Drug Testing

Research article and Analysis

Received: 4 April 2010 Revised: 4 August 2010 Accepted: 15 October 2010 Published online in Wiley Online Library:

([Link]) DOI 10.1002/dta.235

A new technique for spectrophotometric

determination of Pseudoephedrine
and Guaifenesin in syrup and synthetic mixture
Siavash Riahi,a,b∗ Farshad Hadiloo,c Seyed Mohammad R. Milani,c
Nazila Davarkhah,b Mohammad R. Ganjali,b,d Parviz Norouzib,d
and Payam Seyfie
The accuracy in predicting different chemometric methods was compared when applied on ordinary UV spectra and first order
derivative spectra. Principal component regression (PCR) and partial least squares with one dependent variable (PLS1) and
two dependent variables (PLS2) were applied on spectral data of pharmaceutical formula containing Pseudoephedrine (PDP)
and Guaifenesin (GFN). The ability to derivitize in resolved overlapping spectra chloropheniramine maleate was evaluated
when multivariate methods are adopted for analysis of two component mixtures without using any chemical pretreatment.
The chemometrics models were tested on an external validation dataset and finally applied to the analysis of pharmaceuticals.
Significant advantages were found in analysis of the real samples when the calibration models from derivative spectra were
used. It should also be mentioned that the proposed method is a simple and rapid way requiring no preliminary separation
steps and can be used easily for the analysis of these compounds, especially in quality control laboratories. Copyright

c 2011
John Wiley & Sons, Ltd.

Keywords: derivative; partial least squares (PLS); Pseudoephedrine and Guaifenesin; chemometrics; spectrophotometer; principal
component regression (PCR)

Introduction PDP and GFN hydrochlorides are widely used in combination

with other drugs for the clinical treatment of common cold,
Cough and cold pharmaceutical preparations are one of the sinusitis, bronchitis, and respiratory allergies.[8] Various methods
most extended formulations in the world and have many have been used for the determination of PDP or GFN in cough and
pharmaceutical forms: syrup, suspension, sachets, capsules, cold pharmaceuticals.[9 – 16]
and tablets.[1] These preparations represent complex formulae Different methods were developed to determine of drugs in
containing several active ingredients and a broad spectrum of pharmaceutical products.[17 – 22] Despite the success in the field
excipients, such as flavouring agents, saccharose or aspartame, of pharmaceutical analysis, the methods based on ordinary
acidulants, natural or artificial colouring and flavouring agents, spectrophotometry have proven to be often unsatisfactory for
dyes, sweeteners, and preservatives.[2,3] The majority of these the analysis of multicomponent mixtures, owing to their low
ingredients are present as a mixture of basic amino compounds spectral resolution.[23,24] In contrast, derivative spectrophotometry
and their separation in pharmaceutical forms is quite complicated has been successful in resolving complex mixtures because
due to similarities of their physical and chemical properties.[4] The of its ability in extracting high analytical information from
combination of antihistamine such as pyrilamine maleate (PA) and spectra composed of overlapped bands. This technique provides
chlorpheniramine maleate (CLP) is used to overcome the allergic
effects and reduce or relieve cold symptoms.[5]
∗ Correspondence to: Siavash Riahi, Institute of Petroleum Engineering, Faculty
The assay of the drug product in the stability test sample
of Engineering, University of Tehran, P. O. Box 14155-6455, Tehran, Iran.
needs to be determined using the stability indicating method, as E-mail: riahisv@[Link]
recommended by the International Conference on Harmonization
(ICH) guidelines[6] and USP-26.[7] Although the stability indicating a Institute of Petroleum Engineering, Faculty of Engineering, University of Tehran,
P. O. Box 14155-6455, Tehran, Iran
methods have been reported for the assay of various drugs, most of
them describe assay procedures for drug products containing only b Center for of Excellence in Electrochemistry, Faculty of Chemistry, University of
one active drug substance. Only a few stability indicating methods Tehran, P. O. Box 11365-4563, Tehran, Iran
are reported for the assay of combined drug products, containing
c Faculty of Chemistry, Iran University of Science and Technology, Narmak,
two or more active drug substances. The objective of this work, Tehran 16846, Iran
for the first time, was to develop a simple, precise, and rapid
analytical procedure, which would serve as a stability indicating d Endocrinology & Metabolism Research Center, Tehran University of Medical
Sciences, Tehran, Iran
assay for the combined drug product of Pseudoephedrine (PDP)
319

and Guaifenesin (GFN) in syrup. e RAMOPHARMIN Co. Shishehsazi Mina 4 Km Karaj Makhsoos Road, Tehran, Iran

Drug Test. Analysis 2011, 3, 319–324 Copyright

c 2011 John Wiley & Sons, Ltd.
 Drug Testing
and Analysis S. Riahi et al.

OH OH (a)
H H
O N
OH CH3

CH3
CH3
O
(a) (b)

Figure 1. Structural formulae for (A) Guaifenesin and (B) Pseudoephedrine.

a better resolution of the spectral overlap by magnifying

small differences between the spectral curves.[25 – 29] Moreover,
minimization of the nonspecific matrix interference is reached by
elimination of baseline drifts.[30 – 32] A very recent review describes
derivative spectrophotometry as a powerful tool in pharmaceutical (b)
analysis.[33]
In addition, several multicomponent determinations based on
the application of these chemometrics to spectrophotometric
data have been reported.[34 – 37] These techniques cannot provide
a resolution of the spectral overlap special for real samples, because
they have complex matrix.
The combined use of derivative spectrophotometry and
chemometric techniques has been demonstrated to be a highly
convenient choice in the determination of multicomponent
matrices presenting serious spectra overlapping, thanks to their
common potential ability to exploit minor spectral features.[38 – 41]
This work describes a new method for the simultaneous PDP
and GFN (Figure 1) determination in syrup in the presence of the Figure 2. Analytical curve for the univariate determination of
(A) Guaifenesin (λmax = 222 nm, A=0.035CGFN + 0.015 and R2 = 0.998);
overlapping spectra with the PCR, PLS1, and PLS2.
and (B) Pseudoephedrine (λmax = 207 nm, A=0.038CPDP + 0.016 and
R2 = 0.998).

Experimental
standard solutions, respectively. To a series of 10-ml volumetric
Reagents and chemicals flasks, aliquots of PDP and GFN solutions, containing appropriate
All reagents were of analytical grade and all the solutions amount of these drugs in the range of calibrations, were added
were prepared with distilled water. Phosphate salts (Merck) and and then the solutions were diluted to 10 ml. UV spectra of the
phosphoric acid (Merck) were used to prepare buffer solutions mixture were recorded in the wavelength range 200–350 nm
(PH=5-8). Expectorant syrup (containing PDP and GFN) was versus the blank, while the digitized absorbance was sampled at
manufactured by Behsa pharmaceutical Co. (Arak, Iran) for Zahravi 1.0 nm intervals. All the solutions were freshly prepared.
pharma. Co. with a label of containing 30 mg and 100 mg of the
PDP and GFN, respectively. Procedure to determine Pseudoephedrine and Guaifenesin
in pharmaceutical formulations
Apparatus and software A volume of the syrup equivalent to 5 mg guaifenesin, 1.5 mg
The electronic absorption measurements were carried out on pseudoephedrine was diluted to 100 mL with distilled water.
a Perkin Elmer Lambda 800 dual beam spectrophotometer (slit Further dilution was carried out with the buffer to reach the
width of 2 nm and scan rate of 400 nm/min) using 1.00 cm quartz calibration range of each compound.
cells. The absorption spectra were recorded over the wavelength
range of 200–350 nm against a blank (phosphate buffer with
PH=6). The PLS1, PLS2, PCR and derivative spectrophotometry Chemometrics techniques
(DS) calculations were performed with the software package ‘The Experimental design (ED)
Unscrambler 9. 8 (Camo Software As, Oslo, Norway).[42]
Prediction ability of a regression model is largely affected
by the experimental design in building the calibration set. A
Stock and standard solutions
careful selection of the reference solutions increases the chances
Stock solutions of PDP and GFN, containing 100 µg mL−1 were of improving quality of information from the calibration set.
prepared in 100-ml volumetric flasks, by dissolving 10 mg of Statistical ED is a powerful technique for choosing the component
each compound in some drops methanol and distillated water. concentrations so that to minimize the total number of samples in
Working standard solutions were prepared daily by diluting the covering the defined experimental region.
stock solutions for each drug according to its linear calibration The chemometrics model calibration was optimized with the
range, 5–30 and 5–40 µg mL−1 for PDP and GFN, respectively aid of the orthogonal design method. A set of standard samples
(Figures 2A and 2B). Two sets of standard solutions were prepared, was prepared according to a five-level orthogonal array design
320

the calibration set and prediction set which contained 30 and 5 (OAD) which led to 30 samples, so that the concentration of each

[Link] Copyright

c 2011 John Wiley & Sons, Ltd. Drug Test. Analysis 2011, 3, 319–324
 Drug Testing
A new technique for spectrophotometric and Analysis

Table 1. Concentration data of the different mixtures used in the Table 3. Statistical parameters calculated from application of PCR
calibration set for the determination of Guaifenesin (GFN) and and PLS methods to full spectra of calibration samples
Pseudoephedrine (PDP) concentration values are expressed as µgml−1
PCR PLS1 PLS2
Sample GFN PDP 0D 1D 0D 1D 0D 1D

1 5.00 5.00
RMSEP 0.1065 0.0840 0.1062 0.1059 0.1113 0.0845
2 5.00 10.00
GFN R2 0.9999 0.9999 0.9999 0.9999 0.9999 0.9999
3 5.00 15.00
PC 3 3 3 2 3 3
4 5.00 20.00
5 5.00 25.00 RMSEP 0.1630 0.1076 0.1530 0.1066 0.1533 0.1089
6 5.00 30.00 PDP R2 0.9996 0.9998 0.9997 0.9998 0.9997 0.9998
7 12.00 5.00 PC 6 5 5 5 5 4
8 12.00 10.00
9 12.00 15.00
10 12.00 20.00 is used to calculate the concentrations of an unknown sample
11 12.00 25.00 from its spectrum.
12 12.00 30.00 Both algorithms transform the original variables into a smaller
13 19.00 5.00 number of orthogonal variables called principal components (PCs),
14 19.00 10.00 which are linear combinations of the original variables. When
15 19.00 15.00 multivariate calibration approaches are applied in spectrophoto-
16 19.00 20.00 metric multicomponent analysis, a relationship between spectral
17 19.00 25.00 and concentration data from reference samples, representing the
18 19.00 30.00 variables of the system, is established. A new matrix constituted
19 26.00 5.00 by the new variables PCs and scores is built. The calculation of this
20 26.00 10.00 new matrix is planned by an algorithm specific to the regression
21 26.00 15.00 method adopted. PCs represent the absorptivity values of the sam-
22 26.00 20.00 ples at the various concentration and wavelength values, whereas
23 26.00 25.00 the scores represent the numerical coefficients. Their combination
24 26.00 30.00 allows to build the mathematical model representing the refer-
25 33.00 5.00 ence spectra and able to predict the component concentrations
26 33.00 10.00 of new samples.
27 33.00 15.00
28 33.00 20.00
Determination of PC number
29 33.00 25.00
30 33.00 30.00 The optimal number of principal factors is essential in building
multivariate models. The prediction error decreases with the
number of PCs used to reach an optimal value. Most information
is usually contained in the first PCs but it is not assured that the
Table 2. Concentration data of the different mixtures used in the useful information is exclusively reserved to these factors. Full
calibration set for the determination of GFN and PDP concentration cross validation is the most used validation method, in which one
values are expressed as µgml−1 reference at a time is removed from the calibration set; after that
Sample GFN PDP the same sample is predicted by using the calibration built with
the other references. Several tests have been proposed to select
1 6.00 9.50 the number of PCs. The Root Mean Square Error of Prediction
2 7.50 23.00 (RMSEP) was chosen to express the prediction error when PCR
3 17.50 13.00 and PLS procedures were applied. This parameter represents an
4 22.50 8.00 estimate of the error when other samples are predicted with that
5 37.50 28.00 model. The best prediction ability of the models is reached when
the prediction error is at its lowest value.

0.5
1
n
drug in the resulting solutions was in its own linear dynamic range RMSEP = (xi − x̂i )2 (1)
(Table 1). For the prediction set, five mixtures were prepared, not n
i=1
being included in the calibration set, and were employed as an
independent test (Table 2). where xi is the true concentration of the analyte in the sample i,
x̂i represents the estimated analyte concentration in the sample i,
Multivariate analysis x is the mean of the true concentration in the prediction set and
n is the total number of samples used in the prediction set. The
PCR and PLS are factor analysis methods, based on a two-stage values of RMSEP for the PDP and GFN mixtures are summarized in
procedure: a calibration step, in which a mathematical model is Table 3.
built by using component concentrations and spectral data from a Another important statistical parameter in evaluating the model
321

set of references, followed by a prediction step in which the model quality is R2 . It represents an index of quality in fitting all data to a

Drug Test. Analysis 2011, 3, 319–324 Copyright

c 2011 John Wiley & Sons, Ltd. [Link]
 Drug Testing
and Analysis S. Riahi et al.

1.2
Table 4. Comparison of statistical parameters of cross validation and
test set
1
Calibration Prediction
0.8 Mixture [Link] RMSEtest [Link] RMSEtest
Absorbance

PDP 5ppm 0D
0.6 PLS1 0.1182 0.0931 0.1331 0.2237
GFN 5ppm 1D 0.0880 0.0751 0.0978 0.1567
0.4 0
GFN PLS2 D 0.1235 0.0981 0.1388 0.2310
1D 0.0859 0.0749 0.0955 0.1497
0.2 PCR 0D 0.1193 0.0938 0.1341 0.2249
1D 0.0886 0.0749 0.0978 0.1548
0 0
200 250 300 350 PLS1 D 0.2966 0.2836 0.3807 0.5511
1
Wavelength (nm) D 0.1697 0.1588 0.2138 0.3106
PDP PLS2 0D 0.1138 0.1090 0.1425 0.1491
Figure 3. Absorbance spectra of the Guaifenesin (5 µg mL−1 , λ max = 1D 0.0840 0.0830 0.1028 0.1120
222 nm), Pseudoephedrine (5 µg mL−1 , λ max = 207 nm) and their PCR 0D 0.3049 0.3073 0.4393 0.6173
mixture. 1D 0.0939 0.0833 0.1181 0.1738

0.02
0 a comparative study of the three chemometric approaches was
-0.02 undertaken. The models were validated by full cross-validation and
the statistical parameters RMSEP and R2 were recalculated when
-0.04
dA/dl

Mixture a new factor was added to the models. Table 3 lists the optimal
-0.06 PDP 5ppm PCs and the corresponding statistical parameters. RMSEP values
GFN 5ppm were satisfactory for all the models. PCs resulted ranging between
-0.08
3 and 6, showing the higher values in correspondence of PDP
-0.1 determination. This table represents by applying chemometrics
methods on derivative relative just chemometrics methods not
-0.12
200 250 300 350 only reduce RMSEP but also reduce PC numbers.
Wavelength (nm) Analogously, the values of RMSE of cross-validation and test
for calibration and prediction sets showed that use of derivative
Figure 4. First-order derivative spectra of the Guaifenesin (5 µg mL−1 ), spectra precision of determination significantly can be improved
Pseudoephedrine (5 µg mL−1 ) and their mixture. (Table 4).

straight line and represents the fraction of total variance explained Recovery study
by the model. It is computed as:
In order to test the prediction ability of the constructed models,

n 
n
a series of synthetic mixtures PDP and GFN was analyzed. A
R2 = (x̂i − xi )2 / (xi − x i )2 (2)
prediction set of five samples, containing the drugs in the same
i=1 i=1
concentration ranges used for the calibration set, was randomly
where x represents the mean of the true concentrations in the prepared. The above defined PCR and PLSs calibrations were
prediction set. applied to the prediction set and the models were tested by
comparing accuracy, in terms of percentage recovery reported in
Table 5.
Results and discussion In order to assess the applicability of the proposed method to the
analysis of real samples, determination of the drugs was conducted
Data processing and model building in different mixtures of their pharmaceutical formulations. Thus,
The absorption UV spectra of pure PDP, GFN and their mixture different mixtures of the commercially available PDP and GFN
are shown in Figure 3. The wavelength range 200–300 nm was syrups were prepared and analyzed with the PLS-1, PLS-2 and
considered throughout because the absence of absorbance PCR methods. Each measurement was repeated three times and
after 300 nm for all the drugs. Quantitative determination the obtained mean recovery values are presented in Table 6. It
by ordinary spectrophotometry, which uses measurements at can be observed that in all mixtures the calculated values are in
discrete wavelengths, was proven to be inaccurate because of the satisfactory agreement with the declared values in the derivative
extensive overlap of the spectral curves. With the aim to extract and chemometrics methods. For instance, recovery of PDP in PLS1
the most significant analytical information from the spectral data, method without derivative is 68.93, but with applying derivative
the mathematical derivatives of the absorbance values against to will be 101.5. Furthermore, applying chemometrics methods on
wavelengths (dA/dλ) were calculated. Figure 4 shows the first derivative reduce PC numbers. For example, in GFN, PC numbers
order derivative spectra of the two drugs. in PCR reduce from 13 to 5. According these results we conclude
PCR, PLS1, and PLS2 calibration models were first developed on that applying chemometrics method on derivative spectra largely
322

data from the full-range spectra of ordinary and derivative and improved determination of compound with high overlapping in

[Link] Copyright

c 2011 John Wiley & Sons, Ltd. Drug Test. Analysis 2011, 3, 319–324
 Drug Testing
A new technique for spectrophotometric and Analysis

Table 5. The predicted concentrations obtained by the application of the PLS method and the first-order derivative spectra on the absorption UV
spectra for the simultaneous determination of Guaifenesin (µg mL−1 ) and Psuedoephedrine (µg mL−1 ) in binary mixtures

PLS1:

GFN PDP
Found Recovery Found Recovery
Experimental 0D 1D 0D 1D Experimental 0D 1D 0D 1D

6.00 5.76 5.90 96.0 98.3 9.50 9.49 9.63 99.9 101.4
7.50 7.54 7.57 100.5 100.9 23.00 23.12 23.20 100.5 100.9
17.50 17.56 17.68 100.3 101.0 13.00 13.27 13.24 102.1 101.8
22.50 22.81 22.77 101.4 101.2 8.00 8.03 8.10 100.4 101.2
37.50 37.66 37.49 100.4 100.0 28.00 27.90 27.77 99.6 99.2

PLS2:

GFN PDP
Found Recovery Found Recovery
Experimental 0D 1D 0D 1D Experimental 0D 1D 0D 1D

6.00 5.76 5.84 96.0 97.3 9.50 9.57 9.64 100.7 101.5
7.50 7.55 7.55 100.7 100.7 23.00 23.14 23.12 100.6 100.5
17.50 17.77 17.65 100.5 100.8 13.00 12.77 12.86 98.2 98.9
22.50 22.81 22.73 101.4 101.0 8.00 7.95 7.80 99.4 104.0
37.50 37.69 37.41 100.5 99.8 28.00 25.29 26.79 90.3 95.7

PCR:

GFN PDP
Found Recovery Found Recovery
Experimental 0D 1D 0D 1D Experimental 0D 1D 0D 1D

6.00 5.76 5.88 96.0 98.0 9.50 9.46 9.56 99.6 100.6
7.50 7.54 7.56 100.5 100.8 23.00 23.10 23.15 100.4 100.6
17.50 17.76 17.68 101.5 101.0 13.00 13.28 13.24 102.2 101.8
22.50 22.81 22.76 101.4 101.2 8.00 8.05 8.08 100.6 101.0
37.50 37.66 37.50 100.4 100.0 28.00 27.80 27.76 99.3 99.1

Table 6. Recovery and PC studies of GFN and PDP in real sample using PCR and PLS methods

PLS1 PLS2 PCR

Actual 0D 1D 0D 1D 0D 1D

Concentration 25.00 25.10 24.93 25.29 24.76 27.06 25.07

GFN PC 9 3 10 6 13 5
Recovery 100.4 99.7 101.2 99.0 108.2 100.3
Concentration 7.50 5.17 7.61 6.40 7.72 7.56 7.52
PDP PC 8 5 8 6 9 5
Recovery 68.93 101.5 85.3 102.9 100.8 100.3

PLS1 PLS2 PCR

0 1 0 1 0 1
Actual D D D D D D

Concentration 25.00 31.94 24.93 32.04 24.76 31.92 25.07

GFN PC 3 3 6 6 5 5
Recovery 127.8 99.7 128.2 99.0 127.7 100.3
Concentration 7.50 12.43 7.61 12.48 7.72 12.96 7.52
PDP PC 5 5 6 6 5 5
Recovery 165.7 101.5 166.4 102.9 172.8 100.3
323

Drug Test. Analysis 2011, 3, 319–324 Copyright

c 2011 John Wiley & Sons, Ltd. [Link]
 Drug Testing
and Analysis S. Riahi et al.

spectra special for determination in real samples with complex [8] H. Wold, Research Papers in Statistics, Wiley: New York, 1966.
matrix. [9] M. R. Louhaichi, S. Jebali, M. H. Loueslati, N. Adhoum, L. Monser,
Talanta 2009, 78, 991.
[10] H. Okamoto, T. Nakajima, Y. Ito, T. Aketo, K. Shimada, S. Yamato,
J. Pharm. Biomed. Anal. 2005, 37, 517.
Conclusions [11] V. V. Vaidya, M. Khanolkar, J. N. Gadre, Indian Drugs 2001, 38, 16.
[12] T. Harsono, M. Yuwono, G. Indrayanto, J. AOAC Int. 2005, 88, 1093.
UV–visible spectral data can be transformed in increasing [13] M. L. Wilcox, J. T. Stewart, J. Pharm. Biomed. Anal. 2000, 23, 909.
derivative orders so to obtain more useful information. Application [14] X. Xu, J. T. Stewart, J. Liq. Chromatogr. R. T. 2000, 23, 1.
of chemometric methods to derivative UV spectra can magnify [15] O. W. Lau, Y. M. Cheung, Analyst 1990, 115, 1349.
[16] L. A. Shervington, Anal. Lett. 1997, 30, 927.
the prediction ability of this spectrophotometric technique. [17] A. Beheshti, S. Riahi, E. Pourbasheer, M. R. Ganjali, P. Norouzi, J. Food
A comparative study on the application of the multivariate Sci. 2010, 75, C135.
calibration methods PCR, PLS1, and PLS2 on first order derivative [18] F. Faridbod, M.R. Ganjali, R. Dinarvand, S. Riahi, P. Norouzi, J. Food
spectra of two mixtures of PDP and GFN has been accomplished. Drug Anal. 2009;, 17, 264.
[19] S. Riahi, F. Faridbod, M.R. Ganjali, Sensor Lett. 2009, 7, 42.
According to the results obtained in this work, application of the [20] M. R. Ganjali, T. Razavi, R. Dinarvand, S. Riahi, P. Norouzi, Int. J.
UV spectrophotometric method is an effective and accurate way for Electrochem. Sci. 2008, 3, 1543.
the simultaneous determination of PDP and GFN in binary mixtures [21] M. R. Ganjali, T. Razavi, F. Faridbod, S. Riahi, P. Norouzi, Curr. Pharm.
by multivariate calibration of synthetic and pharmaceutical Anal. 2009, 5, 28.
samples. In addition, the application of this method is simple, [22] S. Riahi, M. R. Ganjali, E. Pourbasheer, F. Divsar, P. Norouzi, Curr.
Pharm. Anal. 2008, 4, 231.
fast, precise and affordable, requiring no complex pretreatment or [23] C. Bosh Ojeda, F. Sanchez Rojas, Anal. Chim. Acta 2004, 518, 1.
chromatographic the sample separations containing analytes. [24] J. Karpinska, Talanta 2004, 64, 801.
According to these studies, multivariate calibration methods [25] R. C. Tena, M. A. R. Delgado, M. A. J. Sanchez, F. G. Montelongo,
(PCR, PLS1, PLS2) coupled with derivative spectral data can Talanta 1997, 44, 673.
[26] N. B. Pappano, Y. C. De Micalizzi, N. B. Debattista, F. H. Ferretti,
be recommended as a very suitable choice to resolve severe Talanta 1997, 44, 633.
overlapped absorption spectra of drug mixtures. This approach is [27] S. Riahi, M. R. Ganjali, E. Pourbasheer, P. Norouzi, Curr. Pharm. Anal.
simple in application, inexpensive, requires an easy treatment of 2007, 3, 268.
the samples, and provides reliable analytical results. [28] E. Dinç, D. Baleanu, J. Pharm. Biomed. Anal. 2003, 31, 969.
[29] K. Karljikovic-Rajic, D. Novovic, V. Marinkovic, D. Agbaba, J. Pharm.
Biomed. Anal. 2003, 32, 1019.
Acknowledgement [30] C. Bosh Ojeda, F. Sanchez Rojas, J. M. Cano Pavon, Talanta 1995, 42,
1195.
We gratefully acknowledge the support of this work by the Institute [31] G. PopoviČ, M. Čakar, D. Agbaba, J. Pharm. Biomed. Anal. 2003, 33,
of Petroleum Engineering, Tehran University Research Councils. 131.
[32] L. Wang, M. Asgharnejad, J. Pharm. Biomed. Anal. 2000, 21, 1243.
[33] P. C. Damiani, G. M. Escandar, A. C. Olivieri, H. C. Goicoechea, Curr.
Pharm. Anal. 2005, 1, 145.
References [34] S. Riahi, M. R. Ganjali, E. Pourbasheer, F. Divsar, P. Norouzi,
M. Chaloosi, Curr. Pharm. Anal. 2008, 4, 231.
[1] I. Caraballo, M. Fernandez-Arevalo, M. A. Holgado, [35] S. Riahi, M. R. Ganjali, A. B. Moghaddam, E. Pourbasheer, P. Norouzi,
J. Alvarezfuentes, A. M. Rabasco Drug Dev. Ind. Pharm. 1995, Curr. Anal. Chem. 2009, 5, 42.
21, 605. [36] S. Riahi, K. Bagherzadeh, B. Akbari-Adergani, M. R. Ganjali,
[2] G. Halstead, J. Pharm. Sci. 1982, 71, 1108. P. Norouzi, Chem. Anal. 2009, 54, 1405.
[3] V. Galli, C. Barbas, J. Chromatogr. A 2004, 1048, 207. [37] S. Riahi, M. Hariri, M. R. Ganjali, P. Norouzi, Spectrosc. Lett. 2010, 43,
[4] Y. M. Dong, X. F. Chen, Y. L. Chen, X. G. Chen, Z. D. Hu, J. Pharm. 226.
Biomed. Anal. 2005, 39, 285. [38] M. C. F. Ferraro, P. M. Castellano, T. S. Kaufman, J. Pharm. Biomed.
[5] A. I. Gasco-Lopez, R. Izquierdo-Hornillos, A. Jiminez, J. Chromatogr. Anal. 2004, 34, 305.
A 1997, 775, 179. [39] A. Abbaspour, R. Mirzajani, J. Pharm. Biomed. Anal. 2005, 38, 420.
[6] ICH, Q1A Stability Testing of New Drug Substances and Products, [40] A. Afkhami, M. Bahram, Spectrochim. Acta. A 2005, 61, 869.
in Proceeding of the International Conference on Harmonization, [41] C. K. Markopoulou, E. T. Malliou, J. E. Koundourellis, IlFarmaco 2004,
Geneva, 1993. 59, 627.
[7] The United States Pharmacopoeia, 26th Ed., US Pharmacopoeial [42] M. I. Saleh, M. I. Meullenet, T. J. Siebenmorgen, Cereal Chem. 2008,
Convention, Rockville, MD, 2003, p. 1151. 85, 787.
324

[Link] Copyright

c 2011 John Wiley & Sons, Ltd. Drug Test. Analysis 2011, 3, 319–324