JP2005137996A - Permselective separation membrane - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a permselective separation membrane constituted so as to reduce the elution of endotoxin contained in the membrane as impurities, to reduce the irritation properties to a living body and especially suitable as a hollow fiber type blood purifier having high water permeability used in the treatment of chronic renal failure. <P>SOLUTION: This permselective separation membrane is characterized in that the amount of endotoxin eluted from the membrane is 150 EU/m<SP>2</SP>or below. The sieve coefficient of alubmin of the permselective separation membrane is 0.01-0.30 and, when the content of a hydrophilic polymer in the vicinity of the inner surface of the membrane is set to A and the content of the hydrophilic polymer of the whole of the membrane is set to B, the relation of A and B is A≥5 mass% and B/A≤0.4 and the clearance of β2MG is 40 mL/min or above. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a selectively permeable separation membrane for the purpose of solute removal and purification of a liquid, and more particularly to a selectively permeable separation membrane used for blood purification, removal of endotoxin in water, and the like.

  In recent years, many membrane separation techniques have been put into practical use, and various selectively permeable separation membranes are used for separating a target substance from a liquid or gas mixture or removing impurities. As a material for the selectively permeable separation membrane, organic polymers are generally used. For example, cellulose, regenerated cellulose, cellulose diacetate, cellulose triacetate and the like are used as natural polymers, and synthetic polymers are used. As such, polysiloxane, polyamide, polysulfone, polyethersulfone, polyolefin, polyacrylonitrile, polymethacrylate and the like are used. Among them, polysulfone polymers such as polysulfone and polyethersulfone are widely used as industrial separation membranes because of their excellent resistance to heating, radiation, and chemicals such as acids and alkalis. It is for showing. Moreover, since it is excellent also in biocompatibility, it attracts attention recently as a raw material of a medical separation membrane, and demand is increasing.

  However, since the polysulfone polymer is a material having high water repellency and difficult to wet with water, it is often formed by adding a hydrophilic polymer for the purpose of obtaining desired film performance. For this reason, when using a membrane, there has been a problem that the hydrophilic polymer is eluted in the liquid to be treated or the fine particles and oligomer components are eluted from the polysulfone-based polymer.

Many attempts to improve these drawbacks have been disclosed (see, for example, Patent Documents 1 to 4). In Patent Literature 1, the hydrophilic polymer is cross-linked and insolubilized by irradiating the membrane with water in a water-containing state, and the membrane is firmly fixed to the membrane to improve elution. In Patent Document 2, a hydrophilic polymer is similarly insolubilized and fixed to the membrane by heat-treating the membrane in a dry state. In addition, a method of treating a membrane with a component that forms a water-insoluble complex with a hydrophilic polymer has also been attempted. In Patent Document 3, an insolubilization treatment using a polyvalent acid such as polycarboxylic acid or polyphenol is performed. Yes.
Japanese Patent Laid-Open No. 4-300636 JP-A-9-103664 Japanese Patent Publication No. 8-32297

On the other hand, for elution of fine particles and oligomer components from the polysulfone polymer, Patent Document 4 discloses a method of purifying a raw material resin by reprecipitation or washing a membrane with an alcohol solvent.
JP-A-5-329345

However, all of them are concerned only with elution of hydrophilic polymers, fine particles and oligomer components, and elution of other impurities was not considered. Polysulfone-based membranes are generally formed by a dry-wet spinning method. However, depending on the state of contamination of water used for coagulation or the coagulation bath, some impurities may be mixed in during the film formation process. The same was true when the modularized membrane was filled with water. Among them, physiologically active substances that stimulate the living body are also included. In particular, the glycolipid components are diverse and powerful physiologically active even in trace amounts, as exemplified by endotoxins and endotoxins. Some have substances. Since these have a high affinity for the surface of the polysulfone polymer, they are easily adsorbed on the membrane, and are difficult to remove by washing. On the other hand, it has a high affinity for a solvent having a high dissolution power such as blood, and it may be eluted to express physiological activity. The following method is disclosed with respect to such a problem (for example, refer patent document 5). Patent Document 5 discloses a method in which the bioactive glycolipid content in the membrane is 1.3 ppb or less and the albumin permeability is 0.5 to 0.001%. However, when albumin permeability is 0.5% or less, β1 microglobulin (hereinafter abbreviated as β2MG), which is a causative agent of dialysis amyloidosis, which is a complication of long-term dialysis patients, and α1 are considered to have an effect of improving pruritus. There is a problem that the removal of low molecular weight protein having a molecular weight (33,000) around microglobulin is not sufficiently achieved. In particular, in patients with septicemia, removal of an appropriate amount of albumin is preferable for prognosis, so it cannot be said that the performance is sufficient.
Japanese Patent Laid-Open No. 11-309356

Although it is preferable that the bioactive glycolipid content in the membrane is 1.3 ppb or less, according to Non-Patent Document 1, endotoxin is said to be 5 to 7 EU / ng, and 1.3 ppb is equivalent to about 7000 EU / L. It becomes. In addition, when nonpatent literatures 1 and 2 assume that the standard weight of a patient is 60 kg, it is suggested that when 300 EU flows into the body, there is a possibility of fever, and it cannot be said that sufficient safety can be ensured. Moreover, it could not be said that the membrane potentially contained impurities to achieve an endotoxin concentration of 100 EU / L or less in the dialysate recommended by the Japanese Dialysis Society.
Thirteenth Amendment Japanese Pharmacopoeia General Test B-62-63 Endotoxin Test Wako Pure Chemicals Vol.70 No.64 P17 Dialysis therapy and endotoxin

  It is an object of the present invention to provide a selectively permeable separation membrane that solves the above-mentioned problems of the prior art, reduces elution of endotoxin contained in the membrane as an impurity, and reduces irritation to a living body.

The present invention provides a selectively permeable separation membrane characterized in that the endotoxin eluted from the membrane is 100 EU / m ² or less. Further, the permselective separation membrane is characterized in that the membrane has a sieve coefficient of albumin of 0.01 or more and 0.30 or less.

  The present invention is a hollow fiber membrane that reduces elution of endotoxin, which is considered to be irritating to a living body, enables proper protein removal, reduces complications and pruritus in long-term dialysis patients, and has a good prognosis. It was possible to form a film.

で示される繰り返し単位をもつポリスルホン高分子やポリエーテルスルホン高分子がポリスルホン系高分子として広く市販されており、入手も容易なため特に好ましい。 The hydrophobic polymer used as the material for the selectively permeable separation membrane of the present invention is not limited in any way. For example, cellulose, regenerated cellulose, cellulose diacetate, cellulose derivatives such as cellulose triacetate, etc. are used as natural polymers. As the synthetic polymer, polysiloxane, polyamide, polysulfone, polyethersulfone, polyolefin, polyacrylonitrile, polymethacrylate and the like are used. Of these, polysulfone polymers such as polysulfone and polyethersulfone are preferred because they exhibit excellent resistance to heating, radiation, and chemicals such as acids and alkalis, and are excellent in biocompatibility. Polysulfone polymer is a general term for resins having a sulfone bond and is not particularly limited.

A polysulfone polymer or a polyethersulfone polymer having a repeating unit represented by the formula (1) is particularly preferred because it is widely available as a polysulfone polymer and is easily available.

  The hydrophilic polymer used in the present invention is not particularly limited, but a polymer that forms a micro phase separation structure in a solution with a polysulfone polymer is preferably used. Polyethylene glycol, polyvinyl alcohol, carboxymethyl cellulose, polyvinyl pyrrolidone and the like can be mentioned, but it is a preferred embodiment that polyvinyl pyrrolidone is used from the viewpoint of safety and economy. The polyvinylpyrrolidone here may contain a copolymer with a vinyl monomer such as vinyl acetate, and can be used for the purpose of controlling the affinity with the polysulfone polymer and the hydrophilicity of the membrane surface. Although these molecular weights are not particularly limited, it is preferable to use those having a weight average molecular weight of at least 100,000 in order to remain and impart appropriate hydrophilicity to the film. More preferably, it is 200,000 or more, More preferably, it is 300,000 or more, More preferably, it is 450,000 or more.

In recent years, it has been suggested that the active use of dialysis fluid purification (endotoxin removal) reduces the risk of dialysis complications. Attention to the removal of endotoxin from dialysis facilities has increased, such as the use of osmotic treatment. Under such circumstances, the Japanese Society for Dialysis Therapy recommends 50 EU / L or less as the target of endotoxin level in the dialysate. However, even if the dialysate, dialysate line, and coupler are cleaned, if the dialyzer that is in direct contact with blood is contaminated with bacteria, the influx of endotoxin into the patient's blood cannot be suppressed. The hemodialyzer is treated with γ-rays, ethylene oxide gas, autoclave sterilization, etc., and the live bacteria are killed. However, endotoxin, which is a dead body of the bacteria and its fragments, may cause symptoms such as fever. . As described above, 5EU / kg is considered to be the minimum fever dose for endotoxin (see Reference 1). Here, if the patient's standard weight is 60 kg, if 300 EU flows into the body, it will cause a strong fever action. Currently, the maximum membrane area of dialyzers used at dialysis sites is around 2m ² (based on the surface area of the hollow fiber membrane), so elution of endotoxin from the membrane should be 150EU / m ² or less. preferable. Including the management of variability, it is more preferable to set it to 100 EU / m ² or less, and 50 EU / m ² or less is further used in order to obtain clinical effects such as an increase in patient hematocrit and reduction in erythropoietin dose and treatment costs. preferable. Recently, muramyl dipeptide, the smallest building block of endotoxin and peptidoglycan, has been suggested to induce cytokine production of human peripheral blood mononuclear cells, suggesting some stimulus from contaminated dialysate in long-term dialysis patients Therefore, 30 EU / m ² or less is more preferable.

  Various chemical species exist in glycolipids (endotoxins) having physiological activity. The term “endotoxin” as used in the present invention refers to microorganisms derived from microorganisms mainly present in process water from the viewpoint of contamination in the dry and wet spinning process. It indicates a toxin. As a means for easily detecting endotoxin, an endotoxin measurement method sold by each company as a kit was used. Endotoxin is a toxin of Gram-negative bacteria and is unsuitable for detection of Gram-positive bacteria, etc., but these measurement methods are cumbersome and not established. It was represented by endotoxin measurement based on the point of view of the increase. In addition, due to the characteristics of the test kit, it can be considered that all reacting fragments are detected.

  By setting the endotoxin extraction amount from the membrane in the above-mentioned range, it is considered that endotoxin eluted from the membrane itself will not adversely affect the human body. When a dialysate contaminated due to this trouble is used, it is preferable that the membrane itself has an endotoxin permeation suppressing function for the purpose of suppressing the inflow of endotoxin from the dialysate to the blood. Endotoxin is a lipopolysaccharide present in the outer cell wall of Gram-negative bacilli as described above, and its molecular weight is several thousand if it is small.

  In the present invention, the pore structure of the hollow fiber membrane and the hydrophilic / hydrophobic balance are controlled by optimizing the spinning conditions as described later to prevent permeation of endotoxin from the dialysate side to the blood side. However, even in the current science and technology, it is difficult to accurately measure the pore diameter of the hollow fiber membrane. In the present invention, the permeation blocking ability of endotoxin is substituted by setting the sieve coefficient of bovine plasma albumin to 0.3 or less. ing. More preferably, the sieving coefficient of bovine plasma albumin is 0.28 or less, more preferably 0.25 or less. The lower the sieving coefficient of bovine plasma albumin (the smaller the pore size), the less permeated endotoxin is preferable, but the removal of low molecular weight proteins such as β2MG, which is considered to be a causative agent of dialysis complications, is preferable. In order to obtain the property at a high level, the sieving coefficient of bovine plasma albumin is preferably 0.01 or more, more preferably 0.02 or more, and further preferably 0.03 or more.

  Further, since endotoxin exists from those having a molecular weight of several thousand as described above, β2MG (molecular weight 11,600) is represented by, for example, a hemodialysis membrane only by controlling the pore diameter of the hollow fiber membrane. It is impossible to achieve both the removal of such a low-molecular protein region and the permeation prevention of endotoxin (having a molecular weight of several thousand). As a result of diligent investigations on this point, the inventors confer endotoxin adsorption ability to the hollow fiber membrane by controlling the presence state of the hydrophilic polymer in the membrane, and can prevent endotoxin permeation regardless of the pore diameter. The present invention has been found. That is, since endotoxin has the property of being easily adsorbed to a hydrophobic material, in the hollow fiber membrane of the present invention, the content of the hydrophilic polymer in the vicinity of the inner surface is A, and the content of the hydrophilic polymer in the entire membrane is B. In this case, the relationship between A and B is preferably A ≧ 5 mass% and B / A ≦ 0.4. If the content of the hydrophilic polymer in the vicinity of the inner surface is 5% by mass or more, not only the priming property and solute permeability will be improved, but also the blood compatibility and performance will be stable over time because the adsorption of proteins in the blood will be suppressed. Sexuality is improved. More preferably, it is 6 mass% or more, More preferably, it is 7 mass% or more, More preferably, it is 8 mass% or more. A higher content of the hydrophilic polymer in the vicinity of the inner surface is preferable in terms of improving the above characteristics, but if it is too much, there is a possibility that the amount of the hydrophilic polymer eluted in the blood may increase. The following is preferred. More preferably, it is 23 mass% or less, More preferably, it is 21 mass% or less, More preferably, it is 19 mass% or less. Further, B / A ≦ 0.4 means that the non-blood contact portion occupying most of the hollow fiber membrane has a low hydrophilic polymer content, that is, these portions are more hydrophobic. Since it is strong, the adsorption capacity of endotoxin is increased, which is preferable. Further, since the content of the hydrophilic polymer is small, it is possible to reduce the amount of elution to the blood side. The B / A ratio is more preferably 0.38 or less, further preferably 0.36 or less, and still more preferably 0.34 or less. However, if B / A is too small, blood protein that permeates the membrane may be adsorbed and deposited, and the solute removability may be reduced. 0.21 or more is more preferable, and 0.24 or more is more preferable.

  The content of the hydrophilic polymer in the near-surface layer described above was measured by surface infrared spectroscopy (surface IR), and the absolute value of the content existing up to a depth corresponding to several hundred nm was evaluated. It is a thing. Further, the content of the hydrophilic polymer in the entire hollow fiber membrane of the hydrophilic polymer is measured and calculated by an elemental analysis method as described later.

Conventionally, in order to reduce long-term dialysis complications, attention has been focused on the removal of β2MG, which is thought to be the causative agent, but more recently, many dialysis patients complained of pruritus during dialysis. Although it has not been specified, it has been clinically experienced that it can be suppressed by removing a substance having a molecular weight of about 30,000, which is higher than β2MG. In this sense, it is desired to develop a selectively permeable separation membrane that can positively remove substances having a large molecular weight of about 30,000 in order to maintain a good prognosis. In order to enlarge the pore diameter of the selectively permeable separation membrane and minimize the leakage of albumin, which is a useful substance, it is necessary to narrow the pore distribution width of the membrane. However, there is a limit to reducing the width of the pore distribution of the membrane, and it is difficult to completely suppress the leakage of albumin (molecular weight 67,000) if pores that can remove substances with a molecular weight of about 30,000 are formed. is there. For example, in a hollow fiber membrane in which the sieve coefficient of bovine plasma albumin as described in Patent Document 5 is 0.005 or less (permeability 0.5% or less), the pore diameter is too small and endotoxin permeation can be suppressed. Therefore, it seems difficult to efficiently remove a substance having a molecular weight of about 30,000. Furthermore, when a hollow fiber membrane having such a pore size that is too small is used for a patient who has developed sepsis, an effective therapeutic effect cannot be expected because the causative bacteria flowing out of the septic lesion cannot be removed in the blood. In the present science and technology, it is difficult to directly measure the pore size of the membrane, and it is difficult to easily obtain plasma proteins having a molecular weight of about 30,000. Therefore, in the present invention, β2MG clearance and albumin The effective pore diameter of the membrane was defined by the relationship with the sieve coefficient. In other words, it was confirmed that if β2MG removal performance (β2MG clearance) measured with a module having a membrane area of 1.5 m ² is 40 mL / min or more, clinical effects such as reduction of pruritus can be expressed.

  As an example of a preferred embodiment for obtaining a selectively permeable separation membrane having the above characteristics, a hollow fiber membrane type selectively permeable separation membrane is exemplified below. The case where polysulfone (PS) or polyethersulfone (PES) is used as the hydrophobic polymer and polyvinylpyrrolidone (PVP) is used as the hydrophilic polymer will be described in detail. The film-forming stock solution consists of PS or PES, PVP and their common solvents. Examples of the solvent include N, N-dimethylformamide, N, N-dimethylacetamide (DMAc), dimethyl sulfoxide, and N-methyl-2-pyrrolidone (NMP). These may be used singly or in an arbitrary ratio. Is preferred. DMAc and NMP are more preferable because they are inexpensive and the safety of the solvent itself is high. Further, for the purpose of inducing microphase separation, a small amount of water or salts may be added as an additive. As the salts, inorganic salts such as lithium bromide are preferably used. Furthermore, in order to form a hollow film, an appropriate viscosity is required. For this purpose, the preferred composition is 15 to 40% by mass of the polysulfone polymer, 0 to 30% by mass of PVP, and the rest Solvent and / or additive.

When the obtained spinning dope and internal coagulation liquid are discharged from a double tube spinning nozzle, passed through an air gap part, and guided to a coagulation bath to obtain a hollow fiber membrane, the spinning dope composition, internal coagulation liquid and air gap part length The atmosphere and external coagulation liquid conditions are the main items for controlling the permeation performance of the membrane. Although detailed means for obtaining the transmission performance of the present invention will be described later, the present invention consists of three points. That is,
1) Suppresses contamination of bacteria (endotoxin) in the process.
2) Adjust the sieving coefficient of albumin.
3) Adjust β2MG clearance.

  The first point for obtaining the hollow fiber membrane of the present invention is preferably to prevent contamination of water used in order to prevent contamination of cells and endotoxin during the membrane formation process. Specifically, as a means for suppressing contamination of bacterial cells and endotoxin, it is preferable to carry out the film forming operation in a class 100,000 clean room. In addition, for example, a washing machine, a dryer, a caulking machine, a polyethylene film, or the like used for forming a film, a device or a member that comes in contact with the hollow fiber membrane uses a 70 to 90 wt% aqueous ethanol solution having a strong bactericidal power every time it is used. It is preferable to sterilize. It is also a preferable aspect that the worker wears gloves subjected to UV sterilization to perform the film forming operation. It is preferable that all water used in the membrane-forming process is permeated through a reverse osmosis (RO) membrane system, and RO water whose endotoxin amount in the endotoxin measurement described later is below detection sensitivity is used. In addition, in order to prevent droplets contaminated with bacteria from adhering to the film-forming equipment and components from the floor, all film-forming process equipment should be installed at a height of 50 cm or more from the floor. Is also a preferred embodiment. Air used for drying is purified through a HEPA filter in the same way as in a clean room. As described above, it is possible to obtain a target hollow fiber membrane by preventing contamination of bacterial cells and endotoxin during the thorough membrane formation process.

  Moreover, although the optimal growth conditions differ according to the origin, the growth of cells in a high humidity state is usually remarkable at around 37 ° C. Therefore, it is a preferable embodiment to avoid these conditions as storage conditions of the dried membrane from the production of the hollow fiber membrane to the module assembly and sterilization treatment. For example, it is preferable to store at room temperature (25 ° C.) and a relative humidity of 50 Rh% or less, and 25 to 50 Rh% is a more preferable embodiment from the viewpoint of workability. Further, storing at a low temperature is also a preferred embodiment, and storing at 20 ° C. or lower is preferable. In the cryopreservation, the range exceeding 0 ° C. is a preferred embodiment because there is concern about the destruction of the membrane structure due to the volume expansion of the remaining water and the cost of equipment. Since management is easy, 5 to 15 degreeC is a more preferable embodiment. It is also a preferred embodiment to reduce the oxygen concentration in the preservation system as aerobic bacteria suppression. In particular, since some anaerobic bacteria often have nitrogen as a breeding factor, it is a more preferable embodiment to replace the storage system with argon. Further, combining the above, and making the storage system in an argon atmosphere and storing at a low temperature is a preferred embodiment in that the antibacterial effect can be expected significantly.

  The second point for obtaining the selectively permeable separation membrane of the present invention is to optimize the pore diameter, pore distribution, and hydrophilic polymer content on the blood contact side so that the sieving coefficient of albumin is 0.01 to 0.3. It is preferable to do this. The means for controlling the pore diameter and pore distribution on the blood contact side is preferably based on a combination of the spinning solution composition, the nozzle temperature, and the internal coagulation solution composition. Specifically, the spinning dope composition is preferably hydrophobic polymer / hydrophilic polymer / solvent / non-solvent = 10-30 / 0-15 / 25-85 / 0-10.

  If the content of the hydrophobic polymer in the spinning dope is too small, it may be difficult to keep the pore size of the selectively permeable separation membrane small enough to suppress albumin permeation. On the contrary, if the content of the hydrophobic polymer is too large, it is difficult to increase the porosity, and the clearance of β2MG may be insufficient. In addition, if the amount of hydrophilic polymer in the spinning dope is too low, the effect of suppressing endotoxin permeation increases, but the hydrophobicity of the selectively permeable separation membrane is too strong, which increases the amount of protein adsorbed in the blood. It may be difficult to obtain albumin sieving coefficient and β2MG clearance. On the other hand, if the content of the hydrophilic polymer is too large, there is a possibility that problems such as a decrease in the strength and elongation of the hollow fiber membrane or the permeation of endotoxin easily occur. Therefore, hydrophobic polymer / hydrophilic polymer / solvent / non-solvent in the spinning dope is more preferably 12 to 28/1 to 14/28 to 82/0 to 9, more preferably 14 to 26/2 to 13/30 to 80 / 0-8 are still more preferred, and 15-25 / 3-12 / 32-78 / 0-8 are even more preferred. The ratio of the hydrophilic polymer to the hydrophobic polymer in the spinning dope is preferably 55% by mass or less, more preferably 50% by mass or less, further preferably 45% by mass or less, and still more preferably 40% by mass or less, 35 mass% or less is especially preferable. If the ratio is too large, the content of the hydrophilic polymer in the hollow fiber membrane increases, so that there are problems such as elution during hemodialysis and the like, and endotoxin easily permeates from the dialysate side to the blood side. May occur. On the other hand, if the ratio is too small, the hydrophilic polymer content in the vicinity of the inner surface of the hollow fiber membrane is reduced, and blood compatibility and dialysis performance may be reduced. Accordingly, the ratio is preferably 5% by mass or more, more preferably 7% by mass or more, further preferably 9% by mass or more, and further preferably 11% by mass or more.

  The nozzle temperature is preferably 30 to 80 ° C. If the nozzle temperature is too low, the viscosity of the spinning dope will increase, which may cause problems such as reduced discharge stability or difficulty in stabilizing the nozzle temperature due to a small difference from the outside air temperature. . On the other hand, if the nozzle temperature is too high, the viscosity of the spinning dope becomes low, so that the spinnability is lowered and the spinning stability, and consequently the performance stability in the length direction of the hollow fiber membrane may be lowered. Therefore, the nozzle temperature is more preferably 35 to 75 ° C, and further preferably 40 to 70 ° C.

  As the internal coagulation liquid, it is preferable to use a mixed solution of water and a solvent, and the concentration of the internal coagulation liquid is preferably 20 to 60% by mass. The lower the internal coagulating liquid concentration, the smaller the pore diameter on the inner surface side of the hollow fiber membrane, but it may be difficult to increase the albumin sieving coefficient to 0.01 or more and increase the β2MG clearance. On the other hand, if the concentration of the internal coagulating liquid is too high, the pore diameter of the dense layer becomes large, and it may be difficult to suppress the albumin sieving coefficient to 0.3 or less and to suppress permeation of endotoxin. Therefore, the internal coagulation liquid concentration is more preferably 25 to 55% by mass, and further preferably 30 to 50% by mass.

The third point for obtaining the hollow fiber membrane of the present invention is that the air gap is added to the above-described means for controlling the albumin sieving coefficient so that the clearance of β2MG is 40 mL / min (1.5 m ² ) or more. It is preferable to optimize the conditions, nozzle diameter, and external coagulation liquid conditions. Specifically, it is preferable to surround the air gap portion with a member that shields it from the outside air. The air gap length is preferably 100 to 1000 mm. The nozzle is preferably a double circular nozzle having an inner diameter of the spinning solution discharge hole of 150 to 400 μm. The external coagulation liquid concentration is preferably 0 to 50% by mass. The external coagulation liquid temperature is preferably 20 to 80 ° C.

  By surrounding the air gap portion with a member that shields it from the outside air, the inside of the air gap is filled with the solvent vapor from the spinning dope and the solvent and water vapor from the external coagulation liquid. That is, by controlling the solvent and the amount of water vapor in the air gap, the pore diameter and the open area ratio and the hydrophilic polymer content on the outer surface of the hollow fiber membrane can be controlled. In addition, when the air gap length is short, the pore diameter on the outer surface of the hollow fiber membrane is reduced, so that the reverse influx of endotoxin is reduced, but it may be difficult to improve the β2MG clearance. On the other hand, if the air gap length is long, the outer surface hole diameter is enlarged and the β2MG clearance is improved, but the porosity becomes too high and the hollow fiber membrane strength may be insufficient. Therefore, the air gap length is more preferably 150 to 900 mm, further preferably 200 to 800 mm, and further preferably 250 to 750 mm.

  The present invention is characterized in that a solidifying liquid is used as the internal coagulating liquid with respect to the spinning dope. When using a liquid having coagulation properties with respect to such a spinning stock solution, the inner diameter of the hollow fiber membrane is determined by the inner diameter of the spinning stock solution discharge hole of the nozzle. Since the spinning stock solution and the internal coagulation solution are simultaneously discharged from the nozzle, the spinning raw solution starts coagulation immediately after being discharged from the nozzle. At this time, if excessive tension and stretching are applied to the spinning dope, the solidified pores may be deformed or cracked. When deformation or cracks occur in the pores, the pore distribution expands, and it becomes difficult to obtain sharp fractionation characteristics. For the same reason, it is preferable that the spinning dope is not substantially stretched after the coagulation and phase separation is completely completed. “Substantially not stretching” means that the hollow fiber membrane does not pull more than a tension that does not cause slack during running in the spinning process. The inner diameter of a general blood purification hollow fiber membrane is about 150 to 250 μm, and the inner diameter of the spinning stock solution discharge hole of the double circular nozzle is more preferably 160 to 350 μm, more preferably 170 to 300 μm, and still more preferably. 180-250 μm.

If the concentration of the external coagulating liquid is too high, phase separation on the outer surface side of the hollow fiber membrane that has passed through the air gap and is immersed in the external coagulating liquid grows, and the hole diameter and the hole area ratio may become too high. That is, the strength of the hollow fiber membrane may be reduced, or the reverse influx of endotoxin may increase. Conversely, if the concentration of the external coagulation liquid is too low, solidification is dominant over phase separation, and the pore diameter from the support layer to the outer surface may be too small. In this case, the reverse influx of endotoxin from the dialysate side to the blood side decreases, but the clearance may not be improved because the permeation resistance of β2MG in the membrane cross-sectional direction increases. Therefore, the external coagulation liquid concentration is more preferably 3 to 40% by mass, further preferably 5 to 35% by mass, and further preferably 7 to 30% by mass.
Also, if the external coagulation liquid temperature is too high, the solvent vapor and water vapor inside the air gap will become mist, and the state of solidification and phase separation will change where it adheres to the spinning stock solution and where it does not adhere. Spots may occur in the surface pore diameter and pore structure. That is, the pore diameter is uneven, and there is a possibility that sharp fractionation characteristics cannot be obtained. On the other hand, if the external coagulation liquid temperature is too low, the solvent exchange on the outer surface side becomes slow, so phase separation does not grow, and the hole diameter and the hole area ratio on the outer surface side may become small. That is, it may be difficult to improve the β2MG clearance. Therefore, the external coagulation liquid temperature is more preferably 30 to 80 ° C, further preferably 40 to 75 ° C, and further preferably 50 to 75 ° C.

  The hollow fiber membrane after film formation is wound into a bag in a wet state and then bundled into 3,000 to 20,000 bundles. Next, it is preferable to wash the obtained hollow fiber membrane bundle to remove excess solvent and hydrophilic polymer. As a method for washing the hollow fiber membrane bundle, in the present invention, it is preferable to treat the hollow fiber membrane bundle by immersing it in an aqueous solution of ethanol or isopropanol at room temperature to 50 ° C. and 10 to 40 vol%. Specifically, the hollow fiber membrane bundle is immersed in an excess aqueous ethanol or isopropanol solution and treated at room temperature to 50 ° C. for 15 to 60 minutes, and then the hollow fiber membrane bundle is taken out and subjected to centrifugal drainage. This operation is repeated 3 or 4 times while renewing the aqueous alcohol solution to perform washing and draining treatment. When the processing temperature is too low, it is necessary to increase the number of times of cleaning, which may lead to an increase in cost. In addition, when the treatment temperature is too high, the decomposition of the hydrophilic polymer is accelerated, and the elution amount of the hydrophilic polymer may not be reduced even by washing. By performing the above washing, not only the cells and endotoxin in the hollow fiber membrane can be removed, but also the content of the hydrophilic polymer in the entire hollow fiber membrane is optimized to suppress endotoxin permeation and hydrophilic polymer. It becomes possible to reduce the elution amount of.

  The albumin sieve coefficient is an index of the pore diameter of the hollow fiber membrane, and β2MG clearance can be replaced with an index of the number of pores and the curvature. Here, the solute permeation characteristic of the membrane is a function of the pore diameter, the number of pores, and the pore curvature of the membrane. That is, the pore diameter is increased to such an extent that albumin (Stokes radius of 35 mm) does not permeate, the porosity of the membrane is increased (increase the number of pores), and the film thickness is reduced to reduce β2MG (Stokes radius of 16 mm). It is possible to improve the property. By taking the film forming conditions, the porosity of the hollow fiber membrane can be 50% or more. A higher porosity is preferable from the viewpoint of solute removal, but if it is too high, the film strength may not be maintained, so 85% or less is preferable. A more preferable porosity is 55 to 85%, and further preferably 60 to 85%. The film thickness is preferably 10 to 50 μm. The film thickness is preferably thin for the above reasons, but if it is too thin, the film strength may not be maintained, so it is more preferably 15 μm or more, further preferably 20 μm or more, and even more preferably 25 μm or more. If the film thickness is too large, the washability of the endotoxin in the film or excess hydrophilic polymer may be lowered, so that it is more preferably 45 μm or less, still more preferably 40 μm or less, and even more preferably 35 μm or less.

  Hereinafter, the effectiveness of the present invention will be described with reference to examples, but the present invention is not limited thereto. In addition, the evaluation method of the physical property in the following examples is as follows.

1. Calculation of membrane area The membrane area of the dialyzer is determined as a reference for the inner diameter of the hollow fiber.
A = n × π × d × L
Here, n is the number of hollow fibers in the dialyzer, π is the circumference, d is the inner diameter (m) of the hollow fiber, and L is the effective length (m) of the hollow fiber in the dialyzer.

2. Measurement of endotoxin concentration (sample preparation)
All sample preparations were performed in a clean room.
1) In the case of a hollow fiber membrane bundle A polyethylene film previously sterilized with alcohol was loaded with a hollow fiber membrane bundle having a predetermined membrane area.
2) In the case of module shape Both ends are excised with a pipe cutter previously sterilized with alcohol, and the hollow fiber membrane bundle is taken out. This hollow fiber membrane bundle was used in which a hollow fiber membrane bundle having a predetermined membrane area was loaded on a polyethylene film sterilized and sterilized with alcohol in advance.
(Sample preparation)
All the instruments used for sample preparation were used after autoclaving at 121 ° C. for 20 minutes. For the sample extract, Japanese Pharmacopoeia water for injection / Otsuka distilled water was used. Prepare 500 mL of water for injection per 1 m ² (inner diameter standard) of hollow fiber membrane bundles in a 1 L Erlenmeyer flask with a wide mouth, adjust the temperature to 25 ° C, hold a polyethylene film with gloves sterilized with a UV germicidal lamp, and hollow Put the hollow fiber membrane into the water for injection so as not to touch the thread membrane bundle at all. In this state, the mixture was stirred lightly every 30 minutes and extracted at 25 ° C. for 2 hours.
(Sample measurement)
The endotoxin concentration was measured using the supernatant of the extract thus obtained. Endotoxin concentration was analyzed using Limulus ESII Test Wako (manufactured by Wako Pure Chemical Industries, Ltd.) according to the method described in the instruction manual.

3. Measurement of sieve coefficient of bovine plasma albumin In the case of hollow fiber membrane bundle, after loading in the module case, both ends are sealed with urethane resin, both ends are opened by cutting to form a hollow fiber membrane module. Carried out. The sample may or may not be gamma sterilized.
(Preparation of 1% bovine plasma albumin solution)
Solution A: 53.72 g of Na ₂ HPO ₄ · 12H ₂ O and 26.30 g of NaCl are dissolved in 3 L of pure water.
Solution B: Dissolve 20.42 g of KH ₂ PO ₄ and 26.30 g of NaCl in 3 L of pure water.
Add solution B to solution A and adjust to pH = 7.5 ± 0.1.
30 g of bovine plasma albumin (Wako Pure Chemical Industries, Ltd.) is dissolved in 3 L of this phosphate buffer. After dissolution, adjust to pH = 7.5 ± 0.1 again using 1N-NaOH.
(Preparation of module)
Pure water was supplied at 500 mL / min for 5 minutes through the dialysis fluid side channel of the module, and then was passed through the blood side channel at 200 mL / min for 5 minutes. Subsequently, the phosphate buffer solution 500 mL / min was passed through the dialysate side channel of the module for 5 minutes, and then the blood side channel was passed at 200 mL / min for 5 minutes. Thereafter, the solution was passed for 3 minutes while filtering from the blood side to the dialysate side.
(Measurement)
Connect the measuring solution circuit to the blood side and discard the priming solution (phosphate buffer solution) on the dialysate side. Place in a 37 ° C thermostatic chamber, seal the dialysate side, and flow the blood side at 200 mL / min for 1 minute to remove the remaining priming solution on the blood side. Next, connect a circuit to the dialysate inlet, set the blood flow to 200 mL / min, set the flow rate of the filtration circuit connected to the dialysate inlet to 30 mL / min, and return the blood permeate and filtrate to the test solution. Conduct the test. A test solution, a blood side permeate, and a filtrate 15 minutes after the start of circulation were collected. The collected sample was diluted 10 times with pure water (the filtrate is preferably undiluted), and the absorbance was measured with a spectrometer at a wavelength of 280 nm. The sieving coefficient of albumin is based on the absorbance of each
SCalb = 2 × Cf / (Cb + Co)
(Where Cf is the absorbance of the filtrate, Cb is the absorbance of the test solution, and Co is the absorbance of the blood permeate, and when diluted, the respective dilution factors are multiplied).

(Beta MG clearance measurement)
A hollow fiber membrane module (dialyzer) having a membrane area of 1.5 m ^{2 on} the basis of the inner diameter is prepared for β2MG clearance measurement. ACD-added bovine plasma prepared at a total protein concentration of 7.0 ± 0.5 g / dL and kept at 37 ° C. is flowed at a blood flow rate of 200 mL / min., Commercial dialysis fluid is flowed at 500 mL / min, and a filtration flow rate is 15 mL / min. The dialyzer outlet plasma is a recycling circuit that returns to the beaker containing the original plasma together with the ACD-added physiological saline for the filtration flow rate. To the bovine plasma, human β2MG is added to a concentration of 0.05 to 0.1 mg / L. Measure the β2MG concentration in the blood inlet, outlet, and dialysate outlet for 20 minutes at intervals of 5 minutes from the start of dialysis. The clearance is calculated using the following formula.
Clearance (β2MG) = 200 × [(200 × CBi) − (185 × CBo)] / (200 × CBi)
Here, CBi: blood inlet concentration, CBo: blood outlet concentration. Also, when% MBE calculated by the following formula exceeds ± 50%, it is not adopted as data.
% MBE = 100 × (MB−MD) / MD
Here, MB: (200 × CBi) − (185 × CBo), MD: 515 × CDo.

(Measurement of hydrophilic polymer content near the inner surface)
A measurement method when PVP is used as the hydrophilic polymer will be exemplified. Measurement was performed by infrared absorption analysis. Among the hydrophilic polymers described above, a measurement sample prepared by the same method as the existing ratio in the outermost layer on the outer surface was used, the measurement in the vicinity of the surface was performed by the ATR method, and the entire membrane was measured by the transmission method. In the ATR method, an infrared absorption spectrum was measured by a method using diamond 45 ° as an internal reflection element. For the measurement, IRμs / SIRM manufactured by SPECTRA TECH was used. In the infrared absorption spectrum, the ratio Ap / As of the peak absorption intensity Ap derived from C═O of PVP near 1675 cm ⁻¹ and the peak absorption intensity As derived from polysulfone-based polymer near 1580 cm ⁻¹ was determined. In the ATR method, since the absorption intensity depends on the measured wave number, a ratio υp / υs between the peak position υs of the polysulfone polymer and the peak position υp (wave number) of the PVP was applied to the measured value as a correction value. The existence ratio of PVP in the vicinity layer of the blood contact surface was calculated by the following formula.
Presence ratio (mass%) of hydrophilic polymer in the surface vicinity layer = Cav × Ap / As × υp / υs
However, Cav is the PVP content obtained by the elemental analysis method described later.

(Measurement of hydrophilic polymer content in the entire hollow fiber membrane)
A measurement method when PVP is used as the hydrophilic polymer will be exemplified. The sample was dried at 80 ° C. for 48 hours using a vacuum dryer, 10 mg of which was analyzed with a CHN coder (manufactured by Yanaco Analytical Co., Ltd., MT-6 type), and the mass ratio of PVP from the nitrogen content was expressed by the following formula: Calculated with
Mass ratio (mass%) of PVP = nitrogen content (mass%) × 111/14

(Measurement of endotoxin inhibition performance)
Endotoxin removal performance was measured as follows. A module filled with a predetermined number of hollow fiber membranes was prepared. One feed liquid inlet was provided on the outside of the hollow fiber of the module and one filtrate outlet was provided on the inside of the hollow fiber of the module, primed with pure water in advance before the measurement, air in the module was removed, and the temperature was kept at 37 ° C. To the Feed side, pure water at 37 ° C. was fed at a rate of 500 mL / min per 1.5 m ² membrane area, and the whole amount was filtered. Filtration was continued for 5 hours to eliminate measurement errors due to endotoxin adhering to the circuit and case. When the endotoxin aqueous solution dissolved in pure water so that the endotoxin concentration becomes 10,000 to 100,000 EU / L is filtered at 10 mL at a filtration rate of 500 mL / min per 1.5 m ^{2 of the} hollow fiber membrane area, Endotoxin concentration was measured. Endotoxin concentration was measured using Limulus Test Wako (Wako Pure Chemical Industries). When the endotoxin concentration in the filtrate was below the detection sensitivity, it was regarded as acceptable.

(Example 1)
18.5% by mass of polyethersulfone (manufactured by Sumika Chemtex, Sumika Excel 5200P), 3.0% by mass of polyvinylpyrrolidone (Collidon K-90, manufactured by BASF), 74.5% by mass of dimethylacetamide (DMAc), RO water 4.0% by mass was uniformly dissolved. The ratio of hydrophilic polymer to hydrophobic polymer was 16% by mass. The obtained spinning stock solution was discharged from a tube-in orifice nozzle heated to 65 ° C. together with a 40% by mass DMAc aqueous solution as an internal coagulation solution, passed through a 500 mm dry section cut off from the outside air by a spinning tube, then 75 ° C., The solution was solidified in a 20% by mass DMAc aqueous solution and rolled up in a wet state. A polyethylene film whose surface on the hollow fiber bundle side was sterilized and sterilized with a 70 vol% aqueous ethanol solution was wound around the bundle of about 10,000 hollow fiber membranes, and then cut into a length of 27 cm. The obtained hollow fiber membrane bundle was immersed in an aqueous isopropanol (IPA) solution at 40 ° C. × 30 vol% for 30 minutes. This operation was performed twice with the solution changed, and endotoxin and excess hydrophilic polymer were washed away. The washed hollow fiber membrane bundle was rinsed lightly with RO water to remove IPA, and then put into a ventilating dryer and dried at 45 ° C. for 12 hours. In addition, all film forming processes other than raw material input were performed in a clean room corresponding to the class 100,000 standard, all equipment was installed 150 cm above the floor, and water used was RO water whose viable bacteria and endotoxin were below the detection sensitivity. A HEPA filter was installed just before the air flowing into the dryer to clean it. All equipment used was sterilized and disinfected with a 70 vol% aqueous ethanol solution before use. The worker wore gloves sterilized with UV and performed the work. The obtained hollow fiber membrane bundle was put into a commercially available aluminum laminate bag (ethylene oxide gas (EOG) sterilized), the bag was sealed with argon, sealed, and stored in a refrigerator at 7 ° C. for 3 months.

  Using the hollow fiber membrane after storage, the endotoxin concentration, the vicinity of the inner surface and the entire hydrophilic polymer content, porosity, and film thickness were measured. The results are shown in Table 1.

  Moreover, the hollow fiber membrane bundle after storage was assembled in a module, and the sieve coefficient of bovine plasma albumin, the clearance of β2MG, and endotoxin permeation were measured. The results are shown in Table 1.

(Comparative Example 1)
A dried hollow fiber membrane bundle was prepared in the same manner as in Example 1 except that the height of the external coagulation liquid tank was 30 cm above the floor, the hollow fiber membrane bundle was not washed, the film / equipment was not sterilized, and the water was ion-exchanged water. Obtained. The obtained hollow fiber membrane bundle was put in a commercially available aluminum laminate bag and stored in a refrigerator at 7 ° C. for 3 months.

When the endotoxin concentration, the inner surface vicinity and the entire hydrophilic polymer content, porosity, and film thickness were measured using the hollow fiber membrane after storage, the endotoxin concentration showed a high value of 250 EU / m ² . In this film formation, in addition to the use of viable bacteria and water having a high endotoxin level, washing before drying was not performed, so that a large amount of endotoxin remained in the film. Other results are shown in Table 1.

  In addition, the hollow fiber membrane bundle after storage was assembled in a module, and the sieve coefficient of bovine plasma albumin and the clearance of β2MG were measured. There were no major problems except for the high endotoxin level in the membrane. The results are shown in Table 1.

(Example 2)
Polysulfone (Amoco, P-3500) 17.5 wt%, Polyvinylpyrrolidone (BASF Kollidon K-60) 9.0 wt%, Dimethylacetamide (DMAc) 72.5 wt%, RO water 1.0 wt% % Was dissolved uniformly. The ratio of the hydrophilic polymer to the hydrophobic polymer was 50% by mass. The obtained spinning stock solution was discharged from a tube-in orifice nozzle heated to 55 ° C. together with a 35% by mass DMAc aqueous solution as an internal coagulation solution, passed through a 650 mm dry section cut off from the outside air by a spinning tube, The solution was solidified in a 10% by mass DMAc aqueous solution and rolled up in a wet state. A polyethylene film whose surface on the hollow fiber bundle side was sterilized and sterilized with a 70 vol% aqueous ethanol solution was wound around the bundle of about 10,000 hollow fiber membranes, and then cut into a length of 27 cm. The obtained hollow fiber membrane bundle was immersed in an aqueous ethanol solution at 30 ° C. × 30 vol% for 30 minutes. This operation was performed three times with the liquid changed, and endotoxin and excess hydrophilic polymer were washed away. The hollow fiber membrane bundle after washing was rinsed lightly with RO water, and ethanol was replaced with RO water. Then, it was placed in a ventilating dryer and dried at 40 ° C. for 16 hours. In addition, all film forming processes other than raw material input were performed in a clean room corresponding to the class 100,000 standard, all equipment was installed 150 cm above the floor, and water used was RO water whose viable bacteria and endotoxin were below the detection sensitivity. A HEPA filter was installed just before the air flowing into the dryer to clean it. All equipment used was sterilized and disinfected with a 70 vol% aqueous ethanol solution before use. The worker wore gloves sterilized with UV and performed the work. The obtained hollow fiber membrane bundle was put in a commercially available polyethylene bag (EOG sterilized), the inside of the bag was purged with argon, and stored in a refrigerator at 7 ° C. for 3 months.

  Using the hollow fiber membrane after storage, the endotoxin concentration, the vicinity of the inner surface and the entire hydrophilic polymer content, porosity, and film thickness were measured. The results are shown in Table 1.

  Moreover, the hollow fiber membrane bundle after storage was assembled in a module, and the sieve coefficient of bovine plasma albumin, the clearance of β2MG, and endotoxin permeation were measured. The results are shown in Table 1.

(Comparative Example 2)
The membrane was formed in the same manner as in Example 2 except that the concentration of the internal coagulation liquid was 65% by mass, the height of the external coagulation liquid tank was 30 cm, the hollow fiber membrane bundle was not washed, and ion-exchanged water was used as the water. A dry hollow fiber membrane bundle was obtained. The obtained hollow fiber membrane bundle was put in a commercially available polyethylene bag (EOG sterilized) and stored in a refrigerator at 7 ° C. for 3 months.

When the endotoxin concentration, the vicinity of the inner surface and the entire hydrophilic polymer content, porosity, and film thickness were measured using the hollow fiber membrane after storage, the endotoxin concentration showed a high value of 290 EU / m ² . In this film formation, in addition to the use of viable bacteria and water having a high endotoxin level, washing before drying was not performed, so that a large amount of endotoxin remained in the film. Other results are shown in Table 1.

  The hollow fiber membrane bundle after storage was assembled into a module, and the sieve coefficient, β2MG clearance, and endotoxin permeation of bovine plasma albumin were measured. The β2MG clearance was high, but the albumin sieving coefficient was as high as 0.37. In addition, permeation of endotoxin was observed. Since the internal coagulating liquid concentration was increased, it seems that not only the inner surface of the hollow fiber membrane but also the pore diameter as a whole became too large. In addition, it seems that endotoxin easily permeates due to the high hydrophilic polymer content in the membrane. Other results are shown in Table 1.

(Example 3)
Polyethersulfone (Sumika Excel 4800P manufactured by Sumika Chemtex Co., Ltd.) 18.0% by mass, polyvinylpyrrolidone (BASF Koridone K-90) 6.0% by mass, dimethylacetamide (DMAc) 73.0% by mass, RO water 3 0.0% by mass was dissolved uniformly. The ratio of the hydrophilic polymer to the hydrophobic polymer was 33% by mass. The obtained spinning stock solution was discharged from a tube-in orifice nozzle heated to 50 ° C. together with a 30% by mass DMAc aqueous solution as an internal coagulation solution, passed through a 400 mm dry section cut off from the outside air by a spinning tube, then 45 ° C., The solution was solidified in a 20% by mass DMAc aqueous solution and rolled up in a wet state. A polyethylene film whose surface on the hollow fiber bundle side was sterilized and sterilized with a 70 vol% aqueous ethanol solution was wound around the bundle of about 10,000 hollow fiber membranes, and then cut into a length of 27 cm. The obtained hollow fiber membrane bundle was immersed in an ethanol aqueous solution of 40 ° C. × 40 vol% for 30 minutes. This operation was performed twice with the solution changed, and endotoxin and excess hydrophilic polymer were washed away. The washed hollow fiber membrane bundle was rinsed lightly with RO water, and ethanol was replaced with RO water. Then, it was placed in a ventilating dryer and dried at 60 ° C. for 9 hours. In addition, all film forming processes other than raw material input were performed in a clean room corresponding to the class 100,000 standard, all equipment was installed 150 cm above the floor, and water used was RO water with viable bacteria and endotoxin below the detection sensitivity. A HEPA filter was installed just before the air flowing into the dryer to clean it. All equipment used was sterilized and disinfected with a 70 vol% aqueous ethanol solution before use. The worker wore gloves sterilized with UV and performed the work. The obtained hollow fiber membrane bundle was placed in a commercially available polyethylene bag (EOG sterilized), the inside of the bag was purged with argon, and then stored in a warehouse at 37 ° C. for 1 month.

  Using the hollow fiber membrane after storage, the endotoxin concentration, the vicinity of the inner surface and the entire hydrophilic polymer content, porosity, and film thickness were measured. The results are shown in Table 1.

  Further, the obtained hollow fiber membrane bundle was assembled into a module, and the sieve coefficient of bovine plasma albumin and the clearance of β2MG were measured. The results are shown in Table 1.

(Comparative Example 3)
Polysulfone (Amoco P-3500) 20.0 mass%, polyvinylpyrrolidone (BASF Kollidon K-90) 1.0 mass%, dimethylacetamide (DMAc) 76.0 mass%, RO water 3.0 mass% Was dissolved uniformly. The ratio of the hydrophilic polymer to the hydrophobic polymer was 0.5% by mass. The obtained spinning stock solution was discharged from a tube-in orifice nozzle heated to 50 ° C. together with RO water as an internal coagulating solution, passed through a 400 mm dry section cut off from the outside air by a spinning tube, and then 45 ° C., 20% by mass. The solution was coagulated in an aqueous solution of DMAc and rolled up in a wet state. A polyethylene film whose surface on the hollow fiber bundle side was sterilized and sterilized with a 70 vol% aqueous ethanol solution was wound around the bundle of about 10,000 hollow fiber membranes, and then cut into a length of 27 cm. The obtained hollow fiber membrane bundle was immersed in an ethanol aqueous solution of 40 ° C. × 40 vol% for 30 minutes. This operation was performed twice with the solution changed, and endotoxin and excess hydrophilic polymer were washed away. The washed hollow fiber membrane bundle was rinsed lightly with RO water, and ethanol was replaced with RO water. Then, it was placed in a ventilating dryer and dried at 60 ° C. for 9 hours. In addition, all the film forming processes were performed outside the clean room, all the equipment was installed 150 cm above the floor, and ion-exchanged water was used as the water. The air flowing into the dryer was not treated with a HEPA filter. All equipment used was sterilized and disinfected with a 70 vol% aqueous ethanol solution before use. The worker wore gloves sterilized with UV and performed the work. The obtained hollow fiber membrane bundle was put in a commercially available polyethylene bag (EOG sterilized) and stored in a warehouse at 37 ° C. for 1 month.

Using the obtained hollow fiber membrane, the endotoxin concentration, the vicinity of the inner surface and the entire hydrophilic polymer content, porosity, and film thickness were measured, and the endotoxin concentration showed a high value of 130 EU / m ² . This is probably because the hollow fiber membrane was produced outside the clean room and the air used for drying was not sterilized during the membrane formation, so that the bacteria adhered during drying and grew during storage. Other results are shown in Table 1.

Further, the obtained hollow fiber membrane bundle was assembled into a module, and the sieve coefficient of bovine plasma albumin and the clearance of β2MG were measured. The albumin sieving coefficient was as small as 0.003, and the clearance of β2MG was as low as 22 mL / min (1.5 m ² ). This is probably because the pore diameter on the inner surface of the hollow fiber membrane became too small because the internal coagulating liquid concentration was lowered, so that the permeation resistance of the solute was increased, and the albumin sieve coefficient and β2MG were decreased. Other results are shown in Table 1.

  The selectively permeable separation membrane of the present invention can provide a selectively permeable separation membrane in which elution of endotoxin contained in the membrane as an impurity is reduced and irritation to a living body is reduced. Therefore, it is important to contribute to the industry.

Claims (8)

A selectively permeable separation membrane characterized in that the endotoxin eluted from the selectively permeable separation membrane is 150 EU / m ² or less.   2. The selectively permeable separation membrane according to claim 1, wherein the sieving coefficient of albumin is 0.01 or more and 0.30 or less. A permselective separation membrane is installed in the module so that the membrane area is 1.5 m ² , and bovine plasma with a total protein concentration of 7.0 g / dL ± 0.5 g is filtered into the module at a blood flow rate of 200 mL / min, dialysate flow rate of 500 mL / min 3. The permselective separation membrane according to claim 1, wherein the clearance of β2 microglobulin measured after 20 minutes is 40 mL / min or more when flowing at a rate of 15 mL / min.   4. The selectively permeable separation membrane according to claim 1, wherein the membrane is mainly composed of a hydrophobic polymer and a hydrophilic polymer.   When the content of the hydrophilic polymer in the vicinity of the blood contact surface of the selectively permeable separation membrane is A and the content of the hydrophilic polymer in the entire membrane is B, A ≧ 5 mass%, B / A ≦ 0.4 The selectively permeable separation membrane according to claim 1, wherein the relationship is satisfied.   6. The permselective separation membrane according to claim 1, wherein the hydrophobic polymer is a polysulfone polymer.   7. The permselective separation membrane according to claim 1, wherein the hydrophilic polymer is polyvinylpyrrolidone.   The selectively permeable separation membrane according to any one of claims 1 to 7, wherein the selectively permeable separation membrane is a hollow fiber membrane.