JPH08285539A - Pattern measuring method and device - Google Patents

Abstract

PURPOSE: To accurately measure the overlapping dimension deviation of the pattern of an LSI wafer with a high step difference by forming each of light images from two patterns onto the image sensor of a detection optical system by applying light to a region on a sample. CONSTITUTION: A lighting optical system 2 is provided with a light source 1, and a detection optical system 8 is provided with an objective lens system 3, a branch optical system 4, relay optical systems 5a and 5b of different light path lengths with a plurality of focusing positions, a plurality of image sensors 6a and 6b corresponding to each light path, and an automatic focusing system 7. A stage system 11 is provided with a traveling stage 9 and a sample stand 10 for placing a sample 13. Then, a control system 12 drives and controls the Z stage of the stage system 11 based on the automatic focusing information on the surface of the sample 13 for the objective lens system 3 obtained from the automatic focusing system 7, and finely moves the surface of the sample 13 to a proper focusing distance to position the sample 13 in Z-axis direction, thus detecting the signals at different focusing positions simultaneously with the sensors 6a and 6b, since the light path length each of the relay optical systems 5a and 5b is different from each other.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and apparatus for measuring the size of a pattern having a height difference on a sample, for example, for measuring the size deviation of the overlay pattern of an LSI wafer so that a pattern with a high level difference can be dealt with. Regarding

[0002]

2. Description of the Related Art As prior art 1, a paper ("MINIMISING
OPTICAL OVERLAY MEASUREMENT ERRORS ", SPIE Vol.1926
pp.450-462), double and triple target patterns are detected at the height where the sample stand is moved up and down and a specific position is in focus, or at multiple heights, and these are processed to obtain a high step. It is described that the dimension of the superposition pattern of is measured. Tool-induced shift (TIS: Tool Induced Shi)
ft) measurement is also described. Further, in Prior Art 2 (Japanese Patent Laid-Open No. 58-48918), an optical image of an alignment pattern formed on a mask and a reflected light image of an alignment pattern formed on a wafer are each provided in a proximity exposure apparatus. A technique for aligning the mask and the wafer relative to each other so that the optical path length is changed by the objective lens and an image is formed on the sensor to obtain the deviation amount of the centers of both patterns from the image signal obtained from the sensor and the deviation is eliminated Has been described. Further, in the prior art 3 (Japanese Patent Laid-Open No. 6-232228), each of the reflected light images of a plurality of patterns having a step formed on the wafer has an optical path length according to the height corresponding to the step. A technique is described in which the relative displacement of a plurality of patterns is measured based on clear image data obtained by forming an image on a sensor by using a different detection optical system.

[0003]

In recent years, the patterns of LSI wafers have become finer and have higher steps, and the overlay accuracy in each process has become severe at 100 nm or less. In addition to this, multi-layer resists are used. Increasing steps are being spurred. In such a situation, in order to ensure the overlay accuracy of 100 nm or less, the base reference pattern and the resist pattern to be measured which serves as a mask for etching the pattern of the next layer formed thereon are to be formed. It has become necessary to measure the deviation accuracy with high accuracy of 10 nm or less and monitor the fluctuation of the process state. However, in the related art 1, it is necessary to move the stage on which the target pattern having the step is placed up and down from the first focus position to the second focus position, and the XY due to the vertical movement of the stage is required.
There is a problem that an error factor is generated due to the deviation of the direction and the occurrence of vibration, and the margin is significantly reduced. Further, in the optical system having a plurality of focal points at the same time as described in Related Art 2, images at a plurality of focal points are overlapped with each other, and an image of a sufficiently focused pattern and an image of a defocused pattern are one. There was a problem in that it was overlapped on the sensor, and sufficient consideration was not given to the dimension measurement accuracy. Further, in the method described in Related Art 3 in which images at a plurality of focal positions are simultaneously detected by a plurality of sensors, there is a problem in that error factors such as drift between sensors and vibration are not sufficiently considered.

An object of the present invention is to solve the above problems.
Overlay dimension deviation of overlay pattern corresponding to high step due to miniaturization of patterns such as LSI wafer is 10 nm
An object of the present invention is to provide the following method and apparatus for measuring a pattern having a height difference on a sample that can be measured with high accuracy. Another object of the present invention is to reduce the overlay dimension deviation of the overlay pattern corresponding to a high step due to the miniaturization of the pattern of the LSI wafer or the like to 10 nm.
An object of the present invention is to provide a measuring device for a pattern having a height difference on a sample which can be measured with the following high accuracy.

[0005]

In order to achieve the above object, the present invention provides a sample in which a first pattern and a second pattern having a height difference in a predetermined region are formed to have the height difference. Accordingly, the optical path length difference between the first optical path and the second optical path in the detection optical system is finely adjusted, and the region on the sample is positioned within the visual field of the detection optical system. To automatically focus the surface of the sample and illuminate an area on the sample so that each of the light images from the first and second patterns is directed by each of the first and second optical paths of the detection optics. By forming an image on the image sensor, detecting an image signal from the image sensor, and setting a range indicating the image signal corresponding to the first and second patterns with respect to the detected image signal, Image signal of pattern and second
And the image signal of the pattern, the center position of the first pattern is calculated from the extracted image signal of the first pattern, and the center position of the first pattern is calculated from the extracted image signal of the second pattern. The center position is calculated, and the dimension between the calculated center of the first pattern and the center of the second pattern is subjected to magnification and offset correction of the detection optical system to calculate an accurate dimension. This is a method for measuring the dimension of a pattern having a height difference on a sample. Further, the present invention provides a first optical path and a second optical path in a detection optical system according to the height difference with respect to a sample on which a first pattern and a second pattern having a height difference in a predetermined area are formed. Fine adjustment of the optical path length difference between, the region on the sample is positioned in the field of view of the detection optics, the surface of the sample is autofocused with respect to the detection optics, the region on the sample And illuminate each of the light images from the first and second patterns to form an image on the image sensor by each of the first and second optical paths of the detection optical system to detect an image signal from the image sensor. Then, the first and second image signals are detected with respect to the detected image signal.
The image signal of the first pattern and the image signal of the second pattern are extracted by setting the range indicating the image signal corresponding to the pattern of the first pattern, and the image signal of the extracted first pattern is first extracted. The center position of the first pattern is calculated by the symmetric pattern matching processing by converting into the digital image signal of the pattern, and the extracted image signal of the second pattern is converted into the digital image signal of the second pattern. The center position of the second pattern is calculated by the symmetric pattern matching process, and the magnification and offset of the detection optical system with respect to the calculated dimension between the center of the first pattern and the center of the second pattern. It is a method for measuring the size of a pattern having a height difference on a sample, which is characterized in that an accurate size is calculated by performing correction.

Further, according to the present invention, the first optical path and the second optical path in the detection optical system are formed according to the height difference with respect to the sample in which the reference mark and the resist pattern having the height difference in a predetermined area are formed. Fine adjustment of the optical path length difference between, position the region on the sample in the field of view of the detection optics, automatically focus the surface of the sample with respect to the detection optics, in the region on the sample By illuminating, each of the light images from the reference mark and the resist pattern is imaged on an image sensor by each of the first and second optical paths of the detection optical system to detect an image signal from the image sensor, The image signal of the reference mark and the image signal of the resist pattern are extracted by setting a range indicating the image signal corresponding to the reference mark and the resist pattern with respect to the detected image signal, and the respective image signals are extracted. The center position of the reference mark is calculated from the image signal of the reference mark obtained, the center position of the resist pattern is calculated from the image signal of the extracted resist pattern, and the center of the calculated reference mark and the center of the resist pattern are calculated. A dimension measuring method of a pattern having a height difference on a sample is characterized in that a magnification of the detection optical system and an offset correction are applied to a dimension between them to calculate an accurate dimension. The present invention is also directed to the first area having a height difference in a predetermined area.
The sample stage on which the sample having the pattern and the second pattern is placed, the illumination optical system for illuminating the area on the sample, and the first optical path and the second optical path having optical path length differences from each other. And a fine control optical system for finely controlling the optical path length difference between the first optical path and the second optical path according to the height difference between the first pattern and the second pattern, A detection optical system that forms an image of each of the light images from the first and second patterns on an image sensor by each of the first and second optical paths and detects an image signal from the image sensor, and the sample. Positioning means for moving the stage to position an area on the sample within the field of view of the detection optical system; automatic focusing means for automatically focusing the surface of the sample with respect to the detection optical system; Detect from optical image sensor The image signal of the first pattern and the image signal of the second pattern are extracted by setting a range indicating the image signals corresponding to the first and second patterns with respect to the generated image signal, and the respective image signals are extracted. The center position of the first pattern is calculated from the extracted image signal of the first pattern, the center position of the second pattern is calculated from the extracted image signal of the second pattern, and the calculated first position is calculated. On the sample, there is provided a calculating means for calculating an accurate dimension by performing magnification and offset correction of the detection optical system with respect to a dimension between the center of the pattern and the center of the second pattern. It is a dimensional measuring device for patterns having height differences.

Further, according to the present invention, in the dimension measuring apparatus for a pattern having a height difference on the sample, a relay optical system is provided in each of the first optical path and the second optical path of the detection optical system. And Further, according to the present invention, in the dimension measuring apparatus for a pattern having a height difference on the sample, each of the first optical path and the second optical path of the detection optical system is provided with a relay optical system having different optical path lengths. Is characterized by. The present invention also provides a dimension measuring apparatus for a pattern having a height difference on the sample, wherein each of the first optical path and the second optical path of the detection optical system is provided with a relay optical system, and the optical path is used as the image sensor. It is characterized by having a plurality corresponding to.

Further, according to the present invention, in the dimension measuring apparatus for a pattern having a height difference on the sample, the fine control optical system in the detection optical system is provided in either the first optical path or the second optical path. It is characterized by Further, the present invention is characterized in that, in the dimension measuring apparatus for a pattern having a height difference on the sample, the fine control optical system in the detection optical system is constituted by a birefringent element.

Further, according to the present invention, a sample stage on which a sample having a first pattern and a second pattern having a height difference in a predetermined region is placed, and the sample stage is moved to move the region on the sample. Positioning means for positioning within the field of view of the detection optical system, automatic focusing means for automatically focusing the surface of the sample with respect to the detection optical system, and illumination optical system for illuminating an area on the sample, A first optical path and a second optical path having a relay optical system are formed so as to have an optical path length difference, and the first optical path and the second optical path are formed in accordance with a height difference between the first pattern and the second pattern. A birefringent optical element for finely controlling a difference in optical path length between the optical path and the second optical path is provided, and each of the optical images from the first and second patterns is provided by each of the first and second optical paths. Bonded on each of the first and second image sensors. A detection optical system for detecting an image signal from each image sensor, a positioning means for moving the sample stage to position an area on the sample within the visual field of the detection optical system, and the detection optical system. Automatic focusing means for automatically focusing the surface of the sample, and a range indicating image signals corresponding to the first and second patterns with respect to the image signals detected by the image sensors of the detection optical system are set. To extract the image signal of the first pattern and the image signal of the second pattern,
The center position of the first pattern is calculated from the image signals of the extracted first patterns, and the center position of the second pattern is calculated from the image signals of the extracted second patterns,
And a calculating means for calculating a correct dimension by applying magnification and offset correction of the detection optical system to the calculated dimension between the center of the first pattern and the center of the second pattern. A dimension measuring apparatus for a pattern having a height difference on a sample, which is characterized in that

Further, according to the present invention, there is provided a sample stage on which a sample having a first pattern and a second pattern having a height difference in a predetermined area is placed, and an illumination optical system for illuminating an area on the sample. , A first optical path and a second optical path having an optical path length difference from each other
And fine control optics for finely controlling the optical path length difference between the first optical path and the second optical path according to the height difference between the first pattern and the second pattern. A system, wherein each of the light images from the first and second patterns is transferred to the first
And a detection optical system that forms an image on the image sensor by each of the second optical paths and detects an image signal from the image sensor, and the sample stage is moved so that an area on the sample is within the visual field of the detection optical system. Positioning means for positioning the sample surface, automatic focusing means for automatically focusing the surface of the sample with respect to the detection optical system, and first and second image signals detected by the image sensor of the detection optical system. The image signal of the first pattern and the image signal of the second pattern are extracted by setting the range indicating the image signal corresponding to the second pattern, and the image signals of the first patterns extracted respectively are The image signal of the second pattern extracted by converting the digital image signal of the first pattern and calculating the center position of the first pattern by symmetric pattern matching processing. To a digital image signal of the second pattern and calculates the center position of the second pattern by symmetric pattern matching processing, and calculates the center position of the second pattern between the calculated center of the first pattern and the center of the second pattern. The dimension measuring apparatus for a pattern having a height difference on a sample, further comprising: a calculating unit that performs the magnification of the detection optical system and the offset correction with respect to the dimension to calculate an accurate dimension.

Further, according to the present invention, a sample stage on which a sample having a first pattern and a second pattern having a height difference in a predetermined area is placed, and the sample stage is moved to move the area on the sample. Positioning means for positioning within the field of view of the detection optical system, automatic focusing means for automatically focusing the surface of the sample with respect to the detection optical system, and illumination optical system for illuminating an area on the sample, A first optical path and a second optical path having a relay optical system are formed so as to have an optical path length difference, and the first optical path and the second optical path are formed in accordance with a height difference between the first pattern and the second pattern. A birefringent optical element for finely controlling a difference in optical path length between the optical path and the second optical path is provided, and each of the optical images from the first and second patterns is provided by each of the first and second optical paths. Bonded on each of the first and second image sensors. A detection optical system for detecting an image signal from each image sensor, a positioning means for moving the sample stage to position an area on the sample within the visual field of the detection optical system, and the detection optical system. Automatic focusing means for automatically focusing the surface of the sample, and a range indicating image signals corresponding to the first and second patterns with respect to the image signals detected by the image sensors of the detection optical system are set. To extract the image signal of the first pattern and the image signal of the second pattern,
The extracted image signals of the first pattern are converted into digital image signals of the first pattern, the center position of the first pattern is calculated by the symmetric pattern matching process, and the extracted second pattern of the second pattern is calculated. The image signal is converted into a digital image signal of the second pattern, the center position of the second pattern is calculated by symmetric pattern matching processing, and the calculated center of the first pattern and the center of the second pattern are calculated. A dimension measuring apparatus for a pattern having a height difference on a sample, characterized by comprising a calculating means for performing a magnification of the detection optical system and an offset correction for a dimension between is there.

[0012]

With the above-mentioned structure, the dimension between these patterns can be measured with high accuracy for a sample having a plurality of patterns with high steps. Further, with the above-described configuration, the reference pattern of the base having a high step height due to the miniaturization of the pattern of the LSI wafer and the resist pattern to be measured serving as a mask for etching the pattern of the next layer formed thereon Deviation accuracy of 10 nm
It has become possible to monitor the process state fluctuations by measuring with the following high precision. Further, with the above-described configuration, a multilayer resist is used, and even if the step height is further increased, the reference pattern of the base and the resist pattern of the measurement object, which serves as a mask for etching the pattern of the next layer formed thereon, are formed. It is possible to measure the deviation accuracy with high accuracy of 10 nm or less and with reproducibility, and it is possible to monitor the variation of the process state corresponding to the multilayer resist.

[0013]

Embodiments of the present invention will be described with reference to the drawings. First, the measurement target for measuring (measuring) the dimensions of the pattern will be described. In recent years, as patterns of LSI wafers have become finer and multilayer resists have been used, step heights have become higher and higher, and the overlay accuracy in each process has become severe at 100 nm or less. For the purpose of measuring the overlay accuracy, the pattern measuring method and apparatus according to the present invention, for example, as shown in FIG. 1, FIG. 8, FIG. 9, and FIG.
The measurement target (measurement pattern) 100 formed on the sample (semiconductor wafer) is a base reference pattern (reference mark) 101 and the measurement target is a mask for etching the pattern of the next layer formed thereon. The deviation accuracy from the resist pattern 102 is measured with an accuracy of 10 nm or less, and the fluctuation of the process state is monitored. Then, this monitor result is fed back to the projection exposure apparatus or the like as necessary. Further, if the measured position shift is large, the work such as removing the resist is performed.

Next, the pattern measurement for the measurement object 100 will be described. Hereinafter, FIG.
The case will be described. 2 and 3 show the basic configuration of the pattern measuring apparatus according to the present invention. The pattern measuring apparatus shown in FIG. 2 corresponds to an illumination optical system 2 having a light source 1, an objective lens system 3, a branching optical system 4, relay optical systems 5a and 5b having a plurality of focal positions and different optical path lengths, respectively. A detection optical system 8 including a plurality of image sensors 6a and 6b and an automatic focusing system 7, a stage system 11 including a moving stage 9 and a sample table 10 on which a sample 13 is placed.
And a control system (computer) 12 including image processing. The pattern measuring apparatus shown in FIG. 3 includes an illumination optical system 2 having a light source 1, an objective lens system 3, a branching optical system 4, a plurality of optical paths 14a and 14b having different optical path lengths, a combining optical system 15, and a relay optical system 5. Including a detection optical system 8 having one image sensor 6 for detecting signals at different focus positions and an autofocus system 7, a stage system 11 including a moving stage 9 and a sample stage 10 on which a sample 13 is mounted, and image processing. And a control system (computer) 12.

First, in both cases of FIG. 2 and FIG. 3, the control system (computer) 12 is based on the autofocus information of the surface of the sample (semiconductor wafer) 13 with respect to the objective lens system 3 obtained from the autofocus system 7. The Z stage of the stage system 11 is drive-controlled to finely move the surface of the sample (semiconductor wafer) 13 to an appropriate focus position and position it in the Z-axis direction. As a result, in the case of FIG. 2, since the optical path lengths of the relay optical systems 5a and 5b are different, the signals of different focus positions can be simultaneously detected by the sensors 6a and 6b. In the case of FIG. 3,
Since the optical paths 14a and 14b have different optical path lengths, the synthetic optical system 1
The optical image synthesized at 5 has a plurality of focal positions, and the sensor 6 can detect signals at different focal positions at the same time. As a result, since the measurement is performed simultaneously, the stage may move up and down with a reproduction greater than the measurement accuracy,
There is no problem even if the detection optical system 8 is vibrating. Only in the case of FIG. 2, the relay optical systems 5a, 5b having different optical path lengths and the plurality of sensors 6a, 6b corresponding to the respective optical paths, and in the case of FIG. 3, the plurality of optical paths 14a, 14b having different optical path lengths are provided. Since the system is different for the focal position, it is affected by stability such as vibration. Therefore, the offset value of the sensor position fluctuation is corrected at regular intervals, for example, at appropriate intervals such as one hour or one day.

Next, a method of setting the offset value β of the difference between the positions of the image sensors 6a and 6b and the magnification correction value Δα will be described. First, in the computer 12, the offset β is set to 0 and the magnification correction value Δα is set to 1. Next, the calibration sample (wafer having a step formed by etching or the like) is placed on the sample table 10, and the interval between the same step portions of the calibration sample is output from the image sensor 6a in the image processing 21 of the computer 12. The magnification correction value Δα can be calculated by measuring based on the detection signal 30a and the detection signal 30b output from the image sensor 6b and calculating the interval ratio. Next, the sample (semiconductor wafer) 13 or the calibration sample is measured by a usual method. That is, in the image processing 21 in the computer 12, the position of the reference mark 101 on the sensor 6a is determined based on the detection signal 30a output from the sensor 6a.
0, the position of the resist pattern 102 on the sensor 6b is C based on the detection signal 30b output from the sensor 6b.
Measure as b0. At this time, the measured value δ0, which is the relative displacement of the resist pattern 102 with respect to the reference mark 101, has the relationship of the following (Equation 1). δ0 = ((Ca0-Cb0) * Δα) + β (Equation 1) where α is the magnification, Δα is the magnification correction value (magnification deviation: slight variation in magnification) (when a telecentric detection optical system is used, Δα is The calculation process can be performed assuming that there is no sensor), and β is a parameter of the relative displacement of the sensor, that is, an offset value.

Next, the sample (semiconductor wafer) 13 or the calibration sample is rotated 180 degrees and the same measurement is performed. That is,
In the image processing 21 of the computer 12, the position of the reference mark 101 on the sensor 6a is set to Ca1, the sensor 6 is set to the sensor 6 based on the detection signal 30a output from the sensor 6a.
The sensor 8b based on the detection signal 30b output from b
The position of the resist pattern 102 above is measured as Cb1. At this time, the measured value δ1, which is the relative displacement of the resist pattern 102 with respect to the reference mark 101, is
The relationship of the following (Equation 2) is obtained. δ1 = ((Ca1-Cb1) * Δα) + β (Equation 2) Here, Δα and β take the same values as the values of the above (Equation 1). By the way, the sensor 6a and the sensor 6b
Is a fixed relationship, and δ0 and δ1 are the same sample (sample (semiconductor wafer) 13 or calibration sample) 18
Since only 0 rotations have been made, the absolute value of the relative positional deviation is the same, and the sign is inverted. As a result, δ0 + δ1
= 0. Therefore, the value of β is obtained by the following equation (3).

Β = −0.5 * ((Ca0−Cb0) + (Ca1−Cb1)) * Δα (Equation 3) Therefore, the magnification correction value Δα is calculated as described above, so the computer 12 The parameter β of the sensor relative position deviation amount can be calculated based on the above equation (3). That is, the offset value β can be calculated.

Next, an embodiment of the present invention corresponding to the basic structure shown in FIG. 2 will be described with reference to FIG. That is, in FIG.
FIG. 1 is an overall configuration diagram showing an embodiment of a pattern measuring device for overlaying LSI wafers. This pattern measuring apparatus includes an illumination optical system 2 having an illumination light source 1 for illuminating the sample 13 and a condenser lens 25, a detection optical system 8, a moving stage 9 and a sample table 10 on which the sample 13 is mounted. It is composed of a control system (computer) 12 including a stage system 11 and image processing 21. The detection optical system 8 includes an objective lens system 3 including an objective lens 23 and a half mirror 24, a branching optical system 4 which is a beam splitter,
Relay optical systems 5a and 5b, which have different optical path lengths so as to have a plurality of focal points and include a wedge 16 for fine adjustment of the optical path length.
And a common sync signal corresponding to each optical path 2
Image sensors 6a and 6b, which are individual TV cameras, and an autofocus system 7 including an air micro are provided.

Reference numeral 41 is an output means for outputting measured data. Reference numeral 42 designates an input means composed of a keyboard or the like for designating calculation ranges 61 and 62 for image processing on the computer 12 and designing a reference mark 101 and a resist pattern 102 formed on a sample (semiconductor wafer) 13. The information including the information of the kind of the sample is input so that it can be associated with the design information of the reference mark 101 and the resist pattern 102. Further, the input means 42 inputs the design information of the detection optical system 8 and the image processing program into the computer 12. Four
A display 3 displays the detection signals 30, 30a, 30b detected by the sensors 6, 6a, 6b and the calculation ranges 61, 62 input by the input means 42.
Reference numeral 44 is an external storage device, and the detection optical system 8 is input to the computer 12 input by the input means 42 or the floppy 41.
Of the reference mark 101 and the resist pattern 102 formed on the sample (semiconductor wafer) 13 and the like. Computer 12
There is a CPU, a memory for storing programs, and the like therein, which controls the detection signals 30, 30a, 3 detected by the sensors 6, 6a, 6b in the image processing 21.
The amount of deviation of the resist pattern 102 with respect to the reference mark 101 is calculated with an accuracy of 10 nm or less by performing arithmetic processing based on 0b. Then, the calculated result may be output by the output unit 41 or may be directly output to the projection exposure apparatus via the network 46.

Next, the operation of the pattern measuring device will be described. First, on the sample table 10, for example, a measurement target 1 in which a resist pattern 102 to be measured is formed on a base reference pattern (reference mark) 101 as shown in FIG.
The sample No. 00 is placed, for example, by aligning the orientation flat with the reference plane so as not to cause a deviation in the rotation direction (θ direction). The measurement target (sample) placed on the sample table 10
As 100, various types in which the thickness of the resist pattern 102 and the surface state of the sample to be matched by the autofocus system 7 are different can be considered. Therefore, as described above, the information of the type of the measurement target (sample) 100 is input to the computer 12 by the input means 42 in accordance with the measurement target (sample) 100 being placed on the sample table 10. . Then, the computer 12 causes the reference mark 101 and the resist pattern 102 to be based on the design information of the reference mark 101 and the resist pattern 102 formed on the sample (semiconductor wafer) 13 corresponding to the type stored in the external storage device 44. Standard mark 1 corresponding to the standard step difference H0 with
01, the change ΔH in the level difference between the resist pattern 102 and the resist pattern 102 is calculated, and the thickness of the optical path length fine adjustment assembly wedge 16 corresponding to the calculated change ΔH in the level corresponding to the product type is calculated.
The difference in optical path length between the relay optical systems 5a and 5b is the standard step H between the reference mark 101 and the resist pattern 102.
It is configured to correspond to 0. Therefore, the computer 12 sets the reference mark 101 according to the type of the sample 13 between the relay optical system 5a and the relay optical system 5b.
And the thickness of the optical path length fine adjustment wedge 16 is adjusted (controlled) so as to obtain a difference in optical path length corresponding to the step of the resist pattern 102, and the focus offset value of the automatic focusing system 7 which is an air micro is adjusted. Adjust (control). That is, since the autofocus system 7 is for adjusting a plurality of focus positions with respect to the measurement target 100, it is necessary to adjust the focus offset value of the autofocus system 7 according to the type of the measurement target 100. Therefore, the computer 12 sets the standard measurement target 10 based on the design information of the measurement target 100 on the sample (semiconductor wafer) 13 corresponding to the type stored in the external storage device 44.
The focus offset value of the autofocus system 7 can be adjusted (controlled) by calculating the focus offset value of the autofocus system 7 corresponding to 0 for the product type.

Next, the computer 12 displays the measurement object 10
A sample (semiconductor wafer) mounted on the sample table 10 by controlling the movement of the moving stage 9 based on the design information of 0.
The center of the measurement target 100 is located substantially on the optical axis of the detection optical system 8. Next, the detection signal 30a as shown in FIG. 5 is detected from the image sensor 6a and is input to the image processing 21 of the computer 12. In the image processing 21 of the computer 12, the detected signal 30a ′ thus detected is
For example, it is converted into a digital signal and converted into a binarized signal with a predetermined threshold value, the center of the binarized signal is obtained, and the center is the center of the detection visual field 51 of the image sensor 6a (the optical axis of the detection optical system 8). The movable stage 9 is moved and positioned so that the movable stage 9 is almost located at the position (a). In the computer 12, the detected detection signal 30a ′ is displayed on the display 43, and the center of the entire displayed detection signal 30a corresponds to the optical axis of the detection optical system 8 set on the screen of the display 43. The measurement target 100 may be positioned by moving the moving stage 9 so as to match the reference mark.

Next, based on the focus information of the surface of the sample 13 detected from the automatic focusing system 7 such as an air micro, the computer 12 finely controls the Z stage of the stage system 11 in the z-axis direction to set the measuring object 100. Position it at the proper focus position. Next, the measurement target 100 illuminated by the illumination optical system 2
The reflected light from the image sensor 6 is captured (focused) by the objective lens system 3 and divided by the branch optical system 4 which is a half mirror.
A and the image sensor 6b simultaneously detect a detection signal 30a and a detection signal 30b indicating the reference mark 101 and the resist pattern 102 at different focus positions. Waveforms of these detected detection signals 30a and 30b are shown in FIGS. 6 is a waveform in the X direction of the central portion of the measurement target 100 in FIG. 1 detected by the image sensor 6a, and FIG. 7 is a waveform in the X direction of the central portion of the measurement target 100 in FIG. 1 detected by the image sensor 6b. The image sensor 6a has the reference mark 101, and the image sensor 6b has the resist pattern 1
The focus is on 02.

In the image processing 21 of the computer 12, each of the detection signals 30a and 30b is converted into a digital signal, and the center position of each mark is calculated based on the following equation (4). In the equation (4), c is a symmetric center position coordinate, the detection signal is folded back at this symmetric center c and compared, and the symmetric center position C with the highest degree of coincidence is determined as the accurate center position of the detection signal. Is a symmetry pattern matching process. That is, first
The folding center c is determined, and the difference between f (c−x) of −x and + x from this folding center c (f (c−x) −f (c + x))
Is integrated by changing the value of x in the set calculation range, then the value of the folding center c is changed, and the same is integrated, and the folding center c at which the integrated value is the minimum is set at the center position C of the mark. And

[0025]

[Equation 4]

In the image processing 21 in the computer 12, when the center position Ca of the reference mark 101 is calculated by performing the symmetric pattern matching processing, for example, the above-mentioned x
When the center position Cb of the resist pattern 102 is calculated by performing the symmetric pattern matching process using the data of the reference mark center position calculation range 61 which is input and set as the value of the range of, The data of the resist pattern center position calculation range 62 input and set as the value of is used. The fiducial mark 101 calculated in the image processing 21 in the computer 12
The center position of Ca is Ca, and the calculated resist pattern 102
The center position of is Cb. Next, the computer 12 calculates the reference marks 101 based on the formula (5) shown below.
Center position Ca of the resist pattern 102 and center position C of the resist pattern 102
The value δ obtained by subtracting the offset value β of the difference between the positions of the image sensors 6a and 6b calculated as described above from the difference with b, and correcting the magnification correction value (magnification deviation: slight variation in magnification) Δα is It is the amount of deviation in the X direction between the reference mark 101 and the resist pattern 102. The Y direction is calculated by the same method. δ = ((Ca−Cb) * Δα) + β (Equation 5) As a result, it is possible to measure the shift amount δ between the reference mark 101 and the resist pattern 102 on the image sensors 6a and 6b in the X and Y directions. The deviation amount on the sample can be calculated by the computer 12 by dividing the deviation amount δ on the image sensor by the magnification α of the detection optical system 8. When the magnification α is 100 times, if the shift amount δ can be calculated on the image sensor with an accuracy of 1 μm or less, the measurement can be performed on the sample with an accuracy of 10 nm or less. It should be noted that the magnification α is obtained by using the input means 42 such as a keyboard to input the design information of the detection optical system 8 to the computer 1.
Determined by entering 2. Further, from the relationship with the dimension measured by the pattern measuring device using a calibration sample whose dimension is known by using SEM or the like, the above (Equation 5)
The magnification α and the like can also be calculated from the relationship of the formula. That is, this allows the computer 12 to determine an accurate magnification α.

As described above, the deviation of the pattern with respect to one measurement object 100 is measured. After that, it moves to the next measurement target to be measured. This is repeated, and after the measurement is completed for all measurement objects, the sample is removed and the measurement is completed. Next, a first modified example will be described. Although the embodiment in which the TV camera is used as the sensors 6a and 6b has been described, two sets of one-dimensional image sensors in the X and Y directions may be used instead of the TV camera. According to this modified example, since the detection time for one screen is faster than that of a TV camera, high-speed measurement can be realized.

Next, a second modification will be described.
Although the embodiment in which the air micro is used as the auto focus system 7 has been described, an optical auto focus system to which triangulation, blur effect, etc. are applied may be used instead of the air micro.
According to this modified example, since the optical sensor is generally faster than the air micro, high-speed measurement can be realized. Next, a third modified example will be described.
By installing multiple TV cameras close to each other,
The rigidity can be kept high. Alternatively, the sensor 6
If the same sensor element is divided into two as a and 6b and two images are detected by one sensor, stability is high and highly accurate measurement can be realized.

Next, a fourth modification will be described.
If the relay optical systems 5a and 5b are installed as the relay optical system 5 between the branch optical system 4 and the objective lens system 3, the configuration can be simplified. Next, a fifth modified example will be described. Relay optical system 5a, 5b or branch optical system 4
An exit side telecentric optical system is used as the relay optical system 5 installed between the objective lens system 3 and the objective lens system 3. The incident side of the objective lens system 3 is usually formed by a telecentric optical system. Therefore, in the case of this embodiment, since the detection optical system 8 is constituted by the both-side telecentric optical system, the magnification does not change even if the focal position changes, so that highly accurate measurement can be realized. Next, a sixth modified example will be described. Instead of setting the optical path length difference using the optical path length fine adjustment wedge 16 as the initial setting, the computer 1
The reference numeral 2 sets the optimum optical path length difference by controlling the optical path length fine adjustment wedge 16 and checking the detection signal waveform state by displaying the detection signals 30a and 30b on the display 43 when detecting the measurement pattern. To do. According to this embodiment,
Even if the step of the measurement target changes, it is possible to follow, so that highly accurate measurement can be realized. In other words, the thickness of the optical path length adjusting wedge 16 can be adjusted by adjusting the step difference between the reference mark 101 and the resist pattern 102 for each product type.
It is possible to support patterns with various steps. Further, by adjusting the focus offset value of the automatic focusing system 7 which is an air micro, it is possible to cope with a pattern having various steps. According to the embodiment described above, it is possible to realize highly accurate measurement of a high step sample and reduction of the device cost. Further, in both cases, images at a plurality of focal positions do not overlap each other, and there is no fear of performance deterioration. Furthermore, correction operation is also performed for relative position fluctuations of multiple sensors,
The performance can be maintained stably.

Next, another embodiment of the present invention corresponding to the basic configuration of the present invention shown in FIG. 3 will be described with reference to FIG. That is, the pattern measuring apparatus mounts the illumination optical system 2 having the illumination light source 1 for illuminating the sample 13 and the condenser lens 25, the detection optical system 8, the moving stage 9 and the sample 13 on which the measurement target 100 is formed. Control system (computer) having a stage system 11 including the sample table 10 and an image processing 21
It consists of 12. The detection optical system 8 includes an objective lens system 3 having an objective lens 23 and a half mirror 24, a branching optical system 4 which is a beam splitter, and a branching optical system 4 which is branched and has different optical path lengths. It is composed of a plurality of optical paths 14a and 14b including adjusting wedges 16, a synthetic optical system 15 which is a mirror, a relay optical system 5, a sensor 6 which is a TV camera, and an autofocus system 7 which is an air micro. It Incidentally, 26 and 27 are optical paths 14a and 14b.
Is a mirror for forming.

The present embodiment is different from the embodiment shown in FIG. 3 only in the configuration of the detection optical system 8, and the measurement of the deviation amount between the reference mark 101 and the resist pattern 102 is the same as the embodiment shown in FIG. To be done. That is, the sample (semiconductor wafer) 1
For each product type of the measurement target 100 formed in No. 3, the computer 12 uses the optical path corresponding to the step between the reference mark 101 and the resist pattern 102 by a predetermined amount between the two optical paths 14a and 14b. The thickness of the optical path length adjusting wedge 16 is adjusted (controlled) so as to have a difference in length, and the focus offset value of the autofocus system 7 which is an air micro is calculated to adjust the focus offset. Next, sample 13
Is mounted on the sample table 10 and the moving stage 9 is moved to a position where the measurement target 100 is located. Next, the computer 12 finely moves the Z stage of the stage system 11 in the Z axis direction based on the autofocus information detected by the autofocus system 7 to position the measurement target 100 on the sample 13 at an appropriate focus position. Next, the reflected light from the measurement target 100 on the sample 13 illuminated by the illumination optical system 2 is condensed (captured) by the objective lens system 3. Then, the light is split by the branch optical system 4 which is a beam splitter and passes through the two optical paths 14a and 14b,
The combining optical system 15 which is a mirror combines these lights having different optical path lengths, and the image sensor 6 simultaneously detects the detection signals 30 of different focal positions via the relay optical system 5. In the image processing 21 of the computer 12, the detection signal 3
Based on 0, the shift amount δ between the reference mark 101 and the resist pattern 102 in the X and Y directions is calculated (measured) based on the equations (4) and (5). After that, it moves to the next measurement target to be measured. This is repeated, and after the measurement of all the measurement objects is completed, the sample is removed and the measurement is completed.

By the way, in the case of this embodiment, since the detection is performed by one sensor 6, it is possible to eliminate the fluctuation of the relative position in the sensor. In the case of the present embodiment, the patterns of the detected focus positions are devised so that they do not overlap each other on the sensor 6. FIG. 12 shows the waveform of the detection signal 30 detected by the sensor 6 for the measurement target 100 shown in FIG.
Shown in FIG. 12 is a signal waveform in the X direction at the center of the measurement target 100 shown in FIG. 1 detected by the sensor 6. Light path 1
The light passing through 4a is focused on the reference mark 101, and the light passing through the optical path 14b is reflected by the resist pattern 102.
Is focused on. The signal waveform detected by the sensor 6 is
It is the sum of these, and both the reference mark 101 and the resist pattern 102 are in focus to some extent. The image processing 21 of the computer 12 calculates the center position of each mark from this waveform by the above-mentioned equation (4).
At this time, for the center position Ca of the reference mark 101, the data of the reference mark center position calculation range 122 calculated and set by the computer 12 is used, and the resist pattern 10 is used.
The center position Cb of 2 is calculated using the data of the resist pattern center position calculation range 121 calculated and set in the computer 12. These reference mark center position calculation range 122 and resist pattern center position calculation range 12
1 indicates the range of the value of x in the equation (4). The calculated center position of the reference mark 101 is Ca and the center position Cb of the resist pattern 102 is set. Then, the computer 12 subtracts the offset value β of the difference between the positions of the optical paths 14a and 14b from the difference between the mark positions (Ca-Cb) based on the equation (5) to obtain a magnification correction value (difference in magnification) Δα. Is corrected to calculate the shift amount δ between the reference mark 101 and the resist pattern 102 in the X direction. The Y direction is calculated by the same method.

Next, a first modified embodiment will be described.
Instead of the embodiment in which the sensor 6 is a TV camera, a one-dimensional image sensor in each of the X and Y directions may be used. In the case of this embodiment, one sensor in the X direction can detect images at two focal positions. On the other hand, by using two sensors in the Y direction, it is possible to detect an image at each focus position. According to this embodiment, since the detection time for one screen is faster than that of a TV camera, high-speed measurement can be realized.

Next, a second modification will be described.
As the autofocus system 7, an optical autofocus system to which triangulation, blur effect, etc. are applied may be used instead of the air micro embodiment. According to this embodiment, since the optical sensor is generally faster than the air micro, high-speed measurement can be realized. Next, a third modified example will be described. As the synthesizing optical system 15, instead of the embodiment of the beam splitter, a mirror having a hole so as to pass only a portion of the reference mark placed at a position substantially conjugate with the image forming position may be used. According to this embodiment, since the reference mark 101 and the resist pattern 102 are accurately focused, accurate measurement can be performed. However, it is necessary to switch the mirror in order to deal with different measurement targets. Also in this embodiment, the thickness of the optical path length adjusting wedge 16 and the focus offset of the automatic focusing system 7 which is an air micro can be adjusted by adjusting the step between the reference mark 101 and the resist pattern 102 for each product type. Therefore, it becomes possible to deal with a measurement target (measurement pattern) having various steps. Further, since the signal waveform is detected by one sensor 6, the configuration of the sensor can be simplified. Next, another embodiment according to the present invention will be described with reference to FIG. The pattern measuring apparatus has an illumination optical system 2 having a light source 1 for illuminating a sample 13 and a condenser lens 25, a detection optical system 8, a stage system 11 including a moving stage 9 and a sample table 10, and an image processing 21. It is composed of a control system (computer) 12. The detection optical system 8 includes an objective lens system 3 having an objective lens 23 and a half mirror 24, an optical path length changing unit 19 including a plurality of glass plates 17 having partially different thicknesses and a switching mechanism 18 thereof, and an image formation. It is composed of an optical system 20, a sensor 6 which is a TV camera, and an autofocus system 7 made of an air micro.

This embodiment is different from the embodiment shown in FIGS. 3 and 11 only in the structure of the detection optical system 8 and is different from the reference mark 1 in FIG.
The amount of deviation between 01 and the resist pattern 102 is measured in the same manner as in the embodiment shown in FIG. That is, the partial thicknesses of the plurality of glass plates 17 are set so that the optical path length difference corresponding to the step between the reference mark 101 and the resist pattern 102 can be obtained in advance by an amount determined in advance for each product type to be measured. Is selected to adjust the focus offset of the automatic focusing system 7 which is an air micro. Next, the sample 13 is placed on the sample table 10, and the moving stage 9 is moved to a certain position of the measurement target 100. Computer 12
Is finely moving the Z stage of the stage system 11 in the Z-axis direction based on the autofocus information detected by the autofocus system 7 to position the measurement target 100 on the sample 13 at an appropriate focus position. Next, the reflected light from the measurement target 100 of the sample 13 illuminated by the illumination optical system 2 is captured by the objective lens system 3. The light path length changing unit 19 partially changes the light path length of this light to change the focus position of a part of the visual field, and the sensor 6 through the imaging optical system 20 simultaneously registers the reference mark 101 and the registration mark of different focus positions. The detection signal 30c with the pattern 102 is detected.
In the image processing 21 of the computer 12, the reference mark 101 and the resist pattern 10 are based on the detection signal 30.
The shift amount δ in the X and Y directions from 2 is calculated (measured) based on the equations (4) and (5). After that, it moves to the next measurement target to be measured. This is repeated, and after the measurement of all the measurement objects is completed, the sample is removed and the measurement is completed.

FIG. 14 shows the shape of one glass plate 17 in the optical path length changing portion 19. The glass plate 17 is a glass plate having different thicknesses in the central portion and the peripheral portion. This is provided near the image forming position or at a position where the detection signal waveforms of the reference pattern 101 and the resist pattern 102 can be sufficiently separated. As a result, the optical path length changes between the central part and the peripheral part.

The waveform of the detection signal 30c detected by the sensor 6 is shown in FIG. FIG. 15 is a waveform in the X direction of the central portion of the measurement target 100 of FIG. 1 detected by the sensor 6. Both the reference mark 101 and the resist pattern 102 are in focus. The image processing 21 of the computer 12 is
From this waveform, the center position of each mark is calculated by the above-mentioned equation (4). At this time, the center position Ca of the reference mark 101 uses the data of the reference mark center position calculation range 151 calculated and set in the computer 12, and the center position Cb of the resist pattern 102 is the center of the resist pattern calculated and set in the computer 12. Calculation is performed using the data of the position calculation range 152. The reference mark center position calculation range 151 and the resist pattern center position calculation range 152 are defined by x in the formula (4).
Indicates the range of values taken by. The calculated center position of the reference mark 101 is Ca, and the center position Cb of the resist pattern 102 is Cb.
And Then, the computer 12 subtracts the numerical offset value β from the mark position difference (Ca−Cb) based on the equation (5), corrects the magnification correction value (difference in magnification) Δα, and sets the reference value. A shift amount δ in the X direction between the mark 101 and the resist pattern 102 is calculated. The Y direction is calculated by the same method.

Also in this embodiment, the first
And the second modified embodiment can be applied. Next, a modification of this embodiment will be described. A birefringent element may be used as the optical path length changing unit 19 instead of the plurality of glass plates 17. As the birefringent element, there are variations such as a pair wedge type shown in FIG. 16 or a lens made of the birefringent element. In the paired wedge, P-polarized light and S-polarized light follow different optical paths inside the birefringent element, and the focal length can be changed. In this case, in order to separate the images at the two focal positions, a polarizing plate is inserted in the illumination optical system 2, the objective lens system 3, the optical path length changing unit 19 or the imaging optical system 20, and for example, the reference mark portion is P polarized light. The resist pattern portion is designed to pass only S-polarized light. In addition, the illumination optical system 2
When a polarizing plate is inserted in the plate, it is necessary to provide P and S polarized light in parallel or perpendicular to the pattern and minimize the rotation of the polarization plane at the step portion of the sample 13. According to this modification, the optical path length changing means for the reference pattern portion and the resist pattern portion and the pattern separating means can be made independent, and the degree of freedom can be increased.

According to this embodiment, the reference mark 1 is provided for each product type.
01 and the resist pattern 102 can be adjusted to adjust the thickness of the wedge 16 for adjusting the optical path length and the focus offset of the auto-focus system 7 which is an air micro, so that the pattern having various steps can be supported. Become. Further, since the waveform is detected by one sensor and one optical path, stable and highly accurate measurement can be performed.

[0040]

According to the present invention, it is possible to measure a sample having a plurality of patterns having a high step difference with a high accuracy by a simple structure. Further, according to the present invention, a reference pattern of a base having a high step due to the miniaturization of a pattern of an LSI wafer or the like and a measurement target serving as a mask for etching a pattern of a next layer formed thereon It is possible to measure the deviation accuracy of the resist pattern with a high accuracy of 10 nm or less and to monitor the fluctuation of the process state.

Further, according to the present invention, a multi-layer resist is used, and it becomes a mask for etching the reference pattern of the base and the pattern of the next layer formed thereon even if the step height is further increased. It is possible to measure the deviation accuracy of the resist pattern to be measured with high accuracy of 10 nm or less and with reproducibility, and it is possible to monitor the fluctuation of the process state corresponding to the multilayer resist.

[Brief description of drawings]

FIG. 1 shows an embodiment of a measurement target according to the present invention,
(A) is a plan view and (b) is a sectional view.

FIG. 2 is a diagram showing a first basic configuration of a pattern measuring apparatus according to the present invention.

FIG. 3 is a diagram showing a second basic configuration of the pattern measuring apparatus according to the present invention.

FIG. 4 is a diagram showing an embodiment corresponding to the first basic configuration of the pattern measuring apparatus according to the present invention.

FIG. 5 is a diagram showing a detection signal waveform detected by a sensor for positioning a measurement target on a sample.

6 is a diagram showing a detection signal waveform focused on a reference mark in the embodiment shown in FIG.

7 is a diagram showing a detection signal waveform focused on a resist pattern in the embodiment shown in FIG.

8A and 8B show another embodiment of the measuring object according to the present invention, in which FIG. 8A is a plan view and FIG. 8B is a sectional view.

9A and 9B show another embodiment of the measuring object according to the present invention, where FIG. 9A is a plan view and FIG. 9B is a sectional view.

10A and 10B show another embodiment of the measuring object according to the present invention, in which FIG. 10A is a plan view and FIG. 10B is a sectional view.

FIG. 11 is a diagram showing an embodiment corresponding to the second basic configuration of the pattern measuring apparatus according to the present invention.

12 is a diagram showing a detection signal waveform focused on a reference mark and a resist pattern in the embodiment shown in FIG.

FIG. 13 is a diagram showing another embodiment of the pattern measuring apparatus according to the present invention.

14 is a perspective view showing a glass plate as an optical path length changing unit used in the apparatus shown in FIG.

15 is a diagram showing a detection signal waveform focused on a reference mark and a resist pattern in the embodiment shown in FIG.

16 is a diagram showing a birefringent element as an optical path length changing unit used in the device shown in FIG.

[Explanation of symbols]

2 ... Illumination optical system, 3 ... Objective lens system, 4 ... Branch optical system,
5, 5a, 5b ... Relay optical system, 6, 6a, 6b ... Sensor 7 ... Automatic focusing system, 8 ... Detection optical system, 9 ... Moving stage,
10 ... Sample stand 11 ... Stage system, 12 ... Control system (computer), 1
3 ... Samples 14a, 14b ... Optical path, 15 ... Combined optical system, 19 ... Optical path length changing unit 21 ... Image processing, 23 ... Objective lens, 41 ... Output means,
42 ... Input means 43 ... Display, 44 ... External storage device 61, 62, 121, 122, 151, 152 ... Calculation range 100 ... Measurement object 101 ... Reference mark 102 ... Resist pattern

Claims (12)

[Claims] 1. A first optical path and a second optical path in a detection optical system according to the height difference for a sample having a first pattern and a second pattern having a height difference in a predetermined region. Finely adjusting the optical path length difference between, the region on the sample is positioned in the field of view of the detection optical system, the surface of the sample is automatically focused with respect to the detection optical system, the region on the sample And illuminate each of the light images from the first and second patterns to form an image on the image sensor by each of the first and second optical paths of the detection optical system to detect an image signal from the image sensor. The image signal of the first pattern and the image signal of the second pattern are extracted by setting a range indicating the image signal corresponding to the first and second patterns with respect to the detected image signal. , Of each of the extracted first patterns The center position of the first pattern is calculated from the image signal, the center position of the second pattern is calculated from the extracted image signal of the second pattern, and the calculated center position of the first pattern and the second pattern are calculated. A method for measuring the size of a pattern having a height difference on a sample, wherein a size between the center of the pattern and a magnification of the detection optical system and offset correction are applied to calculate an accurate size. 2. A first optical path and a second optical path in a detection optical system according to the height difference for a sample having a first pattern and a second pattern having a height difference in a predetermined area. Finely adjusting the optical path length difference between, the region on the sample is positioned in the field of view of the detection optical system, the surface of the sample is automatically focused with respect to the detection optical system, the region on the sample And illuminate each of the light images from the first and second patterns to form an image on the image sensor by each of the first and second optical paths of the detection optical system to detect an image signal from the image sensor. The image signal of the first pattern and the image signal of the second pattern are extracted by setting a range indicating the image signal corresponding to the first and second patterns with respect to the detected image signal. , Of each of the extracted first patterns The image signal is converted into a digital image signal of the first pattern, the center position of the first pattern is calculated by symmetric pattern matching processing, and the digital image of the second pattern is calculated from the extracted image signal of the second pattern. The position of the center of the second pattern is calculated by converting the signal into a signal by symmetric pattern matching processing, and the detection optics is measured with respect to the dimension between the calculated center of the first pattern and the center of the second pattern. A method for measuring the size of a pattern having a height difference on a sample, which comprises performing a system magnification and offset correction to calculate an accurate size. 3. An optical path between a first optical path and a second optical path in a detection optical system for a sample having a reference mark and a resist pattern having a height difference in a predetermined area according to the height difference. Fine-tune the length difference, position the area on the sample within the field of view of the detection optics, autofocus the surface of the sample with respect to the detection optics, and illuminate the area on the sample. Each of the light images from the reference mark and the resist pattern is formed on an image sensor by each of the first and second optical paths of the detection optical system, and an image signal is detected from the image sensor, and the detected image signal is detected. An image signal of the reference mark and an image signal of the resist pattern are extracted by setting a range indicating an image signal corresponding to the reference mark and the resist pattern with respect to the image signal, and the extracted reference signals are extracted. The center position of the reference mark is calculated from the image signal of the reference mark, the center position of the resist pattern is calculated from the extracted image signal of the resist pattern, and the center position of the calculated reference mark and the center of the resist pattern are calculated. A dimension measuring method of a pattern having a height difference on a sample, which is characterized by performing a magnification of the detection optical system and an offset correction on the dimension to calculate an accurate dimension. 4. A sample stage on which a sample having a first pattern and a second pattern having a height difference in a predetermined area is placed, an illumination optical system for illuminating an area on the sample, and an optical path to each other. Forming a first optical path and a second optical path having a length difference, and between the first optical path and the second optical path according to a height difference between the first pattern and the second pattern. A fine control optical system for finely controlling the optical path length difference between the first and second patterns, and forming a light image on the image sensor by each of the first and second optical paths. A detection optical system that detects an image signal from an image sensor, a positioning unit that moves the sample stage to position an area on the sample within a field of view of the detection optical system, and a position of the sample with respect to the detection optical system. Autofocus hand to autofocus the surface A step, and by setting a range indicating an image signal corresponding to the first and second patterns with respect to the image signal detected by the image sensor of the detection optical system, the image signal of the first pattern and the second signal of the second pattern are set. The image signal of the pattern is extracted, the center position of the first pattern is calculated from the image signal of the extracted first pattern, and the center of the second pattern is extracted from the image signal of the extracted second pattern. Calculation for calculating a position and performing accurate magnification and offset correction of the detection optical system with respect to the calculated dimension between the center of the first pattern and the center of the second pattern An apparatus for measuring a dimension of a pattern having a height difference on a sample, comprising: 5. The dimension measurement of a pattern having a height difference on a sample according to claim 4, wherein a relay optical system is provided in each of the first optical path and the second optical path of the detection optical system. apparatus. 6. The height difference on the sample according to claim 4, wherein each of the first optical path and the second optical path of the detection optical system is provided with a relay optical system having different optical path lengths. A pattern measuring device. 7. A relay optical system is provided in each of the first optical path and the second optical path of the detection optical system, and a plurality of relay optical systems are provided as the image sensors corresponding to the optical paths. An apparatus for measuring dimensions of a pattern having a height difference on the sample described. 8. A fine control optical system in the detection optical system,
The dimension measuring apparatus for a pattern having a height difference on a sample according to claim 4, wherein the dimension measuring apparatus is provided in either the first optical path or the second optical path. 9. A fine control optical system in the detection optical system,
The dimension measuring device for a pattern having a height difference on a sample according to claim 4, wherein the dimension measuring device comprises a birefringent element. 10. A sample stage on which a sample having a first pattern and a second pattern having a height difference in a predetermined area is placed, and the sample stage is moved to detect an area on the sample. Positioning means for positioning within the field of view of the system, automatic focusing means for automatically focusing the surface of the sample with respect to the detection optical system, and illumination optical system for illuminating a region on the sample, and optical path length differences from each other. To form a first optical path and a second optical path having a relay optical system, and the first optical path and the second optical path are formed according to a height difference between the first pattern and the second pattern. A birefringent optical element for finely controlling a difference in optical path length between the optical path and the second optical path, and each of the optical images from the first and second patterns is divided into first and second optical paths by the first and second optical paths. Each image is formed on each of the second image sensors. A detection optical system that detects an image signal from an image sensor, a positioning unit that moves the sample stage to position an area on the sample within a field of view of the detection optical system, and a position of the sample with respect to the detection optical system. Automatic focusing means for automatically focusing the surface, and by setting a range indicating image signals corresponding to the first and second patterns with respect to the image signals detected by the image sensors of the detection optical system. The image signal of the first pattern and the image signal of the second pattern are extracted, the center position of the first pattern is calculated from the image signal of the extracted first pattern, and the extracted second image signal is extracted. The center position of the second pattern is calculated from the image signal of the pattern, and the position of the detection optical system with respect to the calculated dimension between the center of the first pattern and the center of the second pattern is calculated. Incidence and offset correction and size measuring device of a pattern difference in height on the sample, characterized in that a calculating means for calculating precise dimensions subjected to. 11. A sample stage on which a sample having a first pattern and a second pattern having a height difference in a predetermined region is placed, an illumination optical system for illuminating a region on the sample, and an optical path to each other. Forming a first optical path and a second optical path having a length difference, and between the first optical path and the second optical path according to a height difference between the first pattern and the second pattern. A fine control optical system for finely controlling the optical path length difference between the first and second patterns, and forming a light image on the image sensor by each of the first and second optical paths. A detection optical system that detects an image signal from an image sensor, a positioning unit that moves the sample stage to position an area on the sample within a field of view of the detection optical system, and a position of the sample with respect to the detection optical system. Autofocus to surface autofocus The image signal of the first pattern and the image signal of the second pattern by setting a range indicating an image signal corresponding to the first and second patterns with respect to the image signal detected by the image sensor of the detection optical system. The image signal of the pattern is extracted, the extracted image signal of the first pattern is converted into a digital image signal of the first pattern, and the center position of the first pattern is calculated by symmetric pattern matching processing. , The extracted image signal of the second pattern is converted into a digital image signal of the second pattern, the center position of the second pattern is calculated by symmetric pattern matching processing, and the calculated center position of the first pattern is calculated. And a calculating means for calculating an accurate dimension by performing magnification and offset correction of the detection optical system with respect to a dimension between the center and the center of the second pattern. An apparatus for measuring a dimension of a pattern having a height difference on a sample, characterized by: 12. A sample stage on which a sample having a first pattern and a second pattern having a height difference in a predetermined region is mounted, and the sample stage is moved to detect a region on the sample. Positioning means for positioning within the field of view of the system, automatic focusing means for automatically focusing the surface of the sample with respect to the detection optical system, and illumination optical system for illuminating a region on the sample, and optical path length differences from each other. To form a first optical path and a second optical path having a relay optical system, and the first optical path and the second optical path are formed according to a height difference between the first pattern and the second pattern. A birefringent optical element for finely controlling a difference in optical path length between the optical path and the second optical path, and each of the optical images from the first and second patterns is divided into first and second optical paths by the first and second optical paths. Each image is formed on each of the second image sensors. A detection optical system that detects an image signal from an image sensor, a positioning unit that moves the sample stage to position an area on the sample within a field of view of the detection optical system, and a position of the sample with respect to the detection optical system. Automatic focusing means for automatically focusing the surface, and by setting a range indicating image signals corresponding to the first and second patterns with respect to the image signals detected by the image sensors of the detection optical system. An image signal of the first pattern and an image signal of the second pattern are extracted, and the extracted image signals of the first pattern are converted into digital image signals of the first pattern to perform symmetric pattern matching processing. Calculate the center position of the first pattern by
Converting the extracted image signal of the second pattern into a digital image signal of the second pattern and calculating the center position of the second pattern by symmetric pattern matching processing,
And a calculating means for calculating a correct dimension by applying magnification and offset correction of the detection optical system to the calculated dimension between the center of the first pattern and the center of the second pattern. An apparatus for measuring a dimension of a pattern having a height difference on a sample, which is characterized in that