- Email: [email protected]
Detection of nicotine based on molecularly imprinted TiO2-modified electrodes
Analytica Chimica Acta 633 (2009) 119–126
Contents lists available at ScienceDirect
Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca
Detection of nicotine based on molecularly imprinted TiO2 -modified electrodes Cheng-Tar Wu a , Po-Yen Chen a , Jian-Ging Chen a , Vembu Suryanarayanan a , Kuo-Chuan Ho a,b,∗ a b
Department of Chemical Engineering, National Taiwan University, Taipei 10617, Taiwan Institute of Polymer Science and Engineering, National Taiwan University, Taipei 10617, Taiwan
a r t i c l e
i n f o
Article history: Received 11 July 2008 Received in revised form 13 November 2008 Accepted 14 November 2008 Available online 25 November 2008 Keywords: Amperometry Biosensor Nicotine Molecularly imprinted TiO2 Scanning electrochemical microscopy Poly(3,4-ethylenedioxythiophene)
a b s t r a c t Amperometric detection of nicotine (NIC) was carried out on a titanium dioxide (TiO2 )/poly(3,4ethylenedioxythiophene) (PEDOT)-modified electrode by a molecular imprinting technique. In order to improve the conductivity of the substrate, PEDOT was coated onto the sintered electrode by in situ electrochemical polymerization of the monomer. The sensing potential of the NIC-imprinted TiO2 electrode (ITO/TiO2 [NIC]/PEDOT) in a phosphate-buffered saline (PBS) solution (pH 7.4) containing 0.1 M KCl was determined to be 0.88 V (vs. Ag/AgCl/saturated KCl). The linear detection range for NIC oxidation on the so-called ITO/TiO2 [NIC]/PEDOT electrode was 0–5 mM, with a sensitivity and limit of detection of 31.35 A mM−1 cm−2 and 4.9 M, respectively. When comparing with the performance of the non-imprinted one, the sensitivity ratio was about 1.24. The sensitivity enhancement was attributed to the increase in the electroactive area of the imprinted electrode. The at-rest stability of the ITO/TiO2 [NIC]/PEDOT electrode was tested over a period of 3 days. The current response remained about 85% of its initial value at the end of 2 days. The ITO/TiO2 [NIC]/PEDOT electrode showed reasonably good selectivity in distinguishing NIC from its major interferent, (−)-cotinine (COT). Moreover, scanning electrochemical microscopy (SECM) was employed to elucidate the surface morphology of the imprinted and non-imprinted electrodes using Fe(CN)6 3− /Fe(CN)6 4− as a redox probe on a platinum tip. The imprinted electrode was further characterized by scanning electron microscopy (SEM) and Fourier transform infrared spectroscopy (FTIR). © 2008 Published by Elsevier B.V.
1. Introduction Nicotine (NIC) belongs to the alkaloid family which exists in tobacco leaves [1], and it is an important drug for treating Alzheimer’s disease (AD) [2]. The cognitive effect and clinical diagnosis of NIC have been summarized [3]. However, the frequent intake of NIC in the human body creates a few side effects such as increased blood pressure and heart beat, acceleration of the respiratory organs, and stimulation of the central nervous system leading to disruption of arteries and cardiovascular risk factors. The toxicity of NIC is about 60 mg, as exhibited by its fatal dose in adults [1]. Currently, more than 1.2 billion people in the world consume some type of tobacco product [4], and this intake affects human metabolism resulting in an increase in the amount of low-density lipoprotein (LDL) [5] and injuries to our health. Hence, it is indeed necessary to develop a quick and precise sensing technique for detecting NIC in the human body.
∗ Corresponding author at: Department of Chemical Engineering, National Taiwan University, Taipei 10617, Taiwan. Tel.: +886 2 2366 0739; fax: +886 2 2362 3040. E-mail address: [email protected] (K.-C. Ho). 0003-2670/$ – see front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.aca.2008.11.038
A number of methods for NIC detection in the human body or in tobacco samples have so far been reported, including highperformance liquid chromatography (HPLC) with different kinds of detectors such as ultraviolet (HPLC-UV) [6] and electrochemical detector (HPLC-EC) [7], and high-sensitivity gas chromatography mass spectrometry (GC–MS) [8]. A molecularly imprinted polymeric (MIP) technique has become important in recent years for product separation with solid-phase extraction (SPE) [9,10] and clinical assays where selectivity is achieved by adjusting the monomer-template molar ratios [11]. The quartz crystal microbalance (QCM) in the thickness-shear-mode (TSM) [12] and surface plasmon resonance (SPR) [13] were also reported, and an inhibition biosensor using electrochemical sensing coupled with enzymes has also been well documented [14,15]. Although recently MIPs involving artificial antibodies have been proposed either for biomolecular recognition [16] or environmental sensing [17], most MIPs have been used for SPE in chromatography [9,10]. In some cases, MIPs are immobilized in a bulk polymer [18,19]. Although innumerable methods have been developed for detecting NIC in the human body and tobacco samples, the HPLC technique is still normally employed in clinical laboratories for accurate sensing [20–22]. However, this method is time-consuming and involves a number of steps.
120
C.-T. Wu et al. / Analytica Chimica Acta 633 (2009) 119–126
When compared to other methods, electrochemical sensing has been considered simple, cheap, and reliable. On this basis, amperometric detection of NIC was carried out by a molecular imprinting technique. NIC was imprinted on a TiO2 matrix and coated on an indium tin oxide (ITO) electrode. To improve the conductivity, 3,4-ethylenedioxythiophene (EDOT) was electrochemically polymerized on the electrode to obtain an NIC-imprinted TiO2 / PEDOT-modified electrode (denoted ITO/TiO2 [NIC]/PEDOT). Subsequently, a non-imprinted PEDOT-modified electrode (denoted ITO/TiO2 /PEDOT) was also fabricated for NIC sensing. Scanning electrochemical microscopy (SECM) was used to elucidate the different surface morphologies of both modified electrodes based on their redox chemistry. In recent years, applications of SECM in the field of electrochemistry have been rapid, and involve imaging and microfabrication [23,24]. In fact, almost all kinds of electrochemical measurements such as potentiometry, cyclic voltammetry, or AC voltammetry can be recorded with an SECM setup. The spatial resolution of SECM greatly increases the utility of electrochemical techniques for characterizing interfaces, especially for electrode materials used as potential electrocatalysts and biosensors [25]. In this work, SECM was employed to elucidate the surface morphology of imprinted and non-imprinted electrodes using Fe(CN)6 3− /Fe(CN)6 4− as the redox couple. The imprinted electrode was further characterized by scanning electron microscopy (SEM) and Fourier transform infrared (FTIR) spectroscopy. 2. Experimental procedures 2.1. Chemicals and conducting glass substrates Acetonitrile (CH3 CN, 99.7%) and nitric acid (HNO3 , 60%) were provided by JT Baker and Nacalai Tesque, respectively. (−)-Nicotine (NIC) (C10 H14 N2 , >99.0%) was furnished by Fluka. (−)-Cotinine (COT) (C10 H12 N2 O, >98%) powder and phosphatebuffered saline (PBS, pH 7.4) tablets were supplied by Sigma. Titanium(IV) isopropoxide (Ti[OCH(CH3 )2 ]4 , 98%) and 4-Å molecular sieves (8–12 mesh) were obtained from Acros. Potassium chloride (KCl, 99.0–100.5%), lithium perchlorate (LiClO4 , >95%), 3,4-ethylenedioxythiophene (EDOT) (C6 H6 O2 S), and potassium ferricyanide (K3 (FeCN)6 , >99%) were all purchased from Aldrich. Tri-sodium citrate was purchased from MP Biomedicals. Swine whole blood was drawn by venipuncture from a healthy swine donor and anti-coagulated with tri-sodium citrate (3.2–3.8 wt.%) in a 9:1 volumetric ratio. Hydrochloric acid (HCl, 32%) and poly(ethylene glycol) (PEG) (H(OCH2 CH2 )n OH, MW 20,000) were obtained from Merck. Deionized water (DIW) was used throughout this research, and all measurements in this study were carried out at room temperature. The ITO-conducting glass substrates, which served as the working electrodes in this work, were supplied by Ritdisplay, Hsinchu Industrial Park, Taiwan (Rsh = 20 −1 ). Prior to use, the ITO substrates (4.0 × 1.0 cm) were ultrasonically cleaned using both 0.1 N HCl and 0.1 N NaOH solutions for 5 min each, and then totally rinsed in ultrasonic DIW for 5 min. Finally, the substrates were dried in air. Before the electrodes were fabricated for the amperometric measurements and topographical mapping, conducting Cu tape (3 M) was applied to one edge of the substrate as the bus bar. The Cu bus bar was helpful in providing a uniform current to the sensing area. 2.2. Apparatus A pH meter (Orion, model 720A) was utilized for the pH measurements. An autoclave composed of a pressure reactor (Parr Instrument, model 4560) linked with a temperature controller (Parr
Instrument, model 4843) was used to grow the TiO2 particles. Electropolymerization of the monomer and amperometric measurements were performed using a potentiostat/galvanostat (CH Instruments, model 440). A three-electrode system was employed for electrochemical sensing. The ITO/TiO2 /PEDOT electrode and the ITO/TiO2 [NIC]/PEDOT-modified electrodes were used as the working electrodes. A Pt plate (4.0 cm × 1.0 cm) and Ag/AgCl/saturated KCl (homemade) served as the counter and reference electrodes, respectively. SECM (CH Instruments, model 900B) was used for the topographical mapping. A typical four-electrode system was assembled for the SECM setup, where a 10-m platinum SECM tip (CH Instruments, model 116p) was used to investigate the topographical mapping of the ITO/TiO2 /PEDOT- and ITO/TiO2 [NIC]/PEDOT-modified electrodes. A Pt wire and Ag/AgCl/saturated KCl (homemade) served as the counter and reference electrodes, respectively. Fourier transform infrared (FTIR) spectroscopy was carried out with an FTIR spectrometer (Thermo Nicolet, model Nexus 470) at the Center for Nano Science and Technology, National Taiwan University, Taipei, Taiwan. 2.3. Preparation of the ITO/TiO2 [NIC]/PEDOT and ITO/TiO2 /PEDOT electrodes The preparation procedure of the modified working electrodes for this work is schematically shown in Fig. 1 and is described as follows. The schematic diagrams for the ITO/TiO2 /PEDOT and ITO/TiO2 [NIC]/PEDOT electrodes are shown in Fig. 1(a) and (b), respectively. Fig. 1(a) depicts a large portion of insulating surface that is covered with mesoporous TiO2 film for the non-imprinted ITO/TiO2 /PEDOT electrode, thus the non-imprinted electrode provides less conductive and continuous paths to the substrate when scanning the tip in the presence of the redox couple Fe(CN)6 3− /Fe(CN)6 4− . In contrast, the imprinted ITO/TiO2 [NIC]/PEDOT electrode, as shown in Fig. 1(b), represents more conductive and continuous paths to the substrate. 2.3.1. Synthesis of TiO2 colloids Colloidal TiO2 was prepared by a sol–gel process in acid medium according to a procedure reported in the literature [26,27] Titanium(IV) isopropoxide (72 mL) was added to 430 mL of a 0.1 M nitric acid aqueous solution with constant stirring while simultaneously being heated to 85 ◦ C for 8 h. When the mixture was cooled down to room temperature, the resultant colloid was filtered, and the filtered colloid was heated in an autoclave at a temperature of 240 ◦ C for 12 h in order to allow the growth of TiO2 particles. The TiO2 colloid was concentrated to 13 wt.%, and finally 30 wt.% (with respect to TiO2 ), and PEG (MW 20,000) was added to prevent the film from cracking during drying. The synthesis step for a TiO2 film and the precise control of its particle size were reported previously in our earlier study [28]. 2.3.2. Preparation of ITO/TiO2 [NIC] imprinting and the ITO/TiO2 electrodes The NIC liquid was added to the TiO2 colloids and stirred homogenously to form the 0.125 M NIC mixed solution. The paste was uniformly coated onto pre-cleaned ITO electrodes with a glass rod, followed by air-drying for 30 min. Then the film was heated to 500 ◦ C at a rate of 20 ◦ C min−1 , and maintained for 30 min before cooling to room temperature. During this heat treatment, all NIC molecules were removed from the substrate (note that the boiling point of NIC is about 247 ◦ C), and there was no need for any extraction procedure using solvents. The complete removal of NIC from the surface was ensured by checking the NIC oxidation current response at 0.85 V (vs. Ag/AgCl/saturated KCl) from the
C.-T. Wu et al. / Analytica Chimica Acta 633 (2009) 119–126
121
Fig. 1. Schematic for the preparation of (a) ITO/TiO2 /PEDOT and (b) ITO/TiO2 [NIC]/PEDOT-modified electrodes.
cyclic voltammogram (CV) of the imprinted electrode. Thus, an NICimprinted TiO2 -modified electrode (ITO/TiO2 [NIC]) was obtained. In the same way, the non-imprinted electrode was fabricated without NIC and was used for the comparative study. 2.3.3. Electropolymerization of EDOT An effective area of 0.5 cm × 0.5 cm on both the imprinted and non-imprinted electrodes was obtained with polyimide insulating tape. Both of the working electrodes were dipped in an anhydrous acetonitrile solution containing 0.01 M EDOT monomer and 0.1 M LiClO4 . Prior to electropolymerization, the electrolyte was deoxygenated by purging with nitrogen for 5 min. A constant potential of 1.2 V (vs. Ag/Ag+ ) for 5 s was applied using a potentiostat. The corresponding charges passed during the electrochemical deposition of PEDOT for the ITO/TiO2 [NIC]/PEDOT and ITO/TiO2 /PEDOT electrodes were 6.0 ± 0.04 and 5.4 ± 0.03 mC, respectively. After electropolymerization, both thin films, namely ITO/TiO2 [NIC]/PEDOT and ITO/TiO2 /PEDOT, were cleaned with acetonitrile, dried with N2 , and then stored in air. From our preparation steps for the imprinted electrode of ITO/TiO2 [NIC]/PEDOT, as seen in Fig. 1, the imprinted cavity is formed first after sintering, then followed by the electropolymerization of EDOT. This means that the electrodeposited charge of PEDOT on the ITO/TiO2 [NIC] electrode would depend only on the original open cavities on the ITO/TiO2 electrode and the imprinted NIC open cavities. 2.4. Electrochemical measurements Since the solution’s pH value does affect the electrochemistry of some biomolecules, it was our intention to eliminate the influence of pH on NIC oxidation using a buffer solution. To do so, the
following experiments were carried out in PBS solution at pH 7.4 containing 0.1 M KCl. We employed a PBS due to its common use and easier for comparison with others. It is known that phosphate interacts very strongly with TiO2 , however, we only compare the difference between the electrochemical responses of non-imprinted and imprinted electrodes. Furthermore, by adding KCl as the supporting electrolyte into the sensing solution, it is expected to increase the electrolytic conductance and thus enhance the reliability for the electrochemical detection. All electrochemical measurements done in this section (Section 2.4) were repeated at least for three times, and the corresponding error bars were calculated based on the standard error of the mean from the statistical analysis. 2.4.1. Catalytic ability test To test the catalytic ability of the NIC oxidation peak on the bare ITO, and ITO/TiO2 [NIC]- and ITO/TiO2 [NIC]/PEDOT-modified electrodes, all electrodes were separately bathed in a 5 mM NIC solution containing 0.1 M KCl. Afterward, a CV was recorded with a scan rate set to 50 mV/s, and the CV was used to determine the NIC oxidation current response of each electrode. 2.4.2. Polarization curve Before the sensing measurement, a suitable operating potential was obtained in the limiting, or maximum, current zone resulting from the mass transfer control regime. This was to ensure that the maximum sensing current would mainly be proportional to the concentration of NIC so as to minimize the influence of any current noise. The ITO/TiO2 [NIC]/PEDOT electrode was kept in the background solution containing 0.1 M KCl/PBS and linear sweep
122
C.-T. Wu et al. / Analytica Chimica Acta 633 (2009) 119–126
voltammograms (LSVs) were obtained at the regulated potentials of 0.6–1.0 V with a scan rate of 1 mV s−1 . The net current was calculated by subtracting the background current from the current obtained after the addition of 5 mM NIC. Thus, a proper operating potential was determined from the polarization curve. Moreover, LSVs were also carried out on the bare ITO electrode in order to verify the catalytic behaviors of the different electrodes, including the bare ITO and ITO/TiO2 [NIC]/PEDOT-modified electrodes. 2.4.3. Amperometric detection of NIC After acquiring a suitable operating potential, a typical chronoamperometric I–t curve was recorded with our sensing experiment. First of all, the background current was equilibrated for 200 s, and then the added NIC amount was adjusted. We used the same electrode to obtain the sensing current as a function of NIC concentration up to 5 mM. The same experiment was repeated for three times with three fresh electrodes. We did not carry out the electrode regeneration for this particular experiment. Concentrations of NIC in the bulk solution were varied from 0 to 5 mM, and at each concentration level, the components were thoroughly mixed with a magnetic bar with a rotation speed of 250 rpm for 20 s. In order to obtain a steady-state current, the cell was kept free of disturbance for 200 s. The net current was obtained in the same manner described above. The calibration curves were constructed by calculating the net steady-state current densities at various NIC concentrations ranging from 0 to 5 mM, for the imprinted and nonimprinted electrodes. The sensitivity was the slope of the calibrated current vs. concentration, and the detection limit (based on S/N = 3) of the sensor was also obtained. The detection limit is calculated based the IUPAC recommendation, which is shown by the following equation DL =
kSB b
(1)
where DL, k, and SB are the detection limit, a constant value of 3 (S/N = signal to noise ratio), and the standard deviation of the blank signals, respectively. b is the slope of the calibration curve. Furthermore, the interference of COT [1] was evaluated by adding 0–5 mM of COT to the same electrolyte solution. The interference effects with the equimolar coexistence of NIC and COT in the mixture were also compared. 2.4.4. At-rest stability test In order to explore the lifetime of the ITO/TiO2 [NIC]/PEDOT sensor, the at-rest stability of the sensor was tested over a period of 3 days to see if a deposable EC MIP biosensor is feasible. The sampling interval and the applied potential were fixed at 12 h and 0.88 V to record the net current response for 5 mM NIC oxidized at the modified ITO/TiO2 [NIC]/PEDOT electrode. Between each measurement, the modified ITO/TiO2 [NIC]/PEDOT electrode was regenerated by flushing with acetonitrile, drying with N2 , and then storing it in air before the next measurement. 2.4.5. Sample test for swine plasma In the case of investigating the practical utility of this assay, we use swine platelet-poor plasma (PPP) as the bulk solution to detection NIC dissolved in it. This swine plasma sample is to provide the preliminary data for the sensor facing the complicated biological fluids and hopefully to give us the drawbacks of this senor in future research. PPP isolated from swine whole blood was done by centrifugation at 3000 × g (10 min, 26 ◦ C). The swine PPP acted as the bulk solution during the amperometric detection of 5 mM NIC. The amperometric detection procedure was followed from Section 2.4.3.
2.5. Topographical mappings by SECM A four-electrode system was employed for the SECM instrument, and SECM mapping topographies were obtained in a solution containing 5 mM K3 Fe(CN)6 as a mediator and 0.1 M KCl as the electrolyte. In order to do topographical mapping, the tip must be kept a distance of 10 m from both the ITO/TiO2 /PEDOT and ITO/TiO2 [NIC]/ PEDOT electrodes. The mapping area was 100 m × 100 m. The recorded data were plotted in a three-dimensional figure, and in addition, the tick labels of the X-, Y-, and Z-axes were representative of the distances of the tip on the X- and Y-axes and the normalized current, respectively. A normalized current (iT /iT,∞ ) is defined as the redox current response at the electrode surface (iT ) divided by the limiting current in the bulk solution (iT,∞ ). Finally, in order to characterize the interaction between NIC and the TiO2 matrix, an FTIR spectrum was generated for the imprinted electrode bathed in 2.5 mM NIC. 3. Results and discussion 3.1. The catalytic effect of the conducting PEDOT The CVs of 5 mM of NIC in a 0.1 M KCl solution containing PBS on (a) ITO/TiO2 [NIC]/PEDOT; (a ) ITO/TiO2 /PEDOT; (b) bare ITO; (c) ITO/TiO2 [NIC]; and (c ) ITO/TiO2 electrodes are shown in Fig. 2. The electrochemical behaviors for NIC oxidation on the ITO/TiO2 /PEDOT and ITO/TiO2 [NIC]/PEDOT electrodes are very similar, as seen in Fig. 2 (curves a and a ), even though the EDOT polymerization on the non-imprinted and imprinted TiO2 electrodes involved different amounts of charges due to the different electroactive areas. The conducting nature of the PEDOT film is mainly responsible for the capacitive currents for both the ITO/TiO2 /PEDOT and ITO/TiO2 [NIC]/PEDOT electrodes. The effect of the electroactive surface area does play an important role when comparing the sensing performance of the ITO/TiO2 /PEDOT electrode with that of the ITO/TiO2 [NIC]/PEDOT electrode. Also, the electrochemical behaviors for NIC oxidation on the bare ITO, ITO/TiO2 [NIC], and ITO/TiO2 electrodes are almost identical, which are shown in Fig. 2 (curves b, c and c ). From Fig. 2, it is clear that the modified ITO electrode electropolymerized with PEDOT enhanced the oxidation response of NIC. The above investigation confirms that PEDOT acts as a strong catalyst for oxidizing NIC, confirming our previous results [29].
Fig. 2. Cyclic voltammograms of 5 mM nicotine (NIC) in a PBS solution on (a) ITO/TiO2 [NIC]/PEDOT; (a ) ITO/TiO2 /PEDOT; (b) bare ITO; (c) ITO/TiO2 [NIC]; and (c ) ITO/TiO2 electrodes. Note that curve (b) overlaps with curves (c) and (c ).
C.-T. Wu et al. / Analytica Chimica Acta 633 (2009) 119–126
Fig. 3. Sensing performance of the ITO/TiO2 /PEDOT and ITO/TiO2 [NIC]/PEDOT electrodes. Calibration curves were obtained from the i–t curves.
3.2. Amperometric detection of NIC and its major interferent, COT 3.2.1. Sensing performance between the imprinted and non-imprinted electrodes It is necessary to choose a suitable potential for comparing the sensing performances between the imprinted and non-imprinted electrodes. The operating potential for amperometric detection of NIC was determined from the LSV curve to be 0.88 V. From the above studies (Section 3.1), it was noted that the PEDOTmodified electrode enhanced the current response and lowered the overpotential for NIC oxidation, compared with the oxidation process on the bare ITO electrode. Hence, the next step was to compare the electrochemical responses of the molecularly imprinted (ITO/TiO2 [NIC]/PEDOT) and non-imprinted electrodes (ITO/TiO2 /PEDOT) for detecting NIC. Fig. 3 shows the calibration curves obtained from typical steadystate current–time responses for NIC oxidation at concentrations ranging from 0 to 5 mM. The experiment in the amperometric sensing mode was repeated three times with a fresh electrode for each measurement, and the result is shown as error bars in the figure. In Fig. 3, it can be noted that the current response carried out on the imprinted electrode was statistically higher than that on the non-imprinted electrode. The limit of detection (LOD) was found to be 4.9 M, and the linear detection range (LDR) was between 0 and 5 mM, with a sensitivity value of 31.4 A mM−1 cm−2 (R2 = 0.997) for the imprinted electrode. The sensing performances (LOD and sensitivity) of the non-imprinted electrode were 6.4 M and 25.2 A mM−1 cm−2 , respectively. However, to now, there is no specific definition to distinguish the performance between MIPs and non-imprinted polymers (NMIPs). In order to quantitatively determine the attribution contributed by the template imprinting, a normalized value of the imprinting efficiency (IE) is defined as the bound amount of MIP divided by the bound amount of NMIP Eq. (2): Imprinting efficiency (IE) =
Bound amount of MIP Bound amount of NMIP
(2)
In an amperometric MIP biosensor system, it is difficult to measure the bound amount on modified MIP electrodes. Therefore, a sensitivity enhancement (SE) factor was modified from the definition of IE described above as the ratio of MIP sensitivity over NMIP sensitivity Eq. (3): Sensitivity enhancement (SE) =
Sensitivity of MIP electrode Sensitivity of NMIP electrode
(3)
123
Therefore, the SE of the electrode produced in this work was about 1.24. The reason for the small extent of sensitivity enhancement (only 1.24) can be explained by the mesoporous structure of the TiO2 matrix, which is accessible for EDOT polymerization, even though the TiO2 acts simply as an inert and insulated matrix. We believe that the sensitivity enhancement of the imprinted electrode was derived from the imprinted NIC’s open cavities. In fact, the difference in current response is caused by those imprinted open cavities, which were formed by NIC and naked on the ITO surface, that could be further electroplated with PEDOT. It is those naked open cavities that show the enhanced sensitivity. A partial list of IE values for non-EC MIP sensors and SE values for electrochemical (EC) MIP sensors reported in literatures [12,30–39] are tabulated in Table 1. From this table, it can be noted that the SE of the electrode produced for this work compares well with the SE obtained from other published EC-based sensors. From Table 1, it can be noted that the SE value lies between 1 and 3 for most EC-based MIP sensors. Although the SE of our work is not as high compared to some others, this work represents a first preliminary example of an amperometric nicotine sensor imprinted on oxides of a transition metal immobilized with PEDOT that shows promising results. 3.2.2. Determination of the electroactive surface area The effect of the area of the electroactive surface might shed light on the observed sensing response. In fact, an increase of 24% in sensitivity (a sensitivity enhancement factor of 1.24) can be approximately explained by an increase in the electroactive area. There are various methods leading to the estimation of the electroactive surface area, however, the values obtained from different methods vary considerably. One of the effective ways to estimate the electroactive surface area is based on the integral form of the Cottrell equation, which describes the chronoamperometric current response to a potential step under diffusion control: 1/2
i(t) =
nFADo Co∗ 1/2 t 1/2
(4)
where i(t) is the transient current response, n is the number of electron transfer involved, F is Faraday’s constant, A is the electroactive area, Do is the diffusion coefficient of EDOT, and Co∗ is the concentration of EDOT. By integrating the Cottrell equation from t = 0, one obtains the cumulative charge, Q(t), passed in electropolymerizing EDOT: 1/2
Q (t) =
2nFADo Co∗ t 1/2 1/2
(5)
In Section 2.3.3, we found that the EDOT polymerization on the imprinted (ITO/TiO2 [NIC]/PEDOT) and non-imprinted (ITO/TiO2 /PEDOT) electrodes involved different amounts of charge during a potential step of 1.2 V for 5 s. The recorded charges for the electropolymerization of EDOT for the imprinted and nonimprinted electrodes were 6.0 ± 0.04 and 5.4 ± 0.03 mC, respectively. Assuming the following parameters: n = 1, F = 96,487 C/equiv., Do = 10−5 cm2 s−1 , Co∗ = 10−5 mol cm−3 , t = 5 s, one obtains the electroactive surface areas of 0.78 and 0.70 cm2 for the imprinted and non-imprinted electrodes, respectively. These values are reasonable since both areas are larger than the geometrical area, which is 0.25 cm2 . 3.2.3. Interference effects In the human body, COT has been reported to be the principal metabolite of NIC in blood and urine [12,22]. Hence, COT was considered the key interferent in this study. The corresponding calibration curve is shown in Fig. 4. Curves (a), (b), and (c) correspond to
124
C.-T. Wu et al. / Analytica Chimica Acta 633 (2009) 119–126
Table 1 A partial list of IE obtained for non-EC and SE for EC MIP sensors from the literatures. Functional monomersa
Transduction/substrateb
Templatec
IE
Ref.
Non-EC MIP sensors/assays MAA/EGDMA MAA/EGDMA MAA/DVB TEOS/PTMOS/APTES MAA/EGDMA AN/St AN/2-Py AN AN/4-Py
HPLC-UV HPLC-UV HPLC QCM QC-TSM/Ag QCM/Au QCM/Au QCM/Au QCM/Au
THO CAF THO PT NIC CAF CAF CAF CAF
2.36 1.84 2.26 1.78 2.44 1.07 1.30 1.73 1.88
[30] [30] [31] [32] [12] [33] [33] [33] [33]
Functional monomersa
Transduction/substrateb
Templatec
SE
Ref.
EC MIP sensors/assays MAA/EGDMA MAA/EGDMA EGDMA/4-Py EGDMA/1-VID/4-VPB EGDMA/1-VID/4-VPB EGDMA/1-VID/4-VPB MAA/Zn2+ /DVB RSC/o-PD TiO2
CV/ITO CV/Au DPV/C Amperometry/CPE Amperometry/CPE Amperometry/CPE Amperometry/CPE Amperometry/Au Amperometry/PEDOT/ITO
THO Atrazine 2,4-D Fru-val m-val m--lys D4NP 2,4-D NIC
2.97 1.74 1-4.67 1.71 1.20 1.44 ∼1.4 2.65 1.24
[34] [35] [36] [37] [37] [37] [38] [39] This work
a Monomers, MAA: methacrylic acid; EGDMA: ethylene glycol dimethacrylate; DVB: divinylbenzene; TEOS: tetraethyl orthosilicate; PTMOS: phenyltrimethoxysilane; APTES: aminopropyltriethoxysilane, AN: acrilonitrile; St: styrene; 2-Py: 2-vinylpyridine; 4-Py: 4-vinylpyridine; p-VBBA: p-vinylbenzeneboronic acid; HEMA: 2-hydroxyethyl methacrylate; 1-VID: 1-vinylimidazole; 4-VPB: 4-vinylphenylboronate; RSC: resorcinol; o-PD: o-phenylenediamine; EDOT: 3,4-ethylenedioxythiophene; TEAMA: N,N,Ntrimethylaminoethyl methacrylate. TRIM: trimethylolpropane trimethacrylate. b Transduction and substrate, FPD: flame photometric detector; TSM: thickness-shear-mode; DPV: different pulse voltammetry; ITO: indium tin oxide electrode; CPE: carbon paste electrode; PEDOT: poly(3,4-ethylenedioxythiophene). c Template, THO: theophylline; CAF: caffeine; PT: parathion; NIC: nicotine; 2,4-D: 2,4-dichlorophenoxyacetic acid; Fru-val: fructosyl valine; m-val: methylvaline; m--lys: methyllysine (Z); D4NP: diethyl(4-nitrobenzyl).
the oxidation current responses of the added NIC, the mixture with equimolar coexistence of NIC and COT, and COT alone, respectively. Curves (a) and (b) reveal that the ITO/TiO2 [NIC]/PEDOT electrode exhibited high current responses particularly for NIC oxidation. Based on the oxidation current responses, it can be noted that the ITO/TiO2 [NIC]/PEDOT film shows higher selectivity for NIC, where NIC is adsorbed onto specific sites at the applied potential of 0.88 V. It should be mentioned here that COT is not electroactive at 0.88 V, and hence would not contribute to the oxidation current response under the equimolar coexistence of NIC and COT. Nevertheless, the formation of hydrogen bonds between COT molecules and hydroxyl groups on the TiO2 surface is considered to be stronger, as compared to that of NIC ones. That is to say, the adsorption capability of COT on the imprinted electrode surface is better than that of NIC. Therefore, we believed that even if COT shows no electrochemical behavior at 0.88 V, COT could be considered as a good interferent, especially under the coexistence of NIC and COT.
The above investigations revealed that NIC molecules occupy specific sites on the TiO2 nanostructure during coating and leave imprinted shapes during the thermal treatment (extraction process). These imprinted sites are only selective for NIC molecules, and the adsorption of COT by these specific sites was almost negligible. Hence, the ITO/TiO2 [NIC]/PEDOT electrode can distinguish NIC from COT in terms of the sensing performance. As shown in curve (b) of Fig. 4, it appears that the oxidation current response of the equimolar coexistence of NIC and COT was somewhat lower than with NIC alone. This may have been due to the fact that COT may slightly block a few imprinting sites on the ITO/TiO2 [NIC]/PEDOT electrode. On the basis of the above results, it was concluded that the imprinted sensor, ITO/TiO2 [NIC]/PEDOT, can be used in an environment in which NIC and COT coexist.
Fig. 4. Calibration curves for (a) nicotine (NIC) alone; (b) equimolar coexistence of an NIC and (−)-cotinine (COT) mixture; and (c) COT alone.
Fig. 5. The stability data for the ITO/TiO2 [NIC]/PEDOT electrode. The electrode was tested in a 5 mM NIC dissolved in a PBS solution for a period of 3 days.
C.-T. Wu et al. / Analytica Chimica Acta 633 (2009) 119–126
Fig. 6. The current responses and their relative current responses for NIC oxidation at 5 mM in PBS and swine PPP at the ITO/TiO2 /PEDOT and ITO/TiO2 [NIC]/PEDOT electrodes.
3.2.4. At-rest stability Since tailor-made biosensors based on molecularly imprinted TiO2 -modified electrode reported in this study could be an alternative to enzyme biosensors, the at-rest stability of the NIC-imprinted biosensor was also obtained. Fig. 5 shows the stability data for the ITO/TiO2 [NIC]/PEDOT electrode tested in a 5 mM NIC dissolved in a PBS solution for a period of 3 days. The current response for 5 mM NIC remained about 85% of its initial value at the end of 2 days. That is to say that the present NIC-imprinted TiO2 -modified electrode not only be a disposable sensor, but can act as a reusable sensor during 2 days used and shows a reasonable response. The sudden decay in response after 2 days probably has to do with the preservation of the electrode at ambient condition. 3.2.5. Sensing results for swine plasma Fig. 6 shows the current responses and their relative current responses (against the molecularly imprinted electrode for detecting NIC in PBS) for NIC oxidation at 5 mM in PBS and swine PPP. The pH value of swine PPP is about 8.1, because tri-sodium citrate was
125
added as the anticoagulant. However, only 34% of the relative current response for NIC was observed in swine PPP. We also performed the standard addition of NIC at lower concentration levels from 0 to 0.75 mM. The relative current responses for the detection of NIC in swine PPP were still about one-third of those in the PBS. This is identical to the result obtained for the detection of 5 mM NIC. Therefore, it is clear that the matrix effect does exist when detecting NIC in swine PPP. It is assumed that the reduction capability of the citrate may reduce the oxidation current in detecting NIC. Moreover, we have simplified the complicated bio-sample. Instead of using swine whole blood, PPP was obtained by centrifugation. Swine PPP still contains platelets, proteins, carbohydrates, electrolytes, fat, vitamins, hormone and metabolites. These chemical compounds are known to influence the obtained results. This means that we need to modify the sensor when dealing with a realistic biological fluid in future study. Although the modified electrode showed 34% relative current response of NIC when dealing with the complicated swine blood plasma, the SE still remained about 1.23, which was consistent with aforementioned result acquired from the PBS system. Therefore, we believed that the imprinting concept does work in various sensing environments investigated in this study. 3.3. Topographical mapping by SECM SECM can be used to elucidate the surface morphology on both active and inactive substrates by means of the increase or decrease in redox current responses. In our study, the surface morphologies of imprinted and non-imprinted electrodes could be distinguished using SECM. Before carrying out topographical mapping, the CV was recorded to obtain the applied potential between the tip and substrate. Here, the applied potential between the tip and substrate was set to 0 and 0.5 V, respectively, to characterize the Fe(CN)6 3− and Fe(CN)6 4− redox coupled reaction. Although the ITO acted as the active substrate for the sensing electrode, negative feedback was seen in the approach curve. It was found that the TiO2 on the substrate acted as an insulating layer causing negative feedback, which is consistent with our aforementioned data. Next, the tip was withdrawn by a distance of 10 m from the substrate surface. This avoided the tip being passivated as a result of direct close contact with the roughened TiO2 surface during topographical mapping.
Fig. 7. 3-D topographical mapping on the (a) ITO/TiO2 /PEDOT and (b) ITO/TiO2 [NIC]/PEDOT electrodes. The applied potential at the tip and substrates were 0 and 0.5 V (vs. Ag/AgCl/saturated KCl), respectively.
126
C.-T. Wu et al. / Analytica Chimica Acta 633 (2009) 119–126
Fig. 7a and b show the topographical mapping of the tip based on the redox reaction that occurred within the 10-m vertical region on both electrodes. The normalized current (iT /iT,∞ ) is indicated on the Z-axis. An irregular pattern was found for the TiO2 /PEDOT electrode without imprinted NIC (Fig. 7a), whereas it was very smooth on the NIC-imprinted surface (Fig. 7b). In addition, the average current response of the NIC-imprinted electrode was higher than that of the non-imprinted one, which was correlated with a higher amount of PEDOT formed in the former system. Moreover, the pore size of the imprinted electrode was also larger than that of the nonimprinted one, which was confirmed by the SEM images. The results of the SEM and SECM images infer that NIC was imprinted on underlying sites that were naked to the ITO electrode. These imprinted sites are consistent with larger pore sizes of the imprinted electrode observed under SEM, which thus resulted in a larger charge capacity for PEDOT during electropolymerization, compared to that using the non-imprinted one with smaller pore sizes. Therefore, a moreuniform and higher average current response was observed for the imprinted ITO/TiO2 [NIC]/PEDOT electrode (Fig. 7b) in contrast to a less-uniform and lower average current response obtained for the non-imprinted ITO/[TiO2 ]/PEDOT electrode (Fig. 7a). 3.4. FTIR studies In order to characterize the interaction between NIC and the TiO2 matrix, an FTIR spectrum was taken for the imprinted electrode bathed in 2.5 mM NIC. The spectrum revealed an intensity peak contributed by hydrogen bond formation between nitrogen atoms of NIC molecules and hydroxyl groups on the TiO2 surface [40]. The result is also consistent with the literature [41]. 4. Conclusions In conclusion, a novel NIC electrochemical sensor by using the concept of NIC-imprinted TiO2 has been investigated in this study. The imprinted (ITO/TiO2 [NIC]/PEDOT) electrode showed a higher response current than the non-imprinted one (ITO/TiO2 / PEDOT). The sensitivity of the imprinted (31.4 A mM−1 cm−2 ) film was about 1.24 times higher than the non-imprinted (25.2 A mM−1 cm−2 ) film, i.e., the SE was about 1.24. The LOD was calculated to be 4.9 M. The linear detection region was from 0 to 5 mM, and this offers a possible method for NIC sensing. Moreover, the oxidation current response for 5 mM NIC still remained about 85% of its initial value at 48 h, indicating reasonable stability of the electrode. According to this study, the NIC-imprinted TiO2 (ITO/TiO2 [NIC]/PEDOT) electrode showed reasonably good selectivity in distinguishing NIC from COT. Besides, according to our preliminary test with swine PPP, the imprinted electrodes could detect NIC in a complicated blood plasma system. It is heartening to notice that the SE in PPP still maintained at 1.23, which is very close to the value obtained in PBS. The imprinted and non-imprinted electrodes were characterized by SECM and SEM. SECM mapping provides a novel usage to recognize the template-imprinted cavities on the surface of the modified electrode. The FTIR spectra confirmed the formation of hydrogen bonds between NIC and TiO2 on the imprinted electrode. Although SE values for EC MIP sensors reported in the literatures are comparable with the SE obtained in this work, more work is now being carried out in our laboratory to improve the SE to 2–3. Future investigations will focus on miniaturing the sensing device on a chip and enhancing the SE of the presented NIC sensor.
Acknowledgements This work was financially supported by the National Research Council of Taiwan under grant numbers NSC95-2221-E-002-307 and NSC96-2220-E-006-015. The electrochemical instrumentation used in this work was supported by the Ministry of Education (MOE)’s Program for Promoting Academic Excellence of University under grant number EX-91-E-FA09-5-4. Appendix A. Supplementary data Both electrochemical and analytical data, including LSV, I–t curves for the NIC oxidation, SEM images, and FTIR spectrum, can be found in the supporting information at doi:10.1016/j.aca. 2008.11.038. References [1] J.S. Meyer, L.F. Quenzer, Psychopharmacology: Drugs, the Brain and Behavior, first ed., Sinauer Associates, Sunderland, 2005. [2] A.L. Wilson, L.K. Langley, J. Monley, T. Bauer, S. Rottunda, E. McFalls, C. Kovera, J.R. McCaretn, Pharmacol. Biochem. Behav. 51 (1995) 509. [3] A.H. Rezvani, E.D. Levin, Biol. Psychiatry 49 (2001) 258. [4] T.T. Denton, X.D. Zhang, J.R. Cashman, J. Med. Chem. 48 (2005) 224. [5] J. Cluette-Brown, J. Mulligan, K. Doyle, S. Hagan, T. Osmolski, J. Hojnacki, Proc. Soc. Exp. Biol. Med. 182 (1986) 409. [6] M. Page-Sharp, T.W. Hale, L.P. Hackett, J.H. Kristensen, K.F. Ilett, J. Chromatogr. Biomed. Sci. Appl. 796 (2003) 173. [7] G.N. Mahoney, W. Al-Delaimy, J. Chromatogr., Biomed. Appl. 753 (2001) 179. [8] H.S. Shin, J.G. Kim, Y.J. Shin, S.H. Jee, J. Chromatogr. Biomed. Sci. Appl. 769 (2002) 177. [9] A. Zander, P. Findlay, T. Renner, B. Sellergren, Anal. Chem. 70 (1998) 3304. [10] W.M. Mullett, E.P.C. Lai, B. Sellergren, Anal. Commun. 36 (1999) 217. [11] H.S. Andersson, J.G. Karlsson, S.A. Piletsky, A.-C. Koch-Schmidt, K. Mosbach, I.A. Nicholls, J. Chromatogr. A 848 (1999) 39. [12] Y. Tan, J. Yin, C. Liang, H. Peng, L. Nie, S. Yao, Bioelectrochemistry 53 (2001) 141. [13] T. Abdallah, S. Abdalla, S. Negm, H. Talaat, Sens. Actuators. A Phys. 102 (2003) 234. [14] L. Campanella, G. Favero, M. Tomassetti, Anal. Lett. 34 (2001) 855. [15] Y. Yang, M. Yang, H. Wang, L. Tang, Anal. Chim. Acta 509 (2004) 151. [16] K. Mosbach, Trends Biochem. Sci. 19 (1994) 9. [17] R.J. Ansell, D. Kriz, K. Mosbach, Curr. Opin. Biotechnol. 7 (1996) 89. [18] A.G. Mayes, K. Mosbach, Anal. Chem. 68 (1996) 3769. [19] K. Haupt, K. Mosbach, Chem. Rev. 100 (2000) 2495. [20] H.H. Maurer, J. Chromatogr., Biomed. Appl. 713 (1998) 3. [21] M. Nakajima, T. Yamamoto, Y. Kuroiwaand, T. Yokoi, J. Chromatogr., Biomed. Appl. 742 (2000) 211. [22] I.N. Papadoyannis, V.F. Samanidou, P.G. Stefanidou, J. Liq. Chromatogr. Relat. Technol. 25 (2002) 2315. [23] F. Zhou, P.R. Unwin, A.J. Bard, J. Phys. Chem. 96 (1992) 4917. [24] M. Tsionsky, A.J. Bard, M.V. Mirkin, J. Phys. Chem. 100 (1996) 17881. [25] M.V. Mirkin, B.R. Horrocks, Anal. Chim. Acta 406 (2000) 119. [26] Ch.J. Barbe, F. Arendse, P. Comte, M. Jirousek, F. Lenzmann, V. Shklover, M. Gratzel, J. Am. Ceram. Soc. 80 (1997) 3157. [27] M.K. Nazeeruddin, R. Humphry-Baker, P. Liska, M. Gratzel, J. Phys. Chem. B 107 (2003) 8981. [28] C.Y. Huang, Y.C. Hsu, J.G. Chen, V. Suryanarayanan, K.M. Lee, K.C. Ho, Sol. Energy Mater. Sol. Cells 90 (2006) 2391. [29] W.M. Yeh, K.C. Ho, Anal. Chim. Acta 542 (2005) 76. [30] J.M. Hong, P.E. Anderson, J. Qian, C.R. Martin, Chem. Mater. 10 (1998) 1029. [31] J. Wang, P.A.G. Cormack, D.C. Sherrington, E. Khoshdel, Angew. Chem. Int. Ed. 42 (2003) 5336. [32] S. Marx, A. Zaltsman, I. Turyan, D. Mandler, Anal. Chem. 76 (2004) 120. [33] T. Kobayashi, Y. Murawaki, P.S. Reddy, M. Abe, N. Fujii, Anal. Chim. Acta 435 (2001) 141. [34] Y. Yoshimi, R. Ohdaira, C. Iiyama, K. Sakai, Sens. Actuators B Chem. 73 (2001) 49. [35] R. Shoji, T. Takeuchi, I. Kubo, Anal. Chem. 75 (2003) 4882. [36] S. Kröger, A.P.F. Turner, K. Mosbach, K. Haupt, Anal. Chem. 71 (1999) 3698. [37] K. Sode, S. Ohta, Y. Yanai, T. Yamazaki, Biosens. Bioelectron. 18 (2003) 1485. [38] Z. Meng, T. Yamazaki, K. Sode, Biotechnol. Lett. 25 (2003) 1075. [39] H.H. Weetall, D.W. Hatchett, K.R. Rogers, Electroanalysis 17 (2005) 1789. [40] H. Park, W. Choi, J. Phys. Chem. B 108 (2004) 4086. [41] V. Arnaud, M. Berthelot, J.-Y. Le Questel, J. Phys. Chem. A 109 (2005) 3767.
Contents lists available at ScienceDirect
Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca
Detection of nicotine based on molecularly imprinted TiO2 -modified electrodes Cheng-Tar Wu a , Po-Yen Chen a , Jian-Ging Chen a , Vembu Suryanarayanan a , Kuo-Chuan Ho a,b,∗ a b
Department of Chemical Engineering, National Taiwan University, Taipei 10617, Taiwan Institute of Polymer Science and Engineering, National Taiwan University, Taipei 10617, Taiwan
a r t i c l e
i n f o
Article history: Received 11 July 2008 Received in revised form 13 November 2008 Accepted 14 November 2008 Available online 25 November 2008 Keywords: Amperometry Biosensor Nicotine Molecularly imprinted TiO2 Scanning electrochemical microscopy Poly(3,4-ethylenedioxythiophene)
a b s t r a c t Amperometric detection of nicotine (NIC) was carried out on a titanium dioxide (TiO2 )/poly(3,4ethylenedioxythiophene) (PEDOT)-modified electrode by a molecular imprinting technique. In order to improve the conductivity of the substrate, PEDOT was coated onto the sintered electrode by in situ electrochemical polymerization of the monomer. The sensing potential of the NIC-imprinted TiO2 electrode (ITO/TiO2 [NIC]/PEDOT) in a phosphate-buffered saline (PBS) solution (pH 7.4) containing 0.1 M KCl was determined to be 0.88 V (vs. Ag/AgCl/saturated KCl). The linear detection range for NIC oxidation on the so-called ITO/TiO2 [NIC]/PEDOT electrode was 0–5 mM, with a sensitivity and limit of detection of 31.35 A mM−1 cm−2 and 4.9 M, respectively. When comparing with the performance of the non-imprinted one, the sensitivity ratio was about 1.24. The sensitivity enhancement was attributed to the increase in the electroactive area of the imprinted electrode. The at-rest stability of the ITO/TiO2 [NIC]/PEDOT electrode was tested over a period of 3 days. The current response remained about 85% of its initial value at the end of 2 days. The ITO/TiO2 [NIC]/PEDOT electrode showed reasonably good selectivity in distinguishing NIC from its major interferent, (−)-cotinine (COT). Moreover, scanning electrochemical microscopy (SECM) was employed to elucidate the surface morphology of the imprinted and non-imprinted electrodes using Fe(CN)6 3− /Fe(CN)6 4− as a redox probe on a platinum tip. The imprinted electrode was further characterized by scanning electron microscopy (SEM) and Fourier transform infrared spectroscopy (FTIR). © 2008 Published by Elsevier B.V.
1. Introduction Nicotine (NIC) belongs to the alkaloid family which exists in tobacco leaves [1], and it is an important drug for treating Alzheimer’s disease (AD) [2]. The cognitive effect and clinical diagnosis of NIC have been summarized [3]. However, the frequent intake of NIC in the human body creates a few side effects such as increased blood pressure and heart beat, acceleration of the respiratory organs, and stimulation of the central nervous system leading to disruption of arteries and cardiovascular risk factors. The toxicity of NIC is about 60 mg, as exhibited by its fatal dose in adults [1]. Currently, more than 1.2 billion people in the world consume some type of tobacco product [4], and this intake affects human metabolism resulting in an increase in the amount of low-density lipoprotein (LDL) [5] and injuries to our health. Hence, it is indeed necessary to develop a quick and precise sensing technique for detecting NIC in the human body.
∗ Corresponding author at: Department of Chemical Engineering, National Taiwan University, Taipei 10617, Taiwan. Tel.: +886 2 2366 0739; fax: +886 2 2362 3040. E-mail address: [email protected] (K.-C. Ho). 0003-2670/$ – see front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.aca.2008.11.038
A number of methods for NIC detection in the human body or in tobacco samples have so far been reported, including highperformance liquid chromatography (HPLC) with different kinds of detectors such as ultraviolet (HPLC-UV) [6] and electrochemical detector (HPLC-EC) [7], and high-sensitivity gas chromatography mass spectrometry (GC–MS) [8]. A molecularly imprinted polymeric (MIP) technique has become important in recent years for product separation with solid-phase extraction (SPE) [9,10] and clinical assays where selectivity is achieved by adjusting the monomer-template molar ratios [11]. The quartz crystal microbalance (QCM) in the thickness-shear-mode (TSM) [12] and surface plasmon resonance (SPR) [13] were also reported, and an inhibition biosensor using electrochemical sensing coupled with enzymes has also been well documented [14,15]. Although recently MIPs involving artificial antibodies have been proposed either for biomolecular recognition [16] or environmental sensing [17], most MIPs have been used for SPE in chromatography [9,10]. In some cases, MIPs are immobilized in a bulk polymer [18,19]. Although innumerable methods have been developed for detecting NIC in the human body and tobacco samples, the HPLC technique is still normally employed in clinical laboratories for accurate sensing [20–22]. However, this method is time-consuming and involves a number of steps.
120
C.-T. Wu et al. / Analytica Chimica Acta 633 (2009) 119–126
When compared to other methods, electrochemical sensing has been considered simple, cheap, and reliable. On this basis, amperometric detection of NIC was carried out by a molecular imprinting technique. NIC was imprinted on a TiO2 matrix and coated on an indium tin oxide (ITO) electrode. To improve the conductivity, 3,4-ethylenedioxythiophene (EDOT) was electrochemically polymerized on the electrode to obtain an NIC-imprinted TiO2 / PEDOT-modified electrode (denoted ITO/TiO2 [NIC]/PEDOT). Subsequently, a non-imprinted PEDOT-modified electrode (denoted ITO/TiO2 /PEDOT) was also fabricated for NIC sensing. Scanning electrochemical microscopy (SECM) was used to elucidate the different surface morphologies of both modified electrodes based on their redox chemistry. In recent years, applications of SECM in the field of electrochemistry have been rapid, and involve imaging and microfabrication [23,24]. In fact, almost all kinds of electrochemical measurements such as potentiometry, cyclic voltammetry, or AC voltammetry can be recorded with an SECM setup. The spatial resolution of SECM greatly increases the utility of electrochemical techniques for characterizing interfaces, especially for electrode materials used as potential electrocatalysts and biosensors [25]. In this work, SECM was employed to elucidate the surface morphology of imprinted and non-imprinted electrodes using Fe(CN)6 3− /Fe(CN)6 4− as the redox couple. The imprinted electrode was further characterized by scanning electron microscopy (SEM) and Fourier transform infrared (FTIR) spectroscopy. 2. Experimental procedures 2.1. Chemicals and conducting glass substrates Acetonitrile (CH3 CN, 99.7%) and nitric acid (HNO3 , 60%) were provided by JT Baker and Nacalai Tesque, respectively. (−)-Nicotine (NIC) (C10 H14 N2 , >99.0%) was furnished by Fluka. (−)-Cotinine (COT) (C10 H12 N2 O, >98%) powder and phosphatebuffered saline (PBS, pH 7.4) tablets were supplied by Sigma. Titanium(IV) isopropoxide (Ti[OCH(CH3 )2 ]4 , 98%) and 4-Å molecular sieves (8–12 mesh) were obtained from Acros. Potassium chloride (KCl, 99.0–100.5%), lithium perchlorate (LiClO4 , >95%), 3,4-ethylenedioxythiophene (EDOT) (C6 H6 O2 S), and potassium ferricyanide (K3 (FeCN)6 , >99%) were all purchased from Aldrich. Tri-sodium citrate was purchased from MP Biomedicals. Swine whole blood was drawn by venipuncture from a healthy swine donor and anti-coagulated with tri-sodium citrate (3.2–3.8 wt.%) in a 9:1 volumetric ratio. Hydrochloric acid (HCl, 32%) and poly(ethylene glycol) (PEG) (H(OCH2 CH2 )n OH, MW 20,000) were obtained from Merck. Deionized water (DIW) was used throughout this research, and all measurements in this study were carried out at room temperature. The ITO-conducting glass substrates, which served as the working electrodes in this work, were supplied by Ritdisplay, Hsinchu Industrial Park, Taiwan (Rsh = 20 −1 ). Prior to use, the ITO substrates (4.0 × 1.0 cm) were ultrasonically cleaned using both 0.1 N HCl and 0.1 N NaOH solutions for 5 min each, and then totally rinsed in ultrasonic DIW for 5 min. Finally, the substrates were dried in air. Before the electrodes were fabricated for the amperometric measurements and topographical mapping, conducting Cu tape (3 M) was applied to one edge of the substrate as the bus bar. The Cu bus bar was helpful in providing a uniform current to the sensing area. 2.2. Apparatus A pH meter (Orion, model 720A) was utilized for the pH measurements. An autoclave composed of a pressure reactor (Parr Instrument, model 4560) linked with a temperature controller (Parr
Instrument, model 4843) was used to grow the TiO2 particles. Electropolymerization of the monomer and amperometric measurements were performed using a potentiostat/galvanostat (CH Instruments, model 440). A three-electrode system was employed for electrochemical sensing. The ITO/TiO2 /PEDOT electrode and the ITO/TiO2 [NIC]/PEDOT-modified electrodes were used as the working electrodes. A Pt plate (4.0 cm × 1.0 cm) and Ag/AgCl/saturated KCl (homemade) served as the counter and reference electrodes, respectively. SECM (CH Instruments, model 900B) was used for the topographical mapping. A typical four-electrode system was assembled for the SECM setup, where a 10-m platinum SECM tip (CH Instruments, model 116p) was used to investigate the topographical mapping of the ITO/TiO2 /PEDOT- and ITO/TiO2 [NIC]/PEDOT-modified electrodes. A Pt wire and Ag/AgCl/saturated KCl (homemade) served as the counter and reference electrodes, respectively. Fourier transform infrared (FTIR) spectroscopy was carried out with an FTIR spectrometer (Thermo Nicolet, model Nexus 470) at the Center for Nano Science and Technology, National Taiwan University, Taipei, Taiwan. 2.3. Preparation of the ITO/TiO2 [NIC]/PEDOT and ITO/TiO2 /PEDOT electrodes The preparation procedure of the modified working electrodes for this work is schematically shown in Fig. 1 and is described as follows. The schematic diagrams for the ITO/TiO2 /PEDOT and ITO/TiO2 [NIC]/PEDOT electrodes are shown in Fig. 1(a) and (b), respectively. Fig. 1(a) depicts a large portion of insulating surface that is covered with mesoporous TiO2 film for the non-imprinted ITO/TiO2 /PEDOT electrode, thus the non-imprinted electrode provides less conductive and continuous paths to the substrate when scanning the tip in the presence of the redox couple Fe(CN)6 3− /Fe(CN)6 4− . In contrast, the imprinted ITO/TiO2 [NIC]/PEDOT electrode, as shown in Fig. 1(b), represents more conductive and continuous paths to the substrate. 2.3.1. Synthesis of TiO2 colloids Colloidal TiO2 was prepared by a sol–gel process in acid medium according to a procedure reported in the literature [26,27] Titanium(IV) isopropoxide (72 mL) was added to 430 mL of a 0.1 M nitric acid aqueous solution with constant stirring while simultaneously being heated to 85 ◦ C for 8 h. When the mixture was cooled down to room temperature, the resultant colloid was filtered, and the filtered colloid was heated in an autoclave at a temperature of 240 ◦ C for 12 h in order to allow the growth of TiO2 particles. The TiO2 colloid was concentrated to 13 wt.%, and finally 30 wt.% (with respect to TiO2 ), and PEG (MW 20,000) was added to prevent the film from cracking during drying. The synthesis step for a TiO2 film and the precise control of its particle size were reported previously in our earlier study [28]. 2.3.2. Preparation of ITO/TiO2 [NIC] imprinting and the ITO/TiO2 electrodes The NIC liquid was added to the TiO2 colloids and stirred homogenously to form the 0.125 M NIC mixed solution. The paste was uniformly coated onto pre-cleaned ITO electrodes with a glass rod, followed by air-drying for 30 min. Then the film was heated to 500 ◦ C at a rate of 20 ◦ C min−1 , and maintained for 30 min before cooling to room temperature. During this heat treatment, all NIC molecules were removed from the substrate (note that the boiling point of NIC is about 247 ◦ C), and there was no need for any extraction procedure using solvents. The complete removal of NIC from the surface was ensured by checking the NIC oxidation current response at 0.85 V (vs. Ag/AgCl/saturated KCl) from the
C.-T. Wu et al. / Analytica Chimica Acta 633 (2009) 119–126
121
Fig. 1. Schematic for the preparation of (a) ITO/TiO2 /PEDOT and (b) ITO/TiO2 [NIC]/PEDOT-modified electrodes.
cyclic voltammogram (CV) of the imprinted electrode. Thus, an NICimprinted TiO2 -modified electrode (ITO/TiO2 [NIC]) was obtained. In the same way, the non-imprinted electrode was fabricated without NIC and was used for the comparative study. 2.3.3. Electropolymerization of EDOT An effective area of 0.5 cm × 0.5 cm on both the imprinted and non-imprinted electrodes was obtained with polyimide insulating tape. Both of the working electrodes were dipped in an anhydrous acetonitrile solution containing 0.01 M EDOT monomer and 0.1 M LiClO4 . Prior to electropolymerization, the electrolyte was deoxygenated by purging with nitrogen for 5 min. A constant potential of 1.2 V (vs. Ag/Ag+ ) for 5 s was applied using a potentiostat. The corresponding charges passed during the electrochemical deposition of PEDOT for the ITO/TiO2 [NIC]/PEDOT and ITO/TiO2 /PEDOT electrodes were 6.0 ± 0.04 and 5.4 ± 0.03 mC, respectively. After electropolymerization, both thin films, namely ITO/TiO2 [NIC]/PEDOT and ITO/TiO2 /PEDOT, were cleaned with acetonitrile, dried with N2 , and then stored in air. From our preparation steps for the imprinted electrode of ITO/TiO2 [NIC]/PEDOT, as seen in Fig. 1, the imprinted cavity is formed first after sintering, then followed by the electropolymerization of EDOT. This means that the electrodeposited charge of PEDOT on the ITO/TiO2 [NIC] electrode would depend only on the original open cavities on the ITO/TiO2 electrode and the imprinted NIC open cavities. 2.4. Electrochemical measurements Since the solution’s pH value does affect the electrochemistry of some biomolecules, it was our intention to eliminate the influence of pH on NIC oxidation using a buffer solution. To do so, the
following experiments were carried out in PBS solution at pH 7.4 containing 0.1 M KCl. We employed a PBS due to its common use and easier for comparison with others. It is known that phosphate interacts very strongly with TiO2 , however, we only compare the difference between the electrochemical responses of non-imprinted and imprinted electrodes. Furthermore, by adding KCl as the supporting electrolyte into the sensing solution, it is expected to increase the electrolytic conductance and thus enhance the reliability for the electrochemical detection. All electrochemical measurements done in this section (Section 2.4) were repeated at least for three times, and the corresponding error bars were calculated based on the standard error of the mean from the statistical analysis. 2.4.1. Catalytic ability test To test the catalytic ability of the NIC oxidation peak on the bare ITO, and ITO/TiO2 [NIC]- and ITO/TiO2 [NIC]/PEDOT-modified electrodes, all electrodes were separately bathed in a 5 mM NIC solution containing 0.1 M KCl. Afterward, a CV was recorded with a scan rate set to 50 mV/s, and the CV was used to determine the NIC oxidation current response of each electrode. 2.4.2. Polarization curve Before the sensing measurement, a suitable operating potential was obtained in the limiting, or maximum, current zone resulting from the mass transfer control regime. This was to ensure that the maximum sensing current would mainly be proportional to the concentration of NIC so as to minimize the influence of any current noise. The ITO/TiO2 [NIC]/PEDOT electrode was kept in the background solution containing 0.1 M KCl/PBS and linear sweep
122
C.-T. Wu et al. / Analytica Chimica Acta 633 (2009) 119–126
voltammograms (LSVs) were obtained at the regulated potentials of 0.6–1.0 V with a scan rate of 1 mV s−1 . The net current was calculated by subtracting the background current from the current obtained after the addition of 5 mM NIC. Thus, a proper operating potential was determined from the polarization curve. Moreover, LSVs were also carried out on the bare ITO electrode in order to verify the catalytic behaviors of the different electrodes, including the bare ITO and ITO/TiO2 [NIC]/PEDOT-modified electrodes. 2.4.3. Amperometric detection of NIC After acquiring a suitable operating potential, a typical chronoamperometric I–t curve was recorded with our sensing experiment. First of all, the background current was equilibrated for 200 s, and then the added NIC amount was adjusted. We used the same electrode to obtain the sensing current as a function of NIC concentration up to 5 mM. The same experiment was repeated for three times with three fresh electrodes. We did not carry out the electrode regeneration for this particular experiment. Concentrations of NIC in the bulk solution were varied from 0 to 5 mM, and at each concentration level, the components were thoroughly mixed with a magnetic bar with a rotation speed of 250 rpm for 20 s. In order to obtain a steady-state current, the cell was kept free of disturbance for 200 s. The net current was obtained in the same manner described above. The calibration curves were constructed by calculating the net steady-state current densities at various NIC concentrations ranging from 0 to 5 mM, for the imprinted and nonimprinted electrodes. The sensitivity was the slope of the calibrated current vs. concentration, and the detection limit (based on S/N = 3) of the sensor was also obtained. The detection limit is calculated based the IUPAC recommendation, which is shown by the following equation DL =
kSB b
(1)
where DL, k, and SB are the detection limit, a constant value of 3 (S/N = signal to noise ratio), and the standard deviation of the blank signals, respectively. b is the slope of the calibration curve. Furthermore, the interference of COT [1] was evaluated by adding 0–5 mM of COT to the same electrolyte solution. The interference effects with the equimolar coexistence of NIC and COT in the mixture were also compared. 2.4.4. At-rest stability test In order to explore the lifetime of the ITO/TiO2 [NIC]/PEDOT sensor, the at-rest stability of the sensor was tested over a period of 3 days to see if a deposable EC MIP biosensor is feasible. The sampling interval and the applied potential were fixed at 12 h and 0.88 V to record the net current response for 5 mM NIC oxidized at the modified ITO/TiO2 [NIC]/PEDOT electrode. Between each measurement, the modified ITO/TiO2 [NIC]/PEDOT electrode was regenerated by flushing with acetonitrile, drying with N2 , and then storing it in air before the next measurement. 2.4.5. Sample test for swine plasma In the case of investigating the practical utility of this assay, we use swine platelet-poor plasma (PPP) as the bulk solution to detection NIC dissolved in it. This swine plasma sample is to provide the preliminary data for the sensor facing the complicated biological fluids and hopefully to give us the drawbacks of this senor in future research. PPP isolated from swine whole blood was done by centrifugation at 3000 × g (10 min, 26 ◦ C). The swine PPP acted as the bulk solution during the amperometric detection of 5 mM NIC. The amperometric detection procedure was followed from Section 2.4.3.
2.5. Topographical mappings by SECM A four-electrode system was employed for the SECM instrument, and SECM mapping topographies were obtained in a solution containing 5 mM K3 Fe(CN)6 as a mediator and 0.1 M KCl as the electrolyte. In order to do topographical mapping, the tip must be kept a distance of 10 m from both the ITO/TiO2 /PEDOT and ITO/TiO2 [NIC]/ PEDOT electrodes. The mapping area was 100 m × 100 m. The recorded data were plotted in a three-dimensional figure, and in addition, the tick labels of the X-, Y-, and Z-axes were representative of the distances of the tip on the X- and Y-axes and the normalized current, respectively. A normalized current (iT /iT,∞ ) is defined as the redox current response at the electrode surface (iT ) divided by the limiting current in the bulk solution (iT,∞ ). Finally, in order to characterize the interaction between NIC and the TiO2 matrix, an FTIR spectrum was generated for the imprinted electrode bathed in 2.5 mM NIC. 3. Results and discussion 3.1. The catalytic effect of the conducting PEDOT The CVs of 5 mM of NIC in a 0.1 M KCl solution containing PBS on (a) ITO/TiO2 [NIC]/PEDOT; (a ) ITO/TiO2 /PEDOT; (b) bare ITO; (c) ITO/TiO2 [NIC]; and (c ) ITO/TiO2 electrodes are shown in Fig. 2. The electrochemical behaviors for NIC oxidation on the ITO/TiO2 /PEDOT and ITO/TiO2 [NIC]/PEDOT electrodes are very similar, as seen in Fig. 2 (curves a and a ), even though the EDOT polymerization on the non-imprinted and imprinted TiO2 electrodes involved different amounts of charges due to the different electroactive areas. The conducting nature of the PEDOT film is mainly responsible for the capacitive currents for both the ITO/TiO2 /PEDOT and ITO/TiO2 [NIC]/PEDOT electrodes. The effect of the electroactive surface area does play an important role when comparing the sensing performance of the ITO/TiO2 /PEDOT electrode with that of the ITO/TiO2 [NIC]/PEDOT electrode. Also, the electrochemical behaviors for NIC oxidation on the bare ITO, ITO/TiO2 [NIC], and ITO/TiO2 electrodes are almost identical, which are shown in Fig. 2 (curves b, c and c ). From Fig. 2, it is clear that the modified ITO electrode electropolymerized with PEDOT enhanced the oxidation response of NIC. The above investigation confirms that PEDOT acts as a strong catalyst for oxidizing NIC, confirming our previous results [29].
Fig. 2. Cyclic voltammograms of 5 mM nicotine (NIC) in a PBS solution on (a) ITO/TiO2 [NIC]/PEDOT; (a ) ITO/TiO2 /PEDOT; (b) bare ITO; (c) ITO/TiO2 [NIC]; and (c ) ITO/TiO2 electrodes. Note that curve (b) overlaps with curves (c) and (c ).
C.-T. Wu et al. / Analytica Chimica Acta 633 (2009) 119–126
Fig. 3. Sensing performance of the ITO/TiO2 /PEDOT and ITO/TiO2 [NIC]/PEDOT electrodes. Calibration curves were obtained from the i–t curves.
3.2. Amperometric detection of NIC and its major interferent, COT 3.2.1. Sensing performance between the imprinted and non-imprinted electrodes It is necessary to choose a suitable potential for comparing the sensing performances between the imprinted and non-imprinted electrodes. The operating potential for amperometric detection of NIC was determined from the LSV curve to be 0.88 V. From the above studies (Section 3.1), it was noted that the PEDOTmodified electrode enhanced the current response and lowered the overpotential for NIC oxidation, compared with the oxidation process on the bare ITO electrode. Hence, the next step was to compare the electrochemical responses of the molecularly imprinted (ITO/TiO2 [NIC]/PEDOT) and non-imprinted electrodes (ITO/TiO2 /PEDOT) for detecting NIC. Fig. 3 shows the calibration curves obtained from typical steadystate current–time responses for NIC oxidation at concentrations ranging from 0 to 5 mM. The experiment in the amperometric sensing mode was repeated three times with a fresh electrode for each measurement, and the result is shown as error bars in the figure. In Fig. 3, it can be noted that the current response carried out on the imprinted electrode was statistically higher than that on the non-imprinted electrode. The limit of detection (LOD) was found to be 4.9 M, and the linear detection range (LDR) was between 0 and 5 mM, with a sensitivity value of 31.4 A mM−1 cm−2 (R2 = 0.997) for the imprinted electrode. The sensing performances (LOD and sensitivity) of the non-imprinted electrode were 6.4 M and 25.2 A mM−1 cm−2 , respectively. However, to now, there is no specific definition to distinguish the performance between MIPs and non-imprinted polymers (NMIPs). In order to quantitatively determine the attribution contributed by the template imprinting, a normalized value of the imprinting efficiency (IE) is defined as the bound amount of MIP divided by the bound amount of NMIP Eq. (2): Imprinting efficiency (IE) =
Bound amount of MIP Bound amount of NMIP
(2)
In an amperometric MIP biosensor system, it is difficult to measure the bound amount on modified MIP electrodes. Therefore, a sensitivity enhancement (SE) factor was modified from the definition of IE described above as the ratio of MIP sensitivity over NMIP sensitivity Eq. (3): Sensitivity enhancement (SE) =
Sensitivity of MIP electrode Sensitivity of NMIP electrode
(3)
123
Therefore, the SE of the electrode produced in this work was about 1.24. The reason for the small extent of sensitivity enhancement (only 1.24) can be explained by the mesoporous structure of the TiO2 matrix, which is accessible for EDOT polymerization, even though the TiO2 acts simply as an inert and insulated matrix. We believe that the sensitivity enhancement of the imprinted electrode was derived from the imprinted NIC’s open cavities. In fact, the difference in current response is caused by those imprinted open cavities, which were formed by NIC and naked on the ITO surface, that could be further electroplated with PEDOT. It is those naked open cavities that show the enhanced sensitivity. A partial list of IE values for non-EC MIP sensors and SE values for electrochemical (EC) MIP sensors reported in literatures [12,30–39] are tabulated in Table 1. From this table, it can be noted that the SE of the electrode produced for this work compares well with the SE obtained from other published EC-based sensors. From Table 1, it can be noted that the SE value lies between 1 and 3 for most EC-based MIP sensors. Although the SE of our work is not as high compared to some others, this work represents a first preliminary example of an amperometric nicotine sensor imprinted on oxides of a transition metal immobilized with PEDOT that shows promising results. 3.2.2. Determination of the electroactive surface area The effect of the area of the electroactive surface might shed light on the observed sensing response. In fact, an increase of 24% in sensitivity (a sensitivity enhancement factor of 1.24) can be approximately explained by an increase in the electroactive area. There are various methods leading to the estimation of the electroactive surface area, however, the values obtained from different methods vary considerably. One of the effective ways to estimate the electroactive surface area is based on the integral form of the Cottrell equation, which describes the chronoamperometric current response to a potential step under diffusion control: 1/2
i(t) =
nFADo Co∗ 1/2 t 1/2
(4)
where i(t) is the transient current response, n is the number of electron transfer involved, F is Faraday’s constant, A is the electroactive area, Do is the diffusion coefficient of EDOT, and Co∗ is the concentration of EDOT. By integrating the Cottrell equation from t = 0, one obtains the cumulative charge, Q(t), passed in electropolymerizing EDOT: 1/2
Q (t) =
2nFADo Co∗ t 1/2 1/2
(5)
In Section 2.3.3, we found that the EDOT polymerization on the imprinted (ITO/TiO2 [NIC]/PEDOT) and non-imprinted (ITO/TiO2 /PEDOT) electrodes involved different amounts of charge during a potential step of 1.2 V for 5 s. The recorded charges for the electropolymerization of EDOT for the imprinted and nonimprinted electrodes were 6.0 ± 0.04 and 5.4 ± 0.03 mC, respectively. Assuming the following parameters: n = 1, F = 96,487 C/equiv., Do = 10−5 cm2 s−1 , Co∗ = 10−5 mol cm−3 , t = 5 s, one obtains the electroactive surface areas of 0.78 and 0.70 cm2 for the imprinted and non-imprinted electrodes, respectively. These values are reasonable since both areas are larger than the geometrical area, which is 0.25 cm2 . 3.2.3. Interference effects In the human body, COT has been reported to be the principal metabolite of NIC in blood and urine [12,22]. Hence, COT was considered the key interferent in this study. The corresponding calibration curve is shown in Fig. 4. Curves (a), (b), and (c) correspond to
124
C.-T. Wu et al. / Analytica Chimica Acta 633 (2009) 119–126
Table 1 A partial list of IE obtained for non-EC and SE for EC MIP sensors from the literatures. Functional monomersa
Transduction/substrateb
Templatec
IE
Ref.
Non-EC MIP sensors/assays MAA/EGDMA MAA/EGDMA MAA/DVB TEOS/PTMOS/APTES MAA/EGDMA AN/St AN/2-Py AN AN/4-Py
HPLC-UV HPLC-UV HPLC QCM QC-TSM/Ag QCM/Au QCM/Au QCM/Au QCM/Au
THO CAF THO PT NIC CAF CAF CAF CAF
2.36 1.84 2.26 1.78 2.44 1.07 1.30 1.73 1.88
[30] [30] [31] [32] [12] [33] [33] [33] [33]
Functional monomersa
Transduction/substrateb
Templatec
SE
Ref.
EC MIP sensors/assays MAA/EGDMA MAA/EGDMA EGDMA/4-Py EGDMA/1-VID/4-VPB EGDMA/1-VID/4-VPB EGDMA/1-VID/4-VPB MAA/Zn2+ /DVB RSC/o-PD TiO2
CV/ITO CV/Au DPV/C Amperometry/CPE Amperometry/CPE Amperometry/CPE Amperometry/CPE Amperometry/Au Amperometry/PEDOT/ITO
THO Atrazine 2,4-D Fru-val m-val m--lys D4NP 2,4-D NIC
2.97 1.74 1-4.67 1.71 1.20 1.44 ∼1.4 2.65 1.24
[34] [35] [36] [37] [37] [37] [38] [39] This work
a Monomers, MAA: methacrylic acid; EGDMA: ethylene glycol dimethacrylate; DVB: divinylbenzene; TEOS: tetraethyl orthosilicate; PTMOS: phenyltrimethoxysilane; APTES: aminopropyltriethoxysilane, AN: acrilonitrile; St: styrene; 2-Py: 2-vinylpyridine; 4-Py: 4-vinylpyridine; p-VBBA: p-vinylbenzeneboronic acid; HEMA: 2-hydroxyethyl methacrylate; 1-VID: 1-vinylimidazole; 4-VPB: 4-vinylphenylboronate; RSC: resorcinol; o-PD: o-phenylenediamine; EDOT: 3,4-ethylenedioxythiophene; TEAMA: N,N,Ntrimethylaminoethyl methacrylate. TRIM: trimethylolpropane trimethacrylate. b Transduction and substrate, FPD: flame photometric detector; TSM: thickness-shear-mode; DPV: different pulse voltammetry; ITO: indium tin oxide electrode; CPE: carbon paste electrode; PEDOT: poly(3,4-ethylenedioxythiophene). c Template, THO: theophylline; CAF: caffeine; PT: parathion; NIC: nicotine; 2,4-D: 2,4-dichlorophenoxyacetic acid; Fru-val: fructosyl valine; m-val: methylvaline; m--lys: methyllysine (Z); D4NP: diethyl(4-nitrobenzyl).
the oxidation current responses of the added NIC, the mixture with equimolar coexistence of NIC and COT, and COT alone, respectively. Curves (a) and (b) reveal that the ITO/TiO2 [NIC]/PEDOT electrode exhibited high current responses particularly for NIC oxidation. Based on the oxidation current responses, it can be noted that the ITO/TiO2 [NIC]/PEDOT film shows higher selectivity for NIC, where NIC is adsorbed onto specific sites at the applied potential of 0.88 V. It should be mentioned here that COT is not electroactive at 0.88 V, and hence would not contribute to the oxidation current response under the equimolar coexistence of NIC and COT. Nevertheless, the formation of hydrogen bonds between COT molecules and hydroxyl groups on the TiO2 surface is considered to be stronger, as compared to that of NIC ones. That is to say, the adsorption capability of COT on the imprinted electrode surface is better than that of NIC. Therefore, we believed that even if COT shows no electrochemical behavior at 0.88 V, COT could be considered as a good interferent, especially under the coexistence of NIC and COT.
The above investigations revealed that NIC molecules occupy specific sites on the TiO2 nanostructure during coating and leave imprinted shapes during the thermal treatment (extraction process). These imprinted sites are only selective for NIC molecules, and the adsorption of COT by these specific sites was almost negligible. Hence, the ITO/TiO2 [NIC]/PEDOT electrode can distinguish NIC from COT in terms of the sensing performance. As shown in curve (b) of Fig. 4, it appears that the oxidation current response of the equimolar coexistence of NIC and COT was somewhat lower than with NIC alone. This may have been due to the fact that COT may slightly block a few imprinting sites on the ITO/TiO2 [NIC]/PEDOT electrode. On the basis of the above results, it was concluded that the imprinted sensor, ITO/TiO2 [NIC]/PEDOT, can be used in an environment in which NIC and COT coexist.
Fig. 4. Calibration curves for (a) nicotine (NIC) alone; (b) equimolar coexistence of an NIC and (−)-cotinine (COT) mixture; and (c) COT alone.
Fig. 5. The stability data for the ITO/TiO2 [NIC]/PEDOT electrode. The electrode was tested in a 5 mM NIC dissolved in a PBS solution for a period of 3 days.
C.-T. Wu et al. / Analytica Chimica Acta 633 (2009) 119–126
Fig. 6. The current responses and their relative current responses for NIC oxidation at 5 mM in PBS and swine PPP at the ITO/TiO2 /PEDOT and ITO/TiO2 [NIC]/PEDOT electrodes.
3.2.4. At-rest stability Since tailor-made biosensors based on molecularly imprinted TiO2 -modified electrode reported in this study could be an alternative to enzyme biosensors, the at-rest stability of the NIC-imprinted biosensor was also obtained. Fig. 5 shows the stability data for the ITO/TiO2 [NIC]/PEDOT electrode tested in a 5 mM NIC dissolved in a PBS solution for a period of 3 days. The current response for 5 mM NIC remained about 85% of its initial value at the end of 2 days. That is to say that the present NIC-imprinted TiO2 -modified electrode not only be a disposable sensor, but can act as a reusable sensor during 2 days used and shows a reasonable response. The sudden decay in response after 2 days probably has to do with the preservation of the electrode at ambient condition. 3.2.5. Sensing results for swine plasma Fig. 6 shows the current responses and their relative current responses (against the molecularly imprinted electrode for detecting NIC in PBS) for NIC oxidation at 5 mM in PBS and swine PPP. The pH value of swine PPP is about 8.1, because tri-sodium citrate was
125
added as the anticoagulant. However, only 34% of the relative current response for NIC was observed in swine PPP. We also performed the standard addition of NIC at lower concentration levels from 0 to 0.75 mM. The relative current responses for the detection of NIC in swine PPP were still about one-third of those in the PBS. This is identical to the result obtained for the detection of 5 mM NIC. Therefore, it is clear that the matrix effect does exist when detecting NIC in swine PPP. It is assumed that the reduction capability of the citrate may reduce the oxidation current in detecting NIC. Moreover, we have simplified the complicated bio-sample. Instead of using swine whole blood, PPP was obtained by centrifugation. Swine PPP still contains platelets, proteins, carbohydrates, electrolytes, fat, vitamins, hormone and metabolites. These chemical compounds are known to influence the obtained results. This means that we need to modify the sensor when dealing with a realistic biological fluid in future study. Although the modified electrode showed 34% relative current response of NIC when dealing with the complicated swine blood plasma, the SE still remained about 1.23, which was consistent with aforementioned result acquired from the PBS system. Therefore, we believed that the imprinting concept does work in various sensing environments investigated in this study. 3.3. Topographical mapping by SECM SECM can be used to elucidate the surface morphology on both active and inactive substrates by means of the increase or decrease in redox current responses. In our study, the surface morphologies of imprinted and non-imprinted electrodes could be distinguished using SECM. Before carrying out topographical mapping, the CV was recorded to obtain the applied potential between the tip and substrate. Here, the applied potential between the tip and substrate was set to 0 and 0.5 V, respectively, to characterize the Fe(CN)6 3− and Fe(CN)6 4− redox coupled reaction. Although the ITO acted as the active substrate for the sensing electrode, negative feedback was seen in the approach curve. It was found that the TiO2 on the substrate acted as an insulating layer causing negative feedback, which is consistent with our aforementioned data. Next, the tip was withdrawn by a distance of 10 m from the substrate surface. This avoided the tip being passivated as a result of direct close contact with the roughened TiO2 surface during topographical mapping.
Fig. 7. 3-D topographical mapping on the (a) ITO/TiO2 /PEDOT and (b) ITO/TiO2 [NIC]/PEDOT electrodes. The applied potential at the tip and substrates were 0 and 0.5 V (vs. Ag/AgCl/saturated KCl), respectively.
126
C.-T. Wu et al. / Analytica Chimica Acta 633 (2009) 119–126
Fig. 7a and b show the topographical mapping of the tip based on the redox reaction that occurred within the 10-m vertical region on both electrodes. The normalized current (iT /iT,∞ ) is indicated on the Z-axis. An irregular pattern was found for the TiO2 /PEDOT electrode without imprinted NIC (Fig. 7a), whereas it was very smooth on the NIC-imprinted surface (Fig. 7b). In addition, the average current response of the NIC-imprinted electrode was higher than that of the non-imprinted one, which was correlated with a higher amount of PEDOT formed in the former system. Moreover, the pore size of the imprinted electrode was also larger than that of the nonimprinted one, which was confirmed by the SEM images. The results of the SEM and SECM images infer that NIC was imprinted on underlying sites that were naked to the ITO electrode. These imprinted sites are consistent with larger pore sizes of the imprinted electrode observed under SEM, which thus resulted in a larger charge capacity for PEDOT during electropolymerization, compared to that using the non-imprinted one with smaller pore sizes. Therefore, a moreuniform and higher average current response was observed for the imprinted ITO/TiO2 [NIC]/PEDOT electrode (Fig. 7b) in contrast to a less-uniform and lower average current response obtained for the non-imprinted ITO/[TiO2 ]/PEDOT electrode (Fig. 7a). 3.4. FTIR studies In order to characterize the interaction between NIC and the TiO2 matrix, an FTIR spectrum was taken for the imprinted electrode bathed in 2.5 mM NIC. The spectrum revealed an intensity peak contributed by hydrogen bond formation between nitrogen atoms of NIC molecules and hydroxyl groups on the TiO2 surface [40]. The result is also consistent with the literature [41]. 4. Conclusions In conclusion, a novel NIC electrochemical sensor by using the concept of NIC-imprinted TiO2 has been investigated in this study. The imprinted (ITO/TiO2 [NIC]/PEDOT) electrode showed a higher response current than the non-imprinted one (ITO/TiO2 / PEDOT). The sensitivity of the imprinted (31.4 A mM−1 cm−2 ) film was about 1.24 times higher than the non-imprinted (25.2 A mM−1 cm−2 ) film, i.e., the SE was about 1.24. The LOD was calculated to be 4.9 M. The linear detection region was from 0 to 5 mM, and this offers a possible method for NIC sensing. Moreover, the oxidation current response for 5 mM NIC still remained about 85% of its initial value at 48 h, indicating reasonable stability of the electrode. According to this study, the NIC-imprinted TiO2 (ITO/TiO2 [NIC]/PEDOT) electrode showed reasonably good selectivity in distinguishing NIC from COT. Besides, according to our preliminary test with swine PPP, the imprinted electrodes could detect NIC in a complicated blood plasma system. It is heartening to notice that the SE in PPP still maintained at 1.23, which is very close to the value obtained in PBS. The imprinted and non-imprinted electrodes were characterized by SECM and SEM. SECM mapping provides a novel usage to recognize the template-imprinted cavities on the surface of the modified electrode. The FTIR spectra confirmed the formation of hydrogen bonds between NIC and TiO2 on the imprinted electrode. Although SE values for EC MIP sensors reported in the literatures are comparable with the SE obtained in this work, more work is now being carried out in our laboratory to improve the SE to 2–3. Future investigations will focus on miniaturing the sensing device on a chip and enhancing the SE of the presented NIC sensor.
Acknowledgements This work was financially supported by the National Research Council of Taiwan under grant numbers NSC95-2221-E-002-307 and NSC96-2220-E-006-015. The electrochemical instrumentation used in this work was supported by the Ministry of Education (MOE)’s Program for Promoting Academic Excellence of University under grant number EX-91-E-FA09-5-4. Appendix A. Supplementary data Both electrochemical and analytical data, including LSV, I–t curves for the NIC oxidation, SEM images, and FTIR spectrum, can be found in the supporting information at doi:10.1016/j.aca. 2008.11.038. References [1] J.S. Meyer, L.F. Quenzer, Psychopharmacology: Drugs, the Brain and Behavior, first ed., Sinauer Associates, Sunderland, 2005. [2] A.L. Wilson, L.K. Langley, J. Monley, T. Bauer, S. Rottunda, E. McFalls, C. Kovera, J.R. McCaretn, Pharmacol. Biochem. Behav. 51 (1995) 509. [3] A.H. Rezvani, E.D. Levin, Biol. Psychiatry 49 (2001) 258. [4] T.T. Denton, X.D. Zhang, J.R. Cashman, J. Med. Chem. 48 (2005) 224. [5] J. Cluette-Brown, J. Mulligan, K. Doyle, S. Hagan, T. Osmolski, J. Hojnacki, Proc. Soc. Exp. Biol. Med. 182 (1986) 409. [6] M. Page-Sharp, T.W. Hale, L.P. Hackett, J.H. Kristensen, K.F. Ilett, J. Chromatogr. Biomed. Sci. Appl. 796 (2003) 173. [7] G.N. Mahoney, W. Al-Delaimy, J. Chromatogr., Biomed. Appl. 753 (2001) 179. [8] H.S. Shin, J.G. Kim, Y.J. Shin, S.H. Jee, J. Chromatogr. Biomed. Sci. Appl. 769 (2002) 177. [9] A. Zander, P. Findlay, T. Renner, B. Sellergren, Anal. Chem. 70 (1998) 3304. [10] W.M. Mullett, E.P.C. Lai, B. Sellergren, Anal. Commun. 36 (1999) 217. [11] H.S. Andersson, J.G. Karlsson, S.A. Piletsky, A.-C. Koch-Schmidt, K. Mosbach, I.A. Nicholls, J. Chromatogr. A 848 (1999) 39. [12] Y. Tan, J. Yin, C. Liang, H. Peng, L. Nie, S. Yao, Bioelectrochemistry 53 (2001) 141. [13] T. Abdallah, S. Abdalla, S. Negm, H. Talaat, Sens. Actuators. A Phys. 102 (2003) 234. [14] L. Campanella, G. Favero, M. Tomassetti, Anal. Lett. 34 (2001) 855. [15] Y. Yang, M. Yang, H. Wang, L. Tang, Anal. Chim. Acta 509 (2004) 151. [16] K. Mosbach, Trends Biochem. Sci. 19 (1994) 9. [17] R.J. Ansell, D. Kriz, K. Mosbach, Curr. Opin. Biotechnol. 7 (1996) 89. [18] A.G. Mayes, K. Mosbach, Anal. Chem. 68 (1996) 3769. [19] K. Haupt, K. Mosbach, Chem. Rev. 100 (2000) 2495. [20] H.H. Maurer, J. Chromatogr., Biomed. Appl. 713 (1998) 3. [21] M. Nakajima, T. Yamamoto, Y. Kuroiwaand, T. Yokoi, J. Chromatogr., Biomed. Appl. 742 (2000) 211. [22] I.N. Papadoyannis, V.F. Samanidou, P.G. Stefanidou, J. Liq. Chromatogr. Relat. Technol. 25 (2002) 2315. [23] F. Zhou, P.R. Unwin, A.J. Bard, J. Phys. Chem. 96 (1992) 4917. [24] M. Tsionsky, A.J. Bard, M.V. Mirkin, J. Phys. Chem. 100 (1996) 17881. [25] M.V. Mirkin, B.R. Horrocks, Anal. Chim. Acta 406 (2000) 119. [26] Ch.J. Barbe, F. Arendse, P. Comte, M. Jirousek, F. Lenzmann, V. Shklover, M. Gratzel, J. Am. Ceram. Soc. 80 (1997) 3157. [27] M.K. Nazeeruddin, R. Humphry-Baker, P. Liska, M. Gratzel, J. Phys. Chem. B 107 (2003) 8981. [28] C.Y. Huang, Y.C. Hsu, J.G. Chen, V. Suryanarayanan, K.M. Lee, K.C. Ho, Sol. Energy Mater. Sol. Cells 90 (2006) 2391. [29] W.M. Yeh, K.C. Ho, Anal. Chim. Acta 542 (2005) 76. [30] J.M. Hong, P.E. Anderson, J. Qian, C.R. Martin, Chem. Mater. 10 (1998) 1029. [31] J. Wang, P.A.G. Cormack, D.C. Sherrington, E. Khoshdel, Angew. Chem. Int. Ed. 42 (2003) 5336. [32] S. Marx, A. Zaltsman, I. Turyan, D. Mandler, Anal. Chem. 76 (2004) 120. [33] T. Kobayashi, Y. Murawaki, P.S. Reddy, M. Abe, N. Fujii, Anal. Chim. Acta 435 (2001) 141. [34] Y. Yoshimi, R. Ohdaira, C. Iiyama, K. Sakai, Sens. Actuators B Chem. 73 (2001) 49. [35] R. Shoji, T. Takeuchi, I. Kubo, Anal. Chem. 75 (2003) 4882. [36] S. Kröger, A.P.F. Turner, K. Mosbach, K. Haupt, Anal. Chem. 71 (1999) 3698. [37] K. Sode, S. Ohta, Y. Yanai, T. Yamazaki, Biosens. Bioelectron. 18 (2003) 1485. [38] Z. Meng, T. Yamazaki, K. Sode, Biotechnol. Lett. 25 (2003) 1075. [39] H.H. Weetall, D.W. Hatchett, K.R. Rogers, Electroanalysis 17 (2005) 1789. [40] H. Park, W. Choi, J. Phys. Chem. B 108 (2004) 4086. [41] V. Arnaud, M. Berthelot, J.-Y. Le Questel, J. Phys. Chem. A 109 (2005) 3767.