enantiomeric separation of β-amino alcohols by capillary electrophoresis using dm-β-cd as chiral...
TRANSCRIPT
Enantiomeric Separation ofb-Amino Alcohols by CapillaryElectrophoresis Using DM-b-CD as Chiral SelectorWei Wei, Guoan Luo*
Department of Chemistry, Tsinghua University, Beijing 100084, P.R.China
Chao Yan
Unimicro Technologies, Inc. 4713 First St., Pleasanton CA 94566, USA
Zengpei Sun
National Institute for the Control of Pharmaceutical and Biological Products, Beijing 100050, P.R. China
Ms received: November 7, 1997; accepted: May 29, 1998
Key Words:Enantiomeric separation;b-amine alcohols; capillary electrophoresis; DM-b-CD
SummaryEnantiomeric separations of severalb-amino alcohol drugs,i.e., phe-nylephrine, epinephrine, norepinephrine, synephrine, and chlorpre-naline were performed by capillary electrophoresis using DM-b-CDas a chiral selector. Five test solutes were baseline resolved in six min-utes. The effects of DM-b-CD concentration, pH value, ionic strengthof the buffer, and the type of b-CDs on resolution were investigated.The results indicated that DM-b-CD is suitable for enantiomericseparation of b-amino alcohols containing a phenyl group on thechiral atom. Enantiorecognition mechanisms for test solutes are alsodiscussed.
1 Introduction
Capillary electrophoresis (CE) has become more and more popu-lar in recent years because of its high efficiency, short analysistime, and low solvent and sample consumption. Chiral separationsin CE have been reported by many authors [1–13]. Such chiralseparations are conveniently performed by the use of chiral selec-tor in a buffer, in which inclusion complexation takes placebetween the enantiomeric solutes and the chiral selector. Enantio-meric separations have been demonstrated for different electroki-netic methods, such as capillary zone electrophoresis [1–9],micellar electrokinetic capillary chromatography [10], capillarygel electrophoresis [11], capillary isoelectrofocusing [12], andcapillary electrochromatography [13–15]. Three classes of chiralselectors are often used in CE,viz.cyclodextrins (CDs) [2–4], tet-racarboxylated crown-ethers [16], and antibiotics [17]. CDs areare most widely used in chiral separations.b-CD is one type of thenative CDs made up of seven glucose units connected to eachother througha(1,4) glucoside bonds. It can easily be modified byreaction of the hydroxyl groups at positions 2, 3, and 6 of the glu-cose moiety to form.e.g., heptakis- (2,6-di-O-methyl)-b-CD(DM-b-CD). DM-b-CD has proved to be a better chiral selectorthan b-CD thanks to its increased solubility and the increaseddepth of its cavity (greater hydrophobicity) [18].
Severalb-amino alcohol drugs,i.e, phenylephrine, epinephrine,norepinephrine, synephrine, and chlorprenaline, are of consider-able pharmaceutical importance. Separations of some of these byCE compounds have been studied. Fanali [1] investigated theenantiomeric separations of norepinephrine and epinephrine byCE on 20 cm6 25lm coated capillary column with di-OMe-b-CD as a chiral selector in 10 mM Tris-H3PO4 (pH 2.4) buffers.Quang [3] improved the chiral separations of epineprine and nor-epinephrine usingb-CD and tetraalkylammonium reagents byCE. Wang [6] reported baseline separation of phenylephrine by
using 100 mMb-CD and 6 M urea in 50 mM Tris-H3PO4 buffer(pH 2.3). Recently, Lurie [7] reported the chiral resolution ofcationic drugs of forensic interest by CE with mixtures of neutraland anionic CDs. Their results indicated that DM-b-CD is suita-ble for the analysis of drugs containing norephedrine. In spite ofintensive investigations in the field, a systematic study on enan-tiomeric separations ofb-amino alcohols containing a substitutedphenyl structural moiety on the chiral atom is still lacking.
This paper describes the successful enantiomeric separations offive structurally relatedb-amino alcohols using DM-b-CD asbuffer additives. The effects of DM-b-CD concentrations and thepH and ionic strength of buffer on the enantiomeric resolutionare investigated. In addition, a mechanism of chiral recognitionof enantiomers byb-CDs is discussed.
2 Experimental
2.1 Apparatus
Experiments were performed on a P/ACE 5500 capillary electro-phoresis system (Beckman, Fullerton, CA, USA). The fusedsilica capillaries used in this work were obtained from YongnianOptical Fiber Factory (Yongnian, Heibei, China). The dimensionof the capillary columns was 20 cm6 75lm i.d. (27-cm totallength), 20 cm from the inlet frit to the detection window. On-line UV detector was operated at 214 nm with detection range of0.05 aufs. All experiments were carried out at 258C.
2.2 Reagents and Chemicals
b-CD, heptakis-(2,6-di-O-methyl)-b-CD (DM-b-CD) and hydro-propyl b-CD (HP-b-CD) were purchased from Sigma (St. Louis,MO, USA). Tris(hydroxymethyl)aminomethane (Tris) wasobtained from Shanghai Reagent Factory (Shanghai, China). Thefive drug analogues of phenylephrine, epinephrine, norepineph-rine, synephrine, and chlorprenaline were obtained from the Chi-nese National Institute for the Control of Pharmaceutical andBiological Products (Beijing, China). The structures of thesedrugs are shown inFigure 1.
2.3 Procedures
The background electrolyte (BGE) was prepared by mixing thedesired pH buffer with an appropriate concentration ofb-CD or
440 J. High Resol. Chromatogr. vol. 21, 440–444 (1998) No. 8i WILEY-VCH Verlag GmbH, D-69451 Weinheim 1998 0935-6304/98/0808–0440$17.50+.50/0
Enantiomeric Separationof b-Amino Alcoholsby CapillaryElectrophoresis
J.High Resol.Chromatogr. VOL. 21,AUGUST1998 441
its derivatives.Tris/H3PO4 buffer of 100mM concentrationwasadjustedto the desiredpH using100-mM H3PO4. The BGE wasfiltered througha 0.25-lm filter anddegassedby ultrasonicationfor aboutfive minutesbeforeit wasused.The capillary columnwasinstalledin a cartridgeof the BeckmanModel 5500P/ACEsystem.Thenewcapillarycolumnwasconditionedasfollows:
Rinse1: 1 M HCl for 20 min
Rinse2: waterfor 10 min
Rinse3: 1 M NaOHfor 20min
Rinse4: waterfor 15 min
Rinse5: runbuffer for 10min
Betweentwo consecutiveinjections, the capillary column wasrinsedwith run buffer for 1 min. Sampleswerepreparedby mix-ing a stock solution of 1 mg/ml with methanol.The concentra-tions of the sampleswere 0.2–0.4mg/ml. Electrokineticinjec-tionswereperformedat 1 kV for 1 sor 3 kV over3 s.
Resolution(Rs) of the enantiomerswas calculatedby using thefollowing theequation
Rs � 1:177tR2ÿ tR1
w0:5�1� � w0:5�2�
wheretR1 and tR2 are the migration timesof the first andsecondpeaksof the enantiomers.w0.5(1) andw0.5 (2) arethe half-widthsofthefirst andsecondpeaksof enantiomers.
3 Results and Discussion
3.1Effectof DM-b-CD Concentration
Theresolutionof enantiomersin CE canbeexpressedas
Rs � 0:177Dlep
EL�lav� leo�Dm
� �1=2
�1�
where Rs is the resolution, Dlep is the mobility differencebetweentheenantiomers,lav is theaverageelectrophoreticmobi-lity, leo is the electroosmoticmobility, E and L are the fieldstrengthandthecapillary length,respectively. Dm is thediffusioncoefficient.
For the neutralCDs, Wren and Rowe [19] derivedan equationfor theapparentmobility, laof oneof theenantiomers,
la �lf � lnK1�CD�
1� K1�CD� �2�
wherelf andln arethe electrophoreticmobilities of the uncom-plexedandcomplexedsolutes,respectively, K1 is theequilibriumconstantfor the CD concentrationinclusioncomplex,and[CD]is theCD concentration.
Resolution is determinedmainly by the mobility differencebetweenthe enantiomersand EOF value accordingto Eq. (1).The larger the mobility differencebetweenthe enantiomers,thelarger the resolution.The lower the EOF, the better the resolu-tion. However, the mobility differencebetweenthe enantiomersis also dependenton the equilibrium constantand the CD con-centrationin Eq. (2).
Figure 2 showsthe relationshipbetweenresolutionand DM-b-CD concentrationfor five test compoundsusing 20 mM Tris-H3PO4 (pH 3.24)buffers.A non-linearrelationshipwasobserved.IncreasingDM-b-CD concentrationcausesa generalincreaseofresolutionfor epinephrine,norpinephrine,andsynephrine,whilefor phenylephrine and chloreprenaline resolution initiallydecreaseswith DM-b-CD concentration,but thenreachesa mini-mum before increasingagain at higher DM-b-CD concentra-tions.
Theresultsfor phenylephrineandchloreprenalinepresentedhereseeminconsistentwith Wren’s findings, in which resolutionver-
Figure 1. Thestructureof testsolutes
Figure 2. RelationshipbetweentheDM-b-CDsconcentrationandresolu-tion: (g) phenylephrine;(f) synephrine;(h) epinephrine;(6) norepi-nephrine; (0) chlorprenaline.Experimental conditions: BGE, 20mMTris-H3PO4 (pH 3.24).
Wei, Luo, Yan,Sun
442 VOL. 21,AUGUST 1998 J.High Resol.Chromatogr.
susCD concentrationplots passthrougha maximum and thendecrease[19]. Themostlikely explanationis that inclusioncom-plexationfor two solutesstartsat a low DM-b-CD concentration
of around 6 mM. On increasingthe DM-b-CD concentrationfrom 6 mM to 14mM, the inclusion complexationbetweentheDM-b-CD andenantiomersmaycontinueandtheapparentmobi-lities of the enantiomersincrease,which results in decreasingresolution. At a DM-b-CD concentration above 14 mM, theincreaseof viscosityof buffer seemsto dominate,which resultsin a decreaseof electroosmoticflow velocity, and resolutionincreasesaccordingly.
Theshapesof Rs vs. [DM-b-CD] for thedifferentcompoundsaredifferent, which indicates the discrimination of interactionsbetweenthesolutesandDM-b-CD.
As anexample, Figure 3.a to Figure 3.erespectivelyshowelec-tropherogramsof the baselineseparationsof phenylephrine,epi-nephrine, norepinephrine, synephrine, and chlorprenalineobtainedon usingDM-b-CD aschiral selector. The first peakinthe electropherograms(a) and(c) correspondsto the R form andthe secondto the S form. Unfortunately, we were not able toobtainthestandardsfor theR andSformsof theotherdrugs.Theoverlappingpair of peaksin Figure 3.c wasattributedto a race-mic impurity.
3.2Effectof pH
Theeffect of thepH of thebuffer on resolutionwasinvestigatedat pH 2.52,3.24,4.24,and5.24of 10 mM Tris-H3PO4buffer con-taining 10mM DM-b-CD. Figure 4 is the relationshipbetweenthe resolutionandpH of buffer for phenylephrine andchlorpre-naline,andthe resultscannotbe shownfor the othersolutesdueto the poor resolution at high pH. Generally, increasingpH causeda decreaseof migration time for testcompoundsdueto the increaseelectroosmoticflow (EOF). Resolutionbecomeslarge at low pH in the caseof a low EOF, accordingto Eq.(1).Optimumresolutionfor testcompoundswasfoundat pH 3.24.
Figure 3.a–e. Electropherogramsof five basicdrugs.Experimentalcon-ditions: (a) phenylephrine.BGE: 30mM Tris-H3PO4 containing10mMDM-b-CD (pH 3.24); (b) epinephrine.BGE: 50mM Tris-H3PO4 contain-ing 10mM DM-b-CD (pH 3.24);(c) norepinephrine.BGE: 20mM Tris-H3PO4 containing 16mM DM-b-CD (pH 3.24); (d) synephrine.BGE:30mM Tris-H3PO4 containing10mM DM-b-CD (pH 3.24);(e) chlorpre-nalineBGE: 50mM Tris-H3PO4 containing10mM DM-b-CD (pH 3.24).Capillarycolumn,20cm6 75lm i.d. (27-cmtotal length),appliedvol-tage12kV; detection,214nm (0.01aufs);temperature,25 8C.
Enantiomeric Separationof b-Amino Alcoholsby CapillaryElectrophoresis
J.High Resol.Chromatogr. VOL. 21,AUGUST1998 443
3.3Effectof Ion Strength
The ion strengthof the buffer was found to havea significanteffect on resolution. Separationswere studied at 10mM to50mM Tris-H3PO4. Generally, high ion strength is used forseparationof the enantiomersin order to decreasethe EOF andobtainbetterpeakshape.However, it is believedthatbuffer mayenterinto theCD cavity. Tris buffer maystronglyhydrogenbondto the primary hydroxyl groups and reduce the retention ofsolutes. Furthermore, Joule heating increases at high ionstrength,thus impairing resolution.It is thereforeimportant forenantiorecognitionto control the ion strengthof the buffer. Theeffect of ion strengthon resolutionis shownin Figure 5. A rela-tively good resolutionwas observedat 30mM Tris-H3PO4 fortestsolutes.
3.4Enantiorecognition Mechanisms
In order to explore the enantiorecognition mechanism,b-CD,DM-b-CD, and HP-b-CD are chosenas chiral selectors.It isbelievedthat enantiorecognitionin CD is basedon the inclusion
of anaromaticor alkyl functionality in thehydrophobiccavity ofCDsandadditionalhydrogenbondingbetweensecondaryhydro-xyl groupsaroundthe cone openingand substituentsof guestmolecule[17, 18]. As a result,CDscanincludeapolarmoleculesof appropriatedimensionsand bind them throughdipole-dipoleinteractionsandhydrogenbonding.
The five selectedbasicdrugsall containa hydroxyl groupandahydrogenatomat thechiral centeranda1CH2 in thea-position.Chiral recognitionof testsolutesis observedwith DM-b-CD. Atthe sameDM-b-CD concentration,the resolution for the fivecompoundsfollows theorderchlorprenaline,phenylephrine,epi-nephrine, synephrine,and norepinephrine.The difference inresolutionshouldbe mainly attribute to the substitutedphenylgroup attachedto the chiral atom. Chlorprenalinepossessesachlorophenylgroupat the stereogeniccenter, which is favorablefor theCD hydrophobiccavity. In contrast,theothersolutescon-tain oneto two hydroxyl groupson the phenylring; the ensuingrelativehydrophilicity opposestheir fitting into theCD cavity.
We also investigatedchiral recognitionby HP-b-CD and b-CDunder the same conditions. The resolution of test solutes ispoorerthanthatobtainedon usingDM-b-CD aschiral selector.
DM-b-CD possessesa deepercavity [18] than native b-CD. InHP-b-CD, the secondaryhydroxyls at the mouth of the CDhydrophobiccavity are substitutedwith hydroxypropylgroups.In spiteof theincreasedcavity depth,thehydrophobicityof DM-b-CD is greaterthan that of HP-b-CD becausethereare fewerhydroxyls in DM-b-CD that in HP-b-CD. Among the testsolutes,chlorprenalineshowsthebestresolutionwith DM-b-CD,HP-b-CD, or b-CD underotherwiseidentical separationcondi-tions. The presenceof the hydrophobicaromaticring probablypromotesinclusion of the aromatic ring in the CD cavity. Itappearsthat the enantioselectivityderivesfrom inclusion com-plexation.
Hydrogenbonding interaction also plays an important role inchiral recognitionbecausethe hydroxylsmay provideadditionalsitesfor enantiorecognition[20]. HP-b-CD andb-CD possessthesame number of hydroxyls. If hydrogen-bondinginteractionwerethemain reasonfor chiral recognition,HP-b-CD andb-CDwould obviously give better resolution than DM-b-CD forsolutescontaininghydroxyl groups.The resultsseemto indicatethat the enantiorecognitionresultsfrom inclusion complexationbetweenCD andenantiomersandhydrogen-bondinginteractionplaysonly a minor role in enantiorecognition.
4 Conclusions
Five b-amino alcohol drugs, i.e., phenylephrine,epinephrine,norepinephrine,synephrine,and chlorprenalinewere success-fully resolvedwithin six minutes by capillary electrophoresisusing DM-b-CD as a chiral selector. The effects of DM-b-CDconcentrationand the pH and ionic strengthof the buffer areinvestigated.Comparedwith HP-b-CD andb-CD aschiral selec-tors, DM-b-CD providesthe bestresolutionfor the test solutesusedin this investigation.The preliminary resultsobtainedwiththe selectedchiral compoundsandunderthe givenexperimentalconditions indicate that the enantiorecognitionmechanismisdominatedmainly by the inclusion complexation betweentheCD and solutes.Hydrogenbondinginteractionseemsto play aweakrole in chiral selection.
Figure 4. The relationshipbetweenthe pH of buffer and resolution:(g)phenylephrine, (0) chlorprenaline. Experimental conditions: BGE,10mM Tris-H3PO4 containing10mM DM-b-CDs, other conditionsasstatedin Figure2.
Figure 5. Therelationshipbetweenthe ion strengthof buffer andresolu-tion: (g) phenylephrine;(f) synephrine;(h) epinephrine;(6) norepi-nephrine;(0) chlorprenaline.Experimentalconditions:BGE, Tris-H3PO4
containing10mM DM-b-CDs (pH 3.24), other conditionsas statedinFigure2.
Wei, Luo, Yan,Sun
444 VOL. 21,AUGUST 1998 J.High Resol.Chromatogr.
Acknowledgments
The authorswish to thank the National Natural ScienceFoundationofChinafor thefinancialsupport.
References[1] S.Fanali,J. Chromatogr. 1989, 441, 446.
[2] W. Schutzner, S.Fanali,Electrophoresis1992, 13, 687.
[3] C.Y. Quang,M.G. Khaledi,Anal.Chem.1993, 65, 3354.
[4] R. Kuhn,C. Steinmetz,T. Bereuter, P. Hass,F. Erni, J. Chromatogr.1994, 666, 121.
[5] R. Kuhn, J. Wagner, Y. Walbroehl, T. Bereuter, Electrophoresis1994, 15, 828.
[6] Z. Wang,Y. Sun,Z. Sun,J. Chromatogr. A. 1996, 735, 295.
[7] I.S.Lurie, R.F.X. Klein, T.A.D. Cason,M.J.Lebelle,R. Brenneisen,R.E.Weinberger, Anal.Chem.1994, 66, 4019.
[8] G. Luo, Y. Wang,A.G. Ewing,J . Cap.Elec.1994, 1, 175.
[9] J.Yu, Y. Wang,G. Luo, ActaPharmaceutialSinica1997, 32, 203.
[10] K. Otsuka,T. Terabe,J. Liq. Chromatogr. 1993, 16, 945.
[11] P. Sun, G.E. Barker, R.A. Hartwick, N. Grinberg, R. Kaliszan,J.Chromatogr. A 1993, 247.
[12] I. Jelinek,J. Snopek,E. Smolkova-Keulemansova,J. Chromatogr.1988, 439, 386.
[13] S.Li, D.K. Lloyd, Anal.Chem.1993, 65, 3685.
[14] S.Li, D.K. Lloyd, J. Chromatogr.A 1994, 666, 321.
[15] S.Li, D.K. Lloyd, J. Chromatogr.A 1995, 694, 285.
[16] R. Kuhn, F. Erni, T. Bereuter, J. Hausler, Anal. Chem. 1992, 64,2815.
[17] D.W. Armstrong,K.L. Rundlett,J.R.Chen.,Chirality 1994, 6, 496.
[18] J. Szejtli, Cyclodextrins, their inclusion complexes, AkademiaiKiado,Budapest,1982.
[19] S.A.C.Wren,R.C.Rowe,J. Chromatogr. 1992, 603, 235
[20] A.M. Stalcup,S.C.Chang,D.W. Amstrong,J.Pitha,J. Chromatogr.1990, 513, 181.