Available online at Journal of Chromatography A, 1178 (2008)

Transcription

1 Available online at Journal of Chromatography A, 1178 (2008) High-performance liquid chromatographic evaluation of a coated cellulose tris(3,5-dimethylphenylcarbamate) chiral stationary phase having a small-pore silica support Yueqi Liu,1, Hanfa Zou National Chromatographic Research and Analysis Center, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian , China Received 22 July 2007; received in revised form 18 November 2007; accepted 21 November 2007 Available online 23 November 2007 Abstract A two-step coating/precipitation synthetic procedure has been developed for the preparation of cellulose tris(3,5-dimethylphenylcarbamate) chiral stationary phase (CSP) having a small-pore silica support. With this synthetic strategy, monodisperse, spherical CSP particles can be produced without the need for a wasteful and time-consuming sieving process. The performance of the synthesized CSP towards a variety of racemates was evaluated in the normal-phase HPLC mode. HPLC separation experiments revealed that the synthesized CSP exhibited a chiral recognition ability fully comparable to the corresponding commercial columns prepared using conventional large-pore silica as the support. Moreover, the synthesized CSP was successfully applied to semipreparative enantioseparation of a new triazole antifungal agent Elsevier B.V. All rights reserved. Keywords: Chiral stationary phase; Cellulose tris(3,5-dimethylphenylcarbamate); Small-pore silica; Enantiomeric separation; HPLC 1. Introduction Chromatographic separation of enantiomers from their racemic mixture using a chiral stationary phase has enjoyed broad success in the past 20 years. Various kinds of chiral stationary phases (CSPs) have been introduced and used for this purpose. Among them, polysaccharide derivatives CSPs appear to be the most suitable for the analytical and preparative separations of enantiomers due to their broad enantioselectivity as well as their high sample loading capacity [1 6]. The polysaccharide-based stationary phase generally consists of a chiral polysaccharide derivative, as for example cellulose phenylcarbamates, coated or immobilized on an underlying inert core support such as silica [3,7]. Although the immobilized polysaccharide-based stationary phases can be compatible with a wider range of eluents, they exhibit lower chiral recognition abilities than the coated CSPs in the traditional normal-phase mode Corresponding authors. Tel.: ; fax: addresses: yqliu2008@yahoo.com (Y. Liu), hanfazou@dicp.ac.cn (H. Zou). 1 Faculty of Pharmaceutical Sciences, Mukogawa Women s University, Japan. [8 10]. Moreover, the preparation procedure of the immobilized CSPs is more complicated and time-consuming than that of the coated CSPs. Therefore, at present the coated polysaccharidebased stationary phases are still the most widely used CSPs for enantiomeric separations. Traditionally, the coated polysaccharide-based CSPs are prepared using large pore spherical silicas as the inert core support [2 4]. The usual loading amount of polysaccharide derivative is 20 25% (w/w) [3]. Silicas containing large pores are available, but routes to their preparation result in high cost, low reproducibility and limited availability. In addition, large pore silicas have low mechanical strength which can affect the lifetime of the column. As new chromatographic needs arose, these often were met by changing the properties of silicas. Several attempts have been reported producing cellulose tris(3,5-dimethylphenylcarbamate) (CDMPC)-based CSPs with small-pore silica supports by conventional one-step evaporation coating method [11 14]. Unfortunately, the resulting CSP particles have broad size distribution due to particle agglomeration after evaporating the coating solvent. When narrow disperse particles are desired, which is often the case for HPLC application, sieving of the synthesized particles through a particular size range sieve (<38 m) is necessary [11,13,14]. However, this /$ see front matter 2007 Elsevier B.V. All rights reserved. doi: /j.chroma

2 Y. Liu, H. Zou / J. Chromatogr. A 1178 (2008) sieving process is time-consuming and wasteful, as a considerable amount of large particles is produced, which is of limited suitability for HPLC separations. Even so, the obtained particles are still polydisperse both in shape and size, which has a negative impact on their chromatographic performance through peak broadening and increased column back-pressure. Hence, for the synthesis of monodisperse CDMPC-coated small-pore silica CSP particles with high selectivity, a novel synthetic technique needs to be developed. We hereby present a method that can be used to produce the monodisperse CDMPC-coated CSP particles having small-pore silica support. This method is based on a two-step coating/precipitation procedure, the first being the coating of CDMPC on the surface of silica gels in good solvent such as tetrahydrofuran followed by precipitation of CDMPC onto silica support in poorer solvent. The composite CSP particles formed in this way are protected from agglomeration during coating process and have good spherical shapes and a narrow size distribution similar to the original silica support beads. The obtained monodispersity enables direct HPLC application of the synthesized CSP particles without further size selection. In the present work, we report the chiral recognition characteristics of the synthesized CSP and its applications in HPLC analytical or preparative scale enantioseparations. 2. Experimental 2.1. Chemicals Microcrystalline cellulose (Avicel, Merck, Germany) was purchased from Fluka (Buchs, Switzerland). 3,5- Dimethylphenyl isocyanate and 3-aminopropyltriethoxysilane (APS) were obtained from Aldrich (Milwaukee, WI, USA). Spherical silica gel (5 m, 100 Å, 270 m 2 /g) was purchased from Fuji Silysia (Tokyo, Japan). The racemic compounds of trans-stilbene oxide, benzoin, 1-phenyl-1-ethanol, flavanone, fenoprofen, mandelic acid, naproxen, warfarin, alprenolol, metoprolol, pindolol, propranolol, hydroxyzine, ephedrine, praziquantel and tröger s base were purchased from Sigma (St. Louis, MO, USA) or Aldrich. 3-Butylphthalide was obtained from Shijiazhuang Pharmaceutical Technology Co. (Shijiazhuang, China). -Dimethyl diphenyl hydrogencarboxylate ( -DDB) and fenpropathrin were obtained from Henan Key Laboratory of Fine Chemicals, Henan Academy of Sciences (Zhengzhou, China). Racemate of ranolazine was obtained from National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). The molecular structures of these solutes are presented in Fig. 1. Racemate of triazole antifungal agent (Fig. 2) was obtained from Shanghai Institute of Materia Medica, Chinese Academy of Sciences (Shanghai, China) Preparation of cellulose tris(3,5-dimethylphenylcarbamate) CSP-100 Preparation of aminopropylsilylated silica gel. Silica gel (Fuji, 6 g) was dried in a vacuum oven at 120 C for 3 h. The silica gel was placed into a 250-mL round bottom flask equipped with a condenser. A mixture of 5 ml of APS and 100 ml of anhydrous toluene was then added into the flask. The suspension was allowed to react at about 100 C for 12 h. The reaction mixture was then cooled to room temperature and the product was collected on a fritted filter funnel (<5 m) by vacuum filtration. The resulting white solid was washed successively with toluene, methanol and diethyl ether. The solid was air dried for 1 h then further dried under vacuum to a constant weight. Synthesis of CDMPC. Dried cellulose (Avicel, 1 g) was refluxed in 20 ml of anhydrous pyridine for 8 h. After cooling the mixture to room temperature, 3.3 g of 3,5-dimethylphenyl isocyanate was added and the reaction was continued at 95 C for 24 h under stirring. The reaction mixture, a dark amber viscous, mostly homogeneous liquid, was cooled to room temperature. The solution was poured into 200 ml of vigorously stirred methanol and stirred for 2 h. The resulting white solid was collected by vacuum filtration and washed several times with methanol. The product was dried in air, then under vacuum to a constant weight. Elemental analysis (C 33 H 37 N 3 O 8 ) n, found: C 64.86, H 6.08, N 6.79; calculated: C 65.66, H 6.18, N IR spectroscopy of CDMPC showed the following characteristic absorptions: υ(urethane C O) 1750 cm 1, υ(urethane N H) 3310 cm 1, υ(phenyl) 1602 cm 1. Substantially no absorption near 3500 cm 1 due to hydroxyl groups of cellulose was observed. Procedure for coating CDMPC on small-pore silica. Cellulose tris(3,5-dimethylphenycarbamate) (0.75 g) was combined with 25 ml of tetrahydrofuran (THF) in a round-bottomed flask equipped with a magnetic stirrer and stirred until the CDMPC was dissolved. Aminopropylsilylated silica (3 g) was added to the flask and stirred for 3 h. The solvent was slowly removed in a rotary evaporator at room temperature and this process was interrupted when the majority of the THF was removed. Then 150 ml cosolvent of n-hexane/phenol 90/10 (v/v) was added dropwise. The solvent was again removed by rotary evaporation to dryness. The obtained composite CSP particles were then dried under vacuum at ambient temperature overnight. The preparation of CSP of this type has been shown to be very reproducible from one batch to another. Several batchs of CSP gave very similar retention and resolution for a series of tested analytes Enantioseparations in HPLC The synthesized CSP-100 particles were then suspended in a n-hexane/2-propanol 70/30 (v/v) mixture and packed into stainless steel columns (150 mm 4.6 mm i.d. for analytical column and 200 mm 8 mm i.d. for semipreparative column) by a conventional slurry packing method. Packing pressure used was 5000 psi, and a n-hexane/2-propanol 80/20 (v/v) mixture was employed as the displacing solvent. HPLC experiments were performed on a Waters 510 pump (Waters, Milford, MA, USA), a Spectra-200 UV detector (Spectra-Physics, San Jose, CA, USA) and a WDL-95 workstation (National Chromatographic R&A Center, Dalian, China). Chromatographic studies were performed at room temperature

3 120 Y. Liu, H. Zou / J. Chromatogr. A 1178 (2008) ( 22 C). The mobile phase consists of various alcohol modifiers (2-propanol, ethanol and methanol) at different percentages in n-hexane. All solvents used were filtered and degassed in an ultrasonic bath before use. The flow rates for analytical Fig. 1. Molecular structures of the tested racemates. and semipreparative resolution were maintained at 0.5 ml/min and 2.0 ml/min, respectively. UV detection was carried out at 254 nm. The injections for each experiment were made in triplicate.

4 Y. Liu, H. Zou / J. Chromatogr. A 1178 (2008) Fig. 2. Molecular structures of antifungal agent CNSH and voriconazole. 3. Results and discussion 3.1. Evaluation of the synthesized CSP-100 Polysaccharide derivative CSPs have been widely used under the mobile phase of n-hexane/alcohol (normal phase mode), and they are also useful in combination with aqueous-organic (reversed phase mode) and purely organic (polar organic mode) mobile phases. Here we report the enantiomeric separation using the synthesized cellulose-based CSP in the normal-phase HPLC mode. In order to obtain a broad perspective on the performance of the synthesized CSP, the resulting CSP-100 was examined in the chiral separation of different kinds of compounds (neutral, acidic and basic analytes). Fig. 3 shows chromatograms of eight analytes eluted with n-hexane/alcohol mixtures from a 150 mm 4.6 mm i.d. column packed with CSP-100. The peaks shown in the chromatograms were quite well resolved in a reasonable period of time. Also Table 1 displays retention factors, selectivity factors, resolution, and plate counts of the racemic mixtures shown in Fig. 1 obtained with the same chiral column. Among the 20 randomly chosen racemic analytes, 18 analytes were baseline-separated. Selectivity factors ranged from 1.16 to 3.43, depending on the analyte. Of particular interest was the high column efficiency associated with CSP prepared by this approach. Plate counts of the first eluted peaks for most tested analytes were larger than 30,000 m 1, which were greater than in literature reports for CSPs obtained by other methods [11 14] and compared favorably to those observed for many commercially available achiral columns of similar dimensions. This could be attributed to the monodispersity of thus synthesized CSP-100 particles. As it is known that resolution between the pair of enantiomers depends on both separation factor and column efficiency, higher column efficiency would allow reduction of column lengths and thus decrease analysis time while adequate resolution is maintained. In addition, the resulting CSP-100 exhibited good flow-through properties. The backpressure on a 150 mm 4.6 mm i.d. column was only 0.86 MPa at a flow-rate of 0.5 ml/min. To discern the efficiency of the synthesized cellulose-based stationary phase comprising a small-pore silica support, we also compared the results obtained with our coated CSP-100 to those reported for the corresponding reference CSPs [13 22], using the same chromatographic mobile phases. As listed in Table 1, higher separation factors as well as higher resolution factors were observed on CSP-100 compared with the CSP developed by Matlin and coworkers [13,14], though the retention factors were different for both CSPs. Compared to the reported data on the commercial Chiralcel OD and Chiralcel OD-H columns on the other hand [15 22], the retention factors were markedly higher on CSP-100 whereas the separation factors were slightly lower on this CSP. However, the resolution factors on CSP-100 were superior or equal to those obtained on the commercial

5 122 Y. Liu, H. Zou / J. Chromatogr. A 1178 (2008) Fig. 3. HPLC enantioseparation examples on a 150 mm 4.6 mm i.d. column packed with CSP-100. Chromatographic conditions were as indicated in Table 1. columns for most analytes tested. This might be due to the higher column efficiency of CSP-100. It is worthwhile noting that the resolutions observed for acidic compounds such as naproxen, warfarin and mandelic acid with the synthesized CSP-100 were significantly greater than in literature reports for commercial columns. These results indicate that the stationary phase developed in this study provides chiral recognition ability fully comparable to the corresponding commercial columns prepared using conventional large-pore silica as the support Applcation to the semipreparative enantioseparation of a new triazole antifungal agent Preparative chromatography is becoming widely used for the purification of enantiomers of synthesis intermediates, at least at the drug development stage, mostly because of the fast method-development and the low cost of this method compared to those of the more conventional approach of designing a steroselective organic synthesis. Voriconazole, (2R,

6 Y. Liu, H. Zou / J. Chromatogr. A 1178 (2008) Table 1 Chromatographic data obtained from HPLC analysis using laboratory-made CSP-100 in comparison to the CSPs described in literature Racemates Mobile phase CSP-100 a Chiralcel OD-H or Chiralcel OD b Small-pore CSPs (Matlin and coworkers) c k 1 α R s N 2 k 1 α R s Ref. d k 1 α R s Neutral racemates Benzoin A [15] Flavanone A [15] Phenyl-1-ethanol B [3] trans-stilbene oxide A [15] e Fenpropathrin C [16] 3-Butylphthalide B Acidic racemates Mandelic acid D [17] Warfarin E [18] Naproxen F [18] Fenoprofen G na f [19] -DDB D Basic racemates Tröger s base A [20] Propranolol H [21] Alprenolol H [21] Metoprolol H [18] Pindolol I [18] Praziquantel J [18] Hydroxyzine J na [22] Ephedrine H [19] Ranolazine I Mobile phases: (A) n-hexane/2-propanol (90/10, v/v); (B) n-hexane/2-propanol (98/2, v/v); (C) n-hexane/2-propanol (95/5, v/v); (D) n-hexane/ethanol/trifluoroacetic acid (90/10/0.1, v/v/v); (E) n-hexane/ethanol/acetic acid (80/20/0.1, v/v/v); (F) n-hexane/2-propanol/trifluoroacetic acid (95/5/0.1, v/v/v); (G) n-hexane/2- propanol/trifluoroacetic acid (80/20/0.1, v/v/v); (H) n-hexane/2-propanol/diethylamine (80/20/0.1, v/v/v); (I) n-hexane/ethanol (80/20, v/v); (J) n-hexane/2-propanol (80/20, v/v). Flow-rate: 0.5 ml/min; UV detection at 254 nm; room temperature ( 22 C). a CSP-100 was CDMPC-coated to 5 m, 100 Å aminopropylsilica gel (column size: 150 mm 4.6 mm i.d.). b Commercial Chiralcel OD-H and Chiralcel OD were CDMPC-coated to 5 mor10 m, large pore aminopropylsilica gel (column size: 250 mm 4.6 mm i.d.). c Small-pore CSPs (Matlin and coworkers) were CDMPC-coated to 5 m or3 m, 120 Å aminopropylsilica gel (column size: 150 mm 4.6 mm i.d. or 100 mm 4.6 mm i.d.). The chromatographic data for small-pore CSPs (Matlin and coworkers) are taken from Refs. [13,14]. d The data source for commercial Chiralcel OD-H and Chiralcel OD. e Not mentioned. f Not available. 3S)-2-(2,4-difluorophenyl)-3-(5-fluoro-4-pyrimidinyl)-1-(1H- 1,2,4-triazole-1-yl)-2-butanol, has recently been introduced by Pfizer as an enantiomerically pure antifungal drug. Drug candidate CNSH (Fig. 2), a structural analogue of voriconazole, is being developed at Shanghai Institute of Materia Medica. Its molecular structure cannot be completely disclosed for proprietary reason. Since chiral compounds can have distinct pharmacokinetic, pharmacological and toxicological properties, they must be characterized individually if racemates are administered or if inter-conversion of stereoisomers is possible. Therefore, the preparative separation of the two pure diastereomers was required to evaluate individually their therapeutic actions. In this study, the semipreparative enantioseparation of the drug candidate CNSH on the synthesized CSP-100 was developed. The choice of an appropriate eluent is important in the development of a preparative separation. In order to evaluate the amount of racemic compound CNSH to apply on the synthesized CSP at semipreparative level, its solubility in different solvents was examined. The findings of this study revealed a low solubility (<3 mg/ml) of compound CNSH in water or isopropanol, and a high solubility (>20 mg/ml) in ethanol. Thus, a n-hexane ethanol 60/40 (v/v) mixture was selected as the mobile phase. This condition was extremely attractive for mg-scale enantioseparations not only for the relative simplicity of the evaporation of eluent but also for a high enantioselectivity toward the target analyte. The semipreparative HPLC enantioseparations were carried out on a 200 mm 8 mm i.d. column packed with the synthesized CSP-100 and two fractions were collected. Some chromatograms of sample injections from 0.2 mg to 10 mg are presented in Fig. 4. It is worth noting that a complete separation of the two enantiomers could be obtained even at the loading amount as high as 10 mg racemate. Total times for a separation run was <20 min. Each collected fraction was combined and dried by rotary evaporation. The individual fractions were then analysed on an analytical chiral column. The analytical assessment of the enantiomeric excess values of the first and second eluted enantiomers, respectively, showed a purity of 99.5% ee and 98.5% ee. The yields of the first and second eluted enantiomers were 90% and 85%, respectively. Polarimetric analysis indicated that the first eluted enantiomer on CSP-100 rotated polarized light in

7 124 Y. Liu, H. Zou / J. Chromatogr. A 1178 (2008) Fig. 4. Loading study of compound CNSH on a 200 mm 8 mm i.d. column packed with CSP-100. Mobile phase: n-hexane/ethanol (60/40, v/v); flow-rate: 2.0 ml/min; UV detection at 254 nm; room temperature ( 22 C). the negative direction in ethanol solution, at a wavelength of 589 nm. 4. Conclusions A two-step coating/precipitation synthetic procedure has been developed for the preparation of cellulose tris(3,5- dimethylphenylcarbamate) chiral stationary phase having a small-pore silica support. With this synthetic strategy, monodisperse, spherical CSP particles can be produced in good yield without the need for wasteful and time-consuming sieving process. The performance of the synthesized CSP towards different kinds of racemic analytes was evaluated in the normal-phase HPLC mode. HPLC separation experiments revealed that the synthesized CSP showed high column efficiency and afforded chiral recognition ability fully comparable to the corresponding commercial columns prepared using conventional large-pore silica as the support. Moreover, the synthesized CSP was successfully applied to semipreparative enantioseparation of a new triazole antifungal agent. Apart from cellulose tris(3,5-dimethylphenylcarbamate) CSP described here, other CSPs based on polysaccharide derivatives as chiral selectors can be also prepared by this robust synthetic approach. Acknowledgement We are grateful to Professor Liren Chen (Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences) for his valuable assistance. References [1] B. Chankvetadze, C. Yamamoto, M. Kamigaito, N. Tanaka, K. Nakanishi, Y. Okamoto, J. Chromatogr. A 1110 (2006) 46.

8 Y. Liu, H. Zou / J. Chromatogr. A 1178 (2008) [2] Y. Okamoto, M. Kawashima, K. Hatada, J. Am. Chem. Soc. 106 (1984) [3] Y. Okamoto, M. Kawashima, K. Hatada, J. Chromatogr. 363 (1986) 173. [4] Y. Okamoto, E. Yashima, Angew. Chem. Int. Ed. 37 (1998) [5] T. Zhang, C. Kientzy, P. Franco, A. Ohnishi, Y. Kagamihara, H. Kurosawa, J. Chromatogr. A 1075 (2005) 65. [6] T. Zhang, D. Nguyen, P. Franco, T. Murakami, A. Ohnishi, H. Kurosawa, Anal. Chim. Acta 557 (2006) 221. [7] X.M. Chen, Y.Q. Liu, F. Qin, L. Kong, H.F. Zou, J. Chromatogr. A 1010 (2003) 185. [8] A. Ghanem, H. Hoenen, H.Y. Aboul-Enein, Talanta 68 (2006) 602. [9] X.M. Chen, F. Qin, Y.Q. Liu, X.D. Huang, H.F. Zou, J. Chromatogr. A 1034 (2004) 109. [10] A. Ghanem, H.Y. Aboul-Enein, J. Liq. Chromatogr. Rel. Technol. 28 (2005) [11] S.J. Grieb, S.A. Matlin, J.G. Phillips, A.M. Belenguer, H.J. Ritchie, Chirality 6 (1994) 129. [12] Y.Q. Liu, W.J. Lao, Y.H. Zhang, S.X. Jiang, L.R. Chen, Chromatographia 52 (2000) 190. [13] S.J. Grieb, S.A. Matlin, A.M. Belenguer, H.J. Ritchie, J. Chromatogr. A 697 (1995) 271. [14] Q.B. Cass, A.L. Bassi, S.A. Calafatti, S.A. Matlin, M.E. Tiritan, L.M. Moreir de Campos, Chirality 8 (1996) 143. [15] Application Guide for Chiral Column Selection, third ed., Daicel Chemical Industries, Tokyo, [16] Z.Y. Li, Z.C. Zhang, Q.L. Zhou, Q.M. Wang, R.Y. Gao, Q.S. Wang, J. AOAC Int. 86 (2003) 521. [17] Y. Okamoto, R. Aburatani, Y. Kaida, K. Hatada, Chem. Lett. (1988) [18] C. Perrin, V.A. Vu, N. Matthijs, M. Maftouh, D.L. Massart, Y. Vander Heyden, J. Chromatogr. A 947 (2002) 69. [19] N. Matthijs, C. Perrin, M. Maftouh, D.L. Massart, Y. Vander Heyden, J. Chromatogr. A 1041 (2004) 119. [20] X.M. Chen, C. Yamamoto, Y. Okamoto, J. Chromatogr. A 1104 (2006) 62. [21] Y. Okamoto, R. Aburatani, K. Hatano, K. Hatada, J. Liq. Chromatogr. 11 (1988) [22] A. Bielejewska, K. Duszczyk, J. Zukowski, J. Chromatogr. A 1083 (2005) 133.