JP3442480B2 - Wafer peeling device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a wafer stripping apparatus for stripping a wafer adhered on a flat polishing plate for polishing a semiconductor wafer (hereinafter referred to as a wafer) after the polishing operation.

[0002]

2. Description of the Related Art As a conventional wafer stripping apparatus, the wafer stripping apparatus disclosed in Japanese Patent Laid-Open No. 5-166771 is known as an example. This wafer stripping device is a chuck that holds a polishing plate to which a wafer is attached, a scraper that is provided above the chuck and that is movable along the polishing plate surface to strip the wafer from the polishing plate,
A moving means for tilting the support surface of the chuck holding the polishing plate from the horizontal position in the cassette direction so that the wafer, which is separated by the scraper, slides into the cassette, has a cassette storage portion provided below the support surface of the chuck. It is equipped with. Further, this wafer peeling apparatus is provided with a reflection type sensor or the like at an appropriate position in order to detect the bonding position and the number of wafers bonded on the polishing plate.

[0003]

However, the above-mentioned wafer stripping apparatus has the following problems. In the reflection type sensor, for example, the chuck is rotated or the reflection type sensor itself is rotated to scan the polishing plate in a circular shape, and the presence or absence of the wafer on the plate is determined from the difference in the reflected light amount between the polishing plate and the wafer. It is only detected, and it cannot be detected that defects such as chipping occur in areas other than the scanned portion of the wafer. In addition, the notch or orientation flat provided on the wafer may not be detected. Further, after the wafer adhered on the polishing plate is subjected to polishing processing, the surface of the wafer is cleaned in the cleaning step and is carried to the wafer peeling device by an appropriate transport mechanism, so that water drops on the polishing plate or the wafer are removed. May be attached. In this case, the light output from the reflection type sensor may hit the water droplets and the light may be scattered so that it may not be possible to accurately detect the wafer attachment position and the number of wafers.
In addition, the polishing plate is made of glass or ceramic material, and the reflection amount of the light output from the reflective sensor changes depending on the material of the plate. There is a problem that it is necessary to change the processing method of the signal from the sensor.
Therefore, the present invention has been made to solve the above problems, and its purpose is to detect damage such as chipping of the wafer, change the material of the polishing plate, or even if water droplets are attached to the wafer. It is an object of the present invention to provide a wafer peeling device that can accurately detect and can perform peeling automatically.

[0004]

The present invention has the following constitution in order to achieve the above object. That is, the wafer peeling apparatus according to the present invention includes a rotary table on which a polishing plate having a wafer adhered on its upper surface can be mounted, and a peeling unit for peeling the wafer from the polishing plate mounted on the rotary table. A camera arranged above the rotary table to receive light from the polishing plate and the wafer and generate voltage data for each pixel forming an image in the field of view; and input voltage data for each pixel The de
Conversion means for converting and outputting digital data, and the conversion
Digital data for each pixel output from the means,
On a logical plane that is an XY orthogonal coordinate system corresponding to the field of view
Image to be stored as image data showing the pixel position in
Image storage means and the wafer to the polishing plate
Includes at least one of shape, number of sticking, and sticking position
Multiple wafer sticking specifications are stored and the sticking finish
Wafer peeling conditions corresponding to each of the
Peeling condition storage means and before outputting from the conversion means
The digital data is stored in the image storage means,
The image data stored in the image storage means is processed to apply the sticking.
The specifications are calculated and obtained, and the obtained adhesion specification and the peeling strip are obtained.
Comparing the pasting specifications stored in the case storage means
The sticking specification is specified, and the above-mentioned corresponding sticking specification is specified.
Rotate the rotary table based on the wafer peeling conditions.
And the peeling means are controlled to rotate the polishing plate.
Control means for peeling the wafer from the wafer
In the wafer peeling device, the polishing plate is disc-shaped
And the polishing plate is pre-determined
According to one of the plurality of the pasting specifications,
The wafer should be wrapped around at least the periphery of the polishing plate.
The control means is attached along the direction.
Each wafer has the same center as the center of the polishing plate.
A circular window of a given diameter that includes a partial area of
Before each wafer is included in the circular window
The center of gravity of the area is calculated, and the center of the polishing plate
From the position of the center of gravity, the number of the wafers to be pasted or the pasting
By calculating at least one of the landing positions
Therefore, the sticking specification is obtained.

Further , in the control means, the position of the center of gravity is
Near the center of the polishing plate and in the area
If the area is larger than the reference area,
It is characterized by judging that there is a wafer in the mind.

Further , the control means controls the calculated center of gravity.
Based on position, any wafer can be used as a reference for other wafers.
The angle interval between the
Reference angle between wafers calculated from the number of wafers attached
Do the work of calculating the error from the interval and calculating this error.
All wafers are used as a reference, and the maximum error is the smallest.
When the temperature becomes smaller, the standard wafer is stripped first.
Decide on the reference wafer and start peeling from the reference wafer
Then, rotate the rotary table by the reference angle intervals.
And control so that the wafers are sequentially peeled off.
To do. Further, the control means is the center of the polishing plate.
If a wafer is present in the
It is characterized by controlling to be separated.

Further , the rotary table is provided above the rotary table.
Block external light from entering the camera to cover the cable.
An external cover is provided for
Arrange lighting means to illuminate the rotating table with diffused light
Then, the surface of water droplets adhering to the rotating table or wafer
Is only illuminated by diffused light, and all the reflected light on the surface of the water droplet is
It is almost uniform over the surface, and the water droplets are colorless and transparent.
Then, you can erase the water drop image from the image captured by the camera.
Wafers can be identified without being affected by water droplets.
It

The illuminating means is the same as the first illuminating means.
The first illuminating means is composed of two second illuminating means.
Is the reverse of the first illuminating means reflected on the turntable.
It is arranged at a position where the incident light directly enters the camera,
The illuminating means of the second illuminating means is a
Place the light reflected by the light means at a position where it does not enter the camera directly.
Then, even if the material of the polishing plate changes,
Good wafer by utilizing the difference in surface reflectance between the wafer and the wafer.
-Ha can be identified.

[0009]

The camera catches the polishing plate and the wafer adhered on the polishing plate, and the image processing means performs image processing using the voltage data for each pixel input from the camera to obtain the shape of the wafer on the polishing plate. , Calculating the sticking specification including at least one of the number of sticking and the sticking position, and rotating the rotary table based on the peeling condition which is determined and stored in advance corresponding to the sticking specification. By controlling the peeling means to peel the wafer from the polishing plate, the peeling work can be automatically performed. Further, since the wafer can be captured as image data, the presence or absence of damage such as chipping of the wafer can be detected.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of a wafer peeling apparatus according to the present invention will be described below with reference to the accompanying drawings. First,
The mechanical configuration of the apparatus of this embodiment will be described with reference to FIGS. Reference numeral 10 denotes a rotary table, on which a polishing plate 14 having a wafer 12 adhered on its upper surface can be placed. Further, the rotary table 10 is provided with a suction function of sucking and fixing the mounted polishing plate 14.
The upper surface of the turntable 10 is painted in a color with low lightness such as black or navy blue.

Reference numeral 16 denotes a peeling means, which is a rotary table 10.
The wafer 12 has a function of peeling the wafer 12 from the polishing plate 14 placed on the wafer and transporting the wafer 12 toward the wafer storage unit 18. The peeling means 16 includes a guide rail 20 arranged on the side of the rotary table 10, a slide body 22 slidably arranged on the guide rail 20, and a slide body 22 along the guide rail 20 in the direction of arrow A ( A slide mechanism 24 that slides in the left-right direction in FIG. 1, a motor 26 that drives the slide mechanism 24, an arm portion 28 that extends from the slide body 22 toward the rotary table 10, and a scraper attached to the arm portion 28. And 30. In addition, the slide body 22 has a built-in first cylinder device 32 that drives the arm portion 28 in the direction of arrow B (vertical direction in FIG. 1). For the slide mechanism 24, for example, a ball screw or an endless chain is generally used. The wafer 12 peeled from the rotary table 10 by the scraper 30 of the peeling means 16 is conveyed toward the wafer storage portion 18 by the scraper 30 and stored in the wafer storage portion 18 via a shooter 33 provided between the wafer storage portion 18 and the wafer storage portion 18. It is stored in a cassette (not shown) provided in the section 18. The wafer 12 is attached to the shooter 33.
There is provided a passage sensor 35 including an optical sensor or the like that detects the passage of the signal and outputs a passage signal to the control means described later.

A base 34 supports the rotary table 10, the guide rails 20 and the like. Reference numeral 36 is a mounting plate to which the rotary table 10 and the guide rail 20 are mounted. The mounting plate 36 is arranged inside the base 34,
One end is pivotally attached to the base 34, and the other end is also the base 3
4 is attached to a second cylinder device 38 provided inside. Therefore, the second cylinder device 38 operates and expands and contracts in the direction of arrow C, whereby the mounting plate 36 rotates within a predetermined angle range, and accordingly, the rotary table 10 and the guide rail 20 also move in the direction of arrow D. Can be rotated within the angle range (about 3 to 10 degrees). In the present embodiment, the guide rail 20 is attached to the mounting plate 36 via an appropriate support and rotates integrally with the mounting plate 36 in the arrow D direction. It may be fixed in advance at a predetermined angle. In addition, together with the upper surface of the turntable 10, each of the cameras 42, 4
The upper surface of the mounting plate 36 and the upper surface of the base 34, which are included in the visual fields of Nos. 4 and 46, are painted in advance with a color having low lightness (black as an example). Reference numeral 40 denotes an external cover arranged above the rotary table 10 of the wafer peeling device so as to cover the rotary table 10. The external cover 40 is for blocking the entrance of external light into a camera described later. is there.

Next, the structure of the processing system of the apparatus is shown in FIG.
~ It demonstrates using FIG. Reference numeral 42 denotes a whole camera, which is attached to the central portion of the upper portion of the outer cover 40.
The field of view of the whole camera 42 is set so as to include the entire turntable 10. Reference numeral 44 denotes a central camera, which is attached to an upper central portion inside the outer cover 40 adjacent to the whole camera 42. The central camera 44 is capable of enlarging and capturing an image as compared with the whole camera 42, and its field of view is set so as to capture only a specific portion at the center of the turntable 10. An outer edge camera 46 is attached to the lower side of the outer cover 40. The field of view of the outer edge camera 46 is set so as to magnify and capture only a specific portion of the peripheral edge of the turntable 10. Each of the cameras 42, 44, and 46 has a rotary table 1
0 is arranged above the polishing plate 14 and the wafer 12
It has a function of receiving the light from and generating voltage data for each pixel forming an image in the visual field. The voltage data increases or decreases according to the brightness.

Reference numeral 47 denotes an image processing means, which performs image processing using the input voltage data for each pixel, and at least one of the shape of the wafer 12 on the polishing plate 14, the number of wafers to be stuck, and the sticking position. Is calculated and obtained, and the rotary table 10 is rotated based on the peeling condition determined and stored in advance corresponding to the sticking specification, and the peeling means 16 is controlled to move the wafer from the polishing plate 14 to the wafer. It has a function of peeling 12. Further, the image processing means 47 will be described in detail. Reference numeral 48 is a conversion means, which has a function of converting the voltage data of each pixel into digital data and outputting the digital data. In the present embodiment, a comparator is used as an example, and has a function of binarizing voltage data based on a predetermined level. An image storage unit 50 has a function of storing the digital data for each pixel output from the conversion unit 48 as image data indicating the position of the pixel on the logical plane which is the XY orthogonal coordinate system corresponding to the visual field. Have. The image storage means 50 is usually composed of a storage element such as a RAM. 52 is a peeling condition storage means,
A plurality of sticking specifications of the wafer 12 to the polishing plate 14 are stored in advance, and a plurality of peeling conditions of the wafer 12 corresponding to each of the sticking specifications are also stored in advance. Usually, it is composed of RAM, ROM and the like.

Explaining the sticking specification, in this embodiment, as an example, the sticking specification includes the shape of the wafer 12, the number of sticking to the polishing plate 14, the sticking position and the like. Wafer 12
The pattern is determined in advance depending on the shape, that is, the size (inch size), and when the wafer 12 is 5 or 6 inches, up to 6 to 10 sheets are attached. At this time, one piece is attached to the center of the polishing plate 14 and the remaining wafers 12 are attached to the periphery of the polishing plate 14 at equal angular intervals on the same circumference having a predetermined radius. In the case of an 8-inch wafer, 3 to 4 wafers are attached to the polishing plate 14 on the same circumference having a predetermined radius at equal angular intervals. Therefore, the reference angle interval of the wafer 12 is obtained in advance for each size of the wafer 12 and the number of wafers to be bonded, and the wafer 1 to be bonded is
If the reference angle interval of 2 is stored in the peeling condition storage means 52 as the attachment position, the image processing is performed by the control means described later, and the angular interval of each wafer 12 on the polishing plate 14 is obtained to obtain the wafer 12 Can be specified. Further, the control procedure of the rotary table 10 and the peeling means 16 to be performed at the time of the peeling work is also performed for each sticking specification.
The peeling conditions of No. 2 are stored in association with each other in advance.

Reference numeral 54 is a control means, which stores the digital data output from the conversion means 48 in the image storage means 50, processes the image data stored in the image storage means 50, and attaches the image data (for example, the wafer 12). Shape, number of stickers,
(Adhesion position etc.) is calculated, and the obtained adhesion specification and the plurality of adhesion specifications stored in the peeling condition storage means 52 are compared to specify the adhesion specification and correspond to the specified adhesion specification. It has the functions of rotating the rotary table 10 and controlling the peeling means 16 to peel the wafer 12 from the polishing plate 14 based on the wafer peeling conditions. Further, it also controls the turning on / off of the lighting means described later. The control means 54 is usually composed of a microcomputer. For example, when performing image processing and peeling work at higher speed, a microcomputer is used for image processing and peeling work control. Alternatively, a sequencer may be used to control the peeling work, and a series of operations may be performed by receiving a command from a microcomputer that performs image processing. Further, 55 is a ROM in which the operation program of the microcomputer is stored in advance. Further, in the ROM 55, diameters and positions of various circular windows described later, reference areas of regions included in the circular windows for each wafer size, reference distances of the center of gravity, and the like are calculated and stored in advance as basic data. .

The structure of the outer cover 40 will be described in more detail. The outer cover 40 is formed in a box shape whose lower surface is open. Further, the inner surface of the outer cover 40 is formed in white. However, the inner surface of the portion in which the cameras 42, 44 and 46 are incorporated is formed in black in order to prevent internal reflection of light. In addition, in the outer cover 40,
A cover for indirect lighting (hereinafter, simply referred to as "lighting cover") 56, which is formed in the shape of a hollow quadrangular pyramid whose lower surface is open using a resin material such as translucent white plastic, is arranged with its apex above. There is. Further, the lower surface opening portion of the lighting cover 56 and the lower surface opening portion of the outer cover 40 are formed in substantially the same shape, and when the lighting cover 56 is housed in the outer cover 40, the lower end edges of the lighting cover 56 come into contact with and closely contact with each other. The resin material forming the illumination cover 56 has a function of diffusing transmitted light and outputting it as substantially uniform light. The whole camera 42 is arranged near and above the apex of the lighting cover 56, and the central camera 44 is arranged above and adjacent to the whole camera 42. The outer edge camera 46 is the lighting cover 5.
6 are arranged at the lower end portions of the respective six. A half mirror 58 is attached to a portion of the illumination cover 56 located below the cameras 42, 44, 46.
By installing the half mirror 58, each camera 42,
44 and 46 are capable of capturing an image on the rotary table 10, and when the polishing plate 14 is placed on the rotary table 10, the camera itself is not reflected on the wafer 12 or the polishing plate 14. To prevent.

Further, reference numeral 60 denotes a first lamp as a first illuminating means which constitutes a part of the illuminating means, and is in a space sandwiched between the outer cover 40 and the illuminating cover 56, and the outer cover 40. Are arranged in the four corners above. Further, reference numeral 62 is a second lamp as a second illuminating means which constitutes a part of the illuminating means, and is in a space sandwiched between the outer cover 40 and the illuminating cover 56, and the outer edge camera 46 is attached. Is attached adjacent to the other two lower ends except the lower end portion and the lower end portion facing the lower end portion. Further, immediately above the second lamp 62, eaves portions 64 are provided so that the light of the second lamp 62 is emitted only downward.

By arranging the first lamp 60 and the second lamp 62 in this way, when only the first lamp 60 is turned on, the light emitted by the first lamp 60 is directly or externally covered. The light is reflected by the white inner surface of 40 and hits the lighting cover 56, and the light is diffused by passing through the lighting cover 56 and is irradiated onto the turntable 10. Since this diffused light irradiates the rotary table 10 as a whole from a relatively high position of the illumination cover 56, the reflected light on the rotary table 10 (the polishing plate 14 and the wafer 12) is directly transmitted through the half mirror 58. It is input to the whole camera 42 and the central camera 44. Further, when only the second lamp 62 is turned on, the light emitted by the second lamp 62 is
The light directly or reflected on the white inner surface of the outer cover 40 hits the lower end portion of the lighting cover 56, and the light thereof is emitted.
The light is diffused by passing through and is irradiated onto the turntable 10. This diffused light irradiates the rotary table 10 from a lower position of the illumination cover 56, so that the incident angle on the rotary table 10 becomes large, and the reflected light of the second lamp 62 reflected on the rotary table 10 is at the center of the entire camera 42. The light does not enter the camera 44 directly. Further, since the illumination cover 56 diffuses the light of the first lamp 60 and the second lamp 62 onto the rotary table 10 and irradiates it from the surroundings, water droplets adhere to the polishing plate 14 and the wafer 12. Even in such a case, there is an effect that it becomes difficult to see the water droplets by illuminating the water droplets with substantially uniform light.

Next, the operation of the wafer peeling device will be described. First, depending on the material of the polishing plate 14,
The operation of switching between the first lamp 60 and the second lamp 62 will be described with reference to FIGS. The material of the polishing plate 14 is roughly classified into ground glass and ceramic. When the polishing plate 14 is made of frosted glass, the operator turns on only the first lamp 60 in advance via the control means 54. The ground plate 14 made of frosted glass is in a transparent state because water adheres to the surface of the polishing plate 14 when it is conveyed from the dipping section 66. Therefore, the diffused light emitted from the upper part of the illumination cover 56 is reflected by the wafer 12, and the reflected light is input to the entire camera 42 and the central camera 44, but is irradiated because the polishing plate 14 is transparent. The light that has passed through the polishing plate 14 reaches the upper surface of the rotary table 10 and is reflected by the upper surface of the rotary table 10. The reflected light passes through the polishing plate 14 and is input to the entire camera 42 and the central camera 44. Since the upper surface of the rotary table 10, the upper surface of the mounting plate 36, the upper surface of the base 34, etc. included in the fields of view of the cameras 42, 44, 46 are pre-painted in a low lightness color (black as an example), the whole camera An image captured by the camera 42 and the central camera 44 has a low lightness in the polishing plate 14 portion and a high lightness in the wafer 12 portion due to reflection of diffused light, as shown in FIG.

When the polishing plate is made of ceramics, the operator preliminarily controls the second lamp 62 only.
Turn it on via 4. The reason why only the second lamp 62 is turned on and the first lamp 60 is turned off is the first
When the lamp 60 is turned on, the polishing plate 14 is made of ceramic and does not become transparent even if moisture adheres to the surface, and since the surface has fine irregularities, light is scattered by the irregularities. , An image with high brightness is obtained similarly to the wafer 12. The amount of light reflected by the polishing plate 14 is not as great as that of the wafer 12 whose surface is mirror-polished, but the difference is small,
Since there is little difference in the voltage data output from the cameras 42, 44, 46, it becomes difficult to perform binarization in the conversion means 48.

Therefore, if only the second lamp 62 is turned on, the incident angle of the diffused light emitted from the second lamp 62 onto the turntable 10 is large, so that the wafer 12 is turned on.
Light reflected from the surface of the camera does not enter the cameras 42, 44 and 46 directly. On the contrary, the light irradiated on the polishing plate 14 is scattered by the unevenness of the surface of the polishing plate 14, and the scattered light is input to the cameras 42, 44, 46, so that the camera 42,
In the images captured by 44 and 46, the brightness of the polishing plate 14 portion is high and the brightness of the wafer 12 is low as shown in FIG. As a result, a difference can be provided in the voltage data between the polishing plate 14 and the wafer 12 output from the cameras 42, 44, and 46, and the binarization in the conversion means 48 can be surely performed.

Next, the peeling operation of the wafer peeling device will be described. Note that the polishing plate 14 is made of frosted glass as an example. In the wafer stripping apparatus, first, the control means 54, when the polishing plate 14 is conveyed onto the rotary table 10, activates the suction function provided on the rotary table 10 to suction and fix the polishing plate 14, and then the second Of the mounting plate 36 by operating the cylinder device 38 of
1 is raised and the mounting plate 36 is tilted 5 degrees (step 100).

Next, the control means 54 performs the work of detecting the size and the number of the wafers 12 attached to the polishing plate 14. The detailed flow of this work will be described below. First, the control means 54 operates the whole camera 42 to take in the image on the polishing plate 14 as digital data via the conversion means 48, which is an XY orthogonal coordinate system. It is stored in the image storage means 50 as image data indicating the position of the pixel on the logical plane (step 102). In this embodiment, as an example, the value of the voltage data output from each of the cameras 42, 44, and 46 is high in the bright part of the image and low in the dark part of the image, and the value of the voltage data is output from the converting means 48. Digital data is output in which the high part is High level and the low part is Low level.
The polarities of data and signals between each component may be appropriately determined.

Subsequently, the control means 54 reads the basic data from the ROM 55 and, as shown in FIG.
A circular window centered at the center and having a predetermined diameter is set in the image storage means 50, and the area of each High level portion (hatched portion) included in the circular window is measured and the barycentric coordinates are calculated ( Step 10
4).

In step 104, even if the circular window is not set, the control means 54 measures the area of the high level portion from the entire image data stored in the image storage means 50 to obtain the barycentric coordinates. Good, but when the inch size of the wafer 12 is large and the gap between the adjacent wafers 12 is small, the image data in which the high level portions of the adjacent wafers 12 are continuous due to the influence of the noise and the stain on the polishing plate 14 described above. In some cases, accurate measurement cannot be performed. For this purpose, the circular window is set, and the polishing plate 1 for the wafer 12 is attached along the circumferential direction.
If the image data of only the end portion in the center direction of 4 is used as the image data for measuring the center of gravity, as shown in FIG. 8, the interval between the high level portions (hatched portions) corresponding to each wafer 12 increases, and some noise is generated. High due to dirt and dirt
It is better because the level parts do not continue.

The control means 54 measures the measured area and the distance between the center of the polishing plate 14 and the coordinates of the center of gravity.
It is determined whether or not it is within a predetermined range centered on a reference area or reference distance that is predetermined and stored in the ROM 55. If it is within the range, the hatched portion in FIG.
Certified as part of 2. In addition, the center of gravity is the polishing plate 1.
4 It is judged that there is a wafer at the center of the polishing plate 14 in the High level portion near the center and larger than the reference area (step 106). As described above, by using the reference area and the reference distance to judge whether or not the High level portion is a part of the wafer 12, noise jumping into the data transmission path of the processing system and stains on the polishing plate 14 are detected. It is possible to reliably determine that the portion erroneously stored as high-level image data in the image storage means 50 is not a wafer due to such contamination.

The reason why the wafer can be recognized by the reference area and the reference distance is that the wafer is classified by its size (5, 6, 8).
Since the wafers are attached to the polishing plate 14 at predetermined angular intervals on the same circumference and at the center position of the polishing plate 14, the diameter of the circular window is constant. If so, the area of the portion of the wafer bonded along the circumferential direction included in the circular window becomes substantially constant for each wafer size even if the number of wafers changes, and this can be used as the reference area. The same applies to the reference distance. In addition, it is possible to judge whether it is 5, 6 inches or 8 inches according to each reference area. Further, as one of the data of the reference area, the entire area of the 5 or 6 inch wafer is stored in the ROM 55 as basic data. If so, it is possible to determine whether the wafer is 5 inches or 6 inches by measuring the area of the wafer attached to the central portion included in the circular window.

The control means 54 obtains the number of high level portions on which the calculation of the barycentric coordinates is based, and determines the polishing plate 14.
Determine the number of wafers stuck on top. The reference angle interval is determined by referring to the basic number data stored in the peeling condition storage means 52 based on the calculated number of sheets (step 1).
08). Based on this reference angular interval, the control means 54 rotates the rotary table 10 for the purpose of shortening the processing time of the peeling work when performing the peeling work. In addition,
The reference angle interval may be obtained by dividing 360 degrees by the number of wafers attached to the polishing plate 14 along the circumferential direction. Further, first, an angular interval about the center of the polishing plate 14 of each center of gravity located on substantially the same circumference is calculated, and the reference included in the data of this angular interval and the basic data stored in the peeling condition storage means 52. The number of wafers to be adhered can also be determined by comparing the angular interval. Also, each H
As shown in FIG. 9, the angular interval between the centers of gravity of the high level portions is such that the center of gravity L of the high level portions is the polishing plate 14
Since it is on a straight line connecting the center M of each wafer 12 and the center N of each wafer 12, it substantially coincides with the angular interval of each wafer 12.
In addition, the polishing plate 1 is related to the reference area and the reference distance.
It is also possible to determine the wafer size of the wafer 12 attached to No. 4.

Next, in the control means 54, step 104
Based on the data obtained in step 1, a reference wafer to be subjected to the peeling work is first determined (step 110). Ideally, the angular spacing of each wafer 12 is a reference angular spacing, and when each wafer 12 is adhered at the standard angular spacing, the first wafer is stripped regardless of which wafer is the standard wafer. After the rotary table 10 is rotated and moved to the position (position E in FIG. 11), the peeling operation by the peeling means 16 and the reference angle of the rotary table 10 are sequentially performed based on the peeling conditions stored in the peeling condition storage means 52. By alternately performing the rotation operation for each interval, it is possible to arrange the wafer to be next peeled so that its center is always at the peeling position E.

However, in actuality, the wafer 12 is attached with some angular error, and in order to improve the processing speed of the peeling work, the wafer 12 to be peeled first is moved to the peeling position E and the peeling work is carried out. Since the rotary table 10 is rotated at the reference angle interval after the start of, the center of the second and subsequent wafers 12 may be displaced from the peeling position E.
It is necessary to reduce this deviation as much as possible. To this end, as shown in FIG. 10, the angle of the adjacent wafer 12 and the angle between the adjacent wafer 12 and the adjacent wafer 12 are sequentially determined with reference to one wafer 12 attached in the circumferential direction. , The calculated angle is compared with an integer multiple (1 or more) of the reference angle interval, and the operation of calculating the error is performed for all the wafers 12, and the reference value when the maximum error is the smallest The wafer may be used as the reference wafer.

As an example, as shown in FIG. 10, the calculation is performed for the case of five wafers (12a to 12e).
Here, the reference angle interval is 360/5 = 72 degrees. First, when the wafer 12a is used as a reference wafer, (1) the angle between the wafer 12a and the wafer 12b and 72
The degrees are compared and the error is calculated. (2) Angle between wafer 12a and wafer 12c and 14
4 (= 72 × 2) degrees are compared, and the error is calculated. (3) Angle between wafer 12a and wafer 12d and 21
The 6 (= 72 × 3) degrees are compared, and the error is calculated. (4) Angle between wafer 12a and wafer 12e and 28
8 (= 72 × 4) degrees are compared, and the error is calculated. The same applies to the error between the counterclockwise angle and 72 degrees. (5) In the above (1) to (4), one wafer is used as a reference.
2b, 12c, 12d and 12e are sequentially changed. (6) Among the errors obtained in (1) to (5) above, the reference wafer having the smallest maximum value is used as the reference wafer.

Then, as shown in FIG. 11, the control means 54 controls the wafers 12 corresponding to the size of the wafers 12 and the number of attached wafers.
A circular window having a predetermined diameter whose center is substantially the center is set in the image storage means 50, and the area of each High level portion (hatched portion) included in the circular window is measured to be a predetermined area or more. If it is, it is determined that there is a wafer. As a result, the wafer is reconfirmed (step 112). Further, at this time, if the diameter of the circular window is made substantially the same in correspondence with the size of the wafer, if the measured area does not reach the predetermined area, the wafer is cracked or damaged. You can also confirm that.

Subsequently, the control means 54 moves the reference wafer 12 to the peeling position E by rotating the rotary table 10 while capturing an image with the whole camera 42 as shown in FIG. 11 (step 114). Then, the peeling means 16 is controlled to peel the wafer 12 at the peeling position E. The separated wafer 12 is transferred to the wafer storage unit 1 by the scraper 30.
The wafer is conveyed in eight directions and is stored in a cassette (not shown) provided in the wafer storage unit 18 via the shooter 33 provided between the wafer storage unit 18 and the wafer storage unit 18 (step 116). When the wafer 12 passes through the shooter 33, the passage sensor 35 outputs a passage signal to the control means 54. Therefore, the control means 54 passes the passage signal within a predetermined time after the peeling means 16 is operated. By confirming the input of, the peeling of the wafer 12 and the transfer to the wafer storage unit 18 are confirmed (step 118).

Subsequently, the control means 16 detects the presence / absence of the wafer 12 at the peeling position E by the whole camera 42 (step 120), recognizes that the normal peeling work has been performed when there is no wafer 12, and rotates the wafer. The table 10 is rotated by the reference angle interval to prepare for the next wafer peeling operation (step 122). Then, the control means 54 detects the presence or absence of the wafer 12 at the peeling position E after the rotation operation by the whole camera 42 (step 124),
Repeat steps 116-124 until the wafer 12 is exhausted. As a result, the wafer 12 adhered in the circumferential direction
Peeling is completed.

In step 120, when the wafer 12 is located at the peeling position E after the peeling operation, the control means 5 determines that the peeling operation is not normally performed for some reason.
4 makes a judgment and performs an abnormal recovery process. In the abnormality recovery process, the rotating operation of the temporary rotary table 10 and the peeling operation by the peeling means 16 are stopped (step 132), and an alarm (not shown) is sounded (step 134) to notify the operator of the abnormality. The operator takes measures such as removing the wafer 12 that has failed to be peeled (step 136), and after the abnormal processing is completed, the temporary stop of the wafer peeling device is released (step 138), and the abnormal recovery processing is completed. After this, the control means 54 returns to step 120.

Subsequently, the control means 54 controls the rotary table 10
Check if there is a wafer in the center of the wafer (Step 12
6) If there is, this wafer is finally peeled off (step 128), and the peeling work of all the wafers 12 is completed. The control means 54 controls the second cylinder device 38 to make the mounting plate 36 horizontal, and releases the suction of the polishing plate 14 by the rotary table 10 (step 130).
The entire peeling operation of the wafer 12 on one polishing plate 14 is completed. If there is no wafer 12 in the center in step 126, step 130 is executed to complete the peeling work. Note that, of the steps described above, step 110 and step 112 can be omitted if the wafer 12 is attached to the rotary table 10 with high accuracy.

An outline of the peeling work by the peeling means 16 will be described. The control means 54 operates the motor 26, moves the slide body 22 in the direction of arrow A through the slide mechanism 24, and attaches it to the arm portion 28. The scraper 30 is moved to the back of the wafer 12 at the peeling position E. At the time of this movement, the first cylinder device 32 is operated to raise the arm portion 28 so that the scraper 30 moves the wafer 12
Avoid contact with. Thereafter, the first cylinder device 32 is operated to lower the arm portion 28 to bring the tip of the scraper 30 into contact with the upper surface of the polishing plate 14, and the motor 26 is operated again to move the slide body 22 toward the wafer storage portion 18. Let This allows scraper 30
Removes the wafer 12 and moves the wafer 1 along the inclined surface of the rotary table 10 that is inclined toward the wafer storage portion 18.
Move 2

Further, in the above-described embodiment, the wafer 12
Although the case where the orientation flat or notch is not provided in the above was explained, when the orientation flat or notch is provided, it is possible to detect the position of this orientation flat or notch and determine the reference wafer. A method of detecting the orientation flat or notch position and determining the reference wafer will be described.

First, the case where the orientation flat is provided will be described with reference to FIGS. 12 shows a general sticking method when the size of the wafer is 5 or 6 inches, and FIG. 14 shows a general sticking method when the size of the wafer is 8 inches. In this embodiment, the wafer is always attached to the center of the polishing plate 14 in the case of 5 and 6 inches, and is not attached to the center in the case of 8 inches.
In the case of 6 inches, the wafer may not be attached to the center of the polishing plate 14. If the wafer 12 is 5 or 6 inches, the wafer 12 adhered along the circumferential direction
The orientation flats of a to 12e are all directed toward the center of the polishing plate 14. For this purpose, the polishing plate 14
The reference wafer is determined by detecting the position of the orientation flat of the wafer 12f attached to the center of the. In the method, the central camera 44 magnifies and captures only the image of the central wafer 12, and sets a plurality of small-diameter circular windows that are substantially inscribed in the circle that forms the outer edge of the wafer 12. After that, the area for each High level portion (hatched portion) included in the circular window is measured. Since the area of the High level portion included in the circular window set in the orientation flat portion is smaller than the area included in the other circular windows, the direction in which the orientation flat is provided can be determined. Thus, for example, the wafer 12c in the circumferential direction at the position opposite to the orientation flat provided on the central wafer 12f with respect to the center of the polishing plate 14 is used as the reference wafer, and the wafer at various arbitrary positions is used as the reference wafer. It is possible.

When the wafer 12 is 8 inches (see FIG. 14), the orientation flats of the wafers adhered along the circumferential direction are the polishing plate 14 except one wafer 12d.
Toward the center of. Therefore, the reference wafer is determined by detecting the difference in the positions of the orientation flats. The method captures an image within a circle having a predetermined diameter centered on the center of the polishing plate 14 in the central camera 44 and including a partial region of the wafers 12a to 12d. After that, the area of each High level portion (hatched portion) included in the circle is measured. Since the area of the high level portion of the wafer 12d where the orientation flat portion is not on the center side of the polishing plate 14 is larger than the area of the high level portions of the other wafers 12a to 12c, the wafer 12d is used as a reference wafer.

Next, the case where the notch is provided will be described with reference to FIGS. In this case, as shown in FIG. 15, the wafer 12a adhered along the circumferential direction.
The notches (cutouts) provided in the to 12d face the outer edge direction of the polishing plate 14 except for one wafer 12a. Therefore, the reference wafer is determined by detecting the difference in the positions of the notches. The method is as follows:
15. In FIG. 15, only the image included in the alternate long and short dash line in FIG. 15 is enlarged and captured, and for the captured image (see FIG. 16), for example, the presence or absence of pixels along the X direction is detected for each pixel unit in the Y direction. , A curve (referred to as an original curve) indicating the boundary shape between the wafer 12 and the polishing plate 14 is calculated. Next, the primary difference and the secondary difference of the original curve are calculated, and the difference between the maximum value and the minimum value of the secondary difference is calculated. If this difference is a certain value or more, it is determined that there is a notch there ( FIG. 17 (a)
reference). When there is no notch, as shown in FIG. 17B, the difference between the maximum value and the minimum value in the quadratic difference is smaller than that when there is a notch. As a result, the wafer 12 having no notch in the image captured by the outer edge camera 46 is obtained.
It is possible to use a as a reference wafer. The reference wafer can be set freely, for example, the first wafer 12b in the clockwise direction may be used as the reference wafer based on the wafer 12a having no notch in the image. Further, when the polishing plate 14 is made of ceramic, the voltage data output from each of the cameras 42, 44, and 46 has the inverted polarity, so that the conversion unit 48 and the control unit 5 are provided.
4. Further, by inverting the logic of the data in the image processing means 52, the peeling work can be performed by the same processing as in the case of the above-mentioned glass.

Although various preferred embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments, and when the resolution of the whole camera is high, the circular window is used as a whole polishing plate. By setting to the whole area and measuring the center of gravity of each wafer,
Therefore, the position and size of the wafer placed on the polishing plate, the presence or absence of damage, and the confirmation at the time of moving to the peeling position may be checked, and so on. Of course, modifications can be made.

[0044]

When the wafer stripping apparatus according to the present invention is used, the polishing plate and the wafer stuck on the polishing plate are captured by the camera, and the voltage data output from the camera is converted into digital data for each pixel by the conversion means. The image data is converted and stored in the image storage means as image data, and the image processing means processes the image data to identify the images of the polishing plate and the wafer, and the wafer shape, size, number, and attachment position as processing information. Since the output is performed, the control unit controls the peeling unit and the rotary table based on the processing information, and the peeling work can be automatically performed. Further, since the wafer can be captured as image data, it is possible to detect the presence or absence of damage such as chipping of the wafer, which is a remarkable effect.

[Brief description of drawings]

FIG. 1 is a partially cutaway front sectional view showing a mechanical structure of a wafer peeling apparatus according to the present invention.

FIG. 2 is a plan view showing a configuration of an outer cover shown in FIG.

FIG. 3 is a right side view showing the configuration of the outer cover of FIG.

FIG. 4 is a block diagram showing a configuration of a processing system of FIG.

FIG. 5 is an explanatory diagram showing a lighting method using the outer cover of FIG. 1.

FIG. 6 is a diagram showing an image captured by the camera when the first glass plate lamp in FIG. 5 is turned on.

FIG. 7 is a diagram showing an image captured by the camera when the second lamp for the ceramic plate in FIG. 5 is turned on.

FIG. 8 is a diagram for explaining the processing in step 104 of the apparatus in FIG.

FIG. 9 is a diagram for explaining the process in step 108 of the apparatus in FIG.

FIG. 10 is a diagram for explaining the processing in step 110 of the apparatus in FIG.

11 is a diagram for explaining the processing in steps 114 to 118 of the apparatus in FIG.

FIG. 12 is an explanatory diagram showing a method of determining a reference wafer by utilizing an orientation flat (in the case of 5 and 6 inch wafers).

FIG. 13 is an explanatory diagram showing a method of detecting the position of the orientation flat of the central wafer.

FIG. 14 is an explanatory diagram showing a method of determining a reference wafer using an orientation flat (in the case of an 8-inch wafer).

FIG. 15 is an explanatory diagram showing a method of determining a reference wafer using a notch (in the case of an 8-inch wafer).

16 is a diagram illustrating one method of detecting the presence or absence of a notch in FIG.

17 is a diagram illustrating one method of detecting the presence or absence of a notch in FIG.

18 is a flowchart showing a peeling procedure of the apparatus of FIG.

[Explanation of symbols]

10 rotating table 12 wafers 14 Polishing plate 16 Peeling means 42 Whole camera 44 Central camera 46 Edge camera 47 image processing means 48 conversion means 50 image storage means 52 peeling condition storage means 54 control means

─────────────────────────────────────────────────── --Continued from the front page (56) Reference JP-A-5-166771 (JP, A) JP-A-4-363047 (JP, A) JP-A-6-94430 (JP, A) JP-A-63- 144956 (JP, A) Actual Kaihei 4-115055 (JP, U) Actual Kaihei 4-19873 (JP, U) (58) Fields investigated (Int.Cl. ⁷ , DB name) H01L 21/304

Claims (6)

(57) [Claims] 1. A rotary table on which a polishing plate having a wafer adhered on its upper surface can be placed, a peeling means for peeling the wafer from the polishing plate placed on the rotary table, and above the rotary table. And a camera that receives light from the polishing plate and the wafer and generates voltage data for each pixel forming an image in the field of view, and a digitized voltage data for each pixel.
Conversion means for converting and outputting digital data and the conversion means
The digital data for each pixel output from
On a logical plane that is an XY Cartesian coordinate system corresponding to the field of view
Image data stored as image data showing the pixel position
Storage means and the shape of the wafer on the polishing plate.
Shape, number of sheets to be pasted, and at least one of the pasting locations.
Multiple sticking specifications of the wafer are stored and the sticking specifications are stored.
Wafer peeling conditions corresponding to each of the
Peeling condition storage means and the output from the conversion means
Storing the digital data in the image storage means,
The image data stored in the storage means is processed to apply the sticking finish.
And the peeling conditions
The pasting specifications stored in the storage means are compared and pasted.
Specify the dressing specification and use the above-mentioned window corresponding to the specified sticking specification.
Rotate the rotary table based on the peeling conditions of the wafer
The polishing plate by controlling the peeling means.
And a control means for peeling the wafer from the wafer.
In the wafer peeling device, the polishing plate is formed into a disc shape and
The plate has a predetermined number of the above-mentioned pasting specifications.
According to one sticking specification, the wafer is at least the
It is attached along the circumferential direction to the peripheral edge of the polishing plate.
The control means has the same center as the center of the polishing plate.
Having a predetermined diameter including a partial area of each of the wafers
Set a circular window of the
The center of gravity of the area contained in the window is calculated and the
The wafer from the center of the polishing plate and the position of the center of gravity
Of at least one of
A wafer peeling apparatus , wherein the sticking specification is obtained by calculating and obtaining . 2. The position of the center of gravity of the control means is the
Area near the center of the polishing plate and in the area
Is larger than the reference area, the center of the polishing plate
2. It is also determined that there are wafers.
The wafer peeling apparatus described. 3. The position of the obtained center of gravity is controlled by the control means.
Position on the basis of any wafer
The angular interval between the
-The standard angle between wafers calculated from the number of attached
Calculation of the error between the distance and all the work to calculate this error
Wafer is used as a reference, and the maximum error is the smallest.
When the wafer becomes a standard
Decide on a quasi-wafer and start peeling from the reference wafer
The rotary table by the reference angle intervals.
It is characterized by controlling so that the wafers are sequentially peeled off.
The wafer stripping apparatus according to claim 1 or 2. 4. The control means is provided in the polishing plate.
If there is a wafer in the heart,
4. Control so that peeling is performed.
Wafer stripping equipment. 5. An outer cover which is provided above the turntable to cover the turntable and blocks external light from entering the camera; 5. A wafer peeling apparatus according to claim 1, further comprising an illuminating unit that illuminates the table with diffused light. 6. The lighting means comprises a first lighting means and a second lighting means.
The second illuminating means, the first illuminating means is disposed at a position where the reflected light of the first illuminating means reflected on the turntable directly enters the camera, 6. The wafer peeling apparatus according to claim 5 , wherein the illuminating means is disposed at a position where the reflected light of the second illuminating means reflected on the turntable does not directly enter the camera.