Pine needle-shaped structure enhanced interface bonding PTFE (Polytetrafluoroethylene) ceramic membrane as well as preparation method and application thereof

Abstract

The invention relates to the field of separation membranes, in particular to a PTFE (polytetrafluoroethylene) / ceramic membrane with a pine needle-shaped structure for enhancing interface bonding as well as a preparation method and application of the PTFE / ceramic membrane. The pine-needle-shaped TiO2 intermediate layer is loaded on the surface of the ceramic membrane through an in-situ hydrothermal synthesis technology, the macroporous surface of the ceramic support body is changed, permeation of polytetrafluoroethylene particles is effectively prevented, and the polytetrafluoroethylene membrane layer is embedded, so that combination between the two layers of membranes is enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of separation membranes, in particular to a PTFE / ceramic membrane with pine needle structure reinforced interface bonding and a preparation method and application thereof. BACKGROUND

[0002] Porous hydrophobic membranes play a crucial role in various separation and purification processes due to their unique surface properties and selective permeability. They can reject liquid water while allowing gaseous water, gases, and oil-containing liquids to pass through, making them very effective in applications such as membrane distillation, gas absorption, oil-water separation, and wastewater treatment. Among these membranes, polytetrafluoroethylene (PTFE) is an ideal material for making hydrophobic membranes due to its excellent hydrophobicity and chemical stability. However, polytetrafluoroethylene membranes prepared by traditional mechanical stretching and electrospinning methods usually have fiber nodes and network pore structures with a wide pore size distribution, which is not conducive to the separation precision and anti-fouling performance of the membranes.

[0003] To address these limitations, solid-state granular sintering has been explored as an alternative method for membrane fabrication. This method forms a porous structure through the uniform packing of particles, which typically results in a narrower pore size distribution. Similarly, sintering polytetrafluoroethylene particles into a membrane layer is expected to result in more precise control of pore size. However, the sintering process of polytetrafluoroethylene membranes typically requires high temperatures (300-400℃), making the use of traditional organic supports challenging. Ceramic membranes are characterized by high mechanical strength, good thermal stability, and resistance to solvents and acids / alkalis, making them ideal supports for organic-inorganic composite membranes.

[0004] For ceramic-supported composite membranes, obtaining a complete and defect-free membrane layer on the surface of a macroporous ceramic is both a key task and a major challenge. However, the hydrophobicity of polytetrafluoroethylene membranes is mainly attributed to their low surface energy, and as such, the interfacial bonding between PTFE particles and between the PTFE membrane layer and the ceramic support is weak, which inhibits the integrity and mechanical strength of the composite membrane. In the prior art, ceramic membranes with a pore size less than 0.20μm are commonly used as supports. Although the small pore size of the ceramic support effectively prevents the penetration of polytetrafluoroethylene particles, it increases the mass transfer resistance and inhibits the permeability of the composite membrane. At the same time, the surface of the small-pore-size ceramic membrane is usually smooth, resulting in weak bonding between the ceramic support and the polytetrafluoroethylene membrane layer, which can lead to the problem of membrane layer peeling.

[0005] The use of macroporous ceramic supports can significantly reduce the mass transfer resistance and improve the flux of the membrane. In addition, the relatively rough surface of the macroporous support improves the interfacial bonding strength with the organic membrane layer. However, the macroporous structure also exacerbates the problem of membrane particle penetration. To solve the above problems, the prior art often uses an intermediate layer or a sacrificial layer to cover the macroporous support, but the intermediate layer disclosed in the prior art reduces the bonding strength between the macroporous support and the organic membrane layer. Therefore, how to avoid membrane particle penetration while solving the bonding strength between the multi-layer membrane is a problem to be solved. SUMMARY

[0006] The present application relates to the field of separation membranes, in particular to a pine needle structure reinforced interface bonding PTFE / ceramic membrane and its preparation method and application. BACKGROUND

[0008] Porous hydrophobic membranes play a crucial role in various separation and purification processes due to their unique surface properties and selective permeability. They can reject liquid water while allowing gaseous water, gases, and oil-containing liquids to pass through, making them very effective in applications such as membrane distillation, gas absorption, oil-water separation, and wastewater treatment. Among these membranes, polytetrafluoroethylene (PTFE) is an ideal material for making hydrophobic membranes due to its excellent hydrophobicity and chemical stability. However, polytetrafluoroethylene membranes prepared by traditional mechanical stretching and electrospinning methods usually have fiber nodes and network pore structures, with a wide pore size distribution, which is not conducive to the separation precision and anti-fouling performance of the membrane.

[0009] To address these limitations, solid-state particle sintering has been explored as an alternative method for membrane fabrication. This method forms a porous structure through the uniform packing of particles, which typically results in a narrower pore size distribution. Similarly, sintering polytetrafluoroethylene particles into a membrane layer is expected to result in more precise pore size control. However, the sintering process of polytetrafluoroethylene membranes usually requires high temperatures (300-400℃), making the use of traditional organic supports challenging. Ceramic membranes are characterized by high mechanical strength, good thermal stability, and resistance to solvents and acids / alkalis, making them ideal supports for organic-inorganic composite membranes.

[0010] For ceramic-supported composite membranes, it is a key task and a great challenge to obtain a complete defect-free membrane layer on the surface of a macroporous ceramic. However, the hydrophobicity of a polytetrafluoroethylene (PTFE) membrane is mainly attributed to its low surface energy, and thus the interfacial bonding force between PTFE particles and between the PTFE membrane layer and the ceramic support is weak, which inhibits the integrity and mechanical strength of the composite membrane. In the prior art, a ceramic membrane with a pore size less than 0.20 μm is often used as a support. Although the small pore size of the ceramic support can effectively prevent the penetration of PTFE particles, it increases the mass transfer resistance and inhibits the permeability of the composite membrane. Meanwhile, the surface of the ceramic membrane with a small pore size is usually smooth, resulting in a weak bonding force between the ceramic support and the PTFE membrane layer and a problem of membrane layer peeling.

[0011] The use of a macroporous ceramic support can significantly reduce the mass transfer resistance and improve the flux of the membrane. In addition, the relatively rough surface of the macroporous support improves the interfacial bonding strength with the organic membrane layer. However, the macroporous structure also exacerbates the problem of membrane particle penetration. To solve the above problems, the prior art often uses an intermediate layer or a sacrificial layer to cover the macroporous support, but the intermediate layer disclosed in the prior art reduces the bonding strength between the macroporous support and the organic membrane layer. Therefore, how to avoid membrane particle penetration while solving the bonding strength between the multi-layer membranes is a problem to be solved. SUMMARY

[0013] In view of the above problems, the present application loads a layer of pine needle-shaped TiO2 intermediate layer on the surface of a ceramic membrane by in-situ hydrothermal synthesis technology, which not only changes the macroporous surface of the ceramic support, effectively prevents the penetration of PTFE particles, but also embeds the PTFE membrane layer, thereby strengthening the bonding between the two membrane layers.

[0014] Specifically, the present application provides a PTFE / ceramic membrane with pine needle-shaped structure enhanced interfacial bonding, which comprises a ceramic membrane as a support and a PTFE separation membrane as a separation layer, and has a pine needle-shaped TiO2 intermediate layer between the ceramic membrane and the PTFE separation membrane.

[0015] Preferably, the average pore size of the ceramic membrane is 1-3 μm, more preferably 1.5-2.5 μm.

[0016] Preferably, the surface roughness factor of the TiO2 intermediate layer is 2-5, preferably 3-4.

[0017] Preferably, the PTFE separation membrane contains carbon nanotubes, and the loading amount of the carbon nanotubes in the PTFE separation membrane is 1-8 wt%.

[0018] The application also provides a method for preparing the PTFE / ceramic membrane with pine needle structure reinforced interface combination as claimed in claim 1, which comprises the following steps:

[0019] A hydrochloric acid solution with a molar ratio of HCl:H2O of 1:6-1:13 is prepared, stirred, then tetrabutyl titanate is added, and stirring is continued for 10-60 min to form a mixed solution;

[0020] The ceramic membrane is immersed in the mixed solution, and reacted at 120-180℃ for 5-25 h to form a ceramic membrane with an intermediate layer;

[0021] PTFE particles, water, a dispersing agent and a film forming agent are mixed to form a film forming solution, and the film forming solution is coated on the ceramic membrane, and after drying, the PTFE / ceramic membrane is formed.

[0022] Preferably, in step (1), the molar ratio of HCl and tetrabutyl titanate is 5-20:1.

[0023] Preferably, in step (2), the ceramic membrane needs to be pretreated before being immersed in the mixed solution, and the pretreatment is to ultrasonically clean the ceramic membrane with ethanol and deionized water in sequence; the reaction temperature is preferably 140-160℃, and the reaction time is preferably 12-17 h.

[0024] Preferably, in step (3), the film forming solution does not contain / contains carbon nanotubes, the dispersing agent is one or more of fatty alcohol polyoxyethylene ether, polyacrylamide, sodium carboxymethyl cellulose, sodium dodecylbenzenesulfonate, fluorinated alkyl ethoxyl alcohol ether, sodium butylphenyl sulfonate, sodium hydroxyethyl sulfate, sodium dodecyl sulfate, polyvinylpyrrolidone, polyacrylic acid salt, and polyethylene carboxylic acid, and the film forming agent is one or more of polyvinyl alcohol, polyvinyl butyral, hydroxypropyl methyl cellulose, ethyl cellulose, polyethylene glycol, polypropylene glycol, and glycerol.

[0025] Preferably, in step (3), the coating time is 10-60 s; the drying is drying at room temperature for 8-16 hours, and then further drying in an oven at 60-80℃ for 8-16 hours; the calcination is sintering the composite membrane at a temperature rising speed of 0.5-2℃ / min to 280-340℃ for 5-60 min; the sintering temperature is preferably 300-310℃, and the sintering time is preferably 10-15 min.

[0026] The PTFE / ceramic membrane with pine needle structure reinforced interface combination provided by the application can be applied in water-N-methyl-2-pyrrolidone system separation.

[0027] Compared with the prior art, the application has the following advantages:

[0028] Firstly, the pine needle-shaped TiO2 intermediate layer of the present application covers the macroporous surface of the ceramic support on one hand, effectively preventing the penetration of PTFE particles and avoiding the problem of particle loss caused by macroporous structure; on the other hand, the pine needle-shaped structure is embedded in the PTFE separation layer to form a mechanical interlocking effect, greatly improving the interfacial bonding strength between the two layers of membranes. Compared with the traditional smooth intermediate layer or sacrificial layer, this structure not only ensures high permeation flux, but also solves the problem of insufficient bonding force of multi-layer membranes.

[0029] Secondly, the present application can grow a uniformly distributed pine needle-shaped TiO2 intermediate layer on the surface of the ceramic membrane in situ through the selection and ratio of acid and precursor, hydrothermal reaction conditions (such as temperature and time). This method does not require complex process or additional additives, and the surface roughness of the formed intermediate layer is appropriate, which not only provides enough anchoring points to enhance the adhesion of the PTFE layer, but also avoids defects caused by excessive roughness, achieving the balance between interfacial bonding and structural integrity.

[0030] In addition, CNTs as reinforcing phase, through its high aspect ratio and network structure, inhibit the shrinkage stress of PTFE during high temperature sintering process, reduce the membrane cracking phenomenon; at the same time, the rigid properties of CNTs further improve the mechanical strength of the composite membrane, so that it maintains the structural stability in long-term use.

[0031] Finally, the PTFE / ceramic composite membrane obtained by the present application has uniform morphology, narrow pore size distribution and excellent thermal stability and mechanical stability, and is especially suitable for harsh separation environment. In the water-NMP system separation, the membrane exhibits high permeation flux and excellent NMP rejection performance under a wide range of feed concentration and operating temperature. Long-term test shows that its structural stability and separation efficiency do not decrease obviously, which verifies its potential in industrial separation application. BRIEF DESCRIPTION OF DRAWINGS

[0032] Figure 1 : Schematic diagram of VMD device and process in the present application;

[0033] Figure 2 : Influence of hydrothermal duration: (a) schematic diagram of microstructure; (b) scanning electron microscope image; (c) EDS composition diagram; (d) white light interferometer image;

[0034] Figure 3 : Influence of molar ratio of HCl to TBOT: (a) surface SEM image, (b) cross-section SEM image, (c) XRD pattern, (d) average roughness, and (e) pore size and pure water permeability;

[0035] Figure 4: Dip-coating PTFE film on ceramic support: (a) PTFE suspension and pore size distribution (PSD) of ceramic support, (b) wetting process of PTFE emulsion droplets on the surface of ceramic support, (c) effect of coating time on PTFE film thickness, (d) FESEM micrographs of PTFE / ceramic composite film surface and pore size distribution, (e) schematic diagram of the effect of hydrothermal interlayer on PTFE film;

[0036] Figure 5 : Effect of CNT doping on microstructure and pore size distribution of PTFE / ceramic composite film: (a) 1%, (b) 2%, (c) 4%, (d) 8%, and (e) schematic diagram of CNT doping preventing cracking;

[0037] Figure 6 : Characterization of PTFE / ceramic composite film: (a) cross-sectional FESEM micrograph and elemental distribution map, (b) nano-scratch test, (c) friction coefficient, and (d) long-term wear resistance;

[0038] Figure 7 : Performance of composite film in SR-MD process: (a) flux and rejection rate at different temperatures, (b) saturated vapor pressure of water and NMP at different temperatures, (c) predicted NMP concentration in permeate and NMP rejection rate, (d) flux and rejection rate at different NMP contents, (e) flux in long-term test, (f) NMP rejection rate in long-term test; DETAILED DESCRIPTION

[0039] Example 1

[0040] The method of this example 1 for preparing PTFE / ceramic film with pine needle-like structure to enhance interfacial bonding includes the following steps:

[0041] Hydrochloric acid (mass concentration 36%) was mixed with deionized water at a volume ratio of 1:1 and stirred for 30 minutes; then, a certain amount of tetrabutyl titanate (TBOT) with a mass concentration of 99% was added dropwise to the mixture (Hcl and TBOT molar ratio was 20:1, 16:1, 13:1 and 10:1), and stirring was continued for another 30 minutes to form a mixed solution;

[0042] Alumina with an average pore size of about 1.7 μm was sequentially ultrasonically cleaned with ethanol and deionized water, and then dried for standby use; 40 mL of the prepared solution and the cleaned alumina support were placed in a 100 mL Teflon-lined stainless steel hydrothermal reactor. The reactor was transferred to an oven and heated at a temperature of 150°C for a specified time (0h, 15h, 20h and 25h). After the reaction was completed, the system was cooled to room temperature, and then the reactor was opened. Finally, the alumina support was taken out, washed thoroughly with deionized water, and dried in an oven at 110°C.

[0043] PTFE nanoparticles and fatty alcohol polyoxyethylene ether were dispersed in deionized water at a mass ratio of 5:1 to achieve a PTFE mass concentration of 2%. The mixture was stirred for 2 hours and then ultrasonicated for 30 minutes. PVA film former was added and uniformly dispersed to form a casting solution. The casting solution was dip-coated on a ceramic support. After coating, it was dried at room temperature for 12 hours and then further dried in an oven at 70°C for 12 hours. Finally, the composite film was sintered in a muffle furnace at a temperature increase rate of 1°C per minute to 310°C for 10 minutes. Thus, a PTFE / ceramic composite film was obtained, denoted as P film.

[0044] Example 2

[0045] Example 2 differs from Example 1 in that step (3) is as follows: PTFE nanoparticles and fatty alcohol polyoxyethylene ether were dispersed in deionized water at a mass ratio of 5:1 to achieve a PTFE mass concentration of 2%. The mixture was stirred for 2 hours and then ultrasonicated for 30 minutes. Carbon nanotubes and F127 were dispersed in deionized water at a mass ratio of 10:1, and then stirred and ultrasonicated to prepare a CNT suspension. The CNT suspension was then mixed with the PTFE suspension in a specific ratio (the amount of CNT added to the film-forming solution was 1 wt%, 2 wt%, 4 wt%, and 8 wt%), and PVA film former was added at a mass ratio of 1% to uniformly disperse and form a casting solution. The casting solution was dip-coated on a ceramic support. After coating, it was dried at room temperature for 12 hours and then further dried in an oven at 70°C for 12 hours. Finally, the composite film was sintered in a muffle furnace at a temperature increase rate of 1°C per minute to 310°C for 10 minutes. Thus, a PTFE / ceramic composite film was obtained, denoted as P / CNT film.

[0046] Comparative Example 1

[0047] Comparative Example 1 differs from Example 1 in that there is no step (1) and (2), and the original alumina support with an average pore diameter of about 1.7 μm was directly used.

[0048] Characterization

[0049] The morphology and elemental composition of the samples were analyzed using a field emission scanning electron microscope equipped with an X-ray energy dispersive spectrometer. The pore size distribution of the film was determined using the bubble gas transmission method. The crystalline phase of the titanium oxide was determined using an X-ray diffractometer. The surface morphology of the alumina support after hydrothermal treatment was characterized using a three-dimensional laser scanning microscope. The particle size distribution of the suspension was measured using a Zetasizer. The wetting properties of the composite film were evaluated using a contact angle measuring instrument with a droplet volume of 3 μL. The chemical composition of the film was analyzed using X-ray photoelectron spectroscopy. The surface morphology of the composite film was further detected using an atomic force microscope. The scratch resistance, adhesion, and mechanical strength of the composite film were tested using a nano-scratch tester and a multi-functional material surface performance tester.

[0050] The pure water permeability of different supports before and after hydrothermal treatment was measured using Equation 1:

[0051]

[0052] where J is the permeate water flux (L-m -2 ·h -1 ·bar -1 ), V is the volume of permeate collected (L), A is the effective membrane area (m 2 ), ΔP w is the transmembrane pressure (bar), and t is the test time (h).

[0053] The gas permeability of different membranes was further tested using nitrogen at room temperature using Equation 2:

[0054]

[0055] where K is the nitrogen permeability (mol-m -2 ·s -1 ·Pa -1 ), Q is the nitrogen flow rate (L-s -1 ), A is the effective membrane area (m 2 ), ΔP g is the transmembrane pressure (Pa), and V m is the molar volume of gas (L-mol -1 ).

[0056] The water / NMP separation performance of the composite membranes was evaluated using vacuum membrane distillation (VMD) Figure 1 ). Different concentrations of water-NMP solutions were prepared by mixing deionized water and N-methyl-2-pyrrolidone (NMP, 99%, Aladdin, China) at different ratios. In the VMD process, different concentrations of water-NMP solutions were used as feed solutions. The preheated feed solution was delivered to the membrane module using a peristaltic pump. The feed temperature was maintained between 60-80 °C. On the permeate side, water vapor permeated through the membrane and was subsequently condensed into liquid water in a cooling condenser, with the cooling water temperature set at about 4 °C. An electronic balance was used to measure the weight of the collected permeate. The NMP concentration in the permeate was analyzed using high-performance liquid chromatography (HPLC, e2695, Waters, USA) with an evaporative light scattering detector (ELS) (2424, Waters, USA).

[0057] The permeate flux was calculated using Equation 3:

[0058]

[0059] where F is the permeate flux (kg-m -2 ·h-1 ), m is the mass of permeate collected (kg), and A is the effective membrane area (m²). 2 ), where t is the test time (h).

[0060] The rejection rate of N-methylpyrrolidone was calculated using Formula 4:

[0061]

[0062] Among them, R j C is the retention rate. p and C f The NMP concentrations in the permeate and feed solution are respectively.

[0063] Characterization results

[0064] In this embodiment, HCl was chosen to promote the formation of pine needle-like TiO2 structures to repair surface defects on the macroporous ceramic support and to construct an intermediate layer with higher surface roughness and specific surface area. The effect of hydrothermal treatment time on the morphology and surface structure of TiO2 was systematically investigated when the mass ratio of acid to precursor was fixed at 16:1. The untreated ceramic support surface had large pores, mainly due to the loose accumulation of alumina particles. With prolonged hydrothermal treatment, the number of pine needle-like TiO2 deposited on the surface gradually increased. The aggregation morphology of these nanorods evolved from sparse pine needle-like structures to a dense cauliflower-like structure. Simultaneously, the distribution of titanium became more uniform, and its content also increased. As the TiO2 layer grew, the macroporous surface of the support was gradually covered, resulting in a significant reduction in surface pore size and porosity. Interestingly, the evolution of surface roughness followed a non-linear trend, initially increasing and then decreasing with prolonged hydrothermal treatment time. Notably, at a hydrothermal treatment time of 15 hours, the support surface was almost completely covered by pine needle-like TiO2 while still retaining a porous network.

[0065] To further investigate the effect of precursor content under an optimized hydrothermal time of 15 hours, different concentrations of precursors were tested while keeping the acid content constant. Figure 3 Similar to the effect of time, increased precursor concentration promoted the growth of more pine needle-like TiO2 nanostructures. Figure 3 (a, b) The formation of these needle-like structures is attributed to the inherent anisotropic crystal growth of rutile TiO2, where the high surface energy of the (001) facet favors rapid longitudinal growth along the

[001] direction. In contrast, the (110) facet has lower surface energy and tends to evolve into a stable transverse facet. In an acidic environment, Cl- ions preferentially adsorb on the (110) facet, further inhibiting transverse growth and promoting extension along the c-axis.

[0066] XRD analysis shows that the intensity of the characteristic diffraction peak corresponding to the (110) plane first increases and then decreases with increasing precursor concentration, while the intensity of the (101) and (211) peaks continues to increase Figure 3 c). This indicates that the preferential orientation gradually disappears with increasing precursor concentration. At low concentrations, crystal growth dominates over nucleation, resulting in strong

[001] oriented growth and enhanced (110) peak intensity. However, at higher precursor concentrations, the nucleation rate exceeds the crystal growth rate, resulting in a large number of randomly oriented small crystals. When the acid-to-precursor molar ratio is maintained at 16:1, the (110) peak intensity reaches a maximum, and the (110) / (101) peak intensity ratio reaches 2.4. This ratio is much higher than the reference value of about 2 in the standard powder diffraction pattern (PDF) card.

[0067] At this time, the surface roughness factor also reaches a peak of about 3.7, indicating that it has strong anisotropic growth and significant surface structure Figure 3 d) Compared with the ceramic carrier, the water droplet contact angle on the surface of the carrier also increases due to the increase in surface roughness after hydrothermal treatment, but still remains hydrophilic. After hydrothermal treatment, the average pore size of the carrier continues to decrease Figure 3 e) Notably, under the condition of 16:1, the macropore size associated with the bubble point decreases significantly, indicating that the surface coverage is more dense and defect-free. Generally, higher surface roughness is beneficial to mechanical interlocking, thereby improving the interfacial adhesion between the carrier and the membrane layer, while the reduction of macropore defects helps to form a continuous and defect-free film. However, when the precursor concentration is further increased and exceeds the optimal point, the macropore size tends to be stable, and the pure water permeability starts to decrease due to excessive pore blockage. Therefore, considering the film quality and permeability, the acid-to-precursor molar ratio of 16:1 is determined as the optimal condition for constructing a robust TiO2 intermediate layer with ideal structure and function.

[0068] The effect of the intermediate layer constructed by the hydrothermal method on the PTFE membrane layer prepared by dip coating was further studied Figure 4 ) Here, the intermediate layer was synthesized under the optimized acid-to-precursor molar ratio of 16:1 and hydrothermal time of 15 hours. Without the intermediate layer, the original ceramic carrier exhibits a wide pore size distribution, with a pore size much larger than the size of the dispersed PTFE nanoparticles Figure 4 a). Therefore, during the dip coating process, PTFE nanoparticles cannot stay on the surface because the emulsion droplets are quickly and completely absorbed into the macropores of the carrier within about 3 seconds Figure 4 b). This rapid permeation hinders the formation of a continuous and complete membrane layer.

[0069] In contrast, the introduction of a hydrothermally grown TiO2 interlayer significantly altered the interfacial behavior. The complete absorption time of the PTFE emulsion droplets increased to approximately 8 seconds, indicating reduced surface permeability after modification. The thickness of the PTFE film gradually increased with increasing dip-coating time. Figure 4 c) The unique surface features of the carrier and intermediate layer are also gradually obscured. Figure 4 d). When the coating time exceeded 18 seconds, a continuous, pinhole-free porous membrane surface composed of densely packed PTFE nanoparticles was observed. XPS analysis of the membrane surface did not reveal a Ti2p characteristic peak from the interlayer, indicating that the TiO2 structure formed by the hydrothermal method was completely covered by the PTFE membrane layer. In contrast, even after 60 seconds of dip coating, the original support could not support the formation of a continuous membrane because most of the PTFE particles penetrated deep into the macropores and did not remain on the surface. The enhanced membrane formation effect on the interlayer-modified support can be attributed to two key mechanisms: (i) the interlayer significantly reduced the macroporous defects of the support, thereby limiting the penetration of the emulsion; (ii) the needle-like TiO2 nanostructure improved the interfacial bridging and anchoring between PTFE nanoparticles, promoting their stable accumulation and network formation on the surface. Figure 4 e). These synergistic effects contribute to more uniform PTFE deposition and better membrane integrity.

[0070] Notably, the pore structure of the obtained PTFE membrane exhibits a bimodal distribution. The dominant pore size is approximately 100 nm, corresponding to the interparticle voids formed by the stacking of PTFE nanoparticles. However, a second type of larger pore (200–400 nm) also appears, with its proportion increasing with coating time. This is likely attributed to the generation of microcracks induced by shrinkage stress between adjacent PTFE particles during sintering. As the membrane thickens, these stresses become more significant and may overcome the mechanical constraints provided by the underlying ceramic support, leading to crack formation and larger interstitial pores in the membrane structure.

[0071] To mitigate cracking caused by thermal shrinkage during PTFE sintering, one-dimensional carbon nanotubes (CNTs) were introduced into the membrane formulation. On one hand, CNTs act as reinforcements similar to steel bars in reinforced concrete, forming a composite network with PTFE particles to enhance the membrane's resistance to shrinkage stress. On the other hand, CNTs themselves have low surface energy, which helps maintain the overall hydrophobicity of the membrane surface. The effects of CNT addition on the microstructure and pore structure of the PTFE membrane were first investigated. Figure 5 At low CNT concentrations, the membrane surface is smooth and uniform, exhibiting a homogeneous porous structure formed by the stacking of PTFE nanoparticles. When the CNT loading is optimized to 2 wt%, the resulting membrane displays a narrow, single-peak pore size distribution with an average pore size of approximately 240 nm. Figure 5b) The pore size is increased compared to the PTFE membrane without CNTs, thus the corresponding N2permeability is also improved. The nitrogen permeability exceeds 42 x 10 -6 mol m -2 s -1 Pa -1 , which is 1.6 times that of the PTFE membrane without CNTs at 0.1 MPa. However, excessive CNTs addition has a negative effect on the uniformity of the pores. The average pore size gradually increases from 190 nm to 330 nm when the CNT content increases from 1 wt% to 8 wt%. When the loading exceeds 4 wt%, local CNT aggregation leads to the formation of larger surface pores and a broader pore size distribution.

[0072] Overall, the addition of CNTs effectively prevents the formation of microcracks compared to the PTFE membrane without CNTs. This improvement is mainly attributed to the excellent mechanical strength and flexibility of CNTs, which help to buffer the internal stress generated during the sintering process of PTFE particles. In addition, the high aspect ratio and entangled network structure of CNTs help to redistribute stress and bridge cracks, thereby enhancing the mechanical integrity of the membrane Figure 5 e) Considering the integrity and pore uniformity of the membrane, a CNT concentration of about 2 wt% is determined to be the optimal value. In the XPS analysis of the membrane surface, no Ti2p characteristic peak from the intermediate layer is found, indicating that the TiO2structure formed by the hydrothermal method has been completely covered. Due to the low surface energy of both PTFE and CNTs, the water droplet contact angle on the surface of the resulting composite membrane is about 142°, showing its good hydrophobicity. The contact angle increases slightly compared to the PTFE membrane without CNTs. This is mainly because the introduction of CNTs slightly increases the roughness of the PTFE membrane surface.

[0073] The cross-sectional microstructure and mechanical bonding properties of the composite membrane are further investigated Figure 6 ). The membrane exhibits a clear three-layer structure consisting of an alumina support, a TiO2intermediate layer, and a top PTFE membrane layer. The corresponding elemental distribution maps reveal the clear distribution of Al, Ti, and F within the layers Figure 6 a) Notably, partial overlap of adjacent element signals is observed in the interfacial region, indicating strong interfacial bonding and mechanical interlocking between the layers. Its mechanical properties are evaluated by nano-scratch testing under increasing load Figure 6b) For the PTFE film without CNTs, the indentation depth of the probe steadily increased with displacement, reaching about 10 pm at a load of 40 mN. This indentation depth is even close to the entire film thickness. In contrast, the CNT-reinforced PTFE film exhibited a significantly reduced indentation depth. At a load of 20 mN, the depth only reached 3.5 pm, and even at 80 mN, the depth was still below 4 pm, indicating its strong resistance to deformation. Microscopically, both the length and depth of the scratch were significantly reduced after the introduction of CNTs, further confirming the enhancement of mechanical properties. Meanwhile, we also evaluated the friction behavior Figure 6 c) After the hydrothermal treatment, the roughness increased due to the pine needle-like TiO2interlayer, leading to an increase in the friction coefficient. After coating with PTFE, the friction coefficient decreased significantly because the PTFE surface effectively masked the underlying rough features. The low surface energy and molecular fluidity of the PTFE chains facilitated the interfacial sliding, reducing the friction between the SiC ball and the film surface. However, after the addition of CNTs, a slight increase in the friction coefficient was observed. This is mainly because the CNTs partially exposed on the film surface disrupted the smooth, low-energy PTFE interface. CNTs have higher surface energy and greater stiffness, introducing nanoscale roughness and increasing the interfacial shear resistance. In addition, the presence of CNTs can reduce the local fluidity of PTFE chains near the surface, slightly weakening the self-lubricating effect of the polymer. Interestingly, the friction coefficients of both the ceramic-supported PTFE film and the CNT-PTFE composite film were slightly lower than that of the commercial PTFE film. This improvement is attributed to the superior mechanical stiffness of the ceramic-supported film, which is less prone to deformation during the friction test. Therefore, the contact area between the SiC ball and the film surface is reduced. The CNT-PTFE composite film also exhibited excellent wear resistance. In the long-time friction test of over 800 minutes, the friction coefficient stabilized at about 0.75, indicating that no film delamination or wear-through occurred, further confirming the superior mechanical strength of the composite system Figure 6 d).

[0074] In this work, we employed the CNT-PTFE composite film to treat NMP-containing wastewater through the sweep gas-assisted membrane distillation (SR-MD) process. This process takes advantage of the volatility difference between water and NMP: water preferentially evaporates and passes through the hydrophobic membrane in vapor form, while the non-volatile NMP is trapped and concentrated on the feed side. At a feed concentration of 10 wt% NMP, we investigated the effect of feed temperature on the membrane performance Figure 7 a) The permeation flux increased with increasing temperature, mainly due to the exponential growth of the water saturation vapor pressure, thereby enhancing the driving force for vapor transport in membrane distillation. However, a slight decrease in NMP rejection was observed with increasing temperature. This is attributed to the different growth rates of the saturation vapor partial pressures of NMP and water, which do not change proportionally with temperature.

[0075] To further elucidate this behavior, we simulated the vapor-liquid equilibrium of the water-NMP system using Aspen Plus software and the UNIQUAC thermodynamic model Figure 7 b) The saturated vapor pressures of both water and NMP can be well fitted by the Antoine equation with a correlation coefficient over 0.99. Although the absolute vapor pressure of NMP is much lower than that of water at all temperatures, its relative growth rate is slightly higher. Assuming ideal vapor transport in the membrane pores and complete condensation on the permeate side, we calculated the NMP concentration in the permeate and the rejection rate from the vapor pressure data Figure 7 c) The calculated values agree well with the experimental results, confirming the slight decrease in rejection performance at high temperatures.

[0076] The effect of different NMP concentrations on the membrane performance was investigated at a constant feed temperature of 70 °C Figure 7 d) With the increase of NMP concentration in the feed, both the permeation flux and NMP rejection rate slightly decreased. This is mainly due to the decrease of water vapor partial pressure and the corresponding increase of NMP vapor pressure, which changes the driving force and mass transfer characteristics of the system. Overall, the CNT-PTFE composite membrane maintains high water flux and NMP rejection rate at a wide range of operating temperatures and NMP concentrations. In a 1200 min continuous operation using a feed containing 10 wt% NMP Figure 7 e,f) The permeation flux stabilized at about 30 kg-m -2 h-1, while the NMP rejection rate consistently exceeded 96%. Even at a high feed concentration of 50 wt%, the membrane still maintains stable performance with a flux higher than 25 kg-m -2 h-1 and a rejection rate over 95%. These results confirm the excellent operational stability and separation performance of the CNT-PTFE composite membrane in treating NMP-containing wastewater in the SR-MD process.

Claims

1. A PTFE / ceramic membrane having a pine needle structure for enhancing interfacial bonding, the PTFE / ceramic membrane comprising a ceramic membrane as a support and a PTFE separation membrane as a separation layer, characterized in that, The ceramic membrane and the PTFE separation membrane have a pine needle-shaped TiO2 intermediate layer.

2. The PTFE / ceramic membrane according to claim 1, wherein, The ceramic membrane has an average pore size of 1-3 μm, preferably 1.5-2.5 μm.

3. The PTFE / ceramic membrane according to claim 1, wherein, The TiO2 intermediate layer has a surface roughness factor of 2-5, preferably 3-4 factors.

4. The PTFE / ceramic membrane of claim 1, wherein, The PTFE separation membrane contains carbon nanotubes, and the loading amount of the carbon nanotubes in the PTFE separation membrane is 1-8 wt%.

5. A method of making the PTFE / ceramic membrane of claim 1 having a needle-like structure reinforced interfacial bonding, characterized by, The method comprises the following steps: An HCl: H2O solution with a molar ratio of 1:6-1:13 is prepared, stirred, then tetrabutyl titanate is added, and stirring is continued for 10-60 min to form a mixed solution; The ceramic membrane is immersed in the mixed solution and reacted at 120-180℃ for 5-25 h to form a ceramic membrane with an intermediate layer; PTFE particles, water, a dispersing agent, and a film-forming agent are mixed to form a film-forming solution, and the film-forming solution is coated on the ceramic membrane, and after drying and calcination, the PTFE / ceramic membrane is formed.

6. The method of claim 5, wherein, In step (1), the molar ratio of HCl to tetrabutyl titanate is 5-20:

1.

7. The method of claim 5, wherein, In step (2), the ceramic membrane needs to be pretreated before being immersed in the mixed solution, and the pretreatment is to ultrasonically clean the ceramic membrane with ethanol and deionized water in sequence; the reaction temperature is preferably 140-160℃, and the reaction time is preferably 12-17 h.

8. The method of claim 5, wherein, In step (3), the film-forming solution does not contain / contains carbon nanotubes, the dispersing agent is one or more of fatty alcohol polyoxyethylene ether, polyacrylamide, sodium carboxymethyl cellulose, sodium dodecylbenzenesulfonate, fluorinated alkyl ethoxyl alcohol ether, sodium butylphenyl sulfonate, sodium hydroxyethyl sulfate, sodium dodecyl sulfate, polyvinylpyrrolidone, polyacrylic acid salt, and polyethylene carboxylic acid, and the film-forming agent is one or more of polyvinyl alcohol, polyvinyl butyral, hydroxypropyl methyl cellulose, ethyl cellulose, polyethylene glycol, polypropylene glycol, and glycerol.

9. The method of claim 5, wherein, In step (3), the coating time is 10-60 s; the drying is carried out at room temperature for 8-16 hours, and then further drying is carried out in an oven at 60-80℃ for 8-16 hours; the calcination is to sinter the composite membrane at a temperature rising rate of 0.5-2℃ / min to 280-340℃ for 5-60 min; the sintering temperature is preferably 300-310℃, and the sintering time is preferably 10-15 min.

10. Application of the PTFE / ceramic membrane with a pine needle-shaped structure and enhanced interface bonding in water / N-methyl-2-pyrrolidone system separation.

Images