Drug Testing
and Analysis
Research Article
Received: 24 September 2009
Revised: 15 November 2009
Accepted: 10 January 2010
Published online in Wiley Interscience: 26 February 2010
(www.drugtestinganalysis.com) DOI 10.1002/dta.117
A validated stability-indicating gas
chromatography method for determination
of divalproex sodium impurities
in pharmaceutical preparation
Acharya Subasranjan,a,b∗ Pulluru Suresh,a C. Srinivasulua
and Raoutray Hemantb
A stability-indicating gas chromatography (GC) method has been developed and validated for the quantitative determination
of divalproex sodium impurities in pharmaceutical preparation. A technique has been developed whereby the peak purity of
a compound with poor UV detection can be determined using a gas chromatograph coupled with a mass spectrometer. The
drug products were subjected to hydrolysis, oxidation, photolysis, and heat to apply stress conditions. The stability-indicating
nature of the method has been proven by establishing peak purity of all stressed samples. The chromatographic separation
was performed on a fused silica capillary (Quadrex-FFAP, 30 meter, 0.32 mm and 1 µm film thickness) column. The method
validation results indicate that the method is specific, accurate, linear, reproducible, rugged, and robust. The effectiveness
of the technique was demonstrated with stability sample analysis of divalproex sodium in its pharmaceutical preparation.
c 2010 John Wiley & Sons, Ltd.
Copyright
Keywords: divalproex sodium; impurities; gas chromatography; stability-indicating method.
Introduction
182
Divalproex sodium is a stable co-ordination compound comprising
sodium valproate and valproic acid (VPA) in a 1 : 1 molar
ratio, as shown in Figure 1 and formed during the partial
neutralization of VPA with 0.5 molar equivalent of sodium
hydroxide. Chemically it is designated as sodium hydrogen bis(2propylpentanoate). VPA (2-propylpentanoic acid) is a C8 branched
carboxylic acid and an anti-epileptic drug widely used for the
treatment of seizure disorder.[1] Many analytical approaches
have been published for the determination of VPA, based on
high performance liquid chromatography (HPLC),[2 – 8] capillary
electrophoresis[9 – 10] in combination with mass spectrometry
(MS),[11 – 17] ultraviolet detection (UV) or fluorescence detection,
usually after dervatization with a suitable chromophore or
fluorophore.[18 – 20] VPA and its pharmaceutical formulation is now
official in USP forum and European Pharmacopoeia[21 – 22] but there
no official or analytical method has appeared in the literature for
impurity profiling of VPA in pharmaceutical formulations.
The pharmaceutical industry follows ethical rules and is bound
to monitor strict control over the impurities when manufacturing
drug substances and drug products. These impurities are
classified as organic, inorganic and residual solvents.[23 – 24] Organic
impurities can originate by alteration of reaction conditions,
such as temperature, pH or in storage conditions (hydrolysis,
oxidation, etc.). An ideal stability-indicating method is one
that quantifies the standard drug alone and also resolves
its degradation products. With the advent of International
Conference on Harmonisation (ICH) guidelines, the requirement
of establishment of stability-indicating method has become more
clearly mandated. The guidelines explicitly require the conduction
Drug Test. Analysis 2010, 2, 182–187
of forced decomposition studies under a variety of conditions,
like pH, light, oxidation, dry heat, etc., and separation of the
drug from degradation products. The establishment of purity of
chromatographic peak in stressed samples is essential for the
validation of the chromatographic method. This is particularly
important when developing a stability-indicating method for
determination of impurities. Currently a number of methods are
employed for peak purity using a photo diode array detector,
which essentially compares the entire UV spectra recorded at
various points across the liquid chromatography (LC) peak with
the spectrum collected at the apex of the peak. Sometimes this
concept is not useful as there are often no spectral data available
for the impurity.
Gas chromatography (GC) is a very well established technique
for determination of residual solvents, drug testing and environmental contaminant identification. GC is very rarely used
for quantitative determination of impurities in pharmaceutical
preparation. Developing a stability-indicating method on GC for
impurity profiling in pharmaceutical formulation is a challenging
task. Complexities involved in development of method are extraction of impurities in presence of polymeric materials, selection of
∗
Correspondenceto:Acharya Subasranjan,DrReddy’sLaboratoriesLtd,Generics
AR&D, Innovation Plaza, Survey Nos. 42,45,46 & 54,Bachupalli, Qutubllapur, RR
dist -500072, AP, India. E-mail: subasranjana@drreddys.com
a Dr Reddy’s Laboratories Ltd, Generics AR&D,Innovation Plaza, Survey Nos.
42,45,46 & 54,Bachupalli,Qutubllapur, RR dist -500072,AP,India
b Ravenshaw University, Department of Chemistry, Cuttack -753003, Orissa, India
c 2010 John Wiley & Sons, Ltd.
Copyright
Drug Testing
and Analysis
A stability-indicating Gas chromatography method for determination of Divalproex Sodium Impurities
Oven temperature gradient was started at 60 ◦ C held for 2 min,
then raised to 230 ◦ C at the rate of 3.5 ◦ C/min and held at 230 ◦ C
for 20 min. Helium was used as carrier gas with a constant flow
rate of 2 ml/min. The injector temperature was kept at 180 ◦ C in
split less mode. The detector temperature was kept at 240 ◦ C. The
specificity study was conducted by using heating oven, stability
chamber and heating mantel (Thermo Lab Thane, India).
Standard and sample preparation
Figure 1. Chemical structures of Valproic Acid and Divalproex Sodium.
a suitable diluent for gas chromatographic analysis, and proving
stability indicating nature of the method.
Several derivatization techniques were developed, in which
VPA was converted into methyl ester derivatives, trimethylsilyl
derivative, tetra-butyldimethylsilyl derivatives or pentafluorobenzyl derivatives.[25] Satisfactory results in terms of sensitivity have
been obtained using GC-MS. However, all these methods are intended for determination of VPA in biological matrices, requiring
multiple sample preparation, derivatization and are time consuming. In our study, a simple stability-indicating GC method was
developed and validated for determination of related substance
of divalproex sodium in pharmaceutical formulation. The peak
purity of stressed samples has been established by comparing
mass ion fragmentation pattern with the VPA reference standard.
All four impurities (valeric acid, diethyl acetic acid, ethyl propyl
acetic acid and diallyl acetic acid content) are well separated from
each other. The method was validated as per ICH guideline[26] and
successfully applied for separation of all compound of interest in
the pharmaceutical formulation.
Experimental
Chemicals and reagents
VPA impurities were obtained from Dr Reddy’s Laboratories,
Hyderabd, India. Acetonitrile (HPLC grade), sulfuric acid (analytical
reagent grade), chloroform (HPLC grade), formic acid, sodium
sulfate, sodium hydroxide, hydrochloric acid, hydrogen peroxide
were from Merck (Darmstadt, Germany). Water was purified by a
Millipore (Bedford, MA, USA) Milli-Q water-purification system and
passed through a 0.22 µm membrane filter (Durapore; Millipore,
Dublin, Ireland) before use. Stock standard solutions of VPA and
impurities were prepared in Milli-Q water as diluent.
Instrumentation
Drug Test. Analysis 2010, 2, 182–187
Validation of the method
The method was validated for specificity, sensitivity, linear range,
accuracy, precision and robustness as per ICH guidelines.[26]
Specificity
A study was conducted to demonstrate the effective separation of
VPA and its impurities. The study was also intended to ensure
the effective separation of degradation peaks of formulation
ingredients at the retention time of VPA and its impurities.
Separate portions of drug product and ingredients were exposed
to following stress conditions to induce degradation.
The drug product was subjected to hydrolysis by refluxing the
test solution in 5 N sodium hydroxide solution at 60 ◦ C for 28 h.
Similarly the acidic hydrolysis was performed by refluxing test
solution in 5N hydrochloric acid solution at 60 ◦ C for 28 h. The
neutral hydrolysis was done in water at refluxing temperature of
60 ◦ C for 24 h. Oxidation studies were performed in 3% hydrogen
peroxide solution at 60 ◦ C for 28 h. On photo stability study, the
drug product was sufficiently spread on Petri-plates (1 mm thick
layer) and exposed to sunlight and UV light at ambient conditions
for 7 days. Humidity study was performed separately by exposing
drug product to humidity at 25 ◦ C, 90% RH for 7 days. Thermal
degradation study was performed by heating drug product at
60 ◦ C for 24 h. Similarly, placebo samples were prepared like a
drug product by exposing formulation ingredients without drug
substance. Stressed samples were injected into the GC system
with FID detector by following test method conditions and same
samples also analyzed in a GC coupled with mass detector to
demonstrate peak purity.
c 2010 John Wiley & Sons, Ltd.
Copyright
www.drugtestinganalysis.com
183
GC was carried out in an Agilent GC-FID consisting of 6890A GC and
5888N auto sampler. Quadrex-FFAP, 30 meter, 0.32 mm and 1 µm
film thickness column was used. The injection volume was 4 µL.
The standard stock solution of VAP was prepared by dissolving
an accurately weighed amount of drug in diluent, resulting in a
concentration of 1 mg/mL. Final standard solution was prepared
by taking 5 mL of above stock solution, mixed with 21 mL of
diluent and extracted three times with 20 mL chloroform each.
Chloroform layer collected in a 100 ml flask after passing through
sodium sulfate bed and volume made up to 100 mL.
The test solution was prepared by taking powdered
tablets/capsules and a known amount of VPA equivalent to
4500 mg was transferred to a 50 mL volumetric flask along with
30 mL diluent. The powdered material was dispersed by mixing in
an ultrasonic bath for 20 min and diluted to 50 mL with diluent.
Above solution was centrifuged at 4000 rpm for 15 min in order
to eliminate insoluble excipients and 20 mL of the supernatant
liquid was taken for extraction. Test solution was mixed with 6 mL
of 20% sulfuric acid solution to convert sodium valproate present
in divalproex sodium into VPA and extracted three times with
20 mL chloroform each time. Chloroform layer was collected in a
100 ml flask after passing through sodium sulfate bed and volume
made up to 100 mL. Standard and test preparation was used for
chromatographic analysis.
Drug Testing
and Analysis
A. Subasranjan et al.
Precision
The precision of test method was evaluated by analyzing six
samples of VPA test preparation spiked with VPA impurities blend
solution to get the concentration of 0.1% of sample concentration
and analyzed as per test method.
Accuracy
A study of accuracy of VPA impurities from spiked samples of test
preparation was conducted. Samples were prepared in triplicate
by spiking impurities in test preparation at the level of limit of
quantification (LOQ) 50%, 100%, 150%, 200% and 300% to the
target concentration of impurities (i.e., About 0.2–80 µg mL−1 ).
Limit of detection (LOD) and LOQ
Figure 3. Chemical structures of Valproic Acid and its fragments.
Results and Discussion
LOD and LOQ values were determined by using the signal-to
noise approach as defined in ICH guideline.[26] Increasingly, dilute
solution of each impurity was injected and signal to noise was
calculated at each concentration. LOD and LOQ values were
calculated with a signal to noise ratio (S/N) 3 and 10 respectively.
Linearity of detector response
A series of solutions of VPA-related compounds in the concentration ranging from about LOQ level to about 300%
(0.2–80 µg mL−1 ) of the target concentration of impurities were
prepared and injected into the GC system.
Real time sample analysis
The method suitability was verified by analyzing both initial
and stability sample of in-house formulated product containing
divalproex sodium as active substance. An accurately weighed
quantity equivalent to 4500 mg of drug was taken and dispersed
in 30 mL of diluent. The dispersed material was kept in an ultrasonic
bath for 20 min and volume made up to 50 mL. Above solution was
centrifuged at 4000 rpm for 15 min in order to eliminate insoluble
excipients and 20 mL of the supernatant liquid was used for
extraction. Test solution was mixed with 6 mL of 20% sulfuric acid
solution and extracted thrice, each time with 20 mL chloroform.
Chloroform layer was collected in a 100-ml flask after passing
through sodium sulfate bed and volume was made up to 100 mL.
Standard and test preparation were used for chromatographic
analysis.
In order to obtain the optimized extraction conditions and best
extraction efficiency, we used the peak area of VPA standard as the
GC response to evaluate the extraction efficiency under different
conditions. To optimize the method, all extractions were initially
carried out on standard stock solution and finally to tablets.
Divalproex sodium is a mixture of VPA and sodium valproate.
Sodium valproate present in divalproex sodium was converted
to VPA as sodium valproate is retained in column so thoroughly
that it is not observed even after 75 min of injection, where
as retention time and peak response of VPA are found to be
satisfactory. VPA, with a pKa of 5, exists in neutral (un-ionized)
form at low pH, and is completely ionized at pH higher than 4 and
thus it has more tendency to dissolve in water. Divalproex sodium
was dissolved in water, which was converted to VPA and sodium
valproate. Subsequently, addition of acidic solution converted
sodium valproate to VPA, as shown in Figure 2. The best results
were obtained by adding 6 mL 20% v/v sulfuric acid solution.
It is essential to select a suitable organic solvent for GC analysis.
The following factors should be considered. First, according to the
theory of ‘like attracts like’, the extraction organic solvent should
have high affinity for the analytes in the sample. Secondly, the
organic solvent should be suitable for the GC analysis and the
solvent peak should be satisfactorily resolved from the analyte
peak. Finally, complete recovery in the presence of other excipient
present in drug product is necessary.
Four solvents, N-Heptane, methanol, n-Hexane and chloroform
were tested to select the best one for extraction of VPA in water
samples with this technique. Preliminary experiments showed
184
Figure 2. Chemical reaction of Divalproex Sodium to Valproic acid.
www.drugtestinganalysis.com
c 2010 John Wiley & Sons, Ltd.
Copyright
Drug Test. Analysis 2010, 2, 182–187
A stability-indicating Gas chromatography method for determination of Divalproex Sodium Impurities
Drug Testing
and Analysis
Figure 4A. Typical Chromatogram of Valproic Acid Standard.
Figure 4B. Typical Chromatogram of Identification Solution.
that chloroform gives the best extraction efficiency of VPA in the
presence of excipient matrix.
Selectivity, sensitivity, resolution, and speed of chromatographic
separation were optimized for the GC method. The retention times
of VPA at 35, valeric acid at 28.96, diethyl acetic acid at 29.58, ethyl
propyl acetic acid at 32.30 and diallyl acetic acid at 38.82 min,
respectively, under the chromatographic conditions described,
and the total run time was 70 min. Chromatograms obtained from
valproic acid standard, identification solution and test preparation
are shown in Figures 4A, 4B and 4C, respectively.
The specificity of the proposed method was verified by injecting
all stressed samples and placebo components. Further, all stressed
samples were injected in to a GC coupled with mass detector
to demonstrate the peak purity by following the standard test
procedure. Upon evaluation, no ion was found overlapping with
the main analyte peak in all the stressed samples. This indicates
that there is no co-elution of degradation product peak with the
main analyte peak. The main analyte peak in all the above samples
was compared for the mass ion fragmentation pattern with the
VPA reference standard. The observed possible fragmentation of
main analyte peak is m/z = 73 and 102. Fragmentation patterns
of all stressed samples and reference standards are the same.
The structures of respective fragments are as given in Figure 3.
The data also revealed that the known impurities along with the
unknown degradation products are well resolved from the main
peak.
Low relative standard deviation (RSD) values (<10%) of precision
and intermediate studies indicate that the method is highly precise
and data of precision study are shown in Table 1. The amount
recovered was ±10% of amount added in accuracy study; this
indicates that the method is highly accurate. The data of recovery
study are shown in Table 2. Sensitivity of the method was verified
185
Figure 4C. Typical Chromatogram of Test solution.
Drug Test. Analysis 2010, 2, 182–187
c 2010 John Wiley & Sons, Ltd.
Copyright
www.drugtestinganalysis.com
Drug Testing
and Analysis
A. Subasranjan et al.
Table 1. Percentage of RSD of impurities in precision study
Impurity Name
Table 4. Correlation coefficient of impurities
Precision (% RSD)
(n = 6)
Intermediate
Precision (% RSD)
(n = 6)
1.8
1.7
1.7
2.0
2.6
3.0
3.3
3.4
Valeric Acid
Diethyl acetic acid
Ethyl propyl acetic acid
Diallyl acetic acid
Parameters
(n = 7)
Slope
Y intercept
Standard error
Correlation Coefficient
Valeric
Acid
Diethyl
Acetic
Acid
Ethyl
propyl
Acetic
Acid
Diallyl
Acetic
Acid
19.82
−21.1
18.46
0.999
16.51
−8.87
7.57
0.999
19.82
−21.10
18.46
0.999
17.04
59.89
59.9
0.997
Table 2. Percentage recovery of impurities at different level
% Recovery % Recovery % Recovery
of Diallyl
of Ethyl
% Recovery of Diethyl
Acetic
propyl
Acetic
of Valeric
Nominal
Acid
Acetic Acid
Acid
Acid
concentrations
LOQ level
50% level
100% level
300% level
99.5
104.7
93.0
103.9
98.5
94.8
94.1
100.9
93.1
107.0
109.0
105.4
92.8
101.2
102.4
105.5
Table 3. Limit of detection (LOD) and limit of qualification (LOQ) of
impurities in µg mL−1
Nominal
concentrations
LOD µg mL−1
LOQ µg mL−1
Precission at LOQ (% RSD)
Valeric
Acid
Diethyl
Acetic
Acid
Ethyl
Propyl
Acetic
Acid
0.1
0.5
8.4
0.1
0.4
4.2
0.05
0.4
4.6
Diallyl
Acetic
Acid
0.1
0.5
0.0
and the method was linear, accurate and precise at LOQ. The
data of LOD and LOQ study are shown in Table 3. The calibration
curve for all impurities was obtained by plotting the peak area
of individual impurity versus the concentration over the range of
about 0.2–80 µg mL−1 , and was found to be linear with r = 0.999.
The data of regression analysis of the calibration curves are shown
in Table 4. Real time analysis data ensured that the developed
method is suitable for drug product analysis and data captured in
Table 5.
An efficient, sensitive, selective, GC method has been developed
and successfully applied in the separation of related substances
of VPA. No interfering peaks were observed in blank and placebo,
indicating that signal suppression or enhancement by the product
matrix was negligible. A challenging task of peak purity has been
established to prove that the method is stability indicating.
Conclusion
186
Although LC is a versatile technique for the impurity analysis
in complex matrices, the presence of interfering substances and
poor UV absorbency makes the separation and quantification difficult. A number of analytical approaches have been previously
described to determine VPA in biological materials and pharmaceutical preparation; however, this is the first study reporting a
validated stability-indicating GC method for impurity quantifica-
www.drugtestinganalysis.com
Table 5. Impurity profile of Depakote 500 mg delayed release tablets
% impurity found
Parameters
Initial
Stability (40◦ /75%RH, 3 Month)
Valeric Acid
Diethyl Acetic Acid
Ethyl Propyl Acetic Acid
Diallyl Acetic Acid
Single Max Unknown
Total Impurity
BLD
BLD
BLD
0.030
0.009
0.04
BLD
BLD
BLD
0.067
0.018
0.12
tion in divalproex sodium formulation. The simple GC method
developed in this study makes it suitable for separation and estimation of impurities without interference from excipients and
other related substances present in the pharmaceutical matrices.
Implementation of this GC method will help us to save organic
solvent which will minimize environmental pollution. Development of a GC method is one of the best solutions to the current
global acetonitrile shortage and will safeguard against future risk.
The analytical performance and the result obtained from real time
analysis demonstrate that the method is reliable. In conclusion,
the sensitivity, selectivity, accuracy and reproducibility of the GC
method developed in this study make it suitable for quality control analysis of complex pharmaceutical preparation containing
divalproex sodium and its impurities.
Acknowledgement
We wish to express our sincere thanks to the management of Dr
Reddy’s Laboratories Ltd, Hyderabad for their support.
References
[1] R. L Macdonald, M Kevien-Kelly, Epilepsia, 1995, 36, 52.
[2] S. A Koing, J Knolle, S Friedewald, W Koelfen, E Longin, T Lenz,
D Hanak, Epilepsia, 2003, 44, 708.
[3] J. P Moody, S. M Allan, Clin. Chim. Acta. 1983, 127, 263.
[4] M Nakamura, K Kondo, R Nishioka, S Kawai, J. Chromatogr. 1984,
310, 450.
[5] F. A. L VanderHorst, G. G Eikelboom, J. J. M Holthuis, J.Chromatogr.
1988, 456, 191.
[6] E Bousquest, V Cavrini, R Gatti, A Spadaro, J. Liq. Chro. 1998, 21,
2873.
[7] K Kushida, T Ishizaki, J. Chromatography. 1985, 338, 131.
[8] T Mino, M Nakajima, H Wakabayashi, S Yamato, K Shimada, Anal.
Sci. 2001, 17, 999.
[9] V Loffe, T Kalendarev, I Rubinstein, G Zupkovitz, J. Pharma. Biomed.
Anal. 2002, 30, 391.
[10] V Pucci, R Mandrioli, M. A Raggi, Electrophoresis 2003, 24, 2076.
c 2010 John Wiley & Sons, Ltd.
Copyright
Drug Test. Analysis 2010, 2, 182–187
A stability-indicating Gas chromatography method for determination of Divalproex Sodium Impurities
[11] N. V. S Ramakrishna, K. N Vishwottam, S Manoj, M Koteshwara,
M Santosh, J Chidambara, B. R Kumar, Rapid Commun. Mass Spectro.
2005, 19, 1970.
[12] M. R Anari, R. W Burton, S Gopaul, F. S Abbot, J. Chromatogra. B.
2000, 742, 217.
[13] Z Hussein, K Patterson, J Lamm, J Cavanaugh, G Granneman, Biopharm. Drug Dispos. 1993, 14, 389.
[14] H Muro, T Tatsuhara, T Sugimoto, M Woo, N Nishida, K Murakami,
Y Yamaguchi, J.Chromatogr.B. 1995, 663, 83.
[15] E Fisher, W Wittfoht, H Nau, Biomed. Chromatogr. 1992, 6, 24.
[16] T Tatsuhara, H Muro, Y Matsuda, Y Imai, J. Chromatogr. 1987, 399,
183.
[17] M Liu, K Brouwer, G Pollack, Drug Metab. Dispos. 1992, 20, 810.
[18] C Lucarelli, P Villa, E Lombaradi, P Prandini, A Brega, Chromatographia, 1992, 33, 37.
[19] S Hara, M Kamura, K Inoue, M Fukuzawa, N Ono, T Kuroda, Biol.
Pharma. Bull. 1999, 22, 975.
Drug Testing
and Analysis
[20] S Rompotis, M Parissi-Poulou, E Gikas, M Kazanis, A Vavayannis,
I Panderi, J. Liq. Chromatogr. Related Technol. 2002, 25, 2833.
[21] The United States Pharmacopoeia, Edn 29. The United States
Pharmacopoeia Convention, Inc., Rockville, MD PF 31(5), Sept-Oct
2005, 1412.
[22] The United States Pharmacopoeia-National Formulary, USP30-NF
25, Rockville, MD, 2007.
[23] ICH 3QA,Guidance for Industry, Impurities in New Drug Substances,
US Food and drug Administration, 1996.
[24] ICH 3QB,Guidance for Industry, Impurities in New Drug Products, US
Food and drug Administration, 1996.
[25] C Deng, N Li, J Ji, B Yang, G Duan, X Zhang, Rapid Commun. Mass
Spectrom. 2006, 20, 1281.
[26] ICH Q2(R1) Validation of Analytical Procedure: Text Methodology, US
Food and drug Administration, 2005.
187
Drug Test. Analysis 2010, 2, 182–187
c 2010 John Wiley & Sons, Ltd.
Copyright
www.drugtestinganalysis.com