JP2012233733A - Checking method and checking device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a checking device and a checking method that can accurately determine, without being affected by any distortion of the mask, whether or not a pattern is acceptable.SOLUTION: At least four arbitrary coordinate points on a mask are measured (S101). Then, the differences between at least these four coordinate points and respective coordinate points of design data corresponding at least these four coordinate points are determined, and then one point with the greatest difference is selected (S102). Next, it is determined whether or not the difference between the selected one point and the coordinate point of any design data surpasses a predetermined threshold (S103). If it does not surpass the threshold, at least all the four coordinate points are used to align an optical image and a reference image with each other (S104). If it surpasses the threshold, this point is eliminated and alignment is accomplished by using the remaining at least three points (S105).

Description

  The present invention relates to an inspection method and an inspection apparatus, and more particularly to an inspection method and an inspection apparatus used for detecting a defect of a pattern formed on an inspection target such as a mask.

  In recent years, the circuit line width required for a semiconductor element has become increasingly narrower as the large scale integrated circuit (LSI) is highly integrated and has a large capacity. The semiconductor element uses an original pattern pattern (a mask or a reticle, which will be collectively referred to as a mask hereinafter) on which a circuit pattern is formed, and the circuit is exposed and transferred onto a wafer by a reduction projection exposure apparatus called a stepper. Manufactured by forming. For manufacturing a mask for transferring such a fine circuit pattern onto a wafer, an electron beam drawing apparatus capable of drawing the fine pattern is used. Attempts have also been made to develop a laser beam drawing apparatus for drawing using a laser beam. The electron beam drawing apparatus is also used when drawing a circuit pattern directly on a wafer.

  Yield improvement is indispensable for the manufacture of LSIs that require a large amount of manufacturing costs. On the other hand, recent typical logic devices are required to form patterns with a line width of several tens of nanometers. Here, as a major factor for reducing the yield, there are mask pattern defects and variations in process conditions during exposure transfer. Until now, with the miniaturization of the LSI pattern dimension formed on a semiconductor wafer, it has been performed to absorb the variation margin of process conditions by increasing the dimensional accuracy of the mask. For this reason, in mask inspection, the dimensions that must be detected as pattern defects are miniaturized, and it is necessary to detect position errors of extremely small patterns. For this reason, high inspection accuracy is required for inspection apparatuses that detect defects in transfer masks used in LSI manufacturing.

  Defect detection methods include a die-to-die inspection method and a die-to-database inspection method. The die-to-die inspection method compares the same patterns of chips with different masks when a plurality of chips having the same pattern configuration are arranged in a part or the whole of the same mask. Inspection method. According to this method, since the mask patterns are directly compared, highly accurate inspection can be performed with a relatively simple apparatus configuration. However, it is impossible to detect a defect that exists in common in both patterns to be compared. On the other hand, the die-to-database inspection method is an inspection method for comparing reference data generated from design data used for mask manufacture and an actual pattern on the mask. Since a mechanism for generating a reference image is required, the apparatus becomes large, but a strict comparison with design data can be performed. This method can only be used when there is only one chip in one mask.

  In die-to-database inspection, light emitted from a light source is irradiated onto a mask to be inspected via an optical system. The mask is placed on a table, and light irradiated as the table moves scans the mask. The light transmitted or reflected through the mask forms an image on the image sensor via the lens, and the optical image captured by the image sensor is sent to the comparison unit as measurement data. In the comparison unit, the measurement data and the reference data are compared according to an appropriate algorithm. If these data do not match, it is determined that there is a defect (see, for example, Patent Document 1 or Patent Document 2).

JP 2008-112178 A JP 2006-266864 A

  In the mask inspection apparatus described in Patent Document 2, an optical image is acquired while moving the stage in the X direction and the Y direction with the mask placed, and compared with a reference image. When the optical image and the reference image are compared, alignment for aligning the positions of both images is performed. However, the mask dimensions do not necessarily match the design dimensions, and each mask has its own distortion. Conventionally, such distortion has been dealt with by performing processing such as trapezoidal correction.

  In recent years, as the pattern becomes finer, the difference in distortion between the alignment mark located at the edge of the mask and the LSI pattern located at the center of the mask cannot be ignored. However, the extent of this difference can only be determined after the mask has been inspected. For this reason, for example, if a part with an alignment mark is extremely distorted compared to the other, a trapezoidal correction process is dragged to the alignment mark, and it is determined that the pattern is a defective pattern even though it is a normal LSI pattern. There was a problem. In addition, if the alignment pattern and the LSI pattern are similarly distorted, there is a problem that the LSI pattern that should originally be determined to be defective is regarded as a non-defective product.

  The present invention has been made in view of these points. That is, the present invention is to provide an inspection method and an inspection apparatus that can accurately determine the quality of a pattern regardless of the influence of mask distortion.

  Other objects and advantages of the present invention will become apparent from the following description.

The first aspect of the present invention includes illuminating light on a sample on which a pattern is formed, forming an image of the sample on an image sensor via an optical system, and obtaining an optical image of the sample;
Creating a reference image from pattern design data;
Measuring the coordinates of at least four points on the sample, obtaining a difference from each coordinate of the design data corresponding to these points, and selecting one point having the largest difference,
When the difference between the selected point and the coordinates of the design data exceeds the threshold value, if it is less than the threshold value, the optical image and the reference image are aligned using the at least four points, and the threshold value is exceeded The present invention relates to an inspection method characterized by aligning an optical image and a reference image at at least three points excluding the selected one point.

The first aspect of the present invention is the step of forming at least four types of quadrilateral or more polygons with at least three points selected from the at least four points and theoretical coordinates obtained from these at least three points;
Comparing each theoretical coordinate in at least four types of quadrilateral or more polygons with each coordinate of design data corresponding to each of the at least four points to obtain a displacement amount;
Of the at least four points, it is preferable that the point corresponding to the point with the largest amount of positional deviation is one point with the largest difference.

The first aspect of the present invention is the step of forming at least four types of quadrilateral or more polygons with at least three points selected from the at least four points and theoretical coordinates obtained from these at least three points;
Compare each area of the at least four types of polygons of the quadrangle or more with the area of the polygons of the quadrangle or more formed from the coordinates of the design data corresponding to each of the at least four points. Identifying a polygon greater than or equal to a large square,
Of the at least four points, the point corresponding to the theoretical coordinates of the specified polygon or more polygon is preferably set to one point having the largest difference.

According to a first aspect of the present invention, after aligning the optical image and the reference image, comparing these images to detect pattern defects,
When the number of defects exceeds a predetermined value, the displacement amount in the X direction, the displacement amount in the Y direction, and the distance from the coordinates obtained from the design data are obtained for each defect, and the pattern displacement of the entire sample is constant. And determining whether or not the tendency is recognized,
If it is determined that there is a certain tendency in the pattern deviation of the entire sample, the optical image and the reference image may be aligned after returning to the process of detecting the pattern defect after correcting this deviation. preferable.

  When the pattern is formed by charged particle beam drawing, after aligning the optical image and the reference image, the actual pattern data at the first and last stages of drawing and the corresponding design data are compared. When these differences exceed a predetermined threshold value, it is determined that the mask is an NG mask, and the subsequent steps are preferably stopped. When the difference is equal to or less than the predetermined threshold value, it is preferable to detect a defect in the entire pattern. When comparing the actual pattern data in the first and last stages of drawing with the corresponding design data, there may be several points for each of the first and last stages, and all points are compared at certain intervals. It doesn't matter.

In addition, for each defect, the amount of deviation in the X direction, the amount of deviation in the Y direction, and the distance from the coordinates obtained from the design data are obtained, and the defect is constant when a certain deviation tendency is recognized in at least one of these. Having a step of determining that it has a tendency to shift,
When the ratio of defects determined to have a certain tendency to shift exceeds a predetermined threshold, it is preferable to determine that a certain tendency is observed in the pattern deviation of the entire sample.

According to a second aspect of the present invention, a portion on which a sample on which a pattern is formed is illuminated with light, and an image of the sample is formed on an image sensor via an optical system to obtain an optical image of the sample;
Create a reference image from the design data of the pattern,
From the coordinates of at least four points on the sample, the difference between the coordinates of the design data corresponding to these points is obtained and one point having the largest difference is selected, and the difference between the selected one point and the coordinates of the design data is selected. And a portion for determining whether or not the threshold value exceeds a threshold value.

  ADVANTAGE OF THE INVENTION According to this invention, the inspection method and inspection apparatus which can determine the quality of a pattern correctly irrespective of the influence of the distortion of a mask are provided.

6 is a diagram for explaining an optical image acquisition procedure according to Embodiment 1. FIG. 3 is a flowchart illustrating an alignment method according to the first embodiment. 6 is a flowchart showing another alignment method according to the first embodiment. 10 is a flowchart showing another alignment method of the first embodiment. It is a figure explaining the process (S302) of FIG. 1 is a system configuration diagram of an inspection apparatus according to Embodiment 1. FIG. It is a conceptual diagram which shows the data flow of the inspection apparatus of FIG. 6 is a diagram for describing filter processing according to Embodiment 1. FIG. 6 is a flowchart illustrating an alignment method according to the second embodiment. 10 is a flowchart showing a method for determining whether or not a characteristic tendency is observed in the defects frequently generated in the step (S404) of FIG. 6 is a schematic diagram showing ideal lattice points in Embodiment 2. FIG. In Embodiment 2, it is a figure which shows typically the relationship between an actual defect coordinate and the theoretical coordinate of the defect location calculated | required from design data.

Embodiment 1 FIG.
In this embodiment mode, a mask used in photolithography is used as an inspection target. However, the present invention is not limited to this.

  The mask is placed at a predetermined position on an XYθ stage provided so as to be movable in the horizontal direction and the rotation direction by motors of XYθ axes. When the pattern formed on the mask is irradiated with light from above the XYθ table, the light transmitted through the mask is imaged as an optical image on the photodiode array, then photoelectrically converted by the photodiode array, and further sensor circuit A / D (analog / digital) conversion is performed.

  FIG. 1 is a diagram illustrating an optical image acquisition procedure. As shown in this figure, the inspection area on the mask 101 is virtually divided into a plurality of strip-shaped inspection stripes 20 having a scan width W in the Y direction. The mask 101 is placed on an XYθ table (not shown), and the operation of the XYθ table is controlled so that each inspection stripe 20 is continuously scanned. The optical image is acquired by moving the XYθ table and continuously inputting the image having the scan width W to the photodiode array.

  For example, an image in the first inspection stripe 20 is input to the photodiode array while the XYθ table moves in the X direction. Next, as the XYθ table moves in the −X direction, the image having the scan width W is continuously input to the photodiode array in the same manner for the second inspection stripe 20. The third inspection stripe 20 is in the direction opposite to the direction in which the image in the second inspection stripe 20 is acquired (−X direction), that is, in the direction in which the image in the first inspection stripe 20 is acquired (X direction). The XYθ table is input to the photodiode array while moving. Thus, an optical image is efficiently acquired by inputting an image continuously.

  The optical image output from the sensor circuit is sent to the comparison circuit together with data indicating the position of the mask 101 on the XYθ table. In the comparison circuit, alignment between the optical image and a reference image serving as a model is performed. Here, the actual mask dimensions do not necessarily match the design dimensions. For this reason, processing for correcting the distortion of the mask is required, but due to the miniaturization of the pattern, the difference in distortion between the alignment mark located at the edge of the mask and the LSI pattern located at the center of the mask cannot be ignored. It has become a thing.

  For example, if a part with an alignment mark is extremely distorted compared to the other, correction processing is dragged to the alignment mark, and it may be determined as a defective pattern even though it is a normal LSI pattern. Alternatively, if the alignment pattern and the LSI pattern are similarly distorted, the LSI pattern that should be originally determined to be defective may be regarded as a non-defective product.

  FIG. 2 is a flowchart showing the alignment method in the present embodiment.

  First, the coordinates of any four points on the mask are measured (S101). Here, the four points to be measured are preferably located outside the inspection region which is the peripheral portion of the mask, but may be located within the inspection region where the LSI pattern is formed.

  Next, for these four points, a difference (position shift amount) from each coordinate of the corresponding design data is obtained, and one point having the largest difference is selected (S102). Next, it is determined whether or not the difference (positional deviation amount) between the selected point and the coordinates of the design data exceeds a preset threshold value (S103). If it is less than or equal to the threshold value, the optical image and the reference image are aligned using all four points (S104). On the other hand, if the threshold value is exceeded, this point is excluded and alignment is performed using the remaining three points (S105).

  For example, the following method can be considered as a method for selecting one point having the maximum difference (positional deviation amount) in S102.

  The four points whose coordinates are measured are point A, point B, point C, and point D, respectively. Then, an arbitrary three points are selected from these to form a quadrangle. Here, the coordinates of the fourth point serving as the apex of the rectangle are theoretical coordinates calculated using the three measured points. The quadrangle can be a square, a rectangle, a parallelogram, or the like.

  For example, three points of point A, point B, and point C are selected, and a fourth point coordinate D ′ is calculated from these points. Next, a point B, a point C, and a point D are selected, and a fourth theoretical coordinate A ′ is calculated therefrom. Similarly, a point C, a point D, and a point A are selected, and a fourth theoretical coordinate B ′ is calculated therefrom. Further, the point D, the point A, and the point B are calculated, and the fourth theoretical coordinate C ′ is calculated therefrom.

  The four theoretical coordinates A ′, B ′, C ′ and D ′ obtained as described above are compared with the coordinates of the design data corresponding to each of the measurement points A, B, C and D. That is, the amount of positional deviation between the design data coordinates of the point A and the theoretical coordinates A ′ is obtained. Similarly, the positional deviation amount between the design data coordinate of the point B and the theoretical coordinate B ′, the coordinate of the design data of the point C and the theoretical coordinate C ′, and the coordinate of the design data of the point D and the theoretical coordinate D ′ is also obtained. . For example, when the positional deviation amount between the design data coordinate of the point A and the theoretical coordinate A ′ is the largest among the obtained four positional deviation amounts, the point A is selected in S102.

  Further, as a method for selecting one point having the largest difference in S102, the following may be mentioned.

  Similarly to the above, the four points whose coordinates are measured are referred to as point A, point B, point C, and point D, respectively. Then, three points are selected from these to form a quadrangle. The fourth coordinate is a theoretical coordinate calculated using three points, as described above. The quadrangle can be a square, a rectangle, a parallelogram, or the like.

  For example, the quadrangle 1 'is formed using the point A, the point B, the point C, and the fourth theoretical coordinate D' calculated therefrom. Next, a quadrangle 2 ′ is formed using the points B, C and D and the fourth theoretical coordinate A ′ calculated from these points. Similarly, a quadrangle 3 'is formed by using the point C, the point D and the point A and the fourth theoretical coordinate B' calculated therefrom. Further, a quadrilateral 4 'is formed by using the point D, the point A and the point B and the fourth theoretical coordinate C' calculated therefrom.

  The quadrilaterals formed from the coordinates of the design data corresponding to the points A, B, C and D, respectively, for the areas of the quadrilaterals 1 ′, 2 ′, 3 ′ and 4 ′ obtained as described above. Compare with the area. Then, a rectangle having the largest area difference is specified. The measurement point corresponding to the theoretical coordinates among the vertices of the rectangle is the point selected in S102. For example, when the difference between the square 1 'and the area of the square formed from the design data is the maximum, the point D is selected in S102.

  According to the method shown in FIG. 2, alignment can be performed except for an alignment mark at an extremely distorted portion. Thereby, it is possible to avoid the influence of a specific alignment mark from acting strongly on the correction process. That is, it is possible to reduce a situation in which a normal LSI pattern is determined to be a defective pattern due to a strong effect of a specific alignment mark even though it is originally a process for correcting mask distortion. . Even if the alignment pattern and the LSI pattern are similarly distorted, the quality of the LSI pattern can be determined without being affected by the alignment pattern.

  In the example of FIG. 2, the coordinates of the four points are measured in S101, but the present invention is not limited to this, and there may be at least four measurement points. That is, the measurement coordinates can be 5 points or more. Even when there are five or more points, alignment can be performed in the same manner as in FIG.

  FIG. 3 is a flowchart showing another alignment method in the present embodiment.

  First, the coordinates of any four points on the mask are measured (S201). Here, the four points to be measured are preferably located outside the inspection region which is the peripheral portion of the mask, but may be located within the inspection region where the LSI pattern is formed.

  Next, for these four points, differences (positional deviation amounts) from the coordinates of the corresponding design data are obtained (S202).

  Next, the difference (positional deviation amount) obtained in S202 is compared with a preset threshold value to determine how many points the difference is equal to or less than the threshold value. In order to perform alignment between the optical image and the reference image, coordinates of at least two points are necessary. Therefore, it is determined whether or not the number of the four points that are equal to or smaller than the threshold is 2 to 4 ( S203).

  If the number of points below the threshold is 2 to 4, the process proceeds to S204 and alignment is performed. For example, if two points are equal to or less than the threshold value, the optical image and the reference image are aligned using these two points. If three points are equal to or less than the threshold value, alignment is performed using these three points. If four points are equal to or less than the threshold value, alignment is performed using all of the measurement points.

  On the other hand, if the number of objects that are equal to or smaller than the threshold value is not 2 to 4 in S203, that is, if all of the measurement points exceed the threshold value or only one point is equal to or smaller than the threshold value, alignment may be performed. Can not. In this case, the process proceeds to S205 to cancel the alignment.

  Also by the method shown in FIG. 3, the alignment can be performed except for the alignment mark at the extremely distorted portion. Therefore, as in the example of FIG. 2, it is possible to avoid the influence of a specific alignment mark from acting strongly on the correction process, so that the situation where a normal LSI pattern is determined as a defective pattern can be reduced. it can. Furthermore, even if the alignment pattern and the LSI pattern are similarly distorted, the quality of the LSI pattern can be determined without being affected by the alignment pattern.

  On the other hand, in the example of FIG. 2, one point with the largest difference (a point with the largest positional deviation amount) is selected, and whether or not the difference between the selected point and the design data exceeds a preset threshold value. judge. If it is below the threshold, all four points are used to align the optical image with the reference image. If the threshold is exceeded, the selected one point is excluded and the remaining three points are used for alignment. Do. That is, in this method, since alignment is performed using at least three points, alignment with high accuracy can be performed. On the other hand, in the example of FIG. 3, it is not necessary to select one point having the largest difference (a point at which the positional deviation amount is maximum), and therefore, the entire process can be simplified.

  In the example of FIG. 3, the coordinates of four points are measured in S201. However, since there are at least four measurement coordinates, it is possible to set five or more points. Even when there are five or more points, alignment can be performed in the same manner as in FIG.

  FIG. 4 is a flowchart showing another alignment method in the present embodiment.

  In this method, first, arbitrary four points on the mask, that is, coordinates of point A, point B, point C, and point D are measured (S301). These points are preferably located outside the inspection area, which is the periphery of the mask, but may be located within the inspection area where the LSI pattern is formed.

Next, the quadrangle S ₁ having the four points as vertices is formed, and the four sides of the quadrangle S ₁ and the coordinates of the design data corresponding to the points A, B, C, and D are formed. The four sides of the rectangle S ₀ are compared. Then, a pair whose corresponding sides are closest is specified. Further, to identify the two points of the above four points forming one side of the rectangle S ₁ constituting the pair (S302).

S302 will be described with reference to FIG. Each vertex of the rectangle _{S 1} shown by the solid line, A point measured in S301, the point B, and point C and the point D. A quadrangle S ₀ indicated by a dotted line is formed from the coordinates of design data corresponding to each of the points A, B, C, and D. And four sides of the square S _1, is compared with the four sides of the square S _0, it can be seen that the sides AB of the rectangle S ₁ is overlaps with one side of the square S _0. From this, the point A and the point B are specified in S302.

The two points specified in S302 are reference points in the example of FIG. In S303, the amount of deviation from the ideal positions of the remaining two points is obtained with respect to this reference point. For example, in FIG. 5, the reference points are point A and point B. Therefore, for each of the points C and D, determining the amount of deviation from the corresponding points in the rectangle S _0. The shift amount, for example, can be defined respectively as side BC and angle theta ₁ with the corresponding square S ₀ of the sides, the angle theta ₂ between the sides of the rectangle S ₀ corresponding to the side AD.

  Next, it is determined whether or not the two deviation amounts obtained in S303 exceed a threshold value (S304). When both the deviation amounts are equal to or less than the threshold value, alignment is performed using all four points measured in S301 (S305). Further, even when only one of them exceeds the deviation amount, alignment is performed using the two points specified in S302 and the point whose deviation amount is equal to or smaller than the threshold (S305). On the other hand, when both the deviation amounts exceed the threshold value, alignment is performed using only the two points specified in S302 (S306).

For example, in FIG. 5, it is assumed that the angle θ ₁ exceeds the threshold value, but the angle θ ₂ is equal to or less than the threshold value. In this case, the optical image and the reference image are aligned using the point A, the point B, and the point D.

  Also by the method shown in FIG. 4, the alignment can be performed except for the alignment mark at the extremely distorted portion. Thereby, it is possible to avoid the influence of a specific alignment mark from acting strongly on the correction process. That is, it is possible to reduce a situation in which a normal LSI pattern is determined to be a defective pattern due to a strong influence of a specific alignment mark, although it is originally a process of correcting mask distortion. Even if the alignment pattern and the LSI pattern are similarly distorted, the quality of the LSI pattern can be determined without being affected by the alignment pattern.

  In the example of FIG. 4, the coordinates of four points are measured in S301. However, since there are at least four measurement coordinates, the measurement coordinates can be five or more. Even when there are five or more points, alignment can be performed in the same manner as in FIG.

  Next, the inspection apparatus and inspection method of this embodiment will be described.

  FIG. 6 is a system configuration diagram of the inspection apparatus according to the present embodiment. As shown in this figure, the inspection apparatus 100 includes an optical image acquisition unit A and a control unit B.

  The optical image acquisition unit A includes a light source 103, an XYθ table 102 that can move in the horizontal direction (X direction, Y direction) and the rotation direction (θ direction), an illumination optical system 170 that constitutes a transmissive illumination system, and magnification optics. A system 104, a photodiode array 105, a sensor circuit 106, a laser length measurement system 122, and an autoloader 130 are included.

  In the control unit B, a control computer 110 that controls the entire inspection apparatus 100 is controlled by a position circuit 107, a first comparison circuit 108, a reference circuit 112, a development circuit 111, and an autoloader control via a bus 120 serving as a data transmission path. The circuit 113, the table control circuit 114, magnetic disk devices 109a and 109b, which are examples of storage devices (storage units), a magnetic tape device 115, a flexible disk device 116, a CRT 117, a pattern monitor 118, and a printer 119 are connected. The XYθ table 102 is driven by an X-axis motor, a Y-axis motor, and a θ-axis motor controlled by the table control circuit 114. As these motors, for example, step motors can be used.

  Design data serving as database-based reference data is stored in the magnetic disk device 109a, read out as the inspection progresses, and sent to the development circuit 111. In the development circuit 111, the design data is converted into image data (bit pattern data). Thereafter, this image data is sent to the reference circuit 112 to be used for generating reference data.

  In FIG. 6, constituent components necessary for the present embodiment are described, but other known components necessary for inspecting the mask may be included.

  FIG. 7 is a conceptual diagram showing a data flow in the present embodiment.

  As shown in FIG. 7, CAD data 201 created by a designer (user) is converted into design intermediate data 202 in a hierarchical format such as OASIS. The design intermediate data 202 stores design pattern data created for each layer and formed on each mask. Here, in general, the inspection apparatus is not configured to directly read OASIS data. That is, unique format data is used for each manufacturer of the inspection apparatus. For this reason, the OASIS data is input to the inspection apparatus 100 of FIG. 6, for example, after being converted into format data 203 unique to each inspection apparatus for each layer. In this case, the format data 203 can be data unique to the inspection apparatus 100, but can also be data compatible with the drawing apparatus.

  The format data 203 is input to the magnetic disk device 109a of FIG. That is, the design pattern data used when forming the pattern of the mask 101 is stored in the magnetic disk device 109a.

  The figure included in the design pattern is a basic figure of a rectangle or a triangle. The magnetic disk device 109a includes information such as coordinates (x, y) at the reference position of the graphic, side length, graphic code that serves as an identifier for identifying graphic types such as a rectangle and a triangle, and each pattern graphic. Graphic data defining the shape, size, position, etc.

  Furthermore, a set of figures existing in a range of about several tens of μm is generally called a cluster or a cell, and data is hierarchized using this. In the cluster or cell, arrangement coordinates and repeated description when various figures are arranged alone or repeatedly at a certain interval are also defined. The cluster or cell data is further arranged in a strip-like region called a frame or stripe having a width of several hundred μm and a length of about 100 mm corresponding to the entire length in the X direction or Y direction of the mask.

  The input design data is read from the magnetic disk device 109a by the development circuit 111 through the control computer 110.

  The expansion circuit 111 expands the design pattern to data for each graphic, and interprets a graphic code, a graphic dimension, and the like indicating the graphic shape of the graphic data. Then, binary or multivalued design image data is developed as a pattern arranged in a grid having a grid with a predetermined quantization size as a unit. The developed design image data calculates the occupancy ratio of the figure in the design pattern for each area (square) corresponding to the sensor pixel. And the figure occupation rate in each pixel becomes a pixel value.

  The design data converted into binary or multi-value image data (design image data) as described above is then sent to the reference circuit 112. In the reference circuit 112, an appropriate filter process is performed on the design image data that is the image data of the received graphic.

  FIG. 8 is a diagram for explaining the filter processing.

  An optical image obtained from the sensor circuit 106, which will be described later, is blurred due to the resolution characteristics of the magnifying optical system 104, the aperture effect of the photodiode array 105, or the like, in other words, a state in which a spatial low-pass filter is activated. is there. Therefore, it is possible to match the optical image 204 by applying the filter process to the design image data which is the image data on the design side, in which the image intensity (light / dark value) is a digital value. In this way, a reference image to be compared with the optical image 204 is created.

  Next, an optical image acquisition method will be described with reference to FIGS.

  In FIG. 6, the optical image acquisition unit A acquires the optical image 204 of the mask 101. Here, the optical image 204 is an image of a mask on which a graphic based on graphic data included in the design pattern is drawn. A specific method for acquiring the optical image 204 is as follows, for example.

  The mask 101 to be inspected is placed on an XYθ table 102 provided so as to be movable in the horizontal direction and the rotation direction by motors of XYθ axes. Then, light is emitted from the light source 103 disposed above the XYθ table 102 to the pattern formed on the mask 101. More specifically, the light beam emitted from the light source 103 is applied to the mask 101 via the illumination optical system 170. Below the mask 101, an enlargement optical system 104, a photodiode array 105, and a sensor circuit 106 are arranged. The light that has passed through the mask 101 forms an optical image on the photodiode array 105 via the magnifying optical system 104.

  Note that the magnifying optical system 104 may be configured so that the focus is automatically adjusted by an automatic focusing mechanism (not shown). Further, although not shown, the inspection apparatus 100 irradiates light from below the mask 101, guides the reflected light to the second photodiode array through the magnifying optical system, and collects the transmitted light and the reflected light simultaneously. It may be configured.

  As shown in FIG. 1, the inspection area is virtually divided into a plurality of strip-shaped inspection stripes 20 having a scan width W in the Y direction, and each of the divided inspection stripes 20 is continuously scanned. Thus, the operation of the XYθ table 102 is controlled, and an optical image is acquired while moving in the X direction. An image having a scan width W as shown in FIG. 1 is continuously input to the photodiode array 105. When an image in the first inspection stripe 20 is acquired, an image having a scan width W is continuously input to the second inspection stripe 20 in the same manner while the XYθ table 102 is moved in the opposite direction. The third inspection stripe 20 is acquired while moving in the direction opposite to the direction in which the image in the second inspection stripe 20 is acquired, that is, in the direction in which the image in the first inspection stripe 20 is acquired. In this way, it is possible to shorten a useless processing time by continuously acquiring images.

  The pattern image formed on the photodiode array 105 in FIG. 6 is photoelectrically converted by the photodiode array 105 and further A / D (analog-digital) converted by the sensor circuit 106. Sensors are arranged in the photodiode array 105. An example of this sensor is a TDI (Time Delay Integration) sensor. While the XYθ table 102 continuously moves in the X-axis direction, the pattern of the mask 101 is imaged by the TDI sensor. Here, the light source 103, the magnifying optical system 104, the photodiode array 105, and the sensor circuit 106 constitute a high-magnification inspection optical system.

  The XYθ table 102 is driven by a table control circuit 114 under the control of the control computer 110, and can be moved by a drive system such as a three-axis (XY-θ) motor that drives in the X, Y, and θ directions. It has become. As these X-axis motor, Y-axis motor, and θ-axis motor, for example, step motors can be used. The movement position of the XYθ table 102 is measured by the laser length measurement system 122 and sent to the position circuit 107. The mask 101 on the XYθ table 102 is automatically conveyed from the autoloader 130 driven by the autoloader control circuit 113, and is automatically discharged after the inspection is completed.

  The optical image 204 output from the sensor circuit 106 is sent to the comparison circuit 108 together with data indicating the position of the mask 101 on the XYθ table 102 output from the position circuit 107. The optical image 204 is, for example, 8-bit unsigned data, and represents the brightness gradation of each pixel. Further, the reference image described above is also sent to the comparison circuit 108.

  The comparison circuit 108 includes a deviation statistic processing unit 108a and a defect detection unit 108b.

  In the deviation amount statistical processing unit 108a, the deviation amount obtained in the alignment process of the present embodiment described with reference to FIGS. 2, 3, and 4 is calculated, and the deviation amount is compared with a threshold value. . The comparison result is stored in the magnetic disk device 109b as a deviation amount report.

  For example, using the process of FIG. 2, the processing performed by the deviation amount statistical processing unit 108a is as follows. That is, first, the measured values of the coordinates of the four points on the mask are sent to the deviation amount statistical processing unit 108a. In addition, the design data read from the magnetic disk device 109a is also sent to the deviation amount statistical processing unit 108a. Then, the deviation amount statistical processing unit 108a calculates a difference (position deviation amount) between the measured values of the four points and the coordinates of the corresponding design data. Next, it is determined whether or not the difference (positional deviation amount) between one point having the largest difference and the coordinates of the design data exceeds a preset threshold value. The determination result is stored in the magnetic disk device 109b as a deviation amount report.

  In the above, the deviation amount statistical processing unit 108a is used to calculate the deviation amount obtained in the alignment step of the present embodiment, but it goes without saying that it may be performed by other mechanisms / software such as the control computer 110. .

  In the defect detection unit 108b, the optical image 204 sent from the sensor circuit 106 and the reference image generated by the reference circuit 112 are compared using an appropriate comparison determination algorithm. As a result of the comparison, when the difference between the two exceeds a predetermined threshold, the portion is determined as a defect. If the defect is determined, the coordinates, the optical image 204 and the reference image that are the basis for the defect determination are stored in the magnetic disk device 109a as the mask inspection result 205.

  The mask inspection result 205 can be sent to the review apparatus 500 as shown in FIG. The review is an operation in which the operator determines whether the detected defect is a problem.

  In the review process, the operator determines whether correction is necessary based on the mask inspection result 205. Specifically, the operator compares and reviews the reference image that is the basis for the defect determination and the optical image including the defect.

  The review apparatus 500 displays an image of a defective portion of the mask using the observation optical system of the inspection apparatus 100 while moving the table on which the mask is placed so that the defect coordinates of each defect can be observed. . At the same time, these are displayed side by side on the screen so that the determination conditions for defect determination and the optical image and reference image that are the basis for determination can be confirmed. As the screen, a screen of the control computer 110 or a computer screen separately prepared is used. By displaying the defects on the mask side by side in the review process, it becomes easy to determine whether or not the mask pattern should be corrected. In addition, the review apparatus 500 uses the observation optical system of the inspection apparatus 100 while moving the table on which the mask is placed, and does not display the image of the defective portion of the mask, and refers to the stored optical image. It is also possible to display images for review.

  Through the above steps, it is determined whether or not there is a defect by the inspection apparatus and the subsequent visual review. Then, after determining whether or not correction is necessary and discriminating the defect to be corrected, this mask is sent to the correction device together with information necessary for correction (defect information list 207). Here, the information necessary for correction includes, for example, coordinates in the mask, whether the defect is convex or concave, that is, whether the shading film is to be shaved or compensated, and the pattern of the portion to be corrected by the correction device Pattern data for recognizing. As the pattern data, the above-described optical image can be used.

  As described above, according to the present embodiment, the alignment is performed except for the alignment mark in the extremely distorted portion, so that it is possible to avoid the influence of the specific alignment mark from acting strongly on the correction processing. Can do. That is, it is possible to reduce a situation in which a normal LSI pattern is determined to be a defective pattern due to a strong effect of a specific alignment mark even though it is originally a process for correcting mask distortion. . Even if the alignment pattern and the LSI pattern are similarly distorted, the quality of the LSI pattern can be determined without being affected by the alignment pattern. Therefore, according to the present embodiment, an inspection apparatus and an inspection method capable of accurately determining the quality of a pattern regardless of the influence of mask distortion are provided.

Embodiment 2. FIG.
In the first embodiment, a method has been described in which an optical image and a reference image are aligned and a defect is detected by removing an alignment mark at an extremely distorted portion. In the present embodiment, a method will be described in which when the tendency of misalignment of the alignment mark and the tendency of misalignment of the main pattern are different, these are separated to increase the inspection accuracy.

  The configuration of the inspection apparatus of this embodiment can be the same as that in FIG. 6 described in Embodiment 1. In addition, a mask used in photolithography is used as an inspection target, but is not limited thereto.

  FIG. 9 is a flowchart showing an alignment method according to this embodiment. As shown in this figure, first, the alignment method described in the first embodiment is performed (S401). For example, any of the methods described in FIGS. 2, 3, and 4 can be performed.

  Next, a main pattern such as an LSI pattern to be inspected is pre-scanned to determine whether the pattern has a defect such as distortion (S402). The alignment performed in S401 is normally performed using measurement points located outside the inspection area, and therefore, in S402, the purpose is to roughly grasp the positional deviation in the inspection area. For this reason, instead of scanning the entire inspection range to obtain an optical image, for example, for the strip-shaped stripe region described in FIG. 1, the first few stripes in the Y direction are scanned to obtain an optical image, and then The last few stripes in the Y direction are scanned to obtain an optical image. Thereby, when a pattern is formed by drawing a charged particle beam such as an electron beam, optical images at the first stage and the last stage of the drawing are obtained.

  In recent years, the circuit line width required for a semiconductor element has become increasingly narrower as the large scale integrated circuit (LSI) is highly integrated and has a large capacity. Correspondingly, since the shot size of the electron beam and the deflection width of the deflector in the electron beam drawing apparatus tend to be small, the time required for drawing is long. Variations in drawing conditions can occur during such a long drawing time.

  Therefore, in S402, the optical images of the first stage and the last stage of drawing are acquired, and the positional deviation of each actual pattern is detected and compared from the comparison with these reference images. If the difference in positional deviation between these actual patterns exceeds a predetermined threshold value, it is determined that there is a pattern distortion that cannot be eliminated by correction, and the inspection is terminated without further scanning. Thereby, the whole inspection time can be shortened.

  On the other hand, in S402, when the difference between the defect at the first stage of drawing and the defect at the last stage is equal to or smaller than a predetermined threshold value, the process proceeds to S403, and the entire pattern is scanned to obtain an optical image. That is, a normal inspection process is performed.

  Next, the inspection result is determined in S404. That is, with reference to the mask inspection result 205 described with reference to FIG. 7 of the first embodiment, it is determined whether or not there are many defects. In addition, the defect said here is a position shift defect. Also, the degree of frequent occurrence can be determined, for example, by comparing the number of defects with a predetermined threshold and whether or not the threshold is exceeded.

  If it is determined in S404 that there are not many defects, the inspection process is terminated.

  On the other hand, if it is determined in S404 that there are many defects, the deviation amount report acquired in S401 is referred to (S405). The deviation amount report is as described in the first embodiment. That is, the comparison result between the shift amount between the optical image and the reference image and the threshold value is stored in the magnetic disk device 109b of FIG.

  In S405, detailed examination is performed on the shift amount acquired in S401. This examination is performed by the deviation amount statistical processing unit 108a in FIG. 6, and as shown in FIG. 10, it is examined whether or not there is a characteristic deviation in the order of each defect, inspection stripe, and mask.

  FIG. 10 is a flowchart showing a method for determining whether or not a characteristic tendency is observed in the defects frequently generated in S404 of FIG. This step is performed in S405 of FIG.

  First, the number of detected defects is grasped by referring to the mask inspection result obtained in S403 (S501). Next, the actual measurement coordinates and the ideal coordinates obtained from the design data are grasped for each defect (S502). Next, it is determined whether a characteristic tendency is observed in each defect (S503).

  For example, (1) the amount of deviation in the X direction, (2) the amount of deviation in the Y direction, and (3) the distance from the ideal coordinates are determined for each defect from the data grasped in S501 and S502.

の関係が成立するとき、一様なＸ方向のずれ量が認められると言える。ＸminとＸmaxは、それぞれ任意に設定可能である。 The amount of deviation in the X direction is the difference between the “measured X coordinate” and the “ideal X coordinate”. If a uniform amount of deviation in the X direction is recognized for all of the detected defects, it can be determined that there is a characteristic deviation in the X direction. Here, the uniform deviation amount can be defined as, for example, a state in which the deviation amount of each defect is within a predetermined range. That is, set Xmin and Xmax,

When the above relationship is established, it can be said that a uniform amount of deviation in the X direction is recognized. Xmin and Xmax can be set arbitrarily.

の関係が成立するとき、一様なＹ方向のずれ量が認められると言える。ＹminとＹmaxは、それぞれ任意に設定可能である。 The amount of deviation in the Y direction is the difference between the “actually measured Y coordinate” and the “ideal Y coordinate”. When a uniform amount of deviation in the Y direction is recognized for all detected defects, it can be determined that there is a characteristic deviation in the Y direction. Here, the uniform shift amount can be defined as a state in which the shift amount of each defect is within a predetermined range as described above. That is, set Ymin and Ymax,

When this relationship is established, it can be said that a uniform amount of deviation in the Y direction is recognized. Ymin and Ymax can be set arbitrarily.

で求められる。検出された欠陥の全てについて、得られた値に一様性が認められた場合、理想的な座標からの距離に特徴的なずれがあると判断できる。例えば、ＤminとＤmaxとを設定し、

の関係が成立するとき、理想的な座標からの距離に特徴的なずれがあると言える。ＤminとＤmaxは、それぞれ任意に設定可能である。 The distance from the ideal coordinates is

Is required. When uniformity is recognized in the obtained values for all the detected defects, it can be determined that there is a characteristic shift in the distance from the ideal coordinates. For example, set Dmin and Dmax,

When this relationship is established, it can be said that there is a characteristic shift in the distance from the ideal coordinates. Dmin and Dmax can be arbitrarily set.

  For each defect, whether or not there is a characteristic tendency in positional deviation from the ideal coordinates is determined by (1) the amount of deviation in the X direction, (2) the amount of deviation in the Y direction, and (3) ideal This is done using each result of distance from coordinates. In this case, if any one of the three items (1) to (3) has a characteristic shift, it can be determined that the defect has the characteristic tendency. Alternatively, when there is a characteristic shift in two or more items, it can also be determined that each defect has the characteristic tendency. Furthermore, it can be determined that there is a characteristic tendency for each defect only when a characteristic shift is observed in all three items. Which is selected can be appropriately selected depending on the case.

  Next, based on the above determination made for each defect, it is determined whether or not there is a characteristic tendency in positional deviation from ideal coordinates for each inspection stripe pattern (S504).

  For example, in a single inspection stripe, when the percentage of defects determined to have a characteristic tendency for deviation in S503 is greater than or equal to a predetermined value, it is determined that the pattern in this inspection stripe has a characteristic deviation tendency. be able to. As an example, it is assumed that the predetermined value is 70% and the number of defects detected in the inspection stripe is 10. Assume that it is determined in S503 that eight of these defects have a characteristic tendency to shift. In this case, the percentage of defects determined to have a characteristic tendency of deviation in this inspection stripe is 80%. Therefore, since the predetermined value of 70% is exceeded, it is determined that this inspection stripe pattern has a characteristic deviation tendency. Other inspection stripe patterns can be similarly determined.

  Next, based on the above determination made for each inspection stripe, it is determined whether or not there is a characteristic tendency in positional deviation from ideal coordinates for the pattern of the entire mask (S505).

  For example, it can be determined that the mask has a characteristic shift tendency when the ratio of the inspection stripes determined to have a shift tendency in S504 is equal to or greater than a predetermined value in the entire mask. That is, assuming that the predetermined value is 70%, the ratio of the inspection stripes determined to have a characteristic deviation tendency in S504 to the entire mask is 80%. In this case, it is determined that the main pattern of this mask has a characteristic shift tendency.

  In the present embodiment, it may be determined whether there is a characteristic shift tendency in the entire mask based on the above determination made for each defect. That is, it is possible to omit the step of determining whether there is a characteristic shift tendency for each inspection stripe.

  For example, assume that the number of defects detected in the entire mask is 1000. Assume that in S503, it is determined that 800 of the defects have a characteristic tendency to shift. In this case, the ratio of defects determined to have a characteristic tendency of deviation in the entire mask is 80%. Therefore, since the predetermined value of 70% is exceeded, it is determined that the main pattern of this mask has a characteristic shift tendency.

  An example in which the mask shows a characteristic shift tendency will be described with reference to FIGS.

  FIG. 11 is a schematic diagram showing ideal lattice points. FIG. 12 is a diagram schematically showing the relationship between the actual defect coordinates and the theoretical coordinates of the defect location obtained from the design data. In this example, it can be seen that the actual defect coordinates are shifted in the direction of the arrow as a whole with respect to the ideal lattice points, that is, the theoretical coordinates of the defect location obtained from the design data. The overall shift in the direction of the arrow is a characteristic shift tendency seen in this mask.

  It is determined by the processes of S501 to S505 shown in FIG. 10 whether or not a characteristic tendency is recognized in the displacement of the entire mask. Based on this result, in S406 of FIG. 9, it is determined whether to end the inspection process or to perform inspection again by performing alignment.

  As a result of the examination in S405, if it is determined that there is no characteristic tendency in the main pattern shift of the entire mask, it is determined in S406 that the inspection is finished. On the other hand, when it is determined that there is a characteristic tendency in the displacement of the entire mask, it is determined that the displacement tendency of the alignment mark of the mask does not match the in-plane distribution of the main pattern. Then, after proceeding to S407 and correcting the deviation of the main pattern, the process returns to S401 and the alignment of the first embodiment is performed again. Taking FIG. 12 as an example, after the shift in the direction of the arrow in FIG. 12 is corrected in S407, the process returns to S401.

  According to the first embodiment, since the alignment of the optical image and the reference image is performed except for the alignment mark at the extremely distorted portion, it is avoided that the influence of the specific alignment mark strongly acts on the correction process. be able to. However, if the misalignment tendency of the mask alignment mark and the in-plane distribution of the mask, that is, the misalignment tendency of the main pattern do not match, it is possible to accurately determine the quality of the pattern simply by removing the alignment mark at the extremely distorted portion It may not be enough to make a decision.

  According to the present embodiment, it is determined whether or not the tendency of the alignment mark of the mask and the in-plane distribution of the mask do not match, and if they do not match, the latter correction is performed. Then, the alignment of the optical image and the reference image is performed except for the alignment mark at the extremely distorted portion. This provides an inspection apparatus and an inspection method that can more accurately determine the quality of a pattern.

  The present invention is not limited to the above embodiments, and various modifications can be made without departing from the spirit of the present invention.

  For example, whether or not a characteristic tendency is observed in the defects frequently generated in S404 in FIG. 9 can be determined not only by the method described in FIG. 10 but also as follows.

  First, the number of detected defects is grasped by referring to the mask inspection result obtained in S403 of FIG.

  Next, for each defect, the actual measurement coordinates and the ideal coordinates obtained from the design data are grasped. The ideal coordinates mentioned here are the coordinates obtained from the design data where the measured coordinates should originally exist (coordinates). The method for obtaining the coordinates from the design data is from paragraph number 0063 to paragraph number 0070. It is obtained through the described process. And the deviation | shift amount from the ideal coordinate calculated | required from the design data of each defect is calculated | required.

  Next, the deviation from the ideal coordinates obtained from the design data of each defect is connected by an approximate curve in stripe units. Excluding defects whose deviation amount exceeds a predetermined threshold from this approximate curve, the measured coordinates of the remaining defects are represented by a two-dimensional map.

  The two-dimensional map of the actual measurement coordinates obtained above is overlaid with the two-dimensional map that is similarly expressed with respect to the ideal coordinates obtained from the design data. The difference between the actually measured coordinates (x, y) of the defect and the ideal coordinates obtained from the design data exists in the same quadrant with a probability equal to or higher than a predetermined threshold, and exceeds the predetermined threshold. In this case, it can be determined that there is a characteristic shift tendency.

  In each of the above embodiments, description of parts that are not directly required for the description of the present invention, such as the device configuration and control method, is omitted. However, the necessary device configuration and control method are appropriately selected and used. Needless to say, you can. In addition, all pattern inspection apparatuses or pattern inspection methods that include elements of the present invention and whose design can be changed as appropriate by those skilled in the art are included in the scope of the present invention.

DESCRIPTION OF SYMBOLS 20 Inspection stripe 100 Inspection apparatus 101 Mask 102 XY (theta) table 103 Light source 104 Magnification optical system 105 Photodiode array 106 Sensor circuit 107 Position circuit 108 Comparison circuit 108a Deviation amount statistics processing part 108b Defect detection part 109a, 109b Magnetic disk apparatus 110 Control computer 111 Expansion circuit 112 Reference circuit 113 Autoloader control circuit 115 Magnetic tape device 116 Flexible disk device 117 CRT
118 Pattern Monitor 119 Printer 120 Bus 122 Laser Length Measuring System 130 Autoloader 170 Illumination Optical System 201 CAD Data 202 Design Intermediate Data 203 Format Data 204 Optical Image 205 Mask Inspection Result 500 Review Device 600 Correction Device

Claims (5)

Illuminating a sample on which a pattern is formed, forming an image of the sample on an image sensor via an optical system, and obtaining an optical image of the sample;
Creating a reference image from the design data of the pattern;
Aligning the sample from at least four points on the sample;
The step of aligning the sample is a step of obtaining a difference between at least four points on the sample and design data corresponding to these points and selecting one point having the largest difference;
A step of aligning the optical image and the reference image at at least three points excluding the selected one point when an error amount between the selected one point and the design data exceeds a threshold value. Inspection method to do. Forming at least four types of quadrilateral or more polygons with at least three points selected from the at least four points and theoretical coordinates obtained from these at least three points;
Comparing each theoretical coordinate in the polygon of at least four types of quadrilateral or more with each coordinate of the design data corresponding to each of the at least four points to obtain a displacement amount;
The inspection method according to claim 1, wherein a point corresponding to a point having the largest positional deviation amount among the at least four points is set as one point having the largest difference. Forming at least four types of quadrilateral or more polygons with at least three points selected from the at least four points and theoretical coordinates obtained from these at least three points;
Compare each area of the at least four types of polygons of the quadrangle or more with the area of the polygons of the quadrangle or more formed from the coordinates of the design data corresponding to each of the at least four points. Identifying a polygon that is equal to or larger than a large rectangle of
2. The inspection method according to claim 1, wherein a point corresponding to the theoretical coordinates of the specified polygon or more of the at least four points is set as one point having the largest difference. After aligning the optical image and the reference image, comparing these images to detect defects in the pattern;
A step of performing at least several pattern inspections in advance before pattern inspection;
When the number of defects exceeds a predetermined value, the deviation of the pattern of the entire sample is obtained by obtaining the deviation amount in the X direction, the deviation amount in the Y direction, and the distance from the coordinates obtained from the design data for each defect. And a step of determining whether or not a certain tendency is recognized,
When it is determined that a certain tendency is found in the pattern deviation of the entire sample, the optical image and the reference image are aligned and the defect of the pattern is detected after correcting the deviation. The inspection method according to claim 1, wherein the inspection method returns to step (a). Illuminating the sample on which the pattern is formed, and forming an image of the sample on an image sensor via an optical system to obtain an optical image of the sample;
A part for creating a reference image from the design data of the pattern;
A portion for aligning the sample from at least four points on the sample;
The part for aligning the sample includes a part for obtaining a difference between at least four points on the sample and design data corresponding to these points and selecting one point having the largest difference, and the selected one point. When the error amount of the design data exceeds a threshold value, the inspection apparatus includes a portion for aligning the optical image and the reference image at at least three points excluding the selected one point.

Images