KR20070015093A - Unevenness inspecting apparatus and unevenness inspecting method - Google Patents

Abstract

An unevenness inspection apparatus and an unevenness inspection method are provided to inspect unevenness of a layer formed on a substrate. An unevenness inspection apparatus includes a maintaining part, a light radiating part(3), a photographing part(41), and an unevenness detection part(7). The maintaining part maintains a substrate on which a light transmissive layer is formed. The light radiating part radiates light to the layer from a predetermined incident angle. The photographing part receives a specific wavelength of interference light reflected from the layer among the light from the light radiating part to obtain an original image of the layer. The unevenness detection part compensates affection varied depending on the layer thickness using sensitivity as a ratio of variation in intensity of the specific wavelength of interference light.

Description

Non-uniformity Inspection Device and Non-Uniformity Inspection Method {UNEVENNESS INSPECTING APPARATUS AND UNEVENNESS INSPECTING METHOD}

1 is a diagram showing a change in the film thickness of the reflectance of a film.

2 is a diagram illustrating a change in sensitivity to a film thickness.

3 is a diagram for explaining the relationship between sensitivity and relative reflection intensity.

4 is a diagram illustrating a relationship between sensitivity and relative reflection intensity.

FIG. 5: is a figure which shows the structure of the nonuniformity inspection apparatus which concerns on 1st Embodiment.

6 is a diagram illustrating a wavelength switching mechanism.

7 is a diagram illustrating a light receiving surface of the imaging unit.

8 is a diagram showing a flow of a process for inspecting a film thickness nonuniformity of a film on a substrate.

9 is a diagram showing a relationship between sensitivity and film thickness in interference light of a plurality of wavelengths.

10 is a diagram showing a change in the background film thickness of relative detection intensity.

11 is a diagram showing a change in the background film thickness of relative detection intensity.

12 is a diagram showing a change in the background film thickness of relative detection intensity.

It is a figure which shows the structure of the nonuniformity inspection apparatus which concerns on 2nd Embodiment.

It is a figure which shows a part of the process flow which examines the film thickness nonuniformity of the film | membrane on a board | substrate.

TECHNICAL FIELD This invention relates to the technique of examining the film thickness nonuniformity of the film formed on the board | substrate.

Conventionally, when inspecting a thin film such as a resist film formed on the main surface of a glass substrate or a semiconductor substrate (hereinafter, simply referred to as a "substrate") for a display device, light from a light source is irradiated with a thin film, The nonuniformity of a film thickness is examined using the optical interference in the reflected light from a thin film. For example, in Japanese Patent Laid-Open No. 2003-194739, an interference image is acquired by an imaging unit by receiving light reflected from a film on a substrate of light irradiated from a light source, and performing image processing on the interference image. In the apparatus for inspecting the film thickness nonuniformity, by varying the wavelength of the light received by the imaging unit selectively, the desired difference in luminance between the portion where the film thickness on the substrate is uniform and the portion where the film thickness nonuniformity exists are desired. The technique to make is shown.

Further, in Patent No. 3335503, in the non-uniformity test of light transmittance of a shadow mask for a color CRT, the gray scale data of an image irradiated with light from one main surface side of the shadow mask and picked up from the main surface side of the other side is applied. By performing a smoothing process using a median filter, the smoothing data is obtained, and the image of the shadow mask which highlights the nonuniformity which should be detected based on the normalized data calculated by dividing the gradation data by the smoothing data is displayed. Techniques for simplifying and improving inspection accuracy have been shown.

By the way, when the image of a film is acquired by receiving the interference light of one wavelength reflected from the film on a board | substrate, and the inspection of a film thickness nonuniformity based on this image, the intensity of interference light will change periodically with respect to a film thickness. Therefore, the sensitivity, which is the ratio of the intensity variation of the interference light to the variation in the film thickness, also changes depending on the film thickness. Therefore, depending on the thickness of the film on the substrate, the degree of film thickness nonuniformity, or the like, the degree to which the film thickness nonuniformity is reflected as a shade in the image becomes uneven due to the effect of sensitivity change, so that the film thickness nonuniformity cannot be accurately inspected. have.

This invention is suitable for the nonuniformity test | inspection apparatus which inspects the film thickness nonuniformity of the film formed on the board | substrate, and an object is to test the film thickness nonuniformity of the film formed on the board | substrate with high precision.

The nonuniformity inspection apparatus according to the present invention includes a holding part for holding a substrate on which a light-transmissive film is formed on a main surface, a light irradiation part for irradiating light to a film at a predetermined angle of incidence, and a light reflected from the film among light in the light irradiation part. The image pickup unit for receiving the original image of the film by receiving the interference light of a specific wavelength, and the effect that the sensitivity, which is the ratio of the intensity variation of the interference light of the specific wavelength to the variation in the film thickness, changes depending on the film thickness. While correcting, a nonuniformity detecting portion for detecting the amplitude of a predetermined spatial frequency band in the original image or the image derived from the original image as a film thickness nonuniformity is provided. According to the present invention, the film thickness nonuniformity can be inspected with high accuracy by correcting the influence of sensitivity that changes depending on the film thickness.

In one preferred embodiment of the present invention, the non-uniformity inspection device further includes a storage unit for storing sensitivity information indicating a relationship between the sensitivity and the intensity of the interfering light of a specific wavelength, wherein the non-uniformity detection unit includes a value of each pixel of the original image. And by detecting the film thickness nonuniformity while correcting the influence of the change in sensitivity with reference to the sensitivity information, it is possible to easily correct the influence of the change in sensitivity.

In another preferred embodiment of the present invention, the imaging unit has a plurality of imaging elements, and the non-uniform detection unit is adapted to the change in the individual sensitivity according to the corresponding imaging element for each pixel of the original image or the image drawn from the original image. By correcting the influence caused by the influence, the influence caused by the difference in the output characteristics of the imaging device can also be corrected.

In another more preferable embodiment of the present invention, the non-uniformity inspection device further includes a storage unit for storing sensitivity information indicating a relationship between sensitivity and film thickness, wherein the non-uniformity detection unit refers to the sensitivity and the thickness information of the film on the main surface thereof. The film thickness nonuniformity is detected while correcting the influence of the change of. More preferably, the nonuniformity inspection device further includes a film thickness measurement section for acquiring the thickness of the film on the main surface.

In one aspect of the present invention, a nonuniformity detecting portion includes a filter processing portion for performing a band pass filter processing of a predetermined spatial frequency band on an original image, and a contrast emphasis portion for emphasizing the contrast of the original image after being processed by the filter processing portion. The correction for the influence of the change in sensitivity is performed by changing the degree of contrast enhancement in the contrast emphasis portion.

This invention is also suitable for the nonuniformity test | inspection method which examines the film thickness nonuniformity of the film formed on the board | substrate.

The above objects and other objects, features, aspects, and advantages will be apparent from the following detailed description of the invention made with reference to the accompanying drawings.

FIG. 1 shows a film in a case where light having a wavelength of one wavelength is irradiated onto a glass substrate (hereinafter, referred to as a "substrate") on which a resist film (hereinafter, simply referred to as "film") is formed. It is a figure which shows the change with respect to the film thickness of (absolute) reflectance. In Fig. 1, the reflectance of the film with respect to incident light of wavelengths (exactly, the half band width of about 10 nm in the extremely narrow wavelength band) 550, 600, and 650 nanometers (nm) is shown in Figs. , 113). Herein, the film on the substrate is present on the chromium layer formed on the substrate, and FIG. 1 can be obtained by calculating the incident angle of incident light at 60 degrees, the refractive index of the film at 1.6, and the refractive index of the chromium layer as (3.77 + 4.8i). have. In addition, since calculation of a reflectance is a well-known technique, description is abbreviate | omitted here.

As shown in Fig. 1, it can be seen that the reflectance of the film changes periodically with respect to the film thickness at either wavelength. This change in reflectance is that the reflected light of the film on the substrate is actually the reflected light on the surface (upper surface) of the film on the substrate of the incident light and the reflected light on the lower surface of the film on the substrate (that is, the interface between the film and the substrate body). This is due to interference light with Here, the reflectance of the film on the substrate is the ratio of the intensity of the interference light to the intensity of the incident light, and the vertical axis in FIG. 1 can be regarded as the intensity of the interference light with respect to the incident light of a constant intensity. The intensity of the interference light in the case of 1 (hereinafter referred to as "relative reflection intensity") will be described. In this case, the relative reflection intensity R _k can be expressed as a formula (1) by the predetermined function f _R as the film thickness d. However, as described, the refractive index of the film on the substrate, the refractive index of the layer below the film, and the incident angle of light are constant.

FIG. 2 is a diagram showing variation in relative reflection intensity when the film thickness fluctuates only 1 nm in FIG. 1. In FIG. 2, the ratio of the variation of the relative reflection intensity of the interference light to the variation of the film thickness of 1 nm is indicated by the vertical axis as the sensitivity, and the sensitivity of the interference light of wavelengths 550, 600, and 650 nm is denoted by reference numeral 121 in FIG. , 122 and 123 are indicated by a line to attach. 1 and 2, the sensitivity becomes very small near the maximum and the minimum of the relative reflection intensity (or reflectance) at either wavelength, and the S / N ratio (signal / It can be seen that the noise ratio) cannot be obtained. When the range of the film thickness at which the sensitivity falls below a predetermined value is referred to as the dead band, there is a dead band having a wide film thickness range near the maximum point of relative reflection intensity, and a dead zone having a narrow film thickness range exists near the minimum point. In addition, the sensitivity S _k can be expressed as in the formula (2) using the function f ' _{R obtained} by differentiating the function f _R in the formula (1) with the film thickness d as the film thickness d.

Here, the relationship between the sensitivity and the relative reflection intensity will be described by focusing on one wavelength of 550 nm. 3 is a diagram for explaining the relationship between sensitivity and relative reflection intensity. The upper left side of FIG. 3 shows the change of the film thickness of the sensitivity, the right side of the top shows the change of the relative reflection intensity of the sensitivity, and the right side of the bottom shows the change of the film thickness of the relative reflection intensity. It is shown as.

The two lines marked with the reference numeral 130 at the right side of the upper end in FIG. 3 are the angles of the two mountain forms existing along the valley portion having a relatively narrow width in the range of the film thickness at the left side of the upper end in FIG. 3. The two lines (corresponding to the peak on the left and the right on the bottom of the combination (the portion on which the relative reflection intensity decreases and the portion on which the relative reflection intensity increases as the film thickness increases on the right side of the lower portion in FIG. 3) respectively correspond to two lines ( 130) almost overlapping. In addition, in comparison between a plurality of combinations of two peaks on the left side of the upper end in FIG. 3, the relationship between the sensitivity corresponding to the left peak and the relative reflection intensity, and the sensitivity and the relative reflection intensity corresponding to the right peak There is little difference in each of the relations. Therefore, the relationship between the sensitivity and the relative reflection intensity can be said to be substantially constant without depending on the film thickness, and can be said to be the same in other wavelengths.

In fact, even when comparing between different wavelengths, the relationship between the sensitivity and the relative reflection intensity becomes roughly the same. Therefore, for example, one of the two lines 130 or the line representing the average of these lines 130 on the right side of the upper part in FIG. 3 is relative to the sensitivity at each wavelength as shown in FIG. 4. It is assumed that the relationship with reflection intensity is shown. In the relation between the sensitivity and the relative reflection intensity, assuming the maximum value of the relative reflection intensities R _kmax, the minimum value of R _kmin, sensitivity when the relative reflection intensity one R _k S _k by using the predetermined function f _s formula ( It can be expressed as 3).

Hereinafter, the details of the nonuniformity test using the relationship between the sensitivity and the relative reflection intensity will be described.

FIG. 5: is a figure which shows the structure of the nonuniformity inspection apparatus 1 which concerns on 1st Embodiment of this invention. The nonuniformity inspection device 1 is a glass substrate 9 that can be used for a display device such as a liquid crystal display device, the image of the film 92 of the resist for pattern formation formed on one main surface 91 It is an apparatus which acquires and checks a film thickness nonuniformity based on this image. The film 92 on the substrate 9 is formed by applying a resist liquid onto the upper surface 91 of the substrate 9.

Here, the film thickness non-uniformity of the film 92 on the substrate 9 depends on how the human eye feels and is not easy to define, but for example, the length of each of the species and the transverse In the substrate 9, which is about 2 m and serves as an assembly part of one display device, variation (nonuniformity) in the thickness of the film whose distance corresponding to one cycle ranges from several mm to several cm causes the film thickness unevenness. The magnitude of the amplitude of the variation can be understood as the film thickness nonuniformity (that is, the degree of nonuniformity). In addition, when the distance corresponding to one period of the variation in the thickness of the film is not included in the area corresponding to the panel of one display device on the substrate 9 on which the patterns of the plurality of panels are formed, The variation does not constitute a film thickness nonuniformity. Therefore, in the image which image | photographed the film | membrane 92 on the board | substrate 9, the magnitude | size of the amplitude (namely, the degree of local contrast fluctuation) of a predetermined spatial frequency band can be grasped | ascertained as film thickness nonuniformity.

As shown in FIG. 5, the nonuniformity inspection apparatus 1 has the main surface 91 (henceforth "top surface 91") in which the film | membrane 92 was formed (top (+ Z) side in FIG. 5). To the stage 2 holding the substrate 9, the light irradiation unit 3 for irradiating light to a film 92 on the substrate 9 held on the stage 2 at a predetermined incident angle, and the light irradiation unit 3. ) Is disposed between the light receiving unit 4, the substrate 9, and the light receiving unit 4, which receive light reflected from the film 92 on the upper surface 91 of the substrate 9. 4. A moving mechanism 21 for moving the wavelength conversion mechanism 5 and the stage 2 that change the wavelength of light received by the light 4 relative to the light irradiation unit 3, the light receiving unit 4, and the wavelength conversion mechanism 5. ), The nonuniformity detecting unit 7 and the nonuniformity detecting unit 7 for detecting the film thickness nonuniformity of the film 92 based on the intensity distribution of the light received by the light receiving unit 4 (distribution corresponding to the area of the upper surface 91). Film thickness in And a storage unit 6 for storing predetermined information used at the time of detection of the nonuniformity, and a control unit 8 for controlling these configurations.

The surface on the (+ Z) side of the stage 2 is preferably black and without any moisture. The movement mechanism 21 is a structure in which a ball screw (not shown) is connected to the motor 211, and the stage 2 moves along the guide 212 so that the stage 2 moves on the substrate 9 by rotating the motor 211. It moves in the X direction of FIG. 5 along the upper surface 91 of the.

The light irradiation section 3 is a halogen lamp 31 which is a light source that emits white light (that is, light including light of all wavelengths in the visible region), in the Y direction in FIG. 5 perpendicular to the moving direction of the stage 2. A circumferential quartz rod 32 that extends and a cylindrical lens 33 that extends in the Y direction are provided. In the light irradiation part 3, the halogen lamp 31 is provided in the edge part of the (+ Y) side of the quartz rod 32, and the light which entered the quartz rod 32 from the halogen lamp 31 is Y direction. Is converted into linear light (ie, light whose cross section is long in the Y direction) and exits from the side surface of the quartz rod 32 to pass through the cylindrical lens 33 to the upper surface of the substrate 9 ( 91). In other words, the quartz rod 32 and the cylindrical lens 33 convert the light from the halogen lamp 31 into linear light perpendicular to the moving direction of the stage 2 so that the top surface 91 of the substrate 9 It becomes the optical system leading to).

In FIG. 5, the optical path from the light irradiation unit 3 to the substrate 9 is indicated by a dashed-dotted line (the same applies to the optical path from the substrate 9 to the light receiving unit 4). A part of the light emitted from the light irradiation part 3 is reflected by the upper surface on the (+ Z) side of the film 92 on the upper surface 91 of the substrate 9.

The film 92 is light-transmissive with respect to the light from the light irradiator 3, and the light that is not reflected from the upper surface of the film 92 among the light from the light irradiator 3 passes through the film 92. Reflected on the upper surface 91 of the substrate 9 (ie, the lower surface of the film 92). In the nonuniformity inspection apparatus 1, the interference light of the light reflected from the upper surface of the film 92 on the substrate 9 and the light reflected from the upper surface 91 of the substrate 9 is the wavelength conversion mechanism 5. The light incident unit 4 enters the light receiving unit 4 via.

The wavelength conversion mechanism 5 has a disk shape for holding a plurality of optical filters (for example, an interference filter having a half width of 10 nm) 51 and a plurality of optical filters 51 that selectively transmit light having a plurality of different wavelengths, respectively. The filter wheel 52 of this invention, and the filter rotation motor 53 which is provided in the center of the filter wheel 52, and rotates the filter wheel 52 are provided. The filter wheel 52 is arrange | positioned so that the normal direction may be parallel to the optical path which reaches from the board | substrate 9 to the light receiving unit 4.

FIG. 6 is a view of the wavelength conversion mechanism 5 viewed along the side perpendicular to the filter wheel 52 from the substrate 9 side. As shown in Fig. 6, six circular openings 521 are formed in the filter wheel 52 at equal intervals in the circumferential direction, and five openings 521 therein have different transmission wavelengths. Five types of optical filters 51 are provided.

In the wavelength switching mechanism 5 shown in FIG. 5, the filter wheel 52 is rotated by the filter rotation motor 53 controlled by the control part 8, and among the five optical filters 51 (refer FIG. 6), Any one optical filter 51 (hereinafter, referred to as "selective optical filter 51a" in order to distinguish it from other optical filters 51) is selected, and the light-receiving unit 4 from the substrate 9 is selected. It is arranged on the light path to reach. Thereby, among the reflected light from the board | substrate 9 (that is, reflected light of white light containing interference light of five transmission wavelengths corresponding to five optical filters 51), the selection optical filter 51a arrange | positioned above an optical path Only interference light of a specific wavelength (hereinafter referred to as "specific wavelength") corresponding to) passes through the selection optical filter 51a and is led to the light receiving unit 4.

Then, when the filter wheel 52 is rotated by the filter rotation motor 53, the selection optical filter 51a disposed above the light path from the light irradiation unit 3 to the light receiving unit 4 among the plurality of optical filters 51. Can be switched to another optical filter 51, and the wavelength (ie, a specific wavelength) of the interference light received by the light receiving unit 4 is changed. In this way, the wavelength conversion mechanism 5 changes the selected optical filter 51a between the plurality of optical filters 51 to thereby select a specific wavelength related to the inspection of the film thickness nonuniformity of the plurality of optical filters 51. It becomes a switching mechanism which switches between transmission wavelengths.

The light receiving unit 4 is provided between the image capturing unit 41 and the image capturing unit 41 and the selection optical filter 51a of the wavelength conversion mechanism 5 to capture the reflected light from the substrate 9. It has a lens 42 leading to.

7 is a diagram illustrating the light receiving surface of the imaging unit 41. As shown in FIG. 7, the image pickup unit 41 is provided with a line sensor 410 having a plurality of image pickup elements (eg, CCD (Charge Coupled Device)) 411 arranged in a linear manner in the Y direction. In the imaging section 41 in FIG. 5, the selective optical filter 51a is transmitted among the linear light emitted by the light irradiation section 3 and reflected from the film 92 on the substrate 9. The interference light of one specific wavelength is received by the imaging element 411, and the intensity distribution of the interference light (that is, the distribution in the Y direction which is an output value from each imaging element 411) is obtained. In reality, the two-dimensional film 92 on the substrate 9 is obtained by repeatedly obtaining the intensity distribution of the interference light by the line sensor 410 of the imaging unit 41 as the substrate 9 moves in the X direction. An image is acquired.

The nonuniform detection unit 7 includes an output reception unit 71 that receives the output from the imaging unit 41 as a multi-gradation image of the film 92 on the substrate 9. In addition, the nonuniformity detection part 7 performs the filter process part 72 and the filter processing part 72 which perform the band pass filter process of the predetermined | prescribed spatial frequency band with respect to the image received by the output reception part 71. Contrast emphasis unit 73 for enhancing the contrast, and evaluation value calculation unit 74 for obtaining a predetermined evaluation value used for determining the presence or absence of film thickness nonuniformity (defective film thickness nonuniformity) from the image after contrast enhancement. It is further provided.

Next, the flow of inspection of the film thickness nonuniformity by the nonuniformity inspection apparatus 1 is demonstrated. FIG. 8: is a figure which shows the flow of the process by which the nonuniformity inspection apparatus 1 examines the film thickness nonuniformity of the film | membrane 92 on the board | substrate 9. FIG. When the film thickness nonuniformity of the film 92 on the board | substrate 9 is examined, the sensitivity information 61 used for the process by the nonuniformity detection part 7 is acquired previously, and stored in the memory | storage part 6 as a preparatory preparation. It is prepared (step S10).

The sensitivity information 61 shows the relationship between the sensitivity which is the ratio of the variation of the relative reflection intensity of the interference light of the specific wavelength to the variation of the film thickness, and the relative reflection intensity of the interference light of the specific wavelength (see Fig. 4). As described with reference to 3, after the relationship between the film thickness and the relative reflection intensity and the relationship between the film thickness and the sensitivity is requested, the sensitivity information 61 is obtained from these relationships. As will be described later, since the specific wavelength can be switched to the transmission wavelength of the optical filter 51 other than the selective optical filter 51a, the sensitivity information for each of the transmission wavelengths of the plurality of optical filters 51 is obtained. Although 61 is required, the relationship between the sensitivity and the relative reflection intensity of the interference light is almost the same in all the other wavelengths used in the nonuniformity inspection device 1 as described above. Only 61 is ready. In addition, the sensitivity information 61 is not limited to the material of the film or the like as long as the film 92 on the substrate 9 is a light-transmissive film. In reality, the process of step S10 is performed at the time of assembling the nonuniformity testing device 1, and is stored in the storage unit 6 as device-specific information so that the normality of the nonuniformity testing device 1 The use of is started from the following processing.

In the nonuniformity inspection apparatus 1, after the board | substrate 9 is hold | maintained on the stage 2 located in the test start position shown by the solid line in FIG. 5, in the (+ X) direction of the board | substrate 9 and the stage 2, Is started (step S11). Subsequently, the linear light emitted from the light irradiation part 3 and incident on the upper surface 91 of the substrate 9 at an incident angle of 60 degrees is a linear irradiation area on the upper surface 91 (hereinafter referred to as the "linear irradiation area"). Is irradiated (step S12), and the linearly-irradiated area is moved relative to the substrate 9.

The light from the light irradiator 3 is reflected by the upper surface 91 of the substrate 9, and only the light (interfering light) of a specific wavelength is picked up by passing through the selective optical filter 51a of the wavelength conversion mechanism 5. After exiting, it is led to the imaging unit 41. In the imaging section 41, the interference light of a specific wavelength is received by the line sensor 410, and the intensity distribution of the interference light in the linear irradiation area on the board | substrate 9 is acquired (step S13). The output from each imaging element 411 of the line sensor 410 is sent to the nonuniform detection part 7 and is received by the output reception part 71.

In the nonuniformity inspection apparatus 1, it is repeated during the movement of the board | substrate 9 whether the control part 8 moved the board | substrate 9 and the stage 2 to the test | inspection end position shown by the dashed-dotted line in FIG. If it is confirmed (step S14) and it does not move to the test | inspection end position, it returns to step S13, receives interference light of a specific wavelength, and acquires the intensity distribution of the light of a specific wavelength in a linear irradiation area. Is repeated. And when the board | substrate 9 and the stage 2 move to the test | inspection end position (step S14), the movement of the board | substrate 9 and the stage 2 by the movement mechanism 21 will be stopped, and irradiation of irradiation light will also stop. (Step S15). In the nonuniformity inspection apparatus 1, while the stage 2 is moving in the (+ X) direction, the operation of step S13 is repeated in synchronization with the movement of the stage 2 to thereby the entire substrate 9. Image data of the film 92 is obtained (that is, imaging is performed).

When the film 92 of the entire upper surface 91 of the substrate 9 is imaged (or in parallel with this operation), each of the accumulated line sensors 410 is output by the output accepting portion 71 of the nonuniformity detecting portion 7. The output from the imaging element 411 is arranged in chronological order while being converted to a value (pixel value) of, for example, 8 bits (may be other than 8 bits) based on a predetermined conversion formula. This is a two-dimensional image for processing in the nonuniform detection unit 7 (in fact, it is an image acquired by the imaging unit 41 and at the same time an image before the processing to be described below is given). (Original image) is generated and output to the filter processing unit 72. As described above, when the thickness of the film 92 is different, the reflectance of the film 92 is also different, so that the intensity of the reflected light received by the line sensor 410 is also different. Therefore, when nonuniformity exists in the distribution of the thickness of the film | membrane 92, ideally, the nonuniformity of the pixel value will arise also in an original image.

In addition, the output accepting unit 71 acquires the maximum value and the minimum value of the relative reflection intensity used in the processing described later for each imaging element 411. Here, generally, in the outer edge part of the film 92 on the board | substrate 9, the thickness of the film 92 will become the inclined part gradually decreasing toward the outer side, and this part at the time of imaging At the time of acquiring intensity distribution, the maximum value and minimum value of the output of the imaging element 411 respectively corresponding to the maximum intensity | strength and minimum intensity of interference light are acquired. Therefore, for example, the pixel value is a pixel in the original image from which the pixel value is obtained based on the output from one imaging device 411, and among the plurality of pixels arranged in one column in the direction corresponding to the X direction, the pixel value is the maximum value and the minimum value. It is required to be, and by dividing these values by 255, respectively, the maximum value and minimum value of the relative reflection intensity with respect to this imaging element 411 are acquired substantially.

In addition, since the maximum value and minimum value of the relative reflection intensity in each imaging element 411 depend on the arrangement | positioning with respect to the light irradiation part 3 of the said imaging element 411, the nonuniformity inspection apparatus 1 like sensitivity information 1 ) May be obtained at the time of assembling, and the value may be stored in the storage unit 6 and used at the time of inspection.

Subsequently, the original image is compressed by the filter processing unit 72 to generate a first image. Here, if the pixel value of a pixel located at coordinates (X, Y) in the original image is represented by F _xy , In the first image generated by compressing the image S _a , the pixel XS _a , and the range of pixels as a unit, the pixel value A _xy of the pixel of interest located at the coordinates (x, y) is determined by the equation (4). Become.

In the present embodiment, since S _a is 4 (pixels), the S / N ratio of the first image is increased by four times the original image by the calculation of equation (4). When the first image (the original image after compression) is generated, a low-pass filter process is performed on the first image, and the influence of the high frequency noise is suppressed from the first image to produce a smoothed second image. Low-window for determining the operation range of the low-pass filter processing, a pixel is a square of side length of the (2S _l +1), the second image pixel coordinate value of the pixel of interest is located at (x, y) in the _xy L Is obtained by equation (5) using the pixel value A (see equation (4)) in the first image of each pixel in the vicinity of the pixel of interest.

Thereafter, a high pass filter process is performed on the second image, and a third image is generated from which the low frequency concentration fluctuations that interfere with the contrast enhancement process described later are removed from the second image. Here, the pixel value H _xy of the pixel of interest located at the coordinates (x, y) uses the pixel value L (see Expression (5)) in the second image of each pixel in the vicinity of the pixel of interest. 6) is obtained.

Equation (6) is a window for determining the calculation range of the high pass filter processing, and represents a case where a square window of (2S ₂ +1) pixels is used for the length of each side centered on the pixel of interest. . As described above, in the filter processing unit 72, after performing the low-pass filter processing on the first image in which the original image is compressed, the high-pass filter processing is performed, whereby the band of the predetermined spatial frequency band is provided. A pass filter process is performed (step S16).

When the processing in the filter processing unit 72 is completed, the contrast emphasis unit 73 performs a contrast enhancement process on the third image to generate an emphasis image (step S17). The pixel value E _xy of the pixel of interest located at the coordinates (x, y) in the emphasis image is the pixel value H _xy contrast coefficient of the pixel of interest in the third image (here, it relates to the contrast width of the visualization target (target) Coefficient) r _c , the value C _xy after normalization by dividing the pixel value A of the pixel of interest in the first image before the band pass filter processing by 255 (or obtained in the storage unit 6). Stored in advance) and the function f _s obtained based on the sensitivity information 61 in FIG. 4 of the maximum value R _kmax and the minimum value R _kmin of the relative reflection intensity of the imaging element 411 which leads the pixel value of the pixel of interest. ) And the background value b, are obtained from equation (7).

In this embodiment, the contrast coefficient r _c is 0.05 and the background value b is 127. In addition, since the third image is a compressed image, the plurality of imaging elements 411 actually correspond to the image of interest, but the relative reflection intensity of any one imaging element 411 at the time of calculation of equation (7). The maximum value of R _kmax and the minimum value of R _kmin can be used.

Furthermore, if with respect to the quantization of the pixel value E _xy, the quality value is smaller than 0, as the pixel value E _xy is zero, and the value is greater than 255, the pixel value E _xy is said to be 255.

In equation (7), the contrast coefficient r _c for each pixel of the third image derived from the original image is multiplied by the sensitivity derived from the pixel value A _xy of the pixel to which the first image, which is the original image, is substantially multiplied. As a result, for each pixel of the third image, the higher the sensitivity at the time of acquiring the original image, the smaller the contrast enhancement is performed. That is, in the contrast emphasis section 73, the degree of contrast enhancement for each pixel of the third image is changed in accordance with the sensitivity, so that the effect of changing the sensitivity in the original image depending on the film thickness is affected by the emphasis image. It is easily corrected (reduced). In the formula (7), since the pixel value of each pixel of the third image is corrected using the maximum value and the minimum value of the relative reflection intensity according to the corresponding imaging element 411, between the plurality of imaging elements 411. It is also possible to simultaneously correct the influence of the difference of the output characteristics in the system.

When the contrast enhancement process is completed, a predetermined calculation process is performed on the emphasis image by the evaluation value calculation unit 74 shown in FIG. 5, and the overall nonuniformity in which the nonuniformity is scattered over the entire surface of the substrate 9. The degree of nonuniformity (so-called nonuniformity) with respect to various nonuniformities in which partial nonuniformity exists partially in the part on the board | substrate 9, and the striped nonuniformity which produces a linear nonuniformity in a row direction or a column direction. Intensity) is quantified, and an evaluation value which is a value representing the degree of nonuniformity (degree of amplitude) derived from the highlighted image is calculated (step S18).

By the way, the reflectance of the film 92 with respect to light of a specific wavelength fluctuates with periodicity with respect to the film thickness, as shown in FIG. 1, and as described with reference to FIGS. 1 and 2, near the maximum point of the reflectance and In the vicinity of the minimum, the sensitivity becomes very small. Therefore, in the case where the thickness of the film 92 is close to the film thickness at which the sensitivity is minute (that is, included in the dead band), the value of the pixel hardly fluctuates in the acquired original image, i.e., the film is uneven. The accuracy of the detection of the fluctuation in thickness) is reduced.

For example, if the thickness of the film 92 on the substrate 9 fluctuates in the dead zone, it is difficult to detect such film thickness nonuniformity with high accuracy only based on this specific wavelength.

Therefore, in the nonuniformity inspection apparatus 1, as described above, one of the optical filters 51 is selected optical filter 51a, and the steps S11 to S18 in which the transmission wavelength of the selective optical filter 51a is a specific wavelength. (Step S19), the filter rotation motor 53 of the wavelength conversion mechanism 5 is driven by the control part 8, the filter wheel 52 rotates, and the other optical filter 51 carries out the board | substrate ( It is arrange | positioned on the optical path which reaches the light receiving unit 4 from 9), and the specific wavelength in the wavelength conversion mechanism 5 is changed (step S20). By changing a specific wavelength, the dead band of a film thickness also moves, as shown in FIG.

Thereafter, the stage 2 is returned to the inspection start position by the moving mechanism 21, and the movement of the stage 2 is started again (step S11). In the nonuniformity inspection apparatus 1, the specificity different from the time of the first nonuniformity detection among the reflected light in the board | substrate 9 of the light from the light irradiation part 3 until the stage 2 reaches the test | inspection end position. After the light of the wavelength is received by the imaging section 41, the intensity distribution of the reflected light from the linearly radiated region on the substrate 9 is repeatedly acquired in synchronization with the movement of the stage 2, and sent to the nonuniform detection section 7, Movement of the stage 2 is stopped (steps S12 to S15).

Then, a band pass filter process of a predetermined spatial frequency band is performed on the first image obtained by compressing the original image obtained in the interference light of the specific wavelength after the change (step S16). Subsequently, the contrast emphasis section 73 performs contrast emphasis processing to generate an emphasis image (step S17), and an evaluation value of the film thickness nonuniformity in the emphasis image is calculated (step S18). When the processing of steps S11 to S18 at the specific wavelength after the change is finished (step S19), another optical filter 51 different from the first and second processes is transferred from the substrate 9 to the light receiving unit 4. It is arrange | positioned on the optical path which reaches, a specific wavelength is changed (step S20), and the process of steps S11-S18 is repeated.

In the nonuniformity inspection apparatus 1, when the process of step S11-S18 is repeated using the transmission wavelength of each of the five optical filters 51 as a specific wavelength (step S19), it acquires using the five optical filters 51, respectively. The large standard deviation of the pixel value of the pixel contained in the part corresponding to the predetermined | prescribed area | region (the area | region corresponded to the display surface in a display apparatus) of the film | membrane 92 on the board | substrate 9 among the five original images which were comprised was 2 Original images are selected as ones with a relatively high contrast. In addition, the selection of an original image having a relatively high contrast may be performed by another method.

Subsequently, of the evaluation values of the film thickness nonuniformity obtained for these two original images, the larger value is compared with a predetermined threshold. If the value is larger than the threshold value, the film 92 on the substrate 9 has a film thickness nonuniformity (hereinafter referred to as "film thickness nonuniformity defect") exceeding an allowable range. When the detected value is less than or equal to the threshold, it is assumed that there is no film thickness nonuniformity (film thickness nonuniformity defect) exceeding the allowable range in the film 92 on the substrate 9 (step S21). The inspection process of the film thickness nonuniformity by this is complete | finished.

In addition, when the inspection process of film thickness nonuniformity is continued with respect to another board | substrate (namely, the board | substrate which thinks the film thickness is equal to the board | substrate 9) of the same lot as the board | substrate 9, 2 selected as a comparatively high contrast The process of steps S11-S18 is performed using each of the transmission wavelengths of the two optical filters 51 corresponding to two images as a specific wavelength. (Same in the second embodiment described later.) In this case, the transmission wavelength of one optical filter 51 is designated as a specific wavelength in the path of the reciprocating movement of the substrate 9 in the X direction. The evaluation value is calculated while acquiring (step Sl9), and the same process is performed in which the transmission wavelength of the optical filter 51 on the other side is set to a specific wavelength in the return path of the substrate 9 (steps S20, S11). S19). And the larger value of the evaluation values acquired in the two wavelengths is compared with a threshold value, and film thickness nonuniformity defect is detected (step S21).

9 is a diagram showing a relationship between the sensitivity and the film thickness in the interference light of the wavelength corresponding to each optical filter 51. In Fig. 9, two wavelengths that are preferably selected according to a predetermined criterion in the corneal thickness are superposed on a horizontally long ridge on a line indicating the relationship between the sensitivity and the film thickness at each wavelength. It is represented by As a criterion for the selection of the two wavelengths, for example, it is said that they do not belong to the dead band within the possible range, and that the inclinations of the sensitivity at the two wavelengths are opposite to each other. As described above, in the nonuniformity inspection apparatus 1, two wavelengths in which the standard deviation of pixel values of the portion corresponding to the predetermined region of the film 92 becomes large in the original image are selected, and the films at these wavelengths are selected. Although the evaluation value of thickness nonuniformity is made into the determination object of the presence or absence of a film thickness nonuniformity defect, when the approximate thickness of the film | membrane 92 of the board | substrate 9 is a known fact, as shown in FIG. 9, each film thickness The two wavelengths which are preferably selected in the above are determined in advance, and two wavelengths to be used are specified based on the approximate thickness of the film 92, and the above treatment is performed using each of these wavelengths as a specific wavelength. Also good.

Next, the case where the same process is performed by the said nonuniformity detection part 7 with respect to a test image is demonstrated. First, a nonuniform image of M rows and M columns representing a film on a substrate having striped nonuniformities extending in a predetermined direction is prepared. In the non-uniform image, the brightness of the pixels belonging to the same row is the same, and the constant thickness (also called the background thickness) of the ideal film on the substrate which does not contain the nonuniformity component and the noise component is B, and the depth of the nonuniformity is _au , the nonuniform wavelength. Λ _u , the row number from the first row to the Mth row of the nonuniform image is y, and the row number of the center of the nonuniform image is y _c , and the function fs of the equation (1) in the interference light at a wavelength of 650 nm is the reflection intensity relative to G _s indicates the non-uniform image including the non-uniformity is obtained using the formula (8). In fact, it is said that a nonuniform image contains only the nonuniformity for one cycle.

Then, to each pixel of the non-uniform image represented by equation (8), the noise component n _s represented by equation (9) is added using a random number rnd of 0.0 to 1.0 as the standard deviation of the magnitude of the noise component as σ _n . And a test image in interference light of 650 nm wavelength.

In this way, the test image is simplified to a background film thickness, nonuniformity and noise, and a plurality of test images are prepared by combining various background film thicknesses and nonuniform depths. In practice, the size of the test image is 1200 x 1200 pixels, the wavelength λ _u of nonuniformity in equation (8) is 64, and the standard deviation σ _n of the magnitude of the noise component in equation (9) is 1 It's called .0.

Each test image is converted into an 8-bit original image, and then compression processing is performed in units of the range of S _a pixels × S _a pixels using Equation (4) to obtain a first image.

At this time, S _a in Formula (4) is assumed to be 4. Subsequently, after the low-pass filter process is performed using the formula (5) on the first image and the second image is obtained, the high-pass filter process using the formula (6) is performed on the second image. By this, a third image is obtained. Then, by performing contrast enhancement using Expression (7) on the third image, an emphasis image is obtained. In addition, in Formula (7), the maximum value R _kmax and the minimum value R _kmin of the relative reflection intensity are used for arbitrary imaging elements 411, and the contrast coefficient r _c is 0.05.

Here, as shown in equation (10), for each of each row in each emphasis image, the average value in a plurality of pixels belonging to the corresponding row of the absolute value of the difference between the pixel value and the background value b is the background. is divided by the value b is requested, it is calculated as the evaluation value of the maximum to the relative detection strength m _o to the film thickness non-uniformity of the stripe pattern. Here, the background value b is 127, as shown in equation (7), and each of hei and wid is M / 4.

The above processing is performed for each of a plurality of test images in which the nonuniform depth is changed to 1, 2, 5, 10, 20, 50, 100 nm, and the background film thickness is various values for each nonuniform depth. Intensity is obtained.

FIG. 10 is a diagram illustrating a change with respect to a background film thickness of relative detection intensity acquired from a plurality of test images. FIG. In addition, FIG. 11 is a comparative example with respect to FIG. 10. In the said contrast enhancement process, when f _s ((beta) _xy ) in Formula (7) is made constant at 1 (that is, the degree of contrast enhancement was made constant. Is a diagram showing a change in the background film thickness of the relative detection intensity acquired from the test image. In addition, in FIG.10 and FIG.11 (and FIG.12 mentioned later), the relative detection intensity whose value is one or more is shown as one.

In the case where the degree of contrast enhancement is made constant, as shown in Fig. 11, for example, the relative detection intensity of the nonuniformity of the depth of 1 nm is detected as 0.2 at the background thickness of 1670 nm and 50 nm in the vicinity of the background thickness of 1600 nm. Despite the nonuniformity of the depth, the relative detection intensity is detected as 0.2 or less. This is because sensitivity changes depending on film thickness, and in this case, it becomes difficult to acquire film thickness nonuniformity accurately. On the other hand, when the degree of contrast enhancement is changed in accordance with the sensitivity, as shown in Fig. 10, for the nonuniformity whose nonuniformity depth is smaller than 50 nm, the change in the relative thickness of the relative detection intensity with respect to the film thickness is different. Approximately flattened (almost constant).

In the nonuniformity inspection device 1, since interference light of five kinds of wavelengths is used as described above, the original image of the interference light having the desired wavelength is selected for each background film thickness, and the contrast is emphasized. As a result of the process of changing the degree of in accordance with the sensitivity, as shown in Fig. 12, at each non-uniform depth, the change in the film thickness of the relative detection intensity is further flattened. In Fig. 12, the relative detection intensity becomes substantially constant at the non-uniform depth of 10 nm or less, while the relative detection intensity is approximately proportional to the non-uniform depth, and at each of the nonuniform depths of 20 and 50 nm, the maximum value of the relative detection intensity is about twice the minimum value. This results in a smaller width of variation in relative detection intensity.

Moreover, at 100 nm nonuniformity depth, relative detection intensity is saturated in most background film thickness, and it can be said that it is stable and a film thickness nonuniformity can be detected.

In the change with respect to the backdrop film thickness of the relative detection intensity in each nonuniformity depth in FIGS. 10-12, the backdrop film thickness whose relative detection intensity is in the range of +/- 20% with respect to the average value of the relative detection intensity in each nonuniformity depth. Table 1 shows the ratio (hereinafter, referred to as "ratio of stability detection range") with respect to the whole of the range of (the full range of the backdrop film thickness in which the test image was prepared). In addition, as mentioned above, FIG. 11 shows the relative detection intensity in the image acquired in the single wavelength, and the correction by the change of a sensitivity is not performed, FIG. 10 is acquired in the single wavelength, The relative detection intensity in the image after correction of the influence due to the change in sensitivity is shown, and FIG. 12 is obtained at a wavelength selected from the plurality of wavelengths according to the film thickness, and in the image after correction of the influence due to the change in sensitivity. Relative detection intensity is shown in Table 1, and in Table 1, the item name indicating the ratio of the stable detection range in FIG. 11 is denoted as "when there is no correction", and the item name indicating the ratio of the stable detection range in FIG. The item name indicating the ratio of the stable detection range in FIG. 12, " wavelength selection and correction " If that denotes that ".

Table 1

Non-uniform depth [nm] One 2 5 10 20 50 100 If there is no correction 11 11 14 11 13 44 79 If there is a correction 64 64 59 43 43 53 89 When there is wavelength selection and correction 91 89 88 86 84 88 95

As shown in Table 1, by correcting the influence of sensitivity that varies depending on the film thickness, the film thickness range in which the relative detection intensity becomes largely unevenly decreases (that is, the ratio of the stable detection range increases). I learned. In addition, when selecting the wavelength according to the film thickness from a plurality of wavelengths while correcting the influence of the change in sensitivity, the ratio of the stable detection range in all cases of the nonuniformity depth 1, 2, 5, 10, 20, 50, 100 nm. This was approximately 90%, and it was found that high accuracy film thickness nonuniformity can be detected in a wider film thickness range.

As described above, in the nonuniformity inspection apparatus 1, the sensitivity is changed depending on the film thickness at the time of acquiring the original image in the imaging unit 41. While easily correcting by referring to the pixel value, the degree of amplitude of the predetermined spatial frequency band in the image derived from the original image is detected as film thickness nonuniformity. This makes it possible to accurately inspect the film thickness nonuniformity of the film 92 formed on the substrate 9. In addition, in the nonuniform detection part 7, correction of the influence by the change of the individual sensitivity according to the corresponding imaging element 411 is performed with respect to each pixel of the 3rd image derived from an original image, Simultaneously correcting the influence of the difference in output characteristics between the imaging elements 411 is also realized.

  In addition, in the nonuniformity inspection apparatus 1, only 2, 3 or 4 optical filters 51 can be used, and in step S19, S20 of FIG. Steps S11 to S18 may be repeated while switching between them (same in the second embodiment described later). Thereby, in the nonuniformity inspection apparatus 1, the nonuniformity inspection process is performed at a higher speed than the above-mentioned processing example in which the movement of the substrate 9 in the X direction is repeated five times using five optical filters 51. It is possible to precisely inspect the film thickness nonuniformity in a wide range.

FIG. 13: is a figure which shows the structure of the nonuniformity inspection apparatus 1a which concerns on 2nd Embodiment.

The nonuniformity inspecting apparatus 1a of FIG. 13 is added to the nonuniformity inspecting apparatus 1 of FIG. 5, for example, by a white interference type film thickness measuring unit 11. The film thickness measuring unit 11 is movable in the Y direction in FIG. 13 along the main surface of the substrate 9 by the measuring unit moving mechanism 12. The other structure is the same as the nonuniformity inspection apparatus 1 of FIG. 5, and attaches | subjects the same code | symbol.

FIG. 14 is a diagram showing a part of the flow of the process in which the nonuniformity inspection device 1a inspects the film thickness nonuniformity of the film 92 on the substrate 9, and is performed between step S10 and step S1 in FIG. 8. Indicates processing. When the nonuniformity inspection device 1a inspects the film thickness nonuniformity of the film 92 on the substrate 9, the sensitivity information 61a indicating the relationship between the sensitivity and the film thickness is prepared as a preliminary preparation and the storage unit 6 ) (FIG. 8: step S10). Here, the sensitivity information 61a in this embodiment shows the relationship between the sensitivity of the interference light of the transmission wavelength and the film thickness in each of the plurality of optical filters 51, as shown in FIG. (However, in Fig. 2, the relationship between the sensitivity and the film thickness of only three kinds of transmission wavelengths is shown.).

Subsequently, while controlling the movement mechanism 21 and the measurement part movement mechanism 12, the measurement position of the film thickness measurement part 11 is matched with the predetermined position of the film 92 on the board | substrate 9, and the said position The thickness of the film 92 in step S is obtained (step S31). In practice, the thickness of the film 92 at a plurality of positions (for example, 25 points) arranged substantially uniformly in the film 92 on the substrate 9 is measured, so that the distribution of the thickness of the film 92 is obtained. .

Thereafter, the stage 2 is returned to the inspection start position by the moving mechanism 21, and the movement of the stage 2 is started (FIG. 8: step S11). In the nonuniformity inspection apparatus 1a, in synchronization with the movement of the stage 2, the intensity distribution of the reflected light from the linear irradiation area on the substrate 9 is repeatedly acquired while light of a specific wavelength is received by the imaging unit 41. When the stage 2 reaches the inspection end position, the movement of the stage 2 is stopped (steps S12 to S15).

Then, after the original image is compressed in the non-uniformity detecting unit 7, a band pass filter process of a predetermined spatial frequency band is performed on the first image that is the compressed original image to obtain a third image (step S16), Thereafter, contrast enhancement processing is performed to generate an emphasis image (step S17). At this time, the thickness of the film 92 at the position corresponding to the pixel of interest is derived from the distribution of the thickness of the film 92 obtained by the film thickness measuring section 11, and from this thickness, the relationship between sensitivity and film thickness The sensitivity obtained by referring to the sensitivity information 61a indicating is used in place of f _s (β _xy ) in the calculation of equation (7) for the pixel of interest in the highlighted image. And the evaluation value of the film thickness nonuniformity in an emphasis image is computed (step S18).

In the nonuniformity inspection apparatus 1a, if the transmission of each of the five optical filters 51 as the specific wavelength is repeated, the processes of steps S11 to S18 are repeated (steps S19 and S20), and the five optical filters 51 are used. In each of the five original images obtained, two original images having a relatively high contrast are selected, and a larger value is compared with a predetermined threshold among evaluation values of film thickness nonuniformity obtained for these two original images. do. If the value is larger than the threshold, the film thickness nonuniformity exceeding the allowable range is detected in the film 92 on the substrate 9, and the film thickness nonuniformity is detected. If the value is less than the threshold, the substrate 9 It is assumed that there is no film thickness nonuniformity over the allowable range in the film 92 on the () (step S21), and the inspection processing for the film thickness nonuniformity by the nonuniformity inspection device 1a is completed. In addition, as shown in FIG. 9, when two wavelengths which are preferable to be selected in each film thickness are predetermined, the two wavelengths to be used are specified based on the thickness of the film 92 obtained. The above treatment may be performed using each of these wavelengths as a specific wavelength.

As described above, in the nonuniformity inspection apparatus 1a of FIG. 13, the thickness 92 of the film 92 and the effect of the sensitivity being changed depending on the film thickness at the time of acquiring the original image in the imaging section 41. While easily correcting by referring to the sensitivity information 61a, the degree of amplitude of the predetermined spatial frequency band in the image derived from the original image is detected as the film thickness nonuniformity. This makes it possible to accurately inspect the film thickness nonuniformity of the film 92 formed on the substrate 9. In addition, in the nonuniformity inspection device 1a, the third image drawn from the original image is provided in the nonuniformity detection portion 7 throughout the preparation of a table showing the relationship between sensitivity and film thickness for each imaging element 411. Each pixel may be corrected for the influence due to the change in the individual sensitivity according to the corresponding imaging element 411.

As mentioned above, although embodiment of this invention was described, this invention is not limited to the said embodiment, A various deformation | transformation is possible.

In the first and second embodiments described above, in order to generalize and grasp the intensity of the interference light, the relative reflection intensity, which is the intensity of the interference light when the intensity of the incident light is 1, has been described. Since the intensity is substantially the same as the intensity of the interference light of a specific wavelength, the sensitivity can be understood as the ratio of the intensity variation of the interference light of the specific wavelength to the variation of the film thickness. In this case, in the first embodiment, The sensitivity information 61 indicates the relationship between the sensitivity and the intensity of the interference light of a specific wavelength.

For example, for each pixel of the original image acquired by the imaging unit 41, a corresponding sensitivity is requested with reference to the sensitivity information from the pixel value (or from the thickness at the position where the film 92 corresponds). By multiplying the coefficient according to the sensitivity with the pixel value, an image in which the influence of the sensitivity change is reduced may be obtained. In that case, the influence by the change of the individual sensitivity according to the imaging element 411 corresponding to each pixel of an original image is correct | amended.

By applying only the band pass filter to this image, an image indicating the degree of amplitude in a predetermined spatial frequency band is output as a detection result of film thickness nonuniformity. In this case, in the nonuniformity detecting section, the correction of the influence due to the change in sensitivity is performed as a process independent from the contrast enhancement process, and the film thickness nonuniformity in the image representing the degree of amplitude of the predetermined spatial frequency band in the original image Is detected. As described above, if the degree of amplitude of the predetermined spatial frequency band in the original image or the image derived from the original image is detected as the film thickness nonuniformity, while correcting the effect of the sensitivity varying depending on the film thickness, the nonuniformity detecting unit Various treatments may be performed, and the detection result of film thickness nonuniformity may also be an image showing the film thickness nonuniformity other than the evaluation value.

In the first embodiment, the sensitivity information 61 indicating the relationship between the sensitivity and the relative reflection intensity is not necessarily calculated by calculation. For example, a substrate having various film thicknesses is prepared, and the relative reflection intensity and The relationship between the sensitivity and the relative reflection intensity may be derived by obtaining the relationship with the film thickness by actual measurement. Furthermore, a function that approximates the relationship between the sensitivity and the relative reflection intensity shown in FIG. 4 may be obtained as sensitivity information, and the influence of the change in sensitivity on the original image may be corrected based on this function. Also in this case, it is possible to correct to some extent the influence of sensitivity that varies depending on the film thickness.

In the first and second embodiments described above, the degree of contrast enhancement in the contrast emphasis section 73 is changed to change the sensitivity of the image drawn from the original image acquired by the imaging section 41. Correction of the effect by the effect is efficiently carried out, but as described above, when the correction of the effect by the change in sensitivity is performed as a process independent from the contrast enhancement process, the above-described patent No. 3333503 It is also possible to partially use the technique of the call publication. In this technique, a smoothing image is requested by performing a smoothing process with a median filter on the original image, and the unevenness to be detected is divided by dividing the pixel value of each pixel of the original image by the corresponding pixel value. An emphasis image is acquired.

In the nonuniformity inspection apparatuses 1 and 1a, the wavelength conversion mechanism 5 switches the optical filter 51 on the substrate 9 side of the light receiving unit 4 so that a plurality of wavelengths having different specific wavelengths are different. Although it can switch between, for example, the light irradiation part 3 is provided with the several light source which emits the light of a different wavelength from each other, and can switch the light source activated by the control of the control part 8, for example. The specific wavelength which is the wavelength of the interference light received by the imaging part 41 may be switched between the some wavelength which differs (that is, the control part 8 becomes a switching mechanism).

In the light output unit, a fiber array in which a plurality of optical fibers are directly arranged in a line in place of the quartz rod 32 may be provided, and the light from the halogen lamp 31 may be converted into linear light by passing through the fiber array. . In addition, instead of the halogen lamp 31 and the quartz rod 32, a plurality of light emitting diodes arranged directly on the line may be provided as a light source for emitting the linear light. In the imaging section 41, when the imaging time of the substrate 9 needs to be shortened, a two-dimensional CCD sensor may be provided instead of the line sensor 410.

The holding part which holds the board | substrate 9 contacts the lower surface of the board | substrate 9, for example, in addition to the stage 2 which supports the board | substrate 9, for example, by holding the outer edge part of the board | substrate 9, the board | substrate 9 May be maintained.

The nonuniformity inspection apparatus according to the above embodiment may be used for detecting film thickness nonuniformity of an insulating film or a conductive film formed on a film other than a resist film, for example, the substrate 9. Further, the nonuniformity inspection apparatus may be used for inspection of film thickness nonuniformity of a film formed on another substrate such as a semiconductor substrate.

Although this invention was described in detail and demonstrated, the above description is illustrative and not restrictive. Accordingly, it is understood that many modifications and embodiments are possible without departing from the scope of this invention.

According to the present invention, the film thickness nonuniformity can be inspected with high accuracy by correcting the influence of sensitivity that changes depending on the film thickness.

Claims (15)

In the nonuniformity inspection device for inspecting the film thickness nonuniformity of the film formed on the substrate, A holding part for holding a substrate on which a light transmissive film is formed on the main surface; A light irradiation part for irradiating light to the film at a predetermined angle of incidence; An imaging unit which receives interference light having a specific wavelength reflected from the film among the light emitted from the light irradiation part to acquire an original image of the film; A predetermined spatial frequency in the original image or an image derived from the original image, while correcting the effect that the sensitivity, which is the ratio of the intensity variation of the interference light of the specific wavelength to the variation in the film thickness, changes depending on the film thickness A nonuniformity inspection device comprising a nonuniformity detection portion that detects the degree of amplitude of a band as a film thickness nonuniformity. The method of claim 1, A storage section for storing sensitivity information indicating a relationship between the sensitivity and the intensity of the interference light of the specific wavelength, And the nonuniformity detecting unit detects the film thickness nonuniformity while correcting the influence of the change in sensitivity by referring to the value of each pixel of the original image and the sensitivity information. The method of claim 1, A storage section for storing sensitivity information indicating a relationship between the sensitivity and the film thickness, And the nonuniformity detecting unit detects the film thickness nonuniformity while correcting the influence of the change in sensitivity with reference to the thickness of the film on the main surface and the sensitivity information. The method of claim 3, A nonuniformity inspection device, further comprising a film thickness measurement portion for acquiring the thickness of the film on the main surface. The method of claim 2, The imaging section has a plurality of imaging elements, and the non-uniformity detection section is a change in the individual sensitivity according to the corresponding imaging element for each pixel of the original image or the image drawn from the original image. Non-uniformity tester to correct the effects of The method of claim 1, A filter processing unit for performing the band pass filter processing of the predetermined spatial frequency band on the original image; A contrast emphasis section that emphasizes the contrast of the original image after being processed by the filter processing section, and correction for the influence of the change in sensitivity is performed by changing the degree of contrast enhancement in the contrast emphasis section. Non-uniformity inspection device. The method of claim 1, And a switching mechanism for switching the specific wavelength between a plurality of different wavelengths. In the nonuniformity inspection method for inspecting the film thickness nonuniformity of a film formed on a substrate, A light irradiation step of irradiating the light-transmitting film formed on the main surface of the substrate at a predetermined angle of incidence by the light irradiation unit; An imaging process of acquiring the original image of the film by receiving the interference light of a specific wavelength reflected from the film among the light from the light irradiation part in the imaging section A predetermined spatial frequency in the original image or an image derived from the original image, while correcting the effect that the sensitivity, which is the ratio of the intensity variation of the interference light of the specific wavelength to the variation in the film thickness, changes depending on the film thickness. A nonuniformity inspection method comprising a nonuniformity detection step of detecting the degree of amplitude of a band as a film thickness nonuniformity. The method of claim 8, And a preparation step of preparing sensitivity information indicating a relationship between the sensitivity and the intensity of the interference light of the specific wavelength before the non-uniform detection step, In the nonuniformity detecting step, a film thickness nonuniformity is detected while correcting the influence of the change in sensitivity with reference to the value of each pixel of the original image and the sensitivity information. The method of claim 9, In the preparation step, a first relationship between the film thickness and the intensity of the interference light of the specific wavelength and a second relationship between the film thickness and the sensitivity are obtained, and the first relationship and the second relationship are obtained. The nonuniformity inspection method in which the sensitivity information is obtained from the relationship. The method of claim 8, Before the heterogeneous detection step, further comprising a preparation step of preparing sensitivity information indicating a relationship between the sensitivity and the film thickness, The nonuniformity detecting method in the nonuniformity detecting step wherein a film thickness nonuniformity is detected while correcting the influence of the change in sensitivity with reference to the thickness of the film on the main surface and the sensitivity information. The method of claim 11, And a film thickness measurement step of acquiring the thickness of the film on the main surface before the nonuniform detection step. The method of claim 9, The imaging section has a plurality of imaging elements, and in the nonuniform detection step, for each pixel of the original image or the image derived from the original image, the influence of the change of the individual sensitivity according to the corresponding imaging element Non-uniformity test method in which correction is performed. The method of claim 8, The non-uniform detection step includes a step of performing a band pass filter process of the predetermined spatial frequency band on the original image; And a step of emphasizing the contrast of the original image after the band pass filter treatment, and the correction for the influence of the change in sensitivity is performed by changing the degree of contrast enhancement in the step of emphasizing the contrast. Nonuniformity Test Method. The method of claim 8, A nonuniformity inspection method further comprising a roughening step of sequentially repeating the light irradiation step to the nonuniformity detection step while switching the specific wavelength between two, three, or four kinds of wavelengths different from each other.

Images