Engineered sorbents and process for selective removal of perfluorinated organics species from water or air

Abstract

A process for purifying an initial water source comprising water and an initial amount of one or more per- and polyfluorinated organics (PFAS), the process comprising the steps of contacting the initial water source with a sorbent composition comprising a sorbent substrate and a surface modifier composition to yield purified water and a waste composition comprising PFAS.

Description

UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 ENGINEERED SORBENTS AND PROCESS FOR SELECTIVE REMOVAL OF PERFLUORINATED ORGANICS SPECIES FROM WATER OR AIR RELATED APPLICATION

[0001] This application claims priority to US Provisional Application 63 / 619,247 filed on January 9, 2024, the content of which is incorporated herein by reference in its entireties for all purposes. FIELD

[0002] The present disclosure relates to a composition and a process for removing per- and polyfluorinated organics (PFAS) from contaminated water or air. In particular, the disclosure relates to the use of a treated, cellulose-based sorbent to remove PFAS from contaminated water or air. BACKGROUND

[0003] PFAS are often referred to as “forever chemicals” because they are water soluble and do not readily break down in the environment. PFAS are very stable manmade chemicals that have properties that allow them to repel both water and oil. There are many types of PFAS chemicals, having different chemical compositions and different properties, , resulting in ranges of toxicity.

[0004] Ingestion of PFAS, e.g., by eating or drinking food or water that contain PFAS, the PFAS are absorbed and can accumulate in bodies or organisms and the environment. Scientists are still learning about the health effects of exposures to PFAS. Some concerns relate to the effects on women who are pregnant or likely to become pregnant and on children.

[0005] PFAS are very expensive to remove via conventional water treatment, due to their high hydrophilicity and their low concentrations in the contaminated water or air.

[0006] Current treatment programs typically use a process of one-pass loading of granulated, active carbon followed by disposal of the granulated activated carbon. In other cases, separation processes, e.g., absorption, ion exchange, or membranes e.g., crystalline organic framework membranes, continue to be developed to meet the challenge of cost- effective and environmentally sound clean-up of environmental contamination.

[0007] Conventional granulated activated carbon one-time sorption is commercially available. Other technologies currently under development relate to polymers andUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 functionalized polymers. However, these often leave much room for improvement with regard to PFAS removal and / or are cost-prohibitive.

[0008] Even in view of conventional treatment processes, the need exists for improved processes for removing PFAS from contaminated water that have a high removal efficiency, operate at a low pressure drop, are potentially able to be recycled / re-used, and / or are cost- effective. SUMMARY OF INVENTION

[0009] Disclosed here is a process for purifying an initial liquid or air source comprising liquid or air mixed with an initial amount of one or more of per- and polyfluorinated organics (PFAS). In some embodiments, the process includes (a) contacting the initial liquid or air source with a PFAS sorbent composition comprising a sorbent substrate and a surface modifier composition, and (b) obtaining a purified liquid or air product comprising a reduced amount of PFAS. A reduced amount of PFAS means the product obtained in step (b) contains lower level of the one or more PFAS than the initial amount of the one or more PFAS in the initial liquid or air source.

[0010] In one aspect, the initial liquid or air source is an initial water source, and the purified liquid or air product is purified water. In another aspect of the present disclosure, the initial liquid or air source is an initial air source, and the purified liquid or air product is purified air. In another aspect, the purified water or air may be subject to further treatment to remove other contaminants.

[0011] In some embodiments, the surface modifier composition comprises an alkyl silane. In one aspect, the alkyl silane is a linear alkyl silane. In another aspect, the alkyl silane comprises a carbon chain having from 6 to 20 carbons. In another aspect, the alkyl silane is one or more of trimethoxy(octadecyl)silane (TMOS), triethoxy(octyl)silane (TEOS), or (3- aminopropyl) triethoxysilane (APTES), or a combination thereof.

[0012] In some embodiments, the sorbent substrate comprises a cellulose-based substrate, or optionally a cellulose sponge.

[0013] In some embodiments, the surface modifier composition comprises charged silanes and uncharged silanes. In one aspect, the ratio (w / w) between charged silanes and uncharged silanes ranges from 1:1 to 0.1:100, or from 0.1:1 to 0.1:10, or from 0.1:2 to 0.1:5.

[0014] In one aspect, the surface modifier composition has a pH between 6 and 9, or between 7 and 9. In another aspect, the surface modifier composition is an alkaline solution.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695

[0015] In some embodiments, the surface modifier composition increases hydroxyl group content on the sorbent substrate.

[0016] In some embodiments, the surface modifier composition comprises a metal oxide, optionally an aluminum oxide.

[0017] In some embodiments, the surface modifier composition comprises a divalent cation, In one aspect, the divalent cation is magnesium or calcium. In one aspect, the concentration of the divalent cation is between 1 mM and 200 mM, or between 50 mM and 100 mM, or about 75 mM.

[0018] In some embodiments, the surface modifier composition imparts a charge to the sorbent substrate.

[0019] In some embodiments, the PFAS is one or more member selected from the group consisting of perfluorooctanoic acid (PFOA), perfluorooctanesulfonic acid (PFOS), perfluorohexanoic acid (PFHxA), perfluorohexanesulfonic acid (PFHxS), perfluorobutanoic acid (PFBS), and / or perfluorobutanesulfonic acid (PFBS).

[0020] In some embodiments, the purified water or air comprises less than 75%, or less than 60%, or less than 50%, or less than 38%, or less than 25% of the initial amount of PFAS in the initial water or air source. In one aspect, more than 25%, or more than 40%, or more than 50%, or more than 62%, or more than 75% of PFAS is removed from the initial water or air source.

[0021] In some embodiments, the purified water stream comprises less than 10,000 ppm PFAS, or less than 1000 ppm PFAS, or less than 200 ppm PFAS, or less than 100 ppm PFAS.

[0022] In some embodiments, the contacting step yields a spent sorbent composition comprising the sorbent composition and the PFAS contaminants. In one aspect, the spent sorbent is treated through a process called regeneration by removing the PFAS to form a regenerated sorbent. In another aspect, the regenerated sorbent may be re-use to treat a second water source. In some embodiments, these steps of contacting and regeneration are repeated many times.

[0023] In some embodiments, the initial liquid or air source is air, and the PFAS sorbent composition is contained within a pre-filter positioned upstream of an apparatus comprising a granular activated carbon (GAC) assembly and / or a catalytic oxidation assembly, wherein the initial air source passes through the PFAS sorbent composition and then through the granular activated carbon (GAC) assembly and / or the catalytic oxidation assembly before treated air is obtained. In one aspect, the treated air has reduced amount of PFAS. In another aspect, theUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 treated air has reduced amount of other contaminants. In another aspect, the air purification system can be used to treat air before it enters a building, a room, or other enclosed space, such as the International Space Station (ISS) or a spacecraft, or an aircraft. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 is schematic representation of (a) organosilane molecules, (b) their solution-phase deposition and chemical interactions with a hydroxylated oxide surface, and (c) the resulting surface chemistry.

[0025] Figure 2 shows streaming zeta potential from pH 2–10 for silicon sensors in 1 mM potassium chloride. Results are displayed with respect to surface functionality.

[0026] Figure 3 shows Water contact angle measurements for unmodified and modified silicon sensors.

[0027] Figure 4 shows AFM topographical scans of silicon Qsensors with (a) an oxygen plasma -treated surface, (b) an APTES-treated surface, (c) a TEOS (C8)-treated surface, and (d) a TMOS (C18)-treated surface.

[0028] Figure 5 shows average surface formation water measurements for unmodified silicon, APTES-, TEOS (C8)-, and TMOS (C18)-functionalized surfaces.

[0029] Figure 6 shows average mass accumulation profiles extracted from QCM-D adsorption experiments, ran in triplicate, for PFBA, PFBS, PFOA, and PFOS on unmodified silicon, APTES-, TEOS (C8)-, and TMOS (C18)-functionalized surfaces. All solutions are tested at concentrations of 5 ppm (mg / L).

[0030] Figure 7 shows QCM-D mass accumulation profile of PFBA, PFBS, PFOA, PFOS, and their hydrocarbon counterparts on unmodified silicon surfaces. All solutions are tested at concentrations of 5 ppm (mg / L).

[0031] Figure 8 shows schematics of long-chain PFAS (left) and short-chain PFAS (right) interactions with (a) unmodified silicon, where, PFAS are oriented tail up due to hydrogen bonding between PFAS head groups and surface hydroxyl groups; (b) APTES- functionalized silicon, where, PFAS are oriented tail up due to electrostatic attraction between PFASs negatively-charged functional group and APTESs protonated amine groups; (c) TEOS (C8)- or TMOS (C18)- functionalized silicon, where, PFAS approach the hydrophobic surface with a tail down configuration and adhere their fluorocarbon chain parallel to the surface while also having the potential to insert into the hydrophobic SAM.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695

[0032] Figure 9 is numerical representation of D(f) slope for (a) a hydrophilic- unmodified interface, (b) a hydrophilic-APTES functionalized interface, (c) a hydrophobic- TEOS (C8) functionalized interface, and (d) a hydrophobic-TMOS (C18) functionalized interface.

[0033] Figure 10 illustrates QSPR models relating molar volume, Vm (left) or maximum electrostatic potential, EPmax (right) to log k for PFAS adsorption for (a) a hydrophilic- unmodified interface, (b) a hydrophilic-APTES functionalized interface, (c) a hydrophobic- TEOS (C8) functionalized interface, and (d) a hydrophobic-TMOS (C18) functionalized interface.

[0034] Figure 11 shows a) Plots of zeta potential versus pH for modified surfaces in 1 mM potassium chloride. b) Adhesion force measurements taken on modified surfaces in deionized water.

[0035] Figure 12 shows Zeta potential measurements modified surfaces and AFM tip functionalizations taken in 1 mM potassium chloride and 75 mM magnesium chloride at pH 6.8.

[0036] Figure 13 shows adhesion force boxplots of each modified surface-PFAS combination in 75 mM sodium chloride, 75 mM magnesium chloride.

[0037] Figure 14 shows a diagram of current TCCS pipeline (top) and disclosed pipeline (bottom).

[0038] Figure 15 shows fluorocarbon capture by alkyl silanes.

[0039] Figure 16 is a diagram showing the process of using modified cellulose for PFAS removal and the regeneration process.

[0040] Figure 17 shows result of cycling experiments of unmodified and functionalized cellulose sponges compared to granular activated carbon. Five consecutive cycles of adsorption in 5ppm PFAS solutions (PFOA, PFOS, PFXxA, PFHxS, PFBA, or PFBS) and regeneration in deionized water were conducted on three replicates of each sponge and GAC sample. Individual bars represent averaged percent removal / recovery for each cycle. DETAILED DESCRIPTION

[0041] This disclosure relates to a new approach for removal of perfluorinated organics (PFAS), so-called "forever chemicals," from contaminated water sources. These species are expensive to remove with conventional water treatment due to their low concentrations in water and their high hydrophilicity. Some treatment programs are known and typically use aUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 process of one-pass loading and then disposal of granulated activated carbon (GAC) that is employed as the sorbent. Known processes often leave much room for improvement with regard to PFAS removal and / or are cost-prohibitive.

[0042] It has been found that some PFAS, e.g., long chain PFAS, have the potential to aggregate and block GAC pore openings. This can substantially reduce the surface area available for adsorption, and further result in premature breakthroughs. In addition, some PFAS, e.g., short chain PFAS, have displayed lower adsorption affinities to GAC due to their increased mobility, which, in use, causes these PFAS to rapidly break through a sorbent bed. Detrimentally, these phenomena often require the GAC sorbent to be replaced or regenerated before treatment can continue.

[0043] In some embodiments, this disclosure relates to a cost effective, surface modifier composition. The surface modifier composition can be deployed on a sorbent substrate, e.g., on a porous structure, to form a selective sorbent composition.

[0044] In some embodiments, the disclosure relates to a process for purifying a contaminated water stream using the aforementioned sorbent composition. As a result, the amount of PFAS material for ultimate disposal is significantly reduced or eliminated. In some cases, a “single cycle” process is employed. In other cases, multiple cycles may be utilized. The approach demonstrates selective removal of PFAS, e.g., greater than 50%, from an initial contaminated water source. The disclosed sorbent substrate may be an inexpensive material, e.g., cellulose sponge, cellulose microfibers, cellulose nanofibers, cellulose nanocrystals, and / or cellulose powder, and the properties of the material, e.g., porosity, surface treatment, functionalization, beneficially, may be tunable.

[0045] In some embodiments, the fluid containing PFAS to be treated is passed through a column that is packed with the sorbent composition. In operation, adsorption of PFAS onto the sorbent composition occurs during the turbulent transport through the void space of the sponge pores.

[0046] In some embodiments, the process comprises a desorption step. The desorption step may involve mechanical compression, e.g., squeezing, of the sorbent composition, which acts to recover water trapped within the sorbent composition. The sorbent is then flushed with a brine solution that weakens the adhesion of PFAS to the sorbent, releasing them from the cellulose matrix.

[0047] Some advantages of the disclosed process and sorbent include:

[0048] Simplicity of the system, e.g., for temporary point of use application;UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695

[0049] Selective removal, e.g., greater than 90% removal in one cycle / pass, even at typical very dilute concentrations of PFAS (in contaminated water);

[0050] Environmental friendliness;

[0051] Potential for no or little additional fluid generated for backflusing sorbent;

[0052] Low pressure drop operation (relative to membranes or ion exchange);

[0053] Ability to recycle or re-use (versus conventional activated carbon);

[0054] Cost effectiveness – the disclosed sorbent may comprise an inexpensive cellulose sponge;

[0055] Ability to tune sorbent properties (porosity, surface treatment, functionalization).

[0056] Applications for the disclosed process and sorbent include, but are not limited to, water treatment and environmental clean-up applications.

[0057] In some embodiments, the disclosure relates to a process for purifying an initial water source comprising water and an initial amount of perfluorinated organics. The process comprises the steps of contacting the initial water source with a sorbent composition comprising a sorbent substrate and a particular surface modifier to yield purified water (comprising a reduced amount of perfluorinated organics) and a waste composition comprising perfluorinated organics. The perfluorinated organics may comprise perfluorooctanoic acid (PFOA), perfluorooctanesulfonic acid (PFOS), perfluorohexanoic acid (PFHxA), perfluorohexanesulfonic acid (PFHxS), perfluorobutanoic acid (PFBS), and / or perfluorobutanesulfonic acid (PFBS).

[0058] The process has been found to be unexpectedly efficient in removing PFAS. For example, the purified water may comprise less than 75% of the initial amount of perfluorinated organics, e.g., less than 65%, less than 50%, less than 35%, or less than 25%. The contacting may remove greater than 25% of the initial amount of perfluorinated organics, e.g., greater than 35%, greater than 50%, greater than 65%, or greater than 75%.

[0059] The process may remove specific PFAS, e.g., PFOA and / or PFOS. In some cases, the contacting removes greater than 10% of the initial amount of PFOA, e.g., greater than 20%, greater than 25%, greater than 35%, greater than 50%, greater than 65%, or greater than 75%. In some cases, the contacting removes greater than 62% of the initial amount of PFOS, e.g., greater than 65%, greater than 70%, or greater than 75%. The purified water stream may comprise less than 1000 ppm perfluorinated organics, e.g., PFOA and / or PFOS.

[0060] The inventors have found that the particular surface modifier compositions work particularly well to remove PFAS materials from water with unexpected efficiency. In someUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 embodiments, the surface modifier composition imparts a charge to the sorbent substrate. For example, the surface modifier composition may impart a negative charge to the sorbent composition, e.g., to the sorbent substrate. In other cases, the surface modifier composition may impart a positive charge to the sorbent composition, e.g., to the sorbent substrate.

[0061] In some embodiments, the surface modifier composition comprises a silane, e.g., an alkyl silane or a linear alkyl silane. For example the surface modifier may comprise trimethoxy(octadecyl)silane (TMOS), triethoxy(octyl)silane (TEOS), or (3-aminopropyl) triethoxysilane (APTES), or a combination thereof. In some embodiments, the alkyl silane comprises a carbon chain having from 6 to 20 carbons, e.g., from 8 to 18 carbons.

[0062] The surface modifier composition of the sorbent composition may, in some cases, comprise charged silanes and uncharged silanes. In some embodiments, the ratio of charged silanes to uncharged silanes may range from 0.1:1 to 10 : 1, e.g., from 0.5:1 to 7:1, from 1:1 to 5:1, or from 2:1 to 4:1. In some cases, the ratio of surface modifier, e.g., APTES, to silane molecules may range from 5:1 to 0.2:1, e.g., from 3:1 to 0.33:1, or from 2:1 to 0.5:1. Upper and lower limits may be derived from these ranges. Such ranges and limits are not legally limiting, and other ranges and limits are contemplated.

[0063] In some embodiments, the ratio may be employed to “tune” the sorbent to the particular water stream, e.g., to tune the adsorption of specific PFAS molecules. As one example, an 8 carbon silane molecule may be surprisingly effective in capturing an 8 carbon PFAS molecule, but not a 4 carbon PFAS.

[0064] In some embodiments, the surface modifier is an alkaline solution. Advantageously, the surface modifier may increase hydroxy group content on the sorbent substrate.

[0065] In some embodiments, the surface modifier comprises a metal oxide, optionally an aluminum oxide.

[0066] In some embodiments, the sorbent composition may contribute to or provide for hydrophobic-hydrophobic interactions and / or electrostatic interactions. It is postulated that the surface modification advantageously reinforces electrostatic interactions between modifiers and contaminants. This has been found to increase tunability.

[0067] The sorbent composition comprises the sorbent substrate and the sorbent modifier composition. The method of application of the sorbent modifier composition to the sorbent substrate may vary widely. In some cases, the sorbent modifier composition may be applied by employing one or more of the following steps.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695

[0068] The surface modification of the sorbent substrate may comprise the step of regeneration, or mercerization, of the sorbent substrate, e.g., cellulose material, to remove any bound impurities therefrom. In some cases, this regeneration include removal of bound materials or chemical compounds / groups / substituents to the functionalities, e.g., hydroxyl groups, on the sorbent substrate.

[0069] It is postulated that these hydroxyl groups play a significant role in covalently binding the silane molecules to the sorbent substrate. In some cases, the regenerated cellulose may rinsed, e.g., with deionized water and dried.

[0070] In some embodiments, the cellulose (optionally dried or significantly dried) may be incubated in a silane solution, e.g,. a 1% solution, for a prescribed time period, e.g., 24 hours. The solvent may vary widely and may comprise ethanol [APTES] or hexane [TEOS,TMOS]. After incubation, the treated sorbent substrates may be are sonicated in fresh solvent (containing no silane) to remove any non-covalently bound silane molecules from the sponge. The sponges are then dried and ready for PFAS adsorption.

[0071] The sorbent substrate may vary widely. In some embodiments, the sorbent substrate is cellulose-based, optionally cellulose sponge, cellulose microfibers, cellulose nanofibers, cellulose nanocrystals, and / or cellulose powder.

[0072] In some embodiments, the process may further comprise the step of modifying the sorbent based on the characteristics of the water stream. This step provides the ability to tune the surface treatment to a particular water stream. Stated another way, the characteristics of the sorbent may be tuned based on the characteristics of the water stream and or the perfluorinated organics contaminants.

[0073] In some embodiments, the contacting is conducted at a low pressure drop.

[0074] In some embodiments, the process generates a small amount of additional fluid, based on the weight of the water stream.

[0075] The surface modification has been found to impart advantageous characteristics to the sorbent composition.

[0076] In some cases, the sorbent composition, optionally the surface modifier, demonstrates a high contact angle. For example, the contact angle may be greater than 60 degrees, e.g., greater than 90 degrees, greater than 100 degrees, greater than 104 degrees, greater than 105 degrees, or greater than 109 degrees.

[0077] In some cases, the sorbent composition demonstrates Zeta potential less than 18 mV at a pH of 6, e.g., less than 16mV, less than 15 mV, less than 12 mV, or less than 10 mV.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695

[0078] In some cases, the sorbent composition demonstrates a high adhesion force between the sorbent composition and the perfluorinated organics.

[0079] The process may employ a single stage, in some cases. In other cases the process may employ multiple stages, including recycle and / or regeneration steps.

[0080] In some embodiments, the contacting step may yield a spent sorbent comprising the sorbent and perfluorinated organics contaminants. In some cases, the process may further comprise the step of regenerating the spent sorbent to form a regenerated sorbent and / or contacting a second water source with the regenerated sorbent to yield second purified water. In some embodiments, the regeneration comprises the step of desorbing perfluorinated organics contaminants from the spent sorbent.

[0081] The process may employ one purification cycle or may employ multiple purification cycles. A purification cycle may comprise the contacting and then regenerating the spent sorbent. In other cases, a purification cycle may comprise supplying a sorbent (either initial or regenerated) and contacting same with a water source.

[0082] In some cases, the disclosure relates to the aforementioned sorbent composition, which may be employed to remove perfluorinated organics from the water stream. The sorbent may be a treated sorbent comprising a cellulose based sorbent, optionally having hydroxy functionality.

[0083] Various embodiments of the instant disclosure may be further illustrated by the following items:

[0084] Item 1: A process for purifying an initial liquid or air source comprising liquid or air and an initial amount of one or more of per- and polyfluorinated organics (PFAS), the process comprising contacting the initial liquid or air source with a PFAS sorbent composition comprising a sorbent substrate and a surface modifier composition, and obtaining a purified liquid or air product comprising a reduced amount of PFAS.

[0085] Item 2: The process of Item 1, wherein the initial liquid or air source is an initial water source, and the purified liquid or air product is purified water.

[0086] Item 3: The process of any of Items 1-2, wherein the surface modifier composition comprises an alkyl silane, optionally a linear alkyl silane, optionally trimethoxy(octadecyl)silane (TMOS), triethoxy(octyl)silane (TEOS), or (3-aminopropyl) triethoxysilane (APTES), or a combination thereof.

[0087] Item 4: The process of any of Items 2-3, wherein the alkyl silane comprises a carbon chain having from 6 to 20 carbons.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695

[0088] Item 5: The process of any of Items 1-4, wherein the sorbent substrate comprises a cellulose-based substrate, optionally a cellulose sponge.

[0089] Item 6: The process of any of Items 1-5, wherein the surface modifier composition comprises charged silanes and uncharged silanes and the ratio (w / w) of charged silanes to uncharged silanes ranges from 0.1:1 to 0.1:10.

[0090] Item 7: The process of any of Items 1-6, wherein the surface modifier composition is an alkaline solution.

[0091] Item 8: The process of any of Items 1-7, wherein the surface modifier composition increases hydroxyl group content on the sorbent substrate.

[0092] Item 9: The process of any of Items 1-8, wherein the surface modifier composition comprises a metal oxide, optionally an aluminum oxide.

[0093] Item 10: The process of any of Items 1-9, wherein the surface modifier composition comprises a divalent cation, optionally the divalent cation is magnesium or calcium, and optionally, the concentration of the divalent cation is between 1 mM and 200 mM, or between 50 mM and 100 mM.

[0094] Item 11: The process of any of Items 1-10, wherein the surface modifier composition imparts a charge to the sorbent substrate.

[0095] Item 12. The process of any of Items 1-11, wherein the sorbent composition demonstrates a contact angle greater than 60 degrees, or optionally greater than 90 degrees.

[0096] Item 13: The process of any of Items 1-12, wherein the sorbent composition demonstrates Zeta potential less than 18 mV at a pH of 6.

[0097] Item 14: The process of any of Items 1-13, wherein the PFAS comprise perfluorooctanoic acid (PFOA), perfluorooctanesulfonic acid (PFOS), perfluorohexanoic acid (PFHxA), perfluorohexanesulfonic acid (PFHxS), perfluorobutanoic acid (PFBS), and / or perfluorobutanesulfonic acid (PFBS).

[0098] Item 15: The process of any of Items 1-14, further comprising the step of modifying the sorbent based on the characteristics of the water stream.

[0099] Item 16: The process of any of Items 1-15, wherein the purified water comprises less than 75% of the initial amount of PFAS.

[0100] Item 17: The process of any of Items 1-16, wherein the contacting of the initial water source with the sorbent composition removes at least 25% of the initial amount of PFOA.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695

[0101] Item 18: The process of any of Items 1-17, wherein the contacting of the initial water source with the sorbent composition removes at least 62% of the initial amount of PFOS.

[0102] Item 19: The process of any of Items 1-18, wherein the purified water stream comprises less than 1000 ppm PFAS.

[0103] Item 20: The process of any of Items 1-19, wherein the contacting is conducted at a low pressure drop.

[0104] Item 21: The process of any of Items 1-20, wherein the process generates a small amount of additional fluid, based on the weight of the water stream.

[0105] Item 22: The process of any of Items 1-21, wherein the contacting step yields a spent sorbent composition comprising the sorbent composition and the PFAS contaminants.

[0106] Item 23: The process of any of Items 1-22, further comprising a step of regenerating the spent sorbent to form a regenerated sorbent.

[0107] Item 24: The process of any of Items 1-23, wherein the regeneration comprises the step of removing PFAS contaminants from the spent sorbent.

[0108] Item 25: The process of any of Items 1-24, further comprising the step of contacting a second water source with the regenerated sorbent to yield second purified water

[0109] Item 26: The process of any of Items 1-25, wherein the steps of contacting, regenerating, and contacting are repeated for at least 5 cycles.

[0110] Item 27: The process of any of Items 1-26, wherein the initial liquid or air source is an initial air source, and the PFAS sorbent composition is positioned as a pre-filter upstream of an apparatus comprising a granular activated carbon (GAC) assembly and a catalytic oxidation assembly, wherein the initial air source passes through the PFAS sorbent composition and then through the granular activated carbon (GAC) assembly and the catalytic oxidation assembly.

[0111] Item 28: A surface-modified sorbent composition for removing one or more of PFAS from an air stream comprising air and one or more of PFAS, the surface-modified sorbent composition comprising a sorbent substrate and a surface modifier composition, wherein the surface modifier composition comprises an alkyl silane, optionally the alkyl silane comprising one or more of trimethoxy(octadecyl)silane (TMOS), triethoxy(octyl)silane (TEOS), or (3-aminopropyl) triethoxysilane (APTES), or a combination thereof.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695

[0112] Item 29: The surface-modified sorbent composition of Item 28, wherein the sorbent composition is contained in a pre-filter assembly positioned upstream of an apparatus comprising a granular activated carbon (GAC) assembly and a catalytic oxidation assembly

[0113] Item 30: A surface-modified sorbent composition for removing one or more of PFAS from a water stream comprising water and one or more of PFAS, the surface-modified sorbent composition comprising a sorbent substrate and a surface modifier composition, wherein the surface modifier composition comprises an alkyl silane, optionally the alkyl silane comprising one or more of trimethoxy(octadecyl)silane (TMOS), triethoxy(octyl)silane (TEOS), or (3-aminopropyl) triethoxysilane (APTES), or a combination thereof.

[0114] Item 31: The surface-modified sorbent composition of any of Items 28-30, having a capacity for PFOA removal higher than 1.2 mg PFOA / g sorbent and a PFOA recovery in regeneration process of higher than 10%, or 20%. Examples

[0115] The disclosure will now be illustrated with working examples, which are intended to illustrate the working of disclosure and not intended to restrictively any limitations on the scope of the present disclosure. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood to one of ordinary skill in the art to which this disclosure belongs. Although methods and materials similar or equivalent to those described herein can be used in the practice of the disclosed methods and compositions, the exemplary methods, devices and materials are described herein.

[0116] Example 1 Adsorption behavior of per- and poly-fluoroalkyl substances (PFAS) on model surfaces

[0117] Experiments in this Example were carried out to demonstrate and elucidate the adsorption of four PFAS (i.e., PFOA, PFOS, PFBA, and PFBS) through an integrated approach using QCM-D and organosilane functionalized surfaces. These four PFAS were selected to probe the effects of (1) alkyl chain length (C4 vs. C8) and (2) the head groups (carboxylic vs. sulfonic acid) on hydrophobic and hydrophilic interfaces.

[0118] Specifically, the interfaces were silicon surfaces functionalized with 3- Aminopropyltriethoxysilane (APTES), Triethoxy(octyl)silane (TEOS (C8)), or Trimethoxy(octa- decyl)silane (TMOS (C18)). QCM-D was used to measure the adsorption profile in real-time for different combinations of PFAS and functionalized surfaces andUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 extracted the adsorption rate, performance, and the primary mass-transfer mechanism. Lastly, to characterize the impact of PFAS structure on solid–liquid interfacial adsorption, a single molecular descriptor (molar volume) was employed to develop a correlation between kinetic parameters, molecular size, and interfacial chemistry.

[0119] Surface preparation was initiated by submerging the SiO2 QSensors in a 2 % (w / v) sodium dodecyl sulfate (SDS) solution for 10 min, followed by rinsing with DI water and drying with a clean nitrogen gas stream. The sensors were then exposed to an oxygen plasma and UV- ozone (UVO) cleaner (Optiglow ACE plasma treatment unit, Tempe, AZ, U.S.) at 50 ⁰C for 5 min. UV-ozone plasma treatment is an essential step in the cleaning procedure that activates the material surface chemically via hydroxylation, while additionally removing traces of organic contaminants. The dry-cleaning process has proved to increase the amount of hydroxyl groups on the surface and form a homogeneous layer with reduced surface roughness. Clean sensors were then immediately submerged in 30 mL of a modification solution containing 1 % volume of silane in solvent. Fig.1(a) illustrates the molecular structure of the silanes used for surface functionalization. Solvents were determined based on the polarity of the silane, where, triethoxy(octyl) silane and trimethoxy(octadecyl)silane are non-polar, therefore, they were treated with hexane, and 3- aminopropyltriethoxysilane is polar so it was treated with ethanol.

[0120] Once the sensors were placed in a test tube containing the desired modification solution, the functionalization process was left to proceed on an orbital shaker for 24 h to ensure a homogeneous, thoroughly mixed environment. Fig.1(b) displays the solution-phase deposition process that occurred in each test tube with the oxide surface. This involved a complex reaction with 3 main steps: (1) hydrolysis, (2) condensation, and (3) covalent linkage of the silane to the surface. The solution-phase deposition was completed with an anhydrous solution (e. g., hexane and ethanol), however, previously adsorbed water on the oxide surface, due to rinsing, was important for initiating hydrolysis, where, meth- or ethoxy groups were substituted for hydroxyl groups. Once hydrolysis was initiated, condensation and hydrolysis reactions could co-occur. Condensation reactions formed siloxane bonds (Si-O-Si) and occurred between two hydrolyzed silane molecules or between a surface hydroxyl group and a hydrolyzed silane molecule.

[0121] After 24 h, the sensors were removed from the modification solution and rinsed with ethanol to hydrolyze the alkoxy group. The sensors were then dried with a clean nitrogen gas stream, resubmerged in 30 mL of the appropriate solvent (ethanol or hexane),UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 and placed in an ultrasonic bath (CPX5800 Ultrasonic Bath 9.5 L, Fisherbrand) for 10 min to remove unbound or physisorbed silane molecules. Once removed from the solvent and dried with a clean nitrogen gas stream, the modification procedure was finalized by baking the sensors at an elevated temperature (≥110 ⁰C) for 1–2 h. Baking the sensors further ensured the silanes had cross-linked to the surface and additionally removed any residuals. Fig.1(c) displays the resulting surface structures of the silicon sensors with covalently linked silane molecules.

[0122] Fig.2 shows the streaming zeta potential as a function of pH for unmodified silicon and silane-functionalized silicon. Zeta potential measurements were used to distinguish changes in surface charging behavior or polarity with the addition of hydrophobic or polar moieties. Moreso, this parameter alludes to the mechanisms occurring at the surface when the stationary phase is in contact with a liquid mobile phase. Experimental pH values of concern ranged from 6.5 to 7 due to PFAS solutions using DI water as the solvent. From the results, it was noted that APTES was the only surface treatment to exhibit a positively charged polar surface in the experimental pH range (+8.83 mV). In contrast, unmodified silicon, TEOS (C8)-, and TMOS (C18)-functionalized silicon all portrayed negatively charged non-polar surfaces (–32.52 mV, -17.94 mV, and -9.80 mV, respectively). Furthermore, zeta potential values will play a critical role when determining how surface charge affects PFAS adsorption due to the parameter establishing a basis for electrostatic attraction or repulsion between surface active sites and the liquid medium.

[0123] Water contact angle (WCA) measurements were taken to assess the hydrophobicity of the silicon surfaces before and after the silanization treatment. An unmodified plasma-treated silicon surface demonstrated an average WCA≤5 degrees, declaring it hydrophilic. Silicon surfaces functionalized with APTES, TEOS (C8), and TMOS (C18) exhibited average WCAs of 74.7 degree, 100.3 degree , and 110.5 degree, respectively (Fig.3). All four measurements corresponding to the average are provided in SI (Table S5). Increased hydrophobicity is a result of the linear alkyl chain present in the tail of the silanes, thus, as the carbon content of the molecule increased, so did its hydrophobicity. In addition, surface free energy (SFE) has a decisive influence on surface wettability, where, SFE is dependent on the polarity of the terminal group present at the inter- face (e.g., –OH, –CH3, or –NH2). When a polar terminal group (e.g., –OH and –NH2) is present at the solid–liquid interface, strong bonds are formed, where, more energy will be required to break them and vice versa for the non-polar terminal group (e.g., CH3). Hence hydrophilic surfaces have highUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 SFE and hydrophobic surfaces have low SFE. Therefore, out of the three silanes, TMOS (C18) acts as the principal hydrophobic modifier.

[0124] AFM imaging was employed to further characterize the morphology and surface properties of silicon substrates. Fig.4 displays the topo- graphical images obtained for silicon Qsensors with (a) an unmodified plasma-treated surface, (b) an APTES-treated surface, (c) a TEOS (C8)- treated surface, and (d) a TMOS (C18)-treated surface. AFM imaging also provided surface roughness measurements to impart the surface’s texture, quality, and functional performance. Hydrophobic interfaces reported higher Rq averages in contrast to those possessing hydrophilic interfaces. Where, TMOS (C18)-treated silica had the highest surface roughness with a Rq of 0.564 ± 0.041 nm, followed by TEOS (C8), un- modified, and APTES, with values of 0.481 ± 0.056, 0.359 ± 0.025, and 0.260 ± 0.009 nm, respectively (Table S6). Surface treatments with TEOS (C8) and TMOS (C18) likely result in rough surfaces due to the increased surface area they create. In other words, the addition of their long molecular geometries (C8 and C18) causes a broader distribution of surface energy sites. Whereas, a smooth surface would have a narrow distribution of site energies. This relationship can be seen in Fig.4, where the topographical images of hydrophobic surfaces, (c) and (d), show an assortment of island-like or globular structures with clustering of silica particles. Fig.4(a) is characterized by a moderate surface roughness due to the surface displaying globular features to an extent, however, the hierarchy structures are much more uniform than those of the hydrophobic surfaces. Fig.4(b) was characterized as the smoothest surface due to the absence of aggregate structures; confirming also a uniform charge distribution at the APTES interface.

[0125] Sorption onto the silica surfaces was carried out using QCM-D. Mass adsorption measurements were obtained from QCM-D experiments by monitoring the changes in resonance frequency (Δf) from the oscillating quartz surface. An additional parameter, dissipation change (ΔD) was measured to quantify the adsorbed layers’ viscoelastic nature. Sauer-brey’s equation (Eq. S1) was applied to translate the Δf to nanoscale mass changes, where a greater Δf indicated a greater presence of molecules adsorbed to the surface..

[0126] Before conducting PFAS adsorption experiments, water formation measurements were taken with QCM-D to further interpret the effects of surface chemistry on solid–liquid interfacial interactions. These measurements were not only a baseline measurement to quantify the amount of control fluid on each solid interface at reference zero (Fig.5); they also gauged how much water the test solutions would be interacting with, to adsorb to theUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 interface. In addition, these results confirmed that the terminal group’s properties can dictate water formation at the sol- id–liquid interface. From Fig.5, it was concluded that hydrophilic surfaces – i.e., unmodified and APTES – exhibited strong water-surface interactions which contributed to a growing water layer. The polar functional groups; amine, and hydroxyl, provided locations for the polar water molecules to hydrogen bond. On the other hand, hydrophobic surfaces – i.e., TEOS (C8) and TMOS (C18) – depicted weak water- surface interactions because of the higher energetic cost for cavity formation in water, resulting in poor surface wettability.

[0127] Silicon surfaces functionalized with hydrophobic silanes were not expected to undergo significant dissociation within the environmentally relevant pH range. However, the chemical behavior of silanol and amine groups on the quartz surface varied due to the properties of their surface layer being dependent on the bulk solution pH. An APTES surface layer is readily protonated when interacting with an adjacent water layer at pH~7, due to amino groups undergoing ionization (NH2 ↔NH3+) in the pH range of 6–9 (pKa ~ 8.5). A silicon surface bares siloxane (Si-O-Si), silanol (Si-OH), and silanolate (Si-O-) groups when interacting with an adjacent water layer at pH~7, where, the latter group is a result of the partial deprotonation of silanols (isoelectric point, pH~2–3). This leaves the surface to bear a net negative charge, however, a population of silanols remains protonated due to the various types of species existing at the silica-water interface (i.e., hydrogen- bonded (pKa ~ 4.5), vicinal or geminal (pKa ~ 8.5), and isolated (pKa ~ 10.5)) [42,43]. Consequently, water molecules interacting with the interface will either form strong hydrogen bonds with the acidic silanols or weak bonds with silanols acting as H-bond donors. Thus rending two populations of sorbed water to exist at the interface.

[0128] After achieving baseline measurements for adsorbed water molecules, we monitored the adsorption of four PFAS onto silicon dioxide (SiO2) Qsensors as a function of time. Unmodified silicon sensors, possessing a hydroxylated surface, were first utilized for adsorption experiments, followed by silicon sensors functionalized with either a non- polar: hydrophobic interface (i.e., TEOS (C8) and TMOS (C18)) or a polar: hydrophilic interface (i.e. APTES). Interchanging feed solutions and surface properties allowed for different adsorbate-adsorbent interactions to be studied. The adsorption data illustrated in Fig.6 displays the average mass accumulation profiles of each PFAS solution adsorbing onto an (a) unmodified silicon interface, (b) APTES-functionalized interface, (c) TEOS (C8)- functionalized interface, and (d) TMOS (C18)- functionalized interface.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695

[0129] Focusing on the QCM-D data in Fig.6(a), there is a clear distinction between the perfluoroalkyl carboxylic acids (PFCAs) and perfluoroalkyl sulfonic acids (PFSAs) regarding their mass accumulation onto unmodified silicon surfaces. The mass profiles of PFOS and PFBS both showed minimal adsorption to the hydroxylated surface, while PFOA and PFBA both caused mass to desorb from the surface. As stated in section 3.2.1, when PFAS solutions are introduced into the system, a population of water already exists on the surface. Therefore, the negative adsorption values in Fig.6(a) could depict surface dehydration rather than dissolution of surface coatings due to two populations of water existing at the silica-water interface. Thus, any subsequent water layers adsorbed to the surface via weak H-bonds pose the potential to diffuse back into the bulk liquid. Therefore, it was hypothesized that PFBA and PFOA penetrate the adsorbed water layer, where the molecules remove and rearrange surface-bound water to adsorb to the surface. Due to an excess amount of water being present on the unmodified interface (11.7 ± 1.55 nmol / cm2), it was hypothesized that only weakly bound, subsequent water layers were desorbed from the surface. Leaving a smaller, tightly bound polar network at the interface. It was then presumed that PFBA and PFOA molecules were interacting with the strongly bound water layer via hydrogen bonding. This polar network was then further broken down into smaller fragments upon PFCAs dissolution via H- bonds. Negative adsorption values recorded were likely due to PFCAs displacing water from the surface at a greater rate than they adsorb, rather than from an absence of PFOA or PFBA being retained on the surface.

[0130] While the former explanation elucidates negative mass values, it does not explain why PFCAs allowed for surface dehydration but PFSAs did not. The main difference between the two resides in their head group, where, carboxylate anions are more solvated than their sulfonate counterpart. The difference in solvation between the two is a result of carboxylates’ greater hydrophilic character, where, its oxygen atoms possess partial charges of - 0.80 a.c.u., compared to sulfonates which is - 0.63 a.c.u. Any interactions between the anion and water molecules are governed by oxygen atoms in the functional group, thus, PFCAs would have increased, and stronger water-ion interactions than PFSAs. Additionally, water molecules in the vicinity of the solid interface can exhibit a water-mediated repulsion towards opposing constituents in the system. Moreso, PFSAs stronger hydrophobic character can affect the structure and dynamics of water in the confined region, where a water-induced repulsive force originates. This would preserve their polar networking and cause PFSAs to accumulate as a subsequent layer.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695

[0131] The different adsorption values displayed by PFOA and PFBA, correlate to the magnitude of interaction between water molecules and fluorinated chains. The complete substitution of carbon–fluorine counterparts induces a strong dipole in PFAS, where the molecules can participate in dipolar interactions and enable polar networking. Compared to PFOA, PFBA would have decreased dipolar interactions due to its shorter fluorinated chain presence. Solvents containing PFBA would then hinder the permeance of water layers and result in increased dewetting. Dewetting could also be associated with cavity expulsion potential due to an absence of concomitant solute - water attractive interactions.

[0132] QCM-D adsorption measurements were also conducted for four hydrocarbon surfactants composed of similar functional groups and carbon chain lengths (i.e., octanoic acid (C8 acid), butanoic acid (C4 acid), octane sulfonic acid sodium salt (C8 salt), butane sulfonic acid sodium salt (C4 salt)). The QCM-D data obtained was used to distinguish adsorption trends associated with molecular structure and hydrodynamic interaction. Fig.7 displays the comparative adsorption of PFOA, PFOS, PFBA, and PFBS with their alkyl counterparts on unmodified silicon quartz surfaces.

[0133] From Fig.7, it was noted that C4 and C8 acids illustrated inverse mass profiles from PFBA and PFOA. The low polarizability of PFOA and PFBA likely rendered both molecules poor hydrogen bond acceptors, compared to their alkyl counterparts, hindering their ionization in aqueous media. In addition, fluorocarbon tails induce steric hindrance, where, its partial charge and large molecular geometry limit bond formation and increase selectivity. The greater partial charge on fluorine atoms also results in greater electrostatic repulsion, leaving fluorine atoms rigid and unable to be densely packed together. Therefore, it was hypothesized that the low polarizability and molecular arrangement of fluorocarbons was enough to offset PFBA and PFOAs greater energetic affinity for the adlayer.

[0134] The results from Fig.7 also demonstrated positive mass accumulation profiles for all sulfonate solutions. Their greater affinity to hydroxylated surfaces stemmed from stronger H-bonding and ion–dipole interaction. Sulfonates higher electronegativity, and additional oxygen atom, helped stabilize its negative charge and provide the molecule with a greater negative inductive effect for ionization and H-bonding. In the case that all surface sites are not occupied by adjacent water molecules. Pearson’s idea of hard- and soft acids and bases would apply. Where, carboxylates are known to be a soft base, while sulfonates are hard; and a hard base gets adsorbed onto oxide surfaces easier due to its hard acid identity. SulfonicUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 acids (hard base) would then display a greater affinity for silicon oxide surfaces (hard acid), thus, compared to carboxylic acids.

[0135] Silane modified surfaces-APTES was highly suitable for surface functionalization due to its ability to introduce nucleophilic functional groups to the surface. Its electron-rich moieties were readily protonated when dissolved in water at a pH of 7, allowing for amine groups (pKa ~ 8.5) to bear a positive charge (NH3+) and serve as surface active sites for anionic PFAS adsorption, section 3.2.1. Therefore, for APTES-functionalized surfaces sorption was facilitated by electrostatic attraction due to the aqueous layer adjacent to the protonated surface consisting of anionic PFAS solutes. PFAS was then immobilized and retained on the surface through the attachment of their negatively charged functional group. Adsorption interactions, concerning electrostatic attraction, were dependent on, the strength of the electrostatic force exerted by the charged surface (e.g., zeta potential) and the number of ions present in the medium around the surface (e.g., concentration). Another potential interaction could be dipole attraction, arising from weak ion dipoles between silica material with amine surface groups and anionic PFAS. If an APTES- functionalized surface was not fully protonated, its amine groups likely rendered the two forms, –NH2 and – NH3+, where PFAS adsorption could occur by either electrostatic or dipole attraction.

[0136] Sorption of PFBA, PFBS, PFOA, and PFOS onto an APTES functionalized surface is shown in Fig.6(b), where each PFAS reached an average equilibrium value of 0.68 ± 0.12 nmol / cm2, 0.17 ± 0.01 nmol / cm2, 0.89 ± 0.10 nmol / cm2, and 0.72 ± 0.07 nmol / cm2, respectively. The adsorbed masses of PFOA and PFBA, were greater than that of their sulfonate counterpart, PFOS and PFBS; suggesting higher PFCA adsorption at a positively charged interface. This difference in mass adsorbed is attributed to decreased electrostatic potential, which can be explained by considering the resonance forms of both head groups. Resonance forms contribute to the molecule’s overall stability by distributing its electron density evenly throughout the molecule. Thus, greater electron delocalization occurs in molecules with a higher number of resonance forms. PFSAs one sulfur and three oxygens have three resonance forms to delocalize negative charge from the head group. In contrast, the PFCA head group comprises one carbon and two oxygens, where only two resonance forms exist for delocalizing the negative charge. Consequently, PFSAs will have decreased electrostatic potential due to less absolute charge in the head group.

[0137] The results observed in Fig.5(b) also indicate electrostatic interactions were minimized with increasing fluorocarbon chain length. This was implied due to the largerUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 mass profiles exhibited by PFCAs at the charged interface, and PFBAs considerable presence when compared to PFOA and PFOS. The former result was due to electron-withdrawing effects brought on by the fluorinated tail. With an increase in chain length comes a broader distribution of the head group’s negative charge along its molecular axis. Therefore, the net charge will essentially get diluted among the molecule as chain length increases, where less ab- solute charge will be available for electrostatic interactions. These findings are consistent with previously reported minimum and maximum electrostatic potentials of PFAS (EPmin and EPmax, respectively) calculated in water. Where EPmin and EPmax decreased as the number of conjugate base resonance forms in the head group or electron-withdrawing fluorines in the tail increased.

[0138] TEOS (C8) and TMOS (C18) surface functionalization also provided heightened adsorbent properties for the silicon by implementing a non- polar interface with increased hydrophobicity. PFASs distinguished fluorocarbon carbon chain complements the molecule with pronounced hydrophobicity; yielding its adsorption to non-polar: hydrophobic surfaces (e.g., TEOS (C8) and TMOS (C18)) due to hydrophobic attraction. Hydrophobic interaction originates from the entropic tendency of non-polar hydrophobes aggregating in an aqueous solution to repel water molecules. With this entropic driving force present, PFAS have been preferentially attracted to the hydrophobic surface rather than staying in the bulk liquid. As PFAS adsorb to the hydrophobic surface with their molecular axis horizontal to the surface they minimize the contact between water molecules and the C–F chains. However, it should be noted; hydrophobic attraction between long-chain PFAS and a non-polar surface will start further than that of a short-chain. This will inhibit the diffusion process of PFBA and PFBS, compared to PFOA and PFOS, resulting in less adsorbed mass.

[0139] Cohesive properties, demonstrated by both water and PFAS, also act as a subsequent force for PFAS adsorption. As PFAS approaches the surface, the fluorocarbon chain attempts to exclude any water molecules in its vicinity; disrupting repulsive VDW forces generated by water molecules at both sides of the fluorocarbon chain. Consequently, water molecules exert a net repulsive VDW force on the chain of the molecule, further driving PFAS to the surface. Therefore, PFAS possessing greater hydrophobic character and larger molecular geometry are conjected to obtain higher adsorption capacities when a hydrophobic interface is implemented.

[0140] QCM-D adsorption data for TEOS (C8)- and TMOS (C18)-functionalized surfaces, Figs.6(c) and (d), show correspondence with the former mechanisms discussed. ForUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 example, QCM-D results for all four PFAS displayed a positive correlation to adsorption with respect to hydrophobicity and fluorocarbon chain length. The former trend was supported by an increase in adsorbed mass when surface hydrophobicity increased (TEOS (C8) < TMOS (C18)), where, PFOS adsorption increased from 0.80 ± 0.16 nmol / cm2 to 0.89 ± 0.08 nmol / cm2 for TEOS (C8) and TMOS (C18), respectively. The trend was further exhibited by PFOA, PFBS, and PFBA, where, their adsorption increased from 0.33 ± 0.04 nmol / cm2 to 0.99 ± 0.06 nmol / cm2, 0.14 ± 0.05 nmol / cm2 to 0.77 ± 0.07 nmol / cm2, and 0.38 ± 0.05 nmol / cm2 to 0.79 ± 0.14 nmol / cm2, respectively, with the latter value representing TMOS (C18). The smaller mass accumulation profiles observed for TEOS (C8)- functionalized surfaces could have also been a result of charge repulsion from the anionic PFAS, where, TEOS (C8) surfaces developed a greater negative charge at the solid–liquid interface, compared to TMOS (C18), section 3.1. Also, the adsorbed mass values coincide with the proposed effects from fluorocarbon chain length, where PFAS adsorption to the hydrophobic surface increased as fluorocarbon content did (PFBA, PFBS<PFOA, PFOS).

[0141] Surface complexation of adsorbed PFAS- For hydrophilic surfaces (i.e., unmodified and APTES), it was postulated that PFAS would bind to the surface with a tail-up configuration due to PFAS-surface interactions being exclusively driven by its hydrophilic head group. Specifically, the surface interactions depicted in Fig.8, where, oxygen atoms in the functional group (CO– or SO- ) undergo (a) hydrogen bonding with surface-bound water or –OH groups on unmodified silica, or (b) participate in electrostatic interactions with protected NH2 groups on the interface of APTESs molecular layer (ML). Conversely, hydrophobic interfaces (i.e., TEOS (C8) and TMOS (C18)) were presumed to attract PFAS to the surface-oriented tail-down; largely driven by the fluorocarbon chains’ greater affinity to the hydrophobic surface than the polar head group. Thus, the following surface orientations were hypothesized in Fig.8(c), where PFAS maximizes interactions between its fluorocarbon chain and the hydrophobic interface either by adhering flat or parallel to the surface, distributing their C-F chains on the surface with different molecular tilts, or by inserting their fluoro-carbon chain into the hydrophobic ML.

[0142] QCM-D parameters, ΔD and Δf, were compared to further elucidate surface- adsorbed PFAS configurations. With time removed as a variable, the slope of ΔD / Δf, or a D(f) curve, reflects information regarding the viscoelastic properties of a PFAS adlayer. A high, or steep slope corresponds to the formation of a viscoelastic, or loosely bound adlayer, characterized with an inherent softness due to weaker PFAS-surface interactions. Whereas, aUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 lower slope indicates a rigid, or stiff adlayer highly retained to the surface. Additionally, surface complexations were transcribed due to the film’s viscoelasticity being dependent on its adhesion to the surface. Fig.9 provides the slope of D(f) curves obtained for PFAS adlayers on (a) unmodified silicon, (b) an APTES- functionalized interface, (c) a TEOS (C8)- functionalized interface, and (d) a TMOS (C18)- functionalized interface.

[0143] Results from the D(f) curves show good agreement with the former speculations. Where, unmodified silicon showed steep D(f) curves indicating PFASs weak hydrogen bonding interactions at the hydrophilic-hydroxylated surface. For the hydrophilic APTES- modfied surface it was noted the slope of the D(f) curves increased as chain length increased, suggesting weaker surface interactions for longer chained PFAS. Where PFOA and PFOS adlayers may be less retained or less densely packed compared to PFBA and PFBS. This result is consistent with previous evidence regarding surfactant configuration

[0049] , where, longer fluorocarbon chains prevent a higher degree of packing due to electrostatic repulsive F–F interactions. In contrast, hydrophobic surfaces revealed D(f) slopes that increased with respect to decreasing fluorocarbon chain length, implying longer-chained PFAS produced strongly retained, rigid films in comparison to their shorter- chained counterparts. This result can be attributed to lateral surfactant–surfactant interactions, where, adjacent fluorocarbon chains exhibit rigid behavior when in close proximity.

[0144] Structural film trends described in Fig.8 were also related directly to ΔD, where, its value indicates the way the molecule is adsorbing to the surface. When a molecule with a long molecular axis (e.g., PFAS) adheres itself flat or parallel to the surface, there will be less energy dissipated from the system (ΔD<1 x 10-6), Fig.8(c); while the same molecule standing up or “tail-up” on the surface (Fig.8(b)) causes higher dissipation (ΔD>1 x 10-6). Therefore, when PFAS molecules undergo “point adsorption” with a positively charged APTES surface

[0058] , they are likely to produce films that give high dissipation. Adlayers that comprise long-chain PFAS on an APTES-functionalized surface will cause higher dissipation than short-chain PFAS due to their greater molecular cross-sectional area.

[0145] Adlayers comprising PFOA or PFOS caused low dissipation when formed on TEOS (C8)- or TMOS (C18)-functionalized surfaces due to the high degree of attraction between non-polar surfaces and non-polar moieties of PFAS via hydrophobic interactions. Cohesive forces exhibited by PFAS tail groups on the surface, and entropy-driven adsorption maximizing the interaction between fluorocarbon chains and the hydrophobic interface, were hypothesized to be the main contributors to the film’s degree of rigidity and low dissipation.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 Since PFBA and PFBS have smaller hydrophobic surface areas than those of PFOA and PFOS, it was presumed their higher ΔD was due to less hydrophobic or entropy- driven interactions keeping them bound to the hydrophobic interface.

[0146] Previous hypotheses regarding hydrodynamic interactions on un- modified silica could also be confirmed with ΔD. For PFSAs, it was presumed the molecules adsorbed as a subsequent layer adjacent to a rigid water layer. The low ΔD recorded for both PFOS and PFBS confirmed this formation was due to repulsive forces exerted by the water layer to preserve its H-bonded polar network. For PFCAs, low ΔD was recorded as well. However, taking into account the mass loss from weakly bound water layers, this result is likely due to the strongly bound water layer on the hydrophilic surface. The lower ΔD depicted from the PFOA layer, compared to PFBAs, also confirmed increased dipolar interactions with the water layer.

[0147] Adsorption at the solid–liquid interface- Adsorption mechanisms for PFAS anions at the solid–liquid interface are examined here in terms of surface hydration, system free energy, and electrostatic potential. When a polar or hydrophilic surface is emersed in solvent, an energy barrier appears due to surface hydration, creating a hydrogen-bonded network with water molecules reoriented in a cohesive structure at the interface. PFAS molecules in the diffusive layer face no mean force and diffuse freely in this region with no major interferences. However, as the molecules attempt to approach the surface layer, a free energy barrier occurs and requires an extra energy cost from the PFAS molecules to disrupt the water structure. Once they overcome this barrier, the molecules can reach a stable adsorption position. The replacement of hydroxyl groups with amine groups significantly enhances the electrokinetic charge density of the surface, Fig.2, where an APTES surface can electrostatically attract the mobile anionic PFAS in the diffuse layer to the surface layer, allowing PFAS molecules to overcome this barrier at a greater extent than they would with an un- modified surface. Seeing that the degree of adsorption for APTES- functionalized surfaces is determined by electrostatic potential and surface charge density, it is postulated that short-range electrostatic interactions will act as the main driving force.

[0148] Conversely, when non-polar: hydrophobic surfaces are emersed in solvent, no energy barrier exists. Water layers adjacent to planar or flat hydrophobic surfaces are characterized with lower density than that of hydrophilic surfaces, due to the dynamic motion of water molecules. This prevents water molecules near the surface from forming any orientational arrangement and facilitates a stable diffusive process for TEOS (C8) and TMOSUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 (C18)-functionalized surfaces. Moreover, the lack of an energy barrier further conforms to entropy-driven hydrophobic interactions as the predominant driving force for TEOS (C8)- and TMOS (C18)-functionalized surfaces.

[0149] The fitting of PFO and PSO models to QCM-D adsorption data enabled the determination of PFAS kinetic parameters. Their application sought to quantitatively interpret the underlying mechanisms behind PFAS adsorption seeing that its behavior varies depending on its molecular structure or the interfacial properties of the adsorbent present. Also, model fittings helped elucidate rate-limiting mechanisms for different stages of PFAS adsorption.

[0150] The PFO model provided insight for the initial or first stage (0–2 h) of PFAS adsorption, whereas the PSO model accurately represented the remaining or second stage (t > 2h) of adsorption. All mass profiles obtained from QCM-D adsorption experiments exhibited rapid PFAS uptake in the initial stage of adsorption, with about 25–30 % of total adsorbed mass being accounted for in the first 2 h. This suggest the initial adsorption rate is concentration dependent; following first-order kinetics limited by the mass transfer resistance of PFAS in the liquid phase. The PSO model was able to represent adsorption data following the 2-hour mark due to the model’s consideration that adsorption rate will mainly be affected by the adsorption mechanism. Thus, this second stage of adsorption was likely due to other kinetic mechanisms or surface interactions dominating adsorption rate, e.g., van der Waals forces, or hydrophobic and electrostatic attraction.

[0151] To further assess PFAS adsorption behavior, and conclude on sorbent efficiency, we employed a QSPR analysis to develop a correlation be- tween kinetic parameters, molecular compositions, and interfacial physicochemical properties. PFAS features, Vm and EPmax, were exercised as singular molecular or electronic descriptors, respectively, to characterize the impact of PFAS structure on solid–liquid interfacial adsorption. Fig.10 illustrates Vm (left) and EPmax (right) QSPR models for PFAS adsorption onto (a) a hydrophilic-unmodified interface, (b) a hydrophilic-APTES functionalized interface, (c) a hydrophobic-TEOS (C8) functionalized interface, and (d) a hydrophobic-TMOS (C18) functionalized interface. The X-axis of the QSPR models are labeled according to the employed descriptor, representing the range of PFAS molar volumes (cm3 / mol), or maximum electrostatic potentials (kJ / mol) implemented in the analysis. The y-axis represents log k values (cm2•nmol-1•h-1) for the solid–liquid interfacial adsorption of PFBA, PFBS,UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 PFOA, and PFOS, obtained from PSO model fittings. Thus, each data point corresponds to PFAS adsorption rate, with respect to their Vm or EPmax.

[0152] PSO rate constants were utilized in the QSPR analysis due to their superior accuracy in describing PFAS behavior across the entire range of adsorption, with respect to adsorption mechanisms. Compensation relations were then compared per QSPR models, with differences in log k concerning sorption mechanisms. For unmodified silicon surfaces, it was concluded that PFAS adsorb to the hydroxyl-terminated surface via H- bonding. From Fig. 10(a), no distinct relationship could be identified between PFAS features and its sorption mechanism. The lack of any parallel between the two makes it difficult to predict the driving force or rate-limiting mechanism that is predominant in PFAS sorption to a hydroxylated- hydrophilic surface.

[0153] The results in Fig.10(b) demonstrate a higher model performance for the EPmax model in contrast to the Vm model. Moreover, the results suggest that the EPmax model can provide robust predictions regarding the influence of electronic effects on PFAS adsorption to an APTES- functionalized surface. The increased predictability of the EPmax model, in comparison to the Vm model, is likely due to the absolute charge of the head group being a dominant factor on reaction rate. Where, the EPmax corresponds to the head group’s net charge distribution, or the lowest absolute charged part of the compound (the end of the tail), and the effect of increasing chain length causes a decrease in the magnitude of EPmax. Thus, this relationship gives the descriptor, EPmax the ability to characterize the reaction rate concerning the nature of the head group charge. Conversely, the Vm descriptor depicts poor performance due to its inability to predict the absolute charge available for electrostatic interactions.

[0154] Figs.10(c) and (d) represent EPmax and Vm QSPR models concerning PFAS adsorption onto a hydrophobic interface. The Vm QSPR model established high correlations with both TEOS (C8) and TMOS (C18) log k values; indicating that molar volume can accurately represent the impact of size on cavity formation and hydrophobic interaction. Furthermore, the performance of the model suggests hydrophobic interaction is the main driving force for adsorption. By treating the X- axis as a hydrophobic index, it was noted that the hydrophobic surface area of the molecule increases from left to right. Thus, with increasing Vm, comes an increase in the molecule’s hydrophobic contact area with the surface; rendering higher adsorption rates due to increased hydrophobic interaction.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695

[0155] Now focusing on the EPmax model for Figs.10(c) and (d), we see a similar trend to that of Fig.10(b), where the EPmax model demonstrates higher performance than the Vm model. Increased model performance from the EPmax descriptor is likely do to the uniformity of a molecule’s electrostatic distribution being closely associated with its hydrophobicity and hydrophilicity. The magnitude of EPmax will decrease with a greater presence of electron-withdrawing fluorines. Thus, as the magnitude of EPmax decreases from left to right on the X-axis, the effects of the head group will diminish, and be outweighed by the tail group. Adsorption will then be more susceptible to the hydrophobic-interaction driven behavior of the tails. Moreover, the EPmax-based model can represent the influence of PFAS structure on hydrophobic driven interactions, as well as have improved prediction capabilities in comparison to the Vm model.

[0156] Inspection of Fig.10 shows that the Vm model performance increased as the hydrophobicity of the surface increased, ((a) < 5◦, R2 = 0.47, (b) 74.7◦, R2 = 0.83, (c) 100.3◦, R2 = 0.92, and (d) 110.5◦, R2 = 0.94), indicating Vm is an excellent molecular descriptor for predicting adsorption rate with a non-polar: hydrophobic interface present. The EPmax model performance did not change significantly with surface hydrophobicity. Instead, the model showed high performance on all modified surfaces (R2 ≥ 0.97); indicating the descriptors’ ability to provide robust predictions concerning electrostatic and hydrophobic driven adsorption. This is a result of both the molecule’s constituents contributing to the magnitude of the electrostatic potential.

[0157] Additionally, when the linear regression lines in Fig.10 are broken down, the slope of the models increases from (b)-(d) (Vm: 0.0026 < 0.0027 < 0.0036; EPmax: 0.0018 < 0.0028 < 0.0037). TEOS (C8)- and TMOS (C18)-functionalized surfaces may possess greater reaction rates (eq. slope) due to hydrophobic interfaces lacking an energy barrier. As stated previously, hydrophilic surfaces present energy barriers that PFAS have to overcome to disrupt water molecules at the surface and stably adsorb. The adsorption behavior of PFAS will correspond to the thermodynamic stability of the system. In contrast to non-polar surfaces, polar surfaces present lower minimum values of free energy

[0058] , indicating that the adsorption stability of PFAS on polar surfaces is lower than that of non-polar surfaces as the motion of water molecules has a greater impact. Therefore, when comparing the kinetics of hydrophobic and electrostatic attraction, the latter of the two will be the rate-limiting mechanism concerning diffusion.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695

[0158] The integrated use of engineered sorbents with QCM-D provided valuable insight into the fundamental mechanisms present for PFAS sorption. By studying these mechanisms at the molecular scale, complex behaviors and interactions could be observed. Hydrophobic and electrostatic interactions played the most prominent role in PFAS adsorption; however, electrostatic interactions were compromised by increasing chain length due to diluted head charge. An increased presence of PFAS was exhibited by all molecules at hydrophobic TMOS (C18)-functionalized interfaces, suggesting the importance of the hydrophobic interaction driving force for solid–liquid interfacial adsorption. This study also probed the molecular scale structural properties for solid–liquid interfaces, where results indicated that PFAS molecular geometry plays a strong role in determining film behavior; whereas surface hydration and adsorption stability were more dependent upon interfacial physicochemical properties. Results from the kinetic study confirmed silane-modified surfaces were more receptive to PFAS than unmodified surfaces, however, each surface induced different adsorption mechanisms. The QSPR analysis implemented established a correlation between PSO rate constants, molecular composition, and interfacial physicochemical properties. The results demonstrated that the EPmax model can provide robust predictions of PFAS adsorption rates at the solid–liquid interface with greatly varying structures, in contrast to the Vm model, which was only successful in predicting measured behavior at hydrophobic: non- polar interfaces. Thus, our results contribute to further understanding PFAS molecular-scale behavior, and its mechanisms of action near sorption sites; specifically, the contributions of electrostatic and hydrophobic interactions and their implications in optimizing PFAS retention to adsorbents, or for sorbent regeneration.

[0159] Example 2 Adhesion properties of PFAS on model surfaces

[0160] To understand the electrostatic contribution to PFAS adhesion force, streaming zeta potential measurements were taken of each functionalized surface. The electrical double layer (EDL) is a combination of the region defined as the layer of adsorbed water molecules and ions on a surface called the stern layer, and the diffuse regime above the stern layer expanding to the electrostatic boundary of a surface deemed the slipping plane. The zeta potential, the electrostatic potential measured at the outer boundary of the slipping plane, is widely utilized to estimate the general attraction or repulsion of molecules to a surface. The zeta potential exists due to acid–base reactions occurring on the material surface and adsorption of ions to the interface in solution. Modulating or designing a surface to carry a specific zeta potential is a modern technique used for many applications including enhancingUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 oil recovery from carbonate rock, rendering biomaterials inert to evade immune response, and separation and elution of polar molecules. The isoelectric point (IEP), the pH at which a surface's zeta potential is equal to zero, is often used to obtain information about the functional behavior of the interface. Fig.11a shows the effect of pH titration on the zeta potential of each wafer sample from pH 2–10 in 1 mM potassium chloride. The APTES sample expectedly showed a basic surface functionality compared to acidic C8, and amphoteric C18. The high isoelectric point (pH 8) of APTES is in close agreement with pKa values and IEP data reported by Bhat et al. who modulated the sorption of nanoparticles by pH titration of an APTES modified silicon wafer. The plasma treated control sample had an IEP of 5 and demonstrated amphoteric behavior, consistent with known pKa values and properties of silanols. Examination of individual zeta potential values revealed that the APTES sample carried a positive electrostatic potential, the C8 surface was negative, and the C18 surface was at its isoelectric point at environmentally relevant pH of 6.8. Hydrophobic surfaces are expected to carry a negative zeta potential. At neutral pH values as water molecules are repelled from the interface, hydroxide ions are attracted to the positive dipoles on the C–H chains that predominantly exist on the surface, generating a net negative electrostatic potential. The lower magnitude of zeta potential and difference in isoelectric point of the C18 sample compared to C8 can be attributed to the longer carbon chain length. The folding over and charge condensation of the C18’s longer carbon chains allows for the reduction in the net dipole moment of the molecule, reducing VdW interactions with hydroxide and thus resulting in a more positive surface potential and higher isoelectric point. Though the C18 film is of comparable hydrophobicity to C8, the longer chains convey lower surface roughness and amphoteric behavior compared to the shorter chain modification.

[0161] AFM was used to generate force curve profiles of PFOS and PFBS on the siloxane films in DI water. The baseline adhesion behavior of the polar-hydrophobic fluorocarbon chains in DI water is presented in Fig.11b. PFAS molecules comprised of either four or eight carbons, and sulfonate head groups (PFOS and PFBS) were utilized to determine if fluorocarbon chain length or surface composition was the dominant contributor to the adhesion force. Unmodified AFM tips were not used as a control comparison in this study as previous experimentation yielded significant adhesion between the silicon nitride tip and amine terminated APTES functionalization. It is important to note that because the sulfonate groups were used to functionalize the molecules to the AFM tip, we hypothesize that the resulting force values would mostly correspond to the interaction between the exposedUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 fluorocarbon tails and the substrates. Adhesion of both PFOS and PFBS was the weakest on the plasma treated surface. The high energy, hydrophilic surface of the plasma treated silicon wafer has little affinity for the hydrophobic PFAS tails. Of the surface modified samples, adhesion was lowest on the APTES film likely due to the excluded or inaccessible sulfonate groups, which minimize electrostatic attraction with the amine terminated surface. The negative dipoles of the fluorocarbon backbone, however, can interact with the amine groups, resulting in a Van der Waals adhesion force. The hydrophilicity of the APTES sample would also result in a lower hydrophobic interaction between the geometrically hydrophobic PFAS tails compared to the C8 and C18 surfaces. The PFBS functionalized tip had slightly higher adhesion on all surfaces (APTES: 129.21%, C8: 138.75%, C18: 161.52%) compared to PFOS. It is possible that the smaller molar volume and thus lower hydrophobicity of the fluorocarbon tail allows PFBS molecules to penetrate the adsorbed water layer more effectively on the sample surfaces.

[0162] Next, electrolyte solutions were introduced to examine how the presence of ions influence PFAS behavior at the interface. The specific behavior and magnitude of electrostatic interactions in the bulk solution and at the solid–liquid interface are impacted by the concentration of ions in solution. This behavior was interpreted through additional zeta potential and force measurements in 75 mM NaCl, and 75 mM MgCl2 solutions. Magnesium and sodium chloride were chosen as they are simple, low cost monovalent and divalent salts. Magnesium was selected over calcium chloride due to its smaller Van der Waals radius and ability to impart a kosmotropic ordering effect with its tightly coordinated water molecules. A concentration of 75 mM was chosen as it is believed that ion concentrations greater than 10 mM are capable of generating Hofmeister effects. Fig.12 shows the zeta potential values at pH 6.8 ± 0.3 of each siloxane film and AFM tip functionalization in 1 mM KCl or 75 mM MgCl2. The AFM tip functionalizations carry positive zeta potentials in 1 mM KCl likely due to the attraction of hydronium ions to the negative dipoles on the fluorine atoms of the fluorocarbon tail. The shift towards negative zeta potential values in the presence of 75 mM magnesium chloride has been reported previously and is believed to be the result of the attraction of chloride and hydroxide ions to surface adsorbed magnesium ions.

[0163] AFM adhesion force data are compiled for each surface PFAS-solution combination in Fig.13. Overall, the adhesion force of all PFAS molecules on each surface remained the lowest in deionized water. In deionized water, adsorbed water molecules on the film surface are at a maximum, potentially restricting access or lowering surface affinity toUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 hydrophobic molecules. Without ions in solution to adsorb to the film surface, the formation of a stable zeta potential and thus surface charge is not possible. The absence of ions also prevents bridging of dipoles on the C–F backbones to the functionalized surfaces. In DI water there is no significant driving force for PFAS adhesion.

[0164] The addition of sodium chloride in solution provided a negligible impact on PFAS adhesion on the plasma treated, APTES, and C8 surfaces which corresponds to similar findings discussed in literature. Sodium, compared to magnesium, has a lower charge density conveying a lower affinity for surfaces and molecules than magnesium. The C18 surface, however, showed a minor increase in adhesion force for both PFOS and PFBS in the 75 mM sodium chloride solution. In the presence of 75 mM MgCl2 the adhesion of PFOS and PFBS increased substantially on all surfaces. The increase in adhesion can be explained by several phenomena. The larger charge density and the water structuring effects of the magnesium ions compared to sodium creates a kosmotropic ordering effect. Magnesium interacting with both tip and interfacial surface carry between 6 and 32 strongly coordinated water molecules in their hydration shells depending on ionic strength. The increased structuring of the water hydrogen bonding network generated by magnesium ions creates a more pronounced Hofmeister effect resulting in greater expulsion of hydrophobic molecules from the bulk solution. This could also be responsible for enhancing PFAS sorption at the solid–liquid interface by structuring a hydrogen bonding network between tip and surface.

[0165] Ion-dipole or cation bridging interactions arising from the addition of divalent magnesium ions may also contribute to the adhesion of PFAS chains at the interface. While the fluorocarbon backbone is geometrically non-polar, individual carbon-fluorine bonds have strong permanent dipoles (−1.43 Debyes) that can interact with the magnesium ions and potentially bridge the C–F tails to the interface. A recent study by Huang et al. has demonstrated that introducing sites for ion-dipole interactions on hydrophobic fluorocarbon polymers can increase their adhesion strength by 62%. Divalent cations are also known for enhancing the partitioning of PFAS to solid–liquid and air–liquid interfaces. The suggested mechanisms for which are charge suppression of anionic PFAS headgroups that would be electrostatically repelled from negatively charged surfaces and cation bridging between PFAS and surfaces. Because only the PFAS tails are believed to be exposed in this example, cation bridging is the more likely explanation for the observed change in adhesion in this example.

[0166] The PFAS adhesion trends seen in Fig.13 can also be described in terms of the polarizability of each film. The zeta potential data shown in Fig.12 is in agreement withUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 previous research that has concluded that the polarizability of alkyl molecules increases with carbon chain length. It is also well understood that the strength of induced dipole interactions increases with increasing magnitude of zeta potential. The PFAS adhesion patterns in Fig.13 show that with increasing polarizability of probe-surface combination, a stronger adhesion force is generated. The C18 film, however, is an exception to this explanation as it has the highest zeta potential, and thus polarizability, but low adhesion force compared to other sample combinations. Because AFM measurements are physical in nature, the surface mechanical properties play a significant role. It is likely that because the C18 film is soft compared to the C8 and APTES films, during contact with the AFM tip, a mechanically stable interaction does not form. In addition, the C18 film had a roughness value lower than C8 and comparable to the plasma treated silicon wafer indicating again that possibly a layering or folding phenomenon may be at play. Further investigation into the influence of film viscoelastic properties is required to fully understand PFAS adhesion to complex film surfaces.

[0167] In summary, this example demonstrates the potential of utilizing cation solutions to manipulate PFAS adhesion. Examination of model surfaces composed of siloxane films yielded that adhesion of the PFAS fluorocarbon backbone can be enhanced with the addition of magnesium chloride. AFM force spectroscopy and zeta potential measurements revealed that the addition of 75 mM magnesium chloride provided the greatest enhancement to adhesion for both four and eight carbon PFAS on all sample surfaces. The proposed mechanism for the augmentation of PFAS adhesion is threefold. First, magnesium ions bind to the film surfaces and interact with the dipoles on the PFAS fluorocarbon tails, bridging the PFAS to the surface upon contact. Second, the addition of magnesium ions induces a kosmotropic ordering effect in the bulk solution, creating a higher energy water structure that enhances the hydrophobic effect. Lastly, the impact of magnesium on the zeta potential and thus polarizability of the film-tip interface increases the magnitude of induced dipole strength, enhancing Van der Waals forces upon contact.

[0168] Example 3 Novel pre-filter for efficiency improvement of Trace Contaminant Control Subassembly

[0169] This Example shows use of a pre-filter system to efficiently adsorb fluorocarbon in the air before it enters a building, a room, or other enclosed space, such as the International Space Station (ISS) or a spacecraft, or an aircraft, which requires relatively high air quality.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695

[0170] Air quality is a primary metric aboard the International Space Station (ISS) that is constantly maintained and monitored to ensure crew safety during missions in low earth orbit. The Trace Contaminant Control Subassembly (TCCS) (Figure 14, Top) is a key component of the air revitalization system aboard the ISS that is responsible from remediating ammonia, carbon monoxide, and other trace contaminants from the recirculating cabin air supply. The TCCS works in conjunction with other air revitalization assemblies to ensure that air delivered to the crew meets NASA and Russian Space Agency (RSA) quality standards. It is composed of three main elements: a fixed granular activated carbon (GAC) bed, a thermal catalytic oxidation assembly (COA), and a sorbent bed assembly (SBA). The GAC bed exists to capture incoming ammonia and volatile organic compounds (VOCs) produced by the crew or via off-gassing of materials. The COA serves to oxidize any incoming methane and VOCs that break through the GAC into carbon dioxide and water. Any acid gases formed by the oxidation of VOCs in the COA are captured by the SBA, allowing a clean stream of air to recirculate through the spacecraft. The TCCS is currently limited by its inability to safely remediate certain contaminants that persist in the cabin air supply. The disclosed design comprises a pre-filter to the GAC assembly (Figure 14, Bottom) to shield the downstream components of the TCCS and other life support assemblies in the event of a fluorocarbon leak.

[0171] Current State of GAC Technology-At present, the efficiency of the TCCS heavily relies upon a bed of granular activated carbon (GAC) to protect the catalysts within the COA from irreversible contamination. Certain molecules such as halocarbons, sulfur containing molecules, and siloxanes have been documented to prevent either the catalyst itself from properly functioning or to interfere with a corresponding oxidation reaction. Most of the time, these problematic molecules are captured and retained by the GAC, however, over time there is increasing potential for them to become saturated and break through column. Once the GAC within the TCCS becomes saturated and breakthrough occurs, it must be replaced with new media. The current cost to send a single GAC replacement cannister for the TCCS on a SpaceX Falcon 9 rocket is approximately $62,000. The GAC within the TCCS is currently aboard the ISS is not regenerated, resulting in unnecessary financial loss if it becomes prematurely saturated.

[0172] Although GAC can effectively shield the COA from problematic contaminants, it has high mass compared to alternative sorbents making it expensive to launch, it requires energetically intensive regeneration, and it is non-selective for the molecules that it absorbs.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 The current paradigm of using GAC as a protectant has significant potential for improvement by low complexity enhancements that would expand its lifetime, thereby lowering the cost of TCCS operation and maintenance.

[0173] The unique properties of fluorocarbons make them ideal candidates as working fluids for thermal control. Currently several alternative perfluorinated compounds are being tested to replace ammonia and water in the ISS heat exchange loops. A recent assessment into the environmental impacts of each candidate fluorocarbon concluded that all of them adsorb onto the GAC beds of the TCCS modules and the zeolites within the carbon dioxide regeneration assembly (CDRA). While this seems like a benefit for environmental control in the event of another leak incident, several of the selected fluids have low loading capacities and rapidly foul sorbent media due to their large size and low vapor pressures. Fluorocarbons and other halocarbons produced via off gassing of new materials delivered to the ISS have also been demonstrated to poison the methane oxidation reaction within the COA. While again, most can be adsorbed by the GAC assembly, certain halocarbons are able to reach saturation more rapidly than others and break through to the COA interfering with its normal function. Residual acid gases produced by halocarbons that are successfully oxidized, are neutralized by lithium hydroxide within the SBA, leading to eventual premature exhaustion. Estimates by Lockheed Martin and experimentation by NASA has determined that the SBA as an approximate life span of 3.6 years under ideal conditions, however it has the potential to be lengthened if a lower fluorocarbon and halocarbon flux through the GAC and COA is achieved.

[0174] Each of these cases directly results in the premature exhaustion of GAC and subsequently the SBA, leading to additional dependance on consumables. Improvements to sorbent and catalyst protection are imperative to maintaining air quality to NASA and RSA standards and to ensure that the ecological life support assemblies are protected from undesirable premature contaminant saturation. As fluorocarbons continue to be utilized in long term spaceflight, there is critical need for complete and regenerable fluorocarbon control methods.

[0175] To increase the lifespan of the GAC within the TCCS and to better protect other air revitalization assemblies from fluorocarbon contamination, a new upstream sorbent media is added. A fluorocarbon pre-filter to the TCCS acts as a fail-safe to prevent rapid saturation of GAC in the event of a leak in the thermal control system and provides additional protection against other halocarbons. Adsorption of fluorocarbons, prior to contact with GACUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 and the downstream components of the TCCS, prolongs the lifespan of consumables and protects the catalysts within the COA from contamination, thereby saving crew time, launching costs, and ISS energy expenditures.

[0176] Cellulose sponges are used in the sorbent application as they are commercially available, lightweight, have tailorable porosity, high surface area, and are readily modified. The adsorption of fluorocarbons on a sorbent media is achieved by maximizing electrostatic and Van Der Waals interactions between the adsorbent matrix and contaminant molecules. Results in the Examples above have demonstrated the capability to modify cellulose sponges with alkyl silanes. The modified cellulose sponge has a combination of high surface area, porosity, and selectivity which make it an ideal sorbent for fluorocarbon contaminant control.

[0177] Commercially available and synthesized cellulose sponges are modified with a combination of trimethoxy(octadecyl)silane (TMOS), triethoxy(octyl)silane (TEOS), and 3- aminopropyltrichlorosilane (APTES) to generate a hydrophobic, slightly positively charged surface that is complementary the properties of perfluorinated molecules (Figure 15). Because the interactions between the fluorinated molecules and adsorbent is relatively weak, the sponge is regenerable in low vacuum at ambient temperatures.

[0178] The strength of adsorption interactions between the selected silanes (TMOS, TEOS, and APTES) and fluorocarbons is investigated.

[0179] Cellulose sponges are fabricated by lyophilization from alpha cellulose and cross- linking molecules. Bench-scale and full-scale sponges are surface modified under the conditions determined in previous experiments. Porosity and BET specific surface area are determined by SEM and the nitrogen gas method respectively. The mass ratio of fluorocarbons removed to quantity of modification is used as a metric for removal efficiency. All adsorption capacities are compared to the previously generated Polyani plots and to GAC for reference.

[0180] Candidate filters from previous experiments are assessed for vacuum regeneration capability at varying pressures, temperatures, and relative humidities. Adsorption and regeneration are repeatedly tested to evaluate the life cycle of the sorbent matrix.

[0181] Example 4 Modified cellulose sponges for PFAS adsorption

[0182] Cellulose is one of the most abundant polymer on earth and can be easily and economically extracted from any plant biomass. It can be transmuted into many forms including microcrystals, microfibers, microcrystals, nanofibers, hydrogels, and three- dimensional scaffolds such as sponges. The variability of structures that can be manufacturedUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 with cellulose allows for tailoring mediums to a specific need as has been recently done for PFAS adsorption applications. Functionalization of cellulose for PFAS adsorption is an increasingly utilized technique being examined in the literature, particularly with micro / nanocrystalline cellulose.

[0183] While previous studies are pioneering in terms of providing a low cost and regenerable alternative to granular activated carbon (GAC), each study focuses on grafting cationic molecules to cellulose for enhancement of electrostatic interactions with PFAS. However, the impact of hydrophobic surface modifications on cellulose and their resulting impact on PFAS sorption remains unknown.

[0184] In this Example, as shown in Fig.16, commercially available cellulose sponges were surface modified with cationic, short chain hydrophobic, and long chain hydrophobic silane molecules to observe their impact on the adsorption of six different PFAS. The baseline regeneration capabilities of each modified sponge were assessed by using deionized water and 0.5M potassium nitrate in 10% methanol as regeneration solvents. This work serves as a foundation for investigating combinations of surface modifications and regenerant solutions that aim to optimize the PFAS adsorption and sorbent regeneration processes.

[0185] Regeneration of PFAS sorbents is a widely desired characteristic of novel materials. To ascertain the baseline PFAS desorption qualities of the cellulose matrices, a cycling study was conducted over ten days. Both adsorbed and desorbed PFAS concentrations were quantitated via LC-MS / MS. The process was repeated for five adsorption and desorption cycles. Individual cycle data is presented in Figure 17 while averaged removal and recovery data is provided in Table 1. The total combined PFAS removal values for the unmodified sponge, APTES, C8, C18 modified sponges, and GAC were 87.6%, 72.3%, 66.2%, 86.0%, and 88.8% respectively. Interestingly, both the unmodified sponge and C18 modified sponge had comparable PFAS removal to GAC.

[0186] Examining individual PFAS, unmodified cellulose sponge had the highest PFOA removal (84.3% ± 4.9) of all samples. APTES and C8 modified surfaces demonstrated the highest PFOS removal (97.3% ± 0.9 and 99.2% ± 0.1) while C18 had the greatest PFHxS removal (99.5% ± 0.3). PFHxA, PFBS, and PFBA adsorbed best onto GAC with 91.4% ± 4.6, 93.9% ± 2.0 and 96.3% ± 1.6 removal respectively. Each sample type had a unique affinity for specific PFAS that either increased or decreased over the sorption cycles. PFAS removal increased in multiple cases e.g. – PFOA on the unmodified sponge seen in Figure 17, likely due to multilayering of PFAS on the sorbent surface, resulting in further fluorine-UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 fluorine interactions and thus more favorable adsorption in subsequent cycles. Reductions in removal are also observed, exemplified by the PFBA adsorption on the APTES modified sponge. Because PFBA is the least hydrophobic of the assessed PFAS, it is the least likely to form multiple layers.

[0187] Table 1. Averaged five-cycle mass removal and recovery data in 0.5M KNO3 in 10% MeOH Removal Capacity (mg / g) Sample PFOA PFOS PFBA PFBS Unmodified1.183 ± 0.974 0.710 ± 0.260 0.875 ± 0.197 0.356 ± 0.162APTES1.357 ± 0.673 0.745 ± 0.2151.22 ± 0.271 0.417 ± 0.118 C182.130 ± 0.553 0.793 ± 0.165 1.061 ± 0.251 0.450 ± 0.058GAC2.423 ± 0.828 0.635 ± 0.153 1.892 ± 0.386 1.299 ± 0.104% Recovery (%) Sample PFOA PFOS PFBA PFBS Unmodified38.245 ± 20.167 96.42 ± 37.493 4.873 ± 3.220 23.577 ± 0.162APTES25.837 ± 5.193 86.102 ± 24.914 4.857 ± 2.169 16.99 ± 9.160C1830.929 ± 15.672 65.911 ± 29.755 5.264 ± 1.915 26.11 ± 11.750GAC1.622 ± 0.979 7.458 ± 7.203 40.049 ± 9.599 27.388 ± 3.808

[0188] Unmodified cellulose and C18 modified cellulose had comparable PFAS adsorption to GAC while having a fraction of the surface area. Each of these surfaces relied on hydrophobic interaction for adsorption while the APTES modified sponge utilized electrostatic attraction. Regeneration was attempted with pure DI water to establish a baseline for PFAS desorption in idealized conditions. Different PFAS were able to be recovered from the sponges compared to GAC depending on surface modification. The C8 modified sponge had the best recovery for the majority of PFAS while APTES had better recovery of short chain molecules. Overall PFAS recovery in DI water expectedly was low on average as compared to other desorption solutions reported in literature. Subsequently, regeneration in methanolic potassium nitrate was also attempted. Significant increases to long chain PFAS regeneration on all cellulose samples were observed while GAC saw only increases in regeneration to short chain molecules. The results of this study show that cellulose modified materials have comparable PFAS removal to GAC over the tested number of adsorption- regeneration cycles with increases in PFAS recovery in DI water as compared to PFAS recovery in GAC. This study serves as the groundwork for investigation of further cellulose modification and regeneration experiments. This work demonstrates the potential forUOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 cellulose sponges as a candidate for large scale PFAS remediation and contributes to the understanding of cellulosic materials.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 Cited Literature All references cited in this disclosure, including but not limited to those listed below, are hereby incorporated by reference into this disclosure. ACS Omega 2019, 4, 8001-8006 J. Molecular Liquids 326, 15 March 2021, 115336 Separation and Purification Technology 256, 1 Feb 2021, 117787 Rapid Removal of Poly- and Perfluorinated Alkyl Substances by Poly(ethylenimine)- Functionalized Cellulose Microcrystals at Environmentally Relevant Conditions; Mohamed Ateia, Mohamed F. Attia, Amith Maroli, Nishanth Tharayil, Frank Alexis, Daniel C. Whitehead, and Tanju Karanfil; Environmental Science & Technology Letters; 20185 (12), 764 – 769 Efficient PFAS Removal by Amine-Functionalized Sorbents: Critical Review of the Current Literature; Mohamed Ateia, Alaaeddin Alsbaiee, Tanju Karanfil, and William Dichtel; Environmental Science & Technology Letters; 20196 (12), 688 – 695 US Patent Publication No.20200399147A1 US Patent Publication No.20110024361A1 Amani, P., & Firouzi, M. (2022). Effect of Divalent and Monovalent Salts on Interfacial Dilational Rheology of Sodium Dodecylbenzene Sulfonate Solutions. Colloids and Interfaces, 6(3), Article 3. https: / / doi.org / 10.3390 / colloids6030041Lunkenheimer, K., Geggel, K., & Prescher, D. (2017). Role of Counterion in the Adsorption Behavior of 1:1 Ionic Surfactants at Fluid Interfaces—Adsorption Properties of Alkali Perfluoro-n-octanoates at the Air / Water Interface. Langmuir, 33(39), 10216–10224. https: / / doi.org / 10.1021 / acs.langmuir.7b00786UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 Kotlarz, Nadine, James McCord, David Collier, C. Suzanne Lea, Mark Strynar, Andrew B. Lindstrom, Adrien A. Wilkie, et al. “Measurement of Novel, Drinking Water-Associated PFAS in Blood from Adults and Children in Wilmington, North Carolina.” Environmental Health Perspectives 128, no.7 (July 2020): 077005. https: / / doi.org / 10.1289 / EHP6837. Perng, Wei, Dorothy Nakiwala, and Jaclyn M.Does Not Stay In Utero: A Review of Evidence for Prenatal Epigenetic Programming by Per- and Polyfluoroalkyl Substances (PFAS) in Infants, Children, and Adolescents.” Current Environmental Health Reports 10, no.1 (March 1, 2023): 35–44. https: / / doi.org / 10.1007 / s40572- 022-00387-z.L., A. Beneito, M. Casas, A. Colles, L. Dalsager, E. Den Hond, C. Dereumeaux, et al. “PFAS Levels and Exposure Determinants in Sensitive Population Groups.” Chemosphere 313 (February 2023): 137530. https: / / doi.org / 10.1016 / j.chemosphere.2022.137530. Cakmak, Sabit, AnnaRobert Dales. “The Association between Blood PFAS Concentrations and Clinical Biochemical Measures of Organ Function and Metabolism in Participants of the Canadian Health Measures Survey (CHMS).” Science of The Total Environment 827 (June 2022): 153900. https: / / doi.org / 10.1016 / j.scitotenv.2022.153900. Blomberg, Annelise J., Line S. Haug, Christian Lindh, Azemira Sabaredzovic, Daniela Pineda, Kristina Jakobsson, and Christel Nielsen. “Changes in Perfluoroalkyl Substances (PFAS) Concentrations in Human Milk over the Course of Lactation: A Study in Ronneby Mother-Child Cohort.” Environmental Research 219 (February 15, 2023): 115096.https: / / doi.org / 10.1016 / j.envres.2022.115096. Belkouteb, Nadine, Vera Franke, Philip McCleaf, Stephan Köhler, and Lutz Ahrens. “Removal of Per- and Polyfluoroalkyl Substances (PFASs) in a Full-Scale Drinking Water Treatment Plant: Long-Term Performance of Granular Activated Carbon (GAC) and Influence of Flow-Rate.” Water Research 182 (September 2020): 115913. https: / / doi.org / 10.1016 / j.watres.2020.115913.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 Siriwardena, Dinusha P., Ryan James, Kavitha Dasu, Jonathan Thorn, Ramona Darlington Iery, Franco Pala, Denise Schumitz, Stephanie Eastwood, and Nathan Burkitt. “Regeneration of Per- and Polyfluoroalkyl Substance-Laden Granular Activated Carbon Using a Solvent Based Technology.” Journal of Environmental Management 289 (July 2021): 112439. https: / / doi.org / 10.1016 / j.jenvman.2021.112439.James F. Reuther, Navid B. Saleh, Arjun K. Venkatesan, and Onur G. Apul. “Thermal Regeneration of Spent Granular Activated Carbon Presents an Opportunity to Break the Forever PFAS Cycle.” Environmental Science & Technology 55, no.9 (May 4, 2021): 5608–19. https: / / doi.org / 10.1021 / acs.est.0c08224. Karbassiyazdi, Elika, Medha Kasula, SwetaKalantari, Ali Altaee, Milad Rabbani Esfahani, and Amir Razmjou. “A Juxtaposed Review on Adsorptive Removal of PFAS by Metal-Organic Frameworks (MOFs) with Carbon-Based Materials, Ion Exchange Resins, and Polymer Adsorbents.” Chemosphere 311 (January 2023): 136933. https: / / doi.org / 10.1016 / j.chemosphere.2022.136933.Van den Bergh, Matthias, Andraž Krajnc, Stefan Voorspoels, Sergio Rodrigues Tavares, Steven Mullens, Isabelle Beurroies, Guillaume Maurin, Gregor Mali, and Dirk E. De Vos. “Highly Selective Removal of Perfluorinated Contaminants by Adsorption on All-Silica Zeolite Beta.” Angewandte Chemie 132, no.33 (2020): 14190–94.https: / / doi.org / 10.1002 / ange.202002953. Ateia, Mohamed, Alaaeddin Alsbaiee, Tanju Karanfil, and William Dichtel. “Efficient PFAS Removal by Amine-Functionalized Sorbents: Critical Review of the Current Literature.” Environmental Science & Technology Letters 6, no.12 (December 10, 2019): 688–95. https: / / doi.org / 10.1021 / acs.estlett.9b00659.Murray, Conner C., Robert E. Marshall, Charlie J. Liu, Hooman Vatankhah, and Christopher L. Bellona. “PFAS Treatment with Granular Activated Carbon and Ion Exchange Resin: Comparing Chain Length, Empty Bed Contact Time, and Cost.” Journal of Water Process Engineering 44 (December 2021): 102342. https: / / doi.org / 10.1016 / j.jwpe.2021.102342.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 Brown, Douglas, and Alan LeBlanc. “Navigating PFAS Treatment With GAC and Ion Exchange.” Journal AWWA 114, no.7 (2022): 36–43. https: / / doi.org / 10.1002 / awwa.1959. Gagliano, Erica, Massimiliano Sgroi, Pietro P. Falciglia, Federico G. A. Vagliasindi, and Paolo Roccaro. “Removal of Poly- and Perfluoroalkyl Substances (PFAS) from Water by Adsorption: Role of PFAS Chain Length, Effect of Organic Matter and Challenges in Adsorbent Regeneration.” Water Research 171 (March 15, 2020): 115381. https: / / doi.org / 10.1016 / j.watres.2019.115381.and Hans-Jörg Schneider. “The Hydrophobic Effect Revisited-Studies with Supramolecular Complexes Imply High-Energy Water as a Noncovalent Driving Force.” Angewandte Chemie International Edition 53, no.42 (October 13, 2014): 11158–71.https: / / doi.org / 10.1002 / anie.201310958Yoon, Roe-Hoan, Darrin H. Flinn, and Yakov I. Rabinovich. “Hydrophobic Interactions between Dissimilar Surfaces.” Journal of Colloid and Interface Science 185, no.2 (January 15, 1997): 363–70. https: / / doi.org / 10.1006 / jcis.1996.4583. Cui, Xin, Jing Liu, Lei Xie, Jun Huang, Qi Liu, Jacob N. Israelachvili, and Hongbo Zeng. “Modulation of Hydrophobic Interaction by Mediating Surface Nanoscale Structure and Chemistry, Not Monotonically by Hydrophobicity.” Angewandte Chemie 130, no.37 (2018): 12079–84.https: / / doi.org / 10.1002 / ange.201805137. Tabor, Rico F., Franz Grieser, Raymond R. Dagastine, and Derek Y. C. Chan. “The Hydrophobic Force: Measurements and Methods.” Phys. Chem. Chem. Phys.16, no.34 (2014): 18065–75.https: / / doi.org / 10.1039 / C4CP01410C. Moelbert, Susanne, B. Normand, and Paolo De Los Rios. “Kosmotropes and Chaotropes: Modelling Preferential Exclusion, Binding and Aggregate Stability.” Biophysical Chemistry 112, no.1 (December 1, 2004): 45–57. https: / / doi.org / 10.1016 / j.bpc.2004.06.012.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 Ball, Philip, and John E. Hallsworth. “Water Structure and Chaotropicity: Their Uses, Abuses and Biological Implications.” Physical Chemistry Chemical Physics 17, no.13 (2015): 8297– 8305. https: / / doi.org / 10.1039 / C4CP04564E.“Influence of Ionic Strength on Hydrophobic Interactions in Water: Dependence on Solute Size and Shape.” The Journal of Physical Chemistry B 124, no.46 (November 19, 2020): 10326–36. https: / / doi.org / 10.1021 / acs.jpcb.0c06399. Gregory, Kasimir P., Gareth R. Elliott, Hayden Robertson, Anand Kumar, Erica J. Wanless, Grant B. Webber, Vincent S. J. Craig, Gunther G. Andersson, and Alister J. Page. “Understanding Specific Ion Effects and the Hofmeister Series.” Physical Chemistry Chemical Physics 24, no.21 (June 1, 2022): 12682–718. https: / / doi.org / 10.1039 / D2CP00847E.Luo, Peng, Yanqin Zhai, Erkan Senses, Eugene Antonio Faraone. “Influence of Kosmotrope and Chaotrope Salts on Water Structural Relaxation.” The Journal of Physical Chemistry Letters 11, no.21 (November 5, 2020): 8970–75. https: / / doi.org / 10.1021 / acs.jpclett.0c02619. Yang,for the Impact of Ionic Liquids on Biocatalysis.” Journal of Biotechnology, Industrial Biotechnology: Current Status and Future Development for the Sustainability of Human Society, 144, no.1 (October 12, 2009): 12–22.https: / / doi.org / 10.1016 / j.jbiotec.2009.04.011.and Dispersion of Small Hydrophobic Particles in Aqueous Electrolyte Solutions.” The Journal of Physical Chemistry B 110, no.45 (November 1, 2006): 22736–41.https: / / doi.org / 10.1021 / jp064475+.Butt, Hans-Jürgen, Brunero Cappella, and Michael Kappl. “Force Measurements with the Atomic Force Microscope: Technique, Interpretation and Applications.” Surface Science Reports 59, no.1–6 (October 2005): 1–152. https: / / doi.org / 10.1016 / j.surfrep.2005.08.003.Jin, Tonghui, Mohammad Peydayesh, Hanna Joerss, Jiangtao Zhou, Sreenath Bolisetty, and Raffaele Mezzenga. “Amyloid Fibril-Based Membranes for PFAS Removal from Water.”UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 Environmental Science: Water Research & Technology 7, no.10 (2021): 1873–84. https: / / doi.org / 10.1039 / D1EW00373A. Mohona, Tashfia M., Ning Dai, and Prathima C. Nalam. “Comparative Degradation Kinetics Study of Polyamide Thin Films in Aqueous Solutions of Chlorine and Peracetic Acid Using Quartz Crystal Microbalance.” Langmuir 37, no.48 (December 7, 2021): 14214–27. https: / / doi.org / 10.1021 / acs.langmuir.1c02835. Mohona, Tashfia M., Zhijiang Ye, Ning Dai, and Prathima C. Nalam. “Adsorption Behavior of Long-Chain Perfluoroalkyl Substances on Hydrophobic Surface: A Combined Molecular Characterization and Simulation Study.” Water Research 239 (July 2023): 120074. https: / / doi.org / 10.1016 / j.watres.2023.120074.Kobayashi, Tomoya Ueda, and Masaru Tanaka. “Effect of Bound Water Content on Cell Adhesion Strength to Water-Insoluble Polymers.” Acta Biomaterialia 134 (October 2021): 313–24. https: / / doi.org / 10.1016 / j.actbio.2021.07.058.

Claims

UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 Claims 1. A process for purifying an initial liquid or air source comprising liquid or air and an initial amount of one or more of per- and polyfluorinated organics (PFAS), the process comprising: contacting the initial liquid or air source with a PFAS sorbent composition comprising a sorbent substrate and a surface modifier composition, and obtaining a purified liquid or air product comprising a reduced amount of PFAS.

2. The process of claim 1, wherein the initial liquid or air source is an initial water source, and the purified liquid or air product is purified water.

3. The process of claim 1, wherein the surface modifier composition comprises an alkyl silane, optionally a linear alkyl silane, optionally trimethoxy(octadecyl)silane (TMOS), triethoxy(octyl)silane (TEOS), or (3-aminopropyl) triethoxysilane (APTES), or a combination thereof.

4. The process of claim 2, wherein the alkyl silane comprises a carbon chain having from 6 to 20 carbons.

5. The process of claim 1, wherein the sorbent substrate comprises a cellulose-based substrate, optionally a cellulose sponge.

6. The process of claim 1, wherein the surface modifier composition comprises charged silanes and uncharged silanes and the ratio (w / w) of charged silanes to uncharged silanes ranges from 0.1:1 to 0.1:

10.

7. The process of claim 1, wherein the surface modifier composition is an alkaline solution.

8. The process of claim 7, wherein the surface modifier composition increases hydroxyl group content on the sorbent substrate.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 9. The process of claim 1, wherein the surface modifier composition comprises a metal oxide, optionally an aluminum oxide.

10. The process of claim 1, wherein the surface modifier composition comprises a divalent cation, optionally the divalent cation is magnesium or calcium, and optionally, the concentration of the divalent cation is between 1 mM and 200 mM, or between 50 mM and 100 mM.

11. The process of claim 1, wherein the surface modifier composition imparts a charge to the sorbent substrate.

12. The process of claim 1, wherein the sorbent composition demonstrates a contact angle greater than 60 degrees, or optionally greater than 90 degrees.

13. The process of claim 1, wherein the sorbent composition demonstrates Zeta potential less than 18 mV at a pH of 6.

14. The process of claim 1, wherein the PFAS comprise perfluorooctanoic acid (PFOA), perfluorooctanesulfonic acid (PFOS), perfluorohexanoic acid (PFHxA), perfluorohexanesulfonic acid (PFHxS), perfluorobutanoic acid (PFBS), and / or perfluorobutanesulfonic acid (PFBS).

15. The process of claim 1, further comprising the step of modifying the sorbent based on the characteristics of the water stream.

16. The process of claim 1, wherein the purified water comprises less than 75% of the initial amount of PFAS.

17. The process of claim 1, wherein the contacting of the initial water source with the sorbent composition removes at least 25% of the initial amount of PFOA.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 18. The process of claim 1, wherein the contacting of the initial water source with the sorbent composition removes at least 62% of the initial amount of PFOS.

19. The process of claim 1, wherein the purified water stream comprises less than 1000 ppm PFAS.

20. The process of claim 1, wherein the contacting is conducted at a low pressure drop.

21. The process of claim 1, wherein the process generates a small amount of additional fluid, based on the weight of the water stream.

22. The process of claim 1, wherein the contacting step yields a spent sorbent composition comprising the sorbent composition and the PFAS contaminants.

23. The process of claim 1, further comprising a step of regenerating the spent sorbent to form a regenerated sorbent.

24. The process of claim 23, wherein the regeneration comprises the step of removing PFAS contaminants from the spent sorbent.

25. The process of claim 23, further comprising the step of contacting a second water source with the regenerated sorbent to yield second purified water.

26. The process of claim 25, wherein the steps of contacting, regenerating, and contacting are repeated for at least 5 cycles.

27. The process of claim 1, wherein the initial liquid or air source is an initial air source, and the PFAS sorbent composition is positioned as a pre-filter upstream of an apparatus comprising a granular activated carbon (GAC) assembly and a catalytic oxidation assembly, wherein the initial air source passes through the PFAS sorbent composition and then through the granular activated carbon (GAC) assembly and the catalytic oxidation assembly.UOAZ Ref. No. UA21-161Atty. Ref. No.: UOAZ.P2145WO / 00628695 28. A surface-modified sorbent composition for removing one or more of PFAS from an air stream comprising air and one or more of PFAS, the surface-modified sorbent composition comprising a sorbent substrate and a surface modifier composition, wherein the surface modifier composition comprises an alkyl silane, optionally the alkyl silane comprising one or more of trimethoxy(octadecyl)silane (TMOS), triethoxy(octyl)silane (TEOS), or (3- aminopropyl) triethoxysilane (APTES), or a combination thereof.

29. The surface-modified sorbent composition of claim 28, wherein the sorbent composition is contained in a pre-filter assembly positioned upstream of an apparatus comprising a granular activated carbon (GAC) assembly and a catalytic oxidation assembly.

30. A surface-modified sorbent composition for removing one or more of PFAS from a water stream comprising water and one or more of PFAS, the surface-modified sorbent composition comprising a sorbent substrate and a surface modifier composition, wherein the surface modifier composition comprises an alkyl silane, optionally the alkyl silane comprising one or more of trimethoxy(octadecyl)silane (TMOS), triethoxy(octyl)silane (TEOS), or (3- aminopropyl) triethoxysilane (APTES), or a combination thereof.

31. The surface-modified sorbent composition of claim 30, having a capacity for PFOA removal higher than 1.2 mg PFOA / g sorbent and a PFOA recovery in regeneration process of higher than 10%, or 20%.

Images