An asymmetric hydrophilic PVDF filter membrane for virus removal and a preparation process and membrane filter thereof

Abstract

The application relates to an asymmetric hydrophilic PVDF filter membrane for removing viruses, a preparation process of the filter membrane and a membrane filter. The filter membrane comprises a porous body, the porous body has non-directional winding channels, the porous body comprises an inlet surface and an outlet surface, the porous body is connected with continuous fibers, the average pore diameter of the porous body gradually decreases from the inlet surface to the outlet surface as measured by SEM, the porous body comprises a pre-filter layer and a separation layer for intercepting colloidal gold with a diameter of 20 nm, the separation layer comprises the outlet surface, the average pore diameter of the outlet surface is 60-160 nm as measured by SEM, and the thickness of the separation layer is 15-45 mu m. The application further discloses a preparation process of the filter membrane, and further discloses a membrane filter comprising the filter membrane. The separation layer of the filter membrane has a large pore diameter and thickness, a long winding channel is used to ensure a high interception effect on small-size viruses, the separation layer with a large average pore diameter greatly reduces the protein adsorption rate, and thus high virus interception effect and protein yield are obtained.

Description

Technical Field

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

[0002] Membrane separation technology refers to a separation technique that uses a separation membrane as its core and external pressure or concentration gradient as the driving force to separate, concentrate, and purify components in a feed solution. Compared to conventional separation techniques, membrane separation technology offers high separation efficiency, requires no external reagents, and can separate systems that conventional techniques cannot. More importantly, membrane separation technology is a purely physical separation technique with mild separation conditions. In the biopharmaceutical field, this characteristic makes it less likely to cause denaturation of active substances in the feed solution, which are often expensive to produce. Therefore, membrane separation technology is particularly well-suited for separation processes in the biopharmaceutical industry.

[0003] The production processes of various biological agents are extremely complex, often requiring steps such as cultivation, purification, cleaning, passivation, extraction, freezing, and lyophilization. It is difficult to avoid introducing various viruses during these processes, and if these viruses are injected into patients along with the biological agents, the consequences could be disastrous. Therefore, both the new edition of the Chinese Pharmacopoeia and relevant documents such as Q5A "Viral Safety Assessment of Biological Products" issued by the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) explicitly require the viral safety of biological agents. When submitting drug applications, a report on the viral safety assessment test results of the biological agent must be attached to prove its viral safety. The necessity of viral removal in biological agents, combined with the characteristic of membrane separation technology that it does not easily cause denaturation of active substances during virus filtration, has led to the widespread application of membrane separation technology in biopharmaceutical companies.

[0004] Currently, common membrane materials for virus-free filter membranes include cellulose, PES, and PVDF. PVDF (polyvinylidene fluoride) is a semi-crystalline polymer with good mechanical properties, weather resistance, and chemical stability, making it one of the mainstream membrane materials for ultrafiltration and microfiltration membranes. Compared to cellulose-based membranes, PVDF membranes often have better mechanical properties. Furthermore, the tendency of cellulose membranes to shrink after drying necessitates storage and transportation in a moist state, resulting in higher costs and greater complexity. Compared to PES membranes, PVDF membranes often exhibit higher protein yields. This is because hydrophilically modified PVDF materials have better hydrophilicity than hydrophilically modified PES membranes, leading to lower protein adsorption rates—a significant advantage for biopharmaceutical companies.

[0005] For example, Japanese invention patent JP1984204911A discloses a regenerated cellulose membrane (RC membrane). This regenerated cellulose membrane has good clearance ability for HIV (approximately 100 nm), but its clearance ability for smaller viruses, such as hepatitis B virus (approximately 42 nm), nAnB hepatitis virus (30–60 nm), and mouse parvovirus (approximately 20 nm), which are 20–100 nm in size, is poor and can no longer meet the current stringent requirements for virus clearance. In addition, the cumbersome storage and transportation conditions of cellulose-based filter membranes also limit their development and application.

[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 retention. The other side of the pre-filtration layer and the other side of the separation layer are transitioned by continuous fibers. This PES membrane filters larger particles through the large-pore pre-filtration layer and further retains viruses through the small-pore separation layer, exhibiting good virus retention capacity, with a logarithmic removal rate of over 4. However, the poor hydrophilicity of PES material itself results in a often low protein yield, and since protein production costs are very high, low protein yield is an unacceptable drawback for biopharmaceutical companies.

[0007] For example, Chinese invention patent CN105980037B discloses a virus-removing membrane comprising a hydrophilized synthetic polymer for removing viruses from a protein-containing solution. The virus-removing membrane has a first surface for supplying the protein-containing solution and a second surface for discharging permeate through the membrane. In the cross-section of the wetted membrane, the region capturing colloidal gold with a diameter of 20 nm is located from the first side in an area of ​​25% to 85% of the membrane thickness, and the region capturing colloidal gold with a diameter of 15 nm is located from the first side in an area of ​​60% to 100% of the membrane thickness. It does not capture colloidal gold with a diameter of 10 nm. The pore size of this filter membrane is reduced and then kept constant, with the densest layer near the second side, resulting in good retention of colloidal gold with diameters of 20 nm or even 15 nm. However, although this filter membrane does not capture 10nm colloidal gold, its logarithmic removal rate for 10nm colloidal gold still reaches an LRV of 0.03–0.09 (i.e., the yield of 10nm colloidal gold is approximately 81–93%), according to this patent. Figure 5As recorded in the table, since 10nm colloidal gold is mainly used to characterize the retention effect of filter membranes on proteins (igG proteins, approximately 10nm), the yield of 10nm colloidal gold is only 81-93%, which means that the yield of igG proteins is also low. It is generally believed that the yield of igG proteins is lower than that of 10nm colloidal gold. This is because, even after hydrophilic modification, PVDF still has a certain degree of hydrophobicity. Combined with the small pore size regions of the filter membrane, it will have a strong adsorption effect on proteins, leading to a decrease in protein yield. In addition, this filter membrane is prepared by the pure TIPS method. The intrinsic characteristics of the pure TIPS method determine that the filter membrane often has a low porosity. For feed solutions with high protein concentrations, the high protein rejection rate combined with the high protein concentration can easily clog the filter membrane, leading to a decrease in flux and service life.

[0008] Based on the aforementioned problems, obtaining a filter membrane that has both a high virus rejection rate and a high protein yield for small-sized viruses is an urgent issue to be addressed. Summary of the Invention

[0009] The application provides an asymmetric hydrophilic PVDF filter membrane for virus removal, its preparation process, and a membrane filter. The separation layer of the filter membrane has a large average pore size and thickness, and the long tortuous path ensures a high retention effect for small-sized viruses. The separation layer with a large average pore size significantly reduces the protein adsorption rate, thereby achieving a high virus retention effect and a high protein yield.

[0010] This application provides an asymmetric hydrophilic PVDF filter membrane for virus removal, its preparation process, and the technical solution adopted by the membrane filter is as follows:

[0011] In a first aspect, this application provides an asymmetric hydrophilic PVDF filter membrane for virus removal, employing the following technical solution:

[0012] A PVDF asymmetric hydrophilic filter membrane for virus removal comprises a porous body with non-directional tortuous pathways. The side of the porous body closer to the feed solution is the inlet surface, and the side farther from the feed solution is the outlet surface. The porous body transitions with continuous fibers in the membrane thickness direction. The average pore size of the porous body, measured by SEM, gradually decreases from the inlet surface to the outlet surface. The porous body includes a pre-filtration layer and a separation layer for retaining colloidal gold with a diameter of 20 nm. The separation layer includes the outlet surface, and the average pore size of the outlet surface, measured by SEM, is 60–160 nm. The thickness of the separation layer is 15–45 μm.

[0013] Preferably, the protein yield of the filter membrane is not less than 95%. More preferably, the protein yield of the filter membrane is not less than 98%.

[0014] While the hydrophilic PVDF membrane, modified with the aforementioned technology, exhibits superior hydrophilicity compared to commercially available hydrophilic modified PES membranes, it still falls short of cellulose-based membranes. Therefore, hydrophilic PVDF membranes often still adsorb proteins, leading to a decrease in protein yield. Furthermore, protein adsorption can easily clog the membrane's pore structure, resulting in a rapid reduction in membrane flux. For biopharmaceutical companies, the high production costs of active protein substances make low protein yields unacceptable, and the rapid decline in flux also significantly impacts production efficiency and must be avoided.

[0015] Building upon the above, for virus-removing membranes, to reduce the risk of virus leakage, it is necessary to improve the membrane's virus filtration capacity. Furthermore, with increasingly stringent virus risk control requirements, the membrane's ability to filter small-sized viruses becomes particularly crucial. However, the size of small viruses is quite similar to that of proteins; for example, the diameter of a typical small parvovirus, mouse parvovirus, is approximately 20 nm, and the diameter of a typical small-sized phage, PP7, is approximately 27 nm. In contrast, the diameter of a typical protein, such as IgG (immunoglobulin), is approximately 10 nm. Therefore, a high filtration capacity for small-sized viruses often implies a high filtration capacity for proteins, leading to a decrease in protein yield. Furthermore, even after hydrophilic modification, PVDF still retains some hydrophobicity, maintaining some adsorption for proteins. This means that improving the virus filtration capacity of PVDF membranes often comes at the cost of a decreased protein yield.

[0016] To address the aforementioned problems, the inventors of this application unexpectedly discovered that for hydrophilic PVDF filter membranes, when the membrane has an asymmetric structure (with the largest average pore size measured by SEM at the inlet surface and the smallest average pore size measured by SEM at the outlet surface), and the outlet pore with the smallest average pore size measured by SEM has a relatively large diameter, approximately 60–160 nm; and the separation layer of the filter membrane has a relatively large thickness, approximately 15–45 μm, the filter membrane not only exhibits a high retention efficiency for small-sized viruses but also achieves a high protein yield. This differs from general understanding, which generally holds that for PVDF filter membranes, in order to efficiently retain small-sized viruses of around 20 nm, the pore size of the filter membrane should be relatively small, at least the pore size at the outlet surface with the smallest pore should not exceed 40 nm. If the average SEM pore size of the outlet surface of the filter membrane is greater than 40 nm, or even reaches a relatively large 60–160 nm (due to the pore size gradient of the filter membrane, the pore size of the separation layer should be even larger), the retention efficiency of the filter membrane for small-sized viruses is often poor, failing to meet practical application requirements. The fact that the filter membrane in this application has a high virus rejection rate and a high protein yield that are generally considered difficult to obtain simultaneously for PVDF filter membranes is quite an unexpected result.

[0017] This may be because, although PVDF material has a better hydrophilic effect after hydrophilic treatment (better than hydrophilic modified PES material), it still has a certain degree of hydrophobicity. Therefore, the pore walls of the tortuous pathways inside the filter membrane still have a certain adsorption capacity for proteins in the feed solution (this can be seen from the fact that the protein yield of the PVDF filter membrane in CN105980037B after hydrophilic modification is only about 81-93%), making it very difficult to improve the protein yield.

[0018] Generally, except for viruses, the actual pore size of the filter membrane is larger than the size of the protein. Therefore, the pore size screening effect of the filter membrane for small proteins is not significant, even if the filter membrane has a long and tortuous path inside, its protein screening and retention is still relatively small. However, the filter membrane in CN105980037B can even retain colloidal gold as small as 15 nm, which is very close to the size of a protein. In order to obtain efficient retention of small-sized viruses (colloidal gold), the filter membrane must have regions with a small average pore size measured by SEM. Although the pore size of these regions with a small average pore size measured by SEM is still larger than the size of the protein, these small pore regions are more likely to generate a large amount of protein adsorption, leading to a decrease in protein yield. In other words, for hydrophilic PVDF filter membranes, although reducing the pore size can improve its virus retention effect, with further reduction of the pore size, the adsorption of proteins by the smaller pore structure of the filter membrane may increase rapidly, resulting in a rapid decrease in protein yield.

[0019] When the liquid outlet surface, which has the smallest average pore size measured by SEM, also has a relatively large average pore size (60-160 nm), the pore size of the internal pore structure of the filter membrane is relatively large. This not only makes it difficult to achieve a protein sieving effect, but also significantly reduces the adsorption force on proteins. The adsorption force of the filter membrane on proteins may need to be greater than a certain threshold to adsorb proteins onto the pore walls. Therefore, the decrease in adsorption force caused by the increase in the average pore size measured by SEM at the liquid outlet surface will significantly improve the protein yield, reaching 95% or even 98% or more. This is a huge improvement for currently common PVDF filter membranes.

[0020] It's important to note that for hydrophilic PVDF filter membranes, which inherently possess a certain protein adsorption rate, the adsorption of proteins is inevitable. Therefore, the closer the protein yield gets to 100%, the more difficult it becomes, and the difficulty doesn't increase linearly. For example, increasing the protein yield from 80% to 90% is relatively easy, while increasing it from 90% to 95% is drastically more difficult, and increasing it from 95% to 98% is even more challenging. Once the protein yield has reached over 98%, further increases become exponentially more difficult.

[0021] Building upon a high protein yield, this method increases the thickness of the separation layer, which has virus-retention capabilities, and extends the path of the tortuous pathways within the separation layer. Multiple layers of relatively large pore structures are continuously stacked on top of each other in the membrane thickness. The non-overlapping portions of these stacked pore structures are much smaller than the average pore size measured by SEM (the solid portion of the lower layer blocks the pore structure of the upper layer, preventing the passage of the feed solution and various particulate matter within it; therefore, the effective pathway size for virus passage is often smaller than the actual size of the filter membrane pores). Since the effective pathway for the feed solution formed by the stacked pore structures is much smaller than the average pore size measured by SEM, increasing the thickness of the separation layer in conjunction with appropriately increasing its pore size further enhances the path of the stacked pore structures. This ensures that the stacked pore structure has a low protein retention rate but a high retention rate for slightly larger viruses. This effect is achieved by simultaneously increasing both the pore size and the path length of the effective pathway after stacking, thus simultaneously obtaining high protein yield and high virus retention rate. This is quite unexpected, contrary to the common belief that large-pore hydrophilic PVDF filter membranes cannot retain small-sized viruses.

[0022] Of course, a larger SEM average pore size at the liquid outlet is not always better. As the SEM average pore size at the liquid outlet increases, on the one hand, the overall SEM average pore size of the filter membrane also increases. Even if the tortuous path is extended and the multilayer pore structure is continuously stacked in terms of membrane thickness, an excessively large SEM average pore size still cannot ensure that the effective path size after stacking is small enough to trap small viruses, and the filtration effect on small viruses cannot be guaranteed. Furthermore, the effect of further increasing the SEM average pore size on the adsorption force on proteins has a diminishing marginal effect, so further increasing the SEM average pore size has no significant effect on improving protein yield. On the other hand, as the SEM average pore size of the filter membrane increases, the mechanical strength of the filter membrane may decrease. When subjected to greater pressure from the filtered liquid during use, it may deform under pressure, leading to deformation of the internal pore structure of the filter membrane, and may even cause the filter membrane to become clogged or ruptured. Therefore, for a specific hydrophilic PVDF filter membrane, considering both protein yield and the retention effect on small-sized viruses, the average pore size measured by SEM at the liquid outlet of the filter membrane is preferably 60–160 nm. This is a completely different technical approach from the common practice of controlling the average pore size measured by SEM at the liquid outlet of the filter membrane to be less than 40 nm in order to obtain a high virus retention rate, and it is also a very unexpected result.

[0023] In this application, the separation layer is defined as the region capable of retaining 20nm colloidal gold. Retaining 20nm colloidal gold means that the prepared filter membrane retains 20nm colloidal gold, and the distribution of 20nm colloidal gold in the membrane thickness direction is measured. The distribution measurement can be performed according to the test method in Chinese Patent CN105980038B - Membrane for Virus Removal: the filter membrane after retaining colloidal gold is sliced, and the brightness distribution of multiple sites in the colloidal gold-stained portion at the interface of the slice is measured. Colloidal gold is opaque; therefore, the brightness displacement value can be used to characterize the amount of colloidal gold captured. It should be noted that, if necessary, background noise can be removed by the brightness distribution. Then, a graph is generated with the horizontal axis representing membrane thickness and the vertical axis representing brightness displacement; thus, the region where 20nm colloidal gold is retained in the membrane thickness direction is obtained. See the schematic diagram for details. Figure 6 Furthermore, by observing the SEM images of the filter membrane thickness cross-section after the 20nm colloidal gold retention experiment, it can also be seen that the separation layer region of the filter membrane is significantly blocked by the 20nm colloidal gold, resulting in the disappearance of the pore structure. See the schematic diagram for details. Figure 7 .

[0024] Due to various reasons (such as adsorption and blind pore retention), even in areas like the pre-filter layer, a very small amount of colloidal gold may remain. However, this area cannot be considered as effectively retaining 20nm colloidal gold. Therefore, in this application, when measuring the brightness shift value in the film thickness direction, the difference between the brightness constant 255 and the measured brightness distribution is used to obtain the brightness shift spectrum. The maximum shift peak in this spectrum is the location where the amount of 20nm colloidal gold retained is the maximum. Areas where the ratio to the maximum shift peak is less than 10% are considered to be only a small amount of colloidal gold residue or error, and are not considered to be areas where 20nm colloidal gold is actually retained.

[0025] In other words, although there are some regions along the membrane thickness direction where a small amount of 20nm colloidal gold is trapped, the amount of 20nm colloidal gold trapped in these regions is extremely small and cannot be considered as regions where 20nm colloidal gold is truly trapped; they are merely errors or small residues. Therefore, the separation layer in the filter membrane should be a region that continuously and extensively captures 20nm colloidal gold along the membrane thickness direction.

[0026] It is understood that the so-called "non-directional tortuous pathways" in this application refer to groove structures with irregular orientation and / or discretely distributed pore structures within a porous body, and these non-directional tortuous pathways are interconnected, so that the feed liquid can penetrate the filter membrane through the interconnected pathways. Viruses and large particles in the feed liquid are trapped in the non-directional tortuous pathways inside the filter membrane or the porous body, thereby achieving the effect of filtering out viruses.

[0027] In this application, "continuous fiber transition" means that all fibers in the porous body along the membrane thickness direction are integrally formed and interconnected as a whole. No additional adhesives or other substances are needed to bond the fibers together. Unless torn or peeled by external force, the three-dimensional network of fibers will not separate from each other. Simultaneously, the continuously transitioning three-dimensional network of fibers is also interconnected with the first and second outer surfaces.

[0028] The parameters such as pore size, layer thickness, and fiber diameter mentioned in this application refer to the average values ​​calculated after characterizing the membrane structure using a scanning electron microscope (SEM) and then measuring them using computer software (such as Matlab, NIS-Elements, etc.) or manually. Significantly smaller or larger dimensions are not considered during measurement. It should be noted that porosity can also be calculated using computer software (such as Matlab, NIS-Elements, etc.) or measured by gravimetric methods. Regarding the testing of average pore size using SEM, in addition to analyzing SEM images, the average pore size of each layer can be directly analyzed using an SEM average pore size distribution analyzer, or the average pore size can be tested using the bubble pressure method. The above measurement methods for each parameter are merely examples; it is understood that those skilled in the art can obtain these parameters using other measurement methods.

[0029] Optionally, the separation layer includes separation fibers, which are interconnected to form a three-dimensional network structure of the separation layer, and the average diameter of the separation fibers measured by SEM is 30-50 nm; the pre-filter layer includes support fibers, which are interconnected to form a three-dimensional network structure of the pre-filter layer, and the average diameter of the support fibers measured by SEM is 30-55 nm.

[0030] By adopting the above technical solution, since the average pore size measured by SEM at the liquid outlet surface in this application is relatively large, the average pore size measured by SEM in the separation layer and the pre-filtration layer is also relatively large. For the separation layer, if the average diameter measured by SEM of the separation fibers is too small, it may lead to insufficient self-support performance of the separation layer, causing the pore structure to collapse under external pressure. Once the pore structure of the separation layer collapses, it may cause the separation layer to lose its virus retention capacity (whether the separation layer is squeezed and blocked or ruptured, it will lose its virus retention capacity).

[0031] Furthermore, the separation layer plays a major role in retaining proteins and viruses in the filter membrane, while the solid portion of the separation layer is responsible for protein adsorption. The separation fibers in the solid portion of the separation layer intersect and connect, forming a three-dimensional network structure with a porous structure. The separation fibers act as the pore walls of this porous structure. When the pore structure of the separation layer remains unchanged, the distance between the separation fibers is relatively large, resulting in a larger pore size. This makes it less likely to reach the protein adsorption threshold, thus reducing protein adsorption and increasing protein yield. However, once the pore structure of the separation layer collapses, the distance between the separation fibers shortens significantly, forming a smaller pore structure. Even if the separation layer still has a good virus retention effect, its protein yield will decrease dramatically.

[0032] If the average diameter of the separating fibers measured by SEM is too large, although the pore structure of the separating layer can be better supported, the feed liquid mainly encounters resistance from the solid part when flowing through the filter membrane. An excessively large average diameter of the separating fibers measured by SEM is likely to lead to excessive resistance to the feed liquid, thereby reducing the flux of the filter membrane. Therefore, the average diameter of the separating fibers measured by SEM should not be too large or too small.

[0033] Compared to the separation layer, the pre-filter layer, with its larger SEM-measured average pore size, exhibits lower retention rates for both viruses and proteins. However, the larger pore size of the pre-filter layer often implies poorer mechanical properties. As the area where the filter membrane directly contacts the feed liquid, the pre-filter layer is easily deformed under pressure. Deformation due to pressure can lead to collapse of the pore structure, resulting in a decrease in membrane flux and loading. Furthermore, the reduced distance between supporting fibers after pore collapse leads to more significant protein retention and adsorption, reaching the adsorption threshold and causing a decrease in protein yield. Therefore, the SEM-measured average diameter of the supporting fibers should not be too small to provide better support for the pore structure of the pre-filter layer. However, the average diameter of the supporting fibers measured by SEM should not be too large. This is because a large average diameter of the supporting fibers often indicates increased resistance of the pre-filtration layer to the feed liquid, leading to a decrease in flux. Furthermore, a large average diameter of the supporting fibers means a higher proportion of the solid portion of the pre-filtration layer, naturally reducing the proportion of the pore structure used to trap and accommodate large particles. Since the filter membrane in this application has a relatively thick separation layer, if large particles leak from the pre-filtration layer, they are likely to be screened and trapped by the separation layer or by the long, tortuous path, leading to blockage of the separation layer and a rapid decrease in filter membrane flux and load. Therefore, the average diameter of the supporting fibers measured by SEM should also not be too large or too small.

[0034] Optionally, the ratio of the average diameter of the support fiber measured by SEM to the average diameter of the separated fiber measured by SEM is 0.8 to 1.5.

[0035] By adopting the above technical solutions, it is generally believed that there is a positive correlation between the pore size of the filter membrane and the size of the fiber structure constituting the filter membrane. Therefore, the pre-filtration layer, with a larger average pore size measured by SEM, should have a larger fiber structure, while the separation layer, with a smaller average pore size measured by SEM, should have a smaller fiber structure. That is, approximating the asymmetry of the filter membrane's pore size, the fiber diameter of the filter membrane should also be asymmetrically distributed. However, the inventors of this application unexpectedly discovered that the filter membrane in this application has approximately the same fiber size in the membrane thickness direction, with very small differences in fiber size. That is, the average diameter of the supporting fibers in the pre-filtration layer measured by SEM is approximately the same as the average diameter of the separating fibers in the separation layer measured by SEM, and no obvious asymmetric structure is exhibited. This may be one of the reasons why the filter membrane has a high protein yield. Compared to the general filter membrane where the fiber size gradually decreases from the pre-filtration layer to the separation layer, forming a more obvious asymmetric structure, the fiber size of the filter membrane in this application does not change significantly from the pre-filtration layer to the separation layer. Therefore, the separation layer in this application has a larger fiber structure.

[0036] Since finer fibers have a larger specific surface area and thus a higher protein adsorption rate, the separation layer composed of separation fibers with a larger average diameter as measured by SEM in this application exhibits a lower protein adsorption rate, thereby further improving the protein yield of the filter membrane. Furthermore, the specific thick and large-pore separation layer structure in this application places higher demands on the self-supporting performance of the separation layer. Separation fibers with a larger average diameter as measured by SEM can form a more robust three-dimensional network structure with stronger self-supporting capabilities, thereby improving the mechanical strength of the separation layer and even the filter membrane, and further enhancing the pressure resistance of the filter membrane.

[0037] Optionally, the average length of the separated fiber measured by SEM is 80–150 nm, and the average length of the supporting fiber measured by SEM is 120–180 nm.

[0038] By adopting the above technical solution, since the pore structure of the separation layer is mainly formed by the separation fibers, the average length of the separation fibers measured by SEM can characterize the pore structure size of the separation layer to a certain extent. Similarly, the pore structure of the pre-filtration layer is mainly formed by the support fibers, and the average length of the support fibers measured by SEM can also characterize the pore structure size of the pre-filtration layer to a certain extent. Therefore, both the average length of the separation fibers and the average length of the support fibers measured by SEM can characterize the pore structure of the filter membrane to a certain extent.

[0039] Furthermore, it should be noted that since the average diameters of the separating fibers and the supporting fibers measured by SEM are similar, and their average lengths are also similar, the aspect ratios of the separating and supporting fibers are not significantly different. This means that the self-supporting performance of the filter membrane changes little in the thickness direction, and there are no obvious weak points in the mechanical properties of the filter membrane in the thickness direction. The filter membrane is less likely to deform due to weak points when subjected to external pressure, thus achieving higher pressure resistance. Neither the separating nor the supporting fibers are easily compressed and brought closer together under pressure, which would cause the pore structure to collapse. This reduces the possibility of increased protein adsorption due to pore structure collapse, thereby achieving a higher protein yield.

[0040] Optionally, the ratio of the average pore size measured by SEM of the separation layer to the average diameter measured by SEM of the separation fiber is 2 to 7; the ratio of the average pore size measured by SEM of the pre-filter layer to the average diameter measured by SEM of the support fiber is 3.5 to 8.

[0041] Preferably, the ratio of the average pore size measured by SEM of the separation layer to the average diameter measured by SEM of the separation fiber is 3 to 5; the ratio of the average pore size measured by SEM of the pre-filter layer to the average diameter measured by SEM of the support fiber is 4.5 to 6.5.

[0042] By adopting the above technical solutions, both the separation layer and the pre-filtration layer require stronger support to ensure the dimensional stability of the pore structure under high pressure, given the relatively large average pore size measured by SEM and the high pressure (30 psi or even 50 psi) exerted on the virus-removing membrane during use. Since the fiber structure plays a major supporting role in the pore structure, the ratio of the pore structure size to the fiber size largely characterizes the pressure resistance of the pore structure. This is because larger pore structures often require greater support, and greater support often requires coarser fiber structures; therefore, larger pore structures typically require coarser fiber structures.

[0043] For the separation layer, if the ratio of the average pore size measured by SEM to the average diameter measured by SEM of the separation fibers is too high, the pore structure of the separation layer often cannot be well supported and is prone to deformation under pressure. Since the separation layer is the main factor affecting virus retention and protein yield, insufficient pressure resistance of the pore structure of the separation layer can easily lead to a decrease in the virus retention and protein yield of the filter membrane. If the ratio of the average pore size measured by SEM to the average diameter measured by SEM of the separation layer is too low, it indicates that the solid portion of the separation layer accounts for too high. When the feed liquid flows within the separation layer, it encounters greater resistance from the solid portion, which can easily lead to a significant decrease in flux and affect filtration efficiency.

[0044] For the pre-filter layer, if the ratio of the SEM average pore size of the pre-filter layer to the SEM average diameter of the supporting fibers is too high, the pore structure of the pre-filter layer often cannot be well supported. Once the pore structure of the pre-filter layer collapses, it will lose its ability to retain large particles in the feed liquid, and the filter membrane will easily become clogged. If the ratio of the SEM average pore size of the pre-filter layer to the SEM average diameter of the supporting fibers is too low, it also indicates that the proportion of the solid part of the pre-filter layer is too high, resulting in insufficient loading of the pre-filter layer, which cannot effectively retain and accommodate large particles in the feed liquid, leading to a decrease in flux.

[0045] Optionally, the ratio of the average pore size measured by SEM of the separation layer to the average length measured by SEM of the separation fiber is 1.1 to 1.4; the ratio of the average pore size measured by SEM of the pre-filter layer to the average length measured by SEM of the support fiber is 1.2 to 1.7.

[0046] By adopting the above technical solution, since the pore structure is mainly formed by surrounding fiber structures, the more fibers that make up the pores, the shorter the fiber length (the outer edge of the pore can be considered as a polygonal structure; when the pore size remains constant, the more polygonal sides that make up the pore, the shorter the side length). The ratio of the pore size to the fiber length is larger, and correspondingly, the pore structure is closer to a circle. Because the fluid experiences less resistance in a circular channel, the filter membrane often has a higher flux. Furthermore, for pores of the same area, the perimeter of a circle is smaller than that of a polygon. Therefore, the closer the pore structure is to a circle, the less contact there is between the feed liquid and the pore wall, the lower the protein adsorption rate, and the higher the protein yield of the filter membrane. Therefore, whether it is a separation layer or a pre-filtration layer, the ratio of the pore size to the fiber length should not be too low to obtain higher flux and protein yield.

[0047] However, regardless of whether it's the separation layer or the pre-filtration layer, the ratio of pore size to fiber length should not be too high. This is because a high ratio indicates a large number of fibers forming the pores. Although the pore structure is closer to a circle, the proportion of solid fibers is also often increased. When the feed liquid flows through the filter membrane, although the closer-to-circular pores reduce the resistance to the feed liquid, the greater resistance from the solid fiber structure can lead to a decrease in flux. Therefore, the ratio of pore size to fiber length needs to be controlled within a reasonable range for both the separation layer and the pre-filtration layer. For the separation layer, this ratio is approximately 1.1 to 1.4, and for the pre-filtration layer, it is approximately 1.2 to 1.7.

[0048] Optionally, the average pore size measured by SEM on the liquid inlet surface is 200–500 nm, the ratio of the average pore size measured by SEM on the liquid inlet surface to the average pore size measured by SEM on the liquid outlet surface is 1.2–6, and the specific surface area of ​​the porous body is 6–12 m². 2 / g.

[0049] By adopting the above technical solution, the ratio of the average pore size measured by SEM at the inlet surface to the average pore size measured by SEM at the outlet surface of the filter membrane largely characterizes the degree of asymmetry of the filter membrane in the thickness direction. For the filter membrane with a large thickness and large pore size separation layer specific to this application, the degree of asymmetry in the membrane thickness has a significant impact on the performance of the filter membrane. If the asymmetry of the filter membrane is too high (e.g., the ratio of the average pore size measured by SEM at the inlet surface to the average pore size measured by SEM at the outlet surface is greater than 5), it indicates that the pore structure size on the side of the filter membrane near the outlet surface is too large. On the one hand, given the large average pore size measured by SEM at the outlet surface of the filter membrane, it may be impossible to guarantee the retention effect of the filter membrane for small-sized viruses. On the other hand, the overall pore structure size of the filter membrane is large, which may cause structural collapse when subjected to large external forces, leading to a decrease in the virus retention effect and protein yield of the filter membrane. If the asymmetry of the filter membrane is too low (e.g., the ratio of the average pore size measured by SEM on the inlet surface to that on the outlet surface is less than 1.5), it indicates that the pore structure of the pre-filtration layer of the filter membrane is small, which may not be able to ensure the retention and containment of large particles, leading to filter membrane clogging and a decrease in flux. Therefore, the ratio of the average pore size measured by SEM on the inlet surface to that on the outlet surface of the filter membrane must be strictly controlled between 1.2 and 6. This allows for the formation of a separation layer with a small gradient, large pore size, and large thickness, effectively retaining viruses and reducing protein adsorption, thereby obtaining a filter membrane with high virus retention efficiency and high protein yield.

[0050] In addition, since the pore walls of the pore structure are the primary adsorbents of proteins, the specific surface area of ​​the filter membrane can largely characterize the amount of pore walls. If the specific surface area of ​​the filter membrane is too large, the adhesion of the membrane to proteins increases, leading to a decrease in protein yield. Controlling the specific surface area of ​​the filter membrane within a relatively small range can ensure a high protein yield. Furthermore, because filter membranes with low specific surface areas adsorb less protein, they are less prone to clogging, resulting in slower flux decay and the ability to maintain a high flux for a longer period.

[0051] It should be noted that the specific surface area in this application was measured by the BET specific surface area test method.

[0052] Optionally, the average pore size variation gradient of the separation layer measured by SEM is no greater than 8 nm / μm, and the average pore size of the separation layer measured by SEM is 100–200 nm.

[0053] By adopting the above technical solution, the average pore size of the filter membrane measured by SEM is relatively large, and the ratio of the average pore size measured by SEM of the filter membrane inlet and outlet is relatively small, which to a certain extent reflects the high overall symmetry of the filter membrane.

[0054] The gradient of the average pore size variation and the average pore size measured by SEM of the separation layer can characterize the pore size and pore size variation trend of the separation layer. Due to the relatively large thickness of the separation layer (15–45 μm), the pore structure of the separation layer has a long-path stacked structure in the membrane thickness direction. If the average pore size measured by SEM of the separation layer is too large, small-sized viruses can still easily penetrate through the stacked pore structure, resulting in the filter membrane's virus retention effect failing to meet usage requirements. If the average pore size measured by SEM of the separation layer is too small, although it can achieve good retention of small-sized viruses, proteins and small-sized viruses are not significantly different in size. The stacked, long-path, small-pore structure of the separation layer can easily lead to excessive protein retention, resulting in a decrease in protein yield. Therefore, considering the thickness of the separation layer (15–45 μm) and its average pore size measured by SEM (100–200 nm), it is possible to ensure that the filter membrane simultaneously achieves a high virus retention rate and protein yield.

[0055] Furthermore, the average pore size measured by SEM of the separation layer is relatively large. Based on this, the gradient of the overall average pore size variation measured by SEM of the separation layer is small, not exceeding 8 nm / μm. This indicates that there are no abrupt changes in pore size in the thickness direction of the separation layer. Abrupt changes in pore size, due to their dense and rapidly changing layered pore structure, would extensively trap particles of different sizes, such as viruses and proteins, leading to a decrease in flux and protein yield. Therefore, the separation layer with a small gradient of average pore size variation and a large average pore size measured by SEM has almost no high adsorption region for proteins in the thickness direction, ensuring that the filter membrane has a high protein yield.

[0056] It is understandable that the gradient of the average pore size change measured by SEM = (average pore size measured by SEM on the first porous surface - average pore size measured by SEM on the second porous surface) / thickness. The larger the value, the faster the pore size changes, and the smaller the value, the smaller the pore size changes.

[0057] Optionally, the side of the separation layer closest to the liquid inlet is considered 0%, and the liquid outlet is considered 100%. The separation layer is divided into ten equal parts along the film thickness direction. The ratio of the average pore size measured by SEM in the region where the film thickness of the separation layer is 0-10% to the average pore size measured by SEM in the region where the film thickness of the separation layer is 40-50% is K1, and K1 is not greater than 2.2.

[0058] By employing the above technical solution, the pore structure of the separation layer has a significant impact on the overall performance of the filter membrane. If the pore structure of the separation layer contains regions of rapid pore size change, regardless of whether these regions are located above or below the separation layer, it may lead to a substantial decrease in membrane flux and protein yield. Therefore, although the overall average pore size change gradient measured by SEM for the separation layer is relatively small, if there are both rapidly changing and slowly changing pore size regions along the thickness direction of the separation layer, even though the combination of the two still results in a relatively small average pore size change gradient measured by SEM, the membrane flux and protein yield may still experience a significant decrease.

[0059] The region of 0–10% of the separation layer thickness is considered the upper 10% of the separation layer thickness, while the region of 40–50% is considered the middle 50% of the separation layer thickness. The ratio K1 of the SEM average pore size of these two regions characterizes the variation trend of the SEM average pore size in the upper and middle parts of the separation layer. Since K1 is not greater than 2.2, it means that the pore size change rate in the upper part of the separation layer is slower, and there are no regions with rapid pore size changes. This ensures that the upper part of the separation layer does not have high resistance areas for the feed solution or high adsorption areas for proteins. Combined with the small gradient of the overall SEM average pore size change of the separation layer, the possibility of a decrease in membrane flux and protein yield due to regions with rapid pore size changes can be greatly reduced.

[0060] Understandably, for filter membranes, the side closer to the inlet liquid surface is considered "upper," while the side closer to the outlet liquid surface is considered "lower." That is, the upstream of the liquid flow direction is "upper," and the downstream of the liquid flow direction is "lower."

[0061] Optionally, the ratio of the average pore size measured by SEM in the region where the separation layer film thickness is 40-50% to the average pore size measured by SEM at the liquid outlet surface is K2, where K2 is not greater than 1.5.

[0062] By adopting the above technical solution, the region of 40-50% of the separation layer thickness is considered the upper 50% thickness region of the separation layer, and the liquid outlet surface is considered the middle 100% thickness region of the separation layer. The ratio K2 of the SEM average pore size of these two regions can characterize the SEM average pore size variation trend in the middle and bottom of the separation layer. Since K2 is not greater than 1.5, it means that the pore size change rate in the lower half of the separation layer is slower, and there are no regions with rapid pore size changes. Therefore, the lower half of the separation layer does not have high resistance regions for the feed liquid or high adsorption regions for proteins. Combined with the relatively small SEM average pore size change gradient of the separation layer as a whole and the slower pore size change rate in the upper half of the separation layer, the possibility of a decrease in membrane flux and protein yield due to rapid pore size changes in various regions of the separation layer can be greatly reduced.

[0063] Optionally, K1 is greater than K2, and the ratio of K1 to K2 is 1.05 to 1.6.

[0064] By adopting the above technical solution, K1 is greater than K2 and the ratio between the two is 1.05 to 1.6, indicating that the average pore size change rate of the separation layer measured by SEM is fast at first and then slows down, but the overall difference is not large. Combined with the small gradient of the overall average pore size change of the separation layer measured by SEM, it indicates that the separation layer has high overall symmetry and only small gradient asymmetry.

[0065] Furthermore, because the average pore size of the separation layer, as measured by SEM, decreases relatively quickly initially, the size of the effective pathways formed by the stacked pore structure decreases rapidly, resulting in good retention of small-sized viruses and ensuring a high virus rejection rate for the filter membrane. At this stage, the effective pathway size is still relatively large, allowing even smaller proteins to pass through smoothly. However, as the average pore size of the separation layer decreases further, the size of the effective pathways also gradually decreases. If the effective pathway size becomes too small, it will not only trap viruses but may also trap proteins, leading to a decrease in protein yield. The slower pore size change rate in the lower half of the separation layer indicates further improved symmetry, resulting in smaller changes in the size of the effective pathways formed by the stacked pore structure. This ensures good virus retention while reducing protein retention, thus achieving a good protein yield and a high virus rejection rate for the filter membrane.

[0066] Optionally, the pre-filter layer includes the liquid inlet surface, the thickness of the pre-filter layer is 1-15 μm, the average pore size of the pre-filter layer measured by SEM is 150-300 nm, and the average pore size variation gradient of the pre-filter layer measured by SEM is 5-20 nm / μm.

[0067] By adopting the above technical solution, the pre-filtration layer thickness of the filter membrane in this application is relatively small (compared to the thicker separation layer), and the average pore size measured by SEM of the pre-filtration layer is not significantly different from that of the separation layer. This is because the separation layer thickness in this application is relatively large, and the feed liquid encounters greater resistance from the solid portion of the separation layer when flowing through its long pore structure, leading to a decrease in flux. If the pre-filtration layer thickness is too large, the flow path of the feed liquid will be further extended, resulting in even greater resistance and potentially a significant decrease in flux. Therefore, in order to achieve a higher flux in the filter membrane, the thickness of the pre-filtration layer should be appropriately reduced, given the already large separation layer thickness.

[0068] Based on this, it is necessary to ensure that the pre-filtration layer effectively retains large particles to reduce the possibility of filter membrane clogging. The pore structure of the separation layer near the pre-filtration layer has a larger size, which can also effectively retain and accommodate larger particles of impurities. In other words, the specific large-pore separation layer in this application also has a certain pre-filtration effect, which can supplement and assist the pre-filtration layer. Therefore, even if the pre-filtration layer of the filter membrane in this application is relatively thin, with the synergistic effect of the specific large-thickness, large-pore separation layer, it can still effectively retain large particles without causing rapid clogging of the filter membrane.

[0069] Furthermore, the relatively small average pore size (150–300 nm) measured by SEM in the pre-filter layer, coupled with a relatively large gradient in the average pore size (5–20 nm / μm), indicates that the pore size of the pre-filter layer decreases at a relatively rapid rate near the liquid outlet. This region, where the pore size decreases rapidly, combined with the already small pore structure (compared to large particles), effectively traps large particles. Since the pore size in this region is still relatively large (compared to viruses and proteins), it does not trap smaller viruses and proteins. Large particles not trapped in the upper part of the pre-filter layer are likely trapped in the lower part, where the pore size is smaller, thus reducing their impact on the separation layer.

[0070] It should be noted that, due to the relatively small thickness of the pre-filter layer, it is inevitable that a small amount of large particles will leak out. The front half of the separation layer, with its relatively large pore size (smaller than that of viruses and proteins compared to large particles) and small gradient, can intercept and contain the large particles leaking from the pre-filter layer. Because the pore size of the separation layer is larger on the side closer to the pre-filter layer, the capacity is larger, and a small amount of large particles will not clog the separation layer.

[0071] Optionally, the flux of the filter membrane is not less than 50 L·h -1 ·m -2 @30psi, the filter membrane has a log removal rate of greater than 2 for PP7 phage; the tensile strength of the filter membrane is 5-20 MPa.

[0072] Optionally, the elongation at break of the filter membrane is 50-200%.

[0073] By adopting the above technical solutions, the internationally recognized standard is that membrane filters need to achieve a logarithmic removal rate of 4 or higher for small-sized viruses (such as the typical PP7 bacteriophage, and the indicator virus for small-sized viruses in PDA guidance document TR41). Currently, the number of membranes in membrane filters is not limited to only one. For membrane filters with only one membrane, the membrane in this application can achieve a high logarithmic removal rate by adjusting parameters such as the thickness and pore size of the separation layer, such as a logarithmic removal rate of not less than 4 for PP7. For membrane filters with multiple membranes, the membrane in this application can achieve a relatively low logarithmic removal rate (although multiple membranes used in series can still ensure a logarithmic removal rate of 4 or higher) and a high flux by adjusting parameters such as the thickness and pore size of the separation layer.

[0074] Furthermore, it is generally believed that filter membranes with a large separation layer ratio tend to have lower flux and filtration efficiency because the smaller pore size structure provides greater resistance to the feed liquid, leading to a decrease in membrane flux. However, the specific large-pore size and thick separation layer combined with a thinner pre-filtration layer structure in this application allows the filter membrane to maintain a high flux even with a relatively thick separation layer. This may be because, although the separation layer is thick, the overall small SEM measurement average pore size gradient (not greater than 8 nm / μm) and relatively stable pore size gradient (K1 to K2 ratio of 1.05–1.6) indicate that there are no abrupt pore size changes in the separation layer that significantly affect flux. Therefore, the flow rate of the feed liquid in each region of the separation layer is relatively stable, and there are no "bottleneck regions" that significantly affect flux. Combined with the separation layer's inherently large pore size structure, although the stacked pore structure can effectively trap viruses, the effective pathways formed by the stacked pore structure have relatively little resistance to the feed liquid itself. Because the large-pore separation layer has less overall resistance to the feed liquid, and there are no areas of rapid pore size change that significantly affect the flux within the separation layer, the filter membrane still has a high flux even with a thick separation layer.

[0075] Furthermore, due to the high thickness of the separation layer in this application, and the fact that the fiber structure hardly decreases with the reduction of the pore size of the separation layer, the separation layer possesses a robust, self-supporting three-dimensional network structure, resulting in excellent pressure resistance for both the separation layer and the filter membrane. It not only exhibits high tensile strength (5–20 MPa, compared to approximately 3.5–8 MPa for commonly used PVDF filter membranes) but also good elongation at break, demonstrating excellent mechanical properties. This makes it resistant to mechanical damage during membrane fabrication, storage, transportation, and use, thereby reducing the risk of virus leakage.

[0076] It should be noted that the filter membrane is easily damaged by mechanical impact during the assembly of the membrane filter. Since the separation layer of the filter membrane in this application is thick, even if a small amount of damage occurs on the surface of the separation layer, there is still a thick separation layer area that can trap the virus. Therefore, the filter membrane with a thick separation layer in this application can significantly reduce the risk of virus leakage caused by mechanical damage.

[0077] Optionally, the average pore size of the liquid outlet surface measured by SEM is 60–120 nm, the thickness of the separation layer is 25–45 μm, and the logarithmic removal rate of the filter membrane for PP7 phage is not less than 4.

[0078] By adopting the above technical solution, when the thickness of the separation layer is taken as a large value and the average pore size measured by SEM at the liquid surface is taken as a small value, the three-dimensional network structure of the separation layer forms an effective pathway with a longer path and a smaller pore size after stacking. This can form a long-term retention of viruses in the liquid, thereby achieving a higher virus retention effect and further reducing the risk of virus leakage.

[0079] Furthermore, although the separation layer forms an effective pathway with a long path and small pore size, the large pore size of the separation layer's pore structure means that even with the layered structure of the three-dimensional network forming a relatively small effective pathway, the pore size of the three-dimensional network itself remains large, making it difficult to achieve high protein adsorption. In addition, hydrophilic PVDF inherently possesses good hydrophilicity, resulting in relatively weak adsorption force on the pore walls for proteins. Therefore, the adsorption force of the pore walls of the large-pore structure for proteins may not reach the protein adsorption threshold, allowing smaller, less readily adsorbed proteins to pass through effectively. Therefore, the filter membrane of this application, by using a larger value for the thickness of the separation layer and a smaller value for the average pore size measured by SEM at the liquid outlet, can achieve a high virus rejection rate while still maintaining a high protein yield.

[0080] Optionally, the average pore size of the liquid outlet surface measured by SEM is 90–160 nm, the thickness of the separation layer is 15–30 μm, and the logarithmic removal rate of the filter membrane for PP7 phage is greater than 2 and less than 4.

[0081] By adopting the above technical solution, when the thickness of the separation layer is relatively small and the average pore size measured by SEM at the liquid outlet is relatively large, the three-dimensional network structure of the separation layer forms an effective pathway with a shorter path and a larger pore size after stacking. This significantly reduces the resistance of the feed liquid, resulting in a higher throughput. In addition, because the contact between the protein and the pore wall is significantly reduced, the probability of protein adsorption is further reduced, leading to a higher protein yield. Although the virus retention effect of this type of filter membrane is somewhat reduced, a low risk of virus leakage can still be ensured through multi-layer use, and this type of filter membrane is highly suitable for certain specific filtration conditions.

[0082] Secondly, this application provides a process for preparing a filter membrane, which adopts the following technical solution:

[0083] A process for preparing a filter membrane includes the following steps:

[0084] S1. Prepare a casting solution and cast it onto a carrier to form a liquid film; the casting solution comprises the following components by weight: 18-30 parts of PVDF resin, 5-25 parts of hydrophilic additives, 20-50 parts of small molecule additives, and 20-40 parts of good solvents; the mass percentage of PVDF resin in the casting solution is greater than 20%.

[0085] S2. A gas flow with a relative humidity of 60-90% is blown onto the surface of the liquid film to process it, thereby obtaining a green film; wherein the relative velocity between the gas flow and the liquid film is 0.1-6 m / s, and the duration is not less than 5s; and the temperature of the gas flow is 5-20℃ lower than the temperature of the casting solution.

[0086] S3. Immerse the biofilm in an extraction bath for further extraction and solidification to obtain a solid membrane;

[0087] S4. Perform hydrophilic post-treatment on the solid membrane to obtain the finished membrane.

[0088] Optionally, the number-average molecular weight of the PVDF resin is 300,000 to 1,200,000; the hydrophilic additive is at least one of polyethylene glycol, polyvinylpyrrolidone, and polyvinyl alcohol; and the small molecule additive is at least one of LiCl, NH4Cl, nano-silica, acetone, butanone, and tetrahydrofuran.

[0089] Optionally, polyethylene glycol is one of PEG-2000, PEG-4000, and PEG-6000; polyvinylpyrrolidone is one of PVP(K30), PVP(K60), and PVP(K90); and polyvinyl alcohol is one of PVA-117, PVA-205, and PVA-103 from Kuraray, Japan.

[0090] Optionally, the mass percentage of PVDF resin in the casting solution is 20-35%.

[0091] Optionally, in step S3, the processing time is not less than 3 minutes, and the extraction bath includes at least one of water and a small molecule alcohol.

[0092] Optionally, the good solvent is at least one selected from N-methylpyrrolidone, dimethylformamide, dimethylacetamide, trimethyl phosphate, triethyl phosphate, and γ-butyrolactone; the small molecule alcohol is at least one selected from ethanol and isopropanol.

[0093] By adopting the above technical solution, in preparing the PVDF virus-removing membrane of the present invention, a casting solution is first prepared. The casting solution includes PVDF resin, hydrophilic additives, small molecule additives, and a good solvent. PVDF resin is a film-forming polymer with good film-forming processing properties, resulting in a final film with good mechanical properties and anti-fouling properties, suitable for use in the field of virus removal. Furthermore, the solid content of PVDF resin in the casting solution needs to be strictly controlled, and its solid content needs to be greater than 20%. This is because the inventors of this application have found that if the solid content of PVDF resin in the casting solution is too low, it is difficult to obtain a virus-removing membrane with an ideal membrane structure. The good solvent is at least one of N-methylpyrrolidone, dimethylformamide, dimethylacetamide, trimethyl phosphate, triethyl phosphate, and γ-butyrolactone (it can be one of these solvents or a mixed solvent obtained by mixing multiple solvents). The good solvent is used to fully dissolve the PVDF resin, thereby forming a uniform, stable, and clear casting solution, which is then used to form a filter membrane with an ideal pore size through subsequent phase separation and curing processes.

[0094] Preferably, the PVDF resin has a number-average molecular weight of 300,000 to 1,200,000. This molecular weight is beneficial for forming a uniform, stable casting solution with high solids content, and for achieving films with high mechanical properties. Simultaneously, small-molecule additives are added to the casting solution. These additives are at least one of LiCl, NH4Cl, nano-silica, acetone, butanone, and tetrahydrofuran. These small-molecule additives not only help to give the final filter membrane an ideal pore size, achieving higher flux and maintaining high virus retention, but also improve the uniformity of the filter membrane's pore structure, thereby enhancing its tensile strength and other mechanical properties. In addition, a hydrophilic additive is added to the casting solution. This hydrophilic additive is at least one of polyethylene glycol, polyvinylpyrrolidone, and polyvinyl alcohol. The addition of the hydrophilic additive not only improves the hydrophilicity of the filter membrane but also promotes phase separation in the casting solution, increasing the membrane porosity. The hydrophilic additive works synergistically with the extraction solution to promote more rational phase separation and extraction solidification, thus facilitating the formation of an ideal membrane pore structure, reducing the likelihood of defects and macropores, and resulting in relatively uniform pore size. It should be noted that the membrane fabrication processes for different filter membranes vary significantly. The casting solution system in this application is only suitable for PVDF membrane formation and not for polyethersulfone or cellulose-based filter membranes. The inventors speculate that this is related to the inherent properties of the membrane-forming materials themselves.

[0095] A suitable casting solution formulation has a significant impact on the structure and performance of the final filter membrane, affecting factors such as pore size distribution, thickness, and flow rate (flux). A proper casting solution formulation ensures that the final filter membrane has an appropriate thickness and achieves an ideal pore size. Furthermore, the casting solution with a high solids content in this application has a reasonable viscosity, is easy to handle, and can be manually cast (e.g., by hand pouring, casting, or spreading on a casting surface) or automatically cast (e.g., by pouring or additional casting on a moving bed). Various devices known in the art can be used for casting. Casting equipment includes, for example, mechanical coaters, which include coating knives, doctor blades, or spray / pressurization systems. Various casting speeds are suitable, as known in the art, such as casting speeds of approximately 2 to 6 feet per minute (fpm), depending on the specific circumstances.

[0096] After the casting solution is cast into a liquid film, the liquid film undergoes a phase separation treatment. A high-humidity, relatively low-temperature airflow is blown onto the liquid film surface to promote phase separation and form a growing film. It is well known that the faster the phase separation rate, the smaller the membrane pores. In this invention, through the synergistic effect of airflow humidity (60-90%), airflow temperature (5-20°C lower than the casting solution temperature), and a suitable casting solution formulation, the liquid film undergoes relatively rapid phase separation on the air side. This results in smaller pores on the air side (but due to the limited water vapor content in the airflow, the pores will not be too small, and the film will not have pores with a diameter of approximately 20 nm). This also facilitates the formation of a relatively thick separation layer. Furthermore, the relative velocity between the airflow and the liquid film is 0.1-6 m / s, and the duration is not less than 5 seconds. This phase separation treatment aims to obtain the ideal pore size and thickness of the filter membrane and also facilitates the formation of an ideal membrane structure, i.e., the formation of ideal fibers.

[0097] The biofilm is then immersed in an extraction bath for further extraction and solidification. The extraction bath is at least one of water and a small molecule alcohol (C1-C4 small molecule alcohol), preferably at least one of ethanol and isopropanol. Under the action of the extraction bath, PVDF will be precipitated more completely, and the processing time is not less than 3 minutes, to ensure that a solid membrane with an ideal membrane structure is obtained.

[0098] Next, to improve the hydrophilicity of the solid membrane, a grafting method was used to treat it. The hydrophilic treatment solution contained 8% hydroxypropyl acrylate, 23% 3-butanol, and the balance water by volume percentage. The prepared hydrophilic treatment solution was bubbled and stirred to remove oxygen for 20 minutes, maintaining the temperature of the hydrophilic treatment solution at 45°C during treatment. Subsequently, the solid membrane was placed in a nitrogen atmosphere and cooled to below -60°C, then irradiated with at least 25 kGy of gamma rays using cobalt-60 as the radiation source. After irradiation, the solid membrane was placed under a low pressure of approximately 13.4 Pa and allowed to stand for 15 minutes. The previously stirred and deoxygenated hydrophilic treatment solution was then brought into contact with the solid membrane for reaction, and the reaction was allowed to stand for 1 hour. After the reaction, the membrane was washed with 2-propanol and vacuum dried at 60°C to obtain the finished membrane.

[0099] The final PVDF filter membrane has an ideal membrane structure, high flux, high tensile strength, and can efficiently retain viruses, as well as a high protein yield.

[0100] Thirdly, this application provides a membrane filter, which adopts the following technical solution:

[0101] A membrane filter includes the aforementioned filter membrane, and the membrane filter satisfies the following conditions:

[0102] A. The PP7 phage log removal rate of the filter membrane is greater than or equal to 4, and the membrane filter includes 1 to 2 of the filter membranes.

[0103] B. The PP7 phage log removal rate of the filter membrane is less than 4, and there are 2 to 3 filter membranes in the membrane filter.

[0104] Preferably, the membrane filter can be a capsule filter or a needle filter.

[0105] By adopting the above technical solution, for filter membranes with a logarithmic removal rate of 4 or higher, one filter membrane can be directly made into a membrane filter or two filter membranes can be stacked together to make a membrane filter, thereby further reducing the risk of virus leakage.

[0106] For filter membranes with a log removal rate of less than 4, in order to achieve a log removal rate of greater than 4, 2 to 3 filter membranes can be stacked to reduce the risk of virus leakage.

[0107] In summary, this application includes at least one of the following beneficial technical effects:

[0108] 1. The filter membrane in this application controls the filter membrane to have a separation layer with a large pore size and a large thickness. The separation layer forms a stacked pore structure with a longer path in the membrane thickness direction. The size of the effective channel for virus passage formed by the multi-layered pore structure is much smaller than the actual size of the pore structure, thereby achieving high efficiency in the interception of small-sized viruses. Even with the smaller pore structure after stacking, there is less interception and adsorption of even smaller protein layers, thus achieving both high protein yield and high virus rejection rate. This is different from the general belief that a small pore size of the separation layer is required to achieve a high virus rejection rate, and that large-pore hydrophilic PVDF filter membranes are difficult to intercept small-sized viruses. This is a different technical approach and has yielded very unexpected results.

[0109] 2. In this application, the fiber structure of the filter membrane does not undergo significant size changes from the pre-filtration layer to the separation layer. This is different from the common perception that the larger the pore size of the filter membrane, the thicker the fiber. The specific fiber structure not only ensures low adsorption of proteins by the fiber structure, but also has better mechanical properties because the pressure resistance of the filter membrane is relatively uniform throughout and there are no weak points.

[0110] 3. The filter membrane of this application has a low overall symmetry and a low symmetry of the separation layer. This not only ensures the uniformity of the overall mechanical properties of the filter membrane, but also indicates that there is no abrupt change in pore size in the thickness direction of the filter membrane. In contrast, abrupt change in pore size, due to its dense and rapidly changing layered pore structure, is likely to result in high protein retention and adsorption, leading to a decrease in protein yield.

[0111] 4. In the preparation process of the filter membrane of this application, by controlling the formulation of the casting solution, the mass percentage of PVDF resin is controlled to be greater than 20%, and hydrophilic additives and small molecule additives are added to the casting solution to promote the phase separation process and improve the hydrophilicity of PVDF resin. Combined with subsequent wet air purging, extraction curing and hydrophilic post-treatment, a filter membrane with an ideal pore structure can be obtained, so as to simultaneously obtain high virus rejection rate, high protein yield and high mechanical properties. Attached Figure Description

[0112] Figure 1 This is a cross-sectional SEM image of the filter membrane prepared in Example 1 of this application, with a magnification of 10K×.

[0113] Figure 2 The cross-sectional SEM image of the filter membrane near the liquid outlet surface obtained in Example 1 of this application has a magnification of 10K×.

[0114] Figure 3 This is a SEM image of the liquid inlet side of the filter membrane prepared in Example 3 of this application, with a magnification of 10K×.

[0115] Figure 4 This is a SEM image of the liquid outlet side of the filter membrane prepared in Example 3 of this application, with a magnification of 10K×.

[0116] Figure 5 This is a cross-sectional SEM image of the filter membrane prepared in Example 3 of this application, with a magnification of 2K×.

[0117] Figure 6 The figure shows the distribution curve of 20nm colloidal gold in the filter membrane of this application, which was determined by a 20nm colloidal gold retention experiment. The lower side of the filter membrane is the liquid inlet surface and the upper side is the liquid outlet surface. This figure is only a schematic diagram and is not a filter membrane of a specific embodiment.

[0118] Figure 7 This is a SEM image of the cross-section of the filter membrane after a 20nm colloidal gold retention experiment. The magnification is 5K×. The upper side of the filter membrane in the image is the liquid inlet surface, and the lower side is the liquid outlet surface. This image is only a schematic diagram and is not a filter membrane of a specific embodiment. Detailed Implementation

[0119] The following is in conjunction with the appendix Figures 1-7 This application will be described in further detail.

[0120] To more clearly illustrate the overall concept of this application, detailed descriptions are provided below using examples. Unless otherwise specified, the raw materials and equipment used in the following examples and comparative examples are all available through conventional commercial channels. Specifically, a Hitachi S-5500 scanning electron microscope was used to characterize the structural morphology of the filter membrane.

[0121] Example 1

[0122] A process for preparing an asymmetric hydrophilic PVDF filter membrane for virus removal includes the following steps:

[0123] S1. Prepare the casting solution and cast the prepared casting solution onto the carrier to form a liquid film; the casting solution is composed of the following components by weight: 23 parts PVDF resin, 15 parts hydrophilic additive, 35 parts small molecule additive and 30 parts good solvent; the mass percentage of PVDF resin in the casting solution is 22.3%.

[0124] In the casting solution, the number average molecular weight of PVDF resin is 800,000, the hydrophilic additive is a mixture of PEG-4000 and PVP (K30) in equal mass ratio, the small molecule additive is a mixture of LiCl and acetone in equal mass ratio, and the good solvent is a mixture of N-methylpyrrolidone and dimethylformamide in equal volume.

[0125] S2. A gas flow with a relative humidity of 68% is blown onto the surface of the liquid film to produce a raw film. The relative velocity between the gas flow and the liquid film is 1.5 m / s, the duration is 24 s, and the temperature of the gas flow is 10 ℃ lower than the temperature of the casting liquid (i.e., the liquid film).

[0126] S3. The prepared biofilm is immersed in an extraction bath for further extraction and solidification to obtain a solid membrane. The extraction bath is a mixture of deionized water and a small molecule alcohol in equal volume ratio, wherein the small molecule alcohol is ethanol, and the treatment time is 6 min.

[0127] S4. The obtained solid membrane undergoes hydrophilic post-treatment, specifically by grafting. The hydrophilic treatment solution comprises 8% hydroxypropyl acrylate, 23% 3-butanol, and the balance water by volume percentage. The prepared hydrophilic treatment solution is bubbled and stirred to remove oxygen for 20 minutes, maintaining the temperature of the hydrophilic treatment solution at 45°C during treatment. Subsequently, the solid membrane is placed in a nitrogen atmosphere and cooled to approximately -65°C, then irradiated with approximately 27 kGy of gamma rays using cobalt-60 as the radiation source. After irradiation, the solid membrane is placed under a low pressure below approximately 13.4 Pa and allowed to stand for 15 minutes. The membrane is then contacted with the aforementioned hydrophilic treatment solution after stirring and deoxygenation, and allowed to react for 1 hour. After the reaction, the membrane is washed with 2-propanol and vacuum dried at 60°C to obtain the finished membrane.

[0128] Examples 2-7

[0129] The main difference between Examples 2-7 and Example 1 lies in the formulation of the casting solution and the process parameters for each step, as detailed in the table below:

[0130]

[0131]

[0132] In the casting solution, the number average molecular weight of PVDF resin is 80W, the hydrophilic additive is a mixture of PEG-4000 and PVP (K30) in equal mass ratio, the small molecule additive is a mixture of LiCl and acetone in equal mass ratio, and the good solvent is a mixture of N-methylpyrrolidone and dimethylformamide in equal volume.

[0133] in,

[0134] In the casting solution of Example 2, the hydrophilic additive is a mixture of PEG-2000 and PVP (K60) in equal mass ratio, the small molecule additive is a mixture of nano-silica and acetone in equal mass ratio, and the good solvent is a mixture of N-methylpyrrolidone and dimethylacetamide in equal volume.

[0135] In the casting solution of Example 3, the hydrophilic additive is PEG-6000, the small molecule additive is NH4Cl, the good solvent is N-methylpyrrolidone, and the small molecule alcohol is ethanol.

[0136] In the casting solution of Example 4, the hydrophilic additive is a mixture of PEG-2000 and PVP (K30) in equal mass ratio, the small molecule additive is a mixture of LiCl and butanone in equal mass ratio, the good solvent is a mixture of N-methylpyrrolidone and trimethyl phosphate in equal volume, and the small molecule alcohol is ethanol.

[0137] In the casting solution of Example 5, the hydrophilic additive is a mixture of PEG-2000 and PVA-117 in equal mass ratio, the small molecule additive is a mixture of LiCl and acetone in equal mass ratio, the good solvent is a mixture of N-methylpyrrolidone and dimethylformamide in equal volume, and the small molecule alcohol is isopropanol.

[0138] In the casting solution of Example 6, the hydrophilic additive is a mixture of PEG-6000 and PVA-103 in equal mass ratio, the small molecule additive is a mixture of LiCl and acetone in equal mass ratio, the good solvent is a mixture of N-methylpyrrolidone and dimethylformamide in equal volume, and the small molecule alcohol is ethanol.

[0139] In the casting solution of Example 7, the hydrophilic additive is a mixture of PVP (K30) and PVA-103 in equal mass ratio, the small molecule additive is a mixture of LiCl and acetone in equal mass ratio, the good solvent is a mixture of N-methylpyrrolidone and dimethylformamide in equal volume, and the small molecule alcohol is isopropanol.

[0140] Comparative Example

[0141] Comparative Example 1

[0142] A process for preparing an asymmetric hydrophilic PVDF filter membrane for virus removal includes the following steps:

[0143] S1. Prepare the casting solution and cast the prepared casting solution onto the carrier to form a liquid film; the casting solution is composed of the following components by weight: 40 parts PVDF resin, 10 parts hydrophilic additive, 10 parts small molecule additive and 30 parts good solvent; the mass percentage of PVDF resin in the casting solution is 44.4%.

[0144] In the casting solution, the number average molecular weight of PVDF resin is 80W, the hydrophilic additive is a mixture of PEG-4000 and PVP (K30) in equal mass ratio, the small molecule additive is a mixture of LiCl and acetone in equal mass ratio, and the good solvent is a mixture of N-methylpyrrolidone and dimethylformamide in equal volume.

[0145] S2. The liquid film is immersed in a pretreatment solution for pretreatment. The pretreatment solution is a mixture of deionized water and N-methylpyrrolidone in a volume ratio of 3:2. The pretreatment time is 10s. The temperature of the pretreatment solution is the same as the temperature of the liquid film. After pretreatment, a raw film is obtained.

[0146] S3. The prepared biofilm is immersed in an extraction bath for further extraction and solidification to obtain a solid membrane. The extraction bath is a mixture of deionized water and a small molecule alcohol in equal volume ratio, and the small molecule alcohol is ethanol.

[0147] S4. The obtained solid membrane undergoes hydrophilic post-treatment, specifically by grafting. The hydrophilic treatment solution comprises 8% hydroxypropyl acrylate, 23% 3-butanol, and the balance water by volume percentage. The prepared hydrophilic treatment solution is bubbled and stirred to remove oxygen for 20 minutes, maintaining the temperature of the hydrophilic treatment solution at 45°C during treatment. Subsequently, the solid membrane is placed in a nitrogen atmosphere and cooled to approximately -65°C, then irradiated with approximately 27 kGy of gamma rays using cobalt-60 as the radiation source. After irradiation, the solid membrane is placed under a low pressure below approximately 13.4 Pa and allowed to stand for 15 minutes. The membrane is then contacted with the aforementioned hydrophilic treatment solution after stirring and deoxygenation, and allowed to react for 1 hour. After the reaction, the membrane is washed with 2-propanol and vacuum dried at 60°C to obtain the finished membrane.

[0148] Comparative Example 2

[0149] A process for preparing an asymmetric hydrophilic PVDF filter membrane for virus removal includes the following steps:

[0150] S1. Prepare the casting solution and cast the prepared casting solution onto the carrier to form a liquid film; the casting solution is composed of the following components by weight: 15 parts PVDF resin, 15 parts hydrophilic additive, 35 parts small molecule additive and 30 parts good solvent; the mass percentage of PVDF resin in the casting solution is 15.8%.

[0151] In the casting solution, the number average molecular weight of PVDF resin is 80W, the hydrophilic additive is a mixture of PEG-4000 and PVP (K30) in equal mass ratio, the small molecule additive is a mixture of LiCl and acetone in equal mass ratio, and the good solvent is a mixture of N-methylpyrrolidone and dimethylformamide in equal volume.

[0152] S2. A gas flow with a relative humidity of 40% is blown onto the surface of the liquid film to produce a green film. The relative velocity between the gas flow and the liquid film is 7 m / s, the duration is 60 s, and the temperature of the gas flow is 25°C lower than the temperature of the casting solution (i.e., the liquid film).

[0153] S3. The prepared biofilm is immersed in an extraction bath for further extraction and solidification to obtain a solid membrane. The extraction bath is a mixture of deionized water and a small molecule alcohol in equal volume ratio, wherein the small molecule alcohol is ethanol, and the treatment time is 20 min.

[0154] S4. The obtained solid membrane undergoes hydrophilic post-treatment, specifically by grafting. The hydrophilic treatment solution comprises 8% hydroxypropyl acrylate, 23% 3-butanol, and the balance water by volume percentage. The prepared hydrophilic treatment solution is bubbled and stirred to remove oxygen for 20 minutes, maintaining the temperature of the hydrophilic treatment solution at 45°C during treatment. Subsequently, the solid membrane is placed in a nitrogen atmosphere and cooled to approximately -65°C, then irradiated with approximately 27 kGy of gamma rays using cobalt-60 as the radiation source. After irradiation, the solid membrane is placed under a low pressure below approximately 13.4 Pa and allowed to stand for 15 minutes. The membrane is then contacted with the aforementioned hydrophilic treatment solution after stirring and deoxygenation, and allowed to react for 1 hour. After the reaction, the membrane is washed with 2-propanol and vacuum dried at 60°C to obtain the finished membrane.

[0155] Performance testing and performance parameters

[0156] 1. Structural Characterization

[0157] The required data can be obtained by characterizing the membrane structure of the filter membranes obtained in Examples 1-7 and Comparative Examples 1-2 using scanning electron microscopy; the morphological parameters of the inlet surface, outlet surface, and pre-filtration layer structure of Examples 1-7 and Comparative Examples 1-2 are recorded in the following table:

[0158]

[0159] The morphological parameters of the separated layer structures in Examples 1-7 and Comparative Examples 1-2 are recorded in the following table:

[0160] It should be noted that, in the table above, since the filter membrane prepared in Comparative Example 2 has a large overall pore size, it cannot effectively retain 20nm colloidal gold. Therefore, there is no clear separation layer and pre-filtration layer structure, and no relevant morphological parameters are provided. Only the surface morphology of the inlet and outlet surfaces is characterized.

[0161] II. Virus interception capability

[0162] The filter membranes prepared in each example or comparative example were used as samples for virus challenge testing. The LRV detection method for the filter membranes followed the guidance document TR41 issued by the PDA. During the test, PP7 bacteriophage was used as the intercepted virus, the feed stream was immunoglobulin IVIG, and the buffer system was PBS. The pressure of the feed solution during the virus challenge test was 30 psi. LRV was calculated by detecting the titer of PP7 bacteriophage in the challenge solution and filtrate; protein yield was calculated by detecting the protein concentration in the challenge solution and filtrate; and flux was calculated by recording the flow rate and time.

[0163] It is important to note that when conducting virus challenge tests, samples should first be prepared using a single-layer filter membrane. If the LRV measured by the single-layer filter membrane is less than 4, samples should be prepared using a double-layer filter membrane to further determine the LRV and protein yield under the double-layer filter membrane.

[0164] III. Mechanical Properties

[0165] The filter membranes prepared in each embodiment and comparative example were used as samples, and their tensile properties were tested by a universal tensile testing machine to measure the tensile strength of the filter membranes.

[0166] The mechanical properties and virus retention capacity of the filter membranes in Examples 1-7 and Comparative Examples 1-2 are recorded in the following table:

[0167]

[0168]

[0169] It should be noted that since Examples 5-7 and Comparative Example 1 have an LRV > 4 for PP7 phage (virus rejection rate > 99.99%), which has already reached the internationally recognized virus rejection rate, no virus challenge test was conducted on the double-layer filter membrane. Therefore, Examples 5-7 and Comparative Example 1 do not have double-layer filter membrane LRV data or double-layer filter membrane protein yield data.

[0170] In Comparative Example 2, the filter membrane had an LRV < 1 for PP7 phage. Although its flux and protein yield were high, the risk of virus leakage was too great, rendering it useless. Therefore, no further virus challenge test was conducted on the double-layer filter membrane. Consequently, Comparative Example 2 also lacked LRV and protein yield data for the double-layer filter membrane.

[0171] in conclusion

[0172] By comparing the data from Examples 1-7 and Comparative Example 1, it is easy to see that the filter membrane with a thick, large-pore separation layer in this application not only achieves a high virus rejection rate (for filter membranes with an LRV of 2-4, 2-3 filter membranes can be used in series to obtain a high LRV), but also a high protein yield (for Examples 1-4, even when 2 filter membranes are used in series, the protein yield still reaches 95% or even 98% or more). In contrast, the filter membranes in Comparative Example 1 and CN105980037B, while achieving a high virus rejection rate (LRV > 6 or even LRV > 7) with their small-pore separation layer structure, have a lower protein yield of only about 80-90%.

[0173] Furthermore, by further analyzing the properties of the filter membrane prepared at a relative ratio of 1, it is evident that the membrane exhibits significantly poor mechanical properties. This is likely due to the fact that the diameter ratio of the fiber structure in the pre-filtration layer to the separation layer reaches approximately 3.9, indicating a noticeable asymmetric distribution of the fiber structure along the membrane thickness direction. This implies that the mechanical properties of the membrane are not uniform along the thickness direction, making it prone to weak points and leading to a decline in mechanical performance. This is quite similar to the filter membrane in CN105980037B, which can only be used at a relatively low pressure of 15 psi.

[0174] Furthermore, the filter flux of Comparative Example 1 was also significantly lower. The main region affecting the filter flux was the separation layer. The ratio of the average pore size of the separation layer measured by SEM to the average diameter of the separation fibers measured by SEM was small, only about 1.8. This resulted in a large resistance of the separation layer to the feed solution. In addition, the pore size of the separation layer in Comparative Example 1 was even smaller, leading to a significant decrease in flux and a marked decrease in protein yield. This illustrates that, for common hydrophilic PVDF filter membranes with high asymmetry, high virus rejection rates are often accompanied by a decrease in protein yield.

[0175] 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 hydrophilic PVDF filter membrane for virus removal, comprising a porous body having non-directional tortuous pathways within the porous body, wherein the side of the porous body closer to the feed liquid is the inlet surface, and the side of the porous body farther from the feed liquid is the outlet surface, and the porous body transitions with continuous fibers in the membrane thickness direction, characterized in that: The average pore size of the porous body, as measured by SEM, gradually decreases from the inlet surface to the outlet surface. The porous body includes a pre-filtration layer and a separation layer for retaining colloidal gold with a diameter of 20 nm. The separation layer includes the outlet surface, and the average pore size of the outlet surface, as measured by SEM, is 60–160 nm. The thickness of the separation layer is 15–45 μm.

2. The asymmetric hydrophilic PVDF filter membrane for virus removal according to claim 1, characterized in that: The separation layer includes separation fibers, which are interconnected to form a three-dimensional network structure of the separation layer. The average diameter of the separation fibers measured by SEM is 30-50 nm. The pre-filter layer includes support fibers, which are interconnected to form a three-dimensional network structure of the pre-filter layer. The average diameter of the support fibers measured by SEM is 30-55 nm.

3. The asymmetric hydrophilic PVDF filter membrane for virus removal according to claim 2, characterized in that: The ratio of the average diameter of the supporting fibers measured by SEM to the average diameter of the separated fibers measured by SEM was 0.8 to 1.

5.

4. The asymmetric hydrophilic PVDF filter membrane for virus removal according to claim 2, characterized in that: The average length of the separated fiber measured by SEM is 80–150 nm, and the average length of the supporting fiber measured by SEM is 120–180 nm.

5. The asymmetric hydrophilic PVDF filter membrane for virus removal according to claim 2, characterized in that: The ratio of the average pore size measured by SEM of the separation layer to the average diameter measured by SEM of the separation fiber is 2 to 7; the ratio of the average pore size measured by SEM of the pre-filter layer to the average diameter measured by SEM of the support fiber is 3.5 to 8.

6. The asymmetric hydrophilic PVDF filter membrane for virus removal according to claim 2, characterized in that: The ratio of the average pore size measured by SEM of the separation layer to the average length measured by SEM of the separation fiber is 1.1 to 1.4; the ratio of the average pore size measured by SEM of the pre-filter layer to the average length measured by SEM of the support fiber is 1.2 to 1.

7.

7. The asymmetric hydrophilic PVDF filter membrane for virus removal according to claim 1, characterized in that: The average pore size measured by SEM at the liquid inlet surface is 200–500 nm, the ratio of the average pore size measured by SEM at the liquid inlet surface to the average pore size measured by SEM at the liquid outlet surface is 1.2–6, and the specific surface area of ​​the porous matrix is ​​6–12 m². 2 / g.

8. The asymmetric hydrophilic PVDF filter membrane for virus removal according to claim 1, characterized in that: The average pore size variation gradient of the separation layer measured by SEM is no greater than 8 nm / μm, and the average pore size of the separation layer measured by SEM is 100–200 nm.

9. The asymmetric hydrophilic PVDF filter membrane for virus removal according to claim 1, characterized in that: The side of the separation layer closest to the liquid inlet is considered 0%, and the liquid outlet is considered 100%. The separation layer is divided into ten equal parts along the film thickness direction. The ratio of the average pore size measured by SEM in the region where the film thickness of the separation layer is 0-10% to the average pore size measured by SEM in the region where the film thickness of the separation layer is 40-50% is K1, and K1 is not greater than 2.

2.

10. The asymmetric hydrophilic PVDF filter membrane for virus removal according to claim 9, characterized in that: The ratio of the average pore size measured by SEM in the region where the separation layer film thickness is 40-50% to the average pore size measured by SEM at the liquid outlet surface is K2, where K2 is not greater than 1.

5.

11. The asymmetric hydrophilic PVDF filter membrane for virus removal according to claim 10, characterized in that: K1 is greater than K2, and the ratio of K1 to K2 is 1.05 to 1.

6.

12. The asymmetric hydrophilic PVDF filter membrane for virus removal according to claim 1, characterized in that: The pre-filter layer includes the liquid inlet surface, the thickness of the pre-filter layer is 1-15 μm, the average pore size of the pre-filter layer measured by SEM is 150-300 nm, and the average pore size variation gradient of the pre-filter layer measured by SEM is 5-20 nm / μm.

13. The asymmetric hydrophilic PVDF filter membrane for virus removal according to claim 1, characterized in that: The flux of the filter membrane is not less than 50 L·h -1 ·m -2 @30psi, the filter membrane has a log removal rate of greater than 2 for PP7 phage; the tensile strength of the filter membrane is 5-20 MPa.

14. The asymmetric hydrophilic PVDF filter membrane for virus removal according to any one of claims 1 to 13, characterized in that: The average pore size of the liquid outlet surface measured by SEM is 60–120 nm, the thickness of the separation layer is 25–45 μm, and the logarithmic removal rate of the filter membrane for PP7 phage is not less than 4.

15. The asymmetric hydrophilic PVDF filter membrane for virus removal according to any one of claims 1 to 13, characterized in that: The average pore size of the liquid outlet surface measured by SEM is 90–160 nm, the thickness of the separation layer is 15–30 μm, and the logarithmic removal rate of the filter membrane for PP7 phage is greater than 2 and less than 4.

16. A process for preparing a filter membrane according to any one of claims 1 to 15, characterized in that: The process includes the following steps: S1. Prepare a casting solution and cast it onto a carrier to form a liquid film; the casting solution comprises the following components by weight: 18-30 parts of PVDF resin, 5-25 parts of hydrophilic additives, 20-50 parts of small molecule additives, and 20-40 parts of good solvents; the mass percentage of PVDF resin in the casting solution is greater than 20%. S2. A gas flow with a relative humidity of 60-90% is blown onto the surface of the liquid film to process it, thereby obtaining a green film; wherein the relative velocity between the gas flow and the liquid film is 0.1-6 m / s, and the duration is not less than 5s; and the temperature of the gas flow is 5-20℃ lower than the temperature of the casting solution. S3. Immerse the biofilm in an extraction bath for further extraction and solidification to obtain a solid membrane; S4. Perform hydrophilic post-treatment on the solid membrane to obtain the finished membrane.

17. The filter membrane preparation process according to claim 16, characterized in that: The number-average molecular weight of the PVDF resin is 300,000 to 1,200,000. The hydrophilic additive is at least one of polyethylene glycol, polyvinylpyrrolidone and polyvinyl alcohol; The small molecule additive is at least one of LiCl, NH4Cl, nano-silica, acetone, butanone, and tetrahydrofuran.

18. The filter membrane preparation process according to claim 16, characterized in that: In step S3, the processing time is not less than 3 minutes, and the extraction bath includes at least one of water and small molecule alcohol.

19. The filter membrane preparation process according to claim 18, characterized in that: The good solvent is at least one of N-methylpyrrolidone, dimethylformamide, dimethylacetamide, trimethyl phosphate, triethyl phosphate, and γ-butyrolactone; The small molecule alcohol is at least one of ethanol and isopropanol.

20. A membrane filter, characterized in that: The membrane filter comprises the filter membrane according to any one of claims 1 to 15, and the membrane filter satisfies the following conditions: A. The PP7 phage log removal rate of the filter membrane is greater than or equal to 4, and the membrane filter includes 1 to 2 of the filter membranes. B. The PP7 phage log removal rate of the filter membrane is less than 4, and there are 2 to 3 filter membranes in the membrane filter.

Images