Patent Application
20240001340
None.
The present disclosure relates to silica-based granular media, a process for producing the same, and a method of degrading organic compounds using the silica-based granular media, as well as reactors employing the silica-based granular media.
Due to the ubiquitous and harmful presence of per- and polyfluoroalkyl substances (PFASs) in the environment, these compounds have been designated as emerging contaminants of concern. Utilization in various applications from Teflon production to firefighting foams has led to widespread contamination. PFASs are extremely resistant to degradation, bioaccumulative, and persistent once they enter into the environment. As a result, bioaccumulation and rapid uptake occurs within food chains. Primary exposure pathways for people include drinking water sources, fish consumption, and food packaging, with perfluorooctane sulfonate (PFOS) being one of the main concerns for human exposure. Such exposure can cause chronic health impacts and inhibit child development.
Prior treatment techniques removed most PFASs from ground water using pump-and-treat systems such as IX or granular activated carbon (GAC) treatment. One of the primary drawbacks of IX and GAC is the need to dispose of the regenerant solutions containing concentrated PFASs. Currently, spent GAC or IX resin materials produced at groundwater remediation sites must be transported off-site and typically hauled long distances to licensed facilities for disposal or regeneration. Regeneration of IX resins has been known to produce, on average, five bed volumes of concentrated PFAS solutions, proprietary solvents, and brines per 1,000 beds of volume treatment. This large volume of highly concentrated solution is typically further concentrated, followed by incineration. Incineration methods can produce undesirable by-products and smaller-chained PFASs. If complete mineralization and defluorination is not achieved, toxic shorter-chain PFASs can form as by-products, which are typically harder to treat and more mobile in the environment.
Although incineration is the primary mechanism for destruction of PFASs in concentrated waste streams, additional destructive technologies have been documented. Advanced chemical oxidation using hydroxyl radical-based chemistry, which is typically employed to degrade persistent organic chemicals, is reported to be inefficient for PFASs, as the C—F bonds resist complete reduction. Adsorption using activated carbon, photocatalysis, photolysis, thermolysis, and other promising technologies have been proposed but are either minimally effective in removing short-chained perfluoroalkyl acids (PFAAs) or are energy-intensive, requiring high temperatures or pressures that are difficult to implement in the field.
Advanced oxidation processes for destruction of PFASs have been developed to systematically target PFAS degradation. However, the techniques shown in this field lack the ability to completely mineralize PFASs or are difficult to implement in the field at large capacities. The major drawback of photocatalytic slurries is the need to filter out the photocatalyst after treatment for recovery and reuse. Salts present in groundwater or concentrated wastes can also inhibit the capabilities of the photocatalysts, and any amount of turbidity impacts activation of the photocatalyst.
Degradation initiated by nucleophiles has been shown to completely mineralize PFASs with almost complete recovery of aqueous fluoride. The ability to upscale these treatment technologies to continuously treat large quantities of concentrated waste is very limited due to the high energy required for defluorination and the extended contact time necessary to fully treat solution at less extreme conditions.
There is currently a need for a non-thermal, destructive, practical and cost-effective technology that are scalable, cost-effective and low-energy, and can easily be commercialized to degrade (e.g., mineralize) PFASs present in concentrated liquid waste streams, including legacy aqueous film forming foam (AFFF), ion exchange (IX) resin regenerant, landfill leachate, industrial wastewater, and more.
One aspect of the present disclosure is a photocatalytic silica-based granular media for degrading organic compounds formed from a three-dimensional polymer and comprising cross-linked silicon-oxygen (Si—O—Si) bonds formed through hydrolysis of an alkoxide precursor and a photocatalyst, wherein the media comprises a distribution of pore spaces.
Also provided herein is a process for producing a photocatalytic silica-based granular media, the process comprising introducing a photocatalyst to an alkoxide precursor with heat and/or agitation to form a photocatalyst mixture, hydrolyzing and condensing the photocatalyst mixture until a polymer gel is formed, removing excess solution to fuse the gel into a granular media, and adding a foaming agent to create a distribution of internal pore space within the granular media.
A further aspect of the present disclosure is a method for degrading one or more organic compounds, the method comprising introducing the one or more organic compounds to the silica-based granular media and irradiating the compound with electromagnetic radiation, preferably UV radiation.
Another aspect of the present disclosure is a reactor to degrade a composition comprising one or more organic compounds, the reactor comprising an inlet to allow the passage of an incoming stream containing the one or more organic compounds, at least one media area, wherein the media area is packed with the silica-based granular media, at least one UV light source exposed to a treatment area, and an outlet to allow the passage of an outgoing waste stream at least partially depleted of the one or more organic compounds.
Other objects and features will be in part apparent and in part pointed out hereinafter.
Corresponding reference characters indicate corresponding parts throughout the drawings.
The present disclosure seeks to combine advanced oxidation processes with nucleophilic attack. In particular, a photocatalytic porous silica-based granular media (SGM) is described herein. The SGM is capable of combining both photocatalytic and nucleophilic treatment processes and can be utilized in a packed-bed column system for continuous, passive treatment. The photocatalyst is immobilized in the media, and preloaded nucleophiles may be diffused from within the pore space. The high porosity of the media allows the PFAS degradation products to enter the SGM while filtering out turbidity without creating fouling on the surface. The resultant media is cost-effective, has a low energy consumption, can be scaled to fit the volume needed, and requires no pre- or post-treatment of the waste stream.
One aspect of the present disclosure is directed to a photocatalytic silica-based granular media (SGM). The granular media is typically useful in applications requiring the degradation of organic compounds. The media is formed from a three-dimensional polymer and comprises cross-linked silicon-oxygen bonds (i.e., Si—O—Si bonds). The silicon-oxygen bonds are formed through hydrolysis of an alkoxide precursor and a photocatalyst. The cross-linked polymer structure of the media immobilizes the photocatalyst. A foaming agent can be added, and the three-dimensional polymer can then be fired (for example, from about 200° C. to about 600° C.) to form the granular media. The media generally includes a distribution of pore spaces.
The SGM contains properties similar to lightweight aggregate in terms of density, absorption, and strength. The cross-linked polymer structure allows for the development of micro and meso pore space within the SGM. The pore space can aid the photocatalytic degradation of organic compounds by diffusing out preloaded electrophiles, nucleophiles, or salts when in contact with a liquid waste stream.
The resulting granular media typically contains a distribution of pore spaces. For example, the granular media can have a porosity of at least about 30%, at least about 40%, at least about 50%, at least about 60%, at least about 70% or more. Preferably, the granular media has a porosity of at least about 40%. In various embodiments, the granular media has a porosity of from about 30% to about 90%, from about 40% to about 90%, from about 30% to about 70%, from about 40% to about 60%, or from about 40% to about 50%. Preferably, the granular media has a porosity of from about 40% to about 60%. The fired media generally has a tortuosity of at least about 0.5, or from about 0.5 to about 2.0. In preferred embodiments, the granular media has a tortuosity of from about 0.8 to about 1.5. The media can have an overall size distribution of from about 1 mm to about 30 mm. Additionally, the media can have an internal pore size distribution of from about 100 nm to about 50,000 nm.
The alkoxide precursor can comprise, for example, a silica-containing alkoxide precursor. In other embodiments, the alkoxide precursor does not contain silica. In various embodiments, the alkoxide precursor comprises tetraethyl orthosilicate (TEOS), tetramethyl orthosilicate (TMOS), titanium isopropoxide (TTIP), or a combination thereof. Preferably, and especially when the alkoxide precursor does not contain silica, silicic acid, or another form of silica, also contributes to the Si—O—Si bonds, alone or in combination with the alkoxide precursor, such that at least a portion of the silica present within the silicon-oxygen bonds is provided by silicic acid or another form of silica independent of the alkoxide precursor. In embodiments that do not contain a silica-containing alkoxide precursor, silicic acid, or another form of silica (e.g., silica fumes, colloidal silica, and the like), is required in order to form the Si—O—Si bonds.
The photocatalyst can comprise, for example, a metal oxide. The metal oxide can comprise TiO2, TinO2n, wherein n is an integer from 1 to 10. Bi2O3, BiPO4, In2O3, Ga2O3, Sb2O3, ZnO, or a combination thereof. The photocatalyst can be combined with a dopant comprising, for example, Au, Ag, Al, C, Pt, Si, W, or any combination thereof.
Optionally, the pores of the SGM can include a surface charge, which can be achieved by an acidic or basic rinse through the addition of, for example, water, nitric acid, sulfuric acid, sodium hydroxide, potassium hydroxide, or a combination thereof. The media may optionally be treated by loading the pores of the media with amendments comprising nucleophiles, electrophiles, salts, or a combination thereof, for example nitric acid, sulfuric acid, hydrochloric acid, potassium hydroxide, sodium hydroxide, calcium hydroxide, sodium thiosulfate, or a combination thereof. It will be understood by the skilled person that loading the pores does not require that all pores be loaded. As will be further understood by the skilled person, the overall surface charge of the media is particularly important when the media is used to degrade certain PFAS compounds. As specific examples, perfluorosulfonic acids and perfluoroalkyl acid precursors degrade under basic and acidic amendments and as such, the amendments preferably comprise nitric acid, sulfuric acid, hydrochloric acid, sodium thiosulfate, potassium hydroxide, sodium hydroxide, or a combination thereof. In contrast, perfluorocarboxylic acids degrade under acidic amendments and as such, the amendments preferably comprise sulfuric acid, nitric acid, hydrochloric acid, or a combination thereof.
Also provided herein is a process for producing a photocatalytic silica-based granular media. The process generally comprises introducing a photocatalyst to an alkoxide precursor with heat and/or agitation to form a photocatalyst mixture, hydrolyzing and condensing the photocatalytic mixture until a polymer gel is formed, adding a foaming agent to create internal pore space in the media, and removing excess solution to fuse the gel into a granular media.
Thus, the novel SGM technology described herein develops a porous structure through a cross-linked matrix obtained through the hydrolysis and condensation processes. Three-dimensional cross-links are retained in the polymer structure through the introduction of a foaming agent.
The photocatalyst and alkoxide precursor are introduced, preferably with heat and/or agitation, which leads to the hydrolysis/condensation reaction. Thus, as will be readily understood by one of ordinary skill in the art, the hydrolyzing and condensing step will overlap with the introduction step. The introduction step combines the photocatalyst and alkoxide precursor. The heat and/or agitation can be provided, for example, by introducing a heat source and heating the mixture to from about 20° C. to about 110° C. and/or stirring the mixture at a range of from about 10 rpm to about 800 rpm.
The alkoxide precursor can comprise, for example, a silica-containing alkoxide precursor. In various embodiments, the alkoxide precursor comprises tetraethyl orthosilicate (TEOS), tetramethyl orthosilicate (TMOS), titanium isopropoxide (TTIP), or a combination thereof. Preferably, silicic acid, or another form of silica, is also included in the introducing step, along with the alkoxide precursor and photocatalyst. In embodiments that do not contain a silica-containing alkoxide precursor, silicic acid, or another form of silica, is required in order to form the Si—O—Si bonds. The silicic acid or other form or silica is utilized as a weak acid catalyst to favor the forward hydrolysis/condensation reaction.
The photocatalyst can be introduced to the alkoxide precursor in the form of a solid or in solution with the solvent. The solution can contain the photocatalyst in a dissolved, colloidal, or suspended state. As non-limiting examples, the solvent can comprise methanol, ethanol, nitric acid, or a combination thereof. In various embodiments, the solution can also include maleic anhydrate and/or tetrahydrophthalic anhydride.
The photocatalyst can comprise, for example, a metal oxide. The metal oxide can comprise TiO2, TiO2n, wherein n is an integer (e.g., from 1 to 10), Bi2O3, BiPO4, Bi2XO6, wherein X is a dopant, In2O3, Ga2O3, Sb2O3, ZnO, or a combination thereof. The dopant X can comprise Au, Ag, Al, C, Pt, Si, W, or any combination thereof.
In general, the photocatalyst mixture can include from about 5 wt. % to about 50 wt. %, and more particularly, from about 10 wt. % to about 40 wt. %, of the alkoxide precursor. Thus, the photocatalyst mixture can include from about 5 wt. % to about 20 wt. %, and more particularly, from about 10 wt. % to about 20 wt. % of total silica content.
In some embodiments, a stabilizing agent is also added to the photocatalyst mixture during the introducing step. The stabilizing agent can comprise, for example, dilute nitric acid, acetic acid, hydrochloric acid, potassium hydroxide, sodium hydroxide, calcium hydroxide, sodium thiosulfate, or a mixture thereof. The stabilizing agent can also be introduced with a surfactant, for example, dish soap, butadiene, styrene, benzene, or a combination thereof, or any other suitable surfactant known in the art.
As noted above, introduction of the photocatalyst to the alkoxide precursor, with the addition of heat and/or agitation, initiates the hydrolysis and condensation reaction that produces the polymer gel. Water can be added throughout the process to the gel in order to ensure complete hydrolyzed polymers. This results in the partial hydrolysis of the alkoxide precursor to form reactive monomers, condensation of the monomers to form colloid-like oligomers, and additional hydrolysis to promote polymerization and cross-linking thereby leading to a three-dimensional matrix (gel formation).
The photocatalyst mixture can be rapidly gelled or slowly gelled in order to produce a varied or aligned pore structure in the resulting polymer gel.
After gelation occurs, polymerization is completed, and a foaming agent can then be introduced in order to displace the remaining solvent. In some embodiments, the foaming agent can comprise a hydroxyl source. Non-limiting hydroxyl sources include, for example, sodium hydroxide, potassium hydroxide, ammonium hydroxide, or a combination thereof.
Excess solution can be removed from the polymer gel through firing, desiccation, drying, or exposure to ambient environmental conditions. Thus, the process preferably comprises the step of firing the polymer gel at a low temperature (e.g., from about 200° C. to about 600° C., from about 200° C. to about 550° C., or from about 200° C. to about 500° C.) in order to obtain the porous granular media.
The rate at which the excess solution is removed from the gel or the manner in which the removal occurs is important for producing interconnected pores. Slower removal processes produce more interconnected pores though the required time can be lengthy. Faster removal processes produce less interconnected pore spaces and longer diffusion times but faster production times.
Further, silica content plays a role in the stability of the SGM post-firing, as well as the ability to fixate the catalyst within the media without embedding it. Including too little silica may result in large void formation during firing that creates a non-homogenous pore size and distribution throughout the structure. Thus, large void spaces formed during the rapid evaporation and activation of the foaming agent during firing are typically more readily observed in SGM having lower concentrations of silica. Foaming agent-induced pores decrease in abundance and relative size as silica content is increased. Because of this, increased silica content lends itself to an increase in the tortuosity and permeability of the pore space. While the cross-link formed pore-spaces appear smaller in average size, they also can be more interconnected, homogenous in distribution, and uniform in size. The increase in cross-linked structures is caused by the increase in nucleation sites the silica content brings, thus creating a more durable SGM. Increases in durability and strength of the SGM are demonstrated to improve as the degassing void space decreases and silica concentration during gelation increases.
Optionally, the process can also include, after the removing step, making surficial charge adjustments by an acidic or basic rinse through the addition of water, nitric acid, sulfuric acid, or a combination thereof. The process can also optionally include adding amendments, preferably nucleophiles, electrophiles, salts, or a combination thereof, through loading the media pore space with nitric acid, sulfuric acid, hydrochloric acid, potassium hydroxide, sodium hydroxide, calcium hydroxide, sodium thiosulfate, or a combination thereof. As will be understood by the skilled person, the type of amendment used determines the overall surface charge of the resulting media, and is particularly important when the media is used to degrade certain PFAS compounds. As specific examples, perfluorosulfonic acids and perfluoroalkyl acid precursors degrade under basic and acidic amendments and as such, the amendments preferably comprise nitric acid, sulfuric acid, hydrochloric acid, sodium thiosulfate, potassium hydroxide, sodium hydroxide, or a combination thereof. In contrast, perfluorocarboxylic acids degrade under acidic amendments and as such, the amendments preferably comprise sulfuric acid, nitric acid, hydrochloric acid, or a combination thereof.
The pore space dictates the ability of an acid or base added after formation of the SGM (e.g., in the surficial charge adjustment or amendment step) to leach and the rate at which it diffuses from the SGM.
Acid stabilization methods can also be employed after the removing step to extend the life cycle of the media and give the outside surface of the SGM a positive charge, which improves reactivity. This process also dissolves free sodium hydroxide radicals on the surface of the SGM, thereby opening a direct path for UV interaction. Importantly, excess hydroxyls from the SGM synthesis will still remain in the pore structure (see
The present disclosure also relates to a method of degrading one or more organic compounds. The method generally comprises introducing the organic compound(s) to the silica-based granular media described herein and irradiating the combination with electromagnetic radiation. The organic compound can comprise, for example, a perfluoroalkyl compound, a polyfluoroalkyls compound, a pharmaceutical compound (such as rifampin, acetaminophen, or a combination thereof), a textile dye (such as methylene blue, rhodamine red, azure A, methyl orange, or a combination thereof), or any combinations thereof.
As aforementioned, it is important that the granular media have the correct surface charge by, for example, addition of amendments, particularly in situations where PFAS compounds are degraded. In some circumstances, when the organic compound comprises a PFAS compound, the process of making the granular media may require cycling between acidic and basic amendments in order to fully degrade the PFAS compound.
The organic compound(s) can be provided and introduced to the granular media in any acceptable form. Preferably, the organic compound(s) are in a solution or in the form of an aerosol. For example, the SGM can be used in HVAC or air recycle systems in which aerosols or charged particulates can be attracted to the SGM under electrostatic means, sorbed to the surface, and then treated. The SGM can also be utilized as a thin film or coating when extruded.
The SGM is reusable for multiple cycles of treating, breaks down very slowly, and produces few by-products during degradation. Although some by-products may be produced, they are generally non-hazardous. In preferred embodiments, at least about 80%, at least about 85%, at least about 90%, at least about 95%, or at least about 99% of the organic compound(s) is fully degraded (e.g., mineralized).
Typically, the electromagnetic radiation that the one or more organic compounds is exposed to is ultraviolet radiation with a wavelength of from about 100 nm to about 400 nm.
The present disclosure is also directed to a reactor for degrading the one or more organic compounds. That is, the reactor can be used in the above-described methods to introduce a solution, aerosol, or other appropriate composition form containing at least one of the organic compounds described above to the silica-based granular media described herein to degrade and remove the one or more organic compounds from the composition.
Referring in particular to
Both types of reactor allow for passage of a composition containing the organic compound(s) through a treatment area packed with the silica-based granular media described herein, such as column 12 depicted in
The continuous flow reactor can contain one column 12 or a series of columns (e.g., two, three, or four columns) adjacent to one or more UV light sources. Alternatively, the UV light source may be embedded in treatment 14 area to create a serpentine path in the treatment area through which the incoming stream can flow. The UV light source may be embedded in or adjacent to the recirculation reactor.
In this way, the one or more organic compounds pass over the SGM and are exposed to a UV light source 22 in the reactor, typically having a wavelength of from about 100 nm to 400 nm, to activate the SGM and degrade the organic compound(s). The UV light can contain ozone (185/254 nm wavelength) or can be ozone free (254 nm wavelength). More than one UV light source can be included in the reactor and is typically dependent on the size (e.g. length or diameter) of the reactor. That is, a longer reactor may require an increased number of UV light sources.
In various embodiments, the UV light source can comprise a standard low-pressure mercury lamp, an amalgam lamp, a combination thereof, or any other acceptable light source known in the art. Typically, the UV emission from the light source is from about 1 watt to about 50 watts. The UV light source can be encased in a quartz sleeve.
As mentioned, the reactors contain an inlet 18 to allow incoming passage of a waste stream (containing the one or more organic compounds) and an outlet 20 to allow removal of the outgoing stream (with at least a portion of the one or more organic compounds removed). A pump 24 can also be used with the inlet to keep a steady flow of waste stream.
Having described the invention in detail, it will be apparent that modifications and variations are possible without departing from the scope of the invention defined in the appended claims.
The following non-limiting examples are provided to further illustrate the present invention.
The following Example evaluates the ability to use SGM as a destructive technology for PFAS, and specifically PFOS. PFOS was selected as a surrogate compound for a concentrated aqueous waste stream due to the relative difficulty of degradation when compared to PFOA. Testing was carried out in polypropylene batch reactors with a borosilicate cover to minimize evaporation (see
The single layer of SGM consisted of approximately 0.10 g of photocatalyst with reactive species only on or near the surface within a direct path of the UV light. UVA/B/C lights operating over a wide range of wavelength spectra from 550 nm to 250 nm were precisely placed 10.16 cm (4 in.) above the targeted SGM. Aqueous stock solution of 500 mg/L PFOS were prepared for serial dilutions utilized in batch reactors. Dilutions of the stock solution were prepared with DI water, 1 M sodium hydroxide, or 1 M sodium thiosulfate. The final analyte concentration of 50 mg/L PFOS was fixed throughout all experimental testing. While the SGM is capable of diffusing out preloaded salts or nucleophiles during treatment, dilutions with those solutions occurred during batch reactor experiments. This can be attributed to the small mass or dose of SGM used for testing compared to what would be present in a packed column. Therefore, nucleophile solution concentrations were calculated to equate to the diffusive abilities of the SGM during column reactors for proportionate pore volumes.
Reactor contact time ranged from 0 to 360 min, with aliquot sampling of 600 μL (100 μL for LC/MS and 500 μL for IC) occurring at various time intervals. Samples were contained in polypropylene microcentrifuge tubes and stored in a dark room at 4° C. prior to analysis. All analysis occurred within a maximum 24-h window after extraction from batch reactor experiments were completed in order to mitigate any external influence of contamination of the samples.
Sample preparation and LC/MS methodologies were modified from ASTM D7979-20 and EPA 537. Each 100-μL sample was centrifuged for 20 min at 14,000 rpm while maintained at a temperature of 4° C. A 5-μL aliquot of the supernatant was extracted and diluted in 995 μL of solvent S (1:1 MeOH/H2O+0.1% glacial acetic acid), thoroughly mixed, and again centrifuged. A 40-μL aliquot of the supernatant was then further diluted with 160 μL of solvent S. MPFOS (sodium perfluoro-1-[1,2,3,4-13C4]-octanesulfonate) was added to solvent S prior to the second dilution as an internal standard. The final resulting concentration was 50 ng/mL MPFOS in each sample. Each sample mixture was cortex mixed for 30 s and then injected into the UHPLC.
LC/MS analysis was performed on a SHIMADZU Nexera XR (40-Series) UHPLC system coupled with a SHIMADZU 9030 Q-ToF Mass Spectrometer+DUIS ionization source. A RESTEK Raptor ARC-18 (100 mm length, 2.1 mm internal diameter, 1.8 gm particle size, 90 Å pore size) analytical column was used solely for these experiments and stored between batches in order to eliminate contamination. Analyte elution was performed using gradient elution with 25-mM solution of ammonium acetate in water containing 3% (v/v) acetonitrile (Solvent A) and 100% acetonitrile (Solvent B) at a flow rate of 0.35 mL/min. The temperature of the autosampler and column oven were set at 4° C. and 40° C., respectively. LC/MS analyses were performed using 1-μL sample injections using negative ionization mode-based detection.
Throughout experimental analysis, samples were analyzed for PFOS reduction and screen for possible by-products. Untargeted analysis was performed on each sample using liquid chromatography quadruple time-of-flight mass spectrometry (LC-QToF-MS). While all m/z values observed were reviewed as a possible by-product, perfluoroalkyl carboxylic acids (PFCAs) were specifically screened for by-products and were quantified using Wellington standards.
Free fluoride in solution, or aqueous fluoride, was measured using a DIONEX IC System (ICS-90) with an automated sampler (AS40) Chromeleon 6.80. The ICS-90 system contained a 4×250-mm AS23 analytical standard bore column (Part #064149), An AG23 guard standard column (064147), coupled with an AMMS 300 chemically driven suppressor (064558), and a D5 stabilizer conductivity cell. A 50-μL injection loop was used as a standard for all samples and standards. Individual samples of 0.5 mL were diluted to 5 mL with DI water in order to reduce the solution pH below 10 S/U. This preparation was done in part to allow the bicarbonate eluent to buffer the injected solution but also to reduce the peak-to-peak interference between fluoride and chloride, and to extend the baseline near the water dip. Eluent stock solution was prepared consisting of 450 mM of sodium carbonate and 80 mM of sodium bicarbonate. 30 mL of eluent stock solution was then diluted to 2,000 mL in a mixture of DI water with 3.5% methanol by volume to minimize organic buildup within the system. The increased eluent concentration from the traditional 100×dilution from stock was chosen to preserve baseline conductivity, create better peak separation in the chromatography, and optimize the eluent buffering capacity. Regenerant solution was diluted from 75 mL of 2.0 N sulfuric acid to 2,000 mL with DI water. A 5% relative standard deviation (RSD) was used for the triplicate analysis of analytes measured with the ICS-90 following the standard method. Fluoride calibration standards of 0.1, 1.0, 5.0, 10.0, and 25.0 mg/L were diluted from a 1,000 mg/L stock and ran prior to the analysis of each set of samples. Each sample run time was increased to 32.5 min to ensure peak separation with an average pressure of 1,900 psi. A blank/wash of DI water was run between each sample to ensure no contamination in the peak area from the previous sample occurred. Sample and standard preparation and analysis, along with quality control, were consistent with EPA method 300.0.
SGM was adhered to the bottom of the polypropylene reactors 24 h prior to testing. Initial batch reactor experimentation was performed on the SGM with water as an addition to determine the effect of filling the pore space with solution in comparison to a dry media. This experiment was performed to separate the adsorption to the SGM from the absorption. DI water was poured on the single layer of SGM 12 h before testing. The media was dried to saturated surface dry (SSD) condition just before testing was initiated to ensure all pores were filled but excess water was not on the surface.
Considering that the absorption capacity of the SGM only allowed for a total volume of 1 mL to be present in the pore space, it can be determined that the increased rate was not due to dilution.
Analysis performed using LC-QToF-MS and ICS-90 was performed in triplicated and is presented in
The preferred amendments for the SGM combined degradation pathway and complete mineralization of PFOS are nucleophiles; however, the instability of the Si—O bonds at elevated and sustained pH above 12.0 S/U is problematic. Sodium hydroxide was chosen to challenge the durability of the SGM in sustained elevated pH conditions. Sodium thiosulfate, a weaker nucleophile, was selected because it raises pH less than 10 S/U, allowing for a sustained dual attack. While the NaOH addition did yield higher removal of PFOS than the SGM without any addition, the outer shell of the SGM began to dissolve in the high pH. As this continued, the solution became very turbid and hindered the photocatalytic attack, resulting in lower capacitive removal than the Na2S2O3 addition. Na2S2O3 addition consistently performed at high efficiency, with >99% removal in 30 min.
One line of evidence that defluorination and therefore destruction of PFOS is occurring is the measurement of anionic fluoride using IC. Na2S2O3 addition performed the best of all treatment types, and samples of this treatment were then analyzed for free fluoride in solution over time, along with no addition treatment as a comparison (
In order to render a positive charge to the surface cell of the SGM, media was stabilized in acidic conditions. Further experimental testing was performed in triplicate to validate that the acid stabilization of SGM did not hinder the photocatalytic degradation of PFOS. Experimentation included analysis of both PFOS removal (
The presence and production of PFOS by-products were analyzed using LC-QToF-MS. During LC/MS analysis, the m/z value associated with C7F13O2− was initially produced in the no addition treatment at 15 min of experimental treatment. While analysis was performed to identify the production of other by-products, only C2F3O2− was consistently present at all time intervals. The degradation pathway theorized, from the by-products detected, is a combined free hydroxyl and free radical attack that inundates the C—S bond, which is replaced with an alcohol. This results in a PFCA upon stabilization. The proposed degradation pathway is supported by the presence of perfluoroheptanoic acid. Perfluoroheptanoic acid was not identified; however, this could be due to the rapid degradation of the PFOS prior to the first time increment. The free radicals generated by the UV/TiO2/H2O interaction and the nucleophiles develop a dual attack on the Cx carboxylate chain in a stepwise systematic release of HF until defluorination. The only by-product observed in the nucleophile addition treatments was C2F3O2− and aqueous fluoride. This is attributed to the rapid degradation of PFOS and the affinity for by-products to enter the SGM once the functional head is removed. This phenomenon occurs as the free radicals begin to degrade the C—F chains, and fluoride begins to bond to the silica in the SGM. Free radical generation continues to interact with the degraded chains and break down the by-products as they are produced. As the defluorination of these C—F chains continues, fluoride is released into solution, which bonds to the silica present in the SGM. Assuming the C—F chains degrade at a constant rate, F− is released rapidly at the same time interval, which explains why a large peak of F− concentration can be seen at about 60 min.
While multiple by-products were observed throughout the experiment, only the aqueous fluoride in solution was quantified. The total organic fluoride in the solution was calculated to be 32.3 mg/L. Peak fluoride recovery is reflected in
Scanning electron microscopy (SEM) and energy dispersive x-ray spectroscopy (EDS) analysis was performed and is presented in
After confirming that the fluoride was present in the SGM, further SEM forensic analysis was performed to detect the possible presence of precipitated by-products in the SGM. Considering the relative size difference of angstrom-scale PFOS compared to the SGM, precipitates of by-products and fluoride were found by zooming in on hot spots of fluoride and toggling between elemental mapping. Once an area could be verified as C—F chains, the SEM was switched to secondary electron (SE) mode to achieve clearer imagery.
Initial SEM/EDS imagery confirmed that the by-products have an affinity to diffuse into the SGM and continue to degrade. To detect precipitated or mineralized fluoride after it breaks from the C—F chains and renters the SGM, further SEM analysis was performed. A no addition treatment SGM specimen was selected that had been in contact with the 50 mg/L PFOS in the presence of UV light for 120 min. According to the IC analysis presented previously, aqueous fluoride should have re-entered the SGM at this time.
An SGM was developed using tetraethyl orthosilicate as the alkoxide precursor and titanium dioxide as the photocatalyst. Sodium hydroxide was introduced after formation of the polymer network as the foaming agent. The media was then fired. The resulting SGM was washed in a weak acid to dissolve surficial sodium hydroxide and bring the external surface of the SGM to a neutral pH.
Multiple column reactors were constructed to determine optimal SGM treatment conditions and reaction kinetics and assess scalability. Four variations of amended SGM (denoted Treatment A-D) were compared against four controls:
Each experimental treatment was conducted over the course of four hours within each column providing one hour of contact time, so that treatment variables could be compared at different contact times as well as amendment types.
High basicity columns (Treatments C and D) were packed with a lightweight porous aggregated preloaded with sodium hydroxide. All amendments were preloaded into their respective media by soaking them in 1 M solutions of NaOH or Na2S2O3 for 72 hours prior to testing. Following loading of the SGM with Na2S2O3, media was placed in a 105° C. oven and allowed to dry for 72 hours. Drying the media in this manner left residual salt precipitates within the pore space, which then rehydrated in contact with the filtrate solution. The lightweight aggregate media soaked in high pH solutions were decanted, dried to saturated surface dry conditions to minimize dilution effects, and then packed within the column prior to proceeding with experiments.
Columns were assembled in a series of four DWK LIFE SCIENCES KIMBLE KONTES FLEXCOLUMN Economy Columns. Each column measured 15 mm in diameter, 200 mm long, and held a volumetric capacity of 35 mL. Once packed, columns were oriented in parallel such that a 20-μm pore size filter disc was positioned in the outlet of each column to prevent media migration, plugging of tubing, or fouling of sampling ports. Glass components of the column bodies were manufactured from 33 expansion, low extractable borosilicate glass conforming to USP Type I and ASTM E438, Type I, Class A requirements. 33 expansion borosilicate glass has a low-potassium content in order to yield a very high UV transmission, second only to quartz-based glass in UV transmission. Each column was assembled in an up-flow reactor configuration to release trapped air within each column reactor. At the outlet and inlet of each reactor a three-way stopcock valve was installed, with each outlet valve connected to the next inlet valve by high density polyethene tubing. Each series of four columns was leveled and fixed to a UNISTRUT rack system. SAVIO Skimmer UV lights (57-Watt lamps) from AQUA ULTRAVIOLET were placed on either side of each column. A 100-mL polypropylene syringe was filled with the aqueous film forming form (AFFF)-impacted stormwater and attached to a NE-300 JUST INFUSION syringe pump.
Each column was packed with 10 grams of granular media yielding an average pore volume of 20 mL. Therefore, the syringe was set to pump at 20 mL/h allowing for a sample aliquot to be taken from the effluent of the column after an hour. In Treatments C and D, a full packed pore volume of filtrate (20 mL) was passed through the reactor prior to the start of the experiment, in order to mitigate the impact of dilution of the waste stream by NaOH solution. The second filtrate pore volume (initial 20 mL plus subsequent 20 mL) was sampled and reported. To facilitate direct comparison of results, data presented below are representative of the second pore volume of filtrate. After flushing of residuals was completed, discrete sample aliquots were collected at the outlet sampling port of each column every hour for four hours. Samples were stored in compatible polypropylene micro-centrifuge containers at 4° C. until analysis. All experiments were performed in duplicate with minimum and maximum values reported as observed.
Modified methods ASTM 7979 and EPA 537 were followed for sample preparation and LC-MS analysis of PFAS. Each sample aliquot was centrifuged for 20 minutes at 14,000 revolutions per minute (rpm) while maintained at a temperature of 4° C. Aliquots of 20 μL were added to 180 μL of solvent S (1:1 MeOH/H2O+0.1% glacial acetic acid) that contained 0.01 μg/mL of MPFAC (WELLINGTON LABORATORIES) as the internal standard. The mixture was vortexed for 30 seconds and again centrifuged. The supernatant was injected into the LC-MS system. In addition to the internal standard, one sample in each batch was performed in duplicate to ensure statistically invariance.
Separation of analytes was carried out on a SHIMADZU Nexera XR (40-Series) UHPLC system by injecting 5 μL of sample into a RESTEK Raptor ARC-18 analytical column (100 mm length, 2.1 mm internal diameter, 1.8 μm particle size, 90 Å pore size). Gradient elution of 25 mM solution of ammonium acetate in water containing 3% (v/v) acetonitrile (solvent A) and 100% acetonitrile (solvent B) at a flow rate of 0.35 ml/minute was used as the mobile phase. The gradient began with isocratic flow of 5% solvent B for the first 3 minutes which was followed by a linear gradient of 5% to 95% solvent B from 3 to 28 minutes; the gradient decreased to 5% solvent B at 28.1 minutes and maintained constant until the minute 30. The autosampler and the column were maintained at 4° C. and 40° C., respectively, during analysis.
Untargeted analysis and further detection of PFAS reduction in sample aliquots was performed on a SHIMADZU 9030 Q-TOF Mass Spectrometer coupled to the UHPLC system. The mass spectrometer was equipped with a DUIS ionization source and was operated in negative ionization mode. Concentration calibrations were developed and quantified using WELLINGTON LABORATORIES standards and quantified on the LC-QToF-MS. Relative concentrations were determined for all analytes not found in the standards.
Aqueous fluoride was measured using a DIONEX ion chromatography system (ICS-90). Individual samples of 0.5 mL were diluted to 5 mL with DI water in order to reduce the solution pH below 9. If the pH of the sample was greater than 11, the 0.5 mL aliquot was diluted in 0.1 M nitric acid. Sample and standard preparation and analysis, along with quality control samples, were consistent with EPA Method 300.0.
The stormwater was analyzed on a LC-QToF-MS for PFAS discussed in this section. Table 1 summarizes the identified PFAS and their concentrations in untreated stormwater. Untargeted analysis identified 17 possible PFAS through suspect screening. Of the 17 identified PFASs, 11 were PFAAs and 7 were PFAA precursors. Fluorotelomer sulfonates accounted for 83% of PFAS mass in the stormwater, while PFSAs and PFCAs accounted for 17% of the total PFAS, 10% and 7% respectively. Fluorotelomer sulfonates can transform into biologically inert PFAAs. External lab testing verified the presence of the 204 PFSAs, PFCAs, 8:2 FTS, and 6:2 FTS (Table 1). Studies have shown 6:2 fluorotelomer thioamido sulfonate (6:2 FtTAoS) is one of the primary PFAS present in AFFF from multiple manufacturers. Biotransformation of the manufactured compounds was predicted based on the large concentration of 6:2 FTS. Two known transformation products were detected—6:2 fluorotelomer sulfoxide amido sulfonate (6:2 FtSOAoS) 208 and 6:2 fluorotelomer sulfone amido sulfonate (6:2 FtSO2AoS). Further transformation into PFCAs likely occurred, based on the relatively high concentrations of perfluorohexanoic acid (PFHxA) and perfluoropentanoic acid (PFPeA). 6:2 fluorotelomer sulfonyl propanoic acid (6:2 FtSO2PA) has previously been reported in AFFF-impacted waters and has been identified as a fluorosurfactant ingredient utilized in some AFFF mixtures. 6:2 fluorotelomer sulfonamido propyl betaine (6:2 FTSA-PrB) has been identified at multiple sites and is another fluorosurfactant used in AFFF to replace 215 PFOS. No other 8:2 fluorotelomers and no 4:2 fluorotelomers were identified. In addition to the fluorotelomers identified, a large concentration of PFSAs were identified. No other PFAS were identified in the parts per billion (ppb) range; however, it is possible for other compounds to be present at lower concentrations. Based on the presented compounds, it is likely that multiple AFFF mixtures were released at the site. This is a common scenario at many AFFF sites where multiple products and formulations were used over time.
+number of perfluroinated carbons in the molecule
All 17 PFAS identified in the stormwater were evaluated over the four-hour experiments.
The degradation of PFSAs, depicted in
PFCAs constituted the lowest concentration of PFAS in the untreated stormwater.
In addition to the four treatments, a series of controls were performed on the mini-column reactors to help identify the reaction mechanisms occurring in the system. Variables of the treatment system included adsorption to column reactor, adsorption to SGM, UV photolysis, photocatalysis, nucleophile addition, and heat from the lamp. Results from three of the controls are depicted in
The first control, denoted as No UV/No SGM, quantified PFAS adsorption to the column reactor. This control column yielded a 13% decrease in PFAS in the first pore volume of solution after running through all four columns. Although a significant decrease was observed, the effect was only observed during the first pore volume flush, validating the explanation that all available sorption sites in the reactor are fully utilized during the first pass of solution. Therefore, since the treatment data depicted in
The second control consisted of stormwater passed through the column reactor with UV lamps on and is labeled as photolysis in
The third and most important control is PFAS adsorption to SGM media without UV activation, denoted No UV in
A fourth control was performed to determine the effect of heat from the lamps on experimental data. Equivalent heat result without irradiation yielded a 5% reduction in PFAS compared to data collected during the No UV control. This can be attributed to residual nucleophiles in the SGM from the foaming agent reacting with the heat to aid in PFAS degradation. However, results indicate that PFAS degradation is predominantly from photocatalytic treatment and is enhanced by the addition of more nucleophiles.
A summary of total PFAS degraded in all four treatments is shown in
Aqueous fluoride in solution was measured in the effluent of each column reactor. A theoretical organic fluoride content of 1065.83±39.33 ppb was calculated from the 17 identified PFAS compounds; however, other unidentified fluorinated compounds could be present in the stormwater. Table 2 presents the average aqueous fluoride for each treatment over the 4-hour reaction period. While it may be assumed that the fluoride will continually increase over the four hours, with the fourth column effluent yielding the sum of total defluorination, free fluoride in solution has shown the ability to bond to free silica in the SGM. This was verified by fluctuating fluoride concentrations between each column reactor in the stormwater (Table 2). The high affmity of silica to fluoride results in fluoride mineralization by bonding to silica and therefore total defluorination cannot be quantified without a complete fluoride mass balance within the stormwater and SGM. However, high concentrations of aqueous fluoride were still measured, validating degradation and defluorination of PFAS in the stormwater. In Treatments C and D, defluorination of fluorotelomers is assumed because the measured fluoride is above the theoretical fluoride concentration of 209.32±3.16 ppb available from the PFSAs and PFCAs. All treatments showed recoverable aqueous free fluoride concentrations greater than the theoretical amount of the PFCAs, indicating defluorination of PFSAs and PFAA precursors. The minimum defluorination in each treatment can be determined from the theoretical organic fluoride calculated and the aqueous fluoride recovered in the column 4 effluent. The average minimal defluorination in Treatments A-D are 12.4%, 47.9%, 91.2%, and 69.6%, respectfully. The maximum aqueous fluoride recovery (1024.8 ppb) was seen in Treatment C. Maximum PFAS reduction in Treatment C after 4 hours was 90%, therefore, the theoretical amount of fluoride that could be liberated or defluorinated was 959.25 ppb, which would equate to 106.83% defluorination. This indicates the presence of some unidentified fluorinated compounds.
Evidence of PFAS degradation during treatment was provided by the increase in PFCA intermediate transformation products and generation of fluoride. Stormwater used in this study had a complex and only partially characterized mixture of PFAS. PFAS adsorption to the SGM further complicated the ability to use trends in PFAS concentrations to understand the reaction mechanism.
Treatment C and D both exhibited rapid degradation of 6:2 FtSO2AoS, 6:2 FtSOAoS, 6:2 FtTAoS, 6:2 FTSA-PrB, and 6:2 FtSO2PA. A significant reduction of 6:2 FTS was also observed, but not until after column 1. Therefore, it is theorized that 6:2 fluorotelomers degraded to form 6:2 FTS. 6:2 fluorotelomer unsaturated acid (6:2 FTUA) was measured in the effluent of columns 1-4, but was not present in the initial solution, suggesting it formed during degradation of 6:2 FTS. Based on general concentration trends, 6:2 FTUA is thought to further degrade to form PFHXA and PFPeA. This pathway is also supported by relatively high concentrations of PFHxA and PFPeA in the initial solution, indicating that PFAA precursors may have already began biotransforming into PFCAs.
It is more difficult to trace the degradation pathway of PFSAs, due to complications from the solution matrix. SGM have previously been demonstrated to show that surrogate PFOS degrades into PFCAs which continued to defluorinate to form shorter-chained PFCAs until the compound has been completely degraded. The ability to rapidly degrade PFSAs using a photocatalytic technology comes from coupling the attack with a nucleophile attack. The PFAS is able to both absorb to the SGM and adsorb to the photocatalysis, thus creating a direct transport pathway for free radicals and free hydroxyls to attack the functional 395 group and subsequent C—F chain. With PFOS and PFHxS accounting for the highest concentrations of PFSAs present, the temporary increase in PFOA and PFHxA concentrations, followed by a decrease in concentrations, are consistent with this degradation pathway. PFCAs likely degrade to form shorter-chained PFCAs until defluorination is complete. Additionally, trifluoroacetic acid and significant concentrations of aqueous fluoride were detected in the effluent of the columns, verifying degradation.
Process characterization and discrete particle testing were performed using thermal gravimetric analysis and mercury intrusion porosimetry. A diffusion pore space capable of storing an absorbed solution is observed in the cross-section of SGMs in
Thermal gravimetric analysis (TGA) was performed on the SGM both before and after the firing process. From 25° C. to 800° C., 50% of the weight of the sol-gel was lost due to the evaporation of the liquid matrix surrounding the cross-linked solid network (
Although structural and thermal strengths are important for SGM longevity, the purpose of the SGM is to function as a photocatalytic granular media for water treatment. Photocatalytic reactivity testing was performed on all four variations of SGM utilizing methylene blue and the same reactor setup, as described in the technical approach of Example 1.
Photocatalytic degradation of methylene blue in SGM is shown to follow first-order degradation kinetics.
To show both the photocatalytic sustainability and longevity of the SGM, cyclic testing of two variations of the SGM was performed with methylene blue. The initial reactor volume of 35 mL was replaced with new solution 10 mg/L methylene blue every 60 minutes (
Hydroxyl diffusivity was measured for the media with various amendment types. Hydroxyl diffusion is important for a dual nucleophile and photocatalytic attack of PFAS and to demonstrate the viability of the material to perform both attacks in a coupled manner. Two nucleophiles, sodium thiosulfate (weak) and sodium hydroxide (strong), were utilized for the dual attack. The coupled degradation mechanisms allow for rapid cleaving of the sulfate functional group from the rest of the C—F chained backbone. In addition, nitric acid and sulfuric acid amendments were examined for diffusivity because elevated pH hinders the degradation of PFCAs due to repulsion of anionic PFAS and the photocatalyst. pH diffusion tests were performed over 6 hours; results are depicted in
This Example uses a similar process to the synthesis of an SGM with titanium dioxide. However, because of the density of bismuth trioxide, the photocatalyst is first dissolved in nitric acid before a silica acid mixture and alkoxide precursor (tetraethyl orthosilicate) are added. In general, the process requires adding 100 mL HNO3, 1-10 g Bi2O3, 0.1-0.2 g SiO2, 50 mL water, and 150 mL TEOS.
Once the gel is completely polymerized in a cross-linked structure, sodium hydroxide is added as a foaming agent. It is theorized that the form of bismuth at this point is most likely bismuth hydroxide (Bi3(OH)3). Bi-SGM can be used in-line with Ti-SGM or as a replacement. Bi3O3 (2.1-2.8 eV) has a lower band gap than TiO2 (3.2-3.5 eV), thus having the potential to have an increased oxidation power. Bi-SGM has been validated in batch reactors, mini-column reactors, and submerged lamp reactors. Bi-SGM has been show to defluorinate a variety or per- and poly-fluoroalkyl substances at an irradiation wavelength of 254 nm.
When introducing elements of the present invention or the preferred embodiment(s) thereof, the articles “a”, “an”, “the” and “said” are intended to mean that there are one or more of the elements. The terms “comprising”, “including” and “having” are intended to be inclusive and mean that there may be additional elements other than the listed elements.
In view of the above, it will be seen that the several objects of the invention are achieved and other advantageous results attained.
As various changes could be made in the above materials, processes, and methods without departing from the scope of the invention, it is intended that all matter contained in the above description and shown in the accompanying drawings shall be interpreted as illustrative and not in a limiting sense.
This application claims the benefit of U.S. Provisional Application No. 63/114,291 filed Nov. 16, 2020, the contents of which are incorporated by reference in their entirety.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/US2021/059504 | 11/16/2021 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 63114291 | Nov 2020 | US |
