Charged particle beam device and processor system

Abstract

To provide a technique capable of reducing deterioration caused by a dark pulse of a back scattered electron (BSE) detector.SOLUTION: A charged particle beam device comprises a plurality of BSE detectors each detecting a BSE from a specimen, and a controller. The controller acquires within a period a first peak time of a first peak included in an output signal of a first BSE detector in the plurality of BSE detectors and a second peak time of a second peak included in an output signal of a second BSE detector other than the first BSE detector, determines that the first peak is caused by the BSE in a case where there is a second peak for which a time difference between the first peak time and the second peak time is within a threshold, and determines that the first peak is caused by a dark pulse in a case where there is no second peak for which the time difference is within the threshold.SELECTED DRAWING: Figure 12

Description

[Technical Field]

[0001] The present invention relates to the technology of charged particle beam devices. [Background technology]

[0002] Charged particle beam devices detect signal electrons such as secondary electrons generated by irradiating a sample with a charged particle beam of primary electrons, and obtain a microscopic image of the sample. Examples of charged particle beam devices include scanning electron microscopes (sometimes written as SEMs). SEMs are used as evaluation and measurement tools for semiconductor devices.

[0003] In recent years, semiconductor device structures have become increasingly miniaturized and three-dimensional, and the evaluation values ​​required by semiconductor device manufacturers are becoming more diverse. As device structures become more three-dimensional, there is a need to measure with high precision the dimensions of the shapes of structures such as holes and trenches on semiconductor substrate surfaces in order to improve yield. Dimensions that need to be measured include the dimensions in the height and depth directions of the sample (e.g., vertical direction) and the bottom dimensions in the in-plane direction of the sample (e.g., horizontal direction).

[0004] In measurements using the SEM described above, when a charged particle beam of primary electrons is irradiated onto the surface of a sample, signal electrons with various energies are emitted in various directions due to the interaction between the primary electrons and the sample. These signal electrons carry different information about the sample depending on their emission energy and angle. The discrimination and detection of these signal electrons is essential for a wide range of measurements. The signal electrons carry information in the sense that such information can be obtained by examining the signal electrons detected as images or waveforms.

[0005] Generally, signal electrons emitted with an energy of 50 eV or less are called secondary electrons (sometimes written as SE). Signal electrons emitted with an energy greater than that of secondary electrons and closer to that of primary electrons are called backscattered electrons (sometimes written as BSE). These signal electrons are distinguished.

[0006] Secondary electrons are sensitive to the surface shape and electrical potential of a sample, and are effective for measuring the dimensions of surface structures such as the pattern width of semiconductor device structures. However, secondary electrons are difficult to detect and measure for three-dimensional structures such as holes and grooves because they are absorbed by the sidewalls of the structure and do not easily escape from the structure.

[0007] On the other hand, backscattered electrons contain information about the composition and three-dimensional shape of the sample, and can provide information about the three-dimensional structure, the difference in composition between the surface and bottom of the sample, etc. In addition, because backscattered electrons have high energy, they easily escape from the bottom of hole or groove structures by penetrating the sidewalls, making it possible to detect and measure signal electrons from the bottom of the structure.

[0008] As an example of prior art, a backscattered electron detector (BSE detector) that detects backscattered electrons is described in Patent Document 1. The large angle detector 25 in Fig. 2 includes four segments 251 to 254. [Prior art documents] [Patent documents]

[0009] [Patent Document 1] US Patent Publication No. 2019 / 0088444 Summary of the Invention [Problem to be solved by the invention]

[0010] A backscattered electron detector (sometimes referred to as a BSE detector) is composed of, for example, a scintillator and a photomultiplier tube (photomultiplier tube) or a semiconductor photodetector. The scintillator is a device that detects backscattered electrons and converts them into photons. The photomultiplier tube or semiconductor photodetector is a device that converts the generated photons into a backscattered electron signal (sometimes referred to as a BSE signal) that has a pulse waveform.

[0011] The waveform of the BSE signal is a pulse-like signal with a fast rise time of about 10 ns and a slow fall time of about 100 ns. The semiconductor photodetectors used in BSE detectors generate dark current noise even when there are no incident photons. This dark current noise is also called dark pulse, dark noise, or dark count.

[0012] This dark pulse is difficult to distinguish from the waveform of the BSE signal. For this reason, in conventional SEMs, the dark pulse generated by the BSE detector deteriorates the signal-to-noise ratio (sometimes referred to as S / N) of the detected image. This deterioration in S / N makes it impossible to obtain detailed information on three-dimensional structures such as holes and grooves, reduces measurement accuracy, and increases instrumental variability.

[0013] An object of the present invention is to provide a technique for the above-mentioned charged particle beam device that can reduce deterioration of a BSE detector due to dark pulses, in other words, a technique that can improve measurement accuracy. [Means for solving the problem]

[0014] A representative embodiment of the present disclosure has the following configuration: A charged particle beam device according to the embodiment includes a plurality of BSE detectors that detect backscattered electrons (BSE) from a sample, and a controller, and in order to determine whether a first peak included in an output signal of a first BSE detector among the plurality of BSE detectors is caused by BSE or a dark pulse, the controller acquires a first peak time of the first peak within a period of time, acquires a second peak time of a second peak included in an output signal of a second BSE detector other than the first BSE detector among the plurality of BSE detectors within the period of time, and determines that the first peak is caused by BSE if there is a second peak whose time difference between the first peak time and the second peak time is within a threshold, and determines that the first peak is caused by a dark pulse if there is no second peak whose time difference between the first peak time and the second peak time is within the threshold. [Effects of the Invention]

[0015] According to a representative embodiment of the present disclosure, with regard to the technology of the charged particle beam device, it is possible to reduce deterioration of the BSE detector due to dark pulses, in other words, to improve measurement accuracy, etc. Problems, configurations, effects, etc. other than those described above are described in the description of the embodiment of the invention. [Brief explanation of the drawings]

[0016] [Figure 1] 1 shows an example of the configuration of an entire system (semiconductor measurement system) including a charged particle beam device (for example, SEM) according to the first embodiment. [Figure 2] 1 shows an example of the configuration of the main body of a charged particle beam device (for example, SEM) according to a first embodiment. [Figure 3] In the first embodiment, a schematic explanatory diagram is shown regarding scanning of a sample with a primary electron beam of an SEM. [Figure 4] In the first embodiment, a schematic explanatory diagram of the arrangement of a plurality of BSE detectors in a BSE detector of an SEM is shown. [Figure 5]In the first embodiment, an example of the configuration of a BSE detector will be shown. [Figure 6] In the first embodiment, a configuration example of a BSE detection circuit will be shown. [Figure 7] In the first embodiment, an example of the configuration of a processor system that is a controller for an SEM is shown. [Figure 8] In the first embodiment, an example of the configuration of a processing unit in a controller will be shown. [Figure 9] In the first embodiment, an example of the structure of the sample is shown. [Figure 10] In the first embodiment, an example of emission of secondary electrons and reflected electrons in response to incidence of primary electrons on the structure of a sample will be shown. [Figure 11] In the first embodiment, an example of the configuration of the dark pulse processing unit will be shown. [Figure 12] In the first embodiment, an example of a BSE detection signal of multiple channels is shown. [Figure 13] In the first embodiment, an example of comparison of combinations of pulse signals between channels will be shown. [Figure 14] In the first embodiment, the processing flow of the controller is shown. [Figure 15] In the first embodiment, examples of data and information to be saved will be shown. [Figure 16] In the second embodiment, a configuration example of a dark pulse processing unit will be shown. [Figure 17] In the second embodiment, an example of a signal in a period of a plurality of frames in a BSE detection signal of a plurality of channels will be shown. [Figure 18] In the second embodiment, an example will be shown in which signals over a plurality of frame periods are combined into one. [Figure 19] In the second embodiment, an example of comparison of combinations of pulse signals between channels will be shown. [Figure 20] In the third embodiment, an example will be shown in which signals over a plurality of frame periods in a BSE detection signal of a single channel are combined into one signal. [Figure 21] In the fourth embodiment, an example of the configuration of the processing unit will be shown. [Figure 22]In the fourth embodiment, an example of the structure on the sample surface and its relationship with a four-channel BSE detector will be shown. [Figure 23] In the fourth embodiment, an example of the detection rate of a four-channel BSE detector according to the aspect ratio of the structure is shown. [Figure 24] In the fourth embodiment, an example of correction by weighting the determination result will be shown. [Figure 25] In the fourth embodiment, the processing flow of the controller is shown. [Figure 26] 10 shows an example of a GUI screen that can be applied to each embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0017] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. In the drawings, identical parts are generally designated by the same reference numerals, and repeated explanations will be omitted. In the drawings, the representation of components may not represent their actual positions, sizes, shapes, ranges, etc., in order to facilitate understanding.

[0018] For the sake of explanation, when describing processing by a program, the program, functions, processing units, etc. may be described as the main components, but the main hardware components are the processor, or a controller, device, computer, system, etc. that is configured with the processor, etc. A computer executes processing according to a program read into memory using resources such as memory and communication interfaces as appropriate, through the processor. This realizes functions, processing units, etc. The processor is configured, for example, with semiconductor devices such as a CPU or GPU. Processing is not limited to software program processing, and can also be implemented using dedicated circuits. Dedicated circuits such as FPGAs, ASICs, and CPLDs can be used.

[0019] The program may be pre-installed as data on the target computer, or may be distributed as data from a program source to the target computer. The program source may be a program distribution server on a communication network, or a non-transitory computer-readable storage medium (e.g., a memory card). The program may be composed of multiple modules. The computer system may be composed of multiple devices. The computer system may be composed of a client-server system, a cloud computing system, an IoT system, etc. Various data and information may be composed of structures such as tables and lists, for example, but are not limited to these. Expressions such as identification information, identifiers, IDs, names, and numbers are interchangeable.

[0020] [Solution, etc.] The charged particle beam device of the embodiment makes it possible to discriminate and detect dark pulses and BSE signals, which may occur in a BSE detector.

[0021] The charged particle beam device of the embodiment is a system that uses multiple BSE detection systems (sometimes referred to as channels) for detecting BSE, and includes, for example, multiple BSE detectors (in other words, multiple BSE detection units) as BSE detectors. Each BSE detector includes, for example, a scintillator and a semiconductor photodetector.

[0022] Depending on the shape and structure of a target sample such as a semiconductor wafer, multiple BSEs are generated according to the number of primary electrons irradiated within a certain period of time (τ). Therefore, in a system using multiple BSE detection systems, photons are incident on each semiconductor photodetector of the multiple BSE detectors almost simultaneously. As a result, the multiple BSE detection systems receive a pulse-like BSE signal with a fast rise time of, for example, about 10 ns and a slow fall time of, for example, about 100 ns.

[0023] On the other hand, dark pulses, which are noise, occur randomly in response to the reception of the BSE signal. In the present embodiment, attention is focused on the randomness of these dark pulses. A system including a charged particle beam device according to the present embodiment estimates and determines that a pulse signal is caused by BSE, i.e., a BSE signal, if the pulse signal is received by, for example, multiple BSE detectors within a certain time τ. On the other hand, if a pulse signal is received by, for example, only a single BSE detector within the certain time τ, the pulse signal is estimated and determined to be caused by a dark pulse, i.e., a dark pulse. The present system then removes the estimated and determined dark pulse from the detection signal.

[0024] The controller of the charged particle beam device of the embodiment checks the output signals of the multiple C BSE detectors, in other words, the pulse waveforms of the detection signals, for the same pixel corresponding to the target position on the sample surface among all the multiple C BSE detectors, within a time T. If pulse signals are received or appear almost simultaneously in the output signals of a first number (referred to as N1, a set value in the range of 2 to C) or more of all the multiple C BSE detectors, the controller infers or determines that the pulse signals are caused by BSE, in other words, are BSE signals.

[0025] For a pixel of interest, if pulse signals are not received or appear substantially simultaneously in the output signals of a first number N1 or more of the BSE detectors within a time T among the multiple C output signals from the multiple C BSE detectors, in other words, if pulse signals are received or appear substantially simultaneously in the output signals of BSE detectors less than the first number N1, the controller estimates and determines that the pulse signal is caused by a dark pulse, in other words, that it is a dark pulse. For example, in the first embodiment, if a pulse signal is received from only one BSE detector, it is determined to be a dark pulse.

[0026] The time T is a determination target period corresponding to the aforementioned fixed time τ, and may be, for example, a one-pixel period corresponding to the target pixel. The time T may be a value set by design or user setting. The time T may be a period covering a range of multiple pixels including the target pixel as the time center, or may be a one-frame period corresponding to an image including the target pixel.

[0027] In other words, the pulse signal is a peak waveform, and is expressed by a peak position and a peak time.

[0028] Furthermore, the above determination has been described as a determination having a condition for determining a BSE signal. Conversely, this determination condition can also be described as a condition for determining a dark pulse. In that case, it becomes as follows: The controller of the charged particle beam device of the embodiment determines, for a target pixel, whether pulse signals are received and appear almost simultaneously in the output signals of less than a first number N1 of BSE detectors among the multiple C output signals from the multiple C BSE detectors within a time T, and if such a condition is met, it infers and determines that the pulse signal is caused by a dark pulse, in other words, that it is a dark pulse.

[0029] In the above determination, more specifically, whether the pulse signals are received and appear almost simultaneously can be determined by, for example, determining whether the time difference between the peak positions of the pulse signals from each BSE detector is within a threshold value. The threshold value for the time difference between the peak positions can also be set as one of the determination reference values, either by design or by user setting. In other words, the determination is made as follows: If the number of BSE detector output signals whose time difference between the peak positions of the pulse signals is within the threshold value is equal to or greater than a first number N1, the corresponding pulse signal is determined to be a BSE signal.

[0030] Furthermore, the first number N1, which is one of the criteria for discriminating between the BSE signal and the dark pulse, can be a value set by the design or by the user. In the first embodiment, for example, C=4 and N1=2, but N1 may also be set to 3 or 4. For example, when C=4 and N1=4, the determination can be described as follows: For a target pixel, the controller of the charged particle beam device of the embodiment determines that a pulse signal is a BSE signal if, within time T, pulse signals are received and appear almost simultaneously in the output signals of all of the multiple C BSE detectors from all of the multiple C BSE detectors; and determines that a pulse signal is a dark pulse if pulse signals are received and appear only in the output signals of less than the multiple C BSE detectors.

[0031] <First Embodiment> A charged particle beam device according to a first embodiment of the present disclosure will be described using FIGS. 1 to 14 and the like. In the following first embodiment and the like, an example will be described in which the charged particle beam device is applied to an SEM. The present invention is not limited to this, and the embodiments can be similarly applied to other types of devices as long as they are configured to include a BSE detector. The following also describes a case in which an SEM is used to measure dimensions such as height, depth, and bottom of a three-dimensional structure on the surface of a semiconductor device sample, such as a hole or groove, for example, to measure the CD (critical dimension) of the pattern. The present invention is not limited to this, and the embodiments can be similarly applied to the observation, evaluation, inspection, etc. of the sample.

[0032] [Semiconductor measurement system] FIG. 1 shows a semiconductor measurement system as an example of the configuration of an entire system including a charged particle beam device 1 according to the first embodiment. In the semiconductor measurement system of FIG. 1, the charged particle beam device 1, a client terminal 5, a manufacturing execution system (MES) 6, and a defect inspection device 7 are connected to a communication network 9 such as a LAN. The charged particle beam device 1 is, for example, an SEM, and includes a main body 2 and a controller 100. The controller 100 is a processor system that controls the main body 2 and executes processing. A user such as an operator operates the controller 100 to operate the SEM 1. Although the example shows a case where the MES 6 and the defect inspection device 7 are external devices to the charged particle beam device 1, the present invention is not limited to this.

[0033] The client terminal 5 is an information processing terminal device such as a PC that is connected to a communication network 9 and can communicate with the controller 100 of the SEM 1, etc., via the communication network 9. A user such as an operator can operate the client terminal 5 to access the controller 100, etc., and use the functions of the SEM 1, etc., remotely.

[0034] The MES 6 is a system that executes and manages the manufacture of semiconductor devices, and has, for example, design information for semiconductor devices that are samples and management information for manufacturing flows. The defect inspection device 7 inspects semiconductor devices that are samples for defects and stores the inspection results as defect detection information. The defect detection information includes information indicating the positions of defects detected on the sample surface. The charged particle beam device 1 may refer to the defect detection information from the defect inspection device 7 and observe the positions of defects on the sample surface indicated by the defect detection information. Other examples of external devices include semiconductor manufacturing equipment such as an etching device. The controller 100 may appropriately refer to necessary data and information from an external device such as the MES 6, or may output the data and information to the external device.

[0035] [SEM] 2 shows an example of the configuration of the main body 2 of the SEM 1. The main body 2 includes a column 104, which is an electron microscope column, a stage mechanism 106, an electron gun 101, a deflector 108, an objective lens 107, an SE detector 115, and four BSE detectors 110A to 110D as a BSE detector 110. The optical axis of the electron gun 101 is arranged on the Z axis, which corresponds to the vertical direction.

[0036] A sample 3 is placed and fixed on a stage, which is a sample table of the stage mechanism 106. The movement of the stage of the stage mechanism 106 is driven and controlled by the controller 100. The stage is movable at least in the horizontal direction, which is the direction of the X axis and Y axis, but is not limited to this.

[0037] An electron gun 101, a deflector 108, an objective lens 107, and the like are arranged inside a column 104, which is in a vacuum environment. A primary electron beam is emitted from the electron gun 101 by primary electrons 102 under conditions of voltage Vx and current Ix. The emitted primary electron beam travels along a beam optical axis pointing downward along the Z axis. The trajectory of the primary electron beam is adjusted by the deflector 108, and the primary electron beam is focused on the surface of the sample 3 by the objective lens 107. At this time, a negative voltage is applied to the sample 3, and the primary electrons 102 of the primary electron beam collide with the sample 3 with less energy than when they were generated by the electron gun 101. The collision of the primary electrons generates backscattered electrons 105 and secondary electrons 103 from the sample 3.

[0038] [scanning] FIG. 3 is a schematic diagram illustrating scanning of the surface of a sample 3 with a primary electron beam. The surface of the sample 3, in other words, the front or top surface, is defined as an XY plane consisting of the X and Y axes. To acquire information about the sample 3, primary electrons 102 are sequentially scanned across the surface of the sample 3, as shown in FIG. 3. This example illustrates line-sequential scanning, in which line scanning along the X axis is repeated in the Y axis direction. Multiple target positions for measurement, etc., exist on the XY plane of the sample 3, and each target position corresponds to a pixel in the detected image. Each pixel is also referred to as a target pixel. For example, pixel 30 at a certain target position is represented by two-dimensional position coordinates (x, y) in the detected image. Furthermore, pixel 30 at the target position has time information corresponding to the time of scanning. Note that the primary electron beam is generally irradiated from the Z axis, which is perpendicular to the XY plane of the sample 3 on the stage, but this is not limiting. For example, a configuration in which the optical axis of the beam irradiation is set in a direction oblique to the Z axis, in other words, the tilt direction, is also possible.

[0039] The BSE detector 110 in Fig. 2 is a detector that mainly detects backscattered electrons 105. The BSE detector 110 has four channels, in other words, four detection systems, and four BSE detectors, 110A, 110B, 110C, and 110D. Note that Fig. 2 shows the arrangement positions of the four BSE detectors 110A to 110D in a schematic manner, and in reality, the four BSE detectors 110A to 110D are arranged in a plane at the same height position along the Z axis, as shown in Fig. 4.

[0040] [Deployment of multiple BSE detectors] FIG. 4 is a schematic explanatory diagram of an example of the arrangement of four BSE detectors 110A-110D, showing the XY plane as viewed from above at the horizontal cross section of a column 104. The column 104 and other components have, for example, a cylindrical shape, symmetrical with respect to the Z axis. As shown in FIG. 4, the four BSE detectors 110A-110D are arranged at positions in the east-west, north-south, and east-west directions, spaced 90 degrees apart on a circle with a certain radius centered on point O corresponding to the Z axis, on a plane at the same height. BSE detector 110A is arranged in the north (N) position as the first channel ch1. BSE detector 110B is arranged in the south (S) position as the second channel ch2. BSE detector 110C is arranged in the west (W) position as the third channel ch3. BSE detector 110D is arranged in the east (E) position as the fourth channel ch4. The number and arrangement positions of the BSE detectors 110 are not limited to these.

[0041] Let C be the number of channels in the BSE detector 110. C is the number of channels in the BSE detection system; in this example, C=4. Let F be the number of scanning frames. F is the number of frames scanned, as shown in Figure 3, for each target pixel at the target position. In this example, F is a number greater than or equal to 2. Let P be the number of pixels in each frame. P is the total number of pixels in the image corresponding to the frame. Of the number of channels C, let c be the channel of interest. Of the number of scanning frames F, let f be the frame of interest, or in other words, the order from the first frame. Of the number of pixels P, let p be the pixel of interest, or in other words, the order from the first pixel. 1≦c≦C, 1≦f≦F, 1≦p≦P. Let G be a certain pixel. Then, in Figure 3, pixel G in a certain order in a certain channel c, a certain frame f, can be expressed as (Gc,f,p). For example, (G2,15,50) is the 50th pixel in the 15th frame of the second channel ch2.

[0042] 2, backscattered electrons 105 and secondary electrons 103 generated from a sample 3 by the collision of a primary electron 102 fly within a column 104 according to their respective emission energies and emission angles. The SE detector 115 is a detector that mainly detects the secondary electrons 103. The SE detector 115 is disposed at a position where it is easy to capture the secondary electrons 103, for example, at a position higher than the height position of the BSE detector 110. The BSE detector 110 is a detector that mainly detects the backscattered electrons 105. Four BSE detectors 110A to 110D that make up the BSE detector 110 are disposed at positions where it is easy to capture the backscattered electrons 105, for example, at a certain height position.

[0043] The four BSE detectors 110A to 110D are arranged in the column 104 at equal distances from the beam axis and perpendicular to each other, as shown in FIG. 4 . With this arrangement, backscattered electrons 105 generated on the surface of the sample 3 are input to the four BSE detectors 110 of ch1 to ch4 with approximately equal probability. In this BSE detector 110 configuration, when the voltage Vx and current Ix of the electron gun 101 are sufficiently high, the number of generated backscattered electrons 105 is sufficiently large, and the backscattered electrons 105 are input to and received by all four-channel BSE detectors 110. On the other hand, when the voltage Vx and current Ix of the electron gun 101 are low or when the position on the sample 3 irradiated with the primary electrons 102 has a structure such as a deep hole or groove, the backscattered electrons 105 are input to and received by the BSE detectors 110 with fewer channels, for example, two or three channels.

[0044] [BSE detector] FIG. 5 shows the configuration of each BSE detector 110 in the first embodiment. In the first embodiment, each BSE detector 110 includes a scintillator 51 and a semiconductor photodetector 52. Although FIG. 5 shows the configuration inside the BSE detector 110A of the first channel ch1, the BSE detectors of the other channels have a similar configuration. Each BSE detector 110 is driven and controlled based on a control voltage 112 (112a to 112d) from the controller 100. When a backscattered electron 105 collides with the scintillator 51 of each BSE detector 110, the backscattered electron 105 is converted into a photon 105b. The converted photon 105b is converted into a BSE detection signal 111 by the semiconductor photodetector 52 of each BSE detector 110.

[0045] The BSE detection signals 111 (111a to 111d) output from the BSE detectors 110 (110A to 110D) are transmitted to the BSE detection circuits 211 (211A to 211D). The BSE detection circuit 211 has BSE detection circuits 211A to 211D corresponding to the respective channels.

[0046] [BSE detection circuit] FIG. 6 shows an example of the configuration of the BSE detection circuit 211 (211A to 211D). FIG. 6 shows the configuration of the BSE detection circuit 211A of the first channel ch1 (North), but the BSE detection circuits of the other channels have the same configuration. The BSE detection circuit 211 has, in order from the input side, an I / V conversion circuit 201, an ADC 202, and a digital signal processing circuit 205. The I / V conversion circuit 201 converts a current signal corresponding to the input detection signal 111 into a voltage signal. The ADC 202 converts the analog signal, which is the voltage signal, into a digital signal.

[0047] The digital signal processing circuit 205 includes a waveform shaping circuit 203 and a peak detection circuit 204. The waveform shaping circuit 203 shapes the waveform of the digital signal into a waveform that makes peaks easier to detect in the subsequent peak detection circuit 204. Specifically, since the waveform of the digital signal based on the BSE detection signal 111 has a slow falling edge compared to its rising edge, as described above, this shaping is performed to speed up the falling edge. The peak detection circuit 204 detects peaks corresponding to the pulse waveform based on the shaped waveform signal. The peaks are expressed and identified, for example, by peak times, or in other words, peak positions. The signal 301 (301a to 301d) output from the BSE detection circuit 211 is transmitted to the controller 100 and processed by a processing unit 300 (FIG. 8) of the controller 100.

[0048] Meanwhile, secondary electrons 103 are captured by SE detector 115 in Fig. 2. SE detector 115 converts secondary electrons 103 into detection signal 121 and outputs it. The output detection signal 121 is transmitted to SE detection circuit 116. SE detection circuit 116 obtains signal 310, which is an SE detection signal, based on detection signal 121. The output signal 310 is transmitted to controller 100 and processed by processing unit 300 (Fig. 8) of controller 100.

[0049] [controller] 7 shows an example of the hardware and software configuration of a processor system that is the controller 100. The controller 100 is mainly composed of a computer 700. The computer 700 includes a processor 701, a memory 702, a communication interface device 703, an input / output interface device 704, and the like. These components are connected to a bus and can communicate with each other. An input device 705 and an output device 706 are externally connected to the computer 700 via the input / output interface device 704.

[0050] The processor 701 is configured with a semiconductor device such as a CPU, MPU, or GPU. The processor 701 includes a ROM, a RAM, and various peripheral functions. The processor 701 executes processing in accordance with a control program 711 stored in a memory 702. This realizes functions such as an SEM control function 721, a semiconductor measurement function 722, and a BSE processing function 723. The SEM control function 721 controls the main body 2 of the SEM 1. The semiconductor measurement function 722 is a function of the semiconductor measurement system that measures the dimensions of a pattern on a sample 3. The BSE processing function 723, which corresponds to the processing unit 300 described below (FIG. 8), discriminates between dark pulses and BSE signals based on a BSE detection signal and generates an image from which the dark pulse has been removed. The semiconductor measurement function 722 measures the dimensions of a semiconductor device pattern using the processing results of the BSE processing function 723. This enables high accuracy in measurement.

[0051] The memory 702 stores a control program 711, setting information 712, image data 713, processing data 714, measurement result data 715, etc. The control program 711 is a program for causing the processor 701 to execute processing to realize each function. The setting information 712 is system setting information and user setting information for each function. The image data 713 is data such as detected images and generated images acquired from the SEM 1. The processing data 714 is data generated in the processing process of each function. The measurement result data 715 is data including measured dimensions and the like obtained as a processing result by the semiconductor measurement function 722.

[0052] The communication interface device 703 is a device equipped with a communication interface for the main body 2 of the SEM 1, the communication network 9, etc. The input / output interface device 704 is a device equipped with an input / output interface, and an input device 705 and an output device 706 are externally connected. Examples of the input device 705 include a keyboard and a mouse. Examples of the output device 706 include a display and a printer. The input device 705 and the output device 706 may be built into the processor system that is the controller 100. A user such as an operator may use the controller 100 through the operation of the input device 705 or the screen display of the output device 706. The user may use the SEM 1 by accessing the controller 100 from the client terminal 5 of FIG. 1 via the communication network 9.

[0053] An external storage device (for example, a memory card or a disk) may be connected to the controller 100, and input / output data of the controller 100 may be stored in the external storage device. The controller 100 may also input / output data to / from an external server device via communication.

[0054] When using functions through client-server communication between the controller 100 of the SEM 1 in FIG. 1 and a user's client terminal 5, for example, it can be realized as follows: The user accesses the server function of the controller 100 from the client terminal 5. The server function of the controller 100 sends data such as a web page including a GUI (graphical user interface) to the client terminal 5. The client terminal 5 displays the web page or the like on a display based on the received data. The user views the web page or the like, checks information related to semiconductor measurement, etc., and inputs settings and instructions as necessary. The client terminal 5 sends the information entered by the user to the controller 100. The controller 100 executes processing related to semiconductor measurement, etc. based on the information entered by the user, and saves the results. The controller 100 sends data such as a web page including the processing results, etc. to the client terminal 5. The client terminal 5 displays the web page including the processing results, etc. on a display. The user checks the processing results, etc.

[0055] [Processing section] Fig. 8 shows an example of the configuration of a processing unit 300 and the like related to the BSE processing function 723, which is realized based on the configuration of the controller 100 in Fig. 7. Fig. 8 includes a processing unit 300, a control unit 400, and a GUI 500. The processing unit 300 and the control unit 400 are realized by program processing by the processor 701 of the controller 100. The GUI 500 is realized by displaying as a screen on a display serving as the output device 706 in Fig. 7 (or the client terminal 5 in Fig. 1), for example, based on screen data generated in the form of a Web page or the like based on processing by the processor 701 of the controller 100.

[0056] The processing unit 300 has, as functional blocks, a dark pulse processing unit 304 and an image generation unit 307. The dark pulse processing unit 304 has, as functional blocks, a signal peak position detection unit 302, a dark pulse determination unit 303, a dark pulse removal unit 331, and a determination criterion storage unit 309. Each of the functional blocks can be realized by program processing by the processor 701, but some of the functional blocks can also be implemented by dedicated circuits or the like.

[0057] The control unit 400 controls the processing unit 300. The control unit 400 receives input information such as instructions and settings from the user, and control information 801, via the GUI 500. The control unit 400 transmits a control signal 811 generated based on the control information 801 to the processing unit 300. The processing unit 300 controls each unit in accordance with the control signal 811. The processing unit 300 appropriately transmits a status signal 812 indicating the processing result, processing status, etc. to the control unit 400. The control unit 400 grasps the processing based on the status signal 812 and generates status information 802 to be output to the GUI 500. The control unit 400 displays information such as the processing result, processing status, etc. on the screen of the GUI 500 based on the status information 802.

[0058] 8 is a configuration example for realizing online processing. As online processing, the controller 100 performs processing such as discriminating dark pulses almost in real time based on the BSE detection signal.

[0059] The signal peak position detection unit 302 receives the four-channel signals 111a to 111d of the detection signal 111 from the BSE detection circuit 211 and performs processing to detect the peak times as the peak positions of the signals. The signal peak position detection unit 302 outputs a peak position signal 306, which contains information for the four channels, and is input to the dark pulse determination unit 303 and also transmitted to the control unit 400 as status information.

[0060] Dark pulse determination unit 303 determines whether a pixel at a target position is a dark pulse or a BSE signal based on peak position signal 306 and determination criterion information 341. Determination result information signal 342 output from dark pulse determination unit 303 is input to dark pulse removal unit 331.

[0061] The dark pulse removal unit 331 removes the dark pulse portion from the four-channel detection signals in accordance with the determination result information signal 342. The signal 332 after the dark pulse removal is input to the image generation unit 307. The dark pulse removal unit 331 also includes a unit that performs S / N calculation, which will be described later (FIG. 11). Based on the S / N information of the calculation result, the determination criterion information 341 in the determination criterion storage unit 309 is updated. The determination criterion information 341 is stored and set in the determination criterion storage unit 309.

[0062] The image generating unit 307 generates an image for measuring the dimensions of the pattern on the sample 3 based on the signal 332 after dark pulse removal and the signal 310, which is an SE detection signal. The image generated by the processing of the image generating unit 307 is stored as image data 713 in, for example, the memory 702 in FIG. 7. The generated image can also be displayed on the screen of the GUI 500 in FIG. 8.

[0063] [SE and BSE] Here, using Figure 9 etc., we will explain the general behavior and characteristics of secondary electrons 103 (SE) and backscattered electrons 105 (BSE) when, for example, there is a deep trench structure on the surface of a sample 3 that is the object of measurement, observation, etc. by SEM 1.

[0064] 9 shows a case where a deep trench structure is exposed in the XY plane, which is the surface of sample 3 in the SEM. (A) shows a schematic diagram of the XY plane when a deep trench structure 901 on the surface of sample 3 is enlarged and observed from above. (B) shows a schematic diagram of the XZ plane corresponding to line AA in (A), which is a cross section including the height and depth directions of the deep trench structure 901. (C) shows a schematic diagram of the YZ plane corresponding to line BB in (A).

[0065] In this example (A), the opening of deep trench structure 901 in the XY plane is roughly rectangular, with width 901a in the X direction and width 901b in the Y direction. Width 901a in the X direction is also referred to as the short side, and width 901b in the Y direction is also referred to as the long side.

[0066] In (B), the deep trench structure 901 has a depth 901c in the downward direction of the Z axis, where Z=0 is the height relative to the top surface 902 of the sample 3. Furthermore, the deep trench structure 901 has a generally trapezoidal cross section as shown in the figure, in accordance with an example of a typical processing process. The area of ​​the opening at height Z=0, e.g., the width 901a of the short side, is smaller than the area of ​​the bottom at depth 901c, e.g., the width 901d of the short side.

[0067] In (C), the deep groove structure 901 has a similar structure on the long side. The deep groove structure 901 has a width 901b of the long side of the opening at height Z=0, and a width 901e of the long side of the bottom 903 at depth 901c, where the latter is smaller.

[0068] In addition, this deep groove structure 901 has a depth 901c that is sufficiently larger than the width 901a of the short side, for example, and such a structure is referred to as a deep groove or deep hole. Also, depending on the pattern, it is not limited to the example of this deep groove structure 901, and there are also cases where the groove or hole continues longer in the X direction or Y direction, etc. Furthermore, the opening is not limited to a rectangular shape, and may be elliptical, etc. These can be collectively considered to be a groove structure or a hole structure.

[0069] 10 is a schematic explanatory diagram showing an example of the trajectories of incidence and emission of SEs and BSEs corresponding to the deep trench structure 901 of FIG. 10A shows an example of the trajectories of reflected electrons among the signal electrons generated when primary electrons, for example, the primary electrons 401 and 402, are incident on the deep trench structure 901 in a cross section including the short side of the deep trench structure 901, similar to FIG. 9B. 10B shows an example of the trajectories of secondary electrons among the signal electrons generated when primary electrons, for example, the primary electrons 401 and 402, are incident on the deep trench structure 901 in a cross section including the short side of the deep trench structure 901, similar to FIG. 9B.

[0070] 10A and 10B, when primary electrons 401 are incident on the top surface 902 near the deep trench structure 901 of the sample 3, backscattered electrons B1 are generated in (A), and secondary electrons S1 are generated in (B). On the top surface 902, the number of generated secondary electrons S1 is generally greater than the number of generated backscattered electrons B1.

[0071] On the other hand, when primary electrons 402 reach bottom 903 of deep trench structure 901, reflected electrons B2 are generated in (A), and secondary electrons S2 are generated in (B). In (A), primary electrons 402 reach bottom 903 and collide with bottom 903, generating reflected electrons B2, which pass through sidewall 904 in the X direction and are emitted upward from top surface 902 in the Z direction. In (B), primary electrons 402 reach bottom 903 and collide with bottom 903, generating secondary electrons S2, but secondary electrons S2 cannot pass through sidewall 904 in the X direction and are not emitted upward from top surface 902 in the Z direction.

[0072] When primary electrons 402 reach the bottom 903 of deep trench structure 901, reflected electrons B2 have higher energy than secondary electrons S2, and therefore, as shown in (A), reflected electrons B2 can pass through sidewall 904 and often exit from top surface 902. These reflected electrons B2 can be captured by the BSE detector 110. On the other hand, secondary electrons S2 have lower energy than reflected electrons B2, and therefore, as shown in (B), often cannot pass through sidewall 904 and do not exit from top surface 902. These secondary electrons S2 cannot be captured by the SE detector 115.

[0073] That is, for the deep trench structure 901, it is easy to obtain image information from reflected electrons, but it is difficult to obtain image information from secondary electrons.

[0074] Figure 10(C) shows an example of a trajectory when a primary electron 403 is incident on a cross section including the long side of a deep trench structure 901, as in Figure 9(C), collides with the bottom 903, generates a reflected electron B3, and the reflected electron B3 passes through a side wall 904 in the Y direction and is emitted upward from the top surface 902.

[0075] When the primary electrons 403 are incident up to the bottom 903 of the deep trench structure 901, the distance over which the reflected electrons B3 pass through the sidewall 904 of the sample 3 is shorter in the Y direction corresponding to the long side than in the X direction corresponding to the short side. For example, distance 906 is shorter than distance 905. In other words, in the Y direction, the width 901e of the bottom 903 is larger than the width 901d, and the width 901e is larger relative to the depth 901c, so that the reflected electrons B3 are more likely to escape upward in the Z direction from the bottom 903. Therefore, the number of reflected electrons that can be captured by the BSE detector 110 is greater for the reflected electrons B3 in (C) than for the reflected electrons B2 in (A).

[0076] That is, when image information is obtained from a three-dimensional structure such as the deep trench structure 901 using backscattered electrons, the shape and dimensions of the structure, its positional relationship with the BSE detector, and the like affect the detection.

[0077] The above explanation ignores the influence of the arrangement positions and orientations of the multiple BSE detectors 110A to 110D of the BSE detector 110. In more detail, as will be described later, the ease of detecting reflected electrons varies depending on the arrangement positions and orientations of the BSE detectors.

[0078] 2 is processed by the SE detection circuit 116 and output as a signal 310 suitable for imaging. The signal 310 is imaged by the image generation unit 307 in FIG.

[0079] Meanwhile, the BSE detection signals 111 (111a to 111d) output from the BSE detectors 110 (110A to 110D) in FIG. 2 are each processed by a BSE detection circuit 211 (211A to 211D) in FIG. 6. For example, the detection signal 111a is converted into a voltage by an I / V conversion circuit 201, and then converted into a digital signal by an ADC 202. After the pulse waveform fall time of the digital signal is shortened by a waveform shaping circuit 203, the peak positions of the pulse waveform, in other words, the peak times and amplitudes, are detected by a peak detection circuit 204. The detection signals 111 of the other channels are processed in the same way. The peak positions and amplitudes of these BSE detection signals 111 (111a to 111d) of the four channels are input as signals 301 (301a to 301d) to a dark pulse processing unit 304 in FIG. 8.

[0080] [Dark pulse processing section] The processing of the dark pulse processing unit 304 will be described with reference to Fig. 11 etc. Fig. 11 shows a more detailed configuration of the dark pulse processing unit 304 in Fig. 8. The signal peak position detection unit 302 in the dark pulse processing unit 304 has a peak position comparison unit 305 and a memory 311.

[0081] The peak position comparison unit 305 receives the signals 301 (301a to 301d) transmitted from the four-channel BSE detection circuit 211. The signals 301 contain information such as the peak positions, in other words, the peak times (denoted as t), and amplitudes. The peak position comparison unit 305 compares pulse signals contained in one pixel of the signals 301 between channels, and calculates the time difference (denoted as ΔT) between the peak positions. The peak position comparison unit 305 stores information including the peak positions and amplitudes of the compared pulse signals, as well as the time difference ΔT between the peak positions, in memory 311 and outputs the information as signal 306.

[0082] The dark pulse determination unit 303 compares the time difference ΔT of the signals 306 between each pulse signal with the determination reference value of the determination reference information 341. If the time difference ΔT of the signals 306 is equal to or less than the determination reference value, the dark pulse determination unit 303 determines that the compared pulse signal is a BSE signal, and if the time difference ΔT of the signals 306 exceeds the determination reference value, the dark pulse determination unit 303 determines that the compared pulse signal is a dark pulse. The dark pulse determination unit 303 outputs a signal 342 indicating the determination result, i.e., the result of discrimination between a BSE signal and a dark pulse, along with the pulse signal. The dark pulse determination unit 303 may store information including the determination result in memory or transmit it to the control unit 400 as status information.

[0083] The dark pulse removal unit 331 receives the pulse signal and the determination result signal 342, and removes the part of the pulse signal that is determined to be a dark pulse, as indicated by the signal 342. The dark pulse removal unit 331 outputs the removal result signal 332 and transmits it to the image generation unit 307.

[0084] Here, the dark pulse removal unit 331 may perform an S / N calculation to confirm and evaluate the S / N of the signal after dark pulse removal. In the first embodiment, this S / N calculation is also performed by an S / N calculation unit 331b within the dark pulse removal unit 331. As a result of the S / N calculation, information indicating the S / N of the signal after dark pulse removal is obtained. The dark pulse removal unit 331 compares this S / N value with a target value 344 set and stored in an S / N target value storage unit 345. If the comparison shows that the S / N is less than the target value 344, the dark pulse removal unit 331 adjusts the judgment criterion value in the judgment criterion storage unit 309. The dark pulse removal unit 331 stores the adjusted judgment criterion information 343 in the judgment criterion storage unit 309. In this case, the dark pulse judgment unit 303 performs dark pulse judgment again in the same manner using the adjusted judgment criterion information 343 as the judgment criterion information 341.

[0085] If the S / N obtained by the dark pulse removal unit 331 is equal to or greater than the target value 344, the dark pulse determination unit 303 transmits the signal 332 after the dark pulse removal to the image generation unit 307. The image generation unit 307 generates an image based on the signal 332 after the dark pulse removal.

[0086] The dark pulse removal unit 331 can remove dark pulses using, for example, the following method: The dark pulse removal unit 331 sets the pulse signal portion of the four-channel signal 301 that is determined to be a dark pulse to a zero value. This allows the dark pulse to be removed.

[0087] The above-described configuration of the dark pulse processing unit 304 is capable of online processing, i.e., signal processing in nearly real time. Furthermore, the first embodiment also supports offline dark pulse determination and dark pulse removal. Offline processing refers to performing processes such as dark pulse determination at any time after the generation of the signal 111. Therefore, the dark pulse processing unit 304 also transmits data and information corresponding to the signal 301 (301a to 301d) including the peak position and amplitude, and the signal 306 including the time difference ΔT output by the signal peak position detection unit 302, to the control unit 400. The control unit 400 stores the data and information in a storage resource such as the memory 702. When performing offline processing, the control unit 400 references the data and information stored in the storage resource and performs the same processes as described above, such as dark pulse determination, through program processing.

[0088] [Dark pulse detection] 12 and 13 are schematic explanatory diagrams showing specific examples of processing by the dark pulse processing unit 304 in FIG. 11. FIG. 12 shows example waveforms of signals 301 (301a to 301d), which are detection signals of four channels (ch1, ch2, ch3, ch4) for the same pixel at a certain target position obtained by scanning the surface of the sample 3. The horizontal axis is the time axis. This waveform contains a BSE signal waveform and a dark pulse waveform. For the same frame, the same pixel in each channel is represented as G1,f,p, G2,f,p, G3,f,p, and G4,f,p, respectively. These pixels have the same period, one pixel period PT.

[0089] In this example, signal 301a received on first channel ch1 has waveforms W1 and W2 as pulse signals. Signal 301b received on second channel ch2 has waveform W3 as a pulse signal. Signal 301c received on third channel ch3 has waveform W4 as a pulse signal. Signal 301d received on fourth channel ch4 has waveform W5 as a pulse signal. Each waveform has a peak time, which is the peak position. For example, the peak time, which is the peak position of waveform W1, is t1.

[0090] In the first embodiment, the data of the signal 301 sent to the signal peak position detection unit 302 includes the peak position and amplitude of each waveform as described above, but in FIG. 12, for ease of understanding, it is illustrated as a pulse waveform.

[0091] As described above, depending on the voltage Vx and current Ix of the electron gun 101 and the state of the sample 3, backscattered electrons may be detected as BSE detection signals by all four-channel BSE detectors 110 at approximately the same time, or may be detected by fewer channels. However, these BSE detection signals are input to and received by the BSE detectors 110 of the multiple channels at approximately the same time. In contrast, dark pulses occur randomly. For this reason, the probability that dark pulses will occur simultaneously in the BSE detectors of the multiple channels is low. In the first embodiment, such a mechanism is utilized.

[0092] The dark pulse determination unit 303 of the dark pulse processing unit 304 determines whether the time difference ΔT between the peak positions of pulse signals in the BSE detectors 110 of any of a first number (e.g., N1=2) of channels out of a total of four channels is less than or exceeds a judgment reference value (α). If the peak time difference, which is the time difference ΔT between peak positions, becomes less than the judgment reference value α almost simultaneously in the first number N1 or more channels, for example, if this occurs within one pixel period PT, the dark pulse determination unit 303 determines that the corresponding pulse signal is a BSE signal.

[0093] On the other hand, if the peak time difference is equal to or less than the judgment reference value α and occurs substantially simultaneously in less than a first number of channels (e.g., N1=2), the dark pulse judgment unit 303 judges that the corresponding pulse signal is a dark pulse. The dark pulse judgment unit 303 compares the pulse signals of the four channels, and if the peak time difference is greater than the judgment reference value α, it judges that the corresponding pulse signal is a dark pulse.

[0094] 13 shows an example of a method for comparing the peak time differences between the waveforms of the pulse signals. The dark pulse determination unit 303 comprehensively and without overlaps checks combinations of two waveforms W1 to W5 that are included as peak positions within one pixel period PT in the four-channel signal 301 (301a to 301d). FIG. 13 shows an example of comparing such combinations. The dark pulse determination unit 303 basically determines the peak time differences between waveforms in different channels, and if there are multiple waveforms in the same channel, it also determines the peak time differences between waveforms within the same channel.

[0095] In this example, first, the waveform W1 of the first channel ch1 is used as the reference waveform, and the waveforms W3 of the second channel ch2, W4 of the third channel ch3, and W5 of the fourth channel ch4 are used as the comparison waveforms. The comparison pairs between the reference waveform and the comparison waveform are (W1-W3), (W1-W4), and (W1-W5). Second, the waveform W2 of the first channel ch1 is used as the reference waveform, and the comparison waveforms are waveforms W3, W4, and W5. Third, the waveform W3 of the second channel ch2 is used as the reference waveform, and the comparison waveforms are waveforms W4 and W5. Fourth, the waveform W4 of the third channel ch3 is used as the reference waveform, and the comparison waveform is waveform W5.

[0096] The dark pulse determination unit 303 compares the peak time difference for each of these pairs with the determination reference value α. In this example, as shown on the right, of all the combinations, the peak time difference for the pair of waveforms W2 and W3 is the time difference ΔT between peak time t2 and peak time t3, which is less than or equal to the determination reference value α (ΔT≦α). For the other pairs, the time difference ΔT exceeds the determination reference value α. The number of channels corresponding to ΔT being less than or equal to α is 2.

[0097] For this reason, the dark pulse determination unit 303 determines that the set of waveforms W2 and W3 is a BSE signal. Regarding the other waveforms (W1, W4, W5) that do not fall into this set, the dark pulse determination unit 303 also determines that they are all dark pulses because the time difference ΔT for each combination exceeds the determination reference value α. Similar processing is performed on other frames and other pixels in chronological order.

[0098] As described above, the dark pulse determination unit 303 checks the peak time difference of the pulse signals between channels, and can distinguish between BSE signals and dark pulses based on whether the conditions using the determination reference value α and the number of channels (e.g., N1=2) are satisfied. Note that as a parameter constituting the condition, the number of channels (e.g., N1=2) at which the time difference ΔT is equal to or less than the determination reference value α may also be stored as a set value in the determination reference storage unit 309.

[0099] As described above, the first embodiment utilizes the fact that in the four-channel BSE detector 110 (110A to 110D), there is a correlation between the positions and times at which BSE signals occur among the channels, and that there is no correlation between the positions and times at which dark pulses occur among the channels. Based on this mechanism, the first embodiment can distinguish between BSE signals and dark pulses by comparing the peak positions of the detected pulses.

[0100] [Processing flow] 14 shows a processing flow mainly performed by the controller 100 in the charged particle beam system 1 of the first embodiment. This flow has steps S101 to S113. Note that this flow corresponds to the processing flow of the processing unit 300 in FIG. 8, and the processing contents correspond to each other. Some of the processing contents are performed by the BSE detection circuit 211 in FIG. 6.

[0101] In step S101, the controller 100 (more specifically, the processor 701; the same applies below) starts BSE detection. This start corresponds to the start of measurement of the sample 3, etc. In step S102, the charged particle beam device 1 detects output signals from the four-channel BSE detection system. That is, the charged particle beam device 1 obtains signals 111 (111a to 111d) output from the four BSE detectors 110 (110A to 110D) mentioned above.

[0102] In step S103, the charged particle beam device 1 detects the peak time and amplitude value, which are the peak position, for the pixel (Gc, f, p) at each target position in each frame of each channel based on the signal 111 (111a to 111d) by the BSE detection circuit 211. The controller 100 inputs and acquires the signal 301 (301a to 301d) in which the peak position and the like have been detected.

[0103] In step S104, the controller 100 calculates the peak time difference (ΔT) of each pulse signal for each target pixel based on the signal 301 (301a to 301d).

[0104] In step S105, the controller 100 stores information including peak positions, amplitudes, peak time differences, etc. for each pixel of interest based on the signal 301 in memory.

[0105] In step S106, the controller 100 sets a judgment reference value (α) for the peak time difference (ΔT) or refers to a judgment reference value (α) that has already been set.

[0106] In step S107, the controller 100 compares the peak time differences between the channels for each pixel of interest, and determines whether the time difference ΔT is equal to or less than the reference value α. If it is equal to or less than the reference value α (Y), the process proceeds to step S108, and if it exceeds the reference value α (N), the process proceeds to step S112. More specifically, as described above, whether the number of channels corresponding to a set of pulse signals whose time difference ΔT is equal to or less than the reference value α is equal to or greater than a first number (for example, N1=2) is also determined as a condition.

[0107] In step S108, the controller 100 determines that the set of pulse signals that satisfies the condition in step S107 is a BSE signal.

[0108] In step S109, the controller 100 determines that the set of pulse signals that does not satisfy the condition in step S107 is a dark pulse. In other words, the controller 100 determines that the pulse signals other than the pulse signals determined to be BSE signals in step S108 are dark pulses.

[0109] Furthermore, the process proceeds from step S112 to step S113. In step S113, the controller 100 removes the portion determined to be a dark pulse from the detection signal.

[0110] In step S109, the controller 100 checks whether the above processing has been completed for all target frames and all pixels, and if not (N), returns to step S103 and repeats the same process. If completed, proceeds to step S110.

[0111] In step S110, the controller 100 performs the S / N calculation described above to confirm the effect after dark pulse removal. The controller 100 checks whether the S / N value obtained by the calculation is equal to or greater than the S / N target value. If the target value is not reached (N), the process returns to step S106. In this case, in step S106, the controller 100 resets the judgment reference value α. If the target value is reached (Y), the process proceeds to step S111.

[0112] In step S111, the controller 100 performs imaging using the image generator 307 based on the processed BSE signal (signal 332 in FIG. 8), resulting in the generation of an image. Also in step S111, the controller 100 outputs related data and information handled in the above processing, such as peak position, amplitude, peak time difference, judgment criteria, and image. This output includes saving to memory and output to the user as external output, as shown in FIG. 1. For example, an image based on BSE detection may be displayed on the display screen along with a GUI.

[0113] In the above embodiment, the BSE detector 110 has four channels, in other words, four detection systems, but the present invention is not limited to this and can be applied to cases where the number of channels is more or less than four.

[0114] 15 shows an example of data and information related to the BSE processing function that the controller 100 stores in memory, for example, as a database table. The upper table in FIG. 15 includes items such as the ID of the sample 3, pixels Gc, f, p, pulse signal waveform ID, peak position (in other words, peak time t), amplitude, and image file name. The lower table includes items such as the ID of the pulse signal waveform combination, peak time difference (ΔT), a judgment reference value (e.g., α), and a judgment result indicating whether the signal is a dark pulse or a BSE signal.

[0115] The controller 100 may display the information in the table of Fig. 15 on a screen via a GUI. The table on the bottom side may be a table that is temporarily generated during the processing.

[0116] [Effects of the First Embodiment] As described above, the charged particle beam device 1 of the first embodiment can reduce degradation of detection due to dark pulses generated in the BSE detector 110; in other words, it can improve measurement accuracy. According to the first embodiment, dark pulses and BSE signals can be discriminated from each other in the output signal of the BSE detector 110, and dark pulses can be removed from the detection results. As a result, according to the first embodiment, the S / N ratio of the image acquired based on the BSE detector 110 is improved, detailed information on three-dimensional structures such as holes and grooves can be obtained, high measurement accuracy can be achieved, and instrumental differences can be reduced.

[0117] In the case of online processing mentioned above, only the data and information necessary for online processing is acquired and temporarily stored in storage resources, and processing is carried out in near real time. The temporarily stored data that has been processed is deleted as appropriate. In this case, the system requires relatively few storage resources. In the case of offline processing mentioned above, all data and information necessary for offline processing is acquired and stored in storage resources, and then the stored data and information is referenced and processed collectively when necessary. In this case, it is acceptable for the calculation to take some time.

[0118] <Embodiment 2> The charged particle beam device of the second embodiment will be described with reference to Figure 16 and subsequent figures. The basic configuration of the second embodiment etc. is the same as that of the first embodiment, and the following mainly describes the components of the second embodiment etc. that are different from the first embodiment. The main differences between the second embodiment and the first embodiment are as follows.

[0119] In the first embodiment, the dark pulse determination unit 303 in FIG. 8 is configured to determine whether a signal is a dark pulse or a BSE signal for each pixel period PT (time T) using the signal of a target pixel in one frame. In the first embodiment, a separate determination is made for each frame period. In contrast, in the second embodiment, the dark pulse determination unit is configured to determine whether a target pixel is a dark pulse or a BSE signal using signals from periods corresponding to multiple frames including the pixel. In the second embodiment, similar to the first embodiment, peak time differences are compared between combinations of pulse signal waveforms across channels for pulse signals generated approximately simultaneously. In the second embodiment, the time range to be compared, in other words, the determination period, is extended to a period corresponding to multiple frames. In the second embodiment, for each pixel at a target position, peak time differences are compared between combinations of pulse signal waveforms across channels within a period corresponding to multiple frames to distinguish between a BSE signal and a dark pulse.

[0120] [Dark pulse processing section] 16 shows an example of a functional block configuration of signal peak position detection unit 302B, which is located before dark pulse determination unit 303 in dark pulse processing unit 304 according to embodiment 2. This signal peak position detection unit 302B has, in order from the input side, peak position storage unit 312, peak position synthesis unit 330, and peak position comparison unit 313, and also has determination frame number storage unit 314. Peak position storage unit 312 includes a memory.

[0121] The dark pulse processing unit 304 sets the determination frame number 351 as one of the determination reference values ​​in the determination frame number storage unit 314. The determination frame number 351 is set as FD. The determination frame number storage unit 314 stores information including the determination frame number FD. The determination frame number FD is a number smaller than the maximum frame number F, and can be a value set by design or user setting. The control unit 400 may specify the determination frame number FD in advance based on user settings or the like, and set it in the determination frame number storage unit 314.

[0122] The peak position storage unit 312 refers to the determination frame number 351 (FD) set in the determination frame number storage unit 314 for the input signal 301. For a target pixel in the signal 301 of multiple channels, the peak position storage unit 312 stores the peak position of each waveform of the pulse signal included in one pixel period corresponding to the target pixel within a range of multiple frames corresponding to the determination frame number FD. That is, for each target pixel (Gc,f,p), the number of pixels for which peak positions are stored is the number multiplied by FD if the waveform appears in each frame. For all pixels and all channels, the number is the number multiplied by C×P×FD.

[0123] The signal 352 containing the peak positions and other information temporarily stored in the peak position storage unit 312 is input to the peak position synthesis unit 330. The peak position synthesis unit 330 performs synthesis processing on the peak positions of each waveform within the determination period determined by the determination frame number FD. This synthesis processing is processing for comparison in the peak position comparison unit 313 at the subsequent stage.

[0124] FIG. 17 and other figures are explanatory diagrams of the synthesis process performed by the peak position synthesis unit 330. FIG. 17 shows an example of pulse signal waveforms in three frames, frames f1, f2, and f3, as examples of multiple frames, for one pixel period PT corresponding to the p-th pixel in each frame as a pixel at a certain target position, for a four-channel signal 301 (301a to 301d) obtained by scanning the surface of the sample 3. In this example, the number of determination frames FD is 3. For example, for the first channel ch1, the target pixels G1,f,p in the three frames f1, f2, and f3 can be expressed as pixel G1,1,p, pixel G1,2,p, and pixel G1,3,p. The same applies to the other channels.

[0125] These signals 301 contain BSE signals and dark pulses. In this example, frame f1 contains waveform w1 of the first channel ch1 and waveform w2 of the second channel ch2, frame f2 contains waveform w3 of the third channel ch3 and waveform w4 of the fourth channel ch4, and frame f3 contains waveform w5 of the first channel ch1 and waveform w6 of the second channel ch2. Each waveform has a peak position, or peak time t. For example, the peak position of waveform w1 is peak time tw1.

[0126] For example, when frame f2 is used as the reference for a pixel at a target position, multiple frames (FD=3) that make up the determination target period can be obtained by referencing the previous frame f1 and the next frame f3. However, this is not limiting. For example, when frame f1 is used as the reference, multiple frames (FD) can be obtained by referencing the frames that come later in time, such as the next frame f2 and the frame that comes two frames later, such as the frame f3. For example, when frame f3 is used as the reference, multiple frames (FD) can be obtained by referencing the frames that come earlier in time, such as the frame f2 and the frame f1 that comes two frames earlier.

[0127] The peak position synthesizer 330 synthesizes waveforms of multiple (FD=3) frames (f1, f2, f3) for each channel as a waveform group for each pixel period PT. The peak position synthesizer 330 outputs a signal 353 as the synthesis result.

[0128] FIG. 18 shows an example of the results of combining waveforms for each channel based on signal 301 in FIG. 17. The result of combining multiple frames is referred to as a combined frame 1701, and the result of combining multiple pixel periods is referred to as a combined pixel period 1702. The combined signals in FIG. 18 are shown for each channel as signals 353a, 353b, 353c, and 353d. Signal 353a in combined pixel period 1702 for the first channel ch1 contains waveforms w1 and w5. Waveforms w1 and w5 overlap at approximately the same time. Signal 353b in combined pixel period 1702 for the second channel ch2 contains waveforms w2 and w6. Signal 353c in combined pixel period 1702 for the third channel ch3 contains waveform w3. Signal 353d in combined pixel period 1702 for the fourth channel ch4 contains waveform w4.

[0129] Here, if the number of backscattered electrons 105 detected by the BSE detector 110 is small, the BSE detectors 110 (110A to 110D) of multiple channels may not receive the backscattered electrons as pulse signals corresponding to BSE signals. In this case, with the method described in the first embodiment, since the number of pulse signals corresponding to BSE signals appearing in multiple channels is small, it may be difficult to distinguish between BSE signals and dark pulses.

[0130] Therefore, in the second embodiment, the determination period is expanded to a composite pixel period 1702 of a composite frame 1701 consisting of multiple (FD) frames to facilitate determination even when the number of detected backscattered electrons 105 is small. By combining signals from multiple frames for each channel in this manner, it is expected that more pulse signals corresponding to BSE signals will be received within the composite pixel period 1702 of the composite frame 1701. In this composite pixel period 1702, it is expected that pulse signals corresponding to BSE signals will occur almost simultaneously in multiple channels. On the other hand, because dark pulses occur randomly, it is expected that the probability of dark pulses occurring almost simultaneously in each channel will be low even within the composite pixel period 1702 of the composite frame 1701. Based on this mechanism, in the second embodiment, a determination is made to distinguish between a BSE signal and a dark pulse using the combined signal under the same conditions as in the first embodiment.

[0131] The peak position comparison unit 313 calculates the peak time difference (ΔT) for each combination of channels for the waveforms of the pulse signals of each channel based on the synthesized signal 353 (353a to 353d), and outputs a signal including information such as the peak position and peak time difference of the pulse signals of each pixel of each channel as signal 306.

[0132] As in the first embodiment, the dark pulse determination unit 303 compares the peak time difference of the signal 306 with the determination reference value α, and if the number of channels corresponding to which the time difference ΔT is equal to or less than the determination reference value α is equal to or greater than a first number (e.g., N1=2), the dark pulse determination unit 303 determines that the compared pulse signal is a BSE signal. If the pulse signal does not satisfy this condition, the dark pulse determination unit 303 determines that the pulse signal is a dark pulse. The subsequent processing by the dark pulse removal unit 331 and the like is the same as in the first embodiment.

[0133] Fig. 19 shows an example of checking the peak time difference between combinations of waveforms based on the combined signal 353 in Fig. 18. As in the first embodiment, the peak time difference is checked comprehensively for all combinations of waveforms without overlap. Multiple waveforms in different frames in the same channel (for example, waveform w1 and waveform w5) are also checked as a pair.

[0134] In this example, the peak time differences of the three waveforms, waveform w1 in frame f1 of the first channel ch1, waveform w5 in frame f3 of the first channel ch1, and waveform w4 in frame f2 of the fourth channel ch4 in Fig. 18, are less than or equal to the judgment reference value α. Therefore, the dark pulse judgment unit 303 judges the waveforms w1, w4, and w5 to be BSE signals, and judges the waveforms w2, w3, and w6 to be dark pulses.

[0135] [Processing flow] The processing flow of the controller 100 in the second embodiment is generally similar to the flow of the first embodiment shown in FIG. 14, but differs as follows: In step S102 described above, the controller 100 first starts processing the frames F in order, starting with the first frame. As shown in FIG. 16 described above, the controller 100 performs processing while storing information for each frame in memory. In step S103, for each target pixel, the controller 100 acquires a signal for the pixel period in the frame including the target pixel, and detects the peak position, etc.

[0136] In step S104, the controller 100 acquires signals for each target pixel over a range of frames of a plurality of FDs, including the preceding and following frames, while referencing information for each frame stored in memory, and calculates peak time differences between pulse signals, etc. In step S105, the controller 100 stores the calculated peak time differences, etc. in memory. The controller 100 similarly repeats the processes of steps S102 to S105 for each frame in sequence.

[0137] 17 based on the information in the memory, and then performs the processes from step S106 onward on the combined signal. In other words, the combination process combines pixel periods of different frames into a combined pixel period 1702 of a single combined frame 1701, thereby making it possible to compare peak times.

[0138] [Effects of the second embodiment] In the second embodiment, the BSE detector 110 with multiple (four) channels utilizes the fact that the positions and times of BSE signal occurrence in multiple (FD) frames are correlated between the same channel and between individual channels, and that the positions and times of dark pulse occurrence are uncorrelated between individual channels. Based on this mechanism, the second embodiment can distinguish between BSE signals and dark pulses by comparing the peak positions of pulses in the detection signals of multiple channels. Furthermore, in the second embodiment, even when there is a correlation between signals not only between different channels but also between different frames of the same channel, as shown in the example of FIG. 18, it is possible to distinguish between BSE signals and dark pulses. This improves the detection accuracy of BSE signals even when there is a small number of backscattered electrons detected.

[0139] <Third Embodiment> A charged particle beam device according to the third embodiment will be described with reference to FIG. 20 and subsequent figures. The third embodiment can also be considered as a modified example of the second embodiment. In the second and first embodiments, the target of determination for discrimination was a plurality of detection signals from the BSE detectors 110 of a plurality of channels. In the third embodiment, the target period of determination is extended to a period of frames of a plurality of FDs, as in the second embodiment. In the third embodiment, discrimination is performed not on a plurality of detection signals from the BSE detectors 110 of a plurality of channels, but on a single detection signal from the BSE detector 110 of a single channel, as in the second embodiment. The controller 100 performs discrimination between a BSE signal and a dark pulse for a pixel at a target position, using a single detection signal from a single BSE detector 110 as the target, over a range of frames of a plurality of FDs.

[0140] [Single BSE detector] 20 is a schematic explanatory diagram relating to dark pulse determination in embodiment 3. In the example of embodiment 3, a case will be described in which, for example, only the detection signal 111a (signal 301a in FIG. 6) of the BSE detector 110A of the first channel ch1 of the BSE detector 110 in FIG. 2 is used as the determination target. Note that this is not limiting, and as a modified example, even if only a single BSE detector is provided in the charged particle beam device, similar determination can be performed using that single BSE detector as the target.

[0141] In the third embodiment, the dark pulse processing unit 304 of the controller 100 has the same configuration as that shown in FIGS. 11 and 16. The only difference is that the peak position memory unit 312 shown in FIG. 16 receives only a detection signal from a single BSE detector, such as signal 301a, and processes the signal for each frame in a time series. For each pixel at a target position, the peak position memory unit 312 calculates the peak position of the pulse signal for each frame over a target period of multiple frames corresponding to the number of target frames FD, and stores the calculated peak position in memory. The peak position memory unit 312 outputs signals, such as the peak positions of the pulse signal for each frame, based on the detection signal from the single BSE detector as a single series of signals 352.

[0142] Based on this signal 352, the peak position synthesis unit 330 synthesizes, for each target pixel, signals from pixel periods of multiple FD frames into one signal for a synthesis pixel period 1902 in a synthesis frame 1901, as in the example of Figure 20, in the same way as in the second embodiment. The example of Figure 20 corresponds to an extraction of the signal 301a of the first channel ch1 from Figures 17 and 18. The peak position synthesis unit 330 outputs a synthesized signal 353.

[0143] The peak position comparator 313 compares the peak time difference (ΔT) between the waveforms of the pulse signals included in the composite pixel period 1902 with a threshold value as a judgment reference value (α) for each target pixel based on the composite signal 353, and discriminates between a BSE signal and a dark pulse. This judgment is made as follows.

[0144] The controller 100 of the charged particle beam device 1 of the third embodiment targets pulse signals received and appearing within a period of multiple FD frames as time T for each pixel at a target position for output signals from a single system from a single BSE detector. When multiple (at least two) pulse signals are received and appearing within a period of multiple FD frames, the controller 100 determines whether the peak time difference between the waveforms of those pulse signals is equal to or less than a threshold. When the peak time difference between the waveforms of those pulse signals is equal to or less than the threshold, the controller 100 infers and determines that those pulse signals are caused by BSE, in other words, are BSE signals.

[0145] Conversely, if only a single pulse signal is received or appears within a period of multiple FD frames, or if multiple pulse signals are received or appear but the peak time difference between the waveforms exceeds a threshold, the controller 100 infers or determines that those pulse signals are caused by dark pulses, in other words, that they are dark pulses.

[0146] [Effects of the Third Embodiment] As described above, according to the third embodiment, even when the detection signal of a single BSE detector or a single BSE detection system is used in an SEM or the like, it is possible to discriminate between BSE and dark pulses.

[0147] As a modification of the third embodiment, when an SEM or the like is provided with multiple BSE detectors and multiple BSE detection systems as shown in Fig. 2, it is possible to perform a determination similar to that of the third embodiment on the output signal of any one BSE detector selected from those systems. It is also possible to perform a determination similar to that of the third embodiment independently for each output signal of each BSE detector.

[0148] As a modification of the third and second embodiments, the determination period is not limited to a period of frames of multiple FDs, but may be a time T including the target pixel, such as a period of multiple pixels. As a modification of the third embodiment, the number of waveforms of pulse signals received and appearing within the determination period may be determined. That is, for the output signal from a single BSE detector, the controller 100 determines, for each pixel at the target position, whether the peak time difference between waveforms of a certain number (e.g., three) or more pulse signals within the determination period is equal to or less than a threshold value. The number of waveforms can also be set as one of the determination criteria information, either by design or user setting.

[0149] <Fourth Embodiment> A charged particle beam device according to the fourth embodiment will be described with reference to Figure 21 and subsequent figures. The fourth embodiment can be configured based on either the first or second embodiment, which includes a BSE detector with multiple channels, and has the feature of being added or modified to the first or second embodiment. The fourth embodiment improves accuracy in discriminating between BSE signals and dark pulses by referring to not only the signal from the BSE detection system but also the signal from the SE detection system as an auxiliary and reflecting it as a correction. In the fourth embodiment, based on the first embodiment, an example will be described in which both the determination method using a single frame in the first embodiment and the determination method using multiple frames in the second embodiment are used.

[0150] [Processing section] Fig. 21 shows an example of the configuration of the processing unit 308 of the controller 100 in the fourth embodiment. The configuration in Fig. 21 differs from that in the first embodiment mainly in that a shape estimation unit 317 is added to the processing unit 308, and a weighting unit 316 is added to the dark pulse processing unit 304.

[0151] In the fourth embodiment, the SE detection signal 310 output from the SE detector 115 in FIG. 2 via the SE detection circuit 116 is input to the shape estimation unit 317. The shape estimation unit 317 estimates the position and shape of a structure on the surface of the sample 3 based on the SE detection signal 310. As shown in FIG. 10 above, the ease of detecting secondary electrons differs between the top surface 902 and the deep trench structure 901, for example. Therefore, the position and shape of the structure can be estimated from the SE detection signal 310, for example, from the difference in brightness. For example, as shown in FIG. 9A, the shape of the opening of the deep trench structure 901 on the top surface 902 of the sample 3 can be estimated to be approximately rectangular, and the dimensions of the long and short sides, the depth 901c, and the like can also be estimated. The shape estimation unit 317 outputs a signal 318 representing the estimated shape, etc.

[0152] Based on signal 318, weighting unit 316 weights the number of BSE signals obtained between the same pixels in signals 301 (301a to 301d) of the four-channel BSE detection system. This improves the dark pulse determination rate. Weighting unit 316 outputs signal 323 representing the weighting. Dark pulse elimination unit 331 corrects signal 322, the determination result from dark pulse determination unit 303, based on the weighting value of signal 323. Dark pulse elimination unit 331 removes dark pulses based on the corrected result.

[0153] As shown in Figure 10, the number of secondary electrons obtained from structures such as holes and grooves in the sample 3, for example from the bottom, is small, whereas the number of secondary electrons obtained from the sample surface is greater than the number of backscattered electrons. Therefore, by using the SE detection signal, it is possible to estimate and determine where structures such as holes and grooves are located on the sample surface.

[0154] In this case, for pixels related to the identified structure, the BSE signal is discriminated from the dark pulse using the determination period consisting of multiple FD frames as shown in embodiment 2, and for other pixels corresponding to locations on the sample surface other than the identified structure, the BSE signal is discriminated from the dark pulse using the determination period of a single frame as shown in embodiment 1. This has the effect of reducing the processing time and the capacity of required storage resources (for example, the memory in the peak position storage unit 312).

[0155] A method for improving discrimination accuracy based on structural identification will be described with reference to Figures 21 and 22. In Figure 21, peak position comparison unit 305 of signal peak position detection unit 302 inputs signals 321 (321a to 321d) such as the peak positions of pulse signals contained in each pixel, based on signals 301 (301a to 301d) of each of the four channels, to dark pulse determination unit 303. Similar to the determination method in embodiment 1 or 2, dark pulse determination unit 303 discriminates between BSE signals and dark pulses by comparing the peak time difference with determination reference value α.

[0156] Signals 322 (322a to 322d) of each channel resulting from the discrimination are input to a dark pulse removal unit 331. Here, the dark pulse removal unit 331 uses a weighting signal 323 from the weighting unit 316 to perform weighting correction on the number of BSE signals resulting from the discrimination, and adjusts the ratio of the number of BSE signals to dark pulses for each channel. This weighting is determined based on the detection sensitivity of each channel according to the structure of the sample 3.

[0157] 22 and 23 are schematic explanatory diagrams relating to weighting. Fig. 22 shows an example of the location and shape of a structure 2100, such as a deep trench structure, on the surface of a sample 3, and the positional relationship of a four-channel BSE detector 110 (110A to 110D) with respect to the structure 2100. Fig. 22 is a schematic explanatory diagram on the XY plane, showing an enlarged view of the structure 2100 located at a central point O corresponding to the beam optical axis, and may be considered to have a detected image 2101 as a top-view image similar to that on the left, as shown on the right.

[0158] In this example, the structure 2100 has a roughly rectangular shape and a depth to the bottom, similar to the deep trench structure 901 in FIG. 9. The opening of the structure 2100 has a short side in the X direction and a long side in the Y direction. Here, the length of the short side is D and the length of the long side is H. The position of the structure 2100 can be expressed by position coordinates (x, y) in the coordinate system of the sample 3, for example. Note that various coordinate systems, such as the coordinate system of the sample 3, the coordinate system of the stage, and the coordinate system of the detected image, have corresponding relationships and can be converted as appropriate.

[0159] In this example, the long side of structure 2100 extends along the Y axis, and along the Y axis, BSE detector 110A of first channel ch1 (North) is located on one side (upper side in FIG. 22) of center point O, and BSE detector 110B of second channel ch2 (South) is located on the other side (lower side in FIG. 22). Along the X axis along which the short side extends, BSE detector 110C of third channel ch3 (West) and BSE detector 110D of fourth channel ch4 (East) are located.

[0160] 10, there is a relationship in which a relatively large number of reflected electrons are emitted in the direction of the long sides of the deep trench structure 901, whereas a relatively small number of reflected electrons are emitted in the direction of the short sides. This relationship can be expressed using D / H as the aspect ratio of the structure 2100. In this example, the parameter of this aspect ratio is a ratio in which the length H of the long side in the Y direction is used as the denominator and the length D of the short side in the X direction is used as the numerator. It can be said that the shape of the structure 2100 in which the sides in the Y direction are longer and the value of D / H is smaller relative to 1, the higher the BSE detection sensitivity of the BSE detectors 110A and 110B arranged on the Y axis.

[0161] Therefore, in the fourth embodiment, the BSE detection sensitivity is corrected by weighting based on the position and shape of the structure 2100 as described above and the positional relationship with the BSE detector 110 of each channel. In detail, this weighting correction is not a weighting correction of the BSE detection system itself including the BSE detector 110 or the judgment itself of the dark pulse judgment unit 303, but a weighting correction of the judgment result.

[0162] The table in Figure 23 shows an example of the proportion of the number of BSEs emitted in four directions (east, west, north, and south) in a four-channel BSE detector 110 for the structure 2100 and received by each BSE detector 110. In this example, the smaller the value of D / H, the longer the structure 2100 extends in the Y direction, and the higher the proportion of BSEs received by the BSE detectors 110A and 110B arranged in the Y direction. On the other hand, the larger the value of D / H, the longer the structure 2100 extends in the X direction, and the higher the proportion of BSEs received by the BSE detectors 110C and 110D arranged in the Y direction. When D / H is close to 1, the lengths in the X and Y directions are approximately the same, and the proportion of BSEs received by the BSE detectors 110 of each channel is approximately the same.

[0163] As an example, for the shape of the structure 2100, when D / H=0.6, the reception rate at the North BSE detector 110A and the South BSE detector 110B in the Y direction is 0.4, and the reception rate at the East BSE detector 110D and the West BSE detector 110C in the X direction is 0.1. If the total is 1, then 0.4+0.1+0.1+0.4=1. In the table of FIG. 23, the rows where D / H is "0.5 to 0.7" show the BSE reception rate at each of the North, East, West BSE detectors 110.

[0164] The rate at which BSE is received on each channel, in other words, the detection sensitivity, can also be expressed as a probability. For example, when D / H=0.6, the probability of BSE reception on each of the N, E, W, and S BSE detectors 110 can be expressed as {40%, 10%, 10%, 40%}. The rates shown in the table of FIG. 23 may be set in advance as table data.

[0165] As in the above example, the shape estimation unit 317 estimates the position and shape of the structure 2100 based on the SE detection signal, and the weighting unit 316 estimates the BSE detection rate and detection sensitivity for each channel based on the estimation results. The weighting unit 316 may refer to the rate for each channel from the table in FIG. 23 based on the D / H value, for example, or may calculate the rate each time. Then, the weighting unit 316 determines a weight for correcting the determination result for each channel in the determination result signal 322 (322a to 322d) based on the rate for each channel.

[0166] The following method can be used to determine the weighting, for example. The signal 322 (322a to 322d) of the determination result by the dark pulse determination unit 303 contains information on the estimation and determination result of whether each pulse signal included therein is a BSE signal or a dark pulse for each channel. The weighting is determined so that the ratio approaches 0.4 for signals 322a and 322b corresponding to BSE detectors 110A and 110B in the Y direction (N, S), and the ratio approaches 0.1 for signals 322c and 322d corresponding to BSE detectors 110C and 110D in the X direction (W, E), and is output as signal 323. The dark pulse removal unit 331 corrects the number of BSE signals by re-determining whether pulse signals determined to be BSE signals in the determination result signal 322 (322a to 322d) are BSE signals or dark pulses according to the weighting signal 323.

[0167] In this embodiment, the dark pulse elimination unit 331 is configured to perform the re-determination according to the weighting, but the present invention is not limited to this. A functional block that performs the re-determination according to the weighting may be provided within the dark pulse determination unit 303, or the functional block may be provided between the dark pulse determination unit 303 and the dark pulse elimination unit 331.

[0168] 22, the detection rate is assumed to be the same for the four BSE detectors at the position of the structure 2100, but is not limited to this. If the four BSE detectors are at different distances from the position of the structure 2100, the detection rate may be varied depending on the distance. That is, the closer the distance, the higher the detection rate may be.

[0169] 22, the detection ratio is assumed to be based on the Z direction, which is perpendicular to the surface, of the beam incident on the structure 2100, but is not limited to this. When the beam is incident on the structure 2100 from an oblique direction, the detection ratio may be varied depending on the oblique direction. Examples of when the beam is incident on the structure 2100 from an oblique direction include when the beam optical axis is in the tilt direction or when the beam scanning direction is taken into consideration.

[0170] 24 shows a specific example of re-determination and correction using the weighting, as follows. For example, as shown in the table of FIG. 24, for a period of a certain target pixel, the dark pulse determination unit 303 determines that the number of pulse signals determined to be BSE signals in the first channel ch1 (North) is AS, and the number of pulse signals determined to be dark pulses is AD. Similarly, the number of pulse signals determined to be BS and BD in the second channel ch2 (South), CS and CD in the third channel ch3 (West), and DS and DD in the fourth channel ch4 (East).

[0171] Regarding the BSE determination result, suppose the ratio of AS:BS:CS:DS is different from the aforementioned ratio of, for example, 0.4:0.1:0.1:0.4. In this case, the dark pulse processing unit 304 corrects the ratio of AS:BS:CS:DS by re-determining some of the pulse signals (AD, BD, CD, DD) that have been determined as dark pulses as BSE signals, so that the ratio approaches 0.4:0.1:0.1:0.4.

[0172] For example, suppose the number of BSE signals is AS = 2, BS = 0, CS = 1, and DS = 2, the ratio of the BSE signals is 0.2:0:0.1:0.2, and the number of dark pulses is AD = 3, BD = 3, CD = 4, and DD = 1. In this case, the number of BSE signals is increased (AS = 2 + 2 = 4, BS = 0 + 1 = 1, CS = 1 + 0 = 1, and DS = 2 + 1 = 3), and the number of dark pulses is decreased (AD = 3 - 2 = 1, BD = 3 - 1 = 2, CD = 4 - 0 = 4, and DD = 1 - 1 = 0), and a re-classification is performed to transfer some of the pulse signals identified as dark pulses to BSE signals. In the re-classification, weighting correction is performed so that the ratio of AS:BS:CS:DS approaches the BSE detection rate of {0.4:0.1:0.1:0.4}. The ratio of BSE signals before correction is AS:BS:CS:DS={0.2:0:0.1:0.2}, and after correction it becomes, for example, {0.4:0.1:0.1:0.3}, approaching the BSE detection ratio {0.4:0.1:0.1:0.4}.

[0173] As described above, in the fourth embodiment, taking into consideration the ratio of BSE detection in each channel according to the three-dimensional structure of the sample 3, it is assumed that the judgment result of discriminating between BSE signals and dark pulses will naturally be close to that ratio, and correction is made to the judgment result by weighting.

[0174] [Processing flow] 25 shows a processing flow in the fourth embodiment. This flow is based on the flow in the first or second embodiment, with step S411 and the like related to the SE detection system added. Steps S401 and S402 are similar to the flow in the first or second embodiment, and therefore details will be omitted. In step S401, BSE detection is started for the target sample 3 using the four-channel BSE detector 110.

[0175] In step S402, the controller 100 performs a determination for each target pixel to distinguish between a BSE signal and a dark pulse using the method described in the first or second embodiment. In this case, as an example of how to use different methods, as described above, the method of the first embodiment is applied when the target is the surface of the sample 3, and the method of the second embodiment is applied when the target is a deep trench structure or the like. However, this is not limiting, and as a modified example, only the method of the first or second embodiment may be applied without using different methods.

[0176] On the other hand, steps S411 to S414 can be performed in parallel with steps S401 to S402. In step S411, the controller 100 starts SE detection for the target sample 3 using the SE detector 115. In step S412, the controller 100 estimates the position and shape of structures such as holes and grooves on the surface of the sample 3, as with the shape estimation unit 317 described above. In step S413, the controller 100 estimates the BSE detection ratio of the four-channel BSE detector 110 according to the estimated structure, as shown in FIGS. 22 and 23 described above. In step S414, the controller 100 determines weighting of BSE signals and dark pulses based on the estimated ratio. The weighting information is, for example, information indicating an increase or decrease in the number of BSE signals or dark pulses for each channel.

[0177] On the other hand, in step S403, the controller 100 performs weighting correction on the signals resulting from the determination in step S402 so that the ratio of BSE signal determinations approaches the ratio of BSE detections, as shown in Fig. 24. The weighting correction is a re-determination in which part of the number of dark pulses is transferred as the number of BSE signals, as in the example described above.

[0178] In step S404, the controller 100 removes dark pulses from the determination result that reflects the above correction, as in the first embodiment, etc. Thereafter, although not described further, imaging of the detection signal is performed as described above.

[0179] [Effects of the fourth embodiment] As described above, in the fourth embodiment, the detection signals of the SE detection system are used to perform weighting correction on the determination results, assuming the proportion of BSE signal detection based on the relationship between the structure of the sample 3 and the BSE detectors of multiple channels. As a result, the fourth embodiment can improve the accuracy of discrimination between BSE signals and dark pulses, and the S / N ratio of the acquired image can be improved.

[0180] [GUI example] FIG. 26 shows an example of a screen display of the GUI 500 in FIG. 8, which is applicable to each embodiment. The screen in Fig. 26 has a setting value field 2601 and an output value field 2602. The setting value field 2601 displays various setting values ​​discussed in the embodiment. The user can check the setting values ​​on this screen and, if necessary, can input and change the setting values. In this example, the BSE / dark pulse determination reference values ​​displayed include the aforementioned determination reference value α, which is the threshold value for the peak time difference, the aforementioned first number N1, which is the number of BSE determination channels, and the aforementioned determination frame number FD related to the determination period. The determination reference value α can be set for each channel.

[0181] In the output value column 2602, the S / N value before and after dark pulse removal is displayed in comparison. Also displayed are images without a signal, acquired images of pattern images, etc. The acquired image may be selected and displayed from a detected image before dark pulse removal and a detected image after dark pulse removal, or both may be displayed in parallel. The screen display data is not limited to these.

[0182] [Note] Although the first to fourth embodiments have been described as examples in which the present invention is applied to an SEM, the present invention is not limited to this and can be similarly applied to an inspection device / system that uses X-rays, such as a CT device (CT: Computed Tomography).

[0183] The processor system serving as the controller 100 of the charged particle beam device in each embodiment has been described as a processor system provided in an SEM, but is not limited to this and may be a processor system separate from the charged particle beam device such as an SEM. The processor system may refer to and acquire data and information corresponding to the above-mentioned BSE detection signal and the like from the device such as an SEM via communication or a storage medium, and perform the same processing.

[0184] A semiconductor measurement method or other method corresponding to the charged particle beam device of each embodiment, in other words, a backscattered electron detection processing method or a dark pulse detection processing method, can be configured as follows, for example: The method corresponding to the first embodiment is a method having steps executed in a charged particle beam device including multiple BSE detectors for detecting BSE from a sample and a controller, the method including the steps of: the controller acquiring a first-peak time of a first peak within a period to determine whether a first peak included in an output signal of a first BSE detector of the multiple BSE detectors is caused by BSE or a dark pulse; the controller acquiring a second-peak time of a second peak included in an output signal of a second BSE detector other than the first BSE detector of the multiple BSE detectors within the period; and the controller determining that the first peak is caused by BSE if a second peak exists whose time difference between the first peak time and the second peak time is within a threshold; and determining that the first peak is caused by a dark pulse if a second peak exists whose time difference between the first peak time and the second peak time is within the threshold.

[0185] As a program corresponding to the apparatus and method of each embodiment, a computer program for realizing a processor system that is a controller of a charged particle beam apparatus can be configured as a program that causes a processor to execute processing corresponding to each step of the above method. Data corresponding to the program of the embodiment may be provided in a form stored in a non-transitory computer-readable storage medium.

[0186] Although the embodiments of the present disclosure have been specifically described above, they are not limited to the above-described embodiments and can be modified in various ways without departing from the spirit of the present disclosure. Except for essential components, components can be added, deleted, or replaced in each embodiment. Unless otherwise specified, each component can be singular or plural. A combination of each embodiment is also possible. [Explanation of symbols]

[0187] 1...charged particle beam device, 2...main body, 3...sample, 51...scintillator, 52...semiconductor photodetector, 100...controller (processor system), 102...primary electrons, 105...backscattered electrons, 110 (110A, 110B, 110C, 110D)...BSE detector, 115...SE detector, 300...processing unit, 304...dark pulse processing unit, 400...control unit, 500...GUI

Claims

1. A charged particle beam device comprising: a plurality of backscattered electron (BSE) detectors for detecting BSE from a sample; and a controller, In order to determine whether a first peak included in an output signal of a first BSE detector among the plurality of BSE detectors is caused by a BSE or a dark pulse, The controller acquiring a first peak time of the first peak within a period; acquiring, within the period, a second peak time of a second peak included in an output signal of a second BSE detector other than the first BSE detector among the plurality of BSE detectors; If there is a second peak where the time difference between the first peak time and the second peak time is within a threshold, it is determined that the first peak is caused by BSE; If there is no second peak where the time difference between the first peak time and the second peak time is within a threshold value, it is determined that the first peak is caused by a dark pulse. Charged particle beam device.

2. A charged particle beam device comprising: a first backscattered electron (BSE) detector for detecting backscattered electrons (BSE) from a sample; and a controller, In order to determine whether a first peak included in the output signal of the first BSE detector is caused by a BSE or a dark pulse, The controller a first pixel of a first frame corresponding to the first peak is identified within a period of a plurality of frames as a target period of the output signal of the first BSE detector, and a first sub-signal including a first peak time corresponding to the first pixel is selected; a second pixel in a second frame different from the first frame, the second pixel being located at the same pixel position as the first pixel, is identified within the period; and a second sub-signal including a second peak time corresponding to the second pixel is selected; determining that the first peak is caused by a BSE when there is a second sub-signal in which the time difference between the first peak time of the first sub-signal and the second peak time of the second sub-signal is within a threshold; If there is no second sub-signal in which the time difference between the first peak time of the first sub-signal and the second peak time of the second sub-signal is within a threshold value, it is determined that the first peak is caused by a dark pulse. Charged particle beam device.

3. 2. The charged particle beam device according to claim 1, The controller If, within the period, a condition that the time difference between peaks is within the threshold value is satisfied in detection signals from a first number (N1) or more of the plurality of BSE detectors, the peaks that satisfy the condition are determined to be caused by BSE. Charged particle beam device.

4. 2. The charged particle beam device according to claim 1, The controller identifying a first pixel in a first frame corresponding to the first peak within a period of a plurality of frames, the period being targeted for the output signals of the plurality of BSE detectors, and selecting a first sub-signal including a first peak time corresponding to the first pixel; a second pixel in a second frame different from the first frame, the second pixel being located at the same pixel position as the first pixel, is identified within the period; and a second sub-signal including a second peak time corresponding to the second pixel is selected; determining that the first peak is caused by a BSE when there is a second sub-signal in which the time difference between the first peak time of the first sub-signal and the second peak time of the second sub-signal is within a threshold; If there is no second sub-signal in which the time difference between the first peak time of the first sub-signal and the second peak time of the second sub-signal is within a threshold value, it is determined that the first peak is caused by a dark pulse. Charged particle beam device.

5. 2. The charged particle beam device according to claim 1, a secondary electron (SE) detector for detecting SE from the sample; The controller estimating the position and shape of structures on the sample based on the output signal of the SE detector; estimating a probability of reception of a BSE at each BSE detector of the plurality of BSE detectors based on the location and shape of the structure; determining whether the first peak is due to a BSE or a dark pulse based on an output signal of the first BSE detector corresponding to the position of the structure; When making the determination, a correction is made using a weighting according to the probability. Charged particle beam device.

6. 6. The charged particle beam device according to claim 5, the controller estimates an aspect ratio of the structure, and determines weights according to the probabilities according to the aspect ratio and a positional relationship between the structure and the plurality of BSE detectors; Charged particle beam device.

7. 3. The charged particle beam device according to claim 1, the controller removes the portion of the output signal determined to be the dark pulse based on the result of the determination. Charged particle beam device.

8. 3. The charged particle beam device according to claim 1, The controller calculates a signal-to-noise ratio (S / N) of the output signal based on the result of the determination, and if the S / N value is less than a target value, sets the threshold again and performs determination again. Charged particle beam device.

9. A processor system comprising one or more processors and a memory, The processor system includes: referring to output signals of a plurality of backscattered electron (BSE) detectors that detect BSE from the sample; In order to determine whether a first peak included in an output signal of a first BSE detector among the plurality of BSE detectors is caused by a BSE or a dark pulse, acquiring a first peak time of the first peak within a period; acquiring, within the period, a second peak time of a second peak included in an output signal of a second BSE detector other than the first BSE detector among the plurality of BSE detectors; If there is a second peak where the time difference between the first peak time and the second peak time is within a threshold, it is determined that the first peak is caused by BSE; If there is no second peak where the time difference between the first peak time and the second peak time is within a threshold value, it is determined that the first peak is caused by a dark pulse. Processor system.

Images