JP4253023B2 - Charged particle beam apparatus and scanning electron microscope control apparatus - Google Patents

Description

  The present invention relates to a charged particle beam apparatus equipped with a sample stage driving device on which a sample to be measured (including observation) is mounted, and particularly in a scanning electron microscope, an optical microscope, etc. The present invention relates to a charged particle beam apparatus such as a scanning electron microscope suitable for automatically observing a sample that is known in advance.

  When observing a sample whose shape or internal structure / arrangement is known in advance with a scanning electron microscope, using this drawing, photo, and coordinates for indicating the position of the observation part in the sample, this and the movement of the sample stage There has been known a method for automatically and quickly moving a portion to be observed on a sample within a field of view by associating them with a certain method. There is a technique described in Japanese Patent Publication No. 63-94548 or the like as a technique considered to be effective in these methods. This is intended to reduce the amount of observation work by manually specifying the position of the part to be observed, but in the current scanning electron microscope, this method is further advanced and automation is performed by a computer. Currently, in this method, when a plurality of chips having the same design pattern on the same support called a wafer and having a somewhat accurate geometrical arrangement relationship are regarded as samples, The purpose is to observe the minute patterns that are desired to be observed and cannot be identified and confirmed on the chip design drawing.

  Once the wafer is actually mounted on the sample stage, the position temporarily fixed to the sample stage of the wafer at this time is set as the reference position of the observation position indicating coordinates on the wafer. On the scanning electron microscope control device side, the chip size, placement, rotation direction within the wafer surface, etc. are square on the wafer based on the positional relationship based on the chip design data according to the coordinates at this time. The plurality of rows are virtually registered as a grid. Based on this lattice, the coordinate value of the observation pattern portion in the chip thereafter is determined.

  Here, the sample is moved manually while observing with a scanning electron microscope while actually mounted on the sample stage, and the position of the portion to be observed is manually designated on the actual sample in advance. The scanning electron microscope controller converts the sample stage coordinate value at that time and the position of the part to be observed from the chip array into the in-chip coordinate value, and associates it with the observed partial scanning electron microscope image. Remember. Thus, in the subsequent observation, the scanning electron microscope control device side obtains the coordinate value of the position where the sample stage should be automatically moved from the in-chip coordinate value of the observation portion position and the chip arrangement position registered in advance, The observation part is automatically specified by image collation with the observation partial image stored in association with this part.

  In the prior known example, the position on the drawing based on the design data and the position on the actual sample of two portions showing an arbitrary characteristic sample structure on the actual sample before observation. As a result of the association, coordinate calibration between the sample stage coordinates and the coordinates in the sample drawing, that is, the coordinates used by the coordinate indicating means is performed. Thereafter, when the position of the observation unit is designated by the coordinate designating means, the observation position can be accurately entered into the field of view. In particular, when mounted on the same sample support that can indicate the observation location on the same design drawing (for example, IC patterns arranged for pattern comparison on the sample stage), a switching function by a sample selection switch is provided. By giving, it is possible to indicate the observation position of a plurality of samples by the same drawing.

  In a scanning electron microscope used in a recent semiconductor manufacturing process, a sample (IC pattern) arranged on the same support (sample stage) in the above-described known example is arranged on a wafer. It is replaced with a combination for the IC pattern. Here, the two feature points are not limited to those on the same sample (same chip), but can be specified over a plurality of samples (multiple chips). As a result, coordinates on the coordinate indicating means that can accommodate all the arranged samples are set, and coordinate calibration is performed between the coordinates and the stage coordinates. In such a case, since multiple samples (multiple chips) are geometrically arranged to some extent on the same support (wafer), the observation position (x, y) is the chip arrangement pitch (px, py). From the arrangement (nx, ny) and the in-chip position (xd, yd), it can be obtained by the following [Equation 1].

[Equation 1]
x = px × nx + xd
y = py × ny + yd

In this method, two places having a characteristic structure are registered as preparation before observation. Based on the mounting position of the wafer on the sample stage at this time, two-dimensional coordinates defined by two characteristic points on the mounting wafer are created as the indicating coordinates on the coordinate indicating means. On this coordinate, a drawing used by the coordinate designating means is created from an array of grids representing the chip size and arrangement defined according to the design data. In the subsequent observation, when another wafer on which the same sample (chip) is placed is mounted on the sample stage for observation, the position of the observation portion is instructed to the sample stage based on the indicated coordinates. However, the mounting position of the wafer on the sample stage is determined by the mechanical contact between the stage and the wafer, which causes a slight deviation from the previous two feature point registration, and this amount is different each time. Become. In order to correct this, coordinate calibration is performed by comparing the same feature point at the same position as the previously registered two points with the position at the time of initial registration. As a result, the coordinates at the time of registration and the stage coordinates at the time of observation, which are the coordinates of the coordinate indicating means, are calibrated. Furthermore, the coordinate and the sample indicating the situation in which the coordinate calibration performed by matching the two points in the sample in the previous known example is geometrically arranged on the same support (wafer) of a plurality of samples (chips) The same applies to the coordinates indicating the position of the observation portion in the (chip).
Japanese Patent Publication No. 63-94548

  However, in actuality, particularly in the process of baking chips (individual samples) on the wafer (support) in the semiconductor manufacturing process, the factor that determines the positional accuracy of the chip arrangement on the wafer is the stepper. The factor that determines the position accuracy of the observation portion in the chip depends on the distortion of the stepper lens. In addition, when the chip arrangement is indicated by only one point in each chip, the position of the observation portion in the chip is also shifted due to the shift in the rotation direction in the plane coordinate plane from the sample stage coordinates of the entire chip. Produces the same effect. In particular, when an attempt is made to observe a portion different from the portion used for coordinate calibration, the deviation increases as the distance from the coordinate calibration position increases due to the rotation of the entire chip. This is because the factors that determine the accuracy of the two are different, so that only the coordinate calibration for the coordinates for a set of coordinate instruction means can perform the field-of-view of the observation position by the coordinate instruction with sufficiently high position accuracy. Indicates that it is not possible. However, such misalignment at the time of visual field tends to tend to be in the direction and amount of misalignment when the visual field by moving the stage is performed multiple times at the same position. That is, the actual visual field stop position is often concentrated around a certain position away from the target position in a certain direction. This indicates that the position accuracy is low but the position reproducibility is good.

  On the other hand, when the sample stage of the observation apparatus is moved, there are many cases where an inherent position shift occurs in each apparatus. For example, FIG. 17 shows an example of the sample stage position accuracy of the apparatus A, and FIG. This is a measurement of the trajectory actually moved by the stage at each intersection on the grid when the stage is moved along a straight line on a two-dimensional plane. When both devices are compared, the positional deviation from the target position at each intersection does not become the same in both direction and amount. In addition, it can be seen that it varies depending on the target position even in the same apparatus.

  Although these depend on the linear motion performance of the linear motion guide etc. which comprise a stage, it is difficult to produce the guide which performs perfect linear motion. Therefore, when the microscope visual field is moved by such a movement of the stage, even if an attempt is made to position the target position on the sample at the center of the visual field, a positional deviation will occur therefrom.

  However, such misalignment at the time of visual field tends to have a direction and amount of misalignment when the visual field by moving the stage is performed at the same position several times. That is, the actual visual field stop position is often concentrated around a certain position away from the target position in a certain direction. This indicates that the position accuracy is low but the position reproducibility is good.

As described above, in the case of being caused by the sample side, one or two of the two factors in the case of being caused by the sample stage of the microscope are generated at the same time. However, it can be seen that, regardless of the cause, the state is “position accuracy is low but position reproducibility is good”.
In the past, in order to correct such positional deviation, several coordinate calibration methods have been proposed between the indicated coordinates of the observation position and the stage coordinates that are actually the reference when moving.

  For example, when the stage is moved based on a reference indicated by a broken line in FIG. 17 or FIG. 18, for example, an instruction to move to the actual stage is obtained based on the reference indicated by a solid line and an instruction value is given. It is. Since the solid line shows how the stage actually moves, the movement of each grid to the vicinity of the intersection point has achieved some results. However, a position away from this intersection, for example, an arbitrary point in the vicinity of the center of gravity of each grid is farther than any calibration point, and actual accurate calibration is not necessarily performed. Here, if this grid is made as small as possible, the number of calibration points increases and the interval between them becomes narrow. By doing so, the distance from any point to the calibration point is reduced, and the above problems are solved to some extent. However, if the number of grid points, that is, calibration points is increased, it is necessary to register the number of calibration points, and there is a limit to increase the number of grid points due to the convenience of the calibration work performed by the operator.

In fact, when the field-of-view with the conventional method is performed with a scanning electron microscope, it has been confirmed that the amount of visual field deviation increases as the position moves away from the two feature points used for coordinate calibration.
Currently, this large amount of deviation can be dealt with by ensuring a wide search area by reducing the image magnification of the scanning electron microscope image during the position search using the image collation based on the observed partial scanning electron microscope image. It was. However, this requires a search time because it is necessary to perform image matching for all of the many objects displayed in a wide area. In addition, as a feature of the scanning electron microscope, when observing an image at a low magnification, it is easily affected by image disturbance due to charge-up caused by irradiating primary electrons, resulting in an unstable scanning electron microscope image. There was a problem that there were many search errors due to misrecognition.

  Since a scanning electron microscope used for observation in recent semiconductor manufacturing processes has a high observation magnification, it is necessary to move the sample stage with high position indication accuracy in order to bring the observation object into the field of view. On the other hand, high equipment availability and process management by automatic observation are also required.

  The present invention has been made in view of the above-described problems of the prior art, and in particular, in a charged particle beam apparatus such as a scanning electron microscope used for observing chip pattern defects in a semiconductor manufacturing process, a quick and accurate observation field of view. The purpose is to allow the observation position to move inward.

  In the present invention, a means for detecting a positional difference between a target position on the chip pattern in the microscope observation field after moving the sample stage and a predetermined position in the field, means for storing the result, and subsequent observation When moving to a position, there is provided means for determining a new movement target position in consideration of the previously stored position difference and the movement target position employed at that time.

  On the other hand, in the present invention, a means for detecting a position difference between the target position on the chip pattern in the microscope observation field and a predetermined position in the field after the sample stage is moved, and the result is adopted at that time. The means for correcting the movement target position that has been corrected to obtain a new movement target position, the means for storing the new movement target position each time, and the previous movement when moving to the position from the next observation. The same effect can be obtained by providing means for determining a new movement target position in consideration of the corrected movement target position.

  When observing another wafer printed with the same pattern as the previous pattern or a pattern on the same wafer as the previous pattern, the previous observation field position shift amount registered in the corresponding observation position is taken into account. The movement target instruction position to the sample stage up to the previous time is changed, and the sample stage is moved according to the instruction position.

  The coordinates on the coordinate indicating means for indicating the position in the sample (chip) (hereinafter referred to as the position indicating coordinates in the sample) and the coordinates on the coordinate indicating means for indicating the chip arrangement (hereinafter referred to as arrangement position indicating coordinates) Provided separately. These sample stage coordinates are the sample (chip) at the time when two characteristic structures used for coordinate calibration using the characteristic structure performed before observation and a plurality of sample structures (chips) are registered. The coordinates are obtained by setting an arbitrary position on the wafer as the origin with reference to the fixed position of the mounted support (wafer) with respect to the sample stage. On the other hand, when the wafer is mounted again on the sample stage for observation after registration at two locations, the mechanical origin alignment accuracy between the wafer and the sample stage is limited. Produce. At this time, the amount of deviation from the desired position of the observation portion after the movement of the sample stage is detected, and this result or the movement target position corrected using the result is sequentially stored and stored. The past shift amount or the statistical processing result of the corrected movement target position is reflected in the movement instruction position to be newly determined, or the movement instruction position is re-determined based on the movement instruction position.

  On the other hand, when the observation is automatically performed, these positional deviation amounts are recorded in association with the observation position in a procedure file in which the observation procedure (including the observation position and the position) is recorded in advance. When the observation is executed according to the procedure file, the positional deviation amount is reflected on the movement instruction position. The method is the same as above. On the other hand, the microscope apparatus in which the positional deviation recorded in these procedure files has occurred is distinguished from other apparatuses. For example, the serial number of the apparatus is recorded at the same time as the misalignment detection amount.

  After creating the procedure file, when these files are used in another device, the serial number is automatically read, and the device or the device in which the misregistration detection result is generated is automatically recognized. The movement instructing position is determined in consideration of only the recording of the movement position target position that is corrected each time using the positional deviation amount that has been confirmed to be that of the own apparatus or the result thereof. As a result, it is possible to prevent the correction using the positional deviation amount detected by another apparatus. By the above means, it is possible to provide a sample stage having high position indication accuracy even in an environment in which observation procedure files are exchanged between different apparatuses.

That is, a charged particle beam apparatus according to the present invention converges a charged particle beam source that generates a charged particle beam, a sample stage that holds and moves the sample, and the charged particle beam emitted from the charged particle beam source onto the sample. A lens, a deflector that deflects a charged particle beam, an image detection unit that detects an image of the sample, an image display unit that displays the detected image, a coordinate instruction unit that indicates a position on the sample, and a coordinate instruction A charged particle beam apparatus comprising: means for moving the sample stage to the position of the sample stage corresponding to the coordinate value designated on the coordinate instruction means and the coordinate value designated on the coordinate instruction means so that the coordinate value on the means and the coordinate value on the sample stage can be calibrated When an arbitrary observation position on the sample is to be observed, the sample stage is moved so that the observation target position matches the movement target position indicated by the coordinate instruction means, and the image detection hand is moved after the movement is completed. A positional deviation amount calculating means for calculating a positional deviation amount between a predetermined position of the sample detected by the image detecting means and a predetermined position of the image detecting means, and a storage means for storing the calculated positional deviation amount, and calculating the positional deviation amount. Based on the amount of displacement obtained by the means, the movement target position coordinate value adopted when moving to the observation position corresponding to the observation position or the same observation position from the next time onward is controlled, and the sample position when the sample stage is stopped is controlled. There is provided a positional deviation correcting means which operates so that the predetermined position and the predetermined position of the image display means coincide with each other.
In the charged particle beam apparatus of the present invention, the stopping accuracy of the sample stage is improved by determining the sample stage moving target position in consideration of the amount of visual field shift that occurs when moving to that position before the observation time. Is possible.

  The charged particle beam apparatus according to the present invention also converges a charged particle beam source that generates a charged particle beam, a sample stage that holds and moves the sample, and the charged particle beam emitted from the charged particle beam source to the sample. A lens, a deflector that deflects a charged particle beam, an image detection unit that detects an image of the sample, an image display unit that displays the detected image, a coordinate instruction unit that indicates a position on the sample, and a coordinate instruction A charged particle beam apparatus comprising: means for moving the sample stage to the position of the sample stage corresponding to the coordinate value designated on the coordinate instruction means and the coordinate value designated on the coordinate instruction means so that the coordinate value on the means and the coordinate value on the sample stage can be calibrated When observing an arbitrary observation position on the sample, the sample stage is moved so that the observation target position and the movement target position designated by the coordinate instruction means coincide with each other. A positional deviation amount calculating means for calculating a positional deviation amount between the predetermined position of the sample detected and the predetermined position of the image display means, and the movement target employed at that time using the calculated positional deviation amount. A corrected moving target position coordinate value stored in the storage means having means for correcting the position coordinate value to obtain a corrected moving target position coordinate value and a storing means for storing the calculated corrected moving target position coordinate value Based on the value, the movement target position coordinate value to be adopted when moving to the observation position corresponding to the observation position or the same observation position from the next time onward is controlled, and the predetermined position of the sample when the sample stage is stopped and the image display means It is characterized by having a misalignment correcting means that operates so as to coincide with a predetermined position.

  The charged particle beam apparatus according to the present invention further includes an observation procedure storage device that stores an observation planned position, an observation partial image, and an observation order thereof, and is associated with the observation planned position coordinate value in the observation procedure storage device. Alternatively, the corrected movement target position coordinate value may be stored.

  By recording and registering the record of the movement target position corrected in accordance with the visual field deviation amount or the deviation amount in this way at the same time as the observation procedure recording, when observing the pattern on the same wafer repeatedly, and on the same kind of wafer. It is possible to provide a simple and appropriate method when the pattern is observed repeatedly.

  In addition, when controlling the movement target position coordinates when moving to an arbitrary observation position, use the statistical processing result of the positional deviation amount or the corrected movement target position coordinate value obtained before that multiple times. Can do. The statistical process can be an average process. The statistical processing can also be a weighted average processing in which weighting is heavily applied to the positional deviation detection result obtained at a time close to the present time. In this way, it is possible to stabilize the stopping accuracy of the sample stage with high accuracy by using the averaging method or the weighted averaging method that heavily weights the closer to the present time in the statistical processing method applied to the past visual field deviation amount. It becomes.

There may be provided means for designating in advance the number of misregistration amount calculation points that is effective retroactively by the misregistration amount calculation means. Furthermore, it has means for storing in advance the positional deviation amount calculation points that are effective retroactively by the positional deviation amount calculation means in association with the observation procedure storage means, and only for the calculation points set at the time of automatic observation It is also possible to validate the positional deviation amount calculation result.
In this way, by adopting a predetermined number of movement target positions that are corrected based on the visual field shift amount or the shift amount, the required storage capacity can be appropriately suppressed.

  Further, the apparatus includes a device identification unit that identifies the charged particle beam device from which the displacement amount or the corrected movement target position coordinate value is obtained, and the displacement amount storage unit or the corrected movement target position coordinate value storage unit is the device identification. For each charged particle beam device identified by the means, a positional deviation amount or a corrected movement target position coordinate value is stored in association with the scheduled observation position, and when the movement target position of the sample stage is determined by the positional deviation correction means, The movement target position of the sample stage can also be determined based on the result of statistical processing of the corrected movement target position coordinate value in which the detected amount of displacement is reflected.

Thus, by providing the device identification symbol for identifying the scanning electron microscope device, it becomes possible to share the observation procedure between different devices.
Means for switching the misalignment correction means between valid and invalid may be provided. Furthermore, it is possible to have a means for storing the validity or invalidity designation of the positional deviation correction means in advance in association with the observation procedure storage means, and to control the validity or invalidity of the positional deviation correction means during automatic observation.

  Means for switching the positional deviation amount calculation means between valid and invalid may be provided. Furthermore, it has means for storing the validity or invalidity designation of the positional deviation amount calculation means in advance in association with the observation procedure storage means, and controls the validity or invalidity of the positional deviation amount calculation means during automatic observation. Also good.

  The charged particle beam apparatus according to the present invention has a sample stage movable in a two-dimensional direction and a coordinate value indicating means, and can calibrate the coordinate values on the coordinate value indicating means and the coordinate values on the sample stage. In the charged particle beam apparatus configured to move the sample stage to the position of the sample stage corresponding to the coordinate value specified on the coordinate value indicating means, the target position and the sample specified by the coordinate value indicating means Target position deviation detecting means for detecting in the microscope visual field the amount of positional deviation between the stage after movement and the position deviation amount storage for storing the position deviation detection result by the target position deviation detecting means in association with the target position. And the target position stored in the positional deviation amount storage means when the target position is designated by the coordinate value indicating means and the sample stage is moved to the target position. Based on the positional deviation detection result statistically processed result; and a positional deviation correcting means for determining a movement target position of the sample stage.

This charged particle beam apparatus can improve the stopping accuracy of the sample stage by determining the sample stage moving target position in consideration of the amount of visual field shift that occurs when moving to that position before the observation time. Become. The target position deviation detection means can be realized by a length measurement function that measures the distance between two points on the charged particle beam device image obtained at the time of observation.
The charged particle beam apparatus according to the present invention is characterized in that the amount of positional deviation between the target position specified by the coordinate value indicating means and the position after the movement of the sample stage gradually decreases as the movement of the sample stage is repeated. .

  According to the present invention, there is provided a sample stage having a high position indicating accuracy required for a charged particle beam apparatus such as a scanning electron microscope having a high observation magnification used for observation in a recent semiconductor manufacturing process. Is possible. On the other hand, it is possible to manage the process by high equipment availability and automatic observation.

An embodiment of the present invention will be described below with reference to the drawings. Here, a scanning electron microscope will be described as an example of the charged particle beam apparatus.
FIG. 1 shows an outline of a scanning electron microscope which is an example of a charged particle beam apparatus according to the present invention. A primary electron beam is generated from the electron source 18, converged by the electron lens 19, and scanned on the surface of the sample 21 by the deflector 20. The secondary electrons generated at this time are detected by the detector 22. This signal reflects, in particular, the shape and components of the sample surface, and is input to the image generator 23. The detector signal is observed as a scanning electron microscope (hereinafter, SEM) image by an image generator 23 that creates a two-dimensional image in synchronization with the operation of a deflection amplifier not shown in the figure. These mechanisms are the same as other scanning electron microscopes.

  The sample stage 1 is driven in the orthogonal XY directions by motors 2 and 3. In the sample stage 1, nx × ny pieces (3 × 3 = 9 in FIG. 1) of a plurality of samples (chips) 4a, 4b... Created by baking on the same support (wafer) in the movable range. It is installed. The sample stage control device 5 can move the sample stage 1 to an arbitrary position by driving the motors 2 and 3 by a required distance on the basis of the mechanical origin S1 (0,0) of the sample stage 1. . In this embodiment, only 3 × 3 chip patterns (samples) 4 baked on the wafer (support) 21 in the semiconductor manufacturing process are arranged.

  Here, the control panel (A) 7 shown in FIG. 2 can be instructed to specify the coordinates by a chip arrangement and in-chip coordinates. When the observation position is already known as the coordinate value, the coordinate value is input to the coordinate value input window 8. When the coordinate value is unknown or the position of the pattern to be observed / measured is not accurately determined, the pointing pen 27 gives an instruction on the SEM image. At that time, the chip at the current stage position automatically becomes the chip arrangement designated. Therefore, it is necessary to move the stage by instructing to the desired chip position on the chip arrangement coordinates 10 before instructing the position. A chip array layout diagram based on the design data is drawn at the chip array designating coordinates 10. A chip layout based on the design data is also drawn on the in-chip position indicating coordinates 11. The in-chip position instruction may be performed only on the chip arrangement instruction coordinates 10 and the actual instruction may be performed within the field of view visible in the SEM image 9.

  On the other hand, the misregistration correction function, which is a feature of the present invention, is particularly effective for a case where the same type of chip baking is performed on a large number of wafers, such as a DRAM having a large production number as an observation sample. That is, when a certain amount of sampling inspection is performed from a large number of wafers, the number of inspections increases. It is considered that the chips manufactured by the same semiconductor process were baked under substantially the same conditions up to the chip position accuracy on the wafer. When it is necessary to observe a large number of wafers baked by the same process, the same portion of the chip corresponding to each other is repeatedly observed on different wafers. Any misalignment occurring at the sample stage position may be corrected from the next time. As a result, the occurrence of misalignment in the first time or particularly in the first test is corrected from the next time by the function of the present invention, so that the stage feed can be performed with high accuracy in the subsequent observation.

  However, in other cases, such as ASIC, when the number of produced sheets is small, the number of inspections is small, and in some cases, the printing conditions may be determined by one inspection. In such a case, the function of the present invention which cannot necessarily increase the position accuracy at the first time is not effective. Therefore, this function must be turned off in some cases.

  When the ON / OFF switching of the function of the present invention causes a high or low position accuracy of the sample stage, a wafer that needs to be automatically searched for an observation pattern by image processing after the movement of the sample stage is completed There is a difference in the size of the area on the surface. That is, after the sample stage is stopped with high position accuracy, the stop position has already reached a position close to the observation pattern, so that the area that needs to be searched is inevitably narrowed. On the other hand, after the sample stage is stopped with low position accuracy, a large distance remains between the stop position and the observation pattern, so that the area that needs to be searched is inevitably widened.

  In the present invention, in order to cope with this, a function for switching ON / OFF of the misalignment correction function is attached. As shown in FIG. 3, the control panel (B) 15 is provided with a switch for turning the position deviation correction function ON or OFF. By operating this, an operator can operate the scanning electron microscope of the present invention with the positional deviation correction function being enabled or disabled. Further, as shown in FIG. 2, the control panel (A) 7 is provided with a misalignment correction function execution state display 65 for transmitting the current execution state of the misalignment correction function to the operator. The operator can perform the operation while checking whether there is an error in selecting the function depending on the type of sample to be observed. On the other hand, in the scanning electron microscope of the present invention, in addition to the manual operation, observation may be automatically performed by a control device such as a computer according to the procedure recorded in the observation procedure storage device 40 (recipe). In the present invention, the observation procedure storage device 40 is provided with a positional deviation correction function ON / OFF registration storage device. As shown in FIG. 6, an ON / OFF state is registered in the misregistration correction function ON / OFF registration storage device. When the observation procedure storage device 40 is used for automatic observation, the execution status ON / OFF of the misregistration correction function is registered. It is possible to switch OFF.

  Further, when executing the positional deviation correction function, the result of the correction effect can be changed depending on the point in time of the past positional deviation amount measurement result used for correction depending on the type of sample being observed. That is, for example, when it is easy to charge up by irradiating an electron beam on the observation surface, the potential distribution on the surface of the sample observed by the scanning electron microscope depends on the distortion of the potential distribution in which the stagnant charges are generated and the state that changes every moment. A phenomenon that changes every moment occurs. This is called drift. In the observation of a sample having a high frequency and a large drift phenomenon, an uncertain fluctuation always occurs in the relative positional relationship between the sample stage and the scanning electron microscope field of view. And move every moment. When the effect of the drift phenomenon on the positional deviation is large, if the past positional deviation amount measurement result is traced back over time in the function of the present invention, the deviation included in the deviation amount becomes large, and the original purpose It is not possible to feed the stage to an accurate position. Therefore, in the present invention, the number of positional deviation amount measurement results going back in the past, that is, since they are sampled at regular intervals, the positional deviation is determined by the operator's judgment depending on the sample to be observed for a time corresponding to the number of measurement points. Adds a function to adjust the number of measurement points. In the present invention, as shown in FIG. 2, the effective position data correction point pcj is displayed on the control panel (A), and at the same time, the correction effective data point input having an input window for inputting pcj by the operator is provided. A window 66 is provided.

In addition, as in the case of switching ON / OFF of the positional deviation correction function, observation may be automatically performed by a control device such as a computer according to the procedure recorded in the observation procedure storage device 40 (recipe). In the present invention, as shown in FIG. 1, the observation procedure storage device 40 includes a misregistration amount correction use point registration storage device 62. In this misregistration amount correction use point registration storage device 62, as shown in FIG. 7, the misregistration correction effective data point pcj can be read when necessary, and is recorded in a writable state when it is desired to store. ing.
In the present invention, the position shift amount is continuously measured for a fixed period in the past, and a new movement target position is determined by correcting the movement target position using the measurement result from a certain point in time. There is also a method for improving the sample stage position accuracy.

  In the present invention, this is enabled by the function of turning on or off the positional deviation amount measuring function. As shown in FIG. 3, the control panel (B) 15 is provided with a switch for turning the position deviation measuring function ON or OFF. The scanning electron microscope of the present invention can be operated by enabling or disabling the positional deviation amount measuring function by operating the operator.

  Further, as shown in FIG. 2, the control panel (A) 7 is provided with a positional deviation amount measuring function execution state display 68 for informing the operator of the execution state of the positional deviation amount measuring function at that time. Yes. The operator can perform the operation while checking whether there is an error in the selection of the function in accordance with the purpose of observation each time.

  On the other hand, in the scanning electron microscope of the present invention, in addition to manual operation, there is a case where observation is automatically performed by a control device such as a computer in accordance with the procedure recorded in the observation procedure storage device 40 (recipe) shown in FIG. is there. In the present invention, the observation procedure storage device 40 is provided with a positional deviation amount measurement function ON / OFF registration storage device 68. The positional deviation amount measurement function ON / OFF registration storage device 68 is registered with an ON / OFF state as shown in FIG. 21, and the positional deviation amount measurement function when the observation procedure storage device 40 is automatically observed is used. It is possible to switch the execution state ON / OFF.

  These may be realized on software operating the SEM control device 6 of the scanning electron microscope main body. Accordingly, in the present invention, not only a control panel is provided as an actual product as shown in FIG. The operator can view the corresponding display on the screens of the control panels (A) 7 and (B) 15 and give necessary instructions there. The in-chip coordinate value (xs, ys) of the observation portion registered in advance in the control panel (A) 7 and the arrangement position (nx, ny) of the sample to be observed are output. Then, the coordinate value converter 12 outputs the distance from the origin position S2 (0, 0) of the designated coordinates Xa-Ya at that position.

  In addition, in order to move the sample stage by manual operation, a trackball 13 or the like is connected to the stage controller 5 as switches. When the trackball 13 is operated, the motor 2 or 3 is driven so that the sample stage 1 moves according to the rotation.

  When a movement command is executed in response to an operator instruction on the trackball 13 or the control panel (A) 7, the stage control device 5 receives the coordinate read value of the linear scale 24 from the linear scale controller 14, and this coordinate value. Motors 2 and 3 are driven until the value coincides with the value corresponding to the movement instruction value (xst, yst).

  The operator presses various switches on the control panel (B) 15 as necessary to execute necessary functions each time. FIG. 3 shows the mounting status of various switches. When the operator presses the placement coordinate calibration point AP1 registration switch, the SEM control device 6 reads the sample stage position at that time with the linear scale 24 and registers and stores the value in the calibration point registration device 16. When the arrangement coordinate calibration point AP2 registration switch is pressed, registration and storage are performed in the same manner as AP1. On the other hand, when the length measuring point registration MP switch is pressed, the sample stage position at that time is read by the linear scale 24 and the value is registered and stored in the length measuring point registration storage device 17 as in the case of the calibration point registration. In the above switches, when the operator presses the switch, the sample stage position at that time is registered and recorded for each purpose.

  When the operator presses the coordinate calibration AP1 when the wafer is loaded again for observation after the registration operation, the position stored and registered in the operation and the position of the sample stage at that time are displayed. Association is performed. Similarly, the coordinate calibration AP2 associates the position stored and registered with the sample stage position at that time. When the association is completed for both AP1 and AP2, coordinate calibration between the designated coordinates Xa-Ya and stage coordinates Xst-Yst for automatically specifying the observation position is performed.

  When the operator presses the misalignment correction function ON / OFF switch, the misalignment correction function can be turned ON or OFF. Further, when the operator presses the position deviation amount measurement function ON / OFF switch, the function of measuring the position deviation amount of the visual field position every time the movement of the sample stage is completed can be turned ON or OFF.

  On the other hand, the observation procedure storage device 40 of FIG. 12 stores various data for executing the automatic length measurement sequence. The calibration point registration storage device 16 and the measurement point registration recording device 17 included therein include, for example, in-chip position coordinate values (xa, ya) related to the calibration point AP and the measurement point MP when the automatic length measurement sequence is executed. ) (xm, ym), chip array position (nxa, nya) (nxm, nym), and SEM image recording used when performing position identification by image collation are registered according to the intention of the operator.

  FIG. 4 shows an example of the calibration point registration storage device. In the calibration point registration storage device 16, AP1 and AP2 as many as the number of calibration points and information necessary for performing automatic observation for each are registered. For example, in AP1, the observation position obtained by moving to a certain chip including the observation point AP1 on the wafer 21 and giving a specific instruction on the SEM image with the pointing pen 27 is converted into an in-chip coordinate value. And register. That is, on the SEM control device 6 side, the position coordinates of the portion to be observed from the stage coordinate values (xst, yst) and the chip arrangement (nx, ny) at that time are in-chip coordinate values of the position coordinates indicated by the pointing pen ( xa, ya). The in-chip coordinate values (xa, ya) and the chip arrangement (nx, ny) obtained in this way are registered and stored as shown in FIG. 5 together with the observation partial image used for image collation. The same applies to AP2.

  FIG. 5 shows an example of the measuring point registration storage device. In the measurement point registration storage device 17A, information necessary for performing automatic observation for each of MP1, MP2,..., MPj and the number of measurement points is registered. For example, in MP1, the observation position obtained by moving to a certain chip including the observation point MP1 on the wafer 21 and giving a specific instruction on the SEM image with the pointing pen 27 is converted into an in-chip coordinate value. And register. That is, on the SEM control device 6 side, the position of the portion desired to be observed from the stage coordinate value (xst, yst) and the chip arrangement (nx, ny) at that point in time is the in-chip coordinate value (xm , ym). The in-chip coordinate values (xm, ym) and the chip arrangement (nx, ny) obtained in this way are registered and stored as shown in FIG. 5 together with the observation partial image used for image collation. The same applies to MP2,.

  By the way, if the sample stage movement to the observation position represented by the in-chip coordinate value (xm, ym) and the chip arrangement (nx, ny) is repeated a predetermined number of times, it remains after the movement of the sample stage to the movement target position is completed. The visual field position shift amount is stored and registered in accordance with the number of observations for each of the measurement points MP1, MP2,. For example, in MP1, (dxm11, dym11), (dxm12, dym12), ..., (dxm1i, dym1i) and the visual field position shift amount for i movements are registered and stored. The same applies to MP2,.

  On the other hand, the observation procedure storage device 40 includes a positional deviation correction function ON / OFF registration storage device 61 and a positional deviation correction use point registration storage device 62 for realizing the positional deviation correction function of the present invention during execution of the automatic length measurement sequence. Yes. They are shown in FIGS. These are executed in the sequence shown in FIG. 8 on the automatic length measurement sequence. That is, it is as follows.

Step SA001: Reading the ON / OFF registration contents stored in the observation procedure storage device Step SA002: Checking the ON / OFF based on the result of Step SA001 Step SA003: Reading the registration information of the position correction function execution status of the control panel (A) Step SA004: ON / OFF check based on the result of Step SA003 Step SSA001: ON / OFF function for misalignment correction
Step SSA002: Misalignment correction ON / OFF function OFF

  In addition, the observation procedure storage device 40 is provided with a positional deviation amount measurement function ON / OFF registration recording device for determining or setting in advance whether or not the positional deviation amount measurement is performed in parallel when the automatic length measurement sequence is executed. . This is shown in FIG. This is executed in the sequence shown in FIG. That is, it is as follows.

Step SD001: Position deviation amount measurement function execution state stored in the observation procedure storage device ON / OFF reading Step SD002: ON / OFF check based on the result of step SD001 Step SD003: Position deviation amount measurement function execution of control panel (A) Reading status registration contents Step SD004: Checking the ON / OFF manual switch based on the result of step SD003 Step SSD001: Turning ON / OFF the position deviation measurement function
Step SSD002: Position deviation amount measurement function ON / OFF OFF

  FIG. 9 shows a mechanism for obtaining the measurement point position by image collation. In addition to the target pattern, various images are displayed on the image memory 9 in which the SEM image is taken. In the image memory 9, an area (hereinafter referred to as an image frame 30) having the same size as the SEM image for specifying the measurement point position registered in advance in the measurement point registration storage device 17, for example, in MP 1 is referred to as the above image memory. 9 Cut out from an arbitrary position according to an instruction from the image frame position indicator 31 from above, and send it to the image collating device 32.

  The image collation device 32 obtains the degree of coincidence by obtaining the image 26-2 registered in the measurement point registration storage device 17 and its correlation value. If this exceeds a certain level value and is judged to be “match”, the ON signal adj at the time of image matching is turned ON and the image position (xg, yg) indicated by the image frame position indicator 31 at this time from the gate 33 is turned ON. ) To the SEM control device 6. However, (xg, yg) is a value at coordinates XG-YG on the image memory 9 including the entire image corresponding to the primary electron beam scanning range.

  On the other hand, the deflection offset amount indicator 34 outputs coordinate values (xdof, ydof) indicating the position of the primary electron beam scanning range on the sample at this time to the deflection indicator 35. In the deflection indicator 35, the offset amount added to the “sawtooth wave” signal having the amplitude value linked to the SEM image magnification generated by the SEM control device 6 is changed so as to be linked to the coordinate value (xdof, ydof). The scanning position of the primary electron beam scanned by the deflector 20 on the sample is arbitrarily changed. Actually, in order to prevent problems such as errors generated during amplification by the amplifier 36, a deflector is prepared separately from the above, and the deflection of the primary electron beam corresponding to the offset amount is performed there. . At this time, a “sawtooth wave” having a magnification interlocking amplitude is directly applied to the deflector.

  According to the series of operations described above, if the “match” of the image is not obtained, the ON signal adj at the time of the image collation match is turned OFF, the gate 33 is closed, and the position signal (xg, yg) is not output. Further, the image frame position indicator 31 at this time is shifted in the x direction or the y direction by one step, and the above-described series of image collation operations are repeated again. If all the areas on the image memory are covered by the movement of the image frame 30 and there is no new movement destination, the image collating device 32 changes the offset amount to the deflection offset amount indicator 34. An instruction signal is output.

  The deflection offset amount indicator 34 changes the offset amount of the deflection signal in the x direction or the y direction only for one step. If the SEM image “coincidence” is still not obtained even when the scanning range is moved over the entire scanning range of the primary electron beam, the SEM controller sets the sample stage position in the x direction or by one step. The indicated position to the stage controller 5 is changed so as to move in the y direction. An SEM image in which a characteristic pattern for specifying the position of the measurement point is copied is searched for by the series of operations described above.

  When the target pattern is extracted on the SEM image, the position (cross mark) of the calibration point on the SEM image stored in the calibration point registration storage device 16 is specified. The distance (ddix, ddiy) from the origin of the coordinates on the image frame 30 (for example, the center of gravity of the image frame) is automatically obtained by the attached scale 38. In this case, (ddix, ddiy) is output to the SEM controller 6 by the scale controller 37 that controls the scale 38. On the other hand, the deflection offset amount indicator 34 outputs a signal (xdof, ydof) indicating the offset amount of the deflection scanning range at this time to the SEM control device 6.

  Through the series of operations so far, all the coordinate values necessary for obtaining the position of the length measurement point to be obtained are input to the SEM control device 6. They are (xg, yg), (xdof, ydof), (ddx, ddy). First, the position (xm, ym) of the image frame represented by the designated coordinates Xa-Ya placed on the wafer is obtained from (xg, yg) and (xdof, ydof) by [Equation 2]. Note that f1 and f2 are functions representing conversion from the right side of the expression to the left side.

[Equation 2]
xm = f1 (xdof, xg)
ym = f2 (ydof, yg)

  Further, the position (xxst, yyst) of the length measurement point also represented by the designated coordinates Xa-Ya is obtained by [Equation 3] from (xm, ym) and (ddix, ddiy) obtained as described above. Note that f3 and f4 are functions representing conversion from the right side of the expression to the left side.

[Equation 3]
xxst = f3 (xm, ddix)
yyst = f4 (ym, ddiy)

  FIG. 11 is a flowchart showing the above operation sequence. This flowchart shows a flow of a series of processes for obtaining a visual field shift amount after the stage movement to the target position is completed.

Step S001: End of stage movement to target position Step S002: Is there a target pattern in the field of view?
Step S003: Position deviation detection function status ON / OFF
Step S004: Detect visual field deviation amount Step S005: Output visual field deviation amount with respect to target position Step S006: Visual field deviation amount storage / registration step S007: Is it possible to newly move the image frame within the visual field?
Step S008: Move the image frame by a certain amount within the field of view Step S009: Is it possible to newly move the scanning range by image shift?
Step S010: Image shift constant amount movement step S011: Stage constant amount movement step S012: Move to the next observation position

Further, the above search operation is basically the same for specifying the coordinate calibration point position.
FIG. 10 shows a mechanism of coordinate value conversion between the stage coordinates Xst-Yst and the designated coordinates Xa-Ya. The inner side of the two-dot chain line shows the structure of the coordinate value converter 12.

  In the SEM control device 6, a signal (xs, ys) (nx, ny) instructing the observation position from the control panel (A) 7 is sent from the measurement point registration storage device 17A to the MP.NO of the measurement point NO, the observation position. (Xs, ys) (nx, ny) indicating the field of view displacement that occurred at the position corresponding to the previous observation (dx, dy), and receiving the SEM image used to identify the position of the measurement point ing. If necessary, these values are output from the SEM control device 6 in the opposite direction to the above when the values are displayed on the display board or when it is necessary to register and store them.

  In response to these data, the SEM control device 6 first outputs (xs, ys) (nx, ny) indicating the observation position to ADD1 via Buff2 and Buff3. In ADD1, provisional designated coordinate values (xaa, yaa) are output based on the pitch (px, py) and offset amounts (xoffset, yoffset), which are chip design information. The visual field shift amount (dx, dy) output via Buff1 is arithmetically operated by ADD2 together with the temporary designated coordinate value (xaa, yaa) to become the designated coordinate value (xa, ya). The indicated coordinate value (xa, ya) is converted into a stage coordinate value (xst, yst) by CONV1 and output to the sample stage control device 5.

  The above-described series of flows shows a flow when the coordinate value of the observation position is instructed from the SEM control device 6 to the stage control device 5. Here, on the contrary, it is necessary to return the stage coordinate value obtained by reading the linear scale 24 to the SEM control device 6 side. First, the stage coordinate value (xxst, yyst), which is the read coordinate value of the linear scale 24, is converted into the indicated coordinate value (xxa, yya) by CONV2. DIV1 receives the pitch (px, py) and offset amount (xoffset, yoffset), which are chip design information, from SEM control device 6 in the same manner as ADD1, and is indicated by (xxa, yya) based on this. The current stage position is converted so as to be represented by a chip array (nnx, nny) and in-chip coordinate values (xxs, yys). The in-chip coordinate value (xxs, yys) is the difference (ddx, ddy) in DIF1 from the coordinate value instructed to the stage control device 5 of the SEM control device 6 at the observation position received by Buff6. Used for arithmetic. This difference (ddx, ddy) represents the distance at which the position of the observation is actually generated although the observation position is indicated. The (ddx, ddy), (xxs, yys), and (nnx, nny) thus obtained are stored and registered in the measurement point registration storage device 17A, the calibration point storage device 16 and the like.

  As described above, the above-described embodiment is used for correcting the sample stage movement target position from a certain point in time by storing the amount of positional deviation obtained in a past fixed period. On the other hand, each time the movement of the sample stage to the observation position is completed, the movement target position that was adopted at the time of movement is corrected each time using the measured displacement amount, and a new corrected movement target position is obtained. There is an embodiment by the method of continuing to memorize. In the embodiment, at the next movement to the same location or the corresponding observation position, a new movement target position is obtained by statistical processing using the past corrected movement target position stored so far. Ask.

  This embodiment is different from the embodiment of the present invention shown in FIG. That is, not the visual field position shift amount (dxm, dym) but the history of the corrected movement target position (xmt, ymt) is stored for each length measuring point. FIG. 23 shows an example of a measuring point registration storage device.

  In the measuring point registration storage device 17, MP1, MP2,..., MPj and information necessary for performing automatic observation for each of the measuring points are registered. For example, in MP1, the observation position obtained by moving to a certain chip including the observation point MP1 on the wafer 21 and giving a specific instruction on the SEM image with the pointing pen 27 is converted into an in-chip coordinate value. And register. That is, on the SEM control device 6 side, the position of the portion desired to be observed from the stage coordinate value (xst, yst) and the chip arrangement (nx, ny) at that point in time is the in-chip coordinate value (xm , ym). The in-chip coordinate values (xm, ym) and the chip array (nx, ny) obtained in this way are registered and stored as shown in FIG. 23 together with the observation partial images used for image collation. The same applies to MP2,. These mechanisms are the same as in the example shown in FIG.

  By the way, if the sample stage movement to the observation position represented by the in-chip coordinate value (xm, ym) and the chip arrangement (nx, ny) is repeated a predetermined number of times, it remains after the movement of the sample stage to the movement target position is completed. The visual field position shift amount is measured every time. At the same time, these values are used as correction values, and the movement target position employed at that time is corrected and redetermined. These corrected movement target position coordinate values (xmt, ymt) are stored and registered in accordance with the number of observations for each of the measurement points MP1, MP2,. For example, in MP1, (xmt11, ymt11), (xmt12, ymt12),..., (Xmt1i, ymt1i) and the visual field position shift amount for i movements are registered and stored. The same applies to MP2,.

  On the other hand, in the scanning electron microscope of the present invention, in addition to manual operation, observation may be automatically performed by a control device such as a computer according to the procedure recorded in the observation procedure storage device 40B (recipe) shown in FIG. is there. In the present invention, the observation procedure storage device 40B is provided with a positional deviation amount measurement function ON / OFF registration storage device 68. The positional deviation amount measurement function ON / OFF registration storage device 68 has ON / OFF states registered as shown in FIG. 21, and executes the positional deviation amount measurement function when performing automatic observation using the observation procedure storage device 40B. It is possible to switch the state ON / OFF. These basic mechanisms are in the case of the above-described method of the present invention, which is “a method for storing a positional deviation amount acquired in a fixed period in the past and using it for correction of a sample stage movement target position from a certain point”. It is the same. The difference is that the measuring point registration storage device provided there is 17B instead of 17A, and the corrected moving target position (xmt, ymt) is not stored, but the visual field position shift amount (dxm, dym) is not stored. The point is to remember.

  The mechanism for obtaining the measurement point position by image collation is the same as that in the embodiment shown in FIG. Again, the movement target position is corrected using the visual field position shift amount obtained by the method of this embodiment. These may be realized on software operating the SEM control device 6 of the scanning electron microscope main body. Accordingly, the present invention includes not only a separate control panel as shown in FIG. 1 but also a control panel realization method in all forms.

  FIG. 25 shows a mechanism of coordinate value conversion between the stage coordinates Xst-Yst and the designated coordinates Xa-Ya. The inner side of the two-dot chain line shows the structure of the coordinate value converter 12. In the SEM control device 6, a signal (xs, ys) (nx, ny) instructing the observation position from the control panel (A) 7 is sent from the measurement point registration storage device 17B to the MP.NO of the measurement point NO, the observation position. (Xs, ys) (nx, ny) indicating the field of view displacement that occurred at the position corresponding to the previous observation (dx, dy), and receiving the SEM image used to identify the position of the measurement point ing. If necessary, these values are output from the SEM control device 6 in the opposite direction to the above when the values are displayed on the display board or when it is necessary to register and store them.

  In response to these data, the SEM control device 6 first outputs (xs, ys) (nx, ny) indicating the observation position to ADD1 via Buff2 and Buff3. In ADD1, provisional designated coordinate values (xaa, yaa) are output based on the pitch (px, py) and offset amounts (xoffset, yoffset), which are chip design information. The corrected movement target position (xmt, ymt) output via Buff1 is converted by CONV3 together with the temporary designated coordinate value (xaa, yaa) to become the designated coordinate value (xa, ya). The indicated coordinate value (xa, ya) is converted into a stage coordinate value (xst, yst) by CONV1 and output to the sample stage control device 5.

  The above-described series of flow shows a flow when the coordinate value of the observation position is instructed from the SEM control device 6 to the stage control device 5. Here, on the contrary, it is necessary to return the stage coordinate value obtained by reading the linear scale 24 to the SEM control device 6 side. First, the stage coordinate value (xxst, yyst), which is the read coordinate value of the linear scale 24, is converted into the indicated coordinate value (xxa, yya) by CONV2. DIV1 receives the pitch (px, py) and offset amount (xoffset, yoffset), which are chip design information, from SEM control device 6 in the same manner as ADD1, and is indicated by (xxa, yya) based on this. The current stage position is converted so as to be represented by a chip array (nnx, nny) and in-chip coordinate values (xxs, yys). The in-chip coordinate value (xxs, yys) is the difference (ddx, ddy) in DIF1 from the coordinate value instructed to the stage control device 5 of the SEM control device 6 at the observation position received by Buff6. Used for arithmetic. This difference (ddx, ddy) represents the distance at which the position of the observation is actually generated although the observation position is indicated. Here, the SEM control device 6 performs correction using the visual field position shift amount (ddx, ddy) with respect to the moving target position employed at that time, and sets a new corrected moving target position (xmt, ymt). Ask. Thus obtained (xmt, ymt), (xxs, yys), (nnx, nny) are subjected to storage registration processing in the measuring point registration storage device 17B, the calibration point storage device 16 and the like.

  Hereinafter, an operation procedure in the present embodiment will be described. FIG. 1 shows how the wafer is actually mounted on the sample stage. At this time, the positional reference is based on the positional relationship between the stage coordinates Xst-Yst and the designated coordinates Xa-Ya that are temporarily fixed. Thereafter, the coordinate value (Xa, Ya) indicating the position to be observed is converted into the coordinate value (Xst, Yst) indicating the moving distance of the stage.

  On the scanning electron microscope control device side, the chip size, arrangement, rotation direction in the wafer surface, etc. are quadrangular based on the positional relationship based on the chip design data on the wafer with reference to the coordinates Xa-Ya at this time. Using the chip outer shape as a frame, the plurality of arrangements are virtually registered as a grid. In FIG. 1, the arrangement is 3 × 3. Two of the chips represented by these lattices include feature points 4a and 4b (indicated by crosses) for position determination suitable for coordinate calibration. The other grid includes 4c (indicated by a cross) indicating the pattern position to be observed.

  The position coordinate value of the observed portion (pattern) in the chip thereafter is expressed by the in-chip coordinates Xt-Yt defined for each of these grids as a reference. In FIG. 1, it is represented by (Xs, Ys). When observing different positions of the chip even in the same observation portion, the coordinate value of the observation position is obtained by considering the offset corresponding to the chip arrangement pitch in the coordinate value within the chip.

  First, feature points for coordinate calibration between the stage coordinates Xst-Yst and the designated coordinates Xa-Ya are registered. By operating the trackball 13, the characteristic location AP1 of the first sample 4a (a part on the chip pattern or an exposure alignment tag or the like) is moved to the center of the field of view of the microscope, and the coordinate calibration point AP1 registration switch To operate. As a result, the stage control device 6 enters a state waiting for input of the designated coordinate value from the control panel (A) 7. Next, on the SEM image 9 of the control panel (A) 7, the location of the feature point used for coordinate calibration is indicated with the ponting pen 12. The control panel (A) 7 outputs the coordinate values of the visual field coordinates Xm-Ym projected in the SEM image 9. The coordinates Xm-Ym at this time have an origin at the center of the visual field, for example, and coincide with the stop position of the sample stage. In the SEM control device 6, the coordinate position for calibration coordinates AP1 registration position (xch1, ych1) is calculated by arithmetic operation of the coordinate value (xm1, ym1) and the read value (xL1, yL1) of the linear scale 24 indicating the position of the sample stage. Ask and remember this. Arithmetic is the sum of both, for example, [Equation 4].

[Equation 4]
xch1 = xm1 + xL1
ych1 = ym1 + yL1

  In addition to the processing related to these coordinate positions, the SEM image at that time is simultaneously recorded, and at this time, AP1 and the image are associated and registered. This state is shown in the calibration point registration storage device 16 in FIG. For example, in AP1, which is calibration point registration NO1, the in-chip coordinate value of the observation point is (xa1, ya1) and the chip arrangement is (nxa1, nya1). The pattern position for image matching is indicated by a cross in the SEM registered image. As a result, in subsequent observations, the sample stage is automatically moved from the in-chip coordinate value (xa, ya) of the feature point position and the pre-registered chip arrangement position (nxa, nya) on the scanning electron microscope controller side. A position coordinate value (xst, yst) to be obtained is obtained, and a coordinate calibration point is automatically specified by image collation with a calibration point partial image 26-1 stored in association with this portion. Next, registration is also performed for the characteristic location AP2 of the second sample 4b in the same manner, and (xch2, ych2) is stored. Also, an SEM image is recorded in the same manner as the first feature point AP1.

  Here, while actually observing with a scanning electron microscope while the wafer 21 which is a sample is mounted on the sample stage 1, it is moved manually, for example, by the trackball 13, and the position of the portion to be observed in advance is manually moved on the actual sample. The pointing pen 27 is used to give an instruction. When the measuring point registration MP switch on the control panel (B) 15 is further pressed, the SEM control device 6 side wants to observe the stage coordinate value (xst, yst) and the chip arrangement (nx, ny) at that time. Is converted into in-chip coordinate values (xs, ys) and stored together with the observed partial SEM image. This is shown in the measuring point registration storage device 17 in FIG. For example, in MP1, which is the measurement point registration NO1, the in-chip coordinate value of the observation point is (xm1, ym1) and the chip arrangement is (nxm1, nym1). The position for measuring the length is indicated by a cross in the SEM registered image. Thus, in the subsequent observation, the sample stage should be automatically moved from the in-chip coordinate value (xs, ys) of the observation portion position and the pre-registered chip arrangement (nx, ny) position on the SEM control device 6 side. The coordinate value (xst, yst) is obtained, and the observation part is automatically specified by image collation with the observation partial image 26-2 stored in association with this part. When the registration of the feature points is completed, the next time the sample stage 1 is mounted on the wafer, coordinate calibration is performed by collating the positions of the registered feature points, and the coordinate indication position accuracy is increased.

  Next, an observation stage is entered. First, coordinate calibration between the stage coordinates Xst-Yst and the designated coordinates Xa-Ya is performed. By operating the trackball 7, the characteristic location AP1 of the first sample 4a is moved to the center of the field of view of the microscope, and the coordinate calibration AP1 switch is operated. As a result, the SEM control device 6 waits for input from the control panel (A) 7. Next, on the SEM image 9 of the control panel (A) 7, the location of the feature point used for coordinate calibration is indicated by the pointing pen 12. The control panel (A) 7 outputs the coordinate value of the visual field coordinate Xm-Ym projected in the SEM image. The coordinates Xm-Ym at this time have an origin at the center of the visual field, for example, and coincide with the stop position of the sample stage. In the SEM control device 6, the position coordinate calibration point AP1 position (x1, y1) is calculated by the arithmetic of the coordinate value (xm1, ym1) and the read value (xL1, yL1) of the linear scale 24 indicating the position of the sample stage. It is calculated by the same method as when registering. The second feature point AP2 is also obtained as (x2, y2) by the same method as AP1.

  When (x1, y1) and (x2, y2) are obtained as described above, by comparing and calculating between the positions (xch1, ych1) and (xch2, ych2) at the time of registration, The coordinate calibration between the Xa-Ya coordinate and the sample stage coordinate Xst-Yst is performed. Coordinate calibration can be performed in the same manner as Japanese Patent Publication No. 63-94548.

  The SEM control device 6 that has received the chip arrangement (nx, ny), the in-chip coordinate value (xs, ys), and the visual field shift amount (dx, dy) stored in the vertex registration storage device 17 observes and measures the length. The designated coordinate value (xa, ya) of the power pattern portion is obtained and output to the coordinate value conversion device 12. At this time, conversion according to [Equation 5] is performed.

[Equation 5]
dx = fsx (dxm (1), dxm (2),…)
dy = fsy (dym (1), dym (2),…)

  By the way, since it is necessary to limit the storage capacity to a certain degree in practice, the field-of-view shift amount cannot be held and used in an infinite number in practice. Therefore, [Formula 6] is used.

[Equation 6]
dx = fsx (dxm (1), dxm (2),… dxm (j))
dy = fsy (dym (1), dym (2),… dym (j))

  Here, fsx and fsy are functions representing the statistical processing method in the present invention. Further, j = mpj (mpj: registered misregistration correction effective data points). Here, there is a method of averaging a predetermined number of results as an example of a method that is easily realized among the statistical processing methods. This is as shown in [Equation 7].

[Equation 7]
dx = (dxm (1) + dxm (2) +… + dxm (j)) / j
dy = (dym (1) + dym (2) +… + dym (j)) / j

  However, j = mpj (mpj: registered misregistration correction effective data points). Another example of the statistical processing method is a method of obtaining a heavy load average for a predetermined number of results closer to the present time. This is as shown in [Equation 8].

[Equation 8]
dx = (D1 _・ dxm (1) + D2 _・ dxm (2) +… + Dj _・ dxm (j)) / j
dy = (D1 _・ dym (1) + D2 _・ dym (2) +… + Dj _・ dym (j)) / j

  However, D1 <D2 <... <Dj and (D1 + D2 +... + Dj) / j = 1, and j = mpj (mpj: registered misregistration correction effective data points). The obtained time is dxm (j) closest to the present time, then dxm (j-1) and below ... dxm (2), dxm (1). The numbers assigned to each element for obtaining dy are in a similar sequence.

  By the way, on the other hand, a corrected movement target position is obtained by reflecting the measurement result of the visual field position deviation measured each time the movement of the sample stage is reflected in the movement target position, and this is stored and stored next time. Alternatively, there is an embodiment based on a method of repeating the use when moving to a corresponding place. In this embodiment, a method of indicating the position to be observed using the indicated coordinates Xa-Ya, a method of performing coordinate calibration between the stage coordinates Xst-Yst, (nxm, nym) representing the measurement point position and The basic mechanism such as obtaining (xm, ym) and registering it in the measuring point registration storage device 17B is the same as in the above embodiment. An equation for obtaining the movement target position (xmt (j + 1), ymt (j + 1)) at the next observation from the measured visual field position shift amount (dxm, dym) is [Equation 9].

[Equation 9]
xmt (j + 1) = xmt (j) −dxm (j)
ymt (j + 1) = ymt (j) −dym (j)

By repeating this arithmetic every time the movement of the sample stage to the observation position is completed, the movement target position is obtained in time series. The result is automatically stored in the measuring point registration storage device 17B.
In this way, the movement target position at that time is obtained from the visual field position deviation amount every time the movement of the sample stage is completed. The next movement to the same or corresponding observation point is obtained and stored in time series in the past. A plurality of moving target positions are read out from the measuring point registration storage device 17B, further subjected to statistical processing to obtain a new moving target position, and moved to the moving target position, thereby stabilizing the sample. The stopping accuracy of the stage can be ensured. That is, the SEM control device 6 that has received the chip array (nx, ny), the in-chip coordinate values (xs, ys), and the movement target position (xmt, ymt) stored in the measured vertex registration storage device 17B observes and measures. The designated coordinate value (xa, ya) of the pattern portion to be lengthened is obtained and output to the coordinate value conversion device 12. In this case, conversion according to [Equation 10] is performed.

[Equation 10]
xmt = fssx (xmt (S−1), xmt (S−2),…)
ymt = fssy (ymt (S−1), ymt (S−2),…)

  However, the time is traced back to S representing the time immediately before the movement of the sample stage for the next observation, hereinafter S-1, S-2,. By the way, practically, it is necessary to limit the storage capacity to a certain extent, so in practice it is not possible to hold and use an infinite number of movement target positions. Therefore, [Equation 11] is used.

[Equation 11]
xmt = fssx (xmt (S−1), xmt (S−2), ... xmt (S−j))
ymt = fssy (ymt (S−1), ymt (S−2),… ymt (S−j))

  Here, fssx and fssy are functions representing the statistical processing method in the present invention. Further, j = mpj (mpj: registered misregistration correction effective data points). Here, there is a method of averaging a predetermined number of results as an example of a method that is easily realized among the statistical processing methods. This is as shown in [Equation 12].

[Equation 12]
xmt = (xmt (S−1) + xmt (S−2) +… + xmt (S−j)) / j
ymt = (ymt (S−1) + ymt (S−2) +… + ymt (S−j)) / j

  However, j = mpj (mpj: registered misregistration correction effective data points). Another example of the statistical processing method is a method of obtaining a heavy load average for a predetermined number of results closer to the present time. This is as shown in [Equation 13].

[Equation 13]
xmt = (D1 _・ xmt (S−1) + D2 _・ xmt (S−2) +… + Dj _・ xmt (S−j)) / j
ymt = (D1 _・ ymt (S−1) + D2 _・ ymt (S−2) +… + Dj _・ ymt (S−j)) / j

  However, D1 <D2 <... <Dj and (D1 + D2 +... + Dj) / j = 1, and j = mpj (mpj: registered misregistration correction effective data points). In addition, for example, in xmt, xmt (S−1) is closest to the present time, and is next to xmt (S−2) or less, followed by xmt (S−j + 1) and xmt (S−j). The numbers assigned to each element to obtain ymt are in the same sequence.

  On the other hand, the coordinate value conversion device 12 converts this into a stage coordinate value (xst, yst) indicating the position where the sample stage 1 should actually move, and outputs it to the stage control device 5. In the stage control device 5, the coordinate values (xL, yL) indicated by the linear scale 24 coincide with this coordinate value, or until the sample position 1 coincides with the corresponding position coordinate value, the motor 2 connected to each of the XY directions and 3 is driven. When the movement of the sample stage 1 to a predetermined position is completed, a predetermined operation such as observation or length measurement is performed on the image of the scanning electron microscope, and movement to the next observation point is started after completion.

  In the above method, when the wafer is reloaded from the next observation after the registration of the feature points, the feature points are manually specified using the pointing pen 27 or the like. However, it is also possible to perform automatic recognition by comparing the pattern shape near the feature point projected in the SEM image with the feature point SEM image stored in association with the position when the feature point is registered. It is. This makes it possible to quickly and accurately perform automatic coordinate calibration, automatic observation position extraction, automatic observation, and automatic length measurement in automatic observation of a microscope. Moreover, the automatic recognition by these image collation is similarly applied to the designation of the observation position.

  By implementing the present invention as described above, for example, as shown in FIG. 19, the entire real chip indicated by the solid line in the grid based on the design data indicated by the broken line is always inclined at an angle. As shown in FIG. 20, even when each solid chip is tilted with respect to the broken design data grid, it is displaced from the ideal position due to various factors such as being tilted with each chip. Thus, it is possible to accurately capture the positions of the measurement points mp1 to mp9 that are in the field of view of the scanning electron microscope.

  As described above, according to the embodiment of the present invention, as shown in FIG. Therefore, it is possible to perform a more accurate positioning operation of the sample stage in a microscope or the like suitable for the purpose of repeatedly observing the same sample.

  By the way, in the above embodiment, the case where it is desired to perform automatic observation according to the same observation procedure between different apparatuses is not considered. On the other hand, there are many requests for arranging a plurality of observation devices of the same type and performing automatic observation according to the same observation procedure. In response to such a requirement, it is realized by exchanging and using observation procedure records (including calibration point memory and measurement point memory).

  In many cases, there are variations in the accuracy of the production of the sample stage between a plurality of scanning electron microscopes. Therefore, it is necessary to always clarify with which apparatus the recorded displacement detection result is detected. In the present invention, a device identification symbol (for example, a device manufacturing number) is registered in the observation procedure storage device 40 at the same time as the positional deviation amount (dxm, dym) detection result.

  In FIG. 13, after the tip array (nxm1, nxm2) and the in-chip coordinate values (xm1, ym1) of the target movement position of the sample stage are found at the next observation position before starting the specimen stage movement, the positional deviation amount (dxm , dym) is stored in the observation procedure storage device 40A in association with the device identification symbol (for example, device serial number), and is read into the SEM control device and used as stage position correction data. .

  When the apparatus identification device 52-1 receives the read request signal rreq for misregistration correction data from the SEM control device, the device identification device 52-1 is individually provided in the scanning electron microscope apparatus B (51) of the present invention. The device identification symbol “B” is read from the device identification storage unit 67 and transmitted to the observation procedure storage device 40-1. The observation procedure storage device 40-1 outputs the positional deviation correction data (dxmb1, dymb1), (dxmb2, dymb2),..., (Dxmbi, dymbi) associated with the device number DevNoB to the SEM control device. The positional deviation correction, which is a feature of This is indicated by a solid line arrow. Here, for example, when the device identification symbol “A” is stored from the device identification storage unit 67, the positional deviation correction data (dxma1, dyma1), (dxma2, dyma2) associated with the device number DevNoA as indicated by the broken line arrow. ), ..., (dxmai, dymai) are output to the SEM control device.

The flow of the above processing is shown in the flowchart of FIG.
Step SB001: Sample stage movement command generation step SB002: Deviation amount request generation from SEM control body Step SB003: Deviation amount storage controller reads device name B stored in the observation procedure storage device in advance, and a deviation corresponding to DvNoB Access the location where quantity data is stored.
Step SB004: Send deviation data from the storage location to the SEM control body.
Step SB005: The SEM control main body corrects the sample stage movement target position based on the received data to obtain a new target position.
Step SB006: Start moving the sample stage to the target position

  FIG. 15 shows the positional deviation amount (dxm) that is measured when the tip array (nxm1, nxm2) and the in-chip coordinate values (xm1, ym1) at the observation position have already been stopped after the sample stage is stopped. , dym) is associated with a device identification symbol (for example, a device serial number) and stored in the observation procedure storage device 40 in a block diagram.

  When the device identification device 52-1 receives the write error correction signal wreq from the SEM control device, the device identification device 52-1 is individually provided in the scanning electron microscope apparatus B (51) of the present invention. The device identification symbol “B” is read from the device identification storage unit 67 and transmitted to the observation procedure storage device 40-1. In the observation procedure storage device 40-1, the positional deviation correction data storage locations (dxmb1, dymb1), (dxmb2, dymb2), ..., (dxmbi, dymbi) associated with the device number DevNoB are output from the SEM control device. The amount of displacement is stored sequentially. This is indicated by a solid line arrow. Here, for example, when the device identification symbol “A” is stored from the device identification storage unit 67, the misalignment correction data storage locations (dxma1, dyma1), (dxma2) associated with the device number DevNoA as indicated by the broken line arrows , dyma2),..., (dxmai, dymai), the positional deviation amount is output from the SEM control device.

The flow of the above processing is shown in the flowchart of FIG.
Step SC001: Sample stage movement end step SC002: Observation position deviation amount measurement result confirmation step SC003: Deviation amount storage controller reads device name B stored in the observation procedure storage device in advance, and deviation amount data corresponding to DvNoB Access the location where the is stored.
Step SC004: The positional deviation amount data determined to the storage location is sent from the SEM control main body side.
Step SC005: At the storage location, the received data is stored in order following the previously stored data.
Step SC006: End

  By the way, on the other hand, a corrected movement target position that reflects the measurement result of the visual field position deviation measured every time the movement of the sample stage is reflected in the movement target position is stored, and this is stored and stored next time or There is an embodiment based on a method of repeating the use when moving to a corresponding location. In this embodiment, not the measured visual field position shift amount (dxm, dym) but the movement target position (xmt, ymt) corrected using the shift amount is handled.

  In FIG. 26, after the tip array (nxm1, nxm2) and the in-chip coordinate values (xm1, ym1) of the target movement after correction of the sample stage are found at the next observation position before starting the movement of the sample stage, the corrected movement target has already been obtained. A block when the position (xmt, ymt) is associated with an apparatus identification symbol (for example, an apparatus manufacturing number) and stored in the observation procedure storage apparatus 40B, and is read into the SEM control apparatus and used as stage position correction data The figure is shown.

  When the device identification device 52-1 receives the corrected position target position data read request signal rreq from the SEM control device, the device identification device 52-1 is individually provided in the scanning electron microscope apparatus B (51) of the present invention. The device identification symbol “B” is read from the device identification storage unit 67, which is transmitted to the observation procedure storage device 40-1. The observation procedure storage device 40-1 outputs corrected movement target position data (xmtb1, ymtb1), (xmtb2, ymtb2),..., (Xmtbi, ymtbi) associated with the device number DevNoB to the SEM control device, Misalignment correction, which is a feature of the present invention, is performed. This is indicated by a solid line arrow. Here, for example, when the device identification symbol “A” is stored in the device identification storage unit 67, the positional deviation correction data (xmta1, ymta1), (xmta2, ymta2) associated with the device number DevNoA as indicated by the broken line arrow. ), ..., (xmtai, ymtai) are output to the SEM controller.

The flow of the above processing is shown in the flowchart of FIG.
Step SE001: Sample stage movement command generation step SE002: Deviation amount request generation from the SEM control body Step SE003: The deviation amount storage controller reads the device name B stored in the observation procedure storage device in advance, and corrects it corresponding to DvNoB Access the location where the rear movement target position data is stored.
Step SE004: Send the corrected movement target position data from the storage location to the SEM control main body.
Step SE005: The SEM control main body corrects the sample stage movement target position based on the received data to obtain a new target position.
Step SE006: Start moving the sample stage to the target position

  FIG. 28 shows the positional deviation amount (dxm) that is measured when the tip array (nxm1, nxm2) and in-chip coordinate values (xm1, ym1) at the observation position after the sample stage is stopped are in the observation position. , dym) is a block diagram showing how the corrected movement target position (xmt, ymt) obtained in association with the device identification symbol (for example, device manufacturing number) is stored in the observation procedure storage device 40B.

  When the apparatus identification device 52-1 receives the corrected movement target position data write request signal wreq from the SEM control device, the apparatus identification device 52-1 is individually provided in the scanning electron microscope apparatus B (51) of the present invention. The device identification symbol “B” is read from the device identification storage unit 67, which is transmitted to the observation procedure storage device 40-1. In the observation procedure storage device 40-1, the corrected movement target position data storage locations (xmtb1, ymtb1), (xmtb2, ymtb2), ..., (xmtbi, ymtbi) associated with the device number DevNoB are output from the SEM control device. The corrected corrected movement target positions are sequentially stored. This is indicated by a solid line arrow. Here, for example, when the device identification symbol “A” is stored from the device identification storage unit 67, the corrected movement target position data storage location (xmta1, ymta1), which is associated with the device number DevNoA, as indicated by the broken line arrow, The positional deviation amount is output from the SEM control device to (xmta2, ymta2),..., (xmtai, ymtai).

The flow of the above processing is shown in the flowchart of FIG.
Step SF001: Sample stage movement end step SF002: Observation position deviation amount measurement result determination step SF003: The deviation amount storage controller reads the device name B stored in the observation procedure storage device in advance and moves after correction corresponding to DvNoB. Access the location where the target position data is stored.
Step SF004: Send corrected corrected target position data to the storage location from the SEM control main body side.
Step SF005: At the storage location, the received data is stored in order following the previously stored data.
Step SF006: End

  As described above, for example, an observation procedure record is shared between a scanning electron microscope apparatus A having a stage positioning error as shown in FIG. 17 and a scanning electron microscope apparatus B having a stage positioning error as shown in FIG. However, there is no stage positioning error caused by the difference in the amount of displacement. According to the present invention as described above, it is possible to realize field-of-view with higher positional accuracy with a scanning electron microscope. The present invention can be applied not only to a scanning electron microscope but also to other observation apparatuses and microscopes equipped with a sample stage.

The figure which shows the outline | summary of the scanning electron microscope which is an example of the charged particle beam apparatus by this invention. Explanatory drawing of a control panel (A). Explanatory drawing of a control panel (B). The figure which shows an example of a calibration point registration memory | storage device. The figure which shows an example of the measuring point registration memory | storage device. Explanatory drawing of a position shift correction function ON / OFF registration storage device. Explanatory drawing of a position shift amount correction use point number registration storage device. The flowchart which shows an example of a process procedure. The figure which shows the mechanism which calculates | requires a measurement point position by image collation. The figure which shows the mechanism of the coordinate value conversion between a stage coordinate and an instruction | indication coordinate. The flowchart which shows an example of an operation | movement sequence. Explanatory drawing of an observation procedure memory | storage device. The figure which shows the reading state from an observation procedure memory | storage device. The flowchart which shows an example of an operation | movement sequence. The figure which shows the write-in state to an observation procedure memory | storage device. The flowchart which shows an example of an operation | movement sequence. The figure which shows an example of sample stage position accuracy. The figure which shows the other example of sample stage position accuracy. The figure which shows the state which the whole real chip inclines at a certain angle with respect to design. The figure which shows the state inclined for every real chip | tip with respect to a design data lattice. Explanatory drawing of a positional deviation amount measurement function ON / OFF registration storage device. The flowchart which shows an example of an operation | movement sequence. The figure which shows an example of the measuring point registration memory | storage device. The figure which shows an example of an observation procedure memory | storage device. The figure which shows the mechanism of the coordinate value conversion between a stage coordinate and an instruction | indication coordinate. The figure which shows the reading state from an observation procedure memory | storage device. The flowchart which shows an example of an operation | movement sequence. The figure which shows the write-in state to an observation procedure memory | storage device. The flowchart which shows an example of an operation | movement sequence. The figure which shows a mode that a positional offset amount is converged whenever it repeats the number of observations.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Sample stage, 2 ... Motor X, 3 ... Motor Y, 4 ... Sample (chip), 5 ... Stage control apparatus, 6 ... SEM control apparatus, 7 ... Control panel (A), 8 ... Coordinate value input window, 9 ... SEM image (image memory), 10 ... Chip arrangement instruction coordinates, 11 ... In-chip position instruction coordinates, 12 ... Coordinate value converter, 13 ... Trackball, 14 ... Linear scale controller, 15 ... Control panel (B), 16 ... calibration point registration storage device, 17 ... measurement point registration storage device, 18 ... electron source, 19 ... electron lens, 20 ... deflector, 21 ... wafer, 22 ... detector, 23 ... image generation device, 24-1 ... Y linear scale, 24-2 ... X linear scale, 26 ... registered storage SEM image, 27 ... pointing pen, 28 ... calibration procedure file, 29 ... length measurement procedure file, 30 ... image frame, 31 ... image frame position indicator, DESCRIPTION OF SYMBOLS 2 ... Image collation apparatus, 33 ... Gate, 34 ... Deflection offset indicator, 35 ... Deflection indicator, 36 ... Amplifier, 37 ... Scale controller, 38-1 ... X scale, 38-2 ... Y scale, 40 ... Observation procedure storage device 50... Scanning electron microscope device A 51. Scanning electron microscope device B 52. Device identification device 53. Device symbol recording device 61 61 Position misalignment correction function ON / OFF registration storage device 62 ... misregistration amount correction use point registration storage device, 65 ... misregistration correction function execution status display, 66 ... correction effective data point input window, 67 ... device identification storage unit adj ... ON signal when image collation coincides, pcj (PCJ) ... Position correction effective data points, rreq ... Position correction data request signal at reading, wreq ... Position correction data request signal at writing

Claims (9)

A charged particle beam source that generates a charged particle beam, a sample stage that holds and moves the sample, a lens that converges the charged particle beam emitted from the charged particle beam source onto the sample, and deflects the charged particle beam Deflector, image detecting means for detecting an image of the sample, image display means for displaying the detected image, coordinate indicating means for indicating the position on the sample, coordinate values on the coordinate indicating means, and the sample In a charged particle beam apparatus comprising: means for correlating a coordinate value on a stage so as to be calibrated; and means for moving the sample stage to the position of the sample stage corresponding to the coordinate value designated on the coordinate designating means,
When observing observation positions at the same corresponding position on a plurality of samples manufactured by the same process, the sample stage is moved to a movement target position corresponding to the observation position indicated by the coordinate instruction means. A positional deviation amount calculating means for calculating the positional deviation amount of the observation position by the image detecting means after the movement is completed, and a storage means for storing the calculated positional deviation amount, and obtained by the positional deviation amount calculating means. Based on the positional deviation amount, the movement target position coordinate value adopted when moving to the observation position corresponding to the observation position on the different sample manufactured by the same process after the next time is controlled, and the sample stage is stopped when the sample stage is stopped. A positional deviation correction means that operates so that a predetermined position of the sample and a predetermined position of the image display means coincide;
Device identification means,
The storage means for storing the positional deviation amount stores the positional deviation amount with an apparatus identification symbol, and is identified by the apparatus identification means when the movement target position of the sample stage is determined by the positional deviation correction means. A charged particle beam apparatus characterized in that a movement target position of the sample stage is determined based on a result of statistical processing of a positional deviation amount to which an apparatus identification symbol is assigned. A charged particle beam source that generates a charged particle beam, a sample stage that holds and moves the sample, a lens that converges the charged particle beam emitted from the charged particle beam source onto the sample, and deflects the charged particle beam Deflector, image detecting means for detecting an image of the sample, image display means for displaying the detected image, coordinate indicating means for indicating the position on the sample, coordinate values on the coordinate indicating means, and the sample In a charged particle beam apparatus comprising: means for correlating a coordinate value on a stage so as to be calibrated; and means for moving the sample stage to the position of the sample stage corresponding to the coordinate value designated on the coordinate designating means,
When observing observation positions at the same corresponding position on a plurality of samples manufactured by the same process, the sample stage is moved to a movement target position corresponding to the observation position indicated by the coordinate instruction means. A positional deviation amount calculating means for calculating the positional deviation amount of the observation position by the image detecting means after the movement is completed, and correcting the movement target position coordinate value employed at that time using the calculated positional deviation amount. Means for determining the corrected movement target position coordinate value, and storage means for storing the calculated corrected movement target position coordinate value, based on the corrected movement target position coordinate value stored in the storage means, Control the movement target position coordinate value used when moving to the observation position corresponding to the observation position on a different sample manufactured by the same process after the next time, and when the sample stage is stopped And positional deviation correcting means operates to the predetermined position coincides with a predetermined position and the image display unit of the sample,
Device identification means,
The storage means for storing the corrected movement target position coordinate value stores the corrected movement target position coordinate value with a device identification symbol, and determines the movement target position of the sample stage by the positional deviation correction means. The charged particle is characterized in that the movement target position of the sample stage is determined based on a result of statistical processing of the corrected movement target position coordinate value to which the apparatus identification symbol identified by the apparatus identification unit is assigned. Wire device.   The charged particle beam apparatus according to claim 1 or 2, further comprising an observation procedure storage device that stores an observation scheduled position, an observation partial image, and an observation order thereof, and is associated with the observation planned position coordinate value in the observation procedure storage device. A charged particle beam apparatus that stores the displacement amount or the corrected movement target position coordinate value.   In the charged particle beam apparatus according to claim 1, 2 or 3, when controlling the movement target position coordinates when moving to an arbitrary observation position, the positional deviation amount for a plurality of times obtained before or A charged particle beam apparatus using a statistical processing result of a corrected movement target position coordinate value.   5. The charged particle beam apparatus according to claim 4, further comprising means for preliminarily designating a position shift amount calculation point that is effective retroactively by the position shift amount calculation means.   The charged particle beam apparatus according to claim 1, further comprising a unit that switches the position deviation correction unit between valid and invalid.   7. The charged particle beam apparatus according to claim 6, further comprising means for preliminarily storing designation of validity or invalidity of the positional deviation correction means, and controlling validity or invalidity of the positional deviation correction means during automatic observation. Charged particle beam device.   6. The charged particle beam apparatus according to claim 1, further comprising means for switching the position deviation amount calculating means between valid and invalid.   9. The charged particle beam apparatus according to claim 8, further comprising a unit that stores in advance designation of validity or invalidity of the positional deviation amount calculation unit, and controls validity or invalidity of the positional deviation amount calculation unit during automatic observation. Characterized charged particle beam device.

Images