Utilization of bromination reactions for the determination of carbamazepine using bromate–bromide mixture as a green brominating agent

2.3.1 Titrimetry

A 1.0–7.5 ml aliquot of standard CBZ solution containing 1.00–7.50 mg of CBZ was measured accurately, transferred into a 100 ml Erlenmeyer flask and the total volume was made to 10 ml with water. The solution was acidified by adding 5 ml of 5 M HCl followed by the addition of 10 ml of bromate–bromide mixture 5 mM in KBrO₃ by means of a pipette. The content was mixed well and the flask was kept aside for 10 min with occasional swirling. Then, 5 ml of 5% potassium iodide was added to the flask and the liberated iodine was titrated with 0.03 M sodium thiosulphate to a starch end point. A blank titration was performed taking 10 ml of water–methanol (3:2) mixture in place of drug solution and the drug content in the aliquot was calculated from a knowledge of the amount of KBrO₃ reacted.

2.3.2 Spectrophotometry using methyl orange (method A)

Different aliquots (0.0, 0.5, 1.0, 1.5, … , 3.0 ml) of a standard CBZ (5 μg ml⁻¹) solution were accurately transferred into a series of 10 ml calibrated flasks and the total volume was adjusted to 3.0 ml by adding adequate quantity of water. To each flask, 2 ml of 5 M HCl was added followed by 1.0 ml of bromate–bromide (12.5 μg ml⁻¹ in KBrO₃). The flasks were stoppered, content mixed and allowed to stand for 15 min with occasional shaking. Then, 1.0 ml of methyl orange solution (60 μg ml⁻¹) was added to each flask, and after 5 min, the mixture was diluted to the volume with water and mixed well. The absorbance of each solution was measured at 510 nm against a reagent blank.

2.3.3 Spectrophotometry using indigo carmine (method B)

Varying aliquots (0.0, 0.25, 0.5, 1.0, … , 3.0 ml) of a standard 20 μg ml⁻¹ CBZ solution were transferred into a series of 10 ml calibrated flasks using a micro burette and the total volume was brought to 3 ml by adding water. To each flask 2 ml of 5 M HCl and 1.0 ml of bromate–bromide mixture solution (50 μg ml⁻¹ in KBrO₃) were added. The content was mixed well and the flasks were kept aside for 15 min with intermittent shaking. Finally, 1.0 ml of indigo carmine solution (200 μg ml⁻¹) was added to each flask and the volume was adjusted to the mark with water. The absorbance of each solution was measured at 610 nm against a reagent blank.

2.3.4 Assay procedure for tablets

Ten tablets containing CBZ were accurately weighed and ground into fine powder. A portion of the powder equivalent to 100 mg of CBZ was accurately weighed into a 100 ml calibrated flask, 40 ml of methanol and 20 ml of water were added. The content was shaken for 15–20 min; the volume was diluted to the mark with water, mixed well and filtered using a Whatman No. 42 filter paper. The first 10 ml portion of the filtrate was discarded and a convenient aliquot of the subsequent portion of the filtrate was assayed by titrimetric procedure. The same tablet extract was appropriately diluted with water to get 5.0 and 20 μg ml⁻¹ w.r.t. CBZ for the assay by the spectrophotometric methods A and B, respectively.

2.3.5 Assay procedure for syrup

Five milliliters of syrup (Tegrital 100 mg/5 ml) equivalent to 100 mg of CBZ into a 100 ml calibrated flask, 40 ml of methanol and 20 ml of water were added. The content was shaken for 15–20 min; the volume was diluted to the mark with water and mixed well. The resulting solution was clear and it was assayed by titrimetric procedure. The same solution was appropriately diluted with water to get 5.0 and 20 μg ml⁻¹ w.r.t. CBZ for the assay by the spectrophotometric methods A and B, respectively.

2.3.6 Assay procedure for spiked human urine by the spectrophotometric methods

Five milliliters of CBZ free urine taken in a 125 ml separating funnel was spiked with 5.0 ml of a standard 200 μg ml⁻¹ CBZ solution, the content was mixed well and then extracted with 3 × 10 ml of dichloromethane. The dichloromethane extract was collected in a beaker containing anhydrous sodium sulphate and the water-free organic layer was transferred into a 50 ml dried calibrated flask and the solvent was evaporated on a hot water bath. The residue was dissolved in 2 ml of methanol and the volume was diluted with water up to the mark. An aliquot of the resulting solution was analyzed using the procedures of the methods A and B.

3.2.1 Selection of the solvent

Since carbamazepine is insoluble in water (Merck Index, 2006), different solvents such as ethanol, acetone and methanol were tested to prepare the stock standard solution of CBZ. Ethanol and acetone were unsuitable due to the oxidation of ethanol (Farkas et al., 1949) and bromination of acetone (Lopez-Cueto et al., 2002) by the free bromine as follows:

Among the tested solvents, methanol was found to be an ideal solvent for the preparation of the standard solution of CBZ, and the minimum ratio to get a stable solution was 2:3 (methanol:water) for 1.0 mg ml⁻¹ CBZ.

3.2.2 Optimization of reaction conditions

The reaction conditions as well as the various experimental variables providing accurate and precise results, and affecting the colour development of the measured species were carefully optimized.

3.2.3 Colour stability

The absorbance of the measured colour was constant for more than 24 and 12 h for methods A and B, respectively, even in the presence of the reaction products.

3.3.1 Quantitative data

The titrimetric procedure is applicable over the range of 1.00–7.50 mg of CBZ and the reaction stoichiometry was calculated to be (1:1) for the reaction between CBZ and KBrO₃.

In spectrophotometric procedures, the Beer’s law was obeyed over the concentration ranges of 0.25–1.50 and 0.50–6.00 μg ml⁻¹ of CBZ for methods A and B, respectively. The linear plots gave the following regression equations: $$\begin{array}{cc}\hfill & A=0.0123+0.4953C\text{for method A}\hfill \\ \hfill & A=-0.0031+0.1263C\text{for method B}\hfill \end{array}$$ where A is the absorbance and C is the concentration in μg ml⁻¹. The correlation coefficient (r) of the calibration plots of the methods A and B is calculated to be 0.9980 and 0.9997, respectively, confirming a linear increase in the absorbance with increasing the concentration of the drug. The calculated molar absorptivities are 1.21 × 10⁵ and 2.91 × 10⁴ l mol⁻¹ cm⁻¹ for method A and method B, respectively with corresponding Sandell sensitivity values of 0.0020 and 0.0081 μg cm⁻². The limits of detection (LOD) and limits of quantification (LOQ) for the proposed methods were calculated using the following equations (ICH, 2005): $$\begin{array}{cc}\hfill & \mathrm{LOD}=\frac{3.3\times \sigma }{S}\hfill \\ \hfill & \mathrm{LOQ}=\frac{10\times \sigma }{S}\hfill \end{array}$$ where σ is the standard deviation of replicate determination values under the same conditions as for the sample analysis in the absence of the analyte and S is the sensitivity, namely the slope of the calibration graph. In accordance with the above equations, the LOD values are 0.05 and 0.14 μg ml⁻¹ for methods A and B, respectively with corresponding LOQ values of 0.15 and 0.42 μg ml⁻¹.

3.3.2 Accuracy and precision

To evaluate the accuracy and precision of the methods, pure drug solution at three different levels (within the working limits) was analyzed, each determination being repeated seven times. The relative error (%) and relative standard deviation (%) were less than 3.0 and indicate the high accuracy and precision for the methods (Table 3). For a better picture of reproducibility on a day-to-day basis, a series of experiments were performed in which standard drug solution at three different levels was determined each day for 5 days with all solutions being prepared afresh each day. The day-to-day relative standard deviation values were in the range of 1.17–2.38% and represent the best appraisal of the methods in routine use. These results are also given in Table 3.

3.3.3 Selectivity

In order to evaluate the selectivity of the proposed methods for the analysis of pharmaceutical formulations, the effect of the presence of common excipients, such as talc, starch, lactose, glucose, sodium alginate, calcium gluconate and magnesium stearate was tested for possible interference in the assay by placebo blank. It was confirmed that the change in the titrant value and the absorbance with respect to the water blank was caused only by the analyte. A separate test was performed by applying the proposed methods to the determination of CBZ in a synthetic mixture and the percent recoveries of CBZ were 101.3 ± 2.13, 98.65 ± 1.96 and 97.24 ± 2.02 (n = 5) by titrimetric, spectrophotometric method A and method B, respectively, suggesting that no significant interference by the excipients in the assay of CBZ under the described optimum conditions.

3.3.4 Application to assay of pharmaceutical formulations and spiked urine sample

The proposed methods were successfully applied to the determination of CBZ in formulations (tablet and syrup) (Table 4). The results obtained were statistically compared with those of the official method (British Pharmacopoeia, 1973) by applying the Student’s t-test for accuracy and F-test for precision. The official method consisted of measurement of the absorbance of the alcoholic extract of the tablets at 285 nm. From Table 4, it is clear that the calculated t-value and F-value at 95% confidence level did not exceed the tabulated values of 2.78 and 6.39, respectively, for four degrees of freedom. The results indicated that there is no difference between the proposed methods and the official method with respect to accuracy and precision. The proposed spectrophotometric methods were also applied to the determination of CBZ in spiked human urine sample and the results are presented in Table 5. Accuracy and validity of the proposed methods were further ascertained performing recovery experiments via the standard addition procedure. When the tablet powder or syrup (pre-analyzed) spiked with known amounts of pure CBZ at three different concentration levels (50%, 100% and 150% of the quantity present in the tablet powder or syrup) was analyzed by the proposed methods, the recoveries of pure drug added were quantitative (Table 6).