Skip to main content
NCBI home page
Nature Communications logoLink to Nature Communications
. 2025 Nov 25;16:10481. doi: 10.1038/s41467-025-65493-6

Doctor-blading-assisted interfacial polymerization for green and scalable polyamide membrane fabrication

Guangjin Zhao 1,#, Haohao Liu 2,#, Chengyi Lan 3, Lijun Liang 3, Liangliang Dong 2,✉, Hong Meng 1,4,✉, Menachem Elimelech 5,6,✉
PMCID: PMC12647255  PMID: 41290685

Abstract

Industry-leading polyamide membranes are thin-film composites produced via interfacial polymerization (IP) at an alkane-water interface. However, the current fabrication method results in suboptimal membrane microstructure and compromised performance due to insufficient control of mass and heat transfer within the interfacial reaction zone. Furthermore, the fabrication process utilizes volatile alkane solvents, contributing to a significant environmental burden. Here, we report an IP strategy at an ionic liquid/water interface to synchronously achieve kinetic and thermodynamic control of the interfacial reaction, thereby optimizing the microstructure of polyamide membranes. The high viscosity and low volatility of the ionic liquid facilitate the integration of the industrial doctor blading technique into the IP process, enabling rapid, eco-friendly, and scalable polyamide membrane production. The resulting membrane exhibits an unprecedented combination of high pure water permeance (25.8 LMH/bar) and excellent salt (sodium sulfate) rejection (96.54%), surpassing the performance of commercial benchmark polyamide membranes. This facile fabrication strategy paves the way for the design and production of next-generation, high-performance thin-film composite membranes.

Subject terms: Polymers, Ionic liquids, Pollution remediation

Polyamide membranes are often fabricated using interfacial polymerization methods, though these methods can compromise membrane structure and performance. Here the authors design a polymerization method using ionic liquid and a doctor blading method to optimize membrane fabrication.

Introduction

Dwindling water resources—exacerbated by climate change, population growth, and water pollution—pose a significant challenge to sustainable development. Energy-efficient membrane separation technologies are emerging as a critical solution to global water scarcity1–6. Thin-film composite (TFC) polyamide membranes are the most widely used membranes in desalination and water purification technologies7–10. These membranes feature a thin, selective polyamide layer produced by interfacial polymerization (IP) on a porous polymeric support11–17.

In the past two decades, numerous studies have focused on the development of new materials (e.g., sacrificial interlayers, nanofillers, monomers) to tailor the physicochemical properties of TFC membranes, improving water permeability or water-solute selectivity18–22. However, the polyamide (PA) chemistry of piperazine (PIP) and trimesoyl chloride (TMC) monomers remains the gold standard for commercial TFC membranes. Commercial TFC membranes still rely on outdated manufacturing processes involving the soaking a microporous support in an aqueous solution of PIP and then immersing the PIP-impregnated support membrane into an alkane solution of TMC. The ultrafast reaction kinetics and exothermic nature of the reaction at the alkane-water interface hinder precise control of mass and heat transfer during current IP. This lack of control yields heterogeneous polyamide membranes with suboptimal morphology and microstructure, creating an inherent trade-off between selectivity and permeability23,24. Moreover, the alkane solvents (e.g., n-hexane, n-heptane, and cyclohexane) employed in this process are volatile organic compounds (VOCs), which are environmentally harmful impede the advancement of scalable green fabrication25–27. The remaining liquid organic waste is treated with lye and then incinerated, further hindering a sustainable manufacturing process28. Although much work has been undertaken to introduce new manufacturing technologies, including layer-by-layer deposition29,30, electrospray31,32, and vacuum filtration33,34, their complexity and heavy use of VOCs limit scalability and sustainable manufacturing. Consequently, there is an urgent need to develop robust, scalable, and green manufacturing technologies for next-generation polyamide membranes that can bridge the gap between laboratory innovation and industrial applications.

Ionic liquids (ILs) are nonvolatile, low-flammable substances with potent solvent properties35–37, which can circumvent environmental pollution caused by solvent evaporation and greatly improve greenness of the membrane manufacturing process. A key feature of ILs is their tunable water miscibility, which stems from the unique functional groups of their constituent ions and enables them to form stable interfaces with water. By creating an IL/water interface, the interfacial stability of the IP reaction could be significantly improved, leading to tunable control of the reaction kinetics (interfacial transport of PIP and TMC monomers) and thermodynamics (interfacial heat dissipation). More importantly, the inherent viscosity characteristics of ILs enable integration with the doctor-blading technique for membrane fabrication.

Herein, we introduce a groundbreaking doctor-blading-assisted IP (DBAIP) technology, offering a scalable, industrially adaptable pathway for producing high-performance PA membranes. The structural diversity of ILs enables precise control over heat dissipation and the interfacial transport of PIP and TMC monomers during IP. We systematically investigated critical process parameters, including IL structure, blade coating speed, blade-substrate gap, and reaction time, to optimize the IP process. The fabricated PA membranes exhibit high water permeability and separation efficiency as well as long-term stability. The environmental sustainability of the DBAIP process was rigorously assessed. From both fundamental research and industrial application perspectives, our developed DBAIP process addresses the limitations of conventional PA membrane manufacturing, offering a viable industrial solution for the green production of next-generation TFC membranes.

Results and discussion

DBAIP fabrication of TFC membrane at the ionic liquid/water interface

The fabrication of PA membranes using the DBAIP method is schematically illustrated in Fig. 1a. A polyacrylonitrile (PAN) support was immersed in an aqueous solution of PIP. Subsequently, a TMC solution dissolved in an IL was poured the top of the PAN support, followed by doctor-blading at a controlled speed to create a uniform liquid film. Following a defined reaction time and thermal treatment, a DBAIP membrane was successfully formed.

Fig. 1. Schematic illustration of the DBAIP membrane preparation.

Fig. 1

Open in a new tab

a Schematic representation of the DBAIP method for the fabrication of DBAIP membranes. b Chemical structure and physicochemical properties of ILs used to prepare DBAIP membranes. Photograph of an immiscible IL-water system with a clear interface. The water layer was dyed via Congo red. c Schematic illustration of the DBAIP process where the mass and heat transfer can be synergistically modulated at the IL-water interface. d,e Surface SEM images of PAN support and DBAIP membrane. f Cross-sectional TEM image of the DBAIP membrane. g Probability density function (PDF) of pore diameter distribution for DBAIP membranes prepared with different ILs.

The IL-water interface and the solubility of TMC monomers in ILs were identified as two key factors influencing the IP process. A series of hydrophobic imidazolium-based ionic liquids (IBILs), including 1-ethyl-3-methylimidazolium bis (trifluoromethylsulfonyl) imide (C2) (Figures S1a), 1-butyl-3-methylimidazolium bis (trifluoromethylsulfonyl) imide (C4) (Fig. S1b), 1-octyl-3-methylimidazole bis (trifluoromethylsulfonyl) imide (C8) (Fig. S1c), and 1-dodecyl-3-methylimidazole bis (trifluoromethylsulfonyl) imide (C12) (Fig. S1d), were employed as the organic phase in the DBAIP process, replacing traditional VOCs (Fig. 1b, c). All the ILs exhibited a stable and well-defined interface with water (Fig. S2), creating a robust film-forming environment for the IP reaction. The solubility and stability of TMC in the ILs were monitored using UV spectrophotometry at a wavelength of 700 nm for a period of 6 hours. As illustrated in Fig. S3 and S4a–d, the TMC solutions remained transparent with a transmittance exceeding 90% over the entire 6-hour observation period, confirming the excellent solubility of TMC in the ILs. It is worth noting that PIP also demonstrates excellent solubility in the same IL system (Figs. S5 and S6a–d). This property is essential for the IP process, as dissolved PIP molecules can readily diffuse across the interface and enable the continuous growth of a dense PA layer.

The successful fabrication of DBAIP membranes was confirmed by scanning electron microscopy (SEM) images (Fig. 1d, e and S7, S8a–d). All DBAIP membranes exhibited dense, defect-free layers on top of the nanoporous PAN support surface. Transmission electron microscopy (TEM) images revealed that the formed layer was in direct contact with the PAN support, with a thickness of 87 nm (Fig. 1f and S9a–e). Compared to the PAN substrate, DBAIP membranes exhibited a new FTIR band at 1667 cm-1, attributed to the C-N bond formed by the amidation of acyl chloride by the amine38, confirming the successful fabrication of DBAIP membranes (Fig. S10).

Notably, the physicochemical properties and morphologies of DBAIP membranes fabricated at different IL-water interfaces varied significantly. As the length of the alkyl substituent in the ILs increased, pore size (Fig. 1g, S11, 12), thickness (Fig. 2a), average surface roughness (Ra) (Fig. 2b, S13, 14), and free volume (Fig. S15) of the resulting selective layers decreased noticeably, while the degree of crosslinking exhibited a monotonic increase (Figs. S16–S21, Table S1–S4).

Fig. 2. PIP transport and heat transfer in the DABIP process.

Fig. 2

Open in a new tab

a Radar plot of the thickness of DBAIP membranes prepared with different ILs (C2-C12). b Radar plot of average roughness of DBAIP membranes prepared with different ILs (C2-C12). c,d Monomer concentration detected at the IL–water interface via UV spectroscopy versus interfacial diffusion time. e–h MD simulations of the potential of mean force (PMF) of the PIP molecule at different locations along the x coordinate in the IL-water system with different ILs (C2-C12). i Average heat release rate produced by the addition of PIP/water solution into TMC/IL solution measured by isothermal titration calorimetry. j MD simulations of temperature evolution at the interface in the four IL-water system. Smaller temperature fluctuations of the dotted lines indicate a more stable interface in the IL-water system. k MD simulations of interface stability for four IL-water systems. The dashed line denotes the ideal IL-water interface, and the scattered data points represent the distribution of IL molecules along the X-axis. The vertical distance between the scattered data point and the dashed line indicates the interfacial stability, with a smaller distance signifying a more stable interface. l Interfacial temperature distribution map of four IL-water systems, with red and blue colors representing high and low temperatures, respectively.

Investigations into the viscosity of four room-temperature ILs revealed a correlation between the viscosity of the ILs and the alkyl chain length of the IL cation (Fig. S22 and S23). Concurrently, the surface tension of ILs decreases with increasing alkyl chain length of cations, while the interfacial tension with water tends to increase (Fig. S24), a phenomenon associated with enhanced hydrophobicity39. These observations suggest that ILs with longer alkyl chains create a more stable fluid interface, facilitated by doctor-blading, providing a robust reaction platform for PIP and TMC. Such a stable interfacial environment is crucial for the formation of dense, cross-linked DBAIP membranes. Moreover, as the IL viscosity increases, the diffusion rate of TMC and PIP decreases, significantly influencing the characteristics of the IP process (Figs. 2c, d, Fig. S24 and S25)40. In systems with lower-viscosity ILs, the increased mobility of TMC and PIP enables a rapid reaction at the IL-water interface, leading to the formation of a thicker PA film (~109 nm) (Fig. S9). However, the imbalance between the rapid interfacial polymerization reaction and the slower diffusion rate of PIP across the interface results in a localized depletion of PIP relative to the concentration of TMC. This high concentration of unreacted TMC at the interface significantly increases the risk of hydrolysis41. Consequently, a PA film with a reduced cross-linking degree (42.0%) is formed. Conversely, the use of high-viscosity ILs impedes TMC diffusion, favoring the formation of a denser PA film with an increased crosslinking degree (from 42.0% to 56.3%) and a reduced thickness (from 109 to 87 nm) (Table S1–S4, Fig. S9).

Heat and mass transfer mechanisms at the ionic liquid/water interface

Understanding the heat and mass transfer mechanisms at the multiphase interface of our system is essential for the control of membrane properties. We employed a combined experimental and theoretical approach to investigate mass transport and heat transfer in the DBAIP process. In order to gain a deeper understanding of the DBAIP process, molecular dynamics (MD) simulations were conducted to generate configurations of PIP and TMC in four IL-water systems, and the corresponding diffusion process of PIP monomers and heat transfer in the IL/water interface were analyzed (detailed information on the simulations is provided in the Supporting Information, Fig.e S26, Table S5). MD simulations revealed an equilibrated interfacial structure of C12-water with a thickness of 2.3 nm (Fig. S27a, b), significantly lower than the interfacial thicknesses formed between C8-water (2.5 nm) (Fig. S27c, d), C4-water (2.8 nm) (Fig. S27e, f), and C2-water (3.1 nm) (Fig. S27g, h). The MD simulation results of the potential of mean force (PMF) demonstrate that as the alkyl chain length of ILs increases, the energy barrier for PIP to enter the ILs phase from the interface rises substantially from 10.5 to 15.8 kcal mol-1 (Fig. 2e–h). This implies that increased alkyl chain length of ILs can elevate the mass transfer resistance of PIP into the IL phase, thereby hindering the trans-interface transport of PIP (Fig. S28).

The exothermic nature of the IP process significantly influences the formation of polyamide selective layers. Rayleigh-Bénard convection, induced by the intense heat release, may lead to irreversible damage to the structural integrity of the selective layers42,43. Therefore, controlled heat release during IP is essential to produce structurally intact PA selective layers. The heat release rate during the IP process, measured using isothermal titration calorimetry, is presented in Fig. 2i, j. The results indicate that the C2-water solvent system exhibits the highest heat release rate in the DBAIP process, influencing the rapid diffusion of PIP and the formation of thicker PA-selective layers. To further investigate heat transfer during DBAIP, MD simulations were performed across the IL-water interface (Fig. S29). After 4 ns of the heat input, the atoms near the C2-water system exhibited the highest temperature with a uniform distribution (Figs. S30 and S31). It is worth noting that the intrinsic diffusion rate of the monomer in the C2 medium is more than twice that of the reference system (C12) (see Fig. 2c, d). This intrinsic diffusion advantage, combined with the thermal driving effect, greatly enhances the diffusion of monomers into the interface and intensifies the Marangoni effect, leading to enhanced interfacial disturbance and increased membrane surface roughness (Fig. 2k, Fig. S32)44. Conversely, the C12-water system effectively reduced membrane surface roughness due to its lower overall heat release (Fig. 2l, Figs. S33 and S34).

Scale-up potential of DBAIP membranes

To identify the optimal conditions for synthesizing high-performance DBAIP membranes, we conducted a systematic study of key parameters, including the use of different ILs, blade coating speed, the gap between the blade and substrate, and reaction time. As illustrated in Figs. S35a, b and S36, increasing the length of the alkyl substituent of ILs, blade coating speed, and reaction time resulted in DBAIP membranes with a monotonically decreasing pure water permeability (PWP) but an increasing salt (Na2SO4) rejection. Conversely, increasing the spacing between the blade and substrate led to only a slight increase in PWP and Na2SO4 rejection. Compared to commercial NF membranes (Fig. S37), an optimized DBAIP membrane exhibited the same Na2SO4 rejection but a five-fold increase in water permeance, demonstrating high separation efficiency. Considering both PWP and Na2SO4 rejection, the DBAIP membrane fabricated using C12, with a scraper speed of 20 mm/s, an IP reaction time of 180 s, and a blade-substrate spacing of 250 µm, designated as DBAIP-C12 membrane, was selected as a promising candidate for further characterization and performance evaluation.

To further evaluate its potential for nanofiltration (NF), a systematic performance assessment of the optimized DBAIP-C12 membrane was conducted (Fig. 3a). The membrane displayed low rejection rates for monovalent salt ions due to their smaller hydrated radii compared to the pore size of the membrane. As the hydrated radius of the salt ions increased, the rejection of the DBAIP-C12 membrane began to show an upward trend, indicating that the size-exclusion mechanism played a crucial role during filtration (Figs. 3b and 3c)45. The DBAIP-C12 membrane also displayed higher rejection of salts with divalent anions (Na2SO4) compared to salts with divalent cations (MgCl2 and CaCl2) due to Donnan exclusion by the negatively charged polyamide membrane46 (Figs. S38 and S39). It is noteworthy that the DBAIP-C12 membrane exhibited a high pure water permeance of 26 LMH/bar and Na2SO4 rejection exceeding 98%, significantly surpassing the performance of previously reported membranes (Fig. 3d, Table S6).

Fig. 3. Separation performance of the optimized DBAIP-C12 membrane.

Fig. 3

Open in a new tab

a Pure water permeance (PWP, orange bars, left axis) and rejection of various salts (green circles, right axis) using the DBAIP-C12 membrane. b Pore size distribution of the DBAIP-C12 membrane and the rejection of various cations with relation to their Stokes radius. c Illustration of the size-sieving mechanism of the DBAIP-C12 membrane. d NF performance comparison of DBAIP-C12 membrane with commercial polymeric membranes and emerging MOF, COF, and inorganic membranes. e Large-area DBAIP -C12 membrane fabricated by the DBAIP process. Insert: the SEM images of selected area on DBAIP-C12 membrane. f Schematic structures and photographs of the fabricated 1812 modules. g Salt rejection (right axis) and water permeance (left axis) demonstrating the long-term stability curve of 1812 modules. Error bars represent the SDs of data from at least three replicate measurements.

To demonstrate the scalability of the DBAIP method, we successfully fabricated a DBAIP-C12 membrane with a width of 30 cm and a length of 80 cm (Fig. 3e). The surface of this large-area membrane was highly uniform, suitable for producing a standard commercial 1812 membrane module commonly used in residential water purification (Fig. 3f). Notably, the size of a DBAIP-C12 membrane that can be produced is primarily limited by the dimensions of the coating equipment, indicating that large-area membranes can be easily fabricated. As shown in Fig. 3g, the membrane module maintained over 98% Na2SO4 retention and a stable water permeance during a 50 h long cycle test, demonstrating excellent durability and stability. Furthermore, the membrane module exhibited exceptional structural integrity at high operating pressures, as evidenced by a linear increase in water flux with increasing pressure (Fig. S40).

We also evaluated the antifouling performance of the DBAIP-C12 membrane using humic acid (HA) and bovine serum albumin (BSA) as model organic foulants (Fig. S41 and Table S7). After two operation cycles, the DBAIP-C12 membrane maintained an adequate flux recovery rate of over 85.0%. The membrane surface uniformity (Fig. S8) and low surface roughness (Fig. S14) of DBAIP-C12 membrane are two major characteristics contributing to its strong antifouling performance. Notably, the DBAIP-C12 membrane also showed excellent operational stability under high salt concentration and operating pressure conditions (Fig. S42 and S43, Table S8 and S9). Overall, the pressure, fouling, and high-salinity stability tests demonstrate the great potential of the DBAIP-C12 membrane for practical applications.

Ionic liquid recycling and membrane performance

Beyond scalability, the reusability of materials, particularly solutions, is a significant concern in membrane fabrication processes. To address this, we investigated the reusability of the ILs used for the fabrication of the DBAIP membranes. As illustrated in Fig. S44, after reduced pressure (vacuum) distillation, centrifugation, and vacuum drying, the recovered ILs exhibited similar appearance and structure as fresh ILs, as confirmed by optical photographs, mass spectrometry (Fig. 4a), and ¹H NMR spectroscopy (Fig. 4b, Figs. S45 and S46). Based on the analysis of m/z and 1H NMR spectrometry (details in Supplementary Information), the purity of recovered ILs was determined to be higher than 96.2% (Fig. 4c). Subsequently, the recycled ILs were used to prepare DBAIP membranes for NF testing. In comparison to DBAIP-C12 membranes fabricated with fresh ILs, the membranes prepared using recycled ILs exhibited lower Na2SO4 rejection (Fig. 4d). However, when a mixture of fresh and recycled ILs (1:1 mass ratio) was used, the resulting membranes demonstrated comparable NF performance. Furthermore, pore size analysis and SEM characterization provided additional evidence for the consistency of the microstructure and surface morphology of the membranes prepared using fresh and recovered ILs (Fig. 4e–g).

Fig. 4. Ionic liquid recycling and membrane performance.

Fig. 4

Open in a new tab

a Mass spectra of fresh and recovered ILs (C12). b 1H NMR spectra of fresh and recovered ILs (C12). c Purity of the recovered ILs (C12). d Pure water permeance and salt (Na2SO4) rejection of DBAIP membrane prepared with mixed fresh and recovered ILs at 1:1 mass radio. e Pore size distribution of DBAIP membrane prepared with fresh and recovered ILs (C12). Top surface SEM images of DBAIP membrane prepared with f fresh ILs (top image) and g mixed fresh and recovered ILs at 1:1 mass radio (bottom image). h Environmental impacts of i ILs (C12) j and hexane for green degree evaluation. Environmental impact includes the following nine categories:  global warming potential, ozone depleting potential, photochemical ozone creation potential, acidification potential, eutrophication potential, ecotoxicity potential to water, ecotoxicity potential to air, human carcinogenic toxicity potential to water, and human noncarcinogenic toxicity potential to water. k Comparison of the greenness values of ILs and other volatile organic solvents used for the preparation of TFC membranes by IP. Error bars represent the standard deviations of data from at least three replicate measurements.

By leveraging the characteristic low vapor pressure of ILs, replacing VOCs with ILs can significantly enhance the environmental sustainability of the membrane fabrication process. In this work, we provide calculations regarding the green degree (details in Supporting Information) to quantitatively evaluate the environmental impact of TFC membrane fabrication processes according to the green degree method47. Fig. 4h–j demonstrate that the green degree (Gd) value associated with the DBAIP technique approaches zero (−0.1041 Gd), substantially surpassing that observed in conventional hexane solvent based fabrication techniques for TFC membranes, where the Gd value is significantly lower (−3.9198 Gd) (Fig. 4k, Fig. S47). This observation suggests that replacing volatile organic solvents, such as butanone48, o-xylene49, acetone50, tetrahydrofuran51, and chloroform52, with ILs could effectively mitigate the environmental impact, and that the DBAIP process represents a greener alternative for the fabrication of polyamide membranes (Fig. 4k, Table S10). Moreover, as the membrane preparation was conducted using a programmable logic controller (PLC) system instead of manual operation, labor costs were significantly lower than those of manual systems. Therefore, the developed DBAIP technique demonstrates a compact, economical, scalable, and environmentally friendly approach compared to conventional fabrication techniques. However, to fully evaluate the industrial applicability of the DBAIP technique, the aforementioned analysis is insufficient. A more comprehensive assessment, including a life cycle assessment (LCA) based on raw materials, economic factors, and the energy consumption of the DBAIP process, will be the focus of future work.

We successfully developed a continuous, green, and large-scale DBAIP process for fabricating polyamide membranes using ILs as the organic phase for IP, replacing traditional volatile organic solvents. Defect-free polyamide membranes with a large area of 30 cm × 80 cm and a thin selective layer of 87 nm were prepared using this method. A series of laboratory-scale NF experiments and a pilot-scale trial demonstrated the good NF performance of the fabricated membranes, with high water permeance and an excellent Na2SO4 rejection. The fabrication process is significantly more sustainable compared to other conventional fabrication techniques in terms of process time, chemical usage, and waste generation. This process addresses the limitations of conventional fabrication techniques for producing high-performance TFC membranes and holds the potential for developing membranes for other applications in a compact and environmentally friendly manner.

Materials

Polyacrylonitrile (PAN) ultrafiltration membrane (PAN50, molecular weight cutoff ~50,000 g/mol) was purchased from Guochu Co. Ltd. (Xiamen, China). Piperazine (PIP, ≥99%), trimesoyl chloride (TMC, ≥98%), sodium sulfate (Na2SO4, ≥ 99%), magnesium sulfate (MgSO4, ≥99%), calcium chloride anhydrous (CaCl2, ≥99%), magnesium chloride anhydrous (MgCl2, ≥99%), lanthanum(III) chloride (LaCl3, ≥99%), sodium chloride (NaCl, ≥99%), samarium chloride (SmCl3, ≥99%), and gadolinium chloride (GdCl3, ≥99%) were purchased from Aladdin Reagent Co.Ltd. (Shanghai, China). 1-ethyl-3-methylimidazolium bis (trifluoromethylsulfonyl) imide (C2, ≥98%), 1-butyl-3-methylimidazolium bis (trifluoromethylsulfonyl) imide (C4, ≥98%), 1-octyl-3-methylimidazole bis (trifluoromethylsulfonyl) imide (C8, ≥98%), and 1-dodecyl-3-methylimidazole bis (trifluoromethylsulfonyl) imide (C12, ≥98%) were provided by Macklin Biochemical Co. Ltd. (Shanghai, China). Polyethylene glycols (PEG, ≥99%, 200, 400, 600, and 1000 Da), Hexane (≥98%) were purchased from Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China). The water used in all experiments was deionized water. All chemicals were used as received without further purification.

Preparation of nanofiltration membranes via conventional interfacial polymerization (CIP)

CIP membranes were fabricated on a PAN ultrafiltration substrate. The PAN substrate was first immersed in a 1.0 wt% PIP aqueous solution for 10 minutes to ensure complete pore penetration. Excess solution was then removed using a rubber roller. Subsequently, a 1.6 Wv% TMC hexane solution was poured onto the soaked PAN substrate for 180 s, followed by drying at 60 °C for 3 min to complete cross-linking. The resulting CIP membranes were thoroughly rinsed and stored in deionized water until use.

Preparation of DBAIP membranes

An aliquot of PIP (1.0 g, 11.6 mmol) was dissolved in 100 mL of deionized water. An 8 × 15 cm² PAN membrane was subsequently immersed in the resulting PIP aqueous solution for 10 minutes. The membrane was then promptly transferred to a casting machine. Excess surface water on the PAN membrane was carefully removed to prevent defects during fabrication via a plastic roller. Subsequently, 4 mL of an ILs solution containing 1.6 wt% TMC was then gently poured onto the membrane surface and evenly spread on the membrane surface under the action of the PLC-controlled scraper. Following reaction period at 25 °C, the ILs on the membrane surface was washed away with ethanol. The resulting membrane was finally immersed in deionized water for subsequent use.

Preparation of DBAIP membrane module

Initially, five DBAIP-C12 membranes with effective size of 30 × 80 cm were fabricated. These membranes were subsequently assembled into a spiral-wound module with a feed spacer (40 × 90 cm, thickness: 0.09 cm) and a permeate spacer (50 × 100 cm, thickness: 0.00254 cm). The module, possessing an effective area of approximately 1.2 m², was then enclosed within a membrane shell (model 1812) for subsequent nanofiltration experiments.

Nanofiltration performance of DBAIP membranes

In this study, a cross-flow evaluation device (Flowmwm-0015, Huapu, China) was used to test the nanofiltration performance of DBAIP membrane. Before recording the experimental data, it was pre-run for 1 h under and operating pressure of 4 bar in order to maintain the stability of the membrane performance, and the membrane under each condition was tested three times to ensure the accuracy. The water permeance is calculated using

Permeance=VΔP×A×t 1

where V represents the volume of permeate within a certain time t, A is the effective area of the membrane in the test cell (A = 12.56 cm2), and ΔP represents the operating pressure set during the experiment.

The rejection rate of the membrane is calculated by measuring the electrical conductivity of the feed liquid and the permeate liquid:

Aqueous solutions of inorganic salts were used to evaluate the rejection of the DBAIP membranes. The rejection (R) was calculated using the following equation:

R=1−CPCf×100% 2

where Cf represents the conductivity of the feed solution, and Cp represents the conductivity of the permeate, respectively. In the nanofiltration evaluation experiments, the feed solutions were 1 g/L aqueous solutions of LaCl3, SmCl3, GdCl3, Na2SO4, MgSO4, CaCl2, MgCl2 or NaCl.

Supplementary information

Supplementary Information (8.5MB, pdf)
Transparent Peer Review file (5MB, pdf)

Source data

Source data (4.7MB, xlsx)

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (U23A20688 to H.M. and 22278178 to L.D.), the National Key Research and Development Program of China (2021YFB3802600 to L.D.), the National Science Foundation of Jiangsu Province (BK20230106 to L.D.), and the Fundamental Research Funds for the Central Universities (JUSRP622035 to L.D.).

Author contributions

Conceptualization: H.M. and L.L.D. Supervision: H.M. and L.L.D. Membrane fabrication: G.J.Z. and H.H.L. Characterization and performance tests: G.J.Z. and H.H.L. MD simulations: C.Y.L. and L.J.L. Result analysis: G.J.Z., H.H.L., H.M., L.L.D., and M.E. Writing – original draft: G.J.Z. and H.H.L. Writing – review & editing: H.M., L.L.D., and M.E.

Peer review

Peer review information

Nature Communications thanks Jung-Hyun Lee and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

All data supporting the conclusions of this paper are available within the paper, the Supplementary Information, or from the corresponding author upon request. The source data underlying Supplementary Figs. 4, 6, 10–12, 15–22, 24, 25, 28–32, and 36–46 are provided in the Source Data file. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

These authors contributed equally: Guangjin Zhao, Haohao Liu.

Contributor Information

Liangliang Dong, Email: liangliangdong@jiangnan.edu.cn.

Hong Meng, Email: menghong@xju.edu.cn.

Menachem Elimelech, Email: menachem.elimelech@rice.edu.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-025-65493-6.

References

  • 1.Chisca, S. et al. Polytriazole membranes with ultrathin tunable selective layer for crude oil fractionation. Science376, 1105–1110 (2022). [DOI] [PubMed] [Google Scholar]
  • 2.Yao, Y. et al. High performance polyester reverse osmosis desalination membrane with chlorine resistance. Nat. Sustain.4, 138–146 (2021). [Google Scholar]
  • 3.H. W. Peng, X. F. Liu, Y. F. Su, J. P. Li & Q. Zhao, Advanced lithium extraction membranes derived from tagged modification of polyamide networks. Angew. Chem. Int. Ed. 62, e202312795 (2023). [DOI] [PubMed]
  • 4.Luo, X. Y. et al. Wrinkled metal-organic framework thin films with tunable Turing patterns for pliable integration. Science385, 647–651 (2024). [DOI] [PubMed] [Google Scholar]
  • 5.Guo, B. et al. Double charge flips of polyamide membrane by ionic liquid-decoupled bulk and interfacial diffusion for on-demand nanofiltration. Nat. Commun.15, 2282 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Elimelech, M. & Phillip, W. A. The future of seawater desalination: energy, technology, and the environment. Science333, 712–717 (2011). [DOI] [PubMed] [Google Scholar]
  • 7.Marchetti, P., Solomon, M. F. J., Szekely, G. & Livingston, A. Molecular separation with organic solvent nanofiltration: a critical review. Chem. Rev.114, 10735–10806 (2014). [DOI] [PubMed] [Google Scholar]
  • 8.Yuan, B. B., Zhao, S. C., Hu, P., Cui, J. & Niu, Q. J. Asymmetric polyamide nanofilms with highly ordered nanovoids for water purification. Nat. Commun.11, 6102 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Warsinger, D. M. et al. A review of polymeric membranes and processes for potable water reuse. Prog. Polym. Sci.81, 209–237 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Xie, M., Shon, H. K., Gray, S. R. & Elimelech, M. Membrane-based processes for wastewater nutrient recovery: technology, challenges, and future direction. Water Res.89, 210–221 (2016). [DOI] [PubMed] [Google Scholar]
  • 11.Chowdhury, M. R., Steffes, J., Huey, B. D. & McCutcheon, J. R. 3D printed polyamide membranes for desalination. Science361, 682–685 (2018). [DOI] [PubMed] [Google Scholar]
  • 12.Alhazmi, B. et al. Ultraselective macrocycle membranes for pharmaceutical ingredients separation in organic solvents. Nat. Commun.15, 7151 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Jiang, Z. W. et al. Aligned macrocycle pores in ultrathin films for accurate molecular sieving. Nature609, 58–64 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Tiwari, K. et al. Large area self-assembled ultrathin polyimine nanofilms formed at the liquid–liquid interface used for molecular separation. Adv. Mater.32, 1905621 (2020). [DOI] [PubMed] [Google Scholar]
  • 15.W. X. Jiang, et al. Axial alignment of covalent organic framework membranes for giant osmotic energy harvesting. Nat. Sustain.8, 446–455 (2025).
  • 16.Park, H. B., Kamcev, J., Robeson, L. M., Elimelech, M. & Freeman, B. D. Fructose-driven glycolysis supports anoxia resistance in the naked mole-rat. Science356, eaab0530 (2017). [DOI] [PubMed] [Google Scholar]
  • 17.Zuo, P. P. et al. Near-frictionless ion transport within triazine framework membranes. Nature617, 299–305 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Z. Wang, et al. Nanoparticle-templated nanofiltration membranes for ultrahigh performance desalination. Nat. Commun.9, 2004 (2018). [DOI] [PMC free article] [PubMed]
  • 19.Zhang, W. H. et al. Graphene oxide membranes with stable porous structure for ultrafast water transport. Nat. Nanotechnol.16, 337–343 (2021). [DOI] [PubMed] [Google Scholar]
  • 20.Karan, S., Jiang, Z. & Livingston, A. G. Sub-10 nm polyamide nanofilms with ultrafast solvent transport for molecular separation. Science348, 1347–1351 (2015). [DOI] [PubMed] [Google Scholar]
  • 21.Jimenez-Solomon, M. F., Song, Q., Jelfs, K. E., Munoz-Ibanez, M. & Livingston, A. G. Polymer nanofilms with enhanced microporosity by interfacial polymerization. Nat. Mater.15, 760–767 (2016). [DOI] [PubMed] [Google Scholar]
  • 22.You, X. et al. Electrostatic-modulated interfacial polymerization toward ultra-permselective nanofiltration membranes. iScience24, 102369 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Chen, J., Wang, T., Dai, R., Wu, Z. C. & Wang, Z. W. Trade-off between endocrine-disrupting compound removal and water permeance of the polyamide nanofiltration membrane: phenomenon and molecular insights. Environ. Sci. Technol.58, 9416–9426 (2024). [DOI] [PubMed] [Google Scholar]
  • 24.Wang, Z. G. et al. Microporous polymer adsorptive membranes with high processing capacity for molecular separation. Nat. Commun.13, 4169 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Saalberg, Y. & Wolff, M. VOC breath biomarkers in lung cancer. Clin. Chim. Acta459, 5–9 (2016). [DOI] [PubMed] [Google Scholar]
  • 26.Khan, S., Le Calvé, S. & Newport, D. A review of optical interferometry techniques for VOC detection. Sens. Actuat. A-Phys.302, 111782 (2020). [Google Scholar]
  • 27.Renz, H. & Bruchmann, B. Pathways targeting solvent-free PUR coatings. Prog. Org. Coat.43, 32–40 (2001). [Google Scholar]
  • 28.Nemeth, B., Lang, P. & Hegely, L. Optimisation of solvent recovery in two batch distillation columns of different size. J. Clean. Prod.275, 122746 (2020). [Google Scholar]
  • 29.Gu, J. E. et al. Molecular layer-by-layer assembled thin-film composite membranes for water desalination. Adv. Mater.25, 4778–4782 (2013). [DOI] [PubMed] [Google Scholar]
  • 30.He, R. et al. Unprecedented Mg2+/Li+ separation using layer-by-layer based nanofiltration hollow fiber membranes. Desalination525, 115492 (2022). [Google Scholar]
  • 31.Geng, X. L. et al. Poly(vinyl alcohol)/polydopamine hybrid nanofiltration membrane fabricated through aqueous electrospraying with excellent antifouling and chlorine resistance. J. Membr. Sci.632, 119385 (2021). [Google Scholar]
  • 32.Sun, Y. M. et al. Controllable construction of ultrathin graphene quantum dots/polyamide nanofilms via electrospray interfacial polymerization. Sep. Purif. Technol. 317, 123831 (2023). [Google Scholar]
  • 33.Zagklis, D. P., Vavouraki, A. I., Kornaros, M. E. & Paraskeva, C. A. Purification of olive mill wastewater phenols through membrane filtration and resin adsorption/desorption. J. Hazard. Mater. 285, 69–76 (2015). [DOI] [PubMed] [Google Scholar]
  • 34.Xie, A. T. et al. Graphene oxide/Fe(III)-based metal-organic framework membrane for enhanced water purification based on synergistic separation and photo-Fenton processes. Appl. Catal. B-Environ.246, 118548 (2020). [Google Scholar]
  • 35.Gao, S. Q. et al. The ionic liquid-H2O interface: a new platform for the synthesis of highly crystalline and molecular sieving covalent organic framework membranes. ACS Appl. Mater.Interfaces 13, 36507–36516 (2021). [DOI] [PubMed] [Google Scholar]
  • 36.Yu, G. Q., Jiang, Y. F., Cheng, J. & Lei, Z. G. Structural effect on the vapor-liquid equilibrium of toluene-ionic liquid systems. Chem. Eng. Sci.198, 1–15 (2019). [Google Scholar]
  • 37.Pan, Y. et al. Membranes based on covalent organic frameworks through green and scalable interfacial polymerization using ionic liquids for antibiotic desalination. Angew. Chem. Int. Ed.63, e202316315 (2023). [DOI] [PubMed] [Google Scholar]
  • 38.Peng, H. W. et al. Quaternization-spiro design of chlorine-resistant and high-permeance lithium separation membranes. Nat. Commun.14, 5483 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Qiu, Y. & Molinero, V. Morphology of liquid-liquid phase separated aerosols. J. Am. Chem. Soc.137, 10642–10651 (2015). [DOI] [PubMed] [Google Scholar]
  • 40.Yao, Y. J. et al. More resilient polyester membranes for high-performance reverse osmosis desalination. Science384, 333–338 (2024). [DOI] [PubMed] [Google Scholar]
  • 41.Y. A. Zhang, et al. Ice-confined synthesis of highly ionized 3D-quasilayered polyamide nanofiltration membranes. Science382, 202–206 (2023). [DOI] [PubMed]
  • 42.Zhao, G. J. et al. Polyamide nanofilms through a non-isothermal-controlled interfacial polymerization. Adv. Funct. Mater.34, 2313026 (2024). [Google Scholar]
  • 43.Ma, X. H. et al. Nanofoaming of polyamide desalination membranes to tune permeability and selectivity. Environ. Sci. Tech. Let.5, 123–130 (2018). [Google Scholar]
  • 44.Deshmukh, A., Lienhard, J. H. & Elimelech, M. Heat diffusion during thin-film composite membrane formation. J. Membr. Sci.696, 122493 (2024). [Google Scholar]
  • 45.Sarkar, P., Modak, S. & Karan, S. Ultraselective and highly permeable polyamide nanofilms for ionic and molecular nanofiltration. Adv. Funct. Mater.31, 2007054 (2021). [Google Scholar]
  • 46.Seidel, A., Waypa, J. J. & Elimelech, M. Role of charge (Donnan) exclusion in removal of arsenic from water by a negatively charged porous nanofiltration membrane. Environ. Eng. Sci.18, 105–113 (2001). [Google Scholar]
  • 47.Zhang, X. P., Li, C. S., Fu, C. & Zhang, S. J. Conversion of corncob into biohydrogen by anaerobic fermentation. Ind. Eng. Chem. Res.47, 1085–1094 (2008). [PubMed] [Google Scholar]
  • 48.Esfandian, F., Peyravi, M., Ghoreyshi, A. A., Jahanshahi, M. & Rad, A. S. Fabrication of TFC nanofiltration membranes via co-solvent assisted interfacial polymerization for lactose recovery. Arab. J. Chem.12, 5325–5338 (2019). [Google Scholar]
  • 49.Gupta, R., Singhal, M. & Chaudhary, J. P. Regulation of nanostructured polyaniline synthesis and its properties through organic solvent in interfacial polymerization. Colloid Surf. A.632, 127806 (2022). [Google Scholar]
  • 50.Kong, C. L., Shintani, T., Kamada, T., Freger, V. & Tsuru, T. Co-solvent-mediated synthesis of thin polyamide membranes. J. Membr. Sci.384, 10–16 (2011). [Google Scholar]
  • 51.De Guzman, M. R. et al. Cosolvent-driven interfacial polymerization for superior separation performance of polyurea-based pervaporation membrane. Polym.-Basel13, 1179 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Son, M., Peddie, F., Mulcahy, D., Davey, D. & Haskard, M. R. A microbioreactor based on interfacial polymerisation and application to flow injection analysis of glucose. Sens. Actuat. B-Chem.34, 422–428 (1996). [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Information (8.5MB, pdf)
Transparent Peer Review file (5MB, pdf)
Source data (4.7MB, xlsx)

Data Availability Statement

All data supporting the conclusions of this paper are available within the paper, the Supplementary Information, or from the corresponding author upon request. The source data underlying Supplementary Figs. 4, 6, 10–12, 15–22, 24, 25, 28–32, and 36–46 are provided in the Source Data file. Source data are provided with this paper.

Articles from Nature Communications are provided here courtesy of Nature Publishing Group

RESOURCES