JPH06174428A - Size measuring device - Google Patents

Abstract

(57) [Summary] [Purpose] Achieves stable dimension measurement without being affected by the material and thickness of the pattern, the performance of the image input device, and the like. [Structure] A microscope 1, a television camera 2 for converting an enlarged image of a sample 3 into an electric signal, an image processing device 4 for processing the image signal, and a display device 5 for displaying a measured value by the image processing device. ing. The image processing device 4 converts the image data of the first pattern and the second pattern provided on the substrate into the first pattern image data and the second pattern image data. To generate the first measured pattern image data and the second measured pattern image data, and the first measured pattern image data and the second measured pattern image data. The first measured pattern image inverted data and the second measured pattern image inverted data which are respectively inverted are created, and the first measured pattern image data and the first measured pattern image inverted data are generated. And the correlation function between the second measured pattern image data and the second measured pattern image inversion data, respectively, and the point where each correlation function becomes maximum is detected. Find the distance between the first and second patterns.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a size measuring device, and more particularly to a size measuring device for measuring the distance between two types of patterns.

[0002]

2. Description of the Related Art FIG. 9A is a plan view of a semiconductor substrate, and FIG. 9B is a waveform diagram showing reflected light intensity. An upper layer pattern 20 and a lower layer pattern 21 are formed on the semiconductor substrate to measure the positional deviation between the patterns. Horizontal direction of these patterns 20 and 21 (left and right direction in FIG. 9)
To measure the position shift of the
8 image is taken, the image data is integrated in the vertical direction (vertical direction in FIG. 9), and the reflected light intensity distribution 2 shown in FIG.
Get 9. Based on this intensity distribution 29, the edge coordinates 31 and 32 of the upper layer pattern 20 and the edge coordinates 3 of the lower layer pattern 21.
0 and 33 are detected, and the distance 26 between the center coordinate 22 between the edge coordinates 31 and 32 and the center coordinate 23 between the edge coordinates 30 and 33 is calculated and measured. Here, the edge position is
It is defined as a position 34 having the lowest reflected light intensity shown in FIG. 10A, a falling position 35 shown in FIG. 10B, or a rising position 36 shown in FIG. 10C.

[0003]

However, the relationship between the edge of the pattern and the contrast of the image is not uniform,
Depending on the pattern forming material and thickness, the definition of the edge position must be changed, and depending on the pattern, the reflected light intensity distribution near the edge is shown in FIG. 10 (d).
There is a problem that the measurement reproducibility is deteriorated due to the distortion as shown in FIG. Further, in a pattern in which only a low-contrast image can be obtained, the S / N ratio of the image is poor, and the reproducibility of measurement is deteriorated.

The present invention has been made in view of the above circumstances, and its object is to provide a dimension measuring apparatus capable of performing stable dimension measurement regardless of the material and thickness of the pattern and the shape of the pattern. That is.

[0005]

In order to solve the above-mentioned problems, a dimension measuring apparatus according to the first aspect of the present invention sets a distance between a first pattern and a second pattern provided on a substrate in a predetermined direction. In a dimension measuring device for measuring, image input means for inputting both patterns as an image, first pattern image data and second pattern image data from the image data of both patterns. Measured pattern image data creating means for separating image data and creating first measured pattern image data and second measured pattern image data, and the first measured image data. A pattern-to-be-measured reverse image in which first and second pattern-to-be-measured image reverse data are created by horizontally reversing the measurement pattern image data and the second pattern-to-be-measured image data, respectively. Data creating means and the first target Wherein a constant pattern image data first of the measured pattern - down image reversal de - correlation function of data, the second
Of the measured pattern image data and the second measured pattern
And an inter-pattern distance calculating means for obtaining a distance between the two patterns by detecting a correlation function with each image inversion data and detecting a point where each correlation function becomes maximum.

According to a second aspect of the present invention, there is provided a dimension measuring apparatus for measuring a distance between a first pattern and a second pattern provided on a substrate in a predetermined direction. Image input means for inputting an image as an image, and a first model pattern and a second model pattern which are formed in the same step as the first pattern and the second pattern and whose relative distances are known. From the image data of the first model pattern image data and the second model pattern image data
Model pattern image data that is separated from the model pattern image data and stored as a first pattern model and a second pattern model, respectively, and image data of the first pattern and the second pattern to be measured. From the first pattern image data and the second pattern image data to create the first measured pattern data and the second measured pattern data; Of the pattern model and the first measured pattern data, and the correlation function of the second pattern model and the second measured pattern data, respectively, and detect the point where each correlation function is maximum. And an inter-pattern distance calculation means for obtaining the distance between the two patterns.

[0007]

It is possible to perform stable dimension measurement even for an image with a poor S / N, regardless of the material and thickness of the pattern and the shape of the pattern. Further, in the measurement using the model, stable dimension measurement can be performed even with an asymmetric pattern.

[0008]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is an overall configuration diagram of a dimension measuring apparatus according to an embodiment of the present invention. TV camera 2 for microscope 1
Are connected to convert the magnified image of the sample 3 into an electric signal. The television camera 2 is connected to the image processing device 4, and the image processing device 4 processes the image signal as described later. The image processing device 4 is connected to the display device 5, and the display device 5 displays the measured value.

Next, the operation of the dimension measuring device for measuring the distance (deviation amount) between the upper layer pattern 20 and the lower layer pattern 21 in FIG. 2 will be described.

An enlarged image of the sample 3 observed by the microscope 1 is converted into an electric signal by the television camera 2, and the image signal is sent to the image processing device 4. The operation of the image processing device 4 will be described with reference to FIGS. Figure 2 shows the data when measuring the dimensions.
FIG. 3 is a diagram showing a flow chart during dimension measurement.

First, the image data in the measurement area 28 (FIG. 2A) is vertically integrated. As a result, the two-dimensional image becomes a one-dimensional image (step 21). Next, FIG.
Edge candidates (inflection points) are listed from the horizontal reflected light intensity distribution 6 shown in (b) (step 22). Then, the edge position is detected in pixel units from the edge evaluation function and the design value (step 23). Then, the upper layer pattern image data 9 and the lower layer pattern image data 7 are separated from the edge position data (step 24),
As shown in FIGS. 2C and 2D, inverted data 10 and 8 are created by inverting the left and right of each image data with reference to the center coordinates 280 of the measurement area 28 (step 2).
5). Then, the normal rotation data 9 of the upper layer and the lower layer,
Correlation functions 12 and 11 of 7 and the inversion data 10 and 8 are created (step 26). Then, the point where the correlation functions 12 and 11 of the upper layer and the lower layer take the maximum value is detected (step 27). The point that takes the maximum value is bi, bo (Fig. 2
(E)), both patterns 20 and 21 of the respective forward rotation data 9 and 7 are (bi /
2) and (bo / 2). Therefore, the distance between the upper layer pattern 20 and the lower layer pattern 21 is (bi-bo) / 2. According to this example,
Regardless of the material and thickness of the pattern and the shape of the pattern, stable dimension measurement can be performed even for images with poor S / N.

Next, another embodiment of the present invention will be described.

In the cross-sectional shape of the pattern on the semiconductor substrate, the left and right edges are not always formed symmetrically, and the inclination angles are often different. An example of such a pattern is shown in FIG. 4 (a) is shown in FIG.
FIG. 3B is a sectional view of the semiconductor substrate, FIG. 2B is a plan view, and FIG. 1C is a waveform diagram showing reflected light intensity. Since the right side edge of the upper layer pattern 80 has a larger inclination angle than the left side edge, the fall and rise of the output waveform also become gentle, and as a result, stable dimension measurement cannot be expected.

The features of such a pattern are exactly the same processing steps (photoresist application, exposure, etching, etc.).
The pattern layers formed through the steps appear as the same shape, and the pattern layers having different processing steps appear as different shapes. For example, in the case of the semiconductor substrate of FIG. 4, the pattern 90 formed through the same processing steps as the upper layer pattern 80 also has a large inclination angle of the right edge.
Such a tendency occurs not only with respect to a pattern on one semiconductor substrate, but also between corresponding pattern layers of a plurality of substrates manufactured through the same processing step.

Focusing on the above tendency, the measurement is first performed when pattern dimension measurement is performed at a plurality of locations on one semiconductor substrate or when dimension measurement is performed on a plurality of semiconductor substrates of the same lot. If the data is continuously used for measurement, it is possible to simplify the arithmetic processing.

Therefore, in this embodiment, the upper layer pattern image data 9, the lower layer pattern image data 7 and the distance M between the upper layer pattern and the lower layer pattern M =
(Bi-bo) / 2 is stored as a model, and the calculation of the dimension measurement in a series of patterns is executed using this model. The method will be described in detail below. It should be noted that, instead of storing the distance M = (bi-bo) / 2 as a model, an accurate measurement value of a pattern serving as a model obtained by using a reliable measuring device such as a scanning electron microscope (SEM) in advance. (Distance M) may be stored as a model.

FIG. 5 is a flow chart when creating a model, and FIG. 6 is a diagram showing data when creating a model.

The distance M between the upper layer pattern 80 and the lower layer pattern 81 is measured in advance. First, measurement area 1
The image data in 28 (FIG. 6A) is vertically integrated (step 51). Next, the edge candidates (inflection points) are listed from the reflected light intensity distribution 140 in the horizontal direction (FIG. 6B) (step 52). Then, the edge position is detected pixel by pixel (step 53). Then, as shown in FIGS. 6C and 6D, the upper layer pattern image data 142 and the lower layer pattern image data 141 are separated from the edge position data (step 54). The separated upper layer pattern image data 142 and lower layer pattern image data 141 and the distance M between the upper layer pattern and the lower layer pattern measured in advance are stored as a model in the storage unit of the image processing apparatus 4.

FIG. 7 is a flow chart at the time of dimension measurement using the model, and FIG. 8 is a diagram showing data at the time of dimension measurement using the model.

First, the image data in the measurement area 128 (FIG. 8A) obtained by the microscope 1 and the television camera 2 is displayed.
Data in the vertical direction (step 41). Next, the edge candidates (inflection points) are listed from the horizontal reflected light intensity distribution 113 (FIG. 8B) (step 42).
Then, the edge position is detected in pixel units (step 43). Then, as shown in FIGS. 8C and 8D, the upper layer pattern image data 109 and the lower layer pattern image data 107 are separated from the edge position data (step 44). Then, the upper layer pattern image data
Correlation function 11 between the pattern 109 and the upper pattern model 142
7, lower layer pattern image data 107 and lower layer pattern
The correlation function 115 with the input model 141 is created (step 45). Then, the point where the correlation functions 117 and 115 of the upper layer and the lower layer take the maximum value is detected (step 46). If the points at which the correlation functions 117 and 115 take the maximum values are Bi and Bo (FIG. 8 (e)), the upper layer pattern 90 and the lower layer pattern are calculated using the distance M between the upper layer pattern model and the lower layer pattern model. The distance to 91 is Bi-B
o + M (step 47).

[0022]

As described above, according to the dimension measuring apparatus of the present invention, stable dimension measurement can be performed even for an image having a poor S / N, regardless of the material and thickness of the pattern and the shape of the pattern. . Further, in the measurement using the model, stable dimension measurement can be performed even with an asymmetric pattern.

[Brief description of drawings]

FIG. 1 is an overall configuration diagram of a dimension measuring apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing data when measuring dimensions.

FIG. 3 is a flowchart for measuring dimensions.

FIG. 4 is a diagram showing a cross section, a plane, and a light intensity distribution of a measured pattern when a model is required.

FIG. 5 is a flowchart when creating a model.

FIG. 6 is a diagram showing data when creating a model.

FIG. 7 is a flowchart at the time of dimension measurement using a model.

FIG. 8 is a diagram showing data at the time of dimension measurement using a model.

FIG. 9 is a plan view of a semiconductor substrate and a waveform diagram showing reflected light intensity.

FIG. 10 is a waveform diagram for explaining the definition of an edge position.

[Explanation of symbols]

 1 Microscope 2 Television Camera 4 Image Processing Device 20 Upper Layer Pattern 21 Lower Layer Pattern

Claims (2)

[Claims] 1. A dimension measuring device for measuring a distance between a first pattern and a second pattern on a substrate in a predetermined direction, and image input means for inputting both patterns as images, The first pattern image data and the second pattern image data are separated from the image data of both patterns, and the first measured pattern image data and the second pattern image data are separated. Measured pattern image data creating means for creating the measured pattern image data, and a first left-right inverted version of the first measured pattern image data and the second measured pattern image data. Measured pattern reversal image data creating means for creating one measured pattern image reversal data and second measured pattern image reversal data, the first measured pattern image data and the first measured pattern Correlation with image reversal data , The second measured pattern image data and the second measured pattern image inversion data, respectively, the correlation function is obtained, and the point where each correlation function is maximum is detected to detect both patterns. An inter-pattern distance calculation means for obtaining a distance between the pattern measuring devices. 2. A dimension measuring device for measuring a distance between a first pattern and a second pattern provided on a substrate in a predetermined direction, and image input means for inputting both patterns as an image, From the image data of the first model pattern and the second model pattern, which are formed in the same process as the first pattern and the second pattern and whose relative distances are known, to the first model pattern image data. And a second model pattern image data are separated from each other and stored as a first pattern model and a second pattern model, respectively, and the first pattern and the second pattern to be measured. The first pattern image data and the second pattern image data are separated from the image data of the first measured pattern data and the second measured pattern data. Pattern data creating means for creating data, a correlation function between the first pattern model and the first pattern data to be measured, the second pattern model and the second pattern data to be measured, And the inter-pattern distance calculating means for obtaining the distance between the two patterns by detecting the correlation function of each of the two.