Measurement methods, measuring devices, methods for manufacturing original plates, methods for manufacturing semiconductor devices

Abstract

One embodiment aims to provide a measurement method, a measurement device, a method for manufacturing a master plate, and a method for manufacturing a semiconductor device that can improve the measurement accuracy of the feature quantities of a real pattern. [Solution] According to one embodiment, a measurement method is provided. The measurement method includes acquiring image data containing a plurality of real patterns. The measurement method includes setting a plurality of first measurement regions in the image data. The measurement method includes measuring the feature quantities of each of the plurality of first measurement regions using the image data and forming a first array. The measurement method includes acquiring design data containing a plurality of design patterns. The measurement method includes setting a plurality of second measurement regions in the design data. The measurement method includes measuring the feature quantities of each of the plurality of second measurement regions using the design data and forming a second array. The measurement method includes performing statistical processing using the first array and the second array. The measurement method includes matching the plurality of real patterns with the plurality of design patterns according to the results of the statistical processing.

Description

Technical Field

[0001] This embodiment relates to a measurement method, a measurement device, a method for manufacturing a master, and a method for manufacturing a semiconductor device.

Background Art

[0002] In the manufacturing process of a master including a plurality of actual patterns, it may be necessary to measure the feature amount of the actual pattern and evaluate the master according to the measurement result. In the manufacturing process of the master, it is desirable to improve the measurement accuracy of the feature amount of the actual pattern.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] An object of one embodiment is to provide a measurement method, a measurement device, a method for manufacturing a master, and a method for manufacturing a semiconductor device that can improve the measurement accuracy of the feature amount of an actual pattern.

Means for Solving the Problems

[0005] According to one embodiment, a measurement method is provided. The measurement method includes acquiring image data containing a plurality of real patterns. The measurement method includes setting a plurality of first measurement regions in the image data. The measurement method includes measuring the feature quantities of each of the plurality of first measurement regions using the image data and forming a first array. The measurement method includes acquiring design data containing a plurality of design patterns. The measurement method includes setting a plurality of second measurement regions in the design data. The measurement method includes measuring the feature quantities of each of the plurality of second measurement regions using the design data and forming a second array. The measurement method includes performing statistical processing using the first array and the second array. The measurement method includes matching the plurality of real patterns with the plurality of design patterns according to the results of the statistical processing. [Brief explanation of the drawing]

[0006] [Figure 1] A block diagram showing the configuration of a manufacturing system including a measuring device according to the first embodiment. [Figure 2] A flowchart illustrating the operation of a manufacturing system including a measuring device according to the first embodiment. [Figure 3] A block diagram showing the configuration of the measuring device according to the first embodiment. [Figure 4] A diagram showing the configuration of the measurement mechanism in the first embodiment. [Figure 5] A flowchart illustrating the measurement process in the first embodiment. [Figure 6] A diagram showing the configuration of the object to be measured in the first embodiment. [Figure 7] A flowchart illustrating the dimensional data generation process in the first embodiment. [Figure 8] A figure showing image data and contour data in the first embodiment. [Figure 9] A flowchart showing the matching with the drawing data in the first embodiment. [Figure 10] A figure showing the drawing data in the first embodiment. [Figure 11]A diagram showing the collation of frequency distributions in the first embodiment. [Figure 12] A flowchart showing the dimensional data generation process in the second embodiment. [Figure 13] A diagram showing the contour data in the second embodiment. [Figure 14] A flowchart showing the process of matching with drawing data in the second embodiment. [Figure 15] A diagram showing the drawing data in the second embodiment. [Figure 16] A diagram showing the evaluation of correlation in the second embodiment. [Figure 17] A flowchart showing the measurement process in the third embodiment. [Figure 18] A flowchart showing the matching with drawing data in the third embodiment. [Figure 19] A flowchart showing the operation of a manufacturing system including a measuring device according to the fourth embodiment.

Best Mode for Carrying Out the Invention

[0007] The measuring device according to the embodiment will be described in detail below with reference to the accompanying drawings. Note that the present invention is not limited by these embodiments.

[0008] (First Embodiment) The measuring device according to the first embodiment measures the feature amounts of actual patterns in the manufacturing process of a master plate including a plurality of actual patterns, and evaluates the master plate according to the measurement results. However, a device is provided to improve the measurement accuracy of the feature amounts of the actual patterns.

[0009] [[ID=4,1]] For example, the manufacturing system 100 including the measuring device 1 can be configured as shown in FIG. 1. FIG. 1 is a diagram showing the configuration of the manufacturing system 100 including the measuring device 1.

[0010] The manufacturing system 100 includes a circuit design device 101, a layout design device 102, a master plate creation device 103, a coating device 104, an exposure device 105, a development device 106, a processing device 107, a measurement device 1, and a host controller 108. The circuit design device 101 and the layout design device 102 may be implemented on one computer or may be implemented on a plurality of computers connected to be communicable with each other. The master plate creation device 103 may use, for example, an electron beam lithography device. The circuit design device 101, the layout design device 102, the master plate creation device 103, and the measurement device 1 may be connected to be communicable with each other via a communication line (not shown). The coating device 104, the exposure device 105, the development device 106, the processing device 107, the measurement device 1, and the host controller 108 may be connected to be communicable with each other via a communication line (not shown).

[0011] Also, the manufacturing system 100 operates as shown in FIG. 2. FIG. 2 is a flowchart showing the operation of the manufacturing system 100.

[0012] Based on predetermined design information and / or an instruction from a user, the circuit design device 101 performs circuit design (S1), generates schematic data, and supplies it to the layout design device 102.

[0013] Based on the schematic data and / or an instruction from a user, the layout design device 102 performs layout design (S2) and generates layout data. Further, based on predetermined design information and / or an instruction from a user, the layout design device 102 generates drawing data from the layout data and supplies it to the master plate creation device 103 and the measurement device 1, respectively. The master plate is, for example, a mask, a reticle, or the like.

[0014] The master plate creation device 103 draws a plurality of master plate patterns on a master plate substrate according to the drawing data to create a master plate (S3). The created master plate includes a plurality of master plate patterns (a plurality of actual patterns).

[0015] The created master plate is transported from the master plate creation device 103 to the measuring device 1 by a transport system (not shown). The measuring device 1 performs measurement processing on the master plate using the drawing data (S4).

[0016] For example, the measuring device 1 acquires image data containing multiple real patterns. The measuring device 1 may be a CD-SEM (Critical Dimension Scanning Electron Microscope) having an imaging device. The measuring device 1 may acquire image data IM1 containing multiple real patterns by capturing an SEM image of the original plate using the imaging device. The measuring device 1 sets multiple measurement regions (ROIs) 1 in the image data. The measuring device 1 uses the image data IM1 to measure the feature quantities of each of the multiple measurement regions (ROIs) 1 and forms an array AR1. The feature quantities are, for example, dimensions. Each array AR1 contains multiple feature quantities, each associated with the positional information of the measurement region (ROI).

[0017] The measuring device 1 acquires design data containing multiple design patterns from the layout design device 102. The measuring device 1 may also acquire drawing data as design data from the layout design device 102. Using the design data, the measuring device 1 measures the feature quantities of each of the multiple measurement region ROI2 and forms array AR2. The feature quantities are, for example, dimensions. Array AR1 contains multiple feature quantities, each associated with the positional information of the measurement region ROI2.

[0018] The measurement device 1 performs statistical processing using sequences AR1 and AR2. The measurement device 1 may generate a frequency distribution for sequence AR1 and a frequency distribution for sequence AR2. Depending on the results of the statistical processing, the measurement device 1 matches multiple real patterns with multiple design patterns to obtain desired features. For example, the measurement device 1 may compare the frequency distribution of sequence AR1 with the frequency distribution of sequence AR2 and label sequence AR2 according to the comparison result. The measurement device 1 may extract desired features from among the multiple labeled features according to pre-specified conditions.

[0019] The measuring device 1 evaluates the original plate according to the desired feature quantities and determines whether the original plate meets the criteria (S5). The measuring device 1 may determine that the original plate meets the criteria if the dimensions of the actual pattern on the original plate are within the acceptable range. The measuring device 1 may determine that the original plate does not meet the criteria if the dimensions of the actual pattern on the original plate are outside the acceptable range. The measuring device 1 may notify the host controller 108 of the determination result.

[0020] If the original plate does not meet the criteria (No in S5), you may start over from creating the original plate, as shown by the solid arrow in Figure 2.

[0021] The original plate is transported from the measuring device 1 to the peeling device (not shown) by a transport system (not shown) under the control of the host controller 108, where multiple original plate patterns are peeled off from the original plate substrate. The original plate substrate is transported from the peeling device to the original plate creation device 103 by a transport system (not shown), set in the original plate creation device 103, and steps S3 and S4 are performed again.

[0022] Alternatively, if the original design does not meet the criteria (No in S5), the layout design can be redone from scratch, as shown by the dotted arrow in Figure 2.

[0023] The original plate is transported from the measuring device 1 to the peeling device (not shown) by a transport system (not shown) under the control of the host controller 108, where multiple original plate patterns are peeled off from the original plate substrate. The original plate substrate is transported from the peeling device to the original plate creation device 103 by a transport system (not shown), set in the original plate creation device 103, and steps S2, S3, and S4 are performed again.

[0024] Alternatively, if the original design does not meet the criteria (No in S5), the circuit design may be restarted from scratch, as shown by the dashed arrow in Figure 2.

[0025] The original plate is transported from the measuring device 1 to the peeling device (not shown) by a transport system (not shown) under the control of the host controller 108, where multiple original plate patterns are peeled off from the original plate substrate. The original plate substrate is transported from the peeling device to the original plate creation device 103 by a transport system (not shown), set in the original plate creation device 103, and steps S1, S2, S3, and S4 are performed again.

[0026] If the original plate meets the criteria (Yes in S5), multiple actual patterns from the original plate are transferred to the substrate, and exposure, development, processing, etc. are performed to create the substrate (S6).

[0027] The master plate is transported from the measuring device 1 to the exposure device 105 by a transport system (not shown) and set on the master plate stage of the exposure device 105.

[0028] Meanwhile, the coating apparatus 104 coats a photosensitive material (e.g., resist) onto a substrate (e.g., wafer). The substrate coated with the photosensitive material is transported from the coating apparatus 104 to the exposure apparatus 105 by a transport system (not shown) and placed on the substrate stage of the exposure apparatus 105. The exposure apparatus 105 uses a projection optical system to image the exposure light that has been illuminated by the illumination optical system and passed through the master plate, or the exposure light that has been reflected by the master plate, onto the substrate, transferring multiple real patterns on the master plate to the photosensitive material on the substrate to form a latent image.

[0029] After exposure, the substrate is transported from the exposure apparatus 105 to the developing apparatus 106 by a transport system (not shown). The developing apparatus 106 develops the latent image in the photosensitive material on the substrate. As a result, a pattern corresponding to the design information (drawing data) is developed in each shot area on the substrate 2.

[0030] After development, the substrate is transported from the developing device 106 to the processing device 107 by a transport system (not shown). The processing device 107 uses the developed photosensitive pattern as a mask to perform predetermined processing on the substrate. This forms a pattern in each shot area on the substrate according to the design information (drawing data).

[0031] Furthermore, the measurement process (S4) may be performed after the substrate fabrication (S6). In this case, some of the substrates to be fabricated may be selected and subjected to the measurement process (S4).

[0032] Next, the configuration of the measuring device 1 will be explained using Figure 3. Figure 3 is a block diagram showing the configuration of the measuring device 1.

[0033] The measuring device 1 is a device capable of measuring the shape of a pattern formed on a master substrate by scanning it with an electron beam. Specifically, the measuring device 1 determines the scanning angle of the electron beam relative to the master substrate, scans the master substrate with the electron beam at that scanning angle, and measures the shape of the pattern based on the resulting image. A measuring device 1 capable of performing such measurement processing is configured, for example, as a CD-SEM. An electron beam is an example of a charged particle.

[0034] The measuring device 1 includes a measuring mechanism 11, a controller 2, and a storage unit 28.

[0035] Controller 2 includes an image data acquisition unit 21, a contour point data generation unit 22, an extraction point data generation unit 23, an angle calculation unit 24, a scanning angle determination unit 25, an analysis unit 26, and a determination unit 27. Each part of Controller 2 may be implemented entirely in software, entirely in hardware, or partially in software and partially in hardware. The storage unit 28 may be a storage device such as an SSD or HDD.

[0036] The controller 2 can control the measurement mechanism 11 to perform measurement processing (S4). The controller 2 can store the image data IM1 obtained in the measurement processing (S4) in the storage unit 28. The controller 2 may be implemented as a system-on-a-chip (SoC).

[0037] The measurement mechanism 11 is a physical and mechanical component of the measurement device 1. The measurement mechanism 11 irradiates the original substrate with an electron beam at a predetermined scanning angle. As the predetermined scanning angle, the measurement mechanism 11 can use, for example, a reference 0°, or a scanning angle determined by the scanning angle determination unit 25, which will be described later.

[0038] Figure 4 is a schematic diagram showing an example of the configuration of the measurement mechanism 11. As shown in Figure 4, the measurement mechanism 11 includes a microscope tube 111 in which an electron gun 121 is installed as an irradiation source for an electron beam EB as a charged particle, a sample chamber 112 in which the original plate substrate S is placed, and a control unit that controls each part of the measurement mechanism 11.

[0039] The microscope tube 111 is cylindrical in shape. The microscope tube 111 has a closed upper end and an open lower end to allow the electron beam EB to pass through. The sample chamber 112 is configured to accommodate the original substrate S. The microscope tube 111 and the sample chamber 112 are assembled in an airtight seal. The inside of the microscope tube 111 and the sample chamber 112 are configured to be maintained under reduced pressure by a pump or the like (not shown).

[0040] Inside the microscope tube 111, the electron gun 121, focusing lens 131, objective lens 132, coil 141, and detector 151 are installed in order from near the top end.

[0041] The electron gun 121 irradiates an electron beam EB downward into the lens barrel 111. The electron beam EB emitted from the electron gun 121 travels along the long axis of the lens barrel 111.

[0042] The focusing lens 131 is an electromagnetic coil wound concentrically around the optical axis of the lens barrel 111, and focuses the electron beam EB using a magnetic field.

[0043] The objective lens 132 is an electromagnetic coil wound concentrically around the optical axis of the lens barrel 111, and focuses the electron beam EB emitted toward the original substrate S by the magnetic field.

[0044] Coil 141 is a pair of electromagnetic coils used to deflect the focused electron beam EB or to correct astigmatism, and is arranged symmetrically with respect to the optical axis of the lens barrel 111. Coil 141 causes the focused electron beam EB to scan the device area R of the original substrate S at a predetermined scanning angle. The predetermined scanning angle is, for example, the scanning angle determined by the scanning angle determination unit 25, which will be described later. Coil 141 can also cause the electron beam EB to scan the original substrate S multiple times.

[0045] The detector 151 detects secondary electrons or backscattered electrons obtained as a result of scanning the device region R at a predetermined scanning angle. The detector 151 includes an imaging device 151a. The detector 151 uses the imaging device 151a to acquire the two-dimensional intensity distribution of secondary electrons or backscattered electrons and images the device region R. This generates an image of the device region R.

[0046] A stage 161 on which the original substrate S is placed is installed in the sample chamber 112. An actuator 162 is attached to the stage 161, and the stage 161 is configured to be driven forward, backward, left, and right. When the stage 161 is driven, the desired device region R on the original substrate S can be observed.

[0047] Next, the details of the measurement process (S4) by the measuring device 1 will be explained using Figure 5. Figure 5 is a flowchart showing the operation of the measuring device 1.

[0048] When the original substrate S is placed on the stage 161, the measuring device 1 performs alignment between the original substrate S and the stage 161 (S11). The measuring device 1 may perform alignment using the mark MK provided on the original substrate S and the mark 161a provided on the stage 161.

[0049] For example, the original substrate S has a device region DR and a peripheral region PR, as shown in Figure 6. Figure 6 is a diagram showing the configuration of the object to be measured.

[0050] Multiple master patterns (multiple actual patterns) are formed in the device region DR. If the master is transmissive, the multiple actual patterns are, for example, patterns of a light-shielding film on the master substrate. If the master is reflective, the multiple actual patterns are, for example, patterns of bumps and depressions on the master substrate.

[0051] Multiple marks MK are formed in the peripheral region PR. Figure 6 illustrates four marks MK, but the number of marks MK may be three or fewer, or five or more, as long as alignment is possible. Figure 6 illustrates a configuration in which marks MK are formed at the four corners of the peripheral region PR, but the formation positions of the marks MK may be other positions in the peripheral region PR, as long as alignment is possible.

[0052] As shown in Figure 4, the stage 161 may have a mark 161a around the area where the original substrate S is placed.

[0053] The measuring device 1 includes an imaging device 1a (not shown), and by imaging the mark MK on the original substrate S and the mark 161a on the stage 161 with the imaging device 1a, the relative position of the original substrate S with respect to the stage 161 can be recognized. Depending on the recognition result, the measuring device 1 may determine the deviation ΔPOS of the original substrate S from a reference position and control the actuator 162 to cancel the deviation ΔPOS and drive the stage 161 in the XYZ direction.

[0054] This allows the position of the original substrate S to be aligned to the reference position in the measuring device 1.

[0055] When the measuring device 1 receives a measurement request from the host controller 108, it moves to the measurement coordinates (S12). The measurement request includes multiple measurement coordinates and the conditions to be guaranteed. The measuring device 1 obtains multiple measurement coordinates from the measurement request and selects the measurement coordinate of interest from among the multiple measurement coordinates. The measuring device 1 may determine the amount of movement ΔMV in the XYZ direction to the measurement coordinates, starting from the reference position aligned in S11. The measuring device 1 may drive the stage 161 in the XYZ direction by controlling the actuator 162 to move the original substrate S by the amount of movement ΔMV.

[0056] When the device moves to the measurement coordinates, the measuring device 1 performs a dimension data generation process (S13). The dimension data generation process measures the dimensions of the actual pattern included in the measurement area ROI1 and generates the measurement results as numerical dimension data.

[0057] In the dimension data generation process (S13), steps S21 to S27 shown in Figure 7 may be performed. Figure 7 is a flowchart of the dimension data generation process.

[0058] The measuring device 1 acquires an SEM image of the original plate (S21). The measuring device 1 may also acquire image data IM1 as shown in Figure 8(a) by capturing an SEM image of the original plate using an imaging device. Figure 8 is a diagram showing image data and contour data.

[0059] The image data IM1 shown in Figure 8(a) has a two-dimensional correspondence between pixel position and brightness information for multiple pixels. In Figure 8(a), the horizontal direction may be considered the X direction and the vertical direction may be considered the Y direction. The pixel position at the center of the image data IM1 may be considered the origin of the pixel coordinates.

[0060] Image data IM1 contains multiple real patterns. In Figure 8(a), a line-and-space pattern is exemplified as one of the multiple real patterns, in which multiple lines, each extending in the Y direction, are repeatedly arranged in the X direction with space in between.

[0061] The measurement device 1 acquires contour data IM2 of multiple real patterns using image data IM1 (S22). The measurement device 1 may acquire contour data IM2 as shown in Figure 8(b) by performing edge detection processing on the image data IM1. Known methods such as the Canny method can be used for edge detection processing. Figure 8(b) shows the contour data IM2 of the measurement region ROI1. In Figure 8(b), the horizontal axis represents the X coordinate and the vertical axis represents the Y coordinate.

[0062] The contour data IM2 shown in Figure 8(b) has a two-dimensional correspondence between pixel positions and contour information for multiple pixels. The contour data IM2 corresponds to the image data IM1.

[0063] The contour data IM2 contains multiple contour patterns corresponding to multiple real patterns. If the multiple real patterns are line and space patterns, the multiple contour patterns may also be line contour and space patterns. The line contour indicates the contour of the line pattern.

[0064] The measuring device 1 sets multiple measurement regions (ROIs) 1 for the contour data IM2 (S23). The measuring device 1 may set multiple measurement regions (ROIs) 1, as shown by the dotted lines in Figure 8(c), according to the multiple contour patterns. If the multiple contour patterns are line contours and space patterns, the measuring device 1 may set a measurement region (ROI) 1 for each of the multiple line contours.

[0065] The measuring device 1 may set each of the multiple measurement regions (ROIs) 1 with a shape corresponding to the contour pattern. If the contour pattern is a line contour, the measuring device 1 may set a line-shaped measurement region (ROI) 1.

[0066] The measuring device 1 may identify and retain the central pixel position for each of the multiple measurement regions of interest (ROIs) 1.

[0067] When multiple measurement regions (ROIs) are set, the measurement device 1 selects the measurement region ROI to be processed from among the multiple measurement regions ROIs (S24).

[0068] The measuring device 1 generates dimensional data using the contour pattern of the measurement region ROI 1 to be processed (S25). The measuring device 1 may determine the line width by taking the difference between the pixel positions of adjacent edges in the X direction in the contour pattern of the measurement region ROI 1 to be processed. The measuring device 1 may also determine the line width at multiple different Y positions and take their average. In this way, the measuring device 1 can generate the line width as dimensional data.

[0069] The measuring device 1 stores the generated dimensional data (S26). The measuring device 1 may also store dimensional data by generating or updating array AR1, which contains dimensional data in a form associated with the position information of the measurement region ROI1. When the position information of the measurement region ROI1 is given, array AR1 returns dimensional data. The dimensional data includes, for example, line width.

[0070] If the measuring device 1 finds any unprocessed measurement region ROI1 among the multiple measurement regions ROI1 (Yes in S27), it returns the processing to S24.

[0071] If there are no unprocessed measurement regions (ROIs) among the multiple measurement regions (ROIs) (Yes in S27), the measurement device 1 terminates the process.

[0072] Once the dimension data generation process (S13) is completed, the measuring device 1 performs a comparison with the drawing data (S14). This comparison with the drawing data can be rephrased as a comparison between image data containing multiple actual patterns and drawing data containing multiple design patterns, and can be considered as a matching between multiple actual patterns and multiple design patterns.

[0073] In the comparison with the drawing data (S14), steps S31 to S37 shown in Figure 9 may be performed. Figure 9 is a flowchart of the comparison with the drawing data.

[0074] The measuring device 1 acquires drawing data (S31). The measuring device 1 may acquire drawing data LD1 from the layout design device 102 as shown in Figure 10(a). Figure 10 is a diagram showing the drawing data.

[0075] The drawing data LD1 shown in Figure 10(a) has a two-dimensional correspondence between pixel positions and drawing information for multiple pixels. In Figure 10(a), the horizontal direction may be considered the X direction and the vertical direction may be considered the Y direction. The pixel position at the center of the drawing data LD1 may be considered the origin of the pixel coordinates.

[0076] The drawing data LD1 contains multiple design patterns. In Figure 10(a), a line-and-space pattern is exemplified as one of the multiple design patterns, in which multiple lines, each extending in the Y direction, are repeatedly arranged in the X direction through space. Each design pattern is associated with a design value for line width. In Figure 10(a), light hatched lines are associated with line width LW1. Dark hatched lines are associated with line width LW3. Medium hatched lines are associated with line width LW2.

[0077] The measuring device 1 sets multiple measurement regions (ROIs) 2 for the drawing data LD1 (S32). The measuring device 1 may set multiple measurement regions (ROIs) 2, as shown enclosed by dotted lines in Figure 10(b), according to multiple design patterns. If the multiple design patterns are line-and-space patterns, the measuring device 1 may set a measurement region (ROI) 2 for each of the multiple design patterns. The measuring device 1 may identify and retain the central pixel position for each of the multiple measurement regions (ROIs).

[0078] When multiple measurement regions (ROIs) are set, the measurement device 1 selects the measurement region ROI2 to be processed from among the multiple measurement regions ROIs (S33).

[0079] The measuring device 1 generates dimensional data using the design pattern of the measurement region ROI2 to be processed (S34). The measuring device 1 may also obtain the design value of the associated line width in the design pattern of the measurement region ROI2 to be processed. This allows the measuring device 1 to generate the line width as dimensional data.

[0080] The measuring device 1 stores the generated dimensional data (S35). The measuring device 1 may also store dimensional data by generating or updating an array AR2 containing dimensional data in a form associated with the position information of the measurement region ROI2. When the position information of the measurement region ROI2 is given to array AR2, it returns dimensional data. The dimensional data includes, for example, line width.

[0081] If the measuring device 1 finds any unprocessed measurement region ROI2 among the multiple measurement region ROI2 (Yes in S36), it returns the processing to S33.

[0082] If there are no unprocessed measurement regions ROI2 among the multiple measurement regions ROI2 (Yes in S36), the measurement device 1 performs statistical processing using sequences AR1 and AR2. The measurement device 1 generates a frequency distribution for sequence AR1, generates a frequency distribution for sequence AR2, and performs frequency distribution matching processing to compare the frequency distribution of sequence AR1 and the frequency distribution of sequence AR2 (S37).

[0083] For example, the measuring device 1 sequentially provides positional information of multiple measurement regions (ROIs) 1 to the array AR1, extracts dimensional data from multiple measurement regions (ROIs), and adds it to the frequency-dimensional data value plane. This allows the measuring device 1 to generate a frequency distribution of the array AR1 as shown in Figure 11(a). Figure 11 shows the matching of the frequency distributions.

[0084] In Figure 11(a), the vertical axis represents frequency, and the horizontal axis represents dimensional data values. In the example in Figure 11(a), three distributions are generated, SD1, SD2, and SD3, starting from the lowest dimensional data value. Each distribution, SD1 to SD3, has a bell-shaped spread. The center of each distribution, SD1 to SD3, can be represented by the dimensional data value with the maximum frequency. The width of each distribution, SD1 to SD3, can be represented by the half-width, which corresponds to the difference in dimensional data values ​​where the frequency is half of the maximum.

[0085] The measuring device 1 sequentially provides positional information of multiple measurement regions (ROIs) of array AR2, extracts dimensional data from multiple measurement regions (ROIs), and adds it to the frequency-dimensional data value plane. This allows the measuring device 1 to generate a frequency distribution of array AR2 as shown in Figure 11(b).

[0086] In Figure 11(b), the vertical axis represents frequency, and the horizontal axis represents dimensional data values. In the example in Figure 11(b), three distributions are generated, SD11, SD12, and SD13, starting from the lowest dimensional data value. Each of the distributions SD11 to SD13 is linear and does not spread out, showing the dimensional data value at a pinpoint. Each of the distributions SD11 to SD13 shows line widths LW1, LW2, and LW3, starting from the lowest dimensional data value.

[0087] The measuring device 1 may compare the frequency distribution of sequence AR1 with the frequency distribution of sequence AR2 and label the frequency distribution of sequence AR2 according to the comparison result.

[0088] The measuring device 1 compares the frequency distribution shown in Figure 11(a) with the frequency distribution shown in Figure 11(b) and confirms that both agree in that three distributions, SD1 to SD3 and SD11 to SD13, exist. Accordingly, as shown in Figure 11(c), the measuring device 1 can assign line widths LW1, LW2, and LW3 to distribution SD1, SD2, and SD3, respectively, as representative values ​​of the dimensional data shown by distribution SD3, starting from the lowest dimensional data value. In other words, the measuring device 1 can label distribution SD1, SD2, and SD3 with line widths LW1, LW2, and LW3.

[0089] Once the matching with the drawing data (S14) is complete, the measuring device 1 performs data extraction (S15). The measuring device 1 extracts the conditions to be guaranteed from the measurement request received in S12. The measuring device 1 extracts a group of dimensional data values ​​corresponding to the conditions to be guaranteed in the frequency distribution of array AR2. For example, if the conditions to be guaranteed are the lowest-frequency group of dimensional data values, the measuring device 1 can extract the distribution SD1 shown enclosed by the dotted line in Figure 11(c) as the group of dimensional data values ​​corresponding to the conditions to be guaranteed.

[0090] The measuring device 1 determines whether all of the multiple measurement coordinates obtained from the measurement request have been measured (S16). If there are any measurement coordinates that have not been measured (No in S16), the measuring device 1 returns to S12; if there are no measurement coordinates that have not been measured (Yes in S16), the process ends.

[0091] As described above, in the first embodiment, in the measurement method, the feature quantities of each of the multiple measurement region ROI1 set in image data containing multiple real patterns are measured, and array AR1 is formed. The feature quantities of each of the multiple measurement region ROI2 set in design data containing multiple design patterns are measured, and array AR2 is formed. Statistical processing is performed using arrays AR1 and AR2, and matching of the multiple real patterns and multiple design patterns is performed according to the results of the statistical processing. This makes it possible to numerically match the measurement of feature quantities of real patterns with design data, thereby improving the measurement accuracy of feature quantities of real patterns.

[0092] (Second embodiment) Next, the measurement method according to the second embodiment will be described. The following description will focus on the differences from the first embodiment.

[0093] In the first embodiment, a comparison with plotting data using a frequency distribution is illustrated, while in the second embodiment, a comparison with plotting data using a correlation coefficient is illustrated.

[0094] In the dimensional data generation process (S13) of the measurement process (S4), the following differences from the first embodiment are observed, as shown in Figure 12. Figure 12 is a flowchart of the dimensional data generation process in the second embodiment.

[0095] After steps S21 and S22 are performed in the same manner as in the first embodiment, the measuring device 1 sets a window WD1 and a plurality of measurement regions ROI1 for the contour data IM2 (S41).

[0096] The measuring device 1 may set a window WD1 for the contour data IM2, as shown in Figure 13, enclosed by a dashed line. The X width of window WD1 is smaller than the X width of the contour data IM2. The Y width of window WD1 is smaller than the Y width of the contour data IM2. Window WD1 may be set to include multiple contour patterns.

[0097] The measuring device 1 may set multiple measurement regions (ROIs) 1 within the window WD1. The measuring device 1 may set each of the multiple measurement regions (ROIs) 1 in a shape corresponding to the window WD1. If the window WD1 is set in a rectangular shape, the measuring device 1 may set a rectangular measurement region (ROI) 1.

[0098] Figure 13 illustrates a case where 3 x 3 = 9 measurement regions (ROIs) are set within window WD1, but more measurement regions (ROIs) may be set within window WD1. Multiple rectangular sections of unit length may be set as measurement regions (ROIs) within window WD1.

[0099] Subsequently, steps S24 to S27 are performed in the same manner as in the first embodiment.

[0100] In the measurement process (S4), the matching with the drawing data (S14) is performed in the following ways, as shown in Figure 14, which differs from the first embodiment. Figure 14 is a flowchart of the matching process with the drawing data in the second embodiment.

[0101] After S31 is performed in the same manner as in the first embodiment, the measuring device 1 sets a window WD2 and a plurality of measurement regions ROI2 for the drawing data LD1 (S41).

[0102] The measuring device 1 may set a window WD2 for the drawing data LD1, as shown in Figure 15(a) enclosed by a dashed line. The measuring device 1 may set window WD2 with a shape and size corresponding to window WD1. The X width of window WD2 is smaller than the X width of the drawing data LD1. The Y width of window WD2 is smaller than the Y width of the drawing data LD1. Window WD2 may be set to include multiple design patterns.

[0103] The measuring device 1 may set multiple measurement regions (ROIs) 2 within the window WD2. The measuring device 1 may set each of the multiple measurement regions (ROIs) 2 in a shape corresponding to the window WD2. If the window WD2 is set in a rectangular shape, the measuring device 1 may set a rectangular measurement region (ROI) 2.

[0104] Figure 15(a) illustrates a case where 3 x 3 = 9 measurement regions (ROIs) are set within window WD2, but more measurement regions (ROIs) may be set within window WD2. Multiple unit-length rectangular sections may also be set as measurement regions (ROIs) within window WD2.

[0105] Subsequently, steps S24 to S27 are performed in the same manner as in the first embodiment.

[0106] The measuring device 1, in order to observe the correlation with the drawing data IM2, fixes the position of window WD1 in the contour data IM2 while offsetting the position of window WD2 in the drawing data LD1. This is equivalent to offsetting the position of window WD2 relative to the position of window WD1.

[0107] The measuring device 1 may provide multiple offset amounts as candidates for the offset amount of window WD2. The measuring device 1 may also provide combinations of multiple X offset amounts and multiple Y offset amounts. For example, the measuring device 1 provides six X offset amounts ΔX0, ΔX1, ΔX2, ΔX3, ΔX4, ΔX5. The measuring device 1 also provides six Y offset amounts ΔY0, ΔY1, ΔY2, ΔY3, ΔY4, ΔY5. This provides 6 × 6 = 36 offset amounts. These offset amounts may satisfy the following equations 1 and 2. ΔX0 (=0) < ΔX1 < ΔX2 < ΔX3 < ΔX4 < ΔX5 ... Formula 1 ΔY0 (=0) < ΔY1 < ΔY2 < ΔY3 < ΔY4 < ΔY5 ... Formula 2

[0108] Furthermore, the number of offset quantities is not limited to this number; any number can be used as long as it is appropriate for examining the correlation.

[0109] When multiple offset values ​​are provided, the measuring device 1 selects the offset value to be processed from among the multiple measurement regions of interest (ROI) 2 (S52).

[0110] For example, if X offset amount ΔX0 and Y offset amount ΔY0 are selected, as shown in Figure 15(a), window WD2 is placed at the initial position in the plotting data LD1, and multiple measurement regions of interest (ROIs) 2 are placed within window WD2.

[0111] Alternatively, if an X offset amount ΔX1 and a Y offset amount ΔY0 are selected, as shown in Figure 15(b), the window WD2 is positioned at a location offset by ΔX1 in the X direction from the initial position in the drawing data LD1, and multiple measurement regions of interest (ROIs) 2 are placed within that window WD2.

[0112] Alternatively, if an X offset amount ΔX0 and a Y offset amount ΔY1 are selected, as shown in Figure 15(c), the window WD2 is positioned at a location offset by ΔY1 in the Y direction from the initial position in the drawing data LD1, and multiple measurement regions of interest (ROIs) 2 are placed within that window WD2.

[0113] After steps S33 to S34 are performed in the same manner as in the first embodiment, the measuring device 1 stores the generated dimensional data (S53). The measuring device 1 may store the dimensional data by generating or updating an array AR2 which includes the dimensional data in a manner associated with the X offset amount of window WD2, the Y offset amount of window WD2, and the relative position information of the measurement area ROI2 in window WD2. The array AR2 returns the dimensional data when given the X offset amount of window WD2, the Y offset amount of window WD2, and the relative position information of the measurement area ROI2 in window WD2. The dimensional data includes, for example, the line width.

[0114] S36 is performed in the same manner as in the first embodiment.

[0115] If the measuring device 1 finds any unprocessed offset amounts among the multiple offset amounts (Yes in S53), it returns the process to S52.

[0116] If there are no unprocessed offset amounts among the multiple offset amounts (No in S53), the measuring device 1 performs statistical processing using sequences AR1 and AR2. The measuring device 1 performs a correlation evaluation to determine the correlation coefficient between sequences AR1 and AR2 (S54). The measuring device 1 may also determine the correlation coefficient between sequences AR1 and AR2 for each of the multiple positional offset amounts.

[0117] For example, measuring device 1 selects the offset amount to be processed from among multiple offset amounts. The offset amount to be processed includes the X offset amount and the Y offset amount to be processed.

[0118] The measuring device 1 sequentially provides positional information of multiple measurement regions (ROIs) in window WD1 to array AR1 and extracts dimensional data for the multiple measurement regions (ROIs).

[0119] The measuring device 1 provides the array AR1 with the X offset amount and the Y offset amount to be processed, and sequentially provides the position information of multiple measurement regions ROI2 in window WD2, and extracts the dimensional data of the multiple measurement regions ROI2.

[0120] The measuring device 1 calculates the average value AV1 of the dimensional data from multiple measurement regions ROI1. The measuring device 1 calculates the average value AV2 of the dimensional data from multiple measurement regions ROI2.

[0121] The measuring device 1 calculates the deviation DE1 of the dimensional data of multiple measurement regions ROI1. The measuring device 1 calculates the deviation DE2 of the dimensional data of multiple measurement regions ROI2.

[0122] The measuring device 1 divides the sum of the squares of the deviations DE1 by the number of dimensional data points to obtain the variance σ1 of the dimensional data of multiple measurement regions ROI1. 2The measurement device 1 calculates the variance σ² of the dimensional data of multiple measurement regions ROI2 by dividing the sum of the squares of the deviations DE2 by the number of dimensional data points. 2 We seek.

[0123] The measuring device 1 measures the variance σ1 2 The positive square root of is taken to obtain the standard deviation σ1 of the dimensional data of multiple measurement regions ROI1. The measurement device 1 uses variance σ2 2 Take the positive square root of the given value to obtain the standard deviation σ2 of the dimensional data for multiple measurement regions (ROIs).

[0124] The measuring device 1 divides the product of deviations DE1 and DE2 by the number of dimensional data points and calculates the covariance σ between the dimensional data of multiple measurement areas ROI1 and the dimensional data of multiple measurement areas ROI2. 12 We seek.

[0125] The measuring device 1 measures the covariance σ 12 Divide this value by the product of the standard deviations σ1 and σ2 to obtain the correlation coefficient r between the dimensional data of multiple measurement regions ROI1 and the dimensional data of multiple measurement regions ROI2.

[0126] The measuring device 1 plots the correlation coefficient and Y offset amount as a data series of X offset amounts on the Y offset amount plane.

[0127] The measuring device 1 repeatedly performs these processes while changing the offset amount to be processed, which is selected from among several offset amounts. This allows the measuring device 1 to generate a graph plotting the correlation coefficients for multiple offset amounts, as shown in Figure 16. Figure 16 shows the evaluation of the correlation.

[0128] By referring to the graph in Figure 16, the offset amount (X offset amount, Y offset amount pair) corresponding to the highest correlation coefficient will indicate the set of window WD1 and window WD2 positions with the highest correlation. This allows us to identify the window WD1 and window WD2 positions with the highest correlation.

[0129] In Figure 16, the correlation between the positions of window WD1 and window WD2 is highest when the X offset amount = ΔX2 and the Y offset amount = ΔY3.

[0130] The measuring device 1 may label multiple real patterns based on correlation coefficients. The measuring device 1 may also label multiple real patterns according to the offset amount of a position determined based on correlation coefficients for offset amounts of multiple positions.

[0131] For example, measuring device 1 acquires the contour pattern within window WD1. Measuring device 1 offsets window WD2 with the offset amount (a pair of X offset amount and Y offset amount) that has the highest correlation coefficient, and acquires the design pattern within window WD2. Measuring device 1 can label the contour pattern within window WD1 by comparing the contour pattern within window WD1 with the design pattern within window WD2.

[0132] As described above, in the second embodiment, in the measurement method, the feature quantities of each of the multiple measurement region ROI1 set in image data containing multiple real patterns are measured, and array AR1 is formed. The feature quantities of each of the multiple measurement region ROI2 set in design data containing multiple design patterns are measured, and array AR2 is formed. Statistical processing is performed using arrays AR1 and AR2, and matching of the multiple real patterns and multiple design patterns is performed according to the results of the statistical processing. This makes it possible to numerically match the measurement of feature quantities of real patterns with design data, thereby improving the measurement accuracy of feature quantities of real patterns.

[0133] (Third embodiment) Next, the measurement method according to the third embodiment will be described. The following description will focus on the differences from the first and second embodiments.

[0134] The first and second embodiments illustrate numerical matching, while the third embodiment illustrates how to differentiate between the use of numerical matching and image-based matching.

[0135] In the measurement process (S4), the following differences from the first embodiment are observed, as shown in Figure 17. Figure 17 is a flowchart of the measurement process in the third embodiment.

[0136] After S11 and S12 are performed in the same manner as in the first embodiment, the measuring device 1 refers to the drawing data and determines whether there is a unique pattern or a single design dimension pattern among the multiple design patterns included in the drawing data (S61). The measuring device 1 can determine that there is no unique pattern among the multiple design patterns if the multiple design patterns are regularly repeating patterns such as line and space patterns. The measuring device 1 can determine that there is a unique pattern among the multiple design patterns if the multiple design patterns are irregular patterns such as lines or holes placed near circuit elements. The measuring device 1 can determine that there is a single design dimension pattern if the multiple design patterns are repeating patterns of the same pattern.

[0137] If there is no unique pattern among the multiple design patterns (No in S61), the measuring device 1 performs dimensional data generation processing (S13).

[0138] In the dimension data generation process (S13), the process shown in Figure 18 may be performed. Figure 18 is a flowchart of the dimension data generation process in the third embodiment.

[0139] The measuring device 1 refers to the drawing data and determines whether or not there is line width variation in the multiple design patterns included in the drawing data (S71). If the multiple design patterns are line-and-space patterns in which lines extending in the Y direction are repeatedly arranged in the X direction, the measuring device 1 can determine that there is line width variation in the multiple design patterns if the design line width changes when the lines are viewed in the Y direction. If the measuring device 1 maintains a constant design line width when the lines are viewed in the Y direction, the measuring device 1 can determine that there is no line width variation in the multiple design patterns.

[0140] If there is no line width variation in the multiple design patterns included in the drawing data (No in S71), the measuring device 1 performs the processing shown in Figure 7 in the same manner as in the first embodiment (S72).

[0141] If there is a variation in line width among the multiple design patterns included in the drawing data (Yes in S71), the measuring device 1 performs the processing shown in Figure 12 in the same manner as in the second embodiment (S73).

[0142] Once the dimension data generation process (S13) is complete, the measuring device 1 performs a comparison with the drawing data (S14).

[0143] In the matching with the drawing data (S14), processing as shown in Figure 19 may be performed. Figure 19 is a flowchart of the matching with the drawing data in the third embodiment.

[0144] The measuring device 1 refers to the drawing data and determines whether or not there is line width variation in the multiple design patterns included in the drawing data (S81). If the multiple design patterns are line-and-space patterns in which lines extending in the Y direction are repeatedly arranged in the X direction, the measuring device 1 can determine that there is line width variation in the multiple design patterns if the design line width changes when the lines are viewed in the Y direction. The measuring device 1 can determine that there is no line width variation in the multiple design patterns if the design line width is constant when the lines are viewed in the Y direction.

[0145] If there is no line width variation in the multiple design patterns included in the drawing data (No in S81), the measuring device 1 performs the processing shown in Figure 9 in the same manner as in the first embodiment (S92).

[0146] If there is a variation in line width among the multiple design patterns included in the drawing data (Yes in S81), the measuring device 1 performs the processing shown in Figure 12 in the same manner as in the second embodiment (S83).

[0147] Once the matching with the drawing data (S14) is complete, the measuring device 1 performs S15 in the same manner as the first measured diameter.

[0148] If the measurement device 1 finds a unique pattern among multiple design patterns (Yes in S61), it performs pattern matching (PTM), which is image-based matching (S62).

[0149] In pattern matching (S62), the measuring device 1 acquires an SEM image of the original plate (S21) as image data IM1, similar to the process shown in Figure 7, and uses the image data IM1 to acquire contour data IM2 of multiple real patterns (S22). In the contour data IM2, for example, XY coordinates with the center as the origin are set. The contour data IM2 includes multiple contour patterns.

[0150] The measuring device 1 acquires drawing data LD1 in the same manner as shown in Figure 9 (S31). In the drawing data LD1, for example, XY coordinates with the center as the origin are set. The drawing data LD1 includes multiple design patterns.

[0151] The measuring device 1 uses the pixel positions of the edges of the contour pattern in the contour data IM2 and the pixel positions of the edges of the design pattern in the drawing data LD1 to determine the amount of deviation between the origin of the XY coordinates of the contour data IM2 and the origin of the XY coordinates of the drawing data LD1.

[0152] The measuring device 1 may control the coil 141 to correct the deflection direction of the electron beam EB in order to cancel out the determined amount of deviation, thereby shifting the SEM image. Alternatively, the measuring device 1 may control the actuator 162 to move the stage 161 in the XY direction in order to cancel out the determined amount of deviation.

[0153] This allows the measuring device 1 to perform pattern matching between the contour pattern in the contour data IM2 and the design pattern in the drawing data LD1.

[0154] Once pattern matching (S62) is complete, the measuring device 1 takes an SEM image of the original plate (S63), acquires it as image data IM1, and performs length measurement to measure the dimensions of multiple actual patterns contained in the image data IM1 (S64).

[0155] The measuring device 1 determines whether all of the multiple measurement coordinates obtained from the measurement request have been measured (S16). If there are any measurement coordinates that have not been measured (No in S16), the measuring device 1 returns to S12; if there are no measurement coordinates that have not been measured (Yes in S16), the process ends.

[0156] As described above, in the third embodiment, in the measurement method, if there is no unique pattern among the multiple design patterns, numerical matching is performed, and if there is a unique pattern among the multiple design patterns, image-based matching is performed. This makes it possible to use numerical matching and image-based matching interchangeably.

[0157] While several embodiments of the present invention have been described, these embodiments are presented as examples only and are not intended to limit the scope of the invention. These novel embodiments can be carried out in a variety of other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims of the invention and its equivalents. [Explanation of Symbols]

[0158] 1 measuring device, 2 controllers, 11 measuring mechanisms.

Claims

1. Obtaining image data containing multiple real patterns, Setting a plurality of first measurement regions in the aforementioned image data, Using the aforementioned image data, the feature quantities of each of the plurality of first measurement regions are measured to form a first array. Obtaining design data that includes multiple design patterns, Setting multiple second measurement areas in the aforementioned design data, Using the aforementioned design data, the characteristic quantities of each of the plurality of second measurement regions are measured to form a second array. Performing statistical processing using the first sequence and the second sequence, Based on the results of the statistical processing, the multiple real patterns and the multiple design patterns are matched to obtain the desired feature quantities. Measurement methods including

2. Performing the aforementioned statistical processing means To generate the frequency distribution of the first sequence, To generate the frequency distribution of the second sequence, including The measurement method according to claim 1.

3. The matching process to obtain the desired feature quantities is Comparing the frequency distribution of the first sequence and the frequency distribution of the second sequence, Depending on the comparison results, the frequency distribution of the second sequence is labeled, including The measurement method according to claim 2.

4. Performing the aforementioned statistical processing means This includes determining the correlation coefficient between the first sequence and the second sequence. The measurement method according to claim 1.

5. The matching process to obtain the desired feature quantities is This includes labeling the multiple real patterns based on the correlation coefficient. The measurement method according to claim 4.

6. The setting of the plurality of second measurement areas is performed for each of the offset amounts of a plurality of different positions between the plurality of first measurement areas and the plurality of second measurement areas. Performing the aforementioned statistical processing means This includes determining the correlation coefficient between the first array and the second array for each of the offset amounts of the plurality of positions. The measurement method according to claim 4.

7. The matching process to obtain the desired feature quantities is The offset amount of the position is determined based on the correlation coefficient with respect to the offset amounts of the plurality of positions, Labeling is performed on the plurality of actual patterns according to the offset amount of the determined position, including The measurement method according to claim 6.

8. Measurement mechanism, A controller that controls the aforementioned measurement mechanism, Equipped with, The aforementioned controller, Acquire image data containing multiple real patterns, A plurality of first measurement regions are set in the aforementioned image data, Using the aforementioned image data, the feature quantities of each of the plurality of first measurement regions are measured, and a first array is formed. Acquire design data that includes multiple design patterns, A plurality of second measurement areas are set in the aforementioned design data, Using the design data, the feature quantities of each of the plurality of second measurement regions are measured, and a second array is formed. Statistical processing is performed using the first sequence and the second sequence. Based on the results of the statistical processing, the multiple real patterns and the multiple design patterns are matched to obtain the desired feature quantities. Measuring device.

9. Creating a master version that includes multiple actual patterns, The original plate is imaged to obtain image data including the multiple actual patterns, Setting a plurality of first measurement regions in the aforementioned image data, Using the aforementioned image data, the feature quantities of each of the plurality of first measurement regions are measured to form a first array. Obtaining design data that includes multiple design patterns, Setting multiple second measurement areas in the aforementioned design data, Using the aforementioned design data, the characteristic quantities of each of the plurality of second measurement regions are measured to form a second array. Performing statistical processing using the first sequence and the second sequence, Based on the results of the statistical processing, the multiple real patterns and the multiple design patterns are matched to obtain the desired feature quantities. The original plate is evaluated according to the desired features, A method for manufacturing the original plate, including the plate itself.

10. Creating a master version that includes multiple actual patterns, The original plate is imaged to obtain image data including the multiple actual patterns, Setting a plurality of first measurement regions in the aforementioned image data, Using the aforementioned image data, the feature quantities of each of the plurality of first measurement regions are measured to form a first array. Obtaining design data that includes multiple design patterns, Setting multiple second measurement areas in the aforementioned design data, Using the aforementioned design data, the characteristic quantities of each of the plurality of second measurement regions are measured to form a second array. Performing statistical processing using the first sequence and the second sequence, Based on the results of the statistical processing, the multiple real patterns and the multiple design patterns are matched to obtain the desired feature quantities. The original plate is evaluated according to the desired features, Depending on the evaluation results, the multiple actual patterns on the original plate are transferred to the substrate, A method for manufacturing a semiconductor device containing [a specific component].

Images