METHOD FOR DETERMINING IMPURITIES IN POLYALKYLENE ETHERS OR POLYALKYLENE AMINES AND USE THEREOF

Abstract

Disclosed is a method for the determination of impurities in polyalkylene ethers and polyalkylene amines comprising the steps i) introducing polyalkylene ethers or polyalkylene amines as an analyte into a chromatography column containing monolithic silica gel as a stationary phase, ii) eluting the analyte with a liquid elution agent having such a polarity that the analyte is in adsorptive equilibrium with the stationary phase during chromatography, iii) detecting the components of the analyte at the exit-side end of the chromatography column, receiving a chromatogram, which shows different components of the analyte and its qualitative amount depending on the elution time of the individual components, and iv) identifying bands in the chromatogram having a low height or area compared to the band with the largest height or area as an indication of the presence of impurities in the analyte.

The method allows in an easy manner to identify impurities in the sample. The method can be used for quality control but also for the preparative cleaning of the sample.

Claims

1. A method for the determination of impurities in polyalkylene ethers and polyalkylene amines comprising the steps i) introducing polyalkylene ethers or polyalkylene amines as an analyte into a chromatography column containing monolithic silica gel as a stationary phase, ii) eluting the analyte with a liquid elution agent having such a polarity that the analyte is in adsorptive equilibrium with the stationary phase during chromatography, iii) detecting the components of the analyte at the exit-side end of the chromatography column, receiving a chromatogram, which shows different components of the analyte and its qualitative amount depending on the elution time of the individual components, and iv) identifying bands in the chromatogram having a low height or area compared to the band with the largest height or area as an indication of the presence of impurities in the analyte.

2. The method according to claim 1, characterized in that the polyalkylene ether is a polyethylene glycol or a polypropylene glycol having hydroxyl end groups which may be partially or completely etherified.

3. The method according to claim 2, characterized in that the polyalkylene ether is a polyethylene glycol or a polypropylene glycol containing alkoxy end groups, in particular methoxy end groups.

4. The method according to claim 1, characterized in that the polyalkylene amine is an aliphatic polyoxazoline having amino, hydroxy, alkyl and/or aryl end groups, wherein the amino end groups may be partially or completely alkylated.

5. The method according to claim 1, characterized in that the impurities in the polyalkylene ethers are polyols whose mean molar mass is in the same range or below the mean molar mass of the polyalkylene ethers to be analyzed, preferably alkylene diols and/or di- or trialkylene glycols.

6. The method according to claim 1, characterized in that the impurities in the polyalkylene amines are polyamines whose mean molar mass is in the same range or below the mean molar mass of the polyalkylene amines to be analyzed, preferably alkylene diamines and/or di- or trialkyleneimine diamines.

7. The method according to claim 1, characterized in that the monolithic silica gel is designed as a reversal phase and that the eluent contains water and an organic solvent in such a quantity that the elution of the analyte takes place as adsorption chromatography.

8. The method according to claim 7, characterized in that the eluent contains water and acetonitrile.

9. The method according to claim 1, characterized in that the chromatographic method is a high-pressure liquid chromatography.

10. The method according to claim 1, characterized in that the detection of the components of the analyte at the exit-side end of the chromatography column is carried out by light scattering detection (ELSD) or by UV/VIS spectroscopy.

11. The method according to claim 1, characterized in that the height of the bands with a low height is less than 10%, in particular less than 5% of the height of the band with the highest height.

12. Use of the method according to claim 1 for quality control of polyalkylene glycols or polyalkylene amines or for the preparative purification of polyalkylene glycols or polyalkylene amines.

Description

EXAMPLES

[0091] A chromatography system (Agilent Technologies 1200 Series, Polymer Standards Service GmbH (PSS, Mainz)) was used to perform chromatographic measurements. The system consisted of a column furnace and a light scattering detector (ELSD), which was operated with nitrogen as a carrier gas. Measurements were performed on commercially available polyethylene glycol diols (PEG-diols; PSS, Mainz, Germany and PL, Shropshire, UK) and on synthesized polyethylene glycol monomethyl ethers (mPEG). A high-resolution chromolite column was used (monolithic silica gel with RP-18 end group capped). The chromolite column was obtainedfrom Merck KGaA (Darmstadt, Germany). The length of the column was 100 mm and its inner diameter was 4.6 mm. The column material had a high porosity of more than 80% and had macropores of about 1.1 m in size and mesopores of about 15 nm. The inner surface of the column material was 250 m.sup.2 g.sup.1. The inner surface was determined by mercury porosimetry or by nitrogen adsorption/-desorption isotherms.

[0092] Molar masses (numerical means M.sub.n and weight means M.sub.w) of the synthesized mPEG samples were determined by size exclusion chromatography and showed narrow dispersities (1,1). Details are found in Table 1 below.

TABLE-US-00001 TABLE 1 Overview of the properties of polyethylene glycol diols (PEG) and polyethylene glycol monomethyl ethers (mPEG) product M.sub.n (g mol.sup.1) M.sub.w (g mol.sup.1) PEG 1 375 400 1.07 PEG 2 1840 2010 1.09 PEG 3 2800 3060 1.09 PEG 4 7500 11200 1.51 PEG 5 22100 25800 1.17 PEG 6 34000 42700 1.26 mPEG 1 2300 2400 1.04 mPEG 2 5600 5800 1.04 mPEG 3 7500 7700 1.03 mPEG 4 120000 13200 1.10 mPEG 5 21800 22700 1.04 mPEG 6 34900 37200 1.07 mPEG 7 46200 50100 1.08

[0093] Chemicals and Materials

[0094] The reagents and solvents used were commercially available products purchased from Aldrich or Linde. Ethylene oxide (EO) was stirred in burettes over sodium before distillation. Before their use, 2-methoxyethanol and diphenylmethane were stirred over calcium hydride and then distilled under vacuum. The cleaned reactants were rinsed with argon, stored in Schlenk tubes in a glove box and used within three days. Tetrahydrofuran was dried by heating under reflux over freshly prepared sodium benzophenone until a deep blue color appeared. The dried tetrahydrofuran was subsequently stored in a Schlenk tube under inert gas and used within a short time.

[0095] Acetonitrile of HPLC grade was acquired from Sigma (Taufkirchen, Germany) and ultra-pure water was freshly prepared in a Thermo Scientific Barnstaedt GenPure-xCAD water purification system (Thermo Electron LED GmbH, Langenselbold, Germany). The methoxy polyethylene glycol samples used (mPEG samples) were synthesized. The PEG samples were purchased as SEC standards from PSS (Polymer Standards Service GmbH, Mainz, Germany) and PL (Polymer Laboratories, Shropshire, United Kingdom).

[0096] Manufacture of mPEG Samples

[0097] The production of the initiator potassium-2-methoxy ethanolate took place under inert conditions. 2-Methoxyethanol was dissolved in tetrahydrofuran and diphenylmethyl potassium was added drop by drop until a precipitation of the product could be observed and a slight orange mixture was formed. The product was washed four times with tetrahydrofuran until the orange color had completely disappeared. The product was then dried under vacuum and accumulated as a grey powder.

[0098] The preparation of the initiator solutions for the polymerization of EO to mPEG by living anionic ring opening polymerization (AROP) was carried out with the exclusion of water and air (inert). First, tetrahydrofuran and potassium-2-methoxy ethanolate were added to a GL45 bottle under inert conditions. In order to be able to carry out the entire process under inert conditions, the fine suspension was then transferred via PTFE hoses to a PicoClave glass autoclave reactor (BchiGlasUster, Uster, Switzerland) and cooled to 20 C. by stirring. Subsequently, the corresponding amount of EO was added to the reaction mixture using a mini-CORI-FLOW mass flow control apparatus (Bronkhorst High-Tech B.B., Ruurlo, Netherlands), whereby subsequently was heated to 45 C. within 120 min and was stirred for a further 48 hours. By adding a mixture of ethanol/acetic acid (95/5%, v/v) the polymerization was aborted. For insulation and cleaning, the polymer was filtered in cold diethyl ether and dried under vacuum. The product accumulated as a white powder.

[0099] Determination of the Molecular Weight of the Synthesized mPEG Samples

[0100] Size exclusion chromatographic measurements (SEC) were performed on a Shimadzu SEC system (control unit: CBM-20A VP, degasser: DGU-20A5, pump: LC-10ADVP, automatic sampler: SIL-10AD VP, furnace: TechLab, RI detector: RID-10A), which was operated with a ternary mobile phase consisting of chloroform/-isopropanol/triethylamine (94/2/4%, v/v/v) as an eluent. The PS5 SDV linear S column (5 micron particle size) was operated with a volume flow of 1 mL min.sup.1 at a temperature of 40 C. The system was calibrated in the range of 194 g mol.sup.1 to 106000 g mol.sup.1 using a PEG/PEO standard established by (i) Polymer Standards Service GmbH (PSS, Mainz, Germany) (PEO 106000, 55800, 42700, 26100 g mol.sup.1) and (ii) Polymer Laboratories (PL, Shropshire, UK) (PEG 12600, 7100, 4100, 1470, 960, 600, 440, 194 g mol.sup.1.

[0101] Liquid Chromatography

[0102] Chromatographic measurements were carried out with a modified system of the Agilent Technologies 1200 series from PSS (Polymer Standards Service GmbH, Mainz). Dead volumes of the column were reduced by using 130 m ID hose, which ran from the injector to the column head and from the column outlet to the detector. The injection volume was set to 10 l in all experiments, i.e. 0.6% of the column volume. The column was housed in a TCC 6000 column furnace from PSS (Polymer Standards Service GmbH, Mainz) and was tempered at 30 C. A light scattering detector (ELSD) (Softa Model 400) from PSS (Polymer Standards Service GmbH, Mainz) was connected to the outlet line of the column. Nitrogen was used as carrier gas for the ELSD detector. The temperature of the chamber and the drift pipe was set to 45 and 70 C. respectively. The detector was operated at the maximum data rate of 10 Hz. Elutions were performed on a high-resolution chromolite column (monolithic silica gel with RP-18 end group capped). The chromolite column was obtained by Merck KGaA (Darmstadt, Germany). The column length was 100 mm with a diameter of 4.6 mm. The total dead volume of the system was determined by replacing the column with a hose connector with a dead volume of zero and injecting the smallest and largest PEG sample. It was calculated that the dead volume was about 2% of the total column volume.

[0103] All samples were provided in concentrations between 0.1 and 2.0 mg mL.sup.1 by dissolving them in the respective mobile phase used for the chromatography experiments. Prior to the analysis, samples were filtered using a PTFE filter with a pore size of 0.45 m.

[0104] Experiments of Matrix-Assisted Laser-Desorption-Ionization Flight Time Mass Spectroscopy (MALDI-TOF-MS)

[0105] MALDI-TOF-MS experiments were carried out on the collected elution fractions using an UltraFlex TOF/TOF mass spectrometer (Bruker Daltonics, Bremen, Germany), which was equipped with an Nd-YAG laser. All spectra were recorded in positive reflector mode. Before each measurement the instrument was calibrated with an external PMMA standard from PSS (Polymer Standards Services GmbH, Mainz) in the required mass measuring range. MS data was analyzed with Flex Analysis 3.4 software and isotope patterns were generated using a software (Isotope Pattern Calculator by Bruker Daltonics). For the sample preparation of the MALDI-MS samples, the collected elution fractions, a solution of trans-2-[3-(4-tert.-butylphenyl)-2-methyl-2-propenylidene]malonic acid dinitrile (DCTB, Sigma-Aldrich) in chloroform at a concentration of 30 mg mL.sup.1 and the doping salt sodium iodide dissolved in chloroform at a concentration of 60 mg mL.sup.1 were used. The elution fraction with the sample was first placed on the sample plate and the sample was then allowed to dry. After drying, the solutions of the matrix and the salt were applied to the appropriate place on the sample plate. For each sample, 0.5 l of the sample solution was used, followed by 0.5 l of the matrix-salt mixture solutions.

Example 1: Determination of Elution Times Depending on the Composition of the Eluent

[0106] FIGS. 1 a) and 1 b) show elution times for different mPEG samples (closed symbols) and PEG diol samples (open symbols) depending on the composition of the eluent. Mixtures of acetonitrile with water were used as eluent.

[0107] FIG. 1 a) shows the results for mPEG samples with different molecular weights (M.sub.n=2300 g/mol; M.sub.n=5600 g/mol; M.sub.n=7500 g/mol; M.sub.n=12000 g/mol; M.sub.n21800 g/mol; M.sub.n=34900 g/mol; M.sub.n=46200 g/mol).

[0108] FIG. 1b) shows the results for PEG-diol samples with different molecular weights (M.sub.n=1840 g/mol; M.sub.n=2800 g/mol; M.sub.n=7500 g/mol; M.sub.n=22100 g/mol; M.sub.n=34000 g/mol).

[0109] FIG. 2 shows results of the gradient elution of a PEG-diol sample, a mPEG sample and a mixture of both samples. The upper curve shows the elution of a PEG-diol sample with an average molecular weight M.sub.n=1840 g/mol. The middle curve shows the elution of an mPEG sample with an average molecular mass M.sub.n=2300 g/mol and the lower curve shows the elution of a 50/50 (%, v/v) mixture of these two samples.

[0110] Conditions: Binary composition of the mobile phase with 10% acetonitrile in water which was kept isocratic for 3 minutes and then increased linearly within 50 minutes to a content of 50% acetonitrile in water (%, v/v)

[0111] FIG. 3 shows the elugrams of a mPEG sample of the mean molar weight of M.sub.n12 000 g/mol (middle curve) and of 50/50 (%, v/v) mixtures of this mPEG sample with PEG-diol of the mean molar weight M.sub.n=2800 g/mol (upper curve) or with PEG-diol of the mean molecular weight M.sub.n=22100 g/mol (lower curve). The composition of the mobile phase of water/acetonitrile was 60/40 (%, v/v). The flow rate of the mobile phase was 1 mL/min.

[0112] FIG. 4 (a) shows the elugrams of mixtures of m-PEG with different levels of PEG-diol. The content of mPEG of the mean molecular weight of M.sub.n=2300 g/mol was kept constant at 1 mg/mL and increasing levels of PEG-diol of the mean molecular weight of M.sub.n=1840 g/mol were used.

[0113] FIG. 4 (b) shows a double-logarithmic plot of the band height of the elution signal in relation to the content of PEG-diol in the mixture (with error bars obtained from a three times repeated injection and chromatographic analysis). The measurements were repeated the next day with the same samples (open symbols).

[0114] FIGS. 5a and 5b show the results of elution of mPEG and PEG-diols. The mean molar masses M.sub.n of the mPEG and PEG-diols used range from 1500 to 50000 g/mol. The upper figure shows results obtained using 50/50 to 40/60 acetonitrile/-water (%, v/v) as an eluent. The lower figure shows a clearly enlarged section of FIG. 5a.

[0115] In the upper FIG. 5a, the molar masses (M.sub.n) of PEG-diols (open symbols) and mPEG (closed symbols) are plotted against the elution time at 50/50 acetonitrile/-water (%, v/v), indicating the transition between size exclusion mode (squares) and adsorption/partition mode (diamonds and pentagons). The latter is first achieved for the larger molar masses. The lower diagram 5b with a lower concentration of acetonitrile as elution agent shows that at the content of 40% acetonitrile in the elution agent, the mPEG with a lower molar mass elute significantly later than the PEG-diols. The composition of the mobile phase was 40/60 acetonitrile/water (%, v/v). The flow rate of the mobile phase for both FIGS. 5a and 5b was 1 mL min.sup.1.

[0116] FIGS. 5a and 5b show that a composition of the mobile phase of 50/50 acetonitrile/-water (%, v/v) leads to elution patterns that are influenced by size exclusion effects, i.e. the largest mPEG and PEG-diols elute first. Increase in the polarity of the elution agent of the mobile phase to 42% acetonitrile shows increased elution times for the PEG with larger molar masses, so that all PEG eluate at similar elution times, with virtually no selectivity (circles) over all molar masses. These results demonstrate the limited ability of liquid chromatography under critical conditions (LCCC elution mode) to differentiate between mPEG and PEG-diol, since the difference between alpha-hydroxyl compared to alpha-methoxy is simply too small to allow clear critical adsorption conditions based on the end group character (FIG. 1).

[0117] In the case of eluents containing only 40 to 41% acetonitrile, larger PEG eluate significantly later than the smaller, i.e. at molar masses of more than 10000 g mol.sup.1 (diamonds and pentagons in FIG. 5a) the partition/adsorption of the polymer scaffold is entered, i.e. at molecular weights above 10000 g mol.sup.1, the partition/adsorption mode is clearly occurring (diamonds and pentagons in the upper part of FIG. 5a). Above 10000 g mol.sup.1, the result is determined exclusively by the molar mass dependent adsorption/partition of the polymer scaffold with a gradually decreasing contribution, which derives from the identity of the alpha groups. This is shown in the lower part of FIG. 5a, which relates to the smaller molecular weights (10000 g mol.sup.1, pentagons).

[0118] At this point, partition and adsorption begin to dominate the elution. It should also be noted that small but noticeable differences in the elution time are beginning to develop in the elution times for the populations of the species of mPEG and the PEG-diols. FIGS. 5a and 5b thus describe a unique possibility of separating mPEG and PEG-diol species with very similar molecular weights, i.e. separation in partition/adsorption mode at an acetonitrile content of 40%. While PEG-diols between 1000 and 10000 g mol.sup.1 still elute quite similarly, mPEG show a delayed elution (pentagons in FIGS. 5a and 5b). This is demonstrated, for example, below on separations of mixtures of a mPEG/PEG-diol pair with similar molar masses (FIG. 6a).

[0119] FIG. 6a shows elugrams of PEG-diol (M.sub.n=1840 g mol.sup.1, upper track), mPEG (M.sub.n=2300 g mol.sup.1, middle track) and their mixture of 50/50 (%, v/v) (lower track).

[0120] FIG. 6b shows elugrams of mPEG (M.sub.n=2300 g mol.sup.1, medium track), mixtures of this m-PEG with PEG-diol (M.sub.n=375 g mol.sup.1, upper track) and mixtures of this m-PEG with PEG-diol (M.sub.n=2800 g mol.sup.1, lower track).

[0121] In the experiments shown in FIGS. 6a and 6b, the flow velocity of the mobile phase was 1 mL min.sup.1 and the mobile phase had a composition of 60/40 water/acetonitrile (%, v/v).

[0122] These results show that the mPEG of the example of FIG. 6a elutes as a narrow band that is different from that of a PEG-diol with a similar molar mass. Larger mPEG with molar masses of 20000 g mol.sup.1 elute similarly to model diols with comparable molar mass (FIG. 1). This situation is not surprising and originates from a lost selectivity with respect to the end group character at molar masses greater than 20000 g mol.sup.1, wherein the adsorption of the polymer chain strongly dominates the elution.

[0123] It is noted that when using a diol with significantly different molar mass and molar mass distribution, separation of mPEG and PEG-diol is possible even with larger molar masses.

[0124] To support the pronounced selectivity between mPEG and PEG-diol, the mPEG used in FIGS. 6a and 6b was analyzed using gradient-liquid chromatography (FIG. 2). The experiments showed the expected dispersity of the PEG standards and the mPEG samples due to the appearance of a variety of bands in both the PEG-diol and the mPEG sample. Gradient elution does not improve the discriminability between mPEG and PEG-diol (FIG. 2).

[0125] As a result of different initiation scenarios and reaction kinetics as well as the execution of polymerization, it is unlikely that PEG-diol impurities have the same molar masses values as the mPEG samples. The different initiation probabilities and kinetics of the growing polymer chains with two possible connecting points for monomers need attention. It is therefore a detection desirable whether smaller and in particular larger amounts of PEG-diol impurities are present in mPEG products. In order to imitate such conditions, mPEG samples were examined containing a controlled content of PEG-diol with lower or larger molar mass at the lower and upper end of the molar mass range. Selected elugrams are shown in FIGS. 6b and 3.

[0126] It is clear that diols with lower molar mass differ significantly from the respective mPEG (FIGS. 6b and 3, top traces), but are increasingly difficult to identify with larger molar mass of the diol (FIGS. 6b and 3, lowest trace), although two species can be identified in the elugrams. The example of the PEG-diol with the larger molar mass (FIG. 3, lowest track) elutes in two distinct fractions that come from a bimodal distribution of the molar masses. However, these elute later than the mPEG fraction. This is an inherent result of the elution of the PEG-diols, which is based on the adsorption/partition of the polymer chains of the macromolecules with greater molar mass (FIG. 5). The smaller elution fraction also shows some overlap with the fraction of the mPEG. Notwithstanding and in contrast to the pure mPEG (FIGS. 6a and 6b, middle track), the PEG-diol is displayed by a clear elution pattern as shown in the chromatograms (Fig. S3, lowest track compared to the middle track).

[0127] To illustrate the fundamental suitability of the method for estimating a quantity of existing PEG-diols, experiments were carried out with mixtures of different concentrations of PEG-diol (M.sub.n=1840 g mol.sup.1) with a fixed concentration of mPEG (M.sub.n=2300 g mol.sup.1) (FIG. 4a).

[0128] Although the typical nonlinear dependence of the ELSD is obvious even in the double logarithmic plot of the band height against the concentration of the solution (FIG. 4), the repeatability of the measurements on different days allows an estimation of the concentrations and the content of PEG-diols of less than 1% at a fixed concentration of mPEG 1 mg mL.sup.1 (FIG. 4).

[0129] To demonstrate the performance of the method of the approach described here, a sample of mPEG was analyzed, which had been produced in the presence of a protic contamination, e.g. of water. The water had been added at the beginning of the living anionic ring opening polymerization. In this case, diol contamination is indicated by a clear shoulder in the band in the chromatogram (FIG. 7a).

[0130] FIG. 7a shows an application example for the described method. The same chromatographic conditions as described in FIG. 6 were used.

[0131] FIG. 7a shows the elution trace of 1 mg mL.sup.1 of a product of anionic polymerization containing diol contamination. The representation on the right side of FIG. 7a is an enlargement of the band on the left side of this figure.

[0132] FIG. 7b shows in the left half MALDI-TOF-MS spectra of the collected small elution fractions identified as PEG-diol (shown in black) and the larger elution fraction identified as mPEG (shown in gray). The right half of FIG. 7b the isotope fragmentation pattern of the PEG-diol and the mPEG is shown. An excerpt from the mass spectrum is displayed, where the dashed line is the calculated isotope fragmentation pattern for PEG-diols.

[0133] The elution fraction collected from the chromatogram (FIG. 7a) was identified by using MALDI-TOF-MS as PEG-diol contamination with a wide molar mass distribution of less than 1000 m/z up to 3000 m/z (FIG. 7b, left part, black mass spectrum). The amount of PEG-diol was estimated at about 8%.

[0134] From the data of these examples it can be inferred that under experimental conditions, which do not lead to a selective elution of species based on a distribution of the molar mass (e.g. FIG. 3), but the populations of mPEG and PEG-diols occur in fairly narrow elution bands (e.g. FIG. 3), a rapid identification of PEG-diol impurities in mPEG in the pharmaceutical-relevant molar range is possible.

[0135] Using gradient liquid chromatography coupled with electrospray ionization (ESI), polyoxazolines were investigated (FIG. 8). FIG. 8 shows the individual oligomers in the chromatogram of polyoxazoline, which has a OH and a CH.sub.3 end group. Under the chromatogram, the ESI spectrum of the substance to be examined is shown. The individual oligomers examined in mass spectroscopy are framed. These elute at different times, which is represented in the chromatogram by the respective frames. Mass spectroscopically, CH.sub.3- and H-initiated polyoxazolines could be detected and chromatographically separated using a gradient method. The measurement was carried out with a flow of 0.5 mL/min with aqueous acetonitrile as an eluent. The acetonitrile content was increased from 20% to 40% within 70 min. After a further 10 min, the acetonitrile content was increased to 80%. A chromatography column was used, which contained monolithic silica gel as a stationary phase.