JP5273926B2 - Defect inspection equipment - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To efficiently and surely inspect a defect. <P>SOLUTION: This defect inspection device 1 conveys a substrate W aligned by a conveying part 2 to an end face inspection part 3, to be conveyed sequentially to the end face inspection part 3 and a plane inspection part 4, and inspects an end face, a surface and a reverse face. A notch position of the substrate W is detected based on an end face image obtained by the end face inspection part 3, followed to be converted into a coordinate system using the notch position as a reference. The substrate center with respect to the reference position, and the notch position in a rotational direction are extracted from an image acquired in the surface inspection or the reverse face inspection. The end face image, the surface image and the reverse face image are converted into images of a three-dimensional coordinate system, based on the notch position and the center position of the substrate, so as to be displayed three-dimensional-shapedly on a display 6. <P>COPYRIGHT: (C)2008,JPO&amp;INPIT

Description

  The present invention relates to a defect inspection apparatus used for inspecting defects on an end face of a substrate and on the front surface and the back surface.

For example, in a semiconductor wafer manufacturing process using a photolithography process, a patterned resist layer is formed after a predetermined film is formed on a substrate. If film unevenness or dust adheres when applying the resist, it may cause defects in the line width of the pattern after etching or defects such as pinholes in the pattern. For this reason, an operator visually observes all the substrates before etching to inspect for defects such as pinholes.
Here, when the defect inspection is performed visually, the inspection result is likely to vary due to the difference in the experience of the operator. Therefore, it is desirable that the defect inspection apparatus has a determination function. Furthermore, if an operator enters the clean room, dust may be generated. Therefore, it is preferable to enable observation with the substrate isolated from the operator.

Therefore, a conventional defect inspection apparatus images reflected light when a substrate such as a semiconductor wafer is irradiated with illumination light, and performs defect inspection of the substrate from image data obtained at this time, and also extracts defects by image processing. There is something configured to do. In this defect inspection apparatus, the operator and the substrate can be isolated. By performing defect extraction by image processing, variations in inspection by an operator are prevented.
As such a conventional defect inspection apparatus, for example, as disclosed in Patent Document 1, there is an apparatus in which an illuminating unit and an imaging unit are provided so as to be able to image the back side of a substrate. If the resist wraps around the back surface of the substrate and the peripheral edge of the back surface is raised, dust is attached, or the back surface is scratched, such a defect can be obtained by acquiring the back side image. Can be inspected.
Further, for example, as disclosed in Patent Document 2, there is a configuration in which a plurality of imaging units are arranged corresponding to the inclined portion of the end surface of the substrate so that the end surface of the substrate can be inspected.
International Publication No. 2003/027652 Pamphlet Japanese Patent Laid-Open No. 2001-221749

In recent years, in addition to the front and back surfaces of a substrate, the necessity for end face inspection has increased, and a defect inspection apparatus having a configuration as in Patent Document 1 and a configuration as in Patent Document 2 has been developed. However, in such a configuration, the defect information for each of the front surface, the back surface, and the end surface and the inspection screen are displayed separately. Therefore, the operator must check the defect by switching the screen and the like. It was complicated. Further, when the end face inspection is added, there is a problem that the time required for the inspection becomes longer.
The present invention has been made in view of such circumstances, and has as its main object to enable efficient and reliable defect inspection.

The present invention for solving the above-described problems includes an end surface inspection unit that acquires a two-dimensional image of the entire end surface of the substrate, a planar inspection unit that acquires at least one two-dimensional image of the entire front surface and the entire back surface of the substrate, and the plane A control unit that performs defect inspection on each of the front surface image and the back surface image acquired by the inspection unit, and each of the end surface images acquired by the end surface inspection unit, and creates a three-dimensional image of the substrate including the inspection results; A display unit for displaying the three-dimensional image, wherein the three-dimensional image is obtained by using at least one of the front surface image and the back surface image and the end surface image based on the entire structure data of the substrate. The defect inspection apparatus substrate is a three-dimensional image formed by being associated with a surface having a three-dimensional shape.
This defect inspection apparatus creates a three-dimensional image by associating the end face image with the front and back images when the end face inspection and the front and back surface inspection are performed. When there is a defect on the substrate, the defect is displayed corresponding to the actual position in the three-dimensional image.

  According to the present invention, by creating a three-dimensional image, the state of a defect and the position of the defect on the substrate can be easily confirmed. Since it becomes possible to continuously collect information on defects on the substrate surface layer such as the end face, front surface, and back surface, the inspection becomes easy and the inspection time can be shortened.

Embodiments of the present invention will be described in detail with reference to the drawings.
As shown in FIG. 1, the defect inspection apparatus 1 is capable of inspecting each of the defects on the front surface and the back surface of the substrate W, the transport unit 2 that carries the substrate W, the end surface inspection unit 3 that performs defect inspection on the end surface of the substrate W. A flat surface inspection unit 4, a control unit 5 that controls the entire apparatus, a display unit 6 that displays a result of defect inspection, and an operation unit 7 that is operated by an operator. The inspection target of the defect inspection apparatus 1 is a circular substrate such as a semiconductor wafer, but may be a substrate having another shape. In FIG. 1, solid arrows indicate movement of the substrate W, and broken arrows indicate data flow.

  The transport unit 2 is used to transport the substrate W to each of the end surface inspection unit 3 and the flat surface inspection unit 4, and includes a cassette carry-in / out unit 11, an aligner 12, and a transport arm 13. In the cassette carry-in / out section 11, a cassette (not shown) storing a plurality of substrates W is mounted by an operator or a robot. The transfer arm 13 has a configuration in which one substrate W is taken out from the cassette mounted on the cassette loading / unloading unit 11 and is sequentially transferred to the aligner 12, the end surface inspection unit 3, and the flat surface inspection unit 4. A joint robot is used. The aligner 12 includes a rotation stage, and includes a non-contact type sensor that detects an eccentric position of the substrate W and a rotation angle based on the notch. When the transport arm 13 transports the substrate W to the end surface inspection unit 3 or the flat surface inspection unit 4, the substrate W is positioned based on information on the eccentric amount and rotation angle of the substrate W from the aligner 12.

In the end surface inspection unit 3, a rotary stage 22 is attached to a base 21 so that the substrate W can rotate. The rotary stage 22 has a means for holding the substrate W, for example, a suction hole (not shown). Further, an area sensor camera 23 and an illuminating device 24 are arranged on the side of the rotary stage 22 so that the end face of the substrate W can be imaged.
The plane inspection unit 4 includes a frame-shaped holding member 31 that holds the peripheral edge of the substrate W. The holding member 31 is rotatable about the rotation shaft 32. The rotation shaft 32 extends substantially parallel to the substrate W through the center of the substrate W in plan view.

  An example of the holding member 31 is shown in FIG. The holding member 31 is formed in an octagonal ring shape in which the frame 33 has sufficient rigidity. One support portion 34 is provided on the inner peripheral side of the portion corresponding to the four opposite sides of each side of the frame 33. The support portion 34 is configured to hold the substrate W by a holding portion that projects the peripheral edge portion of the back surface of the substrate W to the inside of the frame 33 provided with a suction portion such as vacuum suction or electrostatic chuck. When it is necessary to take a wide image up to the back surface of the substrate W, a member for holding the end surface of the substrate W on the frame 33 may be provided. The inner periphery of the frame 33 is larger than the outer diameter of the substrate W, and the substrate W can be accommodated in the frame 33 in plan view. Further, the rotating shaft 32 is fixed to a pair of opposite ends of the frame 33. These rotation shafts 32 are set so that the axis centers pass through the center of the substrate W when the substrate W is supported by the frame 33 (substantially equal to the position of the center of gravity) and the center of the substrate W in the thickness direction. . When the holding member 31 is rotated around the rotation shaft 32, the substrate W is rotated together with the frame 33, and the front surface and the back surface can be selectively arranged.

  The holding member 31 is mounted on a stage (not shown), and can reciprocate the substrate W in the X-axis direction shown in FIG. Above the holding member 31, a line light source 36 and a line sensor camera 37 are disposed as an inspection optical system. The line light source 36 illuminates the substrate surface by irradiating parallel light in a line shape in a direction intersecting the X-axis direction in which the holding member 31 moves. The line light source 36 uses a light illumination angle θ0 for capturing a regular reflection (interference light) image of the front or back surface of the substrate W as an initial position, and when capturing an image other than the regular reflection of the front or back surface of the substrate W, It can be rotated within the range of the illumination angle θ1. The line sensor camera 37 is provided at a position facing the line light source 36 with respect to the normal line n, and images reflected light from the substrate W. The line sensor camera 37 has an imaging angle θ0 for capturing a specular reflection image of the front or back surface of the substrate W as an initial position. When capturing an image of scattered light or diffracted light other than specular reflection, the line sensor camera 37 falls within the range of the imaging angle θ2. It is supported so as to be rotatable at an angle.

The control unit 5 includes an image processing unit therein, receives information input by the operator using the operation unit 7, issues a command to each component of the defect inspection apparatus 1, and is connected to the area sensor camera 23 and the line. The image signal output from the sensor camera 37 is received and image data is created. Further, defect extraction is performed by performing image processing on the image data, and the extraction result is displayed on the display unit 6.
The display unit 6 is a display that displays defect inspection information and other information, and may be a liquid crystal display with a touch sensor.
The operation unit 7 is used to instruct a defect inspection operation, and includes, for example, a keyboard and a trackball.

Next, the operation of the defect inspection apparatus 1 will be described.
When performing the inspection, an operator or a robot mounts the cassette in which the substrate W is accommodated in the cassette loading / unloading unit 11. When the operator commands the operation unit 7 to start the operation, the defect inspection is started.
The transport arm 13 takes out one substrate W in the cassette by vacuum suction and transports it to the aligner 12. When the aligner 12 detects the position of the substrate W, the control unit 5 drives the transport arm 13 to correct the position of the substrate W. When the alignment of the substrate W is completed in this way, the transport arm 13 transports the substrate W to the end surface inspection unit 3.

The end surface inspection unit 3 performs a defect inspection on the periphery of the substrate W. When the substrate W is set, the rotary stage 22 holds the central portion of the back surface of the substrate W by vacuum suction. The illumination device 24 illuminates the end surface of the substrate W, and the area sensor camera 23 captures the reflected light from the end surface. The area sensor camera 23 images the regular reflection light from the end surface of the substrate W formed on the imaging surface of the image sensor, and outputs an image signal corresponding thereto.
The control unit 5 takes in image signals sequentially output from the area sensor camera 23 and creates one end face image data including the entire end face of the substrate W. The end face image data is subjected to image processing, and defects on the end face of the substrate W are extracted. The result of extracting the defect is displayed on the display unit 6.

While carrying out the end face inspection of the substrate W, the transfer arm 13 takes out the next substrate W from the cassette in the cassette loading / unloading section 11 and aligns it. The alignment is completed before the end face inspection of the previous substrate W is completed. Here, the alignment is such that when the substrate W is transferred from the transport arm 13 to the planar inspection unit 4 or the end surface inspection unit 3, the center position of the substrate W coincides with the reference positions of the inspection units 3 and 4, and the rotation direction is the same. It is set to be in a predetermined direction.
When the end face inspection of the previous substrate W is completed, the transfer arm 13 holds the inspected substrate W in one hand (not shown) and places the next substrate W held in the other hand on the rotary stage 22. . The end surface inspection unit 3 performs the end surface inspection on the next substrate W newly placed by substrate replacement in the same manner as described above. The transport arm 13 transports the previous substrate W for which the end surface inspection has been completed to the planar inspection unit 4, and transfers the substrate W to the holding member 31 that is waiting in advance at the delivery position of the planar inspection unit 4. The transfer arm 13 takes out the next substrate W from the cassette in the cassette loading / unloading unit 11 and performs alignment of the substrate W until the end surface inspection unit 3 finishes the inspection.

In the planar inspection unit 4, when the substrate W is set on the holding member 31, the support unit 34 sucks the peripheral edge of the back surface of the substrate W. At the same time, the line light source 36 is set to the light illumination angle θ0, and the line sensor camera 37 is set to the imaging angle θ0. The line light source 36 outputs linear illumination light and illuminates the surface of the substrate W at a light irradiation angle θ0. In this state, when the holding member 31 is moved in the forward direction in the X-axis direction at a constant speed, the linear illumination light from the line light source 36 scans the surface of the substrate W at a relatively constant speed.
The reflected light when the line-shaped illumination light is reflected by the surface of the substrate W is taken into the line sensor camera 37. An image signal is output according to the specularly reflected light imaged on the imaging surface of the image sensor of the line sensor camera 37.

  The control unit 5 receives the image signals sequentially output from the line sensor camera 37, and creates one image of the entire surface of the substrate W and bright field image data of regular reflection on the surface. Hereinafter, this image data is referred to as first surface image data. The control unit 5 performs image processing on the first surface image data and extracts defects on the surface of the substrate W. The result of the defect extraction is output to the display unit 6.

  Next, the control unit 5 moves the position of the line light source to a light illumination angle θ1 that is shifted from the light illumination angle θ0 by a predetermined angle. Accordingly, an image other than regular reflection, for example, an incident angle and a reflection angle are set to be different so as to be captured by a dark field, and a dark field image by scattered light can be captured. Instead of changing the light irradiation angle of the line light source, the imaging angle of the line sensor camera may be moved from θ0 to θ2.

  When the setting of the imaging conditions is completed, the holding member 31 is moved at a constant speed in the return path direction in the X-axis direction. The line-shaped illumination light from the line light source 36 scans the surface of the substrate W at a relatively constant speed, and the reflected light is taken into the line sensor camera 37. The control unit 5 receives image signals sequentially output from the line sensor camera 37 and creates one piece of image data for the entire surface of the substrate W. This image data is image data under imaging conditions other than regular reflection. Hereinafter, this image data is referred to as second surface image data. The control unit 5 performs image processing on the second surface image data and extracts defects on the surface of the substrate W. The result of the defect extraction is displayed on the display unit 6.

When performing the back surface inspection, the control unit 5 rotates the holding member 31 about the rotation shaft 32 by 180 ° to invert the substrate W. At this time, the holding member 31 is configured such that the heights of the front surface and the back surface of the substrate W do not change. When an error occurs in the height, the holding member 31 may be configured to be movable in the direction perpendicular to the substrate W, and the center of the substrate W may be aligned with the optical axis of the inspection optical system.
Since the back surface is disposed on the inspection optical system side, the line light source 36 and the line sensor camera 37 are moved to an angle θ0, respectively, so that a specular reflection image can be acquired. The holding member 31 is moved at a constant speed in the forward direction in the X-axis direction, and the line-shaped illumination light is scanned on the back surface of the substrate W relatively and is imaged by the line sensor camera 37. The control unit 5 receives the image signal output from the line sensor camera 37 and creates one piece of image data for the entire back surface of the substrate W in regular reflection. The first back surface image data created in this way is subjected to image processing, and defects on the back surface of the substrate W are extracted. The result of the defect extraction is displayed on the display unit 6.
Further, the installation angle of the line light source 36 or the line sensor camera 37 is changed so as to obtain, for example, a dark field image other than the regular reflection on the back surface of the substrate W. The holding member 31 is moved at a constant speed in the return direction in the X-axis direction, and is imaged by the line sensor camera 37 while being illuminated with line-shaped illumination light. The control unit 5 receives the image signal output from the line sensor camera 37 and creates one piece of image data for the entire back surface of the substrate W other than regular reflection. The second back surface image data created in this way is subjected to image processing, and defects on the back surface of the substrate W are extracted. The result of the defect extraction is displayed on the display unit 6.

  When the imaging of the back surface of the substrate W is completed, the holding member 31 rotates again by 180 ° around the rotation shaft 32. The surface of the substrate W is set on the inspection optical system side. The line light source 36 and the line sensor camera 37 are moved to an angle θ0 that is an initial position. When the holding member 31 is moved to the delivery position, the suction is released. The inspected substrate W is taken out by the hand having the transport arm 13 vacant, and the substrate W after the end surface inspection is newly transferred to the holding member 31. For the new substrate W, the front and back surfaces are inspected in the same manner as described above. The inspected substrate W is transported to the cassette loading / unloading section 11 by the transport arm 13 and returned to the cassette.

The control unit 5 uses the image data and the defect extraction result obtained in this way, and based on the design information of the three-dimensional shape of the substrate W, the front surface, the back surface, and the end surface (hereinafter referred to as the substrate surface layer) of the substrate W. The defect state can be overlaid and displayed on the display unit 6.
An example of the screen at this time is shown in FIG. The screen 40 includes defect information on the surface of the substrate W using the first surface image data and chip design information on the substrate surface (for example, die size, chip size, number of dies in shot, shot offset with respect to the wafer center). A three-dimensional image TD1 to be displayed is displayed. Further, the screen 40 is provided with a menu 41 for realizing a display content changing function, and can be operated by a general three-dimensional CAD (Computer Aided Design) device, for example, a movement or projection plane. Rotation at, 3D rotation at a specified point, magnification change, etc. are possible. FIG. 4 shows an image TD2 in which the magnification is reduced and the back surface is displayed by rotating it three-dimensionally around a designated point with respect to the display of FIG. This image is created using the first back surface image data.

Here, a method of creating and displaying an image of a three-dimensional shape of a substrate from the image data of the end face, the front surface image data, and the back surface image data for three-dimensional display will be described. As the front and back image data, first surface image data that is a bright field image captured by regular reflection and first back image data of the same bright field are used in combination. Further, the second surface image data in the dark field and the second back surface image data in the dark field imaged under the same conditions are used in combination.
Here, the schematic diagram which looked at the board | substrate W from the side surface direction is shown in FIG. The image of the end surface captured by the area sensor camera 23 is the end surface imaging range AR1 of FIG. Corresponding to this, an actually obtained image is shown in FIG. In the image 50, the notch position 51 is an alignment performed prior to the end face inspection, and the Y coordinate is known in advance. Further, since the notch portion is physically recessed, the captured image is dark like the portion 52, and the image is different from the normal end face. Furthermore, the distance per pixel in the Y direction and the amount of rotation of the rotary stage 22 when the notch position 51 is set to 0 ° can be calculated from the clock speed of the area sensor camera 23 and the rotational speed of the rotary stage 22.

Further, a rectangular area AR2 is set avoiding the notch position 51 in the image 50, and luminance information in the rectangular area AR2 is acquired for each pixel. At this time, if an average value of luminance information of a plurality of pixels having the same Y coordinate, that is, a row of pixels extending in the Z direction, is acquired, a profile of luminance information for one line in the Z direction is obtained. An example of this profile is shown in FIG. The maximum value M1 and the minimum value M2 of the brightness of the profile 53 are acquired, the intermediate value M3 is calculated, and the Z coordinates of the intersections m1 and m2 between the intermediate value M3 and the profile 53 are examined. Since these intersections m1 and m2 correspond to the boundary between the end faces of the substrate W, the end face width can be determined from the Z coordinates of the two intersections m1 and m2 in the Z direction. Further, the midpoint of the end face width is calculated, and this is set as the end face edge portion 54.
As a result, in the end face image 50 illustrated in FIG. 6, an arbitrary pixel has a coordinate in the Z direction with the end face edge 54 as the origin, and a circumferential direction in which the notch position 51 is 0 ° in the Y direction. It can be specified by angle. This makes it possible to map each pixel to a three-dimensional coordinate system.

Next, a captured image 61 of the surface is shown in FIG. The shift amount 62 between the image center IG1 and the substrate center WG1 becomes a substantially constant value within an error range of correction by the aligner 12 in the transport unit 2. Similarly, the displacement 63 in the rotational direction of the notch portion 52 is also substantially constant. The notch deviation amount 63 is a rotation angle when the notch position is directly below the captured image 61 and defined as 0 °.
As described above, an arbitrary position of the planar image coordinate system can be mapped to a three-dimensional coordinate system by the shift amount 62 between the image center IG1 and the substrate center WG1 and the shift amount 63 of the notch position. By performing the same process for the back image, it is possible to map a three-dimensional coordinate system.
Here, the information on the angle in the circumferential direction with reference to the notch position can be used in common for the front and back images and the end face image, so the defect positions detected on the end face, front and back surfaces respectively. Can be mapped to a three-dimensional coordinate system to form a three-dimensional image.

Further, the procedure for performing the three-dimensional display will be described with reference to the flowchart of FIG.
First, based on design information such as the diameter, thickness, and notch size of the substrate W, structure data of the three-dimensional substrate W is created with a drawing tool such as CAD (step S101). Actually, the structure of the substrate W is defined by the SEMI standard and the like, and there are only limited types such as 4 inches, 5 inches, 6 inches, 8 inches, and 12 inches. Therefore, it is preferable to create a three-dimensional shape of a substrate having each structure and size in advance.

Next, chip map data is created as an image for overlaying chip design information such as chip layout on the substrate image (step S102). The procedure for creating the chip map data will be described later. Further, the chip map data is read out by using a general-purpose CAD image pasting function and pasted on the surface of the structure data of the substrate created in step S101 (step S103). The pasting position can be automatically performed according to the value of the three-dimensional coordinate system.
Thereafter, defect data is read (step S104). In the defect detection process, generally, as shown in FIG. 10, the image is stored as a binarized image obtained by converting the presence / absence of a defect into two values such as black and white. The defect data image is also pasted by using the CAD image pasting function in the same manner as the chip map data pasting (step S105). In this way, three-dimensional display as shown in FIGS. 3 and 4 can be performed.

Here, for creating the chip map data in step S102, for example, a board design information input screen 71 as shown in FIG. 11 is used. This board design information input screen 71 is created by the processing of the control unit 5 and is displayed on the display unit 6 as input means for accepting an operation from the operation unit 7. As a chip size input field, a width input unit 72A and a height input unit 72B. The width corresponds to the horizontal direction when the substrate W is displayed as shown in FIG. The height corresponds to the vertical direction. The board design information input screen 71 further includes a die size width input section 73A and a height input section 73B, a scribe size width input section 74A and a height input section 74B, a shot layout input section 75A, 75B, matrix layout input sections 76A and 76B, matrix shift input sections 77A and 77B, edge cut amount input section 78, die total number display section 79, and inspection area storage button 80 are arranged. These pieces of information are input by the operator based on the design information. The die is a minimum unit to be cut out as a product in the substrate W and has a repeated pattern. The chip size is the size of the chip contained in the die. The reason why the chip size and the die size are set separately is that the most important inspection area is generally a chip. The chip size and the die size satisfy the following relationship.
Chip size width (height) ≤ die size width (height)
The width and height here correspond to the width and height in FIG.

The scribe size is the size of the scribe included in the die. Scribe is a cutting area when a chip is taken out in a later process. In general, in the substrate inspection, the scribe region is often excluded from the inspection target or only a fatal result such as a large defect is detected. Therefore, in this defect inspection apparatus 1, the scribe line can be displayed by setting the slive size. The relationship between the scribe size and the die size satisfies the following.
Scribe size width (height) + chip size width (height) ≤ die size width (height)
Using this relationship, when the notch of the substrate W is turned downward, the lower left of the die is defined as the origin, and the position of the scribe size is defined as the lower left of the chip.

  Further, the shot layout corresponds to the number of dies in the shot. The matrix layout corresponds to the number of shots in the substrate W. The matrix shift is a shift amount when the center of the shot layout is shifted from the center of the substrate.

  By performing such a definition, as in the design information display screen 81 shown in FIG. 12, the image 82 corresponding to the substrate W has a chip area 83, a scribe area 84, an extra area 85, and an edge cut area 86. Are divided into four areas and displayed. The edge cut region 86 corresponds to an annular portion where the resist is removed at the peripheral edge of the substrate W. The extra region 85 corresponds to the remaining portion on the substrate W excluding the chip region 83, the scribe region 84, and the edge cut region 86. The regions 83 to 86 are displayed with different colors, brightness, patterns, and the like so that they can be visually distinguished. Here, although the chip area 83 is displayed as an inspection area, the chip area 83 can be excluded from the inspection object by clicking with an input means such as a mouse (not shown). The display of the excluded chip area is changed so that it can be distinguished from other inspection objects, for example, as a chip area 87 indicated by a broken line. When the excluded chip area 87 is clicked again, the display is switched and re-registered as the chip area 83 of the inspection area.

  The number of registered chips in the inspection area, that is, the total number of dies including the scribe portion is displayed on the display unit 79 of the board design information input screen 71 in FIG. Finally, when the operator presses the inspection area storage button 80, the chip map image on the design information display screen 81 is stored as chip map data.

  The operator confirms the defect using the three-dimensional image displayed as shown in FIGS. Defects on the front surface, back surface, and end surface are confirmed while appropriately rotating or enlarging or reducing the three-dimensional image. Although it is difficult to confirm with a regular reflection image, there are defects that can be easily confirmed with an image other than regular reflection, so the images are switched as necessary. For example, the first surface image data and the second surface image data are obtained by observing the same substrate W, and the angle in the circumferential direction can be matched with the notch position as a reference, so only the display is switched. Thus, the defect can be easily confirmed.

In this embodiment, since end face inspection, front surface inspection, and back surface inspection are performed and a three-dimensional image is created by associating each image, defect information can be collectively displayed as an image. Since the three-dimensional image can be enlarged / reduced or rotated, the defect can be confirmed promptly and surely. In addition to the correction of the misalignment by the aligner 12, correction using the images obtained by the end face inspection and the front and back surface inspections is performed, so that the misalignment of each image can be prevented and a three-dimensional image can be created with high accuracy.
Furthermore, since the end face inspection and the front surface and back surface inspection can be performed simultaneously, the inspection time can be shortened.

The order of inspecting the substrate W is not limited to the order of the end face, the front surface, and the back surface, and any order may be used. In any case, it is desirable to perform an end surface inspection during the surface or back surface inspection and to perform alignment of the next substrate W during the end surface inspection.
The three-dimensional image shown in FIG. 3 displays an image of defect information as shown in FIG. 10 and an image of chip design information (that is, chip map data) overlaid on the structure data of the substrate W. Only the defect information image may be displayed in the W structure data. Further, the defect coordinates on the front surface and the back surface are converted into positions on the substrate W based on the correction information of the aligner 12 of the transport unit 2. The coordinate conversion may be performed by performing pattern matching using an image of a typical feature region. In this case, coordinates can be converted with higher accuracy.

The configuration and the inspection procedure related to the defect inspection operation on the front surface and the back surface of the transport unit 2, the plane inspection unit 4, the control unit 5, the display unit 6, and the operation unit 7 use the configuration and the inspection procedure described in WO03 / 027652. May be.
The information displayed on the display unit 6 may display not only the defect information on the end face, the front surface, and the back surface, but also an inspection image taken for inspection. In this case, for example, the defect information is displayed in red, the luminance information of the inspection image is displayed in gray scale, and whether or not the defect information is displayed can be switched. This makes it possible to easily confirm the situation of the part where the defect has occurred.

Further, display of chip design information of the substrate W, for example, die size, chip size, number of dies in a shot, shot offset with respect to the center of the substrate, etc., may be displayed overlaid on the back image as shown in FIG. . It becomes possible to confirm whether the defect position on the back surface has an adverse effect on the surface chip as well as the chip position where the defect has occurred on the front surface.
Further, by storing these pieces of information in a data format compatible with a three-dimensional CAD such as the XVL format, the inspection result of the defect inspection apparatus 1 can be easily confirmed by another system. .

(Second Embodiment)
As shown in FIG. 13, in the defect inspection apparatus 101 in this embodiment, the transport unit 2 is not provided with an aligner. The base 21 that indicates the rotary stage 22 of the end surface inspection unit 103 is provided with a moving mechanism (not shown) so that the rotary stage 22 can be moved in two axes in the XY directions orthogonal to the rotary axis of the rotary stage 22. Yes. In addition to the area sensor camera 23 that is the first imaging device, the end surface inspection unit 103 is provided with the line sensor camera 104 and the illumination device 105 that are the second imaging device that images the peripheral portion of the substrate W. ing.

  The inspection sequence of the substrate W is the same as that in the first embodiment, and the substrate W is first transported to the end surface inspection unit 3. However, since there is no aligner in the transport unit 2, the substrate W is transported in a state where alignment is not performed. When inspecting the end face of the substrate W, the rotary stage 22 is rotated to acquire an image of the end face, and an image is taken by the line sensor camera 104 while illuminating the peripheral portion of the substrate W with the illumination device 105. From the image obtained by the line sensor camera 104, information on the notch position and information on the eccentricity of the substrate W are obtained, and the alignment of the substrate W is performed based on these information. When the end face inspection and the alignment are completed, the surface inspection and the back surface inspection are performed by the plane inspection unit 4. Since the inspection result in the planar inspection unit 4 is the result after alignment, it can be handled in the same manner as in the first embodiment. The image in the end face inspection unit 103 is converted based on the correction value when aligned.

  In this embodiment, the alignment in the conveyance unit 2 can be omitted by using the end surface inspection unit 103 having an alignment function. The apparatus configuration can be simplified and the inspection time can be shortened because the alignment and the end face inspection are performed simultaneously. The inspection order may be the first as long as the end face inspection is performed, and the order of the front and back surfaces can be changed.

  Note that the optical system for front surface inspection or back surface inspection may be arranged at a position close to the substrate surface as shown in FIG. The line light source 36 includes a light source 36A and an optical element 36B that is close to the substrate surface. The line sensor camera 37 includes an optical element 37A and an imaging element 37B that are close to the substrate surface. In this case, if a large foreign matter is placed on the substrate W, it interferes with the optical system, so the foreign matter detection sensor 110 is used. The foreign object detection sensor 110 includes a light source 110A that emits a light beam parallel to the substrate surface and a light receiving element 110B with the substrate W interposed therebetween. If foreign matter is placed on the substrate surface, the light flux of the foreign matter detection sensor 110 is blocked, so that the foreign matter can be detected. When foreign matter is detected, the inspection is not performed. A foreign object detection sensor may be further provided along the opposite surface of the substrate W. When the foreign object detection sensor on the opposite side detects a foreign object, the surface is not inspected.

(Third embodiment)
In this embodiment, in the defect inspection apparatuses 1 and 101 shown in FIG. 1 or FIG. 13, an inspection result recording unit such as a hard disk drive for storing the inspection result, and the inspection result stored in the inspection result recording unit include arbitrary information. And a test result analyzing means for displaying a graph using as a key.

  The inspection result analyzing means can search the recorded data and display the data in a graph, and is realized by, for example, a CPU (Central Processing Unit) or a main memory. An example of a screen of a test analysis operation GUI (Graphic User Interface) provided by the test result analysis means is shown in FIG. This analysis screen 121 has an input unit 122 for selecting the type that has been inspected so far, and an input unit 123 for process selection, and data is input to at least one of the input units 122 and 123. . In FIG. 15, only the product type is set, but only the process or both the product type and the process may be specified. In addition, a period input unit 124 is provided so that a period to be analyzed can be designated. The input unit 125 to be analyzed is for inputting the type of process to be analyzed in the production line. It is not always necessary to input data to the input unit 125. However, when data is input to the input unit 125, the corresponding item selection unit 126 is set. When executing the analysis, the analysis start button 127 is pressed.

  FIG. 16 shows an example of a graph displayed during analysis. Each line L1-L5 of this graph has shown the defect occurrence rate in each of an end surface, a surface, and a back surface according to inspection conditions. The condition in which the defect occurrence rate is always small is a condition that makes it difficult to detect a defect, and thus is inappropriate as an inspection condition. For this reason, a defect occurrence rate for a predetermined time can be designated as the effective condition determination threshold E1. The inspection conditions corresponding to the lines L4 and L5 are regarded as inappropriate because there are few portions exceeding the effective condition determination threshold E1. Therefore, the inspection condition is removed from the inspection condition by using an inspection condition deletion unit (not shown). Thus, in the subsequent inspection of the product type “MEMORY-A”, the inspection is performed under the inspection conditions corresponding to the lines L1 to L3. These inspection conditions can be registered in a recipe, and at the time of inspection, inspection of each inspection condition is performed according to the recipe, and a three-dimensional image is created.

  Similarly, a case where an ID of a coater developer (hereinafter referred to as C / D-ID), which is one of manufacturing apparatuses, is selected as an analysis target will be described. The item selection unit 126 selects information specifying C / D such as “C / D # 1” as one of the C / D-IDs. As a result, a graph of the defect occurrence rate as shown in FIG. 17 is obtained. This graph shows the defect occurrence rate for the substrate of the product type “MEMORY-A” created by C / D # 1, for each inspection condition. From the graph, it can be seen that the defect can be detected efficiently if the inspection is performed only with the inspection condition of the line L6 and the inspection condition of the line L7. In this embodiment, only the C / D-ID has been described. However, there is a possibility that the frequency of defect detection changes such as lot ID, slot number, wafer ID, manufacturing apparatus other than C / D, and recipe ID. If it is a key with. These keys are preferably recorded in advance in the inspection result recording means. Further, a key may be added as necessary.

The present invention can be widely applied without being limited to the above-described embodiments.
For example, the plane inspection unit 4 may acquire an image of only one of the front surface and the back surface of the substrate W. In this case, a three-dimensional image is formed using the acquired planar image and end face image.
The first imaging device is not limited to the area sensor camera 23, and a plurality of imaging devices may be arranged according to the shape of the end face, or one imaging device may be arranged to be movable. Similarly, the second imaging device is not limited to the line sensor camera 104.
After the inspection by the plane inspection unit 4, the end surface inspection unit 3 may perform the inspection. In the second embodiment, the front and back surfaces are inspected without alignment, but a three-dimensional image can be created by correcting the image later.

It is a figure which shows schematic structure of the defect inspection apparatus which concerns on embodiment of this invention. It is a top view which shows the state which made the holding member of the plane inspection part hold | maintain a board | substrate. It is the figure which displayed the surface of a board | substrate and the test result of a defect three-dimensionally. It is the figure which displayed the back surface of the board | substrate three-dimensionally. It is a figure which illustrates typically the range of the image acquired in an end surface inspection part. It is a figure which shows an example of the image acquired in an end surface test | inspection part. It is a figure which shows the profile of the brightness | luminance used when specifying an end surface edge part. It is a figure which shows an example of the image acquired by a surface test | inspection image, and demonstrates the process converted into a three-dimensional coordinate. It is a flowchart which produces a three-dimensional image. It is a figure which shows the binarized image which shows a defect. It is a figure which shows an example of a board | substrate design information screen. It is the figure which displayed design information in two dimensions. It is a figure which shows schematic structure of the defect inspection apparatus aligned by an end surface inspection part. It is a figure which shows the example which has arrange | positioned the optical system close to the board | substrate. It is a figure which shows an example of an analysis screen. It is the figure which made the graph so that the defect occurrence rate according to inspection conditions could be compared. It is the figure which graphed the defect occurrence rate about the specific analysis object so that comparison according to inspection conditions was possible.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1,101 Defect inspection apparatus 3 End surface inspection part 4 Plane inspection part 5 Control part 6 Display part 22 Rotation stage 23 Area sensor camera (1st imaging device)
104 Line sensor camera (second imaging device)
W substrate

Claims (5)

An end surface inspection unit that acquires a two-dimensional image of the entire end surface of the substrate;
A plane inspection unit for acquiring a two-dimensional image of at least one of the entire front surface and the entire back surface of the substrate;
A control unit that performs defect inspection on each of at least one of the front image and the back image acquired by the planar inspection unit and the end surface image acquired by the end surface inspection unit, and creates a three-dimensional image of the substrate including the inspection result When,
A display unit for displaying the three-dimensional image,
The three-dimensional image is formed by pasting at least one of the front surface image and the back surface image and the end surface image in association with the surface of a three-dimensional shape based on the entire structure data of the substrate. A defect inspection apparatus characterized by being a three-dimensional image.   The end surface inspection unit includes a rotary stage on which a substrate is placed, a moving mechanism that moves the rotary stage in a direction orthogonal to a rotation axis, a first imaging device that acquires an end surface image, and the rotary stage. And a second imaging device that detects a displacement amount of the substrate with respect to the rotary stage when rotating, and the control unit drives the moving mechanism based on information from the second imaging device to provide a substrate. The defect inspection apparatus according to claim 1, wherein the positional deviation is corrected.   3. The defect inspection apparatus according to claim 2, wherein after the inspection by the end surface inspection unit is completed, the control unit causes the substrate to be conveyed so as to perform the inspection by the planar inspection unit.   The defect inspection apparatus according to claim 1, wherein the three-dimensional image is rotatable on the display unit.   The defect inspection apparatus according to claim 1, wherein the three-dimensional image can be enlarged or reduced on the display unit.

Images