Stimuli-Responsive Membranes for Separations
Synthetic membranes are increasingly being used for a varied range of applications, in industry, medicine, agriculture, environmental sciences, geoengineering, and medical diagnostics, to name just a few. In the chemical industry, membranes are more commonly used in separation processes. Such separations are usually carried out based on one of two basic working principles, i.e., relative permeability and relative sorption. Between the two, permeability-based separations are more common. The physical form and functionality of conventional membranes are not expected to change appreciably during a separation process. This is to ensure reliable performance during separation and thereby ensure consistent product quality. However, the ability to alter membrane permeability in a controllable and reproducible manner increases its scope and range of application. This includes membranes for sequential multiple component separation, controlled release membranes, antifouling membranes, self-cleaning membranes, and membranes for biomedical devices. Such unconventional membranes are generally referred to as “stimuli-responsive” or in some case “environment-responsive” or “smart” membranes. This chapter reviews different kinds of stimuli-responsive membranes, including both permeability-based and sorption-based membranes.
Stimuli-responsive membranes used in separation processes could be broadly categorized into two groups. The first group includes membranes that respond to a stimulus by altering the solute or solvent permeability, i.e., true stimuli-responsive membranes or stimuli-responsive permselective membrane [17, 19, 38, 39, 40]. The second group includes membranes, the selective passage of solutes through which is not governed by “permeability” but by other factors such as physical adsorption or chemical bond formation [13, 29, 41, 42, 43, 44]. Membranes of this group could be called apparent stimuli-responsive membranes or simply stimuli-responsive adsorptive membranes. These are basically adsorbents prepared in a membrane-like format, i.e., in the form of thin microporous sheets.
2 Stimuli-Responsive Permselective Membranes
2.1 Copolymer Blend Membranes
The main advantages of a copolymer blend membrane include the ease and reproducibility of fabrication. Precise amounts of the different polymer components, i.e., the scaffold and the stimuli-responsive polymer, could be blended using appropriate solvent systems and cast into membranes. Different casting methods such as coating followed by solvent evaporation, spray coating, coacervation-phase separation, and phase inversion could be used for preparing such membranes. Alternately, the different polymer components could be cross-linked together using appropriate linker molecules.
The main disadvantages with membranes made using the above approach include low permeability and limited stimuli-responsive properties. This is because the structural and functional properties of these membranes are intertwined. Due to this, the response of the stimuli-responsive polymer could be “quenched” to some degree by the scaffold polymer. Another significant disadvantage of using this approach is the limited range of stimuli that could be used with such membranes. Some of these disadvantages could be overcome by decoupling the fabrication of the scaffold membrane and the stimuli-responsive membrane using the different grafting, filling, and coating methods described below.
2.2 Grafting to Membranes
This is the easiest strategy for making a grafted stimuli-responsive membrane. An appropriate pre-existing porous scaffold membrane is selected, and preformed stimuli-responsive polymers are chemically bonded or physically attached to it [47, 51, 54, 55]. However, physical attachment is less commonly utilized as the resultant membranes tend to be relatively more unstable than membranes prepared by chemical attachment, thereby limiting the operating stimuli windows such as temperature, pH, and ionic strength ranges. The surface of the scaffold membrane is first activated followed by the addition of a polymer possessing an active functional group. The location of the active functional group determines both the density of polymer attachment and the orientation of the responsive polymer. The main advantage of using this approach is the preciseness of the identity and behavior of the stimuli-responsive polymer. After synthesis, a polymer could be purified to a high degree of homogeneity before being attached to the surface of the scaffold. By doing so, the permeability and responsiveness of the resultant membrane could be very precisely controlled. The stimuli-responsive polymer could be selectively attached in a brushlike configuration close to the pore entrance to create a surface (or lid) valve or could be attached in a brushlike configuration within the pores to create an internal valve. If required, the grafted polymer could be further cross-linked to create mesh-like valves. The main advantages of the brushlike configuration include high membrane permeability and a large range for the stimuli-responsive permeability. The main disadvantage with the brushlike configuration is low stability, i.e., drastic change in permeability in response to small change in the stimulus. When the polymer brushes are cross-linked, the magnitude of response is low but better controlled.
The main disadvantage of the grafting to approach is the low grafting density. When attaching preformed polymer chains to surfaces, steric hindrance limits the number of chains attached per unit surface area, particularly when grafting on concave surfaces. This problem can be solved by using the grafting from approach described next.
2.3 Grafting from Membranes
The grafting from approach involves growing polymer chains on the surface of the scaffold membrane [53, 56, 57, 58, 59, 60]. The first step typically involves the attachment of an initiator molecule on the surface. The polymer chain is then grown by controlled monomer addition and reaction. Different chemistries such as atom transfer radical polymerization [53, 56, 57, 58] and reversible addition-fragmentation chain transfer [59, 60] are used for the grafting from approach. Depending on the method used, different approaches are available for obtaining specific monomer composition and chain length of the responsive polymer. In addition to the potential for increasing the graft density, the grafting from approach is generally cheaper than the grafting to approach as the polymer purification step is not necessary.
The main disadvantage of the grafting from approach is the relatively greater heterogeneity of the attached responsive polymer when compared with the grafting to approach. On account of this, the magnitude and responsiveness of permeability are less controllable. Both the grafting to and the grafting from approaches rely on the covalent attachment of the responsive polymer on the membrane. On account of this, the range of material that could be used for preparing stimuli-responsive membranes gets limited. This problem can be overcome using the filled and coated membrane approaches described next where covalent attachment of the responsive polymer to the scaffold membrane is not required.
2.4 Filled Membranes
The main disadvantage of using the pore-filling approach is the relatively low permeability of the resulting membrane. Also, the selectivity of such membranes tends to be low. The problem of low permeability can be solved to some extent using the coated membrane approach discussed next.
2.5 Coated Membranes
3 Stimuli-Responsive Adsorptive Membranes
Conventional membranes are suitable for binary separation, i.e., where a feed material has to be fractionated into two components, the permeate and the retentate. On the other hand, stimuli-responsive membranes whose permeability could be changed in a sequential manner using multiple stimuli are suitable for carrying out multicomponent separations. This opens up endless possibilities about how a separation process could be designed and operated. It also increases process efficiency by replacing a multi-device process with a single device one. Stimuli-responsive membranes could potentially be used for the regulation of concentration of a chemical species within chemical reactors just as the cell membrane regulates the concentration of different species within a cell. The same principle could be applied for regulating species concentration in bioreactors, such as those used for mammalian cell culture or tissue engineering. Membrane fouling is widely considered the Achilles heel of membrane filtration. In most conventional membrane processes, a membrane needs to be frequently cleaned, and this increases the cost of production. Stimuli-responsive membrane can function as good “self-cleaning” or “antifouling” membranes as they could quite easily be cleaned either by forward or back flushing of the foulant in conjunction with an appropriate stimulus which “opens up” the membrane. Using such membranes, separation processes could easily be carried out in a sustained or semicontinuous manner by implementing some simple in-process cleaning strategies. Stimuli-responsive membranes used in separation processes could be broadly categorized into two groups: (a) true stimuli-responsive membranes or stimuli-responsive permselective membranes that respond to a stimulus by altering the solute or solvent permeability, and (b) apparent stimuli-responsive membranes or stimuli-responsive adsorptive membranes, the selective passage of solutes through which is not governed by “permeability” but by other factors such as physical adsorption or chemical bond formation. Stimuli-responsive membranes could be prepared by “copolymer blending,” i.e., by mixing a stimuli-responsive polymer with a standard membrane scaffold polymers and then casting this into a membrane using standard fabrication techniques. However, these membranes tend to have poor permeability and smaller responses to stimuli. Grafted membranes can be prepared using a “grafting to” or a “grafting on approach.” However, grafting limits the range of material that could be used as scaffold or as the stimuli-responsive polymer. This problem could be overcome by preparing filled or coated membranes. Filled and coated membranes are prepared in a similar way, but they differ in terms of the morphology of the stimuli-responsive polymer. In a filled membrane, the responsive polymer forms the bulk domain within the pores, while with coated membranes, the responsive polymer just coats the surface of the scaffold membrane. Overall, the field of stimuli-responsive membranes is a promising new area for research, development, and application.
- 1.N.N. Li, A.G. Fane, W.S.W. Ho, T. Matsuura, Advanced Membrane Technology and Applications (Wiley, New York, 2011)Google Scholar
- 2.W. Ho, K. Sirkar, Membrane Handbook (Springer Science & Business Media, New York, 2012)Google Scholar
- 5.S. Judd, The MBR Book: Principles and Applications of Membrane Bioreactors for Water and Wastewater Treatment (Elsevier, San Diego, 2010)Google Scholar