Shape measuring method and shape measuring instrument

Abstract

To provide a shape measuring method and a shape measuring instrument that are capable of generating combined data with stitching processing errors corrected, when stitching shape measuring data together.SOLUTION: A shape measuring method has: a first combined data generation step of generating first combined data by stitching a plurality of shape measurement data together, a second combined data generation step of generating second combined data by patching the plurality of shape measuring data together, and a correction step of correcting the first combined data based on differential data between the first combined data and the second combined data.SELECTED DRAWING: Figure 6

Description

[Technical field]

[0001] The present invention relates to a shape measuring method and a shape measuring instrument for measuring the shape of a surface to be measured in a non-contact manner. [Background technology]

[0002] As a chuck for holding a wafer, an electrostatic chuck is known that has an embossed wafer holding surface (surface to be measured) and holds a wafer on the wafer holding surface by electrical force (see Patent Document 1). In such an electrostatic chuck, the straightness of the upper surfaces of the many embossments formed on the wafer holding surface is required to meet a certain standard. For this reason, electrostatic chuck manufacturers or semiconductor manufacturing facilities measure the straightness of the upper surface of each embossment. Specifically, the shape of the wafer holding surface is measured using a measuring device capable of measuring the shapes of various surfaces to be measured, and the straightness is measured by extracting only the measurement results of the upper surfaces of the embossments from the shape measurement results.

[0003] For example, when the measuring device is a contact type such as the surface roughness measuring device described in Patent Document 2, the stylus of the measuring device is moved along the wafer holding surface of the electrostatic chuck while being in contact with the wafer holding surface. In this way, the stylus scans the wafer holding surface so as to cross the top surfaces of the multiple embossments, thereby obtaining straightness measurement results for the top surface of each embossment.

[0004] Furthermore, when the measuring device is a non-contact type such as a white light interference microscope or a laser confocal microscope, the measuring device has a limited measurement range (observation field of view). For this reason, the measuring device continuously measures the shape of the wafer-holding surface while moving one of the measuring device and the wafer-holding surface relative to the other in a direction perpendicular to the measurement optical axis of the measuring device, thereby obtaining multiple pieces of shape measurement data. Then, each piece of shape measurement data is stitched together to generate composite data of a length that allows evaluation of straightness (see Non-Patent Document 1). As a result, a straightness measurement result of the top surface of each embossment is obtained based on the composite data. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2018-14515 A [Patent Document 2] Patent No. 6458335 [Non-patent literature]

[0006] [Non-Patent Document 1] Development of stitching algorithm using approximate reference shape: Masato Negishi et al., Journal of the Japan Society for Precision Engineering, 81, 6(2015) 555. Summary of the Invention [Problem to be solved by the invention]

[0007] Incidentally, the embossments are minute protrusions with a diameter of 200 μm to 1 mm and a height of about 10 μm to 20 μm. Therefore, with the contact-type measuring device described in Patent Document 2, it is sometimes impossible to bring the stylus into sufficient contact with the top surface of the embossment, making it difficult to measure the straightness of the top surface of each embossment.

[0008] On the other hand, non-contact measuring devices can measure even minute shapes such as embossments without any problems. However, when stitching each shape measurement data set together using the stitching process described in Non-Patent Document 1, an error component (stitching process error) occurs at the stitched portion, so there is a risk that accurate straightness measurement results for the top surface of each embossment cannot be obtained from the combined data.

[0009] The present invention has been made in consideration of the above circumstances, and has an object to provide a shape measurement method and a shape measuring device that are capable of generating synthetic data in which stitching process errors have been corrected when shape measurement data is stitched together using stitching processing. [Means for solving the problem]

[0010] A shape measurement method for achieving the object of the present invention includes a data acquisition step of acquiring a plurality of shape measurement data continuously measured by a measuring device that measures the shape of the surface to be measured in a non-contact manner while moving the surface to be measured relative to the measuring device in a direction perpendicular to a measurement optical axis thereof, the plurality of shape measurement data having partially overlapping measurement ranges between successive shape measurement data, a first synthetic data generation step of stitching together the plurality of shape measurement data acquired in the data acquisition step by a stitching process to generate first synthetic data, a second synthetic data generation step of stitching together the plurality of shape measurement data acquired in the data acquisition step by a patching process to generate second synthetic data, a differential data generation step of generating difference data between the first synthetic data and the second synthetic data, a first error calculation step of calculating a first error that occurs when the shape measurement data are stitched together by the stitching process from the differential data, and a correction data calculation step of subtracting the first error from the first synthetic data to calculate correction data for the first synthetic data.

[0011] According to this shape measurement method, the first error that occurs when connecting shape measurement data by stitching processing can be obtained from the difference data between the first combined data generated by the stitching processing and the second combined data generated by the patching processing.

[0012] In a shape measurement method according to another aspect of the present invention, the first error calculation step includes a noise removal step of performing a noise removal process on the difference data, and a subtraction step of performing a subtraction process of subtracting a second error, which occurs when connecting shape measurement data by patching and is not removed by the noise removal process, from the difference data, thereby making it possible to obtain the first error.

[0013] In a shape measuring method according to another aspect of the present invention, the method includes a second error acquisition step of acquiring a second error in advance, and a first error calculation step of performing a subtraction process based on the second error acquired in the second error acquisition step, thereby determining the first error.

[0014] In a shape measuring method according to another aspect of the present invention, in the subtraction processing step, subtraction processing is performed on the difference data from which noise has been removed.

[0015] In a shape measuring method according to another aspect of the present invention, the noise removal process is a filtering process that extracts wavelength components having a length at least twice as long as the measurement range from the difference data.

[0016] In a shape measurement method according to another aspect of the present invention, in the data acquisition step, a stage on which the measured surface is placed is moved in a vertical direction while acquiring multiple shape measurement data measured continuously by the measuring device, and stage deflection occurs during the stage movement, and an observation point on the measured surface included in the measurement range is displaced in the optical axis direction of the measurement optical axis according to the stage deflection angle, and a second error indicates a displacement of the observation point for each stage position in the optical axis direction according to the stage deflection angle for each stage position.

[0017] A shape measuring machine for achieving the object of the present invention includes a data acquisition unit that acquires a plurality of shape measurement data continuously measured by a measuring device that measures the shape of a surface to be measured in a non-contact manner while moving the surface to be measured relative to the measuring device in a direction perpendicular to the measurement optical axis of the measuring device, the plurality of shape measurement data having partially overlapping measurement ranges between successive shape measurement data, a first synthetic data generation unit that generates first synthetic data by stitching the plurality of shape measurement data acquired by the data acquisition unit, a second synthetic data generation unit that generates second synthetic data by patching the plurality of shape measurement data acquired by the data acquisition unit, a differential data generation unit that generates difference data between the first synthetic data and the second synthetic data, a first error calculation unit that calculates a first error that occurs when the shape measurement data are stitched together by the stitching process from the differential data, and a correction data calculation unit that subtracts the first error from the first synthetic data to calculate correction data for the first synthetic data.

[0018] In a shape measuring instrument according to another aspect of the present invention, the first error calculation unit includes a noise removal unit that performs noise removal processing on the difference data, and a subtraction processing unit that performs subtraction processing to subtract from the difference data a second error that occurs when connecting shape measurement data by patching processing and is not removed by the noise removal processing. Effect of the Invention

[0019] According to the present invention, when shape measurement data is stitched together by stitching, it is possible to generate composite data in which stitching processing errors have been corrected. [Brief description of the drawings]

[0020] [Figure 1] FIG. 2 is a schematic diagram of a shape measuring machine for measuring the shape of a measurement surface of various objects to be measured. [Diagram 2] FIG. 2 is a top view of an electrostatic chuck that is an object to be measured by the shape measuring machine. [Diagram 3] 2 is an enlarged cross-sectional view of a wafer holding surface of the electrostatic chuck taken along the X direction. FIG. [Figure 4] FIG. 2 is a functional block diagram of a control device. [Diagram 5] 11 is an explanatory diagram for explaining measurement control of shape measurement data by a measurement control unit. FIG. [Figure 6] 11 is a flowchart showing a flow of a straightness measurement process by the form measuring machine. [Figure 7] 7 is a graph showing an example of composite shape measurement data generated in the process of step S3 in FIG. 6. [Figure 8] 7 is a graph showing an example of composite shape measurement data generated in the processing of step S4 in FIG. 6. [Figure 9] 11 is an explanatory diagram for explaining deflection for each position in the X direction of the moving stage. FIG. [Figure 10] 11 is an explanatory diagram for explaining the displacement in the Z direction of the observation point on the wafer holding surface within the measurement range according to the bending angle of the moving stage. FIG. [Figure 11]11 is an explanatory diagram for explaining a Z direction position error contained in the composite shape measurement data. FIG. [Figure 12] FIG. 7 is an explanatory diagram for explaining the generation process of differential data in the process of step S5 in FIG. 6. [Figure 13] FIG. 7 is an explanatory diagram for explaining the noise removal process executed in the process of step S6 in FIG. 6. [Figure 14] FIG. 7 is an explanatory diagram for explaining the subtraction process executed in the process of step S8 in FIG. 6. [Figure 15] 7 is an explanatory diagram for explaining the correction process of composite shape measurement data executed in the process of step S9 in FIG. 6, and the straightness calculation process executed in the process of step S10. FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0021] [Configuration of shape measuring machine] Fig. 1 is a schematic diagram of a shape measuring machine 10 that measures the shape of a measurement surface of various objects. Fig. 2 is a top view of an electrostatic chuck W that is the measurement target of the shape measuring machine 10. Fig. 3 is an enlarged cross-sectional view along the X direction of a wafer holding surface Wa (corresponding to the measurement surface of the present invention) of the electrostatic chuck W. Note that, among the mutually orthogonal XYZ directions in the figure, the XY direction is a direction parallel to the horizontal direction, and the Z direction is a height direction (up and down direction).

[0022] 1 to 3, the wafer holding surface Wa of the electrostatic chuck W is embossed to form a large number of embossments 9. The shape measuring machine 10 measures the shape of the wafer holding surface Wa, and measures the straightness of the upper surface of each embossment 9 based on the shape measurement results (hereinafter, simply referred to as straightness measurement). Note that in FIG. 3 (as well as other figures), the shape, diameter, height, and spacing of the embossments 9 are partially exaggerated to clarify them.

[0023] The shape measuring machine 10 is used to carry out the shape measuring method of the present invention, and includes a base 12 , a linear guide mechanism 14 , a moving stage 16 , a measuring device 18 , and a control device 20 .

[0024] The base 12 has an upper surface parallel to the XY directions, in other words, an upper surface perpendicular to a measurement optical axis OA of a measurement device 18 described later.

[0025] The linear guide mechanism 14 moves the electrostatic chuck W (wafer holding surface Wa) in the X direction relative to a measuring device 18, which will be described later. The linear guide mechanism 14 includes a rail 14a, a block 14b, and an actuator 14c.

[0026] One or more rails 14a are provided on the upper surface of the base 12, and extend in the X direction. Note that a plurality of rails 14a may be provided. The block 14b is held by the rails 14a so as to be movable in the X direction. A moving stage 16 is attached to this block 14b.

[0027] The actuator 14c is, for example, a motor drive mechanism, etc. Under the control of a control device 20 (described later), the actuator 14c moves the block 14b along the X direction, thereby moving the moving stage 16 (electrostatic chuck W) in the X direction.

[0028] Although not shown, the linear guide mechanism 14 is provided with a detection sensor that detects the X-direction position of the block 14b, i.e., the X-direction position of the moving stage 16. This detection sensor sequentially outputs stage position information indicating the X-direction position of the moving stage 16 to the control device 20.

[0029] The moving stage 16 has an upper surface parallel to the XY direction (perpendicular to the optical axis direction of the measurement optical axis OA) and is attached to the block 14b. An electrostatic chuck W is placed on the upper surface of the moving stage 16. When the moving stage 16 is moved in the X direction by the linear guide mechanism 14, the electrostatic chuck W is moved in the X direction relative to a measuring device 18 described later.

[0030] The measuring device 18 is disposed above the base 12 and the moving stage 16 in the Z direction, and has a measurement optical axis OA parallel to the Z direction. For example, a vertical scanning white light interference microscope and a laser confocal microscope are used for this measuring device 18, which measures the shape of the wafer-holding surface Wa in a non-contact manner and outputs shape measurement data D indicating the shape of the wafer-holding surface Wa to the control device 20. Note that the shape measurement data D also includes data used for calculating the shape of the wafer-holding surface Wa, such as an interference signal detected by a white light interference microscope.

[0031] The measurement range R (observation field of view, field of view range) in which the measuring device 18 can measure the shape of the wafer-holding surface Wa in one measurement is smaller than the wafer-holding surface Wa. For this reason, the measuring device 18 continuously measures the shape of the wafer-holding surface Wa within the measurement range R while the moving stage 16 (electrostatic chuck W) is moved in the X direction by the linear guide mechanism 14. This allows for the acquisition of multiple shape measurement data D that indicate the shape of a straightness measurement region, which is a region within the wafer-holding surface Wa with a length required for straightness measurement.

[0032] Fig. 4 is a functional block diagram of the control device 20. As shown in Fig. 4 and the above-mentioned Fig. 1, the control device 20 comprehensively controls the operations of the linear guide mechanism 14 and the measuring device 18, and, as will be described later in detail, calculates and corrects composite shape measurement data DS (see Fig. 7) of the straightness measurement region based on a plurality of shape measurement data D input from the measuring device 18, and calculates the straightness of the upper surface of each embossment 9. In addition to the linear guide mechanism 14 and the measuring device 18, a memory unit 22, an operation unit 24, and a display unit 26 are connected to this control device 20.

[0033] In addition to a control program (not shown), the memory unit 22 also pre-stores a Z-direction position error E2 (see Figure 9), which is a calculation error that occurs when each piece of shape measurement data D is stitched together by patching processing, as will be described in detail later.

[0034] The operation unit 24 is used for various operations of the form measuring machine 10, such as starting various measurements such as straightness measurement, and setting operations for each part of the form measuring machine 10. The display unit 26 displays various setting screens and the measurement results of various measurements such as straightness measurement.

[0035] [Control device functions] The control device 20 is configured by a calculation device such as a personal computer, and includes a calculation circuit configured by various processors and memories, etc. The various processors include a central processing unit (CPU), a graphics processing unit (GPU), an application specific integrated circuit (ASIC), and a programmable logic device (e.g., simple programmable logic devices (SPLD), complex programmable logic devices (CPLD), and field programmable gate arrays (FPGA)). The various functions of the control device 20 may be realized by one processor, or may be realized by multiple processors of the same type or different types.

[0036] The control device 20 executes a control program read from the storage unit 22, thereby functioning as a measurement control unit 30, a data acquisition unit 32, and an arithmetic processing unit 33.

[0037] Fig. 5 is an explanatory diagram for explaining measurement control of shape measurement data D by the measurement control unit 30. As shown in Fig. 5, the measurement control unit 30 drives the linear guide mechanism 14 in response to a measurement start operation on the operation unit 24 to move the moving stage 16 and the electrostatic chuck W in the X direction.

[0038] At the same time, the measurement control unit 30 causes the measuring device 18 to perform shape measurement of the wafer holding surface Wa every time the moving stage 16 moves a predetermined distance in the X direction. As a result, the measuring device 18 performs continuous shape measurement of the wafer holding surface Wa within the measurement range R while relatively moving (scanning) the measurement range R in the X direction within the straightness measurement region of the wafer holding surface Wa. At this time, the measurement control unit 30 adjusts the timing of shape measurement by the measuring device 18 and the moving speed of the moving stage 16 so that the measurement range R of the measuring device 18 partially overlaps before and after successive measurements (see the hatched area in the figure). As a result, the measurement ranges R of successive successive shape measurement data D partially overlap, so that the successive successive shape measurement data D can be joined together in their overlapping areas.

[0039] 4, the data acquiring unit 32 is connected by wire or wirelessly to the measuring device 18 via a communication interface (not shown). Every time the measuring device 18 performs a shape measurement of the wafer holding surface Wa, the data acquiring unit 32 acquires shape measurement data D from the measuring device 18 and acquires stage position information of the moving stage 16 from the linear guide mechanism 14. This makes it possible to obtain a plurality of shape measurement data D indicating the shape within the straightness measurement region and stage position information at the time when each shape measurement data D was measured.

[0040] Based on each shape measurement data D and the positional information for each shape measurement data D acquired by the data acquisition unit 32, the calculation processing unit 33 performs the following operations: (1) generates composite shape measurement data DS (see Figure 7) of the straightness measurement area by stitching processing, (2) calculates the stitching processing error E1 (see Figure 14) generated in the stitching processing, (3) corrects the composite shape measurement data DS using the stitching processing error E1, and (4) calculates the straightness of the top surface of each embossing 9.

[0041] Furthermore, the calculation processor 33 calculates the stitching processing error E1 (see FIG. 14) in (2) above based on the difference data ΔD between the composite shape measurement data DS generated by the stitching processing and the composite shape measurement data DP of the straightness measurement area generated by the patching processing, as shown in FIG. 12 described below. Therefore, the calculation in (2) above includes (2-1) generation of the composite shape measurement data DP by the patching processing, (2-2) calculation of the difference data ΔD between the composite shape measurement data DP, DS, and (2-3) calculation of the stitching processing error E1 from the difference data ΔD.

[0042] To perform the above processes (1) to (4), the calculation processing unit 33 functions as a stitching processing unit 34, a patching processing unit 36, a difference data generating unit 38, a Z-direction position error acquiring unit 40, a stitching processing error calculation unit 42, a correction data calculation unit 44, and a straightness calculation unit 46. In addition, the stitching processing error calculation unit 42 functions as a noise removal unit 42a and a subtraction processing unit 42b.

[0043] The stitching processing unit 34 performs the process (1) above. The patching processing unit 36, the difference data generation unit 38, the Z-direction position error acquisition unit 40, and the stitching processing error calculation unit 42 perform the process (2) above. Specifically, the patching processing unit 36 ​​performs the process (2-1) above, the difference data generation unit 38 performs the process (2-2) above, and the Z-direction position error acquisition unit 40 and the stitching processing error calculation unit 42 (noise removal unit 42a, subtraction processing unit 42b) perform the process (2-3) above. In addition, the correction data calculation unit 44 performs the process (3) above, and the straightness calculation unit 46 performs the process (4) above.

[0044] [Operation of this embodiment] FIG. 6 is a flowchart showing the flow of straightness measurement processing by the shape measuring machine 10 according to the shape measuring method of the present invention.

[0045] As shown in Fig. 6, after the electrostatic chuck W is placed on the moving stage 16, the operator performs an operation to start straightness measurement using the operation unit 24, which activates each unit of the control device 20. First, the measurement control unit 30 controls the linear guide mechanism 14 and the measurement device 18 to perform continuous measurement of shape measurement data D within the straightness measurement area as shown in Fig. 5 (step S1), and the data acquisition unit 32 acquires each piece of shape measurement data D and stage position information at the time of measuring each piece of shape measurement data D (step S2). Note that step S2 corresponds to the data acquisition step of the present invention.

[0046] Fig. 7 is a graph showing an example of the composite shape measurement data DS generated in the process of step S3 in Fig. 6 [the process of (1) above]. As shown in step S3 in Fig. 6 and Fig. 7, the stitching processing unit 34 stitches together the respective shape measurement data D acquired by the data acquisition unit 32 through a stitching process to generate composite shape measurement data DS of the straightness measurement region. Note that the stitching processing unit 34 corresponds to the first composite data generation unit of the present invention, the composite shape measurement data DS corresponds to the first composite data of the present invention, and step S3 corresponds to the first composite data generation step of the present invention.

[0047] Specifically, the stitching processing unit 34 determines overlapping areas between successive previous and next pieces of shape measurement data D in each piece of shape measurement data D, based on stage position information at the time of measurement of each piece of shape measurement data D acquired by the data acquisition unit 32. Then, for each piece of previous and next shape measurement data D, the stitching processing unit 34 performs processing using a known best-fit method on the data of the overlapping areas, and then calculates the amount of rotation [pitching angle around the Y-axis and rolling angle around the X-axis] for smoothly joining the overlapping areas of the previous and next pieces of shape measurement data D.

[0048] Next, the stitching processing unit 34 performs rotation correction for each of the front and rear shape measurement data D by rotating the front and rear shape measurement data D based on the rotation amount previously calculated, and then stitches together the overlapping areas of the front and rear shape measurement data D at the average height of the overlapping areas in the Z direction. At this time, the overlapping areas of the front and rear shape measurement data D use the one with the higher signal strength of the two data. This generates the composite shape measurement data DS.

[0049] In such stitching processing, if there is a difference in the overlapping area between the previous and next shape measurement data D, for example if noise components are included in either of the overlapping areas of the previous and next shape measurement data D, a calculation error will be included in the calculation of the above-mentioned rotation amount. For this reason, when the previous and next shape measurement data D are stitched together, a stitching processing error E1 (corresponding to the first error of the present invention, see FIG. 14), which is an error component such as a waviness component, will be included in the stitched portion (overlapping area).

[0050] Therefore, in this embodiment, the stitching processing error E1 is calculated through the processing of steps S4 to S8 described below [processing (2) above], and then the composite shape measurement data DS is corrected based on the stitching processing error E1 in the processing of step S9 described below [processing (3) above].

[0051] Fig. 8 is a graph showing an example of the composite shape measurement data DP generated in the process of step S4 in Fig. 6 [the process of (2-1) above]. As shown in step S4 in Fig. 6 and Fig. 8, the patching processing unit 36 ​​generates composite shape measurement data DP of the straightness measurement region by patching together the respective shape measurement data D acquired by the data acquisition unit 32. Note that the patching processing unit 36 ​​corresponds to the second composite data generating unit of the present invention, the composite shape measurement data DP corresponds to the second composite data of the present invention, and step S4 corresponds to the second composite data generating step of the present invention.

[0052] Specifically, the patching processor 36 determines overlapping areas between successive front and rear shape measurement data D in each piece of shape measurement data D based on stage position information at the time of measurement of each piece of shape measurement data D, similar to the above-mentioned stitching process. Next, for each piece of front and rear shape measurement data D, the patching processor 36 joins the overlapping areas of the front and rear shape measurement data D in the XY direction at common XY coordinates within the overlapping areas, and joins them in the Z direction at the average height of the two overlapping areas for each XY coordinate. Then, for each piece of front and rear shape measurement data D, the patching processor 36 embeds surface data created by the average value of both data in the overlapping areas of the front and rear shape measurement data D. This generates composite shape measurement data DP.

[0053] Such composite shape measurement data DP includes Z-direction position error E2 (see FIG. 11) caused by the deflection of moving stage 16 during movement in the X-direction.

[0054] FIG. 9 is an explanatory diagram for explaining the deflection of the moving stage 16 for each position in the X direction.

[0055] As shown in Fig. 9, when the moving stage 16 moves in the X direction during continuous measurement of the shape measurement data D, the center of gravity position G of the electrostatic chuck W on the moving stage 16 changes from the start of the movement to the end of the movement. Therefore, the load moment applied to the moving stage 16 changes, and the moving stage 16 is deflected (pitching and rolling). As a result, the deflection angle (pitching angle and rolling angle) of the moving stage 16 changes for each position of the moving stage 16 in the X direction. As a result, the observation point P (see Fig. 10) of the wafer holding surface Wa within the measurement range R changes in the Z direction, i.e., in a direction parallel to the measurement optical axis OA, for each position of the moving stage 16 in the X direction.

[0056] Fig. 10 is an explanatory diagram for explaining the displacement in the Z direction of the observation point P (also called the measurement point) on the wafer holding surface Wa within the measurement range R according to the bending angle of the movable stage 16. Note that the symbol C in Fig. 10 indicates the center position in the X direction of the block 14b (movable stage 16). Also, the actuator 14c (see Fig. 1) is omitted in Fig. 10.

[0057] 10, if the length of block 14b in the X direction is "B", the distance in the X direction from block 14b to the measurement optical axis OA (observation point P) is "x", and the deflection angle (pitching angle, etc.) of moving stage 16 is "i", then observation point P is displaced in the Z direction by a displacement value ΔZ [ΔZ=(x+B / 2)×sin(i)] compared to observation point P0 when the deflection angle i is 0 degrees. This ΔZ changes for each position of moving stage 16 in the X direction.

[0058] Fig. 11 is an explanatory diagram for explaining the Z direction position error E2 included in the composite shape measurement data DP. As shown in Fig. 11, during continuous measurement of shape measurement data D, the Z direction displacement value ΔZ of observation point P changes for each X direction position of moving stage 16, but in the patching process, no correction is made when connecting each piece of shape measurement data D, unlike the stitching process described above. For this reason, the composite shape measurement data DP includes a "Z direction position error E2" equivalent to the displacement value ΔZ for each X direction position of moving stage 16.

[0059] Here, the deflection angle i (pitching angle, etc.) for each X-direction position of the moving stage 16 is found using a known formula, or through experiments, simulations, etc. For this reason, it is possible to previously find the Z-direction position error E2 for each X-direction position of the moving stage 16 by substituting the distance x and the deflection angle i into the above-mentioned formula for calculating the displacement value ΔZ. In this embodiment, the Z-direction position error E2 is previously stored in the storage unit 22.

[0060] The composite shape measurement data DP also contains calculation errors other than the Z direction position error E2. As described above, the patching process does not perform corrections such as the stitching process described above, so a step error, which is a short wavelength (high frequency) error component, occurs in the joints (overlapping areas) of the previous and next shape measurement data D. For this reason, the composite shape measurement data DP contains the Z direction position error E2 and a step error.

[0061] In this embodiment, the composite shape measurement data DS is generated by the stitching process in step S3 and then the composite shape measurement data DP is generated by the patching process in step S4, but the order may be reversed or the two processes may be performed simultaneously.

[0062] Fig. 12 is an explanatory diagram for explaining the generation of differential data ΔD in the process of step S5 in Fig. 6 [the process of (2-2) above]. As shown in step S5 in Fig. 6 and reference symbol XIIA in Fig. 12, the differential data generation unit 38 performs a process of subtracting one of two types of composite shape measurement data DS, DP that are created using different joining methods from the other. In this way, the differential data generation unit 38 calculates differential data ΔD between the composite shape measurement data DS, DP as shown by reference symbol XIIB in Fig. 12. Note that step S5 corresponds to the differential data generation step of the present invention.

[0063] By calculating the difference between the composite shape measurement data DS, DP in this manner, the shape components of the wafer holding surface Wa (each embossment 9) are removed from the difference data ΔD. Therefore, the difference data ΔD includes a stitching process error E1 (see FIG. 14) that occurs in the stitching process, a Z-direction position error E2 and step error that occur in the patching process, and an X-direction position error. Note that the X-direction position error is a short-wavelength (high-frequency) error component that occurs by calculation of the difference data ΔD due to the fact that the X-coordinate values ​​of shape components such as the embossment 9 in the composite shape measurement data DS, DP are not the same, that is, there is an error in the X-coordinate value.

[0064] Next, the stitching processing error calculation unit 42 (noise removal unit 42a and subtraction processing unit 42b) starts the processing from step S6 to step S8 [the processing of (2-3) above] in which the stitching processing error E1 (see FIG. 14) is calculated from the difference data ΔD. Therefore, steps S6 to S8 correspond to the first error calculation step of the present invention, and the stitching processing error calculation unit 42 corresponds to the first error calculation unit of the present invention.

[0065] Fig. 13 is an explanatory diagram for explaining the noise removal process executed in the process of step S6 in Fig. 6. As shown by reference symbol XIIIA in Fig. 13, in the difference data ΔD, the main components are stitching process error E1 (see Fig. 14) and Z direction position error E2, which are long wavelength (low frequency) signal components generated according to the stitching intervals of the shape measurement data D, and the step error and X direction position error, which are short wavelength (high frequency) signal components, appear as noise components.

[0066] Therefore, as shown in step S6 of FIG. 6 and reference symbol XIIIB of FIG. 13, the noise removal unit 42a performs a noise removal process on the difference data ΔD to remove noise components (step error and X-direction position error) from the difference data ΔD (corresponding to the noise removal step of the present invention). As described above, the stitching process error E1 (see FIG. 14) occurs according to the stitching interval of the shape measurement data D. Therefore, the stitching process error E1 can be considered as a signal with a wavelength that is twice or more as long as the measurement range R of one measurement of the measuring device 18. Therefore, the noise removal unit 42a performs a filter process as a noise removal process to extract wavelength components that are twice or more (e.g., three times) as long as the measurement range R from the difference data ΔD. As a result, the difference data ΔD1 including the stitching process error E1 and the Z-direction position error E2 is generated.

[0067] 6, the Z-direction position error acquisition unit 40 acquires the Z-direction position error E2 from the storage unit 22 and outputs it to the subtraction processing unit 42b of the stitching processing error calculation unit 42 (corresponding to a second error acquisition step of the present invention). In this embodiment, the Z-direction position error acquisition unit 40 acquires the Z-direction position error E2 from the storage unit 22, but it may also be acquired from outside the shape measuring machine 10 (for example, a server on the Internet). The timing of step S7 is not particularly limited as long as it is before S8 described later.

[0068] Fig. 14 is an explanatory diagram for explaining the subtraction process executed in the process of step S8 in Fig. 6. As shown in step S8 in Fig. 6 and Fig. 14, the subtraction processing unit 42b performs subtraction process (corresponding to the subtraction process step of the present invention) to subtract the Z direction position error E2 (see symbol XIVB) from the difference data ΔD1 (see symbol XIVA). As a result, by removing the Z direction position error E2 from the difference data ΔD1, the stitching process error E1 (see symbol XIVC) is finally calculated.

[0069] In this embodiment, the noise removal process (step S6) is performed on the difference data ΔD before the subtraction process (step S8). However, the noise removal process may be performed first before the subtraction process.

[0070] Fig. 15 is an explanatory diagram for explaining the correction process of the composite shape measurement data DS executed in the process of step S9 [the process of (3) above] in Fig. 6, and the straightness calculation process executed in the process of step S10 [the process of (4) above]. Note that the symbol ST in Fig. 15 is straightness measurement data of the upper surface of each embossment 9 obtained from the composite shape measurement data DS and the correction data DS1, respectively.

[0071] As shown in step S9 of Fig. 6 and reference symbol XVA of Fig. 15, the correction data calculation unit 44 calculates correction data DS1 by correcting the composite shape measurement data DS by subtracting the stitching process error E1 calculated in step S8 from the composite shape measurement data DS generated in the above-mentioned step S3. This results in the correction data DS1, which is the shape measurement result of the wafer holding surface Wa (straightness measurement area) excluding the stitching process error E1, as shown in reference symbol XVB of Fig. 15. Note that step S9 corresponds to the correction data calculation step of the present invention.

[0072] 6 and symbol XVB in Fig. 15, the straightness calculation unit 46 connects together the upper surfaces of the embossments 9 included in the correction data DS1 to calculate straightness measurement data ST for the upper surface of each embossment 9. Note that instead of calculating the straightness measurement data ST from the correction data DS1, the straightness calculation unit 46 may execute a process of calculating straightness measurement data ST from the composite shape measurement data DS before correction and a process of subtracting a stitching processing error E1 from this straightness measurement data ST.

[0073] As described above, in this embodiment, the stitching process error E1 is calculated from the difference data ΔD between the composite shape measurement data DS generated by the stitching process and the composite shape measurement data DP generated by the patching process, and the composite shape measurement data DS can be corrected based on this stitching process error E1. This allows for the acquisition of corrected data DS1 (corrected composite shape measurement data DS) excluding the stitching process error E1. As a result, the straightness of the top surface of each embossment 9 can be measured with higher accuracy. [others] In the above embodiment, the linear guide mechanism 14 moves the measuring device 18 and the wafer holding surface Wa (moving stage 16) relatively in the X direction, but various relative movement mechanisms may be used to move the measuring device 18 and the wafer holding surface Wa relatively in the X direction. In this case, in step S7, an error (second error of the present invention) that occurs when the shape measurement data D is joined by the patching process and that is not removed by the noise removal process in step S6 is obtained in advance. Then, in step S8, a subtraction process is performed on the difference data ΔD1 based on this obtained error.

[0074] In the above embodiment, an example has been described in which the shape of the wafer holding surface Wa of an embossed electrostatic chuck W and the straightness of the upper surface of each embossment 9 are measured. However, the present invention can be applied to cases in which various shapes of various measurement surfaces of various measurement objects are measured.

[0075] The shape measuring machine 10 in the above embodiment comprises a base 12, a linear guide mechanism 14, a moving stage 16, a measuring device 18 and a control device 20, but the shape measuring machine of the present invention also includes a machine that is composed only of the control device 20 (including software and programs), i.e., an analysis device (programs, etc.) that acquires and analyzes each shape measurement data D measured by an external measuring device / measuring machine. [Explanation of symbols]

[0076] 9. Embossing 10 Shape measuring machine 12 Foundation 14 Linear guide mechanism 14a Rail Block 14b 14c Actuator 16 Moving Stage 18 Measuring Equipment 20 Control device 22 Memory section 24 Control section 26 Display section 30 Measurement control section 32 Data Acquisition Section 33 Processing unit 34 Stitching Processing Section 36 Patching Processing Section 38 Differential Data Generation Unit 40 Z direction position error acquisition section 42 Stitching processing error calculation unit 42a Noise removal section 42b Subtraction processing section 44 Correction data calculation section 46 Straightness calculation section D Shape measurement data DP synthetic shape measurement data DS synthetic shape measurement data DS1 correction data E1 Stitching process error E2 Z direction position error G Center of gravity position OA measurement optical axis P,P0 observation point R Measurement range ST Straightness measurement data W electrostatic chuck Wa Wafer holding surface

Claims

1. a data acquiring step of acquiring a plurality of shape measurement data pieces successively measured by a measuring device that measures the shape of a surface to be measured in a non-contact manner while moving the surface to be measured relative to the measuring device in a direction perpendicular to a measurement optical axis of the measuring device, the plurality of shape measurement data pieces having measurement ranges partially overlapping between adjacent shape measurement data pieces; a first composite data generation step of stitching together the plurality of shape measurement data acquired in the data acquisition step to generate first composite data; a second composite data generating step of generating second composite data by patching the plurality of shape measurement data acquired in the data acquiring step; a correcting step of correcting the first combined data based on difference data between the first combined data and the second combined data; A shape measurement method comprising the steps of:

2. a data acquisition unit that acquires a plurality of shape measurement data pieces that are continuously measured by a measuring device that measures the shape of a surface to be measured in a non-contact manner while moving the surface to be measured relative to the measuring device in a direction perpendicular to a measurement optical axis of the measuring device, the plurality of shape measurement data pieces having measurement ranges that partially overlap with each other between adjacent pieces of the shape measurement data; a first composite data generation unit that stitches together the plurality of shape measurement data acquired by the data acquisition unit to generate first composite data; a second composite data generating unit that generates second composite data by patching the plurality of shape measurement data acquired by the data acquiring unit; a correction unit that corrects the first combined data based on difference data between the first combined data and the second combined data; A shape measuring machine equipped with

Images