- Email: [email protected]
Quantitative determination of salbutamol in tablets by multiple-injection capillary zone electrophoresis
Journal of Chromatography A, 1207 (2008) 181–185
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Quantitative determination of salbutamol in tablets by multiple-injection capillary zone electrophoresis Henrik Lodén a , Curt Pettersson a , Torbjörn Arvidsson a,b , Ahmad Amini a,b,∗ a b
Uppsala University, Division of Analytical Pharmaceutical Chemistry, Biomedical Centre, P.O. Box 574, 751 23 Uppsala, Sweden Medical Products Agency (MPA), P.O. Box 26, Dag Hammarskjölds väg 42, 751 03 Uppsala, Sweden
a r t i c l e
i n f o
Article history: Received 25 June 2008 Received in revised form 11 August 2008 Accepted 14 August 2008 Available online 28 August 2008 Keywords: Content uniformity Oxprenolol Pharmaceutical analysis Salbutamol Validation Ventoline Depot tablets
a b s t r a c t A multiple-injection capillary zone electrophoresis (MICZE) method has been developed for the assay of salbutamol in Ventoline Depot tablets (GlaxoSmithKline). In the developed method, seven sample sets, each consisting of three samples, were sequentially injected into the capillary and analyzed within a single run. This enabled a total of twenty-one sequential injections, i.e., six standards and fifteen samples, containing salbutamol and the injection marker oxprenolol. The injected sample plugs were separated by plugs of background electrolyte, through application of a short-term voltage (30 kV) over the capillary for different time periods, i.e., tPE1 and tPE2 . The samples in each set were isolated from each other by partial electrophoresis for 2.35 min (tPE1 ), while the sample sets were separated for 10.50 min (tPE2 ). After the final injection, all the applied samples were subjected to electrophoresis for a time period corresponding to that in conventional single-injection CZE. The method was validated regarding linearity, accuracy, precision and robustness before it was applied to the determination of salbutamol in 15 tablets of Ventoline Depot with a labeled content of 8 mg salbutamol. The average salbutamol content was determined to 7.8 mg (±0.3 mg) from simultaneous analyses of the 15 different tablets. © 2008 Elsevier B.V. All rights reserved.
1. Introduction The different available modes of capillary electrophoresis (CE), has made it a valuable technique for analysis of various compounds, ranging from small ions to macromolecules with diverse structures and physicochemical properties [1]. Several review articles are devoted to the application of CE for pharmaceutical analysis [2–8]. These applications include assay of drug components and determination of chiral purity. The high separation efficiency of CE enables the technique to compete with well-established methodologies for determination of drugs in fairly simple matrices [4,9]. The main disadvantages of CE are the low concentration sensitivity often encountered when employing UV-detection [2,10] and poor reproducibility [10]. However, transparency of the fused silica for deep UV-radiation permits utilization of UV-detection at lower wavelengths (190–200 nm), where the UV-absorptivity increases [3,11,12]. Reproducibility of the migration times and the peak resolution largely depends on the
∗ Corresponding author. Tel.: +46 18 174619; fax: +46 18 548566. E-mail address: [email protected] (A. Amini). 0021-9673/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2008.08.082
selection of proper conditions for preconditioning of the capillary [10,13,14]. Multiple-injection capillary zone electrophoresis (MICZE) [15] is an alternative approach to perform quantitative determinations [16]. In MICZE, more than one standard or sample is analyzed within a single run which enables simultaneous analysis of the standards, being used for the construction of the calibration curve as well as the samples. This may improve the precision and accuracy of the determinations, since the standards and sample are analyzed under the same experimental conditions. MICZE has previously been employed for determination of the therapeutic peptide buserelin, using the analogue goserelin as an injection marker (IM) [16]. In the present study, salbutamol was selected as a model substance in order to further evaluate the applicability of MICZE for quantitative analysis. Salbutamol, which is a 2 -receptor agonist used for alleviation of bronchoconstriction, has previously been analyzed by capillary electrophoresis [9,11,12,17–23]. This paper presents a novel MICZE mode for the quantitative analysis of salbutamol in Ventoline Depot (GlaxoSmithKline, UK), using oxprenolol as an IM. The IM is added to the samples and standards to compensate for injection volumes fluctuations. The chemical structures of the solutes are shown in Fig. 1.
182
H. Lodén et al. / J. Chromatogr. A 1207 (2008) 181–185
2.4. Method validation The MICZE method was validated with regard to linearity, accuracy, precision and robustness.
Fig. 1. The chemical structures of oxprenolol (A) and salbutamol (B).
2.4.1. Linearity Two of the salbutamol stock solutions were used to prepare standards with the concentrations; 0.199, 0.222, 0.240, 0.261, 0.282 and 0.300 mM. Oxprenolol (IM) was added to all standards yielding a final concentration of 0.409 mM in 5.00 ml. Peak area ratios, i.e., Areasalbutamol /Areaoxprenolol , were plotted versus the corresponding standard concentrations for the construction of the standard curves. The areas of the peaks were not normalized with respect to their migration times.
2. Experimental 2.1. Chemicals The Ventoline Depot tablets, with a labeled content of salbutamol sulfate equivalent to 8 mg salbutamol, were purchased from a local pharmacy (Uppsala, Sweden). Salbutamol reference standards were supplied by the European Directorate for the Quality of Medicines (EDQM, Strasburg, France). Oxprenolol hydrochloride of analytical grade was purchased from LGC (Luckenwalde, Germany). Phosphoric acid and sodium hydroxide of analytical grade were purchased from Merck (Darmstadt, Germany). Triethanolamine (98%) and acetonitrile (ACN) of HPLC reagent grade were purchased from Sigma–Aldrich Chemie (Steinheim, Germany). All solutions were prepared from double distilled water, produced by a Maxima water purification system from USF Elga (High Wycombe, UK). The background electrolytes were filtered through a 0.22 m PVDF (polyvinylidene difluoride) syringe filter (Millipore, Cork, Ireland) before use. 2.2. Capillary electrophoresis Capillary electrophoresis experiments were performed on a ProteomeLab PA 800 system (Beckman Coulter, CA, USA), equipped with a photodiode array (DAD) system that monitored the wavelengths between 190 and 600 nm. Detection was performed at 200 nm. The background electrolyte (BGE) was prepared by adjusting the pH of 100 mM phosphoric acid to 2.6 or 3.1 with triethanolamine, followed by addition of 10% (v/v) ACN. A fused silica capillary of 158 cm (148 cm effective length) × 50 m I.D. (O.D. 375 m) from Polymicro Technologies, Phoenix, AZ, USA, was used. The capillary was cut to the desired length using a SGT Shortix capillary column cutter (Middelburg, The Netherlands). Before the first injection, the capillary was consecutively preconditioned with water and 0.1 M NaOH for 5 min followed by BGE for 10 min at 408 kPa. Injections of standards and samples were performed at 3.4 kPa for 15 s. Between the injections in MICZE mode, the applied plugs were subjected to partial electrophoresis for 2.35 min (tPE1 ) and 10.50 min (tPE2 ), respectively, at +30 kV. Following the last sample injection, a constant voltage of +30 kV was applied. The observed electric current was 15 A. Temperature of the capillary coolant was set to 20.0 ◦ C. Determinations of the peak standard deviation () were performed by dividing the width at the 50% of the peak height by 2.35, i.e., = w50% /2.35 [24]. 2.3. Preparation of stock solutions Stock solutions of salbutamol and the IM oxprenolol were prepared in water. The concentration of the five salbutamol solutions was adjusted to 1.00 mM, while the concentration of the stock solution of oxprenolol was 4.00 mM.
2.4.2. Accuracy and precision The three remaining stock solutions of salbutamol were used to prepare validation samples (5.00 ml) at three concentration levels covering the range of the standard curve, i.e., 0.203, 0.252 and 0.295 mM, respectively. Besides salbutamol, each standard sample contained 0.409 mM of the IM (oxprenolol). 2.4.3. Robustness In order to explore the robustness of the method, separation of salbutamol and oxprenolol was performed in single-injection mode at slightly different conditions than those in the original method. The robustness study was performed in single-injection mode instead of MICZE in order to facilitate interpretation of the electropherograms. The effect of increasing the concentration of triethanolamine in the BGE, and thus increasing the pH from 2.6 to 3.1 on the separation, was evaluated. Additionally, the effect of changing the ACN content of the BGE from 10 to 8 and 12% (v/v), respectively, was also investigated. Further, the results obtained in the original capillary were compared to those using a capillary from a different batch. 2.5. Preparation of the tablet samples Each tablet was weighed individually and ground in a mortar, before being transferred quantitatively to a 100.00 ml volumetric flask. Water was then added and the solutions were placed in a Bransonic 52 ultrasonic bath (Branson Ultrasonics, Danbury, CT, USA) for 15 min in darkness. The solutions were left to adapt to room-temperature before water was added to make up the final volumes. Aliquots of the solutions were filtered, and 3.75 ml was taken to which the injection marker oxprenolol was added to a final concentration of 0.409 mM after making up the final volumes of 5.00 ml with water. 3. Results and discussion 3.1. The separation system In order to apply MICZE to the determination of salbutamol in Ventoline Depot tablets, it was first necessary to design the separation system in the single-injection mode to obtain a suitable time window (tmig ) between the peaks of the IM and salbutamol, respectively. The sample capacity (ns ), which has been defined as the highest number of injections which can be performed in MICZE [25], is affected by three factors, i.e., the migration time of the slow migrating solute, the tmig between the analyte and the IM, as well as the peak efficiency. The migration time of an analyte is dependent not only on its own electrophoretic mobility but also on that of the electroosmotic flow (EOF). The magnitude and direction of EOF can be readily
H. Lodén et al. / J. Chromatogr. A 1207 (2008) 181–185
manipulated by changing the pH or composition of the BGE. The addition of triethanolamine into BGE has been shown to generate an anodic EOF [16,26,27], which results in prolongation of the migration times, which increases the resolution of the solutes and thereby the ns , e.g., Eq. (2). In addition, adsorption of triethanolamine onto the capillary surface, as well as the presence of ACN in the BGE, may suppress interactions between the positively charged solutes and the capillary surface, contributing to the improved peak efficiency and resolution of the solutes [16]. A prerequisite for the successful transfer of the single-injection is the absence of interfering impurities, since additional peaks from impurities reduce the sample capacity (ns ) [25]. However, the purity of pharmaceutical products is generally high and related impurities only constitute a small fraction of the total content. The detection of these impurities is dependent on their concentration as well as their UV-absorptivity. In order to avoid detection of such components, the concentration of the solutes in the sample has to be adjusted. As shown in Fig. 2A, some small peaks besides salbutamol appeared in the electropherogram upon analysis of a tablet sample containing salbutamol at 7 mM. However, these peaks were not detected when the salbutamol concentration was reduced to approximately 0.25 mM, i.e., within the concentration range of the forthcoming determinations, Fig. 2B. Following an initial separation of oxprenolol and salbutamol by conventional CZE in a 148 cm (Leff ) long fused silica capillary (Fig. 3A), the migration times of the peaks were 103.60 and 111.04 min, respectively. This resulted in a migration time difference (tmig ) of 7.44 min between oxprenolol and salbutamol. The peak standard deviation for the broadest peak was calculated to be 0.196 min from the width at 50% of the peak height ( = w0% /2.35). Taking into account the numerical values of tmig and tmig2 seven
183
Fig. 3. Separations of salbutamol and oxprenolol in the single-injection mode (A) and in MICZE mode (B). In MICZE mode, the tPE1 and tPE2 was 2.35 and 10.50 min, respectively. Other conditions as those described in Fig. 2. Peaks: (1) oxprenolol (0.409 mM) and (2) salbutamol (0.252 mM).
sample sets, each consisting of three samples, were possible to sequentially inject into the capillary (Fig. 3B). Within each set, the time window (tmig ) between the oxprenolol and salbutamol peaks from the first injection was used for the detection of the slow migrating solutes, i.e., salbutamol, from the other two injections. Consequently, the corresponding fast migrating peaks from oxprenolol were detected prior to this time window (Fig. 3B). The number of injections was determined by the magnitude of tmig and the time period for the partial electrophoresis (tPE1 ). The tPE1 was determined by the peak standard deviation, Eq. (1) [25]. tPE1 ≥ 12
(1)
Actually, 6 is sufficient to obtain base-line separation of the peaks from the neighboring plugs when the peaks are Gaussian. However, in order to enhance the resolution between the peaks at the boundaries of the injected plugs the distance between the peaks was further increased by 6. Consequently, the minimum value of tPE1 was calculated to be 2.35 min, Eq. (1). The number of injections in each set (n1 ) was determined to three samples using the selected tPE1 , Eq. (2) [25]. n1 =
tmig
(2)
tPE1
In order to obtain complete separation between the sequentially injected sample sets, the time of the partial electrophoresis for the last sample in each set (tPE2 ) has to be longer than tPE1 [25]. Using Eq. (3) the minimum value of tPE2 was calculated to 9.79 min. tPE2 ≥ tmig + 12
(3)
Following adjustment of the tPE2 to 10.50 min, the maximum number of sets (n2 ) was calculated to seven by Eq. (4) [25]. n2 =
tmig2 [tPE2 + (n1 − 1)tPE1 ]
(4)
The total number of injections (ninj ) is the product of n1 and n2 , i.e., 21 in this particular case. Fig. 2. CZE analysis of Ventoline Depot dissolved in water. The concentration of salbutamol was 7 mM (A) and 0.25 mM (B). The BGE consisted of 10% (v/v) ACN in 100 mM phosphoric acid adjusted to pH 2.6 with triethanolamine. Samples were injected at 0.5 psi (3.5 kPa) for 15 s. Separation was performed by applying 30 kV over a 50 m I.D. fused silica capillary with an effective length of 148 cm. UV-detection was performed at 200 nm. (1) Main peak of salbutamol.
ninj = n1 · n2
(5)
The predicted number of injections (ninj ) by Eq. (5) was confirmed by MICZE analysis of 21 standards containing salbutamol and oxprenolol. As shown in Fig. 3B, separation of seven sets, each containing three samples, was achieved.
184
H. Lodén et al. / J. Chromatogr. A 1207 (2008) 181–185 Table 2 Accuracy and precision obtained from MICZE determinations (n = 18)
Table 1 Linearity of the standard curves using corrected peak areas (n = 6 in 2 days) Day
Regression equation
R2
[Standard] (mM)
[Determined] (mM)
Accuracy (%)
RSD (%)
1
Y = 0.00239x − 0.0085 Y = 0.00226x + 0.0165 Y = 0.00218x + 0.0172
0.998 0.998 0.996
0.203 0.252 0.295
0.202 0.250 0.294
99.8 99.4 99.6
2.7 1.9 2.3
2
Y = 0.00233x − 0.0106 Y = 0.00223x + 0.0213 Y = 0.00239x − 0.0142
0.996 0.996 0.997
Samples used for the validation contained 0.409 mM oxprenolol as injection marker. Other conditions as those described in Fig. 3B and Table 1.
The standards contained 0.199, 0.222, 0.240, 0.261, 0.282 and 0.300 mM salbutamol, respectively. All standards contained 0.409 mM of the IM oxprenolol. Other conditions as those given in Fig. 3B. Y: The peak area ratio between the analyte and IM, and x: the analyte concentration.
3.2. Method validation Before employing the MICZE system described above for quantitative analysis of the salbutamol tablets, it was validated with respect to the linearity, precision, accuracy and robustness. Validation of the inter-day linearity, accuracy and precision were performed under the conditions described in Sections 2.4 and 3.1, respectively. The limit of quantification (LOQ, S/N = 10) for salbutamol was determined to be 0.014 mM, i.e., much smaller than the concentrations used for the construction of the standard curves (0.200–0.300 mM). 3.2.1. Linearity Standard solutions of salbutamol containing the IM oxprenolol were analyzed in the MICZE mode. A comparison between the standard curves constructed on different days is presented in Table 1. The consistency of the correlation factors (R2 ) and the agreement of the slopes were good. However, it is noteworthy that an agreement between the slopes is not necessary in MICZE mode, because the concentration of the samples is determined by the standard curve constructed from the simultaneously analyzed standard solutions. Since all samples are analyzed at the same experimental conditions in MICZE, normalization of the peak areas with their corresponding migration times has been found to be unnecessary [16]. In conventional CZE, however, the use of normalized peak areas is a necessity [20,28] since small changes in the EOF may contribute to significant changes of the peak areas [13]. 3.2.2. Accuracy and precision The obtained accuracy and precision from determination of the validation samples containing salbutamol at different concentrations with oxprenolol as IM are given in Table 2. The determined
concentrations deviated by less than 1.0% from the actual concentrations, whereas the variation (RSD) between the determinations was less than 3.0% (n = 18). It should be mentioned that volume measurements during the preparation of the solutions were estimated to contribute to the uncertainty of the determinations by ±1.2%. This implies that the separation system contributed to the variations less than described above. 3.2.3. Robustness In order to study the robustness, separations of salbutamol and oxprenolol were performed in the single-injection mode. The influence of the induced changes in the BGE and the use of a new capillary on the migration times, peak standard deviation () and the migration time differences (tmig ), and thereby on the sample capacity, was investigated (Table 3). The system was found to be robust towards the induced changes, with only slight variation of the migration times as well as the . Slightly increased migration times were observed after increasing the pH of the BGE from 2.6 to 3.1. This may be caused by generation of a stronger anodic EOF as well as by an increase in the viscosity of the BGE. The results showed that readjustments of the tPE1 and the tPE2 are required in order to maintain the number of sample injections (ninj ) at 21, following a change of capillary or the BGE. It is to be noted that the minimum values of tPE2 given in Table 3 may be elongated in order to increase the resolution between the peak sets, while maintaining the sample capacity. However, there is a limit of the tPE2 , after which further prolongation will result in a reduced sample capacity, cf. Eq. (5). The results from the robustness study indeed stress the necessity of performing an initial analysis in the single-injection mode upon any changes in the separation conditions, e.g., replacement of the capillary. This single-injection run can be considered as a system suitability test. It should be mentioned that the use of the long capillary, compared to a short capillary, required more laborious and time consuming method development and system suitability test.
Table 3 Robustness evaluation of the separation system in the single-injection mode (n = 3) pH
ACN (%, v/v)
tmig1 (min)
tmig2 (min)
(min)
tPE1 a (min)
n1 b
tPE2 c (min)
n2 d
ninj e
2.6 2.6 2.6g 2.6 3.1
8 10 10 12 10
101.71 (1%)f 101.05 (1%) 97.00 (1%) 102.71 (2%) 106.79 (1%)
108.38 (1%) 108.46 (1%) 103.69 (1%) 109.87 (2%) 114.52 (1%)
0.183 (2%) 0.195 (3%) 0.192 (2%) 0.200 (4%) 0.208 (1%)
2.20 2.35 2.30 2.40 2.50
3 3 3 3 3
9.10 9.65 9.20 9.60 10.25
7 7 7 7 7
21 21 21 21 21
Separation of salbutamol (0.252 mM) and oxprenolol (0.409 mM) was carried out at different conditions. Other conditions as those described in Fig. 2 and in Section 2.4.3. a The time of partial electrophoresis within each set was calculated using Eq. (1). b The number of injections within each set was calculated using Eq. (2). c The minimum value of tPE2 , calculated by Eq. (3). d The number of applicable sets was calculated using Eq. (4). e The number of injections at the selected tPE1 and tPE2 was calculated using Eq. (5). f Relative standard deviation (RSD). g Separations performed in a capillary from a different batch.
H. Lodén et al. / J. Chromatogr. A 1207 (2008) 181–185
185
ing a mixed sample of the fifteen tablets was determined to 1.0% (RSD). 4. Conclusions
Fig. 4. Assay of 15 salbutamol tablets in MICZE mode. The tablet samples were prepared according to the procedure described in Section 2.5. The concentrations of the six standard solutions were the same as those in Table 1. Other conditions as described in Figs. 2 and 3B. Peaks: (1) oxprenolol and (2) salbutamol.
Table 4 Determination of salbutamol in 15 Ventoline Depot tablets with a labeled content of 8 mg Tablet no.
Repeat
Average (RSD%)
1
2
3
4
5
6
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
7.5 7.6 8.0 8.6 7.8 7.7 8.4 8.0 8.5 7.8 7.7 8.1 8.0 8.0 7.8
7.5 7.3 7.7 8.3 7.8 7.7 7.9 7.4 8.1 7.6 7.5 7.9 7.7 7.7 7.8
7.2 7.3 7.9 8.3 7.3 7.3 7.8 7.8 8.4 7.5 7.3 8.1 7.7 7.9 7.6
7.3 7.6 8.1 8.4 7.6 7.6 7.9 7.7 8.2 7.6 7.6 8.0 7.9 8.0 7.8
7.6 7.3 7.7 8.2 8.0 7.9 8.4 7.7 8.3 7.6 7.4 8.0 7.8 8.3 7.8
7.4 7.2 8.0 8.3 7.8 7.9 8.0 7.6 8.2 7.6 7.5 8.1 7.9 7.9 7.7
7.4 (2.4) 7.3 (2.5) 7.9 (2.1) 8.3 (1.7) 7.7 (3.1) 7.7 (2.9) 8.0 (3.3) 7.6 (2.6) 8.2 (1.8) 7.6 (1.3) 7.5 (1.9) 8.0 (1.0) 7.8 (1.6) 8.0 (2.5) 7.7 (1.1)
Average RSD (%)
8.0 4.1
7.7 3.5
7.7 5.1
7.8 3.6
7.9 4.3
7.8 4.0
7.8 (4.2)
This paper describes a new MICZE method for determination of salbutamol. Following validation, the obtained linearity, accuracy and precision demonstrated that the method could be confidently employed for the determination of salbutamol. The MICZE method was then applied for analysis of the salbutamol content of 15 tablets (Ventoline Depot) within a single-run. The determined content of the tablets deviated by only 2.4% from the declared amount. The inter-run variation of the determined content of each tablet was ≤3% (RSD). Analysis of the mixed tablet sample resulted in a slightly increased precision and a content of salbutamol in the tablets which concurred with the result from the individual determinations. Following preparation of a fresh BGE or a capillary replacement, an initial separation of the analyte and the injection marker should be performed in the single-injection mode in order to readjust the separation parameters, i.e., tPE1 and tPE2 . References
Oxprenolol was added into the tablet samples at a concentration of 0.409 mM. Other conditions as those given in Table 1, Fig. 3 and in Section 2.5, respectively.
3.3. Determination of salbutamol in tablet samples The salbutamol content of 15 Ventoline Depot tablets was determined, employing the validated MIZCE method, see Sections 2.2 and 3.2. The labeled content of salbutamol sulfate was declared to be equivalent to 8 mg salbutamol. An electropherogram from these determinations is shown in Fig. 4 and the results from the determinations are presented in Table 4. The average amount of salbutamol in these 15 tablets was 7.8 mg, with an intermediary precision of 4.2% (RSD). The inter-run precision based on six separate MICZE determinations of the individual tablets was between 1.0 and 3.3% (RSD), respectively. The intra-run precision determined by analyz-
[1] M.G. Khaledi (Ed.), High-Performance Capillary Electrophoresis: Theory, Techniques and Applications, Willey, New York, 1998. [2] L. Suntornsuk, J. Chromatogr. Sci. 45 (2007) 559. [3] K. Altria, A. Marsh, C. Sänger-van de Griend, Electrophoresis 27 (2006) 2263. [4] T.K. Natishan, J. Liq. Chromatogr. Rel. Technol. 28 (2005) 1115. [5] T.G. Morzunova, Pharm. Chem. J. 40 (2006) 158. [6] P.T.T. Ha, J. Hoogmartens, A.V. Schepdael, J. Pharm. Biomed. Anal. 41 (2006) 1. [7] S. Fanali, Z. Aturki, C. Desiderio, Enantiomer 4 (1999) 229. [8] G.K.E. Scriba, Electrophoresis 24 (2003) 2409. [9] M.T. Ackermans, J.L. Beckers, F.M. Everaerts, I.G.J.A. Seelen, J. Chromatogr. 590 (1992) 341. [10] U. Holzgrabe, D. Brinz, S. Kopec, C. Weber, Y. Bitar, Electrophoresis 27 (2006) 2283. [11] K.D. Altria, J. Chromatogr. 634 (1993) 323. [12] K.D. Altria, P. Franke, I. Gill, T. Hadgett, M.A. Kelly, D.R. Rudd, J. Pharm. Biomed. Anal. 13 (1995) 951. [13] J.L. Lambert, D.L. Middleton, Anal. Chem. 62 (1990) 1585. [14] J.E. Gómez, J.E. Sandoval, Electrophoresis 29 (2008) 381. [15] J.D.B. Zavaleta, A. Chinchilla, A. Ramirez, K. Pao, S. Martinez, J.E. Nilapwar, Talanta 71 (2007) 192. [16] H. Lodén, A. Amini, Electrophoresis 28 (2007) 1548. [17] Y. Chen, W. Wang, J.P. Duan, H.Q. Chen, G.N. Chen, Electroanalysis 17 (2005) 706. [18] E. Ekiert, C. Garcia-Ruiz, M.A. Garcia, M.L. Marina, Electrophoresis 24 (2003) 2680. [19] F.S. Felix, M.S.M. Quintino, A.Z. Carvalho, L.H.G. Coelho, C.L. do Lago, L. Angnes, J. Pharm. Biomed. Anal. 40 (2006) 1288. [20] P. Mikuˇs, I. Valáˇsková, E. Havránek, Arch. Pharm. Chem. Life Sci. 338 (2005) 498. [21] J. Vela, E.G. Yanes, A.M. Stalcup, Fresenius J. Anal. Chem. 369 (2001) 212. [22] A. Esquisabel, R.M. Hernandez, A.R. Gascon, M. Igartua, B. Calvo, J.L. Pedraz, J. Pharm. Biomed. Anal. 16 (1997) 357. [23] L. Mälkki-Laine, E. Hartikainen, J. Chromatogr. A 724 (1996) 297. [24] M.J. Mills, J. Maltas, W.J. Lough, J. Chromatogr. A 759 (1997) 1. [25] A. Amini, H. Lodén, C. Pettersson, T. Arvidsson, Electrophoresis 29 (2008) 1. [26] A. Amini, V. Barclay, T. Rundlöf, S. Jönsson, A. Karlsson, T. Arvidsson, Chromatographia 63 (2006) 143. [27] D. Corradini, A. Rhomberg, C. Corradini, J. Chromatogr. A 661 (1994) 305. [28] K.D. Altria, H. Fabre, Chromatographia 40 (1995) 313.
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Quantitative determination of salbutamol in tablets by multiple-injection capillary zone electrophoresis Henrik Lodén a , Curt Pettersson a , Torbjörn Arvidsson a,b , Ahmad Amini a,b,∗ a b
Uppsala University, Division of Analytical Pharmaceutical Chemistry, Biomedical Centre, P.O. Box 574, 751 23 Uppsala, Sweden Medical Products Agency (MPA), P.O. Box 26, Dag Hammarskjölds väg 42, 751 03 Uppsala, Sweden
a r t i c l e
i n f o
Article history: Received 25 June 2008 Received in revised form 11 August 2008 Accepted 14 August 2008 Available online 28 August 2008 Keywords: Content uniformity Oxprenolol Pharmaceutical analysis Salbutamol Validation Ventoline Depot tablets
a b s t r a c t A multiple-injection capillary zone electrophoresis (MICZE) method has been developed for the assay of salbutamol in Ventoline Depot tablets (GlaxoSmithKline). In the developed method, seven sample sets, each consisting of three samples, were sequentially injected into the capillary and analyzed within a single run. This enabled a total of twenty-one sequential injections, i.e., six standards and fifteen samples, containing salbutamol and the injection marker oxprenolol. The injected sample plugs were separated by plugs of background electrolyte, through application of a short-term voltage (30 kV) over the capillary for different time periods, i.e., tPE1 and tPE2 . The samples in each set were isolated from each other by partial electrophoresis for 2.35 min (tPE1 ), while the sample sets were separated for 10.50 min (tPE2 ). After the final injection, all the applied samples were subjected to electrophoresis for a time period corresponding to that in conventional single-injection CZE. The method was validated regarding linearity, accuracy, precision and robustness before it was applied to the determination of salbutamol in 15 tablets of Ventoline Depot with a labeled content of 8 mg salbutamol. The average salbutamol content was determined to 7.8 mg (±0.3 mg) from simultaneous analyses of the 15 different tablets. © 2008 Elsevier B.V. All rights reserved.
1. Introduction The different available modes of capillary electrophoresis (CE), has made it a valuable technique for analysis of various compounds, ranging from small ions to macromolecules with diverse structures and physicochemical properties [1]. Several review articles are devoted to the application of CE for pharmaceutical analysis [2–8]. These applications include assay of drug components and determination of chiral purity. The high separation efficiency of CE enables the technique to compete with well-established methodologies for determination of drugs in fairly simple matrices [4,9]. The main disadvantages of CE are the low concentration sensitivity often encountered when employing UV-detection [2,10] and poor reproducibility [10]. However, transparency of the fused silica for deep UV-radiation permits utilization of UV-detection at lower wavelengths (190–200 nm), where the UV-absorptivity increases [3,11,12]. Reproducibility of the migration times and the peak resolution largely depends on the
∗ Corresponding author. Tel.: +46 18 174619; fax: +46 18 548566. E-mail address: [email protected] (A. Amini). 0021-9673/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2008.08.082
selection of proper conditions for preconditioning of the capillary [10,13,14]. Multiple-injection capillary zone electrophoresis (MICZE) [15] is an alternative approach to perform quantitative determinations [16]. In MICZE, more than one standard or sample is analyzed within a single run which enables simultaneous analysis of the standards, being used for the construction of the calibration curve as well as the samples. This may improve the precision and accuracy of the determinations, since the standards and sample are analyzed under the same experimental conditions. MICZE has previously been employed for determination of the therapeutic peptide buserelin, using the analogue goserelin as an injection marker (IM) [16]. In the present study, salbutamol was selected as a model substance in order to further evaluate the applicability of MICZE for quantitative analysis. Salbutamol, which is a 2 -receptor agonist used for alleviation of bronchoconstriction, has previously been analyzed by capillary electrophoresis [9,11,12,17–23]. This paper presents a novel MICZE mode for the quantitative analysis of salbutamol in Ventoline Depot (GlaxoSmithKline, UK), using oxprenolol as an IM. The IM is added to the samples and standards to compensate for injection volumes fluctuations. The chemical structures of the solutes are shown in Fig. 1.
182
H. Lodén et al. / J. Chromatogr. A 1207 (2008) 181–185
2.4. Method validation The MICZE method was validated with regard to linearity, accuracy, precision and robustness.
Fig. 1. The chemical structures of oxprenolol (A) and salbutamol (B).
2.4.1. Linearity Two of the salbutamol stock solutions were used to prepare standards with the concentrations; 0.199, 0.222, 0.240, 0.261, 0.282 and 0.300 mM. Oxprenolol (IM) was added to all standards yielding a final concentration of 0.409 mM in 5.00 ml. Peak area ratios, i.e., Areasalbutamol /Areaoxprenolol , were plotted versus the corresponding standard concentrations for the construction of the standard curves. The areas of the peaks were not normalized with respect to their migration times.
2. Experimental 2.1. Chemicals The Ventoline Depot tablets, with a labeled content of salbutamol sulfate equivalent to 8 mg salbutamol, were purchased from a local pharmacy (Uppsala, Sweden). Salbutamol reference standards were supplied by the European Directorate for the Quality of Medicines (EDQM, Strasburg, France). Oxprenolol hydrochloride of analytical grade was purchased from LGC (Luckenwalde, Germany). Phosphoric acid and sodium hydroxide of analytical grade were purchased from Merck (Darmstadt, Germany). Triethanolamine (98%) and acetonitrile (ACN) of HPLC reagent grade were purchased from Sigma–Aldrich Chemie (Steinheim, Germany). All solutions were prepared from double distilled water, produced by a Maxima water purification system from USF Elga (High Wycombe, UK). The background electrolytes were filtered through a 0.22 m PVDF (polyvinylidene difluoride) syringe filter (Millipore, Cork, Ireland) before use. 2.2. Capillary electrophoresis Capillary electrophoresis experiments were performed on a ProteomeLab PA 800 system (Beckman Coulter, CA, USA), equipped with a photodiode array (DAD) system that monitored the wavelengths between 190 and 600 nm. Detection was performed at 200 nm. The background electrolyte (BGE) was prepared by adjusting the pH of 100 mM phosphoric acid to 2.6 or 3.1 with triethanolamine, followed by addition of 10% (v/v) ACN. A fused silica capillary of 158 cm (148 cm effective length) × 50 m I.D. (O.D. 375 m) from Polymicro Technologies, Phoenix, AZ, USA, was used. The capillary was cut to the desired length using a SGT Shortix capillary column cutter (Middelburg, The Netherlands). Before the first injection, the capillary was consecutively preconditioned with water and 0.1 M NaOH for 5 min followed by BGE for 10 min at 408 kPa. Injections of standards and samples were performed at 3.4 kPa for 15 s. Between the injections in MICZE mode, the applied plugs were subjected to partial electrophoresis for 2.35 min (tPE1 ) and 10.50 min (tPE2 ), respectively, at +30 kV. Following the last sample injection, a constant voltage of +30 kV was applied. The observed electric current was 15 A. Temperature of the capillary coolant was set to 20.0 ◦ C. Determinations of the peak standard deviation () were performed by dividing the width at the 50% of the peak height by 2.35, i.e., = w50% /2.35 [24]. 2.3. Preparation of stock solutions Stock solutions of salbutamol and the IM oxprenolol were prepared in water. The concentration of the five salbutamol solutions was adjusted to 1.00 mM, while the concentration of the stock solution of oxprenolol was 4.00 mM.
2.4.2. Accuracy and precision The three remaining stock solutions of salbutamol were used to prepare validation samples (5.00 ml) at three concentration levels covering the range of the standard curve, i.e., 0.203, 0.252 and 0.295 mM, respectively. Besides salbutamol, each standard sample contained 0.409 mM of the IM (oxprenolol). 2.4.3. Robustness In order to explore the robustness of the method, separation of salbutamol and oxprenolol was performed in single-injection mode at slightly different conditions than those in the original method. The robustness study was performed in single-injection mode instead of MICZE in order to facilitate interpretation of the electropherograms. The effect of increasing the concentration of triethanolamine in the BGE, and thus increasing the pH from 2.6 to 3.1 on the separation, was evaluated. Additionally, the effect of changing the ACN content of the BGE from 10 to 8 and 12% (v/v), respectively, was also investigated. Further, the results obtained in the original capillary were compared to those using a capillary from a different batch. 2.5. Preparation of the tablet samples Each tablet was weighed individually and ground in a mortar, before being transferred quantitatively to a 100.00 ml volumetric flask. Water was then added and the solutions were placed in a Bransonic 52 ultrasonic bath (Branson Ultrasonics, Danbury, CT, USA) for 15 min in darkness. The solutions were left to adapt to room-temperature before water was added to make up the final volumes. Aliquots of the solutions were filtered, and 3.75 ml was taken to which the injection marker oxprenolol was added to a final concentration of 0.409 mM after making up the final volumes of 5.00 ml with water. 3. Results and discussion 3.1. The separation system In order to apply MICZE to the determination of salbutamol in Ventoline Depot tablets, it was first necessary to design the separation system in the single-injection mode to obtain a suitable time window (tmig ) between the peaks of the IM and salbutamol, respectively. The sample capacity (ns ), which has been defined as the highest number of injections which can be performed in MICZE [25], is affected by three factors, i.e., the migration time of the slow migrating solute, the tmig between the analyte and the IM, as well as the peak efficiency. The migration time of an analyte is dependent not only on its own electrophoretic mobility but also on that of the electroosmotic flow (EOF). The magnitude and direction of EOF can be readily
H. Lodén et al. / J. Chromatogr. A 1207 (2008) 181–185
manipulated by changing the pH or composition of the BGE. The addition of triethanolamine into BGE has been shown to generate an anodic EOF [16,26,27], which results in prolongation of the migration times, which increases the resolution of the solutes and thereby the ns , e.g., Eq. (2). In addition, adsorption of triethanolamine onto the capillary surface, as well as the presence of ACN in the BGE, may suppress interactions between the positively charged solutes and the capillary surface, contributing to the improved peak efficiency and resolution of the solutes [16]. A prerequisite for the successful transfer of the single-injection is the absence of interfering impurities, since additional peaks from impurities reduce the sample capacity (ns ) [25]. However, the purity of pharmaceutical products is generally high and related impurities only constitute a small fraction of the total content. The detection of these impurities is dependent on their concentration as well as their UV-absorptivity. In order to avoid detection of such components, the concentration of the solutes in the sample has to be adjusted. As shown in Fig. 2A, some small peaks besides salbutamol appeared in the electropherogram upon analysis of a tablet sample containing salbutamol at 7 mM. However, these peaks were not detected when the salbutamol concentration was reduced to approximately 0.25 mM, i.e., within the concentration range of the forthcoming determinations, Fig. 2B. Following an initial separation of oxprenolol and salbutamol by conventional CZE in a 148 cm (Leff ) long fused silica capillary (Fig. 3A), the migration times of the peaks were 103.60 and 111.04 min, respectively. This resulted in a migration time difference (tmig ) of 7.44 min between oxprenolol and salbutamol. The peak standard deviation for the broadest peak was calculated to be 0.196 min from the width at 50% of the peak height ( = w0% /2.35). Taking into account the numerical values of tmig and tmig2 seven
183
Fig. 3. Separations of salbutamol and oxprenolol in the single-injection mode (A) and in MICZE mode (B). In MICZE mode, the tPE1 and tPE2 was 2.35 and 10.50 min, respectively. Other conditions as those described in Fig. 2. Peaks: (1) oxprenolol (0.409 mM) and (2) salbutamol (0.252 mM).
sample sets, each consisting of three samples, were possible to sequentially inject into the capillary (Fig. 3B). Within each set, the time window (tmig ) between the oxprenolol and salbutamol peaks from the first injection was used for the detection of the slow migrating solutes, i.e., salbutamol, from the other two injections. Consequently, the corresponding fast migrating peaks from oxprenolol were detected prior to this time window (Fig. 3B). The number of injections was determined by the magnitude of tmig and the time period for the partial electrophoresis (tPE1 ). The tPE1 was determined by the peak standard deviation, Eq. (1) [25]. tPE1 ≥ 12
(1)
Actually, 6 is sufficient to obtain base-line separation of the peaks from the neighboring plugs when the peaks are Gaussian. However, in order to enhance the resolution between the peaks at the boundaries of the injected plugs the distance between the peaks was further increased by 6. Consequently, the minimum value of tPE1 was calculated to be 2.35 min, Eq. (1). The number of injections in each set (n1 ) was determined to three samples using the selected tPE1 , Eq. (2) [25]. n1 =
tmig
(2)
tPE1
In order to obtain complete separation between the sequentially injected sample sets, the time of the partial electrophoresis for the last sample in each set (tPE2 ) has to be longer than tPE1 [25]. Using Eq. (3) the minimum value of tPE2 was calculated to 9.79 min. tPE2 ≥ tmig + 12
(3)
Following adjustment of the tPE2 to 10.50 min, the maximum number of sets (n2 ) was calculated to seven by Eq. (4) [25]. n2 =
tmig2 [tPE2 + (n1 − 1)tPE1 ]
(4)
The total number of injections (ninj ) is the product of n1 and n2 , i.e., 21 in this particular case. Fig. 2. CZE analysis of Ventoline Depot dissolved in water. The concentration of salbutamol was 7 mM (A) and 0.25 mM (B). The BGE consisted of 10% (v/v) ACN in 100 mM phosphoric acid adjusted to pH 2.6 with triethanolamine. Samples were injected at 0.5 psi (3.5 kPa) for 15 s. Separation was performed by applying 30 kV over a 50 m I.D. fused silica capillary with an effective length of 148 cm. UV-detection was performed at 200 nm. (1) Main peak of salbutamol.
ninj = n1 · n2
(5)
The predicted number of injections (ninj ) by Eq. (5) was confirmed by MICZE analysis of 21 standards containing salbutamol and oxprenolol. As shown in Fig. 3B, separation of seven sets, each containing three samples, was achieved.
184
H. Lodén et al. / J. Chromatogr. A 1207 (2008) 181–185 Table 2 Accuracy and precision obtained from MICZE determinations (n = 18)
Table 1 Linearity of the standard curves using corrected peak areas (n = 6 in 2 days) Day
Regression equation
R2
[Standard] (mM)
[Determined] (mM)
Accuracy (%)
RSD (%)
1
Y = 0.00239x − 0.0085 Y = 0.00226x + 0.0165 Y = 0.00218x + 0.0172
0.998 0.998 0.996
0.203 0.252 0.295
0.202 0.250 0.294
99.8 99.4 99.6
2.7 1.9 2.3
2
Y = 0.00233x − 0.0106 Y = 0.00223x + 0.0213 Y = 0.00239x − 0.0142
0.996 0.996 0.997
Samples used for the validation contained 0.409 mM oxprenolol as injection marker. Other conditions as those described in Fig. 3B and Table 1.
The standards contained 0.199, 0.222, 0.240, 0.261, 0.282 and 0.300 mM salbutamol, respectively. All standards contained 0.409 mM of the IM oxprenolol. Other conditions as those given in Fig. 3B. Y: The peak area ratio between the analyte and IM, and x: the analyte concentration.
3.2. Method validation Before employing the MICZE system described above for quantitative analysis of the salbutamol tablets, it was validated with respect to the linearity, precision, accuracy and robustness. Validation of the inter-day linearity, accuracy and precision were performed under the conditions described in Sections 2.4 and 3.1, respectively. The limit of quantification (LOQ, S/N = 10) for salbutamol was determined to be 0.014 mM, i.e., much smaller than the concentrations used for the construction of the standard curves (0.200–0.300 mM). 3.2.1. Linearity Standard solutions of salbutamol containing the IM oxprenolol were analyzed in the MICZE mode. A comparison between the standard curves constructed on different days is presented in Table 1. The consistency of the correlation factors (R2 ) and the agreement of the slopes were good. However, it is noteworthy that an agreement between the slopes is not necessary in MICZE mode, because the concentration of the samples is determined by the standard curve constructed from the simultaneously analyzed standard solutions. Since all samples are analyzed at the same experimental conditions in MICZE, normalization of the peak areas with their corresponding migration times has been found to be unnecessary [16]. In conventional CZE, however, the use of normalized peak areas is a necessity [20,28] since small changes in the EOF may contribute to significant changes of the peak areas [13]. 3.2.2. Accuracy and precision The obtained accuracy and precision from determination of the validation samples containing salbutamol at different concentrations with oxprenolol as IM are given in Table 2. The determined
concentrations deviated by less than 1.0% from the actual concentrations, whereas the variation (RSD) between the determinations was less than 3.0% (n = 18). It should be mentioned that volume measurements during the preparation of the solutions were estimated to contribute to the uncertainty of the determinations by ±1.2%. This implies that the separation system contributed to the variations less than described above. 3.2.3. Robustness In order to study the robustness, separations of salbutamol and oxprenolol were performed in the single-injection mode. The influence of the induced changes in the BGE and the use of a new capillary on the migration times, peak standard deviation () and the migration time differences (tmig ), and thereby on the sample capacity, was investigated (Table 3). The system was found to be robust towards the induced changes, with only slight variation of the migration times as well as the . Slightly increased migration times were observed after increasing the pH of the BGE from 2.6 to 3.1. This may be caused by generation of a stronger anodic EOF as well as by an increase in the viscosity of the BGE. The results showed that readjustments of the tPE1 and the tPE2 are required in order to maintain the number of sample injections (ninj ) at 21, following a change of capillary or the BGE. It is to be noted that the minimum values of tPE2 given in Table 3 may be elongated in order to increase the resolution between the peak sets, while maintaining the sample capacity. However, there is a limit of the tPE2 , after which further prolongation will result in a reduced sample capacity, cf. Eq. (5). The results from the robustness study indeed stress the necessity of performing an initial analysis in the single-injection mode upon any changes in the separation conditions, e.g., replacement of the capillary. This single-injection run can be considered as a system suitability test. It should be mentioned that the use of the long capillary, compared to a short capillary, required more laborious and time consuming method development and system suitability test.
Table 3 Robustness evaluation of the separation system in the single-injection mode (n = 3) pH
ACN (%, v/v)
tmig1 (min)
tmig2 (min)
(min)
tPE1 a (min)
n1 b
tPE2 c (min)
n2 d
ninj e
2.6 2.6 2.6g 2.6 3.1
8 10 10 12 10
101.71 (1%)f 101.05 (1%) 97.00 (1%) 102.71 (2%) 106.79 (1%)
108.38 (1%) 108.46 (1%) 103.69 (1%) 109.87 (2%) 114.52 (1%)
0.183 (2%) 0.195 (3%) 0.192 (2%) 0.200 (4%) 0.208 (1%)
2.20 2.35 2.30 2.40 2.50
3 3 3 3 3
9.10 9.65 9.20 9.60 10.25
7 7 7 7 7
21 21 21 21 21
Separation of salbutamol (0.252 mM) and oxprenolol (0.409 mM) was carried out at different conditions. Other conditions as those described in Fig. 2 and in Section 2.4.3. a The time of partial electrophoresis within each set was calculated using Eq. (1). b The number of injections within each set was calculated using Eq. (2). c The minimum value of tPE2 , calculated by Eq. (3). d The number of applicable sets was calculated using Eq. (4). e The number of injections at the selected tPE1 and tPE2 was calculated using Eq. (5). f Relative standard deviation (RSD). g Separations performed in a capillary from a different batch.
H. Lodén et al. / J. Chromatogr. A 1207 (2008) 181–185
185
ing a mixed sample of the fifteen tablets was determined to 1.0% (RSD). 4. Conclusions
Fig. 4. Assay of 15 salbutamol tablets in MICZE mode. The tablet samples were prepared according to the procedure described in Section 2.5. The concentrations of the six standard solutions were the same as those in Table 1. Other conditions as described in Figs. 2 and 3B. Peaks: (1) oxprenolol and (2) salbutamol.
Table 4 Determination of salbutamol in 15 Ventoline Depot tablets with a labeled content of 8 mg Tablet no.
Repeat
Average (RSD%)
1
2
3
4
5
6
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
7.5 7.6 8.0 8.6 7.8 7.7 8.4 8.0 8.5 7.8 7.7 8.1 8.0 8.0 7.8
7.5 7.3 7.7 8.3 7.8 7.7 7.9 7.4 8.1 7.6 7.5 7.9 7.7 7.7 7.8
7.2 7.3 7.9 8.3 7.3 7.3 7.8 7.8 8.4 7.5 7.3 8.1 7.7 7.9 7.6
7.3 7.6 8.1 8.4 7.6 7.6 7.9 7.7 8.2 7.6 7.6 8.0 7.9 8.0 7.8
7.6 7.3 7.7 8.2 8.0 7.9 8.4 7.7 8.3 7.6 7.4 8.0 7.8 8.3 7.8
7.4 7.2 8.0 8.3 7.8 7.9 8.0 7.6 8.2 7.6 7.5 8.1 7.9 7.9 7.7
7.4 (2.4) 7.3 (2.5) 7.9 (2.1) 8.3 (1.7) 7.7 (3.1) 7.7 (2.9) 8.0 (3.3) 7.6 (2.6) 8.2 (1.8) 7.6 (1.3) 7.5 (1.9) 8.0 (1.0) 7.8 (1.6) 8.0 (2.5) 7.7 (1.1)
Average RSD (%)
8.0 4.1
7.7 3.5
7.7 5.1
7.8 3.6
7.9 4.3
7.8 4.0
7.8 (4.2)
This paper describes a new MICZE method for determination of salbutamol. Following validation, the obtained linearity, accuracy and precision demonstrated that the method could be confidently employed for the determination of salbutamol. The MICZE method was then applied for analysis of the salbutamol content of 15 tablets (Ventoline Depot) within a single-run. The determined content of the tablets deviated by only 2.4% from the declared amount. The inter-run variation of the determined content of each tablet was ≤3% (RSD). Analysis of the mixed tablet sample resulted in a slightly increased precision and a content of salbutamol in the tablets which concurred with the result from the individual determinations. Following preparation of a fresh BGE or a capillary replacement, an initial separation of the analyte and the injection marker should be performed in the single-injection mode in order to readjust the separation parameters, i.e., tPE1 and tPE2 . References
Oxprenolol was added into the tablet samples at a concentration of 0.409 mM. Other conditions as those given in Table 1, Fig. 3 and in Section 2.5, respectively.
3.3. Determination of salbutamol in tablet samples The salbutamol content of 15 Ventoline Depot tablets was determined, employing the validated MIZCE method, see Sections 2.2 and 3.2. The labeled content of salbutamol sulfate was declared to be equivalent to 8 mg salbutamol. An electropherogram from these determinations is shown in Fig. 4 and the results from the determinations are presented in Table 4. The average amount of salbutamol in these 15 tablets was 7.8 mg, with an intermediary precision of 4.2% (RSD). The inter-run precision based on six separate MICZE determinations of the individual tablets was between 1.0 and 3.3% (RSD), respectively. The intra-run precision determined by analyz-
[1] M.G. Khaledi (Ed.), High-Performance Capillary Electrophoresis: Theory, Techniques and Applications, Willey, New York, 1998. [2] L. Suntornsuk, J. Chromatogr. Sci. 45 (2007) 559. [3] K. Altria, A. Marsh, C. Sänger-van de Griend, Electrophoresis 27 (2006) 2263. [4] T.K. Natishan, J. Liq. Chromatogr. Rel. Technol. 28 (2005) 1115. [5] T.G. Morzunova, Pharm. Chem. J. 40 (2006) 158. [6] P.T.T. Ha, J. Hoogmartens, A.V. Schepdael, J. Pharm. Biomed. Anal. 41 (2006) 1. [7] S. Fanali, Z. Aturki, C. Desiderio, Enantiomer 4 (1999) 229. [8] G.K.E. Scriba, Electrophoresis 24 (2003) 2409. [9] M.T. Ackermans, J.L. Beckers, F.M. Everaerts, I.G.J.A. Seelen, J. Chromatogr. 590 (1992) 341. [10] U. Holzgrabe, D. Brinz, S. Kopec, C. Weber, Y. Bitar, Electrophoresis 27 (2006) 2283. [11] K.D. Altria, J. Chromatogr. 634 (1993) 323. [12] K.D. Altria, P. Franke, I. Gill, T. Hadgett, M.A. Kelly, D.R. Rudd, J. Pharm. Biomed. Anal. 13 (1995) 951. [13] J.L. Lambert, D.L. Middleton, Anal. Chem. 62 (1990) 1585. [14] J.E. Gómez, J.E. Sandoval, Electrophoresis 29 (2008) 381. [15] J.D.B. Zavaleta, A. Chinchilla, A. Ramirez, K. Pao, S. Martinez, J.E. Nilapwar, Talanta 71 (2007) 192. [16] H. Lodén, A. Amini, Electrophoresis 28 (2007) 1548. [17] Y. Chen, W. Wang, J.P. Duan, H.Q. Chen, G.N. Chen, Electroanalysis 17 (2005) 706. [18] E. Ekiert, C. Garcia-Ruiz, M.A. Garcia, M.L. Marina, Electrophoresis 24 (2003) 2680. [19] F.S. Felix, M.S.M. Quintino, A.Z. Carvalho, L.H.G. Coelho, C.L. do Lago, L. Angnes, J. Pharm. Biomed. Anal. 40 (2006) 1288. [20] P. Mikuˇs, I. Valáˇsková, E. Havránek, Arch. Pharm. Chem. Life Sci. 338 (2005) 498. [21] J. Vela, E.G. Yanes, A.M. Stalcup, Fresenius J. Anal. Chem. 369 (2001) 212. [22] A. Esquisabel, R.M. Hernandez, A.R. Gascon, M. Igartua, B. Calvo, J.L. Pedraz, J. Pharm. Biomed. Anal. 16 (1997) 357. [23] L. Mälkki-Laine, E. Hartikainen, J. Chromatogr. A 724 (1996) 297. [24] M.J. Mills, J. Maltas, W.J. Lough, J. Chromatogr. A 759 (1997) 1. [25] A. Amini, H. Lodén, C. Pettersson, T. Arvidsson, Electrophoresis 29 (2008) 1. [26] A. Amini, V. Barclay, T. Rundlöf, S. Jönsson, A. Karlsson, T. Arvidsson, Chromatographia 63 (2006) 143. [27] D. Corradini, A. Rhomberg, C. Corradini, J. Chromatogr. A 661 (1994) 305. [28] K.D. Altria, H. Fabre, Chromatographia 40 (1995) 313.