Hundt, J. Chromatogr. B 769 (2002) 1–7. A. Frigerio, K.M. Baker, P.L. Morselli, Adv. Biochem. Psychopharmacol. 7 (1973) 125–134. R.J. Perchalski, B.J. Wilder, Clin. Chem. 20 (1974) 492–493. T.A. Ternes, TrAC Trends Anal. Chem. 20 (2001) 419–434. C. Wang, Z. Wang, D. Han, Q. Wu, X. Zang, Anal. Lett. 39 (2006) 1927–1939. T. Comoglu, N. Gönül, Turk. J. Med. Sci. 35 (2005) 217–221. M.S. Cámara, C. Mastandrea, H.C. Goicoechea, J. Biochem. Biophys. Method 64 (2005) 153–166. D.J. Turk, S.A. McClintock, W.C. Purdy, Anal. Lett. 18 (1985) 2605–2618. Z.Q. Zhang, S.Z. Chen, W.L. Huang, F. Xu, Yao Xue Xue Bao 28 (1993) 312–316. G.S. Moses, K.M. Rao, S. Rao, A. Ramachandraiah, J. Indian Chem. Soc. 72 (1995) 333–337. M.A. García-García, O. Domínguez-Renedo, A. Alonso-Lomillo, M.J. ArcosMartínez, Sensor Lett. 7 (2009) 586–591. S.S. Kalanur, J. Seetharamappa, Anal. Lett. 43 (2010) 618–630. S. Lu, K. Wu, X. Dang, S. Hu, Talanta 63 (2004) 653–657. A. Merkoc¸i, M. Pumera, X. Lopis, B. Pérez, M. del Valle, S. Alegret, TrAC Trends Anal. Chem. 24 (2005) 826–838. Z. Lin, G. Chen, Talanta 70 (2006) 111–115. ˜ ˜ M. José, Pingarrón, Anal. Chim. Acta 622 (2008) 11–47 L. Agüí, P. Yánez-Sede no, (and references cited therein). R.N. Goyal, S.P. Singh, Talanta 69 (2006) 932–937. C. Li, Microchim. Acta 157 (2007) 21–26. A.R.N. Hegde, N.P. Shetti, S.T. Nandibewoor, Talanta 79 (2009) 361–369. H. Jiang, X.-M. Wang, Electrochem. Commun. 11 (2009) 126–129. Y. Wu, S. Ye, S. Hua, J. Pharm. Biomed. Anal. 41 (2006) 820–824. X. Yang, F. Wang, S. Hu, Mater. Sci. Eng. C 28 (2008) 188–194. X. Wen, J. Fei, X. Chen, L. Yi, F. Ge, M. Huang, Environ. Pollut. 156 (2008) 1015–1020. K. Balasubramanian, M. Burghard, Anal. Bioanal. Chem. 385 (2006) 452–468. L.S. Clescerl, A.E. Greenberg, A.D. Eaton (Eds.), Standard Methods for the Examination of Water and Wastewater, American Public Health Association, Washington, DC, 1998. T.A. Ivandini, B.A. Sarada, C. Terashima, T.N. Rao, D.A. Tryk, H. Ishiguro, Y. Kubota, A. Fujishima, J Electroanal. Chem. 521 (2002) 117–126. P.T. Kissinger, W.R. Heineman, Laboratory Techniques in Electroanalytical Chemistry, 2nd ed., Marcel Dekker, New York, 1996. Southampton Electrochemistry Group, Instrumental Methods in Electrochemistry, 1st ed., Ellis Horwood, London, 1985. A.J. Bard, L.R. Faulkner, Electrochemical Methods, Fundamentals and Applications, 2nd ed., Wiley, New York, 2001. K.G. David, Cyclic Voltammetry Simulation and Analysis of Reaction Mechanisms, VCH Publishers, New York, 1993. V. Mirceski, S. Komorsky-Lovric, M. Lovric, Square-wave Voltammetry: Theory and Applications, Springer, Berlin, 2007.
Contents lists available at ScienceDirect
Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca
Ultra-sensitive voltammetric sensor for trace analysis of carbamazepine Alfredina Veiga a , Ana Dordio a , A.J. Palace Carvalho a,c , Dora Martins Teixeira a,b , Jorge Ginja Teixeira a,c,∗ Chemistry Department, Évora University, CLAV Rua Romão Ramalho n.◦ 59, 7000-671 Évora, Portugal Institute of Mediterranean Agricultural and Environmental Sciences (ICAAM), Évora, Portugal c Chemistry Center of Évora (CQE), Évora, Portugal a
b
a r t i c l e
i n f o
Article history: Received 26 April 2010 Received in revised form 18 June 2010 Accepted 23 June 2010 Available online 1 July 2010 Keywords: Carbamazepine Multi-walled carbon nanotubes (MWCNTs) Voltammetric determination Environmental samples Pharmaceutical formulations
a b s t r a c t A multi-walled carbon nanotubes (MWCNTs) film-coated glassy carbon electrode (GCE) was used for the voltammetric determination of carbamazepine (CBZ). The results showed that this simple modified electrode exhibited excellent electrocatalytic activity towards the oxidation of CBZ. The voltammetric response of CBZ at this film-modified electrode increased significantly when compared with that at a bare glassy carbon electrode and the sensor response was reproducible. The proposed method was applied to the quantification of CBZ in wastewater samples, collected in a municipal wastewater treatment plant, and in pharmaceutical formulations. The developed methodology yields results in accord with those obtained by chromatographic techniques commonly used in the quantification of pharmaceutical compounds in real samples. Good recoveries have been obtained and the limits of detection and quantification (40 and 140 nM, respectively) are among the lowest that have been reported to date for this pharmaceutical compound using voltammetric techniques. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Carbamazepine (Fig. 1), chemically known as 5Hdibenzo[b,f]azepine-5-carboxamide, is an anticonvulsant, antiepileptic and mood stabilizing drug used primarily in the treatment of epilepsy and bipolar disorder. It is also used to treat other affective disorders such as resistant schizophrenia, ethanol withdrawal, restless leg syndrome, psychotic behavior associated with dementia and post-traumatic stress disorders [1,2]. Carbamazepine (CBZ) is among the most widely prescribed drugs, in a variety of dosages and pharmaceutical forms [3]. Due to its high consumption rates in modern society [3] it became important to develop and establish new, fast and accurate methodologies for the determination of this drug in applications like quality control analysis of pharmaceutical formulations, clinical control of CBZ consumers, and especially in the field of analytical environmental chemistry, since CBZ is currently considered one of the emergent pollutants to be frequently monitored in ground and surface waters [3–5]. For this reason, CBZ has also been proposed as an anthropogenic marker in water bodies [6]. The introduction of CBZ in the aquatic ecosystem occurs mainly through human excreta of the unabsorbed fraction of CBZ, and dis-
∗ Corresponding author at: Chemistry Department, Évora University, CLAV Rua Romão Ramalho n.◦ 59, 7000-671 Évora, Portugal. Tel.: +351 266 745304; fax: +351 266 745303. E-mail address: [email protected] (J.G. Teixeira). 0003-2670/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2010.06.031
posal of unused or expired drug [3,4,7]. In wastewater treatment plants (WWTP) CBZ is recalcitrant to the conventional treatment processes used in WWTPs and, as it is released in the environment, it proves to be highly resistant to biodegradation [8–10], which helps explain why it is one of the most frequently detected pharmaceutical residues in water bodies around the world. If the pharmaceutical residues are not effectively removed from water by drinking water treatment plants, they will be unintentionally consumed by humans [3]. During wastewater treatment, CBZ was found to be removed by less than 10% [11–13]. As a result, CBZ has been detected at concentrations up to 1.075 g L−1 (4.5 nM) in surface water samples and up to 0.610 g L−1 (2.6 nM) in groundwater samples [3,12,13]. It has also been found in drinking and sea water, but at lower concentration levels [3,14]. In spite of the trace levels detected (ng L−1 to g L−1 ) it is important to take into account the effects of long-term exposure to pharmaceuticals like CBZ on public health and aquatic ecology. Additionally, the unknown synergistic effects of the pharmaceutical mixtures must also be considered another potential problem [14]. Therefore, the occurrence and fate of pharmaceuticals in the aquatic environment has been recognized as one of the emerging issues in analytical and environmental chemistry in the last decade. Several analytical methods have been published concerning the determination of CBZ in pharmaceutical formulations and clinical and environmental samples. The most commonly used are chromatographic techniques, like high performance liquid chromatography (HPLC) with UV [15,16] or diode-array detection [17],
A. Veiga et al. / Analytica Chimica Acta 674 (2010) 182–189
Fig. 1. Molecular structure of carbamazepine (CBZ).
liquid chromatography coupled with mass spectrometry (LC–MS) [18], LC–MS tandem [14,19,20] and gas chromatography coupled with mass spectrometry (GC–MS) [21,22], because of the high sensitivity, selectivity and lower quantification limits that these techniques can achieve (ng–g L−1 range) [23]. Other chromatographic techniques, like micellar electrokinetic chromatography (MEKC) [24], have also been used. Studies using spectrophotometric techniques, such as UV/vis spectrophotometry [5,25,26], and electroanalytical techniques, like voltammetry [5,27–31], are comparatively very scarce and rarely adopted. The few voltammetric methods known rely on cathodic behavior of CBZ at mercury electrodes and on its anodic behavior at carbon electrodes. Among these, the limits of detection vary between 938 and 1800 nM. Despite its less frequent use on the detection and determination of CBZ, it is well known that the voltammetric methods can overcome some of the problems associated to the chromatographic methods. While the latter methods always require the use of expensive and complex equipment, and some of them involve laborious and time-consuming multiple stage pre-concentration steps, the voltammetric methods can be very rapid, simple and sensitive, depending of the electrode materials and techniques used. In addition, they require low cost running time operations and equipments. These equipments can be also portable, which is particularly suitable for on-site monitoring of pharmaceuticals in field analysis of contaminated samples. The aim of this work was to develop a new, simple and ultrasensitive electrochemical methodology, for the quantification of CBZ in environmental samples and pharmaceutical formulations, using voltammetric techniques. The electrochemical sensor was a glassy carbon electrode (GCE) simply modified with multi-walled carbon nanotubes (MWCNTs) [32–35]. Several recent studies have been published dealing with the quantification of pharmaceutical compounds [36–41] using this simple and distinctive sensor. Attending to this and to the fact that very little improvements have been reported on voltammetric determinations of CBZ [5,27–31], a MWCNTs film-coated GCE [35,38,40–43] is proposed in this work, for the determination of this pharmaceutical substance. To the best of our knowledge this was never tried or reported. The proposed method was successfully applied to the quantification of CBZ in commercially available medicinal tablets and in wastewater samples collected from a municipal WWTP in the region of Évora – Portugal. Additionally, the voltammetric methodology yields results that are in accord with the results obtained by HPLC–UV and LC–MS techniques.
183
Clofibric acid (97% purity), atenolol (≥98% purity), diclofenac sodium salt (≥98% purity) and ibuprofen (99.8% purity) were all obtained from Sigma–Aldrich. MWCNTs with >95% purity (7–15 nm o.d. × 3–6 nm i.d. × 0.5–200 m length) were purchased from Sigma–Aldrich (Ref. 694185). These were used as received, without any chemical oxidation or washing step. Dihexadecyl hydrogen phosphate (DHP) was purchased from Sigma–Aldrich (Ref. D2631) and stored at −18 ◦ C. All other chemical reagents used (e.g., Na2 HPO4 , KH2 PO4 , HCl and NaOH) for solutions preparation were reagent grade from Merck and used as received. All solutions were prepared with ultra-pure water (resistivity ≥ 18 M cm) from a Millipore MilliQ system (Simplicity® UV, Millipore Corp., France), and methanol (HPLC gradient grade, Merck). The Proaxen® solution was prepared in a mixture of MeOH:H2 O (4:96, v/v), by dissolving a tablet (300 mg oxcarbazepine) in 20 mL of methanol, followed by its dilution with ultra-pure water on a volumetric flask of 500 mL. The final concentration was 600 mg L−1 oxcarbazepine (or 2.38 × 10−3 M). All other stock solutions of pharmaceutical substances (1.0 mM) were prepared in a mixture of MeOH:H2 O (40:60, v/v). All solutions were sonicated in an ultra-sonic bath for 15 min. Working solutions (0.2 mM) of pharmaceutical substances, including the oxcarbazepine solution, were prepared daily by suitable dilution (with ultra-pure water) of the respective stock solution before the experiments. A phosphate buffer stock solution (Na2 HPO4 /KH2 PO4 , with pH 7.00 and 0.1 M) was used to prepare the supporting electrolyte solutions. Whenever necessary the pH was adjusted either with 4.7 M HCl or 5 M NaOH to the studied values. The wastewater samples were collected after a secondary treatment stage in a wastewater treatment plant (WWTP), serving a small rural community population of ca. 400 inhabitants of Alentejo – Portugal. These samples were collected in amber glass bottles and transported to the laboratory in ice. Prior to analysis, the samples were filtered through a 0.45 m PTFE filter and stored in a refrigerator (at 4 ◦ C). The wastewater samples were characterized by the determination of the following wastewater quality parameters, according to the APHA-AWWA-WPCF methods [44]: total suspended solids (TSS), pH and total and soluble chemical oxygen demand (CODt and CODs ) of the filtered samples. 2.2. Pharmaceutical preparations and spiked wastewater samples The Tegretol® solutions were prepared in a mixture of MeOH:H2 O (20:80, v/v), by dissolving a tablet (200 mg CBZ) in 200 mL of methanol, followed by its dilution with ultra-pure water on a volumetric flask of 1 L. The final concentration was 200 g mL−1 CBZ (or 0.85 × 10−3 M). These solutions were sonicated in an ultra-sonic bath for 15 min, before use. Appropriate aliquots of wastewater samples, unspiked and spiked with 0.500 g mL−1 of CBZ, were subjected to a solid phase extraction (SPE) pre treatment. The samples were percolated through a 3 mL LiChrolut® C18 column, which was previously conditioned with 7.5 mL of methanol and 7.5 mL of ultra-pure water.
2. Experimental
2.3. Electrochemical apparatus and measurements
2.1. Reagents and solutions
Cyclic, linear sweep and square wave voltammetric experiments were carried out using an Autolab PGSTAT 20 (Eco Chemie, Utrecht, Netherlands). The instrument was computer-controlled using GPES (General Purpose Electrochemical System) software, version 4.9. All the measurements were carried out in a three-electrode measuring cell. The working electrodes were a bare and a MWCNT modified glassy carbon disc electrode of 2.0 mm diameter, the auxiliary electrode was a platinum rod, and the reference electrode was Ag/AgCl/3 M KCl. These two electrodes and the glassy car-
CBZ (>99% purity) was obtained from Sigma–Aldrich and the pharmaceutical preparation containing this active compound (Tegretol® tablets with 200 mg CBZ, from Novartis® ) was acquired on a local pharmacy. Oxcarbazepine was used from the pharmaceutical preparation Proaxen® (tablets with 300 mg oxcarbazepine, from Pentafarma/Grünenthal S.A.® ). This product was also acquired on a local pharmacy.
184
A. Veiga et al. / Analytica Chimica Acta 674 (2010) 182–189
bon electrode were purchased from Metrohm. All the potentials in the text are quoted vs. the reference electrode. The used glassy carbon disc electrodes were hand-polished using 0.3 m alumina powder on a polishing cloth followed by successively rinsing with acetone and deionised water, before each experiment or modification. Cyclic voltammetric (CV) measurements were carried out at scan rates of 25, 50, 75, 100, 150, 200, 250 and 300 mV s−1 using a step potential of 2.44 mV. Linear sweep voltammetric (LSV) measurements were carried out at scan rates of 100 mV s−1 . The square wave voltammetric (SWV) measurements were carried out at a frequency of 100 Hz, step potential of 1.2 mV and amplitude of 20 mV. All voltammetric measurements were performed between −0.700 V and some potential value between 1.300 and 1.500 V. All voltammograms were recorded at a constant temperature of 22 ◦ C. Solutions in the cell were purged with oxygen-free nitrogen for 5 min prior to the measurement. A nitrogen blanket was maintained above solutions surface during the measurements. The pH was controlled by a Metrohm pH/mV meter (Model 632) with a glass pH electrode. To homogenize the stock solutions and dispersions, an ultra-sonic bath (Transsonic T660/H Elma® ) was used. 2.4. Preparation of glassy carbon electrode modified with MWCNTs film (MWCNTs–GCE) In the preparation of MWCNTs–GCE, 1 mg MWCNTs was dispersed in 1 mL ultra-pure water containing 1 mg DHP, with the aid of ultra-sonic agitation for about 30 min, to produce a 1 mg mL−1 MWCNTs dispersion. The DHP was used as a dispersing agent to obtain a homogeneous dispersion of multi-walled carbon nanotubes in water and a uniform and stable MWCNTs film over the GCE surface [35]. As previously stated, prior to modification the GCE surface was treated and cleaned. After this, 2–12 L of the MWCNTs dispersion was cast on the GCE surface and dried in a vacuum desiccator. To study the effect of DHP, a DHP–GCE was prepared in a similar way, by casting the GCE surface with a fixed amount of a prepared solution of 1 mg DHP/1 mL ultra-pure water. 2.5. HPLC–UV and LC–ESI-MS/MS analysis High performance liquid chromatography (HPLC) with UV/vis spectrometry detection, using an Elite LaChrom HPLC system equipment (Hitachi, Japan), was used to quantify CBZ in commercial tablets and wastewater samples. The separation was performed in isocratic mode, with a mobile phase composed by 75:25 (%, v/v) acetonitrile:water (acidified with phosphoric acid 0.1%, v/v), at a flow rate of 1.0 mL min−1 , and using a reversed phase analytical column Zorbax Eclipse XDB-C18 with 5 m particle size. The UV detector wavelength was set at 210 nm. LC–ESI-MS/MS analyses of the samples were carried out in a LCQ Advantage ThermoFinnigan mass spectrometer equipped with an electrospray ionisation source and using an ion trap mass analyzer. The conditions of analysis were: capillary temperature of 275 ◦ C, source voltage of 5.0 kV, source current of 100.0 A, and capillary voltage of 15.0 V in positive ion mode. The mass spectrometer equipment was coupled to an HPLC system with autosampler (Surveyor ThermoFinnigan). The analytical column was a reversed phase Thermo Hypersil gold (C18, particle size 5 m, 150 mm × 2.1 mm). The quantification of CBZ was performed with an isocratic program using methanol as eluent A and water acidified with 0.1% (v/v) formic acid as eluent B. The mobile phase was composed by 75% eluent A:25% eluent B (v/v) at a flow rate of 0.3 mL min−1 . The LC–ESI-MS/MS analyses were performed in the SRM (selected reaction monitoring) mode, using the most intense transition described for CBZ (237.0 → 193.9) [19]. Five replicate injections of 20 L were made for each sample previously filtered through a 0.45 m filter. Calibration curves were
Fig. 2. Cyclic voltammetric responses of 6.0 × 10−6 M CBZ in 0.1 M phosphate buffer solution (pH 6.89) at (a) MWCNTs film-coated GCE (8 L of dispersion) and (c) bare GCE. The dotted lines (b) and (d) represent the corresponding blank responses. Scan rate: 100 mV s−1 .
constructed using a set of CBZ standard solutions, prepared with methanol/water (20:80, v/v) or wastewater, with concentrations ranging from 0.25 to 5.0 mg L−1 (1.06–21.16 × 10−6 M). 3. Results and discussion 3.1. Cyclic voltammetric behavior of carbamazepine The electrochemical behavior of CBZ in 0.1 M phosphate buffer with pH 6.89 was investigated by cyclic voltammetry (CV). Fig. 2 depicts the cyclic voltammograms for 6.0 × 10−6 M CBZ at (a) MWCNTs–GCE (8 L of dispersion) and (c) bare GCE. The dotted lines (b) and (d) represent the corresponding blank responses. Within the potential window studied it can be seen that the voltammetric response of CBZ at the MWCNTs–GCE was strongly enhanced, when compared with that of the bare GCE electrode. At this unmodified electrode (Fig. 2c), carbamazepine showed an irreversible behavior with a broad and very weak oxidation peak, around 1.170 V. Above 1.300 V no other CBZ oxidation peak was observed (results not shown), in contrast with previously reported results by Kalanur and Seetharamappa [31]. This was confirmed with further measurements up to concentrations of 15.0 × 10−6 M CBZ. Above this concentration the second oxidation peak, as reported by those authors, begins to be distinguished. Meanwhile, as can be seen from Fig. 2a carbamazepine exhibited at the MWCNTs–GCE a well-defined and very intense irreversible oxidation peak, P+ , around 1.090 V, with no additional post-peak in the forward scan or counterpart peak in the reverse scan. Above this anodic region and at higher CBZ concentrations no further peaks were observable in the voltammograms. The shift of the peak potential to lower anodic potentials and the peak current improvement obtained for CBZ oxidation at the MWCNTs–GCE, in comparison to the GCE response, reflects the decrease of the overpotential for this anodic reaction and the enhancement of the electron transfer between the CBZ molecule and the modified electrode. As expected for these electrodes, these can be ascribed to the strong adsorptive ability and electrocatalytic surface properties of the MWCNTs [35]. On the basis of the few reports about the anodic behavior of the CBZ molecule, the electrode process under discussion can be related with the oxidation of the nitrogen atom in the central ring, which results in the formation of cation radicals (Scheme 1) [31]. Due to the large active surface area, the high electronic reactivity
A. Veiga et al. / Analytica Chimica Acta 674 (2010) 182–189
185
Scheme 1. Probable reaction mechanism for electrooxidation “P+ ” of CBZ at the MWCNTs/GCE (adapted from Ref. [31]).
of the MWCNTs film and the complexity of the oxidation process we believe that the two steps leading to the loss of electrons, as illustrated in Scheme 1, occur simultaneously and cannot be distinguished in the voltammogram. On consecutive cyclic voltammetric sweeps, the electrochemical signals of CBZ at the modified electrode become consecutively poorer (Fig. 3), suggesting that the electrode surface is blocked by the adsorption of the reaction products of CBZ, which reduces the effective reaction sites of the modified electrode surface. As it is
well known, this is a factor that can compromise the voltammetric determination of analytes in routine analysis. However, it was found in this study that the voltammetric response was maintained (RSD = 0.67% for n = 5) if the electrode was kept at an open circuit potential (ocp) for at least 3 min and stirring with a rotation speed of 1000 revolutions min−1 between each run. Similar requirements were deemed necessary for reproducible voltammetric analysis of other tricyclic antidepressant drugs using other electrodes [45]. Following this procedure, the voltammetric determination can be performed with the same modified electrode, without any need for cleaning or discarding it. Probably during the stirring at an open circuit potential, the adsorbed products are released from the electrode surface, and a reproducible quantity of CBZ molecules is accumulated at reactive sites of the MWCNT electrocatalytic layer. Therefore, the voltammetric study of CBZ towards its electrochemical characterization and the analytical quantification was performed with the data obtained from the first voltammetric scan (cyclic voltammetry, CV, or linear sweep voltammetry, LSV). 3.2. Optimization of the experimental variables
Fig. 3. Successive cyclic voltammograms of 6.0 × 10−6 M CBZ at MWCNTs filmcoated GCE. Other conditions are as in Fig. 2.
3.2.1. Influence of the equilibration potential and time In electrochemical studies, where the interaction between the electroactive analyte and the electrode surface can have a strong influence on the quality of the signal to be measured, it is important to fix the potential and time conditions that precede the measurement. In addition to the ocp conditions used, the equilibration potential and time can affect the amount of CBZ that accumulates at the electrode. Bearing this in mind, the effect of both factors on peak current response was also studied. When equilibration poten-
186
A. Veiga et al. / Analytica Chimica Acta 674 (2010) 182–189
Fig. 4. Influence of solution pH on peak potential and peak current of peak P+ . Other conditions are as in Fig. 2.
tial was varied from −0.700 to 0.700 V, it was found that the peak current decreased between −0.500 and 0.700 V (Fig. S1, supplementary data). Hence, an equilibration potential of −0.700 V was adopted. The peak current increased very little when the equilibration time varied from 1 to 300 s (Fig. S1, supplementary data). This indicates that this period of time has little influence on the oxidation peak, when compared with the previous 3 min undertaken at the ocp. As already suggested, it is possible that during this stage, and for a given CBZ solution, the electrode surface becomes saturated with CBZ molecules. To avoid overextending the time of analysis, an equilibration time of 1 s was chosen. 3.2.2. Influence of pH The effect of pH on the anodic response (P+ ) of CBZ at the modified electrode was tested over the pH range 2.0–11.8 using the same supporting electrolyte solution. Within this pH range, the peak potential Ep,a is almost pH independent between pH 2 and 5, and above this value is shifted to less positive values as the pH of the solution increases (Fig. 4). The greater variations occur between pH 5.14 and 7.25 and for pH > 11. In the pH range 5.1–7.3, the Ep,a varies linearly with pH, with a negative slope of 27.9 mV pH−1 . This indicates that the number of protons released from the molecule, in this anodic process and at this pH range, seems to be half the number of electrons transferred to the electrode [45,46]. This is in agreement with the proposed mechanism by Kalanur and Seetharamappa [31] (Scheme 1), since four electrons and two protons are involved, when one dicationic radical dimer is produced. As can be also observed, the peak current is very sensitive to pH changes between pH 5.1 and 8.0, achieving a maximum value around 6.9 (the pH of the phosphate buffer used as supporting electrolyte). Outside this pH range the peak current decreases significantly when pH ≥ 10. Therefore, further studies were performed at pH 6.89. 3.2.3. Influence of the scan rate on peak P+ The dependence of peak potential, Ep , as well as peak current, ip , on the scan rate () were studied in the range 25–300 mV s−1 . The Ep is a function of the scan rate, shifting to a positive direction, and the plot of Ep vs. log (Fig. 5) was linear. It was also found that the peak current, ip , is proportional to the square root of the scan rate (ip (A) = 28.0611/2 (V s−1 )1/2 − 1.6687; R2 = 0.9909; the relation ip vs. was not linear, presenting a R2 = 0.9765). These diagnostic tests prove that the electrode reaction under study is irreversible [47]. The linear relationship found between ip and 1/2 suggests that the electrooxidation is a diffusion-controlled process. However, a lin-
Fig. 5. Cyclic voltammograms of [CBZ] = 6.0 × 10−6 M on MWCNTs-modified GCE at different scan rates (V s−1 ): (a) 0.025; (b) 0.050; (c) 0.075; (d) 0.100; (e) 0.150; (f) 0.200; (g) 0.250; (h) 0.300. Other conditions are as in Fig. 2. The inset shows the plot of anodic peak potential (Ep ) vs. log . The error bars of Ep values are too short to be represented.
ear relationship with a slope close to 0.62, observed between log ip and log (Fig. 6, with ip and expressed in A and V s−1 , respectively) seems to be indicative of some adsorption contributions to the electrode reaction. It can be said that this reaction at the modified GCE, is a “mixed” diffusion–adsorption process [47–49]. This is also in agreement with the proposed mechanism by Kalanur and Seetharamappa [31]. According to these authors, one of the two oxidation reactions is a diffusion-controlled process and the other is an adsorption-limited process. As already noted, at the MWCNTs–GCE, the electrode reaction is a “mixed” diffusion–adsorption process because the two irreversible oxidation processes seem to occur together in a very short time. This, in fact, can be confirmed by square wave voltammetry (SWV) measurements (Fig. 7) as the forward and backward components of the current present only one peak with same height and sign, resulting in an overall voltammogram showing no peaks [50]. 3.2.4. Influence of the amount of MWCNTs/DHP dispersion Fig. 8 illustrates the relationship between the volume of 1 mg mL−1 dispersion used to modify the GCE surface and the anodic peak current of CBZ. It was found that the peak current increased in general with an increasing volume of the dispersion.
Fig. 6. Plot of log ip vs. log . Same conditions as reported in Fig. 5.
A. Veiga et al. / Analytica Chimica Acta 674 (2010) 182–189
Fig. 7. Square wave voltammetric response of 2.7 × 10−6 M CBZ in 0.1 M phosphate buffer solution (pH 6.89) at MWCNTs film-coated GCE (10 L of dispersion). it – total current; if – forward current; and ib – backward current. Effective scan rate: 120 mV s−1 .
Above 10 L the peak current changed only slightly. Since the signal was not improved above this quantity, and because the dispersion spreads beyond the limits of the conductive surface of GCE electrode, the volume of the dispersion chosen in the analytical determination of CBZ was 10 L. 3.3. Calibration curve
Fig. 9. Calibration curve of carbamazepine established by LSV, under the optimized experimental conditions. Between each measurement the modified electrode was maintained at ocp during 3 min and constant rotation of 1000 rpm. Table 1 Comparison of detection limits for CBZ at different electrodes. Electrodes
Detection limits/nM
Refs.
HMDEa HMDEa AgNPs/SPCEb GCEc MWCNTs/GCEd
1500 1000 938 1800 40
[5] [28] [30] [31] Present work
a b
The calibration curve for CBZ (Fig. 9) was established by linear sweep voltammetry (LSV), under the optimized experimental conditions described above. A scan rate of 100 mV s−1 was used. At these conditions, the plot of peak current, ip , vs. CBZ concentration was found to be linear in the range 0.13–1.60 M (30.7–378 g L−1 ). The limits of detection (LOD) and quantification (LOQ) of CBZ were calculated using the typical equations LOD = 3 s m−1 and LOQ = 10 s m−1 (where s is the standard deviation of the peak currents of the blank (five runs) and m is the slope of the calibration curve). The LOD and LOQ obtained were 40 and 140 nM (9.5 and 33.1 g L−1 ), respectively. To the best of our knowledge, this LOD is the lowest that has been reported for carbamazepine using electrochemical techniques (Table 1). At higher scan rates, the linearity between the peak current, ip , and CBZ concentration became unsatisfactory.
187
c d
Hanging mercury drop electrode. Silver nanoparticle-modified carbon screen-printed electrode. Glassy carbon electrode. Multi-walled carbon nanotubes-modified glassy carbon electrode.
3.4. Interferences Table 2 lists the influence of other pharmaceutical substances, which possibly can occur in wastewater [10], on the oxidation signal of CBZ. Under the optimum experimental conditions, it was found that diclofenac and clofibric acid are the strongest interferents. Both pharmaceuticals enhanced the oxidation signal of CBZ, without distorting its shape. In the other three cases, the effect in the peak current response of CBZ was substantially more reduced. For all pharmaceuticals tested, the interferences remained at the same level when the concentration of the compounds was two times the CBZ concentration. Among the tested substances, diclofenac was the only pharmaceutical that produced an additional voltammetric peak (at 0.630 V approx.) at the modified GCE, as noted by Yang et al. [41]. Attending to these effects, it is advisable to perform the analytical determination of CBZ at real samples using the standard addition method, instead of the calibration curve method. 3.5. Analytical determination of CBZ The MWCNTs–GCE was used to determine CBZ in more realistic samples such as tablets and in spiked and unspiked wastewater Table 2 Effect of potential interferents on the voltammetric response of 0.150 M CBZ. Potential interferents in wastewaters (0.05 M)
Fig. 8. Peak current of P+ vs. amount of MWCNT/DHP dispersion on the GCE surface. [CBZ] = 6.0 × 10−6 M in 0.1 mol L−1 phosphate buffer solution (pH 6.89). Same voltammetric conditions as in Fig. 2. The plot includes the equivalent anodic response of CBZ at bare GCE.
Clofibric acid Ibuprofen Atenolol Diclofenac sodium Oxcarbazepine
Signal change (%) ±SD +6.5 +1.6 −3.2 +13.8 −2.8
± ± ± ± ±
1.7 0.6 1.0 2.3 1.2
188
A. Veiga et al. / Analytica Chimica Acta 674 (2010) 182–189
Fig. 10. Standard additions calibration curves for the determination of CBZ in () Tegretol® tablet, spiked wastewater (sww) sample (䊉) and unspiked wastewater (ww) sample (). Each curve shows the analytical signal (peak height) obtained with the respective sample solution at the modified GCE (10 L of MWCNTs dispersion), before and after successive standard additions of 47.3 g mL−1 CBZ working solution. Same voltammetric conditions as in Fig. 9. The inset figure shows the linear sweep voltammograms obtained with the spiked wastewater (sww) sample (without and with the six standard additions (solid lines)). The dashed line is the linear sweep voltammetric signal of the unspiked wastewater sample. Table 3 Wastewater quality parameters of samples used in this work, according to the APHAAWWA-WPCF methods [44]. Chemical characteristics
Values
pH TSS (mg L−1 )a CODt (mg L−1 )b CODs (mg L−1 )c
8.29 57 133 82
a b c
± ± ± ±
0.05 3 2 2
Total suspended solids. Total chemical oxygen demand. Soluble chemical oxygen demand.
samples, by LSV. In addition to carbamazepine as the active ingredient, composition of tablets used in the determination of CBZ includes other inactive ingredients, namely cellulose compounds, dextrates, iron oxides, magnesium stearate, mannitol, polyethylene glycol, sodium lauryl sulfate and titanium dioxide. The wastewater samples were also characterized in terms of some of the most common parameters of wastewater quality (Table 3). In general, organic matter load and suspended solids were somewhat high but still within the Portuguese legal limits for discharge into water bodies. The content of CBZ in all samples was obtained using the standard addition method, at the optimized experimental conditions. In the tablet assays, a suitable aliquot (15 L) of Tegretol® stock solution was diluted in 15 mL of 0.1 M phosphate buffer (pH 6.89). The standard additions were made with the working solution (200 M = 47.3 g mL−1 ) of CBZ. In the wastewater assays, the same procedure was repeated. In this case, an appropriate aliquot
(1.5 mL) of the unspiked or spiked wastewater sample (treated by SPE) was diluted in 13.5 mL of supporting electrolyte solution. Fig. 10 shows the standard addition curves obtained for the Tegretol® solution and the wastewater samples. In Table 4 the results obtained in the voltammetric analysis of five replicates from each sample are summarized. In addition, HPLC–UV and LC–ESIMS/MS were also used in the quantification of CBZ to show the importance and accuracy of this voltammetric method. The results obtained with the developed sensor were in good agreement with the content of CBZ advertised in the label of the tested commercial formulation, and with the known content of the spiked wastewater samples. The good recoveries obtained indicate that the proposed modified electrode and the voltammetric method has a great potential for the practical sample analysis of CBZ. The complex chemical composition of both the tablet ingredients and the wastewater components (after SPE pretreatment of the wastewater samples) do not significantly affect the determination of CBZ. In the analysis of the Tegretol® solution, better recoveries were obtained using this method, when compared with the other chromatographic methods. In the case of the spiked wastewater samples, the concentrations of CBZ obtained using HPLC and LC–MS were substantially higher than the known CBZ content. This fact points out one drawback of these techniques, which is the occurrence of matrix effects. These effects are caused by the high susceptibility of the ESI source and UV detector to other components present in the matrix, which may result in the suppression or enhancement of the signal. As referred above, in order to minimize the influence of interferents in the analytical determination of CBZ at real samples, the standard
Table 4 Determination of carbamazepine in Tegretol® , unspiked and spiked wastewater samples. Sample
Tegretol® 200 Spiked wastewater Unspiked wastewater a b c
RSD = 2.19%. RSD = 5.71%. n.d. – not detected.
CBZ content
200 mg tablet−1 0.500 g mL−1 Unknown
Experimental value ± SD found by
Recovery in this method (%)
This method
HPLC
LC–MS
201.1 ± 4.4a 0.479 ± 0.027b n.d.c
185.2 ± 6.6 0.588 ± 0.036 n.d.c
182.1 ± 4.9 0.558 ± 0.030 n.d.c
100.6 95.8 –
A. Veiga et al. / Analytica Chimica Acta 674 (2010) 182–189
addition method is a suitable alternative. However, that technique is more easily performed and less time consuming using voltammetric methods than in chromatographic methods.
[11] [12] [13] [14]
4. Conclusions
[15] [16] [17]
The MWCNTs–GCE and the voltammetric method presented in this work provide a good technique for the determination of CBZ with good repeatability, reproducibility, and recovery along with a very low detection limit. This electroanalytical technique is a promising alternative to the often reported chromatographic methods owing to its simplicity, rapidity, reliability and low cost of analysis. The multi-walled carbon nanotubes-modified GCE electrode exhibited a markedly electrocatalytic activity towards the oxidation of CBZ, leading to a considerable improvement in the signal response with the effect of significantly lowering the detection and quantification limits in comparison with other voltammetric techniques. The method has been applied with success to the determination of CBZ in real pharmaceutical dosage forms as well as spiked and unspiked wastewaters. For that reason, the authors believe that this can be a valuable analytical tool in quality control for the pharmaceutical industry as well in environmental applications. In the future, it is planned to apply the proposed methodology in the determination of carbamazepine in biological fluids. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.aca.2010.06.031. References [1] J.B. Leikin, F.P. Paloucek, Poisoning and Toxicology Handbook, 4th ed., Informa Health Care, New York, 2008. [2] L.J. Doods (Ed.), Drugs in Use: Clinical Case Studies for Pharmacists, 4th ed., Pharmaceutical Press, London, 2010, p. 357. [3] Y. Zhang, S.U. Geißen, C. Gal, Chemosphere 73 (2008) 1151–1161. [4] K. Kümmerer, A. Schuster, in: K. Kümmerer (Ed.), Pharmaceuticals in the Environment: Sources, Fate, Effects and Risks, 3rd ed., Springer, 2008, p. 46. [5] L. Campanella, A. Ambrosi, F. Bellanti, M. Tomassetti, Curr. Anal. Chem. 2 (2006) 229–241. [6] M. Clara, N. Kreuzinger, B. Strenn, Water Res. 38 (2004) 947–954. [7] K. Fent, A. Weston, D. Caminada, Aquat. Toxicol. 76 (2006) 122–159. [8] K. Stamatelatou, V. Vavilin, G. Lyberatos, Bioresour. Technol. 88 (2003) 131–136. [9] M. Clara, N. Kreuzinger, B. Strenn, O. Gans, H. Kroiss, Water Res. 39 (2005) 97–106. [10] V. Matamoros, C. Arias, H. Brix, J.M. Bayona, Water Res. 43 (2009) 55–66.
[18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44]
[45] [46] [47] [48] [49] [50]
189
T.A. Ternes, Water Res. 32 (1998) 3245–3260. H. Heberer, Toxicol. Lett. 131 (2002) 5–17. T. Heberer, J. Hydrol. 266 (2002) 175–189. L. Hu, H.M. Martin, O. Yarce-Bulted, M.N. Sugihara, K.A. Keating, T.J. Strathmann, Environ. Sci. Technol. 43 (2009) 509–515. Y. Kouno, C. Ishikura, M. Homma, K. Oka, J. Chromatogr. B 695 (1997) 349–353. K.M. Patil, S.L. Bodhanka, J. Pharma. Biomed. Anal. 39 (2005) 181–186. L. Hua, M. Delgado, S.T. Iannaccone, L.J. Forman, C.M. Eggers, Ther. Drug Monit. 15 (1993) 317–327. J.L. Maggs, M. Pirmohamed, N.R. Kitteringham, B.K. Park, Drug Metab. Dispos. 25 (1997) 275–280. N.M. Vieno, T. Tuhkanen, L. Kronberg, J. Chromatogr. A 1134 (2006) 101–111. G.F. van Rooyen, D. Badenhorst, K.J. Swart, H.K. Hundt, T. Scanes, A.F. Hundt, J. Chromatogr. B 769 (2002) 1–7. A. Frigerio, K.M. Baker, P.L. Morselli, Adv. Biochem. Psychopharmacol. 7 (1973) 125–134. R.J. Perchalski, B.J. Wilder, Clin. Chem. 20 (1974) 492–493. T.A. Ternes, TrAC Trends Anal. Chem. 20 (2001) 419–434. C. Wang, Z. Wang, D. Han, Q. Wu, X. Zang, Anal. Lett. 39 (2006) 1927–1939. T. Comoglu, N. Gönül, Turk. J. Med. Sci. 35 (2005) 217–221. M.S. Cámara, C. Mastandrea, H.C. Goicoechea, J. Biochem. Biophys. Method 64 (2005) 153–166. D.J. Turk, S.A. McClintock, W.C. Purdy, Anal. Lett. 18 (1985) 2605–2618. Z.Q. Zhang, S.Z. Chen, W.L. Huang, F. Xu, Yao Xue Xue Bao 28 (1993) 312–316. G.S. Moses, K.M. Rao, S. Rao, A. Ramachandraiah, J. Indian Chem. Soc. 72 (1995) 333–337. M.A. García-García, O. Domínguez-Renedo, A. Alonso-Lomillo, M.J. ArcosMartínez, Sensor Lett. 7 (2009) 586–591. S.S. Kalanur, J. Seetharamappa, Anal. Lett. 43 (2010) 618–630. S. Lu, K. Wu, X. Dang, S. Hu, Talanta 63 (2004) 653–657. A. Merkoc¸i, M. Pumera, X. Lopis, B. Pérez, M. del Valle, S. Alegret, TrAC Trends Anal. Chem. 24 (2005) 826–838. Z. Lin, G. Chen, Talanta 70 (2006) 111–115. ˜ ˜ M. José, Pingarrón, Anal. Chim. Acta 622 (2008) 11–47 L. Agüí, P. Yánez-Sede no, (and references cited therein). R.N. Goyal, S.P. Singh, Talanta 69 (2006) 932–937. C. Li, Microchim. Acta 157 (2007) 21–26. A.R.N. Hegde, N.P. Shetti, S.T. Nandibewoor, Talanta 79 (2009) 361–369. H. Jiang, X.-M. Wang, Electrochem. Commun. 11 (2009) 126–129. Y. Wu, S. Ye, S. Hua, J. Pharm. Biomed. Anal. 41 (2006) 820–824. X. Yang, F. Wang, S. Hu, Mater. Sci. Eng. C 28 (2008) 188–194. X. Wen, J. Fei, X. Chen, L. Yi, F. Ge, M. Huang, Environ. Pollut. 156 (2008) 1015–1020. K. Balasubramanian, M. Burghard, Anal. Bioanal. Chem. 385 (2006) 452–468. L.S. Clescerl, A.E. Greenberg, A.D. Eaton (Eds.), Standard Methods for the Examination of Water and Wastewater, American Public Health Association, Washington, DC, 1998. T.A. Ivandini, B.A. Sarada, C. Terashima, T.N. Rao, D.A. Tryk, H. Ishiguro, Y. Kubota, A. Fujishima, J Electroanal. Chem. 521 (2002) 117–126. P.T. Kissinger, W.R. Heineman, Laboratory Techniques in Electroanalytical Chemistry, 2nd ed., Marcel Dekker, New York, 1996. Southampton Electrochemistry Group, Instrumental Methods in Electrochemistry, 1st ed., Ellis Horwood, London, 1985. A.J. Bard, L.R. Faulkner, Electrochemical Methods, Fundamentals and Applications, 2nd ed., Wiley, New York, 2001. K.G. David, Cyclic Voltammetry Simulation and Analysis of Reaction Mechanisms, VCH Publishers, New York, 1993. V. Mirceski, S. Komorsky-Lovric, M. Lovric, Square-wave Voltammetry: Theory and Applications, Springer, Berlin, 2007.