Microemulsion electrokinetic chromatography – or solvent-modified micellar electrokinetic chromatography?

Microemulsion electrokinetic chromatography – or solvent-modified micellar electrokinetic chromatography?

614 trends in analytical chemistry, vol. 20, no. 11, 2001 Microemulsion electrokinetic chromatography ^ or solvent-modi¢ed micellar electrokinetic c...

188KB Sizes 2 Downloads 242 Views

614

trends in analytical chemistry, vol. 20, no. 11, 2001

Microemulsion electrokinetic chromatography ^ or solvent-modi¢ed micellar electrokinetic chromatography? Steen Honoreè Hansen*, Charlotte Gabel-Jensen, Dina Taw¢k Mohamed El-Sherbiny Department of Analytical and Pharmaceutical Chemistry, Royal Danish School of Pharmacy, Universitetsparken 2, DK-2100 Copenhagen, Denmark

Stig Pedersen-Bjergaard

School of Pharmacy, University of Oslo, Oslo, Norway An overview of the microemulsion electrokinetic chromatography technique and its ¢elds of applications in analytical chemistry are given. The separation mechanisms involved are discussed and the technique is compared to solvent-modi¢ed micellar electrokinetic chromatography. z2001 Elsevier Science B.V. All rights reserved. Keywords: Microemulsion electrokinetic chromatography; Miceller electrokinetic chromatography; Pseudo-stationary phases

1. Introduction A microemulsion is a modi¢cation of a micellar system where a lipophilic organic solvent has been dissolved in the micelles. In order to stabilise the microemulsion by lowering the surface tension between the two liquid phases, a more hydrophilic organic solvent is often added as a co-surfactant. The microemulsions are de¢ned as macroscopically homogeneous, fully optically transparent £uids having more than one liquid phase. The ¢rst description of a transparent mixture of water, a hydrocarbon and a suitable hydrophilic solvent dates back to 1943 [ 1 ] and it was not until 1959 [ 2 ] that such mixtures were called microemulsions. The high solubilising ability of these emulsions [ 3 ] has been extensively used in industry [ 4,5 ]. In separation sciences only a few examples of the use of microemulsions in high-performance *Corresponding author. Tel.: +45 (35) 30-6256; Fax: +45 (35) 30-6010. E-mail: [email protected] Abbreviations: MEEKC, microemulsion electrokinetic chromatography; MEKC, micellar electrokinetic chromatography

liquid chromatography ( HPLC ) have been reported [ 6^8 ]. However, after the ¢rst report [ 9 ] on the use of a microemulsion as a running buffer in capillary electrophoresis ( CE ), a still increasing number of papers describing different applications of microemulsion electrokinetic chromatography (MEEKC ) have been published. MEEKC has been shown to be useful for the separation of very lipophilic substances like polyaromatic hydrocarbons [ 9,10 ], steroids [ 11 ] and fat-soluble vitamins [ 12,13 ], and very hydrophilic substances like water-soluble vitamins [ 12,14 ], sugars [ 15 ] and proteins [ 16 ], as well as pharmaceuticals [ 12,17 ] and natural products [ 18^20 ]. Furthermore, characterisation of the solubility or log P of substances has been a major area of application for MEEKC [ 21^ 25 ]. Two reviews on MEEKC have also been published [ 26,27 ], and as MEEKC may be considered a special case of micellar electrokinetic chromatography (MEKC ), a review on this [ 28 ] should also be consulted. A typical microemulsion used for MEEKC may consist of 0.8% n-octane+3.3% sodium dodecylsulfate (SDS )+6.6% n-butanol+89.3% 10 mM borate buffer pH 9.2. Recently, some more fundamental studies of MEEKC have been performed [ 29^31 ] where a number of system parameters have been investigated and these are discussed in more detail in the following sections.

2. Separation mechanisms In the MEEKC system, outlined in Fig. 1, the neutral solutes will distribute between the lipophilic microemulsion droplet and the external aqueous

0165-9936/01/$ ^ see front matter PII: S 0 1 6 5 - 9 9 3 6 ( 0 1 ) 0 0 1 2 7 - 3

ß 2001 Elsevier Science B.V. All rights reserved.

TRAC 2751 30-11-01

615

trends in analytical chemistry, vol. 20, no. 11, 2001

Fig. 1. Schematic presentation of MEEKC.

phase; it is thus a chromatographic system driven by electroosmosis. As the microemulsion droplets carry a number of negative charges due to the SDS molecules, they migrate towards the anode opposite to the electroosmotic £ow ( EOF ) and thus the more lipophilic substances will exhibit the longest total migration times in a normal instrumental setup. A very lipophilic substance, fully distributed to the core of the microemulsion droplets, will be retained in the system corresponding to the time ( tme ) needed for the EOF to bring the droplets from the injection point to the detection window. Neutral analytes will thus always exhibit a time of analysis tR between t0 ( no distribution to the microemulsion droplets and thus equivalent to EOF ) and tme ( the total time for a microemulsion droplet to travel from the point of injection to the detection window ): 0 1 B 1‡k C tR ˆ @ t0 AWt0 1‡ k tme The k-values, which are the mass distribution ratios of the solutes in the MEEKC system, may be calculated similarly to MEKC: tR 3t0  kˆ  13tR t0 W tme

Separation of solutes will occur when the distribution constants of the solutes differ. Anionic solutes, being hydrophilic in nature and of the same electrical charge as the SDS molecules, have no or only very little af¢nity towards the microemulsion droplets and this is primarily due to repulsion. Therefore the total migration of the anions only depends on the EOF and on the electrophoretic migration of the anions themselves. The separation of anionic solutes is thus more or less similar to what may be observed in a CE system using no additives. The separation mechanisms involved in the separation of cationic solutes are more complex. The cations may form ion-pairs with free SDS molecules in the aqueous phase as well as interact with the surface of the microemulsion droplets. Furthermore, the ion-pairs formed may distribute to the core of the microemulsion droplets. The time of analysis for cationic solutes is thus more dif¢cult to predict as a number of oppositely directed forces are involved in the total migration process. The overall result is that simultaneous separation of anionic, cationic as well as neutral solutes may be performed and a number of parameters may be used to control selectivity.

TRAC 2751 30-11-01

616

trends in analytical chemistry, vol. 20, no. 11, 2001

3. Modi¢cations of selectivity The ability to make changes in and to control selectivity is one of the most important concepts in separation science. It is therefore important to know by which parameters of a separation system changes in selectivity may be induced and also which parameters are of less or no signi¢cance. Studies of the effect on separation by changing the composition of the microemulsion have been performed [ 28 ]. Until recently, the lipophilic organic phase in the microemulsions was expected to provide selectivity changes when substituted by other solvents with different functional groups. However, only minor changes in selectivity are

Fig. 2. Comparison of MEEKC and MEKC. Capillary: 48.5 cm ( 40 cm to detector )U50 Wm; voltage: 20 kV; detection: 215 nm. Electrophoresis medium: ( A ) 0.8% 1-chloropentane+3.3% SDS+6.6% n-butanol+89.3% 10 mM sodium tetraborate pH 9.2 and ( B ) 100 mM SDS in 10 mM sodium tetraborate pH 9.2. Peak identi¢cation: 1, caffeine; 2, terbutaline; 3, tropic acid; 4, cinnamic acid; 5, pindolol; 6, hydrocortisone; 7, prednisolon.

Fig. 3. Analysis of raw opium. The raw opium was dissolved in the electrophoresis medium. Capillary: 32 cmU25 Wm with injection from the detector end ( 8.5 cm to detector ). Running buffer: 0.8% 1-chloropentane+3.3% lithium dodecylsulfate+6.6% n-butanol+89.3% 10mM sodium borate pH 9.2. Voltage: 330 kV. Temperature: 25³C. Detection: 214 nm. Peak identi¢cation: M, morphine; C, codeine; T, thebaine; N, noscapine and P, papaverine.

observed by replacing n-octane with either toluene, 1-chloropentane, diisopropyl ether, 2-octanone, butyl acetate, or n-octanol, and also only minor changes in total migration are observed when changing the amount of the lipophilic organic phase in the 0.5^2.0% range [ 29,30 ]. This suggests that analyte interactions with the lipophilic organic solvent in the core of the microemulsion droplets play a minor role in MEEKC, and for practical work this is a parameter of minor importance. A part of the explanation for this is of course the relatively high concentration ( 3.3%) of the surfactant (SDS ) compared to the concentration ( 0.8%) of the lipophilic organic solvent. Another variable parameter is the type of surfactant used to form the microemulsion. SDS is the preferred surfactant in most cases. It provides the largest separation window as the microemulsion droplet obtains the largest number of charges relative to the size and thus a high migration rate opposite to the EOF. If SDS is partly or fully replaced with zwitterionic, neutral, cationic or other anionic surfactants like 3-(N,N-dimethylmyristylammonium )propanesulfonate, polyoxyethylene sorbitan monolaurate, polyoxyethylene lauryl ether, Ncetyl-N,N,N-trimethylammonium bromide, and dioctylsulfosuccinate, some selectivity changes can be observed [ 29,30 ]. This suggests that the surfactants at the surface of the microemulsion droplets are

TRAC 2751 30-11-01

617

trends in analytical chemistry, vol. 20, no. 11, 2001

Fig. 4. Analysis of the drug substance bumetanide ( B ) for four impurities ( 1^4 ). Bumetanide was dissolved 2 mg / ml in 80% methanol. The four impurities were present corresponding to 0.2% of bumetanide. Capillary: 51 cmU50 Wm ( 42.5 cm to detector ). Running buffer: 0.8% n-octanol+2% lithium dodecylsulfate+6.6% n-butanol+90.6% 10 mM sodium borate pH 9.2. Voltage: +30 kV. Temperature: 25³C. Detection: 214 nm.

more important with respect to control of selectivity than the internal lipophilic core of the droplet. However, the most important factor for changing and controlling the selectivity was found to be the nature of the so-called co-surfactant [ 30,31 ], which is the more hydrophilic organic solvent added to the microemulsion in order to reduce the surface tension. Signi¢cant changes in selectivity and migration order were observed as n-butanol was replaced by methanol, ethanol, 1-propanol, acetonitrile, 2-ethoxy ethanol, or tetrahydrofuran. The majority of the interactions between the microemulsion droplets and the analytes obviously involved the co-surfactant and, for practical work, this is the primary parameter for rapid optimisation of the separation selectivity in MEEKC.

4. Separation ef¢ciency In several of the papers published on MEEKC the technique was found to give a separation ef¢ciency superior to the ef¢ciency obtained in MEKC [ 11,14,20,29,32 ]. The higher ef¢ciency reported in MEEKC is probably due to an improved mass transfer between the microemulsion droplet and the external aqueous phase mediated by the co-surfactant solvent. Also, the separation window is wider in MEEKC ( Fig. 2 ). The reason for this is the addition of the cosurfactant and the effect is similar to what is seen

when organic solvents are added to CE systems in general.

5. Applications MEEKC have been used for the separation of a wide range of test solutes covering most types of small molecules as mentioned in Section 1. However, many of the applications are based on test substances and not on real samples and the introduction of the technique into control laboratories is therefore not widespread. Two examples of the separation capability of MEEKC using real samples are shown in Figs. 3 and 4. Analysis of the opium plant exudate for ¢ve major alkaloids can be done within 120 s. The separation of these ¢ve alkaloids has previously been achieved by using non-aqueous CE [ 33 ] or by adding cyclodextrins to the running buffer [ 34 ]. The very fast separation obtained using MEEKC is only possible due to an exceptional high resolution between the analytes. The optimisation is further improved by: using a shorter capillary with a smaller internal diameter; lowering the conductivity by substituting SDS for lithium dodecylsulfate; and by performing the injection from the detector end of the capillary ( using reversed polarity ). The sample preparation is facilitated by the high solubilising ability of the microemulsion as a clear solution of opium is obtained in the microemulsion.

TRAC 2751 30-11-01

618

trends in analytical chemistry, vol. 20, no. 11, 2001

Fig. 5. Electropherogrammes for some of the 11 test solutes analysed using four different co-surfactants and comparing the MEEKC conditions with the MEKC conditions. MEEKC systems: 0.8% 1-octanol+3.3% SDS+6.6% co-surfactant+89.3% ( 95.9% when no cosurfactant ) 10 mM sodium borate pH 9.2 and MEKC systems: 3.3% SDS+6.6% co-surfactant+90.1% ( 96.7% when no co-surfactant ) 10 mM sodium borate pH 9.2. Capillary: 48.5 cmU50 Wm i.d., 40 cm to the detector; voltage: +20 kV and temperature: 25³C. Peak identi¢cation: ( FA ) formamide; ( 1 ) benzamide; ( 2 ) nicotinic acid methyl ester; ( 3 ) phenylacetamide; ( 7 ) p-chlorobenzamide; ( 8 ) prednisone; ( 9 ) cortisone; ( 10 ) p-bromobenzamide; ( 12 ) prednisolone; ( 13 ) ethyl-3-nitrobenzoate; ( 14 ) betametasone; ( 15 ) nicotinic acid butyl ester.

Also the testing of drug substances for impurities may be an important area of application for MEEKC [ 12 ] and in Fig. 4 one further example is given showing the separation of the impurities in the drug substance bumetanide. If performed by HPLC this separation has to be done either by gradient elution or by ion-pair chromatography.

6. Stability of the microemulsions The stability of the microemulsions may pose a problem as they only are stable within a certain relationship between the ingredients, but mixed

in the right proportions they are stable for months and probably for years. Thus, the microemulsion most frequently used in MEEKC comprises 0.8% n-octane+3.3% SDS+6.6% n-butanol+89.3% 10 mM borate buffer pH 9.2 and is highly stable, whereas the stability decreases as the amount of SDS is decreased. The stability also depends on the nature of the components. It is not possible to obtain a stable microemulsion using 0.8% n-octane and 1% SDS, but when substituting n-octane for n-octanol a very stable microemulsion is obtained [ 30 ]. During selectivity optimisation of microemulsions, the long term stability should be considered in order to develop robust methods.

TRAC 2751 30-11-01

619

trends in analytical chemistry, vol. 20, no. 11, 2001

7. Comparison with solvent-modi¢ed MEKC During the investigations of the importance of the co-surfactant in MEEKC, a number of experiments with MEKC adding the same co-surfactant, the same amount of SDS and the same buffer as used in the microemulsions were performed ( solvent-modi¢ed MEKC ). Also, in solvent-modi¢ed MEKC important selectivity changes were observed using different organic solvents and, moreover, an increased separation window and increased separation ef¢ciency compared to standard MEKC were observed. A comparison of these solvent-modi¢ed MEKC systems with similar MEEKC systems ( Fig. 5 ), where the only difference was the lipophilic organic phase forming the microemulsion, showed them to be very similar with respect to separation selectivity and ef¢ciency [ 31 ].

8. Conclusions During the decade that has passed since its birth, MEEKC has been shown to be a most powerful CE technique for the simultaneous separation of neutral, anionic and cationic solutes. A number of variables in the MEEKC systems make it possible to change and control selectivity and the so-called co-surfactant has been shown to be the most powerful of these. If the advantage of the high solubilising effect of the microemulsions is also utilised, a powerful system for analysis of hydrophobic substances is created. However, recent investigations have shown that solvent-modi¢ed MEKC may provide most of the same advantages in selectivity tuning and high separation ef¢ciency without the problem of the possible lack of stability of the microemulsions. Thus, taking present knowledge into consideration, solvent-modi¢ed MEKC should be preferred over standard MEKC because of the increased separation window and higher separation ef¢ciency. Solvent-modi¢ed MEKC may also be preferred over MEEKC in order to avoid the possible instability of the microemulsions. Although minor selectivity differences still exist between MEEKC and solvent-modi¢ed MEKC, selectivity tuning may easily be performed in solvent-modi¢ed MEKC by changing the hydrophilic organic solvent.

References [ 1 ] T.P. Hoar, J.H. Schulman, Nature 152 ( 1943 ) 102. [ 2 ] J.H. Schulman, W. Stoeckenius, L.M. Price, J. Phys. Chem. 63 ( 1959 ) 1677. [ 3 ] K.R. Wormuth, L.A. Cadwell, E.W. Kaler, Langmuir 6 ( 1990 ) 1035. [ 4 ] D.O. Shah, R.S. Schecter, Improved Oil Recovery by Surfactant and Polymer Flooding, Academic Press, New York, 1977. [ 5 ] M. Kahlweit, Science 240 ( 1988 ) 617. [ 6 ] M.A. Hernandez-Torres, J.S. Landy, J.G. Dorsey, Anal. Chem. 58 ( 1986 ) 744. [ 7 ] A. Berthod, M. De Carvalho, Anal. Chem. 62 ( 1990 ) 1402. [ 8 ] A. Berthod, O. Nicolas, M. Porthault, Anal. Chem. 64 ( 1992 ) 2267. [ 9 ] H. Watarai, Chem. Lett. 231 ( 1991 ) 391. [ 10 ] K.D. Altria, J. Chromatogr. A 892 ( 2000 ) 171. [ 11 ] L. Vomastova, I. Miksik, Z. Deyl, J. Chromatogr. B 681 ( 1996 ) 107. [ 12 ] K.D. Altria, J. Chromatogr. A 844 ( 1999 ) 371. [ 13 ] S. Pedersen-Bjergaard, Ò. N×ss, S. Moestue, K.E. Rasmussen, J. Chromatogr. A 876 ( 2000 ) 201. [ 14 ] R.L. Boso, M.S. Bellini, I. Miksik, Z. Deyl, J. Chromatogr. A 709 ( 1995 ) 11. [ 15 ] I. Miksik, J. Gabriel, Z. Deyl, J. Chromatogr. A 772 ( 1997 ) 297. [ 16 ] G.-H. Zhou, G.-A. Luo, X.-D. Zhang, J. Chromatogr. A 853 ( 1999 ) 277. [ 17 ] S. Terabe, N. Matsubara, Y. Ishihama, Y. Okada, J. Chromatogr. 608 ( 1992 ) 23. [ 18 ] L. Debusschere, C. Demesmay, J.L. Rocca, G. Lachatre, H. Lofti, J. Chromatogr. A 779 ( 1997 ) 227. [ 19 ] R. Szucs, E. van Hove, P. Sandra, J. High Resolut. Chromatogr. 19 ( 1996 ) 189. [ 20 ] G. Li, X. Chen, M. Liu, Z. Hu, Analyst 123 ( 1998 ) 1501. [ 21 ] M.H. Abraham, C. Treiner, M. Roses, C. Rafols, Y. Ishihama, J. Chromatogr. A 752 ( 1996 ) 243. [ 22 ] Y. Ishihama, Y. Oda, N. Asakawa, Anal. Chem. 68 ( 1996 ) 4281. [ 23 ] S.J. Gluck, M.H. Benko, R.K. Hallberg, K.P. Steele, J. Chromatogr. A 744 ( 1996 ) 141. [ 24 ] Y. Ishihama, Y. Oda, N. Asakawa, Anal. Chem. 68 ( 1996 ) 1028. [ 25 ] Y. Ishihama, Y. Oda, K. Uchikawa, N. Asakawa, Anal. Chem. 67 ( 1995 ) 1588. [ 26 ] H. Watarai, J. Chromatogr. A 780 ( 1997 ) 93. [ 27 ] K.D. Altria, J. Chromatogr. A 892 ( 2000 ) 171. [ 28 ] H. Nishii, J. Chromatogr. A 780 ( 1997 ) 243. [ 29 ] S. Pedersen-Bjergaard, C. Gabel-Jensen, S.H. Hansen, J. Chromatogr. A 897 ( 2000 ) 375. [ 30 ] C. Gabel-Jensen, S.H. Hansen, S. Pedersen-Bjergaard, Electrophoresis 22 ( 2001 ) 1330. [ 31 ] S.H. Hansen, C. Gabel-Jensen, S. Pedersen-Bjergaard, J. Sep. Sci., submitted. [ 32 ] Y. Mrestani, N. El-Mokdad, H.H. Ruettinger, R.H.H. Neubert, Electrophoresis 19 ( 1998 ) 2895. [ 33 ] I. BjÖrnsdottir, S.H. Hansen, J. Pharm. Biomed. Anal. 13 ( 1995 ) 1473. [ 34 ] I. BjÖrnsdottir, S.H. Hansen, J. Pharm. Biomed. Anal. 13 ( 1995 ) 687.

TRAC 2751 30-11-01