Patent Application
20080017578
This application is a National Stage filing of International Application PCT/CA2005/000518, filed Apr. 7, 2005, entitled “MEMBRANE STACKS”, claiming priority to and the benefit of U.S. Provisional Patent Application No. 60/560,027, filed Apr. 8, 2004. Both references are expressly incorporated by reference herein, in their entirety.
The invention relates to filtration membranes, particularly stacks of membrane layers separated by spacer layers.
A new generation of membranes suited for membrane chromatography and a variety of other applications including microfiltration, ultrafiltration, membrane absorbers, etc. has been developed. The membranes are useful in e.g. bio-separations where they can be used as disposable devices that have superior performance to existing separation systems.
These membranes comprise a macroporous gel anchored within a suitable non-woven support. In order to achieve high protein and other molecule adsorption capacities in chromatography and related applications, the gels are typically relatively soft, thereby allowing diffusion of the target molecules into the gel phase. Using this approach, dynamic capacities that are a factor of six higher than the best alternative membrane products can be achieved.
There is a downside to using soft macroporous gels. This comes from the fluxes that can be achieved with these membranes, particularly when the membranes are combined to form a stack as is typically used in membrane chromatography applications. Ideally, the flux drop at a given pressure should decrease linearly as additional membranes are added to a stack. It is well known that the flux through a membrane is inversely proportional to the thickness of the active layer. For example, whereas the flux of a single Q-type membrane at 100 kPa is 2500 kg/m2h when eight layers are combined, the resulting flux is 135 kg/m2h, rather than the expected value of 2,500/8=312 kg/m2h. This reduction in flux can be offset by increasing the trans-membrane pressure. But this comes with the added cost of more expensive membrane housings, pumps, etc.
The problem becomes greater in higher capacity gel based membranes with softer gels. Stacks of these membranes exhibit a non-linear pressure flux relationship. In severe cases, the flux reaches a limiting value as the pressure is increased. This effect limits the thickness of the membrane stacks that can be used if high flow rates are to be maintained.
It is known to use certain types of spacers between layers in membrane stacks for particular purposes.
Spacers are common elements in membrane systems or modules used in desalination processes such as reverse osmosis, electrodialysis, electro-deionization and diffusion dialysis. In these systems a “spacer” is a device that provides a generally defined distance between two adjacent membrane sheets to allow cross-flow of a liquid between the two membrane sheets. A feed fluid spacer is a porous spacer layer that provides for the passage of the feed fluid over and parallel to the active side of a membrane. This flow parallel to a membrane surface is termed cross-flow. The feed fluid spacer serves the function of directing the feed fluid to cover the active side of the membrane in a uniform manner. The spacer may also impart turbulence to the feed fluid to provide good mixing in the feed fluid as it travels over the active side of the membrane and to reduce concentration polarization.
A typical spiral wound membrane module comprising such spacers is shown in
The spiral module contains tightly packed membranes sandwiched between mesh spacers and wrapped around a small-diameter central tube. There are two spacers in this module; a permeate carrier layer as well as a feed spacer/distributor.
The membrane stack includes two, long semipermeable membranes with a spacer mesh between them. This is then wound up in a spiral tube with another spacer, the permeate carrier layer, to separate the outer, permeate sides of the stack. The mesh separator or spacer on the feed side allows the feed to be forced in one side of the spiral cylinder and out the other side. Pressure on the feed side forces some of the water (typically less than 15% of the feed water passes through the membrane on a single pass) to pass through the membrane where it is collected in the space between the membranes. The resulting permeate then flows around the spiral where it is collected in the centre of the tube. An early description of a spiral wound device is given in Bray D.T. Reverse osmosis purification apparatus. U.S. Pat. No. 3,417,870. 1968.
The key point in this design of a membrane module is that the spacers on the feed and permeate side allow flow of the two streams parallel to the surfaces of the membranes.
Another common example of membrane stacks is in electrodialysis cells. Electrodialysis is a process that uses a direct electrical current to remove salt, other organic constituents, and certain low molecular weight organics from brackish water. With this technique several hundred flat, ion permeable membranes and water flow spacers are vertically assembled in a stack. Half of the membranes allow positively charged ions, or cations, to pass through them. The other half-anion-permeable membranes—allow negatively charged ions to pass through them. The anion permeable membranes are alternately placed between the cation-permeable membranes. Each membrane is separated from the adjacent membrane in the stack by a polyethylene flow spacer.
An electrical current is established across the stack by electrodes positioned at both ends of the stack. Brackish water is pumped at low pressures into the flow spacers between each membrane and passes through the cell exiting on the opposite side. While passing through the cell, ionic constituents are electrochemically driven through the membranes on either side of the channel. This results in removal of salts from the brackish water stream and the formation of a concentrated salt steam in the intervening channels. The spacers serve to hold the membranes in place and allow the flow of the brackish water and resulting more concentrated streams parallel to the membrane surfaces. Typical stack designs are shown in the patents of Iaconelli (U.S. Pat. No. 3,695,444) and Olsen (U.S. Pat. No. 3,623,610).
It will be appreciated from the above that spacers are conventionally used to permit fluid to flow in the plane of the major surface of a membrane. The spacers used in such devices as spiral wound modules where tangential flow occurs across-the surface of the membranes from a feed channel containing the spacer, are typically quite thick and constructed such that flow channels remain even when sandwiched tightly between two membranes. This flow path is constructed to be tortuous such that movement of the feed through the spacer is turbulent. The design of these spacers is complex with different sized elements making up the mesh like material so as to provide the tortuous path.
But the art does not disclose the use of spacers in instances where the fluid flow path is substantially perpendicular to the plane of a major surface of the membrane layers in a membrane stack.
We have found that, by inserting spacer layers between individual membrane layers, the flux of the membrane stack can be increased dramatically. Surprisingly, the interleaving of spacer layers between the membrane layers in a membrane stack did not lead to a loss of resolution of the stacks in a separation. In fact, the breakthrough curve of a four-membrane stack with interleaving spacer layers was found to be somewhat sharper than the four layers without interleaving. Thus, the invention thus not only improves substantially the flux of a stack at a given pressure but also improves its performance in a separation.
Accordingly, in one aspect, the invention provides a membrane stack, comprising first and second membrane layers; and a spacer layer disposed between the first and second membrane layers; the membrane stack configured such that fluid passes through the membrane stack in a direction substantially perpendicular to the plane of the membrane layers and the spacer layer.
In another aspect, the invention provides a module comprising a membrane stack as described above, the module having a fluid flow path that is substantially perpendicular to the plane of the major surface of the membrane and spacer layers in said membrane stack.
In another aspect, the invention provides a method for separating a substance from a fluid, comprising the step of passing the fluid through a membrane stack or module as described above.
FIGS. 16A and 16B: ESEM images of a nascent (top) and AM610 (bottom) surfaces; (magnification: 5000×).
FIGS. 17A and 17B: ESEM images of the surface of AM611 membrane; (magnification: top—5000×, bottom—3500×).
FIGS. 18A and 18B: ESEM images of membranes AM610 (top) and AM611 (bottom); (magnification 5000×).
Membrane Layers
The membrane layers may be any kind of filtration membrane layers, and the invention is not limited to the use of particular types of membrane layers or combinations thereof.
Preferred membrane layers include those comprising structured gels incorporated into a support, as described in U.S. Pat. No. 6,258,276 (Childs et al.); International Publication No. WO 03/08078 (Childs et al.), R. F. Childs, A. M. Mika, The design of high performance, gel-filled nanofiltration membranes, in A New Insights into Membrane Science and Technology: Polymeric, Inorganic and Biofunctional Membranes@ Elsevier, Edit. A. Butterfield and D Bhattacharyya, 2003 pp 353-375; and U.S. Pat. No. 5,160,627 (Cussler et al.), all of which are incorporated herein by reference. In a particularly preferred embodiment, the membrane layers include macroporous gel-filled membranes as described herein and in International Publication No. WO 2004/073843 (Childs et al.), incorporated herein by reference. These structured gel membranes (MGel membranes) have significant advantages over existing traditional membranes. In terms of UF/MF applications they have very narrow pore-size distributions, have the ability to have dynamic pore-size control, are exceptionally hydrophilic, exhibit much lower fouling than other membranes and can be made at very low cost. In terms of adsorption/specific binding membranes the MGel platform provides membranes with exceptionally high loading capacities, the ready capability of introduction of specific ligands, and the ability to make single membranes with much higher thicknesses than conventional membranes. This latter feature means that fewer layers have to be used to obtain a given total stack thickness. Other useful membrane layers include, without limitation, microfiltration membranes or ultrafiltration membranes as are known in the art.
In one embodiment, a gel-filled membrane is used in combination with a conventional membrane such as a microfiltration or ultrafiltration membrane as mentioned above.
In another embodiment, a conventional microfiltration or ultrafiltration membrane is used in combination with, for example, adsorptive type macroporous gel based membranes. Alternatively, a macroporous gel based size exclusion membrane (microfiltration or ultrafiltration) may be coupled with an affinity type membrane(s). In each case a spacer may be used between the conventional membrane and the gel-filled membrane (as well as between any two or more gel filled membranes forming the stack).
In another embodiment, the membrane stack comprises macroporous gel based membranes with the same functionality but whose pore-size or density of active is or can be different in each layer. This combination may be used in e.g. a tubular arrangement as described herein, so as to even out hydraulic flow through the pores as the radius of the tube increases.
In another embodiment, homogeneous gel-filled membranes are used. These are effective as very tight ultrafiltration membranes and also as nanofiltration type membranes. In one embodiment, at least one such membrane is used in combination with a series of adsorptive type macroporous gel based membranes.
Spacers
Without intending to be bound to any particular theory, it is believed that the spacer layer serves largely to eliminate compression of surface layers, particularly on the macroporous gel based membranes. The macroporous gels are quite soft, mechanically weak and easily deformed. In order to use the macroporous gels as an effective separation membrane, it has been discovered that their mechanical properties could be increased by incorporating the gel into a porous membrane or material, including non-woven fabrics. Inevitably, such membranes will have a surface layer of the gel that is not strengthened by the support. Under conditions of hydraulic flow under pressure, with two membranes put together, it is believed that there is a compression of the unsupported gel surface layers. This compression will lead to a reduction in effective pore-size in this compressed region and a commensurate drop in flux over what would have been expected. Introducing a spacer layer between the two membranes would appear to prevent this compression of the surface gel layers.
The spacer may be any material that can separate the membrane layers and permit hydraulic flow of fluid substantially perpendicularly to the major surfaces of the membrane layers. It is desirable that the spacers do not permit fluid flow tangential to the major membrane surfaces. However, tangential fluid flow may occur across the surface of the first membrane in a stack.
Spacers may desirably possess one or more of the following features or properties:
By way of non-limiting examples, the spacer layer may be a material that is woven, moulded, formed, extruded, knitted, cast or formed in place.
In one embodiment, a thin mesh-like material having uniform thickness of the elements making up the mesh is used. Such a spacer can fit snugly between two membranes and only permit hydraulic flow to occur perpendicularly to the surfaces of the membranes.
Examples of commercially available spacer layers include symmetrical filtration netting made of polypropylene of thickness of 500 μm and mesh opening of 1000 μm (Naltex, Austin, Tex., USA) and polyester mesh of thickness of 350 μm and mesh opening of 500 μm (Polyester Monofilament) Great Lakes Filters Inc. Hillsdale Mich.
Membrane Stacks
Membrane stacks comprise a plurality of layers, including at least one filtration membrane layer and at least one spacer layer. The terms “membrane stack” and “layer” are used with reference to the fluid path through the membrane stack. As shown in
In some membrane stacks, the layers will indeed be composed of separate, discrete sheets of membrane material and spacer material. Such an arrangement is depicted in
Membrane stacks will typically comprise a plurality of filtration membrane layers interleaved with spacer layers, such that each adjacent pair of membrane layers is separated by a spacer layer. However, particularly because the membrane stack may contain other types of layers, or filtration membrane layers that are not especially susceptible to compression, in some embodiments, not every adjacent pair of filtration membrane layers will be separated by a spacer layer. Spacer layers may be used selectively, and included in the membrane stack only where needed to separate membrane layers susceptible to being compressed together.
Non-limiting examples of various possible combinations of membrane layers in the membrane stack are set forth in Table 1 below.
Membrane stacks will usually, although not necessarily, comprise at least one macroporous gel-filled membrane layer as described herein. Some embodiments may involve a combination of one or more macroporous gel-filled membrane layers coupled with one or more conventional membrane layers.
In one embodiment, conventional microfiltration and/or ultrafiltration membrane layers are used in combination with, for example, adsorptive type macroporous gel based membrane layers. Conversely, a macroporous gel based size exclusion membrane layer (microfiltration or ultrafiltration) may be used with one or more affinity type conventional membrane layers. In each case, a spacer layer would be used between the conventional membrane and the gel-filled membrane layer (as well as between any two or more gel filled membrane layers forming the stack).
In another embodiment, the membrane stack comprises one or more macroporous gel based membranes with the same functionality but whose pore-size or density of active material may be different in each layer. This may be useful in, for example, tubular arrangements so as to even out hydraulic flow through the pores as the radius of the tube increases.
In another embodiment, one or more membrane layers are homogeneous gel-filled membrane layers, as defined in R. F. Childs, A. M. Mika, The design of high performance, gel-filled nanofiltration membranes. A New Insights into Membrane Science and Technology: Polymeric, Inorganic and Biofunctional Membranes@ Elsevier, Edit. A. Butterfield and D Bhattacharyya, 2003 pp 353-375; U.S. Pat. No. 6,258,276 (Childs et al.); International Publication No. WO 03/08078 (Childs et al.); and U.S. Pat. No. 5,160,627 (Cussler et al.), the relevant parts of which are incorporated herein by reference. These are useful as very tight ultrafiltration membranes and also as nanofiltration type membranes. They exhibit the same type of surface gel layers as macroporous gel-filled membrane layers described below and the spacer layers can be effective where stacks are formed. Such homogeneous gel-filled membrane layers may be used in combination with a series of adsorptive type macroporous gel based membrane layers.
In one embodiment, the membrane stack is a multifunction membrane stack wherein several membranes are placed one on top of another to provide a composite separation device. These generally comprise a stack of filters/membranes that have size selection properties in the upper layers and adsorption functionality in the bottom layers as depicted in
Modules Comprising Membrane Stacks
Membrane stacks are generally assembled into a module for use. The module contains or carries the membrane stack and is configured such that the fluid is directed through the membrane stack in a fluid path that is substantially perpendicular to the plane of the major surfaces of the layers in the membrane stack. Generally, the module constrains the fluid such that all or a desired portion of the fluid flow must pass through the membrane stack, i.e. not escape around it.
Modules can be e.g. of the spiral wound, plate and frame, tubular, and hollow-fibre type arrangements (Mulder M. Basic principles of membrane technology. Kluwers Academic Publishers (1996) p. 564, and Baker R. W. Membrane technology and applications. New York: McGraw-Hill, (2000) p. 514.) The initial membrane module configurations were of a plate and frame type. This configuration was based on designs in the filtration industry, U.S. Pat. Nos. 3,473,668 and 3,209,915.
In one embodiment, the membrane stack is integrally moulded such that the spacers and the membranes are a single unit that are joined to a feed/distributor system and permeate collection system with flow around the stack being prevented. This type of configuration may take the form of e.g. a flat disk.
In another embodiment, fluid flow around the membrane stack is prevented through the use of a sealing device or sealing means such as pressure on the outside edges of the stack.
In another embodiment, the module comprises a membrane stack configured as a series of touching, concentric rings forming a tubular device. Flow may be from the centre of the tube through the walls of the tube to the outside or the reverse. This type of tubular device does not have to be a strict cylinder. Pleating or other types of folding may be introduced such that the surface area and membrane stack volume can be increased in the module.
The above types of arrangements are illustrated generically in
In another embodiment, where a significant number of layers is desired, the module may take the form of a tubular device such as spiral wound arrangement in which a membrane layer and a spacer layer are wound concentrically. The centre of the resulting tube thus forms one flow channel (feed or permeate), and the outside of the tube forms the second flow channel. In these tubular devices, the ends of the membrane stack may be sealed such that flow has to take place through the stack.
Uses of Membrane Stacks and Modules
The membrane stacks and modules of the invention may be used in all applications where filtration membranes or modules are used. These include, for instance, separation and/or reaction/separation applications. They may also be used for e.g. analytical devices such as multi-well plates or in large scale therapeutic material production.
In one embodiment, the membrane stacks are used for membrane chromatography applications, particularly in the large scale separation and production of bio-molecules. These include the isolation and purification of antibodies, therapeutic proteins, plasmids, etc. Disposable membrane stacks offer significant advantages over the more conventional resin bead chromatography in terms of performance, regulatory compliance and process economics.
In addition to chromatography stacks consisting of a single membrane type, the invention can also greatly improve the performance of multi-element stacks in which membranes of different properties are combined into a single stack. These multi-element stacks allow for a simplification of the bioprocess purification train coupled with enhanced recoveries of the target biomolecules. This may translate into significant process cost savings.
Macroporous Gel-Filled Membranes
As discussed above, the membrane stack of the invention can, in some embodiments, comprise macroporous gel-filled membranes. Macroporous gel-filled membranes can be defined as composite materials that comprise a support member that has a plurality of pores extending through the support member and, located in the pores of the support member and essentially filling the pores of the support member, a macroporous cross-linked gel. In some embodiments, the macroporous gel used is responsive to environmental conditions, providing a responsive membrane.
The macroporous gel fills the pores of the support laterally, i.e. substantially perpendicular to the direction of the flow through the membrane. By “fill” we mean that, in use, essentially all liquid that passes through the membrane must pass through the macroporous gel. A support member whose pores contain macroporous gel to such an amount that this condition is satisfied is regarded as filled. Provided that the condition is met that the liquid passes through the macroporous gel, it is not necessary that the void volume of the support member be completely occupied by the macroporous gel.
The porous support member, or host, may be hydrophilic or hydrophobic and can be, for example, in the form of a membrane, a chromatography bed, or a filtration bed. The support member provides the mechanical strength to support the macroporous gel. The macroporous gel provides a low resistance to hydraulic flow, enabling high flow rates to be achieved with low reductions in pressure across the macroporous gel-filled membrane. The macroporous gel also provides the separating function of the membrane in chromatographic and filtration applications.
A gel is a cross-linked polymer network swollen in a liquid medium. The swelling liquid prevents the polymer network from collapsing and the network, in turn, retains the liquid.
Gels are typically obtained by polymerization of a monomer and a polyfunctional compound (a cross-linker), or by cross-linking a cross-linkable polymer, in a solvent which is a good solvent for the formed polymer network and which swells the polymer network. The polymer chains in such a network can be assumed to be uniformly distributed throughout the whole volume of the network and the average distance between the chains, known as mesh size, is determined by the cross-linking density. As the concentration of the cross-linker is increased, the density of cross-links in the gel also increases, which leads to a smaller mesh size in the gel. The smaller mesh size results in a higher resistance to the flow of liquids through the gel. As the concentration of the cross-linker is increased further, the constituents of the gel begin to aggregate, which produces regions of high polymer density and regions of low polymer density in the gel. Such gels exhibit what has been called microheterogeneity. This aggregation normally causes the gel to display a higher permeability to liquids, as the flow of liquids takes place primarily through the areas in the gel that have a lower polymer density. The low density areas of the gels are defined as draining regions while the higher density aggregates are called non-draining regions. As the concentration of the cross-linker is increased even further, leading to more cross-links, the gel can develop regions in which there is essentially no polymer. These regions are referred to as “macropores” in the present specification.
It is possible to compare the hydrodynamic (Darcy) permeability of a particular membrane with a reference material. The reference material is obtained by filling the pores of a support member identical with that of the macroporous gel-filled membrane with a homogeneous gel of essentially the same chemical composition and the similar mass as the gel of the macroporous gel-filled membrane, that is a gel composed of the same monomers formed in a good solvent, but cross-linked only to such an extent that the gel remains homogeneous and aggregation into regions of high and low polymer density does not occur. Membranes comprising macroporous gels display hydrodynamic (Darcy) permeabilities that are at least one order of magnitude higher than those of the corresponding reference materials, and in some instances the permeabilities are more than two or even more than three orders of magnitude higher. In this specification, a macroporous gel-filled membrane whose hydrodynamic (Darcy) permeability is at least an order of magnitude greater than that of the corresponding reference material is said to have a permeability ratio greater than 10.
The permeability ratio is closely related to the size of the macropores in the macroporous gel-filled membrane. For size-exclusion separations such as ultrafiltration, the permeability ratio can be fairly close to 10. In other applications, for example adsorption, synthesis or cell growth, where larger macropores are used, the permeability ratio can reach, in some embodiments, values of 100 or greater, or even 1000 or greater. In some instances it is possible to calculate the hydrodynamic permeability of homogeneous gels, in accordance with the teachings of Mika A. M. and Childs R. F., Calculation of the hydrodynamic permeability of gels and gel-filled microporous membranes, Ind. Eng. Chem. Res., vol. 40 (2001), p. 1694-1705, incorporated herein by reference. This depends upon data for the particular gel polymer being available.
From the hydrodynamic permeability there can be derived the hydrodynamic radius, defined as the ratio of the pore volume to the pore wetted surface area. It can be calculated from the hydrodynamic (Darcy) permeability using the Carman-Kozeny equation as given, for example, in the book J. Happel and H. Brenner, Low Reynolds Numbers Hydrodynamics, Noordhof of Int. Publ., Leyden, 1973, p. 393, incorporated by reference herein. It is necessary to assume a value for the Kozeny constant and for the purpose of these calculations the inventors assume a value of 5. Macroporous gel-filled membranes are found to have a hydrodynamic radius more than three times as high as the hydrodynamic radius of the corresponding reference material.
From the definition of the hydrodynamic permeability it can be derived that two macroporous gel-filled membranes of the same thickness will have hydrodynamic fluxes at the same pressure that will have the same ratio as their permeability ratio.
The size of macropores in the gel can be within a broad range, from a few nanometers to several hundred nanometers. Preferably, the porous gel constituent of the macroporous gel-filled membrane has macropores of average size between about 10 and about 3000 nm, has volume porosity between 30 and 80% and a thickness equal to that of the porous support member. In some embodiments, the average size of the macropores is preferably between 25 and 1500 nm, more preferably between 50 and 1000 nm, and most preferably the average size of the macropores is about 700 nm.
In the absence of a support member, the macroporous gels may be non-self supporting, and they may change or even lose their porosity when dried. By inserting the macroporous gel within a porous support member, mechanical strength is conferred upon the macroporous gel. The utilization of macroporous gels creates a macroporous gel-filled membrane that permits larger molecules, such as biological molecules, to enter the macropores and the solution containing such molecules to traverse the gel at a high flux.
By a “responsive macroporous gel-filled membrane” is meant a macroporous gel-filled membrane which comprises a macroporous gel whose pore-size can be controlled by varying specific environmental conditions.
General Characteristics of the Macroporous Gel-Filled Membrane
Preferably, the macroporous gel is anchored within the support member. The term “anchored” is intended to mean that the gel is held within the pores of the support member, but the term is not necessarily restricted to mean that the gel is chemically bound to the pores of the support member. The gel can be held by the physical constraint imposed upon it by enmeshing and intertwining with structural elements of the host, without actually being chemically grafted to the host or support member, although in some embodiments, the macroporous gel may become grafted to the surface of the pores of the support member.
It will be appreciated that as the macropores are present in the gel that fills the pores of the support member, the macropores must be smaller than the pores of the support member. Consequently, the flow characteristics and separation characteristics of the macroporous gel-filled membrane are dependent on the characteristics of the macroporous gel, but are largely independent of the characteristics of the porous support member, with the proviso, of course, that the size of the pores present in the support member is greater than the size of the macropores of the gel. The porosity of the macroporous gel-filled membrane can be tailored by filling the support member with a gel whose porosity is primarily controlled by the nature and amounts of monomer or polymer, cross-linking agent, reaction solvent, and porogen, if used. As pores of the support member are filled with the same macroporous gel material, there is achieved a high degree of consistency in properties of the macroporous gel-filled membrane, and for a particular support member these properties are determined largely, if not entirely, by the properties of the macroporous gel. The net result is that the invention provides control over macropore-size, permeability and surface area of the macroporous gel-filled membranes.
The number of macropores in the macroporous gel-filled membrane is not dictated by the number of pores in the support material. The number of macropores in the macroporous gel-filled membrane can be much greater than the number of pores in the support member, although the macropores are smaller than the pores in the support member.
As mentioned above, the effect of the pore-size of the support material on the pore-size of the macroporous gel is generally quite negligible. An exception to this is found in those cases where the support member has a large difference in pore-size and pore-size distribution, and where a macroporous gel having very small pore-sizes and a narrow range in pore-size distribution is sought. In these cases, large variations in the pore-size distribution of the support member are weakly reflected in the pore-size distribution of the macroporous gel. As such it is preferable to use a support member with a somewhat narrow pore-size range in these situations.
The properties of the macroporous gel-filled membrane can be tuned, by adjusting the average pore diameter of the macroporous gel. For some purposes, for example ultrafiltration by means of size exclusion, small pores may be required. For other purposes, for example use as a solid support for a chemical synthesis involving fast-kinetics, large pores may be required. The size of the macropores is mainly dependent on the nature and concentration of the cross-linking agent, the nature or the solvent or solvents in which the gel is formed, the amount of any polymerization initiator or catalyst and, if present, the nature and concentration of porogen.
Generally, as the concentration of cross-linking agent is increased, the size of the macropores in the gel is also increased. For example, the molar ratio of polyfunctional compound(s)(cross-linking agent) to monomer(s) may be in the range of from about 5:95 to about 70:30, preferably in the range of from about 10:90 to about 50:50, and more preferably in the range of from about 15:85 to about 45:55.
The components of the macroporous gel are introduced into the pores of the support member by means of a liquid vehicle, and solvent selection for in situ polymerization or cross-linking plays a role in obtaining porous gels.
Generally, the solvent or solvent mixture should dissolve monomers and polyfunctional compounds, or cross-linkable polymers and cross-linking agents, over a wide range of concentrations. If the solvent is a good solvent for the gel polymer, porosity can only be introduced into the gel by cross-linking or porogen. If, however, there is present a solvent that is a thermodynamically poor solvent or non-solvent, this solvent will act as a porogen. By combining solvents of different affinities to the gel polymer, from a good solvent through a poor solvent to a non-solvent, at different ratios, both porosity and pore dimensions can be altered. In general, the poorer the solvent or the solvent mixture the higher the porosity and the sizes of macropores. Preferably, the solvent or solvent mixture for in situ polymerization contains poor solvent in the range from about 0% to about 100%, more preferably from about 10% to about 90%. Examples of good solvents for poly(2-acrylamido-2-methyl-1-propanesulfonic acid) are water and N,N-dimethylformamide. Examples of poor solvents include dioxane, hydrocarbons, esters, and ketones. An example of a good solvent for poly(acrylamide) is water. Examples of poor solvents include dioxane, alcohols such as methanol, N,N-dimethylformamide, hydrocarbons, esters, and ketones. Preferably, the solvents used are miscible with water.
When the polymerization is carried out using a liquid vehicle that contains non-solvents or poor solvents, the resulting structure is often built of clusters of agglomerated microspheres that form the body of the macroporous gel. The pores in such materials consist of the voids located between clusters (macropores), voids between microspheres in the clusters (mesopores), and pores inside the microspheres themselves (micropores).
Porogens can be broadly described as pore generating additives. Examples of porogens that can be used in the gel-forming reaction include thermodynamically poor solvents or extractable polymers, for example poly(ethyleneglycol), or surfactants, or salts. Porogens are known in the art, and a person skilled can determine, using standard experimental techniques and without exercise of any inventive faculty, which porogens are suitable to prepare macroporous gels for use in a desired membranes.
There is no simple way to predict accurately the structure parameters of porous gels obtained under given conditions, but qualitative rules are available to give some guidance. Generally, the mechanism of porous gel formation via polymerization of one or more monomers and cross-linkers involves, as a first step, an agglomeration of polymer chains to give nuclei. The polymerization continues both in the nuclei and in the remaining solution to form microspheres which grow in size by capturing newly precipitated nuclei and polymers from the solution. At some point, the microspheres become interconnected with each other in large clusters that form the body of the macroporous gel. The poorer the solvent quality the faster nucleation occurs during the gel-forming process. If the number of nuclei formed is very large, as in the case of high concentration of a polymerization initiator, smaller pores may be expected. If, however, the number of nuclei is smaller and the reaction kinetics is such that the nuclei can grow larger, large pores are formed in the gel. High concentration of a cross-linker usually causes early nucleation. The nuclei, however, may be too highly cross-linked to be able to swell with the monomers, grow and coalesce in clusters. This may result in very small pores. Because of the different ways that the polymerization may proceed and the polymerization conditions may affect the gel porous structure, a large variety of structures can be obtained but conditions for each of the structures need to be determined experimentally.
Composition of the Macroporous Gel-Filled Membranes
The macroporous gels can be formed through the in-situ reaction of one or more polymerisable monomers with one or more cross-linkers, or of one or more cross-linkable polymers with one or more cross-linker to form a cross-linked gel that has macropores of a suitable size. Suitable polymerisable monomers include monomers containing vinyl or acryl groups. For Donnan exclusion, there can be used vinyl or acryl monomers containing at least one polar and/or ionic functional group, or functional group that can be converted into ionic group. For biological affinity there can be used vinyl or acryl monomers containing at least one reactive functional group. Preferred polymerisable monomers include acrylamide, 2-acryloxyethyltrimethylammonium chloride, N-acryloxysuccinimide, N-acryloyltris(hydroxymethyl)methylamine, 2-aminoethyl methacrylate hydrochloride, N-(3-aminopropyl)methacrylamide hydrochloride, butyl acrylate and methacrylate, N,N-diethylacrylamide, N,N-dimethylacrylamide, 2-(N,N-dimethylamino)ethyl acrylate and methacrylate, N-[3-(N,N-dimethylamino)propyl]methacrylamide, N,N-dimethylacrylamide, n-dodecyl acrylate, n-dodecyl methacrylate, dodecyl methacrylamide, ethyl methacrylate, 2-(2-ethoxyethoxy)ethyl acrylate and methacrylate, 2,3-dihydroxypropyl acrylate and methacrylate, glycidyl acrylate and methacrylate, n-heptyl acrylate and methacrylate, 1-hexadecyl acrylate and methacrylate, 2-hydroxyethyl acrylate and methacrylate, N-(2-hydroxypropyl)methacrylamide, hydroxypropyl acrylate and methacrylate, methacrylamide, methacrylic anhydride, methacryloxyethyltrimethylammonium chloride, 2-(2-methoxy)ethyl acrylate and methacrylate, octadecyl acrylamide, octylacrylamide, octyl methacrylate, propyl acrylate and methacrylate, N-iso-propylacrylamide, stearyl acrylate, styrene, 4-vinylpyridine, vinylsulfonic acid, N-vinyl-2-pyrrodinone. Particularly preferred monomers include dimethyldiallylammonium chloride, acrylamido-2-methyl-1-propanesulfonic acid (AMPS), (3-acrylamidopropyl) trimethylammonium chloride (APTAC), acrylamide, methacrylic acid (MAA), acrylic acid (AA), 4-styrenesulfonic acid and its salts, acrylamide, glycidyl methacrylate, diallylamine, and diallylammonium chloride.
The crosslinker may be, for example, a compound containing at least two vinyl or acryl groups. Examples of crosslinkers include bisacrylamidoacetic acid, 2,2-bis[4-(2-acryloxyethoxy)phenyl]propane, 2,2-bis(4-methacryloxyphenyl)propane, butanediol diacrylate and dimethacrylate, 1,4-butanediol divinyl ether, 1,4-cyclohexanediol diacrylate and dimethacrylate, 1,10-dodecanediol diacrylate and dimethacrylate, 1,4-diacryloylpiperazine, diallylphthalate, 2,2-dimethylpropanediol diacrylate and dimethacrylate, dipentaerythritol pentaacrylate, dipropylene glycol diacrylate and dimethacrylate, N,N-dodecamethylenebisacrylamide, divinylbenzene, glycerol trimethacrylate, glycerol tris(acryloxypropyl) ether, N,N′-hexamethylenebisacrylamide, N,N′-octamethylenebisacrylamide, 1,5-pentanediol diacrylate and dimethacrylate, 1,3-phenylenediacrylate, poly(ethylene glycol) diacrylate and dimethacrylate, poly(propylene) diacrylate and dimethacrylate, triethylene glycol diacrylate and dimethacrylate, triethylene glycol divinyl ether, tripropylene glycol diacrylate or dimethacrylate, diallyl diglycol carbonate, poly(ethylene glycol) divinyl ether, N,N′-dimethacryloylpiperazine, divinyl glycol, ethylene glycol diacrylate, ethylene glycol dimethacrylate, N,N′-methylenebisacrylamide, 1,1,1-trimethylolethane trimethacrylate, 1,1,1-trimethylolpropane triacrylate, 1,1,1-trimethylolpropane trimethacrylate, vinyl acrylate, 1,6-hexanediol diacrylate and dimethacrylate, 1,3-butylene glycol diacrylate and dimethacrylate, alkoxylated cyclohexane dimethanol dicarylate, alkoxylated hexanediol diacrylate, alkoxylated neopentyl glycol diacrylate, aromatic dimethacrylate, caprolacone modified neopentylglycol hydroxypivalate diacrylate, cyclohexane dimethanol diacrylate and dimethacrylate, ethoxylated bisphenol diacrylate and dimethacrylate, neopentyl glycol diacrylate and dimethacrylate, ethoxylated trimethylolpropane triarylate, propoxylated trimethylolpropane triacrylate, propoxylated glyceryl triacrylate, pentaerythritol triacrylate, tris (2-hydroxy ethyl)isocyanurate triacrylate, di-trimethylolpropane tetraacrylate, dipentaerythritol pentaacrylate, ethoxylated pentaerythritol tetraacrylate, pentaacrylate ester, pentaerythritol tetraacrylate, and caprolactone modified dipentaerythritol hexaacrylate. Particularly preferred cross-linking agents include N,N′,-methylenebisacrylamide, diethylene glycol diacrylate and dimethacrylate, trimethylolpropane triacrylate, ethylene glycol diacrylate and dimethacrylate, tetra(ethylene glycol) diacrylate, 1,6-hexanediol diacrylate, divinylbenzene, poly(ethylene glycol) diacrylate.
The concentration of monomer in the macroporous gel can have an effect on the resiliency of the macroporous gel prepared. A low monomer concentration can lead to a macroporous gel that is non-self supporting. Such non-self supporting gels might be advantageous as adsorbents, as they could lead to gels having greater adsorption capacity. In some embodiments, the monomer concentration is 60% or less, for example about 60, 50, 40, 30, 20, 10 or 5%.
When a cross-linkable polymer is used, it can be dissolved and reacted in-situ in the support with a cross-linking agent to form the macroporous gel. Suitable cross-linkable polymers include poly(ethyleneimine), poly(4-vinylpyridine), poly(vinylbenzyl chloride), poly(diallylammonium chloride), poly(glycidyl methacrylate), poly(allylamine), copolymers of vinylpyridine and dimethyldiallylammonium chloride, copolymers of vinylpyridine, dimethyladiallylammonium chloride, or (3-acrylamidopropyl)trimethylammonium chloride with glycidyl acrylate or methacrylate, of which poly(ethyleneimine), poly(diallylammonium chloride), and poly(glycidyl methacrylate) are preferred. Use of cross-linkable polymers instead of monomers can, in some instances, require a decrease in the concentration of cross-linking agent. In order to retain the large size of the pores in the gel with a lower cross-linking agent concentration, a porogen can be added to the mixture used to prepare the macroporous gel.
The cross-linking agent for reaction with the cross-linkable polymer is selected from molecules containing two or more reactive groups that can react with an atom or group of atoms in the polymer to be cross-linked, such as epoxy groups or alkyl/aryl halides that can react with nitrogen atoms of polyamines, or amine groups that can react with alkyl/aryl halides or epoxy groups of glycidyl-group-containing polymers to be in situ cross-linked. Suitable cross-linkers include ethylene glycol diglycidyl ether, poly(propylene glycol) diglycidyl ether, 1,3-dibromopropane, 1,4-dibromobutane, 1,5-dibromopentane, 1,6-dibromohexane, α,α′-dibromo-p-xylene, α,α′-dichloro-p-xylene, 1,4-dibromo-2-butene, piperazine, 1,4-diazabicyclo[2.2.2]octane, 1,2-diaminoethane, 1,3-diaminopropane, 1,4-diaminobutane, 1,5-diaminopentane, 1,6-diaminohexane, 1,7-diaminoheptane, 1,8-diaminooctane.
It is also possible to modify polymers containing reactive groups such as an amino, hydroxyl, carboxylic acid, carboxylic acid ester, or epoxy groups with reagents to introduce vinyl groups that can be subsequently polymerized by treatment with a polymerization initiator to form a macroporous gel. Examples of suitable vinyl groups that can be introduced include vinylbenzene derivatives, allyl derivatives, acrolyl and methacrolyl derivatives. The cross-linking of these vinyl substituted polymers can in some instances be facilitated by the introduction of further monomers such as acrylamide, N-vinylpyrrolidone, acrylic and methacrylic acids and their salts.
Macromonomers can also be used as monomers or as cross-linking agents. Macromonomers can be polymers or oligomers that have one (monofunctional) or more (cross-linking agent) reactive groups, often at the ends, which enable them to act as a monomer or a cross-linker. For monomers, each macromonomer molecule is attached to the main chain of the final polymer by reaction of only one monomeric unit in the macromonomer molecule. Examples of macromonomers include poly(ethylene glycol)acrylate and poly(ethylene glycol)methacrylate, while examples of polyfunctional macromonomers include poly(ethylene glycol)diacrylate and poly(ethylene glycol)dimethacrylate. Macromonomers preferably have molecular weights of about 200 Da or greater.
Many macroporous gels can be prepared, including neutral hydrogels, charged hydrogels, polyelectrolyte gels, hydrophobic gels, and neutral and functional gels.
If the gel selected is a neutral hydrogel or a charged hydrogel for which water is the swelling liquid medium, the resulting supported macroporous gel is normally quite hydrophilic. Hydrophilic macroporous gel-filled membranes are preferred as they provide better flow characteristics and impart anti-fouling tendencies to the membranes. Examples of suitable hydrogels include cross-linked gels of poly(vinyl alcohol), poly(acrylamide), poly(isopropylacrylamide), poly(vinylpyrrolidone), poly(hydroxymethyl acrylate), poly(ethylene oxide), copolymers of acrylic acid or methacrylic acid with acrylamide, isopropylacrylamide, or vinylpyrrolidone, copolymers of acrylamide-2-methyl-1-propanesulfonic acid with acrylamide, isopropylacrylamide, or vinylpyrrolidone, copolymers of (3-acrylamidopropyl) trimethylammonium chloride with acrylamide, isopropylacrylamide, or N-vinylpyrrolidone, copolymers of diallyldimethylammonium chloride with acrylamide, isopropylacrylamide, or vinylpyrrolidone. Preferred hydrogels include cross-linked poly(vinyl alcohol), poly(acrylamide), poly(isopropylacrylamide) and pol(vinylpyrrolidone) and cross-linked copolymers of neutral monomers such as acrylamide or N-vinylpyrrolidone with charged monomers such as acrylamide-2-methyl-1-propanesulfonic acid or diallyldimethylammonium chloride.
The macroporous gels can be selected to comprise polyelectrolytes. Like the charged hydrogels, polyelectrolyte gels give hydrophilic macroporous gel-filled membranes, and they also carry a charge. The polyelectrolyte gel can be selected, for example, from cross-linked poly(acrylamido-2-methyl-1-propanesulfonic acid) and its salts, poly(acrylic acid) and its salts, poly(methacrylic acid) and its salts, poly(styrenesulfonic acid) and its salts, poly(vinylsulfonic acid) and its salts, poly(alginic acid) and its salts, poly[(3-acrylamidopropyl)trimethylammonium] salts, poly(diallyldimethylammonium) salts, poly(4-vinyl-N-methylpyridinium) salts, poly(vinylbenzyl-N-trimethylammonium) salts, poly(ethyleneimine) and its salts. Preferred charged gels include cross-linked poly(acrylic acid) and its salts, poly(methacrylic acid) and its salts, poly(acrylamido-2-methyl-1-propanesulfonic acid) and its salts, poly[(3-acrylamidopropyl)trimethylammonium] salts, poly(diallyldimethylammonium) salts, and poly(4-vinylpyridinium) salts.
One of the differences between charged gels and polyelectrolyte gels is that the repeating monomer in the polyelectrolyte gel bears a charge, while in the charged gel, the charge is found in a co-polymerized unit that is randomly distributed through the polymer. The monomer used to form the polyelectrolyte gel or the co-polymer in the charged gel that bears a charge usually has a charge bearing group, but it can also be a non-charge-bearing group that can become charged in a post-gelation process (e.g. quaternization of nitrogen bearing groups). Examples of polymers that can become charged include poly(4-vinylpyridine) which can be quaternized with various alkyl and alkylaryl halides. Suitable alkyl halides include those having up to 8 carbon atoms, for example methyl iodide, ethyl bromide, butyl bromide, and propyl bromide. Suitable alkylaryl halides include benzyl halides, especially benzyl chloride and benzyl bromide. Another polymer that can become charged is poly(vinylbenzyl chloride), which can be quaternized with various amines, for example, lower alkylamines or aromatic amines such as triethylamine, pyridine, azabicyclo[2.2.2]octane, N-methylpyrrolidine, and N-methylpiperidine, and lower hydroxyalkylamines, for example triethanolamine. Yet another polymer that can become charged is poly(glycidyl methacrylate) or poly(glycidyl acrylate), which can react with various amines, for example lower alkylamines such as diethylamine and triethylamine, azabicyclo[2.2.2]octane, N-methylpyrrolidine, and N-methylpiperidine. Alternatively, glycidyl moieties can be converted to sulfonic acid groups by reaction with, for example alkali metal sulfites such as sodium sulfite. A person skilled in the art will appreciate that there are other polymers that are, or can be rendered, charge-bearing.
The macroporous gel can be selected to comprise hydrophobic monomers to permit separations in organic solvents, for example hydrocarbons, especially liquid paraffins such as hexanes. Hydrophobic monomers, such as styrene and its derivatives, for example an alkyl substituted styrene derivative such as para-tertbutyl styrene, can be used to prepare hydrophobic macroporous gels. Copolymers of these monomers can be used.
A macroporous gel comprising hydrophobic monomers can be used to capture molecules from fluids passing through the pores by hydrophobic interactions.
As stated above, the macroporous gels can also be selected to comprise reactive functional groups that can be used to attach ligands or other specific binding sites. These functional macroporous gels can be prepared from cross-linked polymers bearing functional groups, for example epoxy, anhydride, azide, reactive halogen, or acid chloride groups, that can be used to attach the ligands or other specific binding sites. Examples include cross-linked poly(glycidyl methacrylate), poly(acrylamidoxime), poly(acrylic anhydride), poly(azelaic anhydride), poly(maleic anhydride), poly(hydrazide), poly(acryloyl chloride), poly(2-bromoethyl methacrylate), poly(vinyl methyl ketone). Functionality that can be introduced can take the form of antibodies or fragments of antibodies, or alternatively, chemical mimics such as dyes. Functional gels are attractive in biomolecule purifications or separations, as they can offer preferential binding to certain molecules by binding to active sites, while being non-reactive to other molecules, even when there is no significant difference in size between the molecules, examples being affinity ligands selected to bind with some proteins but not others. Affinity ligands that can be attached to porous gels via reactive groups include amino acid ligands such as L-phenylalanine, tryptophan, or L-histidine to separate γ-globulins and immunoglobulins, antigen and antibody ligands such as monoclonal antibodies, protein A, recombinant protein A, protein G, or recombinant protein G to separate immunoglobulins from different media, dye ligands such as cibaron blue or active red to separate albumins and various enzymes, metal affinity ligands such as complexes of iminodiacetic acid (IDA) ligand with Cu2+, Ni2+, Zn2+, or Co2+ to separate various proteins such as histidine, lysozyme, or tryptophan from various media.
Responsive Macroporous Gels
Polymers that change their conformation in response to changes in environmental conditions are known. By incorporating the properties of such polymers in the macroporous gel, a macroporous gel-filled membrane with dynamic pore-size is obtained. These membranes having responsive characteristics are substantially the same as the macroporous gel-filled membranes described above, except that at least one of the monomers or polymers that form the macroporous gel has a chemical structure which facilitates changes in pore-size.
The changes in the pore-size of the macroporous gel are due to the physical relationship between the support member and the macroporous gel. The macroporous gel-filled membrane can be described as having three distinct zones: (a) the support member, which ideally does not change shape, (b) the incorporated macroporous gel that “fills” the pores of the support member, and (c) the volume within the macropores of the gel, which volume is filled with water or a solvent and in which is found very little or no gel polymer. Under pressure, hydraulic flow occurs through the macropores of the gel, and the flux through the membrane is related to the number of pores in the macroporous gel, the radius of these pores, and the tortuosity of the path of the pores in the macroporous gel through the macroporous gel-filled membrane.
As the degree of swelling of the macroporous gel is changed by an environmental stimulus, the total volume occupied by the macroporous gel is constrained by the fixed total volume defined by the support member. As the overall volume of the macroporous gel is constrained by the support member, by necessity the volume fraction of the gel expands into the area defined by macropores in the gel. As the number of macropores and their tortuosity remain essentially constant with the change in volume fraction of the macroporous gel, the diameter or radius of the macropores themselves must change. If the macroporous or structured gel were unconfined, the environmentally induced changes would cause the total volume of the swollen gel to change. As such it would not follow in this unconfined case that the changes would result in a controllable change in pore-size of the macroporous gel.
The reason behind the change in volume of the macroporous gel is related to interactions between the polymer structures that form the gels, or the interactions between the polymer chains and the solvents or solutes present in the solvent that diffuse into the gel. The changes in the volume occupied by the gel are linked to the conformation adopted by the polymer chains that form the macroporous gels. The natural tendency of the polymer chains is to coil around themselves, which leads to a gel having a smaller volume. If the polymer chains within the gel can be manipulated to uncoil and form a more rigid backbone, the overall volume of the gel will increase. It is thus this coiling/uncoiling which is affected by the environmental stimuli that are applied to the responsive macroporous gel-filled membrane.
The volume changes of the pores can either be “continuous” or “discontinuous”. A continuous volume change takes place over a relatively large change in the triggering environmental condition and where there exists at least one stable volume near the transition between the swollen and collapsed state. Preferably, a continuous volume change will go through a multitude of stable transition volumes between the swollen and the collapsed state. A discontinuous volume change in gels is characterised by the reversible transition from swollen to collapsed state taking place over an extremely small change in the triggering environmental condition, for example, less than 0.1 pH unit or 0.1 degree Celsius. Gels exhibiting discontinuous volume change are called “phase-transition” gels and systems or devices with such gels are often called “chemical valves”. Preferably, the responsive macroporous gels according to this embodiment of the invention undergo a “continuous” volume change through discrete stable volumes that can be utilized to control the pore-size of the gel.
Of the environmental stimuli that can be used to change the pore-size in the responsive macroporous, mention is made of pH, specific ions, ionic strength, temperature, light, electric fields, and magnetic fields. The effect of each stimulus, and examples of monomers that react to such a stimulus, will be described in more detail below.
One stimulus that can be utilised to change the pore-size of responsive macroporous gel is the pH of the solution being passed through the pores of the gel. A change in the pH of the solution will affect the pore-size of the gel if the gel comprises weak acids or weak bases. In such cases, the natural tendency of the polymer chain within the gel to coil around itself will be balanced by the repulsion between the charged groups (weak acidic or basic groups) along the length of the polymer chain. Variations in the amount of charge along the chain cause large changes in conformation of the polymer chain, which in turn causes changes in the volume occupied by the gel. Changes in the pH of the solution are effective at controlling the amount of repulsion along the polymer chain, as they change the degree of ionisation of the charged groups. A gel comprising weak acid groups becomes less ionised as the pH is lowered and the gel contracts. Conversely, a weak base becomes more ionised as the pH is lowered and the chain elongates or stretches to give a swollen gel.
Examples of monomers that have weak acid functionality include acrylic acid, methacrylic acid, itaconic acid, 4-vinylbenzoic acid, bisacrylamidoacetic acid, and bis(2-methacryloxyethyl) phosphate. Examples of monomers that have weak base functionality include 2-aminoethyl methacrylate hydrochloride, N-(3-aminopropyl)methacrylamide hydrochloride, 2-(tert-butylamino)ethyl methacrylate, diallylamine, 2-(N,N-diethylamino)ethyl methacrylate, 2-(N,N-dimethylamino)ethyl acrylate, 2-(N,N-dimethylamino)ethyl methacrylate, 1-vinylimidazole, and 4-vinylpyridine. Glycidyl methacrylate derivatized hyaluronate-hydroxyethyl acrylate based hydrogels can also be used to prepare macroporous gel-filled membranes that are pH responsive [Inukai M., Jin Y., Yomota C., Yonese M., Chem. Pharm. Bull., (2000), 48:850-854; which is hereby incorporated by reference].
Variations in pH have little effect on the degree of ionisation of strong acids and bases, and as such, only drastic variations in pH can effect pore-size changes in gels comprising these functionalities.
Another stimulus that can be utilised for changing the pore-size of a responsive macroporous gel is the salt concentration of the solution being passed through the pores of the gel. Similarly to variations in pH, variations in salt concentration will effect pore-size variations in macroporous gels that comprise weak acidic or weak basic groups. The reason for the changes in pore-size, however, does differ slightly. The addition of an ionic solute has the ability to shield the charged groups found on the polymer chain in the gel by the formation of ion-pairs. This lessens the coulombic repulsion between the adjacent charged groups, which allows the chain to relax into a coiled conformation. An increase in salt concentration will shield both a weak acid group and a weak base group. Therefore, when the salt concentration is increased, for example by adding a concentrated salt solution to the bulk solution being passed through the macroporous gel-filled membrane, the shielding effect of the additional ions leads to an increase in pore size. Alternatively, a decrease in salt concentration, such as obtained by diluting the bulk solution being passed through the macroporous gel-filled membrane, will lead to less shielding and a smaller pore size.
Changes in salt concentration can also be used with macroporous gels that comprise strong acid groups and strong basic groups, as these groups are also shielded by the presence of free ionic species.
Examples of monomers that bear weak acid or base groups are listed above. Examples of monomers that have strong acid functionality include 2-acrylamido-2-methyl-1-propanesulfonic acid, sodium 2-methyl-2-propene-1-sulfonate, sodium styrenesulfonate, and sodium vinylsulfonate. Examples of monomers that have strong basic functionality include 2-acryloxyethyltrimethylammonium chloride, diallyldimethylammonium chloride, methacrylamidopropyltrimethylammonium chloride, and 3-methacryloxy-2-hydroxypropyltrimethylammonium chloride.
Ionic functionality (weak/strong acids and bases) may also be introduced in macroporous gels that do not originally bear charged functionalities but that bear instead reactive groups that can be converted into ionic or ionisable moieties in a post-polymerization treatment. Suitable monomers with reactive groups include acrylic anhydride, allyl glycidyl ether, bromostyrene, chloromethylstyrene, chlorostyrenes, glycidyl acrylate, glycidyl methacrylate, 4-hydroxybutyl methacrylate, 2-hydroxyethyl acrylate, methacryloyl chloride. For example, a macroporous gel comprising a glycidyl acrylate or methacrylate group can be treated with diethylamine to introduce weak base functionality or with sodium sulfite in an iso-propanol/water mixture to introduce strong acid (sulfonic acid) functionality.
Another stimulus that can be used to change the pore-size of a responsive macroporous gel is the temperature of the gel. Various methods are available for changing the temperature of the macroporous gel, one of which includes changing the temperature of a liquid flowing through the pores of the macroporous gel. While the change in overall gel volume for temperature dependent gels is again due to the control of the coiling or uncoiling of the polymer chains that form the gel, the contraction or expansion of the gel is not linked to the presence of charged groups on the polymer chain. For temperature dependant gels, the amount solvation of the polymer chain controls the conformation of the polymer chain. At lower temperatures the chains are solvated, which allows an elongated conformation of the polymer chain. As the temperature is increased, an entropic desolvation takes place causing the chains to coil and contract. Therefore, increases in temperature lead to larger pore sizes in the gel while decreases in temperature lead to smaller pore sizes.
Macroporous gels that comprise hydrophobic monomers are most suitable for use in temperature dependant systems, as salvation effects are markedly observed for polymers that have hydrophobic functionality. Examples of monomers that have hydrophobic functionality include N-(propyl)acrylamide, N-(tert-butyl)acrylamide, butyl acrylates, decyl acrylates, decyl methacrylates, 2-ethylbutyl methacrylate, n-heptyl acrylate, n-heptyl methacrylate, 1-hexadecyl acrylate, 1-hexadecyl methacrylate, n-hexyl acrylate, n-hexyl methacrylate, and N-(n-octadecyl)acrylamide. Gels displaying thermal response can also be prepared from sulphated hyaluronic acid-based gels (see Barbucci R., Rappuoli R., Borzacchiello A., Ambrosio L., J. Biomater. Sci.-Polym. Ed., (2000), 11:383-399), incorporated herein by reference.
Light is another stimulus that can be used to change the pore-size of the responsive macroporous gel. Light induced changes are due to photoisomerizations in the backbone or the side-chains of the polymer chains that form the gel. These photoisomerizations cause a change in either the conformation and local dipole moment, or in the degree of ionisation through light induced electron transfer reactions. One type of monomer that is suitable for use in light controlled systems comprises unsaturated functionalities than can undergo a trans-cis isomerization on irradiation. Examples of photoresponsive monomers that go through cis-trans conformation and dipole changes include 4-(4-oxy-4′-cyanoazobenzene)but-1-yl methacrylate, 6-(4-oxy-4′-cyanoazobenzene)hex-1-yl methacrylate, 8-(4-oxy-4′-cyanoazobenzene)oct-1-yl methacrylate, 4-[ω-methacryloyloxyoligo(ethyleneglycol)]-4′-cyanoazobenzene, 4-methacryloyloxy-4′-{2-cyano-3-oxy-3-[ω-methoxyoligo(ethyleneglycol)]prop-1-en-1-yl}azobenzene, and methacrylate monomers containing a mesogenic group and a photochromic para-nitroazobenzene group. It is also possible to incorporate the photoresponsive moeity in the crosslinker instead of the monomer. Examples of photoresponsive crosslinkers that go through cis-trans conformation and dipole changes include 4,4′-Divinylazobenzene, N, N′-bis(β-styrylsulfonyl)-4,4′-diaminoazobenzene, 4,4′-bis(methacryloylamino)azobenzene, 4,4′-dimethacryloylazobenzene, and bis((methacryloyloxy)methyl)spirobenzopyran.
The pore-size of the gel can also be altered by subjecting the macroporous gel to an electric field or to an electrical current. The response of the gel to electrochemical current changes is closely related to the pH systems described above. This close relationship is due to the fact that the passage of an electrochemical current through an aqueous system causes a “water splitting” reaction, which reaction leads to changes in the pH of the aqueous system. Electrical current can be passed through a macroporous gel-filled membrane e.g. by placing an electrode at either end of the macroporous gel-filled membrane. When current differential is applied to the electrodes, water molecules will separate and concentrations of H+ and HO− will increase at their respective electrodes. As described earlier, changes in pH can be used to control the pore-size of macroporous gels that comprise weak acid or weak base functionalities, which control is linked to the relationship between the ionisation of these functionalities and the coiling/uncoiling of the polymer chains that form the gel. Examples of weak acidic and weak basic monomers are given above.
Changes in gel volume due to fluctuations in an electrical field have been previously observed, such as in Murdan S., J. Control. Release, (2003), 92:1-17; and in Jensen M., Hansen P. B., Murdan S., Frokjaer S., Florence A. T., Eur. J. Pharm. Sci., (2002), 15:139-148, which are hereby incorporated by reference. While the exact process through which the gel volume is changed by the application of an electrical field is not yet well defined, the volume change itself is well documented. Chondroitin 4-sulphate (CS) is an example of a monomer that is responsive to electrical field fluctuations.
In some embodiments, the various stimuli response systems can be combined to offer gels that respond to more than one stimulus. An example of such a combined system can be prepared by combining a charged polymer (weak/strong acid or base) with a hydrophobic monomer. The macroporous gel resulting from such a combination will display responses to changes in salt concentration, changes in solution pH (when weak acids or bases are used), and changes in temperature. When combining different monomers, it is possible that the responsiveness of the gel to a single of the stimuli will be diminished, as the concentration of the monomer that responds to that particular stimulus will be lowered in the gel.
The magnitude of the response expressed by the macroporous gels, when various stimuli are applied to the gel, depends many different variables, a few of which are discussed below:
The responsiveness of the macroporous gel is dependent on the concentration of the crosslinking agent. Generally, as the concentration of cross-linking agent is increased, the size of the macropores in the responsive gel is also increased, but the range of pore-size changes is decreased. This relationship is fairly straightforward, as a higher concentration of crosslinks within the gel will limit the amount of coiling and uncoiling that will be available to the responsive gel. The molar ratio of crosslinking agent(s) to monomer(s) may be in the range of from about 5:95 to about 40:60, preferably in the range of from about 7.5:92.5 to about 10:90, and more preferably in the range of from about 10:90 to about 25:75.
Certain stimuli naturally evoke a broader range of response in the gel, as they more effectively affect the conformation of the polymer chains that form the gel. For example, variations in pH or temperature evoke a strong response from the appropriate macroporous gels, while changes in salt concentrations and light intensity evoke a slightly smaller response.
The concentration of the responsive monomer in the gel also affects the level of response demonstrated by the gel. Preferably, the responsive macroporous gels are composed of one or more responsive monomers and of one or more neutral monomers. The presence of a neutral monomer is important in those systems that have a very strong response to changes in the environmental conditions, as such systems often display discontinuous responses in pore-size (valve-effects). Addition of a neutral monomer attenuates the response, permitting a more controlled change in pore-size. Preferably, the molar ratio of the neutral monomers to the molar ratio of responsive monomers in the responsive macroporous gel is in the range from 5:95 to 95:5, more preferably in the range from 25:75 to 75:25, and more preferably in the range from 40:60 to 60:40. Suitable neutral monomers include acrylamide, N-acryloylmorpholine, N-acryloxysuccinimide, 2-acrylamido-2-(hydroxymethyl)-1,3-propanediol, N,N-diethylacrylamide, N, N-dimethylacrylamide, 2-(2-ethoxyethoxy)ethyl acrylate, 2-ethoxyethyl methacrylate, 2,3-dihydroxypropyl methacrylate, 2-hydroxyethyl methacrylate, N-(2-hydroxypropyl)methacrylamide, hydroxypropyl methacrylate, methacrylamide, N-[tris (hydroxymethyl)methyl]-1-methacrylamide, N-methylmethacrylamide, N-methyl-N-vinylacetamide, poly(ethylene glycol) monomethacrylate, N-iso-propylacrylamide, N-vinyl-2-pyrrolidone.
Porous Support Member
A variety of materials can be used to form the support member; however, apart from materials such as cellulose and some of its derivatives, most of these materials are strongly or relatively hydrophobic. Hydrophobic filtration membranes are not usually desired for use with aqueous systems, as they can lead to higher membrane fouling tendencies. The more inert and cheaper polymers such as polyolefins, for example (poly(ethylene), poly(propylene) poly(vinylidene difluoride)) can be used to make microporous membranes, but these materials are very hydrophobic. In some embodiments, the hydrophobicity of the support member does not affect the degree of fouling experienced by the macroporous gel-filled membrane as the flow of liquid through the macroporous gel-filled membrane takes place primarily in the macropores of the gel.
In some embodiments, the porous support member is made of polymeric material and contains pores of average size between about 0.1 and about 25 μm, and a volume porosity between 40 and 90%. Many porous substrates or membranes can be used as the support member but the support is preferably a polymeric material, and it is more preferably a polyolefin, which, while hydrophobic, is available at low cost. Extended polyolefin membranes made by thermally induced phase separation (TIPS), or non-solvent induced phase separation are mentioned. Hydrophilic supports can also be used, including natural polymers such as cellulose and its derivatives. Examples of suitable supports include SUPOR® polyethersulfone membranes manufactured by Pall Corporation, Cole-Parmer® Teflon® membranes, Cole-Parmer® nylon membranes, cellulose ester membranes manufactured by Gelman Sciences, Whatman® filter and papers.
In some other embodiments the porous support is composed of woven or non-woven fibrous material, for example a polyolefin such as polypropylene. An example of a polypropylene non-woven material is commercially available as TR2611A from Hollingsworth and Vose Company. Such fibrous woven or non-woven support members can have pore sizes larger than the TIPS support members, in some instances up to about 75 μm. The larger pores in the support member permit formation of macroporous gel-filled membranes having larger macropores in the macroporous gel. Macroporous gel-filled membrane with larger macropores can be used, for example, as supports on which cell growth can be carried out. Non-polymeric support members can also be used, such as ceramic-based supports. The porous support member can take various shapes and sizes.
In some embodiments, the support member is in the form of a membrane that has a thickness of from about 10 to about 2000 μm, more preferably from 10 to 1000μm, and most preferably from 10 to 500 μm. In other embodiments, multiple porous support units can be combined, for example, by stacking. In one embodiment, a stack of porous support membranes, for example from 2 to 10 membranes, can be assembled before the macroporous gel is formed within the void of the porous support. In another embodiment, single support member units are used to form macroporous gel-filled membrane, which are then stacked before use.
Preparation of Macroporous Gel-Filled Membranes
The macroporous gel-filled membranes can be prepared by simple, single step methods. These methods can, in some instances, use water or other benign solvents, such as methanol, as the reaction solvent. The methods also have the benefit of using rapid processes that lead to easier and continuous manufacturing possibilities. The macroporous gel-filled membrane is also potentially cheap.
The macroporous gel-filled membranes can be prepared, for example, by mixing one or more monomers, one or more polymers, or mixtures thereof, one or more cross-linking agents, optionally one or more initiators and optionally one or more porogens, in one or more suitable solvents. The solution produced is preferably homogeneous, but a slightly heterogeneous solution can be used. The mixture is then introduced into a suitable porous support, where a gel forming reaction takes place. Suitable solvents for the gel forming reaction include, for example, water, dioxane, dimethylsulfoxide (DMSO), dimethylformamide (DMF), acetone, ethanol, N-methylpyrrolidone (NMP), tetrahydrofuran (THF), ethyl acetate, acetonitrile, toluene, xylenes, hexane, N-methylacetamide, propanol, and methanol. It is preferable to use solvents that have a higher boiling point, as these solvents reduce flammability and facilitate manufacture. It is also preferable that the solvents have a low toxicity, and they can be readily disposed of after use. An example of such a solvent is dipropyleneglycol monomethyl ether (DPM).
In some embodiments, it is possible to use dibasic esters (esters of a mixture of dibasic acids) as the solvent. Dibasic esters (DBEs) are especially suitable for preparing gels based on polyacrylamide monomers. This solvent system has an unexpected characteristic in that it is poorly soluble in water, which differs from the other solvents used which are essentially completely water miscible. While water miscible solvents offer advantages in terms of solvent removal after fabrication, water immiscible solvents such as DBE's are good replacements, in certain cases, for solvents such as dioxane that are volatile, flammable, and toxic.
In some embodiments, components of the gel forming reaction react spontaneously at room temperature to form the macroporous gel. In other embodiments, the gel forming reaction must be initiated. The gel forming reaction can be initiated by any known method, for example through thermal activation or U.V. irradiation. The reaction is more preferably initiated by U.V. irradiation in the presence of a photoinitiator, as this method has been found to produce larger macropores in the gel, and it accelerates the gel forming reaction more than the thermal activation method. Many suitable photoinitiators can be used, of which 2-hydroxy-1[4-2(hydroxyethoxy)phenyl]-2-methyl-1-propanone (Irgacure 2959*), and 2,2-dimethoxy-2-phenylacetophenone (DMPA) are preferred. Other suitable photoinitiators include benzophenone, benzoin and benzoin ethers such as benzoin ethyl ether and benzoin methyl ether, dialkoxyacetophenones, hydroxyalkylphenones, α-hydroxymethyl benzoin sulfonic esters. Thermal activation requires the addition of a thermal initiator. Suitable thermal initiators include 1,1′-azobis(cyclohexanecarbonitrile) (VAZO® catalyst 88), azobis(isobutyronitrile) (AIBN), potassium persulfate, ammonium persulfate, and benzoyl peroxide.
If the reaction is to be initiated by U.V. irradiation, a photoinititor is added to the reactants of the gel forming reaction, and the support member containing the mixture of monomer, cross-linking agent and photoinitiator is subjected to U.V. irradiation at wavelengths of from 250 nm to 400 nm, for a period of a few seconds to a few hours. With certain photoinitiators, visible wavelength light may be used to inititate the polymerization. To permit the initiation, the support material must have a low absorbance at the wavelength used, to permit transmittance of the UV rays through the support. Preferably, the support and macroporous gel reagents are irradiated at 350 nm for a few seconds to up to 2 hours.
Preferably, thermally initiated polymerization is carried out at 60-80° C. for a few minutes up to 16 hours.
The rate at which polymerization is carried out has an effect on the size of the macropores obtained in the macroporous gel. As discussed earlier, when the concentration of cross-linker in a gel is increased to sufficient concentration, the constituents of the gel begin to aggregate to produce regions of high polymer density and regions with little or no polymer, which latter regions are referred to as “macropores” in the present specification. It is this mechanism which is affected by the rate of polymerization. When polymerization is carried out slowly, such as when a low light intensity in the photopolymerization, the aggregation of the gel constituents has more time to take place, which leads to larger pores in the gel. Alternatively, when the polymerization is carried out at a high rate, such as when a high intensity light source is used, there is less time available for aggregation and smaller pores are produced.
Once the macroporous gel-filled membranes are prepared, they can be washed with various solvents to remove any unreacted components and any polymer or oligomers that are not anchored within the support. Solvents suitable for the washing of the macroporous gel-filled membrane include water, acetone, methanol, ethanol, and DMF.
It must be noted that as used in this specification and the appended claims, the singular forms “a,” “an,” and “the” include plural reference unless the context clearly dictates otherwise. Unless defined otherwise all technical and scientific terms used herein have the same meaning as commonly understood to one of ordinary skill in the art to which this invention belongs.
The following examples are provided to illustrate the invention. It will be understood, however, that the specific details given in each example have been selected for purpose of illustration and are not to be construed as limiting the scope of the invention. Generally, the experiments were conducted under similar conditions unless noted.
The monomers used were acrylamide (AAM), 2-acrylamido-2-methyl-1-propanesulfonic acid (AMPS), (3-acrylamidopropane)trimethylammonium chloride (APTAC), diallyldimethylammonium chloride (DADMAC), ethylene glycol dimethacrylatecrylate (EDMA), glycidyl methacrylate (GMA), N,N′-methylenebisacrylamide (BIS), methacrylic acid (MAA), acrylic acid (AA), and trimethylolpropane triacrylate (TRIM). The polymers used were branched poly(ethylene imine) (BPEI) of an average molecular weight (MW) of 25000 Da, poly(ethylene glycol) (PEG) of average molecular weight of 200, 1000, 2000, 4000 and 10000 Da, and poly(allylammonium hydrochloride) (PAH) of an average molecular weight of 60000 Da. The cross-linker used for BPEI was ethylene glycol diglycidyl ether (EDGE).
The solvents used were di(propylene glycol) methyl ether, 97%, mixture of isomers (DPM) (Aldrich), cyclohexanol (CHX), methylene chloride (CH2Cl2), deionized water, 1,4-dioxane, N,N-dimethylformamide (DMF), dodecanol (DDC), glycerol, methanol, 1-octanol, and 1-propanol.
The free radical polymerization initiators used were 2-hydroxy-1-[4-(2-hydoxyethoxy)phenyl]2-hydroxy-2-methyl-1-propane-1-one (Irgacure® 2959), 2,2-dimethoxy-2-phenylacetophenone (DMPA), and 1,1′-azobis(cyclohexanecarbonitrile) (VAZO® catalyst 88).
The mesh-spacers used were Naltex (Austin, Tex., USA) symmetrical filtration netting made of polypropylene of thickness of 500 μm and mesh opening of 1000 μm and a polyester mesh (Polyester Monofilament)of thickness of 350 μm and mesh opening of 500 μm obtained from Great Lakes Filters Inc. Hillsdale Mich.
Proteins used were bovine serum albumin (BSA), lysozyme, human serum albumin (HSA) and human immunoglobulin (HIgG).
Other chemicals used were acryloyl chloride, hydrochloric acid, sodium azide, sodium chloride, sodium hydroxide, triethylamine, tris(hydroxymethyl)aminomethane (TRIS), 4-morpholineethanesulfonic acid (MES) and buffers (Tris Buffer).
The porous supports used were poly(propylene) thermally induced phase separation (TIPS) membranes PP1545-4 with an average pore diameter of 0.45 μm, thickness of 125 μm, and porosity of 85 vol-%, produced by 3M Company, and PP1183-3X of an average pore diameter of 0.9 μm, thickness of 87 μm, and porosity of 84 vol-%, both produced by 3M Company, and non-woven meltblown poly(propylene) TR2611A of a mean pore flow diameter of 6.5 μm, thickness of 250 μm, and porosity of 89.5 vol-% produced by Hollingworth & Vose Company.
Preparation of Macroporous Gel-Filled Membranes
The macroporous gel-filled membranes can be prepared according to the following general procedure. A weighed support member was placed on a poly(ethylene terephthalate) (PET) or poly(ethylene) (PE)sheet and a monomer or polymer solution was applied the sample. The sample was subsequently covered with another PET or PE sheet and a rubber roller was run over the sandwich to remove excess solution. In situ gel formation in the sample was induced by polymerization initiated by irradiation with the wavelength of 350 nm for the period of 10 to 120 minutes or by heating the sandwich at 60-80° C. for 2 hours. The irradiation was typically carried out using a system containing four 12″ long lamps, approx. 1.5″ spaced and emitting light at 365 nm with the output energy of approx. 0.1 Watt/inch. The system was equipped with a small fan to dissipate the heat (no other temperature control). The irradiated sample was located at approx. 5″ distance from the lamps. In case when preformed polymer and in situ cross-linking was used to form the gel, the sandwich was left at room temperature until the cross-linking reaction was completed, typically for 2-16 hours. The resulting macroporous gel-filled membrane was thoroughly washed with a suitable solvent or a sequence of solvents and stored in a 0.1 wt-% aqueous solution of sodium azide to prevent bacterial growth. In order to determine the amount of gel formed in the support, the sample was dried in vacuum at room temperature to a constant mass. The mass gain due to gel incorporation was calculated as a ratio of an add on mass of the dry gel to the initial mass of the porous support.
Flux Measurements
Water flux measurements through the macroporous gel-filled membranes were carried out after the samples had been washed with water. As a standard procedure, a sample in the form of a disk of diameter 7.8 cm was mounted on a sintered grid of 3-5 mm thickness and assembled into a cell supplied with compressed nitrogen at a controlled pressure. The cell was filled with deionized water or another feed solution and a desired pressure was applied. The water that passed through the macroporous gel-filled membrane in a specified time was collected in a pre-weighed container and weighed. All experiments were carried out at room temperature and at atmospheric pressure at the permeate outlet. Each measurement was repeated three or more times to achieve a reproducibility of ±5%.
The water flux, QH20 (kg/m2h), was calculated from the following relationship:
where m1 is the mass of container with the water sample, m2 is the mass of container, A is the active membrane surface area (38.5 cm2) and t is the time.
The macroporous gel-filled membrane may have water flux values that are smaller than those of the unfilled support member, with possible flux reduction of about a factor of two to about of a factor of a few hundred depending on the application. For ultrafiltration application, the flux may be reduced by a factor of about ten to about a few hundred.
The hydrodynamic Darcy permeability, k (m2) of the membrane was calculated from the following equation
where η is the water viscosity (Pa·s), δ is the membrane thickness (m), dH20 is the water density (kg/m3), and ΔP (Pa) is the pressure difference at which the flux, QH20, was measured.
The hydrodynamic Darcy permeability of the membrane was used to estimate an average hydrodynamic radius of the pores in the porous gel. The hydrodynamic radius, rh, is defined as the ratio of the pore volume to the pore wetted surface area and can be obtained from the Carman-Kozeny equation given in the book by J. Happel and H. Brenner, Low Reynolds Number Hydrodynamics, Noordhof Int. Publ., Leyden, 1973, p. 393:
where K is the Kozeny constant and ε is the membrane porosity. The Kozeny constant K≈5 for porosity 0.5<ε<0.7. The porosity of the membrane was estimated from porosity of the support by subtracting the volume of the gel polymer.
For membrane stacks, similar flux measurements were made. As a standard procedure, a sample in the form of a disk of diameter 4.4 cm was mounted on a sintered grid of 3-5 mm thickness and assembled into a cell supplied with compressed nitrogen at a controlled pressure. The sample consisted of 1 to 4 membrane layers. In some cases layers were separated by the mesh-spacer.
Protein Adsorption/Desorption Experiment
Protein adsorption experiments were carried out with two proteins, namely, bovine serum albumin (BSA) and lysozyme. In the case of experiments with a positively charged macroporous gel-filled membrane, the membrane sample was first washed with distilled water and subsequently with a TRIS-buffer solution (pH=7.8). In an adsorption step, a macroporous gel-filled membrane sample in a form of a single membrane disk of diameter 7.8 cm was mounted on a sintered grid of 3-5 mm thickness in a cell used for water flux measurements and described above. A BSA solution, comprising from 0.4 to 0.5 mg BSA per ml of buffer solution, was poured to the cell to give a 5 cm head over the macroporous gel-filled membrane. This hydrostatic pressure of 5 cm was kept constant by further additions of the BSA solution. In a modification of this method, the cell was pressurised with compressed nitrogen. The flow rate was measured by weighing the amount of permeate as a function of time. Typical values varied between 1 and 5 ml/min. Permeate samples were collected at 2-5 min intervals and analyzed by UV analysis at 280 nm. Following the adsorption step, the macroporous gel-filled membrane in the cell was washed with about 200 ml of the TRIS-buffer solution, and desorption was carried out with a TRIS-buffer solution containing 1M NaCl at 5 cm head pressure or under a controlled pressure of compressed nitrogen. The permeate samples were collected at 2-5 min intervals and tested by UV analysis at 280 nm for BSA content.
For negatively charged macroporous gel-filled membranes, a solution of lysozyme in a MES buffer solution having a pH of 5.5 and a lysozyme concentration of 0.5 g/L was used in a procedure similar to that described above for BSA and positively charged materials. The flow rate during the protein adsorption was again kept within 1-5 ml/min. Prior to the desorption of the protein, the membrane was washed by passing with 200 ml of the buffer solution. The desorption of the protein was carried out using a MES buffer solution (pH=5.5) containing 1 M NaCl in the same way as described above for the desorption of BSA. The lysozyme content in the collected samples was determined by UW spectrophotometry at 280 nm.
In other examples, protein adsorption tests involve stacks of several membranes of diameter of 19 mm mounted into a Mustang® Coin Device manufactured by Pall Corporation and the protein solution was delivered to the membrane stack at controlled flow rate using a peristaltic pump. The permeate fractions were collected and analyzed in the same way as described above. The desorption of the proteins was carried in a similar way as described above, with buffered 1M NaCl delivered to the membrane stack by using the pump instead of gravity or compressed nitrogen pressure.
Protein Separation Experiment
The experimental method used to examine the separation properties of the responsive macroporous gel-filled membranes in protein-protein fractionation processes is based on the pulsed injection ultrafiltration technique and its derivatives developed by Ghosh and his co-workers and described in the following articles: R. Ghosh and Z. F. Cui, Analysis of protein transport and polarization through membranes using pulsed sample injection technique, Journal of Membrane Science, vol. 175, no. 1 (2000) p. 75-84; R. Ghosh, Fractionation of biological macromolecules using carrier phase ultrafiltration, Biotechnology and Bioengineering, vol. 74, no. 1 (2001) p. 1-11; and R. Ghosh, Y. Wan, Z. F. Cui and G. Hale, Parameter scanning ultrafiltration: rapid optimisation of protein separation, Biotechnology and Bioengineering, vol. 81 (2003) p. 673-682 and incorporated herein by reference. The experimental set-up used was similar to that used for parameter scanning ultrafiltration as described in article by R. Ghosh, Y. Wan, Z. F. Cui and G. Hale, Parameter scanning ultrafiltration: rapid optimisation of protein separation, Biotechnology and Bioengineering, vol. 81 (2003) p. 673-682.
A binary carrier phase system was used in the ultrafiltration experiments. The starting carrier phase in all the responsive membrane experiments was one with a low salt concentration (typically 5-10 mM NaCl). In all these experiments the carried phase was switched to one with a high salt concentration (typically 1 M NaCl). The change in salt concentration within the membrane module could be tracked by observing the conductivity of the permeate stream. The change in transmembrane pressure gave an idea about the change in membrane hydraulic permeability with change in salt concentration.
Protein adsorption experiments for the membrane stacks were carried out with bovine serum albumin (BSA). In an adsorption step, a 4-layer macroporous gel-filled membrane sample in a form of a single membrane disk of diameter 4.4 cm was mounted on a sintered grid of 3-5 mm thickness in a cell used for water flux measurements and described above. A BSA solution, comprising of 0.5 mg BSA per ml of 25 mM TRIS buffer solution was poured to the cell and pressure of 35-50 kPa was applied. The flow rate was measured by weighing the amount of permeate as a function of time. Typical values varied between 5 and 7 ml/min. Permeate samples were collected at 2.5-3.0 min intervals and analysed by UV analysis at 280 nm.
This example illustrates effect of pressure on the water flux through the 4-layers of macroporous gel-filled membrane with and without mesh-spacer.
A macroporous gel-filled membrane was prepared according to the following procedure:
A solution containing 2.14 g of (3-acrylamidopropyl)-trimethylammonium chloride (APTAC)monomer as a 75% aqueous solution, 0.06 g of N,N′-methylenebisacrylamide (BIS) as a cross-linker, and 0.02 g of Irgacure ®2959 as a photoinitiator dissolved in 8.75 g of di(propylene glycol)methylether (DPM) was prepared.
A macroporous gel-filled membrane was prepared from the solution and the support TR2611A using the photoinitiated polymerization according to the general procedure described above. The irradiation time used was 5 minutes at 350 nm. The macroporous gel-filled membrane was removed from between the polyethylene sheets, washed with water and TRIS-buffer solution.
Several samples similar to that described above were prepared and averaged to estimate the mass gain of the macroporous gel-filled membrane. The substrate gained 98.5% of the original weight in this treatment.
The single layer macroporous gel-filled membrane produced by this method had a water flux in the range of 1890 kg/m2 hr at 100 kPa. A 4-layers membrane stack with and without spacers was tested for the water flux through it.
As can be seen from
This example illustrates effect of increasing number of layers in a membrane stack on the water flux through the macroporous gel-filled membrane with and without mesh-spacer.
A macroporous gel-filled membrane was prepared and tested for the water flux according to procedures as described above.
These data again show that we are achieving a substantial improvement in flux by using the interleaving approach.
This example illustrates effect of mesh-spacer on the sharpness breakthrough BSA adsorption curve for 4-layers Q-type macroporous gel-filled membrane.
A macroporous gel-filled membrane was prepared according to procedure as described above.
The protein (BSA) adsorption characteristic of the macroporous gel-filled membrane was examined using the general procedure for a multi-layer membrane disk as described above. The concentration of the protein used in this experiment was 0.5 g/L in 25 mM TRIS-buffer. The flow rate was 5-7 ml/min. A plot of the concentration of BSA in the permeate vs. the permeate volume is shown in
It could be expected that insertion of the interleaving layers could lead to a loss of resolution of the stacks in a separation. Surprisingly this is not the case. As is shown in
This example illustrates the formation of an unsupported porous gel, which can be used as the macroporous gel to prepare the macroporous gel-filled membrane.
A solution containing 3.33 g of (3-acrylamidopropane)trimethylammonium chloride (APTAC)monomer as a 75% aqueous solution, 0.373 g of N,N′-methylenebisacrylamide (BIS) cross-linker, and 0.0325 g of Irgacure ®2959 photoinitiator dissolved in 25 ml of a dioxane:dimethylformamide:water mixture, with the solvent volume ratio of 71:12:17, respectively, was prepared. In this solvent mixture, dioxane is a poor solvent while DMF and water are good solvents. A total monomer concentration (APTAC and BIS) of 0.58 mol/L was thus obtained. The cross-linking degree was 20 mol %, based on APTAC. 5 ml of this solution was placed in a glass vial and subjected to UV irradiation at 350 nm for 2 hrs. A white gel was formed which was washed thoroughly with de-ionized water to exchange the reaction solvent and remove the unreacted monomer or soluble oligomers.
The gel formed was mechanically very weak. A sample of the gel was examined using an environmental scanning electron microscope (ESEM) with water vapor present in the sample chamber to prevent drying of the gel. The micrograph, shown in
This example illustrates a method of preparing a positively charged macroporous gel-filled membrane using the monomer solution of composition described in example 4 applied to a sample of the poly(propylene) porous support PP1545-4. The macroporous gel-filled membrane was prepared according to the general procedure described above using UV irradiation at 350 nm for 2 hours. After polymerization, the macroporous gel-filled membrane was washed with with de-ionized water for 48 hrs.
Mass gain of the resulting macroporous gel-filled membrane after drying was 107 wt %, water flux was 1643±5 kg/m2h at 50kPa and Darcy permeability was 9.53×10−16 m2.
The morphology of the gel-incorporated macroporous gel-filled membrane was examined using ESEM in the same manner as described in Example 4. The ESEM micrograph shown in
This example illustrates a method of preparing a negatively charged macroporous gel-filled membrane, with a weak acid functionality.
5.50 g of vacuum-distilled methacrylic acid (MAA) monomer, 0.4925 g of N,N′-methylenebisacrylamide cross-linker and 0.1503 g of Irgacure® 2959 photoinitiator were dissolved in 25 ml of a dioxane:DMF solvent mixture with a volume ratio of 9:1, respectively, to prepare the starting monomer solution. The macroporous gel-filled membrane was prepared using the poly(propylene) PP1545-4 support and the general procedure for the photoinitiated polymerization described above. The irradiation time used was 2 hours and the resulting membrane was washed with DMF for 24 hrs followed by a 48 hr wash with deionized water. The mass gain of the resulting dried membrane was 231 wt %, water flux was 4276±40 kg/m2h at 50 kPa and Darcy permeability was 2.64×10−15 m2.
The protein (lysozyme) absorption/desorption characteristics of the macroporous gel-filled membrane were examined using the general procedure for a single membrane disk outlined earlier. The concentration of the protein used in this experiment was 0.5 g/L in a 10 mM MES buffer at pH 5.5. The flow rate of adsorption experiment was regulated to be 2-4 ml/min. A plot of the concentration of lysozyme in permeate versus the volume of permeate is shown in
This example illustrates the effect of the total monomer concentration and solvent mixture on the hydraulic flow rate (flux) of composite membranes with weak acid functionality of the type described in Example 3.
A series of composite membranes (MAA1 through MAA5) were prepared using monomer solutions of chemical compositions listed in Table 1 and the porous support PP1545-4. The preparation procedure described in Example 3 was employed.
As can be seen from Table 2, the hydraulic flow rate (flux) of composite membranes can be tuned by adjusting the monomer loading in the solution. Contrary to the typical trends found with homogeneous gels, for which an increase in gel density is followed by decrease in permeability, the increase in the mass gain in the membranes of this series results in the flux increase. Further increase in flux is achieved when the concentration of the poor solvent (dioxane) in the solvent mixture is increased (compare samples MAA3 and MAA5).
This example illustrates a method of preparing a negatively charged macroporous gel-filled membrane that has strong acid functionality.
A solution containing 2.50 g 2-acrylamido-2-methyl-1-propanesulfonic acid (AMPS) monomer, 0.372 g N,N′-methylenebisacrylamide cross-linker and 0.0353 g Irgacure® 2959 photo-initiator, dissolved in 25 ml of a dioxane:H2O mixture with a volume ratio 9:1, respectively, was used. A macroporous gel-filled membrane was prepared from the solution and the support PP1545-4 using the photoinitiated polymerization according to the general procedure describe above. The irradiation time used was 1 hour at 350 nm. After polymerization, the membrane was extracted with de-ionized water for 48 hrs. The mass gain of the resulting membrane was 74.0 wt %, water flux was 2559±40 kg/m2h at 50 kPa and Darcy permeability was 1.58×10−15 m2.
This example illustrates further the effect of the solvent mixture composition and the cross-linking degree on the hydraulic flow rate of composite membranes with the strong acid functionality. A series of composite membranes (AMPS1 through AMPS5) was prepared using chemical compositions listed in Table 3 following the general preparation procedure and the irradiation conditions as in example 5.
AMPS3 is the composite membrane prepared in the previous Example.
As can be seen, a similar pattern to that described in example 7 was observed with regards to the relationship between solubility of polymer in the solvent and water flux of composite membranes. Comparison of AMPS2 with AMPS 4 shows that hydraulic flow rate (flux) of a composite membrane can also be adjusted by the degree of cross-linking.
This example illustrates the effect of introducing a neutral co-monomer into a negatively charged macroporous gel-filled membrane.
A solution containing 1.750 g of 2-acrylamido-2-methyl-1-propanesulfonic acid, 0.485 g of acrylamide, 0.868 g of N,N′-methylenebisacrylamide cross-linker, and 0.044 g of Irgacure® 2959 photo-initiator, dissolved in 25 ml of a dioxane:DMF:H2O mixture with a volume ratio 8:1:1, respectively, was prepared. A macroporous gel-filled membrane was prepared from the solution and the support PP1545-4 using the photoinitiated polymerization according to the general procedure describe above. The irradiation time used was 1 hour at 350 nm. After polymerization, the membrane was extracted with de-ionized water for 48hrs.
The mass gain of the resulting membrane was 103 wt %, water flux was 7132±73 kg/m2.h at 100 kPa, and Darcy permeability was 4.40×10−15 m2.
This example illustrates one method of making a positively charged macroporous gel-filled membrane.
A 15 wt-% solution was prepared by dissolving diallyldimethylammonium chloride (DADMAC) monomer and N,N′-methylenebisacrylamide (BIS) cross-linker in a molar ratio of 5:1, respectively, in a solvent mixture containing 37 wt-% water, 45 wt-% dioxane and 18 wt-% DMF. The photo-initiator Irgacure® 2959 was added in the amount of 1% with respect to the mass of the monomers.
A macroporous gel-filled membrane was prepared from the solution and the support PP1545-4 using the photoinitiated polymerization according to the general procedure describe above. The irradiation time used was 30 minutes at 350 nm. The macroporous gel-filled membrane was removed from between the polyethylene sheets, washed with water and TRIS-buffer solution and stored in water for 24 hrs.
Several samples similar to that described above were prepared and averaged to estimate the mass gain of the macroporous gel-filled membrane. The substrate gained 42.2% of the original weight in this treatment.
The macroporous gel-filled membrane produced by this method had a water flux in the range of 2100-2300 kg/m2 hr at 70 kPa and Darcy permeability of 9.87×10−16 m2.
The protein (BSA) adsorption characteristic of the macroporous gel-filled membrane was examined using the general procedure for a single membrane disk described above. The concentration of the protein used in this experiment was 0.4 g/L in 50 mM TRIS-buffer. The flow rate was 2-4 ml/min. A plot of the concentration of BSA in the permeate vs. the permeate volume is shown in
This example illustrates that by adding a neutral monomer to the charged monomer used in Example 9 the protein binding capacity can be substantially increased.
A 10 wt-% solution was prepared by dissolving diallyldimethylammonium chloride (DADMAC) and acrylamide (AAM), in the ratio 80:20, in a solvent mixture containing 63 wt-% dioxane, 18 wt-% water, 15 wt-% DMF, and 4 wt-% dodecanol. N,N′-methylenebisacrylamide cross-linker was added to the monomer solution to obtain 40% (mol/mol) cross-linking degree. The photoinitiator Irgacure® 2959 was added in the amount of 1% with respect to the total mass of monomers.
A macroporous gel-filled membrane was prepared from the solution and the support PP1545-4 using the photoinitiated polymerization according to the general procedure describe above. The irradiation time used was 20 minutes at 350 nm. The macroporous gel-filled membrane was removed from between the polyethylene sheets, washed with water, TRIS-buffer solution and stored in water for 24 hrs.
A similar sample to that described above was prepared and used to estimate the mass gain of the macroporous gel-filled membrane. The substrate gained 80% of the original weight in this treatment.
The macroporous gel-filled membrane produced by this method had a water flux in the range of 250 kg/m2 hr at 70 kPa and Darcy permeability was 1.09×10−16 m2.
The protein (BSA) adsorption characteristic of the macroporous gel-filled membrane was examined using the general procedure for a single membrane disk described above. The protein concentration was 0.4 g/L in a 50 mM TRIS buffer solution. The flow rate of absorption experiment was adjusted to 2-4 ml/min. The macroporous gel-filled membrane had a BSA binding capacity of 104 mg/ml.
This example illustrates the formation of a supported porous gel macroporous gel-filled membrane by cross-linking of a pre-formed polymer.
Three separate solutions were prepared with the following compositions: (A) 20 g of branched poly(ethyleneimine) (BPEI) (25,000 Da) in 50 ml of methanol, (B) 20 g of poly(ethyleneglycol) PEG (˜10,000 Da) in 50 ml of methanol, and (C) ethyleneglycol diglycidyl ether (0.324 g) in 5 ml of methanol.
A mixture of the three solutions was prepared consisting of 2 ml of (A), 3 ml of (B), and 5 ml of (C). A portion of this resulting solution was allowed to stand in a vial overnight when a phase separation was observed. Examination of the morphology of the upper clear gel layer indicated that it was macroporous.
The same mixed solution was spread on a sample of poly(propylene) support PP1545-4 using the techniques described in the general procedure. The membrane was sandwiched between two poly(ethyleneterephthalate) sheets and allowed to stand overnight. The macroporous gel-filled membrane was extracted with methanol at room temperature for 24 h, and a mass gain of 95% was observed. The water flux of the macroporous gel-filled membrane was 6194 kg/m2 h at 100 kPa and Darcy permeability was 4.89×10−10 m2.
The dynamic protein absorption capacity of the macroporous gel-filled membrane was measured using a BSA solution (0.4 mg/mL) in the method for a single membrane disk described in the general section above. It had a capacity of 68 mg/ml before breakthrough.
This example illustrates the effect of monomer mixture composition and the polymerization conditions on the hydraulic properties of macroporous gel-filled membranes prepared by in situ polymerization of glycidyl methacrylate (GMA) with ethylene dimethacrylate (EDMA) used as a cross-linker. The solvents used were dodecanol (DDC), cyclohexanol (CHX), and methanol. A porous polypropylene support membrane PP1545-4 and two modes of initiation of in situ polymerization were used according to the general procedure described above. In the photopolymerization mode, 2,2-dimethoxy-2-phenylacetophenone (DMPA) was used as a photoinitiator while the thermal polymerization was initiated by 1,1′-azobis(cyclohexanecarbonitrile). In both modes, the polymerization was carried out for 2 hours.
The polymerization conditions and properties of the macroporous gel-filled membranes containing porous poly(glycidyl methacrylate-co-ethylene diacrylate) are presented in Table 4.
The mass gain obtained in this series of macroporous gel-filled membranes is proportional to the total monomer concentration in the polymerization mixture. Membranes AM612, AM615, and AM619 were prepared using high concentration of monomers while in membranes AM614 and AM616 the monomer concentration was cut approximately by half (Table 4).
The high values of the pure water flux measured at 100 kPa of transmembrane pressure (Table 4) indicate that the pore-filling material is macroporous. The pure water flux and, consequently, the hydraulic radius are affected not only by the mass gain but also by the polymerization mode. As shown in
By changing the solvents from dodecanol/cyclohexanol 9/91 to methanol, which is cheaper and environmentally more acceptable than the other solvents, a macroporous gel-filled membrane with very high flux was obtained (membrane AM619, Table 4). The macroporous gel-filled membrane was produced from the concentrated monomer mixture and had flux comparable with that of the membrane AM614 which had a mass gain almost twice as low as that of AM619.
This and subsequent examples illustrate a feature of some macroporous gel-filled membranes. With the capability to change the composition and concentration of the monomers and solvents, there can be produced stable macroporous gel-filled membranes with different porous structures. As shown in Table 4, macroporous gel-filled membranes with larger pores can be made in this way.
This example illustrates further the effect of monomer mixture composition on the hydraulic properties of macroporous gel-filled membranes.
A series of macroporous gel-filled membranes have been prepared according to the general procedure described above and containing porous poly(acrylamide) gels formed by in situ photoinitiated polymerization of acrylamide (AAM) and N,N′-methylenebisacrylamide (BIS) as a cross-linker in the pores of a poly(propylene) support membrane. The porous support member used was poly(propylene) TIPS membrane PP1545-4. 2,2-Dimethoxy-2-phenylacetophenone (DMPA) or 1-[4-(2-hydroxyethoxy)phenyl]2-hydroxy-2-methyl-1-propane-1-one (Irgacure® 2959) were used as photoinitiators. Irradiation was carried out at 350 nm for 2 hours. Composition of the pore-filling solutions and the properties of the resulting macroporous gel-filled membranes are summarized in Table 5.
Membranes AM606 through AM609 have been prepared using very similar concentration of monomers (12.0-14.4 wt-%) and a similar, relatively high, degree of cross-linking (16.8-18.0 wt-% of monomers). The mass gains obtained with these macroporous gel-filled membranes are also very similar. As shown in
The high degree of cross-linking in the macroporous gel-filled membrane prepared from aqueous solution without non-solvent (AM606) leads to relatively high permeability. Surprisingly, the addition of methanol or glycerol, which are poor solvents for linear poly(acrylamide), to water, which is a good solvent for the linear polymer, brings about a substantial reduction in the Darcy permeability and the hydrodynamic radius calculated on its basis. The reduction in permeability is higher with glycerol than with methanol and increases with the amount of glycerol in the solution.
The use of mixtures of poor solvents, such as N,N′-dimethylformamide and 1-propanol or 1-octanol, as well as the further increase of the degree of cross-linking and total monomer concentration have been tested in membranes AM610, AM611, and AM617. As shown in Table 5, substantially higher permeabilities and hydraulic radii are obtained with all these macroporous gel-filled membranes as compared to the macroporous gel-filled membranes prepared with water as one of the solvents. This occurred despite an increase of the total monomer concentration; more than double in membranes AM610 and AM617 than that in the membranes prepared with water as one of the solvents. Changing the other solvent from 1-propanol in AM610 to 1-octanol in AM617 also brings about substantial increase in permeability and hydraulic radius.
Microscopic images of the surface of membrane AM610 are shown in FIGS. 15 (AFM) and 16 (ESEM). For comparison, an ESEM image of the nascent porous support member is also shown in
Membrane AM611 was prepared with DMF and 1-propanol but the total monomer concentration was just over half that of AM610. Membrane AM611 shows very high flux and the hydraulic radius three times that of AM610. The ESEM images of the surface of this membrane are presented in
A comparison of surfaces of membranes AM610 and AM611 is presented in
The macroporous gel-filled membranes prepared in this example can serve as ultrafiltration membranes. It has been shown that the pore size of the macroporous gel-filled membrane and, therefore, its separation properties, can be controlled to achieve a wide range of values.
This example illustrates the effect of pore size of the support member on the hydraulic flow rate (flux) through macroporous gel-filled membranes.
Two polypropylene support membranes of pore size 0.45 μm and 0.9 μm, PP1545-4 and PP1183-3X, respectively, were used to produce macroporous gel-filled membranes with the same monomer mixture containing 39.4 wt-% of glycidyl methacrylate and 9.2 wt-% of ethylene diacrylate in methanol, thus having 48.6 wt-% of monomers and 18.9 wt-% of ethylene diacrylate (cross-linker) in the monomer mixture. The photoinitiator used was DMPA in the amount of 1.3 wt-% of monomers.
The macroporous gel-filled membranes were prepared according to the general procedure described above. The irradiation time was 2 hours at 350 nm. The resulting macroporous gel-filled membranes were washed with methanol followed by deionized water. The macroporous gel-filled membranes were tested for water flux at 100 kPa to calculate the Darcy permeability and hydraulic radius. The results are presented in Table 6.
The data shows that the hydraulic radius in both macroporous gel-filled membranes is the same within an experimental error, proving that the macroporous gel-filled membranes contain macroporous gels of similar structure.
This Example illustrates the synthesis of poly(ethylene glycol) (PEG, MW's 4000, 2000, 1,000, and 200) diacrylates, which can be used as cross-linkers to prepare the macroporous gel-filled membrane.
The synthesis procedure used follows that described by N. Ch. Padmavathi, P. R. Chatterji, Macromolecules, 1996, 29, 1976, which is incorporated herein by reference. 40 g of PEG 4000 was dissolved in 150 ml of CH2Cl2 in a 250-ml round bottom flask. 2.02 g of triethylamine and 3.64 g of acryloyl chloride were added dropwise to the flask separately. Initially the reaction temperature was controlled at 0° C. with an ice bath for 3 hrs, and then the reaction was allowed to warm to room temperature and kept for 12 hrs. The reaction mixture was filtered to remove the precipitated triethylamine hydrochloride salt. The filtrate then was poured into an excess of n-hexane. The colorless product, referred to as PEG 4000 diacrylate, was obtained by filtration and drying at room temperature.
The same procedure was used with the PEG's of other molecular weights. The molar ratios of the PEG to acryloyl chloride were kept the same as used above with PEG 4000.
This example illustrates a further method of preparing a negatively charged macroporous gel-filled membrane that has a high adsorption capacity for lysozyme.
A solution containing 0.6g of 2-acrylamido-2-methyl-1-propanesulfonic acid (AMPS), and 0.4g of acrylamide (AAM) as monomers, 0.25 g of N,N′-methylenebisacrylamide (BIS) and 1.0 g of PEG 4000 diacrylate obtained in Example 14 as cross-linkers, and 0.01 g of Iragure ®2959 as a photoinitiator was prepared in 10 ml of solvent consisting of a 80:10:10 volume ratio of dioxane, dimethylformamide (DMF), and water.
A porous poly(propylene) support member in the form of a membrane (PP1545-4) was used and the macroporous gel-filled membrane was prepared according to the general procedure, with the irradiation carried out at 350 nm for 20 minutes. After polymerization, the macroporous gel-filled membrane was washed thoroughly with de-ionized water for 24 hrs.
Mass gain of the resulting macroporous gel-filled membrane after drying was 113.2 wt %, water flux was 366±22 kg/m2h at 100 kPa, and Darcy permeability was 2.26×10−16.
The protein (lysozyme) absorption/desorption characteristics of the macroporous gel-filled membrane were examined using the general procedure for a single membrane disk outlined earlier. The concentration of the protein used in this experiment was 0.5 g/L in a 10 mM MES buffer at pH 5.5. The flow rate of adsorption experiment was regulated to be 2-4 ml/min. A plot of the concentration of lysozyme in the permeate versus the volume of permeate is shown in
This example illustrates preparation of a negatively charged macroporous gel-filled membrane with the same nominal polymer composition as in Example 18 but with much higher hydraulic flows (flux) and good lysozyme uptake capacity.
The monomer solution was produced by dilution of the solution formulated in Example 15 with acetone with the mass ratio of 1:1.
A porous poly(propylene) support member in the form of a membrane (PP1545-4) was used and the macroporous gel-filled membrane was prepared according to the general procedure. The irradiation time used was 2 hours. After polymerization, the macroporous gel-filled membrane was washed thoroughly with de-ionized water for 24 hrs.
The mass gain of the resulting macroporous gel-filled membrane after drying was 51.1 wt % and water flux was 6039±111 kg/m2.h at 100 kPa giving Darcy permeability of 3.73×10−15.
The protein (lysozyme) absorption/desorption characteristics of the macroporous gel-filled membrane were examined using the general procedure for a single membrane disk outlined earlier. The concentration of the protein used in this experiment was 0.5 g/L in a 10 mM MES buffer at pH 5.5. The flow rate of adsorption experiment was regulated to be 2-4 ml/min. A plot of the concentration of lysozyme in permeate versus the volume of permeate is shown in
This example illustrates a further preparation of a negatively charged macroporous gel-filled membrane that has a very high flux but lower protein binding capacity.
The monomer solution was produced by dilution of the solution formulated in Example 15 with acetone with the mass ratio of 1:2.
A porous poly(propylene) support member in the form of a membrane (PP1545-4) was used and the preparation of macroporous gel-filled membrane was carried out according to the general procedure described above. UV initiated polymerization was carried out for 2 hours. After polymerization, the macroporous gel-filled membrane was washed thoroughly with de-ionized water for 24 hrs.
The mass gain of the resulting macroporous gel-filled membrane after drying was 34.4 wt % and water flux was 12184±305 kg/m2h at 100 kPa giving Darcy permeability of 7.52×10−15.
The protein (lysozyme) absorption/desorption characteristics of the macroporous gel-filled membrane were examined using the general procedure for a single membrane disk outlined earlier. (The concentration of the protein used in this experiment was 0.5 g/L in a 10 mM MES buffer at pH 5.5. The flow rate of adsorption experiment was regulated to be 2-4 ml/min.) A plot of the concentration of lysozyme in permeate versus the volume of permeate is shown in
Examples 18, 19 and 20 show that it is possible to control the loading of porous gel into the host membrane thereby controlling the water flux at a defined pressure (100 kPa in the data given in the examples) and also that the lysozyme uptake is related to the mass of incorporated porous gel.
This example illustrates preparation of a negatively charged membrane that has both good protein adsorption capacity and good flux using a macromonomer.
A monomer solution containing 0.6 g of 2-acrylamido-2-methyl-1-propanesulfonic acid (AMPS), 0.4 g of acrylamide (AAM), 0.25 g of N,N′-methylenebisacrylamide (BIS), 0.01 g of Irgacure® 2959, and 1.0 g of PEG 2000 macromonomer obtained in Example 14, dissolved in 10 ml of a dioxane-(DMF)-water mixture with a volume ratio 80:10:10, respectively, was prepared.
A microporous poly(propylene) support member in the form of a membrane, support PP1545-4, was used together with the general procedure described above. The irradiation time used was 20 minutes. After polymerization, the membrane was washed thoroughly with de-ionized water for 24 hrs.
The mass gain of the resulting membrane after drying was 108.4 wt % and water flux was 1048±4 kg/m2h at 100 kPa giving Darcy permeability of 6.47×10−16.
The protein (lysozyme) adsorption/desorption characteristics of the membrane were examined using the general procedure for a single membrane disk outlined earlier. A relatively steep break through curve was obtained. The membrane had a lysozyme binding capacity of 88.7 mg/ml. The desorption experiment indicated that the recovery of protein was 64.0%.
This example in combination with example 18 above further illustrates that the protein binding capacity and flow characteristics of a membrane can be tuned.
The monomer solution was produced by dilution of the solution formulated in Example 21 with acetone with the mass ratio of 1:1.
A porous poly(propylene) support member in the form of a membrane, support PP1545-4, was used along with the general procedure for the preparation of macroporous gel-filled membranes described above. The irradiation time used was 90 minutes. After polymerization, the membrane was washed thoroughly with de-ionized water for 24 hrs.
The mass gain of the resulting membrane after drying was 45.7 wt % and water flux was 7319±180 kg/m h at 100 kPa.
The protein (lysozyme) absorption/desorption characteristics of the membrane were examined using the general procedure for a single membrane disk outlined earlier. A relatively steep break through curve was observed. The membrane had a lysozyme binding capacity of 63.4 mg/ml. The desorption experiment indicated that the recovery of protein was 79.3%.
This example illustrates the effect of a neutral co-monomer on the protein binding capacity of macroporous gel-filled membranes.
PEG2000XL: PEG2000 diacrylate prepared in example 14
Monomer solutions were prepared by dilution of stock solutions S1-S4 in Table 7 with acetone with the mass ratio of 1:1.
Composite membranes M1-M4 were prepared by using the corresponding diluted solutions of stocks S1-S4 and following the general preparation procedure described earlier. The porous support used was PP1545-4 and the irradiation time was 90 minutes. Upon completion of polymerization, the composite membranes were washed with de-ionized water for 24 hrs.
The properties and protein binding capacities of composite membranes were examined and the results shown in Table 8. It is evident that the charge density of polyelectrolyte gels influences significantly protein adsorption onto membranes.
This example illustrates the effect of the chain length of the polyfunctional macromonomers used as cross-linkers (PEG diacrylates) on protein binding capacity of macroporous gel-filled membranes.
A series of stock solutions containing 0.6 g of 2-acrylamido-2-methyl-1-propanesulfonic acid (AMPS), 0.4 g of acrylamide (AAM), 0.10 g of N,N′-methylenebisacrylamide (BIS), 0.01 g of Irgacure® 2959, and 1.0 g of PEG diacrylate with different molecular weights (200, 1000, 2000, 4000), obtained in Example 17, dissolved in 10 ml of a dioxane-DMF-water mixture with a volume ratio 80:10:10, respectively, was prepared. The stock solutions were subsequently diluted with acetone at the mass ratio of 1:1. A series of composite membranes were prepared from these solutions using poly(propylene) support PP1545-4 and by following the general preparation procedure described above. The irradiation time used was set to 90 minutes. Upon completion of polymerization, the composite membranes were washed with de-ionized water for 24 hrs.
The properties and protein binding capacities of composite membranes were examined according to the general procedure for a single membrane disk. The results shown in Table 9 clearly indicate that the gel structure of composite membranes has substantial effect on protein adsorption. Possibly, it is related to the gel structure near the macropore surface, where an extremely loose structure may be formed that can allow protein to penetrate into the gel layer at a certain depth. Another possibility is that by using a longer chain PEG diacrylate the surface area is increased owing to some fuzziness at the surface and thus making more adsorption site available to proteins.
This example illustrates the use of fibrous non-woven support to produce a macroporous gel-filled membrane containing positively charged macroporous gel. A 10 wt-% solution was prepared by dissolving diallyldimethylammonium chloride (DADMAC) and acrylamide (AAM), which were taken in the ratio 80:20, in a solvent mixture containing 65 wt-% of dioxane, 18 wt-% of water, and 17 wt-% of DMF. N,N′-methylenebisacrylamide (BIS) was added to the monomer solution to obtain 40% (mol/mol) cross-linking degree. The photoinitiator Irgacure® 2959 was added in the amount of 1% with respect to the total mass of the monomers.
A sample of the fibrous non-woven polypropylene substrate TR2611A was placed on a polyethylene sheet and filled with the monomer solution. The substrate was subsequently covered with another polyethylene sheet and the resulting sandwich was run between two rubber rollers to press the monomer solution into the pores and remove excess of solution. The filled substrate was irradiated at 350 nm for 20 min for the polymerization process to occur. The macroporous gel-filled membrane was removed from between the polyethylene sheets, washed with water, TRIS-buffer solution and stored in water for 24 hrs. A duplicate sample was used to estimate the mass gain of the macroporous gel-filled membrane. The substrate gained 45% of the original weight in this treatment.
The macroporous gel-filled membrane produced by this method had a water flux in the range of 2320 kg/m2 hr at 70 kPa.
The protein (BSA) adsorption characteristic of a mono-layer of the macroporous gel-filled membrane was examined using the general procedures one for a single membrane disk and one for a multi-membrane stack, as described above. The membrane stack contained 7 membrane layers of total thickness 1.75 mm. In both experiments the protein concentration was 0.4 g/L in a 50 mM TRIS buffer solution, and the flow rate of the protein solution used was 3.1±0.1 ml/min, delivered by peristaltic pump. The breakthrough capacity for BSA was 64 mg/ml in the single membrane experiment and 55±2 mg/ml in the multi-membrane stack experiment.
This example illustrates the use of a mixture of two monomers in making a positively charged macroporous gel-filled membrane.
A 10 wt-% solution was prepared by dissolving diallyldimethylammonium chloride (DADMAC) and (3-acrylamidopropyl)trimethylamonium chloride (APTAC), in the ratio 50:50, in a solvent mixture containing 65 wt-% dioxane, 18 wt-% water and 17 wt-% DMF. N,N′-methylenebisacrylamide (BIS) was added to the monomer solution to obtain 40% (mol/mol) cross-linking degree. The photoinitiator Irgacure® 2959 was added in the amount of 1% with respect to the total mass of the monomers.
A sample of the non-woven polypropylene substrate TR2611A was placed on a polyethylene sheet and filled with the monomer solution. The substrate was subsequently covered with another polyethylene sheet and the resulting sandwich was run between two rubber rollers to press the monomer solution into the pores and remove excess of solution. The filled substrate was irradiated at 350 nm for 20 min for the polymerization process to occur. The macroporous gel-filled membrane was removed from between the polyethylene sheets, washed with water, TRIS-buffer solution and stored in water for 24 hrs.
The macroporous gel-filled membrane produced by this method had a water flux in the range of 2550 kg/m2 hr at 70 kPa. The mass gain determined with a duplicate sample was found to be 45%.
The protein (BSA) adsorption characteristic of the mono-layer macroporous gel-filled membrane was examined using the general procedure for a single membrane disk described above. A solution of BSA concentration of 0.4 g/L in a 50 mM TRIS buffer solution was delivered to the membrane at a flow rate of 2-4 ml/min. The breakthrough capacity of the macroporous gel-filled membrane was 40 mg/ml.
This example illustrates the effect of addition of a neutral monomer to the mixture of charged monomers used in example 26.
A 15 wt-% solution was prepared by dissolving diallyldimethylammonium chloride (DADMAC), (3-acrylamido-propyl)trimethylammonium chloride (APTAC), and acrylamide (AAM), which were taken in the ratio 40:40:20, in a solvent mixture containing of 65 wt-% dioxane, 17 wt-% of water, and 18 wt-% of DMF. N,N′-methylenebisacrylamide (BIS) was added to the monomer solution to obtain 20% (mol/mol) cross-linking degree. The photoinitiator Irgacure® 2959 was added in the amount of 1% with respect to the total mass of the monomers.
A sample of the non-woven polypropylene substrate TR2611A was placed on a polyethylene sheet and filled with the monomer solution. The substrate was subsequently covered with another polyethylene sheet and the resulting sandwich was run between two rubber rollers to press the monomer solution into the pores and remove excess of solution. The filled substrate was irradiated at 350 nm for 20 min for the polymerization process to occur. The macroporous gel-filled membrane was removed from between the polyethylene sheets, washed with water, TRIS-buffer solution and stored in water for 24 hrs.
The macroporous gel-filled membrane produced by this method had a water flux of 550 kg/m2 hr at 70 kPa and a mass gain (determined using a duplicate samples) of 65 wt-%.
The protein (BSA) adsorption characteristic of the mono-layer macroporous gel-filled membrane was examined using the general procedure for a single membrane disk described above. A solution of BSA concentration of 0.4 g/L in a 50 mM TRIS buffer solution was delivered to the membrane at a flow rate of 3.5-4 ml/min. The breakthrough capacity of the macroporous gel-filled membrane was 130 mg/ml.
This example illustrates the formation of unsupported positively charged macroporous gel by cross-linking of a preformed polymer.
A 10% solution of poly(allylamine hydrochloride) PAH was prepared by dissolving the polymer in a solvent mixture containing 60% water and 40% iso-propanol (2-propanol). The polymer was partially deprotonated (40%) by adding 6.67 N NaOH. Ethylene glycol diglycidyl ether (EDGE) was added to this solution to obtain 40% (mol/mol) degree of cross-linking. The solution was kept at room temperature for 3 hours for gel formation by the cross-linking reaction between the amine groups of PAH and epoxy groups of EDGE. After 3 hours, the gel was placed in a water bath for all un-reacted chemicals to leach out.
A sample of the wet gel was examined using ESEM. The micrograph shown in
This example illustrates the making macroporous gel incorporated in a non-woven fabric support.
A 10 wt-% solution of poly(allylamine hydrochloride) (PAH) was prepared as in Example 28. The polymer was partially deprotonated and EDGE added as described in Example 25 and the solution was applied to a sample of the non-woven polypropylene membrane support TR2611A placed between two polyethylene sheets. The resulting sandwich was run between two rubber rollers to press the polymer solution into the pores of the substrate, spread it evenly, and remove the excess solution. The solution-filled support sample was kept at room temperature for 3 hours for cross-linking process to take place resulting in the formation of gel. After that time, the macroporous gel-filled membrane was removed from the sandwich and placed in a water bath for 12 hours to leach out unreacted chemicals.
A wet sample of the resulting composite membrane was examined using ESEM. The micrograph shown in
This example provides a comparison of the protein adsorption by a composite membrane with the commercial Mustang® Coin Q produced by Pall Corporation.
A macroporous gel-filled membrane prepared in example 22 was tested in a multi-membrane stack of 7 membrane layers of a total thickness of 1.75 mm, according to the testing protocol described in Example 25. A Mustang® Coin Q was also tested under similar conditions. The membrane stack was prepared by placing seven (7) layers of the membrane sample in the wet state on top of each other. The assembled membrane stack was lightly compressed pressed to remove excess of water. The membrane stack was then heated in an oven at 60-70° C. for at least 30 min. The thickness of the resulting membrane stack in the dry state was 1.8-1.9 mm. This process produced a stack in which the multiple membrane layers adhered to each other. The results shown in
This example provides the hydrodynamic (Darcy) permeability of reference macroporous gel-filled membranes containing porous support member and homogeneous gels filling the pores of the support. The homogeneous gels were obtained by using thermodynamically good solvents and their homogeneity was assessed based on transparency of simultaneously formed unsupported gels of the same composition. Clear and transparent gels were assumed to be homogeneous, contrary to macroporous gels that were always found opaque.
The macroporous gel-filled membranes containing homogeneous gels of glycidyl methacrylate-co-ethylene glycol dimethacrylate, GMA-co-EDMA, were prepared using 1,4-dioxane as a solvent and 4.7 wt-% of EDMA (cross-linker) in monomer mixture, and different total monomer concentrations. Poly(propylene) support PP1545-4 and the general procedure for preparing the macroporous gel-filled were used. DMPA was used as photoinitiator and the irradiation was carried out at 350 nm for 120 minutes. The hydrodynamic permeability of the membranes was measured and an empirical equation was derived for the relationship between the hydrodynamic permeability, k, and the mass gain of the composite membranes containing poly(GMA-co-EDMA) homogeneous gels in the PP1545-4 support. The equation is as follows:
k=3.62×103×G−9.09
The differences between the measured values of permeability and that calculated from the above equation were found to be less than ±3%. This empirical relationship was subsequently used to calculate permeability of reference macroporous gel-filled membranes at different mass gains.
The macroporous gel-filled membranes containing homogeneous gels of diallyldimethylammonium chloride-co-methylenebisacrylamide, DADMAC-co-BIS, were prepared using water as a solvent and 1.0 wt-% of BIS cross-linker in monomer mixture, and different total monomer concentrations. Poly(propylene) support PP1545-4 was coated with Triton X-114 surfactant by immersing the support samples in 2t-% solution of the surfactant in methanol/water (60/40) mixture for 16 hours followed by drying in air. The general procedure for preparing the macroporous gel-filled membranes used to make homogeneous gel-filled membranes. Irgacure® 2959 was used as photoinitiator and the irradiation was carried out at 350 nm for 30-40 minutes. The hydrodynamic permeability of the series of membranes was measured and an empirical equation was derived for the relationship between the Darcy permeability, k, and the mass gain, G:
k=2.09×10−12×G−4.01
The hydrodynamic permeability of homogeneous poly(acrylamide)-co-methylenebisacrylamide, AAM-co-BIS, was estimated from the empirical relationship between the gel permeability and the gel polymer volume fraction provided by Kapur et al. in Ind. Eng. Chem. Res., vol. 35 (1996) pp. 3179-3185. According to this equation, the hydrodynamic permeability of a poly(acrylamide) gel, kgel=4.35×10−22×Φ−3.34, where Φ is the polymer volume fraction in the gel. In the same article, Kapur et al. provide a relationship between the hydrodynamic permeability of a gel and a porous membrane filled with the same gel. According to this relationship, the permeability of the membrane, kmembrane=(ε/τ)×kgel, where ε is the porosity of the support and τ is the tortuosity of the support pores. The pore tortuosity can be estimated as a ratio of the Kozeny constant, K, for a given porosity, i.e., K=5, and the Kozeny constant for a straight cylindrical capillary equal to 2. Thus, for the poly(propylene) support PP1545-4 with porosity of 0.85, the ratio (ε/τ)=0.85/2.5=0.34.
The polymer volume fraction, Φ, can be converted to mass gain using the partial specific volume, ν2, for poly(acrylamide) and the density, ρ, of poly(propylene). The values of these parameters can be found in Polymer Handbook, edited by Brandrup et al., Chapter VII, Wiley and Sons, New York, 1999. Thus, the mass gain of a macroporous gel-filled membrane containing poly(propylene) support of porosity ε filled with a gel whose polymer occupies the fraction Φ of the pores is given by
The above equation was combined with that of Kapur et al. to give an empirical relationship allowing one to calculate hydrodynamic permeability of reference macroporous gel-filled membranes, k, at different mass gains, G. The combined equation is as follows:
k=1.80×10−12×G−3.34
The equation is valid for ρ=0.91 g/cm3; ε=0.85; ν2=0.7 cm3/g; (ε/τ)=0.34.
The macroporous gel-filled membranes containing homogeneous gels of 2-acrylamido-2-propane-1-sulfonic acid-co-methylenebisacrylamide, AMPS-co-BIS, were prepared using water as a solvent and 10.0 wt-% of BIS cross-linker in monomer mixture, and different total monomer concentrations. Poly(propylene) support PP1545-4 was coated with Triton X-114 surfactant by immersing the support samples in 2t-% solution of the surfactant in methanol/water (60/40) mixture for 16 hours followed by drying in air. The general procedure for preparing the macroporous gel-filled membranes was used to make homogeneous gel-filled membranes. Irgacure® 2959 was used as photoinitiator and the irradiation was carried out at 350 nm for 60 minutes. The hydrodynamic permeability of the series of membranes was measured and an empirical equation was derived for the relationship between the Darcy permeability, k, and the mass gain, G:
k=2.23×10−16×G−1.38
The macroporous gel-filled membranes containing homogeneous gels of (3-acrylamidopropane)trimethylammonium chloride-co-methylenebisacrylamide, APTAC-co-BIS, were prepared using water as a solvent and 10.0 wt-% of BIS cross-linker in monomer mixture, and different total monomer concentrations. Poly(propylene) support PP1545-4 was coated with Triton X-114 surfactant by immersing the support samples in 2t-% solution of the surfactant in methanol/water (60/40) mixture for 16 hours followed by drying in air. The general procedure for preparing the macroporous gel-filled membranes was used to make homogeneous gel-filled membranes. Irgacure® 2959 was used as photoinitiator and the irradiation was carried out at 350 nm for 60 minutes. The hydrodynamic permeability of the series of membranes was measured and an empirical equation was derived for the relationship between the Darcy permeability, k, and the mass gain, G:
k=9.51×10−16×G−1.73
The macroporous gel-filled membranes containing homogeneous gels of branched poly(ethyleneimine) cross-linked with ethylene glycol diglycidyl ether (EDGE) were prepared using methanol solutions of BPEI of different concentrations. The degree of cross-linking used was 10 mol-%. Poly(propylene) support PP1545-4 was used together with the general procedure of fabrication of macroporous gel-filled membranes by in situ cross-linking of cross-linkable polymers described in Example 10. A series of membranes with different mass gains was prepared by changing the concentration of PEI in the solution. The Darcy permeability of the membranes was measured and an empirical equation describing the relationship between the permeability, k, and the mass gain, G, was derived. The equation is as follows:
k=4.38×10−14×G−2.49
This example provides comparison between Darcy permeability of macroporous gel-filled membranes containing supported macroporous gels and the permeability of the reference membranes containing homogeneous gels filling the porous support member.
The comparison is shown in Table 10 below.
These examples illustrate a method of preparing a responsive macroporous gel-filled membrane using photoinitiated free radical polymerization of acrylic acid (AA) (ionic monomer), acrylamide (AA), and trimethylolpropane triacrylate (TRIM) as a cross-linker. The molar ratio of acrylic acid to acrylamide was 1:1 and 1,4-dioxane was used as a solvent in all experiments. Monomer solution compositions and polymerization conditions are given in Table 11. After polymerization, the responsive macroporous gel-filled membrane was washed with de-ionized water for about 16 hrs.
The amount of gel formed in the support member depends on the pore volume available to fill, the total concentration of the monomer mixture, and the degree of conversion in the polymerization. In
This example illustrates the responsiveness of the responsive macroporous gel-filled membranes according to Example 33 to ionic interactions. For this purpose, the macroporous gel-filled membranes were tested with solutions of different pH and/or salt concentrations by measuring the flux at 100 kPa. A typical change in flux taking place with the change of pH from about 3 (1 mM HCl) to about 12 (1 mM NaOH), obtained with membrane AM675 is shown in
The ratio of the flux measured with 1 mM HCl to that measured with 1 mM NaOH has been taken as a measure of membrane response (MR). The results obtained with membranes described in Example 33-40 are shown in Table 12.
The results in Table 12 show that the response of the composite membranes to ionic interaction can also be controlled by the total concentration of monomer mixture and the degree of cross-linking. As the monomer concentration increases, the membrane sensitivity to the environmental changes decreases. Similar effect is found when the degree of cross-linking is increased.
This example illustrates the ability of membranes based on responsive macroporous gel-filled membranes to fractionate proteins. The separation of therapeutic proteins Human Serum Albumin (HSA) and Human Immunoglobulin G (HIgG) was chosen as a case study. Human plasma is the starting material for the production of a number of therapeutic proteins, which are referred to as plasma proteins. The most abundant amongst these are HSA and HIgG, both of which are manufactured in bulk quantities. These proteins are generally fractionated by precipitation based processes which give high product throughput but poor resolution in terms of separation. Membrane based processes such as ultrafiltrations have the potential for giving both high throughput and high resolution.
Two composite membranes containing responsive macroporous gel, duplicates of membrane AM695 (see Tables 11 and 12), were tested for their suitability in separation of these plasma proteins. In the experiments discussed here, the change in membrane pore size with change in salt concentration was utilized to effect protein-protein separation in the manner desirable, i.e. sequential release from the membrane module. Other environmental conditions such as pH could well be used to achieve a similar objective.
A binary carrier phase system was used in the ultrafiltration experiments. The starting carrier phase in all the experiments was one with a low salt concentration (typically 5-10 mM NaCl). In all the experiments the carried phase was switched to one with a high salt concentration (typically 1 M NaCl). The change in salt concentration within the membrane module could be tracked by observing the conductivity of the permeate stream. The change in transmembrane pressure gave an idea about the change in membrane hydraulic permeability with change in salt concentration.
Experiments were carried out using human serum albumin and human immunoglobulin mixtures. The ultrafiltration was started at a low salt concentration (i.e. 10 mM). At this condition, human serum albumin was transmitted while human immunoglobulin G was almost totally retained. The salt concentration was then increased and this increased the pore diameter (as evident from drop in pressure in constant permeate flux ultrafiltration). This in turn led to the transmission of human immunoglobulin G through the membrane. Hence by altering the environmental condition it was possible to sequentially transmit proteins having different sizes through the same membrane. If the initial mixture had also contained a protein significantly larger than human immunoglobulin, it would have been possible to fractionate the three proteins (i.e. human serum albumin, human immunoglobulin and the significantly bigger protein) by appropriately controlling the change in salt concentration. Two of the three fractions obtained here would be in the permeate while the third fraction would be in the retentate.
The results obtained with duplicates of membrane AM694 are shown in
The results presented in
This example illustrates a method for making a positively charged macroporous gel-filled membrane having a high protein binding capacity.
A 15 wt-% solution was prepared by dissolving (3-acrylamidopropyl)-trimethylammonium chloride (APTAC), N-(hydroxymethyl)acrylamide and N,N′-methylenebisacrylamide as cross-linker in a molar ratio of 1:0.32:0.1, respectively, in a solvent mixture containing 10 wt-% water, 60 wt-% di(propylene glycol)methyl ether and 30 wt-% dimethylformamide (DMF). The photo-initiator Irgacure® 2959 was added in the amount of 1% with respect to the mass of the monomers.
A sample of the fibrous non-woven polypropylene substrate TR2611A was placed on a polyethylene sheet and filled with the monomer solution. The substrate was subsequently covered with another polyethylene sheet and the resulting sandwich was run between two rubber rollers to press the monomer solution into the pores and to remove any excess solution. The substrate was irradiated for 5 minutes at 350 nm. The macroporous gel-filled membrane was then removed from between the polyethylene sheets, washed with water and TRIS-buffer solution and stored in water for 24 hrs.
Several samples were prepared according to the above process, and the samples were then dried and weighed. The average mass gain of the macroporous gel-filled membrane was 55.7% of the original weight of the starting support member.
The protein (BSA) adsorption characteristic of a multi-membrane stack of the above macroporous gel-filled membrane was examined using the general procedure for a mono-layer of the macroporous gel-filled membrane, as described earlier. The membrane stack tested contained 4 membrane layers, giving a total thickness 1.05 mm. The protein solution used was a 25 mM TRIS buffer solution with a protein concentration of 0.4 g/L, and the flow rate of the protein solution was 5.0 ml/min at 150 kPa. The breakthrough capacity for BSA was 281 mg/ml. In a subsequent desorption step, approximately 85% of the BSA was recovered.
All references mentioned herein are incorporated herein by reference to the same extent as if each reference were stated to be specifically incorporated herein by reference.
To those skilled in the art, it is to be understood that many changes, modifications and variations could be made without departing from the spirit and scope of the present invention as claimed hereinafter.
All publications and patent applications cited in this specification are herein incorporated by reference as if each individual publication or patent application were specifically and individually indicated to be incorporated by reference. The citation of any publication is for its disclosure prior to the filing date and should not be construed as an admission that the present invention is not entitled to antedate such publication by virtue of prior invention.
Although the foregoing invention has been described in some detail by way of illustration and example for purposes of clarity of understanding, it is readily apparent to those of ordinary skill in the art in light of the teachings of this invention that certain changes and modifications may be made thereto without departing from the spirit or scope of the appended claims.
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/CA05/00518 | 4/7/2005 | WO | 8/2/2007 |
| Number | Date | Country | |
|---|---|---|---|
| 60560027 | Apr 2004 | US |
