Microsoft Word - Contents.doc HUNGARIAN JOURNAL OF INDUSTRIAL CHEMISTRY VESZPRÉM Vol. 34. pp. 1-14 (2006) SIMULATED MOVING BED (SMB) SEPARATION OF PHARMACEUTICAL ENANTIOMERS AND CRYSTALLIZATION G. GÁL1 , L. HANÁK1, J. ARGYELÁN1, A. STRBKA1, ZS. HERSECZKI, T. SZÁNYA1 A. ARANYI2, K. TEMESVÁRI2 and Z. HORVÁTH2 1University of Pannonia, Department of Chemical Engineering Processes H-8201 Veszprém, PO Box 158, HUNGARY, E-mail: gaborgal@freemail.hu 2Gedeon Richter LTD., H-1475 Budapest PO Box. 27, HUNGARY A promising solution for the separation of pharmaceutical enantiomers is the method of SMB liquid chromatography (SMB-LC). An optimised coupling of SMB-LC and crystallization processes can improve the efficiency of enantioseparation. In the first process both enantiomers can be produced in enriched concentrations in extract and raffinate streams of simulated moving bed chromatograph suitable for crystallization. Chromatographic investigations include the experimental determination of adsorption isotherms, bed voidage, NTP, HETP etc. on a suitable chiral stationary phase as well as the simulation and optimization of a the appropriate SMB-LC process. Mathematical models and joined computer programmes were published in our earlier papers had been developed and applied for the calculation of SMB-LC. We assumed isotherm, isochor equilibrium adsorption (competitive multicomponent Langmuir-adsorption equilibrium) in the mathematical model neglecting the effects of axial dispersion. The mathematical model was solved by finite differences, numerical mathematical method using PC. The SMB-LC separations were carried out on a laboratory scale (I.D. = 1 cm, L = 25 cm) four-column open loop eluent system equipment at 1:1:1:1 column configuration. The process variables of the SMB-LC (product purity, yield, productivity, specific solvent consumption) are favourable. The efficiency of the enantiomer separation can be increased by coupled crystallization of the enriched extract and raffinate fractions resulting >99% (w/w) pure enantiomer crystals. „Mother liquids” recirculation from crystallization to the SMB- LC process increases with great benefit the economical parameters of enantiomer separation. Keywords: preparative liquid chromatography, simulated moving bed chromatography, pharmaceutical enantiomers, chiral chromatographic packing, crystallization Introduction Nowadays more than half of the registered medicals have got chiral structures owing special importance in pharmaceutical industry. From pharmaceutical points of view only one enantiomer has got good biological activity, meanwhile the other enantiomer is inactive or toxic. The conclusion is, that the enantiomer purity has determining effect. The classical resolution can be well applied for the production of optically active compounds. For example optically active compounds can be produced by stochiometric catalytic asymmetric synthesis starting from achiral compounds. Enantiomers produced by synthesis can be separated by simulating moving bed liquid chromatography (SMB-LC) process. This last process has growing importance in pharmaceutical industry as it is well applicable in case of the high purity separation of wide scale of racemic mixtures. The advantages of SMB-LC compared to the traditional batch chromatography are: continuous process, constant product purity for both enantiomers, high yields, high productivities, low specific solvent consumptions. The chromatographic packing (CSP) is used in full columns length, so the productivity and yield is high in case of SMB-LC compared to batch chromatography. Disadvantages of SMB-LC are: high investment costs and process parameter sensitivity. In analytical, preparative and SMB liquid chromatography the polysaccharide derivatives proved to be versatile among industrial chiral stationary phases (CSP). The productivity of SMB-LC in case of these CSP packings are as high as 0.2-2 kg racemic mixture/kg packing/day. The coupled crystallization to SMB-LC provides possibility to impurity removal, thus the valuable enantiomers can be achieved with more than 99% (w/w) purity [1-5]. The next crystallization methods can be applied: resolution by entrainment, separation via formation of diastereomer salts and crystallization from optically active solvents. The triangle solubility diagram of two enantiomers/one solvent system presents the possibility of enantiomer enrichment. The yield of crystallization process depends on the above mentioned triangle solubility diagram data and the inlet liquid concentration. This publication is on the SMB-LC and the coupled crystallization of enantiomers. In the first step SMB-LC process is used for enantiomer enrichment, then in the second step pure enantiomer is produced by crystallization. Valuable 2 enantiomer crystallization mother liquid after evaporation was recirculated to the feed of SMB-LC. Determination of optimal concentration of the valuable enantiomer coming from SMB-LC is an important task during the planning of hybrid process (SMB-LC, liquid evaporation, crystallization with cooling, mother liquid recirculation, etc.). Enantiomer separation by chiral liquid chromatography The first and significant step in planning of chiral chromatographic enantiomer separation is to determine the proper moving and stationary phases can be made by screening experiments in analytical scale. These data are used for SMB-LC mathematical models, calculations helping previous planning and size increase of SMB-LC process. Fundamentals of SMB-LC chromatography Nowadays the SMB-LC process is widely used and well applied in various separation problems. The large scale SMB-LC process was developed by UOP in 1960 and since this time it has gone over tremendous development. Since 1980 it has been used especially in fine chemical industry and for the last decade in the field of pharmaceutical industry. On Fig. 1. the classical four column open eluent loop SMB-LC process is presented. The SMB-LC chromatographic process is a multi column system with two inlet (fresh eluent and feed of enantiomers) and three outlet (extract, raffinate and outlet liquid) streams where the liquid and solid phases are moving in counter current direction. The counter current stream of liquid and solid phases is not real, but simulated as the stationary phase is moving by columns at each switching time. Simulated solid phase movement is carried out at the switching time by the proper periodical change of inlet and outlet points of the equipment. These points divide the SMB-LC equipment for four zones (I, II, III, IV). The enantiomer solution (A, B) for separation is fed (F = feed) into the equipment amidst II and III zones. As the component A adsorbs stronger on CSP than component B, the previous is moving rather with the solid phase while the other one is moving with the liquid phase in the SMB-LC equipment. Components A and B in zones II and III separate from each because of their different adsorption affinity and the resulted pure A and B components can be got in extract and raffinate streams respectively (Fig. 1) Solid phase regeneration happens in zone I with fresh eluent, while in zone IV the eluent regeneration is carried out by retaining the less adsorbing component B. Raffinate, R Extract, E Fresh eluent, D Feed, F B A+B A IV III II I LR OUT Fig. 1: Simulated moving bed (SMB) adsorber: I, II, III, IV-zones, respectively HPLC columns; D-desorbent (solvent, eluent); E-extract stream with the stronger adsorbed component A; F-feed stream with the components A and B; R-raffinate stream with the less adsorbed component B; LR OUT- outlet liquid. Working points and the separation range of SMB-LC processes [6, 7] The main task in SMB-LC planning is the proper choice of working conditions. Obviously it means the determination of relative volumetric streams of liquid and solid phases in each zones. Five parameters must be determined in a two component mixture separation task using SMB-LC equipment. These are the liquid volumetric streams in four zones and the so called switching time. Morbidelli and co-workers significantly contributed to the handling of this non trivial planning problem. According to the aboves Morbidelli theory or triangle was used for the initial parameters calculation of SMB-LC. Assume the next component balance equation for equilibrium adsorption without component indexes. ( ) 01 =⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛ ∂ ∂ +⎟ ⎠ ⎞ ⎜ ⎝ ⎛ ∂ ∂ −+⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛ ∂ ∂ z f z f t f t c A t q A z c B εε 3 Where Bf = volumetric velocity (cm 3/min), c = concentration in liquid phase (mg/cm3), z = axial distance (cm), t = time (min), ε = bed voidage (cm3 liquid/cm3 column), Af = column cross section (cm2), q = concentration in solid phase (mg/cm3). Assuming linear adsorption equilibrium isotherm: q=K · c The linear velocity of liquid element with c concentration, uc = (cm/min) after the de Vault equation: ( )K A B u t z f f c c εε −+ ==⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛ ∂ ∂ 1 The column is packed with granular solid adsorbent, which does not fill the full volume. The column can be characterised by the bed voidage (ε). In our case “A” and “B” two component mixture was investigated with KA and KB adsorption equilibrium distribution coefficient. In our case KA >KB, K measure unit is (g compo- nent/cm3 solid)/(g component/cm3 liquid). Determination of the Morbidelli parameters [6,7] Let examine the III zone of the SMB-LC equipment. Presume, that the adsorbent have neither “A” nor “B” components and during “T” switching time “A” and “B” components are fed into the III zone. As the volumetric velocity in the given III zone is D-E+F, so the velocities of “A” and “B” components in III zone are as follows (see Fig. 1): A f A K A FED u )1( εε −+ +− = B f B K A FED u )1( εε −+ +− = The length of the III zone is L (cm), the cross section of the column is Af. “A” is not allowed to run out of the III zone, but “B” must leave it. A f B K T L A FED K )1()1( εεεε −+< +− <−+ A f B KL LT A FED K < − − +− < )1( ε ε AIIIB KmK << mIII – Morbidelli parameter, relative velocity for the III zone The next can be written for the II zone of the SMB- LC equipment: “B” component must leave the II zone and “A” is not allowed to get through: A f B K T L A ED K )1()1( εεεε −+〈 − 〈−+ A f B KL LT A ED K 〈 − − − 〈 )1( ε ε AIIB KmK 〈〈 mII – Morbidelli parameter, relative velocity for the II zone In the I zone of the SMB-LC equipment no “A” can be remained, so regeneration must be perfect. A f I KL LT A D m 〉 − − = )1( ε ε mI – Morbidelli parameter, relative velocity for the I zone “B” component is not allowed to leave the IV zone of the SMB-LC equipment. B f IV KL LT A RFED m 〈 − − −+− = )1( )( ε ε mIV – Morbidelli parameter, relative velocity for the IV zone These conditions are necessary for the separation of a two-component “A”, “B” mixture for pure “A” and “B” components. Summary KA < mI KB < mII < KA KB < mIII < KA mIV < KB 4 Fig. 2: Morbidelli triangle in case of linear adsorption isotherms assuming independent adsorption: Region 1: pure “A” and pure “B”, Region 2: pure “B” in raffinate, impure “A” in extract, Region 3: pure “A” in extract, impure “B” in raffinate, Region 4: impure “A” in extract, impure “B” in raffinate. Theoretical analysis in case of non linear adsorption isotherms [6,7] If linear adsorption equilibrium isotherms and independent adsorption conditions are not existing during the theoretical description of equilibrium adsorption described in the previous chapter, than the Morbidelli triangle modifies. Fig. 3: Modification of Morbidelli triangle in case of competitive Langmuir isotherms It can be seen from the Figure 3, that the modified Morbidelli triangle is also divided into four regions: Region 1: pure “A” and pure “B”; Region 2 pure “B” in raffinate, impure “A” in extract; Region 3: pure “A” in extract, impure “B” in raffinate; Region 4: impure “A” in extract, impure “B” in raffinate Let the two component competitive Langmuir adsorption isotherm be valid: 22111 cbcb ca q iii ++ = 2,1=i Where: qi – solid phase concentration (mg/cm 3 adsorbent) ci – liquid phase concentration (mg/cm 3) ai – Langmuir constant (cm 3 liquid/cm3 adsorbent) bi – Langmuir constant (cm 3 liquid/mg “A” or “B” component). The conditions of simultaneous production of pure “A” and “B” components are as follows: mI > KA mII,min (mII, mIII) < mII < mIII < mIII,max (mII, mIII) mIV < mIV, Kr [ [ ] IIIBIIIIIFBBIIIB IIIII F BBIIIBKrIV mammcbmK mmcbmKm 4)( )( 2 1 2 , −−++ −−++= where F means the feed of SMB-LC equipment (cm3/min). Explanation of Fig 3: Point a KA = KA Point b KB = KB Point f ωG = ωG Point r [ ] ⎥ ⎦ ⎤ ⎢ ⎣ ⎡ − −+−− )( )()()( , 2 FABA FAGBBAGAFG A G KKK KKKKK K ω ωωωωωω Point w [ ] ⎥ ⎦ ⎤ ⎢ ⎣ ⎡ − −+− )( )()( , FAB FBBBAFG A GB KK KKKK K K ω ωωωω Where ωF and ωG (ωG > ωF > 0) are the roots of equation given below (1 + bAcA F + bBcB F)ω2 – [KA(1 + bBcB F) + +KB(1 + bAcA F)]ω + KAKB = 0 It can be seen on Fig. 3 that the shape and area of Morbidelli triangle significantly changes caused by the competitive Langmuir isotherm affected by feeding total and component concentrations, real effects as adsorption kinetics (component transfer resistances), axial mixing phenomena, column packing efficiency (Fig. 4). 5 Fig. 4: Feeding concentration change on Morbidelli triangle in case of racemic mixture (c0 F < c1 F < c2 F < c3 F < c4 F, c0 F ~ 0 g/dm3) Increasing the feeding concentration the triangle area decreases and deforms at constant Langmuir adsorption equilibrium parameters. It is unfavourable for enantiomer separation. The mathematical model of SMB-LC From the above equations, it can be seen that the pure extract-pure raffinate area is smaller by higher feed concentration than the rectangular triangle, consequently we have less possibility by parameter variation. The Morbidelli's area changes, when the following possibilities of intervention vary during the planning: temperature, changing adsorbent, and composition of the solvent (eluent). When we have already selected the chromatographic packing for the separation, the next task is to choose operating variables: fresh eluent, recirculated eluent, feed, and extract, raffinate flow rates, switching time. During calculation, the influence of a given parameter is to be investigated, the others must be considered as constants. Better productivity, product purity, yield, and eluent consumption can be achieved by optimizing operating conditions by computer with the exact mathematical model of the SMB. The basic equation of the mathematical model can be deduced from the component balance of the solid–liquid equilibrium system [8, 9]. The equal balance for the component “k” is: 0)()()1()(0 = ∂ + ∂ ∂ −+ z k z k t k t c t q z c v ∂ εε ∂ ∂ Where v0 = velocity of fluid phase (cm/min), dck/dz = place derivative of liquid concentration of component k, dck/dt = time derivative of liquid concentration of component k, and dqk/dt = time derivative of solid concentration of component k. When we discuss a two-component equilibrium system the competitive multi-Langmuir type isotherm can be written as: 2211 11 1 1 cbcb ca q ++ = 2211 22 2 1 cbcb ca q ++ = The denominator of this isotherm was replaced by “N”: 22111 cbcbN ++= Considering the equal balance and concentration (c1, c2) derivatives of the above Eq. can be written: t c N cba N cbaNa t c z c N vo ∂ ∂ − + − ∂ ∂ = ∂ ∂ − − 2 2 121 2 1111112 ) 1 ( 1 ε ε ε The derivatives were substituted with difference quotients: )() 1 ( 1 2 2 1 111 2 1 112 t c b t c bcaNNa t c z c N vo Δ Δ + Δ Δ − − + Δ Δ = Δ Δ − − ε ε ε Difference Eq. above was rearranged and written for component “k”: )) 1 ( 1 2 1 2 kkkk kk n k kkk ko cbaNNa t c t c bca z c N v − − + Δ Δ = Δ Δ + Δ Δ − − ∑ = ε ε ε The above Eq. is the base of the numeric simulation software. The component transfer is calculated with this Eq. between the so-called equilibrium cascades. In our model, the number of equilibrium cascades are compliant with the number of theoretical plates. Based on the mathematical model, the KROM-N and SMB KROM-N computer programs were written in Delphi computer language [10]. Frontal adsorption experiments and SMB-LC measurements can be simulated by the computer programs. At 200 NTP/column the calculation time on a 3 GHz personal computer is 0,6% of the measurement time. The input data of the software: number of components, feed concentration, adsorption equilibrium properties (Langmuir parameters), column geometrical data (average) NTP/column, total porosity, bulk density, volumetric velocities (eluent, feed, extract, raffinate, recirculation). The initial SMB operating conditions were calculated with the equilibrium triangle method. We considered the maximal flow rate of eluent pump, and minimal column switching time was chosen. Simulated moving bed liquid chromatography (SMB-LC) coupling with crystallization [1-5] Enantiomer purification by crystallization is based on their terner solubility diagrams. According to the type of saturation curves in the phase diagram, three fundamental types (conglomerate, racemic compound, pseudo racemate) forming systems can be identified. Only 5-10% of racemates belong to the conglomerate forming group, 6 90-95% belong to the true racemates and the third group pseudo racemates are relatively few. On Fig. 5 the solubility diagram of two, (S)- and (R)- enantiomers and solvent “H” can be seen on equilateral triangle at different temperatures. Fig. 5: The solubility diagram of two enantiomers, (S) and (R) in “H” solvent. (S): less retained enantiomer, (R): more retained enantiomer, H: n-hexane solvent, 1. Raffinate, 2. Evaporated raffinate, 3. Crystallization mother liquid, 4. (S) crystals, TC = -20 °C cooling temperature, TE = 50 °C evaporation temperature, -.-.- limiting evaporated liquid concentration curve of >99% (w/w) (S) crystals. SMB-LC and joined crystallization processes are recommended for enantiomer separation according to Fig. 5. The (S)-enantiomer in required purity can be achieved, if crystallization started from an asymmetric component content. Joined SMB-LC and crystallization process starts at Point 1 (Raffinate) with (S) enrichment. As the raffinate is a diluted unsaturated liquid (Point 1), it must be evaporated to put Point 2 between evaporation and cooling solubility curves. Cooling the liquid (Point 2) forms crystallization mother liquid (Point 3) and pure (S) crystal is resulted (Point 4). SMB-LC, evaporation, cooling crystallization and crystallization mother liquid recycling The joint process can be seen on the Fig. 6, where both enantiomers can be purified by crystallization. In our work only the (S)-enantiomer was crystallized and its crystallization mother liquid was recycled to the feed. In case of (R)-enantiomer only the eluent was recycled. The original (S) and (R) 50-50% (w/w) feed is to be changed because of the recirculation of (S) crystallization mother liquid and thus concentration of (S) increases over 50%. We have to work out an optimization method, which has got an objective function ((S) productivity maximalization, fresh solvent consumption minimalization) at higher than 99% (w/w) (S) purity, and higher than 99% (S) yield. Fig. 6: SMB-LC, evaporator, cooling crystallization and crystallization mother liquor recirculation. 7 Purity and productivity are determined by the next equations in extract and in raffinate. 100⋅ + = RS S cc c Purity BC R S S V cR P ρ4 ⋅ = BC E R R V cE P ρ4 ⋅ = cS – (S) concentration in extract or raffinate streams (g/dm3) cR – (R) concentration in extract or raffinate streams (g/dm3) P – productivity (mg component/g packing/day) R, E – raffinate and extract volumetric streams (cm3/min) Vc – column volume (cm 3) ρB – bulk density (g/cm 3) In case of the high purity production of (S)- enantiomer only by SMB-LC the (S) productivity is low. Joining SMB-LC and crystallization the (S) productivity can be significantly increased. Experiments Chiral racemic ester mixture separation was investigated as a model system. The two pure (S)- and (R)- enantiomers can not be reached in the market and no data can be found in special literature on them. During our research work both chromatographic and crystallization measurements were carried out. By previous analytical and preparative scale HPLC measurements adsorption isotherms were estimated for chiral stationary phase-eluent systems. The terner system solubility (S, R, solvent) diagrams were determined at different temperatures in laboratory scale experiments. Several laboratory scale SMB-LC and joined crystallization experiments were carried out based on previous chromatographic and solubility data. Further we describe in details the applied experimental methods. Chiral chromatographic packing selection Chiral racemic ester mixture separation was investigated by analytical HPLC columns (I.D. = 0.46 cm, L = 25 cm, dp = 5-10 µm) at 20 °C. Previous selection of CSP was done by DAICEL handbook. Accordingly the next CSP of DAICEL were chosen: Chiralcel OD-H, Chiralcel OJ, Chiralpak AD, Chiralpak IA, Chiralpak AS. The applied eluents are as follows: n-hexane, IPA, Et-OH, AcN, Me-OH, MTBE, DKM in different volumetric ratio with 1 cm3/min volumetric velocity. The 20 µl, 5 g racemic mixture/dm3 eluent concentration sample was separated by analytical HPLC column. The enantiomers were detected by GILSON type UV spectrophotometer at 254 nm wavelength. Investigation of CSP packing properties The SUPELCO (I.D. = 1 cm, L = 25 cm) column was packed with the 20 µm Chiralcel OD by dry vibration method applying 60 min vibration time. Air was removed by n-hexane:IPA = 95:5% (v/v) eluent. After that the NTP was determined by injection method. Applied eluents were: n-hexane:IPA = 80:20, 90:10, 95:5% (v/v). Reaching equilibrium condition with eluent, a 100 µl 50 g racemic mixture/dm3 eluent sample was injected to the column by a Rheodyne type injector at 20°C temperature. Detection of enantiomers happened by WATERS UV detector at 254 nm wavelength or CHIRALYSER type chiral detector. Applied eluent volumetric velocities: 2.5, 5, 10, 15, 20, 30 cm3/min. Evaluation of residence time distribution curves were done by triangle method (determination of σ, tR, NTP, α values). The packing total porosity (ε) was determined by TTBB (Tri-tert-butyl-benzene) method. Bulk density of CSP was calculated from weight measurements of dry packing. Values of Langmuir a* constants were calculated from k’ values. Langmuir b constants were calculated from frontal adsorption- elution measurement data [11]. Frontal adsorption-elution measurements Measurements were carried out by SUPELCO column (I.D.=1 cm, L=25 cm) packed with 20 µm Chiralcel OD at 20 °C temperature using n-hexane:IPA = 95:5, 95:7% (v/v) eluent. 2, 5, 10 g racemic mixture/dm3 eluent sample in 10 cm3 volume was injected to the column with 2.5, 5 cm3/min volumetric velocities. Detection happened by GILSON type UV spectrophotometer at 254 nm wavelength. Based on the above data the Langmuir adsorption isotherm a* and b values were calculated by Felinger et al. [10] method. Description of laboratory scale SMB-LC equipment The SMB-LC equipment with four columns (I.D. = 1 cm, L = 25 cm SUPELCO type stainless steel, mounted with liquid distributor and porous stainless steel plates), four zones and open eluent loop was constructed in the Central Mechanical Workshop of the Pannon University. The joint pipes were made of 1/16” stainless steel material, the liquid stream controlled by VALCO 5 four-ways and 4 two-ways cocks, and by 4 GILSON type HPLC pumps with maximum volumetric velocity: one of them is 5 cm3/min, two are 10 cm3/min and one is 25 cm3/min. Volumetric velocities (D, E, F, R) were measured by EMX 100/211 type digital balance connected to a computer. Pressure gauge was connected after the eluent pump. Extract and raffinate outlet streams were detected on-line by UV or chiral detectors. 8 Parameters of SMB-LC measurements There is no liquid recirculation in open loop SMB-LC equipment, but the liquid from IV zone is collected (LR OUT) separately. The feed composition is 2.5 g (S)/dm3 eluent and 2.5 g (R)/dm3 eluent. The eluent composition is: n-hexane:IPA = 95:5% (v/v), 20 °C temperature, column configuration 1:1:1:1. Volumetric velocities adjusted on different pumps: fresh eluent D = 20 cm3/min, feed F = 0.5-1.5-2.9 cm3/min, extract E = 4 cm3/min, raffinate R = 6 cm3/min. Switching time is 5 min. The SMB-LC equipment was working by the usual four- zone process, in each 5 minutes inlet and outlet points were changed. Examination of SMB-LC product streams Extracts, raffinates and LR OUT liquid streams were collected in the 1st, 2nd, 3rd and 4th cycle separately through 20-20 minutes. During the first switching time of the 5th cycle extract, raffinate and LR OUT streams were fractionated in each 75 seconds for studying concentration transients. The above samples were analysed by GILSON HPLC equipment, on Chiralcel OD-H packing at 20 °C temperature, at UV 254 nm wavelength using n-hexane:IPA = 95:5% (v/v) eluent. Solubility investigations 20 mg 80:20, 85:15, 90:10, 95:5, 100:0% (w/w) (S):(R) samples were solved in 4 cm3 n-hexane at 20 °C temperature, later cooled to -22 °C and kept staying for 8 hours. Crystals and crystallization mother liquid were separated from each other and analyzed by GILSON analytical HPLC equipment. Crystal purity in each case was higher than 99% (w/w) (S). The solubility of pure (S)-enantiomer in n-hexane at -22 °C was 1.2 g/dm3, at 20 °C it was 28.5 g/dm3. Solubility of pure (R) in n-hexane at -22 °C was also 1.2 g/dm3. Solubility of “SR” racemate at -27 °C was 0.735 g/dm3, at 20 °C was 25.45 g/dm3. Solubility of (S) and (R) was examined in n-hexane-IPA mixtures at 0, 5, 10, 20% (v/v) IPA content too. Crystallization investigations Extract and raffinate fractions of the SMB-LC measurements were separately evaporated in vacuum at 40-60 °C temperature in Rotadest equipment according to Fig. 5. Because of the n-hexane-IPA vapour liquid equilibrium data at 95:5% (v/v) n-hexane:IPA concentration during evaporation only n-hexane remained in liquid phase. Evaporation was done on the way, that enantiomer concentration was about 5 g/dm3 in the evaporated liquid. The evaporated liquid was kept in a -20…-27 °C freezer through 8 hours. Crystals and crystallization mother liquid were separated from each other and analyzed by GILSON analytical HPLC equipment. Results CSP packing selection Capacity relations and selectivity coefficients were calculated from analytical HPLC measurements. 0 0' t tt k R − = ' ' S RR S k k =α 0 2 4 tF L d ⋅ ⋅ = π ε Where tR – retention time of the given enantiomer (min) t0 – dead time (min), d = 0.46 cm, inner diameter of column, L = 25 cm, length of column, F = 1 cm3/min, eluent volumetric velocity ε – bed voidage, total porosity (determined by TTBB method) On Table 1 are summarized the experimental data. The best αRS selectivity was given in case of Chiralcel OD-H packing and n-hexane-IPA eluent. Thus this system was used in further experiments. 9 Table 1: Screening measurement results by analytical HPLC equipment Concentration (% v/v) Chiralcel OD-H Chiralcel OJ Chiralpak AS Chiralpak AD Chiralpak IA n-hexane:IPA 70:30 αRS=1 n-hexane:IPA 80:20 αRS=1.17 α R S=1.196 n-hexane:IPA 90:10 αRS=1.173 α R S=1 α R S=1 n-hexane:IPA 97.5:2.5 αRS=1.06 n-hexane:IPA 95:5 αRS=1.19 α R S=1 α R S=1.03 n-hexane:Et-OH 95:5 αRS=1.122 n-hexane:IPA:AcN 80:10:10 αRS=1 AcN 100 k'S=0.24 αRS=1 AcN-Me-OH 80:20 αRS=1 α R S=1 Ethanol 100 αRS=1 n-hexane:MTBÉ 80:20 k'S>20 n-hexane:MTBÉ 60:40 k'S=6.10 n-hexane:MTBE:Et-OH (60:40)+5% Et-OH k'S=1.086 n-hexane:IPA αRS=1.08 n-hexane:Met-OH 99:1 αRS=1.12 n-hexane:Et-OH 99:1 αRS=1.11 n-hexane:DKM 75:25 αRS=1.03 n-hexane:IPA:EtOH 95:2.5:2.5 αRS=1.15 n-hexane:IPA:MetOH 95:2.5:2.5 αRS=1.05 n-hexane:EtOH:MetOH 95:2.5:2.5 αRS=1.05 PackingEluent Investigation of CSP packing properties 0 50 100 150 200 250 300 350 400 450 0 5 10 15 20 25 30 F (cm3/min) N TP / 25 cm n-hexane:IPA=95:5% (v/v), (S)-enantiomer n-hexane:IPA= 90:10% (v/v), (S)-enantiomer n-hexane:IPA= 80:20% (v/v), (S)-enantiomer n-hexane:IPA= 95:5% (v/v), (R)-enantoimer n-hexane:IPA= 80:20% (v/v), (R)-enantiomer n-hexane:IPA= 90:10% (v/v), (R)-enantiomer Fig. 7: Results of elution chromatographic experiments, SUPELCO HPLC column (I.D. = 1 cm, L = 25cm), Chiralcel OD (particle size: 20 μm) packing, T = 20 °C, sample: 50 g chiral racemic mixture/cm3 eluent, 100 μl injection, eluent: n-hexane:IPA. For example the n-hexane:IPA = 95:5% (v/v), (S) isomer experimental curve can be described by the next fitting equitation. 3876,036.643 −= FNTP Where F – volumetric velocity (cm3/min), NTP – Number of theoretical plates/25 cm column, ε = 0.67 bed viodage, total porosity ρB = 0.6 g dry CSP/cm 3 column volume, bulk density 10 Frontal adsorption-elution measurement and simulation On the basis of k’ capacity relation values determined by elution measurements (Chiralcel OD-H, 95:5% (v/v) = n-hexane:IPA) K Morbidelli parameters and Langmuir adsorption a* values were determined at 20 °C temperature, b values of Langmuir adsorption isotherm were calculated by Felinger et al. [11] method using the KROM-N simulation program (Table 2). Table 2: Input data of KROM-N software Data input to the KROM-N software Number of components: 2 Column inner diameter: I.D. = 1 cm Column length: L = 25 cm Free volume coefficient: ε = 0.67 cm3 liquid free volume/cm3 column Bulk density: ρB = 0.6 g packing/cm 3 column Langmuir constants: as given Volumetric velocity: 2-2.5-5 cm3/min Sample feeding time: 2-4-5-6 min (10 cm3) Sample concentration: g (S ) component/dm3 liquid g (R ) component/dm3 liquid Number of Theoretical Plates: NTP = 200/25 cm column End of calculation time: 400 min 55.21FB −−=c 55.21FA −−=c adsorbentcm liquidcm kK RARA 3 3 )()( 998.121 ' = − = ε ε adsorbentcm liquidcm kK SBSB 3 3 )()( 058.101 ' = − = ε ε columncm packingg BULK 3 6.0=ρ 29.1=RSα method) benzol-butil-tert-Tri(67.0 3 3 columncm liquidcm =ε packingg volumefreeliquidcm Ka BULK RARA 3 )( * )( 1489.7 1 = − = ρ ε componentAmg volumefreeliquidcm b RA 3 )( 016.0= packingg volumefreeliquidcm Ka BULK SBSB 3 )( * )( 5320.5 1 = − = ρ ε componentBmg volumefreeliquidcm b SB 3 )( 012.0= KROM-N simulation program was used for simulation and compared to the measurement results. The conclusion is, that small concentration (2 g/dm3 sample) measurements provide the best separation. Results of SMB-LC simulations We examined the inlet concentration effect on Morbidelli triangle. Enantiomers in feed are in equal ratio, (R)=(S) as a racemic mixture. By data given in previous chapter. adsorbentcm liquidcm K R 3 3 998.12= adsorbentcm liquidcm K S 3 3 058.10= Fig. 8 shows, that increase of inlet concentration decreases deforms the pure components separation area of Morbidelli triangle. 7 8 9 10 11 12 13 14 15 7 8 9 10 11 12 13 14 15 mII m III cF=0,0 g/dm3 cF=2,5 g/dm3 cF=5,0 g/dm3 cF=7,5 g/dm3 cF= 10,0 g/dm3 cF = 0.0 g/dm 3 cF = 2.5 g/dm 3 cF = 5.0 g/dm 3 cF = 7.5 g/dm 3 cF = 10.0 g/dm 3 Fig. 8: Changes in Morbidelli triangle in function of feed concentration, cF = cS+cR, cS = cR in feed. The SMB-LC process volumetric velocity by zones and the switching time were determined in linear adsorption isotherm region using the Morbidelli theory. On the basis of that the next relations must be true to produce pure (S)- and pure (R)-enantiomers respectively in product streams: 12.998 = KR < mI 10.058 = KS < mII < KR = 12.998 10.058 = KS < mIII < KR = 12.998 mIV < KS = 10.058 Assuming 5 min switching time by the above relations the fresh eluent (D) must be over 19.47 cm3/min, extract stream between 0.53 and 4.34 cm3/min, the feed stream between 0 and 3.47 cm3/min values. The minimal raffinate stream was 3.24 cm3/min value. During simulation the aim was to achieve 80, 85, 90, 95, 99% (w/w) (S) pure raffinate at >99% (S) yield. The fresh (D) volumetric velocity can not be higher than 20 cm3/min because of the allowed highest 50 bar pressure (CSP packing specification). The working points of the simulations can be found in Morbidelli triangle on Fig. 9, the parameters in Table 3 and the results in Table 4. 11 Table 3: Input data of SMB KROM-N software Number of components: 2 Column inner diameter: I.D. = 1 cm Column length: L = 25 cm Number of columns: N = 4 Free volume coefficient: ε = 0.67 cm3 liquid free volume/cm3 column Bulk density: ρB = 0.6 g packing/cm 3 column SIM 1. SIM 2. SIM 3. SIM 4. SIM 5. SIM 6. SIM 7. Feed: cm3/min 3.47 3.2 2.9 2.6 2.35 1.5 0.5 Fresh eluent: D = 20 cm3/min Extract: E = 4 cm3/min Raffinate: R = 6 cm3/min Recycling: REC = 0 cm3/min Langmuir constants: as given Feed concentration: g (S ) component/cm3 liquid g (R ) component/cm3 liquid Number of Theoretical Plates: NTP = 200/25 cm column Switching time: 5 min Calculation time: 400 min Data input to the SMB-KROM-N software 5.2=FBc 5.2=FAc Table 4: Results of simulations S % (A/A) R % (A/A) S (%) R (%) (mg S / g packing/day) (mg R / g packing/day) (cm3 eluent/mg S ) (cm3 eluent/mg R ) SIM 1. 76.44 99.61 99.97 68.84 268.31 184.16 2.27 3.32 SIM 2. 80.43 99.63 99.97 75.29 245.40 184.15 2.49 3.32 SIM 3. 85.88 99.65 99.98 83.04 222.50 184.06 2.74 3.33 SIM 4. 93.18 99.66 99.98 91.77 199.63 182.37 3.06 3.35 SIM 5. 97.43 99.64 99.97 96.10 180.54 173.61 3.38 3.54 SIM 6. 99.73 99.35 99.96 98.06 115.32 112.43 5.3 5.44 SIM 7. 99.98 98.38 99.96 97.98 38.55 37.41 15.86 16.34 Eluent consumption Simulation ProdictivityYieldPurity 5 10 15 5 10 15 mII m III SIM 1 SIM 2 SIM 3 SIM 4 SIM 5 SIM 6 SIM 7 Fig. 9: The working points of the SMB simulations, simulated with SMB KROM-N program, dash line: non-linear Morbidelli triangle at cF = 5 g/dm 3. Summarized results for (S)-enantiomer (purity, yield, productivity, eluent consumption) can be seen on Fig. 10. 0 50 100 150 200 250 300 0 1 2 3 4 Feed (cm3/min) 0 2 4 6 8 10 12 14 16 18 Purity S% (A/A) Yield (%) Productivity (mg S/g packing/day) Fig. 10: Summarized results for (S)-enantiomer in Raffinate stream. Evaporation of raffinate streams, crystallization, crystallization mother liquid recirculation Simulation calculation were done. Solubility of both (S)- and (R)-enantiomers at -20 °C temperature were 1.2 g/dm3. Raffinates of SMB-LC were evaporated until 5 g (S)+(R)-enantiomers/dm3 concentration followed by cooling to -20 °C temperature. Purity of crystal (S) was assumed 99.9% (w/w). On Table 5 are summarized data for the calculation of crystallization. 12 Table 5: Summarized data for the calculation of crystallization S (g/dm³) R (g/dm³) S :R S (g/dm³) SR (g/dm³) S (g) S (%) 4.00 1.00 80:20 1.2 2.0 1.8 45.0 4.25 0.75 85:15 1.2 1.5 2.3 54.1 4.50 0.50 90:10 1.2 1.0 2.8 62.2 4.75 0.25 95:5 1.2 0.5 3.3 69.5 4.90 0.10 98:2 1.2 0.2 3.6 73.5 5.00 0.00 100:0 1.2 0.0 3.8 76.0 Crystal ή Solution to cooling +20°C, 5 g/dm³ total concentration Solution at -20°C Crystallization mother liquid was evaporated to 5 g (S)+(R)-enantiomers/dm3 concentration, which was mixed to the SMB-LC feed. In the recycled evaporated mother liquid IPA concentration was adjusted at 5% (v/v) value. Evaporation of raffinate streams, crystallization, crystallization mother liquid recirculation was repeated by this modified feed concentration twice and three times. After the third full calculation quasi-stationary condition was reached. Calculated data can be found in Table 6 and 7. Table 6: First step crystallization calculated data cFS c F R (g/dm3) (g/dm3) C I.01 3.80 1.50 1.97 2.82 2.18 C I.02 4.07 1.18 2.02 2.81 2.19 C I.03 4.38 0.81 2.09 2.79 2.21 C I.04 4.65 0.47 2.13 2.77 2.23 C I.05 4.59 0.32 2.03 2.76 2.24 C I.06 3.00 0.15 1.35 2.75 2.25 C I.07 1.00 0.00 0.50 2.50 2.50 Crystallization code 99,9% purity crystal (mg/min) Recirculation (cm3/min) Fresh Feed (cm3/min) Table 7: Calculated data after the third full calculation Raffinate purity Fresh eluent consumption Productivity Rec. Fresh cFS c F R S % (w/w) (cm 3 eluent/mg S ) (mg S /g packing/day) SIM. III.01. C III.01. SIM._III.02. C III.02. SZIM. III.03. C III.03. SZIM.III.04. C III.04. SZIM. III.05. C III.05. SZIM. III.06. C III.06. SZIM. III.07. C III.07. Simulation, Crystallization code Feed (cm3/min) Concentration (g/dm3) 1.37 2.10 2.85 2.15 81.71 3.80 160.96 1.08 2.12 2.83 2.17 84.67 3.77 162.01 0.77 2.13 2.81 2.19 89.07 3.74 163.50 0.47 2.13 2.79 2.21 94.70 3.73 163.43 0.33 2.02 2.78 2.22 97.82 3.94 155.33 0.17 1.33 2.77 2.23 99.76 6.00 102.05 99.98 15.86 38.550.00 0.50 2.50 2.50 On the basis of the above data productivity maximum was at values 2.13 cm3/min fresh feed and 0.77 cm3/min recirculation feed, when raffinate purity was 89.07% (w/w) (S). Fresh eluent consumption (calculated for the total SMB+crystallization system) was 3.74 cm3 fresh eluent/mg (S). Considering the pure (S)-enantiomer 28.5 g (S)/dm3 solubility in n-hexane at 20 °C temperature and (R) and (S) 1.2 g/dm3 solubility in n-hexane at -20 °C temperature, crystallization process was recalculated (See Table 8). During calculation purity of (S) 99.9% (w/w) was assumed in crystals. The so called limiting curve belonging to the 99.9% (w/w) (S) crystal value can be drawn (Fig. 11). Table 8: Summarized data for the calculation of crystallization of 99.9% (w/w) pure (S) S (g/dm³) R (g/dm³) S :R S (g/dm³) SR (g/dm³) S (g) S (%) 4.80 1.20 80:20 1.2 2.40 2.40 50.0 6.80 1.20 85:15 1.2 2.40 4.40 64.7 10.80 1.20 90:10 1.2 2.40 8.40 77.7 22.80 1.20 95:5 1.2 2.40 20.40 89.4 27.93 0.57 98:2 1.2 1.14 26.16 93.7 28.50 0.00 100:0 1.2 0.00 27.30 95.7 Crystal ή Solution to cooling +20°C Solution at -20°C 0 5 10 15 20 25 30 0 5 10 15 20 R enantiomer in solution to cooling (% w/w) L im iti ng c on ce nt ra tio n (g /d m3 ) Total concentration S concentration R concentration Fig. 11: Concentration limiting curve of 99.9% (w/w) (S) crystals, calculated. Using the above so called limiting curve data belonging to the 99.9% (w/w) (S) crystal product this hybrid process (evaporation of raffinate streams, crystallization, crystallization mother liquid recirculation) were recalculated three times. Data can be seen on Table 9 and Fig. 12. By these data the optimum of hybrid system was at 2.32 cm3/min fresh feed, 0.18 cm3/min recirculated feed and 95.6% (w/w) (S) raffinate purity. At optimum value productivity was 177.83 mg (S)/g packing/day, fresh eluent consumption (calculated for the total SMB+crystallization system) was 3.44 cm3 fresh eluent/mg (S). 13 Table 9: The data of hybrid system at optimized evaporation, quasi-stationary state Raffinate purity Fresh eluent consumption Productivity Rec. Fresh cFS c F R S % (w/w) (cm 3 eluent/mg S ) (mg S /g packing/day) SIM. III.08. C III.08. SIM._III.09. C III.09. SZIM. III.10. C III.10. SZIM.III.11. C III.11. SZIM. III.12. C III.12. 99.52 4.18 134.370.05 1.75 2.54 2.46 95.60 3.44 177.83 0.11 2.24 2.55 2.45 97.43 3.64 167.97 0.18 2.32 2.56 2.44 83.81 3.68 166.2 0.44 2.30 2.64 2.36 90.73 3.47 176.01 Simulation, Crystallization code Feed (cm3/min) Concentration (g/dm3) 1.03 2.17 2.77 2.23 0 50 100 150 200 80 82 84 86 88 90 92 94 96 98 100 S% (w/w) 0 5 10 15 20 Productivity at c=5g/dm3 (mg S/packing/day) Productivity at optimized c (mg S/packing/day) Fig. 12: The calculated data of hybrid system at optimized evaporation presented on figure. SMB-LC measurements Table 10 shows planned volumetric velocities. Specific values of SMB-LC processes were calculated after data of 4th full cycle. Difference between the measured and simulated values was caused by the relatively short measuring time (80 min), so the SMB-LC quasi- stationary condition could not be reached. The calculated and measured working points are shown on Fig. 13. Evaporation crystallization joined to SMB-LC (SMB 01 measurement) resulted from 96.68% (w/w) (S) raffinate the 99.33% (w/w) (S) crystal. In case of the SMB 02 and SMB 03 measurements the 99% (w/w) (S) purity could not be reached. Table 10: Results of SMB measurements Identifier Eluent Sample Switching time (min) D E F R LR OUT (S ) (R ) (S ) (R ) (S ) (R ) Purity % (w/w) 96.68 85.7 87.22 84.7 77.43 96 Yield (%) 98.2 99.9 97.4 74 98.7 28 49.3 213.7 61.1 (mg product/g packing/day) Productivity 32.7 30.7 74.7 12.38 2.9 10.2 (cm3 eluent/mg product) Eluent consumption 18.9 20.1 8.18 10.5 11.5 12.9 Enantiomer 0.5 1.5 2.9 6 6 6 n-hexane:IPA = 95:5% (v/v) 5 g racemic/cm3 n-hexane:IPA = 95:5% (v/v) 5 (cm3/min) 20 20 20 4 4 4 SMB 01-03 measurements SMB 01 SMB 02 SMB 03 5 10 15 5 10 15 mII m II I non-linear Morbidelli triangle SMB01 sim. SMB01 meas. SMB02 sim. SMB02 meas. SMB03 sim. SMB03 meas. Fig. 13: The calculated and measured working points in Morbidelli triangle. Measurement results can be seen on Fig. 14 showing the earlier measurement results done at Richter Gedeon Ltd. and Pannon University, together with the limiting curve belonging to the 99.9% (w/w) (S) crystal value. Table 11: Results of evaporation and crystallization measurements (S ) (R ) (S ) (R ) (S ) (R ) (S ) (R ) (S ) (R ) (S ) (R ) Raffinate 96.68 3.32 5.08 0.17 87.22 12.78 5.72 0.84 77.43 22.57 4.21 1.23 E+LROUT 28.81 71.19 1.57 3.88 34.12 65.88 1.69 3.25 Raffinate 86.25 13.75 0.99 0.16 66.35 33.65 0.95 0.48 67.98 22.84 1.28 0.38 E+LROUT 37.83 62.17 0.48 0.79 Raffinate 99.33 0.67 92.44 7.55 83.29 22.31 2.65 0.76 E+LROUT 27.49 72.51 Crystals Purity % (w/w) c (g/dm3) Evaporated liquid Crystallization mother liquor Purity % (w/w) c (g/dm3) Purity % (w/w) c (g/dm3) Evaporation and crystallization SMB 01 SMB 02 SMB 03 14 0 10 20 30 40 50 60 70 80 90 100 0 5 10 15 20 25 R enantiomer in solution before cooling (% w/w) E ff ic ie nc y of c ry st al liz at io n (% ) SMB 01 (>99% (w/w) purity product) SMB 02 (92% (w/w) purity product) SMB 03 (83% (w/w) purity product) Experiment of Pannon University >99% w/w product Experiment of Gedeon Richter LTD. >99% w/w product Efficiency of cristallyzatoin at optimized limiting concentration Efficiency of crystallization at 5g/dm3 Fig. 14: Measurement results and the earlier measurement results done at Richter Gedeon Ltd. and Pannon University, together with the optimized limiting curve belonging to the 99.9% (w/w) (S) crystal value. Summary In this work separation of pharmaceutical enantiomers coupling with crystallization using SMB-LC joined crystallization hybrid system were studied. The aim of this work was the production of valuable (S)-enantiomer of racemic ester mixture from raffinate with higher than 99% (w/w) (S) purity and higher than 99% (S) yield for getting maximal (S) productivity and minimal fresh solvent consumption. The main parts of hybrid system and processes are: enrichment of (S) in SMB-LC raffinate stream, evaporation of raffinate, cooling, separation of (S) crystals and crystallization mother liquid, recirculation of crystallization mother liquid to the feed after (S)+(R) and eluent concentration adjustment. Evaporated eluent was recirculated from raffinate and extract streams to the SMB-LC equipment inlet. After total evaporation of extract stream (R) component was produced could be transformed to (S)-enantiomer by chemical or other processes outside the SMB-LC crystallization hybrid system. By previous laboratory scale measurements fundamental data of SMB-LC and joined crystallization necessary for computer simulation were determined. We concluded, that in SMB-LC crystallization hybrid system productivity could significantly be increased from 115.32-177.83 mg (S)/g packing/day compared to the separation method using only SMB-LC equipment at higher than 99% (w/w) (S) purity and higher than 99% (S) yield. The above hybrid system applicability was proven by laboratory and industrial scale SMB-LC, crystallization experiments. Acknowledgements The research was financed by the Chemical Engineering Institute Cooperative Research Center of the Pannon University, upon the request of Richter Gedeon LTD. The authors express their thanks to these institutions. REFERENCES 1. LORENZ H., SHEEHAN P., SEIDEL-MORGENTERN A., J. Chromatogr. A, 908 (2001) 201-214 2. LORENZ H., PERLBERG A., SAPOUNDJIEV D., ELSNER M. P., SEIDEL-MORGENSTERN A., Chemical Engineering and Processing 45 (2006) 863–873 3. KASPEREIT M., GEDICKE K., ZAHN V., MAHONEY A. W., J.Chromatogr. A 1092( 2005) 43-54 4. STRUBE J., GARTNER R., SCHULTE M., Chem. Eng. J. 85(2002), 273-288 5. HUTHMANN E., JUZA M., J. Chromatogr. A, 908 (2001) 185–200 6. SZÁNYA T., HANÁK L.: Simulated Moving Bed Liquid Chromatography (SMB-LC), Post-graduate engineer course book, University of Veszprém (2001) 7. MIGLIORINI C., MAZZOTTI M., MORBIDELLI M., J. Chromatogr. A, 827 (1998) 161 M. 8. G. GUIOCHON, J. Chromatogr. A. 965 (2002) 129-161. 9. HEUER C., KÜSTENS E., PLATTNER T., SEIDEL- MORGENSTERN A., J. Chromatogr. A 827 (1998) 175. 10. SZANYA T., ARGYELAN J., KOVATS S., HANAK L., J. Chromatogr. A, 908 (2001) 265–272 11. FELINGER A., CAVAZZINI A., GUIOCHON G., J. Chromatogr. A 986 (2003) 207–225 12. HERSECZKI ZS., Diploma piece: Separation Of Optical Isomers With Preparative Liquid Chromatograph (2006) << /ASCII85EncodePages false /AllowTransparency false /AutoPositionEPSFiles true /AutoRotatePages /None /Binding /Left /CalGrayProfile (Dot Gain 20%) /CalRGBProfile (sRGB IEC61966-2.1) /CalCMYKProfile (U.S. Web Coated \050SWOP\051 v2) /sRGBProfile (sRGB IEC61966-2.1) /CannotEmbedFontPolicy /Error /CompatibilityLevel 1.4 /CompressObjects /Tags /CompressPages true /ConvertImagesToIndexed true /PassThroughJPEGImages true /CreateJDFFile false /CreateJobTicket false /DefaultRenderingIntent /Default /DetectBlends true /DetectCurves 0.0000 /ColorConversionStrategy /CMYK /DoThumbnails false /EmbedAllFonts true /EmbedOpenType false /ParseICCProfilesInComments true /EmbedJobOptions true /DSCReportingLevel 0 /EmitDSCWarnings false /EndPage -1 /ImageMemory 1048576 /LockDistillerParams false /MaxSubsetPct 100 /Optimize true /OPM 1 /ParseDSCComments true /ParseDSCCommentsForDocInfo true /PreserveCopyPage true /PreserveDICMYKValues true /PreserveEPSInfo true /PreserveFlatness true /PreserveHalftoneInfo false /PreserveOPIComments true /PreserveOverprintSettings true /StartPage 1 /SubsetFonts true /TransferFunctionInfo /Apply /UCRandBGInfo /Preserve /UsePrologue false /ColorSettingsFile () /AlwaysEmbed [ true ] /NeverEmbed [ true ] /AntiAliasColorImages false /CropColorImages true /ColorImageMinResolution 300 /ColorImageMinResolutionPolicy /OK /DownsampleColorImages true /ColorImageDownsampleType /Bicubic /ColorImageResolution 300 /ColorImageDepth -1 /ColorImageMinDownsampleDepth 1 /ColorImageDownsampleThreshold 1.50000 /EncodeColorImages true /ColorImageFilter /DCTEncode /AutoFilterColorImages true /ColorImageAutoFilterStrategy /JPEG /ColorACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /ColorImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000ColorACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000ColorImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 300 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /GrayImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000GrayACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000GrayImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict << /K -1 >> /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile () /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False /Description << /CHS /CHT /DAN /DEU /ESP /FRA /ITA /JPN /KOR /NLD (Gebruik deze instellingen om Adobe PDF-documenten te maken die zijn geoptimaliseerd voor prepress-afdrukken van hoge kwaliteit. De gemaakte PDF-documenten kunnen worden geopend met Acrobat en Adobe Reader 5.0 en hoger.) /NOR /PTB /SUO /SVE /ENU (Use these settings to create Adobe PDF documents best suited for high-quality prepress printing. Created PDF documents can be opened with Acrobat and Adobe Reader 5.0 and later.) >> /Namespace [ (Adobe) (Common) (1.0) ] /OtherNamespaces [ << /AsReaderSpreads false /CropImagesToFrames true /ErrorControl /WarnAndContinue /FlattenerIgnoreSpreadOverrides false /IncludeGuidesGrids false /IncludeNonPrinting false /IncludeSlug false /Namespace [ (Adobe) (InDesign) (4.0) ] /OmitPlacedBitmaps false /OmitPlacedEPS false /OmitPlacedPDF false /SimulateOverprint /Legacy >> << /AddBleedMarks false /AddColorBars false /AddCropMarks false /AddPageInfo false /AddRegMarks false /ConvertColors /ConvertToCMYK /DestinationProfileName () /DestinationProfileSelector /DocumentCMYK /Downsample16BitImages true /FlattenerPreset << /PresetSelector /MediumResolution >> /FormElements false /GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks false /IncludeInteractive false /IncludeLayers false /IncludeProfiles false /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /DocumentCMYK /PreserveEditing true /UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling /UseDocumentProfile /UseDocumentBleed false >> ] >> setdistillerparams << /HWResolution [2400 2400] /PageSize [612.000 792.000] >> setpagedevice