Stress degradation studies on etamsylate using stability-indicating chromatographic methods

Analytica Chimica Acta 536 (2005) 49–70

Stress degradation studies on etamsylate using stability-indicating chromatographic methods Neeraj Kaul, Himani Agrawal, Abhijit Kakad, S.R. Dhaneshwar∗ , Bharat Patil Department of Quality Assurance Techniques, Bharati Vidyapeeth Deemed University, Poona College of Pharmacy, Erandwane, Pune 411038, Maharashtra, India Received 25 September 2004; received in revised form 15 December 2004; accepted 15 December 2004 Available online 30 January 2005

Abstract Two sensitive and reproducible methods are described for the quantitative determination of etamsylate in the presence of its degradation products. The first method was based on high-performance liquid chromatographic (LC) separation of the drug from its degradation products on the reversed phase, kromasil column [C18 (5-␮m, 25 cm × 4.6 mm, i.d.)] at ambient temperature using a mobile phase consisting of methanol and water (50:50, v/v). Flow rate was 0.6 ml min−1 with an average operating pressure of 180 kg cm−2 and retention (tR ) time was found to be 2.93 ± 0.05 min. Quantitation was achieved with UV detection at 305 nm based on peak area with linear calibration curves at concentration range 10–100 ␮g ml−1 . The second method was based on high-performance thin layer chromatographic (HPTLC) separation followed by densitometric measurement of spots at 305 nm. The separation was carried out on Merck HPTLC aluminium sheets of silica gel 60 F254 using toluene:methanol:chloroform (8.0:4.5:6.0, v/v/v) as mobile phase. This system was found to give compact spots for etamsylate after double development (retention factor, Rf value of 0.23 ± 0.02). The second order polynomial regression analysis data was used for the regression line in the range of 500–6000 ng spot−1 . Both methods have been successively applied to pharmaceutical formulation. No chromatographic interference from the tablet excipients was found. Both methods were validated in terms of precision, robustness, recovery and limits of detection and quantitation. The analysis of variance (ANOVA) and Student’s t-test were applied to correlate the results of etamsylate determination in dosage form by means of HPTLC and LC method. Drug was subjected to acid and alkali hydrolysis, oxidation, dry heat, wet heat treatment and photo-degradation. As the proposed methods could effectively separate the drug from its degradation products, they can be employed as stability indicating one. Moreover, the proposed LC method was utilized to investigate the kinetics of the acidic, alkaline and oxidative degradation processes at different temperatures and the apparent pseudo first order rate constant, half-life and activation energy was calculated. In addition the pH-rate profile of degradation of etamsylate in constant ionic strength buffer solutions with in the pH range 2–11 was studied. © 2005 Elsevier B.V. All rights reserved. Keywords: Etamsylate; High-performance liquid chromatography; Reversed phase; High-performance thin layer chromatography; Method validation; Quantitative analysis; Student’s t-test; Kinetics of degradation; pH-rate profile; Arrhenius plot; Activation energy

1. Introduction Etamsylate (Fig. 1a) chemically diethylamine 2,5-dihydroxybenzenesulphonate is official in British Pharmacopoeia [1]. It is a white or almost white, crystalline powder, very soluble in water, freely soluble in methanol, soluble in ethanol, practically insoluble in methylene chloride. Etamsy∗

Corresponding author. Tel.: +91 20 25437237; fax: +91 20 25439383. E-mail address: [email protected] (S.R. Dhaneshwar).

0003-2670/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2004.12.041

late possesses antihemorrhagic properties and also enhances P-selectin membrane expression in human platelets and cultured endothelial cells [2–9]. Literature survey reveals few analytical methods reported for the quantitative estimation of etamsylate. Thin layer chromatographic method has been reported in British Pharmacopoeia for quantitative estimation of etamsylate and its known impurity (benzene-1,4-diol) (Fig. 1b). Yang et al. [10] have reported chemiluminescence method for the determination of etamsylate in pharmaceutical preparation.

50

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

Gujarat, India, used without further purification and certified to contain 99.65% (w/w) on dried basis. All chemicals and reagents used were of LC grade and were purchased from Merck Chemicals, India. 2.2. Instrumentation and chromatographic conditions

Fig. 1. (a) Structure of etamsylate. (b) Structure of hydroquinone (benzene1,4-diol), a known impurity of etamsylate.

To our knowledge, no article related to the stabilityindicating chromatographic determination of etamsylate in pharmaceutical dosage form has been reported in literature. The International Conference on Harmonization (ICH) guideline entitled “Stability testing of new drug substances and products” requires that stress testing be carried out to elucidate the inherent stability characteristics of the active substance [11]. Acidic, alkaline, oxidative and photolytic stability are required. An ideal stability-indicating method is one that quantifies the standard drug alone and also resolves its degradation products. Consequently, the implementation of an analytic methodology to determine etamsylate in pharmaceutical dosage form in presence of its degradation products is a pending challenge of the pharmaceutical analysis. Therefore, it was thought necessary to study the stability of etamsylate towards acidic, alkaline, oxidative, UV and photodegradation processes. The aim of this work was to develop stability-indicating chromatographic methods for determination of etamsylate in presence of its degradation products and related impurities for assessment of purity of bulk drug and stability of its bulk dosage forms using LC and HPTLCdensitometry. The two methods are simple, accurate, specific, repeatable, stability indicating, reduces the duration of the analysis and suitable for routine determination of etamsylate in tablet dosage form. Both the proposed methods were validated in compliance with ICH guidelines [12,13] and its updated international convention [14]. Furthermore, the developed LC method was used to investigate the kinetics of the acidic, alkaline and oxidative degradation processes by quantitation of drug at different temperatures, and to calculate the activation energy and half-life for etamsylate degradation. The proposed LC method was also utilized for pH-rate profile study of degradation of etamsylate in constant ionic strength buffer solutions with in the pH range 2–11.

2. Experimental 2.1. Materials Pharmaceutical grade of etamsylate (batch no: et-S4220) was kindly supplied as a gift sample by Finecure Pharma Ltd.,

2.2.1. For LC method The LC system consisted of a pump (model jasco PU 1580, intelligent LC pump) with auto injecting facility (AS-1555 sampler) programmed at 20 ␮l capacity per injection was used. The detector consisted of a UV–vis (Jasco UV 1575) model operated at a wavelength of 305 nm. The software used was jasco borwin version 1.5, LC-Net II/ADC system. The columns used were Kromasil C-18 (250 mm × 4.6 mm, 5.0 ␮m) Flexit Jour Laborarories Pvt. Ltd., Pune, India and Finepak SIL-5, C-18 (250 mm × 4.6 mm, 5.0 ␮m) Jasco Corporation, Japan. Different mobile phases were tested in order to find the best conditions for separation of etamsylate in presence of its degradation products. The optimal composition of the mobile phase was determined to be methanol:water (50:50, v/v). The flow rate was set to 0.6 ml min−1 and UV detection was carried out at 305 nm. The mobile phase and samples were filtered using 0.45 ␮m membrane filter. Mobile phase was degassed by ultrasonic vibrations prior to use. All determinations were performed at ambient temperature. 2.2.2. For HPTLC densitometry The samples were spotted in the form of bands of width 6 mm with a Camag 100 microlitre sample (Hamilton, Bonaduz, Switzerland) syringe on precoated silica gel aluminium Plate 60 F254 , (20 cm × 10 cm) with 250 ␮m thickness; (E. Merck, Darmstadt, Germany, supplied by Anchrom Technologists, Mumbai) using a Camag Linomat IV (Switzerland, supplied by Anchrom Technologists, Mumbai). The plates were prewashed by methanol and activated at 110 ◦ C for 5 min prior to chromatography. A constant application rate of 0.1 ␮l s−1 was employed and space between two bands was 6 mm. The slit dimension was kept at 5 mm × 0.45 mm and 10 mm s−1 scanning speed was employed. The monochromator bandwidth was set at 20 nm with K 320 cut off filter, each track was scanned thrice and baseline correction was used. The mobile phase consisted of toluene:methanol:chloroform (8.0:4.5:6.0, v/v/v) and 15 ml of mobile phase was used per chromatography. Linear ascending development was carried out in 20 cm × 10 cm twin trough glass chamber (Camag, Muttenz, Switzerland). Dimensions: length × width × height = 12 cm × 4.7 cm × 12.5 cm. It was saturated (lined on the two bigger sides with filter paper that had been soaked thoroughly with the mobile phase) and the chromatoplate double development was carried out in dark with the mobile phase. The optimized chamber saturation time for mobile phase was 30 min at room temperature (25 ◦ C ± 2) at relative humidity of 60% ± 5. The length

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

of single chromatogram run was 9 cm and approximately 30 min. Subsequent to the development, TLC plates were dried in a current of air with the help of an air dryer in wooden chamber with adequate ventilation. The flow of air in the laboratory was maintained unidirectional (laminar flow, towards exhaust). Densitometric scanning was performed on Camag TLC scanner III in the reflectance–absorbance mode at 305 nm for all measurements and operated by CATS software (V 3.15, Camag). The source of radiation utilized was deuterium lamp emitting a continuous UV spectrum between 190 and 400 nm. Concentrations of the compound chromatographed were determined from the intensity of diffusely reflected light. Evaluation was via peak areas with second order polynomial regression. The shape of calibration curves in thin layer chromatography is generally inherently non-linear due to scattering of light. Calibration curves generally comprise a pseudolinear region at low sample concentration and then departure from linearity begins at higher sample concentrations. The extent of individual ranges of the calibration curves is frequently very different for different substances. In some instances, the pseudolinear range may be adequate for most analytical purposes, in others no reasonable linear range exist [15]. Scattering of light is highly dependent on the type of the TLC plate, measuring wavelength, measuring

Fig. 2. Chromatogram of standard etamsylate (50 ␮g ml−1 ); (tR : 2.93 ± 0.05) measured at 305 nm, mobile phase: methanol and water (50:50, v/v).

51

mode, molar absorptivity and concentration of the sample. The use of HPTLC plates is therefore advantageous since they are less scattering than conventional TLC plates [16]. 2.3. Standard solutions and calibration graphs Stock standard solution was prepared by dissolving 600 mg of etamsylate in 100 ml methanol (6000 ␮g ml−1 ). 2.3.1. For LC method The standard solutions were prepared by dilution of the stock solution with methanol to reach a concentration range 10–100 ␮g ml−1 for etamsylate. Injection of 20 ␮l by triplicate were made six times for each concentration and chromatographed under the conditions described above. The peak areas were plotted against the corresponding concentrations to obtain the calibration graphs. 2.3.2. For HPTLC-densitometric method The standard solutions were prepared by dilution of the stock solution with methanol to reach a concentration range 500–6000 ng ␮l−1 . One microlitre from each standard solution was spotted on the HPTLC plate to obtain final concentration 500–6000 ng spot−1 . Each concentration was spotted six times on the HPTLC plate. The plate was developed on previously described mobile phase. The peak areas were plot-

Fig. 3. Densitogram of standard etamsylate (2000 ng spot−1 ); peak 1 (Rf : 0.23 ± 0.02), mobile phase toluene:methanol:chloroform (8.0:4.5:6.0, v/v/v).

52

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

Table 1 Reproducibility of run time (single development)a

the same LC system under the same conditions using linear regression equation.

Plate conditionb

Run timec (min)

S.D.

R.S.D. (%)

Blank plate Plate spotted with standards Plate spotted with standard along with degraded products

29.94 30.54 31.08

0.45 0.78 1.11

1.32 1.80 1.94

2.4.2. For HPTLC-densitometric method The above stock solution was further diluted to obtain sample solution at three different concentration levels of 2000, 4000 and 6000 ng ␮l−1 , respectively. One microlitre of each sample solution was applied six times to the HPTLC plate to give concentration 2000, 4000, 6000 ng spot−1 for etamsylate. The plate was developed in the previously described chromatographic conditions. The peak area of the spots were measured at 305 nm and concentrations in the samples were determined using multilevel calibration developed on the same plate under the same conditions using second order polynomial regression equation.

a

n = 6. Plates pre-treated with methanol and activated at 110 ◦ C. c Development was performed in the ascending direction at constant run length of 9 cm. b

Table 2 Linear regression data for calibration curves (n = 6) Parameters

TLC densitometry

LC

Linearity range r2 ± S.D. Slope ± S.D. Intercept ± S.D. Confidence limit of slopea Confidence limit of intercepta S.E. of estimation

1000–6000 ng spot−1

10–60 ␮g ml−1 0.9997 ± 0.98 0.42 ± 0.10 2.03 ± 0.18 0.34–0.50 1.88–2.17 0.08

a

0.9998 ± 0.70 0.12 ± 0.04 5.65 ± 1.06 0.09–0.15 4.80–6.49 1.04

2.5. Method validation 2.5.1. Precision Precision of the method was determined with the product. An amount of the product powder equivalent to 100% of the label claim of etamsylate was accurately weighed and assayed. System repeatability was determined by six replicate applications and six times measurement of a sample solution at the analytical concentration. The repeatability of sample application and measurement of peak area for active compound were expressed in terms of % relative standard deviation (R.S.D.) and standard error (S.E.). Method repeatability was obtained from R.S.D. value by repeating the assay six times in same day for intra-day precision. Intermediate precision was assessed by the assay of two, six sample sets on different days (inter-day precision). The intra-day and inter-day variation for determination of etamsylate was carried out at three different concentration levels 2000, 4000, 6000 ng spot−1 and 20, 40, 60 ␮g ml−1 for HPTLC and LC, respectively.

95% confidence limit.

ted against the corresponding concentrations to obtain the calibration graphs. 2.4. Sample preparation To determine the content of etamsylate in conventional tablets (label claim: 250 mg etamsylate per tablet), the 20 tablets were weighed, their mean weight determined and they were finely powdered and powder equivalent to 1000 mg etamsylate was weighed. Then equivalent weight of the drug was transferred into a 100 ml volumetric flask containing 50 ml methanol, sonicated for 30 min and diluted to 100 ml with methanol. The resulting solution was centrifuged at 3000 rpm for 5 min. Supernatant was taken and after suitable dilution the sample solution was then filtered using 0.45-␮m filter (Millipore, Milford, MA).

2.5.2. Robustness of the method 2.5.2.1. For LC method. To evaluate LC method robustness a few parameters were deliberately varied. The parameters included variation of C-18 columns from different manufacturers, flow rate, percentage of methanol in the mobile phase, column temperature and methanol of different lots. Two analytical columns, One (Kromasil C-18 column) from Pune, India and the other (Finepak C-18 column) from Japan, were used during the experiment. Robustness of the method was done at three different concentration levels 20, 40, 60 ␮g ml−1 for etamsylate, respectively.

2.4.1. For LC method The above stock solution was further diluted to get sample solutions at three different concentrations of 20, 40 and 60 ␮g ml−1 respectively. A 20-␮l volume of each sample solution was injected into LC, six times, under the conditions described in Section 2.2.1. The peak area of the spots were measured at 305 nm and concentrations in the samples were determined using multilevel calibration developed on Table 3 Intra- and inter-day precisiona (n = 6) HPTLC densitometry

LC

Intra-day precision

Inter-day precision

Intra-day precision

Inter-day precision

S.D. of areas

% R.S.D.

S.E.

S.D. of areas

% R.S.D.

S.E.

S.D. of areas

% S.D.

S.E.

S.D. of areas

% S.D.

S.E.

1.42

1.28

0.51

1.89

1.58

0.81

1.52

1.21

0.68

1.78

1.46

0.72

a

Average of three concentrations 2000, 4000,

6000 ng spot−1

and 20, 40,

60 ␮g ml−1

for HPTLC and LC, respectively; 95% confidence limit.

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

53

different concentration levels 2000, 4000, 6000 ng spot−1 for etamsylate.

2.5.2.2. For HPTLC-densitometric method. By introducing small changes in the mobile phase composition, the effects on the results were examined. Mobile phases having different composition like toluene:methanol:chloroform (7.5:5.0:6.0, v/v/v), (8.5:4.0:6.0, v/v/v), (7.5:4.5:6.5, v/v/v), (8.5:4.5:5.5, v/v/v), (8.0:4.0:6.5, v/v/v), (8.0:5.0:5.5, v/v/v) and so on were tried and chromatograms were run. The amount of mobile phase, temperature and relative humidity was varied in the range of ±5%. The plates were prewashed by methanol and activated at 60 ◦ C ± 5 for 2, 5, 7 min, respectively prior to chromatography. Time from spotting to chromatography and from chromatography to scanning was varied from 0, 20, 40 and 60 min. Robustness of the method was done at three

2.5.3. Limit of detection and limit of quantitation The detection limit of an individual analytical procedure is the lowest amount of analyte in a sample that can be detected but not necessarily quantitated as an exact value. The quantitation limit of an individual analytical procedure is the lowest amount of analyte in a sample that can be quantitatively determined with suitable precision and accuracy. The quantitation limit is a parameter of quantitative assays for low levels of compounds in sample matrices, and is used partic-

Table 4 (a) Robustness evaluationa of the LC method (n = 6); (b) robustness testingb of HPTLC-densitometric method Chromatographic changes Factorc

Level

tR d

ke

Tf

−1 0 1

3.00 2.93 2.85 2.93 ± 0.07

2.18 2.15 2.12 2.15 ± 0.03

1.41 1.38 1.26 1.35 ± 0.07

B: Percentage of methanol in the mobile phase (v/v) 48 −1 50 0 52 −1 Mean ± S.D. (n = 6)

2.99 2.93 2.90 2.94 ± 0.04

2.12 2.15 2.17 2.15 ± 0.03

1.41 1.38 1.24 1.34 ± 0.09

C: Temperature 24 25 26 Mean ± S.D. (n = 6)

2.95 2.93 2.90 2.93 ± 0.02

2.19 2.15 2.11 2.15 ± 0.04

1.43 1.38 1.36 1.39 ± 0.04

D: Columns from different manufacturers Kromasil Finepak Mean ± S.D. (n = 6)

2.95 2.93 2.94 ± 0.01

2.15 2.17 2.16 ± 0.01

1.38 1.37 1.38 ± 0.01

E: Solvents of different lots First lot Second lot Mean ± S.D. (n = 6)

2.92 2.93 2.93 ± 0.01

2.15 2.18 2.16 ± 0.02

1.38 1.34 1.36 ± 0.03

(a) A: Flow rate (ml min−1 ) 0.55 0.60 0.65 Mean ± S.D. (n = 6)

−1 0 1

Parameter

S.D.g of peak area

% R.S.D.g

(b) Mobile phase composition Amount of mobile phase Temperature Relative humidity Plate pre-treatment Time from spotting to chromatography Time from chromatography to scanning

1.75 1.64 1.16 1.54 0.83 0.60 0.65

1.26 1.32 0.83 1.08 0.60 0.40 0.48

a b c d e f g

Average of three concentrations 20, 40, 60 ␮g ml−1 . n = 6. Four factors were slightly changed at three levels (1, 0, −1); each time a factor was changed from level (0) the other factors remained at level (0). Retention time. Capacity factor. Tailing factor. Average of three concentrations 2000, 4000, 6000 ng spot−1 .

54

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

presence of its degradation products along with other parameters like retention time (tR ), capacity factor (k), tailing or asymmetrical factor (T) etc. Further the specificity of the method was also demonstrated with the help of photodiode array detector (JASCO MD-2010), which employed a 16 ␮l high-pressure flow cell and 10 mm path length to evaluate the peak purity of the pure drug. 2.5.4.2. For HPTLC-densitometric method. The specificity of the method was ascertained by analyzing standard drug and sample. The spot for etamsylate in sample was confirmed by comparing the Rf and spectra of the spot with that of standard. The peak purity of etamsylate was assessed by comparing the spectra at three different levels, i.e., peak start (S), peak apex (M) and peak end (E) positions of the spot. 2.5.5. Recovery studies For both methods recovery studies was carried out by applying the method to drug sample to which known amount of etamsylate corresponding to 80, 100 and 120% of label claim had been added (standard addition method). At each level of the amount six determinations were performed and the results obtained were compared with expected results.

Fig. 4. In situ spectrum of standard and sample etamsylate measured from 190 to 450 nm (r = 0.9997).

2.6. Forced degradation of etamsylate

ularly for the determination of impurities and/or degradation products.

A stock solution containing 100 mg etamsylate in 100 ml methanol was prepared. This solution was used for forced degradation to provide an indication of the stabilityindicating property and specificity of proposed method. In all degradation studies the average peak area of etamsylate after application (100 ␮g ml−1 for LC and 5000 ng spot−1 for HPTLC) of seven replicates was obtained. In order to study the degradation products of etamsylate using HPTLC method most of the study has been carried out by single development of the plate in order to prevent the movement of the non-polar degradates to extreme end of the plate.

2.5.3.1. For LC method. The limit of detection (LOD) and limit of quantitation (LOQ) were separately determined at a signal-to-noise ratio (S/N) of 3 and 10. LOD and LOQ were experimentally verified by diluting known concentrations of etamsylate until the average responses were approximately 3 or 10 times the standard deviation of the responses for six replicate determinations. 2.5.3.2. For HPTLC-densitometric method. In order to estimate the limit of detection (LOD) and limit of quantitation (LOQ), blank methanol was spotted six times following the same method as explained above. The signal-to-noise ratio (S/N) of 3 and 10 was determined for six replicate determinations.

Table 6 Stability of etamsylate in sample solutions (n = 6)

2.5.4. Specificity 2.5.4.1. For LC method. The specificity of the LC method was determined by the complete separation of etamsylate in

Parameter

HPTLC densitometrya

HPLCb

S.D. of area % R.S.D. S.E.

1.46 1.01 0.38

1.12 0.80 0.21

a b

Average of three concentrations 2000, 4000, 6000 ng spot−1 . Average of three concentrations 20, 40, 60 ␮g ml−1 .

Table 5 Standard addition technique for determination of etamsylate by HPTLC densitometry and LC (n = 6) HPTLC densitometry

LC

Excess drug added to the analyte (%)

Theoretical content (ng)

Recovery (%)

% R.S.D.

S.E.

Excess drug added to the analyte (%)

Theoretical content (␮g)

Recovery (%)

% R.S.D.

S.E.

0 80 100 120

2500 4500 5000 5500

99.87 100.12 101.01 99.64

1.51 1.12 1.78 1.45

1.20 1.08 1.25 1.16

0 80 100 120

25 45 50 55

100.25 101.74 100.11 100.54

1.44 1.32 1.98 1.74

1.12 1.07 1.65 1.54

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

2.6.1. Preparation of acid and base induced-degradation product To 10 ml of methanolic stock solution, 10 ml each of 5 M HCl and 0.1 M NaOH were added separately. These mixtures were refluxed for 2 h at 70 ◦ C. The forced degradation in acidic and basic media was performed in the dark in order to exclude the possible degradative effect of light. For LC study the resultant solution was diluted to obtain 100 ␮g ml−1 solution and 20 ␮l were injected into the system. For HPTLC study 10 ␮l of the resultant solutions (5000 ng spot−1 ) were applied on HPTLC plate and the chromatograms were run as described in Section 2.2. 2.6.2. Preparation of hydrogen peroxide induced-degradation product To 10 ml of methanolic stock solution, 10 ml of 6.0% (w/v) and 50% (w/v) of hydrogen peroxide was added. The solution was refluxed for 2.0 h at 70 ◦ C and then heated in boiling water bath for 10 min to remove completely the excess of hydrogen peroxide. For LC study the resultant solution was diluted to obtain 100 ␮g ml−1 solution and 20 ␮l were injected into the system. For HPTLC study 10 ␮l of the resultant solutions (5000 ng spot−1 ) were applied on TLC plate and the chromatograms were run as described in Section 2.2. 2.6.3. Dry heat and wet heat degradation product The standard drug was placed in oven at 100 ◦ C for 2 h to study dry heat degradation and the stock solution was refluxed for 2.0 h on boiling water bath for wet heat degradation. 2.6.4. Photochemical and UV degradation product The photochemical stability of the drug was also studied by exposing the stock solution (1 mg ml−1 ) to direct sunlight for 45 days on a wooden plank and kept on terrace. For UV degradation study, the solution was exposed to UV radiation for 7 days. For LC study, the solution was diluted to 100 ␮g ml−1 and then 20 ␮l of the solution was injected into the system. Five microlitres of the solution (5000 ng spot−1 ) was applied on HPTLC plate and chromatograms were run as described in Section 2.2. Table 7 Applicability of the proposed methods for the determination of etamsylate in commercial tablets (n = 6) Parameters

HPTLC densitometryb

LCc

Label claim (mg/tablet) Drug content (%) ± S.D. % R.S.D. S.E. t-valuea F-valuea

250 100.75 ± 1.44 0.64 0.45 0.76 2.48

250 100.84 ± 1.65 1.03 0.88 0.95 2.94

a The theoretical values for t- and F-values are equal to 2.57 and 5.05, respectively (P = 0.05). b Average of three concentrations 2000, 4000, 6000 ng spot−1 . c Average of three concentrations 20, 40, 60 ␮g ml−1 .

55

2.6.5. Neutral hydrolysis To 10 ml of methanolic stock solution, 10 ml double distilled water was added and the mixture was refluxed for 2 h at 70 ◦ C to study the degradation under neutral conditions. 2.7. Detection of the related impurities The stock solution of hydroquinone (known impurity) was prepared by dissolving 100 mg in 100 ml methanol. The above solution was appropriately diluted to get 100 ␮g ml−1 . Triplicate 20 ␮l of the above solution was injected into the LC system. Also from the stock solution 1 ␮l was spotted on the HPTLC plate to get the concentration of 1000 ng spot−1 and the plate was developed. To determine the related unknown impurity using LC, triplicate 20 ␮l of sample solution (500 ␮g ml−1 ) and standard solution (5 ␮g ml−1 ) were injected and their respective areas were correlated. The related unknown impurity was determined in HPTLC by spotting higher concentrations of the drug so as to detect and quantify them. Etamsylate (2000 mg) was dissolved in 100 ml of methanol, and this solution was termed as sample solution (20 mg ml−1 ). Two millilitres of the sample solution was diluted to 100 ml with methanol and this solution was termed as standard solution (0.4 mg ml−1 ). One microlitre of both the standard (400 ng spot−1 ) and the sample solution (20,000 ng spot−1 ) were applied on HPTLC plate and the chromatograms were run as described in Section 2.2. To confirm the specificity of the developed chromatographic methods, standard (etamsylate) stock solution and hydroquinone (known impurity) were mixed in the ratio of 1:1. The mixture was then analyzed using the proposed methods. Table 8 Summary of validation parameters: statistical data for the calibration graphs of etamsylate by HPTLC densitometric and LC method (n = 6) Parameter

HPTLC densitometric

LC

Linearity range Correlation coefficient Limit of detection Limit of quantitation Recovery (n = 6)

1000–6000 ng spot−1 0.9998 ± 0.70

10–60 ␮g ml−1 0.9997 ± 0.98

100 ng spot−1 400 ng spot−1

0.10 ␮g ml−1 0.50 ␮g ml−1

100.16 ± 0.59

100.66 ± 0.74

1.82

–

Precision (% R.S.D.) Repeatability of applicationa Repeatability of measurementa Inter-day (n = 6) Intra-day (n = 6)

1.47

–

1.58 1.28

1.46 1.21

Robustness Specificity

Robust 0.9997

Robust 2.93 ± 0.05

a

Three concentrations, three replicates each.

56

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

2.8. Kinetic investigation Accurately weighed 100 mg of drug was dissolved in 100 ml methanol. Fifteen millilitres aliquots by separate of this standard solution were transferred into separate 100 ml of double neck round bottom flask and mixed, respectively with 15.0 ml of 0.1 M NaOH, 5 M HCl and 50% (w/v) hydrogen peroxide to get final concentration of 0.75 mg ml−1 . The flasks were refluxed at different temperatures (40, 50, 60, 70, 80 and 90 ◦ C) for acidic, basic and oxidative degradation for different time intervals. At the specified time the contents of the flask were neutralized to pH 7.0 using predetermined volumes of 0.1 M HCl and 5.0 M NaOH and for oxidative degradation the excess of hydrogen peroxide was removed by heating on water bath. The contents of the flask were quantitatively transferred to 50 ml volumetric flasks with the help of microsyringe and appropriately diluted to volume with methanol and estimated by LC method by one point standardization using external standard. Each experiment was

repeated three times at each temperature and time interval. Aliquots of 20 ␮l of each solution were chromatographed under the conditions described above and the concentration of the remaining etamsylate was calculated at each temperature and at time interval for the three replicates. Data was further processed and degradation kinetics constants were calculated. 2.9. pH-rate profile Accurately weighed 100 mg of etamsylate were transferred into 100 ml volumetric flask and diluted to volume with constant ionic strength buffer solutions prepared as per Indian Pharmacopoeia [23]. The pH values of buffer solutions used for the measurement of the pH-rate profile of the degradation of etamsylate were as follows, pH 1.8, 2.8, 3.8, 4.6, 5.7, 6.8, 8.0, 9.2, 9.7 and 10.8. The pH values of these buffer solutions were checked before and after the reaction and were unchanged. The ionic strength of these buffer solutions was

Table 9 Two-way ANOVA test of etamsylate in six independent samples in duplicate by HPTLC and LC Sample

1 2 3 4 5 6

HPTLCa

LCa

First sampling

Second sampling

First sampling

Second sampling

98.14 99.90 99.28 101.14 99.48 98.46

100.81 98.99 99.41 101.20 99.84 100.45

100.24 98.89 101.34 99.40 100.18 99.75

99.18 100.24 98.15 100.78 101.24 101.45

Anova: two-factor with replication HPTLC

LC

Total

Set 1 Count Sum Average Variance

6 596.4 99.4 1.15392

6 599.8 99.96666667 0.706706667

12 1196.2 99.68333333 0.933315152

Set 2 Count Sum Average Variance

6 600.7 100.1166667 0.722146667

6 601.04 100.1733333 1.644146667

12 1201.74 100.145 1.076463636

Total Count Sum Average Variance

12 1197.1 99.75833333 0.992833333

12 1200.84 100.07 1.080218182

ANOVA Source of variation

SS

d.f.

MS

F

P-value

Fcrit

Sample Columns Interaction Within

1.278816667 0.582816667 0.39015 21.1346

1 1 1 20

1.278816667 0.582816667 0.39015 1.05673

1.21016406 0.551528457 0.369205

0.284364917 0.466327734 0.550275639

4.3512500 4.3512500 4.3512500

Total

23.38638333

23

a

The results are presented as (%) of declared amount of etamsylate per tablet. Fstat < Fcrit .

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

adjusted with sodium chloride. Separate 10 ml aliquots of the buffer solution containing etamsylate (200 ␮g ml−1 ) were transferred into separate stoppered round bottomed flasks. The flasks were then refluxed at 70 ◦ C for different time intervals. At the specified time interval the contents of the flasks were neutralized to pH 7.0 using 1 M NaOH or 1 M HCl solution. The contents of the flasks were transferred into 100 ml volumetric flasks and diluted to volume with mobile phase. Aliquots of 20 ␮l of each solution were chromatographed under the conditions described above and the concentration of the remaining etamsylate was calculated at each pH value and time interval.

3. Results and discussion 3.1. Optimization of procedures 3.1.1. Optimization of LC method The LC procedure was optimized with a view to develop a stability-indicating assay method. Pure drug along with its degraded products were injected and run in different solvent systems. Initially methanol and water in different ratios were tried. It was found that methanol:water in ratio of 50:50 (v/v) gave acceptable retention time (tR = 2.93 min) at the flow rate of 0.6 ml min−1 and drug showed typical peak nature and peak were symmetrical at 305 nm (Fig. 2). Tailing factor for etamsylate peak was less than 2% and the resolution of standard in presence of degradation products was satisfactory. Ultimately mobile phase consisting of methanol and water

57

(50:50, v/v) was selected for validation purpose and stability studies. 3.1.2. Optimization of HPTLC-densitometric method Initially toluene and methanol was tried in the ratio of 5.0:5.0 (v/v). The developed spots lack compactness, were diffused and Rf was considerably high. It was found that toluene:methanol in the ratio of 8.0:4.5 (v/v) gave compact spot but Rf was very low. To the above mobile phase different volumes of acetone was added to increase the Rf value. But this lead to poor spot characteristic along with dragging. Acetone was replaced by acetonitrile but still dragging was observed in upward direction. Then acetonitrile was replaced by different volumes of chloroform and it was found that addition of 6 ml of chloroform to the above mixture of toluene:methanol (8.0:4.5, v/v) resulted in excellent spot characteristic (compact spot) and typical peak nature but the Rf was less. Double development of the HPTLC plate in the same mobile phase was carried out in order to increase the Rf value. The final mobile phase consisting of toluene:methanol:chloroform in the ratio of (8.0:4.5:6.0, v/v/v) was optimized and good resolution with Rf value of 0.23 ± 0.02 for etamsylate was obtained when densitomet-

Table 10 Average results of etamsylate determination by HPTLC and LC and their correlation by paired t-test Sample

HPTLCa

LCa

1 2 3 4 5 6

98.29 100.88 99.45 101.14 99.26 98.78

101.21 99.68 100.71 98.47 100.28 100.46

t-Test: paired two sample for means

Mean Variance Observations Pearson correlation Hypothesized mean difference d.f. t-Stat P(T ≤ t) one-tail t-Critical one-tail P(T ≤ t) two-tail t-Critical two-tail t-Stat < t-Critical a

98.29

101.21

99.902 1.09112 5 −0.861934357 0 4 −0.021500543 0.491938073 2.131846486 0.983876145 2.776450856

99.92 0.80135 5

The results are presented as (%) of declared amount of etamsylate per tablet.

Fig. 5. Chromatogram of acid (5 M HCl, reflux for 48.0 h, temperature 70 ◦ C) treated etamsylate (100 ␮g ml−1 ); peak 1 (etamsylate) (tR : 2.93 min), peak 2 (degraded) (tR : 4.70 min), peak 3 (degraded) (tR : 5.00 min).

58

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

ric scanning was performed at 305 nm (Fig. 3). The spot appeared more compact and peak shape more symmetrical when the HPTLC plates were pre-treated with methanol and activated at 100 ◦ C for 5 min. Well-defined spots of standard along with its degradation products were obtained when the chamber saturation time was optimized at 30 min at room temperature. It was required to eliminate the edge effect and to avoid unequal solvent evaporation losses from the developing plate that can lead to various types of random behavior usually resulting in generally lack of reproducibility in Rf values (Table 1). 3.2. Linearity Etamsylate showed good correlation coefficient in concentration range of 500–6000 ng spot−1 (r2 = 0.9998 ± 0.70) and 10–100 ␮g/ml (r2 = 0.9997 ± 0.98) for HPTLC and LC, respectively. Linearity was evaluated by determining six standard working solutions containing 1000–6000 ng spot−1 and 10–60 ␮g ml−1 twice in triplicate for HPTLC and LC, respectively (Table 2). For LC method the linearity of calibration graphs and adherence of the system to Beer’s law was validated by high value of correlation coefficient and the S.D. for intercept value was less than 2%. The Y-intercept for linear regressions of HPTLC methods are often non-linear as would be expected with LC methods. Unlike LC methods, for which linearity of detector response over a wide range of concentrations of analyte can be ob-

Fig. 6. (a) Chromatogram of base (0.1 M NaOH, reflux for 2.0 h, temperature 70 ◦ C) treated etamsylate (100 ␮g ml−1 ); peak 1 (etamsylate) (tR : 2.93 min), peak 2 (degraded) (tR : 3.55 min). (b) Chromatogram of base (0.1 M NaOH, reflux for 6.0 h, temperature 70 ◦ C) treated etamsylate (100 ␮g ml−1 ); peak 1 (etamsylate) (tR : 2.93 min), peak 2 (degraded) (tR : 3.55 min), peak 3 (degraded) (tR : 4.72 min).

Fig. 7. Densitogram of acid (5 M HCl, reflux for 24.0 h, temperature 70 ◦ C) treated etamsylate (5000 ng spot−1 ); peak 1 (standard) (Rf : 0.23), peak 2 (degraded) (Rf : 0.33), peak 3 (degraded) (Rf : 0.48).

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

tained, the calibration curve of UV detector response versus a wide range of concentration for HPTLC often does not follow linear regression but rather polynomial regression. With HPTLC, the analyte interact with the layer surface of the stationary phase where scattering and absorption tend to take place, especially with high concentrations of analyte [17]. These combined processes are not adequately described by Beer–Lambert law, but the Kubelka Munk model [18]. For HPTLC method calibration graph was found to be linear that is adherence of the system to Kubelka Munk theory, which relies on the idea that light is travelling in all directions simultaneously within the precoated HPTLC plate. This is approximated as a flux of light travelling upwards and a flux travelling downwards at any depth in the plate. When this flux passes through a thin layer of material, some of it passes through, some of it is scattered backwards and some of it is absorbed. The relationship between the concentration of the etamsylate and peak area of its spot was investigated. The linear relationship was tested and found to be unaccepted due to poor correlation. The second order polynomial fit was found to be more suitable. The second order polynomial regression equation was found to be: A = 5.65 + 0.12 C − 2.49 × 10−4 C, where A is the peak area of the spot and C is the concentration of etamsylate in ng spot−1 . For both proposed methods no significant difference was observed in the slopes of standard curves (ANOVA; p < 0.05).

Fig. 8. In situ spectrum of standard (λmax = 305 nm) and its major acid degraded product (peak 3) (λmax = 288 nm) measured from 190 to 450 nm.

59

3.3. Precision 3.3.1. For LC method The within-run precision and between-run precision of the proposed LC method were determined by assaying the tablets in six times per day for consecutive six days and expressed as % R.S.D. The intra-day and inter-day precision has been depicted in Table 3. 3.3.2. For HPTLC-densitometric method The repeatability of sample application and measurement of peak area were expressed in terms of % R.S.D. and were found to be 1.82 and 1.47 for etamsylate, respectively. The % R.S.D. values depicted in Table 3 shows that proposed method provides acceptable intra-day and inter-day variation of etamsylate. 3.4. Robustness of the method 3.4.1. For LC method Each factor selected (except columns from different manufacturers and solvents of different lots) to examine were charged at three levels (−1, 0 and 1). One factor at the time was changed to estimate the effect. Thus, replicate injections (n = 6) of mixed standard solution at three concentration levels were performed under small changes of six chromatographic parameters (factors). Results, presented in Table 4a indicate that the selected factors remained unaffected by small variations of these parameters. The results from the

Fig. 9. Densitogram of base (0.1 M NaOH, reflux for 2.0 h, temperature 70 ◦ C) treated etamsylate (5000 ng spot−1 ); peak 1 (base spot) (Rf : 0.04), peak 2 (standard) (Rf : 0.23).

60

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

Fig. 11. Densitogram of hydrogen peroxide (6.0% (w/v), reflux for 2.0 h, temperature 70 ◦ C) treated etamsylate after single development; peak 1 (etamsylate) (Rf : 0.13), peak 2 (degraded) (Rf : 0.33).

two columns indicated that there is no significant difference between the results from the two columns. It was also found that methanol of different lots from the same manufacturer had no significant influence on the determination. Insignificant differences in peak areas and less variability in retention time were observed. 3.4.2. For HPTLC-densitometric method The standard deviation of peak areas was calculated for each parameter and % R.S.D. was found to be less than 2%. The low values of % R.S.D. as shown in Table 4b indicated robustness of the method. 3.5. LOD and LOQ 3.5.1. For LC method The signal-to-noise ratios 3:1 and 10:1 were considered as LOD and LOQ, respectively. The LOD and LOQ were found to be 0.10 and 0.50 ␮g ml−1 , respectively for etamsylate. 3.5.2. For HPTLC-densitometric method The LOD and LOQ were found to be 100 and 400 ng spot−1 , respectively for etamsylate. Fig. 10. (a) Chromatogram of hydrogen peroxide (6.0% (w/v), reflux for 2.0 h, temperature 70 ◦ C) treated etamsylate (100 ␮g ml−1 ); peak 1 (etamsylate) (tR : 2.93 min), peak 2 (degraded) (tR : 3.78 min). (b) Chromatogram of hydrogen peroxide (50.0% (w/v), reflux for 2.0 h, temperature 70 ◦ C) treated etamsylate (100 ␮g ml−1 ); peak 1 (etamsylate) (tR : 2.93 min), peak 2 (degraded) (tR : 3.78 min).

3.6. Specificity 3.6.1. For LC method The specificity of the LC method is illustrated in Figs. 5, 6, 10, 12, 16 and 19, where complete separation

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

61

Fig. 14. Densitogram of sunlight (45 days) treated etamsylate; peak 1 (etamsylate) (Rf : 0.23), peak 2 (degraded) (Rf : 0.53).

Fig. 12. Chromatogram of wet heat (reflux for 2.0 h, temperature 70 ◦ C) treated etamsylate (100 ␮g ml−1 ); peak 1 (etamsylate) (tR : 2.93 min), peak 2 (degraded) (tR : 3.88 min), peak 3 (degraded) (tR : 4.46 min).

of etamsylate in presence of its degradation products was noticed. The average retention time ± standard deviation for etamsylate was found to be 2.93 ± 0.05, for six replicates. The peaks obtained were sharp and have clear baseline sep-

Fig. 13. Densitogram of wet heat (reflux for 1.0 h, temperature 100 ◦ C) treated etamsylate after single development; peak 1 (etamsylate) (Rf : 0.13), peak 2 (degraded) (Rf : 0.18), peak 3 (degraded) (Rf : 0.22), peak 4 (degraded) (Rf : 0.48), peak 5 (degraded) (Rf : 0.53), peak 6 (degraded) (Rf : 0.63).

Fig. 15. Densitogram of neutral (double distilled water, reflux for 2.0 h, temperature 70 ◦ C) treated etamsylate (3000 ng spot−1 ) after double development; peak 1 (degraded) (Rf : 0.09), peak 2 (standard) (Rf : 0.23).

62

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

aration. The photodiode array detector scanned all the components present in mixture in whole wavelength range from 200 to 400 nm and it indicated that there is no degradation peak (hiding) under or unresolved from the analyte peak (pure drug), which also reflected the specificity of the method.

3.6.2. For HPTLC-densitometric method The peak purity of etamsylate was assessed by comparing their respective spectra at peak start, peak apex and peak end positions of the spot i.e., r(S, M) = 0.9997 and r(M, E) = 0.9996. Good correlation (r = 0.9997) was also obtained between standard and sample spectra of etamsylate (Fig. 4).

3.7. Recovery studies Both the proposed methods when used for extraction and subsequent estimation of etamsylate from pharmaceutical dosage form after spiking with additional drug afforded recovery of 98–102% and mean recovery for etamsylate from the marketed formulation are listed in Table 5.

3.8. Stability in sample solution 3.8.1. For LC method Three different concentrations of etamsylate, 20, 40 and 60 ␮g ml−1 were prepared from sample solution and stored at room temperature for 3 days. They were then injected into the LC system and no additional peak was found in the chromatogram indicating the stability of etamsylate in the sample solution (Table 6).

3.8.2. For HPTLC-densitometric method Solutions of three different concentrations, 2000, 4000 and 6000 ng spot−1 for etamsylate were prepared from sample solution and stored at room temperature for 0.5, 1.0, 2.0, 4.0 and 24.0 h, respectively. They were then applied on the same HPTLC plate, after development the densitogram was evaluated as listed in Table 6 for additional spots if any. There was no indication of compound instability in the sample solution.

3.8.2.1. Spot stability. The time the sample is left to stand on the solvent prior to chromatographic development can influence the stability of separated spots and are required to be investigated for validation [19]. Two-dimensional chromatography using same solvent system was used to find out any decomposition occurring during spotting and development. In case, if decomposition occurs during development, peak(s) of decomposition product(s) shall be obtained for the analyte both in the first and second direction of the run. No decomposition was observed during spotting and development.

3.9. Analysis of the marketed formulation 3.9.1. For LC method The peaks at tR 2.93 min for etamsylate was observed in the chromatogram of the drug samples extracted from tablets. Experimental results of the amount of etamsylate in tablets, expressed as percentage of label claim were in good agreement with the label claims thereby suggesting that there is no interference from any excipients, which are normally present in tablets. The drug content was found to be 100.84% ± 1.65 (% R.S.D. of 1.03) for etamsylate. Statistical evaluation was performed using Student’s t-test and the F-ratio at 95% confidence level as shown in Table 7. 3.9.2. For HPTLC-densitometric method The spots at Rf 0.23 for etamsylate were observed in the densitogram of the drug samples extracted from tablets. There was no interference from the excipients commonly present in the tablets. The drug content was found to be 100.75% ± 1.44 (% R.S.D. of 0.64) for etamsylate. It may therefore be inferred that degradation of etamsylate had not occurred in the marketed formulations that were analyzed by this method as shown in Table 7. The low % R.S.D. value indicated the suitability of this method for routine analysis of etamsylate in pharmaceutical dosage form. The summary data of validation parameters are listed in Table 8. 3.10. HPTLC versus LC Six different samples taken during in process control of tablet manufacturing were determined simultaneously by HPTLC and LC methods. Each sample was analyzed in duplicate. To test differences between the proposed HPTLC and LC method statistical tests were performed for the level of confidence 95% (P = 0.05). Two-way ANOVA was applied to test both method–sample interactions (interaction variation) and differences in the method precision (column variation). Since the within cell variation (residual variation) is greater than interaction variation as well as column variations, the method–sample interaction and the differences between the methods are not significant. To test means (averages) a paired t-test was applied. The test removes any variations between samples [20]. The obtained value of tstat is lower than twotail tcrit , which leads to the conclusion that there is no significant difference between the means. The results of two way ANOVA and paired t-test are given in Tables 9 and 10, respectively. 3.11. Stability-indicating property 3.11.1. Acid and base induced-degradation product In LC: The chromatograms of the acid degraded sample showed two additional peaks at tR 4.70 and 5.00 min, respectively and base degraded samples showed one additional peaks at tR 3.55 min and after 6.0 h an additional peak arises

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

63

Table 11 (a) Degradation of etamsylate using LC method; (b) degradation of etamsylate using HPTLC method Condition

Time (h)

% Recovery

tR (min) of degradation products

(a) Acid 5 N HCl, refa (70 ◦ C) Base 0.1 N NaOH, ref H2 O2 6% (w/v), ref H2 O2 50% (w/v), ref Dry heat (100 ◦ C) Wet heat, ref (100 ◦ C) Day light (25 ◦ C) UV light Neutral hydrolysis

48.0 2.0 2.0 2.0 2.0 2.0 1080.0 168 2.0

3.68 16.09 84.87 10.53 99.70 94.46 8.54 97.41 88.14

4.70, 5.00 3.55 3.78 3.78 3.70 3.88, 4.46 3.70 3.70 –

Condition

Time (h)

% Recovery

Rf value of degradation products

(b) Acid 5 M HCl, refa (70 ◦ C) Base 0.1 M NaOH, ref H2 O2 6% (w/v), ref H2 O2 50% (w/v), ref Dry heat (100 ◦ C) Wet heat, ref (100 ◦ C) Day light (25 ◦ C) UV light Neutral hydrolysis

24.0 2.0 2.0 2.0 2.0 2.0 1080.0 168 2.0

50.21 14.65 90.10 8.41 98.89 90.61 6.21 90.54 85.29

0.48 0.04 0.33 0.33 0.53 0.18, 0.22, 0.48, 0.53, 0.63 0.53 0.53 0.10

a

Refluxed.

at 4.72 min which closely matches with the tR 4.70 min of hydroquinone (known impurity) (Figs. 5, 6a and 6b). In HPTLC: The densitogram of the acid degraded sample showed two degraded peaks at Rf 0.33 and 0.48 (Fig. 7).

Fig. 16. Chromatogram of etamsylate and its known impurity (hydroquinone) (each 100 ␮g ml−1 ); peak 1 (etamsylate) (tR : 2.93 min), peak 2 (hydroquinone) (tR : 4.70 min).

Fig. 17. Densitogram of etamsylate and its known impurity (hydroquinone) after single development; peak 1 (etamsylate) (Rf : 0.10), peak 2 (impurity) (Rf : 0.48).

64

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

Fig. 18. In situ spectrum of standard (λmax = 305 nm) and its known impurity (hydroquinone) (λmax = 288 nm) measured from 190 to 450 nm.

Fig. 20. Densitogram of etamsylate (8000 ng spot−1 ) and its unknown impurity after single development; peak 1 (etamsylate) (Rf : 0.13), peak 2 (impurity) (Rf : 0.55).

In situ spectrum of above densitogram shows the λmax of major acid degraded product at 288 nm (Fig. 8). The densitogram of base degraded sample showed base spot at Rf 0.04 (Fig. 9).

Fig. 19. Chromatogram of etamsylate (500 ␮g ml−1 ) and its unknown impurity; peak 1 (etamsylate) (tR : 2.93 min), peak 2 (unknown impurity) (tR : 3.88 min).

Fig. 21. Pseudo first-order plots for the degradation of etamsylate with 5 M HCl at various temperatures using LC method. Key: 90 ◦ C (䊉), 80 ◦ C ( ), 70 ◦ C (×), 60 ◦ C (
), 50 ◦ C (), 40 ◦ C (), Ct , concentration at time t; C0 , concentration at time zero.

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

The concentration of the drug was found to be changing from the initial concentration indicating that etamsylate undergoes degradation under acidic and basic conditions. 3.11.2. Hydrogen peroxide induced-degradation product In LC: The sample degraded with 6 and 50% (w/v) hydrogen peroxide (Fig. 10a and b) showed one additional peak at tR 3.78 min. In HPTLC: The sample degraded with 6 and 50% (w/v) hydrogen peroxide (Fig. 11) showed one additional peak at Rf value of 0.33. The spots of degraded products were well resolved from the drug spot. 3.11.3. Dry heat and wet heat degradation product The samples degraded under dry heat condition showed additional peak at tR 3.70 min and Rf value 0.53 in LC and HPTLC, respectively. The samples degraded under wet heat conditions (Fig. 12) showed additional peaks at tR 3.88 and 4.46 min in LC. In HPTLC densitogram degraded samples showed additional spots at Rf 0.18, 0.22, 0.48, 0.53 and 0.63, respectively (Fig. 13). The spots of degraded products were well resolved from the drug spot. 3.11.4. Photochemical and UV degradation product In LC the photo and UV degraded sample showed one additional peak at tR 3.70 min when drug solution was left in day light for 45 and 7 days, respectively. The HPTLC densitogram showed one additional peak at Rf 0.53 for the photo and UV degraded sample, respectively (Fig. 14).

65

3.11.5. Neutral degradation The LC chromatogram for neutral degradation showed decrease in peak area of standard without corresponding rise in new peak. The HPTLC densitogram for neutral hydrolysis showed one addition peak at Rf 0.09 (Fig. 15) after double development of HPTLC plate. This indicates that the drug is susceptible to acid–base hydrolysis, oxidation dry and wet heat degradation and photodegradation. The results are listed in Table 11a and b. 3.12. Detection of the related impurities The standard (etamsylate) and its known impurty (hydroquinone) were well-resolved using LC (Fig. 16) as well as HPTLC method (Fig. 17). The tR and Rf value of known impurity was found to be 4.70 min and 0.48, respectively. It is clear from Fig. 18 that the λmax of known impurity and major acid degraded product exactly matches. While injecting higher concentration of standard etamsylate drug solution (500 ␮g ml−1 ) in triplicate, an additional peak was observed at tR 3.88 min, which was considered as another unknown impurity (Fig. 19). The spots other than the principal spot (etamsylate) from the sample solution were not more intense than the principal spot from the standard solution. The sample solution showed one additional spot at Rf 0.55 (Fig. 20). However, the peak area of the additional spot was found to be much less as compared to the peak area of principal spot from the standard solution (Table 12). From Table 11a, it can be observed that the tR of acid degraded product matches with the tR of known impurity. Also the tR of peroxide, UV, sunlight, dry heat degraded, first component of acid and wet heat degraded closely resembles with the unknown impurity. From Table 11b, it can be observed that the Rf values of acid degraded, third component of wet heat degraded product closely matches with the Rf value of known impurity. Also the Rf value of sunlight, UV, dry heat and fourth component of wet heat degraded product closely matches with the Rf value of unknown impurity present in the drug. Therefore, it might Table 12 Related impurity (unknown) of etamsylate using LC and HPTLC method

Fig. 22. Pseudo first-order plots for the degradation of etamsylate with 0.1 M NaOH at various temperatures using LC method. Key: 90 ◦ C (䊉), 80 ◦ C ( ), 70 ◦ C (×), 60 ◦ C (
), 50 ◦ C (), 40 ◦ C (), Ct , concentration at time t; C0 , concentration at time zero.

Concentration of drug

tR (min)

Area

In LC 5.0 ␮g ml−1 500.0 ␮g ml−1

2.93 2.93

182544.50 20254452.13

Related impurity 500.0 ␮g ml−1

3.88

12342.18

Concentration of drug

Rf value

Area

In HPTLC 400 ng spot−1 20,000 ng spot−1

0.23 0.23

2736.48 60182.85

Related impurity 10,000 ng spot−1

0.55

315.11

66

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

Fig. 24. Arrhenius plot for the degradation of etamsylate in presence of 5 M HCl () and 0.1 M NaOH ().

Fig. 23. Pseudo first-order plots for the degradation of etamsylate with 50% (w/v) hydrogen peroxide at various temperatures using LC method. Key: 90 ◦ C (䊉), 80 ◦ C ( ), 70 ◦ C (×), 60 ◦ C (
), 50 ◦ C (), 40 ◦ C (), Ct , concentration at time t; C0 , concentration at time zero.

be possible that during processing or storage the drug may have undergone oxidation or hydrolysis to a little extend. 3.13. Degradation kinetics The kinetic of degradation of etamsylate was investigated in 0.1 M NaOH, 5 M HCl and 50% (w/v) hydrogen peroxide, since the decomposition rate of etamsylate at lower strength of HCl and hydrogen peroxide was too slow to obtain reli-

Table 13 Degradation rate constant (Kobs ), half-life (t1/2 ) and t90 for etamsylate in presence of 5 M HCl, 0.1 M NaOH and 50% (w/v) H2 O2 determined by LC method Temperature (◦ C)

Kobs (h−1 )

t1/2 (h)

t90 (h)

In 5 M hydrochloric acid 40 50 60 70 80 90

0.004 0.005 0.012 0.020 0.029 0.037

2.89 2.31 0.96 0.58 0.40 0.31

0.44 0.35 0.15 0.09 0.06 0.05

In 0.1 M sodium hydroxide 40 50 60 70 80 90

0.027 0.032 0.034 0.036 0.041 0.046

0.43 0.36 0.34 0.32 0.28 0.25

0.06 0.54 0.50 0.05 0.04 0.03

In 50% (w/v) hydrogen peroxide 40 0.005 50 0.010 60 0.014 70 0.018 80 0.022 90 0.026

2.17 1.16 0.85 0.63 0.53 0.43

0.33 0.17 0.12 0.09 0.08 0.06

Fig. 25. Arrhenius plot for the degradation of etamsylate in presence of 50% (w/v) hydrogen peroxide.

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

67

Table 14 Summary of degradation kinetic data at 25 ◦ C using LC method Parameters

In 5 M HCl

In 0.1 M NaOH

In 50% (w/v) H 2 O2

Ea (kcal K mol−1 )a K25 (h−1 )b t1/2 (h)c t90 (h)d Ae

10.99 × 10−3 1.465 × 10−2 47 7.13 129.45

2.25 × 10−3 22.30 × 10−2 3.10 0.47 2.80

6.12 × 10−3 20.74 × 10−2 3.34 0.51 368.38

a b c d e

Activation energy. Degradation rate constant. Half-life. Time left for 90% potency. Arrhenius frequency factor.

able kinetic data. Each experiment was repeated three times at each temperature and time interval. The mean concentration of etamsylate was calculated for each experiment. A regular decrease in the concentration of etamsylate with increasing time intervals was observed. At the selected temperatures (40, 50, 60, 70, 80 and 90 ◦ C for acidic, alkaline and oxidative degradation) the degradation process followed pseudo-first order kinetic (Figs. 21–23). From the slopes of the straight lines it was possible to calculate apparent first degradation rate constant, half-life (t1/2 ) and t90 (i.e., time where 90% of original concentration of the drug is left) at each temperature for acidic, alkaline and oxidative degradation processes determined by LC method (Table 13). Data obtained from first order kinetics treatment was further subjected to fitting in Arrhenius equation: log K =

log A − Ea 2.303RT

Fig. 27. In situ spectrum of standard etamsylate (λmax = 305 nm) and its unknown impurity (λmax = 311 nm) measured from 190 to 450 nm.

where K is rate constant, A the frequency factor, Ea the energy of activation (cal mol−1 ), R the gas constant (1.987 cal/K mol) and T is absolute temperature (K). A plot of (2 + log Kobs ) values versus (1/T × 103 ), the Arrhenius plot was obtained

(1)

Fig. 26. pH-rate profile for the decomposition of etamsylate at constant ionic strength buffer solutions at 70 ◦ C.

Fig. 28. Densitogram of hydroquinone (1000 ng spot−1 ) and its unknown impurity after single development; peak 1 (hydroquinone) (Rf : 0.48), peak 2 (unknown impurity) (Rf : 0.55).

68

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70 Table 15 Degradation rate constant (Kobs ), half-life (t1/2 ) and t90 for etamsylate in constant ionic strength buffer at different pH values and a temperature of 70 ◦ C using LC method pH

Kobs (h−1 )

t1/2 (h)

t90 (h)

1.8 2.8 3.8 4.6 5.7 6.8 8.0 9.2 9.7 10.8

0.011 0.004 0.001 0.002 0.003 0.004 0.006 0.009 0.014 0.019

1.05 2.88 11.55 5.78 3.85 2.88 1.93 1.28 0.83 0.61

0.16 0.44 1.75 0.88 0.58 0.44 0.29 0.19 0.13 0.09

(Figs. 24 and 25), which was found to be linear in the temperature range 40–90 ◦ C. The activation energy and the Arrhenius frequency factor were calculated, respectively for acidic, alkaline and oxidative degradation processes determined by LC

method Table 12. The method of accelerating testing of pharmaceutical products based on principles of chemical kinetics was used to obtain a measure of the stability of the drug under said conditions [21,22]. The degradation rate constant at room temperature (K25 ) is obtained by extrapolating to 25 ◦ C (where 1000/T = 3.356) by inserting this into Eq. (1) and t1/2 and t90 are calculated, respectively Table 14. The pH-rate profile of degradation of etamsylate in constant ionic strength buffer solutions was studied at 70 ◦ C using LC method (Fig. 26). The apparent first order degradation rate constant and the half-life were calculated for each pH value (Table 15). From the degradation kinetic data, it can be concluded that the drug is highly susceptible to alkaline and oxidative degradation. The pH-rate profile study shows

Fig. 30. Densitogram of acid degraded product (2000 ng spot−1 ) and its unknown impurity after single development; peak 1 (acid degraded product) (Rf : 0.48), peak 2 (unknown impurity) (Rf : 0.55).

Fig. 31. In situ spectrum of acid degraded product (λmax = 288 nm) and its unknown impurity (λmax = 311 nm) measured from 190 to 450 nm.

Fig. 29. In situ spectrum of hydroquinone (λmax = 288 nm) and its unknown impurity (λmax = 311 nm) measured from 190 to 450 nm.

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

69

Fig. 34. IR spectroscopy of the acid degraded product of etamsylate.

Fig. 32. In situ spectrum of unknown impurity (λmax = 311 nm) of etamsylate, hydroquinone and acid degraded product measured from 190 to 450 nm.

that the etamsylate was found to be most stable at pH of 3.8. From the densitograms, spectrums (Figs. 27–32) and IR data (Figs. 33–35), it can be concluded that the acid degraded product is hydroquinone itself, which is a known impurity of etamsylate as reported in B.P. A strong and broad peak at 3300 cm−1 indicates the presence of an OH group in the acid degraded product. A strong band at 1050 cm−1 confirms the presence of primary alcoholic group. The bands at 2265 and 2920 cm−1 may be due to C H stretching. Two sharp peaks present in the region of 2600–2550 cm−1 due to S H stretching (Fig. 33) are absent in acid degraded product (Fig. 34) as well as in hydroquinone (Fig. 35). Further the melting point of acid degraded was determined and found to be in the range of 171–174 ◦ C that closely matches with melting point of hydroquinone (173 ◦ C). In the present study it is clear that the

Fig. 33. IR spectroscopy of the pure etamsylate.

Fig. 35. IR spectroscopy of the pure hydroquinone.

known impurity (hydroquinone) is not present in the reference sample provided by the manufacturer. But another unknown impurity was found out whose λmax is 311 nm, which is also associated with the hydroquinone and acid degraded product itself.

4. Conclusion The proposed LC and HPTLC methods provide simple, accurate, reproducible and stability indicating for quantitative analysis for determination of etamsylate in pharmaceutical tablets, without any interference from the excipients and in the presence of its acidic, alkaline, oxidative and photolytic degradation products. Both the chromatographic methods were validated as per ICH guidelines. Six real samples of tablets were determined simultaneously by HPTLC and LC methods and the results were correlated. The HPTLC method is simple and uses a minimal volume of solvents, compared to LC method. Statistical tests indicate that the proposed HPTLC and LC methods reduce the duration of analysis and appear to be equally suitable for routine determination of etamsylate in pharmaceutical formulation in quality control laboratories, where economy and time are essential. This study is a typical example of development of a

70

N. Kaul et al. / Analytica Chimica Acta 536 (2005) 49–70

stability-indicating assay, established following the recommendations of ICH guidelines. It is one of the rare studies where forced decomposition was done under all different suggested conditions and the degradation products were resolved. The method can be used to determine the purity of the drug available from various sources by detecting the related impurities and also in stability studies. It is proposed for the analysis of the drug and degradation products in stability samples in industry. The above results showed the suitability of proposed LC method for acid, base and peroxide induced degradation kinetic study of etamsylate. The degradation rate constant, half-life and t90 of etamsylate can be predicted. It was found that etamsylate is rapidly degraded in alkaline medium, while it is more stable in acidic medium. The higher stability of etamsylate was found to be at pH 3.8. It may be extended for quantitative estimation of said drug in plasma and other biological fluids. The method, however, is not suggested to establish material balance between the extent of drug decomposed and formation of degradation products. As the method separates the drug from its degradation products, it can be employed as a stability indicating one.

Acknowledgements The authors thank Dr. L.D. Patel, Professor, L.M. College of Pharmacy, Gujarat, India, for arranging gift sample of etamsylate. Authors are also thankful to Mr. Shailesh Bawaskar (General Manager-Service, Anatek Services Pvt. Ltd., Mumbai, India) and Mr. Dilip Charegaonkar (Managing Director, Anchrom HPTLC Specialists, Mumbai, India), respectively for providing facilities during research work.

References [1] British Pharmacopoeia, Vol. 1, Monographs, 2001, p. 559. [2] M. Alvarez-Guerra, M.R. Hernandez, G. Escolar, C. Chiavaroli, R.P. Garay, P. Hannaert, Thromb. Res. 107 (6) (2002) 329–335.

[3] T.A. Parshina, B.Z. Sirotin, I.M. Davidovich, Ter Arkh 71 (6) (1999) 49–53. [4] K.P. Sanghvi, R.H. Merchant, A. Karnik, A. Kulkarni, Indian Pediatr. 36 (7) (1999) 653–658. [5] D. Elbourne, S. Ayers, H. Dellagrammaticas, A. Johnson, Arch. Dis. Child Fetal Neonatal 84 (2001) F183–F187. [6] L. Kovacs, J. Annus, Contracept. Deliv. Syst. 1 (2) (1980) 155– 159. [7] M. Okuma, H. Takayama, T. Sugiyama, S. Sensaki, H. Uchino, Thromb. Haemost. 48 (3) (1982) 330–333. [8] J. Bonnar, B.L. Sheppard, Br. Med. J. 313 (1996) 579–582. [9] I. Yamboliev, T. Nikova, S.V. Bogdanova, Pharmazie 47 (4) (1992) 282–285. [10] F. Yang, C. Zhang, W.R. Baeyens, X. Zhang, J. Pharm. Biomed. Anal. 30 (3) (2002) 473–478. [11] ICH, Q1A, Stability Testing of New Drug Substances and Products, in: Proceedings of the International Conference on Harmonization, Geneva, October, 1993. [12] ICH, Q2A, Harmonised Tripartite Guideline, Text On Validation of Analytical Procedures, IFPMA, in: Proceedings of the International Conference on Harmonization, Geneva, March, 1994. [13] ICH, Q2B, Harmonised Tripartite Guideline, Validation of Analytical Procedure: Methodology, IFPMA, in: Proceedings of the International Conference on Harmonization, Geneva, March, 1996. [14] ICH Guidance on Analytical Method Validation, in: Proceedings of the International Convention on Quality for the Pharmaceutical Industry, Toronto, Canada, September, 2002. [15] C.K. Poole, S.K. Poole, J. Chromatogr. 492 (1989) 539–584. [16] A. Shimadzu, Dual Wavelength Flying Spot Densitometer CS-9000, User’s Guide. [17] K.E. McCarthy, Q. Wang, E.W. Tsai, R.E. Gilbert, M.A. Brooks, J. Pharm. Biomed. Anal. 17 (1998) 671–677. [18] B. Fried, J. Sherma, Thin-Layer Chromatography, Techniques and Applications, third ed., Marcel Dekker, New York, 1994. [19] P.D. Sethi, High Performance Thin Layer Chromatography, Quantitative Analysis of Pharmaceutical Formulations, CBS Publishers and Distributors, New Delhi, 1996. [20] J.C. Miller, J.N. Miller, Statistics for Analytical Chemistry, second ed., Ellis Horwood, New York, 1992. [21] E.R. Garrett, R.F. Carper, Chemical stability of pharmaceuticals, J. Am. Pharm. Assoc. Sci. 44 (1955) 515–521. [22] J.T. Carstensen, C.T. Rhodes, Drug Stability Principles and Practices, Marcel Dekker Inc., New York, 2000. [23] Indian Pharmacopoeia, Government of India, Ministry of Health and Family Welfare, Vol. 2, A-144–A-147.