KR20100001133A

KR20100001133A - Method for correcting optical proximity effect

Info

Publication number: KR20100001133A
Application number: KR1020080060932A
Authority: KR
Inventors: 최진영
Original assignee: 주식회사 하이닉스반도체
Priority date: 2008-06-26
Filing date: 2008-06-26
Publication date: 2010-01-06

Abstract

PURPOSE: A method for correcting an optical proximity effect is provided to improve the accuracy of the optical proximity effect compensation by using an SEM image of the etched layer pattern. CONSTITUTION: A DI(Develop Inspection) contour image is extracted by using a DI image(S10). A first calibration is performed by using the DI contour image(S20). The FI contour image is extracted by using the FI(Final Inspection) image(S30). An image of the noise pattern excluding the image of the etched layer pattern is removed from the FI contour image(S40). A second calibration is performed by using FI contour image with removed noise pattern image(S60). A modeling is performed by using a first and a second calibration result(S70), and The optical proximity effect compensation is performed(S80).

Description

Method for correcting optical proximity effect

The present invention relates to an optical proximity effect compensation method, and more particularly, to an optical proximity effect compensation method capable of accurately performing calibration after an etching process.

As the degree of integration of semiconductor devices increases, design rules of the devices decrease, and many studies for this have been conducted.

In the exposure process for realizing a fine pattern, an exposure source with a short wavelength is used as an effort to increase resolution, and as the size of the pattern gradually decreases, exposure such as extreme ultraviolet rays having a wavelength similar to the size of the pattern In the case of using a circle, the influence of light diffraction and interference in the exposure process is greatly increased and the distortion of the pattern is intensified.

This phenomenon is called the optical proximity effect, and light is generated from a different optical path than the optical path to achieve the same layout as the designed layout. It is not possible to implement it in a distorted form.

Since the distorted patterns have to be compensated because they lower the yield and performance of the semiconductor device, an optical proximity effect compensation method has been developed for this purpose.

Compensation of the optical proximity effect is generally performed by extracting calibration data for the photosensitive film pattern formed by the exposure developing process.

However, the photoresist pattern is formed by etching the etched layer formed under the photoresist pattern by using the photoresist pattern as an etch mask in the etching process after the exposure development process due to the process index that changes as the pattern becomes more precise and finer. In this case, the etching cannot be performed.

Therefore, since the photoresist pattern formed after the exposure and development process and the etched layer pattern formed after the etching process are not formed in the same shape, the optical proximity to the etched layer pattern is only compensated for by the optical proximity effect performed using only the calibration of the photoresist pattern. Since the effects cannot be accurately corrected, there is still a problem in that the pattern is formed distorted.

In order to solve such a problem, calibration may be performed on the etched layer pattern formed by the etching process performed after the exposure and developing process, and the optical proximity effect may be compensated for using the etched layer pattern, but reliability of data for calibration of the etched layer may be compensated. Apart from this, there is a problem that optical proximity effect compensation cannot be performed accurately.

The reason can be seen through the optical proximity effect compensation method according to the prior art described below.

FIG. 1A is a photograph showing a conventional inspection (DI) image, and FIG. 1B is a photograph showing a conventional inspection (FI) image.

As illustrated in FIG. 1A, the DI image is an image after the exposure and development processes are performed and shows the photoresist pattern 10 formed on the etched layer.

At this time, since the calibration extracted from the photoresist pattern 10 has high accuracy, the accuracy of modeling using the same is also high.

As shown in FIG. 1B, the FI image shows the etched layer pattern 20 formed by etching the etched layer by using the photoresist pattern 10 as an etch mask. As shown in FIG. 1B, a lower pattern is formed around the etched layer pattern. 30 is exposed as it is.

The DI contour image and the FI contour image are extracted using the DI image and the FI image thus obtained (S1).

In this case, the contour image is an image of a line connecting the outer portion of the pattern, and in the case of the DI contour image, the contour image is represented by a line connecting the outer portion of the photoresist pattern 10.

The FI contour image extracted in this manner should appear as a line connecting the outer portion of the etched layer pattern 20, but since the lower pattern is exposed around the etched layer pattern 20, only the etched layer pattern 20 is exposed. Rather, the outer portion of the lower pattern 30 appears as a reflected line.

Next, the target layout and the FI contour image are aligned (S2).

At this time, the target layout and the FI contour image are not aligned correctly, because the exposed subpatterns around the etched layer pattern are recognized as noise in the FI contour image, so that the FI contour image does not match the target layout.

For this reason, accurate information may not be obtained even at the step S3 of performing calibration using alignment information of the FI contour image and the etched layer pattern.

The method currently used to perform the calibration of the etch pattern is to measure the CD for the etch pattern, which is very cumbersome and inaccurate data.

As a result, the optical proximity effect compensation cannot be accurately performed due to the incorrect calibration of the etched layer pattern formed through the etching process.

In the optical proximity effect compensation method of the present invention, the lower pattern exposed during the etching process for forming the etched layer pattern acts as a noise to solve the problem that accurate calibration data cannot be obtained.

The optical proximity effect compensation method of the present invention comprises the steps of extracting a DI contour image using a development inspection (DI) image and

Performing a first calibration using the DI contour image;

Extracting the FI contour image using a final inspection (FI) image, and

Removing an image of a noise pattern except an image of an etched layer pattern from the FI contour image;

Performing a second calibration using the FI contour image from which the image of the noise pattern has been removed; and

Modeling using the first and second calibration results and performing optical proximity effect compensation using the first and second calibration results.

In this case, after performing the optical proximity effect compensation, performing a simulation on the layout in which the optical proximity effect compensation is completed, extracting a simulation image, and performing optical proximity effect compensation until the target layout matches the simulation image; The method may further include repeating the simulation.

The DI image is an image obtained by measuring a photoresist pattern formed by forming an etched layer on an upper surface of a semiconductor substrate, applying a photoresist layer on the etched layer, and performing an exposure and development process using an exposure mask having the target layout. Characterized in that it comprises a.

The DI contour image may include a line connecting an outer portion of the image of the photoresist pattern.

The FI image may include an image obtained by measuring an etched layer pattern and a noise pattern.

The noise pattern may include a lower pattern formed on a lower layer of the etched layer pattern.

The FI contour image may include a line connecting an outer portion of the etched layer pattern image and a line connecting an outer portion of the noise pattern image.

In addition, removing the image of the noise pattern except the image of the etched layer pattern from the FI contour image may match the target layout of the lower pattern with the FI contour image, thereby forming an image of the etched layer pattern in the FI contour image. It is characterized in that it comprises the step of removing the information on the image of the subpattern from the FI contour image leaving information about.

The performing of the second calibration may include aligning the target layout with the FI contour image from which the image of the noise pattern has been removed.

The optical proximity effect compensation method of the present invention can perform accurate calibration using a scanning electron microscope image of an etched layer pattern formed after an etching process, thereby improving the accuracy of optical proximity effect compensation, thereby improving characteristics of a semiconductor device. .

Hereinafter, with reference to the accompanying drawings an embodiment of the present invention will be described in detail.

2 is a flowchart of an optical proximity effect compensation method according to the present invention.

An etched layer is formed on the semiconductor substrate, a photosensitive film is applied on the etched layer, and a photosensitive film pattern is formed using a mask having a target layout.

In this case, a hard mask layer may be further formed on the etched layer to facilitate etching of the etched layer.

The DI image of the present invention is an image obtained by measuring the photoresist pattern 40 as shown in FIG. 3A, wherein the DI image is preferably measured by a scanning electron microscope.

First, the DI contour image is extracted using the DI image (S10).

As illustrated in FIG. 3B, the DI contour image of the present invention may be represented by a dotted line 50 connected to an outer portion of the photoresist pattern.

Next, a first calibration is performed using the DI contour image (S20).

At this time, since there is no singular pattern that may generate noise around the photoresist pattern, it is possible to accurately calibrate using the DI contour image.

Calibration can be understood as a task for minimizing the degree of mismatch by extracting the mismatched parts by matching the actual implementation pattern with the target layout.

It can be easily selected and used by those skilled in the art from conventional calibration methods.

In the present invention, it can be understood as a task for extracting and minimizing a portion where the DI contour image does not match the target layout.

The etched layer is then etched using the photoresist pattern as an etch mask to form an etched layer pattern.

In this case, the etched layer pattern may not have the same shape as the target layout due to various process variables including an etching process.

The FI image of the present invention is an image obtained by measuring the etched layer pattern 60 as shown in FIG. 4A, wherein the FI image is most preferably measured by a scanning electron microscope like the DI image measuring method.

Next, the FI contour image is extracted (S30).

At this time, the FI contour image should be a dotted line connected to the image outline of the etched layer pattern as shown in FIG. 4B, but the noise patterns are exposed around the etched layer pattern, so that the FI contour image is a dotted line connecting the image outline of the etched layer pattern ( 70 as well as dotted lines 80 connected to the image outline of the noise pattern.

In this case, the noise pattern includes a lower pattern formed on the lower layer of the etched layer pattern.

Next, the image of the noise pattern except the image of the etched layer pattern is removed from the FI contour image (S40).

Since the FI contour image reflects both the image of the etched layer pattern and the image of the noise pattern, the FI contour image removes the noise pattern from the FI contour image and performs calibration using only information on the image of the etched layer pattern.

In this case, the step of removing the image of the etched layer pattern and the image of the noise pattern from the FI contour image may be performed through the following process.

Matching the target layout of the noise pattern with the FI contour image, and removing information on the image of the noise pattern from the FI contour image while leaving information on the image of the etched layer pattern in the FI contour image.

As a result, the FI contour image is separated from the image of the etched layer pattern and the image of the noise pattern at the same time, so that only necessary information can be extracted and used separately.

Next, the FI contour image and the target layout are aligned (S50).

In this case, the FI contour image aligned with the target layout becomes a FI contour image in which the image of the noise pattern is removed in step S40.

Next, a second calibration is performed using the alignment information on the FI contour image and the target layout (S60).

That is, since the second calibration is performed using information obtained by matching only the image of the etched layer pattern with the target layout in the FI contour image, the calibration can be accurately performed without the influence of the noise pattern.

In this case, the alignment information may be an error of alignment between the image of the etched layer pattern and the target layout.

Next, modeling is performed using the above-described first and second calibrations (S70).

That is, modeling reliability may be improved by including not only the first calibration using the DI contour image but also the second calibration data using the alignment information on the etched layer pattern and the target layout.

In this case, modeling is a process of extracting an equation or the like for describing that the layout is implemented on the actual wafer.

It is readily available to those skilled in the art from conventional modeling methods.

Herein, it can be understood as a process for describing a layout in which the mismatched parts are minimized through the first and second calibrations on the actual wafer.

The first optical proximity effect compensation is performed using the modeling data obtained as described above (S80).

Next, a simulation is performed on the layout in which the first optical proximity effect compensation is completed (S90).

In this case, the simulation may predict an image in which the layout in which the optical proximity effect compensation is completed is implemented on the actual wafer.

Next, the target layout and the simulated image are compared and verified (S100).

In this case, the simulation image may be referred to as a wafer image that can be implemented later through a photolithography process.

That is, the wafer image may include an image of the final pattern formed on the wafer through a photolithography process using an exposure mask.

If the simulation image does not match when compared with the target layout, the second optical proximity effect compensation is performed again, and the simulation image is compared with the target layout by comparing the simulation image and the target layout for the layout where the second optical proximity compensation is completed. In comparison, it is preferable to repeatedly perform the optical proximity effect compensation until the same shape is achieved so as to implement the desired target layout.

As described above, after performing the etching process by performing the calibration process using the DI contour image as well as the calibration using the FI contour image to compensate for the optical proximity effect, the modeling of the etched layer pattern is precisely performed so that the optical proximity effect compensation is more accurate. Make sure it's done correctly.

In addition, by verifying that the etched layer pattern close to the same shape as the target layout may be implemented through simulation, the characteristics of the semiconductor device may be improved by allowing the etched layer pattern to be implemented without distortion from the target layout of the etched layer pattern.

Figure 1a is a photograph showing a development inspection (DI) image of the prior art.

1b is a photograph showing a prior art final inspection (FI) image.

2 is a flowchart of an optical proximity effect compensation method according to the present invention;

Figure 3a is a photograph showing a development inspection (DI) image of the present invention.

Figure 3b is a photograph showing a DI contour image of the present invention.

4A is a photograph showing a final inspection (FI) image of the present invention.

Figure 4b is a photograph showing a FI contour image of the present invention.

Claims

Extracting a DI contour image using a development inspection (DI) image;

Performing a first calibration using the DI contour image;

Extracting a FI contour image using a final inspection (FI) image;

And modeling using the first and second calibration results and performing optical proximity effect compensation using the first and second calibration results.

The method of claim 1,

After performing the optical proximity effect compensation,

Extracting a simulation image by performing simulation on the layout in which the optical proximity effect compensation is completed; And

And repeatedly performing optical proximity effect compensation and simulation until the target layout matches the simulated image.

The method of claim 1,

The DI image is

And forming an etched layer on the semiconductor substrate, applying a photosensitive film on the etched layer, and performing an exposure and development process using an exposure mask on which the target layout is formed. Optical proximity effect compensation method.

The method of claim 3, wherein

The DI contour image is

The optical proximity effect compensation method comprising a line connecting the outer portion of the image measured the photosensitive film pattern.

The method of claim 1,

The FI image is

And an image obtained by measuring the etched layer pattern and the noise pattern.

The method of claim 5,

The noise pattern is

And a lower pattern formed on the lower layer of the etched layer pattern.

The method of claim 1,

The FI contour image is

And a line connecting an outer portion of the etched layer pattern image and a line connecting an outer portion of the noise pattern image.

The method of claim 6,

Removing the image of the noise pattern except the image of the etched layer pattern from the FI contour image

Matching the target layout of the lower pattern with the FI contour image, and removing information on the image of the subpattern from the FI contour image while leaving information on the image of the etched layer pattern in the FI contour image; Optical proximity effect compensation method characterized in that.

The method of claim 1,

Performing the second calibration is

And aligning the target layout with the FI contour image from which the image of the noise pattern has been removed.