Fluid observation device

Abstract

Provided is a method that enables analysis direction-independent observation of a fluid by using an analysis method based on an S-BOS method. This fluid observation device comprises: an imaging unit (50) that captures a background image (53) via an observation region (S) in which a fluid is present; and a computer (60) that analyzes the flow of the fluid by a background-oriented Schlieren method using the image captured by the imaging unit (50). The background image (53) has a pattern in which light and dark regions repeatedly change in both the lateral direction and the longitudinal direction. The computer (60) is configured to analyze the flow of the fluid in both the lateral direction and the longitudinal direction of the background image (53).

Description

Fluid observation device

[0001] The subject matter disclosed in this specification relates to a fluid observation device.

[0002] Conventionally, an apparatus for cleaning a substrate such as a semiconductor wafer by supplying a processing liquid to the substrate is known. In this type of apparatus, a slight change in the air flow in the chamber may cause a deterioration in the product quality. Therefore, there is a need for a technique that can observe the state of the air flow in the chamber while performing the cleaning process of the substrate.

[0003] As a technique for observing the air flow, for example, the Background Oriented Schlieren (BOS) method is known. The BOS method is a technique in which a background (such as a random dot pattern) arranged through an observation target is imaged by a camera, and a refractive index change associated with a temperature change or a density gradient by the observation target is obtained from an optical distortion of the background distribution. According to the BOS method, by capturing a local refractive index change, the movement of gas and heat flow can be visualized simply and reasonably. A conventional apparatus using the BOS method is described in, for example, Patent Document 1.

[0004] When considering the BOS method as an air flow visualization technique, there were several problems related to the processing method. Specifically, the processing load of the cross-correlation calculation is large, it becomes difficult to output immediately as the number of pixels increases, and it is necessary to appropriately set the inspection area and the search area required for the cross-correlation calculation. Thus, the BOS method had problems to be solved from the viewpoints of image clarity and immediacy, and automation of the technique.

[0005] As a technique for solving the problems of the BOS method, the S-BOS (Simplified-Background Oriented Schlieren) method has been proposed (for example, Non-Patent Document 1). The S-BOS method is a technique for visualizing the air flow by using a periodic pattern (such as a stripe) for the background pattern and calculating the amount of movement (phase change) thereof. Different from the normal BOS method, the S-BOS method can be analyzed by simple trigonometric function calculations, so the processing load is light and it can be applied to real-time observation.

[0006] Japanese Patent Publication No. 2020-143917

[0007] Junichi Akatsuka, Shinji Nagai, Shinji Honami, “Improvement of a visualization method for density gradients based on the Background-Oriented Schlieren method,” JSME Vol. 77 No. 784, 2391-2400 (2011). Philippe Weinzaepfel, Jerome Revaud, Zaid Harchaoui, Cordelia Schmid, “DeepFlow: Large displacement optical flow with deep matching,” HAL-OS pp. 1385-1392 (2013).

[0008] However, the S-BOS method only performs analysis in directions orthogonal to the periodic pattern of the background image, such as stripes. Therefore, it can only obtain analysis results in one direction corresponding to that periodic pattern. Consequently, since the sensitivity and resolution depend on the direction of analysis, it was necessary to determine the direction of analysis in advance. Therefore, in the visualization of fluids with an undefined direction, such as turbulence, the interpretation of analysis results using the S-BOS method was difficult, making its application challenging.

[0009] The object of the present invention is to provide a method that enables fluid observation independent of the analysis direction using an analysis method based on the S-BOS method.

[0010] To solve the above problems, the first embodiment is a fluid observation device for observing the flow of a fluid, comprising: an imaging unit that captures a background image through an observation area where the fluid is present; and a computer that analyzes the flow of the fluid using the image acquired by the imaging unit by the background Schlieren method, wherein the background image has a pattern in which the intensity of light and dark changes repeatedly in both the horizontal and vertical directions, and the computer is configured to analyze the flow of the fluid with respect to both the horizontal and vertical directions of the background image.

[0011] The second embodiment is a fluid observation apparatus according to the first embodiment, wherein the computer synthesizes the analysis results of the fluid flow in both the lateral and vertical directions.

[0012] The third embodiment is a fluid observation apparatus according to the first or second embodiment, wherein the background image has a pattern in which the intensity changes periodically in both the horizontal and vertical directions.

[0013] A fourth embodiment is a fluid observation apparatus according to the third embodiment, wherein the background image has a pattern composed of two figures of different colors or tones that are alternately arranged in both the horizontal and vertical directions.

[0014] The fifth embodiment is a fluid observation apparatus according to the first or second embodiment, wherein the background image has a pattern that does not have periodicity in both the horizontal and vertical directions.

[0015] According to the fluid observation apparatus of the first to fifth embodiments, by obtaining analysis results in two directions, the longitudinal and transverse, the multidimensional characteristics of the fluid flow can be inferred, enabling more accurate analysis. In other words, based on the S-BOS method, it becomes possible to observe the fluid independently of the analysis direction.

[0016] According to the fluid observation device of the second embodiment, by synthesizing the analysis results in the lateral and vertical directions, it becomes possible to intuitively visualize the flow even of a turbulent fluid.

[0017] According to the fluid observation device of the third embodiment, by using a periodic background image, the flow can be analyzed by simple trigonometric function calculations, thereby significantly reducing the processing load.

[0018] This is a longitudinal cross-sectional view of a substrate processing apparatus equipped with a fluid observation device according to one embodiment. This is a diagram showing an example of a background image. This is a diagram showing another example of a background image. This is a block diagram showing the electrical connections between the computer and each part of the substrate processing apparatus. This is a diagram conceptually illustrating the S-BOS method when using the background image shown in Figure 2 (an image having periodicity in the horizontal and vertical directions). This is a diagram conceptually illustrating the S-BOS method when using the background image shown in Figure 3 (a random dot pattern). This is a diagram showing the flow of obtaining an S-BOS analysis image from a time-series measurement image. This is a diagram showing an additional analysis example based on a composite image.

[0019] Embodiments of the present invention will be described below with reference to the attached drawings. Note that the components described in these embodiments are merely illustrative and are not intended to limit the scope of the present invention to them alone. In the drawings, for ease of understanding, the dimensions and number of parts may be exaggerated or simplified as needed.

[0020] <1. Embodiment> Figure 1 is a longitudinal cross-sectional view of a substrate processing apparatus 1 equipped with a fluid observation device according to one embodiment. The substrate processing apparatus 1 is a device that supplies a processing liquid to the surface of a disc-shaped substrate W in the semiconductor wafer manufacturing process to clean the substrate W. As shown in Figure 1, the substrate processing apparatus 1 comprises a chamber 10, a substrate holding unit 20, a rotating mechanism 30, a processing liquid supply unit 40, an imaging unit 50, and a computer 60.

[0021] The chamber 10 is a housing that encloses a processing space 11 for processing the substrate W. The chamber 10 has side walls 12 surrounding the sides of the processing space 11, a top plate 13 covering the upper part of the processing space 11, and a bottom plate 14 covering the lower part of the processing space 11. The substrate holding section 20, the rotating mechanism 30, and part of the processing liquid supply section 40 (specifically, the nozzle 45 and the cup 47) are housed inside the chamber 10. Part of the side wall 12 is provided with an entrance for loading and unloading the substrate W into and out of the chamber 10, and a shutter for opening and closing the entrance.

[0022] The substrate holding section 20 is a mechanism that holds the substrate W in a horizontal position (with the normal to the main surface aligned with the vertical direction) inside the chamber 10. As shown in Figure 1, the substrate holding section 20 has a disc-shaped spin base 21 and a plurality of chuck pins 22. The plurality of chuck pins 22 are arranged at predetermined intervals on the outer circumference of the upper surface of the spin base 21. The substrate W is held by the plurality of chuck pins 22 with the surface to be processed, on which the pattern is formed, facing upward. Each chuck pin 22 contacts the lower surface and outer end surface of the peripheral edge of the substrate W, and supports the substrate W above the spin base 21 with a small gap between it and the upper surface of the spin base 21.

[0023] The rotation mechanism 30 is a mechanism for rotating the substrate holder 20. As shown in Figure 1, the rotation mechanism 30 has a motor 31 and a support shaft 32. The support shaft 32 extends in the vertical direction. The lower end of the support shaft 32 is connected to the motor 31. The upper end of the support shaft 32 is fixed to the center of the lower surface of the spin base 21. When the motor 31 is driven, the support shaft 32 rotates about its axis 320. Along with the support shaft 32, the substrate holder 20 and the substrate W held by the substrate holder 20 also rotate about the axis 320.

[0024] The processing liquid supply unit 40 is a mechanism that supplies processing liquid to the upper surface of the substrate W held by the substrate holding unit 20. The processing liquid supply unit 40 includes a liquid supply source 41, liquid supply piping 42, a valve 43, a pump 44, a nozzle 45, and a cup 47. The liquid supply source 41 is a tank that stores the processing liquid before supply. Examples of processing liquids used include SPM cleaning solution (a mixture of sulfuric acid and hydrogen peroxide), SC-1 cleaning solution (a mixture of ammonia water, hydrogen peroxide, and pure water), SC-2 cleaning solution (a mixture of hydrochloric acid, hydrogen peroxide, and pure water), DHF cleaning solution (dilute hydrofluoric acid), and pure water (deionized water).

[0025] The liquid supply piping 42 is a pipe that connects the liquid supply source 41 and the nozzle 45. The valve 43 and the pump 44 are located along the path of the liquid supply piping 42. The nozzle 45 is positioned facing the upper surface of the substrate W. When the valve 43 is opened and the pump 44 is operated, processing liquid is supplied from the liquid supply source 41 through the liquid supply piping 42 to the nozzle 45. The processing liquid is then discharged from the nozzle 45 toward the center of the upper surface of the substrate W.

[0026] The cup 47 is a substantially cylindrical member. The cup 47 can surround the sides of the substrate holding portion 20 and the substrate W held by the substrate holding portion 20. The cup 47 can also move up and down between a position higher than the substrate W held by the substrate holding portion 20 and a position lower than the substrate W by a lifting mechanism.

[0027] The processing liquid discharged from the nozzle 45 onto the upper surface of the substrate W spreads outwards from the center of the substrate W due to the centrifugal force caused by the rotation of the substrate W. The processing liquid that has scattered outwards from the substrate W is collected in the cup 47 and then discharged to the outside of the chamber 10.

[0028] Although not shown in the figures, the substrate processing apparatus 1 may also be provided with an airflow generating unit that forms a downflow of purified air within the chamber 10.

[0029] The imaging unit 50 is a device that captures a background image 53. The imaging unit 50 has an image sensor composed of an image element such as a CCD (Charge-Coupled Device) or CMOS (Complementary Metal Oxide Semiconductor). The background image 53 is an image placed inside the chamber 10. The background image 53 has a pattern in which the intensity of light and dark changes repeatedly in both the vertical direction and the horizontal direction that intersects (in this case, orthogonal to) the vertical direction. The background image 53 is positioned perpendicular to the optical axis A of the imaging unit 50.

[0030] A transparent window 15 is provided in the side wall 12 of the chamber 10. The imaging unit 50 acquires multi-gradation image data composed of multiple pixels by capturing light incident on the image sensor from the background image 53 through the observation area S and the window 15. The imaging unit 50 also acquires image data (measurement images) along multiple time series by repeatedly capturing images with the image sensor at minute time intervals.

[0031] Figure 2 shows an example of background image 53. The pattern shown in Figure 2 is a checkerboard pattern in which two rectangles of different colors (black and white) are repeatedly arranged alternately in a vertical and horizontal direction. The checkerboard pattern shown in Figure 2 is also an example of a pattern in which the shades of light and dark change periodically in both the horizontal and vertical directions. Furthermore, the checkerboard pattern is an example of a pattern composed of two figures of different colors or tones that are arranged alternately in both the horizontal and vertical directions.

[0032] Figure 3 shows another example of the background image 53. The pattern of the background image 53 shown in Figure 3 is a random dot pattern consisting of a large number of randomly placed dots. The pattern shown in Figure 3 is an example of a pattern that does not have periodicity in both the horizontal and vertical directions.

[0033] The computer 60 shown in Figure 1 is an information processing device that controls the operation of each part of the substrate processing device 1 and analyzes the flow of fluids. Figure 4 is a block diagram showing the electrical connections between the computer 60 and each part of the substrate processing device 1. As shown in Figure 4, the computer 60 has a processor 61 such as a CPU, memory 62 such as RAM, and a storage unit 63 such as a hard disk drive.

[0034] The memory unit 63 stores an operation control program P1 and a fluid observation program P2. The operation control program P1 is a computer program for controlling the operation of each part of the substrate processing apparatus 1 in order to perform processing on the substrate W in the substrate processing apparatus 1. The fluid observation program P2 is a computer program for analyzing the state of the fluid in the chamber 10 based on image data obtained from the imaging unit 50.

[0035] As shown in Figure 4, the computer 60 is electrically connected to the chuck pin 22, motor 31, valve 43, and pump 44, respectively. The computer 60 controls the operation of each of the above parts according to the operation control program P1 stored in the memory unit 63. This allows the cleaning process of the substrate W in the substrate processing apparatus 1 to proceed.

[0036] Furthermore, the computer 60 is electrically connected to the imaging unit 50. The computer 60 performs an analysis of the fluid flow in the chamber 10 based on the image data obtained from the imaging unit 50, according to the fluid observation program P2 stored in the storage unit 63. In this embodiment, the imaging unit 50, the background image 53, and the computer 60 constitute a fluid observation device.

[0037] In this embodiment, the computer 60 analyzes the flow of gas in the observation area S between the imaging unit 50 and the background image 53. However, by setting the observation area to a region where liquid is present, the computer 60 may be made to analyze the flow of liquid in the observation area.

[0038] The computer 60 uses the image data input from the imaging unit 50 to analyze the airflow in the observation area S using the background-oriented Schlieren (BOS) method. In the BOS method, refractive index changes due to temperature changes and density gradients caused by the observed object are obtained from the optical distortion of the background distribution. In this embodiment, the Simplified-Background-Oriented Schlieren (S-BOS) method described in Non-Patent Literature 1 is used.

[0039] In the S-BOS method described in Non-Patent Literature 1, a background image with periodically changing density in one direction is used for imaging, and the fluid density gradient is detected in that direction. In contrast, in this embodiment, a background image 53 with repeatedly changing density in both the vertical (Y direction) and horizontal (X direction) directions is used to detect the fluid density gradient in each of the two directions and analyze the flow. The analysis flow will be explained below.

[0040] In the analysis of fluid in the horizontal and vertical directions, similar to the S-BOS method described in Non-Patent Document 1, the luminance of each pixel in the output image is calculated according to the following formula (1).

[0041]

[0042] In formula (1), "I(i, j) out " is the luminance value of pixel (i, j) in the output image. Pixel (i, j) means the pixel in the i-th column from the left and the j-th row from the top of the image. I(i, j) meas is the luminance value of pixel (i, j) in the measurement image. The measurement image is, specifically, time-series image data acquired by the imaging unit 50. I(i, j) ref is the luminance value of pixel (i, j) in the reference image. The reference image is, for example, the background image 53 or an image obtained by imaging the background image 53 with the imaging unit 50. Note that it is also possible to use the immediately preceding image as the reference image. ΔI(i, j) is the luminance gradient of pixel (i, j), and more specifically, it is the average value of the luminance gradient in the measurement image and the luminance gradient in the reference image. ΔI(i, j) is calculated by, for example, the central difference method using a Sobel filter or the like. Also, I(i, j) offset is an offset term set to give the luminance value at a location with no change, and usually, the median luminance of the background is used (128 in the case of an 8-bit grayscale image).

[0043] FIG. 5 is a diagram conceptually showing the S-BOS method when using the background image 53 (an image having periodicity in the horizontal and vertical directions) shown in FIG. 2. As shown in FIG. 5, when taking the difference between the measurement image distorted by the fluid flow and the reference image without distortion according to formula (1), a change in the luminance value corresponding to the phase difference φ appears. However, since both positive and negative luminance values appear with respect to the center of vibration, only the difference cannot obtain an output corresponding to the displacement direction of the target image. Therefore, as shown in formula (1), the product of the luminance and the gradient of the target pixel is taken. Thereby, the luminance value of the target pixel can be divided into positive and negative with respect to the vibration center according to the displacement direction of the image.

[0044] As described in Non-Patent Document 1, when the luminance change of the background image is a sine wave, if the moving average of the output image obtained according to Equation (1) is performed within the range of half a period (1 / 2 period) of the background image, the luminance of the output image is represented by the following Equation (2).

[0045]

[0046] As shown in Equation (2), the moving average of the luminance in the output image can be calculated as a sine function of the phase difference φ. Thus, when using the periodic background image 53, it becomes possible to analyze the flow by simple trigonometric function calculations. Thereby, the processing load can be significantly reduced.

[0047] FIG. 6 is a diagram conceptually showing the S-BOS method when using the background image 53 (random dot pattern) shown in FIG. 3. In this case, based on Equation (1), by taking the difference between the measured image with distortion and the reference image without distortion, the position displacements of the dots Dt of the background image 53 in the horizontal and vertical directions can be detected respectively.

[0048] As described above, the computer 60 calculates an output image as an analysis result based on the measured image, which is time-series image data, and the reference image using Equation (1). In the following description, the output image obtained based on Equation (1) is referred to as an S-BOS analysis image.

[0049] FIG. 7 is a diagram showing the process of obtaining an S-BOS analysis image from a time-series measured image 71. As shown in FIG. 7, S-BOS analysis images 71X and 71Y in the horizontal direction (X) and the vertical direction (Y) are obtained from the time-series measured image 71 respectively. The computer 60 integrates the time-synchronized S-BOS analysis images 71X and 71Y using a compositing method such as alpha blending, and as a result, generates a composite image 71XY.

[0050] As shown in Figure 7, it is difficult to properly analyze fluids with an undefined flow direction, such as turbulent flow, using only one of the two S-BOS analysis images 71X and 71Y, which are analyzed in either the horizontal or vertical direction. In other words, because irregular and complex flows such as turbulence fluctuate in many directions, the flow can only be partially captured from the analysis results of only one direction, making it difficult to accurately grasp the flow characteristics. In contrast, as in this embodiment, by obtaining analysis results in two directions, it becomes possible to infer the multidimensional characteristics of the flow and enable a more accurate analysis. In particular, by combining the S-BOS analysis images 71X and 71Y, which are the analysis results, it becomes possible to intuitively visualize the flow even in a turbulent fluid state.

[0051] Furthermore, instead of combining the S-BOS analysis images 71X and 71Y, which are the analysis results in both the vertical and horizontal directions, the analysis values ​​obtained from these images may be combined (averaged, summed, or vector summed, etc.) and output as an image. Alternatively, it is possible to utilize the individual analysis results without combining the analysis results in both the vertical and horizontal directions. For example, in turbulent flow, it is possible to evaluate which direction (vertical or horizontal) has a larger airflow. Analysis may also be performed on a specific directional component of interest, such as the diagonal direction.

[0052] Figure 8 shows an example of additional analysis based on the composite image 71XY. As shown in Figure 8, optical flow may be applied as an additional analysis example. In optical flow, the movement of each pixel between temporally continuous composite images 71XY is calculated as a vector. The method described in Non-Patent Literature 2 can be used as the optical flow. This optical flow makes it possible to estimate the fluid's velocity and direction of movement, thereby enabling objective fluid flow imaging.

[0053] Furthermore, as shown in Figure 8, variance analysis may be applied as another additional analysis example. In this example, for a process in which wafer rotation and liquid ejection are sequentially added, starting from a downflow-only state, the variance of brightness in the time-series composite image 71XY is plotted on the vertical axis. In this example, the variance increases with the addition of processing. Thus, variance analysis makes it possible to link process changes. Therefore, it can be expected to have applications such as anomaly detection.

[0054] Furthermore, when additional analysis is required, the computer 60 may accept the designation of an ROI (Region of Interest) for the composite image 71XY and perform analysis only on the accepted ROI. This eliminates the need to process the entire image, significantly reducing the computational load and thereby improving processing speed. It also reduces the use of memory and processing resources, thus greatly reducing the system load. Moreover, by designating an ROI, the influence of noise from areas outside the target region can be reduced, improving the accuracy of the analysis results.

[0055] <2. Modifications> Although embodiments have been described above, the present invention is not limited to those described above, and various modifications are possible.

[0056] In the above embodiment, the background image 53 was fixedly placed on the surface of the structure inside the chamber 10. However, the background image 53 may also be realized by projecting light from outside the chamber 10 onto the surface of the structure inside the chamber 10.

[0057] Furthermore, the above embodiment described an apparatus for performing a cleaning process on the substrate W. However, the apparatus equipped with the fluid observation device may be an apparatus for performing processes other than cleaning. Also, the apparatus equipped with the fluid observation device may be an apparatus for performing processes on objects other than the substrate W.

[0058] However, in a device that supplies processing liquid to a substrate W for precision electronic components such as semiconductor wafers, it is required to control the state of the airflow in the chamber 10 with extreme precision. Therefore, the inclusion of a fluid observation device in a device that supplies processing liquid to the substrate W is particularly significant.

[0059] Although this invention has been described in detail, the above description is illustrative in all respects, and the invention is not limited thereto. It is understood that countless variations not illustrated can be conceived without falling outside the scope of this invention. The components described in each of the above embodiments and variations can be combined or omitted as appropriate, as long as they do not contradict each other.

[0060] 50: Imaging unit 53: Background image 60: Computer S: Observation area

Claims

1. A fluid observation device for observing the flow of a fluid, comprising: an imaging unit that captures a background image through an observation area where the fluid is present; and a computer that analyzes the flow of the fluid using the image acquired by the imaging unit by a background Schlieren method, wherein the background image has a pattern in which the intensity of light and dark changes repeatedly in both the horizontal and vertical directions, and the computer is configured to analyze the flow of the fluid with respect to both the horizontal and vertical directions of the background image.

2. A fluid observation device according to claim 1, wherein the computer synthesizes the analysis results of the fluid flow in both the lateral and vertical directions.

3. A fluid observation device according to claim 1 or claim 2, wherein the background image has a pattern in which the intensity changes periodically in both the horizontal and vertical directions.

4. A fluid observation device according to claim 3, wherein the background image has a pattern composed of two figures of different colors or tones arranged alternately in both the horizontal and vertical directions.

5. A fluid observation device according to claim 1 or claim 2, wherein the background image has a pattern that does not have periodicity in both the horizontal and vertical directions.

Images