Analytica Chimica Acta 549 (2005) 212–220
Analytical study for the charge-transfer complexes
of losartan potassium
Ibrahim A. Darwish ∗
Department of Pharmaceutical Analytical Chemistry, Faculty of Pharmacy, Assiut University, Assiut 71526, Egypt
Received 15 April 2005; received in revised form 30 May 2005; accepted 7 June 2005
Available online 11 July 2005
Abstract
Studies were carried out, for the first time, to investigate the charge-transfer reactions of losartan potassium (LOS-K) as n-electron
donor with the ␴-acceptor iodine and various ␲-acceptors: 7,7,8,8-tetracyanoquinodimethane, 1,3,5-trinitrobenzene, 2,3-dichloro-5,6-dicyano1,4-benzoquinone, p-chloranilic acid, tetracyanoethylene, 2,3,5,6-tetrabromo-1,4-benzoquinone, 2,3,5,6-tetrachloro-1,4-benzoquinone, and
2,4,7-trinitro-9-fluorenone. Different colored charge-transfer complexes and radical anions were obtained. Different variables affecting the
reactions were studied and optimized. The formed complexes and the site of interaction were examined by UV–vis, IR, and 1 H NMR
techniques, and computational molecular modeling. The formation of the colored complexes were utilized in the development of simple,
rapid and accurate spectrophotometric methods for the analysis of LOS-K in pure form as well as in its pharmaceutical tablets. Under the
optimum reaction conditions, linear relationships with good correlation coefficients (0.9985–0.9998) were found between the absorbances
and the concentrations of LOS-K in the range of 2–200 ␮g ml−1 . The limits of assays detection ranged from 0.61 to 19.65 ␮g ml−1 . No
interference could be observed from the co-formulated hydrochlorothiazide (HCTZ), as well as from the additives commonly present in the
tablets. The methods were successfully applied to the analysis of tablets from different manufacturers that contain LOS-K, alone or combined
with HCTZ, with good accuracy and precision; the recovery percentages ranged from 98.96 ± 1.62% to 101.58 ± 1.29%. The results were
compared favourably with the reported method.
© 2005 Elsevier B.V. All rights reserved.
Keywords: Losartan potassium; Charge-transfer complexes; Spectrophotometry; Pharmaceutical analysis
1. Introduction
Losartan potassium (LOS-K); 2-butyl-4-chloro-1-[p-(o1H-tetrazol-5-ylphenyl)benzyl-imidazole-5-methanol potassium, is a new non-peptide angiotensin II receptor antagonist
with antihypertensive activity, due mainly to selective blockade on AT1 receptors and the consequent reduced pressor
effect of angiotensin II [1]. LOS-K is used in the management
of hypertension with a lower incidence of side effects such as
cough, which develops with typical angiotensin-converting
enzyme inhibitors. It can be used alone or combined with the
diuretic hydrochlorothiazide (HCTZ) in patients with moderate heart failure [2].
∗
Tel.: +20 88 2411251; fax: +20 88 2332776.
E-mail address: [email protected].
0003-2670/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.aca.2005.06.023
The methods that have been reported for the analysis of
LOS-K in pure form or in biological fluids include high
performance liquid chromatography [3–7], thin layer chromatography [8], gas chromatography [9], capillary electrophoresis [10,11], and radioreceptor assay [12]. These
I.A. Darwish / Analytica Chimica Acta 549 (2005) 212–220
methods are not simple to perform, time-consuming, and/or
utilize expensive instruments that are not available in
most quality control laboratories. Spectrophotomeric analysis, being simple and common technique, is a convenient
approach. However, few spectrophotometric methods have
been reported for the analysis of LOS-K [13–15]. Unfortunately, these methods were unable to achieve the direct
analysis of LOS-K when present in combination with HCTZ
in the dosage form. Therefore, the aim of the present study
was directed to the development of simple and sensitive spectrophotometric methods for the selective analysis of LOS-K
in its pharmaceutical tablets, irrespective of the presence of
HCTZ.
The molecular interactions between electron donors and
electron acceptors are generally associated with the formation of intensely colored charge-transfer complexes, which
absorb radiation in the visible region [16]. A variety of
electron donating compounds have been reported to yield
charge-transfer complexes with various acceptors. The rapid
formation of these complexes leads to their utility in the
development of simple and convenient spectrophotometric
methods for these compounds [17–19]. The charge-transfer
reaction has not been reported yet for LOS-K, therefore the
aim of the present study was directed to investigate this reaction.
2. Experimental
2.1. Apparatus
A Lambda-3 B (Perkin-Elmer Corporation, Norwalk,
USA) and UV-1601 PC (Shimadzu, Japan) ultraviolet–visible
spectrophotometers with matched 1-cm quartz cells, were
used for all measurements. An infrared spectrometer (IR-470,
Shimadzu, Japan) and a 1 H NMR spectrometer (EM-360,
60 MHz, Varian Instrument Division, Palo, Alto, CA, USA)
were also used.
2.2. Chemicals and reagents
Losartan potassium (Hetero Drugs Ltd., Hyderabad,
India) was used as working standard. Hydrochlorothiazide
was obtained from Kahira Co., Egypt. Iodine, resublimed
(Riedel-De-Haen AG, Germany), was 4 mg ml−1 in 1,2dichloroethane. The solution was stable for at least 1 week
at 4 ◦ C. 7,7,8,8-Tetracyanoquinodimethane (TCNQ; Aldrich
Chem. Co, Milwaukee, USA) was 1 mg ml−1 in acetonitrile,
and the solution was stable for at least 1 week at 4 ◦ C.
1,3,5-Trinitrobenzene (TNB; Prodotti Chimiol Pet Analist
Scientifico, Milano, Italy) was 4 mg ml−1 in acetonitrile,
and the solution was prepared fresh daily. 2,3-Dichloro5,6-dicyano-1,4-benzoquinone (DDQ; Merck, Schuchardt,
Munich, Germany) was 2 mg ml−1 in methanol and it
was prepared fresh daily. 2,5-Dichloro-3,6-dihydroxy-1,4benzoquinone (chloranilic acid, pCA) from BDH Chemicals
213
(Poole, UK) was 4 mg ml−1 in acetonitrile, and the solution
was prepared fresh daily. Tetracyanoethylene (TCNE;
Nacalai Tesque, Kyoto, Japan) was 2 mg ml−1 in acetontrile,
and it was prepared fresh daily. 2,3,5,6-Tetrabromo-1,4benzoquinone (bromanil; Hopkin & Williams Ltd., UK),
and 2,3,5,6-tetrachloro-1,4-benzoquinone (chloranil; Sigma
Chemical Co, St. Louis, USA) were 5 mg ml−1 in acetonitrile,
and the solutions were prepared fresh daily. 2,4,7-Trinitro9-fluorenone (TNF; Fluka, Switzerland) was 5 mg ml−1 in
acetonitrile was prepared fresh daily. All solvents and other
chemicals used throughout this study were of analytical
grade.
2.3. Pharmaceutical tablets
The commercial tablets used in the present investigation
were Cozaar® tablets (Merck Sharp & Dohme, Haarlem, The
Netherlands), labeled to contain 50 mg LOS-K per tablet.
Kanzar® tablets (Alkan Pharma, S.A.E., Egypt), labeled to
contain 25 mg LOS-K per tablet. Losartan® tablets (Amriya
Pharmaceutical Industries, Alexandria, Egypt), labeled
to contain 50 mg LOS-K per tablet. Lozapress® tablets
(Sigma Pharmaceutical Industries, Egypt), labeled to contain
25 mg LOS-K per tablet. Hyzzar® tablets (Merck Sharp
& Dohme, Haarlem, The Netherlands), labeled to contain
50 mg LOS-K and 12.5 mg HCTZ per tablet. Lozapress
H® tablets (Sigma Pharmaceutical Industries, Egypt),
labeled to contain 50 mg LOS-K and 12.5 mg HCTZ per
tablet.
2.4. Preparation of standard and tablets sample
solutions
2.4.1. Preparation of stock standard solution
Into a 50-ml calibrated flask, 20–100 mg of LOS-K
was accurately weighed and dissolved in 2 ml methanol,
completed to volume with the same solvent (for DDQ),
with 1,2-dichloroethane (for iodine), and with acetonitrile
(for the other acceptors). These stock solutions were diluted
with the respective solvents to obtain suitable concentrations that lie in the linear range of each particular assay
method.
2.4.2. Preparation of tablets sample solution
Twenty tablets of each formulation were weighed
and finely powdered. A quantity of the powder equivalent to 100 mg was transferred into a 50-ml calibrated
flask, dissolved in 2 ml methanol, swirled and sonicated
for 5 min, completed to volume with the corresponding solvent (as in stock solutions), shaken well for
15 min, and filtered. The first portion of the filtrate was
rejected, and a measured volume of the filtrate was diluted
quantitatively with a suitable solvent to yield suitable
concentrations lie in the linear range of each particular assay
method.
214
I.A. Darwish / Analytica Chimica Acta 549 (2005) 212–220
Table 1
Optimum conditions for the charge-transfer reaction of losartan potassium with different acceptors
Acceptor
(1.7)a
Iodine
TCNQ (1.7)
TNB (0.7)
DDQ (1.9)
pCA
TCNE (2.2)
Bromanil (1.37)
Chloranil (1.37)
TNF (1.1)
a
Reagent concentration (mg ml−1 )
Solvent
Time (min)
λmax (nm)
4
1
4
2
4
2
5
5
5
1,2-Dichloroethane
Acetonitrile
Acetonitrile
Methanol
Acetonitrile
Acetonitrile
Acetonitrile
Acetonitrile
Acetonitrile
At once
15
30
At once
At once
15
5
5
60
365
842
438
460
520
414
500
463
413
Figures in parenthesis are the electron affinities of the acceptors; from Ref. [16].
2.5. General analytical procedure
One milliliter of the standard or sample solution of LOSK (20–2000 ␮g ml−1 ) was transferred into 10-ml calibrated
flasks. One milliliter of the acceptor solution was added, and
the reaction was allowed to proceed at room temperature
(25 ± 5 ◦ C) for 5 min (in case of bromanil and chloranil),
15 min (in case of TCNQ and TCNE), 30 min (in case of
TNB), and for 60 min (in case of TNF). The reactions in case
of iodine, DDQ, and pCA were achieved instantaneously. The
solutions were diluted to volume with 1,2-dichloroethane (for
iodine), methanol (for DDQ), and with acetonitrile (for the
other acceptors). The absorbances of the resulting solutions
were measured at the wavelengths of maximum absorption
(365, 842, 438, 460, 520, 414, 500, 463, and 413 nm for
iodine, TCNQ, TNB, DDQ, pCA, TCNE, bromanil, chloranil, and TNF, respectively) against reagent blanks treated
similarly.
up comprising different complementary proportions (0:10,
1:9,. . ., 9:1, 10:0, inclusive) in 10-ml calibrated flasks. The
reactions were allowed to proceed for the optimum reaction
time (Table 1) and then the absorbances of the resulting
solutions were measured at the corresponding wavelengths
of maximum absorbances (λmax ).
2.8. Preparation of the complexes for infrared
measurements
To 2 ml of 0.1 M methanolic LOS-K solution, 2 ml of
0.1 M of each acceptor in the appropriate solvent (methanol
for DDQ, 1,2-dichloroethane for iodine, and acetonitrile
for other acceptors), was added in a round-bottom flask
containing ∼30 ml of the appropriate solvent and stirred for
30 min. Solvent was evaporated under reduced pressure, and
the resulting residues were dried over calcium chloride. The
dried residues were used for IR measurements.
2.6. Preparation of losartan from losartan potassium
2.9. Solutions for 1 H NMR measurements
An accurately weighed amount (500 mg) of LOS-K was
dissolved in 5 ml distilled water in a test tube. The solution was transferred quantitatively into a 100-ml separating
funnel, then rendered acidic with 10% HCl. The liberated
losartan (LOS) was filtered off and dried at room temperature
(25 ± 5 ◦ C) over anhydrous calcium chloride. The obtained
residue was used in studying the possible site of n-electron
charge transfer in LOS-K.
A quantity of 0.1 mmol (46 mg) of LOS-K was dissolved
in 1 ml of d6 -DMSO, and 1 ml containing an equimolar
amount of the acceptor in the same solvent was added and
used directly for 1 H NMR measurements.
3. Results and discussion
3.1. Spectral characteristics of the reaction
2.7. Determination of molar ratio
The Job’s method of continuous variation [20] was
employed. Master equimolar solutions of LOS-K and
reagents were prepared. The concentrations of these
solutions were 4.9 × 10−3 M (in acetonitrile for TCNQ),
1.9 × 10−2 M (in acetonitrile for TNB), 8.8 × 10−3 M (in
methanol for DDQ), 1.9 × 10−2 M (in acetonitrile for pCA),
1.6 × 10−2 M (in acetonitrile for TCNE), 1.2 × 10−2 M (in
acetonitrile for bromanil), and 2 × 10−2 M (in acetonitrile
for chloranil). Series of 10-ml portions of the master
solutions of LOS-K with the respective reagent were made
3.1.1. Reaction with σ-acceptor (iodine)
The color of iodine in 1,2-dichloroethane is violet
showing absorption maximum (λmax ) at 500 nm. This
color was immediately changed into lemon yellow, and
the absorption spectrum of LOS-K-iodine reaction product
showed absorption peaks at 290 and 365 nm (Fig. 1).
This change in color, and the appearance of these two
peaks were attributed to the formation of charge-transfer
complex between the LOS-K and iodine, having an ionized
structure DI+ . . . I3 − , taking into account that the absorption
spectrum of I3 − in 1,2-dichloroethane showed the same
I.A. Darwish / Analytica Chimica Acta 549 (2005) 212–220
Fig. 1. Absorption spectra of 50 ␮g ml−1 LOS-K (1), and the reaction
product of 100 ␮g ml−1 LOS-K with iodine (2). Solutions were in 1,2dichloroerthane.
absorption maximum of the reaction product (two absorption
maxima at 290 and 365 nm of molar absoptivity (ε) = 3792
and 2070 l mol−1 cm−1 , respectively). This complex should
originate from an early intermediated outer complex D. . .I2 :
D + I2 D–I+ I− [D–I+ ] +I− outer complex
inner complex
I−
3
tri−iodide ion pair
Further confirmation for the charge-transfer nature of the
reaction, LOS-K was extracted from the complex by shaking
with aqueous mineral acids. The color of iodine in 1,2dichloroethane layer was restored to violet, confirming the
charge-transfer nature of the reaction. Measurements were
carried out at 365 nm due to the interference from the native
UV absorption of LOS-K at 290 nm.
3.1.2. Reaction with π-acceptors
The interaction of LOS-K with polyhaloquinone and
polycyanoquinone ␲-acceptors in non-polar solvents such
as dichloroethane was found to produce colored chargetransfer complexes with low molar absorptivity values. In
polar solvents such as methanol or acetonitrile, complete
electron transfer from the LOS-K (D), as an electron donor,
to the acceptor moiety (A) takes place with the formation of
intensely colored radical ions with high molar absorptivity
values, according to the following scheme:
D + A (D–A)
complex
polar solvent
215
Fig. 2. Absorption spectra of the reaction products of LOS-K with each of
TNB (1), DDQ (2) and TCNQ (3). Concentrations of LOS-K were 60, 50,
and 50 ␮g ml−1 in case of TNB, DDQ, and TCNQ, respectively. Solutions
were in methanol in case of DDQ, and in acetonitrile in case of TNB and
TCNQ.
The predominent chromogen with TCNQ in acetonitrile is
the bluish-green colored radical anion, which exhibits strong
absorption maxima at 842, 825, 762, and 742 nm. These
bands may be attributed to the formation of the radical anion
TCNQ•− , which was probably formed by the dissociation of
an original donor–acceptor (D–A) complex with LOS-K. The
dissociation of the complex was promoted by the high ionizing power of acetonitrile. Further support of this assignment
was provided by the absorption maxima with those of TCNQ
radical anion produced by the iodide reduction method [21].
The complex of LOS-K with TNB showed two absorption
maxima at 438 and 552 nm. The intensity of the first maximum is about 1.75-fold the second one. Therefore, measurements were carried out at 438 nm, at which higher sensitivity
was achieved.
Chloranilic acid (pCA) exists in three ionic forms, the
neutral yellow-orange H2 A at very low pH, the dark purple
HA− which is stable at pH 3 and a colorless A2− , which is
stable at high pH; these transformations are illustrated in the
D•+ + A•−
radical ions
The dissociation of the (D–A) complex was promoted by
the high ionizing power of the polar solvent and the resulting
peaks in the absorption spectra of LOS-K-acceptor reaction
mixtures were similar to the maxima of the radical anions of
the acceptors obtained by the iodide reduction method [21].
The interaction of LOS-K with ␲-acceptors at room temperature gave colored chromogens showing different absorption maxima at 842, 438, 460, 520, 414, 500, 463, and 413 nm
for TCNQ, TNB, DDQ, pCA, TCNE, bromanil, chloranil,
and TNF, respectively (Figs. 2 and 3).
Fig. 3. Absorption spectra of the reaction products of LOS-K with each of
TCNE (1), bromanil (2), and pCA (3). Concentrations of LOS-K were 200,
300, and 75 ␮g ml−1 in case of TCNE, bromanil, and pCA, respectively.
Solutions were in acetonitrile.
216
I.A. Darwish / Analytica Chimica Acta 549 (2005) 212–220
following scheme:
H2 A H+ + HA− (violet),
Table 2
Effect of solvents on the position and intensity of absorption of the reaction
of losartan potassium (100 ␮g ml−1 ) with pCA (4 mg ml−1 )
HA− H+ + A2− (colorless).
Solvent
Dielectric
constanta
λmax
(nm)
Absorbance
Since the interaction of LOS-K with pCA in acetonitrile
gave a violet product, it might be concluded that HA− was
the form of pCA involved in the reaction described herein.
With TCNE, the characteristic shaped absorption band
of TCNE radical anion with reported maximum in acetonitrile at 432 nm was not found. Instead, a duplet at 394
and 414 nm was formed which corresponds to the 1,1,2,3,3pentacyanopropeneide (PCNP) anion, which is more preferable than TCNE anion, in quantitative analysis, in having
higher molar absorptivity [21]. The resulting maxima of
LOS-K with DDQ, bromanil, chloranil, and TNF are similar to that of radical anions of these acceptors obtained by
the reduction method and coincide with the values reported
in the literature [22,23].
The relative sensitivity of the nine acceptors employed in
the present analytical work may be attributed to their difference in electron affinities (Table 1), as well as the conditions
employed in the reaction (reagent concentration, reaction
time, and solvent). The weak and small molar absorptivity values in cases of bromanil, chloranil, and TNF may be
explained on the basis of insufficient ionization of these relatively weak ␲-acceptors that possess lower electron affinities,
than TCNQ and DDQ, exhibiting higher molar absorptivities.
Because of the poor sensitivity of TNF, it has been excluded
from further investigations.
Acetonitrile
Methanol
Ethanol
Acetone
Propan-1-ol
Propan-2-ol
Butan-1-ol
Diethylether
Benzene
Carbon tetrachloride
Toluene
1,4-Dioxane
Xylene
37.5
32.7
24.6
20.7
20.3
19.9
17.5
4.3
2.3
2.2
2.4
2.2
2.0
520
527
525
525
525
530
525
530
535
520
535
520
520
0.614
0.539
0.464
0.423
0.435
0.433
0.408
0.331
0.363
0.336
0.318
0.291
0.287
3.2. Optimization of reaction conditions
3.2.1. Effect of reagent concentration
The results of variations in the reagents concentrations
indicated that 1 ml of the concentrations indicated in Table 1
were the optimum concentrations. The higher concentrations
of the reagents may be useful for rapidly reaching equilibrium, thus minimizing the time required to attain maximum
absorbance at the corresponding wavelengths of maximum
absorbance.
3.2.2. Effect of solvent
In order to select the most appropriate solvent, the reactions were carried out in different solvents. Small shifts in the
position of the maximum absorption peak were observed,
and the absorption intensities were also influenced. 1,2Dichloroethane was found to be an ideal solvent in case of
iodine (Table 1), because it is favourable for the formation
of tri-iodide ion pair (inner complex). Methylene chloride,
chloroform and carbon tetrachloride produced lower absorption readings. Polar solvents were found to be unsuitable
as their blanks with iodine gave higher readings. Methanol
gave maximum sensitivity in case of DDQ. Acetonitrile was
considered as an ideal solvent for the other ␲-acceptors.
This because it offered maximum sensitivity, which was
a
Ref. [24].
attributed to the high dielectric constant of acetonitrile that
promotes maximum yield of radical anions, in addition to
its high solvating power for the acceptors [24]. The results
obtained with pCA, as a representative example, is given in
Table 2.
3.2.3. Effect of reaction time
The optimum reaction time was determined by monitoring the color development at room temperature (25 ± 5 ◦ C).
Complete color development was attained instantaneously
with iodine, DDQ, and pCA, or after 5–60 min with other
acceptors (Table 1). The developed colors remained stable at
room temperature for at least a further 30 min, except in case
of iodine whereas the color decreased dramatically resulting
in high imprecision of the readings.
3.3. Spectroscopic investigations for the structure of the
charge-transfer complexes
The structures of donor–acceptor complexes were investigated by both IR and 1 H NMR spectroscopic techniques. The
majority of IR measurements on the complexes have been
concerned with shifts in the vibrational frequencies in the
donor, acceptor and/or both. The IR spectra of the complexes
showed some differences compared with the sum of the spectra of the two components. These differences have been used
to distinguish between weak charge-transfer complexes and
the products of electron transfer or proton-transfer reactions
[16]. The IR spectrum of the complex formed with iodine was
identical to the spectrum of LOS-K; this was attributed to the
fact that iodine is IR inactive. In the IR spectra of the complexes between LOS-K and various acceptors, decreases in
the vibrational frequency of a particular band were observed
(Table 3). These shifts were used as evidence for the formation of charge-transfer complexes.
In the 1 H NMR spectra of the complexes, generally, the
protons of the donor are shifted to down field [17–19]. The
1 H NMR spectra of the complexes of LOS-K with different
I.A. Darwish / Analytica Chimica Acta 549 (2005) 212–220
217
Table 3
Characteristic bands of infrared spectra of the acceptors and their corresponding charge-transfer complexes with LOS-K
Acceptor
TCNE
TCNQ
TNB
DDQ
pCA
Chloranilc
a
b
c
Frequency (cm−1 )
Stretching group
C N, C O
C N, C C, DSBb
Aryl NO2 , C N
C N, C O
O H, C O, DSBb
C O, C Cl, DSBb
Acceptor
Complex
2195, 1678
2222, 1629, 853
1530, 912
2210, 1665
3225, 1664, 844
1673, 1104, 879
2190, 1627a
2185, 1607, 814
1521, 906
2190, 1627
3210, 1651, 830
1667, 980, 763
Overlapped with the band at the same frequency corresponding to N = N stretch of LOS-K.
DSB is 1,4-disubstituted benzene.
Similar results were obtained with bromanil.
acceptors were recorded in d6 -DMSO and compared with
the spectrum of the free drug (Table 4). The 2-butyl protons
(δ = 0.8, 1.4, and 2.5 ppm) were not affected indicating that
imidazole nitrogen atoms might not contribute by their lone
pairs of electrons in the electron donation. The CH2 protons
of benzyl moiety and hydroxymethyl (δ = 5.5 and 4.6 ppm,
respectively) were slightly downfield shifted (δ = 0.1 ppm).
The aromatic protons (δ = 7.7 and 7.3 ppm) were obviously
downfield shifted (δ = 0.1–0.3 ppm). These results suggested that the electron-donating site in LOS-K is close to
the aromatic protons; most probably the anionic tetrazole
moiety.
3.4. Molar ratio of the reaction and the proposed site of
interaction
Job’s method of continuous variation [20] was used for
determining the molar ratio of LOS-K to each of the analytical reagent employed in the charge-transfer reactions. These
ratios were 1:1 in all cases. This indicates that only one site
is possible for the formation of the complex and a univalent
charged species, is the possible site of the charge-transfer
process. This site, considering 1 H NMR results, was postulated to be the anionic tetrazole moiety of LOS-K. To prove
this suggestion, the charge-transfer reactions were carried out
using the parent losartan (LOS). It was found that LOS did
not exhibit any electron-donating ability indicating that the
anionic tetrazole moiety of LOS-K, but not LOS, is the site
Table 4
1 H NMR spectra of LOS-K and its charge-transfer complexes with various
acceptors
δ, ppm (d6 -DMSO)
LOS-K
7.7 (m, 4H)
7.3 (m, 4H)
5.5 (s, 2H)
4.6 (s, 2H)
2.5 (t, 2H)
1.4 (m, 4H)
0.8 (t, 3H)
Charge-transfer complexes with
TCNE
TCNQ
DDQ
Chloranil
7.9
7.4
5.5
4.6
2.5
1.4
0.8
7.9
7.5
5.6
4.6
2.5
1.4
0.8
8.0
7.5
5.5
4.7
2.5
1.4
0.8
8.0
7.6
5.7
4.7
2.5
1.4
0.8
of charge-transfer interaction with the acceptors. This was
further supported by the acidic behavior of LOS (pKa = 5–6)
[26].
3.5. Molecular modeling for the charge-transfer
complexes
The molecular modeling for the charge-transfer complexes were performed by using CS Chem3D Ultra, version
8 (CambridgeSoft Corporation, Cambridge, MA, USA)
implemented with molecular orbital computations software
(MOPAC), and molecular dynamics computations software
(MM2). The LOS-K and the acceptor (TCNE and TCNQ
as representative examples) were minimized alone and both
together. It was found that both TCNE and TCNQ attack
LOS-K at the area of tetrazole anion. LOS-K has three
possible donating entities (anionic tetrazole ring, and the two
nitrogen atoms of the imidazole ring). It is acceptable that
certain electron density was required for achievement of a
successful electron transfer [16]. The total charge on each particular site was computed, and these were −0.1418, −0.2780,
and −0.5368 for N1-imidazole, N3-imidazole, and anionic
tetrazole ring, respectively. These results, in conjunction with
the LOS-K:acceptor molar ratio (1:1) and the 1 H NMR data,
proved that the source of electron donation can be better from
the anionic tetrazole than nitrogen atoms of the imidazole
ring.
3.6. Development and validation of the analytical
methods
3.6.1. Calibration curves, linearity and sensitivity
Under the specified optimum reaction conditions, the
calibration curves for LOS-K with the different analytical
reagents employed in the present work were constructed. The
regression equations for the results were derived using the
least-squares method. In all cases, Beer’s law plots (n = 5)
were linear with very small intercepts and good correlation coefficients in the general concentration range of 2200 ␮g ml−1 (Table 5). The limits of detection (LOD) and
limits of quantitation (LOQ) were determined [27] using the
formula: LOD or LOQ = ␬S.D.a/b, where ␬ = 3 for LOD and
218
I.A. Darwish / Analytica Chimica Acta 549 (2005) 212–220
Table 5
Quantitative parameters for the charge-transfer reaction of losartan potassium with various acceptors
Acceptor
Range (␮g ml−1 )
Intercept (a)
Slope (b)
Correlation
coefficient (r)
Molar absorptivity ε
(×103 l mol−1 cm−1 )
LOD (␮g ml−1 )
LOQ (␮g ml−1 )
Iodine
TCNQ
TNB
DDQ
pCA
TCNE
Bromanil
Chloranil
10–200
2–60
6–60
8–80
10–120
20–140
40–180
60–180
0.0199
0.0021
0.0167
0.0147
0.0202
0.0063
0.0075
0.0169
0.0044
0.0408
0.0125
0.0262
0.0064
0.0037
0.0023
0.0014
0.9992
0.9994
0.9996
0.9986
0.9992
0.9998
0.9990
0.9985
2.07
7.72
5.19
4.45
2.88
1.75
1.86
0.71
2.91
0.61
1.82
2.28
2.75
6.12
11.90
19.65
8.72
1.84
5.45
6.85
8.24
18.37
35.72
58.96
10 for LOQ, S.D.a is the standard deviation of the intercept,
and b is the slope. Based on the basis of six replicate measurements, the limits of detection were 0.61–19.65 ␮g ml−1 .
3.6.2. Precision
The precisions of the assays (within-assay and betweenassays) were determined at the LOS-K concentrations cited
in Table 6. The within-assay precision was assessed by analyzing six replicates of each sample as a batch in a single
assay run, and the between-assays precision was assessed by
analyzing the same sample, as triplicate, in two separate assay
runs. The assays, except iodine method, gave satisfactory
results; the relative standard deviations (R.S.D.) were less
than 2% (Table 6). This level of precision of the proposed
methods was adequate for the quality control analysis of
LOS-K. Owing to the high imprecision of the iodine method
(R.S.D. > 10%), this method was excluded from further investigations.
3.6.3. Specificity and interference
The proposed spectrophotometric methods have the
advantages that the measurements in all of these methods
are performed in the visible region, away from the UVabsorbing interfering substances that might be co-extracted
from LOS-K-containing dosage forms. The interference from
the congenital HCTZ that is co-formulated with LOS-K in
some of its dosage forms was studied in a ratio, which is
normally present in the dosage form. The good percentage
recoveries (Table 7) revealed that there was no interference
from HCTZ with the proposed methods. This specificity of
the charge-transfer reaction for LOS-K was attributed to its
basic character, which allows the charge transfer, rather than
Table 6
Precision of the proposed methods for analysis of losartan potassium (LOS-K)
Method
LOS-K (␮g ml−1 )
Within-assay, n = 6
Between-assays, n = 6
Mean (␮g ml−1 ± S.D.)
Iodine
TCNQ
TNB
DDQ
pCA
TCNE
Bromanil
Chloranil
100
25
40
50
80
100
200
300
98.58
24.78
40.73
50.51
78.78
100.57
201.57
298.78
±
±
±
±
±
±
±
±
9.58
0.16
0.23
0.52
0.41
0.75
0.84
2.69
R.S.D.
Mean (␮g ml−1 ± S.D.)
9.72
0.95
0.56
1.03
0.52
0.75
0.42
0.90
96.98
25.04
39.55
49.56
80.05
98.96
198.24
301.05
±
±
±
±
±
±
±
±
14.36
0.29
0.35
0.76
0.84
1.51
1.81
4.08
R.S.D.
14.81
1.16
0.88
1.53
1.05
1.53
0.91
1.36
Table 7
Analysis of losartan potassium in the presence of hydrochlorothiazide (HCTZ) and common excipients by different methods
Ingredient
Recoverya (% ± S.D.)
TCNQ
TNB
DDQ
pCA
TCNE
Bromanil
Chloranil
(12.5)
Starch (50)
Glucose (10)
Lactose (10)
Acacia (10)
MSc (10)
99.62 ± 0.82
100.05 ± 1.01
99.81 ± 0.54
100.22 ± 0.85
99.64 ± 1.02
100.10 ± 0.024
100.02 ± 0.56
99.98 ± 0.68
98.52 ± 1.05
100.25 ± 0.58
101.07 ± 0.87
100.13 ± 1.03
98.56 ± 0.35
99.26 ± 0.56
99.58 ± 1.03
101.26 ± 1.05
102.85 ± 0.84
100.84 ± 0.95
101.53 ± 1.02
99.46 ± 0.46
99.58 ± 1.23
100.37 ± 0.87
98.84 ± 0.39
97.97 ± 1.26
102.36 ± 0.82
98.76 ± 0.59
99.52 ± 1.04
101.49 ± 1.15
103.02 ± 0.95
99.01 ± 0.76
100.98 ± 1.24
98.65 ± 1.05
99.59 ± 0.87
98.71 ± 0.4
99.36 ± 1.25
97.02 ± 1.81
102.84 ± 0.58
98.47 ± 1.10
100.87 ± 0.85
102.21 ± 1.64
101.58 ± 1.21
100.43 ± 0.98
Average ± S.D.
Pool S.D.
99.91 ± 0.38
0.71
100.00 ± 0.83
0.80
100.39 ± 1.57
0.80
99.63 ± 1.23
0.87
100.69 ± 1.83
0.89
99.05 ± 1.30
1.19
101.07 ± 0.50
1.20
HCTZb
a
b
c
Values are mean of three determinations.
Figures in parenthesis are the amounts in mg added per 50 mg of losartan potassium.
MS = magnesium stearate.
I.A. Darwish / Analytica Chimica Acta 549 (2005) 212–220
219
Table 8
Analysis of losartan potassium in its tablets by different methods
Method
TCNQ
TNB
DDQ
pCA
TCNE
Bromanil
Chloranil
Reported c
a
b
c
Label claima (% ± S.D.)
Cozaar® tablets
(50 mg/tablet)
Kanzar® tablets
(25 mg/tablet)
Losartan® tablets
(50 mg/tablet)
Lozapress® tablets
(25 mg/tablet)
Hyzzar® tabletsb
(50 mg/tablet)
Lozapress H® tabletsb
(50 mg/tablet)
100.85 ± 1.75
t: 1.02; F: 3.47
100.47 ± 0.96
t: 0.79; F: 1.04
100.21 ± 1.63
t: 0.66; F: 3.01
100.82 ± 1.85
t: 2.00; F: 3.87
99.82 ± 1.95
t: 2.16; F: 4.30
99.86 ± 1.69
t: 2.24; F: 3.23
100.56 ± 1.92
t: 0.87; F: 4.17
100.35 ± 0.94
100.58 ± 1.25
t: 1.55; F: 1.34
100.47 ± 1.28
t: 2.11; F: 1.40
101.22 ± 1.62
t: 1.59; F: 2.25
100.82 ± 1.78
t: 0.21; F: 2.72
100.45 ± 1.95
t: 1.67; F: 3.26
100.68 ± 1.28
t: 1.00; F: 1.40
100.39 ± 1.48
t: 2.31; F: 1.88
100.87 ± 1.08
101.18 ± 1.77
t: 1.77; F: 4.34
99.47 ± 1.26
t: 2.62; F: 2.20
100.21 ± 1.43
t: 1.54; F: 2.83
99.82 ± 1.05
t: 0.65; F: 1.53
100.52 ± 1.94
t: 2.50; F: 5.21
99.34 ± 1.86
t: 2.51; F: 4.79
100.26 ± 1.04
t: 2.24; F: 1.50
99.92 ± 0.85
99.58 ± 1.77
t: 1.55; F: 2.41
100.29 ± 1.32
t: 1.72; F: 1.34
100.21 ± 1.43
t: 1.26; F: 1.57
99.82 ± 0.78
t: 0.83; F: 0.47
99.54 ± 1.58
t: 1.86; F: 1.92
100.09 ± 1.69
t: 0.61; F: 2.20
100.46 ± 1.82
t: 2.10; F: 2.55
99.95 ± 1.14
100.89 ± 1.17
t: 0.73; F: 3.38
100.85 ± 1.56
t: 0.97; F: 2.81
100.81 ± 1.04
t: 1.55; F: 1.18
100.49 ± 1.84
t: 2.40; F: 3.91
101.34 ± 1.58
t: 1.40; F: 2.89
100.86 ± 1.26
t: 1.07; F: 1.84
101.58 ± 1.92
t: 2.20; F: 4.26
101.05 ± 0.93
99.95 ± 1.77
t: 1.55; F: 1.97
98.97 ± 1.85
t: 2.37; F: 2.16
100.05 ± 1.43
t: 2.23; F: 1.29
100.04 ± 1.78
t: 1.90; F: 2.00
99.98 ± 1.58
t: 1.79; F: 1.57
98.96 ± 1.62
t: 2.63; F: 1.65
98.99 ± 1.57
t: 2.55; F: 1.55
99.57 ± 1.26
Values are mean of five determinations. The tabulated values of t and F at 95% confidence limit are 2.78 and 6.39, respectively.
Combined with 12.5 mg of hydrochlorothiazide.
Ref. [14].
HCTZ, that does not have sufficient basicity to achieve charge
transfer (pKa = 7.9) [25]. Potential interference by the excipients in the dosage forms was also studied. Samples were
prepared by mixing known amount (50 mg) of LOS-K with
various amounts of the common excipients such as starch,
glucose, lactose, acacia, and magnesium stearate. The results
(Table 7) revealed that no interference was observed from any
of these excipients with the proposed methods. The absence
of interference from these excipients was attributed to the
extraction with organic solvent prior to the analysis.
3.6.4. Ruggedness and robustness
The ruggedness of the proposed methods was assessed
by applying the procedures using two different instruments
(shown in Section 2) in two different laboratories at different
elapsed time. Results obtained from lab-to-lab and day-today variation were found to be reproducible as R.S.D. did
not exceed 2%. Robustness of the procedures was assessed
by evaluating the influence of small variation of experimental variables: concentrations of acceptor reagent, and reaction
time, on the analytical performance of the method. In these
experiments, one experimental parameter was changed while
the other parameters were kept unchanged, and the recovery
percentage was calculated each time. The small variations in
any of the variables did not significantly affect the results;
recovery percentages were 98.2–102.5% ± 0.84–1.82. This
provided an indication for the reliability of the proposed
methods during routine work.
3.7. Application of the method to the analysis of tablets
The obtained satisfactory validation results made the proposed procedures suitable for the routine quality control analysis of LOS-K. The proposed and the reported methods [14]
were applied to the determination of LOS-K in its tablets. The
results obtained by the proposed methods were statistically
compared with those obtained by the reported method. The
obtained mean values of the labeled amounts ranged from
98.96 ± 1.62% to 101.58 ± 1.29% (Table 8). In the t- and
F-tests, no significant differences were found between the
calculated and theoretical values of both the proposed and
the reported methods at 95% confidence level. This indicated
similar precision and accuracy in the analysis of LOS-K in its
tablets. It is evident from these results that all the proposed
methods are applicable to the analysis of LOS-K in its tablets
with comparable analytical performance. However, the critical recommendations of some of these methods might be
based on the experimental conditions (e.g. reaction time),
and the ultimate sensitivity that determines the amount of
specimen required for analysis. For example, the methods
involving DDQ, pCA, bromanil, and chloranil are recommended whenever rapid analysis is required; this because
they have very short reaction time. The method involving
TCNQ is recommended, as high sensitivity is required on the
expense of the analysis time.
3.8. Conclusions
The charge-transfer complexation reaction of losartan
potassium (LOS-K) as electron donor and some electron
acceptors has been investigated. The obtained complexes
were studied by ultraviolet–visible spectrophotometry, IR,
and 1 H NMR spectroscopic techniques, and by computational molecular modeling. The obtained colored complexes
were utilized in the development of seven simple, rapid
and accurate spectrophotometric methods for the analysis
of LOS-K in pure form as well as in tablets. The proposed
methods are superior to the previously reported UV-based
220
I.A. Darwish / Analytica Chimica Acta 549 (2005) 212–220
spectrophotometric methods, as the measurements are performed in the visible region, away from the UV-absorbing
congenital hydrochlorothiazide or interfering excipients
that might be co-extracted from LOS-K-containing dosage
forms. In addition, the proposed methods used a spectrophotometer, which is available in all quality control laboratories,
and they involve very simple procedures.
Acknowledgment
The author is grateful to Dr. Hamdy M. Abdel-Rahman
(Department of Pharmaceutical Medicinal Chemistry, Faculty of Pharmacy, Assiut University) for performing the
molecular modeling.
References
[1] Martindale, The Complete Drug Reference, 33rd ed., Pharmaceutical
Press, London, 2002, pp. 921–922.
[2] I.A. Reid, in: B.G. Katzung (Ed.), Basic & Clinical Pharmacology,
8th ed., Librairie du Liban, Beirut, Lebanon, 2001, pp. 292–298.
[3] D.L. Hertzog, J.F. McCafferty, X. Fang, R.J. Tyrrell, R.A. Reed, J.
Pharm. Biomed. Anal. 30 (2002) 747–760.
[4] G. Carlucci, G. Palumbo, P. Mazzeo, M.G. Quaglina, J. Pharm.
Biomed. Anal. 23 (2000) 185–189.
[5] T. Iwasa, T. Takano, K. Hara, T. Kamei, J. Chromatogr. B: Biomed.
Sci. Appl. 734 (1999) 325–330.
[6] G.V. Kanumula, B. Raman, Indian Drugs 37 (2000) 38–41.
[7] A.P. Argekar, J.G. Sawant, Anal. Lett. 33 (2000) 869–880.
[8] K.E. McCarthy, G. Wang, E.W. Tsai, R.E. Gilbert, D.P. Ip, M.A.
Brooks, J. Pharm. Biomed. Anal. 17 (1999) 671–677.
[9] H.H. Maurer, T. Kraemer, J.W. Arlt, Ther. Drug Monit. 20 (1998)
706–713.
[10] L. Gonzalez, U. Akesolo, R.M. Jimenez, R.M. Alonso, Electrophoresis 23 (2002) 223–229.
[11] R.C. Williams, R.J. Boucher, J. Pharm. Biomed. Anal. 22 (2000)
115–122.
[12] A. Soldner, H. Spahn-Langguth, D. Palm, E. Mutschler, J. Pharm.
Biomed. Anal. 17 (1998) 111–124.
[13] O.C. Lastra, I.G. Lemus, H.J. Sanchez, R.F. Perez, J. Pharm. Biomed.
Anal. 33 (2003) 175–180.
[14] N. Erk, J. Pharm. Biomed. Anal. 24 (2001) 603–611.
[15] N.R. Lande, B.M. Shektar, S.S. Kadam, S.R. Dhaneshwar, Indian
Drugs 37 (2000) 577–581.
[16] R. Foster, Organic Charge Transfer Complexes, Academic Press,
London, 1969, pp. 51, 387.
[17] G.A. Saleh, H.F. Askal, I.A. Darwish, A.A. El-Shorbagi, Anal. Sci.
19 (2003) 281–287.
[18] G.A. Saleh, H.F. Askal, M.F. Radwan, M.A. Omar, Talanta 54 (2001)
1205–1215.
[19] G.A. Saleh, Talanta 46 (1998) 111–121.
[20] P. Job, Ann. Chem. 16 (1936) 97.
[21] A. Taha, G. Rücker, Arch. Pharm. 310 (1977) 485.
[22] L.R. Melby, in: S. Patai (Ed.), The Chemistry of the Cyano Group,
Interscience Publisher/John Wiley & Sons, New York, 1970, pp.
656–659.
[23] A. Yamagishi, Bull. Soc. Jpn. 48 (1975) 2440.
[24] Vogel’s Textbook of Practical Organic Chemistry, 5th ed., Longman
Group UK Ltd., England, 1989, pp. 1442–1444.
[25] H.P. Deppeler, in: K. Florey (Ed.), Analytical Profiles of Drug
Substances, vol. 10, Academic Press, New York, 1981, pp. 405–
441.
[26] S. Budavari (Ed.), The Merck Index, 13th ed., Merch & Co., Inc.,
New Jersey, 2001, p. 1000.
[27] The United States Pharmacopeia 24, The National Formulary 19,
US Pharmacopeial Convention Inc, Rockville, MD, 2000, pp.
2151–2152.