Positioning detection device and positioning method

Abstract

The application relates to the technical field of semiconductor processing, in particular to a positioning detection device and a positioning method. The device comprises two calibration plates, each of which is provided with a plurality of standard patterns. The interval between two adjacent standard patterns needs to meet the requirement that the camera field of view of a vision system can simultaneously capture at least two complete standard patterns in the length and width directions. The device is provided with the calibration plates with the plurality of standard patterns. The repeated positioning precision of each time of return is measured through visual identification and calculation after each time of movement return. In the position adjustment process, the position change amount is calculated by capturing the standard patterns with different codes.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of semiconductor processing, in particular to a positioning detection device and a positioning method. BACKGROUND

[0002] In a wafer high-precision alignment process, after measuring the spatial position coordinates of the upper and lower wafers, the upper and lower carrier tables carrying the upper and lower wafers need to be moved to the position before measurement. At this time, it is necessary to detect whether the upper and lower carrier tables are accurately moved to the position between measurements, or to detect the difference (PEC, position error compensation) between the two positions.

[0003] On the existing device, the repeated positioning accuracy is usually measured by arranging at least three laser sensors or capacitive sensors. This scheme has good accuracy, but the sensors usually have only a small measurement distance, which needs to be paid special attention when used.

[0004] In addition, in the alignment process, after measuring the position deviation of the upper and lower wafers, one of the wafers needs to be moved to align the positions of the two wafers. At present, the movement amount of the carrier table is usually detected by the encoder of the mechanism driving the carrier table to move, rather than directly measuring the movement amount of the carrier table. From the driving mechanism to the carrier table, there are other parts (such as bearings, springs, air-floating marble motion tables, etc.) in between, and the transmission relationship is relatively complex, which may introduce errors in the intermediate links. SUMMARY

[0005] The purpose of the present application is to provide a positioning detection device and a positioning method.

[0006] In order to solve the above technical problems, the present application provides the following technical solutions:

[0007] A positioning detection device comprises two calibration plates, each of which is provided with a plurality of standard patterns, and the spacing between adjacent two standard patterns needs to meet the requirement that the camera field of view of the vision system can simultaneously capture at least two complete standard patterns in the length and width directions.

[0008] Further, all the standard patterns on the calibration plates are arranged in a matrix. The number of rows and columns is not fixed, and mainly depends on the range, i.e. the expected displacement to be measured.

[0009] Further, the standard patterns on the two calibration plates can be the same or different. Preferably, they are the same, because the learning of the patterns is involved, and in the case of the same patterns, the system only needs to learn once. The shape of the standard pattern is not limited, and any pattern that can be recognized by the camera can be used. As long as the pattern is trained by machine vision, the center point of the pattern can be calculated. In theory, it can be a square, a circle, etc., or even a point.

[0010] Further, all standard patterns on the same calibration plate are the same, and each standard pattern has a unique number. The image algorithm can identify the number to obtain the position of the photographed standard pattern in the calibration plate. Preferably, all numbered standard patterns are the same, so that the machine vision learning is simpler, and only the same pattern and number need to be learned.

[0011] Alternatively, all standard patterns on the same calibration plate are different. In this case, different patterns represent different positions, and the type, number and arrangement of patterns need to be obtained by measurement and analysis, for example, by combining the unique position through the different thickness and spacing of the engraved lines.

[0012] Further, the calibration plate is of transparent material; the standard pattern can be machined on the calibration plate using electron beam, embossing, etching and other processes, and the size and distance of each element in the standard pattern are known.

[0013] Further, the calibration plate is mounted on a fixed seat, the fixed seat is fixed on a support, and the bottom of the support is fixedly connected with a base. The base and the support are both of marble material. Two calibration plates can be mounted on two fixed seats or on the same fixed seat, as long as their installation positions do not interfere with other parts. The support can be a door-shaped support, and the fixed seat is fixed on the door-shaped support by screws, and then the door-shaped support is fixed or glued on the base by screws.

[0014] Further, the fixed seat is provided with a groove, and the calibration plate is embedded or pasted in the groove.

[0015] Further, the base is provided with a lower bearing table and an upper bearing table, and two sets of vision systems are respectively installed on the lower bearing table and the upper bearing table, and the vision systems are used to shoot the standard patterns on the calibration plate.

[0016] The positioning method of the positioning detection device is as follows:

[0017] When used to measure the repeat positioning accuracy, first, the lower bearing table is aligned with the calibration plate when the lower bearing table is at the initial position, and the vision system installed on the lower bearing table takes a picture, calculates the position of the complete standard pattern A in the camera field of view of the vision system, and finally calculates the position difference between the center of the field of view and the standard pattern A.

[0018] After the lower bearing table moves to the far end to complete the required process action, the lower bearing table returns to the initial position again from the far end, and the vision system of the lower bearing table is aligned with the calibration plate again and takes a picture, calculates the position of the complete standard pattern B in the camera field of view of the vision system, and calculates the position difference between the center of the field of view and the standard pattern B.

[0019] According to the positions of the standard pattern A and the standard pattern B on the calibration plate, the distance between the center of the field of view when the vision system takes the first picture and the center of the field of view when the vision system takes the second picture in the X and Y directions can be obtained, and the distance is the repeated positioning error of the lower carrier table before and after moving.

[0020] When the position change for monitoring the relative position adjustment of the upper and lower wafers is used, two sets of vision systems of the upper carrier table or the lower carrier table take pictures of the two calibration plates respectively, each set of vision systems can take the position change before and after moving, and then the X and Y direction displacement of each set of vision systems can be calculated, and finally the X and Y displacement and the deflection angle of the upper carrier table or the lower carrier table can be calculated.

[0021] Compared with the prior art, the positioning detection device and the positioning method have at least the following beneficial effects:

[0022] The positioning detection device of the present application is provided with a calibration plate with a plurality of standard patterns, and the repeated positioning accuracy of each return is measured through visual recognition and calculation after each movement. During the position adjustment process, the position change is calculated by taking pictures of standard patterns with different codes.

[0023] The positioning method of the present application is a new scheme for measuring the repeated positioning accuracy, which is not limited by the working distance. In theory, as long as the calibration plate is large enough, a large distance can be measured, and the image recognition measurement can provide high accuracy. Secondly, in the prior art, when the relative positions of the upper and lower wafers are adjusted, the movement of the corresponding carrier table can only be detected by the encoder of the driving mechanism, and cannot be directly seen in real time. However, the present method can detect the driving mechanism encoder while additionally increasing a direct measurement device with a large range, so that even if a large adjustment is needed, it can still be detected. Thirdly, since the standard pattern is processed by etching process, the position between each pattern is fixed, the standard pattern on the calibration plate has an independent and unique code, and each standard pattern taken represents a unique position, so the present method is an absolute measurement, which can continue to measure after the equipment is accidentally powered off and restored, the data will not be cleared, and no zero reset is needed.

[0024] The positioning detection device and the positioning method of the present application will be further described below with reference to the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 It is a front view of the calibration plate;

[0026] Figure 2 It is a schematic diagram of the installation of the positioning detection device of the present application;

[0027] Figure 3 It is a schematic diagram of the installation of the calibration plate;

[0028] Figure 4 schematic diagram of position difference of two shots;

[0029] Figure 5 schematic diagram of calculation of translation amount of the carrier table;

[0030] Figure 6 schematic diagram of calculation of deflection amount of the carrier table.

[0031] Wherein, 1 - base, 2 - lower carrier table, 3 - upper carrier table, 4 - vision system, 5 - calibration plate, 6 - bracket, 7 - fixed seat. DETAILED DESCRIPTION

[0032] As shown in Figures 1-3 , a positioning detection device includes two calibration plates 5, each of which is provided with a plurality of standard patterns. The spacing between adjacent two standard patterns needs to meet the requirement that the camera field of view of the vision system 4 can simultaneously capture at least two complete standard patterns in the length and width directions. The spacing between adjacent two standard patterns in the calibration plate is very critical, otherwise it may not be able to identify a complete standard pattern. When two complete standard patterns are captured at the same time, one of them can be selected for calculation.

[0033] As shown in Figure 1 , all the standard patterns on the calibration plate 5 are arranged in a matrix form, which is 7 rows by 10 columns. According to actual needs, the number of rows and columns is not fixed, mainly depending on the range, i.e. the displacement to be measured.

[0034] The standard patterns on the two calibration plates 5 are the same. As shown in Figure 1 , all the standard patterns on the same calibration plate 5 are the same, and each standard pattern has a unique number. The standard pattern is a combination of a square and a cross, and the lower right corner of the square is marked with a number.

[0035] The calibration plate 5 is transparent glass material; the calibration plate 5 is installed on the fixed seat 7, the fixed seat 7 is fixed on the bracket 6, and the bottom of the bracket 6 is fixedly connected with the base 1. The base and the bracket are both made of marble. The two calibration plates can be installed on two fixed seats or on the same fixed seat, as long as their installation positions do not interfere with other components. The bracket is selected as a door type bracket, the fixed seat is fixed on the door type bracket with screws, and then the door type bracket is fixed or glued on the base with screws.

[0036] The fixed seat 7 is provided with a groove, and the calibration plate 5 is embedded or pasted in the groove.

[0037] The base 1 is provided with a lower carrier table 2 and an upper carrier table 3, and the lower carrier table 2 and the upper carrier table 3 are respectively installed with two groups of vision systems 4, which are used to shoot the standard patterns on the calibration plate 5.

[0038] As Figure 4 For example, the positioning method of the positioning detection device is specifically:

[0039] When used to measure the repeated positioning accuracy, first, the lower bearing table is at the initial position, the vision system installed on the lower bearing table is aligned with the calibration plate and takes a photo, and the position of the complete standard pattern A (the pattern numbered 23 in Figure 4 ) in the camera field of view of the vision system (i.e. the dotted box on the left in Figure 4 ) is calculated, and finally the position difference (i.e. x1, y1 in Figure 4 ) between the center of the field of view (the thick cross mark in Figure 4 ) and the standard pattern A is calculated;

[0040] After the lower bearing table moves to the far end to complete the required process action (which can be a certain step in the wafer alignment process, such as the lower bearing table with the vision system finding the mark of the upper wafer), the lower bearing table returns to the initial position again from the far end, at which time the vision system of the lower bearing table is aligned with the calibration plate again and takes a photo, and the position of the complete standard pattern B (the pattern numbered 36 in Figure 4 ) in the camera field of view of the vision system (i.e. the dotted box on the right in Figure 4 ) is calculated, and the position difference (i.e. x2, y2 in Figure 4 ) between the center of the field of view and the standard pattern B is calculated;

[0041] According to the positions of the standard pattern A and the standard pattern B on the calibration plate, the distance between the standard pattern A and the standard pattern B (i.e. a, b in Figure 5 ) can be obtained, and then the repeated positioning error of the lower bearing table before and after movement is calculated, i.e. Δx = x1 + x2 + a, Δy = y1 + y2 + b. To avoid the problem of taking positive and negative signs when calculating the position error each time, the coordinate system direction can be agreed, i.e. x1, x2, y1, y2 in the above formula already contain positive and negative signs.

[0042] Similarly to the above method, when used to monitor the position change of the relative position adjustment of the upper wafer and the lower wafer, two sets of vision systems of the upper bearing table or the lower bearing table take photos of two calibration plates respectively, each set of vision system will detect the position change before and after movement, and then the X and Y displacement amounts of each set of vision system are calculated, and finally the translation amount and the deflection amount (the deflection angle in the plane can be synthesized from the X and Y translations) of the upper bearing table or the lower bearing table are calculated. When the upper bearing table moves from the position before adjustment to the position after adjustment, if the X and Y change amounts detected by the left and right vision systems of the upper bearing table are the same, i.e. the X change amount detected by the left vision system (abbreviated as left vision) is equal to the X change amount detected by the right vision system (abbreviated as right vision), and the Y change amount detected by the left vision system is equal to the Y change amount detected by the right vision system, it is considered that the upper bearing table has completed the translation compensation.Figure 6 When the upper carrier moves from the position before adjustment to the position after adjustment, if the X, Y variation detected by the two vision systems on the left and right sides of the upper carrier are different, it is considered that the upper carrier has completed the deflection compensation, as shown in ​ During the adjustment, the left and right vision systems are locked relative to the carrier, so the distance L between the centers of the fields of view of the left and right vision systems is constant, and y1, y2 can be calculated by the vision system, so the deflection angle The same applies to the calculation of x1, x2.

[0043] The above-described embodiments are merely preferred embodiments of the present application and are not intended to limit the scope of the present application. Various modifications and improvements to the technical solutions of the present application made by those of ordinary skill in the art without departing from the design spirit of the present application shall fall within the scope of protection of the present application as defined by the claims.

Claims

1. A positioning method of a positioning detection device, characterized by: The positioning detection device comprises two calibration plates (5), each of which is provided with a plurality of standard patterns, and the interval between adjacent two standard patterns needs to meet the requirement that the camera field of view of the vision system (4) can simultaneously capture at least two complete standard patterns in the length and width directions; the calibration plate (5) is installed on a fixed seat (7); There are a lower bearing table (2) and an upper bearing table (3), and two groups of vision systems (4) are respectively installed on the lower bearing table (2) and the upper bearing table (3), which are used for shooting the standard patterns on the calibration plate (5); All the standard patterns on the same calibration plate (5) are the same, and each standard pattern is provided with a unique number which can be recognized by an image algorithm to obtain the position of the shot standard pattern in the calibration plate; or all the standard patterns on the same calibration plate (5) are different, and different patterns represent different positions, and the type, number and arrangement of the patterns need to be obtained by measurement and analysis; When used for measuring the repeated positioning accuracy, first, the lower bearing table is at the initial position, the vision system installed on the lower bearing table is aligned with the calibration plate, and the position of the complete standard pattern A in the camera field of view of the vision system is calculated, and the position difference between the center of the field of view and the standard pattern A is obtained; After the lower bearing table moves to the far end to complete the required process action, it returns to the initial position again from the far end, at this time, the vision system of the lower bearing table is aligned with the calibration plate again and takes a picture, and the position of the complete standard pattern B in the camera field of view of the vision system is calculated, and the position difference between the center of the field of view and the standard pattern B is obtained; According to the positions of the standard pattern A and the standard pattern B on the calibration plate, the distance between the center of the field of view of the first shot and the center of the field of view of the second shot in the X and Y directions can be obtained, and this distance is the repeated positioning error of the lower bearing table before and after movement; When used for monitoring the position change of the relative position adjustment between the upper wafer and the lower wafer, the two groups of vision systems of the upper bearing table or the lower bearing table shoot two calibration plates respectively, each group of vision systems will shoot the position change before and after movement, and then the X and Y displacement amounts of each group of vision systems are calculated, and finally the X and Y displacement amounts and the deflection angle of the upper bearing table or the lower bearing table are calculated.

2. The positioning method of claim 1, wherein: All the standard patterns on the calibration plate (5) are arranged in a matrix.

3. The positioning method of claim 2, wherein: The standard patterns on the two calibration plates (5) are the same.

4. The positioning method according to any one of claims 1 to 3, characterized in that: The calibration plate (5) is of transparent material.

5. The positioning method of claim 4, wherein: The fixed seat (7) is provided with a groove, and the calibration plate (5) is embedded or pasted in the groove.

6. The positioning method according to any one of claims 1-3, wherein the fixed seat (7) is fixed on a support (6), and the bottom of the support (6) is fixedly connected with a base (1).

Images