CE CURRENTS An Introduction to the Theory and Application of

CE CURRENTS An Introduction to the Theory and Application of Microemulsion Electrokinetic Chromatography 1GlaxoSmithKline P.-E. Mahuzier,1,2 M.S. Aurora Prado,1,3 B.J. Clark,2 E.R.M. Kedor-Hackmann3 and K.D. Altria,4 R&D, Ware, Hertfordshire, UK, 2Drug Design, Bradford School of Pharmacy, University of Bradford, West Yorkshire, UK, 3Department of Pharmacy, College of Pharmaceutical Sciences, University of São Paulo, Brazil, 4GlaxoSmithKline, Harlow, Essex, UK. Microemulsion electrokinetic chromatography (MEEKC) is an electrodriven separation technique. Separations are achieved using microemulsions, which are nanometer-sized oil droplets suspended in an aqueous buffer. The surface tension between the oil and water components is reduced by covering the oil droplet with an anionic surfactant, such as sodium dodecylsulphate (SDS), and a co-surfactant, such as a short-chain alcohol. The number of publications describing the use of MEEKC is currently rising as awareness of the separation possibilities that MEEKC offers increases. Introduction In the last few years microemulsion electrokinetic chromatography (MEEKC) has become an important field of research in capillary electrophoresis offering a large range of applications. MEEKC is an electrodriven separation technique that offers the possibility of highly efficient separation of both charged and neutral solutes covering a wide range of water solubilities. Microemulsions are solutions containing nanometre-size droplets of an immiscible liquid dispersed in an aqueous buffer. The droplets are coated with a surfactant to reduce the surface tension between the two liquid layers allowing an emulsion to form. The surface tension of the droplet is further lowered by the addition of a short-chain alcohol, such as butanol, that stabilizes the system. The diameter of the oil droplets is below 10 nm giving an optically transparent emulsion; larger droplets would scatter light making an opaque emulsion. 2 A microemulsion containing ionic surfactant allows chromatographic separation to be obtained as solutes can partition between the charged oil droplets and the aqueous buffer phase. Water-insoluble compounds will favour inclusion into the oil droplets rather than into the buffer phase. Micelles are used in micellar electrokinetic chromatography (MEKC), which is another capillary electrokinetic technique operating in a similar manner to MEEKC. Solutes are more easily able to penetrate the surface of a droplet than the surface of a micelle, which is much more rigid. This ability allows MEEKC to be applied to a wider range of solutes than MEKC. High-pH buffers such as borate and phosphate are generally used in MEEKC. These buffers generate a high electroosmotic flow (EOF) when the voltage is applied across the capillary filled with the buffer. Surfactant-coated oil droplets are negatively charged (Figure 1) and, therefore, attempt to migrate towards the anode (away from the detector) when the voltage is applied. However, the EOF is sufficiently strong to eventually sweep the oil droplets through the detector to the cathode. If a solute is ionized then it will electrophoretically migrate according to its size and number of charges when the voltage is applied. Repulsion from negatively charged droplets will occur if the solute is also negatively charged. Conversely if the solute is positively charged it may have ion-pair type interaction with the negatively charged droplets. The migration time obtained in MEEKC for ionized solutes reflects a combination of both the electrophoretic and chromatographic behaviour of the solute ions. Applications Some applications of MEEKC are summarized in Table 1, which gives an appreciation of the separation possibilities LC•GC Europe January 2003 CE Currents that MEEKC may offer for the analysis of pharmaceutical and non-pharmaceutical solutes. Solubility (Log P) of a neutral solute can be directly assessed from migration time data obtained in MEEKC.1–3 Extremes of pH (i.e., pH 1.19 and pH 12) can be used to measure the Log P of acids and bases, respectively, in their uncharged forms. Figure 2(a) shows separation of eight neutral solutes, with known Log P values, using a sodium dodecylsulphate (SDS)–octane–butan-1-ol microemulsion. The migration times were plotted (Figure 2(b)) to generate a calibration graph. Pharmaceuticals: Pharmaceutical analysis is the most frequent application3 of MEEKC for which it has been used to separate and quantify a range of pharmaceutical classes. A wide range of water-soluble and insoluble acidic drugs have been resolved by a high-pH MEEKC method using a single set of operating conditions.3 These include a range of related cephalosporins, Pharmaceutical analysis is the most frequent application of MEEKC for which it has been used to separate and quantify a range of pharmaceutical classes. Figure 1: Principles of MEEKC. (Reproduced with permission from reference 3.) Microemulsion velocity ve ve EOF Droplet SDS surfactant Table 1: Summary of selected microemulsion applications. Analytes Solubility determinations Basic drugs 4-hydroxybenzoate preservatives Analgesic cardiac Naphthalene, guaiphenesin, 4-hydroxyacetophenone, paracetamol, sorbic acid, amitriptyline hydrochloride Naproxen and rizatriptan Ingredients of an ointment: diphenyldramine hydrochloride, phenylephrine hydrochloride, hydrocortisone acetate Hop bitter acids Vitamins Water- and fat-soluble vitamins Agrochemicals Pesticides Proteins Bases and nucleosides Sulphated disaccharides derived from glycosaminoglycans Diphenylhydrazones of dicarbonylsugars Separation of saturated fatty acids Separation medium/buffer Heptane–butan-1-ol–SDS–sodium borate buffer (pH 12) or sodium phosphate buffer (pH 7) Octane–butan-1-ol–SDS–sodium borate buffer (pH 9.2) Octane–butan-1-ol–SDS–phosphate buffer (pH 2.1) Heptane–butan-1-ol–SDS–sodium borate buffer (pH 9.2) Octane–butan-1-ol–SDS–sodium borate buffer (pH 9.2) Ref. 1, 2 3 4, 5 6 7 Octane–butan-1-ol–SDS–sodium borate buffer (pH 9.2) Octane–butan-1-ol–propan-1-ol–SDS–sodium borate buffer (pH 9.2) 8 9 Heptane–butan-1-ol–SDS–sodium borate buffer (pH 9.2) Hexane–methanol–sodium phosphate buffer Octane–butan-1-ol–SDS–sodium borate buffer (pH 8.5) Octane–butan-1-ol–SDS–potassium dihydrogen phosphate–sodium borate buffer (pH 7) Heptane–butan-1-ol–SDS–sodium phosphate/borate buffer (pH 7.0) Heptane–butan-1-ol–SDS–sodium borate buffer (pH 9.2) Toluene–butan-1-ol–SDS–sodium carbonate buffer (pH 10) Octane–butan-1-ol–SDS–sodium borate buffer (pH 9.3) Octane–butan-1-ol–SDS–sodium borate buffer (pH 8.0) Heptane–butan-1-ol–sodium cholate–sodium borate buffer (pH 10.2) 10 11 12 13 14 15 16 17 18 19 www.lcgceurope.com 3 CE Currents acetylsalicylic acid and insoluble drugs such as ibuprofen, indomethacin and troglitazone. The method was used to quantify levels of troglitazone in a tablet formulation. A result of 199.4 mg troglitazole was obtained3 compared with the label claim of 200 mg. MEEKC separations of a range of watersoluble and insoluble basic drugs (including terbutaline, bupivacaine and amitryptyline) were also resolved by the same standard operating conditions (Figure 3) with no evidence of peak tailing. The separation of basic analytes achieved in MEEKC is based on solute partitioning into the droplet, ionpair interaction with the surface of the droplet and electrophoretic migration of the positively charged compound. To eliminate the ion-pair and migration aspects it is possible to employ high-pH (pH 13)1,3 microemulsions where the basic drug will be neutral and will separate solely by partitioning with the droplet. Fast separation can be achieved using a low-conductivity microemulsion buffer4 prepared with ethyl acetate. Ethyl acetate Figure 2: (a) Separation of a range of phenones by MEEKC. Separation conditions: 0.81% w/w octane, 6.61% w/w butan-1-ol, 3.31% w/w sodium dodecyl sulphate and 89.27% w/w 10 mM sodium tetraborate buffer, 15 kV, 30 cm 3 50 mm i.d. capillary (detection window at 22 cm), 40 °C, 200 nm (Reproduced with permission from reference 3), and (b) plot of MEEKC migration times versus Log P data for a range of phenones. (a) 2 120 100 Response 4 80 60 40 20 4 5 6 7 8 9 1 3 5 6 7 8 9.83 Migration time (min) (b) 5.0 4.5 4.0 3.5 Log P 3.0 2.5 2.0 1.5 1.0 0.5 0 0 2 4 6 8 10 Migration time (min) Peaks: 1 phenacetine, 2 acetophenone, 3 acetanilide, 4 propiophenone, 5 butyrophenone, 6 valerophenone, 7 hexanophenone, 8 heptanophenone. has a lower surface tension than oils such as heptane and octane. Therefore, a lower concentration of surfactant can be used allowing high voltages to be applied across the capillary without generating excessive current levels. Analysis of neutral components such as guaiphenesin, 4-hydroxyacetophenone and benzophenone (Figure 4) was possible in a very short time.4 When using a high-speed separation buffer the injection repeatability was excellent over a range of 70 runs using the same set of vials. Analgesic cardiac and glycosides steroid drugs are highly insoluble,6 neutral and possess limited chromophores. MEEKC has been successfully used to separate a range of related glycosides6 with detection at low UV wavelength. A single set of MEEKC operating conditions has been shown to be useful for the analysis of sweeteners such as aspartame and saccharin, and preservatives such as parabens. The simultaneous separation of methyl, ethyl, propyl and butyl paraben5 (Figure 5), which are used as antimicrobial preservatives in food products, cometics and pharmaceuticals, was performed using a low-pH microemulsion. A selective separation was achieved against parahydroxybenzoic acid,5 which is the major degradation impurity of the preservatives. The method was employed to assay methyl and propyl paraben (Table 2). Successful validations were obtained (Table 2) for methyl and propyl paraben including linearity and limit of detection. The results obtained (Table 2) demonstrated that the MEEKC method was suitable for determination of paraben concentrations, and no impurities were detected in the batches tested above the limit of detection of 0.1% w/w. Non-pharmaceuticals: Hop bitter acids are present in the hops used to manufacture beer. The levels and composition of these acids affect the quality of the hops and it is, therefore, tested before use in beer manufacture. MEEKC has been shown to give accurate and precise data for this analysis.10 Resolution and separation efficiency was shown to be superior to that obtained by MEKC and the analysis time was shown to be shorter than for HPLC (typically 45–60 min). Water-soluble acidic vitamins such as nicotinic acid and vitamin C can be analysed using capillary electrophoresis (CE) with high-pH borate or phosphate buffers. However, the fat-soluble vitamins such as vitamins A and E are neutral, have LC•GC Europe January 2003 4 CE Currents Figure 3: Separation of a range of water-soluble and insoluble basic drugs. Separation conditions : 0.81% w/w octane, 6.61% w/w butan-1-ol, 3.31% w/w sodium dodecyl sulphate and 89.27% w/w 10 mM sodium tetraborate buffer, 15 kV, 30 cm 3 50 mm i.d. capillary (detection window at 22 cm), 40 °C, 200 nm. (Reproduced with permission from reference 3.) 3 2 20 19 18 1 17 16 15 14 1 2 3 4 Time (min) Peaks: 1 lamidurine, 2 terbutaline, 3 sumatriptan, 4 6 lignocaine, 7 bupivacaine, 8 amitryptyline. alosetron, 5 clenbuterol, 5 6 7 6 4 5 8 7 Figure 4: Separation of guaiphenesin (G), 4-hydroxyacetophenone (H) and benzophenone (Bph) in a high pH microemulsion (pH 5 8.1). The sample was injected for 2 s at 220 mbar (“short-end” of the capillary). Separation conditions: 0.5% w/w ethyl acetate, 1.2% w/w butan-1-ol, 0.6% w/w sodium dodecyl sulphate and 97.7% w/w 100 mM Tris, constant current of 100 µA (equivalent to 225kV), 33 3 50 µm i.d. capillary (detection window 24.5 cm), 40 °C, 214 nm. Elution order guaiphenesin (G), 4-hydroxyacetophenone (H), benzophenone (BPh). (Reproduced with permission from reference 4.) 180 160 140 Response 120 100 80 60 40 70 20 1 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 G H Bph Migration time (min) poor water solubility and require the use of a chromatographic-based CE method. MEEKC has been shown to be useful for the simultaneous determination of waterand fat-soluble vitamins.11,12 In these conditions lypophilic compounds are completely solubilized in the microemulsion and there are no precipitation problems. A complex multivitamin pharmaceutical formulation was successfully resolved using an SDS–octane–butan-1-ol microemulsion. Low-pH microemulsion was also suitable for the separation of vitamins A, E and D3.12 Water-insoluble neutral compounds, such as phenylurea herbicides, require chromatographic-based methods whilst water-soluble acidic compounds, such as chlorinated acids and phenols, can be analysed using high-pH CE buffers. Resolution of six phenylureas and chlorsulfuron was achieved using an SDS–octane–butan-1-ol microemulsion system.13,14 MEEKC has also been used to separate a range of proteins. These were separated, based on their hydrophobicities, using an SDS–heptane–butan-1-ol microemulsion in borate buffer. Resolution of the separated proteins was strongly affected by the SDS concentration. The resolutions obtained for ribonuclease A, carbonic anhydrase II, lactoglobulin A and myoglobulin by MEEKC15 were better than conventional CE using a borate buffer or MEKC. Proteins are generally too large to partition into a micelle but can partition into the microemulsion droplet, which has a larger volume. The MEEKC method resolved both basic and acidic proteins and was applied to the analysis of a range of injection formulations containing various protein mixtures. Bases and nucleosides are well-known compounds as they are fundamental components of deoxyribonucleic acid (DNA) and ribonucleic acid (RNA). The electrophoretic behaviours of five bases and corresponding nucleosides were examined in comparison with those in normal capillary zone electrophoresis (CZE). MEEKC using toluene as the core oil in oil in water (o/w) microemulsion proved to be effective.16 From the dissociation constants of the compounds and the electrophoretic mobilities of the corresponding anions obtained by CZE, the distribution constants between microemulsion droplets and aqueous phase were calculated. The importance of an adsorption mechanism in the MEEKC system was suggested from the correlation between Log KD and Log P. MEEKC has been applied to the analysis 5 www.lcgceurope.com Response CE Currents Figure 5: Separation of a test mixture of parabens in low pH microemulsion (pH 5 2.1). Separation of the test mixture by MEEKC. Sample injected for 3 s at 20 mbar. Separation conditions: 0.81% w/w octane, 6.61% w/w butan-1-ol, 3.31% w/w sodium dodecylsulphate and 89.27% w/w 50 mM phosphate buffer pH 2.1, 211 kV, 33 cm 3 50 µm i.d. capillary (detection window at 25 cm), 40 °C, 200 nm. The running buffer was rinsed through the capillary for 2 min between each run. Elution order butylparaben (BP), propylparaben (PP), ethylparaben (EP), methylparaben (MP), internal standard (ISS) and 4-hydroxybenzoic acid (Ac). (Reproduced with permission from reference 5.) EP PP BP MP Table 2: Assay results for industrial samples of paraben. (Reproduced with permission from reference 5.) Methyl paraben sample 1 Methyl paraben sample 2 Propyl paraben sample 1 Propyl paraben sample 2 Linearity 50–150% w/w methyl paraben Linearity 50–150% w/w propyl paraben 99.3% w/w 100.2% w/w 100.8% w/w 102.3% w/w Slope 0.01063 Constant 0.0721 r2 0.9999 Slope 0.0179 Constant 0.0795 r2 0.9998 300 280 260 240 220 200 Response 180 160 140 120 100 80 60 40 20 0 0 2 ISS de Amparo à Pesquisa do Estado de São Paulo of Brazil for financial support and fellowships (process FAPESP 98/03912-9). References 1. Ac 2. 3. 4. 4 6 8 5. 6. 7. Migration time (min) of sulphated disaccharides derived from glycosaminoglycans.17 Derivatized nonsulphated, mono-, di- and trisulphated disaccharides were completely separated using the microemulsion octane–SDS–butan-1-ol in sodium borate buffer. Agreement of the obtained disaccharides composition with literature values showed that MEEKC can be used for the analysis of glycosaminoglycans. A 10-component carbohydrate mixture that had been derivatized with o-phenylenediamine was resolved by MEEKC18 using an SDS–octane–butan-1-ol microemulsion system. However, a borate–SDS MEKC system only permitted resolution of the test mixture into three multicomponent peaks. The MEEKC method was then successfully used to profile the carbohydrate content in Daedalea quercina cultivates. Fatty acids are difficult to analyse by conventional CE as they have poor aqueous solubility and low UV activity. It is possible to separate them by CE using a 6 high-pH buffer coupled with indirect UV detection but these have limited sensitivity. Derivatives of fatty acids, such as phenacyl esters, are often prepared to give enhanced UV detection possibilities. An MEEKC method with a cholate–heptane–butan-1ol–borate microemulsion19 has been used to separate fatty acid esters ranging from C2–C20. This compared favourably with a cholate-based MEKC method, which was only able to separate acids up to C8. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. Conclusions MEEKC in CE offers fast, non-expensive analysis. A standard set can be used for a wide range of pharmaceutical and nonpharmaceutical applications including drugs, excipients and vitamins. Optimization of microemulsion composition can also be useful for specific applications, and faster separations can be obtained by changing oil and surfactant concentrations and types. S.J. Gluck et al., J. Chromatogr. A, 744, 141–146 (1996). Y. Ishihama, Y. Oda and N. Asakawa, Anal. Chem., 68, 1028–1032 (1996). K.D. Altria, J. Chromatogr. A, 892, 171–186 (2000). P.-E. Mahuzier, B.J. Clark and K.D. Altria, Electrophoresis, 22, 3819–3823, (2001). P.-E. Mahuzier, K.D. Altria and B.J. Clark, J. Chromatogr. A, 924, 465–470, (2001). L. Debusschère et al., J. Chromatogr. A, 779, 223–227 (1997). K.D. Altria, B.J. Clark and P.-E. Mahuzier, Chromatographia, 52, 758–768 (2000). P.-E. Mahuzier et al., J. Sep. Sci., 24, 784–788 (2001). H. Okamoto et al., J. Chromatogr. A, 929, 133–141 (2001). R. Szücs, E. Van Hove and P. Sandra, J. High Resol. Chromatogr., 19, 189–192 (1996). M.S. Bellini et al., J. Chromatogr. B, 741, 67–75 (2000). J.M. Sánchez and V. Salvadò, J. Chromatogr. A, 950, 241–247 (2002). L. Song et al., J. Chromatogr. A, 699, 371–382 (1995). W.L. Klotz, M.R. Schure and J.P. Foley, J. Chromatogr. A, 930, 145–154 (2001). G.-H. Zhou, G.-A. Luo and X.-D. Zhang, J. Chromatogr. A, 853, 277–284 (1999). T. Furumoto et al., Electrophoresis, 22, 3438–3443 (2001). O. Mastrogianni et al., Electrophoresis, 22, 2743–2745 (2001). I. Mikík, J. Gabriel and Z. Deyl, J.Chromatogr. A, 772, 297–303 (1997). I. Mikík and Z. Deyl, J. Chromatogr. A, 807, 111–119 (1998). Acknowledgements The authors would like to thank the Fundação LC•GC Europe January 2003 CE Currents Pierre-Etienne Mahuzier is currently concluding a PhD into the theory and applications of microemulsion CE jointly with GlaxoSmithKline, Ware, UK and Bradford University, UK. María S. Aurora Prado has a PhD in pharmaceutical sciences and is currently a researcher in the Department of Analytical Chemistry at the Chemistry Institute, University of São Paulo-USP. Her area of research focuses on the development and validation of methodologies for the quantification of drugs in pharmaceutical substances. Brian J. Clark is Associate Dean of Research for the School of Life Sciences at the University of Bradford, UK. Erika R.M. Kedor-Hackmann is Professor at the University of São Paulo-USP, São Paulo, College of Pharmaceutical Sciences, (Department of Physical Chemistry, Quality Control of Medications and Cosmetics) with teaching responsibilities in the quality control disciplines. Her area of research focuses on the development of methodologies for the quantification of drugs in pharmaceutical substances and cosmetics. “CE Currents” editor Kevin D. Altria is senior principal scientist at GlaxoSmithKline, Harlow, Essex, UK, and is a member of the Editorial Advisory Board of LC•GC Europe. Direct correspondence about this column to “CE Currents,” K.D. Altria, Building H89 Location 2-030, South Site, Product Line Extensions, GlaxoSmithKline, New Frontiers Science Park, Third Avenue, Harlow, Essex, CM19 5AW, UK, tel. +44 1279 643742, fax +44 1279 643953, e-mail: KDA8029@gsk.com, website: http://www.ceandcec.com/ www.lcgceurope.com 7