Voltammetric determination of niclosamide at a glassy carbon electrode

Abstract

A very sensitive and selective procedure was developed for the determination of niclosamide based on square-wave voltammetry at a glassy carbon electrode. Cyclic voltammetry was used to investigate the electrochemical reduction of niclosamide at a glassy carbon electrode. Niclosamide was first irreversibly reduced from NO₂ to NHOH at −0.659 V in aqueous buffer solution of pH 8.5. Reversible and well defined peaks at −0.164 V and −0.195 V (vs. Ag/AgCl) were obtained that are responsible for two electron peaks between NHOH and NO. Following optimisation of the voltammetric parameters, pH and reproducibility, a linear calibration curve over the range 5 × 10⁻⁸–1 × 10⁻⁶ mol dm⁻³ was achieved. The detection limit was found to be 2.05 × 10⁻⁸ mol dm⁻³ niclosamide. For eight successive determinations of 5 × 10⁻⁷ mol dm⁻³ niclosamide, a relative standard deviation of 2.4% was obtained. This voltammetric method was applied to the direct determination of niclosamide in tablets. The results of the analysis suggest that the proposed method has promise for the routine determination of niclosamide in the products examined.

---

Introduction

Niclosamide (2′,5-dichloro-4′-nitrosalicylanilide, NA), with the structure shown in Fig. 1, is a drug that is effective against all the species of tapeworm infections.^1,2 Its mode of action against tapeworm species is to uncouple oxidative phosphorylation and it blocks the glucose uptake and inhibits respiration in cestodes.^3,4 Formulated as the ethanolamine salt, it is one of the most effective and widely used molluscicides for the control of snail vectors of schistosomiasis, a parasitic disease affecting over 200 million people in more than 70 countries.⁵ Owing to the wide application of NA, analytical methods for its determination need to be both selective and sensitive owing to the presence of potential interferences and to the low concentrations present. For environmental control, monitoring of its concentration levels in ponds and rivers where it is applied for the eradication of snail vectors is also very signifcant.

Fig. 1 Structure of niclosamide (2′,5-dichloro-4′-nitrosalicylanilide).

A number of qualitative and quantitative techniques for the determination of NA have been reported. These include spectrophotomteric techniques for derivatives or complexes of niclosamide such as sodium diethyldithiocarbamate,⁶ sulfamethoxazole and trimethylamine,⁷ ammonium reineckate⁸ and others,^9–12 differential ultraviolet spectrophotometric measurements,¹³ HPLC,^14–16 gas–liquid chromatography forming N,O-dimethyl⁵ and heptafluorobutyl derivatives¹⁷ and polarography.¹⁸ Of these, techniques, the spectrophotometric methods that involve complex formation or derivatisation are non-selective and subject to high interference. The polarographic method has a very narrow linear range (5–10 μg cm⁻³). The HPLC methods are simple and suitable for routine analysis and have low detection limits.

The voltammetric determination of drugs in pharmaceutical preparations is by far the most common use of electrochemistry for pharmaceutical analytical problems. As a rule, many of the active constituents of formulations (in contrast to excipients) are readily oxidised or reduced. The specificity of the method is usually excellent because the compound can be identified by its voltammetric redox potential.¹⁹ A survey of the literature revealed that no attempt has been made to determine NA and study its behaviour at solid electrodes using voltammetric methods.

In this paper, a new sensitive voltammetric method is described for the determination of NA at a glassy carbon electrode using the square-wave voltammetric technique. In 0.1 mol dm⁻³ NH₃–NH₄Cl buffer solution (pH 8.5), a sensitive square-wave voltammetric peak of NA was obtained at −0.148 V (vs. Ag/AgCl). A linear relationship held between the peak current and the concentration of NA in the range 5.00 × 10⁻⁸–1.00 × 10⁻⁶ mol dm⁻³ and the detection limit was 2.05 × 10⁻⁸ mol dm⁻³. The method was applied successfully to the determination of NA in pharmaceutical tablets.

Experimental

Apparatus

Voltammetric measurements were performed using a BAS100B Electrochemical Analyser [Bioanalytical Systems (BAS), USA] and a one-compartment glass cell vial (BAS MR-1208) with a three-electrode configuration (BAS Cell Stand C3). The electrodes used were a glassy carbon disk working electrode with a diameter of 3 mm (BAS MF-2012), a platinum wire auxiliary electrode (BAS MW-1032) and an Ag/AgCl (3 M NaCl) reference electrode (BAS MF-2052). The working electrode was polished with BAS polishing alumina on a micro-cloth pad and rinsed with water before use. The absorbance of NA was determined with a Cary 1E UV–vis spectrophotometer. The pH of the buffer solution was measured with a Hanna Instruments digital pH meter with a glass combination electrode. All potentials are reported with respect to an Ag/AgCl (3 mol dm⁻³ NaCl) reference electrode.

Reagents

Niclosamide was obtained from Sigma (USA), 2-chloro-4-nitroaniline from Aldrich (USA), 5-chloro-2-hydroxybenzoic acid from Merck-Schuchardt (Germany), ammonium chloride and disodium hydrogenphosphate from Saarchem (South Africa), ammonia solution, sodium perchlorate and citric acid from Riedel-de Haën (Germany) and sodium hydroxide and N,N-dimethylformamide (DMF) from ACE (South Africa) and were used as received. Distilled, de-ionized water was used throughout.

Ammonia–ammonium chloride buffers in the pH range 7–11 were prepared from 0.1 mol dm⁻³ ammonia solution and 0.1 mol dm⁻³ ammonium chloride in water. The pH of the solutions was adjusted by adding drops of acetic acid or 1 mol dm⁻³ sodium hydroxide. A stock standard solution of NA (10⁻⁴ mol dm⁻³) was prepared in DMF and kept in the dark. The required concentration of niclosamide in aqueous buffer solutions was then prepared from the stock standard solution.

Procedure

A 20 cm³ volume of supporting electrolyte was placed in the electrochemical cell and the required volume of standard NA solution was added to the cell with a micro-pipette. The same procedure was followed for sample analysis. The solution was deaerated with pure nitrogen (99.999%, Air Products SA). Cyclic voltammetric measurements were run from 0.200 to −0.800 V and back. Square-wave voltammetric measurements were run from −1.150 to 0.200 V using the Osteryoung square-wave voltammetric mode and the net current responses were recorded. After each electrochemical determination, the solution was stirred for 10 s prior to the next measurement. The parameters for square-wave voltammetric measurements were as follows: potential step, 7 mV; square-wave amplitude, 45 mV; and square-wave frequency, 110 Hz. All measurements were carried out at room temperature (22 ± 2 °C). The pK_a of NA in a mixture of 10% DMF (containing 1 × 10⁻⁴ mol dm⁻³ NA) and 90% water (v/v) (containing 0.1 mol dm⁻³ citric acid and 0.2 mol dm⁻³ disodium hydrogenphosphate) was determined following the spectrophotometric method of Albert and Sergeant.²⁰

Analysis of tablets

Five tablets of niclosamide (EPHARM), each weighing 650 mg and containing 500 mg of niclosamide, were ground to powder and thoroughly mixed. From the ground tablets, 42.51 mg were taken and dissolved in 100 cm³ DMF in order to obtain 1 × 10⁻³ mol dm⁻³ NA. This was diluted to 1 × 10 ⁻⁴ mol dm⁻³ with the supporting electrolyte solution. An aliquot of this solution (20 mm³) was spiked into the electrochemical cell that contained 20 cm³ of 0.1 mol dm⁻³ NH₃–NH₄Cl buffer (pH 8.5) and the voltammogram was recorded following the already outlined voltammetric procedure. The standard addition method was then applied, adding successive aliquots of 20 mm³ of 1 × 10⁻⁴ mol dm⁻³ NA DMF solution to the electrochemical cell.

Results and discussion

Electrochemical behaviour of NA at a glassy carbon electrode

Fig. 2 shows the cyclic voltammograms of 5 × 10⁻⁶ mol dm⁻³ NA at a glassy carbon electrode in pH 8.5 NH₃–NH₄Cl buffer for two continuous cycles. During the first cycle an irreversible reduction peak A appeared at −0.659 V on the cathodic sweep. On the reverse scan an oxidation peak B₁ appeared at −0.164 V. No oxidation peak was observed corresponding to peak A. During the second cathodic sweep, a new reduction wave B₂ that is reversible with B₁ was seen at −0.195 V. Thus, a pair of well-defined redox peaks, B₁ and B₂, were obtained that were separated by about 30 mV. The 30 mV peak separation indicates the involvement of two electrons in the electrode process.

Fig. 2 Cyclic voltammogram of 5 × 10⁻⁶ mol dm⁻³ NA in 0.1 mol dm⁻³ NH₃–NH₄Cl buffer (pH 8.5) at a glassy carbon electrode with scan rate 100 mV s⁻¹ and initial potential 200 mV.

Fig. 3 shows the cyclic voltammograms of NA for five repetitive cycles (experimental conditions as in Fig. 2). As the number of cycles increased the peak heights of the reversible redox couple (B₁ and B₂) increased at the expense of the irreversible reduction of peak A. This indicates that the product of the irreversible reduction of NA remained on or near the electrode surface and was oxidised on the anodic sweep. When the potential was scanned while stirring the solution, the peaks due to the redox couple disappeared. Furthermore, when the potential scan was restricted to the potential range 0.200 to −0.500 V, the reversible peaks (B₁ and B₂) disappeared completely. These phenomena indicate that the electrochemically generated product during the irreversible oxidation of NA (peak A) is responsible for the formation of peaks B₁ and B₂. It is evident from the literature that nitrophenyls undergo irreversible four-electron reduction to give N-phenylhydroxylamine.^21,22 It is also reported that p-nitrosophenol could be converted easily into p-hydroxylaminophenol reversibly via a two-electron oxidation/reduction mechanism.^23,24 Hence it is believed that the same electron transfer mechanism can be applied to the electrochemical reduction and oxidation of niclosamide.

RNO₂ + 4e⁻ + 4H⁺ → RNHOH + H₂O (A)

RNHOH ⇌ RNO + 2H⁺ + 2e⁻ (B₁/B₂)

where RNO₂ represents niclosamide.

Fig. 3 Cyclic voltammogram of 5 × 10⁻⁶ mol dm⁻³ NA for five repetitive cycles. The numbers represent the cycles. Other experimental conditions as in Fig. 2.

Cyclic voltammograms were run at various scan rates in order to characterise further the electrode reaction mechanism for the redox couple (B₁ and B₂) and to ascertain the existence of adsorption. At low scan rates (5–100 mV s⁻¹) no shift in the peak potential was detected and the current peak ratio remained constant. Table 1 shows the influence of the scan rate on the currents and potentials of the redox couple. The peak currents of the redox couple increased with increasing scan rate. The ratio of the anodic peak current to the cathodic peak current also increased with increasing scan rate. The cathodic peak (B₂) potential shifted negatively with increasing scan rate whereas the anodic peak (B₁) potential remained nearly the same. The above observed behaviour agrees with the diagnostic criteria for a mechanism of chemical reaction followed by electron transfer (ce mechanism).^22,24 It is known that arylhydroxylamines can react with nitroso compounds and yield diarylazoxy derivatives.²⁵ It is therefore reasonable to suggest that the mechanism of the redox couple follows a ce mechanism. It should be noted that the reverse cathodic peak is apparent even at very low scan rates, indicating that the rate of the electron transfer is faster than the rate of the chemical reaction.

Table 1 Influence of scan rate on the currents and potentials of the redox couple of NA. Conditions: cyclic voltammograms of 5 × 10⁻⁶ mol dm⁻³ NA in 0.1 mol dm⁻³ NH₃–NH₄Cl buffer (pH 8.5); initial potential, 0.2 V; switch potential, −0.8 V; stirring time before each run, 10 s; and equilibration time, 30 s

ν/V s^−1	ipc/μA	ipa/μA	ipa/ipc	Epc/V	Epa/V	ΔEp/mV
0.1	1.98	1.62	0.82	−0.195	−0.167	28
0.2	3.48	2.85	0.82	−0.197	−0.168	29
0.3	4.45	3.69	0.83	−0.200	−0.168	32
0.4	5.48	4.60	0.84	−0.204	−0.169	35
0.5	6.75	5.74	0.85	−0.206	−0.169	37
0.6	7.72	6.64	0.86	−0.209	−0.169	40
0.7	8.60	7.40	0.86	−0.212	−0.170	42
0.8	9.52	8.28	0.87	−0.214	−0.170	44
0.9	9.74	8.47	0.87	−0.217	−0.171	46
1.0	9.80	8.62	0.88	−0.218	−0.171	47

---

The diagnostic tests made for the existence of adsorption for the redox couple did not show the presence of strong adsorption as this can also be ascertained from the shape of the redox peaks. However, the presence of weak adsorption of the oxidised or reduced species could not be ruled out since many electrochemical reactions of organic compounds involve both coupled chemical reactions and adsorption.²²

Experiments were carried out using the square-wave voltammetric technique and Fig. 4 illustrates the square-wave voltammogram of 2 × 10⁻⁶ mol dm⁻³ NA. Two well-defined symmetrical peaks were observed at −0.660 and −0.172 V when the potential was scanned in the positive direction. However, when the potential was scanned in the negative direction, the peak for the reversible couple disappeared completely. As can be seen, the peak current at −0.172 V that is attributed to the reversible redox couple is about six times higher than the peak current corresponding to the irreversible reduction of NA at −0.660 V. Moreover, the peak current was found to be proportional to NA concentration. Hence this peak was systematically studied by square-wave voltammetry for the detection of NA.

Fig. 4 Square wave voltammogram of 2 × 10⁻⁶ mol dm⁻³ NA in 0.1 mol dm⁻³ NH₃–NH₄Cl buffer (pH 8.5) at a glassy carbon electrode with initial potential −1.200 V, final potential 0.200 V, step 4 mV, amplitude 25 mV and frequency 15 Hz.

Influence of buffer and pH of supporting electrolyte

A series of buffer solutions as supporting electrolytes were tested for their suitability in the determination of NA. These include phosphate buffer, KH₂PO₄–Na₂HPO₄; acetate buffer (CH₃COOH–CH₃COONa), borate buffer (H₃BO₃–NaBO₂) and ammonia buffer (NH₃–NH₄Cl). The peak height and shape of the voltammograms were considered for the choice of the supporting electrolytes. There was no voltammetric signal detected in an acidic buffer system. The optimum buffer solution chosen for subsequent studies was ammonia buffer.

The influence of pH on the peak current of NA was investigated over the pH range 7–11. Fig. 5(A) and (B) show the dependence of the peak current and potential on pH for square-wave voltammetry and cyclic voltammetry, respectively. In both techniques the peak current decreased beyond pH 8.7 and below pH 8.2. The maximum peak current was observed at around pH 8.5. The optimum pH adopted for further studies was 8.5.

Fig. 5 (A) Plots of peak current as function of pH (●) and peak potential as function of pH (■) for the square-wave voltammetric responses of 2 × 10⁻⁶ mol dm⁻³ NA. Other experimental conditions as in Fig. 4. (B) Plots of peak current as function of pH (●) and peak potential as function of pH (■) for the cyclic voltammetric oxidation peak of 5 × 10⁻⁵ mol dm⁻³ NA. Other experimental conditions as in Fig. 2.

The shift in the square-wave peak potential as a function of pH was studied and a linear dependence was observed [Fig. 5(A)]. A similar peak potential shift and linear dependence on pH were obtained when the pH was varied using cyclic voltammetry [Fig. 5(B)]. This indicates that the H⁺ ion takes part in the electrode reaction. According to the literature,²⁶, E_p = K − (0.059y/n)pH, where y is the number of H⁺ ions that take part in the electrode reaction and n is the number of electrons, for a reversible process. As can been seen from Fig. 5(A), there are two linear ranges, which are described by the following equations:

E_p/V = −6.50 × 10⁻²pH + 0.481; r = 0.990 (pH 7.5–8.5)

E_p/V = −2.80 × 10⁻²pH + 0.061; r = 0.999 (pH 8.5–10.0)

The dependence of the peak potential on pH has slopes of 65 and 28 mV per pH unit, respectively. Similarly, two linear ranges were obtained using the CV mode with slopes of 60 and 33 mV per pH unit, respectively [Fig. 5(B)]. This implies that the number of protons involved in the redox process changes from two to one as the pH increases (for n = 2). Electrode processes involving a weak acid or weak base have potential–pH variations which show a change in slope at pH = pK_a.²⁶ NA is a weak acid and its redox behaviour is strongly pH dependent. From the intersections of the linear parts of the plots, the pK_a of NA was estimated to be 8.5. The pK_a of NA was studied following the spectrophotometric method of Albert and Sergeant,²⁰ in buffer solutions in the pH range 2.5–9.5. The plot of ΔA/ΔpH as a function of pH gave two peaks at pH 3.8 (small peak) and pH 8.0 (large peak). The latter value is in good agreement with the pK_a value determined from the voltammetric measurements.

Voltammetric parameters

The instrumental parameters in square-wave voltammetry are interrelated and have a combined influence on the peak current.²⁷ Hence, in order to establish the optimum conditions in the determination of NA, the influence of instrumental parameters on the current response was studied.

The influence of the square-wave amplitude on the peak current was studied in the range 10–80 mV. The peak current increased sharply up to 40 mV, then reached a plateau over the range 45–80 mV. The square-wave amplitude was set at 45 mV for the subsequent measurements.

The effect of the potential step on the peak current was investigated in the range 2–16 mV. The plot of the peak current as a function of the potential step gave a bell-shaped curve with a maximum value of about 10 mV (not shown). A potential step of 10 mV was chosen for the analysis.

The square-wave frequency was varied from 5 to 150 Hz. The peak current increased with increase in the frequency. However, the shape of the voltammogram became broader as the frequency increased. A square-wave frequency of 110 Hz was chosen as the optimum value.

The influence of the initial sweep potential on the peak current was examined in the potential range −1200 to −800 mV. The peak current was found to increase linearly and rapidly with increase in the the initial potential from −800 to −1050 V. A maximum response was obtained at about −1150 V (Fig. 6). An initial sweep potential of −1150 mV was adopted for all subsequent measurements.

Fig. 6 Influence of the initial sweep potential on the peak current for the square-wave voltammetric response of 2 × 10⁻⁶ mol dm⁻³ NA, step 7 mV, amplitude 45 mV and frequency 110 Hz.

Linear range and detection limit

Under the optimum conditions, using the square-wave mode the peak current was linearly dependent on NA concentration. The square-wave voltammograms at different concentrations of NA are shown in Fig. 7. A plot of the mean value of the peak current (for five runs) as a function of the concentration of NA was drawn and the response was linear in the concentration range 5.00 × 10⁻⁸–1.00 × 10⁻⁶ mol dm⁻³ NA. For the regression plot of the peak current vs. NA concentration, the slope was 4.013 μA μM⁻¹, the y-intercept was 0.152 μA and the correlation coefficient was r² = 0.999. The detection limit (three times the signal-to-noise ratio) was 2.05 × 10⁻⁸ mol dm⁻³ NA. For eight successive determinations of 5 × 10⁻⁷ mol dm ⁻³ NA, a relative standard deviation (RSD) of 2.4% was obtained.

Fig. 7 Square wave voltammograms of NA at a glassy carbon electrode in 0.1 mol dm⁻³ NH₃–NH₄Cl buffer (pH 8.5). Concentration range, 0–6.5 × 10⁻⁶ mol dm⁻³; the difference in concentration between successive peaks is 0.05 × 10⁻⁶ mol dm⁻³. Other experimental conditions as in Fig. 6.

Interferences

The effect of the concomitants associated with NA in its pure form and its formulations were tested using the developed method. This method does not suffer any interference from commonly associated sweetening and flavouring agents used in the preparation of tablets, such as sucrose, lactose, dextrose, starch, talc, stearic acid and sodium alginate with respect to known amounts of NA. The mean recovery was 99.33%.

Selectivity and ruggedness

Selectivity is the ability to separate the analyte from metabolites, degradation products and co-administered drugs.²⁸ NA does not have known metabolites since it is not significantly absorbed from the gastrointestinal tract.²⁹ The selectivity of the method was evaluated by studying 2-chloro-4-nitroaniline and 5-chlorosalicylic acid. These compounds are disintegration products of NA that are routinely tested in the drug manufacture of NA.^2,4 5-Chlorosalicylic acid did not show any voltammetric peak in the potential range studied, whereas 2-chloro-4-nitroaniline gave an irreversible reduction peak at −0.770 V (CV mode) which is more negative than the irreversible reduction peak of NA. In the potential region where NA is determined (−0.164 to 0.2 V), 2-chloro-4-nitroaniline did not show any redox peaks. The presence of a fivefold molar excess of 2-chloro-4-nitroaniline in NA did not affect the peak current of NA. These observations indicate that the method is very selective for NA.

The effect of different analysts (two) on the results for 1 × 10⁻⁷ and 1 × 10⁻⁶ mol dm⁻³ NA standard samples was evaluated. The results showed good agreement, with RSD in the range 1.04–6.80%.

Analytical application

The proposed method was applied to the determination of niclosamide in tablets by using the standard additions method. The procedure used for the determination of niclosamide as described earlier gave a mean value of 498.6 mg of NA per tablet, which is in very good agreement with the declared value of 500 mg.

The powdered tablet of NA was treated at high temperature (80 °C) and elevated humidity for 12 h and its voltammogram was compared with that of an untreated NA tablet of the same concentration (Fig. 8). The response of the treated tablet decreased by 37% as a result of, presumably, the degradation of NA. No signal was detected due to the degraded products of NA in the potential range studied.

Fig. 8 Square wave voltammograms of NA tablets at a glassy carbon electrode in 0.1 mol dm⁻³ NH₃–NH₄Cl buffer (pH 8.5). (A) 2 × 10⁻⁶ mol dm⁻³ NA without heat and humidity treatment; (B) 2 × 10⁻⁶mol dm⁻³ NA after the powdered tablet had been treated at 80 °C and elevated humidity for 12 h. Other experimental conditions as in Fig. 6.

Conclusion

This study has shown that NA can be determined using the square-wave voltammetric technique on the basis of its reduction process at a glassy carbon electrode. The method was successfully applied to the direct determination of NA in pharmaceutical formulations with adequate reproducibility and sensitivity without any purification. A faster analysis for NA can be performed by direct measurement from the calibration curve. The method is relatively cheap and rapid in comparison with other methods. Chromatographic and HPLC methods for the determination of NA need expensive equipment and materials and also include time-consuming extraction steps to eliminate the excipients. Most spectrophotometric methods include complex reactions that cause contamination and loss of substance and have lower sensitivity than the proposed voltammetric method. The polarographic method previously reported for NA determination has a very narrow linear range and is undesirable owing to the toxic properties of mercury.

Acknowledgements

The authors are grateful for the procurement of instruments by the Ministry of Education for the National University of Lesotho, Chemistry Department. The Department of Chemistry is acknowledged for both material and financial support.

References

1. W. E. Jones, Am. J. Trop. Med. Hyg., 1979, 28, 300.
2. Remington's Pharmaceutical Sciences, Mack, Easton, PA, 16th edn., 1980, p. 1182.
3. A. Korlkovas, Essentials of Medicinal Chemistry, Wiley, New York, 2nd edn., 1988, p. 617.
4. Martindale, The Extra Pharmacopoeia, ed. J. E. F. Reynolds, Pharmaceutical Press, London, 28th edn., 1982, p. 100.
5. F. C. Churchill and D. N. Ku, J. Chromatogr., 1980, 189, 375.
6. N. Bergisadi and D. Sarigul, Acta Pharm. Turc., 1986, 28, 51.
7. S. S. Zarapkar, S. S Mehron and S. R. Rele, Indian Drugs, 1989, 26, 360.
8. F. Onur and N. Tekin, Anal. Lett., 1994, 27, 229.
9. S. S. Zarapkar and P. M. Deshpande, Indian J. Pharm. Sci., 1989, 51, 136.
10. C. P. Sastry, M. Amna and A. R. Rao, Talanta, 1988, 35, 23.
11. C. P. Sastry, M. Amna and A. R. Rao, Indian Drugs, 1988, 25, 348.
12. S. A. Fattah, Spectrosc. Lett., 1997, 30, 795.
13. H. G. Daabees, Anal. Lett., 2000, 33, 639.
14. C. G. Derek and P. Norbert, Int. J. Environ. Anal. Chem., 1980, 8, 1.
15. V. K. Dawson, Can. J. Fish. Aquat. Sci., 1982, 39, 778.
16. M. T. Schreier, K. V. Dawson and A. M. Boogaard, J. Agric. Food Chem., 2000, 48, 2212.
17. S. J. Johnson and B. P. Geoffrey, Pestic. Sci., 1979, 10, 531.
18. A. K. Mishra and K. D. Gode, Indian Drugs, 1985, 22, 317.
19. Laboratory Techniques in Electroanalytical Chemistry, ed. P. T. Kissinger and R. Heineman, Marcel Dekker, New York, 1996, p. 793.
20. A. Albert and E. P. Serjeant, The Determination of Ionization Constants, Chapman and Hall, London, 1971.
21. W. Smith and A. J. Bard, J. Am. Chem. Soc., 1975, 97, 5203.
22. Physical Methods of Chemistry, V. II Electrochemical Methods ed. B. W. Rossiter and J. F. Hamilton, Wiley-Interscience, New York, 1986, p. 859.
23. R. S. Nicholson and I. Shain, Anal. Chem., 1965, 37, 190.
24. Southampton Electrochemistry Group, Instrumental Methods in Electrochemistry, Ellis Horwood, Chichester, 1985, p. 199.
25. P. Zuman and B. Shah, Chem. Rev., 1994, 94, 1621.
26. P. H. Rieger, Electrochemistry, Prentice-Hall, Englewood Cliffs, NJ, 1987, p. 25.
27. J. H. Christie, J. A. Turner and R. A. Osteryoung, Anal. Chem., 1977, 49, 1899.
28. F. Bressolle, M. Bromet-Petit and M. Audron, J. Chromatogr., B, 1996, 686, 3.
29. Isolation and Identification of Drugs, V. II, ed. E. C. Clarke, William Clowes, London, 1978, p. 1067.

---

Footnote

† This paper was submitted for inclusion in the New Directions in Electroanalysis Guest Editor Issue of The Analyst.

---