CN100419503C - Graphic correction device and method of manufacturing display device - Google Patents

Abstract

Provided is a pattern correction device and a method of manufacturing a display device, in which there are pixels in a region surrounded by two gate wirings and two drain wirings, and a short-circuit defect of an adjacent pixel electrode is passed through the gate wiring and the short-circuited portion. The mask of the transmissive pattern corresponding to the pattern of the drain wiring and the pixel electrode is irradiated with laser light to remove the short-circuited portion. The removal of the above-mentioned short-circuit portion can be known by comparing the information sent by the inspection device with the normal pattern, and the pattern defect formed on the substrate can be automatically corrected. Therefore, by applying the above-mentioned method to the manufacturing process of a display device, especially the photoresist pattern forming process, a display device having high-quality absorption characteristics can be realized.

Description

Graphic correction device and method of manufacturing display device

technical field

The present invention relates to a pattern correction technique for correcting missing portions coated with wiring material while using laser light to correct pattern-shaped protrusions of substrates forming predetermined patterns, and is suitable for manufacturing methods of liquid crystal display devices. In addition, the present invention relates to a technology for manufacturing a display device, and particularly relates to a technology for preventing the occurrence of defective circuit patterns produced on a TFT substrate or the like.

Background technique

A liquid crystal display device is a structure in which liquid crystal is sandwiched between two glass substrates, and resins (colored resins) that are alternately coated with blue, green, and red are formed on a glass substrate (also called a color filter (CF) substrate). On another substrate (also known as an active matrix substrate or thin film transistor (TFT) substrate), a pixel circuit composed of thin film transistors, wiring or driving circuits, etc. are formed.

When a pattern defect occurs in a color filter or wiring, it becomes a display abnormality, and the liquid crystal display device becomes a waste product. In display abnormalities, for example, in the color filter substrate, there are color defects (color mixing) caused by the color resin coated on the color filter bleeding to adjacent pixels and uneven film thickness of the resin. The resulting coating unevenness, short circuit and disconnection between wires in the active matrix substrate, etc.

The color filter and wiring of the liquid crystal display device are formed by overlapping several layers of patterns. Therefore, pattern defects must be corrected before forming the upper layer pattern. A general pattern inspection device can be used in the pattern defect detection method.

As a correction method for color filter bleeding of colored resin and wiring short circuit, as shown in Patent Document 1, a general method is to irradiate laser light to the short circuit portion and remove it for correction. There are also occasions where patterns of the same shape are repeatedly formed like a liquid crystal display device. As shown in Patent Document 2, a method of irradiating laser light through a mask having a standard pattern shape is used to remove parts different from the standard pattern for correction. method. As a method of applying a wiring material to a pattern defect portion, as disclosed in Patent Document 3, there is a method of applying a hollow straw with a tapered front port and a small diameter.

In addition, Patent Document 4 discloses a technique of disconnecting and correcting a short-circuit defect generated in a circuit pattern of a semiconductor device by laser processing. In addition, in Cited Document 5, it is disclosed that metal materials such as palladium (the material of the circuit pattern) are made into a liquid state or a gaseous state, and coating or spraying the disconnection defect part has an effect on the disconnection that occurs in the circuit pattern of the semiconductor device. Defect correction technology.

Patent Document 1: Japanese Patent Application Laid-Open No. 9-307217

Patent Document 2: Japanese Patent Application Laid-Open No. 5-27111

Patent Document 3: Japanese Patent Application Laid-Open No. 8-66652

Patent Document 4: Japanese Patent Application Laid-Open No. 10-177844

Patent Document 5: Japanese Patent Application Laid-Open No. 10-324973

Contents of the invention

When correcting a pattern defect by irradiating laser light, the operator performing the correction operation performs laser irradiation after overlapping the defect position with the laser irradiated area. In addition, when laser irradiation is performed through a mask having a standard pattern, the operator must also set the irradiation area so that it overlaps with the actual pattern. In either method, since the operator determines the laser irradiation position, the shape of the pattern after correction is not only determined by the skill of the operator, but also requires an operator for all correction machines, which increases the cost. Therefore, it is desirable that the laser irradiation area can be automatically determined.

As a method of automatically determining the laser irradiation area, there is a method of superimposing patterns by superimposing these reference points after acquiring an image of the laser irradiation area and a solid figure, and detecting previously indicated reference points through image processing. Using this method, automatic corrections can be made.

However, since the reference point must sometimes be set at a position characteristic in shape, the area in which the reference point can be set is limited. Therefore, when there is a defect in the fiducial point and when the fiducial point is separated from the defect position and cannot enter the field of view of the observation optical system at the same time, the fiducial point cannot be found and the patterns cannot be superimposed.

Therefore, it is an object of the present invention to provide a device capable of automatically correcting pattern defects while utilizing a mechanism for aligning a laser irradiation area obtained by laser irradiation through a mask with a solid pattern.

In addition, in the techniques described in the aforementioned Patent Documents 4 and 5, in order to correct a circuit pattern created on a display device, unevenness occurs in the circuit pattern due to the correction. Therefore, it is difficult to obtain circuit patterns with the required quality and precision.

In addition, in the past, the inspection of the circuit pattern was performed after the display device was completed. Since the circuit pattern is produced by laminating a plurality of pattern layers on the display device, defects in the circuit pattern can only be corrected for defects in patterns formed on the uppermost layer of the display device.

The present invention has been made in view of the above circumstances, and another object of the present invention is to prevent occurrence of defective circuit patterns produced on a display device without degrading the quality and precision of the circuit patterns.

In order to solve the above problems, a method for manufacturing a display device of the present invention, after forming an inorganic or organic film on the surface of the substrate, the film is patterned into electrodes or wirings. The manufacturing method is characterized in that it is carried out sequentially: Photoresist coating on the film, exposure of the photoresist through an exposure mask having a transmission pattern corresponding to the above-mentioned electrodes or wirings, and development of the exposed photoresist, the photoresist In the first step of forming a photoresist pattern corresponding to the transmission pattern, the above-mentioned photoresist pattern is visually inspected, and information on the position of a defect having a shape different from that of the transmission pattern in the photoresist pattern is extracted. The second step is to irradiate the defect with laser light based on the position information of the defect, thereby removing the unnecessary part of the above-mentioned photoresist pattern in the defect, and the third step of correcting the defect, and by correcting the above-mentioned defect. The photoresist pattern etches the above-mentioned film to pattern the fourth process of forming the above-mentioned electrodes or wirings. The state of the standard pattern of the same shape as the pattern of the above defect is projected onto the defect of the photoresist pattern.

A manufacturing method of a display device according to the present invention, the inorganic or organic film formed on the surface of the substrate is shaped into the pattern of the photoresist formed on the film, the manufacturing method is characterized in that it is carried out sequentially: Coating of the photoresist on the film, exposure of the photoresist by laser light passing through an exposure mask having a transmission pattern formed thereon, and development of the exposed photoresist form the photoresist into The first process of the photoresist pattern corresponding to the transmission pattern is to perform visual inspection on the above-mentioned photoresist pattern, and the second step is to extract the position information of a defect in the photoresist pattern that has a shape different from that of the transmission pattern. process, a third process of removing an unnecessary portion of the above-mentioned photoresist pattern in the defect by irradiating the above-mentioned defect with laser light based on the position information of the above-mentioned defect, and by removing the above-mentioned unnecessary portion in the above-mentioned third process The above-mentioned photoresist pattern etches the above-mentioned film to form the fourth step of the photoresist pattern. In the above-mentioned third step, the above-mentioned laser beam is processed by forming a processing transmission pattern having the same shape as the above-mentioned transmission pattern. A mask is projected onto the defect of the photoresist pattern.

A method for manufacturing a display device of the present invention, the display device sequentially stacks a first pattern layer formed with an island pattern on the surface of a substrate, and a circuit formed with a circuit extending from the top of the island pattern to the outside of the top surface The second graphic layer of graphics, the manufacturing method is characterized in that: the first graphic layer and the second graphic layer are respectively formed by performing the following first process, second process, and third process in sequence, that is, after forming Coating a photoresist on the inorganic or organic film of the above-mentioned island pattern or the above-mentioned circuit pattern, exposing the photoresist with laser light passing through an exposure mask formed with a transmission pattern, and developing the exposed The photoresist uses the first process of forming a photoresist pattern corresponding to the island pattern or the circuit pattern, and the second process of checking foreign matter mixed into the above photoresist pattern, using the photoresist pattern that has passed the above photoresist pattern. In the third process of forming the above-mentioned film into the above-mentioned island pattern or the above-mentioned circuit pattern by etching, in the above-mentioned second process, when the above-mentioned foreign matter mixed region mixed with the above-mentioned photoresist pattern is detected, before the above-mentioned third process, The above-mentioned photoresist is recoated on and around the foreign matter-incorporated region of the photoresist pattern, and then the re-coated photoresist is irradiated with laser light, and the photoresist is re-formed so as to avoid the foreign-matter mixed-in region. pattern, and using the re-formed photoresist pattern, etch the above-mentioned film in the third step.

In order to solve the above-mentioned problems, mode 1 of the present invention switches and uses a mask having the same standard pattern as the corrected pattern according to the target pattern, and by making the substrate stage for carrying the substrate and the mask stage for carrying the mask as The mechanism of synchronous movement, after using the reference point to make the positions of the substrate and the mask coincide, the substrate stage and the mask stage can be moved synchronously until they reach the defect position. Mask patterns are superimposed with high precision and laser irradiation is performed.

In addition, by using the image near the defect obtained from the inspection device to detect defects and patterns, and predetermine the reference point and laser irradiation area from the positional relationship, it is possible to solve the problem of not being able to find the reference point of the pattern on the substrate and due to large defects. While solving the problem of misalignment, it is possible to shorten the time to search for the fiducial point of the pattern on the substrate.

In addition, mode 2 of the present invention is a method of manufacturing a display device having a plurality of pattern layers by etching a layer on which a photoresist pattern is formed. Previously, an inspection step of the photoresist pattern and a correction step of correcting the photoresist pattern according to the inspection result of the inspection step are performed.

Here, the correction step may include disconnection of the short-circuit defect in the resist pattern by laser processing when the inspection result in the inspection step shows a short-circuit defect in the resist pattern. In addition, when the inspection result in the above-mentioned inspection process shows a disconnection defect of the photoresist pattern, recoating of the photoresist material to the short-circuit defect portion of the photoresist pattern and recoating the photoresist material through the recoating process may also be included. Laser processing is performed partially and the reformation of the photoresist pattern is partially completed.

Since the present invention can not only improve the accuracy of defect correction, but also automate the correction, it is possible to improve the display quality and reduce the cost. In addition, by automating correction, inspection and correction can be performed in continuous transfer, and cost reduction due to reduction in processing time of the inspection and correction process and improvement in yield can be achieved.

In addition, according to the present invention, since the photoresist pattern is corrected, the influence of pattern defect correction on the pattern layer can be reduced. In addition, since the photoresist pattern is inspected and corrected in each formation process of the pattern layer of the display device, pattern defects in each pattern layer can be corrected. Therefore, it is possible to prevent the occurrence of defective circuit patterns produced on the display device without degrading the quality and precision of the circuit patterns.

Description of drawings

Other features, objects and advantages of the present invention will become more apparent from the detailed description with reference to the following drawings, in which:

FIG. 1 is a schematic plan view illustrating an example of pixels formed on an active matrix substrate of a liquid crystal display device.

FIG. 2 is a schematic plan view showing an example in which there is a giant short across a plurality of pixel electrodes.

FIG. 3 is an explanatory diagram of a mask having a laser transmission pattern having the same shape as a wiring pattern of a short circuit portion.

Fig. 4 is an explanatory view showing the configuration of the laser irradiation optical system of the present invention.

Fig. 5 is an explanatory diagram of the system configuration of the correction device of the present invention.

Fig. 6 is an explanatory diagram of a first example of a method of obtaining reference points of a mask and a real pattern.

Fig. 7 is an explanatory diagram of a second example of a method of obtaining reference points of a mask and a real pattern.

Fig. 8 is an explanatory diagram of a third example of a method of obtaining reference points of a mask and a real pattern.

Fig. 9 is an explanatory diagram of a fourth example of a method of obtaining reference points of a mask and a real pattern.

FIG. 10 is an explanatory diagram of a configuration of a laser irradiation optical system for realizing another method of registering a relative positional relationship between a reference point of a wiring pattern and a reference point of a mask image.

FIG. 11 is an explanatory diagram of a method of planning a laser irradiation area.

FIG. 12 is an explanatory diagram of extracted graphic reference points.

FIG. 13 is an explanatory diagram of a detection method of a laser irradiation area.

FIG. 14 is an explanatory diagram of a method of setting a laser irradiation area.

FIG. 15 is an explanatory diagram of the shape of the laser irradiation area.

FIG. 16 is an explanatory diagram of mask alignment in the middle of a correction process of the wiring including the defect shown in FIG. 11 .

Fig. 17 is an explanatory diagram of another setting method of the laser irradiation area.

Fig. 18 is an explanatory diagram of still another setting method of the laser irradiation area.

FIG. 19 is an explanatory diagram of a configuration example of a mask holder.

FIG. 20 is an explanatory diagram of a configuration of a laser irradiation optical system for realizing still another method of registering a relative positional relationship between a reference point of a wiring pattern and a reference point of a mask image.

Fig. 21 is an explanatory diagram of another configuration example of the defect correction device of the present invention.

FIG. 22 is an explanatory diagram showing an example of the overall configuration of the defect correction device of the present invention.

FIG. 23 is an explanatory diagram showing another example of the overall configuration of the defect correction device of the present invention.

Fig. 24 is an explanatory diagram of attachments of the optical unit.

FIG. 25 is an explanatory diagram of a correction flow when performing foreign matter removal.

Fig. 26 is a process diagram illustrating Embodiment 2 of the present invention.

Fig. 27 is a process diagram illustrating Embodiment 3 of the present invention.

Fig. 28 is a process diagram illustrating Embodiment 4 of the present invention.

FIG. 29 is a diagram illustrating correction when there is a disconnection in the wiring pattern of the liquid crystal display device.

FIG. 30 is a diagram showing the configuration of a preferred correcting apparatus for implementing the disconnection defect correcting method of Embodiment 5. FIG.

Fig. 31 is an enlarged view of a material application mechanism illustrating a state of application of a material for disconnection correction.

Fig. 32 is an explanatory diagram of a material application mechanism.

Fig. 33 is an explanatory diagram of a state where a correction material is applied by a material application mechanism to the portion of the disconnection defect in Fig. 29 .

Fig. 34 is an explanatory diagram of a correction method for a disconnection defect.

Fig. 35 is a diagram illustrating the flow of correction work as Embodiment 5 of the present invention.

Fig. 36 is a diagram illustrating an inspection and correction system using a laser electronic circuit pattern correction device having a mask exchange function.

Fig. 37 is a diagram illustrating another inspection and correction system using a laser electronic circuit pattern correction device having a mask exchange function.

Fig. 38 is a diagram illustrating another arrangement form of the electronic circuit board.

Fig. 39 is a process chart for explaining a circuit pattern forming process according to Embodiment 8 of the present invention.

FIG. 40 is a process diagram for explaining the resist pattern correction step (S3906) shown in FIG. 39. FIG.

FIG. 41 is a diagram for explaining a short defect correction process.

Fig. 42 is a diagram for explaining a disconnection defect correction process.

Fig. 43 is a diagram for explaining a foreign matter entry defect correction step.

FIG. 44 is a process diagram for explaining the circuit pattern forming step (S3903) shown in FIG. 39. FIG.

Fig. 45 is a diagram for explaining a case where exposure is performed by partially changing an exposure mask.

46 is a diagram showing an example of a resist pattern inspection and correction system used in the resist pattern inspection step and the resist pattern correction step of the eighth embodiment.

FIG. 47 is a schematic configuration diagram of the resist pattern correcting apparatus shown in FIG. 46. FIG.

Fig. 48 is a schematic sectional view of the dispenser.

FIG. 49 is a diagram showing a modified example of the recoated photoresist material.

Detailed ways

Preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings of the examples. In addition, in the description of the embodiment, the wiring correction of the liquid crystal display device was described as an example, but it can also be applied to the correction of patterns formed on a plane in general, and is not limited to the liquid crystal display device.

[Example 1]

Generally, a liquid crystal display device is a structure in which liquid crystal is sandwiched between two glass substrates, and the molecules of liquid crystal in each pixel are affected by the electric field in the capacitor formed by the pixel electrodes of multiple images arranged in a matrix, the opposite electrode and the liquid crystal. The orientation is controlled to generate an electronic latent image, and in a transmissive liquid crystal display device, the electronic latent image is visualized to display an image by controlling the transmittance of light from a backlight provided on the back. A color image is displayed by forming a circuit for controlling voltage applied to pixel electrodes on an active matrix substrate, and forming, for example, three-color color filters on a color filter substrate.

FIG. 1 is a schematic plan view illustrating an example of pixels formed on an active matrix substrate of a liquid crystal display device. Two adjacent pixels are shown in FIG. 1 . Various wirings and electrodes formed on the active matrix substrate are composed of thin-film multilayer circuits with insulating layers interposed therebetween.

In FIG. 1 , a plurality of gate wirings 31 are formed in parallel in one direction on a substrate 9 preferably made of glass. On a part of the gate wirings 31 , a gate electrode 31A of a thin film transistor is formed protruding into the pixel. An island-shaped portion of a semiconductor layer (here, an a-Si layer) 32 as an active layer is patterned over the gate electrode 31A.

A gate insulating layer (not shown) is formed to cover the gate wiring 31 , and a plurality of drain wirings 33 insulated by the gate insulating layer are formed in parallel in the other direction crossing the gate wiring 31 . One pixel is formed in a region surrounded by two gate wirings 31 and two drain wirings 33 . A part of the drain wiring 33 extends over the semiconductor layer 32 to become the drain electrode 33A of the thin film transistor. In addition, on the semiconductor layer 32 , a source electrode 33B forming a channel of the thin film transistor is formed in the same layer as the drain electrode 33A and closely faces the drain electrode 33A.

On the upper layer of the gate wiring 31 and the drain wiring 33, a passivation layer film is formed, and the pixel electrode 34 is formed thereon. The pixel electrode 34 is preferably a transparent electrode made of ITO, and is electrically connected to the source electrode 33B through a contact hole not shown in the middle. In addition, the drain electrode 33A and the source electrode 33B can be interchanged during operation, but they are described as above for convenience of description.

The gate wiring 31 is a scanning wiring, and the drain wiring 33 is a signal wiring. The thin film transistor connected to the gate wiring 31 selected by the scanning signal is turned ON (conducting), and the display of the drain wiring 33 is generated and supplied to the pixel electrode 34. data corresponding to the voltage. An electric field having a magnitude corresponding to the voltage generated in the pixel electrode 34 is generated between the pixel electrode 34 and a counter electrode (not shown). This electric field is used to control the molecular orientation of the liquid crystal to control the transmission amount of the illuminating light emitted from the backlight to form a visible picture.

Such a thin-film multilayer circuit is generally formed in this order by using a photolithography technique, including a gate wiring 31 , a gate insulating film, a semiconductor, a drain electrode, and a pixel electrode. Due to overlapping parts of each layer, pattern abnormalities such as short circuit and disconnection must be corrected before forming the next layer.

In forming wiring by photolithography, first, the wiring material is formed into a uniform thin film over the entire substrate, and a photosensitive resin is applied as a photoresist. Afterwards, the photoresist is exposed to light through a mask for forming a circuit pattern. When using a positive photoresist, the photosensitive part can be removed by development to form a photoresist pattern. Wiring is then formed through an etching process and a photoresist stripping process.

A short, which is one of pattern abnormalities, is a portion remaining without being etched, and is caused by remaining photoresist and insufficient etching. In particular, foreign matter adhered during the coating of the photoresist gathers the photoresist around the foreign matter due to surface tension, resulting in a large amount of photoresist residue and a wiring short across several pixels.

FIG. 2 is a schematic plan view showing an example in which there is a giant short circuit 21 across a plurality of pixel electrodes. It shows the state where the pixel electrode 34 is short-circuited across four adjacent pixels. A general laser processing machine can irradiate with a rectangular laser beam and remove the short-circuited part several times. However, as shown in FIG. 2, since the original wiring pattern is not known when the pattern of the pixel electrode is completely destroyed, it is difficult to restore the shape of the electrode with high precision.

Therefore, when removing the short-circuit 21 by irradiation with laser light through the mask 5 having the same wiring pattern shape of the short-circuit 21 portion as in FIG. High-precision correction is possible.

Since the circuit of the liquid crystal display device is formed by laminating several layers of patterns, when laser irradiation is performed on parts other than pattern abnormalities, the already formed lower layers may be affected. A case where the pattern of the pixel electrode 34 is abnormal will be described as an example. Generally, ITO (indium tin oxide) is used for the pixel electrode 34, and aluminum is used for the gate wiring 31 and the drain wiring 33. ITO has a high absorption rate at a long wavelength, for example, 200 to 300 nm, while aluminum has a high reflectance at this wavelength.

Therefore, only ITO can be removed when irradiated with laser light having a wavelength of 200 to 300 nm. When laser correction is performed using a mask with a predetermined wiring pattern, even parts other than the abnormal part of the correction pattern are irradiated with laser light, and only the part that should be corrected can be removed by utilizing the difference in laser absorption characteristics of the material.

Fig. 4 is an explanatory view showing the configuration of the laser irradiation optical system of the present invention. In addition, FIG. 5 is an explanatory diagram of the system configuration of the correction device of the present invention. The beam expander 2 expands the beam diameter of the irradiation laser 100 emitted from the laser oscillator 1 to a predetermined value, and the homogenizer 3 ensures uniformity of laser intensity over the entire laser irradiation area. Make the shaped laser pass through the mask 5 arranged on the mask stage 4, and after passing through the imaging lens 6 and the objective lens 7, the wiring patterns 10 or electrodes on the glass substrate 9 mounted on the substrate stage 8 etc. Correction point 11 is irradiated.

Imaging lens 6 and objective lens 7 are configured so that the image of mask 5 can be projected onto glass substrate 9, and with the ratio of the focal length of imaging lens 6 and objective lens 7 (the focal length of M=objective lens 7/the focal length of imaging lens 6) A multiple of the size projects the mask image onto the glass substrate 9 . With this optical system configuration, laser light can be irradiated to the area of the transmission portion of the reduction mask 5 .

On the mask 5 is formed a mask pattern 16 which is 1/M times the standard pattern of the wiring pattern 10 formed on the glass substrate 9 . The mask pattern 16 can be formed of a material with a high reflectivity to the laser light 100, such as aluminum.

The mask stage 4 and the substrate stage 8 can move in a plane perpendicular to the optical axis direction of the laser irradiation optical system, and can be moved in synchronization with each other. Since the image of the mask 5 mounted on the mask stage 4 is projected onto the glass substrate 9 at a scale reduced by 1/M, when the movement amount of the mask stage 4 is V, the projected image moves by V/ M.

Therefore, by synchronizing the movement of the mask stage 4 and the substrate stage 8 and making the ratio of the movement amount of the mask stage 4 to the movement amount of the substrate stage 8 equal to the ratio of the size of the mask pattern 16 to the wiring pattern 10 , the wiring pattern 10 and the laser irradiation pattern can be kept consistent and moved.

As described above, the ratio of the sizes of the mask pattern 16 and the wiring pattern 10 is equal to the ratio of the focal lengths of the imaging lens 6 and the objective lens 7 . Generally, since the imaging lens 6 is fixed, the size ratio of the mask pattern 16 and the wiring pattern 10 is determined by the objective lens 7 . Therefore, a mechanism capable of changing the ratio of the movement amount of the mask stage 4 to the movement amount of the substrate stage 8 in conjunction with the exchange mechanism of the objective lens 7 may be added.

In order to make the position of the laser pattern irradiated through the mask 5 and the wiring pattern 10 on the glass substrate 9 coincide, generally, reference points are set on each pattern, and the positions are overlapped by making the relative position of each reference point reach a predetermined value. . Common methods such as pattern matching can be used for the detection of reference points.

Pattern matching is a method of registering the image of the characteristic part of the pattern as a template in advance, setting a reference point on the image, and detecting the reference point by detecting the place consistent with the registered image from the image of the real pattern. The reference point of the wiring pattern is easy to detect by using this pattern matching method.

Next, the detection method of the reference point of the laser pattern irradiated through the mask 5 will be described with reference to FIG. 4 . First, when the illumination light irradiated from the light source 12 is condensed by the lens 13 and irradiates the mask 5 through the semi-transparent mirror 48a, the image of the mask 5 (mask image, mask pattern) is reflected on the glass by the imaging lens 6 and the objective lens 7. imaging on substrate 9.

When storing this mask pattern, since it must not be possible to write the part other than the mask image, it can be projected on the pattern projection part 20 on the substrate stage 8 or a plate-shaped jig having the same thickness as the glass substrate 9 can be placed onto the substrate stage 8 and project the mask image onto the fixture surface.

Since the imaging position is the surface of the glass substrate 9, the height of the figure projection part 20 is consistent with the surface of the glass substrate. In order to correspond to the change of the thickness of the glass substrate 9, the distance between the objective lens 7 and the glass substrate 9 can be changed. mechanism. When the substrate stage 8 is at a position where the glass substrate 9 is loaded/unloaded, the pattern projecting unit 20 is attached to a position where the pattern projecting unit 20 enters the field of view. As a result, the position of the laser irradiation area can be checked every time a glass substrate to be corrected is loaded on the substrate stage 8 and erroneous correction can be prevented.

Laser light for processing may also be used as the illumination light for illuminating the mask 5 . Generally, for the imaging lens 6 and the objective lens 7 used for irradiating laser light, devices of complementary colors matching the wavelength of the irradiating laser light are used. Therefore, when light having different wavelengths is used as illumination light, the position of the mask image will shift due to the difference in refractive index. When using a processing laser, there is an advantage that such a deviation does not exist. It can be easily achieved by reducing the laser output or using unprocessable materials in the mask image projection part.

When the figure image projected on the figure projection part 20 is captured by the CCD camera 14 through the half mirror 48b, in order to capture an image of the entire mask, it is captured by a single-step feed method. At this time, the movement amount of the mask stage 4 is associated with the imaged image and stored in the mask image storage unit 60 .

The system configuration of the correction device in Fig. 5 includes: a laser oscillator 1 and a laser controller 1A; a lens switching machine 40; a lens controller 40A; a mask stage 4; a mask stage controller 4A; image input device 14; substrate stage 8; substrate stage controller 8A; substrate stage encoder 8B; mask image storage unit 60; reference point storage unit 70 and integrated stage controller 74.

It also includes: an image data acquisition unit 71; an irradiation area planning unit 72; a device controller 73 and a foreign matter removal nozzle 75; each component is connected according to the control relationship and the controlled relationship shown by the arrows in the figure. The reference point of the laser pattern is set at a position close to the reference point in the wiring pattern set in the mask image stored in the mask image storage unit 60 .

In addition, it is also possible to adopt a mechanism in which the orientation of the semi-transparent mirror 48b in FIG. 4 can be rotated by 90 degrees, switched, and the lens 15 can be changed to a lens with the same focal length as the objective lens 7, so that the mask image is directly imaged on the CCD. The method of shooting on the camera 14.

Next, a method for obtaining the relative positional relationship between the reference point of the laser pattern and the reference point of the wiring pattern when the laser pattern and the wiring pattern overlap will be described with reference to FIGS. 4 and 6 . In addition, FIG. 6 is an explanatory diagram of a first example of a method of obtaining reference points of a mask and a real pattern. The glass substrate 9 is placed on the substrate stage 8 of FIG. 4 , and the pattern image (A) of normal wiring is captured by the CCD camera 14 and displayed on the monitor 64 . At this time, as shown in FIG. 6 , the imaged wiring pattern image (A) and the image (B) stored in the mask image storage unit 60 are displayed as a superimposed composite image (C) of the translucent image.

The composite image (C) in which the semi-transmissive images are superimposed can be easily displayed by calculating the average value of the luminance value of the wiring pattern image (A) and the luminance value of the mask image (B). The mask image on the screen is displayed in synchronization with the mask stage 4, and when the mask stage 4 is moved, the mask image on the screen of the monitor 64 also moves. The operator moves the substrate stage 8 or the mask stage 4 so that the two images coincide, and when they coincide, the reference points of the pattern image (A) of the wiring and the reference points of the mask image (B) are stored. relative position.

Fig. 7 is an explanatory diagram of a second example of a method of obtaining reference points of a mask and a real pattern. As another method of simultaneously displaying the graphic image (A) of the wiring and the mask image (B), as shown in FIG. 7, the mask edge image (B) of the outline extracted from the mask image (B) ') is a display method superimposed on the graphic image (A) of the wiring. The mask edge image (B') is composed of a portion through which illumination light passes and a portion that is shielded, and is an image with high contrast.

Therefore, its contour extraction can use a general edge detection method. As a method of superimposing the wiring pattern image (A) and the mask edge image (B'), when the wiring pattern image (A) is a bright (high luminance value) image, the image at the outline position is displayed in black. (a value whose luminance value is close to 0), and when the wiring pattern image (A) is a dark (low luminance value) image, it is displayed in white (a value whose luminance value is close to the maximum value).

Fig. 8 is an explanatory diagram of a third example of a method of obtaining reference points of a mask and a real pattern. In addition, when the graphic image (A) of the wiring is a bright (high luminance value) image, the image at the outline position is represented by black (a value close to 0 luminance value). An explanatory diagram of the fourth example of the reference point method is another reference point acquisition method using the wiring pattern image (A) and mask image (B) explained in FIG. 6 . In FIG. 8 , a part of the mask image (B) is displayed superimposed on the pattern image (A) of the wiring, and the overlapping of the images at the boundary between the two images is compared.

In addition, in FIG. 9 , a part of the mask image (B) is superimposed on the left and right with the wiring pattern image (A) and displayed, and the superimposition of the images at the boundary between the two images is compared.

FIG. 10 is an explanatory diagram of a configuration of a laser irradiation optical system for realizing another method of registering a relative positional relationship between a reference point of a wiring pattern and a reference point of a mask image. The same reference numerals as in FIG. 4 denote the same functional parts, and overlapping descriptions are omitted. FIG. 10 is a configuration in which a light source 12b, a CCD camera 14b, and lenses 13b and 15b are added to the configuration in FIG. 4 . In FIG. 10 , first, the mask image 1 imaged on the pattern projection unit 20 is imaged by the CCD camera 14 a while the mask image illuminated by the light source 12 b is imaged on the imaging element of the CCD camera 14 b to be imaged. . The images captured by the CCD cameras 14a and 14b are displayed on the same screen by the method shown in FIGS. 6 to 9, and the positions of the CCD cameras 14a and 14b are adjusted to match.

After that, the glass substrate 9 is placed on the substrate stage 8, and the light source 12b is turned on so that the wiring pattern on the glass substrate 9 can be photographed. The images captured by the CCD cameras 14a, 14b are displayed on the same screen using the method shown in FIGS. The relative positions of the reference point of the graphic image and the reference point of the mask image are stored.

When the corrected part is larger than the laser irradiation area, the laser can be irradiated several times to remove the corrected part. At this time, correction is performed while moving in steps with the reference point of the figure as the starting point. When the fiducial point of the figure is included in the correction part, since the fiducial point cannot be found by pattern matching, the fiducial point of the pattern image of the adjacent wiring is used. Generally, when the reference point cannot be found at the initial position, a method of searching until the reference point is found can be adopted by moving the field of view in a spiral shape.

However, it takes time to search until a valid reference point is found, and sometimes a defect cannot be corrected within the specified time, and a defect may be mistakenly determined as a reference point. Therefore, the reference point and step area to be used are planned and set in advance using the defect vicinity image acquired from the inspection device.

The following describes how to plan the laser irradiation area. First, a method of detecting a figure reference point will be described. 11 is an explanatory diagram of a planning method of a laser irradiation area, FIG. 11( a ) is a plan view of a wiring pattern of a pixel having a defective portion, and FIG. 11( b ) is an explanatory diagram of an extracted defective portion. In addition, FIG. 12 is an explanatory diagram of extracted graphic reference points. While performing image processing on the wiring pattern in FIG. 11( a ) to extract the defect image shown in FIG. 11( b ), pattern reference point candidates 81 a to 81 e are detected by pattern matching shown in FIG. 12 .

In this pattern matching, since the detection error becomes larger when there is a defect within the range of the matched template, a match that does not include the defect 21 is selected from the positional relationship between the detected matching regions 80a, 80b, 80c, 80d, 80e and the defect 21 area. According to this rule, in the example of FIG. 12 , the matching area 80 c is excluded because it contains the defect 21 . Select the pattern fiducial point to be used from the remaining matching areas 80a, 80b, 80d, 80e, for example, to shorten the moving time from the fiducial point, select the pattern fiducial point 81a closest to the defect. In addition, when two or more reference points are used, in order to further improve the accuracy of position overlap, graphic reference points 81a and 81d may also be selected.

An example of a method of setting a laser irradiation area will be described below. FIG. 13 is an explanatory diagram of a detection method of a laser irradiation area. The defects 21 shown in Fig. 11 are given by binarized pixels. As shown in FIG. 13, the projection value on the X-axis can be calculated by calculating the number of pixels contained in the defect for each X-coordinate value. Similarly, the projection value is also calculated for the Y axis. The defect area 85 can be detected from the start point and the end point obtained from the projected values on the X axis, Y axis. Generally, the laser irradiation area 86 is set to be larger than the defect area 85 by a predetermined value in order to prevent the occurrence of unprocessed parts. Here, the relative positional relationship between the graphic reference point 81 and the laser irradiation area 86 is stored.

FIG. 14 is an explanatory diagram of a method of setting a laser irradiation area. In addition, FIG. 15 is an explanatory diagram of the shape of the laser irradiation region. When the defect 21 is too large to irradiate the entire defect with one laser irradiation, as shown in FIG. 14( a ), laser irradiation may be performed several times. Since the size of the laser irradiation region 86 is known, the laser irradiation regions 86 a , 86 b , 86 c , and 86 d are set starting from a point shifted by a predetermined amount from the upper left point of the defect region 85 so that the defect region 85 is completely covered. In order to prevent the occurrence of unprocessed parts, each laser irradiation area is set to overlap by a certain amount. The relative positional relationship between the graphic reference point 81 and each laser irradiation area is stored as in the case of one laser irradiation.

The laser irradiation area 86 may also be hexagonal as shown in FIG. 14( b ). Since the field of view of a general optical system is circular, compared with a square shape, a hexagonal shape has a larger area to be irradiated at one time, which has the effect of reducing the number of beams. Since the laser irradiation areas 86 are set to overlap in order to prevent the occurrence of unprocessed parts, there are areas where laser irradiation is performed multiple times. In order to avoid this, it is possible to adopt a chamfered shape as shown in FIG. 15( a ) and a circular shape as shown in FIGS. 15( c ), (d).

When laser irradiation is divided into several times, the step movement accuracy of the laser irradiation area must be high. In the manufacture of liquid crystal display devices, the size of the substrate is increasing day by day, and the substrate used is from more than 1 m to 2 m. At the same time, the substrate stage is also enlarged, and in order to ensure the sub-micron step movement accuracy, the movement speed is reduced and the cost is increased. In correction devices, therefore, a realignment generally has to be carried out in the correction position. However, as described above, since there is a restriction on the position of the reference point used for alignment, it is not necessarily limited to the existence of the reference point at the corrected position. Therefore, alignment is performed using the wiring pattern and the already processed wiring pattern.

FIG. 16 is an explanatory diagram of mask alignment in the middle of a correction process of the wiring including the defect shown in FIG. 11 . For this defect, as shown in FIG. 14, the defect 21 was removed by four laser irradiations. FIG. 16 is an explanatory diagram after laser irradiation is performed on the laser irradiation region 86 a of FIG. 14( a ). Thereafter, the laser irradiation region 86b is irradiated with laser light and moved step by step, but if the movement accuracy is insufficient, the laser irradiation region 86b is precisely aligned using the pattern of the wiring portion 87a and the normal portion 87b formed by correction.

With this method, the alignment does not deviate even with several step movements. In addition, even for defects larger than the laser irradiation area, fine alignment can be performed by partially overlapping the corrected pattern.

Fig. 17 is an explanatory diagram of another setting method of the laser irradiation area. In this setting method, the laser irradiation area 86 of the pattern reference point 81 is set in advance, and the laser irradiation area is selected by comparing with the defect image in FIG. 11( b ). In this way, since the laser irradiation area can be set in advance, for example, as shown in FIG. High-precision processing, and the overlapping portion of the laser irradiation area can be set to a place where the influence of laser irradiation is small.

Fig. 18 is an explanatory diagram of still another setting method of the laser irradiation area. The preset laser irradiation area 86 may have the shape shown in FIG. 18 . FIG. 18 is an example of setting laser irradiation areas only in areas where defects may occur. This is an effective area designation method when there is an area where laser irradiation cannot be performed. In this example, too, the laser irradiation area can be selected in comparison with the defect image of FIG. 11( b ).

FIG. 19 is an explanatory diagram of a configuration example of a mask holder. As mentioned above, the TFT substrate of the liquid crystal display device is formed by overlapping several layers of thin films. Therefore, there are several kinds of correction patterns. If only one pattern can be corrected in one device, it is inefficient to replace the mask every time the correction layer is changed or the type is changed.

Therefore, a mask stage having the structure shown in FIG. 19( a ) can be used. Several kinds of masks 5 are mounted on the mask holder 51 provided on the X stage 52 . The mask holder 51 is movable relative to the X stage 52, and can switch the mask used for the correction object.

The X stage 52 and the Y stage 53 are used to align the positions of the wiring patterns. The θ stage 54 is used when correcting the inclination of the wiring pattern, stores the adjustment value when the mask 5 is set, and performs rotation correction using the stored value when the mask is switched. Slits with simple shapes such as rectangles and circular oblique lines are opened in the slit plate 56, which can be used for fine-tuning the shape of a pattern corrected by a mask or for removing foreign matter by concentrated laser irradiation.

Fig. 19(b) is another example of a mask stage. The mask holder 51 is fixed on the X stage 52 , and the mask can be switched by moving the X stage 52 and the Y stage 53 .

In addition, in FIG. 4, instead of the mask 5, a liquid crystal display device may be used to form a necessary pattern. Since arbitrary patterns can be formed by using a liquid crystal display device, the above-mentioned mask switching is unnecessary. At this time, in order to reduce the damage of the liquid crystal display device by laser light, the reduction ratio of the pattern can be increased. For example, when the reduction ratio is 1/100, damage can be reduced because the energy density in the liquid crystal display device at the mask position is 1/10000.

Fig. 20 is an explanatory diagram of another laser irradiation optical system configuration in which a laser irradiation region having the same shape as the wiring pattern is set. The same reference numerals as in FIG. 4 and FIG. 6 correspond to the same functional parts, and repeated explanations are omitted. In FIG. 20, the reflection mirror 18 is made by vapor-depositing a material with high laser reflectivity such as aluminum on a transparent substrate such as quartz glass. The reflective portion made of aluminum or the like forms a pattern that becomes the laser irradiated area in FIG. 3 when irradiated with laser light, like the mask 5 .

In addition, the reflective surface 17 of the reflective mirror is formed at an angle with respect to the reflective mirror 18 so that the obliquely incident laser light is irradiated perpendicularly to the substrate. Also, by using a DMD (Digital Micromirror Device) instead of the mirror 18, the DMD can be controlled to have a graphic shape generated from CAD data, and laser light having an irradiation area of an arbitrary shape can be irradiated.

Fig. 21 is an explanatory diagram of another configuration example of the defect correction device of the present invention. The same reference numerals as those in Fig. 4, Fig. 10, and Fig. 20 correspond to the same functional parts, and duplicate descriptions will be omitted. In FIG. 21 , the point-shaped laser light output from the laser oscillator 1 is reflected by the galvano mirror 98 , and the laser light is vertically incident on the mask surface by the Fθ lens 97 . By changing the angle of the galvano mirror 98, the entire surface of the mask 5 can be irradiated with laser light. In this method, since the light spot is scanned, it is possible to irradiate the laser beam with a uniform intensity within the surface of the mask 5 .

FIG. 22 is a diagram showing an example of the overall configuration of the defect correction device of the present invention. In FIG. 22( a ), glass substrate 9 is set on substrate stage 8 , and substrate stage 8 has a drive shaft that moves in the Y direction. The optical unit 101 is a unit in which the optical systems shown in FIGS. 4 , 10 , 20 , and 21 are installed, and has a drive shaft that moves in the X direction and the Z direction. The positioning of the laser irradiation position is carried out by the Y-axis movement of the substrate stage 8 and the X-axis movement of the optical unit 101 , and the focus is adjusted on the Z-axis of the optical unit 101 . This configuration, on the contrary to the drawback that the footprint becomes larger, can perform positioning with high precision because the two axes can be independently controlled.

FIG. 22( b ) is another device configuration example, in which the optical unit 101 is mounted on the optical unit stage 102 . The optical unit stage 102 has a drive shaft that moves in the Y direction, and the optical unit 101 has a drive shaft that moves in the X direction and the Z direction. This configuration has the advantage that the footprint of the device can be reduced.

FIG. 23 is a diagram showing another example of the overall configuration of the defect correction device of the present invention. In FIG. 23( a ), an optical unit stage 103 is attached to a rail 104 provided on a bed. Generally, since the substrate 9 is disposed on the vibration-absorbing table 105, not only is it a structure that does not transmit the vibration generated when the optical unit 101 moves to the glass substrate, but the vibration-absorbing table 105 can be made very small. As in FIG. 22( a ), a structure in which the substrate stage 8 is moved and the optical unit stage 102 is fixed may be used.

FIG. 23( b ) shows a structure in which the substrate stage 8 is tilted, and the glass substrate 9 is adsorbed on the substrate stage 8 and moved. For loading and unloading of the glass substrate, there are a method of making the substrate stage 8 horizontal, a method of using a transfer device which can keep the glass substrate 9 upright, and the like. In order to reduce the platform area of the device, that is, the coverage surface, there is no limit to the range of the angle θ formed by the substrate stage 8 and the horizontal plane, especially in the range of 80 to 95 degrees, which can prevent processing chips generated during laser processing reattachment and lens contamination.

FIG. 24 is an explanatory diagram of attachments of the optical unit 101. FIG. Fig. 24(a) shows an example of a lens guard. In order to prevent processing chips and smoke generated during laser processing from adhering to the surface of the lens 110, such a lens protective cover is provided in front of the lens. Since the protective cover itself is contaminated when the lens protective cover is used fixedly, the film-shaped protective cover 112 is installed to be wound on the reel 113, and is wound up at the end of laser irradiation so that a new part can be used at the time of laser irradiation.

In order to facilitate the exchange of the lens 110 , a lens guard guide 111 is provided to form a gap between the lens guard 112 and the lens 110 . In order to replace the protective cover regularly when the processing object is a processed object with few spatters, a cover-shaped protective cover that can be attached can be attached to the front of the lens. In addition, the method of installing a pipe for sucking smoke near the lens and the method of blowing nitrogen or air in front of the lens to prevent dirt from adhering can also be used. Fig. 24(b) is an example of a foreign matter removing nozzle. Foreign matter is also included in the defects detected by the pattern inspection, but there is also foreign matter 116 adhering only to the surface of the glass substrate 9 . Such foreign matter 116 can be easily removed by blowing nitrogen or air from the foreign matter removing nozzle 115 .

FIG. 25 is an explanatory diagram of a correction flow when performing foreign matter removal. First, the glass substrate is loaded on the correction device (step 1, described as S1 below), and after moving to the defect position (S2), an image is acquired (S3). After blowing nitrogen or air to remove foreign matter (S4), an image is acquired again (S5). Compare the images before and after blowing nitrogen or air (S6), and if there is a difference in the image to determine that the foreign matter has been removed, laser irradiation is not performed, and when there is no difference in the image to determine that the foreign matter is fixed, laser irradiation is performed (S7). It is checked whether there is a next defect (S8), and if there is a defect, the above steps are repeated, and when there is no defect, the glass substrate is unloaded (S9).

Whether defect correction is necessary can be judged by using the image near the defect shown in Fig. 11 . As shown in FIG. 12 , since the pattern reference points 81 a to 81 e can be detected, it is easy to determine whether it is a defect that needs to be corrected or a defect that does not need to be corrected from the relative positional relationship with the reference point. Laser irradiation is not required for defects that do not require correction. According to this embodiment, even if there are short-circuit defects spanning several pixels, they can be collectively corrected, so it is possible to improve correction accuracy while reducing costs.

[Example 2]

Fig. 26 is a process diagram illustrating Embodiment 2 of the present invention. FIG. 26 shows a manufacturing process of a glass substrate forming a switching element thin film transistor used in a liquid crystal panel in Example 2. FIG. The glass substrate on which this switching element is formed is generally referred to as a "TFT substrate" or "array substrate", and is hereinafter referred to as a TFT substrate. In Example 2, first, an inorganic film or an organic film is formed on a glass substrate provided when manufacturing a TFT substrate (S10).

A typical inorganic substance includes a metal material constituting the wiring of the TFT substrate. A photoresist is coated on the film-formed glass substrate and fired (S11).

Here, the so-called photoresist is a photosensitive material, which is used to process the film-forming material into a predetermined shape. Thereafter, by exposing the photoresist to light (S12), the photoresist is exposed to light using the wiring pattern constituting the TFT substrate. After that, by developing the photoresist (S13), the same photoresist pattern as the wiring pattern constituting the TFT substrate remains on the film-formed glass substrate.

Thereafter, an appearance inspection is performed on the photoresist pattern formed on the glass substrate (S14). In the visual inspection, the contrast between the film-forming material and the photoresist pattern formed into the film on the glass substrate is obvious. Therefore, it is possible to discriminate the portion where the shape of the photoresist is abnormal due to nucleation of foreign matter or the like. After that, based on the visual inspection of the photoresist pattern, information on fatal defect positions is extracted (S15).

Pattern defects with abnormal shape and abnormal size exceeding general management standards in visual inspection are extracted as fatal defect candidates. Abnormalities in the shape of the photoresist can be broadly classified into short-circuit defects in which multiple parts that should be independent become connected, and disconnection defects in which parts that should be connected are missing. In Example 2, the short-circuit defect will be described as an object in particular.

Afterwards, based on the design information of the wiring pattern constituting the liquid crystal panel, the position information of the fatal area is obtained, and the position of the real fatal defect is extracted from the photoresist pattern shape, size, and position information obtained from the appearance inspection of the photoresist pattern. information. Here, the information on the shape, size, and position information of the fatal photoresist pattern is obtained from the shape and electrical characteristics of the wiring pattern, and is different from the design specifications of the TFT substrate.

Then, resist pattern correction is carried out (S16). In resist pattern correction, processing specifications are determined using the shape and size of critical defects obtained at the time of extracting critical defect position information. The resist defect correction in this case is the same as that already described in the other embodiments, and the short-circuited resist is removed using a laser or the like. Since there are no defects on the glass substrate after the photoresist pattern correction, in the etching (S17) which is the next step, the part not covered with the photoresist among the film-formed materials is processed.

Then, by completing the stripping of the photoresist (S18), the film-formed material remains on the glass substrate in a regular pattern shape (S19). A TFT substrate is completed by repeating a series of processes from film formation ( S10 ) to resist stripping ( S18 ) a predetermined number of times.

[Example 3]

Fig. 27 is a process diagram illustrating Embodiment 3 of the present invention. In FIG. 27, a manufacturing process of a TFT substrate used in a liquid crystal panel is shown. In Example 3, first, an inorganic film or an organic film is formed on a glass substrate provided when manufacturing a TFT substrate (S30). Here, since a transparent metal ITO (indium tin oxide) provided on a TFT substrate can be mentioned as a typical inorganic material for film formation, the following description will be made using an ITO film as a representative example.

At the stage where film formation is completed, visual inspection is performed (S31). Here, since ITO is transparent, it can be seen that there are foreign matter on ITO and foreign matter in the lower layer of ITO. The results of the visual inspection at the film formation completion stage are accumulated in the process of depositing the visual inspection results after film formation ( S41 ).

After that, a photoresist is coated on the TFT substrate, and after sintering (S32), exposure (S33), and development (S34), an appearance inspection is performed on the photoresist pattern (S35). The results of the visual inspection of the resist are accumulated in data by the process of depositing the results of the visual inspection of the resist pattern ( S42 ).

Here, in the results of the visual inspection of the resist pattern, both information on the location of the abnormal shape of the resist on the formed film and the presence of foreign matter under the transparent film despite the absence of the resist were observed. Then, only the portion with an abnormal shape of the resist pattern occurring after the application and firing of the resist is separated and extracted by difference processing (S43). Extraction of fatal defect position information from the separated and extracted photoresist pattern shape abnormal part and the design specification of the TFT substrate is carried out (S36).

Then, by performing photoresist pattern correction (S37), performing etching (S37), and performing photoresist stripping (S39), the film-formed material remains on the glass substrate in a regular pattern shape. By repeating a series of processes from film formation (S30) to resist stripping (S39) a predetermined number of times, the TFT substrate is completed.

[Example 4]

Fig. 28 is a process diagram illustrating Embodiment 4 of the present invention. In FIG. 28, first, an inorganic film or an organic film is formed on a glass substrate provided when manufacturing a TFT substrate (S50). Thereafter, an appearance inspection is performed (S51). Here, foreign matter that exists on the film and is larger than a predetermined management size is extracted. Thereafter, foreign substance removal is performed (S52).

A contact foreign matter removal method and a non-contact foreign matter removal method can be applied to the foreign matter removal here. As a contact-type removal method, there is a method in which the brush is brought into contact with the foreign matter on the film to remove it, typified by rotating the brush or reciprocating vibration.

In addition, a method of removing foreign matter using instruments represented by tweezers, and a method of removing foreign matter using a needle-shaped structure or a structure similar to a blade can be used. Then, at this time, the storage process of the foreign object removal coordinate information is implemented (S63). On the other hand, laser beams and high-pressure fluids can be cited as non-contact foreign matter removal methods.

After that, a photoresist is coated on the TFT substrate, and after sintering (S53), exposure (S54), and development (S55), an appearance inspection is performed on the photoresist pattern (S56). Afterwards, after the fatal defect position information extraction (S5), the correction of the photoresist pattern (S58) and the storage of the correction result of the photoresist pattern (S64) are carried out. Then, etching (S59) and photoresist stripping (S60) are performed.

After that, it proceeds to the judgment of the correction number management standard (S61). Here, foreign substance removal coordinate information and resist pattern correction results accumulated since the inspection has been performed are used. If the number of foreign objects corrected based on the foreign object removal coordinate information does not satisfy the management standard, it means that there are many foreign objects during film formation, and a process management warning is issued (S65), prompting inspection and countermeasures in the film forming apparatus.

In addition, when the result of storing (S64) the correction result of the photoresist pattern does not satisfy the correction number management standard, it means that a fault has occurred between the photoresist coating and sintering (S53) and the development (S55). In the same manner as before, a process management warning (S65) is issued to prompt inspection and countermeasures in corresponding devices.

Of course, when the correction number satisfies the control standard, the following film formation is performed on the TFT substrate (S62) to form a TFT substrate. As a result, the process can be constantly monitored, not only for corrections, but also for timely maintenance of equipment.

[Example 5]

Next, Embodiment 5 of wiring correction of a liquid crystal display device will be described. In Embodiment 5, as in Embodiment 1, wiring correction of a liquid crystal display device is described as an example, but it is also applicable to correction of general patterns formed on a plane, and is not limited to a liquid crystal display device.

FIG. 29 is a diagram illustrating correction when there is a disconnection in the wiring pattern of the liquid crystal display device. In the formation process of the TFT array on the TFT substrate, especially in the formation process of electrodes and wirings, as shown in FIG. Therefore, for example, after the drain wiring 33 is formed, whether or not the leak wiring 33 is disconnected is checked by visual inspection or the like, and when a disconnection is found, correction is performed as necessary.

In addition, the gate wiring 31 may also be disconnected in the same way. As described below, it can be corrected in the same way as the disconnection correction of the drain wiring. However, since it is the initial stage of the TFT substrate manufacturing process, it is also possible to peel off all the wiring and electrode patterns. Remove and recreate.

FIG. 30 is a diagram showing the configuration of a preferred correcting apparatus for implementing the disconnection defect correcting method of Embodiment 5. FIG. This device is a device in which the material application mechanism 206 for disconnection correction is added to the structure of the automatic correction device capable of short circuit correction by laser shown in the first embodiment. In addition, reference numeral 201 denotes a correction integrated controller, 204 denotes a laser head drive shaft, 207 denotes a laser beam, and 208 denotes a defective portion.

In FIG. 30, there is shown a structure in which the material application mechanism 206 is arranged obliquely with respect to the optical axis of the optical system 202 of the pattern correction device, and the disconnection correction material is applied from this oblique direction. Fig. 31 is an enlarged view of a material application mechanism illustrating a state of application of a material for disconnection correction. According to Embodiment 5, the coating position and material coating state can be checked in real time using the observation image of the optical system 202 of the pattern correction device, and can be controlled by the coating mechanism control device 203 .

For example, as shown in FIG. 31 , when the material application mechanism 206 is contacted with an electronic circuit substrate 210 (for example, a TFT substrate) for application, it is necessary to detect the contact state of the material application mechanism 206 with the substrate 210 so as not to Excessive contact with the substrate 210 can damage the substrate 210 or damage the material application mechanism 206 . The coating material can be supplied in a preferred contact state by monitoring with the pattern correction device optical system 202 .

Next, when a part of the wiring of the TFT substrate is missing, that is, when a disconnected state is taken as an example, the procedure for correcting the disconnection defect will be described in detail. Here, the case of correcting the disconnection defect 216 shown in FIG. 29 will be described as an example. The TFT substrate 210 having a disconnection defect 216 detected by an inspection device (not shown) is transferred to a correction device by a transfer robot (not shown) or the like, and placed on a stage 209 .

On the other hand, the network 205 of the production line receives defect position information detected by the inspection device, and drives the stage 209 based on the information to reproduce the disconnection defect position 216 within the field of view of the optical system of the correction device.

Thereafter, the entire optical system is moved in the Z direction perpendicular to the surface of the TFT substrate 210 on which the stage 209 is installed by an autofocus mechanism (not shown) to bring the focal point into focus on the surface of the TFT substrate 210 . The substrate 210 may also be moved in the Z direction by the substrate stage 209 . When the optical system 202 is moved, the optical axis of the laser optical system can be kept constant by also moving the laser oscillator and the irradiation optical system integrally.

Here, it is determined whether or not it is a correctable disconnection defect 216 from an image captured by a CCD camera installed in the laser optical system. When it is judged that the disconnection defect 216 can be corrected, the source electrode 33 is coated with the coating material (liquid) 243 .

When the disconnection defect 216 is generated due to foreign matter and the foreign matter remains, the disconnection defect 216 is corrected after the foreign matter is removed by the pulse laser of the correction device. In addition, if necessary, the oxide film of the wiring connected through the coating material 243 can be removed by laser irradiation or the like to reduce the connection resistance.

The front end of the material application mechanism 206 is placed in the storage container so that it will not be solidified by the application material 243 . This is to keep the front end of the material application mechanism 206 in a constant state. Move the position of the disconnection defect 216 to approximately the center of the field of view of the pattern correction device optical system 202 and move the front end of the material application mechanism 206 to reach the center of the field of view. The material application mechanism 206 has the function of allowing only this part to move slightly, and can automatically move to the position of the disconnection defect 216 based on image recognition.

When the material application mechanism 206 is slowly lowered from this state, the front end portion comes into contact with the surface of the drain wiring 211 . When it is further lowered, the tip portion of the material application mechanism 206 is bent by the elastic force, and deviates in the direction of the tip within the observation field of view. By observing this deviation, it can be confirmed that the front end portion of the material application mechanism 206 is in contact with the drain wiring 211 . By continuously monitoring a certain amount of deviation, the amount of descent of the material application mechanism 206 can be stabilized.

In addition, when the amount of offset is too large, the source electrode 33 may be damaged due to a force applied to the source electrode 33 , so the amount of offset is set to a value of several μm. After the contact is confirmed, the material is supplied by the material application mechanism 206 .

Fig. 32 is an explanatory diagram of a material application mechanism. FIG. 33 is an explanatory diagram of a state where a correction material is applied to a portion of the disconnection defect 216 in FIG. 29 by a material application mechanism. The material application mechanism 206 is a straw filled with a metal complex as a raw material of the metal film, for example, a glass straw made of a glass material.

As shown in FIG. 32( a ), the material application mechanism 206 has a structure in which the suction tube is filled with a liquid application material 243 . The material 243 in the straw is extruded with a mechanical tool 238 in the manner shown in FIG. Coating material 243 (FIG. 33, FIG. 31). When coating the coating material 243, as shown in FIG. 31, in order to sufficiently secure the connection portion, the coating material 243 is also supplied to the upper surface of the normal portion to reduce the wiring contact resistance and sufficiently perform wiring connection.

Fig. 34 is an explanatory diagram of a correction method for a disconnection defect. After the coating material 243 is supplied, as shown in FIG. 34( a ), shaping of the coating shape is performed. In this molding, as shown in Example 1, shaping processing of the shape using a mask is performed. Generally, since the metal film is processed by thermal processing, it must have strong processing energy, and the underlying layer may be damaged depending on the processing conditions and the lamination state of the TFT substrate 210 . Therefore, when the coating material 243 is in the state of a metal complex, it is preferable to perform image formation by processing mainly by molecular dissociation by a photochemical reaction.

When the volume change during the formation of the metal film by annealing (heat treatment) is large, it is better if the volume change after annealing is small if the shaping treatment is performed after one treatment during annealing. The annealing treatment can be performed using an infrared lamp, a substrate heater, or laser irradiation.

In the laser irradiation, it is preferable to select a laser that absorbs in the coating material 243 . In order not to perform processing for removing the coating material 243 by laser irradiation, it is preferable to perform continuous heat treatment using continuous oscillation of the laser.

In addition, in the annealing process using such laser irradiation or the like, by supplying an inert gas to the disconnection correction part, oxidation during annealing and a change in the material of the coating material 243 before the formation of the metal film are suppressed, and highly reliable wiring can be performed. connect.

The electronic circuit pattern is shaped by mask processing (FIG. 34(a)). In mask processing, as shown in the examples, by using pulsed laser light with a pulse width of several nm, processing with little thermal influence can be performed. Thereafter, the metal film is deposited by annealing, and the disconnection correction is completed to obtain the corrected wiring 219 ( FIG. 34( b )). If necessary, a wiring shaping step may be added after the metal film is formed.

What has been described above is the correction of the disconnection defect 216 in the process of the source electrode 33 of the TFT substrate 210, but the same process can be used for pattern correction of defects in other TFT layers. In addition, it can also be applied to the pattern of the intermediate layer necessary for the production process that does not remain on the TFT substrate. Even in this case, when the material of the coating material 243 changes due to photochemical reaction, the coating material 243 and the material coating mechanism 206 are shielded from light, and a stable material supply can be performed.

Fig. 35 is a diagram illustrating the flow of correction work as Embodiment 5 of the present invention. Using the above-mentioned material application mechanism 206, a correction system as shown in FIG. 35 can be obtained. Various pattern defects of electronic circuit boards detected and classified by a defect detection device based on visual inspection etc. are corrected in the following procedure.

[Graph A]··Short circuit defect

In FIG. 35, the procedure of short defect correction is demonstrated. First, a mask conforming to the pattern of the TFT substrate 210 is selected (S1A). Processing conditions such as laser energy, wavelength, and number of beams are selected using the material and pattern (laminated structure) of the correction target layer (S2A). Multiple conditions can be applied in stages as needed. The pattern is overlaid on the electronic circuit board, and mask processing by laser is performed (S3A). When the correction area cannot be included in one laser irradiation area, it is divided into a plurality and corrected sequentially. The processing status is monitored in real time by using the images captured by the optical system 202 of the pattern correction device, and it is judged that the correction is completed or reprocessed (S4A).

[Graph B]··Foreign matter defect

When there is a foreign object, even if the location and size of the foreign object have no effect on the layer to be corrected, it may affect the next layered figure. Therefore, it is preferable to remove it in advance in terms of the production process of the TFT substrate 210 . The same is true for short-circuit correction, and the processing conditions are selected according to the type of foreign matter (judged by color and shape), the position of foreign matter, and the laminated structure (S1B).

In this case, mask correction can be performed in the same way as short circuit correction, but in order to effectively use laser energy, it is effective to shrink the laser light into a rectangle or a circle to process the foreign matter (S2B). The processing is monitored by the observation optical system, and it is judged whether the correction is completed (S3B). If there is a foreign object in the circuit pattern and the removal processing is performed, it may affect the pattern. In this case, the disconnection correction described below is performed.

[Graph C]··Disconnection defect

The procedure for correcting the disconnection defect will be described below. First, the material application mechanism (S1C) is performed. The coating state is monitored by an observation optical system, and it is judged whether or not the coating material is full of disconnection defects (S2C). (Determine whether a disconnection defect is found in the coating material (S2C).) When the volume of the material fluctuates greatly, in order to improve the connection reliability and make the material more stable, a heat treatment is temporarily performed by annealing (S2C).

Mask correction is carried out in the above-mentioned step of [Pattern A], and it is shaped to conform to the shape of the electronic circuit pattern (S4C). After shaping, a metal film is deposited in annealing (S5C). The correction state is monitored in real time by the correction device optical system 202, and it is judged that the correction is completed (S6C). The TFT substrate 210 corrected by the above correcting steps is transferred to the next process.

[Example 6]

In the non-contact material application mechanism 206 shown in FIG. 32 , since the mask correction method described in Embodiment 1 can be used to shape the pattern after material application, it is not necessary to carry out communication with the source as in the past. The material application of the width of the electrode 33 coincides with high precision. A material is applied to a wide area including a defect portion by a non-contact material application mechanism 206, and a pattern is formed by mask processing. That is, unnecessary parts can be removed by laser processing, and the coating material 243 remains on the wiring with the disconnection defect 216, and metal wiring can be formed by annealing this.

In this way, since the material application mechanism 206 can be applied without contacting the substrate 210, and the control amount of the material application mechanism 206, that is, the control time can be shortened, the correction processing time can be shortened. At this time, the position accuracy of the material coating mechanism 206 can be obtained by implementing the position correction of the laser correction optical system 202 and the material coating mechanism 206 in advance, so that the coating material 243 is on the broken line defect part 216 and the coating material 243 is in the same position as the normal one. The positional accuracy of the degree of partial overlap coating.

The application region does not necessarily have to be within the wiring width of the source electrode 33 because of the shaping process by mask processing. The disconnection defect portion 216 may be covered. In FIG. 32( b ), the material application mechanism 206 is arranged in the vertical direction relative to the substrate, and the application material 243 is applied by injection. However, as described in FIG. 31 , it may also be performed from an oblique direction. The coating state is monitored using the observation image of the optical system 202 of the pattern correction device, and the injection amount and position are controlled. In this way, compared with the contact type material application mechanism 206, the processing speed of the disconnection defect correction can be increased.

[Example 7]

Fig. 36 is a diagram illustrating an inspection and correction system using a laser electronic circuit pattern correction device having a mask exchange function. Here, a description will be given of a processing system that performs inspection and correction by continuous transport without stopping the substrate 210 . In the inspection process 250 , the substrate 210 is photographed by the imaging element 225 , such as the line sensor 227 having the CCD camera 14 element, etc., and the defect detection image processing is performed by the image processing device 225 to make the defect visible.

In order to be able to inspect large substrates, a plurality of line sensors are arranged in one axis direction, and a plurality of rows are arranged in a staggered manner. Illumination, by applying downlight (coaxial illumination with the imaging element), oblique illumination 228, and transmissive illumination 229 alone or in combination, through illumination switching, not only can make it easier to visualize defects, but also can detect defects Types are classified. In addition, if necessary, by polarized light illumination and polarized light detection, it is possible to easily visualize thin films made of organic substances with variable polarization characteristics and foreign objects. The detected image is sent to the image processing device 225, where necessary image processing is performed to detect defects.

Disconnection, short circuit, and foreign object defect type and coordinate information detected by these inspection steps 250 are sent to the inspection and correction management server 224 via the network 205 of the production line. Here, the defect that can become fatal in production, that is, wiring short circuit and disconnection defect, foreign matter that becomes a problem in the process, etc. are extracted, and the determination and correction method of the defect to be corrected is determined.

These data are sent to the correction control PC 226 of the correction process 251 . In addition, when the number of defects exceeds a certain value and it is estimated that there is a problem in the production process, the production management PC (not shown) 230 sends a message through the network 205 and issues a warning to implement process countermeasures.

The correction process 251 is composed of a foreign matter removal head (laser optical system) 230 , a material application head (material application mechanism) 206 , and a short defect correction and image forming head (laser optical system) 232 as three correction heads. In order to improve the processing efficiency in continuous transmission, each correction head performs correction independently. They are moved in a parallel direction with respect to the substrate conveying direction to correct defect locations. The substrate is constantly being conveyed, and it also moves in the conveying direction in order to move, but there is a mechanism that returns to the origin when it moves a certain amount in the conveying direction as long as the amount of movement can keep up with the conveying speed of the substrate.

When the defect position flows when returning to the origin, it is tracked so that the correction head moves in the conveying direction. The inspection and correction management server 224 specifies the coordinates from which defect to perform correction, and the correction procedure of the target defect, in order to ensure correctness. The foreign matter removal laser optical system 230 basically functions as foreign matter removal, but it can also be used for short circuit correction if necessary.

The inspection step 250 performs processing within a certain period of time, and the processing time of the correction step 251 depends on the number of defects to be corrected. Therefore, when the correction process 251 requires processing time, the conveyance of the substrate 210 may be stagnated. Therefore, during the transfer of the substrate 210 , an interval of about half a substrate to one substrate is vacated for transfer, and the processing time difference of the correction step 251 is adjusted by this transfer interval.

That is, a speed control function 233 for realizing different conveying speeds is provided in the checking step 250 and the correcting step 251 . It is also possible to increase the processing speed by installing multiple correction heads according to the number of correction objects. In addition, reference numeral 209 denotes a stage.

Fig. 37 is a diagram illustrating another inspection and correction system using a laser electronic circuit pattern correction device having a mask exchange function. The same reference numerals as in FIG. 36 correspond to the same functional parts. This configuration shows an example in which the transport system 209 of the correction process 251 has two branches. As in FIG. 36 , each correction head moves in a direction perpendicular to the transport direction, and moves to the defect position coordinates for correction.

The transmission system is a two-branch system, but the correction head is a unit that moves across the two transmission systems for correction. For example, due to the time for short-circuit correction processing for a defect, the foreign matter removal head 230 and the material application head 206 are idle, so that the foreign matter removal (short correction can also be performed) and the material of the disconnected part of the other substrate can be performed using these idle heads. Apply.

In addition, the speed control of the conveying systems 209 can be carried out independently. In the correction process, since there are two conveying systems, the conveying adjustment can also be performed, and the processing speed of the inspection process 250 can be kept constant. These can utilize the speed control function 233 to continuously monitor and control the transmission speed, and the inspection and correction management server 224 determines the correction method and the transmission method of the correction process 251 .

The system described above is a system that performs inspection and correction steps using continuous conveyance, and inspection and correction can be performed even with step conveyance. At this time, when the line sensor is not used in the inspection camera, imaging is performed using the area sensor. Also, when a line sensor is used as in continuous transfer, moving images are shot during step transfer. In the correction process, correction is performed when the step is stopped. In this case, the inspection row and the transfer row move at different steps and speeds. In addition, even if the inspection process is continuous conveyance and the correction process is step conveyance, the above-mentioned inspection and correction can be performed.

The above description is for the case where the stage 209 on which the electronic circuit board 210 is placed is horizontal with respect to the device installation surface, but when the size of the electronic circuit board 210 is increased, the installation area of the device is enlarged. Therefore, it may also be as shown in FIG. 38 . That is, FIG. 38 is a diagram illustrating another arrangement form of the electronic circuit board. The correction unit 230 on which the foreign matter removal head and the material application head are placed is installed on the moving stage 235 and can correct any point on the substrate 210 .

By arranging the transmission form of the inspection and correction system described so far in a vertical or near-vertical orientation as shown in FIG. 38, the installation area of the device can be reduced. At this time, by configuring the optical system in the inspection and correction process also in the normal direction of the substrate 210, inspection and correction can be performed similarly to horizontal transfer.

[Example 8]

As an eighth embodiment of the present invention, a method for manufacturing a TFT substrate will be described as an example.

The manufacturing process of the TFT substrate includes the process of forming a gate electrode on a glass substrate (gate electrode forming process), the process of forming a gate insulating film on a glass substrate on which a gate electrode is formed (gate insulating film forming process), forming amorphous silicon, etc. The process of TFT active layer (island portion) (island portion forming process), the process of forming drain-source electrode (drain-source forming process), and the process of forming a protective film (protective film forming process).

Here, at least the gate electrode forming step, the island portion forming step, and the drain-source electrode forming step (these steps are referred to as a circuit pattern forming step), in the formation of the circuit pattern (gate electrode, island portion, drain-source electrode) Using photoresist pattern etching.

In Embodiment 8, in each of these circuit pattern forming steps, before etching, the photoresist pattern is inspected, and by correcting the photoresist pattern according to the inspection result, the possible defects in the circuit pattern are eliminated. It is corrected before the circuit pattern is formed.

Fig. 39 is a process chart for explaining a circuit pattern forming process according to Embodiment 8 of the present invention. Herein, the case where the circuit pattern is used as the drain-source electrode is taken as an example for description.

First, a drain-source film is formed on a glass substrate (intermediate product) on which a gate electrode, a gate insulating film, and an island portion are formed by a film forming step (S3901). Afterwards, a photoresist solution is applied and fired in a photoresist film coating and sintering process to form a photoresist film on the drain source film (S3902). After that, the exposure process described later is carried out, and the correction history (S3903) and development (S3904) of the photoresist pattern for the circuit pattern (gate electrode, island) of the lower layer are reflected in the mask projection pattern, and the drain source film A photoresist pattern is formed on it.

Thereafter, an appearance inspection of the photoresist pattern is performed to detect defects such as short-circuit defects, disconnection defects, and foreign matter contamination (S3905). Among them, in the visual inspection apparatus of the photoresist pattern, for example, an existing pattern matching technique can be used. That is to say, after the removable foreign matter on the photoresist pattern is blown away by blowing etc., the photographed image of the photoresist pattern and the normal image of the photoresist pattern prepared in advance are compared to detect the inconsistency between the two .

After that, the detected inconsistent portion is compared with a defect image of a photoresist pattern prepared in advance for each defect type. Then, the kind of defect corresponding to the most similar defect image is detected as a defect caused by an inconsistent portion.

In S3905, if there is no abnormality in the visual inspection of the photoresist pattern, etching (S3907) and stripping of the photoresist pattern (S3908) are performed to form drain-source electrodes.

On the other hand, in S3905, if there is an abnormality in the visual inspection of the resist pattern, a resist pattern correction step described later is performed to correct the resist pattern for the drain-source electrodes and generate The revision history of the photoresist pattern of the drain-source electrode (S3906). Thereafter, drain-source electrodes are formed by performing etching (S3907) and photoresist pattern stripping (S3908).

FIG. 40 is a process diagram for explaining the resist pattern correction step (S3909) shown in FIG. 39. FIG.

First, the type of defect detected in the resist pattern inspection step (S3905) shown in FIG. 39 is confirmed (S4001). When the defect type is a short-circuit defect, it proceeds to the short-circuit correction process of S4002 to S4004, and to proceed to the disconnection defect correction process of S4005 to S4007 when it is an open defect, and to proceed to S4008 to S4011 when foreign matter is mixed. foreign matter mixed into the defect correction process.

(1) Short-circuit defect correction process

In the photoresist pattern inspection process (S3905) shown in FIG. 39, when the photoresist patterns 4100 and 4101 shown in FIG. 41(B), a processing mask 4102 for forming an opening 4103 for irradiating only the portion A with laser light as shown in FIG. 41(B) may be selected (S4002).

By performing laser irradiation through a mask 5 having a laser transmission pattern 22 having the same shape as the wiring pattern of the short circuit 21 shown in FIG. 3, the short circuit 21 is removed to separate the pixels from each other. After that, processing conditions such as light intensity, wavelength, and number of beams (pulses) of the laser are selected according to the material and film thickness of the photoresist film to be corrected (S4003).

Afterwards, as shown in FIG. 41(C), while the selected processing mask 4102 is overlapped with the portion A, laser light is irradiated to the short-circuit defect position through the processing mask 4102 according to the selected processing conditions. At this time, the part A is photographed by the photographing device, and the spectral waveform (for example, RGB intensity) analysis is performed on the image of the photographed part A to study the prioritization characteristics of the part A, and the thickness of the remaining film of the part A is measured from the spectral characteristics, according to Feedback control may be performed by changing the number of processing conditions (beams) of the laser light as a result of the measurement.

For example, it is also possible to measure the change in the film thickness for each beam, and use the measurement result to calculate the number of beams necessary to remove the residual film to change the processing conditions. Alternatively, the thickness distribution of the residual film in part A may be measured, and the light intensity of a part of part A thicker (smaller) than other parts may be increased (decreased) to change the processing conditions. As a result, as shown in FIG. 41(D), the portion A is removed to separate the resist patterns 4100, 4101 (S4004).

(2) Disconnect defect correction process

In the photoresist pattern inspection process (S3905) shown in FIG. 39, a disconnection defect formed due to the separation of parts B of the photoresist patterns 4200 and 4201 that should be connected as shown in FIG. 42(A) is detected. 42(B), partial recoating and sintering of the photoresist solution 4202 may be performed on the portion B as shown in FIG. 42(B) (S4005).

At this time, part B is photographed by the photographing device, and the spectral characteristics of part B are studied by analyzing the spectral waveform of the photographed image of part B, and the film thickness of the photoresist solution reapplied to part B is measured from the spectral characteristics. Or hardness, and feedback control may be performed by changing coating conditions such as the number of injections of photoresist solution and sintering conditions such as heating temperature based on the measurement results.

For example, the change in film thickness per shot may be measured, and the processing conditions may be changed by calculating the number of shots necessary to form a predetermined photoresist film using the measurement results. Alternatively, the hardness distribution of the photoresist film formed on the portion B may be measured, and the processing conditions may be changed by increasing (decreasing) the heating temperature of portions of the portion B that are less (larger) hardened than other portions.

In addition, before S4005, the portion B may be irradiated with laser light to form fine recesses or roughness. In this way, the photoresist solution can be self-aligned to the coating position.

After that, the reforming mask 4203 for reforming the resist pattern on the portion B as shown in FIG. 42(C) is selected (S4006). In addition, according to the material and film thickness of the photoresist film to be recoated, the processing conditions such as the wavelength of the laser beam and the number of beams are selected. Afterwards, as shown in FIG. 42(D), while the selected reshaping mask 4203 is aligned with the position of the portion B, the portion B is irradiated through the partial exposure mask 4203 according to the selected processing conditions. Laser shaping (S4007).

As a result, as shown in FIG. 41(E), a photoresist film is formed on the portion B to connect the photoresist patterns 4200, 4201 (S4007).

In addition, the application of the photoresist solution is carried out with approximate accuracy, and as described above, shape shaping can also be performed by irradiating laser light through a mask pattern having the same shape as the circuit pattern. That is, by irradiating laser light through the mask 5 having the same laser transmission pattern 22 as the wiring pattern of the short circuit 21 in FIG. 3 , the short circuit 21 can be removed to separate the pixel electrodes 34 of each pixel.

(3) Foreign matter mixed defect correction process

In the photoresist pattern inspection process (S3905) shown in FIG. 39, when the foreign substance inclusion defect in the part C of the photoresist pattern 4300 shown in FIG. 43(B), partial recoating and sintering of the photoresist solution 4203 can be performed on the region D including the part C and the part where the serpentine pattern is formed (S4008).

At this time, the region D is photographed by the imaging device, the spectral characteristics of the region D are studied by analyzing the spectral waveform of the photographed image of the region D, and the film thickness of the photoresist solution reapplied to the region D is measured from the spectral characteristics. Or hardness, and feedback control may be performed by changing coating conditions such as the number of injections of photoresist solution and sintering conditions such as heating temperature based on the measurement results.

For example, the change in film thickness per shot may be measured, and the processing conditions may be changed by calculating the number of shots necessary to form a predetermined photoresist film using the measurement results. Alternatively, the hardness distribution of the photoresist film formed on the region D may be measured, and the processing conditions may be changed by increasing (decreasing) the heating temperature of the portion in the region D that is less (higher) hardened than the other portions.

After that, a detour mask 4302 for detouring the portion C to form a resist pattern as shown in FIG. 43(C) is selected (S4009). In addition, processing conditions such as light intensity and wavelength of laser light are selected according to the material and film thickness of the photoresist film to be recoated. Afterwards, as shown in FIG. 43(D), while the selected detour mask 4202 is aligned with the region D, the region D is irradiated with laser light through the detour mask 4202 according to the selected processing conditions. Perform shaping (S4010). As a result, a detour pattern 4304 is formed in the area E as shown in FIG. 43(E).

In addition, before S4009, the portion C may be irradiated with laser light to remove the part of the photoresist pattern in which the foreign matter is mixed. In this case, when the etching layer (underlayer) does not cause film peeling, etc., the same treatment (S4005-S4007) as in the case of the (2) disconnection defect correction step can be performed to form a photoresist again on the part C. graphics.

FIG. 44 is a process diagram for explaining the circuit pattern forming step (S3903) shown in FIG. 39. FIG.

First, it is examined whether or not S4008 to 4010 in FIG. 40 (foreign matter entry defect correction process) is performed on the lower circuit pattern layer. For example, when the circuit pattern to be formed is the drain-source electrode, whether or not to perform S4008 to S4010 in FIG. 40 is examined for the island layer or the gate electrode layer as the lower circuit pattern layer (S4401).

When it is judged not to be executed in S4401, the normal circuit pattern process is passed (S4402). On the other hand, when it is judged to be executed in S4401, it is necessary to form a detour circuit. In the exposure step for forming a circuit pattern, the correction performed in S4008 to S4010 of FIG. 40 is not used in order to form a normal circuit (state where no detour circuit is formed) pattern.

Then, apply photoresist to this coordinate part (Use photoresist to planarize the normal circuit pattern for reformation) (S4403), and use laser processing to form a detour circuit overlapping with the new lower layer pattern (S4404).

In addition, in the maskless process using a DMD, a liquid crystal display device, etc. in the exposure process, the following method is effective. That is, when it is determined not to execute in S4401, exposure is performed using a standard exposure mask prepared for the circuit pattern to be formed. On the other hand, when it is judged to be executed in S4401, the exposure mask is partially changed to overlap with the detour mask used in the lower-level circuit patterning device to perform exposure.

For example, in the island-shaped portion layer (TFT active layer) formation process, as a result of performing S4008 to S010 in FIG. 40, as shown in FIG. 45(A), the island-shaped portion 4501 is at a distance from the original formation position E When forming at the offset position F of H, in the formation process of the drain source electrode layer, perform S4403 in FIG. It is formed by extending a distance H from the original electrode end 4504, and the exposure mask for the drain electrode 4502 and the source electrode 4503 is partially changed. Then, exposure is performed using this exposure mask.

Next, the resist pattern inspection and correction system used in the resist pattern inspection step (S3905) and the resist pattern correction step (S3906) shown in FIG. 39 will be described.

46 is a diagram showing an example of a resist pattern inspection and correction system used in the resist pattern inspection step and the resist pattern correction step of the eighth embodiment.

As shown in the figure, the photoresist pattern inspection and correction system has a photoresist pattern inspection system that inspects the photoresist pattern formed on the semi-finished product (TFT substrate with the photoresist pattern formed on the uppermost layer) 485 flowing over the TFT substrate on the stage 480. The photoresist pattern inspection device 460, according to the inspection results in the photoresist pattern inspection device 460, corrects the photoresist pattern defects of the photoresist pattern formed on the semi-finished product 485 of the TFT substrate flowing on the stage 480 The correction device 470 and a network 490 such as a LAN connected to the resist pattern inspection device 460 and the resist pattern correction device 470 .

The photoresist pattern inspection device 460 has a line sensor 461 and a defect detection device 462 . The line sensor 461 is formed by arranging a plurality of image pickup elements (such as CCD cameras), and can take pictures of at least one row-sized semi-finished TFT elements included in the semi-finished TFT substrate 485 .

The defect detection device 462 is a computer that detects defects of the photoresist pattern formed on the semi-finished product 485 of the TFT substrate. In the defect inspection device 462, for the TFT element produced on the TFT substrate, the photoresist pattern for the gate electrode, the photoresist pattern for the gate insulating film, the photoresist pattern for the island portion, and the source and drain patterns are registered in advance. The electrodes are patterned with photoresist for the respective normal images.

In addition, for each of the photoresist pattern for the gate electrode, the photoresist pattern for the gate insulating film, the photoresist pattern for the island portion, and the photoresist pattern for the drain-source electrode, the defect Each type (short circuit, disconnection, foreign matter, etc.) is registered in advance.

The defect detection device 462 compares the photoresist pattern image of the semi-finished product of each TFT element captured by the line sensor 461 with the normal image of the photoresist pattern, and detects the inconsistency between the two by pattern matching technology. Then, the detected inconsistency is compared with the defect image of the photoresist pattern registered in advance for each defect type, and the defect type corresponding to the most similar defect image is detected as a defect generated in the inconsistency. Afterwards, defect information including defect type, coordinate information of defect occurrence location, and captured image of defect occurrence location is generated and sent to photoresist pattern correction device 470 via network 490 .

In addition, the defect detection device 462 may compare each pixel of the photoresist pattern image of the semi-finished product of each TFT element captured by the line sensor 461 with adjacent pixels (normal parts) to determine the presence of a defect. In addition, it is also possible to determine the type of defect using feature quantities (shape, color, film thickness, etc.) extracted from the resist pattern image. In addition, it is also possible to identify the presence or absence of defects and types by coordinating with the above-mentioned registered images.

The photoresist pattern correction device 470 has a coating and sintering mechanism 471 for partially coating and sintering the photoresist solution; the position of the defect part of the semi-finished product 485 of the mask and the TFT substrate is overlapped, and the photoresist pattern is corrected by laser processing. A photographing laser optical system 472 for correcting defective parts; and a control device 473 . The control device 473 is a computer that controls the coating and sintering mechanism 471 and the imaging laser optical system 472 .

FIG. 47 is a schematic configuration diagram of a resist pattern correcting device 470 .

The coating and firing mechanism 471 (see FIG. 46 ) has a coating mechanism and a firing mechanism. The sintering mechanism has a heater 4715 provided with a heating surface that can change heat locally. The application mechanism has a dispenser 4711 that holds the photoresist solution, and a drive device 4712 that drives the dispenser 4711 to discharge the photoresist solution from the dispenser 4711 . FIG. 48 is a schematic cross-sectional view of a dispenser 4711 .

As shown in the figure, the distributor 4711 has a holding part 47111 forming a discharge port; a piston part 47112; and an assist gas guide part 47113. A photoresist solution 47114 is held in the holding portion 47111 . The piston part 47112 is driven by the driving device 4712, and pushes out the photoresist solution 47114 held in the holding part 47111 from the ejection port. In addition, an inert gas 47115 for stabilizing the material of the resist liquid 47114 such as nitrogen gas may be filled between the resist liquid 47114 held in the holding portion 47111 and the holding portion 47111 .

The guide part 47113 is formed around the holding part 47111, and ejects inert gas such as nitrogen gas supplied from the driving device 4712 toward the direction in which the photoresist solution 47114 held in the holding part 47111 is pushed out from around the discharge port of the holding part 47111. . By providing the guide portion 47113, even when the dispenser 4711 is disposed at a position separated from the defective portion of the TFT substrate semi-finished product 485, the photoresist solution 47114 can be applied (spread) to the defective portion.

In addition, the dispenser 4711, as shown in FIG. As shown, it may be arranged at a position that does not overlap with the optical axis L of the imaging optical system and the laser optical system.

The imaging laser optical system 472 (see FIG. 46 ) has a laser optical system and an imaging optical system. Here, the laser optical system has the same optical axis L as the photographing optical system.

The laser optical system includes a laser oscillator 4721 ; a beam shaping mechanism 4722 including a beam expander, a homogenizer, etc.; a mask mechanism 4723 including a mask stage; an imaging lens 4724 ; and an objective lens 4725 . The laser beam irradiated from the laser oscillator 4721 is expanded to a predetermined beam diameter by the beam shaping mechanism 4722, and is shaped so that the laser intensity distribution in the laser irradiation area becomes a predetermined distribution.

Afterwards, the shaped laser is shaped into a mask projection pattern corresponding to the mask set on the mask mechanism 4741, and then irradiates the semi-finished product 485 of the TFT substrate set on the stage 480 through the imaging lens 4724 and the objective lens 4725. on the defective part.

The shooting optical system has: a shooting device (such as a CCD camera) 4731 provided with an AF (autofocus) mechanism; a lens 4732; a semi-transparent mirror 4733; Lens 4735; semi-transparent mirror 4736; illuminating device for mask shooting (for example, optical fiber for light source) 4737; condenser lens 4738; semi-transparent mirror 4739; illuminating device for transmitted illumination (for example, optical fiber for light source) 4740; 4741.

The imaging device 4731 captures the image of the mask installed in the mask mechanism 4723 through the semi-transparent mirror 4733 and the lens 4732 . At this time, the brightness of the image of the mask is adjusted by irradiating the light of the illuminating device 4737 for mask imaging onto the mask provided in the mask mechanism 4723 through the condenser lens 4738 and the half mirror 4739 . In addition, the imaging device 4731 captures an image of a half 485 of the TFT substrate placed on the stage 480 via a half mirror 4733 and a lens 4732 .

At this time, the brightness of the image of the semi-finished product 485 is adjusted by irradiating the light of the illuminating device 4734 for photographing the photoresist pattern to the semi-finished product 485 of the TFT substrate placed on the stage 480 through the condenser lens 4735 and the semi-transparent mirror 4736. The brightness of the image of the semi-finished product 485 is adjusted by irradiating the light of the illuminating device 4740 for transmission illumination from the back side through the condenser lens 4741 to the semi-finished product 485 of the TFT substrate placed on the stage 480 .

The control device 473 controls each part of the photoresist pattern correction device 470 through the network 490 according to the defect information sent from the defect detection device 462, and performs partial coating and sintering of the photoresist solution and defect correction of the photoresist pattern.

(1) Coating and sintering of local photoresist solution

The control device 471, when the defect type contained in the defect information received from the defect detection device 462 is a disconnection defect or a foreign matter incorporation defect, performs a correction to the photoresist pattern correction device 470 for the application and sintering of the local photoresist solution. control of each department. First, the stage 480 is controlled, and the semi-finished product 485 of the TFT substrate placed on the stage 480 is moved so that the position indicated by the coordinate information of the defect occurrence site included in the defect information and the coating of the photoresist solution generated by the dispenser 4711 are aligned. The placement is the same.

Afterwards, the driving device 4712 is controlled to spray the photoresist solution from the dispenser 4711, and the photoresist solution is coated to cover the defective part of the photoresist pattern of the semi-finished product. At this time, the spectral waveform analysis is performed on the image of the photoresist pattern defect part captured by the imaging device 4731, and the spectral characteristics of the defect part are studied. The relational information measures the film thickness of the photoresist solution reapplied on the defective portion, and performs feedback control based on the measurement result to change the application conditions such as the number of injections of the photoresist solution.

For example, the change in film thickness per shot is measured, and the driving device 4712 is controlled using the measurement result to discharge the resist liquid from the dispenser 4711 in the number of shots required to form a predetermined resist film. In addition, the film thickness of the photoresist solution can also be measured using a change in the focus position of the AF mechanism included in the imaging device 4731 . In addition, it is also possible to provide a means for pushing out the photoresist solution to be reapplied (press spoon, etc.), and to adjust the film thickness of the photoresist solution by using the pushing out mechanism.

Afterwards, the heater 4715 is controlled to sinter the photoresist solution coated on the photoresist defective part of the semi-finished product 485 of the TFT substrate. At this time, the spectroscopic waveform analysis is performed on the image of the recoated part of the photoresist solution captured by the imaging device 4731 to study the spectral characteristics of the recoated part. Measure the degree of curing of the photoresist solution in the recoated part, and perform feedback control to change the sintering conditions of the photoresist solution according to the measurement result.

For example, the hardness distribution of the photoresist film in the recoated part is measured, and the heater 4715 is controlled so that the heating temperature of the part with a lower (larger) degree of hardening than other parts in the recoated part is increased (decreased) and changed. Processing conditions.

(2) Defect correction of photoresist pattern

The control unit 471 controls each part of the resist pattern correcting unit 470 for correcting the defect portion of the resist pattern according to the defect information received from the defect detecting unit 462 . First, a photographed image of a photoresist pattern defect portion included in the defect information received from the defect detection device 462 is displayed on a display device (not shown), and the operator is asked to use a mask (for processing) to correct the defect portion. A mask, a reforming mask, and a detour mask) are set on the mask mechanism 4723 .

Afterwards, the mask placed on the mask mechanism 4723 and the semi-finished product 485 of the TFT substrate placed on the stage 480 are photographed by the imaging device 4731, for example, in order to make the fiducial mark set in the mask and the semi-finished product 485 specified The position coincides, and by controlling the stage 480, the half-finished product 485 of the TFT substrate set on the stage 480 is moved.

After that, the laser oscillator 4721 is controlled to shape the photoresist pattern of the semi-finished product 485 of the TFT substrate. At this time, the spectroscopic waveform analysis is performed on the image of the defect portion of the photoresist pattern captured by the imaging device 4731, and the spectral characteristics of the defect portion are studied. Measure the film thickness of the photoresist film on the defective portion, and perform feedback control to change the processing conditions of the laser according to the measurement result.

For example, the change in the film thickness for each beam is measured, and the laser oscillator 4721 is controlled using the measurement result so as to output laser pulses of the number of beams necessary for removing the residual film. Alternatively, the thickness distribution of the remaining film in the defective portion is measured, and the beam shaping mechanism 4722 is changed so that the laser intensity of the portion having a larger (smaller) thickness than other portions of the defective portion is increased (weakened).

In addition, the resist pattern correcting apparatus 470 may include a foreign matter removal mechanism 4750 for removing foreign matter mixed in the resist pattern by blowing, photolysis, thermal and chemical treatment decomposition, etc., as shown in FIG. 47 .

The eighth embodiment has been described above. In Embodiment 8, since the photoresist pattern is corrected, the influence of pattern defect correction on the pattern layer can be reduced. In addition, since the photoresist pattern is inspected and corrected during the formation process of each pattern layer of the TFT substrate, pattern defects of each pattern layer can be corrected. Therefore, it is possible to prevent the occurrence of defective circuit patterns formed on the TFT substrate without degrading the quality and precision of the circuit patterns.

In addition, in the eighth embodiment, the photoresist liquid is recoated on the defective part of the photoresist pattern, but the photoresist material may not be liquid. For example, a thin-film photoresist material 4901 shown in FIG. 49(A) can be used. Alternatively, a particulate photoresist material 4902 shown in FIG. 49(B) may also be used. In addition, in FIG. 49, reference numeral 4903 is a photoresist film, and reference numeral 4904 is an etching layer (underlayer).

While several embodiments according to the present invention have been shown and described above, it should be appreciated that changes and modifications may be made to the disclosed embodiments without departing from the scope of the invention. Therefore, the invention is not limited to the details shown and described above, but includes all changes and modifications covered by the appended claims.

Claims (8)

1. the manufacture method of a display device after substrate surface forms inorganics or organic film, becomes electrode or wiring with this film composition, and this manufacture method is characterised in that successively carries out:

Carry out successively on above-mentioned film photoresist coating, passed through to have the homology figure corresponding with above-mentioned electrode or wiring exposure mask this photoresist exposure and to the development of the photoresist that exposed, this photoresist is configured as first operation of the photoresist figure corresponding with this homology figure

Above-mentioned photoresist figure is carried out visual examination, extracts second operation of the positional information of the defective that is the shape different in this photoresist figure with this homology figure,

According to the positional information of above-mentioned defective with the above-mentioned defective of laser radiation, thereby remove the unwanted part of the above-mentioned photoresist figure in this defective, revising the 3rd operation of this defective, and

By the above-mentioned photoresist figure of having revised above-mentioned defective above-mentioned film is carried out etching, forms the 4th operation of above-mentioned electrode or wiring with composition,

In above-mentioned the 3rd operation, above-mentioned laser has formed this defective that is projected to this photoresist figure with the state of the identical shaped test pattern of the figure of the above-mentioned defective that is used to revise above-mentioned photoresist figure to utilize mask or catoptron.

2. the manufacture method of display device as claimed in claim 1 is characterized in that:

In above-mentioned the 3rd operation, use the aforementioned mask that is formed with the homology figure corresponding with above-mentioned electrode or wiring, make laser form above-mentioned test pattern by making above-mentioned laser by this homology figure.

3. the manufacture method of display device as claimed in claim 1 is characterized in that:

Before above-mentioned first operation, implement to be formed at the above-mentioned inorganics on the aforesaid substrate or the visual examination of organic film, and when removing the foreign matter of in this visual examination, finding, carry out the operation that this foreign matter in this film of storage is removed the information of coordinate,

In above-mentioned the 3rd operation, when removing above-mentioned unwanted part and this graphics shape be modified to normal shape from above-mentioned photoresist figure, the positional information of storing the figure correction coordinate of this photoresist,

When the lift-off processing of the above-mentioned etching in utilizing above-mentioned the 4th operation, the above-mentioned photoresist of following is processed into predetermined pattern with above-mentioned film, when the positional information of removing the information of coordinate and above-mentioned photoresist figure correction coordinate according to the above-mentioned foreign matter of storage is earlier managed the generation state of above-mentioned foreign matter, in surpassing of this foreign matter under the situation of certain management standard, give a warning the indication countermeasure.

4. the manufacture method of display device as claimed in claim 1 is characterized in that:

When the mode with the disappearance of the above-mentioned photoresist in the part of above-mentioned photoresist figure produces above-mentioned defective, in above-mentioned the 3rd operation, on the disappearance part of above-mentioned photoresist, apply this photoresist again, after burning till, above-mentioned defective is shone above-mentioned laser and carried out shaping via forming again with mask.

5. the manufacture method of display device as claimed in claim 1 is characterized in that:

When producing above-mentioned defective owing to the foreign matter that has adhered on the aforesaid substrate surface, in above-mentioned the 3rd operation, periphery in this defective applies above-mentioned photoresist again, after burning till with this defective of above-mentioned laser radiation, form the figure of circuitous this foreign matter on the photoresist that applied again.

6. the manufacture method of a display device will be formed at the inorganics of substrate surface or organic film is configured as the photoresist that forms on this film figure, and this manufacture method is characterised in that successively carries out:

Carry out the coating of the photoresist on above-mentioned film successively, based on the exposure of this photoresist of the laser of the exposure mask that has passed through to be formed with homology figure and to the development of the photoresist that exposed, this photoresist is configured as first operation of the photoresist figure corresponding with this homology figure

Above-mentioned photoresist figure is carried out visual examination, extracts second operation of the positional information of the defective that is the shape different in this photoresist figure with this homology figure,

According to the positional information of above-mentioned defective with the above-mentioned defective of laser radiation, thereby remove the 3rd operation of the unwanted part of the above-mentioned photoresist figure in this defective, and

By the above-mentioned photoresist figure of in above-mentioned the 3rd operation, having removed above-mentioned unwanted part above-mentioned film is carried out etching, with the 4th operation of this photoresist figure that is shaped,

In above-mentioned the 3rd operation, above-mentioned laser is projected on this defective of this photoresist figure with mask by being formed with the processing of the identical shaped processing of above-mentioned homology figure with homology figure.

7. the manufacture method of a display device, this display device is at the second graph layer that has stacked gradually first graph layer that is formed with island shape figure on the substrate surface and be formed with the circuitous pattern that extend in the outside above this from this island shape figure, and this manufacture method is characterised in that:

Form above-mentioned first graph layer and above-mentioned second graph layer respectively by carrying out following first operation, second operation and the 3rd operation successively, promptly

On inorganics that will be configured as above-mentioned island shape figure or foregoing circuit figure or organic film, apply photoresist, with the laser of the exposure mask that has passed through to be formed with homology figure this photoresist that exposes, and first operation of this photoresist that has exposed of developing with the photoresist figure corresponding that be shaped with this island shape figure or this circuitous pattern

Inspection is blended into second operation of the foreign matter of above-mentioned photoresist figure,

The etching that above-mentioned photoresist figure has been passed through in utilization is configured as the 3rd operation of above-mentioned island shape figure or foregoing circuit figure with above-mentioned film,

In above-mentioned second operation, when detecting the above-mentioned foreign matter of sneaking into above-mentioned photoresist figure and sneak into the zone, before above-mentioned the 3rd operation, sneak into the zone at this foreign matter of this photoresist figure and apply above-mentioned photoresist on every side again with it, the photoresist that has applied again with laser radiation then, with avoid this foreign matter sneak into the zone mode form this photoresist figure again, and

Use the photoresist figure formed again, carry out the etching of the above-mentioned film in the 3rd operation.

8. the manufacture method of display device as claimed in claim 7 is characterized in that:

In being used to form above-mentioned second operation of above-mentioned first graph layer, when detecting the above-mentioned foreign matter of sneaking into above-mentioned photoresist figure and sneak into the zone, sneak into the zone in order to avoid this foreign matter, from the offset determined by above-mentioned exposed mask the position form the above-mentioned island shape figure of this first graph layer

After above-mentioned first operation that forms above-mentioned second graph layer, on the zone above the above-mentioned island shape figure that arrives the position that is offset from the photoresist figure corresponding, apply above-mentioned photoresist again with the foregoing circuit figure, and the photoresist that has applied again with laser radiation, make this photoresist figure extend to this island shape figure above, then, in above-mentioned the 3rd operation that forms above-mentioned second graph layer, the top photoresist figure that utilization extends to above-mentioned island shape figure comes the above-mentioned film of etching, thereby forms the foregoing circuit figure.

Images