WO1999037702A1

WO1999037702A1 - Process for purifying polyketones

Info

Publication number: WO1999037702A1
Application number: PCT/GB1998/003892
Authority: WO
Inventors: Derek Alan Colman
Original assignee: Bp Chemicals Limited
Priority date: 1998-01-23
Filing date: 1998-12-23
Publication date: 1999-07-29
Also published as: GB9801433D0

Abstract

A process for purifying a contaminated polyketone comprising (a) mixing the polyketone with a solvent in which the impurities are soluble and the polyketone is insoluble; (b) passing the resulting mixture of polyketone and solvent along a membrane, and (c) applying a pressure differential across the membrane so that a portion of the solvent containing soluble impurities permeates through the membrane.

Description

PROCESS FOR PURIFYING POLYKETONES The present invention relates to a process for purifying a contaminated polyketone and, in particular, to a process for removing a solvent containing soluble impurities from a polyketone using a membrane.

For the purposes of this patent, polyketones are defined as linear polymers having an alternating structure of (a) units derived from carbon monoxide and (b) units derived from one or more olefinically saturated compounds. Such polyketones have the formula:

O [(CR₂-CR₂)C]_m

where the R groups are independently hydrogen or hydrocarbyl groups and m is a large integer; they are disclosed in several patents e.g. US 3694412. Processes for preparing the polyketones are disclosed in US 3694412 and EP 181014 and EP 121965. Although for the purposes of this patent polyketones correspond to this idealised structure, it is envisaged that materials corresponding to this structure in the main, but containing small regimes (i.e. up to 10 wt%) of the corresponding homopolymer or copolymer derived from the olefinically unsaturated compound, also fall within the definition.

To obtain a polyketone having acceptable quality and chemical stability it is often necessary to remove soluble impurities (e.g. polymerisation reaction byproducts, catalyst residues and co-catalyst residues) from the polyketone. This may be achieved by contacting the polyketone with a solvent in which the impurities are soluble and the polyketone is insoluble and removing the solvent and dissolved impurities from the polyketone using hydrocyclone, centrifugal separation or filtration techniques. Problems arise with the use of such conventional separation techniques in that the separation efficiency may be significantly influenced by the morphology of the polyketone particles an-d/or the density difference between the polyketone and the solvent. Thus during conventional filtration, conditions may be such that the morphology of the polyketone leads to the formation of a dense, almost impermeable filter cake which results in unacceptably low filtration rates. For separation techniques that rely on a difference in density between the polyketone and the solvent, such as hydrocyclones and centrifugal separators, effective separation is difficult to achieve when this density difference is small, for example less than 0.1 kg/litre. The morphology of the polyketone particles has also been observed to influence the separation efficiency of hydrocyclone and centrifugal separators.

Surprisingly, it has now been found that when a solvent together with soluble impurities is removed from a polyketone using a cross flow membrane filtration process that the separation efficiency is less dependent on the morphology of the polyketone and/or density difference between the polyketone and the solvent.

Thus, according to the present invention there is provided a process for purifying a contaminated polyketone which process comprises the steps of (a) mixing the polyketone with a solvent in which the impurities are soluble and the polyketone is insoluble; (b) passing the resulting mixture of polyketone and solvent along a membrane, and (c) applying a pressure differential across the membrane so that a portion of the solvent containing soluble impurities permeates through the membrane. As noted above for the purposes of this patent, polyketones are defined as linear polymers having an alternating structure of (a) units derived from carbon monoxide and (b) units derived from one or more olefinically unsaturated compounds. Suitable olefinic units are those derived from C₂ to Cι₂ alpha-olefins or substituted derivatives thereof or styrene or alkyl substituted derivatives of styrene. It is preferred that such olefin or olefins are selected from C₂ to Cβ normal (straight chain) alpha-olefins and it is particularly preferred that the olefin units are either derived from ethylene or most preferred of all from a mixture of ethylene and one or more C₃ to Cβ normal alpha-olefin(s) especially propylene or butylene. In these most preferable materials it is f rther preferred that the molar ratio of ethylene units to C₃ to Cβ normal alpha-olefin units is greater than or equal to 1 most preferably between 2 and 30. Typically, the polyketone will be a copolymer of ethylene/propylene/CO or ethylene/butylene/CO where the units derived from propylene or butylene are present in an amount in the range 5-8 % e.g. 6 % by weight of the polymer. The polyketone will suitably have a number average molecular weight of between 20,000 and 1,000,000 relative to a polymethyl methacrylate standard, preferably between 40,000 and 500,000, more preferably between 50,000 and 250,000, for example, 60,000 to 150,000. A preferred polyketone is an ethylene/propylene/CO terpolymer or an ethylene/butylene/CO terpolymer having a number average molecular weight in the range 60,000 to 150,000.

The polyketone will suitably have a particle size in the range 1 μm to 2000 μm, preferably 10 μm to lOOOμm and most preferably 50 μm to 750 μm.

The polyketone will suitably have a density in the range 1.10 kg/litre to 1.30 kg/litre, preferably 1.23 kg/litre to 1.25 kg/litre. Typical soluble impurities include catalyst residues, co-catalyst residues and polymerisation reaction by-products. Catalyst residues and co-catalyst residues are for the purposes of this patent defined as active catalyst and co-catalyst used in the preparation of the contaminated polyketone as well as inactive species derived from the catalyst and co-catalyst. Polymerisation reaction by-products include low molecular weight polyketones which are soluble in the solvent used in the process of the present invention.

Catalyst compositions for preparing polyketones typically comprise a Group Vm metal compound containing at least one ligand capable of coordinating to the Group VIII metal and a co-catalyst (or promoter). A typical catalyst composition would be that described in EP 121965 and

EP 314309, which is derived from (a) a palladium compound (b) a source of an anion which is either non-coordinating or only weakly coordinating to palladium (co-catalyst) and (c) a bisphosphine of formula (I) R'R^-R-PR'R⁴ where R¹ to R⁴ are independently aryl groups which can optionally be polar substituted and R is a divalent organic bridging group such as -(CH₂)_n- (n = 2-6). A source of the anion is typically its conjugate acid.

Also suitable are catalyst compositions as detailed in EP 619335 which comprise

(a) a Group VIII metal compound, containing at least one ligand capable of coordinating to the Group VIII metal and (b) a boron hydrocarbyl compound, preferably a Lewis acid of the formula BXYZ where at least one of X Y and Z is a monovalent hydrocarbyl group (co-catalyst). Typically the boron hydrocarbyl compound is a compound of the formula BR₃ where R is a Ci-Cβ alkyl, or an aryl group, for example, a substituted or unsubstituted phenyl group, for example, C5Η5, ClCβ-H , or C5F5.

Typically the ligand capable of coordinating to the Group VHt metal is a bidentate phosphine ligand, for example, a ligand of the formula (I) herein above. Alternatively the ligand may be a bidentate phosphine ligand having at least two phosphorus atoms joined by a bridging group of the formula -(N)_x-(P)_y-N- where x is 0 or 1 and y is 0 or 1, in particular, a bridging group of the formula -(NR²)_X- (PR³)_y-NR²- where each R² is the same or different and R² and R³ represent a monovalent organic group. A preferred ligand has the formula (II) R¹ ₂P-(NR²)_X- (PR³)_y-NR²-PR¹ ₂ where each R¹ is independently an aryl, alkyl, alkoxy, amido or substituted derivative thereof, R² is a hydrogen, a hydrocarbyl or hetero group, R³ is a hydrocarbyl or hetero group. For any of the catalyst systems described herein above preferred bidentate ligands are (o-anisyl)₂-X-P(o-anisyl)₂ where X = - (CH₂)„- n = 2-4, or X = N(R) R = Cι-C₆ alkyl or aryl.

The solvent which is used in the process of the present invention is one in which the polyketone is insoluble or virtually insoluble. Since polyketones tend to be insoluble in most solvents a wide choice of materials is available. Typical examples of preferred solvents are aliphatic alcohols (e.g. isopropanol, methanol or butanol), ketones (e.g. methyl isobutyl ketone, methyl ether ketone or acetone), ethers (e.g. diethyl ether or dimethyl ether), saturated hydrocarbons (e.g. pentane, hexane, heptane), chlorinated solvents (e.g. dichloromethane) and water. It is preferred to use the solvent in which the polyketone was prepared since on a commercial scale this greatly simplifies the manufacturing process.

Preferably, a portion of the solvent in which the polyketone was prepared (containing catalyst residues, co-catalyst residues and soluble polymerisation reaction by-products) may be separated from the contaminated polyketone using steps (b) and (c) of the present invention prior to mixing the polyketone with solvent in step (a).

Preferably, the mixing and separation steps are repeated until an acceptable level of soluble impurities is attained in the mixture of polyketone and solvent which is retained by the membrane (the retentate). Typically, the mixing and separation steps are repeated until the level of soluble impurities falls below 25%, preferably below 15%, more preferably below 10%, still more preferably below 5%, most preferably below 2.5%, for example, below 1% of their original level. Preferably, the mixing and separation steps are repeated until the level of soluble impurities falls below the limits of detection for the impurities. Preferably, during the solvent removal step (c) of the present invention a concentrated mixture of polyketone and solvent is retained by the membrane (retentate). Preferably, the polyketone is maintained in suspension in the retentate and is thereby prevented from forming a filter 'cake' on the surface of the membrane. Typically, the amount of polyketone suspended in the retentate is at least 1% by weight, preferably at least 5% by weight, more preferably at least 10% by weight, most preferably, at least 15% by weight, based on the total weight of the retentate.

The membranes used in the present invention are chosen so that their pore size is small enough to prevent permeation of the polyketone through the membrane. Typically the pore size is less than the size of the smallest polyketone particles, preferably the pore size is substantially less than the size of the smallest polyketone particles (to avoid plugging of the membrane pores with polyketone particles). Preferably the pore size of the membrane is less than or equal to 1.0 μm, more preferably less than or equal to 0.5 μm, most preferably less than or equal to 0.05 μm. Where the membrane is defined by its molecular weight cut off, the membrane preferably has a molecular weight cut off of 15 000.

The membranes of the present invention may be made of any suitable material which is resistant to the solvent, for example, the membrane material may be selected from the group consisting of glass, a ceramic material, carbon, silicon carbide, and a polymeric material such as PTFE. A preferred membrane material is ceramic. Typically, the ceramic membrane material is α-Al₂O₃, ZrO₂ or a mixed oxide of SiO₂ and Al₂O₃ or ZrO₂. Suitable ceramic membranes are those produced by Kerasep (having a molecular weight cut off of 15 000) and Ceramem (having a pore size of 0.05 μm) and SCT (having a pore size of 0.02 μm). The membranes may be flat, tubular (including circular, square, rectangular, triangular cross section), or spiral wound. Preferably, membrane modules comprising at least one ceramic membrane monolith may be employed.

Typically the mixture of polyketone and solvent is passed along the membrane at a flow rate of at least 1 ms^"1, more preferably at least 3 ms^'1. Typically the pressure differential across the membrane is in the range 0.5 to 30 bar, preferably 0.5 to 10 bar, more preferably 0.5 to 3 bar, most preferably 1 to 1.5 bar.

Preferably the flux (which for the purposes of this patent is defined as the volume of solvent permeating the membrane per hour through 1 m² of filtration area) is at least 80 1/m /h, more preferably in the range 80 to 250 1/m /h, most preferably in the range 100 to 150 l/m²/h.

In order to prevent any build up of polyketone polymer on the surface of the membrane the differential pressure across the membrane may be periodically removed or the membrane may be periodically back-flushed with solvent. The process of the present invention may be batch or continuous, preferably continuous, in particular, where the contaminated polyketone is prepared using a continuous polymerisation process.

The process of the present invention may be used in continuous diafiltration, successive staged flow or continuous counter-current flow configurations.

A continuous diafiltration process is illustrated in Figure 1. A feed stream comprising a mixture of polyketone, solvent and soluble impurities (1) is fed to a membrane separation unit (2) containing a membrane (3). A filtrate (4) comprising solvent and soluble impurities permeates through the membrane and is removed from the membrane separation unit (1). A retentate (5) comprising polyketone and a portion of the solvent and soluble impurities is retained by the membrane. The retentate (5) is recycled to the membrane separation unit (1) after being mixed with fresh solvent (6). When an acceptable level of soluble impurities has been attained in the retentate (5), the process is stopped and purified polyketone and solvent is removed from the membrane separation unit (2).

A typical successive staged flow process is illustrated in Figure 2. A feed stream comprising a mixture of polyketone, solvent and soluble impurities (1) is fed to a first membrane separation unit (2) containing a membrane (3). A filtrate (4) comprising solvent and soluble impurities permeates through the membrane and is removed from the membrane separation unit (2). A retentate (5) comprising polyketone and a portion of the solvent and soluble impurities is retained by the membrane and is fed to a second membrane separation unit (7) after being mixed with fresh solvent (6). A filtrate (8) comprising solvent and soluble impurities is removed from the second membrane separation unit (7). A retentate (9) comprising polyketone, solvent and a reduced amount of soluble impurities is fed to a third membrane separation unit (10) after being mixed with fresh solvent (6). A filtrate (11) comprising solvent and trace amounts of soluble impurities and a product stream (12) comprising purified polyketone and solvent are removed from the third membrane separation unit (10). The number of membrane separation units employed in a successive staged flow process is dependent on the level of impurities in the polyketone feed stream (1), the amount of solvent (6) which is mixed with the retentate streams removed from the membrane separation units and the surface area of the membrane (3) in each membrane separation unit. At least two membrane separation units should be employed, preferably 3 to 6 membrane separation units.

Atypical continuous counter-current flow process is illustrated in Figure 3. A feed stream comprising a mixture of polyketone, solvent and soluble impurities (1) is fed to a first membrane separation unit (2) containing a membrane (3). A filtrate (4) comprising solvent and soluble impurities permeates through the membrane and is removed from the membrane separation unit (2). A retentate (5) comprising polyketone and a portion of the solvent and soluble impurities is retained by the membrane and is fed to a second membrane separation unit (7). A filtrate (8) comprising solvent and a lower amount of soluble impurities is removed from the second membrane separation unit (7) and is mixed with the polyketone feed stream (1). A retentate (9) comprising polyketone, solvent and a reduced amount of soluble impurities is fed to a third membrane separation unit (10) after being mixed with fresh solvent (6). A filtrate (11) comprising solvent and trace amounts of soluble impurities is mixed with the retentate (5) removed from the first membrane separation unit (2). A product stream (12) comprising purified polyketone and solvent is removed from the third membrane separation unit (10). The number of membrane separation units employed in a continuous counter- current flow configuration process is dependent on the level of impurities in the polyketone feed stream (1) and the amount of fresh solvent (6) which is employed in the process and the surface area of the membrane (3) in each membrane separation unit. At least two membrane separation units should be employed, preferably 3 to 6 membrane separation units. Soluble impurities are removed with the filtrate from the first membrane separation unit while fresh solvent is mixed with the retentate fed to the last membrane separation unit. The filtrate removed from the second and subsequent membrane separation units is mixed with the retentate fed to the preceding membrane separation unit. The present invention is further illustrated by way of the following Examples: Materials

A mixture of ethylene/propylene/CO terpolymer (polyketone) and dichloromethane (4% by weight of solids) containing soluble impurities (polyketone slurry) was used in the following experiments. Under optical microscopic analysis the polyketone particles had a primary particle size range of 0.1 to 1 μm. The density of the polyketone was 1.24 kg/1. Comparative Example A The mixture of polyketone and dichloromethane (polyketone slurry) was filtered to a wet 'cake' type consistency using a Nutsche type filter/dryer fitted with a polypropylene cloth filter medium having a pore size of 0.5 μm. The differential pressure across the filtration medium was 2 bar. The solvent filtration rate was found to be 501/hr/m² of filtration area. However, this was the average filtration rate over the filtration period. The filtration rate was found to decrease rapidly after filtration was commenced owing to the formation of a dense filter 'cake' of polyketone over the whole area of the filtration medium. The filtration rate at the termination of the experiment was estimated to be 2 1/hr/m². This was an unacceptably slow rate. Comparative Example B

The polyketone slurry was mixed with twice its volume of isopropanol and filtered using the procedure of Example A. An improvement in the initial filtration rate was observed. It is believed that this is due to a dilution effect associated with the addition of the isopropanol rather than a true improvement in filtration rate. The filtration rate was observed to rapidly decline in a similar manner to Example A. Example 1

A sample of polyketone slurry in dichloromethane (20 litres) was removed from a polymerisation reactor and was contacted with a ceramic microfiltration membrane having a pore size of 0.05 μm and a filtration area of 0.14 m². The slurry was contacted with the ceramic membrane at a flow rate along the membrane of 20 1/min (a velocity across the membrane of about 1.5 m/s), using a small lobe pump, which also supplied the pressure differential across the membrane required to achieve filtration. The differential pressure across the membrane was kept constant at 1.5 bar and a solvent flux of 100 l m²/hr was achieved. Example 2

The procedure of Experiment 1 was repeated except a sample of polyketone slurry in dichloromethane (10 litres) was mixed with 20 litres of isopropanol and the pressure differential across the membrane was varied between 0.375 and 2.8 bar during the experiment. Fluxes of between 100 and 150 l m²/hr were achieved. Example 3

The procedure of Experiment 2 was repeated except that the pressure differential across the membrane was maintained constant at 1.5 bar. From the polyketone slurry was removed 26 litres of solvent at a flux of 90 l/m² hr. The remaining 4 litres of polyketone slurry was mixed with 8 litres of isopropanol and contacted with the ceramic microfiltration membrane as described above with 10 litres of solvent being removed. The remaining 2 litres of polyketone slurry was mixed with 10 litres of isopropanol and contacted with the ceramic membrane as described above, with 10 litres of solvent being removed. The remaining 2 litres of polyketone slurry in predominantly isopropanol was mixed with three consecutive 10 litre volumes of water. 10 litres of water were removed from each consecutive mixture by contacting the slurry with the ceramic membrane as described above. Water flux rates of 90 l/m²/hr were achieved. Example 4

Table 1 demonstrates the amount of solvent required to reduce the relative level of impurities from 1 to 0.0444 for a continuous diafiltration process ("continuous mem"), a successive staged flow process having six membrane separation units ("6 stage cross flow wash") and continuous counter-current flow processes having 6 and 3 membrane separation units ("6 stage counter mem" and "3 stage counter mem" respectively). It can be seen that the successive staged flow process employs the highest amount of solvent while the continuous counter- current flow process having 6 membrane separation units employs the least amount of solvent. Table 1

Base case:

Initial contaminant cone CO 1

Initial volume of current wash Vi 300

Time based on cont. membrane Hr 9.34

Initial concentration factor of polymer 1

Desired final contaminent cone. C 0.0444

6 stage 6 stage 3 stage continuous cross flow counter counter mem wash mem mem

Final concentration C 0.0444 0.0444 0.0443 0.0435

Control 0.00000 0.00000 0.00006 0.00089

Flux (/hr) 100 131 265 231

Total wash flow vol 934 1225 413 720

10

Claims

Claims:

1. A process for purifying a contaminated polyketone having an alternating structure of (a) units derived from carbon monoxide and (b) units derived from one or more olefinically unsaturated compounds which process comprises the steps of (a) mixing the polyketone with a solvent in which the impurities are soluble and the polyketone is insoluble; (b) passing the resulting mixture of polyketone and solvent along a membrane, and (c) applying a pressure differential across the membrane so that a portion of the solvent containing soluble impurities permeates through the membrane.

2. A process as claimed in claim 1 wherein the polyketone is a copolymer of ethylene/propylene/CO or ethylene/butylene/CO.

3. A process as claimed in claim 1 or claim 2 wherein the polyketone has a number average molecular weight in the range 20,000 to 1,000,000.

4. A process as claimed in any one of the preceding claims wherein the polyketone has a particle size in the range 1 ╬╝m to 2000 ╬╝m.

5. A process as claimed in any one of the preceding claims wherein the polyketone has a density in the range 1.10 kg/litre to 1.30 kg/litre.

6. A process as claimed in any one of the preceding claims wherein the solvent is selected from the group consisting of aliphatic alcohols, ketones, ethers, saturated hydrocarbons, chlorinated solvents and water.

7. A process as claimed in any one of the preceding claims wherein a portion of the solvent in which the polyketone was prepared is separated from the contaminated polyketone via steps (b) and (c) prior to mixing the polyketone with solvent in step (a).

8. A process as claimed in any one of the preceding claims wherein the mixing and separation steps are repeated until the level of soluble impurities in the mixture

11 of polyketone and solvent which is retained by the membrane falls below 25% of their original level.

9. A process as claimed in claim 8 wherein the mixing and separation steps are repeated until the level of soluble impurities in the mixture of polyketone and solvent which is retained by the membrane falls below 10% of their original level.

10. A process as claimed in claim 9 wherein the mixing and separation steps are repeated until the level of soluble impurities in the mixture of polyketone and solvent which is retained by the membrane falls below 2.5% of their original level.

11. A process as claimed in any one of the preceding claims wherein the pore size of the membrane is less than or equal to 1.0 ╬╝m.

12. A process as claimed in any one of the preceding claims wherein the membrane has a molecular weight cut off of 15 000.

13. A process as claimed in any one of the preceding claims wherein the membrane is made of a material selected from the group consisting of glass, a ceramic material, carbon, silicon carbide and a polymeric material.

14. A process as claimed in any one of the preceding claims wherein the membrane is flat, tubular or spiral wound.

15. A process as claimed in any one of the preceding claims wherein the mixture of polyketone and solvent is passed along the membrane at a flow rate of at least 1 ms^'1.

16. A process as claimed in any one of the preceding claims wherein the pressure differential across the membrane is in the range 0.5 to 30 bar.

17. A process as claimed in any one of the preceding claims wherein the flux is at least 80 l m²/h.

18. A process as claimed in any one of the preceding claims wherein the differential pressure across the membrane is periodically removed or the membrane is periodically back-flushed with solvent.

19. A process a claimed in any one of the preceding claims wherein the process is carried out in a continuous diafiltration, successive staged flow or continuous counter-current flow configuration.

12