KR980012210A - Rotation direction detection method, measurement position determination method and apparatus therefor - Google Patents

Abstract

The present invention provides a technique capable of detecting the rotational direction of a wafer without using the rotating mechanism of the wafer stage and determining the position of the wafer measurement point even when the rotational direction of the wafer is unclear.

By taking an image of a part of the imaging area on the surface of the wafer to be measured and analyzing the straight portion included in the image, four equivalent rotation angles different from each other by an integer multiple of 90 degrees are determined as the rotation angles that the wafer under measurement can take. The matching pattern is detected by performing pattern matching using a model pattern, and one of four equivalent rotation angles is selected based on the direction of the matching pattern. The rotation direction of the wafer under measurement is determined from the selected equivalent rotation angle. In order to determine the position of the measurement point of the wafer, the reference angle is used to register the rotation angle θ1, the position of the alignment reference point RP and the positions of the plurality of measurement points PM1 to PM15 in advance. Further, the first model pattern is acquired near the alignment reference point RP, and the second model pattern is acquired near the measurement point. Regarding the wafer to be measured, first, by performing a pattern matching process using the first model pattern in the region near the alignment reference point RP, the rotation direction θ2 of the wafer to be measured and the alignment reference frame RP are determined. Determine your location. Based on this information, the predicted position of the measuring point is determined, and the actual position of the measuring point is determined by performing a pattern matching process using the second motel pattern in the region near the predicted position of the measuring point.

Description

Rotation direction detection method, measurement position determination method and apparatus therefor

1 is a block diagram showing a configuration of a measuring apparatus having a function of performing alignment processing (alignment processing) of a semiconductor wafer by applying an embodiment of the present invention.

2 is a block diagram showing an internal configuration of an image processing unit 50. FIG.

3 is a flow chart showing a procedure of preprocessing using a reference wafer.

4 is a conceptual diagram showing the arrangement of a plurality of chips formed in the semiconductor wafer WF.

5 is a conceptual diagram showing an enlarged vicinity of the center of the wafer.

6 is an explanatory diagram showing a process of detecting linear edge information by a one-dimensional projection method.

7 is an explanatory diagram showing a method of image processing by a Sobel operator

FIG. 8 is an explanatory diagram showing a method of calculating the angle of the straight portion of the image from the horizontal edge value and the vertical support value obtained using the Sobel operator. FIG.

9 is an explanatory diagram showing an example of a multi-gradation image to be processed and a bar graph of angles detected by the Sobel operator method in the multi-gradation image.

10 is an explanatory diagram showing four equivalent rotation angles.

11 is an explanatory diagram showing a state in which the center of view is moved to the intersection position of the scribe line SL.

12 is an explanatory diagram showing the registration form of the model pattern MPa

FIG. 13 is an explanatory diagram showing a process content of step S10. FIG.

14 is a flowchart showing a procedure for determining the rotation angle of a wafer under measurement.

15 is an explanatory diagram showing an example of a field of view set on a wafer under measurement.

FIG. 16 is an explanatory diagram showing a method of pattern matching for a wafer under measurement.

FIG. 17 is an explanatory diagram showing the relationship between the preliminary rotation angle α2pr of the wafer and the inquiry full-angle angle α2 in the wafer under measurement.

18 is an explanatory diagram showing a relationship between two fields of view set on a wafer under measurement.

19 is an explanatory diagram showing a method of obtaining a high precision relative rotation angle.

20 is an explanatory diagram showing a method of obtaining a relative relative angle of rotation.

21 is an explanatory diagram showing an outline of alignment processing in the embodiment.

22 is a flowchart showing the entire procedure of counterfeit matching in the embodiment;

23 is a flowchart showing a procedure of pre-alignment preprocessing using the reference wafer WF1.

24 is a flowchart showing the procedure of pre-alignment preprocessing using the reference wafer WF1.

25 is an explanatory diagram showing a registration form of a model pattern MPa.

FIG. 26 is an explanatory diagram showing a process content in step S10. FIG.

FIG. 27 is a flow chart showing a procedure of fine alignment preprocessing using the reference wafer WF1. FIG.

FIG. 28 is an explanatory diagram showing a visual field W (i) including the i-th measurement point PM i in the fine alignment preprocessing of the reference wafer WF1.

29 is a flowchart showing a procedure of pre-alignment processing of a wafer under measurement.

30 is a flowchart showing a procedure of pre-alignment processing of a wafer under measurement.

FIG. 31 is an explanatory diagram showing a method of pattern matching for a wafer under measurement. FIG.

32 is an explanatory diagram showing the relationship between the reference point Qc of the matching pattern MPc and the first scribe line intersection Pc.

33 is an explanatory diagram showing a relationship between two fields of view set on a wafer under measurement.

34 is a flowchart showing a procedure of fine alignment processing using a wafer under measurement.

35 is an explanatory diagram showing an image after rotation obtained in the vicinity of a measurement point in a fine alignment process using a wafer under measurement;

36 is an explanatory diagram showing a processing method in the second embodiment.

37 is an explanatory diagram showing a processing method in the third practical example.

* Explanation of symbols for main parts of the drawings

30: Zeo operation unit 31: display unit

32: control unit 33: control unit

34: Stage drive unit 35: Stage coordinate reading unit

36: XY stage 36a, 36b ...: Wafer holding arm

38: communication path 40: optical unit

41: camera 42: light source

43: half mirror 44: objective lens

50: image processing unit 110: CPU

112: bus line 114: ROM

116: RAM 136: Monitor

138: magnetic disk 139: alignment information file

140: input and output interface 150: equivalent rotation angle determination means

152: imaging position determining means 154: pattern matching means

156: angle selection means 158: rotation direction determination means

160: reference position determining means 162: measurement position determining means

DL1, DL2 ...: Reference point connection direction Dw1, Dw2 ...: Reference direction of the wafer coordinate system

Ds ... Reference direction of stage coordinate system a1, a2 ... Turning angle

θ1, θ2 ...: High precision rotation angle

The present invention relates to a method for detecting a rotation direction of a semiconductor wafer, a method for determining a measurement point, and an apparatus thereof. A semiconductor wafer is processed and measured by various apparatuses and measuring apparatuses in the manufacturing process. Depending on the apparatus, it is necessary to recognize the rotation direction (orientation) of the wafer. For example, in the film thickness measuring apparatus of the wafer, the optical head is moved to a predetermined measuring point on the wafer during the measurement. For this reason, first, the rotation direction of a wafer is detected and alignment processing (alignment process) is performed according to this rotation direction. As a conventional technique for detecting the rotational direction of a wafer, there exist some which were described, for example in Unexamined-Japanese-Patent No. 8-23023 disclosed by this applicant. In this technique, the mechanism which rotates the stage of a wafer is provided, and the position of the notch and orientation flat in the outer periphery of a wafer is detected by a sensor, rotating a wafer. Then, the rotational direction of the wafer is corrected by rotating the stage so that the detected position is a predetermined position. In addition, at the measurement distance, " alignment process (alignment process) " is performed to accurately position the measurement probe (optical element, electrode, etc.) at a predetermined measurement point on the wafer.

[Problem to Solve Invention]

Even with the above-described prior art, it is possible to detect the rotational direction of the semiconductor wafer. However, this technique takes time because the operation of rotating the wafer is required, and there is a problem that only the apparatus having the rotating mechanism of the stage of the wafer can be applied. Further, depending on the measuring device, the wafer may take any rotational direction (orientation) when the wafer is placed on the support of the measuring device. In such a measuring apparatus, since the rotation direction of the wafer is unclear, it was difficult to position the measuring point on the wafer.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problems in the prior art, and it is possible to detect the rotational direction of the wafer without using the rotating mechanism of the wafer stage, and to determine the position of the measuring point of the wafer even when the rotational direction of the wafer is unclear. It aims to provide the technology that there is.

[Means to solve the problem, its action and effect]

A first method for solving at least part of the above-described problems includes the steps of: (a) preparing a model pattern to be used for pattern matching of an image in a method of detecting a rotation direction of a wafer under measurement mounted on a table; (b) imaging a first image of a portion of the imaging area on the surface of the wafer under measurement; (c) processing the first image to detect a first straight portion included in the first image; Determining, from the direction of one straight portion, a first set of four equivalent rotation angles different from each other by an integer multiple of 90 degrees as a rotation angle at which the wafer under measurement on the table can rotate; (d) the first image; Pattern matching is performed on the model pattern to detect a matching pattern in the first image, and the fourth equivalent rotation angle is formed as a first direction indicator angle based on a direction of the matching pattern. And a step of selecting one from one set, and (e) determining a rotation direction of the wafer under measurement on the table according to the first direction indicator angle.

In the first invention, the first image is processed to detect a first straight portion included in the first image, and as a rotation angle at which the wafer under measurement on the table can rotate from the direction of the first straight portion. A process of determining a first set consisting of four equivalent rotation angles different from each other by an integral multiple of 90 degrees, pattern matching is performed on the first image with the model pattern to detect a matching pattern in the first image, and the matching Any process of selecting one of the first sets of the four equivalent rotation angles as the first direction indicator angle based on the direction of the pattern can be performed by image processing. Therefore, the rotation direction of the wafer can be detected without using the rotating mechanism.

In the first invention, the step (a) includes (1) providing a reference wafer having a surface substantially the same as the surface of the wafer under measurement on the table, and (2) partial imaging of the reference wafer surface. (3) processing the second image to detect a second straight portion included in the second image, and from the direction of the second straight portion, the reference wafer on the table Determining a second set consisting of four equivalent rotation angles different from each other by an integral multiple of 90 degrees as a rotation angle that can be rotated, (4) extracting a model pattern from a part of the second image, and (5 And a step of selecting one of the second sets of the four equivalent rotation angles as a second direction indicator angle based on the direction of the model pattern, wherein step (e) comprises the first and the first 2 direction indicator angle And calculating a difference to determine a relative rotation angle of the wafer under measurement with respect to the reference wafer. This allows the relative rotation angles of the reference wafer and the wafer to be measured to be determined.

The second invention relates to a method of detecting a rotational direction of a wafer under measurement, comprising the steps of: (a) preparing a model pattern for pattern matching of an image; and (b) a first imaging area which is a part of the instantaneous wafer. (C) processing the first image to detect a first straight portion included in the first image, and rotating the wafer under measurement on the table from the direction of the first straight portion; Determining a first set consisting of four equivalent rotation angles different from each other by an integral multiple of 90 degrees, and (d) performing pattern matching with the pattern on the first image to perform the pattern matching in the first image. Detecting a matching pattern and selecting one of the first sets of four equivalent rotation angles as a first direction indicator angle based on a direction of the matching pattern; and (e) the first direction indicator angle; Based on By specifying a predetermined search direction in the first imaging area, specifying at least one other imaging area away from the first imaging area by a predetermined distance from the first imaging area, and at the same time, the image of each of the specified imaging areas is specified. A step of picking up an image, (f) performing pattern matching on the image of the image pickup area with the model pattern to detect a matching pattern, and (g) the first image picking area arranged along the search direction; And determining a first connection direction for connecting the predetermined reference positions of the matching patterns detected in each of the plurality of imaging areas with each other, and determining the rotational direction of the wafer under measurement only based on the first connection direction. It is characterized by.

In the second invention, a plurality of matching patterns are obtained by image processing, and the rotation direction of the wafer is determined based on the connection direction connecting the reference positions of the plurality of matching patterns. Therefore, the rotation direction of the wafer can be detected without using the rotating mechanism. In addition, since the connection direction is accurately determined from the reference positions of the plurality of matching patterns, the rotational direction of the wafer can be detected with relatively high accuracy.

In the second invention, the step (a) comprises the steps of (1) providing a reference wafer having the same surface as the surface of the wafer under measurement on the table, and (2) a second portion of the reference wafer surface. (3) detecting the second straight portion included in the second image by processing the second image, and from the direction of the second straight portion, the reference on the table Determining a second set of four equivalent rotation angles different from each other by an integer multiple of 90 degrees as a rotation angle at which the wafer can rotate; (4) extracting the model pattern from a part of the second image; (5) selecting one of the second sets of the four equivalent rotation angles as the second direction indicator angle based on the direction of the model pattern; and (6) based on the second direction indicator angle. By the second shot Specifying a predetermined search direction in the area, specifying at least one other imaging area spaced apart from the second imaging area by a predetermined distance along the search direction, and simultaneously picking up an image of each specified imaging area And (7) detecting the matching pattern by performing pattern matching on each of the imaging regions on the reference wafer with the model pattern, and (8) the second imaging region arranged along the search direction. And determining a second connection direction for connecting the predetermined reference positions of the matching mates detected in each of the plurality of imaging areas including each other, wherein step (g) is performed by the first and second direction indicator angles. Preferably, the method includes determining a relative rotation angle of the wafer under measurement with respect to the reference wafer by obtaining an angle difference. In this way, the relative rotation angles of the reference wafer and the wafer to be measured can be determined with relatively high accuracy.

In the second invention, the reference is based on coordinate values of the reference positions of the plurality of matching patterns on the reference wafer and coordinate values of the reference positions of the plurality of matching patterns on the wafer under measurement. It is preferable to further comprise the step of determining the relative relationship between the first coordinate system of the wafer and the second coordinate system of the wafer under measurement. This allows not only the relative direction of rotation of the wafer, but also the relative coordinates.

The third invention is a method for determining the position of a measuring point on a wafer under measurement mounted on a table, comprising: (a) registering in advance the positional relationship between the alignment reference point on the surface of the wafer to be measured and the measuring point; Preparing a first model pattern used for pattern matching on an image near the alignment reference point on the measurement wafer surface, and a second model pattern used for pattern matching on an image near the measurement point; b) imaging a first image in a first imaging area near the alignment reference point; and (c) performing a first process on the first image, including a pattern matching process using the first model pattern. Determining the rotational direction of the wafer to be measured and the position of the alignment reference point; (d) the position of the rotational direction and the alignment reference point determined in the step (c); Estimating the position of the measurement point based on the positional relationship between the fit reference point and the measurement point, (e) imaging the second image in the second imaging area near the predicted position; And performing a second process including a pattern matching process using a second model pattern on the second image to determine the actual position of the measurement point.

In the pattern matching process, the position of the image portion and the direction of the image portion which are almost the same as the model pattern can be determined. The first process including the pattern matching process may determine the rotation direction and the position of the alignment reference point of the wafer under measurement. From this rotation direction and the position of the alignment reference point, the correspondence between the coordinate system fixed to the measuring device and the coordinate system fixed to the wafer can be determined. Since the positional relationship between the alignment reference point and the measurement point is registered in advance, the position of the measurement point can be predicted from the correspondence of the two coordinate systems. When the second process including the pattern matching process is performed in the vicinity of the predicted position, the position of the measurement point can be determined. Therefore, even if the rotation direction of the wafer is unclear, the position of the measuring point of the wafer can be accurately determined.

In the third invention, the wafer to be measured is provided with a plurality of the measuring points on the surface, and for the specific measuring points within the predetermined range from the alignment reference point among the plurality of measuring points, the process (e) and the It is preferable to omit the process of process (f), and to use the said predicted position with respect to the said specific measurement point obtained by the said process (d) as a measured position of the said specific measurement point. In this way, the amount of deviation between the predicted position and the measured position is relatively small for the measurement point within the predetermined range from the alignment reference point. Therefore, by eliminating steps (e) and (f) for the measurement points within this range, the processing time of the entire alignment process can be shortened while securing a certain alignment accuracy. Further, in the third invention, the plurality of measurement points are provided on the surface of the wafer under measurement, and among the plurality of measurement points, a predetermined number of the measurement points is applied at a first measurement point at which the processing of step (e) and step (f) is performed. The processing of steps (e) and (f) is omitted for the second measuring point within the range, and the predicted position of the second measuring point is determined based on the shift amount between the predicted position and the measured position at the first measuring point. By correcting, the actual measurement position of the second measurement point may be determined. In this way, the processing time of the entire alignment process can be shortened while ensuring a certain degree of alignment accuracy for the above reason. In the method, the step (c) includes a step of imaging a plurality of the first images for a plurality of regions near the alignment reference point, and the step (d) includes the plurality of first images. It is preferable to determine the rotational direction of the wafer under measurement and the position of the alignment reference point by executing the first processing for each. By performing a pattern matching process on a plurality of images, it is possible to more accurately determine the rotation direction of the wafer under measurement and the position of the alignment reference point.

[Invention is another aspect]

The present invention also includes other aspects as follows. In the first aspect, the process for determining the rotational direction in the step (c) of the method includes (i) processing the first image to detect a first straight portion included in the first image, and the first straight line. Determining, from the direction of the portion, a first set consisting of four equivalent rotation angles different from each other by an integral multiple of 90 degrees as a rotation angle at which the wafer under measurement on the table can rotate; and (ii) with respect to the first image. Pattern matching is performed on the first model pattern to detect a matching pattern in the first image, and 1 of the first sets consisting of the four equivalent rotation angles as a first direction indicator angle based on the direction of the matching pattern. A step of selecting a dog, and (iii) determining a rotation direction of the wafer under measurement on the table according to the first direction indicator angle.

In a second aspect, in the first aspect, the step (a) includes (1) providing a reference wafer on the table having a surface substantially the same as that of the wafer under measurement, and (2) the Imaging a third image of the imaging area on the wafer surface; (3) processing the third image to detect a second straight portion included in the second image; and from the direction of the second straight portion, Determining a second set of four equivalent rotation angles different from each other by an integral multiple of 90 degrees as a rotation angle at which the reference wafer on the table can rotate; and (4) the first model pattern is obtained from a portion of the third image. And (5) selecting one of said second sets consisting of said four equivalent rotation angles as second direction indicator angles, based on the direction of said first model pattern, said process (c) is the first First by calculating the difference between the two direction indicators angle includes a step for determining a relative rotation angle of the wafer to be measured with respect to the reference wafer.

In a third aspect, the process for determining the rotational direction in step (c) of the method comprises: (i) processing the first image to detect a first straight portion included in the first image, wherein the first straight line Determining, from the direction of the portion, a first set consisting of four equivalent rotation angles different from each other by an integral multiple of 90 degrees as a rotation angle at which the wafer under measurement on the table can rotate; and (ii) with respect to the first image. Pattern matching is performed on the first model pattern to detect a matching pattern in the first image, and one of the first sets consisting of the four equivalent rotation angles is selected as a first direction indicator angle of the matching pattern. And (iii) specifying a predetermined search direction in the first imaging area based on the first direction indicator angle, and at least one separated from the first imaging area by a predetermined distance along the search direction. Picking up another image pickup area and simultaneously picking up an image of each image picked-up area; and (iv) performing pattern matching on the image of each image pickup area by the first model pattern to detect a matching pattern. And (v) determining a first connection direction connecting each other to a predetermined reference position of a matching pattern detected in each of the plurality of imaging areas including the first imaging area arranged along the search direction, and And determining a rotation direction of the wafer under measurement based on the first connection direction.

In a fourth aspect, in the third aspect, the step (a) comprises (1) providing a reference wafer on the table having a surface substantially the same as the surface of the wafer under measurement, and (2) the reference wafer surface. Picking up a third image of the third imaging area which is a part of (3) from the direction of the second straight portion, in order to detect the second straight portion included in the third image by processing the third image, Determining a second set of four equivalent rotation angles different from each other by an integer multiple of 90 degrees as a rotation angle at which the reference wafer on the table can rotate; and (4) the first model from a portion of the third image. Extracting a pattern; (5) selecting one of said second sets of said four equivalent rotation angles as a second direction indicator angle based on the direction of said first model pattern; ) The angle of the second direction indicator Specifying a predetermined retrieval direction in the third imaging area, specifying at least one other imaging area away from the third imaging area by a predetermined distance along the retrieval direction, and at the same time (7) performing pattern matching on the image of each image capturing area on the reference wafer with the first model pattern to detect a matching pattern, and (8) the search direction. And forming a second connection direction for connecting the predetermined reference positions of the matching pattern detected in each of the plurality of imaging areas including the third imaging area arranged along with each other, wherein the step (c) includes: Determining a relative rotation angle of the wafer under measurement with respect to the reference wafer by obtaining an angle difference between the first and second connection directions.

In the fifth aspect, in the fourth aspect, the step (a) includes the coordinate values of the reference positions of the plurality of matching patterns on the reference wafer and the reference positions of the plurality of matching patterns on the wafer under measurement. And determining a relative relationship between the first coordinate system of the reference wafer and the second coordinate system of the wafer under measurement based on the coordinate value.

A sixth aspect is an apparatus which realizes each step of each of the above inventions as corresponding means, respectively, in order to detect a rotational direction of a wafer under measurement mounted on a table.

A seventh aspect is an apparatus which realizes each of the above steps of the invention as corresponding means, respectively, in order to determine the position of the measurement point of the wafer under measurement mounted on the table.

An eighth aspect is a recording or storage medium on which a computer program is recorded or stored which executes the functions of each process or each means of the invention by a computer. As a recording medium, a portable storage medium that can be read by a computer such as a flexible disk or CD-ROM, an internal storage device (memory such as RAM or ROM) and an external storage device of a computer system, or a medium on which a computer program other than this is recorded. Many media can be read by a computer system.

A ninth aspect is a program supply device for supplying a computer through a computer program for executing the functions of each process or each means of the present invention.

[Embodiment of the Invention]

A. Configuration of the device

EMBODIMENT OF THE INVENTION Next, embodiment of this invention is described based on an Example. 1 is a block diagram showing a configuration of a measuring apparatus having a function of performing alignment processing (alignment processing) of a semiconductor wafer by applying an embodiment of the present invention. This measuring apparatus is provided with the control operation unit 30, the optical unit 40, and the image processing unit 50. As shown in FIG. The control operation unit 30 includes a display unit 31, an operation unit 32, a control unit 33, a stage drive unit 34, a stage coordinate reading unit 35, and an XY stave 36. Doing. As the display part 31, a monitor, a liquid crystal display, etc. are used, for example. As the operating column 32, for example, a keyboard or a mouse is used. The semiconductor wafer WF is mounted on the XY stage 36. On the surface of the semiconductor wafer WF, a plurality of rectangular semiconductor chips arranged in a tile shape are formed. In addition, this apparatus does not have a mechanism for rotating the XY stage 36.

The optical unit 40 includes a camera 41, a light source 42, a half mirror 43, and an objective lens 44.

The half mirror 43 reflects light emitted from the light source 42 toward the objective lens 44 and irradiates light onto the semiconductor wafer WF on the XY stage 36. Light reflected from the surface of the semiconductor wafer WF passes through the objective lens 44 and the half mirror 43 and enters the camera 41. That is, the camera 41 picks up an image of the surface of the semiconductor wafer WF. As the image, it is preferable to read a multi-gradation image (gray image). In this embodiment, the size of the field of view of the camera 41 is smaller than the size of one semiconductor chip formed on the surface of the semiconductor wafer. As described later, the multi-gradation image of the semiconductor wafer WF is processed by the image processing unit 50, whereby the rotational direction of the semiconductor wafer WF is detected. The monitor 136 of the image processing unit 50 displays a multi-gradation image of a part of the imaging area of the semiconductor wafer WF.

When the user operates the operation unit 32 and inputs a movement command to the XY stage 36, the control unit 33 controls the stage driving unit 34 according to the command to move the XY stage 36 in the X and Y directions. Move to. When the coordinate reading command of the stage is input by the operation unit 32, the stage coordinate information at that time is read by the stage coordinate reading unit 35 and supplied to the control unit 33. Stage coordinate information is displayed on the display part 31 as needed. The stage coordinate information is also supplied from the control unit 33 to the image processing unit 50 via the bidirectional communication path 38. As will be described later, the image processing unit 50 determines the rotational direction and the measurement position of the wafer by using the rotational direction recognized by the image processing and this stage coordinate information.

2 is a block diagram showing an internal configuration of the image processing unit 50. As shown in FIG. This image processing unit 50 is composed of a computer system in which a CPU 110, a ROM 114, a RAM 116, and an input / output interface 140 are connected to a bus line 112. The monitor 136, the magnetic disk 138, and the communication path 38 are connected to the input / output interface 140.

The RAM 116 includes an equivalent rotation direction determining means 150, an imaging position determining means 152, a pattern matching means 154, an angle selecting means 156, a rotation direction determining means 158, and a reference positioning means. An application program for realizing the 160 and the measuring position determining means 162 is stored. When determining only the rotation direction, the reference positioning means 160 and the measurement positioning means 162 may be omitted, and the function of each of these means will be described later.

Further, a computer program (application program) for realizing the functions of each of these means is provided in the form of a recording on a portable recording medium (a recordable medium) such as a floppy disk or a CD-ROM. Is transferred to main memory or external memory. Then, it is stored in the RAM 116 at the time of execution. Alternatively, the computer program may be supplied from the program supply apparatus to the computer system through a communication path. In the present specification, the computer system refers to an apparatus including hardware and an operation system and operating under the control of the operation system. The application program realizes the functions of the above-described parts in such a computer system. In addition, some of the functions described above may be implemented by the operation system rather than the application program.

B. Determination of Wafer Rotation Direction

Determination of the wafer rotation direction is roughly divided into a pretreatment process using a reference wafer and a process process on a wafer under measurement. Below, each of these processes is demonstrated, respectively.

B-1. Preprocessing Using Reference Wafer: FIG. 3 is a flowchart showing the procedure of preprocessing using reference wafer. The reference wafer is a wafer in which a pattern similar to the wafer under measurement to be processed for rotation direction detection is formed. In general, one of a plurality of wafers processed in the same lot is used as a reference wafer, and another wafer becomes a wafer under measurement.

In step S1 of FIG. 3, the user inputs the chip dimensions of the wafer and the number of chips in the X-axis direction and the Y-axis direction.

4 is a conceptual diagram showing the arrangement of chips formed on the surface of a semiconductor wafer. On the surface of the semiconductor wafer WF, a plurality of rectangular chips CP of the same size are arranged in a tile shape. There are four combinations of even-numbered and odd-numbered chips in the X- and Y-axis directions: even-even, odd-even, even-odd, and odd-odd. FIG. 3 (A) is an example of even-odd, and FIG. 3 (B) is an example of even-odd. The chip position near the center can be calculated based on the center O of the wafer from the information of any of these four combinations and the pitches LX and LY of the length and width of the chip. Therefore, at step S1, information indicating at least one of the four combinations of the number of chips and information indicating the pitches LX and LY of the chips are input.

3, in step S2, the multi-gradation image (gray image) is imaged by the camera 41 at the center position of the wafer. When the wafer is initially placed on the XY stage 36, as shown in FIG. 4, the outer circumference of the wafer is held by the wafer holding arms 36a and 36b of the XY stage 36, and the XY stage 36 Is located almost in the center. In this state, when imaging is performed with the camera 41, an image near the center of the wafer can be obtained.

5 is a conceptual diagram showing an enlarged vicinity of the center of the wafer. In this embodiment, it is assumed that the characteristic pattern PT which is not present in the other three corners is formed at the upper right corner of each chip CP. An image portion including this pattern PT is used as a model pattern in pattern matching described later. Chip CP is divided by orthogonal scribe lines SL. In the multi-tone image obtained by imaging the wafer surface, the scribe line SL may be identified as a dark region or may be identified as a bright region. In any case, the scribe line SL may be It can be identified as an area different in brightness from the chip CP.

5 illustrates the positions of the views W1 to W4 of the camera 41 according to a combination of four types of chips. As described above, since the field of view of the camera 41 is smaller than the size of one chip, not all of the chips are included in the field of view. The first field of view W1 corresponds to the imaging area at the center of the wafer when the number of chips is even-even.

This visual field W1 is located in the substantially center of the intersection of the scribe line SL. The second field of view W2 corresponds to the imaging area at the center of the wafer in the case where the number of chips is even-odd. This field of view W2 is in the position fitted to two chips. The third field of view W3 corresponds to the imaging area at the center of the wafer in the case where the number of chips is odd-even. This visual field W3 is also in the position fitted to the chip | tip. The fourth field of view W4 corresponds to the imaging area at the center of the wafer in the case where the number of chips is odd-odd. This field of view W4 is at the nearly center position of the chip. In addition, since the reference wafer is actually rotated at the position shown in FIG. 5, the visual fields W1 to W4 are inclined from the normal orientation in the scribe line SL.

In step S2 of FIG. 3, since the image is picked up at the center position of the wafer, an image is obtained at any position in the fifth-view visual field W1 to W4. This image is used to recognize a straight portion (scribe line SL or the like) included in the image in the next step S3. If the number of chips is odd-odd, there is a high possibility that the straight portion is not included as in the fourth field of view (W4) of FIG. In this case, therefore, only half of the pitch (LX or LY) of the chip may be imaged at a position away from the center of the wafer.

In step S3 of FIG. 3, the equivalent rotation direction determining means 150 (FIG. 2) detects the linear edge information included in the image, and determines the coarse rotation angle of the reference wafer from the linear edge information. The term " coarse rotation angle " means a relatively low precision rotation angle obtained from linear edge information. As the extraction method of the linear edger information, the one-dimensional projection method, the Sobel Operator method, and the like described below can be used.

6 is an explanatory diagram showing a process of detecting linear edge information by a one-dimensional projection method. 6 shows a two-dimensional multi-gradation image in which straight portions exist only in the horizontal direction. In the one-dimensional projection method, this two-dimensional multi-gradation image is projected one-dimensionally in various directions and the pixel values are added. When projecting in a direction parallel to the straight portion, the pixel value at the coordinate where the straight portion is present has a large pitch. On the other hand, when projecting in a direction not parallel to the straight portion, the peak value of the added pixel value is smaller than this. In this way, the projection direction in which the peak value of the accumulated value of the pixel values is maximized by one-dimensional projection of the two-dimensional image in various directions can be determined as the direction of the straight portion. The inquiry full angle is determined from the direction of the straight line portion. For example, a predetermined direction (e.g., the 3 o'clock direction of the clock) of the stage coordinate system (coordinate system fixed to the XY stage 36) is referred to as the reference direction, and the angle measured from the reference direction to the direction of the linear portion in the counterclockwise direction is inquired. You can do a full-angle view.

7 to 9 are explanatory views showing the detection processing of the linear edge information by the Sobel operator method. 7 shows a method of image processing by the Sobel operator. In the Sobel operator method, first, a pixel block of a predetermined size (3x3 block in the example of FIG. 7) including an edge pixel as shown in FIG. 7A-1 or A-2 is selected from among multi-gradation images. do. Here, in the "edge pixel", "the pixel value of at least one pixel among 8 pixels is different from the pixel value of oneself (the center pixel of FIG. 7 (A-1), (A-2)), and also the boundary of an image. Pixel not in an image ". In Fig. 7A, a pixel in the corner portion of the image is recognized as an edge pixel. In Fig. 7A-2, a pixel in the straight portion is recognized as an edge pixel. The identification of the edge pixel is performed by scanning a 3x3 block in a multi-gradation image and determining whether the center pixel of the block conforms to the above definition.

7 (B-1) and (B-2) show the Sobel operators in the horizontal direction and the vertical direction, respectively. By operating these horizontal operators and vertical operators for 3x3 blocks including edge pixels, the horizontal and vertical edge values are obtained, respectively. FIG. 7C shows an example of calculation when the horizontal operator is operated. The horizontal edge value is obtained by applying the horizontal operator to the pixel value of the 3x3 block, and the vertical edge value is obtained by applying the vertical operator to the pixel value of the 3x3 block.

8 is an explanatory diagram showing a method of calculating the angle of the straight portion of the image from the horizontal edge value x and the vertical support y obtained using the Sobel operator. As shown in Fig. 8A, the angle α of the straight portion of the image is given by tan ⁻¹ (y / x). Here, the angle α is the angle measured counterclockwise from the horizontal right direction (3 o'clock of the clock). For example, in the example of FIG. 8B, since the vertical edge value is 0 and the horizontal edge value is 4, it can be determined that the angle α is 0 degrees. In addition, in the example of FIG. 8C, since either the vertical edge value and the horizontal edge value are 1, it can be determined that the angle α is 45 degrees. The angle α is assumed to have a value in the range of 0 to 180 degrees. The range of 180 degrees to 360 degrees is equivalent to the range of 0 degrees to 180 degrees.

9 is an explanatory diagram showing an example of a multi-gradation image and a bar graph of an angle α detected from the multi-gradation image by the Sobel operator method. In the multi-gradation image shown in FIG. 9 (A), a 3x3 block having edge pixels like the 7th (A-1) or 7th (A-2) as the center pixel is detected, and the edge pixels are detected. The angle α is determined by the method shown in FIG. 8 for each 3 × 3 block to be included. 9 (B) is a bar graph showing the frequency of the angle α obtained in this way for a large number of 3x3 blocks. In this example, the peak exists at the position of 40 degrees and 130 degrees, and the peak at the position of 40 degrees is the largest. At this time, the angle? 1 of the maximum peak position is employed as the inquiry full-angle indicating the direction of rotation of the linear part in the multi-gradation image.

Moreover, four equivalent angles different by the integral multiple of 90 degrees exist in the inquiry full-angle angle (alpha) 1 detected using the one-dimensional projection method or the Sobel operator method mentioned above. In other words, the inquiry full-angle α1 has an indeterminate degree of 1/4. 10 is an explanatory diagram showing four equivalent rotation angles. As shown in FIG. 10 (A), the case where the image of the intersection of the scribe line SL is seen in the visual field W of the camera 41 is considered. In this embodiment, since the field of view of the camera 41 is smaller than the size of the chip, any one of four types of the rotational directions of the chip in the tenth (B) to (E) degrees cannot be identified from the image data. Thus, the correct rotation angle of the wafer is one of four equivalent rotation angles at intervals of 90 degrees. In step S3 of FIG. 3, at least one of these four equivalent rotation angles is detected as an inquiry full angle. If one of the equivalent rotation angles can be detected, it can be considered that other equivalent rotation angles can also be detected.

In addition, the linear edge information obtained at step S3 and the inquiry full-angle are those of the scribe line SL in most cases. However, not only the scribe line SL but also the linear full edge information regarding the linear image portion existing in the multi-gradation image of the wafer may be detected. The linear portion of the circuit in the chip is mostly parallel to the scribe line SL. Therefore, even if the linear image portion other than the scribe line SL is detected, the inquiry full angle of the wafer can be obtained. In step S4, it is determined whether or not the linear edge information detected in step S3 is reliable. For example, when the one-dimensional projection method shown in FIG. 6 is used, this determination can be performed depending on whether or not the peak value of the accumulated pixel value is equal to or greater than a predetermined threshold value. In addition, when the Sobel operator method shown in FIGS. 7-9 is used, it can determine based on whether the peak value of the bar graph of FIG. 9 (B) is more than predetermined threshold value. Alternatively, if the image of the wafer is displayed on the monitor 136, the user can visually determine whether or not a certain straight edge is included in the image. If the straight edge information is unreliable, in step S5, the XY stage 36 is moved by a predetermined amount (e.g., 1 o'clock), and the multi-gradation image is captured at another position near the wafer center. Then, by executing step S3 again, the linear edge information is detected and the inquiry full angle? 1 is obtained. When the inquiry full-angle α1 is obtained in this way, in step S6, the target position coordinates of the XY stage 36 are included so that the imaging position determining means 152 includes the intersection position of the scribe line SL in the vicinity of the center of the waiter. Calculate and move. As described above, the initial seeding position at the center of the wafer according to the four types of combinations of the number of chips in the X-axis direction and the Y-axis direction (even-even, odd-even, even-odd, odd-odd is shown in FIG. 5). It is almost determined by the positions of the four visual fields W1 to W4 shown in Fig. 6. The imaging position determining means 152 (Fig. 2) includes the inquiry full-angle α obtained in step S3 and the dimensions (pitch LX, LY) of the chip. It is calculated from the information on the number of chips in the X direction and the Y direction to determine how much the intersection position of the scribe line SL can be moved to the position included in the field of view. (FIG. 1) and move the XY stage 36. Then, the multi-gradation image is captured again by the camera 41. Since the inquiry full-angle α1 has 1/4 uncertainty, The multi-gradation image is imaged again by one movement. Because of the uncertainty of 1/4, it cannot be said that the center of view can reach the intersection position of the scribe line SL by one movement. By changing the direction of movement in the rotational direction and moving the same distance, the center of the field of view (that is, the imaging area) can be moved to the intersection position of the scribe line SL. As shown in Fig. 11, the linear portion of the wafer (the direction of the scribe line SL is rotated by the reference angle α1 in the reference direction Ds of the stage coordinate system).

The image picked up by the camera 41 in step S6 is displayed on the monitor 136. In step S6, the imaging position determining means 152 further calculates an actual measured value (coordinate value) of the intersection point of the scribe line SL. The coordinates of the intersection positions of the scribe lines SL are used later when imaging at the corresponding intersection positions on the wafer under measurement. The intersection position of the scribe line SL of a reference wafer is represented by the coordinate of the center point Pa of the visual field Wa of the camera 41 shown, for example in FIG. The position of this point Pa can be designated by the user moving a cursor using a pointing device, such as a mouse, on the image displayed on the monitor 136. FIG. Or it is also possible to automatically determine the coordinate of the center position of the intersection of the scribe line SL by processing the multi-gradation image image | photographed with the camera 41. FIG. In the case of obtaining the center position of the intersection point in the image processing, first, the straight edge is detected in the same manner as in the above-described step S3. And the straight line which approximates the edge of the scribe line SL is calculated | required. In addition, the center position of the four corner parts which consist of these approximation lines is determined as the intersection position of the scribe line SL. In addition, the coordinate of the center position of the visual field Wa is the coordinate (coordinate fixed to a stage) of the stage coordinate system picked up by the stage coordinate reading part 35 (FIG. 1). The coordinates of the stage coordinate system at any position in the visual field Wa (that is, the picked-up image) can be easily calculated from this coordinate value.

In step S7 of FIG. 3, image processing for rotating the image picked up in S6 clockwise by the reference angle? 1 is performed, and the pattern matching model pattern (also called a "template image") is cut out from the rotated image and registered. . 12 is an explanatory diagram showing the registration form of the model pattern MPa. In step S7, first, the multi-gradation image (Fig. 12 (A)) is rotated clockwise by the inquiry angle α1 as shown in Fig. 12B at the intersection position of the scribe line SL, and then rotated. An image is displayed on the monitor 136. It is executed by Affine Transformaition of an image.

The pattern which can be used as the model pattern MPa by observing the displayed image means that the user can select one of four equivalent rotation angles which are equivalent as the pre-inquiry angle α1 in the direction of the pattern. As the model pattern MPa, a pattern having no rotational symmetry of an integral multiple of 90 degrees is preferable. In other words, a pattern having any one of rotational symmetry (90, 180, and 270 degrees) of rotation symmetry of 90 degrees is inappropriate as the model pattern MPa. In the visual field Wa near the intersection point of the scribe brain SL, each corner part of four adjacent chips is included, so that a unique pattern included in only one of these four corner parts is modeled. It can register as (MPa).

If there is no pattern that can be used as the model pattern MPa in the current field of view Wa, the XY stage 36 is moved little by little while displaying and observing the image captured by the camera 41 on the monitor 136. Let's do it. And the pattern which can be used as model pattern MPa is set to the state which enters into a visual field. If a pattern that can be used as the model pattern MPa exists in the current visual field Wa, as shown in Figs. 12B and 12C, the model pattern is cut out in the recall after rotation. The range of the model pattern MPa is specified by the user using a pointing device such as a mouse. The model pattern MPa is preferably present near the intersection of the scribe line SL, but may not necessarily be present near the intersection.

In step S8 of FIG. 3, the predetermined angle of the model pattern MPa rotated and cut out (for example, the 3 o'clock direction of the clock) is defined as the reference direction (0 degree direction) Dwl of the wafer coordinate system, whereby the inquiry total angle (α1). In the image rotated clockwise by (), the 3 o'clock direction of the clock is set to the reference direction Dwl of the wafer coordinate system. In addition, the 3 o'clock direction of the clock may be automatically set as the reference direction Dwl without the user's designation. The rotation angle of the wafer is an angle from the reference directions D and S of the stage coordinate system to the reference direction Dwl of the wafer coordinate system. In the case of the twelfth (B), the rotation angle of the reference wafer is equal to the inquiry full angle? 1. In addition, when the reference direction of the wafer coordinate system is selected to a direction other than the three o'clock direction of the clock, the rotation angle of the reference wafer is different from α1. However, even in this case, the rotation angle of the reference wafer is the value obtained by adding or subtracting a predetermined value to the inquiry total angle? 1. For example, in the state of FIG. 12B, the rotation angle of the reference wafer in which the 12 o'clock direction of the clock is selected as the reference direction of the wafer coordinate system is (α1 + 90).

In step S9 of Fig. 3, the cut model pattern MPa is registered with the inquiry full-angle angle? 1. At this time, the coordinates of the reference point (for example, the point Qa on the upper left) at the predetermined position of the model pattern MPa are also registered (FIG. 1C). In addition, the coordinate of reference point Qa is represented by the coordinate value which made the center of a wafer the origin, for example.

In step S10, the XY stage 36 is moved so that an imaging area may be located at the scribe line intersection position of the adjacent chip, and an image is imaged. Pattern matching is performed on this image to detect a pattern (matching pattern) similar to the model pattern MPa. 13 is an explanatory diagram showing the contents of the process in step S10. In this example, the visual field Wb is moved from the intersection position where the model pattern MPa is registered to the intersection position adjacent to the lower right side at an angle. The scribe line intersection positions of adjacent chips may be adjacent to each other in the vertical, horizontal, and inclined directions. The matching pattern MPb matching the model pattern MPa is detected in the image in this field of view (imaging area) Wb.

In step S10, after detecting the matching pattern MPb, the coordinate of the reference point Qb is also calculated. In addition, as the direction of the straight line L1 connecting the reference points Qa and Qb of the patterns MPa and MPb, the direction (connection direction of the reference point) DL1 from the second reference point Qb toward the first reference point Qa ). The rotation angle (angle measured in the counterclockwise direction from the reference direction Ds of the stage coordinate system) θ1 of the connection direction DL1 is calculated. In addition, since the coordinates of the reference points Qa and Qb are obtained as the coordinates of the stage coordinate system, the rotation angle θ1 of the connection direction DL1 can be obtained from these coordinates with simple steps.

It is also possible to use the rotation angle θ1 of the connection direction DL1 of the reference point as the rotation angle of the reference wafer instead of the inquiry full angle α1 shown in FIG. 12 (B). The difference between the two rotation angles is due to which direction is selected as the reference direction of the wafer coordinate system, and which may be defined as the rotation angle. The rotation angle [theta] 1 in the connection direction of the reference point of the matching pattern has an advantage that high accuracy can be determined even when looking at the inquiry full angle? 1.

In step S10, the coordinates of the center point Qab of the reference points Qa and Qb of the two matching patterns are also calculated. The following data is registered by preprocessing concerning the abnormal reference wafer.

(a) rotation angle α1 of the reference wafer;

(b) image data (template data) of the model pattern MPa;

(c) coordinate values of predetermined points Pa and Pb of the two imaging areas;

(d) coordinate values of the reference points Qa and Qb of the two matching patterns;

(e) coordinate values of the center point Qab of the reference points Qa and Qb of the two patterns;

(f): Rotation angle (θ1) in the connecting direction connecting the two reference points (Qa, Qb)

These data are used at the time of processing the wafer to be described below.

In addition, with respect to the reference wafer, predetermined processing such as various measurements is performed using the apparatus shown in FIG. For example, when the apparatus shown in FIG. 1 is a film thickness meter, the film thickness of the wafer surface is measured at a plurality of measurement positions in the reference wafer. At this time, the coordinates of the respective measurement positions are registered as the coordinates of the wafer coordinate system.

B-2, Processing of the Measured Wafer

The same wafer (measurement of film thickness) is performed at the same position as the reference wafer. However, since the rotation angle of the wafer under measurement is unclear when the wafer under measurement is placed on the XY stage 36, the position of the XY stage 36 cannot be moved so that the reference wafer is at the same measurement position. Therefore, the rotation angle of the wafer under measurement is determined according to the following procedure before executing the measurement distance for each wafer under measurement. If the coordinate of the measurement position is corrected (coordinate change) using this rotation angle, the measurement can also be performed at the same measurement position as the reference wafer.

14 is an explanatory diagram showing a procedure for determining the rotation angle of a wafer under measurement. The processing from steps S1 to S6 is the same as the processing relating to the reference wafer shown in FIG. 3. As a result, an image of the scribe line intersection near the wafer center is imaged. 15 shows an example of the field of view set on the wafer under measurement. Here, in the visual field Wc S3, the inquiry full-angle angle? 2pr is detected. In addition, this inquiry full angle? 2pr has an uncertainty of 90 degree integer multiples. In the wafer under measurement, the inquiry full angle before removing the uncertainty is also referred to as a "preliminary rotation angle". This name implies a preliminary angle of rotation that includes uncertainty.

In step S21, the pattern matching means 154 (FIG. 2) performs pattern matching processing using the model pattern MPa registered in advance in the preprocessing of the reference wafer with respect to the image in this field of view Wc, and preliminary rotation angle alpha 2pr. Eliminate uncertainty

16 is an explanatory diagram showing a pattern matching method for a wafer under measurement. First, the read image shown in FIG. 16 (A) is rotated clockwise by the preliminary rotation angle α2pr by affine change to create an image as shown in FIG. 16 (B). Then, the pattern matching with the model pattern MPa in the rotated image is detected by the pattern matching process. At this time, as shown in FIG. 16C, it is preferable to create four model patterns rotated by 90 degrees in advance as model patterns. The model pattern with the highest matching degree is determined among these four model patterns, and the position coordinates of the pattern (matching pattern) matched with this are detected. In the example of FIG. 16 (B), the matching degree of the model pattern of 180 degree rotation is the highest. Therefore, it is determined that the inquiry full-angle angle α2 of the wafer under measurement is (α2pr + 180). That is, it is possible to resolve the uncertainty of the preliminary rotation angle α2pr by template matching to determine the value of the inquiry full angle α2.

FIG. 17 is an explanatory diagram showing the relationship between the preliminary rotation angle α2pr of the wafer and the inquiry full angle α2 in the wafer under measurement. The preliminary rotation angle α2pr is an angle from the reference direction Ds of the stage coordinate system to the direction of the straight portion (scribe line SL) of the wafer. The inquiry full-angle α2 is an angle measured counterclockwise from the reference direction Ds of the stage coordinate system to the reference direction Dw2 of the wafer coordinate system. The reference direction Ds of the wafer coordinate system is defined to be a direction toward the three o'clock direction of the clock when the matching pattern MPc is upright (in the direction of the first template in FIG. 16C). Since the preliminary rotation angle α2pr has been determined only at the straight portion of the wafer, in this example, the preliminary rotation angle α2pr and the inquiry total angle α2 differ by 180 degrees. Of course, these angles α2pr and α2 may be the same.

The measurement position in the wafer under measurement can be determined at the clamping angle α2 and the center position of the wafer. However, by measuring the rotation angle relative to the reference wafer more accurately, the measurement position can be determined more accurately. For this reason, in step S21, the coordinate of the reference point Qc (FIG. 17) of the detected matching pattern MPs is calculated and registered.

In step S22, the XY stage 36 is moved so that the intersection position of the scribe line adjacent to the connection direction (direction of angle (theta) 1) of the reference point calculated | required from the reference wafer is imaged, and an image is imaged. At this time, the connection direction of the reference point in the wafer under measurement is obtained from the difference between the inquiry full-angle α1 of the reference wafer and the inquiry full-angle α2 of the wafer under measurement. The stage coordinates for making the intersection position of the second scribe line a visual field are calculated from the coordinates of the first intersection position Pc and the coordinates of the two intersection positions Pa and Pb in the reference wafer.

18 is an explanatory diagram showing two viewing relationships established in the wafer under measurement. The positional relationship between the intersection positions Pc and Pd of the two scribe lines of the wafer under measurement is the same as the positional relationship of the intersection positions Pa and Pb of the two scribe lines of the reference wafer in the wafer coordinate system. Therefore, the second intersection position Pd is present in the direction along the direction of the straight line L2 corresponding to the straight line L1 of the reference wafer at the first intersection position Pc. The movement amount for moving the XY stage 36 to the second intersection position Pd is equal to the difference between the coordinate values of the two reference points Qa and Qb of the reference wafer. In this way, the second field of view Wd in FIG. 18 is set, and the swap image is read.

In step S23 of FIG. 14, pattern matching with the model pattern MPa is performed on the image in this field of view Wd.

The coordinates of the reference point Qd of the matching pattern MPd that most closely match the template image MPa are obtained. 2

The coordinates of the center point Qcd of the dog reference points Qc and Qd are also calculated.

In step S24, the rotation direction determining means 158 (FIG. 2) is connected to the connection direction DL ₂ of the two reference points Qc and Qd.

The rotation angle θ2 is obtained. The rotation angle θ2 is an angle measured counterclockwise from the reference direction Ds of the stage coordinate system to the connection direction DL2 of the reference point. By using the rotation angle θ2 of the connection direction DL2 of the reference point on the wafer under measurement and the connection direction DL1 of the reference point on the reference wafer, the relative rotation angles of both can be determined with high accuracy.

19 is an explanatory diagram showing a method of obtaining a high precision relative rotation angle. FIG. 19A shows a straight line L1 connecting two reference points Qa and Qb obtained with respect to the reference wafer. The connection direction DL1 which connects these reference points Qa and Qb is taken in the direction toward the 1st reference point Qa from 2nd reference points Qa and Qb. The rotation angle θ1 of the connection direction DL1 is an angle measured counterclockwise from the reference direction Ds of the stage coordinate system to the connection direction DL1. FIG. 19B shows a straight line L2 connecting two reference points Qc and Qd obtained for the wafer under measurement. The connecting direction DL2 connecting these reference points Qa and Qd is also taken in the direction from the second reference point Qd to the first reference point Qc. The rotation angle θ2 of the connection direction DL2 is also an angle measured counterclockwise from the reference direction Ds of the stage coordinate system to the connection direction DL2. In this manner, the rotation angle θ1 of the connection direction DL1 on the reference wafer and the rotation angle θ2 of the connection direction DL2 on the wafer under measurement are also determined according to the same definition. Therefore, by calculating these differences Δθ = θ2-θ1, this can be employed as a relative rotation angle between the wafer and the wafer under measurement. By the way, another method can also be considered as a method of determining the rotation angle (rotation direction) of the to-be-measured wafer. 20 is a descriptive explanatory diagram of the approximate relative rotation angles using the inquiry full angles a1 and a2. The inquiry full angles a1 and a2 are angles measured counterclockwise from the reference direction Ds of the stage coordinate system to the reference directions Dw1 and Dw2 of the wafer coordinate system. The difference Δα = α2-α1 of the inquiry angle can be made to be a relative rotation angle between the reference wafer and the wafer to be measured. However, since the rotation angles θ1 and θ2 described above have higher precision than the inquiry angles α1 and α2, the relative rotation angle Δθ is also higher than the relative rotation angle Δθ determined by the inquiry full angle. High precision

As the method for determining the rotation angle (rotation direction) to be measured by the rotation direction determining means 158, various methods as follows can be considered, including the above-described method.

Method 1: The relative rotation angle (rotation direction) of both sides is determined from the difference Δθ between the high precision rotation angle θ1 of the reference wafer and the high precision rotation angle θ2 of the wafer under measurement. This method 1 shows FIG. According to this method, there is an advantage that the relative rotation angle (rotation direction) can be determined with high precision.

Method 2: The relative rotation angle (rotation direction) of the reference wafer is determined from the difference [Delta] [alpha] between the inquiry full angle [alpha] 1 of the reference wafer and the inquiry full angle [alpha] 2 of the wafer under measurement. This method 2 is shown in FIG. When this method is used, pattern matching may be performed on at least one image on the reference wafer or the wafer under measurement. Therefore, there is an advantage that the processing can be speeded up.

Method 3: The high precision rotation angle θ2 itself of the wafer under measurement is used as the rotation angle (rotation direction) of the wafer under measurement. As can be seen from FIG. 19 (B), the high precision rotation angle θ2 is the rotation angle from the reference direction Ds of the stage coordinate system to the connection direction DL2 of the wafer coordinate system. Therefore, it is possible to think that the wafer under measurement is rotating by θ2 in the reference direction Ds of the stage coordinate system. In addition, the method 3 may add a fixed value to the high-precision rotation angle (theta) 2 as a deformation | transformation, or may make it the rotation angle (rotation direction) of a wafer under measurement. According to this method 3, not only the relative rotation angle in the reference wafer but also the rotation angle (rotation direction) based on the predetermined reference direction Ds of the stage coordinate system can be determined with high accuracy.

Method 4: The inquiry full angle α1 of the wafer under measurement is set as the rotation angle (rotation direction) of the wafer under measurement. Also in this case, the value obtained by adding or subtracting a fixed value to the rotation angle α2 may be the rotation angle (rotation direction) of the wafer under measurement as in the method 3. This method 4 has the advantage that it is possible to determine not only the relative rotation angle on the reference wafer but also the rotation angle (rotation direction) based on the predetermined reference direction Ds of the stage coordinate system at high speed.

Moreover, this invention is not limited to the said Example and embodiment, It is possible to implement in various aspects in the range which does not deviate from the summary, for example, the following modification is also possible.

(1) In the above embodiment, when obtaining highly accurate rotation angles θ1 and θ2, two matching patterns were detected by performing pattern matching in two imaging areas. However, the imaging pattern for pattern matching is not limited to two, but the matching pattern may be detected in three or more imaging regions lined up in the same direction (search direction). In this case, the straight line L2 connecting each reference point of three or more matching patterns may be obtained by the least square method. In this way, the rotation angles θ1 and θ2 can be determined with higher accuracy.

(2) A predetermined point on a straight line L1 that connects two reference points Qa and Qb of the matching patterns MPa and MPb of the reference wafer (for example, the center point Qab of the reference point shown in FIG. 19A). The relative positional coordinates of the reference wafer and the wafer to be measured may be obtained from the coordinates and the coordinates of the point (the center point Qcd shown in FIG. 19B) of the wafer under measurement. Relative position coordinates as well as angles can be obtained.

(3) In the above embodiment, the case where the rotation mechanism of the wafer is not provided in the apparatus is described, but the present invention can be applied to the apparatus having the rotation mechanism. According to the present invention, since the rotation direction (rotation angle) of the wafer can be detected by image processing even in the apparatus having the rotation function, there is an advantage that the detection process in the rotation direction can be processed easily and at high speed. .

C. Alignment

C-1. Overview of justification:

21 is an explanatory diagram showing an outline of alignment processing in the alignment processing embodiment. The reference wafer (Fig. 21 (A)) is a wafer on which a pattern similar to the wafer under measurement (Fig. 21 (B)) to be subjected to the alignment process is formed. Generally, one sheet of a plurality of wafers processed in the same lot is used as the reference wafer WF1, and the other wafer becomes the wafer to be measured WF2. Positioning reference points and a plurality of measurement points PM1 to PM15 (indicated by white circles) are set on one wafer. As also shown in Figs. 21A and 21B, when the wafer is placed on the XY stage 36, any rotation direction can be taken. Since the measuring apparatus shown in FIG. 1 does not have a rotating mechanism for rotating the wafer mounted on the XY stage 36, the rotation direction of the wafer is recognized and corrected by image processing.

22 is a flowchart showing the overall procedure of alignment processing in the alignment processing embodiment. The alignment process of the wafer measurement points includes prealignment preprocessing (step T1) using the reference wafer, fine alignment preprocessing (step T2) using the reference wafer, and prealignment processing using the wafer under measurement (step). T3) and the pile alignment process (step T4) using the to-be-measured wafer. The alignment process in steps T1 and T3 is executed in cooperation with other means 150, 152, 154, 156, 158 other than the measurement positioning means 162 under the control of the reference positioning means 160 (FIG. 2). The fine alignment processing in steps T2 and T4 is executed by the imaging positioning means 152 and the maton matching means 154 under the control of the measurement positioning means 162.

In the pre-alignment preprocessing using the reference wafer WF1 in step T1 of FIG. 22, prealignment information including the rotation angle θ1 of the reference wafer WF1 and the position of the alignment reference point RP is registered. The rotation angle θ1 of the reference wafer WF1 is an angle formed of the reference direction Ds of the wafer coordinate system and the predetermined direction DL1 of the reference wafer WF1. The reference direction Ds of the wafer coordinate system is a direction fixed with respect to the XY stage 36, and is set to, for example, the X direction on the XY stage 36. The predetermined direction DL1 of the reference wafer WF1 is a fixed direction with respect to the reference wafer WF1. The determination method of this direction DL1 is mentioned later. In addition, the setting method of these directions Ds and DL1 is arbitrary, and definition other than this is also possible.

In step T1 of FIG. 22, first, the first model pattern for pattern matching is acquired near the alignment reference point RP. In addition, pattern matching is performed in another region near the alignment reference point RP to determine the position of the matching pattern that substantially matches the first model pattern. Based on the position of the first model pattern, the rotation angle θ1 of the reference wafer and the position of the alignment reference point RP are determined.

In the fine alignment preprocess using the reference wafer WF1 in step T2, the positions of the plurality of measurement points PM1 to PM15 on the reference wafer WF1 are registered. At this time, first, a second model pattern used for pattern matching of an image is obtained near a plurality of measurement points. Then, the positional relationship between the second model pattern and each measurement point is registered. At this time, the positional relationship between the position of the alignment reference point RP and each measurement point is also registered.

In the pre-alignment process using the to-be-measured webpee WF2 in step T3 of FIG. 22, the rotation angle (theta) 2 and the position of the alignment reference point RP of the wafer WF2 are determined. At this time, the pattern matching process is performed by using the first model pattern in two areas near the alignment reference point RP to determine the positions of the two matching patterns. Then, the positions of the wafer rotation angle θ2 and the alignment reference point RP to be measured are determined from the positions of these matching patterns.

In the fine alignment process using the wafer WF2 to be measured in step T4 of FIG. 22, the actual measurement position of the measurement point is determined by performing a pattern matching process using the second model pattern in the vicinity of the plurality of measurement points.

C-2. Pre-alignment pretreatment using reference wafers;

23 and 24 are flow charts showing the procedure of pre-alignment normal processing using the reference wafer WF1. In S1 of FIG. 23, the user inputs the chip dimensions of the wafer and the number of chips in the X-axis direction and the Y-axis direction.

Step S2 of FIG. 23 is similar to Step S2 of FIG. 3 except that the reference wafer WF1 is used instead of the wafer in Step S2 of FIG. 23, and FIG. 5 includes the characteristic pattern PT not found in the other three corners. The only difference in FIG. 3 is that the image portion to be used is used as the first model pattern instead of the model in pattern matching described later.

Steps S3 to S6 of FIG. 23 are the same as those of steps S3 to S6 of FIG. 3, respectively, and thus detailed description thereof is omitted here.

In step S7 of FIG. 23, image processing for rotating the image picked up in step S6 by the inquiry full-angle angle? 1 in the clockwise direction is performed. In step S8, the imaging positioning means 152 obtains the measured value of the exact position (coordinate value) of the intersection Pa of the scribe line SL, and saves it. The intersection position coordinates of the scribe line SL are used later to determine the alignment reference point of the reference wafer WF1. The intersection position of the scribe line SL of a reference wafer is represented by the coordinate of the center point Pa of the visual field Wa of the camera 41 shown, for example in FIG. The position of the point Pa can be specified by the user moving the cursor using a pointing device such as a mouse on the image displayed on the monitor 136. Or by processing the multi-gradation image image | photographed with the camera 41, it is also possible to automatically determine the coordinate of the center position of the intersection of the scribe line SL. When the center position of the intersection is obtained by image processing, firstly, the linear energy is detected in the same manner as in step S3 described above. And the straight line which approximated the edge to the scribe line SL is calculated | required. In addition, the four corners and the central position constituted by these approximate straight lines are obtained by approximating the edges to the scribe line SL. In addition, the four corner parts and the center position which consist of these approximation lines are determined as the intersection position of the scribe line SL. In addition, the coordinate of the center position of the visual field Wa is the coordinate (coordinate fixed to a stage) of the wafer coordinate system picked up by the stage coordinate reading part 35 (FIG. 1). The coordinates of the wafer coordinate system at any position in the visual field Wa (that is, the image picked up) can be easily calculated from this coordinate value.

The offsets? X and? Y of the coordinate values between the reference point Qa of the first model pattern MPa and the scribe line intersection Pa are stored in the alignment information file 139.

In step S9, the first model pattern (called a model pattern) for pattern matching is cut out from the image rotated in step S7 and registered. 25 is an explanatory diagram showing a registration form of the first model pattern MPa. In step S7, first, the multi-gradation image (second (A)) is rotated clockwise by the inquiry full angle α1 as shown in FIG. 25 (B) at the intersection of the scribe line SL, and the image after the rotation is rotated. Displayed on the monitor 136. Rotation of the image is performed by affine transformation. The user observes the displayed image to determine whether there is an image pattern that can be used as the model pattern MPa. The image pattern which can be used as the model pattern MPa means an image pattern which can select one of four equivalent rotation angles which are equivalent as inquiry full-angle angle (alpha) 1 from the direction of the image pattern. As the model pattern MPa, an image pattern without rotation symmetry of an integral multiple of 90 degrees is preferable. In other words, an image pattern having any one of the rotational symmetry of 90 degrees (the rotation symmetry of 90 degrees, 180 degrees, and 270 degrees) is inappropriate as the model pattern MPa. Since four adjacent corner portions are included in the visual field Wa near the intersection of the scribe line SL, a unique volcanic pattern contained only within one of these four corner portions is used as the first model pattern MPa. You can register. If there is no image pattern that can be used as the model pattern MPa in the current field of view Wa, the XY stage 36 is moved little by little while displaying and observing the image captured by the camera 41 on the monitor 136. Let's do it. And the image pattern which can be used as model pattern MPa is set to the state into which it enters into a visual field.

If there is an image pattern that can be used as the model pattern MPa in the current visual field Wa, as shown in Figs. 25 (B) and (C), registration as a model pattern MPa is made from the image after rotation. Cut out the area. The range of the model pattern MPa is specified by the user using a pointing device such as MauTM. The model pattern MPa is preferably present in the vicinity of the intersection of the scribe line SL, but may not necessarily be present in the vicinity of the intersection.

In step S10 of FIG. 23, the coordinates of the image of the first model pattern MPa and the reference point (for example, the upper left point Qa shown in FIG. 25C) at the predetermined position of the model pattern MPa are the magnetic disk 138. Is registered in the alignment information file 139 (FIG. 2). In addition, the coordinate of the reference point Qa is represented by the coordinate value of a wafer coordinate system, for example.

In step S11 of FIG. 24, the user rotates and sets the predetermined direction (e.g., the three o'clock direction of the clock) of the model pattern MPa as the reference direction (zero movement direction) Dw1 of the wafer coordinate system, so that Eliminate uncertainty For example, as shown in FIG. 25 (B), in the image rotated clockwise by the inquiry full-angle angle? 1, the 3 o'clock direction of the clock is set as the reference direction Dwl of the wafer coordinate system. In addition, the 3 o'clock direction of the clock may be automatically set to the reference direction (Dwl) without user designation. The rotation angle of the wafer is an angle from the reference direction Ds of the wafer coordinate system to the reference direction Dwl of the wafer coordinate system. Therefore, in the case of the 25th (B), the rotation angle of the reference wafer is equal to the inquiry full angle? 1. In addition, when the reference direction of the coordinate system of the wafer is selected in a direction other than the three o'clock direction of the clock, the rotation angle of the reference wafer is different from α1. However, also in this case, the rotation angle of the reference wafer is a value obtained by adding or subtracting a predetermined value to the inquiry full-angle α1. For example, in the state of Fig. 25 (B), when the 12 o'clock direction of the clock is selected as the reference direction of the wafer coordinate system, the rotation angle of the reference wafer becomes (? 1 + 90 degrees). In step S12 of FIG. 24, the value of this rotation angle (alpha) 1 is stored in the alignment information file 139. FIG.

In step S13, the XY stage 36 is moved so that an imaging area may be located at the scribe line intersection position of the adjacent chip, and an image is imaged. In step S14, an image pattern (matching pattern) similar to the first model pattern MPa is detected by performing pattern matching on this image. 26 is an explanatory diagram showing the process contents of steps S13 and S14. In this example, the visual field Wb is shifted from the intersection position where the first model pattern MPa is registered to the intersection position adjacent to the lower right side at an angle. The scribe line intersection positions of adjacent chips may be adjacent to each other in the vertical, horizontal and inclined directions. The matching pattern MPb matching the first model pattern MPa is detected in the image of this field of view (imaging region) Wb.

In step S14, after detecting the matching pattern MPb, the coordinate of the reference point Qb is also calculated. The direction of the straight line L1 connecting the reference points Qa and Qb of the two image patterns MPa and MPb to the first reference point Qa from the second reference point Qb (connection direction of the reference point). (DL1) is specified. The rotation angle (angle measured in the counterclockwise direction from the reference direction Ds of the coordinate system of the wafer) θ1 of the connection direction DL1 is calculated. In addition, since the coordinates of the reference points Qa and Qb are obtained as the coordinates of the wafer coordinate system, the rotation angle θ1 of the connection direction DL1 can be obtained from these coordinates by simple calculation.

In this embodiment, the rotation angle θ1 of the reference point connection direction DL1 is used as the rotation angle of the reference wafer instead of the inquiry full angle α1 shown in FIG. 25 (B). The difference between the two rotation angles α1 and θ1 is due to which direction is selected as the reference direction of the wafer coordinate system, and which may be defined as the rotation angle. However, there is an advantage that the rotation angle θ1 in the reference point connection direction of the image pattern can be determined with higher precision than the inquiry full angle? 1. The rotation angle θ1 shown in FIG. 21 (A) is the rotation angle defined by the connecting direction DL1 of this reference point.

In step S16 of FIG. 24, the position of the second scribe line intersection Pb (FIG. 13) is determined and stored in the alignment information file 139. FIG. For example, it is assumed that the positional relationship between the second scribe line intersection Pb and the matching pattern Mpb is the same as the positional relationship between the model pattern MPa of the first scribe line correction Pa. Therefore, the position of the second scribe line intersection Pb is based on the position of the reference point Qb of the matching pattern MPb and the relative position of the first scribe line intersection Pa and the reference point Qa of the model pattern MPa. Is calculated.

Alternatively, the position of the second scribe line intersection Pb may be determined by the same method as the determination direction of the first scribe line intersection Pa. That is, the user may designate the position of the second scribe line intersection Pb, or may automatically determine the position of the second scribe line intersection Pb by analyzing the image in the second field of view Wb. .

In step S17 of FIG. 24, the coordinates of the center point Pab of the two scribe line intersections Pa and Pb are calculated and stored in the alignment information file 139 as the alignment reference point (point RP in FIG. 21 (A)). . Since the coordinates of this alignment reference point Pab are determined from the coordinates of two scribe line intersections Pa and Pb in the diagonal directions of the grating | lattice prescribed | regulated by a scribe line, the position can be set with high precision.

As the alignment reference point, there are other various setting methods. For example, the center point Qab of the reference points Qa and Qb of the two image patterns MPa and MPb may be used as the alignment reference point. It is also possible to select one point among the scribe line intersections Pa and Pb and the reference points Qa and Qb as the alignment reference point.

The following information is registered in the alignment information file 139 by the preprocessing regarding the abnormal reference wafer.

(a) inquiry total angle α1 of the reference wafer and high-precision rotation angle θ1;

(b) image data of the first model pattern MPa;

(c) coordinate values of the model pattern reference points Qa and Qb;

(d) offsets δx and δy of coordinates from the model pattern reference points Qa and Qb to respective scribe line intersections Pa and Pb;

(e) Coordinate value of alignment reference point (Pab).

These pieces of information are information used to determine the correspondence between the stage coordinate system and the wafer coordinate system (called "coordinate system correspondence determination information"). By using the correspondence determination information for the coordinates, the stage coordinate system and the wafer coordinate system of the reference wafer WF1 can be coordinate-transformed with each other by an infine transformation. As will be described later, this coordinate system correspondence determination information can be used to obtain a correspondence relationship between the stage coordinate system and the wafer coordinate system of the wafer under measurement.

Since the stage coordinate system and the wafer coordinate system can be easily converted to each other by an affine transformation, some coordinate values in the coordinate system correspondence determination information may be registered as coordinate values of the stage coordinate system, and also registered as coordinates of the wafer coordinate system. You may also do it. In either case, if the conversion coefficient of the affine transformation is registered, the affine transformation can be easily performed.

C-3. Fine alignment pretreatment using a reference wafer;

FIG. 27 is a flowchart showing a procedure of fine alignment preprocessing using the reference wafer WF1. In the fine alignment preprocess, the position (relative positional relationship with the alignment reference point RP) of the plurality of measurement points PM1 to PM2 (indicated by the white circles in FIG. 21 (A)) on the reference wafer WF1 is registered as follows. do.

In step S21, a user specifies the position which includes a measurement point (for example, PM1 of FIG. 21 (A)) in a visual field, and moves the XY stage 36. FIG. 28 is an explanatory diagram showing a field of view W (i) including the i-th measurement point PMi. This visual field W (i) shows an image after being rotated by the inquiry full-angle angle? 1 of the reference wafer WF1 by affine switching.

In step S22 of FIG. 27, the user instructs the position of the first measurement point PMi on the screen shown in FIG. 28 using a pointing device such as a mouse. The coordinate values Xi and Yi of the measuring point PMi are stored in the alignment information file 139. In step S23, the user searches for an image pattern suitable for the second model pattern MRa in the vicinity of the measuring point PMi. It is preferable that the image pattern suitable as the second model pattern MRa is an image pattern without rotation symmetry of an integral multiple of 90 degrees.

If an appropriate image pattern is found, the user adjusts the position of the XY stage 36 so that the image pattern is in the center of the screen in step S24. In step S25, the image in the field of view is picked up, and the image is rotated by the inquiry full-angle α1.

In step S26, the user designates the area of the second model pattern MPa, thereby cutting out the image data of the second model pattern MRa and storing it together with the coordinates (x, y) of the reference point Ra.

The second model pattern MRa may be the same as the first model pattern MPa (25 (C)) used for prealignment preprocessing. For example, the first model pattern MPa near each measurement point. If this exists, instead of step S26, the pattern matching means 154 (FIG. 2) performs the pattern matching process using the 1st model pattern MPa, and determines the position of the reference point Ra.

In step S27, the offsets (ΔX, ΔY) of the X and Y coordinates from the reference point Ra of the second model pattern MRa to the i th measurement point PMi are obtained and stored in the alignment information file 139. . The position of the measuring point PMi is specified by the user. The offsets ΔX and ΔY of the coordinates are information indicating a positional relationship with the I-th measurement point PMi of the second model pattern MRa existing near the i-th measurement point PMi.

In step S28, it is determined whether there are other measuring points. If there are other measuring points, the process returns to step S21 and the processes of the above-described steps S21 to S27 are repeated. On the other hand, when the process of steps S21-S27 is complete about all the measurement points, the fine alignment preprocess in the existing wafer WF1 is complete | finished.

As the second model pattern MPa, one image pattern common to all measurement points may be used, or different image patterns may be registered for each measurement point.

In the fine alignment preprocessing of the reference wafer WF1 described above, the following information is registered in the alignment information file 139 with respect to the plurality of measurement points, respectively.

(a) Image data of the second model pattern MRa:

(b) the position of the reference point Ra of the second model pattern MRa (ie, the relationship between the reference point Ra and the alignment reference point RP);

(c) Offsets (ΔX, ΔY) of the coordinates with the reference point Ra measuring point PMi of the second model pattern MRa.

The above information is information used for determining the position of each measurement point PMi in the wafer coordinate system, and hereinafter referred to as "measurement positioning information". By using this measurement positioning information, the position of each measurement point on the wafer under measurement can be obtained.

Since the stage coordinate system and the wafer coordinate system can be coordinated with each other, the coordinate values included in the measurement positioning information may be registered as coordinate values of the stage coordinate system or may be registered as coordinates of the wafer coordinate system.

In addition, the reference wafer WF1 is subjected to predetermined processing such as various measurements using the apparatus shown in FIG. For example, when the apparatus shown in FIG. 1 is a film thickness meter, the thickness of the wafer surface is measured at the plurality of measuring points PM1 to PM15 in the reference wafer WF1.

C-4. Pre-alignment with wafers to be measured:

The same measurement process (for example, bar thickness measurement) is performed on the measurement target wafer WF2 at the same measurement point as the reference wafer WF1. However, when the measured wafer WF2 is placed on the XY stage 36, the rotation angle of the measured wafer WF2 is unclear, so that the XY stage (for example) can be positioned at the same measuring point as the reference wafer WF1. 36) cannot be moved. Therefore, before performing the measurement process on the wafer WF2 to be measured, the rotation angle and the position of the alignment reference point of the wafer WF2 are first determined by the pre-alignment process described below. The rotation angle and the alignment reference point position of the wafer under measurement WF2 may be regarded as information indicating a correspondence relationship between the wafer coordinate system of the wafer under measurement and the stage coordinate system. And the coordinate of each measuring point is correctly determined by the fine alignment process mentioned later.

29 and 30 are explanatory diagrams showing the procedure of pre-alignment processing of the wafer under measurement. The processing from steps S1 to S6 is the same as the prealignment preprocessing for the reference wafer shown in FIG. As a result, an image of the scribe line intersection near the center of the wafer is imaged. 15 shows an example of the field of view set on the wafer under measurement. In this case, an image is taken with the visual field Wc as the imaging area. In step S3 of FIG. 29, the inquiry full-angle angle? 2pr shown in FIG. 15 is detected. This inquiry full angle? 2pr has an uncertainty of an integer multiple of 90 degrees. In the wafer under test, the inquiry full-angle angle α2pr before removing the uncertainty is also called a "preliminary rotation angle". This name implies a preliminary angle of rotation that includes uncertainty.

In step S31, the pattern matching means 154 (FIG. 2) performs the pattern matching process using the 1st model pattern MPa registered by the preprocess of the reference wafer with respect to the image in this visual field Wc.

31 is an explanatory diagram showing a method of pattern matching for a wafer under measurement. First, the read image shown in FIG. 31 (A) is rotated clockwise by the preliminary rotation angle α2pr by an affine transformation to create an image as shown in FIG. 31 (B). Then, the image pattern matching the first pattern model MPa in the rotated image is detected by the pattern matching process. At this time, as shown in FIG. 31 (C), it is preferable to prepare four model patterns rotated by 90 degrees in advance. The model pattern having the highest matching degree is determined among these four model patterns, and the reference point coordinates of the image pattern (matching pattern) matched with this are determined. In the example of FIG. 31 (B), the matching degree of the model pattern of 180 degree rotation is the highest. Therefore, it is determined that the inquiry full-angle angle? 2pr of the measurement target wafer is (? 2pr + 180 degrees). In other words, the uncertainty of the preliminary rotation angle α2pr can be eliminated by pattern matching using the model pattern, thereby determining the value of the inquiry full angle α2pr. The angle selected by pattern matching among the angles (0 degrees, 90 degrees, 180 degrees, 270 degrees) related to the four model patterns which are rotationally symmetrical is referred to as " matching angle " below.

17 is an explanatory diagram showing the relationship between the preliminary rotation angle α2pr and the inquiry full angle α2 of the wafer under measurement. The preliminary rotation angle α2pr is an angle measured counterclockwise from the stage coordinate system reference direction Ds to the direction of the straight portion (scribe line SL) of the wafer under measurement. The inquiry full angle α2 is an angle measured counterclockwise from the reference direction Ds of the stage coordinate system to the reference direction Dw2 of the wafer coordinate system. The reference direction Dw2 of the wafer coordinate system is defined as the direction toward the 3 o'clock direction of the clock when the matching pattern MPc is established (in the direction of the first model pattern in FIG. 30C). The preliminary rotation angle α2pr is determined not only from the straight portion of the wafer, but in this example, the preliminary rotation angle α2pr and the inquiry total angle α2 differ by 180 degrees. Of course, these angles α2pr and α2 may be the same.

In step S32 of FIG. 29, the coordinates of the reference point Qc of the matching pattern MPc are stored in the alignment information file 139. In step S33 of FIG. 30, the uncertainty of the rotation angle of the wafer under measurement is eliminated, and the relative rotation angle Δα is obtained. The relative rotation angle Δα of the wafer under measurement is defined as the difference α2-α1 between the inquiry full-angle az of the non-measured wafer and the inquiry full-angle α1 of the reference wafer.

The position of each measuring point in the wafer under measurement can also be determined using this relative rotation angle Δα. In this embodiment, however, the position of each measurement point is determined more accurately by accurately calculating the relative rotation angle with the reference wafer in the following procedure.

In step 34, the position of the first scribe line intersection Pc (Fig. 17) in the vicinity of the matching pattern MPc is calculated from the result of pattern matching. 32 is an explanatory diagram showing the relationship between the reference point Qc of the matching pattern MPc and the first scribe line intersection Pc. As described above, in the pattern matching processing, it is confirmed that the image pattern is matched with any one of the four matching angles shown in Figs. 32 (a) to (d). Coordinates Xc and Yc of the scribe line intersection point Pc are respectively calculated as follows according to the matching angle.

(a) 0 degree matching angle

Xc = Xs + δx, Yc = Ys-δy

(b) 90 degree matching angle

Xc = Xs + δy, Yc = Ys-δx

(c) When the matching angle is 180 degrees

Xc = Xs-δx, Yc = Ys-δy

(d) When the matching angle is 270 degrees

Xc = Xs-δy, Yc = Ys + δx

Here, δx and δy are offsets of the coordinates between the reference point Qa (Fig. 13) of the model pattern MPa and the scribe line intersection point Pc in the vicinity thereof, which are obtained in the pre-alignment preprocessing of the reference wafer described above. to be. By using the offsets δx and δy of the coordinates, the coordinates of the scribe line intersection Pc can be calculated as described above at the reference point Qc of the matching pattern MPc. Moreover, since the relationship shown in FIG. 32 is in the state which rotated the wafer WF2 to be measured by the preliminary rotation angle (alpha) 2pr, the affine transformation was performed so that it may rotate by the preliminary rotation angle (alpha) 2pr as each coordinate value of said four formulas. The value is used.

In step S35 of FIG. 30, the wafer under measurement WF2 is moved to a position including the second scribe line intersection in the field of view. 18 is an explanatory diagram showing two viewing relationships established in the wafer under measurement. The positional relationship of the two scribe line intersections Pc and Pd of the wafer under measurement is the same as the positional relationship of the two scribe line intersections Pc and Pd of the reference wafer in the wafer coordinate system. Therefore, the second scribe line intersection Pd exists in the direction along the direction of the straight line L2 corresponding to the straight line L1 of the reference wafer at the first scribe line intersection Pc. The movement amount for moving the XY stage 36 to the second scribe line intersection Pd is equal to the difference between the coordinate values of the two reference points Qa and Qb of the reference wafer. In this way, the second field of view Wd in FIG. 25 is set.

In step S36 of FIG. 30, the image of the second field of view Wd is read, and the image is rotated by an affine transformation by the inquiry full angle α2, and pattern matching is performed on the image after the rotation. In this pattern matching, the coordinates of the reference point Qd (Fig. 18) of the matching pattern MPd that most closely matches the first model pattern MPa are obtained.

In step S27, the rotational angle θ2 of the connecting direction DL2 of the two reference points Qc and Qd is determined by the rotation direction determining means 158 (FIG. 2), and the reference point connecting direction (s) in the reference direction Ds of the stage coordinate system. The angle measured counterclockwise to DL2).

In step S38, the coordinates of the second scribe line intersection Pd are calculated from the coordinates of the reference point Qd of the second matching pattern MPd. This operation is as shown in FIG. In step S39, the coordinates of the center point Pcd (FIG. 18) of the scribe line intersection points Pc and Pd at 1 point and 2 points are calculated | required. This center point Pcd becomes the origin in the wafer coordinate system. Moreover, it is used as alignment reference point RP (FIG. 21 (B)) in the fine alignment process demonstrated below.

The relative rotation angle between the reference wafer WF1 and the wafer to be measured WF2 is the rotation angle θ2 of the connecting direction DL2 of the reference point in the wafer under measurement and the rotation angle θ1 of the connecting direction DL1 of the reference point in the reference wafer. Can be determined with high precision.

19 is an explanatory diagram showing a method of obtaining a high precision relative rotation angle. FIG. 19 (A) shows a straight line L1 connecting two reference points Qa and Qb obtained with respect to the reference wafer. The connection direction DL1 which connects these reference points Qa and Qb is taken in the direction toward the 1st reference point Qa from the 2nd reference point Qb. The rotation angle θ1 of the connection direction DL1 is an angle measured counterclockwise from the reference direction Ds of the stage coordinate system to the connection direction DL1.

19B shows a straight line L2 connecting two reference points Qc and Qd obtained with respect to the wafer under measurement. The connecting direction DL2 connecting these reference points Qc and Qd is also taken in the direction from the second reference point Qd toward the first reference point Qc. The rotation angle θ2 of the connection direction DL2 is an angle measured counterclockwise from the reference direction Ds of the stage coordinate system to the connection direction DL2. In this manner, the rotation angle θ1 of the connection direction DL1 in the reference wafer and the rotation angle θ2 of the connection direction DL2 in the wafer under measurement are also determined according to the same definition. Therefore, by calculating these differences Δθ = (θ2) − (θ1), this can be employed as the relative rotation angle between the reference wafer and the wafer under measurement. By the way, another method can also be considered as a method of determining the rotation angle (rotation direction) of the wafer under measurement. 20 is an explanatory diagram showing a method of determining an approximate relative rotation angle using the inquiry full angles α1 and α2. The inquiry full angles α1 and α2 are angles measured counterclockwise from the reference direction Ds of the stage coordinate system to the reference directions Dw1 and Dw2 of the wafer coordinate system. Therefore, the difference [Delta] [alpha] = [alpha] 2- [alpha] 1 of the inquiry full angle can be set as the relative rotation angle between the reference wafer and the wafer under measurement. However, since the rotation angles α1 and α2 described above are higher than the inquiry angles α1 and α2, the relative rotation angle Δθ is also higher than the relative rotation angle Δα determined by the inquiry full angle. . As the method for determining the rotation angle (rotation direction) of the wafer under measurement, the rotation direction determining means 158 includes various methods as described below including the above-described method.

Method 1: From the difference [Delta] [theta] between the high precision rotation angle [theta] 1 of the reference wafer and the high precision rotation angle [theta] 2 of the wafer under measurement, the relative rotation angle (rotation direction) of both is determined. This method 1 is shown in FIG. According to this method, there is an advantage that the relative rotation angle (rotation direction) can be determined with high accuracy.

Method 2: The relative rotation angle (rotation direction) of the reference wafer is determined from the difference Δα between the inquiry full-angle α1 of the reference wafer and the inquiry full-angle α2 of the wafer under measurement. This method 2 is shown in FIG. When this method is used, pattern matching may be performed on at least one image on the reference wafer or the wafer under measurement. Therefore, there is an advantage that the processing can be speeded up.

Method 3: The high precision rotation angle θ2 of the wafer under measurement is itself used as the rotation angle (rotation direction) of the wafer under measurement. As can be seen from FIG. 19 (B), the high precision rotation angle θ2 is the rotation angle from the reference direction Ds of the stage coordinate system to the connection direction DL2 to the wafer coordinate system. Therefore, it is possible to think that the wafer to be measured is rotated by θ2 in the reference direction Ds of the stage coordinate system. As a modification of Method 3, a value obtained by adding or subtracting a fixed value to the high precision rotation angle θ2 may be used as the rotation angle (rotation direction) of the wafer under measurement. According to this method 3, not only the relative rotation angle in the reference wafer but also the rotation angle (rotation direction) based on the reference direction Ds of the stage coordinate system can be determined with high accuracy. In particular, when the coordinates of the alignment reference point and the plurality of measurement points of the reference wafer are previously converted to the coordinates of the stage coordinate system, the rotation angle θ2 can be used as it is as the rotation angle of the wafer under measurement.

Method 4: The inquiry full angle? 2 of the wafer under measurement is itself the rotation angle (rotation direction) of the wafer under measurement. Also in this case, the value obtained by adding or subtracting a fixed value to the rotation angle α2 may be the rotation angle (rotation direction) of the wafer under measurement as in the method 3. According to this method 4, not only the relative rotation angle on the reference wafer is used. There is an advantage that the rotation angle (rotation direction) based on the reference direction Ds of the stage coordinate system can be determined at high speed. However, since the position and rotation angle θ2 of the alignment reference point RP (FIG. 21) were determined by pre-aligning the wafer under measurement, the coordinates of the stage coordinate system and the wafer coordinate system of the wafer under measurement were used using these. Coordinate transformation can be performed. By performing such coordinate transformation, the coordinates of each measuring point of the wafer under measurement can be predicted. However, when determining the coordinates of each measurement point from the result obtained by the pre-alignment process, there are the following problems.

The first problem is the rotation angle θ2 error of the wafer under measurement. As described above, the rotation angle θ2 is determined at the positions of the reference points Qc and Qd obtained by two pattern matching. Therefore, an error occurs in the value of the rotation angle θ2 due to the alignment error in pattern matching. Because of this error, the measurement point present at a position far from the alignment reference point RP has a problem that the gap between the predicted position of the measurement point predicted by the coordinate transformation and the actual position (actual position) becomes large, resulting in poor alignment accuracy. It is as large as the measurement point far from the alignment reference point RP due to the error of the rotation angle θ2.

In particular, when the coordinates of the alignment reference point and the plurality of measurement points of the reference wafer are previously converted to the coordinates of the stage coordinate system, the rotation angle θ2 can be used as it is as the rotation angle of the wafer under measurement.

The second problem is the influence of the mechanical precision of the XY stage 36. Due to the mechanical precision of the XY stage 36, an error of orthogonality of the stage coordinate system and an error of nonlinear distortion (bending) occur. Due to these errors, the accuracy of positioning each measurement point is poor. The error due to this mechanical precision is also as large as the measurement point at the alignment reference point RP.

The third problem is the effect of the warping of the wafer itself. Since the shape taken from the reference wafer and the wafer under measurement is different, a alignment error occurs due to this difference. As described above, since the alignment error of the measurement point is caused by various causes, the position of each measurement point is accurately determined in the fine alignment process described below in this embodiment.

C-5. Fine alignment using wafers to be measured:

34 is a flowchart showing a procedure of fine alignment processing using the wafer under measurement. In step S41, the coordinates of the first measurement point are predicted using the information (pre-alignment information) obtained by the pre-alignment process, and the XY stage 36 is moved to the position of the model pattern (matching image) in the vicinity of the measurement point. The coordinates of the measuring point in the stage coordinate system are determined by the relative positional relationship between the rotation angles α2 and θ2 of the wafer under measurement and the stage coordinates of the alignment reference point RP and the alignment reference point RP on the reference wafer and the respective measuring points. Can be predicted based on

The predicted value of such a measurement point coordinate is called "prediction position" or "prediction value" below. In addition, the value of the measurement point coordinate determined by the fine alignment process demonstrated below is called "measured position" or "measured value".

In step S42, the image of the vicinity of a measuring point is picked up, and an image is rotated by the inquiry full angle (alpha) 2 calculated | required by the pre-alignment process. 35 is an explanatory diagram showing an image after rotation. This image is almost the same as the image of FIG. 28 in the pre-alignment process of the reference wafer mentioned above.

In step S43 of FIG. 34, the pattern matching using the second model pattern MRa (FIG. 18) is performed to determine the matching pattern MRb (FIG. 35). In step S44, the coordinate of the reference point Rb of is determined. In step S45, the actual position of the measurement point PMi is determined by obtaining the coordinates of the reference point Rb from the reference wafer and correcting them as offsets ΔX and ΔY of the coordinates of the measurement point PMi and the reference point Ra. The actual position of this measuring point PMi is converted into the coordinate of a stage coordinate system. Therefore, the measuring probe can be accurately determined at the measuring point PMi in accordance with the measured position of the measuring point PMi.

In step S46 of FIG. 34, it is determined whether there are other measuring points. If there are other measuring points, the process returns to step S41 and the processes of steps S41 to S45 are repeated. In this way, by repeatedly executing the processes of steps S41 to S45 for the plurality of measuring points PM1 to PM15 (FIG. 21) on the wafer to be measured, each measuring point measured position can be determined accurately. The measurement distance (eg film thickness measurement) at each measurement point can be performed between steps S45 and S46. Alternatively, steps S41 to SS45 may be repeatedly performed for all measuring points, and then measurement processing may be performed at each measuring point.

As described above, in the above embodiment, the coordinates of the alignment reference point RP (= Pcd) of the wafer under measurement are determined by the pre-alignment process and the rotation angles α2 and θ2 are determined. Then, the positions of the plurality of measurement points of the wafer under measurement were predicted according to the positional relationship between the prealignment information, the alignment reference point RP registered on the reference wafer, and each measurement point. Further, by performing fine alignment at this predicted position, the actual measurement position of each measurement point was determined accurately. Therefore, it is possible to reduce various alignment errors as described above and to position the measuring point with high accuracy.

C-6. Second Embodiment

In the first embodiment described above, fine alignment processing is performed for each measurement point, so that high-precision positioning can be performed at all measurement points, but requires considerable processing time. However, if the predicted position of the measuring point of the wafer under measurement calculated from the pre-alignment information has the necessary alignment accuracy, it is possible to omit the fine alignment process.

However, as described in the first embodiment, in the pre-alignment process, pattern matching is performed in the vicinity of two scribe line intersections Pc and Pd near the wafer center to obtain two matching reference points Qc and Qd, respectively. Was deciding. Moreover, the position of alignment reference point RP (= Pcd) was calculated | required from the position of two scribe line intersection points Pc and Pd. At this time, there are some errors in the positions of the reference points Qc and Qd of the matching pattern. Therefore, as the measurement point moves away from these matching patterns, it is considered that the amount of misalignment (alignment error) caused by the matching error increases.

36 is an explanatory diagram showing a relationship between the distance from the alignment reference point RP to the measurement point PMi and the alignment error. For simplicity, we can ignore errors other than pattern matching here. At this time, for example, as shown in FIG. 36 (B), the distance from the alignment reference point RP to the measurement point PMi is about 30 mm, and the alignment error in the vicinity of the measurement point (between the predicted position and the measured position) Difference) is assumed to be ± 5 μm. This alignment error is mainly caused by the error (delta) of rotation angle (theta) 2. Thus, the alignment error at each measurement point is proportional to the distance at the alignment reference point RP. For example, the alignment error is ± 16.7 μm at the measuring point PM1 about 100 m away from the alignment reference point RP. For example, assuming that the allowable alignment accuracy is ± 10 μm, the required alignment accuracy is not satisfied at a position 100 mm apart, but the required accuracy can be satisfied up to a position about 60 mm apart.

Therefore, in the second embodiment, when the distance from the alignment reference point RP to the measurement point PMi is less than or equal to the predetermined value Lmax, the fine alignment process is omitted and the prealignment information (the position and rotation angle α2, The predicted position calculated using [theta] 2) is used as the measured position of the measuring point PMi. On the other hand, when the distance from the alignment reference point RP to the measurement point PMi is larger than the predetermined value Lmax, fine alignment process is performed with respect to the measurement point PMi, and the actual measurement position is determined.

For example, as shown in Fig. 36 (A), 4 in the range of 60 mm (inside of the circle drawn with broken lines in the drawing) on the wafer within the radius of the positioning reference point RP among 15 measurement points of PMl to PM15. For the two measuring points PM5, PM8, PM11, and PM13, the fine alignment process is omitted and the measurement point actual position is determined by the calculation using only the pre-alignment information. On the other hand, for other measurement points, file alignment processing is performed to determine the actual position.

In this second embodiment, since the file alignment process is omitted for the measurement point within the predetermined range of the alignment reference point RP, the processing time of the entire alignment process can be shortened while satisfying the alignment accuracy at each measurement point. .

C-7. Third Embodiment

Also in the third embodiment, as in the second embodiment, it is determined whether or not fine alignment processing is performed in accordance with the position of the measurement point. However, the difference from the second embodiment is that the fine alignment processing is omitted for the other measuring points within the predetermined range from the measuring points on which the fine alignment processing has been performed.

37 is an explanatory diagram showing a processing method in the third embodiment. The measurement point position is the same as in Fig. 36 (A). In the third embodiment, it is assumed that after performing the pre-alignment process, the station is executed while aligning 15 measurement points of PM1 to PM15 in this order. In addition, when the distance between the i-th measuring point PMi and the measuring point which performed the fine alignment process previously is less than predetermined value Lmax (60 mm), the position which is prescribed | required even if the fine alignment process of the measuring point PMi is abbreviate | omitted Assume that a fitting precision is obtained.

In Fig. 37, since the distance from the alignment reference point RP to the chlchdml measurement point PM1 is larger than the predetermined value Lmax ((60 mm), fine alignment processing is performed on this measurement point PM1 to determine the actual position. The range Lmax at the measuring point PM1 is indicated by a broken line in the drawing The distance from the first measuring point PM1 to the second measuring point PM2 is less than or equal to Lmax, so for the second measuring point PM2 the fine alignment In the following, it is determined whether or not fine alignment processing is omitted for each measuring point, so that the measuring points at which fine alignment processing is performed are shown as measuring points (PM1, PM4, PM6, PM8, PM10, PM12 and PM13), and fine alignment processing can be omitted at other measurement points. To the call referred to as "fine alignment measurement point not" fine alignment process is performed the measurement point the measurement point is referred to as [a fine alignment measurement point.

The actual position of the fine alignment omitted measurement point is determined by ensuring the predicted position of the fine alignment omitted measurement point based on the amount of the misalignment between the predicted position and the actual position. The predicted position of the fine aligned measurement point is obtained by coordinate transformation using prealignment information. As a method of correcting a coordinate value, for example, there is a method of obtaining an affine transformation coefficient indicating a correspondence relationship between a predicted position of a fine aligned measuring point and a measured position, and affine transforming the predicted position of a fine alignment omitted measurement point using the affine transformation coefficient. Alternatively, the alignment reference point position and the rotation angle of the wafer may be corrected from the amount of the misalignment between the predicted position of the fine aligned measurement point and the actual measurement position, and based on these, the actual position of the fine alignment omitted measurement point may be calculated.

Also in the third embodiment, fine alignment processing may not be performed on the measurement point where the distance from the alignment reference point RP is equal to or less than Lmax, as in the second embodiment.

As described above, in the third embodiment, the fine alignment processing is omitted for the measurement points within a predetermined range from the fine aligned measurement points, and the prediction position of the measurement points is corrected based on the staggered amount between the prediction position and the actual measurement position of the fine alignment measurement points. It was made. Therefore, the number of measuring points which perform fine alignment processing can be reduced. As a result, it is possible to shorten the processing time of the entire alignment process while satisfying the alignment accuracy at each measurement point.

In addition, this invention is not limited to the said Example and the form, It is possible to implement in various aspects in the range which does not deviate from the summary, For example, the following modification is also possible.

(1) In the above embodiment, a part of the configuration realized by hardware may be replaced by software, or conversely, a part of the configuration realized by software may be replaced by hardware.

(2) The first pattern matching for determining the coordinates of the alignment reference point RP may be performed with respect to the image of the area near the alignment reference point RP, and pattern matching is performed on the image of the area including the alignment reference point RP. There is no need to do it. Similarly, the second pattern matching for determining the coordinates of each measurement point does not need to be pattern-matched for the image of the area including each measurement point, but may be performed for the image in the area near each measurement point.

(3) In the above embodiment, pattern matching is performed in two areas when determining the position of the alignment reference point RP and the rotation angle θ2. However, the pattern matching is performed in one area by the pattern matching. The position and rotation angle θ2 may be determined. Alternatively, by performing pattern matching on three or more regions, the position of the alignment reference point RP and the rotation angle θ2 may be determined. In addition, when determining the position of each measuring point, pattern matching may be performed with respect to a plurality of areas. In general, as the number of regions for pattern matching increases, the accuracy of coordinates and rotation angles determined thereby increases.

(4) In the above embodiment, the case where the rotating mechanism of the wafer of the apparatus is not provided has been described, but the present invention can also be applied to the apparatus having the rotating mechanism. According to the present invention, since the rotation direction (rotation angle) of the wafer can be detected by image processing even in an apparatus having a rotation function, there is an advantage that the measurement point positioning process is simple and can be processed at high speed.

Claims (17)

A method of detecting the rotational direction of a wafer under measurement mounted on a table, comprising the steps of: (a) preparing a model pattern used for pattern matching of an image; (C) processing the first image to detect a first straight line portion containing the first image, and rotating the wafer under measurement on the table from the direction of the first straight portion; Determining a first set consisting of four equivalent rotation angles different from each other by an integral multiple of 90 degrees as a rotation angle that can be rotated; and (d) performing pattern matching on the first image with the model pattern to perform the first image. Detecting a matching pattern therein, and selecting one of the first sets of the four equivalent rotation angles as a first direction indicator angle based on the direction of the matching pattern; (e) the first Depending on the direction indicators angle rotation method for detecting orientation of a wafer to be measured mounted on a table, it characterized in that it comprises a step of determining the rotation direction of the measurement on the wafer table. The process according to claim 1, wherein the step (a) comprises: (1) providing a reference wafer having a surface substantially the same as the surface of the wafer under measurement on the table, and (2) a part of the imaging area of the reference wafer surface. And (3) processing the second image to detect a second straight portion included in the second image, and from the direction of the second straight portion, the reference wafer on the table Determining a second set consisting of four equivalent rotation angles that are rotatable rotation angles that are different from each other by an integral multiple of 90 degrees, (4) extracting a model pattern from a portion of the second image, and (5) And a step of selecting one of said second sets of said four equivalent rotation angles as a second direction indicator angle, based on the direction of said model pattern, and said step (e) comprises: said first and second 2 direction indicator angle By calculating this rotation direction detection method of a wafer to be measured it mounted on a table, comprising the step of determining a relative rotation angle of the wafer to be measured with respect to the reference wafer. A method of detecting a rotational direction of a wafer under measurement, comprising: (a) preparing a model pattern used for pattern matching of an image; and (b) a first image of a first imaging area that is a part of the surface of the wafer under measurement. (C) processing the first image to detect a first linear portion included in the first image, and the wafer under measurement on the table can rotate from the direction of the first linear portion. Determining a first set consisting of four equivalent rotation angles different from each other by an integer multiple of 90 degrees as the rotation angle; and (d) performing pattern matching on the first image with the model pattern to match the pattern in the first image. And detecting one of the first set of the four equivalent rotation angles as the first direction indicator angle based on the direction of the matching pattern, and (e) the first direction indicator angle. Specifying a predetermined search direction in the first imaging area, specifying at least one other imaging area away from the first imaging area by a predetermined distance from the first imaging area, and at the same time Image pickup; (f) performing pattern matching with the model pattern on the images of the respective imaging areas, and detecting a matching pattern; and (g) the first imaging arranged along the search direction. Determining a first connection direction for connecting the predetermined reference positions of the matching pattern detected in each of the plurality of imaging areas including the area to each other, and determining the rotational direction of the wafer under measurement based on the first connection direction. And a rotation direction detecting method of the wafer under measurement. The process according to claim 3, wherein the step (a) comprises: (1) providing a reference wafer having a surface substantially the same as the surface of the wafer under measurement on the table; and (2) a second part of the reference wafer surface. (3) processing the second image to detect a second straight portion included in the second image, and from the direction of the second straight portion, the reference on the table Determining a second set of four equivalent rotation angles different from each other by an integer multiple of 90 degrees as a rotation angle at which the wafer can rotate; (4) extracting the model pattern from a part of the second image; (5) selecting one of the second sets of the four equivalent rotation angles as the second direction indicator angle based on the direction of the model pattern; and (6) based on the second direction indicator angle. By the second A predetermined search direction is specified in the image area, at least one imaging area separated from the second imaging area by a predetermined distance along the search direction, and at the same time, an image of each of the specified imaging areas is imaged. And (7) performing pattern matching on the image of each imaging area on the reference wafer with the model pattern to detect a matching pattern; and (8) the second imaging area arranged along the search direction. And determining a second connection direction for connecting the predetermined reference positions of the matching pattern detected in each of the plurality of imaging areas including each other, wherein the step (g) includes the first and second direction indicators. Determining an angle of rotation of the wafer under measurement with respect to the reference wafer by obtaining an angle difference of the angle. Rotation direction detection method. 5. The method according to claim 4, wherein the reference wafer is based on coordinate values of the reference positions of the plurality of matching patterns on the reference wafer and coordinate values of the reference positions of the plurality of matching patterns on the measurement target wafer. And determining a relative relationship between the one coordinate system and the second coordinate system of the wafer under measurement. A method of determining the position of a measurement point on a wafer to be mounted mounted on a table, the method comprising: (a) registering a positional relationship between an alignment reference point and a measurement point on the surface of the wafer to be measured; Preparing a first model pattern used for pattern matching on an image near the alignment reference point, and a second model pattern used for pattern matching on an image near the measurement point; (b) the Imaging the first image in the first imaging area in the vicinity of the alignment reference point; and (c) performing a first process on the first image, the first process including pattern matching processing using the first model pattern and performing the measurement. Determining the rotational direction of the wafer and the position of the alignment reference point; (d) the position and the position of the rotational direction and the alignment reference point determined in step (c). A step of predicting the position of the measurement point based on the positional relationship between the fitting reference point and the measurement point, (e) imaging the second image in the second imaging area in the vicinity of the predicted position; And performing a second process including a pattern matching process using two model patterns on the second image to determine the measured position of the measurement point. How to decide. The process (e) and the process (f) according to claim 6, wherein the wafer to be measured is provided with a plurality of measuring points on the surface, and for a specific measuring point within a predetermined range from the alignment reference point among the plurality of measuring points. O), and further including the step of using the predicted position with respect to the specific measuring point obtained in the step (d) as an actual measuring position of the specific measuring point. How to determine the measurement position. The said measured wafer is provided with the said some measurement point on the said surface, The predetermined range in the 1st measurement point in which the process (e) and the process (f) were performed among the said some measurement point is carried out. The processing of steps (e) and (f) is omitted for the second measuring point within the range, and the predicted position of the second measuring point is determined based on the shift amount between the predicted position and the measured position at the first measuring point. And determining a measurement position of the second measurement point by correcting the measurement position of the wafer under measurement mounted on the table. The said process (c) includes the process of imaging the several said 1st image with respect to the several area | region near the said alignment reference point, The said process (d) is a said 1st some And a step of determining the rotational direction of the wafer under measurement and the position of the alignment reference point by performing the first processing on the image, respectively. 7. The process according to claim 6, wherein the step (c) (i) processes the first image to detect a first linear portion included in the first image, and from the direction of the first linear portion on the table Determining a first set of four equivalent rotation angles different from each other by an integer multiple of 90 degrees as a rotation angle at which the wafer under test can rotate; and (ii) pattern matching with the first model pattern with respect to the first image. Detecting a matching pattern in the first image and selecting one of the first sets of the four equivalent rotation angles as a first direction indicator angle based on the direction of the matching pattern; iii) determining the rotational direction of the wafer under measurement on the table in accordance with the first direction indicator angle. 11. The process according to claim 10, wherein the step (a) comprises: (1) providing a reference wafer having a surface substantially the same as the surface of the wafer under measurement on the table; And (3) processing the third image to detect a second straight portion included in the second image, and from the direction of the second straight portion, the reference wafer on the table Determining a second set consisting of four equivalent rotation angles which are different from each other by an integer multiple of 90 degrees as a rotation angle which can be rotated; (4) extracting the first model pattern from a part of the third image; (5) a step of selecting one of said second sets of said four equivalent rotation angles as a second direction indicator angle, based on the direction of said first model pattern, and said step (c) The first and second directions By calculating the difference between the rotational angle table method for detecting orientation of a wafer to be measured it mounted on a table, comprising the step of determining a relative rotation angle of the wafer to be measured with respect to the reference wafer. 7. The process according to claim 6, wherein the step (c) (i) processes the first image to detect a first straight line portion included in the first image, and from the direction of the first straight line portion, Determining a first set of four equivalent rotation angles different from each other by an integer multiple of 90 degrees as a rotation angle at which the wafer under test can rotate; and (ii) pattern matching with the first model pattern with respect to the first image. Detecting a matching pattern in the first image and selecting one of the first sets consisting of the four equivalent rotation angles as the first direction indicator angle based on the direction of the matching pattern; At least one other imaging direction specified in the first imaging area based on the first direction indicator angle, and spaced apart from the first imaging area by a predetermined distance along the search direction. Specifying the inverse and simultaneously picking up an image of each of the picked-up image capturing regions; and (iv) performing a pattern matching with the first model pattern on the image of the respective picked-up regions to detect a matching pattern. And (v) determine a first connection direction for connecting the predetermined reference positions of the matching pattern detected in each of the plurality of imaging areas including the first imaging area arranged along the search direction, to each other; 1. A method for determining a measurement position of a wafer under measurement mounted on a table, comprising the step of determining a rotation direction of the wafer under measurement based on a connection direction. 13. The process according to claim 12, wherein the step (a) comprises: (1) providing a reference wafer having a surface substantially the same as the surface of the wafer under measurement on the table; and (2) a third that is a part of the reference wafer surface. Imaging the third image in the imaging area; and (3) processing the third image to detect a second straight portion included in the third image from the direction of the second straight portion on the table. Determining a second set of four equivalent rotation angles different from each other by an integral multiple of 90 degrees as the rotation angle at which the reference wafer can rotate; and (4) extracting the first model pattern from a portion of the third image. (5) selecting one of said second sets of said four equivalent rotation angles as a second direction indicator angle based on the direction of said first model pattern, and (6) said second 2 direction indicator angle In this case, a predetermined search direction is specified in the third imaging area, at least one other imaging area spaced apart from the third imaging area by a predetermined distance along the search direction, and at the same time, Image pickup; (7) performing pattern matching on the image of each image pickup area on the reference wafer with the first model pattern to detect a matching pattern; and (8) along the search direction. And determining a second connection direction for connecting the predetermined reference positions of the matching pattern detected in each of the plurality of imaging areas including the arranged third imaging areas to each other, wherein the step (c) includes: Determining an angle of rotation of the wafer under measurement with respect to the reference wafer by obtaining an angle difference between the first and second connection directions. The wafer to be measured measurement location method mounted on the table. The method of claim 13, wherein the step (a) is based on coordinate values of the reference positions of the plurality of matching patterns on the reference wafer and coordinate values of the reference positions of the plurality of matching patterns on the wafer under measurement. And determining a relative relationship between the first coordinate system of the reference wafer and the second coordinate system of the wafer to be measured. An apparatus for detecting a rotational direction of a wafer under measurement mounted on a table, comprising: means for preparing a model pattern used for pattern matching, means for imaging a first image of a part of an imaging area on the surface of the wafer under measurement; The first image is processed to detect a first straight portion included in the first image, and an integer multiple of 90 degrees from each other as a rotation angle at which the wafer under measurement can rotate from the direction of the first straight portion. Means for determining a first set of four other equivalent rotation angles, and pattern matching on the first image with the model pattern to detect a matching pattern in the first image, based on the direction of the matching pattern Means for selecting one of the first sets of the four equivalent rotation angles as a first direction angle, and according to the first direction angle And a means for determining the rotational direction of the wafer under measurement on the table. An apparatus for detecting a rotational direction of a wafer under measurement, comprising: means for preparing a model pattern used for pattern matching, means for photographing a first image that is part of the surface of the wafer under measurement, and processing the first image The first linear portion included in the first image is detected, and four equivalent rotation angles different from each other by an integer multiple of 90 degrees as a rotation angle at which the wafer under measurement can rotate from the direction of the first linear portion. Means for determining a first set and performing pattern matching on the first image with the model pattern to detect a matching pattern in the first image, based on the direction of the matching pattern, as a first direction indicator angle. Means for selecting one of said first sets of said four equivalent rotation angles and said first imaging area based on said first direction indicator angle; Means for specifying a positive retrieval direction, specifying at least one other imaging area spaced apart from the first imaging area by the predetermined distance from the first imaging area, and imaging the image of each of the specified imaging areas; A matching pattern detected in each of a plurality of imaging areas including means for detecting a matching pattern by performing pattern matching with the model pattern on the image in the imaging area, and the first imaging area arranged along the search direction; Means for determining a first connection direction for connecting the predetermined reference positions of each other, and for determining a rotational direction of the wafer under measurement based on the first connection direction. Device. An apparatus for determining the position of a measurement point on a wafer under measurement mounted on a table, wherein the positional relationship between the alignment reference point on the surface of the wafer to be measured and the measurement point is registered in advance, and an image near the alignment reference point is obtained. Means for preparing a first model pattern used for pattern matching, a second model pattern used for pattern matching on an image near the measurement point, and a first imaging area near the alignment reference point Means for determining the rotational direction of the wafer to be measured and the position of the alignment reference point by performing a first process including means for photographing an image and a pattern matching process using the first model pattern on the first image. And the position of the rotation direction and the alignment reference point determined by the means for executing the first processing, and the alignment reference point. Means for predicting the position of the measurement point based on the positional relationship between the measurement point and the measurement point, means for imaging a second image in the second imaging area near the predicted position, and pattern matching processing using the second model pattern. And a means for determining the actual position of the measurement point by performing a second process including the second image on the second image. ※ Note: The disclosure is based on the initial application.