Sponge roller for substrate cleaning

Abstract

To provide a sponge roller for substrate cleaning that improves cleaning performance, minimizes the number of particles remaining on the substrate surface, and has excellent manufacturability.SOLUTION: In a cylindrical sponge roller 1A for substrate cleaning, formed of a resin sponge having a continuous porous structure and hydrophilic properties, with a cleaning surface 10 on the outer circumference and a shaft insertion hole 20 on the inner circumference, the cleaning surface 10 has a plurality of grooves 11 that are straight when developed in a flat plane and extend in at least two directions, and the grooves 11 are spaces with both side walls 12 perpendicular to the cleaning surface 10 and of a predetermined depth.SELECTED DRAWING: Figure 4

Description

Technical Field

[0001] The present invention relates to a sponge roller for substrate cleaning used in a cleaning process of a substrate, such as a silicon wafer or a glass substrate, which requires a high degree of surface flatness.

Background Art

[0002] Conventionally, in a semiconductor manufacturing process, in order to flatten the surface of a silicon wafer as a substrate, a polishing method called CMP (Chemical Mechanical Polishing), which uses a combined action of chemical action and mechanical polishing, is known.

[0003] In this CMP, the surface of a silicon wafer is polished using a slurry containing an abrasive and a grindstone. After the polishing is completed, various particles such as abrasive grains, cutting chips, and polishing chips remain on the surface of the silicon wafer, so it is necessary to clean them.

[0004] Therefore, a sponge roller for substrate cleaning for cleaning the surface of a silicon wafer after polishing is formed by forming a resin sponge into a cylindrical shape, pressing the outer peripheral side thereof against the polished silicon wafer as a cleaning surface and rotating it relatively, and flowing a cleaning liquid over it. For example, it is described in Japanese Patent Application Laid-Open No. 98 / 020987 (Patent Document 1), Japanese Patent Application Laid-Open No. 2015-089531 (Patent Document 2), Japanese Patent Application Laid-Open No. 2017-535418 (Patent Document 3), etc.

[0005] These conventionally known sponge rollers for substrate cleaning are those in which convex cylindrical protrusions are regularly arranged as shown in Patent Documents 1 and 2, or those in which grooves are formed as shown in Patent Document 3. Thus, instead of simply making the cleaning surface on the outer peripheral side a smooth surface, efforts have been made to improve the cleaning efficiency and cleaning performance by forming protrusions and grooves.

[0006] However, it is difficult to claim that the cleaning performance of the conventionally known sponge roller for substrate cleaning is still sufficient, and particles may remain on the surface of the silicon wafer after cleaning. In the extremely miniaturized semiconductor process rules in recent years, the remaining of even a small number of particles has been an issue to be resolved.

Prior Art Documents

Patent Documents

[0007]

Patent Document 1

Patent Document 2

Patent Document 3

Summary of the Invention

Problems to be Solved by the Invention

[0008] An object of the present invention is to provide a sponge roller for substrate cleaning that has better cleaning performance than conventional sponge rollers for substrate cleaning, can minimize the number of particles remaining on the substrate surface as much as possible, and is excellent in manufacturability.

Means for Solving the Problems

[0009] The present invention made to solve the above problems is a cylindrical sponge roller for substrate cleaning formed of a resin sponge having an open-cell structure and hydrophilicity, with the outer peripheral side as the cleaning surface and the inner peripheral side as a shaft insertion hole. The cleaning surface has a plurality of grooves that are linear when developed in a plane and extend in at least two directions, and the grooves are spaces having side walls perpendicular to the cleaning surface and a predetermined depth.

[0010] According to the present invention, by holding and guiding polishing liquid, cleaning liquid, etc. through the grooves and discharging them to the outside, good cleaning results can be obtained. And by forming linear grooves, it is excellent in productivity when manufacturing, for example, by cutting using a lathe or the like.

[0011] In the present invention, when the groove is a space having both side walls perpendicular to the cleaning surface and a predetermined depth, at the right-angle edge portion between the cleaning surface and the groove, an edge effect of scraping polishing liquid, cleaning liquid, etc. like a scraper can be exerted.

[0012] In the present invention, when the groove is composed of an oblique groove intersecting the axial direction and a reverse oblique groove intersecting the oblique groove in the opposite direction to the axial direction, according to the relative rotation with the substrate, each groove can hold and guide cleaning liquid, foreign matter, etc. while making it easy to discharge them to the outside.

[0013] In the present invention, when the groove is composed of parallel grooves parallel to the axial direction and intersecting grooves intersecting the axial direction, according to the relative rotation with the substrate, each groove can hold and guide cleaning liquid, foreign matter, etc. while making it easy to discharge them to the outside. At the same time, the parallel grooves parallel to the axial direction have a structure that is easy to process by cutting.

[0014] In the present invention, when the resin sponge is a polyvinyl alcohol sponge (PVA sponge), it has very high hydrophilicity and excellent water absorption and water retention ability, so it is particularly desirable.

Effects of the Invention

[0015] According to the present invention, it is possible to provide a sponge roller for substrate cleaning having excellent cleaning performance and productivity.

Brief Description of the Drawings

[0016]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Embodiments for Carrying Out the Invention

[0017] Hereinafter, embodiments of the present invention will be described based on the drawings.

[0018] FIG. 1 is a perspective view showing a sponge roller 1A for substrate cleaning according to the first embodiment of the present invention, FIG. 2 is a front view of the sponge roller 1A for substrate cleaning, FIG. 3 is a view showing a state where the cleaning surface of the sponge roller 1A for substrate cleaning is developed in a plane, and FIG. 4 is a partially enlarged cross-sectional view of the sponge roller 1A for substrate cleaning.

[0019] The sponge roller 1A for substrate cleaning has a cylindrical shape with a cleaning surface 10 on the outer peripheral side and a shaft insertion hole 20 on the inner peripheral side, and is entirely formed of a resin sponge having a continuous pore structure in which continuous pores are exposed on the cleaning surface 10 and having hydrophilicity. For example, in a cleaning step after a polishing step such as CMP for polishing a substrate such as a silicon wafer or a glass substrate, a cleaning liquid is poured while pressing the cleaning surface against the surface of the substrate to be polished, and the sponge roller is relatively rotated to clean slurry, polishing debris, etc. remaining on the surface of the substrate.

[0020] The dimensions of the sponge roller 1A for substrate cleaning are in a long shape with a length of about 700 mm, an outer diameter of φ50 mm, and an inner diameter of φ26 mm.

[0021] Regarding the material of the sponge roller 1A for substrate cleaning, more specifically, polyvinyl alcohol sponge (PVA sponge) is preferable. Also, the pore diameter is preferably about 180 to 180 μm. PVA sponge is a resin sponge having hydrophilicity with a three-dimensional network structure having continuous pores, rich in water absorption and water retention, and excellent in chemical resistance and abrasion resistance. Therefore, it is a very suitable material when used as a resin sponge for substrate cleaning. However, for example, other polyvinyl resin continuous pore bodies or other resin sponges other than PVA sponge may be used.

[0022] The cleaning surface 10 has a plurality of grooves 11 that are linear when developed in a plane and extend in two directions. The grooves 11 include an oblique groove 11a that intersects the axial direction and a reverse oblique groove 11b that intersects the axial direction in the opposite direction to the oblique groove 11a (see FIG. 3). And the grooves 11 are spaces where both side walls 12 are perpendicular to the cleaning surface 10 and have a predetermined depth (see FIG. 4).

[0023] In this way, by adopting a diamond-shaped groove pattern formed by intersecting groove angles in two directions, it is possible to form a structure that guides the cleaning liquid, foreign matter, etc. while holding them in each groove according to the relative rotation with the substrate and facilitates discharge to the outside.

[0024] Moreover, by making the edge angle of the groove 11 a right angle, an edge effect of scraping the polishing liquid, cleaning liquid, etc. like a scraper can be exerted at the edge portion.

[0025] Then, by guiding while holding the cleaning liquid, foreign matter, etc. in the space of the groove 11 and discharging them to the outside, good cleaning results can be obtained. Also, by making the groove a linear groove, it is excellent in manufacturability when manufacturing by cutting using, for example, a milling machine.

[0026] In addition, when the groove 11 is formed at an obtuse angle with respect to the cleaning surface 10, the above-described edge effect is small and the volume of the space of the groove 11 decreases, so the amount of cleaning liquid that can be held decreases. On the contrary, when the groove 11 is formed at an acute angle with respect to the cleaning surface 10, although the above-described edge effect increases, a flow of the cleaning liquid toward the depth of the groove 11 is generated, so the update of the cleaning liquid is not smooth and it becomes more difficult to process because it becomes a dovetail groove shape.

[0027] Therefore, the case where the edge angle between the cleaning surface 10 and the groove 11 is a right angle as in the present embodiment contributes to the improvement of cleaning performance and is particularly suitable because cutting processing also becomes easy. However, this edge angle is not strictly limited to a right angle and may be angled in the obtuse angle direction or the acute angle direction.

[0028] The shaft insertion hole 20 may be a simple circular hole or a hole formed with a notch for shaft engagement (not shown).

[0029] The steps for manufacturing the substrate cleaning roller 1A are as follows A to C. A. Using a mold, a long cylindrical resin sponge is molded. B. Rotate the resin sponge and cut the entire outer circumference to remove the skin layer remaining during molding and achieve a predetermined outer diameter. C. While rotating the resin sponge as needed, cut a groove in the outer circumference to form a cleaning surface.

[0030] The above steps are particularly desirable because by processing in a wet state in which the resin sponge is impregnated with a liquid, it is possible to mold it according to the dimensions when actually used in the cleaning process. As the liquid to be impregnated, water or the cleaning liquid to be used can be used.

[0031] During the cutting process in step B, for example, lathe processing can be used, and during the cutting process in step C, for example, milling processing can be used.

[0032] In this way, by cutting the entire outer circumference of the long cylindrical resin sponge, it is possible to easily expose the continuous pores on the surface of the cleaning surface from the state where the texture of the mold surface is transferred and the continuous bubbles are buried when molded into a long cylindrical shape.

[0033] And the formation of the groove 11 can be continuously performed by cutting regardless of the length of the long cylindrical resin sponge, and it can also be easily cut into a specified length from the state of being formed in a long cylindrical shape to obtain a product, so the manufacturability is extremely excellent.

[0034] FIG. 5 is a perspective view showing a sponge roller 2A for substrate cleaning according to the second embodiment of the present invention, FIG. 6 is a front view of the sponge roller 2A for substrate cleaning, FIG. 7 is a view showing the cleaning surface of the sponge roller 2A for substrate cleaning in a state of being developed on a plane, and FIG. 8 is a partially enlarged cross-sectional view of the sponge roller 2A for substrate cleaning.

[0035] This sponge roller 2A for substrate cleaning has a cylindrical shape with a cleaning surface 10 on the outer peripheral side and a shaft insertion hole 20 on the inner peripheral side, and is entirely formed of a resin sponge having a continuous pore structure in which continuous pores are exposed on the cleaning surface 10 and having hydrophilicity, which is the same as the sponge roller 1A for substrate cleaning in this regard, but the groove shape is different.

[0036] In the sponge roller 2A for substrate cleaning, the cleaning surface 10 has a plurality of grooves 11 that are linear when developed in a plane and extend in two directions. The grooves 11 are composed of parallel grooves 11c parallel to the axial direction and intersecting grooves 11d intersecting the axial direction (see Fig. 7). And the grooves 11 are spaces where both side walls 12 are perpendicular to the cleaning surface 10 and have a predetermined depth (see Fig. 8).

[0037] In this way, by forming a groove pattern in the shape of a parallelogram mesh with intersecting groove angles in two directions, it is possible to form a structure in which the cleaning liquid, foreign matter, etc. are held and guided by each groove according to the relative rotation with the substrate and are easily discharged to the outside.

[0038] And, similar to the sponge roller 1A for substrate cleaning, the formation of the grooves 11 can be continuously performed by cutting regardless of the length of the long cylindrical resin sponge, and it can be easily cut into a specified length from the state of being formed in a long cylindrical shape to obtain a product. Furthermore, since it is not necessary to rotate the resin sponge during cutting for the parallel grooves 11c parallel to the axial direction, it has a structure that is easy to process by cutting.

[0039] Regarding the dimensions (outer diameter, inner diameter, length, groove angle, etc.) of the sponge roller for substrate cleaning of the present invention, it can be appropriately adjusted and manufactured according to the device to be used and the purpose of use. For example, when it is desired to increase the contact area, the outer diameter can be increased, or the length can be increased according to the size of the substrate. Conversely, the outer circumference or the length can be decreased.

[0040] Also, the groove shape of the sponge roller for substrate cleaning of the present invention is not limited to the disclosed one. For example, those having grooves in three directions may be used (not shown).

[0041] Figure 9 shows a third embodiment of the sponge roller for substrate cleaning according to the present invention. This sponge roller 1B for substrate cleaning has the same diamond-shaped mesh groove shape as the sponge roller 1A for substrate cleaning.

[0042] Figure 10 is a diagram showing a fourth embodiment of the sponge roller for substrate cleaning according to the present invention. This sponge roller 2B for substrate cleaning has the same parallelogram-shaped mesh groove shape as the sponge roller 2A for substrate cleaning.

[0043] The dimensions of the sponge rollers 1B and 2B for substrate cleaning are approximately 150 mm in length, 60 mm in outer diameter, and 32 mm in inner diameter, and they are shorter and larger in diameter compared to the sponge rollers 1A and 2A for substrate cleaning.

Example

[0044] Hereinafter, in order to verify the cleaning ability of the sponge roller for substrate cleaning of the present invention, four kinds of experiments were conducted, and the content and results are described below.

[0045] Figure 11 is a schematic diagram showing a polishing apparatus 2 used as an experimental apparatus. This polishing apparatus 2 has a rotatable surface plate 3 and a rotatable shaft 4 provided above the surface plate 3. An object to be polished is set on the surface plate 3, a sponge roller for substrate cleaning is attached to the shaft 4, and by rotating one or both of the surface plate 3 and the shaft 4, the object to be polished and the sponge roller for substrate cleaning are relatively rotated to perform a cleaning operation. In this experiment, the rotation directions of both the surface plate 3 and the shaft 4 are clockwise. Note that the vertical distance between the surface plate 3 and the shaft 4 is configured to be adjustable.

[0046] FIG. 12 is a diagram showing each sample used by setting on the polishing apparatus 2, (a) a conventional sponge roller 1S for cleaning a substrate with protrusions (hereinafter, "comparative example"), (b) a sponge roller 1B for cleaning a substrate of the present invention (hereinafter, "Example 1"), and (c) a sponge roller 2B for cleaning a substrate of the present invention (hereinafter, "Example 2"). In this experiment, relatively short sponge rollers 1B and 2B for cleaning a substrate were used in accordance with the size of the polishing apparatus 2.

[0047] <Experiment 1: Verification of surface pressure distribution using pressure-sensitive paper> As a first experiment, pressure-sensitive paper was attached on a surface plate, a sponge roller for cleaning a substrate was attached to a shaft, and both the surface plate and the shaft were rotated to conduct experiments on each of the comparative example, Example 1, and Example 2 regarding what kind of surface pressure distribution was drawn.

[0048] The experimental result of Experiment 1 is shown in FIG. 13. As shown in this figure, in the comparative example, the entire contact area is colored, and in addition to the relatively high surface pressure acting on the whole, concentric stripe patterns also appear. It can be presumed that the boundary between the protrusion portion with high pressure and the groove portion with low pressure appears as the stripe pattern.

[0049] On the other hand, in Example 1, only the entire contact area is very lightly colored, and it can be seen that the surface pressure is evenly distributed. Further, in Example 2, the entire contact area is colored slightly more than in Example 1, and it can be seen that an appropriate surface pressure is generated.

[0050] As a result of Experiment 1, in Example 1 and Example 2, compared with the pressure-sensitive paper showing the experimental result of the comparative example, the coloring is overall evenly, and it can be observed that the surface pressure is dispersed in both cases, and it can be expected to perform cleaning without unevenness.

[0051] <Experiment 2: Verification of surface pressure distribution using a polishing liquid (cleanser)> As a second experiment, a PVC sheet was attached to the surface plate as the object to be polished, a sponge roller for substrate cleaning was attached to the shaft, a cleanser was dropped onto the surface of the PVC sheet as the polishing liquid, and both the surface plate and the shaft were rotated to conduct an experiment using a comparative example and Example 1 to observe how the PVC sheet was polished.

[0052] The experimental result of Experiment 2 is shown in Fig. 14. As shown in this figure, in the comparative example, concentric stripe patterns appeared on the surface of the PVC sheet as in Experiment 1, and it can be inferred that the boundary between the high-pressure bump portion and the low-pressure groove portion appeared as stripe patterns. In addition, concentric stripe patterns with thick and thin circles were also observed, which is presumably due to the uneven height of the bumps.

[0053] On the other hand, in Example 1, a uniform pattern appeared on the surface of the PVC sheet throughout the contact area, and it was observed that the surface pressure was dispersed. There is no concern about damage caused by local pressure generation, and it is expected that cleaning can be performed without unevenness.

[0054] <Experiment 3: Verification of cleaning ability using polishing liquid (colloidal silica)> As a third experiment, a PVC sheet was attached to the surface plate as the object to be polished, a sponge roller for substrate cleaning was attached to the shaft, colloidal silica was dropped onto the surface of the PVC sheet as the polishing liquid, and both the surface plate and the shaft were rotated to conduct an experiment for each of the comparative example, Example 1, and Example 2 to observe how the polishing liquid was removed from the PVC sheet.

[0055] The experimental result of Experiment 3 is shown in Fig. 15. As shown in this figure, in the comparative example, it was observed that droplet-like residues appeared from the middle to the outer peripheral side of the contact area on the PVC sheet. This is presumably because the conventional sponge roller 1S for substrate cleaning with bumps has a weak effect of discharging liquid outward from the rotation center of the surface plate, and liquid and particles remain on the surface of the bumps, and the liquid and particles remained as they were at the end of the experiment. In addition, concentric stripe patterns with thick and thin circles were also observed, which is presumably due to the uneven height of the bumps.

[0056] On the other hand, in Example 1, it was observed that droplet-like residues appeared on the outermost peripheral side of the contact area on the PVC sheet, but no residues were found at the intermediate position from the center. It is presumed that the sponge roller 1B for substrate cleaning is excellent in discharging liquid outward from the rotation center of the surface plate, and at the intermediate position from the center, the liquid could be sufficiently guided and discharged by the action of the groove.

[0057] In Example 2, it was observed that residues appeared annularly on the outside of the contact area on the PVC sheet. That is, it can be seen that the liquid was successfully discharged outward over the entire circumference of the sponge roller 2B for substrate cleaning. It is presumed that the sponge roller 2B for substrate cleaning is even more excellent in discharging liquid outward from the rotation center of the surface plate, and the liquid could be sufficiently guided and discharged by the action of the groove over the entire circumference of the contact area.

[0058] <Experiment 4: Verification of Cleaning Ability Using Simulated Slurry (Aqueous Hot Cake Mix Solution)> As the fourth experiment, a PVC sheet was attached to the surface plate as the object to be polished, a sponge roller for substrate cleaning was attached to the shaft, and as the simulated slurry, a viscous aqueous hot cake mix solution was thinly applied in a disk shape on the surface of the PVC sheet. For each of the comparative example, Example 1, and Example 2, an experiment was conducted on how the simulated slurry was removed from the PVC sheet by rotating both the surface plate and the shaft for 30 seconds.

[0059] The simulated slurry (aqueous hot cake mix solution) is a commercially available hot cake mix mainly composed of wheat flour, containing other ingredients such as sugar, glucose, powdered maltose, salt, and baking powder, dissolved in an appropriate amount of water. Since the hot cake mix is easily available and inexpensive, and has appropriate viscosity when dissolved in water, it was used in this experiment as a substitute for the abrasive (slurry) used in actual CMP.

[0060] The experimental results of Experiment 4 are shown in Fig. 16. As shown in this figure, in the comparative example, it was observed that many residues appeared concentrically from the middle to the outer peripheral side of the contact area on the PVC sheet. This is presumably because the conventional ribbed sponge roller 1S for substrate cleaning has a weak effect of discharging liquid outward from the rotation center of the surface plate, the simulated slurry stays between the ribs, and the simulated slurry remained at the end of the experiment as it was.

[0061] On the other hand, in Example 1, residues appeared annularly outside the contact area on the PVC sheet, and at the middle position from the center of the contact area on the PVC sheet, it was observed that droplet-like residues appeared at the contact portion with the cleaning surface. This is presumably because the sponge roller 1B for substrate cleaning has an excellent effect of discharging liquid outward from the rotation center of the surface plate, and most of the simulated slurry could be guided and discharged by the action of the grooves, but only a very small amount of the simulated slurry remained on the cleaning surface.

[0062] In Example 2, residues appeared annularly outside the contact area on the PVC sheet, and at the middle position from the center of the contact area on the PVC sheet, it was observed that slightly droplet-like residues appeared at the contact portion with the cleaning surface. This is presumably because the sponge roller 2B for substrate cleaning that cleans the surface of the workpiece to be polished has an even more excellent effect of discharging liquid outward from the rotation center of the surface plate, and the simulated slurry could be sufficiently guided and discharged by the action of the grooves over the entire circumference of the contact area.

[0063] <Evaluation of Experimental Results> The conventional ribbed sponge roller 1S for substrate cleaning, which is a comparative example, is a molded product, the height of the ribs varies due to manufacturing tolerances, the ribs contact the substrate to be cleaned unevenly, the guidance of the liquid is also insufficient, the effect of discharging to the outside of the contact area is poor, and liquid and particles also remain on the rib surface. Thus, it is clear from the above experiments that the cleaning performance is lacking. Furthermore, since there is a pressure distribution difference, there is also a risk that the substrate to be cleaned may be damaged at the portion with high pressure.

[0064] On the one hand, the sponge rollers 1B and 2B for substrate cleaning according to the present invention in the embodiments are those in which the outer peripheral surface is cut by cutting and grooves are cut to form a cleaning surface. Compared with the comparative examples, it can be seen from Experiment 1 and Experiment 2 that since the height of the cleaning surface is uniform, the substrate to be cleaned can be evenly contacted with the cleaning surface, and it is clear from Experiment 3 and Experiment 4 that the grooves can guide liquids and particles and smoothly discharge them outside the contact area.

[0065] Since the pressure distribution is uniform, it is possible to avoid the risk of damage to the substrate to be cleaned in the portion with high pressure as in the comparative example.

[0066] In addition, by exposing continuous pores on the cleaning surface, it is also conceivable that the pores take in liquids and particles, that is, the wiping property is exhibited, and it is clear that this leads to less residue on the surface of the substrate to be cleaned, that is, high particle removal performance, as shown in Experiment 3 and Experiment 4.

[0067] Although there are some differences between the sponge rollers 1B and 2B for substrate cleaning depending on the difference in groove shape, both exhibit high cleaning performance. An optimal groove shape can be selected according to the properties such as the abrasive grain size and viscosity of the polishing liquid (slurry), the usage amount, or the polishing target. In addition, it can also be adjusted according to parameters such as the rotation direction, rotation speed, and pressing amount.

[0068] As described above, according to the present invention, it is possible to provide a sponge roller for substrate cleaning having excellent cleaning performance and manufacturability.

Explanation of symbols

[0069] 1A, 1B, 2A, 2B Sponge rollers for substrate cleaning, 10 Cleaning surface, 11 Groove, 11a Oblique groove, 11b Reverse oblique groove, 11c Parallel groove, 11d Cross groove, 12 Side wall, 20 Shaft insertion hole

Claims

1. A sponge roller for substrate cleaning, which is formed of a resin sponge having a continuous pore structure, and has a cylindrical shape with a cleaning surface on the outer peripheral side and a shaft insertion hole on the inner peripheral side, wherein the cleaning surface has a plurality of grooves that are linear when developed on a plane and extend in at least two directions.

2. The sponge roller for substrate cleaning according to claim 1, wherein the groove is a space having side walls perpendicular to the cleaning surface and a predetermined depth.

3. The sponge roller for substrate cleaning according to claim 1, wherein the groove is composed of an oblique groove intersecting the axis and a reverse oblique groove intersecting the axis in the opposite direction to the oblique groove.

4. The sponge roller for substrate cleaning according to claim 1, wherein the groove is composed of a parallel groove parallel to the axis and an intersecting groove intersecting the axis.

5. The sponge roller for substrate cleaning according to claim 1, wherein the resin sponge is a polyvinyl alcohol sponge (PVA sponge).

Images