An asymmetric regenerated cellulose fiber virus removing filter membrane and a preparation process thereof

Abstract

The application relates to an asymmetric regenerated cellulose virus-removing filter membrane and a preparation process thereof, the filter membrane comprising a porous main body, the porous main body comprising a pre-filter layer, a pore size transition layer and a separation layer, the PMI average pore size of the filter membrane being 15-40 nm; the thickness of the pre-filter layer being 20-80 mu m; the pore size transition layer having a pore size variation coefficient S (0～0.1) / (0.9～1) not more than 5, the thickness of the pore size transition layer being not more than 20 mu m; the pore size variation coefficient The filter membrane provided by the application uses a hydrophilic cellulose material as a film-forming material, has a high protein yield, has a good interception effect on viruses with small and medium sizes, has a large pollution capacity due to the large thickness of the pre-filter, the pore size of the pore size transition layer does not change suddenly, and is not prone to local blockage due to concentrated interception, and thus has a large load, and the substantially symmetrical structure of the separation layer makes the filter membrane have a large flux.

Description

Technical Field

[0001] This application relates to the field of membrane separation technology, and in particular to an asymmetric regenerated cellulose virus removal filter membrane and its preparation process. Background Technology

[0002] Biological agents generally refer to those made from various microorganisms (such as bacteria, bacteriophages, rickettsiae, etc.), microbial metabolites, parasites, human or animal blood or tissues, etc., using modern biotechnology and chemical methods, which have good diagnostic, preventive and therapeutic effects on specific infectious diseases and immune diseases.

[0003] Biologics have experienced rapid development over the past few decades due to their unique targeting, high efficacy, and low side effects. With this rapid development, increasing the concentration of active substances (proteins) in biologics is currently a recognized direction for further development. High-concentration biologics are generally administered via subcutaneous injection. Compared to intravenous injection, subcutaneous injection offers shorter treatment time and less patient discomfort. Furthermore, the reduced volume of high-concentration biologics significantly reduces production and transportation costs.

[0004] In the production of biopharmaceuticals, viruses can be introduced into both raw materials and various production processes. Both the latest Chinese Pharmacopoeia and ICH Q5A, "Viral safety assessment of biological products," set forth clear requirements for the viral safety of biopharmaceuticals. Viral safety assessments and test reports directly impact drug application and approval results. Therefore, all biopharmaceutical companies must incorporate virus clearance and / or virus inactivation steps when producing various biopharmaceuticals to ensure that the biopharmaceuticals are free of viral safety issues.

[0005] Membrane separation technology is widely used in the production of various biological agents due to its high separation efficiency, low energy consumption, small footprint, and especially its ability to avoid denaturing active protein substances when removing viruses. It is used to filter out viruses of various sizes in biological agents and improve the viral safety of biological agents. The core of membrane separation technology is the filter membrane.

[0006] For example, Chinese invention patent application CN113842792A discloses an asymmetric PES filter membrane for virus removal. This PES filter membrane comprises a main body, including a pre-filtration layer and a separation layer for virus interception. The other side of the pre-filtration layer and the other side of the separation layer are transitioned by continuous fibers. This PES filter membrane exhibits good virus interception efficiency (LRV > 4). However, the inherently poor hydrophilicity of PES material (even after hydrophilic modification, its hydrophilicity remains poor) determines its high adsorption capacity for proteins. If the protein concentration in the biological agent is high, the proteins adsorbed by the filter membrane can easily cause membrane clogging, rapid flux decay, and a decrease in the concentration of the biological agent, affecting its quality.

[0007] For example, US Patent 20200238221A1 filed by Sartorius discloses a porous monolayer polymer membrane. At least one major surface of this polymer membrane has a surface porosity of at least 40%, and the total porosity of the polymer membrane is 0.8 to 1.4 times the surface porosity of at least 40%; furthermore, the polymer membrane has an asymmetry factor of 1.5 to 10. This filter membrane is primarily used to filter viruses, proteins, or macromolecules, and it can only retain large particles of hundreds of nanometers. Although this filter membrane, due to its large pore size, is less prone to retaining smaller proteins and is less susceptible to clogging by viruses, resulting in a slower flux decay rate and higher loading capacity, its large pore size means it has no filtering effect on small-sized viruses, such as those with a diameter of 20 nm. Therefore, it cannot achieve the desired retention effect for small-sized viruses.

[0008] For example, Chinese patent CN1759924B, filed by Milippel, discloses a multilayer composite ultrafiltration membrane; the composite ultrafiltration membrane includes at least one first porous membrane layer having a first side and an equivalent second side, and at least one second porous membrane layer having an equivalent first side and a second side, wherein the connection between the first layer and the second layer is superimposed and has a porosity connection transition region from the equivalent first side of the second layer to the equivalent second side of the first layer, wherein at least one of the layers is an asymmetric ultrafiltration membrane.

[0009] At least one of the ultrafiltration membranes mentioned above is an asymmetric ultrafiltration membrane. An asymmetric ultrafiltration membrane means that the ultrafiltration membrane layer has a macropore side with a larger pore size and a micropore side with a smaller pore size. In order to reduce the risk of virus leakage, the micropore side of the asymmetric ultrafiltration membrane needs to have a good virus retention effect (if the macropore side has a good retention effect, the micropore side with a smaller pore size will likely lead to a significant decrease in the membrane flux). In addition, the virus retention position of the micropore side should not be too close to the second surface of the asymmetric ultrafiltration membrane (otherwise the risk of virus leakage from the second surface will be greatly increased). However, because the ultrafiltration membrane is asymmetric, the pore size near the second surface on the small pore side should be smaller than that near the first surface. Furthermore, the membrane-forming material in all embodiments of this patent is polyethersulfone (PES). The characteristics of PES mean that even after hydrophilic modification, its hydrophilicity remains poor. The combination of a poorly hydrophilic membrane-forming material and the small pore size near the second surface of the asymmetric membrane likely results in a high protein adsorption rate but a low protein yield. In the biomedical field, the high cost of proteins makes a low protein yield unacceptable. Furthermore, the asymmetric nature of the ultrafiltration membrane means that viruses and other pathogens are likely to be trapped in a small area of ​​the membrane thickness, potentially leading to membrane blockage and rapid flux decay, thus resulting in a low loading capacity. Therefore, while this membrane may have good virus retention, its loading capacity and protein yield are often low.

[0010] Furthermore, the preparation process of the aforementioned ultrafiltration membrane mainly involves casting two solutions layered on a carrier surface, both having a lower critical solution temperature (LCST) and the ultrafiltration layer having a higher LCST, or the ultrafiltration layer having no LCST or no measurable LCST, while the microporous layer solution has an LCST; heating the cast multilayer liquid sheet to a predetermined temperature higher than the LCST of the microporous layer and lower than the LCST of the ultrafiltration layer, and then immersing it in a precipitation bath to form a wet multilayer ultrafiltration membrane; or, supplying each feed outlet with different polymer solutions and coating the solutions onto the moving carrier surface to form a multilayer coating layer on the carrier, distributing the multilayers, and before coating each successive layer, only partial phase separation occurs in the previous layer, and performing phase separation treatment on the multilayers to complete phase separation and form a wet multilayer ultrafiltration membrane. In other words, the phase separation mechanism of the polymer solution is that it is a homogeneous liquid at low temperatures. As the temperature rises, it reaches the LCST (lowest temperature limit) of the microporous layer, causing the materials in the microporous layer to become immiscible and forming phase separation solidification. The ultrafiltration layer, on the other hand, undergoes phase separation solidification through a subsequent precipitation bath. The phase separation mechanisms of the ultrafiltration layer and the microporous layer are different. The microporous layer rapidly reaches above the LCST temperature under the action of the carrier, thus achieving phase separation solidification, while the ultrafiltration layer undergoes phase separation solidification under the action of the precipitation bath after the microporous layer has completely solidified.

[0011] This means that, in order to achieve "high-temperature-induced microporous layer phase separation" and "precipitation bath-induced ultrafiltration layer phase separation" in the above-mentioned ultrafiltration membrane preparation process, the microporous layer solution must first be cast onto the carrier, and then the ultrafiltration layer solution must be cast onto the microporous layer solution. Otherwise, the ultrafiltration layer solution will not be able to directly contact the precipitation bath, and the carrier must first heat the ultrafiltration layer solution before further heating the microporous layer solution. This will result in the inability to form a multilayer ultrafiltration membrane. This preparation process has many limitations, and during the preparation process, the ultrafiltration membrane layer remains in a liquid film state for a long time, which may lead to a decrease in the uniformity of the ultrafiltration membrane layer, and consequently, a decrease in the uniformity of the pore structure after the phase separation and solidification of the ultrafiltration membrane layer. This is also one of the defects of this preparation process. In addition, the different phase separation mechanisms of the two casting solutions result in certain differences in their pore structure and fiber structure, which is reflected in the relatively obvious boundary line in the cross-sectional SEM of the above-mentioned ultrafiltration membrane.

[0012] For various biological agents with proteins, peptides and their derivatives as the main active substances, filter membranes with high virus rejection efficiency often have a fast flux decay rate and low protein yield, while filter membranes with slow flux decay rate often fail to achieve the required virus rejection effect. This is a problem that needs to be solved but is difficult to solve for virus removal filter membranes used in the virus removal process of various biological agents. Summary of the Invention

[0013] This application provides an asymmetric regenerated cellulose virus-removing filter membrane and its preparation process. The filter membrane uses hydrophilic cellulose as the film-forming material, resulting in a high protein yield. The average pore size (PMI) of the filter membrane is 15–40 nm, thus exhibiting good retention of small to medium-sized viruses. The pre-filtration layer is relatively thick, thus providing good retention of large particles and a large contaminant capacity. The pore size transition layer of the filter membrane is thin and has a small pore size variation coefficient, preventing abrupt changes in the pore size and reducing the risk of concentrated retention, local blockage, and rapid flux decay due to abrupt pore size changes, resulting in a large loading capacity. The basic symmetrical structure of the separation layer ensures good retention of small to medium-sized viruses while avoiding blockage and rapid flux decay due to the absence of micropores. The absence of micropores also contributes to the high flux of the filter membrane. Therefore, the filter membrane in this application not only has good virus retention but also high flux, loading capacity, and protein yield.

[0014] Firstly, the asymmetric regenerated cellulose virus removal filter membrane provided in this application adopts the following technical solution:

[0015] An asymmetric regenerated cellulose virus-removing filter membrane includes a porous body with non-directional tortuous pathways. One side of the porous body is a liquid inlet surface, and the other side is a liquid outlet surface. The average pore size measured by SEM on the liquid inlet surface is larger than that measured by SEM on the liquid outlet surface. The porous body includes a pre-filtration layer, a pore size transition layer, and a separation layer sequentially from the liquid inlet surface to the liquid outlet surface. One side of the pre-filtration layer is the liquid inlet surface, and one side of the separation layer is the liquid outlet surface.

[0016] The average pore size of the filter membrane is 15–40 nm.

[0017] The thickness of the pre-filter layer is 20–80 μm;

[0018] The pore size variation coefficient S of the pore size transition layer (0～0.1) / (0.9～1) The pore size transition layer is not greater than 5, and its thickness is not greater than 20 μm.

[0019] The separation layer has a basically symmetrical structure;

[0020] The pore size transition layer is considered to have a film thickness of 0 on the side near the liquid inlet surface and a film thickness of 1 on the side near the liquid outlet surface.

[0021] Aperture variation coefficient S x / y = Average pore diameter measured by SEM in region x of the pore transition layer / Average pore diameter measured by SEM in region y of the pore transition layer.

[0022] By adopting the above technical solution, the separation layer in this application refers to the region in the thickness direction of the filter membrane that has a retention effect on small and medium-sized viruses, while the pre-filtration layer refers to the region in the thickness direction of the filter membrane that has a retention effect on medium and large-sized viruses, and the pore size transition layer refers to the region between the pre-filtration layer and the separation layer.

[0023] The PMI (Polymer Surface Area) of a filter membrane can reflect its retention capacity for particles of a corresponding size to a certain extent. Filter membranes with different surface morphologies (such as different pore sizes, fiber structures, and thicknesses) have different PMIs. Different applications require filter membranes with different virus retention capacities, and in such cases, filter membranes with different PMIs can be selected based on actual needs. When filtering large particles or viruses, filter membranes with small PMIs are not suitable because the small pores are easily clogged, leading to a rapid decrease in flux. Conversely, when filtering small particles or viruses, filter membranes with large PMIs are not suitable because small particles and viruses can easily penetrate, leading to a risk of virus leakage. Therefore, to ensure that the filter membrane has a good retention effect on the viruses to be filtered, it is necessary to select a filter membrane with an appropriate PMI.

[0024] For example, when the average pore size (PMI) of the filter membrane is 15–25 nm, the membrane often exhibits good retention of small-sized viruses (such as PP7 bacteriophage with a particle size of approximately 25 nm) with a diameter of only about 20–30 nm. Conversely, when the average pore size (PMI) of the filter membrane is 25–40 nm, it exhibits good retention of medium-sized viruses (such as hepatitis B virus with a particle size of approximately 42 nm) with a diameter of approximately 30–45 nm. The filter membrane in this application specifically defines an average pore size (PMI) of 15–40 nm. Therefore, it exhibits good retention of small to medium-sized particles and viruses (such as those with a particle size of approximately 20–45 nm), and the required average pore size (PMI) can be selected according to different application scenarios.

[0025] It is understandable that larger viruses or particulate impurities (such as HIV viruses with a particle size of 80nm or larger) are generally filtered out in the previous process, leaving only a very small amount of residue, so there is no need to specifically treat large viruses or particulate impurities; however, medium and large viruses or particulate impurities are not completely filtered out in the previous process, so there may still be some medium and large viruses or particulate impurities in the feed solution.

[0026] Furthermore, as mentioned above, medium to large-sized viruses or particulate impurities often remain in the feed solution. The separation layer of the filter membrane mainly serves to trap small to medium-sized viruses or particulate impurities. If medium to large-sized viruses or particulate impurities leak into the separation layer, they are likely to clog the smaller pore structures in the separation layer. Once the smaller pore structures in the separation layer are clogged, not only will the filter membrane flux decrease rapidly, but the risk of small to medium-sized viruses leaking through the larger pore structures in the separation layer will also increase. The filter membrane in this application has a relatively thick pre-filtration layer (not less than 20 μm), which ensures that the pre-filtration layer has a sufficiently large contaminant-holding space to trap medium-to-large-sized particles or viruses in the feed solution. Of course, the thickness of the pre-filtration layer should not be too large (e.g., greater than 80 μm). This is because, although further increasing the thickness of the pre-filtration layer can improve the filter membrane's contaminant-holding capacity for medium-to-large-sized particles or viruses to some extent, the filtration method of virus-free filter membranes is generally dead-end filtration. During the filtration process, external pressure is the main driving force for filtration. Therefore, the filter membrane is subjected to a large pressure from the feed solution. The pore structure of the pre-filtration layer has a large pore size. Combined with the inherently soft texture of cellulose material, the pore structure is prone to collapse under the pressure of the feed solution. If the thickness of the pre-filtration layer is too large, the possibility of structural collapse will be greatly increased (large pore structures are more prone to defects, and an increase in the thickness of the pre-filtration layer means an increase in the amount of large pores, which also increases the possibility of defects). Once the pore structure of the pre-filtration layer collapses, both the ability to guide the feed solution and the contaminant-holding capacity will be greatly reduced.

[0027] Furthermore, this application specifically controls the thickness and pore size variation coefficient of the pore size transition layer to ensure that the filter membrane has a large loading capacity and flux. This is because if the thickness of the pore size transition layer is too large, it indicates that there are areas in the filter membrane that do not effectively retain medium-to-large-sized viruses (not that they have no retention capacity), nor can they retain small-to-medium-sized viruses. Although increasing the thickness of these areas can reduce the impact of medium-to-large-sized viruses on the separation layer, an excessively thick pore size transition layer often means excessive resistance to the feed liquid, resulting in a lower flux of the filter membrane. In addition, since the overall pore size variation coefficient of the pore size transition layer is no greater than 5, it indicates that the pore size change of the pore size transition layer is not a rapid abrupt change, but a slow gradual change. Because there are no abrupt changes in the pore size of the pore size transition layer, it is less likely to cause local blockage of the filter membrane due to concentrated local retention, thereby causing a rapid decrease in the flux of the filter membrane.

[0028] In this application, the separation layer of the filter membrane is specifically designed with a basically symmetrical structure (basically symmetrical means that the average pore size change gradient measured by SEM is no greater than 5 nm / μm, not that the pore size is completely constant). This ensures that the separation layer has good virus retention capacity while also ensuring higher contaminant holding capacity and flux. This is because a separation layer with a large degree of asymmetry retains viruses by continuously reducing the pore size. When the pore size is reduced to a certain extent, it will have a strong retention effect on viruses of the corresponding size. Therefore, for a separation layer with a large degree of asymmetry, viruses tend to be retained in a relatively small thickness range. Combined with the small pore size at this point, the large amount of virus retention can easily lead to local blockage of the filter membrane, resulting in a rapid decrease in flux.

[0029] A nearly symmetrical separation layer tends to have a wider thickness range for virus retention. This is because the relatively small pore size of the separation layer, combined with its inherent thickness, results in a continuous stacking of pores along the thickness direction. The retention effect is smaller than the actual size of the pore structure. For example, the projection surface of the upper pore structure may resemble the fibrous structure of the lower pores, which separates the upper pore structure, thus improving the retention of smaller particles. This stacking of pore structures along the membrane thickness direction ensures good retention of even smaller particles. A more uniform distribution of viruses makes the separation layer less prone to clogging, resulting in slower flux decay. Furthermore, compared to separation layers with greater asymmetry, a nearly symmetrical separation layer does not introduce extremely small pore structures, and the rate of decrease in contaminant capacity and flux is much greater than the rate of decrease in pore size. For example, if we consider the pore structure as a virtual sphere, when the pore size is reduced to half, the volume of the virtual sphere will be reduced to one-eighth. This means that the pore size that the pore structure can accommodate small viruses decreases significantly, and the flow channel space through which the feed liquid passes also decreases significantly (even considering only the cross-sectional area of ​​the flow channel, it will be reduced to one-quarter). Therefore, a slight reduction in the pore size will lead to a significant decrease in the pore structure's contaminant holding capacity and flux, while the effect of a slight reduction in the pore structure on the virus retention effect is likely not significant (i.e., there is a diminishing marginal effect).

[0030] Furthermore, the filter membrane provided in this application uses cellulose-based materials as the film-forming raw material, thus exhibiting good hydrophilicity and low protein adsorption, thereby improving protein yield. Combined with the basic symmetrical structure of the separation layer (the separation layer with small pores is the easiest part of the filter membrane to adsorb proteins), the separation layer does not have extremely small pores, making it even less prone to protein adsorption, thus further improving protein yield.

[0031] In summary, the filter membrane in this application exhibits both high flux and high loading capacity due to its relatively thick pre-filtration layer (less prone to clogging), thin pore transition layer without abrupt pore size changes (less prone to clogging, low resistance), and essentially symmetrical separation layer (less prone to clogging, low resistance). Furthermore, its suitable PMI average pore size results in high virus retention for viruses of corresponding sizes. The essentially symmetrical separation layer (no small pores, no abrupt changes) combined with the use of hydrophilic fiber materials as the film-forming substance leads to a high protein yield (the protein yield of the filter membranes in this application can reach over 98%). Therefore, the filter membrane in this application combines excellent virus retention, high flux, and high loading capacity.

[0032] It is understandable that the so-called non-directional tortuous path refers to a randomly oriented groove structure and / or a discretely distributed pore structure, and each non-directional tortuous path is interconnected. During filtration, the liquid flows within the tortuous pore structure, and impurities in the liquid are intercepted through sieving, adsorption and other methods.

[0033] It should be noted that although the filter membrane in this application has a multilayer structure, it is not obtained by compositing multiple membranes together, and the filter membrane in this application does not have obvious dividing lines, so layer separation is not likely to occur. In addition, the pressure for colloidal gold retention test and virus challenge test in this application is 30 psi. Since the retention area of ​​the filter membrane for particles of different sizes may change under different test pressures, and the risk of virus leakage will also be different, it is necessary to limit the test pressure.

[0034] The measurement of various surface morphology parameters of filter membranes (such as fiber diameter and pore size) can be achieved by characterizing the membrane structure using a scanning electron microscope (SEM), followed by measurement and calculation using computer software (such as Matlab, NIS-Elements, etc.) or manually. During membrane fabrication, in the direction perpendicular to the membrane thickness (planar if the membrane is flat; perpendicular to the radius if it is hollow fiber), its characteristics, such as pore size distribution, are generally uniform and consistent. Therefore, the average pore size of the entire plane can be reflected by measuring the average pore size of a portion of the corresponding plane. In actual measurement, the membrane surface (or cross-section) can be characterized using an electron microscope to obtain the corresponding SEM image, and a certain area, such as 1 μm, can be selected. 2 (1μm x 1μm) or 25μm 2 (5μm by 5μm), the specific area size depends on the actual situation. Then, use appropriate computer software or manual measurement to determine the diameter of all holes in the area, and then calculate to obtain the average diameter of the area. Of course, those skilled in the art can also obtain the above parameters through other measurement methods. The above measurement methods are for reference only.

[0035] Optionally, the pore size variation coefficient S of the pore size transition layer (0～0.1) / (0.4～0.5) The value is 1.1 to 1.6; and / or, the pore size variation coefficient S of the pore size transition layer is... (0.5～0.6) / (0.9～1) It ranges from 1.1 to 1.6.

[0036] By adopting the above technical solution, the pore size variation coefficient of the first half of the pore size transition layer and the pore size variation coefficient of the second half of the pore size transition layer do not exceed 1.6. This indicates that there is no abrupt change in pore size in the first half and the second half of the pore size transition layer. This greatly reduces the possibility of blockage caused by concentrated interception due to abrupt change in pore size in the pore size transition layer.

[0037] Optionally, the pore size variation coefficient S of the pore size transition layer (0～0.1) / (0.4～0.5) The pore size variation coefficient S of the pore size transition layer (0.5～0.6) / (0.9～1) The ratio is 0.8 to 1.2.

[0038] By adopting the above technical solution, the pore size variation coefficients of the first half (i.e., the region with a pore size transition layer thickness of 0–0.5 mm) and the second half (i.e., the region with a pore size transition layer thickness of 0.5–1 mm) of the pore size transition layer are both 1.1–1.6, and the ratio between the two is only 0.8–1.2. This indicates that the phase separation dynamics of the entire pore size transition layer do not change significantly during the phase separation process. This further demonstrates that the first and second halves of the pore size transition layer are less prone to abrupt changes in pore size. This reduces the possibility of pore size transition layer blockage and ensures that the filter membrane has a high loading capacity.

[0039] Optionally, the difference between the average SEM-measured pore size of the 0.3–0.5 region of the pore size transition layer and the average SEM-measured pore size of the 0.5–0.7 region of the pore size transition layer is 30–100 nm.

[0040] By adopting the above technical solution, the pore size of the pore transition layer varies gradient along the thickness direction, with the 0.3–0.7 mm thickness region being the relatively intermediate area of ​​the pore transition layer. Since the entrapment portion and the latter half of the pore transition layer both have relatively small pore size variation coefficients, and considering that the total thickness of the pore transition layer does not exceed 20 μm and the pore size variation in the intermediate region does not exceed 100 nm, it indicates that the pore size variation coefficients and absolute values ​​of the pore size variations in the first, middle, and latter halves of the pore transition layer are all small. Therefore, the overall structure of the pore transition layer does not exhibit abrupt changes in pore size, naturally reducing the likelihood of impurity concentration and retention due to abrupt changes in pore size. This lowers the possibility of clogging of the pore transition layer leading to rapid decline in filter membrane flux.

[0041] Optionally, the average pore size of the pore size transition layer measured by SEM is 100-300 nm, and the average pore size of the pore size transition layer measured by SEM is smaller than the average pore size of the pre-filter layer and smaller than the average pore size of the separation layer measured by SEM.

[0042] By adopting the above technical solution, the average pore size measured by SEM in the pore size transition layer should not be too large or too small. If the average pore size measured by SEM in the pore size transition layer is too large, it indicates that the average pore size measured by SEM in the pre-filtration layer is even larger. Although the stacking of pore structures can still trap large particulate impurities, the risk of leakage of large particulate impurities is higher. Since the average pore size measured by SEM in the pore size transition layer is also relatively large, the probability of large particulate impurities penetrating both the pre-filtration layer and the pore size transition layer is greatly increased. Large particulate impurities are likely to clog the small-diameter pore structures in the separation layer, leading to a rapid decline in membrane flux and an increased risk of virus leakage. Conversely, if the average pore size measured by SEM in the pore size transition layer is too small, it indicates that the average pore size measured by SEM in the separation layer is even smaller. Although reducing the pore size of the separation layer can further improve the LRV of the membrane, it greatly increases the resistance to the feed liquid, leading to a decrease in membrane flux. In addition, reducing the pore size of the separation layer will also lead to a decrease in the dirt-holding capacity of the separation layer, thereby increasing the rate of membrane flux decline.

[0043] Optionally, the SEM measurement fiber diameter in the 0-0.5 region of the pore size transition layer is 20-50 nm; the SEM measurement fiber diameter in the 0.5-1 region of the pore size transition layer is 20-45 nm.

[0044] By adopting the above technical solution, the pore size gradient of the pore size transition layer changes, and the pore size of the first half of the pore size transition layer is larger than that of the second half. If the SEM measurement of the fiber diameter of the fiber structure in the first half of the pore size transition layer is less than 20 nm, it is very likely that the pore structure in the first half of the pore size transition layer will not be well supported. If the SEM measurement of the fiber diameter of the fiber structure is greater than 50 nm, although the pore structure can be well supported, the resistance generated by the fiber structure as a solid part to the liquid is too large, which can easily lead to a decrease in throughput. Similarly, for the pore structure in the second half of the pore size transition layer, the SEM measurement of the fiber diameter of the fiber structure needs to be controlled between 20 and 45 nm.

[0045] Optionally, the difference between the SEM measured fiber diameter in the 0.3–0.5 region of the pore size transition layer and the SEM measured fiber diameter in the 0.5–0.7 region of the pore size transition layer is no greater than 10 nm.

[0046] By adopting the above technical solution, the difference in fiber diameter measured by SEM in the middle part of the filter membrane pore size transition layer in this application is small. The difference between the fiber diameter measured by SEM in the 0.3-0.5 pore size transition layer and the fiber diameter measured by SEM in the 0.5-0.7 pore size transition layer is no more than 10 nm. This means that not only does the pore size of the pore structure transition in a relatively gentle manner, but the diameter of the fiber structure in the pore size transition layer also transitions in a relatively gentle manner. This is reflected in the SEM image of the filter membrane as the absence of obvious boundary lines or delamination phenomena in the cross-section.

[0047] When observing SEM images of a filter membrane cross-section, its surface morphology is a major factor affecting the visual quality of the image. Factors such as pore structure and fiber structure are crucial. If the pore size and fiber diameter differ significantly between the first and second halves of the pore transition layer, a clear boundary will appear in the SEM image. Although the filter membrane in this application exhibits a slight difference in pore size between the first and second halves of the pore transition layer, their fiber structures are nearly identical. This means that during filtration, the feed liquid will not experience abnormal resistance due to abrupt structural changes in the solid portion (i.e., fiber structure). The feed liquid flows smoothly and is less prone to concentrated retention in the pore transition layer, which significantly improves the membrane's loading capacity.

[0048] Optionally, the SEM measured average pore size of the pre-filter layer is 200–800 nm; the SEM measured average pore size of the separation layer is 30–80 nm; and the thickness of the separation layer is 20–50 μm.

[0049] By employing the above technical solution, the pre-filter layer primarily functions to trap large particles in the feed liquid. Therefore, the pre-filter layer needs to have a large dirt-holding capacity and excellent trapping effect for large particles. If the average pore size of the pre-filter layer measured by SEM is too large (e.g., greater than 800 nm), although the pre-filter layer with a larger pore structure generally has a greater dirt-holding capacity, its trapping effect for large particles is insufficient. Once large particles leak from the pre-filter layer, causing blockage of the pore transition layer or separation layer, it will lead to a rapid decline in the membrane flux. Furthermore, an excessively large average pore size measured by SEM also means a reduction in the pressure resistance of the pre-filter layer. Once the pre-filter layer collapses under the pressure of the feed liquid, both the membrane flux and loading capacity will decrease. Therefore, although increasing the average pore size measured by SEM of the pre-filter layer within a certain range is beneficial for increasing the membrane loading capacity and reducing the rate of flux decline, the average pore size measured by SEM of the pre-filter layer should not be too large, otherwise it will lead to a decrease in the membrane loading capacity and an accelerated rate of flux decline. The average pore size of the pre-filter layer measured by SEM should not be too small (e.g., less than 200 nm). As a pre-filter layer that plays a role in pre-filtering large particles in the feed liquid, it needs to have a large dirt-holding space for large particles to avoid being quickly blocked by them.

[0050] The separation layer of the filter membrane primarily functions to trap small- to medium-sized particles. Considering the high requirements of the biomedical field regarding the risk of viral leakage, it is essential to ensure that the filter membrane is free from viral leakage. Therefore, the separation layer must have a good retention effect on small- to medium-sized viruses. Furthermore, to ensure a high loading capacity and a slow flux decay rate, the separation layer also needs ample space to accommodate small- to medium-sized particles, preventing rapid clogging of the pore structure. Since the pore size of the separation layer in this application remains essentially constant along the thickness direction, the trapping of small- to medium-sized viruses is mainly achieved through the stacking of pore structures in the thickness direction. Therefore, if the average pore size measured by SEM is too large (e.g., greater than 80 nm) and the separation layer thickness is too small (e.g., less than 20 μm), it often indicates an increased risk of viral leakage. Conversely, while a small average pore size measured by SEM (e.g., less than 30 nm) and an excessively large separation layer thickness (e.g., greater than 20 μm) can indeed further improve the virus retention effect of the filter membrane and reduce the risk of viral leakage, this will also lead to a significant decrease in the filter membrane flux.

[0051] It is important to note that the generally accepted standard both domestically and internationally is that if a virus-removing filter membrane achieves an LRV > 6, it is considered to have essentially removed all viruses. Further improving the membrane's retention rate (e.g., by further reducing the average pore size measured by SEM at the filter's outlet), while theoretically possible, often comes at the cost of a significant decrease in flux. Furthermore, there is a diminishing marginal return effect to improving virus retention; this means that if the membrane's retention is already good, even a small further improvement may require a substantial decrease in flux. However, when the average pore size measured by SEM at the filter's outlet is within the range of 30–45 nm, not only is the risk of virus leakage significantly reduced, but the flux decrease is also minimal. In this case, the average pore size measured by SEM at the outlet can be adjusted appropriately according to actual needs to achieve higher virus retention or higher flux.

[0052] Optionally, the average pore size change gradient of the pre-filter layer measured by SEM is at least 2 nm / μm greater than that of the separation layer measured by SEM, and the average pore size change gradient of the separation layer measured by SEM is no greater than 3 nm / μm.

[0053] By adopting the above technical solution, as mentioned earlier, changes in pore size will lead to greater changes in pore cross-sectional area and pore volume. This means that smaller pore sizes in the filter membrane have a greater impact on the filter membrane's flux. In this application, the average pore size variation gradient of the filter membrane separation layer measured by SEM is no greater than 3 nm / μm. Combined with the average pore size of the separation layer measured by SEM (30–80 nm), this indicates that there are no extremely small pore structures in the separation layer, ensuring that the filter membrane has a large flux. The average pore size variation gradient of the pre-filter layer measured by SEM should not be too small. This is because if the pre-filter layer forms a basically symmetrical structure similar to the separation layer, a larger thickness is required to ensure the retention effect on large particles. If the thickness of the pre-filter layer is too large, not only will the resistance to the feed liquid increase, but the possibility of structural collapse under the pressure of the feed liquid will also increase. A pre-filter layer with a pore size gradient not only ensures the retention effect on large particles, but also, combined with the larger pore size (200–800 nm) of the pre-filter layer, ensures sufficient space to accommodate large particles.

[0054] Optionally, the SEM measurement fiber diameter of the pre-filter layer is 60–150 nm, and the SEM measurement fiber diameter of the separation layer is 20–40 nm.

[0055] By adopting the above technical solution, the filter membrane in this application is made of a relatively soft fiber material, and the average pore size of the pre-filtration layer measured by SEM is 200–800 nm. As the layer structure directly bearing the pressure of the feed liquid, the pre-filtration layer is prone to structural collapse due to the pressure. Based on this, the pre-filtration layer has a larger fiber structure than the separation layer and the pore size transition layer, which can provide stronger support for the larger pore size of the pre-filtration layer. This ensures that the pre-filtration layer structure, which directly bears the pressure of the feed liquid, has strong self-supporting performance and is not prone to structural collapse due to the pressure, thus ensuring that the filter membrane has high loading capacity and flux.

[0056] Compared to the larger fiber structure of the pre-filtration layer, the fiber structure of the separation layer should not be too large. This is because the pore size of the separation layer is relatively small, the three-dimensional network structure formed by the fiber structure is more compact, and the connection between fibers is tighter. Therefore, the separation layer inherently has better self-supporting ability than the pre-filtration layer. In addition, compared to the pre-filtration layer which directly bears the pressure of the feed liquid, the separation layer is closer to the liquid outlet surface and does not directly bear the pressure of the feed liquid. Therefore, the fiber diameter of the separation layer does not need to be too large. Furthermore, further increasing the fiber diameter of the separation layer will lead to a further increase in the resistance encountered by the feed liquid at the separation layer (the resistance of the feed liquid by the solid part increases). As mentioned above, the surface morphology (pore structure, fiber structure) of the separation layer of the filter membrane is an important factor affecting the filter membrane flux. Therefore, further increasing the fiber diameter of the separation layer may lead to a significant decrease in the filter membrane flux, resulting in a decrease in filter membrane efficiency and reduced practicality.

[0057] Optionally, the liquid inlet surface includes a plurality of strip-shaped first fibers, the plurality of first fibers being interconnected and adjacent first fibers surrounding each other to form a first hole;

[0058] The average pore size of the first hole, as measured by SEM, is 350–2000 nm; and / or,

[0059] The liquid outlet surface has several circular second holes;

[0060] The average pore size of the second hole, as measured by SEM, is 25–45 nm.

[0061] By adopting the above technical solution, the average pore size of the first pore measured by SEM should not be too large (e.g., exceeding 2000 nm), so as to avoid the porous body near the liquid inlet surface of the filter membrane from collapsing under pressure due to insufficient pressure resistance, thereby leading to a decrease in filter membrane flux and loading capacity; the average pore size of the first pore measured by SEM should also not be too small (e.g., less than 350 nm), so as to avoid the feed liquid experiencing greater resistance and insufficient pre-filtration dirt-holding capacity, leading to a decrease in flux and loading capacity.

[0062] The average pore size of the second pore, measured by SEM, should also not be too large (e.g., exceeding 45 nm). This is because the separation layer of the filter membrane in this application has a pore structure with a basically constant pore size. Although the thickness of the separation layer can still form a good interception of small and medium-sized viruses, the risk of leakage of small and medium-sized viruses increases significantly with the extension of filtration time. The porous body near the liquid outlet of the filter membrane is the last barrier of the filter membrane. The smaller average pore size of the second pore, measured by SEM, can ensure efficient interception of a small number of small and medium-sized viruses that leak from the separation layer. The average pore size of the second pore, measured by SEM, should also not be too small (e.g., less than 30 nm). This is because a pore structure with too small a pore size at the liquid outlet will have too much resistance to the feed liquid, resulting in a significant decrease in the flux of the filter membrane.

[0063] Optionally, among the second holes, those with a diameter greater than 30 nm as measured by SEM shall account for no more than 20% of the total number of second holes; and among the second holes, those with a diameter less than 20 nm as measured by SEM shall account for no more than 20% of the total number of second holes.

[0064] By adopting the above technical solution, as mentioned above, the most important factor affecting the flux of the filter membrane is the pore structure in the filter membrane. Although the pore structure has a better interception effect on viruses, a slight reduction in the pore size will lead to a significant reduction in the cross-sectional area of ​​the pore structure, resulting in a significant reduction in flux.

[0065] The inventors of this application have discovered that, for a filter membrane with a specific structure as described in this application, if the proportion of second pores larger than 30 nm (SEM measured average pore size) exceeds 20%, the risk of virus leakage after process dwell time (e.g., during top-wash LRV) will be significantly increased. This is because the proportion of large-diameter pores in the second pores is relatively high. After process dwell time, viruses trapped in the separation layer are likely to be released due to Brownian motion or other reasons. In this case, re-pressurizing and filtration may cause the released viruses in the separation layer to leak from the high proportion of large-diameter second pores, leading to an increased risk of virus leakage after process dwell time. While the risk of virus leakage is reduced if the proportion of second pores smaller than 20 nm (SEM measured average pore size) exceeds 20%, these high-proportion, small-diameter second pores will have a significant impact on the filter membrane flux. When the proportion of second pores larger than 30 nm (SEM average pore size) and smaller than 20 nm (SEM average pore size) is less than 20%, it indicates that the proportion of large-diameter and small-diameter pore structures in the second pores is low. Therefore, the filter membrane not only has a high flux but also a lower risk of virus leakage.

[0066] Optionally, in the second aperture, the proportion of second apertures with a deviation from the mean of no more than 25 nm is no less than 70%; the deviation from the mean is the absolute value of the difference between the SEM measured aperture of the second aperture and the SEM measured average aperture of the second aperture.

[0067] By adopting the above technical solution, the proportion of second pores with a deviation of no more than 25nm is relatively high, which can more stably trap the virus. Furthermore, since the pore size distribution of the second pores is mainly near the average pore size measured by SEM, the resistance encountered by the liquid at various points is relatively small, making it less likely to cause concentrated blockage of small-diameter second pores, and also less likely to leak from large-diameter second pores.

[0068] Optionally, the flux retention coefficient of the filter membrane is T, wherein T is not less than 0.4@(10g / L), and T is calculated by the following formula:

[0069]

[0070] In the above formula, V is the filter membrane loading capacity, and the loading capacity when the flux decreases by a% is Va; V25 is the filter membrane loading capacity when the flux decreases by 25%, that is, V50 is the filter membrane loading capacity when the flux decreases by 50%, and V75 is the filter membrane loading capacity when the flux decreases by 75%.

[0071] By adopting the above technical solution, T not less than 0.4@(10g / L) means that in a system with a protein concentration of 10g / L, the flux retention coefficient T of the filter membrane is not less than 0.4. It should be noted that compared with the currently common system with a protein concentration of only 1g / L, testing with a protein concentration of 10g / L poses a greater challenge to the filter membrane's load capacity and is more likely to cause filter membrane clogging.

[0072] The inventors of this application have discovered that, while ensuring the filter membrane has a good retention effect on small and medium-sized viruses (high LRV), the flux decay rate of the filter membrane can be well characterized by the above formula. Compared with determining the flux decay rate of the filter membrane in a more qualitative way, characterizing and judging the flux decay rate of the filter membrane in a quantitative way has higher reference value.

[0073] It should be noted that the measurement system for the above data is as follows: protein concentration 10 g / L, buffer system is 50 mM acetic acid + 100 mM NaCl, pH 5.0, Cond 30.5 μs / cm, density 1 g / mL, and this system is pre-filtered with a 0.22 μm sterile membrane; in this system, the protein is IVIG.

[0074] Optionally, the top-wash LRV retention coefficient of the filter membrane is not less than 80%; the filter membrane flux at 75% decay is not less than 120 L / m³. 2The protein yield of the filter membrane is not less than 98%; the flux of the filter membrane is not less than 60 L / (h·m). 2 @30psi.

[0075] By adopting the above technical solution, due to the special structure of the filter membrane of this application, it is possible to ensure that the filter membrane not only has a high flux, but also a high loading capacity and LRV. Moreover, the filter membrane of this application can be used for a long time under a pressure of 30 psi. Furthermore, since the proportion of large-diameter pore structure and small-diameter pore structure in the second pore of the filter membrane is low, the filter membrane can have a top wash LRV retention coefficient of not less than 80% after the process stay.

[0076] The term "process hold" refers to the process of removing the external pressure and adding buffer to the filter membrane when the flux drops to 25% or when only a small amount of the challenge solution remains at the bottom during a virus challenge test. The filter is then pressurized again and the virus challenge test is performed after the process hold. The filtrate obtained from the test is collected separately and the virus titer is determined. The LRV after the process hold is then calculated.

[0077] The top-wash LRV retention factor refers to the ratio of the LRV of the filter membrane before and after top washing during a virus challenge test, and is expressed as a percentage.

[0078] Secondly, this application provides a process for preparing the aforementioned filter membrane, which adopts the following technical solution:

[0079] A process for preparing a filter membrane includes the following steps:

[0080] S1. Casting: The first casting solution and the second casting solution are sequentially cast onto the carrier to form a double-layer liquid film. The first casting solution comprises the following parts by weight of raw materials: 10-30 parts of the first film-forming polymer and 30-150 parts of the first solvent system. The second casting solution comprises the following parts by weight of raw materials: 10-20 parts of the second film-forming polymer and 70-300 parts of the second solvent system. The viscosity ratio of the first casting solution to the second casting solution is (2-20):1. The solid content of the first casting solution is not less than 15%, and the surface tension of the first solvent system is not greater than 30 dyne / cm.

[0081] S2. Phase separation and solidification: Immerse the double-layer liquid film in a coagulation bath to allow the casting liquid to separate and solidify until the double-layer liquid film is completely solidified. The coagulation bath is water or ethanol to obtain the raw film.

[0082] S3. Regeneration: The biofilm is immersed in a regeneration bath for regeneration to obtain a regenerated biofilm.

[0083] By adopting the above technical solution, the filter membrane in this application employs a double-layer casting process. Specifically, a first casting solution and a second casting solution with different solid content and viscosity form the small-pore and large-pore structures of the filter membrane, respectively. Since the phase separation mechanism in this application is non-solvent-induced phase separation (NIPS), a coagulation bath is required to penetrate into the casting solution to induce phase separation and solidification. Therefore, the phase separation start times of the first and second casting solutions are not the same (the second casting solution, closer to the coagulation bath, begins phase separation earlier). Furthermore, the largely identical formulation systems of the first and second casting solutions allow for mass transfer between them. Combined with the aforementioned difference in the phase separation start times of the first and second casting solutions, this means that after sequentially casting them onto the carrier, before the first casting solution separates, mass transfer occurs between the first and second casting solutions to form a mixed casting solution. This mixed casting solution creates a solid content and viscosity gradient from the first casting solution to the second casting solution. Because although the first casting solution, the mixed casting solution, and the second casting solution are different, their systems are roughly the same and their separation mechanisms are the same. The morphology and size of their fiber structure and the morphology and size of their pore structure do not undergo obvious abrupt changes. Therefore, the cross-section of the final filter membrane does not have obvious boundary lines or layered structures after being characterized by SEM.

[0084] The first casting solution, with higher solids content and viscosity, mainly forms a separation layer after phase separation and solidification, which has a good retention effect on small and medium-sized viruses. The higher solids content of the first casting solution makes it easier to form a small pore structure. The second casting solution, with lower solids content and viscosity, mainly forms a pre-filtration layer after phase separation and solidification, which has a good retention effect on large particles. The relatively lower solids content of the second casting solution makes it easier to form a large pore structure. The mixed casting solution formed by mixing the first and second casting solutions forms the pore size transition layer of the filter membrane, and a pore structure with a certain pore size gradient is formed in the pore size transition layer.

[0085] In the filter membrane preparation process of this application, the viscosity ratio of the first casting solution and the second casting solution should not be too high or too low. If the viscosity ratio of the first casting solution and the second casting solution is too high (e.g., exceeding 20:1), it indicates that the viscosity of the first casting solution is too high or the viscosity of the second casting solution is too low. When the viscosity of the first casting solution is too high, the coagulation bath is difficult to penetrate into the interior of the first casting solution, resulting in uneven distribution of the coagulation bath inside the first casting solution and uneven phase separation in various parts of the first casting solution. The resulting pore structure is naturally less uniform, and the risk of virus leakage is higher. If the viscosity of the first casting solution is so high that the coagulation bath is difficult to penetrate, it is very likely to lead to membrane formation failure. Furthermore, a significant viscosity difference between the first and second casting solutions also affects the mass transfer between them. If a mixed casting solution is not formed between the first and second casting solutions, a pore size abrupt change region will appear at the interface between the small pore structure formed by the first casting solution and the large pore structure formed by the second casting solution. This pore size abrupt change region is easily blocked by the concentrated retention of particles of different sizes, leading to a rapid decline in the filter membrane flux. When the viscosity of the second casting solution is too low, a pre-filter layer with an excessively large pore size structure is easily formed, resulting in poor pressure resistance and a tendency for structural collapse under pressure.

[0086] If the viscosity ratio of the first casting solution and the second casting solution is too small (e.g., below 2:1), it indicates that the viscosity of the first casting solution is too low or the viscosity of the second casting solution is too high. When the viscosity of the first casting solution is too low, although the coagulation bath can easily enter the first casting solution, the coagulation bath is excessively diluted by the solvent system in the first casting solution, thus easily forming a separation layer with a large pore size. This makes it difficult for the separation layer to achieve the required virus retention effect, resulting in an excessive risk of virus leakage. When the viscosity of the second casting solution is too high, the coagulation bath encounters greater resistance at the second casting solution, making it difficult to penetrate and immerse itself in the casting solution to promote phase separation, resulting in uneven phase separation. Furthermore, the high viscosity of the second casting solution easily forms a pre-filtration layer with a small pore size, resulting in a low dirt-holding capacity of the pre-filtration layer and a corresponding decrease in the filter membrane's loading capacity.

[0087] It is important to note that the surface tension of the first solvent system should not be too high (below 30 dyne / cm). This is because the non-solvent-induced phase separation mechanism used in this application determines that the coagulation bath, as the primary driving force for phase separation, has a significant impact on the phase separation rate of the casting solution due to its immersion and penetration speed. To improve the phase separation uniformity throughout the casting solution, it is necessary to reduce the resistance of the coagulation bath in the casting solution, thereby improving the dispersion uniformity of the coagulation bath and reducing the uneven phase separation rate caused by uneven coagulation bath dispersion. Uneven phase separation rate leads to a decrease in the uniformity of pore size distribution and a larger standard deviation of pore size. For the separation layer, a decrease in pore size distribution uniformity means a significantly increased risk of virus leakage from larger pore structures. Since the viscosity of the first casting solution is much greater than that of the second casting solution, and the first casting solution is located below the second casting solution, it is more difficult for the coagulation bath to penetrate evenly into the interior of the first casting solution. Therefore, the surface tension of the first solvent system is more critical for the first casting solution.

[0088] By controlling the surface tension of the first solvent system, it is possible to ensure that the coagulation bath can be more uniformly immersed and penetrate into the interior of the first casting solution, thereby making the phase separation of the first casting solution more uniform and obtaining a basically symmetrical and highly uniform pore structure. Since the proportion of extremely small pore structures and the proportion of large pore structures formed by the high solid content first casting solution are both low, it can not only reduce the risk of virus leakage, but also ensure that the filter membrane has a relatively large flux.

[0089] Optionally, step S2 specifically includes:

[0090] S21. Pretreatment: Immerse the double-layer liquid membrane in a pretreatment bath for pretreatment. The pretreatment time is 0.5 to 10 seconds. The pretreatment bath is a 40 to 100% aqueous solution of solvent to obtain a pretreated membrane.

[0091] S22. Curing: Immerse the pretreated membrane in a coagulation bath until the pretreated membrane is completely phase-separated and cured to obtain a green membrane. The coagulation bath is water or ethanol.

[0092] By adopting the above technical solution, the inventors of this application have discovered that it is difficult to form pores on the surface of cellulose materials. If the casting solution is directly immersed in the coagulation bath, even if the solid content of the second casting solution is low, it is easy to form a relatively dense skin structure. This may be related to the high sensitivity of cellulose materials to the coagulation bath. That is, when cellulose materials encounter a high concentration of coagulation bath, they will quickly separate phases in a short time. It is generally believed that the faster the phase separation rate, the smaller the pore size of the formed pore structure. Therefore, the surface of the second casting solution in contact with the coagulation bath is likely to form a skin structure. During the process of the coagulation bath penetrating into the casting solution after passing through the skin layer, the solvent system in the casting solution dilutes the coagulation bath, thereby reducing the concentration of the coagulation bath and slowing down the phase separation rate of the casting solution. Combined with the lower solid content of the second casting solution, the casting solution inside the skin layer forms a pore structure with a larger pore size. Furthermore, since the larger pore structure can accommodate a larger amount of coagulation bath, this part of the coagulation bath can penetrate into the mixed casting solution and the first casting solution more quickly, promoting the rapid phase separation of the mixed casting solution and the first casting solution.

[0093] This application specifically involves pretreatment before immersion in the coagulation bath. The pretreatment bath can be a mixture of solvent and water or a pure solvent. When the pretreatment bath is a pure solvent, after immersion in the bilayer liquid film, the surface of the second casting liquid is diluted by the pure solvent, resulting in a certain low-solids content region on the surface of the second casting liquid (with a lower solids content than the second casting liquid itself), thereby greatly reducing the possibility of skin formation after immersion in the coagulation bath. When the pretreatment bath is a mixture of solvent and water, compared to a high-concentration coagulation bath, the gelation phase separation effect is significantly reduced due to the addition of a certain amount of solvent in the pretreatment bath. Therefore, the second casting liquid gels and separates at a significantly slower rate in the pretreatment bath. Combined with the already low solids content of the second casting liquid and the dilution effect of the solvent in the pretreatment bath, the possibility of skin formation in the second casting liquid can also be greatly reduced.

[0094] It is understood that a 40-100% aqueous solution of solvent in the pretreatment bath means that the volume percentage of solvent in the pretreatment bath is 40-100%, that is, the pretreatment bath can be a mixture of solvent and water, or it can be a pure solvent. The solvent refers to a substance capable of dissolving the film-forming polymer. In this application, the solvent can be one of acetone, dioxane, dimethylacetamide, N-methylpyrrolidone, acetic acid, propionic acid, butyric acid, and valeric acid.

[0095] Optionally, the surface tension of the second solvent system is not greater than 30 dyne / cm; the surface tension of the first solvent system is less than the surface tension of the second solvent system.

[0096] By adopting the above technical solution, in this application, the first casting solution has a higher viscosity. Higher viscosity not only means increased coagulation bath resistance but also increased mass transfer resistance between the first and second casting solutions. The specific double-layer casting membrane fabrication process in this application needs to avoid abrupt changes in pore size. Therefore, the first and second casting solutions need to transfer mass to each other to form a mixed casting solution with a viscosity gradient. Furthermore, ensuring that the surface tension of the second solvent system in the second casting solution is no greater than 30 dyne / cm ensures better mass transfer between the first and second casting solutions, reducing the possibility of abrupt changes in pore size within the filter membrane.

[0097] Furthermore, the higher viscosity of the first casting solution, coupled with its location away from the coagulation bath, makes it more difficult for the coagulation bath to penetrate and permeate into the first casting solution. If excessive mass transfer resistance leads to significant inhomogeneity of the coagulation bath within the first casting solution, the uniformity of phase separation velocities throughout the solution will decrease, making it difficult to form a uniform, essentially symmetrical pore structure. By limiting the surface tension of the first solvent system to be lower than that of the second solvent system, the mass transfer resistance of the coagulation bath within the first casting solution can be significantly reduced, thereby improving the phase separation uniformity of the first casting solution and ensuring a lower risk of virus leakage into the separation layer.

[0098] Optionally, the viscosity of the first casting solution is 10,000 to 30,000 cp, and the viscosity of the second casting solution is 1,000 to 4,000 cp.

[0099] By adopting the above technical solution, as mentioned earlier, the viscosity of the first casting solution should not be too high or too low, and similarly, the viscosity of the second casting solution should not be too high or too low, and the ratio of their viscosities needs to be controlled within a certain range. Based on this, when the viscosity of the first casting solution is 10,000–30,000 cp and the viscosity of the second casting solution is 1,000–4,000 cp, and the coagulation bath is water or ethanol; the first and second casting solutions can not only form a mixed casting solution through mutual mass transfer, thereby obtaining a pore size transition layer structure with a pore size gradient to reduce the possibility of rapid flux decay caused by abrupt changes in pore size; but also ensure that the coagulation bath is dispersed in the casting solution (especially in the first casting solution) in a relatively rapid and uniform manner, thereby ensuring that the pore structure at the separation layer has a more uniform pore size.

[0100] Understandably, the pore size uniformity of the separation layer is more important than that of the pre-filter layer. This is because the pre-filter layer, with its larger pore size, is mainly used to trap large particles, with less impact on trapping smaller viruses. Furthermore, the pre-filter layer has a certain pore size gradient and a large thickness, so even if the pore size distribution uniformity of the pre-filter layer is relatively poor, as the pore size decreases and the pore structure is continuously stacked in the thickness direction, it can still effectively trap large particles. The pore size transition layer can also trap a small amount of large particles that leak from the pre-filter layer.

[0101] The separation layer is the main area in the filter membrane that traps viruses. In order to ensure high contaminant capacity and high throughput, the separation layer of the filter membrane in this application has a basically symmetrical structure. This means that the risk of virus leakage from the separation layer of the filter membrane in this application is relatively higher. Once the pore size distribution in the separation layer is uneven, it is very likely to lead to virus leakage. The risk of virus leakage must be strictly controlled. Therefore, it is relatively more important to control the casting solution system, phase separation conditions, etc. to improve the uniformity of pore size distribution in the separation layer.

[0102] Optionally, the first film-forming polymer is at least one of cellulose diacetate, cellulose triacetate, cellulose propionate, cellulose acetate phthalate, cellulose acetate butyrate, and cellulose acetate propionate.

[0103] The second film-forming polymer is at least one of cellulose diacetate, cellulose triacetate, cellulose propionate, cellulose acetate phthalate, cellulose acetate butyrate, and cellulose acetate propionate.

[0104] Optionally, the first solvent system is a mixture of a first good solvent and a first pore-forming agent in a mass ratio of (1-8):(5-22);

[0105] The second solvent system is a mixture of a second good solvent and a second pore-forming agent in a mass ratio of (3-13):(4-17);

[0106] The first good solvent is at least one selected from acetone, dioxane, dimethylacetamide, N-methylpyrrolidone, acetic acid, propionic acid, butyric acid, and valeric acid;

[0107] The second good solvent is at least one selected from acetone, dioxane, dimethylacetamide, N-methylpyrrolidone, acetic acid, propionic acid, butyric acid, and valeric acid;

[0108] The first pore-forming agent is at least one of ethanol, 1-propanol, hexafluoroisopropanol or trifluoroethanol;

[0109] The second pore-forming agent is at least one of ethanol, 1-propanol, hexafluoroisopropanol or trifluoroethanol.

[0110] By adopting the above technical solutions, nitrocellulose, cellulose acetate, cellulose propionate and other esters of cellulose can be regenerated by hydrolysis and degreasing under certain conditions. Regenerated cellulose is currently a more mainstream cellulose-based film-forming material and is more environmentally friendly and green than cuprammonium fiber.

[0111] Both the first solvent system and the second solvent system are obtained by mixing a good solvent and a pore-forming agent. It can be understood that a good solvent refers to a solvent with good solubility for the film-forming polymer, while a pore-forming agent refers to a component that, although it has poor solubility for the film-forming polymer, can promote the formation of pore structures during the film-forming process and promote the coagulation bath to penetrate into the casting solution.

[0112] Because the first and second casting solutions have different casting sequences and solid contents, their resistance to the coagulation bath also differs. To promote rapid and uniform mass transfer in the coagulation bath between the first and second casting solutions, the first casting solution, with a higher solid content and located further from the coagulation bath, requires more pore-forming agent to facilitate mass transfer. The inventors of this application have discovered that, for the specific double-layer casting process of this application, when the first solvent system is a mixture of a first good solvent and a first pore-forming agent at a mass ratio of (1–8):(5–22), and the second solvent system is a mixture of a second good solvent and a second pore-forming agent at a mass ratio of (3–13):(4–17), a filter membrane with high LRV and high loading capacity can be obtained.

[0113] This application specifically uses a low surface tension pore-forming agent. Compared with commonly used PVP and PEG pore-forming agents, the low surface tension pore-forming agent can promote the penetration of the coagulation bath into the casting solution and improve the phase separation uniformity of each region inside the casting solution. Although PVP, PEG and other pore-forming agents can indeed form pores after removal, these pore-forming agents have high viscosity and a large mass transfer resistance to the coagulation bath. Considering that the first casting solution has a high solid content and high viscosity, if PVP and PEG, which can form pores, are added further, they will have a significant impact on the phase separation uniformity of the first casting solution and are more likely to lead to the formation of large gradient and non-uniform pore structures.

[0114] Optionally, in step S3, the regeneration bath is an aqueous sodium hydroxide solution, and the temperature of the regeneration bath is 20–40°C, with a regeneration time of 30–120 min.

[0115] Optionally, after step S3, step S4, crosslinking, is performed, in which the regenerated membrane is placed in a crosslinking agent for crosslinking treatment. After the crosslinking treatment is completed and the membrane is cleaned, the finished membrane is obtained. The crosslinking agent is at least one of halogenated epoxides, dihalogenated alkanes, and dihalogenated alcohols.

[0116] By adopting the above technical solution, it is found that due to the soft texture of cellulose, the regenerated fiber membrane obtained after hydrolysis and regeneration often has poor mechanical properties. Virus filtration often employs dead-end filtration, which subjectes the membrane to significant pressure from the feed solution. If the membrane's pressure resistance is insufficient, it may cause the membrane's pore structure to collapse. Once the membrane's pore structure collapses due to pressure, its virus retention capacity and flux will be significantly affected. Therefore, after step S3, a crosslinking treatment with a crosslinking agent is performed to improve the membrane's pressure resistance and reduce the possibility of structural collapse under pressure.

[0117] In summary, this application includes at least one of the following beneficial technical effects:

[0118] 1. The filter membrane of this application uses cellulose raw materials with good hydrophilicity as the film-forming material, which has a high protein yield; the average pore size of the filter membrane is 15-40 nm, thus it has a good retention effect on small and medium-sized viruses; the pre-filtration layer is relatively thick, thus it has a good retention effect on large particles and a large dirt holding capacity; the pore size transition layer of the filter membrane is relatively thin and the pore size change coefficient is relatively small, so the pore size of the pore size transition layer does not change abruptly, and it is not easy to cause concentrated retention and local blockage and rapid decrease in flux due to abrupt changes in pore size, thus it has a large loading capacity; the basic symmetrical structure of the separation layer ensures that the separation layer has a good retention effect on small and medium-sized viruses, and since no micropore structure is introduced, it is not easy to cause blockage and rapid decrease in flux due to the retention of a large number of viruses in a small area, and the absence of micropore structure also makes the filter membrane have a large flux; therefore, the filter membrane of this application not only has a good virus retention effect, but also has high flux, loading capacity and protein yield.

[0119] 2. This application further discloses a filter membrane preparation process, which adopts a double-layer coating process, using casting liquids of different viscosities to form a separation layer with a small pore structure and a pre-filtration layer with a large pore structure. Since the two casting liquids have roughly the same system, mass transfer will occur between them after casting, forming a mixed casting liquid with a viscosity gradient. This greatly reduces the possibility of a boundary line in the filter membrane and reduces the possibility of rapid flux decay due to local concentrated retention. Attached Figure Description

[0120] Figure 1 This is a SEM image of the inlet surface of the filter membrane in Embodiment 6 of this application, and the magnification of the image is 20k×.

[0121] Figure 2 This is a SEM image of the liquid outlet surface of the filter membrane in Example 6 of this application, and the magnification of the image is 50k×.

[0122] Figure 3 This is a cross-sectional SEM image of the filter membrane of Embodiment 6 of this application, and the magnification of the image is 1k×.

[0123] Figure 4 This is a cross-sectional SEM image of the pore size variation layer of the filter membrane in Embodiment 6 of this application, and the magnification of the image is 10k×.

[0124] Figure 5 This is a cross-sectional SEM image of the filter membrane near the liquid inlet surface of Embodiment 6 of this application, and the magnification of the image is 10k×.

[0125] Figure 6 This is a cross-sectional SEM image of the filter membrane near the liquid outlet surface of Embodiment 6 of this application, and the magnification of the image is 50k×. Detailed Implementation

[0126] The following is in conjunction with the appendix Figures 1-6 This application will be described in further detail.

[0127] Example 1

[0128] This application discloses a process for preparing an asymmetric cellulose virus-removing filter membrane, comprising the following steps:

[0129] S1. Casting: The materials are mixed and prepared according to the proportions in Table 1 to obtain the first casting solution and the second casting solution. The first and second casting solutions are then sequentially cast onto a carrier, with the first casting solution positioned on the carrier side and the second casting solution above it. After being cast onto the carrier, the first and second casting solutions form a double-layer liquid film. In this embodiment, both the first and second film-forming polymers are cellulose diacetate, both the first and second good solvents are dimethylacetamide, and both the first and second pore-forming agents are ethanol.

[0130] S2, Phase Separation Curing, specifically includes:

[0131] S21. Pretreatment: Immerse the double-layer liquid membrane in a pretreatment bath for 7 seconds. The pretreatment bath is an 85% acetone aqueous solution. After the pretreatment, the pretreated membrane is obtained.

[0132] S22. Curing: Immerse the pretreated membrane obtained after pretreatment into a coagulation bath until the pretreated membrane is completely phase-separated and cured to obtain a green film. In this embodiment, the coagulation bath is water.

[0133] S3. Regeneration: Immerse the fully cured green membrane in a regeneration bath at 30°C for 70 minutes to hydrolyze cellulose acetate into regenerated cellulose. Remove the hydrolyzed membrane and wash it with water until the pH is neutral to obtain the regenerated membrane. In this embodiment, the regeneration bath is a 0.05 mol / L sodium hydroxide aqueous solution at 50°C.

[0134] S4. Crosslinking: The regenerated membrane is placed in a crosslinking agent for crosslinking treatment. After the crosslinking treatment is completed, the finished membrane is obtained. In this embodiment, the crosslinking agent is an aqueous solution of epichlorohydrin with a concentration of 10 wt%, the crosslinking time is 25 min, and the crosslinking temperature is 50 °C.

[0135] Examples 2-6

[0136] The main difference between Examples 2-6 and Example 1 is that the process parameters for each step are different and the composition ratio of the casting solution is different, as detailed in Table 1.

[0137] Example 7

[0138] The main difference between Example 7 and Example 1 is that in Example 7, a pretreatment bath was not used to pretreat the bilayer structure. Instead, the bilayer liquid film was directly immersed in the coagulation bath to allow the bilayer liquid film to solidify through phase separation. Apart from this, the process parameters and composition ratio of the casting solution for each step are detailed in Table 1. Step S2 specifically includes the following process steps:

[0139] S2. Phase separation and solidification: The double-layer liquid film is immersed in a coagulation bath to allow the casting liquid to separate and solidify until the double-layer liquid film is completely solidified, thus obtaining a raw film. In this embodiment, the coagulation bath is water.

[0140] Example 8

[0141] The main difference between Example 8 and Example 1 is that the regenerated membrane was not crosslinked in Example 8. Apart from that, the process parameters and composition ratio of the casting solution for each step are detailed in Table 1.

[0142] Comparative Example

[0143] Comparative Example 1

[0144] The main difference between Comparative Example 1 and Example 1 is that Comparative Example 1 uses a single-layer casting process to prepare the filter membrane. Specifically, in Comparative Example 1, only the first casting solution is cast onto the carrier; a second casting solution is not cast on top of the first. Apart from this, the process parameters and composition ratios of the casting solutions for each step are detailed in Table 1. The filter membrane preparation process in Comparative Example 1 includes the following steps:

[0145] S1. Casting: The materials are mixed and prepared according to the proportions in Table 1 to obtain the first casting solution, and the first casting solution is cast onto the carrier to form a single-layer liquid film. In this comparative example, the first film-forming polymer is cellulose diacetate, the first good solvent is N-methylpyrrolidone, and the first pore-forming agent is 1-propanol.

[0146] S2, Phase Separation Curing, specifically includes:

[0147] S21. Pretreatment: Immerse the single-layer liquid film in a pretreatment bath for 5 seconds. The pretreatment bath is an 80% acetone aqueous solution. After the pretreatment, a pretreated film is obtained.

[0148] S22. Curing: Immerse the pretreated membrane obtained after pretreatment into a coagulation bath until the pretreated membrane is completely phase-separated and cured to obtain a green membrane. In this comparative example, the coagulation bath is water.

[0149] S3. Regeneration: The fully cured green membrane is placed in a regeneration bath at 30°C for 60 minutes to hydrolyze cellulose acetate into regenerated cellulose. The hydrolyzed membrane is then removed and washed with water until the pH is neutral to obtain the regenerated membrane. In this comparative example, the regeneration bath is a 0.05 mol / L sodium hydroxide aqueous solution at 50°C.

[0150] S4. Crosslinking: The regenerated membrane is placed in a crosslinking agent for crosslinking treatment. After the crosslinking treatment is completed, the finished membrane is obtained. In this comparative example, the crosslinking agent is an aqueous solution of epichlorohydrin with a concentration of 10 wt%, the crosslinking time is 25 min, and the crosslinking temperature is 50℃.

[0151] Comparative Example 2

[0152] The main difference between Comparative Example 2 and Example 1 is that, although Comparative Example 2 still uses a double-layer casting process to prepare the filter membrane, the pore-forming agent in both the first and second solvent systems of Comparative Example 2 is PEG-400. Apart from that, the process parameters and composition ratio of the casting solution for each step are detailed in Table 1.

[0153] Comparative Example 3

[0154] The main difference between Comparative Example 3 and Example 1 is that, although Comparative Example 3 still uses a double-layer casting process to prepare the filter membrane, the formulation of the first casting solution in Comparative Example 3 is adjusted so that the solid content of the first casting solution is less than 15% and the viscosity is less than 10 × 10⁻⁶. 3 In addition, the formulation of the second casting solution was adjusted to make the viscosity of the second casting solution higher, thereby making the viscosity ratio of the first casting solution and the second casting solution smaller; the specific composition ratio of the first casting solution and the second casting solution is detailed in Table 1.

[0155] Table 1. Formulation and process parameters for each embodiment and comparative example.

[0156]

[0157] Performance testing and performance data

[0158] I. Virus Challenge Test

[0159] The viral challenge test was performed according to the relevant specifications in PDA TR41, using PP7 bacteriophage or hepatitis B virus as the model virus, IVIG as the model protein, and PBS as the buffer. During the test, the changes in flux and load over time were recorded, yielding the membrane's LRV, flux, and load. When the membrane flux decreased to 25% or when only a small amount of the challenge solution remained at the bottom, a top wash was performed. Specifically, the external pressure was removed, buffer was added, and the mixture was allowed to stand for 15 minutes. Then, the pressure was increased to 30 psi for the post-process hold viral challenge test. The collected filtrate was the top wash filtrate. This filtrate was collected separately, and the viral titer was determined. The LRV after top washing was calculated to further determine the top wash LRV retention factor.

[0160] The surface morphology parameters of each embodiment and comparative example are recorded in Table 2:

[0161] Table 2 Surface morphology parameters of each embodiment and comparative example

[0162]

[0163]

[0164] The virus retention data of the filter membranes in each embodiment and comparative example are recorded in Table 3:

[0165] Table 3. Virus retention data for each embodiment and comparative example.

[0166]

[0167] in conclusion

[0168] The filter membranes in Examples 1-6 of this application all exhibit both high flux and virus retention efficiency. The filter membrane in Example 7, due to the lack of pretreatment, has a significantly smaller pore size at the inlet surface, resulting in a lower loading capacity. In Example 8, the lack of cross-linking results in a filter membrane with a larger pore size buffer structure, but a lower loading capacity. This may be due to the poor pressure resistance of the uncross-linked fiber structure, leading to a certain degree of pore collapse under pressure.

[0169] The filter membrane in Comparative Example 1 was fabricated using a single-layer casting process. Because there was no mass transfer between two different casting solutions to form a mixed casting solution with varying viscosity and solid content, a pore size transition layer structure was not formed. Observing the SEM image of the filter membrane cross-section in Comparative Example 1, it was found that the pore size of the membrane first decreases and then increases again in the thickness direction, which differs from the pore size change trend of the filter membranes prepared in the various embodiments of this application, which is generally from large to small. Furthermore, in preparing this filter membrane, to ensure good virus retention, the casting solution was controlled to have a higher viscosity and solid content (compared to the viscosity and solid content of the second casting solution in the various embodiments). A higher viscosity casting solution makes surface pore opening more difficult; therefore, a longer pretreatment time was used to ensure that a macroporous structure could be formed on the surface of the casting solution. However, although the long pretreatment time forms a macroporous structure on the surface of the casting solution, the excessively large pore structure causes the porous body near the liquid inlet of the filter membrane to be unable to withstand high pressure. Under a pressure of 30 psi, the porous body near the liquid inlet of the filter membrane collapses. Combined with the fact that the filter membrane has a large proportion of small pores, the flux and loading capacity are both small, resulting in low practicality.

[0170] Although the filter membrane in Comparative Example 2 employed a double-layer casting process, the use of PEG as a pore-forming agent resulted in a higher surface tension compared to the low-surface-tension pore-forming agent used in this application. Furthermore, the viscosity of the PEG pore-forming agent was significantly higher; the viscosity ratio of the first casting solution to the second casting solution in Comparative Example 2 was greater than 20:1. This high viscosity combined with the high viscosity ratio made mass transfer between the first and second casting solutions difficult, thus preventing the formation of a distinct pore size variation layer structure in Comparative Example 2. This resulted in a relatively significant abrupt change in pore size between the pre-filtration layer with its macroporous structure and the separation layer with its microporous structure, making localized clogging more likely and leading to a lower filter membrane loading. In addition, the high viscosity of the casting solution also resulted in higher resistance to the coagulation bath within the casting solution, leading to poor dispersion uniformity of the coagulation bath and an increased risk of virus leakage.

[0171] Although the filter membrane in Comparative Example 3 also employed a double-layer casting process, the solid content of the first casting solution was less than 15%, while the second casting solution had a higher solid content and viscosity. Consequently, the viscosity ratio between the first and second casting solutions was only 1.4:1. Because the coagulation bath encountered significant resistance at the second casting solution, it struggled to penetrate and infiltrate the interior of the casting solution, resulting in poor dispersion uniformity. Furthermore, the higher solid content of the second casting solution resulted in a smaller pore size and lower dirt-holding capacity, leading to a lower filter membrane loading capacity. In addition, the delayed phase separation start time of the first casting solution reduced the material uniformity within it, resulting in poor pore size uniformity in the separated layer structure after phase separation and solidification, thus increasing the risk of virus leakage.

[0172] This specific embodiment is merely an explanation of this application and is not intended to limit it. After reading this specification, those skilled in the art can make modifications to this embodiment without contributing any inventive step, but such modifications are protected by patent law as long as they fall within the scope of the claims of this application.

Claims

1. An asymmetric regenerated cellulose virus-removing filter membrane, comprising a porous body having non-directional tortuous pathways within the porous body, one side surface of the porous body being a liquid inlet surface and the other side surface of the porous body being a liquid outlet surface, characterized in that: The average pore size measured by SEM at the liquid inlet surface is greater than the average pore size measured by SEM at the liquid outlet surface; the porous body comprises a pre-filtration layer, a pore size transition layer, and a separation layer in sequence from the liquid inlet surface to the liquid outlet surface, with one side of the pre-filtration layer being the liquid inlet surface and one side of the separation layer being the liquid outlet surface; The average pore size of the filter membrane is 15–40 nm. The thickness of the pre-filter layer is 20–80 μm; The pore size variation coefficient S of the pore size transition layer (0～0.1) / (0.9～1) The pore size transition layer is not greater than 5, and its thickness is not greater than 20 μm. The pore size variation coefficient S of the pore size transition layer (0～0.1) / (0.4～0.5) The pore size variation coefficient S of the pore size transition layer is 1.1 to 1.

6. (0.5～0.6) / (0.9～1) It ranges from 1.1 to 1.6; The average pore size of the pore size transition layer measured by SEM is 100-300 nm. The average pore size of the pore size transition layer measured by SEM is smaller than the average pore size of the pre-filter layer but larger than the average pore size of the separation layer measured by SEM. The separation layer has a basically symmetrical structure, which means that the average pore size change gradient measured by SEM is no greater than 5 nm / μm, and the thickness of the separation layer is 20 to 50 μm. The pore size transition layer is considered to have a film thickness of 0 on the side near the liquid inlet surface and a film thickness of 1 on the side near the liquid outlet surface. Aperture variation coefficient 2. The asymmetric regenerated cellulose virus-removing filter membrane according to claim 1, characterized in that: The pore size variation coefficient S of the pore size transition layer (0～0.1) / (0.4～0.5) The pore size variation coefficient S of the pore size transition layer (0.5～0.6) / (0.9～1) The ratio is 0.8 to 1.

2.

3. The asymmetric regenerated cellulose virus-removing filter membrane according to claim 1, characterized in that: The difference between the average SEM-measured pore size of the 0.3–0.5 region of the pore size transition layer and the average SEM-measured pore size of the 0.5–0.7 region of the pore size transition layer is 30–100 nm.

4. The asymmetric regenerated cellulose virus-removing filter membrane according to claim 1, characterized in that: The SEM measurement of the fiber diameter in the 0-0.5 region of the pore size transition layer is 20-50 nm; the SEM measurement of the fiber diameter in the 0.5-1 region of the pore size transition layer is 20-45 nm.

5. The asymmetric regenerated cellulose virus-removing filter membrane according to claim 1, characterized in that: The difference between the SEM measured fiber diameter in the 0.3–0.5 region of the aperture transition layer and the SEM measured fiber diameter in the 0.5–0.7 region of the aperture transition layer is no greater than 10 nm.

6. The asymmetric regenerated cellulose virus-removing filter membrane according to claim 1, characterized in that: The average pore size of the pre-filter layer measured by SEM is 200–800 nm; the average pore size of the separation layer measured by SEM is 30–80 nm.

7. The asymmetric regenerated cellulose virus-removing filter membrane according to claim 1, characterized in that: The average pore size change gradient of the pre-filter layer measured by SEM is at least 2 nm / μm greater than that of the separation layer measured by SEM, and the average pore size change gradient of the separation layer measured by SEM is no greater than 3 nm / μm.

8. The asymmetric regenerated cellulose virus-removing filter membrane according to claim 1, characterized in that: The SEM measurement of the pre-filter layer shows a fiber diameter of 60–150 nm, and the SEM measurement of the separation layer shows a fiber diameter of 20–40 nm.

9. The asymmetric regenerated cellulose virus-removing filter membrane according to claim 1, characterized in that: The liquid inlet surface includes several strip-shaped first fibers, and the several first fibers are interconnected and adjacent first fibers surround each other to form a first hole; The average pore size of the first hole, as measured by SEM, is 350–2000 nm; and / or, The liquid outlet surface has several circular second holes; The average pore size of the second hole, as measured by SEM, is 25–45 nm.

10. The asymmetric regenerated cellulose virus-removing filter membrane according to claim 9, characterized in that: Of the second pores, those with a diameter greater than 30 nm as measured by SEM account for no more than 20% of the total number of second pores. Of the second type of pores, those smaller than the average pore diameter measured by SEM by 20 nm account for no more than 20% of the total number of second pores.

11. The asymmetric regenerated cellulose virus-removing filter membrane according to claim 9, characterized in that: In the second aperture, the proportion of second apertures with a deviation from the mean of no more than 25 nm is no less than 70%; the deviation from the mean is the absolute value of the difference between the SEM measured aperture of the second aperture and the SEM measured average aperture of the second aperture.

12. The asymmetric regenerated cellulose virus-removing filter membrane according to claim 1, characterized in that: The flux retention coefficient of the filter membrane is T, and T is not less than 0.4@(10g / L). Wherein, T not less than 0.4@(10g / L) means that in a system with a protein concentration of 10g / L, the flux retention coefficient T of the filter membrane is not less than 0.4; The value of T is calculated using the following formula: In the above formula, V is the filter membrane loading capacity, and the loading capacity when the flux decreases by a% is Va; V25 is the filter membrane loading capacity when the flux decreases by 25%, that is, V50 is the filter membrane loading capacity when the flux decreases by 50%, and V75 is the filter membrane loading capacity when the flux decreases by 75%.

13. The asymmetric regenerated cellulose virus-removing filter membrane according to claim 1, characterized in that: The top-wash LRV retention coefficient of the filter membrane is not less than 80%; the filter membrane flux at 75% decay is not less than 120 L / m³. 2 The protein yield of the filter membrane is not less than 98%; the flux of the filter membrane is not less than 60 L / (h·m). 2 @30psi.

14. The process for preparing the filter membrane according to any one of claims 1 to 13, characterized in that: The process includes the following steps: S1. Casting: The first casting solution and the second casting solution are sequentially cast onto a carrier to form a double-layer liquid film. The first casting solution comprises the following raw materials in parts by weight: 10-30 parts of a first film-forming polymer and 30-150 parts of a first solvent system. The second casting solution comprises the following raw materials in parts by weight: 10-20 parts of a second film-forming polymer and 70-300 parts of a second solvent system. The viscosity ratio of the first casting solution to the second casting solution is (2-20):

1. The solid content of the first casting solution is not less than 15%, and the surface tension of the first solvent system is not greater than 30 dyne / cm. S2. Phase separation and solidification: Immerse the double-layer liquid film in a coagulation bath to allow the casting liquid to separate and solidify until the double-layer liquid film is completely solidified. The coagulation bath is water or ethanol to obtain the raw film. S3. Regeneration: The biofilm is immersed in a regeneration bath for regeneration to obtain a regenerated biofilm.

15. The filter membrane preparation process according to claim 14, characterized in that: Step S2 specifically includes: S21. Pretreatment: Immerse the double-layer liquid membrane in a pretreatment bath for pretreatment. The pretreatment time is 0.5 to 10 seconds. The pretreatment bath is a 40 to 100% aqueous solution of solvent to obtain a pretreated membrane. S22. Curing: Immerse the pretreated membrane in a coagulation bath until the pretreated membrane is completely phase-separated and cured to obtain a green membrane. The coagulation bath is water or ethanol.

16. The process for preparing the filter membrane according to claim 14 or 15, characterized in that: The surface tension of the second solvent system is no greater than 30 dyne / cm; the surface tension of the first solvent system is less than that of the second solvent system.

17. The process for preparing the filter membrane according to claim 14 or 15, characterized in that: The viscosity of the first casting solution is 10,000 to 30,000 cp, and the viscosity of the second casting solution is 1,000 to 4,000 cp.

18. The process for preparing the filter membrane according to claim 15 or 16, characterized in that: The first film-forming polymer is at least one of cellulose diacetate, cellulose triacetate, cellulose propionate, cellulose acetate phthalate, cellulose acetate butyrate, and cellulose acetate propionate. The second film-forming polymer is at least one of cellulose diacetate, cellulose triacetate, cellulose propionate, cellulose acetate phthalate, cellulose acetate butyrate, and cellulose acetate propionate.

19. The process for preparing the filter membrane according to claim 14 or 15, characterized in that: The first solvent system is a mixture of a first good solvent and a first pore-forming agent in a mass ratio of (1-8):(5-22); The second solvent system is a mixture of a second good solvent and a second pore-forming agent in a mass ratio of (3-13):(4-17); The first good solvent is at least one selected from acetone, dioxane, dimethylacetamide, N-methylpyrrolidone, acetic acid, propionic acid, butyric acid, and valeric acid; The second good solvent is at least one selected from acetone, dioxane, dimethylacetamide, N-methylpyrrolidone, acetic acid, propionic acid, butyric acid, and valeric acid; The first pore-forming agent is at least one of ethanol, 1-propanol, hexafluoroisopropanol or trifluoroethanol; The second pore-forming agent is at least one of ethanol, 1-propanol, hexafluoroisopropanol or trifluoroethanol.

20. The process for preparing the filter membrane according to claim 14 or 15, characterized in that: In step S3, the regeneration bath is an aqueous sodium hydroxide solution, and the temperature of the regeneration bath is 20-40°C, with a regeneration time of 30-120 min.

21. The process for preparing the filter membrane according to claim 14 or 15, characterized in that: After step S3, step S4, crosslinking, is performed. The regenerated membrane is placed in a crosslinking agent for crosslinking treatment. After the crosslinking treatment is completed and the membrane is cleaned, the finished membrane is obtained. The crosslinking agent is at least one of halogenated epoxides, dihalogenated alkanes, and dihalogenated alcohols.

Images