METHOD / DEVICE FOR TARGET COMPOUND PURIFICATION

Abstract

The present invention relates to a separation method comprising: i) providing an aqueous solution comprising a target compound; ii) applying a separation step to the aqueous solution, thereby providing a plurality of fractions of the aqueous solution: iii) determining a concentration parameter indicating a concentration of the target compound in at least part of the fractions; iv) determining a nuclear magnetic resonance (NMR) parameter by applying an NMR measurement to the fractions, the NMR parameter indicating a nuclear magnetic spin relaxation in said at least part of the fractions; and v) determining a target parameter of said at least part of the fractions based on the concentration parameter and the nuclear magnetic resonance parameter. The present invention further relates to separation systems, uses, preparations, and methods related thereto.

Claims

1. A separation method comprising: i) providing an aqueous solution comprising a target compound; ii) applying a separation step to the aqueous solution, thereby providing a plurality of fractions of the aqueous solution; iii) determining a concentration parameter indicating a concentration of the target compound in at least part of the fractions; iv) determining a nuclear magnetic resonance (NMR) parameter indicative of a transverse nuclear magnetic spin relaxation of water by applying an NMR measurement to the fractions; and v) determining a target parameter of said at least part of the fractions based on the concentration parameter and the NMR parameter.

2. The method of claim 1, wherein the NMR parameter is indicative of a transverse nuclear magnetic spin relaxation of protons in water.

3. The method of claim 1, wherein the NMR parameter comprises at least one of a transverse nuclear magnetic spin relaxation time T.sub.2 and a transverse nuclear magnetic spin relaxation rate R.sub.2.

4. The method of claim 1, wherein the concentration parameter is directly proportional to the concentration of the target compound.

5. The method of claim 1, wherein the concentration parameter is not an NMR parameter indicative of a transverse nuclear magnetic spin relaxation of water.

6. The method of claim 1, wherein the target parameter is directly proportional to the concentration parameter and to the NMR parameter being the transverse relaxation time (T.sub.2).

7. The method of claim 1, wherein the target parameter is directly proportional to the product of the concentration parameter and the NMR parameter being the transverse relaxation time (T.sub.2).

8. The method of claim 1, further comprising step vi) identifying, based on the target parameter, fractions comprising the target compound at a desired purity.

9. The method of claim 8, further comprising step vii) combining at least two of the fractions comprising the target compound at a desired purity.

10. The method of claim 1, wherein the separation step in step ii) comprises a chromatographic separation, wherein the fractions are fractions of an eluate of the chromatographic separation.

11. The method of claim 1, wherein the separation step in step ii) comprises size-exclusion chromatography or ion exchange chromatography.

12. The method of claim 1, wherein the method is a continuous in-line method and/or wherein, in step ii), the fractions are generated as a continuous stream of liquid.

13. The method of claim 1, wherein the target compound comprises, in an embodiment is, a compound selected from the group consisting of (i) a polypeptide, (ii) a polynucleotide, (iii) a complex of one of (i) or (ii); and (iv) a conjugate of one of (i) to (iii).

14. The method of claim 1, wherein the target compound comprises, in an embodiment is, a polypeptide.

15. The method of claim 1, wherein the target compound comprises, in an embodiment is, a non-aggregated polypeptide.

16. The method of claim 1, wherein said target compound is a virus or a virus-like particle.

17. The method of claim 1, wherein said concentration parameter is determined in a direct photometric assay.

18. The method of claim 1, wherein the concentration parameter is the absorption, extinction, or fluorescence of the target compound.

19. A separation system, comprising: a) a separation device configured for applying a separation step to an aqueous solution comprising a target compound, thereby providing a plurality of fractions of the aqueous solution; b) a concentration determining device configured for determining a concentration parameter indicating a concentration of the target compound in at least part of the fractions; c) a nuclear magnetic resonance (NMR) measurement device configured for determining an NMR parameter indicative of a transverse nuclear magnetic spin relaxation of water by applying an NMR measurement to the fractions, and d) an evaluation device configured for determining a target parameter of said at least part of the fractions based on the concentration parameter and the NMR parameter.

20. The system of claim 19, the system further comprising: e) a liquid distribution element for combining at least two of the fractions comprising the target compound and/or an output device outputting the values of the target parameter determined in step d).

21. (canceled)

22. A method of production of a target compound, in particular a polypeptide, comprising the steps of the method according to claim 1.

23. A preparation of a target compound produced or producible according to the method according to claim 1.

24. A method for increasing concentration of a target compound in a solution by diafiltration, comprising A) determining a first gradient of a nuclear magnetic resonance parameter (NMR parameter) indicative of a transverse nuclear magnetic spin relaxation of water over time in said solution or in a fraction thereof, B) determining a second gradient of said NMR parameter over time in said solution or in a fraction thereof, and C) at least temporarily decreasing diafiltration rate and/or increasing stirring rate in case the value of said second gradient of said NMR parameter deviates at least 10%, from the value of said first gradient of said NMR parameter.

Description

SHORT DESCRIPTION OF THE FIGURES

[0199] Further optional features and embodiments will be disclosed in more detail in the subsequent description of embodiments, preferably in conjunction with the dependent claims. Therein, the respective optional features may be realized in an isolated fashion as well as in any arbitrary feasible combination, as the skilled person will realize. The scope of the invention is not restricted by the preferred embodiments. The embodiments are schematically depicted in the Figures.

[0200] Therein, identical reference numbers in these Figures refer to identical or functionally comparable elements.

[0201] In the Figures:

[0202] FIG. 1 shows schematically an exemplary setup of a separation system, comprising a separation device 110, e.g. configured as a chromatography column having a sample and mobile phase inlet 118: a concentration determining device 112 and an NMR measurement device 114 performing in-line measurements on the eluate from separation device 110 and providing measurement data to evaluation device 116, said evaluation device 116 optionally directing a valve in a liquid distribution element 120; as an alternative, liquid distribution element 120 may operate autonomously and may collect fractions 122 e.g. over predefined intervals.

[0203] FIG. 2 shows an overview of compounds (Antibody Like Proteins (ALPs)) used in the Examples; M: metal.

[0204] FIG. 3 shows a graphical representation of a variety of parameters in purification of a therapeutic antibody by cation exchange chromatography; parameters are: norm c[protein]: protein concentration normalized to the highest measured protein concentration; norm T2: T2 value normalized to the highest measured T2 value; norm (T*C[protein]): normalized product of T2 and c[protein]; norm T2 (buffer w/o protein): normalized T2 value from a test run in the absence of the antibody; and norm T2-norm T2 (buffer w/o protein): norm T2 from which norm T2 (buffer w/o protein) was subtracted.

[0205] FIG. 4 shows wNMR relaxation T2 (y-axis) in dependence of increase of protein concentration (as fold volume reduction during the process, x-axis) for non-stressed (solid line) and artificially stressed (dashed line) protein solutions. The inserts show the appearance of the solutions, arrows indicate starting aggregation.

[0206] FIG. 5 Dependency of T2 on the DNA-loading of AAV particles of the serotype 2 (AAV2).

[0207] FIG. 6 Dependency of T2 on the DNA-loading of AAV particles of the serotype 6 (AAV6).

[0208] FIG. 7 Dependency of T2 on the DNA-loading of AAV particles of the serotype 8 (AAV8).

DETAILED DESCRIPTION OF THE EMBODIMENTS

Example 1: Materials and Methods

[0209] Antibody Like Proteins

[0210] An overview of the antibody Like Proteins (ALPs) used in the Examples is given in FIG. 2. Cibisatamab (ALP1) is a 2+1 bispecific antibody targeting carcinoembryonic antigen (CEA) and CD3, as described in WHO Drug Information (International Nonproprietary Names for Pharmaceutical Substances), Recommended INN: List 80, 2018, vol. 32, no. 3, p. 438; IgG-IL2 (ALP2) is a fusion protein between an IgG and two IL-2 peptides fused to the heavy chains, respectively, as described in WO 2015/118016 A1. PD1-IL2 (ALP3) is an anti-PD1 antibody C-terminally fused to an IL-2 peptide, as described in WO 20181184964 A1. ALP4 is a 2+1 bispecific fusion protein comprising an IgG structure fused to V.sub.HVL-domains and to a metal complex.

[0211] NMR

[0212] Transverse relaxation rates (R2) or times (T2) were recorded with a Bruker minispec mq20 spectrometer (20 MHz; Bruker BioSpin GmbH, Rheinstetten, Germany). The spectrometer was equipped with a 0.47 T magnet and a H20-10-25AVGX4 probe. At least 4 acquisitions were measured for each sample with sample volumes of 900 μl at 20° C. To determine T2 or R2, signal decay was followed for at least 5 sec.

[0213] Preparation of Protein Load Solutions

[0214] The proteins used were expressed from stably transfected CHO cells. Fermentation supernatants were applied to MabSelect SuRe columns (GE Healthcare Europe GmbH, Freiburg, Germany) and eluted with acidic buffer solutions with a pH<5. After elution the protein pool was adjusted to a pi of 5.0 to 5.5 by adding adequate amounts of 1 M Tris solution.

[0215] Chromatography

[0216] SEC

[0217] Prior to size exclusion chromatographies the concentration of the protein A pools was increased to >20 mg/mi by ultrafiltration using an Amicon stirring cell with ultrafiltration discs having a 10 kDa or 30 kDa cut-off (Merck KGaA, Darmstadt, Germany). Size exclusion chromatography was done by loading 30 mg to 180 mg of protein A purified protein samples to a Superdex 20W prep grade column (60 cm column height, 1.6 cm diameter; GE Healthcare Europe GmbH, Freiburg. Germany). Chromatography runs were performed with a flow rate of 1.6 ml/min in 20 mM His/His-HCL, 140 mM NaCl, pH 5.5. During elution the absorption at 280 nm was recorded and fractions were collected.

[0218] CEX

[0219] For cation exchange chromatography (CEX) the proteins were loaded on 10 ml Poros XS columns (20 cm column height, 0.8 cm diameter; Thermo Fisher, Scientific, Waltham, Mass., USA) with a load density between 25 mg/ml and 30 mg/ml.

[0220] For conductivity gradient elution, the column was equilibrated with 40 mM acetate pH 5.5. Then the protein A purified protein solution was loaded with 2.0 ml/min and the column was washed with 3 column volumes (CV) of equilibration buffer. For elution of the bound proteins an acetate gradient up 500 or 750 mM within 20 or 15 CV, respectively, and a flow rate of 2.5 ml/min was used at pH 5.5 or 6.5. During elution adsorption at 280 nm was recorded and fractions were collected.

[0221] For pH gradient elution, the column was equilibrated with 20 mM citrate, 20 mM sodium phosphate, 20 mM Tris, pH 5.0 and a NaCl concentration between 50 mM and 100 mM. Then the protein A purified protein solution was loaded with 2.0 ml/min and the column washed with 5 CV equilibration buffer. For elution of the bound proteins a pH gradient from 5.0 to 8.5 within 20 to CV and a flow rate of 2.5 ml/min was used. During elution adsorption at 280 nm was recorded and eluate fractions were collected.

[0222] In addition, buffer control runs were performed for conductivity and pH gradient elutions. For these runs, the above described procedures were used, but no protein loaded to the column.

[0223] Analytical SEC

[0224] Pools and fractions of the chromatographies were analyzed by analytical SEC. To that end, approximately 25 μg (but not more than 100 μl) of protein were injected on a TSKgel UP-SW3000 column (Tosoh Bioscience, Griesheim. Germany) and subjected to isocratic chromatography in 200 mM potassium phosphate, 250 mM potassium chloride, pH 6.2 with a flow rate of 0.25 ml/min. Purities were calculated based on absorption at 280 nm with the Software Chromeleon 7 (Thermo Fisher, Scientific, Waltham, Mass., USA).

[0225] UF

[0226] Ultrafiltration was done with a manual benchtop device with a 400 ml vessel and a 30 kDa Sartorius (Sartorius A G, Göttingen, Germany) membrane. The device was regenerated with 0.5 M NaOH, flushed with water and equilibrated with the corresponding buffer until pH and conductivity of the equilibration buffer was constant. Subsequently, the equilibration buffer was exchanged by the protein solution. Ultrafiltration was run under pressure control of 1-1.2 bar.

[0227] Samples were taken after defined volume reductions. Protein concentration. SE-HPLC and wNMR T2 were determined for each sample. Two sets of experiments were run. One at RT and one under artificial stress conditions (50° C.). Artificial stress was applied to induce rapid HMW formation at a specific concentration.

Example 2

[0228] As shown in FIG. 1, in the separation system the sample 118 is applied to the separation device 118, which may be an LC column, membrane, or filter, an ultracentrifuge or centrifuge beaker, an ultrafiltration device, or the like. By the separation step and by means of liquid distribution element 120, fractions 122 are formed of which at least two differ in composition. In at least part of the fractions or in aliquots thereof, concentration of the target compound or of the class of compound the target compound belongs to is determined by the concentration determining device 112, which may in particular be a photometer, e.g. a UV-photometer. Also in at least part of the fractions or in aliquots thereof, an NMR parameter is determined by the NMR measurement device 114, which may in particular be a benchtop NM R device. Determination of the concentration parameter and/or the NMR parameter may be performed before the actual formation of fractions by the liquid distribution element 120, as shown in FIG. 1, e.g. by flow-through measurement, or may be performed after formation of the fractions in the established fractions or aliquots thereof. Based on the values of the concentration parameter and of the NMR parameter, the evaluation device 116 determines a target parameter.

Example 3

[0229] FIG. 3 shows a graphical representation of the change of a variety of parameters over the elution volume in purification of a therapeutic antibody by cation exchange chromatography. As will be appreciated, the standard measurement of protein concentration by absorption measurement at 280 nm (solid line) provides a broad peak making it impossible to determine which fractions comprise the desired monomeric antibody. In contrast, the graphical representation of the normalized product of T2 and c[protein] (short-dashed line) show a much narrower peak, excluding in particular undesired aggregates of the antibody.

Example 4

[0230] Four different Antibody Like Proteins (ALPs, FIG. 2, Table 1) were applied to LC purification by cation exchange chromatography using either an acetate or a pH gradient, or by size exclusion chromatography (SEC). After chromatography, fractions were pooled according to c[protein] determination (UV measurement) alone, according to normalized T2 alone, or according to the normalized product of T2 and c[protein] (Table 2). The composition of the pools used for chromatography is shown in Table 1. As is clear from the data, using a combination of T2 and c[protein] as the target parameter provides for improved purity of the product and in particular improved removal of high and/or low molecular weight contaminants.

TABLE-US-00001 TABLE 1 Compositions of the protein solutions applied to the indicated LC columns; ALP: Antibody Like Protein; HMW: High molecular weight (aggregates), LMW: low molecular weight (degradation products), na: not applicable/not determined, SEC: size exclusion chromatography. Target compound Column HMW Dimer Monomer LMW ALP1 SEC 15.6 na 81.4 2.93 ALP1 Poros XS 10.2 na 85.0 4.81 ALP2 SEC 9.53 na 90.0 0.44 ALP3 Poros XS 5.58 na 93.1 1.34 ALP4 SEC 12.6 19.3 68.0 na ALP4 Poros XS 10.5 21.75 67.7 na

TABLE-US-00002 TABLE 2 Compositions of the protein solutions applied to the indicated LC steps after pooling according to c[protein] determination (UV measurement) alone (c[protein]), according to normalized T2 alone (normalized T2), and according to the normalized product of T2 and c[protein] (T2 and c[protein]): HMW: high molecular weight (aggregates), LMW: low molecular weight (degradation products), na: not applicable/not determined, SEC: size exclusion chromatography. c[protein] normalized T2 T2 and c[protein] Target Mono- Mono- Mono- com- Col- Yield HMW Dimer mer LMW Yield HMW Dimer mer LMW Yield HMW Dimer mer LMW pound umn Gradient [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] [%] ALP1 Poros pH 84 4.7 na 93 2.2 69 3.5 na 95 1 43 2.1 na 97 1.0 XS ALP2 SEC isocratic 84 1.0 na 99 1.0 ALP3 Poros pH 72 4.0 na 95 1.4 69 2.4 na 96 1.4 43 1.3 na 98 1.0 XS 61 1.4 2.6 96 na ALP4 SEC isocratic ALP4 Poros acetate 82 7.0 21 72 na 79 7.1 21 72 na 62 6.1 17 77 na XS ALP4 Poros pH 88 7.0 21. 72 na 81 5.6 21 73 na 64 2.5 14 83 na XS

Example 4

[0231] Protein solutions at various concentrations were artificially stressed to induce aggregation, or not, and wNMR-T2 values were measured. As shown in FIG. 4, starting aggregation causes a rapid drop in the T2 signal, which can be used to detect starting aggregation in ultrafiltration methods.

Example 5: Filled and Empty AAV Capsids

[0232] Materials and Methods

[0233] AAV Particles

[0234] DNA-loaded AAV particles as well as empty AAV particles of different serotypes were purchased from Virovek, Hayward, Calif., United States of America. For filled AAV particles, the concentration of vg/mL (virus genomes per mL) equals the concentration of vp/mL (virus particles per mL) as one genome is packaged in one particle; for empty AAV particles, the number of vp/mL can be determined e.g. by comparing the amount of capsid protein(s) to the amount of capsid protein(s) in a preparation of filled AAV particles with a known concentration in vg/mL. The DNA is of the DNA-loaded AAV particles comprised an expression cassette for green fluorescent protein with a CMV promoter (CMV-GFP):

[0235] AAV8-empty (Lot 19-564E); AAV8-CMV-GFP (Lot 18-737) in 1×PBS buffer containing 0.001% pluronic F-68, and 0.22 μm filter sterilized.

[0236] AAV2-empty (Lot 19-604E)/AAV2-CMV-GFP (Lot 17-600) in 1×PBS buffer containing 0.001% pluronic F-68, 100 mM sodium citrate and 0.22 μm filter sterilized.

[0237] AAV6-empty (Lot 19-540E)/AAV2-CMV-GFP (Lot 19-718) in 1×PBS buffer containing 0.001% pluronic F-68, 100 mM sodium citrate and 0.22 μm filter sterilized.

[0238] NMR

[0239] Transverse relaxation rates (R.sub.2) or times (T.sub.2) were recorded as specified herein above in Example 1.

[0240] Preparation of AAV Samples

[0241] The samples were generated by mixing full and empty AAV2, AAV6 and AAV8 particles, respectively, having the same concentration (2*10.sup.13 vg/ml) dissolved in the same aqueous solution comprising 1×PBS buffer containing 0.001% (w/v) pluronic F-68 (for AAV8) and containing 0.001% (w/v) Pluronic F-68 as well as, for AAV2 and AAV6, 100 mM sodium citrate at different ratios. Samples were measured without further longtime storage e.g. <0° C.

Example 5.1

[0242] Determination of the Transverse Relaxation Times for Different Ratios of DNA-Loaded and Empty AAV Particles

[0243] DNA-loaded AAV particles and empty AAV particles were mixed in aqueous solution (AAV2 and AAV6: 1×PBS buffer containing 0.001% a (w/v) pluronic F-68, 100 mM sodium citrate; AAV8 same buffer without sodium citrate) to result in different full/empty ratios spanning the range from 0% to 100% DNA-loaded AAV particle. This has been done for AAV particles of the to serotypes 2, 6 and 8. For the individual samples, the transverse relaxation time, (T.sub.2) were recorded as outlined in the Materials and Methods section above. The results are presented in the following Tables.

TABLE-US-00003 TABLE 3 T.sub.2 values for differently DNA-loaded AAV2 particle samples. concentration fraction transverse relaxation time [ms] ratio of DNA-loaded AAV2 standard full/empty particles experiment 1 experiment 2 average deviation  0:100 0 vg/mL 0.00 1559.9 1566.9 1563.4 3.5 25:75 0.5 * 10.sup.13 vg/mL 0.25 1605.8 1568.0 1586.9 18.9 50:50 1 * 10.sup.13 vg/mL 0.5 1627.0 1602.4 1614.7 12.3 75:25 1.5 * 10.sup.13 vg/mL 0.75 1677.6 1662.4 1670.0 7.6 100:0  2 * 10.sup.13 vg/mL 1.00 1750.6 1770.3 1760.5 9.9

TABLE-US-00004 TABLE 4 T.sub.2 values for differently DNA-loaded AAV6 particle samples. concentration fraction transverse relaxation time [ms] ratio of DNA-loaded AAV6 standard full/empty particles experiment 1 experiment 2 average deviation  0:100 0 vg/mL 0.00 1629.9 1614.2 1622.1 7.9 25:75 0.5 * 10.sup.13 vg/mL 0.25 1664.6 1629.1 1646.9 17.8 50:50 1 * 10.sup.13 vg/mL 0.5 1700.7 1073.3 1687.3 13.5 75:25 1.5 * 10.sup.13 vg/mL 0.75 1688.5 1688.5 1688.5 0.0 100:0  2 * 10.sup.13 vg/mL 1.00 1757.5 1786.1 1771.8 14.3 The value for the 75:25 ratio was excluded for the fitting calculation due to being deemed to be an experimental error.

TABLE-US-00005 TABLE 5 T.sub.2 values for differently DNA-loaded AAV8 particle samples. transverse ratio concentration fraction relaxation time full/empty of DNA-loaded AAV8 particles [ms]  0:100   .sup.  0 vg/mL 0.00 1704.2 25:75 0.5*10.sup.13 vg/mL 0.25 1668.0 50:50   1*10.sup.13 vg/mL 0.5 1635.6 75:25 1.5*10.sup.13 vg/mL 0.75 1608.9 100:0    2*10.sup.13 vg/mL 1.00 1497.3

[0244] The obtained transverse relaxation times were fitted using linear and second order polynomial functions. The results are presented in the following Table 6.

TABLE-US-00006 TABLE 6 Fitting results. For the serotype 6 the fitting is shown including (in brackets) and excluding the data for the 75:25 ratio. AAV serotype function R.sup.2 value linear fitting 2 1.9088*x + 1543.7 0.9208 6 1.5298*x + 1615.1 0.9905 (1.3646*x + 1615.1) (0.8984) 8 −1.8916*x + 1717.4   0.9026 2.sup.nd order polynomial fitting 2 0.0184*x.sup.2 + 0.0642*x + 1566.7 0.996  6 0.0047*x.sup.2 + 1.0479*x + 1620.9 0.9987 (0.0089*x.sup.2 + 0.4749*x + 1626.2) (0.9318) 8 −0.0166*x.sup.2 − 0.2333*x + 1696.7   0.9633

[0245] FIG. 5 shows the dependency of T2 on the DNA-loading of AAV particles of the serotype 2 (AAV2), i.e. the dependency of T2 on the concentration of DNA-containing AAV2 particles in an aqueous sample. It can be seen that the absolute difference in the relaxation time between empty AAV2 particles (0% full) and completely DNA-loaded AAV2 particles (100% full) at a concentration of 2*10.sup.13 particles/mL is 197.1±6.4 ms and the relative difference is 12.6±0.4%. The data points can be fitted with a second order function with an R.sup.2 value of 0.996.

[0246] FIG. 6 shows the dependency of T2 on the DNA-loading of AAV particles of the serotype 6 (AAV6), i.e. the dependency of T2 on the concentration of DNA-containing AAV6 particles in an aqueous sample. It can be seen that the absolute difference in the relaxation time between empty AAV6 particles (0% full) and completely DNA-loaded AAV6 particles (100% full) at a concentration of 2*10.sup.13 particles/mL is 149.8±22.1 ms and the relative difference is 9.2±1.4%. The data points can be fitted with a second order function with an R.sup.2 value of 0.9987.

[0247] FIG. 7 shows the dependency of T2 on the DNA-loading of AAV particles of the serotype 8 (AAV8), i.e. the dependency of T2 on the concentration of DNA-containing AAV8 particles in an aqueous sample. It can be seen that the absolute difference in the relaxation time between empty AAV8 particles (0% full) and completely DNA-loaded AAV8 particles (100% full) at a concentration of 2*10.sup.13 particles/mL is 206.9 ms and the relative difference is 14%. The data points can be fitted with a second order function with an R.sup.2 value of 0.9633.

LITERATURE

[0248] Feng et al. (2015), Chem. Commun. 51: 6804 [0249] Hills et al. (1989), Molecular Physics, 67(4):903-918) [0250] Metz and Mäder (2008), International Journal or Pharmaceutics 364:170-175 [0251] Shigemitsu et al. (2016), Analytical Biochemistry 498:59-67 [0252] Taraban et al. (2015), Journal of Pharmaceutical Sciences 104:4132-4141 [0253] Taraban et al. (2017), Analytical Chemistry 89:5494-5502 [0254] Taraban et al. (2019), 7th Annual PANIC Conference, Poster presentation P35: “Water Flow-NMR—A Prospective Contact-Free In-Line Analytical Tool for Continuous Biomanufacturing” [0255] Taraban et al. (2019). Anal Chem 91(6):4107 [0256] WO 2012/015912A1 [0257] WO 2014/169229 A1 [0258] WO 2015/118016 A1 [0259] WO 2018/102681 A1 [0260] WO 2018/184964 A1 [0261] WO 2019/016154A1

LIST OF REFERENCE NUMBERS

[0262] 110 separation device [0263] 112 concentration determining device [0264] 114 NMR measurement device [0265] 116 evaluation device [0266] 118 sample (aqueous solution)/mobile phase inlet [0267] 120 liquid distribution element [0268] 122 fractions