Rinse-free molecular layer-by-layer polyamide membrane fabrication

Abstract

Provided herein are methods for fabricating a polyamide reverse osmosis and nanofiltration membrane. The methods include depositing a first monomer solution comprising trimesoyl chloride (TMC) onto a substrate, depositing a second monomer solution comprising m-phenylenediamine (MPD) or piperazine (PIP) onto the substrate, and repeating the first and second depositing steps one or more times to form a polyamide film, wherein the first and second monomer solutions have concentrations ranging from about 0.01% to about 0.1% by mass fraction, and wherein the method does not include rinsing steps between the monomer depositing steps. Also provided herein are systems for fabricating a polyamide reverse osmosis membranes.

Description

RINSE-FREE MOLECULAR LAYER-BY-LAYER POLYAMIDE MEMBRANE FABRICATIONCROSS REFERENCE TO RELATED APPLICATION

[0001] The application claims the benefit of U.S. Provisional Application No. 63 / 790,386 filed on April 17, 2025, the entire contents of which are incorporated by reference herein.STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

[0002] This invention was made with government support under 2232686 awarded by the National Science Foundation (NSF). The government has certain rights in the invention.FIELD OF THE INVENTION

[0003] The present disclosure relates to reverse osmosis (RO) and nanofiltration (NF) membrane fabrication, and more particularly to rinse-free molecular layer-by-layer deposition methods for producing polyamide-based RO and NF membranes.BACKGROUND OF THE DISCLOSURE

[0004] Reverse osmosis (RO) and nanofiltration (NF) membranes play a crucial role in water purification and desalination processes. These semipermeable membranes allow water molecules to pass through while rejecting dissolved ions / salts and other contaminants. Polyamide-based thin film composite membranes have become widely used for reverse osmosis applications due to their excellent salt rejection and water permeability characteristics.

[0005] Conventional methods for fabricating polyamide reverse osmosis and nanofiltration membranes typically involve interfacial polymerization between an aqueous diamine solution and an organic multifunctional acid chloride solution. While effective, this approach can result in membranes with relatively high surface roughness and heterogeneous structures, which may impact performance and fouling resistance.

[0006] Molecular layer-by-layer deposition has emerged as an alternative technique for producing polyamide membranes with more controlled structures. Molecular layer-by-layer deposition involves sequentially depositing alternating layers of complementary monomers to1130761.00568 / 157813095v.1build up the membrane one molecular layer at a time. However, traditional molecular layer-by-layer processes often require multiple rinsing steps between monomer depositions to remove excess unreacted material.

[0007] The rinsing steps in conventional molecular layer-by-layer deposition can be timeconsuming and resource-intensive. They may also introduce variability into the membrane fabrication process. There is interest in developing simplified molecular layer-by-layer approaches that can produce high-quality polyamide membranes while reducing or eliminating intermediate rinsing steps.

[0008] Advances in membrane fabrication techniques that allow for greater control over membrane structure and properties while also improving manufacturing efficiency could benefit reverse osmosis and nanofiltration applications across water treatment, desalination, and other separation processes. Certain membrane characteristics, such as thickness, porosity, crosslinking density, and surface morphology can significantly impact separation performance, fouling resistance, and other key properties.

[0009] As global demand for clean water continues to rise, there is an ongoing need for innovations in reverse osmosis membrane technology and fabrication methods. Improvements in membrane performance, durability, and manufacturing could help expand access to water purification and desalination capabilities worldwide.SUMMARY

[0010] This summary is provided to introduce a selection of concepts in a simplified form that are further described below in the detailed description. This summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used as an aid in determining the scope of the claimed subject matter.

[0011] Various aspects of the present disclosure are directed to methods for fabricating a polyamide reverse osmosis membrane. Methods according to various aspects of the disclosure include depositing a first monomer solution comprising trimesoyl chloride (TMC) onto a substrate, depositing a second monomer solution comprising m-phenylenediamine (MPD) onto the substrate, and repeating the depositing steps one or more times to form a polyamide film. The first and second monomer solutions may have concentrations ranging from about 0.01% to about 0.1% by mass fraction. In some instances, methods according to the disclosure do not include rinsing steps2130761.00568 / 157813095v.1between monomer depositions. Tn some instances, methods according to the disclosure preferably do not include rinsing steps between monomer depositions. Resulting polyamide films have tunable thicknesses and exhibit high salt rejection, boron rejection, and water permeance. The resulting polyamide films, which exhibit inherently low surface roughness values, also exhibit greatly reduced inorganic scaling and organic fouling, as well as high crosslink densities. The high crosslink density of the resulting films enables enhanced boron rejection without the need for pH adjustment. Various aspects of the disclosure also provide systems for implementing the methods, including deposition mechanisms for the monomer solutions and automated deposition processes to generate the polyamide membrane without intermediate rinsing steps.

[0012] In some instances, methods according to various aspects of the disclosure may include one or more of the following features. The first and second monomer solutions may have a concentration of about 0.01% by mass fraction. The method may further include transferring the polyamide film to a porous support. The porous support may be a polyacrylonitrile (PAN) or porous Matrimid® support. Two methods can be used for fabricating the polyamide membranes. The first method may further include depositing a release layer on the substrate prior to depositing the first monomer solution, performing rinse-free molecular layer by layer deposition, and then transferring the polyamide film to a porous support to obtain a RO or NF membrane. The release layer may comprise a polyelectrolyte, such as, e.g., poly(sodium 4-styrenesulfonate) (PSS). The second method may include depositing a layer of polyelectrolytes on the porous support, and performing rinse-free molecular layer-by-layer deposition directly on the porous support. This method eliminates the need for polyamide film transfer.

[0013] Various aspects of the present disclosure are also directed to polyamide reverse osmosis membranes, polyamide reverse osmosis membranes according to various aspects of the disclosure include a polyamide film formed by depositing alternating layers of trimesoyl chloride (TMC) and m-phenylenediamine (MPD) without intermediate rinsing steps. The TMC and MPD can be deposited from solutions having concentrations ranging from about 0.01% to about 1% by mass fraction.

[0014] Various aspects of the present disclosure are also directed to polyamide nanofiltration membranes. Polyamide nanofiltration membranes according to various aspects of the disclosure include a polyamide film formed by depositing alternating layers of trimesoyl chloride (TMC) and3130761.00568 / 157813095v.1piperazine (PIP) without intermediate rinsing steps. The TMC and PIP can be deposited from solutions having concentrations ranging from about 0.01% to about 1% by mass fraction.

[0015] Membranes according to various aspects of the disclosure exhibit various properties. For example, membranes according to the disclosure exhibit salt rejections of at least about 99% for a 1000 mg / L NaCl solution. Also for example, membranes according to the disclosure exhibit water permeances of at least about 0.5 L m‘2h-1bar'1. In some instances, membranes according to the disclosure may be free-standing membranes. In some instances, membranes according to the disclosures may further include a porous support, wherein the polyamide fdm is disposed on the porous support. In some instances, the porous support can be, for example, a polyacrylonitrile (PAN) support or a polysulfone support.

[0016] Various aspects of the present disclosure are also directed to systems for fabricating a polyamide reverse osmosis membrane, comprising a substrate; a first deposition mechanism configured to deposit a trimesoyl chloride (TMC) solution onto the substrate; a second deposition mechanism configured to deposit an m-phenylenediamine (MPD) solution onto the substrate; and a controller programmed to control the first and second deposition mechanisms to repeatedly deposit the TMC and MPD solutions without intermediate rinsing steps, wherein the TMC and MPD solutions have concentrations of about 0.01% to about 0.1% by mass fraction. In some instances, the first and second deposition mechanisms comprise a spin coater configured to distribute the TMC and MPD solutions across the substrate. In some instances, the first and second deposition mechanisms comprise programmable syringe pumps configured to dispense solutions of varying volumes onto the substrate. In some instances, the controller is further programmed to control spin speeds of the spin coater based on the concentrations of the TMC and MPD solutions. In some instances, the system further comprises a transfer mechanism configured to transfer the polyamide film from the substrate to a porous support. In some instances, the system further comprises a release layer deposition mechanism configured to deposit a release layer, such as a poly(sodium 4-styrenesulfonate) (PSS) release layer, on the substrate prior to deposition of the TMC and MPD solutions.

[0017] Various non-limiting aspects of the present disclosure are as follows.

[0018] In some instances, a first aspect of the disclosure can be described as a method for fabricating a polyamide reverse osmosis and nanofiltration membrane, comprising depositing a first monomer solution comprising or consisting essentially of trimesoyl chloride (TMC) onto a4130761.00568 / 157813095v.1substrate; depositing a second monomer solution comprising or consisting essentially of m-phenylenediamine (MPD) and / or piperazine (PIP) onto the substrate; and repeating the first and second depositing steps one or more times to form a polyamide film, wherein the first and second monomer solutions have concentrations ranging from about 0.01% to about 0.1% by mass fraction, and wherein the method does not include rinsing steps between the monomer depositing steps.

[0019] In some instances, a second aspect of the disclosure can be described as a method according to the first aspect, wherein the first and second monomer solutions have concentrations of about 0.01% by mass fraction.

[0020] In some instances, a third aspect of the disclosure can be described as a method according to the first or second aspect, further comprising transferring the polyamide film to a porous support.

[0021] In some instances, a fourth aspect of the disclosure can be described as a method according to the third aspect, wherein the porous support is a polyacrylonitrile (PAN) support or a polysulfone support.

[0022] In some instances, a fifth aspect of the disclosure can be described as a method according to any one of the first through fourth aspects, further comprising depositing a release layer on the substrate prior to depositing the first monomer solution.

[0023] In some instances, a sixth aspect of the disclosure can be described as a method according to the fifth aspect, wherein the release layer comprises, consists essentially of, or consists of poly(sodium 4-styrenesulfonate) (PSS).

[0024] In some instances, a seventh aspect of the disclosure can be described as a method according to any one of the first through sixth aspects, wherein the polyamide film has a thickness ranging from about 10 to about 200 nm.

[0025] In some instances, an eighth aspect of the disclosure can be described as a polyamide reverse osmosis membrane comprising a polyamide film formed by depositing alternating layers of trimesoyl chloride (TMC) and at least one of m-phenylenediamine (MPD) and piperazine (PIP) without intermediate rinsing steps.

[0026] In some instances, a ninth aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to the eighth aspect, wherein the TMC and at least one of MPD and PIP are deposited from solutions having TMC and MPD and / or PIP concentrations of about 0.01% to about 1% by mass fraction.5130761.00568 / 157813095v.1

[0027] Tn some instances, a tenth aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to the eighth or ninth aspect, wherein the polyamide film comprises a surface functionalized with terminating functional groups.

[0028] In some instances, an eleventh aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to the tenth aspect, wherein the terminating functional groups impart the polyamide film surface with antifouling properties.

[0029] In some instances, a twelfth aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to the tenth or eleventh aspect, wherein the terminating functional groups impart the polyamide film surface with antiscaling properties.

[0030] In some instances, a thirteenth aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to any one of the tenth through twelfth aspects, wherein the terminating functional groups impart the polyamide film surface with enhanced solute rejection properties as compared to corresponding polyamide films without the surface functionalization.

[0031] In some instances, a fourteenth aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to any one of the tenth through thirteenth aspects, wherein the terminating functional groups are selected from carboxyls or functional groups comprising one or more carboxyl groups, amines or functional groups comprising one or more amines, polyethylene glycols or functional groups comprising one or more polyethylene glycol groups, fluorocarbons or functional groups comprising one or more fluorocarbon groups, and zwitterionic groups functional groups comprising one or more zwitterionic groups.

[0032] In some instances, a fifteenth aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to any one of the eighth through fourteenth aspects, wherein the polyamide film has a thickness of about 10 to about 200 nm.

[0033] In some instances, a sixteenth aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to any one of the eighth through fifteenth aspects, wherein the membrane has a salt rejection of at least about 99% for a 1000 mg / L NaCl solution.

[0034] In some instances, a seventeenth aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to any one of the eighth through sixteenth aspects, wherein the membrane has a water permeance of at least about 0.5 L nr2h'1bar1.

[0035] In some instances, an eighteenth aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to any one of the eighth through seventeenth6130761.00568 / 157813095v.1aspects, wherein the membrane exhibits a steady-state salt rejection or a steady-state water permeance within about 4 hours of operation.

[0036] In some instances, a nineteenth aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to any one of the eighth through seventeenth aspects, wherein the membrane exhibits a steady-state salt rejection and a steady-state water permeance within about 4 hours of operation.

[0037] In some instances, a twentieth aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to any one of the eighth through nineteenth aspects, wherein the polyamide film exhibits a root-mean-square (RMS) surface roughness of less than about 10 nm.

[0038] In some instances, a twenty-first aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to any one of the eighth through twentieth aspects, further comprising a porous support, wherein the polyamide film is disposed on the porous support.

[0039] In some instances, a twenty-second aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to any one of the eighth through twenty-first aspects, wherein the membrane exhibits reversible silica scaling, the reversible silica scaling characterized by recovery of at least about 90% of initial water permeance following exposure to a silica-containing feed solution and subsequent rinsing with deionized water.

[0040] In some instances, a twenty-third aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to any one of the eighth through twenty-second aspects, wherein the membrane exhibits resistance to organic fouling, the resistance to organic fouling characterized by at least about 90% recovery of water permeance after exposure to organic foulants and subsequent cleaning with an aqueous solution.

[0041] In some instances, a twenty-fourth aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to any one of the eighth through twenty-third aspects, wherein the membrane exhibits a boron rejection of at least about 90% at a feed solution having a pH ranging from 7 to 10.

[0042] In some instances, a twenty-fifth aspect of the disclosure can be described as a system for fabricating a polyamide reverse osmosis membrane, the system comprising a substrate, a first deposition mechanism configured to deposit a trimesoyl chloride (TMC) solution onto the7130761.00568 / 157813095v.1substrate, a second deposition mechanism configured to deposit an m-phenylenediamine (MPD) and / or a piperazine (PIP) solution onto the substrate, and a controller programmed to control the first and second deposition mechanisms to repeatedly deposit the TMC solution and the MPD and / or PIP solutions without intermediate rinsing steps.

[0043] In some instances, a twenty-sixth aspect of the disclosure can be described as a system according to the twenty-fifth aspect, wherein the TMC solution and the MPD and / or PIP solutions have concentrations of about 0.01% to about 0.1% by mass fraction.

[0044] In some instances, a twenty-seventh aspect of the disclosure can be described as a system according to the twenty-fifth or twenty-sixth aspect, wherein the first deposition mechanism comprises one or more spin coaters configured to distribute the TMC solution across the substrate.

[0045] In some instances, a twenty-eighth aspect of the disclosure can be described as system according to the twenty-seventh aspect, wherein the controller is further programmed to control the spin speeds of the one or more spin coaters of the first deposition mechanism based on the concentration of the TMC solution.

[0046] In some instances, a twenty-ninth aspect of the disclosure can be described as a system according to any one of the twenty-fifth through twenty-eighth aspects, wherein the second deposition mechanism comprises one or more spin coaters configured to distribute the MPD solution and / or PIP solution across the substrate.

[0047] In some instances, a thirtieth aspect of the disclosure can be described as a system according to the twenty-ninth aspect, wherein the controller is further programmed to control the spin speeds of the one or more spin coaters of the second deposition mechanism based on the concentration of the MPD and / or PIP solutions.

[0048] In some instances, a thirty-first aspect of the disclosure can be described as a system according to any one of the twenty-fifth through thirtieth aspects, further comprising a transfer mechanism configured to transfer the polyamide film from the substrate to a porous support.

[0049] In some instances, a thirty-second aspect of the disclosure can be described as a system according to the thirty -first aspect, wherein the porous support is a polyacrylonitrile (PAN) support or a polysulfone support.8130761.00568 / 157813095v.1

[0050] Tn some instances, a thirty-third aspect of the disclosure can be described as a\ system according to the thirty-first or thirty-second aspect, wherein the controller is further programmed to control the transfer mechanism.

[0051] In some instances, a thirty -fourth aspect of the disclosure can be described as a system according to any one of the twenty-fifth through thirty-third aspects, further comprising a release layer deposition mechanism configured to deposit a release layer on the substrate prior to deposition of the TMC solution and the MPD or PIP solutions.

[0052] In some instances, a thirty-fifth aspect of the disclosure can be described as a system according to the thirty-fourth aspect, wherein the release layer deposition mechanism is configured to deposit a poly(sodium 4-styrenesulfonate) (PSS) release layer.

[0053] In some instances, a thirty-sixth aspect of the disclosure can be described as a system according to the thirty-fourth aspect or thirty-fifth aspect, wherein the controller is further programmed to control the release layer deposition mechanism.

[0054] In some instances, a thirty-seventh aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to any one of the eighth through twenty-fourth aspects prepared by a method according any of one the first through seventh aspects.

[0055] In some instances, a thirty-eighth aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to any one of the eighth through twenty-fourth aspects prepared by a system according any of one the twenty-fifth though thirty-sixth aspects.

[0056] In some instances, a thirty-eighth aspect of the disclosure can be described as a polyamide reverse osmosis membrane according to any one of the eighth through twenty-fourth aspects prepared by a method according any of one the first through seventh aspects wherein the method is performed, at least in part, by a system according any of one the twenty-fifth though thirty-sixth aspects.DETAILED DESCRIPTION

[0057] FIG. 1 is a schematic illustration of a modified molecular layer-by-layer (mLbL) deposition technique according to various aspects of the disclosure for the fabrication of polyamide membranes.

[0058] FIG. 2 is a schematic illustration of a floating film transfer process, for transferring a polyamide membrane from a substrate to a porous support.9130761.00568 / 157813095v.1

[0059] FIG. 3 is a schematic illustration of an exemplary process for reacting a polyamide membrane with various compounds to functionalize the membrane surface with terminating functional groups.

[0060] FIG. 4 provides graphical displays of Zeta potentials of a PAN50 support, an as-cast mLbL, a TMC-terminated mLbL, and a commercial brackish water RO membrane measured as a function of pH (left) and static water contact angle measurements of the same membranes at pH 7 and pH 10 (right). The error bars represent one standard deviation (n = 3).

[0061] FIG. 5 provides three-dimensional AFM images of polyamide films according to the disclosure on silicon wafers. The as-cast mLbL film (left) and the film reacted with TMC (right) have root-mean-square (RMS) roughness of (1.1 ± 0.05) nm, and (1.7 ± 0.05) nm, respectively. The error bars represent one standard deviation (n=3).

[0062] FIG. 6 provides FTIR spectra of a commercial membrane (top left), an as-cast membrane according to the disclosure (top middle), and a TMC-functionalized membrane according to the disclosure (top right) showing characteristic amide I (1718 cm ') and amide II (1546 cm ') peaks under different chlorination conditions, and graphical displays of elemental ratios O / N (bottom left), C / N (bottom middle) and C / O (bottom right) of the commercial membrane, an as-cast membrane and TMC-functionalized membrane derived from XPS analyses.

[0063] FIG. 7 provides graphical displays comparing normalized water permeance and normalized salt rejection of commercial, and rinse-free as-cast (according to the disclosure) and TMC functionalized mLbL(according to the disclosure) polyamide membranes after chlorine exposure at pH 4.0 (left), pH 7.0 (middle), and pH 10.0 (Right). Chlorination dose is expressed as ppm h. The error bars represent one standard deviation (n = 3).

[0064] FIG. 8 provides graphical displays comparing water flux decline of commercial, as-cast (according to the disclosure), and TMC-functionalized rinse-free mLbL (according to the disclosure) membranes during fouling tests with sodium alginate (SA) (left) and bovine serum albumin (BSA) (right) feed solutions at pH 7. The error bars represent one standard deviation (n =3)-

[0065] FIG. 9 provides graphical displays comparing water flux recovery following static soaking cleaning of commercial, as-cast (according to the disclosure), and TMC (according to the disclosure) membranes after fouling with sodium alginate (SA) (left) and bovine serum albumin (BSA) (right). Three cleaning conditions were employed for 24 hours: deionized water (DI),10130761.00568 / 157813095v.1alkaline solution (pH 10), and alkaline solution with 250 ppm NaOCl The error bars represent one standard deviation (n=3).

[0066] FIG. 10 provides graphical displays comparing water flux of commercial, as-cast, and TMC-functionalized membranes after fouling with sodium alginate (SA) or bovine serum albumin (BSA) in the presence of 0 mM Ca2+, 1 mM Ca2+, and 2 mM Ca2+(left), and zeta potential of SA and BSA solutions at the same Ca2+concentrations, showing charge neutralization with increasing Ca2+that corresponds to the observed water flux decline (right). The error bars represent one standard deviation (n = 3).

[0067] FIG. 11 is a graphical display comparing the flux recovery (%) of commercial, as-cast(according to the disclosure), and TMC-functionalized(according to the disclosure) membranes following SA and BSA fouling under three cleaning conditions: (a) DI water, (b) alkaline solution at pH 10, and (c) alkaline solution (pH 10) containing 250 ppm NaOCl. Results are shown for fouling in feeds with 0 mM Ca2, 1 mM Ca2+, and 2 mM Ca2+. The error bars represent one standard deviation (n=3).

[0068] FIG. 12 provides graphical displays comparing AFM force-separation curves (adhesive force) (top) and corresponding interaction energy distributions (bottom) for the absence of BSA and Ca2+(top and bottom left), presence of BSA without Ca2+(top and bottom middle), and presence of BSA with Ca2+(top and bottom right).

[0069] FIG. 13 provides schematic illustrations of amino-silane-treated QCM-D sensors prepared for mLbL processes according to the disclosure, b) sequential deposition of trimesoyl chloride (TMC) and m-phenylenediamine (MPD) to form polyamide films according to the disclosure, followed by surface functionalization, and (c) subsequent silica scaling experiments.

[0070] FIG. 14 is a graphical display comparing normalized permeance and flux recovery after DI water rinsing for as-cast rinse-free mLbL according to the disclosure and commercial (DuPont brackish water BWXLE30) polyamide membranes exposed to silica scaling.

[0071] FIG. 15 provides graphical displays comparing XDLVO analyses, which reveals electrostatic interaction as the dominant mechanism through which silica particles in the form of negatively charged oligomers interact with the surface to cause scaling. PEG-amine and TMC surfaces demonstrate superior anti-scaling properties due to hydration layer formation and electrostatic repulsion, respectively11130761.00568 / 157813095v.1

[0072] FIG. 16 provide three-dimensional AFM images of polyamide films according to the disclosure on silicon wafers. The as-cast mLbL film and the films reacted with TMC, MPD, PEG-NH2, and perfluorooctyl amine have root-mean-square (RMS) roughness of (1.1 ± 0.1) nm, (1.7 ± 0.1) nm, (2.4 ± 0.1) nm, (3.3 ± 0.2) nm, and (4.4 ± 1.1) nm, respectively.

[0073] FIG. 17 provides graphical displays comparing Zeta potential of mLbL polyamide films (as-cast and reacted with TMC, MPD, PEG-NFb, and perfluorooctyl amine) measured at pH 6.8 using 0.1 mM KC1 solution (left) and static contact angles of DI water, glycerol, and diiodomethane on mLbL polyamide films (as-cast and reacted with TMC, MPD, PEG-NH2, and perfluorooctyl amine) (right). Error bars represent one standard deviation from three measurements.

[0074] FIG. 18 provides graphical displays comparing desalination performance of rinse-free mLbL polyamide membranes with different surface chemistries, specifically water permeance as a function of time over a 4 h filtration experiment using a 1 g / L NaCl solution at 30 bar (left) and steady-state water permeance and NaCl rej ection for each membrane surface (right). The error bars represent one standard deviation.

[0075] FIG. 19 provides graphical displays comparing silica scaling performance of rinse-free mLbL polyamide membranes with different surface chemistries, specifically water permeance over time, showing the initial desalination run, followed by two consecutive silica scaling cycles, each lasting three hours, with a 10 min DI water rinse in between (top left), normalized permeance after the first silica scaling cycle and corresponding flux recovery following DI water rinsing (top right), Real-time silica deposition over 24 h as measured by QCM-D (bottom left) and correlation between water flux decline after the second silica scaling cycle and total silica deposition (bottom right). The error bars represent one standard deviation.

[0076] FIG. 20 provides graphical displays showing water flux decline as a function of hydration energy (left) and wetting function f(0) (right). The error bars represent one standard deviation.

[0077] FIG. 21 is a graphical display showing seta potential of 11.2 mM silicic acid solution measured in a QCMD feed module over 0, 12, and 24 h.

[0078] FIG. 22 provides graphical displays showing relationships between water flux decline and zeta potential (left), electrostatic interaction energy (Gei / KbT) (middle), and total interaction energy (Gtot / KbT) (right) for rinse-free mLbL polyamide membranes according to the disclosure.12130761.00568 / 157813095v.1

[0079] FIG. 23 provides graphical displays comparing QCM-D silica scaling deposition on polyamide membranes with different terminal surface chemistries over 24 h. Each panel represents independent replicate according to the disclosure runs. For PFA surfaces, only three runs are shown here, as the remaining replicates exhibited no detectable silica deposition even after achieving a stable baseline.DETAILED DESCRIPTION

[0080] The following description of the embodiments is merely exemplary in nature and is in no way intended to limit the subject matter of the present disclosure, their application, or uses.

[0081] As used throughout, ranges are used as shorthand for describing each and every value that is within the range. Any value within the range can be selected as the terminus of the range. Unless otherwise specified, all percentages and amounts expressed herein and elsewhere in the specification should be understood to refer to percentages by weight.

[0082] For the purposes of this specification and appended claims, unless otherwise indicated, all numbers expressing quantities, percentages or proportions, and other numerical values used in the specification and claims, are to be understood as being modified in all instances by the term “about.” The use of the term “about” applies to all numeric values, whether or not explicitly indicated. This term generally refers to a range of numbers that one of ordinary skill in the art would consider as a reasonable amount of deviation to the recited numeric values (i.e., having the equivalent function or result). For example, this term can be construed as including a deviation of ±10 percent, alternatively ±5 percent, alternatively ±1 percent, alternatively ±0.5 percent, and alternatively ±0.1 percent of the given numeric value provided such a deviation does not alter the end function or result of the value. Accordingly, unless indicated to the contrary, the numerical parameters set forth in this specification and attached claims are approximations that can vary depending upon the desired properties sought to be obtained by the present invention.

[0083] As used in this specification and the appended claims, the singular forms “a,” “an,” and “the,” include plural references unless expressly and unequivocally limited to one referent. As used herein, the term “include” and its grammatical variants are intended to be non-limiting, such that recitation of items in a list is not to the exclusion of other like items that can be substituted or added to the listed items. For example, as used in this specification and the following claims, the terms “comprise” (as well as forms, derivatives, or variations thereof, such as “comprising” and “comprises”), “include” (as well as forms, derivatives, or variations thereof, such as “including”13130761.00568 / 157813095v.1and “includes”) and “has” (as well as forms, derivatives, or variations thereof, such as “having” and “have”) are inclusive (i.e., open-ended) and do not exclude additional elements or steps. Accordingly, these terms are intended to not only cover the recited element(s) or step(s), but may also include other elements or steps not expressly recited. Furthermore, as used herein, the use of the terms “a” or “an” when used in conjunction with an element may mean “one,” but it is also consistent with the meaning of “one or more,” “at least one,” and “one or more than one.” Therefore, an element preceded by “a” or “an” does not, without more constraints, preclude the existence of additional identical elements.

[0084] The present disclosure relates to methods for fabricating polyamide membranes using modified molecular layer-by-layer (mLbL) deposition techniques. In some instances, modified mLbL deposition techniques according to various aspects of the disclosure utilize and reduced number of rinse steps between monomer depositions as compared to the number of rinse steps required in conventional molecular layer-by-layer deposition techniques. For example, in some instances, modified mLbL deposition techniques according to various aspects of the disclosure provide for the ability to conduct two or more (such as three or more, or four or more) monomer depositions before any rinse step, after which additional monomer depositions may then resume. In some instances, modified mLbL deposition techniques according to various aspects of the disclosure completely eliminates the need for rinse steps between monomer depositions.

[0085] Conventional mLbL processes typically involve sequential deposition of monomer solutions followed by rinse steps to remove excess unreacted monomers. These rinse steps may consume significant amounts of time and resources during membrane fabrication. Unlike conventional mLbL processes, modified mLbL deposition techniques according to the disclosure allow for controlled polyamide film growth without intermediate rinsing.

[0086] In some instances, modified mLbL deposition techniques according to various aspects of the disclosure allow for careful tuning and control of reagent (for example, reactive monomer solutions) volumes and concentrations, as well as deposition parameters. The ability to provide such control in fabrication allows for, for example, minimization of excess monomer delivery to the growing film surface, which enables the minimization or complete elimination of rinse steps while still producing polyamide membranes with desirable properties such as controlled thickness, low surface roughness, and high crosslink density. Eliminating rinse steps may significantly reduce membrane fabrication time and improve reagent efficiency compared to conventional mLbL14130761.00568 / 157813095v.1methods. Modified mLbL deposition techniques according to various aspects of the disclosure therefore offer a path towards more scalable production of polyamide-based membranes for applications such as reverse osmosis desalination and water purification.

[0087] Modified mLbL deposition techniques according to various aspects of the disclosure utilize monomer solutions prepared with specific concentrations to enable controlled polyamide film growth without intermediate rinsing steps. In some instances, the monomer solutions may comprise, but are not limited to, trimesoyl chloride (TMC), m-phenylenediamine (MPD), and piperazine (PIP), or any combination thereof, dissolved in suitable solvents. Various types of organic solvents or organic co-solvent systems may be used. In some instances, suitable solvents include, but are not limited to, non-polar organic solvents such as toluene, polar organic solvents such as tetrahydrofuran (THF), isoparaffmic hydrocarbon solvents, or any combination thereof. Generally, any solvents that are compatible with the monomers and facilitate controlled deposition of the monomers may be used.

[0088] The concentration of monomers in the solutions may be carefully tuned to minimize excess monomer delivery to the growing fdm surface. In some instances, monomer solutions with monomer mass fractions (WB) ranging from about 0.01% to about 1%, and any value therebetween, may be prepared and used in the fabrication of a polyamide membrane. The monomer mass fraction (WB) of a monomer solution may be calculated as the mass of the monomer divided by the total mass of the solution as a dimensionless ratio or as a percentage (by multiplying 100%).

[0089] In some cases, the TMC and MPD (and / or PIP) solutions may be prepared at identical concentrations for a given deposition process. For example, both the TMC and MPD solutions may be prepared at WB of about 0.01% to about 0.05% in a solvent, such as toluene, for use in a particular mLbL deposition sequence. In some instances, the use of such low monomer concentration solutions may better allow for more precise control over the amount of monomer delivered to the surface during each deposition step, further enabling the elimination of rinse steps between monomer depositions while still achieving controlled polyamide fdm growth. Solutions with WB values of about 0.05% and about 0.1% may provide intermediate levels of control over monomer delivery and fdm formation. In some instances, monomer solutions prepared at higher concentrations such as WB ranging from about 0.4% to about 1% may result in excessive monomer deposition and uncontrolled fdm growth when used without intermediate rinsing. Such high concentration solutions, however, may still be useful without intermediate rinsing, however, by15130761.00568 / 157813095v.1modifying other fabrication parameters, however, such as reducing the amount (volume) and / or rate of monomer solution deposition in individual deposition steps.

[0090] By systematically evaluating different monomer concentrations, individual parameters of modified mLbL deposition techniques according to various aspects of the disclosure may be varied to achieve different membranes with certain properties which may be desirable for different end uses of application said membranes, such as controlled thickness, low surface roughness, and high crosslink density, without the need for rinse steps between depositions during fabrication of said membranes.

[0091] Modified mLbL deposition techniques according to various aspects of the disclosure may utilize various substrates for polyamide mLbL film formation and deposition thereof. Exemplary substrates include, but are not limited to silicon wafers with native oxides, polyacrylonitrile (PAN) substrates, polysulfone substrates, and aluminum oxide substrates. In some instances, the use silicon wafers with native oxide as substrates may be preferable for depositing mLbL films.

[0092] For experiments involving a final rinse with water, such as those designed to test for the presence or removal of unreacted monomers, surfaces of the substrates may be treated, prior to mLbL deposition, with a reagent to provide the substrate surface with functional groups that provide for covalent bonding between the mLbL film and the substrate surface. Such covalent bonding may enhance adhesion of the mLbL film to the substrate surface as well as enhance the stability of the mLbL film. In some instances, the use of an aminosilane as the reagent may be preferred.

[0093] The substrate preparation process may be tailored to accommodate different characterization techniques and experimental requirements. For example, silicon wafers may be used for thickness measurements and surface characterization, while gold-coated substrates may be preferred for spectroscopic analysis. The choice of substrate and its preparation may influence the properties and behavior of the deposited mLbL films.

[0094] Modified mLbL deposition techniques according to various aspects of the disclosure may be carried out using customizable environmental spin coaters. In some cases, the spin coaters may be purged with clean, dry air to reduce exposure to water vapor prior to and / or during the deposition process.16130761.00568 / 157813095v.1

[0095] Monomer solutions may be loaded into gas-tight syringes for deposition. In some instances, the gas-tight syringes may be outfitted with in-line filters such as, for example 0.45 pm polytetrafluoroethylene (PTFE) syringe filter. Such filters may help remove any particulates or contaminants from the monomer solutions prior to deposition on the substrate surface.

[0096] The deposition process may involve sequentially applying the trimesoyl chloride (TMC) and m-phenylenediamine (or piperazine) monomer solutions to the substrate surface. In some instances, a volume of about 0.01 to about 0.2 mb of each monomer solution may be deposited during each half-cycle of the process. The volume of monomer solution(s) deposited during any deposition step may be varied based target properties of the final polyamide membrane.

[0097] For example, a single deposition cycle may comprise the following steps:

[0098] 1. Deposition of TMC solution (for example, 0.06 mL) onto the substrate;

[0099] 2. Spinning of the substrate to distribute the TMC solution on the substrate surface;

[0100] 3. Deposition of MPD (and / or PIP) solution (for example, 0.06 mL) onto the TMC solution-deposited substrate surface; and

[0101] 4. Spinning of the substrate to distribute the MPD (or PIP) solution on the TMC solution-deposited substrate surface.

[0102] This cycle may be repeated multiple times to build up the desired polyamide film thickness. The number of cycles may be varied depending on the target membrane thickness and the monomer solution concentrations used.

[0103] In some instances, the deposition process may be carried out without any intermediate rinsing steps between monomer applications. By carefully controlling the monomer solution concentrations and deposition parameters, excess unreacted monomers may be minimized, potentially eliminating the need for rinsing steps.

[0104] The spin speed and duration used during the deposition process may be optimized to achieve uniform monomer distribution and controlled film growth. In some cases, different spin parameters may be used for the TMC and MPD (or PIP) deposition steps to account for differences in solution properties.

[0105] The time between successive monomer depositions may also be controlled to allow for sufficient reaction time while maintaining an efficient overall process. In some cases, the interval between TMC and MPD (or PIP) depositions may be on the order of seconds to minutes.17130761.00568 / 157813095v.1

[0106] By systematically varying parameters such as monomer concentration, deposition volume, spin speed, and cycle count, the mLbL process may be optimized to achieve desired membrane properties without the need for intermediate rinsing steps. This modified approach significantly reduces overall fabrication time and chemical consumption compared to conventional mLbL methods that incorporate rinsing between each monomer deposition.

[0107] FIG. 1 is a schematic illustration of an exemplary modified molecular layer-by-layer (mLbL) deposition process 100 according to various aspects of the disclosure for the fabrication of polyamide membranes. In the process 100 of FIG. 1, a series of steps are described. Although specific steps are described in the method 100, in some instances, processes according to the disclosure for the fabrication of polyamide membranes may have more or less steps than the described steps without departing from the scope of the disclosure. According to various aspects of the disclosure, the process 100 can start at step 110.

[0108] In step 110, in preparation for the formation of a polyamide membrane thereon, a substrate is washed with one or more organic solvents, such as ethanol and toluene, to remove impurities that may be present on the substrate, the presence of which may be detrimental to the mLbL deposition process. Once step 110 is complete, the process may proceed to step 120.

[0109] In step 120, the washed substrate is subjected to a nitrogen gas purge to remove any excess solvent, from step 110, remaining on the substrate. Once step 120 is complete, the process may proceed to step 130.

[0110] In step 130, the substrate from 120 is subjected to UV-ozone irradiation for further cleaning of the substrate surface. Once step 130 is complete, the process may proceed to step 140. In some instances, step 130 may be omitted and the process may proceed to step 140 from step 120 instead.

[0111] In step 140, aqueous poly(styrene sulfonate) (PSS) solution is deposited on the substrate surface and dried to form a PSS release layer on the substrate surface. Once step 140 is complete, the process may proceed to step 160.

[0112] In step 150, multiple deposition cycles of sequential application the trimesoyl chloride (TMC) and m-phenylenediamine (or piperazine) monomer solutions to the substrate surface is conducted as described elsewhere herein to form a polyamide fdm layer of the PSS release layer. Once step 150 is complete, the process may proceed to step 160.

[0113] In step 160, a polyamide film applied on a substrate is retrieved.18130761.00568 / 157813095v.1

[0114] FIG. 2 is a schematic illustration of a floating film transfer process 200 for transferring a polyamide film from a substrate (for example, as obtained in step 160 of FIG. 1), to a porous support. In the process, a substrate 210, having a polyamide membrane 230, is inserted into a deionized water bath. During insertion, the polyamide membrane 230 releases from the substrate 210 and is transferred to the surface of a porous substrate 260. In some instances, the porous substrate 260 may itself be deposited on a different substrate 250 such as, for example, a fabric or textile.

[0115] FIG. 3 is a schematic illustration of an exemplary process for reacting a polyamide membrane with various compounds to functionalize the membrane surface with terminating functional groups. Various compounds may be used. In some instances, it may be preferable to use compounds that have functional groups that provide the polyamide membrane terminating functional groups such as carboxyls or functional groups comprising one or more carboxyl groups, amines or functional groups comprising one or more amines, polyethylene glycols or functional groups comprising one or more polyethylene glycol groups, fluorocarbons or functional groups comprising one or more fluorocarbon groups, or zwitterionic groups functional groups comprising one or more zwitterionic groups. The addition of terminating functional groups on a polyamide membrane surface may impart one or more of antifouling properties, antiscaling properties, and / or enhancement of solute rejection properties of the polyamide membrane.

[0116] The modified molecular layer-by-layer (mLbL) polyamide films may be characterized using various analytical techniques to assess properties such as thickness, chemical composition, and surface morphology.

[0117] By systematically characterizing films prepared under different deposition conditions, relationships between processing parameters and film properties may be established. Such systematic characterization enables optimization of the modified mLbL process to achieve desired membrane characteristics for specific applications.

[0118] The modified molecular layer-by-layer (mLbL) polyamide membranes may be evaluated for desalination performance using various testing methods and apparatus. In some cases, the mLbL films may be transferred from their original deposition substrate to a porous support material prior to performance testing.

[0119] In some cases, the mLbL polyamide films may be removed from the substrate and transferred to the surface of a porous support such as, for example, macroporous polyacrylonitrile19130761.00568 / 157813095v.1(PAN) support or a porous polysulfone support. In some instances, the PAN support may have a pore size of about 20-500 nm, which may facilitate efficient water transport through the membrane. To enable the transfer process, a thin layer of, for example, poly(styrene sulfonate) (PSS) may be deposited onto the original substrate prior to mLbL deposition. The PSS layer may act as a release layer, allowing the mLbL film to be separated from the original substrate and transferred to the PAN support.

[0120] The membrane performance may be evaluated using a stirred dead-end filtration cell or crossflow cell. The cell may be operated at an applied pressure of about 5-80 bar during desalination and water purification tests.

[0121] For desalination performance testing, an aqueous sodium chloride (NaCl) solution may be used as the feed. In some cases, the NaCl concentration in the feed solution may be about 1000 mg / L. This concentration may be selected to represent brackish water or seawater conditions and allow for comparison with other membrane technologies.

[0122] The desalination performance of the membranes may be characterized by measuring key metrics such as salt rejection and water permeance. Salt rejection may be determined by measuring the conductivity of the feed and permeate solutions using a conductivity meter. Water permeance may be calculated based on the volume of permeate collected over time, normalized by the membrane area and applied pressure.

[0123] In some cases, the desalination tests may be conducted over a period of several hours or days to assess the stability and consistency of membrane performance. Multiple samples may be tested to evaluate reproducibility and calculate statistical measures such as standard deviation for the performance metrics.

[0124] The stirred dead-end filtration cell may incorporate a stirring mechanism to minimize concentration polarization effects near the membrane surface during testing. The stirring speed may be optimized to balance between reducing concentration polarization and avoiding excessive shear stress on the membrane surface.

[0125] Temperature control may be implemented during the desalination tests to maintain consistent operating conditions. In some cases, the feed solution temperature may be maintained at a specific value, such as 25°C, using a temperature control system.

[0126] By systematically evaluating membrane performance under controlled conditions, relationships between fabrication parameters and desalination capabilities may be established.20130761.00568 / 157813095v.1This may enable optimization of the modified mLbL process to achieve desired salt rejection and water permeance characteristics for specific water treatment applications.

[0127] The modified molecular layer-by-layer (mLbL) process may involve complex interactions between various elements to produce high-performance polyamide membranes without intermediate rinsing steps. The interplay between monomer concentration, deposition parameters, and resulting membrane properties may be carefully tuned to achieve desired outcomes.

[0128] In some cases, the concentration of monomers in the deposition solutions may play a crucial role in controlling film growth and properties. Lower monomer concentrations, such as about 0.01 to 0.05% by mass fraction, may allow for more precise delivery of monomers to the growing film surface. This controlled delivery may minimize excess unreacted monomers, potentially eliminating the need for rinsing steps between depositions.

[0129] The deposition process parameters may interact with monomer concentration to influence film formation. Factors such as spin speed, deposition volume, and time between successive monomer applications may be optimized based on the monomer concentration used. For example, lower monomer concentrations may require longer spin times or multiple depositions to achieve a desired film thickness.

[0130] In some instances, the absence of rinsing steps may affect the reaction kinetics and crosslinking density of the polyamide film. Without intermediate rinsing, unreacted monomers from previous deposition steps may remain available for further reaction. This may lead to increased crosslinking and potentially alter the membrane's chemical structure and performance characteristics.

[0131] The substrate surface properties may interact with the monomer solutions and influence film growth. Surface treatments such as UV-ozone cleaning or surface functionalization (with, for example, an aminosilane) may affect monomer adsorption and initial film formation. These substrate-monomer interactions may be particularly important in the absence of rinsing steps, as they may influence the uniformity and adhesion of the growing polyamide film.

[0132] In some cases, the sequential deposition of trimesoyl chloride (TMC) and m-phenylenediamine (or piperazine) without rinsing may lead to the formation of a reaction zone at the interface between the two monomers. The thickness and characteristics of this reaction zone21130761.00568 / 157813095v.1may depend on factors such as monomer concentration, diffusion rates, and reaction kinetics. Control over this interfacial region may be important for achieving desired membrane properties.

[0133] The overall film thickness may result from the combined effects of monomer concentration, number of deposition cycles, and absence of rinsing steps. Lower monomer concentrations may lead to slower film growth rates, potentially requiring more deposition cycles to achieve a target thickness. However, the elimination of rinsing steps may allow for more efficient use of deposited monomers, potentially offsetting the slower growth rate.

[0134] In some instances, the interactions between these various elements may influence the final membrane properties such as salt rejection and water permeance. The balance between crosslinking density, film thickness, and surface characteristics may determine the membrane's separation performance. By carefully controlling monomer concentration and deposition parameters, membranes with high salt rejection and adequate water permeance may be achieved without the need for intermediate rinsing steps.

[0135] The modified mLbL process may also involve interactions between the growing polyamide film and the surrounding environment. Factors such as temperature, humidity, and exposure to air during deposition may influence reaction rates and film properties. Control over these environmental factors may be important for achieving consistent and reproducible membrane fabrication without rinsing steps.

[0136] In some instances, the absence of rinsing steps may lead to the incorporation of small amounts of unreacted monomers or oligomers within the polyamide film. These residual species may interact with the crosslinked polymer network and potentially influence membrane properties such as free volume, hydrophilicity, or charge density. Understanding and controlling these interactions may be important for optimizing membrane performance.

[0137] By carefully considering and optimizing the interactions between monomer concentration, deposition process parameters, and environmental factors, high-performance polyamide membranes may be produced using the modified mLbL technique without the need for intermediate rinsing steps. This approach may offer advantages in terms of reduced fabrication time and chemical consumption while maintaining desirable membrane properties for applications such as water desalination.Examples22130761.00568 / 157813095v.1

[0138] Example 1 - Evaluation of molecular layer-by-layer (mfblf) polyamides for resistance towards organic fouling.

[0139] Materials. Trimesoyl chloride (TMC, molecular weight: 265.48 g / mol, purity: >98%) and 3-aminopropyldimethylethoxysilane (APDES, molecular weight: 161.29 g / mol) were purchased from Sigma-Aldrich and Gelest respectively and used as received. The aqueous poly(sodium 4-styrenesulfonate) (PSS) solution (15% by mass in water, monomer molecular weight: 206.20 g / mol, >98% purity) used to form the sacrificial release layer was obtained from Sigma-Aldrich. Toluene and ethanol were purchased from Fisher Scientific. Toluene was dried over 3 A molecular sieves and confirmed to contain less than 10 ppm water using a Karl Fischer CS20X coulometric titrator (Mettler Toledo). Silicon wafers (orientation <100>, University Wafers, 2 cm x 2 cm) were cleaned by sequential rinsing with toluene and ethanol, dried under nitrogen, and treated with ultraviolet-ozone (Model 342, Jelight) for 20 minutes to activate the surface. A sacrificial release layer of PSS was then deposited by spin-coating a 1 wt% aqueous PSS solution at 2000 rpm for 30 seconds.

[0140] SA (>99% purity, Sigma-Aldrich) and BSA (>98% purity, Sigma- Aldrich) were selected as model foulants to represent polysaccharide- and protein-based organic fouling, respectively. Stock solutions of 2 g / L were prepared by dissolving each foulant in deionized (DI) water (18 MQ cm, Barnstead Nanopure) and stirring at room temperature for at least 24 hours to ensure complete dissolution. The solutions were then filtered using 0.45 pm hydrophilic membrane filters (VWR, Radnor, PA) and stored in sterilized glass bottles at 4 °C until use. To study ionbridging, calcium chloride dihydrate (CaCE ^FEO; >99% purity, Sigma- Aldrich) was added at 1 mM concentration, and the ionic strength was adjusted to 10 mM using sodium chloride.

[0141] Fabrication and Functionalization od mLbL Polyamide Film. Polyamide films were prepared using the rinse-free mLbL method. Monomer solutions of 0.02 % by mass fraction of TMC and 0.01 % by mass fraction of MPD were prepared in toluene. Silicon wafers were rinsed with toluene, ethanol, and DI water, dried with nitrogen gas, and exposed to ultraviolet-ozone for 20 minutes, followed by coating with 1% PSS in water (~50 nm) to form a release layer. A polyamide film of approximately 20 nm thickness was then deposited via 25 deposition cycles.

[0142] Functionalization was carried out by immersing the films in a 1% reagent solution in toluene at 75 °C for 24 hours, followed by rinsing with toluene and drying with nitrogen gas. As-cast with no functionalization, underwent the same process but without any reagent. For fouling23130761.00568 / 157813095v.1experiments, the film-floating method was used to transfer the polyamide film onto a porous polyacrylonitrile (PAN50) support following literature protocols.

[0143] The membranes were left to dry overnight to ensure proper adhesion before being trimmed and mounted onto a dead-end filtration cell for performance testing. The PAN50 support was selected for its well-defined pore size and spacing, enabling efficient water transport. Its high porosity reduces the diffusion path for water molecules, enhancing permeance while maintaining selectivity.

[0144] Characterization of mLbL Surfaces. Chemical functionalization of the mLbL films was confirmed via X-ray photoelectron spectroscopy (XPS, AXIS Ultra DLD Spectrometer, Kratos Analytical). The surface charge of the polyamide surfaces was measured via the streaming potential technique using the SurPASS™ 3 electrokinetic analyzer with 0.1 mM KC1 at pH 3-10. Atomic force microscopy (AFM) imaging of the polyamide films was performed using a Multimode 8 AFM (Bruker). Before imaging, the samples were rinsed with ethanol to remove potential contaminants and dried with ultrapure N2. Imaging was conducted with scan sizes of 5 pm, 1 pm, and 200 nm using a ScanAsyst-Air probe (Bruker), and the root-mean-square (rms) roughness of the films was determined at different locations on each sample. The DI water contact angle was measured using a goniometer (rame-hart 260-U4) via the sessile drop method. Before each contact angle measurement, the membranes were rinsed with ethanol and ultrapure water, and then dried with ultrapure N2 gas. A 1-pL droplet of the aqueous solution was placed on the sample using a pipette, allowing the droplet to remain on the surface for a few seconds to 1 minute to establish a stable solution-surface interface, while avoiding complete evaporation. The static contact angle was determined by measuring the tangent angle at the liquid-surface interface after equilibrium was reached. Three contact angle measurements were taken at 3 different points across each surface (diagonally, n=9).

[0145] Fouling and Cleaning Experiments. To evaluate membrane performance, a stirred dead-end fdtration cell (HP4750, Sterlitech) with an active membrane area of 4.16 cm2was used. Three sets of membranes were fabricated, with each set tested in duplicate. A commercial brackish water RO membrane (BWXLE 30, DuPont FilmTec) was included as a performance benchmark.

[0146] All the membranes were compacted by filtering DI water at 20.7 bar (300 Psi) for 5 hours until steady-state flux was achieved. Following compaction, membranes were subjected to chemical exposure to identify a suitable hypochlorite cleaning condition known as “screening”24130761.00568 / 157813095v.1experiments. For this, the membranes were immersed in three separate 250 ppm NaOCl solutions, with pH adjusted to 4, 7, or 10 using 1 M hydrochloric acid (HC1) or sodium hydroxide (NaOH). To evaluate the effect of exposure duration, membranes were soaked for 4 hours and 24 hours at ambient conditions without applied pressure. After each exposure, the membranes were thoroughly rinsed with DI water, remounted in the filtration cell, and tested for water flux.

[0147] Following this, fouling experiments were conducted by feeding solutions containing 20 mg / L of model foulants (SA and BSA) into the stirred dead-end filtration cell and water flux was studied for 5 hours at a controlled pH of 7 and an operating pressure of 30 bar. CaCL solution (1 mM and 2 mM) was introduced in select tests to probe ion-bridging effects. After each fouling phase, the three sets of membranes were processed separately for cleaning: (a) the first set was rinsed with deionized (DI) water, (b) the second set of membranes were cleaned with pH 10, and (c) the third set was cleaned using the optimal sodium hypochlorite (NaOCl) solution which showed negligible water flux loss and highest salt rejection from the screening experiments. The dead-end filtration cells were filled with the respective cleaning solutions, with continuous stirring at 250 RPM maintained throughout the cleaning period under no applied pressure. After the soaking period, membranes were tested using DI water as the feed solution to evaluate water flux loss and recovery over a 2 h period. This duration was selected to allow the membranes to compact and reach steady-state conditions prior to measuring water flux.

[0148] AFM Force Measurements. Force-distance measurements were performed using a carboxylate-modified latex (CML) probe to quantify foul ant-membrane interactions on polyamide surfaces under controlled aqueous conditions. The CML probe, serving as a surrogate for carboxyl-rich organic foulants, was prepared following the same protocol as previously described (i.e., attachment of a 10 pm CML sphere to a cantilever, followed by cleaning, UV-ozone treatment, and spring constant calibration). Prior to measurements, the probe was equilibrated in buffer (1 mM NaCl and 0.2 mM NaHCCF, pH 7.5), and force measurements were conducted in a fluid cell after at least 30 min equilibration with the membrane surface. In this study, the interaction experiments were systematically conducted under three distinct conditions to decouple physicochemical contributions: (i) baseline interaction between the CML probe and the pristine polyamide surface in the absence of foulants and calcium, (ii) interaction in the presence of BSA as the sole model foulant, prepared following the same protocol described above (2 g / L stock,25130761.00568 / 157813095v.1filtered, diluted to 20 mg / L), and (iii) interaction in the presence of both BSA and calcium (2 mM CaCL) to probe ion-bridging effects.

[0149] All measurements were performed using identical AFM parameters (approach / retract velocity: 500 nm / s; dwell time: 1 s; ramp size: 500 nm), and a minimum of 192 force curves were collected at randomly distributed locations across each membrane surface to ensure statistical robustness. The raw deflection-displacement data obtained from AFM were first converted to force-separation curves using the calibrated cantilever spring constant and deflection sensitivity. Contact point determination was performed using a threshold-based algorithm to accurately identify the onset of tip-surface interaction, followed by baseline correction to remove any cantilever drift. From the retract curves, the maximum adhesion force (pull-off force) was determined as the minimum force required to detach the probe from the surface, while the work of adhesion (interaction energy) was calculated by integrating the area under the retract curve (negative force region) using numerical trapezoidal integration.

[0150] Surface Roughness, Wetting Properties and Surface Charge of the Polyamide Membranes. The surface charge and wettability of the substrate, as-cast mLbL, TMC -terminated mLbL, and commercial brackish water RO membranes were evaluated, as shown in FIG. 4, while the corresponding morphologies of the mLbL membranes are presented in FIG. 5. Both mLbL films displayed a smooth morphology with a root-mean-square (RMS) roughness below 2 nm, a result of controlled polymer growth and the stoichiometrically limited nature of the mLbL process that promotes uniform film formation. In contrast, commercial brackish water RO membranes exhibit a much higher surface roughness of 45.6 ± 2.3 nm.23 This pronounced roughness originates from the interfacial polymerization (IP) process, which is diffusion-limited and kinetically uncontrolled. The rapid diffusion of aqueous amine monomers into the organic phase during IP generates localized variations in crosslinking density, producing the characteristic “ridge-and-valley” structures that enhance surface heterogeneity and increase roughness.

[0151] The zeta potential data in FIG.4 (left) offer valuable insight into the surface-charge behavior of the PAN50 support, as-cast mLbL, TMC-terminated mLbL, and commercial RO membranes. The PAN50 support exhibits a negative charge, likely due to the polar nature of the nitrile groups, which preferentially adsorb hydroxide ions from water, resulting in a net negative surface potential. Both the as-cast and commercial membranes show similar pH-dependent trends, reflecting the presence of unreacted carboxylic acid groups (-COOH) that deprotonate to26130761.00568 / 157813095v.1carboxylates (-COO") and residual amine groups (-NH2, pKa~4-l 1) that protonate to -NHv at lower pH.14 In contrast, the TMC-terminated mLbL surface exhibits a substantially more negative zeta potential across the same pH range, attributed to enrichment in deprotonated carboxylates (pKa= 6-9) and the consumption of unreacted amines by the terminating chemistry.

[0152] The contact angle results in FIG. 4 (right) highlight clear differences in surface wettability among the membranes. At both pH 7 and 10, the TMC-terminated mLbL surface exhibits the lowest water contact angle, reflecting its strong hydrophilicity. This behavior is consistent with XPS measurements (FIG. 6), where the TMC surface shows an O / N ratio of 3.21 compared to 1.3 for the as-cast film, confirming a surface enriched in carboxyl groups. The abundance of deprotonated COO enhances interactions with water, thereby lowering the contact angle. On the other hand, the as-cast and commercial membranes also display slightly lower contact angles at pH 10, which can be attributed to the complete deprotonation of carboxylic groups, further enhancing water interaction and surface wettability.

[0153] Surface Functionalization and Chlorine Resistance. To evaluate membrane performance stability, rinse-free mLbL membranes and commercial membranes were exposed to free chlorine at different pH conditions, revealing a strongly pH-dependent response. At low pH (pH 4), where hypochlorous acid (HOC1) predominates, all membranes including commercial, as-cast, and TMC-functionalized exhibited rapid deterioration in NaCl rejection, with losses exceeding 90% (FIG. 7, left), indicating extensive and irreversible degradation of the polyamide selective layer. In contrast, at alkaline pH (—10), where OC1 is the dominant species, all membranes maintained high salt rejection (>95%) over extended exposure (Figure 2c), demonstrating minimal structural damage. At pH 7, moderate performance loss is observed at higher chlorine doses (FIG. 7, middle), suggesting a slower reaction.

[0154] Chlorination of aromatic polyamide membranes proceeds through a well-established sequence. Free chlorine species (HOC1 / OC1 ) first induce N-chlorination of the amide linkage (-CONH- — > -CONC1-), which disrupts intermolecular hydrogen bonding and weakens the polymer network. The N-chlorinated amide is chemically unstable and, particularly in the presence of HOC1, can progress toward amide bond scission, resulting in irreversible chain cleavage and loss of selectivity. Under alkaline conditions, however, N-chlorination is largely reversible and its progression to bond scission is significantly suppressed, consistent with the minimal performance decline observed at pH 10.27130761.00568 / 157813095v.1

[0155] For the commercial membrane, moderate performance loss is observed at higher chlorine doses at pH 7 (FIG. 7, middle), indicating that chlorination progresses beyond reversible N-chlorination and begins to advance toward irreversible amide bond cleavage. FTIR spectra (FIG. 6, top left, top middle and top right) show a decrease in the -1718 cm1band, assigned to-COOH groups, when normalized to the aromatic C=C band (-1612cm suggesting a reduction in acid-containing functionalities relative to the aromatic backbone. Since normalization is performed using the aromatic ring signal, FTIR cannot distinguish the removal of specific structural motifs (e.g., chain ends vs. ring-containing fragments) but instead indicates a relative depletion of oxygen-rich groups.

[0156] XPS provides further insight into these changes. For the commercial membrane, the O / N ratio increases upon chlorination, indicating a relative enrichment of oxygen-containing functionalities compared to nitrogen. This trend suggests that amine-rich chain ends are preferentially removed during chlorination, leading to a relative increase in oxygen content. Consistently, changes in the C / N ratio indicate a reduction in nitrogen relative to carbon, further supporting the pruning of nitrogen-rich (amine-dominated) moieties. Together, these results point toward selective removal of chemically labile, amine-rich regions within the network. This structural modification is sufficient to reduce salt rejection (-60% at high chlorine dose), indicating that bond cleavage has begun to compromise the transport-selective layer. At pH 10, membrane performance remains stable (>95% rejection, FIG. 7, right), indicating that while similar chemical modifications occur, progression to irreversible backbone scission is largely suppressed. FTIR still shows a decrease in the -1718 cm1band and a slight shift of the amide I band (-1680 cm-1), consistent with N-chlorination and reduced hydrogen bonding (N-H — > N-Cl), but these changes do not translate to measurable performance loss.

[0157] In contrast, for the rinse-free mLbL membranes, the response to chlorination reflects their inherently different surface composition compared to the commercial membrane. Notably, the O / N ratios of the mLbL membranes are higher than those of the commercial membrane prior to chlorination, indicating a greater relative abundance of oxygen-containing functionalities, consistent with a higher density of acid-rich chain ends or surface moieties. FTIR trends support this observation, with a decrease in the -1718 cm1band indicating a reduction in acid-containing functionalities relative to the aromatic backbone. Upon chlorination, XPS shows that the O / N ratio decreases, indicating a loss of oxygen-containing functionalities relative to nitrogen. Concurrently,28130761.00568 / 157813095v.1the decrease in C / N indicates a relative enrichment of nitrogen compared to carbon, while the slight decrease in C / O might suggest pruning off chain ends from the removal of acid functionalities. Membrane performance remains largely unchanged at pH 10 (>95% rejection), indicating that these modifications are mostly limited to the surface and do not disrupt the selective layer. At pH 7, some performance loss is observed, suggesting that the same chemical processes extend further into the membrane network and begin to affect transport pathways. The similar FTIR and XPS trends at pH 7 and pH 10 indicate that the same reactions (e.g., N-chlorination and removal of surface groups) occur under both conditions, but to different extents leading to irreversible damage at pH 7 while remaining mostly reversible at pH 10. Based on these findings, 250 ppm chlorine at pH 10 for 24 h was selected as the cleaning condition for antifouling studies, as it induces measurable chemical modification without causing irreversible performance loss.

[0158] Fouling Performance and Reversibility. To assess the fouling behavior of the membranes, a stirred dead-end filtration cell was used, where 20 mg / L of foulant solutions (SA and BSA) were introduced in separate set of experiments. Distinct water flux trends were observed depending on the foulant and membrane type (see FIG. 8). In both tests, the TMC-functionalized rinse-free mLbL membrane consistently maintained higher water flux compared to the as-cast and commercial membranes, while the commercial membrane showed the steepest and most rapid decline, approaching near-complete water flux loss within 4 h. This pronounced fouling tendency can be attributed to the ridge-and-valley morphology characteristic of IP membranes, which greatly increase the effective surface area and create microscale valleys and crevices where foulants readily accumulate, leading to more severe fouling.

[0159] Between the two foulants, BSA resulted in approximately 8-10% greater water flux reduction than SA across all membranes. This stronger fouling tendency is attributed to the denaturation of BSA molecules. Upon contact with the membrane surface, BSA undergoes conformational changes that expose hydrophobic and charged residues, enhancing adhesion and the formation of a dense, compact fouling layer. In contrast, SA fouling arises primarily through the deposition of polysaccharide chains, which form a more open and hydrated gel-like layer that imposes less resistance to water transport than denatured protein aggregates.

[0160] Effect of Contact Time. To evaluate the effect of contact time on fouling removal, membranes were soaked in an alkaline solution (pH 10) for 4 and 24 hours. Increasing the exposure duration enhanced fouling removal. For commercial membranes, extending the soaking time from29130761.00568 / 157813095v.14 to 24 hours resulted in a substantial improvement in permeate recovery, rising from approximately 30 ± 1.5% % to 60 ± 0.2% %. This observation aligns with established findings in the membrane fouling literature, where extended static soaking allows cleaning solutions to infiltrate and degrade compacted foulant layers gradually restoring flux more effectively than shorter contact times, particularly when physical shear is absent and chemical diffusion governs removal effectiveness. In contrast, the rinse-free mLbL membranes demonstrated consistently high cleaning efficiency independent of contact time, achieving around 88 ± 1.1% flux recovery even after only 4 hours of static soaking. This performance advantage can be attributed to their ultrasmooth surface (roughness <2nm) and well-defined surface chemistry, which minimize the strength of foulant attachment and thereby enable easier detachment during cleaning.

[0161] Effect of the Type of Cleaning Solution. The type of cleaning solution also played a critical role in determining the extent of cleaning and fouling reversibility. For commercial membranes, switching from DI water to an alkaline solution at pH 10 resulted in minimal improvement (-10%). However, exposure to an alkaline hypochlorite solution (pH 10, 250 ppm NaOCl) led to a two-fold increase in water flux recovery, underscoring the strong effect of oxidative cleaning in degrading foulant layers. In contrast, both as-cast and TMC-terminated rinse-free mLbL membranes consistently achieved high recovery, regardless of the cleaning solution used.

[0162] Mechanistically, alkaline cleaning enhances foulant removal by modulating charge and increasing solubility. Polyamide membranes contain carboxylic acids (pKa-6-9) that are fully deprotonated at pH 10, increasing negative surface charge (Figure la). Similarly, SA (isoelectric point -3.5) and BSA (isoelectric point -4.7) are both negatively charged at pH 7-10 (see FIG. 9), resulting in electrostatic repulsion that reduces their adhesion to the membrane surface. In addition, for SA, increasing pH to 10 dramatically reduces the viscosity and promotes the dissolution of deposited layers. Similarly, proteins like BSA, when exposed to pH values far from their isoelectric point, become more prone to unfolding or swelling, disrupting compact aggregates and increasing solubilization.

[0163] NaOCl cleaning amplifies these effects through oxidative degradation. At elevated pH levels, NaOCl exists primarily as hypochlorite ion (OCF), which oxidizes organic foulants. For proteinaceous foulants like BSA, NaOCl targets amino acid side chains, leading to protein unfolding, cleavage of disulfide bonds, and partial depolymerization. This chemical degradation30130761.00568 / 157813095v.1results in the breakdown of the compact protein layer into smaller, more water-soluble fragments, making them easier to remove. Polysaccharides like alginate are also susceptible to oxidative modifications, with OCI introducing carbonyl and carboxyl groups that weaken hydrogen bonding and gel structure, facilitating their release from the membrane. The dual benefit of chemical degradation and electrostatic repulsion under alkaline NaOCl cleaning results in more complete removal of complex organic fouling layers (-100% removal), often restoring flux more effectively than alkaline water alone.

[0164] Calcium Induced Ion-Bridging and Reversibility. To investigate the impact of background ions on fouling, Ca2+was added to the feed solution at concentrations of 1 mM and 2 mM. The introduction of Ca2+into the feed markedly altered foulant charge and aggravated fouling across all the polyamide membranes (see FIG. 10). Adding Ca2+caused a reduction in the zeta potential of SA and BSA solution due to the binding of the divalent cations to the carboxylate functional groups, neutralizing their charges. This charge neutralization diminishes the electrostatic repulsion between foulant molecules and the negatively charged polyamide membrane surface, leading to increased foulant aggregation and heavier mass deposition.

[0165] Experimentally, this was reflected in a decline in water flux for all three membranes (commercial, as-cast, and TMC-functionalized) with the severity of the decline increasing as Ca2concentration in the feed rose from 0 to 2 mM. For example, commercial membranes showed the highest water flux drop, around 60% while TMC-functionalized rinse-free mLbL polyamide membranes exhibited the lowest drop in water flux (-20%). Such water flux losses are consistent with literature reports that divalent cations can induce the formation of compact foulant layers that severely impede flow across the membrane matrix. Beyond simple screening, calcium promotes dense cross-linked layers by coordinating simultaneously with carboxyl groups on adjacent foulant chains and with similar groups on the membrane surface. In the case of SA, this leads to the formation of rigid, calcium-induced gels, while for BSA, Ca2+fosters protein aggregation and enhances binding to the polyamide surface. Such ion bridges create foulant structures that are thicker, less porous, and more resistant to shear removal, reflected by rapid water flux losses observed in the presence of Ca2+.

[0166] Calcium-rich fouling significantly alters cleaning efficacy. Compared to foulant-only conditions, the presence of Ca2' led to markedly reduced water flux recovery across all membranes (FIG. 11). Using DI water, commercial membranes recovered less than 50% of their initial water31130761.00568 / 157813095v.1flux, while the as-cast and TMC-functionalized rinse-free mLbL membranes achieved -60% and -80% recovery, respectively. Alkaline cleaning improved performance modestly, with commercial membranes recovering -60%, as-cast membranes -80-85%, and TMC membranes >90%. In contrast, alkaline hypochlorite cleaning proved most effective for all three membranes, with water flux recovery exceeding 90% in every case. Notably, the TMC-functionalized mLbL membranes exhibited near-complete recovery.

[0167] Foulant layers formed without Ca2+are easier to reverse; DI water or mild alkaline cleaning (pH 10) can swell the organic layer through deprotonation and electrostatic repulsion, enabling nearly complete water flux recovery. However, calcium-bridged layers are highly cohesive, and physical rinsing alone recovers only a small fraction of permeability. Alkaline cleaning improves hydrophilicity and disrupts some carboxyl-Ca2+interactions, enabling detachment of the fouling layer, but significant water flux reductions are still observed, demonstrating that ion bridges remain largely unbroken. This aligns with studies showing that base cleaning alone fails to fully dissolve calcium-crosslinked alginate or protein layers (see FIG. 11). The most effective cleaning strategy combines high pH with oxidative chemistry.

[0168] NaOCl in alkaline conditions cleaves organic macromolecules including alginate and BSA breaking polymer chains, disrupting structural integrity of the fouling matrix, and chemically eliminating bridging points. Under these conditions, salt bridges are disrupted and oxidized chains disperse readily, restoring permeability. Lastly, at higher pH, both membranes and foulants are negatively charged, creating strong electrostatic repulsion, which further aids cleaning. Among the tested membranes, the TMC-functionalized mLbL fdms exhibited the most negative surface charge at pH 10, directly correlating with their near-complete recovery, even in the presence of elevated Ca2+concentrations.

[0169] Validating mLbL Anti-Fouling Superiority using Carboxylate-Modified Latex (CML) Probe Interactions. AFM colloidal-probe force spectroscopy was employed to quantify foulantmembrane interactions and establish a direct mechanistic link to macroscopic fouling and cleaning behavior (FIG. 12). This approach is well-established in membrane science, where interfacial forces measured using AFM, particularly adhesion forces and interaction energies have been shown to correlate directly with fouling propensity and cleaning efficiency under varying solution chemistries. In this framework, AFM does not serve as a standalone characterization tool, but rather as a quantitative probe of the physicochemical interactions governing foulant deposition,32130761.00568 / 157813095v.1accumulation, and removal. Tn this study, interaction energy is reported using a signed convention, where more negative values correspond to lower detachment work (i.e., more repulsive), while more positive values indicate stronger adhesion. Complementarily, the maximum pull-off force reflects the peak adhesive interaction during detachment, providing a localized measure of foulant binding strength.

[0170] Under experimental conditions in the absence of foulant and Ca2+, the TMC-functionalized membrane shows the weakest interaction with the CML probe, which is attributed to its highly negative surface charge arising from a high density of COO groups. In contrast, the as-cast membrane contains both amine NHs and COO groups, resulting in a chemically heterogeneous surface where localized attractive interactions partially offset electrostatic repulsion, leading to slightly stronger interactions. The commercial membrane exhibits similar surface chemistry to the as-cast membrane but with significantly higher roughness (50-100* greater). This rough, ridge-and-valley morphology increases the effective contact area and creates localized regions where the probe can interact more strongly, resulting in adhesion comparable to that of the smooth as-cast mLbL surface.

[0171] In the presence of 20 mg / L BSA and absence of Ca2+, the interaction mechanism shifts from direct probe-membrane contact to interactions mediated by a foulant layer. Under these conditions, the as-cast surface exhibits weaker interaction compared to the TMC-functionalized surface. This behavior can be explained by the chemistry of the foulant used in the AFM measurements. BSA, unlike SA, contains both COO and NHL groups at neutral pH (FIG. 12, top middle), resulting in a lower net negative charge and a more heterogeneous charge distribution. While the COO -rich TMC-functionalized surface is expected to repel negatively charged foulants, the presence of NHv groups in BSA enables localized electrostatic attraction with the surface, promoting stronger adsorption and a more stable foulant layer. In contrast, the as-cast membrane contains both NIL and COO groups; the NHs groups can repel NHf sites in BSA, reducing favorable interactions and resulting in weaker overall adhesion. Surface roughness does not significantly contribute to this difference, as both mLbL membranes are ultrasmooth, with the TMC-functionalized surface (1.7 ± 0.1 nm) only slightly rougher than the as-cast surface (1.1 ± 0.1 nm). This difference is insufficient to affect physical entrapment or contact area, indicating that the observed trends are dominated by surface chemistry. The commercial membrane follows33130761.00568 / 157813095v.1the same trend established earlier, exhibiting the strongest interactions due to its rough morphology.

[0172] When 2 mM Ca2+is introduced alongside 20 mg / L BSA, the interaction behavior is further altered due to ion-bridging effects. Calcium ions can bind to carboxylate (COO ) groups present on both the foulant and the membrane surface, forming bridges that enhance cohesion within the foulant layer and strengthen its attachment to the membrane. Among the mLbL surfaces, the as-cast membrane exhibits the weakest interaction, which can be attributed to its heterogenous surface chemistry containing both NHs+and COO groups. The presence of NFL groups disrupt effective Ca2+bridging by introducing local charge repulsion somehow reducing the availability of adjacent COO sites required for stable coordination, resulting in weaker foulant attachment. In contrast, the TMC-functionalized membrane, which is enriched in COO groups, provides a more favorable environment for Ca2+-mediated bridging. The higher density of COO groups facilitate stronger coordination with Ca2+, promoting the formation of more cohesive and stable foulant structures and leading to increased interaction with the probe. Consistent with the earlier discussion, the commercial membrane exhibits the strongest interactions under these conditions due to its rough morphology.

[0173] Conclusions. In this example, the superiority of mLbL polyamide surfaces according to various aspects of the disclosure over commercial membranes in mitigating fouling and improving cleaning efficiency is demonstrated. This improvement arises from the combined effects of reduced surface roughness, controlled surface chemistry, and enhanced hydration behavior. Systematic filtration experiments were conducted, which showed that the ultrasmooth mLbL membranes limit foulant accumulation and enable improved flux recovery, whereas the commercial membrane, characterized by a rough ridge-and-valley morphology, promotes stronger foulant attachment and more severe water flux decline. To further elucidate the underlying fouling mechanisms, AFM colloidal-probe force spectroscopy was employed to quantify foulant-surface interactions. These measurements revealed that surface chemistry plays a dominant role in governing interfacial interactions. Among the mLbL surfaces, differences in interaction behavior arise from the distribution of COO and NFL+groups, which control electrostatic interactions and Ca2+-mediated bridging. Specifically, AFM results indicate that the TMC-functionalized surface exhibits stronger interactions with BSA due to localized attraction between COO groups on the membrane and NFL+groups in the foulant, while the as-cast surface shows comparatively weaker34130761.00568 / 157813095v.1interactions due to NHv-NH repulsion and disruption of favorable binding. However, to validate these findings under realistic operating conditions, dynamic filtration experiments were performed, which revealed a contrasting trend. Despite exhibiting stronger interactions in AFM measurements, the TMC-functionalized membrane demonstrated superior antifouling performance during filtration. This difference can be attributed to the formation of a stable hydration layer associated with the high density of COO groups, which weakens foulant adhesion and facilitates removal. These results highlight that AFM interaction measurements alone are insufficient to predict fouling behavior, and that hydration and hydrodynamic effects play a critical role during filtration.

[0174] Example 2 - Tailorins terminal chemistries of molecular layer-by-layer (mLbL) deposited polyamide reverse osmosis membranes for resistance towards silica scaling

[0175] Materials and Equipment. Trimesoyl chloride (TMC, molecular mass: 265.48 g / mol, purity: >98%), m-phenylenediamine (MPD, molecular mass: 108.14 g / mol, purity: >99 %), and poly(ethylene glycol) bis(amine) (PEG-amine, number average molecular mass: -3400 g / mol) were purchased from Sigma- Aldrich and used without further purification. 1H,1H-Perfluorooctylamine (PF A, molecular mass: 399.10 g / mol, purity: >97 %) was obtained from Alfa Aesar, and 3-aminopropyldimethylethoxysilane (APDES, molecular mass: 161.29 g / mol: purity: 97%) was obtained from Gelest, both used as received. An aqueous poly(sodium 4-styrenesulfonate) (PSS) solution (30 % by mass in water, molecular mass: 200,000 g / mol, purity: >98 %), used as a release layer, was obtained from Sigma-Aldrich, and diluted to 1 %. Toluene, ethanol, and tetrahydrofuran (THF) were sourced from Fisher Scientific. Toluene was dried over 3 A molecular sieves, which maintained a water content below 10 ppm, as measured using a Karl Fischer CS20X coulometric titrator (Mettler Toledo). THF was prepared by drying it using a solvent purification system from LC Technology Solutions. The water content in the THF was less than 25 ppm, as measured using a Karl Fisher CS20X coulometric titrator (Mettler Toledo), ensuring its suitability for use in the experiments.

[0176] Fabrication and functionalization of the mLbL polyamide films. We used four different substrates for specific experimental purposes: silicon wafers ((orientation <100, University Wafers) for contact angle measurements, gold-coated silicon wafers for film thickness measurements, silicon dioxide (SiC>2)-coated QCM (QuartzPro, 5 MHz 14 mm Cr / Au / SiCL Q Sense) sensors for silica nucleation studies, and porous PAN50 supports (Sepro Membranes, Inc.)35130761.00568 / 157813095v.1for membrane performance testing. For contact angle measurements, silicon wafers were diced into 1.5 cm x 1.5 cm squares and cleaned by rinsing with toluene and ethanol, drying with nitrogen, and treating with ultraviolet-ozone (Model 342, Jelight) for 20 min. To ensure covalent attachment of the mLbL polyamide fdms and prevent fdm detachment during experiments, the substrates were functionalized by submerging them in a 1% by mass solution of APDES in dry toluene, followed by gentle heating at 50 °C for 12 h. After modification, the APDES-treated samples were rinsed with toluene, ethanol, and deionized water (18 MQ cm, Barnstead Nanopure), then dried in a vacuum oven for 24 h.

[0177] For QCM-D measurements, SiCh-coated QCM sensors were also treated with APDES following the protocol described above. Polyamide films on these QCM substrates were prepared using the conventional intermediate rinse-based mLbL protocol (Chan et al. Review of Scientific Instruments 2012, 83, (11)). All solutions and rinse solvents were loaded into gas-tight syringes, passed through 0.45 pm syringe filters, and dispensed using a spin coater (Laurell Technologies) purged with dry air to minimize water exposure. Monomer concentrations were 0.1 % by mass (TMC and MPD) in toluene, and each deposition cycle consisted of sequential deposition and spinning of TMC (0.06 mL), toluene rinse (0.24 mL), MPD (0.06 mL), and THF rinse (0.24 mL) at room temperature. The thickness of the mLbL polyamide films was determined by the number of deposition cycles, which in all cases was set to 50.5 cycles. The final half-cycle was performed to ensure that the surface of the polyamide film was predominantly saturated with acid chlorides, enabling subsequent reaction with reagents of choice. Post-deposition, functionalization was performed by immersing films in 1 % by mass solutions in toluene at 75 °C for 24 h, followed by rinsing with toluene, ethanol, and DI water, and drying with nitrogen gas. For each surface functionalization, samples were prepared in triplicate (see FIG. 13).

[0178] Characterization of the mLbL polyamide films. The thickness of the mLbL polyamide films was measured using variable angle spectroscopic ellipsometry (M-2000XI, J. A. Woollam). Spectra were obtained at incidence angles between 50° and 70°, in 10° increments, and over a wavelength range of (210 to 1700) nm. The film thickness was modeled based on a Cauchy dispersion relation using the CompleteEASE software (J. A. Woollam). Chemical functionalization was previously confirmed via X-ray photoelectron spectroscopy (XPS). The surface charge of the polyamide surfaces was measured via the streaming potential technique using the SurPASS™ 3 electrokinetic analyzer with 0.1 mM KC1 at pH 6.8. We also performed atomic force microscopy36130761.00568 / 157813095v.1(AFM) imaging of the polyamide films using a Multimode 8 AFM (Bruker). Before imaging, the samples were rinsed with ethanol to remove potential contaminants and dried with ultrapure N2. Imaging was conducted with scan sizes of 5 pm, 1 pm, and 200 nm using a ScanAsyst-Air probe (Bruker), and the root-mean-square (rms) roughness of the films was determined at different locations on each sample.

[0179] Contact angles were measured using a goniometer (rame-hart 260-U4) via the sessile drop method using 3 different liquids, DI water, glycerol, and diiodomethane; to allow quantification of polar and nonpolar surface energies of the polyamide films. Before each contact angle measurement, the membrane samples (control, MPD, TMC, PEG-amine, and PF A) were rinsed with ethanol and ultrapure water, and then dried with ultrapure N2. A 1 pL droplet of the probe fluid was placed on the sample using a pipette, allowing the droplet to remain on the surface for a few seconds to 1 minute to establish a stable solution-surface interface, while avoiding complete evaporation. The static contact angle was determined by measuring the tangent angle at the liquid-surface interface after equilibrium was reached. Three contact angle measurements were taken at 3 different points (n=9) across each surface for each probe fluid.

[0180] For membrane performance testing, polyamide RO membranes were prepared using a rinse-free mLbL approach on PSS-coated silicon wafers, following our previous work 13. This method eliminates the need for intermediate rinsing after each cycle, thereby reducing both fabrication time and solvent / reagent use. Low monomer concentrations of 0.01 % by mass (TMC and MPD in toluene) were used, allowing deposition cycles to be carried out without intermediate solvent rinsing. Despite the adoption of the rinse-free fabrication approach, the water contact angles and surface morphology of the resulting fdms were indistinguishable from those prepared using the conventional mLbL process. Polyamide fdms of ~20 nm thickness were obtained after 25.5 deposition cycles, with the final half-cycle again producing an acyl chloride-dominated surface for subsequent functionalization. Functionalization was carried out by immersing polyamide-coated silicon wafers in 1 % by mass solutions of the reagent of choice in toluene at 75 °C for 24 h, followed by rinsing with toluene, and drying with nitrogen gas. The polyamide films were then transferred onto porous PAN50 using a film-floating method. The membranes were left to dry overnight, ensuring proper adhesion before being trimmed and mounted onto a dead-end filtration cell for performance testing. The PAN50 support was selected for its well-defined pore37130761.00568 / 157813095v.1size and spacing, enabling efficient water transport. Control samples (as-cast with no functionalization) underwent the same process but without any reagent.

[0181] Desalination and scaling experiments. Membrane performance was evaluated using a stirred dead-end filtration cell (HP4750, Sterlitech) with an active membrane area of 4.16 cm2. The membranes were first tested against aqueous solutions containing 1000 mg / L NaCl at an operating pressure of 30 bar (435 psi or 3 MPa). Salt concentration and rejection were measured using an Orion 3 Star Plus conductivity meter (Thermo Scientific), and experiments were performed in duplicate over a 4 h period. Once steady state was achieved, the membranes were subjected to silica scaling. For scaling experiments, 11.2 mM silica solutions were prepared using sodium metasilicate pentahydrate (purity: >98 %), 65 mM sodium chloride (NaCl, purity: >99 %), and 10 mM 3-(N-morpholino) propanesulfonic acid (MOPS, purity: >99 %) buffer, with the pH adjusted to 6.8 using 1 M hydrochloric acid (HC1, purity: >37 %). The buffer solution, consisting of 65 mM sodium chloride and 10 mM MOPS buffer, was adjusted to a pH of 6.8 using 1 M sodium hydroxide (NaOH, purity: >98 %). The saturation index (c) quantifies the thermodynamic driving force for the nucleation of silica and embodied by equation (1) :where [H4SiO4] is silicic acid concentration and Ksp is the equilibrium solubility product of amorphous silica gel(Ksp = IO’271, or 1.9* 10'3M at 25 °C).

[0182] To evaluate the scaling propensity of membranes, freshly prepared silica solution was fed into the filtration cell in two consecutive scaling cycles. In the first cycle, compacted membranes were exposed to the oversaturated 11.2 mM silica solution for 3 h. The membranes were then rinsed with DI water for 10 min and remounted for a second 3 h scaling cycle using the same silica solution recipe. This second cycle was used to assess water flux recovery after rinsing. All scaling experiments were performed in duplicate.

[0183] QCM-D scaling measurement. Silica scaling kinetics were studied using QCM-D experiments, which were performed in triplicate for each sensor (QuartzPro, 14 mm, 5 MHz, Cr / Au / SiCh) on a Biolin Scientific QSense Analyzer. The sensors were coated with polyamide films and functionalized with different functional groups following the procedures described earlier. Ultrapure water from the Milli-Q (Millipore, 18 MQ cm) water system was used for all38130761.00568 / 157813095v.1solution preparations and throughout all experiments. To prevent silica leaching from glass containers into the solutions, plastic polyethylene containers were used for preparing all scaling solutions.

[0184] Prior to inserting the sensors into the QCM-D modules, the sensors were first rinsed with toluene, ethanol, and DI water followed by drying with ultrapure N2. To establish a stable baseline, the sensors were first rinsed with water, followed by a 24 h exposure to the buffer solution. The silica solution was prepared according to the recipe described above and immediately introduced into the QCM-D modules at a flow rate of 0.1 mL / min, to allow silica deposition for an additional 24 h. The deposited silica mass was quantified using the Sauerbrey equation (2):where C = 17.7 ng Hz’1cm’2for a 5 MHz quartz crystal, n is the overtone number. When switching solutions during the QCM-D experiments, a stable baseline was determined to be when frequency shift Af is less than 0.2 Hz / 10 min.

[0185] Classical nucleation theory in predicting silica scaling. Classical Nucleation Theory (CNT) describes phase transition by predicting the energy barrier that must be overcome for a stable nucleus to form. It differentiates between homogeneous nucleation, which occurs in bulk solution, and heterogeneous nucleation, where a surface lowers the energy barrier, making nucleation more favorable. Following the method proposed by Oss et al. (Chemical reviews 1988, 88, (6), 927-941), we calculated the surface tension components and parameters of macromolecule-coated surfaces through contact angle measurements with three probe liquids. The Lifshitz-van der Waals (or non-polar) surface tension component (yiLW), electron-acceptor parameter (yf), electron-donor parameter (yf), and total surface tension (yitot) for the three probe liquids are presented in Rales 1-2. The electron-donor (y-) and electron-acceptor (y+) components describe polar interactions, including hydrogen bonding. The surface tension components of the surfaces were derived using the Young-Dupre equation (3):(l+COS0).yJot-2' (3)39130761.00568 / 157813095v.1

[0186] Here, 0 represents the measured contact angle of a probe liquid on a macromolecule-coated surface, and ymLW, ym , and ym’ are the Lifshitz-van der Waals surface tension component, electron-acceptor parameter, and electron-donor parameter of the mLbL surfaces, respectively. The subscripts m and 1 refer to the mLbL surface and the probe liquid, respectively. The ynJ^', ym, and ym components of polyamide surfaces, obtained by solving the Young-Dupre equation, quantify the surface’s non-polar and polar interactions with different probe liquids. The hydration energy was calculated using equation (4):~(lri~€C3S6)y^(4)

[0187] Here, 0 is the measured water contact angle, and YwOtis the total surface tension of the water. The extent of free energy barrier reduction can be quantified using the wetting function f(9) as described in equation (5):(5)

[0188] Here, AG^etand AG;*0?nrepresent the free energy barriers for heterogeneous nucleation and homogeneous nucleation, respectively. Wetting function f(9) is calculated using equation (6):4 (6)

[0189] Here, 0 is the contact angle between a silica nucleus and the mLbL surface. Following this, cos9 was calculated using equation (7):cosG ¥as-v.-'Yss(8)

[0190] The subscripts m, w, and s refer to the mLbL surface, water, and silica, respectively. The interfacial tensions between the two phases are calculated based on the yLW, y+, and y- values of each phase.ytotai (mJ / m2) yLWy+y yabWater 72.8 21.8 25.5 25.5 51 Glycerol 63.4 34 3.92 57.4 63.4 Diiodomethane 50.8 50.8 0 0 0130761.00568 / 157813095v.1Silica 47.76 47.14 0.002 47.87 0.62 Table 1. The surface tension components of water, glycerol and diiodomethane, and silicayLWy y- TMC 32.01 0.28 25.48 MPD 42.49 0.75 19.75 PFA 17.49 0.783 6.03 PEG-amine 40.6 0.95 34.38As-cast 33.15 0.22 27.37Table S2. The surface energy components of the mLbL surfaces calculated from the contact angle measurements of three probe liquids (i.e., water, glycerol and diiodomethane).

[0191] XDLVO analysis of different surface chemistries. The Extended Derj aguin-Landau-Verwey-Overbeek (XDLVO) theory is an advanced model for analyzing interactions between particles and surfaces in colloidal and interfacial systems. It expands on the classical DLVO theory, which balances van der Waals attractions and electrostatic double-layer repulsions to predict particle stability. While DLVO successfully explains many colloidal behaviors, it overlooks critical short-range interactions that dominate in certain conditions, such as those influenced by surface chemistry. XDLVO addresses this by incorporating additional forces, notably the Lewis acid-base (AB) interactions, alongside the traditional van der Waals (LW) and electrostatic (EL) components. The XDLVO theory has been utilized to quantify the surface interaction energies arising from the interaction between silica particles and various surface chemistries in an 11.2 mM sodium metasilicate pentahydrate solution. To calculate the total interaction energy, equation (9) has been used:as G«1 -r G&B 4. 01W

[0192] where Gel, GAB, and GLWindicate electrostatic, acid-base and Lifshitz-Van der Waals interactions respectively. Electrostatic interaction is expressed as equation (10):41130761.00568 / 157813095v.1?r£K^(rJ « sv ill . T Hit X &&P ( I t(10)

[0193] Here, d represents the separation distance between the nanoparticle (NP) and the surface, while TN and TS denote the surface charges of the nanoparticle and the collector surface, respectively. The term 1 / K corresponds to the double layer thickness, as defined in equation (11):(11)

[0194] Here, s is the permittivity of the medium, e represents the charge of an electron, k is the Boltzmann constant, T is the temperature of the solution at which the QCM-D experiment is performed, zi denotes the valency of ion i, and m is the number of such ions per unit volume.

[0195] The Lifshitz-Van der Waals (AGLW) components of the free energy of interaction between a nanoparticle and a surface are calculated using the extended DLVO theory as shown in equation (12):W „ ..... H pr(4*r) . ^f2?i(12)

[0196] Here, d is the separation distance between the nanoparticle (NP) and the surface, and r is the nanoparticle radius. H, the effective Hamaker constant for the nanoparticle-collector-water system, can be expressed as equation (13):H 24zrd| (\ / ~~ VrfF) * (ypp'’ ~~V - / \ - / (13)

[0197] Here, do is the minimum separation distance between the nanoparticle and the surface. Usually, it is considered to be 0.158 nm for various surfaces, and d is considered as the separation distance in nm. GABis characterized as a short-range interaction potential that decreases exponentially with distance. The range of this interaction is primarily governed by the decay length I, measured in nanometers (nm). The range of X, varies from 0.2 nm to up to 13 nm. For our analysis, we considered a range of values for 1 from 1 nm to 60 nm at an interval of 0.5 nm. We noticed that any value equal or lower than 1 nm yields quantitatively similar trends of GABfor all the surfaces. Here, AGABcan be expressed as equation (14):— 7 ( &&& ( CAB OS)- -z Kv . V hv J ’ Ks . V )(14)130761.00568 / 157813095v.1

[0198] Here, the term denotes the polar component of the surface free energy where N,W, S represents nanoparticle, water and surface respectively.can be calculated using equation (15). The interaction energy plots were generated in the software OriginPro 2024, with the interaction energy values expressed in kT units (IkT = 4.05xl0'21J). The total interaction energy, described in Equation S6, is calculated for the interaction between a nanoparticle (NP) and a surface as a function of the distance and the hydrophobic correlation length X.n — vn u(15)

[0199] The interaction energies relative to various surfaces are presented in FIG. 14. The XDLVO theory results align with experimental observations, highlighting the dominance of EL interactions over LW and AB interactions (FIG. 15). For instance, at a separation distance of 1 nm, TMC exhibits the strongest EL interactions with charged silica species. As the separation distance extends to approximately 4 nm, nearly all interactions diminish to zero, indicating that these interactions are most significant within the near-surface region, specifically between 1 nm and 2 nm. Interestingly, X serves as a critical factor as it depicts a short-range characteristic which arises due to the highly localized nature of hydrogen bonding and electron donor-acceptor interactions, which primarily occur at the molecular level. As reported in previous studies, as the X increases, the AB interactions diminish exponentially, with X typically ranging from 0.6 to 2 nm. In our analysis, when X is set to 1 nm: the AB interactions across different surfaces overlap, exhibiting nearly identical magnitudes (FIG. 15). At the same time, when the separation distance increases, the AB interactions diminish significantly.

[0200] Surface roughness, wetting properties, and surface charge of the mLbL polyamide films. The surface morphology and wetting properties of the polyamide surfaces terminated with different functional groups are shown in FIGS. 2 and 3, respectively. All surfaces exhibited a smooth topology (FIG. 2) with a root-mean-square (RMS) roughness below 1 nm, indicating that modifications to the terminating functional group had minimal impact on surface morphology. The smoothness of the polyamide films arises from controlled polymer growth and the stoichiometrically limited nature of the mLbL process.

[0201] The zeta potential measurements (FIG. 3, left) provide insight into the surface charge characteristics of polyamide films functionalized with different terminal groups. The as-cast43130761.00568 / 157813095v.1polyamide surface, consisting of unreacted amines and acid chlorides, served as the control and exhibited a moderately negative zeta potential at pH 6.8. Upon TMC termination, the surface became more negatively charged, due to the dominant presence of deprotonated carboxylate groups (-COO") and the consumption of unreacted amines. MPD-functionalized surfaces are rich in amine groups, whose pKa spans approximately 4 to 11. As a result, a significant fraction of these groups is protonated (-NH2 — > -ME ) at pH 6.8, giving the surface a net positive charge. The PFA functionalized membranes, despite having non-polar fluorinated groups at the surface, exhibits a negative zeta potential (-18 mV). This negative charge arises from the highly electronegative C-F bonds, where fluorine draws electron density, alters the structure of interfacial water, and promotes the accumulation of hydroxide (OH ) ions at the interface, resulting in a net negative zeta potential 27.

[0202] The contact angle measurements presented in FIG. 3 (right) provide a comparison of the wettability of polyamide surfaces functionalized with different chemical groups at pH 6.8, using DI water, glycerol, and diiodomethane as the three probe fluids with decreasing degrees of polarity. Among these surfaces, PEG-amine functionalization yields the lowest contact angles, particularly with DI water and glycerol, indicating a highly hydrophilic surface. This strong wettability is attributed to the affinity of PEG chains for water, which promotes extensive hydrogen bonding interactions with polar liquids. In comparison, PFA-functionalized surfaces are distinctly hydrophobic, exhibiting high contact angles and minimal wetting due to the dominance of nonpolar fluorinated groups. As for surfaces functionalized with deprotonated carboxyl groups (e g., TMC) and protonated amine groups (e.g., MPD), water interaction at pH 6.8 is primarily driven by hydrogen bonding and charge-dipole interaction. In addition to assessing surface hydrophobicity, the 3-probe liquid contact angles are further used to calculate the surface tension components of the polyamide films with various surface chemistries (See Classical nucleation theory in predicting silica scaling and XDLVO analysis of different surface chemistries (above) for detailed methodology), to explore the underlying mechanisms of silica nucleation. As discussed above, The Lifshitz-van der Waals (or non-polar) surface tension component (yiLW), electron-acceptor parameter (yi+), electron-donor parameter (yf), and total surface tension (yito1) for the three probe liquids are presented in Tables 1-2.

[0203] Desalination performance of membranes functionalized with various surface chemistries. To assess the desalination performance of the mLbL membranes, the functionalized44130761.00568 / 157813095v.1membranes were mounted in a stirred dead-end fdtration cell and tested with a 1 g / L NaCl feed at 30 bar (~ 435 psi) for 5 h. In our previous study, rinse-free mLbL membranes were reported to reach steady-state salt rejection within just 4 h, eliminating the need for overnight DI water compaction; a process commonly employed in most membrane studies. This rapid stabilization was attributed to their highly cross-linked and dense polyamide network, which reduces free volume and minimizes structural rearrangement under pressure, thereby allowing the membranes to compact more quickly than commercial interfacial polymerization-based membranes that exhibit greater surface-to-depth heterogeneity. All the mLbL-deposited membranes exhibit a gradual decline in permeance before reaching steady state, regardless of surface functionalization (FIG. 18, left), indicating a similar compaction behavior. The stabilized permeance values vary slightly across different terminal chemistries, ranging from 0.7 to 0.9 L m2h1bar1. In addition to similar water permeance, all functionalized membranes achieve consistently high NaCl rejection (>99.6 %) over the same period (FIG. 18, right), which is attributed to the high degree of crosslinking in the polyamide films.

[0204] Surface-chemistry dependent water permeance decline and recovery under silica scaling. The susceptibility of rinse-free mLbL membranes to silica scaling was evaluated through two consecutive 3-h scaling cycles, separated by a 10-minute DI water rinse to assess permeance reversibility. In the initial silica scaling cycle, distinct trends of water permeance were observed for membranes with different terminating chemistries. The PEG-amine terminated membrane showed water permeance decline similar to the as-cast membrane. In comparison, the MPD-terminated membrane exhibited the most severe scaling with the sharpest decline in permeance, whereas the TMC- and PFA-terminated membranes showed the least decline, maintained higher permeance throughout the test (FIG. 19, top left).

[0205] Following the DI water rinse, each surface chemistry exhibited distinct flux recovery behavior during the second silica scaling cycle (FIG. 19, top right). The MPD-terminated membrane showed the lowest flux recovery, while PF A- and TMC-terminated surfaces maintained the highest recovery and overall permeance stability. Overall, membranes that experienced the greatest permeance decline also exhibited the lowest flux recovery. For all membranes except the MPD-terminated surface, silica scaling was largely reversible upon rinsing. This observation contrasts with commercial IP membranes, where mineral scaling is typically reported as nearly irreversible due to their rough ridge-and-valley morphology that traps and anchors sealants. This45130761.00568 / 157813095v.1behavior was further validated by comparing the rinse-free mLbL membranes with a commercial DuPont brackish water BWXLE30 polyamide membrane. Despite exhibiting similar surface charge (commercial polyamide membrane: -33.9 ± 1.5 mV; rinse-free mLbL as-cast membrane: -33 ± 0.8 mV) (FIG. 14), the commercial membrane showed 2.25-fold lower flux recovery than the rinse-free mLbL membrane, emphasizing the advantage of the smoother and more uniform mLbL surface in enabling reversible silica scaling.

[0206] Membrane permeance decline correlates with silica mass deposition measured in QCM-D. We further measured the kinetics of silica mass deposition on QCM-D sensors coated with polyamide with different terminal chemistries (FIG. 19, bottom left). MPD-terminated films exhibited the highest and most rapid silica accumulation, while TMC- and PFA-terminated films showed the lowest and slowest rates of deposition. A strong correlation was observed between silica deposition measured by QCM-D and water flux decline measured in dead-end filtration cells (R2= 0.97, p = 0.03) (FIG. 19, bottom right). Membranes with greater silica deposition experienced correspondingly larger water flux decline, whereas those with minimal deposition exhibited the smallest flux loss. We note that our dead-end cells were operated under very slow stirring, which was intentionally adjusted to induce just enough turbulence at the membrane-solution interface to mimic the shear forces present in crossflow systems. The minimized concentration polarization effects are consistent with that the water flux decline is primarily driven by silica nucleation and deposition.

[0207] Limitations of classical nucleation theory in predicting silica scaling. Our membrane scaling experiments and QCM-D measurements together illustrates the role of terminating functional groups in controlling silica scaling. To determine whether silica scaling follows the predictions of the classical nucleation theory (CNT), water flux decline were plotted against hydration energy and the wetting function f(0) (FIG. 20). CNT predicts that both parameters control the thermodynamic driving forces for heterogeneous nucleation. However, our results indicate that neither hydration energy nor f(9) alone adequately describe the observed scaling behavior. Hydration energy was calculated from water contact angles (See Classical nucleation theory in predicting silica scaling (above) for details). Lower water contact angles correspond to more hydrophilicity, with increasingly negative hydration energies. It is well known that stronger hydration layers hinder the attachment of organic foulants. However, this relationship does not hold for silica scaling, as no trends were observed between water flux decline and the hydration46130761.00568 / 157813095v.1energy of surfaces. This apparent inconsistency can be rationalized by considering the chemistry of silicic acid and its silanol groups (Si-OH) and their interactions with various surface chemistries. Small silicic acid molecules and oligomeric species can diffuse through the hydration layer and grafted polymer films, despite the presence of an interfacial water barrier. Once near the membrane surface, these silanol groups are capable of forming strong hydrogen bonds (-Si-OH -O-C, -Si-OH O-H) with functional groups on the grafted polymer and underlying polyamide layer. Such hydrogen bonding interactions can initiate the condensation of silicic acid into nuclei directly at the surface, thereby promoting scaling even on hydrophilic polymers. Consequently, no clear correlation was observed (FIG. 20, left), consistent with a previous study.

[0208] For example, PEG-amine, despite exhibiting the highest hydrophilicity (-128.6 mJ / m2), still shows considerable water flux decline compared to TMC and PF A, both of which possess significantly lower hydration energies. In contrast, the MPD-terminated surface, with a hydration energy similar to PEG-amine (-114.5 mJ / m2), exhibits 99% water flux decline and nearly five times higher silica deposition, representing a clear deviation from the expected trend ( FIG. 19, bottom left, and FIG. 20, left). These observations suggest that hydration energy cannot explain the scaling behavior. Instead, additional factors such as membrane surface chemistry and charge interactions play a critical role in governing silica scaling kinetics on polyamide surfaces, as will be discussed in the following subsection.

[0209] The wetting function (or known as potency factor), f(9), was calculated to estimate the heterogenous silica nucleation barrier of on each tested polyamide surface. According to CNT, higher f(0) values should correspond to higher free energy barriers to heterogeneous nucleation, and therefore reduced silica scaling. However, as shown in FIG. 20 (right), no such trend was observed. Surfaces with comparable f(0) values displayed markedly different scaling behaviors, and overall, water flux decline did not decrease systematically with increasing f(0). For example, the hydrophobic PFA surface, which exhibits the lowest nucleation barrier, showed the least water flux decline, whereas the PEG-amine surface, with the highest nucleation barrier, experienced nearly 2.5 times greater water flux decline than PFA.

[0210] Overall, these results highlight that silica scaling does not strictly follow the predictions of heterogeneous nucleation theory. Rather, both heterogeneous and homogeneous pathways are involved, with the kinetics of silicic acid polymerization and the formation of colloidal aggregates. This interpretation is supported by Malvern Analytical Zetasizer47130761.00568 / 157813095v.1measurements, which detected hydrodynamic diameters ranging from 600 nm to over 3000 nm within 10-20 minutes of silicic acid solution preparation. Such large aggregates are capable of depositing directly onto the membrane surface, representing a homogeneous nucleation pathway that contributes to water flux decline alongside surface-mediated heterogeneous nucleation.

[0211] Dominance of electrostatic interactions and silica particle deposition in silica scaling. Silica scaling is strongly influenced by the interactions between negatively charged oligomeric silica species and the surface charge of membranes. At pH 6.8, orthosilicic acid (HtSiCL) remains largely uncharged due to its relatively high pV, of 9.2. However, polymerization produces oligomeric species such as disilicic acid (H«Si2O7) and hydroxylated trisiloxanes (HsCLSis), which possess lower pAavalues and are therefore more likely to ionize under these conditions. As a result, a substantial fraction of silica species in solution carries negative charge (see FIG. 21 for zeta potential measurement) and can interact electrostatically with membrane surfaces.

[0212] Among the tested membranes, the MPD-terminated surface showed the most severe scaling, with nearly complete flux loss (FIG. 22, left) and the highest silica mass deposition (FIG.19, bottom left). This outcome is attributed to its positive zeta potential (+28 mV), which strongly attracts negatively charged silica oligomers. In contrast, the TMC-terminated surface exhibited a highly negative zeta potential (-55 mV), resulting in strong electrostatic repulsion against silica species. Consequently, TMC showed minimal silica deposition and only -20% water flux decline. The as-cast polyamide surface displayed intermediate scaling behavior, with -40% water flux decline and -50% less silica deposition compared to MPD. This trend reflects the presence of both unreacted carboxylates (-COO"), which repel negatively charged silica oligomers, and protonated amines (-NIL ), which attract them.

[0213] To further elucidate the mechanisms behind these observations, an extended DLVO (XDLVO) analysis was performed. The results showed that Gibbs free energy contributions from electrostatic interactions overwhelmingly outweighed those from van der Waals and Lewis acidbase interactions (see XDLVO analysis of different surface chemistries (above) for detailed methodology and FIG. 15). As shown in FIG. 15 (top left), electrostatic interactions dominate the energy landscape, with MPD exhibiting strong attractive potentials and TMC showing strong repulsion at short separation distances. In comparison, van der Waals forces (FIG. 15, top right) and acid-base interactions (FIG. 15, bottom left) contributed far less to the total interaction energy and decayed rapidly beyond a few nanometers. However, deviations from this trend highlight the48130761.00568 / 157813095v.1influence of additional mechanisms. For example, the PFA surface exhibited a lower Gei compared to as-cast but still showed nearly three times lower water flux decline. Correspondingly, the correlation analysis between water flux decline and Gei yielded only a moderate trend (R2= 0.61, p = 0.12), with the lack of statistical significance confirming that Gei alone does not fully account for the variability in silica scaling. This necessitates consideration of other interfacial interactions, as elaborated in the following subsection.

[0214] The hydrophobic PFA surface in this study exhibited markedly different behavior, with negligible silica deposition and only minimal water flux decline. While electrostatic repulsion and smooth morphology contributed, we hypothesize that the disruption of interfacial water structuring by the high-electronegativity fluorinated groups reduced the stability of hydrogen bonding between silica species and the membrane, preventing persistent adhesion. This aligns with prior findings that perfluoroalkyl chains create highly electronegative, weakly polarizable surfaces with intrinsically low surface energy and poor wettability by aqueous solutions. At the water interface, the fluorinated groups disrupt hydrogen-bond networks and suppress orientational ordering of interfacial water molecules. Simulations of heterogeneous ice nucleation have shown that pre-ordering of interfacial water can facilitate crystal formation, whereas surfaces that frustrate such ordering raise the nucleation barrier. We surmise that the weakly structured, low-density interfacial water on perfluoroalkyl surfaces likely reduces the availability of nucleation sites for silica. Any residual interactions that do occur are likely dominated by weak London dispersion forces, which are insufficient to sustain scaling.

[0215] On the other hand, for the hydration-dominant PEG-amine surface, the hydrophilic PEG chains formed a hydration barrier that initially delayed silica nucleation, as observed by an induction period in QCM-D measurements (FIG. 23). Despite this delay, PEG-amine still showed considerable silica deposition. This can be attributed to its net positive charge (+10 mV) resulting from bis-amine chemistry, where one amine group reacts with TMC during film formation, while the second amine remains unreacted and protonated. The presence of this protonated amine attracts negatively charged silica oligomers, accelerating deposition after the induction phase.

[0216] Designing Scaling-Resistant RO Membranes. We here systematically explored the mechanisms of silica scaling by combining membrane performance testing, time-resolved measurement of silica scaling kinetics, and surface interaction analyses. Using the rinse-free mLbL technique, we fabricated ultra-thin (~20 nm), ultra-smooth (<1 nm RMS), and highly cross-linked49130761.00568 / 157813095v.1polyamide membranes with excellent control over terminal surface chemistries. This level of structural precision provided a controlled framework to isolate how specific surface properties influence silica deposition and removal.

[0217] The results demonstrate that silica scaling cannot be attributed to a single factor. While electrostatic interactions are dominant, other factors such as water molecule orientation at the membrane-silica solution interface, silanol chemistry, and hydrogen bonding also play critical roles. Beyond surface chemistry, membrane morphology is equally important. The ultra-smooth mLbL films facilitated the detachment of silica deposits through simple rinsing, in sharp contrast to conventional interfacial polymerization (IP) membranes, whose heterogeneous ridge-and-valley structures trap silica deposits and typically undergo irreversible scaling, as reported in the literature.

[0218] A number of examples have been described. Nevertheless, it will be understood that various modifications may be made without departing from the spirit and scope of the disclosure. Accordingly, other implementations are within the scope of the following claims.50130761.00568 / 157813095v.1

Claims

CLAIMS1. A method for fabricating a polyamide reverse osmosis and nanofdtration membrane, comprising:depositing a first monomer solution comprising trimesoyl chloride (TMC) onto a substrate; depositing a second monomer solution comprising m-phenylenediamine (MPD) or piperazine (PIP) onto the substrate; andrepeating the first and second depositing steps one or more times to form a polyamide film, wherein the first and second monomer solutions have concentrations ranging from about 0.01% to about 0.1% by mass fraction, andwherein the method does not include rinsing steps between the monomer depositing steps.

2. The method of claim 1, wherein the first and second monomer solutions have concentrations of about 0.01% by mass fraction.

3. The method of claim 1, further comprising transferring the polyamide film to a porous support.

4. The method of claim 3, wherein the porous support is a polyacrylonitrile (PAN) support or a polysulfone support.

5. The method of claim 1, further comprising depositing a release layer on the substrate prior to depositing the first monomer solution.

6. The method of claim 5, wherein the release layer comprises poly(sodium 4-styrenesulfonate) (PSS).

7. The method of claim 1, wherein the polyamide film has a thickness ranging from about 10 to about 200 nm.

8. A polyamide reverse osmosis membrane comprising:a polyamide film formed by depositing alternating layers of trimesoyl chloride (TMC) and at least one of m-phenylenediamine (MPD) and piperazine (PIP) without intermediate rinsing51130761.00568 / 157813095v.1steps, wherein the TMC and at least one of MPD and PTP are deposited from solutions having concentrations of about 0.01% to about 1% by mass fraction.

9. The polyamide reverse osmosis membrane of claim 8, wherein the polyamide film comprises a surface functionalized with terminating functional groups to impart one or more of antifouling properties, antiscaling properties, and enhance solute rejection properties.

10. The polyamide reverse osmosis membrane of claim 9, wherein the terminating functional groups are selected from carboxyls, amines, polyethylene glycols, fluorocarbons, and zwitterionic groups.

11. The polyamide reverse osmosis membrane of claim 8, wherein the polyamide film has a thickness of about 10 to about 200 nm.

12. The polyamide reverse osmosis membrane of claim 8, wherein the membrane has a salt rejection of at least about 99% for a 1000 mg / L NaCl solution.

13. The polyamide reverse osmosis membrane of claim 8, wherein the membrane has a water permeance of at least about 0.5 L m’2h'1bar1.

14. The polyamide reverse osmosis membrane of claim 8, wherein the membrane exhibits a steadystate salt rejection and water permeance within about 4 hours of operation.

15. The polyamide reverse osmosis membrane of claim 8, wherein the polyamide film exhibits a root-mean-square (RMS) surface roughness of less than about 10 nm.

16. The polyamide reverse osmosis membrane of claim 8, further comprising a porous support, wherein the polyamide film is disposed on the porous support.

17. The polyamide reverse osmosis membrane of claim 8, wherein the membrane exhibits reversible silica scaling, the reversible silica scaling characterized by recovery of at least about 90%52130761.00568 / 157813095v.1of initial water permeance following exposure to a silica-containing feed solution and subsequent rinsing with deionized water.

18. The polyamide reverse osmosis membrane of claim 8, wherein the membrane exhibits resistance to organic fouling, the resistance to organic fouling characterized by at least about 90% recovery of water permeance after exposure to organic foulants and subsequent cleaning with an aqueous solution.

19. The polyamide reverse osmosis membrane of claim 8, wherein the membrane exhibits a boron rejection of at least about 90% at a feed solution having a pH ranging from 7 to 10.

20. A system for fabricating a polyamide reverse osmosis membrane, comprising:a substrate;a first deposition mechanism configured to deposit a trimesoyl chloride (TMC) solution onto the substrate;a second deposition mechanism configured to deposit an m-phenylenediamine (MPD) or a piperazine (PIP) solution onto the substrate; anda controller programmed to control the first and second deposition mechanisms to repeatedly deposit the TMC solution and the MPD or PIP solutions without intermediate rinsing steps, wherein the TMC solution and the MPD or PIP solutions have concentrations of about 0.01% to about 0.1% by mass fraction.

21. The system of claim 20, wherein the first and second deposition mechanisms comprise spin coaters configured to distribute the TMC solution and the MPD or PIP solutions across the substrate.

22. The system of claim 21, wherein the controller is further programmed to control spin speeds of the spin coaters based on the concentrations of the TMC solution and the MPD or PIP solutions.

23. The system of claim 20, further comprising a transfer mechanism configured to transfer the polyamide film from the substrate to a porous support.53130761.00568 / 157813095v.

124. The system of claim 23, wherein the porous support is a polyacrylonitrile (PAN) support or a polysulfone support.

25. The system of claim 20, further comprising a release layer deposition mechanism configured to deposit a release layer on the substrate prior to deposition of the TMC solution and the MPD or PIP solutions.

26. The system of claim 25, wherein the release layer deposition mechanism is configured to deposit a poly(sodium 4-styrenesulfonate) (PSS) release layer.54130761.00568 / 157813095v.1

Images