Carbon nanotube-Web modified electrodes for ultrasensitive detection of organophosphate pesticides

Carbon nanotube-Web modified electrodes for ultrasensitive detection of organophosphate pesticides

Electrochimica Acta 101 (2013) 209–215 Contents lists available at SciVerse ScienceDirect Electrochimica Acta journal homepage: www.elsevier.com/loc...

2MB Sizes 0 Downloads 28 Views

Electrochimica Acta 101 (2013) 209–215

Contents lists available at SciVerse ScienceDirect

Electrochimica Acta journal homepage: www.elsevier.com/locate/electacta

Carbon nanotube-Web modified electrodes for ultrasensitive detection of organophosphate pesticides Mustafa Musameh a,∗ , Marta Redrado Notivoli a , Mark Hickey a , Chi P. Huynh a , Stephen C. Hawkins a , Jumana M. Yousef b , Ilias Louis Kyratzis a a b

Materials Science and Engineering, Commonwealth Scientific and Industrial Research Organisation (CSIRO), Clayton, VIC 3168, Australia Monash Institute of Pharmaceutical Sciences, Monash University, 381 Royal Parade, Parkville, VIC 3052, Australia

a r t i c l e

i n f o

Article history: Received 8 July 2012 Received in revised form 7 November 2012 Accepted 7 November 2012 Available online 19 November 2012 Keywords: Carbon nanotube-Web Organophosphates Methyl parathion Ultrasensitive Adsorption Reproducibility

a b s t r a c t A novel carbon nanotube-Web (CNT-Web) modified glassy carbon (GC) electrode is prepared and utilized for ultrasensitive electrochemical detection of methyl parathion (MP) organophosphate pesticide (OP) in aqueous solutions. The electrode was prepared by placing and securing 30 CNT-Web layers onto the GC electrode surface drawn with the aid of an electrical winding device. CNT webs comprise bundles of long (∼450 ␮m) multiwalled CNTs with high degree of alignment, porosity and rigidity. These CNT webs are highly electrocatalytic compared to the bare GC electrode. Due to their hydrophobic nature, electrical conductivity and absence of binders they show high affinity toward the adsorption and electrochemical detection of MP by cyclic and differential pulse voltammetries at trace levels compared to other conventional electrodes. Regeneration of the CNT web surface was possible by chemical methods due to the high stability and strong adhesion of the CNT to the GC electrode surface which did not alter the reproducible measurement of MP (RSD 4.7%, n = 10). Operational parameters, including the amount of CNT, incubation time, initial scanning potential and pH of the incubation medium have been optimized. The CNT-Web modified electrode yields well-defined, undistorted and interference free voltammetric response with good linearity in the range of 20–1000 nM (R2 = 0.993), 1–10 ␮M (R2 = 0.993) and 10–50 ␮M (R2 = 0.991) reflecting its high surface area. A detection limit of 1 nM is estimated based on a signal-to-noise ratio of 3 after incubation for 2 min along with an average sensitivity of 1.44 ␮A/␮M based on three calibration curves. The detection limit was further improved to 1 pM by using 10 min adsorption time. This new electrode nano-material will open the doors for the design of miniaturized and highly rigid sensors suitable for operation in harsh environments. Crown Copyright © 2012 Published by Elsevier Ltd. All rights reserved.

1. Introduction Due to the high toxicity of organophosphate pesticides (OPs) even at low levels, there is a need for sensitive and rapid quantification of these compounds particularly in aqueous environments. Traditional methods such as GC, HPLC, mass spectroscopy, and biological methods such as immunoassay can be used for the analysis of OPs [1,2]. However such analyses are generally performed in the lab rather than in the field, require expensive analytical instruments, involve several sample preparation steps and are time consuming. Field deployable kits for OPs detection are commercially available and offer several advantages, including portability, rapid analysis times, and low cost [3]. However, these kits do not meet the current needs for sensitive detection of OPs

∗ Corresponding author. Tel.: +61 3 9545 8149. E-mail addresses: [email protected], [email protected] (M. Musameh).

(measures only ppm levels), still need several sample preparation steps, usually are not very accurate, and are prone to interferences. Biosensors for the determination of OPs based on inhibition and noninhibition assays using the enzymes AChE [4,5] or OP hydrolase [6], respectively, have been fabricated. Despite their novelty, these systems suffer from instability, need for enzyme regeneration, tedious handling protocols, slow enzyme reaction and low sensitivity. Therefore there is a need for a more robust, stable, easy to handle and operate, rapid and sensitive sensors that would address the above issues. Due to their high electrical conductivity and electrocatalytic activity, unique surface chemistry including its ␲-conjugative structure with a highly hydrophobic surface and high porosity, carbon nanotube (CNT) based sensors were employed for the adsorption and electrochemical detection of OPs with high efficiencies compared to other conventional sensors. CNTs based sensors have been constructed from pure CNT [7], CNT/polymer composites [8], CNT/Ionic liquid (IL) composites [9], CNT/Metal composites [10] and CNT/mediator films [11].

0013-4686/$ – see front matter. Crown Copyright © 2012 Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.electacta.2012.11.030

210

M. Musameh et al. / Electrochimica Acta 101 (2013) 209–215

Zhang et al. [12] reported a highly linear response in the concentration range of 500 nM–60 ␮M of methyl parathion (MP) with a detection limit of 100 nM using CNT/gold composite electrodes. Another composite electrode made of CNT and chitosan [13] showed enhanced electrochemical response after 5 min accumulation of MP where the response was highly linear over the range of 0.05–2.0 ␮g/mL, with a detection limit of 5 ng/mL. Composite electrodes made of CNTs and ILs [9] also showed a highly linear response over the range of 2 nM to 4 ␮M MP after 3 min accumulation with a detection limit of 1 nM. Most of the reported CNT based sensors used for OPs detection require the use of binders to improve the stability and/or the conductivity of the film. Despite its usefulness, the presence of these binders can alter some of the properties of the CNTs such as their electrical conductivity and surface adsorption properties which would affect their analytical performance. Therefore there is a need for a more stable form of CNTs based sensors that would omit the need for binders and at the same time would have improved analytical performance for OPs adsorption and detection. CNT webs are a novel form of CNTs produced by drawing CNTs away from the front face of specially grown ‘forests’ of aligned CNTs [14,15]. CNTs are drawn as a continuous pure CNTs ‘web’ of around 20 ␮m in thickness with high porosity (optical transmission ∼80%). The web can adhere to a solid surface and get densified to about 50 nm in thickness by wetting in a solvent and drying. CNT webs mainly comprise highly pure and wellaligned multiwalled CNTs (MWCNTs), with some occasional cross fibers (mean diameter ∼ 10 nm) and are electrically conductive [16]. In this paper we describe the design of a novel and binder free CNT web modified GC electrode for the detection of MP OP with high sensitivity, stability, selectivity, and linearity. In addition, the CNT web modified electrode combines both the ease of fabrication, low background current, and the improved electrochemical response compared to other conventional electrodes with and without binders. 2. Experimental 2.1. Chemicals and reagents Methyl parathion, Nafion (5 wt%), graphite powder (<20 ␮m) and mineral oil were purchased from Sigma–Aldrich. Stock solutions of methyl parathion were dissolved in acetonitrile (ACN) and diluted with 0.1 M phosphate buffer (pH 7.0). Alumina ceramic plates were purchased from CoorTek Inc. (Grand Junction, CO 81505, USA). Multiwalled carbon nanotubes (MWCNT), 95 wt%, 5–15 nm diameter, 1–10 ␮m length were obtained from BuckyUSA (Houston, TX 77074, USA) and were used as supplied without any further purification. Carbon tape was obtained from SPI supplies (West Chester, PA 19380, USA). Other reagents were commercially available and were of analytical reagent grade. All solutions were prepared from double-distilled water. 2.2. Electrode preparation Drawable CNT forest synthesis is based on an optimized process described in detail elsewhere [14]. The CNT growth was performed in a 3 zone furnace quartz tube (95 mm reactor). The reaction was carried out at atmospheric pressure using flows of C2 H2 , He and H2 gases. The process is modified, with an acetylene and H2 concentration of only 2.5% (He/H2 /C2 H2 = 4000/100/100 sccm) and running time of 20 min. Analysis by TGA, EDX, and TEM of the CNT forest shows very low amorphous carbon content and no trace of the iron catalyst (data not shown).

A prototype CNT web electrode (Figs. 1 and 2) was constructed using drawn CNT webs laid down onto alumina strips (33.8 mm × 10.0 mm × 0.635 mm) or glass slips (8.5 mm × 26 mm × 0.16 mm) and solvent densified/bound to the surface with acetone to enable good adhesion with the support. Sets of six matched samples were prepared simultaneously using the winding device shown in Fig. 1 with the amount of CNT being controlled by the width of the web, the degree of insulation, and the number of turns applied. Typically, about 8.0 mm wide and 30 turns are used. In order to modify GC electrodes (3.0 mm diameter) (CH Instruments, Austin, TX, USA) with the prepared CNT webs, after placement of 30 CNT web layers on a glass slip or alumina plate, they were removed from the support after densification and immediately placed on the GC surface. A few drops of acetone were added for further densification. Insulation was made using parafilm or other polymeric insulators. Other electrodes were prepared for comparison purposes as summarized below: a. 30 CNT-Web layers/GC – as described above. b. MWCNT/Nafion/GC – 20 ␮l of 10 mg/ml MWCNT suspended in 1% Nafion in ethanol was cast on the GC surface and allowed to dry at room temperature for 2 h. c. Graphite/GC – 20 ␮l of 1 mg/ml graphite suspended in DMF was cast on the GC surface and allowed to dry at room temperature for 2 h. d. Carbon paste – 30 mg MWCNT was mixed with 70 mg mineral oil and a portion of the resulting paste was then packed firmly into the electrode cavity (2-mm-diameter, 2-mm-depth) of a glass sleeve. Electrical contact was established via a copper wire. e. Carbon tape/GC – a small piece of the carbon tape was placed on the GC electrode surface and the rest of the tape resting on the electrode body was insulated with parafilm. 2.3. Apparatus Differential pulse voltammetry (DPV) and cyclic voltammetry (CV) measurements were performed using a computer controlled Autolab PGSTAT302N potentiostat (Ecochemie, NL) with a threeelectrode configuration. The 30 CNT-Web layers modified GC electrode was used as the working electrode, the reference electrode (Ag/AgCl electrode, Model RE-5B, MF-2079 Bioanalytical Systems Inc. (BASi), West Lafayette, IN 47906, USA), and the counter electrode (a bright platinum wire) were inserted into the 50-mL glass cell (Amber glass, homemade) through holes in its Teflon cover. Scanning electron microscope (SEM) images were obtained using a Philips XL30 unit at an applied electron accelerating voltage of 5 kV. Optical images were obtained using Kombestereo optical microscope. 2.4. Procedure A 30 CNT-Web layers/GC electrode was dipped into a stirring 0.1 M phosphate buffer (pH 7.0) solution containing the desired concentration of MP pesticides for 2 min, washed with distilled water carefully, and transferred to a 50-mL electrochemical cell containing 0.1 M phosphate buffer (pH 7.0) solution. DPV measurements were performed from −0.8 to +0.3 V with a step potential of 5 mV, an amplitude of 25 mV, an interval time of 0.25 s and a scan rate of 20 mV/s. Baseline correction of the resulting voltammogram was performed by background subtraction. The cyclic voltammogram was recorded between −1.0 and +1.0 V at a scan rate of 100 mV/s. Both DPV and CV measurements were performed under quiescent conditions and at room temperature.

M. Musameh et al. / Electrochimica Acta 101 (2013) 209–215

211

Fig. 1. Photograph of the unit used to prepare the CNT-Web electrodes using alumina substrates (left) and a schematic describing the process of modifying the GC electrode with CNT-Web layers (right). (a) A spinnable CNT forest is used, (b) a proper support (e.g. alumina plates) is used and a winding device is used to continuously draw the CNT-Web and place it on the support, (c) the support with the required CNT-Web layers is removed, (d) the CNT-Web is peeled off the support, and (e) placed on the GC electrode surface, densified with acetone and secured with parafilm.

Cleaning and regeneration of the electrode surface from any adsorbed pesticide were performed by incubating the electrode for 3 min in ACN followed by washing thoroughly with double distilled water. A clean DPV baseline was ensured before each new incubation experiment with MP. 3. Results and discussion CNT-Webs have high degree of flexibility that enables the design of electrodes with different geometries, such as planar, yarn, ring, and ribbon [16]. The simplest configuration of these is the planar or disc electrode which is shown in Fig. 2. This configuration is suitable for the mass production of electrodes using the system shown in Fig. 1 and will be used throughout this study. This is followed by peeling the CNT webs off the substrate, placing them on the GC electrode surface, and securing and insulating with parafilm. To maintain better compactness between the CNT web layers and to have good adhesion to the GC electrode surface, the CNT web layers were densified by using a few drops of a volatile organic solvent such as acetone. Upon densification, the thickness of the CNT web layers decreased to one-third of the original thickness as was reported earlier [15]. The amount of CNT can be controlled by the number of CNT-Web layers determined by the number of winding device turns (Fig. 1) or by the electrode electroactive surface area controlled by the insulation process afterwards. It is apparent from

Figs. 1 and 2 that CNT webs maintain a high degree of alignment with occasional cross CNT fibers in between. CNTs shown in the SEM image of Fig. 2b are grouped together in the form of larger bundles and each single CNT within the webs is made of an average of seven inner walls (multiwall); about 450 ␮m long and around 10 nm in diameter. The optimal number of CNT web layers that showed the highest electrocatalytic activity and highest signal to background ratio was 30 which was used for all subsequent work. Fig. 3 compares cyclic voltammograms at the 30 CNT-Web layers/GC electrode (a) and the bare GC electrode (b) using 10 mM K4 Fe(CN)6 as the redox probe. The 30 CNT-Web layers/GC electrode showed enhanced electrochemical activity toward the oxidation and reduction of K4 Fe(CN)6 compared to the poor response at the bare GC electrode. This is clearly seen in the well defined shape of the redox peaks at the 30 CNT-Web layers/GC surface (peak-to-peak separation of 300 mV) with more than 10 times enhancement in oxidation and reduction peak currents compared to the bare GC electrode (peak-to-peak separation of 650 mV). Due to their extended length (∼450 ␮m), large surface area (200 m2 /g), and defect free surface (free from graphitic and metallic impurities), CNT webs are characterized by highly hydrophobic surfaces. This high hydrophobicity besides their electrical conductivity makes them suitable for the adsorption and electrochemical detection of toxic hydrophobic compounds such as OPs. Using a model OP

Fig. 2. An optical microscope image of a single CNT-Web layer on alumina substrate (a) and SEM image of 30 CNT-Web layers (b).

212

M. Musameh et al. / Electrochimica Acta 101 (2013) 209–215

Fig. 3. Cyclic voltammograms in 10 mM potassium ferricyanide using 30 CNT-Web layers/GC (a) and bare GC (b) electrodes. Potential scan rate 100 mV/s. Electrolyte 0.1 M KCl.

compound such as MP, the general protocol of detection involves first the adsorption of MP on the 30 CNT-Web layers/GC surface by incubating the electrode for a short period of time (∼2 min) in the MP solution. This is followed by reducing the nitrophenyl group of MP to hydroxylamine via a four-electron reduction process by applying a low negative potential or using a cathodic sweeping scan, and then the hydroxylamine is oxidized to the nitroso compound during the anodic scan which is used to quantify for the target OP [17]. The CV in Fig. 4a shows two well-defined redox peaks at −0.05 V and −0.18 V for the oxidation and reverse reduction of the nitroso group on MP at the 30 CNT-Web layers/GC electrode. No redox activity was observed in absence of MP as shown in Fig. 4b. The inset of Fig. 4 shows the corresponding DPVs of the 30 CNT-Web layers/GC electrode in presence (a) and absence (b) of MP. Similar to the oxidation peak observed in CV of Fig. 4, the differential pulse stripping peak is sharp and well defined which represents the oxidation of the hydroxylamine group of MP that occurred at −0.12 V. No oxidation peak was observed in absence of MP. To compare the performance of the 30 CNT-Web layers/GC electrode with other carbon based electrodes with and without binders on the adsorption and detection of MP, the responses at these electrodes were recorded as shown in Fig. 5. It is very clear that no or negligible response can be seen on these electrodes compared to

Fig. 4. Cyclic voltammograms of MP/30 CNT-Web layers/GC (a) and 30 CNT-Web layers/GC (b) in 0.1 M phosphate buffer solution (pH 7.0). Potential scanning rate, 100 mV/s. MP was adsorbed by dipping the electrodes in stirring 0.1 M of phosphate buffer solution (pH 7.0) containing 30.0 ␮M MP for 2 min and carefully washing with distilled water before electrochemical measurement. Bottom inset is corresponding differential pulse stripping voltammograms. Top inset shows the chemical structure of methyl parathion. DPV conditions: Scanning potential range, −0.8 to +0.2 V; scan rate 20 mV/s.

Fig. 5. Differential pulse stripping voltammograms of 30 CNT-Web layers/GC (a), Graphite/GC (b), MWCNT paste (c), carbon tape/GC (d), and GC (e) electrode after 2 min adsorption in stirring 0.1 M phosphate buffer solution (pH 7.0) containing 1.0 ␮M MP. DPV conditions: scanning potential range, −0.8 to +0.2 V; scan rate 20 mV/s.

the 30 CNT-Web layers modified electrode reflecting the unique structure and geometry of the CNT webs surface that showed high sensitivity and specific adsorption capability of MP compared to other forms of carbon based electrodes. This is greatly attributed to the increased surface area of the CNT webs as compared for example to graphite (Fig. 5b) along with enhanced signal to background ratio and the absence of any binders that may alter the response as the case for the MWCNT paste electrode (Fig. 5c) and carbon tape/GC electrode (Fig. 5d). Furthermore, the bare GC electrode did not show any signs of MP adsorption as shown in Fig. 5e. An important requirement for performing repetitive measurements using a single electrode is to ensure the absence of any memory effects of analyte from previous measurement. For processes involving the adsorption of compounds onto the electrode surface, cleaning or regeneration of the electrode surface after each measurement is important to avoid false results and poor reproducibility. In electrochemical sensing protocols, this is usually achieved by electrochemical cleaning of the surface after each measurement to completely remove any remaining analyte species adsorbed on the electrode surface. As shown in Fig. 6a, electrochemical cleaning of the 30 CNT-Web layers/GC electrode surface by repetitive electrochemical stripping in fresh phosphate buffer solution resulted in reducing the response of bound MP. However, after the 10th electrochemical oxidation stripping cycle, 78% of the bound MP was successfully removed from the surface and 22% was still bound. This indicated the strong affinity of MP to the CNT web surface, which made it hard to remove all of the bound MP from the surface even with using a small concentration MP such as 100 nM. Another method for cleaning the surface is by chemically solubilizing MP using a proper organic solvent. For the practical utilization of this approach, it is important that the organic solvent used in the cleaning process does not disintegrate or affect the physical properties of the electrode surface material which otherwise would affect its analytical performance and reproducibility. Incubating the 30 CNT-Web layers modified GC electrode in ACN for 3 min after measurement of MP resulted in complete removal of MP from the surface as indicated from the second electrochemical stripping voltammogram shown in Fig. 6b. This clearly showed that ACN was very effective in solubilizing and removing all the bound MP from the surface in a relatively short period of time compared to the electrochemical cleaning process. In order to verify that chemical cleaning of the CNT web modified electrode with ACN will not affect the reproducibility of MP measurement, 10 repetitive measurements of 100 nM MP after cleaning with ACN was conducted as shown in Fig. 7. A stable signal is

M. Musameh et al. / Electrochimica Acta 101 (2013) 209–215

Fig. 6. Differential pulse stripping voltammograms of the electrochemical (a) and chemical (3 min in ACN) (b) regeneration processes of the 30 CNT-Web layers/GC electrode after 2 min adsorption in 100 nM MP. Other conditions, same as Fig. 5.

observed during the entire operation using the 30 CNT-Web layers/GC electrode with RSD of 4.7%. The electrode showed a stable response when tested over one month on daily basis and stored in dry conditions when not in use. This stable response showed that the CNT webs physical structure and adhesion to the GC electrode were not affected by the repetitive cleaning with ACN or continuous measurement. This is attributed to the rigid structure of CNT webs along their extended length, organization into larger bundles, high degree of compactness and strong adhesion to the GC electrode surface. On the other hand, films of disorganized and short MWCNTs powder mixed with Nafion polymer were not stable even after the first washing with ACN (data not shown).

Fig. 7. Differential pulse stripping voltammograms of 10 successive adsorptions of 100 nM MP for 2 min followed by chemical regeneration (3 min in ACN) of the 30 CNT-Web layers/GC electrode. Other conditions, same as Fig. 5.

213

Before using the 30 CNT-Web layers/GC electrode for analytical determination of different MP concentrations, it was essential to optimize a number of analytical variables to maximize the voltammetric response as shown in Fig. 8. The amount of MWCNT laid on the electrode surface determined by the number of CNT web layers affected the voltammetric response of MP due to the increase in surface area which increased the amount of MP adsorbed. At the same time the increase of CNT content also led to an increase in electrochemical activity and hence response at the CNT web modified electrode. As shown in Fig. 8a after 2 min adsorption of 1 ␮M MP, the response increased rapidly for the first 20 layers of CNT webs and increased slowly after that to level off at 30 CNT web layers. Going beyond 30 CNT web layers resulted in an increase in background current which affected the measured electrochemical signal. Based on that 30 CNT web layers were chosen for the subsequent work. The effect of incubation time on response was also studied as shown in Fig. 8b. Upon increasing the incubation time, more MP was adsorbed onto the CNT web surface with a rapid increase in response up to 1 min, followed by a slower change in response up to 2 min, and much slower thereafter. As a compromise between sensitivity and measurement time, an accumulation time of 2 min was used in the subsequent work. The initial scanning potential has a profound effect on the stripping voltammetric response resulting from the ability to reduce the nitrophenyl group of the MP to hydroxylamine that undergoes subsequent oxidation during the anodic scan. Fig. 8c shows that at potentials higher than or equal to −0.4 V it was not possible to reduce the nitrophenyl group in order to generate a measurable signal. Upon decreasing the potential below −0.4 V, the signal was increasing rapidly up to −0.6 V and leveled off after that. Finally the effect of pH of the incubation solution on the voltammetric response was also studied as shown in Fig. 8d. According to this figure, the highest voltammetric response was obtained at a pH of 7. At a pH higher than 8, the response was greatly reduced due to complete hydrolysis of MP in basic medium. At lower pH (<7) the response was slightly smaller compared to pH 7 but still significant and can also be attributed to partial and slow degradation of MP with time in acidic medium. A pH of 7 for the incubation solution was used for the subsequent work. After complete optimization of analysis parameters, the analytical performance of the 30 CNT-Web layers/GC electrode was investigated using a series of MP concentrations as shown in Fig. 9. The electrode had good linearity in the range 20–1000 nM (R2 = 0.993) (a), 1–10 ␮M (R2 = 0.993) (b) and 10–50 ␮M (R2 = 0.991) (c) reflecting its high surface area and capacity to adsorb MP. A detection limit of 1 nM was estimated based on a signal-to-noise ratio of 3 after incubation for 2 min along with an average sensitivity of 1.44 ␮A/␮M based on the three calibration curves. The detection limit was further improved to 1 pM by increasing the incubation time to 10 min (left inset, Fig. 9a). The obtained detection limits and linear ranges are comparable or even better than those reported earlier using CNT and other carbon-based electrodes for the detection OP compounds [9,13,17]. The effect of different interfering species and electrolytes on the response of MP was evaluated as shown in Fig. 10. Voltammetric responses were recorded after incubation with 1 ␮M of MP for 2 min or after mixing 1 ␮M of MP with 1 ␮M of p-nitrophenol, phenol, or nitrobenzene separately. The results have shown minimal interference from p-nitrophenol and nitrobenzene (<6%) while the presence of 1 ␮M phenol resulted in an increase in response by 12%. The exact reason for that is not fully understood. On the other hand, the effect of other ionic electrolyte species present in the incubation solution at a concentration of 0.1 M on MP signal was also evaluated. The presence of non-oxygenated ionic salts such as NaCl caused only a marginal decrease in MP response by <3%. However, the presence of both Na2 SO4 and CH3 COONa caused an

214

M. Musameh et al. / Electrochimica Acta 101 (2013) 209–215

Fig. 8. Effect of number of CNT-Web layers (a), incubation time (b), initial scanning potential (c) and pH of the incubation medium (d) on the adsorption of MP. The concentration of methyl parathion in incubation medium was 5.0 ␮M. The adsorption experiments were performed in a pH 7.0 of 0.1 M phosphate buffer solution. Other conditions, same as Fig. 5.

increase in response by 20%. It is apparent that the presence of these oxygenated electrolyte species at high concentration is enhancing the concentration of MP on the more hydrophobic CNT web surface. None of the tested interfering species was electrochemically

active in the measured potential range. Most phenolic compounds can be detected by electrochemical oxidation at potentials higher than 0.3 V [18] where they would not overlap with MP signal at −0.12 V.

Fig. 9. Differential pulse stripping voltammograms of increasing concentrations of MP in the range 20–1000 nM (a), 1–10 ␮M (b) and 10–50 ␮M (c) after 2 min adsorption on the 30 CNT-Web layers/GC electrode. Right inset of each shows the corresponding calibration curve. Left inset of (a) shows the DPV after 10 min adsorption of 100 pM methyl parathion. Other conditions, same as Fig. 5.

M. Musameh et al. / Electrochimica Acta 101 (2013) 209–215

215

Acknowledgment This work was supported by CSIRO Materials Science and Engineering. References

Fig. 10. Differential pulse stripping voltammograms of 1 ␮M MP after 2 min adsorption on the 30 CNT-Web layers/GC electrode in the absence and presence of 1 ␮M p-nitrophenol, 1 ␮M phenol, 1 ␮M nitrobenzene, 0.1 M NaCl, 0.1 M Na2 SO4 and 0.1 M CH3 COONa, respectively. Other conditions, same as Fig. 5.

4. Conclusions CNT web modified electrodes were prepared by using a simple setup and have shown to be useful for the adsorption and ultrasensitive detection of MP (a model OP compound) over a short period of time. Due to their unique physical and chemical structure, CNT web modified electrodes have shown very high adsorption capacity and affinity to MP in the absence and presence of different interfering species. In addition these modified electrodes were shown to maintain good stability in absence of any binders in aqueous and organic mediums which resulted in high reproducibility of electrochemical measurements. Due to their flexibility CNT webs can be suited for the design of different electrode geometries such as yarn and ribbon based electrodes that can serve as miniaturized and flow through detectors. Their high affinity for MP besides their electrochemical activity makes them suitable for the detection of other electroactive OP compounds and nonelectroactive OP compounds by incorporating selective biological recognition elements such as enzymes and antibodies. CNT webs are a novel and emerging form of nano-material with features that will help in designing sensors suitable for operation in complex and harsh matrices such as blood and sewage water, non-aqueous mediums, and high temperature and acidity environments.

[1] J.K. Miller, D.E. Lenz, Development of an immunoassay for diagnosis of exposure to toxic organophosphorus compounds, J. Appl. Toxicol 21 (2001) S23–S26. [2] J. Sherma, PESTICIDES, Anal. Chem 65 (1993) R40–R54. [3] The EnviroLogix Cholinesterase Screening Test (EP 014). EnviroLogix Inc., www.envirologix.com [4] D. Du, J.W. Ding, J. Cai, A.D. Zhang, Electrochemical thiocholine inhibition sensor based on biocatalytic growth of Au nanoparticles using chitosan as template; Sens, Actuators, B 127 (2007) 317–322. [5] J.M. Gong, L.Y. Wang, L.Z. Zhang, Electrochemical biosensing of methyl parathion pesticide based on acetylcholinesterase immobilized onto Aupolypyrrole interlaced network-like nanocomposite; Biosens, Bioelectron 24 (2009) 2285–2288. [6] A. Mulchandani, W. Chen, P. Mulchandani, J. Wang, K.R. Rogers, Biosensors for direct determination of organophosphate pesticides, Biosens. Bioelectron 16 (2001) 225–230. [7] P.L.A. Sundari, S.P. Palaniappan, P. Manisankar, Enhanced sensing of carbendazim, a fungicide on functionalized multiwalled carbon nanotube modified glassy carbon electrode and its determination in real samples, Anal. Lett 43 (2010) 1457–1470. [8] P. Manisankar, P.L.A. Sundari, R. Sasikumar, S.P. Palaniappan, Electroanalysis of some common pesticides using conducting polymer/multiwalled carbon nanotubes modified glassy carbon electrode, Talanta 76 (2008) 1022–1028. [9] S.S. Fan, F. Xiao, L.Q. Liu, F.Q. Zhao, B.Z. Zeng, Sensitive voltammetric response of methylparathion on single-walled carbon nanotube paste coated electrodes using ionic liquid as binder, Sens. Actuators, B 132 (2008) 34–39. [10] M. Siswana, K.I. Ozoemena, T. Nyokong, Electrocatalytic detection of amitrole on the multi-walled carbon nanotube - Iron (II) tetra-aminophthalocyanine platform, Sensors 8 (2008) 5096–5105. [11] J.Y. Li, Y.M. Chi, Determination of carbendazim with multiwalled carbon nanotubes-polymeric methyl red film modified electrode, Pestic. Biochem. Physiol 93 (2009) 101–104. [12] Y. Zhang, T.F. Kang, Y.W. Wan, S.Y. Chen, Gold nanoparticles-carbon nanotubes modified sensor for electrochemical determination of organophosphate pesticides, Microchim. Acta 165 (2009) 307–311. [13] D. Du, M.H. Wang, J.M. Zhang, H. Cai, H.Y. Tu, A.D. Zhang, Application of multiwalled carbon nanotubes for solid-phase extraction of organophosphate pesticide, Electrochem. Commun 10 (2008) 85–89. [14] C.P. Huynh, S.C. Hawkins, Understanding the synthesis of directly spinnable carbon nanotube forests, Carbon 48 (2010) 1105–1115. [15] M. Zhang, S.L. Fang, A.A. Zakhidov, S.B. Lee, A.E. Aliev, C.D. Williams, K.R. Atkinson, R.H. Baughman, Strong, transparent, multifunctional, carbon nanotube sheets, Science 309 (2005) 1215–1219. [16] M. Musameh, M.R. Notivoli, M. Hickey, I.L. Kyratzis, Y.A. Gao, C. Huynh, S.C. Hawkins, Carbon nanotube webs: a novel material for sensor applications, Advanced Materials 23 (2011) 906. [17] G.D. Liu, Y.H. Lin, Electrochemical stripping analysis of organophosphate pesticides and nerve agents, Electrochem. Commun 7 (2005) 339–343. [18] J. Wang, R.P. Deo, M. Musameh, Stable and sensitive electrochemical detection of phenolic compounds at carbon nanotube modified glassy carbon electrodes, Electroanalysis 15 (2003) 1830–1834.