NO871595L - COMPOSITE MEMBRANE. - Google Patents

Description

Different composite membranes have been developed for certain separation processes. These can be used for areas that include gas separation, liquid/liquid separation (pervaporation), reverse osmosis and ultrafiltration. These composite membranes comprise a porous membrane

as a support for a permeable or microporous film. For gas separation and pervaporation, the film is selectively permeable to a component of the feed mixture. For reverse osmosis and ultrafiltration, films are used that differ from each other in the size of the pores, in ultrafiltration the pore diameter is generally in the range 2 - 2000nm, while reverse osmosis requires a pore size below 2nm in diameter. In all cases, it is desirable that the permeable or microporous film be as thin as possible to achieve high flow rates. In general, a film-forming material is applied to a porous carrier membrane in an amount sufficient to block any defects in the porous carrier, which could otherwise reduce selectivity.

In order to achieve a very thin film, but sufficiently supported, permeable or microporous film, the pores in the carrier membrane should be relatively fine.

Conventionally, an asymmetric organic polymer membrane of the ultrafiltration type has been used. However, such membranes have low porosity (often less than 1%) on them

its surface (skins), which means that the effective area for the composite membrane is low. Furthermore, ultrafiltration membranes have defects with a size often up to 1[im. When sufficient film-forming material is applied to block these, the result is a thicker film which tends to slow down the flow rate.

Currently, polymeric permeable or microporous films are used which are often formed from polymers deposited using solvent casting techniques. The polymers that can be used are limited by the requirement that the current solvents must not significantly degrade the carrier material. In a similar way, the structure of carrier material can be affected if plasma polymerization or curing techniques, heat or ultra-violet irradiation are used.

According to a feature of the invention, a composite membrane is provided comprising a porous, inorganic membrane which is asymmetric as a result of it having a system of large pores extending in from one side and a system of smaller pores extending in from the other side, and where the system of large pores is united with the system of smaller pores, as well as at least one permeable or microporous polymer film that lies over a surface of the porous membrane, preferably on the surface with smaller pores.

According to another feature of the invention, a composite membrane comprising a porous, anodic aluminum oxide membrane and at least one permeable or microporous polymer film is provided.

As porous supports, anodic alumina membranes exhibit a number of advantages compared to conventional materials.

It is well known that when an aluminum substrate is made anodic in an electrolyte that has a certain dissolving ability for aluminum oxide, a porous oxide film is formed on the surface of the metal, where the pore diameter and spacing can be controlled by the anode voltage and the film thickness of the electrolysis time.

During the past two to ten years, techniques have been developed to separate anodic oxide films from their metal substrate. The resulting membranes in most cases have partially or through-extending pores of substantially constant diameter along their length. For example, K.N.Rai and E. Ruckenstein (Journal of Catalyses, preparation no. 40 (1975) pages 117-123) describe the preparation of anodic oxide films which are porous in that they have essentially cylindrical pores extending all the way through the film. By appropriately controlling the anodizing conditions, it is possible to achieve uniform diameters in the range 10 - 2000nm.

In EPA No. 1788 31, anodic aluminum oxide membranes with an asymmetric structure are described. In these membranes, a system of larger pores extends in from one side and a system of smaller pores extends in from the other side, and the system of larger pores is connected to the system of smaller pores, so that the inner ends of one or several small pores are united to the inner end of a larger pore, and there are essentially no blinding larger pores. These membranes are particularly suitable for use as carriers for permeable or microporous films. The small pores extending from one side can be arranged to have mean diameters as small as 2nm, making the surface ideal as a support for a very thin active film. But the length of such small pores is very short, because they are connected to the larger pores, so that the porous carrier membrane nevertheless has a low resistance to flow. The asymmetric structure can also be important when controlling the thickness of the films that are deposited from solutions. Viscous braking effect means that capillary rise is much slower from the small-pored side.

Conventional porous anodic oxide membranes have a non-porous barrier layer between the bottom of the pores and the metal/oc interface and the thickness of the barrier layer is dependent on the anodization voltage. The asymmetric membranes according to EPA No. 178831 are produced by a controlled technique by slow voltage reduction, which aims to make the barrier layer thinner by forming progressively finer pores in it and finally to release any remaining barrier layer at the metal/oxide interface.

Other asymmetric porous inorganic membranes are described in the literature, for example aluminum oxide or other metal oxide membranes produced by sol-gel techniques, as described by A.F.M. Leenars et al in J-Mat.Sci., 19

(1984) 1077-1088.

An average pore diameter of the porous membrane at the surface where it is in contact with the permeable or microporous film should preferably be less than 100nm and preferably less than 30nm or 10nm. Membranes formed by the sol-qel techniques can have pore sizes down to lnm. In general, the thickness of the permeable or microporous film should be at least as large as the pore diameter of the support to ensure that the film is adequately supported. Thus, a membrane carrier with fine pores will enable the application of a thin film on it.

The porous membrane may itself be composite. For example, it can comprise a porous anodic aluminum oxide membrane, either asymmetric or of a conventional type, or be a tape-cast membrane and a porous film deposited on the membrane using a sol-gel technique. Porous membranes of this type are described in Norwegian patent application no. 87.1596. An advantage of such porous membranes is that they can be arranged to have relatively fine pores on the surface such as the permeable or microporous

watch movie is on. The nature of the permeable or microporous polymer film is not critical to the invention. Although any film can be used which is currently used in a supported state for the purposes contemplated, two materials in particular are particularly considered.

These are: Organic polymer films (included in the term "organic polymers" are, for example, polymers based on silicon which can also contain organic groups, see "J. Membrane Science", 22 (1985) pages 257 - 258) which will generally be permeable , but not porous, as well as polymers with layers that are limited to a thickness of a few molecules, and which can be produced, for example, by Langmuir-Blodgett techniques.

There is an extensive literature regarding thin permeable polymer films formed on porous supports. The subject is covered in a review article entitled "Separation of Gases with Synthetic Membranes" by S.L. Matson et al in Chemical Engineering Science, 38.4 (1983) pages 503-524. The polymers may be linear or cross-linked or may be prepared by plasma polymerization, and may be intended for use above or below their glass transition temperature. An extensive list of polymers is given in the examples in EPA 174918. Application of a film of the polymer to a porous membrane can conveniently be achieved by casting from a solution in an organic solvent. It is necessary that the film forms bridges over the pores in the carrier, but does not fill the pores. Techniques for achieving this are described in the literature, including control of the viscosity of the polymer solution and by immersing the porous membrane in the solvent.

The molded polymer film can advantageously be made as thin as possible, subject to the requirement that it bridges all the pores in the carrier.

Other known techniques for applying films to porous supports include:

- lamination,

- dip coating, spin coating, roller coating of po- the lymer solution with subsequent drying, coating with monomer or prepolymer solution, and

curing by means of heat or radiation, plasma polymerization,

interfacial polymerization on the support surface.

It has been proposed to use crosslinked Langmuir-Blodgett films for separation purposes, for example for reverse osmosis, pervaporation and gas separation (K. Heckmann et al, Thin Solid Films, 99 (1983) pages 265 to 269). Such films can include, for example, stearic acid or a metal stearate, for example magnesium stearate or esters of stearic acid with polymerizable groups, for example vinyl stearate. These monomolecular multilayer films can also be prepared using other techniques. These films, which are only a few molecules thick, can also be applied to porous membranes to give the composite membranes according to the invention. Although formation of one or more monomer monolayers by dip coating, followed by polymerization of the monolayers is possible, it is preferred to polymerize or crosslink the monolayers in Langmuir-Blodgett before dip coating. Repeated dip coatings can be used to produce films of the desired thickness, generally lnm - 20nm or more. In the pre-polymerized state, these films are capable of bridging pores that are significantly wider than the thickness of the individual films.

In the review article by S.L. Matson et al as mentioned above describe homogeneous films which can be permeable to gases, as well as microporous films through which gases can pass by different mechanisms. Films rated as microporous generally have a pore size of at least lnm, films with pores below this size can better be considered non-porous, but still permeable. At the other end of the scale, films that are considered micro-porous will generally have a pore size that is not significantly larger than the mean free path for the gas in them, films with pores above this size can better be considered porous. Depending on the gas composition, temperature and pressure, microporous films can have pore sizes in the range of up to lOOnm or even more. Applications of composite membranes comprising films generally considered permeable but not porous include reverse osmosis, hyperfiltration, pervaporation and gas separation. Applications of composite membranes comprising microporous films include ultrafiltration, dialysis and diafiltration. For the production of composite membranes according to the invention, porous anodic aluminum oxide membranes, in particular asymmetric membranes as described in EPA 178831, and membranes coated using sol-gel techniques, for example described in Norwegian patent application no. 87.1596, offer the following advantages: a) Smaller pores with narrow pore size distribution. This allows the application of thin films to provide high flow rate composite membranes. Polymeric ul-traf iltration membranes have a much wider pore size distribution, b) High porosity which leads to a high flow rate. The finely porous surface of asymmetric, anodic membranes has a porosity of approx. 20%, while asymmetric polymeric ultrafiltration membranes generally exhibit less than 1% surface porosity. c) Ferre defects leading to high selectivity. For example, asymmetric anodic membranes will retain bacteria that have a size greater than 0.3 µm, while polymeric ultrafiltration membranes with the same average pore size are considered not to be sufficiently perfect to safely retain such bacteria, d) Resistance to solvents, temperature and radiation, allowing a wide choice of film materials and application techniques. Such membranes tolerate temperatures above 400°C, UV and other radiation and a wide range of organic solvents. In general, polymeric membranes cannot withstand temperatures above 200°C, and they can be affected by UV radiation and behave variably in contact with different organic solvents, e) Smoothness which leads to fewer defects in the applied films, f) Stiffness that reduces deformation under pressure. Polymeric membranes are easily deformed, which can have a damaging effect on the structure and the effectiveness of composite membranes formed from them.

In the following examples, the porous inorganic membranes are asymmetric anodic oxide membranes prepared according to the technique in EPA 178831. The membranes were approx. 60 Mm thick, and had a system of larger pores (approx. 200 nm), extending from one face and connected by a system of smaller pores (approx. 25 nm) extending in from the other side. Unless otherwise stated, the permeable or microporous film was applied to the microporous surface of the membrane.

Example 1.

This example describes the production of composite membranes by interfacial polymerization. Asymmetric anodic films were dipped in a 0.33% aqueous solution of polyethyleneimine (PEI), rinsed and then immersed in a 1% solution of terephthaloyl chloride in hexane for 1 min.

for cross-linking the polymer. The samples were then allowed to dry and then cured at 120°C for 10 min.

Examination using a scanning electron microscope typically shows a 0.4 µm thick PEI layer, but the very thin condensed film could not be released. There were certain signs that the coating had penetrated into the fine pores (approx. 25 nm diameter) in the separated anodic film. It also appeared that the membrane covered the openings of the wider pores (diameter approx. 0.2 µm) on the opposite surface.

Example 2.

The composite membrane prepared according to example 1 was used as a semipermeable membrane that separated two cells. One contained water and the other contained a copper sulphate solution diluted to the same level. Change in the solution level and the copper ion concentration in the two cells was followed for 22 hours. The change in solution level corresponded to a water permeability of 0.068 ml/cmV hour. Absorption at 810 nm due to copper ions determined using a spectrophotometer. The absorbance decreased from 0.371 to 0.313 in the copper sulfate solution and increased from 0 to 0.002 in the water. Thus, the water penetrated through the membrane driven by the osmotic pressure, while the copper sulphate was largely retained. It is believed that these membranes would be useful for reverse osmosis.

Example 3.

A composite membrane prepared according to example 1 was placed in a double cell apparatus where one cell contained water and the other cell contained a mixture of water and 21.3% methanol. After 4 1/2 hours the water had penetrated the water/methanol mixture at a rate of 0.38 ml/cm^/hour. A very low level of ethanol was detected chromatographically in the cell that originally contained pure water. This corresponded to a penetration rate of 0.025 ml/cm<2>/hour.

Example 4.

A composite membrane prepared according to Example 1 was used in the double cell apparatus, separating methanol from propanol. After 1 1/2 hours there was a net increase in volume of 8.5 ml in the half-cell containing propanol. This represents a net permeability of 1.6 ml/cm<2>/ hour. Gas chromatography was used to determine the composition in the two half-cells, and the proportions of propanol and methanol are given below.

These data indicate selective permeability to methanol,

and which indicates that the composite membrane may be useful in the pervaporation of two separate alcohols.

Example 5.

An example describes the production of a composite membrane using plasma polymerization.

Asymmetric anodic films were coated by exposure with a plasma generator from a trifluoromethane monomer. This was carried out in a plasma etcher (Model No. PE80) Plama Technology Ltd. The anodic films were placed on the lower electrode near the gas supply port. The aluminum electrodes had a diameter of 30 cm and were 5 cm apart. After evacuation to approx. 10 torr for 30 min. the samples were exposed to an argon plasma (10 7 KHz, 0.2 torr, 140W) for 10 min. and then for CHF3plasma (107 kHz,

0.2 torr, 200W, 1.IA) for approx. 20 min. with a CHF3 flow rate of 100 cm.3/min.

Example 6.

The composite membrane prepared according to example

5 was investigated with regard to gas permeability. It was placed in a cell where the feed gas passed over a surface at 40 ml/min. and a carrier gas which passed the second surface at 85 ml/min. The feed gas was a mixture of helium and the gas whose permeability was to be determined. The carrier gas was also helium. The total pressures on each side of the membrane were equalized. The cell was kept at 21°C. A mass spectrometer was used to determine the proportion of either oxygen, nitrogen, carbon dioxide or methane in the feed and permeate streams. These data were compared with the data obtained for an uncoated, separate anodic film.

For both the uncoated and coated porous anodic membranes, there appears to be a linear relationship between the proportion of gas in the permeate and the feed streams for 12 to 100% of the feed stream. These ratios and the proportion of gas in the permeate for the composite membrane relative to the uncoated separate anodic film were 0.25 for oxygen, 0.29 for nitrogen, 0.42 for carbon dioxide and 0.26 for methane. Although there is little difference in the results for oxygen, nitrogen and methane, it seems that the composite membrane is selectively permeable to carbon dioxide. This can thus be used to remove carbon dioxide from a mixture of hydrocarbons. When the feed stream consisted of 100% carbon dioxide, the permeate contained 5.6% carbon dioxide.

Example 7.

A composite membrane prepared according to Example 5 was attached to the end of a glass tube using an epoxy adhesive - 10 ml of ethanol was introduced into the tube and the top was covered to prevent evaporation. Gas chromatography was used for analysis. The end of the tube with the membrane was immersed in a beaker containing 150 ml of water. After 24 hours, the concentration of ethanol in the tube was 99% compared to the original concentration. This difference was not significant. However, the ethanol had permeated through the membrane and into the water at a rate of 0.07 ml/cm^/hour.

The membrane is hydrophobic and was not expected to be substantially permeable to water. The experiment indicated selective permeability to ethanol.

Example 8.

This example shows the production of a composite membrane by plasma polymerization.

Deposition of a separate anodic film was carried out in a tubular reactor, 25 cl long with a diameter of 5 cm, to which 13.65 MHz radio frequency power was inductively coupled via a 9-winding copper coil which was externally wound around a 10 cm length of the tube. Trifluoroethylene was introduced into the reactor at 0.4 7 cm 2 (S.T.P.)/min. while a vacuum pump

pe maintained a pressure of 0.06 mbar. A power of 10 W was applied for 10 min. while the separated anodic film was placed 6 cm before the winding.

Examination using a scanning electron microscope showed that a 0.14 µm thick coating had formed on the surface of the separated anodic film, with no evidence of scaling inside the substrate's pores with a diameter of 25 mn.

Example 9.

A composite membrane, prepared according to example

8, was investigated with respect to gas permeability. The membrane was placed in a permeability cell with an argon feed gas passing a surface at 25 ml/min.

and an argon carrier gas which passed the second surface at 100 ml/min. (all pressure determinations corrected to standard pressure and temperature). The total pressures on each side of the membrane were equalized before introducing streams of oxygen, nitrogen, hydrogen, carbon dioxide or methane (flow rates from 5 to 10 ml/min.) to the feed gas. A mass spectrometer was used to determine the proportion of each gas in the feed and in the permeate streams, but the pressure difference across the membrane as a result of added gas was monitored. Permeability data and the flow rate of gas through the membrane are given in the following tables. The results indicate that these composite membranes make it possible to achieve

high gas flow rates.

* The flow rate measurements are corrected for the unit membrane area and partial pressure difference of 1 kPA.

The membranes show high permeability for H2 and can be used for hydrogen separation. The membranes also exhibit selectivity for nitrogen compared to oxygen, and can therefore be used for air separation.

Example 10.

This example describes the preparation of a composite membrane by the deposition of monomolecular multilayers on the surface of separated anodic films using the Langmuir-Blodgett technique.

Pentacosa-10,12-dienoic acid was prepared by standard methods (B. Tieke, G. Wegner, D. Naegela and H. Ringsdorf, Angew. Chem. Int. Ed. Engl., 15, 764 (1976), B . Tieke and G. Wegner in "Topics in Surface Chemistry", Plenum,

New York, 1987, p 121, B. Tieke, G. Leiser and G. Wegner, J. Polym. Sei., Polym. Chem. Ed., 17, 1631 1979). The multilayers were deposited using a 1000 cm Langmuir trough (Nima Technology Ltd). This was equipped to maintain a constant surface pressure during deposition. The subphase in the trough was a 0.005 mol/l high purity aqueous solution of cadmium chloride adjusted to pH 6.0 and kept at 20°C. Initially, the separated anodic film was immersed through the surface of the subphase. A monolayer of diacetylene monomer was dispersed from a 1 mg/ml solution of a 4:1 hexane/chloroform mixture, compacted to 25 MN/m and polymerized by exposure to 254 nm/ultraviolet light for 10 min. The first monolayer of polymerized diacetyl was deposited by raising the anodic film vertically at a speed of 5 mm/min. through the surface. The surface pressure was kept constant at 25 mN/m during the deposition. Using the above-described procedure, a new polymerized monolayer of diethylene was produced on the subphase, and two additional layers were deposited on the dry anodic film by dipping at 5 mm/min. Newly polymerized monolayers were prepared prior to deposition of the 4, 6, 8 etc, on the anodic film.

Examination using a scanning electron microscope showed that an 11-layer thick coating was approx. 20 nm thick and extended over the pores with a diameter of 25 nm in the substrate.

Example 11.

Composite membranes were produced by the method according to example 10. On a sample A, 11 layers were deposited, which gave a hydrophobic surface, and were used to separate ethanol from water. The second sample B had 10 deposited layers which made it hydrophilic, and was used to separate water from an ethanol/water mixture. The following table shows the results of the permeability tests. Diffusion of ethanol appears independent of the nature of the membrane surface. However, the water permeability is influenced by the wettability of the surface. Both of the membranes shown exhibit a selectivity that can be useful for pervaporation processes.

Example 12.

A separated anodic film was prepared with an alumina coating formed on its surface using sol/gel techniques. Using this, a composite membrane was produced by using the interfacial polymerization method according to example 1. From an electro-optical examination of the PEI layer, this appears to be similar to that in example 1, except that it was only 0.2 µm thick. The composite membrane was used in example 3. The water permeability rate was 0.27 ml/cm/hour,

while the permeability to ethanol was 0.0 2 ml/cm^/hour. The selectivities were essentially the same as those shown

in example 3.

Example 13.

A composite membrane corresponding to that according to example 12 was used as described in example 2. The water permeability into the copper sulfate solution was 0.054 ml/cm^/hour, while no copper ion could be detected in the cell containing water. It seems that this membrane provides superior ion retention than that according to example 2, coupled with a lower water diffusion. This may reflect the membrane's degree of perfection.

Example 14.

Composite membranes were prepared according to Example 10 with 11 layers deposited on the surface to give a uniform 0.02 µm thick coating. These samples were examined for gas permeability using the method similar to that indicated in Example 9. The flow rates and measurement techniques were also as indicated in Example 9. The results for permeability and the gas flow rate through the composite membrane for nitrogen, oxygen, carbon dioxide and methane is given in the following table.

For all gases, the composite membrane exhibited a high re-flow rate with good gas permeability. The selectivity of the membrane for CH4 over CO2 and N2 over 02 indicates that they can be useful for biogas treatment and air separation respectively.

Example 15.

165 g of aluminum oxide and 0.42 g of magnesium oxide were suspended in a liquid system consisting of 78 g of trichlorethylene,

32 g ethanol, 3.8 g corn oil, 8.4 polyvinyl butyral and 14.2

g polyethylene glycol. The slurry ink was cast on a glass substrate using a coating sheet to a film with a width of 173 mm, which was dried in air to give a flexible tape with a thickness of 0.14 mm. Discs of 26 mm diameter were cut from the strip and partially sintered to give a porous ceramic material with an average pore size of 0.3 µm. The discs were pre-treated by immersion for 5 s in a 5% solution of sodium metasilicate, and then spray-coated with viscous boemisol with a concentration of 15 g/l, where the water had been replaced by ethylene glycol. The coated porous substrate was heated

for 1 hour at 450°C to convert the gel layer into a stable -A^O^ film containing pores with a slot width of 4.2 nm.

An aqueous boehmitsol with a concentration of 30 g/l was then deposited on the first sol-gel layer by spray coating. This was followed by heating for 1 hour at 450°C which formed a stable )'-Al 2 O 3 film with a slot width of 2.8 nm.

By using this porous membrane, a membrane was produced by using the interfacial polymerization method according to example 1. This was used in the same way as in example 3. The water permeability rate was 0.2 ml/ cm 2 /hour, while that for ethanol was 0.02 ml/ cm 7/hour. These rates are very similar to those shown in examples 1 and 12, where the same type of thin polymer film was used, but on different supports.

Claims (10)

1. Composite membrane, characterized by a porous inorganic membrane which is asymmetric in that it has a system of large pores extending in from one face, and a system of small pores extending in from the other face, and where the system of large pores are connected to the system of small pores, as well as at least one permeable or microporous polymer film overlying one surface of the porous membrane. 2. Composite membrane, characterized by a porous anodic aluminum oxide membrane and at least one permeable or microporous polymer film. 3. Composite membrane according to claim 2, characterized in that the porous anodic aluminum oxide membrane is asymmetrical in that it has a system of large pores that extend in from one surface and a system of smaller pores that extend in from the other surface, and that the system of large pores is connected to the system of smaller pores. 4. Composite membrane according to claim 1, characterized in that the porous membrane consists of aluminum oxide. 5. Composite membrane according to any one of claims 1-4, characterized in that the permeable or microporous film lies over small-porous surfaces of the porous membrane. 6. Composite membrane according to any one of claims 1-5, characterized in that the average pore diameter of the porous membrane, at the surface which is overlain by the permeable or microporous film, is in the range 1 - 100 nm. 7. Composite membrane according to any one of claims 1-6, characterized in that the film is permeable non-porous organic polymer. 8. Composite membrane according to any one of claims 1-6, characterized in that the film is applied using a Langmuir-Blodgett technique. 9. Composite membrane according to claim 7, characterized in that the film is applied to the membrane by means of plasma polymerization or by interfacial polymerization at the surface of the membrane. 10. Use of the composite membrane according to claims 1-9 for separation purposes.