JP2004157139A - Circuit pattern inspection device and circuit pattern inspection method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To make a condition set efficiently, to shorten an inspection time, and to enhance reliability of inspection, in a circuit pattern inspection device for inspecting a defect, a foreign matter, a residue or the like in the same design pattern of a semiconductor device on a wafer under a production process for the semiconductor device, by an electron beam. <P>SOLUTION: This device is provided independently with a detection signal processing circuit for forming a large current high-speed image for detecting the existence of the defect, and a detection signal processing circuit for forming an image of a specified narrow portion detected by the defect detecting inspection hereinbefore. Alternatively, the device stores the first electronic optical system for the defect detecting inspection and the second electronic optical system exclusive for a review for observing the specified narrow portion detected by the defect detecting inspection, in the same vacuum container. Alternatively, the device is provided with the first detector for the defect detecting inspection and the second detector exclusive for a review for observing the specified narrow portion detected by the defect detecting inspection. <P>COPYRIGHT: (C)2004,JPO

Description

  The present invention relates to an inspection apparatus and an inspection method for a circuit pattern, and more particularly to an inspection apparatus and an inspection method for a circuit pattern such as a wafer in a semiconductor device manufacturing process.

  2. Description of the Related Art In a manufacturing process of a semiconductor device, as an inspection method of comparing and inspecting a defect of a circuit pattern formed on a wafer, an image of a similar pattern of two or more LSIs on one wafer is obtained and compared. Inspection devices have been put to practical use.

  In particular, a pattern comparison and inspection apparatus using an electron beam is described in Patent Documents 1 and 2, Non-Patent Documents 1 to 3, and the like. There, it is necessary to acquire images at a very high speed in order to obtain a practical throughput. In order to secure the S / N ratio of the image acquired at high speed, the electron beam current of 100 times or more (10 nA or more) of the ordinary scanning electron microscope is used, and the S / N ratio of the image is secured while maintaining a practical inspection speed. are doing. The beam diameter is considerably wider than that of a normal scanning electron microscope, and is about 0.05 μm to 0.2 μm. This is because the beam current is large and the energy width of the electrons is widened, thereby increasing chromatic aberration, limiting the brightness of the electron gun, and being limited by the Coulomb effect.

  An image formed by such an electron optical system is sent to an image processing system, and a comparison inspection of adjacent images of the same pattern portion is performed. If there is a portion having different brightness when comparing the images, the portion is regarded as a defect and its coordinates are stored.

  With the above configuration, it is possible to detect a defect having a size of about 0.1 μm.

  Patent Documents 3 to 8 and the like disclose devices for inspecting semiconductor defects by applying a voltage to a sample and an electrode near the sample to reduce energy of an electron beam. However, there is no description of the review function described below or a statement that it also has an energy analysis function.

JP-A-59-192943

JP-A-5-258703 JP-A-7-078855 JP 09-181139 A JP-A-10-019538, JP 10-027834 A JP 10-027835 A JP 11-029501 A J. Vac. Sci. Tech. B, Vol. 9, No. 6, pp. 3005-3009 (1991) J. Vac. Sci. Tech. B, Vol. 10, No. 6, pp. 2511-2515 (1992) SPIE (The International Society for Optical Engineering) Vol. 2439

  When an inspection is started using the above-described apparatus, there are many parameters that need to be set in advance. First, the setting parameters of the electron optical system include the irradiation energy of the electron beam, the gain of the secondary electron (or charged particles such as backscattered electrons) signal forming system that forms the image, the pixel size (minimum element of the image), and the beam current. and so on. On the other hand, when the image processing apparatus compares two images of the same pattern adjacent to each other, there is a threshold for determining whether or not the image is a defect. When this threshold value is set low, the defect detection sensitivity improves, but the possibility that non-defect portions are regarded as defects increases. On the other hand, when the threshold value is increased, the detection sensitivity decreases.

  The optimum values of the above parameters vary depending on the process to be inspected, the pattern size, and the type of defect to be detected. Therefore, it is necessary to perform a trial inspection before the inspection, display an image of the detected defect coordinates, and set the above parameters to an optimum value while checking whether the defect to be detected is detected.

  Further, there is a need to obtain an image of the defect coordinate portion after the inspection is completed, and the operator needs to confirm what kind of defect has been detected.

  That is, in order to detect the presence of a defect, an image is acquired at high speed, and not only the defect is detected by image processing, but also a specific narrow visual field is imaged and observed with the naked eye, similarly to a normal scanning electron microscope. The function of doing is also essential. In the present specification, this is hereinafter referred to as a review. In the case where it is necessary to particularly distinguish the review and the inspection by the high-speed image acquisition for detecting the presence of a defect over a relatively large area, the latter is used. Will be referred to as a defect detection inspection.

  At the time of review, it is not necessary to form an image at a higher speed than at the time of defect detection inspection. On the other hand, since it is necessary to recognize not only the presence or absence of a defect but also the shape and type of the defect to some extent, a high-resolution image is required.

  However, in the conventional apparatus, the electron optical system is designed optimally for acquiring an image by high-speed scanning with a large current, and a sufficient resolution cannot be obtained as a review image. Therefore, the accuracy for determining whether the detected defect is a true defect or an erroneous detection caused by improper parameter setting is low. Therefore, the inspection is often performed without setting the setting parameter to an optimum value.

  SUMMARY OF THE INVENTION An object of the present invention is to improve the inspection conditions in an apparatus for inspecting a sample, for example, by inspecting an electron beam for defects, foreign matter, residues, etc. of the same design pattern of a semiconductor device on a wafer in the process of manufacturing the semiconductor device. An object of the present invention is to provide an inspection apparatus that can be set well.

The inspection apparatus of the present invention achieves the above object by the following configuration.
That is, a first invention according to claim 1 of the present invention provides an electron optical system which converges an electron beam emitted from an electron source by an electron lens and irradiates a sample with the electron beam to scan the sample. A stage that is mounted and movable to irradiate a desired position on the sample with an electron beam; a detector that detects backscattered electrons or secondary electrons obtained from an irradiated portion of the sample; A first circuit for amplifying the output from the detector at a first amplification factor, an amplifier for amplifying the output from the detector at an amplification factor greater than the first amplification factor, and a high-frequency component cut filter. A second circuit, an image forming apparatus that forms an image of the sample based on a detection signal from the detector, and an image signal obtained by the image forming apparatus, the same from another location on the sample. And compare the difference between the two image signals. And an output from the detector for scanning the electron beam over the sample at a relatively large area, a relatively large current, and a relatively high speed in the first case. From the detector in the second case of scanning the electron beam over the sample at a small area smaller than the relatively large area, a small current smaller than the relatively large current, and a low speed slower than the relatively high speed. And a switching means for switching and outputting the output of the second circuit to the second circuit.
According to a second aspect of the present invention, in the inspection apparatus according to the first aspect, in the second case, the pixel size is a minimum unit of an image formed in the first case. An inspection apparatus that forms an image with a smaller pixel size.
According to a third aspect of the present invention, in the inspection apparatus according to the first or second aspect, in the second case, the spot size of the electron beam on the sample is equal to the first size. An inspection apparatus characterized in that the inspection apparatus is smaller than the above.
According to a fourth aspect of the present invention, in the inspection apparatus according to any one of the first to third aspects, the electron lens includes an objective lens closest to the sample, and the electron source. A second lens disposed on the side, wherein in the second case, the second lens is adjusted and the magnification of the electron optical system is set to be small.
According to a fifth aspect of the present invention, in the inspection apparatus according to any one of the first to fourth aspects, the electron optical system includes a stop having an opening having a plurality of diameters. In the second case, the inspection apparatus is characterized in that the diameter of the stop is switched to a smaller one.
According to a sixth aspect of the present invention, in the inspection apparatus according to the fourth aspect, the second lens is a condenser lens, and in the second case, the electron beam is formed. The inspection apparatus is characterized in that the position of the crossover is moved.
According to a seventh aspect of the present invention, in the inspection apparatus according to any one of the first to sixth aspects, in the second case, the scanning deflection area of the electron beam is the first deflection area. An inspection apparatus characterized in that the deflection area is reduced to half or less as compared with the case of (1).
According to an eighth aspect of the present invention, in the inspection apparatus according to any one of the first to seventh aspects, the second circuit has a higher response speed than the first circuit. Is an inspection apparatus characterized by being slow.
According to a ninth aspect of the present invention, in the inspection apparatus according to any one of the first to eighth aspects, means for detecting a difference between the two image signals in the first case is provided. Means for switching the magnification of the electron optical system, the current value of the electron beam, the scanning speed of the electron beam, and the number of image additions in the image forming apparatus in an interlocking manner in accordance with the output result. An inspection apparatus characterized in that:
A tenth invention according to claim 10 of the present invention has a first electron source that emits a first electron beam, a first focal length, and a first electron beam that emits a first electron beam. A first objective lens that focuses on the sample located at the sample position, and a first scanning deflector that scans the sample located at the first sample position with the first electron beam. A first electron optical system; a first detector for detecting secondary charged particles generated from a sample located at the first sample position; a second electron source for emitting a second electron beam; A second objective lens having a second focal length that is shorter than the first focal length and focusing the second electron beam on a sample located at a second sample position; A second scanning optical deflector for scanning a sample located at the second sample position on the sample; and the second electron optical system; A second detector for detecting secondary charged particles generated from the sample located at the sample position; and the first and second detectors based on outputs from the first detector and the second detector. An image forming apparatus for imaging each of the samples located at the sample position, and a stage for moving the sample from the first sample position to the second sample position, wherein the first electron optical system, A second electron optical system, the first detector, the second detector, and the stage are housed in the same vacuum vessel, and the first detector detects a sample located at the first sample position. Moving a specific position on the sample selected based on the image data formed based on the output to the second sample position by moving the stage, and moving the specific position to the second sample position. This is an inspection device that performs magnified observation using an electron optical system.
According to an eleventh aspect of the present invention, in the inspection apparatus of the tenth aspect, the acceleration voltage for the second electron beam in the second electron optical system is the first electron beam. An inspection apparatus characterized in that the inspection voltage is lower than an acceleration voltage for the first electron beam in an optical system.
According to a twelfth aspect of the present invention, in the inspection apparatus of the tenth or eleventh aspect, the electron beam is decelerated to a desired energy immediately before the stage and the sample on the stage. Therefore, the negative voltage applied to the stage and the sample is a negative voltage at the second sample position smaller than the negative voltage applied at the first sample position. It is an inspection device.
According to a thirteenth aspect of the present invention, in the inspection apparatus according to any one of the tenth to twelfth aspects, an X-ray generated from a sample located at the second sample position is detected. An inspection device comprising an X-ray detector.
According to a fourteenth aspect of the present invention, in the inspection apparatus according to any one of the tenth to twelfth aspects, the secondary electron emitted from the sample located at the second sample position is provided. An inspection apparatus comprising an energy filter for filtering energy, and detecting a potential at a position irradiated with the second electron beam.
According to a fifteenth aspect of the present invention, there is provided an electron optical system which converges an electron beam emitted from an electron source by a lens, irradiates the sample with a beam, and scans the sample. A stage movable to irradiate a desired position on the sample with an electron beam, and a first stage for detecting backscattered electrons or secondary electrons obtained from an irradiated portion of the sample with a first sensitivity. A second detector for detecting backscattered electrons or secondary electrons obtained from the irradiated portion of the sample at a second sensitivity higher than the first sensitivity, and detection from the detector An image forming apparatus for forming an image of the sample based on the signal, and comparing the image signal obtained by the image forming apparatus with image signals similarly obtained from other places on the sample, Means for detecting the difference between the signals; and When scanning the sample at a first area, a first current value, and a first speed, the backscattered electrons or the secondary electrons are guided to the first detector, and the electron beam is scanned on the sample on the sample. When scanning at an area smaller than the first area, at a current value smaller than the first current value, and at a lower speed than the first speed, deflection for guiding the backscattered electrons or secondary electrons to the second detector. This is an inspection apparatus provided with a circuit.
According to a sixteenth aspect of the present invention, there is provided an electron optical system for irradiating and scanning a sample on which a circuit pattern is formed with a converged electron beam, and a back side obtained from an irradiated portion of the sample. A detector that detects scattered electrons or secondary electrons, an image forming apparatus that forms an image of the sample based on a detection signal from the detector, and an image signal obtained by the image forming apparatus that is used as a reference image. Providing a difference detecting means for detecting a difference between the two image signals by comparing with a signal; and scanning a relatively large area of the sample with the electron beam at a relatively large current and a relatively high speed. The output obtained from the detector obtained by the amplification by an amplifier having a first amplification factor is supplied to the image forming apparatus to form an image signal. Obtained from other areas on the sample Detecting the difference between the two image signals by comparing with the similar image signal to determine the coordinates of the location where the difference occurs; and the relatively large area including the coordinate position of the sample. An output from the detector obtained by scanning an area having a smaller area with the electron beam at a small current smaller than the relatively large current and at a low speed slower than the relatively high speed is used as the first output. Supplying an image signal to the image forming apparatus through a circuit including an amplifier that amplifies at an amplification factor greater than the amplification factor and a high-frequency component cut filter to form an image signal and observing a location where the difference occurs. This is a circuit pattern inspection method including the above.

Since the present invention is as described above, the following effects can be obtained. A defect detection inspection by high-speed image acquisition for detecting the presence of a defect over a relatively large area, and a review of imaging a specific narrow part detected by the defect detection inspection and observing it with the naked eye, respectively. An image signal signal having a sufficient SN ratio can be obtained by performing the detection through a detection circuit having independently optimized paths, and the inspection conditions can be set efficiently. As a result, the speed of the inspection process is increased and the reliability of the inspection result is improved.
Further, the first electron optical system for the defect detection inspection for detecting the presence of the defect and the second electron optical system dedicated for review for observing a specific narrow portion detected by the defect detection inspection are the same. The inspection apparatus according to the present invention, which is arranged in a vacuum vessel and can switch between the defect detection inspection and the review only by moving the stage on which the sample is placed, enables quick and highly reliable inspection.
Also, a first detector for defect detection inspection for detecting the presence of a defect, a second detector exclusively for review for observing a specific narrow part detected by the defect detection inspection, and an electron beam irradiation The inspection apparatus of the present invention, which has a deflection circuit for guiding backscattered electrons or secondary electrons generated from the sample to the first detector at the time of defect detection inspection and to the second detector at the time of review, has low noise. A signal with little deterioration in high-frequency characteristics is obtained, enabling highly reliable inspection.

(Example 1)
A first embodiment of the present invention will be described with reference to FIGS. 1 and 2 are configuration diagrams of the present invention.
The inspection apparatus is roughly divided into an electron optical system 101, a sample chamber 102, a control unit 104, and an image processing unit 105.
The electron optical system 101 includes an electron source 1, an electron beam extraction electrode 2, a condenser lens 4, a blanking deflector 17, a scanning deflector 8, a diaphragm 5, and an objective lens 7.
FIG. 1 shows a case where the secondary electron detector 9 is located below the objective lens 7, but the position of the detector 9 may be provided on the objective lens 7. In that case, the secondary electrons reach the detector 9 after passing through the objective lens 7. When the detector 9 is located below the objective lens 7, the deflection range of the objective lens 7 can be widened, but the resolution at the time of review is limited due to the large aberration of the objective lens 7. On the other hand, when the detector 9 is provided on the objective lens 7, it is possible to reduce the aberration of the objective lens 7, so that the resolution at the time of review can be further increased. However, there is a disadvantage that the deflection range is reduced. Both have advantages and disadvantages, and either configuration may be selected as needed. A positive high voltage is applied to the detector 9. Further, the ground levels of the preamplifier 12 and the detection circuit 30 connected to the detector 9 are also on the positive high potential applied to the detector 9. The output signal of the secondary electron detector 9 is amplified by the preamplifier 12 and converted into digital data by the detection circuit 30. This signal is converted into light by the light transmitter 34, received by the light receiver 35 at the ground potential level, converted into an electric signal, and sent to the image processing unit 105.
In FIG. 1, the condenser lens 4 is constituted by an electromagnetic lens, but may be constituted by an electrostatic lens as shown in FIG. 2. The electrostatic lens is constituted by, for example, three electrodes. A so-called Einzel lens, which applies a positive or negative voltage to the center electrode, is optimal, but the same is not particularly true for an Einzel lens. By using an electrostatic lens, the size of the electron optical system can be reduced.
Returning to FIG. 1, the sample chamber 102 includes a stage 24, an optical sample height measuring device 29, and a stage position measuring device 31 using a laser length measuring device or a linear scale.
The image processing unit 105 includes image storage units 18 and 19, a calculation unit 20, and a defect determination unit 21. The captured electron beam image is displayed on the monitor 22. Operation commands and operation conditions of each part of the inspection apparatus are input and output from the control unit 104. Conditions such as an acceleration voltage at the time of generation of an electron beam, an electron beam deflection width, a deflection speed, a moving speed of a sample table, and a timing for capturing a signal of a detector are input to the control unit 104 in advance. In addition, a correction signal is generated from the signal of the optical sample height measuring device 29 and the signal of the stage position measuring device 31 and the objective lens power supply 7 and the scanning signal generator 13 are so controlled that the electron beam 6 is always irradiated to a correct position. Send a correction signal.

FIG. 3 is a diagram showing a flow of semiconductor pattern appearance inspection using an electron beam. A flow for carrying out a pattern appearance inspection of a semiconductor wafer according to the present invention will be described with reference to FIG.
First, after loading a wafer, rough alignment is performed by an optical microscope. This is because the scanning range is narrow in an electron beam scanning image, and there is a possibility that the alignment mark does not fall within the scanning range of the electron beam only by mechanical positioning during loading. Therefore, alignment is performed with an accuracy of about several tens of μm using a low-magnification optical microscope having a wide field of view.
Next, conditions for the electron optical system are set. The irradiation energy of the electron beam, the pixel size (minimum element of the image), the beam current, and the like are set according to the type of the pattern of the wafer to be inspected. The optimum electron beam irradiation energy differs depending on the pattern material. Normally, it is preferable to set the energy to a higher value of several keV or more so that a high resolution can be obtained in a conductive material, and to set the energy to 1.5 keV or less in a pattern including an insulator to prevent electrification.
Next, a pixel size corresponding to the defect size of particular interest is set. Insignificantly reducing the pixel size increases the time required for inspection. Next, the beam current is set. The default value is unique to the apparatus, but it is preferable to set a smaller current especially for a wafer which is easily charged. If it is known in advance that the contrast of the defect to be detected is large, the current can be made smaller and the electron beam can be made smaller so that high-sensitivity inspection can be performed.
Next, an image of the wafer to be inspected by an electron beam is displayed, and focus and astigmatism are corrected. This can be carried out automatically on a computer or by using an image processing device dedicated thereto by taking in an image.
After the above setting is completed, fine alignment of the image by the electron beam is performed next. As a result, the coordinate system of the stage and the coordinate system of the wafer accurately match, and a desired position on the wafer can be inspected by measuring the coordinates of the stage with a laser length measuring device or the like.

  Next, an area to be inspected is input from a control screen such as a display of a control workstation.

Next, a test inspection is performed. The trial inspection is performed to determine whether the various parameters set above are appropriate. First, a part of the pattern to be inspected is specified. This is the area designation of the trial inspection. Then, a test inspection is performed. In the test inspection, an image is detected by an electron beam in the same operation as in the actual defect detection inspection, but all the acquired images are loaded into the memory without passing through the image processing circuit.
Next, image processing conditions are set. This condition is a threshold value used by the image processing apparatus to determine whether or not there is a defect when comparing two adjacent images of the same pattern. When this threshold value is set low, the defect detection sensitivity improves, but the possibility that non-defect portions are regarded as defects increases. On the other hand, when the threshold value is increased, the detection sensitivity decreases.
In FIG. 3, the brightness threshold is a threshold of a difference between luminance signals of two images. An image formed by an electron beam is rougher than an optical microscope image (that is, the SN ratio of the image is poor). Therefore, even if the inspection objects to be compared are the same and the average brightness is the same, there is a slight variation in brightness when viewed from pixel to pixel. Therefore, a threshold value of the light-dark difference is provided, and by comparing the images and detecting only a portion where the light-dark difference is larger than the threshold value as a defect, the roughness of the image is not recognized as a defect.
The position deviation threshold value is provided so that incomplete registration of two images to be compared is not erroneously recognized as a defect. It is considered that the image position deviation exists about the threshold value set here, and only a case where a further difference occurs is regarded as a defect.
The filter type is a filter to be applied to the acquired image, and exists to reduce the roughness of the detected image and enhance the edge. In this method, an optimum one is selected by trial and error according to the type of pattern and the defect to be detected.

After setting the above parameters, image processing is performed. In the image processing in the trial inspection, a comparative inspection is performed using a series of already acquired images. This inspection detects a defect, stores its coordinates, and displays the location of the defect on a control screen. If no defect is detected at this time, the sensitivity is increased by lowering the threshold value of the image processing parameter, and the comparison inspection is performed again using the two-dimensional information on the memory.
Next, an image of the coordinates detected in the trial inspection is observed. This is a review. Since the image of that portion has already been stored in the memory, the image is first displayed on a display of a control workstation or a display dedicated to image display, and the type of defect detected is visually observed. Here, for a considerably large defect, the shape and type of the defect can be easily determined visually. However, a fine defect close to the performance limit of the device and a defect having a small contrast cannot often be determined sufficiently visually. Therefore, the condition of the electron optical system is switched to the review condition to display a high-resolution image. This makes it possible to easily determine whether the defect detected in the test inspection is a true defect or a defect that the user wants to detect. From the result, it is possible to find out a condition for reliably detecting a defect to be detected by repeatedly setting the parameters of the image processing and repeating the review.
Conventional equipment, which does not have high-resolution review conditions, is designed to form an image with a large current electron beam. May make it impossible to observe the image. Further, the purpose is to confirm the presence or absence of a defect, and the resolution is insufficient for judging the details of the defect by visual observation. For this reason, trial and error had to be repeated in the review and the setting of the image processing parameters, which required a great deal of time. However, according to the present invention, since the quality of the review screen is good, the quality of the selection of the image processing parameter can be accurately determined, and the number of trial and error has been significantly reduced. As a result, the time required for manned work has been significantly reduced, and the overall operation efficiency of the apparatus has been improved.

  When the review for setting the conditions is completed, the defect detection inspection is started. This defect detection inspection is performed automatically and unattended. When the defect detection inspection is completed, the review is performed again. After confirming whether the defect detected by this review is a true defect, the wafer is unloaded. If necessary, move to another analysis / inspection apparatus to perform various inspection / analysis.

When automatically inspecting a wafer for which defect detection inspection conditions have already been established, the defect detection inspection and review are performed according to the flow shown in FIG. That is, as soon as the wafer is loaded, a defect detection inspection is performed under the existing condition settings, and a review is performed after the completion of the defect detection inspection. In this review, the detected defects are classified according to the types of defects such as foreign matter, electrical defects, and shape defects.
Next, the specification, differences, and features of the review conditions and defect detection inspection conditions that are the points of the present invention will be described.
FIG. 5 compares the performance of the defect detection inspection condition and the review condition.
Under the defect detection inspection conditions, an extremely high-speed image acquisition is possible because the inspection of the entire wafer is performed in a practical time. That is, the time for forming an image of 100 μm square with a pixel size of 0.1 μm is 20 msec or less. The stage is continuously moved for high-speed image acquisition, and the electron beam is scanned in a direction substantially orthogonal to the stage. The beam current was designed to be 20 nA or more in order to realize image formation with an SN ratio that can withstand inspection at high speed. Specifically, the current density obtained from the electron source was set to 0.2 mA / sr or more. In addition, the aperture of the lens is larger than that of the high-resolution SEM, and the angle of capture of the electron beam from the electron source α (that is, the electron beam emitted from the electron source at a wide angle, Beam divergence angle) was set to 6 mrad or more. The pixel size, which is the minimum unit of the image, was set equal to the minimum defect size to be detected. At present, the smallest defect that causes a problem is approximately 0.05 μm, so the pixel size is set to 0.05 μm or more. The meaning of 0.05 μm or more is described to indicate that the pixel size can be set to be larger than 0.05 μm because a defect size larger than 0.05 μm may be seen depending on a wafer. Also, depending on the process of the wafer to be inspected, the amount of secondary electrons generated is so small that sufficient brightness may not be obtained by only one electron beam scanning. In that case, it is also possible to add images obtained by scanning the electron beam a plurality of times. However, there is a problem that the inspection time is delayed, so the smaller the number of additions, the better. The larger the image forming range is, the better.

  On the other hand, under the review conditions, a high current value is not required since image formation is not required to be so fast, and a current value of 5 nA or less is sufficient. The pixel size needs to be 1/5 or less of the defect size in order to observe the shape of the defect in detail. Therefore, the thickness is set to 0.02 μm or less. The addition of images is 2 n times, and there is no particular upper limit on the number. In the review, the stage is stationary and the electron beam is scanned two-dimensionally to observe a specific place carefully. In order to further enlarge the defect, the image forming range needs to be small and is 50 μm or less.

In summary, there is a big difference in the concept of pixel size between inspection and review mode. In the case of inspection, there is a defect even if the probe size is larger than the defect size. Or not. That is, it is sufficient that the probe size and the pixel size are equal or larger. If the pixel size is small, there is a problem that the amount of electrons applied to a unit area is small, so that irradiation is performed for a long time, and the clock is slow, so that high-speed inspection cannot be performed. On the other hand, at the time of review, it takes a long time to collect an image of the defect, and a probe of at least 1/5 of the defect size is required. An object of the present invention is to change the probe size at the time of inspection and review, and to change the pixel size at the same time.
Next, how to switch between the defect detection inspection condition and the review condition will be specifically described.
First, the components of the electron optical system and the details of the operation will be described with reference to FIG. The electron source 1 is preferably a field emission electron source, particularly a diffusion-supply type thermal field emission electron source. A so-called Zr / O / W type electron source having a tungsten chip provided with a coating layer made of zirconium and oxygen is used. Was. This electron source can emit electrons stably for a long time. In addition, the emission angular current density can be freely set in the range of 0.001 mA / sr to 1 mA / sr by changing the extraction voltage. However, when the current density is increased, the energy width of the emitted electrons also increases, so that the chromatic aberration increases. At the time of inspection, it was necessary to obtain an electron beam current of 20 nA or more. Therefore, the emission angular current density of the electron beam emitted from the electron source 1 is used in the range of 0.2 to 1 mA / Sr. The electron beam 6 is extracted from the electron source 1 by applying a voltage to the extraction electrode 2. The electron beam 6 is accelerated by applying a high negative potential to the electron source 1. The acceleration voltage can be set to 10kV or more. This is because the use of a large current electron beam suppresses the chromatic aberration caused by the increase in the energy width of the electron beam described above, and suppresses the phenomenon that the large current electron beam spreads due to the interaction of electrons and cannot be focused (Coulomb effect). That's why. The electron beam 6 travels in the direction of the stage 24 with energy corresponding to about 10 kV, is converged by the condenser lens 4, further narrowed down by the objective lens 7, and mounted on the stage 24 (wafer or chip, etc.). ). The distance between the objective lens 7 and the substrate 10 to be inspected (operating distance) was set to 25 mm in order to obtain a deflection area of 50 μm square or more and obtain an image without distortion even in the peripheral part. As a result, the focal length of the objective lens 7 is as long as about 30 mm.
The substrate to be inspected 10 has negative voltage applying means for applying a negative voltage from the high voltage power supply 25. By adjusting the high-voltage power supply 25, it becomes easy to adjust the irradiation energy of the electron beam to the inspection target substrate 10 to an optimum value. For example, if this voltage is -9.5 kV and the acceleration voltage of the electron beam is 10 kV, the irradiation energy to the sample is 500 eV.
For the image formation, a method was employed in which the electron beam 6 scans only one dimension and the stage 24 is continuously moved in a direction orthogonal to the scanning direction.
Signal detection for image formation is performed as follows. That is, secondary electrons are generated by the electron beam 6 applied to the sample 10. Since these secondary electrons are rapidly accelerated by the potential applied to the sample 10, it is very difficult to directly draw them into the detector 9. Therefore, a deflector combining an electric field and a magnetic field, for example, an ExB deflector 14, and a conversion electrode 11 for converting accelerated secondary electrons into low-speed secondary electrons are provided between the substrate 10 to be inspected and the detector 9.
The ExB deflector 14 is a deflector in which the electric field and the magnetic field are orthogonal to each other. For the primary electron beam 6 incident on the ExB deflector 14 from above, the deflecting action by the magnetic field and the deflecting action by the electric field are reversed. This deflector is configured to cancel the secondary electrons coming into the ExB deflector 14 from below, and to add the deflection action by the magnetic field and the deflection action by the electric field.
After the secondary electrons are deflected by the ExB deflector 14, they are irradiated on the conversion electrode 11. Then, secondary electrons generated from the conversion electrode 11 are detected by the detector 9. As the detector 9, a PIN-type semiconductor detector was used to realize high-current and high-speed detection of 10 nA or more. The secondary electron signal detected here is amplified by the preamplifier 12, is AD-converted by the detection circuit 30, and stores the data corresponding to the two-dimensional image of the sample obtained from the secondary electron signal. It is sent to 19. The substrate to be inspected 10 is moved by the movement of the stage 24, the electron beam 6 from the field emission cathode, which is the electron source 1, is irradiated on the substrate to be inspected 10, and the comparison inspection is executed using the secondary electron signal emitted.

  According to the above-described method, a defect is detected by comparing and judging an image of a repeated portion of the same pattern adjacent to the semiconductor pattern at a distance of several μm at a high speed by the calculation unit 20 and the defect determination unit 21. Defects can also be detected by comparing images of different chips of the same pattern on different chips.

  Next, a method of switching to the review condition will be described.

  The factors that limit the resolution under the defect detection inspection conditions are mainly 1) chromatic aberration caused by focusing electrons with varied energy by a lens, and 2) Coulomb caused by repulsion of electrons due to high current density. Effect, 3) The finite light source diameter caused by the fact that the tip of the electron source is not a point light source. The basic concept of switching to the review conditions is to reduce the electron beam current lower than that at the time of defect detection inspection to suppress factors that hinder narrowing of the electron beam, such as optical system aberration and Coulomb effect, And thereby to obtain a high-resolution image. Specific methods are listed below.

The first is a method of reducing the current density drawn from the electron source. FIG. 6 shows a configuration diagram of the electron gun. The electron source 1 is a so-called Zr / O / W type Schottky electron source in which a coating layer made of zirconium and oxygen is provided on a tungsten chip. The filament of the electron source 1 is heated by passing a current from the heating power supply 405. The tip of the electron source 1 protrudes from a central hole of the suppressor electrode 403 for suppressing unnecessary thermoelectrons. A negative potential with respect to the electron source 1 is given to the suppressor electrode 403 by the Vs power supply 407. The tip of the electron source 1 faces the extraction electrode 2, and a positive potential with respect to the electron source 1 is applied to the extraction electrode 2 by the extraction power supply 406. Each of the power supplies 405, 406, and 407 is floated negatively by a desired acceleration voltage by the acceleration power supply 408. As a result, the electron beam extracted from the electron source 1 is accelerated until the anode electrode 404 at the ground potential. In order to reduce the electron current density extracted in the electron gun having the above configuration, the extraction electric field at the tip of the electron source may be reduced. This can be realized by a method of increasing the voltage value of the power supply 407 connected to the suppressor electrode 403 and a method of decreasing the potential of the extraction electrode 2.
Next, the reason why aberration can be reduced by reducing the current density will be described. FIG. 7 shows the current density per emission angle of the electrons emitted from the electron source and the half width (hereinafter referred to as energy width) of the variation in the energy of the beam. The energy width causes the chromatic aberration of the lens of the electron optical system, and the relation is that the energy width and the chromatic aberration are proportional as shown in Expression 1. Therefore, if the energy width is reduced to １ ／, the chromatic aberration is also reduced to １ ／. Under the defect detection inspection conditions, the emission angular current density is used at 0.1 mA / sr to 1 mA / sr, and at the time of review, it is used at 0.001 to 0.01 mA / sr. As a result, the beam current is about 100 nA at the time of defect detection inspection and about 5 nA or less at the time of review, and the energy width is reduced to 1/3 and the chromatic aberration is reduced to 1/3 from FIG. As a result, the beam diameter becomes about 1/3, and the resolution is improved.

  A second method of improving the resolution by reducing the current is a method of reducing the aberration by replacing the diameter of the stop (for example, the stop 5 in FIG. 1) with a smaller diameter. As shown in Equation 1, the chromatic aberration of the optical system is proportional to the beam opening angle α. Therefore, if the diameter of the stop for determining the beam opening angle is reduced to １ ／, the chromatic aberration is also reduced to １ ／. If the opening angle is too small, the diffraction aberration in inverse proportion to the opening angle increases. However, if the opening angle is about 1 mrad or more, the diffraction aberration is not a problem, so that the beam diameter is almost halved. However, this method requires a mechanical diaphragm moving mechanism, and has many problems in terms of reliability and simplicity.

  As a third method, a method will be described in which the focal length of a lens is changed to change the magnification of an optical system, thereby reducing the beam opening angle without moving the stop. As a result, the size of the light source of the electron source can be reduced, which is superior to the second method. This will be described with reference to FIG. The electron beam emitted from the electron source is once focused (referred to as crossover) by the condenser lens under the defect detection inspection condition, and is again focused on the sample by the objective lens. On the other hand, at the time of review, by reducing the lens action of the condenser lens to zero, the take-in angle α from the electron source becomes narrow, and the irradiation angle β of the objective lens also becomes narrow. As a result, the chromatic aberration of the objective lens is reduced along with the decrease in the current value, and the resolution is improved. In addition, the magnification of the entire optical system at the time of defect detection inspection was (b / a) * (d / c), but at the time of review it was reduced to d / (a + b + c). As a result, the resolution can be further improved. The condenser lens uses an electromagnetic lens and includes a coil and a magnetic path. Since the strength of the lens action can be controlled by the value of the current flowing through the coil, the current may be set to zero when the lens action is made zero. In addition, the case where the action of the condenser lens is set to zero has been described. However, similar effects can be expected by setting the lens condition so weak that a crossover is not formed even if the action is not completely set to zero.

It is also possible to improve the resolution by increasing the strength of the condenser lens as compared with the defect detection inspection and moving the crossover position above the stop. This is shown in FIG. The effect obtained is the same as when the lens action of the condenser lens is made zero. That is, the magnification of the optical system is represented by (b / a) * (d / c). As can be seen by comparing FIGS. 8 and 9, b is smaller and c is larger than at the time of the defect detection inspection. Becomes smaller. Further, the irradiation angle β can be reduced.
The method of forming a high-resolution image at the time of review by reducing the beam current has been described above. On the other hand, when the beam current is reduced, the signal that can be detected by the detector is greatly reduced. Therefore, if the amplification factor of the detection system is the same as that at the time of the defect detection inspection, the input signal becomes less than the minimum bit of the detection circuit 30, and the electron beam is scanned on the sample a plurality of times to perform image addition. However, a good image cannot be obtained.
Therefore, another circuit was prepared from the preamplifier to the input to the AD converter. That is, two signal paths are provided in the detection circuit 30 in FIG. The configuration is shown in FIG. At the time of defect detection inspection, a signal from the detector 9 is directly input to the A / D converter 304 via the low gain amplifier 301. On the other hand, at the time of review, the signal from the detector 9 decreases because the beam current value decreases. The signal path is switched to the high gain amplifier 302 having a higher gain. Further, the signal is input to the AD converter 304 via the high-frequency cut filter 303. The operation of this circuit will be described with reference to FIG. 16 which is a schematic diagram of signal waveforms at the time of defect detection inspection and at the time of review.

The case where the beam current at the time of defect detection inspection is 100 nA and the beam current at the time of review is 500 pA is shown as an example. The gain ratio between the low gain amplifier 301 and the high gain amplifier 302 was set to 200 which is equal to the ratio of the beam current. As a result, the average value of the outputs of both amplifiers becomes equal (arbitrary unit 100). FIG. 16A shows a signal waveform at the time of a defect detection inspection when an image is formed on a non-patterned portion of a semiconductor. The noise amplitude is small because a large current beam is used. On the other hand, FIG. 16B shows a signal waveform at the time of review of the same part.
The average of the signal is almost 100, but since a small signal is amplified with a high gain, the noise is large and swings from -100 to 300. When the analog signal fluctuates to the minus side, the digitized value becomes zero. Also, if it is too large, it will exceed the full scale of the A / D converter. Accordingly, there is a problem that an accurate waveform cannot be obtained even when digital signals are averaged. To solve this, a high-frequency cut filter 303 is placed after the high-gain amplifier. The noise of the input signal to the A / D converter 304 is reduced as shown in the waveform of “with filtering” in FIG. Further, after this waveform is digitized, an image signal having a sufficient SN ratio can be obtained by scanning the electron beam a plurality of times and adding the signals.
The three methods for improving the resolution and the method for solving the problem caused by reducing the beam current have been described above. In the present embodiment, in order to improve the resolution, the third method of mainly changing the excitation condition of the condenser lens is used, and the first method of varying the current value drawn from the electron source is used in an auxiliary manner. Specific numerical examples are shown below.
At the time of defect detection inspection, the emission angular current density from the electron source was 0.5 mA / sr, the magnification of the optical system was 1, and the beam current was 100 nA. The resolution of the image at this time was about 0.08 μm. FIG. 10 shows a line profile of the image of the line and space pattern at this time.
In the review, the condenser lens current value was increased and the magnification was set to 0.2. Furthermore, the high-voltage power supply of the electron gun was controlled, and the emission angular current density from the electron source was reduced to 0.1 mA / sr. As a result, the beam current was reduced to 500 pA, so the gain of the preamplifier of the detection system was switched to 200 times, and the beam passed through a filter that reduced the frequency characteristic to about 1/10. The scanning speed of the electron beam was set to about 1/10 of that during the defect detection inspection. Further, in order to compensate for the deterioration of the SN ratio, the image addition is performed 64 times. The resolution of the image at this time was about 0.02 μm. The line profile of the image at this time is shown in FIG. It can be seen that the rising of the line profile at the time of review is sharp and the amplitude is also large. From this, it can be confirmed that the resolution is clearly improved and even a fine object can be observed.
(Example 2)
In the first embodiment, the same detector is used for the review and the defect detection inspection. However, the gain of the preamplifier 12 needs to be increased by a factor of 200 to 1000 by the rate at which the current decreases because the detection signal sharply decreases during the review. Further, if only the gain is multiplied while the frequency characteristic is high as in the defect detection inspection, the circuit oscillates or the noise amplitude of the circuit becomes larger than the average signal amount. For this purpose, it is necessary to add a circuit for cutting high frequencies, which complicates the circuit. As a result, there is a possibility that noise is mixed in the detection signal at the time of the defect detection inspection or the frequency characteristic is deteriorated. In order to solve this, in the present embodiment, a detector dedicated for review is provided. FIG. 11 shows the configuration diagram. A review detector is provided at a position substantially symmetrical with respect to the optical axis of the electron beam of the defect detection / inspection detector 9. This detector is used in an ordinary scanning electron microscope, and includes a scintillator 52 and a photomultiplier tube 51. The response characteristic is DC to 20 MHz. Secondary electrons emitted from the sample 10 are deflected by the ExB deflector 14 in the direction of the defect detection / inspection detector 9 at the time of defect detection inspection, and in the direction of the review detector at the time of review. The secondary electrons are applied to the conversion electrodes facing the respective detectors, and the secondary electrons emitted thereby are detected. A high positive voltage is applied to the front surface of the scintillator 52 by a power supply 53 to accelerate and collide secondary electrons to convert them into light. The light is multiplied by a photomultiplier tube 51 and detected. The switching of the secondary electron deflection direction is achieved by reversing the polarity of the control power supply 31 of the ExB deflector 14 (reversing the polarity of the applied voltage and the deflection coil current). In conjunction with this, a signal from one of the detectors is displayed on the monitor 22 via the memory 18. The switching is performed by the control unit 104.

(Example 3)
Depending on the type of defect, the detected defect contrast may disappear if the review conditions are not made appropriate. For example, there are cases where the contrast of the defect disappears during the irradiation of the beam a plurality of times, among the conduction defect of the conduction hole. Therefore, when reviewing such defects, it is necessary to limit the number of additions of images. Since the number of additions is limited, it is necessary to increase the beam current in order to obtain an image with a good SN ratio. Also, depending on the type of defect, the contrast of the defective portion is greater when the electric field (electric field for decelerating the electron beam applied to the inspection wafer) is stronger in the case of a large current beam during defect detection inspection. In some cases, the contrast is enhanced by weakening the electric field. Hereinafter, a condition setting method for reviewing a conduction defect will be described as an example.

Semiconductor processes include the step of drilling holes and embedding a conductive material therein to establish vertical conduction between the upper and lower layers. If there is a defect in the drilling or embedding, it becomes completely electrically open or has a very high resistance value. When such a process is inspected for defect detection, conductive holes are bright and non-conductive holes are dark. Further, in the case of a delicate high resistance value, the brightness becomes intermediate, so that it is detected as a defect by comparing with a hole having perfect conduction. If such defects are reduced in current during the review and the image addition is performed several tens of times, the brightness becomes the same as that of a hole that is completely conductive, and no information is often obtained even after the review.
Therefore, when conducting a defect detection inspection of a conduction hole, the brightness level is stored in combination with defect coordinates for each detected defect, and when a gray defect is reviewed, the beam current is automatically reduced to 5 nA or less. The optical conditions were set so as not to cause the above, and the number of image additions was suppressed to 10 or less. On the other hand, when reviewing defects at very dark levels, the beam current was reduced as much as possible, and the review was performed under conditions of 500 pA or less.

  The above is only an example, and the point is that the types of defects are automatically classified using information obtained from the image at the time of defect detection inspection, that is, brightness, defect size, shape, etc. , The scanning range, beam current, scanning speed, and number of image additions are automatically associated. At this time, in the present invention, the change of the optical condition is not accompanied by the mechanical movement, that is, the movement of the stop, etc., so that the condition can be instantaneously switched for each defect.

(Example 4)
As a fourth embodiment of the present invention, the review electron optical system is independent of the defect detection / inspection electron optical system. This will be described with reference to FIG. FIG. 12 is a configuration diagram of the present embodiment.

  The inspection apparatus is roughly composed of a defect detection / inspection electron optical system 101, a review electron optical system 200, a sample chamber 102, a control unit 104, and an image processing unit 105. Except for the review electron optical system 200, it is substantially the same as the first embodiment.

  The review electron optical system includes an electron source 201, an electron beam extraction electrode 202, a condenser lens 204, a scanning deflector 208, an aperture 205, and an objective lens 207. The secondary electron detector 209 is located above the objective lens 207, and the output signal of the secondary electron detector 209 is amplified by the preamplifier 212 and sent to the low-speed image display circuit 218.

  A retarding voltage is applied to the sample stage 24, but when the stage moves from under the defect detection / inspection optical system 101 to under the reviewing electronic optical system, the retarding voltage is once interrupted by the structure of the stage. In some cases, it is preferable to move after that, so that the selection may be made according to the applied retarding voltage or the type of the substrate to be inspected.

  Next, the specifications, differences and features of the review electron optical system and the defect detection / inspection optical system which are the points of this embodiment will be described.

  The electron optical system 101 for defect detection inspection can acquire an image at a very high speed because the entire surface of the wafer is inspected in a practical time. That is, the time for forming an image of 100 μm square with a pixel size of 0.1 μm or less is 20 msec or less. The stage is continuously moved for high-speed image acquisition, and the electron beam is scanned in the direction orthogonal to the stage. The beam current was designed to be 20 nA or more in order to realize an image formation with an SN ratio that can withstand inspection at high speed. Specifically, first, the current density obtained from the electron source 1 was set to about 1 mA / sr, which is a limit value that can be obtained stably. This is about 20 times higher resolution SEM. Further, the aperture of the lens is made larger than that of the high-resolution SEM, and the sample 10 of the electron beam emitted from the electron source 1 at a wide angle (ie, the electron beam emitted from the electron source 1 at a wide angle) is irradiated. Opening angle of the electron beam to be used) was set to about 20 times.

  In addition, if the peripheral portion of the image is distorted or the resolution is lower than that of the central portion, the inspection sensitivity becomes non-uniform. Therefore, it is necessary to provide a margin in the scanning region of the electron beam. For this reason, we adopted an objective lens whose focal length and operating distance were much longer than those of a normal high-resolution SEM.

  Further, when the electron beam opening angle β of the electron beam becomes large, the depth of focus becomes shallow. Since β is obtained by dividing α by the magnification M of the entire optical system, the magnification M cannot be made very small. Therefore, the magnification is considerably larger than that of the high-resolution SEM.

  With the above design, a high current required for high-speed operation can be obtained while securing the resolution required for the defect detection inspection. As described above, even if the electron beam current is simply reduced, this electron optical system cannot obtain the same resolution as a high-resolution SEM, because of the magnification of the optical system, the focal length of the objective lens, and the operating distance. it is obvious.

On the other hand, high speed is not so important in the review electron optics. Also, since the coordinates of the place to be observed are accurately known, a field of view of 20 μm is sufficient. Therefore, the electron beam current may be small. In addition, since the focal length and operating distance of the objective lens may be short and the magnification of the optical system can be reduced, a high-resolution image similar to a normal high-resolution SEM can be obtained.
Next, the review electron optical system 200 will be described. As the electron source 201, a Zr / O / W type electron source, which is a diffusion-supply type thermal field emission electron source similar to the electron optical system for defect detection inspection, is used. When the current density is increased, the energy width of the emitted electrons also increases, and the chromatic aberration increases. In the review electron optical system, the beam current may be 100 pA or less, so the emission angular current density is set to 0.05 mA / Sr or less. As a result, the energy width of the electron beam 206 is reduced to about 1/3 to 1/4 as compared with the case of the electron optical system for defect detection and inspection, and the chromatic aberration is reduced accordingly. The electron beam 206 is extracted from the electron source 201 by applying a voltage to the extraction electrode 202. The acceleration of the electron beam 206 is performed by applying a high negative potential to the electron source 201. In the review optical system 200, the acceleration voltage can be varied from 500V to 10kV. In the review optical system 200, since the current may be small, the chromatic aberration is small and the Coulomb effect can be neglected as described above. Therefore, a sufficiently small beam diameter can be obtained with a low accelerating voltage, and the accelerating voltage is set to 2 kV as a standard. The electron beam 206 advances toward the sample stage 24 with an energy equivalent to 2 kV, is converged by the condenser lens 204, further narrowed down by the objective lens 207, and mounted on the stage 24 (wafer or chip). Is irradiated. The objective lens 207 was very close to the substrate 10 to be inspected, and the operating distance was 5 mm. The focal length of the objective lens 207 is 8 mm. As a result, an objective lens with small aberration was realized. Resolution is mostly determined by chromatic aberration. Therefore, the resolution is about 1/12, because the aberration coefficient is about 1/4 and the energy width of the electron beam is about 1/3 of that of the defect detection and inspection optical system compared with the defect detection and inspection objective lens. The substrate to be inspected 10 is common for the defect detection inspection and for the review, and therefore a negative voltage can be applied by the high voltage power supply 25. By adjusting the high-voltage power supply 25, the irradiation energy of the electron beam to the substrate 23 to be inspected can be adjusted to an optimum value. Since images having different contrasts can be obtained by changing the irradiation energy, information such as whether the defect is caused by a difference in shape, by a material, or by electrical continuity can be obtained from various angles.
For image formation, the electron beam 206 is scanned two-dimensionally. That is, the stage 24 is fixed. Secondary electrons generated by the electron beam 206 applied to the substrate 10 to be inspected are accelerated by a potential applied to the substrate 10 to be inspected, similarly to the optical system 101 for defect detection and inspection. The detector 209 is provided above the objective lens 207 because the objective lens 207 is very close to the substrate under inspection 30 mm or less, and secondary electrons pass through the center of the objective lens 207. Since the potential applied to the substrate 10 to be inspected is lower than that of the optical system 101 for defect detection and inspection, the energy of secondary electrons is weak, but it is still difficult to directly draw the electrons into the detector 209. Then, the conversion electrode 211 is irradiated, and secondary electrons emitted therefrom are detected by the detector 209. As the detector 209, a detector 209 using a phosphor and a photomultiplier used in a normal SEM was used. The secondary electron signal detected here is amplified by the preamplifier 212 and AD-converted by the AD converter 230, displayed on the monitor 22 for image observation, and stored as an image file in the external storage device 219 such as a disk as necessary. It can be printed out.

Although the fourth embodiment has been described above, the numerical values of each function are merely examples. In short, an electron optics system for review that can achieve a beam diameter of about 1/10 compared to the electron optics system for defect detection and inspection should be arranged side by side on the same sample chamber as the electron optics system for defect detection and inspection. It is the essence of the present embodiment that the operation of the defect detection inspection and the review can be quickly switched only by the parallel movement of the sample stage.
(Example 5)
In the fifth embodiment, an X-ray detector 240 is incorporated in the lower magnetic path of the objective lens 207 of the review electron optical system 200 of the fourth embodiment so that characteristic X-rays generated by electron beam irradiation can be detected. did. As a result, EDX (Energy-Dispersive X-ray) analysis becomes possible, and the material of the defect can be specified at the time of review. FIG. 13 is a diagram showing the vicinity of an objective lens when an annular X-ray detector having a hole at the center is incorporated.

Also, the X-ray analyzer may be inserted between the objective lens and the sample by making the operating distance of the objective lens slightly longer than in the first embodiment.
(Example 6)
In the inspection using an electron beam, it is possible to inspect not only shape defects but also electrical continuity and non-conduction. This is because the conductive portion is not charged, and the energy and orbit of the secondary electron generated by charging only the non-conductive portion changes, resulting in a different brightness of the image. This occurs as a particularly large contrast due to the irradiation of the high-current electron beam. However, in the review electron optical system, the difference in electrification is small because the current is small, and it is difficult to appear as a contrast difference. Therefore, a secondary electron energy filter capable of filtering even a small difference in the energy of secondary electrons with high sensitivity is provided. The energy filter was provided on the objective lens. A diagram of this analyzer is shown in FIG. A hemispherical mesh 220 is provided above the objective lens 207 so as to block the secondary electron trajectory. A hole through which the primary electron beam 206 passes is formed in the center of the mesh 220. A potential in the range of about ± 20 V from the potential of the substrate 10 to be inspected is applied to the mesh 220 by the power supply 221. As a result, the state of charging of the substrate to be inspected can be imaged. The secondary electrons have an energy distribution with a peak at about 2 eV with respect to the potential φ at the place of emission. Therefore, the energy peak of the secondary electrons viewed from the ground potential is (−φ + 2) eV. For example, assuming that a retarding potential of 500 V is applied to the substrate to be inspected 10, the peak of the secondary electron energy is 502 eV. If the primary beam irradiation position is originally a place to be electrically connected to the substrate 10 and only that part is non-conductive due to a pattern defect, the part is negatively charged. Then, the energy of the secondary electrons emitted therefrom is higher by the charged voltage. If the charging voltage is 5 V, it is 507 eV. Therefore, if -505 V is applied to the mesh 220, most of the secondary electrons from the uncharged portion cannot pass through the mesh 220. Therefore, only the charged portion looks bright. In this way, a slight charge can be obtained as a contrast difference. Here, only the case of negative charging has been described, but in an actual semiconductor, the mechanism of charging by electron beam irradiation is complicated, and positive charging may occur in some cases. In that case, the potential applied to the mesh 220 may be equal to or slightly higher than the retarding potential. In that case, the portion charged positively becomes dark and the other portions become bright.

  The resolution of the energy analyzer of this embodiment can be about 0.1 V. Therefore, at the time of review, even if the charge amount of the defective portion is small because the electron beam current is small, the contrast by the potential is emphasized, so that the defect can be easily specified.

FIG. 1 is a diagram illustrating an example of a device configuration according to a first embodiment of the present invention. FIG. 4 is a diagram illustrating a modification of the device configuration according to the first embodiment of the present invention. It is a figure explaining an example of the flow of the inspection of the present invention. It is a figure explaining other examples of the inspection flow of the present invention. FIG. 3 is a diagram illustrating conditions for inspection and review. FIG. 2 is a diagram illustrating a part of the first embodiment of the present invention. FIG. 3 is a diagram illustrating the operation principle of the first embodiment of the present invention. FIG. 3 is a diagram illustrating the operation principle of the first embodiment of the present invention. FIG. 3 is a diagram illustrating an example of an operation principle of the first embodiment of the present invention. FIG. 3 is a diagram illustrating the effect of the first embodiment of the present invention. FIG. 6 is a diagram illustrating a configuration of a second exemplary embodiment of the present invention. FIG. 9 is a diagram illustrating a configuration of a third exemplary embodiment of the present invention. FIG. 9 is a diagram illustrating a configuration of a fourth embodiment of the present invention. FIG. 14 is a diagram illustrating a configuration of a fifth embodiment of the present invention. FIG. 2 is a diagram illustrating a configuration of a detection circuit according to the first embodiment of the present invention. FIG. 3 is a diagram for explaining signal waveforms at the time of defect detection inspection and at the time of review in Example 1 of the present invention.

Explanation of reference numerals

1: electron source, 2: extraction electrode, 3: insulator, 4: condenser lens, 5: aperture, 6: electron beam, 7: objective lens, 8: scanning deflector, 9: secondary electron detector, 10: coated Inspection board, 11: conversion electrode, 12: preamplifier, 13: scanning signal generator, 14: ExB deflector, 17: blanking deflector, 18: image storage unit, 19: image storage unit, 20: arithmetic unit, 21 : Defect judgment part, 22: Monitor, 24: Stage, 25: High voltage power supply, 26: Objective lens power supply, 27: Correction control circuit, 29: Optical sample height measuring instrument, 30: AD converter, 31: Stage position Measuring instrument, 32: Preamplifier, 33: AD converter, 34: Transmitter, 35: Receiver, 101: Inspection electron optical system, 102: Sample chamber, 104: Control unit, 105: Image processing unit, 200: For review Electron optical system, 201: electron source, 202: extraction electrode, 204: condenser lens, 205: aperture, 206: electron beam, 207: objective lens, 208: scanning deflector, 209: detector, 211: conversion electrode, 212 : Preamplifier, 218: Image storage unit, 219: external storage device, 220: mesh, 221: power supply, 230: AD converter, 240: X-ray detector, 403: suppressor electrode, 404: anode, 405: heating power supply, 406: extraction power supply, 407: suppressor power supply, 408: acceleration power supply.

Claims (7)

In a circuit pattern inspection method using a scanning electron microscope operable in both a first observation mode and a second observation mode having a higher resolution than the first observation mode,
Observing the circuit pattern with the scanning electron microscope,
Detecting a specific portion of the circuit pattern,
Switching the operation mode of the scanning electron microscope from the first mode to the second mode;
Observing the specific location in the second observation mode,
A method for inspecting a circuit pattern, comprising displaying an image of the observed specific portion on a display means. The circuit pattern inspection method according to claim 1,
The sample stage holding the sample on which the circuit pattern is formed is continuously moved,
Observing the circuit pattern in the first mode,
After detecting a specific portion of the circuit pattern,
Switching an operation mode of the scanning electron microscope to the second mode based on input information of a device user;
Stationary the sample stage,
A method for inspecting a circuit pattern, comprising: observing the specific portion in the second observation mode while the sample stage is stationary. The circuit pattern inspection method according to claim 1,
In the first mode, the beam current of the electron beam is set to a first value, and the electron beam is scanned at a first speed,
In the second mode, the beam current of the electron beam is set to a smaller value than in the first mode, and the electron beam is scanned at a second speed lower than the first speed. Inspection method of the characteristic circuit pattern. The circuit pattern inspection method according to claim 1,
A circuit pattern inspection method, wherein observation conditions of a scanning electron image including radiant energy of an electron beam current and a pixel size are set before starting observation in the first mode. The circuit pattern inspection method according to claim 1,
A method for inspecting a circuit pattern, comprising: after detecting a specific portion of the circuit pattern, reducing the size of the electronic probe to be smaller than the specific portion of the circuit pattern. The circuit pattern inspection method according to claim 4,
A circuit pattern inspection method, wherein the size of the electronic probe is reduced to one fifth or less of the size of the specific portion. The circuit pattern inspection method according to claim 1,
In the first mode, the sample stage is moved in a first direction,
And an electron beam used in the scanning electron microscope is scanned on the circuit pattern in a second direction orthogonal to the first direction.

Images