WO2017072542A1

WO2017072542A1 - Method for the purification of cyclosporine a

Info

Publication number: WO2017072542A1
Application number: PCT/HU2016/050049
Authority: WO
Inventors: László LORÁNTFY; László NÉMETH
Original assignee: Rotachrom Technológiai Kft.
Priority date: 2015-10-26
Filing date: 2016-10-26
Publication date: 2017-05-04
Also published as: HUP1500502A2

Abstract

The object of the invention relates to a method for the purification of cyclosporine A using a centrifugal partition chromatography method. More precisely, in addition to removing the other contaminants, the method according to the invention especially relates to the removal of the contaminant dihydro-cyclosporine A. The method according to the invention can be preferably carried out by using the specially constructed extraction cells described in the invention and modular CPC rotors containing these.

Description

METHOD FOR THE PURIFICATION OF CYCLOSPORINE A

The object of the invention relates to a method for the purification of cyclosporine A using a centrifugal partition chromatography method. More precisely, in addition to removing the other contaminants, it especially relates to the removal of the contaminant dihydro- cyclosporine A using the method according to the invention.

The technical background of the invention:

Cyclosporine A (hereinafter: CsA) is an immunosuppressant medicine used in transplantation to prevent the rejection of the transplanted organ. In addition it may be used to good effect in the treatment therapies of various autoimmune diseases. Chemically it is a peptide that is produced by a fungus found in soil called Tolypocladium inflatum Gams. It is not only built up by amino acid building proteins. Its immunosuppressive effect was discovered in 1972 and it has been used in kidney and liver transplantations since 1983.

Cyclosporine, as a cyclic peptide, can be most economically produced via fermentation. Hungarian patent number 204 101 presents such a fermentation method. Usually, numerous by-products with chemical structure and physical properties similar to the target compound are also formed in these processes. The by-products created in this way may not get into the pharmaceutical preparations so they must be removed.

Removing certain contaminants from cyclosporine A using methods commonly used in organic chemistry, such as crystallization and distillation, is not possible because the physical-chemical properties and solubility of the contaminants are similar, so it is not possible to obtain a product of sufficient purity using crystallization. Distillation is also not a feasible method due to the size of the molecules and the lack of thermal stability. Therefore various chromatography methods have been elaborated for the purification of cyclosporine A. According to the inventors of patent number EP 0920447, CsA can be separated from the contaminants CsB, CsC, CsL, CsU and CsD by using normal phase high performance liquid chromatography (HPLC). According to patent number US 5382655, CsA can be separated from the CsB and CsC contaminants by using normal phase high performance liquid chromatography. HPLC is a very costly technique in which there is a great deal of solvent consumed and the silica gel used has to be replaced at certain intervals.

According to the inventors of patent number US6306306 simulated moving bed high performance liquid chromatography (SMB) is suitable for separating CsA from the contaminants CsC, CsB, CsL, CsG and CsD. SMB is a more economic technology than HPLC, however, it is less robust and its configuration takes much more time and money.

None of the inventors of the above patents deals with the removal of the contaminant dihydro-cyclosporine A (hereinafter: H2CsA), this may appear during the fermentation production of cyclosporine, and in the case of its content being over the limit value, the product obtained is unusable for pharmaceutical industry purposes if the contaminant cannot be economically removed. Dihydro-cyclosporine A differs from CsA only in that the double bond in one of the side chains in CsA is saturated in H2CsA. With respect to both their physical and chemical properties the two compounds are exceptionally similar.

Patent number US6620325 presents an industrial counter-flow extraction based method for the purification of CsA. The system containing n-heptane, acetone and water is able to remove the dihydro-cyclosporine A from the main component of cyclosporine A. According to the patent the distribution of cyclosporine A and dihydro-cyclosporine A between light and heavy phases is 0.8 (CsA) and 0.1 (H2CsA) in this system at 40 °C, in other words selectivity is 1.25. This would make it possible to remove this contaminant using an approximately 25-metre high counter-flow extractor with a theoretical plate count of 30. Although at first glance this seems to be an efficient, industrially applicable method, it still has numerous disadvantages:

- According to the patent the purification may be carried out in two steps.

- The individual grades require mixed columns that are 25 metres high, which are expensive to construct and require a special architectural structure due to their dimensions.

- The solvent demand for the production of 1 kg of purified product is about 1500 litres.

- A separate problem is presented by the fact that regeneration of heptane is difficult.

- 1 kg of the final product can be extracted by evaporation from approximately 500 I of solvent, which involves a further heat burden on the product, furthermore - according to the patent the extraction itself takes place at a temperature preferably between 30-60 °C, in other words the columns must be heated. This is made probable by that the partition coefficients were carried out in solvents at 40 °C. The fact that the heating of the extractor must be precisely controlled and the extractor must also be insulated raises the already high investment and operation costs even further. This is also supported by the fact that according to our measurements at room temperature, the selectivity of a heptane - acetone - water system with a volume ratio of 2:2.5:1, which system has a composition nearly identical to that of the patent, is only 1.1.

- Furthermore, although from line 56 to line 63 of column 4 patent number US6620325 promises to produce CsA with 99.7% purity at industrial quantities with this method, this is not supported by the practical examples. The pilot system produced a product with an HPLC purity of 98.5%, while in the other two examples the product produced had purity of 99.5%. At the same time, the chromatographic patterns presented are not integrated, therefore the real content of the contaminants cannot be calculated, neither in the initial, the intermediate nor final product. Not only is the content of all contaminants limited in the case of the purity of pharmaceutical active substances, but the content of the individual contaminants is also limited, i.e. of H2CsA as well. According to current regulations, this limit is 0.3%. Judging by the data described in the case of the method according to the patent, it is not probable that such a product is produced. The counter-flow extraction described in the patent was simulated using ProMise 2 software (ProMISE2 version 1.2.0.1, J. de Folter 2014). The simulation was set up with a sample feed of 1 ml/min., the bottom phase feed and the bottom phase feed rate was 13 ml/min. and the volume of the column was 100 ml. The efficiency with 160 CCD extraction elements was set to 25% and the initial H2CsA content was set to 0.48%. If the selectivity is 1.25 (K CsA= 0.8 K H2CsA=l), then the system would really be able to produce a product with 0.26% H2CsA content with 93% recovery. However, selectivity greater than 1.1 was not measuredby us. With 1.1 selectivity (K CsA=0.909 K H2CsA=l) the method would give a product with 0.318% H2CsA content with 75% recovery, which quality is insufficient. If the feed of the bottom phase were to be reduced to 12 ml/min., then according to the simulation it would be possible to obtain a product with 0.25% H2CsA content, but only with 56% recovery. However, with the efficiency of 25% of the 160 elements used for the calculation, this would only be possible using a 25-metre high column. And in this case, in the 25-metre column, in which the first extraction step of the above patent is performed, only the apolar contaminants would be removed, while the polar contaminants would have to be removed in a subsequent extraction step. The simulation provides a result for an ideal case, which neglects to take into consideration further circumstances that degrade the separation. This means that in practice an even worse result can be expected than the separation performance obtained from the simulation.

On the basis of the above a person skilled in the art would by all means improve the method described in the patent by changing the column at least for the purpose of using less solvent and increasing the level of purity, or by using a solvent mixture that makes better separation possible. However, an experimental column is a significant investment, which a person skilled in the art would only risk in the hope of probable success.

Therefore, there is a demand for the elaboration of a new purification method that overcomes the above problems, or that significantly reduces their effect, and that may be applied on industrial scales. The objective set was the elaboration of a method with which dihydrochloride cyclosporine A (hereinafter: H2CsA) can be removed from raw cyclosporine A (hereinafter: CsA) obtained via fermentation to such an extent that the amount of H2CsA in the CsA is less than 0.03%. A further objective was to reduce the amount of solvent used for the purification, to use a simple and easily reproduced method that may, in a given case, be automated, which uses simple equipment that may be simply replaced and expanded if required, and which method, if possible, does not create solid waste that needs to be disposed of and that does not use a large amount of energy. The optimisation of purification methods is made even more difficult by the fact that for reasons that are not fully understood, when producing CsA the level of H2CsA is generally between 0.2-0.5%. Sometimes it complies with the limit value and sometimes it is just above it. In such a case, especially if the value is above 0.32%, HPLC purification is no longer economic.

Surprisingly we found that the above objectives can be achieved if centrifugal partition chromatography (CPC) is used for the purification of cyclosporine A. In a preferable embodiment of the method, a CPC device is used that is equipped with improved, further developed cells. Using this method raw cyclosporine containing as much as 0.48% of H2CsA can be purified with a yield of more than 60%. As during our pilot investigations while purifying raw cyclosporine containing 0.48% of H2CsA with the method according to the invention, a product was obtained that contained 0.22% H2CsA with a recovery rate of 62%.

Brief description of the invention: The essence of the invention is a method for purifying cyclosporine A using centrifugal partition chromatography (CPC). This method is a liquid-liquid phase separation based method in which the one phase is in the device as the stationary phase and the other, mobile phase is made to flow through the stationary phase. In a preferable embodiment of the method, a CPC device is used that is equipped with improved, further developed cells. We surprisingly found that the disadvantages of patent number US6620325 can be significantly reduced if a method is used based on liquid-liquid phase separation, in which one of the phases is the stationary phase and the other phase is a mobile phase. With respect to that the basic principle of the two methods, the distribution of the material to be purified and its contaminants between the two liquid phases and the subsequent separation of the phases, is the same, it is surprising that the desired pure product can be produced with the use of little solvent, at a high level of purity in a simple way. It was even more surprising that the set objective could be achieved with centrifugal partition chromatography (CPC) even in the case that the selectivity of the solvent system used is much lower with respect to CsA - H2CsA than in the case described in the state of the art (1.25).

Detailed description of the invention:

The object of the invention relates to a centrifugal partition chromatography (CPC) method for purifying cyclosporine A. The method being carried out by

a.) filling the CPC rotor with liquid constituting the stationary phase, then

b.) injecting the raw cyclosporine A solution while the CPC rotor is rotating, then following this

c. ) having mobile phase flow through the CPC rotor while it is rotating, following which d. ) stationary phase is once again filled into the CPC rotor while it is rotating, then, in a given case, the steps b.), c.) and d.) are repeated until the desired amount of cyclosporine solution to be purified has been injected, and in the meantime a fraction or fractions are taken from the mixture exiting from the CPC rotor in steps c.) and d.) and then that fraction or those fractions are selected in the mixtures of which the H2CsA content is < 0.3%, preferably < 0.27%, more preferably < 0.25% with respect to the CsA content, then the cyclosporine A is separated from the selected fraction/fractions mixture.

The separation of the purified cyclosporine A from the solvent belongs to the general knowledge of a person skilled in the art. It may also be separated by distilling off the obtained fraction.

Figure 1 presents a schematic picture of the CPC rotor device. The method essentially comprises filling the CPC rotor with stationary phase, preferably while the rotor is rotating, then after this injecting the raw CsA solution via the sample inlet unit, then starting the feeding of the mobile phase via the liquid pump system. After inputting a determined amount of mobile phase the inputting of the stationary phase is started with the liquid pump unit, with this expelling the remaining mobile phase from the CPC rotor, as well as that part of the stationary phase which contains active substance and/or contaminants from the sample. In the meantime fractions are taken from the expelled solvent mixture, via an interposed detector, and the fractions of the appropriate purity are combined, and then the pure CsA is separated from them. As by the end of the cycle the device is filled with clean stationary phase, a new cycle can be started by injecting more raw CsA. In this way the method can be made quasi-continuous.

The raw CsA solution may be prepared for input by dissolving the raw CsA in the mixture forming the stationary or mobile phase or in one of the solvents forming the stationary and mobile phases, preferably in the solvent in which it dissolves better. If it has been dissolved in pure solvent, then this is diluted with one of the phases if necessary. For example, in a hexane-acetone-water biphase system both the light and heavy phases contain acetone, therefore the raw CsA can be dissolved in acetone, dissolved in the heavier phase and acetone or dissolved in the lighter phase and acetone mixture. It was seen that by increasing the proportion of the well dissolving component present in both phases - acetone in the case of the hexane-acetone-water system - although the ability of the mixture to dissolve is improved, the selectivity obtained is reduced, as are the plate number and resolution. However, the sample solution may also be made in such a common component, e.g. in acetone in a more concentrated solvent mixture, and this does not significantly influence the above parameters, but this makes it possible to use more concentrated solutions, which results in a reduction of solvent consumption. As an example, in a hexane-acetone-water biphase system, the raw CsA solution may be preferably produced by dissolving 100 grams of raw CsA in 100 ml of acetone then making it up to 1 litre with the heavier phase, or alternatively by dissolving 100 g of raw CsA in 1.6 litres of lighter phase. Any material optionally precipitated when making the solution can be left to settle then filtered out and used in the next sample mixture. The fact that even sample input and fraction picking can be ensured makes it significantly easier to operate the device and so the device may be easily automated.

During experiments it was found that it is preferable to perform the method according to the invention in step b.) by feeding in raw cyclosporine solution equivalent to 2-10%, preferably 4-5% of the volume of the column, in step c.) by feeding in mobile phase equivalent to 50-150%, preferably 60-80% of the volume of the column, and in step d.) by feeding in stationary phase equivalent to 20-100%, preferably 30-40% of the volume of the column.

Numerous solvent systems were examined in the hope that it is possible to find a system that has better selectivity than the heptane-acetone-water system described in patent number US6620325. Upon examining the solvent systems used in the patent and other systems similar to this, no system was found among the aforementioned and similar systems in which the selectivity was over 1.10. From the density of the solvents and the heptane- acetone-water phase diagram in figure 11, which contains the data of the ternary system measured at room temperature, it maybe determined that the system used in patent US6620325 largely conforms to the heptane-acetone-water system with a volume ratio of 2:2.5:1 approximately with respect to the composition of the two phases. As the composition of the lower phase of the system in the patent is 50 m/m% acetone 50 m/m% water. Expressed in volume per cent this is 55% and 45%. And the upper phase consists of 78% heptane and 22% acetone. If these two lines are connected in the phase diagram it can be seen that the composition of 2:2.5:1 lies close to it. By using the counter-flow extraction technology described in patent number US6620325, the selectivity of 1.10 results in requiring a very large extractor, approximately 100 m high, if another solvent mixture were to be used in the place of the mixture containing heptane used there, or if the column were not to be heated. Therefore it is obvious for a person skilled in the art that counter-flow extraction cannot be economically used in the case when the main objective is to remove the contaminant H2CsA. The following mixtures are given as examples from among the examined mixtures:

Explanation:

• MTBE = methyl-tert-butyl-ether

• MeOH = methanol

• EtAc = ethyl-acetate

• Partition coefficient: concentration in the upper phase / concentration in the lower phase

It was surprising to find that by using centrifugal partition chromatography (CPC), it possible to achieve the set objectives even if the selectivity of the solvent system used wi respect to CsA - H2CsA is much lower than in the case described in the prior art (1.25). It is especially preferable that the method used by us can be characterised by an at least 10 times theoretical plate number, therefore it is able to replace a 250 m high extractor. Apart from this it is able to process much more concentrated solutions, meaning that solvent use for 1 kg of purified product is in the region of 500 litres. As is mentioned above, during CPC chromatography a stationary and a mobile phase are made to come into contact with each other. As the stationary and mobile phases in the method according to the invention a biphase mixture of a lighter and a heavier phase is used made up of a mixture of liquids in balance with each other, immiscible with each other or only partially miscible, which phases have partition selectivity for CsA - H2CsA greater than 1.05, preferably equal to or greater than 1.1.

An apolar aprotic solvent may be used as the one component of the solvent mixtures forming the two phases in balance, preferably an aromatic, aliphatic or cycloaliphatic hydrocarbon, aliphatic ether or a mixture thereof, and water may be used as the other component and optionally a polar protic and/or dipolar aprotic solvent may be used as a further component of the mixture.

Pentane, hexane, heptane, octane or their isomers, most preferably hexane are used as the aliphatic hydrocarbon in the mixture, alkyl ethers, preferably methyl tert butyl ether (MTBE) is used as the aliphatic ether in the mixture, aliphatic alcohols, preferably alcohols with C1-C4 carbon chains, even more preferably methanol (MeOH), ethanol and isopropyl alcohol (IPA), n-butyl alcohol are used as the polar protic solvent in the mixture, preferably ketones, esters of carboxylic acids with C1-C4 carbon chains formed with alcohols with C1-C4 carbon chains, more preferably ethyl acetate (EtAc) or acetone, and most preferably acetone are used as the dipolar aprotic solvent.

The method according to the invention may be also carried out with the selected mixture from the solvent mixtures forming two phases in balance below where the volume ratios of the individual components have been indicated in brackets in the order of the components: Hexane - MTBE - MeOH - Water (4:5:5:5), Hexane - MTBE - MeOH - Water (4:6:6:5), Hexane - Acetone - Water (3:5:2), Hexane - Acetone - Water (2:6:2), Hexane - IPA - Water (4:4:2) Hexane - EtAc - MeOH - Water (4:5:5:4), Hexane - EtAc - MeOH - Water (4:6:6:4) Hexane - Acetone - Water(2:3:l), Pentane - Acetone - Water (2:3:1), Heptane - Acetone - Water (2:3:1) Heptane - Acetone - Water (2:2.5:1).

According to the most preferable embodiment of the invention, the system forming the two phases contains a mixture of n-hexane, acetone and water and the density of the upper phase is between 0.68 and 0.73, preferably between 0.700 and 0.702 g/ml at 20 °C and the lower phase is in balance with this and its density is between 0.82 and 0.90 g/ml, preferably between 0.885 and 0.890 g/ml at 20 °C.

The method according to the invention is performed at a temperature between -10 and +50 °C, preferably between 0-30 °C and most preferably between 15-20 °C, as long as the upper phase is the mobile phase and the lower phase is the stationary phase. It is surprising that the lower temperature favours separation, and the method works much more efficiently at room temperature, or even cooled a little. All the more so, because as presented above, the CsA - H2CsA selectivity rose with the increase in temperature in the case of heptane- acetone-water presented above. This is contrary to the operation temperature preferably between 30-60 °C described in patent number US6620325, where selectivity is 1.25, the present method is optimal at 15-20 °C, and in this case efficient separation becomes possible even in the case of large injected amounts, even considering that the selectivity in this case is just 1.10. At the same time it is experienced that the method according to the invention may be carried out at a temperature between +10 and +50 °C, preferably between 20-40 °C and most preferably between 25-35 °C, as long as the lower phase is the mobile phase and the upper phase is the stationary phase.

In the method according to the invention the rotation of the CPC rotor must maintained during extraction at a speed so that the stationary phase does not travel together with the mobile phase, but stays in the cell into which it was originally filled as long as the mobile phase is flowing.

Therefore, according to the invention the CPC rotor is rotated at a speed so that the centripetal acceleration occurring due to the rotation at the point furthest from the axis of rotation in the cell in the rotor that is the furthest from the axis of rotation if necessary is greater than 150 g, preferably between 180-500 g, even more preferably between 200-350 g and most preferably at a value between 200-250 g. (1 g~ 9.8 m/s²) So the method according to the invention may be most preferably carried out by dissolving a sample of 100 grams of raw cyclosporine A in 1.6 litres of lighter phase. The column volume of the CPC rotor used is 1.4 litres. With the device rotated at 50-100 g the column was filled with stationary phase (heavier phase) at a flow rate of 200 ml/min. In the meantime the device was gradually accelerated to 220 g. The sample was injected for 1 minute at a flow rate of 50 ml/min. Following this the mobile phase was pumped for 2 minutes at a flow rate of 20 ml/min. then for 25 minutes at a flow rate of 40 ml/min. Following this the pumping was switched over to the stationary phase, which was pumped at a rate of 40 ml/min., and after 5 minutes nine 50-ml fractions were taken. The process was repeated 20 times from injecting the sample. The ambient temperature was 19-20 °C volt. The H2CsA content of the initial sample was 0.50%.

After grouping and combining the obtained fractions CsA of the following purities were obtained with the following yields:

CsA with 80% yield (H2CsA content 0.28%/ HPLC)

CsA with 60% yield (H2CsA content 0.25%/ HPLC)

CsA with 40% yield (H2CsA content 0.24%/ HPLC).

A biphase system was used for the stationary and mobile phase components in the experiment in which the density of the hexane-acetone mixture (upper, lighter phase) used is between 0.68-0.72 g/ml, preferably between 0.69-0.71 g/ml, most preferably between 0.699-0.701 g/ml, while the density of the balanced lower, heavier phase is 0.885-0.890 g/ml at 20 °C.

With the method according to the example 1 kg of product containing 0.27% H2CsA was obtained from 2 kg of raw CsA containing 75% active substance. The rotor used contains 96 tube-like cells. The length of one cell is 10-15 cm, and its diameter for 200 ml/minute is 20 mm. The distance of the cells measured from the axis of rotation is 45-55 cm. It would be even more preferable if a rotor containing 128 cells were to be used. The method according to the invention may be made even more preferable with a specially constructed rotor. As during our research work it was found that if the present, completely general practice of arranging the cells of the rotor so that the stationary and mobile phases in the device are interchangeable is not adhered to, then with a special cell structure a much better separation result may be achieved in the pharmaceutical industry and in the purification of cyclosporine A.

During our investigations the surprising result was found that from the point of view of atomisation a small cell inlet cross-section is preferable while from the point of view of settling, a larger cell outlet cross-section is preferable, in other words a CPC extraction cell which from the point of view of reversibility of flow direction is asymmetrical provides better flow as compared to a similar traditional symmetrical extraction cell. Accordingly, the asymmetrical CPC extraction cell according to the invention has differing inlet and outlet cross-sections.

Furthermore, during our investigations we came to another surprising conclusion regarding that the contact interface between the stationary phase and the mobile phase is significantly increased if, contrary to its traditional structure, instead of the CPC extraction cell having a smooth internal wall, a roughened internal surface is formed on two facing sidewalls. From the point of view of the roughening, a step-like or saw tooth surface structure is especially preferable. During even further investigations it was determined that by suitably selecting the geometry of the CPC extraction cell, back-mixing in the cell can be minimised, furthermore the volume ratio of the stationary phase and the contact interface between the two phases can be maximised. For this the cell geometry in use today is preferably changed by forming a collection pool between the extraction space of the cell and its outlet. In addition, if the outlets and inlets of the CPC extraction cells according to the invention, as well as the channels connecting the individual cells are manufactured to have standard chromatography connections, then the disassembly of a CPC rotor assembled from them becomes very simple, and also the possibility of leakages and other faults is minimised.

Furthermore, the radius of gyration of the extraction cells of CPC rotors widely in use today is generally a maximum of 300 mm, usually falling between 50 mm and 300 mm . The main reason for this is that almost without exception the extraction cells in question a re arranged on a disc with an annular cross-section; in practice the cells are formed in the material of the disc by CNC machining (e.g. milling, cutting). In other words the manufacture of the widely used, disc-type CPC rotor structures with large dimensions, i.e. with a radius of gyration well in excess of 300 mm is very difficult.

During our investigations we also recognised that the effect of the Coriolis force exerted in the extraction cells drops as the axis of gyration r of the CPC rotor increases in line with the physical definition of the Coriolis force

in other words it may be preferable to select the radius of gyration of the CPC rotor (in other words the distance from the axis of rotation at the centre-point of the rotor to where the extractions cells are located) to be relatively large, i.e. in excess of the 300 mm used today.

According to our investigations it is preferable to appropriately change the production technology in accordance with the volume of the work space of the available CNC machines or 3D printers, and instead of producing the whole annular disc as a single work piece, manufacturing it in the form of segments that may be combined to form the disc, where the individual segments comply with the the volume of the work space of the available traditional CNC machines or 3D printers, therefore, by making effective use of this space, the imprecision of machining of CNC machines or 3D printers with larger work spaces can be avoided.

The complete disc forming the CPC rotor is produced preferably from 4-30 segment pieces, optimally from 4-12 segment pieces, which following this are suitable for being combined to form an annular CPC rotor in a way known to a person skilled in the art. Selecting the number of segments to be greater than this means that too few extraction cells can be formed in the individual segments from the point of view of optimum operation . Selecting the number of segments to be smaller than this means that the capacity and separation variability of the extraction cells constructed on a segment drops. In the case of the modular construction in question the distance of the extraction cells from the centre of rotation is 300-1000 mm, and 400-500 mm in an optical case in the applied cut-up structure.

A further advantage of these new cells is that their disassembly, cleaning and possible replacement may be easily carried out, even segment-by-segment. In the case of CPC rotors with a traditional disc structure the material leaking in between the sealing plates used when assembling the discs and the discs into rotors during the operation of the rotors results in the rotors becoming contaminated, which in the light of the strict quality assurance prescriptions of the pharmaceuticals industry, for example, is unacceptable. This significantly prevents the application of CPC rotors in the pharmaceuticals industry. This is especially true above a certain size limit (10-litre internal volume), as due to the pharmaceuticals industry prescriptions the separation of different types of material in the same CPC assembly is only possible following the sequential stopping of the assembly, the complete disassembly of the rotor, the thorough cleaning of each component and then assembly once again. Figures:

An embodiment of the new, very preferable CPC rotors used for the method according to the invention is presented in detail in the following with reference to the Figures, where

Figure 1 is a schematic block diagram of a CPC assembly used today for performing a CPC separation process, where the arrows represent the flow of liquid between the components of the assembly and its direction;

Figure 2 is a schematic illustration of the operation of an extraction cell used in the CPC rotor unit of the assembly according to Figure 1;

Figure 3 is a simulated flow pattern existing in a known extraction cell, which illustrates the degrading back-mixing effect of the Coriolis force in the cell (indicated with an arrow); Figure 4A and 4B are simulated flow patterns that illustrate the competition between chromatographic parameters in the case of two different cell geometries in use today; Figure 5 illustrates a preferable exemplary embodiment of the new type asymmetrical CPC extraction cell according to the invention produced using FDM 3D technology in longitudinal cross section, as well as in partially cutaway schematic perspective view;

Figure 6 is the top view of the CPC extraction cell illustrated in Figure 5 viewed from the direction of arrow A in figure 5;

Figures 7A and 7B are simulated flow patterns of the flow occurring in the extraction cell according to Figure 5 (using the SC/Tetra vll software package) with different outlet cross- sections (1.7 mm and 2.0 mm, the left hand and right hand image in Figure 7A respectively), and with different cell axis tilts (16.5° and 20.0°, the left hand and right hand image in Figure 7B respectively); beside the designated parameter the other parameters used in the simulation remained unchanged in both cases;

Figure 8 shows an example of a preferable embodiment of a new-type asymmetrical CPC extraction cell produced using CNC milling in longitudinal cross section as well as in a schematic, perspective view; Figure 9 shows a schematic depiction of a module of a modularly constructed CPC rotor according to the invention containing several CPC extraction cells according to the invention linked closely together; and

Figure 10 depicts a large sized (r > 300 mm) disc CPC rotor constructed using the CPC rotor modules depicted in Figure 9 with an annular cross-section. Figure 11 depicts the phase diagram of the heptane-acetone-water ternary mixture at room temperature.

Figure 12 is a chromatogram obtained during the HPLC examination of purified CsA, in which CsU = Cyclosporine U; H2CsA = Dihydro-cyclosporine A.

Figure 13 is a chromatogram obtained during the HPLC examination of purified CsA, in which Abu5Cs = Abu5-Cyclosporine A (contaminant name, Abu amino acid can be found in position 5). CsB = Cyclosporine B.

Therefore the method according to the invention may be carried out most preferably by using a new type of CPC rotor that has a. ) at least one extraction cell 100, 200, a cell wall 120, 220 delimiting an enclosed extraction chamber 150, 250 and the differing cross-sections of the inlet 115, 215 and the outlet 140, 240 located at substantially opposing parts of the cell wall 120, 220 formed to ensure the liquid connection between the extraction chamber 150, 250 and the space outside of the extraction cell 100, 200 are such that when the centrifugal partition chromatograph is in operating status the passage cross-section of the outlet 140, 240 exceeds the passage cross- section of the inlet 115, 215 and in a given case the inlet 115, 215 is formed as at least one inlet branch 115a, 115b, furthermore b. ) the cell wall 120, 220 forms a substantially rectangular based inclined prism, the angle of inclination of which, with the extraction cell 100, 200 in its position in the centrifugal partition chromatograph, being selected to minimise the Coriolis force occurring as a result of the rotation when the extraction cell 100, 200 is in its operation status, and preferably this angle of inclination is between 5° and 30°, more preferably between 15° and 18°, furthermore c.) the collection pool 130 is formed as a protruding hemispherical part of the cell wall 120, and the outlet 140 is located on this hemisphere, or the collection pool 230 is formed as a protruding section of the extraction chamber 250 delimited by a part of the cell wall 220 extending in an inclined way in the direction of the outlet 240 and in a given case d.) at least a part of the surface 125 of the cell wall 120 bordering the extraction chamber 150 is roughened, preferably the roughening is ensured by steps or saw teeth on the surface 125.

The method then may be performed with a device constructed of modules 300 containing the above cells 100, 200, which has a modular construction realised with essentially identical modules 300, where all of the modules 300 contain more than one extraction cell 100, 200 connected by channels 330 providing a fluid connection between them, furthermore the individual modules 300 are connected in series with each other via tubes 430.

Consecutively Figures 5 and 6 show an example of a preferable embodiment of a new-type asymmetrical CPC extraction cell 100 in longitudinal cross-section in partially cutaway perspective view and in outline top view. This embodiment of the extraction cell 100 is preferably produced using FDM ("fused deposition modelling") 3D printing technology, although, as is known to a person skilled in the art, it may also be produced using lost core injection moulding. Preferably e.g. peek, i.e. poly[phenyl-(4-phenylene doxy phenyl)ketone resin is used for the 3D printing, although other inert materials suitable for performing 3D printing may also be used, which is obvious for a person skilled in the art. The extraction cell 100 with its geometry to be detailed below is primarily suitable for carrying out CPC processes where the rate of flow of the mobile phase is a maximum of 20 ml/minute.

The extraction cell 100 has a cell wall 120 marking out the extraction chamber 150, which may be basically viewed as a three-dimensional (3D) shape with width "a", thickness/depth "b" and height "c", the dimensions of which preferably comply with the following relationships a > b > 0.5a and 3a > c > 2a, and at least the edges of which perpendicular to the a-c plane are rounded; in the case of a parallelepiped shape the geometric axis of the extraction cell 100 (not depicted in the drawing) when in its position in the CPC rotor (operation state) points essentially in the direction of the centre point of the CPC rotor (see the centre point O of the CPC rotor 400 constructed as an annular disc presented in Figure 10), i.e. it coincides with the radius of the CPC rotor.

At the same time, in its position in the CPC rotor (operation state), the geometric axis of the extraction cell 100 essentially in the shape of an inclined prism is at a determined angle to the radial line drawn from the centre point O of the CPC rotor to the geometric centre point of the extraction cell 100 (tilted cell); where the angle of inclination of the extraction cell (depending on the length of the gyration radius and the planned speed of rotation of the CPC rotor) falls between 5° and 30°; with a gyration radius of 450 mm and a speed of rotation of 750 1/min. the angle of inclination of the extraction cell 100 is preferably between 15° and 18°, and even more preferably 16.5°. The extraction chamber 150 of the extraction cell 100 has an inlet 115 and an outlet 140. The inlet 115 is structured so as to be divided into one or more inlet branches, where each inlet branch is connected to the extraction chamber through a circular inlet opening. Every one of these inlet openings is the same size, their diameters are preferably between 0.5 mm and 1.0 mm. In particular, Figures 5 and 6 show an extraction cell 100 that has two inlet branches 115a, 115b, and, accordingly, two inlet openings 115al, 115bl. According to our investigations, the inlet 115 may be preferably divided into between two and ten inlet branches; the number of inlet branches is preferably between two and four, and is most preferably two. The inlet branches used open into the extraction chamber 150 perpendicularly. The centre points of the inlet openings belonging to the inlet branches are essentially located along a straight line, the straight line in question lies in the a-b plane and is essentially perpendicular to the width "a" of the extraction cell 100. The one or more inlet openings of the extraction cell 100 are essentially formed halfway along the width "a" of the cell, this position may be changed by a maximum of 10% along the width "a" of the cell. From a closer aspect, the division of the inlet 115 into several inlet branches improves atomisation, in other words, it increases the size of the contact interface between the two phases present in the extraction cell 100, due to which the rate of material transport in the cell accelerates, which in practice, from the point of view of chromatography, means an increase in plate number.

The internal surfaces 125 of the extraction cell 100 according to the invention defined by the planes b-c are preferably not smooth but roughened. The roughening is preferably formed by steps or saw teeth created on the surfaces 125, the height of which is preferably between 0.1 mm and 0.4 mm. According to our investigations, the roughening of the internal surfaces 125 slightly increases the atomisation of the mobile phase and reduces the adhesion of the mobile phase to the surfaces 125. The outlet 140 of the extraction cell 100 has one branch, i.e. it is not divided, also it has a circular cross-section. The size of the flow cross-section of the outlet 140 always exceeds the flow cross-section of an individual inlet branch. The simulation tests aimed at determining the flow pattern have clearly proven that an outlet 140 larger than the inlet 115 significantly reduces the dead volume occurring in the extraction cell 100. The asymmetric construction is a result of the different cross-sections of the outlet

140 and the inlet branches as well as due to there being a collection pool 130 of a determined size established between the extraction chamber 150 and the outlet 140. As a result of this, the liquid phase leaving the extraction chamber 150 flows through this collection pool 130 before leaving through the outlet 140. In the light of the simulation tests, the collection pool 130 in question is preferably hemispherical, the radius of which hemisphere exceeds the diameter of the outlet 140, however, it is smaller than any of the "a", "b" and "c" dimensions of the body containing the extraction cell 100. The diameter of the collection pool 130 is preferably equal to a half of the width "a" of the extraction cell 100. Using the outlet 140 with the collection pool 130 significantly reduces the amount of back-mixed mobile phase, and also improves the settling efficiency, i.e. the mobile phase volumeratio drops. This, from a chromatography point of view, reduces dead volume ratio and increases the stationary phase volume ratio. The solution in question also makes a greater flow rate possible, which increases the theoretical plate number and, with the increase in speed, increases the productivity of the CPC assembly.

On the basis of the simulation results (see Table 1), the technical solutions according to the invention, besides reducing dead volume, increase the size of the contact interface between the stationary and mobile phases, therefore two competing parameters are simultaneously improved.

Table 1. Comparison of small volume extraction cells (planned for a 250 ml CPC column volume)

Figure 8 depicts an example of another preferable embodiment of a new type asymmetrical CPC extraction cell 200 according to the invention in longitudinal cross-section and in perspective schematic view. This embodiment of the extraction cell 200 is preferably produced from peek material plates using CNC milling. The production work process is, to a certain extent, similar to the method presented in the US publication document No. US2010/0200488, where the halves of the cells are milled in individual plates, then following this, the whole cells are created by clamping together the two plates containing the two half-cells.

As compared to this solution, the main difference of the production process used by us is that the plates containing the milled cell halves are coated with a thin layer of a fluoropolymer (for example, by the heat-actuated continued polymerisation of a partially polymerised dispersion), then the coated plates are clamped together and subjected to heating, due to which the molecules of the polymer coating partially diffuse into one another and they adhere to one another forming appropriate insulation/sealing. In this way a single component is created in which the problem of leakage of the solutions between the layers does not appear. With respect to its geometry, the extraction cell 200 obtained in this way is very similar to the extraction cell 100 presented previously. The extraction cell 200 is primarily suitable for performing CPC processes where the flow rate of the mobile phase is a maximum of 275 ml/min. The structure of the extraction cell 200 is very similar to the extraction cell 100 presented in Figure 5. Accordingly, the cell wall 220 of the extraction cell 200 determines an extraction chamber 250, which extraction chamber 250 has an inlet 215 that opens into the extraction chamber 250 essentially perpendicularly, as well as an outlet 240 that serves to permit the liquid phase to leave the extraction chamber 250.

The inlet 215 may have one or more inlet branches, where all the inlet branches are each connected to the extraction chamber 250 via a circular inlet opening. All of the inlet openings in question are of the same size, their diameter is preferably between 0.5 mm and 2.0 mm. Particularly, Figure 8 shows an extraction cell 200 that has one inlet branch and, accordingly, one inlet opening. In the case of the latter embodiment the cross-section of the inlet branch in question perpendicular to the inlet direction is in the shape of a rectangle with rounded off corners (oval), where the dimensions of the rounded off rectangle are preferably between 1.0 mm and 2.0 mm. In such a case a transfer zone is preferably formed at the end of the inlet 215 more distant from the extraction chamber 250 - as illustrated in Figure 8 - which "transforms" the usually circular cross-sectioned channel connecting the extraction chamber 200 in question in series to another such extraction cell in a CPC rotor (see for example Figures 9 and 10) into the aforementioned inlet branch with an oval cross- section. It should be noted that this transfer zone is for creating a liquid connection between the internal flow cross-section of the aforementioned channel and the inlet branch in question of the inlet 215, and this usually has a geometrical structure of a narrowing nature.

The outlet 240 of the extraction cell 200 is also single-branched, i.e. it is not divided and also has a circular cross-section. The flow cross-section of the outlet 240 exceeds the flow cross-section of the inlet. From the point of view of reversibility of flow direction, the extraction cell 200 according to the invention is also asymmetric, in other words, when performing CPC processes, the direction of flow in the cell cannot be reversed. The asymmetric construction is a result of the different cross-sections of the outlet 240 and the inlet 215, as well as due to there being a collection pool 230 of a determined size established between the extraction chamber 250 and the outlet 240. As a result of this, the liquid phase leaving the extraction chamber 250 flows through this collection pool 230 before leaving through the outlet 240. In the light of the simulation tests, the collection pool 230 in question, unlike the hemispherical collection pool 130 used in the case of extraction cell 100, preferably designates an inclined surface extending along the entire width of the bottom of the extraction chamber 250, which inclined surface is connected to the cell wall 240 without any distinct angles.

On the basis of the simulation tests performed (see Table 2) the extraction cell 200 according to the invention produced by CNC milling and by being fused together has significant advantages as compared to the currently available cells, as beside reducing dead volume, it increases the size of the contact interface between the stationary and mobile phases, therefore two competing parameters are simultaneously improved. Apart from this, it is able to operate at a much larger flow rate as compared to the volume of the cell, therefore, its productivity is much higher than the productivity of traditional cells.

Table 2. Comparison of medium volume cells (planned for a 1 litre CPC column volume)

Cell Gyration Speed of Flow rate Specific Mobile phase radius (mm) rotation (ml/min.) interface volume

(rpm) (m ¹)

Reference 105 1400 25 241 10.09%

(Armen)

Reference 105 1400 30 333 25.19%

(Armen)

Own 450 750 275 824 16.43%

(No. 21)

Own 450 750 200 638 13.5%

(No. 22)

Figure 9 depicts a module 300 of a modularly constructed CPC rotor according to the invention containing several CPC extraction cells 100, 200 connected to each other in series. The module 300 in question has channels 330 providing fluid links for the extraction cells 100, 200 formed on a carrier 310 or in the carrier 310 as well as a single liquid input 320 and a single liquid output 340. Figure 10 depicts a large (r > 300 mm) disc CPC rotor 400 with an annular cross-section constructed using the CPC rotor modules 300 depicted in Figure 9. The modules 300 positioned on the sector segments 410 of the CPC rotor 400 are connected in series through tubes 430, where the liquid input of a selected module 300 is connected by tube 420 preferably to the liquid inlet located at the main axis of the CPC rotor 400, while the liquid outlet of the neighbouring module 300 is preferably connected by tube 420' preferably to the liquid output located at the main axis of the CPC rotor 400. In the case of the CPC rotor 400 according to the invention, the sum total of the extraction cells, channels and connections is produced from a single piece by plastic FDM 3D printing, or using similar technology. With this it is possible for the cell network to be made from fewer connected pieces as compared to the solutions according to the state of the art.

Implementation using plastic FDM 3D printing technology is very similar to a CNC process (robot with 3 or more axes), however, it is not a subtractive but an additive process, due to which the amount of waste created is significantly less, therefore this production process is more environmentally friendly and economical. Similarly to CNC procedures, devices suitable for working on large pieces are either very expensive or do not have the required degree of precision and speed. It is easy to realise that the series of cells positioned in annular circular sector shapes may have an external housing, with the help of which the elements may be easily positioned into an annular disc.

The advantages of the invention as compared to the state of the art are summarised in the following: According to the state of the art only various chromatography methods (HPLC and SMB) and liquid-liquid extraction methods were available for the purification of cyclosporine A. All three methods use an exceptionally large amount of solvent. A further disadvantage of chromatographic methods is that the material of the solid phase must be replaced from time to time, which as compared to the amount of product usually means the production of a large amount of solid waste that is contaminated at least by solvent. The advantage of the present invention as compared to these solutions is that it provides CsA at a very high level of purity, uses much less solvent and does not create solid waste.

The liquid-liquid counter-flow extraction method described in patent number US6620325 is much more efficient than the chromatographic methods, but this also has many disadvantages. The present invention overcomes these disadvantages, or significantly reduces their effect in such a way that the quality of the product created is better than in the case of the method described in the prior art. Namely,

- in the method according to the present invention cyclosporine A may be produced in one extraction step as opposed to the counter-flow method described in the prior art, in the case of which in the first step the contaminants less polar than cyclosporine A are removed and then in the second step the contaminants more polar that cyclosporine A are removed,

- the present invention does not require columns 25 metres high, or the related special architectural and engineering facilities, - it requires the use of 500 litres of solvent per kg of finished product, significantly less than the 1500 litres per kg required by counter-flow extraction, which in the case of 1 tonne of finished product reduces solvent consumption by 1000 tonnes, and also reduces the regeneration/disposal costs and the related environmental load and risk. - the present method does not require the use of heptane, and

- it is carried out at room temperature or at a temperature a few degrees lower, which means a huge reduction in investment and heating costs in the present method as compared to the counter-flow method.

- What is more, in addition to the above technology, economic and environmental protection advantages, using the present method high purity CsA can be reliably produced containing less than 0.3% H2CsA with a high level of yield even from a raw product containing more than 0.48% H2CsA.

The method according to the invention is presented in the following examples without the scope of protection being limited to these examples. 1. The identification of the CsA, H2CsA content using HPLC and the measurement of their quantity:

Measurement method:

• Sample preparation: Pipetting the solvent mixture in equal volumes, sample adding, mixing. After balance is achieved, sample taken with pipette diluted with eluent. · Device: HPLC-UV.

• Column used: Agilent Zorbax SB-C8 100x2.1 mm (1.8 μιη).

• Column temperature: 80 °C

• Eluent A: 40% acetonitrile, 53% water, 7% MTBE

• Eluent B: 52% acetonitrile, 41% water, 7% MTBE

· Gradient: full linear gradient in 15 minutes, then equilibration

• Detecting wavelength: 210 nm.

The selectivity measurement error is less than 2%. The error is derived from that the automatic pipette takes different volumes from the two phases. Volume measurement error is <5%. However, as selectivity is in focus here, which is the ratio between two partition coefficients, therefore on the basis of Gauss error propagation the selectivity error is approximately the square root of the partition coefficient error. Therefore, the error of the measured selectivities is certain to be under 2% (1.10 +/- 0.02). The HPLC chromatograms measured with the above method (page 12).

Example 1: purification of cyclosporine A

Production of the biphase:

Production of the upper phase:

The necessary amount of hexane and acetone are mixed in proportions so that the density is between 0.68 and 0.73, preferably between 0.700 and 0.702 g/ml at 20 °C.

Production of the lower phase:

The necessary amount of acetone and water are mixed in proportions so that the density of the mixture is between 0.82 and 0.90 g/ml, preferably 0.875 and 0.885 g/ml at 20 °C.

The lighter phase is used as the mobile phase and the heavier phase and the stationary phase in the following separation.

CPC extraction:

100 grams of raw cyclosporine A was dissolved in 1.6 litres of lighter phase. The column volume used was 1.4 litres. With the device rotated at 50-100 g the column was filled with stationary phase (heavier phase) at a flow rate of 200 ml/min. In the meantime the device was gradually accelerated to 220 g. The sample was injected for 1 minute at a flow rate of 50 ml/min. Following this the mobile phase was pumped for 2 minutes at a flow rate of 20 ml/min. then for 25 minutes at a flow rate of 40 ml/min. Following this the pumping was switched over to the stationary phase, which was pumped at a rate of 40 ml/min., and after 5 minutes nine 50-ml fractions were taken. The process was repeated 20 times from injecting the sample. The ambient temperature was 19-20 °C volt. The H2CsA content of the initial sample was 0.50%. After grouping and combining the obtained fractions CsA of the following purities were obtained with the following yields:

CsA with 80% yield (H2CsA content 0.28%/ HPLC)

CsA with 60% yield (H2CsA content 0.25%/ HPLC)

CsA with 40% yield (H2CsA content 0.24%/ HPLC).

With the method according to the example 1 kg of product containing 0.27% H2CsA was obtained from 2 kg of raw CsA containing 75% active substance.

Claims

1. Method for purifying cyclosporine A, characterised by that a centrifugal partition chromatography method is used for the purification.

2. The method according to claim 1, characterised by the method being carried out by a. ) filling the CPC rotor with liquid constituting the stationary phase, then b. ) injecting the raw cyclosporine solution while the CPC rotor is rotating, then following this c. ) having mobile phase flow through the CPC rotor while it is rotating, following which d. ) stationary phase is once again filled into the CPC rotor CPC rotor while it is rotating, then, optionally, the steps b.), c.) and d.) are repeated until the desired amount of cyclosporine solution has been injected, and in the meantime a fraction or fractions are taken from the mixture exiting from the CPC rotor in steps c.) and d.) and then that fraction or those fractions are selected in which the H2CsA content is < 0.3%, preferably < 0.27%, more preferably < 0.25% with respect to the CsA content, then the cyclosporine A is separated from the selected fraction/fraction mixture.

3. The method according to claim 2, characterised by feeding in cyclosporine solution equivalent to 2-10%, preferably 4-5% of the volume of the column in step b.), by feeding in mobile phase equivalent to 50-150%, preferably 60-80% of the volume of the column in step c), and feeding in stationary phase equivalent to 20-100%, preferably 30-40% of the volume of the column in step d.).

4. The method according to any of claims 1 to 3, characterised by that a biphase mixture of a lighter and a heavier phase is used as the stationary and mobile phases made up of a mixture of liquids in balance with each other, immiscible with each other or only partially miscible, which phases have partition selectivity for CsA - H2CsA greater than 1.05, preferably equal to or greater than 1.1.

5. The method according to any of claims 1 to 4, characterised by that the lighter phase is used as the mobile phase.

6. The method according to any of claims 1 to 5, characterised by that the partition selectivity of CsA - H2ScA with respect to the lighter phase is greater than 1.05, preferably 1.1 or larger.

7. The method according to claims 4 to 6, characterised by that an apolar aprotic solvent is used as the one component of the solvent mixtures forming the two phases in balance, preferably an aromatic, aliphatic or cycloaliphatic hydrocarbon, aliphatic ether or a mixture thereof, and water is used as the other component and, optionally, as a further component of the mixture a polar protic and/or dipolar aprotic solvent is also used.

8. The method according to claim 7, characterised by that pentane, hexane, heptane, octane or their isomers, most preferably hexane are used as the aliphatic hydrocarbon in the mixture, alkyl ethers, preferably methyl tert butyl ether is used as the aliphatic ether in the mixture, aliphatic alcohols, preferably alcohols with C1-C4 carbon chains, even more preferably methanol, ethanol and isopropyl alcohol, n-butyl alcohol are used as the polar protic solvent in the mixture, preferably ketones, esters of carboxylic acids with C1-C4 carbon chains formed with alcohols with C1-C4 carbon chains, more preferably ethyl acetate or acetone, and most preferably acetone are used as the dipolar aprotic solvent.

9. The method according to claim 8, characterised by that the composition of the solvent mixtures forming two phases in balance and their preferable volume ratios are selected form among the following: Hexane - MTBE - MeOH - Water (4:5:5:5), Hexane - MTBE - MeOH - Water (4:6:6:5), Hexane - Acetone - Water (3:5:2), Hexane - Acetone - Water (2:6:2), Hexane - IPA - Water (4:4:2) Hexane - EtAc - MeOH - Water (4:5:5:4), Hexane - EtAc - MeOH - Water (4:6:6:4) Hexane - Acetone - Water (2:3:1), Pentane - Acetone - Water (2:3:1), Heptane - Acetone - Water (2:3:1) Heptane - Acetone - Water (2:2.5:1).

10. The method according to claim 7, characterised by that the system forming the two phases contains a mixture of n-hexane, acetone and water and the density of the upper phase is between 0.68 and 0.73, preferably between 0.700 and 0.702 g/ml at 20 °C and the lower phase is in balance with this and its density is between 0.82 and 0.90 g/ml, preferably between 0.875 and 0.885 g/ml at 20 °C.

11. The method according to any of claims 1 to 10, characterised by that the method is performed at a temperature between -10 and +50 °C, preferably between 0-30 °C and most preferably between 15-20 °C, as long as the upper phase is the mobile phase and the lower phase is the stationary phase, or at a temperature between +10 and +50 °C, preferably between 20-40 °C and most preferably between 25-35 °C, as long as the lower phase is the mobile phase and the upper phase is the stationary phase.

12. The method according to any of claims 1 to 11, characterised by that the CPC rotor is rotated at a speed so that the centripetal acceleration occurring due to the rotation at the point furthest from the axis of rotation in the cell in the rotor that is the furthest from the axis of rotation > 150 g, preferably between 180-500 g, even more preferably between 200- 350 g and most preferably at a value between 200-250 g.

13. The method according to claims 1 to 12, characterised by that a CPC rotor is used for the method that has a. ) at least one extraction cell (100, 200), a cell wall (120, 220) delimiting an enclosed extraction chamber (150, 250) and the differing cross-sections of the inlet (115, 215) and the outlet (140, 240) located at substantially opposing parts of the cell wall (120, 220) formed to ensure the liquid connection between the extraction chamber (150, 250) and the space outside of the extraction cell (100, 200) are such that when the centrifugal partition chromatograph is in operating status the passage cross-section of the outlet (140, 240) exceeds the passage cross-section of the inlet (115, 215) and optionally the inlet (115, 215) is formed as at least one inlet branch (115a, 115b), furthermore b. ) the cell wall (120, 220) forms a substantially rectangular based inclined prism, the angle of inclination of which, with the extraction cell (100, 200) in its position in the centrifugal partition chromatograph, being selected to minimise the Coriolis force occurring as a result of the rotation when the extraction cell (100, 200) is in its operation status, and preferably this angle of inclination is between 5° and 30°, more preferably between 15° and 18°, furthermore c.) the collection pool (130) is formed as a protruding hemispherical part of the cell wall (120), and the outlet (140) is located on this hemisphere, or the collection pool (230) is formed as a protruding section of the extraction chamber (250) delimited by a part of the cell wall (220) extending in an inclined way in the direction of the outlet (240) and optionally d.) at least a part of the surface (125) of the cell wall (120) bordering the extraction chamber (150) is roughened, preferably the roughening is ensured by steps or saw teeth on the surface (125).

14. The method according to claim 13, characterised by that the method is performed with a CPC rotor which has a modular construction realised with essentially identical modules (300), where all of the modules (300) contain more than one extraction cell (100, 200) connected by channels (330) providing a fluid connection between them, furthermore the individual modules (300) are connected in series with each other via tubes (430).