2503

Electrophoresis 2007, 28, 2503–2526

Kimberly A. Kahle Joe P. Foley Department of Chemistry, Drexel University, Philadelphia, PA, USA

Received December 11, 2006 Revised April 4, 2007 Accepted April 13, 2007

Review

Review of aqueous chiral electrokinetic chromatography (EKC) with an emphasis on chiral microemulsion EKC The separation of enantiomers using electrokinetic chromatography (EKC) with chiral microemulsions is comprehensively reviewed through December 1, 2006. Aqueous chiral EKC separations based on other pseudostationary phases such as micelles and vesicles or on other chiral selectors such as CDs, crown ethers, glycopeptides, ligand exchange moeities are also reviewed from both mechanistic and applications perspective for the period of January 2005 to December 1, 2006. Keywords: CE / Chiral EKC / Chiral separation / Cyclodextrins / Microemulsion EKC DOI 10.1002/elps.200600808

1

Introduction

CE as a separation technique has experienced widespread growth since its introduction in 1967 by Hjertén [1]. Since then, many advancements in instrumentation and additives have led to its industrial and academic acceptance. In the

Correspondence: Professor Joe P. Foley, Department of Chemistry, Drexel University, 3141 Chestnut Street, Philadelphia, PA 19104, USA E-mail: [email protected] Fax: 11-215-895-1265 Abbreviations: BOH, 1,10 -bi-2-naphthol; BNP, 1,10 -binaphthyl2,20 -diylhydrogenphosphate; BZN, benzoin; CDhm, 6-deoxy-6[2-(4-imidazolyl)ethylamino]-b-CD; CDmh, 6-deoxy-6-[4-(2-aminoethyl)imidazolyl]-b-CD; CD-NH3Cl, mono-6-ammonium-6deoxy-CD chloride; CE-
-CD, carboxyethyl-b-CD; 18C6H4, (1)-(18crown-6)-2,3,11,12-tetracarboxylic acid; CM-
-CD, carboxymethyl-b-CD; DDCV, dodecoxycarbonylvaline; DM-
-CD, dimethyl-b-CD; HDAS-
-CD, heptakis(2,3-diacetyl-6-sulfato)-CD; HDMS-
-CD, heptakis(2,3-dimethyl-6-sulfato)-CD; HMAS-
-CD, heptakis(2-O-methyl-3-O-acetyl-6-O-sulfo)cyclomaltoheptaose; HMS-CD, heptakis(2-O-methyl-6-O-sulfo)cyclomaltoheptaose; HPTMA-CD, 6-O-(2-hydroxy-3-trimethylammoniopropyl)-CD; HP-x-CD, hydroxypropyl-x-CD (x = a, b, g, or d); HpS-
-CD, heptakis-6-sulfato-CD; HS-
-CD, highly sulfated beta-CD; HS-ª-CD, highly sulfated g-CD; MEEKC, microemulsion EKC; poly-L-SUCL, poly(sodium N-undecenoxycarbonyl-L-leucinate; PSP, pseudostationary phase; QA-
-CD, quaternary ammonium-CD; S-CD, sulfated-CD; SDLIL, sodium N-[4-n-dodecyloxybenzoyl]-L-isoleucinate; SDLL, sodium N-[4-n-dodecyloxybenzoyl]-L-leucinate; SDLV, sodium N-(4-n-dodecyloxybenzoyl)-L-valinate; SI-S-
-CD, heptakis-(2,3-dihydroxy-6-O-sulfo)-CD; TM-
-CD, heptakis(2,3,6tri-O-methyl)-CD; L-UCLB, undecenoxycarbonyl-L-leucinol bromide; L-UCPB, undecenoxycarbonyl-L-pyrrolidinol bromide

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

1980s, several key developments were made including [1]: use of small inner diameter glass capillaries (75 mm) by Jorgenson et al. in 1981; the advent of EKC by Terabe et al. in 1984 (micellar EKC (MEKC)); the first chiral CE separation by Gassmann et al. in 1985; the first commercial instrumentation in 1988; and the first use of CDs in CE by Guttman et al. in 1988. The characteristics of CE that make it so appealing have been frequently noted [1–5] and include: high efficiencies, high resolving power, short equilibration times, short analysis times, ease of additive use, and low consumption of analytical reagents and sample. One area in which CE has been shown to be particularly useful is in chiral separations. The multitude of chiral reagents that can simply be added to the electrolyte solution continues to increase and allows many types of chiral analytes to be resolved. Some examples of chiral selectors include: surfactant aggregates, CDs, crown ethers, and glycopeptides. Due to the vast exploration in this subject, numerous reviews detailing fundamentals and applications can be found in the literature prior to 2005 [6–11] and over the past two years [12–15]. Reviews have also covered methods developed specifically for pharmaceutical compounds [16–22], biological compounds [23], chiral LC-MS/ CE-MS [24], pesticides [25], pollutants [26], and food additives [27, 28]. The literature review conducted herein focuses on chiral electrokinetic chromatographic techniques with charged chiral selectands (micelles, vesicles, microemulsions, CDs, crown ethers, glycopeptides, and ligand exchange). Publications from January 2005 to December 1, 2006 have been included with the exception of chiral microemulsion EKC (MEEKC) where all reports to date have been reviewed. The basic mechanisms of chiral recognition for each selector are also described. www.electrophoresis-journal.com

2504

2

Chiral separation mechanisms

There are two main approaches to chiral separations, indirect and direct. Indirect relies on derivatization of enantiomers to form diastereomers that can then be separated without chiral selectors and as such does not fall into the separation category of chiral EKC. The direct approach is the strategy employed in EKC where a temporary diastereomeric interaction occurs between the enantiomers and the chiral selectand. The two enantiomers can be separated when either their binding constants with the chiral agent or the mobilities of their diastereomeric complexes are different. In cases where one chiral selector is unable to provide sufficient enantioresolution, a second chiral agent can be added. Assuming that neither chiral selectand is inert with respect to the chiral analytes, the two selectands will either work constructively or destructively to improve or worsen the overall resolution, respectively. It is generally recognized that a three-point-interaction must take place for chiral discrimination [29]. The nature of these interactions can be electrostatic, hydrogen bonding, steric hindrance, p–p, ion– dipole, dipole–dipole, dipole-induced dipole, or van der Waals (in order of decreasing strength). Efforts to simplify chiral method development have taken the directions of theoretical model development and spectroscopic mechanism elucidation. Mathematical models for optimizing chiral selector conditions with CDs have been developed by Wren and Rowe [30–33]. Kafri and Lancet [34] examined chromatographic data and proposed the string model for enantiorecognition (SMED). Their goal was to provide guidelines for selection of a chiral agent to reduce the screening process. NMR is a popular approach for defining specific interactions leading to chiral recognition, typically for CD-analyte complexes [35, 36] but also for micellar studies [37]. One last aspect to note in chiral CE is the importance of elution order. Many separation methods are quantitative; to improve quantitation it is beneficial to have a minor peak elute prior to the major peak. This can be accomplished with chiral surfactant aggregates by switching to the opposite selector enantiomer, if available. However, the situation is not as easy with other types of selectands such as CDs. A recent review by Chankvetadze [38] focused on this special consideration.

3

Electrophoresis 2007, 28, 2503–2526

K. A. Kahle and J. P. Foley

Surfactant-based chiral selectors

Chiral selectors formed from chiral surfactants can be divided into three main groups: micelles, vesicles, and microemulsions, with the simplest being micelles. In general, the resolution of neutral analytes as well as charged ones can be achieved with surfactant-based pseudostationary phases (PSPs) where differential partitioning into the aggregate results in separation. The enantiomeric separations described in this section are restricted to those © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

achieved solely with chiral surfactant systems. Instances where a dual-chiral selectand scheme was used are detailed in Section 4.1.3. 3.1 Micelles In MEKC, a surfactant (anionic, cationic, or zwitterionic) above its CMC (CMC, minimum concentration required for the formation of micelles) is utilized as a PSP. Selectivity can be modified by changing surfactant concentration and/or identity; buffer concentration, identity, and/or pH; and additive content (including organic solvent). Typically in chiral MEKC, the chiral portion of the surfactant is located close to the surfactant polar head group, i.e., near the outer rim of the PSP. Thus, chiral recognition takes place in the outer portion of the micelle. Bile salts such as sodium cholate were among the first chiral surfactants used in MEKC [39]. Glycosidic surfactants have also been employed as chiral selectors [40]. In addition, synthetic chiral surfactants have been created, many of which are derivatized with amino acid functionalities. More recently, micelle polymers have appeared in the literature for enantiomeric separations [41, 42]. In contrast to conventional micelles, micelle polymers are formed via covalent bonds and do not have a CMC. Thus, they are more rigid and remain structurally stable in the presence of high concentrations of organic solvents, which also facilitates the adjustment and control of their concentration. The difference between conventional micelles and micelle polymers can be seen in Fig. 1 where the resulting PSPs were used for MEKC-ESI-MS analyses [43]. A large variety of enantiomeric separations have been published in the field of MEKC, with some utilizing multiple chiral selectors. This section of the review will cover chiral MEKC analyses where only a chiral micelle was employed. One such application used chiral bile salt micelles (sodium taurocholate) for the resolution of metyrosine enantiomers [44]. BGE composition was optimized for surfactant concentration, pH, buffer identity, buffer concentration, applied voltage, and temperature. The developed method was able to reproducibly separate metyrosine with a resolution of 2.1 and an LOD of 1.0 mg/mL. Enantiomers of a diethyleneglycol diester benzoporphyrin derivative were successfully separated using sodium cholate micelles in a borate buffer with 30% ACN (to prevent analyte aggregation) [45]. Better resolution was obtained when the surfactant concentration was increased to 150 mM; still higher concentrations had a negative effect on the separation and caused precipitation. Also reported in this study was an attempt to discriminate the enantiomers of tin ethyl etiopurpurin dichloride; however, baseline resolution could not be obtained even when a variety of CDs and cholate surfactants were employed separately and together. Subsequently in the literature, sodium cholate micelles were shown to effectively separate the flavanone-7-O-glycoside epimers of neohesperidin and naringin [46]. The chromatographic data were used to simulate the www.electrophoresis-journal.com

Electrophoresis 2007, 28, 2503–2526

CE and CEC

2505

Figure 1. Comparison of (A) low-molecular-mass monomeric (unpolymerized) micelles and (B) high-molecular-mass micelle polymer introduced to ESI-MS [43]. Reproduced with permission.

epimer interconversion process. Another application where the selector was sodium cholate micelles was the analysis of palonosetron hydrochloride stereoisomers by Tian et al. [47]. The electrolyte solution was optimized with respect to organic modifier, pH, selector concentration, and buffer concentration. Additionally, the applied voltage was varied to provide baseline separation of the target enantiomers. The use of chiral polymeric micelles as chiral selectors in MEKC-ESI-MS, the second such study, was published in 2005 by Akbay et al. [43]. Poly(sodium N-undecenoxycarbonyl-L-leucinate (poly-L-SUCL) was synthesized and employed for the concurrent testing of eight b-blockers. Optimization was performed for both separation and detection conditions. In terms of the MEKC system, an ammonium acetate buffer with triethylamine (TEA) had to be used for MS compatibility. During the optimization of surfactant concentration, it was observed that, as the surfactant concentration was increased, the resolution of hydrophilic bblockers increased whereas the resolution of the hydrophobic b-blockers decreased. Although seemingly contradictory, both results are most likely due to the well-known trend of increasing retention factor with increasing surfactant concentration; in the case of the hydrophilic compounds, their initially small retention factors increased to more optimal values whereas in the case of the hydrophobic compounds their already large retention factors increased to still less optimal values. A comparison was performed between L-SUCL and poly-L-SUCL, with the polymeric micelle © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

providing larger S/N, better enantioresolution, higher efficiencies, and a different elution order for each pair of enantiomers. Rizvi and Shamsi [48] conducted another set of experiments using micelle polymers with a focus on alkyl chain length and degree of polymerization (DP); the analytes were again b-blocker enantiomers. Chiral surfactants with chain lengths of 8, 9, 10, and 11 (C11 monomers with a terminal double bond and a terminal triple bond) were polymerized to form micelles. Characterization of the aggregates showed an increase in CMC with decreasing chain length and increase in degree of unsaturation. The most hydrophobic analytes were best separated with shorter chain surfactants at a lower concentration. In general, the micelle polymer formed from the monomer with a terminal triple bond provided higher efficiency than its double bond terminated analog. Efficiencies with a constant micellar concentration decreased with increasing chain length; the C11 triple bond terminated surfactant was an exception to this trend, giving better efficiency than the C10 surfactant. Seven pairs of enantiomers were successfully discriminated in this study, with the shortest chain surfactant performing the best. Variations in selectivity can be observed with the different micelle polymers for the separation of these enantiomers, see Fig. 2. Molecular micelles containing polymerized amino acid-based surfactants were examined for the separation of seven pairs of enantiomers (norlaudanosoline, laudanosoline, laudanosine, chlorthalidone, and three benzoin derivatives) [49]. A www.electrophoresis-journal.com

2506

K. A. Kahle and J. P. Foley

Electrophoresis 2007, 28, 2503–2526

Figure 2. Comparison of 25 mM poly-L-SOcCL, poly-L-SNoCL, poly-L-SDeCL, and poly-L-SUCL (all polymerized at 56CMC) for simultaneous enantioseparation of seven chiral b-blockers (1,10 = atenolol, 2,20 = metoprolol, 3,30 = pindolol, 4,40 = oxprenolol, 5,50 = talinolol, 6,60 = alprenolol and 7,70 = propranolol) [48]. MEKC conditions: pH 8.8, 25 mM NH4OAc/25 mM TEA, 257C. Pressure injection: 40 mbar/s; 20 kV applied for separations; UV detection at 214 nm. Reproduced with permission.

total of 18 different PSPs were tested including dipeptide surfactants with a single chiral center (six variants) or a double chiral center (nine variants) and single amino acid surfactants with a single chiral center (three variants). The compound set was divided into two groups based on steric hindrance. The aspects investigated included: surfactant head group impact (single vs. dipeptide), dipeptide amino acid order effect, chiral selector head group steric effects, and the role of selectand chiral center location. The results indicated that enantioselectivity could be improved with the usage of dipeptide containing PSPs. In terms of analyte structure, more sterically hindered compounds were better separated with less sterically hindered micelle polymers and vice versa. Another study on MEKC-ESI-MS was conducted using micelle polymers for the enantioresolution of oxazepam, lorazepam, and nefopam enantiomers [50]. Six different molecular micelles were screened (three each of carbamate-type and amide-type) with poly(sodium N-undecenoyl© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

L-leucinate) (poly-L-SUL) providing some enantioselectivity for analytes. The effect of organic modifier was studied using poly-L-SUL and showed some improvements with the addition of ACN; however, baseline resolution was still not obtained for all enantiomers. Therefore, further method development was performed using the dipeptide polymeric micelle (poly(sodium N-undecenoxy carbonyl-L,L-leucyl-valinate) (poly-L,L-SUCLV). Optimization of electrolyte and instrumental parameters led to sufficiently resolved peaks. An interesting observation was that MS signal intensities were PSP-dependent as well as analyte-dependent. Specifically, the MS signal depended on both the PSP identity and analyte proton affinity where poly-L-SUCL gave the lowest S/ N ratio for all analytes and dipeptide-based PSPs generally gave higher S/N ratios. The first use of chiral ionic liquids in MEKC was reported by Rizvi and Shamsi in 2006 [51]. The synthesis and characteristics of two chiral ionic liquids,

www.electrophoresis-journal.com

CE and CEC

Electrophoresis 2007, 28, 2503–2526

undecenoxycarbonyl-L-leucinol bromide (L-UCLB) and undecenoxycarbonyl-L-pyrrolidinol bromide (L-UCPB), were described. The resulting cationic chiral selectors were employed for the separation of acidic enantiomers (including phenoxypropionic acid and a-bromophenylacetic acid) with conventional and polymeric micelles. The flexibility difference between the two surfactants led to enhanced resolution for certain selectand/analyte combinations. Analysis of a-bromophenylacetic acid was better with L-UCPB and monomeric micelles; micelle polymers resulted in reduced resolution. Phenoxypropionic acid enantiomers were better resolved with L-UCLB and did not change significantly when micelle polymers were used instead of monomeric micelles. Lastly, a comparison was made between polymeric anionic chiral surfactants of polysodium N-undecenoxycarbonyl-L-leucine sulfate (polyL-SUCLS) and poly-L-SUCL and the two new cationic surfactants (L-UCLB and L-UCPB). Only slight resolution (Rs of 0.62) could be achieved with poly-L-SUCLS or poly-LSUCL, due to charge repulsion between the anionic selector and the negative analytes. 3.2 Vesicles Vesicles are slightly more complex, larger in size, and more rigid than micelles. A drawback to using this PSP is that the larger size can decrease LOD because of increased scattering. One of the advantages is that larger elution ranges can be obtained due to the vesicles having a higher charge density. These aggregates are formed by either combining oppositely charged surfactants in an appropriate ratio or by using double tailed surfactants. A bilayer surfactant structure is formed that envelops a water core, as shown in Fig. 3, and the resulting solution typically has a bluish hue. Among the surfactant-based PSPs, vesicles have received less attention with only four publications total detailing chiral separations. There have been two papers on chiral vesicle EKC for the 2005–2006 time period.

2507

Mohanty and Dey [53] reported a chiral vesicle composed of sodium N-(4-n-dodecyloxybenzoyl)-L-valinate (SDLV) for the atropisomeric/enantiomeric separation of 1,10 -bi-2naphthol (BOH), 1,10 -binaphthyl-2,20 -diamine (BDA), 1,10 binaphthyl-2,20 -diylhydrogenphosphate (BNP), benzoin (BZN), and Tröger’s base (TB). The analysis of BOH and BDA was conducted with an SDLV concentration of 2 mM, whereas BNP required 5 mM for adequate separation. The applied voltage was studied with a value of 15 kV chosen as a compromise between long retention times and degraded peak shapes from excessive Joule heating while maintaining decent resolution. The buffer concentration and pH were optimized to improve resolution and efficiency with the best conditions varying slightly from analyte to analyte. The best separation of BZN enantiomers (Rs = 1.61) was found with 4 mM SDLV in 60 mM borate buffer at a pH of 10.3. The final chiral compound in this study, TB, could not be baseline resolved even under optimized conditions (Rs = 1.06). The authors state that the resolutions achieved with SDLV vesicles were higher than those obtained with other chiral monomeric surfactant systems. In 2006, Mohanty and Dey [54] studied sodium N-[4-ndodecyloxybenzoyl]-L-leucinate (SDLL) and sodium N-[4-ndodecyloxybenzoyl]-L-isoleucinate (SDLIL) chiral vesicles, using the same analytes as noted above, with more focus on chiral recognition mechanisms. The researchers optimized the buffer concentration and pH with the best separations obtained under identical conditions as their previous study with SDLV. Surfactant concentration variations changed the resolution but not the selectivity for the different analytes. Enantioselectivities with SDLL, SDLIL, and SDLV were compared to gain insight into the chiral discrimination process. For the systems used, hydrophobic-type interactions did not contribute significantly to chiral selectivity; steric hindrance near the chiral center of the surfactant, however, played a large role in enantiomer differentiation. If the surfactant head group is chiral and bulky, analytes cannot access the PSP interior, leading to increased association with the chiral center and improved enantioselectivity. Although the data support this conclusion for SDLL, SDLIL, and SDLV vesicles, the authors noted that contradictory results exist for BNP separations in the literature for other types of chiral systems. It was hypothesized that the nature of the chiral selector (vesicle vs. polymer micelles for example) is responsible for this discrepancy, although the overall recognition mechanism was believed to be similar. Lastly, the presence of two chiral centers in SDLIL had a negative impact on the resolution of BOH and BZN stereoisomers. 3.3 Microemulsions

Figure 3. Illustration of the bilayer nature of unilamellar vesicles as compared to a single-layered micelle [52]. Reproduced with permission.

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Microemulsions have been utilized in EKC since 1991. They are more complex than other surfactant aggregates, typically being comprised of three components – a surfactant, a cosurfactant, and an oil – that must be combined in a suitable ratio to achieve an optically transparent, thermowww.electrophoresis-journal.com

2508

K. A. Kahle and J. P. Foley

dynamically stable nanodroplet (,1–5 nm). The general structure and separation scheme for this PSP are given in Fig. 4. Microemulsions can sometimes be formulated without cosurfactant, but their use is infrequent since they are generally less stable. In comparison to the other surfactant PSPs, microemulsions are less rigid, provide higher efficiencies, can solubilize more hydrophobic compounds, have a more tunable elution range (migration window), and offer a wider variety of parameters to optimize. The field of chiral MEEKC (where the only chiral selector(s) is/are one or more components of the microemulsion) has been minimally researched with only 12 publications appearing in peer-reviewed journals (see Table 1). Three recent reviews on MEEKC have included chiral applications [55–57]. The first separation via chiral MEEKC was reported by Aiken and Huie et al. [58] in 1993 and utilized a chiral oil, (2R, 3R)-di-n-butyl tartrate, in conjunction with an achiral surfactant (SDS) and achiral cosurfactant (1-butanol) to achieve a selectivity of 2.6 for the enantiomers of ephedrine. It was also shown that the presence of 1-butanol (the cosurfactant) was essential – no enantioseparation could be achieved when it was omitted. The second chiral MEEKC separation was reported by Pascoe and Foley [59] in 2002 and utilized a microemulsion based on a chiral surfactant (dodecoxycarbonylvaline, DDCV), an achiral cosurfactant (1-butanol), and an achiral low-interfacial-tension oil (ethyl acetate). Nine pairs of pharmaceutical enantiomers were tested with this chiral surfactant-based PSP and results were compared to those from MEKC using the same chiral surfactant. One main difference between the techniques was the approximately 2.5-fold greater elution range observed for MEEKC. Enantioselectivities were slightly greater with MEEKC than MEKC while efficiencies were lower and resolution remained essentially unchanged. The underlying cause for the lower efficiency of the MEEKC system was later determined to be buffer related [60]. The main advantage to using a microemulsion over a micelle was shown to be the threefold reduction in analysis time, as exemplified by the simultaneous separation of ephedrine and methylpseudoephedrine enantiomers in less than 4 min.

Electrophoresis 2007, 28, 2503–2526

Two subsequent publications of chiral microemulsion EKC also used the chiral surfactant DDCV. One study examined the impact of oil identity and concentration [61]. Three low-interfacial-tension oils of methyl formate, methyl acetate, and methyl propionate were used in microemulsion formulations and were compared to the original ethyl acetate system using both equimolar (51 mM) and equal v/v% (0.5%) concentrations for the separation of 14 pharmaceutical compounds. The aggregates prepared with methyl formate were not stable, reverting back to the individual components after 1 week. Ethyl acetate performed better than the other oils (compared at constant v/v%) in terms of elution range, enantioselectivity, and number of pairs of baselineresolved enantiomers. Efficiency did not vary drastically with the changes in oil identity. The results showed that the identity of the achiral oil had an impact on the quality of enantioseparation, and that the ethyl acetate system was the best overall. The next published experiments focused on the combination of derivatized CDs (HP-b-CD, sulfated-CD (S-bCD), and highly sulfated b-CD (HS-b-CD)) with both DDCV and SDS microemulsions [62]. The efficiencies observed with the CD-DDCV microemulsion systems were evaluated with respect to three buffers: ACES, Tris, and phosphate. ACES exhibited excessive UV absorbance at the desired wavelengths and was excluded from further study; although acceptable baselines and efficiencies could be obtained for a few compounds using Tris, phosphate provided better results for all compounds. Next to appear was the first report on chiral polymeric MEEKC [63]. The microemulsion was prepared using a polymeric chiral surfactant (poly-D-SUV), 1-butanol as the cosurfactant, and n-heptane as the oil. Barbiturate, binaphthyl, and paveroline enantiomers were analyzed. Interestingly, chiral separations were obtained for BNP in the absence of oil but not in the absence of cosurfactant. It was suggested that hydrogen-bonding interactions between the analyte and PSP were enhanced by 1-butanol, allowing enantioresolution. This effect was analyte-dependent, with differing trends for other binaphthyl derivatives (BOH and BNA). Optimization of surfactant, cosurfactant, and oil con-

Figure 4. Schematic presentation of MEEKC [55]. A typical running buffer consists of 0.8% n-octane 3.3% SDS 6.6% 1-butanol in borate buffer, pH 9.2.

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.electrophoresis-journal.com

CE and CEC

Electrophoresis 2007, 28, 2503–2526

2509

Table 1. Chiral microemulsion formulations published to date

Microemulsion composition Surfactant

Cosurfactant

Oil

Buffer

Comments and analytes

Ref.

0.6% w/w SDS

1.2% w/w 1-Butanol

0.5% w/w (2R, 3R)-din-Butyl tartratea) 0.5% v/v Ethyl acetate

15 mM THAM pH 8.1

Ephedrine

[58]

1.0% w/v DDCVa) 1.2% v/v 1-Butanol (R- and S-) 1.0% w/v R-DDCVa)

1.2% v/v 1-Butanol

1.0% w/v DDCVa) 1.2% v/v 1-Butanol 0.6% w/v SDS

0.25–0.76% w/w Poly-D-SUVa)

1.00–7.00% w/w 1-Butanol

1, 2, 3, and 4% w/v R-DDCVa)

1.2% v/v 1-Butanol

3.5% w/w SDS

R-2-Pentanola) R-2-Hexanola) R-2-Heptanola) 1.2% v/v 1-Butanol

a) 2% w/v DDCVa) b) 4% w/v DDCVa)

1% w/v DDCVa) 4% w/v DDCVa)

1.2% v/v 1-Butanol

50 mM ACES pH 7.0

Atenolol, ephedrine, indapamide, methyl[59] pseudoephedrine, metoprolol, norphenylephrine, octopamine, pseudoephedrine, and synephrine [61] 50 mM ACES pH 7.0 Study on oil identity and concentration 0.5% v/v or 51 mM Pseudoephedrine, ephedrine, methylephedrine, Ethyl acetate, metoprolol, norphenylephrine, indapamide, methyl acetate, synephrine, atenolol, octopamine, epinephrine, methyl arterenol, propranolol, chlorpheniramine, and propionate, or verapamil methyl formate CD-modified MEEKC (S-b-CDa) and HP-b-CDa) [62] 0.5% v/v Ethyl acetate 50 mM ACES pH 7.0, 50 mM phosphate pH 7.0, at 1.0 and 1.5% w/v) 100 mM Tris pH 7.0, Pseudoephedrine, ephedrine, methylephedrine, 50 mM phosphate pH 8.0, metoprolol, indapamide, synephrine, octopamine, 100 mM phosphate pH 8.0 propranolol, and chlorpheniramine [63] 0.21–1.60% w/w 25 mM Phosphate pH 7.0 Chiral polymer formed under microemulsion n-Heptane conditions Secobarbital, pentobarbital, binaphthyl derivatives, and paveroline derivatives (laudanosoline and norlaudanosoline) [60] 0.5% v/v Ethyl 50 mM ACES pH 7.0 Study on surfactant concentration and buffer acetate 50 mM Phosphate pH 7.0 identity. 46phase (4.0% DDCV, 4.8% butanol, 2.0% ethyl acetate) also examined Epinephrine, arterenol, ephedrine, atenolol, methylephedrine, metoprolol, synephrine, octopamine, norphenylephrine, indapamide, and pseudoephedrine 0.8% w/w Octane 10 mM Borate pH 9.2 Cosurfactant concentration was varied [69] Norephedrine, ephedrine, nadolol, and propranolol 0.5% v/v Ethyl acetate

50 mM Phosphate pH 7.0

0.5% v/v Ethyl acetate

50 mM Phosphate pH 7.0

2% w/v DDCVa) 0.5% v/v Ethyl acetate 50 mM Phosphate pH 7.0 1.65% S-2-Hexanola) (R- and S-) 0.5% v/v Ethyl acetate 50 mM Phosphate pH 7.0 2% w/v R-DDCVa) 1.20% v/v 1-Butanol 1.42% v/v 1-Pentanol 1.42% v/v 2-Pentanol 1.65% v/v 1-Hexanol 1.65% v/v 2-Hexanol 1.19% v/v Cyclopentanol 1.39% v/v Cyclohexanol 2% w/v DDCVa) 1.65% v/v 2-Hexanol 1.23% v/v Dibutyl 50 mM Phosphate pH 7.0 (R- and S-) tartratea) (D- and L-)

Comparison of MEEKC, solvent-modified MEKC, and MEKC Epinephrine, arterenol, ephedrine, atenolol, methylephedrine, metoprolol, synephrine, octopamine, norphenylephrine, indapamide, and pseudoephedrine Temperature study Epinephrine, ephedrine, atenolol, methylephedrine, metoprolol, synephrine, norphenylephrine, indapamide, and pseudoephedrine Dual-chirality microemulsion N-Methyl ephedrine and pseudoephedrine Achiral/racemic cosurfactant study, equimolar concentrations (131 mM) Atenolol, ephedrine, N-methyl ephedrine, pseudoephedrine, metoprolol, and synephrine

[65]

Dual-chirality microemulsion Atenolol, ephedrine, N-methyl ephedrine, pseudoephedrine, metoprolol, and synephrine

[68]

[64]

[66] [67]

a) Denotes chiral selector.

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.electrophoresis-journal.com

2510

K. A. Kahle and J. P. Foley

centrations was performed for the various types of enantiomers. Increasing the concentration of cosurfactant led to varying effects on the chiral resolution for anionic and neutral vs. cationic enantiomers. Results were compared with MEKC and solvent-modified MEKC conditions. Larger efficiency and resolution were obtained with the MEEKC system for barbiturates. The chiral surfactant DDCV was employed in chiral MEEKC in six additional publications. One study focused on the impact of surfactant concentration and buffer identity [60]. As previously described, the identity of the buffer was found to have a drastic impact on the quality of separations. Subsequent to those initial observations, improvements in system stability were found when phosphate was used instead of ACES. In addition, with phosphate buffer a larger increase in efficiency was achieved as the percentage of DDCV was increased. Overall resolution increased as the surfactant concentration was increased as a result of more optimal retention factors. A system containing four times as much microemulsion (4% w/v DDCV, 4.8% v/v 1-butanol, and 2.0% v/v ethyl acetate) was also tested and resulted in a decreased elution range. Although analyte efficiency was better than that with 4% DDCV, reduced resolutions and noisy baselines were observed. The importance of surfactant concentration optimization in chiral MEEKC was demonstrated. DDCV-based microemulsions were next used in a temperature study [64]. Nine pairs of enantiomers were used to determine the effect of temperature on enantioselectivity and resolution as well as to conduct a van’t Hoff (thermodynamic) analysis. Enantioselectivity and resolution were confirmed to decrease at higher temperature, as predicted. van’t Hoff plots (ln Keq vs. 1/T) were linear for the 1% DDCV formulation but not the 4% version. It was hypothesized that temperature had more of an impact on the phase ratio and/ or microemulsion structure when a larger amount of surfactant was used. Negative enthalpies of transfer were observed for all compounds, representing a preference for the PSP over the buffer. Entropies of transfer were positive or negative, where positive values signified an increase in disorder for analyte partitioning into the PSP and negative values showing a decrease in disorder for penetration into the microemulsion. Examination of Gibb’s free energy changes with enthalpy and entropy revealed that chiral discrimination mechanisms rely more heavily on the entropic contribution. Finally, the existence of enthalpy/entropy compensation was demonstrated with a compensation temperature of 227 K (at which DG7 is similar for all enantiomers). Thermodynamic results were compared to DDCV micellar data and showed that a larger achiral attraction exists for the micelles. A comparison of chiral MEKC, solvent-modified chiral MEKC, and chiral MEEKC was conducted using the chiral surfactant DDCV with identical concentrations [65]. The baselines obtained using micelles and butanol-modified micelles at 4% w/v DDCV were unstable, in contrast to the © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Electrophoresis 2007, 28, 2503–2526

stable baselines observed with the equivalent microemulsions. Irrespective of surfactant concentration, the elution range decreased from micelles to solvent-modified micelles to microemulsion, due mainly to the PSP electrophoretic mobility (micelles had the smallest size and largest mobility). Retention factors of the chiral analytes were somewhat larger using micelles and similar between microemulsions and butanol-modified micelles. At 2% DDCV, the efficiency obtained with microemulsions and butanol-modified micelles was comparable and somewhat better than that obtained with the plain micelles, probably due to the formers’ larger aggregate fluidity and improved mass transfer from the presence of the alcohol modifier; at 4% DDCV the average efficiency doubled to nearly 70 000 using the microemulsion but remained the same for the micelles and butanol-modified micelles. Finally, a slightly better average resolution was obtained using micelles at 2% DDCV and using butanol-modified micelles at 4% DDCV, although the microemulsion finished a close second (out of three) at both concentrations. Moreover, in a comparison of all results achieved with a stable baseline, the optimal microemulsion provided the highest average efficiency and resolution with a slightly lower enantioselectivity. The next DDCV publication was the first to describe the simultaneous incorporation of two chiral selectors into one microemulsion; it was subsequently employed for the separation of N-methyl ephedrine and pseudoephedrine enantiomers [66]. Each chiral surfactant enantiomer (R- and S-DDCV) was combined with one enantiomer of a chiral cosurfactant (S-2-hexanol). Interestingly, the microemulsion prepared with S-DDCV and S-2-hexanol provided larger efficiency and resolution values for both analytes. Enantioselectivity for N-methyl ephedrine was unaffected by the chiral surfactant stereochemistry, but pseudoephedrine showed a higher value with R-DDCV. Enantioseparations of pseudoephedrine with DDCV/2-hexanol microemulsions are shown in Fig. 5 where RX, RS, and SS represent R-DDCV/racemic 2hexanol, R-DDCV/S-2-hexanol, and S-DDCV/S-2-hexanol microemulsions, respectively. Thermodynamic synergies (positive and negative) with dual-chirality microemulsions were identified and postulated to be a result of either (a) interaction of the chiral surfactant and chiral cosurfactant or (b) a three-way interaction between the chiral microemulsion components and the enantiomers. In addition, racemic 2-hexanol was tested as a cosurfactant and found to greatly improve efficiency over the previously used 1-butanol. Another DDCV study compared the effects of seven different achiral cosurfactants (primary, secondary, and cyclic alcohols at equimolar concentrations) on the enantioseparation of six pairs of pharmaceutical enantiomers [67]. The main trends identified were: higher enantioselectivity with cyclic and short chain primary alcohols (1-butanol and cyclopentanol), lower enantioselectivity with longer chain primary alcohols, largest overall efficiency with 1-hexanol, and largest overall resolution with 1-butanol. Changing the cosurfactant was also shown to modify the separation selectivity. Methylene www.electrophoresis-journal.com

CE and CEC

Electrophoresis 2007, 28, 2503–2526

2511

Figure 5. Effect of cosurfactant stereochemical configuration on the enantiomeric separations of pseudoephedrine [66]. (A) RX microemulsion, (B) RS microemulsion, and (C) SS microemulsion. Peak identification: 1S,2S = (1S,2S)-pseudoephedrine; 1R,2R = (1R,2R)-pseudoephedrine. Surfactant (DDCV) concentration, 2% w/v; cosurfactant (2-hexanol) concentration, 1.65% v/v; 0.5% v/v ethyl acetate; 50 mM phosphate buffer, pH 7.0; detection wavelength, 215 6 5 nm; capillary dimensions: Ltot = 32 cm, Leff = 23.6 cm, id = 50 mm; hydrodynamic injection = 25 mbar for 2 s; applied voltage, 11.5 kV. Reproduced with permission.

selectivity (aggregate hydrophobicity) decreased with an order of 1-hexanol .2-hexanol .1-pentanol .2-pentanol
cyclohexanol
1-butanol .cyclopentanol, demonstrating that the cosurfactant has a significant impact on analyte partitioning and microemulsion polarity. The most recent DDCV study described results obtained with the first combination of a chiral surfactant (DDCV) with a chiral oil (dibutyl tartrate) [68]. For ephedrine-based analytes, two-chiral-component microemulsions with the surfactant and oil in opposite stereochemical configurations provided higher enantioselectivities than the single-chiral-component microemulsion. Enantioselective synergies were determined using a thermodynamic model and showed that the second proposed explanation reported in the first dual-chirality publication ([66], see above) was the most probable. Efficiencies observed with microemulsions comprised of racemic dibutyl tartrate or dibutyl-D-tartrate were significantly higher than those obtained with dibutyl-L-tartrate. The dual-chirality system containing R-DDCV and dibutyl D-tartrate provided significantly better resolution than the remaining one- and twochiral-component microemulsions for the ephedrine-based compounds, but only slightly better or equivalent resolution for non-ephedrine compounds. The final chiral MEEKC paper in the literature utilized chiral cosurfactants (secondary alcohols) in conjunction with an achiral surfactant (SDS) and an achiral oil (n-octane) for the analysis of five pharmaceutical compounds [69]. Enantioresolution could not be achieved with R-2-butanol and © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

improved going from R-2-pentanol to R-2-hexanol. The use of R-2-heptanol allowed for the only separation of propranolol enantiomers, but degraded the resolution of other analytes. N-Methyl ephedrine could not be resolved with the systems tested, most likely due to its structure. Furthermore, use of the other cosurfactant enantiomer, S-2-hexanol, effectively reversed the elution order. Variations of the oil phase demonstrated that octane performed better than ethyl acetate for this microemulsion formulation; removal of the oil from this formulation resulted in a complete loss of enantioresolution. The research in chiral MEEKC has established that optimization of microemulsion conditions is complex with the concentration and identity of the aggregate components, as well as analyte identity, requiring consideration.

4

Complex-forming selectors

A wide variety of chiral selectors that do not form aggregate structures have been employed for the difficult task of enantioseparation. In general, these types of selectands form complexes with the analytes. A separation is obtained when the binding affinity for one enantiomer is higher than the other or when the complexes have different mobilities. The literature is dominated by CD usage, see ref. [70–72] for recent reviews. In many cases, CDs are combined with other PSPs to either produce or improve resolution. www.electrophoresis-journal.com

2512

K. A. Kahle and J. P. Foley

Electrophoresis 2007, 28, 2503–2526

4.1 Charged CDs

4.1.1.1 Carboxyalkyl-derivatized anionic CDs

Native CDs, which are neutral, are available in three “flavors” which differ in cavity size: a, b, and g, having 6, 7, and 8 glucopyranose units, respectively. Their UV transparency is one characteristic of CDs that have popularized their use. In order to enhance solubility and alter selectivity, many derivatized CD forms have been synthesized, creating a wide variety of neutral and charged options. The utilization of such charged CDs for pharmaceutical enantiomeric separations was reviewed by de Boer et al. [73] in 2000. A distinction in the literature has been made in which separations obtained using neutral and charged CDs are classified as CD-CZE and CD-EKC, respectively [74]. In the preparation of modified CDs, the nature and size of the derivatizing group have a dramatic impact on selectivity: large groups can make access to the cavity more difficult and charged groups add an opportunity for electrostatic interactions. An important aspect of charged CD use is that the BGE ionic strength increases with the addition of this chiral selector, thereby increasing Joule heating. In some instances, problems can arise from inclusion of BGE additives in the CD cavity, which decreases enantioselectivity. One approach to fine-tuning chiral selectivity has been the simultaneous use of neutral and charged CDs; Fillet et al. [75] reviewed this literature in 2000. The mechanism of chiral recognition with CDs has been studied using NMR, indicating that a host–guest complex is typically formed via interactions within the CD cavity and with the outer rim [29].

A novel anionic b-CD, 6-O-succinil-b-CD (CDsuc6), was synthesized by Cucinotta et al. [77] and was demonstrated to provide enantioresolution of four pairs of catecholamine enantiomers. Although this selector is not charged at high pH, it still provided enantioselectivity for one analyte under alkaline conditions, terbutaline. A method for pheniramine enantiomer analysis by CE was developed and validated with carboxyethyl-b-CD (CE-bCD) as the chiral selector [78]. The BGE composition was fine-tuned to reduce dispersion effects and increase resolution. The combination of CITP with CZE, using CE-b-CD, was shown to reduce interferences and concentrate the sample for analysis of enantiomers in urine [79]. The pH was adjusted to optimize CD ionization and enantioresolution. The method was validated and ng/mL LODs were obtained. A strategy for screening various CDs for chiral analysis of basic enantiomers was reported by Sokoließ and Köller [83]. A drug candidate was used to evaluate a variety of different neutral and anionic CDs; the best enantioresolution was achieved with carboxymethyl-b-CD (CM-b-CD). The authors also addressed capillary conditioning, validated the technique, and successfully tested drug formulations with excipients. A combination of anionic and neutral CDs (CM-b-CD and a-CD) was able to differentiate the enantiomers of 3-(4methylbenzylidene)-camphor, a sunscreen agent [85]. Enantioresolution was improved when a low temperature (157C) was used along with a high concentration of a-CD (120 mM). Changing the sample solvent to a mixture of methanol and water reduced peak tailing. Prior to implementation of the method for quantitation, the capillary conditioning steps were optimized and the procedure evaluated for linear range, LOD and quantitation, precision, accuracy, and selectivity. Quantitative analysis of 3-(4-methylbenzylidene)-camphor enantiomers in commercial products and in skin absorption was performed. The effect of temperature on enantioselectivity for some neutral and basic compounds was studied by Westall et al. [86]. Two neutral CDs and one anionic CD (HPb-CD and dimethyl-b-CD (DM-b-CD), and CM-b-CD, respectively) gave the expected trend of decreasing chiral selectivity with increasing temperature. HS-b-CD, on the other hand, showed an increase in enantioselectivity for some analytes with increasing temperature; the discussion focused on the chiral selector concentration as the underlying cause. Citalopram and its chiral metabolites were the target for chiral separation by Berzas-Nevado et al. [87]. A number of CDs were included in their preliminary experiments, with CM-g-CD chosen for further optimization. An entangled polymer solution of hydroxypropylmethylcellulose provided a means to improve selectivity. The final method was thoroughly validated (Plackett–Burman approach for robustness) and used for urine analysis. As part of Trapp’s unified kinetic theory for pseudo-first-order reactions, simulated and

4.1.1 Anionic CDs Negatively charged CDs have been produced using either carboxyalkyl or sulfur-containing groups. Derivatives containing a sulfated group are permanently charged, whereas the ionization of carboxyalkyl-CDs varies with pH. More recently, single isomer CDs have experienced increased interest due to reductions in batch-to-batch variability. Chiral separations that have been performed with anionic CDs during the specified review period are listed in Table 2. The bulk of the literature contains method optimization steps focusing on: buffer identity, buffer pH, buffer concentration, CD identity, CD concentration, applied voltage, and/or temperature. In many cases, a wide spectrum of selectors was screened with anionic CDs outperforming neutral CDs because of the former’s countermigration property and ability to electrostatically interact with positive enantiomers. One exception was the negatively charged enantiomers of etodolac, in which a series of eight neutral CDs and five anionic CDs were studied with only HP-b-CD and g-CD showing promise [76]; a method using HP-b-CD was subsequently optimized and validated. © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.electrophoresis-journal.com

CE and CEC

Electrophoresis 2007, 28, 2503–2526

2513

Table 2. Chiral EKC applications utilizing anionic CDsa)

CD(s)

Sample

Other CDs screened and/or comments

Ref.

CDsuc6

Catecholamines

Novel anionic CD

[77]

Carboxyethyl-CDs CE-b-CD CE-b-CD

Pheniramine Antihistaminic pharmaceuticals

Native CDs, S-b-CD, and IMA-b-CD also evaluated Coupled CITP-CE

[78] [79]

CM-b-CD

Basic pharmaceuticals

CM-b-CD CM-b-CD and a-CD CM-b-CD, HP-b-CD, DM-b-CD, and HS-b-CD CM-g-CD

Eight chiral pesticides 3-(4-Methylbenzylidene)-camphor Basic and neutral enantiomers, including three amino alcohols Citalopram and metabolites

b-CD and HP-b-CD also tested Dual-CD system HP-b-CD, S-b-CD, and DM-b-CD also examined Validated method b-CD, g-CD, M-b-CD, DM-b-CD, HP-b-CD, bile salts, and SDS were screened General chiral method development strategy with ten CDs (neutral and charged) CM-g-CD, S-b-CD, and succinylated-CD (Succ-b-CD) also tested Dual-CD system Temperature study

[80] [81]

CM-g-CD

Glucopyranosyl-based enantiomers Four isomers of a benzoaxathiepin derivative Citalopram

[87]

CM-b-CD

Thalidomide

[89]

Carboxymethyl-CDs CM-b-CD CM-b-CD and HP-g-CD

[82]

[83] [84] [85] [86]

CM-b-CD

Bevantolol enantiomers

Entangled polymer solution also used Several selectors employed in preliminary studies Unified equation development for reaction mechanism and rate information HPLC method also developed

HP-b-CD

Etodolac enantiomers

Eight neutral and five anionic CDs evaluated

[76]

Five pharmaceutical enantiomers

Different types of sulfated CDs used. Compared to HPLC method

[90]

a-CD, b-CD, and g-CD tested

[91] [92]

HS-g-CD also screened

[93]

Chiral recognition described

[94]

Equine plasma analysis LIF detection HS-a-CD and HS-g-CD also evaluated Analysis time reduction study HS-a-CD and HS-b-CD also provided resolution Validated method HS-a-CD and HS-g-CD also screened CE-MS

[95] [96]

Highly sulfated CDs HS-b-CD, S-b-CD, and HpS-b-CD HS-g-CD HS-g-CD HS-a-CD and/or HS-b-CD HS-a-CD, HS-b-CD, and HS-g-CD HS-b-CD HS-b-CD HS-g-CD HS-g-CD HS-b-CD HS-g-CD HS-g-CD HS-g-CD

HS-b-CD HS-b-CD HS-a-CD, HS-b-CD, and HS-g-CD HS-b-CD Sulfated CDs S-b-CD and a-CD

Moxifloxacin hydrochloride Butorphanol tartrate and cycloamine enantiomers g-Amino butyric acid, baclofen, phaclofen, saclofen, and hydroxysaclofen Benzothiazolinone and benzoxazolinone derivatives Ketamine and norketamine enantiomers Baclofen enantiomers Amphetamine and related compounds Tramadol enantiomers Phenylalanine enantiomers and synthesis by-products Amphetamine derivatives, methadone, and tramadol Methamphetamine derivatives, precursors, and impurities Tolterodine and two developmental basic pharmaceuticals Sumanirole Rivastigmine Propranolol, atenolol, chlorpheniramine, tryptophan Cis and trans nucleosides

[97] [98] [99] [100] [101]

g-CD, HP-g-CD, S-b-CD, HpS-b-CD, HS-a-CD, HS-b-CD, [102] heptakis(2,3-diacetyl-6-sulfato)-CD (HDAS-b-CD), and heptakis(2,3-dimethyl-6-sulfato)-CD (HDMS-b-CD) also screened Original CE method used neutral CD, but was replaced with HS-b-CD [103] [104] CD choice varied with analyte identity [105] HPLC method gave better performance HS-a-CD and HS-g-CD also evaluated

Substituted 2,3-dihydro[1,4]dioxinol[2,3Dual-CD system b]pyridine derivatives and intermediates

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

[88]

[106]

[107]

www.electrophoresis-journal.com

2514

Electrophoresis 2007, 28, 2503–2526

K. A. Kahle and J. P. Foley

Table 2. Continued

CD(s)

Sample

Other CDs screened and/or comments

Ref.

S-b-CD and HP-b-CD

S-b-CD

Cetirizine

S-b-CD

Tamsulosin hydrochloride

S-b-CD

Tamsulosin hydrochloride

Dual-CD system Methods with HP-b-CD and DM-b-CD also described Validated method a-CD, b-CD, g-CD, HDMS-b-CD, and DM-b-CD also examined a-CD, b-CD, HP-b-CD, CE-b-CD, 6I-deoxy-6I-monomethylamino-CD (1MA-b-CD) also screened Validated method b-CD, CM-b-CD, DM-b-CD, and HP-b-CD also tested Validated method a-CD, b-CD, g-CD, DM-b-CD, HP-b-CD, and CM-b-CD also tested 50 Analytes examined Single and dual-CD

[108]

S-b-CD

Ephedrine and pseudoephedrine enantiomers Aziridine

S-b-CD Basic pharmaceuticals S-b-CD and/or HP-b-CD, b-CD, Five phenothiazines or g-CD SI-S-b-CD and b-CD Hydrobenzoin and related enantiomers S-b-CD

Biperiden

S-b-CD and HP-b-CD

Hydroxychloroquine and metabolites

S-b-CD S-b-CD S-b-CD

Phenyloxirane and phenylethanediol Propafenone and metabolites b-Aminoketones

S-b-CD

Verapamil and norverapamil

S-b-CD S-b-CD and a-CD

Catecholamines and related enantiomers Etomidate and related enantiomers

S-b-CD

Epinephrine, isoproterenol, octopamine, and norephedrine Disopyramide

S-b-CD Sulfur-containing CDs DAS-b-CD

Methamphetamine and ephedrine derivatives Basic pharmaceuticals

Dual-CD system with single isomer anionic CD Randomly substituted S-b-CD also analyzed Validated method b-CD, g-CD, CM-b-CD, and sulfobutyl-b-CD also tested Dual-CD, validated method DM-b-CD, heptakis(2,3,6-tri-O-methyl)-CD (TM-b-CD), CM-b-CD, and maltodextrin also examined Focus on microbiological process Method development and application HPLC method developed as well a-CD, b-CD, g-CD, M-b-CD, DM-b-CD, TM-b-CD also tested Validated method b-CD, g-CD, HP-b-CD, M-b-CD, sulfobutylether-b-CD, CE-b-CD, HS-a-CD, HS-b-CD, and DOC micelles also screened Randomly and singly sulfated CDs were used Dual-CD system, Augmented Plackett-Burman, Validated method b-CD and HP-b-CD also tested Temperature study Electrochemiluminescence detection

[109] [110] [111] [112] [113] [114] [115] [116] [117]

[118] [119, 120] [121] [122]

[123] [124] [125] [126]

CE-MS Native and DM-b-CD, S-b-CD, and DMS-b-CD also tested HMAS-b-CD Single isomer CD Aqueous and nonaqueous BGEs HMS-b-CD Basic pharmaceuticals Single isomer CD Aqueous and nonaqueous BGEs Sulfonated-b-CD Arylalcohols Dual-chiral system with silver colloid Sulfato-b-CD, CM-g-CD, Flavanones and flavanone-7-O-glyclosides Several neutral and anionic CDs screened in addition to chiral methyl-b-CD, and HP-g-CD micelles HDAS-b-CD Cetirizine Validated method CM-b-CD, HDMS-b-CD, HP-b-CD, and TM-b-CD also evaluated HpS-b-CD Atomoxetine enantiomers, positional Chiral HPLC analysis was main focus isomers, and impurities Sulfobutyl ether-CD (SB-b-CD) Frovatriptan Validated method b-CD, DM-b-CD, TM-b-CD, and HP-b-CD also tested HpS-b-CD N-(3R)-1-Azabicyclo[2.2.2]oct-3-ylfuro[2,3- Native b-CD, HP-b-CD, DM-b-CD, TM-b-CD, S-b-CD, HDMS-b-CD, c]pyridine-5-carboxamide HDAS-b-CD, and HS-CDs were also screened

[127, 128] [129] [130] [131] [46] [132] [133] [134] [135]

a) Review period from January 1, 2005 to December 1, 2006.

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.electrophoresis-journal.com

Electrophoresis 2007, 28, 2503–2526

experimental reaction profiles were employed to study the enantiomerization barrier for thalidomide [88]. Chiral separations were carried out with EKC using CM-b-CD as the selector, baseline resolution was not required because the plateau between the peaks contained data of interest. The proposed models provided an efficient means to gather reaction mechanism and rate information. Long et al. [89] devised chiral separation methods for both EKC and HPLC for bevantolol enantiomers. Comparison of the two liquid separation techniques showed higher efficiency and resolution with EKC but at the expense of longer retention times. Additionally, accuracy, precision, and the LOD were better with HPLC (S-enantiomer LODs of 0.1 and 0.05% for EKC and HPLC, respectively). 4.1.1.2 Sulfur-derivatized, anionic CDs Rocheleau [90] compared S-b-CDs with different degrees of substitution for five pharmaceutical compounds. Enantioseparations were better when the selectand was more substituted. A validated method for N-benzoyl methyl piperazine was presented along with a basic strategy for the development of chiral separations. Analysis of the same enantiomers via chiral HPLC was more precise and sensitive and provided lower LOD and LOQ values. A validated method for the resolution of the four stereoisomers of moxifloxacin hydrochloride with highly sulfated g-CD (HS-g-CD) was published by Cruz and Hall [91]. In order to maintain adequate separation, the BGE and instrument parameters needed to be maintained within 1% of the specified value. The LOQ was 0.05% for the active ingredient and final product. In addition, analyses of placebos and stability samples were performed. Enantiomers of butorphanol tartrate and cycloamine were successfully discriminated with HS-g-CD [92]. Neutral CDs were first evaluated with g-CD providing sufficient resolution but the elution order was such that the main peak migrated before the minor one. The highly sulfated version of g-CD gave the desired migration order and resolution at a very low concentration (0.02% w/v). This method was shown to be effective for commercial product analysis and met regulatory requirements. Highly sulfated CDs (a and b) were used to enantioresolve baclofen analogs [93]. The effect of selector concentration was examined and a reversal in migration order was observed when the CD cavity size was changed. This report also focused on pKa calculations from CE data. All four pairs of enantiomers could be baseline resolved with either HS-a-CD and/or HS-b-CD but not with HS-g-CD. Migration orders were correlated to analyte affinity for the selector. In a different study, the chiral recognition mechanism between a negatively charged CD (HS-CDs) and cationic enantiomers (benzothiazolinone and benzoxazolinone derivatives) was investigated [94]. The fact that analyte structure has a major impact on complex formation was confirmed. It was demonstrated that the phenyl ring in the analytes is included in the CD cavity as part of the enantioresolution © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

CE and CEC

2515

process. Binding constants with HS-CDs were approximately 60 times greater than those with neutral CDs, signifying the importance of electrostatic interactions. Rudaz and co-workers [97] employed HS-g-CD and was able to speed up the separation of amphetamine and related chiral analytes. In addition to UV detection, a method for mass spectrometric detection was developed. Three approaches were undertaken to reduce retention times: (i) high selector concentration (2.5% CD) with suppressed EOF (reverse polarity), (ii) high selector concentration (2.5% CD) with high EOF (normal polarity), and (iii) low selector concentration (0.2% CD) with high EOF (normal polarity). For (i), short-end injection and constant pressure were separately tried to further reduce analysis times, but the resolution decreased to unacceptable levels. A pressure gradient was found to speed up the analysis from ,23 to 12 min while maintaining baseline resolution. Attempts to employ an additional high concentration of an anionic CD with normal polarity were not fruitful. The use of low chiral selector concentrations with the partial filling technique resulted in 11-min separations. Finally, the method with 0.2% CD was adapted for CE-MS with a 6-min analysis time obtained. Another example of amphetamine analysis (enantiomers, precursors, impurities) using HS-g-CD was reported by Iwata et al. [101]. Electrokinetic injection provided better peak height ratios for the impurities (higher S/N ratio) than hydrodynamic. Two factors contributed to the enhanced peak heights with electrokinetic injection: (a) the analytes were positively charged thus electrokinetic injection at 15 kV would be biased for the sampling of cations and (b) the main component of the sample mixture, methamphetamine, was more highly complexed with the chiral selector than the trace impurities and as such under the specified injection conditions was not sampled to the same extent as the impurities. A different cavity size of HS-CD, HS-b-CD, was required to enantioseparate sumanirole [103]. This particular CE method had been under development/modification for several years where a neutral CD had originally been employed. However, a later trial with HS-b-CD led to better separation, and the validated method gave a lower LOD than chiral HPLC. The employment of HS-b-CD as the chiral resolving agent for rivastigmine enantiomers was reported by Wang et al. [104]. During method development, an increase in buffer pH was shown to improve detection sensitivity and the addition of a capillary coating step increased efficiency. The optimized method was validated. All three highly sulfated versions of a, b, and g CD were used by Matthijs and Vander Heyden [105] to discriminate the enantiomers of four basic pharmaceuticals. The methods were optimized using factorial design with the goal of quantitating impurities at the 0.1% level. Resolution was shown to vary depending on migration order of the major and minor components; baseline resolution (Rs = 1.5) was not always sufficient for impurity determination when peaks were asymmetric. Equations for resolution were compared and the method was validated using internal standards. www.electrophoresis-journal.com

2516

K. A. Kahle and J. P. Foley

The combination of a single isomer anionic CD (heptakis-(2,3-dihydroxy-6-O-sulfo)-b-CD (SI-S-b-CD)) with neutral b-CD proved to be beneficial for the enantioseparation of hydrobenzoin and related compounds [115]. Hydrobenzoin was found to bind more strongly with the single isomer CD than with its randomly substituted counterpart. The migration order for the three pairs of enantiomers could be altered by varying the concentration of neutral b-CD while maintaining a low, fixed anionic CD concentration in borate buffer. Binding constants were calculated to better understand the complex interactions between the analyte, dual-CD system, and borate ions. The data showed that the interaction between hydrobenzoin and borate contributes significantly to the chiral recognition process. A validated method using HS-b-CD for the determination of propafenone enantiomers and their metabolites was reported by Afshar and Thormann [119]. Three analytical figures of merit were tested during validation: migration time reproducibility, range of linearity, and LOQ. FDA requirements were met and the procedure was applied to human serum and in vitro metabolism samples. The authors published a second paper based on this chiral separation method [120] but focusing on the metabolism of propafenone by human cytochrome P-450 and the reaction kinetics. The anionic CD, S-b-CD, was utilized for b-aminoketone analysis [121]; additionally, a dual-neutral-CD system was also developed for the resolution of these enantiomers. Comparison among these two different CD-based methods and HPLC showed that CE provided better selectivity and S-b-CD was less robust than the dual-CD system but offered higher resolution and faster analysis time. Randomly and singly sulfated b-CDs were employed by Lin et al. [123] under four sets of BGE conditions based on CD type, CD concentration, buffer concentration, buffer pH, and CE polarity. As shown in Fig. 6, mode (I) was performed under normal polarity with a low concentration of randomly substituted S-b-CD at a low pH; mode (II) was implemented with reversed polarity, moderately high concentration of randomly substituted S-b-CD, and a low pH; mode (III) utilized normal polarity, a moderately high concentration of randomly substituted S-b-CD, and neutral pH; and mode (IV) conditions consisted of normal polarity, low to moderate concentrations of singly substituted S-b-CD, and a low pH. For each mode, the polarity was chosen such that the direction of analyte mobility allowed detection. The target chiral analytes in this report were cationic (catecholamines and similar compounds); differences between the separation schemes and chiral recognition mechanisms were discussed. A temperature study was conducted for the chiral separation of four basic analytes (epinephrine, isoproterenol, octopamine, and norephedrine) with randomly substituted S-b-CD [125]. For three of the test analytes, the authors observed an unexpected increase in enantioselectivity with increasing temperature using low selector concentrations with reversed polarity; this was explained to be a difference in the impact of temperature on the individual enantiomers. Further, singly substituted S-b-CD did not give the same phenomenon. A © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Electrophoresis 2007, 28, 2503–2526

Figure 6. Schematic of different separation schemes available with anionic CDs [123]. (I) low MI-S-b-CD concentration with a high-concentration BGE; at low pH in the normal polarity mode; (II) high MI-S-b-CD concentration at low pH in the reversed polarity mode; (III) high MI-S-b-CD concentration at neutral pH in the normal polarity mode; (IV) low to moderately high SI-S-b-CD concentration with a high-concentration BGE at low pH in the normal polarity mode. MI and SI refer to random and single substitution, respectively. Reproduced with permission.

unique approach to chiral CE was taken by Fang et al. [126] for the discrimination of disopyramide enantiomers (S-bCD) by using electrochemiluminescence detection with the visualization agent tris-(2,20 -bipyridyl)-ruthenium. In another departure from the generally utilized set-up, the solutions in the inlet and outlet buffer vials were optimized with their final conditions varying in pH and ionic strength (chiral selector was present only in the inlet vial). The described procedure was able to separate the target analyte in contrast to using identical solutions in the inlet and outlet vials. The method was demonstrated for the analysis of plasma samples. The first single isomer sulfated CD, heptakis(2-O-methyl-3-O-acetyl-6-O-sulfo)cyclomaltoheptaose www.electrophoresis-journal.com

CE and CEC

Electrophoresis 2007, 28, 2503–2526

(HMAS-b-CD) was synthesized by Busby and Vigh [129]. It is substituted with non-identical groups on carbons 2, 3, and 6. The new chiral selector was applied to separate 24 weakly basic pharmaceutical compounds under both aqueous and nonaqueous conditions. The majority of enantiomers could be baseline resolved with selector concentration in the range of 2.5–30 mM. When the CD concentration increased, the EOF also increased, demonstrating some CD adsorption to the capillary wall. The authors explored the theoretical selectivity using the charged resolving agent migration model (CHARM) and found two discontinuous regions. For some analytes, the nonaqueous buffer produced sharper peaks (less electromigration dispersion). The same group prepared another single isomer anionic CD: heptakis(2-O-methyl-6-Osulfo)cyclomaltoheptaose (HMS-b-CD) [130]. As with their previous report, basic pharmaceutical enantiomers were used to investigate the selectivity of the new CD in both aqueous and nonaqueous acidic buffers. Although the same range of selector concentrations was used, HMS-b-CD solutions lowered the EOF at higher concentrations, in contrast to HMAS-b-CD. Most of the chiral analytes were baseline resolved in aqueous media (19 out of 23) but not in nonaqueous buffers (four out of 23). The mobility and binding characteristics for the analytes with both HMS-b-CD and its precursor were described and compared. Choi et al. [131] took a slightly different approach to chiral CE by incorporating a silver colloid into a buffer system containing sulfonated-b-CD. The addition of colloidal particles greatly enhanced enantioselectivity but increased

2517

analysis times. The resulting method was capable of individually separating three arylalcohols. Cyclic voltammetry was used to further explore the chiral recognition mechanism. Ramstad and Johnson [135] developed a CE method for the enantiomeric separation of a basic drug candidate. Following the initial screening of several CDs, a method employing heptakis-6-sulfato-CD (HpS-b-CD) was further optimized and validated: an internal standard was added, the capillary ends were rinsed with water to remove sample, and inlet/outlet vials were used for only eight injections. 4.1.2 Cationic CDs In distinct contrast to the plethora of CD-EKC publications with anionic selectors, cationic CDs have yet to be as frequently utilized as shown by the number of publications during the review period, see Table 3. Typically, some sort of amine functionality is attached to the CD during derivatization to impart the positive charge. During the review period, single isomer cationic selectands were the most widely employed. An important caveat for cationic CDs is that like all positively-charged species they can electrostatically interact with (adsorb onto) the negatively charged walls of silica capillaries, thereby reducing or even reversing the EOF; such CD–wall interactions should be carefully monitored and controlled. Researchers at Nanyang Technological University have contributed several papers to the field of cationic CD synthesis and application [136–143]. A very brief study described

Table 3. Chiral EKC applications utilizing cationic CDsa)

CD(s)

Sample

Comments

Ref.

Imidazolium, pyridinium, and quaternary ammonium substituted b-CDs BuAM-b-CD

Aromatic carboxylic acids

Novel cationic CDs

[136]

Hydroxy acids, carboxylic acids, and amphoteric compounds Dansyl amino acids

Novel single isomer cationic CD

[137]

Novel single isomer cationic CD

[138]

Eight single isomer cationic CDs Novel single isomer cationic CD Novel single isomer cationic CD and a modified synthesis of CD precursor Application of existing CD

[139] [140] [141]

Novel single isomer cationic CD

[143]

Chiral recognition study utilizing NMR Synthesis of highly charged CD. Results compared to QA-b-CD Low degree of substitution (DS)

[144] [145]

Chiral analysis with TM-b-CD, vancomycin, and nano-LC also demonstrated

[147]

Mono-6-N-allylammonium-6-deoxy-b-CD chloride (ALAM-CD) 6-Mono(alkylimidazolium)-b-CDs PeAM-b-CD b-CD-NH3Cl b-CD-NH2 Mono-6^-propylammonium-6^-deoxy-CD (PrAM-b-CD) QA-b-CD HPTMA-b-CD 6-O-(Hydroxypropyltrimethylammonia)-CD (GTA-b-CD) 6-Monodeoxy-6-monoamino-CD (6-MA-b-CD)

Dansyl amino acids Anionic and ampholytic acids Anionic and ampholytic acids Hydroxy acids, carboxylic acids, and amino acids Hydroxy acids, carboxylic acids, and amphoteric compounds Two dianionic enantiomers Arylpropionic acids, racemic mandelic acid, and 3-phenyllactic acid Acidic pharmaceuticals Clofibric acid derivatives

[142]

[146]

a) Review period of January 1, 2005 to December 1, 2006.

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.electrophoresis-journal.com

2518

Electrophoresis 2007, 28, 2503–2526

K. A. Kahle and J. P. Foley

the synthesis of three different cationic b-CDs: imidazolium, pyridinium, and quaternary ammonium substituted b-CDs [136]. BuAM-b-CD (mono-6^-butylammonium-6^-deoxy-b-CD tosylate), a single isomer CD, was created and shown to be an effective chiral selector [137] for three types of chiral analytes: hydroxy acids, carboxylic acids, and ampholytic acids with the latter two classes, in general, shown to weakly bind with the positive CD. Resolution was studied as a function of CD concentration and implications for the chiral recognition mechanism were discussed in relation to the analyte types. The optimal selector concentration range was higher for hydroxy acids (.20 mM) than carboxylic acids (2–5 mM). Eight different versions of alkylimidazolium derivatized b-CD tosylates/chlorides were synthesized by Ong et al. [139]. The tosylate groups decreased the UV transparency of the chiral agent and as such did not display as desirable characteristics as the chloride analogs. Longer alkyl chains, i.e., decyl, provided lower enantioresolutions due to steric hindrances and the alkyl substituent of butyl gave the most promising separations. The novel single isomer CD mono-6^-N-pentylammonium-6^-deoxy-b-CD chloride (PeAM-b-CD) was prepared and evaluated as a chiral selector for anionic and ampholytic acids [140]. The optimal selector concentration varied based on analyte type and identity; in general, carboxylic acids required low concentrations of CD due to their high binding constants. The amino acids did not interact effectively enough with the chiral agent to produce well-resolved peaks. Also in 2005, a new synthesis scheme for mono(6-amino-6-deoxy)-b-CD (bCD-NH2) was introduced [141]. Prior synthesis schemes required long reaction times, costly reagents, and/or high pressures whereas the new reaction reduced the reaction time and eliminated the expensive catalyst and harsh experimental conditions. Ultimately, the single isomer cationic CD bmono-6-ammonium-6-deoxy-CD chloride (CD-NH3Cl) was produced and employed for the analysis of several anionic and ampholytic acids. The excellent chiral discrimination ability of this novel selector was demonstrated with the simultaneous separation of eight pairs of enantiomers (anionic and ampholytic) in under 35 min, see Fig. 7, where the optimized conditions included a buffer pH of 6.0 and selector concentration of 20 mM.

As a follow-up to an EKC experiment, the chiral discrimination process for two pairs of dianionic analytes with quaternary ammonium-CD (QA-b-CD) (quaternary ammonium-b-CD) was investigated via NMR by Liu et al. [144]. The analyte-to-selectand ratio was confirmed to be 1:1. The chemical shifts from NMR spectra were described in detail and used to explain the chiral recognition mechanism. The aromatic ring of the enantiomer was found to enter the CD cavity and the negative functional group of the enantiomers (sulfate) interacted with the positive CD functional group. The positively charged selectand, 6-O-(2-hydroxy-3trimethylammoniopropyl)-CD (HPTMA-b-CD), was prepared by Lin et al. [145]. The ability of this CD to separate arylpropionic acids, racemic mandelic acid, and 3-phenyllactic acid was evaluated and the results were compared to those obtained using QA-b-CD. Adjustment of CD concentration displayed the expected optimum concentration behavior described by Wren et al. [30, 31]. For all 11 chiral analytes, HPTMA-b-CD provided resolution whereas QA-b-CD could not. The underlying cause was deemed to be the steric hindrance of QA-b-CD at the rim with partial blocking of the cavity likely. Another difference between these selectors was their adsorptive tendency and resulting impact on EOF and polarity mode: HPTMA-b-CD adsorbed somewhat to the capillary wall and reduced but did not reverse the EOF, allowing the normal polarity to be maintained, in contrast to QA-b-CD which adsorbed to a greater degree thereby reversing the EOF and thus requiring a reversed polarity. 4.1.3 CD-modified EKC In order to improve enantioresolution, CDs have frequently been employed in conjunction with another chiral selector. As previously described, the identity and concentration ratio of the chiral agents must be examined in order to improve rather than degrade enantioselectivity. Several reports utilized the combination of a neutral CD with another chiral selector (Table 4). Several others have employed either a charged or neutral CD with an achiral PSP such as SDS micelles; the latter are beyond the scope of this review. 4.2 Crown ethers

Figure 7. Baseline separation of a mixture of eight acids by the use of b-CD-NH3Cl [141]. Reproduced with permission.

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Crown ethers are another class of chiral selectors that form inclusion complexes with chiral analytes (see [148] for a recent review). In particular, they are capable of resolving enantiomers containing a primary amino group. The mechanism of recognition is hydrogen bonding between oxygens on the planar chiral agent and the hydrogens of the amine group of the analyte [148]. The planar geometry of the selectand allows the formation of two diastereomeric complexes based on interactions with the faces of the crown ether that have different binding constants [149]. The structure of the main chiral crown ether employed in chiral EKC analyses, (1)-(18-crown-6)-2,3,11,12-tetracarboxylic acid (18C6H4), is shown in Fig. 8. www.electrophoresis-journal.com

CE and CEC

Electrophoresis 2007, 28, 2503–2526

2519

Table 4. Chiral EKC applications using neutral CDs in combination with other chiral selectors

Charged chiral selector

Neutral CD(s)

Analyte(s)

Comments

Ref.

DDCV microemulsion

HP-b-CD

S-b-CD also used

[62]

Chiral crown ether, 18C6H4 DOC micelles Chiral crown ether, 18C6H4 Sodium cholate micelles

HP-b-CD b-CD DM-b-CD Native, methyl, and HP-CDs b-CD

Pseudoephedrine, ephedrine, methylephedrine, metoprolol, indapamide, synephrine, octopamine, propranolol, and chlorpheniramine Amino acids and derivatives Serine enantiomers 1-Phenylethylamine and 1-cyclohexylethylamine Flavanones

Sodium taurodeoxycholate micelles Chiral crown ether, 18C6H4

DM-b-CD

Derivatized amino acids Small amines

Figure 8. The chemical structure of 18C6H4 [151]. Reproduced with permission.

Amino acids and some amino acid derivatives were enantioseparated by Salami et al. [149] using 18C6H4. They found that larger groups at the analyte chiral center allowed for chiral discrimination (i.e., benzyl), whereas smaller functionalities precluded resolution (i.e., methyl). Separations were improved via small increases in selector concentration, pH (increased ionization), modifier (ACN), and additive (tetrabutyl ammonium bromide). To further enhance enantioselectivity, a secondary chiral agent (HP-bCD) was added. The same chiral crown ether was employed by Gong et al. [150] for enantioresolution of underivatized amino acids. Contactless conductivity detection was used in this study, thereby allowing facile analysis of non-UV absorbing analytes. Kuwahara et al. [151] also investigated the potential of 18C6H4 for amino acid separations. In order to detect some of the analytes, indirect photometric detection was used. The ions evaluated as visualization agents (for UV detection) were chrysoidine, quinine, and 2-aminobenzimidazole, with the latter two eliminated from consideration after excessive baseline instability was observed. Investigation of chrysoidine concentration (which can interact with the chiral crown ether because of its primary amine functionalities) showed analyte enantioselectivity was achievable within the range of 1–5 mM (crown ether concentration of 4 mM). Nine out of 14 pairs of amino acid enantiomers could be resolved with the developed method. The combination of © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Neural samples Conductivity detection SDS micelles with neutral CDs also reported Resolution improvement due mainly to EOF effects Contactless conductivity detection

[149] [177] [178] [46] [179] [152]

a chiral crown ether (18C6H4) with a neutral CD (DM-b-CD) was reported for the resolution of small amines [152]; contactless conductivity detection was utilized due to the amines’ lack of a UV chromophore. Measurement of enantiomeric excess and separation on a chip were demonstrated for a representative pair of enantiomers. A novel crown ether, (S,S)-1,7-bis(4-phenyl-5-hydroxy-2oxo-3-zapentyl)-1,7-diaza-12-crown-4, was synthesized by Wang et al. [153]. The derivatized enantiomers of carnitine were successfully differentiated when sulfonated-b-CD was combined with the crown ether. 4.3 Glycopeptides Glycopeptides employed medicinally as antibiotics were first utilized by Armstrong as a chiral stationary phase substituent for HPLC. The complex nature of these selectors has precluded the exact determination of the chiral recognition mechanism. Specifically, the large number of chiral centers (over 18 for vancomycin) makes it difficult to pinpoint the location of chiral interactions. Another complicating factor is the presence of several ionizable groups. Therefore, pH has a major influence on enantiorecognition. The three most common examples of antibiotic chiral agents are vancomycin, ristocetin, and teicoplanin. In chiral EKC, partial filling of the capillary is typically performed with these selectors to avoid interference with detection. The use of macrocyclic glycopeptides as chiral selectors has been reviewed by Ward and Farris [154]. Bauvais et al. [155] performed a theoretical study on the mechanism of vancomycin. Two methods were used to probe the conformations and enantiomer interactions: molecular modeling and quantitative structure–property relationships. The researchers were able to accurately predict elution order and retention times for 17 pairs of enantiomers. Interpretation of simulated data led to the following conclusions: the stereochemistry of an enantiomer impacts its ability to fit in the vancomycin (selector) cavity, enantiomers interact with www.electrophoresis-journal.com

2520

Electrophoresis 2007, 28, 2503–2526

K. A. Kahle and J. P. Foley

different regions of the selector, and chiral separations are achieved by a docking process. A variant of vancomycin, balhimycin, was used to study the effect of chlorine groups on enantioresolution [156]. Chiral separations of dansyl amino acids and 2-arylpropionic acids were also carried out with dechlorobalhimycin, a non-chlorinated analog of balhimycin. The structures of these selectors are shown in Fig. 9. The CE setup included a coated capillary to reduce antibiotic adsorption, partial filling with the selector, and analysis under co-EOF conditions. The non-chlorinated version of the chiral agent consistently provided lower enantioresolutions and mobility differences as demonstrated in Fig. 10 for dansyl-methionine and ketoprofen, although the mechanistic rationale was undetermined. Enantiomer binding was determined to occur in the aglycon cavity of the antibiotic dimer and favored the R-enantiomer. The dimerization constant decreased when the two chlorines of balhimycin were replaced with hydrogen. Vancomycin was employed under partial filling conditions by Fantacuzzi et al. [147] for the enantioresolution of clofibric acid derivatives. A slightly different setup from the CD analyses (Section 4.1.2) was reported where the capillary was coated with polyacrylamide for use with this selector. The enantioseparation was not greatly impacted by a change in pH in the range of 4–6. Another application of vancomycin was reported for the separation of dimethyl diphenyl bicarboxylate derivative atropisomers [157]. Three analytes containing carboxylic groups were baseline separated using optimized conditions. As in previously referenced studies, the capillary was coated (with hexadimethrine bromide in this case), resulting in reversed EOF, decreased analysis times, and higher efficiencies. In contrast to other publications, the partial filling technique was not implemented to avoid detection interferences; instead, the detection wavelength was adjusted.

Figure 10. Enantioresolution of enantio-enriched L-dansylmethionine (A) and S-(1)-ketoprofen (B) by CE [156]. Conditions: running buffer, 50 mM Tris-phosphate buffer solution (pH 6.0) containing 0.001% w/v HDB; 2.0 mM balhimycin (bromobalhimycin or dechlorobalhimycin) dissolved in running buffer was partially filled with a pressure of 70 mbar for 40 s; fused-silica capillary, 50 mm id (370 mm od)647 cm (37 cm to detection window); injection by a pressure of 20 mbar for 4 s; column was thermostated at 257C; applied voltage, 215 kV; detection at l = 214 nm. Reproduced with permission.

4.4 Ligand exchange

Figure 9. Molecular structures of the glycopeptide antibiotics balhimycin (I), bromobalhimycin (II), and dechlorobalhimycin (III) [156]. Reproduced with permission.

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

The basic principle of ligand exchange chiral separations is the formation of a complex with a transition metal, typically copper(II), a chiral selector, and the chiral analytes. Only enantiomers containing two polar groups can be resolved using ligand exchange motifs. This technique was recently reviewed by Schmid and Gübitz [158]. A simple confirmation of the ligand exchange mechanism is the observation of a complete loss of resolution when the transition metal ion is removed [159, 160]. www.electrophoresis-journal.com

CE and CEC

Electrophoresis 2007, 28, 2503–2526

The enantioseparation of halogenated amino acids was achieved via complexation with copper(II) and amino acids/ amino acid derivatives (L-4-hydroxyproline, L-histidine, N-(2hydroxypropyl)-L-4-hydroxyproline and N-(2-hydroxyoctyl)L-4-hydroxyproline) [161]. The derivatized versions of L-4hydroxyproline provided much better resolution, at lower concentrations, than either native L-4-hydroxyproline or L-histidine. The best separations were obtained with N-(2hydroxyoctyl)-L-4-hydroxyproline, pH 4.5, copper to chiral agent ratio of 1:2, and no organic modifier. Another report utilized copper (II) but with a novel chiral ligand, L-ornithine [162]. Several pairs of enantiomers were tested including: amino acids, dansyl amino acids, amino alcohols, and a dipeptide. Low efficiencies were obtained when an uncoated capillary was employed. Once the capillary was coated, the efficiencies improved but resolution decreased. These findings imply that ligand exchange associations are faster in free solution, i.e., when the ligand is not absorbed on the capillary wall. Additionally, LE-MEKC with SDS micelles was explored, but resulted in longer migration times and decreased selectivity. A slightly different ligand exchange electrophoretic setup was reported by Nakajima et al. [163] for amino acid analysis. In addition to the conventional capillary format, a microchip was employed with LIF detection of fluorescently labeled analytes. The chiral complexing system contained copper(II) and L-prolinamide. Under typical CE conditions, five pairs of derivatized amino acid enantiomers were resolved. The same analytes were tested in ligand exchange microchip electrophoresis. Comparison of the data showed better resolution and selectivity for the microchip with shorter analysis times. The elution order for some analytes was found to differ between the two techniques; the underlying cause was still under investigation. Pantothenic acid enantiomers were effectively differentiated via ligand exchange-CE where the chiral selector system was composed of borate and (S)-3-amino-1,2-propanediol [164]. NMR analysis revealed the existence of several cyclic borate complexes. Thermodynamic calculations showed a large entropic difference between the diastereomeric complexes. The structures of complexes were optimized using molecular modeling and were discussed with respect to the migration order. The D-pantothenic acid complex eluted later, which agrees with the proposed structure (more open, larger frictional drag). 4.4.1 Ligand exchange – CD-EKC and ligand exchange – MEKC A CD derivative containing a diamino group (3A,3B-diamine3A,3B-dideoxy-2A(S),2B(S),3A(R),3B(R)-b-CD, (AB3NH2)) was utilized with copper(II) to further probe the mechanism of chiral interactions [165]. NMR spectroscopy provided information on the CD structure. Amino acids of alanine and tryptophan served as analytes to determine their interaction with the copper–CD complex and investigate binding. Elec© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

2521

trophoretic separations of tryptophan, tyrosine, and phenylalanine with the ligand exchange selector gave baseline resolution for tyrosine only. The neutral, derivatized CD CDen, 6^-(2-aminoethylamino)-6^-deoxy-b-CD, was synthesized and used in conjunction with copper(II) for the separation of aromatic amino acids [166]. The chiral recognition was classified as a combination of chelation and inclusion complexation. Five derivatized b-CDs containing an amine functional group under ligand exchange conditions (copper(II)) were compared for chiral separations of phenylalanine, tryptophan, and tyrosine [159]. Comparison of the results using different chiral selector systems gave varying trends for resolution depending on CD identity, selector concentration, and analyte. Additionally, the stability of the analyte/copper/CD complexes affects the elution order. Figure 11 shows the complexes of 6-deoxy-6-[4(2-aminoethyl)imidazolyl]-b-CD (CDmh) with L-tryptophan (a) and D-tryptophan (b) and 6-deoxy-6-[2-(4-imidazolyl)ethylamino]-b-CD (CDhm) with D-tryptophan (c) and L-tryptophan (d), where (a) and (c) are more stable due to the interaction of the indole group of the analyte with the CD cavity. Thus, enantiomer elution order reversal for tryptophan can be accomplished by changing CD identity. Ligand exchange MEKC was performed by Kodama et al. [160] for the enantioseparation of three chiral diols. The chiral selector complex contained borate and (1S, 2S, 3R, 5S)pinanediol in a 1:1 ratio (other ratios examined), 200 mM each. However, the chiral complex alone did not elicit chiral discrimination, therefore an additional PSP, SDS micelles, was included. The partitioning of the diastereomeric complexes into the surfactant PSP introduces a second separation mechanism that influences the ligand exchange process and aids in enantiomer resolution. Enantiomer elution order was reversed by using the opposite enantiomer of pinanediol. The mechanism of chiral interaction appears to include (i) complex formation between the borate and diol; and (ii) hydrophobic interactions between the chiral ligand and analyte. Similar chiral ligands differing in charge were tested but did not separate the three pairs of diol enantiomers. 4.5 Miscellaneous The role of steric hindrance was investigated using three ureaderivatized tergurides (synthetic ergot alkaloid-based chiral selectors) for the chiral analysis of dansyl amino acids, pesticides, and mandelic acid [167]. The synthesized dimethyl, diethyl, and diisopropyl terguride analogs demonstrated low solubility and required THF for dissolution. The capillary was coated and filled with chiral selector solution; the inlet and outlet vials did not contain the chiral agent due to its UV absorbance thereby creating a zone-sharpening effect. Increased chain length for the urea group on the terguride resulted in decreased enantiorecognition for dansylated amino acids. As shown in Fig. 12, the separation of dansyl aspartic acid enantiomers degraded as the steric hindrance of the selector increased (a dimethyl derivative .a diethyl derivative .a diwww.electrophoresis-journal.com

2522

Electrophoresis 2007, 28, 2503–2526

K. A. Kahle and J. P. Foley

Figure 11. Schematic structures of the complexes (charges omitted): (a) [Cu(CDmh)(L-Trp)]; (b) [Cu(CDmh)(D-Trp)]; (c) [Cu(CDhm)(D-Trp)]; (d) [Cu(CDhm)(L-Trp)] [159]. Reproduced with permission.

isopropyl derivative). Importantly, amino acids with a chiral center at the b position to the carboxylic group were resolved. Contrary to the amino acid trend, mandelic acid was best separated with the most hindered chiral selector (diisopropyl derivatized). Overall, the pesticides displayed approximately the same enantioresolution with the dimethyl and diethyl tergurides but significantly lower values with diisopropyl. A sensitive chiral CE method with fluorescence detection for the separation of serine enantiomers (naphthalene-2,3dicarboxaldehyde tagged) in neural samples was successfully created by Quan et al. [168]. The addition of a saccharide (D(1)-glucose) to a neutral CD (HP-g-CD) improved the separation. Potentially interfering compounds were analyzed and found to elute separately from the target analyte (D-serine) under the optimized conditions. The distribution of Dserine in rat brain was determined. Previous publications that utilized saccharides as chiral selectors demonstrated that a combination of hydrophobic interactions and hydrogen bonding are responsible for chiral recognition [169] with specific requirements for D-glucose chiral selectivity being that the C-1 oxygen is in the a-configuration, the C-4 hydroxyl group is oriented downward, and that either a methyl or hydroxymethyl group be present at C-5 [170]. Three publications by Martinez-Gomez et al. [171–173] utilized HSA as a chiral selectand (affinity EKC) during the review period. The first report detailed the separation of oxprenolol enantiomers [171]. The partial filling technique was used because the chiral selector has the disadvantage of being UV absorbing. The highest resolution was 1.47 and the © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

method was applied to commercial product analysis. The second study was expanded to include three additional chiral pharmaceuticals (alprenolol, propranolol, and promethazine) and multivariate optimization (Box-Behnken) [172]; pH, selector concentration, HSA plug length, and their cross-products were examined through experimental design (pareto charts, partial least squares regression, and multiple linear regressions performed) to increase resolution. Hydrophobic analytes required lower HSA concentrations and smaller plug lengths, whereas the opposite was found for more hydrophilic compounds. In the last publication, HSA was explored as the chiral selectand for 28 pairs of enantiomers [173]. The authors investigated the underlying interactions responsible for chiral discrimination with HSA and found that hydrophobicity, electronic character, and steric effects all played important roles. In addition, general guidelines were given for the potential of HSA to produce enantioresolution of target analytes. Nine chiral flavonoids were tested with chiral selectors of either neutral cyclosophoraoses or highly sulfated cyclosophoraoses with and without SDS micelles [174]. Cyclosophoraoses are oligosaccharides, specifically cyclic, unbranched (1 ? 2)b-D-glucans, which are produced by microorganisms. The neutral version of the chiral agent was prepared from cell cultures and then derivatized to create the highly sulfated analog. The exact structure and recognition mechanism for this type of chiral agent has yet to be elucidated. Three of the analytes, catechin, isosakuranetin, and neohesperidin, could be resolved to varying degrees when the chiral selector was combined with micelles. The first aptamer use for chiral CE was published by Ruta et al. in 2006 [175]. Anti-arginine L-RNA (partial sequence containing the necessary binding region) acted as the chiral selector (partial filling mode) to differentiate arginine enantiomers. In order to elute the more strongly bound D-enantiomer, the temperature had to be raised above 507C to sufficiently decrease its interaction with the aptamer. Thermodynamic and kinetic contributions to efficiency were discussed. The influence of sample load was examined and the data explained with relation to desorption kinetics and the number and strength of selector binding sites. The conformations of the chiral selector were also studied and deemed to be contributors to binding site heterogeneity. The chiral ionic liquid (R)-N,N,N-trimethyl-2-aminobutanol-bis-(trifluormethanesulfon)imidate was reported for the enantioresolution of 15 pairs of enantiomers [176]. Additionally, the same chiral agent was tested for GC and HPLC enantioseparations. Resolution with CE ranged from 0.6 to 6.8 for the analytes investigated.

5

Concluding remarks

The wide variety and sheer number of publications detailing novel chiral selectors and applications of existing chiral selectors demonstrate the importance of this field of separawww.electrophoresis-journal.com

Electrophoresis 2007, 28, 2503–2526

CE and CEC

2523

Figure 12. Influence of the size of the urea side chain on the chiral separation of racemic Dns-aspartic acid. L-Enantiomer always eluted prior to D-enantiomer [167]. Reproduced with permission.

tion science. The ability to successfully develop and validate chiral EKC methods has been demonstrated more frequently in recent years, with approved methods now in place for regulated environments such as pharmaceuticals (i.e., USP and EP), agrochemicals, and biological sample analysis. Careful examination of solution phenomena can be used to elucidate the separation mechanisms and in turn improve the separation quality. As research continues in chiral EKC, additional chiral agents will undoubtedly be discovered for especially challenging samples. Additional studies on recognition mechanisms can only further the knowledge base and lead to faster method development. Many opportunities still exist for separation scientists to fill in gaps and improve existing chiral EKC technology.

[5] Pyell, U., Electrokinetic Chromatography: Theory, Instrumentation and Applications, John Wiley & Sons, West Sussex 2006, p. 552. [6] Gübitz, G., Schmid, M. G., J. Chromatogr. A 1997, 792, 179– 225. [7] Gübitz, G., Schmid, M. G., Electrophoresis 2000, 21, 4112– 4135. [8] Maier, N. M., Franco, P., Lindner, W., J. Chromatogr. A 2001, 906, 3–33. [9] Rizzi, A., Electrophoresis 2001, 22, 3079–3106. [10] Gübitz, G., Schmid, M. G., Methods Mol. Biol. (Totowa, NJ) 2004, 243, 1–28. [11] Ward, T. J., Hamburg, D.-M., Anal. Chem. 2004, 76, 4635– 4644. [12] Van Eeckhaut, A., Michotte, Y., Electrophoresis 2006, 27, 2880–2895. [13] Gübitz, G., Schmid, M. G., Mol. Biotechnol. 2006, 32, 159– 179.

6

References

[1] Weinberger, R., Practical Capillary Electrophoresis, Academic Press, San Diego, CA 2000, p. 462. [2] Landers, J. P., Handbook of Capillary Electrophoresis, CRC Press, Boca Raton, FL 1997, p. 894. [3] Khaledi, M. G., High-Performance Capillary Electrophoresis: Theory, Techniques, and Applications, John Wiley, New York 1998, p. 1047. [4] Miller, J. M., Chromatography: Concepts and Contrasts, Wiley, Hoboken, NJ 2005, p. 490.

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

[14] Ward, T. J., Anal. Chem. 2006, 78, 3947–3956. [15] García-Ruiz, C., Marina, M. L., Electrophoresis 2006, 27, 195– 212. [16] Scriba, G. K. E., J. Pharm. Biomed. Anal. 2002, 27, 373–399. [17] Scriba, G. K. E., Electrophoresis 2003, 24, 2409–2421. [18] Ha, P. T. T., Hoogmartens, J., Van Schepdael, A., J. Pharm. Biomed. Anal. 2006, 41, 1–11. [19] Altria, K., Marsh, A., Sänger-van de Griend, C., Electrophoresis 2006, 27, 2263–2282. [20] Holzgrabe, U., Brinz, D., Kopec, S., Weber, C. et al., Electrophoresis 2006, 27, 2283–2292.

www.electrophoresis-journal.com

2524

K. A. Kahle and J. P. Foley

Electrophoresis 2007, 28, 2503–2526

[21] Mandrioli, R., Raggi M. A., Electrophoresis 2006, 27, 213– 221.

[56] Marsh, A., Clark, B., Broderick, M., Power, J. et al., Electrophoresis 2004, 25, 3970–3980.

[22] Natishan, T. K., J. Liq. Chromatogr. Relat. Technol. 2005, 28, 1115–1160.

[57] Huie, C. W., Electrophoresis 2006, 27, 60–75.

[23] Schou, C., Heegaard Niels, H. H., Electrophoresis 2006, 27, 44–59. [24] Erny, G. L., Cifuentes, A., J. Pharm. Biomed. Anal. 2006, 40, 509–515. [25] Zhang, Y.-P., Yuan, Z.-B., Sepu 2001, 19, 417–422. [26] Hernandez-Borges, J., Rodriguez-Delgado, M. A., GarcíaMontelongo, F. J., Cifuentes, A., Electrophoresis 2005, 26, 3799–3813. [27] Cifuentes, A., Electrophoresis 2006, 27, 283–303. [28] Peace, R. W., Gilani, G. S., J. AOAC Int. 2005, 88, 877–887. [29] Berthod, A., Anal. Chem. 2006, 78, 2093–2099. [30] Wren, S. A. C., Rowe, R. C., J. Chromatogr. 1992, 603, 235– 241. [31] Wren, S. A. C., J. Chromatogr. 1993, 636, 57–62.

[58] Aiken, J. H., Huie, C. W., Chromatographia 1993, 35, 448– 450. [59] Pascoe, R., Foley, J. P., Analyst 2002, 127, 710–714. [60] Mertzman, M. D., Foley, J. P., Electrophoresis 2004, 25, 3247– 3256. [61] Mertzman, M. D., Foley, J. P., Electrophoresis 2004, 25, 723– 732. [62] Mertzman, M. D., Foley, J. P., Electrophoresis 2004, 25, 1188– 1200. [63] Iqbal, R., Rizvi, S. A. A., Akbay, C., Shamsi, S. A., J. Chromatogr. A 2004, 1043, 291–302. [64] Mertzman, M. D., Foley, J. P., J. Chromatogr. A 2005, 1073, 181–189. [65] Mertzman, M. D., Foley, J. P., Electrophoresis 2005, 26, 4153– 4163.

[32] Wren, S. A. C., Rowe, R. C., Payne, R. S., Electrophoresis 1994, 15, 774–778.

[66] Kahle, K. A., Foley, J. P., Electrophoresis 2006, 27, 896–904.

[33] Wren, S. A. C., Electrophoresis 1995, 16, 2127–2131.

[67] Kahle, K. A., Foley, J. P., Electrophoresis 2006, 27, 4321–4333.

[34] Kafri, R., Lancet, D., Chirality 2004, 16, 369–378.

[68] Kahle, K. A., Foley, J. P., Electrophoresis 2007, 28, 1723–1734.

[35] Kahle, C., Deubner, R., Schollmayer, C., Scheiber, J. et al., Eur. J. Org. Chem. 2005, 8, 1578–1589.

[69] Zheng, Z.-X., Lin, J.-M., Chan, W.-H., Lee, A. W. M. et al., Electrophoresis 2004, 25, 3263–3269.

[36] Morris, K. F., Froberg, A. L., Becker, B. A., Almeida, V. K. et al., Anal. Chem. 2005, 77, 254A–263A.

[70] Shpigun, O. A., Ananieva, I. A., Budanova, N. Y., Shapovalova, E. N., Russ. Chem. Rev. 2003, 72, 1035–1054.

[37] Morris, K. F., Becker, B. A., Valle, B. C., Warner, I. M. et al., J. Phys. Chem. B 2006, 110, 17359–17369.

[71] Evans, C. E., Stalcup, A. M., Chirality 2003, 15, 709–723.

[38] Chankvetadze, B., Electrophoresis 2002, 23, 4022–4035. [39] Otsuka, K., Terabe, S., J. Chromatogr. A 2000, 875, 163–178.

[72] Blaschke, G., Chankvetadze, B., Drugs Pharm. Sci. 2003, 128, 175–209.

[40] El Rassi, Z., J. Chromatogr. A 2000, 875, 207–233.

[73] de Boer, T., de Zeeuw, R. A., de Jong, G. J., Ensing, K., Electrophoresis 2000, 21, 3220–3239.

[41] Palmer, C. P., Electrophoresis 2000, 21, 4054–4072.

[74] Terabe, S., Electrophoresis 1994, 15, 1227–1228.

[42] Yarabe, H. H., Billiot, E., Warner, I. M., J. Chromatogr. A 2000, 875, 179–206. [43] Akbay, C., Rizvi, S. A. A., Shamsi, S. A., Anal. Chem. 2005, 77, 1672–1683. [44] Hefnawy, M. M., J. Liq. Chromatogr. Relat. Technol. 2005, 28, 439–452. [45] Peng, X., Sternberg, E., Dolphin, D., Electrophoresis 2005, 26, 3861–3868. [46] Wistuba, D., Trapp, O., Gel-Moreto, N., Galensa, R. et al., Anal. Chem. 2006, 78, 3424–3433. [47] Tian, K., Chen, H., Tang, J., Chen, X. et al., J. Chromatogr. A 2006, 1132, 333–336. [48] Rizvi, S. A. A., Shamsi, S. A., Electrophoresis 2005, 26, 4172– 4186. [49] Billiot, F. H., Billiot, E. J., Ng, Y. K., Warner, I. M., J. Chromatogr. Sci. 2006, 44, 64–69. [50] Hou, J., Rizvi, S. A. A., Zheng, J., Shamsi, S. A., Electrophoresis 2006, 27, 1263–1275. [51] Rizvi, S. A. A., Shamsi, S. A., Anal. Chem. 2006, 78, 7061– 7069. [52] Owen, R. L., Strasters, J. K., Breyer, E. D., Electrophoresis 2005, 26, 735–751. [53] Mohanty, A., Dey, J., J. Chromatogr. A 2005, 1070, 185–192. [54] Mohanty, A., Dey, J., J. Chromatogr. A 2006, 1128, 259–266. [55] Hansen, S. H., Electrophoresis 2003, 24, 3900–3907.

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

[75] Fillet, M., Hubert, P., Crommen, J., J. Chromatogr. A 2000, 875, 123–134. [76] De Pablos, R. R., García-Ruiz, C., Crego, A. L., Marina, M. L., Electrophoresis 2005, 26, 1106–1113. [77] Cucinotta, V., Giuffrida, A., Maccarrone, G., Messina, M. et al., J. Pharm. Biomed. Anal. 2005, 37, 1009–1014. [78] Mikus, P., Valaskova, I., Havranek, E., J. Pharm. Biomed. Anal. 2005, 38, 442–448. [79] Mikus, P., Kubacak, P., Valaskova, I., Havranek, E., Talanta 2006, 70, 840–846. [80] Shi, Y., Huo, C., Yao, H., Gao, R. et al., J. Chromatogr. A 2005, 1072, 279–282. [81] Beaufour, M., Morin, P., Ribet, J.-P., J. Sep. Sci. 2005, 28, 529– 533. [82] Berzas Nevado J. J., Guiberteau-Cabanillas, C., Villasenor Llerena M. J., Rodriguez-Robledo, V., J. Chromatogr. A 2005, 1072, 249–257. [83] Sokoließ, T., Köller, G., Electrophoresis 2005, 26, 2330–2341. [84] García-Ruiz, C., Alvarez-Llamas, G., Puerta, A., Blanco, E. et al., Anal. Chim. Acta 2005, 543, 77–83. [85] Gomara, B., García-Ruiz, C., Marina, M. L., Electrophoresis 2005, 26, 3952–3959. [86] Westall, A., Malmstroem, T., Petersson, P., Electrophoresis 2006, 27, 859–864.

www.electrophoresis-journal.com

Electrophoresis 2007, 28, 2503–2526 [87] Berzas-Nevado, J. J., Villasenor-Llerena, M. J., GuiberteauCabanillas, C., Rodriguez-Robledo, V., Electrophoresis 2006, 27, 905–917.

CE and CEC

2525

[117] Cardoso, C. D., Jabor, V. A. P., Bonato, P. S., Electrophoresis 2006, 27, 1248–1254.

[88] Trapp, O., Electrophoresis 2006, 27, 2999–3006.

[118] Liu, Y., Sha, Q., Wu, S., Wang, J. et al., J. Ind. Microbiol. Biotechnol. 2006, 33, 274–282.

[89] Long, P. H., Trung Tran, Q., Oh Joung, W., Kim Kyeong, H., Arch. Pharm. Res. 2006, 29, 808–813.

[119] Afshar, M., Thormann, W., Electrophoresis 2006, 27, 1517– 1525.

[90] Rocheleau, M.-J., Electrophoresis 2005, 26, 2320–2329.

[120] Afshar, M., Thormann, W., Electrophoresis 2006, 27, 1526– 1536.

[91] Cruz L. A., Hall, R., J. Pharm. Biomed. Anal. 2005, 38, 8–13. [92] Kvasnicka, F., Biba, B., Cvak, L., Kratka, J. et al., J. Chromatogr. A 2005, 1081, 87–91. [93] Vaccher, M.-P., Lipka, E., Bonte, J.-P., Foulon, C. et al., Electrophoresis 2005, 26, 2974–2983. [94] Danel, C., Lipka, E., Bonte, J.-P., Goossens, J.-F. et al., Electrophoresis 2005, 26, 3824–3832. [95] Theurillat, R., Knobloch, M., Levionnois, O., Larenza, P. et al., Electrophoresis 2005, 26, 3942–3951. [96] Kavran-Belin, G., Rudaz, S., Veuthey, J.-L., J. Sep. Sci. 2005, 28, 2187–2192. [97] Rudaz, S., Geiser, L., Souverain, S., Prat, J. et al., Electrophoresis 2005, 26, 3910–3920. [98] Zandkarimi, M., Shafaati, A., Foroutan, S. M., Oliaee, N. H. et al., Daru 2005, 13, 143–147. [99] Zhao, Y., Yang, X.-B., Jiang, R., Sun, X.-L. et al., Chirality 2006, 18, 84–90. [100] Schappler, J., Guillarme, D., Prat, J., Veuthey, J.-L. et al., Electrophoresis 2006, 27, 1537–1546. [101] Iwata, Y. T., Inoue, H., Kuwayama, K., Kanamori, T. et al., Forensic Sci. Int. 2006, 161, 92–96. [102] Ramstad, T., J. Chromatogr. A 2006, 1127, 286–294. [103] Ramstad, T., Mills, K. A., Walker, G. A., J. Liq. Chromatogr. Relat. Technol. 2006, 29, 1703–1724. [104] Wang, Z.-Y., Xu, X.-X., Hu, Z.-D., Kang, J.-W., Chin. J. Chem. 2006, 24, 1384–1387. [105] Matthijs, N., Vander Heyden, Y., Biomed. Chromatogr. 2006, 20, 696–709.

[121] Simplicio, A. L., Matias, P., Gilmer, J. F., Clancy, J. M., J. Chromatogr. A 2006, 1120, 89–93. [122] Ha, P. T. T., Sluyts, I., Van Dyck, S., Zhang, J. et al., J. Chromatogr. A 2006, 1120, 94–101. [123] Lin, C.-E., Cheng, H.-T., Fang, I. J., Liu, Y.-C. et al., Electrophoresis 2006, 27, 3443–3451. [124] Hammitzsch, M., Rao, R. N., Scriba, G. K. E., Electrophoresis 2006, 27, 4334–4344. [125] Lin, C.-H., Kuo, C.-M., Liu, Y.-C., Cheng, H.-T. et al., Electrophoresis 2006, 27, 4345–4350. [126] Fang, L., Kang, J., Yin, X.-B., Yang, X. et al., Electrophoresis 2006, 27, 4516–4522. [127] Iio, R., Chinaka, S., Takayama, N., Hayakawa, K., Anal. Sci. 2005, 21, 15–19. [128] Iio, R., Chinaka, S., Takayama, N., Hayakawa, K., J. Health Sci. 2005, 51, 693–701. [129] Busby, M. B., Vigh, G., Electrophoresis 2005, 26, 1978–1987. [130] Busby, M. B., Vigh, G., Electrophoresis 2005, 26, 3849–3860. [131] Choi, S.-H., Noh, H.-J., Lee, K.-P., Bull. Korean Chem. Soc. 2005, 26, 1549–1554. [132] Van Eeckhaut, A., Michotte, Y., Electrophoresis 2006, 27, 2376–2385. [133] Sellers, J. A., Olsen, B. A., Owens, P. K., Gavin, P. F., J. Pharm. Biomed. Anal. 2006, 41, 1088–1094. [134] Khan, M., Viswanathan, B., Rao, D. S., Reddy, G. S., J. Pharm. Biomed. Anal. 2006, 41, 1447–1452. [135] Ramstad, T., Johnson, R. L., J. Sep. Sci. 2006, 29, 2056– 2064.

[106] Lipka, E., Len, C., Rabiller, C., Bonte, J. P. et al., J. Chromatogr. A 2006, 1132, 141–147.

[136] Muderawan, I. W., Ong, T.-T., Tang, W.-H., Young, D. J. et al., Tetrahedron Lett. 2005, 46, 1747–1749.

[107] Lazar, S., Soukri, M., El Haddad, M., Akssira, M. et al., Chirality 2005, 17, 30–36.

[137] Tang, W.-H., Muderawan, I. W., Ong T.-T., Ng, S.-C., Electrophoresis 2005, 26, 3125–3133.

[108] Phinney, K. W., Toshihide, I., Sanders, L. C., J. Chromatogr. A 2005, 1077, 90–97.

[138] Tang, W.-H., Muderawan, I. W., Ong, T.-T., Ng, S.-C., Anal. Chim. Acta 2005, 546, 119–125.

[109] Bitar, Y., Degel, B., Schirmeister, T., Holzgrabe, U., Electrophoresis 2005, 26, 2313–2319.

[139] Ong, T.-T., Tang, W.-H., Muderawan, W., Ng, S.-C. et al., Electrophoresis 2005, 26, 3839–3848.

[110] Mikus, P., Valaskova, I., Havranek, E., J. Sep. Sci. 2005, 28, 1278–1284.

[140] Tang, W.-H., Muderawan, I. W., Ong, T.-T., Ng, S.-C., J. Chromatogr. A 2005, 1091, 152–157.

[111] Kavalirova, A., Pospisilova, M., Karlicek, R., Farmaco 2005, 60, 834–839.

[141] Tang, W.-H., Muderawan, I. W., Ng., S.-C., Chan H. S. O., J. Chromatogr. A 2005, 1094, 187–191.

[112] Maier, V., Horakova, J., Petr, J., Tesarova, E. et al., J. Pharm. Biomed. Anal. 2005, 39, 691–696.

[142] Tang, W.-H., Muderawan, I. W., Ng, S.-C., Chan, H. S. O., Anal. Chim. Acta 2005, 554, 156–162.

[113] Yang, G. S., Chen, D. M., Yang, Y., Tang, B. et al., Chromatographia 2005, 62, 441–445.

[143] Tang, W.-H., Muderawan, I. W., Ng, S.-C., Chan, H. S. O., Anal. Chim. Acta 2006, 555, 63–67.

[114] Lin, C.-E., Liao, W.-S., Cheng, H.-T., Kuo, C.-M. et al., Electrophoresis 2005, 26, 3869–3877.

[144] Liu, Q., Tillekeratne, L. M. V., Kim, Y.-W., Kirchhoff, J. R. et al., Chirality 2005, 17, 570–576.

[115] Lin, C.-E., Lin, S.-L., Cheng, H.-T., Fang, I. J. et al., Electrophoresis 2005, 26, 4187–4196.

[145] Lin, X., Zhao, M., Qi, X., Zhu, C. et al., Electrophoresis 2006, 27, 872–879.

[116] Mandrioli, R., Musenga, A., Lasaponara, S. S., Saracino, M. A. et al., Anal. Chim. Acta 2006, 560, 57–63.

[146] Zhao, M. G., Hao, A. Y., Li, J., Lin, X.-L., Chin. Chem. Lett. 2006, 17, 407–410.

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.electrophoresis-journal.com

2526

K. A. Kahle and J. P. Foley

Electrophoresis 2007, 28, 2503–2526

[147] Fantacuzzi, M., Bettoni, G., D’Orazio, G., Fanali, S., Electrophoresis 2006, 27, 1227–1236.

[164] Aizawa, S.-I., Yamamoto, A., Kodama, S., Electrophoresis 2006, 27, 880–886.

[148] Schmid, M. G., Gübitz, G., Methods Mol. Biol. (Totowa, NJ) 2004, 243, 317–321.

[165] Cucinotta, V., Giuffrida, A., Maccarrone, G., Messina, M. et al., Dalton Trans. 2005, 16, 2731–2736.

[149] Salami, M., Jira, T., Otto, H. H., Pharmazie 2005, 60, 181– 185.

[166] Wu, B., Wang, Q., Liu, Q., Xie, J. et al., Electrophoresis 2005, 26, 1013–1017.

[150] Gong, X. Y., Kubánˇ, P., Tanyanyiwa, J., Hauser, P. C., J. Chromatogr. A 2005, 1082, 230–234.

[167] Honzatko, A., Cvak, J., Vaingatova, S., Flieger, M., J. Sep. Sci. 2005, 28, 673–677.

[151] Kuwahara, Y., Nagata, H., Nishi, H., Tanaka, Y. et al., Chromatographia 2005, 62, 505–510.

[168] Quan, Z., Song, Y., Feng, Y., LeBlanc, M. H. et al., Anal. Chim. Acta 2005, 528, 101–106.

[152] Gong, X. Y., Hauser, P. C., Electrophoresis 2006, 27, 4375– 4382.

[169] Kano, K., Minami, K., Horiguchi, K., Ishimura, T. et al., J. Chromatogr. A 1995, 694, 307–313.

[153] Wang, C.-Y., Wang, D.-H., Leng, T.-H., Yu, Q.-S., J. Heterocycl. Chem. 2005, 42, 1043–1045.

[170] Nakamura, H., Sano, A., Sumii, H., Anal. Sci. 1998, 14, 375– 378.

[154] Ward, T. J., Farris III, A. B., J. Chromatogr. A 2001, 906, 73– 89.

[171] Martinez-Gomez, M. A., Martinez-Pla, J. J., Sagrado, S., Villanueva-Camanas, R. M. et al., J. Pharm. Biomed. Anal. 2005, 39, 76–81.

[155] Bauvais, C., Barbault, F., Zhu, Y., Petitjean, M. et al., SAR QSAR Environ. Res. 2006, 17, 253–264. [156] Jiang, Z., Bertazzo, M., Suessmuth, R. D., Yang, Z. et al., Electrophoresis 2006, 27, 1154–1162. [157] Gao, W., Kang, J., J. Chromatogr. A 2006, 1108, 145–148. [158] Schmid, M. G., Gübitz, G., Methods Mol. Biol. (Totowa, NJ) 2004, 243, 375–386. [159] Cucinotta, V., Giuffrida, A., Maccarrone, G., Messina, M. et al., Electrophoresis 2006, 27, 1471–1480. [160] Kodama, S., Yamamoto, A., Iio, R., Aizawa, S.-I. et al., Electrophoresis 2005, 26, 3884–3889.

[172] Martinez-Gomez, M. A., Villanueva-Camanas, R. M., Sagrado, S., Medina-Hernandez, M. J., Electrophoresis 2005, 26, 4116–4126. [173] Martinez-Gomez, M. A., Villanueva-Camanas, R. M., Sagrado, S., Medina-Hernandez, M. J., Electrophoresis 2006, 27, 4364–4374. [174] Park, H., Jung, S., Electrophoresis 2005, 26, 3833–3838. [175] Ruta, J., Ravelet, C., Grosset, C., Fize, J. et al., Anal. Chem. 2006, 78, 3032–3039. [176] Yuan, L. M., Han, Y., Zhou, Y., Meng, X. et al., Anal. Lett. 2006, 39, 1439–1449.

[161] Koidl, J., Hoedl, H., Schmid, M. G., Pantcheva, S. et al., Electrophoresis 2005, 26, 3878–3883.

[177] Zhao, S., Song, Y., Liu, Y.-M., Talanta 2005, 67, 212–216.

[162] Zheng, Z.-X., Wei, Y., Lin, J.-M., Electrophoresis 2005, 26, 1007–1012.

[178] Gong, X. Y., Hauser, P. C., J. Chromatogr. A 2005, 1094, 196– 199.

[163] Nakajima, H., Kawata, K., Shen, H., Nakagama, T. et al., Anal. Sci. 2005, 21, 67–71.

[179] Chen, F., Zhang, S., Qi, L., Chen, Y., Electrophoresis 2006, 27, 2896–2904.

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.electrophoresis-journal.com