Anal. Chem. 2001, 73, 5103-5108
Micellar Electrokinetic Chromatography-Mass Spectrometry Using a Polymerized Chiral Surfactant Shahab A. Shamsi*
Department of Chemistry, Center of Biotechnology and Drug Design, Georgia State University, 38 Peach Tree Center Avenue, Atlanta, Georgia 30303
The coupling of chiral micellar electrokinetic chromatography (CMEKC) to mass spectrometry (MS) using conventional surfactant [above the critical micelle concentration (cmc)] is very challenging. Preliminary investigation in this laboratory indicates that the use of a chiral polymeric surfactant provides one possible solution to this difficult coupling. This is because of many positive attributes of micelle polymers which include zero cmc, lower surface activity, low volatility, high electrophoretic mobility, and function as a suitable separation medium even at lower concentrations of pseudophases. In this work, the feasibility of using poly(sodium N-undecanoylL-valinate (poly-L-SUV) in CMEKC-MS is demonstrated. After CMEKC separation, enantiomers of 1,1′-binaphthol (BOH) were detected using electrospray ionization mass spectrometry (ESI-MS) by selected ion monitoring (SIM) in the negative ion mode. Although in the SIM mode ESIMS parameters (nebulizer pressure, drying gas flow rate, drying gas temperature, and sheath liquid flow rate) affected only the signal-to-noise ratio of (()BOH, two of the ESI-MS parameters (nebulizer pressure, sheath flow rate) were found to have a significant impact on chiral resolution of (()BOH. At the optimum ESI-MS conditions, the enantioseparation of (()BOH was successfully accomplished by varying the buffer pH, concentration of the volatile background electrolyte, and poly-L-SUV. Resolution and structural identification of enantiomeric compounds are necessary steps for studying chiral drug interactions. A recent survey indicated that the annual sale of the world market for chiral drugs exceeds $100 billion in 2000 and is anticipated to increase at a good pace in this millennium.1 Analytical separation methods such as high-performance liquid chromatography (HPLC) are predominantly used to control the enantiomeric purity of starting materials and products.2-6 However, important progress has been made in the past five years on chiral separation based * Corresponding author: (tel) (404) 651-1297; (fax) 404-651-2751; (e-mail)
[email protected]. (1) Stinson, C. Cheml. Eng. News 2000, (Oct 23), 55-78. (2) Gu ¨ bitz, G. Chromatographia 1990, 30, 555-564. (3) Tang, Y. Chirality 1996, 8, 136-142. (4) Ahuja, S., Ed. Chiral Separations Applications and Technology; American Chemical Society: Washington, DC, 1997. (5) Haginaka, J. J. Chromatogr., A 2001, 906, 253-274. (6) Tachibana, K.; Ohnishi, A. J. Chromatogr., A 2001, 906, 127-154. 10.1021/ac0105179 CCC: $20.00 Published on Web 09/22/2001
© 2001 American Chemical Society
solely on capillary electrophoresis (CE).7-13 Speed, simplicity, low amounts of sample and chiral selector consumption, and good compatibility with biological samples are some additional advantages of chiral CE over the more established enantioseparation techniques such as chiral HPLC. In capillary zone electrophoresis (CZE) and micellar electrokinetic chromatography (MEKC), UV absorbance is the most commonly used detection method. However, electropherograms obtained from the analysis of unknown samples or samples containing contaminants may result in several peaks. Despite the fact that the photodiode array detector is now commonly used for peak purity identification in both HPLC and CE, only limited structural information is obtained from the UV spectra. Another challenge to the UV detection method for CE is the determination of trace levels of pharmaceuticals and metabolites in pharmacokinetic studies. Thus, more sensitive detectors are highly desirable. Coupling of chiral CE with mass spectrometry (MS) may solve the identification problems associated with unknown chiral compounds. This is because MS provides important information not only about mass but also about the structure of the separated compounds. This benefit of CE-MS is particularly useful for the separation and structure confirmation of metabolites in clinical and biopharmaceutical analysis. Fanali and co-workers reported the potential of chiral CE-MS for stereoselective analysis of the phase I and phase II metabolites of the nonsteroidal antiinflammatory drug etodolac in human urine.14 The work of these researchers illustrates the value of on-line chiral CE-MS coupling for peak identification and peak purity testing of each enantiomeric peak. It should be mentioned that fraction collection for further analysis is time-consuming and still remains a challenge in CE compared to HPLC. Hence, on-line coupling of a high-resolution separation technique such as chiral CE with the powerful structure elucidation technique of MS has been of great interest in the past few years. (7) Chankvetadze, B.; Blaschke, G. J. Chromatogr., A 2001, 906, 309-364. (8) Fanali, S. J. Chromatogr. 1991, 545, 437-444. (9) Nishi, H. J. Chromatogr., A 1996, 735, 345-351. (10) Stalcup, A. M.; Gahm, K.-H. Anal. Chem. 1996, 68, 1360-1368. (11) Armstrong, D. W.; Rundlett, K. L.; Chen, J.-R. Chirality 1994, 6, 496-509. (12) Armstrong, D. W.; Gasper, M. P.; Rundlett, K. L. J. Chromatogr., A 1995, 689, 285-304. (13) Gasper, M. P.; Berthod, A.; Nair, U. B.; Armstrong, D. W. Anal. Chem. 1996, 68, 2501-2514. (14) Fanali, S.; Desiderio, C.; Schulte G.; Heitmeier, S.; Strickmann, D.; Chankvetadze, B.; Blaschke, G. J. Chromatogr., A 1998, 800, 69-76.
Analytical Chemistry, Vol. 73, No. 21, November 1, 2001 5103
One modern mode of CE that can be used for assessing the potential of micelles to recognize both neutral and charged (anions and cations) chiral molecules is chiral micellar electrokinetic chromatography (CMEKC).15-18 Shamsi and Warner19 reviewed several chiral surfactants that are successfully used in MEKC with UV detection. However, these monomerics (unpolymerized chiral micelles) seem to cause problems when CMEKC is coupled to MS. The appearance of conventional chiral micelles in the ionization source of the MS reduces the sensitivity since they often compete with an analyte for available charge. In addition, relatively high concentrations of chiral surfactants are necessary to achieve chiral separations in MEKC which in turn increases the background noise of the mass spectrometer. In recent years, there has been rapid growth in the development of a new type of micellar system, i.e., “polymerized surfactants” or simply “a micelle polymer” as a pseudostationary phase for both achiral20-24 and chiral separations25-34 in CE. These micelle polymers have provided several distinct advantages over conventional (unpolymerized) micelles in MEKC. First, the presence of covalent bonds formed between monomers after polymerization eliminates the dynamic equilibrium that exists between surfactant monomers and the micelle aggregates, thus, simplifying and enhancing the process of binding between the micelle and the solute.25,26 Second, due to zero critical micelle concentration (cmc), polymeric surfactants can be used at any concentrations at which they are soluble. In contrast, normal micelles require higher surfactant concentrations (at least 2-10 times the cmc) for effective separations. As a result, Joule heating is much higher in micellar-mediated MEKC than in MEKC with micelle polymers. Third, in buffers modified with a higher volume fraction of organic solvents, the chromatographic selectivity is superior to that with SDS micelle.24 For example, as high as 75% (v/v) acetonitrile or methanol can be used with micelle polymers, whereas the SDS micelle can only tolerate ∼30-40% (v/v) of these (15) Nishi, H.; Terabe, S. J. Chromatogr., A 1996, 735, 3-27. (16) Fanali, S. J. Chromatogr., A 1996, 735, 77-121. (17) St. Claire, R. L., III. Anal. Chem. 1996, 68, 569R-586R. (18) Williams, C. C.; Shamsi, S. A.; Warner, I. M. Chiral Micelle Polymers for Chiral Separations in Capillary Electrophoresis. In Advances in Chromatography; Brown, P. R., Grushka, E., Eds.; Marcel Dekker: New York, 1997; Vol. 37, pp 363-419. (19) Shamsi, S. A.; Warner, I. M. Electrophoresis 1997, 18, 853-872. (20) Palmer, C. P.; McNair, H. M. J. Microcolumn Sep. 1993, 4, 509-514. (21) Palmer, C. P.; Khaled, M. Y.; McNair, H. M. J. High. Resolut. Chromatogr. 1992, 15, 756-762. (22) Palmer, C. P.; Terabe, S. J. Microcolumn Sep. 1996, 8, 115-121. (23) Palmer, C. P.; Terabe, S. Anal. Chem. 1997, 69, 1852-1860. (24) Shamsi, S. A.; Akbay, C.; Warner, I. M. Anal. Chem. 1998, 70, 3078-3083. (25) Wang, J.; Warner, I. M. Anal. Chem. 1994, 66, 3773-3776. (26) Wang, J.; Warner, I. M. J. Chromatogr. 1995, 711, 297-304. (27) Agnew-Heard, K. A.; Sanchez Pena M.; Shamsi, S. A.; Warner, I. M. Anal. Chem. 1997, 69, 958-964. (28) Shamsi, S. A.; Macossay, J.; Warner, I. M. Anal. Chem. 1997, 69, 29802987. (29) Macossay, J.; Shamsi, S. A.; Warner, I. M. Tetrahedron Lett. 1999, 40, 577580. (30) Billiot, E.; Macossay, J.; Shamsi, S. A.; Warner, I. M. Anal. Chem. 1998, 70, 1375-1381. (31) Billiot, E.; Agbaria, R. A.; Shamsi, S. A.; Warner, I. M. Anal. Chem. 1999, 71, 1252-1256. (32) Billiot, E.; Thibodeaux, S.; Shamsi, S. A.; Warner, I. M. Anal. Chem. 1999, 71, 4044-4049. (33) Haddadian, F.; Billiot, E.; Shamsi, S. A.; Warner, I. M. J. Chromatogr., A 1999, 858, 219-227. (34) Yarabe, H. H.; Shamsi, S. A.; Warner, I. M. Anal. Chem. 1999, 71, 39923999.
5104
Analytical Chemistry, Vol. 73, No. 21, November 1, 2001
solvents.22-24 Fourth, MEKC with micelle polymers provides a more extended elution window than MEKC; therefore, higher peak capacity is observed.35 The coupling of MEKC with MS using conventional micelles (above the cmc) is very difficult if not impossible. Currently, there are only a few reports on the use of achiral surfactants (cationic, anionic, or nonionic) with CE-ESI-MS.36-39 Cole and Varghese37 used CE-ESI-MS for the separation and detection of some cationic peptides using a cationic surfactant, cetyltrimethylammonium bromide, over the concentration range of 0.5-2.5 mM. Kirby et al.40 reported that cationic polypeptides can be CE separated and detected with ES-MS using a mixed surfactant system containing 1.25 mM each of SDS (which is below the cmc) and a nonionic surfactant, poly(ethylene glycol)lauryl ether.40 However, no polypeptide signal was observed when SDS was used alone. The researchers hypothesize that the nonionic surfactant may have sequestered the SDS, reducing its charge density and the ion suppression effect. To this investigator’s knowledge, a MEKCMS method using a monomeric surfactant above its cmc (i.e., in the micelle form) has not yet been developed. Several researchers have reported a fouling of the ionization source in the ESI interface when conventional micelles are used. For example, the use of high concentrations of anionic surfactant, SDS, produces strong signals in both positive and negative ion modes. At SDS concentrations required for micellar-mediated CE separations, SDS-analyte adduct formation suppresses the ionization efficiency of the source, resulting in poor signal-to-noise ratios (S/N) for the cationic analytes.40 For these reasons, Ozaki and co-workers41 explored the possibility of detecting cationic solutes using butyl acrylate-butyl methacrylate-methacrylic acid (BBMA) as an anionic micelle polymer in MEKC-MS. They suggested that the improved S/N for analyte ion can be attributed to the low volatility of the polymeric surfactant employed as pseudostationary phase. Another recent study by Rundlett and Armstrong42 indicated that the ability of surfactants to quench ES ionization may be directly related to the surface activity and the charge of the surfactant. The authors suggested that low surface activity of polymeric surfactant could be the reason for enhancement in MS signal intensity for the solute of interest. Thus, there is sufficient evidence in the literature that polymeric surfactant should have considerable advantages over monomeric surfactants when employed as a pseudophase for micellar-mediated MEKC separations that allow detection of the separated analyte ions by ESI-MS. However, there are no reports on the use of polymeric chiral surfactants using MEKC coupled to MS. (35) Shamsi, S. A.; Akbay, C.; Warner, I. M. Electrophoresis 1999, 20, 145-151. (36) Wendsjo, S.; Ornskov, E. AMEKC with Monomeric Micelles. Presented at the 6th International Symposium on High Performance Capillary Electrophoresis; San Diego, CA, Jan 31-Feb 3, 1994. (37) Cole, B. R.; Varghese, J. J. Chromatogr. 1993, 652, 369-376. (38) Dekkers, S. E. G.; Tjaden, U. R.; Greef J. V. D. J. Chromatogr. 1995, 712, 201-209. (39) Lu, W.; Poon, G. K.; Carmichael, P. L.; Cole, B. R. Anal. Chem. 1996, 68, 668-674. (40) Kirby, D.; Greve, K. F.; Foret, F.; Vourous, P.; Karger, B. L. Capillary Electrophoresis Electrospray Ionization Mass Spectrometry Utilizing Electrolytes Containin Surfactants. Presented at the 42nd ASMS Conference on Mass Spectrometry; Chicago, IL, May 29-June3, 1994: pp 1014-1015. (41) Ozaki, H.; Itou, H.; Terabe, S.; Tahada, Y.; Shaairi, M.; Koizumi, H. J. Chromatogr. 1995, 716, 69-79. (42) Rundlett, K. L.; Armstrong, D. W. Anal. Chem. 1996, 68, 3493-3497.
In the present study, the use of chiral polymeric surfactants as one possible solution for analysis of chiral compounds via CMEKC-MS is proposed. The feasibility of the approach is demonstrated by a systematic method development of (()-1,1′bi-2-naphthol [(()BOH] using poly(sodium N-undecanoyl-L-valinate) (poly-L-SUV) as chiral selector. EXPERIMENTAL SECTION Instrumentation. All MEKC-ESI-MS experiments were carried out with an Agilent CE system interfaced to a quadrupole mass spectrometer, Agilent 1100 series MSD (Agilent Technologies, Palo Alto, CA). To interface the CE instrument with the mass spectrometer, a CE-MS adapter kit and a CE-ESI-MS sprayer kit (all from Agilent Technologies) were used. The CE-MS adapter kit includes a capillary cassette, which allows thermostating a part of the capillary within the instrument, a capillary safety sleeve, and a plastic UV capillary alignment interface to permit simultaneous UV-MS detection. The sprayer kit consists of a triple tube CE-MS orthogonal sprayer which provides both a coaxial sheath liquid makeup and a nebulization gas to aid in droplet formation. Both the nebulizing gas and the drying gas were nitrogen and were delivered using a nitrogen generator. To maintain electrical contact, the coaxial sheath liquid was delivered using an Agilent 1100 series isocratic HPLC pump. This pump, which is equipped with 1:100 splitter, allows sheath liquid to be delivered in the microliter per minute range. The Agilent ChemStation and CE-MSD add-on software was used for instrument control and data analysis. CE-ESI-MS Conditions and Procedures. Polyimide-coated fused-silica capillaries of 50-µm i.d. and 375-mm o.d. was obtained from Polymicro Technologies (Phoenix, AZ). The capillaries were cut to various lengths from a spool of a stock capillary bundle, prepared by burning a 2-3-mm segment of polyimide coating to create a UV detection window (22 cm from inlet to UV alignment interface), and installed in Agilent cassettes for CE-MS or CEUV-MS. A new capillary was conditioned for 1 h with 1 N NaOH at 60 °C, followed by a 10-min rinse with 1 N NH4OH and triply distilled water prior to the insertion into the sprayer needle. After the installation into the sprayer, the capillary was flushed with ammonium acetate (NH4OAc) at a pH of 9.2 for 30 min. The capillary was filled with the running MEKC buffer for 3 min before each MEKC run. The enantiomers of (()BOH were injected with pressure injection of 50 mbar for 3-5 s, involving volumes between 3 and 5 nL. All CE separations were performed at a constant voltage of 20 kV, and the capillary temperature was thermostated at 25 ( 1 °C. To assist electrospray, the sheath liquid containing 5 mM NH4OAc in 50% (v/v) methanol/water was delivered at various flow rates of 1-10 µL/min. A potential of -2500 V was applied to the sprayer tip for optimal electrospray performance. To optimize the fragmentor voltage in the scan mode, the m/z range was scanned from 50 to 3000 m/z units at the rate of 1.2 s/scan. The MS detection was performed in the selective ion monitoring (SIM) mode. Since (()BOH exist as anions in basic solutions, negative [M - H]- ions were monitored at 285 m/z with a dwell time of 1000 ms. Materials. The individual (R)-(+) and (S)-(-) enantiomers of BOH were obtained from Aldrich (Milwaukee, WI). HPLCgrade methanol was purchased from EM Science (Gibbstown, NJ).
The SigmaUltra grade (99%) NH4OAc used as a background electrolyte (BGE) was obtained as a 7.5 M solution from Sigma (St. Louis, MO). The poly-L-SUV was prepared according to the procedure previously published.25,28 Preparation of Analyte and Electrolyte Solution. The BGE was prepared by diluting the stock 7.5 M NH4OAc solution in triply deionized water and then adjusting to various pH values (9.2-11.2) with ammonium hydroxide. Various amount of polyL-SUV was added to the above BGE. All MEKC buffer containing surfactants was filtered through a 0.45-µm syringe filter (Nalgene, Rochester, NY) by creating a vacuum inside the syringe. This was followed by ultrasonication for ∼5 min to ensure properly degassed MEKC buffers. Sheath liquid solution was prepared by simply diluting the 7.5 M stock solution of NH4OAc to provide a final concentration of 5 mM NH4OAc in 50/50% (v/v) of methanol/ water. RESULTS AND DISCUSSION Using poly-L-SUV, chiral micellar electrokinetic chromatography-mass spectrometry (CMEKC-MS) was optimized for the separation and detection of a model chiral analyte, (()BOH. This compound is relatively hydrophobic and has a chiral plane with C2 symmetry instead of a chiral center (atropisomer). First, the electrospray parameters (fragmentor voltage, nebulizing gas pressure, drying gas temperature, drying gas flow rate, sheath liquid flow rate) were evaluated to achieve the highest sensitivity. After that, the chiral separation conditions were studied in terms of pH, micelle polymer concentration, and volatile background electrolyte concentrations to obtain the best possible chiral separation without sacrificing ESI-MS sensitivity. Evaluation of the Electrospray Ionization Experimental Parameters. A manual tuning on the mass spectrometer was performed to optimize the fragmentor voltage. In a typical procedure, a solution of 0.1 mg/mL (()BOH was introduced continuously (∼5 min) from the CE inlet vial by using ∼1 bar pressure. After the initial spray chamber parameters were set, the scan was started by selecting a scan range from 50 to 3000 m/z. Once the MS signal intensity was optimized to 100% abundance of the largest analyte peak, fragmentor voltage was ramped. From this manual tune procedure, the skimmer voltage of 104 V was found to produce a signal of maximum sensitivity. Effects of Nebulizing Gas Pressure. To study the effect of nebulizer gas pressure on MS signal intensity and chiral resolution, experiments were performed at nebulizer pressure from 6 to 20 psi. Figure 1 shows that the ESI response (measured as the S/N) increases but chiral resolution of (()BOH decreases with increasing nebulizer pressure. Note that migration times were also increased by decreasing the nebulization pressure. This suggests that the nebulizer is generating a suction resulting in a laminar flow inside the capillary. Therefore, as a compromise between chiral resolution and ESI response, the nebulizer pressure was set to its moderate setting (10 psi). Certainly, the optimization of this parameter is essential for successful development of a CMEKC-MS methodology. Effects of Drying Gas Flow Rate and Drying Gas Temperature. The flow rate of the drying gas (N2) and its temperature can influence the S/N of the electrospray. Panels A and B of Figure 2 show that S/N increases with decreasing gas flow rate whereas the reverse occurred when the gas temperature was Analytical Chemistry, Vol. 73, No. 21, November 1, 2001
5105
Figure 3. Effects of sheath liquid flow rate on chiral resolution and MS signal intensity of (()BOH. Conditions same as Figure 1 except the drying gas flow rate was set at 6 L/min and drying gas temperature at 200 °C. Table 1. Optimum ESI-MS Experimental Parameters for (()BOH in the SIM Mode Figure 1. Effects of nebulizer pressure on chiral resolution and MS signal intensity of (()BOH. Conditions: run buffer, 0.25% w/v polyL-SUV/10 mM NH4OAc, pH 9.2; capillary, fused-silica, total length 106 cm, 50 µm i.d.; +20 kV, 90 mbar/s pressure injection; sheath liquid, 5 mM NH4OAc in 50% methanol, 5 µL/min; drying gas, nitrogen, 8 L/min, 200 °C; acquisition, negative mode, Vcap -2.5 kV, fragmentor voltage 110 V, PW 0.25 min, SIM at 285 m/z.
Figure 2. (A) Effects of drying gas flow rate and (B) effects of drying gas temperature on MS signal intensity of (()BOH. Conditions same as Figure 1 except the nebulizer pressure was set at 10 psi.
increased. However, unlike nebulization pressure, these two parameters had no influence on migration time and chiral resolution of (()BOH(data not shown). Thus, for subsequent CMEKC-MS experiments, the drying gas flow rate and the temperature were maintained at 6 L/min and 200 °C, respectively. Effects of Sheath Liquid Flow Rate. The choice of a sheath liquid flow rate was also found to be critical in developing the chiral CE-ESI-MS method. Recent studies indicated that a sheath liquid composition of 5 mM ammonium acetate in 50% methanol 5106 Analytical Chemistry, Vol. 73, No. 21, November 1, 2001
nebulizing gas pressure drying gas flow rate drying gas temperature sheath liquid flow rate fragmentor voltage electrospray voltage peak width
N2, 10 psi N2, 6 L/min N2, 200 °C 4 µL/min 104 V -25 kV 0.15 min
provides the highest sensitivity for amino acids.43 Therefore, at this composition, the effects of sheath liquid flow rates on chiral resolution and sensitivity of (()BOH was studied. As shown in Figure 3, initially increasing the flow rate (from 1 to 4 µL/min) increases the S/N by a factor of 3. However, at a flow rate greater than 4 µL/min, a general trend of decreasing S/N was evident. While the differences in chiral resolution of (()BOH is not as dramatic (Rs ∼0.5 and 1.0 was observed at 1 and 10 µL/min, respectively) the differences were significant. However, the differences in Rs values are not significant enough to allow us to draw any real conclusions. It should be noted that, unlike nebulization pressure, the effects of sheath liquid flow rate has no impact on migration time of the separated chiral enantiomers (data not shown). As a compromise between chiral resolution and S/N ratio, the flow rate of the sheath liquid (4 µL/min) might be appropriate. Further studies are planned with a variety of other types of chiral analytes and various other sheath liquid compositions. The evaluation of electrospray parameters in the SIM mode resulted in the optimized conditions for MS detection of (()BOH. The optimum value of each experimental parameter is listed in Table 1. Optimization of the Micellar Electrokinetic Chromatography Conditions. To transfer a chiral MEKC-UV method to a chiral MEKC-MS method, the choice and concentration of the BGE and the micellar solution as well as the pH are very essential considerations. For example, the use of a nonvolatile BGE and surfactants results in strong background ions in the mass spectra as has been observed with SDS.37,40 In addition, the ion spray process is impaired using the high ionic strength of the BGE. (43) Soga, T.; Heiger, D. N. Anal. Chem. 2000, 72, 1236-1241.
Figure 4. Effects of polymer concentration on chiral resolution and MS signal intensity of (()BOH. MS conditions same as optimized in Table 1. MEKC conditions: 35 mM NH4OAC, pH 9.2, 0.02-1.0% (w/v) poly-L-SUV; injection 250 mbar/s of 0.2 mg/mL +BOH, +20 kV, SIM (285 m/z).
Therefore, optimizing these parameters to find a compromise between a good MEKC separation and a high MS sensitivity is essential. A running MEKC buffer containing 0.25% poly-L-SUV and 10 mM NH4OAc was studied over the range of pH between 9 and 11.5. However, unlike MEKC-UV detection, where chiral resolution is best seen at pH 10.2 (using Tris buffer),31,32 with MEKC-MS, the best combination of chiral resolution and MS sensitivity was observed at a pH of 9.2 using NH4OAc as BGE (data not shown). Effects of Polymer Concentration on Chiral Resolution and MS Signal Intensity of (+)BOH. Figure 4 shows the effect of micelle polymer (poly-L-SUV) on chiral resolution and MS signal intensity. It is not too surprising that an increase in micelle polymer concentration decreases the chiral resolution of (()BOH.25 However, note that even the introduction of a 1% (w/v) poly-L-SUV solution into the MEKC-MS system generated a stable electrospray and did not impair the MS detection sensitivity significantly. It appears that the high electrophoretic mobility of the micelle polymer toward the anodic end (an injection end in positive-polarity CE) prevents the chiral selector from entering the MS interface. Because of the limited chiral separation obtained in the MEKC-MS-compatible buffers, it is important to gain as much resolution as possible by optimum surfactant concentration. Therefore, a conclusive compromise between chiral resolution and S/N appears to be around 0.2% (w/v) poly-L-SUV. Effects of Background Electrolyte (NH4OAc) Concentration. Even higher chiral resolution in MEKC-MS can be achieved at low sample concentration and injection size and high ionic strength of the volatile BGE. This is shown in Figure 5 where chiral resolution of (()BOH as high as 3.2 was obtained using 50 mM NH4OAc and 90 mbar/s injection (compared to 250 mbar/s injection in Figure 5). However, this occurred at the
Figure 5. Effects of BGE concentration on chiral resolution and MS signal intensity of (()BOH. MEKC-MS conditions same as optimized in Figure 3 except for 0.20% (w/v) poly-L-SUV and 90 mbar/s injection of 0.1 mg/mL +BOH.
expense of longer separation times and decrease in MS sensitivity. Therefore, the conclusive compromise of the background electrolyte concentration is around 20 mM NH4OAc. This BGE concentration appears to be a reasonable tradeoff between chiral resolution and analysis time, while maintaining a high sensitivity in MS. Simultaneous UV and MS Detection. The simultaneous use of both UV (diode array) and MS detectors in series with CE can provide a wealth of complementary information that would be beneficial for many clinical studies. For example, there are many cases in pharmacokinetic studies where a major metabolite and its degradation products have very similar chemical structure (e.g., differing by only CH3 or OH group). In such cases, the UV-visible spectra will be very similar between the two analytes, and it may be very difficult to identify an impurity based on UV-visible spectra. It is also possible to have impurities that have the same mass, especially at low molecular weights. However, having both UV-visible spectra and the mass of the two compounds be identical would be very unusual. Thus, in such cases using a combination of UV detector and a mass-selective detector is more effective than using either one alone. Figure 6A shows examples of chiral separationof (()BOH with simultaneous UV and MS detection. Because the length of the capillary up to the diode array detector was only 25 cm, the BOH enantiomers were not baseline resolved when UV detector was used. However, the enantiomers were more than baseline separated (Rs ) 2.6) with high sensitivity by MS detection. Furthermore, it should be noted that the (S)-(-) enantiomer of BOH eluted faster than its corresponding (R)-(+) form. This suggests that the (R)-(+)-BOH has a higher affinity with the (L) form of Analytical Chemistry, Vol. 73, No. 21, November 1, 2001
5107
Figure 6. (A.1, A.2) Chiral MEKC-UV-MS electropherogram of (()BOH. (B.1,B.2) Mass spectrum of (S)-(-)and (R)-(+) enantiomers eluted at 36.075 and 37.045 min, respectively. All conditions same as Figure 4.
the micelle polymer, poly-L-SUV. Figure 6B shows the mass spectrum of the BOH enantiomers, the data clearly show that the identical masses of solute that are resolved in a chiral CE system are very reliable signatures of the presence of two optical isomers. Note that the relative mass abundance of each enantiomer is consistent with the peak heights shown in Figure 6B.1 and B.2.
SUV, 20 mM NH4OAc), baseline separation of 0.1 mg/mL (()BOH was observed in less than 30 min with a S/N of ∼93. Future efforts in this laboratory will explore the application of this valuable technology to a wide range of chiral cationic (e.g., β-adrenergic blocking agents, benzodiazepinones) anionic (e.g., warfarin, coumachlor), and neutral (e.g., benzoin derivatives) analytes.
CONCLUSIONS On-line CMEKC-MS using poly-L-SUV was explored. During the chiral method development of (()BOH, consideration was given to optimize both ESI-MS and CMEKC parameters. No detrimental effect of the presence of poly-L-SUV was observed at any stage of the method development. Although the presence of a high concentration [i.e., 1% (w/v) poly-L-SUV] tends to lower S/N for ESI-MS detection, still a reasonable compromise can be reached to permit adequate separation with sensitive detection. Under the optimum CMEKC-MS conditions (nebulizer pressure 10 psi, drying gas flow rate 6 L/min, drying gas temperature 200 °C, sheath liquid flow rate 4 µL/min, pH 9.2, 0.20% (w/v) poly-L-
ACKNOWLEDGMENT is made to the donors of the The Petroleum Research Fund, administered by the ACS, for support of this research. Georgia State University (Research Initiation Grant) is also acknowledged for the research funding. The author thank Dr. Isiah M. Warner (Louisiana State University) for providing access to the polymerization source.
5108 Analytical Chemistry, Vol. 73, No. 21, November 1, 2001
Received for review May 7, 2001. Accepted August 20, 2001. AC0105179
