CN-121986593-A - Method for automatically detecting, classifying and removing defects of flat substrate

CN121986593ACN 121986593 ACN121986593 ACN 121986593ACN-121986593-A

Abstract

A method for automatically detecting, classifying and removing defects of a flat substrate is provided, and includes a first step of photographing a surface of the flat substrate including a wafer or a mask by a photographing unit to detect defects, a second step of determining whether the defects can be removed and one of a plurality of removing methods of the defects according to a preset reference and automatically classifying the defects, a third step of moving a defect removing module to a target position where the defects are located if the defects are of a removable type, a fourth step of selecting one of the defect removing modules to remove the defects according to the defect removing method, and a fifth step of confirming whether the defects are removed.

Inventors

JIN XIANZHU
LI ZHENYUAN
CHENG MINGJUN
ZHAO DEGAO
Kang Hulong
JIN YINGZHI

Assignees

株式会社考恩斯特

Dates

Publication Date: 20260505
Application Date: 20240829
Priority Date: 20240528

Claims (7)

1. A method for automatically detecting, classifying and removing defects in a flat substrate, comprising: a first step of photographing a surface of a flat substrate including a wafer or a mask by an image pickup unit to detect defects, A second step of determining whether the defect can be removed and one of a plurality of removal methods of the defect according to a preset reference, and automatically classifying the defect, A third step of moving a defect removal module to a target position where the defect is located if the defect is of a removable type, A fourth step of selecting one unit of the defect removal module to remove the defect according to the defect removal method, and And a fifth step of confirming whether the defect is removed.
2. The method for automatically detecting, classifying and removing defects of a flat substrate according to claim 1, In the first step, the defect is detected by comparing the following images with each other: 1) A reference image which is an image obtained by preprocessing an original image taken on the surface of the flat substrate, and 2) An image to be compared, which is an image in which a difference in brightness is minimized by applying a color filter to a scanned image obtained by color scanning a surface of a flat substrate.
3. The method for automatically detecting, classifying and removing defects of a flat substrate according to claim 1, In the second step, the removing method includes one of the following removing methods according to the type of the defect: a first removing method of injecting a gas toward the target position and a peripheral portion thereof, A second removing method of irradiating a laser beam to the target position, and A third removal method uses both the gas and the laser beam.
4. The method for automatically detecting, classifying and removing defects of a flat substrate according to claim 3, In the third step, the defect removal module includes: and a purge unit that, in response to the first removal method, desorbs the defect (foreign matter) in a process of rapidly contracting and then rapidly expanding the defect by performing physical collision of the gas on the defect.
5. The method for automatically detecting, classifying and removing defects of a flat substrate according to claim 3, In the third removing method, the laser beam is irradiated to remove the defect while the gas is sprayed to cool the flat substrate.
6. The method for automatically detecting, classifying and removing defects of a flat substrate according to claim 1, Between the fourth step and the fifth step, further comprising: step 4-1, sucking in the suspended matter generated during the removal of the defect.
7. The method for automatically detecting, classifying and removing defects of a flat substrate according to claim 6, The suspension is sucked in by the suction unit, A suction portion is formed at one side of the suction unit, the suction portion being recessed toward the inside of the suction unit, and the suction portion having a circular cross section so that suspended matter is sucked in a radial direction.

Description

Method for automatically detecting, classifying and removing defects of flat substrate Technical Field More particularly, the present invention relates to a method for automatically performing the entire process of inspecting a flat substrate including a wafer or a mask, determining various defects on the flat substrate and classifying the defects accordingly, and finally removing the defects. Background In general, an optical inspection device is used to inspect a base material such as a semiconductor wafer for defects. Conventionally, optical inspection of a semiconductor wafer is performed by a method of photographing a black-and-white image with a monochrome (Mono) camera and comparing the photographed images. However, this method has a problem that it is difficult to accurately determine various types of defects, and thus automatic classification of defects is difficult and reliability of detection is lowered. In addition, when defects are found on the wafer, removal may be performed by a cleaning device. However, this method relies on manual operations by operators, or if it cannot be removed, the wafers are directly discarded, resulting in problems of reduced yield of products, loss of manufacturing costs, and the like. Further, the operator inevitably participates in the entire process (i.e., from wafer inspection to classification and removal of defects according to the result), and causes problems of poor productivity, labor consumption, and the like according to the proficiency thereof, and the like. This is also similar for masks including photomasks and the like. Disclosure of Invention Problems to be solved by the invention The embodiments of the present invention have been made to solve the above-described problems, and an object thereof is to provide a method capable of automatically detecting defects on a flat substrate including a wafer or a mask, automatically classifying the defects, and automatically removing the defects. The invention aims to improve the reliability of a detection result of a flat substrate by an algorithm capable of minimizing erroneous judgment in a detection step. In the classifying step, defects can be accurately classified according to a preset reference. In the removing step, it is intended to provide a module capable of selecting different modes according to the type of defects on the flat substrate. The invention aims to provide a module capable of preventing suspended matters and the like generated in the defect removing process from remaining on a flat substrate. Means for solving the problems In order to solve the problems described above, an embodiment of the present invention provides a method for automatically detecting, classifying and removing defects of a flat substrate, which includes a first step of photographing a surface of the flat substrate including a wafer or a mask by an image pickup unit to detect defects, a second step of determining whether the defects are removable and one of a plurality of removal methods of the defects according to a preset reference and automatically classifying the defects, a third step of moving a defect removal module to a target position where the defects are located if the defects are removable, a fourth step of selecting one of the defect removal modules according to the defect removal method to remove the defects, and a fifth step of confirming whether the defects are removed. In the first step, the defect is detected by comparing 1) a reference image, which is an image obtained by preprocessing an original image obtained by photographing the surface of the flat substrate, and 2) an image to be compared, which is an image obtained by applying a color filter to a scanned image obtained by color scanning the surface of the flat substrate, with each other. Preferably, in the second step, the removing method includes one of a first removing method of spraying a gas toward the target position and a peripheral portion thereof, a second removing method of irradiating a laser beam to the target position, and a third removing method of simultaneously using the gas and the laser beam according to the type of the defect. Preferably, in the third step, the defect removing module includes a purge unit, responsive to the first removing method, for desorbing the defect (foreign matter) in a process of rapidly contracting and then rapidly expanding the defect by performing physical collision of the gas to the defect. Preferably, in the third removing method, the gas is sprayed to cool the flat substrate while the laser beam is irradiated to remove the defect. Preferably, a 4-1 th step of sucking in suspended matters generated during the removal of the defect is further included between the fourth step and the fifth step. Preferably, the suspended matter is sucked by a suction unit, a suction portion is formed at one side of the suction unit, the suction portion is recessed toward the inside of the suction unit, and the suction p