JP2008241506A - Surface shape measurement apparatus and surface shape measurement method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a surface shape measurement apparatus and a surface shape measurement method for accurately measuring a surface shape. <P>SOLUTION: The surface shape measurement apparatus of the invention comprises a sensor head 12 provided with a light source 21, and an optical sensor 25 for receiving a reflection light reflected by a sample 11; an X-drive mechanism 17 for changing a relative position between the sensor head 12 and the sample 11; a position extracting section 31 for extracting the incident position of the reflection light on the optical sensor 25; a transform section 32 for transforming position data for mapping the relative position to the incident position of the reflection light into frequency domain data; a filter section 33 for filtering a component of the frequency domain data at a fixed frequency or higher, or in a predetermined frequency band; and an inverse transform section 34 for inverse-transforming the filtered frequency domain data into spatial domain data. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a surface shape measuring device and a surface shape measuring method, and more particularly to a surface shape measuring device and a surface shape measuring method for measuring the shape of the surface of a sample.

  In the semiconductor manufacturing process, defects such as conveyance due to wafer deformation may occur. In order to include this defect, it is desired to evaluate the warpage (deflection) of the wafer. In order to evaluate the warpage of the wafer, the surface shape of the wafer may be measured. In order to measure the surface shape of the wafer, a technique for measuring the length of the reflected light reflected from the wafer surface according to the incident position on the optical sensor is disclosed (Patent Documents 1 to 3).

  For example, in Patent Document 2, light is incident on the sample from an oblique direction. Therefore, the incident position of the reflected light on the position detection sensor changes according to the distance between the sensor and the sample surface. Therefore, the distance from the sample can be measured by obtaining the incident position of the reflected light on the position detection sensor.

Further, a method for evaluating the wafer by measuring the shape of the front and back surfaces of the wafer is disclosed (Patent Document 4). In this evaluation method, the shape of the front surface and the back surface is measured, and the profile is subjected to frequency analysis. Specifically, amplitude or power spectral density is obtained by frequency analysis, and quantitative evaluation is performed using this as a parameter. Furthermore, a defect inspection apparatus is disclosed in which data obtained by measuring fluctuations in the height direction is subjected to Fourier transform using an autofocus function, and a low-order spectral component is extracted as a warp amount (Patent Document 5).
JP 59-164910 A JP 63-7626 A JP-A 63-85311 JP-A-9-311108 JP-A-5-249655

  In the above technique, for example, the length is measured based on the peak position and the center position of the profile of the reflected light. However, when measuring length according to the light receiving position on the sensor, there are the following problems. Various patterns are formed on the semiconductor wafer. Accordingly, the reflectance of light varies depending on the presence or absence of a pattern or the type of pattern. For this reason, for example, when light enters the boundary portion of the pattern, the reflected light profile at the optical sensor changes.

  That is, the amount of reflected light increases at a portion where the reflectance is high, and the amount of reflected light decreases at a portion where the reflectance is low. Therefore, the peak position and center position of the reflected light profile change according to the difference in the reflectance on the sample surface. Specifically, the peak position and the center position are shifted to a portion with high reflectivity. For this reason, the incident position of the reflected light on the optical sensor is shifted and determined. Therefore, the above technique has a problem that it is difficult to accurately measure the surface shape. Similar problems occur not only in semiconductor wafers but also in photomasks used in semiconductor manufacturing processes.

  The present invention has been made to solve such problems, and an object thereof is to provide a surface shape measuring device and a surface shape measuring method capable of accurately measuring the surface shape of a sample. .

  A surface shape measuring apparatus according to a first aspect of the present invention is a surface shape measuring apparatus for measuring a surface shape of a sample, and includes a light source that irradiates light to the sample, and the sample out of light from the light source. A sensor head provided with an optical sensor for receiving reflected reflected light, a drive mechanism for changing a relative position between the sensor head and the sample, and an incident position of the reflected light on the optical sensor are extracted. A position extraction unit, a conversion unit that converts position data in which the relative position is associated with the incident position of the reflected light into frequency domain data, and a component having a frequency equal to or higher than a certain frequency with respect to the frequency domain data A filter unit configured to perform filtering so as to cut off a component of the predetermined frequency band; and an inverse conversion unit configured to perform inverse conversion on the filtered frequency domain data into spatial domain data. Than is. Thereby, it can measure correctly.

  The surface shape measuring apparatus according to the second aspect of the present invention is the above-described measuring apparatus, wherein the frequency of the component filtered by the filter unit is set based on the period of the pattern provided in the sample. It is what. As a result, the influence on the pattern boundary can be reduced, and accurate measurement can be performed.

  A surface shape measuring apparatus according to a third aspect of the present invention is the above-described measuring apparatus, wherein the position extracting unit extracts the peak position of the received light intensity on the optical sensor as the incident position, When a plurality of peak positions exist with respect to the detection result at the relative position, the peak position closest to the peak position at the adjacent relative position is extracted as the incident position. Thereby, it can measure more correctly.

  A surface shape measurement apparatus according to a fourth aspect of the present invention is the above-described measurement apparatus, wherein an incident position of the reflected light on the optical sensor is determined according to a distance between the sample surface and the sensor head. The sensor head is provided with a slit for restricting the passage of the reflected light in a changing direction. Thereby, since stray light can be removed, it can measure correctly.

The surface shape measurement apparatus according to the fifth aspect of the present invention is the above-described measurement apparatus, further comprising a stage that supports the sample at an end portion of the sample, and with respect to the back surface and the surface of the sample, Based on the calculated surface shape, a deformation amount of the sample including a warp amount due to the weight of the sample is obtained. Thereby, the deformation amount of the sample including the warp amount due to its own weight can be accurately measured.
A surface shape measuring apparatus according to a sixth aspect of the present invention is the above-described measuring apparatus, which obtains the amount of warpage due to the weight of the sample and obtains the amount of deformation of the sample excluding the amount of warpage due to the weight. . Thereby, the deformation amount of the sample excluding the warp amount due to its own weight can be obtained accurately.

  A surface shape measuring method according to a seventh aspect of the present invention is a sensor head provided with a light source that irradiates a sample with light and an optical sensor that receives reflected light reflected by the sample out of light from the light source. And a surface shape measuring method by a surface shape measuring device comprising a drive mechanism for changing a relative position between the sensor head and the sample, the step changing the relative position between the sensor head and the sample; Extracting the incident position of the reflected light on the optical sensor; converting position data in which the relative position is associated with the incident position of the reflected light into frequency domain data; and Filtering the data having a predetermined frequency or higher or the predetermined frequency band so as to block the filtered frequency region, The data are those comprising the steps of: inverse transforming the data in the spatial domain. Thereby, it can measure correctly.

  A surface shape measuring method according to an eighth aspect of the present invention is the above-described measuring method, wherein the frequency of the filtered component is set based on a period of a pattern provided in the sample. is there. As a result, the influence on the pattern boundary can be reduced, and accurate measurement can be performed.

  A surface shape measurement method according to a ninth aspect of the present invention is the above-described measurement method, wherein the position extraction unit extracts the peak position of the received light intensity on the optical sensor as the incident position, and When a plurality of peak positions exist with respect to the detection result at the relative position, the peak position closest to the peak position at the adjacent relative position is extracted as the incident position. Thereby, it can measure more correctly.

  A surface shape measurement method according to a tenth aspect of the present invention is the above-described measurement method, wherein an incident position of the reflected light on the photosensor is determined according to a distance between the sample surface and the sensor head. The sensor head is provided with a slit for restricting the passage of the reflected light in a changing direction. Thereby, the amount of warpage due to its own weight can be accurately measured.

A surface shape measurement method according to an eleventh aspect of the present invention is the above-described measurement method, wherein the sample is placed on a stage that supports the sample at an end portion of the sample, and the back surface and the surface of the sample On the other hand, the amount of warpage due to the weight of the sample is obtained based on the calculated surface shape. Thereby, since stray light can be removed, it can measure correctly.
A surface shape measuring method according to a twelfth aspect of the present invention is the above-described measuring method, wherein the amount of warpage due to the weight of the sample is obtained, and the amount of deformation of the sample excluding the amount of warpage due to the own weight is obtained. . Thereby, the deformation amount of the sample excluding the warp amount due to its own weight can be obtained accurately.

  ADVANTAGE OF THE INVENTION According to this invention, the surface shape measuring apparatus and surface shape measuring method which can measure the surface shape of a sample correctly can be provided.

  In the present embodiment, the surface shape of the sample is measured by irradiating the sample with light and detecting the reflected light. Thereby, the curvature (deflection) of a sample can be evaluated. Typical samples include semiconductor wafers and photomasks.

  The structure of the surface shape measuring apparatus according to this embodiment will be described with reference to FIG. Fig.1 (a) is a top view which shows the structure of a surface shape measuring apparatus, FIG.1 (b) is the side view. As shown in FIG. 1, the surface shape measuring device 10 includes a sensor head 12, a stage 13, a processing device 15, a Y drive mechanism 16, and an X drive mechanism 17. The surface shape measuring apparatus 10 measures the surface shape of the sample 11. In FIG. 1, the vertical direction is the Z direction, and the direction perpendicular to the Z direction is the X direction. Furthermore, the Z direction and the direction perpendicular to the X direction are defined as the Y direction. Therefore, the XY plane is a horizontal plane. The sample 11 is disposed substantially parallel to the XY plane.

  The sample 11 is placed on the stage 13. A predetermined pattern film 11 c is formed on the lower surface of the sample 11. That is, the sample 11 includes a region where the pattern film 11c is formed and a region where the pattern film 11c is not formed. The stage 13 supports the sample 11 at three points. That is, the stage 13 contacts the lower surface of the sample 11 at three points at the end of the sample 11. Accordingly, the lower side of the entire sample 11 is open. That is, the lower side of the sample 11 is a space except for three points where the stage 13 is provided. As the stage 13, for example, a hollow stage provided with three support columns can be used. The stage 13 is connected to the Y drive mechanism 16. The Y drive mechanism 16 moves the stage 13 in the Y direction. Accordingly, the sample 11 on the stage 13 is scanned in the Y direction by the Y drive mechanism 16.

  The sensor head 12 is disposed below the sample 11. That is, the sensor head 12 is arranged on the pattern surface side of the sample 11. Based on the detection by the sensor head 12, the surface shape of the sample 11 is measured. The sensor head 12 outputs a detection signal corresponding to the detection result to the processing device 15. The sensor head 12 performs detection by applying the principle of triangulation. The configuration of the sensor head 12 will be described later. The sensor head 12 is connected to the X drive mechanism 17. The X drive mechanism 17 moves the sensor head 12 in the X direction. Therefore, the position of the sensor head 12 with respect to the sample 11 on the stage 13 changes in the X direction. Thus, the sensor head 12 is scanned in the X direction by the X drive mechanism 17. For example, the X drive mechanism 17 reciprocates the sensor head 12 at a constant speed. That is, the sensor head 12 moves in the X direction at a constant speed except in the vicinity of both ends of the scanning region.

  The X drive mechanism 17 and the Y drive mechanism 16 include an actuator such as a motor, a linear guide, and the like. Thereby, it can move with sufficient straightness. The X drive mechanism 17 and the Y drive mechanism 16 are connected to the processing device 15. The processing device 15 controls driving of the X drive mechanism 17 and the Y drive mechanism 16. That is, the X drive mechanism 17 and the Y drive mechanism 16 are driven by the drive signal from the processing device 15.

  The sensor head 12 moves at a higher speed than the stage 13. Then, the stage 13 is sent out in the Y direction while moving the sensor head 12 in the X direction. Thus, measurement can be performed on the entire sample surface. That is, the entire sample 11 can be measured by raster scanning or zigzag scanning of the measurement position. Specifically, the X drive mechanism 17 moves the sensor head 12 in the + X direction from one end of the sample 11 to the other end. Thereby, the measurement for one line of the sample 11 is performed. When the movement for one line in the X direction is completed, the X drive mechanism 17 reverses the movement direction. Further, while the moving direction is reversed, the Y drive mechanism 16 sends the stage in the Y direction by a predetermined interval. Then, the second line is measured. Therefore, in the measurement of the second line, the sensor head 12 moves in the reverse direction (−X direction). The sensor head 12 moves at substantially the same speed except when the sensor head 12 is reversed. That is, the sensor head 12 moves at a constant speed while measuring one line.

  A detection signal from the sensor head 12 is input to the processing device 15. The processing device 15 stores a detection signal from the sensor head 12. In addition, the Y drive mechanism 16 outputs a position signal indicating the position of the stage 13 in the Y direction to the processing device 15. Further, the X drive mechanism 17 outputs a position signal indicating the position of the sensor head 12 in the X direction to the processing device 15. Thereby, the coordinate of the detection position which the sensor head 12 is detecting in the sample 11 can be specified. The processing device 15 stores the coordinates of the detection position in association with the detection signal from the sensor head 12. Thereby, the detection signal at each coordinate of the sample 11 is stored.

  The processing device 15 is an information processing device such as a personal computer, and performs predetermined arithmetic processing on the detection signal. That is, the processing device 15 is a computer having a storage area such as a CPU or a memory. For example, the processing device 15 includes a CPU (Central Processing Unit) that is an arithmetic processing unit, a ROM (Read Only Memory) and RAM (Random Access Memory) that are storage areas, a communication interface, and the like, and measures an asserted shape. Execute the processing necessary to For example, the ROM stores an arithmetic processing program for performing arithmetic processing, various setting data, and the like. Then, the CPU reads out the arithmetic processing program stored in the ROM and develops it in the RAM. Then, the program is executed in accordance with the setting data and the output from the sensor head or the like. Further, the processing device 15 has a monitor or the like for displaying the calculation processing result. The processing by the processing device 15 will be described later.

  Next, the configuration of the sensor head 12 will be described with reference to FIG. FIG. 2 is a diagram showing the configuration of the optical system of the sensor head 12. The sensor head 12 is provided with a light source 21, a lens 22, a slit 23, a lens 24, and an optical sensor 25. In FIG. 2, the configuration shown in FIG. 1 is shown upside down. In FIG. 2, the surface of the sample 11 is shown as sample surfaces 11a and 11b. Note that the distance from the sensor head 12 varies between the sample surface 11a and the sample surface 11b. That is, the sample surface 11b is more distant from the sensor head 12 than the sample surface 11a.

  The light source 21 is a laser light source or the like, and irradiates the sample 11 with laser light. Therefore, the light from the light source 21 enters the surface 11 a of the sample 11. The surface 11a of the sample 11 is illuminated with light from the light source 21 at a predetermined spot. Further, the laser light from the light source 21 is incident on the sample 11 at an angle. That is, the laser light propagates in a direction inclined from a direction perpendicular to the surface 11a of the sample 11 and enters the sample 11 at a predetermined incident angle. Part of the light incident on the sample 11 from the light source 21 is reflected by the sample surface 11a. Of the light incident on the sample 11 from the light source 21, the reflected light reflected on the sample surface 11 a enters the lens 22. The lens 22 refracts the reflected light and collects it. Thereby, the reflected light becomes a parallel light flux.

  The reflected light refracted by the lens 22 enters the slit 23. The slit 23 is disposed at the pupil position of the lens 22. The slit 23 has a predetermined width and shields reflected light incident on the outside thereof. That is, the reflected light that enters the line-shaped opening provided in the slit 23 passes through the slit 23, and the reflected light that enters the outside of the opening is shielded. Thereby, stray light in the optical system can be shielded. Therefore, an error due to diffracted light from the patterned sample 11 can be reduced, and accurate detection becomes possible. The center of the slit 23 is arranged to be the center of the reflected light. The reflected light that has passed through the slit 23 enters the lens 24. The lens 24 refracts the reflected light and collects it on the light receiving surface of the optical sensor 25.

  The optical sensor 25 is, for example, a line sensor. Specifically, a one-dimensional CCD can be used as the optical sensor 25. Therefore, the light receiving pixels are arranged in one row on the light receiving surface of the optical sensor 25. The optical sensor 25 outputs a detection signal corresponding to the reflected light received by each light receiving pixel to the processing device 15. That is, the profile of the received light intensity on the optical sensor is output as a detection signal. Here, when the distance between the sensor head 12 and the surface of the sample 11 changes, the light receiving position of the optical sensor 25 changes. The light receiving pixels of the optical sensor 25 are arranged in a direction in which the light receiving position is shifted depending on the distance between the sample surface and the sensor head 12. For example, the reflected light from the sample surface 11 a is incident on the pixel at the substantially center of the light receiving surface of the optical sensor 25, but the reflected light from the sample surface 11 b is incident on a pixel that is shifted from the center of the light receiving surface of the optical sensor 25. To do. Here, the reflected light from the sample 11b is incident on a pixel shifted to the left from the center pixel. As described above, since the light from the light source 11 is obliquely applied to the sample 11, the incident position of the light on the surface of the sample 11 changes according to the height of the surface of the sample 11. Therefore, the position where the light is reflected on the surface of the sample 11 changes. When the distance between the optical sensor 25 and the sample 11 changes, the light receiving position on the optical sensor 25 changes. Therefore, the distance between the optical sensor 25 and the surface of the sample 11 can be obtained based on the incident position of the reflected light on the light receiving surface of the optical sensor 25.

  Further, the slit 23 restricts the passage of the reflected light in the direction in which the incident position of the reflected light on the optical sensor 25 changes according to the distance between the sample surface and the sensor head 12. Specifically, light that is shifted to both sides from the center in the optical system is blocked, so that the reflected light passes through the center of the optical system. Therefore, the symmetry of the profile detected by the optical sensor 25 can be maintained. Therefore, the incident position on the optical sensor 25 can be accurately extracted.

  Then, the relative position between the sensor head 12 and the sample 11 is changed by the X drive mechanism 17 and the Y drive mechanism 16. Thereby, the position of the sensor head 12 with respect to the sample 11 changes in the horizontal direction, and the measurement position on the sample 11 changes. Then, the distance from the surface of the sample 11 to the sensor head 12 is measured while changing the relative position. The distances from the surface of the sample 11 to the sensor head 12 at a plurality of relative positions are joined together. By doing so, the surface shape of the sample 11 can be measured. That is, the amount of deviation of the incident position with reference to the light receiving pixel at the center of the optical sensor 25 corresponds to the distance between the sample 11 and the sensor head 12. The distance between the sample 11 and the optical sensor 25 is measured according to the amount of deviation of the incident position. Then, by measuring the distance from the optical sensor 25 to the surface of the sample 11 with respect to the entire sample 11, the unevenness, warpage, bending, undulation, etc. of the sample 11 can be obtained. Here, processing for measuring the surface shape by the detection signal from the sensor head 12 is executed in the processing device 15.

  Next, the structure of the processing apparatus 15 is demonstrated using FIGS. FIG. 3 is a block diagram showing a configuration of the processing device 15. 4 and 5 are profiles showing the detection results of the optical sensor 25 at arbitrary coordinates on the sample 11. 4 and 5, the horizontal axis indicates the position on the optical sensor 25, and the vertical axis indicates the amount of received light. FIG. 4 shows a profile at a location where the pattern film 11c is not present, and FIG. 5 shows a profile at the boundary of the pattern film 11c. FIG. 6 is a diagram for explaining the processing by the processing device 15. In the processing apparatus 15, in order to measure a surface shape correctly, the 1st surface shape measurement process and the 2nd surface shape measurement process are performed. First, the first surface shape measurement process will be described below.

  As illustrated in FIG. 3, the processing device 15 includes a position extraction unit 31, a conversion unit, a filter unit 33, an inverse conversion unit 34, a shape measurement unit 35, a warp amount calculation unit 36, a sensor head control unit 41, a storage unit 42, A stage control unit 43 is provided. The sensor head control unit 41 outputs a drive signal and controls the driving of the sensor head 12 by the X drive mechanism 17. The stage control unit 43 outputs a drive signal and controls the drive of the stage 13 by the Y drive mechanism 16. Thereby, the detection of the sample 11 at an arbitrary coordinate can be performed. The storage unit 42 stores the detection signal from the sensor head 12 in association with the coordinates on the sample 11.

  The position extraction unit 31 extracts the peak position of the detection signal profile based on the detection signal from the sensor head 12. That is, the position extraction unit 31 extracts the peak position of the profile as the incident position of the reflected light. For example, the position extraction unit 31 performs curve fitting with a quadratic curve on data in the vicinity of a light receiving pixel having a high light receiving intensity. Thereby, the peak position of the profile on the light receiving surface of the optical sensor 25 can be accurately extracted. Thus, the position extraction unit 31 extracts the peak position of the received light intensity profile on the optical sensor 25 as the incident position. Of course, the peak position extraction method is not limited to this.

  For example, a profile in a place where the pattern film 11c is not shown is as shown in FIG. In a place where the pattern film 11c is not present, the peak position of the profile in the optical sensor 25 is L0. However, at the boundary of the pattern film 11c, the reflectance of the sample surface varies depending on the presence or absence of the pattern film 11c. For example, the portion where the pattern film 11c exists has a high reflectance, and the portion where the pattern film 11c exists has a low reflectance. Therefore, the profile shown in FIG. 4 is broken at the boundary of the pattern film 11c. For example, as shown in FIG. 5, the profile is broad and the peak position is shifted to the light receiving pixel having the higher reflectance. Therefore, when the spot of the laser beam from the light source 11 straddles the end of the pattern film 11c, the peak position is shifted. In this case, an error from the height of the sample surface at the actual spot center of the laser beam occurs with respect to the detected peak position. That is, even if the height of the portion where the pattern film 11c is not present is the same, the peak position is largely shifted at the boundary of the pattern film 11c. When the entire spot of the laser beam is incident on the pattern film 11c, for example, the profile is the same as that in FIG. 4, and the peak height changes.

  Therefore, the peak position changes abruptly at the boundary of the pattern film 11c. For example, as shown in FIG. 6A, it is assumed that the pattern film 11c is formed on the sample 11 at a constant pitch. FIG. 6A is a plan view showing the configuration of the surface of the sample 11. In this case, when the sensor head 12 is moved in the direction of the arrow in FIG. 6A, laser light is incident on the boundary of the pattern film 11c at regular intervals. That is, the boundary of the pattern film 11c becomes a measurement position at a constant period.

  Therefore, the change in the peak position when the sensor head 12 is moved is as shown in FIG. FIG. 6B shows peak position data with respect to the measurement coordinates. That is, in FIG. 6B, the horizontal axis indicates the measurement position on the sample 11, and the vertical axis indicates the peak position extracted by the position extraction unit 31. Data indicating this change in peak position is defined as position data. Therefore, the position data is composed of incident positions (peak positions) of reflected light extracted at a plurality of measurement positions. In the position data, the relative position (measurement position) of the sensor head 12 with respect to the sample 11 is associated with the incident position. In FIG. 6, the horizontal axis represents the measurement position on the sample with the X coordinate. If the sensor head 12 is moving in the X direction at a constant speed, the horizontal axis can be taken as the time axis. Therefore, the position data shown in FIG. 6B is data in the space domain or the time domain.

  As shown in FIG. 6B, the peak position changes abruptly at the edge portion of the pattern film 11c. As described above, this is because the reflectance changes depending on the presence or absence of the pattern film 11c. Therefore, if the surface shape is measured based on the position data shown in FIG. 6B, the measurement error at the boundary of the pattern film 11c increases. That is, the peak position protrudes greatly at the boundary of the pattern film 11c. Therefore, in the present embodiment, the conversion unit 32 converts the position data in the spatial domain illustrated in FIG. 6B into frequency domain data.

  Here, the frequency domain data converted by the conversion unit 32 is shown in FIG. In FIG. 6C, the horizontal axis indicates the frequency, and the vertical axis indicates the signal intensity. The converting unit 32 performs, for example, discrete Fourier transform (DFT). Specifically, Fast Fourier Transform (FFT) is executed to convert the spatial domain data shown in FIG. 6B into frequency domain data. The Fourier transform is performed by Equation 1.

  If the horizontal axis in FIG. 6B is taken as a time axis, the conversion unit 32 converts time-domain position data into frequency-domain data. As described above, the conversion unit 32 converts the position data corresponding to the incident position of the reflected light into frequency domain data using Fourier transform. Here, a peak appears at fd in the frequency domain data. This fd corresponds to the time interval (spatial distance) at which the boundary of the pattern film 11c appears. That is, the frequency at which the boundary of the pattern film 11c is the measurement position is fd. Of course, depending on the arrangement of the pattern film 11c, peaks may appear at two or more locations.

The filter unit 33 filters components having a certain frequency (f ₀ ) or higher. As a result, the frequency domain data is as shown in FIG. Here, f ₀ is made to a frequency lower than the fd. Therefore, the filter unit 33 filters high frequency components including fd. For example, the filter unit 33 cuts the _{f 0} or more components. Therefore, the signal intensity becomes zero at a frequency greater than f _0. Thereby, the component of the high frequency band including fd is removed, and the influence due to the boundary of the pattern film 11c can be removed. In this way, the filter unit 33 functions as a low-pass filter that passes a low-frequency signal. Therefore, at frequencies lower than f _0, the signal strength does not change.

  Then, the inverse transformer 34 performs inverse transformation on the filtered frequency domain data. As a result, the data in the frequency domain is inversely transformed into data in the spatial domain (time domain). Specifically, inverse Fourier transform is performed to calculate spatial domain data. The data inversely transformed into the space domain is as shown in FIG. In this way, the influence of the pattern film 11c is reduced. Accordingly, as shown in FIG. 6E, the peak position smoothly changes even at the boundary of the pattern film 11c. That is, the data protruding at the boundary of the pattern film 11c is deleted. Therefore, the measurement error at the boundary of the pattern film 11c can be reduced.

  The influence of the pattern film 11c is a short period with respect to the warp of the wafer as the sample 11. For this reason, when performing inverse Fourier transform, inverse Fourier transform is performed using equation 3 instead of the commonly used equation 2.

In Equation 3, M is a value not longer than the signal period due to the influence of the pattern. Thereby, a high frequency component is cut. f ₀ is set based on the period of the pattern provided on the sample 11. Therefore, in other words, information on the warp of the sample 11 is included in the band of f ₀ or less. f ₀ to be to cut frequencies, it is possible to reduce the influence by the pattern of the short period in comparison with the warp of the wafer (deflection). By doing in this way, the process by the filter part 33 and the process by the inverse transformation part 34 can be performed at once. Therefore, the measurement error due to the influence of the pattern can be easily reduced. Of course, the cut frequency f ₀ may be variable according to the sample 11. f ₀ may be set according to the width of the pattern and the gap between the patterns.

  Then, the shape measuring unit 35 measures the surface shape based on the incident position in the inversely converted data. Specifically, the shape measuring unit 35 measures the length based on the incident position of the reflected light on the optical sensor 25. That is, the shape measuring unit 35 measures the distance between the sample 11 and the sensor head 12 based on the incident position in the inversely converted data. Based on the amount of deviation of the incident position, the height of the sample surface is calculated. The height of the sample surface corresponds to the amount of displacement of the surface of the sample 11 in the Z direction. Then, measurement is performed by changing the relative position of the sensor head 12 with respect to the sample 11, that is, the measurement position. The shape measuring unit 35 connects the displacement amounts in the Z direction of the surface of the sample 11 at the respective measurement positions. Thereby, the surface shape of the sample 11 can be measured based on the length measurement result. Moreover, the influence of the boundary part of the pattern film 11c can be reduced by cutting the high frequency component. For this reason, accurate surface shape measurement is possible. The shape measuring unit 35 may measure the length based on the position data before being converted by the converting unit 32. That is, the distance between the sample surface and the optical sensor 25 may be calculated before the Fourier transform is performed by the conversion unit 32. In this case, the displacement amount of the sample surface when the measurement position is changed becomes the position data. Then, position data corresponding to the amount of displacement at each measurement position is Fourier transformed to the frequency domain.

  Next, the calculation processing result will be described with reference to FIGS. FIG. 7 shows a measurement result before the influence of the boundary portion of the pattern film 11c is reduced. FIG. 8 shows the measurement result after reducing the influence of the boundary portion of the pattern film 11c. That is, FIG. 7 shows a waveform of position data before Fourier transform, and FIG. 8 shows a waveform of data after inverse Fourier transform. In other words, FIG. 7 shows the measurement result inversely converted by Equation 2, and FIG. 8 shows the measurement result inversely converted by Equation 3. 7 and 8, the horizontal axis indicates time, and the vertical axis indicates the amount of displacement of the sample surface based on the incident position. When the time on the horizontal axis is converted from the moving speed of the sensor head 12 in the X direction to the spatial distance on the sample 11, 25 msec is 1.56 mm.

  If the influence of the pattern film 11c is not removed, the measurement result at the edge portion of the pattern film 11c appears as a protruding noise as shown in FIG. On the other hand, when the influence of the pattern film 11c is removed, the measurement result at the edge portion of the pattern film 11c does not appear as noise as shown in FIG. Therefore, measurement errors due to the influence of the pattern can be reduced.

  Next, the second surface shape measurement process will be described with reference to FIGS. FIG. 9 is a profile showing a detection result of the optical sensor 25 at an arbitrary coordinate on the sample 11. In FIG. 9, the horizontal axis indicates the position on the optical sensor 25, and the vertical axis indicates the amount of received light. FIG. 10 is a diagram for explaining the processing by the processing device 15.

  When the pattern film 11c is formed on the sample 11, the profile as shown in FIG. 9 may be obtained depending on the pattern film 11c. That is, in the place where the pattern film 11c exists, a plurality of peaks may appear due to the interference of the laser light. In such cases, there are usually low peaks on either side of the high peak. That is, the position of the three peaks appearing in the profile, the peak at the center is the highest peak. As described above, a plurality of peak positions are calculated at a location where the pattern 11c exists.

  Assume that measurement is performed on a region where pattern films 11c as shown in FIG. 10A are provided at regular intervals. Then, the position data in this area is as shown in FIG. That is, there are three values of the peak position at the location where the pattern film 11c is formed. Accordingly, at the position where the pattern film 11c is formed, the peak position value is calculated above and below the peak position value corresponding to the central peak at a certain arbitrary measurement position. On the other hand, only one peak position is calculated at a location where there is no pattern film 11c. In FIG. 10B, the horizontal axis indicates the measurement position on the sample 11, and the vertical axis indicates the peak position extracted by the position extraction unit 31.

  Therefore, in this measurement process, the value of the peak position continuous from the adjacent measurement position is extracted as the true incident position. That is, for the three peak positions, the peak position having a value closest to the peak position at the adjacent measurement position is extracted as the incident position. Accordingly, only one peak position is extracted from the measurement position where only one peak position is calculated and the adjacent measurement position. Therefore, even if three peak positions are calculated at the measurement position where the pattern film 11c is not formed and the adjacent measurement position, only the peak position corresponding to the central peak can be extracted. Then, by shifting the measurement position in order and executing the above processing, only one peak position is extracted at each measurement position.

  That is, the positions other than the peak position corresponding to the central peak are not recognized as the incident positions. For this reason, one incident position is specified at each measurement position. Therefore, one peak position can be accurately extracted even at the measurement position where the pattern film 11c exists. Thus, when the value of the continuous peak position from the adjacent measurement position is extracted as the true peak position, the position data is as shown in FIG. In FIG. 10C, the horizontal axis indicates the measurement position on the sample 11, and the vertical axis indicates the peak position extracted by the position extraction unit 31.

  In this way, the peak position that is continuous with the peak position at the adjacent measurement position is selected in order. When a peak position is selected from a place where the pattern film 11c is not present, one incident position is specified even in a place where the pattern film 11c is present. Therefore, the incident position can be accurately extracted. For this reason, accurate surface shape measurement becomes possible.

  Of course, you may perform the process of a 1st surface shape measurement process and a 2nd surface shape measurement process. In this case, after the second surface shape measurement process is performed on the position data of the spatial region, the first surface shape measurement process is performed. As a result, it is possible to reduce the influence at the boundary portion of the pattern film 11c in a state where the waveform of the space region is continuous. Therefore, more accurate shape measurement is possible, and the warp (deflection) of the sample 11 can be accurately evaluated.

  Further, the warpage amount calculation unit 36 calculates the warpage amount (deflection amount) due to the weight of the sample 11. By performing the above surface shape measurement on both surfaces of the sample 11, the amount of warpage due to its own weight can be calculated. For example, after measuring the surface shape with respect to the front surface (pattern surface) of the sample 11, the sample 11 is inverted and placed on the stage 13. Then, the surface shape of the back surface of the sample 11 is measured. Thereby, the surface shape of both surfaces (front surface and back surface) of the sample 11 can be measured. Then, for each measurement position, the difference between the front surface measurement distance and the back surface measurement distance is taken and divided by two. Here, the difference in the measurement distance at the same measurement position is taken and divided by two. This data is the amount of warpage of the sample 11 due to its own weight. In addition, the order of the shape measurement on the front surface and the back surface may be reversed.

  Based on the amount of warpage calculated by the amount of warpage calculator 36, the surface shape in a state where no self-weight is applied can be estimated. Thereby, for example, the surface shape in the state where the sample 11 is arranged in the vertical direction is estimated. That is, the surface shape when a semiconductor wafer as a sample is placed vertically is estimated. On one side, the amount of warpage is added to the measured surface shape, and on the other side, the amount of warpage is subtracted from the measured surface shape. Then, a surface shape obtained by adding or subtracting the amount of warpage to the measured surface shape is estimated as a surface shape in a state where no weight is applied. The unevenness of the thickness of the semiconductor wafer is about several μm, and the amount of warpage due to its own weight is about 70 to 100 μm. Therefore, the change in the amount of warp due to its own weight is sufficiently larger than the thickness unevenness. Therefore, based on the amount of warpage due to its own weight, it is possible to estimate the surface shape in a state where no own weight is applied. Thus, the amount of warpage due to the weight of the sample 11 can be obtained from the measured value or the estimated value. Then, the deformation amount of the sample 11 excluding the warpage amount due to the determined weight can be obtained.

  In the above description, the optical sensor 25 is described as a line sensor, but it may be an area sensor in which light receiving pixels are arranged in a matrix or a PSD (Positive Sensitive Detector) sensor. In other words, any device that can detect the incident position of the reflected light may be used. Further, although the incident position of the reflected light has been described as the peak position of the profile, it may be the center of gravity position of the profile. Moreover, although the filter part 33 demonstrated that the component more than fixed frequency was filtered, it does not restrict to this. For example, the filter unit 33 may filter a component in a predetermined frequency band. That is, the filter unit 33 may block a component near a specific frequency. Also in this case, the frequency of the component to be filtered is preferably determined by the period of the pattern. In this case, the filter unit 33 functions as a bandpass filter.

  Note that the processing device 15 is not limited to a physically single device. Some of the position extraction unit 31, the conversion unit 32, the filter unit 33, the inverse conversion unit 34, the shape measurement unit 35, and the warp amount calculation unit 36 may be incorporated in different apparatuses. For example, the detection of the incident position by the position extraction unit 31 may be performed by the controller of the optical sensor 25. That is, the position extraction unit 31 may be provided in the optical sensor 25. In this case, the peak position at each measurement position is input to the processing device 15. And the processing apparatus 15 performs the process after the conversion part 32. FIG. Further, the second shape measurement process may be performed by the controller of the optical sensor 25.

  The surface shape can be accurately determined by the above surface shape measuring method. It is particularly suitable for measuring the surface shape of a patterned substrate. Therefore, it is possible to accurately evaluate warpage, deflection, waviness, etc. of a semiconductor silicon wafer or a photomask used in a semiconductor manufacturing process. Therefore, defective wafers and photomasks can be removed, and semiconductor productivity can be improved.

It is a figure which shows the structure of the surface shape measuring apparatus concerning this Embodiment. It is a figure which shows the structure of the sensor head used for the surface shape measuring apparatus concerning this Embodiment. It is a block diagram which shows the structure of the processing apparatus of the surface shape measuring apparatus concerning this Embodiment. It is a figure which shows the profile of the detection result in the location without a pattern film | membrane. It is a figure which shows the profile of the detection result in the sample in which a pattern film | membrane exists. It is a figure for demonstrating the 1st process with the processing apparatus of the surface shape measuring apparatus concerning this Embodiment. It is a figure which shows the position data before performing a 1st shape measurement process. It is a figure which shows the data after performing a 1st shape measurement process. It is a figure which shows the profile of another detection result in the sample in which a pattern film | membrane exists. It is a figure for demonstrating the 2nd process with the processing apparatus of the surface shape measuring apparatus concerning this Embodiment.

Explanation of symbols

11 Wafer, 11a Sample surface, 11b Sample surface, 11c Pattern film,
12 sensor heads, 13 stages, 15 processing equipment,
16 Y drive mechanism, 17 X drive mechanism,
21 light source, 22 lens, 23 slit, 24 lens, 25 optical sensor,
31 position extraction unit, 32 conversion unit, 33 filter unit, 34 inverse conversion unit,
35 shape measurement unit, 36 warpage amount calculation unit,
41 sensor head control unit, 42 storage unit, 43 stage control unit,

Claims (12)

A surface shape measuring device for measuring the surface shape of a sample,
A sensor head provided with a light source for irradiating the sample with light, and an optical sensor for receiving reflected light reflected by the sample among the light from the light source;
A drive mechanism for changing a relative position between the sensor head and the sample;
A position extraction unit for extracting an incident position of the reflected light on the optical sensor;
A conversion unit that converts position data in which the relative position is associated with the incident position of the reflected light into frequency domain data;
A filter unit for filtering the frequency domain data so as to block a component of a predetermined frequency or higher, or a component of a predetermined frequency band;
A surface shape measurement apparatus comprising: an inverse transform unit configured to inversely transform the filtered frequency domain data into spatial domain data.   The surface shape measuring apparatus according to claim 1, wherein a frequency of a component filtered by the filter unit is set based on a period of a pattern provided in the sample. The position extraction unit extracts the peak position of the received light intensity on the optical sensor as the incident position,
The peak position closest to the peak position at the adjacent relative position is extracted as the incident position when there are a plurality of peak positions with respect to the detection result at the arbitrary relative position. Surface shape measuring device.   The sensor head is provided with a slit that restricts the passage of the reflected light in a direction in which the incident position of the reflected light on the optical sensor changes according to the distance between the sample surface and the sensor head. The surface shape measuring device according to any one of claims 1 to 3. A stage for supporting the sample at an end of the sample;
5. The deformation amount of the sample including the amount of warpage due to the weight of the sample is obtained based on the calculated surface shape with respect to the back surface and the front surface of the sample. The surface shape measuring apparatus as described.   The surface shape measuring apparatus according to claim 5, wherein the amount of warpage due to the weight of the sample is obtained and the amount of deformation of the sample excluding the amount of warpage due to the own weight is obtained. A sensor head provided with a light source for irradiating the sample with light, and an optical sensor for receiving reflected light reflected by the sample among the light from the light source;
A surface shape measuring method using a surface shape measuring device comprising: a drive mechanism that changes a relative position between the sensor head and the sample;
Changing the relative position of the sensor head and the sample;
Extracting an incident position of the reflected light on the optical sensor;
Converting position data in which the relative position is associated with the incident position of the reflected light into frequency domain data;
Filtering the frequency domain data so as to cut off a component of a predetermined frequency or higher or a component of a predetermined frequency band;
Back-converting the filtered frequency domain data into spatial domain data.   The surface shape measurement method according to claim 7, wherein the frequency of the component to be filtered is set based on a period of a pattern provided in the sample. The position extraction unit extracts the peak position of the received light intensity on the optical sensor as the incident position,
The peak position closest to the peak position at the adjacent relative position is extracted as the incident position when there are a plurality of peak positions with respect to the detection result at the arbitrary relative position. Surface shape measurement method.   The sensor head is provided with a slit that restricts the passage of the reflected light in a direction in which the incident position of the reflected light on the optical sensor changes according to the distance between the sample surface and the sensor head. The surface shape measuring method according to claim 6, wherein: The sample is placed on a stage that supports the sample at an end of the sample,
The surface shape measuring method according to any one of claims 7 to 10, wherein the amount of warpage due to the weight of the sample is obtained based on the calculated surface shape with respect to the back surface and the surface of the sample.   The surface shape measuring method according to claim 11, wherein the amount of warpage due to the weight of the sample is obtained and the amount of deformation of the sample excluding the amount of warpage due to the weight of the sample is obtained.

Images