Comparison of the performance of chiral stationary phase for separation of fluoxetine enantiomers

Comparison of the performance of chiral stationary phase for separation of fluoxetine enantiomers 1 2 first-author †E-mail:jie_0822@yahoo.com.cn †E-mail:jie_0822@yahoo.com.cn jie_0822@yahoo.com.cn 1 3 1 Institute of Pharmaceutical Engineering, Department of Chemical Engineering, Zhejiang University, Hangzhou 310027, China Institute of Pharmaceutical Engineering, Department of Chemical Engineering, Zhejiang University, Hangzhou 310027, China School of Pharmacy, Zhengzhou University, Zhengzhou 450001, China School of Pharmacy, Zhengzhou University, Zhengzhou 450001, China Department of Biological and Pharmaceutical Engineering, Ningbo Institute of Technology, Zhejiang University, Ningbo 315100, China Department of Biological and Pharmaceutical Engineering, Ningbo Institute of Technology, Zhejiang University, Ningbo 315100, China Separation of fluoxetine enantiomers on five chiral stationary phases (chiralcel OD-H, chiralcel OJ-H, chiralpak AD-H, cyclobond І 2000 DM and kromasil CHI-TBB) was investigated. The optimal mobile phase compositions of fluoxetine separation on each column were hexane/isopropanol/diethyl amine (98/2/0.2, v/v/v), hexane/isopropanol/diethyl amine (99/1/0.1, v/v/v), hexane/isopropanol/diethyl amine (98/2/0.2, v/v/v), methanol/0.2% triethylamine acetic acid (TEAA) (25/75, v/v; pH 3.8) and hexane/isopropanol/diethyl amine (98/2/0.2, v/v/v), respectively. Experimental results demonstrated that baseline separation ( R S > 1.5) of fluoxetine enantiomers was obtained on chiralcel OD-H, chiralpak AD-H, and cyclobond І 2000 DM while the best separation was obtained on the last one. The eluate orders of fluoxetine enantiomers on the columns were determined. The first eluate by chiralcel OJ-H and kromasil CHI-TBB is the S -enantiomer, while by chiralpak AD-H and cyclobond I 2000 DM is the R -enantiomer. Fluoxetine Chiral stationary phase Enantiomeric separation High performance liquid chromatography (HPLC) INTRODUCTION 1 Fluoxetine (Fig. 1 1 ) is a selective serotonin-re-uptake inhibitor used for treating depression and obsessive-compulsive disorders. So far the drug used is racemate, but the individual optical isomers do not have identical activity (Agranat and Cancer, 1999 1999 ): R -isomer is effective for treating depression while S -isomer is for migraine headaches. Study on the separation of racemic fluoxetine hydrochloride is of important significance. A few chromatographic methods for the separation of fluoxetine enantiomers have been reported. Indirect methods of GC (Torok-Both et al., 1992 1992 ) and LC (Guo et al., 2002 2002 ; 2004 2004 ; Wang et al., 2004 2004 ; Eap et al., 1996 1996 ; Potts and Parli, 1992 1992 ) require the preparation of derivatives. The disadvantage of methods using chiral mobile phases (Piperaki et al., 1995 1995 ) is the consumption of many chiral mobile phases. HPLC using chiral stationary phases (CSPs) is more effective (Gatti et al., 2003 2003 ; Kaddoumi et al., 2001 2001 ; Olsen et al., 1998 1998 ; Wang and Yun, 1997 1997 ; Bakhtiar and Tse, 2000 2000 ; Shen et al., 2002 2002 ; Yee et al., 2000 2000 ; Piperaki and Poulou, 1993 1993 ; Berthod et al., 1990 1990 ; Yu et al., 2002 2002 ). By using tris (3,5-dimethylphenyl carbamate) cellulose (chiralcel OD-H), Gatti et al.( 2003 2003 ) obtained a separation factor ( α ) of 1.10. Olsen et al.( 1998 1998 ) obtained a resolution of 1.4 by using ovomucoid (Ultron ES-OVM). A separation factor of 1.08 was obtained on urea derivative (Wang and Yun, 1997 1997 ), and 1.09 on Teicoplanin (Bakhtiar and Tse, 2000 2000 ) CSP. Furthermore, baseline separation was achieved by using α-acid glycoprotein (chiral AGP) or β-cyclodextrin (cyclobond I) (Piperaki and Poulou, 1993 1993 ; Berthod et al., 1990 1990 ). The present work was aimed at studying the performance of four CSPs for the separation of fluoxetine enantiomers, i.e. chiralcel OJ-H, chiralpak AD-H, cyclobond I 2000 DM and kromasil CHI-TBB. The performance of chiralcel OD-H was also determined for comparison. EXPERIMENTAL DETAILS 1 Reagents 2 HPLC-grade hexane, methanol, acetonitrile and isopropyl alcohol were from B & J Brand (Muskegon, MI). Ultra pure water was obtained by Milli-Q system (Millipore, Milford, MA). Acetic acid, diethylamine (DEA) and triethylamine (TEA) were filtered with 0.45 µm solvent filter and ultrasonically degassed. Samples of racemic fluoxetine hydrochloride were from Lijing Ltd. (Taizhou, China). Apparatus 2 The chromatograph system consisted of a Waters 1525 Binary HPLC pump, a Waters 717 plus multiple autosampler and a Waters 2487 Dual λ UV detector (Waters, USA). The detection wavelength was 226 nm and the detector temperature was 20 °C. Chromatograms were recorded at 1 Hz, and peak-areas were determined by using a Waters Breeze data acquisition system. Columns of chiralcel OD-H, chiracel OJ-H, chiralpak AD-H and cyclobond I 2000 DM were from chiral technologies (Exton, PA, USA). Kromasil CHI-TBB column was from Eka Chemicals AB (Bohus, Sweden). The columns were all of the same size of 25 cm×4.6 mm i.d., and the CSP particle size was 5 µm. Dead volumes of the columns were determined with 1,3,5-tri-tert-butylbenzene for chiralcel OD-H, chiralcel OJ-H and chiralpak AD-H, toluene for cyclobond I 2000 DM, and trifluoroethene for kromasil CHI-TBB. Sample preparation 2 Five milligrams of fluoxetine HCl was dissolved in 25 ml mobile phase. The injection size was 10 µl. RESULTS AND DISCUSSION 1 Comparison of separation performance of different chiral columns 2 It is well known that the performance of a definite column is affected by the composition and the flow rate of the mobile phase. Experiments were carried out to find the optimal composition and flow rate for each column. The best separation results of the five columns at optimal conditions are listed in Table 1 1 , the corresponding chromatograms are shown in Fig. 2 2 . Mobile phase compositions of Nos. 1, 3 and 5 are hexane/isopropanol/DEA, 98/2/0.2 (v/v/v), of No. 2 are hexane/isopropanol/DEA, 99/1/0.1 (v/v/v), of No. 4 are methanol/0.2% TEAA (triethylamine acetic acid), 25/75 (v/v), pH 3.8; Flow rate of mobile phase of Nos. 1 and 4 is 0.8 ml/min, of Nos. 2, 3, and 5 is 0.5 ml/min; Column temperature of No. 1 is 15 °C, of Nos. 2 and 3 is 20 °C, of No. 4 is 30 °C, of No. 5 is 23 °C 2 No. 2 Stationary phase 2 Retention time (min) 2 Capacity factor 2 Separation factor ( α ) 2 Resolution ( R S ) t R1 t R2 k 1 ′ k 2 ′ 1 Chiralcel OD-H 18.4 20.3 4.10 4.63 1.13 1.74 2 Chiralcel OJ-H 19.9 20.9 2.27 2.44 1.07 0.99 3 Chiralpak AD-H 19.2 21.4 2.18 2.54 1.16 1.79 4 Cyclobond I 2000 DM 16.4 19.2 3.32 4.05 1.22 2.30 5 Kromasil CHI-TBB 11.4 12.1 2.12 2.22 1.05 0.84 Chromatogram numbers are the same as in Table 1 1 The performance of a stationary phase is usually evaluated by two criteria: resolution and capacity factor. The resolution should be acceptable if the enantiomers are baseline separated; while the capacity factor should be small so that the retention time is short and the mobile phase consumption is low. Fig. 2 2 of the results shows that fluoxetine enantiomers are baseline separated by three of the CSPs studied: chiralcel OD-H, chiralpak AD-H and cyclobond I 2000 DM. Among them the best separation is obtained by cyclobond I 2000 DM (5-dimethyl-β-cyclodextrin), with a resolution of 2.30 (Table 1 1 ), much greater than that of all other CSPs reported in the literature. This is perhaps due to the hydrogen bond, dipole-dipole and π-π interactions, and especially due to the hydrophobic cavity of β-cyclodextrin affording a chiral environment. As compared to cyclobond I (β-cyclodextrin bonded CSP), the selectivity is probably enhanced by the addition of dimethyl groups. The performance of chiralpak AD-H [tris (3,5-dimethylphenyl carbamate) amylose] is also satisfactory. Its capacity factor is much less than the other two CSPs while the resolution is good enough. In consequence, its mobile phase consumption is much less while the retention time is comparable to the other two CSPs. The performance of chiralcel OD-H [tris (3,5-dimethylphenyl carbamate) cellulose] is in-between that of the other two. Its resolution is similar to that of chiralpak AD-H while its capacity factor is similar to cyclobond I 2000 DM. Regarding the chemical nature of the CSPs, the resolution on cyclodextrin is the best, the mobile phase consumption by amylose is the least, and the performance of cellulose is somewhat in-between. Fluoxetine enantiomers are not well separated by chiralcel OJ-H and kromasil CHI-TBB (0,0′-bis 4-tert-butylbenzoyl)-N,N′-diallyl-L-tartar diamide). The structures of chiralcel OJ-H [tris (4-dimethylphenyl carbamate) cellulose] and chiralcel OD-H are similar, differing only by one functional group, and provides an example of the significant effect of functional groups on separation performance. Elution order of fluoxetine on different chiral columns 2 To the authors’ knowledge, the elution order of fluoxetine enantiomers on chiralcel OJ-H, chiralpak AD-H, cyclobond I 2000 DM and kromasil CHI-TBB have never been reported. In the present work the elution order of three CSPs was examined. Gatti et al.( 2003 2003 ) reported that the S -enantiomer of fluoxetine is eluted earlier than the R -enantiomer on chiralcel OD-H with hexane/isopropanol as mobile phase. The first eluate of multiple injections on chiralcel OJ-H, cyclobond I 2000 DM and kromasil CHI-TBB were collected and evaporated. Then the samples were injected into a chiralcel OD-H column to obtain the chromatograms under the same conditions (mobile phase: hexane/isopropanol/DEA, 98/2/0.2 (v/v/v); flow rate: 0.5 ml/min; ambient temperature). Fig. 3 3 of chromatograms shows that the first eluate from chiralcel OJ-H and kromasil CHI-TBB is the S -enantiomer, and that from cyclobond I 2000 DM is the R -enantiomer. CONCLUSION 1 Fluoxetine enantiomers were separated for the first time on four columns: chiralpak AD-H, chiralcel OJ-H, cyclobond I 2000 DM and kromasil CHI-TBB. The resolution order of the studied CSPs for separation of fluoxetine enantiomers was: cyclobond I 2000 DM>chiralpak AD-H≈chiralcel OD-H>chiralcel OJ-H>kromasil CHI-TBB. Among them fluoxetine enantiomers are not resolved on the last two. The first eluate by chiralcel OD-H, chiralcel OJ-H and kromasil CHI-TBB is the S -enantiomer, while by chiralpak AD-H and cyclobond I 2000 DM, it is the R -enantiomer. References I Agranat , H Caner . Intellectual property and chirality of drugs . Drug Discov Today , 1999 . 4 ( 7 ): 313 - 321 . 10377509 DOI: 10.1016/S1359-6446(99)01363-X http://doi.org/10.1016/S1359-6446(99)01363-X . Intellectual property and chirality of drugs R Bakhtiar , F Tse . High-throughput chiral liquid chromatography/tandem mass spectrometry . Rapid Commun Mass Spectrom , 2000 . 14 ( 13 ): 1128 - 1135 . 10867688 DOI: 10.1002/1097-0231(20000715)14:13<1128::AID-RCM1>3.0.CO;2-5 http://doi.org/10.1002/1097-0231(20000715)14:13<1128::AID-RCM1>3.0.CO;2-5 . High-throughput chiral liquid chromatography/tandem mass spectrometry A Berthod , HL Jin , TE Beesley , JD Duncan , DW Armstrong . Cyclodextrin chiral stationary phases for liquid chromatography separations of drug stereoisomers . J Pharm Biomed Anal , 1990 . 8 ( 2 ): 123 - 130 . 2094413 DOI: 10.1016/0731-7085(90)80018-K http://doi.org/10.1016/0731-7085(90)80018-K . Cyclodextrin chiral stationary phases for liquid chromatography separations of drug stereoisomers CB Eap , N Gaillard , K Powell , P Baumann . Simultaneous determination of plasma levels of fluvoxamine and of the enantiomers of fluoxetine and norfluoxetine by gas chromatography-mass spectrometry . J Chromatogr B , 1996 . 682 265 - 272 . . Simultaneous determination of plasma levels of fluvoxamine and of the enantiomers of fluoxetine and norfluoxetine by gas chromatography-mass spectrometry G Gatti , I Bonomi , R Marchiselli , C Fattore , E Spina , G Scordo , R Pacifici , E Perucca . Improved enantioselective assay for the determination of fluoxetine and norfluoxetine enantiomers in human plasmaby liquid chromatography . J Chromatogr B , 2003 . 784 ( 2 ): 375 - 383 . DOI: 10.1016/S1570-0232(02)00820-6 http://doi.org/10.1016/S1570-0232(02)00820-6 . Improved enantioselective assay for the determination of fluoxetine and norfluoxetine enantiomers in human plasmaby liquid chromatography X Guo , T Fukushima , F Li , K Imai . Determination of fluoxetine enantiomers in rat plasma by pre-column fluorescence derivatization and column-switehing high-performance liquid chromatography . Analyst , 2002 . 127 ( 4 ): 480 - 484 . 12022644 DOI: 10.1039/b111053p http://doi.org/10.1039/b111053p . Determination of fluoxetine enantiomers in rat plasma by pre-column fluorescence derivatization and column-switehing high-performance liquid chromatography XJ Guo , Y Xu , FM Li . Enantiomeric separation of derivatized fluoxetine on chiralcel OD column . Chin J Anal Chem , 2004 . 32 ( 10 ): 1353 - 1355 . . Enantiomeric separation of derivatized fluoxetine on chiralcel OD column A Kaddoumi , MN Nakashima , K Nakashima . Fluorometric determination of DL-fenfluramine, DL-norfenfluramine and phentermine in plasma by achiral and chiral high-performance liquid chromatography . J Chromatogr B , 2001 . 763 79 - 90 . . Fluorometric determination of DL-fenfluramine, DL-norfenfluramine and phentermine in plasma by achiral and chiral high-performance liquid chromatography BA Olsen , DD Wirth , JS Larew . Determination of fluoxetine hydrochloride enantiomeric excess using high-performance liquid chromatography with chiral stationary phases . J Pharm Biomed Anal , 1998 . 17 ( 4-5 ): 623 - 630 . 9682145 DOI: 10.1016/S0731-7085(98)00050-8 http://doi.org/10.1016/S0731-7085(98)00050-8 . Determination of fluoxetine hydrochloride enantiomeric excess using high-performance liquid chromatography with chiral stationary phases S Piperaki , MP Poulou . Use of cyclodextrin as chiral selectors for direct resolution of the enantiomers of fluoxetine and its metabolite norfluoxetine by HPLC . Chirality , 1993 . 5 ( 4 ): 258 - 266 . DOI: 10.1002/chir.530050414 http://doi.org/10.1002/chir.530050414 . Use of cyclodextrin as chiral selectors for direct resolution of the enantiomers of fluoxetine and its metabolite norfluoxetine by HPLC S Piperaki , SG Penn , DM Goodall . Systematic approach to treatment of enantiomeric separations in capillary electrophoresis and liquid chromatography II: a study of the enantiomeric separation of fluoxetine and norfluoxetin . J Chromatogr A , 1995 . 700 ( 1-2 ): 59 - 67 . DOI: 10.1016/0021-9673(95)00113-2 http://doi.org/10.1016/0021-9673(95)00113-2 . Systematic approach to treatment of enantiomeric separations in capillary electrophoresis and liquid chromatography II: a study of the enantiomeric separation of fluoxetine and norfluoxetin BD Potts , CJ Parli . Analysis of the enantiomers of fluoxetine and norfluoxetine in plasma and tissue using chiral derivatization and normal-phase liquid chromatography . J Liq Chromatogr , 1992 . 15 ( 4 ): 665 - 681 . . Analysis of the enantiomers of fluoxetine and norfluoxetine in plasma and tissue using chiral derivatization and normal-phase liquid chromatography Z Shen , S Wang , R Bakhtiar . Enantiomeric separation and quantification of fluoxetine in human plasma by liquid chromatography/tandem mass spectrometry using liquid-liquid extraction in 96-well plate format . Rapid Commun Mass Spectrom , 2002 . 16 ( 5 ): 332 - 338 . 11857715 DOI: 10.1002/rcm.580 http://doi.org/10.1002/rcm.580 . Enantiomeric separation and quantification of fluoxetine in human plasma by liquid chromatography/tandem mass spectrometry using liquid-liquid extraction in 96-well plate format GA Torok-Both , GB Baker , RT Coutts . Simultaneous determination of fluoxetine and norfluoxetine enantiomers in biological samples by gas chromatography with electron-capture detection . J Chromatogr , 1992 . 579 99 - 106 . 1447355 . Simultaneous determination of fluoxetine and norfluoxetine enantiomers in biological samples by gas chromatography with electron-capture detection ZX Wang , ZH Yun . Separation of fluoxetine enantiomers by HPLC with urea derivative as chiral stationary phase . Chin J Anal Chem , 1997 . 25 ( 4 ): 464 - 467 . . Separation of fluoxetine enantiomers by HPLC with urea derivative as chiral stationary phase TJ Wang , J Li , Y Wang , XJ Guo , CH Zhang , FM Li . Enantiomeric separation of derivatized fluoxetine by HPLC on amylose stationary phase . Chin J Pharm Anal , 2004 . 24 ( 3 ): 318 - 320 . . Enantiomeric separation of derivatized fluoxetine by HPLC on amylose stationary phase L Yee , S Wong , V Skrinska . Chiral high-performance liquid chromatography analysis of fluoxetine and norfluoxetine in rabbit plasma, urine and vitreous humor using an acetylated β-cyclodextrin column . J Anal Toxicol , 2000 . 24 651 - 655 . 11043675 . Chiral high-performance liquid chromatography analysis of fluoxetine and norfluoxetine in rabbit plasma, urine and vitreous humor using an acetylated β-cyclodextrin column HW Yu , CB Ching , P Fu , SC Ng . Enantioseparation of fluoxetine on a new β-cyclodextrin bonded phase column by HPLC . Separation Sci Technol , 2002 . 37 ( 6 ): 1401 - 1415 . DOI: 10.1081/SS-120002618 http://doi.org/10.1081/SS-120002618 . Enantioseparation of fluoxetine on a new β-cyclodextrin bonded phase column by HPLC 10.1631/jzus.2007.B0056 2005-11-01 2006-02-14 2006-12-18 2022-01-05 Copyright © 2007, Journal of Zhejiang University Science 2007 Journal of Zhejiang University-SCIENCE B (Biomedicine and Biotechnology) 01 08