Ascorbate amperometric determination using conducting copolymers from aniline and N-(3-propane sulfonic acid)aniline

Transcription

1 Talanta 71 (2007) Ascorbate amperometric determination using conducting copolymers from aniline and N-(3-propane sulfonic acid)aniline Jorge Yánez Heras, Ana F. Forte Giacobone, Fernando Battaglini INQUIMAE, Departamento de Química Inorgánica, Analítica y Química Física, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Ciudad Universitaria, Pabellón 2, C1428EHA Buenos Aires, Argentina Received 30 May 2006; received in revised form 28 July 2006; accepted 30 July 2006 Available online 7 September 2006 Abstract The sequential electrochemical polymerization of aniline and N-(3-propane sulfonic acid)aniline (PSA) is proposed to construct a sensor able to detect ascorbate at physiological conditions. Compared to poly(aniline) modified electrode, a device with improved conducting and electrochemical properties at neutral ph is obtained. The electrochemical copolymerization of the same starting materials is also carried out. For a PSA:aniline ratio of 10:90, a polymer with a similar electrochemical behavior to the one grown in the sequential mode is observed. The detection of ascorbate was tested for both configurations at ph 7.2, the modified electrode is able to determine ascorbate at 0 mv versus Ag/AgCl; an optimized sensor constructed by sequential polymerization can easily detect ascorbate concentrations with a detection limit of 2.2 M. Uric acid and dopamine does not interfere in the ascorbate determination Elsevier B.V. All rights reserved. Keywords: Poly(aniline); N-(3-Propane sulfonic acid)aniline; Ascorbate; Amperometric detection 1. Introduction Ascorbate is a relevant biomolecule involved in the immune response, wound healing and the absorption of iron [1]; its concentration can be used to assess the stress in human as well as in plants [2,3]. It is present in many fruits and vegetables, and used in pharmaceutical preparations; therefore, its determination is important in many areas. Ascorbic acid can be electrocatalytically oxidized at polyaniline electrodes; its detection using conducting polymer modified electrodes has been recently reviewed [4]. It has been shown that, in a slightly acidic solution, the anodic peak for electro-oxidation of ascorbic acid shifts from 0.36 V versus Ag/AgCl at a bare platinum electrode to 0.13 V versus Ag/AgCl at a polyaniline modified electrode [5]. Also, polyaniline modified electrodes can be used for ascorbate determination at neutral ph; an autocatalytic mechanism has been proposed to explain the ability of polyaniline to electrocatalyze the oxidation of ascorbate at ph conditions where PANI is present in its undoped and nonconducting form [6]. For this type of modified electrode, an Corresponding author. Tel.: ; fax: address: battagli@qi.fcen.uba.ar (F. Battaglini). operating potential window of V was used and a 50 M limit of detection was achieved, this detection limit is likely due to PANI protonation dependence with the analyte. Poly(aniline) behaves as a conductor only in the halfoxidized form (emeraldine) when it is protonated. Above ph 5, the emeraldine deprotonates becoming an insulator. For a poly(aniline) film, the deprotonation of the emeraldine form is associated with the egress of both, protons and the associated anions from the film. This is only possible if the anions are small and mobile, for example chloride or bisulfate anions. If instead of mobile ions, long chain polymeric counter ions are used, they become trapped within the poly(aniline) film, and the overall process changes. As a consequence of this change, the conductivity of poly(aniline) can be maintained at a much higher ph [7]. This strategy has been used with several negatively charged polymers [8 12]. In particular, Bartlett and Wallace presented a poly(vinylsulfonate)/polyaniline composite electrode able to oxidize ascorbate at neutral ph using a working potential of 0.14 V versus Ag/AgCl [10]. Another approach that may aid the detection at neutral ph would be to modify the N atom in the backbone with an alkyl sulfonate group. Our group have modified PANI with 3-propane sultone to obtain a copolymer, poly(aniline-co-n-propane sulfonic acid aniline), with good conducting and electrochemical /$ see front matter 2006 Elsevier B.V. All rights reserved. doi: /j.talanta

2 J.Y. Heras et al. / Talanta 71 (2007) properties at physiological ph [13]. The synthesis of this copolymer involves two steps, the aniline electrochemical polymerization, followed by the chemical reaction between propane sultone and the PANI modified electrode. This second step could be troublesome if the reaction has to be carried out on small dimension electrodes. To overcome this drawback, it would be convenient to bind the propane sulfonate moiety to the polymer by electrochemical means. To achieve that goal, the synthesis of N-(3-propane sulfonic acid)aniline (PSA) from propane sultone and aniline was carried out. The new aniline derivative was electrochemical polymerized in different conditions, in contrast to previous works in which propane sultone is reacted with PANI in solution [14] or in heterogeneous phase [13]. Through the electrochemical polymerization of PSA with aniline, the construction of modified electrodes able to work at physiological ph was achieved. The construction was carried out in a sequential order, first the electrochemical polymerization of aniline followed by the electrochemical polymerization of PSA. The modified electrode obtained presents an improved conducting behavior at neutral ph compared with PANI modified electrodes. Also, the copolymerization from a solution containing PSA and aniline in a molar ratio of 10:90 is able to produce a modified electrode with improved properties compared to polyaniline at neutral ph. The modified electrodes are used as amperometric sensors for the detection of ascorbate. They are able to work at 0 mv versus Ag/AgCl with a limit of detection of 2.2 M in the case of the sequential electrochemical polymerization. Furthermore, the presence of other electroactive compounds, like uric acid or dopamine, does not interfere in the ascorbate determination. In the case of uric acid, no signal is observed. For dopamine a signal at higher potentials is obtained, which is easily distinguishable from the ascorbate signal. 2. Experimental 2.1. Reagents Aniline and propane sultone were from Aldrich. All other reagents used were analytical grade. Aniline was distilled prior to use Equipment Electrochemical measurements were performed using a PINE Instruments AFRDE 5 bipotentiostat. Signals were recorded in a computer using a LabPC 1200 data acquisition card (Texas Instruments). A Ag/AgCl electrode was used as reference and Pt as the counter electrode. Three millimeter diameter glassy carbon electrodes and home-built dual carbon band electrodes [13] were used as working electrodes Conductivity measurements The resistance of the polymers synthesized on the band electrodes was measured modifying a technique presented by Wrighton and co-workers [13,15]. A cell of two working electrodes, the dual carbon bands (WE), one reference electrode and one counter electrode is used. The two carbon bands are 12 m apart from each other, separated by an insulating gap. The polymer grows on both WE until the two bands are joined by it. Then, the modified electrodes are immersed in a solution at a given ph. One of the WE (let us say WE1) is kept at a constant potential with respect to the reference, while the potential of the other WE (WE2) is swept up and down by 20 mv with respect to WE1. The current flowing through the cell is the result of the electrochemical process plus the current that flows between the two WE due to the potential difference between them. By analogy with a physical transistor, the current flowing between the two working electrodes is called the drain current, the potential of WE1 with respect to the reference is the gate voltage, and the potential of WE2 referred to WE1 is the drain voltage. The current measured in any of the working electrodes is the sum of the current produced by a redox process and the current forced to flow between the two WE by the drain voltage (drain current). For such small potential changes (20 mv), the contribution to the current produced by the redox processes can be neglected with respect to the drain current. The current flowing through the polymer can be assumed to be the result of electronic conduction of the film, which is inversely proportional to the resistance. A source of error in the determination is the capacitive current, which can be avoided by stopping the voltage sweep in the positive (or negative) limit until the value of the current is constant Hydrodynamic techniques A wall jet cell was made in acrylic, with a 0.5 mm nozzle and 1 mm nozzle to electrode distance. A 3 mm diameter glassy carbon electrode was employed. The counter electrode was part of the stainless steel outlet tubing, and the reference electrode was a Ag/AgCl electrode placed downstream. The working potential was set at 0 mv versus Ag/AgCl. The sample (10 ml) circulated continuously through the cell at 1.5 ml min 1. For the flow injection system, a 200 L loop was used with the same flow rate Synthesis of PSA The synthesis of PSA was carried by dissolving 0.7 g of propane sultone in 2.5 ml of aniline under stirring at 30 C. After a few minutes a white precipitate is formed and the product is recrystalized from methanol/acetone, giving a white powder. 1 H NMR; 500 MHz, D 2 O; δ values were 2.15 (m, 2H), 2.95 (t, 2H), 3.55(t, 2H) and 7.5(m, 5H). M + : Polymerization Sequential polymerization was carried out by cyclic voltammetry between 0.2 and 0.85 V versus Ag/AgCl at 50 mv s 1. The first step was carried out by cycling the electrode in a solution of 0.09 M aniline in 1.8 M H 2 SO 4 for five times, unless stated otherwise; then, the electrode was rinsed with water and

3 1686 J.Y. Heras et al. / Talanta 71 (2007) immersed in a solution of 0.01 M PSA in 1.8 M H 2 SO 4 and cycled at 50 mv s 1 ; the number of cycles for this step is indicated in each experiment. Electrochemical copolymerizations were carried out by cyclic voltammetry between 0.2 and 0.85 V versus Ag/AgCl at 50 mv s 1, from a solution of PSA and recently distilled aniline in 1.8 M H 2 SO 4 with a total concentration of co-monomers equal to 0.1 M. The growth of the polymer on dual band electrodes was carried out by immersing the electrode in a 0.52 M aniline in 1.8 M H 2 SO 4 solution at fixed potential of 0.8 V versus Ag/AgCl for 3 min. The union of the band was checked by conductance measurements in acid medium and then, the electrochemical polimerization of PSA was carried out Ascorbate detection The determination of ascorbate was carried out in different ways. Cyclic voltammetry at 2 mv s 1 was used to establish the response and sensitivity to ascorbate and the effect of potential interferences of the different modified electrodes prepared. The results of these studies were used to choose the type of sensor and the working potential for the determinations carried out in the hydrodynamic techniques. 3. Results and discussion 3.1. Sensor construction and properties The reaction of aniline with propane sultone produces the aniline derivative (Scheme 1) in a few minutes. The compound is soluble in water and methanol giving a colorless solution. The electrochemical response of this compound in acid medium is depicted in Fig. 1. It presents peak currents at 0.52 and 0.7 V and the signal increases with continuous cycling until the 8th cycle; then, the current begins to decay and the solution turns green, like emeraldine, which can be attributed to the formation of a soluble polymer. Copolymerization of PSA with aniline was carried out from solutions with the following ratios of PSA:aniline: 75:25, 50:50, 25:75 and 10:90. The electrochemical polymerization of the three first solutions leads to cyclic voltammetries with peaks at 0.52 and 0.7 V, similar to the pure PSA. The growth rate of the polymer depends on the ratio of PSA:aniline; the higher is the proportion of PSA, the faster is the growth rate of the polymer. In all the cases the signal increases until the 12th cycle. However, unlike pure PSA, the signal remains stable, suggesting that the joint polymerization of PSA and aniline produces a Fig. 1. Cyclic voltammetry of electropolymerized PSA on a glassy carbon electrode in 1.8 M sulfuric acid at 50 mv s 1. polymer that remains anchored to the electrode surface. In none of these cases, the typical current peak for the first oxidation process of PANI, at 0.24 V, is observed. When a solution with a PSA:aniline ratio equal to 10:90 is used, the voltammogram shows two peaks in the first scans at 0.52 and 0.7 V versus Ag/AgCl; then, a peak at 0.25 V begins to grow faster than the others. The electrochemical response of this polymer after 20 scans is shown in Fig. 2. The presence of a broad new peak compared to the typical voltammogram of PANI indicates the incorporation of PSA as part of the electrically conducting polymer. In the sequential polymerization method, PANI was grown by cycling the potential five times at 50 mv s 1 between 0.20 and 0.85 V (Fig. 3, gray line). Then, the electrode was immersed in a 10 mm PSA solution in 1.8 M sulfuric acid and cycled four times between the same potentials (Fig. 3, thin black line). It can be observed that the peak current at 0.25 V increases from 60 A to practically 80 A and a broad anodic peak around 0.5 V develops. If cycling is continued, after 7 cycles (Fig. 3, bold black line), the first anodic peak shifts to a higher potential Scheme 1. Structure of N-(3-propane sulfonic acid)aniline (PSA). Fig. 2. Cyclic voltammetry in 1.8 M sulfuric acid of the polymerization product obtained from a solution containing 90 mm aniline and 10 mm PSA in 1.8 M H 2 SO 4. Sweep rate: 10 mv s 1.

4 J.Y. Heras et al. / Talanta 71 (2007) the conducting potential region follows the oxidation process for the formation of emeraldine. The maximum conductivity of this polymer decreases 25 times from acid to neutral medium Ascorbate determination Fig. 3. Cyclic voltammetries in 1.8 M H 2 SO 4 of PANI modified electrode (gray line), PANI + 4 cycles of PSA (thin black line) and PANI + 7 cycles of PSA (bold black line). Sweep rate: 10 mv s 1. but it does not increase further, suggesting that the oxidation process becomes slightly slower. In addition, a broader peak is observed in the range of V. If the cycling of the potential in presence of PSA continues after 15 cycles, the first anodic peak does not change, while the broad peak between 0.4 and 0.6 increases and the voltammogram shows a larger capacitive response (data not shown). The modified electrodes obtained either by copolymerization from a PSA:aniline solution (10:90) or by sequential polymerization, show a quasi-reversible behavior at ph 7.2. In both cases, the electrodes show a stable signal; the peak currents are around 70% of the original values for both types of electrodes after 40 min of cycling between 0.2 and 0.6 V at 50 mv s 1. Fig. 4 depicts the quasi-reversible response of an electrode modified in sequential order (PSA 7 cycles). The resistance of the same polymer grown between two band electrodes can also be observed; The equilibrium potential of the couple ascorbate dehydroascorbate is V versus Ag/AgCl, but oxidation at bare glassy carbon or platinum electrodes requires potentials of 0.4 and 0.6 V, respectively. These high over potentials result in electrode fouling, poor reproducibility and low selectivity when these materials are used for analytical applications. Due to these problems several groups have developed modified electrodes to catalyze the electrochemical oxidation of ascorbate; among them, Bartlett and Wallace have shown that it is possible to oxidize ascorbate at lower potentials using a polyaniline polyvinylsulfonate composite coated electrode [10]. For all the electrode configurations presented in this work, a catalytic response to the presence of ascorbate can be observed. As an example, Fig. 5 shows a cyclic voltammogram in buffer at ph 7.2, for a modified electrode constructed in a sequential mode (dotted line). When ascorbic acid is added to a final concentration of 125 M, a new peak at 25 mv appears due to its oxidation (solid line), while the cathodic wave of PANI dramatically decreases due to its reduction by ascorbate. Cyclic voltammetry was also used to study the sensitivity of the ascorbate response for the different electrode configurations. The electrode modified by copolymerization of aniline and PSA shows the poorest response and worst linear range (white squares in Fig. 6). The response improves in the case of the electrodes modified by subsequent polymerizations. In all of them, PANI was grown by cycling the potential five times between 0.2 and 0.85 V at 10 mv s 1 ; then, different amounts of PSA were polymerized. The best sensitivity and linear range is observed in those electrodes where the polymerization of PSA was carried Fig. 4. Cyclic voltammetry of a modified electrode by subsequent polymerization of aniline and PSA (7 cycles) in 0.1 M phosphate buffer ph 7.2 at 10 mv s 1 (currents on left axis); and resistance behavior of the same polymer in 0.1 M phosphate buffer ph 7.2, grown between two band electrodes at a distance of 12 m (right axis). Fig. 5. Cyclic voltammetry for a modified electrode by subsequent polymerization of aniline and PSA (7 cycles) in 0.1 M phosphate buffer ph 7.2 (dotted line), in the presence of 125 M ascorbate (solid line). Sweep rate: 2 mv s 1.

5 1688 J.Y. Heras et al. / Talanta 71 (2007) Fig. 6. Ascorbate response for different electrode configurations: copolymer from aniline and PSA (90:10) (white squares); PANI + 4 PSA cycles (diamonds); PANI + 7 PSA cycles (black squares); PANI + 15 PSA cycles (triangles). The values plotted are the currents observed at 20 mv vs. Ag/AgCl from a cyclic voltammetry at 2 mv s 1. Fig. 7. Cyclic voltammetry for a modified electrode by subsequent polymerization of aniline and PSA (7 cycles) in the presence of 125 M ascorbate (solid line) and 125 M dopamine (dotted line). Sweep rate: 2 mv s 1. out by 7 and 15 cycles (circles and triangles in Fig. 6). It can be observed for the last two electrodes that the response is practically the same. This is probably due to the fact that the addition of PSA to the polymer does not scale linear with the number of cycles. As it is was stated before, after 7 cycles the peak corresponding to 0.25 V does not increase further, indicating that the electroactive part of the polymer that takes part in the ascorbic detection remains the same. The effect of possible interferences was studied, the response to ascorbate in the presence of uric acid and dopamine was determined. The modified electrode constructed by the polymerization of aniline (5 cycles) followed by the polymerization of PSA (7 cycles) was used in these experiments. The presence of uric acid did not affect the response of the modified electrode in a working potential range from 0.2 to 0.5 V, in spite of uric acid presents an oxidation peak potential at 0.4 V on a bare carbon electrode. A similar result was obtained by O Connell et al. [16] using only polyaniline. These authors suggested that the polyaniline layer is acting as a permselective membrane avoiding the oxidation of uric acid on the electrode surface. In Fig. 7, the cyclic voltammogram for ascorbate at ph 7.2 (solid line) is compared to the response for ascorbate plus dopamine, at the same concentrations (dotted line). Dopamine is oxidized at higher potentials, with no effect on the ascorbate signal, allowing the possibility of simultaneous determination of both species, e.g. by means of square wave voltammetry [17]. A similar modified electrode than the one used for the interferences studies was tested for ascorbate determinations at low concentrations. In this case the working potential was fixed at 0 mv. The electrode was part of a wall jet cell with the sample continuously circulating. The calibration curve presents a linear range from 5 to 50 M(Fig. 8). Table 1 shows the equation obtained for the calibration graph and the regression coefficient. The detection limit was calculated on the basis of 3σ (σ being the residual standard deviation of the intercept), yielding a value of 2.2 M. Fig. 8. Ascorbate amperometric response in a flow system. Applied potential: 0 mv vs. Ag/AgCl, flow rate 1.5 ml min 1. Sensor constructed by consecutive electrochemical polymerization of aniline and PSA (7 cycles). The response in a flow injection system with a 200 L loop was also studied. For this experiment PANI was grown by cycling the electrode three times at 10 mv s 1, and then in presence of PSA four times more. The sensor, constructed in these conditions, shows good stability at ph 7.2 and the time response is faster than for the other electrodes previously shown, since a Table 1 Parameters for the calibration curve shown in Fig. 8 Parameter Value Slope (na M 1 ) 1.96 Intercept (na) 2.7 Correlation coefficient Limit of detection ( M) 2.2 Linear range ( M) 5 50

6 J.Y. Heras et al. / Talanta 71 (2007) thinner layer is used. Using this system and a linear range up to 100 M was obtained. 4. Conclusions The sequential polymerization of aniline and PSA produces a polymer with improved conducting and electroactive properties compared to PANI at ph 7. These properties make them very suitable to determine ascorbate at physiological conditions, which reacts with the polymer at lower potentials compared to glassy carbon, platinum or other modified electrodes presented in the literature. The low potential value for the ascorbate oxidation allows its determination in presence of uric acid and dopamine. In contrast to our previous report where PANI was chemically modified with propane sultone [13], these modified electrodes can be built completely by electrochemical means. Therefore their size can be severely reduced and they can be part of an array of different modified electrodes. For the last, the construction of the modified electrode would not even require the handling of small volumes, since it will only need to impose the right potential to the element of the array which is going to be modified. Acknowledgements The financial support of Universidad de Buenos Aires and ANPCyT is acknowledged. References [1] R. Berkow, A.J. Fletcher (Eds.), El Manual Merck de Diagnóstico y Terapéutica, 9th ed., Mosby/Doyma Libros, Barcelona, Spain, 1996, p (in Spanish). [2] P.L. Conklin, E.H. Willlams, E.L. Last, Proc. Natl. Acad. Sci. U.S.A. 93 (1996) [3] I. Koshiishi, T. Imanari, Anal. Chem. 69 (1997) [4] A. Malinauskas, R. Garjonyte, R. Mazeikiene, I. Jureviciute, Talanta 64 (2004) [5] S.L. Mu, J.Q. Kan, Synth. Meth. 132 (2002) [6] I. Jurevciute, K. Brazdziuviene, L. Bernotaite, B. Salkus, A. Malinauskas, Sens. Actuators B 107 (2005) [7] P.N. Bartlett, Y. Astier, J. Chem. Soc., Chem. Commun. (2000) [8] J. Liu, S. Tian, W. Knoll, Langmuir 21 (2005) [9] O.A. Raitman, E. Katz, A.F. Bückmann, I. Willner, J. Am. Chem. Soc. 124 (2002) [10] P.N. Bartlett, E.N.K. Wallace, Phys. Chem. Chem. Phys. 3 (2001) [11] P.N. Bartlett, E. Simon, Phys. Chem. Chem. Phys. 2 (2000) [12] P.N. Bartlett, E.N.K. Wallace, J. Electroanal. Chem. 486 (2000) [13] D. Raffa, K.T. Leung, F. Battaglini, Anal. Chem. 75 (2003) [14] S.A. Chen, G.-W. Hwang, J. Am. Chem. Soc. 117 (1995) [15] E.W. Paul, A.J. Ricco, M.S. Wrighton, J. Phys. Chem. 89 (1985) [16] P.J. O Connell, C. Gormally, M. Pravda, G.G. Guilbault, Anal. Chim. Acta 431 (2001) [17] P.R. Roy, T. Okajima, T. Ohsaka, Bioelectrochemistry 59 (2003)