JPH04269614A - Pattern-position detecting method and executing apparatus thereof - Google Patents

Abstract

PURPOSE:To make it possible to detect the position of an object pattern accurately even if a beam is located at a position which is deviated from a target position. CONSTITUTION:In an inspection and measuring method using a charged beam, two-dimensional images are received, and then the image signals are added in one direction. Matching is performed for a one-dimensional waveform comprising the signal line obtained by the addition and a reference waveform obtained by the same procedure beforehand. Then, the position of an objective pattern in the two-dimensional image is detected. The charged beam undergoes line scanning at a position passing the detected position. The size of the pattern is measured based on the waveform of the two-dimensional signal obtained by the line scanning. The time is shortened to a large extent in comparison with the pattern matching in the intact state of the two-dimensional image, and the size of the pattern can be measured accurately and automatically.

Description

[Detailed description of the invention]

0001

[Field of Industrial Application] The present invention relates to a method for detecting a pattern position and an apparatus for implementing the same, particularly for measuring pattern dimensions, inspecting pattern shapes, etc. in the manufacturing process of very large scale semiconductor integrated circuit devices (VLSI), etc. The present invention relates to a method for detecting the position of a target pattern for a purpose, or a method for detecting a mark therefor, and an apparatus for implementing the same.

The present invention also relates to a method of automatically measuring dimensions such as the diameter of a circular pattern such as a minute through hole by using this pattern position detection method.

0003

[Prior Art] In a dimension measuring device using an electron beam,
By scanning the electron beam and detecting reflected electrons or secondary electrons in synchronization with this scanning signal, a secondary electron signal waveform in a direction perpendicular to the measurement pattern is obtained, and the pattern dimensions are measured. A widely used method for obtaining pattern dimensions from this signal waveform is to set an appropriate slice level for the secondary electron signal waveform, binarize it, and measure the pattern dimension from the interval. In addition, an apparatus has been proposed that linearly approximates each of the edges and baselines and obtains pattern dimensions from the distance between their intersection points (see ``Dimension Measuring Apparatus'' in Japanese Unexamined Patent Publication No. 61-80011).

[0004] When measuring such pattern dimensions,
First, recognize the target pattern, set the slice level from the outside (inside) of the pattern, and start the convex (
(Concave) Detects the edges of the pattern and measures the dimensions.

[0005]

[Problem to be Solved by the Invention] However, when automatically performing such pattern recognition or edge recognition, the position to capture the waveform is specified in advance, but the accuracy of stopping the stage and the alignment of the electron beam may be affected. deviates from the target position due to misalignment, positional variations of the chips themselves on the wafer, etc. Therefore, as shown in Figure 6, when a line scan is performed in the x direction at a pre-specified waveform acquisition position, if there is a shift in the y direction, the target pattern will not be detected by the line scan. I can't. In particular, fine through-holes and contact holes often have a nearly circular shape, and the center must be measured. In such a case, the waveform acquisition position specified in advance does not pass through the center of the through hole, so accurate measurement cannot be performed.

[0006] For this reason, there is a method of detecting patterns and finding the center positions of through holes and contact holes by pattern matching using two-dimensional images, but in this case, there is a problem that pattern detection takes time. .

The present invention has been made in order to solve the above-mentioned problems, and an object of the present invention is to solve the problem when the chip is placed at a position deviated from the target position due to stage stopping accuracy, misalignment, dispersion in the position of the chip itself, etc. An object of the present invention is to provide a technique that can accurately detect the position of a target pattern even when a beam is present.

Another object of the present invention is to accurately and automatically detect the center position of a nearly circular pattern such as a minute through hole or contact hole, and accurately measure its diameter and other dimensions. The goal is to provide technology that enables

The above and other objects and novel features of the present invention will become apparent from the description of this specification and the accompanying drawings.

0010

[Means for Solving the Problems] In order to achieve the above object, in the present invention, (1) in an inspection measurement method using a charged beam, after capturing a two-dimensional image, the image signal is transmitted in one direction. The most important thing is to detect the position of the target pattern in the two-dimensional image by performing matching processing between the one-dimensional waveform consisting of the signal sequence obtained by the addition and the reference waveform obtained in advance by the same procedure. The characteristics are as follows.

(2) A charged beam is line-scanned at a position passing through the position detected by the pattern position detection method, and the dimension of the pattern is measured from the secondary electron signal waveform obtained by the line-scanning. .

(3) In the inspection and measurement device using the charged electric beam, means for capturing a two-dimensional image, and an array for adding each signal component of the two-dimensional image in the X and Y directions and storing the added values. a memory, a means for extracting a signal string of a specified length from this array memory, and calculating a difference in each component between the extracted signal string and a signal string stored in an array memory different from the array memory. means for calculating the sum of the absolute values of the differences and storing this sum in correspondence with the position from which the signal string is extracted; and means for detecting the position of the signal string where the sum is the minimum. It is characterized by having the following.

[0013]

[Operation] According to the above-mentioned means (1), after capturing a two-dimensional image, a one-dimensional waveform consisting of a signal sequence obtained by adding the image signals in one direction, a reference sample, and a similar sample are used. The position of the target pattern in the two-dimensional image is detected through matching processing with a reference waveform obtained in advance using the same procedure, so even if the pattern is nearly circular, such as a minute through hole or contact hole, the center of the pattern can be detected. Position can be detected automatically and accurately.

According to the means (2), the central position of the nearly circular pattern is automatically and accurately detected, and a charged beam is scanned on a line passing through the detected position.
Since the dimensions of the target pattern are measured by obtaining a two-dimensional electronic signal, the center position of even almost circular patterns such as minute through holes and contact holes can be automatically detected accurately, and dimensions such as the diameter can be measured. Can be measured accurately.

According to the means (3), the means for capturing a two-dimensional image adds each signal component of the two-dimensional image in the X and Y directions, stores the added values in an array memory, and stores the added values in an array memory. Extracts a signal string of the specified length from the array memory,
Calculate the difference in each component between the extracted signal string and a signal string stored in an array memory different from the array memory, find the sum of the absolute values of the differences, and calculate this sum at the position where the signal string is extracted. Since the position of the signal train where the total amount is the minimum is detected, the center position can be accurately and automatically determined even for patterns with almost circular shapes such as minute through holes and contact holes. can be detected. Furthermore, compared to detecting edges by setting slice levels in a one-dimensional waveform, patterns can be detected with higher accuracy. In addition, since noise is removed when creating a 1D waveform from a 2D image, pattern detection can be performed in a shorter time and by removing the effects of noise compared to pattern matching in 2D. can.

[0016]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the drawings.

In the pattern position detection method of this embodiment, after the stage on which the sample is mounted has been moved and focused,
The target pattern is detected by pattern matching the reference waveform captured in advance and the waveform of the pattern to be measured.

FIG. 1 is a flowchart showing the processing procedure of a pattern position detection method according to an embodiment of the present invention. Here, as a method for cutting out a one-dimensional waveform from a two-dimensional image,
The gradation information of the two-dimensional image is divided into an X-direction component and a Y-direction component and added to form a one-dimensional waveform. The same length as the reference waveform is cut out from this waveform, and the position where the matching amount (the sum of the differences in the signal amounts of the two waveforms) is the minimum is detected. The pattern detection processing procedure will be explained with reference to FIG.

Step 101: Capture a two-dimensional SEM image.

Step 102: A signal obtained by adding each component S (i, j) of this two-dimensional image signal in the x and y directions,
Sx(i)=ΣS(i,j),
Find Sy(j)=ΣS(i,j).

Step 103: Sx(i), Sy(j)
are taken out as one-dimensional waveforms obtained by scanning the beam in the x and y directions, respectively.

Step 104: To detect the x-coordinate of the target pattern, the one-dimensional waveform Sx (
Use i). In order to perform matching processing between this waveform and a reference waveform created in advance, a one-dimensional waveform Sx scanned in the x direction according to the number of pixels m of the reference waveform
(i) Cut out a waveform of m pixels from (n pixels).

Step 105: Reference waveform Rx(i),
Before calculating the matching amount of the cut out waveforms, in order to match the levels of both signals, the values of the cut out one-dimensional waveforms are converted so that the maximum and minimum values of both signals are the same.

Step 106: Calculate the amount of matching between the reference waveform and the extracted one-dimensional waveform. This calculation may be performed using the sum of the absolute values of the signal differences between the two, or the sum of the squares of the signal differences.

Step 107: The matching amount is calculated while changing the waveform cutting position.

Step 108: Calculate the position where the matching amount is the minimum (most matches the reference waveform).

Step 109: If the change in the waveform extraction position in step 7 is large, it is selected whether to perform fine adjustment. When performing fine adjustment, the change in the waveform cutout position is made small, the vicinity of the matching position calculated in step 8 is renewed, and the processes in steps 104 to 107 are repeated.

Step 110: Calculate the amount of pattern positional deviation in the captured one-dimensional waveform.

[0029] Through the above processing procedure, it is possible to detect the amount of pattern shift in the x direction within the captured two-dimensional image. The same applies to the y direction. In this way, the position of the target pattern can be detected.

Note that when measuring the pattern width on a long line in one direction, it is not necessary to detect the position in the longitudinal direction of the line. In this case, the one-dimensional waveform only needs to be created in one direction.

As can be seen from the above description, according to the pattern position detection method of this embodiment, the processing time can be reduced compared to two-dimensional pattern matching. In other words, in two-dimensional processing, the number of calculations required for pattern matching is proportional to 4(n-m)2m2, but
In the method of this embodiment, it is proportional to 2(n-m)m. For example, in the case of n = 512 and m = 128, two-dimensional processing requires approximately 1010 calculations, whereas the method of this embodiment only requires approximately 105 calculations, resulting in a significant time reduction. .

FIG. 2 is a diagram for explaining an example of detecting an alignment mark using the pattern position detection method of this embodiment, which consists of the processing steps described above. This is an example of using it as an alignment mark. In this method, as shown in Figure 6,
Even if the position of the pattern is shifted, if the target pattern is present in the captured two-dimensional image, it is possible to detect the stepped portion in the one-dimensional waveform created by addition. However, if the pattern is spread out on one side as in the case of FIG. 2, and the center of the pattern is toward the edge, the S/N ratio will be poor. Therefore, a pattern extending in both directions, such as a cross pattern, is advantageous because the same S/N ratio can be obtained anywhere within the screen.

FIG. 3 is a diagram for explaining an application example of the pattern position detection method of this embodiment, which consists of the above-mentioned processing procedure, and is an example of detecting the position of an x mark in the diagram. Here, the positions on the left side of the two wires are detected. in this case,
A signal waveform including two wires is prepared as a reference waveform, and a position (x mark) to be detected within the signal waveform is also stored in the reference waveform. After capturing a two-dimensional scanning electron microscope (SEM) image, the processing is the same as in FIGS. 1 and 2 to prepare a waveform that is added in one-dimensional direction. While cutting out this waveform to the same length as the reference waveform,
Calculate the matching amount and find the matching position. Through this procedure, a position including two wires is detected. Next, by adding the amount by which the detection position (x mark) in the reference waveform is shifted from the center of the reference waveform, the position to be detected can be detected in the two-dimensional SEM image. in this way,
The method of marking the target pattern in the reference waveform has the effect of making it possible to specify the position in detail, such as measuring which of a plurality of wires to measure.

FIG. 4 is a diagram for explaining a method for measuring dimensions such as the diameter of a circular pattern such as a minute through hole, to which the pattern position detection method of the present invention is applied. In the same manner as in FIG. 1 or 2, the center position of the through hole is calculated from the signal waveforms added in the x and y directions. Next, the charged beam is scanned at a position passing through the detected center position, and the diameter of the through hole is measured. in this way,
If the through hole is in the SEM image, its center position can be detected accurately, and the charged beam is scanned and measured at a position passing through the center of the through hole, so the diameter can be accurately measured. It is also possible to accurately determine the area of the through-hole portion by determining the dimensions while sequentially shifting the scanning position of the charged beam from the end.

Note that pattern matching processing is not necessarily necessary for detecting the center position of the through hole. In the signal waveform added in the x and y directions, the portion where the signal amount is small corresponds to the through hole (where secondary electrons are difficult to detect). Therefore, a portion where the signal amount continues to be small near the minimum value of the signal waveform may be recognized as a through hole. However, even if the one-dimensional matching process of this embodiment was used, there was no significant difference in processing speed experimentally.

FIG. 5 is a block diagram showing a schematic configuration of a pattern position detecting device according to an embodiment for carrying out the pattern position detecting method of the present invention, and is an example of a length measuring device using an electron beam. In FIG. 5, 1 is a two-dimensional scanning electron microscope (SEM), 2 is a frame memory, 3 and 4 are adder circuits, 5 to 10 are first to sixth memories for arrays, 1
1 is a reference waveform database, 12 and 14 are difference circuits, 1
3 and 15 are, for example, sum calculation circuits consisting of comparison circuits; 16;
1 is an address designation circuit, 17 is a memory for corresponding to a signal string extraction position, and 18 is a minimum position detection circuit.

Next, the operation of the pattern position detection device of this embodiment will be explained. A scanning electron microscope (SEM) irradiates a sample with an electron beam to capture a secondary electron image (SEM image), and each signal of this SEM image is input into a frame memory 2 . Each component S(i
, j) are respectively added by the addition circuit 3 in the X direction and the addition circuit 4 in the Y direction, and the value Sx(i) of the output of the addition circuit 3 in the X direction is stored in the first memory 5. The value Sy(j) of the output of the circuit 4 is stored in the second memory 6. The respective lengths of these first and second memories 5 and 6 are:
It is determined by the captured pixels of the frame memory 2, and normally values such as 512 and 1024 are used. On the other hand, apart from these memories, a signal waveform of a pattern having the same shape as the target pattern is previously captured as a reference waveform. This reference waveform is
Typical patterns for each pattern and material are stored in a reference waveform database 11, from which the same type as the detected pattern is selected and stored in the fifth and sixth memories 9.
, 10. The length of each of the fifth and sixth memories 9 and 10 is, for example, a value of about 128 to 256 (not necessarily a multiple of 2), since only the target pattern portion is required. The fifth of these
, a signal string having the same length as the length of the sixth memories 9, 10 is extracted from the first and second memories 5, 6 and is stored in the third and fourth memories.
Store in memories 7 and 8. 3rd memory 7 and 5th memory 9
, the difference between each component of the fourth memory 8 and the sixth memory 10 is calculated by the difference circuits 12 and 14, and the sum of the absolute values is calculated by the addition circuit 1.
3 and 15 respectively. The obtained summation amounts are stored in the memory 17 corresponding to the signal string extraction position along with the position where the signal string is extracted (starting position) while sequentially shifting the position from which the signal string is extracted. When the signal string is extracted to the end, the minimum position detection circuit 18 consisting of a comparison circuit detects the position where the total sum is minimum, thereby detecting the pattern position. Of course, these calculations may be performed using a computer.

Although the present invention has been specifically explained above based on examples, it goes without saying that the present invention is not limited to the above-mentioned examples and can be modified in various ways without departing from the gist thereof. .

[0039]

[Effects of the Invention] As explained above, according to the present invention, since a two-dimensional image is converted into a one-dimensional waveform and pattern matching processing is performed, the time required for pattern matching is significantly longer than that of performing pattern matching on a two-dimensional image. It is shortened to . Further, even if the beam position deviates from the target position due to stage stopping accuracy, misalignment, variations in the position of the chip itself, etc., the pattern can be detected as long as there is a pattern within the screen. Therefore, the pattern dimensions at the center of almost circular patterns such as through holes and contact holes can be accurately and automatically measured.

[Brief explanation of the drawing]

FIG. 1 is a flowchart showing the processing procedure of a pattern position detection method according to an embodiment of the present invention.

FIG. 2 is a diagram for explaining an example of detecting an alignment mark using the pattern position detection method of the present embodiment.

FIG. 3 is a diagram for explaining an application example of the pattern position detection method of this embodiment.

FIG. 4 is a diagram for explaining a pattern dimension measurement method to which the pattern position detection method of the present invention is applied.

FIG. 5 is a block diagram showing a schematic configuration of a pattern position detection device according to an embodiment for carrying out the pattern position detection method of the present invention.

FIG. 6 is a diagram for explaining problems in a conventional pattern position detection method.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1... Two-dimensional scanning electron microscope (SEM), 2... Frame memory, 3, 4... Addition circuit, 5-10... First memory to sixth memory, 11... Reference waveform database, 12, 14...
Difference circuit, 13, 15...Sum calculation circuit, 16...Address designation circuit, 17...Memory for corresponding to signal string extraction position, 18
…Minimum position detection circuit

Claims (3)

[Claims] Claim 1: In an inspection measurement method using a charged beam, after capturing a two-dimensional image, the image signals are added in one direction, and a single
A pattern position detection method characterized by performing matching processing between a dimensional waveform and a reference waveform obtained in advance in the same procedure to detect the position of a target pattern in a two-dimensional image. 2. Line-scanning the charged beam at a position passing through the position detected by the pattern position detection method according to claim 1, and measuring the dimension of the pattern from the secondary electron signal waveform obtained by the line-scanning. A pattern dimension measurement method characterized by: 3. An inspection and measurement device using a charged beam, comprising means for capturing a two-dimensional image, an array memory for adding each signal component of the two-dimensional image in the X and Y directions, and storing the added values. , means for extracting a signal string of a specified length from this array memory, and means for calculating the difference in each component between the extracted signal string and a signal string stored in an array memory different from the array memory. and means for calculating the sum of the absolute values of the differences and storing this sum in correspondence with the position from which the signal string is extracted, and means for detecting the position of the signal string where the sum is the minimum. A pattern position detection device characterized by: