JP5542095B2 - Inspection device - Google Patents

Description

  The present invention relates to a technique for inspecting a height variation of a sample such as a semiconductor device.

  A technique for measuring the state of a circuit pattern in the middle of a process in a semiconductor device manufacturing line is known. This technology is positioned as an indispensable technology for improving manufacturing yield and quality control. One technique for measuring the state of a circuit pattern is a length measuring SEM (Scanning Electron Microscopy) that measures the dimensions of the circuit pattern. The length measurement SEM uses the high resolution and deep focal depth peculiar to the SEM to measure the dimension of the circuit pattern in the lateral direction (in the in-plane direction of the circuit pattern), thereby realizing sub-nm measurement accuracy.

  Further, as one of the techniques for measuring the state of a circuit pattern, there is a mirror projection type inspection apparatus that detects an electrical characteristic defect of a circuit pattern (for example, Patent Document 1). The mirror projection type inspection apparatus irradiates a semiconductor wafer with a planar electron beam and repels and pulls back electrons near the surface of the circuit pattern of the semiconductor wafer to which a voltage is applied (hereinafter referred to as mirror electrons). An image is formed by a lens, and this captured image (hereinafter referred to as a mirror image) is observed. A portion having a defective electrical characteristic of the circuit pattern is charged differently from a portion having a normal electrical characteristic. For this reason, the mirror electrons pulled back near the location where the electrical characteristics are defective draw a different trajectory from the mirror electrons pulled back near the location where the electrical characteristics are normal, and form an image. Therefore, by observing this mirror image, it is possible to detect a defective electric characteristic portion.

  One technique for measuring the state of a circuit pattern is AFM (Atomic Force Microscopy) that measures the shape of the circuit pattern in the height direction. The AFM scans the surface of the object to be inspected using a thin needle having a tip curvature of sub-μm, and the height of the object to be inspected is imaged by synchronizing the movement amount of the needle in the height direction with the scanning. Measure.

Japanese Patent Application No. 11-108864

  According to the AFM, it is possible to detect variations in the height direction of the inspection object. However, the measurement accuracy of AFM greatly depends on the scanning speed of the needle, and the accuracy decreases as the scanning speed is increased. For this reason, it takes a lot of time to detect the variation in the height direction of the inspection object over a wide range with high accuracy.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to detect variations in the height direction of an object to be inspected.

  In order to solve the above-described problems, the present invention performs a charge control process as necessary so that the charged state of an object to be inspected is uniform, and uses a mirror projection method to inspect the object having a uniform charged state. The mirror electrons repelled and pulled back in the vicinity of the surface of the lens are imaged by a lens, and this mirror image is observed.

For example, the present invention is an inspection apparatus for inspecting variation in the height direction of a sample,
Holding means for holding the sample;
Charging processing means for charging the sample held in the holding means;
Voltage applying means for applying a voltage to the sample held in the holding means;
An electron optical system that irradiates an electron beam toward a sample to which a voltage is applied by the voltage applying unit, and forms an image of mirror electrons pulled back near the surface of the sample;
Image processing means for image processing a mirror image obtained by imaging the mirror electrons,
The image processing means includes
A process of generating a difference image representing a difference between a mirror image obtained by imaging the mirror electrons and a mirror image of the standard product;
The pixels constituting the difference image to generate a binary image by binarizing, calculate the percentage of pixels of the two values constituting the binarized image, the ratio of the pixel of the binarized image And processing for outputting the information on the variation in the height direction of the sample according to the difference image .

  According to the present invention, when there is a difference in height on the surface of the object to be inspected, the mirror electrons pulled back in the vicinity of different heights draw different trajectories and form images at different focal positions according to the height difference. To do. Therefore, it is possible to detect the height difference by observing this mirror image. Therefore, it is possible to detect variations in the height direction of the inspection object by comparing the height difference indicated by the mirror image of the standard product prepared in advance with the height difference indicated by the mirror image of the inspection object. it can.

FIG. 1A and FIG. 1B are diagrams schematically showing a locus of mirror electrons in the vicinity of a circuit pattern of a semiconductor wafer. FIG. 2 is a schematic view of a mirror projection type inspection apparatus to which the first embodiment of the present invention is applied. FIG. 3 is a flowchart for explaining inspection recipe creation processing of the mirror projection type inspection apparatus to which the first embodiment of the present invention is applied. FIG. 4 is a flowchart for explaining the filter pattern creation processing. FIG. 5 is a diagram showing an example of a recipe creation GUI screen displayed by the console terminal 28 in the inspection recipe creation process shown in FIG. 3, and FIG. 5A shows a case where a filter pattern is registered as a height variation inspection template. An example GUI screen is shown, and FIG. 5B shows an example GUI screen when a reference pattern is registered as a height variation inspection template. FIG. 6 is a flowchart for explaining the height variation detection processing of the mirror projection type inspection apparatus to which the first embodiment of the present invention is applied. FIG. 7 is a flowchart for explaining the filter pattern method. FIG. 8 is a flowchart for explaining the reference pattern method. FIG. 9 is a view showing an example of the inspection result GUI screen displayed on the console terminal 28 in the height variation inspection processing shown in FIG. 6, and FIG. 9A shows an example of the inspection result GUI screen by the filter pattern method. FIG. 9B shows an example of a test result GUI screen by the reference pattern method. FIG. 10A is a schematic diagram (a front view and a bottom view) of an example of a standard product for height correction, and FIG. 10B is a schematic diagram of an example of a standard sample 52 constituting the standard product for height correction. (A front view, a bottom view, and a side view). FIG. 11 is a flowchart for explaining the electron optical system correction amount calculation processing of the mirror projection type inspection apparatus to which the second embodiment of the present invention is applied. FIG. 12 is a diagram showing an example of an optical system adjustment GUI screen displayed on the console terminal 28 in the electron optical system correction amount calculation processing shown in FIG. FIG. 13 is a schematic view of a mirror projection type inspection apparatus to which the third embodiment of the present invention is applied. FIG. 14 (A) is a flowchart for explaining the inspection process of the height variation of the mirror projection type inspection apparatus to which the third embodiment of the present invention is applied, and FIG. 14 (B) is the third embodiment of the present invention. It is a flowchart for demonstrating the review process of height difference using AFM of the mirror projection type inspection apparatus to which the embodiment is applied. FIG. 15 is a diagram schematically showing a mirror image of a review place and a measurement result. FIG. 16 is a diagram showing an example of the review GUI screen displayed by the console terminal 28 in the processing shown in FIGS. 14 (A) and 14 (B). FIG. 17 is a diagram illustrating an example of a semiconductor device manufacturing process. FIG. 18 is a diagram showing an example of a data management GUI screen displayed on the console terminal 28 in data management of inspection results, and FIG. 18A is a GUI screen for analyzing data of one inspection result. 18 (B) is a GUI screen for statistically analyzing the inspection result of the same process.

<Principle>
Prior to the description of the embodiments of the invention, the principle of detecting the height difference of the inspection object using the mirror projection method will be described.

  FIG. 1A and FIG. 1B are diagrams schematically showing a locus of mirror electrons in the vicinity of a circuit pattern of a semiconductor wafer. FIG. 1A shows a locus of mirror electrons in the vicinity of a defective electric characteristic of a circuit pattern generated in a contact process of manufacturing a semiconductor wafer, and FIG. 1B shows the vicinity of a circuit pattern having a height difference. The locus of mirror electrons is shown.

  As shown in FIG. 1A, for example, when the electrical characteristic normal part 10 of the circuit pattern is charged to 0V and the electrical characteristic defective part 11 is charged to -1V, the equipotential surface 12a near the electrical characteristic defective part 11 is curved. To do. As a result, the mirror electrons 13a bounced in the vicinity of the defective electrical property portion 11 draw a different locus from the mirror electrons 13b bounced in the vicinity of the normal electrical property portion 10, and are imaged at different focal positions in the height direction. The

  Similarly, as shown in FIG. 1B, since the equipotential surface 12b is curved according to the height difference 14 of the circuit pattern, the mirror electrons 13c draw different trajectories according to the height difference 14 of the circuit pattern. The images are formed at different focal positions in the height direction.

  Therefore, in the present invention, the object to be inspected is charged as necessary so that the charged state of the circuit pattern is uniform (so that the normal electrical characteristic portion and the poor electrical property portion are in the same charged state). Then, the height difference of the inspection object is detected by observing the mirror image of the inspection object using the mirror projection method.

<First Embodiment>
FIG. 2 is a schematic view of a mirror projection type inspection apparatus to which the first embodiment of the present invention is applied.

  As shown in the figure, the mirror projection type inspection apparatus 2 of the present embodiment includes an electron optical system 20, a sample chamber 21 in which a sample (semiconductor wafer) 29 is placed, a charge control unit 23, a stage control unit 24, and an electron An optical system control unit 25, a system control unit 26, an image processing unit 27, and a console terminal 28 for receiving instructions from an operator and displaying information to the operator.

  The electron optical system 20 includes an irradiation optical system 201, an imaging optical system 202, and a projection optical system 203. The electron optical system 20 is always placed in an atmosphere evacuated by an evacuation device (not shown).

  The irradiation optical system 201 includes an electron source 2011, a condenser lens 2012, and a diaphragm 2013. The imaging optical system 202 includes a beam separator 2021 and an objective lens 2022. The projection optical system 203 includes an intermediate lens 2031, a projection lens 2032, and a detector 2033.

  The electron beam emitted from the electron source 2011 is converged by the condenser lens 2012, and the trajectory is limited by the stop 2013 so as to be a desired irradiation region. Thereafter, a focal point is formed on the focal plane of the objective lens 2022 located around the beam separator 2021. The focal plane of the objective lens 2022 is determined by the excitation of the objective lens 2022 and the action of the electrostatic lens by retarding described later. By forming a focal point on this focal plane, the electron beam deflected by the beam separator 2021 becomes a planar electron beam whose irradiation direction is perpendicular to the surface of the sample 29 via the objective lens 2022.

  A negative potential slightly higher than the acceleration voltage of the electron beam is applied to the sample 29 by a retarding power source 242 described later of the stage control unit 24. Therefore, an electric field reflecting the shape and charge state of the circuit pattern formed on the surface of the sample 29 is formed near the surface of the sample 29. This electric field functions as a decelerating means, and most of the electron beam is pulled back before colliding with the sample 29, and the shape and charging of the circuit pattern of the sample 29 in the direction according to the shape and charging state of the sample 29. It becomes mirror electrons that move with an intensity corresponding to the state.

  The mirror electrons are converged by the objective lens 2022 and deflected by the beam separator 2021 so as to be aligned with the optical axis of the projection optical system 203. Thereafter, an image representing a local change in the charged state and a height difference of the surface of the sample 29 is formed on the detector 2033 by the intermediate lens 2031 and the projection lens 2032. Here, the electron source 2011 preferably has a narrow energy width of emitted electrons. Which equipotential surface of the electric field formed on the surface of the sample 29 is pulled back is determined by the energy of the electrons incident on the sample 29. Therefore, if there is an energy width of the emitted electrons, the mirror is pulled back on a different equipotential surface. Electrons are imaged. This degrades the resolution of the image. The detector 2033 is composed of, for example, a TDI (Time Delay Integration) sensor or a CCD (Charge Coupled Devices) sensor, obtains an image formed on the detection surface at a high speed, and obtains the image according to the amount of mirror electrons. The signal is converted into a signal and transmitted to the image processing unit 27.

  The sample chamber 21 includes a holder 211 on which the sample 29 is placed, a stage 212 that moves the sample 29 in a direction perpendicular to the incident direction of the electron beam, a stage driving device 244 such as a motor, and an optical microscope 213. .

  The holder 211 is electrically insulated from the stage 212 so that a retarding voltage can be applied to the holder 211 and the sample 29 placed on the holder 211. A retarding voltage output from a retarding power source 242 disposed outside the sample chamber 21 (stage control unit 24) is applied to the holder 211 via a high-pressure feedthrough (not shown). In addition, the interior of the sample chamber 21 is always evacuated by a vacuum evacuation device (not shown), like the electron optical system 20.

  The optical microscope 213 is used when searching for a desired location in the sample 29 placed on the holder 211 or creating an inspection recipe.

  The charging control unit 23 measures the charged state of the sample 29 placed on the holder 211 and charges the sample 29 so that the charged state becomes uniform. The charging process is performed by applying a potential to the charge control electrode 231 disposed in the vicinity of the sample 29. The charged state of the sample is measured by measuring the value of the current flowing through the charge control electrode 231. By performing this process, the distortion of the electric field due to the non-uniformity of the charged state can be removed from the electric field formed on the surface of the sample 29 by another electron beam apparatus or the like. Only the distortion of the electric field due to the height difference of the surface 29 can be detected. Note that a surface potential meter or the like may be used for measuring the charged state of the sample 29.

  The stage control unit 24 includes a main control device 241, a retarding power source 242, and a stage position measurement device 243.

  The main control device 241 receives a control signal related to stage movement from the system control unit 26, and based on the received control signal, the main control device 241 uses the stage 212 measured by the stage position measuring device 243 as a reference for the stage driving device 244. The stage 212 is moved under control. As a result, the holder 211 placed on the stage 212 moves in the direction perpendicular to the incident angle of the electron beam, and acquires a mirror image of a desired region of the sample 29 placed on the holder 211. Can do. The main controller 241 transmits the positions of the stage 212, the charging controller 23, the optical microscope 213, and the electron optical system 20 measured by the stage position measuring device 243 to the system controller 26. The system control unit 26 selects the desired sample 29 placed on the holder 211 based on the positions of the stage 212, the charging control unit 23, the optical microscope 213, and the electron optical system 20 received from the main control device 241. A control signal for moving the stage 212 is generated so as to be located immediately below any of the charging control unit 23, the optical microscope 213, and the electron optical system 20, and this is transmitted to the main controller 241. As a result, the main controller 241 moves the desired region of the sample 29 placed on the holder 211 so as to be directly below any one of the charge controller 23, the optical microscope 213, and the electron optical system 20. Can do.

  The electron optical system control unit 25 includes a main control device 251, an irradiation optical system control device 252, and a projection optical system control device 253.

  The main control device 251 receives a control signal related to electron beam irradiation from the system control unit 26, and based on the received control signal, the irradiation optical system control device 252 sends each part of the irradiation optical system 201 and the imaging optical system 202. (Electron source 2011, lenses 2012, 2022, beam separator 2021, astigmatism corrector, deflector, etc. not shown) are controlled so that mirror electrons 5 are imaged at a desired contrast toward the sample 29. Irradiate with an electron beam. At the same time, the retarding power source 242 of the stage control unit 24 is controlled. In addition, a control signal related to the imaging of the mirror image is received from the system control unit 26, and based on the received control signal, each part of the projection optical system 203 (lenses 2031, 2032, detector 2033) is sent to the projection optical system control device 253. ) Is controlled, and the mirror electrons pulled back in the vicinity of the sample 29 are detected.

  The image processing unit 27 includes an image forming device 271 and an image processing arithmetic device 272.

  The image forming apparatus 271 synchronizes the signal corresponding to the intensity of the mirror electrons detected by the detector 2033 of the projection optical system 203 with the coordinate signal of the stage 212 sent from the main controller 241 of the stage controller 24. Capture. The captured signal is AD converted and transmitted to the image processing arithmetic device 272. The image processing arithmetic device 272 performs image processing for extracting variation in the height direction from the mirror image represented by the signal received from the image forming device 271, and sends the result together with the mirror image to the system control unit 26 as necessary. Send.

  The system control unit 26 generates a control signal for moving the stage 212 in accordance with a user instruction received via the console terminal 28 and transmits the control signal to the stage control unit 24, and is mounted on the charge control unit 23 on the holder 211. The sample 29 is charged so that the charged state of the placed sample 29 is uniform. In addition, the height variation detection result or the mirror image and the height variation detection result received from the image processing unit 27 are displayed on the display screen of the console terminal 28 and presented to the operator.

In the mirror projection type inspection apparatus having the above configuration, the charging control unit 23, the stage control unit 24, and the electron optical system control unit 25 are integrated logic ICs such as ASIC (Application Specific Integrated Circuits) and FPGA (Field Programmable Gate Array). It can be realized in hardware, or DSP (Digital Signal Process)
or) may be realized by software. In addition, the system control unit 26, the image processing unit 27, and the console terminal 28 read information from an external storage device such as a CPU, a memory, and an HDD, and a portable storage medium 409 such as a CD-ROM and a DVD-ROM. In a general computer having a reader, an input device such as a keyboard and a mouse, an output device such as a display, a communication device for communicating with a partner device via a communication line, and a bus connecting these devices This can be realized by the CPU executing a predetermined program loaded on the memory.

  Next, the operation of the mirror projection type inspection apparatus having the above configuration will be described. The operation of the mirror projection type inspection apparatus having the above-described configuration includes information necessary for inspecting the height variation of the sample 29 (alignment information of the electron optical system 20, charge control information of the charge control device 23, and height variation detection). Therefore, the process can be classified into two processes: a process for creating an inspection recipe including the standard height difference information of the sample 29 and a process for detecting the height variation of the sample 29 using the test recipe.

  FIG. 3 is a flowchart for explaining inspection recipe creation processing of the mirror projection type inspection apparatus to which the first embodiment of the present invention is applied.

  The system control unit 26 receives input of basic information of the sample 29 to be measured from the operator via the console terminal 28 (S101). When the sample 29 is a semiconductor wafer, the type of the semiconductor wafer, the name of the manufacturing process, and the like correspond to the basic information. The system control unit 26 stores this basic information in a memory or the like in order to manage inspection recipes to be created.

  Next, the system control unit 26 receives input of layout information of the circuit pattern of the sample 29 to be measured from the operator via the console terminal 28, and associates it with the basic information of the sample 29 received in S101. (S102). Here, the layout information of the circuit pattern may be input to the console terminal 28 while the operator confirms the optical image obtained by the optical microscope 213 or the dimensions described in the CAD data of the sample 29. And may be input to the console terminal 28.

  Next, the system control unit 26 receives from the operator via the console terminal 28 whether or not it is necessary to perform a charge control process on the sample 29 to be measured (S103). Here, whether or not charging control processing is necessary is determined and set by the operator. However, it is also possible to automatically determine and set from the basic information of the sample 29 received in S101. There are two types of automatic determination methods. One refers to the inspection recipe of the sample 29 having the same basic information created in the past and stored in the memory or the like, and if the necessity of the charge control processing described in this inspection recipe is “necessary” In this case, it is determined that the charge control process is necessary again this time. The other is that correspondence table data indicating the necessity of charge control processing is stored in advance in a memory or the like for each basic information of the sample, and the necessity of charge control processing is determined using this correspondence table data. For example, when the sample 29 is a semiconductor wafer, the structure and material on the wafer surface are reflected in the process name. Therefore, correspondence table data indicating the necessity of charge control processing is created for each process name and stored in a memory or the like. The system control unit 26 determines from the correspondence table data according to the necessity of the charging control process associated with the same name as the basic information.

  If the charging control process is unnecessary (NO in S104), the system control unit 26 proceeds to S107. On the other hand, if necessary (YES in S104), the system control unit 26 receives an input of the set voltage used for the charging control processing from the operator via the console terminal 28, and uses this as the basic information of the sample 29 received in S101. The information is registered in the memory or the like in association (S105) and the process proceeds to S106. Here, the set voltage depends on the pattern so as to make the charged state uniform over the entire sample 29, that is, “there is no uneven charging in the circuit pattern” and “a problem in measuring the height variation”. It is determined that no charging contrast occurs. In a mirror projection type inspection apparatus in which electrons are not incident on a circuit pattern during imaging, the charged state of most of the samples 29 can be made uniform by charge removal processing.

  In S106, the system control unit 26 registers the set voltage registered in the standard product of the sample 29 placed on the holder 211 (the sample serving as a reference for height variation determination) in association with the basic information of the sample 29 in S105. Is applied, and the charge control processing of the standard sample 29 is performed. Then, it progresses to S107.

  In step S <b> 107, the system control unit 26 performs registration processing for an alignment template used in an alignment process for adjusting the positional relationship between the coordinates of the stage 212 and the coordinates of the circuit pattern of the sample 29 placed on the holder 211. In this embodiment, the optical image is roughly used for alignment (first alignment process), and then the mirror image is used for detailed alignment (second alignment process). Therefore, the system control unit 26 obtains an optical image of a standard product of the sample 29 imaged in advance in accordance with an instruction received from the operator via the console terminal 28, and accepts this in step S101 as a first alignment template. The information is registered in a memory or the like in association with the basic information of the sample 29. Similarly, a mirror image of a standard image of the sample 29 imaged in advance is obtained, and this is registered as a second alignment template in the memory or the like in association with the basic information of the sample 29 received in S101.

  Next, the system control unit 26 moves the stage 212 by controlling the stage control unit 24 in accordance with an instruction received from the operator via the console terminal 28, and aligns the sample 29 (first alignment and second alignment). The position coordinates of the place to be performed are registered in the memory or the like in association with the basic information of the sample 29 received in S101 (S108). Here, in order to accurately adjust the positional relationship between the coordinates of the stage 212 and the coordinates of the sample 29 (circuit pattern) placed on the holder 211, it is necessary to perform alignment in at least two places. Therefore, in the present embodiment, two templates (optical image and mirror image) for the first alignment and the second alignment are registered in S107 described above, and at least two locations for acquiring images used for alignment in S108. Two different coordinates on the sample 29 are registered for each of the first alignment and the second alignment. Naturally, the position coordinates for acquiring the images for the first alignment and the second alignment need to be different from the position coordinates for acquiring the template image.

  Next, the system control unit 26 controls the stage control unit 24 to move the stage 212 to the two position coordinates registered in S108. The optical microscope 213 and the detector 2033 display the alignment images. Take an image. Then, alignment is performed by comparing the obtained image (optical image, mirror image) with the alignment template registered in association with the basic information of the sample 29 in S107 (S109). The alignment image may be acquired in step S108 for registering position coordinates. In that case, the obtained image is stored in the memory in association with the alignment image index, and the image is called in step S109 to execute the alignment.

  Specifically, first, the system control unit 26 performs a first alignment process. In accordance with an instruction received from the operator via the console terminal 28, one of the first alignment templates is selected from the first alignment templates registered in S107, and two (or a plurality) of optical images captured in S109. Select one of the images. Then, the selected first alignment template and optical image are displayed on the console terminal 28. The operator compares the first alignment template displayed on the console terminal 28 with the optical image, and observes a shift between them (for example, a shift in feature points such as an edge portion of the circuit pattern). This process is performed for all the optical images acquired in S109 and for all of the first alignment templates. The deviation amount obtained as a result is input to the console terminal 28. The system control unit 26 receives the shift amount received from the operator via the console terminal 28 as a first alignment amount (offset amount) for matching the coordinates of the stage 212 with the coordinates of the circuit pattern of the sample 29 in S101. The information is registered in the system control unit 26 in association with the basic information of the sample 29.

  Next, the system control unit 26 performs a second alignment process. According to the instruction received from the operator via the console terminal 28, one second alignment template is selected from the second alignment templates registered in S107, and two (or a plurality) of mirror images captured in S109. Select one of the images. Then, the selected second alignment template and mirror image are displayed on the console terminal 28. At this time, a mirror image is displayed by reflecting the first alignment amount on the coordinates of the circuit pattern of the sample 29. The operator compares the second alignment template displayed on the console terminal 28 with the mirror image, and observes a shift between them (for example, a shift in feature points such as an edge portion of the circuit pattern). This process is performed for all the optical images acquired in S109 and for all of the second alignment templates. The deviation amount obtained as a result is input to the console terminal 28. The system control unit 26 receives the shift amount received from the operator via the console terminal 28 as a second alignment amount (offset amount) for aligning the coordinates of the stage 212 with the coordinates of the circuit pattern of the sample 29 in S101. The information is registered in the system control unit 26 in association with the basic information of the sample 29.

  After the above alignment is completed, the system control unit 26 performs a registration process of a measurement location search template used for searching for a measurement location of the circuit pattern of the sample 29 (S110). In this embodiment, as a measurement location search template, a mirror image of a measurement location imaged in advance at a low magnification and the coordinates of the stage 212 of this measurement location are received from the operator via the console terminal 28, and this is received in S101. The information is registered in a memory or the like in association with the basic information of the sample 29. Here, in the case of a wafer inspection apparatus, the low magnification means a magnification that can specify a repeating unit of a circuit pattern formed on a wafer, for example, a location such as an exposure shot, a die, or a memory mat on the wafer. Means. In the current semiconductor wafer, the repeating unit of the circuit pattern has a size of about 50 μm to 200 μm, and the magnification of the acquired image is about 500 to 2000 times.

  Next, the system control unit 26 performs registration processing of a height variation inspection template for inspecting height variation at the measurement location of the circuit pattern formed on the sample 29 (S111). Specifically, the coordinates of the stage 212 are corrected using the first alignment amount and the second alignment amount registered in association with the basic information of the sample 29. Next, the stage control unit 24 is controlled using the corrected coordinates of the stage 212, and the stage 212 is placed at the measurement location of any measurement location search template registered in association with the basic information of the sample 29 in S110. The electron optical system controller 25 is controlled to cause the electron optical system 20 to capture a low-magnification mirror image. Then, pattern matching is performed between the captured mirror image and the template mirror image to determine the coordinates of the stage 212 at the measurement location. Next, the electron optical system control unit 25 is controlled to capture a mirror image of the measurement location specified by the determined coordinates. This mirror image has an imaging magnification that is substantially the same as the imaging magnification of the mirror image used when inspecting the height variation. Then, a height variation inspection template at this measurement location is created from the captured mirror image, and is registered in a memory or the like in association with the basic information of the sample 29. Details of the height variation inspection template creation processing will be described later.

  Now, the system control unit 26 inquires of the operator whether or not to continue the registration of the height variation inspection template via the console terminal 28, and when continuing (NO in S112), returns to S110 and does not continue ( If YES in S112), that is, if the registration of the height variation inspection templates at all measurement locations is completed, this flow is terminated.

  As described above, the inspection recipe information of the sample 29 is registered in the memory or the like in association with the basic information of the sample 29.

  Next, a process for creating a height variation inspection template (S111 in FIG. 3) will be described. The height variation inspection template creation process includes a process of creating a filter pattern for filtering information about height difference from a mirror image of a standard sample of the sample 29 (filter pattern creation process), and a standard product of the sample 29. There is a process of creating a reference pattern for comparing with a mirror image of the inspection sample 29 from the mirror image (reference pattern creating process). Hereinafter, each of the filter pattern creation process and the reference pattern creation process will be described.

  First, the filter pattern creation process will be described.

  FIG. 4 is a flowchart for explaining the filter pattern creation processing.

  First, the image processing unit 27 converts a signal corresponding to the intensity of the mirror electrons detected by the detector 2033 of the projection optical system 203 into a coordinate signal of the stage 212 sent from the main control device 241 of the stage control unit 24. Capture in sync. Then, the captured signal is AD converted to generate a mirror image (S1111). The standard sample 29 is preferably one having as little height variation as possible. When a mirror image is acquired with the sample 29 having a large height variation, the characteristics of the reference filter pattern are deteriorated, and the measurement accuracy of the subsequent height variation is deteriorated.

  Next, the image processing unit 27 selects a region (trimming window) having a high 2n × 2n pixel size from the mirror image and performs Fourier transform (fast Fourier transform) to generate a Fourier transform image (S1112). ). Then, the generated Fourier transform image is transmitted to the system control unit 26. The system control unit 26 displays the Fourier transform image received from the image processing unit 27 on the console terminal 28.

  Next, the system control unit 26 receives a filter pattern created based on the Fourier transform image from the operator via the console terminal 28. The operator enlarges the Fourier transform image displayed on the console terminal 28 as necessary, and confirms the spot 301 due to the periodicity of the original mirror image (S1113). Next, the operator selects each spot using a pointing device such as a mouse. The console terminal 28 generates a filter pattern for masking each spot selected by the operator (S1114) and transmits it to the system control unit 26. Here, if the size of the mask is too large, the height variation information may be lost during the inverse Fourier transform described later. For this reason, it is preferable to prepare a plurality of mask sizes for concealing one spot on the console terminal 28 according to the application. A small mask may be used when it is desired to measure subtle height variations with high accuracy, and a large mask may be used when only protruding height variations are to be extracted. By performing inverse Fourier transform by masking all the spots that can be confirmed in the Fourier transform image, the periodic component of the height difference of the standard sample 29 (the periodic component of the original height of the circuit pattern to be inspected) Can be removed.

  Next, the system control unit 26 transmits the filter pattern received from the console terminal 28 to the image processing unit 27. The image processing unit 27 combines the filter pattern received from the system control unit 26 and the Fourier transform image, and masks the Fourier transform image (S1115). Then, the masked Fourier transform image is subjected to inverse Fourier transform, and the inverse Fourier transform image is transmitted to the system control unit 26 (S1116). The system control unit 26 displays the inverse Fourier transform image received from the image processing unit 27 on the console terminal 28. When the periodic component of the height difference of the standard sample 29 (periodic component of the height inherent in the circuit pattern to be inspected) can be removed, the inverse Fourier transform image is a flat (uniform) image. Therefore, the operator evaluates the contrast of the inverse Fourier transform image and determines whether or not to register the filter pattern from the evaluation result. Then, the determination result is input to the console terminal 28. When the determination result input to the console terminal 2 is the registration of the filter pattern, the system control unit 26 registers the above filter pattern in the memory or the like in association with the basic information of the sample 29. On the other hand, when recreating the filter pattern, the process returns to S1114.

  Next, the reference pattern creation process will be described.

  The reference pattern includes a mirror image itself and an inverse Fourier transform image generated by performing an inverse Fourier transform on only a periodic component of a circuit pattern included in the Fourier transformed image of the mirror image.

  When the reference pattern is the mirror image itself, the image processing unit 27 receives a signal corresponding to the intensity of the mirror electrons detected by the detector 2033 of the projection optical system 203 from the main control device 241 of the stage control unit 24. Captured in synchronism with the coordinate signal of the incoming stage 212. The captured signal is AD-converted to generate a mirror image, which is transmitted to the system control unit 26. The system control unit 26 registers the mirror image received from the image processing unit 27 as a reference pattern in a memory or the like in association with the basic information of the sample 29.

  On the other hand, when the reference pattern is an inverse Fourier transform image, the image processing unit 27 performs Fourier transform on the generated mirror image, filters the Fourier transform image to extract a periodic component of the circuit pattern, and extracts the period of the extracted circuit pattern. Inverse Fourier transform the component. Then, the inverse Fourier transform image is transmitted to the system control unit 25. The system control unit 26 registers the inverse Fourier transform image received from the image processing unit 27 as a reference pattern in a memory or the like in association with the basic information of the sample 29.

  FIG. 5 is a diagram showing an example of a recipe creation GUI screen displayed by the console terminal 28 in the inspection recipe creation process shown in FIG. FIG. 5A shows an example of a GUI screen when a filter pattern is registered as a height variation inspection template, and FIG. 5B is a GUI screen when a reference pattern is registered as a height variation inspection template. An example is shown.

  As shown in FIGS. 5A and 5B, the registration GUI screen for the height variation inspection template is a layout display for displaying the layout of the circuit pattern formed on the sample 29 (for example, a semiconductor wafer). A window 41, an image processing designation field 42 for designating a height variation extraction method and a measurement method, an acquired image display window 43 for displaying various images including a mirror image and a height variation inspection template, And a parameter setting field 44 for setting image processing parameters for extracting and measuring height variations.

  The layout display window 41 is used for designating a circuit pattern measurement location. When the operator operates a pointing device such as a mouse to operate the pointer 45 and designates the measurement location of the circuit pattern, the console terminal 28 transmits information on the measurement location designated by the pointer 45 to the system control unit 26. . The system control unit 26 controls the stage control unit 24, the optical microscope 213, and the image processing unit 27, acquires image data of the optical image at the measurement location designated by the pointer 45, and transmits this to the console terminal 28. . The console terminal 28 displays the acquired image data of the optical image on the layout display window 46. When the mirror image is not displayed on the acquired image display window 43, the console terminal 28 activates the display window 46 and updates the optical image in real time. The operator can register a measurement location (measurement region) by designating a desired region on the optical image with the pointer 45. It should be noted that the magnification of the optical image displayed on the display window 46 can be changed so that a desired region can be designated from the optical image for various patterns. The registered measurement location is displayed by the marker 47 on the layout display window 41 together with the ID number indicating the registration order by the console terminal 28.

  The image processing designation field 42 has check boxes for designating a height variation extraction method and measurement method. In the present embodiment, either a fast Fourier transform method (the above-described filter pattern creation process) or a difference image method (the above-described reference pattern creation process) can be designated as a method for extracting height variation. Further, either a histogram method or an area ratio method can be designated as a method for measuring the height variation. FIG. 5A shows the case where the fast Fourier transform method is designated as the height variation extraction method, the histogram method is designated as the height variation measurement method, and FIG. 5B shows the height variation. The case where the difference image method is designated as the variation extraction method and the area ratio method is designated as the height variation measurement method is illustrated. A method for measuring the height variation will be described later.

  The acquired image display window 43 displays an image that the operator refers to when creating a height variation inspection template at a measurement location. When the fast Fourier transform method is designated as the height variation extraction method, as shown in FIG. 5A, a mirror image, a Fourier transform image of the mirror image, and a trimming window 431 for trimming the Fourier transform image. Then, the inverse Fourier transform image obtained by performing inverse Fourier transform by masking the filter pattern and the Fourier transform image with the filter pattern is displayed. When the difference image method is designated as the height variation extraction method, as shown in FIG. 5B, only the mirror image (reference pattern), or the mirror image, the Fourier transform image of the mirror image, Then, an inverse Fourier transform image (reference pattern) obtained by performing inverse Fourier transform on the periodic component extracted from the Fourier transform image is displayed. At the bottom of the acquired image display window 43, there is an optical image / mirror image switching button. When an optical image is selected by the switching button, the optical image displayed in the layout display window 46 is updated in real time. The On the other hand, when the mirror image is selected by the switching button, the mirror image in the acquired image display window 43 is updated in real time.

  When the fast Fourier transform method is designated as the height variation extraction method, the parameter setting field 44 includes image processing (“FFT”, “IFFT”, “Cancel”, and “Cancel”, as shown in FIG. In order to designate the size of the filter pattern mask to be displayed on the acquired image display window 43, the designation button 441 for accepting designation of "any of registration"), the filter pattern creation window 442 used when creating the filter pattern And a histogram display window 444 for displaying a histogram of the inverse FFT transformed image displayed in the acquired image display window 43. The console terminal 28 enlarges the FFT converted image in the trimming window 431 displayed in the acquired image display window 43 and displays it on the filter pattern creation window 442. The trimming window 431 can be moved to an arbitrary position by an operator's operation. The size of the designation button 443 can be designated in stages by pulling down the operator. When the operator selects an arbitrary location in the trimming window 431 with the pointer 45, the console terminal 28 sets a mask at the selected location in the FFT converted image, and adds this mask to the filter pattern creation window 442. . When IFFT is selected by the designation button 441, the console terminal 28 displays the filter pattern displayed in the filter pattern creation window 442 in the acquired image display window 43, and displays the filter pattern together with the IFFT instruction in the system. It transmits to the control part 26. Then, the inverse FFT transformed image received from the system control unit 26 is displayed on the acquired image display window 43 and the histogram of the frequency component is displayed on the histogram display window 444. When the registration is selected by the designation button 441, the console terminal 28 transmits the filter pattern displayed in the filter pattern creation window 442 to the system control unit 26 as a height variation inspection template. On the other hand, when cancellation is selected by the designation button 441, the filter pattern and the inverse FFT converted image displayed in the acquired image display window 43 are cleared, and the contents of the filter pattern creation window 442 and the histogram display window 444 are cleared. To do.

  When the difference image method is designated as the height variation extraction method, the parameter setting field 44 designates image processing (either “cancel” or “registration”) as shown in FIG. And a slider 446 for setting a threshold value used for binarization of the difference image. When registration is selected by the designation button 445, the console terminal 28 uses the reference pattern displayed in the acquired image display window 43 as a template for height variation inspection and the binarization threshold set by the slider 446. To the system control unit 26. On the other hand, when cancel is selected by the designation button 445, the mirror image displayed in the acquired image display window 43 is cleared.

  FIG. 6 is a flowchart for explaining the height variation detection processing of the mirror projection type inspection apparatus to which the first embodiment of the present invention is applied.

When the system control unit 26 receives input of basic information of the sample 29 to be measured from the operator via the console terminal 28, the system control unit 26 reads out an inspection recipe stored in a memory or the like in association with the basic information (S201). ). Next, the system control unit 26 determines whether or not the charging control process is necessary by checking whether or not the read inspection recipe includes the set voltage for the charging control process (S202). If it is determined that the charging control process is unnecessary (NO in S203), the process proceeds to S206. On the other hand, when it is determined that the charging control process is necessary (YES in S203), the set voltage of the charging control process is read from the inspection recipe (S204). Then, the system control unit 26 controls the charge control unit 23 to apply the set voltage to the sample 29 placed on the holder 211 and perform a charge control process on the sample 29. Thereafter, the process proceeds to S206.

  In S206, the system control unit 26 reads out the alignment amounts (the first alignment amount and the second alignment amount) from the inspection recipe. Then, using the alignment amount, the coordinates of the stage 212 are corrected and alignment is performed (S207). Here, the alignment means control for adding an offset amount to the stage position by the amount of alignment stored in the recipe.

  After the above-described alignment is completed, the system control unit 26 reads a measurement location search template used to search for a measurement location of the circuit pattern formed on the sample 29 from the inspection recipe (S208). Next, the system control unit 26 controls the stage control unit 24 using the aligned coordinates of the stage 212, moves the stage 212 to the measurement location of the measurement location search template read in S208, and the electron optical system. The controller 25 is controlled to cause the electron optical system 20 to capture a low-magnification mirror image. Then, pattern matching is performed between the captured mirror image and the measurement location search template, and the coordinates of the stage 212 of the measurement location are determined (S209).

  Next, the system control unit 26 reads out a height variation inspection template for inspecting height variation at the same measurement location as the measurement location search template read out in S208 from the inspection recipe (S210). Then, the electron optical system control unit 25 is controlled to capture a mirror image for height variation inspection at the measurement location determined in S209. Then, using the captured mirror image and the height variation inspection template, the height variation at the measurement location determined in S209 is inspected (S211). Details of the height variation inspection process will be described later.

  Now, the system control unit 26 determines whether or not the height variation has been inspected using all the measurement location search templates registered in the inspection recipe. If there is a measurement location search template that is not used for the height variation inspection (NO in S212), the process returns to S208 to read this measurement location search template. On the other hand, if all the measurement location search templates have been used for the height variation inspection (YES in S212), this flow ends.

  As described above, the height variation of the sample 29 is inspected.

  Next, the height variation inspection process (S211 in FIG. 6) will be described. The height variation inspection process includes a filter pattern method using a filter pattern as a height variation inspection template and a reference pattern method using a reference pattern. Hereinafter, each of the filter pattern method and the reference pattern method will be described.

  FIG. 7 is a flowchart for explaining the filter pattern method.

  First, the image processing unit 27 converts a signal corresponding to the intensity of the mirror electrons detected by the detector 2033 of the projection optical system 203 into a coordinate signal of the stage 212 sent from the main control device 241 of the stage control unit 24. Capture in sync. Then, the captured signal is AD converted to generate a mirror image (S2111). Next, the image processing unit 27 performs Fourier transform (fast Fourier transform) on the generated mirror image to generate a Fourier transform image (S2112). Then, the generated Fourier transform image is combined with the filter pattern received from the system control unit 26 to mask the Fourier transform image (S2113). Then, the masked Fourier transform image is subjected to inverse Fourier transform to generate an inverse Fourier transform image (S2114). Then, the image processing unit 27 performs a spectrum analysis process on the inverse Fourier transform image and generates histogram distribution data indicating the result (S2115). Next, the image processing unit 27 transmits the mirror image, the inverse Fourier transform image, and the histogram distribution data to the system control unit 26. The system control unit 26 transmits these data received from the image processing unit 27 to the console terminal 28. When the variation in the height of the circuit pattern of the sample 29 is small compared to the standard sample 29, the inverse Fourier transform image is a superimposed (uniform) image, and the width of the histogram distribution data is small. On the other hand, when the height variation of the circuit pattern of the sample 29 is larger than that of the standard sample 29, the inverse Fourier transform image is a rough (non-uniform) image, and the width of the histogram distribution data is increased. Therefore, the operator can confirm the degree of variation in the height of the sample 29 by confirming the width of the histogram distribution data (for example, it can be expressed by the half width (the width of the frequency component at the half value of the frequency)). The console terminal 28 may measure the width of the histogram distribution data and display it as height variation data.

  FIG. 8 is a flowchart for explaining the reference pattern method.

  First, the image processing unit 27 converts a signal corresponding to the intensity of the mirror electrons detected by the detector 2033 of the projection optical system 203 into a coordinate signal of the stage 212 sent from the main control device 241 of the stage control unit 24. Capture in sync. Then, the captured signal is AD-converted to generate a mirror image (S2211). Next, the image processing unit 27 generates a difference image (mirror image-reference pattern) between the generated mirror image and the reference pattern (S2212). Then, the luminance value of each pixel constituting the difference image is binarized with a predetermined threshold (S2213). Then, the image processing unit 27 transmits the mirror image, the difference image, and the binarized image to the system control unit 26. The system control unit 26 transmits these data received from the image processing unit 27 to the console terminal 28. When the variation in height of the circuit pattern of the sample 29 is small compared to the standard sample 29, the binarized image becomes a flat (uniform) image, and the white area (or black area) of the black area (or white area) The area ratio with respect to (region) becomes small. On the other hand, when the variation in height of the circuit pattern of the sample 29 is large compared to the standard product of the sample 29, the binarized image becomes a rough (non-uniform) image, and the white region (or white region) of the black region (or white region) Alternatively, the area ratio to the black region) increases. Therefore, the operator can confirm the degree of height variation of the sample 29 by confirming the area ratio of the black region (or white region) to the white region (or black region) in the binarized image. The console terminal 28 may measure the ratio of the number of black pixels (or white pixels) to the number of white pixels (or black pixels), and display this as height variation data.

  FIG. 9 is a diagram illustrating an example of an inspection result GUI screen displayed on the console terminal 28 in the height variation inspection process illustrated in FIG. 6. FIG. 9A shows an example of an inspection result GUI screen by the filter pattern method, and FIG. 9B shows an example of an inspection result GUI screen by the reference pattern method.

  As shown in FIGS. 9A and 9B, the inspection result GUI screen is designated by a layout display window 51 for displaying a circuit pattern layout of a sample 29 (for example, a semiconductor wafer) and an inspection recipe. It has an image processing method display field 52 for displaying the height variation extraction method and measurement method, a processing status display window 53 for displaying the status of processing, and a result display window 54 for displaying the progress status.

  The layout display window 51 displays in real time which region of the sample 29 is being measured. The system control unit 26 transmits information on the measurement position of the sample 29 set in S209 of FIG. The console terminal 28 displays a marker 511 at the measurement location in the layout display window 51. Prior to the start of the inspection, the system control unit 26 transmits all measurement positions of the sample 29 to the console terminal 28, and the console terminal 28 transmits in real time all the measurement positions previously received from the system control unit 26. For example, the measured area 512 and the scheduled measurement area 513 may be displayed in a color-coded manner so that the measured area 512 and the scheduled measurement area 513 can be identified.

  In the image processing method display field 52, the height variation extraction method and measurement method specified in the inspection recipe are displayed. In the present embodiment, either a fast Fourier transform method (the above-described filter pattern method) or a difference image method (the above-described reference pattern method) is displayed as a method for extracting height variation. In addition, either a histogram method or an area ratio method is displayed as a method for measuring the height variation. FIG. 9A shows a case where the fast Fourier transform method is designated by the inspection recipe as the height variation extraction method, and the histogram method is designated by the inspection recipe as the height variation measurement method. B) illustrates a case where the difference image method is designated by the inspection recipe as the height variation extraction method and the area ratio method is designated by the inspection recipe as the height variation measurement method.

  In the processing status display window 53, an image acquired in the height variation inspection and a statistical value of the inspection result are displayed. When the fast Fourier transform method is specified in the inspection recipe as a method for extracting the height variation, as shown in FIG. 9A, the mirror image, the Fourier transform image of the mirror image is masked with a filter pattern, and inverse Fourier transform is performed. The average value of the transformed inverse Fourier transform image, the histogram of the inverse Fourier transform image, and the height variation data (for example, the half width of the histogram) obtained from the histogram of the measured region is displayed. Further, when the difference image method is specified in the inspection recipe as a method for extracting the height variation, as shown in FIG. 9B, the mirror image, the difference image between the mirror image and the reference pattern, and the difference image 2 are displayed. An average value of the height variation data obtained from the binarized image and the binarized image of the measured region (for example, the area ratio of the white region (or black region) to the black region (or white region)) is displayed.

  The result display window 54 displays the measurement date and time, the file name of the inspection recipe, the file name of the inspection result, the progress of measurement, and the like. The result display window 54 is provided with a designation button 541 for accepting the start and stop of the inspection from the operator. The operator can interrupt and restart the measurement at any time by operating the designation button 541.

  The first embodiment of the present invention has been described above.

  When there is a height difference in the circuit pattern of the sample 29, the mirror electrons drawn back in the vicinity of different heights draw different trajectories and form images at different focal positions according to the height difference. Therefore, it is possible to detect the height difference by observing this mirror image. According to the present embodiment, the sample 29 is compared by comparing the height difference indicated by a standard captured image (filter pattern, reference pattern) prepared in advance with the height difference indicated by the captured image of the inspection object. Can detect variations in the height direction.

Second Embodiment
Next, a second embodiment of the present invention will be described. In the present embodiment, in the first embodiment described above, fluctuations in the height direction other than the height difference of the circuit pattern of the sample 29 (for example, macroscopic height fluctuations due to warpage of the semiconductor wafer, etc.) Correct so that it does not affect.

  In the first embodiment shown in FIG. 2, the mirror projection type inspection apparatus according to this embodiment includes a standard product for height correction placed on the holder 211 separately from the sample 29, and a sample placed on the holder 211. 29 and a height sensor for measuring the height of the standard product for height correction.

  The height sensor measures the average height of an area of several μm to several mm, which is a resolution lower than the defect area that can be detected by the mirror projection inspection device in the plane of the sample 29 and the standard product for height correction. I can do it. Thereby, a macroscopic height variation of the sample 29 (for example, a warp of several hundred μm of the semiconductor wafer) can be detected. As such a height sensor, for example, an optical height inspection device (Z sensor) or the like can be used.

  FIG. 10A is a schematic diagram (a front view and a bottom view) of an example of a standard product for height correction. As shown in the figure, the standard product for height correction has a plurality of steps 50 on the surface, and a standard sample 52 is provided on each step surface 51. The height from the reference point of the holder 211 to each standard sample 52 is measured with high accuracy when the height correction standard product is attached to the holder 211. In this example, three standard samples 52 are arranged on different step surfaces 51, but the number of standard samples 52 is naturally not limited to this number.

  FIG. 10B is a schematic diagram (a front view, a bottom view, and a side view) of an example of a standard sample 52 constituting a standard product for height correction. As shown in the figure, the standard sample 52 is configured by arranging sample pieces 53 having different sizes and heights in a matrix. In this example, the sample pieces 53 having the same size are arranged for each row, and the sample pieces 53 having the same height are arranged for each column. The reason why the sample pieces 53 having different sizes are prepared is that the focus conditions of the electron optical system 20 differ if the sizes are different even at the same height. For example, when the sample 29 is a semiconductor wafer, the size of the sample piece 53 may be matched with the design rule of the semiconductor process.

  Next, the correction amount calculation process of the electron optical system 20 using the above-described standard product for height correction will be described.

  FIG. 11 is a flowchart for explaining the electron optical system correction amount calculation processing of the mirror projection type inspection apparatus to which the second embodiment of the present invention is applied.

  First, the system control unit 26 controls the charging control unit 23, the stage control unit 24, the electron optical system control unit 25, and the image processing unit 27, and has an arbitrary size among the sample pieces 53 of the standard sample 52 serving as a reference. A mirror image is taken for each of the sample pieces 53 having a thickness (S301).

  In the mirror image, the contrast and size of the defect can be arbitrarily changed by adjusting the excitation of the objective lens 2022. Therefore, the operator can adjust the excitation of the objective lens 2022 so that the desired height difference of the standard sample 52 is emphasized. That is, the excitation of the objective lens 2022 can be adjusted so that the contrast of the mirror image continuously changes according to the height of the sample piece 53 for each sample piece 53 having the same size. For example, when a semiconductor wafer having a circuit pattern half pitch of 100 nm is used as the sample 29, the objective lens 2022 is formed so that the contrast of the mirror image of the sample piece 53 having a size of 100 nm changes continuously according to the height. It is preferable to adjust the excitation.

  Next, the system control unit 26 receives the mirror images of each of the sample pieces 53 having the same size from the image processing unit 27, and these mirror images are registered in the optical conditions of the mirror image capturing, the memory or the like in advance. Are registered in association with the height information of the standard sample 52 serving as a reference and the size information of the sample piece 53 (S302).

  Next, the system control unit 26 controls the charging control unit 23, the stage control unit 24, the electron optical system control unit 25, and the image processing unit 27, and performs imaging in S301 under the same optical conditions as the standard sample 52 as a reference. A mirror image of each sample piece 53 having the same size as the sample piece 53 is taken (S303). When the captured mirror image is received from the image processing unit 27, the mirror image of these sample pieces 53 and the mirror image of each sample piece 53 of the standard sample 52 as a reference registered in S303 are displayed on the console terminal 28. Send. The console terminal 28 displays these mirror images and allows the operator to determine the necessity of adjusting the optical conditions (S304 to S308). First, the operator compares the contrast between the mirror image of the reference standard sample 52 and the mirror image of the standard sample 52 to be compared. If they are different (No in S304), the excitation of the intermediate lens 2031 is adjusted (S305). The system control unit 26 captures a mirror image under the adjusted optical conditions (S308), and returns to S304. This process is repeated until the contrast between the two becomes the same.

  Next, the operator compares the magnification and rotation of the mirror image of the standard sample 52 as a reference and the mirror image of the standard sample 52 to be compared. If they are different (No in S306), the excitation of the projection lens 2032 is adjusted (S307). The system control unit 26 captures a mirror image under the adjusted optical conditions (S308), and returns to S304. This process is repeated until both magnifications and rotations are the same.

  If the mirror image of the standard sample 52 to be compared becomes the same as the mirror image of the reference standard sample 52 in all of contrast, magnification, and rotation (both YES in S304 and S306), The optical conditions are registered in association with the height information of the standard sample 52 to be compared and the size information of the sample piece 53 registered in advance in a memory or the like (S309).

  Next, the system control unit 26 ends this flow when registration of optical conditions for all the standard samples 52 is completed (YES in S310). If not completed (NO in S310), the process returns to S303 and continues.

  12 is a diagram showing an example of an optical system adjustment GUI screen displayed on the console terminal 28 in the electron optical system correction amount calculation process shown in FIG.

  As shown in FIG. 12, the GUI screen for adjusting the optical system has a layout display window 61 showing the approximate position of the holder 211, and a sample image display for displaying images of the standard sample 52 as a reference and the standard sample 52 to be adjusted. A window 62, a parameter adjustment column 63 for setting parameters of the electron optical system 20, an image mode instruction reception column 64 for receiving an instruction to switch the image mode of the standard sample 52, the height of the standard sample 52 to be adjusted, In addition to displaying the size of the standard piece 53 and the parameter setting values of the electron optical system 20, it has a registration instruction receiving field 65 for registering these.

  The layout display window 61 is used for specifying the standard sample 51. When the operator operates a pointing device such as a mouse to operate the pointer 66 and designates the measurement location of the standard sample 52 as a reference, the console terminal 28 sends the measurement location designated by the pointer 66 to the system control unit 26. Send. The system control unit 26 controls the stage control unit 24, the optical microscope 213, and the image processing unit 27, acquires an optical image or mirror image of the measurement location designated by the pointer 66, and transmits this to the console terminal 28. . The console terminal 28 displays the acquired image data of the standard sample 52 as a reference in the sample image display window 62. Similarly, when the operator operates a pointing device such as a mouse to operate the pointer 66 and designates the measurement location of the standard sample 52 to be adjusted, the console terminal 28 performs system control on the measurement location designated by the pointer 66. It transmits to the part 26. The system control unit 26 controls the stage control unit 24, the optical microscope 213, and the image processing unit 27, acquires an optical image or mirror image of the measurement location designated by the pointer 66, and transmits this to the console terminal 28. . The console terminal 28 displays the acquired image data of the standard sample 52 to be adjusted on the sample image display window 62.

  In the sample image display window 62, the image data of the reference standard sample 52 and the image data of the standard sample 52 to be adjusted are displayed side by side. At the time of parameter adjustment of the electron optical system 20, the image data of the standard sample 52 to be adjusted is updated every time the parameter setting value in the parameter adjustment column 63 is updated.

The parameter adjustment column 63 has a scale bar for adjusting parameters of various lenses of the electron optical system 20. When the operator operates a scale bar by operating a pointing device such as a mouse, the console terminal 28 transmits parameter values of various lenses of the electron optical system 20 indicated by the scale bar to the system control unit 26. The system control unit 28 transmits the parameter value received from the console terminal 28 to the electron optical system control unit 25 and causes the electron optical system control unit 25 to adjust the electron optical system 20. Thereby, the image data of the standard sample 52 to be adjusted is adjusted. The operator can adjust the scale bar so that the image data of the standard sample 52 to be adjusted is the same as the image data of the standard sample 52 serving as a reference.

  The image mode instruction reception column 64 includes a focus button for selecting an optical image, a focus button for selecting a mirror image, and a wobbler button for receiving adjustment of the electro-optical system 20. When the operator operates a focus button for selecting an optical image by operating a pointing device such as a mouse, the console terminal 28 transmits an optical image acquisition instruction to the system control unit 26. In response to this, the system control unit 28 controls the optical microscope 213 to cause the optical microscope 213 to capture an optical image. This optical image is transmitted to the system control unit 26 via the image processing unit 27, and is transmitted from the system control unit 26 to the console terminal 28. Further, when the operator operates a focus button for selecting a mirror image by operating a pointing device such as a mouse, the console terminal 28 transmits a mirror image acquisition instruction to the system control unit 26. In response to this, the system control unit 28 controls the electron optical system 20 to cause the electron optical system 20 to capture a mirror image. This mirror image is transmitted to the system control unit 26 via the image processing unit 27, and is transmitted from the system control unit 26 to the console terminal 28. When the operator operates a pointing device such as a mouse to operate a wobbler button, the console terminal 28 activates the parameter adjustment field 63 so that adjustment of the parameters of the electron optical system 20 can be accepted from the operator.

  The registration instruction reception field 65 includes a current value button for displaying the height of the standard sample 52 to be adjusted, the size of the standard piece 53, and the parameter setting value of the electron optical system 20, and the electronic device registered immediately before. A previous value button for displaying the parameter setting value of the optical system 20, a save button for registering the height of the standard sample 52 to be adjusted, the size of the standard piece 53, and the parameter setting value of the electron optical system 20 And having. When an operator operates a pointing device such as a mouse to operate a save button, the console terminal 28 causes the height of the standard sample 52 to be adjusted, the size of the standard piece 53, and parameter setting values (optical) of the electron optical system 20 to be adjusted. Condition) is transmitted to the system control unit 26. In response to this, the system control unit 28 stores these pieces of information in a memory or the like as macroscopic height fluctuation correction information.

  Next, an inspection recipe creation process and a height variation detection process of the mirror projection type inspection apparatus to which this embodiment is applied will be described.

  In the inspection recipe creation process of the present embodiment, the height of the sample 29 is measured by the height sensor in S111 (the height variation inspection template reception / registration process) of the inspection recipe creation process of the first embodiment shown in FIG. Then, the system control unit 26 causes the electron optical system control unit 25 to connect the electron optical system 20 using the optical conditions registered in the memory or the like in association with the height and the size of the circuit pattern of the sample 29. Control and cause mirror image to be taken. Others are the same as the inspection recipe creation processing of the first embodiment shown in FIG.

  In the height variation detection process of the present embodiment, the height of the sample 29 is measured by the height sensor in S211 (height variation inspection) of the height variation detection process of the first embodiment shown in FIG. The system control unit 26 causes the electron optical system control unit 25 to control the electron optical system 20 using the optical conditions registered in the memory or the like in association with the height and the size of the circuit pattern of the sample 29. Then, a mirror image is taken. The rest is the same as the height variation detection process of the first embodiment shown in FIG.

  The second embodiment of the present invention has been described above.

  According to the present embodiment, in addition to the effects of the first embodiment described above, fluctuations in the height direction other than the height difference of the circuit pattern of the sample 29 (for example, macroscopic height fluctuations due to warpage of the semiconductor wafer, etc.) Correction can be made so as not to affect the height variation inspection result.

<Third Embodiment>
Next, a third embodiment of the present invention will be described. In this embodiment, the height difference of the circuit pattern of the sample 29 can be measured using the AFM in the first embodiment or the second embodiment.

  FIG. 13 is a schematic view of a mirror projection type inspection apparatus to which the third embodiment of the present invention is applied. As shown in the figure, the mirror projection type inspection apparatus of this embodiment is obtained by adding an AFM material chamber 31 in the first embodiment or the second embodiment shown in FIG.

  In the sample chamber 31, the AFM 32, the holder 33 for placing the sample 29, the stage 34 for moving the sample 29 in the direction perpendicular to the needle of the AFM 32, the AFM optical microscope 35, and the sample 29 from the holder 211 to the holder 33. And a stage control unit (not shown) for controlling the stage 34. Unlike the sample chamber 21, the sample chamber 31 does not need to be evacuated. The inside of the material room 31 may be an atmospheric environment.

  The AFM 32 scans the surface of the object to be inspected using a thin needle having a tip curvature of sub μm, and images the image in synchronism with the movement amount of the needle in the height direction. Measure the shape. An existing device can be used for the AFM 32. The AFM optical microscope 35 is used to search for a measurement location by the AFM 32.

  Next, an inspection process of the mirror projection type inspection apparatus to which the third embodiment of the present invention is applied will be described. The inspection process of the mirror projection type inspection apparatus of the present embodiment can be divided into an inspection process for variation in height using a mirror image and a review process for difference in height using an AFM.

  FIG. 14 (A) is a flowchart for explaining the inspection process of the height variation of the mirror projection type inspection apparatus to which the third embodiment of the present invention is applied, and FIG. 14 (B) is the third embodiment of the present invention. It is a flowchart for demonstrating the review process of height difference using AFM of the mirror projection type inspection apparatus to which the embodiment is applied.

  With reference to FIG. 14A, an inspection process for height variation will be described. First, the sample 29 is carried into the sample chamber 21 by a transport unit (not shown) and placed on the holder 211 (S401). The system control unit 26 performs the flow shown in FIG. 6 and inspects the height variation of the sample 29 (S402). Then, the inspection result is output to the console terminal 28, and the necessity of review by AFM is received from the operator via the console terminal 28 (S403).

  When the necessity of review by the AFM received from the console terminal 28 is “necessary” (YES in S403), the system control unit 26 designates the review location of the sample 29 from the operator via the console terminal 28 (for example, a mirror of the review location). Image designation) is accepted (S404). Then, the sample 29 is transported from the sample chamber 21 to the sample chamber 31 by a transport unit (not shown) (S405).

  On the other hand, if the necessity of review by the AFM received from the console terminal 28 is “No” (NO in S403), if there is a remaining sample 29 (YES in S406), the process returns to S401. If not (NO in S406), this flow ends.

  The height difference review process will be described with reference to FIG. First, the sample 29 is carried into the sample chamber 31 by a transport unit (not shown) and placed on the holder 33 (S411). The system control unit 26 controls the stage control unit (not shown) to move the stage 34, thereby causing the AFM 32 to measure the height difference of the review place received in S404 of the sample 29. Then, the measurement result is transmitted to the console terminal 28 together with the mirror image of the review place acquired in the flow of FIG. In response to this, the console terminal 28 displays the mirror image of the review location and the measurement result (S412). Note that the coordinate system of the stage 212 and the coordinate system of the stage 34 have a correspondence relationship in advance.

  Here, the adjustment of the correspondence between the coordinate system of the stage 212 and the coordinate system of the stage 34 will be described.

  In the height variation inspection by the mirror projection method, coordinate correction between the stage 212 and the sample 29 placed on the holder 211 and coordinate correction between the optical microscope 213 and the electron optical system 20 are performed. The purpose of the coordinate correction between the stage 212 and the sample 29 placed on the holder 211 is to correct a positional deviation that occurs when the sample 29 is placed on the holder 211. The coordinate correction value is obtained by the alignment process described above. (Dxiopt, dyiopt) is measured. This process needs to be performed every time the sample 29 is placed on the holder 211. On the other hand, the coordinate correction between the optical microscope 213 and the electron optical system 20 is for the purpose of measuring the relative distance between the optical microscope 213 and the electron optical system 20, and needs to be performed every time the sample 29 is replaced. Absent. This relative distance is measured when adjusting the optical microscope 213 and the electron optical system 20. The coordinate correction values (dxiopt, dyopt) between the stage 212 and the sample 29 placed on the holder 211 are used as part of the inspection recipe file of the sample 29, and the coordinate correction values between the optical microscope 213 and the electron optical system 20 ( dxins, dyins) are individually stored in the memory or the like of the system control unit 26 as adjustment data of the mirror projection type inspection apparatus.

  Even in the measurement of the height difference by the AFM, the coordinate correction between the stage 34 and the sample 29 placed on the holder 33 and the coordinate correction between the AFM optical microscope 35 and the AFM 32 are performed. The purpose of the coordinate correction between the stage 34 and the sample 29 placed on the holder 33 is to correct a positional shift that occurs when the sample 29 is placed on the holder 33, and the same process as the alignment process described above. As a result, the coordinate correction values (dxaopt, dyaopt) are measured. This process needs to be performed every time the sample 29 is mounted on the holder 33. On the other hand, the coordinate correction between the AFM optical microscope 35 and the AFM 32 is intended to measure the relative distance between the AFM optical microscope 35 and the AFM 32, and need not be performed every time the sample 29 is replaced. The relative distance is measured when adjusting the AFM optical microscope 213 and the AFM 32. The coordinate correction values (dxaopt, dyaopt) of the sample 29 mounted on the stage 34 and the holder 33 are part of the inspection recipe file of the sample 29, and the coordinate correction values (dxafm, dyafm) of the AFM optical microscope 35 and the AFM 32. ) Is individually stored in the memory or the like of the system control unit 26 as AFM adjustment data.

The stage position measurement device 241 measures the coordinates of the stage 211 as (Xi + dxiopt + dxins, Yi + dyiopt + dyins). However, the correction coordinates (Xi, Yi) of the stage 241 can be calculated by storing the correction values (dxiopt, dyiopt) and (dxins, dyins). When reviewing with AFM, the correction coordinates (Xi, Yi) of the stage 211 are set to the coordinates (Xa, Ya) of the stage 34, and the above correction values (dxaopt, dyaopt), (dxafm, dyafm) are used. 34 correction coordinates are calculated. As a result, it becomes possible to perform a review with AFM based on the coordinates of the defective portion detected by the height variation inspection.

  Now, when the review is completed, the sample 29 is unloaded from the sample chamber 31 by a transport unit (not shown) (S413). Then, the process returns to S411 and waits for a new sample 29 to be reviewed to be carried into the sample chamber 31 (S411).

  FIG. 15 is a diagram schematically showing a mirror image of a review place and a measurement result. Here, reference numeral 71 is a mirror image of the review location, reference numeral 72 is a graph (line profile) showing a measurement result of height difference by AFM in the A1-A2 section in the mirror image 71, and reference numeral 72 is a mirror image 71 It is a graph (line profile) which shows the measurement result of the height difference by AFM of a B1-B2 section. The contrast of the mirror image 71 is measured as the height difference of the surface of the sample 29 in the line profiles 72 and 73. From this measurement result, the operator can confirm the height difference of the surface of the sample 29 from which the height variation of the circuit pattern is detected by the height variation detection process.

  FIG. 16 is a diagram showing an example of the review GUI screen displayed by the console terminal 28 in the processing shown in FIGS. 14 (A) and 14 (B). FIG. 16A shows an example of the review condition setting GUI screen when setting the review condition in S404 of FIG. 14A, and FIG. 16B displays the review result in S412 of FIG. 14B. The review result display GUI screen example to be shown is shown.

  As shown in FIG. 16A, the review condition setting GUI screen includes a layout display window 81 indicating the position of the review target area, an image processing designation field 82 for designating a method for extracting and measuring a height variation. , A review condition setting field 83 for setting a review condition, and a registration setting field 84 for registering a result.

  The layout display window 81 displays a height variation inspection area 811 and an alignment location 812. In the image processing designation field 82, the result input by the operator when creating the inspection recipe is displayed as it is.

  The review condition setting field 83 includes an input field for receiving a review time (the maximum time allowed for the review) and a sampling rate (the ratio of defects to be reviewed among all defects detected in the height variation inspection process). ) Input field for receiving, sampling method (how to sample defects detected by height variation inspection processing), input field for receiving designation of AFM mode, and sampling number (number of defects of review candidates) An input field for receiving and a display field for displaying the maximum number of defects that can be reviewed (maximum number of review points) are provided.

Here, the maximum number of review points is determined by the review time, the sampling rate, and the number of samplings. When all the defects of review candidates can be reviewed at the sampling rate determined by the operator, the product of the number of samplings and the sampling rate is the maximum number of review points. On the other hand, if the console terminal 28 cannot review all sampling candidate defects within the review time at the sampling rate determined by the operator, the system control unit 26 performs sampling so that all sampling candidate defects can be reviewed within the review time. Automatically adjust the rate. When “Random” is designated as the sampling method, the system control unit 26 determines a defect to be randomly reviewed from all the defects detected in the height variation inspection process and performs the review. In addition, when “specific die” is designated, the system control unit 26 automatically adjusts the sampling rate so that the review is completed within a specified time for only the designated die. Furthermore, the command designated by this sampling method can be freely created by the operator. By appropriately selecting a sampling method, an effective review can be performed in a shorter time.

  The registration setting field 84 has a check button for accepting designation of a method for storing the inspection result of height variation and the review result by AFM. The system control unit 26 registers the inspection result of the height variation and the review result by the AFM in the memory or the like by the storage method (data format) accepted from the operator via the registration setting field 84 displayed on the console terminal 28. To do.

  As shown in FIG. 16B, the review result display GUI screen includes a layout display window 85 indicating a review position, an image processing designation field 86 for designating a height variation extraction method and a measurement method, and a measurement result. Are displayed in real time, and an attribute display window 88 for displaying attribute information.

  In the layout display window 85, a review position 851 of the sample 29 is displayed. The system control unit 26 transmits the position information under review to the console terminal 28. The console terminal 28 displays the review position 911 specified by the position information sent from the system control unit 26 on the layout display window 85. In addition, you may display on the layout display window 85 so that the area | region designated with the test | inspection recipe can be identified. The image processing designation field 86 displays the result input by the operator when creating the inspection recipe.

  The result display window 87 is provided with a selection column 871 for selecting the type of result to be displayed on the display window. With this selection field 871, the operator can select a plurality of arbitrary images from “mirror image”, “inverse FFT transformed image”, “optical image”, “histogram”, “AFM image”, and “line profile”. it can. The system control unit 26 displays the result selected by the operator in the result display window 87 of the console terminal 28 in real time. In the example shown in FIG. 16B, three types of results are displayed: an “inverse FFT transformed image” acquired by the height variation inspection process, an “optical image” by the AFM optical microscope 35, and a “line profile” by the AFM 33. ing.

  The attribute display window 88 displays the review execution date, inspection recipe file name, inspection result storage file name, and review progress. These pieces of information are transmitted from the system control unit 26 to the console terminal 28. The attribute display window 88 is provided with an instruction button for receiving an instruction to start and stop the review from the operator.

  The third embodiment of the present invention has been described above.

  According to the present embodiment, in addition to the effects of the first embodiment, it is possible to confirm the height difference of the sample 29 in which the height variation is detected.

<Fourth embodiment>
In the present embodiment, a case where the mirror projection inspection apparatus according to any one of the first to third embodiments is applied to an inline process monitor will be described.

  FIG. 17 is a diagram illustrating an example of a semiconductor device manufacturing process. In this example, each of steps S501 to S511 from the pattern formation on the substrate to the planarization after embedding in the Shallow Trench Isolation (STI) step is shown in a simplified manner. In the present embodiment, in the semiconductor manufacturing process shown in FIG. 17, the lithography process (S502 to S504) for forming a pattern and the planarization process (S509 to S511) after filling are performed in the first to third embodiments. A case will be described in which any one of the mirror projection type inspection apparatuses in the form is applied.

  The semiconductor device manufacturing process shown in FIG. 17 will be briefly described.

  First, a hard mask is formed as a mask during substrate etching (S501). This hard mask also serves as an end point determination (stopper) for the chemical and mechanical polishing (hereinafter abbreviated as CMP) process in the planarization step (S509 to S511). Next, after applying a resist (S502), the pattern is exposed using a stepper (S503), and the resist at a portion to be etched is removed by a development process (S504). Thus far, the pattern is formed on the surface of the substrate.

  Next, the portion where the resist is removed is etched to remove the hard mask (S505), and then the substrate is etched (S506). Next, after an oxide film is formed (S507), an insulating material embedding process is performed (S508).

  Finally, after a planarization process by CMP (S509), the hard mask is removed by wet etching (S510), and after cleaning (S511), the wafer is advanced to the next step.

  The case where the mirror projection type inspection apparatus according to any one of the first to third embodiments is applied to the lithography process (S502 to S504) will be described.

  In the lithography process (S502 to S504), there are problems such as mixing of fine particles (foreign matter in the film) at the time of applying the resist, scratching of the resist surface, and surface unevenness. These cause deterioration of the resolution of exposure. Therefore, in the process after applying the resist or after the development, the height variation is inspected by using the mirror projection type inspection apparatus according to any one of the first to third embodiments, so that the foreign matter in the film and the surface Unevenness can be detected. A function as a monitor of the semiconductor manufacturing process can be realized by feeding back the detection result to the resist coating process.

  The case where the mirror projection type inspection apparatus according to any of the first to third embodiments is applied to the flattening step (S509 to S511) will be described. In the present embodiment, a case will be described in which a height variation is inspected by using the mirror projection type inspection apparatus according to any one of the first to third embodiments in the cleaning step after the planarization step by the CMP process. . However, the height variation can be inspected by using the mirror projection type inspection apparatus according to any of the first to third embodiments in a planarization process other than CMP such as etch back.

  In the cleaning step S511, the remaining cleaning of the slurry used during the CMP process, scratches generated during the CMP process, and local surface irregularities due to the pattern shape of the substrate become problems. These cause the resolution degradation in the subsequent lithography process and the electrical characteristic failure in the completed device. As in the case of the lithography process, in the planarization process, the height variation is inspected using the mirror projection type inspection apparatus according to any one of the first to third embodiments. Accurate management can be realized. For example, since the remaining slurry has a cause in the cleaning process, when this is detected, the cleaning process needs to be reviewed. On the other hand, since scratches and surface irregularities are caused by the CMP process, if these are detected, it is necessary to review the CMP process.

  Therefore, by detecting these defects using the mirror projection inspection apparatus according to any one of the first to third embodiments described above and feeding back to the semiconductor manufacturing process, it is possible to improve the yield in the semiconductor manufacturing process at the development stage. Realized and mass production stage semiconductor processes can detect yield deterioration at an early stage. In addition, although resist materials and low-k materials used in the current semiconductor manufacturing process are concerned about deterioration due to electron beam irradiation, the mirror projection type inspection apparatus pulls back the electron beam very close to the wafer surface. There is an advantage that no deterioration occurs. That is, a damageless inline process monitor can be realized by using any one of the mirror projection type inspection apparatuses of the first to third embodiments.

  FIG. 18 is a diagram showing an example of a data management GUI screen displayed on the console terminal 28 in data management of inspection results. FIG. 18A is a GUI screen for analyzing the data of one inspection result. As in the development stage of the semiconductor manufacturing process, the inspection result is analyzed in detail rather than inspecting a lot of wafers. Suitable for finding process problems early. On the other hand, FIG. 18B is a GUI screen for statistically analyzing the inspection result of the same process, and is suitable for analyzing the inspection result as a process monitor of the semiconductor manufacturing process in the mass production stage. The data analysis is performed by the console terminal 28 or the system control unit 26.

  As shown in FIG. 18A, a GUI screen for analyzing data of one inspection result includes a layout display window 91 indicating the position of the inspection target area, and an inspection result designation field 92 for selecting the inspection result. And a review result display window 93 for displaying the result of the review, and an analysis result display window 94 for displaying the analysis result of the inspection result.

  In the layout display window 91, the distribution of defects as inspection results in the wafer surface is displayed. When the operator operates the selection button provided on the layout display window 91 and selects either “wafer” or “die”, the entire wafer can be displayed or one die can be displayed. it can.

  The inspection result designation field 92 accepts selection of a file of inspection results to be displayed from the operator. When a pull-down button in the inspection result designation field 92 is selected, a list of past inspection results is displayed, and the operator can arbitrarily select an inspection result.

  The review result display window 93 displays the result of the defect review together with the item selected on the review condition setting GUI screen described above. FIG. 18A shows a case where a mirror image storage and a line profile by AFM are selected as the review conditions. By selecting a pull-down button in the review result display window 93, a list of review results stored in the system control unit 26 is displayed for each ID number, and the operator selects a defect to be arbitrarily displayed from this list display. can do.

  The analysis result display window 94 displays information necessary for the operator to analyze the inspection result. In FIG. 18A, the relationship between the half width of the histogram and the occurrence frequency is displayed as information used for analysis. Here, the histogram half width indicates the histogram half width of the inspection result obtained by the above-described filter pattern method, and indicates the variation in contrast of the acquired image region. The occurrence frequency indicates how many regions with the same half width of the histogram exist. Furthermore, when the function of performing defect determination is enabled from the mirror image obtained at the time of inspection, the distribution of each defect is displayed in a superimposed manner. The operator can arbitrarily set the defect determination threshold by sliding the marker on the graph or selecting it with the pull-down button while looking at the analysis result display window 94 and the review result display window 93. The defect occurrence location is reflected in the distribution of the defects in the layout display window 91 on the wafer surface.

  As shown in FIG. 18B, the GUI screen for statistically analyzing the inspection result includes a file structure display window 95 for displaying the file structure of the inspection result, and an inspection result display window 96 for displaying the inspection result. And an analysis result display window 97 for displaying the analysis result, and a statistics / analysis method selection column 98 for selecting a statistics / analysis method.

  In the file structure display window 95, the inspection result files stored in the memory or the like of the system control unit 26 are hierarchically displayed. In FIG. 18B, folders are provided for each product type and process, such as a root layer 951, a product type folder 952, and a process folder 953. An inspection result file 954 of the same product type and the same process is stored under the process folder 953. The operator can select a desired inspection result file 954 to be analyzed from the file structure display window 95. In FIG. 18B, “A” to “D” of the inspection defect file 954 are selected, and the analysis result is displayed in the analysis result display window 97.

  The inspection result display window 96 is provided with a pull-down button for selecting one of the inspection result files 954 selected in the file structure display window 95. When the operator selects an inspection result file with a pull-down button, the contents are displayed.

  The analysis result display window 97 displays the statistics / analysis results of the inspection result file 954 selected in the file structure display window 95 in accordance with the statistics / analysis method selected in the analysis method selection field 98 described later. In FIG. 18B, two types of statistics / analysis results are displayed separated into an upper stage and a lower stage. In the upper part, the defect classification results are displayed in a table for each inspection result, and in the lower part, the relationship between the inspection date and the defect classification is displayed in a graph.

  The statistic / analysis method selection column 98 is provided with a pull-down button and a check list for designating the statistic / analysis method of the inspection result file to be displayed on the statistic / analysis result display window 97. The operator can arbitrarily determine the number of windows on the statistics / analysis result display screen 97 with a pull-down button. As a result display method, either a table display or a graph display can be selected with a check box.

  The embodiment of the present invention has been described above.

  The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope of the gist. For example, the GUI screen described with reference to FIGS. 16 and 18 may be applied to each device described in the first to fourth embodiments. Further, the GUI screen described in each of the above embodiments is merely an example, and the GUI screen may be configured by other elements. Alternatively, each control unit of the mirror projection type inspection apparatus of each of the above embodiments may be realized by one computer system, or may be realized by linking a plurality of computer systems via a network. May be.

20: Electron optical system, 21: Sample chamber, 23: Charge control unit, 24: Stage control unit, 25: Electron optical system control unit, 26: System control unit, 27: Image processing unit, 28: Console terminal, 29: Sample: 31: Sample chamber, 32: AFM, 33: Holder, 34: Stage, 35: Optical microscope for AFM, 201: Irradiation optical system, 202: Imaging optical system, 203: Projection optical system, 211: Holder, 212 : Stage, 213: optical microscope, 231: charging control electrode, 241: main control device, 242: retarding power supply, 243: stage position measuring device, 244: stage drive device, 251: main control device, 252: irradiation optical system Control device, 253: Projection optical system control device, 271: Image forming device, 272: Image processing arithmetic device

Claims (5)

An inspection apparatus for inspecting variation in the height direction of a sample,
Holding means for holding the sample;
Charging processing means for charging the sample held in the holding means;
Voltage applying means for applying a voltage to the sample held in the holding means;
An electron optical system that irradiates an electron beam toward a sample to which a voltage is applied by the voltage applying unit, and forms an image of mirror electrons pulled back near the surface of the sample;
Image processing means for image processing a mirror image obtained by imaging the mirror electrons,
The image processing means includes
A process of generating a difference image representing a difference between a mirror image obtained by imaging the mirror electrons and a mirror image of the standard product;
Each pixel constituting the difference image is binarized to generate a binarized image, the ratio of the binary pixels constituting the binarized image is calculated, and the ratio of the pixels of the binarized image is calculated. And a process of outputting the information as variation information in the height direction of the sample according to the difference image. The inspection apparatus according to claim 1,
The image processing means includes
An inspection apparatus further comprising a process of calculating and outputting a ratio of white pixels and black pixels constituting the binarized image. The inspection apparatus according to claim 1 or 2,
Storage means for storing the optical conditions of the electron optical system in association with the height information of the sample;
Measuring means for measuring the height of the sample at a resolution lower than the region of the sample having a variation in the height direction that can be identified from the information according to the difference;
An inspection apparatus, further comprising: a setting unit that sets an optical condition stored in the storage unit in association with the height measured by the measurement unit in the electron optical system. The inspection apparatus according to claim 3,
It also has a standard product for height correction consisting of sample pieces with different heights,
The holding means is
Further holding the standard product for height correction,
In the storage means,
Optical information when a mirror image of each sample piece of the standard product for height correction held by the holding means is captured by the electron optical system is stored in association with the height of each sample piece. Characteristic inspection device. The inspection apparatus according to any one of claims 1 to 4,
AFM holding means for holding the sample;
Inspection apparatus characterized by further having a AF M measuring the height direction of the shape of the sample held by the holding means for the AFM.

Images