JP2020204497A - Inspection device - Google Patents

Abstract

To provide an inspection device for performing inspection at high speed.SOLUTION: An inspection device in one embodiment of the present invention comprises: an inspection unit for obtaining the first inspection image of a first test area of a test object that includes a first inspection area and a second test area adjacent to the first test area in the scan direction and having a pattern common to the first test area, comparing the first inspection image with a reference image, and when a first defect is detected at the first coordinate of the firs test area, moving to the second coordinate of the second test area that corresponds to the first coordinate of the first test object during inspection of the first test area and obtaining a second inspection image of only a pattern of the second coordinate; and a determination unit for determining, from the first inspection image at the first coordinate and the second inspection image at the second coordinate, whether the first defect is a defect caused by the mask used for drawing the first test area or a defect attributable to a process for drawing the first test area.SELECTED DRAWING: Figure 1

Description

The present invention relates to an inspection device.

With the increasing integration and capacity of large scale integrated circuits (LSIs), the circuit dimensions required for semiconductor devices are becoming narrower. For example, in a logic device, pattern formation with a line width of 10 nm or less is required.

Improving the yield in the manufacturing process is indispensable for LSIs that require a large manufacturing cost. Here, in the semiconductor element, an original image pattern (referring to a mask or a reticle, hereinafter collectively referred to as a mask) in which a circuit pattern is formed by a reduced projection exposure device called a stepper or a scanner in the manufacturing process thereof is on the wafer. It is exposed and transferred to. The major factors that reduce the yield of the semiconductor element (chip) include defects caused by the mask pattern and defects caused by the process.

In cases such as EUV (Extreme Ultraviolet) masks, where the pattern is fine and it is difficult to detect defects with conventional optical system inspection equipment, defects are inspected by EB (Electric Beam) inspection after transfer to the wafer, and this wafer defect inspection data is used. It has been proposed to feed back to mask defects. It is necessary to distinguish whether the defects found in the wafer defect inspection are caused by the mask or other processes, but the defects caused by the mask also occur on each chip transferred on the wafer. , Separation is performed by inspecting multiple chips as a whole. However, since this method requires a plurality of inspections, an increase in inspection time is an issue.

JP-A-2017-090063

Therefore, one aspect of the present invention provides an inspection device.

The inspection device of one aspect of the present invention includes a first inspected area and a second inspected area which is adjacent to the first inspected area in the scanning direction and has a pattern common to the first inspected area. The first inspection image of the first inspection area to be inspected is obtained, the reference image is compared with the first inspection image, and the first defect is detected at the first coordinates of the first inspection area. When this is done, during the inspection of the first inspected area, the second coordinate of the second inspected area corresponding to the first coordinate of the first inspected area is moved to only the pattern of the second coordinate. The inspection unit for obtaining the second inspection image and the process for drawing the first inspection area, whether the first defect is a defect caused by the mask used for drawing the first inspection area or the first inspection area. An inspection device characterized by having a determination unit for determining whether or not the defect is caused from the first inspection image at the first coordinate and the second inspection image at the second coordinate is preferable.

In addition, the first criterion for determining whether the first defect is a mask-induced defect or a process-induced defect is used as a result of determining whether the first defect is a mask-induced defect or a process-induced defect. A calculation unit for changing to the second reference is further provided, the object to be inspected includes a third area to be inspected having a pattern in common with the first area to be inspected, and the inspection unit is a third area to be inspected. A third inspection image is obtained, the reference image is compared with the third inspection image, and the second defect is detected.

The determination unit determines whether the second defect is a mask-induced defect used to draw the third inspected area using the second criterion or is a process-derived defect for drawing the third inspected area. An inspection device characterized by determining whether or not it is a defect is preferable.

Further, the determination unit uses the defect information of the inspection data of the mask that draws the first area to be inspected, and whether the first defect is a defect caused by the mask used to draw the first area to be inspected. An inspection device characterized by determining whether or not the defect is caused by a process for drawing the first area to be inspected is preferable.

In addition, the first criterion for determining whether the first defect is a mask-induced defect or a process-induced defect is used as a result of determining whether the first defect is a mask-induced defect or a process-induced defect. A calculation unit for changing to the third standard is further provided, the inspection unit detects a third defect other than the first defect included in the first inspected area, and the determination unit uses the third standard. Therefore, an inspection device characterized in determining whether the third defect is a mask-induced defect or a process-induced defect is preferable.

Further, it is preferable that the inspection apparatus further includes a reference image correction unit that reflects the mask-induced defect information obtained in the determination unit on the reference image and adds the mask-induced defect information to the reference image.

According to one aspect of the present invention, it is possible to provide an inspection device for performing high-speed inspection.

It is a block diagram of the inspection apparatus in 1st Embodiment. It is a schematic diagram of the object to be inspected in 1st Embodiment. It is an enlarged schematic diagram of a part of the subject to be inspected in 1st Embodiment. It is a schematic diagram of the 1st inspection image and the 2nd inspection image in 1st Embodiment. It is an enlarged schematic diagram of a part of the subject to be inspected in 1st Embodiment. It is a flowchart of the inspection method in 1st Embodiment. It is an enlarged schematic diagram of a part of the subject to be inspected in the 2nd Embodiment. It is an enlarged schematic diagram of a part of the subject to be inspected in the 2nd Embodiment. It is a flowchart of the inspection method in 2nd Embodiment. It is a flowchart of the inspection method in 2nd Embodiment. It is a schematic diagram of the inspection image of the mask in the 3rd Embodiment. It is a flowchart of the inspection method in 3rd Embodiment. It is an enlarged schematic diagram of a part of the subject to be inspected in 4th Embodiment. It is a flowchart of the inspection method in 4th Embodiment. It is a flowchart of the inspection method in 4th Embodiment. It is a schematic diagram of the reference image in 5th Embodiment. It is a flowchart of the inspection method in 5th Embodiment.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

Hereinafter, in the embodiment, a case where an electron beam is used as an example of a charged particle beam will be described. However, it is not limited to this. Other charged particle beams such as ion beams may be used.

The inspection device of one aspect of the present invention includes a first inspected area and a second inspected area which is adjacent to the first inspected area in the scanning direction and has a pattern common to the first inspected area. The first inspection image of the first inspection area to be inspected is obtained, the reference image is compared with the first inspection image, and the first defect is detected at the first coordinates of the first inspection area. When this is done, during the inspection of the first inspected area, the second coordinate of the second inspected area corresponding to the first coordinate of the first inspected area is moved to only the pattern of the second coordinate. The inspection unit for obtaining the second inspection image and the process for drawing the first inspection area, whether the first defect is a defect caused by the mask used for drawing the first inspection area or the first inspection area. It has a determination unit for determining whether or not the defect is caused from the first inspection image at the first coordinate and the second inspection image at the second coordinate.

(First Embodiment)
FIG. 1 is a configuration diagram showing a configuration of an inspection device according to the first embodiment. In FIG. 1, the inspection device 100 that inspects the pattern to be inspected is an example of a particle beam inspection device. The inspection device shown in FIG. 1 is a device that inspects with a single beam, but the inspection device may be a multi-beam inspection device that inspects with a multi-beam. The inspection device 100 includes a control unit 110 and an inspection unit 150.

The control unit 110 includes a control computer 121, a bus 122, a storage device 131, a monitor 132, a printer 133, a memory 134, a position calculation unit 135, a stage control unit 136, a deflection control unit 137, a blanking control unit 138, and a lens control unit 139. , A reference image creation unit 140, a determination unit 141, an image correction unit 142, and a calculation unit 143.

The inspection unit 150 includes a drive unit 151, a laser length measurement system 152, a detection unit 153, a pattern memory 154, an inspection image creation unit 155, an electron beam column 161 (electron lens barrel), and an inspection room 162. An electron gun 163, an electromagnetic lens 164, a blanking deflector 165, a limiting aperture array substrate 166, an electromagnetic lens 167, a main deflector 168, a sub deflector 169, and a detector 170 are arranged in the electron beam column 161. ..

In the examination room 162, for example, a stage 171 and a mirror 172 that can move on the xy plane are arranged. The target X to be inspected is arranged on the xy stage 171. The inspection target X is a wafer on which a pattern is formed, a mask for forming a pattern on the wafer, and the like. The inspection target X is arranged on the stage 171 with the pattern forming surface facing upward (electron gun 163 side), for example. Further, on the stage 171 is a mirror 172 that reflects the laser beam for laser length measurement emitted from the laser length measurement system 152 arranged outside the examination room 162.

The detector 170 is connected to the detection unit 153 outside the electron beam column 161. The detection unit 153 is connected to the pattern memory 154. The pattern memory 154 is connected to the inspection image creation unit 155.

In the control unit 110, the control computer 121, which is a computer, transmits the storage device 131, the monitor 132, the printer 133, the memory 134, the position calculation unit 135, the stage control unit 136, the deflection control unit 137, and the blanking control via the bus 122. It is connected to a unit 138, a lens control unit 139, a reference image creation unit 140, a determination unit 141, an image correction unit 142, and a calculation unit 143. Further, the control computer 121 is connected to the inspection image creating unit 155 via the bus 122. The inspection result and control information are displayed on the monitor 132. The inspection result can be printed by the printer 133.

The stage control unit 136 is connected to the drive unit 151. The stage control unit controls the drive unit 151. The position of the stage 171 is moved by the drive unit 151 controlled by the stage control unit 136. The drive unit 151 is composed of a drive system such as a three-axis (xy−θ) motor that drives in the x direction, the y direction, and the θ direction, for example. With these drive systems, the stage 171 can be moved in any direction. As the x-motor, y-motor, and θ-motor (not shown), for example, a step motor can be used. The stage 171 can be moved in the horizontal direction and the rotational direction by the motor of each axis of xyθ. Then, the moving position of the stage 171 is measured by the laser length measuring system 152 and supplied to the position calculation unit 135, and the positions of the stage 171 and the inspection target X are calculated. The laser length measuring system 152 measures the position of the stage 171 by the principle of the laser interferometry method, for example, by receiving the reflected light from the mirror 172.

A high-voltage power supply circuit (not shown) is connected to the electron gun 163, and an acceleration voltage from the high-voltage power supply circuit is applied between the filament (not shown) in the electron gun 163 and the extraction electrode, and a voltage of a predetermined extraction electrode is applied and predetermined. By heating the cathode (filament) at the temperature of, the group of electrons emitted from the cathode is accelerated and emitted as an electron beam. Both the electromagnetic lens 164 and the electromagnetic lens 167 are controlled by the lens control unit 139. The blanking deflector 165 is composed of a group of electrodes having at least two poles, and is controlled by the blanking control unit 138. The main deflector 168 and the sub deflector 169 are each composed of a group of electrodes having at least four poles, and are controlled by the deflection control unit 137.

When the object X to be inspected is a semiconductor wafer on which a plurality of chip (die) patterns are formed, the pattern data of the chip (die) patterns is input from the outside of the inspection device 100 and stored in the storage device 131. .. When the object X to be inspected is an exposure mask, design pattern data that is a basis for forming a mask pattern on the exposure mask is input from the outside of the inspection device 100 and stored in the storage device 131. Further, the storage device 131 stores the past inspection image, the defect pattern and the type of the defect (caused by the mask, the cause of the process), the defect information of the mask used for drawing the inspection target X, and the like. Is preferable.

Here, FIG. 1 describes a configuration necessary for explaining the first embodiment and other embodiments. The inspection apparatus 100 may usually have other necessary configurations. The calculation unit 143 and the image correction unit 142 in FIG. 1 will be described in another embodiment.

The primary electron beam 200 emitted from the electron gun 163 (emission source) is refracted by the electromagnetic lens 164 and proceeds to the electromagnetic lens 167 while forming an intermediate image and a crossover. Then, the electromagnetic lens 167 focuses the primary electron beam 200 on the inspection target X. The primary electron beam 200 focused on the X-plane to be inspected by the electromagnetic lens 167 is deflected by the main deflector 168 and the sub-deflector 169, and is irradiated to a desired position on the X-plane to be inspected. Here, when the primary electron beam 200 is irradiated to the object X to be inspected while the stage 171 continuously moves, the irradiation position of the primary electron beam 200 is deflected by the main deflector 168 so as to follow the movement of the stage 171. A tracking operation is performed.

When the primary electron beam 200 is deflected by the blanking deflector 165, the position is displaced from the central hole of the limiting aperture array substrate 166 like a long chain line, and the primary electron beam 200 is shielded by the limiting aperture array substrate 166. To. On the other hand, the primary electron beam 200 not deflected by the blanking deflector 165 passes through the central hole of the limiting aperture array substrate 166 as shown by the solid line shown in FIG. By turning ON / OFF of the blanking deflector 165, blanking control is performed and ON / OFF of the beam is controlled. In this way, the limiting aperture array substrate 166 shields the primary electron beam 200 deflected by the blanking deflector 165 so that the beam is turned off. Then, the primary electron beam 200 for inspection (for image acquisition) is formed by the beam that has passed through the limiting aperture array substrate 166 formed from when the beam is turned on to when the beam is turned off.

When the primary electron beam 200 is irradiated to a desired position of the target X to be inspected, the primary electron beam 200 corresponds to the primary electron beam 200 from the target X to be inspected due to the irradiation of the primary electron beam 200. Secondary electrons (secondary electron beam 201) are emitted. The secondary electron beam 201 (broken line) may contain backscattered electrons.

The secondary electron beam 201 emitted from the subject X to be inspected is projected onto the detector 170. The detector 170 detects the projected secondary electron beam 201. The intensity signal, which is the secondary electron image data detected by the detector 170, is output to the detection unit 153.

FIG. 2 shows a schematic diagram of the object to be inspected X in the first embodiment. The target X to be inspected in FIG. 2 has a plurality of regions to be inspected having a common pattern represented by a rectangle. When the object X to be inspected is a wafer, the area to be inspected becomes a semiconductor chip when cut from the wafer. That is, the plurality of target areas to be inspected have substantially the same pattern except for some defects that are not common. In FIG. 2, two second inspected areas B1 and B2 sandwiching the first inspected area A and the first inspected area A are shown. The scanning direction of the first area A to be inspected at the time of inspection is the x direction. The area to be inspected adjacent to the arrow direction above the first area to be inspected A is the second area to be inspected B1. The area to be inspected adjacent to the lower side of the first area to be inspected A in the direction of the arrow is the second area to be inspected B2. The second inspected areas B1 and B2 are inspected areas adjacent to the scanning direction of the first inspected area A. When the scanning direction is other than the x direction, for example, the y direction, the second area to be inspected is arranged at a position different from that in FIG.

For example, a conventional inspection device inspects the entire inspection region a of the inspection target Z and the entire inspection region b adjacent to the inspection region a to evaluate defects. That is, in the conventional inspection method, it takes time to inspect two inspected areas in order to evaluate a defect in one inspected area.

FIG. 3 shows an enlarged schematic view of a part of the object X to be inspected according to the embodiment. The schematic diagram of FIG. 3 shows three regions to be inspected. The second inspected areas B1 and B2 are arranged so as to sandwich the first inspected area A. The first area to be inspected A is the area to be inspected. The second areas to be inspected, B1 and B2, are areas for confirmation. A cross pattern P, which is a pattern common to the three inspected areas, is drawn in each inspected area. In FIG. 3, defects d1 to d7 are indicated by black circles. The broken line in FIG. 3 is a portion where an image is acquired by scanning. The solid line in FIG. 3 is a portion where only movement is performed and no image is acquired. Arrows are drawn on the broken lines and solid lines, and the arrows indicate the scanning direction (movement direction). The first scan is the first scan S1 and up to the tenth scan S10 in FIG. It wraps after the first scan S1. Following the first scan S1, there is a second inspection S2 in the direction opposite to that of the first scan S1. The image is acquired, for example, in a region having a rectangular frame as a unit, but the frame is not shown in FIG. In the frame (not shown), for example, a plurality of rectangular frames in the x direction are continuously arranged from one end (left in the figure) of the area to be inspected to the other end (right in the figure) in the x direction. A plurality of these rectangular frames are continuously arranged in the y direction from one end (upper figure) to the other end (lower figure). The scanning direction is the forward direction of the inspection order of the inspection frames arranged in the inspection order, and is one direction in each scan (for example, S1 to S10, respectively). Since scanning is performed by folding back, the nth scanning direction and the n + 1th scanning direction are opposite directions.

A first inspection image of the first inspection area A is obtained. The first area to be inspected A in FIG. 3 contains three defects d1 to d3. The three defects d1 to d3 are those detected as defects by comparing the reference image with the first inspection image. The reference image will be described in detail later with reference to FIG. 6, but when performing a Data to Database inspection, the reference image creation unit 140 is based on the design data of the mask that draws the inspection target X (inspected area). Create a reference image at. In addition, when performing so-called Data to Data inspection, the past inspection image is used as a reference image.

The second area to be inspected B1 in FIG. 3 contains two defects d4 and d6. The second area to be inspected B2 contains two defects d5 and d7. The first coordinate of the first defect d1 in the first area A to be inspected is c1. In the second area B1 to be inspected, the second coordinate corresponding to the first coordinate c1 of the first defect d1 is c2. The second areas B1 and B2 to be inspected may not include defects. The defects in the second inspected areas B1 and B2 are used not for comparison with the reference image but for comparing the first inspected area A with the first inspection image obtained as the inspection target. Then, the type of defect in the first area A to be inspected is determined.

When the first defect d1 is detected at the first coordinate c1 in the first scanning S1, the inspection position is moved without performing the inspection. Then, the inspection unit 150 acquires an image of only the pattern of the second coordinate c2 of the second inspection region B1. Then, the second coordinate c2 is folded back after acquiring the image of the pattern, and moves without acquiring the image until it enters the region of the first inspected region A in the direction opposite to the first scanning S1. Since it is only, the inspection can be performed at a higher speed. Even if the speed of moving in the second area B1 to be inspected is the same as the speed of acquiring an image, the image is moved to the second coordinate c2 without changing the scanning direction, the image is acquired, the image is folded back, and the image is acquired again. Since it returns to the first area to be inspected A, the inspection of the embodiment is faster than inspecting the entire area B1 to be inspected in addition to the entire area A to be inspected. The inspection range at the second coordinate c2 is an area equal to or larger than the pattern of the first defect d1, and is determined in consideration of the shape, size, position, and the like of the first defect d1. All the defects in the first inspection area A are defined as the first defect (defect group), and these coordinates are defined as the first coordinates (coordinate group). Further, the coordinates corresponding to the first coordinates in the second inspected areas B1 and B2 are set as the second coordinates (coordinate group).

In the eighth scan S8, two first defects d2 and d3 are included. Let the coordinates of the first defect d2 be c3 and the coordinates of the first defect d3 be c4. The inspection of the coordinates c5 and c6 corresponding to the coordinates c3 and c4 of the second area to be inspected B2 adjacent to each other in the direction opposite to that of the first scan S1 is performed in the same manner as in the first scan S1. In the second area to be inspected B1 in FIG. 3, no defect is detected at the coordinate c5, and a defect d7 is detected at the coordinate c6. In the eighth scan S8 as well, except for the coordinates c5 and the coordinates c6, the inspection target X is moved without acquiring an image, so that the inspection time can be further shortened.

Here, with respect to a different inspection target from FIG. 3, shortening of the inspection time will be described using a mathematical formula. The length in the scanning direction (distance in the x direction of the area to be inspected) is L, the scanning speed during inspection is Vs, the time required for folding back is Ta, the total number of scans is N, and the moving speed during non-inspection is kVs (k × Vs). , K >> 1), the total number of frames in the x direction is M (M >> 1), and the number of frames required to inspect one defect is m (1 ≦ m << M). In the calculation, the distance between the areas to be inspected (between A and B1 and between A and B2) is set to zero in order to simplify the evaluation. The frame required to inspect one defect is a frame that is continuous in the x-direction, the y-direction, or the x-direction and the y-direction. The formula shown below is for comparing the inspection times, and does not indicate the exact inspection time. Further, since the defect to be inspected is the worst condition in the method of this embodiment, it is assumed that the defect is located at the right end of the inspection area A when scanning in the right direction and at the left end of the inspection area A when scanning in the left direction. In the conventional case, the inspection time Tio can be expressed as Tio = ((2L / Vs) + Ta) × N. The time Ta required for turning back is a very short time as compared with the inspection time. Next, in the case of the embodiment, the inspection time Tin can be expressed as Tin = ((L / Vs) + (2L / kVs) + ((m / M) × (L / Vs)) × N. Difference in inspection time. Is Tio-Tin. Tio-Tin = (1-((2 / k) + (m / M)) × (L / Vs) × N. Here, the defect is relative to the area to be inspected. Since it occupies a very small area, M >> m (for example, M is 100 and m is 3), and (m / M) ≈ 0. Further, the movement speed during non-inspection is the movement during inspection. K, which indicates how many times the speed is, is much faster than that at the time of inspection, and specifically, k >> 1 (for example, k ≧ 10). Then, Tio-Tin is one region to be inspected. It is an approximate value of (L / Vs) x N, which is close to the inspection time. The moving speed at the time of inspection is "the distance in the scanning direction for one frame" [the image of one frame is acquired and the adjacent frame is obtained. It is calculated by dividing by the time to travel to. Therefore, in the inspection of the embodiment, even if there are all defects at the end in the scanning direction of each scan, which is inefficient, the inspection is performed to nearly half of the conventional inspection time. It shows that the time can be shortened in many ways.

FIG. 4 shows a schematic diagram of an inspection image obtained by the above inspection. The image obtained with the first inspection area A as the inspection target is the first inspection image C. The inspection image including the whole image of the first inspection area A is referred to as inspection image E. The images of the second inspection areas B1 and B2 are three of the second inspection images D1, D2 and D3. The second inspection image D obtained by inspecting the confirmation area is the data of only the portion in which the defect is detected in the first inspected area A and the area corresponding to the periphery thereof, so that the inspection time It shows that it has been shortened.

By evaluating with the determination unit 141 using the obtained inspection image, whether the defect in the first inspected area A is a defect caused by the mask used for drawing the first inspected area A. Alternatively, it is determined whether the defect is caused by the process for drawing the first area to be inspected. In the area to be inspected drawn with a common mask, defects caused by the mask are commonly present. Further, as for the defect caused by the process, for example, the presence or absence of the defect differs depending on the area to be inspected. Information determined to be a defect caused by a mask can be used as defect information for a mask that draws an area to be inspected, or can be used as a criterion for determining a defect in an area to be inspected for more efficient defect inspection. Can be done.

Next, the inspection device 100 will be described in more detail by explaining the inspection method of the inspection device 100. FIG. 6 shows a flowchart of the inspection method. The inspection method in the first embodiment includes an inspection step (S101), a defect detection step (S102), a defect confirmation step (S103), and a defect determination step (S104). The operation of the inspection device 100 including the steps in the inspection method shown in FIG. 6 is controlled by the control computer 121.

<First inspection step (S101)>
In the inspection method, first, the first inspection step (S101) is performed. In the first inspection step (S101), the first electron beam 200 of the subject X to be inspected is irradiated with the primary electron beam 200, and the inspection image is taken from the emitted secondary electron beam 201 (including backscattered electrons). Is the process of obtaining.

An inspection is performed for each inspection frame of the first inspection area A, and an inspection image is created for each inspection frame. Specifically, the detection unit 153 converts the analog signal of the secondary electron beam 201 incident on the detector 170 into a digital signal. The detection unit 153 is, for example, a sensor such as CMOS or CCD. The obtained digital signal is stored in the pattern memory 154. Then, the inspection image creating unit 155 obtains the first inspection image C by combining the signal strength information and the inspection position information. The obtained first inspection image C is stored in the storage device 131 or the memory 134. It is created by the inspection image creation unit 155 of another inspection image.

<First defect detection step (S102)>
The first defect detection step (S102) is a step of detecting defects by comparing the first inspection image C including the inspection result for each inspection frame with the reference image. When performing a so-called Data to Database inspection, the reference image creation unit 140 creates a reference image from the design data of the mask that draws the inspection target X (inspection area). In addition, when performing so-called Data to Data inspection, the past inspection image is used as a reference image. In addition, for example, an image created by connecting a part or all of inspection images in a row of inspection frames continuously arranged in the scanning direction (x direction) can be used as the first inspection image C. Alternatively, the image of one inspection frame can be used as the first inspection image C. Therefore, although it depends on the base image, the reference image can be a part of the image data cut out.

The reference image and the first inspection image C are compared by an arbitrary algorithm, and for example, when the determination index is equal to or greater than the threshold value, it is determined that there is a defect (defect is detected). When a defect is detected, the first defect confirmation step (S103) and the defect type determination step (S104) are performed. When a defect is not detected in the first defect detection step (S102), the inspection target in the first defect detection step (S102) is the final scan of the area A to be inspected (10th scan S10 in FIG. 3). When the inspection is completed, when the defect is not detected in the first defect detection step (S102), the inspection is the final scan of the first inspection area A (10th scan S10 in FIG. 3). ), For example, the stage is moved so as to be the next scanning position, and the inspection step (S101) is performed again. Then, the inspection is repeated until the final scan is reached. The detected defect is designated as the first defect d1, and the position of the first defect d1 is stored in the storage device 131 or the memory 134 as the first coordinate c1.

<First defect confirmation step (S103)>
In the first defect confirmation step (S103), when a defect is detected in the first defect detection step (S102), the second inspected area B1 adjacent to the scanning direction of the first inspected area A Inspect (or B2). In the inspection of the second inspected area B1, only the image of the pattern of the second coordinate c2 of the second inspected area B1 corresponding to the first coordinate c1 of the first defect d1 is acquired.

For example, it is assumed that a defect is found in the first inspection area A from the 50th frame to the 53rd frame of the first scan S1. At this time, the position corresponding to the 50th to 53rd frames of the first scan S1 of the second area B1 to be inspected is set as the second coordinate c2. Then, in this example, inspecting only the three frames from the 50th to the 53rd of the second inspected area B1 is "a second inspected area corresponding to the first coordinate c1 of the first defect d1". It means "inspect only the pattern of the second coordinate c2 of B1 (B2)". Then, in this example, the image of only three frames from the 50th to the 53rd of the second inspection area B1, that is, the second inspection corresponding to the first coordinate c1 of the first defect d1. An image of only the pattern of the second coordinate c2 of the region B1 is referred to as a second inspection image D (second inspection image D1 in FIG. 4). If a defect is detected while scanning in the direction opposite to the scanning direction described above, the image is acquired by moving to the coordinates corresponding to the defect in the second area B2 to be inspected.

<First defect type determination step (S104)>
In the first defect type determination step (S104), the first inspection image C and the second inspection image D are compared, and the first defect d1 is "used to draw the first inspection area A". It is determined whether the defect is caused by the mask that was present or is caused by the process of drawing the first area A to be inspected. The first criterion (detection parameter) is used for this determination. In determining the type of defect, the positional relationship between the pattern P and the defect, whether or not the defect is detected in the second inspection image D for confirmation, the shape of the defect, the size of the defect, and the like are taken into consideration. Do. Determining the type of defect is done by an arbitrary algorithm. The type of defect optionally includes information such as the positional relationship between the pattern P and the defect, whether or not the defect is detected in the second inspection image D for confirmation, the shape of the defect, and the size of the defect. It is preferable to store in.

After performing the first defect type determination step (S104), when the inspection target in the first defect type determination step (S104) is not the final scan of the area A to be inspected (10th scan S10 in FIG. 3). For example, the stage is moved and turned back so as to be at the next scanning position, and the object X to be inspected is moved relative to the primary electron beam 200 in the direction opposite to the direction of the first scanning S1. Then, the first inspection step (S101) is performed at the position of the second scan S2. Then, the inspection is repeated until the final scan (for example, the tenth scan S10) is reached.

When inspecting the first area A to be inspected in FIG. 3, scanning is performed 10 times from scanning S1 to scanning S10. Since no defects have been detected in the second scan S2 to the seventh scan S7, the movement to the second inspection areas B1 and B2 and the acquisition of images are not performed. Then, in the eighth scan S8, it is assumed that the area A to be inspected detects two defects d2 and d3. At this time, it moves to the second inspection area B2 adjacent to the first inspection area A in the scanning direction of the eighth scanning S8. The scanning directions of the first scan S1 and the eighth scan S8 are opposite. Depending on the scanning direction, only the pattern of the positions of the coordinates c5 and c6 of the second inspected area B2 corresponding to the coordinates c3 and c4 of the defects d2 and d3 of the first inspected area A is inspected, and the first inspected area is inspected. Confirm the defect in the inspection area A. As a result of this inspection, for example, no defect is detected at the coordinate c5, and a defect d7 is detected at the coordinate c6. Then, based on this inspection result, the types of defects d2 and d3 can be determined.

After the inspection is completed, the inspection image creating unit 155 can obtain an inspection image E including the entire image of the first inspection region A by partially overlapping and joining the obtained first inspection images C. .. In addition, you may join them without overlapping a part. The inspection image E including the entire image of the first inspection region A is preferably stored in the storage device 131.

FIG. 5 shows a schematic diagram of the area to be inspected. When the second area to be inspected B1 is arranged only in one scanning direction of the first area A to be inspected and the second area to be inspected B2 is not arranged in the opposite direction as shown in FIG. The inspection carried out in the second inspected area B2 is carried out in the second inspected area B1 by moving a part of the inspected second inspected area B1 in the −y direction during the inspection. , It may move in the y direction and return to the original scanning position. For example, the defects d2 and d3 detected in the first area A to be inspected during the scanning of the scanning S8 of FIG. 5 are adjacent to the scanning direction of the scanning S9 during the scanning of the next scanning S9. In the second inspected area B1, the second inspected area B1 corresponding to the coordinates c3 and c4 of the defects d2 and d3 detected in the first inspected area A by returning to the −y direction for one frame. A quick inspection can be performed by moving to the coordinates c5 and c6 and inspecting only at the coordinates.

As described above, for example, in the second area B1 to be inspected, the inspection positions of the first to 49th frames of the first scan S1 are moved at high speed (speed kVs) without acquiring an image, and 50 Only the frames from the th to the 53rd are inspected, and after the inspection is completed, the scanning direction is folded back, the position of the second scanning S2 is moved in the x direction, and the second scanning of the first inspection area A is performed. Inspect the position of S2. The inspection of the embodiment makes it possible to significantly reduce the inspection time without adversely affecting the inspection accuracy of the evaluation of the first inspection area A.

(Second Embodiment)
The inspection device 100 of the second embodiment is a modification of the first embodiment. The description of the contents common to the inspection device and the inspection method of the first embodiment and the inspection device and the inspection method of the second embodiment will be omitted. In the second embodiment, the calculation unit 143 determines whether the first defect is a mask-induced defect or a process-induced defect, and uses the result of determining whether the first defect is a mask-induced defect or a process-induced defect. The first criterion for determining the presence is changed to the second criterion, and the inspection target includes a third inspection region having a pattern common to the first inspection region A, and the inspection unit 150 includes the third inspection region. A third inspection image of the third inspection region is obtained, the reference image is compared with the third inspection image to detect the second defect, and the determination unit 141 uses the second criterion to perform the second inspection. It is determined whether the defect of is a defect caused by the mask used for drawing the third area to be inspected or a defect caused by the process for drawing the third area to be inspected.

In the second embodiment, an inspected object X'that is in a region different from the inspected region A of the first embodiment and includes a third inspected region having a pattern common to the inspected region A is used. .. The target X'inspected in the second embodiment may be the same wafer or mask as that inspected in the first embodiment, or the area to be inspected although not inspected in the first embodiment. Wafers and masks having a pattern common to A may be used. When the object to be inspected X'is the same wafer or mask as the one inspected in the first embodiment, the inspected area A is a region different from the third inspected area. FIG. 7 shows a schematic diagram of the object to be inspected X'in the second embodiment. The test target X'includes a third test area F having a pattern in common with the first test area A. In addition, all the objects to be inspected of the object to be inspected X'may have a pattern common to the first area A to be inspected.

The calculation unit 143 uses the result of determining whether the first defects d1 to d3 are mask-induced defects or process-induced defects in the inspection of the first inspected area A, respectively, and is caused by the mask. The first criterion, which is a criterion for determining whether the defect is a defect or a process-induced defect, is changed to the second criterion. The first criterion is whether the detected defect is a mask-induced defect or a process-induced defect by using the comparison result of comparing the detected defect with the pattern of the inspection area for confirmation. This is the standard for judgment.

In the determination using the first criterion, the defect information of the first inspection area A, the inspection result for confirmation, and the first criterion are required. By using the judgment standard (second standard) in which the judgment using the first standard is added to the judgment standard (first standard), the image of the region for confirming the type of the detected defect is not acquired. Can be judged as.

FIG. 8 shows an enlarged schematic view of the third area to be inspected F of the object to be inspected X'. FIG. 8 shows a scanning number and a broken line indicating a portion where an image is acquired, a solid line in which only scanning (movement) is performed without acquiring an image, and an arrow indicating a scanning (movement) direction, as in FIG. .. In the second embodiment, the inspection unit 150 detects the second defects d8 to d10 with the third area F to be inspected as the inspection target. In the third area F to be inspected, the defect is detected by comparing the inspection image of only the third area F to be inspected with the reference image regardless of whether the defect is detected or not. Then, the determination unit 141 determines whether the second defects d8 to d10 are mask-induced defects or process-induced defects using the second criterion.

Whether the defect is caused by the defect pattern information (positional relationship between the pattern P and the defect, the shape of the defect, the size of the defect, etc.) and the mask of the first defects d1 to d3 in the first inspection area A in the calculation unit 143. Alternatively, arithmetic processing is performed using the result of determining whether the defect is caused by the process. Then, based on the result obtained by this calculation, the first criterion for determining whether the defect is a mask-induced defect or a process-induced defect is changed to the second criterion. That is, in determining the defect type of the second defects d8 to d10 found in the inspection of the third inspected area F, the inspection results of the first inspected area A and the second inspected areas B1 and B2 are used. It can be used. Therefore, in order to determine the types of defects d8 to d10 of the second defects d8 to d10 included in the third area F to be inspected, an image of the area to be inspected adjacent to the scanning direction of the third area F to be inspected is acquired. You don't have to. When determining the result type of the third inspected area F using the second criterion, it is said that the mask used for drawing the first inspected area A and the third inspected area F is a common mask. , It becomes easy to determine that the defect is caused by the mask. In addition, in the calculation unit 143, the calculation for changing the determination criterion inspects the result obtained in another embodiment and the inspection target having a pattern that is not common to the other inspection region or the first inspection region. The result of the calculation can be included in the calculation target.

In the inspection device of the second embodiment, by using the inspection result performed before, it is possible to inspect the target inspection area without performing the confirmation inspection and determine the type of defect. Therefore, the inspection time can be shortened by the amount that the confirmation inspection is not performed, and the inspection can be performed at a higher speed. For example, an inspection device capable of performing inspection at a higher speed by quickly and accurately determining a defect of a common shape at coordinates determined to be a defect caused by a mask in the first embodiment as a defect caused by a mask. It becomes.

Next, the second embodiment will be described in more detail by explaining the inspection method of the inspection device 100 of the second embodiment. FIG. 9 shows a flowchart of the inspection method of the second embodiment. The inspection method in the second embodiment includes a first inspection step (S101), a first defect detection step (S102), a first defect confirmation step (S103), and a first defect type determination step (S104). It includes a first calculation step (S111), a second inspection step (S112), a second defect detection step (S113), and a second defect type determination step (S114). The first inspection step (S101), the first defect detection step (S102), the first defect confirmation step (S103), and the first type defect determination step (S104) are common to the first embodiment.

<First calculation process (S111)>
The first calculation step (S111) is calculated using the result of determining whether the first defects d1 to d3 in the first area A to be inspected are mask-induced defects or process-induced defects. , This is a step of changing the first criterion, which is a criterion for determining the type of defect, to the second criterion based on the result obtained by this calculation. In the first defect type determination step (S104), the first inspection region A and the second inspection regions B1 and B2 for confirmation are inspected to be included in the first inspection region A. The type of defect is determined. By learning this determination result, it is possible to determine the type of defect in the third area F to be inspected without inspecting the inspection area for confirmation. In this step, the inspection result of the first inspection area A is reflected in the defect type determination criteria. The calculation unit 143 performs the first calculation step (S111). For example, in the calculation unit 143, the result for changing the determination standard can be obtained by learning (machine learning) the image of the inspection result. It is possible to learn a plurality of images obtained by repeating the first inspection step (S101) to the first defect type determination step (S104). After the learning in the first calculation step (S111) is completed, the determination criteria are changed, and in the inspections from the second inspection step (S112) to the second defect type determination step (S114), the changed second criterion is used. , The type of defect can be determined. By determining the defect type based on the second criterion, it is possible to make a determination more quickly or with higher accuracy than the determination based on the first criterion. The judgment criteria can be cumulatively updated by further learning using the image of the inspection result and the judgment result of the defect type when judging by the second criterion. The second reference is stored in the storage device 131 or the memory 134 so that the determination unit 141 can make a determination using the obtained second reference. The second criterion can be a criterion that can refer to all the inspection results in the first embodiment including the first criterion, the first inspection image C, and the second inspection image D.

<Second inspection step (S112)>
In the second inspection step (S112), the primary electron beam 200 is irradiated to the third inspected region F of the inspected object X', and the emitted secondary electron beam 201 (including backscattered electrons) is inspected. This is the process of obtaining an image. The second inspection step (S112) is common to the first inspection step (S101) except that the inspection target is different. In the second inspection step (S112), as in the first inspection step (S101), an image is acquired for each inspection frame of the third inspection area F, and an inspection image (third inspection) is obtained for each inspection frame. (Image) is created, and the created third inspection image is stored in the storage device 131 or the memory 134.

<Second defect detection step (S113)>
In the second defect detection step (S113), a defect is found by comparing the third inspection image obtained in the second inspection step (S112) with the third inspection area F as the inspection target and the reference image. This is the process of detection. Similar to the first inspection image C, the third inspection image is also an image created by joining, for example, a part or all of the inspection images of one row of inspection frames continuously arranged in the scanning direction (x direction). Can be used as the third inspection image, or the image of one inspection frame can be used as the third inspection image. Therefore, although it depends on the base image, the reference image can be a part of the image data cut out.

The reference image and the third inspection image are compared by an arbitrary algorithm, and for example, when the determination index is equal to or greater than the threshold value, it is determined that there is a defect (defect is detected). When a defect is detected, the second defect type determination step (S114) is performed. When a defect is not detected in the second defect detection step (S113), the inspection target in the second defect detection step (S113) is the final scan of the area A to be inspected (10th scan S10 in FIG. 8). When is, the inspection is finished. Further, when a defect is not detected in the second defect detection step (S113) and the inspection is not the final scan of the third area F to be inspected (10th scan S10 in FIG. 8), the next scan is performed. For example, the stage is moved so as to be in the position, and the second inspection step (S112) is performed again. Then, the inspection is repeated until the final scan is reached. The detected defects are designated as the second defects d8 to d10, and the obtained information such as the coordinates of the second defects d8 to 10 and the third inspection image is stored in the storage device 131 or the memory 134.

<Second defect type determination step (S114)>
The second defect type determination step (S114) is a step of determining whether the second defects d8 to d10 are mask-induced defects or process-induced defects using the second criterion. By using the second criterion, it is possible to determine the type of defect by using the result of inspecting the confirmation inspection area without inspecting the confirmation inspection area. In the first embodiment, the inspection time is shortened by moving to the coordinates of the inspection target at high speed without performing the inspection, but in the second embodiment, the confirmation inspection is not performed after learning. , From the inspection of the third area F to be inspected to the determination of the defect type can be performed even faster.

The second defects d8 and d10 are defects existing at the first coordinates c1 and c4. If the second defect d8 is a defect having the same pattern as the first defect d1, it can be determined that the second defect d8 and the first defect d1 are of the same defect type. The second criterion has a criterion for making such a determination. Further, if the second defect d8 is a defect having a pattern different from that of the first defect d1, the type of the defect can be determined by the defect information of the second defect d8 and the second criterion. The second defect d10 can also be compared with the first defect d3. The second defect d9 is information not included in the first defect. The inspection image corresponding to the coordinate c7 of the second defect d9 in the first inspection region A can be arbitrarily used as a judgment criterion, and the defect type of the second defect d9 can be determined using the second criterion. ..

In the flowchart of FIG. 9, after performing an inspection (second inspection step (S112)) and defect detection (second defect detection step (S113)) for each scan, a second defect type determination step (S114) Is doing. FIG. 10 is a flowchart of a modified example of the inspection method of the second embodiment. In the inspection method shown in the flowchart of FIG. 10, after the inspections from the first scan to the tenth scan S10 in FIG. 8 are continuously performed, it is determined whether or not the third region F to be inspected contains a defect. Then, the type of defect included in the third area F to be inspected is determined. In the method of the flowchart of FIG. 10, the time from the inspection start (S112) of the third inspection area F to the second defect type determination step (S114) becomes long. In both the method of the flowchart of FIG. 9 and the method of the flowchart of FIG. 10, the calculation unit 143 calculates and learns in the time from the second inspection step (S112) to before the second defect type determination step (S114). Therefore, even if the calculation takes time, the third area F to be inspected can be inspected at high speed. In the method of the flowchart of FIG. 10, a long time can be taken for the learning calculation.

In the inspection method of the embodiment, the inspection can be determined using the past inspection results by calculating and learning with the calculation unit 143, and the inspection time can be shortened. Such an inspection method is an excellent inspection method in that high-speed inspection can be performed because the inspection time can be shortened even when inspecting other inspection areas.

(Third Embodiment)
The inspection device 100 of the third embodiment is a modification of the first embodiment. The description of the contents common to the inspection device and the inspection method of the first embodiment and the inspection device and the inspection method of the third embodiment will be omitted. In the third embodiment, the determination unit 141 is used by the first defect to draw the first inspected area A by using the defect information of the inspection data of the mask that draws the first inspected area A. It is determined whether the defect is caused by the mask or the defect caused by the process for drawing the first area A to be inspected.

In the third embodiment, a mask for drawing the first inspected area A when determining the type of the first defect d1 to d3 in the first inspected area A of the first embodiment. Use the defect information of. From the result of inspecting the mask itself that draws the first area A to be inspected, mask defect information suspected to be a defect or defect of the mask can be obtained. By referring to and comparing the mask defect information in the determination unit 141, it is possible to perform a faster and more accurate inspection.

FIG. 11 shows a schematic view of the inspection image of the mask G. It is known from the inspection of the mask G that two mask defects d11 to d12 are present or suspected to be present in the mask G. The fact that two mask defects d11 to d12 exist or are suspected to exist in the mask G is the defect information of the mask G that draws the first area to be inspected. Since the pattern of defects caused by the mask generated in the first inspected area A can be predicted from the defect information of the mask that draws the first inspected area A, the mask defect information is the first defect type. Useful for judgment. Comparing the first object A to be inspected and the mask G in FIG. 3, the positions of the defects d1 and d3 of the first coordinates c1 and c4 of the first object A to be inspected correspond to each other. That is, the defect information of the mask G means that there is a high possibility that a defect caused by the mask will occur in the first coordinates c1 and c4 of the first object A to be inspected. When determining the type of defect in the determination unit 141, the defect type in the first inspected area A is taken into consideration by considering the condition that there is a possibility (high) that a defect caused by a mask occurs in the coordinates c1 and c4. It is possible to increase the reliability of the judgment of. Therefore, if a defect is detected in the first coordinates c1 and c4 of the first area A to be inspected, the possibility of determining the defect due to the mask increases. Determining the defect type with reference to the defect information of the mask G is preferably performed when determining the defect type in other embodiments and other areas to be inspected. For example, in the inspection of the third area F to be inspected, the defect information of the mask G may be referred to.

Next, the third embodiment will be described in more detail by explaining the inspection method of the inspection device 100 of the third embodiment. FIG. 12 shows a flowchart of the inspection method of the third embodiment. The inspection method in the second embodiment includes a first inspection step (S101), a first defect detection step (S102), a first defect confirmation step (S103), a mask information reference step (S121), and a third. The defect type determination step (S122) is included. The first inspection step (S101), the first defect detection step (S102), and the first defect confirmation step (S103) are common to the first embodiment.

<Mask information reference step (S121)>
The mask information reference step (S121) is a step of referring to the defect information of the mask G on which the first inspected area A is drawn in order to use it for determining the defect type of the first inspected area A. By storing the position and shape of the defect of the mask G in the memory 134 or the storage device 131 as comparable information, it can be used in the third defect type determination step (S122) later. The mask information reference step (S121) may be performed at any time before the third defect type determination step (S122).

<Third defect type determination step (S122)>
In the third defect type determination step (S122), it is first inspected whether the first defect detected in the first defect detection step (S102) is a mask-induced defect or a process-induced defect. This is a step of determining using the mask defect information for drawing the area A, that is, using the mask defect information and the first criterion. Since the mask defect information is a highly reliable judgment standard for determining that the first defect is a defect caused by the mask, the time required for determining the defect type can be shortened and the reliability of the defect type determination can be shortened. You can increase the degree. The third defect type determination step (S122) and the first defect type determination step (S104) are common steps except whether or not mask defect information is used. By using the mask defect information in the second defect determination step (S114), the time required for determining the defect type can be similarly shortened, and the reliability of the defect type determination can be increased.

(Fourth Embodiment)
The inspection device 100 of the fourth embodiment is a modification of the first embodiment or the second embodiment. The description of the contents common to the inspection device and the inspection method of the first embodiment and the inspection device and the inspection method of the fourth embodiment will be omitted. In the fourth embodiment, the calculation unit 143 uses the result of determining whether the first defect is a mask-induced defect or a process-induced defect, and uses the result of determining whether the first defect is a mask-induced defect or a process-induced defect. The first criterion for determining whether or not the defect is the third criterion is changed to the third criterion, and the inspection unit 150 detects a third defect other than the first defect included in the first inspection area A, and detects the third defect. The determination unit 141 determines whether the third defect is a mask-induced defect or a process-induced defect using the third criterion.

In the fourth embodiment, the first inspected region A having the third defect d13 is set as the inspected region. FIG. 13 shows an enlarged schematic view of a part of the object X to be inspected. The schematic diagram of FIG. 13 is similar to the schematic diagram of FIG. The difference between the schematic diagram of FIG. 13 and the schematic diagram of FIG. 3 is that in the schematic diagram of FIG. 13, the third defect d13 exists at the coordinate c8 of the first inspected region A.

The calculation unit 143 uses the result of determining whether the first defects d1 to d3 are mask-induced defects or process-induced defects in the inspection of the first inspected area A, respectively, and is caused by the mask. The first criterion, which is a criterion for determining whether the defect is a defect or a process-induced defect, is changed to a third criterion. The first criterion is a criterion for comparing the detected defect with the pattern of the inspection area for confirmation and determining whether the detected defect is a defect caused by a mask or a defect caused by a process. In the determination using the first criterion, the defect information of the first inspection area A, the inspection result for confirmation, and the first criterion are required. By using the judgment standard (third standard) in consideration of the judgment result using the first standard as the judgment standard (first standard), for example, the first defect d1 in the first standard. If it is determined that the pattern is derived from the process, the criterion is changed to the third criterion in consideration of this result. If a defect of the same pattern is found, it is determined at high speed that the determination is derived from the process. Can be done. In the first inspected area A, the defect type of the third defect d13 detected other than the first defects d1 to d3 can be determined without inspecting the area for confirmation.

When the calculation in the calculation unit 143 is completed, that is, the determination of the type of defect required for changing the determination criterion is performed the required number of times, and the calculation for learning this result is completed, the determination criterion is changed from the first criterion. It will be changed to the third standard. Then, during the inspection of the inspection area inspected to change the determination standard, the type of defect is determined by the third standard which is a new inspection standard. Defects for which the type of defect is determined by the third criterion can be inspected at a higher speed than the defect determined by the first criterion because the type of defect is determined without performing the inspection for confirmation. .. In the second embodiment, after the inspection of the first area A to be inspected, the inspection is performed with a new criterion to shorten the inspection time, but in the fourth embodiment, the sequential criterion is set in real time during the inspection. By changing it, the inspection time will be further shortened. When some defects whose type of defect is difficult to determine by the third criterion are detected, the coordinates corresponding to the coordinates of the third defect d13 in the second inspected areas B1 and B2 adjacent to the scanning direction. High-precision inspection can be performed by acquiring only the image.

Next, the fourth embodiment will be described in more detail by explaining the inspection method of the inspection device 100 of the fourth embodiment. FIG. 14 shows a flowchart of the inspection method of the fourth embodiment. The inspection method in the fourth embodiment includes a first inspection step (S101), a first defect detection step (S102), a first defect confirmation step (S103), and a first defect type determination step (S104). It includes a second calculation step (S131), a third inspection step (S132), a third defect detection step (S133), and a fourth defect type determination step (S134). In the second embodiment, after the inspection of one inspection area is completed, another inspection area is inspected by the second criterion obtained by the first calculation step (S111), but the fourth embodiment is performed. In the form, when inspecting one inspection area, the inspection criteria can be sequentially changed (improved) in real time during the inspection. The timing of the sequential change may be for each defect, for each of a plurality of locations, for each scan, or the like.

In the fourth embodiment, for example, when the ninth scan S9 is completed, the calculation unit 143 completes the calculation of the inspection results from the first scan S1 to the eighth scan S8, which is a new criterion improved. Third criterion is obtained. Then, when a third defect is found in the tenth scan S10, the third defect type determination step (S122) determines the third defect based on this new third criterion. In this way, even if a third defect is found in the tenth scanning S10 (during inspection), the third defect type determination step (S122) without inspecting the second inspection areas B1 and B2 for confirmation. ) Determines the type of the third defect. Hereinafter, this method will be described for each step.

The first inspection step (S101), the first defect detection step (S102), the first defect confirmation step (S103), and the first defect type determination step (S104) themselves are the same as those in the second embodiment. However, when the first defect is not detected in the first defect detection step (S102), instead of determining whether or not it is the final scan, the second calculation step (S131) in the calculation unit 143 The difference is that it determines whether or not is finished.

When the first defect is not detected in the first defect detection step (S102), it is confirmed whether the second calculation step (S131) is completed. Since the calculation is not completed during the execution of the process of the second calculation step (131), the determination of whether or not the calculation is completed is "NO". Whether or not the second calculation step (S131) is completed in the next process of the first defect type determination step (S104) was detected earlier than the defect whose defect type was determined immediately before. It is determined whether or not the operation for learning the defect is completed. Further, when there is no defect in the first defect detection step (S102), determining whether or not the second calculation step (S131) is completed in the next process is to learn the defects detected in the past. Judge whether or not the calculation to be performed is completed.

When the second calculation step (S131) is completed, the inspection by the third inspection step (S132) is performed in order to determine by the third criterion. If the second calculation step (S131) is not completed, the process returns to the next scanning position and the inspection by the first inspection step (S101) is performed in the same manner. Since the accuracy of determining the defect type is improved in the third criterion as compared with the first criterion, the defect type can be determined faster by determining with the third criterion than in the case of determining with the first criterion. Can be determined. It is preferable to learn many defect patterns in order to enhance the judgment system of the third criterion.

After the first defect is detected in the first defect detection step (S102) and the types of the first defects d1 to d3 are determined in the first defect type determination step (S104), the second calculation step It is confirmed whether (S131) is completed. When the second calculation step (S131) is completed, the inspection by the third inspection step (S132) is performed in order to determine by the third criterion. If the second calculation step (S131) is not completed, the process returns to the next scanning position and the inspection by the first inspection step (S101) is performed in the same manner.

<Second calculation process (S131)>
The second calculation step (S131) is calculated using the result of determining whether the first defects d1 to d3 in the first area A to be inspected are mask-induced defects or process-induced defects. This is a step of changing the first criterion, which is a criterion for determining the type of defect, to the third criterion. The difference between the first calculation process (S111) and the second calculation process (S131) is that in the second calculation process (S131), it is referred to whether or not the calculation is completed, and if the calculation is completed, the calculation is completed. If so, the judgment criteria will be reflected in real time during the inspection and will become a new third criterion. When the inspection is performed as shown in the schematic diagram of FIG. 13, the second calculation step is completed in the ninth scan S9. The second calculation step (S131) makes it possible to determine the type of defect from the characteristics of the defect.

<Third inspection step (S132)>
The third inspection step (S132) is a step of obtaining a first inspection image C by using an electron beam to inspect the first inspection region A as in the first inspection step (S101).

<Third detection step>
The third defect detection step (S133) is a step of comparing the reference image with the first inspection image C to detect the third defect in the same manner as in the first defect detection step (S101). This step is performed until the final scan regardless of whether or not a defect is detected. In FIG. 13, in the tenth scan S10, the third defect d13 is detected at the coordinate c8.

<Third defect type determination step (S134)>
The third defect type determination step (S134) is a step of determining the type of the third defect d13 detected in the third defect detection step (S133) based on the third criterion. The type of the defect of the third defect d13 is determined by the third criterion without performing the defect confirmation step after the third defect detection step (S133). Since the second calculation step (S131) is performed in real time during the inspection of the first inspected area A, the type of defect is determined using the calculation result during the inspection of the first inspected area A. Can be done. If a defect is detected after the calculation is completed, the inspection time can be shortened.

In the flowchart of FIG. 14, after the completion of the second calculation step (S131), the inspection and the defect detection are performed in each scan, and then the third defect type determination step (S134) is performed. FIG. 15 shows a flowchart of a modified example of the inspection method of the fourth embodiment. In the inspection method shown in the flowchart of FIG. 15, after the completion of the second calculation step (S131), all the remaining scans are performed, and then all the types of the third defects d13 detected in the remaining scans are determined. It is collectively performed in the fourth defect type determination step (S134). Further, as a further modification example, it is also preferable to sequentially perform the second calculation step (S131) and change the inspection standard a plurality of times during the inspection of one inspection area.

(Fifth Embodiment)
The inspection device 100 of the fifth embodiment is a modification of the first embodiment. The description of the contents common to the inspection device and the inspection method of the first embodiment and the inspection device and the inspection method of the fifth embodiment will be omitted. In the fourth embodiment, the image correction unit 142 reflects the mask-induced defect information obtained by the determination unit 141 in the reference image, and adds the mask-induced defect information to the reference image.

In the third embodiment, the mask defect information is used for determining the type of defect, but in the fifth embodiment, the mask defect information is reflected in the reference image created by the reference image creating unit 140. The mask defect information is added to the reference image when the reference image is created from the database of the die to database inspection.

For example, the defect information (position, defect pattern) of the mask is obtained by referring to the inspection result of the mask itself for drawing the first inspection area A stored in the storage device 131. Then, the image correction unit 142 adds a defect existing or may be present in the mask to the reference image. FIG. 16 shows a schematic view of a reference image to which defect information is added. In the schematic view of the reference image of FIG. 16, the reference image H having the pattern p generated from the original image pattern is shown. For example, it is assumed that the mask defect information has or may have a mask defect at coordinates c9 and c10, respectively. Therefore, two mask defects d14 and d15 are drawn in the reference image H of FIG. The coordinates c9 and c10 of the mask defects d14 and d15 correspond to the first coordinates c1 and c4 of the first area A to be inspected.

By comparing the reference image including the mask defect information with the first inspection image, defect detection can be performed at higher speed. Further, when determining the type of defect, since the defect information of the mask is input in the processing system that performs the determination, the inspection time can be shortened and the reliability of the inspection can be improved.

Next, the fifth embodiment will be described in more detail by explaining the inspection method of the inspection device 100 of the fifth embodiment. FIG. 17 shows a flowchart of the inspection method of the fifth embodiment. The inspection method in the fifth embodiment includes a first inspection step (S101), a fourth defect detection step (S142), a first defect confirmation step (S103), a first defect type determination step (S104), and The reference image correction step (S141) is included. The first inspection step (S101), the first defect confirmation step (S103), and the first defect type determination step (S104) are common to the first embodiment.

<Image correction step (S141)>
The image correction step (S141) is a step of adding defect information of the mask for drawing the first inspection target A to the reference image. In the fourth defect detection step (S142), defect information is added to the reference image so that it can be distinguished whether it is a pattern (original such a pattern, not a defect) or a defect. The reference image H to which the defect information of the mask is added is stored in the memory 134 or the storage device 131.

<Fourth defect detection step (S142)>
The fourth defect detection step (S142) is a step of comparing the reference image H with the first inspection image C to detect a defect in the first inspection region A. In the reference image H, since the defect information and the pattern information include information in a distinguishable form, the pattern and the defect are distinguished and compared by comparing the first inspection image C and the reference image H. By doing so, it is possible to shorten the time required for defect detection and improve the accuracy of defect detection.

The evaluation of the object to be inspected can be performed in a short time by preferably inspecting two or three areas to be inspected for one object to be inspected by the inspection of the form including the first embodiment. You can.

In the above description, each "~ part" includes hardware, software and hardware, and a combination of software.

The embodiments of the present invention have been described above with reference to specific examples. The above embodiments are merely given as examples, and do not limit the present invention. In addition, the components of each embodiment may be combined as appropriate.

In the embodiment, the description of the configuration of the inspection device, the manufacturing method thereof, the inspection method, and other parts not directly required for the description of the present invention is omitted, but the required configuration of the inspection device and the inspection method is appropriately configured. It can be selected and used. In addition, all inspection devices and inspection methods that include the elements of the present invention and can be appropriately redesigned by those skilled in the art are included in the scope of the present invention. The scope of the present invention is defined by the scope of claims and the scope of their equivalents.

100 Inspection device 110 Control unit 121 Control computer 122 Bus 131 Storage device 132 Monitor 133 Printer 134 Memory 135 Position calculation unit 136 Stage control unit 137 Deflection control unit 138 Blanking control unit 139 Lens control unit 140 Reference image creation unit 141 Judgment unit 142 Image correction unit 143 Calculation unit 150 Inspection unit 151 Drive unit 152 Laser length measurement system 153 Detection unit 154 Pattern memory 155 Inspection image creation unit 161 Electron beam column 162 Examination room 163 Electron gun 164 Electromagnetic lens 165 Blanking deflector 166 Limited aperture array Substrate 167 Electromagnetic lens 168 Main deflector 169 Sub-deflector 170 Detector 171 Stage 172 Mirror 200 Primary electron beam 201 Secondary electron beam

Claims (5)

The first subject to be inspected, including a first region to be inspected and a second region to be inspected that is adjacent to the first inspected region in the scanning direction and has a pattern common to the first inspected region. When the first inspection image of the inspection area is obtained, the reference image is compared with the first inspection image, and the first defect is detected at the first coordinates of the first inspection area, the first defect is detected. During the inspection of the first area to be inspected, it moves to the second coordinate of the second area to be inspected corresponding to the first coordinate of the first area to be inspected, and only the pattern of the second coordinate is moved. The inspection department that obtains the second inspection image of
Whether the first defect is a defect caused by a mask used for drawing the first area to be inspected or a defect caused by a process for drawing the first area to be inspected is determined. A determination unit that determines from the first inspection image at one coordinate and the second inspection image at the second coordinate,
An inspection device characterized by having. First, it is determined whether the first defect is a mask-induced defect or a process-derived defect by using the result of determining whether the mask-induced defect or the process-induced defect. Further equipped with a calculation unit that changes the standard of
The subject to be inspected includes a third region to be inspected having a pattern common to the first region to be inspected.
The inspection unit obtains a third inspection image of the third area to be inspected, compares the reference image with the third inspection image, and detects a second defect.
Using the second criterion, the determination unit draws the third inspected area to see if the second defect is a defect caused by the mask used to draw the third inspected area. The inspection apparatus according to claim 1, wherein it is determined whether or not the defect is caused by a process for the purpose. The determination unit is a defect caused by the mask used by the first defect to draw the first inspected area by using the defect information of the inspection data of the mask that draws the first inspected area. The inspection apparatus according to claim 1 or 2, wherein it determines whether the defect is due to a process for drawing the first area to be inspected. First, it is determined whether the first defect is a mask-induced defect or a process-derived defect by using the result of determining whether the mask-induced defect or the process-induced defect. Further equipped with a calculation unit that changes the standard of
The inspection unit detects a third defect other than the first defect included in the first inspection region, and detects the third defect.
Of claims 1 to 3, the determination unit determines whether the third defect is a defect caused by the mask or a defect caused by the process by using the third criterion. The inspection device according to any one of the claims. According to claim 1, the reference image correction unit is further provided by reflecting the mask-induced defect information obtained by the determination unit on the reference image and adding the mask-induced defect information to the reference image. The inspection device according to any one of 4.