The present invention relates to a membrane for membrane distillation, a membrane for forward osmosis, and a process for treating a high salinity feed. More specifically, the present invention is concerned with a membrane for membrane distillation comprising a microporous surface-modified mat of electrospun nanocomposite nanofibers; a thin film composite membrane for forward osmosis comprising a mat of electrospun nanocomposite nanofibers as a support layer and a nanocomposite rejection layer; and a process for treating a high salinity feed, such as fracking wastewater, combining microfiltration or ultrafiltration, followed by forward osmosis, and then followed by membrane distillation.
A massive source of natural gas, called “shale gas”, exists in pockets of underground porous rocks. Hydraulic fracturing made these underground porous rocks a viable natural gas source. Hydraulic fracturing, also called “fracking”, is a process comprising drilling and injecting fluid into the ground at high pressure in order to crack shale rocks, releasing the natural gas. In this process, a sand/water suspension and proppants (chemicals) are pumped, at high pressure, into the shale layer. As a result, natural gas is released and flows back up to the surface with the drilling fluids.
Currently shale gas is being produced in many regions of the United States. The production of shale gas through hydraulic fracturing has been criticized because of its negative environmental impacts and of the management implications of used hydraulic fracturing fluids, also known as “fracking wastewater”. High-salinity is the main characteristic of fracking wastewater, which contains different types of inorganic salts obtained from underground brines. Shale gas wastewater also contains dissolved organic compounds, oil, and sand. The discharge of such highly saline fracking wastewater is of great concern. The management of fracking wastewater is crucial to shale gas development and to the preservation of the environment and human health. A possible solution to these issues is to treat fracking wastewater before discharge/reuse. The treatment of highly saline fracking wastewater is both challenging and energy intensive.
On another subject, various desalination techniques are known in the art.
For example, forward osmosis (FO) is a desalination process in which a feed solution is treated by osmotic pressure rather than hydraulic pressure. The primary principle behind this process is osmosis, the natural diffusion of water (water flux) through a semi-permeable membrane from a low salinity feed solution into a high salinity draw solution due to the osmotic pressure gradient between these two solutions. This technique exploits the natural process of osmosis, which is the diffusion of salt due to different salinities on either side of a semi-permeable membrane. In contrast, the reverse osmosis process uses hydraulic pressure as the driving force for separation, which serves to counteract the osmotic pressure gradient that would otherwise favor water flux from the permeate to the feed. Hence significantly more energy is required for reverse osmosis compared to forward osmosis.
Another desalination technique is membrane distillation (MD), which is a thermally driven, membrane-based technology. MD is an emerging technology that utilizes low-grade heat or industrial waste-heat at a temperature of ˜50° C. to drive separation. MD is a thermally driven process in which water vapor transport occurs across a non-wetted microporous hydrophobic membrane. The driving force behind the MD process is the vapor pressure gradient, which is generated by a temperature difference across the membrane. Compared with reverse osmosis (RO) membrane process, which is the most popular water purification membrane technology, there is no need in MD to exert high operation pressure. Therefore, the energy cost of MD can be significantly reduced. In MD, a hydrophobic microporous membrane is used, which lets water vapors pass through, but repels liquid water. The driving force in MD is the vapor pressure gradient across the membrane derived from temperature difference between the hot feed and cold permeate streams. Low working temperature (30-80° C.) distinguishes MD from conventional thermal distillation, making it possible to utilize low-grade heat such as waste heat or solar thermal energy.
Despite such attractive advantages, MD is still in its embryonic stage and has not been widely applied in industrial and commercial development due to unresolved challenges. Membrane fouling and wetting are two major obstacles leading to MD operation failures when treating challenging wastewater sources. The potential membrane fouling and wetting constraint conventional hydrophobic MD membranes to the treatment of relatively clean feed solutions that are free of hydrophobic and amphiphilic substances. First, membrane fouling is a serious problem that affects MD performance and can cause major damage and costs, especially over long-term operation. The foulant layer formed on the hydrophobic membrane surface can block the membrane pores, and consequently cause significant decrease in water vapor flux and membrane wetting. Generally, humic acid, proteins and oily substances can easily attach on the membrane through hydrophobic-hydrophobic interaction. Moreover, deposition of inorganic species (scaling) on the membrane surface causes pores plugging. In addition to fouling, membrane wetting is another challenge that affects stable flux performance and salt rejection. Membrane wetting occurs when the feed liquids penetrate the pores, for example when the trans-membrane pressure exceeds the liquid entry pressure (LEP), which is affected by liquid surface tension, membrane hydrophobicity, pore size and pore shape. To avoid pore wetting, the hydrostatic pressure must be lower than the LEP. However, even though the operating pressure is lower than the LEP, MD membranes can be easily wetted by low surface tension contaminants (oil, alcohols, and surfactants) which are widely present in feeds, thus contaminating the distillate and undermining their rejection properties.
In accordance with the present invention, there is provided:
In the appended drawings:
Membrane distillation (MD) is a thermally driven, membrane-based desalination technology. Typically, MD is carried out as shown in
MD membranes are microporous and hydrophobic. Indeed, MD membranes must repel liquid water and only allow water vapor through. In other words, the MD membrane must remain non-wetted during use. Membrane wetting occurs when the feed liquids penetrate the membrane pores. Unfortunately, conventional MD membranes are rather easily wetted when low surface tension contaminants (oil, alcohols, and surfactants) are present in the feed. Another problem of MD membrane is membrane fouling in which foulants deposit on the membrane surface and block the membrane pores thus undesirably decreasing water vapor flux and favoring membrane wetting. As a result of the above, conventional MD membranes are limited to the treatment of relatively clean feeds that are free of hydrophobic and amphiphilic substances (e.g. low surface tension contaminants).
The present inventors aimed to produce a MD membrane that could be used to treat more challenging feeds, such feeds containing low surface tension contaminants, hydrophobic and amphiphilic substances, etc. They therefore sought to produce a membrane less prone to fouling and wetting. Conventionally, most efforts to achieve this goal aimed to modify the surface of existing membranes (more specifically membranes previously used for microfiltration and having a structure and surface not designed for MD) to make them superhydrophobic. Others have sough to produce amphiphobic MD membranes. However, these efforts have led to membranes suffering from one or more drawbacks such as:
In a first aspect of the invention, there is provided a membrane for membrane distillation. As shown in Example 1, the MD membrane of the invention is superhydrophobic and amphiphobic, exhibits enhanced stability and durability, presents low fouling and low wetting without compromising permeation (water vapor) flux and salt rejection, and which can therefore be used for treating highly saline feeds containing low surface tension substances. Further, the MD membrane of the invention can easily be manufactured by electrospinning followed by dip-coating.
Herein, “superhydrophobicity” (synonym of “ultrahydrophobicity”) means having a static water contact angle greater than 150°. Additionally, the superhydrophobic membranes of the invention have a water sliding angle of less than 10°. This helps maintaining an air gap between liquid droplets and the surface. This air gap provides an opportunity to increase allowable pore sizes prior to pore wetting occurrence, consequently allowing higher flux. Moreover, superhydrophobicity is believed to reduce the interaction between the feed and the membrane surface thereby reducing the fouling propensity. Because they repulse water, nearly spherical droplets form on the membrane surface and roll away, possibly taking foulants away.
Herein, amphiphobicity means having contact angles larger than 90° with both water and low surface tension liquids. Amphiphobicity allows the membranes of the invention to more effectively prevent contact between contaminants and membrane surface.
The membrane for membrane distillation of the invention comprises a microporous mat of electrospun nanofibers, wherein the nanofibers are made of a nanocomposite of reduced graphene oxide dispersed in a hydrophobic polymer, and wherein the surface of the nanofibers is grafted with:
Herein, a “mat of electrospun nanofibers” is a mat made by electrospinning. In the case of the above composite, electrospinning produces a mat with a surface presenting asperities and reentrant structures, the mat comprising nanofibers randomly arranged in an interconnected open microporous structure. Herein, a “open microporous structure” is a structure presenting micropores that are connected to each of other through the material.
In embodiments, the membrane for membrane distillation is in the shape of a sheet, either flat or curved, preferably flat.
In preferred embodiments, the hydrophobic polymer is polyvinylidene fluoride (PVDF), polytetrafluoroethylene (PTFE), polypropylene (PP), or poly (vinylidene fluoride-co-hexafluoropropylene) (PVDF-HFP). In more preferred embodiments, the hydrophobic polymer is poly (vinylidene fluoride-co-hexafluoropropylene) (PVDF-HFP).
In embodiments, the hydrophobic polymer has a molecular weight (Mw) of about 400 kDa.
In preferred embodiments, the reduced graphene oxide (rGO) is in the form of single-layer reduced graphene oxide nanosheets, preferably with a thickness of 0.7-1.2 nm and length of 300-800 nm.
In preferred embodiments, the concentration of reduced graphene oxide in the nanocomposite is between about 0.15 and about 0.25 wt %, preferably about 0.15 wt %, based on the total weight of the nanocomposite.
In preferred embodiments, the hydrophobic nanoparticles are titanium dioxide, silver, alumina, or silica nanoparticles that have been surface-modified as needed to have a hydrophobic surface. Such surface treatments include, for example, grafting a silane coupling agent on the surface of the nanoparticles. In preferred embodiments, the hydrophobic nanoparticles are surface-modified silica nanoparticles, for example silica nanoparticles with a silane coupling agent grafted on the surface of the silica nanoparticles
In embodiments, the silane coupling agent (which coats the mat, or which is grafted to the surface of the nanoparticles) is of formula Rm—Si—Xn, wherein:
R is alkyl, alkenyl, haloalkyl, or haloalkenyl,
X is alkoxy or halogen, and
m and n are integers between 1 and 4, such that m+n=4.
In preferred embodiments, R is alkyl, alkenyl, perhaloalkyl, or perhaloalkenyl; preferably alkyl, alkenyl, perfluoroalkyl, or perfluoroalkenyl; more preferably alkyl, alkenyl, perhaloalkyl, and yet more preferably methyl, vinyl, or perfluorododecyl or perfluorododecyl.
In preferred embodiments, X is methoxy, ethyoxy or chloro.
In preferred embodiments, m is 1 or 2 and n is 2 or 3, preferably m is 1 and n is 3.
In preferred embodiments, the silane coupling agent is a haloalkyltrialkoxysilane, a dialkyldihalosilane, alkenyltrialkoxysilane, or alkyltrialkoxysilane, preferably a haloalkyltrialkoxysilane, and more preferably a perhaloalkyltrialkoxysilane. In more preferred embodiments, the silane coupling agent is perfluorooctyltriethoxysilane (POTS), dimethyldichlorosilane (DDS), vinyltrimethoxysilane (VTS), methyltriethoxysilane (MTES), perfluorododecyltrichlorosilane, or perfluorodecyltrimethoxysilane, preferably perfluorooctyltriethoxysilane.
The MD membrane can be manufactured in two easy steps.
First, a mat of electrospun nanofibers is produced by electrospinning a dope solution of the hydrophobic polymer in which the rGO is suspended. This dope solution can be prepared by:
At this stage, a highly hydrophobic nanofiber mat is produced. In addition, Example 1 shows that the PVDF-HFP membranes with rGO exhibited improved stability and durability with satisfactory distillate quality compared with pristine PVDF-HFP membranes.
Then, surface superhydrophobicity and amphiphobicity are further constructed surface modification, i.e. grafting of the silane coupling agent, and/or hydrophobic nanoparticles. Indeed, the surface of the nanofibers can be easily modified by dip-coating to achieve the desired grafting.
For example, the nanofibers mat can be immersed in a solution of the silane coupling agent thus allowing reaction of the silane coupling agent with the surface of the nanofibers to achieve the desired grafting, rinsing, and then heating (for example at 120° C. for 4 h) to complete the reaction of the silane coupling agent with the surface of the nanofibers.
Alternatively, the nanofibers mat can be immersed in a suspension of the hydrophobic nanoparticles, such as the abovementioned surface-modified silica nanoparticles, thus allowing the grafting of the hydrophobic nanoparticles on the surface of the nanofibers, rinsing, and then heating (for example at 120° C. for 4 h) to complete the reaction of the silane coupling agent with the surface of the nanofibers. The suspension of hydrophobic surface-modified nanoparticles can be prepared by providing a suspension of nanoparticles, adding the silane coupling agent to the suspension and allowing the grafting of the silane coupling agent to the nanoparticles (via a Si—O—Si covalent bond). The nanofibers mat can be immersed directly into this reaction mixture, which will lead to a coating comprising the hydrophobic surface-modified nanoparticles, and optionally remaining unreacted silane coupling agent (which, after the above immersion step, will be either rinsed away, or react during the heat treatment).
It should be noted that the nanofibers mat is not simply coated with the silane coupling agent, and/or the hydrophobic surface-modified nanoparticles. Rather, these are grafted, i.e. attached, to the surface of the nanofibers. Without being limited by theory, it is believed that a condensation reaction occurs between the silane molecules (either free or attached to the surface of the SiO2 nanoparticles) and oxides functions at the surface of the nanofibers. The Si—OR bonds of these silane molecules hydrolyze readily with water to form silanol Si—OH groups, which could then condense with each other and with hydroxyl groups on the nanofiber surface surface to form polymeric structures. Similarly, the unreacted Si—OH on the SiO2 nanoparticles could form new Si—O bond on the nanofiber surface.
No matter with mechanism is at work, possibly in part because of the fact that the silane coupling agent, and/or the hydrophobic surface-modified silica nanoparticles are grafted on the surface of the nanofibers, the MD membrane of the invention exhibit robust chemical and mechanical stability and enhanced durability as shown in Example 1.
There is also provided a membrane distillation process using the above membrane for membrane distillation.
More specifically, this process comprises the steps of:
Forward osmosis (FO) is a membrane-based technology that allows separating water from dissolved solutes. Typically, FO is carried out as shown in
An ideal FO membrane has high liquid water permeability, high solute rejection, and high chemical and mechanical stability with a low propensity toward fouling. However, conventional FO membranes, some being somewhat hydrophobic, others being somewhat hydrophilic, have serious disadvantages including:
In another aspect of the invention, there is provided a forward osmosis membrane. As shown in Example 2, the thin-film composite FO membrane of the invention exhibits a high water flux with enhanced water permeability, improved antifouling properties and reduced ICP effect as well as high mechanical strength (compared to commercial FO membranes). This combination of properties is highly desirable for a FO membrane and crucial for successful application in the forward osmosis process.
It is believed that the support layer possesses high mechanical strength due to the synergistic effect between the interconnected (chemical connected) spider-web like structure of the electrospun N6 nanofiber mat and the integrated network structure of SiO2 nanoparticles. The interconnected spider-web like structure is unexpected in electrospun mats; rather a random stack of fiber is expected.
It is believed that the enhanced water permeability (high water flux), the improved antifouling properties, high porosity of the support layer, and the reduced ICP effects are due at least in part to the presence of SiO2 nanoparticles in both the support and rejection layers and to the electrospun nature of the support layer.
The FO membrane of the invention can be easily manufacture by electrospinning technique followed by interfacial polymerization on the surface of electrospun nanofiber mat.
The membrane for forward osmosis of the invention comprises a microporous support layer and a rejection layer formed on one side of the support layer, wherein:
Membranes comprising a support layer with a rejection layer formed on one side of the support layer are conventionally referred to as “thin film composite” FO membranes. One interesting feature of the membrane for forward osmosis of the invention is that both its support layer and its rejection layer advantageously comprise hydrophilic nanoparticles.
Since the rejection layer is formed on one side of the support layer, the forward osmosis membrane can be said to have an rejection layer side (i.e. the side of membrane where the rejection layer is formed) and a support layer side (i.e. the opposite side of the membrane).
In the present invention (and as described in the next section), the rejection layer is formed by interfacial polymerization of precursors to form the crosslinked meta-aramid of formula (I) on the support layer. Thus, it can be said that the rejection is interfacially polymerized on the support membrane. The rejection layer, while formed on the microporous support layer, is not porous. Rather, it is dense with few or no pores. The rejection layer allows water through via dissolution and diffusion of the water in the crosslinked meta-aramid of formula (I) of the rejection layer. Then, the water reaches the support layer where is migrates through the pores to reach the support layer side of the membrane. The salts and other solutes do not dissolve in the rejection layer cannot cross the membrane.
In embodiments, the membrane for forward osmosis is in the shape of a sheet, either flat or curved, preferably flat.
In embodiments, the support layer has a porosity of more than 90%, preferably of about 95%.
In preferred embodiments, the hydrophilic polymer is polyacrylic acid, polyvinyl alcohol, nylon 6, nylon 6.6, a proteins, cellulose, a polyethylene glycol ether, or a polyacrylic amide. In more preferred embodiments, the hydrophilic polymer is nylon 6. Herein, “nylon 6” refers to the polymer also known as polycaprolactam, polyamide 6, and poly(hexano-6-lactam) (IUPAC name), which has the following formula:
In yet more preferred embodiments, the nylon 6 is present in the nanofibers as a semi-crystalline polymer containing crystals of α and γ-form.
Herein, a “mat of electrospun nanofibers” is a mat made by electrospinning. In the case of the above hydrophilic nanoparticles/hydrophilic polymer composite, electrospinning produces a support layer that is a mat of highly interconnected nanofibers forming a spiderweb-like open microporous structure. Herein, a “spiderweb-like” structure is a network of fibers interconnected with each other via chemical interactions, for example ionic and/or hydrogen bonds, so as to form a web.
Herein, “aramid” has is usual meaning in the art, i.e. it designates aromatic polyamides, which are polymers with repeat units in which amide groups (—CO—NH—) directly bind two aromatic rings together (i.e. —Ar1-CO—NH—Ar2-CO—NH—). Aramids can be categorized as meta or para aramids depending on the attachment of the amide groups on the aromatic rings. A well-known meta aramid is poly(m-phenylene isophthalamide) (MPIA, Nomex™) which has the following formula:
The above crosslinked meta-aramid of formula (I) is quite similar to MPIA, except that the phenyl ring of the isophthalamide bears an extra functional group allowing crosslinking. Note that in formula (I), as per usual for crosslinked polymers, the open bonds indicate crosslinks to other repeat units of the crosslinked polymer.
In preferred embodiments, the concentration of silica nanoparticles in the nanocomposite with the hydrophilic polymer is between 10 wt % and 20 wt %, preferably of about 20 wt % (based on the total weight of the N6/silica nanoparticles nanocomposite).
In preferred embodiments, the concentration of silica nanoparticles in the nanocomposite with the crosslinked meta-aramid of formula (I) is between 1 wt % and 6 wt %, preferably of about 4 wt % (based on the total weight of the aramid/silica nanoparticles nanocomposite).
The hydrophilic nanoparticles in the support layer and in the rejection layer may be the same or different, preferably they are the same. In preferred embodiments, the hydrophilic nanoparticles are graphene oxide, montmorillonite, carboxylated gold, carboxylated silver, zinc oxide, titanium dioxide, or silica nanoparticles. In more preferred embodiments, the hydrophilic nanoparticles are silica nanoparticles.
In embodiments, silica nanoparticles range in size from about 10 to about 80 nm. Typically, the silica nanoparticles are smaller than a) the nanofiber diameter and b) the thickness of rejection layer. Preferably, the silica nanoparticles are about 10 to about 30 nm in size.
The FO membrane can be manufactured in two easy steps. First, the support layer is manufacture and then, the rejection layer is formed on one side of the support layer.
First, a mat of electrospun nanofibers is produced by electrospinning a dope solution of the hydrophilic polymer in which the hydrophilic nanoparticles are suspended. This dope solution can be prepared by:
The mat of electrospun nanofibers is the support layer. Then, the rejection layer is formed on one side of this mat by interfacial polymerization. The crosslinked meta-aramid of formula (I) can indeed be produced by polymerization between one or more aromatic di- or polyfunctional amines and one or more aromatic di- or polyfunctional acyl chlorides. This polymerization is carried out in the presence of silica nanoparticles, which result in the incorporation of the silica nanoparticles in the rejection layer.
For example, the rejection layer can be produced by:
Note that above, in step a), one side and the edge of the support layer are protected so that during steps b) and d), only the unprotected side of the support layer is in contact with the solutions of the first and second monomers.
Both m-phenylenediamine (MPD) and 1,3,5-benzenetricarbonyl trichloride (TMC) are precursors of the crosslinked meta-aramid of formula (I). Steps b) and c results in a support layer with MPD and silica particles deposited on its exposed side. In step d), this MPD reacts with the TMC to form the crosslinked meta-aramid of formula (I)—reaction presented in Example 2. The purpose of step f) is to complete internal cross-linking of the remaining un-reacted precursors. After step f), a rejection layer comprising silica nanoparticles dispersed in crosslinked meta-aramid of formula (I) (i.e. a nanocomposite) is formed.
There is also provided a forward osmosis process using the above membrane for forward osmosis.
More specifically, this process comprises the steps of:
Optionally, the process can comprise a further step of separating water from the diluted draw solution resulting from step c).
Process for Treating a High-Salinity and/or High-Strength Feed, Such As Fracking Wastewater
As noted above, the discharge of highly saline fracking wastewater produced by hydraulic fracturing is of great concern due to both human health and environmental effects. However, the high salinity and the contaminants present in the fracking wastewater make its treatment quite challenging. Therefore, in another aspect of the invention, there is provided a process for treating high-salinity and/or high-strength feeds, such as fracking wastewater. This process combines microfiltration, forward osmosis and membrane distillation. This process allows producing fresh water from high-salinity and/or high-strength feeds.
Indeed, as shown in Example 3, the process of the invention was successfully applied for the first time to treat fracking wastewater. Microfiltration as a pre-treatment removed ˜52% of total organic carbon (TOC) and ˜98.5% of turbidity. High average water fluxes (19.98 LMH for NaCl and 30.97 LMH for NaP draw solutions) with high solute rejection were obtained via the FO process using a nanocomposite membrane. In addition, 98.5% of the initial water flux was recovered with the nanocomposite membrane after desalination of the fracking wastewater. Membrane distillation as a downstream separator allowed recycling the FO draw solution, along with the production of pure water.
The process of the invention is thus a process for treating a high-salinity and/or high-strength feed, such as fracking wastewater, comprising:
Herein, a “high-salinity feed” is a feed that that a salinity of about 35 g/kg or more (i.e. about 35 g or more of salts in 1 kg of feed). Note that the “salinity” of a feed or solution is defined as the concentration of all the salts dissolved in the feed or solution and that the average ocean salinity is about 35 g/kg. Non-limiting examples of high-salinity feeds include fracking wastewater, textile and lather industry effluents, effluents from the petroleum refinery industry, and effluents from the agro-food industry.
Herein, a “high-strength feed” is a feed that that a chemical oxygen demand (COD) of about 10,000 mg/L or more. Note that the “chemical oxygen demand” is an indicative measure of the amount of oxygen that can be consumed by reactions in a measured solution. It is commonly expressed in mass of oxygen consumed over volume of solution which in SI units is milligrams per litre (mg/L). COD is as well-known test that easily quantifies the amount of organics in water. In fact, the most common application of COD is in quantifying the amount of oxidizable pollutants found in surface water. Non-limiting examples of high-strength feeds include fracking wastewater.
The first step of the process is the microfiltration or ultrafiltration of the high-salinity and/or high-strength feed to produce a pre-treated feed as a filtrate. Pre-treatment of fracking wastewater (or other high-salinity and/or high-strength feeds) was found to increase the efficiency and life expectancy of the FO membrane by minimizing fouling. It also removed sand particles and oil from fracking wastewater (or other high-salinity and/or high-strength feeds) thus producing as a filtrate, a pre-treated feed more suitable for FO.
Microfiltration (MF) and ultrafiltration are both size exclusion-based filtration technologies. Typically, MF and UF are carried out as shown in
More specifically, step a) of the process of the invention comprises the sub-steps of:
MF and UF membranes are porous semipermeable membranes that allow various particle sizes to either flow through or be trapped by the membrane, and the degree of separation largely depends on particle size. The main difference between MF and UF membranes is pore size with microfiltration membranes having a pore size ranging from 0.1 to 10 μm, while ultrafiltration membranes have a pore size ranging from 0.1 to 0.01 μm.
The MF and UF membranes used in the process of the invention prevents particles such as sediment, algae, protozoa or large bacteria from passing through. Depending on the non-dissolved contaminants—such as sand and oil—in the high-salinity and/or high-strength feed, ultrafiltration or microfiltration will be used as pre-treatment technology. In general, MF, with its larger pore-sized membrane, allows water, monovalent and multivalent ions, and viruses through its barrier while blocking certain bacteria and suspended solids. In contrast, ultrafiltration, with its smaller pore size, blocks everything microfiltration can in addition to viruses, silica, proteins, plastics, endotoxins, and smog and/or fumes. The pore size of the MF or UF membrane used will also be selected depending on the non-dissolved contaminants in the feed.
In preferred embodiments, step a) of the process of the invention comprises subjecting the high-salinity and/or high-strength feed to microfiltration. Indeed, microfiltration can be more convenient than ultrafiltration, since it typically requires lower pressures and has typically higher permeability. Thus, unless the nature of the contaminants in the high-salinity and/or high-strength requires the use of ultrafiltration, microfiltration is preferred. In particular, it has been shown in Example 3 below that microfiltration could successfully be used when treating fracking wastewater.
The exact nature of the MF and UF membranes is no crucial to the invention. Commercial MF and UF membranes can be used. MF and UF membranes can be constructed with polymers, such as polypropylene, cellulose acetate, and polysulfone, but they can also be constructed of ceramic or stainless steel. In preferred embodiments, the MF membrane is an electrospun Nylon 6 nanocomposite membrane as described in Example 3, or a commercial polysulfone (PSf) membrane such as the HT 200 membrane (Pall Corporation, USA). In more preferred embodiments, the MF membrane is an electrospun Nylon 6 nanocomposite membrane as described in Example 3.
The next step of the process is to subject the pre-treated feed (filtrate) obtained from step a) to forward osmosis. Indeed, forward osmosis can desalinate the pre-treated feed using fairly straightforward and economic, low-pressure equipment.
The forward osmosis process has been generally described above and this general description applies here. As noted in the previous sections, forward osmosis requires the use of a draw solution and results in a water-diluted draw solution.
More specifically, step b) of the process of the invention comprises the sub-steps of:
The exact nature of the FO membrane is no crucial to the invention. Commercial FO membranes can be used. In embodiments, the FO membrane is cellulose triacetate membrane or a thin-film composite polyamide membrane. In preferred embodiments, the FO membrane is a membrane for forward osmosis as described in the previous sections, preferably as described in Example 2 hereinbelow, or a commercial polyamide membrane such as that provided by Hydration Technology Innovations (HTI, Albany, Oreg., USA), such as 40161507 Filter membranes, Basic TFC Forward Osmosis Membranes kit. In more preferred embodiments, the FO membrane is a membrane for forward osmosis as described in the previous sections, preferably as described in Example 2 hereinbelow.
The draw solution can comprise a single simple salt or multiple simple salts or a substance specifically tailored for forward osmosis applications. In embodiments, the draw solution is an aqueous NaCl or NaP solution. In more preferred embodiments, the draw solution is an aqueous 4.0 M NaCl solution from an aqueous 4.6 M NaP solution, as long as the draw solution has a salinity higher than the salinity of the pre-treated feed. In yet more preferred embodiments, the draw solution is an aqueous 4.6 M NaP solution.
The next step of the process is to subject the water-diluted draw solution to membrane distillation to produce water. Membrane distillation is indeed used as a separator downstream of the FO process to recycle the water-diluted draw solution produced by the forward osmosis process. Indeed, since membrane distillation removes water from the water-diluted draw solution, it regenerates the (more concentrate) draw solution and produces water. This renders the process more economical.
The membrane distillation process has been generally described above and this general description applies here. More specifically, membrane distillation step c) of the process of the invention comprises the sub-steps of:
The exact nature of the MD membrane is no crucial to the invention. Commercial MD membranes can be used. Highly hydrophobic membranes are nevertheless preferred. In embodiments, the MD membrane is a membrane for membrane distillation as described in the previous sections, preferably as described in Example 1 hereinbelow, a commercial poly(vinylidene fluoride) (PVDF) membrane, such as Durapore®, Millipore, USA (mean pore size 0.22 μm, porosity 75%). In more preferred embodiments, the MD membrane is a membrane for membrane distillation as described in the previous sections, preferably as described in Example 1 hereinbelow.
In embodiments, the process further comprises the step of reusing the draw solution regenerated in step c) in the forward osmosis treatment of step b).
The use of the terms “a” and “an” and “the” and similar referents in the context of describing the invention (especially in the context of the following claims) are to be construed to cover both the singular and the plural, unless otherwise indicated herein or clearly contradicted by context.
The terms “comprising”, “having”, “including”, and “containing” are to be construed as open-ended terms (i.e., meaning “including, but not limited to”) unless otherwise noted.
Recitation of ranges of values herein are merely intended to serve as a shorthand method of referring individually to each separate value falling within the range, unless otherwise indicated herein, and each separate value is incorporated into the specification as if it were individually recited herein. All subsets of values within the ranges are also incorporated into the specification as if they were individually recited herein.
Herein, the terms “alkyl”, “alkenyl” and their derivatives (such as alkoxy, etc.) have their ordinary meaning in the art. For more certainty, herein:
It is to be noted that, unless otherwise specified, the hydrocarbon chains of the above groups can be linear or branched. Further, unless otherwise specified, these groups can contain between 1 and 18 carbon atoms, more specifically between 1 and 12 carbon atoms, between 1 and 6 carbon atoms, between 1 and 3 carbon atoms, or contain 1 or 2, preferably 1, or preferably 2 carbon atoms.
All methods described herein can be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context.
The use of any and all examples, or exemplary language (e.g., “such as”) provided herein, is intended merely to better illuminate the invention and does not pose a limitation on the scope of the invention unless otherwise claimed.
No language in the specification should be construed as indicating any non-claimed element as essential to the practice of the invention.
Herein, the term “about” has its ordinary meaning. In embodiments, it may mean plus or minus 10% or plus or minus 5% of the numerical value qualified.
Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
Other objects, advantages and features of the present invention will become more apparent upon reading of the following non-restrictive description of specific embodiments thereof, given by way of example only with reference to the accompanying drawings.
The present invention is illustrated in further details by the following non-limiting examples.
We developed two easy-to-produce superhydrophobic and amphiphobic nanofibrous membranes for membrane distillation. These membranes comprise a nanocomposite of reduced graphene oxide in poly (vinylidene fluoride-co-hexafluoropropylene).
In fact, two superhydrophobic and amphiphobic membranes, which could repel both water and low surface tension liquids (e.g. oil), were produced by electrospinning followed by surface modification.
First, highly hydrophobic nanofiber mats were prepared by electrospinning a blend polymer of poly (vinylidene fluoride-co-hexafluoropropylene) (PVDF-HFP) and reduced graphene oxide (rGO). The rGO incorporated membranes exhibited improved stability and durability with satisfactory distillate quality compared with pristine PVDF-HFP membranes.
Surface superhydrophobicity and amphiphobicity were further increased by grafting a fluoroalkylsilane of low surface energy or silica nanoparticles to the rGO incorporated membranes. Indeed, perfluorooctyltriethoxysilane (POTS) and hydrophobic SiNPs were grafted on the membrane surface through a simple dip-coating process. The functionalization of the membranes with POST and SiNPs is shown in
We investigated the morphology, surface robustness and anti-wetting properties of the modified nanofibrous membranes and examined the impacts of surface modification on permeate flux and salt rejection in the direct contact membrane distillation (DCMD) process.
Various techniques such as SEM, AFM, XPS, liquid entry pressure (LEP) and contact angle measurement were used to explore the effects of surface modification on the morphology and structure of the membranes. The analysis results revealed changes in the electrospun nanofiber diameters, surface roughness, elemental composition and hydrophobicity.
The membranes produced showed excellent superhydrophobicity and amphiphobicity, as demonstrated by their wetting resistance with water and low surface tension organic solvents. In particular, both the hydrophobic SiO2 nanoparticles grafted membrane (PH-rGO-SiNPs membrane) and POTS-grafted membrane (PH-rGO-POTS membrane) modified membrane displayed superhydrophobicity with water contact angle larger than 150° and sliding angle lower than 2°, indicating their self-cleaning properties. Moreover, two as-prepared membranes exhibit large diiodomethane contact angles of 146.5° and 145.5°, respectively. These two membranes also exhibited excellent amphiphobic chemical, thermal and mechanical stability even after the challenging treatments including 4 h boiling in DI water, 110 h etching in strong HCI and NaOH solution, and sonication for 1 h demonstrating that POTS and SiO2 could adhere to PVDF-HFP/rGO nanofibers firmly and withstand various harsh treatment.
The novel nanofibrous membranes also exhibited excellent anti-wetting and anti-fouling performances in membrane distillation. Indeed, we challenged the stability of the amphiphobic nanofibrous membrane with a model surfactant—sodium dodecyl sulfate (SDS) containing feed saline solution. We demonstrated the membrane properties during dynamic membrane distillation operation, in which the membranes were used to purify water from a 35 g/L sodium chloride solution in the presence of SDS. The modified membranes exhibited enhanced stability and durability of MD performance in both high permeation flux and salt rejection. The SiO2 nanoparticles-grafted amphiphobic membrane presented a robust dynamic performance with a relatively higher water flux and desired permeate conductivity in the presence of 0.3 mM SDS during the DCMD process, compared with the pristine membrane without SiO2 nanoparticles grafting, demonstrating the outstanding anti-wetting property of amphiphobic membranes.
The present Example refers to the following documents, all of which are incorporated herein by reference.
Poly (vinylidene fluoride-co-hexafluoropropylene) (PVDF-HFP, Mw: ˜400000), dimethylacetamide (DMAc), perfluorooctyltriethoxysilane (POTS, 97%), tetraethyl orthosilicate (TEOS), ethyl alcohol (C2H5OH), ammonium hydroxide (NH4OH), and sodium dodecyl sulfate (SDS) were purchased from Sigma-Aldrich (Oakville, ON, Canada) and used without any pre-treatment. All other agents including sodium chloride (NaCl≥99.5%) and acetone (ACS reagent grade) were purchased from Fisher-Scientific (St Laurent, QC, Canada). Single layer reduced graphene oxide (rGO) nanosheets, with a thickness of 0.7-1.2 nm and length of 300-800 nm, were purchased from Cheap Tubes Inc. (Grafton, Vt., USA). Deionized (DI) water was prepared using a Milli-Q purification system (Millipore, Billerica, Mass.).
PVDF-HFP/rGO electrospun mats (PH-rGO) were first prepared. A dope solution was prepared by dissolving PVDF-HFP (3.0 g) in 20 mL of a mixture of DMAc/acetone (8/12, V/V). A mixture of rGO and PVDF-HFP was prepared by suspending 30 mg (0.15 wt %) of rGO in the aforementioned mixture of DMAc/acetone by probe sonication (Branson 3510, Shanghai, China) for 10 min followed by the addition of the same amount of PVDF-HFP. The mixture was stirred overnight on a hot plate at 45° C. 20 mL of each solution was loaded into a Luer-lock syringe (Vitaneedle, Mass.).
Electrospinning of the dope solutions was conducted using a Nanospinner (NE300, Inovenso, Turkey). Polymeric solutions were delivered to the metallic nozzle at a 2.0 mL/h flow rate. A high voltage (25 KV) was applied between the nozzle and the electrically grounded metallic drum. The distance between nozzle tip and collector (12 cm), temperature (24° C.) and relative humidity (25%) were held constant during the process.
The electrospun mats were modified by POTS as shown in
The electrospun mats were modified with SiNPs as shown in
First, a hydrophobic silica nanoparticles (SiNPs) suspension was prepared as described in Reference 1.1. Ammonia (2.4 mL) and ethanol (30 mL) were mixed to form a homogenous solution, and TEOS (2.8 mL) was then added. After 8 h of magnetic stirring, 0.4 mL POTS was added to the reaction solution. The reaction was stirred for another 24 h at room temperature to form a hydrophobic silica particulate sol. Under this synthesis condition, the silica particles were present in the suspension at a concentration of 1.5 wt %.
The PVDF-HFP-rGO mats were immersed in the silica particulate suspension (silica particle concentration, 1.5 wt %) for 36 h to apply silica nanoparticles to the membrane surface. After rinsing with DI water, the treated mat was then dried at 120° C. for 45 min.
The morphology of the membranes was observed using a FEI Quanta 450 Environmental Scanning Electron Microscope (FE-SEM; FEI company, USA). Samples were coated with a thin 4 nm layer of palladium (Pd) before observation by microscopy. Silicon (Si) mapping was obtained using an EDS apparatus. The fiber diameter distribution and frequency were measured via ImageJ software. The elemental composition of membranes was evaluated by X-ray photoelectron spectroscopy (XPS; SK-Alpha). Surface roughness and topography of membranes were investigated by atomic force microscopy (AFM; NanoINK Inc. Skokie, Ill., USA). Fourier transform infrared spectroscopy (FT-IR) was performed with a Nicolet 6700/Smart iTR (Thermo Scientific, Waltham, Mass., USA) equipped with an attenuated total reflectance (ATR) single logic accessory to observe the functional group changes on PH-rGO mat surface after grafting of POTS and silica nanoparticles.
The surface wetting resistance of the membranes was evaluated by contact angle measurements of DI water (γ=72.1 mN/m), diiodomethane (γ=50.1 mN/m), and glycerol (γ=64 mN/m) using video contact angle system (VCA; AST Products, Inc., Billerica, Mass., USA). The static contact angles were measured by using the system software (VCA optima XE). The water sliding angles were measured by tilting the membrane samples that were fixed on the stage until the water droplet (10 μL) started to move on the surface. At least three desiccator-dried samples were used for contact angle measurements and for each sample, about three points were tested. The data was averaged between the samples.
The liquid entry pressure of water (LEPw) was measured by placing the membrane in a dead-end filtration cell. The setup for LEP measurements is shown in
The mean flow pore size (MFP), pore size distribution of the as-prepared membranes were characterized by using a capillary flow porometer (CFP-1500AE, Porous Materials Inc. (PMI), Ithaca, N.Y., USA) based on the wet/dry flow method, where the membranes were firstly wetted with wetting liquid called Galwick (surface tension: 15.9 mN/m) and then placed in a sealed chamber through which gas flows. The membrane porosity was determined by the gravity method reported previously.
The thermal, mechanical and chemical stability of surface amphiphobicity were evaluated under challenging conditions including boiling water (DI water, 100° C.) for 4 h, sonication for 60 min, acidic (HCI solution, pH=2) and basic conditions (NaOH solution, pH=12) at room temperature for 110 h. Water and diiodomethane contact angles of the top electrospun membrane surface were then measured at room temperature following the above procedure.
The DCMD performance experiment was conducted with the apparatus shown in
A flat-sheet membrane, with an effective area of 34 cm2, was tightly fixed into a PTFE membrane cell (CF042P-FO, Sterlitech Corporation, USA). A hot feed solution was maintained at a constant temperature using a water bath. The feed solution and a cold solution were moved at the same speed across the bottom and upper face of the membrane cell respectively with the help of two gear pumps (GH-75211-10, Cole-parmer, Canada) at around 0.8 psi (GH-68930-12, Cole-Parmer, Canada). The circulation feed rate and permeate rate were detected by two flowmeters (0.1-1 LPM, McMaster-CARR, Canada) and held constant at 0.75 LPM. The operational temperature was monitored at the inlet and outlet of the module using four thermocouples (SCPSS-032u-6, OMEGA, Canada) connected to a thermometer (EW-91427-00, Cole-Parmer, Canada). The inlet temperature of the hot feed varied from 50 to 75° C., while the cold side was kept at a constant 25.0° C. The conductivity of NaCl in the distillate was investigated with an electric conductivity meter (Oakton Instruments, Vernon Hills, Ill., USA).
The experiments were first carried out with DI water to determine the pure water flux of the membranes. Subsequently, 35 g/L of NaCl solution was employed as feed solution to investigate the salt rejection.
The permeate flux, J, of the prepared membrane was calculated using the following equation:
where J is the permeate flux (kg/m2 h), ΔM is the quantity of distillate (kg), A is the effective membrane area (m2) and Δt is the operation time (h). The salt rejection R was calculated using the following equation:
where Cf and Cp are the concentration of the feed and permeate, respectively.
The wetting propensity of the modified nanofibrous membrane was investigated in the presence of SDS surfactant in the feed solution. For the initial 60 min of MD runs, a 3.5 wt % NaCl solution was used as a feed solution, and the mass and conductivity of the cold DI water side were recorded constantly, and thus the real-time flux and salt rejection were calculated and monitored. The SDS was then added to the feed stream solution with a final concentration of 0.3 mM to reduce the surface tension of the solution and thereby to induce pore wetting, if the membranes were wettable. If the membranes were wetted, the feed saline solution would permeate through the wetted portions of the membrane to the distillate side, leading to a significant water flux decline and a salt rejection loss.
The pristine PH-rGO membrane was used as a control sample.
Surface morphology and chemical composition of the membrane surface are first discussed. Electrospinning was applied to fabricate PVDF-HFP-rGO (PH-rGO) nanofibrous membranes. Moreover, electrospun nanofibrouss membranes with multilevel surface roughness provides a re-entrant structure, which improves surface amphiphobicity.
SEM was used to observe the morphology of PVDF-HFP-rGO electrospun membrane before and after surface modification as shown in
The modified membranes had a radically different surface morphology compared to the pristine PVDF-HFP/rGO membrane due to the presence of POTS and the SiNPs. It is evident from the SEM images that POTS molecules and SiNPs cover the surface of the nanofibers and that some form aggregates, which can sustain a metastable Cassie-Baxter thermodynamic state—see Reference 1.3.
The presence of POTS molecules and SiN Ps were further examined by XPS analysis and FTIR—see
The respective FTIR spectra of the pristine and modified PVDF-HFP/rGO membranes are shown in
A possible mechanism of the condensation reaction between silane molecules (POTS and SiO2 nanoparticles) and an oxide surface (PH-rGO) is shown in
To further analyse the oxygen state on POTS-grafted or hydrophobic SiO2 nanoparticles grafted membrane, high-resolution XPS spectra were collected at the binding energy from 536 to 528 eV. The Si—O peak at 533.9 eV appeared on the two modified membranes (
The surface wettability was characterized using the static contact angle using water (
In addition to superhrdrophobicity, the POTS molecules and SiO2 nanoparticles modified membranes presented strong oleophobicity with a sharp increase in diiodomethane contact angle from 52.3° for pristine PVDF-HFP/rGO to 145.5° and 146.5°, respectively (
Conventionally the long-term stability of amphiphobic surfaces remains a challenge for their practical applications. Generally, due to the poor adhesion between the coating and the hydrophobic porous PVDF-HFP/rGO support, dip-coating method is perceived as being not effective as it produces coatings that tend to easily peel off from the support—see Reference 1.10. The membrane grafted with POTS and those grafted with hydrophobic SiO2 nanoparticles unexpectedly possess robust amphiphobicity possibly due to the condensation reaction sites provided by the rGO with POTS molecules and SiO2 nanoparticles. To further investigate the chemical and mechanical stability of the modified membranes, the membranes were challenged with critical conditions including boiling in water and exposure to strong acid and base solutions, and then their contact angles were determined against water and diiodomethane, respectively. The results reported in
As can be seen from
The DCMD test was first conducted using NaCl solution as the feed to assess the performance characteristics of the two modified membranes. The effect of different parameters such as feed and coolant temperature, the feed salt concentration on product distillate water were analyzed.
In
It can be seen from
To investigate the impact of grafted silica nanoparticles on the membrane performances, the water flux and salt rejection of PH-rGO-SiNPs nanofibrous were tested in the DCMD process over a long-term operation using 3.5 wt % NaCl with addition of 0.3 mM SDS solution as a feed solution maintained at 75° C. and DI water was used as a permeate cooling solution maintained at 25° C.
Sodium dodecyl sulfate (SDS) is a characteristic popular surfactant in wastewater, often remarkably decreasing the surface tension of wastewater, which normally wets the MD membrane and breaks its performance immediately—see Reference 1.21 and 1.22.
Here, 0.3 mM SDS was introduced into 3.5 wt % NaCl feed solution to challenge the membrane stability during the DCMD process.
It is interesting that the pristine membrane was not wetted immediately once the SDS was added into the feed, meaning it already had some resistance against low surface tension solutions and could be a good candidate for substrate aiming for amphiphobic membranes. The highly hydrophobic electrospun PVDF-HFP/rGO nanofiber substrate with the reentrant structure exhibited glycerol contact angle around 130°. The capability to prevent surfactant wetting should also be correlated with its surface superhydrophobicity. Similar phenomenon has also been observed on highly hydrophobic polymer surfaces (Teflon AF, Parafilm and PP) with ionic surfactants such as cationic dodecyltrimethylammonium bromide (DTAB) and anionic sodium dodecyl sulfate (SDS)—see Reference 1.24. Therefore, the membrane with highly amphiphobic grafted can also effectively prevent wetting from SDS in the feed.
The SiNPs grafted PVDF-HFP/rGO amphiphobic membrane showed a striking contrast and presented a relatively stable water flux with no obvious decrease and keep constant 100% salt rejection. There was almost no NaCl penetration across the membrane (99.99% of salt rejection) and the modified membrane possessed high wetting resistance against SDS.
The stronger anti-wetting property of PH-rGO-SiNPs nanofibrous membrane as compared to PVDF-HFP/rGO is attributable to the second scale of roughness introduced by the SiN Ps on the fiber surface, which rendered the local wetting of individual fibers more difficult. The importance of multiscale reentrant structure on enhancing the superhydrophobicity or amphiphobicity of membranes is relatively well understood—see References 1.25 and 1.26. In previous studies, the additional local reentrant structure was imparted by coating fluorinated TiO2 or SiO2 nanoparticles onto the existing membrane substrates, which has been shown to be very effective in mitigating membrane wetting—see Reference 1.27. Here, we demonstrate an effective anti-wetting MD membrane with hierarchical roughness can be fabricated on the rGO based polymeric substrate without any pretreatment to create a second scale of reentrant structure. In conclusion, the SiNPs grafted improved the dynamic anti-wetting property of PVDF-HFP/rGO nanofibrous membrane against SDS surfactant in the DCMD process.
A high flux and antifouling thin-film composite (TFC) forward osmosis (FO) membrane containing silica (SiO2) nanoparticles was fabricated using a facile electrospinning technique followed by interfacial polymerization on surface of electrospun nanofiber mat. The successful fabrication of the TFC membrane was confirmed via FE-SEM, TEM, XRD, FTIR, and AFM analyses.
Both the electrospun nylon 6 (N6) substrate and the polyamide (PA) active layer contained superhydrophilic SiO2 nanoparticles enhancing the hydrophilicity of the fabricated FO membrane. The fabricated electrospun N6/SiO2-supported TFC FO membrane with a PA/SiO2 composite active layer was robust (tensile strength of 22.3 MPa) with a water contact angle of 14°.
In the FO process, the fabricated TFC membrane exhibited a high water flux (27.10 LMH) with a low specific reverse salt flux (5.9×10−3 mol·L−1). The fabricated membrane also showed high antifouling propensity in FO process for the model foulants of sodium alginate and calcium sulfate. The initial water flux recovery for this membrane was 98% for sodium alginate and 94% for calcium sulfate.
Moreover, a strong interaction between the electrospun substrate and the active layer demonstrated the structural stability of the fabricated TFC membrane. Experimental
The present Example refers to the following documents, all of which are incorporated herein by reference.
Nylon 6 (N6), tetraethyl orthosilicate (TEOS), ethyl alcohol (C2H5OH), ammonium hydroxide (NH4OH), m-phenylenediamine (MPD), 1,3,5-benzenetricarbonyl trichloride (TMC), and hexane were obtained from Sigma-Aldrich, USA. Both formic acid and acetic acid were acquired from Fisher Scientific, USA. A commercial flat-sheet TFC forward osmosis membrane was purchased from Hydration Technology Innovations (HTI, Albany, Oreg., USA). De-ionized (DI) water was obtained from a Millipore Integral 10 water system (Millipore, Billerica, Mass.).
N6 (21% by weight) was dissolved in a mixture of formic and acetic acids (80% formic acid and 20% acetic acid by volume) using magnetic stirring (rpm 350) for 5 h at room temperature. Separately, a SiO2 solution was prepared by mixing TEOS, ethanol, and water at a molar ratio of 1:2:2, respectively, in the presence of an NH4OH catalyst and stirred at 25° C. for 4 h. SiO2 nanoparticles were then separated from the mixture through centrifugation. Subsequently, SiO2 nanoparticles were dispersed in a formic acid (80% by volume) and acetic acid (20% by volume) mixture under sonication for 20 min. An appropriate ratio of SiO2 dispersion was added to the N6 solution and sonicated for 5 min and then stirred for 5 h at ambient conditions to make a N6/SiO2 solution with 20% SiO2 content. In the electrospinning process, high-voltage electricity (Nanospinner NE300, Inovenso, Turkey) was applied to the prepared N6/SiO2solution in a syringe (volume 20 mL, inside diameter 19.05 mm) via an alligator clip attached to the syringe nozzle. The applied voltage was adjusted to 30 kV. The solution was delivered to the nozzle tip via a syringe pump to control the solution flow rate (0.18 mL/h). Fiber mats were collected on an electrically grounded metallic drum placed 8.8 cm above the nozzle tip—see References 2.1 and 2.2. Temperature (25° C.) and relative humidity (40%) were controlled by the electrospinning machine itself and a dehumidifier (RECUSORB DR-010B), respectively, throughout the fabrication process.
A N6 solution without SiO2 nanoparticles was also electrospun to fabricate a pristine N6 nanofiber mat as a substrate of the TFC membrane.
An active layer of PA/SiO2 nanoparticle composite was formed on the electrospun N6/SiO2 substrate by an interfacial polymerization reaction. First, the electrospun substrate was put on a glass plate and then each side of the substrate was tapped with the glass plate very well. The electrospun substrate with the glass plate was immersed in an aqueous MPD/SiO2 solution (1% MPD and 1, 2, 4 and 6% SiO2 with respect to MPD) for 2 min. Excess MPD solution was removed from the substrate surface using an air knife. The MPD/SiO2 substrate was then dipped into a solution of 0.15 wt % TMC in hexane for 1 min (to form an ultrathin PA/SiO2 composite as active layer by an interfacial polymerization reaction between MPD and TMC) followed by removal of the excess TMC solution from the top surface of the substrate using an air knife. The electrospun substrate with the PA/SiO2composite active layer was then heated at ˜75° C. in an oven for 10 min to complete internal cross-linking of the remaining un-reacted precursors of interfacial polymerization reaction—see References 2.3 to 2.5.
A schematic representation of the interfacial polymerization between MPD and TMC is presented below:
A polyamide active layer was also fabricated on the electrospun N6/SiO2 substrate without adding SiO2 nanoparticles into MPD solution during the interfacial polymerization between MPD and TMC as the same protocol mentioned above. Moreover, both polyamide and polyamide/SiO2 composite (4% SiO2 content as regards MPD) active layers were fabricated on the electrospun N6 substrate.
All the fabricated TFC membranes were stored in DI water until they were tested. These TFC membranes are summarized in Table 2.
A N6 substrate was also prepared by casting and phase inversion method. A 21% N6 solution (by weight) in 80% formic acid and 20% acetic acid mixture (by volume) was manually cast on a clean glass plate using a casting knife with the thickness of 85 μm at ambient condition. After casting, the film was dried for 24 h at ambient condition and then it was peeled from the glass plate. Finally, the film was immersed into DI water for another 24 h in order to remove the remaining solvent.
Field emission-scanning electron microscopy (FE-SEM) (QUANTA FEG 450) with a platinum coating on the sample surface was performed to observe the morphology of the substrates and TFC membranes. Cross-sectional morphology and thickness of the TFC membrane were measured using FE-SEM and a TMI instrument (Testing Machines, Inc.), respectively—see Reference 2.6.
Transmission electron microscopy (TEM) (TF20) was conducted to examine the morphology of the electrospun substrates. A structural study of the electrospun substrates was conducted via the use of X-ray diffraction (XRD) (Bruker D8 Discover with a copper X-ray source and equipped with a Vantec area detector), and Fourier transform infra-red (FTIR) (NICOLET 6700 FT-IR) spectrometry. The wettability and surface roughness of both the substrates and TFC membrane were investigated using a VCA optima instrument (AST Products, Inc.), and an atomic force microscope (AFM) (BRUKER, NanoScopeRV), respectively. The wettability of SiO2 nanoparticles was also investigated using the VCA optima instrument (AST Products, Inc.).
The analysis was performed using the same protocol for SiO2 nanoparticles described elsewhere—see References 2.6 and 2.7. Briefly, a few drops of SiO2 nanoparticle suspension, in ethanol, were placed on a glass slide and then dried in an oven at 80° C. for 30 min followed by cooling the SiO2 nanoparticles on the glass slide at room temperature. Then, the VCA optima instrument was used to determine the water contact angle of the SiO2 nanoparticles placed on the glass slide. The tensile strength of the substrates and TFC membrane were investigated using an Instron instrument (Mini 44), USA.
The gravimetric method was used to investigate the porosity of the electrospun substrate, using the following equation:
where Ww and Wd are the weight of the wet and dry substrates, respectively; ρw is the water density (0.998 g cm−3); A is the effective area of the substrate; and L is the substrate thickness—see Reference 2.8 to 2.12.
The mean pore size of the substrate was determined via the filtration velocity method. The volume of permeate water was obtained using a dead-end stirred cell filtration device (Millipore stirred ultra-filtration cells, 8010, USA, effective area of 0.0003 m2) connected to a nitrogen gas cylinder. The mean pore size (rm) was calculated using the Guerout-Elford-Ferry equation:
where ε is the substrate porosity; η is the water viscosity (8.9×10−4 Pa s); I is the substrate thickness; QT is the permeate volume per unit time; ΔP is the applied pressure (1 bar); and A is the effective area of the substrate—see References 2.9 to 2.12.
The maximum pore size (Rmax) was determined via the bubble point method. The bubble point pressure was determined using the aforementioned dead-end stirred cell filtration system—see Reference 2.6. The substrate was immersed in DI water for 4 h and then fitted on the dead-end cell. The output tube of the dead-end cell was immersed in DI water so that the bubble point pressure could be read. The maximum pore size was calculated according to Laplace's equation:
where σ is the surface tension of water (72.80×10−3 Nm−1); θ is the contact angle of water on the substrate; and P is the minimum bubble point pressure—see References 2.6 and 2.9.
A flat-sheet TFC membrane was used to conduct all the forward osmosis experiments. The water permeability coefficient (A) and salt permeability coefficient (B) for the TFC membrane were investigated using a bench-scale cross-flow RO test system. A piece of the membrane with an effective surface area of 19.94 cm2 was placed in a stainless-steel test cell with the active surface of the membrane facing the feed stream. Using a high-pressure positive displacement pump (Hydra-cell pump), the feed solution was re-circulated at the velocity of 52.6 cm/s. DI water was used as the feed stream to investigate A, and a 20 mM solution of NaCl was used as the feed stream to investigate R (rejection) and B for the TFC membrane.
A, R and B for the membrane were determined using the following equations:
where J is the pure water flux; Am is the effective membrane area; ΔV is the permeate volume; Δt is time; ΔP is the hydraulic pressure difference across the membrane; Cf is the salt concentration of the feed solution; Cp is the salt concentration of the permeate solution; and Δπ is the osmotic pressure of the feed solution—see References 2.8 and 2.13 to 2.15.
The pressure was increased in 0.345 MPa increments from 0.345 to 1.034 MPa in order to investigate A of the TFC membrane. Constant pressure was applied at each increment for 8 h. The water flux through the membrane was obtained from a liquid flow sensor (Sensirion, The Sensor Company) directly connected to a computer. To investigate R and B, 1.896 MPa pressure was applied to the RO cell. Conductivity of the feed and permeate solutions was investigated using a calibrated conductivity meter (Oakton, Eutech Instruments) to calculate solute rejection. This experiment was conducted at a constant temperature of 24° C. using a chiller (Polystat, Cole-Parmer).
A bench-scale FO test system was used to determine the structural parameter (S) of the TFC membrane by applying the following equation:
where Jw is the FO water flux for the draw solutions. In this approach, de-ionized water was used as the feed solution, while 1 M NaCl was used as the draw solution—see References 2.8, 2.13, 2.16 and 2.17.
A bench-scale experimental setup (shown in
where Cf and Vf are the salt concentration and total volume of the feed, respectively, at the end of the tests; and Cf,i and Vf,i are the initial salt concentration and total volume of the feed, respectively—see Reference 2.18.
Sodium alginate (SA) and calcium sulfate (CaSO4) were used as model organic and inorganic foulants, respectively, to investigate the antifouling properties of the FO membranes. The membrane coupon was placed into the FO cell with the active layer facing the feed side. The membrane coupon was immersed in DI water for 24 h before conducting the antifouling test. First, the FO experiment was conducted for 6 h at a flow rate of 26.3 cm/s using 1 M NaCl as draw solution and DI water as feed. Then, 1 M NaCl, as draw solution, and DI water with SA (200 mg/L) and CaCl2 (1 mM), as feed solution, were used to conduct the antifouling test for 6 h at the same flow rate (26.3 cm/s) using a new membrane coupon. To investigate antifouling propensity in relation to CaSO4, 1 M NaCl as draw solution and DI water with CaSO4 (2000 mg/L) as feed solution, were used to conduct the antifouling test. This experiment was also conducted for 6 h at a flow rate of 26.3 cm/s using a new membrane coupon. Weight changes of the feed solution throughout the FO experiments were monitored precisely using a digital weight balance at fifteen-minute intervals. Then, for 2 h, DI water with a flow rate of 52.6 cm/s was applied to physically clean the membrane active surface of the both fouled membranes (fouled by SA and CaSO4). In the FO experiment, the water flux through the cleaned membranes was measured using 1 M NaCl and DI water as draw and feed, respectively, in order to investigate flux recovery for these membranes. These experiments were also conducted for 6 h at a flow rate of 26.3 cm/s in which the weight changes of the feed solution were monitored using a digital weight balance at fifteen-minute intervals.
The FE-SEM images of N6 and N6/SiO2 composite electrospun substrates with 20 wt. % SiO2 content are shown in
The electrospun N6 substrate showed a fibrous morphology in which the diameter of the fibers ranged between 80 to 160 nm (
The SEM-EDX spectra of the electrospun substrates for the pristine N6 and N6/SiO2 composite are shown in
The EDX analysis suggests the presence of C, N and O atoms of N6 (
TEM images of the electrospun substrates of pristine N6 and N6/SiO2 composites are shown in
The incorporation of SiO2 into N6 caused a reduction in the peak intensity and a smaller peak appeared as a result of the crystal structure splitting from γ (2θ=21.2°) into the α-form at 2θ=23.5°—see Reference 2.21.
The FTIR spectra are shown in
The wettability of the electrospun substrates of pristine N6 and N6 with 20 wt. % SiO2content is shown in
The porosities of the electrospun substrates of pristine N6 and N6/SiO2 (20 wt. %) composite are shown in Table 3. The electrospun N6 substrate with 21 wt % of N6 solution exhibited high porosity (86%) due to the high surface area to volume ratio of the nanofibers of the substrate. However, the incorporation of SiO2 nanoparticles (20 wt %) increased the porosity of the electrospun N6 substrate by ˜10%. The average and maximum pore sizes of the electrospun N6 substrate were 406 and 575 nm, respectively, while those values were 478 nm (average) and 661 nm (maximum) for the electrospun N6/SiO2 composite substrate (Table 3). It is assumed that the higher pore sizes are due to higher fiber diameters of the electrospun N6/SiO2 composite substrate as compared to the electrospun N6 substrate.
The tensile strength of the fabricated electrospun substrates is also shown in Table 3. The electrospun N6 substrate showed a tensile strength of 19.0 MPa. The high tensile strength of the electrospun N6 substrate was due to the highly interconnected spider-web like structure in the substrate. The ionic species of the N6 solution form stronger hydrogen bonds because of the extra available charge on them in the presence of high applied voltage during the electrospinning process. The protonated amide group of ionic N6 can effectively form hydrogen bonds with oxygen atoms of a N6 molecule in the main fiber and form another hydrogen bond between an oxygen atom between the ionic molecule and a hydrogen atom from the amide group of another main fiber to form the interconnected spider-web like substrate. The incorporation of SiO2 nanoparticles enhanced the tensile strength of the electrospun N6 substrate (21.40 MPa), likely due to the integrated network structure of SiO2 (see the schematic representation of the electrospun N6/SiO2 composite provided above).
A N6 substrate was prepared by the phase inversion method. The casted N6 substrate was almost nonporous (see
The top surface FE-SEM images of the electrospun N6/SiO2 composite supported TFC membranes with pristine PA and PA/SiO2 composite active layers are exhibited in
The cross-section of the fabricated electrospun N6/SiO2 supported TFC membrane with 4% SiO2 content in the PA active layer is shown in
The SEM-EDX spectra of the surfaces of the electrospun N6/SiO2 supported TFC membranes with the pristine PA and the PA/SiO2 composite active layers are shown in
The surface roughness of the electrospun N6/SiO2 supported TFC membranes with the pristine PA and the PA/SiO2 composite active layers was also investigated through AFM and the result of this investigation is shown in
The surface roughness of the electrospun substrates was investigated through AFM and the result of this investigation is shown in
The wettability of the fabricated and the commercial TFC membranes is reported in Table 4. The water contact angles of the fabricated E.Spun N6-PA and E.Spun N6-PA/SiO2 TFC membranes were 63° and 47°, respectively. A lower water contact angle was obtained for the TFC membrane with PA/SiO2 composite active layer. The water contact angle of the fabricated E.Spun N6/SiO2-PA TFC membrane was 32°, however, the water contact angle was only 14° when incorporating 4% SiO2 nanoparticles (as regards MPD during interfacial polymerization) into the PA active layer. The water contact angle decreased due to superhydrophilic properties of the incorporated SiO2 nanoparticles into the PA active layer. In Table 4, it is also observed that the wettability of the fabricated TFC membranes increased with increasing wettability of the substrates, while the active layers remained the same. In fact, the highly wettable substrate induced the very thin active layer to be more wettable. The water contact angle of the fabricated E.Spun N6/SiO2-PA/SiO2 TFC membrane was 0.56 times lower as compared to that of the commercial TFC membrane (water contact angle 25°). The obtained water contact angle of the commercial TFC membrane was comparable to the literature value (water contact angle)24° for the same type of membrane—see Reference 2.26.
22 ± 0.4
The tensile strength of the fabricated membranes—as well as commercial TFC membranes—is also reported in Table 4. The tensile strength of E.Spun N6-PA and E.Spun N6-PA/SiO2 TFC membranes were 19.4 and 19.5 MPa, respectively. The fabricated E.Spun N6/SiO2-PA TFC membrane showed a tensile strength of 22 MPa. The tensile strength of the E.Spun N6/SiO2-PA/SiO2 TFC membrane was almost same as the fabricated E.Spun N6/SiO2-PA TFC membrane. The very small quantity of incorporated SiO2 nanoparticles into the PA active layer could not provide any contribution to enhance mechanical strength of the E.Spun N6/SiO2-PA/SiO2 TFC membrane. However, the tensile strength of the electrospun substrates slightly increased after fabricating active layers on it due to fiber binding effect of the active layer (Table 3 and Table 4). The obtained tensile strength of the commercial TFC membrane was much lower (8.2 MPa) as compared to the fabricated TFC membranes (Table 4).
A cross-flow RO cell was used to investigate pure water permeability of the fabricated as well as a commercial TFC membranes, and the obtained water permeability values were 20.1, 23.3, 28.2, 45, and 32.5 LMH/MPa for E.Spun N6-PA, E.Spun N6-PA/SiO2, E.Spun N6/SiO2-PA, E.Spun N6/SiO2-PA/SiO2, and commercial TFC membranes, respectively. The obtained water permeability value for the commercial TFC membrane is very near to the literature value (31.6 LMH/MPa) for the same type of membrane—see Reference 2.27. The fabricated E.Spun N6-PA and E.Spun N6-PA/SiO2 TFC membranes were not considered for further FO performance investigations due to their lower water permeability compared to those of the other two fabricated membranes. The structural parameters of the fabricated and the commercial TFC membranes were determined through the investigation of salt rejection and salt permeability coefficient in a cross-flow RO cell (Table 5).
The salt rejections of the fabricated membranes were 98% for E.Spun N6/SiO2-PA and 98.5% for E.Spun N6/SiO2-PA/SiO2, whereas it was 97.27% for the commercial TFC membrane. The salt permeability coefficient of the fabricated membranes were 1.04 LMH for E.Spun N6/SiO2-PA, 1.24 LMH for E.Spun N6/SiO2-PA/SiO2, which were lower than that of the commercial membrane (1.65 LMH). FO water fluxes for the fabricated and the commercial TFC membranes are presented in Table 5. In order to obtain water flux, 1 M NaCl and DI water were used as draw solution and feed, respectively, in the FO process. The use of 1 M NaCl as draw solution and DI water as feed is a common practice in FO process. The obtained FO water fluxes for the fabricated E.Spun N6/SiO2-PA/SiO2TFC membrane was higher (27.10 LMH) than those of the other fabricated E.Spun N6/SiO2-PA (17.50 LMH) and the commercial (20.82 LMH) TFC membranes at the same experimental conditions. Compared to the fabricated E.Spun N6/SiO2-PA and the commercial TFC membranes, the higher FO water flux was obtained for the fabricated E.Spun N6/SiO2-PA/SiO2 TFC membrane due to its higher hydrophilicity with lower structural parameters as presented in Table 6 and Table 5.
Table 6 presents a comparison between the intrinsic permeation properties of lab-made TFC membranes and the literature TFC flat sheet membranes under both FO and RO conditions.
The reverse salt flux and specific reverse salt flux of the TFC membranes used in FO processes are shown in
The antifouling propensity of the fabricated membrane, as well as commercial TFC membrane, was studied in the presence of two separate foulants, namely SA (model organic foulant) with calcium ions (as bridging agent) and CaSO4 (model inorganic foulant). The fouling behavior of the TFC membranes with these two foulants is illustrated in
The decline in water flux for the E.Spun N6/SiO2-PA, E.Spun N6/SiO2-PA/SiO2, and the commercial TFC membranes were 18%, 13%, and 23%, respectively, when CaSO4 was used as the foulant (
A combined process, comprised of microfiltration, forward osmosis and membrane distillation was successfully applied to the treatment of fracking wastewater. In fact, both insoluble and soluble contaminants were removed by microfiltration and forward osmosis, respectively. After applying this combined process, fresh water was obtained from the fracking wastewater.
Microfiltration as a pre-treatment process followed the emerging forward osmosis coupled with membrane distillation—used as a downstream separator to recycle FO draw solutions as well as to produce pure water—as post-treatment processes were successfully applied for the first time to the treatment of fracking wastewater. Microfiltration as a pre-treatment removed ˜52% of TOC and ˜98.5% of turbidity. High average water fluxes (19.98 LMH for NaCl and 30.97 LMH for NaP draw solutions) with high solute rejection were obtained via the FO process using a nanocomposite membrane, while these water fluxes were 14.39 LMH for NaCl and 23.79 LMH for NaP draw solutions when using a PA membrane. High solute rejection was obtained by both membranes (nanocomposite and PA) in the FO treatment of pre-treated fracking wastewater. This research also demonstrated that 98.5% and 97% of initial water flux can be recovered by the nanocomposite and PA membranes, respectively, after desalination of fracking wastewater. In membrane distillation, permeate fluxes were about 10.40 LMH for NaCl and about 13.82 LMH for NaP with approximately 99.99% solute rejection, producing “pure” water. This result indicates a successful implementation of membrane distillation as a downstream separator in the FO process.
The present Example refers to the following documents, all of which are incorporated herein by reference.
Sodium chloride (NaCl) and sodium propionate (NaP) were purchased from Sigma-Aldrich, USA.
Nanocomposite microfiltration membranes were produced by our laboratory (see details below) and polysulfone (PSf) microfiltration membranes were purchased from Pall Corporation, USA (Part number: S80065, Description: HT 200 membrane, 8 inch-10 inch sheet; Base material: Unsupported polysulfone (HT), pore size: 0.2 micrometer, thickness: 114.3-190.5 micrometer).
The flat-sheet thin-film composite (TFC) FO membranes were as described in Example 2 above [nanocomposite membrane] and #40161507 Filter membranes, Basic TFC Forward Osmosis Membranes by Hydration Technology Innovations (HTI, Albany, Oreg., USA) [polyamide (PA) membrane].
Millipore, USA, provided poly(vinylidene fluoride) (PVDF) membrane (Durapore®) (mean pore size 0.22 μm, porosity 75%) for membrane distillation.
Sample fracking wastewater was obtained from Canbriam Energy Inc., Calgary, Alberta, Canada. The composition of dissolved inorganic solids in this wastewater is provided in the Table 7. De-ionized (DI) water was supplied from a Millipore Integral 10 water system (Millipore, Billerica, Mass.).
Nylon 6 (N6), tetraethyl orthosilicate (TEOS), ethyl alcohol (C2H5OH), polyvinyl acetate (Mw 140,000), ammonium hydroxide (NH4OH), acetone and sodium dodecyl sulfate (SDS) were obtained from Sigma-Aldrich, USA. Both formic acid and acetic acid were received from Fisher Scientific, USA. Machine oil (90% base oil with 10% additives, density of 881.4 kg/m3 at 20° C., kinematic viscosity of 271.62 mm2/s at 20° C., and surface tension of 29.8 mN/m at 20° C.) was received from Canadian Tire (Canada). De-ionized (DI) water was obtained from a Millipore Integral 10 water system (Millipore, Billerica, Mass.).
N6 (21% by weight) was dissolved in a mixture of formic and acetic acids (80% formic acid and 20% acetic acid by volume) using magnetic stirring (rpm 350) for 5 h at room temperature. Separately, a SiO2 solution was prepared by mixing TEOS, ethanol and water at a molar ratio of 1:2:2, respectively, in the presence of an NH4OH catalyst and stirred at 25° C. for 4 h. The SiO2 nanoparticles were then separated from the mixture through centrifugation. Subsequently, the SiO2 nanoparticles were dispersed in a formic acid (80% by volume) and acetic acid (20% by volume) mixture under sonication for 20 min. An appropriate ratio of SiO2 dispersion was then added into the N6 solution and sonicated for 5 min and then stirred for 5 h at ambient condition.
Electrospinning. High-voltage electricity (Nanospinner NE300, Inovenso, Turkey) was applied to the prepared solutions in a syringe (volume 20 mL, inside diameter 19.05 mm) via an alligator clip attached to the syringe nozzle. The applied voltage was adjusted to 30 kV. The solution was delivered to the nozzle tip via a syringe pump to control the solution flow rate (0.18 mL/h). Fiber mats were collected on an electrically grounded metallic drum placed 8.8 cm above the nozzle tip [12, 30]. Temperature (25° C.) and relative humidity (40%) were controlled throughout the fabrication process.
Coating, drying and washing. A PVAc coating layer was applied onto the electrospun nanofiber mat through casting and then phase inversion techniques. PVAc was dissolved in acetone under magnetic stirring for 3 h to make a 10% casting solution. The N6 nanofiber mat was first soaked in DI water before coating in order to minimize the penetration of the PVAc solution into the nanofiber mat. After making the coating, the resulting two-tier composite membrane was dried for 4 h at ambient conditions and then immersed in de-ionized water for 24 h in order to remove the excess solvent from the membrane.
Films of pristine N6 and PVAc were also prepared to investigate water contact angles. A 21% nylon 6 solution (by weight) in 80% formic acid and 20% acetic acid mixture (by volume) was used to fabricate nylon 6 film. The nylon 6 solution was casted manually on a clean glass plate using a casting knife with the thickness of 60 μm at ambient condition. After casting, the film was dried for 24 h at ambient condition and then it was removed from the glass plate. A 10% PVAc solution (by weight) in acetone was used to make PVAc film. The PVAc solution was casted manually on an aluminium foil putting on glass plate using a casting knife with the thickness of 60 μm at ambient condition. The PVAc film was dried for 24 h at ambient condition after casting and then it was removed from the aluminium foil.
The thicknesses of all membranes were measured using a TMI instrument (Testing Machines, Inc.)—see Reference 3.1. The wettability and tensile strength of the membranes were investigated using a VCA optima instrument (AST Products, Inc.) and Instron (Mini 44), USA, respectively. Field emission-scanning electron microscopy (FE-SEM) (QUANTA FEG 450) with a platinum coating on the sample surface was performed to examine the morphology of the FO membranes.
The gravimetric method was used to investigate porosity (ε) of the MF membranes using the following equation:
where Ww and Wd are the weight of the wet and dry membranes, respectively; ρw is the water density (0.998 g cm−3); Am1 is the effective area of the membrane and L1 is the membrane thickness—see References 3.2 to 3.6.
The mean pore size of the MF membranes was determined via the filtration velocity method. The volume of permeate water was obtained using a dead-end stirred cell filtration device (Millipore stirred ultra-filtration cells, 8010, USA, effective area of 0.0003 m2) connected to a nitrogen gas cylinder. The mean pore size (rm) was calculated using the Guerout-Elford-Ferry equation:
where is the water viscosity (8.9×10−4Pa s); L1 is the membrane thickness; QT is the permeate volume per unit time; ΔP is the applied pressure (1 bar) and A is the effective area of the membrane—see References 3.3 to 3.6.
Pure water flux for the MF membranes was measured using a dead-end stirred cell filtration device (Millipore stirred ultra-filtration cells, 8010, USA, effective area of 0.0003 m2) connected to a nitrogen gas cylinder. The membrane was pre-compacted at an applied pressure of 0.28 bar until a constant water flux was achieved. Pure water flux at a temperature of 25° C. was measured at applied pressures of 0.28, 0.55, 0.83, 1.1 and 1.38 bar. The equations below were used to calculate pure water permeability for the MF membranes:
where J0, V, Am1, A1, Δt1 and ΔP1 are the pure water flux/permeate flux, permeated water volume, membrane effective area, water permeability, measurement time, and applied pressure across the membrane, respectively—see References 3.2 and 3.7.
The water permeability (A) for the FO membranes was investigated using a flat-sheet bench-scale cross-flow RO test system. A piece of the membrane with an effective surface area of 19.94 cm2 was placed in a stainless-steel test cell with the active surface of the membrane facing the feed stream. Using a high-pressure positive displacement pump (Hydra-cell pump), the feed solution was re-circulated at 1.0 L/min. DI water was used as the feed stream to investigate water permeability for the FO membranes. Water permeability values for the membrane were calculated using the following equations:
where J is the pure water flux, Am is the effective membrane area, ΔV is the permeate volume, Δt is time, and ΔP is the hydraulic pressure difference across the membrane—References 3.2 and 3.8 to 3.10. The pressure was increased in 3.45 bar increments from 3.45 to 10.34 bar in order to investigate A of the FO membranes. Constant pressure was applied at each increment for 8 h. The water flux through the membrane was obtained from a liquid flow sensor (Sensirion, The Sensor Company) that was directly connected to a computer.
The treatment of fracking wastewater involved three steps: microfiltration, then forward osmosis and finally recovery of draw solution and pure water production by membrane distillation. The fracking wastewater treatment process is shown schematically in
Microfiltration for fracking wastewater water was conducted using both nanocomposite and PSf membranes in a dead-end stirred cell filtration device (Millipore stirred ultra-filtration cells, 8010, USA, effective area of 0.0003 m2) connected to a nitrogen gas cylinder. The membranes were pre-compacted using DI water at an applied pressure of 0.28 bar until a constant water flux was achieved. Then, microfiltration using the fracking wastewater as a feed was conducted for 12 h at a stirring rate of 500 rpm and an applied pressure of 0.28 bar. Turbidity, total organic carbon (TOC), conductivity and pH of the fracking wastewater after sample collection and after microfiltration were investigated using a MicroTPW Turbidimeter (HF, Scientific, Inc., USA), a TOC analyzer (TOC VCPH/CPN, Shimadzu Corp., Japan), a calibrated conductivity meter (Oakton, Eutech Instruments) and a calibrated pH meter (Oakton, Eutech Instruments), respectively.
The osmotic pressure of the fracking wastewaters after microfiltration was also investigated in the following way. The van Laar equation was used to calculate the osmotic pressure of the fracking wastewater:
where πo.p.f is the osmotic pressure, Ri is the ideal gas constant, T is the absolute temperature, V0 is the molar volume of solvent and a1 is the water activity—see References 3.11 and 3.12.
The water activity was calculated using the following equation:
where P and P0 are the vapor pressures of the fracking wastewater and DI water, respectively, at 24° C. The vapor pressures of the fracking wastewater and DI water were investigated using a U-Tube Manometer (Tenaquip, Canada)—see References 3.11 and 3.13.
Water flux recovery for the MF membranes after pre-treatment of fracking wastewater was also investigated. After filtering the fracking wastewater, the membranes were cleaned by rinsing with DI water for 30 min, and the pure water flux was then measured again using the equation (3.3) at the same applied pressure (0.28 bar). Water flux recovery (FR) was calculated according to the following equation:
where Jy and Jx are the pure water flux of membrane before and after filtration of fracking wastewater, respectively—see References 3.3, 3.5, and 3.14 to 3.16.
A bench-scale FO experimental setup (
To investigate water flux recovery for the FO membranes fouled by pre-treated fracking wastewater, the weight changes of feed solution throughout the FO experiments was monitored closely (30 minute interval) using a digital weight balance. After FO, DI water (in the both feed and draw side) was applied for 2 h with a flow rate of 1 L/min to physically clean the active surface of the fouled membranes. The water flux through the cleaned membranes was finally measured using 4.6 M NaP and pre-treated fracking wastewater as draw and feed solutions, respectively, in the same FO experiment set-up in order to investigate water flux recovery for these membranes. These experiments were conducted for 6 h at the flow rate of 0.5 L/min in which the weight changes of feed solution were monitored using a digital weight balance at thirty minute intervals. The same experiment was also conducted for NaCl draw solution (4.0 M).
Membrane distillation was used as a downstream separator to recycle the FO draw solutions. A Sterlitech membrane test cell system with a membrane active area of 34 cm2 was used to conduct the membrane distillation experiment. In this experiment, the draw solutions NaCl (4.0 M) and NaP (4.6 M) (used for pre-treated fracking wastewater using nanocomposite FO membrane) were used as feed solutions and DI water (conductivity <15 μS) was used as the coolant in the permeate side. To conduct the experiment, the feed solution and the permeate were placed in two separate 2.0 L reservoirs. The permeate container was placed on a digital analytical balance. Each experiment was conducted for 3 h, maintaining the feed and permeate temperatures of 50° C. and 20° C., respectively. Weight changes and conductivity of the permeate were monitored using the digital weight balance and a calibrated conductivity meter (Oakton, Eutech Instruments), respectively, at 30 min intervals. Initial conductivity of the feed solution was also measured using the calibrated conductivity meter (Oakton, Eutech Instruments). Concentration of the feed solution was determined using the gravimetric method at 60 min intervals during the MD experiment. Permeate flux and solute rejection (in terms of conductivity) were calculated using the following equations:
where J1, V1, Am2, Δt2, R, Cf, and Cp are the permeate flux, permeated water volume, membrane effective area, measurement time, solute rejection, feed concentration and permeate concentration, respectively—see References 3.18 to 3.20.
The characteristics of the MF membranes used for pre-treatment of fracking wastewater are reported in Table 7. The thicknesses of the nanocomposite and PSf membranes were almost identical (thickness 155 μm for nanocomposite membrane and 160 μm for PSf membrane). The porosities of the nanocomposite and PSf membranes were also almost identical (porosity 78% for nanocomposite membrane and 75% for PSf membrane). However, the mean pore size of the nanocomposite membrane was 1.18 times lower than that of the PSf membrane (mean pore size 170 nm for nanocomposite membrane and 200 nm for PSf membrane). The water contact angle of the nanocomposite membrane was 21°, while it was 2.14 times higher for the PSf membrane (water contact angle 45°). Due to higher hydrophilicity, a much higher water permeability was obtained for the nanocomposite membrane (water permeability 4814 LMH/bar) as compared to the PSf membrane (water permeability 2728 LMH/bar).
The characteristics of the FO membranes used for desalination of fracking wastewater are also reported in Table 7. The thicknesses of the nanocomposite and the PA membranes were similar (thickness 85 μm for nanocomposite membrane and 82 μm for PA membrane). The water contact angle of the nanocomposite membrane was 14°, while it was 1.79 times higher for the PA membrane (water contact angle 25°). Due to higher hydrophilicity, higher water permeability was obtained for the nanocomposite membrane (water permeability 4.5 LMH/bar) as compared to the PA membrane (water permeability 3.25 LMH/bar).
A MD process was used downstream to recover and recycle the draw solution in the FO process. The characteristics of the membrane used in the MD process are also reported in Table 7. An hydrophobic (water contact angle 123°) and microporous (mean pore size 220 nm) PVDF membrane was used in the MD process. The thickness, porosity and tensile strength of this membrane were 158 μm, 75% and 6.5 MPa, respectively.
14 ± 0.5
The pure water permeability values of the nanocomposite and PSf membranes are presented in
Water permeability as a function of time in pre-treatment of fracking wastewater by microfiltration is presented in
The dt/dV versus V filtration curves for the fouling stage in the microfiltration of fracking wastewater for each membrane are showed in
where dV is the permeate volume in the time of dt and q is a constant—see References 3.1 and 3.21.
The specific cake resistances for the fouling stage for the nanocomposite and the PSf membranes in microfiltration of fracking wastewater are shown in
The antifouling properties, in terms of water flux recovery, of the nanocomposite and the PSf membranes in microfiltration of fracking wastewater are shown in
The turbidity, TOC, conductivity, and pH of the fracking wastewater were 106 NTU, 853 mg/L, ˜67 mS, and ˜5.0, respectively, after collection of the sample (Table 9). The turbidity and TOC were due to the presence of oil and dissolved organic compounds in the fracking wastewater. The turbidity was reduced to 1.6 NTU and 2 NTU by the nanocomposite and PSf membranes, respectively, after conducting microfiltration. The TOC decreased to 409 mg/L and 413 mg/L after conducting microfiltration by the nanocomposite and PSf membranes, respectively. The decrease in turbidity and TOC were due to the removal of oil from the fracking wastewater by microfiltration. The TOC values of 409 mg/L (for nanocomposite membrane) and 413 mg/L (for PSf membrane) after microfiltration were due to the presence of dissolved organic compounds in the wastewater. Both of the membranes (nanocomposite and PSf) removed ˜52 TOC and ˜98% turbidity from the fracking wastewater by microfiltration process. The osmotic pressure of the fracking wastewater before and after microfiltration was measured and the value of this parameter was ˜128.3 bar (Table 9).
Water flux through the nanocomposite and PA membranes as a function of time for the raw fracking wastewater used as feed in FO process is shown in
Water flux through the nanocomposite and PA membrane, as a function of time for the pre-treated fracking wastewater used as feed in FO process is exhibited in
The composition of fracking wastewater and draw solutions in terms of TDS and TOC were investigated before and after desalination (Table 10. The TDS values of feed solutions were slightly higher before desalination as compared to those values after desalination when NaP was used as draw solution. The TDS values of NaP draw solution were also slightly higher after desalination as compared to those values before desalination. These observations indicate that very small quantities of solute (NaP) might pass through the membrane from feed to draw side during desalination by FO. On the other hand, TDS values of feed solutions slightly lower before desalination as compared to those values after desalination when NaCl was used as a draw solution. TDS values of the NaCl draw solution were little bit lower after desalination as compared to those values of before desalination. These scenarios might be due to higher reverse salt flux for this draw solution as compared to the organic draw solutions. The TOC values of feed solutions were slightly higher after desalination as compared to those values before desalination for NaP draw solution. The higher TOC values obtained were likely due to reverse salt flux of the organic draw solutions during FO process. However, TOC values were almost identical before and after desalination for the NaCl draw solution, indicating greater than 99% rejection of dissolved organic compounds in fracking wastewaters during FO.
Fouling behaviours of the membranes after FO were investigated through FE-SEM. FE-SEM images of virgin nanocomposite and PA membranes are shown in
The FE-SEM images of the fouled nanocomposite and PA membranes when NaCl was used as a draw solution, are provided in
Fouling behaviours of the membranes were further investigated by SEM-EDX (spectra shown in
Post-desalination water flux recovery of the nanocomposite and the PA membranes was studied after treatment by FO. Water flux declined 4.5% and 9% over 6 h (with initial water fluxes of 31.78 LMH for the nanocomposite and 24.83 LMH for the PA membranes) for the nanocomposite and the PA membranes, respectively, when 4.6 M NaP was used as draw solution against pre-treated fracking wastewater (
FO water flux recovery data of the nanocomposite and the PA membranes, when NaCl was used as draw solution against pre-treated fracking wastewater, are shown in
The draw solutions (obtained from FO) were used as feed solutions in the MD process, by which the separation of these draw solutions was conducted to recycle draw solute for reuse in further FO process. In the MD process, the permeate fluxes were approximately 10.40 LMH for NaCl and 13.82 LMH for NaP where a ˜99.99% solute rejection rate was obtained for both draw solution (
Note, in
The scope of the claims should not be limited by the preferred embodiments set forth in the examples but should be given the broadest interpretation consistent with the description as a whole.
The present description refers to a number of documents, the content of which is herein incorporated by reference in their entirety. These documents include, but are not limited to, the following:
This application claims benefit, under 35 U.S.C. § 119(e), of U.S. provisional application Ser. No. 62/732,781, filed on Sep. 18, 2018. All documents above are incorporated herein in their entirety by reference.
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/CA2019/051324 | 9/18/2019 | WO | 00 |
Number | Date | Country | |
---|---|---|---|
62732781 | Sep 2018 | US |