JP5196723B2 - Defect correction apparatus and defect correction method - Google Patents

Description

  The present invention relates to a defect correction apparatus and a defect correction method suitable for automatically correcting a defect in a device pattern or a wiring pattern formed on a substrate such as a TFT (Thin Film Transistor) substrate.

  Conventionally, a TFT substrate used for a flat panel display such as a liquid crystal display or an organic EL (Electro-Luminescence) is manufactured by forming a fine device pattern or wiring pattern on a glass substrate. When such a device pattern or wiring pattern is formed, a defect may occur in the pattern being formed, for example, by performing a process process with dust or water droplets adhering to the glass substrate. A TFT substrate in which such a defect has occurred becomes a defective device and decreases the yield. Therefore, in order to stabilize the yield of the production line at a high level, pattern repair is performed using a defect correction device (also referred to as a “repair device”) and then the flow is passed to the next process.

  In order to correct the pattern defect, first, it is necessary to check whether or not there is a defect on the substrate. If there is a defect, it is necessary to specify the position of the defect. Therefore, a defect inspection is performed for this purpose. There are mainly two types of defect inspection methods: electrical defect inspection and optical defect inspection.

  Electrical defect inspection is a method of inspecting short circuit and open circuit as a pattern by detecting a continuity and capacity abnormality by flowing a minute current through device pattern and wiring pattern and comparing with the characteristic that should be as an electric circuit. It is.

  This electric defect inspection has an advantage that a defect causing an abnormal operation can be surely found, but it is not good at specifying the defect position accurately. With regard to the defect of the device pattern, it is possible to know at which pixel the defect has occurred, but it is not possible to know where it is in the pixel. In addition, since the signal lines and gate lines in the pixels that are generated in the liquid crystal display are protruded, a pixel electrode (transparent conductive film) region called a window portion is chipped, so that the amount of light for that pixel is insufficient. This type of defect cannot be found by electrical defect inspection alone.

  On the other hand, the optical defect inspection is a method of inspecting a pattern defect by capturing a device pattern or a wiring pattern with an apparatus such as an optical microscope and comparing it with an image that should be originally.

  The optical defect inspection can find a pattern defect regardless of electrical characteristics, and can accurately detect the position of the defect portion. Furthermore, it is possible to specify the type of defect by extracting a feature quantity such as the shape of the defect site and comparing it with information stored in a database in advance.

  However, in order to detect the detailed position and shape of the defective portion on the substrate in the optical defect inspection, it is necessary to increase the imaging magnification. When the imaging magnification increases, the number of times of imaging must be increased in the case of an area camera, and the imaging time in the case of a line camera must be increased, so that the inspection time required for inspecting the entire surface of the substrate becomes longer. In an actual manufacturing process, since inspection time is limited, optical defect inspection is normally performed with a resolution corresponding to 2 to 5 μm of one pixel of a captured image. Even if a defect can be detected, it is difficult to detect a detailed shape.

  Based on the defect information obtained by such electrical defect inspection and optical defect inspection, defect correction is performed using a defect correction device. Recently, various automation attempts have been made to improve productivity. ing. For example, by attaching a digitizer to the defect correction device, after the operator has taught the fine correction position about the detected defect position, the defect correction device can automatically perform the defect correction work. ing.

  Further, as a method that does not involve an operator at all, an optical defect inspection is performed on a defect found based on an electrical defect inspection, whereby a defect position is identified and automatically corrected ( For example, see Patent Document 1.)

In addition, a method has been proposed in which an image pickup device is attached to the defect correction device and data for operating the defect correction device is created from the image of the imaged defect (see, for example, Patent Documents 2, 3, and 4).
JP-A-10-177844 JP-A-7-86722 JP 2000-208902 A JP 2004-297452 A

  However, the conventional defect correction methods described in Patent Documents 1 to 4 have a problem that they can deal only with defects that are relatively easy to repair.

  First, the one described in Patent Document 1 only corrects a defect detected by an electrical defect inspection, and cannot deal with a simple pattern protrusion defect without a short circuit.

In addition, those described in Patent Documents 2 and 4 are corrected if the difference image between the defect image (inspected image) and the reference pattern is simply corrected as a defect range unless the type of the defect is grasped. There is a problem that it is easy to fail.
In other words, if the information displayed in the difference image is a foreign object, or is a pattern protruding, or if it is a protruding defect, the composition and film thickness of the protruding pattern must be known. Setting is difficult. For example, in the method of repeatedly correcting a plurality of times until it can be corrected as described in Patent Document 2, it takes too much time to complete the correction, or the first time is excessively shaved and damages the lower layer. May end up.

  Further, Patent Document 3 discloses a method for searching for a defect to be corrected while identifying a defect type to be corrected in the defect inspection apparatus and comparing it with a defect that matches with an image detection mechanism included in the defect correction apparatus. Although this method has been proposed, since it is necessary to image the defect location with a sufficient imaging magnification in the defect inspection device in advance and the defect correction device must search for the defect to be corrected, etc., from the inspection It may take a long time to correct.

  In addition, since the repeatability of a general XY stage built in a defect inspection apparatus is about several μm due to the presence of backlash, the size is about several μm when defect correction is performed based on a captured defect image. The defect cannot be corrected with high accuracy, the size and angle of the optical slit is fixed, or it is not compatible with automation. There was a problem that there were many defects.

  The present invention has been made in view of such a point, and an object thereof is to perform defect correction with high accuracy even for a defect having a high degree of difficulty that requires a complicated correction operation.

In order to solve the above problems, a defect correction apparatus of the present invention is a defect correction apparatus that corrects a defect on a substrate on which a plurality of wiring patterns are formed, and a defect extraction unit that extracts a defect portion on the substrate; , a position information extraction unit that extracts position information in the substrate of the defective portion extracted by the defect extracting unit, to obtain the process information at that time, was extracted with the process information and the positional information extraction section the based on the position information of the defect site, the design information of the substrate, and the film information acquisition portion for obtaining a film information indicating the state of the peripheral or the lower the defect site at that time, was extracted by the defect extracting unit wherein A feature extraction unit that quantifies the feature quantity of the defect site; a defect classification unit that classifies a defect type based on the feature information digitized by the feature extraction unit; the position information of the defect site; the feature information; Serial defect type and the correcting method generating unit for generating a defect correction method of the defective portion on the basis of the film information, defect correction by operating the defect correction mechanism based on the defect correction method that is generated by the modified method of generating unit And a control unit for performing the above.

According to the above configuration, the defect correction method is performed by acquiring and analyzing the image of the defective part based on the defect position information of the substrate, and further using the film information indicating the periphery of the defective part or the state of the ground at that time. Therefore, for example, effective correction can be automatically performed with high accuracy even for a minute defect of several μm or less or a defect having a complicated shape.
In addition, since the information on the underlying film of the defect is grasped and an appropriate laser process is selected, there is little possibility of damaging the substrate or leaving behind the defect.

  Further, according to one aspect of the defect correction apparatus of the present invention, the defect correction mechanism includes a laser transmitter, an opening area through which laser light passes and a slit that changes an opening angle in a plane orthogonal to the optical axis, and the substrate. And a stage that moves in a plane orthogonal to the optical axis, and an optical axis moving unit that moves the optical axis minutely, while holding the position of the stage during laser light irradiation, the optical axis moving unit It is characterized by adjusting the laser beam irradiation position.

  According to the above configuration, since the laser movement position is adjusted by finely moving the laser optical axis, the laser movement position can be moved regardless of the positioning accuracy of the XY stage.

  ADVANTAGE OF THE INVENTION According to this invention, it can correct with high precision also to the defect with high difficulty which requires a complicated correction operation.

  Hereinafter, an embodiment of the present invention will be described with reference to FIGS.

FIG. 1 is a diagram showing a configuration example according to an embodiment of a defect correction apparatus according to the present invention.
In the present embodiment, a defect correction apparatus for correcting a pattern defect on a TFT substrate such as a liquid crystal display will be described as an example.

  Although the defect correction apparatus 1 according to the present embodiment is a so-called laser repair apparatus, it is also possible to adopt a configuration capable of performing a wiring correction technique such as a laser CVD (Chemical Vapor Deposition) method. It is not limited.

  As a result of the defect inspection performed by the defect inspection apparatus 100 in advance, the defect correction apparatus 1 uses the position information of the defect of the TFT substrate 6 as an overall control unit (hereinafter referred to as “control unit”) in the defect correction apparatus 1. ) Receive it at 2. The accuracy of the position information may be such accuracy that the defect portion is surely within the imaging range by the imaging device 17 built in the defect correction device 1. The control unit 2 is a computer (arithmetic processing unit) such as a personal computer, and constitutes a correction method generation unit together with the image analysis and correction method generation unit 3.

  First, the control unit 2 sends a command to the stage control unit 7 to move the XY stage 5 on which the TFT substrate 6 is mounted so that the defect location detected by the defect inspection apparatus 1 is directly below the objective lens 8. Next, the focus stage 10 is moved to adjust the distance between the objective lens 8 and the TFT substrate 6 so that the imaging device 17 can capture a focused image of the light transmitted through the optical lens 14g. Here, an image with appropriate brightness is obtained by epi-illumination by the half mirrors 15a and 15b, the optical lens 14a, and the lamp 9.

FIG. 3 is an example of a defect image including a protruding defect obtained in this manner.
In the present embodiment, the pixel pattern 110 is a schematic configuration of one pixel pattern of a glass substrate from which, for example, a color filter of a liquid crystal display in an active matrix system is removed, and is largely divided into a signal line 111, a scanning line 112, a storage capacitor element. (Capacitor) 113 and pixel electrode 114. In this example, a protruding defect 115 occurs in the signal line 111. The captured defect image is temporarily stored in the defect image memory 18 of the image analysis and correction method generation unit 3.

  Next, the control unit 2 sends a command to the stage control unit 7 to move the XY stage 5, and the position moved from the defective part to the position where several repeated patterns, that is, exactly the same pixel pattern, is directly below the objective lens 108. In this manner, a reference image having no defect is captured and stored in the reference image memory 19 of the image analysis and defect generation method generation unit 3. An example of the reference image 120 obtained in this way is shown in FIG.

  In the defect extraction unit 20, the defect image stored in the defect image memory 18 and the reference image stored in the reference image memory 19 are aligned, and a difference image is generated to generate a difference image. Extract. FIG. 5 shows the defect 115 extracted from the difference image between the defect image 110 shown in FIG. 3 and the reference image 120 shown in FIG. The defect extraction unit 20 outputs the extracted defect site information to the detailed position information extraction unit 21 and the feature extraction unit 22.

  The detailed position information extraction unit 21 calculates the exact position of the extracted defect on the TFT substrate 6 from the current position of the XY stage 5 and the defect image, and the information is obtained from the film information code addition unit 24 and the correction method generation unit 27. Send to.

  The film information code adding unit 24 extracts pattern design information of the position where the defect exists from the design information database based on the accurate position of the defect from the detailed position information extracting unit 21 and the process information at that time, and the film information code The assigning unit 24 outputs the film information code.

A substrate such as a TFT substrate is generally configured by laminating a plurality of circuit layers. Therefore, the state of the substrate differs for each film formation process or pattern formation process at that time. For example, even at the same position in the same wiring pattern, the type and presence / absence of the wiring pattern and the members in contact with the periphery differ depending on the process at that time. For this reason, the thickness of the defect portion, the strength against heat, and the like are different, and it is necessary to determine whether laser irradiation is possible, the laser irradiation position, the laser output, etc. in consideration of these conditions. In order to carry out defect correction with high accuracy, the accurate position of the defect and the process information at that time are acquired, and the substrate state, that is, the film information (design), such as the state of the periphery of the defect at that time or the state of the base of the defect. Information or circuit information) and need to be reflected in the defect correction method.
The process information may be obtained from a master computer (not shown) or may be obtained directly from the substrate manufacturing apparatus. Examples of the process information include information such as the current layer number or lot number.

  Also, the feature extraction unit 22 digitizes the extracted feature information such as the color, size, contrast, and shape of the defect and outputs it to the comparison / classification code providing unit 26.

  The comparison / classification code assigning unit 26 compares the digitized feature information of the defect with the feature information data registered in the classification database 25, and assigns a matching feature classification code. The classification code is associated with the defect type, and the type of the defect can be determined by examining the classification code.

  In the present embodiment, the defect type is classified by the classification database 25 and the comparison / classification code assigning unit 26, but the classification method is various known techniques, for example, Japanese Patent Laid-Open Nos. 5-108800 and 2003-233807. Alternatively, a classification method described in JP 2001-168160 A can be applied.

The correction method generation unit 27 defines the details of how to operate the correction mechanism unit from these detailed position information, feature information, design information (film information code), and defect type information (classification code). Generate method.
Specifically, for example, in the case of the defect 115 shown in FIG. 3, since it can be determined that the signal line pattern existing in the window portion is protruding, the laser irradiation output suitable for removing the composition of the signal line 111 and The irradiation time and the laser irradiation size, angle, position, etc. on the TFT substrate 6 are calculated, and their information is output to the control unit 2.

  The control unit 2 sends a command to the correction mechanism control unit 16 according to the defect correction method generated by the correction method generation unit 27, operates each unit in the correction mechanism unit 4, and automatically corrects the defect.

  The correction mechanism unit 4 can change the irradiation size and angle by passing the variable slit 12 after correcting the laser beam irradiated from the laser light source 13 with the optical lenses 14b and 14c.

  The variable slit 12 is, for example, a so-called XY-θ slit as shown in FIG. 6 and is a slit that can change the opening length of the rectangle in the X and Y directions and the rotation angle θ. It can be driven by a drive signal.

  The laser beam shaped by the variable slit 12 is irradiated to the TFT substrate 6 through the objective lens 8 after passing through the galvanometer mirrors 11a and 11b, the various optical lenses 14d, 14d and 14f, and the half mirror 15a.

  The galvanometer mirrors 11a and 11b are two-dimensionally variable angle mirrors, and are driven by the correction mechanism control unit 16 to move the optical axis of the laser beam within the field of view of the objective lens 8 without moving the XY stage 5. That is, the irradiation position can be moved.

  In the defect correction apparatus 1 including the variable slit 12 and the galvanometer mirrors 11a and 11b, since the laser beam can be irradiated with sufficient positional accuracy with respect to the defect, the pattern defect can be corrected with higher accuracy than before. Become.

  Since the defect inspection apparatus 1 can use an optical inspection machine as a method for searching for a defect, it is possible to correct a pattern defect whose conduction state is normal.

  In addition, since the defect position information, feature information, and type information are extracted using the highly accurate image of the defect site included in the defect correction apparatus 1 and the correction method is comprehensively determined from these information, It becomes possible to correct defects with high accuracy, and there is no need to repeat corrections.

  Further, the defect inspection apparatus 100 only needs to be able to detect at least the position and does not require highly accurate image acquisition of the defective part. Therefore, the inspection with a resolution corresponding to 2 to 5 μm of normal one pixel is sufficient, and a long inspection time is required. do not need.

  Next, the defect correction process by the defect correction apparatus 1 will be described with reference to the flowchart of FIG.

  First, the control unit 2 finely adjusts the XY stage 5 and the focus stage 10 to capture an image to be inspected (defect image) (step S1) and store it in the defect image memory 18. Next, the image to be inspected (defect image) stored in the defect image memory 18 and the reference image registered in the reference image memory 19 are compared by the defect extraction unit 20, and a defect is extracted from the difference (step S2). ).

  Subsequently, the detailed position information extraction unit 21 acquires detailed position information on the TFT substrate 6 of the defective part (step S3). This detailed position information is sent to the film information code applying unit 24, and the film information code providing unit 24 obtains design information of the TFT substrate 6 at the defect site based on the detailed position information of the defect and the process information at that time. The membrane information code shown is acquired (step S4). This design information (membrane information code) is sent to the correction method generation unit 27.

  On the other hand, the feature extraction unit 22 extracts information obtained from the defect site image extracted by the defect extraction unit 20 as a feature amount of the defect site and digitizes the information (step S5). Then, the defect type of the defect is classified based on the feature information digitized by the comparison / classification code assigning unit 26, and the classification code is output (step S6). This classification code is sent to the correction method generation unit 27.

  The correction method generation unit 27 receives detailed position information, feature information, design information (film information code), and defect type information (classification code) of the defect, and an appropriate defect correction method for the defect site in consideration of each condition. Is generated and determined (step S7). The generated defect correction method is sent to the control unit 2. Then, based on the generated defect correction method, the control unit 2 controls the correction mechanism control unit 16 to drive the correction mechanism unit 4 to perform optimal defect correction (step S8). If there is another defect, the series of processes consisting of steps S1 to S8 is repeated.

Here, removal of pattern defects by a laser will be described.
A defect 131 in FIG. 7A is an example of a defect extracted by the defect extraction unit 20. When this defect 131 is corrected using a conventional defect correction apparatus, as shown in FIG. 7B, the opening area 132 of the variable slit 12, that is, the irradiation region so as to include all the defects to be removed. 132 was adjusted and the laser was irradiated. However, in this irradiation method, a laser is irradiated on many parts other than the defect 131. Irradiation of the laser beam to a place other than the defect 131 may damage the underlying glass substrate.

  Therefore, if the two irradiation regions 133 and 134 are irradiated with laser in two steps as shown in FIG. However, since the conventional defect correction apparatus uses the XY stage to move the laser irradiation position, it has only a repeated position accuracy of about several μm. For this reason, even if a defect correction as shown in FIG. 7C is attempted, an error is likely to occur in the irradiation position, such as the irradiation areas 135 and 136 being separated as shown in FIG. 7D, and automation is difficult.

  In the defect correction apparatus 1 according to the present embodiment, the accuracy of the laser irradiation position can be kept within 1 μm by moving the optical axis of the laser beam using the galvanometer mirrors 11a and 11b. it can.

  FIG. 8 shows an example in which FIG. 7C is further developed. It is considered that a mesh is applied to the laser irradiation region set so as to include the entire defect with respect to the defect 141, and the laser is irradiated to several tens of the divided areas to remove the defect. In this case, since the laser is not irradiated except for the defect 141, the defect can be corrected neatly. However, since the laser diameter is small, it takes time to complete defect correction (longer tact time). Further, since the irradiation area per one time is smaller than the laser irradiation areas shown in FIGS. 7B to 7C, the movement accuracy of the laser irradiation position is not so required.

  On the other hand, as shown in FIGS. 7B to 7C, when the laser irradiation region is relatively large, the movement accuracy of the laser irradiation position is required. However, according to the configuration of the present embodiment, the laser irradiation region can be moved with high accuracy. A large area can be irradiated with a laser at one time, and the time required for defect correction can be shortened (shortening of the tact time).

  In addition to the galvanometer mirror, a mechanism for moving the laser beam includes a method of moving the laser beam by attaching a piezo stage to some lenses included in the optical system.

  In this embodiment, a captured image in the vicinity of the defective portion is used to obtain a reference image. However, an ideal image that should originally be a defective portion synthesized from pattern design information can be used as a reference image. is there

  As described above, the embodiment of the defect correcting apparatus for executing the defect correcting method according to the present invention has been described based on the defect position information of the TFT substrate by the defect inspection apparatus, as described in the case of correcting the defect of the TFT substrate or the like. Since the image acquisition and analysis of the part is performed, and the defect correction method is determined using the film information of the defect position, effective correction is possible even for micro defects of several μm or less and defects with complex shapes. It can be done automatically with high accuracy.

  In addition, since the underlying film information is grasped and an appropriate laser process is selected, there is little possibility of damaging the substrate and leaving defects.

  In addition, it has a laser beam scanning mechanism such as a galvanometer mirror that moves the optical axis by moving the optical axis, so that the laser irradiation area and angle can be selected appropriately, and the laser is maintained while maintaining the position of the XY stage. Since the beam can be moved, regardless of the positioning accuracy of the XY stage, it is possible to automatically perform pattern protrusion by laser irradiation, removal of foreign matters, or wiring of a disconnected portion by laser CVD or the like with high accuracy and high accuracy.

  Even if a more complex defect correction method is determined by detailed analysis of the defect and substrate status, it will be meaningless unless a defect correction mechanism capable of accurately executing the complicated defect correction method is provided. . The defect correction mechanism of the defect correction apparatus according to the present invention can follow the request for such complicated defect correction processing, and can realize highly accurate defect correction.

  In addition, the numerical conditions such as the materials used, the amount thereof, the processing time, and the dimensions mentioned in the above description are only suitable examples, and the dimensional shapes and arrangement relationships in the drawings used for the description are also schematic. That is, the present invention is not limited to this embodiment.

  In addition, when using a laser CDV apparatus instead of the laser repair apparatus of the above-described embodiment, the XY slit opening area and opening angle optimum for pattern redrawing, optimum laser output, optimum CVD source gas state, The defect correction method is determined by obtaining the optimal irradiation position, the optimal number of irradiations, and the like.

  Further, in the above-described embodiment, the case of correcting the defect of the design pattern formed on the glass substrate of the flat panel display has been described. However, the correction target is not limited to this example, for example, a semiconductor wafer, a photo The present invention can be applied to a mask, a magnetic disk or the like in which a predetermined pattern is formed on a correction target substrate.

It is a block diagram of the defect correction apparatus which concerns on one embodiment of this invention. It is a flowchart of the defect correction process which concerns on one embodiment of this invention. It is a figure which shows the example of a protrusion defect. It is a figure which shows the example of a reference image. It is a figure which shows the example of the extracted defect. It is a figure which shows XY-theta slit. It is a figure which shows the example of removal of a pattern defect. It is a figure which shows the example of removal of a pattern defect.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Defect correction apparatus, 2 ... Overall control part (control part), 3 ... Image analysis and correction method production | generation part, 4 ... Correction mechanism part, 5 ... XY stage, 6 ... TFT substrate (inspection object), 7 ... Stage Control unit, 11a, 11b ... Galvano mirror (optical axis moving unit), 12 ... Variable slit, 13 ... Laser light source, 14a-14g ... Optical lens, 16 ... Correction mechanism control unit, 17 ... Imaging device (imaging unit), 18 Defect image memory (inspected image storage unit) 19 Reference image memory (reference image storage unit) 20 Defect extraction unit 21 Detailed position information extraction unit 22 Feature extraction unit 23 Design information database ( Substrate design information acquisition unit), 24 ... Film information code provision unit (substrate design information acquisition unit), 25 ... Classification database (defect classification unit), 26 ... Comparison / classification code provision unit (defect classification unit), 27 ... Correction method Generator

Claims (3)

A defect correction apparatus for correcting defects on a substrate on which a plurality of wiring patterns are formed,
A defect extraction unit for extracting a defect site on the substrate;
A position information extraction unit that extracts position information in the substrate of the defective portion extracted by the defect extracting unit,
Get process information at that time, on the basis of the position information of the defect site extracted by the process information and the position information extraction unit, from the design information of the substrate, or near the base of the defect site at that time A membrane information acquisition unit for obtaining membrane information indicating the state of
A feature extraction unit that quantifies the feature amount of the defect site extracted by the defect extraction unit;
A defect classification unit for classifying defect types based on the feature information digitized by the feature extraction unit;
A correction method generating unit that generates a defect correction method for the defective part based on the position information of the defective part, the feature information, the defect type, and the film information ;
The correction method generating unit operates the defect correction mechanism based on the generated defect correction method in by defect correction apparatus Ru and a control unit that performs defect correction. The defect correcting mechanism is:
A laser transmitter;
A slit for changing an opening area in a plane orthogonal to the optical axis and the opening area through which the laser beam passes;
A stage mounted with the substrate and moving in a plane perpendicular to the optical axis;
An optical axis moving unit that moves the optical axis minutely,
When the laser beam irradiation, while maintaining the position of the stage, defect correction apparatus of claim 1 adjust the laser beam irradiation position by using the optical axis moving unit. A defect correction method using a defect correction apparatus for correcting defects on a substrate on which a plurality of wiring patterns are formed,
Extracting a defect site on the substrate;
Acquiring positional information on the substrate of the defect site;
Get process information at that time, on the basis of the position information of the defect site and the process information, the design information of the substrate, to obtain a film information indicating the state of the peripheral or the lower the defect site at that time Process,
Quantifying the feature quantity of the extracted defect site and generating feature information;
A process of classifying defect types based on the feature information;
Generating a defect correcting method for the defective part based on the position information of the defective part, the feature information, the defect type and the film information ;
Ru defect correction method name and a process of performing the allowed by defect correction operation defect correction mechanism based on the generated defect correction method.