- Email: [email protected]
Electrochemical conversion of poly-aniline into a redox polymer in the presence of nordihydroguaiaretic acid
Journal of Electroanalytical Chemistry 626 (2009) 143–148
Contents lists available at ScienceDirect
Journal of Electroanalytical Chemistry journal homepage: www.elsevier.com/locate/jelechem
Electrochemical conversion of poly-aniline into a redox polymer in the presence of nordihydroguaiaretic acid Grzegorz Milczarek * Institute of Chemistry and Technical Electrochemistry, Poznan University of Technology, Piotrowo 3, PL-60-965 Poznan, Poland
a r t i c l e
i n f o
Article history: Received 12 August 2008 Received in revised form 28 November 2008 Accepted 4 December 2008 Available online 11 December 2008 Keywords: Electropolymerization Redox polymer Poly-aniline Nordihydroguaiaretic acid NADH
a b s t r a c t Thin films of poly-aniline (PANI) undergo electrochemical restructuring when subjected to potential cycling in the presence of nordihydroguaiaretic acid (NDGA). As a consequence, the properties of PANI change from those of a conducting polymer to those of a redox polymer. Thus formed composite PANI–NDGA films show redox activity characterized by two new electrochemical couples observed at mid-peak potentials of 0.32 and 0.52 V (vs. Ag/AgCl, 0.5 M H2SO4). The peak currents exhibit a linear dependence on the potential scan rate as expected for surface-confined redox species and the peak potentials shift toward lower potentials with increasing solution pH at a rate of ca. 60 mV/pH unit. The PANI– NDGA composite film was found to electrocatalyze the oxidation of NADH in a nearly neutral (pH 7.4) phosphate buffer. Ó 2008 Elsevier B.V. All rights reserved.
1. Introduction Poly-aniline (PANI) has been attracting the attention of electrochemists for many years because of its easy preparation, high electroconductivity as well as attractive electrochemical and optical properties [1,2]. The main drawback of PANI is however the limited pH range in which PANI retains its electrochemical activity. It usually becomes insulating and electrochemically inactive at pH above 4 [3], which limits its use as a sensing platform for applications requiring neutral or alkaline electrolytes. Many approaches have been used to overcome this problem. These are based on (i) postpolymerization grafting of sulfonic groups on PANI chains by fuming sulfuric acid treatment [4], copolymerization of parent aniline with –COOH or –SO3H functionalized aniline derivatives [5–8], or performing (electro)polymerization in the presence of various organic acids [9–14]. All these approaches by introducing an ionogenic acidic group to the PANI structure hinder the deprotonation of the conducting form of PANI (emeraldine salt) and thus extend its electroactivity toward less acidic pHs. PANIs when deposited on common electrodic materials as thin films exhibit valuable electrocatalytic properties and have been used as transducers for electrochemical sensing of different organic (bio)molecules. Electrochemical detections of ascorbic acid (AA) and reduced nicotinamide adenine dinucleotide (NADH) are the most representative examples of such applications [9,10,12,14– * Tel.: +48 61 66 52 158; fax: +48 61 66 52 571. E-mail address: [email protected] 0022-0728/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jelechem.2008.12.002
18]. PANI was also found to enhance the reversibility of electrochemical oxidation/reduction of reversible systems such as Fe2+/ Fe3+ and hydroquinone/quinone [19]. However, researchers overlooked one important reaction that may take place along with this redox cycling, namely the covalent attachment of electrogenerated quinone (Q) to the PANI chain, which generates new redox active groups. This process was observed both for parent catechol and electrogenerated catechol-sulfonic acid, leading in the latter case to the generation of self-doped PANI showing a new voltammetric couple in cyclic voltammograms, electroactivity in a wide pH range and electrocatalytic properties toward the oxidation of AA [20]. It seems likely that the above mentioned reaction due to the commercial availability of a large number of catechol derivatives may become a simple and effective entry point for post-polymerization modification of PANI, giving new materials with attractive (electro)chemical properties. In this paper a method of converting ultrathin PANI films to a redox active material by an electrochemical reaction with nordihydroguaiaretic acid (1,4-bis-(3,4-dihydroxybenzene)-2,3dimethylbutane, NDGA) is presented and discussed in detail. 2. Experimental 2.1. Chemicals Nordihydroguaiaretic acid (NDGA) was obtained from Sigma, aniline from POCH (Gliwice, Poland). Aniline was distilled under reduced pressure before use. Doubly distilled water was used to
144
G. Milczarek / Journal of Electroanalytical Chemistry 626 (2009) 143–148
prepare all solutions. All buffer materials were reagent grade and used without further purification. 2.2. Apparatus The electrochemical experiments were carried out using a
lAutolab electrochemical analyzer (EcoChemie, Utrecht – Netherlands) connected to a PC for control, data acquisition and storage. An Au electrode (BAS MF2014, 1.6 mm diameter) was used as the substrate electrode for all experiments. The counter electrode was a platinum wire. All potentials reported in this paper are referenced to an Ag/AgCl electrode with no regard for the liquid junction potential. A rotating disc electrode (EcoChemie) with an Au active disc (3.0 mm diameter) was used for electrochemical measurement under hydrodynamic conditions. 2.3. Electrode preparation Before its modification, the working surface was polished with alumina slurries of 1 and 0.05 lm on Buehler polishing cloth with water as a lubricant, rinsed with doubly distilled water, and sonicated in a water bath for 3 min. PANI electrodeposition was performed using commonly applied conditions. Briefly 0.05 mM aniline in 0.5 M HCl was used as the plating solution and the potential of the working electrode was scanned between -0.2 and 0.9 V at 100 mV s1 for 2 scans to initiate the electropolymerization, which was followed by scanning between –0.2 and 0.78 V for a desired number of cycles. After the modification procedure, the PANI-modified electrode was thoroughly rinsed with water and cycled between –0.2 and 0.8 V in 0.5 M H2SO4/acetonitril (1:1 v/v) containing 0.2 mM NDGA until a steady cyclic voltammogram was obtained, i.e. typically for 50 scans. Thus obtained PANI– NDGA electrode was characterized electrochemically as described in the next section.
Fig. 1. Cyclic voltammograms of 0.2 mM NDGA at bare gold (A) and PANI-modified electrode (B). Electrolyte; 0.5 M H2SO4/acetonitril (1:1 v/v), scan rate; 20 mV s1. (C) Shows the effect of scan rate on cyclic voltammetry of the electrode obtained in (B) after transferring to pure 0.5 M H2SO4. Scan rate are 10, 20, 30, 40 and 50 mV s1 (from inner to outer).
2.4. AFM imaging For surface analysis PANI and PANI–NDGA films were prepared using the same conditions as those for voltammetric electrodes. A gold coated silicon wafer (Aldrich) was used as the substrate. Before deposition of polymeric films the substrate was cleaned by sonication in acetone for 5 min. The AFM imaging was accomplished using a Nanosurf EasyScan 2 AFM microscope. 2.5. FT-IR spectra RT-IR spectra of thin films of PANI and PANI–NDGA were recorded using a Bruker IFS 113v spectrometer working in an ATR mode. For this characterization the films were deposited electrochemically on a gold plated silicon wafer (Aldrich). Sample size was ca. 1.5 2 cm. 3. Results and discussion 3.1. Electrochemistry of NDGA at bare Au and PANI-modified Au electrodes Fig. 1 shows cyclic voltammograms of NDGA at bare Au (A) and PANI-modified Au electrodes (B) as well as a PANI electrode cycled in an NDGA solution after transferring to pure supporting electrolyte (C). As may be seen in Fig. 1A, at the applied scan rate the cyclic voltammogram of dissolved NDGA exhibited a typical quasireversible behavior characterized by the mid-peak potential (Eo0 ) of 0.52 V and peak separation (DEp) of 70 mV. This may be attributed to the reversible oxidation/reduction of catechol moieties of
NDGA to corresponding quinones ðQH2 () Q þ 2Hþ þ 2eÞ. Moreover, the anodic-to-cathodic peak ratio is nearly a unity as expected for a fully reversible redox system. The same process at the PANI layer is characterized by smaller DEp (40 mV), which is probably the effect of adsorption of hydrophobic NDGA molecules on organic PANI film, changing the electrochemistry of NDGA from diffusion-controlled to that typical for surface-confined systems. On the other hand, the peak ratio is no longer a unity and significantly increases, which indicates that the Q form of NDGA undergoes a follow-up reaction on this film. Yet the decrease of two couples assignable to PANI, i.e. those corresponding to leucoemeraldine/emeraldine (Eo0 of ca. 0.12 V) and emeraldine/pernigraniline (Eo0 of ca. 75 V) transitions may be seen, along with the development of a new redox system at Eo0 of 0.32 V. All these observations imply a significant restructuring of the PANI film and the formation of new redox active moieties within it. Note also a positive shift of the first PANI couple simultaneous with a negative shift of the second couple, leading finally to their merging with the NDGA peaks. The same shifts were noticed for poly-acrylic acid doped [21] PANI, and substituted PANIs [22], and were attributed to steric and electronic effects provided by bulky organic doping ions. The same effect seems likely for relatively large NDGA molecules incorporated into the PANI film. This assertion can also be supported by literature data showing significant shifts of PANI peaks during postpolymerization cycling in the presence of aminobezonitriles [23] or diamino-methylbenzoate [24], that lead to a covalent modification of the PANI structure. As the cycling continues, a steady-state voltammogram is reached after ca. 50 scans. On washing the modified electrode and transferring it into an NDGA-free electrolyte,
G. Milczarek / Journal of Electroanalytical Chemistry 626 (2009) 143–148
145
the two types of NDGA molecules is. This can be answered by a careful analysis of the observed complex cyclic voltammogram. Fig. 2 shows the anodic sweep of the cyclic voltammogram after background correction. This curve can be deconvoluted into four Gaussian curves with peak values at 0.25, 0.36, 0.50 and 0.59 V, respectively. Peaks I and IV correspond to two aforementioned transitions of PANI, while peaks II and III to QH2/Q transitions of bounded NDGA. The molar ratio of single bounded (ns) to double bounded (nd) NDGA molecules can be calculated from the integrals of curves II and III using the following equation:
ns 2 AIII ¼ nd AII AIII
ð1Þ
where AII and AIII are surface areas of peaks II and III, respectively. The calculation gave a value of 20.3, indicating that ca. 1 per 20 attached molecules is double bounded. Such molecules act as bridges between PANI chains and cause an apparent increase of the molecular weight of the modified PANI with respect to the starting material. Surface coverage values (C) for the two new redox systems calculated from the peak surface area were 6.65 109 and 6.03 109 for peak II and III, respectively.
Fig. 2. Deconvolution of anodic sweep of the cyclic voltammogram of PANI–NDGA electrode into Gaussian curves. Electrolyte; 0.5 M H2SO4, scan rate; 50 mV s1.
the cyclic voltammogram revealed peaks assigned to the newly formed PANI–NDGA composite, indicating permanent modification of PANI film (Fig. 1C). The new redox systems, i.e. these assigned to the formation of the Q form of NDGA and PANI-attached NDGA partly overlapping with parent PANI peaks, may be seen at Eo0 values of ca. 0.45 and 0.32 V, respectively. The peak currents exhibit a linear dependence on the scan rate as expected for surface-grafted systems. Since NDGA is a bifunctional compound, it can be attached to PANI chains as a single or double bonded molecule. The single bonded NDGA shows two redox transitions, one of them being assigned to the QH2/Q transition of the unreacted catechol moiety and the other to an analogous transition involving a catechol moiety attached to a nitrogen atom of the PANI chain. The presence of electron-donating nitrogen shifts the redox potential of the QH2/Q transition to significantly lower potentials. Double bounded NDGA molecules show only the latter type of electroactivity. The question which arises now is what the ratio between
3.2. AFM analysis Atomic force microscopy (AFM) has been proved to be an excellent tool for studying morphological properties of conducting polymers [25,26]. Therefore this technique was selected to analyze the effect of electrochemical cycling in the presence of NDGA on the morphology of thin PANI films. The representative AFM images of PANI films before and after restructuring with NDGA are shown in Fig. 3. For the parent PANI film the image reveals dense packing of ordered polymer bundles with an average width of ca. 100 nm. This observation is consistent with results reported by other research groups [27]. The restructuring of PANI with NDGA resulted in an apparent increase of the root mean square (r.m.s.) roughness from ca. 3.7 nm for thus prepared PANI to around 7.7 nm for NDGA modified films. The roughening was probably the result of insertion of bulky NDGA molecules to the outermost layer of PANI, followed by the covalent attachment of these molecules to PANI chains, which could result in a distortion of the coil-like structure characteristic for parent PANI.
Fig. 3. AFM images of PANI films before (A) and after reaction with NDGA (B). Taping mode, image size; 4 4 lm.
146
G. Milczarek / Journal of Electroanalytical Chemistry 626 (2009) 143–148
3.3. FT-IR analysis To justify the mechanism proposed above, the ex situ FT-IR spectra of the native and NDGA-functionalized PANI were recorded and are shown in Fig. 4. As it may be seen, both spectra show intense bands between 1100 and 1600 cm1 due to bending of C–H of 1,4-disubstituted aromatic ring (ca. 1110 cm1), stretching of C–N of aromatic amine (ca. 1305 cm1), stretching of benzenoid C@C (ca. 1500 cm1) and stretching of quinoid C@C (ca. 1590 cm1), respectively [23]. These bands do not change appreciably upon NGDA functionalization of PANI. More dramatic changes are seen at higher wavelength numbers. For the NDGAfunctionalized PANI film, new absorption bands appeared at 1720 and ca. 2900 cm1 which can be assigned to carbonyl C@O and aliphatic C–H stretching modes. This suggests that oxidized form of NDGA is incorporated into the PANI film. However the formation of covalent C–N bond between the two species is justified by the disappearance of a minor band at ca. 3380 cm1 assignable to the N–H stretching mode. The disappearance of this band clearly confirms the formation of a new covalent bond between N atoms of PANI and the incoming NGDA molecules. A similar bond formation mechanism was suggested to occur for PANI-lignin composite [28]. Note that during electrochemical oxidation, lignins form o-quinones [29] as NDGA does [30]. 3.4. Electrochemical and electrocatalytic properties of the PANI–NDGA modified gold electrode
Fig. 5. Effect of pH on cyclic voltammetry of PANI-NDGA modified electrode in a series of B–R buffers. Scan rate; 10 mV s1.
As expected for quinone-type modifiers, the formal potentials of the redox systems of the PANI–NDGA composite are significantly affected by the pH of electrolyte. Fig. 5 shows a series of cyclic voltammograms of the composite-modified electrode recorded in a series of Britton–Robinson (B–R) buffers with pH ranging from 1.8 to 8.6. As it may be seen, a progressive shift of the overlapped voltammetric peaks toward lower potentials is apparent over the whole applied pH range. Although a direct determination of peak potentials and thus Eo0 values from these voltammograms is not
easy due to the overlapping of two redox systems, a rough estimation of the cathodic and anodic peak potentials reveals that the mid-peak potentials (equal to Eo0 ) shifted negatively with pH with a slope of ca. 60 mV/pH unit for both redox systems for pH lower than 7.5, which is very close to the theoretical value of 59 mV/pH unit predicted by the Nernst equation for the two-proton/twoelectron process. It must be emphasized here that PANI–NDGA composite films obtained only from relatively thin PANI layers showed well pronounced electroactivity in the selected pH range. Fig. 6 shows the effect of the starting PANI film thickness (expressed as the charge of the leucoemeraldine/emeraldine peak,
Fig. 4. FT-IR spectra of PANI (A) and PANI-NDGA (B) thin films along with assignment of the most important bands.
Fig. 6. Effect of PANI thickness (expressed as charge) on the cyclic voltammetry of PANI–NDGA composite. Electrolyte; phosphate buffer (pH 7.4), scan rate; 10 mV s1.
G. Milczarek / Journal of Electroanalytical Chemistry 626 (2009) 143–148
QPANI) on the cyclic voltammetry of the final composite in B–R buffer of pH 7.4. It is evident that for films having QPANI higher than ca. 0.5 mC cm2 voltammograms show progressive increase of the peak(s) separations and highly ohmic character. This result can be explained if one supposes that due to the relatively large molecular size NDGA molecules are attached mainly to the outermost layer of the PANI film and the PANI–NDGA composites obtained from thicker PANI films behave as if they were a bilayer assembly, consisting of an inner layer of a conducting polymer (unmodified PANI) and another layer of a redox polymer. On pH rising above four the inner PANI layer becomes deprotonated and thus insulating, making it difficult to transfer electrons between the electrode support and the outer redox polymer. In contrast, thin films are almost completely converted into the redox polymer and thus by close contact with the gold electrode are able to exchange electrons in a wide pH range. It seems also likely that the other effect responsible for extension of the pH range in which the composite remains active is the protonation of imine nitrogens of PANI with hydroxyl group of NDGA. The same type of interaction was suggested for PANI-lignin complexes [31]. Reversible redox systems confined to the electrode surface have been used as electron shuttles for solution species showing sluggish electroactivity at unmodified electrodes. The mediated electrochemical process, however, requires certain positions of Eo0 of the solution species respective to the Eo0 of the mediator. For a mediated electrooxidation the redox potential of the reducing solute must be located below the Eo0 of the electrocatalyst. For the mediated electroreduction process, the reverse is the case. A biologically important coenzyme – reduced nicotinamide adenine dinucleotide (NADH) which is characterized by a fairly negative formal potential of –0.56 V (vs. SCE) [32] was selected to study the electrocatalytic properties of the PANI–NDGA composite film deposited on the gold electrode. Fig. 7 shows cyclic voltammograms of the PANI–NDGA modified gold electrode recorded in phosphate buffer (pH 7.4), in the
147
absence and presence of different amounts of NADH. The capability of the PANI–NDGA film to electrocatalyze the oxidation of this coenzyme is clearly observed. The electroxidation reaction starts at ca. –0.05 V and reaches a peak value at 0.1–0.2 V (depending on the NADH concentration). The anodic catalytic peak current increases linearly with an increase of the NADH level within the applied concentration range. Note that at an unmodified gold electrode the NADH electrooxidation is observed at a fairly positive potential of 0.71 V (not shown), meaning that the overpotential reduction provided by the electrocatalyst is at least 0.5 V. The observed catalytic currents are comparable with those observed for electropolymerized NDGA [30] and other catechol derivatives, but the electrooxidation process starts at a potential by ca. 0.1 V lower, which is presumably due to the presence of an attached electron-donating nitrogen atom and the formation of new iminoquinone redox active moieties, which are known to be very active electrocatalysts of NADH electrooxidation [17,33,34]. 3.5. Stability of the PANI–NDGA modified electrode The composite electrode stability was investigated by occasional recording of the cyclic voltammograms during storage in either 0.5 M H2SO4 or phosphate buffer (pH 7.4) at room temperature. After being stored for 7 days, the observed redox currents dropped to 82% and 64% of the initial values for the former and latter storage conditions, respectively. 4. Conclusions It has been proved that thin PANI films may be modified by subsequent potential cycling in the presence of NDGA. The eventually obtained new, modified polymer may be considered as a redox polymer. After modification the film character changes from a mixed metallic and redox type, observed at low pH, to a purely redox type, apparent in neutral or slightly alkaline electrolytes. The catalytic activity toward selected biologically active analytes and the easy preparation of the composite film offer interesting opportunities for the development of electrochemical sensors. Acknowledgement Ministry of Science and Higher Education of the Polish Government is acknowledged for financial support under Project No. N205 0857 33. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16]
Fig. 7. Cyclic voltammograms of PANI–NDGA modified gold electrode in the absence (dashed line) and the presence of NADH (solid lines). Scan rate; 10 mV s1.
[17] [18] [19] [20]
K. Hyodo, Electrochim. Acta 39 (1994) 265. J. Huang, Pure Appl. Chem. 78 (2006) 15. A.F. Diaz, J.A. Logan, J. Electroanal. Chem. 111 (1980) 111. C. Li, S. Mu, Synth. Met. 149 (2005) 143. D.-M. Zhou, J.-J. Xu, H.-Y. Chen, H.-Q. Fang, Electroanalysis 9 (1997) 1185. C. Barbero, H.J. Salavagione, D.F. Acevedo, D.E. Grumelli, F. Garay, G.A. Planes, G.M. Morales, M.C. Miras, Electrochim. Acta 49 (2004) 3671. A.A. Karyakin, A.K. Strakhova, A.K. Yatsimirsky, J. Electroanal. Chem. 371 (1994) 259. J.-J. Xu, D.-M. Zhou, H.-Y. Chen, Fresenius J. Anal. Chem. 362 (1998) 234. J.-J. Sun, D.-M. Zhou, H.-Q. Fang, H.-Y. Chen, Talanta 45 (1998) 851. S. Mu, J. Kan, Synth. Met. 132 (2002) 29. S. Mu, Synth. Met. 139 (2003) 287. L. Zhang, S. Dong, J. Electroanal. Chem. 568 (2004) 189. E.M. Barrios, G.A. Mulica, C.L. Velasquez, J. Electroanal. Chem. 586 (2006) 128. L. Zhang, Electrochim. Acta 52 (2007) 6969. P.N. Bartlett, E. Simon, Phys. Chem. Chem. Phys. 3 (2001) 1491. O.A. Raitman, E. Katz, A.F. Buckmann, I. Willner, J. Am. Chem. Soc. 124 (2002) 6487. S.A. Kumar, S.-M. Chen, Sensors 8 (2008) 739. P.N. Bartlett, E. Simon, Phys. Chem. Chem. Phys. 2 (2000) 2599. Z. Mannic, L. Duic, J. Electroanal. Chem. 403 (1996) 133. G. Milczarek, Electrochem. Commun. 9 (2007) 123.
148
G. Milczarek / Journal of Electroanalytical Chemistry 626 (2009) 143–148
[21] L.T. Cai, H.Y. Chen, J. Appl. Electrochem. 28 (1998) 161. [22] M.T. Nguyen, P. Kasai, J.L. Miller, A.F. Diaz, Macromolecules 27 (1994) 3625. [23] M. Sato, I. Fushimi, K. Takahashi, J. Nakaya, J. Chem. Soc. Chem. Commun. (1994) 2373. [24] G. Inzelt, E. Csahok, V. Kertesz, Electrochim. Acta 46 (2001) 3955. [25] K.D. O’Neil, O.A. Semenikhin, J. Phys. Chem. 111 (2007) 14823. [26] D.H. Han, S.-M. Park, J. Phys. Chem. B 108 (2004) 13921. [27] M. Sharma, D. Kaushik, R. Raj Singh, R.K. Pandy, J. Mater. Sci.–Mater. Electron. 17 (2006) 537.
[28] P.C. Rodrigues, M.P. Cantao, P. Janissek, P.C.N. Scarpa, A.L. Mathias, L.P. Ramos, M.A.B. Gomes, Eur. Polym. J. 38 (2002) 2213. [29] G. Milczarek, Electroanalysis 19 (2007) 1411. [30] A. Ciszewski, G. Milczarek, Anal. Chem. 72 (2000) 3203. [31] P.C. Rodrigues, M. Muraro, C.M. Garcia, G.P. Souza, M. Abbate, W.H. Schreiner, M.A.B. Gomes, Eur. Polym. J. 37 (2001) 2223. [32] C.R. Raj, B.K. Jena, Chem. Commun. 15 (2005) 2005. [33] L. Gorton, E. Dominguez, Rev. Mol. Biotechnol. 82 (2002) 371. [34] X. Tu, Q. Xie, Z. Huang, Q. Yang, S. Yao, Electroanalysis 19 (2007) 1815.
Contents lists available at ScienceDirect
Journal of Electroanalytical Chemistry journal homepage: www.elsevier.com/locate/jelechem
Electrochemical conversion of poly-aniline into a redox polymer in the presence of nordihydroguaiaretic acid Grzegorz Milczarek * Institute of Chemistry and Technical Electrochemistry, Poznan University of Technology, Piotrowo 3, PL-60-965 Poznan, Poland
a r t i c l e
i n f o
Article history: Received 12 August 2008 Received in revised form 28 November 2008 Accepted 4 December 2008 Available online 11 December 2008 Keywords: Electropolymerization Redox polymer Poly-aniline Nordihydroguaiaretic acid NADH
a b s t r a c t Thin films of poly-aniline (PANI) undergo electrochemical restructuring when subjected to potential cycling in the presence of nordihydroguaiaretic acid (NDGA). As a consequence, the properties of PANI change from those of a conducting polymer to those of a redox polymer. Thus formed composite PANI–NDGA films show redox activity characterized by two new electrochemical couples observed at mid-peak potentials of 0.32 and 0.52 V (vs. Ag/AgCl, 0.5 M H2SO4). The peak currents exhibit a linear dependence on the potential scan rate as expected for surface-confined redox species and the peak potentials shift toward lower potentials with increasing solution pH at a rate of ca. 60 mV/pH unit. The PANI– NDGA composite film was found to electrocatalyze the oxidation of NADH in a nearly neutral (pH 7.4) phosphate buffer. Ó 2008 Elsevier B.V. All rights reserved.
1. Introduction Poly-aniline (PANI) has been attracting the attention of electrochemists for many years because of its easy preparation, high electroconductivity as well as attractive electrochemical and optical properties [1,2]. The main drawback of PANI is however the limited pH range in which PANI retains its electrochemical activity. It usually becomes insulating and electrochemically inactive at pH above 4 [3], which limits its use as a sensing platform for applications requiring neutral or alkaline electrolytes. Many approaches have been used to overcome this problem. These are based on (i) postpolymerization grafting of sulfonic groups on PANI chains by fuming sulfuric acid treatment [4], copolymerization of parent aniline with –COOH or –SO3H functionalized aniline derivatives [5–8], or performing (electro)polymerization in the presence of various organic acids [9–14]. All these approaches by introducing an ionogenic acidic group to the PANI structure hinder the deprotonation of the conducting form of PANI (emeraldine salt) and thus extend its electroactivity toward less acidic pHs. PANIs when deposited on common electrodic materials as thin films exhibit valuable electrocatalytic properties and have been used as transducers for electrochemical sensing of different organic (bio)molecules. Electrochemical detections of ascorbic acid (AA) and reduced nicotinamide adenine dinucleotide (NADH) are the most representative examples of such applications [9,10,12,14– * Tel.: +48 61 66 52 158; fax: +48 61 66 52 571. E-mail address: [email protected] 0022-0728/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jelechem.2008.12.002
18]. PANI was also found to enhance the reversibility of electrochemical oxidation/reduction of reversible systems such as Fe2+/ Fe3+ and hydroquinone/quinone [19]. However, researchers overlooked one important reaction that may take place along with this redox cycling, namely the covalent attachment of electrogenerated quinone (Q) to the PANI chain, which generates new redox active groups. This process was observed both for parent catechol and electrogenerated catechol-sulfonic acid, leading in the latter case to the generation of self-doped PANI showing a new voltammetric couple in cyclic voltammograms, electroactivity in a wide pH range and electrocatalytic properties toward the oxidation of AA [20]. It seems likely that the above mentioned reaction due to the commercial availability of a large number of catechol derivatives may become a simple and effective entry point for post-polymerization modification of PANI, giving new materials with attractive (electro)chemical properties. In this paper a method of converting ultrathin PANI films to a redox active material by an electrochemical reaction with nordihydroguaiaretic acid (1,4-bis-(3,4-dihydroxybenzene)-2,3dimethylbutane, NDGA) is presented and discussed in detail. 2. Experimental 2.1. Chemicals Nordihydroguaiaretic acid (NDGA) was obtained from Sigma, aniline from POCH (Gliwice, Poland). Aniline was distilled under reduced pressure before use. Doubly distilled water was used to
144
G. Milczarek / Journal of Electroanalytical Chemistry 626 (2009) 143–148
prepare all solutions. All buffer materials were reagent grade and used without further purification. 2.2. Apparatus The electrochemical experiments were carried out using a
lAutolab electrochemical analyzer (EcoChemie, Utrecht – Netherlands) connected to a PC for control, data acquisition and storage. An Au electrode (BAS MF2014, 1.6 mm diameter) was used as the substrate electrode for all experiments. The counter electrode was a platinum wire. All potentials reported in this paper are referenced to an Ag/AgCl electrode with no regard for the liquid junction potential. A rotating disc electrode (EcoChemie) with an Au active disc (3.0 mm diameter) was used for electrochemical measurement under hydrodynamic conditions. 2.3. Electrode preparation Before its modification, the working surface was polished with alumina slurries of 1 and 0.05 lm on Buehler polishing cloth with water as a lubricant, rinsed with doubly distilled water, and sonicated in a water bath for 3 min. PANI electrodeposition was performed using commonly applied conditions. Briefly 0.05 mM aniline in 0.5 M HCl was used as the plating solution and the potential of the working electrode was scanned between -0.2 and 0.9 V at 100 mV s1 for 2 scans to initiate the electropolymerization, which was followed by scanning between –0.2 and 0.78 V for a desired number of cycles. After the modification procedure, the PANI-modified electrode was thoroughly rinsed with water and cycled between –0.2 and 0.8 V in 0.5 M H2SO4/acetonitril (1:1 v/v) containing 0.2 mM NDGA until a steady cyclic voltammogram was obtained, i.e. typically for 50 scans. Thus obtained PANI– NDGA electrode was characterized electrochemically as described in the next section.
Fig. 1. Cyclic voltammograms of 0.2 mM NDGA at bare gold (A) and PANI-modified electrode (B). Electrolyte; 0.5 M H2SO4/acetonitril (1:1 v/v), scan rate; 20 mV s1. (C) Shows the effect of scan rate on cyclic voltammetry of the electrode obtained in (B) after transferring to pure 0.5 M H2SO4. Scan rate are 10, 20, 30, 40 and 50 mV s1 (from inner to outer).
2.4. AFM imaging For surface analysis PANI and PANI–NDGA films were prepared using the same conditions as those for voltammetric electrodes. A gold coated silicon wafer (Aldrich) was used as the substrate. Before deposition of polymeric films the substrate was cleaned by sonication in acetone for 5 min. The AFM imaging was accomplished using a Nanosurf EasyScan 2 AFM microscope. 2.5. FT-IR spectra RT-IR spectra of thin films of PANI and PANI–NDGA were recorded using a Bruker IFS 113v spectrometer working in an ATR mode. For this characterization the films were deposited electrochemically on a gold plated silicon wafer (Aldrich). Sample size was ca. 1.5 2 cm. 3. Results and discussion 3.1. Electrochemistry of NDGA at bare Au and PANI-modified Au electrodes Fig. 1 shows cyclic voltammograms of NDGA at bare Au (A) and PANI-modified Au electrodes (B) as well as a PANI electrode cycled in an NDGA solution after transferring to pure supporting electrolyte (C). As may be seen in Fig. 1A, at the applied scan rate the cyclic voltammogram of dissolved NDGA exhibited a typical quasireversible behavior characterized by the mid-peak potential (Eo0 ) of 0.52 V and peak separation (DEp) of 70 mV. This may be attributed to the reversible oxidation/reduction of catechol moieties of
NDGA to corresponding quinones ðQH2 () Q þ 2Hþ þ 2eÞ. Moreover, the anodic-to-cathodic peak ratio is nearly a unity as expected for a fully reversible redox system. The same process at the PANI layer is characterized by smaller DEp (40 mV), which is probably the effect of adsorption of hydrophobic NDGA molecules on organic PANI film, changing the electrochemistry of NDGA from diffusion-controlled to that typical for surface-confined systems. On the other hand, the peak ratio is no longer a unity and significantly increases, which indicates that the Q form of NDGA undergoes a follow-up reaction on this film. Yet the decrease of two couples assignable to PANI, i.e. those corresponding to leucoemeraldine/emeraldine (Eo0 of ca. 0.12 V) and emeraldine/pernigraniline (Eo0 of ca. 75 V) transitions may be seen, along with the development of a new redox system at Eo0 of 0.32 V. All these observations imply a significant restructuring of the PANI film and the formation of new redox active moieties within it. Note also a positive shift of the first PANI couple simultaneous with a negative shift of the second couple, leading finally to their merging with the NDGA peaks. The same shifts were noticed for poly-acrylic acid doped [21] PANI, and substituted PANIs [22], and were attributed to steric and electronic effects provided by bulky organic doping ions. The same effect seems likely for relatively large NDGA molecules incorporated into the PANI film. This assertion can also be supported by literature data showing significant shifts of PANI peaks during postpolymerization cycling in the presence of aminobezonitriles [23] or diamino-methylbenzoate [24], that lead to a covalent modification of the PANI structure. As the cycling continues, a steady-state voltammogram is reached after ca. 50 scans. On washing the modified electrode and transferring it into an NDGA-free electrolyte,
G. Milczarek / Journal of Electroanalytical Chemistry 626 (2009) 143–148
145
the two types of NDGA molecules is. This can be answered by a careful analysis of the observed complex cyclic voltammogram. Fig. 2 shows the anodic sweep of the cyclic voltammogram after background correction. This curve can be deconvoluted into four Gaussian curves with peak values at 0.25, 0.36, 0.50 and 0.59 V, respectively. Peaks I and IV correspond to two aforementioned transitions of PANI, while peaks II and III to QH2/Q transitions of bounded NDGA. The molar ratio of single bounded (ns) to double bounded (nd) NDGA molecules can be calculated from the integrals of curves II and III using the following equation:
ns 2 AIII ¼ nd AII AIII
ð1Þ
where AII and AIII are surface areas of peaks II and III, respectively. The calculation gave a value of 20.3, indicating that ca. 1 per 20 attached molecules is double bounded. Such molecules act as bridges between PANI chains and cause an apparent increase of the molecular weight of the modified PANI with respect to the starting material. Surface coverage values (C) for the two new redox systems calculated from the peak surface area were 6.65 109 and 6.03 109 for peak II and III, respectively.
Fig. 2. Deconvolution of anodic sweep of the cyclic voltammogram of PANI–NDGA electrode into Gaussian curves. Electrolyte; 0.5 M H2SO4, scan rate; 50 mV s1.
the cyclic voltammogram revealed peaks assigned to the newly formed PANI–NDGA composite, indicating permanent modification of PANI film (Fig. 1C). The new redox systems, i.e. these assigned to the formation of the Q form of NDGA and PANI-attached NDGA partly overlapping with parent PANI peaks, may be seen at Eo0 values of ca. 0.45 and 0.32 V, respectively. The peak currents exhibit a linear dependence on the scan rate as expected for surface-grafted systems. Since NDGA is a bifunctional compound, it can be attached to PANI chains as a single or double bonded molecule. The single bonded NDGA shows two redox transitions, one of them being assigned to the QH2/Q transition of the unreacted catechol moiety and the other to an analogous transition involving a catechol moiety attached to a nitrogen atom of the PANI chain. The presence of electron-donating nitrogen shifts the redox potential of the QH2/Q transition to significantly lower potentials. Double bounded NDGA molecules show only the latter type of electroactivity. The question which arises now is what the ratio between
3.2. AFM analysis Atomic force microscopy (AFM) has been proved to be an excellent tool for studying morphological properties of conducting polymers [25,26]. Therefore this technique was selected to analyze the effect of electrochemical cycling in the presence of NDGA on the morphology of thin PANI films. The representative AFM images of PANI films before and after restructuring with NDGA are shown in Fig. 3. For the parent PANI film the image reveals dense packing of ordered polymer bundles with an average width of ca. 100 nm. This observation is consistent with results reported by other research groups [27]. The restructuring of PANI with NDGA resulted in an apparent increase of the root mean square (r.m.s.) roughness from ca. 3.7 nm for thus prepared PANI to around 7.7 nm for NDGA modified films. The roughening was probably the result of insertion of bulky NDGA molecules to the outermost layer of PANI, followed by the covalent attachment of these molecules to PANI chains, which could result in a distortion of the coil-like structure characteristic for parent PANI.
Fig. 3. AFM images of PANI films before (A) and after reaction with NDGA (B). Taping mode, image size; 4 4 lm.
146
G. Milczarek / Journal of Electroanalytical Chemistry 626 (2009) 143–148
3.3. FT-IR analysis To justify the mechanism proposed above, the ex situ FT-IR spectra of the native and NDGA-functionalized PANI were recorded and are shown in Fig. 4. As it may be seen, both spectra show intense bands between 1100 and 1600 cm1 due to bending of C–H of 1,4-disubstituted aromatic ring (ca. 1110 cm1), stretching of C–N of aromatic amine (ca. 1305 cm1), stretching of benzenoid C@C (ca. 1500 cm1) and stretching of quinoid C@C (ca. 1590 cm1), respectively [23]. These bands do not change appreciably upon NGDA functionalization of PANI. More dramatic changes are seen at higher wavelength numbers. For the NDGAfunctionalized PANI film, new absorption bands appeared at 1720 and ca. 2900 cm1 which can be assigned to carbonyl C@O and aliphatic C–H stretching modes. This suggests that oxidized form of NDGA is incorporated into the PANI film. However the formation of covalent C–N bond between the two species is justified by the disappearance of a minor band at ca. 3380 cm1 assignable to the N–H stretching mode. The disappearance of this band clearly confirms the formation of a new covalent bond between N atoms of PANI and the incoming NGDA molecules. A similar bond formation mechanism was suggested to occur for PANI-lignin composite [28]. Note that during electrochemical oxidation, lignins form o-quinones [29] as NDGA does [30]. 3.4. Electrochemical and electrocatalytic properties of the PANI–NDGA modified gold electrode
Fig. 5. Effect of pH on cyclic voltammetry of PANI-NDGA modified electrode in a series of B–R buffers. Scan rate; 10 mV s1.
As expected for quinone-type modifiers, the formal potentials of the redox systems of the PANI–NDGA composite are significantly affected by the pH of electrolyte. Fig. 5 shows a series of cyclic voltammograms of the composite-modified electrode recorded in a series of Britton–Robinson (B–R) buffers with pH ranging from 1.8 to 8.6. As it may be seen, a progressive shift of the overlapped voltammetric peaks toward lower potentials is apparent over the whole applied pH range. Although a direct determination of peak potentials and thus Eo0 values from these voltammograms is not
easy due to the overlapping of two redox systems, a rough estimation of the cathodic and anodic peak potentials reveals that the mid-peak potentials (equal to Eo0 ) shifted negatively with pH with a slope of ca. 60 mV/pH unit for both redox systems for pH lower than 7.5, which is very close to the theoretical value of 59 mV/pH unit predicted by the Nernst equation for the two-proton/twoelectron process. It must be emphasized here that PANI–NDGA composite films obtained only from relatively thin PANI layers showed well pronounced electroactivity in the selected pH range. Fig. 6 shows the effect of the starting PANI film thickness (expressed as the charge of the leucoemeraldine/emeraldine peak,
Fig. 4. FT-IR spectra of PANI (A) and PANI-NDGA (B) thin films along with assignment of the most important bands.
Fig. 6. Effect of PANI thickness (expressed as charge) on the cyclic voltammetry of PANI–NDGA composite. Electrolyte; phosphate buffer (pH 7.4), scan rate; 10 mV s1.
G. Milczarek / Journal of Electroanalytical Chemistry 626 (2009) 143–148
QPANI) on the cyclic voltammetry of the final composite in B–R buffer of pH 7.4. It is evident that for films having QPANI higher than ca. 0.5 mC cm2 voltammograms show progressive increase of the peak(s) separations and highly ohmic character. This result can be explained if one supposes that due to the relatively large molecular size NDGA molecules are attached mainly to the outermost layer of the PANI film and the PANI–NDGA composites obtained from thicker PANI films behave as if they were a bilayer assembly, consisting of an inner layer of a conducting polymer (unmodified PANI) and another layer of a redox polymer. On pH rising above four the inner PANI layer becomes deprotonated and thus insulating, making it difficult to transfer electrons between the electrode support and the outer redox polymer. In contrast, thin films are almost completely converted into the redox polymer and thus by close contact with the gold electrode are able to exchange electrons in a wide pH range. It seems also likely that the other effect responsible for extension of the pH range in which the composite remains active is the protonation of imine nitrogens of PANI with hydroxyl group of NDGA. The same type of interaction was suggested for PANI-lignin complexes [31]. Reversible redox systems confined to the electrode surface have been used as electron shuttles for solution species showing sluggish electroactivity at unmodified electrodes. The mediated electrochemical process, however, requires certain positions of Eo0 of the solution species respective to the Eo0 of the mediator. For a mediated electrooxidation the redox potential of the reducing solute must be located below the Eo0 of the electrocatalyst. For the mediated electroreduction process, the reverse is the case. A biologically important coenzyme – reduced nicotinamide adenine dinucleotide (NADH) which is characterized by a fairly negative formal potential of –0.56 V (vs. SCE) [32] was selected to study the electrocatalytic properties of the PANI–NDGA composite film deposited on the gold electrode. Fig. 7 shows cyclic voltammograms of the PANI–NDGA modified gold electrode recorded in phosphate buffer (pH 7.4), in the
147
absence and presence of different amounts of NADH. The capability of the PANI–NDGA film to electrocatalyze the oxidation of this coenzyme is clearly observed. The electroxidation reaction starts at ca. –0.05 V and reaches a peak value at 0.1–0.2 V (depending on the NADH concentration). The anodic catalytic peak current increases linearly with an increase of the NADH level within the applied concentration range. Note that at an unmodified gold electrode the NADH electrooxidation is observed at a fairly positive potential of 0.71 V (not shown), meaning that the overpotential reduction provided by the electrocatalyst is at least 0.5 V. The observed catalytic currents are comparable with those observed for electropolymerized NDGA [30] and other catechol derivatives, but the electrooxidation process starts at a potential by ca. 0.1 V lower, which is presumably due to the presence of an attached electron-donating nitrogen atom and the formation of new iminoquinone redox active moieties, which are known to be very active electrocatalysts of NADH electrooxidation [17,33,34]. 3.5. Stability of the PANI–NDGA modified electrode The composite electrode stability was investigated by occasional recording of the cyclic voltammograms during storage in either 0.5 M H2SO4 or phosphate buffer (pH 7.4) at room temperature. After being stored for 7 days, the observed redox currents dropped to 82% and 64% of the initial values for the former and latter storage conditions, respectively. 4. Conclusions It has been proved that thin PANI films may be modified by subsequent potential cycling in the presence of NDGA. The eventually obtained new, modified polymer may be considered as a redox polymer. After modification the film character changes from a mixed metallic and redox type, observed at low pH, to a purely redox type, apparent in neutral or slightly alkaline electrolytes. The catalytic activity toward selected biologically active analytes and the easy preparation of the composite film offer interesting opportunities for the development of electrochemical sensors. Acknowledgement Ministry of Science and Higher Education of the Polish Government is acknowledged for financial support under Project No. N205 0857 33. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16]
Fig. 7. Cyclic voltammograms of PANI–NDGA modified gold electrode in the absence (dashed line) and the presence of NADH (solid lines). Scan rate; 10 mV s1.
[17] [18] [19] [20]
K. Hyodo, Electrochim. Acta 39 (1994) 265. J. Huang, Pure Appl. Chem. 78 (2006) 15. A.F. Diaz, J.A. Logan, J. Electroanal. Chem. 111 (1980) 111. C. Li, S. Mu, Synth. Met. 149 (2005) 143. D.-M. Zhou, J.-J. Xu, H.-Y. Chen, H.-Q. Fang, Electroanalysis 9 (1997) 1185. C. Barbero, H.J. Salavagione, D.F. Acevedo, D.E. Grumelli, F. Garay, G.A. Planes, G.M. Morales, M.C. Miras, Electrochim. Acta 49 (2004) 3671. A.A. Karyakin, A.K. Strakhova, A.K. Yatsimirsky, J. Electroanal. Chem. 371 (1994) 259. J.-J. Xu, D.-M. Zhou, H.-Y. Chen, Fresenius J. Anal. Chem. 362 (1998) 234. J.-J. Sun, D.-M. Zhou, H.-Q. Fang, H.-Y. Chen, Talanta 45 (1998) 851. S. Mu, J. Kan, Synth. Met. 132 (2002) 29. S. Mu, Synth. Met. 139 (2003) 287. L. Zhang, S. Dong, J. Electroanal. Chem. 568 (2004) 189. E.M. Barrios, G.A. Mulica, C.L. Velasquez, J. Electroanal. Chem. 586 (2006) 128. L. Zhang, Electrochim. Acta 52 (2007) 6969. P.N. Bartlett, E. Simon, Phys. Chem. Chem. Phys. 3 (2001) 1491. O.A. Raitman, E. Katz, A.F. Buckmann, I. Willner, J. Am. Chem. Soc. 124 (2002) 6487. S.A. Kumar, S.-M. Chen, Sensors 8 (2008) 739. P.N. Bartlett, E. Simon, Phys. Chem. Chem. Phys. 2 (2000) 2599. Z. Mannic, L. Duic, J. Electroanal. Chem. 403 (1996) 133. G. Milczarek, Electrochem. Commun. 9 (2007) 123.
148
G. Milczarek / Journal of Electroanalytical Chemistry 626 (2009) 143–148
[21] L.T. Cai, H.Y. Chen, J. Appl. Electrochem. 28 (1998) 161. [22] M.T. Nguyen, P. Kasai, J.L. Miller, A.F. Diaz, Macromolecules 27 (1994) 3625. [23] M. Sato, I. Fushimi, K. Takahashi, J. Nakaya, J. Chem. Soc. Chem. Commun. (1994) 2373. [24] G. Inzelt, E. Csahok, V. Kertesz, Electrochim. Acta 46 (2001) 3955. [25] K.D. O’Neil, O.A. Semenikhin, J. Phys. Chem. 111 (2007) 14823. [26] D.H. Han, S.-M. Park, J. Phys. Chem. B 108 (2004) 13921. [27] M. Sharma, D. Kaushik, R. Raj Singh, R.K. Pandy, J. Mater. Sci.–Mater. Electron. 17 (2006) 537.
[28] P.C. Rodrigues, M.P. Cantao, P. Janissek, P.C.N. Scarpa, A.L. Mathias, L.P. Ramos, M.A.B. Gomes, Eur. Polym. J. 38 (2002) 2213. [29] G. Milczarek, Electroanalysis 19 (2007) 1411. [30] A. Ciszewski, G. Milczarek, Anal. Chem. 72 (2000) 3203. [31] P.C. Rodrigues, M. Muraro, C.M. Garcia, G.P. Souza, M. Abbate, W.H. Schreiner, M.A.B. Gomes, Eur. Polym. J. 37 (2001) 2223. [32] C.R. Raj, B.K. Jena, Chem. Commun. 15 (2005) 2005. [33] L. Gorton, E. Dominguez, Rev. Mol. Biotechnol. 82 (2002) 371. [34] X. Tu, Q. Xie, Z. Huang, Q. Yang, S. Yao, Electroanalysis 19 (2007) 1815.