Defect inspection apparatus

Abstract

A defect inspection apparatus 100 comprises: a processor 400; a secondary detection optical system OS2 having an optical device; and a memory 401 that stores related information between operation conditions for the optical device and defect types of samples 8. The processor 400 identifies the defect type of a sample 8, sets an operation condition for the optical device by looking up the identified defect type in the related information, generates a first image of a defect region of the sample 8 including a defect by irradiating, with a primary charged particle beam, the defect region in a state where the optical device is operated under said operation condition, and outputs information about the defect of the sample 8 using the first image.

Description

Defect Inspection Equipment

[0001] The present invention relates to a defect inspection apparatus, and more particularly to a defect inspection apparatus for irradiating a sample with an electron beam and inspecting the sample for defects.

[0002] Semiconductor devices are manufactured by repeatedly transferring patterns formed on wafers using photomasks through lithography and etching processes. To improve yield and stabilize the manufacturing process, it is essential to detect defects through in-line wafer inspection, quickly analyze the defects, and take measures to address the defects.

[0003] In inline wafer inspection, defects are typically detected using an optical defect inspection system, and then re-detected and observed in detail using a review system placed after the defect inspection system. In order to quickly translate the inspection results into countermeasures for defects, it is important to have technology that can review a large number of defects at high speed and classify them by their cause.

[0004] However, as semiconductor devices become smaller, the size of defects also becomes smaller, making it difficult for optical review devices to perform high-speed review and to obtain high-resolution images.

[0005] For this reason, review devices (review SEMs) using scanning electron microscopes (SEMs) have been commercialized. Review SEMs enable high-speed reviews and provide high-resolution images. Furthermore, with the increasing automation of observation work in semiconductor mass production lines, review SEMs are equipped with automatic defect review (ADR) as a function for automatically collecting images of defect locations within a sample.

[0006] Patent Document 1 discloses a method in which an inspection device uses ADR to output defect positions, captures an image based on the coordinates of the defect positions, and detects defects from the image.The method also describes a method in which an image of an area where the same structural pattern as the defect portion is formed is used as a reference image, and the image of the defect portion is compared with the reference image to detect the defect.

[0007] In recent years, as semiconductor devices have become increasingly miniaturized, the number of locations requiring inspection and measurement has increased, creating a demand for high-speed review equipment. In addition, as semiconductor devices have become increasingly three-dimensional, with deep holes and trenches, there is a growing demand for improved visibility of a wide variety of defects.

[0008] Patent Document 2 discloses a method for determining the defect position with high accuracy under low magnification conditions at the imaging position of the observation object, and then switching the electron beam conditions to low current and imaging the defect at high magnification, thereby enabling high-speed imaging while maintaining high resolution.

[0009] Patent Document 3 discloses a technology for improving the efficiency of defect detection by optimizing the voltage conditions of electrodes for controlling charging from a histogram of an SEM image and determining inspection conditions in order to increase the contrast due to the distribution of the charging state of a structure pattern.

[0010] In order to improve the visibility of various structural patterns and various defects, it is effective to use multiple detectors to detect signal electrons emitted from a sample with different exit elevation angles or exit energies. Patent Document 4 describes a technology for acquiring an image that emphasizes the unevenness information of the sample shape by applying an electric field that accelerates signal electrons near a wafer and discriminating and detecting the exit elevation angle and azimuth angle of signal electrons generated from the sample by irradiating it with an electron beam.

[0011] Patent Document 5 discloses a classifier that compares a defect image with a reference image, calculates the detection rate of wiring pattern edges within a defect area, and determines whether the defect exists on the wiring pattern or below the wiring pattern based on the edge preservation rate.

[0012] Japanese Patent Application Laid-Open No. 2001-189358 Japanese Patent Application Laid-Open No. 2007-248360 Japanese Patent Application Laid-Open No. 2005-292076 Japanese Patent Application Laid-Open No. 2006-228999 Japanese Patent Application Laid-Open No. 2007-184565

[0013] For example, consider the case where a defect that is difficult to see at the bottom of a deep trench and a minute pattern defect on the surface of the sample exist on the same sample, and these defects need to be reviewed. When imaging the defect at the bottom of the trench, it is possible to highlight the defect by discriminating the signal electrons emitted perpendicular to the sample and detecting signal electrons emitted at low elevation angles. Furthermore, depending on the type of defect, it may be effective to discriminate the emission energy of the signal electrons in addition to discriminating the emission elevation angle of the signal electrons.

[0014] However, discriminating and detecting signal electrons in this way reduces the yield of signal electrons, so to obtain an image with a high S / N (ratio of signal electrons to noise), it is necessary to increase the electron beam dose and imaging time.

[0015] On the other hand, for minute pattern defects, there is no need to selectively discriminate signal electrons, and the yield of signal electrons can be increased. Therefore, visibility can be improved by a high-quality defect image with a high S / N ratio. Furthermore, if visibility is sufficient, the increased yield of signal electrons allows the electron beam irradiation dose to be reduced, and the imaging time can also be shortened.

[0016] In this way, it is necessary to image defects under imaging conditions of signal electrons that are optimized for each defect type.

[0017] In Patent Document 2, high-speed imaging is possible while maintaining high resolution by switching the electron beam current conditions between low-magnification imaging and high-magnification imaging at the imaging position of the observation object. However, when imaging a sample that has various types of structural defects, the visibility of the defects cannot be optimized by simply switching the current conditions.

[0018] Furthermore, if the setting of imaging conditions is dependent on the experience of the operator, the reproducibility of the examination cannot be improved, and there is also the problem that highly sensitive imaging conditions cannot necessarily be set.

[0019] As described above, there is a need for a defect inspection device that can maintain high defect visibility for various types of defects present on a sample such as a semiconductor wafer and can acquire high-resolution images in a short imaging time. Other objects and novel features will become apparent from the description of this specification and the accompanying drawings.

[0020] A defect inspection apparatus according to one embodiment includes a processor that generates an image based on secondary charged particles obtained by irradiating a sample with a primary charged particle beam, a secondary detection optical system having an optical device that acts on the secondary charged particles, and a memory that stores first association information between operating conditions of the optical device and defect types of the sample. The processor identifies the defect type of the sample, sets first operating conditions of the optical device by referring to the identified defect type in the first association information, irradiates a first region of the sample containing the defect with the primary charged particle beam while the optical device is operating under the first operating conditions, generates a first image of the first region, and outputs defect information of the sample using the first image.

[0021] According to one embodiment, it is possible to provide a defect inspection apparatus that can maintain high visibility of defects present on a sample and can acquire high-resolution images in a short imaging time.

[0022] Fig. 4 is a schematic diagram showing a defect inspection device in embodiment 1. Fig. 5 is a flow diagram showing an inspection method in embodiment 1. Fig. 6 is a flow diagram showing determination of optimal imaging conditions in embodiment 1. Fig. 7 is an explanatory diagram showing an example of linking the imaging conditions in Fig. 3. Fig. 8 is an explanatory diagram showing an example of linking the imaging conditions in Fig. 3. Fig. 9 is a flow diagram showing an inspection method in embodiment 2. Fig. 10 is a flow diagram showing determination of optimal imaging conditions in embodiment 2. Fig. 11 is a schematic diagram showing a defect inspection device in embodiment 3.

[0023] Hereinafter, embodiments will be described in detail with reference to the drawings. In all drawings for explaining the embodiments, components having the same functions are designated by the same reference numerals, and repeated explanations thereof will be omitted. In the following embodiments, explanations of the same or similar parts will not be repeated unless particularly necessary.

[0024] (Embodiment 1) <Defect Inspection Apparatus> A defect inspection apparatus 100 according to embodiment 1 will be described below with reference to FIG. 1 . The defect inspection apparatus 100 is a charged particle beam apparatus, such as a review apparatus (review SEM) using a scanning electron microscope. However, the defect inspection apparatus 100 may also be an imaging apparatus using charged particles such as ions. The term "charged particle beam apparatus" broadly includes apparatuses that capture images of a sample using a charged particle beam. The defect inspection apparatus 100 also includes a system in which defect inspection apparatuses 100 are connected via a network, or a composite apparatus made up of multiple charged particle beam apparatuses.

[0025] In this specification, a "sample" refers to a semiconductor wafer on which various structural patterns are formed. The "sample" includes, for example, a semiconductor substrate on which multiple p-type impurity regions and multiple n-type impurity regions are formed, semiconductor elements such as transistors, resistors, or capacitors formed on the semiconductor substrate, and multiple insulating films and multiple wirings formed on the semiconductor substrate.

[0026] In this specification, the term "structural pattern" includes a groove pattern formed in a semiconductor substrate, a plurality of conductor patterns used for electrodes of semiconductor elements, wiring patterns, via patterns connecting each wiring pattern, via patterns connecting conductor patterns and wiring patterns, and via patterns connecting a semiconductor substrate and a wiring pattern.

[0027] In this specification, the "defect type" includes foreign matter present on the structure pattern, defects in the structure pattern, deformations in the structure pattern, and dimensional abnormalities in the structure pattern.

[0028] In this specification, a "defect image" refers to an image that is the subject of defect inspection (an image to be inspected), and includes not only images of real defects but also images of defect candidates or pseudo defects. Furthermore, a "reference image" in this specification refers to an image that is used for comparison with a "defect image" for defect extraction, and is an image of a structural pattern that is the same as the structural pattern in which a defect exists, and is an image of an area that is presumed to be free of defects (a normal area).

[0029] In this specification, "secondary electrons" refer to electrons that have an energy of 50 eV or less and are emitted from a sample by irradiation with an electron beam. In addition, "backscattered electrons" refer to electrons that have an energy of 50 eV or more and are emitted from a sample by irradiation with an electron beam. In addition, in this specification, "signal electrons" refer to electrons that are emitted from a sample by irradiation with an electron beam, and are a combination of "secondary electrons" and "backscattered electrons."

[0030] As shown in FIG. 1, the defect inspection apparatus 100 includes an image pickup device 101 , a processor 400 , a memory 401 , a power supply unit 402 , an input / output unit 403 for an external storage medium, a user interface unit 404 , and a bus 500 .

[0031] The imager 101 irradiates an electron beam (primary charged particle beam) and captures an image. The memory 401 is a storage device including a magnetic disk or semiconductor memory, and stores various information used in the defect inspection device 100. The power supply unit 402 is a source of voltage or current used in the defect inspection device 100. The external storage medium input / output unit 403 inputs and outputs information to and from an external storage medium electrically connected to the defect inspection device 100. The user interface unit 404 is an operating device including a keyboard, mouse, display, etc., and is operated by the user.

[0032] The image capture device 101, processor 400, memory 401, power supply unit 402, input / output unit 403 for external storage media, and user interface unit 404 are electrically connected via a bus 500, and can exchange information with one another.

[0033] The processor 400 is an arithmetic processing unit including a plurality of semiconductor devices such as CPUs. The processor 400 is electrically connected to and controls the image pickup device 101, memory 401, power supply unit 402, external storage medium input / output unit 403, and user interface unit 404. In other words, each control and operation performed in the defect inspection apparatus 100 is performed by the processor 400.

[0034] The imager 101 includes a primary charged particle optical system OS1, a secondary detection optical system OS2, and a separator 5. The primary charged particle optical system OS1 has an electron beam axis (optical axis) 200. In the primary charged particle optical system OS1, an electron beam (primary charged particle beam) is irradiated onto the sample 8 with the electron beam axis 200 as a reference. The secondary detection optical system OS2 has a signal electron axis 300 that is at a different angle from the electron beam axis 200. In the secondary detection optical system OS2, signal electrons (secondary charged particles) emitted from the sample 8 are detected with the signal electron axis 300 as a reference.

[0035] The primary charged particle optical system OS1 has an electron source (charged particle source) 1, an accelerating electrode 2, a first condenser lens 3, a second condenser lens 4, a deflector 6, an objective lens 7, a movable stage 9, and an axially symmetric detector 11. A sample 8 can be mounted on the movable stage 9. The axially symmetric detector 11 can detect signal electrons. When inspecting the sample 8, the sample 8 is mounted on the movable stage 9 when an electron beam is irradiated onto the sample 8.

[0036] The secondary detection optical system OS2 has an axial detector 14 and an axially symmetric detector 15 that can detect signal electrons, and has a converging lens 13 for signal electrons, an energy discrimination electrode 16, an energy discrimination electrode 17, and an electrode 18 as optical devices that act on the signal electrons.

[0037] The separator 5 is a deflector for separating the electron beam from the signal electrons. The separator 5 may be an electrostatic deflector or a magnetic deflector, a Wien filter that combines an electrostatic deflector and a magnetic deflector, or a multi-stage deflection type deflector that combines three or more deflectors.

[0038] An electron beam emitted from an electron source 1 is accelerated by an accelerating electrode 2. The current amount of the electron beam and the magnification of the primary charged particle optical system OS1 are controlled by a first condenser lens 3 and a second condenser lens 4. The electron beam is then focused onto a sample 8 by an objective lens 7 into a microscopic spot of a few nanometers. The electron beam is also scanned two-dimensionally on the sample 8 by a deflector 6.

[0039] Signal electrons emitted from the sample 8 by irradiation with the electron beam travel in the opposite direction to the electron beam and pass through the objective lens 7. In order to improve the resolution of the electron beam and the ability to distinguish the azimuth angles of the signal electrons, a negative voltage may be applied to the sample 8, or a positive voltage may be applied to some electrodes of the objective lens 7.

[0040] Among the signal electrons, the high elevation angle component 10 of the reflected electrons propagates and is detected by an axially symmetric detector 11. By separating the axially symmetric detector 11 into multiple detectors in the azimuth direction, it is also possible to discriminate the azimuth angle information of the signal electrons.

[0041] The axially symmetric detector 11 is configured in an annular shape, and a hole is provided in the center of the axially symmetric detector 11. Therefore, low elevation angle components 12 of secondary electrons and backscattered electrons among the signal electrons pass through the hole in the center of the axially symmetric detector 11. Thereafter, the low elevation angle components 12 are separated from the electron beam by the separator 5 and deflected toward the signal electron axis 300. The secondary detection optical system OS2 is configured to detect the signal electrons after separation by the separator 5.

[0042] The signal electrons deflected to the signal electron axis 300 are converged by the convergent lens 13 for the signal electrons and detected by the on-axis detector 14 or the axisymmetric detector 15. The axisymmetric detector 15 is configured in an annular shape, and a hole is provided in the center of the axisymmetric detector 15. By separating the axisymmetric detector 15 in the azimuthal direction into multiple detectors, it is also possible to discriminate azimuthal angle information of the signal electrons.

[0043] After the signal electrons are separated from the electron beam, they are controlled by the optical devices of the secondary detection optical system OS2, so that the conditions of the signal electron trajectory can be controlled independently of the electron beam performance (electron beam trajectory, electron beam spot size, electron beam current amount, etc.) As an optical device of the secondary detection optical system OS2, an electrostatic deflector or a magnetic deflector may be used to deflect the signal electron trajectory (not shown).

[0044] The energy discriminatory electrodes 16 and 17 are configured as mesh electrodes. When a voltage is applied to the energy discriminatory electrodes 16 and 17, the energy discriminatory electrodes 16 and 17 act as potential barriers. Therefore, the energy discriminatory electrodes 16 and 17 can discriminate the energy of the signal electrons detected by the on-axis detector 14 or the axisymmetric detector 15.

[0045] The signal electron trajectory 301 is the trajectory when the converging lens 13 for the signal electrons operates under weak converging conditions and most of the signal electrons pass through the hole in the center of the axially symmetric detector 15 and are converged onto the on-axis detector 14. By increasing the yield of signal electrons in the on-axis detector 14, it becomes possible to obtain an image with a high S / N ratio with the on-axis detector 14.

[0046] The signal electron trajectory 302 is the trajectory when the converging lens 13 for the signal electrons operates under a strong converging condition and the signal electrons are spread on the plane of the axially symmetric detector 15. In this way, the low elevation angle components are detected by the on-axis detector 14. Moreover, the high elevation angle components are discriminated from the low elevation angle components and detected by the axially symmetric detector 15.

[0047] The axisymmetric detector 11, the on-axis detector 14, and the axisymmetric detector 15 may be, for example, a phosphor, an Everhart-Thornley (ET) detector which is a combination of a light guide and a photomultiplier tube, or a semiconductor detector. The signal electrons detected by these detectors are subjected to waveform shaping or analog-to-digital conversion processing by an analog circuit provided in the defect inspection apparatus 100, and generated as a digital image. That is, the processor 400 generates an image based on the signal electrons obtained by irradiating the sample 8 with an electron beam.

[0048] <Inspection Method> The inspection method for the sample 8 performed using the defect inspection device 100 will be described below with reference to the flow chart of Fig. 2. Figs. 3 to 5 will also be used as necessary.

[0049] As shown in FIG. 2, first, in step S100, various inspection conditions are input using the user interface unit 404. The inspection conditions include the electron beam current, electron beam energy, and the field of view (magnification) for imaging defects. Normally, combinations of various inspection conditions are compiled into a database as an inspection condition file (recipe) and stored in memory 401, so the user can set the inspection conditions simply by selecting a recipe. The memory 401 also stores various structural patterns formed on the sample 8.

[0050] The inspection conditions also include inspection information of the sample 8 that has been output in advance by another inspection device, such as an optical defect inspection device, different from the defect inspection device 100. The inspection information of the sample 8 includes, for example, defect coordinates, defect size, the result of defect type classification by the other inspection device, inspection conditions for the other inspection device, and defect images by the other inspection device.

[0051] In step S101, the sample 8 to be inspected is placed on the movable stage 9.

[0052] In step S102, the processor 400 sets the electron beam irradiation conditions for each component based on the inspection conditions input in step S100. Here, the electron beam irradiation conditions particularly include the acceleration voltage of the electron beam, the current amount of the electron beam, and the spot diameter and focal depth on the sample 8. In other words, the electron beam irradiation conditions include the acceleration voltage of the accelerating electrode 2, and the focusing conditions of the first condenser lens 3, the second condenser lens 4, and the objective lens 7.

[0053] Next, an image of a predetermined location on the sample 8 is acquired, and the focus and astigmatism are adjusted from that image. At the same time, the height of the sample 8 is determined by a height detector (not shown), and the correlation between the height information and the focusing conditions for the electron beam is determined. As a result, when acquiring subsequent images, the focusing conditions are automatically adjusted based on the height detection results, without having to perform focusing each time. This makes it possible to acquire successive images at high speed.

[0054] In step S103, steps S104 to S109 are performed if there are any defects that have not been inspected among the many defects present in the sample 8. That is, steps S103 to S110 are repeated until inspection of all defects present in the sample 8 is completed.

[0055] In step S104, the imaging conditions determined as standard conditions are set. The imaging conditions here include the operating conditions of the separator 5, the operating conditions of each optical device of the secondary detection optical system OS2, the number of images to be captured, the scanning speed, the imaging field of view, etc. These imaging conditions are set so that their influence on the electron beam irradiation conditions set in step S102 can be ignored.

[0056] In step S105, the movable stage 9 is moved so that the coordinates of the defect are included in the imaging field of view, and a defect image is captured at a low magnification with a field of view of about 3 μm to 10 μm. That is, the processor 400 generates a defect image of the defect area by irradiating an electron beam onto a defect area in the sample 8 that includes a defect, while operating each optical device of the secondary detection optical system OS2 under the imaging conditions set in step S104.

[0057] In step S106, similar to step S105, a reference image is captured at a low magnification with a field of view of approximately 3 μm to 10 μm. While the sample 8 contains various structural patterns, the reference image is an image of a structural pattern identical to the structural pattern in which the defect exists, and is an image of an area that does not contain any defects. In the semiconductor chip manufacturing process, multiple identical structural patterns are contained within a semiconductor substrate. Therefore, a reference image can be obtained by capturing an image of a semiconductor chip adjacent to a chip containing a defect, or a similar structural pattern formed near the defect.

[0058] That is, the processor 400 generates a reference image of the defect-free area by irradiating the electron beam onto a defect-free area of ​​the sample 8, in which a structural pattern similar to that of the defect area imaged in step S105 is formed. The defect area and the reference image are stored in the memory 401.

[0059] Instead of capturing a reference image of a defect-free area, a pseudo reference image may be created by, for example, calculating a weighted average using multiple defect images captured of locations where the same structural pattern is formed. Alternatively, a pseudo reference image may be synthesized from defect images by utilizing the periodicity of the structural pattern. Alternatively, a pseudo reference image created by SEM simulation based on design information of a circuit corresponding to the observation location may be used.

[0060] These pseudo reference images are stored in the memory 401 before the start of the inspection. Therefore, the imaging in step S106 may not be performed, and the pseudo reference images stored in advance in the memory 401 may be used as reference images.

[0061] In step S107, optimal imaging conditions for imaging the defect at high magnification are determined. Fig. 3 is a flow diagram for determining the imaging conditions. As shown in Fig. 3, the defect image captured in step S105 and the reference image captured in step S106 are input to a defect classifier. The defect classifier may be the classifier disclosed in Patent Document 5.

[0062] In step S200, the defect image and the reference image are input to a defect classifier. The defect classifier extracts features of the defect image and the reference image and classifies the defect type. In step S201, the defect classifier outputs the defect type. Note that the defect classifier is part of the functions of the processor 400. That is, the processor 400 identifies the defect type by comparing the reference image with the defect image.

[0063] In step S202, imaging conditions corresponding to the defect type are output. For example, when defect type A is output, imaging condition A is automatically selected, and when defect type B is output, imaging condition B is automatically selected.

[0064] 4 and 5 each show an example of the association of imaging conditions corresponding to the defect types in FIG.

[0065] 4 shows a database of imaging conditions corresponding to defect types created in advance by a user. This database stores operating conditions of the optical devices of the secondary detection optical system OS2 and irradiation conditions of the electron beam for various imaging conditions such as imaging condition A and imaging condition B.

[0066] In order to create the database, the user determines the conditions for each defect type in advance using the sample to be observed or the sample for creating the database. For example, if defect type A is a foreign substance present at the bottom of a deep trench, imaging condition A is set to a condition for selectively acquiring backscattered electrons at a low elevation angle, which have less interaction with the sidewall of the trench, from among the signal electrons generated from the bottom.

[0067] In this case, the voltage and current values ​​for the signal electron converging lens 13, the energy discriminator electrode 16, and the energy discriminator electrode 17 are recorded in a database as the operating conditions of the optical device. When signal electrons are selectively detected and conditions that result in a low yield are used, conditions such as slowing down the scanning speed of the electron beam or increasing the number of images to be acquired are recorded in the database. This allows images with a high S / N ratio to be obtained.

[0068] In the example of FIG. 5, the correspondence between the defect type and the operating conditions of the optical device is calculated in advance using a map of the exit elevation angle and exit energy of the signal electrons as parameters, and a database is created.

[0069] The image visibility can be set in advance for each expected defect type using an SEM image simulator that uses the Monte Carlo method, etc. In this case, it is possible to optimize which region of the signal electrons should be discriminated and detected in the map of the signal electron exit elevation angle and exit energy to obtain an image with a high S / N ratio for the target defect.

[0070] The correspondence between the discrimination of regions in the map of the signal electron emission elevation angle and emission energy and the optical device conditions can also be calculated in advance by using an electromagnetic field calculation using the finite element method or an electron trajectory calculation simulator.

[0071] 5 illustrates, for each defect type, standard conditions, conditions for defect type A, and conditions for defect type B. The conditions for defect type A are conditions for capturing an image with emphasis on reflected electrons. The conditions for defect type B are conditions for separately detecting secondary electrons and reflected electrons.

[0072] First, under standard conditions, backscattered electrons at low elevation angles are detected by the axial detector A100 (detection area A103). Other signal electrons are detected by the axially symmetric annular detector A200 (detection area A203). Here, the energy discrimination electrode A101 is operated so that signal electrons of 50 eV or less are not detected by the axial detector A100.

[0073] Under the conditions for defect type A, the axial detector A100 detects backscattered electrons at a low elevation angle, and the axisymmetric annular detector A200 also detects only backscattered electrons. Here, the energy discrimination electrode A101 is operated so that signal electrons of 50 eV or less are not detected by the axial detector A100, and the energy discrimination electrode A201 is operated so that signal electrons of 50 eV or less are not detected by the axisymmetric annular detector A200.

[0074] Under the conditions of defect type B, a converging lens for signal electrons is used to detect all secondary electrons and backscattered electrons at low elevation angles in the axial detector A100. Other backscattered electrons are detected in the axisymmetric annular detector A200. Generally, the yield of secondary electrons is overwhelmingly greater than the yield of backscattered electrons at low elevation angles, so secondary electron information is emphasized in the axial detector A100, and backscattered electron information is emphasized in the axisymmetric annular detector A200. Note that, in addition to the exit elevation angle and exit energy of the signal electrons, the exit azimuth angle may also be used as a parameter for the correspondence relationship between the defect type and the operating conditions of the optical device.

[0075] The memory 401 stores information relating to the defect type in the sample 8, the operating conditions of each optical device in the secondary detection optical system OS2, and the irradiation conditions of the electron beam. The processor 400 identifies the defect type in the sample 8 in step S107, and sets the operating conditions of the optical devices and the irradiation conditions of the electron beam in step S108 by referring to the identified defect type in the associated information in the memory 401. That is, the processor 400 determines the exit elevation angle of the signal electrons emitted from the sample 8 by controlling each optical device in the secondary detection optical system OS2 in accordance with the defect type.

[0076] In step S108, the operating conditions of each optical device of the secondary detection optical system OS2 determined in step S107 and the irradiation conditions of the electron beam are set, i.e., the voltage and current values ​​of each optical device used in the secondary detection optical system OS2 are set, and any of the scanning speed, scanning direction, imaging field of view, and number of scanning frames of the electron beam used in the primary charged particle optical system OS1 is set.

[0077] In step S109, an image of the defect is captured at a high magnification. The processor 400 generates a defect image of the defect area by irradiating an electron beam onto a defect area of ​​the sample 8 that includes a defect while operating each optical device of the secondary detection optical system OS2 under the operating conditions set in step S108. The processor 400 irradiates the electron beam under the irradiation conditions set in step S108.

[0078] The defect area in step S109 is an area included in the defect area in step S105, and the defect image in step S109 is captured at a higher magnification than the defect image in step S105. The processor 400 outputs defect information of the sample 8 using the defect image in step S109.

[0079] In step S110, if there is a defect that has not been inspected among the many defects present on the sample 8, the process returns to step S103, and steps S104 to S109 are performed again.

[0080] As described above, according to the first embodiment, defect imaging can be performed automatically using imaging conditions optimized for each defect type. That is, by using the defect inspection apparatus 100 of the first embodiment, high defect visibility can be maintained for various types of defects present on the sample 8, and high-resolution images can be acquired in a short imaging time. This makes it easy to identify the cause of the defect. Furthermore, since the imaging conditions can be set automatically without relying on the user's experience, the reproducibility of the inspection results is high.

[0081] 6 and 7, a defect inspection device 100 according to a second embodiment will be described below. Note that the following description will mainly focus on differences from the first embodiment, and descriptions of points that overlap with the first embodiment will be omitted.

[0082] 6, in the second embodiment, steps S100 to S104, S109, and S110 are the same as those in the first embodiment, but capturing a defect image in step S105 and capturing a reference image in step S106 are omitted. Also, in the second embodiment, steps S307 and S308 are performed instead of steps S107 and S108 in the first embodiment.

[0083] In step S307, the optimum imaging conditions for imaging the defect at high magnification are determined. Fig. 7 is a flow chart for determining the imaging conditions.

[0084] 7, in the second embodiment, the imaging conditions are optimized using an algorithm for determining the imaging conditions. First, in step S400, input is made to the algorithm for determining the imaging conditions.

[0085] Inputs to the imaging condition determination algorithm include, for example, layout design information of the sample 8, such as a GDS file, or inspection information of the sample 8 output by an inspection device other than the defect inspection device 100, such as an optical defect inspection device. The inspection information of the sample 8 includes at least the defect positions and the results of defect type classification by the other inspection device. The inspection information of the sample 8 also includes other information such as the defect size, the inspection conditions by the other inspection device, and defect images by the other inspection device. The layout design information and inspection information of the sample 8 described above are stored in the memory 401 via the input / output unit 403 of the external storage medium.

[0086] In step S401, the imaging condition determination algorithm outputs the operating conditions of each optical device of the secondary detection optical system OS2 and the irradiation conditions of the electron beam in response to the input of step S400 so that the S / N of the defect image and the defect detection rate are increased depending on the defect type.

[0087] For example, a model using reinforcement learning, which is a type of machine learning, is prepared in advance using a sample to be observed or a learning sample, with the S / N ratio of the defect image and the defect detection rate as rewards. Furthermore, a map of defect types and the exit elevation angle and exit energy of signal electrons, as described in FIG. 5, may be used in an intermediate layer between the input and output in the algorithm for determining the imaging conditions.

[0088] In recent years, improvements in the stopping accuracy of movable stages and high-resolution (high-pixel) imaging have made it possible to capture defects with a single defect image capture. Therefore, it may be possible to identify the defect's location and type using inspection information for the sample 8 output by another inspection device. By effectively utilizing this inspection information, the defect type can be identified in a short time. Furthermore, by using layout design information, the defect's location can be grasped more accurately, making it easier to predict the defect type in the structural pattern corresponding to the defect's location.

[0089] The algorithm for determining the imaging conditions is one of the functions of the processor 400. That is, the processor 400 identifies the defect type based on the layout design information and the inspection information of the sample 8.

[0090] In step S308, the operating conditions of each optical device of the secondary detection optical system OS2 determined in step S307 and the irradiation conditions of the electron beam are set, i.e., the voltage and current values ​​of each optical device used in the secondary detection optical system OS2 are set, and any of the scanning speed, scanning direction, imaging field of view, and number of scanning frames of the electron beam used in the primary charged particle optical system OS1 is set.

[0091] Thereafter, steps S109 and S110 are carried out in sequence in the same manner as in the first embodiment.

[0092] As described above, in the second embodiment as well, it is possible to automatically capture images of defects under imaging conditions optimized for each defect type. Furthermore, in the second embodiment, it is possible to omit capturing the defect image in step S105 and capturing the reference image in step S106, thereby reducing the defect inspection time compared to the first embodiment.

[0093] On the other hand, if the information on the defect position or defect type output by the other inspection device is unclear, the accurate defect position or defect type can be derived by comparing the defect image in step S105 with the reference image in step S106, as in embodiment 1. In other words, embodiment 1 is superior to embodiment 2 in determining the defect position or defect type with high accuracy.

[0094] (Embodiment 3) A defect inspection device 100 according to embodiment 3 will be described below with reference to Fig. 8. In the following description, differences from embodiment 1 will be mainly described, and descriptions of points that overlap with embodiment 1 will be omitted.

[0095] In the first embodiment, the separator 5 separates the electron beam from the signal electrons. In the third embodiment, as shown in FIG. 8 , the signal electrons are not separated. The signal electron axis of the secondary detection optical system OS2 is the same as the electron beam axis 200. The secondary detection optical system OS2 in the third embodiment has an axially symmetric detector 11 and, as an optical device, a converging lens 210 for the signal electrons. The converging lens 210 for the signal electrons is provided at the object point position of the objective lens 7, and is, for example, an electrostatic Einzel lens or an electromagnetic lens combining a magnetic pole and a coil.

[0096] 8 illustrates a central trajectory 201 of the electron beam in addition to the electron beam axis 200. The central trajectory 201 of the electron beam is extracted from the electron source 1 by the acceleration electrode 2, forms a first cross (not shown) by the first condenser lens 3, and forms a second cross (not shown) by the second condenser lens 4. The central trajectory 201 of the electron beam is focused by the objective lens 7 into a minute spot of several nm on the sample 8, with the second cross as the object plane and the sample 8 as the image plane.

[0097] After signal electrons 220 are generated from the sample 8, the signal electrons 220 travel in a direction opposite to the electron beam while spreading, and are detected by the axisymmetric detector 11. In addition, since a hole is provided in the center of the axisymmetric detector 11 as a passage for the electron beam, low elevation angle components of the signal electrons 220 pass through the hole in the axisymmetric detector 11 and are not detected by the axisymmetric detector 11.

[0098] The converging lens 210 for the signal electrons is provided between the objective lens 7 and the axially symmetric detector 11, and is provided at the converging position of the electron beam, so that the trajectory conditions of the signal electrons 220 can be controlled independently of the performance of the electron beam.

[0099] A signal electron trajectory 221 is the trajectory of the signal electrons 220 when the converging lens 210 for signal electrons is made to act as a lens. Compared to when the converging lens 210 for signal electrons is not made to act, the signal electrons 220 are converged toward the center of the axially symmetric detector 11 and are detected by the axially symmetric detector 11.

[0100] That is, only components with lower elevation angles or higher energy components are detected than the trajectory of the signal electrons 220. For example, by setting the converging lens 210 for the signal electrons under conditions in which secondary electrons with a narrow energy range are converged within the center of the axially symmetric detector 11, only reflected electrons can be detected by the axially symmetric detector 11. Therefore, the processor 400 determines the exit elevation angle of the signal electrons emitted from the sample 8 by controlling the converging lens 210 for the signal electrons in accordance with the defect type.

[0101] The inspection method of the third embodiment can be applied to the flowchart of Fig. 2 of the first embodiment or the flowchart of Fig. 6 of the third embodiment. As a method for identifying the defect type, the method described in Figs. 3 to 5 of the first embodiment or the method described in Fig. 7 of the second embodiment can be used. In this way, also in the third embodiment, the defect can be automatically imaged under imaging conditions optimized for each defect type.

[0102] The present invention has been specifically described above based on the above embodiment, but the present invention is not limited to the above embodiment and can be modified in various ways without departing from the spirit of the present invention.

[0103] REFERENCE SIGNS LIST 1 Electron source (charged particle source) 2 Acceleration electrode 3 First condenser lens 4 Second condenser lens 5 Deflector (separator) for separating signal electrons 6 Deflector 7 Objective lens 8 Sample 9 Movable stage 10 High elevation angle component 11 Axisymmetric detector 12 Low elevation angle component 13 Converging lens for signal electrons 14 On-axis detector 15 Axisymmetric detector 16 Energy discrimination electrode 17 Energy discrimination electrode 18 Electrode 100 Defect inspection device 101 Imager 200 Electron beam axis (optical axis) 201 Central trajectory of electron beam 210 Converging lens for signal electrons 220 Signal electron 221 Signal electron trajectory when using a converging lens for signal electrons 300 Signal electron axis 301 Signal electron trajectory 302 Signal electron trajectory when using a converging lens for signal electrons 400 Processor 401 Memory 402 Power supply unit 403 Input / output unit for external storage medium 404 User interface unit 500 Bus A100 Axial detector A101 Energy discrimination electrode A103 Detection area in map of exit elevation angle and exit energy of signal electrons A200 Axisymmetric annular detector A201 Energy discrimination electrode A203 Detection area in map of exit elevation angle and exit energy of signal electrons A300 Secondary electrons A301 Backscattered electrons OS1 Primary charged particle optical system OS2 Secondary detection optical system

Claims

1. A defect inspection apparatus comprising: a processor that generates an image based on secondary charged particles obtained by irradiating a sample with a primary charged particle beam; a secondary detection optical system having an optical device that acts on the secondary charged particles; and a memory in which first association information between operating conditions of the optical device and a defect type of the sample is stored, wherein the processor identifies the defect type of the sample, sets first operating conditions of the optical device by referring to the identified defect type in the first association information, irradiates the primary charged particle beam onto a first region of the sample containing a defect while the optical device is operating under the first operating conditions, thereby generating a first image of the first region, and outputting defect information of the sample using the first image.

2. A defect inspection apparatus as described in claim 1, wherein the memory stores a structural pattern formed on the sample, and the processor irradiates the primary charged particle beam onto a second region of the sample including the first region to generate a second image of the second region, and identifies the defect type by comparing the second image with a third image in which a structural pattern similar to that of the second region is captured.

3. A defect inspection apparatus as described in claim 2, wherein the third image is pre-stored in the memory or is generated by irradiating the primary charged particle beam onto a third region of the sample in which a structural pattern similar to that of the second region is formed.

4. A defect inspection apparatus as described in claim 1, wherein the memory stores layout design information of the sample and inspection information of the sample output by an inspection device other than the defect inspection apparatus, and the processor identifies the defect type based on the layout design information and the inspection information.

5. A defect inspection apparatus according to claim 4, wherein the inspection information includes at least the result of classification of the defect type in the other inspection apparatus.

6. A defect inspection device according to claim 1, further comprising a separator for separating the primary charged particle beam and the secondary charged particles, and the secondary detection optical system is configured to detect the secondary charged particles after separation by the separator.

7. A defect inspection device as described in claim 6, wherein the secondary detection optical system has a signal electron axis that is at a different angle from the optical axis of the primary charged particle beam, and the separator deflects the secondary charged particles from the optical axis of the primary charged particle beam to the signal electron axis.

8. A defect inspection apparatus according to claim 1, wherein the secondary charged particles are secondary electrons and reflected electrons, and the optical device is a lens, an electrode or a deflector.

9. A defect inspection apparatus as described in claim 1, wherein the secondary detection optical system has a plurality of detectors capable of detecting the secondary charged particles, and the processor determines the emission elevation angle of the secondary charged particles emitted from the sample by controlling the optical device according to the defect type.

10. A defect inspection apparatus as described in claim 1, further comprising an objective lens for focusing the primary charged particle beam on the sample, the secondary detection optical system having a detector capable of detecting the secondary charged particles and a focusing lens for the secondary charged particles, the focusing lens for the secondary charged particles being provided between the objective lens and the detector and at a focusing position of the primary charged particle beam, and the processor determining the emission elevation angle of the secondary charged particles emitted from the sample by controlling the focusing lens in accordance with the defect type.

11. A defect inspection apparatus as described in claim 1, wherein the defect types include foreign matter present on a structural pattern formed on the sample, defects in the structural pattern, deformations in the structural pattern, and dimensional abnormalities in the structural pattern.

12. A defect inspection device comprising: a processor that generates an image based on secondary charged particles obtained by irradiating a sample with a primary charged particle beam; and a memory in which first association information between irradiation conditions of the primary charged particle beam and a defect type of the sample is stored, wherein the processor identifies the defect type of the sample; sets first irradiation conditions of the primary charged particle beam by referring to the identified defect type in the first association information; irradiates a first region of the sample including a defect with the primary charged particle beam under the set first irradiation conditions, thereby generating a first image of the first region; and outputting defect information of the sample using the first image.

13. A defect inspection apparatus according to claim 12, wherein the first irradiation condition is any one of the scanning speed, scanning direction, imaging field of view, and number of scanning frames of the primary charged particle beam.

Images