JPH0630332B2 - Positioning device and substrate positioning method using the device - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a mask, reticle, wafer, etc. positioning apparatus and method used in the manufacture of semiconductor devices such as ICs and LSIs, and more particularly to a wafer prober and a wafer scriber. The present invention relates to a suitable positioning device and a positioning method.

(Background of the Invention) A positioning device for a substrate, such as a wafer, in which a plurality of chip patterns are two-dimensionally arranged is disclosed in, for example, Japanese Patent Laid-Open No.
As disclosed in Japanese Patent Laid-Open No. 54648/1994, a spot of a laser beam is scanned on a wafer to detect photoelectrically scattered light from a pattern on the wafer, and waveform data of the photoelectric signal and a reference stored in advance. There is known an apparatus that detects a positional deviation of a wafer by comparing the waveform data (template) with the pattern matching by pattern matching. In this apparatus, the template is created by extracting from a predetermined local area on the first wafer, and the second and subsequent wafers are two-dimensionally positioned at the same position as the first wafer. By the way, when the wafer is rotated and placed on the apparatus for such positioning, the desired positioning accuracy cannot be obtained. Therefore, it is necessary to correct the rotation of the wafer by some method.
Conventionally, a wafer is placed on a two-dimensional moving stage via a rotatable wafer holder. Then, in order to correct the rotation of the wafer, the wafer holder is rotated, but some correction is required to correct this rotation.
It took time. Especially, in a simple apparatus, the rotational positioning accuracy of the wafer holder may be low, and the rotation correction time has a disadvantage in that it makes up a considerable proportion of the entire positioning time, thus reducing the throughput.

(Object of the Invention) An object of the present invention is to provide a positioning device that shortens the time for rotation correction of a substrate having a chip pattern such as a wafer or a mask, and improves accuracy, and a substrate positioning method by the device. To do.

(Summary of the Invention) The present invention is an apparatus for positioning a substrate (wafer, mask, etc.) in which a plurality of chips of a predetermined size are two-dimensionally arranged at a predetermined pitch along an orthogonal coordinate system xy. The first information according to the pattern in the first local area and the second information according to the pattern in the second local area that is separated from the first local area in the chip arrangement direction by a distance that is approximately an integer multiple of the pitch.
Pattern information extraction means for extracting the information, and a rotation error for detecting the rotation error with respect to the coordinate system xy of the substrate based on the correlation between the first information and the second information and rotating the substrate so as to correct the rotation error. A correction means, a means for detecting a residual rotation error of the rotation-corrected substrate, and a means for correcting and positioning the substrate in the x and y directions of the coordinate system xy in accordance with the residual rotation error are provided. It is a technical point.

Next, the configuration of the positioning device according to the embodiment of the present invention will be described with reference to FIGS. 1 and 2. The wafer 6 is placed on a stage 7 having a position sensor 12 such as a laser interferometer. The stage 7 is two-dimensionally moved by a drive mechanism 8 including a motor and the like. On the other hand, from the light source including the laser oscillator 1 and the cylindrical lens 2, a converged laser light flux having an elliptical cross section is emitted, and this laser light flux is bent by the reflection mirror 3 and is formed into an elongated slit-shaped spot on the surface of the wafer 6. Image as light 13. The laser light scattered by the uneven pattern on the surface of the wafer 6 reaches the photoelectric element 4. The photoelectric signal of the photoelectric element 4 is amplified by the amplifier 9 by a predetermined amount and applied to the next A / D converter (hereinafter referred to as ADC) 10. Therefore, ADC
The digital signal D ₂ output from 10 is proportional to the intensity or amount of scattered light generated from the wafer 6. In addition, a microscope 14 for observing the pattern on the wafer 6
Is provided to check if the aligned position is correct. The position sensor 12 is, for example, a spot light 1
The position signal D ₁ regarding the two-dimensional position of the stage 17 based on the irradiation position of _{No. 3} is generated, and the signal D of the ADC 10 is generated.
_It is input together with ₂ to the arithmetic processing unit 11. Arithmetic processing unit 11
Performs an operation required for alignment based on the signals D ₁ and D ₂ , and controls the drive mechanism 8 according to the result,
The wafer 6 is moved to a predetermined position.

For details of the arithmetic processing unit 11, see Japanese Patent Laid-Open No. 58-5.
Since it is disclosed in Japanese Patent No. 4648, the description thereof is omitted here, but scattered light generated from the wafer 6 is stored corresponding to the position, and matching between the previously stored reference pattern and the scattered light pattern is performed. Is used to detect a position with good matching.

By the way, the size of the spot light 13 in the longitudinal direction is the same as the size of one side of a standard chip when the same chip pattern is arranged periodically such as a wafer or a mask to be aligned. Make them almost equal. According to this definition, when the stage 7 shown in FIG. 1 is moved in the direction orthogonal to the longitudinal direction of the spot light 13, the photoelectric element 4 causes the fine pattern edge (concavo-convex) existing in the irradiation portion of the spot light 13. Since the scattered light generated in all directions from the step portion) is received almost all over, the photoelectric signal hardly changes even when the pattern changes minutely. This is the third
This will be described in more detail with reference to the drawings. This figure shows the relationship between two adjacent chips 50 on the wafer 6 and the spot light 13. As mentioned earlier, spot light 13
Moves in a direction orthogonal to the longitudinal direction with respect to the wafer 6. At this time, as shown in the figure, the movement is performed along the direction of the arrangement of the chips 50. Therefore, spot light 13
Moves while keeping its longitudinal direction parallel to a band width portion of about 50 to 100 μm, which is usually called a street line 53 provided between the chips 50. Further, inside the chip 50, an area 51 where an actual circuit pattern is formed and a plurality of bonding pads 52 located around the area 51 are provided. Now, assuming that the chip 50 size is 5 × 5 mm and the spot light 13 size is 5 mm × 50 μm, the spot light 1
When 3 illuminates the area 51, the circuit pattern has a two-dimensionally complicated shape, and scattered light is generated in all directions. Therefore, in the photoelectric signal of the photoelectric element 4, the spot light 13 is in the area 51.
The average value is high while moving inside. Also,
When the spot light 13 irradiates the vicinity of the spot 52, the width of the pad row in which the spots 52 are arranged is usually about 100 μm, so that the magnitude of the photoelectric signal fluctuates with the movement of the spot light 13. Further, when the spot light 13 irradiates the street line 53, the street line 53 has the characteristic of totally reflecting the normal light, that is, the characteristic of hardly scattering, so that the photoelectric signal has an extremely small value. The change state of the photoelectric signal described above is reproduced almost at any position on the wafer 6 as long as the longitudinal direction of the spot light 13 is longer than the length of one side of the chip 50. By making the spot light 13 elongated like this, the change of the pattern density on the wafer 6 (for example, the number of fine line patterns existing within the range irradiated by the spot light 13) is smoothed. Can be detected.

Further, as shown in FIG. 3, the spot light 13 moves along the arrangement direction of the chips 50 while keeping parallel to the street lines 53. At this time, the relationship between the spot light 13 and the wafer 16 is the fourth. It becomes like the figure. A plurality of chips 50 having the same pattern are formed on the wafer 16 in a matrix shape. One of the chips, that is, the chips 50 in the left-right direction on the paper surface is generally linearly cut out at a part of the periphery of the wafer 6. Lined up in parallel with the flat F. This flat F serves as a reference end for rough positioning of the wafer 6 (called pre-alignment). And
A rotation correction mechanism for mounting and rotating the wafer 6 is incorporated in the stage 7, and during the pre-alignment, the flat F is set parallel to one moving direction of the stage 7, for example, the x direction shown in FIG. As a result, the two moving directions x and y of the stage 7 which are orthogonal to each other and the two street lines 53 which are orthogonal to each other on the wafer 6 are parallel to each other. In this pre-aligned state, if the stage 6 is moved in the x direction or the y direction while irradiating the wafer 6 with the spot light 13, the surface of the wafer 6 is scanned by the spot light 13. become. At this time, as shown in FIG. 4, when the stage 7 is moved in the x direction, the spot light 13 has a slit shape elongated in the y direction, and when moved in the y direction, the slit light 13 is elongated in the x direction. State This is done, for example, by rotating the cylindrical silica lens 2 in the generatrix direction by 90 °.

Now, the wafer 6 must be placed on the coordinate system xy of the stage 7 without rotational deviation. Therefore, a principle explanation of the detection of the rotation deviation will be given with reference to FIGS. 5 and 6. FIG. 5 shows a state in which the wafer 6 is positioned in the coordinate system xy by substantially no pre-alignment, and further shows two chips 50 on the wafer 6 which are separated by a distance L in the x direction and are aligned in the y direction. . Here, the distance L is chip 5.
It is determined to be as large as possible by an integer multiple of 0 in the x-direction array pitch. First, the spot light 13 is formed into a slit shape extending in the x direction, and the stage 7 is arranged so that the spot light 13 overlaps the chip 50 on the left side of the wafer 6.
To position. At this time, the signal processing unit 11 moves to the stage 7
The coordinate value of is read from the position sensor 12 and stored. The values are set as positions x ₀ and y ₀ . Next, the signal processing unit 11 scans the stage 7 in the y direction by a distance longer than the array pitch of the chip 50 in the y direction, and extracts and stores the intensity distribution of scattered light accompanying the movement of the spot light 13. Figure 6 (a)
Is a diagram showing the intensity distribution A at this time, in which the vertical axis represents the scattered light intensity I and the horizontal axis represents the position in the y direction. Wafer 6
At the street line 53 extending in the x-direction between the chips 50 on the left side of the chip 50, the spot light 13 is considerably lower than the scattered light intensity when the chip 50 is scanning the chip 50. Therefore, the intensity I becomes the bottom (minimum value) at the position y ₁ during the scanning from the positions y ₀ to y ₂ . Next, the signal processing unit 11 returns the stage 7 to the coordinate value (x ₀ , y ₀ ), moves the stage 7 in the x direction by the distance L without changing the y coordinate value from that position, and the spot light 13
Is controlled so as to overlap with the chip 50 located on the right side of the wafer 6. Here, the stage 7 is again scanned to the position y _{2 in the} y direction, and the intensity distribution of scattered light is extracted. FIG. 6 (b) is a diagram showing the intensity distribution B at this time, and the vertical axis and the horizontal axis are the same as in FIG. 6 (a). Here, it can be seen that the street line 53 between the chips on the right side of the wafer 6 is at the position y ₃ between the positions y ₀ and y ₂ . Then, the signal processing unit 11 performs a correlation calculation between the two intensity distributions A and B, and a shift amount in the y direction between the intensity distributions A and B when the bottom of the waveform due to the street line 53 is matched, that is, y ₁ -Y ₃ is calculated.

As a method of detecting this deviation amount, for example, the intensity distribution of scattered light is extracted in advance from a certain distance (a value smaller than the array pitch) before and after including the street line 53, and the waveform of this distribution is stored as a reference pattern. , Intensity distribution A, B
For each of the above, the degree of correlation with the reference pattern may be calculated to obtain the positions y ₁ and y _3, and then the calculation of y ₁ -y ₃ may be performed. When the shift amount is obtained as described above, the rotation amount θ of the wafer 6 is approximated by the following equation (1).

θ = (y ₁ −y ₃ ) / L (1) Therefore, in principle, the wafer 6 should be rotationally corrected by this amount of rotation θ, but if only this method is used, Such problems arise. The first problem is that, even if the rotation amount θ is calculated as described above, if the accuracy of the mechanism for rotating the wafer 6 does not follow, the rotation amount of the wafer 6 is corrected again after the rotation correction of the wafer 6. This is because it is necessary to repeat the operation of detecting the rotation speed, and it may take a long time to correct the rotation. The second problem is that when the size of the chip 50 is extremely small with respect to the distance L, the stage 7 is moved to x
The point is that when the wafer 6 is sent by the distance L in the direction, the spot light 13 does not scan the left and right chips on the same row arranged in the x direction due to the rotation of the wafer 6. This second problem causes a significant error. That is, the wafer 6
Pattern matching is performed between the chips on the left side and the chips on the right side that are offset by an integer multiple (not including zero) of the array pitch in the y direction, and rotation is performed by an integer multiple of the array pitch with respect to the distance L. The error remains.

Therefore, an embodiment of the present invention for solving the second problem will be described with reference to FIG. FIG. 7 is an extreme representation of the wafer 6 rotating by θ with respect to the coordinate system xy, and the coordinate system αβ is the array coordinates of the chips on the wafer 6. First, the spot light 1 is set so that one chip on the right side of the wafer 6 and the street line in the x direction can be scanned.
Position 3a. Then, the stage 7 is scanned in the y direction for a predetermined distance to extract the data of the intensity distribution of scattered light,
By pattern matching with the reference pattern (template), the position y ₅ of the street line associated with the chip is detected. Next, the arrangement pitch (or the size of the chip) of the chip in the β-axis (y-axis) direction is CP, and the rotation error of the wafer 6 which can occur depending on the accuracy of pre-alignment is ε.
₀ , when a coefficient greater than ₀ and less than or equal to 0.5 is a, and an integer greater than or equal to 1 is n, the signal processing unit 11 calculates
The distance d that simultaneously satisfies (3) is calculated.

d <a · CP / ε ₀ (2) d = n · CP (3) The distance d is next to the spot light 13a of the spot light 13b that should be positioned to extract the intensity distribution of the scattered light. It represents the distance in the x direction. Thus, the distance d from the spot light 13a
After the spot light 13b is positioned at the position, the stage 7 is scanned in the y direction to extract the intensity distribution data of the scattered light. At this time, since the distance d is determined by the above equations (2) and (3), the street line existing in the scanning range of the spot light 13b is the same as the street line existing in the scanning range of the spot light 13a. Is. That is, the y-direction intensity distribution of scattered light is extracted at two positions separated by a distance d with respect to a plurality of chips arranged in a line along the α axis of the coordinate system αβ. Now, the position y ₆ of the street line is detected by pattern matching from the data extracted by scanning the spot light 13b.

Next, the signal processing unit 11 determines the shift amount between the positions y ₅ and y ₆ and the distance d.
The rotation amount θ ₀ is calculated by the equation (4) based on

θ ₀ = (y ₅ −y ₆ ) / d (4) The rotation amount θ ₀ includes an error because the distance d is shorter than the distance L on the wafer 6. Therefore,
The error amount (detection accuracy) in the y direction at the position where the rotation amount θ ₀ is detected at this distance d is set to δ ₀ as shown in FIG. In FIG. 8, the position of the spot light 13a in the x direction is P ₁ , and the position of the spot light 13b in the x direction is P ₂ . And apart from the position P ₁ to a distance L x-direction, and angular position P ₃ determined by the rotation amount theta ₀ with respect to the x-axis
When the spot light 13c is positioned at the position, the intensity distribution of the scattered light in the y direction is extracted, and matching with the reference pattern is performed, and the deviation amount between the intensity distribution and the reference pattern is obtained, the possible deviation amount is δ. _{If the} distance is ₁ , the distance L is expressed by the equation (5) from the geometrical relation as shown in FIG.

L = d · (δ ₁ / δ ₀ ) (5) Further, if the shift amount δ ₁ is also smaller than 1/2 of the array pitch (or chip size) CP in the y direction, then δ ₁ <a · CP ... Represented by (6).

Therefore, from the above equations (5) and (6), the signal processing unit 11
An optimum distance L that satisfies (7) and (8) at the same time is calculated.

L <(a · CP · d) / δ ₀ (7) L = n · CP (8) Next, the signal processing unit 11 moves the distance from the position P ₁ (the position of the spot light 13a) in the x direction. L only moves the stage 7, the amount of further determined by the amount of rotation theta ₀ in the y direction (L ·
It is stopped at the position P ₃ which is moved by θ ₀ ). Here, the spot light 13c is scanned in the y direction to extract the intensity distribution of the scattered light, and the position y ₇ of the street line attached to the chip in the y direction is detected by pattern matching. The street line detected here is the same as the street line detected by the scanning of the spot light 13a, which means that the scanning of the spot light 13a and 13c is performed on the left and right chips in the same row. Then, the signal processing unit 11 uses the following approximate expression (9) to accurately calculate the wafer 6 based on the position y ₅ previously obtained by scanning the spot light 13a and the position y ₇ obtained by scanning the spot light 13c. The rotation amount θ is calculated.

θ = (y ₅ −y ₇ ) / L (9) If the rotation amount θ is obtained by the above method, even if the size of the chip is very small with respect to the distance L, the left and right sides in the same row Since the pattern matching can be performed with respect to the chip, the second problem described above is solved.
The signal processing unit 11 only has to correct the rotation of the wafer 6 by the rotation amount θ. However, the rotation error Δθ cannot be corrected by only one rotation correction due to the accuracy of the rotation correction mechanism. Remains. It is possible to repeatedly correct the rotation error Δθ to drive the error Δθ to zero, but this causes the first problem described above. Therefore, the embodiment of the present invention which solves the first problem will be continuously described.

First, the rotation correction of the wafer 6 is performed only once according to the calculated rotation amount θ. Then, the signal processing unit 11 again sets x on the wafer 6.
A local portion including two chips located at a distance L (however, L = n.CP) in the direction and a street line associated with the chip is scanned by the spot light 13 in the y direction as shown in FIG. 5, for example. Then, the intensity distribution of scattered light generated from the local portion is extracted, and the positional deviation amount Δy in the y direction of both local portions is calculated by pattern matching. Then, from the relationship with the distance L, the residual rotation error Δθ is calculated by the following equation (1
0).

Δθ = Δy / L (10) Next, as shown in FIG.
The spot light 13d extending in the x-direction and the spot light 13e extending in the y-direction represent the local portion including the predetermined position P ₆ with respect to the two positions P ₄ and P ₅ for which y is obtained, respectively in the y-direction. Scanning is performed in the x direction, and a two-dimensional intensity distribution of scattered light generated from the local portion is extracted. The relative position of the position P ₆ with respect to the positions P ₄ and P ₅ is the same for all wafers to be aligned. Then, the signal processing unit 11 uses the data of the two-dimensional intensity distribution of scattered light extracted in advance from the local portion corresponding to the position P ₆ of the first wafer as a template and the template 6 and the wafer 6. Pattern matching is performed with the intensity distribution data of the local portion at the position P ₆ to obtain the amount of deviation from the template. If positioning the stage 7 so as to correct the deviation amount, the position corresponding to the position P ₆ of the first wafer, the position P ₆ of the second or subsequent wafer 6 Supotsuto light 13
Will be aligned with respect to. That is, the two-dimensional position of one wafer is reproduced for the second and subsequent wafers 6. Then, if the stage 7 is two-dimensionally moved with the chip at the reproduced position P ₆ as the origin, the desired chip on the wafer 6 can be positioned within the observation field of the microscope 14 or directly below the probe needle of the prober. However, the further away from the position P _{6, the} larger the positional deviation due to the residual rotation error Δθ. By the way, if the size (or array pitch) of each chip on the wafer 6 in the x direction and the y direction is α ₀ and β ₀ , for example, the array coordinate system αβ is set so that the chip center at the position P ₆ is the origin. When determined, the center of each chip is the coordinate system αβ, where m and n are integers (mα ₀ , n
It is distributed like a matrix at the coordinate value of β ₀ ). Therefore, when the position in the coordinate system αβ is converted into the position in the coordinate system xy in consideration of the residual rotation error Δθ, the following equation (1
It is represented by 1) and (12).

x = m · α ₀ · cos (Δθ) + n · β ₀ · sin (Δθ) …… (1
1) y = n ・ β ₀・ cos (Δθ) -m ・ α ₀・ sin (Δθ) …… (1
2) Since the error Δθ is extremely small, cos (Δθ) ≈ 1, sin
When (Δθ) ≈Δθ, the equations (11) and (12) are approximated to the equations (13) and (14), respectively.

x = m · α ₀ + n · β ₀ · Δθ (13) y = n · β ₀ −m · α ₀ · Δθ (14) Coordinate value (x determined by the formulas (13) and (14) ，
According to y), when positioning the stage 7 so that the chip at the position (m · α ₀ , n · β ₀ ) in the coordinate system αβ is positioned directly under the microscope 14 or the probe needle, On the other hand, the position of the stage 7 is (n ・ β ₀・ Δ
θ, −m · α ₀ · Δθ).

This state is shown in FIGS. 10 and 11. FIG. 10 shows a state when the stage 7 is positioned without correcting the residual rotation error Δθ, and FIG. 11 shows the residual rotation error Δθ.
Shows the state when the stage 7 is positioned by correcting the. The rectangles shown by the dotted lines in FIGS. 10 and 11 represent the regions 30 in which the probe needles are located, corresponding to each chip. When the stage 7 is positioned without correction, the chip C _{1 at} the position P ₆ is well aligned because it has only a rotation error of itself, but the chip C ₁ is separated from the chip C _{1 in the} x direction (α direction). With respect to C ₂ and C ₃ , the positional deviation from the region 30 gradually increases with respect to the error Δθ. However Yotsute to be corrected as FIG. 11, the center of each chip _{C 1,} C 2, _{C 3} is located on the x-axis, the position and the chip _{C 1,} C 2, _{C 3} and the region 30 At the time of alignment, only the relative rotation error between the chip and the region 30 remains, and the same alignment accuracy can be obtained for all the chips on the wafer 6.

(Effects of the Invention) As described above, according to the present invention, the rotation error of the substrate such as the wafer is detected, and then the rotation of the substrate is corrected only once, which is extremely high speed. Moreover, after the rotation correction, the residual rotation error of the substrate is detected again, and when positioning along the coordinate system xy, the position to be positioned is corrected in the xy direction according to the residual rotation error. It is possible to obtain the effect that the chips can be aligned with high accuracy equivalent to all the chips on the substrate.

Further, in the positioning method of the present invention, according to the method of detecting the positions of three chips on the substrate, when detecting the rotation error of the substrate by detecting the positions of the two chips which are largely separated on the substrate, The problem that the chips are not located in the same row is solved, and the rotation error of the substrate can always be detected accurately even when the chips are extremely small.

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram of an alignment apparatus suitable for an embodiment of the present invention, FIG. 2 is a perspective view of an optical system for forming spot light, and FIG. 3 is a plan view showing a relationship between spot light and a chip. Figure,
FIG. 4 is a plan view showing a two-dimensional relationship between a wafer and spot light, FIG. 5 is a view for explaining the principle of the present invention, and FIG.
FIG. 7 is a waveform diagram showing the intensity distribution of scattered light used for pattern matching, FIG. 7 is a diagram for explaining an embodiment of the present invention, FIG. 8 is a diagram for explaining a method for calculating the amount of rotation of a wafer,
FIG. 9 is a diagram for explaining an operation for alignment, and FIG.
FIG. 11 and FIG. 11 are views for explaining the operation of correcting the residual rotation error. [Explanation of Signs of Main Parts] 5 ... Laser beam, 6 ... Wafer, 7 ... Stage, 1
1 ...... signal processing unit, 12 ...... position sensor, 13 ...... Supotsuto _{light, 50, C 1, C 2} , C 3 ...... chips, 53 ...... street line

Claims (4)

[Claims] 1. A rectangular coordinate system xy on which a substrate in which a plurality of rectangular chips are two-dimensionally arranged at a predetermined pitch is placed.
An apparatus for positioning an arbitrary chip on the substrate at a predetermined position in the Cartesian coordinate system xy, which has a stage that moves two-dimensionally along a detection center at a predetermined position in the Cartesian coordinate system xy. And generating a first signal according to a pattern in the first local region by photoelectrically scanning a first local region including a characteristic pattern shape formed on the substrate, By photoelectrically scanning the second local region, which is separated from the first local region by a distance that is an integer multiple of the pitch in the array direction of the chips, the second local region corresponding to the pattern in the second local region is photoelectrically scanned. Pattern information generating means for generating a signal; computing means for calculating an overall rotation error of the substrate in the rectangular coordinate system xy based on the correlation between the first signal and the second signal; The calculated whole A rotation mechanism for rotating the substrate on the stage in order to correct various rotation errors; and, after correction of the overall rotation error of the substrate, obtained from each of two local regions on the substrate. Residual difference calculating means for calculating the residual error of the entire rotation error of the substrate based on the correlation of two signals; and when sequentially positioning any chip on the substrate at the predetermined position, A positioning device provided with a control means for correcting and moving the stage in the x-direction and the y-direction of the Cartesian coordinate system xy according to the residual error of the overall rotation error. 2. The pattern information generating means is a portion of the periodic pattern structure in the arrangement direction of a plurality of chips on the substrate, which generates a characteristic light intensity distribution by photoelectric scanning, Set as the second local area,
The apparatus according to claim 1, further comprising a photoelectric element that detects a light intensity distribution of each of the first and second local regions. 3. A stage for mounting a substrate for manufacturing a semiconductor device, in which a plurality of rectangular chips are arranged two-dimensionally at a substantially constant pitch, and moving two-dimensionally along an orthogonal coordinate system xy. , A rotation mechanism for rotating the substrate on the stage, and a local range including a characteristic pattern shape formed on a part of the substrate, having a detection center at a predetermined position on the orthogonal coordinate system xy. , The x direction of the coordinate system xy, or y
A photoelectric detection unit that outputs a photoelectric signal whose intensity changes in accordance with the pattern shape when scanning in the direction, and an arithmetic unit that calculates the position of the chip on the substrate with respect to the orthogonal coordinate system xy based on the photoelectric signal. And a substrate positioning method using a positioning device including a control unit that controls driving of the stage based on the calculated value, the position of a chip at a characteristic position on the substrate being detected by the photoelectric detection. Means for detecting a plurality of chips using the means and the calculating means; a step of obtaining an overall rotation error in the θ direction of the substrate from each position of the detected plurality of chips; Rotating the substrate in the θ direction by the rotating mechanism to correct the rotation error in the θ direction as a whole; When sequentially positioning at predetermined positions on the Cartesian coordinate system xy, the control means corrects the amount of movement in the xy directions according to the residual error of the rotation error of the substrate in the θ direction as a whole. A positioning method characterized by the above. 4. A rectangular coordinate system xy on which a substrate in which a plurality of rectangular chips are two-dimensionally arranged at a predetermined pitch is placed.
A stage that moves two-dimensionally along the axis, a rotation mechanism that rotates the substrate on the stage, a detection center at a predetermined position on the Cartesian coordinate system xy, and a characteristic pattern formed on the substrate When scanning a local area that includes a shape,
A photoelectric detection unit that outputs a light intensity distribution that changes according to the pattern shape, a calculation unit that calculates the position of the chip on the Cartesian coordinate system xy based on the light intensity distribution, and a calculated position. A method of positioning a substrate using a positioning device having a control means for controlling the driving of the stage based on a chip group arranged in the x direction on the substrate, the chip group being located around the substrate. Designating a first chip and detecting a first light intensity distribution according to the pattern shape in the y direction within a first local range including a part of the first chip by the photoelectric detection means; Within a second local range that includes a part of the second chip and specifies a second chip that is located on the center side of the substrate and is separated from the first chip by a distance d in the x direction. In the y direction Detecting a second light intensity distribution according to a turn shape by the photoelectric detection means; and based on the first light intensity distribution and the second light intensity distribution, the first chip and the second chip The respective positions in the y direction are obtained, and the coordinate system xy is calculated based on the respective positions and the distance d.
Calculating a rotation amount θ ₀ of the array axis of the one row of chip groups with respect to the coordinate axis of; and a distance L larger than the distance d in the one row of chip groups based on the calculated rotation amount θ _0. Third away
Designating a chip and detecting a third light intensity distribution according to the pattern shape in the y direction within a third local range including a part of the third chip by the photoelectric detection means; The y-direction position of the third chip is obtained based on the light intensity distribution, and the orthogonal coordinate system xy of the substrate is calculated based on the difference between the position and the y-direction position of the first chip and the distance L. And a step of calculating an accurate rotation error within the positioning method.