PROCESS FOR SEPARATING ONE OR MORE COMPONENTS FROM A MIXTURE

Abstract

The invention relates to a process for separating one or more components from a mixture by a membrane separation in whichdepending on the component to be separatedan acid or a base is added to the mixture before the membrane separation.

Claims

1. A process for separating a homogeneous catalyst system that contains at least one metal from groups 8 to 10 of the periodic table of elements or a compound thereof and a phosphorus-containing ligand, from a reaction solution by a membrane separation that results in the homogeneous catalyst system being depleted in the resulting permeate stream and enriched in the resulting retentate stream, wherein at least one component of the homogeneous catalyst system contains at least one acidic or basic functional group, wherein the reaction solution prior to the membrane separation contains an acid in an amount of 0.1% by weight, based on the total weight of the mixture, when the at least one component to be separated has a basic functional group, or contains a base in an amount of 0.1% by weight, based on the total weight of the mixture, when the at least one component to be separated has an acidic functional group.

2. The process according to claim 1, wherein the acid or base is added to the mixture in an amount of 0.1% by weight and <5% by weight.

3. The process according to claim 1, wherein the acid is a Brnsted acid having a pKa 5, or a Lewis acid having an LAU value of more than 25.

4. The process according to claim 1, wherein the acid is a Brnsted or Lewis acid, the Brnsted acid being selected from the group consisting of perchloric acid, sulfuric acid, phosphoric acid, methylphosphonic acid or a sulfonic acid and the Lewis acid being selected from the group consisting of aluminium triflate, aluminium chloride, aluminium hydride, trimethylaluminium, tris(pentaflurophenyl)borane, boron trifluoride, boron trichloride, titanium(IV) isopropoxide, Bu.sub.2SnO, BuSn(O)OH, or mixtures thereof.

5. The process according to claim 1 wherein the base is selected from the group consisting of alkali metals, alkali metal oxides, alkali metal or alkaline earth metal acetates, alkali metal or alkaline earth metal oxides, alkali metal or alkaline earth metal alkoxides, such as NaEtOH or MgEtOH, and alkali metal carbonates such as K.sub.2CO.sub.3 or Cs.sub.2CO.sub.3.

6. The process according to claim 1, wherein the phosphorus-containing ligand of the homogeneous catalyst system has a mono- or bidentate structure.

7. The process according to claim 1, wherein the phosphorus-containing ligand is a benzene-based diphosphine compound.

8. The process according to claim 1, wherein a ceramic membrane material is used in the membrane separation.

9. The process according to claim 8, wherein the ceramic membrane material comprises or consists of aluminium oxide, SiO.sub.2, TiO.sub.2, ZrO.sub.2 or mixtures of these materials.

10. The process according to claim 1, wherein a polymer-based membrane material is used in the membrane separation and wherein the polymer-based membrane material includes at least one separation-active layer.

11. The process according to claim 10, wherein the polymer-based material as the separation-active layer includes a material selected from the group consisting of polyimide (PI), polydimethylsiloxane (PDMS), polyetherimide (PEI), poly(2,6-dimethyl-1,4-phenylene oxide) (PPO), polyvinyl alcohol (PVA), polyacrylic acid (PAA), polyurethanes (PU), poly(l-trimethylsilyl-1-propyne) (PTMSP), polybenzimidazoles (PBI), polydicyclopentadienes (pDCPD), polyaryl ether ketones (PAEK) and mixtures of these materials.

12. The process according to claim 1, wherein the separation-active layer is composed of a PAEK polymer.

13. The process according to claim 12, wherein the separation-active layer is composed of PEEK, preferably having a degree of sulfonation of less than 20%.

14. The process according to claim 1, wherein, in the alkoxycarbonylation, a C2 to C20 hydrocarbon having at least one olefinic double bond is reacted with an alcohol and carbon monoxide in the presence of the homogeneous catalyst system.

15. The process according to claim 14, wherein the alcohol used in the alkoxycarbonylation is a mono- or polyol (two or more OH groups) having 1 to 50 carbon atoms or a mixture of two or more mono- and/or polyols.

16. The process according to claim 1, wherein the acid is a Brnsted acid having a pKa 3, or a Lewis acid having an LAU value of more than 29.

17. The process according to claim 2, wherein the acid is a Brnsted acid having a pKa 3, or a Lewis acid having an LAU value of more than 29.

18. The process according to claim 12, wherein the separation-active layer is composed of PEEK, preferably having a degree of sulfonation of less than 10%.

19. The process according to claim 2, wherein the acid is a Brnsted or Lewis acid, the Brnsted acid being selected from the group consisting of perchloric acid, sulfuric acid, phosphoric acid, methylphosphonic acid or a sulfonic acid and the Lewis acid being selected from the group consisting of aluminium triflate, aluminium chloride, aluminium hydride, trimethylaluminium, tris(pentaflurophenyl)borane, boron trifluoride, boron trichloride, titanium(IV) isopropoxide, Bu.sub.2SnO, BuSn(O)OH, or mixtures thereof.

20. The process according to claim 2, wherein the base is selected from the group consisting of alkali metals, alkali metal oxides, alkali metal or alkaline earth metal acetates, alkali metal or alkaline earth metal oxides, alkali metal or alkaline earth metal alkoxides, such as NaEtOH or MgEtOH, and alkali metal carbonates such as K.sub.2CO.sub.3 or Cs.sub.2CO.sub.3.

Description

EXAMPLES

Example 1

Increase in Membrane Retention for Various Membrane Materials

[0046] The tests were carried out in a commercially available dead-end batch-filtration cell of the METcell model from Evonik MET with two commercially available membranes from Evonik (Evonik DuraMem 300, polyimide-based) and Borsig (Borsig oNF-2, polydimethylsiloxane-based). The PEEK membrane according to the invention was produced in accordance with the publication J. da Silva Burgal et al.; Journal of Membrane Science, vol. 479 (2015), pp. 105-116 (see also WO 2015/110843 A1).

[0047] Tests were carried out under the following conditions: 56.7 cm.sup.2 active membrane surface area, 20 bar transmembrane pressure, 250 rpm stirrer speed.

[0048] The principal constituents of the mixture investigated are methanol in a content of 43% by weight and methyl octanoate in a content of 57% by weight. To determine the retention, the ligand 4,4-di-tert-butyl-2,2-dipyridyl (44tB22) was added to the test mixture in a content of 0.01% by weight. A first measurement was then carried out. 0.5% by weight of aluminium triflate (based on the total weight of the mixture) was then added as acid and the measurement of permeability and retention was repeated.

[0049] The METcell was filled with 200 ml of the above mixture, the operating pressure was then applied and 100 ml of permeate was then driven out, with continuous recording of the permeate weight. At the end of the test, once 100 ml of permeate had permeated out of the METcell through the membrane, samples were taken for GC and HPLC analyses. The reported permeability is an average across the 100 ml of permeate collected. After each run of the METcell, the permeate driven out was returned to the METcell, i.e. the permeate was mixed again with the residual retentate in the cell.

[0050] The retention was determined by HPLC-UV on a C18 column. Any method known to those skilled in the art in which the individual components of the mixture are reliably separated by chromatography and can thus be distinguished is suitable for the analysis.

TABLE-US-00001 TABLE 1 Comparison of permeability Example Membrane P.sub.0 (before addition) P.sub.(after addition) Ligand Acid w(acid) L/m.sup.2h.sup.1bar.sup.1 L/m.sup.2h.sup.1bar.sup.1 I PEEK 44tB22 Al(OTf).sub.3 0.5% by wt. 0.3 0.3 II Borsig oNF-2 44tB22 Al(OTf).sub.3 0.5% by wt. 0.7 1.6 III PuraMem S 44tB22 Al(OTf).sub.3 0.5% by wt. 0.7 2.0

TABLE-US-00002 TABLE 2 Comparison of membrane retention Example Membrane R.sub.0 (before addition) R.sub.(after addition) Ligand Acid w(acid) % % I PEEK 44tB22 Al(OTf).sub.3 0.5% by wt. 54 81 II Borsig oNF-2 44tB22 Al(OTf).sub.3 0.5% by wt. 20 94 III PuraMem S 44tB22 Al(OTf).sub.3 0.5% by wt. 57 78

[0051] The examples show clearly that membrane retention is increased by addition of an acid, irrespective of the membrane material used. Whereas the permeability of the type oNF-2 and PuraMem S membranes initially shows an increase that is advantageous in principle, stability problems become apparent in the long term. The PEEK membrane shows both a considerable increase in the retention of the component to be retained and acceptable permeability that does not change appreciably as a result of the addition of acid.

Example 2

[0052] Increase in Membrane Retention for Various Phosphine Ligands with the PEEK Membrane

[0053] The tests were carried out in a continuously operated test system with complete recycling of the permeate in a closed loop. The system essentially comprises a high-pressure through-flow loop pressurizable up to 60 bar having a flat-channel membrane cell. The loop is fed from a reservoir filled with a feed solution that is mechanically mixed and blanketed with argon. An HPLC pump is used to bring the feed solution to the operating pressure of the membrane loop and thus to the high-pressure region of the test system. The high-pressure region of the test system consists essentially of a liquid loop, which is operated by means of a circulation pump, and a flat-membrane test cell and also the necessary sensors (e.g. pressure measurement, temperature measurement). The liquid flow penetrating through the membrane is withdrawn from the membrane module as permeate and recycled into the reservoir. The amount of permeate is measured on a balance. The excess feed volume (supply stream to the high-pressure pump minus the total permeate of the membrane) is likewise recycled into the reservoir. This recycling is effected by means of a mechanical supply pressure regulator, which is also used to set the supply pressure for the nanofiltration stage. The loop is heated by means of a thermostat in order to ensure a defined temperature for the separation.

[0054] The tests were carried out under the following conditions: active membrane surface area per module 84.5 cm.sup.2, transmembrane pressure 45 bar, separation temperature 50 C. The membrane material used was PEEK (produced in accordance with example 1).

[0055] A feed solution consisting of 43% by weight of methanol and 57% by weight of methyl octanoate (molar ratio 4:1) was investigated. Determination of the retention was carried out using in each case 0.1% of the phosphine ligands listed in table 3. The retention for the ligands was first determined over a period of 70 h. 0.5% by weight of aluminium triflate (based on the total weight of the mixture) was then added as the acid and the retention measurement was repeated. The retention of the ligands was determined by HPLC with UV detector (220 nm) on a C18 column.

TABLE-US-00003 TABLE 3 Comparison of the retention of various phosphine ligands with aluminium triflate R.sub.0 (before R .sub.(after .sub.addition) .sub.addition) Example Ligand % % IV Diphenyl(2-pyridyl)phosphine [00001] 60 85 V 1,2-Bis[(2-pyridyl)tert- butylphosphinomethyl]benzene [00002] 80 92 VI 1,1-Bis[(2-pyridyl)tert- butylphosphinomethyl]ferrocene [00003] 80 95 VII 2,6-Bis(di-tert- butylphosphinomethyl)pyridine [00004] 80 99

[0056] Table 3 shows that an increase in retention occurs after addition of acid, irrespective of the ligand structure and of the number of phosphorus centres.

Example 3

[0057] The entire test procedure was repeated in accordance with example 2, but with addition of sulfuric acid instead of aluminium triflate. The results are shown in table 4.

TABLE-US-00004 TABLE 4 Comparison of the retention of various phosphine ligands with sulfuric acid R.sub.0 (before R .sub.(after .sub.addition) .sub.addition) Example Ligand % % IV Diphenyl(2-pyridyl)phosphine [00005] 59 84 V 1,2-Bis[(2-pyridyl)tert- butylphosphinomethyl]benzene [00006] 80 93 VI 1,1-Bis[(2-pyridyl)tert- butylphosphinomethyl]ferrocene [00007] 79 95 VII 2,6-Bis(di-tert- butylphosphinomethyl)pyridine [00008] 81 99

[0058] Table 4 shows that an increase in ligand retention occurs for sulfuric acid to the same degree.