Amd

BD

a

ARRAA

KWEMMA

1

tmycoth

ww[lc

h0

Sensors and Actuators B 196 (2014) 539–545

Contents lists available at ScienceDirect

Sensors and Actuators B: Chemical

jo ur nal home page: www.elsev ier .com/ locate /snb

n electrochemical sensor based on multiwall carbon nanotubes andolecular imprinting strategy for warfarin recognition and

etermination

ehzad Rezaei ∗, Omid Rahmanian, Ali Asghar Ensafiepartment of Chemistry, Isfahan University of Technology, Isfahan 84156-83111, Islamic Republic of Iran

r t i c l e i n f o

rticle history:eceived 28 November 2013eceived in revised form 28 January 2014ccepted 10 February 2014vailable online 19 February 2014

eywords:arfarin determination

lectro-sensorolecularly imprinted polymerultiwall carbon nanotubes

a b s t r a c t

A sensitive electrochemical sensor for warfarin was prepared based on molecular imprinting strategy byelectropolymerization qof o-phenylenediamine (o-PD) on a glassy carbon electrode via cyclic voltam-metry (CV). In order to enhance the electrode sensitivity and electronic transmission, multiwall carbonnanotubes (MWCNT) containing carboxylic functional group (f-MWCNTs) were introduced on glassycarbon electrode (GCE). Thin film of molecularly imprinted polymer (MIP) with specific binding sites forwarfarin was cast on the modified electrode using electrochemical deposition. In order to form a dou-ble layer with MIP layer as an insulating electrolyte, Au nanoparticles (AuNPs) was introduced at theMIP surface to form final modified electrode (AuNP/MIP/f-MWCNT/GCE). The properties of AuNP/MIP/f-MWCNT/GCE were studied in the presence of K3Fe(CN)6 as a probe for signal transduction and also bythe use of electrochemical impedance spectroscopy (EIS). AuNP/MIP/f-MWCNT/GCE exhibits fast binding

u-nanoparticles kinetics and good selectivity to template due to their high ratio of surface imprinted sites, large surface-to-volume ratios and large affinity to template. The modified electrode was used to detect the concentrationof warfarin with a linear range and detection limit (S/N = 3) of 0.031–0.616 ng mL−1 and 0.024 ng mL−1,respectively. Finally, the modified electrode was successfully applied to determine warfarin in humanserum sample.

© 2014 Elsevier B.V. All rights reserved.

. Introduction

Warfarin is widely used as an oral anticoagulant in the preven-ion and treatment of venous and arterial thromboembolism. It is

ost likely to be the drug popularly referred to as a “blood thinner,”et this is a misnomer, since it does not affect the thickness or vis-osity of blood. Instead, it acts on the liver to decrease the quantityf a few key proteins in blood that allow blood to clot [1]. Therapeu-ic concentration of warfarin is about 2.0–5.0 �g mL−1. Warfarinas a long half-life. The half-life is 20–60 h (mean 40 h) [2].

Several methods have already been described to determinearfarin such as high performance liquid chromatography (HPLC)ith fluorescence detection [3,4], LC and HPLC with UV detection

5–8], HPLC-MS/MS [9–11], multi-mode UPLC (ultra performanceiquid chromatography)-MS/MS method [12,13], LC-MS/MS [10],hiral capillary electrophoresis with spectrophotometric detection

∗ Corresponding author. Tel.: +98 3113912351; fax: +98 3113912350.E-mail addresses: rezaei@cc.iut.ac.ir, rezaeimeister@gmail.com (B. Rezaei).

ttp://dx.doi.org/10.1016/j.snb.2014.02.037925-4005/© 2014 Elsevier B.V. All rights reserved.

[14], chiral stationary-phase liquid chromatography-fluorescencesystem coupled with on-line circular dichroism detector [15] andtime-correlated single-photon counting [16]. However, there maybe several disadvantages for these methods. For example some ofthem are time consuming and tedious, while others use large bio-logical fluid volumes and need expensive instruments and toxicsolvents, plus time consuming procedures.

As a capable element for specific recognition, molecularlyimprinted polymers (MIPs) have paying attention in recent years[17]. Molecular imprinting technique is an approach to synthesiz-ing a polymer matrix with molecular recognition sites, which arespecific in shape and size to the target molecular, exhibiting specifichigh binding behaviors to the target molecules. Due to low chemicaland physical stability of the antibodies or enzymes avoids their usein harsh circumstances and at high temperatures. MIPs have somespecific advantages such as low cost of preparation, high affinity to

the template, robustness and long-term stability [18,19]. MIPs havebeen used for molecular recognition variety from tiny molecules[20–24] to large molecules [25–28]. To improve some limitationof MIPs, such as low density of imprinted sites, slow binding time,

5 Actua

hps[an[tlpsda[mr

MbbctdtotEuist

2

2

((mut

P(wm((pp

2

aNcMeutm1a

40 B. Rezaei et al. / Sensors and

eterogeneous distribution of imprinted sites and insultingolymer in electrochemically application of MIPs, several studiesuggested the formation of thin MIP layer on nanomaterials surface29–32]. Own to their unique structure, high chemical stabilitynd high surface-to-volume ratio, carbon nanotubes, as new classanomaterials, are extremely attractive in electrochemical sensors33]. In order to enhance the electrode sensitivity and electronicransmission, MWCNTs and AuNPs were introduced in both of insu-ating MIP layer. Electrochemical synthesis of MIP is one of the mostromising techniques with which to develop electrochemical sen-ors [34]. On the other hand, the incorporation of electrochemicalevices and MIPs, which exhibit good sensitivity and selectivity, isn attractive approach for the development of biochemical sensors34–43]. Some of the previous works using electropolymerization

ethod for the fabrication of imprinted sensing system wereeported [20,34–44].

Here in we attempt to investigate the feasibility of fabricating aIP sensor. The sensor, AuNP/MIP/f-MWCNT/GCE, was fabricated

y electrochemical deposition of o-PD at the surface of glassy car-on electrode (GCE), surface modified with MWCNTs containingarboxylic functional group (f-MWCNTs) and followed by deposi-ion of AuNPs using a potentiostatic method. The electrochemicaleposition method has many benefits such as simple prepara-ion, easy control of the film thickness, and good reproducibilityf identical polymer films [45]. f-MWCNTs and AuNPs were usedo enhance the electrode sensitivity and electronic transmission.lectrochemical impedance spectroscopy, cyclic voltammetry weresed to study of the sensor. Both of the nanoparticles were

ntroduced in both sides of the MIP layer to form a double layertate. K3Fe(CN)6 was used an electrochemical probe for signalransduction.

. Experimental

.1. Chemicals

Warfarin, hydrochlorothiazide (HCT), and phenazopyridinePAP) were supplied from Sigma–Aldrich. DimethylformamideDMF), o-PD, phosphoric acid, sodium acetate, acetic acid, and

ethanol were purchased from Merck. All chemicals and reagentssed were of analytical grade and doubly distilled water was usedhroughout.

MWCNTs were brought from Iran’s Research Institute ofetroleum Industry (synthesized by chemical vapor depositionCVD) with the purity of 95% and the average thickness ofall and length of the MWCNT were about 40 nm and severalicrometers, respectively). Trisodium citrate and gold chloride

HAuCl4) were supplied from Aldrich. Phosphate buffer solutionPBS, 0.03 mol L−1) with different pH values was used to study ofH effect. pH values were measured using a Metrohm (Model 827H-lab) pH meter.

.2. Electrochemical measurements

Electrochemical measurements were performed with potentiostat–galvanostat �AutoLab (Echo Chemie, B.V.,etherlands, NOVA software), and a conventional three-electrodeonfiguration, containing 50 mL glass cell, AuNP/MIP/f-WCNT/GCE or AuNP/NIP/f-MWCNT/GCE as the working

lectrode, and platinum wire as the auxiliary electrode wassed. Potentials were measured versus saturated calomel elec-

rode (SCE). Electrochemical impedance spectroscopic (EIS)

easurement was performed with the Autolab system (PGSTAT2, Eco Chemie B.V., Utrecht, Netherlands). The system was run on

PC using GPES and FRA 4.9 software. In each measurement, the

tors B 196 (2014) 539–545

modified electrode was immersed into the solution of warfarin inPBS (pH 6.5) for 90 s and then the electrochemical measurementswere done in a solution containing 5.0 mmol L−1 K3Fe(CN)6 in0.1 mol L−1 KCl as a probe, and completed to the mark with PBSat pH 6.8 [20]. The cyclic voltammetric curve obtained under apotential range of −0.10 V to 0.50 V and at a scan rate of 50 mV s−1.After each analysis, the MIPCNT or NIPCNT electrode was recov-ered with methanol to remove the adsorbed molecules. Then, therecovered electrode was employed to the subsequent rebinding.For impedance measurements, a frequency range of 100 kHz to5.0 MHz was employed. The AC voltage amplitude and the poten-tial were 5.0 mV and 0.20 V, respectively. All of electrochemicalexperiments were carried out at room temperature.

2.3. Preparation of AuNPs

AuNPs were prepared by following reported procedure [46].Briefly, 4.0 mL 1.0% (w/v) sodium citrate was added rapidly to thesolution obtained by dissolving 1.0 mL 1.0% (w/v) HAuCl4 and 99 mLwater and then, keeping the mixture in an oven for 90 min at 60 ◦C.The color solution changed from pale yellow to blue and finally tored–violet. The prepared AuNPs were stored in a dark bottle at 4 ◦Cuntil further use.

2.4. Preparation of the modified-imprinted sensor

MWCNTs containing carboxylic functional group (f-MWCNTs)was prepared as follow; 500 mg of crude MWCNTs were added intoa glass reactor containing 60 mL of HNO3. The mixture was keptunder ultrasonicating for 15 min. This mixture was refluxed for 22 hat 80 ◦C and then cooled, filtered and washed by passing throughabout 5 L of distilled water. The prepared f-MWCNTs were used as asupporting material for the formation of the polymer on the surfaceof GCE.

Prior to use, the surface of GCE was polished carefully to amirror-finished with slurry alumina (0.05 �m) at polishing clothfor 2 min. Then, the electrode was rinsed with distilled water andcleaned ultrasonically into a mixture of water:methanol (1:1) for3 min to remove particles that might be trapped at the surface. A5.0 mg of f-MWCNTs were added into 2.5 mL DMF and f-MWCNTswere dispersed by ultrasonic treatment for 6 min. Then, 3.0 �Lof the dispersed mixture was poured on the GCE surface andallowed to evaporate solvent and formed f-MWCNTs on the elec-trode surface.o-PD functionalized on the electrode surface modifiedwith f-MWCNTs were prepared firstly by immersing the f-MWCNTselectrode into a acetic acid/acetate buffer solution (ABS, pH 5.2)containing 5.0 mmol L−1 o-PD and 1.7 mmol L−1 warfarin, for 13 h.The MIP/f-MWCNT/GCE sensor was fabricated by electrochemicaldeposition via cyclic voltammetry in the potential range of −0.40 Vto 1.00 V at a scan rate 50 mV s−1 for 20 cycles. Then, potentiostaticdeposition method was applied to prepare the Au-nanoparticlesmodified onto MIP/f-MWCNT/GCE (AuNP/MIP/f-MWCNT/GCE) byimmersing of MIP/f-MWCNT/GCE into 1.5 × 10−4 mol L−1 HAuCl4and 0.1 mol L−1 KNO3 solution, using a potential of −0.20 V forthe duration of 180 s [47]. The template was extracted from theimprinted film by immersing the obtained electrode under gentlestir in methanol for 90 min. The completely removal of warfarinwas checked electrochemically, so that after 90 min as an incuba-tion time, the redox peak of the probe was reached to maximumvalue and stayed constant. Therefore, 90 min was chosen as an incu-

bation time for complete removing the template by methanol. Thenon-imprinted polymer (NIP) sensor was also prepared by using thesame method without addition of warfarin. The entire schematicrepresentation is illustrated in Fig. 1.

B. Rezaei et al. / Sensors and Actuators B 196 (2014) 539–545 541

FM

2

og14fatwia

3

3

ip2tFcteo

F5bIc

Fig. 3. Cyclic voltammograms of (a) bare GCE at the present of warfarin andcyclic voltammograms of 5.0 mmol L−1 K3Fe(CN)6 and 0.1 mol L−1 KCl at: (b) bareGCE; (c) f-MWCNT/GCE; (d) MIP/f-MWCNT/GCE; (e) AuNP/MIP/f-MWCNT/GCE; (f)AuNP/MIP/f-MWCNT/GCE after removing warfarin from polymeric network; and (g)

−1

ig. 1. The simplified sketch for the fabrication process of AuNP/MIP/f-WCNT/GCE.

.5. Sample preparation

A 0.5 mL human serum sample (taken from the health centerf Isfahan University of Technology) was spiked with warfarin toive a working concentration of 30.8 ng mL−1, 61.6 ng mL−1 and54.0 ng mL−1. This sample was placed into a glass vial including.0 mL of methanol, vortexed for 10 s, and centrifuged at 4000 rpmor 20 min. A 1.0 mL aliquot of the methanol layer was placed intonother glass vial and 3.0 mL of methanol was added and cen-rifuged again at 4000 rpm for 10 min. A 0.50 mL of the supernatantas diluted to 10.0 mL with PBS solution (pH 6.5) and after 90 s

ncubation time (open circuit) for adsorption of warfarin, CV waspplied by K3Fe(CN)6 probe to measure reduced current.

. Results and discussion

.1. Preparation and characterization of the sensor

Cyclic voltammetry was employed to electrodepositingmprinted polymer on the surface of f-MWCNTs electrode in theotential range from −0.40 V to 1.00 V, scan rate of 50 mV s−1 for0 cycles. A typical cyclic voltammogram recorded during the elec-ropolymerization in the presence of o-PD and warfarin is shown inig. 2. The peak current decreased significantly under consecutive

yclic scans, indicating that the insulating polymer was formed tohe f-MWCNTs electrode surface. In order to compare f-MWCNTsffect, electropolymerization was also performed in the absencef f-MWCNTs (Fig. 2, inset) at the surface of GCE. As can be seen

ig. 2. Cyclic voltammograms for the electrochemical polymerization of.0 mmol L−1 o-PD containing 1.7 mmol L−1 warfarin at f-MWCNT/GCE in acetateuffer solution (pH 5.2). Conditions: Scan rate, 50 mV s−1; number of cycles, 20.

nset is the electrochemical polymerization at a bare GCE as the same mentionedonditions.

the same electrode as (f) after 90 s incubating in 0.310 �g mL warfarin solution.Conditions: Potential scan range, −0.10 V to +0.50 V; scan rate, 50 mV s−1 in PBS (pH6.8).

in this figure, the peak currents were increased remarkably atpresent of f-MWCNTs vs. in the absence of the carbon nanotubes.Therefore, for the preparation of the sensor, we used f-MWCNTsbefore treatment of the electrode for polymerization.

To characterize the different prepared electrodes, the elec-trochemical behavior of the stepwise production process wasinvestigated at pH 6.8 (PBS) containing 5.0 mmol L−1 K3Fe(CN)6,performed as an electrochemical probe. Fig. 3 shows the CV of elec-trochemical probe at different cases. CV of bare GCE at the presentof warfarin was recorded to demonstrate that warfarin is not elec-trochemically active at the ferocyanide oxidation potential. As canbe seen (Fig. 3a) nothing peaks was never observed in this appliedpotential. At the first, f-MWCNTs was successfully introduced atthe GCE surface to increase surface area accessible for formation ofMIP layer, as described before. The CV of bare GCE was obtained(Fig. 3b) and after solvent evaporation, the CV of modified GCEwith f-MWCNTs (f-MWCNT/GCE) was recorded as same condition(Fig. 3c). It is obvious that the current response at f-MWCNT/GCEbecame larger which would attribute to the enhancement of theconduction of the electrode by deposited f-MWCNTs. When theelectrode was coated with polymeric film at the f-MWCNTs sur-face, the pair of the redox peaks of the probe disappeared in Fig. 3d.This is due to the fact that the polymeric film was not conductiveand there were almost no channels for the active probe to accessthe electrode surface. With introducing AuNPs on the polymericfilm (AuNP/MIP/f-MWCNT/GCE) the peak current value increasedagain (Fig. 3e). Once the template extracted from the MIP sen-sor, the redox peaks of the probe was occurred again in Fig. 3f.It may be due to the cavity of the removed warfarin molecules,forming some channels for the probe to pass through the cavitydue to its little size. In order to show the capability of the modi-fied electrode to bind with MIP sites, the modified electrode beingincubated in 0.310 �g mL−1 warfarin in PBS (pH 6.5) and CV wasobtained (Fig. 3g). As can be seen, the oxidation peak of the probewas reduced in comparison with Fig. 3f. This result can be attributedto bind of the template molecules to the sites of the MIP, and there-fore, the channels to probe arrival are blocked by the templatemolecules.

3.2. EIS measurements

EIS is a helpful method for probing the features of surfacemodified electrodes. The impedance spectra comprise a semicircle

542 B. Rezaei et al. / Sensors and Actuators B 196 (2014) 539–545

Fig. 4. EIS of 5.0 mmol L−1 K3Fe(CN)6 and 0.1 mol L−1 KCl at: (a) f-MWCNT/GCE;(b) AuNP/MIP/f-MWCNT/GCE after extraction of warfarin from polymeric network;and (c) the same electrode (b) after 90 s incubating in 0.310 �g mL−1 warfarin solu-tion. Inset: Effect of AuNPs deposition on MIP/f-MWCNT/GCE on Rct. Conditions:p5

pltsatoecsaAbstif(tTis

3

pow3astcnf

3

eep

optimum incubation time of 90 s was chosen for the determinationof warfarin.

otential, 0.20 V; frequency range, 5.0 mHz to 100 kHz; and AC voltage amplitude,.0 mV.

art at high frequencies, corresponding to the electron transferimited process and a linear part at lower frequencies, indicatinghe diffusion-limiting step of the electrochemical process. Theemicircle diameter equals to the electron transfer resistance (Rct)t the electrode surface. Fig. 4a shows EIS signal of f-MWCNT/GCEhat almost straight line was revealed, which was characteristicf a diffusion-limiting step of the electrochemical process. Asxpected, by formation of the polymeric film under the electro-hemical deposition onto f-MWCNT/GCE electrode surface, Rct wasignificantly expanded which indicated the polymeric film actss a definite kinetic barrier for the electron transfer (Fig. 4, inset).fter potentiostatic deposition of AuNPs, Rct was reduced that maye due to nanometer-sized gold particles participate a key roleimilarly to a conducting wire or electron-conducting tunnel andherefore promoting electron-transfer reactions of probe (Fig. 4,nset). By removing template from the polymeric network, Rct wasurther reduced in comparison with AuNP/MIP/f-MWCNT/GCEFig. 4b). Finally, after the template extraction, the modified elec-rode was incubated into 0.310 �g mL−1 warfarin (PBS, pH 6.5).he results showed that Rct increased since warfarin embeddedn binding sites and blocked some arrival channels to electrodeurface (Fig. 4c).

.3. Effect of monomer to template ratio

Commonly, the molar ratios of monomer to template in thereparation process influence the affinity and imprinting efficiencyf MIP. The performance of AuNP/MIP/f-MWCNT/GCE electrodeas examined by changing o-PD to warfarin molar ratio as 1:1,

:1 and 5:1. Three types of the imprinted sensors were constructedccording the proposed process, respectively. The results demon-trated that the prepared electrode with the ratio of 3:1 revealedhe largest and fastest current response. Therefore, it might be aapable applicant for efficient electrochemical sensor for the recog-ition of warfarin and it was selected for the imprinted sensor

abrication.

.4. Effect of pH

The pH of the solution in adsorption step has a significant influ-nce on the warfarin oxidation at the surface of the modifiedlectrode. To make the convenient configuration of warfarin, theH of the sample should be adjusted. For this purpose, a series of

Fig. 5. Effect of pH of working solution on warfarin adsorption in MIP cavities.

PBS solutions in the pH range of 5.0–8.0 were selected contain-ing 0.310 �g mL−1 warfarin, using 90 s as an incubation times. Asshown in Fig. 5, the current intensity of the oxidation peak reachesits maximum at a pH value of 6.5, and then decreased as the pHincreases further. The reason may be that the warfarin was pro-tonated in acidic environment and cannot be fitting into cavities.While, the stability of the imprinted film will be diminished inhigh alkalinity and the warfarin was formed in anionic structure.Therefore, pH 6.5 was chosen as the suitable pH for further studies.

3.5. Incubation time

The incubation time is typically a simple and useful approachto improve the sensitivity of the imprinted sensor. After templatemolecules were removed from the imprinted film, the modi-fied electrode was incubated in 0.310 �g mL−1 warfarin solutioncontaining PBS (pH 6.5) for different incubation times and the suc-cessive decrease of the peak current was draw vs. the incubationtime (Fig. 6). The results show that by increasing the incubationtime, the oxidation peak current of the probe reduced sharply upto 90 s as an incubation time, and then it’s leveled off gradually.The decreasing of the oxidation peak current indicates that theamount of the adsorbed warfarin in the cavities of the imprintedfilm was growing with increasing the incubation time. Thus, the

Fig. 6. Influence of incubation time on the response of the imprinted sensor.

B. Rezaei et al. / Sensors and Actuators B 196 (2014) 539–545 543

Table 1Results of warfarin determination in real sample analysis (n = 3).

Warfarin added (ng mL−1) Average of warfarin found (ng mL−1) Recovery (%) RSD (%)

Serum sample − Not detected − −30.8 27.8 90.3 2.261.6 65.6 106.5 2.3

154.0 150.8 98.0 3.1

3

tcntod

siiwtrtreta

3

ageliscom

Fto0i0−

.6. Figures of merit

Under the optimized experimental conditions, a dependence ofhe relative oxidization peak current of the probe vs. warfarin con-entration was examined with amperometric detection (Fig. 7). Theet currents were linearly increased with the analyte concentra-ion in the range of 0.031–0.616 ng mL−1 with regression equationf Ip (�A) = 2.9622 C (ng mL−1) + 0.6358 (R2 = 0.9913). The limit ofetection based on S/N = 3 was 0.024 ng mL−1 warfarin.

In order to evaluate the reproducibility of the imprinted sen-or, the change in the net response of the sensor before and afterncubation into 0.310 �g mL−1 warfarin solutions was measuredn four replicates. After each experiment, the elution performed

ith methanol to remove warfarin trapped in the cavities. Then,he next experiment was implemented at the same electrode. Theelative standard deviation (RSD) was estimated as 3.2%. This elec-rode was tested several times every day and for three weeks. Theesults showed (with recovery higher than 96%) that the modifiedlectrode could be used several times without major variety. Sohe proposed imprinting sensor was expected to be regeneratednd have good reproducibility.

.7. Selectivity

In order to assess the selectivity of the new sensor, HCT, PAPnd Aspirin (ASP) were selected since these drugs have functionalroups, which are able to bind with MIP cavities, but with differ-nt conformation. Warfarin may reason strict bleeding that can beife threatening and even cause death. The danger of bleeding isncreased when warfarin is accompanied with antiplatelet drugs

uch as ASP, or other nonsteroidal anti-inflammatory drugs andan interacts with many usually used drugs, and the metabolismf warfarin varies seriously between patients. Therefore, deter-ination of warfarin with good selectivity could be attractive. As

ig. 7. Calibration curve corresponding to the response of the MIP sensoro the concentration of warfarin. Inset is the square wave voltammogramsf AuNP/MIP/f-MWCNT/GCE sensor at present of 5.0 mmol L−1 K3Fe(CN)6 and.1 mol L−1 KCl after immersion in different warfarin concentration from outer to

nner: 0.031 ng mL−1, 0.062 ng mL−1, 0.123 ng mL−1, 0.185 ng mL−1, 0.246 ng mL−1,.339 ng mL−1, 0.493 ng mL−1 and 0.616 ng mL−1. Conditions: Potential scan range,0.10 V to +0.50 V; scan rate, 50 mV s−1 in PBS (pH 6.8) and 90 s incubation time.

Fig. 8. Evaluate selectivity of warfarin (0.310 �g mL−1) against HCT (0.297 �g mL−1),PAP (0.213 �g mL−1) and ASP (0.180 �g mL−1) (at equivalent of 1.0 �mol L−1 concen-trations for all of them) in PBS (pH 6.5) and compared with NIPCNT electrode.

shown in Fig. 8, the MIP sensor (AuNP/MIP/f-MWCNT/GCE) wastested in 0.297 �g mL−1 HCT, 0.213 �g mL−1 PAP, 0.180 �g mL−1

ASP, 0.310 �g mL−1 warfarin (at equivalent of 1.0 �mol L−1 con-centrations for all of them), and these results compared to theNIP electrode (AuNP/NIP/f-MWCNT/GCE). For NIP electrode, eitherwarfarin or interfering species, there were produced small currentresponses. This might explained that there were no suitable sites atNIP electrode for these examined drugs. The current response of theMIPs electrode for warfarin was significantly higher than that of theother interfering species, illuminating that the MIPs electrode hadgood selectivity to warfarin and the imprinting efficiency was 4.52.This can be explicating by the size and conformation of functionalgroup of cavities matching with warfarin in the MIPs film.

3.8. Real sample analysis

In order to investigate the capability of the proposed sensor fordetermination of warfarin in complex matrix of real clinical sam-ples, and since there are no patient samples available that theywould use warfarin drug, a spike method was chosen. Three mea-surements were performed for each concentration. As can be seenin Table 1, good recoveries and RSD were obtained revealing thatthe recommended method has capability in determination of war-farin serum samples. In order to use of the proposed method forthe reality, serum may be required to dilute several times.

4. Conclusions

In this work, a novel sensitive molecularly imprinted elec-trochemical sensor was prepared for determination of warfarinusing electrochemical deposition of MIP layer within MWCNTs andAuNPs as a double layer. The modified electrode was used to achievea selective warfarin sensor with high recoveries. Furthermore, theprepared sensor showed very high recognition ability toward war-

farin as low as 0.024 ng mL−1. The imprinted sensor shows fastbinding kinetics to the template due to their high ratio of surfaceimprinted sites, large surface-to-volume ratios (because of MWC-NTs and AuNPs introducing) and has shown large affinity to the

5 Actua

te

A

CBn

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

44 B. Rezaei et al. / Sensors and

emplate. Eventually, the capability of the proposed sensor wasvaluated by determination of warfarin in serum samples.

cknowledgement

We wish to express our gratitude to the Research Council, theenter of Excellence in Sensor and Green Chemistry, Institute ofiotechnology and Bioengineering of Isfahan University of Tech-ology (IUT) for financial support of this work.

eferences

[1] M. Ufer, Comparative pharmacokinetics of vitamin K antagonists: war-farin, phenprocoumon and acenocoumarol, Clin. Pharmacokinet. 44 (2005)1227–1246.

[2] A. Walfisch, G. Koren, The warfarin window in pregnancy: the importance ofhalf-life, J. Obstet. Gynaecol. Can. 10 (2010) 988–989.

[3] C. Banfield, M. Rowland, Stereospecific fluorescence high-performance liquidchromatographic analysis of warfarin and its metabolites in plasma and urine,J. Pharm. Sci. 73 (1984) 1392–1396.

[4] V.K. Boppana, W.H. Schaefer, M.J. Cyronak, High-performance liquid-chromatographic determination of warfarin enantiomers in plasma withautomated on-line sample enrichment, J. Biochem. Biophys. Methods 54 (2002)315–326.

[5] T. Uno, T. Niioka, M. Hayakari, K. Sugawara, T. Tateishi, Simultaneous deter-mination of warfarin enantiomers and its metabolite in human plasma bycolumn-switching high-performance liquid chromatography with chiral sepa-ration, Ther. Drug Monit. 29 (2007) 333.

[6] P.R. Ring, J.M. Bostick, Validation of a method for the determination of (R)-warfarin and (S)-warfarin in human plasma using LC with UV detection, J.Pharm. Biomed. Anal. 22 (2000) 573–581.

[7] A. Osman, K. Arbring, T.L. Lindah, A new high-performance liquid chromato-graphic method for determination of warfarin enantiomers, J. Chromatogr. B826 (2005) 75–80.

[8] I. Locatelli, V. Kmetec, A. Mrhar, I. Grabnar, Determination of warfarin enan-tiomers and hydroxylated metabolites in human blood plasma by liquidchromatography with achiral and chiral separation, J. Chromatogr. B 818 (2005)191–198.

[9] G.P. Miller, D.R. Jones, S.Z. Sullivan, A. Mazur, S.N. Owen, N.C. Mitchell,A. Radominska-Pandya, J.H. Moran, Assessing cytochrome P450 and UDP-glucuronosyltransferase contributions to warfarin metabolism in humans,Chem. Res. Toxicol. 22 (2009) 1239–1245.

10] W. Naidong, P.R. Ring, C. Midtlien, X. Jiang, Development and validation of asensitive and robust LC-tandem MS method for the analysis of warfarin enan-tiomers in human plasma, J. Pharm. Biomed. Anal. 25 (2001) 219–226.

11] M.J. Kwon, H.J. Kim, J.W. Kim, K.H. Lee, K.H. Sohn, H.J. Cho, Y.K. On, J.S. Kim, S.Y.Lee, Determination of plasma warfarin concentrations in Korean patients andits potential for clinical application, Korean J. Lab. Med. 29 (2009) 515–523.

12] D.R. Jones, G. Boysen, G.P. Miller, Novel multi-mode ultra performance liquidchromatography-tandem mass spectrometry assay for profiling enantiomerichydroxywarfarins and warfarin in human plasma, J. Chromatogr. B 879 (2011)1056–1062.

13] M.A. Radwan, G.A. Bawazeer, N.M. Aloudah, B.T. AlQuadeib, H.Y. Aboul-Enein,Determination of free and total warfarin concentrations in plasma using UPLCMS/MS and its application to a patient samples, Biomed. Chromatogr. 26 (2012)6–11.

14] M. Wang, Z. Cai, L. Xu, Coupling of acetonitrile deproteinization and salting-out extraction with acetonitrile stacking in chiral capillary electrophoresisfor the determination of warfarin enantiomers, J. Chromatogr. A 1218 (2011)4045–4051.

15] H. Takahashi, T. Kashima, S. Kimura, N. Muramoto, H. Nakahata, S. Kubo,Y. Shimoyama, M. Kajiwara, H. Echizen, Determination of unbound warfarinenantiomers in human plasma and 7-hydroxywarfarin in human urine by chi-ral stationary-phase liquid chromatography with ultraviolet or fluorescenceand on-line circular dichroism detection, J. Chromatogr. B 701 (1997) 71–80.

16] A.M. Rosengren, B.C.G. Karlsson, I. Näslund, P.O. Andersson, I.A. Nicholls, In situdetection of warfarin using time-correlated single-photon counting, Biochem.Biophys. Res. Commun. 407 (2011) 60–62.

17] L. Ye, K. Mosbach, Molecular imprinting: Synthetic materials as substitutes forbiological antibodies and receptors, Chem. Mater. 20 (2008) 859–868.

18] F. Lanza, B. Sellergren, The application of molecular imprinting technology tosolid phase extraction, Chromatographia 53 (2001) 599–611.

19] L.I. Andersson, Molecular imprinting for drug bioanalysis: A review on theapplication of imprinted polymers to solid-phase extraction and binding assay,J. Chromatogr. B 739 (2000) 163–173.

20] J. Zhang, Y. Wang, R. Lv, L. Xu, Electrochemical tolazoline sensor based on gold

nanoparticles and imprinted poly-o-aminothiophenol film, Electrochim. Acta55 (2010) 4039–4044.

21] B.R. Hart, K.J. Shea, Synthetic peptide receptors: Molecularly imprinted poly-mers for the recognition of peptides using peptide–metal interactions, J. Am.Chem. Soc. 123 (2001) 2072–2073.

[

[

tors B 196 (2014) 539–545

22] R. Shoji, T. Takeuchi, I. Kubo, Atrazine sensor based on molecularlyimprinted polymer-modified gold electrode, Anal. Chem. 75 (2003)4882–4886.

23] T. Jing, H. Xia, J. Niu, Y. Zhou, Q. Dai, Q. Hao, Y. Zhou, S. Mei, Determina-tion of trace 2,4-dinitrophenol in surface water samples based on hydrophilicmolecularly imprinted polymers/nickel fiber electrode, Biosens. Bioelectron.26 (2011) 4450–4456.

24] Z. Jie, H. Xiwen, Study of the nature of recognition in molecularly imprintedpolymer selective for 2-aminopyridine, Anal. Chim. Acta 381 (1999) 85–91.

25] A. Bossi, S.A. Piletsky, E.V. Piletska, P.G. Righetti, A.P.F. Turner, Surface-graftedmolecularly imprinted polymers for protein recognition, Anal. Chem. 73 (2001)5281–5286.

26] X. Wang, L. Wang, X. He, Y. Zhang, L. Chen, A molecularly imprinted polymer-coated nanocomposite of magnetic nanoparticles for estrone recognition,Talanta 78 (2009) 327–332.

27] D. Lakshmi, A. Bossi, M.J. Whitcombe, I. Chianella, S.A. Fowler, S. Subrah-manyam, E.V. Piletska, S.A. Piletsky, Electrochemical sensor for catechol anddopamine based on a catalytic molecularly imprinted polymer-conductingpolymer hybrid recognition element, Anal. Chem. 81 (2009) 3576–3584.

28] T. Zołek, P. Lulinski, D. Maciejewska, A computational model for selectiv-ity evaluation of 2-(3,4-dimethoxyphenyl)ethylamine (homoveratrylamine)imprinted polymers towards biogenic compounds, Anal. Chim. Acta 693 (2011)121–129.

29] X. Kan, Y. Zhao, Z. Geng, Z. Wang, J. Zhu, Composites of multiwalled carbonnanotubes and molecularly imprinted polymers for dopamine recognition, J.Phys. Chem. C 112 (2008) 4849–4854.

30] P. Chen, P. Nien, C. Hu, K. Ho, Detection of uric acid based on multi-walledcarbon nanotubes polymerized with a layer of molecularly imprinted PMAA,Sens. Actuators B: Chem. 146 (2010) 466–471.

31] Z. Zhang, Y. Hu, H. Zhang, S. Yao, Novel layer-by-layer assembly molecularlyimprinted sol–gel sensor for selective recognition of clindamycin based on Auelectrode decorated by multi-wall carbon nanotube, J.Colloid Interface Sci. 344(2010) 158–164.

32] Z. Zhang, Y. Hu, H. Zhang, L. Luo, S. Yao, Electrochemical layer-by-layer mod-ified imprinted sensor based on multi-walled carbon nanotubes and sol–gelmaterials for sensitive determination of thymidine, J. Electroanal. Chem. 644(2010) 7–12.

33] Y.J. Zou, C. Xiang, L.X. Sun, F. Xu, Glucose biosensor based on electrode-position of platinum nanoparticles onto carbon nanotubes and immobi-lizing enzyme with chitosan-SiO2 sol–gel, Biosens. Bioelectron. 23 (2008)1010–1016.

34] M.C. Blanco-Lopez, M.J. Lobo-Castanon, A.J. Miranda-Ordieres, P. Tunon-Blanco,Electrochemical sensors based on molecularly imprinted polymers, TrendsAnal. Chem. 23 (2004) 36–48.

35] A. Aghaei, M.R.M. Hosseini, M. Najafi, A novel capacitive biosensor for choles-terol assay that uses an electropolymerized molecularly imprinted polymer,Electrochim. Acta 55 (2010) 1503–1508.

36] W. Song, Y. Chen, J. Xu, X.R. Yang, D.B. Tian, Dopamine sensor based on molec-ularly imprinted electrosynthesized polymers, J Solid State Electrochem. 14(2010) 1909–1914.

37] C. Xie, S. Gao, Q. Guo, K. Xu, Electrochemical sensor for 2,4-dichlorophenoxyacetic acid using molecularly imprinted polypyrrole membrane as recognitionelement, Microchim. Acta 169 (2010) 145–152.

38] A. Gómez-Caballero, A. Ugarte, A. Sánchez-Ortega, N. Unceta, M.A. Goicolea,R.J. Barrio, Molecularly imprinted poly[tetra(o-aminophenyl)porphyrin] as astable and selective coating for the development of voltammetric sensors, J.Electroanal. Chem. 638 (2010) 246–253.

39] J. Huang, X. Zhang, Q. Lin, X. He, X. Xing, H. Huai, W. Lian, H.Zhu, Electrochemical sensor based on imprinted sol–gel and nanomate-rials for sensitive determination of bisphenol A, Food Control 22 (2011)786–791.

40] J. Huang, X. Xing, X. Zhang, X. He, Q. Lin, W. Lian, H. Zhu, A molecularly imprintedelectrochemical sensor based on multiwalled carbon nanotube-gold nanopar-ticle composites and chitosan for the detection of tyramine, Food Res. Int. 44(2011) 276–281.

41] X.C. Fu, X. Chen, Z. Guo, C.G. Xie, L.T. Kong, J.H. Liu, X.J. Huang, Stripping voltam-metric detection of mercury(II) based on a surface ion imprinting strategy inelectropolymerized microporous poly(2-mercaptobenzothiazole) films modi-fied glassy carbon electrode, Anal. Chim. Acta 685 (2011) 21–28.

42] X. Kan, T. Liu, H. Zhou, C. Li, B. Fang, Molecular imprinting polymer electrosensorbased on gold nanoparticles for theophylline recognition and determination,Microchim. Acta 171 (2010) 423–429.

43] Y. Liu, Q.J. Song, L. Wang, Development and characterization of an ampero-metric sensor for triclosan detection based on electropolymerized molecularlyimprinted polymer, Microchem. J. 91 (2009) 222–226.

44] X. Shen, Y. Cui, Y. Pang, H. Qian, Pre-concentration and in situ electrochemicalsensing of 1-hydroxypyrene on an electrodeposited poly(3-methylthiophene)film modified electrode, J. Electroanal. Chem. 667 (2012) 1–6.

45] L. Feng, Y. Liu, Y. Tan, J. Hu, Biosensor for the determination of sorbitol basedon molecularly imprinted electrosynthesized polymers, Biosens. Bioelectron.19 (2004) 1513–1519.

46] G. Frens, Controlled nucleation for the regulation of the particle size inmonodisperse gold suspensions, Nat. Phys. Sci. 241 (1973) 20–22.

47] S. Liu, Y. Li, J. Li, L. Jiang, Enhancement of DNA immobilization and hybridizationon gold electrode modified by nanogold aggregates, Biosens. Bioelectron. 21(2005) 789–795.

Actua

B

nology, Iran (2008). Currently, he received his PhD degree

B. Rezaei et al. / Sensors and

iographies

Behzad Rezaei is professor of chemistry, Isfahan Uni-versity of Technology (IUT), Isfahan, Iran. BS of Isfahan

University, 1988; MS of Isfahan University in AnalyticalChemistry, 1991; PhD of Isfahan University of Technol-ogy in Analytical Chemistry, 1999. His research interestscover electrochemical, flow injection method of analysisand electrochemical sensors.

tors B 196 (2014) 539–545 545

Omid Rahmanian received his BS degree in chemistryfrom Damghan University of Basic Science, Iran in 2005and completed MS degree in Isfahan University of Tech-

in Department of Chemistry in Isfahan University of Tech-nology, Iran. His interested fields include application ofmolecularly imprinted polymer (MIP) and nanomaterialsin electrochemistry and other detection systems.