Analytica Chimica Acta 575 (2006) 32–38

Fabrication of a new polyaniline grafted multi-wall carbon nanotube modified electrode and its application for electrochemical detection of hydrogen peroxide P. Santhosh a , K.M. Manesh a , A. Gopalan a,b , Kwang-Pill Lee a,∗ a

Advanced Analytical Science and Nanomaterials Lab, Department of Chemistry Education, Kyungpook National University, Daegu 702-701, South Korea b Department of Industrial Chemistry, Alagappa University, Karaikudi 630003, India Received 13 August 2005; received in revised form 14 October 2005; accepted 22 May 2006 Available online 2 June 2006

Abstract A modified electrode is fabricated by grafting polyaniline (PANI) chains onto multiwalled carbon nanotube (MWNT) for the detection of hydrogen peroxide (H2 O2 ). PANI grafted MWNT modified electrode (PANI-g-MWNT-ME) displays excellent electrocatalytic response to the detection of H2 O2 in a concentration range of (1.0–20) × 10−8 M showing linear response to current, with an extended lower detection limit down to 1 nM. This modified electrode exhibits an accelerated electron transfer at the interface with minimized surface fouling and surface renewability. Further, electrochemical analysis of H2 O2 performed in the presence of common interferents such as uric acid, ascorbic acid and acetaminophen with the modified electrode reveals that there is no overlapping signal from the interferents. The combined presence of MWNT and PANI in the modified electrode provides high sensitivity and selectivity. © 2006 Elsevier B.V. All rights reserved. Keywords: Multiwall carbon nanotube; Grafting; Modified electrode

1. Introduction Carbon nanotubes (CNT) represent an important group of nanomaterials with attractive electronic, chemical and mechanical properties [1,2]. The high surface area and hollow geometry, combined with electronic conductivity and mechanical properties make CNTs to promote electron transfer reactions when CNTs are fabricated as electrodes for electrochemical reactions [3]. CNT modified electrodes have been shown to have improved sensitivity and selectivity toward analytes, including cytochrome c [4], NADH [5], norepinephrine [6], nucleic acids [7], hydrazine compounds [8] or hydrogen peroxide [9]. Higher sensitivity and stability have been demonstrated for CNTs due to the interfacial accumulation of analytes onto their surfaces and minimization of surface fouling, respectively. Organic conducting polymers have emerged as promising materials in the development of compact and portable probes

∗

Corresponding author. Tel.: +82 53950590; fax: +82 539528104. E-mail addresses: algopal [email protected] (A. Gopalan), [email protected] (K.-P. Lee). 0003-2670/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2006.05.075

for the detection of biologically significant molecules [10,11]. However, conducting polymers exhibit surface fouling in electrochemical devices and have lesser sensitivity at submicron levels probably due to the large volumetric capacitance [12]. Composites of conducting polymer and CNT have shown properties of the individual components with a synergistic effect [13–15]. Variety of conducting polymers has been tried for making CNT/polymer composites toward various applications [16–18]. Amperometric electrodes fabricated by dispersing CNTs in conducting polymer matrixes have been reported [19–22]. However, in these studies, the CNT/conducting polymer modified electrodes have been fabricated as a bilayer matrix with conducting polymer layer over-coated on CNT surface. The top formed conducting polymer layer is expected to block the functional capabilities of CNT. Hence, the electrochemical activity of CNT could not be realized to the fullest extent. On the other hand, if a modified electrode is fabricated by grafting the conducting polymer to CNT in the electrode fabrication, such a modified electrode will have the potential effects of both CNT and conducting polymer. However, preparation of modified electrodes in which conducting polymer is present as connected unit to CNT has not been reported so far.

P. Santhosh et al. / Analytica Chimica Acta 575 (2006) 32–38

In the present study, we present a simple and effective fabrication of polyaniline (PANI) grafted multiwalled carbon nanotube (MWNT) modified electrode (PANI-g-MWNT-ME). PANI-gMWNT-ME is expected to posses the combined advantages of PANI and MWNT and can show better sensitivity and selectivity for electrochemical detection of analytes. The attractive features of PANI-g-MWNT-ME over PANI-ME towards the detection of hydrogen peroxide (H2 O2 ) are detailed here. 2. Experimental 2.1. Reagents and materials Aniline (Aldrich) was distilled and used. Multi-walled carbon nanotubes (MWNTs) (10–50 nm in diameter) obtained

33

from CNT Co. Ltd., Incheon, Korea, were rinsed with double-distilled water and dried. Poly(ethylene glycol) bis(3aminopropyl), hydrogen peroxide, acetaminophen, ascorbic acid and uric acid of analytical grades were used as received. Double-distilled water was used throughout the experiments. Aqueous solutions of H2 O2 were prepared in phosphate buffer (pH 7) afresh at the time of experiments. For electrochemical studies, indium tin oxide (ITO) coated glass plate (specific surface resistance of about 10
) was used as the working electrode. Before each of the electrochemical experiment, the ITO electrode was degreased with acetone and rinsed with distilled water. For each experiment, a new ITO coated glass plate was used. A constant effective surface of 1 cm2 was maintained by masking the other part of the working electrode through lacquering.

Scheme 1. (a) Functionalization of MWNT and (b) fabrication of PANI-g-MWNT-ME.

34

P. Santhosh et al. / Analytica Chimica Acta 575 (2006) 32–38

2.2. Modified electrode fabrication via grafting of PANI chains onto MWNTs Grafting of PANI onto MWNT was carried out by surface modifying the MWNT with amine functional groups and grafting the PANI chains through amine functional groups by electrochemical polymerization of aniline. Fabrication of PANI-g-MWNT-ME involved two steps (Scheme 1). In the first step, MWNT was amine functionalized (MWNT-NH2 ) using sequential procedures (Scheme 1a). In the second step, cyclic voltammetry was used to deposit PANI-g-MWNT as film on the surface of working electrode (ITO) to obtain PANI-g-MWNT-ME (Scheme 1b). Following procedure was adopted for the preparation of amine functionalized MWNTs. Fifty milligrams of MWNT was refluxed in 4 M HNO3 for 24 h and filtered through a 0.2 ␮m pore size polycarbonate membrane. The residue, MWNT-COOH (carboxylated MWNT) was washed with deionized water and dried under vacuum at 60 ◦ C for 12 h. Fifty milligrams of MWNTCOOH was refluxed in 100 mL of thionyl chloride at 65 ◦ C for 24 h to get MWNT-COCl. MWNT-COCl was filtered, washed with THF and dried under vacuum at room temperature. To prepare the amine functionalized MWNTs, MWNT-COCl was refluxed with poly(ethylene glycol) bis(3-aminopropyl) using THF at 60 ◦ C for 24 h. The amine functionalized MWNT (MWNT-NH2 ) were separated by filtration, dried under vacuum. Fifty milligrams of MWNT-NH2 was homogeneously dispersed in 40 mL of 0.5 M cetyltrimethyl ammonium bromide using ultrasonication (BRANSON Digital Sonifier) for 3 h. 10 mL of 50 mM of aniline (in 0.5 M H2 SO4 ) was added to the above solution. This makes the final concentration of MWNTNH2 , aniline, CTAB and sulphuric acid as 0.1% (w/v), 10 mM, 0.4 M and 0.1 M, respectively. 10 mL of this solution containing MWNT-NH2 (0.1%) and aniline (10 mM) in 0.1 M H2 SO4 and 0.4 M CTAB was subjected to electrochemical polymerization by continuous potential cycling in the potential range of 0–900 mV. Green coloured deposit (PANI-g-MWNT) was seen on the surface of the electrode. After polymerization, the modified electrode was washed with water and stored in phosphate buffer (pH 7) under nitrogen atmosphere.

state conditions in a phosphate buffer (pH 7.0) by applying a constant potential of −0.3 V to the working electrode. The amperometric experiment was performed in a standard singlecompartment electrochemical cell that contained modified electrode, an SCE reference electrode, and a platinum wire auxiliary electrode. The background response of the modified electrode was allowed to decay to a steady state with stirring. When the background current became stable (after 20 s), a solution of H2 O2 was injected into the electrolytic cell, and its response was measured. 3. Results and discussion 3.1. Electrochemistry of modified electrode formation Cyclic voltammetry was used for the fabrication of PANIg-MWNT-ME through the electro-polymerization of aniline in presence of MWNT-NH2 . Fig. 1 displays the cyclic voltammetric profiles recorded during the electrochemical polymerization of a solution containing mixture of MWNT-NH2 and aniline at a scan rate of 100 mV s−1 . CV profiles of polymer film deposited from the MWNT-NH2 and aniline mixture showed oxidation waves around 200, 510, 600 and 705 mV with cathodic counterparts around 650, 500, 420 and 100 mV (appearing as a weak shoulder), respectively. In the case of electrochemical polymerization of aniline (in the absence of MWNT-NH2 ) under a similar condition to the mixture, four identifiable anodic peaks around 220, 500, 590 and 820 mV with cathodic counter parts around 800, 550, 480 and 80 mV, respectively, could be noticed [23]. The peak around 200 mV corresponds to formation of radical

2.3. Instrumentation All the electrochemical measurements were carried using EG&G PAR 283 Electrochemical Analyzer. Hydrodynamic voltammetry was performed in the flow cell for a series of injections of H2 O2 containing solutions into the analyte. The applied potential was stepped from +0.1 to −0.6 V in +0.05 V increments for subsequent injection. The current–potential traces were recorded after 2 min stabilization time at each potential. In the case of flow analysis, current versus time for variable H2 O2 concentration injections were recorded for injection of a series of concentrated solutions of H2 O2 to the phosphate buffer solution (pH 7.0). The amperometric response of the H2 O2 on modified electrodes (PANI-g-MWNT and PANI) was recorded under steady-

Fig. 1. Cyclic voltammograms recorded during the electrochemical polymerization of the solution containing aniline (10 mM) and poly(ethyleneglycol) bis(3-aminopropyl) functionalized MWNT (0.1%, w/v), [CTAB]: 0.4 M and [H2 SO4 ]: 0.1 M (inset: CV of PANI-g-MWNT-ME (a) and PANI-ME (b) in the background electrolyte). Scan rate: 100 mV s−1 .

P. Santhosh et al. / Analytica Chimica Acta 575 (2006) 32–38

cations, the second peak at 510 mV corresponds to the formation of benzoquinone (as hydrolysis product), the peaks at 600 and 705 mV correspond to the oxidation of head to tail dimer and conversion of emeraldine to pernigraniline structure, respectively [23,24]. Few distinct differences could be noticed on comparing the CV profiles of electropolymerization of simple aniline and mixture of aniline and MWNT-NH2 . The initial (first) oxidation occurs at a lower potential (200 mV) for the mixture of MWNTNH2 and aniline in comparison to the case of simple aniline (220 mV). Also, the peak current at the first oxidation state was significantly higher for the mixture (MWNT-NH2 and aniline). These two facts favor the simultaneous oxidation of amine sites at MWNT-NH2 and aniline in the case of electropolymerization with a mixture of MWNT-NH2 and aniline. The cation radicals thus generated from MWNT-NH2 and aniline can interact to nucleate grafting of PANI chains onto MWNT. Hence, we anticipate the oxidation of amine sites in MWNT-NH2 along with amine groups in aniline. This could cause a cross-reaction between oxidized amine groups in MWNT-NH2 and aniline cation radical. As a result of this, the formation of PANI grafted chain over MWNT is expected. The redox characteristics representing the growth of PANI-g-MWNT on the working electrode were thus different in the case of electrochemical polymerization with simple PANI. Additionally, a gradual increase in peak current values was noticed with increase in number of cycles during electropolymerization of MWNT-NH2 and aniline. This confirms the continuous build up of film of PANI-g-MWNT on the surface of working electrode on increasing in number of cycles. It is also important to note that the conversion of emeraldine to pernigraniline form was more facile (705 mV) in PANI-g-MWNT than in simple PANI (820 mV).

3.2. Electrochemical behaviour of PANI-g-MWNT-ME Electrochemical behaviour of PANI-g-MWNT-ME was studied in 0.1 M H2 SO4 by cyclic voltammetry (inset in Fig. 1, curve a). Two prominent couples of redox peaks centered around 200

35

and 700 mV corresponding to leucoemeraldine/emeraldine and emeraldine/pernigraniline redox transitions, respectively, were observed for PANI-g-MWNT-ME in the background electrolyte. For comparison, CV profile of PANI-ME (inset in Fig. 1, curve b) was also recorded. The redox current of PANI-g-MWNT-ME is larger than that of PANI-ME that indicates PANI-g-MWNTME has larger effective surface area than PANI-ME. To obtain further evidence for the grafting of PANI chains onto MWNT in the ME, high resolution transmission electron microscopy, HRTEM images of MWNT-ME (Fig. 2a) and PANI-g-MWNTME (Fig. 2b) were recorded. HRTEM image (Fig. 2b) reveals the presence of PANI over the entire surface of MWNT with a thickness of about 30–50 nm. This is in contrast to the previous observations of PANI modified MWNT electrode. A partial masking of CNT with PANI has been previously reported [21]. 3.3. Electrochemical reduction of H2 O2 Hydrodynamic voltammograms were recorded for the reduction process at the PANI-g-MWNT-ME and compared with the similar process at PANI-ME (Fig. 3). Significant increase in peak currents at potentials of H2 O2 reduction was noticed with PANI-g-MWNT-ME in comparison at PANI-ME. The combined presence of MWNT and PANI in PANI-g-MWNTME provides a three-dimensional electron-conductive network, which extended throughout the CNT matrix and resulted in the improvement of its electronic and ionic transport capacity. This was ascertained by performing control experiments for H2 O2 reduction with MWNT-NH2 -ME and PANI-ME electrodes. The other two electrodes (MWNT-NH2 -ME and PANI-ME) did not enhance the peak current significantly in comparison to PANIg-MWNT-ME. Besides, at the PANI-g-MWNT-ME, lowering of H2 O2 overpotential (>200 mV) was noticed. This may be expected by the augmented conduction provided by MWNT. These changes were indicative of more efficient electron transfer at PANI-g-MWNT-ME than that at PANI-ME. An improved path for charge transport is envisaged for PANI-g-MWNT-ME that can favor in electron transfer and accompanying ion movement. The enhanced charge-transport capacity of the PANI-g-MWNTME causes efficient reduction for H2 O2 .

Fig. 2. HRTEM image of MWNT (a) and PANI-g-MWNT (b) (scale bar is 50 nm).

36

P. Santhosh et al. / Analytica Chimica Acta 575 (2006) 32–38

Fig. 3. Hydrodynamic voltammograms for 1 ␮M H2 O2 at PANI-g-MWNT-ME (a) and PANI-ME (b). (A) Chronoamperometric signals measured at −0.3 V of H2 O2 from 1 × 10−8 M (b) to 7 × 10−8 M (h) at PANI-g-MWNT-ME. Dashed line (a) indicates the blank solution. (B) Stability of the chronoamperometric response for repetitive measurements at −0.3 V of 1 ␮M H2 O2 at the PANI-g-MWNT-ME (a) and PANI-ME (b); electrolyte, phosphate buffer (pH 7.0).

3.4. Stability and sensitivity Static and dynamic measurements were made to determine the stability and sensitivity for the detection of H2 O2 at PANI-gMWNT-ME. Fig. 3B compares the stability of the response for 30 repetitive chronoamperometric measurements of 1 ␮M H2 O2 at PANI-g-MWNT-ME (a) and PANI-ME (b). PANI-ME showed a rapid loss of its activity (upto 68% decrease of the response and a RSD of 32%). In contrast, a highly stable signal was observed over the entire operation at the PANI-g-MWNT-ME (RSD 2%). Also, it is important to note that PANI-g-MWNT-ME allows to detect a lower concentration (1 × 10−8 M) of H2 O2 through amperometric measurements (Fig. 3A). However, PANI-ME could not go down to such a lower detection level. Fig. 4A displays the flow injection response for H2 O2 solutions of increasing concentrations ((1–6) × 10−8 M; a–f). Well-defined peaks with response currents proportional to the H2 O2 concentration were observed with a low noise level. The calibration plot (figure not shown) is perfectly linear (slope, 3 nA/10−8 M), with a correlation coefficient of 0.997. Fig. 4 compares the stability of the response for repetitive flow injection measurements of 6 × 10−8 M H2 O2 at PANI-g-MWNT-ME (B) and 1 ␮M H2 O2 at PANI-ME (C). PANI-ME showed a gradual decrease of the response (with a 70% decrease and a RSD of 35%; n = 15). In contrast, a stable signal was observed over the entire period of operation at the PANI-g-MWNT-ME (RSD 0.8%). The amperometric responses at the PANI-g-MWNT-ME for each successive addition of 1 × 10−8 M H2 O2 are presented in Fig. 5; the inset shows the calibration curve. Well-defined current responses for H2 O2 reduction were obtained at the PANI-g-

MWNT-ME. Fast dynamic equilibrium attainment was noticed at the PANI-g-MWNT-ME upon each addition of H2 O2 solution, generating a steady-state current signal within 8 s. The response at the PANI-g-MWNT-ME to H2 O2 is linear between the con-

Fig. 4. (A) Flow injection amperometric response of the PANI-g-MWNT-ME to H2 O2 solutions of increasing concentrations ((1–6) × 10−8 M; a–f). (B and C) Flow injection response to successive injections of 6 × 10−8 M H2 O2 at PANIg-MWNT-ME and 1 ␮M H2 O2 at PANI-ME, respectively. Operating potential: −0.3 V; flow rate: 1.0 mL min−1 .

P. Santhosh et al. / Analytica Chimica Acta 575 (2006) 32–38

37

was observed at the potential of −300 mV. The subsequent injection of relevant physiological levels (1 ␮M) of UA, AA and AP did not show any additional signal or modify the current response (Fig. 6). The accelerated electron transfer reaction due to the presence of MWNT and PANI allows H2 O2 detection at very low potential. The lower overpotential for H2 O2 detection at this electrode limits the effects from interferents (Fig. 6), indicating high selectivity toward the H2 O2 . 4. Conclusions

Fig. 5. Amperometric responses of PANI-g-MWNT-ME to successive additions of 1 × 10−8 M hydrogen peroxide (inset shows the calibration curve). Potential: −0.3 V; electrolyte, phosphate buffer (pH 7.0).

centration range of 1 and 20 × 10−8 M of H2 O2 . The response curve is very sensitive at low detection limits for H2 O2 due to lower signal-to-noise ratio at PANI-g-MWNT-ME. PANIg-MWNT-ME displayed a lower detection limit of 1 nM at −300 mV. This limit is much lower than noticed with other modified electrodes [25–28]. Recently, the sultonated PANI microelectrode [29] for H2 O2 determination has been reported. Such film electrodes have displayed a low detection limit of 1 ppm at 0 V, although the linear dynamic range of the determination was rather limited to the lower concentration. 3.5. Influence of interferences Fig. 6 represents the amperometric response for H2 O2 (1 ␮M), uric acid (UA), ascorbic acid (AA) and acetaminophen (AP) at the PANI-g-MWNT-ME. A well-defined H2 O2 response

Fig. 6. Amperometric responses of PANI-g-MWNT-ME to hydrogen peroxide (1 ␮M) (a), uric acid (b), ascorbic acid (c) and acetaminophen (d) at −0.3 V. Electrolyte, phosphate buffer (pH 7.0).

We have demonstrated a new approach for the fabrication of modified electrode based on grafting of PANI chains onto MWNT for the selective and sensitive detection of H2 O2 . This new modified electrode has augmented catalytic performance from the combined influence of MWNT and PANI. PANI-gMWNT-ME displayed a linear range of response (between 1 and 20 × 10−8 M), better sensitivity (upto 1 nM), improved multiple use capability and faster response time (∼8 s) compared to PANI-ME. Also, PANI-g-MWNT-ME eliminates the potential interferences and preferentially detects H2 O2 . Though the method of fabrication of the electrode seems to be time consuming, the better sensitivity, faster response and minimum interference with PANI-g-MWNT-ME outweigh this drawback. The enhanced catalytic activity exhibited by PANI-g-MWNTME provides ample scope for using this modified electrode for a wide range of sensing applications. References [1] C.N.R. Rao, B.C. Satishkumar, A. Govindaraj, M. Nath, Chem. Phys. Chem. 2 (2001) 782. [2] R.H. Baughman, A. Zakhidov, W.A. de Heer, Science 297 (2002) 787. [3] Q. Zhao, Z. Gan, Q. Zhuang, Electroanalysis 14 (2002) 1609. [4] J. Wang, M. Li, Z. Shi, N. Li, Z. Gu, Anal. Chem. 74 (2002) 1993. [5] M. Musameh, J. Wang, A. Merkoci, Y. Lin, Electochem. Commun. 4 (2002) 743. [6] J. Wang, M. Li, Z. Shi, N. Li, Z. Gu, Electroanalysis 14 (2002) 225. [7] J. Wang, A.M. Kawde, M. Musameh, Analyst 128 (2003) 912. [8] Y. Zhao, W.D. Zheng, H. Chen, Q.M. Luo, Talanta 58 (2002) 529. [9] J. Wang, M. Musameh, Y. Lin, J. Am Chem. Soc. 125 (2003) 2408. [10] D. McQuade, A.E.P. Tyler, T.M. Swager, Chem. Rev. 100 (2000) 2537. [11] P.N. Bartlett, Y. Astier, Chem. Commun. (2000) 105. [12] Z. Cai, C.R. Martin, J. Electroanal. Chem. 300 (1991) 35. [13] M.A. Hamon, J. Chen, H. Hu, Y.S. Chen, M.E. Itkis, A.M. Rao, Adv. Mater. 11 (1999) 834. [14] J.E. Riggs, Z.X. Guo, D.L. Carroll, Y.P. Sun, J. Am. Chem. Soc. 122 (2000) 5879. [15] M. Hughes, G.Z. Chen, M.S. Shaffer, D.J. Fray, A.H. Windle, Chem. Mater. 14 (2002) 1610. [16] Q. Xiao, X. Zhou, Electrochim. Acta 48 (2002) 575. [17] K. Lota, V. Khomenko, E.J. Frackowiak, Phys. Chem. Solids 65 (2004) 295. [18] S. Carrara, V. Bavastrello, D. Ricci, E. Stura, C. Nicolini, Sens. Actuators B 109 (2005) 221. [19] J. Wang, M. Musameh, Anal. Chim. Acta 539 (2005) 209. [20] K.P. Loh, S.L. Zhao, W.D. Zhang, Diamond Relat. Mater. 12 (2004) 1075. [21] M. Guo, J. Chen, J. Li, B. Tao, S. Yao, Anal. Chim. Acta 532 (2005) 71. [22] F. Qu, M. Yang, J. Jiang, G. Shen, R. Yu, Anal. Biochem. 344 (2005) 108.

38

P. Santhosh et al. / Analytica Chimica Acta 575 (2006) 32–38

[23] T.C. Wen, Y.H. Chen, A. Gopalan, Mater. Chem. Phys. 77 (2002) 559. [24] W.C. Chen, T.C. Wen, C.C. Hu, A. Gopalan, Electrochim. Acta 47 (2002) 1305. [25] M.S.M. Quintino, H. Winnischofer, K. Araki, H.E. Toma, L. Angnes, Analyst 130 (2005) 221.

[26] Y. Hasebe, T. Gu, J. Electroanal. Chem. 576 (2005) 177. [27] Y. Xu, W. Peng, X. Liu, G. Li, Biosens. Bioelectron. 20 (2004) 533. [28] X. Wang, H. Zhang, E. Wang, Z. Han, C. Hu, Mater. Lett. 58 (2004) 1661. [29] D. Raffa, K.T. Leung, F. Battaglini, Anal. Chem. 75 (2003) 4983.