JPH11183154A - Electron beam system inspection or measurement apparatus and its method and optical height detecting apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable inspection and length measurement with high precision and high reliability on the basis of an electron beam image (SEM), by reducing image distortion caused by deflection and aberration of an electron optical system, and deterioration of resolution due to defocus, and improving the quality of the electron beam image (SEM). SOLUTION: This measuring apparatus is provided with an electron beam image detection optical system 104 detecting a secondary electron beam image generated from an object to be inspected by an electron beam cast from an electron optical system, optical height detecting apparatus 200a, 200b detecting optically the surface height, a focus control means 109 converging the electron beam on the object in the focusing state by controlling an objective 103 of the electron optical system, a deflection control means 108 correcting image distortion containing magnification error of an electron beam image which is generated on the basis of focus control, and an image processing means 124 performing inspection or measurement of a pattern which is formed on the object on the basis of the secondary electron beam image.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an inspection or measuring apparatus and method using a charged particle beam such as an electron beam or an ion beam for a semiconductor substrate such as a semiconductor wafer, and a charged particle beam image observation and processing apparatus. And an optical height detecting device.

[0002]

2. Description of the Related Art As an apparatus using a charged particle beam, an automatic inspection system for inspecting and measuring a fine circuit pattern formed on a semiconductor wafer or the like using an electron microscope image will be described as an example. Defect inspection of fine circuit patterns formed on semiconductor wafers and the like is performed by comparing the pattern to be inspected with a non-defective pattern or the same type of pattern on the wafer to be inspected, and many apparatuses using optical microscope images have been put into practical use. ing. Similarly, when an electron microscope image is used, a defect inspection is performed by comparing pattern images. Also, in a length measuring device using a scanning electron microscope that measures the line width and the hole diameter of a fine circuit pattern used for setting the manufacturing process conditions of a semiconductor device or for monitoring, etc., the length measurement is automated by image processing. Have been done. As described above, in the case of a comparative inspection for detecting a defect by comparing electron beam images of similar patterns, or when processing an electron beam image to measure the line width of a pattern, the obtained electron beam image Quality greatly affects the reliability of the test results. The quality of an electron beam image is degraded due to image distortion caused by deflection or aberration of the electron optical system, or a decrease in resolution due to defocus. These deteriorations in image quality degrade the performance of comparative inspection and length measurement. When the inspection is performed with an electron beam image input under the same conditions over the entire range when the height of the sample surface is not constant, the electron beam image changes depending on the inspection location as shown in FIG. If the inspection is performed by comparing the focused image (a) with the defocused images (b) and (c), correct results cannot be obtained. Further, in these images, the width of the pattern changes and the result of edge detection of the image cannot be obtained stably, so that the line width and hole diameter of the pattern cannot be measured stably. Conventionally, focusing with an electron microscope is performed by adjusting the control current of the objective lens while observing the electron beam image, but in addition to taking a lot of time, the electron beam scans the sample surface many times. As a result, damage to the sample may be a problem.

In the case of a processing apparatus using a convergent charged particle beam, focusing of the charged particle beam affects the processing accuracy, and is a very important problem as in the observation apparatus. In these devices, focusing cannot be performed by observing an image like an electron microscope. Therefore, it is necessary to detect the height of the wafer by another method and feed it back to the control current of the objective lens or the height of the sample stage. Examples of the processing apparatus include an electron beam type semiconductor pattern exposure apparatus and a circuit correction apparatus using a focused ion beam. Further, as the height of the sample changes, not only the degree of focusing but also generally the magnification of the charged particle optical system changes. For this reason, it is necessary to detect the height of the sample and add feedback to the scanning width of the charged particle optical system.

[0004] Conventionally, for example, Japanese Patent Application Laid-Open No. 58-168906.
As shown in (2), it is widely used to determine the height of a sample by irradiating a spot light onto a sample surface, forming an image of the specularly reflected light, and detecting the position of the image with a detector. Was. As the detector, any of a detector called a PSD which outputs a signal proportional to a position where light is applied, a two-division photodiode, and a one-dimensional linear image sensor has been conventionally used. However, this method has a problem that a large height detection error occurs when the sample surface has a distribution of the reflectance and the spot light hits a boundary having a different reflectance. This is because the true center position of the spot does not match the center of gravity of the spot because the spot light detection waveform is distorted. This is because they cannot. If a one-dimensional linear image sensor is used, the detection value can be brought closer to the true center position of the spot by devising a signal processing method.
It is difficult to accurately and stably detect the contour of a spot image when there is a large uneven reflectance on the sample surface. The cause of the uneven reflectance of the sample surface is not only the difference in the material of the surface, but also the effect of scattering due to the fine pattern on the surface, especially when the sample has a fine circuit pattern drawn on the surface of a semiconductor device. This effect is severe. If the displacement of the sample surface is large, the imaging relationship between the projected light, the sample surface, and the detector changes. However, there is a problem that is larger. Since the semiconductor wafer of the inspection object in the middle of the manufacturing process is deformed due to heat treatment and the like, the amount of deformation is the worst and several hundred μm,
It is important to stably maintain accuracy even in a wide range of detection.

As a measure against an error caused by a change in the state of the sample surface, as disclosed in Japanese Patent Application Laid-Open No. Sho 59-195728, a plurality of slits are used as a wafer height sensor for focusing an optical projection exposure apparatus. A method has been proposed in which an image is projected onto a wafer from an oblique direction and the position of specular reflection light is detected by a line image sensor to detect the height and inclination of the wafer surface with high accuracy. However, even this method does not provide a sufficient solution to the instability at the time of wide-area detection as described above. This method has a Z stage, and is a method used for an optical projection exposure apparatus that performs exposure after controlling the wafer surface to a constant height by the stage, and stable height detection over a wide range is not necessarily required. . However, charged particle beam devices such as electron microscopes have problems with the reliability and cost of the movable mechanism in the vacuum sample chamber, and generally do not have a Z-stage. Is important. As other known examples, US Pat. No. 4,477,185 and US Pat.
As shown in FIG. 88, a method is disclosed in which a spot light is projected from the left and right onto the same point on the sample in a superimposed manner, and the positions of the respective specularly reflected lights are combined to detect the same. This method makes use of the fact that the detection error due to the distribution of the reflectance of the sample surface becomes opposite when the detection error is detected from the opposite direction, and is canceled out by the addition. In this method, the detection error becomes very small when the sample is at the reference height and the position of the spotlight is accurately superimposed, but when the height of the sample changes, the superimposition of the spotlight becomes impossible. Then, a detection error occurs due to the distribution of the reflectance of the sample surface. Therefore, as in the case of the above-mentioned method, this method is effective when the height of the sample is always kept constant at the Z stage. However, a configuration in which the focal position is changed by controlling the objective lens without using the Z stage. There was a problem as a height sensor for this device.

A height detecting method having a function of moving an optical system in response to a displacement of a sample surface is also disclosed in Japanese Patent Laid-Open No. 6-67727.
However, there is a problem that a high-precision movable mechanism is required and the device becomes large-scale. Even if stable height detection can be achieved by these optical methods, the field of view of the charged particle optical system that performs focus adjustment using them must match the measurement position of the height detector. If so, the focus adjustment function will not work. In particular, when controlling the focus while continuously moving the stage on which the sample is placed,
Since a position shift occurs due to a detection time delay of the height detector, a countermeasure is required.

[0007]

SUMMARY OF THE INVENTION It is an object of the present invention to improve the quality of an electron beam image by reducing image distortion due to deflection or aberration of an electron optical system or reduction in resolution due to defocus. Another object of the present invention is to provide an electron beam type inspection or length measuring apparatus and a method for performing inspection and length measurement using an electron beam image with high accuracy and high reliability. Another object of the present invention is to realize real-time detection of the height of the surface of the object to be inspected and control of the electron optical system, thereby realizing a high-resolution electron beam without image distortion due to continuous stage movement. It is an object of the present invention to provide an electron beam type inspection or measuring apparatus and method for obtaining an image, improving inspection performance and stability thereof, and shortening the inspection time. Further, another object of the present invention is to improve the quality of a charged particle beam image by reducing image distortion caused by deflection or aberration of a charged particle optical system or a decrease in resolution due to defocus, thereby improving the quality of a charged particle beam image. It is an object of the present invention to provide a charged particle beam observation device which enables observation using an image.

Another object of the present invention is to provide an electron beam exposure apparatus and a focused ion beam processing apparatus capable of exposing and processing an extremely fine pattern at high resolution without image distortion. Another object of the present invention is to provide an optical height detecting device capable of detecting the height of the surface of an object with high accuracy by a simple configuration.

[0009]

To achieve the above object, the present invention provides an electron beam source, a deflecting element for deflecting an electron beam emitted from the electron beam source, and an electron beam deflected by the deflecting element. And an electron lens having an objective lens for focusing and irradiating the object on the object to be inspected, and secondary electrons generated from the object to be inspected by the electron beam which is deflected by the electron optical system, focused and irradiated. An electron beam image detecting optical system for detecting a line image, a projection optical system for projecting a lattice-like (for example, slit-like) light beam onto the object to be inspected from obliquely above, and a grid-like image projected by the projection optical system. A detection optical system which forms an image of a lattice-like light beam reflected from the surface of the object to be inspected by the light beam, receives a lattice-like optical image corresponding to this imaging state, converts the optical image into a lattice-like signal, and detects it. Example obtained from a lattice signal detected by the detection optical system If the detected error is small imaging condition based on the outline level of the surface of the inspected object (e.g., (number 34)
A focusing lattice sf) is searched based on the equation, and a weighting process (for example, a weighted average process using a weight or a product-sum process using a weight) adapted to an imaging state in which the detected detection error is small. Is applied to the grid-like signal detected by the detection optical system, so that the amount of movement or the amount of phase change with respect to the reference signal corresponding to the reference height as the grid-like signal can be calculated by, for example, Equation (35) or (Equation 35). 37) an optical height detecting device having calculating means for obtaining information according to the surface height of the inspection object obtained based on the equation (37), and an inspection device obtained by the optical height detecting device. Focus control means for controlling a current or a voltage to be applied to the objective lens of the electron optical system based on information corresponding to the height of the surface of the object to focus the electron beam on the object to be inspected in a focused state. And the electron beam image detecting optical system Based on the secondary electron beam image detected is an electron beam inspection or measurement device characterized by comprising an image processing means for performing inspection or measurement of the pattern formed on the object to be inspected.

Further, according to the present invention, an electron beam source, a deflecting element for deflecting the electron beam emitted from the electron beam source, and an electron beam deflected by the deflecting element are focused and irradiated on an object to be inspected. An electron optical system having an objective lens, and an electron beam image detection optical system that detects a secondary electron beam image generated from the inspection object by an electron beam that is deflected by the electron optical system, focused, and irradiated. A projection optical system for projecting a lattice-like (for example, slit-like) light beam onto the inspection object from obliquely above, and a grating reflected from the surface of the inspection object by the lattice-like light beam projected by the projection optical system. Detection optical system that forms a lattice-like light beam, receives a lattice-like optical image corresponding to this imaging state, converts it into a lattice-like signal, and detects the lattice-like signal, and a lattice-like signal detected by the detection optical system Selective processing to reduce the detection error An optical height detecting device having calculating means for obtaining a movement amount or a phase change amount with respect to a reference signal as a lattice signal to obtain information corresponding to the height of the surface of the object to be inspected; The electron beam is controlled by controlling a current or a voltage applied to the objective lens of the electron optical system based on information corresponding to the surface height of the object to be inspected obtained by the optical height detecting device. Focus control means for focusing on an object in a focused state, and an image for inspecting or measuring a pattern formed on the object to be inspected based on a secondary electron beam image detected by the electron beam image detecting optical system An electron beam type inspection or measurement device comprising a processing unit.

According to the present invention, in the electron beam type inspection or measuring apparatus, the selective processing in the calculating means of the optical height detecting apparatus is performed by a weighting function (w (x) or w (x)).
(S)) (for example, a weighted average process using weights or a product-sum process using weights). Further, the present invention provides an electron beam source, a deflection element for deflecting the electron beam emitted from the electron beam source, and an objective lens for focusing and irradiating the electron beam deflected by the deflection element onto the inspection object. And an electron beam image detection optical system that detects a secondary electron beam image generated from the inspection object by an electron beam that is deflected by the electron optical system, focused, and irradiated. As shown in FIG. 30, FIG. 32, and FIG. 33, each of a plurality of slit-like or spot-like luminous fluxes is obliquely placed at a plurality of different places close to each other in a region on the inspection object irradiated with the electron beam by the electron optical system. A projection optical system that forms an image from above and a plurality of light beams that are reflected by a plurality of light beams that are formed and projected on a plurality of locations on the inspection object by the projection optical system to form an image. Received multiple luminous flux images A detection optical system that converts and detects the number of signals, and obtains a movement amount or a phase change amount with respect to a reference signal corresponding to a reference height in each of the plurality of detection signals detected by the detection optical system, and Electron beam shifted by the amount of displacement by obtaining information corresponding to the surface height and interpolating (including interpolation and extrapolation) the obtained information corresponding to the surface height at a plurality of locations. Calculating means for calculating information corresponding to the height of the surface at the location where the electron beam is irradiated, and the surface at the location where the electron beam calculated by the optical height detection apparatus is irradiated Focus control means for controlling a current flowing through an objective lens of the electron optical system or a voltage to be applied based on information corresponding to the height of the electron beam to focus an electron beam on an object to be inspected in a focused state; Detected by line image detection optics That is an electron beam inspection or measurement device characterized by comprising an image processing means for performing inspection or measurement of a pattern formed on the inspection object based on the secondary electron beam image.

Further, according to the present invention, an electron beam source, a deflecting element for deflecting the electron beam emitted from the electron beam source, and an electron beam deflected by the deflecting element are focused and irradiated on an object to be inspected. An electron optical system having an objective lens, and an electron beam image detection optical system that detects a secondary electron beam image generated from the inspection object by an electron beam that is deflected by the electron optical system, focused, and irradiated. A projection optical system for imaging and projecting each of a plurality of slit-like or spot-like luminous fluxes from obliquely above at a plurality of different places close to each other in a region on the inspection object irradiated with an electron beam by the electron optical system; The projection optical system forms an image of each of the plurality of light beams reflected by the plurality of light beams imaged and projected on a plurality of locations on the inspection object, receives the formed plurality of light beam images, and Convert to multiple signals and detect Information corresponding to the height of the surface at the plurality of locations by calculating the amount of movement or the amount of phase change with respect to the reference signal corresponding to the reference height in each of the plurality of detection signals detected by the output optical system and the detection optical system. Interpolation (including interpolation and extrapolation) of the information corresponding to the obtained surface heights at the plurality of locations is performed, and the height of the surface at the location irradiated with the electron beam shifted by the positional shift is obtained. An optical height detecting device having a calculating means for calculating information corresponding to the information, and from information corresponding to the surface height at a location where the electron beam calculated by the optical height detecting device is irradiated, A focus control current or a focus control voltage is calculated based on a calibration parameter between the information corresponding to the height of the surface and the focus control current or the focus control voltage, and the calculated focus control current or the focus control voltage is calculated by the electronic control unit. Optical system pair Focus control means for controlling the beam to be applied to the lens to focus the electron beam on the object to be inspected in a focused state; and an object to be inspected based on the secondary electron beam image detected by the electron beam image detecting optical system. An electron beam type inspection or measurement device comprising: an image processing means for inspecting or measuring a pattern formed on an object.

[0013] The present invention also provides the electron beam inspection or measurement apparatus, further comprising: information based on a surface height at a location irradiated with the electron beam calculated by the optical height detection apparatus. And a deflection control unit for correcting an amount of deflection of the electron optical system to the deflection element and correcting an image distortion including a magnification error of the electron beam image generated based on the focus control. In addition, the present invention, in advance, is mounted on a stage, at intervals determined based on the array information of the pattern formed on the inspection target in the region on the inspection target irradiated with the electron beam A height detecting device for detecting a surface height, and an estimating means for estimating a surface height at an arbitrary position between the intervals by interpolating the surface height at the intervals detected by the height detecting device. An electron beam source, a deflecting element for deflecting the electron beam emitted from the electron beam source, and an electron beam deflected by the deflecting element focused and irradiated on an inspection object mounted on the stage. An electron optical system having an objective lens and a deflection by the electron optical system,
An electron beam image detecting optical system for detecting a secondary electron beam image generated from the object to be inspected by the focused and irradiated electron beam, and an electron beam is irradiated by the electron optical system estimated by the estimating means. Control the current or voltage applied to the objective lens of the electron optical system based on information on the surface height of the inspection object at an arbitrary position to focus the electron beam on the inspection object in a focused state. A focus control unit for performing the inspection, and an image processing unit for inspecting or measuring a pattern formed on the inspection object based on the secondary electron beam image detected by the electron beam image detection optical system. It is an electron beam type inspection or measurement device.

Further, according to the present invention, in the electron beam type inspection or measurement apparatus, the focus control means may include a relative position in a height direction between a focus position of the electron optical system and a table on which the inspection object is placed. The position may be controlled. Further, according to the present invention, a projection optical system for moving a test object in at least a predetermined direction and projecting a lattice-like (for example, slit-like) light beam onto the test object from obliquely above, The lattice-like light beam reflected from the surface of the inspection object is imaged by the lattice-like light beam, and a lattice-like optical image corresponding to the imaging state is received, converted into a lattice-like signal, and detected. An imaging state in which a detection error is small based on a grid-like signal detected by the detection optical system by an optical height detection device having a detection optical system (for example, a focused grating sf based on Expression 34). , And performing a weighting process suitable for an imaged state in which the detected error is small to the lattice-like signal detected by the detection optical system to obtain a reference height as a lattice-like signal. For the reference signal according to The amount of movement or the amount of phase change is determined based on, for example, Expression (35) or Expression (37), and the height of the surface in the region on the inspection object irradiated with the electron beam is detected. The electron beam emitted from the electron beam source is controlled by controlling the current flowing through the objective lens of the electron optical system or the voltage to be applied based on the height of the electron beam. Focused in a focus state, a secondary electron beam image generated from the inspected object by the deflected, focused and irradiated electron beam in the focused state is detected by an electron beam image detecting optical system, and the detected An electron beam inspection or measurement method characterized by inspecting or measuring a pattern formed on an object to be inspected based on a secondary electron beam image.

Further, according to the present invention, the object to be inspected is moved at least in a predetermined direction.
As shown in FIG. 2, a projection optical system for forming and projecting a plurality of slit-shaped or spot-shaped light beams from a diagonally upper position on a plurality of adjacent different places on the inspection object, and the projection optical system on the inspection object Detecting each of a plurality of light beams reflected by a plurality of light beams imaged and projected at a plurality of locations, receiving the formed plurality of light beam images, converting them into a plurality of signals, and detecting them An optical height detecting device having an optical system and a moving amount or a phase change amount with respect to a reference signal corresponding to a reference height in each of the plurality of detection signals detected by the detecting optical system. Information obtained according to the height of the surface is obtained, and the obtained heights of the surface at a plurality of locations are interpolated (including interpolation and extrapolation) as shown in FIGS. 31 and 34, for example. The electron beam shifted by the shift The electron beam emitted from the electron beam source by calculating the height of the surface at the place where the light is projected, and controlling the current or the voltage applied to the objective lens of the electron optical system based on the calculated height of the surface Is deflected by a deflecting element of an electron optical system to be focused on an object to be inspected in a focused state,
An electron beam image detecting optical system detects a secondary electron beam image generated from the inspected object by the electron beam that has been deflected and focused in a focused state, and the detected secondary electron beam image is detected. An inspection or measurement method for an electron beam, wherein an inspection or measurement of a pattern formed on an object to be inspected is performed based on the inspection method.

Further, according to the present invention, in the electron beam inspection or measurement method, the electron beam is obtained by shifting a focus control current or a focus control voltage applied to an objective lens of the electron optical system by a calculated positional shift. The calculation is performed based on a calibration parameter between the information corresponding to the surface height and the focus control current or the focus control voltage from the surface height at a certain location. Further, the present invention is based on the arrangement information of the pattern formed on the inspection object in a region on the inspection object which is placed on the stage and irradiated with the electron beam by the height detection device in advance. The height of the surface at any position between the intervals is determined by detecting the surface height at the interval determined in advance and obtaining the height distribution, and interpolating the height distribution (the surface height at the interval). The electron beam emitted from the electron beam source is controlled by controlling the current or the voltage applied to the objective lens of the electron optical system based on the estimated surface height at the position where the electron beam is irradiated. Secondary electrons generated from the object to be inspected by the electron beam irradiated by being deflected by the deflecting element of the optical system to be focused on the object to be inspected in a focused state, and then deflected and focused in the focused state. Line image is detected by electron beam image detection optical system An electron beam inspection or measurement method and performing inspection or measurement of a pattern formed on the inspection object based on the secondary electron beam image to be issued 該検.

According to the present invention, a charged particle beam source such as electrons and ions, a deflecting element for deflecting a charged particle beam emitted from the charged particle beam source, and a charged particle beam deflected by the deflecting element are observed. A charged particle optical system having an objective lens that focuses and irradiates the object, and charged particles generated from the observation target object by the charged particle beam that is deflected by the charged particle optical system and focused and irradiated A charged particle beam image detection optical system for detecting a line image, a projection optical system for projecting a lattice-like light beam onto the object to be observed from obliquely above the object to be observed, and a lattice-like image projected by the projection optical system. A detection optical system that forms a lattice-like light beam reflected on the surface of the object to be observed by the light beam and detects the position of the optical image, and an optical image composed of the lattice-like light beam detected by the detection optical system The observed pair based on the position change of An optical height detection device configured to detect the height of the surface in the region of Butsujo optically,
The charged particle beam is inspected by controlling the current or the voltage applied to the objective lens of the charged particle beam optical system based on the height of the surface on the object to be observed detected by the optical height detection device. A charged particle beam image observation apparatus characterized by having focus control means for focusing on an object in a focused state and display means for displaying a charged particle beam image detected by the charged particle beam image detection optical system. is there.

Further, according to the present invention, in the charged particle beam image observation apparatus, the deflection of the electron optical system is further performed based on a height of a surface on the inspection object detected by the optical height detection apparatus. And a deflection control unit for correcting an image distortion including a magnification error of the charged particle beam image generated based on the focus control by correcting a deflection amount to the element. Further, the present invention provides the charged particle beam image observation device, further comprising, from the height of the surface on the inspection object detected by the optical height detection device, the height of the surface on the inspection object. Control current or focus control voltage is calculated based on a calibration parameter between the focus control current and focus control voltage, and the calculated focus control current or focus control voltage is provided to the objective lens of the charged particle optical system. Focus control means for controlling the charged particle beam to be focused on the object to be inspected in a controlled manner, and display means for displaying a charged particle beam image detected by the charged particle beam image detection optical system. It is characterized by the following.

According to the present invention, a charged particle beam source such as electrons and ions, a deflecting element for deflecting a charged particle beam emitted from the charged particle beam source, and a charged particle beam deflected by the deflecting element are observed. An electron optical system having an objective lens that focuses and irradiates the object, and a charged particle beam image generated from the observation target object by a charged particle beam that is deflected by the electron optical system, focused, and irradiated A charged particle beam image detecting optical system for detecting the object, a projection optical system for projecting a lattice-like light beam onto the object to be observed from obliquely above the object to be observed, and a lattice-like light beam projected by the projection optical system. A detection optical system that forms a lattice-like light beam reflected on the surface of the object to be observed and detects the position of the optical image, and the position of the optical image composed of the lattice-like light beam detected by the detection optical system On the observed object based on the change An optical height detection device configured to optically detect the height of the surface in the region, and the electronic height based on the height of the surface of the object to be observed detected by the optical height detection device. Focus control means for controlling the relative position in the height direction between the focal position of the optical system and the table on which the object to be observed is placed to focus the charged particle beam on the object to be observed in a focused state; Display means for displaying a charged particle beam image detected by the charged particle beam image detection optical system.

Further, the present invention provides a charged particle beam source such as an electron and an ion, a deflecting element for deflecting a charged particle beam emitted from the charged particle beam source, and a charged particle beam deflected by the deflecting element. A charged particle optical system having an objective lens that focuses and irradiates the target object, and charged particles generated from the workpiece by the charged particle beam that is deflected by the charged particle optical system and focused and irradiated. A charged particle beam image detection optical system for detecting a line image, a projection optical system for projecting a lattice-like light beam onto the object to be processed from obliquely above the object to be processed, and a lattice-shaped image projected by the projection optical system A detection optical system that forms a lattice-like light beam reflected on the surface of the object to be processed by the light beam and detects the position of the optical image, and an optical image composed of the lattice-like light beam detected by the detection optical system The workpiece pair based on the change in the position of An optical height detection device configured to detect the height of the surface in the region of Butsujo optically,
The charged particle beam is inspected by controlling a current or a voltage applied to the objective lens of the charged particle beam optical system based on the height of the surface on the workpiece detected by the optical height detection device. A charged particle beam image processing apparatus comprising: a focus control unit that focuses on a target object in a focused state; and a unit that processes the surface of the target target object by a charged particle beam that is converged and irradiated. .

Further, according to the present invention, in the charged particle beam image processing apparatus, furthermore, the deflection of the electron optical system is performed based on a surface height on the inspection object detected by the optical height detection apparatus. The image forming apparatus further includes a deflection control unit that corrects an amount of deflection to the element and corrects image distortion including a magnification error of the charged particle beam image generated based on the focus control. Further, the present invention provides a projection optical system that projects a lattice-like light beam onto an object from obliquely above, and a lattice-like light beam reflected from the surface of the object by the lattice-like light beam projected by the projection optical system. A detection optical system that forms an image, receives a lattice-shaped optical image corresponding to the image-forming state, converts the light into a lattice-like signal, and detects the signal, based on the lattice-like signal detected by the detection optical system. An image formation state with a small detection error is searched for, and a weighting process adapted to the image formation state with a small detection error is performed on a grid-like signal detected by the detection optical system, thereby forming a grid-like state. Calculating means for obtaining a movement amount or a phase change amount with respect to the reference signal corresponding to the reference height as a signal to obtain information corresponding to the height of the surface of the object. It is a detection device. Further, the present invention provides a projection optical system that projects a lattice-like light beam onto an object from obliquely above, and a lattice-like light beam reflected from the surface of the object by the lattice-like light beam projected by the projection optical system. A detection optical system which forms an image, receives a lattice-shaped optical image corresponding to the image-forming state, converts the light into a lattice-like signal, and detects the lattice-like signal. Calculating means for obtaining a movement amount or a phase change amount with respect to a reference signal as a lattice signal by performing selective processing such that a detection error is reduced, and obtaining information corresponding to the height of the surface of the object; It is an optical height detecting device characterized by having:

Also, the present invention provides a projection optical system for forming and projecting each of a plurality of slit-like or spot-like luminous fluxes from obliquely above at a plurality of different places close to each other on an object;
The projection optical system forms an image of each of the plurality of light beams reflected by the plurality of light beams imaged and projected on a plurality of locations on the inspection object, receives the formed plurality of light beam images, and A detection optical system that converts the signals into a plurality of signals and detects the signals, and a moving amount or a phase change amount with respect to a reference signal corresponding to a reference height in each of the plurality of detection signals detected by the detection optical system; The information corresponding to the height of the surface at the arbitrary position is obtained, and the obtained information corresponding to the height of the surface at a plurality of positions is interpolated (including interpolation and extrapolation) to obtain the information of the surface at an arbitrary position. A calculating means for calculating information according to the height.

As described above, according to the above configuration,
An electron beam image (SEM image) by reducing image distortion due to deflection and aberration of the electron optical system and reduction in resolution due to defocus without being affected by the surface state of the inspection object.
Inspection and length measurement based on an electron beam image (SEM image) can be performed with high accuracy and high reliability by improving the quality of the image. Further, with the above configuration, the detection of the surface height of the inspection object and the control of the electron optical system can be executed in real time, and a high-resolution electron beam image (SEM image) free from image distortion due to continuous stage movement. As a result, the inspection performance and its stability can be improved, and the inspection time can be reduced. Further, with the above-described configuration, the quality of the charged particle beam image is improved by reducing the image distortion due to the deflection and aberration of the charged particle optical system and the reduction in resolution due to defocus, and a good image is used. Observations can be made. Further, with the above configuration, it is possible to realize an electron beam exposure apparatus and a focused ion beam processing apparatus capable of exposing and processing an extremely fine pattern with high resolution without image distortion.

[0024]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments according to the present invention will be described with reference to the drawings. An embodiment of an automatic inspection system for inspecting and measuring a fine circuit pattern formed on a semiconductor wafer or the like as an object to be inspected according to the present invention will be described. Defect inspection of a fine circuit pattern formed on a semiconductor wafer or the like is performed by comparing a pattern to be inspected with a non-defective pattern or a pattern of the same type on a wafer to be inspected. Also in the case of an appearance inspection using an electron microscope image (SEM image), a defect inspection is performed by comparing images of patterns. In addition, the length measurement (SE) by a scanning electron microscope for measuring the line width and the hole diameter of a fine circuit pattern used for setting the manufacturing process conditions of a semiconductor device or for monitoring, etc.
M measurement), the length measurement is automated by image processing. Thus, in the case of a comparative inspection for detecting the defect by comparing the electron beam images of the same pattern, or when processing the electron beam image to measure the line width of the pattern,
The quality of the obtained electron beam image greatly affects the reliability of the inspection result. The quality of an electron beam image is degraded due to image distortion caused by deflection or aberration of the electron optical system, or a decrease in resolution due to defocus. These deteriorations in image quality degrade the performance of comparative inspection and length measurement. If the height of the surface of the object to be inspected is not constant and the inspection is performed under the same conditions over the entire range, as shown in FIG. Line image (SE
M image). As a result, FIG.
(Electron beam image of area A (height za))
a and the defocused image shown in FIG.
(Electron beam image of height zb) b and defocused image shown in FIG. 1D (electron beam image of region C (height zc))
If the inspection is performed by comparing with c, a correct inspection result cannot be obtained. Further, in these images, the width of the pattern changes and the result of edge detection of the image cannot be obtained stably, so that the line width and hole diameter of the pattern cannot be measured stably.

Next, an embodiment of the electron beam apparatus according to the present invention will be described with reference to FIG. An electron beam apparatus 100 composed of an electron beam column that irradiates an electron beam onto a test object (sample) 106 includes an electron beam source 101 that emits an electron beam.
And a deflection element 102 for two-dimensionally deflecting the electron beam emitted from the electron beam source 101, and an objective lens 103 controlled to focus the electron beam on a sample 106. That is, the electron beam emitted from the electron beam source 101 passes through the deflecting element 102 and the objective lens 103, and the sample 10
6 is focused and illuminated. The sample 106 is placed on an XY stage 105, and the position is measured by a laser measuring system 107. Further, in the case of the SEM device, secondary electrons emitted from the sample 106 are detected by the secondary electron detector 104, and the detected secondary electron signal is converted by the A / D converter 122 into an S / D converter 122.
The image is converted into an EM image, and the converted SEM image is processed by the image processing unit 124. For example, SE for length measurement
In the case of M, the image processing means 124 measures the distance between the patterns in the designated image. In the case of the observation SEM (appearance inspection based on the SEM image), the image processing unit 124 performs processing such as image enhancement.

Next, means for preventing the electron beam image from deteriorating in the electron beam apparatus (observation SEM apparatus, length measuring SEM apparatus) will be described. That is, the quality of the electron beam image is degraded due to image distortion caused by deflection or aberration of the electron optical system, or a decrease in resolution due to defocus. As means for preventing such deterioration of image quality, a height detecting device 200 including a height detecting optical device 200a and a height calculating device 200b, a focus control device 109, a deflection signal generating device 108, and an overall control device 120 are used. Be composed. The height detection device 200 including the height detection optical device 200a and the height calculation means 200b is configured in substantially the same manner as the embodiment described later with reference to FIG.
Projection optical system 270 around the optical axis 110 of the electron beam
And the detection optical system 271 are installed on the left and right. The projection optical system 270 of each height detecting optical device 200a includes a light source 201, a condenser lens 202, a mask 203 having a multi-slit pattern, and a projection lens 210. The detection optical system of each height detection optical device 200 a enlarges the detection lens 215 and the intermediate multi-slit image formed by the projection lens 215 to form an enlargement lens 264 formed on the line image sensor 214. , Mirror 206 and cylindrical lens (cylindrical lens) 213
And a line image sensor 214.

The electron beam is irradiated by the illumination optical system of the height detecting optical device 200a as described above,
A multi-slit pattern is projected onto a measurement position 110 on the sample 106 where an image is to be detected, and the specular reflection image is formed by the detection optical system of each height detection optical device 200a and detected as a multi-slit image. First, the light source 201
Is emitted by the condenser lens 202.
The light source image is focused on the pupil of the detection lens. The light further illuminates the mask 203 on which the multi-slit pattern is drawn, and the sample 10
6 is projected. The multi-slit pattern projected on the sample is specularly reflected and forms an image before the magnifying lens 264 through the detection lens 215 arranged on the opposite side to the optical axis 110 of the electron beam. This intermediate image is displayed on a magnifying lens 26.
4 forms an image on the line image sensor 214. In this embodiment, a cylindrical lens 213 is placed in front of the line image sensor 214, and the longitudinal direction of the slit is compressed to narrow the slit on the line image sensor 214. Assuming that the magnification of the detection optical system is m, the height of the sample is z
, The entire multi-slit image shifts by 2 mz · sin θ. Utilizing this, height calculation means 2
00b calculates the shift amount of each slit image from the signal of the multi-slit image detected from the detection optical system of each height detection optical device 200a, and shifts the calculated shift amount of the multi-slit image by the shift amount. The height of the sample surface is calculated by giving a weight according to the amount and performing weighted averaging, and the height of the electron beam at the optical axis 110 is obtained.

Although the height detecting optical device 200a has been described in the case where it is substantially the same as the embodiment shown in FIG. 11, the embodiment shown in FIG. 28 or FIG.
It is obvious that the optical system of the embodiment shown in FIG. 30 or the embodiment shown in FIGS. 30 and 41 may be used. The focus control device 109 drives and controls the electromagnetic lens or the electrostatic lens 103 based on the height data 190 obtained by the height calculation means 200b, and focuses the electron beam on the surface of the sample 106. Here, as the height data, FIG.
3, height information detected in advance by means of the embodiment described with reference to FIGS. 44 and 45 may be used. The deflection signal generator 108 generates a deflection signal 141 to the deflection element 102. At this time, the image magnification variation and the electromagnetic lens 10
The deflection signal 141 is corrected based on the height data 190 obtained by the height calculating means 200b so as to compensate for the image rotation accompanying the control of No. 3. Note that if an electrostatic lens is used instead of the electromagnetic lens as 103, the image rotation associated with focus control does not occur.
This eliminates the need to correct the image rotation. Also, 10
3 is composed of a combination of an electromagnetic lens and an electrostatic lens,
If the main focusing function is provided to the electromagnetic lens and the focal position is adjusted by the electrostatic lens, it is not necessary to correct the image rotation by the height data 190. The stage 105 is an XYZ stage, and the focus controller 109 controls the electromagnetic lens or the electrostatic lens 103.
Instead of directly controlling the focal position of the stage 105, the height of the stage 105 may be controlled.

The overall control device 120 includes an electron beam device (SE
(M apparatus) The entire control is performed, and the processing result processed by the image processing unit 124 is displayed on the display unit 143 and stored in the storage unit 142 together with the coordinate data on the sample. In addition, the overall control device 120 includes a height calculating unit 200.
b, focus control device 109, and deflection signal generator 10
8 to realize high-speed autofocus control in the electron beam apparatus and image magnification correction and image rotation correction accompanying the focus control. In addition, the overall control device 120 also executes calibration of a height detection value described later. FIG. 3 shows the SEM
1 shows a configuration of an embodiment of a visual inspection device using images. That is, a visual inspection apparatus using an SEM image includes an electron beam source 101 that generates an electron beam, a beam deflector 102 that scans a beam to form an image, and an inspection target that includes an electron beam from a wafer or the like. Objective lens 103 to form an image on object 106, objective lens 103 and inspected object 1
06 and the inspection object 1
, A stage 105 for scanning and positioning the same, a secondary electron detector 104 for detecting secondary electrons generated from the inspection object 106, a height detection optical device 200a, and an objective lens. A focus position control device 109 for adjusting the focus position of the electron beam source 103; a source potential adjusting means 121 for controlling the voltage of the electron beam source; and a deflection control device (deflection signal generation) for controlling the beam deflector 102 to realize beam scanning. Device) 108, grid potential adjusting means 127 for controlling the potential of the grid 118, sample stage potential adjusting means 125 for adjusting the potential of the sample stage, and A for A / D converting a signal from the secondary electron detector 104. / D converter 122 and A /
Image processing circuit 12 for processing a digital image after D conversion
4, the image memory 123 for this, and the stage 105
Stage control means 126 for controlling the operation, the overall control unit 120 for controlling the whole of these, and the vacuum sample chamber (vacuum vessel) 1
00. The height detection value 190 of the height detection sensor 200 is always fed back to the focus position control device 109 and the deflection control device (deflection signal generation device). When inspecting the inspection object 106, the overall control unit 120 continuously moves the stage 105 by issuing an instruction to the stage control device 126. At the same time, the overall control unit 120 issues a command to a deflection control device (deflection signal generation device) 108, and the deflection control device 108 drives the beam deflector 102 to scan the electron beam in a direction orthogonal to the direction. At the same time, the deflection controller 108 receives the height detection value 190 obtained from the height calculator 200b and corrects the deflection direction and the deflection width. Further, the focus position control means 109 adjusts the electromagnetic lens or the electrostatic lens 103 according to the height detection value 190 obtained from the calculation means 250.
To correct the focus height of the electron beam. At this time, the secondary electrons generated from the sample 106 are detected by the secondary electron detector 10.
4 and is input to the A / D converter 122 and focused on S
EM images are obtained continuously. As described above, in the case of the appearance inspection based on the SEM image, a 2 ×
Since it is necessary to capture a two-dimensional SEM image, while continuously moving the stage 105, the beam deflector 102 is driven to scan the electron beam in a direction substantially perpendicular to the moving direction of the stage 105, and the secondary electron detector 104 is scanned. It is necessary to detect a two-dimensional secondary electron image signal. That is, stage 10
5 while continuously moving the beam deflector 5 in the X direction, for example.
02, the electron beam is scanned in the Y direction substantially perpendicular to the moving direction of the stage 105, and then the stage 105 is moved stepwise in the Y direction. The secondary electron detector 104 must detect the two-dimensional secondary electron image signal by driving the 102 and scanning the electron beam in the Y direction substantially orthogonal to the moving direction of the stage 105. Then, the image processing circuit 1
Reference numeral 24 realizes a comparison inspection by comparing the electron beam image delayed by the image memory 123 with the image directly input from the A / D converter 122 to compare the corresponding image of the repetition pattern. it can. The overall control unit 120 controls the image processing circuit 124 and receives the inspection result at the same time, displays the inspection result on the display unit 143, and stores the inspection result in the storage unit 142. In the embodiment shown in this figure, the objective lens 193 having excellent responsiveness is used.
Focusing is performed by controlling the control current flowing through
Alternatively, the stage 105 may be moved up and down. But,
If the stage 105 is moved up and down for control, the response will be poor.

Further, an appearance inspection apparatus using SEM images will be described with reference to FIGS. S according to the present invention
FIG. 4 shows an embodiment of a visual inspection device using an EM image. Here, the object to be inspected 106 such as a wafer is scanned by the electron beam 112, electrons generated from the object to be inspected 106 by irradiation of the electron beam are detected, and an electron beam image of the scanned portion is detected based on the intensity change. And pattern inspection is performed using the electron beam image. As the inspection object 106, for example, there is the semiconductor wafer 3 as shown in FIG. On the semiconductor wafer 3, a large number of chips 3a which are finally the same product are arranged. The pattern layout inside the chip 3a is composed of a memory mat section 3c in which memory cells are regularly arranged two-dimensionally at the same pitch, and a peripheral circuit section 3b, as shown in the enlarged view of FIG. When applied to the inspection of the pattern of the semiconductor wafer 3, a detection image of a certain chip (for example, the chip 3d) is stored and compared with a detection image of another chip (for example, the chip 3e) (hereinafter, referred to as a chip). Alternatively, a detected image in a certain memory cell (for example, the memory cell 3f) is stored and compared with a detected image in another cell (for example, the cell 3g) (hereinafter, referred to as "cell comparison"). ")" To recognize the defect. If the repetitive patterns of the inspection object 106 (chips or cells in the case of a semiconductor wafer, for example) are strictly equal and if the same detected images can be obtained, they will not match when the images are compared. Since only the defect is detected, the defect can be recognized.

However, actually, there is a mismatch between the two images even in the normal part. The mismatch in the normal part includes a mismatch caused by the object to be inspected and a mismatch caused by the image detection system. The inconsistency caused by the inspection object is caused by a subtle difference between repeated patterns generated during a wafer manufacturing process such as exposure, development, and etching. This appears as a slight difference in pattern shape and a difference in gradation value on the detected image. The inconsistency caused by the image detection system is caused by variations in the amount of illumination light, vibration of the stage, various electrical noises, deviations in the detection positions of the two images, and the like. These appear on the detected image as differences in the gradation values of the partial images, pattern distortions, and image displacements. In the first embodiment according to the present invention, a detected image (first two-dimensional image) in which the gradation value of coordinates (x, y) aligned in pixel units is f1 (x, y) and coordinates The threshold value (permissible value) at the time of determining a defect by comparing with a comparison image (second two-dimensional image) whose gradation value of (x, y) is g1 (x, y) is determined by the pattern position. It is set for each pixel in consideration of a shift, a difference in gradation value, and the like, and a defect is determined based on a threshold value (allowable value) set for each pixel.

As shown in FIGS. 4 and 7, the pattern inspection system includes a detecting unit 115, an image extracting unit 121,
The image processing unit 124 includes an overall control unit 120 that controls the entire system. The pattern inspection system includes an inspection room 100 in which the room is evacuated, and a spare room (not shown) for loading and unloading the inspection object 106 into and from the inspection room 100. Preliminary room is inspection room 100
It is configured to be able to evacuate independently of the vacuum pump. First, the detection unit 115 will be described with reference to FIGS. That is, the inside of the inspection room 100 in the detection unit 115
It is roughly composed of an electron optical system 116, an electron detection unit 117, a sample chamber 119, and an optical microscope unit 118.
The electron optical system 116 includes an electron gun 31 (101), an electron beam extraction electrode 11, a condenser lens 32, a blanking deflector 13, a scanning deflector 34 (102), an aperture 14,
It comprises an objective lens 33 (103), a reflector 17, an ExB deflector 15, and a Faraday cup (not shown) for detecting a beam current. The reflection plate 17 was formed in a conical shape to have a secondary electron multiplication effect. Electronic detector 117
Among them, an electron detector 35 (104) for detecting electrons such as secondary electrons and reflected electrons is installed in the inspection room 100, for example, above the objective lens 33 (103). The output signal of the electronic detector 35 is amplified by an amplifier 36 installed outside the inspection room 100.

The sample chamber 119 includes a sample stage 30, an X stage 31, a Y stage 32, a position monitor length measuring device 107, and a substrate height measuring device 200 for inspection. In addition,
The stage may be provided with a rotary stage. The length monitor 134 for position monitoring includes the stages 31 and 32 (stage 1).
05) and the like, and the result is transmitted to the overall control unit 12.
0. The drive system of the stages 31 and 32 is also controlled by the overall control unit 120. As a result, the overall control unit 120 can accurately grasp the region and position where the electron beam 112 is irradiated based on these data. Inspection board height measuring instrument 200
Is the inspection object 1 placed on the stages 31 and 32
06 is measured. Then, based on the measurement data measured by the inspected substrate height measuring device 200,
The objective lens 33 (10) for narrowing down the electron beam 112 finely.
The focal length of 3) is dynamically corrected, so that the electron beam can be irradiated while the inspection area is always in focus. In FIG. 7, the height measuring device 200 is
Although installed inside the inspection room 100, the inspection room 100
And a method of projecting light into the inspection room 100 through a glass window or the like.

The optical microscope section 118 is installed near the electron optical system 116 in the room of the inspection room 100 and at a position away from the electron optical system 116 so as not to affect each other. Is naturally a known value. The X stage 31 or the Y stage 32 is configured to reciprocate a known distance between the electron optical system 116 and the optical microscope unit 118. The optical microscope section 118 includes a light source 61, an optical lens 62, and a CCD camera 63. The optical microscope unit 118 detects an optical image of a circuit pattern formed on, for example, the semiconductor wafer 3, which is the inspection object 106, and calculates a rotational shift amount of the circuit pattern based on the detected optical image. The calculated rotation deviation amount is transmitted to the overall control unit 120. Then, the overall control unit 120 can correct the rotation deviation amount by, for example, rotating the rotation stage. Further, the overall control unit 120 sends the rotational deviation amount to a correction control circuit 120 ′, and the correction control circuit 120 ′ corrects, for example, the scanning deflection position of the electron beam by the scanning deflector 34 based on the rotational deviation amount. It is possible to correct the rotational deviation. The optical microscope unit 118 detects an optical image of a circuit pattern formed on, for example, the semiconductor wafer 3 that is the inspection target object 106, and displays and observes the optical image on the monitor 50, for example. The overall control unit 1 uses the input means to input the coordinates of the inspection area based on the image.
20 to input the inspection area to the overall control unit 1.
20 can also be set. For example, the pitch between chips in a circuit pattern formed on the semiconductor wafer 3 or the repetition pitch of a repetition pattern such as a memory cell can be measured in advance and input to the overall control unit 120. The optical microscope section 118
Is inside the inspection room 100 in FIG.
0, and an optical image of the semiconductor wafer 1 may be detected through a glass window or the like.

As shown in FIG. 4 and FIG.
The electron beam that has exited (101) is
2. Through the objective lens 33 (103), the beam diameter on the sample surface is reduced to about the pixel size. At this time, the ground electrode 38 (118) and the retarding electrode 37
By applying a negative potential to the sample and decelerating the electron beam between the objective lens 33 (103) and the inspection object (sample) 106, high resolution in a low acceleration voltage region is achieved. When the electron beam is irradiated, electrons are generated from the inspection object (wafer 3) 106. Synchronous with the repetitive scanning of the electron beam in the X direction by the scanning deflector 34 (102) and the continuous movement of the inspection object (sample) 106 in the Y direction by the stage 2 (105), from the inspection object 106 By detecting the generated electrons, a two-dimensional electron beam image of the inspection object can be obtained. Electrons generated from the inspected object are detected by the detector 35 (10
4) and are amplified by the amplifier 36. here,
In order to enable high-speed inspection, an electrostatic deflector having a high deflection speed is used as the deflector 34 (102) for repeatedly scanning the electron beam in the X direction.
As 1 (101), use is made of a thermal field emission type electron gun capable of shortening the irradiation time because the electron beam current can be increased, and a semiconductor detector capable of high-speed driving is used as the detector 35 (104). It is desirable.

Next, the image extracting section 140 will be described with reference to FIGS. 4, 7 and 8. That is, the electron detection signal detected by the electron detector 35 (104) in the electron detector 117 is amplified by the amplifier 36, and is converted into digital image data (gradation image data) by the A / D converter 39 (122). Is done. Then, the A / D converter 39 (1
The output of 22) is transmitted by an optical conversion means (light emitting element) 23, a transmission means (optical fiber cable) 24, and an electric conversion means (light receiving element) 25. According to this configuration, the transmission unit 24 transmits the A / D converter 39 (12
The transmission speed may be the same as the clock frequency of 2). A
The output of the / D converter 39 is the light conversion means (light emitting element) 23
Is converted into an optical digital signal, transmitted optically by a transmission means (optical fiber cable) 24, and converted into digital image data (gradation image data) by an electric conversion means (light receiving element) 25. The conversion into the optical signal and the transmission of the light signal in this way are performed in order to guide the electrons 52 from the reflection plate 17 to the semiconductor detector 35 (104). This is because it is necessary to float the device 35, the amplifier 36, the A / D converter 39, and the light converting means (light emitting element) 23 to a positive high potential by a high voltage power supply (not shown). More precisely, only the semiconductor detector 35 needs to be set to a positive high voltage.
However, it is desirable that the amplifier 36 and the A / D converter 39 are located in the immediate vicinity of the semiconductor detector in order to prevent noise contamination and signal deterioration, and it is difficult to keep only the semiconductor detector 35 at a positive high voltage. , And all the components are set to a high voltage. That is, since the transmission means (optical fiber cable) 24 is formed of a highly insulating material, the light conversion means (light emitting element) 23
In, after the image signal at the positive high potential level has passed through the transmission means (optical fiber cable) 24, an output of the ground level image signal is obtained from the electric conversion means (light receiving element) 25.

The pre-processing circuit (image correction circuit) 40 includes a dark level correction circuit 72, a fluctuation correction circuit 73 for an electron source, a shading correction circuit 74, and the like. Digital image data (gradation image data) 71 obtained from the electric conversion means (light receiving element) 25 is subjected to image correction such as dark level correction, electron source fluctuation correction and shading correction in a preprocessing circuit (image correction circuit) 40. Is performed. As shown in FIG. 9, the dark level correction in the dark level correction circuit 72 is performed based on the detection signal 71 during the beam blanking period extracted based on the scan line synchronization signal 75 obtained from the overall control unit 120. Correct the level. That is, the reference signal for correcting the dark level is updated for each scanning line by setting the average of the gradation values of a specific number of pixels at a specific position, for example, during the beam blanking period as the dark level. As described above, in the dark level correction circuit 72, the detection signal detected during the beam blanking period is subjected to the dark level correction to the reference signal updated for each line. As shown in FIG. 9, the fluctuation correction of the electron source in the fluctuation correction circuit 73 of the electron source is performed by converting the detection signal 76 subjected to the dark level correction into a Faraday detection for detecting the beam current at a correction cycle (for example, 100 kHz line unit). This is performed by normalizing the beam current 77 monitored by a cup (not shown). Since the fluctuation of the electron source does not fluctuate rapidly,
The beam current detected several lines before may be used. As shown in FIG. 9, the shading correction in the shading correction circuit 74 corrects a light amount variation due to the beam scanning position 79 obtained from the overall control unit 120 with respect to the detection signal 78 corrected for the fluctuation of the electron source. .
That is, the shading correction performs correction (normalization) for each pixel based on the reference brightness data 83 detected in advance. The shading correction reference data 83 is detected in advance, and the detected image data is temporarily stored in an image memory (for example, 147).
It is transmitted to a computer provided in the general control unit 120 or a higher-level computer connected to the general control unit 120 via a network, and transmitted to a computer provided in the general control unit 120 or a higher-level computer connected to the general control unit 120 via a network. Is created by processing with software on a computer. The shading correction reference data 83 is stored in the overall control unit 120.
Calculated and held in advance by a higher-level computer connected to the network, downloaded at the start of the inspection,
The CPU of the shading correction circuit 74 may take in the downloaded data. Regarding the correspondence to the entire visual field width, the number of pixels (for example, 102
Time that is outside the inspection area (time from the end of one visual field inspection to the start of the next one visual field inspection) with two correction memories (4 pixels)
By switching the memory. The correction data has the maximum number of pixels (for example, 5000 pixels) of the maximum width of the electron beam, and may be rewritten by the CPU in each correction memory by the end of the next one visual field inspection.

As described above, the digital image data (gradation image data) 71 obtained from the electric conversion means (light receiving element) 25 is corrected for the dark level (the dark level is corrected based on the detection signal 71 during the beam blanking period). to correct.),
After correcting the fluctuation of the electron beam current (monitoring the beam current intensity and normalizing the signal with the beam current) and performing shading correction (correcting the light intensity variation due to the beam scanning position), the corrected digital image data (gradation) The filtering processing circuit 81 performs filtering processing on the (image data) 80 using a Gaussian filter, an average value filter, an edge emphasis filter, or the like to improve image quality. Further, if necessary, the image distortion is corrected. These pre-processes are for converting the detected image so as to be advantageous in the subsequent defect determination process.

A delay circuit 4 composed of a shift register or the like
1 delays a digital image signal (gradation image signal) 82 whose image quality has been improved from the preprocessing circuit 40 by a fixed time. If it is time to move (d1 in FIG. 5), the delayed signal g0 and the undelayed signal f0 become image signals at the same location on adjacent chips, and the above-described chip comparison inspection is performed. Alternatively, if the delay time is obtained from the overall control unit 120 and the stage 2 is moved by the pitch of the memory cell (d2 in FIG. 5),
The delayed signal g0 and the non-delayed signal f0 become image signals at the same location of the adjacent memory cell, and the cell comparison test described above is performed. As described above, the delay circuit 41 is configured such that an arbitrary delay time can be selected by controlling the pixel position to be read out based on the information obtained from the overall control unit 120. As described above, the digital image signals (gradation image signals) f0 and g0 to be compared are extracted from the image extracting unit 120. Hereinafter, f0 is called a detected image and g0 is called a comparative image. As shown in FIG. 7, the comparison image signal g0 is stored in a first image storage unit 46 composed of a shift register, an image memory and the like, and the detected image signal f0 is stored in a first image storage unit composed of a shift register and an image memory. It may be stored in the two-image storage unit 47. As described above, the first image storage unit 46 may constitute the delay circuit 41, and the second image storage unit 47 is not always necessary. The electron beam image captured in the preprocessing circuit 40 and the second image storage unit 47 or the optical image detected by the optical microscope unit 118 is displayed on the monitor 5.
0 is displayed and can be observed.

Next, the image processing section 124 will be described with reference to FIG. From the preprocessing unit 40, the inspection object 106
A detection image f0 (x, y) represented by a tone value (shade value) for a certain inspection area above is obtained, and a certain inspection area on the inspection object 106 serving as a reference to be compared is obtained from the delay circuit 41. , A comparison image (reference image: reference image) g0 (x, y) expressed by the gradation value (shade value) of is obtained. In the pixel-based positioning unit 42, the detected image f0 (x,
y), the displacement amount of the comparison image g0 (x, y) is 0 to 1
In other words, the position where the "matching degree" between f0 (x, y) and g0 (x, y) is maximum is 0-1.
For example, the position of the comparison image is shifted so as to be between the pixels. As a result, for example, as shown in FIG.
(x, y) and the comparison image g0 (x, y) are aligned in units of one pixel or less. In FIG. 6, square eyes indicated by chain lines indicate pixels. This pixel is a unit that is detected by the electronic detector 35, sampled by the A / D converter 39 (122), and converted into a digital value (gradation value: gray value). That is, the pixel unit indicates the minimum unit detected by the electronic detector 35. In addition, the following (Equation 1) or the like can be considered as the “degree of matching”.

Max | f0−g0 |, ΣΣ | f0−g0 |, ΣΣ (f0−g0) ² (Equation 1) max | f0−g0 | is the detected image f0 (x, y) and the comparison image g0
Indicates the maximum value of the absolute value of the difference from (x, y). ΣΣ | f0-g
0 | is the sum in the image of the absolute value of the difference between the detected image f0 (x, y) and the comparison image g0 (x, y). ΣΣ (f0−g0)
² indicates a value obtained by squaring the difference between the detected image f0 (x, y) and the comparison image g0 (x, y) and integrating the difference in the x and y directions.
The processing content changes depending on which of the above (Equation 1) is adopted. Here, a case where ΣΣ | f0−g0 | is adopted will be described. The shift amount in the x direction of the comparative image g0 (x, y) is mx, and the shift amount in the y direction is my (where mx, my
Is an integer), and each of e1 (mx, my) and s1 (mx, my) is defined as shown in the following (Formula 2) and (Formula 3).

E1 (mx, my) = ΣΣ | f0 (x, y) −g0 (x + mx, y + my) | (Equation 2) s1 (mx, my) = e1 (mx, my) + e1 (mx + 1, my ) + E1 (mx, my + 1) + e1 (mx + 1, my + 1) (Formula 3) In the formula (2), ΣΣ represents the sum in the image. What we want to find here is the shift amount in the x direction that minimizes s1 (mx, my) is the value of mx, and the shift amount in the y direction is the value of my.
And my are changed to ± 0, 1, 2, 3, 4... N, in other words, the comparative image g0 (x, y) is shifted at the pixel pitch, and the s1 (mx , My). Then, a value mx0 of mx and a value my0 of my that minimize the value are obtained. Note that the maximum shift amount n of the comparison image needs to be set to a larger value as the position accuracy is lower, according to the position accuracy of the detector 115. From the pixel-by-pixel alignment unit 42, the detected image f0 (x, y) remains unchanged.
The comparison image g0 (x, y) is output shifted by (mx0, my0). That is, f1 (x, y) = f0 (x, y), g1
(x, y) = g0 (x + mx0, y + my0).

In a displacement detection unit (not shown) for pixels or less, images f 1 (x, y) and g 1 aligned on a pixel-by-pixel basis.
(x, y) is divided into small areas (for example, a partial image composed of 128 * 256 pixels), and a positional shift amount equal to or less than a pixel (the positional shift amount is a real number of 0 to 1) for each of the divided areas (partial images). Is calculated). The division into small regions is for dealing with image distortion, and it is necessary to set the region to a region where distortion is negligible. As a measure of the degree of consistency, there is also an option as shown by the following equation (1). Here, an example in which the third “sum of squares of difference” (ΣΣ (f0−g0) ² ) is used is shown. . An intermediate position between f1 (x, y) and g1 (x, y) is defined as a position shift amount 0, and as shown in FIG.
δx, −δy in y direction, g1 is + δx in x direction, +1 in y direction
Assume that it is shifted by δy. That is, the shift amount between f1 and g1 is considered to be 2 * δx in the x direction and 2 * δy in the y direction. Since δx and δy are not integers, it is necessary to define a value between pixels to shift by δx and δy. f1
F shifted by + δx in the x direction and + δy in the y direction
2. An image g2 in which g1 is shifted by -δx in the x direction and -δy in the y direction is defined as in the following equations (4) and (5).

F2 (x, y) = f1 (x + δx, y + δy) = f1 (x, y) + δx (f1 (x + 1, y) −f1 (x, y)) + δy (f1 (x, y + 1) − f1 (x, y)) (Equation 4) g2 (x, y) = g1 (x-.delta.x, y-.delta.y) = g1 (x, y) +. delta.x (g1 (x-1, y) -g1 (x, y)) + δy (g1 (x, y-1) -g1 (x, y)) (Equation 5) (Equation 4) (Equation 5) is a so-called linear interpolation. The degree of matching e2 (δx, δy) between f2 and g2 is given by the following (Equation 6) when “sum of squares of the difference” is adopted. e2 (δx, δy) = {(f2 (x, y) −g2 (x, y)) ² (Equation 6)} is the sum in the small area (partial image). The purpose of the displacement detection unit (not shown) for pixels or less is e2 (δx, δy)
Is to determine the value Δx0 of Δx and the value Δy0 of Δy that takes the minimum value. This can be done by setting the equation obtained by partially differentiating the above equation (6) with δx and δy as 0 and solving it for δx and δy. The result is as shown in the following equations (7) and (8).

Δx = {(ΣΣC0 * Cy) * (ΣΣCx * Cy) − (ΣΣC0 * Cx) * (ΣΣCy * Cy)} / {(ΣΣCx * Cx) * (ΣΣCy * Cy) − (ΣΣCx * Cy) * (ΣΣCx * Cy)} (Equation 7) δx = {(ΣΣC0 * Cx) * (ΣΣCx * Cy) − (ΣΣC0 * Cy) * (ΣΣCx * Cx)} / {(ΣΣCx * Cx) * (ΣΣCy * Cy) − (ΣΣCx * Cy) * (ΣΣCx * Cy)} (Equation 8) Here, each of C0, Cx and Cy is represented by the following (Equation 9):
Equation (10) and equation (11) are obtained.

C0 = f1 (x, y) −g1 (x, y) (Equation 9) Cx = {f1 (x + 1, y) −f1 (x, y)} − {g1 (x−1, y) − g1 (x, y)} (Equation 10) Cy = {f1 (x, y + 1) -f1 (x, y)}-{g1 (x, y-1) -g1 (x, y)} (Equation 11) In order to obtain each of δx0 and δy0, as shown in Expressions (7) and (8), the various statistics 種 々 Ck * Ck
(However, Ck = C0, Cx, Cy) must be obtained. The statistic calculation unit 44 includes a detection image f1 (x, y) composed of tone values (shading values) aligned in pixel units obtained from the pixel-by-pixel alignment unit 42 and a comparison image g1 (x, y). Based on the above, the various statistics ΔCk * Ck are calculated. The sub CPU 45 calculates the above equation (7) and the equation (8) using ｋCk * Ck calculated by the statistic calculation unit 44.
The calculation of the expression is performed to obtain Δx0 and Δy0. The delay circuits 46 and 47 constituted by shift registers and the like are provided for delaying the image signals f1 and g1 by the time required to obtain δx0 and δy0 by a displacement detector (not shown) for pixels or less. is there.

The difference image extracting circuit (difference extracting circuit: distance extracting unit) 49 calculates the difference image (distance image) sub (x, y) between f1 and g1 having a positional shift of 2 * δx0 and 2 * δy0. Ask for. The difference image (distance image) sub (x, y) is represented by the following equation (12). sub (x, y) = g1 (x, y) -f1 (x, y) (Equation 12) The threshold value operation circuit (permissible range operation unit) 48 outputs the image signals f1 and g1 which Further, the defect determination circuit (defect determination unit) 50 uses the difference image extraction circuit (difference extraction circuit: distance Extraction unit) difference image (distance image) sub (x,
Two thresholds (allowable values indicating an allowable range) thH (x, y) and thL for determining whether or not a defect is a candidate according to the value of y)
(x, y). thH (x, y) is a threshold value (allowable value indicating an allowable range) defining the upper limit of the difference image (distance image) sub (x, y), and thL (x, y) is
This is a threshold value (allowable value indicating an allowable range) that defines the lower limit of the difference image (distance image) sub (x, y). FIG. 5 shows the configuration of the threshold value operation circuit 48. The contents of the operation in the threshold value operation circuit 48 are expressed by the following equations (Equation 13) and (Equation 14).

ThH (x, y) = A (x, y) + B (x, y) + C (x, y) (Equation 13) thL (x, y) = A (x, y) −B (x, y) −C (x, y) (Equation 14) where A (x, y) is expressed by the relationship of Equation (15), and is substantially obtained by using the positional deviation amounts δx0 and δy0 below the pixel. This is a term for correcting the threshold value according to the value of the difference image (distance image) sub (x, y). Further, B (x, y) is expressed by the relationship of Expression (16), and a slight displacement of a pattern edge (a slight difference in pattern shape and a pattern distortion) between the detected image f1 and the comparative image g1. This is a term for allowing a small positional deviation of the pattern edge when viewed locally.

Further, C (x, y) is expressed by the following equation (17), and allows a slight difference in gradation value (shade value) between the detected image f1 and the comparison image g1. Is the term. A (x, y) = {dx1 (x, y) * δx0−dx2 (x, y) * (− δx0)} + {dy1 (x, y) * δy0−dy2 (x, y) * (− δy0 )} = {Dx1 (x, y) + dx2 (x, y)} * δx0 + {dy1 (x, y) + dy2 (x, y)} * δy0 (Equation 15) B (x, y) = | {dx1 ( x, y) * α-dx2 (x, y) * (− α)} | + | {dy1 (x, y) * β−dy2 (x, y) * (− β)} | = | {dx1 ( x, y) + dx2 (x, y)} * α | + | {dy1 (x, y) + dy2 (x, y)} * β | (Formula 16) C (x, y) = ((max1 + max2) / 2 Here, α and β are real numbers of 0 to 0.5, γ is a real number of 0 or more, and ε is an integer of 0 or more.

Dx1 (x, y) is expressed by the relationship of equation (18), and represents the change in the gradation value (shade value) between the detected image f1 (x, y) and the image adjacent to the x direction + 1. Show. Also dx2
(x, y) is expressed by the relationship of equation (19), and the comparison image g1
It shows the change in the gradation value (shade value) between (x, y) and the image adjacent in the x direction minus one. Dy1 (x, y) is given by (Equation 20)
It is shown by the relationship of the expression, and shows the change of the gradation value (shade value) between the detected image f1 (x, y) and the next image in the y direction + 1. Dy2 (x, y) is expressed by the relationship of equation (21).
It shows the change in the tone value (shade value) of the comparative image g1 (x, y) with respect to the image adjacent in the y direction minus one in the y direction. dx1 (x, y) = f1 (x + 1, y) -f1 (x, y) (Equation 18) dx2 (x, y) = g1 (x, y) -g1 (x-1, y) (Equation 19) dy1 (x, y) = f1 (x, y + 1) -f1 (x, y) (Equation 20) dy2 (x, y) = g1 (x, y) -g1 (x, y-1) (Equation 21) max1 is represented by the relationship of Expression 22, and the detected image f1 (x,
The maximum gradation value (shade value) of the image adjacent to the x direction + 1 and the image adjacent to the y direction + 1, including itself in y).

Max2 is expressed by the relationship of the following equation (23).
-1 in the x direction, including itself in the comparison image g1 (x, y)
The maximum gradation value (shade value) of the next image and the next image in the y direction minus 1 are shown. max1 = max {f1 (x, y), f1 (x + 1, y), f1 (x, y + 1), f (x + 1, y + 1)} (Equation 22) max2 = max {g1 (x, y), g1 ( x-1, y), g1 (x, y-1), g (x-1, y-1)} (Equation 23) First, threshold values thH (x, y) and thL (x, y) are calculated. calculate
The first term A (x, y) in Equations (13) and (14)
Will be described. That is, thresholds thH (x, y), thL
The first term A (x, y) in the equations (13) and (14) for calculating (x, y) corresponds to the displacements δx0 and δy0 below the pixel obtained by the displacement detection unit 43. To correct the threshold value. For example, d expressed by equation (18)
Since x1 is the local change rate of the tone value of f1 in the x direction,
Dx1 (x, y) * δx0 shown in the equation (15) has a position of δ
It can be said that this is a predicted value of a change in the gradation value (shade value) of f1 when there is a shift of x0. Therefore, the first equation (15)
The term {dx1 (x, y) * δx0−dx2 (x, y) * (− δx0)} is
A value that predicts, for each pixel, how much the gradation value (shading value) of the difference image (distance image) between f1 and g1 changes when the position of f1 is shifted by δx0 and the position of g1 is shifted by −δx0 in the x direction. It can be said. Similarly, the second term can be said to be a value predicted in the y direction. That is, {dx1 (x, y) + dx
2 (x, y)} * δx0 is the local change rate {dx1 in the x direction in the difference image (distance image) between the detected image f1 and the comparison image g1.
(x, y) + dx2 (x, y)} is multiplied by the displacement δx0 to determine how much the gradation value (shade value) of the difference image (distance image) between f1 and g1 changes in the x direction. It is a value predicted for each pixel, and {dy1 (x, y) + dy2 (x, y)} * δy0
Is the difference image (distance image) between the detected image f1 and the comparison image g1
Local change rate in the y direction at {dy1 (x, y) + dy2
(x, y)} multiplied by the displacement δy0 to obtain f1 and g1
Is a value that predicts, for each pixel, how much the gradation value (shade value) of the difference image (distance image) changes in the y direction.

As described above, the threshold value thH (x,
The first term A (x, y) in y) and thL (x, y) is a term for canceling the known positional shifts δx0 and δy0.

Next, threshold values thH (x, y) and thL (x, y)
The following describes the second term B (x, y) in the equations (13) and (14) for calculating the following equation. That is, the threshold value thH (x,
y), Equation (13) for calculating thL (x, y) and (Equation 14)
The second term B (x, y) in the expression allows a slight displacement of the pattern edge (small difference in pattern shape and pattern distortion results in a small displacement of the pattern edge when viewed locally). It is a term for. As is apparent from a comparison between the equation (15) for obtaining A (x, y) and the equation (16) for obtaining B (x, y), B (x, y) is determined by the displacement α and β. This is the absolute value of the change prediction of the gradation value (shade value) of the difference image (distance image). If it is assumed that the displacement is canceled by A (x, y), adding B (x, y) to A (x, y) is more effective than adding the pattern shape or pattern from the state of alignment. This means that the position is shifted by α in the x direction and by β in the y direction in consideration of a slight displacement of the pattern edge due to a slight difference based on the distortion. That is, as a slight displacement of the pattern edge due to a slight difference based on the pattern shape or the pattern distortion, +
+ B (x, y) shown in the above equation (13) allows + β in the α and y directions. Further, subtracting B (x, y) from A (x, y) as shown in the above equation (Equation 14) means that the position is further shifted by −α in the x direction and −β in the y direction from the aligned state. It means shifting. It is -B (x, y) shown in the above equation (Equation 14) that allows a displacement of -α in the x direction and -β in the y direction. As shown in the above equations (Equation 13) and (Equation 14), the threshold value has an upper limit thH (x,
y), the lower limit thL (x, y) provides ± α, ±
This means that a position shift of β is allowed. Then, in the threshold value calculating circuit 48, the input parameters α and β are set to appropriate values, so that the permissible positional deviation due to a minute difference based on the pattern shape and the pattern distortion (the pattern edge It is possible to freely control the minute displacement.

Next, threshold values thH (x, y) and thL (x, y)
The following describes the third term C (x, y) in Expressions (13) and (14) for calculating. Threshold thH (x, y), t
The third term C (x, y) in the equations (13) and (14) for calculating hL (x, y) is a gradation value (grayscale value) between the detected image f1 and the comparison image g1. ) Is a term for allowing a slight difference. As shown in Expression (13), C (x,
The addition of y) means that the gradation value (shade value) of the comparative image g1 is allowed to be larger than the gradation value (shade value) of the detected image f1 by C (x, y). As shown by the expression (14), the subtraction of C (x, y) is performed by the gradation value (shade value) of the comparative image g1.
Means that C (x, y) is smaller than the gradation value (shade value) of the detected image f1. Here, C
(x, y) is a sum of a value obtained by multiplying a representative value (here, a max value) of a gradation value in a local region by a proportional constant γ and a constant ε, as shown in Expression (17). However, it is not necessary to stick to this function, and if the way of changing the gradation value is known, it is better to use a function suitable for it. For example, if it is known that the variation width is proportional to the square root of the gradation value, C (x, y) = (square root of (max1 + max2)) * γ + ε should be used instead of Expression (17). In the threshold value operation circuit 48, B
Similarly to (x, y), it is possible to freely control the allowable difference in gradation value (shade value) by the input parameters γ and ε.

That is, the threshold value calculating circuit (permissible range calculating unit) 48 receives the detected image f 1 (x, y) composed of the gradation value (shade value) inputted from the delay circuit 46 and the inputted image from the delay circuit 47. Image g1 (x,
y) and (Equation 18) and (Equation 19)
A first arithmetic circuit for performing the operation of {dx1 (x, y) + dx2 (x, y)}, and {dy1 by the expression (20) and the expression (21)
(x, y) + dy2 (x, y)}, a second arithmetic circuit for performing the arithmetic operation, and (max1 + max) by the equations (22) and (23).
And 2) a third arithmetic circuit for performing the arithmetic operation 2). Furthermore,
The threshold operation circuit 48 is obtained from the first operation circuit.
Based on {dx1 (x, y) + dx2 (x, y)}, δx0 obtained from the pixel or less displacement detection unit 43, and the input α parameter, a part of Expression 15 and Expression 16 are obtained. ({Dx1 (x, y) + dx2 (x, y)} * δx0 ± | {dx1 (x,
y) + dx2 (x, y)} | * α) and {dy1 (x, y) obtained from the second arithmetic circuit.
y) + dy2 (x, y)}, δy0 obtained from the displacement detection unit 43 for pixels or less, and the input β parameter.
A part of the equation (15) and a part of the equation (16) ((dy1
(x, y) + dy2 (x, y)} * δy0 ± | {dy1 (x, y) + d
y2 (x, y)} | * β) based on (max1 + max2) obtained from the third operation circuit and input γ and ε parameters, for example, 17) a sixth arithmetic circuit for performing an operation of (((max1 + max2) / 2) * γ + ε) according to the equation (17). Further, the threshold value operation circuit 48 is obtained from the fourth operation circuit ({dx
1 (x, y) + dx2 (x, y)} * δx0 + | {dx1 (x, y) +
dx2 (x, y)} | * α) and ({dy1 (x, y) + dy2 (x, y)} * δy0 + | {dy1 (x,
y) + dy2 (x, y)} | * β) and (((max1 + max2) / 2) * γ + ε) obtained from the sixth arithmetic circuit, and the upper threshold value thH (x, y) And (((max1 + max2) / 2) * γ obtained from the sixth arithmetic circuit.
+ Ε) obtained from a subtraction circuit and a sixth operation circuit ({dx1 (x, y) + dx2 (x, y)} * δx0− | {dx
1 (x, y) + dx2 (x, y)} | * α) and ({dy1 (x, y) + dy2 (x, y)} * δy0− | {dy1
(x, y) + dy2 (x, y)} | * β) and − (((max1 + max2) / 2) * γ + ε) obtained from the subtraction circuit are subjected to + operation to obtain a lower limit threshold thL (x, y) ) Is output.

It should be noted that the threshold value calculation circuit 48 can also be realized by software processing by the CPU. The parameters α, β, γ,
ε may be input using input means (for example, composed of a keyboard, a recording medium, a network, etc.) provided in the overall control unit 120. In the defect determination circuit (defect determination unit) 50, a difference image extraction circuit (difference extraction circuit) 4
9, a difference image (distance image) sub (x, y) obtained from the threshold value calculation circuit 48, a lower limit threshold value (allowable value indicating a lower limit allowable range) thL (x, y), and Upper threshold value (allowable value indicating the upper allowable range) thH
If (x, y) is used and the relationship of the following equation (24) is satisfied, the pixel at position (x, y) is a non-defective candidate; if not, the pixel at position (x, y) is a defective candidate. Is determined. The defect determination circuit 50 outputs, for example, 0 for non-defective candidate pixels and def (x, y) having a value of 1 or more indicating a mismatch amount for defective candidate pixels. thL (x, y) ≦ sub (x, y) ≦ thH (x, y) (Equation 24) In the feature extraction circuit 50a, noise removal processing (for example, de
Reduction / expansion processing is performed on f (x, y). For example, 3
If all the x3 pixels are not defective candidate pixels at the same time, the central pixel is set to, for example, 0 (non-defective candidate pixel), reduced, erased, and expanded to return to the original pixel. Do. ) For noise (for example, 3
Not all of the × 3 pixels are simultaneously defective candidate pixels. Then, after the output is deleted, the merging process is performed on the defect candidate portions that combine neighboring defect candidate portions into one. Thereafter, for each group, the barycentric coordinates and the XY projection length (the maximum lengths in the x direction and the y direction are indicated.
The square root of (the power + the square of the Y projection length) is the maximum length. ),
A feature amount 88 such as an area is calculated and output.

As described above, the image processing unit 124 controlled by the general control unit 120 irradiates an electron beam and detects the object (sample) 106 to be detected by the electron detector 35 (105). The feature amount of the defect candidate portion according to the coordinates (for example, barycentric coordinates, XY projection length, area, etc.)
88 will be obtained. The overall control unit 120 converts the position coordinates of the defect candidate portion on the detected image into a coordinate system on the inspection object (sample) 106, deletes a pseudo defect, and finally, deletes the inspection object ( The defect data including the position on the (sample) 106 and the feature amount calculated from the feature extraction circuit 50a of the image processing unit 124 is collected.

According to the present embodiment, a positional deviation of the entire small area (partial image), a small positional deviation of each pattern edge, and a small difference of a gradation value (shading value) are allowed. Therefore, the normal part is not erroneously recognized as a defect. Further, by setting the parameters α, β, γ, and ε to appropriate values, it is possible to easily control the allowable amount of displacement and change in the gradation value. Further, in the present embodiment, since a pseudo-aligned image is not generated by interpolation, the interpolation does not receive an unavoidable smoothing effect of the image. There is also the advantage of being advantageous. In fact, according to experiments performed by the inventors, a pseudo-aligned image is generated by interpolation using a result of detection of a position shift of a pixel or less, and then a position shift and a gradation are performed in the same manner as in the present embodiment. Comparing the result of calculating a threshold value that allows a change in the value and performing a defect determination with the result of performing the defect determination according to the present embodiment, the present embodiment has a defect detection performance of 5% or more in the present embodiment. Improvements were seen. The means for preventing the electron beam image from deteriorating in the electron beam apparatus described above will be further described. That is, the quality of the electron beam image is degraded due to image distortion caused by deflection or aberration of the electron optical system, or a decrease in resolution due to defocus. As means for preventing such image quality deterioration, a height detecting optical device 200a and a height calculating means 200
b, the focus control device 10
9. Deflection signal generator 108 and overall controller 12
0.

Next, an embodiment of the height detecting optical device 200a will be described. FIGS. 10 and 11 are views showing a first embodiment of the height detecting optical device 200a according to the present invention. That is, the height detecting optical device 200 according to the present invention.
a denotes a light source 201 and a mask 203 on which the same pattern illuminated with light from the light source 201, for example, a pattern in which a rectangular pattern is repeated (repeated), a projection stop 211, and S-polarized light are emitted. Polarizing filter 240
And a projection lens 210, for example, a projection optical system 270 that projects a multi-slit pattern with S-polarized light at an angle θ (θ = 60 degrees or more) from perpendicular to the sample surface 106, Lens 215 for forming an image of the specular reflection light on the light receiving surface of line image sensor 214
Lens 213 and line image sensor 2 for converging the detection diaphragm 216 and the longitudinal direction of the multi-slit pattern to the light receiving pixels of line image sensor 214
14 and a detection optical system 271 composed of Among these, the projection optical system 270 and the detection optical system 27
The height detection optical system 272 is a combination of the two. In addition,
The height calculating means 200b is provided with a line image sensor 214
Is to detect the height of the sample surface 106a from the shift amount of the multi-slit image detected from.

The light emitted from the light source 201 illuminates the mask 203 on which the same pattern, for example, a multi-slit pattern obtained by repeating a rectangular pattern is drawn. If the surface of the mask 203 is perpendicular to the optical axis of the projection optical system, the multi-slit pattern has an imaging plane parallel to the mask surface near the intersection of the optical axis of the projection optical system and the sample surface 106a. At this time, as shown in FIG. 10, if the intersection line between the plane including the optical axis of the projection optical system and the optical axis of the detection optical system (the plane parallel to the paper surface in FIG. 10) and the mask surface is the s-axis, the slit that is repeated The patterns are arranged in the direction of the s-axis.

Here, the multi-slit pattern drawn on the mask 203 does not necessarily have to be a slit pattern, and any pattern such as an ellipse or a square may be used as long as the same pattern is repeated. Good. These multi-slit light beams are reflected on the sample surface 106,
An image is formed again on the line image sensor 214 such as a CCD by the detection lens 215. Assuming that the magnification of the detection optical system is m, when the height of the sample surface 106a changes by z, the multi-slit image is shifted as a whole by 2z · sin θ · m. By utilizing this, the height of the sample surface 106a can be detected from the shift amount of the multi-slit image obtained based on the signal received by the line image sensor 214. Here, 110 is a position where the height is to be detected. When the height detecting device 200a is used as a height sensor for autofocus, 11 is used.
0 is the optical axis of the upper observation system. The pitch of the multi-slit pattern projected on the sample surface 106a is p, and the pitch of the multi-slit pattern of the projected image is p / p.
cosθ, the pitch of the multi-slit pattern formed on the mask 203 is p when the magnification of the projection optical system is m ′.
/ (Cos θ · m ′). In addition, the image sensor 21
The pitch of the pattern on 4 is p · m.

As shown in FIGS. 11A and 11B, the sample 10
In the case where the height is detected on the boundary having different reflectances on the line 6, the intensity distribution of the signal detected on the line image sensor 214 is affected by the reflectance distribution of the sample. However, by using a multi-slit pattern that is as thin as possible so as to maintain a clear image in the height detection range, a detection error due to the reflectance distribution on the surface of the object can be suppressed to a small value. This is because the detection error is caused by the deviation of the center of gravity of the slit image due to the reflectance distribution of the sample, and the absolute value of the deviation increases in proportion to the width of the slit. In the example shown in FIG. 11, the third slit from the left is affected by the unevenness of the reflectance due to the boundary of the sample, but the detection error is small due to the narrow slit width. Further, by using the height detection values obtained by the plurality of slits, it is possible to reduce a detection error and a detection variation caused by the object.

In the embodiment shown in FIG. 10, the polarizing filter 240 is disposed in front of the projection lens 210 to selectively project S-polarized light. This has an effect of suppressing a position shift due to multiple reflection in the transparent film and an effect of reducing a difference in reflectance between regions. As shown in FIG. 12, when the surface of the sample is covered with a film that is transparent to light, such as an insulating film, the projection light causes multiple reflections in the transparent film and the position of the projection light is reduced. A shift phenomenon occurs. Since the S-polarized light is more likely to reflect on the surface of the transparent film than the P-polarized light, insertion of the polarizing filter 240 makes it less likely to cause multiple reflections. On the other hand, FIG. 13 shows a graph of the reflectance of a resist and silicon as an example of the transparent film. Rs is the reflectance of S polarized light, Rp is the reflectance of P polarized light, and R is the reflectance of random polarized light. As described above, the difference between the materials in the reflectance of the S-polarized light is smaller. Also, as can be read from this graph, the reflectance increases as the incident angle increases, and the difference between the materials decreases. That is, an error is less likely to occur at the pattern boundary. Therefore, it is better that the incident angle θ is as large as possible. While an angle of 80 ° or more is ideal, it is desirable to use an incident angle of at least 60 ° or more. Note that the position of the polarizing filter 240 does not necessarily need to be in front of the projection lens 210,
Approximately the same effect can be obtained regardless of the position between the detector and the detector 214. The light source 201 may be a laser light source or a light emitting diode, but a broadband lamp such as a halogen lamp, a metal halide lamp, or a mercury lamp is more preferable. Alternatively, lasers and light-emitting diodes of a plurality of wavelengths may be mixed using a dichroic mirror. This is because the monochromatic light causes multiple interference in the transparent film and shifts the projected light, and the difference in reflectance due to the material and the pattern on the sample becomes large, so that a large error easily occurs. In the embodiment shown in FIG. 10, a cylindrical lens 213 is provided in front of the line image sensor 214. This is to condense light on the line image sensor 214, increase the amount of detected light, and reduce errors by averaging the reflected light from a large area on the sample. However, the use of the cylindrical lens 213 is not an indispensable condition, and its use should be determined according to necessity.

Next, one embodiment of an algorithm for detecting the height of the sample surface 106a by the height calculating means 200b connected to the height detecting device 200a having the height detecting optical system 272 having the structure shown in FIG. Will be described with reference to FIG. Here, for the sake of simplicity, a case where the magnification of the detection optical system is m = 1 will be described. p if m = 1
May be set to m · p.

It is assumed that the total number of slits is n, the pitch is p, and the detected waveform is y (x). The position of the peak corresponding to each slit when the height z = 0 is represented by ygo (i) (i = 0,..., N−
1). (There is a relationship of ygo (i) = ygo (0) + p * i.) Here, the total number n of the slits and the pitch p are, for a known value, the total control unit 12 for the height calculating means 200b.
Given from 0. Also, the peak position [ygo (i) (i = 0,...) Corresponding to each slit when the height z = 0.
·, N-1)], the height calculating means 200b
For example, when the height z = 0 set from the overall control unit 120 using the standard sample 130, the height may be set based on the detection waveform y (x) obtained from the height detection optical system 272. The position [ygo (i) (i = 0,..., N-1)] of the peak corresponding to each slit when the height z = 0 is determined by, for example, using the standard sample 130 as a whole control unit. When the height z = 0 set from 120, the height detection optical system 272
The detected waveform y (x) obtained from
It is also possible to manually set in the overall control unit 120, for example, by displaying it on the display means 143 provided at 0, and to provide it to the height calculating means 200b.

Next, an algorithm for detecting the height of the inspection object (sample) 106 by the height calculating means 200b as shown in FIG. 15 will be described. 1. In step S151, the signal waveform y (x) obtained from the linear sensor 214 is scanned, and the position xmax at which the maximum value is obtained is obtained. 2. In step S152, the approximate position of each peak is obtained by following the pitch p from left to right from xmax. 3. By the way, if the leftmost peak position is x0, the approximate position of the peak i is x0 + p * i. The positions of the left and right valleys are (x0 + p * ip / 2, x0 + p * i + p / 2).

Here, for simplicity, the magnification m of the detection optical system is
= 1 will be described. When m is not 1, p may be set to m · p.

Then, ymin is determined by the larger one of the left and right valleys as shown by the following equation (30). ymin = max (y (x0 + p * ip-2), y (x0 + p * i + p / 2)) (Equation 30) From the above relationship, in step S153, the threshold value (threshold value) yth is expressed as follows ( It is set by the equation (31). yth = ymin + k * (y (x0 + p * i) -ymin) (Formula 31) k is a constant around 0.3.

4. In step S154, (x0 +
The center of gravity of y (x) -yth is determined for a point where y (x)> yth between (p * ip / 2) and (x0 + p * i + p / 2), and this is defined as yg (i). I do. 5. In step S155, Δg (i) = (yg (i)
−ygo (i)) and the image shift (the amount of movement of the slit [ΣΔg (i) /
n]).

Then, the height z is calculated by multiplying the image shift by the detection gain (1 / (2 m · sin θ)) and adding an offset, based on the following equation (32).

Z = [weighted average of Δg (i)] × (1 / (2m · sin θ)) + offset value (Equation 32) Here, the reason for performing the weighted average in step S155 and the detailed implementation of the method Examples will be described later. In this embodiment, the peak of the slit image is used, but a valley between the slit images may be used instead. That is, y
For the point where (x) <yth, the barycenter of yth-y (x) is obtained and set as the barycenter of each valley image, and the entire image shift is obtained by averaging the movement amounts of these valley images. This produces the following effects. Since the detected waveform is determined by the product of the projected waveform and the reflectance of the sample surface, the bright portion of the slit image is more affected by the variation in reflectance, and the shape of the detected waveform is likely to change. On the other hand, the valley portion of the waveform is hardly affected by the reflectance of the sample surface. For this reason, it is possible to further reduce the detection error caused by the surface state of the target object by the height detection algorithm based on the measurement of the movement amount of the valley between the slit images.

In the example shown in FIG. 11, if the width of the slit is reduced, the detection error is reduced in proportion. However, there is a limit to this. The image is not sharply formed on the sensor 214, and the contrast is reduced. Therefore, as described below, it is necessary to design the height detecting optical system 272 in advance. That is, step S156 shown in FIG.
Shows conditions of design specifications for designing the height detection optical system 272 in advance. Assuming that the target height detection range is ± zmax, the multi-slit image on the image sensor 214 is defocused by ± 2zmax · cos θ at this time. On the other hand, when the period of the multi-slit pattern projected on the sample surface 106a is p and the NA (Numerical Aperture) of the detection lens 215 is NA, the depth of focus is ± a · 0.61p / NA. Therefore, it is always a design condition that the slit period p satisfies the relationship of the following (Equation 31) so that a multi-slit image can always be clearly detected. Here, a is a constant determined by how far the depth of focus is defined in a state where the amplitude is reduced. When defining the depth of focus in a state where the amplitude is reduced to １ ／,
Is about 0.6. As described above, in steps S151 to S155 shown in FIG.
Is always set to a design condition (relationship of the formula (33)) for clearly detecting the multi-slit image. (2zmax · cos θ) <(a · 0.61p / NA) (Equation 33) In the embodiment shown in FIG. It is in focus. This is because the projection lens 210 and the detection lens 215 are made telecentric on the sample side to eliminate magnification fluctuation due to up and down movement of the sample 106.

Here, the influence of the unevenness of the reflectance of the sample surface on the height detection when the height displacement of the sample surface is large will be described. For simplicity, an example in which there is one slit light beam shown in FIG. 16 will be described. FIG. 16A shows a case where the focal point 273 of the height detection optical system coincides with the sample surface 106, and FIG. 16B shows a case where defocus occurs due to a change in the height of the sample surface. As shown in FIG. 16B, when the height of the sample surface 106 changes by Z, the optical path length of the entire height detection optical system 272 changes. At this time, since the projection lens 210 and the detection lens 215 are telecentric on the sample side, magnification does not change, but the imaging position 2 of the multi-slit image formed by the projection optical system 270 is not changed.
73, the focal point position 274 conjugate with the sensor surface in the detection optical system 271, and the sample surface 275 do not match. As shown in FIG. 16A, when the image formation position of the projection optical system 270 and the focus position of the detection optical system 271 are aligned on the sample surface, detection is performed by the pattern image on the surface. Although the signal waveform 276 of the slit light beam collapses, the width of the signal waveform of the slit light beam detected by defocusing does not increase. In contrast,
When the light is projected onto the sample surface 275 in a defocused state as shown in FIG. 16B, the light beam reflected at the low reflectivity portion becomes dark, and the intensity distribution in the light flux is reduced as shown in FIG. Will have. Further, since the detection optical system is not in focus, the signal waveform 277 of the detected slit light beam has a reduced amplitude and an increased slit width due to defocus. In such a case, the pattern on the sample surface hardly resolves, but a phenomenon occurs in which the entire slit image shifts due to the influence of the intensity distribution in the light beam, so the focus shown in FIG. 16B, a larger detection error occurs in the defocused state of FIG. When an image shift caused by such defocus occurs, it is very difficult to accurately determine the true center position of the slit light beam. As described above, the magnitude of the detection error caused by the uneven reflectance of the sample surface changes depending on the positional relationship between the focal positions of the projection optical system and the detection optical system and the sample surface. For this reason, there arises a problem that the detection accuracy varies depending on the sample surface height, and this becomes a problem when a wide height detection range is required. A similar phenomenon occurs when detecting the multi-slit light beam shown in FIG.

Next, the relationship between the sample surface height and the multi-slit light beam in the optical system shown in FIG. 10 will be described with reference to FIG. FIG. 17A shows a state where the height of the sample surface 106 is at the reference position, and FIG. 17B shows a case where the height of the sample surface changes Z. In the state of FIG. 17A, the image forming position of the slit light beam of the projection optical system 270 and the detection optical system 27
The in-focus positions of 1 coincide with each other, and the focal position (position of the intermediate image) B ′ of the central slit light beam B is on the sample surface. Focus positions A ′, B ′ of the slit light beams A, B, C,
C ′ and the imaging positions A ″, B ″, C ″ on the sensor are in an optically conjugate relationship. FIG. 17B shows only the optical paths of the slit light beams B, C for simplicity. In this case, the focal positions of the projection optical system 270 and the detection optical system 271 do not match due to a change in the optical path length.
The intermediate position between the two focal positions of 70 and the detection optical system 271 is shown by B 'and C' as the focal positions of the entire height detection optical system. The three slit light beams A, B, and C in FIG.
, The focal positions of the projection optical system 270 and the detection optical system 271 coincide with each other. As a result, the height detection error caused by uneven reflectance is reduced as compared with other slit light beams. On the other hand, FIG.
When the height displacement Z occurs as shown in FIG. 7, the focal position B ′ of the slit light beam B as the height detection optical system 272 is separated from the projection position 275 on the sample surface, and instead, the height of the slit light beam C becomes high. The focal position C ′ as the detection optical system 272 comes to coincide with the projection position on the sample surface. In this case, the image of the slit light beam C has the smallest detection error (however, since the height detection optical system 272 has an effect of defocus as a whole, FIG.
In this case, the detection error is slightly larger than the slit light beam B). As described above, the slit light beam that minimizes the detection error caused by the uneven reflectance changes depending on the height of the sample surface, so that the slit light beam with a small detection error is selectively processed according to the height of the sample surface. If the height is detected in this manner, the height can be detected relatively stably even in a wide detection range. Here, the focal position 2 in the slit light beam at which the detection error is minimized at each sample surface height
79 is as shown in FIG. That is, the focal position 279 in the slit light beam at which the detection error of the height detection optical system 272 is minimized is always constant regardless of the height of the sample surface.

Next, a method for solving such a problem will be described. A multi-slit image is obtained from the height detection device 200a including the height detection optical system 272 shown in FIG.
The above problem can be solved by processing it as described below in the height calculating means 200b. That is, as shown in FIG. 18, the slit numbers 0, ± 1, ± 2,. . , ±
s,. . , ± S, the slit number sf at which the focal position coincides with the sample surface is obtained by the following equation (34). Here, ze is the approximate height of the sample surface, p is the slit pitch projected on the sample surface, m 'is the magnification of the projection optical system 270, and this slit sf (z) is hereinafter referred to as a focusing slit ( In the case shown in FIG. 18, the focusing slit sf is the slit of the number +1).

Sf (ze) = ze · sin θ / p (sf (ze) is an integer) (Equation 34) That is, as shown in FIG. 19, the height calculating means 200 b
For example, FIG. 1 obtained from the line image sensor 214
4 from the signal waveform y (x) shown in FIG.
The approximate height ze of the sample surface 106a is calculated based on a height detection algorithm including steps S191 to S195 similar to steps S191 to S155. For example, the moving amount [Δg (i) of each slit image y (x) obtained by the method shown in FIG.
= (Yg (i) -ygo (i))], it is possible to know the approximate height ze of the sample surface 106a although there is an influence of an error. In particular, in step S155, the approximate height z of the sample surface 106a is calculated based on the above equation (32).
Calculate e. Then, step S197 shown in FIG.
In the above, based on the calculated ze, the above (Equation 3)
The in-focus slit sf (ze) is obtained from equation (4).

The magnitude of the detection error caused by uneven reflectance in each slit light beam is determined by the focusing slit sf
As the distance from (ze) increases, it continuously increases.
Therefore, in step S198 shown in FIG.
Detection value y (x) by multi-slit light beam as shown in FIG.
, A weighting function w (s) that gives a large weight to the slit light beam with a small detection error is determined, and a selective function for the slit light beam with a small detection error is determined based on the following equation (35). By performing weighting using the weight function w (s), which is a process, and performing weighted averaging, highly accurate height detection z is performed.

As the weight function w (s), for example, FIG.
As shown in FIG. 7, a target function 280 is prepared on the axis of s = 0 such that the maximum value is obtained at s = 0 and the weight decreases as the distance increases. z = [Δg (i) = (yg (s) −ygo (s))
Weighted average by weighting based on the weight function w (s-sf)] × (1 / (2 m · sin θ)) + OFFSET value z = Σ (Δg (s) · w (s-sf)) / Σ (w (s −sf)) × (1 / (2m · sinθ)) + OFFSET value (Σ is −S ≦ s ≦ S) (Expression 35) where Δg (s) = (yg (s) −ygo (s)) ygo ( s) (s =-((n-1) / 2),-((n-1) / 2) +
1, ..., 0, ..., ((n-1) / 2) -1, ((n-1)
/ 2)) indicates the position of the peak corresponding to each slit when the height z = 0. n is the total number of slit light beams, and p is the pitch of the slit light beams on the sample surface. y (x) is a detection waveform detected from the image sensor 214. y
g (s) is determined by y (x)> yth obtained in step 194.
Is the center of gravity of (y (x) -yth) for the point The threshold value yth can be obtained by the above-described equations (Equation 30) and (Equation 31) in steps S191 to S193 as in steps S151 to S153.

As described above, by performing weighted averaging with emphasis on the amount of movement of the focusing slit sf using equation (35), it is relatively stable even when the displacement of the sample surface height is large. Height detection accuracy can be obtained. At this time, as shown in step S196, in order for the focusing slit to exist in the entire height detection range, the number of slits n, the slit pitch p, and the height detection range zrange (zmax) are expressed by the following (Equation 36). Expression must be satisfied. Step S196 indicates conditions of design specifications for designing the height detection optical system 272 in advance, similarly to the case of step S156 in FIG.

(N−1) · p / sin θ ≧ zrange (Equation 36) Further, a detailed example of the method of the weighted average will be described. First, how to select the weight function w (s) will be described. As described above, in the height detection optical system 272 using the multi-slit light beam shown in FIG. 10, the manner in which errors occur in each slit light beam changes depending on the height of the sample surface 106a. Therefore, by using a multi-slit light beam and selecting only the slit light beam with the smallest detection error at each height, a relatively stable height over a wide range compared to the conventional case using only one slit light beam. Detection accuracy can be realized. In this case, since only the slit light beam with the minimum error is selected, w2 (s) 281 shown in FIG. 20 is used as the weight function w (s).
Becomes The weighting function w2 (s) 281 is 1 at s = 0, and is 0 except s = 0.

However, the weight function w as shown in FIG.
(S) By taking a weighted average of the detection values of a plurality of slits using 280, more stable and highly accurate height detection can be realized. Next, the weight function w (s) 280 in FIG.
An example in which is used will be described. The error caused by the unevenness in the reflectance of the sample surface changes due to the defocus of the height detection optical system.However, even if the defocus of the height detection optical system occurs, if the reflectance of the sample surface is uniform. No error occurs. That is, as shown in FIG. 11, an error due to uneven reflectance occurs only when slit light is projected on a boundary portion of a pattern or the like. For this reason, even if the slit is defocused, the average processing is performed using many slits as long as there is no influence from the reflectance unevenness of the sample surface, thereby reducing the influence of the error at the boundary. And the effects of errors due to other noises and the like can be reduced. If it is possible to determine whether or not the projected slit light is on the pattern boundary using the symmetry of the detected waveform, etc., highly accurate height detection can be performed by removing the detected value of that slit. 16
As shown in (b), in the state where the height detection optical system 272 is defocused, the image of the sample surface is not resolved, so it is difficult to determine the influence of the pattern boundary.
On the other hand, if weighting is performed on the slit light flux in the defocused state that causes a large error due to uneven reflectance, and the weight is increased on the slit light flux in the focused state that does not easily cause an error, and weighted averaging is performed. In addition, it is possible to prevent the influence of an increase in error due to defocusing, and to obtain an effect of reducing error by using a large number of slits. That is, by performing a weighting process (for example, a weighted average) based on the weighting function w (s) so as to conform to the imaging state of the light image of the slit-like light beam (lattice-like light beam), the detection error can be reduced. .

FIG. 21 is a diagram showing an example of a height detection result when this method is used. FIG. 21A shows a weighting function w (s−s) when the focusing slit number sf = −3, which indicates an image formation state of a light image of a slit light beam (lattice light beam).
It is the figure which showed f), and a horizontal axis | shaft shows slit number s.
FIG. 21B shows a slit movement amount [yg (s) -ygo corresponding to the height of the sample surface when there is a pattern boundary on the sample surface at the projection position of the focusing slit sf (-3).
(s)]. Since the sample surface is flat and the average slit movement amount when there is no detection error is shown as 0, the vertical axis is the detection error. Squares indicate detection errors due to each slit number. In the slit number sf projected on the pattern boundary, a relatively large detection error occurs due to the effect of the pattern despite the fact that the slit is in focus. FIG. 21C shows the slit movement amount [yg (s) -ygo (s)] corresponding to the height of the sample surface when the pattern boundary is at the projection position of the slit number +4. Since the sample surface is flat and the average slit movement amount when there is no detection error is shown as 0, the vertical axis is the detection error. Squares indicate detection errors due to each slit number. Since the pattern boundary indicated by the slit number +4 is far from the focusing slit sf and is in a defocused state, a large detection error occurs compared to FIG. 21B.

In each of the cases shown in FIGS. 21B and 21C, the slit movement amount [Δg (s) = (yg
(s) −ygo (s))] is indicated by a dashed line, and the weighting as shown in the above equation (35) is performed using the weight function w (s−sf) shown in FIG. The result of the averaging is indicated by a dashed line. In the case shown in FIG. 21B, a large weight is given to the focused slit light beam sf = −3 where an error occurs, so that the detection error in the case of the weighted average is slightly larger than the simple average. Luminous flux sf
The error is reduced from the detection error (detection error indicated by a square) when only = -3 is used. In contrast, FIG.
In the case shown in (c), a very large detection error occurs because the slit light beam at the pattern boundary is defocused, and the detection error cannot be sufficiently reduced by simple averaging. On the other hand, as a result of performing the weighted average as shown in the above equation (35) using the weight function w (s-sf) shown in FIG. 21A, the detection error becomes almost zero. As described above, by performing weighting so as to conform to the focusing slit sf which is an imaging state obtained according to the height (ze) as shown in the above equation (34),
Stable and highly accurate height detection can be realized in a wide detection range. Comparing the result of the simple averaging with the result of the height detection in the case of performing the weighted averaging according to the present embodiment, the error in the pattern boundary portion can be reduced to 50% or less in the present embodiment.

Next, the weighting function w (s) may be of various types such as a type in which several lines around the focusing slit are selected and a type in which the standard deviation of the Gaussian function is changed, in addition to the two types shown in FIG. Things are conceivable. These weights may be set so that the weights for the slit light beams that cause unacceptable errors caused by defocusing are sufficiently low. Usually, since the influence of defocus changes continuously, a more stable result can be obtained by using detection values from many slits using a continuous function such as w (s) 280. If an error caused by defocus is measured in each slit, it can be used as an index of reliability of a detection value of each slit. For example, if a standard deviation of a detection error is obtained, and its reciprocal is used as an evaluation value of reliability and weighted averaging is performed using a function that gives a weight in proportion to the reliability, a highly accurate height detection value can be obtained. Thus, an appropriate weighting function may be selected according to the specifications of the optical system and the target accuracy. At this time, if the width of the slit light beam is the same (the defocus condition is the same), as the slit pitch p is smaller, the number n of slits that can be used increases, and
A large effect of the multi-slit can be obtained. Further, as described above, by using a slit having a width as narrow as possible, it is possible to reduce errors caused by uneven reflectance.

Next, the effect on the height detection when the sample surface 106a is inclined will be described with reference to FIG. FIG. 22 is a partially enlarged view of FIG. 10, where 210 is a projection lens and 215 is a detection lens. Detection lens 215
Assuming that the conjugate plane or the in-focus plane of the image sensor 214 is 218, the shift of the projection light on the conjugate plane 218 is detected on the image sensor 214. Sample surface 10
When the height of 6a is increased by z, the detection light is reflected on the sample in a defocused state. Further, when the sample surface 106a is tilted by the angle εrad, the detection light reflected at 217 by the so-called optical lever effect is further tilted by 2εrad. Then, the position of the detection light on the conjugate plane 218 of the image sensor moves by 2εz · cos (π−2θ) / cosθ. If s sin θ is multiplied to convert this movement amount into height, the height detection value changes by −2εz / tan 2θ depending on the inclination εrad of the sample 106. On the other hand, the slit light projection position 21 caused by the inclination εrad
7, and the height displacement between the intersection 282 of the sample surface 106a and the focal plane 218 of the detection optical system is also -2z / tan2 ?.
That is, the actual height change caused by the movement and inclination of the detected image due to the optical leverage cancel each other out.
In a minute range near the area where the slit light beam is projected, the sample surface is considered to be substantially flat. Therefore, such a condition always holds, and the inclination ε of the sample surface 106a
Irrespective of the focal plane 218 of the detection optical system and the sample plane 106
The height corresponding to the intersection 282 of a is detected. this is,
The same applies to the case where height detection is performed using a multi-slit, and the movement amounts of the plurality of slits all indicate the height of the same point. Here, it is necessary to pay attention to the following points. As shown in FIG. 23, when there is a step on the sample surface between the intersection point 282 and the projection position 217 of the slit light beam, this step is added to the detected height. As described above, the variation in the moving amount of the multi-slit detected by the optical system shown in FIG. 10 is caused by the step and the detection error of each slit projection position, and is not affected by the inclination. If the height detection value by the multi-slit reflects the height of each slit light projection position, if the weighted averaging is performed, an offset occurs in the detection value due to the inclination. There is no problem because of the influence of the inclination for the reason.

As shown in FIG. 10, the projection position 217 of the slit light beam is shifted by z · tan θ from the position 110 where the height is to be detected due to the height change z of the sample surface. Since the detection value of each slit detects a step on the sample surface, a height detection error occurs due to a projection position shift. However, as can be seen from FIG. 18, since the projection position of the weighted slit (focusing slit) is always constant irrespective of the height of the sample surface, the weighted average of the previous weighted average based on the above equation (35) is obtained. By performing the processing, the correction of the projection position shift can be realized.

When the step existing on the sample surface is small with respect to the required detection accuracy, the projection position shift of the slit light beam caused by the inclination of the sample surface shown in FIG. 22 does not cause any particular problem. In such a case, the detection accuracy can be improved by removing only those with large errors from each of the movement amounts Δg (i) of the multi-slit. As described above, the variation in the amount of movement of the multi-slit is due to a step at each slit projection position and a detection error. Therefore, if the step is small enough to cause no problem, it is estimated that, among the obtained movement amounts of the multi-slit, a movement amount having a large value greatly includes a detection error at a pattern boundary. Therefore, the moving amount of each slit obtained first is simply averaged, the one with the largest difference from the average value is judged to include the error caused by uneven reflectance, and this is removed and the average is calculated again. Do. This is a constant value (threshold) where the difference between the maximum and minimum movements of the remaining slits is
By repeating the process until the following, the process excluding the slit having a large error can be performed. This threshold value may be determined in consideration of the size of the step existing on the sample surface. In this process,
Since the slit to be selected becomes indefinite, the effect of correcting the projection position deviation as in the case of performing weighted averaging cannot be obtained. In addition, when there is a large step on the sample surface or when an error due to other factors such as electric noise of the detector is large, the result may be unstable. In such a case, FIG. The method using the weighted averaging process as shown in (1) gives better results.

Next, an embodiment of the weighted averaging process in the above embodiment will be described. In order to find a focusing slit and to perform weighting at the time of weighted averaging, the height ze of the sample surface is used.
Is necessary (see the equation (34)).
Since e may be a rough value for selecting a slit, the height detection value obtained by simple averaging or the height obtained last time may be used as ze. An example of the configuration of the height detecting means for realizing these processes will be described below. If the processing time does not matter, the processing equivalent to these can be realized by software.

FIG. 24 is a diagram showing an embodiment in which the height calculating means 200b calculates the height of the sample surface by selecting a focusing slit by simple averaging of the slit movement amounts. The multi-slit image detection signal y (x) 248 detected by the line image sensor 214 is processed by the slit position detection unit 283, and the position 284 of each slit is calculated. Here, the slit position detection unit 283 is a unit that detects each slit position by image processing, and for example, performs steps S191 to S191 illustrated in FIG.
By performing the processing shown in S194, the slit movement amount y
g (i) is calculated. These calculated slit positions yg
The difference between (i) 284 and the slit position ygo (i) 285 at the previously stored reference height is obtained by the subtractor 286, and the amount of movement of each slit [Δg (i) = (yg (i) −
ygo (i))] 287. This slit movement amount 28
7 is temporarily stored in the storage unit 289, and these values are input to the weighted average calculation unit 290. Weighted average calculator 2
In 90, first, a simple average (weighted average in which all weights are equal) [ΣΔg (i) / n = Σ (yg (i) −ygo (i))
/ N]. Although the result of the simple averaging includes a detection error, the slit moving amount corresponding to the approximate value ze of the height of the sample surface 106a (for example, based on the expression (34), the focusing slit s
Calculate f (ze). ) 296 are obtained. The result of the simple averaging (for example, the focusing slit sf (ze)) is input to the selector 292, and the weighting factor selector 293 determines the weighting factor w (s-sf) 294 based on the result. The weight coefficient w (s) is stored in the table 295 in advance, and the slit movement amount (for example, the focusing slit sf (z
e)) Select an appropriate one according to 296. The selected weighting coefficient w (s-sf) 294 is calculated by
When input to 90, the weighted average calculator 290 calculates the weight coefficient w (s−sf) 294 and the stored slit movement amount [Δg (s) = (yg (s) −ygo (s))] 287. And the weighted average [Σ (Δg (s) · w (s-sf)) / Σ
(W (s-sf))], the slit movement amount (Δg) 29
7 is calculated. The result (Δg) is multiplied by [1 / (2 m · sin θ)] using a multiplier 298, and added to an offset value to convert the result into height information z 299.
0 and output to the focus control device 109.

FIG. 25 is a diagram showing an embodiment in which the height of the sample surface is calculated by performing weighting using the detected value at the previous time by the height calculating means 200b. As in FIG. 24, the shift amount of the slit [Δg (i) = (yg (i) −yg)
o (i))], but does not store the detected value, and inputs the result to the weighted average unit 290. On the other hand, the weight coefficient selection unit 293 stores the slit movement amount [Δg (s)] 301 at the previous time delayed by the delay circuit 300.
Is input, the weight coefficient w (s-sf) 294 is determined, and the weight coefficient is input to the weighted average unit 290. Thus, the weight coefficient w determined based on the slit movement amount 301 one time ago.
Using (s-sf) 294, a weighted average [Σ (Δg
(s) · w (s−sf)) / Σ (w (s−sf))],
A shift amount (Δg) 291 of the slit is detected, and the result (Δg) is calculated by using a multiplier 298 as [1 / (2 m · sin).
θ)], add an offset value, and add height information z29.
9 is output to the overall control device 120 and the focus control device 109. In other words, since it is possible to suppress a variation of about several tens μm or less in terms of the height of the sample surface between the slit movement amount before one time and the slit movement amount at the present time, the slit movement amount before one time Even if the weight coefficient w (s-sf) 294 is determined based on 301, there is no problem and a detection error of about several μm or less at the height of the sample surface can be realized.

Next, an embodiment for obtaining the two-dimensional distribution of the height of the sample 106 will be described with reference to FIG. Light source 201
Is illuminated on a mask 203 ″ on which a repetitive pattern of, for example, a rectangular pattern is drawn. This is projected at a position 217 on the sample 106 by the projection lens 210. The light is projected on the sample. The multi-slit pattern 217 is imaged by a detection lens 215 on a two-dimensional image sensor 214 ″ such as a CCD. Assuming that the magnification of the detection system is m, when the height of the sample changes by z, the slit image shifts by 2 mz · sin θ. Since this shift amount reflects the height of the point where the slit irradiates the sample,
By utilizing this, it is possible to detect the height distribution of the sample surface 106a in the irradiation range of the multi-slit.
In the embodiment shown in FIG. 26, the stop 211 is located at the front focal position of the projection lens 210 and the stop 216 is located at the detection lens 2.
It is located at the back focal point of the 15th position. This is because the lenses 210 and 215 are made telecentric on the sample side to eliminate magnification fluctuation due to up and down of the sample 106. Thus, a change in magnification due to a change in the height of the sample surface 106 can be suppressed, and the detection linearity can be improved. Also, as in the embodiment shown in FIG.
May be added before the projection lens 210 to selectively project S-polarized light. This is because when the pattern formed on an insulating film or the like is inspected based on an SEM image, since the insulating film is a transparent film, multiple reflection in the transparent film is prevented, and the reflectance between the materials is reduced. This is because the difference can be suppressed and the inspection can be performed. The position of the polarizing filter 240 does not necessarily have to be in front of the projection lens 210, and the same effect can be obtained anywhere between the light source 201 and the detector 214 ″.

Here, for simplicity, the magnification m of the detection optical system is m
= 1 will be described. When m is not 1, p may be set to m · p.

Next, a description will be given of a multi-slit moving amount detection algorithm executed by the height calculating means 200b, for an embodiment different from FIGS. FIG. 27 shows a method of detecting the phase change φ of the periodic waveform. Assuming that the pitch of the multi-slit pattern is p, the phase change φ (rad) corresponds to the amount of movement pφ / 2π, which corresponds to the height change pφ / (2πm · sinθ). This results in detecting the phase change of the periodic waveform. The height detection in the height calculating means 200b can be realized by a product-sum operation. That is,
Let the detected waveform be y (x) and the function g (x) = w (x) ex
The product sum with p (i2πx / p) may be obtained, and the resulting phase may be obtained. Here, i is an imaginary unit, w (x) is a weight function, and is an appropriate real correlation function. When the correlation function w (x) is a Gaussian function, w (x) is particularly called a Gabor filter. However, w (x) may be any function as long as the function disappears smoothly at both ends. In the above description, a complex function was used, but when expressed in real numbers, it is as follows.

That is, gr (x) = w (x) · cos (i2
πx / p) and gi (x) = w (x) · sin (i2πx /
p) with y (x), and the result is R
And I. Then, the phase of y (x) is shown below (Equation 3)
It is obtained from the relationship of the expression 7). Naturally, the phase change (Equation 3)
7) is obtained based on the equation. φ = arctan (I / R) (Equation 37) However, since this phase is folded back from −π to π,
Tracking is performed so that no jump occurs from the previous detected phase, and the phases are joined, or the approximate position of the peak is separately obtained, and the approximate value of the phase in the 2π order is obtained. The weighting function w (x) may be determined so as to give weight to the focusing slit, which is the optical image of the slit light beam (lattice light beam), similarly to the weighted average in the previous embodiment. Good. Next, another embodiment of the slit movement amount measurement algorithm executed in the height calculation means 200b will be described with reference to FIG. In the embodiments shown in FIGS. 15 and 19, the displacement of the slit image is measured using the center of gravity. However, in this method, the height is converted based on the position of the edge of the slit image. First, the approximate position of the peak of each slit (x0 + p * i) as in the embodiment shown in FIGS.
And the positions of the valleys on both sides (x0 + p * ip / 2, x0 + p * i
+ P / 2), and the amplitude [y (x0 + p * i) -y (x
0 + p * ip / 2) or y (x0 + p * i) -y (x0
+ P * i + p / 2)] to determine an appropriate threshold value yth.

Next, two points sandwiching the threshold value yth are searched for (x
_j , y _j ) and (x _{j + 1} , y _{j + 1} ). Then, the x coordinate of the point where the threshold intersects with the line segment connecting these two points is x _j + {(x
_j + 1 is represented by _{-x j) * (yth-y} j) / (y j + 1 -y j)}. This operation is performed for each of the left and right inclined portions of the slit, and the position of the intersection of the threshold value and the line segment is obtained, and the middle point is defined as the position yg (i) of the slit light beam i. All the above methods have been described on the premise that the position yg (i) of the slit light beam is determined. It is also possible to determine the height z of the sample. This produces the following effects. The disturbance of the waveform of the detected multi-slit pattern due to the reflectance distribution on the sample surface is greater when the reflectance boundary matches the peak of the multi-slit image than when it matches the valley. This is because the detected light amount distribution is determined by the product of the light amount distribution when the sample reflectance is constant and the sample reflectance. Therefore, a change in the detected light amount with respect to the same change in the reflectance is more likely to occur in a bright portion. Therefore, it is possible to detect the position of the slit image with a small error and to detect the height of the sample by finding the position of the valley where the waveform is small, regardless of the reflectivity of the sample. become. As a method of detecting the position of the valley, an algorithm for calculating the center of gravity shown in FIGS. 15 and 19 for the waveform −y (x) whose sign is inverted, and an algorithm for obtaining a point crossing the threshold value yth shown in FIG. 28 by interpolation. Etc. may be used.

Next, a description will be given of a detection error due to a measurement position shift. In the embodiment shown in FIG. 10, the center of the electron optical system visual field coincides with the height measurement position 110. However, as shown in FIG.
If there is a deviation between 10, an error will occur due to the inclination of the sample surface 106a. Therefore, an embodiment of a height detecting method capable of correcting the detection value and obtaining the correct height of the center of the field of view of the electron optical system when the amount of displacement of the measurement position is clear will be described below. As described in the embodiment shown in FIG. 22, in the method in which the slit light beam projected from the oblique direction is reflected on the sample surface 106a and the height z of the sample surface is obtained from the moving amount of the image, the inclination ε of the sample surface 106a Regardless of the height, the height detection position was the intersection 282 between the sample surface 106a and the conjugate surface 218 of the sensor. Therefore, as shown in FIG. 30, the two slit light beams A and B are projected onto the sample surface 106a using the slit 203, and each of the reflected lights is detected by each of the sensors 214a and 214b. At this time, the optical path length is changed by the glass plates 302 and 303, and the projection positions (close different positions) on the sample surface of the respective slit light beams A and B are respectively changed by the image forming position of the projection optical system 270 and the detection optical system. The focal point position of the system 271 is matched (A ′ and B ′). In this way, two heights can be detected simultaneously by one height detection optical system. Each of these sensors 214a,
Simply averaging the detected values from 214b, A 'and B'
Of the intermediate point C can be detected. FIG.
As shown in FIG. 1, the inclination ε of the sample surface 106a is obtained based on the heights zA and zB of two points detected by the optical system. Since the sample surface is considered to be flat in a narrow range around the two detection positions, the sample surface is in the vicinity, and is a clear arbitrary point based on the amount of displacement of the electron optical system visual field center 291 with respect to the height detector measurement position 110. (Indicated by a black circle in FIG. 31).

When only one slit light is used for one measurement point as in the embodiment shown in FIG. 30, as shown in FIG. 32, in addition to a linear sensor, a split sensor or a PSD is used as a detector. Can also be used. In the embodiment shown in FIG. 32, two detectors 214'a and 214'b each composed of a divided sensor having a different actual sensor position can be used to detect each of the slit light beams A and B. Becomes In the case of this embodiment, the glass plates 302, 30
3 becomes unnecessary. When a wide range of height detection is required, each of the two slit light beams A and B obtained from the slit 203 in FIG.

Although the detection accuracy is lower than that of the embodiment shown in FIG. 30, a method for correcting the detection value from two heights is shown in FIG. In FIG. 33, the light flux A to be projected is 1
As a book, a slit light beam reflected on the sample surface 106a is split by a beam splitter 305, and two linear sensors 214 "a and 214" b are arranged at different distances. At this time, the projection light forms an image on the sample surface 106a.
In the detection optical system, two sensors 214 ″ a, 21
The two sensors 21 are arranged so that the respective conjugate planes (focusing planes 1 and 2) of 4 ″ b are symmetrical about the projection position of the slit light.
4 ″ a, 214 ″ b are arranged. Then, the detection positions 1 and 2 detected by the two sensors 214 ″ a and 214 ″ b respectively correspond to the conjugate planes (focusing planes 1 and 2) and the sample plane 1
As shown in FIG. 33, the height of each of the two points is detected as shown in FIG. 33.
The focus plane 2 and the detection 2 correspond to b. ). Therefore, similarly to the embodiment of FIG. 31, these detected values can be corrected and the height of a desired position can be detected. Unlike the embodiment of FIG. 30, in the embodiment of FIG. 33, the image formation position of the projection optical system 270 does not coincide with the focal point of the detection optical system 271. Is slightly larger. When only one slit light A is used as in the embodiment of FIG. 33, a PSD, a split sensor, or the like can be used as a detector in addition to a linear sensor. The accuracy can be improved by making the slit in FIG. 33 a multi-slit and performing weighted averaging as in the previous embodiment. Also, as in the embodiment of FIG.
The two-dimensional height distribution of the sample surface 106a can be obtained using the two-dimensional sensor 214 ″.

Next, an embodiment of an electron beam type inspection apparatus using such a position correction function will be described. In the case of the appearance inspection based on the SEM images shown in FIGS. 3 and 4,
Since it is necessary to capture a two-dimensional SEM image over a wide area to some extent, the beam deflector 102 is driven to move the
It is necessary to scan in a direction substantially perpendicular to the direction of movement of the electron beam 05 and detect a two-dimensional secondary electron image signal with the secondary electron detector 104. That is, while continuously moving the stage 105 in the X direction, for example, the beam deflector 102 is driven to scan the electron beam in the Y direction substantially orthogonal to the moving direction of the stage 105, and then the stage 105 is step-moved in the Y direction. Let
Thereafter, while continuously moving the stage 105 in the X direction, the beam deflector 102 is driven to scan the electron beam in the Y direction substantially orthogonal to the moving direction of the stage 105, and the secondary electron detector 104 performs two-dimensional secondary It is necessary to detect an electronic image signal.

Also in this embodiment, a correct inspection result is obtained by performing automatic focus control by obtaining the height of the surface of the inspection object 106 for which the secondary electron image signal is always detected by the height detection device 200. There is a need.

However, the height detecting optical device 200a
, The focus control is delayed due to the image accumulation time of the image sensor 214, the calculation time of the height calculating means 200b, the responsiveness of the focus position control device 109, and the like.
Therefore, even if a delay occurs in the focus control, it is necessary to accurately focus on the surface of the inspection object 106 for detecting the secondary electron image signal. Since the stage 105 moves during the time delay of the focus control, the same positional deviation as in the example shown in FIG. 29 occurs. Therefore, Figure 3
The position shift due to the focus control time delay may be corrected by the position shift correction using the height detection device shown in FIG. For example, in FIG. 30, the stage 105 is continuously moved from right to left. In this case, as shown in FIG. 34, the height calculation means 200b calculates the height slightly to the right of the upper observation system visual field center 110 in consideration of the delay time, and based on the calculated height. The focus control may be performed by controlling the focus control current or the focus control voltage to the objective lens 103 by the focus control device 109. The required shift amount of the detection position is the product VT of the delay time T and the scanning speed (moving speed) V of the stage 105. This deviation amount VT
May be corrected by the method shown in FIG. By such means, even if the focus control is delayed, the height calculating means 200b calculates the height of the surface of the inspection object 106 for detecting the secondary electron image signal. The focus control current or the focus control voltage to the objective lens 103 can be controlled by 109 to accurately focus on the surface of the inspection target 106 for detecting the secondary electron image signal. At this time, even if the detection position shift caused by the time delay occurs, as shown in FIG.
The sensors 214a to 214 "a and 214b to 214" make the detection position 341 fall between two detection positions (the detection position 342 of the slit A and the detection position 343 of the slit B).
If the detection position interval of b is set, the positional deviation correction is always interpolation of two points, and stable detection is possible.

In the embodiment shown in FIG. 34, the detection time delay correction method on the premise that the scanning direction of the stage 2 and the projection-detection direction of the multi-slit are almost parallel has been described. A method for correcting a detection time delay that can be used regardless of the projection-detection direction will be described. The line image sensors 214, 214 "
Since the image signal accumulated during a certain time T1 is output,
It can be considered that an average image in the period of T1 is obtained. That is, the data obtained from the line image sensors 214, 214 "has a time delay of T1 / 2. Further, in order for the image to be processed by the height calculating means 200b including a computer, a certain time T2 is required. Required, in total,
The height detection value indicates past information only for the time of (T1 / 2) + T2. The height calculation means 200b
As shown in FIG. 35, detection values obtained at regular intervals are represented by Z
_{-m, Z - (m-1} ), ... Z -2, Z -1, When Z _0, it is possible to estimate the current height Z _c from these data. For example, as shown in FIG. 35, it can be obtained by the following (Equation 38) by extrapolating the latest detection value Z ₀ and the detection value of the immediately preceding eye with a straight line.

[0103] _{Z c = Z 0 + ((} Z 0) - (Z -1)) × ((T 1/2) + T 2) / T 1 ( number 38) Of course, extrapolation straight line 3 months or more Points Z _-m , Z- _(m-1) , ...
The error may be determined by making the error small with respect to Z _-2 , Z _-1 , and Z ₀ , or a quadratic function for these points,
It may be obtained by applying a cubic function or the like. These extrapolation methods are mathematically well known, and an appropriate method may be selected according to the magnitude of the change in the height detection value and the magnitude of the variation. As another embodiment, a case where a height detection value is interpolated and output will be described. When the height detection value changes stepwise at the interval T1, the feedback of the electron beam is performed using the detected value, and the quality of the electron beam image becomes equal to the interval T1.
It may not be preferable because it changes suddenly at 1. In this case, it is better to be gradually changed in a short time interval by interpolating the Z _c. In this case, in addition to the extrapolation height detection value Z _c,
Further, a height extrapolation interpolated value Z _c ′ after the time a by T1 is similarly obtained. In the embodiment shown in FIG.
9) Determined by the equation.

Z _c = (Z ₋₁ ) + (((Z ₋₁ ) − (Z ₋₃ )) / (2T ₁ )) × 2.5T ₁ Z _c ′ = (Z ₀ ) + (((Z ₀ ) − (Z ₋₂ )) / (2T ₁ )) × 2.5T ₁ (Expression 39) Using these Z _c and Z _c ′, the height Z ₁ after t from time a is interpolated. And can be obtained by the following equation (40). Z ₁ = by _{Z c + (Z c '-Z} c) t / T 1 ( number 40) or to correct the detection time delay caused by the CCD storage time and height calculation time, the height of the inspected object 106 Even when changes every moment, a height detection value with a small error is obtained,
Feedback can be applied to an electron optical system that stably controls an electron beam.

Next, a method for obtaining the height distribution of the sample surface in advance will be described. In the embodiments described above, the height of the sample surface is detected at the time of inspection or measurement, and focusing and deflection correction of the electron beam are performed using the information. However, if the height cannot be detected fast enough or the time delay cannot be corrected sufficiently, the height distribution of the sample surface is checked and stored in advance, and the focus is determined based on the information. Adjustments can also be made. In the electron beam inspection apparatus shown in the embodiment of FIG. 2, the stage control device 1 is controlled based on a command from the overall control unit 120 before the start of the inspection.
The sample is moved by the stage 105 by controlling 26. At each point that has moved, the height of the sample surface is detected by the height detection device 200, and the height detection result is displayed together with the stage position obtained by the laser length measurement system 107 on the overall control system 120.
Via the storage device 142. At this time, the height detection need not be performed at all points on the sample surface, but may be measured at appropriate intervals. Depending on the shape and holding method of the target sample, if the height change is relatively gradual, detect the height at appropriate intervals according to the cycle of the height change and interpolate them. Can obtain the height distribution of the entire sample surface. During the inspection, the sample position can always be known by the laser measuring system 107. Therefore, the height of the position is input to the focus control device 109 from the height distribution recorded in advance by the overall control unit 120. Thus, the focus adjustment and the deflection correction of the electron beam can be performed in the same manner as in the case where the height detection is performed simultaneously with the inspection.

The focus adjustment method in which height detection information is obtained in advance can be applied to a case where height detection means cannot be provided as in the embodiment shown in FIG. For example, if the optimum focus control current or focus control voltage of the electron optical system can be obtained for each sample surface height instead of the height distribution, these values can be interpolated in the same manner as the height information to obtain the same. The effect of can be obtained. At each point, a focus control current or a focus control voltage at which a secondary electron image signal (SEM image signal) as a charged particle beam image detected by the secondary electron detector 104 as a charged particle detector becomes the sharpest is controlled. It is obtained from the device 109 and measured. At this time, the sharpness of the secondary electron image (SEM image), which is a charged particle beam image, is detected by the secondary electron detector 104 and converted into a digital SEM image signal or a digital SEM image signal converted by the A / D converter 39 (122). Digital SEM preprocessed by preprocessing circuit 40
The image signal is input to the overall control unit 120 and the display unit 14
3 or stored in the image memory 47 and displayed on the display means 5.
The SEM image displayed at 0 is visually determined, or is determined by image processing or the like for obtaining a change rate of an image at an edge portion in the SEM image input to the overall control unit 120. Focus control can also be performed by interpolating the optimum control current or control voltage value obtained in this manner. However, a calibration pattern 130 as shown in an embodiment shown in FIG. When used in the apparatus, the nonlinearity between the sample surface height and the control current can be corrected, and the height information of the sample surface can be obtained without a height detector.

FIG. 37 shows an embodiment in which height detection and interpolation are performed on a sample whose surface height changes gradually. FIG. 37A shows a cross-sectional shape of the sample 106. In addition to the undulation of the entire sample, there is a step 307 due to the pattern 306 on the surface. FIGS. 37 (b) and (d) show the height measurement points 308 in the cross-sectional view of FIG. 37 (a), and FIGS. 37 (c) and (e) show these detection results as 1 The result 309 is the result of the next interpolation, and the sectional shape 310 is indicated by a dotted line. In FIGS. 37 (b) and (c),
Since the height is always detected only above the step, the interpolation result is good. However, FIGS. 37 (d) and (e)
In this case, since the height measurement position includes both the top and bottom of the step, the cross-sectional shape cannot be obtained well. For this reason, it can be seen that, when obtaining the height distribution of the sample surface by interpolation of the detected height values, it is necessary to select and detect points having the same step on the surface. Note that FIG.
Although the result of the primary interpolation is shown for simplicity, a smoother and more accurate result can be obtained by performing a higher-order interpolation. As shown in FIG. 5, a large number of chips 3a are arranged on the surface of a semiconductor wafer targeted by the electron beam inspection apparatus of the present invention. The chip 3a includes a memory mat section 3c and a peripheral circuit section 3b, and the two areas usually have different surface heights. Therefore, the height measurement position is determined according to the layout of the sample surface. In order to select a point that is the same area in all chips, the interval between the selected points may be equal to the interval between the chips 3a or an integer multiple thereof.

However, since the interval between the chips of the semiconductor device is usually about several tens of millimeters, if the height changes rapidly, it is necessary to detect the height at a finer interval. In this case, as shown in FIG. 38, for example, the height may be detected at the same pitch but at a different phase, with the interval between the chips 3a as the basic pitch. At this time, to select only the upper part of the pattern, the phase of the height measurement point may be determined as follows. As shown in FIG. 38, first, when there are a plurality of memory mat units in a chip, these are sequentially stored in a memory mat unit 1, a memory mat unit 2,. . . , A memory mat portion i, their width is wi, and the distance between the memory mat portions w1 and wi is d1i. With the center of the memory mat section 1 as a reference position, the basic pitch of the measurement points is equal to the chip interval, and this is pc, the phase ph
Is given by the following (Equation 41).

Ph = {x | d1i−wi / 2 <x <d1i + wi / 2 |} (i = 1, 2,..., Mm) (Equation 41) where mm is the chip interval pc. This is the number of memory mats in the chip at that time. If measurement points are selected at appropriate intervals within the range of (Equation 41), a point in the shape of the memory mat can always be selected as a measurement point. In general, an apparatus for inspecting the appearance of a semiconductor device performs inspection using a condition file in which inspection conditions are recorded for each product or process of a wafer to be inspected. This inspection condition file contains information such as the chip of the wafer to be inspected and the layout of the memory mat portion in the chip. By using such information, the range of the pitch and phase of the height detection position can be obtained. Can be determined. At the start of the inspection, the sample is positioned using the electron beam image or the image of the optical microscope 118 if the configuration shown in FIG. 7 is used. Therefore, it is necessary to determine which position on the sample corresponds to the position of the inspection condition file. You can also know. After the positioning of the sample is completed, the height detection position is selected according to the state of the sample surface, so that the height distribution of the sample surface can be obtained stably. In FIG. 38,
Although only one direction and the x direction are shown, the height measurement position may be selected in the y direction in the same manner.

If the height is detected only at the top of the step as shown in FIG. 39, the focus will not be adjusted when observing the bottom of the step. It is sufficient if the step is of such a degree that there is no problem with respect to the depth of focus of the electron microscope. Therefore, as shown in FIG.
By shifting the phase, the height of the point 311 at the bottom of the pattern is also detected at the same time. Next, among the obtained height detection values, interpolation is performed only on the upper part of the pattern to obtain the height distribution of the entire sample surface. By obtaining a height estimation value at a point 311 corresponding to the bottom of the pattern from the height distribution obtained by interpolation, and comparing this with a value actually measured by the height detection device 200,
The step due to the pattern can be calculated. When the steps are uniform over the entire sample surface, the steps can be obtained by averaging these steps. By using this step information, only when observing the bottom of the pattern, focusing may be performed by subtracting the step from the height distribution of the upper part of the pattern as the sample surface height. When the step changes over the entire surface of the sample, a height distribution may be obtained by using only the detected height at the bottom in the same manner as at the top of the step, and these may be used in combination. The selection of the height detection position at the pattern bottom may be determined by the above (Equation 41) using the width of the pattern bottom and the position with respect to the reference.

The height distribution information obtained in this manner can be notified to the operator by using the display means 143. For example, as shown in FIG. 40, contour display 312 of the height distribution of the sample surface is performed, or an arbitrary cross section is designated,
It is possible to display the cross-sectional shape 313 and the maximum-minimum 314 of the displacement. Further, when an unexpected state such as a very large displacement or inclination of the sample surface exceeding the design specification of the inspection apparatus is detected, a function of displaying a warning to that effect can be provided. In particular, when the sample is held by a method that does not change the shape, information on the shape of the sample can be given to the operator by such a function.

Next, the observation SE including the appearance inspection SEM device shown in FIG. 2, FIG. 3, FIG. 4, or FIG.
A description will be given of an embodiment relating to a method of calibrating a focus control current or a focus control voltage of a charged particle optical system (objective lens 103) and a focus position in an M device, a SEM device for length measurement, and the like. When the relationship between the control current and the focal position is non-linear, it is necessary to correct the non-linearity. A method for evaluating the linearity and determining the correction value will be described. A calibration standard pattern 130 as shown in FIG. 42 is fixedly arranged on a sample stage on a stage 105 holding an object to be inspected 106 as shown in FIG. Standard pattern for calibration 130
Is made of a conductive material so as not to be charged by scanning of the electron beam 112 which is a charged particle beam.

At the time of calibration, the stage control device 126 is controlled based on a command from the overall control unit 120 to move the calibration standard pattern 130 to an observation area centered on the optical axis 110 of the upper observation system. The overall control unit 120 includes:
Using this standard pattern 130, a focus at which a secondary electron image signal (SEM image signal) as a charged particle beam image detected by the secondary electron detector 104 as a charged particle detector at each point is the sharpest. The control current or the focus control voltage is obtained from the focus control device 109 and measured. At this time,
The definition of the secondary electron image (SEM image), which is a charged particle beam image, is detected by the secondary electron detector 104, and the A / D
The digital SEM image signal converted by the converter 39 (122) or the digital SE preprocessed by the preprocessing circuit 40
The M image signal is input to the overall control unit 120 and the display unit 1
43, or stored in the image memory 47, displayed on the display means 50, and visually determined, or determined by image processing or the like for obtaining a change rate of an image at an edge portion in the SEM image input to the overall control unit 120. . By the way, since the actual height dimension of the calibration sample surface (calibration standard pattern 130) is known, by inputting this height information using input means (not shown), the overall control unit 120 The relationship between the actual height of the sample surface and the optimum focus control current or focus control voltage can be obtained by the above measurement as shown in FIG. At the same time, the height of the standard pattern 130 for calibration is measured by the height detecting optical device 200a and the height calculating means 200b, so that the overall control unit 12
0 is a calibration curve representing the relationship between the actual height of the sample surface and the height detection value measured by the height detection optical device 200a and the height calculation means 200b as shown in FIG. 43 (b). From these two calibration curves, the overall control unit 120
Is a height detecting optical device 200a and a height calculating means 20
The optimum focus control current or focus control voltage value for capturing a focused charged particle beam image can be found from the detection value of 0b. In addition, the overall control unit 120 uses two sets of calibration curves of the sample surface height and the detection value obtained by the height detecting optical device 200a, the sample surface actual height, and the focus control current or the focus control voltage. Alternatively, a calibration curve between the value detected by the height detecting optical device 200a or the like shown in FIG. 43C and the focus control current or the focus control voltage may be directly obtained. In this case, it is not necessary to know the actual height dimension of the standard pattern 130 for calibration.

That is, as shown in FIG. 45, calibration is performed using the calibration standard pattern 130. In step S30, calibration is started. In step S31,
The overall control unit 120 commands the stage control unit 126 to bring the position n of the calibration sample 130 to the optical axis 110 of the electron optical system. Next, step S32 and steps S33 to S38 are executed in parallel. In step S32, the overall control unit 120 issues a height detection command to the height calculation means 200b to obtain an uncorrected height detection Zdn.
At the same time, in step S33, the overall control unit 120
Sends an instruction to the focus control device 109 to adjust the focus control signal of the electron optical system (objective lens 103) to Ii. Next, in step S34, the overall control unit 120 issues an instruction to the deflection control device 108 to scan the electron beam one-dimensionally or two-dimensionally. Next, in step S35, the overall control unit 120 instructs the image processing unit 124 to process the acquired SEM image to obtain the image sharpness Si. Next, in step S36, the focus control signal Ii of the electron optical system (objective lens 103) is changed to i = i + 1, and until i ≦ Nn in step S37.
Steps S33 to S35 are repeated to determine the sharpness Si of the image in each focus control signal Ii. Next, if i ≦ Nn is NO in step S37, in step S38, the overall control unit 120 obtains the focus control signal In that maximizes the sharpness Si of the image.

Next, in step S39, the overall control unit 120 instructs the image processing unit 124 to perform image distortion correction parameters such as image magnification correction and image rotation correction at each height Zn on the calibration sample 130. Is obtained and stored in the storage unit 142. Next, in step S40, the position n on the sample piece 130 is changed to n = n + 1, and in steps S31 to S31 until n ≦ Nn in step S41.
By repeating S39, the focus control signal In that maximizes the sharpness of the image at the height Zdn of each sample piece and the image distortion correction parameters including the image magnification correction and the image rotation correction are obtained. Next, when n ≦ Nn is NO in step S41, in step S42, the overall control unit 120
The calibration curve shown in FIG. 43C is obtained from the uncorrected height detection value Zdn and the focus control signal In at which the sharpness of the image at the height Zdn of each sample piece is maximized, or the sample piece 13 is obtained.
If the actual height dimension Zn at each position n of 0 is known, Zdn,
Calibration curves shown in FIGS. 43A and 43B are obtained from Zn and In. Then, in step S43, the overall control unit 12
0 is a parameter of the calibration curve (for example, a coefficient of a polynomial approximation), which is stored in the storage unit 142 and terminated (S4).
4). The calibration standard pattern 130 shown in FIG. 42 has flat ends at both ends, and by performing calibration at these two locations, it is also possible to calibrate gain and offset. This calibration standard pattern 130
Is effective in that although the shape of the calibration curve is stable, quick calibration can be performed when only the gain and offset drift. When the shape of the calibration curve is very stable and can be calibrated by another method, as shown in FIG. 44 (a), the optical height detection optical device 200a and the objective lens 103 are provided in a standard pattern having one step. Calibration of the gain and offset between the current and the control current may be performed. If the calibration curve has a simple shape that can be approximated by a quadratic function, a standard pattern having two steps may be used as shown in FIG. 44 (b). 42 and 44 when the particle beam device has a Z stage.
The height detection optical device 200 is obtained by detecting the height by moving the Z stage and evaluating the image by using only a normal pattern having no steps, instead of the standard pattern shown in FIG.
The calibration of the control current to the objective lens 103 can be performed. In this case, the focus can be adjusted by the Z stage. However, if the response speed of the stage is not sufficient to change the observation position, the stage is fixed and the focus is adjusted by the control current of the objective lens 103. It is also possible to do. Calibration using the calibration parameters determined as described above and the appearance inspection based on the SEM image in the SEM device shown in FIG. 2 or FIG. 3 will be described based on the processing flow shown in FIG.
That is, it is started in step S70. Next, in step S71, the overall control unit 120 takes out the calibration parameters from the storage unit 142, loads the height detection unit calibration parameters into the height calculation unit 200b, and sends the height-focus control signal to the focus control unit 109. The calibration parameters are loaded, and the image distortion correction parameters such as the image magnification correction are loaded into the deflection control device 108.

Next, in step S72, the overall control unit 120 instructs the stage control unit 126 to move the stage to the stage scanning start position. Next, steps S73, S74, S75, and S
And 76 are executed in parallel. In step S37, the overall control unit 120 instructs the stage control device 126,
The stage control device 126 controls the stage 2 on which the inspection object 106 is mounted at a constant speed. At the same time, in step S74, the overall control unit 120 instructs the height calculating unit 200b to perform real-time height detection and height detection device calibration parameters obtained from the height detecting optical device 200a by the height calculating unit 200b. The information 190 on the correction detection height based on the focus control device 109 and the deflection control device 1
08. At the same time, in step S75, the overall control unit 120 instructs the focus control device 108 and the deflection control device 109,
And deflection control device 109, respectively, for electron beam scanning and focus control based on a height-focus control signal calibration parameter based on the correction detection height, and deflection distortion correction using image distortion correction parameters such as image magnification correction based on the correction detection height. Is performed continuously. At the same time, in step S76, the overall control unit 120 instructs the image processing unit 124 to perform SEM images continuously obtained by the image processing unit 124 to perform an appearance inspection. Next, in step S77,
At the stage scanning end position, the overall control unit 120 displays the inspection result received from the image processing unit 124 on the display unit 143.
, Or stored in the storage unit 142. Next, if the inspection is not completed in step S78, the process returns to step S72. Step S78
, The inspection ends when the inspection ends (step S7).
9).

In the embodiment described above, the SEM device (electron beam device) has been described. However, the present invention can be applied to other focused charged beam devices such as a focused ion beam device. In that case, the electron gun 101 may be replaced with an ion source. In this case, the secondary electron detector 104 is not always necessary, but a secondary electron detector or a secondary ion detector may be provided at the position of 104 in order to monitor the processing state by the ion beam. Further, the present invention can be applied to a broadly-defined processing apparatus including a drawing apparatus using an electron beam. In this case, the secondary electron detector 104 is not necessarily required, but it is desirable to use the secondary electron detector 104 in the same manner for monitoring the processing state and aligning the sample. Also, in an optical device such as a normal optical microscope, an optical appearance inspection device, and a light exposure device, if there is a mechanism for controlling the focal position, it is similarly possible to configure an automatic focusing mechanism using the present height detection device. it is obvious. In the case of a device that changes the focal position of the optical system instead of moving the sample up and down for focusing, the effect of the characteristic that this height detection device can perform high-precision height detection over a wide range is particularly effective. growing. FIG.
FIG. 7 is a diagram showing an embodiment in this case. Only different points from FIG. 2 will be described. Reference numeral 191 denotes a light source of the optical device, and illumination light is applied to the sample 106 through a lens 196, a half mirror 195, and an objective lens 193. This image passes through the objective lens 193, is reflected by the half mirror 195,
An image is formed on the image detector 194 via the lens 197.
At this time, it is necessary to focus the objective lens 193 on the surface of the sample 106. At this time, if the height detector 200 is provided, high-speed focusing can be realized. In the embodiment shown in this figure, focusing is performed by moving the objective lens 193 up and down, but the stage 105 may be moved up and down instead. However, when the objective lens 193 is moved up and down, the effect of the characteristic of obtaining high accuracy over a wide measurement range of the height detector 200 can be more exhibited. Alternatively, 191, 193, 195, 196, 197, 194
Of course, focusing may be performed by raising and lowering the entire optical system composed of. An optical appearance inspection apparatus may be formed by adding the image processing means 124 shown in FIGS. 2 and 3 to the structure shown in FIG. Further, it goes without saying that the laser processing machine may be configured using the configuration of the embodiment shown in FIG.

[0118]

According to the present invention, the quality of an electron beam image (SEM image) can be improved by reducing image distortion due to deflection or aberration of an electron optical system or a decrease in resolution due to defocus. As a result, there is an effect that inspection and length measurement based on an electron beam image (SEM image) can be executed with high accuracy and high reliability. Further, according to the present invention, the height information of the surface of the inspection object and the focus control current or focus control voltage of the electron optical system detected by the optical height detection device, and image distortion such as an image magnification error. By obtaining the calibration parameters during the period, the sharpest electron beam image (SEM image) without image distortion is obtained from the inspection object, and inspection and length measurement based on the electron beam image (SEM image) can be performed. , With high accuracy,
In addition, an effect that can be executed with high reliability is obtained. Further, according to the present invention, in the electron beam type inspection apparatus, the detection of the surface height of the inspection object and the control of the electron optical system can be executed in real time, so that the image distortion due to the continuous stage movement can be reduced. Inspection can be performed by obtaining a high-resolution electron beam image (SEM image), which can improve the inspection performance and its stability, and shorten the inspection time. In particular, shortening the inspection time is effective for increasing the diameter of the object to be inspected, such as a semiconductor wafer.
Further, according to the present invention, a similar effect can be obtained in an observation processing apparatus using a convergent charged particle beam.

[Brief description of the drawings]

FIG. 1 is a diagram for explaining the necessity of focusing on an object to be inspected such as a semiconductor wafer in an electron beam image inspection according to the present invention.

FIG. 2 is a schematic configuration diagram showing an embodiment of an electron beam apparatus (SEM apparatus) according to the present invention.

FIG. 3 is a schematic configuration diagram showing an embodiment of an electron beam inspection apparatus (SEM inspection apparatus) according to the present invention.

FIG. 4 is a specific configuration diagram showing an embodiment of an electron beam inspection apparatus (SEM inspection apparatus) according to the present invention.

FIG. 5 is a view showing a semiconductor wafer on which a semiconductor memory according to the present invention is formed.

FIG. 6 is a diagram showing a detected image f1 (x, y) and a comparative image g1 (x, y) to be compared and inspected in the electron beam inspection apparatus (SEM inspection apparatus) according to the present invention.

FIG. 7 is a specific configuration diagram showing another embodiment of the electron beam inspection apparatus (SEM inspection apparatus) according to the present invention.

FIG. 8 is a diagram specifically showing the pre-processing circuit shown in FIGS. 4 and 7;

FIG. 9 is a diagram for explaining contents to be corrected by the pre-processing circuit shown in FIG. 8;

FIG. 10 is a diagram showing a first embodiment of a height detecting optical device in the height detecting device according to the present invention.

FIG. 11 is a diagram for explaining a principle of reducing a detection error by a multi-slit.

FIG. 12 is a diagram for explaining a detection error caused by multiple reflection by a transparent film such as an insulating film existing on a semiconductor wafer or the like.

FIG. 13 is a diagram showing a change in reflectance with respect to an incident angle of silicon and a resist (a transparent film such as an insulating film) existing on a semiconductor wafer or the like.

FIG. 14 is a diagram for explaining an embodiment of a height detection algorithm processed by height calculation means in the height detection device according to the present invention.

FIG. 15 is a process flowchart illustrating an embodiment of a height detection algorithm processed by a height calculation unit in the height detection device according to the present invention.

FIG. 16 is a diagram for explaining an error caused by defocus of a height detection optical system and uneven reflectance on a sample surface in the height detection device according to the present invention.

FIG. 17 is a diagram for explaining the relationship between the multi-slit and the focal position of the height detection optical system in the height detection device according to the present invention.

FIG. 18 is a view showing a slit light beam with a minimum detection error in the height detecting optical device according to the embodiment of the present invention.

FIG. 19 is a process flowchart for explaining a weighting process based on a weight function, which is another embodiment of the height detection algorithm processed by the height calculation means in the height detection device according to the present invention.

FIG. 20 is a diagram illustrating an example of a weighting function when a weighted average of the detection values of each slit is performed in the height detection device according to the present invention.

FIG. 21 is a diagram for explaining an effect of reducing a height detection error by performing a weighted average according to a sample surface height in the height detection device according to the present invention.

FIG. 22 is a diagram for explaining that there is no detection error due to the inclination of the sample in the embodiment of the height detecting optical device according to the present invention.

FIG. 23 is a diagram for explaining the influence of a sample surface step at a slit light projection position in an embodiment of the height detecting optical device according to the present invention.

FIG. 24 is a diagram specifically showing a height detection processing unit based on a weighted average of multiple slits in the height detection optical device according to the present invention.

FIG. 25 is a diagram specifically illustrating a height detection processing unit based on a weighted average of a multi-slit using a detection value at a previous time in the height detection optical device according to the present invention.

FIG. 26 is a diagram showing a configuration example of measuring a height distribution on a surface by the height detecting optical device according to the present invention.

FIG. 27 is a view for explaining an embodiment in which the position of a multi-slit pattern is detected by a Gabor filter, which is a height detection algorithm processed by a height calculation means in the height detection device according to the present invention.

FIG. 28 is a diagram showing an embodiment for measuring a slit edge position, which is a height detection algorithm processed by a height calculation means in the height detection device according to the present invention.

FIG. 29 is a diagram for explaining a height detection error caused by a measurement position shift in the height detection optical device according to the present invention.

FIG. 30 is another embodiment of the height detecting optical device according to the present invention, and is a view showing a configuration of a height detecting optical device capable of detecting two heights at the same time.

FIG. 31 is a diagram for explaining a method of correcting a measurement position shift using two height detection values in the height detection device according to the present invention.

FIG. 32 is an embodiment of the height detecting optical device according to the present invention, and is a view showing another configuration of the height detecting optical device capable of detecting two heights at the same time.

FIG. 33 is a view showing another embodiment of the height detecting optical device according to the present invention, and showing a configuration of the height detecting optical device capable of detecting two heights at the same time.

FIG. 34 is a diagram illustrating a method of correcting a measurement position shift caused by a delay in height detection time using two height detection values in the height detection device according to the present invention.

FIG. 35 is a diagram for explaining a detection time delay correction method that can be used in the height detection device according to the present invention regardless of the scanning direction of the stage and the projection-detection direction of the multi-slit.

FIG. 36 is a diagram for explaining a detection time delay correction method that can be used in the height detection device according to the present invention regardless of the scanning direction of the stage and the projection-detection direction of the multi-slit.

FIG. 37 is a diagram for explaining the relationship between the height measurement position and the interpolation result of the detected value in the method for detecting the height of the sample surface according to the present invention.

FIG. 38 is a view for explaining a method of selecting only the upper part of the pattern as a measurement point in the method for detecting the height of the sample surface according to the present invention.

FIG. 39 is a diagram for explaining a method of selecting the top and bottom of a pattern as measurement points in the method for detecting the height of a sample surface according to the present invention.

FIG. 40 is a diagram showing an example of information on the height distribution of the sample surface shown to the operator in the automatic inspection device according to the present invention.

FIG. 41 is a diagram showing an electron beam apparatus or the like in which a standard pattern for calibration is arranged on an XY stage.

FIG. 42 is a perspective view showing an example of a calibration standard pattern having an inclined portion.

FIG. 43 is a diagram for explaining a calibration curve obtained by using a standard pattern for calibration in an electron beam apparatus or the like according to the present invention.

FIG. 44 is a perspective view showing another embodiment of the standard pattern for calibration.

FIG. 45 is a diagram showing a processing flow for obtaining parameters for calibration.

FIG. 46 is a view showing a schematic flow of performing inspection while correcting using a calibration parameter by driving a stage at a constant speed in the electron beam inspection apparatus according to the present invention.

FIG. 47 is a schematic configuration diagram showing an embodiment of an optical appearance inspection apparatus according to the present invention.

[Explanation of symbols]

8 ground electrode, 24 transmission means, 32 condenser lens, 37 retarding electrode, 38 ground electrode, 40 preprocessing circuit, 100 electron beam device (inspection room), 101 (31) electron source, 102 (34): deflection element (scanning deflector); 103 (33): objective lens, 1
04 secondary electron detector, 2, 105 stage, 106
… Sample (object to be inspected), 107… Laser measuring system, 10
8 Deflection control device, 109 Focus control device, 110 Optical axis of upper observation system, 115 Detection unit, 116 Electron optical system, 117 Electron detection unit, 118 Grid, 119
Sample chamber, 120: overall control unit, 120 ′: correction control circuit, 121: source potential adjusting means, 122 (39): A /
D converter 123, image memory 124 image processing means (image processing circuit) 125 sample stage potential adjusting means 12
Reference numeral 6: stage control device, 127: grid potential adjusting means, 130: standard pattern (calibration sample), 142: storage device, 143: display means, 200: height detecting device, 2
00a: height detecting optical device, 200b: height calculating means,
201: light source, 202: condenser lens, 203: multi-slit mask (205: half mirror, 206)
... reflecting mirror, 210 ... projection lens, 211 ... projection stop, 2
13 ... cylindrical lens, 214 ... line image sensor, 214 "... two-dimensional sensor, 214a, 214"
a, 214b, 214 "b ... linear sensor (detector),
214'a, 214'b: split sensor, 215: detection lens (imaging lens), 216: detection aperture, 217: slit light projection position, 240: polarization filter, 280: weight function w (s), 281 ... Weight function w2 (s), 30
2, 303: glass plate, 305: beam splitter,
306: pattern portion, 307: step, 308: height measurement point, 309: height measurement result, 310: sample surface cross-sectional shape,
311: Detection position of step bottom 312: Height distribution contour display, 313: Cross-section shape display, 314: Cross-section maximum value-minimum value display

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Yasutoshi Usami 882 Ma, Hitachinaka-shi, Ibaraki Pref.Hitachi, Ltd.Measurement Division, Hitachi, Ltd. Measuring Instruments Division

Claims (25)

[Claims] An electron beam source, a deflecting element for deflecting an electron beam emitted from the electron beam source, and an objective lens for converging and irradiating an electron beam deflected by the deflecting element onto an object to be inspected. An electron optical system having the same; an electron beam image detecting optical system for detecting a secondary electron beam image generated from the inspection object by an electron beam deflected, focused, and irradiated by the electron optical system; A projection optical system for projecting a grid-like light beam onto the object from obliquely above and an image of the grid-like light beam reflected from the surface of the inspection object by the grid-like light beam projected by the projection optical system. A detection optical system that receives a lattice-shaped optical image corresponding to an image formation state, converts the light into a lattice-like signal, and detects the image, and forms an image with a small detection error based on the lattice-like signal detected by the detection optical system. The state is searched, and the detected imaging state with a small detection error is detected. Weighting process is performed on the lattice-like signal detected by the detection optical system to obtain the amount of movement or phase change with respect to the reference signal corresponding to the reference height as the lattice-like signal. An optical height detection device having calculation means for obtaining information corresponding to the height of the surface of the inspection object; and an optical height detection device corresponding to the height of the surface of the inspection object obtained by the optical height detection device. A focus control unit that controls a current or a voltage to be applied to an objective lens of the electron optical system based on information to focus an electron beam on an object to be inspected in a focused state; and the electron beam image detection optical system. An electron beam inspection or measurement device comprising: an image processing means for inspecting or measuring a pattern formed on an inspection object based on a detected secondary electron beam image. 2. An electron beam source, a deflecting element for deflecting an electron beam emitted from the electron beam source, and an objective lens for converging and irradiating the electron beam deflected by the deflecting element onto an object to be inspected. An electron optical system having the same; an electron beam image detecting optical system for detecting a secondary electron beam image generated from the inspection object by an electron beam deflected, focused, and irradiated by the electron optical system; A projection optical system for projecting a grid-like light beam onto the object from obliquely above and an image of the grid-like light beam reflected from the surface of the inspection object by the grid-like light beam projected by the projection optical system. A detection optical system that receives a lattice-shaped optical image corresponding to an imaging state, converts the light into a lattice-like signal, and detects the lattice-like signal. By performing various selective processing An optical height detecting device comprising: calculating means for obtaining an amount of movement or a phase change with respect to the quasi-signal to obtain information corresponding to the height of the surface of the inspection object; and The electron beam is focused on the inspected object by controlling the current or the voltage applied to the objective lens of the electron optical system based on the information corresponding to the height of the surface of the inspected object. A focus control unit for performing the inspection, and an image processing unit for inspecting or measuring a pattern formed on the inspection object based on the secondary electron beam image detected by the electron beam image detection optical system. Characteristic electron beam inspection or measurement equipment. 3. The electron beam type inspection or measurement apparatus according to claim 2, wherein the selective processing in the calculation means of the optical height detection apparatus is a weighting processing. 4. The method according to claim 1, further comprising: correcting a deflection amount of the electron optical system to a deflection element based on information corresponding to a surface height of the inspection object obtained by the optical height detection device. 2. A deflection control means for calibrating image distortion including magnification error of an electron beam image generated based on focus control.
Or the electron beam type inspection or measurement device according to 2. 5. An electron beam source, a deflecting element for deflecting the electron beam emitted from the electron beam source, and an objective lens for converging and irradiating the electron beam deflected by the deflecting element onto an object to be inspected. An electron beam image detection optical system that detects a secondary electron beam image generated from the inspection object by an electron beam that is deflected, focused, and irradiated by the electron optical system; A projection optical system for forming and projecting a plurality of slit-like or spot-like luminous fluxes from a diagonally upper position on a plurality of different places close to each other in a region on the inspection object irradiated with an electron beam by the system; and the projection optical system. A plurality of light fluxes reflected by a plurality of light fluxes imaged and projected at a plurality of locations on the inspection object are formed, and the plurality of light flux images thus formed are received to form a plurality of signals. A detection optical system for converting and detecting The information corresponding to the height of the surface at the plurality of locations is obtained by obtaining the amount of movement or the amount of phase change with respect to the reference signal corresponding to the reference height in each of the plurality of detection signals detected by the output optical system. The information corresponding to the height of the surface at the location irradiated with the electron beam shifted by the positional deviation by interpolating the information corresponding to the height of the surface at a plurality of locations (including interpolation and extrapolation). An optical height detecting device having a calculating unit for calculating the height of the electron beam, and the electron optical system based on information corresponding to a surface height at a location where the electron beam is irradiated calculated by the optical height detecting device. Focus control means for controlling the current flowing through the objective lens or the applied voltage to focus the electron beam on the inspection object in a focused state; and a secondary electron beam image detected by the electron beam image detection optical system. On the inspection object based on Electron beam inspection or measurement device characterized by comprising an image processing means for performing inspection or measurement of the made patterns. 6. An electron beam source, a deflecting element for deflecting an electron beam emitted from the electron beam source, and an objective lens for converging and irradiating the electron beam deflected by the deflecting element onto an object to be inspected. An electron beam image detection optical system that detects a secondary electron beam image generated from the inspection object by an electron beam that is deflected, focused, and irradiated by the electron optical system; A projection optical system for forming and projecting a plurality of slit-like or spot-like luminous fluxes from a diagonally upper position on a plurality of different places close to each other in a region on the inspection object irradiated with an electron beam by the system; and the projection optical system. A plurality of light fluxes reflected by a plurality of light fluxes imaged and projected at a plurality of locations on the inspection object are formed, and the plurality of light flux images thus formed are received to form a plurality of signals. A detection optical system for converting and detecting The information corresponding to the height of the surface at the plurality of locations is obtained by obtaining the amount of movement or the amount of phase change with respect to the reference signal corresponding to the reference height in each of the plurality of detection signals detected by the output optical system. The information corresponding to the height of the surface at the location irradiated with the electron beam shifted by the positional deviation by interpolating the information corresponding to the height of the surface at a plurality of locations (including interpolation and extrapolation). An optical height detection device having a calculating means for calculating the height of the surface from information corresponding to the height of the surface at a location where the electron beam is calculated by the optical height detection device. A focus control current or a focus control voltage is calculated based on a calibration parameter between the corresponding information and the focus control current or the focus control voltage, and the calculated focus control current or the focus control voltage is used as an objective lens of the electron optical system. Give to Focus control means for controlling the electron beam to be focused on the object to be inspected in a focused manner, and on the object to be inspected based on the secondary electron beam image detected by the electron beam image detection optical system. An electron beam inspection or measurement device, comprising: an image processing means for inspecting or measuring a formed pattern. 7. A deflection amount of the electron optical system to a deflection element is corrected based on information calculated by the optical height detection device and corresponding to a surface height at a location where the electron beam is irradiated. 7. The electron beam inspection or measurement apparatus according to claim 5, further comprising a deflection control unit for calibrating an image distortion including a magnification error of the electron beam image generated based on the focus control. 8. A surface which is previously mounted on a stage and irradiated with an electron beam in an area on the object to be inspected, at an interval determined based on arrangement information of a pattern formed on the object to be inspected. A height detecting device for detecting a height, and an estimating means for estimating a surface height at an arbitrary position during the interval by interpolating the surface height at the interval detected by the height detecting device. An electron beam source, a deflecting element for deflecting the electron beam emitted from the electron beam source, and an electron beam deflected by the deflecting element is focused and irradiated onto an inspection object mounted on the stage. An electron optical system having an objective lens; an electron beam image detecting optical system for detecting a secondary electron beam image generated from the inspection object by an electron beam that is deflected by the electron optical system, focused, and irradiated. Before being estimated by the estimating means Based on information on the surface height of the object to be inspected at an arbitrary position where the electron beam is irradiated by the electron optical system, the current applied to the objective lens of the electron optical system or the voltage applied is controlled to control the electron beam. Focus control means for focusing on the inspection object in a focused state; and inspecting or measuring a pattern formed on the inspection object based on a secondary electron beam image detected by the electron beam image detection optical system. An electron beam type inspection or measurement apparatus, comprising: 9. The deflection of the electron optical system based on information on the surface height of the inspection object estimated by the height detection device at an arbitrary position where the electron beam is irradiated by the electron optical system. 7. An electron beam system according to claim 5, further comprising a deflection controller for correcting an amount of deflection to the element and correcting an image distortion including a magnification error of the electron beam image generated based on the focus control. Inspection or measurement device. 10. An electron beam source, a deflecting element for deflecting an electron beam emitted from the electron beam source, and an objective lens for converging and irradiating the electron beam deflected by the deflecting element onto an object to be inspected. An electron optical system having the same; an electron beam image detecting optical system for detecting a secondary electron beam image generated from the inspection object by an electron beam deflected, focused, and irradiated by the electron optical system; A projection optical system for projecting a grid-like light beam onto the object from obliquely above and an image of the grid-like light beam reflected from the surface of the inspection object by the grid-like light beam projected by the projection optical system. A detection optical system that receives a lattice-shaped optical image corresponding to an imaging state, converts the light into a lattice-like signal, and detects the image, and forms an image with a small detection error based on the lattice-like signal detected by the detection optical system. Search for the state, and change to the imaging state where the detected error is small. Calculating a moving amount or a phase change amount with respect to a reference signal corresponding to a reference height as a lattice signal by performing a suitable weighting process on the lattice signal detected by the detection optical system. An optical height detection device having calculation means for obtaining information according to the height of the surface of the inspection object; and an optical height detection device that obtains information according to the surface height of the inspection object obtained by the optical height detection device. And controlling the relative position in the height direction between the focal position of the electron optical system and the table on which the inspection object is placed based on the obtained information to focus the electron beam on the inspection object in a focused state. Focus control means, and image processing means for inspecting or measuring a pattern formed on the inspection object based on a secondary electron beam image detected by the electron beam image detection optical system. Electron beam inspection or measurement equipment . 11. An electron beam source, a deflecting element for deflecting an electron beam emitted from the electron beam source, and an objective lens for focusing and irradiating the electron beam deflected by the deflecting element onto an object to be inspected. An electron beam image detection optical system that detects a secondary electron beam image generated from the inspection object by an electron beam that is deflected, focused, and irradiated by the electron optical system; A projection optical system for forming and projecting a plurality of slit-like or spot-like luminous fluxes from a diagonally upper position on a plurality of different places close to each other in a region on the inspection object irradiated with an electron beam by the system; and the projection optical system. A plurality of light fluxes reflected by a plurality of light fluxes imaged and projected at a plurality of locations on the inspection object are formed, and the plurality of light flux images thus formed are received to form a plurality of signals. A detection optical system that converts and detects The information corresponding to the height of the surface at the plurality of locations is obtained by calculating the amount of movement or the amount of phase change with respect to the reference signal corresponding to the reference height in each of the plurality of detection signals detected by the detection optical system. The information corresponding to the height of the surface at the location irradiated with the electron beam shifted by the positional deviation by interpolating the information corresponding to the height of the surface at a plurality of locations (including interpolation and extrapolation). An optical height detecting device having a calculating unit for calculating the height of the electron beam, and the electron optical system based on information corresponding to a surface height at a location where the electron beam is irradiated calculated by the optical height detecting device. Focus control means for controlling the relative position in the height direction between the focal position of the object and the table on which the object to be inspected is placed to focus the electron beam on the object to be inspected in a focused state; Secondary detected by detection optics Electron beam inspection or measurement device characterized by comprising an image processing means for performing inspection or measurement of a pattern formed on the inspection object based on the sagittal image. 12. A projection optical system for moving an object to be inspected at least in a predetermined direction and projecting a lattice-like light beam onto the object to be inspected from obliquely above, and a lattice-like light beam projected by the projection optical system. A detection optical system that forms a lattice-like light beam reflected from the surface of the inspection object, receives a lattice-like optical image corresponding to the image-forming state, converts the light into a lattice-like signal, and detects the signal. An optical height detection device searches for an imaging state with a small detection error based on a lattice-like signal detected by the detection optical system, and weights suitable for the searched imaging state with a small detection error. The processing is performed on the lattice-like signal detected by the detection optical system, and the electron beam is irradiated by calculating the amount of movement or the amount of phase change with respect to the reference signal corresponding to the reference height as the lattice-like signal. Area on inspected object Detecting the height of the surface of the electron optical system, and controlling the current or the voltage applied to the objective lens of the electron optical system based on the detected surface height to control the electron beam emitted from the electron beam source to the electron optical system. Is deflected by the deflecting element and focused on the object to be inspected in a focused state. The secondary electron beam image generated from the object to be inspected by the deflected, focused and irradiated electron beam in the focused state An electron beam image detecting optical system, and inspecting or measuring a pattern formed on the inspection object based on the detected secondary electron beam image. Method. 13. A projection optical system for moving an object to be inspected at least in a predetermined direction and projecting a lattice-like light beam onto the object to be inspected from obliquely above, and a lattice-like light beam projected by the projection optical system. A detection optical system that forms a lattice-like light beam reflected from the surface of the inspection object, receives a lattice-like optical image corresponding to the image-forming state, converts the light into a lattice-like signal, and detects the signal. The moving amount or the phase with respect to the reference signal as a lattice signal by performing selective processing such that a detection error is reduced with respect to the lattice signal detected by the detection optical system by the optical height detection device. The height of the surface in a region on the object to be inspected to which the electron beam is irradiated is detected by calculating the amount of change, and the current or the current applied to the objective lens of the electron optical system is applied based on the detected height of the surface. Control the voltage to The electron beam emitted from the source is deflected by a deflecting element of the electron optical system to be focused on the object to be inspected in a focused state, and is deflected, focused in the focused state, and irradiated by the irradiated electron beam. A secondary electron beam image generated from the inspection object is detected by an electron beam image detection optical system, and inspection or measurement of a pattern formed on the inspection object is performed based on the detected secondary electron beam image. An electron beam type inspection or measurement method characterized by being performed. 14. Projection for moving an object to be inspected in at least a predetermined direction, and projecting each of a plurality of slit-like or spot-like luminous fluxes from a diagonally upper position onto a plurality of adjacent different places on the object to be inspected. The optical system and the projection optical system form an image of each of the plurality of light beams reflected by the plurality of light beams imaged and projected on a plurality of locations on the inspection target object to form a plurality of the formed light beam images. A reference signal corresponding to a reference height in each of the plurality of detection signals detected by the detection optical system by an optical height detection device having a detection optical system that receives light, converts the light into a plurality of signals, and detects the signals. The information corresponding to the heights of the surfaces at the plurality of locations is obtained by calculating the amount of movement or the amount of phase change with respect to the height, and the obtained heights of the surfaces at the plurality of locations are interpolated (including interpolation and extrapolation). By shifting the position The height of the surface at the location where the generated electron beam is irradiated is calculated, and based on the calculated height of the surface, the current flowing to the objective lens of the electron optical system or the voltage to be applied is controlled to emit light from the electron beam source. The deflected electron beam is deflected by a deflecting element of the electron optical system to be focused on the object to be inspected in a focused state. Detecting a secondary electron beam image generated from the object by an electron beam image detecting optical system, and inspecting or measuring a pattern formed on the inspection object based on the detected secondary electron beam image. Electron beam inspection or measurement method. 15. A projection for moving an object to be inspected at least in a predetermined direction, and projecting each of a plurality of slit-like or spot-like luminous fluxes from a diagonally upper position onto a plurality of adjacent different places on the object to be inspected. The optical system and the projection optical system form an image of each of the plurality of light beams reflected by the plurality of light beams imaged and projected on a plurality of locations on the inspection target object to form a plurality of the formed light beam images. A reference signal corresponding to a reference height in each of the plurality of detection signals detected by the detection optical system by an optical height detection device having a detection optical system that receives light, converts the light into a plurality of signals, and detects the signals. The information corresponding to the heights of the surfaces at the plurality of locations is obtained by calculating the amount of movement or the amount of phase change with respect to the height, and the obtained heights of the surfaces at the plurality of locations are interpolated (including interpolation and extrapolation). By shifting the position The height of the surface at the location where the electron beam is irradiated is calculated, and from the calculated height of the surface, a calibration parameter between information corresponding to the height of the surface and the focus control current or the focus control voltage is calculated. A focus control current or a focus control voltage is calculated based on the calculated focus control current or the focus control voltage, and the calculated focus control current or the focus control voltage is controlled so as to be applied to an objective lens of the electron optical system, and an electron beam emitted from an electron beam source is subjected to electron optics. A secondary electron beam generated from the object to be inspected by the irradiated electron beam which is deflected by the deflecting element of the system and focused on the object to be inspected in a focused state, and is deflected and focused in the focused state. Detecting an image by an electron beam image detecting optical system, inspecting or measuring a pattern formed on the inspection object based on the detected secondary electron beam image, or an electron beam inspection or Measuring method. 16. A method according to claim 1, wherein said height detecting device is based on arrangement information of a pattern formed on the inspection object in a region on the inspection object to be irradiated with the electron beam, which is mounted on the stage. Detecting a surface height at a predetermined interval; estimating a surface height at an arbitrary position during the interval by interpolating the surface height at the detected interval; The electron beam emitted from the electron beam source is controlled by controlling the current or the voltage applied to the objective lens of the electron optical system based on the height of the surface at the position where the light is irradiated, and the electron beam is deflected by the deflection element of the electron optical system. An electron beam image detecting optical system that focuses on the inspected object in a focused state, and deflects and converges in the focused state to emit a secondary electron beam image generated from the inspected object by the irradiated electron beam. And the detected two An electron beam inspection or measurement method, wherein an inspection or measurement of a pattern formed on an object to be inspected is performed based on a secondary electron beam image. 17. A charged particle beam source such as an electron or ion, a deflecting element for deflecting a charged particle beam emitted from the charged particle beam source, and a charged particle beam deflected by the deflecting element on an object to be observed. A charged particle optical system having an objective lens for focusing and irradiating, and a charged particle beam image generated from the observation target object by the charged particle beam deflected by the charged particle optical system and focused and irradiated is detected. A charged particle beam image detection optical system, a projection optical system for projecting a lattice-like light beam onto the object to be observed from obliquely above the object to be observed, and an object to be observed by the lattice-like light beam projected by the projection optical system. A detection optical system that forms a lattice-like light beam reflected on the surface of the object to detect the position of the optical image, and detects a change in the position of the optical image composed of the lattice-like light beam detected by the detection optical system. Based on the area on the observed object An optical height detection device configured to optically detect the height of the surface to be measured, and the charged particles based on the height of the surface on the observation target detected by the optical height detection device. Focus control means for controlling a current flowing through an objective lens of the line optical system or an applied voltage to focus the charged particle beam on the object to be inspected in a focused state, and is detected by the charged particle beam image detecting optical system A charged particle beam image observation apparatus comprising a display means for displaying a charged particle beam image. 18. The method according to claim 17, further comprising: correcting an amount of deflection of the electron optical system to a deflection element based on a height of a surface on the inspection object detected by the optical height detection device to perform the focus control. 2. The apparatus according to claim 1, further comprising a deflection control unit configured to calibrate an image distortion including a magnification error of the charged particle beam image generated based on the image data.
8. The charged particle beam image observation device according to 7. 19. The method according to claim 19, further comprising determining a height of the surface on the object to be observed and a focus control current or a focus control voltage from the height of the surface on the object to be inspected detected by the optical height detecting device. The focus control current or the focus control voltage is calculated based on the calibration parameters during the period, and the calculated focus control current or the focus control voltage is controlled so as to be applied to the objective lens of the charged particle optical system, and the charged particle beam is irradiated. 19. A focus control device for focusing on an inspection object in a focused state, and a display device for displaying a charged particle beam image detected by the charged particle beam image detection optical system. The charged particle beam image observation apparatus according to the above. 20. A charged particle beam source such as an electron or an ion, a deflecting element for deflecting a charged particle beam emitted from the charged particle beam source, and a charged particle beam deflected by the deflecting element on an object to be observed. An electron optical system having an objective lens that focuses and irradiates, and a charged beam that is deflected by the electron optical system and detects a charged particle beam image generated from the observation target object by the charged particle beam irradiated by focusing. A particle beam image detection optical system, a projection optical system for projecting a lattice-like light beam onto the object to be observed from obliquely above the object to be observed, and the object to be observed by the lattice-like light beam projected by the projection optical system A detection optical system that forms a lattice-like light beam reflected on the surface to detect the position of the optical image, based on a change in the position of the optical image composed of the lattice-like light beam detected by the detection optical system. Table in the area on the object to be observed An optical height detection device configured to optically detect the height of the object, and the electron optical system based on the height of the surface of the observation target detected by the optical height detection device. Focus control means for controlling a relative position in a height direction between a focal position and a table on which the object to be observed is placed, so as to focus the charged particle beam on the object to be observed in a focused state; and the charged particle beam. Display means for displaying a charged particle beam image detected by the image detection optical system. 21. A charged particle beam source such as an electron or an ion, a deflecting element for deflecting a charged particle beam emitted from the charged particle beam source, and a charged particle beam deflected by the deflecting element on an object to be processed. A charged particle optical system having an objective lens that focuses and irradiates, and a charged particle beam image generated from on the workpiece by the charged particle beam deflected by the charged particle optical system and focused and detected is detected. A charged particle beam image detecting optical system, a projection optical system for projecting a lattice-like light beam onto the object to be processed from obliquely above the object to be processed, and a lattice-shaped light beam projected by the projection optical system. A detection optical system that forms a lattice-like light beam reflected on the surface of the object to detect the position of the optical image, and detects a change in the position of the optical image composed of the lattice-like light beam detected by the detection optical system. Based on the area on the workpiece An optical height detector configured to optically detect the height of the surface to be processed, and the charged particles based on the height of the surface on the workpiece detected by the optical height detector. Focus control means for controlling a current or a voltage applied to the objective lens of the line optical system to focus the charged particle beam on the inspected object in a focused state; and A charged particle beam image processing apparatus comprising means for processing a surface of a processing object. 22. A method for correcting the amount of deflection of the electron optical system to a deflecting element based on a height of a surface on an object to be inspected detected by the optical height detecting device, and controlling the focus control. 3. A deflection control means for correcting an image distortion including a magnification error of a charged particle beam image generated on the basis of the deviation.
2. The charged particle beam image processing apparatus according to 1. 23. A projection optical system for projecting a lattice-like light beam onto an object from obliquely above, and a lattice-like light beam reflected from the surface of the object by the lattice-like light beam projected by the projection optical system. A detection optical system that receives a lattice-shaped optical image corresponding to the image formation state, converts the light into a lattice-like signal, and detects the lattice-like optical image; and detects based on the lattice-like signal detected by the detection optical system. An imaging state with a small error is searched for, and a weighting process adapted to the imaging state with a small detection error is performed on a grid-like signal detected by the detection optical system, thereby forming a grid-like state. Calculating means for obtaining an amount of movement or a phase change with respect to the reference signal corresponding to the reference height as a signal to obtain information corresponding to the height of the surface of the object; optical height detection apparatus. 24. A projection optical system for projecting a grid-like light beam onto a target object from obliquely above, and a grid-like light beam reflected from the surface of the target object by the grid-like light beam projected by the projection optical system. A detection optical system that receives a lattice-like optical image corresponding to the image formation state, converts the light into a lattice-like signal, and detects the lattice-like optical image; and detects the lattice-like signal detected by the detection optical system. Calculating means for obtaining a movement amount or a phase change amount with respect to a reference signal as a lattice signal by performing selective processing such that an error is reduced to obtain information corresponding to the height of the surface of the object. An optical height detecting device, comprising: 25. A projection optical system for forming and projecting a plurality of slit-like or spot-like luminous fluxes from a diagonally upper direction onto a plurality of different places close to each other on an object; Detecting each of a plurality of light beams reflected by a plurality of light beams imaged and projected at a plurality of locations, receiving the formed plurality of light beam images, converting them into a plurality of signals, and detecting them An optical system, and a movement amount or a phase change amount with respect to a reference signal corresponding to the reference height in each of the plurality of detection signals detected by the detection optical system is determined, and information corresponding to the surface height at the plurality of locations is obtained. Then, the information corresponding to the surface height at an arbitrary position is calculated by interpolating (including interpolation and extrapolation) the information corresponding to the obtained surface heights at a plurality of locations. Means. Optical height detector.