JP5656958B2 - Pattern inspection method and pattern inspection apparatus - Google Patents

Description

  The present invention relates to a pattern inspection method and a pattern inspection apparatus using a scanning electron microscope image.

  In a lithography process of a semiconductor manufacturing process, a pattern formed on a photomask is transferred onto a semiconductor wafer (hereinafter referred to as a wafer) using an exposure apparatus. If there is a defect or distortion in the pattern transferred onto the wafer in this way, the semiconductor device may fail or the characteristics may vary.

  Therefore, inspection using a SEM image of a wafer photographed using a scanning scanning electron microscope is performed.

  As one of inspections using such an SEM image, a pattern on design data for manufacturing a photomask (hereinafter referred to as a design pattern) and a pattern on an SEM image (hereinafter referred to as an SEM pattern) There is a way to compare. In this method, the magnitude of the shift of the edge position of the design pattern and the SEM pattern is obtained along the measurement line drawn in the normal direction of the SEM pattern, and the deformation amount of the pattern is evaluated.

  However, if there are fine irregularities on the edge of the SEM pattern, the direction of the measurement line varies, and the deformation amount of the SEM pattern may not be evaluated with good reproducibility.

JP 2008-164593 A

  Therefore, an object of the present invention is to provide a pattern inspection method and a pattern inspection apparatus that can evaluate the deformation amount of an SEM pattern with high reproducibility.

According to one aspect, a step of selecting a comparison candidate pattern from SEM patterns on an SEM image, a step of extracting a design pattern on design data corresponding to the comparison candidate pattern, the design pattern and the comparison candidate pattern A step of setting a plurality of contour points on one of the edges, a step of selecting any one of the plurality of contour points as a measurement point, and (a) the selected measurement Extracting a predetermined number of contour points around the point; (b) determining a regression line for the measurement point and a predetermined number of contour points around the measurement point; and (c) perpendicular to the regression line Obtaining a measurement line passing through the measurement point; and (d) the SEM pattern and design pattern along the measurement line. The steps of (a) to (d) are repeated while sequentially moving the contour points as the measurement points, and the SEM pattern edges and design patterns for all the contour points. Determining a distance from the edge of the pattern.

  According to the pattern inspection method of the above aspect, a contour line as a measurement point and a regression line for a plurality of contour points before and after the measurement point are obtained, and an SEM pattern and a design pattern are measured using a measurement line perpendicular to the regression line. Find the distance between the edges.

  In this way, since the measurement line is obtained using the regression line obtained from a plurality of contour points, it is possible to suppress the variation in the direction of the measurement line even when there are fine irregularities on the edge of the pattern. Thereby, the deformation amount of the SEM pattern can be evaluated with good reproducibility.

FIG. 1 is a block diagram of the pattern inspection apparatus according to the first embodiment. FIG. 2 is a flowchart showing an outline of the pattern inspection method according to the first embodiment. FIG. 3 is a flowchart for explaining a specific procedure of the SEM pattern processing (step S10) of FIG. 4A to 4D are diagrams illustrating an example of a method for selecting a comparison candidate pattern. FIGS. 5A to 5E are diagrams showing a method for extracting a contour from an SEM pattern. FIG. 6 is a flowchart for explaining a specific procedure for processing the design data (step S20) in FIG. FIG. 7A is a diagram showing a design data cut-out method, and FIG. 7B is a diagram showing a method of combining design data fracturing figures. FIGS. 8A to 8C are diagrams showing a design data alignment method. FIG. 9 is a diagram illustrating a method of associating the comparison candidate pattern with the design pattern. FIG. 10A is a diagram showing a shape comparison method (comparative example) based on the normal vector of the edge of the SEM pattern, and FIG. 10B is a diagram showing problems in the method of FIG. It is. FIG. 11 is a flowchart showing how to obtain the deformation amount of the SEM pattern according to the first embodiment. 12A to 12C are views for explaining how to obtain the deformation amount of the SEM pattern based on the method shown in FIG. FIG. 13 is a flowchart showing how to obtain the deformation amount of the SEM pattern according to the second embodiment. 14A to 14C are diagrams showing how to obtain the deformation amount of the SEM pattern based on the method shown in FIG.

  Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

(First embodiment)
FIG. 1 is a block diagram of a scanning electron microscope according to the embodiment.

  The scanning electron microscope 100 is roughly divided into an electronic scanning unit 1 and a control unit 10.

  The electron scanning unit 1 includes an electron gun 3, and electrons are emitted from the electron gun 3 at a predetermined acceleration voltage. The electrons emitted from the electron gun 3 are converged by the condenser lens 4 to become a primary electron beam 3a. The electron beam 3a is deflected by the deflection coil 5 and then focused by the objective lens 6 to be on the surface of the sample 8. Irradiated. The electron scanning unit 1 deflects the electron beam 3 a with the deflection coil 5, thereby scanning the electron beam 3 a within the observation region on the surface of the sample 8.

  Secondary electrons are emitted from the surface of the sample 8 by the irradiation of the primary electron beam 3a, and the emitted secondary electrons are detected by a plurality of electron detectors 9 provided above the sample stage 7.

  The amount of secondary electrons detected by the electron detector 9 is input to the signal processing unit 11 of the control unit 10. The signal processing unit 11 converts the amount of input secondary electrons into a digital amount using an AD converter. Then, the amount of secondary electrons and the deflection position of the primary electron beam 3a by the deflection coil 5 are matched and arranged on a two-dimensional array to generate image data (SEM image).

  The SEM image generated by the signal processing unit 11 is stored in the storage unit 20, and a part of the image is replaced with a luminance signal and displayed on the display unit 22. The storage unit 20 stores design data of the pattern of the sample to be observed along with the SEM image generated by the signal processing unit 11.

  In addition, the control unit 10 includes a design data processing unit 13 that processes the design data into a form that can be easily compared with the SEM image, a SEM image processing unit 12 that performs processing such as contour extraction of an SEM pattern from the SEM image, A comparison processing unit 14 that compares the SEM image is provided.

  Hereinafter, a pattern inspection method using the pattern inspection apparatus 100 shown in FIG. 1 will be described.

  FIG. 2 is a flowchart showing an outline of the pattern inspection method according to the present embodiment.

  In the pattern inspection method of the present embodiment, as shown in FIG. 2, SEM image processing (step S10), design data processing (step S20), and SEM pattern and design pattern comparison processing (step S30) and comparison are performed. Processing proceeds in the order of result display (step S40). Hereinafter, steps S10 to S30 will be further described.

  FIG. 3 is a flowchart showing a specific procedure for processing (step S10) of the SEM image of FIG.

  First, in step S11 of FIG. 3, an SEM image is acquired using the pattern inspection apparatus 100 (see FIG. 1).

  Next, in step S12, the SEM image processing unit 12 of the control unit 10 designates a comparison candidate pattern to be compared with design data from the SEM patterns on the SEM image acquired in step S11.

  4A to 4D are diagrams illustrating an example of a method for specifying a comparison candidate pattern.

  As shown in FIG. 4A, the SEM image 41 acquired in step S11 is displayed on the display unit 22 of the pattern inspection apparatus 100. Reference numerals 42a to 42f in the figure represent SEM patterns on the SEM image. Since these patterns 42a to 42f are made of a material different from that of the other portions, the SEM patterns 42a to 42f are displayed with a lightness (tone) different from that of the other portions in the SEM image 41.

  Here, the comparison candidate pattern is designated using a tone specific to the SEM pattern.

  First, as shown in FIG. 4A, the operator uses the input unit 21 to move the pointer 43 over the SEM pattern 42b to specify the tone of the SEM pattern 42b. Along with this, the SEM image processing unit 21 searches the SEM image 41 for an area having the same tone as the SEM pattern 42b.

  Then, a comparison candidate pattern is designated by the method shown in any of FIGS. 4B to 4D below. In the figure, a pattern highlighted by a thick line represents an SEM pattern designated as a comparison candidate pattern, and a pattern indicated by a broken line represents an SEM pattern that is not a target.

  In the case of FIG. 4B, the SEM image processing unit 12 designates all patterns 42a to 42f having the same tone as the SEM pattern 42b designated by the pointer 43 as comparison candidate patterns.

  In the case of FIG. 4C, the SEM image processing unit 12 designates only the SEM pattern 42b designated by the pointer 43 as a comparison candidate pattern.

  In the case of FIG. 4D, the SEM image processing unit 12 uses the SEM pattern 42b in which the pointer 43 is arranged and the SEM patterns 42a and 42c similar in shape to the SEM pattern 42b as comparison candidate patterns. specify.

  The determination as to whether or not the other SEM pattern is similar in shape to the SEM pattern 42b is made based on, for example, an aspect ratio that is a ratio of the vertical and horizontal sizes of the SEM pattern. Further, it may be determined whether the shapes are similar based on whether the area of the SEM pattern is close instead of the aspect ratio of the SEM pattern.

  As described above, the designation of the comparison candidate pattern in step S12 (see FIG. 3) is completed.

  Next, in step S13 of FIG. 3, the SEM image processing unit 12 performs contour extraction of the comparison candidate pattern. Hereinafter, the outline extraction method will be described by taking the SEM pattern 42b as an example.

  FIGS. 5A to 5E are diagrams showing a method for extracting the contour of the SEM pattern.

  First, as shown in FIG. 5A, a point M included in the SEM pattern 42b for detecting a contour point is set in the SEM image. The point M can be set based on the tone, and the position can be anywhere in the SEM pattern 42b.

  Next, a plurality of lines are drawn while changing the angle as indicated by an arrow C with the point M as a starting point. Then, line profiles along these lines are obtained, respectively, and the edge of the SEM pattern 42b is searched.

Here, as an example, the edge E ₀ on the line profile drawn in the X direction is detected as the first contour point.

Next, as shown in FIG. 5B, a line 81 extending in the X direction is set at a position separated from the edge E _{0 in the} Y direction by a distance twice the designated step ΔL. The designation step ΔL is a parameter for designating the contour point extraction interval.

Subsequently, a line profile along the line 81 is obtained, and the provisional detection edge P ₁ is obtained from the line profile.

Next, as shown in FIG. 5C, a line 82 passing through the edge E ₀ and the provisional detection edge P ₁ is set. Subsequently, a line 83 orthogonal to the line 82 is set along the line 82 at a position separated from the edge E ₀ by the designated step ΔL. In addition, a line 84 orthogonal to the line 82 is set along the line 82 at a position separated from the edge E _{0 by} a distance twice as large as the designated step ΔL.

Thereafter, the edge E ₁ is detected as the second contour point based on the line profile along the line 83. Further, the redetection edge R ₁ is detected based on the line profile along the line 84.

Next, as shown in FIG. 5D, a line 85 passing through the edge E ₁ and the redetection edge R ₁ is set. Subsequently, a line 86 orthogonal to the line 85 is set along the line 85 at a position away from the edge E _{1 by} twice the designated step ΔL. Then, the second temporary detection edge P ₂ is detected from the line profile along the line 86.

Next, as shown in FIG. 5E, a line 87 passing through the edge E ₁ and the provisional detection edge P ₂ is set. Subsequently, a line 88 orthogonal to the line 87 is set along the line 87 at a position separated from the edge E _{1 by} the distance of the designated step ΔL. Based on the line profile along the line 88, the third edge E ₂ is detected.

Further, a line 89 orthogonal to the line 87 is set along the line 87 at a position separated from the edge E _{1 by} a distance twice as large as the designated step ΔL. Then, based on the line profile along the line 89, to detect the second redetection edge R _2.

Thereafter, by repeating the procedure described in FIGS. 5D and 5E, contour points are detected so as to go around the SEM pattern 42b. Finally the edge E _n detected immediately before the edge E ₁ as the final edge, SEM pattern 42b of the contour extraction is completed.

  By the above method, contour points are detected at the edges of the SEM pattern 42b at intervals substantially equal to the designated step ΔL.

  Furthermore, the method described above is repeated for all comparison candidate patterns to obtain contour points.

  The contour points extracted by the above method are collected for each SEM pattern and stored in the storage unit 20 (see FIG. 1) as contour point sequence data.

  Thus, the processing of the SEM image in step S10 in FIG. 2 is completed.

  Next, the design data is processed in step S20.

  FIG. 6 is a flowchart showing a specific procedure for processing the design data in step S20 of FIG. FIG. 7A is a diagram showing the cutout of the design data, and FIG. 7B is a diagram showing the combination of the fracturing figures of the design data.

  First, in step S <b> 21 of FIG. 6, the design data processing unit 13 reads design data corresponding to the SEM image 41 from the storage unit 20.

  As shown in FIG. 7A, the design data 50 may store a wider range of data than the SEM image 41. In such a case, in step S21, the design data processing unit 13 cuts out the design data 51 of a portion indicated by a broken line corresponding to the field of view of the SEM image 41.

  The SEM image 41 may be rotated due to the accuracy of the stage 7 (see FIG. 1) of the pattern inspection apparatus 100.

  Therefore, in step S <b> 22 of FIG. 6, the design data processing unit 13 rotates the design data 51 cut out by the same angle as the rotation angle of the SEM image 41.

  In step S23, the design data processing unit 13 combines the fracturing figures of the design pattern.

  As shown in FIG. 7B, the design pattern 58 having a relatively complicated shape is represented by fracturing (trapezoidal division) by rectangular small regions 58a to 58d. Since the comparison with the contour of the SEM pattern cannot be performed as it is, the fracturing figure is combined as shown in the figure.

  Next, the design data 51 and the SEM image 41 are aligned. This is because the positions of the SEM image 41 and the design data 51 may be slightly shifted depending on the accuracy of the stage 7.

  FIGS. 8A to 8C are diagrams showing a design data alignment method.

  Here, first, in step S24 (see FIG. 6), the design data processing unit 13 extracts, from the design data 51, an alignment pattern that serves as a reference when aligning the position of the design data 51 with the position of the SEM image. .

  This alignment pattern can be determined using, for example, the pointer 43 used when designating the comparison candidate pattern.

  As shown in FIG. 8A, the case where the pointer 43 is arranged on the SEM pattern 42b on the SEM image will be described as an example. In this case, as shown in FIG. 8B, the corresponding pointer 53 is arranged on the design data 51 based on the position coordinates of the pointer 43. Then, the design pattern 52b including the pointer 53 is extracted as an alignment pattern.

  If the pointer 53 is not included in the design pattern, the design pattern closest to the pointer 53 is extracted as the alignment pattern.

  Next, in step S25 (see FIG. 6), the design data processing unit 13 extracts the edge of the alignment pattern 52b, the position of the contour point sequence of the corresponding SEM pattern 42b, and the alignment pattern 52b. The amount of movement of the design data 51 that minimizes the difference in edge position is obtained.

  8C, the design data is moved so as to minimize the positional deviation between the alignment pattern 52b and the corresponding SEM pattern 42b.

  Thus, the alignment of the design data 51 is completed.

  Thereafter, in step S26 of FIG. 6, the design data processing unit 13 associates the comparison candidate pattern with the design pattern.

  FIG. 9 is a diagram illustrating a method of associating the comparison candidate pattern with the design pattern.

  As shown in FIG. 9, pointers 43a, 43b, and 43c are arranged at the centers of SEM patterns 42a, 42b, and 42c designated as comparison candidate patterns on the SEM image 41.

  Next, in the design data 51, the pointers 53a, 53b, and 53c are arranged at the same positions as the position coordinates of the pointers 43a, 43b, and 43c.

  The design patterns 52a, 52b, and 52c including the pointers 53a, 53b, and 53c are associated with the SEM patterns 42a, 42b, and 42c designated as the comparison candidate patterns, respectively.

  Thus, the design data processing in step S20 of FIG. 2 is completed.

  Next, in step S30, the SEM pattern designated as the comparison candidate pattern is compared with the design pattern.

  FIG. 10A is a diagram illustrating a shape comparison method (comparative example) based on the normal vector of the edge of the SEM pattern, and FIG. 10B is a diagram illustrating a method using the normal vector of FIG. It is a figure which shows a problem.

  In the method shown in FIG. 10A, first, the SEM pattern 42b and the corresponding design pattern 52b are overlaid. Then, measurement points are set at predetermined intervals on the edge of the SEM pattern 42b.

  Next, as shown by the arrows in the figure, the normal line of the edge of the SEM pattern 42b is drawn from the measurement point of the SEM pattern 42b, and the edge of the SEM pattern 42b and the edge of the design pattern 52b are aligned along the normal direction. Find the distance.

  Thereafter, by summing the distances between the edges of the SEM pattern 42b and the design pattern 52b obtained by the above method, the amount of deformation of the SEM pattern 42b is quantified and evaluated.

  However, as shown in FIG. 10B, fine unevenness may occur on the edge of the SEM pattern 42b. If there is such unevenness, the normal direction of the edge varies due to local unevenness at the position where the measurement point is arranged. Therefore, even if the position of the measurement point is changed slightly, the total result of the distance between the edges of the SEM pattern 42b and the design pattern 52b changes, and the amount of deformation cannot be obtained with high reproducibility.

  Therefore, in the present embodiment, as described below, the distance between the edges of the SEM pattern 42b and the design pattern 52b is obtained using a plurality of contour points extracted from the SEM pattern 42b instead of the edges of the SEM pattern 42b. .

  FIG. 11 is a flowchart showing a procedure for measuring the deformation amount of the design pattern of the SEM pattern according to this embodiment, and FIG. 12 is a view showing a method for measuring the deformation amount of the SEM pattern.

  The deformation amount of the SEM pattern according to the present embodiment is evaluated by the comparison processing unit 14 (see FIG. 1) of the pattern inspection apparatus 100. First, in step S301 in FIG. 11, the comparison processing unit 14 selects the first SEM pattern 42b for performing deformation amount evaluation from the comparison candidate patterns.

  Next, in step S302, the comparison processing unit 14 acquires the design pattern 52b corresponding to the SEM pattern 42b selected in step S301 and the contour point sequence data of the SEM pattern 42b from the storage unit 20. Then, the design pattern 52b and the outline point sequence are overlaid.

  FIG. 12A shows an example in which the SEM pattern 42b and the design pattern 52b are selected as comparison targets as the first pattern. As illustrated, in step S302, the outline point sequence 47b of the SEM pattern 42b is overlaid with the design pattern 52b.

  Next, in step S303, the comparison processing unit 14 selects the first contour point as the measurement point from the contour point sequence 47b.

  The first contour point may be any contour point in the contour point sequence 47b. In the example of FIG. 12B, the leftmost contour point 200 is selected as the first contour point.

  Next, in step S304, the comparison processing unit 14 extracts four contour points around the contour point selected in step S303.

  In the case of FIG. 12B, two contour points 198 and 199 before the contour point 200 and contour points 101 and 102 behind the contour point 200 are extracted.

  Note that the number of contour points extracted here is not limited to four. The number of contour points extracted is set so that at least the length of the area covered by the measurement points and the contour points extracted in step S304 is greater than the period of the irregularities as noise components of the edges of the SEM pattern. Is done.

  The number of contour points may be set manually based on the observation result of the SEM pattern, or may be automatically obtained and set by the comparison processing unit 14.

  In the case of obtaining automatically, the relationship between the number of contour points to be extracted in advance and the change in the regression line is obtained, and the number of contour points is determined so that the change in the regression line with respect to the increase / decrease in the contour point is less than a certain value Just do it.

  By extracting the number of contour points satisfying such conditions, the direction of a regression line described later is stabilized, and variations in measurement results can be suppressed.

  Next, in step S305, the comparison processing unit 14 obtains a regression line for the contour point selected in step S303 and the contour point extracted in step S304.

  That is, as shown in FIG. 12B, the regression line 48a is obtained based on the position coordinates of the contour points 200, 101, 102, 198, and 199. In this embodiment, this regression line 48a is regarded as the direction of the tangent at the contour point 200 of the pattern 42b. Since the regression line 48a is obtained using a plurality of contour points, it is less susceptible to local unevenness of the edge of the SEM pattern, and variation in the tangential direction is suppressed.

  In the next step S306, the comparison processing unit 14 obtains a measurement line 48b that is perpendicular to the regression line 48a obtained in step S305 and passes through the contour point 200 that is the measurement point selected in step S303.

  In the present embodiment, the measurement line 48b is regarded as a normal line at the contour point 200 of the SEM pattern 42b.

  Next, in step S307, the comparison processing unit 14 obtains an intersection 501 between the measurement line 48b and the edge of the design pattern 52b. Subsequently, the distance between the intersection 501 and the contour point 200 selected in step S303 is obtained.

  Thereby, the distance between the edges of the SEM pattern 42b and the design pattern 52b at the contour point 200 is obtained.

  Next, in step S308, the comparison processing unit 14 determines whether or not the contour point selected as the measurement point is the last.

  If it is determined in step S308 that the contour point selected as the measurement point is not the last contour point (NO), the process proceeds to step S309. Then, as shown in FIG. 12C, the comparison processing unit 14 selects the next contour point 101 as a measurement point.

  Thereafter, the processing described in steps S304 to S307 is performed to obtain the distance between the edges of the SEM pattern 42b and the design pattern 52b for the next contour point 101.

  The comparison processing unit 14 repeats the processes in steps S304 to S307 described above until the process for the last contour point is completed for the measurement point, so that the edges of the SEM pattern 42b and the design pattern 52b are obtained for all the contour points. Find the distance between.

  When the processes of steps S304 to S307 are completed for the last contour point, the comparison processing unit 14 determines YES in step S308, and proceeds to step S310.

  In step S310, the comparison processing unit 14 determines whether or not the calculation of the distance between edges has been completed for all comparison candidate patterns.

  If it is determined in step S310 that the calculation of the distance between edges has not been completed for all the comparison candidate patterns (NO), the process proceeds to step S311 and the comparison processing unit 14 selects the next comparison candidate pattern. To do. And the process of step S302-S309 is performed.

  On the other hand, when it is determined that the calculation of the inter-edge distance has been completed for all the comparison candidate patterns (YES), the comparison processing by the comparison processing unit 14 is ended.

  Thus, the comparison process in step S30 in FIG. 2 is completed.

  Thereafter, in step S40 of FIG. 2, the control unit 10 displays a comparison result between the SEM pattern and the design pattern.

  For example, the control unit 10 calculates the distance between the edges at each contour point of the SEM pattern, obtains the deformation amount of the SEM pattern, and displays the deformation amount together with the SEM pattern on the SEM image.

  Further, an SEM pattern in which the distance between the edges of the SEM pattern and the design pattern exceeds a predetermined value may be colored and displayed on the SEM image as a defect candidate pattern.

  As described above, in the pattern inspection method of the present embodiment, a regression line is obtained for a plurality of contour points on the edge of the SEM pattern, and the SEM pattern and design are designed along the measurement line in the direction perpendicular to the regression line. The distance between edges between the pattern is obtained.

  In this way, by obtaining the distance between edges using a plurality of contour points, even if there are fine irregularities on the edge of the SEM pattern, it is possible to suppress variations in measurement results, and to reduce the deformation amount of the SEM pattern. It can be measured with good reproducibility.

(Second Embodiment)
In the first embodiment, how far the design pattern is far from the SEM pattern is obtained, but in this embodiment, how far the SEM pattern is far from the design pattern is obtained.

  Hereinafter, the pattern inspection method of this embodiment will be described.

  The pattern inspection method of the present embodiment is also performed by the pattern inspection apparatus 100 shown in FIG. 1, and the processes described with reference to FIGS. 2 to 9 except for the comparison process are the same. Description is omitted.

  FIG. 13 is a flowchart illustrating the procedure of the comparison process according to the second embodiment.

  First, as shown in step S321 in FIG. 13, the comparison processing unit 14 selects the first comparison target pattern.

  Next, in step S322, the comparison processing unit 14 extracts the image data of the comparison target pattern (SEM pattern 42b) selected in step S321 and the design pattern 52b from the storage unit 20.

  Since the design pattern 52b extracted here is defined by contour points arranged only at the corners, the number of contour points is small, and accurate measurement cannot be performed as it is.

  Therefore, in step S323, the comparison processing unit 14 supplements the contour points at regular intervals along the edge of the design pattern 52b.

  As a result, as shown in FIG. 14A, a contour point sequence 57b composed of a plurality of contour points is arranged along the edge of the design pattern 52b.

  Next, the process proceeds to step S324, and the comparison processing unit 14 selects the first contour point as the measurement point from the contour point sequence 57b.

  The method of selecting the first contour point is not particularly limited, but in the example of FIG. 14B, the contour point 571 at the lower left corner of the design pattern 52b is selected as a measurement point.

  Next, in step S325, the comparison processing unit 14 extracts a predetermined number of contour points around the contour point 571 selected as the measurement point.

  In the example of FIG. 14B, two contour points 598 and 599 before the contour point 571 that are measurement points and two contour points 572 and 573 behind the contour point 571 are extracted.

  Next, in step S326, the comparison processing unit 14 obtains a regression line for the contour points and measurement points extracted in step S325.

  Thereby, as shown in FIG. 14B, a regression line 58a based on the five contour points 598, 599, 571, 572, and 573 is obtained.

  Next, in step S327, the comparison processing unit 14 calculates a measurement line 58b that is perpendicular to the regression line 58a and passes through the measurement point (contour point 571).

  Next, in step S328, the comparison processing unit 14 obtains the distance between the edges of the design pattern 52b and the SEM pattern 42b along the measurement line 58b.

  Here, as shown in FIG. 14B, an intersection 401 between the measurement line 58b and the edge of the SEM pattern 42b is obtained. And the distance between the edge of the design pattern 52b and the SEM pattern 42b along the measurement line 58b is calculated | required by calculating | requiring the distance of the intersection 401 and the outline point 571. FIG.

  Next, the process proceeds to step S329, and the comparison processing unit 14 determines whether or not the measurement point is the last contour point. Here, in FIG. 14B, the last contour point is the contour point 599 on the near side of the first contour point 571. If the comparison processing unit 14 determines in step S329 that the measurement point is not the last contour point (NO), it is determined that the processing has not been completed for all contour points, and the process proceeds to step S330.

  In step S330, the comparison processing unit 14 selects the next contour point as a measurement point. That is, in the example of FIG. 14C, the contour point 572 immediately after the first contour point 571 is selected as the measurement point. Thereafter, the processes in steps S325 to S328 are repeated to obtain the distance between the edges of the design pattern and the SEM pattern for the contour point 572.

  The above steps S325 to S330 are all repeated until the last contour point 599 is reached.

  When the process for the last contour point 599 is completed, in step S329, the comparison processing unit 14 determines that the measurement point is the last contour point (YES), and proceeds to step S331.

  In step S331, the comparison processing unit 14 determines whether comparison processing with the design pattern has been completed for all comparison candidate patterns.

  If it is determined in step S331 that the comparison process with the design pattern has not been completed for all the comparison candidate patterns, the process proceeds to step S332, the next comparison candidate pattern is selected, and then the process proceeds to step S322.

  Then, the deformation amount of the SEM pattern viewed from the design pattern is obtained for the next comparison candidate pattern.

  On the other hand, if it is determined in step S331 that the comparison process between the design pattern and the SEM pattern is completed for all the comparison candidate patterns, the comparison process is terminated.

  Thus, the comparison process of this embodiment is completed.

  As described above, according to the present embodiment, the amount of deformation of the SEM pattern viewed from the design pattern can be obtained by complementing the contour point on the design pattern side. Also in this embodiment, since the measurement line is obtained using a plurality of contour points, the deformation amount of the SEM pattern with respect to the design pattern can be measured with good reproducibility.

  In the above description, the regression line 58a is obtained using five contour points, but the present embodiment is not limited to this. Since the design data is not affected by noise, the regression line 58a may be obtained using three contour points.

  Further, the number of contour points used for calculating the regression line 58a may be five or more. For example, the regression line 58a is made up of a large number of contour points so that the change of the regression line 58a with respect to the increase / decrease of the contour points is reduced to some extent. It may be calculated.

  DESCRIPTION OF SYMBOLS 1 ... Electron scanning part, 2 ... Chamber, 3 ... Electron gun, 3a ... Electron beam, 4 ... Converging lens, 5 ... Deflector, 6 ... Objective lens, 7 ... Stage, 8 ... Sample, 9 ... Detector, 10 ... Control unit, 11 ... Signal processing unit, 12 ... SEM image processing unit, 13 ... Design data processing unit, 14 ... Comparison processing unit, 20 ... Storage unit, 21 ... Input unit, 22 ... Display unit, 41 ... SEM image, 42a ~ 42f ... SEM pattern, 43, 43a-43c, 53, 53a-53c ... Pointer, 47b, 57b ... Outline point sequence, 48a, 58a ... Regression line, 48b, 58b ... Measurement line, 50, 51 ... Design data, 52a ˜52f, 58 ... design pattern, 58a-58d ... small area, 100 ... scanning electron microscope (pattern inspection apparatus), 101-199, 571-599 ... measurement points, 401, 402, 501, 02 ... intersection.

Claims (6)

Selecting a comparison candidate pattern from among SEM patterns on the SEM image;
Extracting a design pattern on design data corresponding to the comparison candidate pattern;
Setting a plurality of contour points on one of the edges of the design pattern and the comparison candidate pattern;
Selecting any one of the plurality of contour points as a measurement point;
(A) extracting a predetermined number of contour points around the selected measurement point;
(B) obtaining a regression line for the measurement point and a predetermined number of contour points around the measurement point;
(C) obtaining a measurement line perpendicular to the regression line and passing through the measurement point;
(D) obtaining a distance between edges of the SEM pattern and the design pattern along the measurement line;
Steps of obtaining the distance between the edge of the SEM pattern and the edge of the design pattern by repeating the steps (a) to (d) while sequentially moving the contour points as the measurement points;
A pattern inspection method comprising:   The number of contour points extracted from the periphery of the measurement point is such that the length of the portion occupied by the measurement point and the extracted contour point is greater than the period of unevenness as a noise component appearing at the edge of the SEM pattern. The pattern inspection method according to claim 1, wherein the pattern inspection method is set to be long.   Prior to the step of obtaining the distance between the edges, a SEM pattern for positioning and a design pattern corresponding to the SEM pattern for positioning are extracted from the SEM patterns on the SEM image, and these patterns are extracted. 3. The pattern inspection method according to claim 1, wherein the design data or the SEM image is moved so as to minimize a difference in edge position, and the previous SEM pattern and the design pattern are aligned. .   4. The SEM pattern having a portion where a distance between edges of the SEM pattern and the design pattern along the measurement line is a predetermined value or more is detected as a defect candidate pattern. The pattern inspection method according to claim 1. Before selecting as the measurement point,
Determining the relationship between the number of contour points extracted in advance and the change in the regression line;
5. The method of claim 1, further comprising: setting the number of contour points to be extracted so that a change in the regression line with respect to the number of extracted contour points is equal to or less than a predetermined value. The pattern inspection method according to any one of the above. The electronic scanning unit that irradiates the sample surface with an electron beam to capture the SEM image, the SEM pattern of the SEM image, and the design pattern on the design data for producing the sample are compared, and the SEM pattern A pattern inspection method performed by a pattern inspection apparatus including a control unit that detects a deformation amount with respect to design data ,
The control unit selects a comparison candidate pattern from SEM patterns on an SEM image, extracts a design pattern on design data corresponding to the comparison candidate pattern, the design pattern and the comparison candidate pattern A step of setting a plurality of contour points on any one of the edges, a step of selecting any one of the plurality of contour points as a measurement point, and (a) the selected measurement point Extracting a predetermined number of contour points around, (b) determining a regression line for the measurement points and a predetermined number of contour points around the measurement points, and (c) perpendicular to the regression line Determining a measurement line passing through the measurement point; and (d) an error of the SEM pattern and the design pattern along the measurement line. The steps of (a) to (d) are repeated while sequentially moving the contour points as the measurement points and the steps of (a) to (d). And a step of obtaining a distance from the edge to obtain a deformation amount of the design data of the SEM pattern.

Images