a n a l y t i c a c h i m i c a a c t a 6 2 6 ( 2 0 0 8 ) 1–9

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/aca

Electrochemical detection of celecoxib at a polyaniline grafted multiwall carbon nanotubes modified electrode Kalayil Manian Manesh a , Padmanabhan Santhosh a , Shanmugasundaram Komathi a , Nam Hee Kim a , Jong Wook Park a , Anantha Iyengar Gopalan a,b , Kwang-Pill Lee a,b,∗ a b

Department of Chemistry Graduate School, Kyungpook National University, Daegu 702-701, South Korea Nano Practical Application Center, Daegu 704-230, South Korea

a r t i c l e

i n f o

a b s t r a c t

Article history:

A modified electrode is fabricated by grafting polyaniline (PANI) chains onto multiwall

Received 25 May 2008

carbon nanotubes (MWNTs) and utilized for the adsorptive reduction of celecoxib (CEL).

Received in revised form

PANI-g-MWNTs modified electrode appreciably enhances the sensitive detection of CEL in

25 July 2008

extremely lower concentrations (1 × 10−11 M). Square wave stripping voltammogram (SWSV)

Accepted 28 July 2008

shows a reduction peak at −1.08 V with a high peak current for SW frequency of 100 Hz,

Published on line 8 August 2008

amplitude of 25 mV and step height of 6 mV. The high surface area of PANI-g-MWNTs is effectively utilized for the adsorption of CEL to preconcentrate at the electrode. The PANI

Keywords:

chains covalently linked to MWNTs mediate the electron transfer processes. The present

Carbon nanotubes

finding open-up the scope for extending on the use of other conducting polymers grafted

Polyaniline

MWNTs modified electrodes for the detection of compounds that do not have surface-active

Electrochemical grafting

properties at conventional electrodes.

Modified electrode

© 2008 Elsevier B.V. All rights reserved.

Celecoxib Square wave stripping voltammetry

1.

Introduction

Non-steroidal anti-inflammatory drugs (NSAIDs) have been widely used to alleviate symptoms of arthritis, arthritisassociated disorders, fever, postoperative pain, migraine, and so on [1]. NSAIDs vary in their potency duration of action and mechanism of removal from the body. NSAIDs induce ulcer and life-threatening toxicity in gastrointestinal tract. Prostaglandins are the related chemicals that cause such side effects. Prostaglandins are produced within the body’s cell by the enzyme cyclooxygenase (COX). There are two kinds of COX, cyclooxygenase-1 (COX-1) and cyclooxygenase-2 (COX-2). Out of the two isoforms of COX, COX-1 is responsible for mediating the production of prostaglandin. COX-2

is primarily associated with pain, fever and inflammation [2]. The more a NSAID blocks COX-1, greater is the chance of the drug to induce ulcer. COX-2 selective NSAIDs have minimum drug related side effects since they spare COX-1 activities. Celecoxib (CEL), 4-[5-(4-methylphenyl)-3(trifluoromethyl)-1H-pyrazol-1-yl]benzenesulfonamide, is one of the selective inhibitor for COX-2, and used for the relief from symptoms of ankylosing spondylitis [3]. It is important to develop simple and suitable analytical methods for the quantitative determination of CEL. A survey of literature reveals that several analytical methods have been utilized for the quantitative determination of CEL in pharmaceutical formulation and human plasma. These methods include voltammetry [4], spectrophotometry

∗ Corresponding author at: Department of Chemistry Graduate School, Kyungpook National University, Daegu 702-701, South Korea. Tel.: +82 53 950 5901; fax: +82 53 95 28104. E-mail address: [email protected] (K.-P. Lee). 0003-2670/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2008.07.050

2

a n a l y t i c a c h i m i c a a c t a 6 2 6 ( 2 0 0 8 ) 1–9

[5], thin-layer chromatography [6], liquid chromatography [7], solid-phase extraction and high-performance liquid chromatography [8], micellar electro kinetic chromatography [9], liquid chromatography–mass spectrometry [10], high-performance liquid chromatography with UV [11] and fluorimetric [12] detections. Among the various methods, electrochemical methods have attracted interest because of their quick response, high sensitivity and the possibilities for miniaturization. However, reports on the use of nanomaterials like carbon nanotubes (CNTs) towards the determination of CEL are scarce. CNTs are one of the attractive nanomaterials due to their specific electronic, chemical, and mechanical properties. CNTs can facilitate electron transfer between the electroactive species and electrodes and are utilized for the fabrication of modified electrodes. Recently, CNTs based modified electrodes have been used for the sensitive detection of a few biologically important analytes such as ascorbic acid [13], hydrogen peroxide [14], glucose [15,16], etc. Electrochemical studies have clearly demonstrated that CNTs based modified electrodes could exhibit a sensitive and highly selective electrochemical detection [13–18]. Further, CNTs could minimize the electrode fouling [14]. The CNTs based electrodes have been utilized for the determination of few drugs. The high surface area of CNTs is beneficial to preconcentrate the drugs through adsorption, before the electrochemical detection. Nanocomposites of CNTs with polymers have been first reported by Ajayan et al. [19]. After that, flurry of research activities have then been directed on the preparation of nanocomposites combining CNTs with varieties of polymers. Functional nanocomposites have been prepared with superior properties. The use of CNTs-conducting polymer (CP) composites as a layer has been demonstrated to improve the electrochemical performance for the sensitive detection of several analytes that include carbon monoxide, and oxygen [14,18,19]. Importantly, CNTs/CP composite have shown to possess properties of the individual components with synergistic effects. CNTs/CP composites modified electrodes were generally prepared by dispersing CNTs onto the surface of an electrode and subsequently depositing a layer of CP over the CNTs’ layer using any of the electrochemical methods. As a consequence, CNTs/CP modified electrodes have a bilayer configuration with a layer of CP existing over the layer of CNTs [20–22]. In the bilayer configuration, the top formed layer of CP is expected to block the functional capabilities of CNTs. As a result, the characteristics of CNTs could not be fully exploited for electrochemical performance. However, if chains of CP are covalently linked to the surface of CNTs, the CP linked CNTs could exhibit synergistic properties of CNTs and CP. The chains of CP could be covalently linked to CNTs by the process known as “grafting”. CNTs grafted with chains of CP are expected to have the synergistic effects of CNTs as well as the CP. Keeping these prospective in view, modified electrodes were fabricated by grafting the chains of different kinds of CPs onto the surface of CNTs [14,17]. Recently, we have demonstrated that the modified electrode fabricated by poly(diphenylamine) (PDPA) grafted multiwall carbon nanotubes (MWNTs) could exhibit superior electrochemical characteristics over the MWNT/PDPA bilayer electrode [23].

In the present study, we have fabricated a modified electrode, PANI-g-MWNT-ME (ME represents modified electrode) by modifying the surface of indium tin oxide (ITO) coated glass electrode with PANI-g-MWNT. The PANI-g-MWNT-ME was utilized for the electrochemical detection of CEL. The high surface area of PANI-g-MWNTs was effectively utilized to preconcentrate CEL at the electrode through prior adsorption. The covalently linked PANI units in PANI-g-MWNTs are expected to mediate the electron transfer processes. Thus, PANI-g-MWNT-ME possesses the combined properties of PANI and MWNTs and exhibits superior performance for the electrochemical detection of CEL. Also, films of MWNTs and PANI were formed successively to fabricate the (bilayer) electrode over the surface of ITO, MWNT/PANI-ME. The electrochemical detection of CEL at the MWNT/PANI-ME was also assessed and compared with PANI-g-MWNT-ME. The sensitivity and selectivity of PANI-g-MWNT-ME were found to be much superior than at the MWNT/PANI (bilayer) electrode. The details are discussed.

2.

Experimental

2.1.

Chemicals

MWNTs with diameters in the range of 40–60 nm, length: 0.5–500 ␮m, surface area: 40–390 m2 g−1 were obtained from CNT Co., Ltd., Incheon, Korea. The purity was 95 wt.%, and the impurities included in MWNTs were iron, aluminum, cobalt, nickel and amorphous carbon. MWNTs were rinsed well with double-distilled water and dried. All the other chemicals were obtained from Aldrich and used without any purification. Celecoxib (CEL), under the brand name celebrex, was obtained from SEARLE® . Britton–Robinson buffer (BRb) solution was prepared by mixing solutions of phosphoric acid, acetic acid and boric acid (0.04 M, each) with the appropriate volumes of a solution of 0.2 M sodium hydroxide. Deionised water was used for all the preparations and measurements.

2.2.

Preparation of amine functionalized MWNTs

The amine functionalized MWNTs (MWNT-NH2 ) were prepared using sequential chemical modifications of MWNTs [24]. Typically, 50 mg of MWNTs were refluxed in 4 M HNO3 for 24 h and filtered through a polycarbonate membrane (pore size: 0.2 ␮m). The residue, MWNT-COOH (carboxylated MWNT), was washed with water and dried under vacuum at 60 ◦ C for 12 h. 50 mg of MWNT-COOH was refluxed in 100 ml of thionyl chloride at 65 ◦ C for 24 h. The residue (MWNTCOCl) was filtered, washed with tetra hydrofuran (THF) and dried under vacuum at room temperature. MWNT-NH2 was prepared by refluxing MWNT-COCl with O,O -bis(3aminopropyl)polyethylene glycol in THF at 60 ◦ C for 24 h. The residue (MWNT-NH2 ) was washed with water and dried under vacuum.

2.3.

Fabrication of modified electrodes

Modified electrodes such as PANI-g-MWNT-ME, MWNT/PANIME (bilayer), and PANI-ME were fabricated as detailed below.

a n a l y t i c a c h i m i c a a c t a 6 2 6 ( 2 0 0 8 ) 1–9

2.3.1.

PANI-g-MWNT-ME

PANI chains were grafted onto the walls of MWNTs through electrochemical polymerization of aniline in the presence of MWNT-NH2 . Cyclic voltammetry was employed to deposit PANI grafted MWNTs (PANI-g-MWNTs) as film on the surface of working electrode, ITO (1 cm × 2 cm; surface resistance: 10 ). A typical procedure for the electrochemical deposition of PANI-g-MWNTs [14] is outlined. 50 mg of MWNT-NH2 was dispersed in 0.5 M cetyltrimethyl ammonium bromide (CTAB) using ultrasonication (BRANSON Digital Sonifier) for 3 h. 10 mM of aniline (in 0.1 M H2 SO4 ) was added to the above solution. The solution containing MWNT-NH2 and aniline was subjected to electrochemical polymerization by continuous cycling of potentials in the range of 0–900 mV. A green coloured deposit (PANI-g-MWNTs) was seen on the surface of ITO.

2.3.2.

MWNT/PANI-ME and PANI-ME

A black suspension of MWNTs was prepared by dispersing 50 mg of MWNTs (unfunctionalized) in 0.5 M CTAB with the aid of ultrasonication (3 h). About 500 ␮l of the black suspension was dropped on the surface of ITO electrode and dried at 60 ◦ C for 12 h. The electrode was then rinsed with water and stored under nitrogen atmosphere. PANI was electrochemically deposited onto MWNTs coated ITO by electrochemical polymerization of aniline (10 mM in 0.1 M H2 SO4 ). Thus, the bilayer modified electrode, MWNT/PANI-ME was fabricated. PANI-ME was fabricated by depositing PANI film on the surface of ITO electrode from the solution containing aniline (10 mM) in 0.5 M CTAB and 0.1 M H2 SO4 . In all the cases, area of the electrode was kept as 1 cm2 . After fabrication, the electrodes were held at a potential of +900 mV for 30 s in the background electrolyte solution. Then, the electrodes were removed, washed with water and stored in BRb (pH 7) under nitrogen atmosphere.

2.4.

Instrumentation

The morphology of the modified electrodes was examined by field emission scanning electron microscope (FESEM)-Hitachi S-4300. The room-temperature electrical conductivity of the films was determined using Jandel four-probe conductivity meter (of 0.3% accuracy). Elemental compositions of pristine MWNTs and MWNT-NH2 (% of carbon, nitrogen, hydrogen) were determined by Fision EA H10 elemental analyzer equipped with flash combustion furnace. Cyclic voltammetry (CV), electrochemical impedance spectroscopy (EIS), constant potential chronoamperometry, differential pulse stripping voltammetry (DPSV) and square wave stripping voltammetry (SWSV) measurements were performed by using EG&G PAR Electrochemical Analyzer with Frequency Response Analyzer 1025 in the electrochemical cell consist of modified electrode as working electrode, Ag/AgCl as reference and a platinum wire as auxiliary electrodes. All electrochemical experiments were performed at 25 ◦ C inside the Faraday cage in order to minimize the contribution of background noise to the analytical signal.

2.5.

Electrochemical determination of CEL

Stock solutions of CEL were prepared in methanol and stored in dark bottles at 4 ◦ C. The stock solutions were diluted using

3

methanol to prepare the required concentration of CEL. Solutions of CEL were prepared afresh before electrochemical measurements. Electrochemical analysis of the CEL was carried out as follows. An aliquot of CEL solution in BRb was taken in an electrochemical cell and electrochemical measurements were performed. The optimum operating conditions used for differential pulse stripping voltammetry (DPSV) experiments were: scan rate = 10 mV s−1 and pulse amplitude = 25 mV. Square wave stripping voltammetry (SWSV) experiments were carried out with a frequency of 100 Hz (f), scan increment (dE) of 6 mV and SW pulse amplitude (E) as 25 mV.

3.

Results and discussion

3.1.

Fabrication of modified electrodes

Fig. 1a displays the cyclic voltammograms (CVs) recorded during the electro-deposition of PANI-g-MWNTs onto the surface of ITO electrode. The CVs exhibit four anodic waves around 200, 510, 600 and 760 mV with cathodic counterparts around 650, 500, 420 and 100 mV, respectively. The peak at 200 and 510 mV are assigned to the formation of radical cations and benzoquinone, respectively. The peaks at 600 and 760 mV correspond to the oxidation of head to tail dimers (intermediate) and conversion of emeraldine to pernigraniline structure, respectively [25,26]. Redox features of PANI-g-MWNT-ME, MWNT/PANI-ME, PANI-ME are compared (Fig. 1a–c). The redox characteristics representing the growth of PANI-g-MWNTs on the ITO were entirely different from the formation of PANI over MWNTs coated ITO (Fig. 1b) and the growth of simple PANI (Fig. 1c). The redox features of PANI in MWNT/PANI-ME and PANI-ME are similar to pristine PANI reported in literature [25,26]. The changes in redox features of PANI in PANI-g-MWNT-ME as compared to pristine PANI are attributed to the following reasons. The amine groups in MWNT-NH2 as well as aniline were simultaneously oxidized during the electro-polymerization of a solution containing both MWNT-NH2 and aniline. The crossreactions between the amine cation radicals formed from aniline as well as –NH2 groups in MWNT-NH2 resulted in grafting of PANI chains onto MWNTs. Field emission transmission electron microscopy (FESEM) image of PANI-g-MWNT-ME adds further evidence for the grafting of PANI chains onto MWNTs. Fig. 2 shows the FESEM images of PANI-g-MWNT-ME (a), MWNTs coated ITO (b), MWNT/PANI-ME (c) and PANI-ME (d). FESEM image of PANI-gMWNT-ME reveals that PANI exists as a layer over the surface of MWNTs. The average thickness of PANI was determined to be 30–50 nm (Fig. 2a). Importantly, nearly a uniform layer of PANI was found on MWNTs because of grafting of PANI chains onto MWNTs. However, one could observe a dense coverage of PANI layers in the FESEM image of MWNT/PANI-ME. As a result, MWNTs could not be seen on the surface (Fig. 2b). The morphology of PANI-g-MWNT-ME in which PANI exists as a layer over the MWNTs is beneficial to derive synergistic electrochemical influence from PANI and MWNTs as compared to MWNT/PANI-ME (bilayer electrode) [20–22]. Besides the morphology, electrical conductivity also plays a key role

4

a n a l y t i c a c h i m i c a a c t a 6 2 6 ( 2 0 0 8 ) 1–9

Fig. 1 – Cyclic voltammograms recoded during the deposition of (a) PANI-g-MWNT over ITO, (b) PANI over MWNTs coated ITO and (c) PANI over ITO eletrodes (even number of cycles are presented).

in deciding the electrochemical characteristics of the modified electrodes and hence conductivity measurements were made. The conductivity values of PANI-ME, MWNT/PANI-ME and PANI-g-MWNT-ME were 3.4, 15.32 and 128.3 mS cm−1 , respectively. Interestingly, the electrical conductivity of PANIg-MWNT-ME is higher than that of MWNT/PANI-ME and PANI-ME. In order to gain insight into the extent of grafting of PANI chains onto MWNTs, elemental analysis of pristine MWNTs and MWNT-NH2 was made. MWNT-NH2 showed 5.2% of nitrogen with carbon to nitrogen ratio as 13:1 (i.e. one carbon out of 13 carbon atoms is linked to nitrogen); while pristine MWNTs did not contain nitrogen (only low value of <0.05% was obtained). The amine sites in MWNT-NH2 are the nucleating sites for grafting of PANI chains. Hence, a maximum of 8% grafting of PANI onto MWNTs is expected. However, the exact grafting % of PANI onto MWNTs could not be determined, probably due to presence of different lengths of PANI chains in the grafted units.

3.2. Electrochemical characteristics of PANI-g-MWNT-ME Electrochemical behaviour of the three electrodes, (PANI-gMWNT-ME, MWNT/PANI-ME and PANI-ME) were compared

using ferricyanide (FeIII (CN)6 3− ) as an electrochemical marker. Fig. 3 shows the CVs of the three electrodes in the aqueous solution containing 5 mM K3 Fe(CN)6 and 0.1 M KCl (scan rate: 20 mV s−1 ). Oxidation/reduction characteristics of Fe(CN)3−/4− could be seen in all the cases (Fig. 3). However, higher peak separation between anodic and cathodic processes was observed at all the electrodes for a low scan rate. In general, the anode-cathode peak separation (Ep ) for a reversible oneelectron transfer reaction is in the range of 59 mV. However, the electrodes under investigation gave Ep of much higher over 59 mV (Fig. 3). The larger Ep might originate from the uncompensated bulk resistances in the films. Interestingly, PANI-g-MWNT-ME had an Ep value of 230 mV, which is the lowest compared to PANI/MWNT-ME and PANI-ME. Also, the redox peak current at PANI-g-MWNT-ME is much higher than that at MWNT/PANI-ME and PANI-ME. The effective surface area of PANI-g-MWNT-ME was determined and compared with PANI-ME electrode from the CVs obtained in potassium hexacyanoferrate(III) solution [27]. The surface area of PANI-ME and PANI-g-MWNT-ME was found to be 0.36 and 12.68 cm2 , respectively. Furthermore, electrochemical impedance spectra (EIS) of the modified electrodes were recorded in BRb in the frequency range of 100 Hz to 0.1 MHz at amplitude of 10 mV.

5

a n a l y t i c a c h i m i c a a c t a 6 2 6 ( 2 0 0 8 ) 1–9

Fig. 2 – FESEM images of (a) PANI-g-MWNT-ME, MWNTs coated ITO electrode before (b) and after (c) the deposition of PANI and (d) PANI deposited on ITO electrode.

Table 1 – Electrochemical parameters collected from Nyquist plot in BRb (pH 7) for the modified electrodes Electrodes PANI-g-MWNT-ME MWNT/PANI-ME PANI-ME

Fig. 3 – Cyclic voltammograms recorded at (a) PANI-g-MWNT-ME, (b) MWNT/PANI-ME and (c) PANI-ME in the solution containing 5 mM K3 Fe(CN)6 and 0.1 M KCl; scan rate: 20 mV s−1 .

Rs ( cm−2 ) 3.48 6.72 1.21

Rct (k cm−2 ) 32.4 73.4 121.7

Cdl (␮F cm−2 ) 895 326 128

The interfaces of three electrodes show typical impedance characteristics with well-defined semicircles at the high frequency corresponding to the charge transfer process followed by a straight line in the low-frequency region. Although the impedance spectra have similar shapes, the diameters of the semicircles differ for each of the electrodes. Interestingly, the impedance spectrum of PANI-g-MWNT-ME shows a semicircle with smaller diameter as compared to MWNT/PANI-ME and PANI-ME. In the absence of any redox species in the electrolyte medium, the Rct is ascribed to the doping/de-doping redox reactions of PANI in the electrode. The impedance parameters were calculated based on the Randle’s equivalent circuit model and presented in Table 1. PANI-g-MWNT-ME has a lower Rct value as compared to other electrodes. The higher value of Cdl (double layer capacitance) for PANI-g-MWNT-ME provides additional evidence for the increase in the electrode surface

6

a n a l y t i c a c h i m i c a a c t a 6 2 6 ( 2 0 0 8 ) 1–9

peak current shows a linear dependence on the scan rate (log ip = 0.9196 log  − 0.8334 with R2 = 0.9942 (n = 4)). The slope of peak current vs. scan rate was 0.92, a value nearer to the theoretically predicted value (unity) for the surface confined adsorption reaction at the electrode [28].

3.3.2.

Influence of supporting electrolyte

Several supporting electrolytes, such as BRb, potassium sulphate, potassium chloride, potassium nitrate were tested at various pHs for the efficiency of preconcentration and detection of CEL at PANI-g-MWNT-ME. The best electrochemical response measured in terms of the highest analytical signal and improved reproducibility was obtained in BRb (Fig. 5(ii)). Typically, a current signal of 7.9 ␮A was noticed at PANI-gMWNT-ME for CEL (1 ␮M) whereas, the current values with potassium sulphate, potassium chloride, potassium nitrate were comparatively lower (0.96, 0.58 and 0.05 ␮A, respectively) than observed in BRb (pH 7). Hence, BRb was selected as supporting electrolyte to perform further electrochemical measurements. Fig. 4 – Cyclic voltammograms recorded at PANI-g-MWNT-ME for 1 ␮M CEL in BRb (pH 7) at scan rate of 100 mV s−1 (a) without accumulation; (b and c) after 30 and 90 s at Eacc = −0.5 V.

area which probably due to the presence of PANI as grafted chains to MWNTs. Thus, it is demonstrated that the PANI-g-MWNT-ME has a larger surface area over the other electrode, MWNT/PANI-ME (bilayer). PANI-g-MWNT-ME is therefore expected to exhibit enhanced electrochemical properties that are beneficial for the electrochemical detection of CEL.

3.3.

Electrochemical behavior of CEL

Fig. 4 shows the CVs recorded at PANI-g-MWNT-ME for 1.0 ␮M CEL in BRb (pH 7) at a scan rate of 100 mV s−1 . A peak was observed at −1.08 V corresponding to adsorptive reduction of CEL (line a). The reduction wave was not accompanied by any anodic wave, which indicates that the reduction process at PANI-g-MWNT-ME is totally irreversible. Further, CVs were recorded after the accumulation of CEL (E: −0.5 V) at the electrode for different time intervals (30 and 90 s). As the accumulation time increases (Eacc = −0.5 V), a sharp with increase in peak current was observed (lines b and c). The increase in peak current indicates that the CEL gets adsorbed onto the surface PANI-g-MWNT-ME (at E = −0.5 V) and preconcentrate at the surface of the electrode. The preconcentration of CEL at the electrode effectively augments the reduction reaction. MWNTs provide high surface area for the adsorption of CEL and PANI contributes electroactive sites for the reduction process. The voltammetric results clearly show that PANIg-MWNT-ME can effectively be used for the electrochemical detection of CEL.

3.3.1.

Influence of scan rate

The influence of scan rate on the peak current (ip ) corresponding to adsorptive reduction of CEL was studied in the range between 10 and 200 mV s−1 (Fig. 5(i)). The

3.3.3.

Influence of pH

The influence of pH on the electrochemical adsorptive detection of CEL was also examined by cyclic voltammetry in BRb of different pH values. The solution of CEL was preconcentrated (Eacc = −0.5 V) at the PANI-g-MWNT-ME separately for different accumulation times (30 and 90 s). The CVs were recorded after the preconcentration step. The plot of peak current vs. pH informs that a maximum peak current was attained at pH 7 (BRb) in both cases (Fig. 5(iii)).

3.4.

Constant potential chronoamperometry

We compared the amperometric responses of CEL at PANIg-MWNT-ME, MWNT/PANI-ME and PANI-ME in BRb to the successive injections of 0.1 ␮M CEL at an applied potential of −1.20 V, PANI-g-MWNT-ME and MWNT/PANI-ME show current response to the addition of CEL (Fig. 6a and b). However, PANI-ME did not electrochemically respond to CEL (figure not shown). Nevertheless, PANI-g-MWNT-ME shows higher current response to CEL (ip (␮A) = 1.0023 + 0.002C (␮M)) than the MWNT/PANI-ME (ip (␮A) = 0.1404 + 0.0004C (␮M)). The higher current response observed at PANI-g-MWNT-ME for the adsorptive reduction of CEL is due to large surface area provided by grafting of PANI chains onto MWNTs. In addition to this, the combined presence of MWNTs and PANI as a single unit accounts for the enhanced electroactivity of PANI-g-MWNT-ME. Hence, the attention was finally focused on PANI-g-MWNT-ME electrode for the sensitive detection of CEL.

3.5. Electrochemical behavior of CEL at PANI-g-MWNT-ME by DPSV and SWSV Fig. 7 shows the differential pulse stripping voltammograms (DPSV) for 1 × 10−11 M CEL recorded at a scan rate of 10 mV s−1 . The DPSV shows a reduction peak at −1.01 V (Ip = 3 nA) after an accumulation time of 30 s (Eacc = −0.5 V). As the accumulation time increases to 90 s, the peak current increases sharply. This again confirms the attainment of adsorptive preconcen-

a n a l y t i c a c h i m i c a a c t a 6 2 6 ( 2 0 0 8 ) 1–9

7

Fig. 5 – (i) Influence of scan rate on the peak current of 1 ␮M CEL in BRb (pH 7); tacc = 30 s. (ii) Influence of supporting electrolytes on the peak current of 1 ␮M CEL; (a) BRb, (b) potassium sulphate, (c) potassium chloride and (d) potassium nitrate of pH 7, tacc = 30 s; scan rate: 100 mV s−1 . (iii) Influence of pH on the peak current of 1 ␮M CEL after the pre-concentration at −0.5 V for 30 (a) and 90 s (b) in BRb; scan rate: 100 mV s−1 .

tration at the PANI-g-MWNT-ME. On the contrary, square wave stripping voltammogram (SWSV) shows a reduction peak at −1.08 V with a much higher peak current (Ip = 31 nA) as compared to that noticed in DPSV. The signal intensity of SWSV was found to be 10 times higher that that of DPSV. Similar observation was found with SWSV for the trace analysis of few other drugs [29–32].

Fig. 6 – Amperometric response of (a) PANI-g-MWNT-ME and (b) MWNT/PANI-ME at −1.20 V upon successive addition of 0.1 ␮M CEL in BRb (pH 7).

3.6. Optimization of experimental parameters in SWSV The peak current obtained in SWSV is dependent on various instrumental parameters, such as SW amplitude, SW frequency, step height and quit time. These parameters are interrelated and have a combined effect on the signal (current response) [33]. SWSV was performed on identifying the influence of these parameters on the electrochemical signal corresponding to the drug, CEL. It was found that these parameters did not influence the peak potential of the reduction of CEL, however, they have strong influence on the peak current values (Table 2). An increase in peak current was witnessed while increasing the amplitude in the range 10–25 mV. There was an increase in the peak width, when the amplitude was greater than 25 mV. Hence, 25 mV was chosen as the optimum SW amplitude. The step height together with the frequency determines the effective scan rate. Hence, an increase in either the frequency or the step height results in an increase in the effective scan rate. The current signal of CEL increased with a SW frequency up to 100 Hz. The current signal was unstable and obscured with a large residual current beyond 100 Hz. The SW frequency was fixed as 100 Hz and the effect of step height on the current response of CEL was studied. At the step height of 6 mV, the detection of current response was sharp and reproducible. Hence, the step height was fixed as 6 mV. In all, the optimum SWSV parameters for effective response to CEL at PANI-g-MWNT-ME were found to be, SW

8

a n a l y t i c a c h i m i c a a c t a 6 2 6 ( 2 0 0 8 ) 1–9

Fig. 7 – Square wave stripping voltammograms for 1 × 10−11 M CEL with frequency (f) = 100 Hz, scan increment (dE) = 6 mV and pulse amplitude (E) = 25 mV. Inset shows the differential pulse stripping waveform recorded at scan rate = 10 mV s−1 ; pulse amplitude = 25 mV; accumulation time: 30 s (a) and 90 s (b).

frequency = 100 Hz; SW amplitude = 25 mV, step height = 6 mV. The influence of SWSV parameters on the current signal of CEL is summarized in Table 2. The effect of accumulation time on the peak current for different concentrations of CEL (1 × 10−11 to 1 × 10−6 M) at PANI-g-MWNT-ME was studied. Fig. 8(i) presents the representative SWSVs recorded at different accumulation time for CEL (1 × 10−11 M) in BRb (a plot of tacc vs. ip for 1 × 10−11 M CEL is also shown). Also, calibration plot(s) was made to study the effect of concentration on the peak current values of CEL (Fig. 8(ii)). At the lower concentrations between 1 × 10−11 and 1 × 10−8 M,

Table 2 – Influence of parameters on square wave stripping voltammograms for 1 × 10−11 M CEL at PANI-g-MWNT-ME in BRb (pH 7) Constant parameters

Varied parameters

f = 100 Hz dE = 6 mV QT = 20 s

Pulse amplitude, E (mV)

E = 25 mV dE = 6 mV QT = 20 s

Frequency, f (Hz)

E = 25 mV f = 100 Hz QT = 20 s

Scan increment, dE (mV)

E = 25 mV f = 100 Hz dE = 6 mV

Quit time, QT (s)

a

ip (nA) 10 15 25

17.8 21.1 29.6a

60 80 100

23.7 21.5 28.5a

4 6 8

21.6 29.8a 31.6

5 10 20

19.4 21.8 30.3a

Optimized conditions selected for the detection of CEL (1 × 10−11 M) through SWSV at PANI-g-MWNT-ME.

Fig. 8 – (i) Effect of accumulation time on the SWSV for 1 × 10−11 M CEL with frequency (f) = 100 Hz, scan increment (dE) = 6 mV and amplitude (E) = 25 mV (tacc = (a → k) 5, 10, 20, 30, 40, 50, 60, 75, 90, 100 and 120, respectively). (ii) Calibration plot(s).

the peak current showed a linear increase with accumulation time, however a non-linear increase in current was noticed when the concentration was increased further. On the contrary, the current response was found to be linear within the concentration levels from 1 × 10−11 to 1 × 10−6 M (Fig. 8(ii)) and the peak current was found to be gradually decreases and saturates at the higher concentrations of CEL (>␮M). This is due to the saturation of electrode area with adsorbent molecules at higher concentration. Important aspects such as repeatability of the measurement and reproducibility of the responses were evaluated

a n a l y t i c a c h i m i c a a c t a 6 2 6 ( 2 0 0 8 ) 1–9

at the PANI-g-MWNT-ME. The repeatability of the measurements at PANI-g-MWNT-ME was tested with five calibration plots by using the same PANI-g-MWNT-ME. A R.S.D. value of about 1.38% was obtained from the slope of the plot, which indicates the good repeatability of the measurements at the electrode. The reproducibility of the responses was measured with five different electrodes of same configuration. Reproducible current responses were noticed for each electrode. Results from five different electrodes yielded R.S.D. value of only 7.28%, which indicates the methodology adapted for the fabrication of PANI-g-MWNT-ME was reliable and reproducible. Modification of ITO with a PANI-g-MWNTs layer leads to a strong interfacial accumulation of CEL and as a result, highly sensitive adsorptive stripping measurements could be made. Thus, by controlling the accumulation time, we could effectively utilize the PANI-g-MWNT-ME for the sensitive detection of CEL. It is important to note that, the detection level of CEL (1 × 10−11 M) at PANI-g-MWNT-ME is much higher than reported using hanging mercury drop electrode [4] and also by other techniques such as spectrophotometry, chromatography and fluorimetric methods [5–12].

4.

Conclusions

In the present investigation, a modified electrode based on grafting of PANI chains onto MWNTs, was fabricated. PANIg-MWNT-ME has adequate electrochemical characteristics in terms of high surface area, higher electronic conductivity and large capacitance and exhibits a facile electrochemical detection of CEL. A validated, precise and specific voltammetric procedure for the trace determination of CEL is developed. PANI-g-MWNT-ME significantly exhibits a sensitive detection of CEL in extremely lower concentrations (1 × 10−11 M). The enhanced electrochemical activity of PANIg-MWNT-ME towards the detection of drug molecule arises from the synergistic contribution from PANI and MWNTs. The PANI-g-MWNT-ME suits well for the trace analysis of drugs in clinical and quality control laboratories. Thus, the use of conducting polymer grafted carbon nanotubes based electrode provides ample scope for a wide range of trace analysis.

Acknowledgments This work was supported by Korean Research Foundation Grant (KRF-2006-J02402 and KRF-2006-C00001). The authors acknowledge the Korea Basic Science Institute (Daegu) and Kyungpook National University Center for Scientific Instrument. We thank the anonymous reviewers for many insightful comments and suggestions for useful experiments.

9

references

[1] S.H. Roth, Ann. Intern. Med. 109 (1988) 353. [2] W. Xie, J.G. Chipman, D.L. Robertson, R.L. Erikson, D.L. Simmons, Proc. Natl. Acad. Sci. U.S.A. 88 (1991) 2692. [3] A. Weber, A. Casini, A. Heine, D. Kuhn, C.T. Supuran, A. Scozzafava, G. Klebe, J. Med. Chem. 47 (2004) 550. [4] M.M. Ghoneim, A.M. Beltagi, Talanta 60 (2003) 911. [5] R.N. Saha, C. Sajeev, P.R. Jadhav, S.P. Patil, N. Srinivasan, J. Pharm. Biomed. Anal. 28 (2002) 741. [6] L.I. Bebawy, A.A. Mostafa, N.F. Talib, J. Pharm. Biomed. Anal. 27 (2002) 779. [7] M.K. Srinivasu, C.L. Narayana, D.S. Rao, G.O. Reddy, J. Pharm. Biomed. Anal 22 (2000) 949. [8] H.S. Chow, N. Anavy, D. Salazar, D.H. Frank, D.S. Alberts, J. Pharm. Biomed. Anal. 34 (2004) 167. [9] M.K. Srinivasu, D.S. Rao, G.O. Reddy, J. Pharm. Biomed. Anal. 28 (2002) 493. [10] M.A. Hamid, L. Novotny, H. Hamaz, J. Chromatogr. B 753 (2001) 401. [11] H. Jalalzadeh, M. Amini, V. Ziaee, A. Safa, H. Farsam, A. Shafiee, J. Pharm. Biomed. Anal. 35 (2004) 665. [12] F. Schonberger, G. Heinkele, T.E. Murdter, S. Brenner, U. Klotz, U. Hofmann, J. Chromatogr. B 768 (2002) 255. [13] K.-P. Lee, A. Gopalan, P. Santhosh, K.M. Manesh, J.H. Kim, K.S. Kim, J. Nanosci. Nanotechnol. 6 (2006) 1575. [14] P. Santhosh, K.M. Manesh, A. Gopalan, K.P. Lee, Anal. Chim. Acta 575 (2006) 32. [15] K.M. Manesh, J.H. Kim, P. Santhosh, A. Gopalan, K.-P. Lee, H.-D. Kang, J. Nanosci. Nanotechnol. 7 (2007) 3365. [16] K.M. Manesh, H.T. Kim, P. Santhosh, A.I. Gopalan, K.-P. Lee, Biosens. Bioelectron. 23 (2008) 771. [17] P. Santhosh, K.M. Manesh, A. Gopalan, K.-P. Lee, Sens. Actuators B 125 (2007) 92. [18] K.M. Manesh, P. Santhosh, A.I. Gopalan, K.-P. Lee, Electroanalysis 18 (2006) 1564. [19] P.M. Ajayan, O. Stephan, C. Colliex, D. Trauth, Science 265 (1994) 1212. [20] J. Wang, M. Musameh, Anal. Chim. Acta 539 (2005) 209. [21] K.P. Loh, S.L. Zhao, W.D. Zhang, Diam. Relat. Mater. 13 (2004) 1075. [22] M. Guo, J. Chen, J. Li, B. Tao, S. Yao, Anal. Chim. Acta 532 (2005) 71. [23] P. Santhosh, K.M. Manesh, K.-P. Lee, A. Gopalan, Electroanalysis 18 (2006) 894. [24] P. Santhosh, A. Gopalan, K.-P. Lee, J. Catal. 238 (2006) 177. [25] C.Y. Chung, T.C. Wen, A. Gopalan, Mater. Chem. Phys. 71 (2001) 148. [26] W.C. Chen, T.C. Wen, C.C. Hu, A. Gopalan, Electrochim. Acta 47 (2002) 1305. [27] J. Ni, H. Ju, H. Chen, D. Leech, Analyst 123 (1998) 2895. [28] E. Laviron, J. Electroanal. Chem. 112 (1980) 1. [29] P. Santhosh, N. Senthilkumar, M. Renukadevi, A.I. Gopalan, T. Vasudevan, K.-P. Lee, Anal. Sci. 23 (2007) 475. [30] J.C. Vire, N. Abo El-Maali, G.I. Patriarche, G.D. Christian, Talanta 35 (1988) 997. [31] A.M.S. Roque da Silva, J. Lima, M.T. Oliva Teles, A.M.O. Brett, Talanta 49 (1999) 611. [32] A. Radi, A.M. Beltagi, M.M. Ghoneim, Talanta 54 (2001) 283. [33] N.P. Saravanan, S. Venugopalan, N. Senthilkumar, P. Santhosh, B. Kavita, H.G. Prabu, Talanta 69 (2006) 656.