JP6662654B2 - Image acquisition method and electron beam inspection / length measuring device - Google Patents

Description

  The present invention relates to an image acquisition method and an electron beam inspection / length measurement device, and more particularly to an image acquisition method for acquiring an image of a predetermined region on a sample as a panoramic image by connecting a plurality of child field-of-view images.

  Inspection and length measurement devices for inspecting and measuring fine patterns on semiconductor wafers and masks use electron beams in a vacuum. With the recent miniaturization of semiconductor devices, inspection and measurement of wafers and masks have become increasingly difficult. Its importance is growing. As a length measuring / inspection apparatus using such an electron beam, an electron beam inspection apparatus (EBI: Electron Beam Inspection) for detecting a pattern defect, and a defect candidate detected by the electron beam inspection apparatus (EBI) are observed in detail. There are a review device (Review SEM) and a dimension measuring device (CD-SEM) for measuring a pattern dimension.

  The electron beam inspection apparatus (EBI) has a problem that the inspection speed is extremely slow although the resolution is higher than that of the conventional optical system. On the other hand, the review device (Review SEM) also needs to increase the review speed and detect the electron beam in order to reliably detect minute pattern defects while detecting various signals such as reflected electrons, secondary electrons, and X-rays. It is required to increase the resolution by reducing the beam diameter. On the other hand, a dimension measuring device (CD-SEM) is required to reduce the beam diameter of an electron beam to a limit in order to accurately measure a dimension.

  In particular, in an electron beam inspection apparatus (EBI), in order to increase the inspection speed, it is required to widen the field of view (FOV) as much as possible, select a large beam current and scan the inspection area at high speed. A region where the probability of occurrence of a defect is predicted in advance by a computer or the like is roughly observed with a large field of view (FOV), a large beam diameter, and a large beam current. (Review SEM) or the like can observe the area with a small field of view (FOV) and a small beam diameter, and can observe defects with high accuracy. In this manner, the inspection accuracy is generally improved.

  When a pattern becomes finer, in optical lithography, correction of a complicated pattern shape such as OPC or multi-patterning is required. When such a complicated correction is performed, a defect (systematic defect) is likely to occur in a specific fragile pattern if the defocus of the scanner (light exposure device) or the exposure amount fluctuates. In order to detect a defect based on such an LSI pattern design, systematic defect detection (D: DB inspection) is important by comparing an inspection apparatus image or a review apparatus image with a theoretical simulation image of a wafer pattern. Has become. This D: DB inspection requires high-resolution image information. For this purpose, a small field of view (FOV) free from the influence of deflection aberration is used. However, even if a narrow area is observed with high precision and analyzed using the D: DB function, a good analysis result cannot be obtained because systematic defects are affected by a wide area. This is because a distant pattern has a strong influence on the pattern of interest. For this reason, from the viewpoint of high-precision inspection of D: DB, it has been required to acquire an image with a large field of view (FOV).

  As a conventional example, for example, Patent Literature 1 discloses that a plurality of images having a relatively small field of view (FOV) are obtained, and a plurality of images having a small field of view (FOV) are connected to obtain a panoramic image. . Here, a predetermined region on the sample is divided into a plurality of regions, and the divided regions are irradiated with an electron beam and scanned to obtain an image with a small field of view (FOV). Then, the stage supporting the sample is moved in a direction (X direction and Y direction) orthogonal to the optical axis, and the other partitioned area is irradiated with an electron beam and scanned to form a next small field of view (FOV). ) To get the image. By repeating such steps, a plurality of images having a small field of view (FOV) are acquired. However, when a panoramic image is obtained in this way, the time for finely moving the stage and the time for re-adjustment of the electron beam due to the movement of the stage are added by the number of images with a small field of view (FOV) to be obtained. However, there is a problem that it takes a long time to obtain a panoramic image. This is because, in a large area field of view (FOV), the pattern is distorted in the periphery and the resolution is reduced. Therefore, an image is obtained by dividing the small area field of view (FOV) into sections, and the small area field of view (FOV) is obtained. This is because a panoramic image has to be formed by synthesizing the FOV) image. With such a method, an image with a large field of view (FOV) can be obtained, but from the viewpoint of speeding up, it is a completely backward method.

  In order to obtain an image with a large field of view (FOV), it is conceivable to use a deflector having a large deflection width. However, to increase the deflection width, it is necessary to apply a high deflection voltage to the deflector. . The scanning speed of the electron beam decreases as the deflection voltage increases, and the processing speed decreases. Also, when trying to obtain an image of a field of view (FOV) of a large area, the farther away from the center of the field of view (the closer to the periphery of the above-mentioned field of view), the more the image is distorted due to deflection distortion, and the beam is blurred due to field curvature, The beam shape changes due to the deflection astigmatism, and a high-resolution image cannot be obtained.

JP 2010-67516 A

  In view of the above problems, an object of the present invention is to provide an image acquisition method and an electron beam inspection and length measurement device capable of acquiring an image with a large field of view (FOV) at high speed.

  In order to solve the above-described problems, an image acquisition method of the present invention for acquiring an image of a predetermined region on a sample with an inspection / length measuring device using an electron beam as a panoramic image by connecting a plurality of child field-of-view images, A predetermined area on the sample is divided into a plurality of areas, and the divided area is used as a child field of view to obtain a child field of view image, and a panorama is obtained by connecting the obtained plural child field of view images. A panoramic image acquiring step of acquiring an image, wherein the sub-field image acquiring step includes the step of scanning the electron beam in the sub-field of view by a sub-deflector provided in an electron optical system including a source of an electron beam. And a second step of moving the electron beam to the next child field region by the main deflector having a larger deflection width than the sub deflector further provided in the electron optical system, And repeating the second step to acquire each child visual field image based on at least one of the secondary electron information and the reflected electron information generated from the sample, the control information of the sub deflector, and the control information of the main deflector. Features.

  According to the present invention, when acquiring an image of a certain child field of view, the electron beam is scanned by the sub deflector. Since the deflection width of the sub deflector is smaller than that of the main deflector, the deflection voltage applied to the sub deflector can be reduced, and the electron beam can be scanned at high speed by the sub deflector. Then, when acquiring an image of the next child viewing area, the electron beam is moved by the main deflector. Since the deflection width of the main deflector is larger than that of the sub deflector, the deflection voltage applied to the main deflector increases, and the moving speed of the electron beam by the main deflector becomes slower than the scanning speed of the electron beam by the sub deflector. However, as compared with the conventional example, the moving speed of the electron beam by moving the stage (physical method) is overwhelmingly high. Specifically, while the electron beam moving time by the stage movement between the child visual field regions according to the conventional example is about 1 sec, the electron beam moving time by the main deflector between the child visual field areas is about 0.5 μsec. Overwhelmingly short. As described above, by scanning and moving the electron beam by the electric method of double deflection of the sub deflector and the main deflector, a panoramic image can be obtained at a higher speed than in the above-described conventional example.

By the way, when an electron beam is deflected in a large field of view, a change in pattern position due to deflection distortion caused by the deflection, a field curvature, and a decrease in resolution due to deflection astigmatism become remarkable. For this reason, in the above-described conventional example, a panoramic image is formed by combining the acquisition of an image by dividing it into a small deflection child field of view (FOV) and the movement of the stage. On the other hand, in the present invention, the child visual field image acquiring step divides the child visual field area so that the variation amount of the deflection aberration falls within a predetermined allowable range, and acquires an image of the child visual field area, and then the main deflector. The electron beam is moved to the next child viewing area. Before the movement of the electron beam to the next child field area is completed , deflection aberration correction is performed. Here, in the present invention, the deflection aberration refers to deflection distortion, field curvature and deflection astigmatism. In this case, the deflection distortion, the field curvature and the deflection astigmatism accompanying the large deflection by the main deflector are corrected by exchanging the data of the deflection distortion, the field curvature and the deflection astigmatism.

  Further, in the present invention, a Z-map creation step of creating a Z-map defining a Z-value in the sample plane by defining a direction orthogonal to the upper surface of the stage as a Z-direction and a height position in the Z-direction as a Z-value, Using a calibration pattern formed on a stage that can be freely moved, or a plurality of calibration patterns formed on the stage and having different Z values, corresponding to discrete Z values, the deflection aberration is reduced within the predetermined area. Preferably, the method further includes a correction map creating step of measuring each of the measured values and creating a corrected deflection aberration map defining a correction amount of the deflection aberration from these measured values for each Z value. Note that the Z map is preferably created and stored before image acquisition. However, if height detection can be performed at high speed, the Z value in a predetermined area may be measured in real time at the time of image acquisition. Then, the Z value of the predetermined area is obtained from the Z map, and the deflection aberration is corrected for each child field area in the predetermined area using the corrected deflection aberration map corresponding to the obtained Z value. According to this, in accordance with the Z value (height) of the predetermined area, deflection distortion, curvature of field, and deflection astigmatism are corrected for each child field area in the predetermined area. For this reason, deflection aberration accompanying large deflection by the main deflector is corrected, and a high-resolution panoramic image can be obtained. Moreover, by using the corrected deflection aberration map, the deflection aberration can be corrected in a short time until the electron beam is moved to the next child field area.

  Further, in the present invention, when image information is acquired a plurality of times and integrated to obtain one image information, a part of an adjacent child field of view may be shifted in the X and Y directions at the time of the first and the plurality of image acquisitions. It is preferable to obtain an image at a time. According to this, it is possible to reduce a joint error of the child visual field region and obtain a high-precision panoramic image, which is advantageous.

In order to solve the above problems, an electron optical system including a source of an electron beam, and a stage that holds a sample to be irradiated with the electron beam and is movable in a direction orthogonal to the optical axis of the electron optical system. An electron beam inspection / measurement device according to the present invention, comprising: an image acquisition unit that acquires an image based on secondary electrons generated from a sample; and an inspection unit that inspects and measures the length based on the image acquired by the image acquisition unit. The long device divides a predetermined area on the sample into a plurality of sections, uses the partitioned area as a sub-field of view, a sub-deflector for scanning an electron beam in each field of view, and moves the electron beam between the sub-fields of view. A main deflector having a larger deflection width than the sub deflector, further comprising a deflection control unit for controlling the sub deflector and the main deflector, wherein the image acquisition unit is configured to output a secondary electron signal from a sample. And at least one of the reflected electron signals When, in Rukoto to synchronize with the control information of the sub-deflector and the main deflector by the deflection control means, the image is acquired a plurality of times, in which obtaining a single image by integrating the child adjacent the viewing area partially characterized that you configured to acquire an image to be shifted in the X and Y directions in the first and at a plurality time of image acquisition.

  In the present invention, a care area acquiring means for acquiring a region to be inspected and measured for a sample as a care area, and a section which divides the care area into a plurality of predetermined areas and divides the plurality of predetermined areas into a plurality of child visual field areas. Means, a Z-direction in a direction perpendicular to the upper surface of the stage, and a Z-value in the Z-direction, and a Z-value acquiring means for acquiring a Z-value in a predetermined area, and a correction amount of deflection aberration for each discrete Z-value. Correction map creating means for creating a prescribed deflection aberration correction map, and corresponding to the Z value acquired by the Z value acquiring means during or before or after moving the electron beam to the next child field area by the main deflector. It is preferable to further include a deflection aberration correction unit that corrects the deflection aberration of the next child field region with reference to the deflection aberration correction map to be performed. In the present invention, the predetermined area refers to an area for acquiring a panoramic image.

FIG. 2 is a diagram illustrating the concept of an image acquisition method according to the embodiment of the present invention. FIG. 2 is a view for explaining the concept of the image acquisition method according to the embodiment. FIG. 1 is a diagram schematically illustrating a schematic configuration of an electron beam inspection / length measuring device according to an embodiment. FIG. 1 is a diagram schematically illustrating an electron beam inspection / length measurement device according to an embodiment. FIG. 4 is a schematic diagram illustrating an example of a calibration mark used for creating a deflection aberration correction map. FIG. 3 is a diagram illustrating the concept of a shifted image acquisition method of CFOV. FIG. 3 is a diagram illustrating the concept of a shifted image acquisition method of CFOV. (A)-(c) is a figure which shows the modification of deflection typically. The figure explaining the application example of the present invention.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. The concept of the image acquisition method according to the present embodiment will be described with reference to FIGS. First, a region to be inspected using an electron beam inspection device (EBI) will be described. In FIG. 1, IB0 is a semiconductor chip as a sample. If the entire semiconductor chip IB0 is inspected using an electron beam inspection device (EBI), it takes an enormous amount of time and is not realistic. Therefore, using a simulation device (computer) independent of the electron beam inspection device, a region to be inspected (hereinafter referred to as “care area”) is extracted from the semiconductor chip IB0 on the data, and the extracted care area is electronically extracted. Input to the beam inspection device (EBI). Here, as the care area, (1) a so-called “hot spot” predicted by a verification tool after OPC (Optical Proximity Correction), (2) an area including a pattern causing a problem in the past or a pattern similar thereto, (3) An area including a pattern recognized as dangerous by a semiconductor device designer, (4) an area including another dangerous pattern, and the like can be exemplified. A computer inspection technique for efficiently and accurately extracting such a care area from data has been developed recently. The pattern of the care area is a potential systematic defect. The care area on the data is physically formed on the wafer through a semiconductor manufacturing process, and the physically formed care area is referred to as a “physical care area” here. FIG. 1 shows three physical care areas IB1, IB2, and IB3. When the physical care area IB1 is inspected, the physical care area IB1 is divided into a plurality of predetermined areas PFOV1, PFOV2,. It is divided into PFOV (n-1) and PFOVn. The plurality of predetermined regions PFOV1,..., PFOVn are regions for acquiring a panoramic image (hereinafter, referred to as “panoramic viewing region”). The panoramic image may be square or rectangular.

  Next, for example, as shown in FIG. 2, the panoramic visual field region PFOV1 is partitioned into, for example, 64 matrixes in the x direction and 64 matrixes in the y direction. The divided areas CFOV101, CFOV102,..., CFOV6464 are used as child visual field areas, electron beams are scanned in these child visual field areas to obtain images, and the obtained images of the child visual field areas are connected. Then, an image of the panoramic visual field (hereinafter, referred to as “panoramic image”) is acquired. The shape of the child field of view CFOV is not limited to a square, but may be a rectangle. Hereinafter, the term “CFOV” includes the case of “CFOV image”. Further, the term “PFOV” includes the case of “PFOV image”.

  The dimension d1 of one child field region CFOV is a size in which the amount of change in deflection aberration (deflection distortion, field curvature, and deflection astigmatism) falls within a predetermined allowable range, and is set to, for example, 3 μm. In the case of FIG. 2, since the number of CFOVs is 64 in each of the X direction and the Y direction, the dimension d2 of the panoramic visual field region PFOV is 192 μm (= 3 μm × 64). Assuming that the pixel size of the CFOV is 3 nm, the number of pixels required to acquire one CFOV image is 1 k × 1 k, and the number of pixels required to acquire one PFOV image (panoramic image) is 64 k. × 64k.

  To obtain a panoramic image, an electron beam is scanned to obtain an image of the CFOV 101, and when that is completed, the electron beam is electrically moved to the next CFOV 102, an image of the CFOV 102 is obtained, and this is continued. To go. The switching speed (moving speed of the electron beam) from the CFOV 101 to the CFOV 102 is about 0.5 μsec, and the moving speed is overwhelming compared to the case where the electron beam is moved mechanically (by moving the stage) as in the above-described conventional example. Speed is fast. Although details will be described later, the deflection aberration of the next CFOV 102 is corrected during this 0.5 μsec. The clock frequency in the CFOV 101 can be set to, for example, about 100 MHz, and the scan time from pixel to pixel is 10 nsec.

  Next, with reference to FIG. 3, a schematic configuration of the electron beam inspection / length measurement device of the present embodiment will be described. In FIG. 3, reference numeral 100 denotes an electron beam inspection and length measuring device, and 101 denotes an optical axis of the electron optical system EO. The electron optical system EO includes a sub-deflector 102, a return sub-deflector 103, a main deflector 104, a return main deflector 105, a dynamic focus lens 106, an objective lens 107, and a secondary electron detector 109. In the drawing, reference numeral 108 denotes an inspection surface (sample surface) of the wafer W as a sample. The electron beam inspection / length measuring apparatus 100 scans an electron beam with a sub-deflector DAC / amplifier 111 that drives the sub-deflectors 102 and 103 and main deflectors 104 and 105 as a control system of the electron optical system EO, and , A main deflector DAC / amplifier 112 for supplying a signal for dynamic astigmatism correction, a DAC / amplifier 113 for a dynamic focus lens driving the dynamic focus lens 106, and deflection control for controlling the entire deflection system / signal system. A circuit 120 is provided. In the electron beam inspection / length measuring apparatus 100, the main deflectors 104 and 105 have a function of dynamic astigmatism correction. However, apart from the main deflectors 104 and 105, a dynamic astigmatism corrector and its DAC are provided. -An amplifier may be provided.

  The secondary electron signal detected by the secondary electron detector 109 becomes image information and is input to the signal control circuit 110. The signal control circuit 110 is synchronized with the deflection control circuit 120. That is, the secondary electron signal from the sample surface 108 is synchronized with the control information of the sub deflectors 102 and 103 and the main deflectors 104 and 105 to acquire image information, and the acquired image information is transmitted from the signal control circuit 110. It is input to the overall control device 130. The electron beam inspection device 100 is controlled by the overall control device 130. The input 142 to the overall controller 130 includes a pixel size, a beam diameter of the electron beam EB, a dimension d2 of the panoramic field area PFOV, a dimension d1 of the child field area CFOV, a scanning speed, and a stage (not shown) for holding the sample W. Basic input information such as a movement method (step & repeat or continuous movement). The input 143 is inspection information of the semiconductor chip IB0 such as the care area (physical care area), GDS data of the pattern, and design information. The output 141 is information related to the inspection result. Information input from the overall control circuit 130 to the deflection control circuit 120 is beam scanning parameters 132, stage position information 133, and wafer Z map information 134 described later. The wafer Z map information is the Z map information of the entire wafer W or the Z map information of the physical care areas IB1, IB2, IB3.

  Note that since the inspection information and the design information of the semiconductor chip IB0 are input to the overall control device 130, the overall control device 130 determines a physical care area from the input information, and converts the determined physical care area into a panoramic visual field region PFOV. Can also be divided.

  The sub deflectors 102 and 103 perform relatively small (eg, about 3 μm) deflection, while the main deflectors 104 and 105 perform relatively large (eg, about 200 μm) deflection. The voltage generated by the sub deflector DAC amplifier 111 is about one digit lower than the voltage generated by the main deflector DAC amplifier 112, and the speed of the DAC amplifier can be increased as the output voltage is lower. High-speed operation is possible. The inventor of the present invention obtains the panoramic image acquisition time by scanning one deflector (single deflection) as in the conventional example, and scans the main deflector and the sub deflector by the panoramic image acquisition method of the present invention. In order to compare the time to obtain a panoramic image by double deflection), experiments were performed using various LSI patterns. According to the experiment, it was confirmed that the conventional example takes about three times as long as the present invention. That is, it has been found that when acquiring a panoramic image with the same accuracy (resolution), the method of the double deflection of the present invention can acquire the panoramic image at approximately three times the speed of the method of the single deflection.

  Next, an embodiment of an electron beam inspection / length measurement device 200 having a more detailed configuration than the above-described electron beam inspection / length measurement device 100 will be described with reference to FIG. In the electron beam inspection / length measurement device 200, the same reference numerals are given to the elements corresponding to the electron beam inspection / length measurement device 100, and the detailed description is omitted.

  The electron beam inspection / length measuring device 200 includes an electron gun 201 that emits an electron beam EB. As the electron gun 201, a Schottky type having a known structure can be used. The electron beam EB emitted from the electron gun 201 is shaped by an aperture 202, and a crossover 204 is formed by a condenser lens 203. A blanking electrode 205 is arranged so that the crossover 204 becomes the center of deflection, and a blanking aperture plate 206 is arranged below the blanking electrode 205. When turning on the electron beam on the sample surface 424 of the wafer W, a signal (voltage) is not applied to the blanking electrode 205, and the electron beam passes through the hole of the blanking aperture plate 206. On the other hand, when turning off the electron beam, a voltage is applied to the blanking electrode 205 to deflect the electron beam so that the electron beam hits a portion other than the hole of the blanking aperture plate 206. A condenser lens 207 for further reducing the crossover is provided below the blanking aperture plate 206, and a convergent half angle variable mechanism 208 is provided below the condenser lens 207.

The convergence half-value angle changing mechanism 208 includes a convergence half-value angle aperture plate 209 provided on the optical path of the electron beam, a deflector 210 disposed above the convergence half-value angle aperture plate 209, and a convergence half-value angle aperture plate 210. Deflectors 211 and 212 are provided, and an electric control circuit 230 that drives the deflectors 210, 211 and 212 is provided. Although not shown in detail, the convergence half-value aperture plate 209 has a plurality (for example, five pieces) of sizes (hole diameters) corresponding to the convergence half-value angles α _opt , α ₀ , α ₁ , α ₂ , and α ₃ of the electron beam. ) Apertures A _opt , A ₀ , A ₁ , A ₂ , A ₃ are opened. It is preferable that the aperture A _opt opened in the central portion of the convergence half angle aperture plate 209 has a size corresponding to the convergence half angle α _opt that gives the minimum electron beam diameter d _min . In order to form an electron beam with the minimum electron beam diameter d _min , it is necessary to minimize the influence of the aberration created by the variable convergence half-value angle mechanism 208, and therefore, it is necessary to use the aperture plate shown in FIG. It is advantageous to place the aperture A _opt in the central part of 209). In the present embodiment, the one-stage deflector 210 and the two-stage deflectors 211 and 212 are arranged before and after the convergent half-value angle aperture plate 209, however, how the deflectors are arranged is appropriately set. can do. Then, an optimal pixel size is selected in accordance with the size of the LSI pattern to be measured and the purpose (inspection, review, or length measurement), and which convergence half-value angle aperture is selected in accordance with the selected pixel size. . The electron beam is deflected by the deflector 210 at the front stage to illuminate the selected half-angle aperture. At this time, it is advantageous to dispose the deflector so that the crossover 213 is formed at the deflection center of the deflector 210 because the positions of the crossovers 214 and 215 do not change even when the aperture is switched. The electron beam that has passed through the convergence half-value angle aperture is turned back by the deflector 211, and the angle is adjusted by the deflector 212 so as to be coincident with or parallel to the optical axis. In this way, an optimal convergence half-value angle aperture can be electrically selected instead of mechanical movement. A projection lens 216 is arranged below (at a later stage) the convergence half-value angle limiting aperture plate 209, and the crossover 213 is imaged at the position of the crossover 214 by the projection lens 216.

  Here, it is known that when the electron beam flowing through the lens barrel (not shown) changes, the focal length changes and the electron beam blurs due to the Coulomb effect acting between the electrons. When the focal length fluctuates depending on the beam current, it can be corrected by controlling the voltage applied to the electrostatic lens type dynamic focus lens 106. That is, when the convergence half-value angle aperture having a different diameter is changed, the fluctuation of the focal length (focus position) can be corrected by changing the lens strength of the dynamic focus lens 106. In addition, the dynamic focus lens 106 is used to correct not only the curvature of field when the deflection is performed by the main deflectors 104 and 105 but also the correction of the change of the focus position due to the Coulomb effect and the unevenness of the sample surface. It can also be used for correcting a change in the focus position. The objective lens 107 forms an image of the crossover 214 on the sample surface 108. A retarding voltage may be applied to the sample surface 108 for the purpose of improving the resolution and reducing the charge on the sample surface 108, so that the acceleration voltage of the electron beam on the sample surface 108 is reduced to a level of 1 kV or less. Is done. Further, a boosting voltage is applied to the objective lens 107 in addition to the retarding voltage to extract secondary electrons generated on the sample surface 108. The extracted secondary electrons are accelerated by the retarding voltage and the boosting potential, enter the lens barrel, advance in the lens barrel, and are detected by the detector 109. The sample wafer W is fixed by the electrostatic chuck 217. The wafer W is moved on an XY plane perpendicular to the optical axis by an XY stage 218 movable in the X and Y directions. Above the XY stage 218, a Z stage 219 using a piezo element and movable in the Z direction is provided, so that the wafer W can be moved in the Z direction. Further, Z sensors 220 (light projecting unit) and 221 (light receiving unit) for measuring the position of the sample surface 108 in the Z direction are provided. A laser interferometer mirror 222 is provided on the Z stage 219, and the laser interferometer 223 can precisely measure the positions of the stages 218, 219 in the X and Y directions. A calibration mark 301 having the same height position as the wafer W is provided on the Z stage 219. The calibration mark 301 has a retarding voltage as in the case of the wafer W. Although a reference mirror is not shown in FIG. 4, a known mirror may be provided at an appropriate place such as a lower part of a lens barrel.

  The control system of the electron beam inspection / length measuring device 200 processes image information output from the electric control circuit 230 for driving the deflectors 210, 211, 212 of the high-speed convergent half-value angle varying mechanism 208 and the detector 109. A signal control circuit 110, an electronic control circuit 121 for controlling a sub deflector DAC / amplifier 111 for driving the sub deflectors 102 and 103, and a main deflector DAC / amplifier 112 for driving the main deflectors 104 and 105. Control circuit 122 for controlling the DAC / amplifier 113 for the dynamic focus lens for driving the dynamic focus lens 106, an electronic control circuit 231 for controlling the Z sensors 220 and 221, a booth Electronic control circuit 232 for applying a starting voltage or a retarding voltage, A control circuit 233 for controlling the laser 219, a control circuit 234 for controlling the XY stage 218, a control circuit 235 for processing signals of the laser interferometer 223, an overall control device 130 for integrally controlling these, and a storage means 140. Prepare. The overall control device 130 creates a Z map from the measured values of the Z sensors 220 and 221 and the signal (X coordinate, Y coordinate) of the laser interferometer 223, and stores the created Z map in the storage unit 140. The storage unit 140 stores, in addition to the Z map, a deflection aberration correction map, image information, and various data input to the overall control device 130, which will be described later. Note that the electronic control circuits 121, 122, and 123 may be configured as the deflection control circuit 120, similarly to the electron beam inspection / length measuring device 100.

  Next, an image acquisition method using the electron beam inspection / length measurement device 200 will be described with reference to an example in which a panoramic image of the panoramic visual field region PFOV1 shown in FIG. 2 is acquired.

  First, a wafer W as a sample is transferred by a transfer robot (not shown) and chucked by an electrostatic chuck 217. Even with such chucking, it is impossible to keep the sample surface 108 of the wafer W flat, and non-flatness remains. Therefore, before the operation such as inspection and length measurement, that is, before acquiring a panoramic image, the stage 218 is moved, the Z value is measured by the Z sensors 220 and 221, and the measured value and the laser interferometer 223 are measured. A Z map defining the Z value of the sample surface 108 of the wafer W is created from the X coordinate and the Y coordinate of the signal stage 218, and the created Z map is stored in the storage unit 140. The stored Z map is used when correcting deflection aberration described later. Here, if the dimension d2 of the predetermined area (panoramic FOV) is set to about 200 μm, the non-flatness in the Z direction in the predetermined area can be made ± 100 nm or less, and thus it can be considered as flat. For this reason, the Z map creates a mesh with the size d2 of the predetermined area, measures the Z value for each mesh (that is, for each predetermined area), and creates a Z map from the measured values.

  The overall control device 130 reads the physical care area IB1 corresponding to the wafer W input from the computer, or the physical care area IB1 obtained from the inspection information or design information input from the computer, and reads a plurality of panoramic visual field areas PFOV1,. .. Divide into PFOVn. Next, the plurality of panoramic viewing areas are divided into a plurality of child viewing areas CFOV101,..., CFOV6464,. At this time, the panoramic visual field area PFOV1 is divided into child visual field areas CFOV101 to CFOV6464.

  An electron beam EB is generated from the electron gun 201, and the sub-deflectors 102 and 103 scan the electron field EB in the child field region CFOV 101 on the sample W (first step). The secondary electron generated from the sample surface 108 of the sample W is detected by the detection unit 109, and the image of the CFOV 101 is obtained by synchronizing the detected secondary electron signal with the deflection information of the sub deflectors 102 and 103. . Here, the scanning start point of the electron beam is set to the upper left corner of the child viewing area CFOV101, but may be another corner such as the upper right corner or the center of the child viewing area CFOV101. . Further, the secondary electron signal from the sample W and the deflection information are synchronized, but an image may be obtained by synchronizing at least one of the secondary electron signal and the reflected electron signal from the sample W with the deflection information. .

  After scanning the inside of the child field area CFOV 101, the electron beam EB moves to the next child field area CFOV 102 by the main deflectors 104 and 105 (second step).

  Here, since the deflection widths of the main deflectors 104 and 105 are relatively large, the deflection aberration becomes a problem. In the present invention, the deflection aberration is a total aberration caused by the deflection of the objective lens 107 and the main deflectors 104 and 105, and is a general term for deflection distortion, field curvature, and deflection astigmatism. Hereinafter, the correction of the deflection aberration will be described.

The deflection aberration DA can be written as follows at the position (X, Y, Z) in the three-dimensional space.
DA = DA (X, Y, Z)
As described above, the dimension d2 of the panoramic image of the panoramic visual field region PFOV acquired by the present invention is at most about 200 μm, and the height variation is at most about 100 nm. For this reason, since the panoramic viewing area PFOV can be regarded as being in the Z = Zn plane, the deflection aberration DAP in the panoramic viewing area can be written as the following equation.
Deflection aberration in panoramic viewing area = DAP (X, Y, Zn)

If the deflection aberration of the electron beam is corrected for each pixel, it must be performed in a very short time, and the data transfer speed cannot keep up, which is not practical. Therefore, in the present invention, the deflection aberration is corrected for each child field region CFOV. For that purpose, it is required that the aberrations in the child visual field region CFOV are approximately equal everywhere. Assuming that the position of the center of the M-th child field area CFOVm is (Xmc, Ymc), the deflection aberration DACm of CFOVm can be written as the following equation. The position of the corner of CFOVm may be used instead of the position of the center of CFOVm.
Deflection aberration in child field area CFOVm = DACm (Xmc, Ymc, Zn)
If the above-mentioned deflection aberration is corrected so as to cancel each child field region CFOV, a highly accurate CFOV image can be obtained over the entire panoramic field region, and the resolution of the panoramic image can be improved. In order to execute this, as described above, it is necessary to set the dimension d1 of the CFOV small so that the deflection aberration within the CFOV falls within an allowable range. According to the simulation calculation, the maximum dimension of the CFOV was approximately 8 μm for a device having a design rule of 20 nm. Therefore, for advanced devices having a design rule of 20 nm or less, the dimension d1 of the CFOV may be set to 8 nm or less (for example, 3 nm).

  Here, it is not realistic to correct the deflection aberration in real time at the time of inspection. Therefore, according to the present invention, a deflection aberration correction map that defines the deflection aberration correction amount of each CFOV in a PFOV corresponding to discrete Z values (Z1, Z2,. By storing them in 140, deflection distortion, curvature of field, and deflection astigmatism can be immediately given to each CFOV. As described above, by creating a Z map of each PFOV in advance, deflection distortion, curvature of field, and deflection astigmatism can be immediately given to each CFOV of the PFOV. This is realistic because the deflection aberration can be corrected simultaneously with the deflection by 104 and 105 (during the movement of the electron beam to the next CFOV or before and after the movement).

The deflection aberration correction map can be created by using the calibration mark 301 created on the stage. As shown in FIG. 5, as the calibration mark 301, a mark in which cross patterns (cross marks) 302 are arranged in a PFOV in which an electron beam can be deflected by the main deflectors 104 and 105 can be used. . One of the cross patterns 302 is selected, and an image of the selected cross pattern 302 is obtained. From the acquired image of the cross pattern 302, the image center coordinate position (RXimage, RYimage) of the cross pattern can be determined. Then, the stage 218 is moved to position the cross pattern 302 at the center of the PFOV. Since the position (X coordinate, Y coordinate) of the stage 218 is measured using the laser interferometer 223, the true movement distance of the stage 218, that is, the true movement distance (RXL, RYL) of the cross pattern 302 is determined. I understand. The distortion (ΔX, ΔY) of the measurement point is represented by the following equation.
ΔX = RXimage−RXL
ΔY = RYimage-RYL
If the distortion of the cross pattern 302 arranged over the entire PFOV is measured, a deflection distortion correction map can be created from the measured values. The distortion at the center position of each CFOV can be corrected using this deflection distortion correction map. In the case of field curvature, as described above, the half-width of the rising edge of the secondary electron signal obtained from the edge of the cross pattern 302 arranged over the entire PFOV and the half-width of the signal at the time of Just Focus are described. Is detected, a focus shift ΔF is detected. If the defocus is measured over the entire PFOV, a field curvature correction map can be created. Similarly, for the deflection astigmatism, if an image of the cross pattern 302 arranged over the entire PFOV is acquired and the astigmatism is measured, a deflection astigmatism correction map can be obtained.

  If gold fine particles are arranged in the cross pattern 302 of the calibration mark 301, the diameter and shape of the electron beam EB can be measured using the gold fine particles. Even if gold fine particles are used instead of the cross pattern 302, a field curvature correction map and a deflection astigmatism correction map can be created. It is preferable to use gold fine particles for creating the deflection astigmatism correction map.

  As described above, the case where the deflection aberration correction map (the deflection distortion correction map, the field curvature correction map, and the deflection astigmatism correction map) is created in the case of the predetermined Z value has been described as an example. Since the deflection distortion, the field curvature, and the deflection astigmatism change with Z, the deflection distortion correction map, the field curvature correction map, and the deflection astigmatism correction map corresponding to discrete Z1, Z2,. Each is created and stored in the storage means 140 such as a memory. At this time, the operation of changing the Z value of the calibration mark 301 by moving the Z stage 219 in the Z direction and creating a deflection aberration correction map corresponding to the changed Z value is repeated. Note that a plurality of calibration patterns (calibration marks 301) having different Z values may be formed on the Z stage 219, and a deflection aberration correction map may be created using each calibration pattern. Creation of these deflection aberration correction maps may be performed at the time of starting up or calibrating the electron beam inspection and length measuring apparatus 200, and need not be performed frequently.

  By creating the Z map and the deflection aberration correction map in this manner, the Z value of the PFOV can be determined from the Z map, and the CFOVs in the PFOV are referred to by referring to the deflection aberration correction map corresponding to the determined Z value. By obtaining the deflection distortion, curvature of field, and deflection astigmatism, the deflection aberration can be corrected for each CFOV. Specifically, the controller 130 obtains a correction amount to be applied to correction devices such as the main deflectors 104 and 105, the dynamic focus lens 106, and the dynamic astigmatism corrector, and corrects a correction signal corresponding to the correction amount. Sent to the device. According to this, the deflection aberration can be corrected during or shortly before and after moving the electron beam to the next CFOV 102. After the correction of the deflection aberration, the electron beam in the CFOV 102 is scanned by the sub-deflectors 102 and 103 in the same manner as described above, and the secondary electron signal detected at this time is compared with the sub-deflectors 102 and 103 and the main deflection. By synchronizing the deflection information of the devices 104 and 105, an image of the CFOV 102 is obtained. Hereinafter, similarly, the first step and the second step are repeated to acquire images of CFOV103 to CFOV6464. Then, a panoramic image is acquired by connecting the acquired images of the plurality of child visual field regions CFOV101 to CFOV6464. Thereafter, the overall control device 130 performs inspection and length measurement using the acquired panoramic image. Inspection methods and length measurement methods using the acquired images are known, and thus description thereof is omitted here. Note that the overall control device 130 corresponds to “image acquisition unit”, “inspection unit”, “partition unit”, “correction map creation unit”, and “deflection aberration correction unit” in the claims.

  By the way, in the electron beam inspection and length measuring apparatus 200, an image having a good SN ratio may not be obtained by one image acquisition. The present inventors have conducted intensive research and confirmed that it is effective to shift the CFOV image to obtain a panoramic image in the case of inspection and length measurement. In this case, the images acquired a plurality of times are integrated and used for the inspection measurement. In the case of integrating N times (for example, 5 times), an image is obtained by shifting the CFOV in the X direction and the Y direction by 1 / N (for example, N = 5) of the dimension d1 of the CFOV, and averaging seams of the CFOV. By doing so, the joint can be made invisible. As such an image acquisition method, for example, as shown in FIG. 6, images CFOV1-1, CFOV1-2,..., CFOV1-5 of CFOV1 are acquired five times while shifting by d1 / N, and The images of the batches are stored in the storage means (memory) 140 and integrated, whereby an image of CFOV1 can be obtained. In the CFOV image after the integration, the remaining deflection aberrations are averaged, so that the accuracy is improved. As a result, the seam of the CFOV is also smoothed by the averaging effect, and the seam is not erroneously recognized as a defect. Here, the image of the CFOV is acquired five times, but the number of times is not limited to this, and can be appropriately set in consideration of the averaging effect at the joint. As shown in FIG. 7, after acquiring the PFOV 1-1, the panoramic images PFOV1-2 and 1-3 are shifted in the X direction and the Y direction by 1 / N (for example, N = 5) of the dimension d1 of the CFOV. , 1-4, 1-5 may be acquired and integrated. Also in this case, the number of times of integrating the PFOV is not limited to five, and can be appropriately set in consideration of the averaging effect at the joint.

  In the above example, the pixels are moved over the entire CFOV. However, in the CFOV, the electron beam is randomly deflected (moved) from the pixel 401 to the pixel 402 as shown in FIG. Is also good. The pixel size d3 can be set to, for example, 3 nm. In the CFOV, as shown in FIG. 8B, the deflection can be freely controlled, such as multiplexing the pixels 403 and scanning the lines 405 and 406 so that the lines 405 and 406 partially overlap. In addition, it is not necessary to acquire a CFOV image over the entire PFOV, and as shown in FIG. 8C, by skipping CFOVs that do not need to be inspected (for example, by deflecting from the CFOV 407 to the CFOV 408), it becomes necessary. An image can be acquired by selecting only a portion.

Hereinafter, an application example of the electron beam inspection and length measuring device M will be described. In this example, the pixel size is 12 nm, the number of pixels is (1/4) k bits, and the size d1 of the CFOV is 3 μm. The panoramic image is 192 μm × 192 μm. Such a panoramic image is obtained using the above-described image obtaining method, and is compared with a model-based simulation pattern (reference pattern). That is, the analysis is performed using the D: DB function. When the actual image (the pattern thereof) deviates from the reference pattern by more than the allowable value, it is determined that there is a high possibility that the pattern is a defect (for example, a potential defect). In the example shown in FIG. _{9, CFOVa 1 b 1 c 1} d 1 and CFOVa _n b _n c _n d _n is the CFOV it is determined that there is a high possibility that there is a defect. A re-inspection (Review) is performed on the CFOV determined in this manner with an increased inspection accuracy (with a small pixel size). In this reinspection, the pixel size can be set to 3 nm and the number of pixels to 1 k. At this time, in order to obtain a beam size corresponding to a pixel size of 3 nm, the deflectors 210, 211, and 212 are controlled to select an optimum half angle of convergence α. Since the crossover 213 is formed at the center of deflection of the upper deflector 210 for selecting the convergence half value angle, the position of the crossover 213 does not move even if the convergence half value angle setting aperture is electrically changed. That is, when a rough inspection is performed based on a panoramic image, and a condition where a defect is suspected is observed with a high-precision CFOV by changing conditions, the position of the beam remains unchanged even if the conditions are changed. This is an essential requirement for re-inspection of images with significant potential defects.

  CFOV: child field area, child field image, EB: electron beam, EO: electron optical system, IB1 to IB3: physical care area (care area), PFOV: panoramic field area (predetermined area), panoramic image, W: wafer ( Sample), 100, 200: electron beam inspection / length measuring device, 102, 103: sub deflector, 104, 105: main deflector, 108: sample surface, 110: signal control circuit (image acquisition unit), 120: deflection Control circuit (deflection control means), 130 ... Overall control device (image acquisition unit, inspection unit, partitioning means, correction map creation means, deflection aberration correction means), 201 ... Electron gun (source of electron beam), 219 ... Z Stage (stage), 220, 221... Z sensor (Z value acquiring means), 301... Calibration mark (calibration pattern).

Claims (6)

An image acquisition method for acquiring an image of a predetermined region on a sample with an inspection / length measuring device using an electron beam as a panoramic image by connecting a plurality of child field-of-view images,
A predetermined area on the sample is divided into a plurality of areas, and the divided area is used as a child field of view to obtain a child field of view image, and a panorama is obtained by connecting the obtained plural child field of view images. look contains a panoramic image acquisition step of acquiring an image, the child field image obtaining step, first scans the electron beam in a child viewing area by the sub-deflector provided in the electronic optical system including a source of an electron beam 1) and a second step of moving the electron beam to the next child field region by the main deflector having a larger deflection width than the sub deflector further provided in the electron optical system. Repeat second step, obtaining the child field image based on the at least one control information of the control information and the main deflector of the sub-deflector of the secondary electron signal and the backscattered electron signal generated from the sample , Image information is acquired a plurality of times, in which obtain one image information by integrating,
An image acquisition method, wherein an image is acquired such that a part of an adjacent child field of view is shifted in the X and Y directions at the time of the first and a plurality of image acquisitions .   2. The image acquisition method according to claim 1, wherein in the child field image acquiring step, the child field area is partitioned such that a variation amount of deflection aberration falls within a predetermined allowable range. Before SL before movement of the electron beam to the next child field area is completed, the image acquisition method according to claim 1 or 2, wherein the correcting the deflection aberration of the next child viewing area. A Z-map creation step for creating a Z-map defining the Z-value in the sample plane by setting the direction orthogonal to the upper surface of the stage as the Z-direction and the height position in the Z-direction as the Z-value. Using the formed calibration pattern or a plurality of calibration patterns formed on the stage and having different Z values, the deflection aberrations are respectively measured within the predetermined area in accordance with the discrete Z values, and these measurements are performed. A correction map creation step of creating a correction deflection aberration map defining a correction amount of the deflection aberration from the values for each Z value.
Acquiring a Z value of the predetermined area from a Z map, and correcting a deflection aberration for each child field area in the predetermined area using a corrected deflection aberration map corresponding to the obtained Z value. The image acquisition method according to claim 3 . An electron optical system including a source of electron beam,
A stage that holds the sample to be irradiated with the electron beam and is movable in a direction orthogonal to the optical axis of the electron optical system;
An image acquisition unit that acquires an image based on at least one of a secondary electron signal and a reflected electron signal generated from the sample,
Met electron beam inspection-measuring device and a test unit for performing inspection and measurement on the basis of the image acquired by the image acquisition unit,
The electron optical system divides a predetermined region on the sample in a plurality, and a sub deflector for scanning the electron beam in each sub-field regions of the sectioned areas as child viewing area, the electron beam between the child viewing area A main deflector having a larger deflection width than the sub deflector,
Deflection control means for controlling the sub deflector and the main deflector, further comprising:
Wherein the image acquiring unit is a Rukoto synchronized with at least one of the secondary electron signal and the backscattered electron signal from the sample, and the control information for the sub-deflector and the main deflector by the deflection control means, a plurality of images Times and integrates to get one image,
An electron beam inspection and length measurement device , wherein an image is acquired such that a part of an adjacent child field region is shifted in the X and Y directions at the time of the first and a plurality of image acquisitions . Care area acquisition means for acquiring a region to be inspected and measured for a sample as a care area,
Division means for dividing the care area into a plurality of predetermined areas, and dividing the plurality of predetermined areas into a plurality of child visual field areas,
Z value acquisition means for acquiring a Z value in a predetermined area by setting a direction orthogonal to the upper surface of the stage as a Z direction and a height in the Z direction as a Z value;
Correction map creating means for creating a deflection aberration correction map that defines the amount of deflection aberration correction for each discrete Z value;
Before the movement of the electron beam to the next child field area by the main deflector is completed , the deflection vector correction map corresponding to the Z value acquired by the Z value acquisition means is referred to, and the next child field is read. 6. The electron beam inspection and length measuring apparatus according to claim 5 , further comprising a deflection aberration correcting unit that corrects a deflection aberration of the area.

Images