Double-sided wafer polishing process

Abstract

Double-sided polishing process of a wafer (W) in which the wafer (W), which has been placed into a wafer loading hole (10a) of a carrier (10), is pressed and held together with the carrier (10) by an upper support plate (2) and a lower support plate (3), and the upper support plate (2) and the lower support plate (3) are rotated while a slurry is supplied to the wafer (W), the process comprising the following steps: Measuring an inclination value of a main surface of each of several supports (10) near the edge of the wafer loading hole (10a); Selecting from the multiple supports (10) for which the inclination value is equal to or less than a threshold, based on the measurement results of the inclination value; and Applying double-sided polishing to the wafer (W) using the selected carrier (10).

Description

Technical field

[0001] The present invention relates to a double-sided wafer polishing process and in particular a double-sided polishing process of a wafer using a double-sided polishing support with a special shape. Technical background

[0002] A silicon wafer, used as a substrate material for semiconductor devices, is manufactured by sequentially applying processes such as external circumferential grinding, cutting, lapping, etching, double-sided polishing, single-sided polishing, cleaning, etc., to a silicon single-crystal ingot grown using the Czochralski method. Among the processes mentioned above, double-sided polishing is necessary to obtain a wafer of a predetermined thickness and to improve its flatness. This process utilizes a double-sided polishing unit to polish both sides of the wafer simultaneously.

[0003] As a technique related to the double-sided polishing process, JP 2014-50 913 A and DE 11 2013 003 643 T5, for example, describe polishing both sides of the wafer while maintaining the flatness of the inner edge surface of a resin insert of a support holding the wafer at 100 µm or less and the verticality of this inner edge surface at 5° or less. This is done to suppress any deterioration of the wafer's flatness after polishing, such as sagging of the outer periphery (edge ​​drop). Furthermore, JP 2008-23 617 A describes using a titanium support for a double-sided polishing device and setting its surface roughness Ra to 0.14 µm or more to prevent sagging of the wafer's outer periphery after double-sided polishing and thus improve flatness.

[0004] Furthermore, JP 2003-19660A describes a double-sided polishing process which involves holding a frame body with a thickness greater than that of a wafer (object to be processed) in the opening of a carrier plate of a double-sided polishing device and polishing the wafer held in the frame body. That is, when the double-sided polishing of the wafer is performed, which is held in the opening of the carrier plate between upper and lower support plates that are integrally driven into rotation by a polishing cloth, the wafer is held in the frame body with a thickness greater than that of the frame body and mounted together with the frame body in the opening of the carrier plate.

[0005] Other polishing methods with a carrier plate and upper and lower support plates are known from JP 2000 - 210 863 A, JP 2014 - 188 668 A and JP 2015 - 104 771 A. Brief description of the invention [Problem to be solved by the invention]

[0006] In a double-sided polishing device, one or more wafers are mounted in each of several carriers for double-sided polishing; however, the outer periphery of the wafer is significantly deformed under the influence of a thickness profile of the carrier.

[0007] Accordingly, to suppress variation in the outer peripheral shape between wafers, the thickness of each of the carriers is conventionally measured, followed by sorting the carriers by thickness to minimize variation in thickness between a group of carriers that are mounted in the double-sided polishing device at once, and in this state the double-sided polishing is carried out.

[0008] However, even when double-sided polishing is performed using one group of carriers, where thickness variation is suppressed by sorting the carriers by thickness, the outer peripheral shape will still vary from one wafer to another, and one or more wafers with poorer flatness will inevitably exist. Therefore, improvements in this regard are desirable.

[0009] The objective of the present invention is to provide a double-sided polishing method for a wafer which is capable of reducing variation in flatness by suppressing edge drop of the wafer. [Means to solve the problem]

[0010] To solve the above problem, a double-sided polishing method for a wafer is provided according to the present invention, in which the wafer, which has been placed in a wafer loading hole of a carrier, is compressed and held together with the carrier by an upper support plate and a lower support plate, and the upper support plate and the lower support plate are rotated while a slurry is supplied to the wafer, the method comprising the following steps: measuring an inclination value of a major surface of each of several carriers near the edge of the wafer loading hole; selecting from the several carriers for which the inclination value is equal to or less than a threshold, based on the inclination value measurement results; and applying double-sided polishing to the wafer using the selected carrier.

[0011] According to the present invention, the inclination value is calculated within a certain area extending outwards from the inner peripheral edge of the wafer loading hole, and the inclination value of a carrier actually to be used in the double-sided polishing process is limited to an area equal to or less than the threshold, thereby suppressing edge drop of the wafer after polishing, so that the wafer can be improved with respect to an outer peripheral shape distribution.

[0012] In the present invention, the threshold is preferably set to 0.25 × 10 -3 and preferably at 0.2 × 10 -3 specified. If the inclination value within a certain range from the inner peripheral edge of the wafer loading hole is equal to or less than 0.25 × 10 -3If the wafer's edge drop can be reliably suppressed, this is achieved by using a support whose slope value is equal to or less than 0.2 × 10⁻⁶. -3 It is possible to produce a wafer that is more excellent with respect to an outer peripheral flatness, where ESFQRmax is equal to or less than 25 nm.

[0013] In the present invention, the measuring range of the inclination value is preferably defined in a region between the inner peripheral edge of the support and a position 2 mm inwards from there. The inclined shape of the main surface in the region between the inner peripheral edge of the support and a position 2 mm inwards from there has a significant influence on the wafer outer peripheral shape, so that by achieving the inclined shape of the support in this region, the flatness of the wafer outer peripheral part can be considerably improved.

[0014] In the present invention, the inclination value of the main surface of the support near the edge of the wafer loading hole is preferably the inclination value at a single position of the inner peripheral edge of the wafer loading hole, or an average of the inclination values ​​at multiple positions of the inner peripheral edge of the wafer loading hole, and more preferably the average of the inclination values ​​at multiple positions defined at equal intervals along the inner peripheral edge of the wafer loading hole. By appropriately measuring the inclination value at multiple positions, the reliability of the inclination value can be improved.

[0015] In the present invention, the slope value of the main surface of the support near the edge of the wafer loading hole is preferably a slope of a regression line derived from the thickness distribution of the support within a certain region from the inner peripheral edge. The edge slope of the wafer loading hole of the support is formed on both the upper and lower main surfaces of the support, so that by calculating the support slope value as a rate of change of thickness near the edge of the support, the slope value of both the front and rear surfaces of the support can be taken into account, and the support slope value can be easily calculated. Furthermore, by restricting the slope value of the support to a region equal to or less than a threshold, the edge slope of the wafer after polishing can be suppressed, thus improving the wafer with respect to its outer peripheral shape distribution.

[0016] In the present invention, the support is preferably made mainly of a metal, such as stainless steel, or a resin, such as glass epoxy resin. Furthermore, the support preferably consists of a combination of a metal support with a circular opening and an annular resin insert provided along the inner periphery of the opening of the support body, and the width of the resin insert is equal to or greater than 2 mm.

[0017] A double-sided polishing method of the wafer according to the present invention is a method which uses several supports to apply double-sided polishing to several wafers at once, wherein the respective inclination values ​​of the main surfaces of the several respective supports near the edge of a wafer loading hole are all preferably equal to or less than a threshold, and a variation in thickness between supports preferably falls within ±4 µm.

[0018] If all multiple carriers used in a batch process meet the above conditions, the flatness of the outer peripheral parts of the individual wafers can be sufficiently improved.

[0019] Furthermore, a double-sided polishing method of a wafer according to the present invention is a method which clamps a wafer, which is placed in a wafer loading hole of the carrier between an upper and lower support plate, together with a carrier and applies double-sided polishing to the wafer by rotating the upper and lower support plate while a slurry is supplied to the wafer, wherein only one carrier whose inclination value within a certain range from the inner peripheral edge of the wafer loading hole is equal to or less than 0.25 × 10 -3 , when the carrier is used.

[0020] According to the present invention, the inclination value is calculated within a certain area extending outwards from the inner peripheral edge of the wafer loading hole, and the inclination value of a carrier to be actually used in the double-sided polishing process is reduced to an area equal to or less than 0.25 × 10 -3 limited, which can suppress edge falloff of the wafer after polishing, thus improving the wafer with respect to an outer peripheral shape distribution. [Advantageous effects of the invention]

[0021] According to the present invention, a double-sided polishing process of a wafer can be provided which is able to reduce variation in flatness by suppressing edge drop of the wafer. Brief description of the drawings Fig. Figure 1 is a schematic cross-sectional side view illustrating the configuration of a double-sided polishing device according to an embodiment of the present invention. Fig. 2 is a top view of the Fig. 1 illustrated double-sided polishing device. Fig. 3A to 3C are views that illustrate the configuration of the carrier, where Fig. 3A is a top view, Fig. 3B is a cross-sectional side view and Fig. Figure 3C is a partially enlarged view showing the proximity of the inner peripheral edge (resin insert) of the wafer loading hole of the carrier. Fig. Figure 4 is a flowchart to explain the double-sided polishing process of the wafer, including a carrier selection process. Fig. 5A and Fig. Figure 5B are schematic diagrams to explain the relationship between the inclined shape of the main surface of the support 10 near the edge of the wafer loading hole 10a and the edge shape of the wafer after polishing, wherein Fig. 5A illustrates the inclined surface according to a conventional technique and Fig. 5B illustrates the inclined surface according to the present invention. Fig. 6A and Fig. 6B are views explaining a procedure for measuring the carrier inclination value, wherein Fig. 6A is a top view to explain line scan measurement positions of the carrier and Fig. 6B is a cross-sectional side view to explain the image of a beam to be measured and a position for calculating the inclination of the beam. Fig. Figure 7 is a scatter plot illustrating the relationship between the carrier inclination value and ESFQDmean. Fig. Figure 8 is a scatter plot illustrating the relationship between the carrier inclination value and ESFQRmax. Fig. Figure 9 is a table showing the relationship between the gap and ESFQDmean and ESFQRmax of the wafer after polishing for each inclination value of the support near the edge of the wafer loading hole (support hole). Fig. Figure 10 is a scatter plot illustrating the relationship between the inclination value of a beam with a beam lifetime of 0 min (before use) and the inclination value of a beam with a beam lifetime of 40000 min (after use). Detailed description of the embodiments

[0022] Preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0023] Fig. Figure 1 is a schematic cross-sectional side view illustrating a configuration of a double-sided polishing device according to an embodiment of the present invention. Fig. 2 is a top view of the Fig. 1 illustrated double-sided polishing device, wherein Fig. 1 a cross-sectional view along line RR' from Fig. 2 is.

[0024] As in Fig. 1 and Fig. As illustrated in Figure 2, a double-sided polishing device 1 comprises an upper support plate 2 and a lower support plate 3, positioned vertically opposite each other. Polishing cloths 4 and 5 are attached to the lower surface of the upper support plate 2 and the upper surface of the lower support plate 3, respectively. A sun gear 6 is positioned at the central part of the device between the upper support plate 2 and the lower support plate 3, and an internal gear 7 is positioned at the peripheral part. Each wafer W, for example a silicon wafer, is positioned between the upper support plate 2 and the lower support plate 3 in an intermediate state, having been placed in a wafer loading hole 10a of a carrier 10.

[0025] As in Fig. As illustrated in Figure 2, five supports 10 are arranged around the sun gear 6, and the outer peripheral teeth 10b of each support 10 engage with corresponding teeth of the sun gear 6 and the inner gear 7. The upper support plate 2 and the lower support plate 3 are driven by a drive source (not shown), causing each support 10 to rotate around the sun gear 6. During this rotation, the wafers W, placed in the wafer loading holes 10a of the supports 10, are held by the supports 10, and both surfaces of the wafer are simultaneously polished by contact with the upper and lower polishing cloths 4 and 5. A polishing fluid is supplied from a nozzle (not shown) during polishing. For example, an alkaline solution containing dispersed colloidal silicon dioxide can be used as the polishing fluid.

[0026] Fig. 3A to 3C are views illustrating the configuration of carrier 10, where Fig. 3A is a top view, Fig. 3B is a cross-sectional side view and Fig. Figure 3C is a partially enlarged view showing the proximity of the inner peripheral edge (resin insert 12) of the wafer loading hole 10a of the carrier 10.

[0027] As in Fig. 3A and Fig. As illustrated in Figure 3B, the support 10 includes a metal support body 11, which is made of a metal and has a circular opening 11a that is larger than the wafer W, and a resin insert 12, which has a ring shape and is arranged along the inner peripheral edge of the opening 11a of the support body 11.

[0028] The carrier body 11 is a disc-like element, and the outer peripheral teeth 11b are provided on its outer peripheral portion. Although a typical material for the carrier body 11 is SUS, titanium or another metallic material can be used. The thickness D of the carrier body 10 is determined based on a target wafer thickness W after double-sided polishing. For example, the thickness of the carrier 10 used for a wafer with a diameter of 300 mm is set to approximately 0.8 mm, and the dimensioning / polishing is performed to reduce the thickness of the wafer W, which is approximately 1 mm thick before processing, to a thickness equivalent to that of the carrier 10.Since the central position of the opening 11a is offset from the central position of the carrier body 11, the wafer W, which is placed in the opening 11a, is moved eccentrically with the center of the carrier body 11 as the axis of rotation, thereby improving polishing efficiency and polishing uniformity.

[0029] The resin insert 12 lies between the outer peripheral surface of the wafer W and the inner edge surface of the opening 11a of the carrier body 11 and serves to prevent contact between them. An inner opening 12a of the resin insert 12 forms the wafer loading hole 10a (see Fig. 2) of the support 10, and the outer peripheral surface of the wafer W is in contact with the inner edge surface of the resin insert 12. The lateral width (ring width) of the resin insert 12 is, for example, 2 mm or more and is determined taking into account the size of the opening 11a of the support body 11 and the size of the wafer W. The thickness of the resin insert 12 is preferably equal to the thickness D of the support body 11.

[0030] As in Fig. As illustrated in Figure 3C, both the upper and lower corners of the inner peripheral portion of the inner opening 12a of the resin insert 12 are not perpendicular and have a sloping shape. If the wafer W, with a thickness of approximately 1 mm, is polished to a thickness equivalent to that of the support 10 as described above before processing, the support 10 is also polished along with the wafer W, so that a bevel inevitably occurs in the inner peripheral edge of the resin insert 12, which is made of a soft material.

[0031] In the present embodiment, an inclination angle of the main surface of the support 10 near the edge of the wafer loading hole 10a refers to the slope value of an upward gradient extending from the inner peripheral edge of the resin insert 12 to its outer peripheral edge. The edge slope is formed in the upper and lower main surfaces of the resin insert 12, so the support inclination value is first calculated by determining a thickness distribution of the resin insert 12 from the inner peripheral edge to the outer periphery, and then by calculating the slope of a regression line derived from the thickness distribution within a certain region of the inner peripheral edge. That is, the support inclination value can be calculated as a rate of change of thickness near the edge of the support.

[0032] Assuming that the thickness of the inner peripheral edge of the resin insert 12, calculated from the regression line derived from the thickness distribution of the resin insert 12, is y1, and that the thickness of the outer peripheral edge of the resin insert 12 at a position separated from the inner peripheral edge by a distance x is y2, the inclination value tanθ of the beam is (y2 - y1) / x. That is, the inclination value tanθ of the beam is calculated as the sum of a front inclination value tanθ1 = h1 / x and a back inclination value tanθ2 = h2 / x. The angle θ is a sum of θ1 and θ2, and normally θ1 is approximately equal to θ2.

[0033] To improve the outer peripheral shape of the wafer W after double-sided polishing, the inclination value tanθ of the support 10 must be equal to or less than 0.25 × 10 -3This means that in the double-sided wafer polishing process according to the present embodiment, a carrier whose inclination value is 0.25 × 10 -3 If the value exceeds the specified limit, it is not used. To improve the reliability of the tilt value of carrier 10, it is preferable to use an average value of tilt values ​​measured at several positions around the wafer loading hole.

[0034] Fig. Figure 4 is a flowchart to explain the double-sided polishing process of the wafer, including a carrier selection process.

[0035] As in Fig. As illustrated in Figure 4, the double-sided polishing process of the wafer according to the present embodiment includes a step (S1) of prior measurement of the inclination value of the support 10 to be used in the double-sided polishing device, and a step (Yes in S2, S3) of selecting a support whose inclination value is equal to or less than a threshold (0.25 × 10 -3 ) is, as a carrier that can be used in the double-sided polishing process, and a step (S4) of applying double-sided polishing to the wafer using the selected carrier. A carrier whose slope value exceeds the threshold is excluded from the carriers to be used (No in S2, S5).

[0036] As described above, the wafer can be processed by limiting the inclination value of the support 10, which is to be used in the double-sided polishing of the wafer, to 0.25 × 10 -3or less so with regard to an outer peripheral shape distribution after polishing.

[0037] Fig. 5A and Fig. Figure 5B are schematic diagrams to explain the relationship between the inclined shape of the main surface of the support 10 near the edge of the wafer loading hole 10a and the edge shape of the wafer after polishing. Fig. 5A illustrates the inclined surface according to a conventional technique and Fig. Figure 5B illustrates the inclined surface according to the present invention.

[0038] In double-sided polishing, the wafer W lies between the upper and lower support plates 2 and 3 and is polished while being subjected to pressure from the polishing cloths 4 and 5, which each have a thickness of approximately 1 mm and are positioned between the upper support plate 2 and the wafer W or between the lower support plate 3 and the wafer W, respectively, so that if the inclination value of the support 10 is large, as in Fig. Figure 5A illustrates a large depression between the support 10 in which wafer W is formed, and the polishing cloths 4 and 5 sink into the depression, resulting in an increased polishing area at the edge of wafer W. This means that the stress (polishing pressure) exerted on the edge of wafer W is increased, resulting in a large edge drop of wafer W after polishing.

[0039] However, as in Fig. Figure 5B illustrates that if the inclination value of the support 10 is small, the depression formed between the support 10 and the wafer W is small, resulting in a reduction of the polishing extent of the edge of the wafer W. That is, the stress (polishing pressure) exerted on the edge of the wafer W is reduced, so that the edge drop of the wafer W after polishing is small.

[0040] As the polishing process progresses to reduce the thickness of the wafer W, the thickness difference (gap) between the wafer W and the support 10 becomes small; however, if the thickness around the wafer loading hole is small, the outer peripheral shape of the wafer W is subject to edge degradation, even as the polishing progresses. However, in the present embodiment, the outer peripheral shape of the wafer W can be improved after polishing by: measuring the thickness of the support 10 around the wafer loading hole; calculating the slope value of the support 10 around the wafer loading hole based on the result of the measurement; and achieving a slope value of the wafer 10 to be used equal to or less than a threshold value.

[0041] The dominant factor influencing wafer edge drop (W) is the thickness profile around the wafer loading hole. Even if the thickness of the support wafer varies, a good result is achieved if the slope value around the wafer loading hole is small. However, the influence of the slope value around the wafer loading hole is significant, so if the wafer is polished to a thickness equivalent to the support thickness, causing the wafer thickness to vary relative to the support thickness after polishing, edge drop cannot be prevented.

[0042] As described above, in the double-sided polishing process of a wafer according to the present embodiment, which compresses and holds the wafer W, which is placed in the wafer loading hole 10a of the carrier 10, together with the carrier 10 between the upper and lower support plates 2 and 3, and applies double-sided polishing to the wafer W by rotating the upper and lower support plates 2 and 3 while a slurry is delivered to the wafer W, only a carrier is used whose inclination value within a certain range from the inner peripheral edge of the wafer loading hole 10a of the carrier 10 is equal to or less than 0.25 × 10 -3 is, so that the sloping shape of the wafer outer peripheral part can be suppressed, thereby allowing a reduction in the variation in the flatness of the wafer outer peripheral part.

[0043] While the present invention has been described based on the preferred embodiment, it is not limited to the embodiments described above, and various modifications may be made within the scope of protection of the present invention. Accordingly, all such modifications are included in the present invention.

[0044] For example, while in the above embodiment the support 10 consists of the metal support body 11 and the resin insert 12, the entire support can be made of a single resin, meaning that the support body 11 and the resin insert 12 can be integrated together. Alternatively, the entire support 10 can be made of a single metal.

[0045] Furthermore, while a support 10 has one wafer loading hole 10a and holds the wafer W in the embodiment above, a support 10 can have multiple wafer loading holes 10a. In this case, the slope value of the main surface near the edge of each of the multiple wafer loading holes 10a must be equal to or less than 0.25 × 10 -3 and the variation in thickness between the multiple supports is preferably ±4 µm or less.

[0046] Furthermore, the configuration of the double-sided polishing device 1 in the present embodiment is merely illustrative, and any of the different types can be used. Moreover, the wafer to be polished is not limited to a silicon wafer, and wafers of different types can be used as the wafer to be polished. [Examples]

[0047] A large number of support samples with varying support inclination values ​​were used to evaluate the edge flatness of a 300 nm diameter silicon wafer after double-sided polishing. For this evaluation test, 150 supports were prepared, and a line scan measurement was then performed on these 150 prepared supports using a laser misalignment measuring device. The inclination value of the main surface of each support was then calculated near the wafer loading hole.

[0048] Fig. 6A and Fig. Section 6B presents views explaining a procedure for measuring the beam inclination value. Fig. Figure 6A is a top view to explain the line scan measurement positions of the support and the Fig. Figure 6B is a cross-sectional side view to explain the image of a beam to be measured and a position for calculating the inclination of the beam.

[0049] As in Fig. 6A and Fig. As illustrated in Figure 6B, eight line scan measurement positions were established around the wafer loading hole. The slope values ​​at each measurement position were then calculated, and the average of these values ​​was used as the beam slope value. A measurement length x, used in calculating the slope value at each beam measurement position, was set to 2 mm, which is the width of the area where the resin insert exists.

[0050] The line scan measurement was performed as illustrated from the inside of the wafer loading hole to outside the support, and the slope of the regression line (regression coefficients) was calculated when the measurement length and support thickness were set to x and y, respectively.

[0051] The slope 'a' of the regression line is obtained by a correlation coefficient · (standard deviation of y / standard deviation of x) and assuming that the mean values ​​of x and y arex or m y If they are, the following expression will be obtained. α=∑(x−mx)(y−my)∑(x−mx)2

[0052] The carriers, whose inclination values ​​were measured, were used to apply double-sided polishing to the silicon wafer. The double-sided polishing device used in the evaluation test is the device made of Fig. Two trays, into which five carriers can be placed at once, were used, allowing the polishing process to be carried out with five carriers as a single group. Polishing conditions included: using a 1.0 mm thick urethane foam pad; using an alkaline slurry containing colloidal silicon dioxide as the abrasive grain; setting a backing plate rotation speed of 20 to 30 rpm; and setting a processing surface pressure of 300 g / cm². 2The thickness of the silicon wafer before polishing was set to 790 µm, the thickness of the substrate was set to 778 µm, and the target thickness of the silicon wafer after polishing was set to between 778 µm and 782 µm. A laser displacement measuring device was used to measure the thickness of the silicon wafer.

[0053] The wafer's ESFQD and ESFQR values ​​were then measured after double-sided polishing, confirming an influence of the support inclination value on the wafer's outer periphery shape. ESFQD (Edge Site Flatness Front Reference Least Square Deviation) and ESFQR (Edge Site Flatness Front Reference Least Square Range) are each an evaluation index for flatness (spot flatness) that focuses on the wafer's edge, where flatness deteriorates slightly, and each indicates the amount of edge flatness loss. The wafer's edge flatness is calculated for each unit area (spot) obtained by uniformly dividing the annular outer periphery area, defined in a region (sector length: 30 mm) from, for example, 2 mm to 32 mm from the wafer's outermost periphery, in the periphery direction. A flatness measuring instrument (WaferSight2, manufactured by KLA-Tencor Corporation) was used for wafer flatness measurement.During the measurement, the measuring range was set to 296 mm (excluding a 2 mm section from the outer periphery). Furthermore, for the edge measurement, the number of sectors (number of location sites) was set to 72 and the sector length to 30 mm.

[0054] Fig. 7 is a scatter plot illustrating the relationship between the carrier inclination value and ESFQDmean and Fig. Figure 8 is a scatter plot illustrating the relationship between the carrier inclination value and ESFQRmax.

[0055] ESFQDmean is the mean value of ESFQDs across all locations, and ESFQD refers to a deviation with a larger absolute value between a maximum deviation (α) and a minimum deviation (-β) from a reference surface (best surface approximation of the location), obtained by a least-squares method from a thickness distribution at the location. For example, if α is greater than β, then ESFQD is α, whereas if α is less than β, then ESFQD is -β. ESFQD is an index that can have positive and negative values, and the larger ESFQD is on the negative side, the greater the edge drop of the wafer.

[0056] ESFQRmax is the maximum value among the ESFQRs across all locations, and ESFQR refers to a difference (α - (-β)) between the maximum deviation (α) and the minimum deviation (β) from the best surface approximation of the location. The larger ESFQR is, the greater the edge slope of the wafer, while the outer peripheral flatness improves as ESFQD approaches 0.

[0057] As in Fig. 7 and Fig. As illustrated in Figure 8, ESFQDmean on the negative side becomes larger the larger the beam inclination value becomes. In particular, when the beam inclination value is equal to or greater than 0.25 × 10 -3As the slope value decreases, the variation in the distribution of ESFQDmean increases, and consequently, the variation in the distribution of ESFQRmax also increases. Conversely, as the support slope value decreases, the ESFQDmean values ​​tend to move from the negative to the positive side, resulting in a smaller wafer edge drop and improved ESFQRmax. From the above, it can be seen that to improve ESFQR and ensure stable production, the slope value should ideally be equal to or less than 0.25 × 10⁻⁶. -3 is. Furthermore, the graph reveals Fig. 8, that if the target value of ESFQRmax is set to be equal to or less than 25 nm, improved outer peripheral flatness can be maintained by using a carrier with a slope value equal to or less than 0.2 × 10 -3 is used.

[0058] Then the influence of the “gap” and the support inclination value on the wafer flatness was evaluated, where the “gap” is a difference (wafer thickness after polishing - wafer thickness) between the mean value (average thickness) at calculation positions for the thickness of the support near the edge of the wafer loading hole (around the support hole) and the wafer thickness after polishing.

[0059] Fig. Figure 9 is a table showing the relationship between the gap and ESFQDmean and ESFQRmax of the wafer after polishing for each inclination value of the support near the edge of the wafer loading hole (support hole).

[0060] As in Fig. As shown in Figure 9, this evaluation test was performed in the gap range from 0 µm to +4 µm and in this range it was possible to make ESFQRmax of any wafer equal to or less than 25 µm by making the support inclination value equal to or less than 0.25 × 10 -3was set. Conversely, it was possible if the slope value was equal to or greater than 0.3 × 10 -3 The goal was to make ESFQRmax equal to or less than 25 µm. At that time, when the thickness of an area (a large part of the support) where generally no resin is used was achieved / used within a range of ±4 mm of the thickness of the slope value calculation position (near the edge), no influence of flatness was confirmed. This reveals that the thickness profile near the edge of the wafer loading hole, which is the slope value calculation position, determines the shape of the wafer edge drop this time, and that, accordingly, achieving this shape near the edge of the wafer loading hole is important.

[0061] Fig. Figure 10 is a scatter plot illustrating the relationship between the inclination value of a beam with a beam lifetime of 0 min (before use) and the inclination value of a beam with a beam lifetime of 40000 min (after use).

[0062] As in Fig. As illustrated in Figure 10, it was confirmed that the slope value of the beam was reduced by 0% to approximately 30% after use, relative to the beam before use. This is because the slope value decreases as the beam's service life progresses due to a reduction in the overall beam thickness. This indicates that by limiting the slope value at an initial load, it is possible to continue processing wafers with a high degree of flatness. [List of reference symbols] 1 double-sided polishing device 2 upper support plate 3 lower support plate 4, 5 polishing cloths 6 sun wheel 7 Internal gear 10 carriers 10a Wafer loading hole 10b Outer peripheral teeth 11 support bodies 11a Opening of the carrier body 11b Outer peripheral teeth of the carrier body 12 Resin application 12a inner opening of the resin insert W Wafer

Claims

A double-sided polishing process for a wafer (W), wherein the wafer (W), which has been placed into a wafer loading hole (10a) of a carrier (10), is pressed and held together with the carrier (10) by an upper support plate (2) and a lower support plate (3), and the upper support plate (2) and the lower support plate (3) are rotated while a slurry is supplied to the wafer (W), the process comprising the following steps: measuring an inclination value of a principal surface of each of several carriers (10) near the edge of the wafer loading hole (10a); selecting from the several carriers (10) for which the inclination value is equal to or less than a threshold, based on the inclination value measurement results; and applying double-sided polishing to the wafer (W) using the selected carrier (10). Double-sided polishing method of a wafer (W) according to claim 1, wherein the threshold is set to 0.25 × 10-3. Double-sided polishing method of a wafer (W) according to claim 1 or 2, wherein the measuring range of the inclination value is defined in a region between the inner peripheral edge of the carrier (10) and a position 2 mm inwards from there. Double-sided polishing method of a wafer (W) according to one of claims 1 to 3, wherein the inclination value of the main surface of the carrier (10) near the edge of the wafer loading hole (10a) is the inclination value at one position of the inner peripheral edge of the wafer loading hole (10a) or an average of the inclination values ​​at several positions of the inner peripheral edge of the wafer loading hole (10a). Double-sided polishing method of a wafer (W) according to one of claims 1 to 4, wherein the inclination value of the main surface of the support (10) near the edge of the wafer loading hole (10a) is an inclination of a regression line derived from the thickness distribution of the support (10) within a certain area from the inner peripheral edge. Double-sided polishing method of a wafer (W) according to any one of claims 1 to 5, wherein the carrier (10) consists of a combination of a carrier body (11) made of metal with a circular opening (11a) and an annular resin insert (12) provided along the inner periphery of the opening of the carrier body (11), and the width of the resin insert (12) is equal to or greater than 2 mm. Double-sided polishing method of a wafer (W) according to any one of claims 1 to 6, wherein in a method of using multiple supports (10) for applying double-sided polishing to multiple wafers (W) at once the respective inclination values ​​of the main surfaces of the multiple supports (10) near the edge of the wafer loading hole (10a) are equal to or less than a threshold and a variation in thickness between the multiple supports (10) falls within ±4 µm.

Images