Particle separation device and separation method

Abstract

To provide a particle separator and a particle separation method improving particle separation efficiency in a PFF (Pinched Flow Fractionation) technique.SOLUTION: In a particle separator, it is found out that a particle separation-treating amount improves nearly proportionately to the duct width 51 of a narrow part in a y-axis direction by elongating the above duct width 51 when defining a flow direction (longitudinal direction) of the duct as x-axis, a direction vertical to an interface between the sample and sheath liquids as z-axis and a direction orthogonal to x- and z-axes as y axis. Besides, it is found that both treatment amount and separability are improved by restraining length variation of the duct width 51 of the narrow part in a z-axis direction. Further, it is found out that the separability together with the treatment amount improves by making the sample and sheath liquids flow in the narrow part so that a flow ratio of both liquids becomes constant in the narrow part.SELECTED DRAWING: Figure 4

Description

[Technical field]

[0001] The present invention relates to an apparatus for separating particles contained in a fluid and a method for manipulating particles contained in a fluid using said apparatus. [Background technology]

[0002] Techniques for separating particles and cells on the microscale or smaller by size are used in biochemistry, diagnostic medicine, and various industrial fields. For example, hematopoietic stem cells that differentiate into other cells and stem cells such as ES cells need to be separated from many cell groups to study cell differentiation. In addition, testing for the presence of circulating tumor cancer cells (CTCs) in a patient's blood to diagnose cancer is attracting attention as a new method of cancer screening. These stem cells and CTCs generally differ in size from other cells, and it is expected that they can be separated from other cell groups by utilizing the size difference, so techniques for separating particles and cells on the microscale or smaller by size difference are extremely important.

[0003] Techniques for separating particles and cells include centrifugation and filtration, which are relatively easy to operate, and methods that use electric or magnetic fields that utilize external forces. However, when used for cell separation, centrifugation and filtration impart physical stimuli, which can lead to a decrease in survival rate and low separation accuracy. In addition, methods that use electric or magnetic fields may require operations such as sample modification, and require devices or equipment that apply external forces, making the operation complicated.

[0004] Meanwhile, in recent years, particle and cell separation techniques using minute flow channels (microchannels) with widths of several microns to several hundred microns have been widely researched and developed using microfabrication techniques such as photolithography and 3D printing (Non-Patent Documents 1 to 3). Since laminar flow is stably maintained within a microchannel, the fluid within the channel can be precisely controlled, and particles and cells suspended in the fluid can be separated with high accuracy.

[0005] For example, Patent Document 1 reports a technique called pinched flow fractionation (hereinafter referred to as PFF). This technique does not require complicated equipment or devices, and by introducing a fluid into a flow channel in a microchip, particles or cells can be continuously separated by size, and separation into multiple fractions is also possible. In other words, for example, when separating by size, separation can be performed not only in two stages, large and small, but also into three or more groups by size.

[0006] PFF is a separation method that focuses on the fact that a fluid is introduced from multiple branches into a structure (hereinafter referred to as a narrow section) where the flow path width is locally narrowed, the position of the particles in a direction perpendicular to the flow, and then the direction of the force that the flow exerts on the particles varies depending on the particle's position in the narrow section, and is a technology that enables continuous and precise separation. In PFF, as described in Patent Document 1, a structure with two inlets can be used, and for example, a liquid containing the target particles (hereinafter referred to as a sample liquid) is flowed from one inlet and a liquid not containing the target particles (hereinafter referred to as a sheath liquid) is flowed from the other inlet, and the particles can be separated by pressing them against the wall of the narrow section with the sheath liquid to align them.

[0007] Here, in the narrow portion of the overhead view (Figure 1(a)), if the flow direction (longitudinal direction) of the channel is the x-axis, the direction perpendicular to the interface between the sample liquid and the sheath liquid is the z-axis, and the direction perpendicular to the x-axis and z-axis is the y-axis, and the channel width in the z-axis direction is 50 and the channel width in the y-axis direction is 51, the particle separation ability in a PFF is generally determined by the channel width 50 in the z-axis direction and the flow rate ratio of the sample liquid and the sheath liquid, and the throughput is improved by the flow rates of the sample liquid and the sheath liquid, or by parallelizing the microchips. [Prior art documents] [Patent documents]

[0008] [Patent Document 1] JP 2005-205387 A [Non-patent literature]

[0009] [Non-Patent Document 1] Analytical Chemistry, 76, 5465-5471, 2004. [Non-Patent Document 2] Lab on a chip, 5, 1233-1239, 2005. [Non-Patent Document 3] Lab on a chip, 9, 939-948, 2009. [Non-Patent Document 4] Lab on a chip, 5, 778-784, 2005. Summary of the Invention [Problem to be solved by the invention]

[0010] However, when the flow rates of the sample liquid and sheath liquid are increased to improve the throughput, problems arise such as breakage of the flow path and leakage from the piping connections due to the increased pressure loss. In addition, when the material of the flow path is flexible (e.g., polydimethylsiloxane (PDMS)), the flow path width 50 in the z-axis direction of the narrow part widens due to the increased shear force caused by the increased flow rate, and there is a concern that the separation ability will decrease, so there is a limit to the increase in throughput due to the increased flow rate. On the other hand, when microchips are parallelized, the throughput increases in proportion to the number of parallels, but the material and manufacturing costs of the microchips also increase in proportion to this, which can be a barrier to industrial use.

[0011] The present invention provides a particle separation device and method that utilizes the same separation principle as PFF, and that is simpler and less expensive while improving the amount of particle separation processing per unit time. [Means for solving the problem]

[0012] In view of the above problem, the present inventors have determined that the flow channel width 51 in the y-axis direction of the narrow portion ofIt has been found that by lengthening the width 51 of the passage in the y-axis direction, the particle separation throughput is improved in approximately proportion to the width 51 of the passage in the z-axis direction of the narrow portion. of We found that by suppressing the variation in length, both the throughput and the resolution were improved. Here, the channel width 51 refers to the distance between two pairs of opposing wall surfaces of the channel that face each other at a position perpendicular to the interface of the fluid formed in the channel, and is indicated by the reference symbol 51 in Fig. 1(b) . Moreover, the channel width 50 refers to the distance between two pairs of opposing wall surfaces of the channel that face each other at a position parallel to the interface of the fluid formed in the channel, and is indicated by the reference symbol 50 in Fig. 1(a) . Furthermore, it was found that the separation ability as well as the throughput can be improved by allowing the flow ratio of the sample liquid and the sheath liquid flowing into the narrow portion to be constant within the narrow portion.

[0013] Therefore, in the above respects, the present invention relates to the following. [1] A particle separation device comprising two or more branch flow paths each having a fluid inlet at an end thereof, and a flow path formed by the joining of the branch flow paths, wherein a fluid containing particles to be separated is introduced through the fluid inlet of at least one of the branch flow paths, (1) The downstream side of the expansion start point on the terminal side of the flow path formed by the merging is expanded, and at least one recovery port is provided at the terminal of the expanded flow path; (2) The particle separation device, characterized in that, of the two sets of opposing wall surfaces of the flow path formed by the merging, the distance between the wall surfaces opposing each other in a direction perpendicular to the interface between at least two or more fluids formed in the flow path is longer than the distance between the wall surfaces opposing each other in a horizontal direction. [2] The particle separation device described above, characterized in that at least one of the branched flow paths has a branched structure. [3] The particle separation device described above, characterized in that the distance between opposing wall surfaces in a direction perpendicular to the interface is at least three times longer than the distance between opposing wall surfaces in a horizontal direction. [4] The particle separation device described above, characterized in that it has one or more pillar structures placed so as to be in contact with the wall surfaces that face in a parallel direction to the interface between at least two or more fluids formed within the flow path, out of the two sets of opposing wall surfaces of the flow path formed by the merging of the fluids. [5] The particle separation device described above, characterized in that when multiple pillar structures are placed, the spacing between the pillar structures is relatively narrow near the wall surfaces that face perpendicularly to the interface of at least two or more fluids formed in the flow path, of the two sets of opposing wall surfaces of the flow path formed by the merging, and / or one or two pillar structures are protruding structures from the wall surfaces. [6] The particle separation device described above, characterized in that the flow path formed by the merging of the fluids is a multilayer structure formed of two or more layers, and the interface between at least two or more fluids formed within the flow path is parallel to the two or more layers. [7] A particle separation method comprising the steps of: introducing a fluid containing particles to be separated suspended in one fluid inlet; introducing a fluid not containing particles in the other fluid inlet; and causing the particles to slide along the wall of a flow path where the fluids join together, the method comprising the steps of: A particle separation method, characterized in that, in a flow path where the fluids join, the length of each fluid in a direction parallel to the interface direction formed by the fluids joining is longer than the length of each fluid in a direction perpendicular to the interface direction formed by the fluids joining. Effect of the Invention

[0014] The present invention provides a particle separation device and method that utilizes the same separation principle as PFF, but is simpler, lower cost, and achieves a high particle separation throughput. This makes it possible to achieve both precise separation by particle size and a high particle separation throughput, which was previously difficult. [Brief description of the drawings]

[0015] [Figure 1] FIG. 1 shows the principle of particle separation in PFF. FIG. 1(a) is an overhead view of the flow channel, and FIG. 1(b) is a cross-sectional view of particles taken along line A-A' in FIG. 1(a). [Diagram 2]2A, 2B, and 2C are schematic perspective views of three structures including an embodiment of a particle separation apparatus and method according to an embodiment of the present invention, in which FIG. 2(a) shows a first stage structure, FIG. 2(b) shows a second stage structure, and FIG. 2(c) shows a third stage structure. [Diagram 3] Figure 3(a) is a schematic perspective view of a structure formed by inverting and stacking the second and third stage structures on the first stage structure shown in Figure 2. Figure 3(b) is a schematic diagram of an xz cross section of Figure 3(a), where the part surrounded by a thick line represents the flow channel structure. Figure 3(c) is a schematic diagram of a yz cross section of the area surrounded by a dotted line in Figure 3(b). [Figure 4] Figure 4(a) is an explanatory schematic diagram showing the behavior of the fluid introduced from the two inlets and the separation of particles in the cross-sectional view shown in Figure 3(b), and Figure 4(b) is a schematic diagram that enlarges the area surrounded by dotted lines in Figure 4(a) and shows the separation in detail. [Diagram 5] In particle separation experiments using the fabricated microchip, Figure 5(a) is a graph summarizing the particle recovery rate from each outlet when the sample liquid flow rate was 50 μL / min and the sheath liquid flow rate was 1000 μL / min (Example 1), Figure 5(b) is a graph summarizing the particle recovery rate from each outlet when the sample liquid flow rate was 150 μL / min and the sheath liquid flow rate was 3000 μL / min (Example 2), and Figure 5(c) is a graph summarizing the particle recovery rate from each outlet when the sample liquid flow rate was 250 μL / min and the sheath liquid flow rate was 5000 μL / min (Example 3). [Figure 6] 6(a) and (b) are schematic cross-sectional views of three structures comprising an embodiment of a particle separation apparatus and method according to another embodiment of the present invention, showing that the particle size exiting from the outlets 32a / b can be adjusted by switching the inlets 30a / b introducing particle-free fluid 100N and particle-containing fluid 100P. [Figure 7]7(a) and (b) are schematic cross-sectional views of two structures comprising an embodiment of a particle separation apparatus and method according to another embodiment of the present invention, showing that the particle size exiting from the outlets 32a / b can be adjusted by switching the inlets 30a / b introducing particle-free fluid 100N and particle-containing fluid 100P. [Figure 8] Fig. 8(a) is a schematic diagram of a pillar protrusion structure with an embodiment of a particle separation apparatus and method according to another embodiment of the present invention, and Fig. 8(b) is an enlarged view of region A in Fig. 8(a) from above. [Figure 9] 9 is a graph summarizing particle separation experiments (Example 4) when the pillar structure shown in FIG. 8 was modified. [Figure 10] In a particle separation experiment using a microchip with a modified bank structure, Figure 10(a) is a graph summarizing the particle recovery rate from each outlet when the sample liquid flow rate was 500 μL / min and the sheath liquid flow rate was 10,000 μL / min (Example 5), Figure 10(b) is a graph summarizing the particle recovery rate from each outlet when the sample liquid flow rate was 800 μL / min and the sheath liquid flow rate was 16,000 μL / min (Example 6), and Figure 10(c) is a graph summarizing the particle recovery rate from each outlet when the sample liquid flow rate was 1,000 μL / min and the sheath liquid flow rate was 20,000 μL / min (Example 7). [Figure 11] In a continuous separation experiment of larger diameter particles, Figure 11(a) is a graph summarizing the particle recovery rate from each outlet when the sample liquid flow rate was 250 μL / min and the sheath liquid flow rate was 5000 μL / min (Example 8), and Figure 11(b) is a graph summarizing the particle recovery rate from each outlet when the sample liquid flow rate was 400 μL / min and the sheath liquid flow rate was 8000 μL / min (Example 9). [Figure 12] FIG. 13 is a schematic diagram showing an experimental system for adjusting the fluid outflow ratio of outlets 32a, b (Example 10). [Figure 13] 13 is a graph showing the recovery rate of each particle at each outflow rate in a particle separation experiment in which the fluid outflow ratio of outlets 32a, b was adjusted (Example 10). [Figure 14] 14A and 14B are schematic perspective views of three structures including an embodiment of a particle separation device and a method thereof according to one embodiment of the present invention (FIG. 6). FIG. 14(a) shows a first-stage structure, FIG. 14(b) shows a second-stage structure, and FIG. 14(c) shows a third-stage structure. The particle separation device of FIG. 6 is constructed by inverting and stacking the second-stage and third-stage structures on the first-stage structure. [Figure 15] 15 is a graph showing the recovery rate of each particle in the particle separation experiment described in Example 11 using the particle separation device shown in FIGS. [Figure 16] 15 is a graph showing the recovery rate of each particle in the particle separation experiment described in Example 12 using the particle separation device shown in FIGS. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0016] Hereinafter, the embodiments of the present invention will be described in detail with reference to the drawings. However, the present invention can be embodied in many different forms, and is not limited to the following embodiments and examples.

[0017] 2 shows a schematic diagram of the channel structure of each stage of a three-stage microchip equipped with an embodiment of the particle separation device and method of the present invention. The x-axis direction in FIG. 2 indicates the flow (longitudinal) direction, the z-axis direction is perpendicular to the interface between the sample liquid and the sheath liquid, and the direction perpendicular to the x-axis and z-axis is shown as the y-axis. Thus, the channel width in the y-axis direction is defined by walls 35a and 35b, and the channel width (depth) in the z-axis direction corresponds to the height of pillar 19 in FIG. 2(b).

[0018] The structures formed on the flow channel substrates 10, 16, and 22 in FIG. 2 have a rectangular cross-sectional shape at any point, and the flow channel depth is uniform. As described above, the cross-sectional shape is preferably rectangular from the viewpoint of ease of fabrication of the flow channel structure, but may be circular, elliptical, polygonal, or other shapes, and may be partially other than rectangular. Similarly, the depth of the flow channel structure is preferably constant from the viewpoint of ease of fabrication, but may be partially different in depth. On the other hand, if it is desired to send the fluid downstream more accurately, it may be preferable to provide a structure with partially different depths, since it is possible to adjust the pressure loss by providing a structure with partially different depths and suppress backflow into the flow channel into which the sample liquid is introduced, or to suppress flow rate fluctuations caused by external factors such as head pressure.

[0019] The first, second and third stage flow path substrates for constructing the three-stage stacked microchip 29 shown in FIG. 2 are formed by inverting and stacking the flow path substrate 16 on the flow path substrate 10, and then inverting and stacking the flow path substrate 22 on top of the flow path substrate 10. The second stage pillar portion 19 overlaps the bank structure 13 portion of the flow path substrate 10, and the gap between the pillars is used as a narrow structure in which the flow path width is locally narrowed. The width (depth) in the z-axis direction of the narrow structure depends on the distance between the wall 34a of the bank structure 13 of the flow path substrate 10 and the wall 34b of the narrow portion of the flow path substrate 16, and corresponds to the height of the pillar structure of the second stage flow path substrate. In other words, the resolution of this device and method can be adjusted by the height of the second stage pillar. Therefore, in the narrow structure (narrow portion 31) formed in the above-mentioned three-layer stacked microchip, the interface formed by the fluids introduced from each inlet is formed parallel to the surface of each layer.

[0020] 2(a) shows a schematic diagram of the first stage of the channel structure of the three-stage stacked channel of the present invention. The branched structures shown in 12 and 14 are structures consisting of one or more branching parts for uniformly introducing fluid into the channel in the microchip (particularly into the narrow part 31). Since the more branching parts there are, the more uniformly the fluid can be introduced into the channel, it is preferable to adjust the number of branches according to the channel width. This is a measure based on the phenomenon that the flow velocity distribution is generally different near the wall surface and in the center of the channel in a channel having a structure in which at least one or more lengths in the x, y, and z axis directions in FIG. 2 are on the micro to millimeter level, and the branched structure does not necessarily have to be used as long as it achieves the purpose of uniformly introducing the fluid, and other structures may be used, for example, a mode in which one or more pillar structures are provided.

[0021] Here, since the bank structure 13 is one of the structures that constitute the narrow portion, the xy plane of the bank structure 13 is preferably flat, and this xy plane becomes the wall 34a in the z-axis direction of the flow channel of the narrow portion.

[0022] FIG. 2(b) shows a schematic diagram of the second-stage flow channel structure of a three-stage stacked flow channel. The width (depth) in the z-axis direction of the through-holes shown in 18 and 20 is preferably several hundred micrometers or less in order to keep the laminar flow in the stacked flow channel stable. For the same reason, the length in the x-axis direction of the through-holes is also preferably several hundred micrometers or less. For the convenience of fabrication, such as bonding and adhesive strength, the length in the x-axis direction of the pillar structure 19 is preferably approximately the same as the length in the x-axis direction of the bank structure 13. The walls 35a and 35b in the xz plane in FIG. 2(b) define the width of the flow channel in the y-axis direction of the narrow part.

[0023] 2(c) shows a schematic diagram of the third-stage flow channel structure of a three-stage stacked microchip used in the present invention for particle separation. The branched structures shown in 25 and 26 are structures for uniformly introducing fluid into the flow channels in the microchip, and since the more branches there are, the more uniformly the fluid can be introduced into the flow channels, it is advisable to adjust the number of branches according to the width of the flow channel. Also, as with the branched structures shown in 12 and 14, it is not necessary to use the branched structures shown above as long as the purpose of uniformly introducing the fluid is achieved, and other structures may be used.

[0024] Figure 3(a) shows a microchip 29 used in the present invention, which integrates the three-stage structure shown in Figure 2. The microchip 29 is a microchip for separating particles according to their size and collecting particles in a sample liquid by particle size by using an appropriate recovery system, and may have a flat plate-like structure formed by three microstructures 10, 16 and 22 made of a polymer material such as PDMS, or a hard material with little bending such as glass or metal.

[0025] Here, the recovery system refers to a structure or mechanism used to recover the separated particles, and includes, for example, the outlets 32a and 32b in FIG. 3(b), as well as resin or metal pipes connected to the outlets and containers for recovering the separated particles. A container for recovering the separated particles may be provided at the downstream outlet of the pipes. In this case, the distribution ratio of the fluid flowing into the two outlets may change due to the resistance of the fluid passing through the outlets and pipes, so it is preferable to appropriately adjust the inner diameter and length of the outlets and pipes.

[0026] As a technique used for producing the microchip 29, for example, a production technique using a mold such as molding or embossing is preferable because it allows accurate and easy production of a flow path structure, but other production techniques such as photolithography, soft lithography, wet etching, dry etching, nanoimprinting, laser processing, electron beam direct drawing, mechanical processing, and 3D printer production techniques can also be used. However, when producing a mold and a flow path by general photolithography, a highly accurate microfabrication technique is required to produce a structure with an aspect ratio of 3 or more (structure depth / structure width; here, the structure depth refers to the structure length in the direction of travel of ultraviolet light that exposes the resist in photolithography, and the structure length in the direction perpendicular to the travel direction) and this leads to a complicated operation for producing the flow path and an increase in costs. Therefore, it is preferable to use the above-mentioned laminated structure so that the narrow portion 31 can be formed even in a structure with a low aspect ratio.

[0027] In addition, various polymer materials such as PDMS and acrylic can be used as materials for producing the microchip 29, and it is also possible to use a combination of any two types of substrates from these materials. However, in order to produce the flow path itself at low cost and provide a disposable device, it is preferable to use a polymer material.

[0028] FIG. 3(b) shows a schematic diagram of the xz cross section of the microchip 29. 30a is an inlet consisting of 11, 17, and 23 in FIG. 2, and 30b is an inlet consisting of 24 in FIG. 2. 31 is a narrow portion in which the width (depth) of the flow path in the z-axis direction is locally reduced by the bank structure of 13 in FIG. 2, 32a is an outlet consisting of 15, 21, and 28 in FIG. 2, and 32b is an outlet consisting of 27 in FIG. 2. The inlets 30a and 30b are fluid inlets for a fluid containing particles to be measured and a fluid not containing particles to be measured, respectively, and the outlets 32a and 32b are fluid outlets. It is preferable that the fluid containing particles to be measured (or the sample liquid) is introduced from either the inlets 30a or 30b, but in the embodiment shown in FIG. 3, it is preferable that it is introduced from the inlet 30b. Furthermore, the fluid (or sheath liquid) not containing particles to be measured is preferably introduced from either inlet 30a or 30b, but in the embodiment shown in Fig. 3, it is preferably introduced from inlet 30a. In narrow section 31, an interface is formed between the introduced sample liquid and sheath liquid. The distance between walls 34a and 34b facing each other in a direction parallel to the interface (y-axis direction) corresponds to the flow path width (depth) of narrow section 31 in the z-axis direction. Furthermore, although inlets 30a and 30b are provided at the same height in Figs. 2 and 3, they may be provided at different heights.

[0029] FIG. 3(c) shows a yz cross-sectional view of the narrow portion 31. The bank structure 13 and the pillar structure 19 in FIG. 2 overlap, and the gap between them is used as a flow path of the narrow portion. The width (depth) of the flow path in the z-axis direction of the narrow portion 31 depends on the height of the pillar structure 19, so it can be set to any value, but when separating microscale particles, it is preferable that the height is tens of microns or less. In order to maintain the strength of adhesion when overlapping with the bank structure 13, the width of the pillar structure 19 in the y-axis direction is preferably several micrometers or more, and the distance between each pillar is preferably several times the pillar width. The pillar structure 19 is provided for the purpose of making the particle separation ability uniform in the narrow portion 31 by making the width of the narrow portion in the z-axis direction of FIG. 3 constant. For example, when the material constituting the microchip 29 is glass or a polymer material with low flexibility, the pillar structure 19 is not necessarily required. However, in the case of a flexible material such as PDMS, the pillar structure 19 is preferable because it is possible to suppress bending of the structure forming the narrow portion. In addition, since the flow rate generally tends to be low near the wall surface of a flow channel having a length, width, and height on the micro to millimeter level, the interval between each pillar in the pillar structure 19 near the wall surface may be relatively shortened to adjust the flow rate ratio of the sample liquid and the sheath liquid in the narrow portion 31 to be uniform in the y-axis direction of FIG. 3. In addition, based on the idea of ​​adjusting the flow rate ratio of the sample liquid and the sheath liquid to be uniform in the y-axis direction of FIG. 3, it is more preferable to locally narrow or widen the interval between each pillar in the pillar structure 19, or to connect the pillar structures near the walls 35a and 35b, or to connect them to the wall surface to form a partial protruding structure. In the narrow portion 31, an interface is formed between the introduced sample liquid and the sheath liquid. The interface is formed in a direction parallel to the bank structure 13. The distance between walls 35a and 35b facing each other in the direction perpendicular to the interface (z-axis direction) corresponds to the flow path width in the y-axis direction of narrow portion 31. The flow path width in the y-axis direction of narrow portion 31 is preferably longer than the flow path width (depth) in the z-axis direction of narrow portion 31, and can be set to be, for example, 3 times or more, 5 times or more, 10 times or more, 50 times or more, 100 times or more, 500 times or more, 1000 times or more, 1500 times or more, or 5000 times or more longer.

[0030] The total length of the flow path, i.e., the length from one end where inlet 30a is located to the other end where outlet 32a is located, can be set to any value equal to or greater than 1 μm, but from the viewpoint of ease of flow path fabrication and pressure loss, it is preferable to set it to approximately several μm to several hundred mm. From the viewpoint of achieving particle alignment, the narrow portion 31 preferably has a lower limit of length of 100 nm or more, more preferably 1 μm or more, and preferably has an upper limit of length of 10 mm or less, more preferably 500 μm or less.

[0031] Next, the separation principle of the present invention will be described in detail with reference to FIG. 4. FIG. 4(a) shows a schematic diagram of particle separation when a fluid 100N, which is a fluid not containing particles to be separated and shown with diagonal lines, is introduced from an inlet 30a of a microchip 29, and a fluid 100P, which is a fluid containing particles to be separated, is introduced from an inlet 30b. FIG. 4(b) is an enlarged view of an area 33. Particles 300a and 300b indicate relatively large particles and relatively small particles, respectively. In FIG. 4(b), arrows 400a and 400b indicate the motion vectors of the large particle 300a and the small particle 300b, respectively.

[0032] A fluid 100P containing particles and a fluid 100N not containing particles are continuously supplied from two inlets (fluid introduction ports) 30b, 30a, respectively, using a liquid delivery mechanism such as a syringe pump. At this time, in the narrow portion 33, each fluid flows while maintaining a stable laminar flow, forming an interface. Then, by adjusting the flow rates of the fluid 100P containing particles and the fluid 100N not containing particles, the depth of the fluid 100P in the narrow portion 33 in the z-axis direction is made smaller than the particle size of the smallest particle to be separated. By this operation, all particles to be separated slide along the upper surface of the narrow portion 33, and the position of the particle in the direction perpendicular to the upper surface of the narrow portion 33 along which the particle slides can be made constant depending on the size of the particle.

[0033] Then, for particles that pass through the narrow section 33, the distance between the center position of the particle and the top surface of the narrow section 33 varies depending on the particle size, so the closer the center position is to the top surface of the narrow section, the shallower the position of the flow channel in the area where the flow channel depth is expanded downstream of the narrow section 33 (upper side in Fig. 4), and the closer the center position is to the bottom surface of the narrow section, the deeper the position of the flow channel (lower side in Fig. 4). Therefore, the depth to which the particle flows out is determined by the particle size. By adjusting the flow ratio of the particles that flow out to the shallow position and the particles that flow out to the deep position from each outlet, it is possible to classify them.

[0034] The region where the flow channel depth is expanded refers to a region where the depth is greater than the narrow portion 33. More generally, it refers to a region where at least one of the expansion angles 200a, 200b with the flow channel downstream of the narrow portion is less than 180°, as shown in FIG.

[0035] Here, the phrase "particles gliding along the wall" refers to particles flowing near the wall, and does not necessarily mean that they are in contact with the wall. It also includes particles rotating or jumping near the wall. Regarding the flow rates of the sample liquid and the sheath liquid, it is preferable that the sheath liquid flow rate is larger, more preferably at least two times larger, and most preferably at least ten times larger. In addition, the separation ability can also be adjusted by flowing a large amount of fluid into one of the outlets based on the principle of "Asymmetric Pinched Flow Fractionation (AsPFF)" (Non-Patent Document 4). More specifically, this is the principle that by appropriately designing the resistance value of the flow path and by considering the relationship between the flow path width of the narrow part 33 and the particle diameter, particles of a certain size can be introduced only into a specific outlet. When the concept of AsPFF is applied to the present invention, for example, the resistance value of each flow path may be appropriately designed so that a large amount of fluid flows into the outlet 32b, and the number of outlets may be increased as appropriate. A space through which fluid 100N not including particles to be separated flows is one branch flow path, and a space through which fluid 100P including particles to be separated flows is another branch flow path, and a space where an interface between fluid 100N and fluid 100P exists formed by the joining of the two branch flow paths described above, and a flow path through which particles to be separated slide along any face or side of the space where the interface exists is the narrow portion 33. However, each of the branch flow paths described above refers to a space existing upstream of the narrow portion.

[0036] As a method for adjusting the amount of introduction to achieve the flow rate conditions inside the flow channel, it is preferable to introduce the solution from the inlet using a syringe pump or the like because it is easy to operate, but it is also possible to use a method using other pumps such as a peristaltic pump, a constant pressure liquid delivery method using a cylinder or a pressure device, a method using a liquid level difference, a liquid delivery method using electroosmotic flow or centrifugal force, etc. Also, a pressure device that applies negative pressure from the outlet can be used.

[0037] In order to achieve stable and efficient separation of particles, it is desirable to maintain a stable laminar flow within the flow path. Specifically, it is preferable to carry out the liquid transfer operation under conditions where the Reynolds number is 2300 or less, and more preferably 1000 or less.

[0038] Fluorescent polystyrene beads with diameters of 3.2 μm and 15 μm were used as the particles to be separated, but depending on the purpose, polymer particles such as polystyrene, metal particles, ceramic particles, or particles whose surfaces have been subjected to physical or chemical treatment can be used. In addition, since the force exerting external pressure or deformation on the particles is small in the present invention, it is possible to treat microparticles and cells, which would burst or be destroyed by conventional separation methods such as filters, biological particles such as organelles, viruses, bacteria, extracellular vesicles such as exosomes, proteins, and protein aggregates, which are capable of being treated. As long as the maximum diameter of the particles to be classified is less than the depth of the narrow portion 31 in the microchip 29, particles of any size can be separated. For example, when a flow channel with a narrow portion 31 having a depth of 100 μm is used, the diameter of the particles to be separated is less than 100 μm.

[0039] As the fluid in which the particles are suspended, a 0.5% (v / v) Tween 80 aqueous solution was used, but any liquid or gas can be used, and various solutions can be used depending on the purpose. For example, when polymer particles or metal particles are used as the particles, in addition to aqueous solutions containing various chemical substances, organic solutions, ionic fluids, etc. can be used. Furthermore, when biological particles such as cells are used as the particles, it is preferable to use an aqueous solution that is isotonic with the cells, such as a cell culture solution or a physiological buffer solution. However, in the case of cells that are resistant to relatively hypotonic or hypertonic solutions, such as bacteria or plant cells, it is not necessarily necessary to be isotonic. Furthermore, for convenience of operation, a system in which the difference between the density of the solution and the density of the particles is not large is more preferable. Regarding the viscosities of the sample liquid and the sheath liquid, a system in which the difference in these viscosities is small is more preferable, but as long as the system allows the processing of particles, there may be a difference in viscosity. Note that both the fluid containing particles and the fluid not containing particles may be composed of the same fluid, or each may be composed of two or more different fluids. The densities of the sample liquid and the sheath liquid may be equal or different. From the viewpoint of forming a laminar flow in which the sample liquid and the sheath liquid are stacked vertically, the density of the liquid forming the lower laminar flow can be made heavier.

[0040] In addition, as shown in FIG. 6 and FIG. 14, the present invention may use a structure that forms the narrow portion 31 without using the bank structure 13. The first-stage flow path substrate 10 is a flat substrate without a flow path structure. The second-stage flow path substrate 16 has a wall 34b that forms the narrow portion, and forms the flow path of the narrow portion 31 together with the surface (wall 34a) of the first-stage flow path substrate 10, and gives the narrow portion 31 a flow path width (depth) in the z-axis direction corresponding to the thickness of the pillar structure 19. In this case, a material such as glass having a flatter surface can be used for the wall surface against which the particles are pressed based on the principle of PFF. The second-stage flow path substrate 16 has through structures 18 and 20, and communicates with the flow path formed in the third-stage flow path substrate 22 via the through structures 18 and 20. The second-stage flow path substrate has branch structures 12 and 14, and uniformly introduces a fluid into the flow path formed by the second-stage flow path substrate 16 and the first-stage flow path substrate 10. The second-stage flow path substrate 16 has a portion 17 of the inlet 30a and a portion 21 of the outlet 30a that communicate with the branching structures 12 and 14, which penetrate the second-stage flow path substrate 16 and form the inlet 30a and the outlet 32a, respectively, together with a portion 23 of the inlet 30a and a portion 28 of the outlet 32a of the third-stage flow path substrate 22. The third-stage flow path substrate 22 has branching structures 25 and 26, and uniformly introduces fluid into the flow path formed by the third-stage flow path substrate 22 and the second-stage flow path substrate 16. The third-stage flow path substrate 22 has a portion 24 of the inlet 30b and a portion 27 of the outlet 32b that communicate with the branching structures 25 and 26. Furthermore, by selecting the inlets 30a / b for introducing the fluid 100N not containing particles and the fluid 100P containing particles, it is possible to select which of the z-axis walls 34a / b of the flow passage in the narrow section the particles will slide along, thereby adjusting the particle size flowing out of the outlets 32a / b. Here, the fluid (or sample liquid) containing the particles to be measured is preferably introduced from either inlet 30a or 30b. In the embodiment shown in Figures 6 and 14, if wall surface 34a / b is formed from a material having a flatter surface, it is preferable to introduce it from inlet 30a, and if wall surface 34b is formed from a material having a flatter surface, it is preferable to introduce it from inlet 30b.

[0041] Furthermore, as shown in Fig. 7, the present invention may use an embodiment in which a two-tiered microchip is formed without using the branched structures 12, 14, 25, 26 and the through structures 18, 20. By selecting the inlets 30a / b for introducing the particle-free fluid 100N and the particle-containing fluid 100P, it is possible to select which of the walls 34a / b in the z-axis direction of the narrow passage of the flow path the particles are caused to slide along, thereby adjusting the particle size flowing out from the outlets 32a / b. Here, it is preferable that the fluid (or sample liquid) containing the particles to be measured is introduced from either the inlets 30a or 30. In the embodiment shown in Fig. 7, when the wall surface 34a / b is made of a material having a flat surface, it is preferable that the fluid is introduced from the inlet 30a, and when the wall surface 34b is made of a material having a flat surface, it is preferable that the fluid is introduced from the inlet 30b.

[0042] The three-layer stacked microchip described above was fabricated by fabricating three layered flow channel substrates (10, 16, 22) based on a general photolithography technique and bonding the surfaces of the layers together, but the structures shown in FIG. 3(a) and FIG. 6 may be fabricated in one go using 3D printing technology. Therefore, the present invention does not need to be fabricated by stacking multiple layers, which is preferable in terms of reducing costs such as labor costs involved in the flow channel manufacturing process. On the other hand, in terms of reducing the cost of materials through mass production, it is preferable to form a multilayer structure of two or more layers using the above-mentioned photolithography, metal processing, soft lithography using a metal mold fabricated by microfabrication technology, and injection molding. EXAMPLES

[0043] Manufacturing Example: Fabrication of a Layered Channel The microstructure of each stage of the three-stage microchip 25 equipped with the embodiment of the continuous classification method according to the present invention was fabricated using general photolithography and soft lithography techniques. The specific procedure is as follows:

[0044] Photoresist SU-8 3000 series (Microchem) was dropped onto a 4-inch bare silicon wafer (Filtech Co., Ltd.), and a thin photoresist film was formed using a spin coater (MIKASA Co., Ltd.). A channel pattern was formed on the photoresist film using a mask aligner (Ushio Inc.) and a film mask with an arbitrary pattern, and the channel pattern was developed using SU-8Developer (Microchem Co., Ltd.) to create a mold for the channel to be used. Next, a mixture of uncured siloxane monomer and polymerization initiator (weight ratio 10:1) prepared using SILPOT184 (Dow Corning Toray Co., Ltd.) was poured into the mold and heated at 85°C for 30 minutes to produce polydimethylsiloxane (PDMS) with the shape of the flow channel transferred to it. The cured PDMS was carefully peeled off from the mold and molded to the desired size with a cutter, and then a port on the inlet side of the flow channel and an outlet were formed using a puncher. The surface of each peeled PDMS was treated with an oxygen plasma generator (Meiwa Force Co., Ltd.), and the alignment was adjusted and the PDMS of each stage was bonded together to produce a microchip29. The structure of the microchip produced is as shown in Figures 2 and 3.

[0045] The size of each structure of the microchip 29 used is expressed as length in the x-axis direction, width in the y-axis direction, and depth in the z-axis direction in Fig. 2. The bank structure 13 is 300 μm long, 1 cm wide, and 300 μm deep, the through structures 18 and 20 are both 300 μm long, 1 cm wide, and 300 μm deep, the branch structures 12 and 14 are 500 μm wide and 320 μm deep channels, and the branch structures 25 and 26 are 500 μm wide and 80 μm deep channels branched into two, with each structure repeated four times, and finally branched into 16 channels. Furthermore, the pillar structure 19 is 50 μm long, 50 μm wide, 20 μm deep, and the wall distance between the pillars is 75 μm.

[0046] Continuous particle separation experiment Example 1 Fluid 100P containing particles was prepared by diluting 3.2 μm diameter fluorescent polystyrene (Fluoro-Max; Thermo Scientific) to 5 μg / mL and 15 μm diameter fluorescent polystyrene-divinylbenzene particles (Fluoro-Max; Thermo Scientific) to 200 μg / mL, using an aqueous solution of 18 wt% sucrose and 0.5 wt% Tween 80. Fluid 100N not containing particles was prepared by diluting 18 wt% sucrose and 0.5 wt% Tween 80.

[0047] Using the microchip 29 (Figs. 2 and 3) produced in the manufacturing example, fluid 100N was delivered from inlet 30a, and fluid 100P was delivered from inlet 30b at flow rates of 1000 and 50 μL / min, respectively, by adjusting the flow rates with a syringe pump. Recovery tubes were set in outlets 32a and 32b, and the outflowing liquid was recovered in a microtube. The amount of outflowing liquid was measured from the difference between the tare weight and the mass after recovery. Then, the fluorescent particle concentration in the two recovered liquids was measured using a hemocytometer. The recovery rate of particles of each size that flowed out from each outlet was calculated from the amount and concentration of the outflowing liquid. The separation experiment was performed three times, and the numbers in the graph represent the average data.

[0048] Figure 5(a) is a graph showing the particle recovery rate from each outlet. The horizontal axis represents outlets 32a and 32b, and the vertical axis represents the recovery rate of each particle. More than 95% of the 15 μm particles flowed out from outlet 32a, and more than 95% of the 3.2 μm particles flowed out from outlet 32b, achieving high-precision separation.

[0049] Here, in the particle separation experiment described in Non-Patent Document 4, the length in the direction perpendicular to the interface between the sample liquid and the sheath liquid was 20 μm, and the processing amount of the sample liquid under roughly the same conditions as in this embodiment was 20 μL / hour, which shows that the sample processing amount of the present invention is 150 times higher than that of the conventional PFF.

[0050] Example 2 A particle separation experiment was carried out in the same manner as in Example 1, except that the flow rates of fluid 100N and fluid 100P were changed to 3000 and 150 μL / min, respectively. A graph showing the recovery rate is shown in Figure 5(b). More than 90% of the 15 μm particles flowed out from outlet 32a, and more than 90% of the 3.2 μm particles flowed out from outlet 32b, respectively. Compared to Example 1, high-precision separation was achieved even under conditions where the total flow rate was three times higher.

[0051] Example 3 A particle separation experiment was carried out in the same manner as in Example 1, except that the flow rates of fluid 100N and fluid 100P were changed to 5000 and 250 μL / min, respectively. A graph showing the recovery rate is shown in Figure 5(c). More than 95% of the 15 μm particles flowed out from outlet 32a, and more than 90% of the 3.2 μm particles flowed out from outlet 32b, respectively. Compared to Example 1, high-precision separation was achieved even under conditions where the total flow rate was five times higher.

[0052] Pillar structure change Example 4 A particle separation experiment was carried out in the same manner as in Example 3, except that a pillar protruding structure 61 with a protruding length of 200 μm from the walls 35a, b of the bank structure 13 was connected to the wall surface (see FIG. 8). A graph showing the recovery rate is shown in FIG. 9. More than 97% of the 15 μm particles flowed out from the outlet 32a, and more than 98% of the 3.2 μm particles flowed out from the outlet 32b, respectively, achieving a higher degree of separation accuracy compared to Example 3.

[0053] Change in width of embankment structure Example 5 A particle separation experiment was carried out in the same manner as in Example 1, except that the width of the bank structure 13 was changed from 1 cm to 3 cm, and the flow rates of the fluids 100N and 100P were changed to 10,000 and 500 μL / min, respectively. A graph showing the recovery rate is shown in FIG. 10(a). More than 88% of the 15 μm particles flowed out from the outlet 32a, and more than 90% of the 3.2 μm particles flowed out from the outlet 32b, respectively. Compared to Example 1, high-precision separation was achieved even under conditions where the total flow rate was 10 times higher.

[0054] Example 6 A particle separation experiment was carried out in the same manner as in Example 5, except that the flow rates of fluid 100N and fluid 100P were changed to 16,000 and 800 μL / min, respectively. A graph showing the recovery rate is shown in Figure 10(b). More than 90% of the 15 μm particles flowed out from outlet 32a, and more than 92% of the 3.2 μm particles flowed out from outlet 32b, respectively. Compared to Example 1, high-precision separation was achieved even under conditions where the total flow rate was 16 times higher.

[0055] Example 7 A particle separation experiment was carried out in the same manner as in Example 5, except that the flow rates of fluid 100N and fluid 100P were changed to 20,000 and 1,000 μL / min, respectively. A graph showing the recovery rate is shown in Figure 10(c). More than 84% of the 15 μm particles flowed out from outlet 32a, and more than 90% of the 3.2 μm particles flowed out from outlet 32b, respectively. Compared to Example 1, high-precision separation was achieved even under conditions where the total flow rate was 20 times higher.

[0056] Continuous separation experiment of larger diameter particles Example 8 The length of the pillar structure 19 was changed to 150 μm, the width to 75 μm, the depth to 70 μm, the protruding length from the wall of the pillar protruding structure to 555 μm, the fluid 100P containing the particles was diluted with 18 wt% sucrose and 0.5 wt% Tween 80 aqueous solution to 5 μg / mL of fluorescent polystyrene (Fluoro-Max; Thermo Scientific) with a diameter of 15 μm and 200 μg / mL of fluorescent polystyrene-divinylbenzene particles (Fluoro-Max; Thermo Scientific) with a diameter of 50 μm, and the flow rates of the fluids 100N and 100P were changed to 5000 and 250 μL / min, respectively. A particle separation experiment was performed in the same manner as in Example 1. A graph showing the recovery rate is shown in FIG. 11(a). More than 96% of the 50 μm particles flowed out from outlet 32a, and more than 94% of the 15 μm particles flowed out from outlet 32b. Compared to Example 1, highly accurate separation was achieved even for particles with diameters about three times larger.

[0057] Example 9 A particle separation experiment was carried out in the same manner as in Example 1, except that the flow rates of fluid 100N and fluid 100P were changed to 8000 and 400 μL / min, respectively. A graph showing the recovery rate is shown in Figure 11(b). More than 97% of the 50 μm particles flowed out from outlet 32a, and more than 92% of the 15 μm particles flowed out from outlet 32b, respectively. Compared to Example 8, high-precision separation was achieved even under conditions where the total flow rate was 1.6 times higher.

[0058] Adjustment of the fluid outflow ratio of the outlets 32a,b Example 10 A 200 mm length of silicon tubing with an outer diameter of 3 mm and an inner diameter of 2 mm was connected to the outlet 32a via an L-shaped metal pipe, and 10 mm from the end of the silicon tubing was fixed to a jack with tape. The height of the jack was adjusted to adjust the fluid outflow ratio of the outlets 32a and b using the potential energy of the silicon tube (FIG. 12). Fluid 100P containing particles was diluted with an 18 wt% sucrose, 0.5 wt% Tween 80 aqueous solution to 20 μg / mL of fluorescent polystyrene (Fluoro-Max; Thermo Scientific) with a diameter of 3.2 μm and 300 μg / mL of fluorescent polystyrene-divinylbenzene particles (Fluoro-Max; Thermo Scientific) with a diameter of 9.9 μm, respectively. A particle separation experiment was performed in the same manner as in Example 1, except that the flow rates of fluid 100N and fluid 100P were changed to 2000 and 100 μL / min, respectively (FIG. 13, Table 1). The jack height was adjusted based on the reference (0 cm) being the surface where the bottom of the device was in contact with the experimental bench (the z-axis origin side in Figure 2(a)). As a result of the verification, it was found that the particle separation performance was best (jack height 2 cm) when the proportion of fluid flowing to recovery port 1 (recovery port 1 outflow rate) was 8.4%, and that high-precision separation was possible by adjusting the outflow rate from the above.

[0059] [Table 1]

[0060] Particle separation using structures without bank structures Example 11 The microchip shown in FIG. 6 was produced by the method described in the manufacturing example. A more detailed structure is shown in FIG. 14. In each structure, the through structures 18 and 20 are both 300 μm long, 1 cm wide, and 360 μm deep, the branched structures 12 and 14 are each a structure in which a channel with a width of 100 μm and a depth of 40 μm is branched into two, which is repeated five times, and finally branched into 32 channels, and the branched structures 25 and 26 are each a structure in which a channel with a width of 100 μm and a depth of 160 μm is branched into two, which is repeated four times, and finally branched into 16 channels. Furthermore, the pillar structure 19 is 100 μm long, 50 μm wide, and 20 μm deep, the wall distance between the pillars is 75 μm, and the pillar protruding structure is 50 μm.

[0061] Fluid 100N without particles was prepared by diluting sugar solution OptiPrep (manufactured by Cosmo Bio) with pure water to a specific gravity of 1.05 g / mL, and preparing an aqueous solution containing 0.05 v / v% Tween 20. Fluid 100P with particles was prepared by diluting fluorescent polystyrene particles (manufactured by Polyscience) with a diameter of 2.2 μm to 1×10^7 particles / mL and fluorescent polystyrene particles (manufactured by Polyscience) with a diameter of 10.2 μm to 1×10^6 particles / mL, respectively, with fluid 100N without particles.

[0062] Using the microchip (Figure 6) fabricated by the above method, fluid 100N was fed from inlet 30b with the flow rate adjusted by a pressure pump (Dolomite), and fluid 100P was fed from inlet 30a with the flow rate adjusted by a syringe pump at flow rates of 2000 and 100 μL / min, respectively. Recovery tubes were set at outlets 32a and 32b, and the outflowing liquid was collected in a microtube. The outflowing liquid volume was measured from the difference between the tare weight and the mass after collection. After that, the fluorescent particle concentration in the two collected liquids was measured using a hemocytometer. The recovery rate of particles of each size flowing out from each outlet was calculated from the outflowing liquid volume and concentration. A graph showing the recovery rate is shown in Figure 15. It was confirmed that particle separation was possible, with 99% or more of the 10.2 µm particles flowing out from outlet 32b and 88% or more of the 2.2 µm particles flowing out from outlet 32a.

[0063] Example 12 The microchip used was the same as that used in Example 11, except that the branched structures 12 and 14 had a depth of 20 μm, the branched structures 25 and 26 had a depth of 80 μm, the through structures 18 and 20 were both 530 μm long, 1 cm wide, and 360 μm deep, the pillar structure 19 had a length of 700 μm, a width of 20 μm, and a depth of 5 μm, the wall distance between the pillars was 40 μm, and the pillar protruding structure was 175 μm. Fluid 100P containing particles was made of fluorescent polystyrene particles (Polystyrene) having a diameter of 0.5 μm. The particle separation experiment was carried out in the same manner as in Example 11 except that the fluids 100N and 100P were fed at flow rates of 1500 and 50 μL / min, respectively, and the concentration of the collected particles was calculated by the method described in International Application No. PCT / JP2018 / 004900. FIG. 16 shows a graph showing the recovery rate. More than 97% of the 2 μm particles flowed out from the outlet 32b, and more than 91% of the 0.5 μm particles flowed out from the outlet 32a, respectively, and high-precision separation was achieved even for particles with smaller diameters compared to Example 11. [Explanation of symbols]

[0064] 10 1st stage channel board 11 Part of inlet 30a 12 Branched Structure 13 Embankment structure 14 Branched Structure 15 Part of Outlet 32a 16 2nd stage channel board 17 Part of inlet 30a 18 Penetration structure 19 Pillar structure 20 Penetration structure 21 Part of Outlet 32a 22 3rd stage channel board 23 Part of Inlet 30a 24 Part of inlet 30b 25 Branched Structure 26 Branched Structure 27 Part of Outlet 32b 28 Part of Outlet 32a 29 Three-layer stacked microchip 30a, 30b inlet 31 Narrow area 32a Outlet (Collection Port 2) 32b Outlet (Collection Port 1) 33 Part of Figure 4(b) 34a, b Parallel to the interface At the position Opposing Wall 35a, b Perpendicular to the interface At the position Opposing Wall 50 Width of the narrow passage in the z-axis direction 51 Channel width in the y-axis direction of the narrow section 60 Area A 61 Pillar protruding structure 62 Overhang length 100N particle-free fluid Fluid containing 100P particles 200a, 200b Enlarged angle 300a, 300b particles 400a, 400b Motion vectors of particle 300a and particle 300b

Claims

1. Fluid at end 1 A particle separation device comprising: two or more branch flow paths each having an inlet; and a flow path formed by joining the branch flow paths; wherein a fluid containing particles to be separated is introduced through a fluid inlet of at least one of the branch flow paths, (1) The downstream side of the expansion start point on the terminal side of the flow path formed by the merging expands in the depth direction, and at least one recovery port is provided at the terminal of the expanded flow path; (2) Of the two pairs of opposing wall surfaces of the flow path formed by the merging of the sample liquids, the distance between the opposing wall surfaces at a position perpendicular to an interface between the sample liquid and the sheath liquid formed in the flow path is longer than the distance between the opposing wall surfaces at a position parallel to the interface, The particle separation device is a multi-layer structure formed of two or more layers, and the interface is parallel to the two or more layers; one or more pillar structures are placed in contact with a wall surface facing the interface between the sample liquid and the sheath liquid formed in the flow channel in a parallel position; 3. The particle separation device according to claim 1, wherein the pillar structure is a protruding structure extending from a wall surface facing the interface between the sample liquid and the sheath liquid in a position parallel to the interface.

2. 2. The particle separator according to claim 1, wherein at least one of said branched flow paths has a branched structure.

3. 3. The particle separator according to claim 1, wherein the distance between opposing wall surfaces in a position perpendicular to the interface is at least three times longer than the distance between opposing wall surfaces in a position parallel to the interface.

4. A particle separation device as described in any one of claims 1 to 3, characterized in that, of the two sets of opposing wall surfaces of the flow path formed by the merging of the liquids, the spacing between the pillar structures is relatively narrow near the opposing wall surfaces that are perpendicular to the interface between the sample liquid and sheath liquid formed in the flow path.

5. A particle separation method using a particle separation device according to any one of claims 1 to 4, comprising introducing a sample liquid having particles to be separated suspended therein from one fluid inlet and introducing a sheath liquid not containing particles from the other inlet, and causing the particles to slide along a wall of a flow path where the sample liquid and the sheath liquid join together, The end side of the flow channel formed by the merging expands in the depth direction downstream from the expansion start point, a flow path in which the sample liquid and the sheath liquid join together, the width of each of the flows of the sample liquid and the sheath liquid in a direction parallel to an interface direction formed by the joining of the sample liquid and the sheath liquid is longer than the width of each of the flows of the sample liquid and the sheath liquid in a direction perpendicular to the interface direction formed by the joining of the sample liquid and the sheath liquid.

Images