Membranes for dewatering acid mixtures

Abstract

The invention provides a membrane suitable for dewatering acidic mixtures, comprising a bridged organosilica directly applied on a macroporous support in the absence of an intermediate mesoporous or finer layer. The bridged organic silica comprises divalent C.sub.1-C.sub.9 organic groups A.sup.2 and/or trivalent C.sub.1-C.sub.9 organic groups A.sup.3 directly bound to the silicon atoms of the organosilica. In particular, the membrane comprises bis-silylmethane or bis-silylethane groups. The membranes effectively separate water from acidic mixtures at high temperatures and without decrease in performance for at least several months.

Claims

1. A supported membrane suitable for dewatering acidic mixtures, comprising a microporous separating layer and a macroporous ceramic support, wherein the microporous separating layer comprises a bridged organosilica directly applied on the macroporous ceramic support, in the absence of an intermediate mesoporous or finer layer, the bridged organosilica comprising divalent C.sub.1-C.sub.9 organic groups and/or trivalent C.sub.1-C.sub.9 organic groups directly bound to the silicon atoms of the organosilica.

2. The membrane according to claim 1, wherein the organosilica has the formula Si.sub.(1-x)M.sub.xA.sup.1.sub.iA.sup.2.sub.jA.sup.3.sub.kO.sub.1.25-1.85, wherein: M is one or more metal(s); A.sup.1 is a monovalent C.sub.1-C.sub.9 hydrocarbyl or fluorohydrocarbyl group; A.sup.2 is a divalent organic group having the formula C.sub.mH.sub.n; A.sup.3 is a trivalent organic group having the formulaC.sub.mH.sub.(n-1)<; i=0-0.67; j=0-0.75; k=0-0.33, 0.5<i+2j+3k<1.1; and 2j+3k>0.3; m=1-9 and n=2(m?p) and 0?p<m, wherein the values m, n and p may be different between formulas A.sup.1 and A.sup.2; x=0-0.2.

3. The membrane according to claim 2, wherein one or more of the following values apply: A.sup.1 is a monovalent C.sub.1-C.sub.9 hydrocarbyl group; i=0-0.33; j=0.25-0.5; k=0-0.1, m=1-6; x=0-0.1.

4. The membrane according to claim 3, wherein one or more of the following values apply: A.sup.1 is a monovalent C.sub.1-C.sub.4 hydrocarbyl group; i=0-0.2; j=0.35-0.5; m=1-6; x=0-0.1.

5. The membrane according to claim 4, wherein one or both more of the following values apply: j=0.45-0.5; m=1 or 2.

6. The membrane according to claim 2, wherein A.sup.1 has the formula C.sub.qH.sub.rF.sub.s, wherein q=1-4, and r+s=2(q?t)+1 and 0?t<q.

7. The membrane according to claim 1, wherein the macroporous support comprises a alumina.

8. The membrane according to claim 1, which has a stable water/acetic acid separation factor in the presence of 90% acetic acid in water at 100? C. of at least 100 for at least 120 days.

9. A process for producing a membrane of claim 1, wherein the membrane is suitable for dewatering acidic mixtures, comprising one or more cycles of (a) directly applying a bridged organosilica on a macroporous support in and (b) drying the support with the applied organosilica, until the membrane is essentially microporous, the bridged organosilica comprising divalent C.sub.1-C.sub.9 organic groups and/or trivalent C.sub.1-C.sub.9 organic groups directly bound to the silicon atoms of the organosilica.

10. The process according to claim 9, wherein the cycle of applying and drying of the organosilica is performed 2-4 times.

11. The process according to claim 9, further comprising (c) calcining the dried support with applied membrane.

12. The process according to claim 9, wherein the bridged organosilica is applied by hydrolysis of a precursor having the formula (RO).sub.3SiC.sub.mH.sub.nSi(OR).sub.3 and/or (RO).sub.3Si C.sub.mH.sub.(n?1)[Si(OR).sub.3].sub.2 wherein R is C.sub.1-C.sub.4-alkoxy, m=1-9 and n=2(m?p) and 0?p<m.

13. The process according to claim 9, wherein the bridged organosilica further comprises monovalent groups of formula C.sub.qH.sub.rF.sub.s, wherein q=1-4, and r+s=2(q?t)+1 and 0 ?t<q, wherein the process comprises co-hydrolysis of a precursor having the formula (RO).sub.3Si-C.sub.qH.sub.rF.sub.s, wherein R is C.sub.1-C.sub.4-alkoxy.

14. The process according to claim 9, wherein the macroporous support comprises ?-alumina.

15. A process for dewatering an aqueous acidic mixture, comprising contacting a feed mixture containing at least water and an acid with a membrane according to claim 1 to obtain a permeate and a retentate, and recovering the retentate.

16. The process according to claim 15, wherein the feed mixture comprises 20-95 wt% of acid and 5-50 wt.% of water, and the retentate comprises less than 3% water.

17. The process according to claim 16, wherein the retentate comprises less than 1% water.

18. The process according to claim 15, wherein the acid comprises a carboxylic acid.

19. The process according to claim 15, wherein the acidic mixture results from the production of the acid.

20. The process according to claim 15, wherein the mixture results from an esterification process.

21. The process according to claim 19, wherein the mixture comprises between 20 and 80 wt.% of an alcohol and/or an ester of the acid and alcohol.

22. The process according to claim 15, wherein the mixture results from an acetalisation process.

23. The process according to claim 15, wherein the acidic mixture is a recycle stream in the production of another compound.

24. The process according to claim 15, wherein said contacting is performed at a temperature between 90 and 150 ? C.

Description

DESCRIPTION OF THE FIGURES

(1) FIG. 1 shows the water flux in an endurance test in concentrated acetic acid at 100? C. of a membrane not according to the invention (having a ?-Al.sub.2O.sub.3 layer).

(2) FIG. 2 shows the purity of water in the permeate using a feed of concentrated acetic acid at 100? C. separated by a membrane not according to the invention at 100? C.

(3) FIG. 3 show an endurance tests in concentrated acetic acid at 100? C. of a BTESE membrane according to the invention (without ?-Al.sub.2O.sub.3 layer) in 90/10 w/w HAc/H.sub.2O.

(4) FIG. 4 shows a photograph of a prior art membrane having an intermediate gamma-alumina layer.

(5) FIG. 5 shows a photograph of a membrane without intermediate layer according to the invention.

(6) FIG. 6 shows the long-term water flux and the purity of the permeate of a membrane of the invention in the separation of a 90:10 HAc/H.sub.2O mixture without and with added methanesulfonic acid.

EXAMPLES

(7) Sol Preparation:

(8) 1,2-Bis(triethoxysilyl)ethane (BTESE, purity 96%, ABCR) and 1,2-bis(triethoxysilyl)-methane (BTESM, purity 96%, ABCR) were used as a precursor for making a sol. The required amounts of ethanol absolute, distilled water and nitric acid (65 wt %, Aldrich) were mixed and added to the precursor mixture under continuous stirring. The mixtures had a [H+]:[Si] ratio of 0.043 and a [H.sub.2O]:[precursor] ratio of 6 and a [Si] concentration of 1.5 M. Finally 3.98 ml BTESE or 3.76 ml BTESM were added to 10 ml of the stock solution under continuous stirring and refluxed for 3h at 60? C.

Comparative Example

(9) A sol with BTESM as described above was coated directly on a tubular mesoporous ?-Al.sub.2O.sub.3 substrate with an average pore size of 3-4 nm. After drying for 12 hours at room temperature, the membrane was exposed to a heat treatment under N.sub.2 atmosphere by heating with 0.5? C./min to 250? C. with a dwell of 2h and cooling down again to room temperature with 1? C./min. Three individual BTESM membranes and one BTESE membrane were made according to Example 1 of WO 2010/008283 and tested in the dewatering of acetic acid of different compositions (70, 80 and 90 HAc/H.sub.2O w/w). The membrane subjected to a composition of HAc/H.sub.2O of 80/20 w/w lost selectivity after 21 days of operation. Three other individual membranes subjected to a similar concentration showed shorter life times. FIG. 4 shows a SEM of a tubular membrane with mesoporous ?-Al.sub.2O.sub.3 and microporous BTESE layer.

(10) It can be seen in FIG. 1 that the flux increases after 5 days for BTESM and 17 days for BTESE (90/10) and after 15 days for BTESM (70/30), indicative of a structure that is opening up. Under steady performance, commonly an initial reduction is followed by a constant water flux. This is the first indication of low stability.

(11) FIG. 2 shows that the concentration of water in permeate resulting from 70/30, 80/20 and 90/10 w/w HAc/H.sub.2O at 100? C. decrease over time in case of BTESM and BTESE. The selectivity losses after about 4 days and 14 days (90/10), 6 days (70/30) and 24 days (80/20) is clearly shown.

Example 1

(12) A sol with BTESE as described above was coated directly on a macroporous tubular ceramic support of ?-Al.sub.2O.sub.3 with average pore size of approximately 0.17 ?m. After drying for 12 hours at room temperature, the membrane was exposed to a heat treatment under N.sub.2 atmosphere by heating with 0.5? C./min to 250? C. with a dwell of 2h and cooling down again to room temperature with 1? C./min. This process was performed for a total of three times until a microporous BTESE layer was formed on top of the macro-porous support. Two individual BTESE membranes were made according to Example 1 of WO 2010/008283 (with the exception of being deposited on the ?-alumina) and were tested in dewatering of acetic acid. The configuration of a tubular membrane without intermediate ?-alumina layer is illustrated in FIG. 5. Pervaporation in mixtures of acetic acid and water of 90/10 wt % at 100? C. showed that the performance of the membrane according to the invention was unchanged after a period of more than 337 days (FIG. 3). The purity of the water in the permeate was always above 97%, and did not show any tendency to reduce over time. The separation factor increased from an initial value of over 200 to a value of over 3000 after 180 days. The procedure was repeated with a second membrane and showed essentially the same result.

(13) The membrane without the ?-alumina layer (FIG. 5) was tested in acetic acid and water of 90/10 wt % at 100? C. for 148 days. After this period 2 wt % of methanesulfonic acid was added to the feed mixture during operation. The membrane had a separation factor of more than 300 during the complete measurement period of 442 days. The results are shown in FIG. 6.

Example 2

(14) Example 1 was repeated with the only difference that BTESM was used instead of BTESE. Two individual BTESM membranes were made and were tested in the dewatering acetic acid. The results are similar to those of Example 1.

REFERENCES CITED

(15) Agirre, et al. 2012, J Chem Technol Biotechnol 87, 2012, 943-954 Aminabhavi et al. 2003: Designed Monomers & Polymers, 6 (3), 2003, 211-236 Campaniello et al. 2004: Chem. Commun. 2004, 834-835. Castricum et al. 2008-a: Chem. Commun. 2008, 1103-1105; J. Mater. Chem. 2008, 18, 1-10 Castricum et al. 2008-b: J. Membr. Sci. 2008, 324 (1-2) 111 Iglesia et al. 2007: Chemical Engineering Journal 2007, 131 (1-3) 35-39 Ishida et al. 2006: J. Chemical Engineering Japan 2006, 38 (11): 903-907 Khajavi et al. 2010, J, Membr. Sci. 2010, 356 (1-2), 1-6 Kreiter et al. 2008: J. Membrane Science 2008, 319 (1-2): 126-132. Kreiter et al. 2009: ChemSusChem 2009, 2(2), 158. Masuda et al. 2003: Separation and Purification Technology 2003, 32: 181-189. Sanz et al. 2006: Chemical Engineering Journal 2006, 123: 9-14. Tsai et al. 2000: J. Membrane Science 2000, 169: 255-268 Tsuru et al. 2012: J Membr. Sci. http://dx.doi.org/10.1016/j.memsci.2012.06.012 Van Veen et al. 2011: J. Membr. Sci. 2011, 380, 124-31 Zhou et al. 2006-a: Industrial & Engineering Chemistry Research 2006, 45: 1787-1796. Zhou et al. 2006-b: Polymer 2006, 47: 280-288.