D. HUANG et al., Raceic -cyclohexyl-andelic Acid Resolution across, Che. Bioche. Eng. Q. 22 (4) 447 452 (2008) 447 Raceic -cyclohexyl-andelic Acid Resolution across Hollow Fiber Supported Liquid Mebrane D. Huang, a K. Huang, a,* S. Chen, b S. Liu, a and J. Yu a a School o Cheistry and Cheical Engineering, Central South University, Changsha 410083, China b Key Laboratory o New Material and Technology or Package, Hunan University o Technology, Zhuzhou 412008, China Original scientiic paper Received: Septeber 30, 2007 Accepted: Septeber 29, 2008 This paper deals with the resolution o raceic -cyclohexyl-andelic acid containing copper(ii) N-dodecyl-(L)-hydroxyproline (CuN 2 ) as a chiral carrier across hollow iber supported liquid ebrane. A atheatical odel o transport and resolution o chiral copounds was deduced, the observed partition coeicient between the eed phase and the ebrane phase, the stripping phase and the ebrane phase, ass transer resistance o boundary layer in strip phase inside the hollow ibers, boundary layer in eed phase and ass transer resistance o the ebrane phase are taken into account in the odel equations. Using the experiental results, several paraeters o the proposed odel have been achieved by a nonlinear itting ethod. It is a siply atheatical odel which can be easily used to predict the concentration o the enantioers and the separation actor o the resolution process. Key words: Resolution, hollow iber supported liquid ebrane, -cyclohexyl-andelic acid, odel Introduction Over recent years, attention has increased in the use o supported liquid ebrane (SLM) as selective separation barriers. 1 3 A SLM achieves resolution through a higher ainity o the liquid ebrane or a particular enantioer, it oers any advantages such as high selectivity, high eiciency o separation, high enrichent eiciency, iniu product containation, low cost, no phase separation proble, easy scale up option or coercial applications and less use o the organic phase than in the classical solvent extraction process. 4 6 There are three conigurations available or a SLM: lat sheet SLM, hollow iber SLM and spiral wound SLM. Hollow iber SLM containing active chiral carriers provide a siple, stable eans or studying the transport and selective separation o chiral drugs ro dilute aqueous solutions. The ost requent applications o hollow iber liquid ebrane systes or raceic resolution have been in the concentration and selective resolution o chiral aino acids. 7 9 Hollow iber SLMs are usually prepared by illing the pores o these ebranes with the organic liquid coposed o the chiral carrier and the diluent. This siple geoetry, which is econoical or research purposes, uses a very sall * Corresponding author: Fax: +86-731-8879850; E-ail: klhuang@ail.csu.edu.cn aount o organic solution to transport large aounts o drugs. Optical pure -cyclohexyl-andelic acid (HCHMa) is a kind o iportant precursor o chiral drugs and can be used to synthesize ultiplicate chiral drugs with vital biological activity and well curative eect such as oxybutynin. 10 Chiral extraction or resolution o enantioers has great potentialities, and it has been highly regarded in recent years. 11 13 So it is necessary to resolute raceic HCHMa into R- and S-enantioers. In this work, a hollow iber supported liquid ebrane was used to resolute the raceic HCHMa and a new atheatical odel was presented or analyzing the transport o enantioers. It provides a new way to resolute the raceic HCHMa and a siply atheatical odel which can be easily used to siulate and predict the concentration o the enantioers and the separation actor o the resolution process. Theory In this paper we propose a echanis or the transport o HCHMa enantioers through a hollow iber SLM. Consider a porous carrier-acilitated SLM used or separating R- and S--cyclohexyl-andelic acid o the raceic ixtures. The ebrane consists o a chiral carrier dissolved in a water iiscible organic diluent. 9. sijełanj 2009 10:11:53 67
448 D. HUANG et al., Raceic -cyclohexyl-andelic Acid Resolution across, Che. Bioche. Eng. Q. 22 (4) 447 452 (2008) In order to sipliy the atheatics and odel developent, the ollowing assuptions are ade: 1. An ideal syste exists under coplete ixing and constant teperature operation; 2. Constant physical and transport properties; 3. The ebrane phase is copletely iiscible with the aqueous phase; 4.The volue o the liquid ebrane phase is neglected relative to the volue o the eed phase and the strip phase. 5. The observed partition coeicient include (a) partition o the enantioers and (b) cobination o the enantioers with the carrier. Solute partition equilibriu A ass balance to the hollow iber odule and vessel can be used to deterine the change in solute concentration with tie: V d j dt KA * j j ( ) (1) Were K is the overall ass transer coeicient, V represents the aqueous phase volue, A is the ebrane transer area, and is the solute ass concentration. The subscripts j reers to the enantioers o the chiral drugs, the subscripts reers to the eed phase (shell side), and the superscript * reers to the solute concentration in the aqueous phase, in equilibriu with the solute concentration in the stripping phase (tube side). I it is assued that the solute partition between the two phases is adequately described by a constant observed partition coeicient P then, integration o eq. (1) and * j sj j j P2 / P1 (2) Eq. (2) indicates the solute concentration in the aqueous phase, in equilibriu with the solute concentration in the stripping phase, and there is an organic ebrane phase between these two aqueous phases, so the total partition process includes two processes o P 1 and P 2. P 1 is observed partition coeicient between the eed phase and the ebrane phase, P 2 is observed partition coeicient between the stripping phase and the ebrane phase, the observed partition coeicient between these two aqueous phases should be expressed as the ratio o P 1 and P 2. The subscripts s reers to the stripping phase (tube side). The solute concentration in the stripping phase ( sj ), obtained by ass balance: V V V V (3) 0j s s0j j s sj An analytical expression o sj versus t with K as the unknown paraeter can be received as: V s d sj dt V s KA 0j s0j V sj PVPV 1 2 P V (4) j s j 1 j Evaluation o ass transer correlations Three individual ass transer resistances ay be considered in hollow iber resolution process: The boundary layer resistance in the shell side liquid phase; The ebrane resistance to solute diusion across the ebrane pores and the boundary layer resistance in the tube side liquid phase. The overall ass transer resistance, or resolution process using a hollow iber supported liquid ebrane syste, can be expressed as: 1 1 1 P2 j (5) KA ka Pk A PkA j i o 1 j 1j s i Where k, k and k s are the individual ass transer coeicients on the eed phase side, ebrane and stripping phase side, respectively, and A i, A o and A are the ibers internal, external and logarithic ean areas, respectively. k D /( ) (6) 2/ 1 (7) Where,, are the thickness o the ebrane ibers, porosity o the ebrane, tortuosity actor o the ebrane pores which takes into account the dierence between the eective thickness and the physical thickness. D is the ree bulk diusion coeicient or the solute calculated using the Wilke-Chang 14 equation: D74. 10 8 ( M ) 0. 5 T/( V 0. 6 ) B B A (8) In the present work a Lévèque type equation 15,16 will be used to correlate both the tube side and the shell side ass transer coeicients: b c d o Sh a Sc Re ( d / L) (9) b c d s s s s i s s s Sh a Sc Re ( d / L) (10) Where Sh, Sc and Re are the diensionless nubers o Sherwood, Schidt and Reynolds, respectively, a, b, c and d are the unknown paraeters. The diensionless nubers are deined as: Sh k d / D (11) Red u / (12) Sc/( D) (13) Where k, d, u,,, D are the ass transer coeicient, diension, liquid velocity, density, liquid viscosity and diusion coeicient, respectively. 9. sijełanj 2009 10:11:53 68
D. HUANG et al., Raceic -cyclohexyl-andelic Acid Resolution across, Che. Bioche. Eng. Q. 22 (4) 447 452 (2008) 449 Integrate eq. (4), the solute concentration in the stripping phase will be expressed as: P1j( V 0jVss0j) sj P2jV P1jVs (14) P2jVs0j P1jV 0 P V P V j 2j 1j s exp AK jt P2jVP1jV s PV 1 j V s And the expression o separation actor () can be received as: Experiental Materials sr / ss (15) Raceic HCHMa was purchased ro Synergetica Changzhou, Ltd. in China. L- hydroxyproline was obtained ro Xinghu Biology o Science and Technology Zhaoqing Ltd. in China, the speciic optical rotation [ ] 20 D 85.2, the purity was above 98 %. Aldehyde C-12 lauric was purchased ro Fluka Cheical Copany, the purity was above 97 %. Palladiuon on carbon catalyst was ro Shanghai Cheical Reagent Copany in China. The polyvinylidene luoride (PVDF) hollow iber ebrane odules were purchased ro Mo-tian Mebrane Engineering and Technology Co. Ltd., Tianjin, China. The PVDF hollow iber ebrane used in this experient has an eective area o A0095. 2. The coniguration is shown in Table 1. Table 1 Characteristics o PVDF hollow iber ebrane External/Internal diaeter Analytical ethod Teperature range Pore size PH range Operating pressure 1.2/0.8 5~45 C 0.22 2~10 0.12 MPa HPLC was perored with a LC-2010A SHIMADAZU syste controller (Kyoto Japan), a saple loop injector o 20 L, a Shiadzu C-R3A Chroatopac, a Kroasil RP-18, 5, 4.6 150 colun was used or the analysis. Copper ion concentrations in aqueous phase were deterined by voluetric titration with EDTA using 1-(2-pyriclylaze)-2-naphthol (PAN) indicator dye. The copper concentration in the organic phase was deterined by irst diluting the saple with ethanol and then titrating with EDTA using PAN. HN was synthesized in our lab. 17 The concentrations o enantioers in the aqueous were easured by chiral obile phase HPLC. Chroatographic conditions: 18 concentration o -cyclodextrin c = 9.5 ol L 1, V [(aqueous KH 2 PO 4 solution (0.075 ol L 1 )]: V (ethanol): V (acetonitrile) = 65:20:15, ph 4.8, low rate Q = 1Lin 1 and roo teperature. Resolution o HCHMa The experients were carried out on the resolution o raceic HCHMa using copper(ii) N-dodecyl-(L)-hydroxyproline (CuN 2 ) as a chiral carrier in a ebrane solvent (octanol). The optial operating conditions were deterined by chiral extraction studies detailed in Feipeng and Kelong: 19 [Cu 2+ ] = 5 ol L 1 ; [HN] = 10 ol L 1 ; [NaAc] = 0.1 ol L 1 ; ph 4.0. Then, the partition coeicients o S- HCHMa and R- HCHMa are 0.995 and 0.693 respectively. The organic phase was prepared by adding HN (10 ol L 1 ) to octanol, and was then contacted with an equal volue o aqueous copper acetate solution (Cu 2+ : 5 ol L 1 )atph 4.0. The aqueous copper acetate solution was prepared by adding cupric sulate to acetate buer. The organic and aqueous solutions were contacted or 48 h. The ebrane was soaked in the resulting organic phase or at least 3 h. The whole solution was iobilized in the pores o the polyvinylidene luoride polyeric ebrane separating two tube- and shell-side aqueous phases. The solution o chiral selectors in the ebrane pores was iobilized by puping the solution o chiral selectors into the tube-side o the hollow iber ebrane odule. This solution was circulated or 24 h in order to distribute the dissolved CuN 2 olecules into the ebrane pores. The eed and strip solutions were prepared by adding perchloric acid to deionized water until the solution ph was 4.0. A raceic ixture o the HCHMa was dissolved in the eed solution (solute concentration: 10 ol L 1 ). The eed phase was puped into the shell-side while the strip solution into the tube-side. At various ties, 100 L saples were reoved ro the strip solution and the R and S-enantioers concentrations were deterined using HPLC. Results and discussions Experiental result o enantioers concentration and separation actor The experiental data o the enantioers diensionless concentration and separation actor o resolution process are shown by dot in Figs. 1 and 2. It can be seen that both enantioer concen- 9. sijełanj 2009 10:11:53 69
450 D. HUANG et al., Raceic -cyclohexyl-andelic Acid Resolution across, Che. Bioche. Eng. Q. 22 (4) 447 452 (2008) concentration dierence is deterined by the ratio o the observed partition coeicients o the two enantioers. In this case, the axiu is shown at the tie o 20 h. This result can be used as a guideline to deterine the optiu operation conditions or the resolution process. Fig. 1 Diensionless concentration proiles o enantioers in the stripping phase. and experiental data; Solid line: odel prediction o diensionless concentration o the S-enantioer; Dot line: odel siulation o diensionless concentration o the R-enantioer. Model siulation o enantioers concentration proiles There are eight unknown constants o a, b, c and d in the stripping phase and eed phase. According to the research o any workers, 20 22 there are soe sei-experiential orulas about the eqs. (10) and (11), and soe constants o the are the sae: b b 033., d 033., d 1 (16) s s Then integrating eq. (5) to eq. (13), the overall ass transer coeicient will be expressed as a unction o our unknown paraeters: a, a s, c and c s : 1 K j a A i / A c 2 du D 3 L o 033. A A P P i / 2j/ 1j D P a du c s D 2 s 1 j s L d 2 033. (17) Using the experiental data o concentration and separation actor which are shown in Figs. 1 and 2, the our unknown paraeters can be obtained by nonlinear itting ethod as: Fig. 2. Separation actor o the enantioseparation process. experiental data o separation actor; Solid line: results coputed by eq. (15). trations o the strip solution increase rapidly in the irst ew hours, and the S- HCHMa concentration is higher than that o the R- HCHMa. The slope o S- HCHMa is steeper than the slope or R- HCHMa. It is because that the total ass transer ipetus is the highest one at the beginning since that the partition process is ar ro the equilibriu, and the ass transer ipetus o the S- HCHMa is higher than that o the R- HCHMa because that the observed partition coeicient o the S- HCHMa is higher than that o the R- HCHMa.The concentration dierence between the S- and the R- HCHMa in the strip solution shows a axiu. This axiu a 439127., a., c 235., c 033. (18) s s The siulation curves o the enantioers diensionless concentration are shown by line in Fig. 1, and the correlation coeicients o curve siulation o S-enantioer concentration and R-enantioer concentration are 0.996. Model prediction and experiental result o separation actor Substituting the paraeters obtained by nonlinear itting ethod into eqs. (14) and (15), the atheatical odel o separation actor can be obtained. Fig. 2 shows the experiental result and odel prediction o the separation actor o resolution process. It can be seen that the coputational results o separation actor are shown by a solid line which is in good agreeent with the experiental data. Thereore, it can be concluded that the paraeters 9. sijełanj 2009 10:11:54 70
D. HUANG et al., Raceic -cyclohexyl-andelic Acid Resolution across, Che. Bioche. Eng. Q. 22 (4) 447 452 (2008) 451 achieved in eq. (18) can be used to calculate the separation actor o this resolution process. It also indicated that the initial separation actor is the highest one achieved during the run, and it declines rapidly in the irst ew hours, then it approaches to the value o 1 as tie passes. According to the eq. (15), the initial separation actor is deined as: K li a 0 K R S (19) It eans that the ratio o the overall ass transer coeicient or the S- and R- enantioer ay be used to predict the axiu degree o separation that could be expected under ideal conditions. For this case, the initial separation is calculated to be 1.45. Conclusions Resolution o raceic -cyclohexyl-andelic acid containing copper(ii) N-dodecyl-(L)-hydroxyproline (CuN 2 ) as a chiral carrier across hollow iber supported liquid ebrane was carried out successully. A atheatical odel was developed to analyze the ass transer and the separation actor. The observed partition coeicient between the eed phase and the ebrane phase, the stripping phase and the ebrane phase, ass transer resistance o boundary layer in strip phase inside the hollow ibers, boundary layer in eed phase and ass transer resistance o the ebrane phase were taken into account in the odel equations. Using the experiental results o the enantioers concentration, several paraeters o the proposed odel had been achieved by a nonlinear itting ethod. The atheatical odel was used to predict the separation actor o the resolution process and the coputational results o separation actor were in good agreeent with the experiental data. It is a siple atheatical odel which can be used to predict the concentration and the separation actor o the resolution process. ACKNOWLEDGEMENTS This work was supported by the National Nature Science Foundation o China (No.20576142) Noenclature A area, c 2 c enantioer concentration, ol L 1 d diaeter o the hollow iber, D diusion coeicient, c 2 s 1 k ass transer coeicient, c s 1 K overall ass transer coeicient, c s 1 L length o the iber, c P observed partition coeicient Q volue low rate, L in 1 Re Reynolds nuber Sc Schidt nuber Sh Sherwood nuber t tie, h u liquid velocity, c s 1 V volue o aqueous phase, c 3 Greek letters separation actor solute ass concentration, g c 3 density, g L 1 liquid viscosity, cp ebrane thickness, c porosity o the ebrane tortuosity o the ebrane pores Subscripts i j o R S s denotes eed phase tube side o the hollow iber enantioeric or ebrane shell side o the hollow iber contactor R-enantioer o the raceic ixture S-enantioer o the raceic ixture denotes stripping phase Reerences 1. Ata, O. N., Che. Bioche. Eng. Q. 19 (2005) 25. 2. Maxiini, A., Chiel, H., Holdik, H., Maier, M. W., J. Mebr. Sci. 276 (2006) 221. 3. Stankovic, V., Che. Bioche. Eng. Q. 21 (2007) 33. 4. Qian Yang, Tai-shung Chung, J. Mebr. Sci. 294 (2007) 127. 5. Uebereld, J., Parthasarathy, N., Zbinden, H., Gisin, N., Bule, J., Analytical Cheistry 74 (2002) 664. 6. Van de Voorde, I., Pinoy, L., De Ketelaere, R. F., J. Mebr. Sci. 234 (2004) 11. 7. Hadik, P., Szabó, L.-P., Nagy, E., Farkas, Zs., J. Mebr. Sci. 251 (2005) 223. 8. Clark, J. D., Han, B. B., Bhown, A. S., Wickraasinghe, S. R., Sep. Puri. Technol. 42 (2005) 201. 9. Hadik, P., Kotsis, L., Eniszné-Bódogh, M., Szabó, L. P., Nagy, E., Sep. Puri. Technol. 41 (2005) 299. 10. Prelog, V., Stojanac,., Kovaèeviæ, K., Helv. Chi. Acta 65 (1982) 377. 11. Zhongzhe, C., Shuihong, C., Journal Cheical Industry and Engineering (China) 51 (2000) 418. 12. Prelog, V., Kovaèeviæ, M., Egli, M., Angew. Che. 28 (1989) 1147. 9. sijełanj 2009 10:11:54 71
452 D. HUANG et al., Raceic -cyclohexyl-andelic Acid Resolution across, Che. Bioche. Eng. Q. 22 (4) 447 452 (2008) 13. Jérôe, L., Catherine, G. G., Sonya, T. H., Jonathan, J. J., Angew. Che. 39 (2000) 3695. 14. Reid, R. C., Prausnitz, J. M., Sherwood, T. K., The properties o Gases and Liquids, 3rd edn., McGraw-Hill, New York, 1977. 15. Dahuron, L., Cussler, E. L., AIChE J. 34 (1988) 130. 16. Prasad, R., Sirkar, K. K., AIChE J. 34 (1988) 177. 17. Ding, H. B., Carr, P. W., Cussler, E. L., AIChE J. 38 (1992) 1493. 18. Shanshan, H., Yizu, W., Meiren, Sh., Fine Cheical. 21 (2004) 731. 19. Feipeng, J., Kelong, H., Xia, Y., Xuehui, Zh., Jingang, Y., J. Bio. Sci. 9 (2006) 1149. 20. Ortiz, I., Galan, B., Roan, F. S., Ibanez, R., AIChE J. 47 (2001) 895. 21. Gawronski, R., Wrzesinska, B., J. Mebr. Sci. 168 (2000) 213. 22. Wu, J., Chen, V., J. Mebr. Sci. 172 (2000) 59. 9. sijełanj 2009 10:11:54 72