TWI439798B - Method of inspecting photomask defect - Google Patents

Description

Method for detecting mask defects

The present invention relates to a method of detection, and more particularly to a method of detecting a defect in a reticle.

In the integrated circuit process, the pattern of each material layer on the wafer or on the wafer is mainly formed by the transfer of the mask pattern. A typical method is to first form a photoresist layer on a wafer, and then project a pattern of the mask onto the photoresist layer through an exposure process, and then form a desired pattern in the photoresist layer through a development process, and thereafter, via the formation of a desired pattern in the photoresist layer. The etch process transfers the pattern in the photoresist layer to a layer of material on the wafer or wafer.

As components continue to become smaller, integrated circuits become more complex, and the accuracy of the pattern on the reticle becomes more and more important. The current reticle manufacturing technology does not guarantee that the reticle formed is perfect. Usually, the new reticle that has just been fabricated still has defects and must be tested to determine whether the reticle is acceptable, needs to be repaired or must be reworked. On the other hand, even if the finished reticle is acceptable, in order to ensure the stability of the process before the wafer patterning process, the reticle used must also be tested to avoid the reticle process. Defects generated in the process affect the accuracy of the pattern to be formed.

A typical method for detecting a defect of a reticle is to form a photoresist layer on a wafer, perform an exposure and development process, and directly transfer the pattern of the reticle to the photoresist layer, thereby judging the reticle by the pattern in the photoresist layer. Whether to pass the test. However, this method is very time consuming and uneconomical and does not meet the demand.

SUMMARY OF THE INVENTION The present invention is directed to a method of detecting a defect in a reticle that can be detected without actually performing an exposure and development process on the wafer.

SUMMARY OF THE INVENTION The present invention is directed to a method for quickly detecting reticle defects.

SUMMARY OF THE INVENTION The present invention is directed to a method for economically and efficiently detecting reticle defects.

The present invention provides a method of detecting a defect in a reticle. This method includes creating a database that stores a plurality of immersive modules. The immersive module is established by a correction process that actually acquires data from the wafer, and the correction process includes baking parameters. Next, a mask defect image is obtained. Then, the simulated contour of the reticle defect image is generated from the immersive module of the database, and then the acceptability of the reticle is determined.

According to the embodiment of the present invention, in the method for detecting a defect of the reticle, the factor considered in the process of establishing the immersive module further includes a machine effect.

In the above method for detecting a reticle defect, the immersive module includes a plurality of optical proximity effect correction simulation modules.

According to an embodiment of the invention, the method for detecting a defect of the reticle further comprises performing an optical rule check after generating the simulated contour of the reticle defect image.

According to the embodiment of the invention, in the method for detecting a defect of the reticle, the optical rule check comprises checking the overlapping of the simulated image of the reticle defect image and the layout of the front reticle, or verifying the simulated image of the reticle defect image. The overlap of the rear reticle layout.

According to the embodiment of the invention, in the above method for detecting a defect of the reticle, the step of determining the acceptability of the reticle is determined according to the result of the optical rule inspection.

According to the embodiment of the invention, in the method for detecting a defect of the reticle, after obtaining the reticle defect image, the method further includes digitizing the reticle defect image.

According to the embodiment of the invention, in the method for detecting a defect of the reticle, the step of generating the simulated contour of the reticle defect image is generated based on all the digital reticle defect images.

According to the embodiment of the invention, in the method for detecting a defect of the reticle, the step of generating the simulated contour of the reticle defect image is generated according to a part of the digitized reticle defect image.

According to the embodiment of the invention, in the method for detecting a defect of the reticle, the step of determining the acceptability of the reticle is determined according to the simulated contour of the reticle defect image generated.

According to an embodiment of the invention, in the method for detecting a defect of a reticle, the reticle defect image includes a defect image.

According to an embodiment of the invention, in the method for detecting a defect of a reticle, the reticle defect image includes a geometric layout image and a defect image.

According to an embodiment of the invention, in the method for detecting a reticle defect, the geometric layout image includes an optical proximity image corrected geometric layout image.

According to an embodiment of the invention, in the method for detecting a defect of a reticle, the geometric layout image includes a geometric layout image that is not corrected by an optical proximity effect.

According to the embodiment of the invention, in the method for detecting a defect of the reticle, before the reticle defect image is obtained, the defect of the reticle is further detected to obtain the defect image.

According to an embodiment of the invention, the method for detecting a defect of the reticle further comprises storing the digital reticle defect image in a database.

The method for detecting a defect of the reticle of the present invention can be detected without actually performing an exposure and development process on the wafer.

The method of detecting a defect of the reticle of the present invention can quickly detect whether the defect on the reticle is acceptable.

The method of detecting reticle defects of the present invention is time-saving, economical, and effective.

The above and other objects, features and advantages of the present invention will become more <RTIgt;

The method for detecting the defect of the reticle according to the embodiment of the present invention is to obtain the reticle defect image by using the actual parameter conditions such as exposure, development and baking parameters which are subsequently scheduled, by having the actual data acquired from the wafer. Correct the database of immersive modules established by the process to simulate the actual process of the wafer, such as exposure and development, to determine whether the reticle can pass or must be repaired or reworked. A detailed description of the embodiments is given below.

FIG. 1 is a block diagram of a method for detecting a defect of a reticle according to an embodiment of the invention.

Referring to Figure 1, block 102 detects defects in the provided reticle. The reticle provided may be a reticle that has just been fabricated by the reticle factory or a reticle that has been actually processed in a semiconductor factory. The step 102 of detecting the reticle defect can use the inspection machine to find the light through the alignment between the wafer and the wafer, or by the interference of the light beam passing through the wafer, or by comparing the design database. A potential defect on the cover. If the inspection machine detects that there is no defect on the reticle, it means that the inspection of the reticle passes, and the pattern on the reticle can be directly printed on the wafer for subsequent lithography process.

If the inspection machine detects a defect on the reticle, it is necessary to further judge whether the defect on the reticle is negligible or must be repaired or reworked. Typically, the inspection machine used can output the detected reticle defect image, causing the executor to obtain a reticle defect image, block 104, for subsequent detection steps. The reticle defect image described herein may include a geometric layout image as well as a defective image or only a defective image. The geometric layout image may be a geometric layout image that is not corrected by optical proximity effects, such as a less complex or less dense integrated circuit, the geometry of the reticle is usually the same as the circuit pattern of the corresponding layer on the wafer. The same, without being corrected by the optical proximity effect. The geometric layout image can also be a geometric layout image corrected by optical proximity effect. For example, a complicated or dense integrated circuit, the size of the circuit has almost reached the optical limit. To compensate for the proximity effect, the mask is usually Some optical proximity correction patterns, such as serifs, hammerheads, biases, and aids bars, are included. The defective image may be an image of various possible defects, and the defect may be located in an opaque region or a light transmitting region of the reticle. For example, a pinhole defect or an edge protruding defect located on an opaque region, or a spot defect or an edge intrusion defect located on a light transmitting region. Or a geometry bridge defect that connects the two opaque regions on the opaque region, or a geometric break defect that cuts the light-transmitting region into two parts.

After obtaining the reticle defect image, the reticle defect image may be digitized into a data file to facilitate subsequent immersive simulation, block 106. The stored data may include geometric layout images as well as the penetration, size, and shape of the defective image. There is no restriction on the format of the data file. Thereafter, the digitized mask defect image can be stored in the database.

Thereafter, block 108 is performed to process the data of the digitized mask defect image and the actual parameter conditions to be subsequently performed, such as the parameter conditions of the exposure process to be subsequently scheduled, the parameter conditions of the subsequent development process, and the subsequent baking process. The parameter conditions and the difference of the machine are input to the image simulation device to generate a simulated contour of the reticle defect image from at least one immersive module in the database of the image simulation device. The data of the digitized mask defect image input may be all digitalized mask defect image data on the mask. The data of the digitized mask defect image input may also be a part of the digitized mask defect image data, and the method may select a part of the digitized mask defect image as a predetermined simulation area. Then, the simulated contour of the reticle defect image is generated according to the predetermined simulation area.

The input exposure parameters may include the numerical aperture (NA) of the exposure system, the coherency value of the exposure system, the wavelength and dose of the source used by the exposure system, the defocus of the exposure, and the lens aberration (lens aberration). ), wafer conditions, critical dimensions, etc. The input development parameters may include the kind of the developer, the concentration of the developer, the time of development, the temperature, and the like. The baking parameters entered may include the number of bakings, the temperature of baking, and the time of baking.

A plurality of sets of immersive modules are stored in the database of the image simulation device. The immersive module is the one that established the correction process for actually obtaining data from the wafer. Typically, the immersive module of the database is created prior to mask inspection, block 100. The method of establishing the immersive module can firstly use the test reticle with various patterns to carry out the pattern of the photoresist layer on the wafer according to various exposure, development and baking parameters according to different exposure and development machines. Chemical. Exposure parameters may include the numerical aperture of the exposure system, the coherence value of the exposure system, the wavelength of the source used by the exposure system, as well as energy, exposure defocus, lens aberrations, wafer conditions, critical dimensions, and the like. The development parameters may include the kind of the developer, the concentration of the developer, the time of development, the temperature, and the like. The baking parameters may include the number of bakings, the temperature of baking, and the time of baking. After that, the relevant information of the patterned photoresist layer on the wafer is measured, and then the relevant data is stored to establish a plurality of immersive modules. In an embodiment, the immersive module includes an optical proximity correction analog module. Since these immersive modules are all established by the correction process, that is, the correction process is the data actually transferred from the pattern on the test reticle to the wafer. Therefore, the actual situation faced by the pattern on the reticle transferred to the photoresist layer on the wafer, including optical effects, chemical effects, and machine effects, has been fully considered. The optical effects are, for example, the scattering phenomenon caused by the light beam passing through the reticle, the interference phenomenon caused by the light beam passing through different patterns on the reticle, and the reflected light generated by the uneven wafer. Chemical effects include the effects of developing and baking parameters. The machine effect indicates the impact of the difference between different machines. Therefore, after the immersive simulation, the simulated contour of the reticle defect image generated in the database of the image simulation device can simulate the pattern on the reticle being transferred onto the wafer.

Thereafter, based on the simulated profile of the resulting reticle defect image, the effect of the possible defect area on the process can be known to determine the acceptability of the reticle, block 112. When the reticle passes, a subsequent lithography process can be performed. When the mask does not pass, it can be repaired or reworked as needed.

In general, defects on the reticle may not only affect the layer of the wafer, but may also affect the upper and lower layers of the layer. Therefore, after generating the simulated contour of the reticle defect image, a Lithography Rule Check (LRC) may be performed, block 110. The optical rule inspection step includes verifying the overlap of the simulated image of the reticle defect image of the layer with the reticle layout of the front layer, or verifying the stack of the simulated image of the reticle defect image of the layer and the reticle layout of the back layer For the situation. After the optical rule test, the acceptability of the mask can be determined based on the results of the optical rule test.

The method for detecting a defect of the reticle of the present invention can be detected without actually performing an exposure and development process on the wafer.

The method of detecting a defect of the reticle of the present invention can quickly detect whether the defect on the reticle is acceptable.

The method of detecting reticle defects of the present invention is time-saving, economical, and effective.

Although the present invention has been disclosed in the above embodiments, it is not intended to limit the present invention, and the present invention may be modified and modified without departing from the spirit and scope of the invention. The scope is subject to the definition of the scope of the patent application attached.

100~112. . . Step block

FIG. 1 is a block diagram of a method for detecting a defect of a reticle according to an embodiment of the invention.

100~112. . . Step block

Claims (16)

A method for detecting a defect of a reticle includes: establishing a database storing a plurality of immersive modules, wherein the immersive modules are actually established by a correction process for obtaining data from the wafer, and the correction process is established Including baking parameters; obtaining a mask defect image; generating an analog contour of the mask defect image from the immersive modules of the database; and determining the acceptability of the mask. The method for detecting reticle defects according to claim 1, wherein the factors considered by the immersive modules in the process of establishment include a machine effect. The method for detecting a reticle defect according to claim 1, wherein the immersive modules comprise a plurality of optical proximity correction analog modules. The method for detecting a reticle defect as described in claim 3, further comprising performing an optical rule inspection after generating the simulated contour of the reticle defect image. The method of detecting a reticle defect according to claim 4, wherein the optical rule inspection comprises verifying a superposition of an analog image of the reticle defect image and a front reticle layout, or verifying the reticle defect The overlapping of the image of the image with the layout of a back mask. The method of detecting a reticle defect as described in claim 5, wherein the step of determining the acceptability of the reticle is determined based on the result of the optical rule inspection. The method for detecting a reticle defect according to claim 1, wherein the obtaining the reticle defect image further comprises digitizing the reticle defect image. The method of detecting a reticle defect as described in claim 7, wherein the step of generating an analog profile of the reticle defect image is generated based on all digitized reticle defect images. The method of detecting a reticle defect as described in claim 7, wherein the step of generating an analog profile of the reticle defect image is generated based on a portion of the reticle defect image. A method of detecting a reticle defect as described in claim 1, wherein the step of determining the acceptability of the reticle is determined based on the simulated contour of the reticle defect image produced. The method of detecting a reticle defect according to claim 1, wherein the reticle defect image comprises a defect image. The method of detecting a reticle defect according to claim 1, wherein the reticle defect image comprises: a geometric layout image; and a defect image. The method of detecting a reticle defect according to claim 12, wherein the geometric layout image comprises an optical proximity image corrected geometric layout image. The method of detecting a reticle defect as described in claim 12, wherein the geometric layout image comprises a geometric layout image that is not corrected by optical proximity effects. The method for detecting a reticle defect according to claim 1, wherein the reticle defect image is detected before the reticle defect image is acquired to obtain the defect image. The method for detecting a reticle defect according to claim 1, further comprising storing the digital reticle defect image in the database.