VANHERCK, Katrien (Nekkerspoelstraat 349, Mechelen, B-2800, BE)
VANKELECOM, Ivo (Brainestraat 7, Blanden, B-3052, BE)
VERBIEST, Thierry (Speelbergweg 23, Veltem-Beisem, B-3020, BE)
HERMANS, Sanne (Boudewijnlaan 32-bus 5, Maasmechelen, B-3630, BE)
VANHERCK, Katrien (Nekkerspoelstraat 349, Mechelen, B-2800, BE)
VANKELECOM, Ivo (Brainestraat 7, Blanden, B-3052, BE)
VERBIEST, Thierry (Speelbergweg 23, Veltem-Beisem, B-3020, BE)
| CLAIMS 1. A membrane based separation process of substances characterised in that the separation membrane comprises specific additives capable of at least partially absorbing electromagnetic irradiation, and wherein said separation membrane is irradiated with an appropriate electromagnetic wave during separation and said specific additives in said separation membrane convert the absorbed electromagnetic energy into heat energy. 2. The process according to claim 1 wherein said specific additives make up less than 10 wt% of said separation membrane. 3. The process according to claims 1 or 2 wherein said specific additives are selected from the group consisting of precious metal, metal or semiconductor nanoparticles. 4. The process according to claim 3 wherein said nanoparticles have a surface plasmon resonance wavelength that is categorized in the infrared, visible or ultraviolet spectrum. 5. The process according to claim 4 wherein said nanoparticles are silver, gold, CdTe or CdSe nanoparticles. 6. The process according to claim 5 wherein said nanoparticles are gold nanoparticles, gold nanorods or gold core-shell nanoparticles. 7. The process according to claims 3 to 6 wherein the electromagnetic wave has a frequency that belongs in the infrared, visible or ultraviolet spectrum, corresponding to the surface plasmon resonance frequency of the nanoparticles. 8. The process according to claims 7 wherein the electromagnetic radiation is from a laser, a continuous wave laser or a LED light source. 9. The process according to claims 1 or 2 wherein said specific additive is graphite or a dielectric molecule. 10. The process according to claim 9 wherein said electromagnetic wave is a microwave. 1 1 . The process according to claims 1 or 2 wherein said specific additive is a dye. 12. The process according to claim 1 1 wherein said electromagnetic wave has a frequency that belongs in the infrared, visible or ultraviolet spectrum. 13. The process according to claims 1 to 12 wherein the separation process is selected from a group consisting of filtration, microfiltration, ultrafiltration, nanofiltration and solvent resistant nanofiltration, reverse osmosis (super filtration), dialysis, electrodialysis, electrolysis (membrane method), gas permeation and pervaporation. 14. The process according to claims 1 to 13 wherein the separation membrane is selected from a group consisting of a densified membrane, a porous membrane, a complex membrane, an phase inversion membrane, a sintered polymer membrane, an oriented polymer membrane, an ion exchange membrane, a polyelectrolyte membrane, a mixed matrix membrane, an inorganic membrane and a hybrid membrane. 15. The process according to claim 14, wherein the separation membrane is a mixed matrix membrane comprising an organic polymer continuous phase and a dispersed phase made up of precious metal, metal or semiconductor nanoparticles. 16. A method to increase the flux of a membrane while retaining membrane selectivity in a membrane based separation process characterised in that the separation membrane comprises specific additives capable of at least partially absorbing electromagnetic irradiation, and wherein said separation membrane is locally heated during separation by electromagnetic irradiation upon conversion of said electromagnetic irradiation into heat energy by said specific additives capable of at least partially absorbing electromagnetic irradiation. 17. A membrane separation apparatus comprising (i) a separation membrane having specific additives incorporated, dissolved or dispersed in the membrane polymer matrix, wherein said additives are capable of at least partially converting electromagnetic irradiation in to heat energy; (ii) a source of electromagnetic irradiation; and (iii) means for allowing said membrane to be irradiated by said source of electromagnetic irradiation during separation. |
FIELD OF THE INVENTION
The present invention relates to novel membrane based separation processes for liquid and gas separations with enhanced membrane fluxes without compromising selectivity. These novel separation processes make use of membranes with additives, that are incorporated in the membrane polymer matrix and that are susceptible to electromagnetic irradiation. Upon irradiation with an electromagnetic source, such as light, said specific additives convert the electromagnetic radiation into heat energy, thus inducing a local heating of the membranes.
BACKGROUND OF THE INVENTION
Membrane separation processes are an increasingly important field in the art of separation science. They can be applied in the separation of components in the gas phase or liquid phase [1-3]. While water purification remains the largest market for membrane applications, a large variety of membranes are now available for a wide range of industrial separation processes in the food, pharmaceutical and (petro)chemical industries as well as for biotechnological and medical applications [1 -3].
Because of the generally low energy use and minimal waste generation, membrane separations can be a sustainable alternative for other separation processes such as solvent extraction and distillation. Other advantages of membranes include the easy fabrication and handling, the straightforward implementation of the membrane separation unit in existing industrial processes and the possibility to tailor membranes for very specific separations.
Membrane separations take place under the influence of a driving force, such as a pressure or concentration gradient, a partial pressure difference or a potential difference. As in any other activated process, mass transfer through a membrane has to overcome a number of resistances, such as the friction between the permeating molecules and the membrane material, so that a higher driving force is only partially translated into a higher flux.
Problems in membrane separations include a usual trade-off between flux and selectivity. Indeed, for instance, in the case of solvent resistant nanofiltration when compounds with a very low molecular weight need to be separated, very dense (polyimide) membranes are required that generally have very low fluxes. Other difficulties are the existence of performance reducing phenomena such as concentration polarisation and fouling [1]. These phenomena add to the resistances that permeating molecules have to overcome while migrating through the pores or free volumes of a membrane. There are different ways of dealing with these problems, such as pre-treating the feed to remove fouling agents [4], adapting the membrane materials [5,6] and membrane surfaces [7,8] or altering the process hydrodynamics [1].
There have been several studies investigating the influence of feed temperature on the membrane flux and rejection, producing variable results [e.g. 9-13]. Depending on the type of membrane process, an increased membrane flux with increasing temperature is attributed to the effect of the temperature increase on the solvent viscosity, sorption and diffusion coefficients and on the polymer chain flexibility and/or the membrane pore size. However, the solute permeability usually also increases with temperature, leading to a decrease in rejection and a decreased membrane selectivity. Moreover, in an industrial process where large volumes are filtrated, it is neither efficient nor realistic to heat up the entire feed.
Electromagnetic radiation takes the form of self-propagating waves that consist of an electric and a magnetic field component that oscillate perpendicular to each other and perpendicular to the direction of energy propagation. Electromagnetic radiation is classified according to the wave frequency: radiowaves, microwaves, terahertz, infrared, visible light, ultraviolet light, X-rays and gamma-rays.
Microwaves (300MHz - 300GHz) can cause certain molecules to absorb energy and heat up, a phenomenon that is exploited in microwave ovens. Polar molecules have an electric dipole moment and will align themselves when brought into an electromagnetic field. When the field is oscillating, as in electromagnetic waves, the molecules will rotate to try and continuously align themselves with it. A microwave field oscillates at high frequencies and the rotating molecules are hindered in their attempt to keep aligning with the field by their surrounding molecules. Through the resulting friction and collisions, energy is passed to adjacent molecules and atoms in the material. The resulting average increase in kinetic energy of all molecules and atoms in the material is measured as the temperature increase of the material. The rate of energy dissipated in a material by dielectric heating per unit of volume, Q, is given by the following formula:
Q = 2πί E 2 ε 0 ε tan6
Where f is the electromagnetic field frequency, E is the field strength, ε 0 is the absolute permittivity of vacuum, ε is the dielectric constant of the material and tan5 is the loss tangent or dissipation factor. Most polymers have a dielectric constant ε of 2 to 4 and a dissipation factor of 0.0001 to 0.01.
Nakai et al. [14,15] and US Pat 6,706,088 treated membranes by microwave irradiations during gas separations. The dipolar cellulose acetate, hydroxypropyl cellulose and poly(methyl metacrylate) membranes showed increased gas permeabilities under microwave irradiation, while permeability of the apolar polystyrene membrane was not affected. In this context, Nakai and coworkers [14] and US Pat 6,706,088 teaches that the observed increased gas permeabilities result from accelerated molecular motion of polar functional groups present in the polar membrane under microwave irradiation (or stated differently, from small structural changes in the membrane) rather than to an increased temperature in the membrane, since these authors observed permeability differences between separations with or without microwave irradiation but performed at the same temperature. Likewise, US Pat 6,706,088 refers to microwave irradiation as an external operation of a molecular valve and stated that the permeability of the gas to be separated upon microwave irradiation is altered according to an increase or decrease in the size of the molecular pores. However, microwave treatment of the membrane to enhance permeability cannot be used on apolar membranes and is also ill suited in combination with polar feeds.
There hence remains a clear need in the art for novel membrane based separation processes with enhanced flux, while at the same time retaining a good selectivity.
Surprisingly, the inventors developed novel membrane separation processes comprising a localized heating of the separation membrane whereby membrane flux is increased without lowering membrane selectivity, which is a highly desired but rarely found effect in membrane technology.
SUMMARY OF THE INVENTION
The present invention relates to novel methods for increasing the membrane permeability but retaining membrane selectivity in a membrane separation process as well as to novel membrane based separation processes comprising the local heating of a membrane comprising specific additives sensitive to electromagnetic irradiation during a separation process by electromagnetic irradiation, typically microwaves, infrared, visible or ultraviolet light. The present invention also relates to a membrane separation apparatus suitable for performing the methods according to the present invention. Thus, a first aspect of the present invention provides a membrane based separation process of substances characterised in that the separation membrane comprises specific additives capable of at least partially absorbing electromagnetic irradiation, and wherein said separation membrane is irradiated with an electromagnetic wave during separation and said specific additives in said separation membrane convert the absorbed
electromagnetic energy into heat energy.
A second aspect of the present invention relates to a method to increase the flux of a membrane while retaining membrane selectivity in a membrane based separation process characterised in that the separation membrane comprises specific additives capable of at least partially absorbing electromagnetic irradiation, and wherein said separation membrane is locally heated during separation by electromagnetic irradiation upon conversion of said electromagnetic irradiation into heat energy by said specific additives capable of at least partially absorbing electromagnetic irradiation.
A third aspect of the present invention provides a membrane separation apparatus comprising (i) a membrane having specific additives incorporated, dissolved or dispersed in the membrane polymer matrix, wherein said additives are capable of at least partially converting electromagnetic irradiation in to heat energy; (ii) a source of electromagnetic irradiation; and (iii) means for allowing said membrane to be irradiated by said source of electromagnetic irradiation during separation.
In the different objects of the present invention, said specific additives preferably make up less than 10 wt%, such as 8 wt%, 6 wt% or 4 wt%, of said separation membrane and include dyes, dielectric molecules, graphite and/or nanoparticles.
In a first preferred embodiment, said specific additives are selected from the group consisting of precious metal, metal or semiconductor nanoparticles. Preferably, said separation membrane is a mixed matrix membrane comprising an organic polymer continuous phase and a dispersed phase made up of precious metal, metal or
semiconductor nanoparticles. More preferably, said nanoparticles have a surface plasmon resonance wavelength that is categorized in the infrared, visible or ultraviolet spectrum. More preferably, said nanoparticles are silver, gold, CdTe or CdSe nanoparticles, most preferably said nanoparticles are gold nanoparticles, gold nanorods or gold core-shell nanoparticles. Preferably, said electromagnetic radiation in this first preferred embodiment of the present invention comprises a wavelength or frequency that belongs in the infrared, visible or ultraviolet spectrum. More preferably, said electromagnetic radiation comprises a wavelength or frequency which is similar or identical to the surface plasmon resonance wavelength or frequency of said nanoparticles. Preferred sources of light irradiation are a laser, a continuous wave laser or a LED light source.
In a second preferred embodiment, said specific additive is graphite or a dielectric molecule, which is dissolved or dispersed in the membrane polymer matrix. Preferably, said electromagnetic wave in this second preferred embodiment of the present invention is a microwave.
In a third preferred embodiment, said specific additive is a dye, which is dissolved in, dispersed in or covalently linked to the membrane polymer matrix. Preferably, said electromagnetic radiation in this third preferred embodiment of the present invention comprises a frequency that belongs in the infrared, visible or ultraviolet spectrum. More preferably, said electromagnetic wave comprises a wavelength or frequency, that is absorbed by said dye.
The different aspects and embodiments of the present invention can be used with all types of membranes, including a densified membrane, a porous membrane, a complex membrane, an phase inversion membrane, a sintered polymer membrane, an oriented polymer membrane, an ion exchange membrane, a polyelectrolyte membrane, a mixed matrix membrane, an inorganic membrane and a hybrid membrane, and having specific additives, such as dyes, dielectric molecules and/or (nano)particles, incorporated into their structure.
The different objects and embodiments of the present invention are essentially applicable to all membrane separation processes, including filtration, microfiltration, ultrafiltration, nanofiltration and solvent resistant nanofiltration, reverse osmosis (super filtration), dialysis, electrodialysis, electrolysis (membrane method), gas permeation and pervaporation.
DETAILED DESCRIPTION
Legends to the figures
Figure 1 is a schematic representation of a membrane filtration apparatus with glass window or lens allowing for the irradiation of a membrane by laser light of the infrared, visible or ultraviolet spectrum during the filtration. It is here exemplified as a dead-end filtration cell.
Figure 2 shows the heating response of gold nanoparticles (GNP) containing cellulose acetate membranes in wetted (A) and dry (B) conditions under continuous wave Argon-ion laser (514 nm) irradiation. Membrane codes as in Table 1. Figure 3 shows the heating response of GNP containing polyimide membranes in wetted (A) and dry (B) conditions under continuous wave Argon-ion laser (514 nm) irradiation. Membrane codes as in Table 2.
Figure 4 shows the isopropanol permeabilities of PI-0 (A), PI-05 (B), PI-1 (C) and PI-2 (D) in filtrations at 5 bar. Continuous wave Argon laser was switched on after 2 samples (1 sample for PI-1 ) and directed onto a membrane at an intensity of 0.25W/cm 2 .
Figure 5 shows the permeabilities of isopropanol (A), ethanol (B) and water (C) of CA-2 membrane in filtrations at 5 bar. Continuous wave Argon laser was switched on after 2 samples and directed onto a membrane at an intensity of 0.2W/cm 2 , and switched off after 2 (5B) or 3 (5A.5C) more samples.
Figure 6 shows the permeances of NF1 , NF2 and NF3 in filtrations of (A) IPA with 35μΜ BTB; (B) IPA with 17μιη RB and (C) ethanol with BTB with laser (white) and without laser (grey) irradiation. The average percentual increase is given between square brackets.
Figure 7 shows the rejections of NF1 , NF2 and NF3 in filtrations of (A) IPA with 35μΜ BTB; (B) IPA with 17μηη RB and (C) ethanol with BTB with laser (white) and without laser (grey) irradiation.
Figure 8 shows the percentual differences in permeance and rejection of ethanol + methyl orange mixtures obtained under laser irradiation for PRE and ISR membranes.
Description
The present invention relates to methods for the membrane based separation of substances, such as liquids or gasses, whereby said separation membrane comprise specific additives that are susceptible to electromagnetic irradiation, and whereby said separation membrane is at least partially irradiated with an electromagnetic beam during said separation of substances and locally heated upon conversion of the electromagnetic irradiation into heat energy by said specific additives that are susceptible to
electromagnetic irradiation. Alternatively, the present invention relates to methods for enhancing or controlling the permeability of a substance through a separation membrane, while retaining the separation selectivity of the membrane in a separation process of substances comprising the step of localized heating of said separation membrane by converting electromagnetic irradiation into heat energy during separation by specific additives that are susceptible to electromagnetic irradiation and that are present in said separation membrane. A local heating of the membrane by electromagnetic irradiation of a membrane that comprises additives capable of at least partially, and preferably strongly, converting said electromagnetic irradiation to heat, is much more efficient than heating up the entire feed or the entire membrane, and yields an increased membrane flux without lowering its selectivity or negatively affecting the rejection:
Obviously, less energy is required for the localized heating of the membrane to obtain a similar increase in membrane permeability than in case of heating the whole feed.
Furthermore, the higher temperature in the membrane will locally decrease the friction between the permeating molecules and the membrane (polymeric) constituents, while the kinetic energy of the permeating molecules is increased, resulting in a higher flux. Indeed, when heating the membrane or rather specific materials therein, this heat is transferred in a membrane from the bulk to the pore walls or free volume, where the friction between the membrane and the permeating molecules will be decreased. The heat transfer results in a better mass transfer and eventually in higher fluxes. However, since the temperature of the feed is less affected, the kinetic energy and permeability of the solute is only affected to a little extent, thus retaining membrane selectivity. As an additional effect, the exothermic sorption of molecules onto the membrane surface or pore walls will be decreased, resulting in reduced fouling.
In addition, the incorporation of specific additives susceptible to electromagnetic irradiation in the membrane allows for a more efficient localized and remote (non-contact) heating of the membrane matrix by irradiation with such suitable electromagnetic radiation, that is independent of or not affected by the properties of the membrane polymers, such as their polarity or their sensitivity to such radiation, or the properties (e.g. polarity) of the feed. Also, the use of additives that strongly absorb electromagnetic irradiation with specific frequency, allows for a lower intensity source of irradiation. Furthermore, since rejection is mostly determined at the membrane surface based on size exclusion, low membrane affinity or electrostatic repulsion, a direct heating of the membrane such as by acting on the membrane polymeric matrix may negatively affect membrane surface properties and hence rejection as well.
In the different embodiments and aspects of the present invention, the separation membrane comprises specific additives capable of at least partially absorbing
electromagnetic irradiation and transforming the absorbed energy into heat energy, thus increasing the temperature of the membrane and altering the permeability of the
substances to be separated, but not the membrane selectivity. Said specific additives include dyes, dielectric molecules, graphite and/or nanoparticles. Typically, said specific additives susceptible to electromagnetic irradiation are present in the membrane at additive-to-polymer ratios (or expressed as % of membrane weight) between 0 and 10 wt%, preferably between 0.1 and 6 wt%, such as between 0.2 and 5 wt%, more preferably, between 0.3 and 4 wt%; such as between 0.4 and 3.5 wt%, most preferably between 0.5 wt% and 3.0 wt%. Said additives can be dissolved or dispersed in the membrane.
Preferably, said specific additives susceptible to electromagnetic irradiation include light absorbing nanoparticles, such as semi-conductor nanoparticles and silver and gold nanoparticles. These additives can be added to the membrane casting solution, or, such as in the case of the metal and noble metal nanoparticles, such as gold nanoparticles, they can be formed in situ during or after membrane preparation e.g. by reduction of gold containing precursor molecules. In case of non-soluble additives, very small nanoparticles are preferred to obtain a good dispersion in the membrane matrix and to avoid the creation of defects in the membrane, particular in the thin separating top layer of the membrane. Preferably, said additives are only present in the micrometer thin membrane top layer. The electromagnetic wave used in the different embodiments of the present invention has a wavelength (or frequency) that falls in the infrared (between about 700 nm extending conventionally to 300 μιτι), visible (between about 400 nm and 700 nm) or ultraviolet (between about 10 nm to 400 nm) spectrum or is a microwave.
The frequency or wavelength band of the irradiation during separation is selected depending on the additives incorporated in the membrane. Preferably, the wavelength or wavelength band of the electromagnetic irradiation used in the different embodiments and objects of the present invention comprises at least one wavelength that is absorbed, preferably strongly absorbed, and subsequently converted into heat energy, by said specific additives sensitive to electromagnetic irradiation. Preferably, the electromagnetic irradiation used in the different embodiments of the present invention is not or only to a limited extent absorbed by the membrane polymers making up the bulk of the membrane matrix or by the molecules making up the (bulk of the) feed solution. Hence, said electromagnetic irradiation has little to no direct effect on the membrane polymers and their structural properties, nor on the temperature of the feed. The separation membrane can be continuously irradiated during separation by said electromagnetic source or the
electromagnetic source can be periodically turned off and on during separation. Said electromagnetic source can be any microwave source or any light source. Preferred light sources include but are not limited to lasers, such as continuous wave or one-wavelength lasers, appropriately designed lamps or light emitting diodes (LEDs). Preferably, the separation membrane is irradiated by solar radiation or sunlight. This may be of interest for membranes used for desalting purposes. One embodiment of the membrane based separation process according to the present invention relates to microwave technology combined with specific additives, dispersed or dissolved in the membrane for a localized heating of the membrane independent of the polarity of the membrane. Typically, the frequencies of microwaves that can be used include but are not limited to frequency ranges that are allotted as Industrial, Scientific and Medical frequencies, such as the wavelength frequencies with center frequency at 915, 2450 and 5800 MHz. Preferred additives are dielectric molecules, such as water, or graphite. Typically, said graphite particles are still small enough to avoid inducing defects in the membrane structure.
Instead of using microwave technology, photothermal effects can also be used to locally heat a membrane. Indeed, the polarity of the feed has little if any impact on light irradiation and photothermal heating of membranes comprising light sensitive molecules. Also, light energy is largely independent of membrane matrix polymer composition and polarity, instead relying on the presence of the specific light sensitive additives. Furthermore, light can be guided to the membranes with glass fibre technology, thus easily focusing the light on the membrane. When using the photothermal effect to influence membrane filtrations, temperatures are required that are low enough to avoid destruction of the polymer matrix but high enough to induce a significant flux increase.
Thus, a preferred embodiment of the present invention relates to methods for the membrane based separation of substances, such as liquids or gasses, whereby said separation membrane comprises specific additives absorbing infrared, visible or ultraviolet light and whereby said separation membrane is irradiated with the appropriate light source during the separation of substances and locally heated upon conversion of the light into heat energy by said specific additives absorbing infrared, visible or ultraviolet light.
Preferably, a coloured membrane or a membrane containing a dye can absorb light energy and thus in principle be heated. In principle, any light absorbing dye, dissolved in or covalently linked to the membrane matrix polymer, can be used to locally heat the membrane.
Another preferred embodiment of the present invention relates to methods for the membrane based separation of substances, such as liquids or gasses, whereby said separation membrane is a composite separation membrane comprising light absorbing nanoparticles, such as semi-conductor nanoparticles and silver and gold nanoparticles, dispersed in the membrane polymer matrix, and whereby said separation membrane is irradiated with an appropriate light source during the separation of substances and thus locally heated upon conversion of the light into heat energy by said light absorbing nanoparticles. At certain wavelengths, semi-conductor nanoparticles, such as the .
chalcogenides ll-VI materials CdTe and CdSe, and silver and gold nanoparticles strongly absorb light and turn it into heat very efficiently, thus allowing the localized heating of the separation membrane during separation. Typically, said nanoparticles are still small enough to avoid inducing defects in the membrane structure and have dimensions in the order of 0.1 to 50 nm, preferably between about 0.25 to 25 nm or between about 0.5 to 20 nm, more preferably between 0.5 and 15 or between 0.5 and 10 nm, most preferably between 1 and 8 nm or between 1 and 6 nm.
In a more preferred embodiment of the present invention, said separation membrane comprises gold nanoparticles (GNPs). GNPs have several advantages over organic dyes and over other particles that have a surface plasmon resonance, such as semi-conductor nanoparticles (CdTe, CdSe) and metal nanoparticles (MNPs). GNPs have a higher photothermal effect than organic dyes and they very efficiently transform light energy into heat energy. Furthermore, unlike organic dyes, GNPs have a high photothermal stability. The semi-conductor nanoparticles have a much smaller heat generation rate. Although silver nanoparticles can generate heat from light about ten times stronger than GNPs [16], unlike Ag and other metal nanoparticles, GNPs are inert and insensitive to photobleaching, which is very important for the durability of the membrane. Moreover, the plasmon wavelength of GNPs can be easily tuned within the visible spectrum, using e.g. spherical GNPs and gold nanorods, or shifted into the infrared spectrum (e.g. gold nanoshells)
[16,17]. Selective heating of the gold nanoparticles (GNPs) built into separation
membranes can be achieved by irradiating the membrane with visible to infrared light. Gold nanoparticles, nanorods and nanoshells have a very pronounced surface plasmon resonance, which means they absorb light of certain wavelengths (tunable from ultraviolet or visible to infrared light) and they very effectively transform this electromagnetic energy into heat. In a membrane, this heat is transferred from the bulk to the pore walls or free volume, where the friction between the membrane and the permeating molecules will be decreased. The heat transfer results in a better mass transfer and thus in higher fluxes.
Thus, a more preferred embodiment of the present invention relates to methods for the membrane based separation of substances, such as liquids or gasses, whereby said separation membrane comprises gold nanoparticles, such as spherical GNPs, gold nanorods or gold core nanoshells, and whereby said separation membrane is irradiated with an appropriate light source during the separation of substances and thus locally heated upon conversion of the light into heat energy by said gold nanoparticles.
Alternatively, the present invention relates to methods for enhancing or controlling the permeability of a substance through a separation membrane, while retaining the separation selectivity of the membrane in a separation process of substances comprising the step of localized heating of said separation membrane by converting light irradiation into heat energy during separation by gold nanoparticles that are present in said separation membrane.
APPLICATIONS
The improved membrane based separation processes of substances according to the present invention comprising the local heating of the separation membrane by conversion of electromagnetic irradiation into heat energy during separation by specific additives, that are incorporated, dissolved or dispersed in the membrane polymer matrix and that are susceptible to such electromagnetic irradiation is applicable in a wide range of membrane separation processes, such as microfiltration, ultrafiltration, nanofiltration and solvent resistant nanofiltration, reverse osmosis, dialysis, electrodialysis, electrolysis (membrane method), gas permeation and pervaporation.
The improved membrane based separation processes of substances comprising the local heating of the separation membrane by conversion of electromagnetic irradiation into heat energy during separation by specific additives, incorporated in the membrane, that are susceptible to such electromagnetic irradiation is applicable in a wide range of membranes, such as densified membranes, porous membranes, complex membranes, phase inversion membranes, sintered membranes, oriented polymer membranes, ion exchange
membranes, polyelectrolyte membranes, mixed matrix membranes, inorganic membranes, hybrid membranes.
In this text, the term "phase inversion" refers to the controlled transformation of a thermodynamically stable membrane solution to a solid phase by liquid-liquid-demixing. It can be carried out by immersion of the cast membrane in a bath containing a non-solvent for the polymer, possibly after a solvent evaporation step during which a certain phase inversion can take place already or not; or by contacting the cast membrane with a vapour phase containing a non-solvent for the polymer or by thermal precipitation.
As used herein, a "mixed matrix membrane" refers to a separation membrane prepared of an inorganic or organic dispersed phase interspersed in a continuous organic matrix, wherein the dispersed phase may consist of molecular sieves, adsorbents, nanoparticles, carbon nanotubes, or polymers.
In a preferred embodiment of the present invention, the improved membrane based separation processes of substances as set out above comprises a mixed matrix separation membrane or a composite separation membrane comprising an organic polymer continuous phase and a dispersed phase made up of precious metal, metal or
semiconductor nanoparticles, more preferably comprising gold nanoparticles.
Preferably, the membrane based separation process according to the different
embodiments of the present invention is a membrane based separation process of solvents that have low effusivity or thermal conductivity values and using relatively dense membranes.
Another aspect of the present invention further refers to a membrane filtration or separation apparatus comprising: (i) a membrane having specific additives incorporated, dissolved or dispersed in the membrane polymer matrix, wherein said additives are capable of at least partially converting electromagnetic irradiation in to heat energy; (ii) a source of electromagnetic irradiation; (iii) means for allowing said membrane to be irradiated by said source of electromagnetic irradiation during separation and (iv) feed and pressure inlets and permeate outlet.
Said source of electromagnetic irradiation can be any microwave source or any light source. Preferred light sources include but are not limited to lasers, such as continuous wave or one-wavelength lasers, appropriately designed lamps or LED lights.
Said means for allowing said membrane to be irradiated by said source of electromagnetic irradiation during separation includes but is not limited to one or more waveguides, transparent walls or windows in a wall of the apparatus (e.g. made up of glass or quartz or other transparent materials known in the art), lenses or mirrors. In addition, optical fibres above or beneath the membrane or integrated inside the membrane module, the membrane support or the spacers, can be used to conduct electromagnetic irradiation, particular light irradiation to the membrane surface. Optionally, membrane irradiation during separation occurs as bottom-up irradiation, sideways irradiation or laser scanning.
Optionally, probes or sensors can be incorporated in the module or membrane
filtration/separation apparatus of this object of the present invention to enable online and in situ temperature measurements of the membrane and solvent and monitoring of the separation process.
Said membrane filtration/separation apparatus can be any filtration/separation apparatus, including cross flow systems, that are adapted for irradiation with electromagnetic irradiation during separation.
The present invention is further illustrated in the following non-limiting Examples. Example 1. Cellulose acetate membranes
Preparation
Cellulose acetate (CA) casting solutions with gold to polymer weight ratios varying between 0, 0.5, 1.0 and 2.0 wt% were prepared according to Table 1.
Table 1. Compositions of CA membranes
The solutions were stirred until homogeneous, left to stand until no air bubbles were present and then cast onto a porous non-woven polypropylene/polyethylene support (Novatexx), saturated with a 5:1 acetone:formamide solution, by use of an automated casting knife set at 250μηι thickness. The membranes were left under a hood for 10 minutes to evaporate a part of the solvent and were immersed in a water bath at room temperature until solidified. After the membrane synthesis was complete, the membranes were immediately immersed into a NaBH4 bath to reduce the gold to form GNPs inside the solid polymer matrix. The membrane colour turned from a light yellow into the typical dark red colour associated with GNPs. The synthesis method is based on Huang et al. [19],
Characterisation
Scanning Electron Microscopy of the cross-sections of the different CA membranes showed that the formation of the (low levels of) GNPs inside the membrane after membrane synthesis had little effect on the membrane structure.
Transmission electron microscopy (TEM) of CA membrane cross-sections indicated that the GNPs are very well dispersed in the membrane matrix and have a mean particle size of approximately 5 nm. Some larger particles with a size up to 20 nm are present, particularly in the CA-2 membrane. These particles are still small enough to avoid inducing defects in the membrane structure. However, they may influence the photothermal behavior by absorbing light at slightly higher wavelengths compared to the 5 nm particles.
In the diffuse reflectance spectroscopy (DRS) spectrum of the membrane surface, all three GNP containing membranes show a maximum absorbance at λ=536 nm. After redissolving the membranes in acetone, the GNP absorbance maximum in the UV-visible spectrum for CA-1 and CA-2 were both measured at λ=532 nm. The absorption maximum for CA05 could not be measured accurately due to the very low concentration of GNPs in the dilute membrane solution. The SPR wavelengths obtained by UWIS are slightly lower than the maximum indicated by DRS. Inside the membrane, the GNPs are trapped at closer distances in a solid matrix, which may increase their SPR wavelength. Teranishi et a/.[20] have reported GNPs with a size of >2 nm to have an absorption maximum at 530 nm. This is in accordance with the GNP size as determined here by TEM.
Photothermal Effect
To determine the degree of heating that can be obtained in the CA membranes containing GNPs by use of irradiation with a continuous wave Argon-ion laser beam (514 nm), flat pieces of membranes (both in the dry and wetted (H20) state) were heated by aiming the laser beam at the membrane surface. The temperature of the membrane surface was detected by an infrared thermometer (Elcometer). The laser intensity was varied by using a lens at different distances from the membrane surface. The heating response is given in Figure 2A and 2B.
The temperature increased immediately upon irradiation when wetted as well as in dry state. The temperature of CA-2 for instance increased to over 80°C when dry at a laser intensity of 0.35 W/cm 2 and to over 40°C when wetted at a laser intensity of 0.5 W/cm 2 .
Overall, the temperature of GNP containing membranes increases close to linearly with the laser intensity. The temperature of the unfilled CA membranes (CA-0) remained below 21 °C. For laser intensities of 0.5 W/cm 2 and higher, temperature measurements by the infrared distance thermometer became statistically inaccurate, due to the small size of the laser beam. Burn spots were induced by the laser in the GNP containing dry membranes starting at an intensity of 2.5 W/cm 2 . Example 2. Polyimide membranes
Preparation
PVP-protected GNPs were prepared in dimethylacetamide (DMA) according to the procedure described by Mertens et al. [18]. Solutions of HAuCI 4 .3H 2 0 (0.05, 0.1 and 0.2 mmol) and an amount of PVP (MW of 10000g/mol; molar ratio monomeric units of PVP : gold = 2) were prepared in DMA (6g). Then, a freshly prepared NaBH 4 solution in DMA (2g) was added under vigorous stirring (molar ratio NaBH :gold = 5) and immediately a colour change from yellow to dark red occurred in the solution, indicating the reduction of gold into nanoparticles. The solution was characterized by a Perkin Elmer UV-VIS spectrophotometer and the typical dark red colour showed as a large peak at 530 nm, corresponding to the plasmon absorbance band of the GNPs.
PI (2g) was added to the GNP solutions in DMA, resulting in three casting solutions with different weight ratios gold:PI (0.5, 1.0 and 2.0%) with compositions given in Table 2 (Pl- 05, PI-1 , PI-2). A similar polymer solution PI (2g) in pure DMA (8g) was prepared as a reference (PI-0). Another solution was prepared containing PI and a red dye, disperse red 1 1 , similar as PI-1 but with 0.2 wt% dye instead of GNP (PI-DR). The solutions were stirred at room temperature until a homogeneous mixture was obtained. The solutions were then allowed to stand until air bubbles had disappeared and were cast onto a non-woven support material (Novatex) that had been saturated with DMA. An automated casting knife (250 μιτι slid) was used and the resulting polymer films were immediately immersed into a water bath to obtain solid membranes. The reference membrane was yellow, the membranes containing GNPs were light pink to red-brown in colour. All membranes were kept in water for no more than 1 day and then in IPA until further use. The nanofiltration membranes NF1 , NF2 and NF3 were prepared similarly, containing 1wt% of GNPs and membrane compositions are listed in Table 2.
Table 2. Compositions of PI membranes
PI-2 Matrimid 20 79.6 0 0 0 0.4
NF1 P84 22 44.56 0 33 0 0.44
NF2 Matrimid 20 47.6 0 0 32 0.4
NF3 P84 22 0 20.56 58 0 0.44
Characterisation
A strong peak at 530nm, corresponding to the surface plasmon resonance (SPR) of the GNPs in the DMA solution, was observed in the UV-VIS spectrum. This is in accordance with Teranishi et al. [20], who have indicated that, in that case, such GNPs have a size of >2nm. This is in line with the TEM analysis of the GNP particle sizes in the DMA solution, showing a mean particle size of approximately 5 nm.
DRS spectra were taken of the GNP containing membrane. Although the intensity of the peaks is low, it is clear that compared to the GNP solution in DMA, the SPR band has red- shifted and broadened considerably. This is most probably due to the solid environment capturing the GNPs at closer distances from each other compared to in a dissolved state. The SPR band of PI1 is found at the smallest wavelength (537nm) which may indicate a better size distribution compared to PI05 (557nm) and PI2 (544nm).
TEM pictures of the membrane cross-sections clearly visualized the rise in concentration of GNPs in the membranes (PI05<PI1 <PI2). The mean size of the GNPs has not significantly changed compared to their size in the original solution (~2nm), but there is a clear difference in dispersion between the membranes. Compared to PI2, the GNPs in PI1 have a better size distribution and are well dispersed, which confirms the DRS results where PI1 has the lower SPR wavelength. In PI2, at the highest GNP/polymer ratio, there are clusters of GNPs visible, with cluster sizes increasing up to 20 nm. Overall, the membrane synthesis procedure had little effect on the GNP size distribution and allows for a good dispersion of the GNPs in the membrane matrix. SEM pictures were taken of the membrane cross-sections as well as a close up of the membrane top layer to assess the effect of PVP-stabilized GNP incorporation on the membrane structure (not shown). For Pl- 05 and PI-1 , the difference in membrane structure is small, with similar sublayer structures, a high occurrence of finger-like macrovoids and a densified toplayer. In PI-2, however, the macrovoids are less pronounced, the substructure matrix is more porous and a dense skin layer is no longer visible. Photothermal Effect
The membranes were irradiated by laser light as in Example 1. Wet-state and dry-state membrane slabs were attached to a non-conductive surface. A continuous green argon laser beam (514nm) was directed by mirrors to fall perpendicularly onto the membrane surface. To vary the laser intensity, the power was adjusted (0.4, 0.7 and 1.0W) and the laser beam was passed through a lens held at different distances from the membrane surface. The intensity was measured as the laser power divided by the illuminated surface. An infrared thermometer (Elcometer) was used to determine the temperature of the illuminated membrane. The temperature profile of the membranes is given in Figure 3. The pure PI membranes reacted very slowly to the increase of laser intensity in their dry state and did not show any significant temperature rise at all in their wet state. For the GNP containing membranes, however, the temperature increased immediately upon irradiation and rose almost linearly with the increasing laser intensity both when wet and dry. The temperature of PI-2 increased to over 70°C when dry and over 40°C when wet at a laser intensity of 0.5W/cm 2 .
The temperature increases with the gold concentration in the membrane. The dry GNP containing PI membranes reached significantly higher temperatures than the wet membranes. In solvents with lower heat capacities than water, this effect will presumably be reduced. Temperature increases close to linearly with the laser intensity. For laser intensities higher than 0.5 W/cm 2 , temperature measurements by the infrared distance thermometer became statistically inaccurate, due to the small size of the laser beam. Burn spots were induced by the laser in the dry membranes with GNPs starting at an intensity of 2.5W/m 2 while PI0 did not burn until an intensity of 13W/m 2 was reached. In the wet membranes, burn spots were induced at 2.5, 13 and 23 W/m 2 for PI2, PI1 and PI05, respectively, but the wetted PI0 did not burn at all.
Example 3. Mixed matrix membranes
Polydimethylsiloxane (PDMS) in toluene, polyvinyl alcohol (PVA) in water and PI (PI) in dimethylformamide solutions were prepared with dispersed graphite particles and a pure PI solution without graphite particles. The PDMS and PVA solutions were spread onto a glass plate and the solvent was evaporated overnight to obtain dense membrane slabs. The PI solutions were cast onto a non-woven fabric by use of an automated casting knife set at 250μηι thickness, immersed into a water bath at room temperature and dried overnight. Pieces of membrane were heated in a microwave reactor vessel containing toluene by use of a CEM discovery microwave reactor oven. The microwave reactor with fiberoptic temperature sensor was set to reach 100°C at a maximum power of 150W within a time frame of 5 minutes. The influence of the presence of a membrane on the heating process was investigated.
Pure toluene without presence of a membrane reached 95°C after 5 minutes under the set conditions.
The presence of a pure PI membrane in the reaction vessel, had no notable influence on the heating of toluene. The graphite containing PI slab however induced an increase in the heating rate of toluene, reaching 100°C within 2.5 minutes. When the PI membrane was attached to the fiberoptic sensor and was prevented from contact with the toluene, however, a temperature of 100° was measured within 2.5 minutes, showing that the membrane did heat when exposed to microwaves. The temperature of the graphite containing PI slab could then reach 100°C within 1.5 minutes.
A graphite containing PVA slab induced a very fast heating rate of toluene, reaching 100°C within 1 minute. When the slab is attached to the fiberoptic and prevented from contact with toluene, it even burns up after only 15s. The graphite containing PDMS slab however had no influence on the toluene heating rate. With the PDMS membrane attached to the fiberoptic, a temperature of 100°C was reached within 3 minutes.
The experiment shows that mixed matrix membranes as well as pure polymer membranes can be locally heated by microwaves. A device as built by Nakai et al. [14] can be used to exploit the heating of membranes under microwaves with the purpose of inducing flux increases in membrane separation processes.
Example 4. Filtration with GNP containing PI membranes
Flux Enhancement
The membranes described in Example 2 were used in a liquid filtration under laser light irradiation. Dead-end membrane separations were carried out in the specially made glass filtration cell (Figure 1 ). A transparent glass window was built in the top to allow a diffracted laser beam to pass and illuminate a part (0.0004m 2 ) of the active membrane surface
(0.001736m 2 ). Before each filtration, the membranes were immersed in isopropanol for at least one day. Pure solvent dead-end filtrations were then carried out at 5 bar with ethanol and isopropanol and with and without laser irradiation. The laser was set at different intensities. Permeances were calculated as the amount of solvent (V) that passed through the membrane per unit of time (t), membrane surface (A) and applied pressure (ΔΡ) so that:
Permeance = V ■ t ~ 1 ■ 1 ■ ΔΡ 1
Pure solvent dead-end filiations were carried out with PIO, PI05, PI 1 and PI2 with isopropanol and ethanol as solvents. Figure 4 (A-D) show the filtration data (isopropanol, average of 3 repetitions) for PI-0, PI-05, PI-1 and PI-2 respectively. The laser (0.2W/cm 2 ) was switched on after the first two samples were taken (one sample for PI-1 ), and was switched off after three more samples. At this laser power density, 30% of the active membrane surface was effectively irradiated by the laser beam. For each gold
concentration, three pieces of membrane were filtrated. It is clear that all GNP containing membranes show an increase in permeance after the point where the laser was switched on. The mean percentual difference between the average permeance without laser for an individual membrane compared to the highest permeance reached under laser irradiation for this same membrane was 15, 31 and 20% for PI05, PI1 and PI2, respectively. The pure PI membrane however shows no such flux increase.
It should be noted that in the current set-up, the laser intensities were limited to 0.2W/cm 2 and no more than 30% of the active membrane surface was being irradiated by the beam. Generally, a larger effect of the laser on the membrane flux can be expected when higher laser intensities are used and when the entire membrane surface is irradiated. Assuming that the flux through the non irradiated part of the membrane is unchanged, the permeance increase can be recalculated from the data to take into account only the irradiated part of the membrane. For the membranes PI05, PI1 and PI2, the permeance increase for the irradiated part was calculated as 76%, 168% and 90%, respectively. This shows that the efficiency of photothermal heating will be considerably enhanced by irradiating the entire membrane surface.
Decreasing the laser intensity to 0.1W/cm 2 by reducing the laser power (0.5W) so that the irradiated surface remained the same, decreased the flux increase to a similar extent. The increase in flux is thus rising with the laser intensity as well. Again, this flux behavior corresponds well with the heating behavior of the membranes.
Similar results were obtained for pure solvent filtrations with ethanol.
Rejections To study the effect of laser heating on solute rejection, three different GNP containing NF membranes were used in the dead-end filtration setup as described above with dilute bromothymol blue (BTB, 35μΜ) solutions in ethanol and isopropanol and with a Rose Bengal (RB, 17 μΜ) solution in isopropanol. Rejections were calculated as the percentage of the feed concentration that was retained:
Rejection = 100 ■ [ 1 - ( c p ■ cf 1 ) ]
Wherein c p is the permeate concentration and Cf is the feed concentration.
The results are given in Figures 6 and 7.
From these data, it becomes clear that the effect of the laser irradiation is always stronger for low permeances, i.e. with denser membranes or slower permeating solvents. In the IPA filtrations of NF3, i.e. the membrane with the lowest permeance, percentual increases were measured up to 171 % (BTB) and 404% (RB). Since the ethanol permeance was higher, the percentual increase for NF3 was still significant but reduced to 42%. Similar effects were noted for NF1 and NF2. Presumably, a higher flow rate through the membrane reduced the energetic efficiency of the localized heating effect, which is probably due to a higher energy loss by convection.
No significant differences were noted for the rejections with and without laser irradiation (Figure 7). Indeed, as local heating is supposed to lower the friction between the permeating molecules and the surrounding polymer chains unselectively (for both solvent and solute molecules), rejection - which is mostly determined at the membrane surface based on size exclusion, preferential adsorption or electrostatic repulsion - should not undergo too much influence.
Example 5. Filtration with GNP containing CA membranes
Flux Enhancement
The CA-2 membrane described in Example 1 was used in dead-end filtrations of water, ethanol and isopropanol while being irradiated by laser light. Isopropanol (IPA) and ethanol were chosen as solvents to explicitly prove the concept because of their relevance in solvent resistant nanofiltration and their lower heat capacities (2.6 and 2.4 kJ kg "1 K "1 ) compared to water.
The results are summarized in figure 5A (isopropanol), 5B (ethanol) and 5C (water). After two samples had been taken, the laser was switched on. In each case, an increase in flux was noted until the laser was switched off again after three more samples (or two in B). The laser (0.25 W/cm 2 ) was switched on after the first two samples were taken, and was switched off again after three more samples. It should be mentioned that at this laser power density, only 45% of the active membrane surface was effectively irradiated by the laser beam under the current experimental conditions (see experimental section). As shown in Figure 5C, an average percentual increase in the water permeance of 15% was measured when the laser was switched on. The results for ethanol and IPA filtrations are summarized in Figure 5B and 5A, respectively. In both cases, average percentual increases in permeability under laser irradiation were noted of 400%. A while after the laser was turned off, the permeances dropped back to the equilibrium value reached before the laser was switched on. This indicated that the effect of laser irradiation was due to the (reversible) heating and not to any kind of membrane modification or pore formation.
The thermal conductivity and the heat capacity of a solvent can be combined in the effusivity parameter e, indicating the capacity of the solvent to exchange heat energy with its environment, as used for heat exchangers. When taking this parameter into account and without being bound by theory, it seems that the difference in effusivity largely explains why the permeance increase is substantially higher for the alcohols than for water. This is again a strong indication that the permeance increase under laser irradiation is due to thermal effects only.
When comparing the membrane fluxes obtained under laser irradiation with those obtained by increasing feed temperature, it can be estimated by extrapolation that the permeance obtained via laser heating for the same feed and membrane, is reached at a feed temperature of 37°C in the comparison experiment. For the duration of the filtration, the theoretical energy input necessary to heat the ethanol feed from 20°C to 37°C and maintain it at this temperature was calculated to be 0.490Wh. When using laser heating, the total energy input during radiation was 0.183 Wh, indicating that, to create a similar flux, it was about 2.5 times more efficient to use the laser than to heat the feed. It is expected that the laser heating will be even more efficient when the entire membrane surface is irradiated, since the average laser induced permeance should be considerably higher in such a case. It should be stated that for both systems, only the energy input into the membrane or into the feed were taken into account.
Since the flowing solvent will remove heat from the membrane by convection, it is expected that this effect will be dependent on the flow rate of the solvent through the membrane. To investigate this, filtrations were run at different pressures. Overall, a higher energy input is necessary to create an identical flux increase when the filtration pressure is increased. Thus, the convective energy loss is more important at the higher pressures, resulting in an increased energy input to obtain the same permeance increases.
Rejections
The filtrations of CA-2 were repeated with a solution of 35μΜ bromothymol blue in ethanol to assess the effect of the photothermal heating of the membrane on the solute rejection. Averaged over three measurements a percentual increase of permeance was measured of 167%. The average rejection did not change statistically, varying from 80.5% (± 25 1.3%) to 82.5% (± 2.8%). It can be concluded that there is no change in rejection when the laser is switched on. Indeed, as local heating is supposed to lower the friction between the permeating molecule and the surrounding polymer chains, rejection (which is mostly determined at the membrane surface based on size exclusion, low membrane affinity or electrostatic repulsion) should not undergo too much influence. To increase fluxes of a given membrane without lowering its selectivity is a highly desired but rarely found effect in membrane technology.
Example 6. GNP containing PI membranes: in situ chemical reduction vs
incorporation of preformed GNPs
Preparation
Incorporation of pre-synthesized GNPs to form GNP-PI membranes ("PRE-membranes") was according to Example 2. PI was added to a solution of preformed polyvinylpyrrolidon protected GNPs in DMA resulting in four casting solutions with different gold - PI ratios (1.0; 2.0; 3.0 & 4.0 wt%). Membrane casting and phase inversion was as in Example 2.
Formation of GNP-PI membranes using an in situ chemical reduction ("ISR-membranes") was similar to the method described in Example 1 for CA membranes. In short,
HAuCI .3H 2 0 was added to a polyimide solution prepared in a mixture of DMA and THF to obtain casting solutions with gold to polymer weight ratios of 1.0, 2.0, 3.0 and 4.0 wt%. A similar polymer solution PI was prepared without HAuCI 4 .3H 2 0, as a reference. Following membrane casting, a short solvent evaporation step (30s) and phase inversion into a water coagulation bath, the membranes were moved immediately into a solution of NaBH4 in water to reduce the gold to nanoparticles, upon which the membrane color turned from yellow to dark red. The membranes were further kept in IPA and IPA:glycerol and then dried. Characterisation
In the ISR-membranes, the GNPs are formed inside the solid membrane matrix, wherein the polymer itself acts as a stabilizer. In the PRE-membranes, PVP-stabilized GNPs are present in the polymer solution before the membrane is cast and synthesized by phase inversion. As such, the membrane synthesis procedure may have an influence on the GNP size and distribution, but on the other hand the GNPs may have an influence on the membrane structure as well.
SEM of cross-sections of the ISR- and PRE-membranes showed clear differences. In the case of ISR-membranes, a clear dense skin layer is visible on the membrane with cross- sections with increasing roughness with increasing gold content. The synthesis of larger amounts of GNPs in a solid membrane can cause a disruption of the membrane structure, particularly in the top layer. In the case of the PRE-membranes, cross-sections are more smooth than for the ISR-membranes. A dense skin layer is only visible for the PRE- membranes with lower gold content. PRE-membranes with higher gold content were more porous, presumably due to the presence of higher PVP levels in the casting solution.
DRS characterization: For the ISR membranes, regardless of the gold concentration, the size and dispersion of GNPs in the membrane top layer seem to a large extent very similar. In contrast, for the PRE-membranes, DRS and the UWIS wavelength are clearly increasing with increasing gold concentration. This indicates that both in the top and sublayer, larger GNPs may have been formed at higher gold concentrations.
TEM data is in line with the DSR and UWIS data. The mean GNP particle size is about 3 - 5 nm, irrespective of the preparation method. However, in the ISR membranes, hardly any aggregation of GNPs is visible either in the top layer or in the substructure, at any concentration of gold (GNP particle size ranging between 1 and 8 nm for the lower GNP concentrations and ranging between 2 and 12 nm for the higher GNP concentrations). In the PRE membranes, clustering of GNPs occurs both in the skin layer and the
substructure, particularly at the higher gold concentrations (GNP particle size similar to the ISR membranes, but with clusters up to ca. 20 nm). The in-situ synthesis methods thus leads to a better dispersion and less aggregation of the GNPs compared to the use of pre- synthesized GNPs. The TEM pictures further indicated that higher gold contents in the membranes lead to broader particles size distributions but that the mean particle size remains constant up to 3wt% gold, regardless the incorporation method. To study the effect of light-induced local photothermal heating of membrane on filtration behavior, dead-end filtrations of methyl orange in ethanol were performed for both PRE- and ISR-membranes under laser irradiation, according to the setup described above. The laser was set at an intensity of 0.2W/cm 2 . The performance under laser irradiation was compared to the original performance of the membranes and the results are visualised in Figure 8. For both the ISR and the PRE membranes, the percentual difference in permeance induced by plasmonic heating increases at higher gold contents. Although there is a clear difference in the membrane behavior, the GNP distribution and GNP aggregation for PRE and ISR membranes, the absolute differences in permeance are similar for both methods. The rejection is in neither case strongly affected by the laser irradiation, the differences fall within the expected experimental error.
CONCLUSIONS
Membranes comprising specific additives or materials that are susceptible to
electromagnetic irradiation can be heated locally by irradiating the membrane with an electromagnetic source such as light or microwaves and subsequent conversion of the electromagnetic energy into heat energy, thus inducing a flux increase. The effect is comparable to an increased process temperature, an often applied strategy in which case the entire feed has to be heated to a certain temperature. However, heating the membrane locally, according to the methods of the present invention, is energetically and economically more efficient, and retains membrane selectivity. The membrane heating effect is then translated into improved membrane fluxes, effected by presumably an increased kinetic energy of the permeating molecules, a decreased viscosity and a reduced friction with the membrane pore walls and/or free volume. Overall, this leads to an increase in flux and possibly a reduction of fouling as well.
Other embodiments of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.
REFERENCES
1. Mulder, M., Basic Principles of Membrane Technology - Second Edition. 1996,
Dordrecht: Kluwer Academic Publishers. 564. 2. Pabby, A.K., S.S.H. Rizvi, and A.M. Sastre, Handbook of Membrane Separations:
chemical, pharmaceutical, food and biotechnological applications. 2008: Taylor & Francis Group, LCC. 1 164.
3. Noble, R.D. and S.A. Stern, Membrane separations technology : principles and applications. Third Edition ed. Membrane Science and Technology Series, 2. 1995,
Amsterdam: Elsevier Science B.V. 718.
4. Speth, T.F., A.M. Gusses, and R.S. Summers, Evaluation of nanofiltration pretreatments for flux loss control. Desalination, 2000. 130(1): p. 31-44.
5. Bartels, C, et al., New generation of low fouling nanofiltration membranes. Desalination, 2008. 221 (1-3): p. 158-167.
6. Wilf, M. and S. Alt, Application of low fouling RO membrane elements for reclamation of municipal wastewater. Desalination, 2000. 132(1-3): p. 1 1 -19.
7. Kochkodan, V.M., et al., Effect of the surface modification of polymer membranes on their microbiological fouling. Colloid Journal, 2006. 68(3): p. 267-273.
8. Kull, K.R., M.L. Steen, and E.R. Fisher, Surface modification with nitrogen-containing plasmas to produce hydrophilic, low-fouling membranes. Journal of Membrane Science, 2005. 246(2): p. 203-215.
9. Machado, D.R., D. Hasson, and R. Semiat, Effect of solvent properties on permeate flow through nanofiltration membranes. Part I: investigation of parameters affecting solvent flux. Journal of Membrane Science, 1999. 163(1): p. 93-102.
10. Tsuru, T., et al., Temperature effect on transport performance by inorganic
nanofiltration membranes. Aiche Journal, 2000. 46(3): p. 565-574.
1 1. Ratanatamskul, C, et al., Effect of operating conditions on rejection of anionic pollutants in the water environment by nanofiltration especially in very low pressure range. Water Science and Technology, 1996. 34(9): p. 149-156.
12. Scarpello, J.T., et al., The separation of homogeneous organometallic catalysts using solvent resistant nanofiltration. Journal of Membrane Science, 2002. 203(1-2): p. 71-85.
13. J.A. Zhou, H. Zhou, H.J. Fan, F.F. Yang, J.X. Yuan, R.W. Liu, Application of Thermal Sensitive Polyurethane Membranes for Selective Separation of Substances with Different Sizes, Acta Polymerica Sinica. 2010, 308-315. 14. Nakai, Y., H. Yoshimizu, and Y. Tsujita, Enhanced gas permeability of cellulose acetate membranes under microwave irradiation. Journal of Membrane Science, 2005. 256(1-2): p. 72-77.
15. Nakai, Y., H. Yoshimizu, and Y. Tsujita, Enhancement of gas permeability in HPC, CTA and PMMA under microwave irradiation. Polymer Journal, 2006. 38(4): p. 376-380.
16. Govorov, A.O. and H.H. Richardson, Generating heat with metal nanoparticles. Nano Today, 2007. 2(1 ): p. 30-38.
17. Daniel, M.C. and D. Astruc, Gold nanoparticles: Assembly, supramolecular chemistry, quantum-size-related properties, and applications toward biology, catalysis, and nanotechnology. Chemical Reviews, 2004. 104(1 ): p. 293-346.
18. Mertens, P.G.N., et al., Catalytic oxidation of 1,2-diols to alpha-hydroxy-carboxylates with stabilized gold nanocolloids combined with a membrane-based catalyst separation. Catalysis Letters, 2005. 102(1-2): p. 57-61.
19. Huang, Y.J., et al., Semipermeable membrane embodying noble metal nanoparticles and its electrochemical behaviors. Journal of Electroanalytical Chemistry, 2005. 579(2): p.
277-282.
20. T. Teranishi, I. Kiyokawa and M. Miyake, Advanced Materials, 105 1998, 10, 596-599.