How to create a reference spectral library used for estimating the film thickness of a workpiece.

Abstract

This technology expands the range of estimable film thicknesses by broadening the range of multiple reference spectra used to estimate the film thickness of workpieces such as wafers. [Solution] The method for creating a reference spectrum library involves extrapolating a pre-polishing reference spectrum corresponding to a pre-polishing time tm-1 prior to the start of polishing tm of the reference workpiece, and a post-polishing reference spectrum corresponding to a post-polishing time tn+1 prior to the end of polishing tn of the reference workpiece. Multiple reference film thicknesses T-1 to T+1 corresponding to the pre-polishing reference spectrum, multiple actual reference spectra, and the post-polishing reference spectrum are calculated, and the reference spectrum library is created by associating the multiple reference film thicknesses T-1 to T+1 with the pre-polishing reference spectrum, multiple actual reference spectra, and the post-polishing reference spectrum, respectively.

Description

Technical Field

[0001] The present invention relates to a method for creating a reference spectrum library used for estimating the film thickness of workpieces such as wafers and substrates. In particular, it relates to a method for creating a reference spectrum library including a plurality of reference spectra to be compared with the measurement spectrum of the reflected light from the workpiece.

Background Art

[0002] An optical film thickness measurement device is configured to generate a measurement spectrum of the reflected light from a wafer, determine the reference spectrum in the reference spectrum library that is closest in shape to the measurement spectrum, and determine the film thickness pre-associated with the determined reference spectrum.

[0003] FIG. 12 is a diagram for explaining a process of determining the film thickness from the comparison between the measurement spectrum and a plurality of reference spectra. The optical film thickness measurement device compares the measurement spectrum generated during the polishing of the wafer with a plurality of pre-prepared reference spectra, determines the reference spectrum that is closest in shape to the measurement spectrum, and determines the film thickness pre-associated with the determined reference spectrum. The reference spectrum that is closest in shape to the measurement spectrum is the spectrum with the smallest difference in shape between the reference spectrum and the measurement spectrum.

[0004] The plurality of reference spectra are obtained by pre-polishing a reference wafer having the same surface structure as the wafer to be polished. Each reference spectrum is associated with the film thickness when the reference spectrum is obtained. That is, the plurality of reference spectra are obtained when the film thicknesses of the reference wafers are different, and the plurality of reference spectra correspond to a plurality of different film thicknesses. Therefore, by specifying the reference spectrum that is closest in shape to the measurement spectrum, the current film thickness of the wafer can be estimated.

[0005] An example of a process for obtaining multiple reference spectra and corresponding film thicknesses is described below. First, a reference wafer having the same surface structure as the wafer to be polished is prepared. The reference wafer is transported to a film thickness measuring instrument, and the initial film thickness of the reference wafer is measured by the film thickness measuring instrument. Next, the reference wafer is transported to a polishing machine, and the reference wafer is polished by the polishing machine. During the polishing of the reference wafer, light is shone on the surface of the reference wafer, and the spectrum of the reflected light from the reference wafer (i.e., the reference spectrum) is generated. The reference spectrum is generated periodically during the polishing of the reference wafer. Therefore, multiple reference spectra are obtained as the film thickness decreases during the polishing of the reference wafer. After the polishing of the reference wafer is completed, the reference wafer is transported again to the film thickness measuring instrument, and the film thickness of the polished reference wafer (i.e., the final film thickness) is measured.

[0006] Figure 13 is a graph showing the relationship between the film thickness of a reference wafer and the polishing time. Under the condition that the polishing rate (also called the removal rate) of the reference wafer is constant, the film thickness decreases linearly with polishing time from the initial film thickness T0 to the final film thickness Tf, as shown in Figure 13. In other words, the film thickness can be expressed as a linear function in which polishing time is a variable. The polishing rate can be calculated by dividing the difference between the initial film thickness T0 and the final film thickness Tf by the difference between the polishing time tfin at the final film thickness Tf and the polishing time tini at the initial film thickness T0. Polishing rate = [T0 - Tf] / [tfin - tini]

[0007] The times t1, t2, ..., tn at which multiple reference spectra are generated are within the range from polishing time tini to polishing time tfin. The film thickness corresponding to the reference spectrum can be calculated from the initial film thickness T0, the final film thickness Tf, and the times t1 to tn at which the reference spectra were generated. For example, the film thickness corresponding to the reference spectrum generated at time t2 can be calculated using the following formula. T0-[[T0-Tf] / [tfin-tini]]·[t2-tini]

[0008] In this way, multiple reference spectra corresponding to different film thicknesses are obtained. Each reference spectrum is associated with (linked to) a corresponding film thickness. The optical film thickness measuring device can estimate the current film thickness of the wafer from the film thickness associated with the reference spectrum by identifying the reference spectrum whose shape is closest to the measurement spectrum during wafer polishing. [Prior art documents] [Patent Documents]

[0009] [Patent Document 1] Japanese Patent Publication No. 2022-10553 [Patent Document 2] Special Publication No. 2016-510953 [Overview of the Initiative] [Problems that the invention aims to solve]

[0010] As can be seen from Figure 13, the reference spectrum is generated while the film thickness of the reference wafer decreases from the initial film thickness T0 to the final film thickness Tf. Therefore, reference spectra corresponding to film thicknesses greater than the initial film thickness T0, and reference spectra corresponding to film thicknesses less than the final film thickness Tf, are not obtained. If the current film thickness of the wafer being polished is greater than the initial film thickness T0, or less than the final film thickness Tf, the current film thickness of the wafer cannot be estimated because there is no corresponding reference spectrum.

[0011] Therefore, the present invention provides a technique for expanding the range of estimable film thickness by broadening the range of multiple reference spectra used for estimating the film thickness of a workpiece such as a wafer. [Means for solving the problem]

[0012] In one embodiment, a method for creating a reference spectrum library used for estimating the film thickness of a workpiece is provided, comprising: measuring the initial film thickness of a reference workpiece before polishing the reference workpiece; polishing the reference workpiece with a polishing device; generating multiple actual reference spectra of reflected light from the reference workpiece at multiple actual polishing times during the polishing of the reference workpiece; measuring the final film thickness of the reference workpiece after polishing the reference workpiece; calculating the polishing rate by dividing the difference between the initial film thickness and the final film thickness by the time from the start to the end of polishing of the reference workpiece; generating a pre-polishing reference spectrum corresponding to the pre-polishing time before the start of polishing of the reference workpiece and a post-polishing reference spectrum corresponding to the post-polishing time after the end of polishing of the reference workpiece by extrapolation; calculating multiple reference film thicknesses corresponding to the pre-polishing reference spectrum, the multiple actual reference spectra, and the post-polishing reference spectrum, respectively, based on the polishing rate; and creating a reference spectrum library by associating the multiple reference film thicknesses with the pre-polishing reference spectrum, the multiple actual reference spectra, and the post-polishing reference spectrum, respectively.

[0013] In one embodiment, the pre-polishing reference spectrum and the post-polishing reference spectrum are generated by an extrapolation model. In one embodiment, the pre-polishing reference spectrum is output from the extrapolation model by inputting the pre-polishing time into the extrapolation model, and the post-polishing reference spectrum is output from the extrapolation model by inputting the post-polishing time into the extrapolation model. In one embodiment, the extrapolation model is a trained model constructed by machine learning using training data including the multiple real reference spectra and the multiple real polishing times. In one embodiment, the machine learning is long-shorter-time memory (LSTM). In one embodiment, the extrapolation model is a statistical model that analyzes the plurality of actual reference spectra and the plurality of actual polishing times according to a statistical algorithm and generates the pre-polishing reference spectrum and the post-polishing reference spectrum based on the analysis results. [Effects of the Invention]

[0014] The reference spectrum library includes multiple real reference spectra obtained during the polishing of the reference workpiece, as well as pre-polish and post-polish reference spectra generated by extrapolation. The pre-polish reference spectrum is a hypothetical reference spectrum corresponding to a film thickness greater than the initial film thickness of the reference workpiece, and the post-polish reference spectrum is a hypothetical reference spectrum corresponding to a film thickness less than the final film thickness of the reference workpiece. Therefore, the range of reference spectra included in the reference spectrum library is broadened, and the range of film thicknesses that can be estimated during the polishing of the workpiece is broadened. [Brief explanation of the drawing]

[0015] [Figure 1] This is a schematic diagram showing one embodiment of a polishing apparatus. [Figure 2] This is a cross-sectional view showing the detailed configuration of an optical film thickness measuring device. [Figure 3] This is a schematic diagram showing an example of a measurement spectrum generated from light intensity measurement data. [Figure 4] This is a schematic diagram showing one embodiment of a workpiece processing system, including a polishing device and a film thickness measuring device, used in a method for creating a reference spectral library. [Figure 5] This figure shows an example of multiple real reference spectra generated at different polishing times during the polishing of a reference workpiece. [Figure 6] This graph shows an example of the decrease in film thickness of a reference workpiece due to the polishing time of the reference workpiece. [Figure 7] This graph shows an example of a pre-polishing reference spectrum corresponding to the time before polishing, and multiple actual reference spectra generated during the polishing of the reference workpiece. [Figure 8] This graph shows an example of a post-polishing reference spectrum corresponding to the time after polishing, and multiple actual reference spectra generated during the polishing of the reference workpiece. [Figure 9]A graph showing an example of a plurality of reference spectra including a pre-polishing reference spectrum and a post-polishing reference spectrum shown in FIGS. 7 and 8, and a plurality of actual reference spectra generated during the polishing of a reference workpiece. [Figure 10] A graph showing an example of a plurality of reference film thicknesses at a plurality of times corresponding to the plurality of reference spectra shown in FIG. 9. [Figure 11] A flowchart for explaining one embodiment of a method for creating a reference spectrum library. [Figure 12] A diagram for explaining a process of determining a film thickness from a comparison between a measurement spectrum and a plurality of reference spectra. [Figure 13] A graph showing the relationship between the film thickness and the polishing time of a reference wafer.

Mode for Carrying Out the Invention

[0016] Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a schematic diagram showing an embodiment of a polishing apparatus. As shown in FIG. 1, the polishing apparatus 100 includes a polishing table 3 that supports a polishing pad 2, a polishing head 1 that presses a workpiece W against the polishing pad 2, a table motor 6 that rotates the polishing table 3, a polishing liquid supply nozzle 5 for supplying a polishing liquid such as slurry onto the polishing pad 2, and an operation control unit 9 for controlling the operation of the polishing apparatus 100. The upper surface of the polishing pad 2 constitutes a polishing surface 2a for polishing the workpiece W.

[0017] The workpiece W has a film forming a wiring structure on its surface. Examples of the workpiece W include a wafer, a substrate, a wiring board, and a square substrate used in the manufacture of semiconductor devices. In one example, the workpiece W is a product wafer on which a multilayer film is formed.

[0018] The polishing head 1 is connected to the head shaft 10, which is connected to the polishing head rotating device 15. The polishing head rotating device 15 is configured to rotate the polishing head 1 together with the head shaft 10 in the direction indicated by the arrow. The configuration of the polishing head rotating device 15 is not particularly limited, but in one example, the polishing head rotating device 15 includes an electric motor, a belt, a pulley, etc. The polishing table 3 is connected to the table motor 6, which is configured to rotate the polishing table 3 and the polishing pad 2 in the direction indicated by the arrow. The polishing head 1, the polishing head rotating device 15, and the table motor 6 are connected to the operation control unit 9.

[0019] The workpiece W is polished as follows: The table motor 6 and the polishing head rotating device 15 rotate the polishing table 3 and the polishing head 1 in the direction indicated by the arrows in Figure 1, while polishing fluid is supplied from the polishing fluid supply nozzle 5 to the polishing surface 2a of the polishing pad 2 on the polishing table 3. As the workpiece W is rotated by the polishing head 1, the workpiece W is pressed against the polishing surface 2a of the polishing pad 2 by the polishing head 1 while polishing fluid is present on the polishing pad 2. The surface of the workpiece W is polished by the chemical action of the polishing fluid and the mechanical action of the abrasive grains contained in the polishing fluid and / or the polishing pad 2.

[0020] The operation control unit 9 comprises a storage device 9a in which a program is stored, and an arithmetic unit 9b that performs calculations according to the instructions contained in the program. The operation control unit 9 is composed of at least one computer. The storage device 9a comprises a main memory such as random access memory (RAM) and an auxiliary storage device such as a hard disk drive (HDD) or solid state drive (SSD). Examples of arithmetic units 9b include a CPU (central processing unit) and a GPU (graphics processing unit). However, the specific configuration of the operation control unit 9 is not limited to these examples.

[0021] The polishing apparatus 100 is equipped with an optical film thickness measuring device 20 for measuring the film thickness of the workpiece W. The optical film thickness measuring device 20 includes a light source 22 that emits light, an optical sensor head 25 that irradiates the workpiece W with light from the light source 22 and receives reflected light from the workpiece W, a spectrometer 27 connected to the optical sensor head 25, and a processing system 30 that determines the film thickness of the workpiece W based on the spectrum of the reflected light from the workpiece W. The optical sensor head 25 is located in the polishing table 3 and rotates together with the polishing table 3.

[0022] The processing system 30 comprises a storage device 30a in which a program is stored, and an arithmetic unit 30b that performs calculations according to the instructions contained in the program. The processing system 30 consists of at least one computer. The storage device 30a comprises a main memory such as random access memory (RAM) and an auxiliary storage device such as a hard disk drive (HDD) or solid-state drive (SSD). Examples of arithmetic units 30b include a CPU (central processing unit) and a GPU (graphics processing unit). However, the specific configuration of the processing system 30 is not limited to these examples.

[0023] The motion control unit 9 and the processing system 30 may each be composed of multiple computers. For example, the motion control unit 9 and the processing system 30 may each be composed of a combination of edge servers and cloud servers. In one embodiment, the motion control unit 9 and the processing system 30 may be composed of a single computer.

[0024] Figure 2 is a cross-sectional view showing the detailed configuration of the optical film thickness measuring device 20. The optical film thickness measuring device 20 includes a light-emitting optical fiber cable 31 connected to a light source 22 and a light-receiving optical fiber cable 32 connected to a spectrometer 27. The tip 31a of the light-emitting optical fiber cable 31 and the tip 32a of the light-receiving optical fiber cable 32 constitute an optical sensor head 25. That is, the light-emitting optical fiber cable 31 guides the light emitted by the light source 22 to the workpiece W on the polishing pad 2, and the light-receiving optical fiber cable 32 receives the reflected light from the workpiece W and transmits it to the spectrometer 27.

[0025] The spectrometer 27 is connected to the processing system 30. The light-emitting optical fiber cable 31, the light-receiving optical fiber cable 32, the light source 22, and the spectrometer 27 are mounted on the polishing table 3 and rotate together with the polishing table 3 and the polishing pad 2. The optical sensor head 25, consisting of the tip 31a of the light-emitting optical fiber cable 31 and the tip 32a of the light-receiving optical fiber cable 32, is positioned facing the surface of the workpiece W on the polishing pad 2.

[0026] The optical sensor head 25 is positioned so that it crosses the surface of the workpiece W on the polishing pad 2 each time the polishing table 3 and polishing pad 2 rotate once. The polishing pad 2 has a through hole 2b located above the optical sensor head 25. The optical sensor head 25 irradiates the workpiece W with light through the through hole 2b and receives reflected light from the workpiece W through the through hole 2b each time the polishing table 3 rotates once.

[0027] In one embodiment, a flow of pure water may be formed within the through-hole 2b of the polishing pad 2 to prevent polishing fluid and polishing debris from coming into contact with the optical sensor head 25. Light is guided from the optical sensor head 25 through the pure water to the workpiece W, and reflected light from the workpiece W is received by the optical sensor head 25 through the pure water. In another embodiment, a transparent window (not shown) may be fitted within the through-hole 2b of the polishing pad 2. The transparent window is made of a material that allows light transmission (e.g., a transparent resin). In this case, light is guided from the optical sensor head 25 through the transparent window to the workpiece W, and reflected light from the workpiece W is received by the optical sensor head 25 through the transparent window.

[0028] The light source 22 is a flash light source that emits light repeatedly at short time intervals. An example of the light source 22 is a xenon flash lamp. The light source 22 is electrically connected to the motion control unit 9 and emits light in response to a trigger signal sent from the motion control unit 9. More specifically, while the optical sensor head 25 moves across the surface of the workpiece W on the polishing pad 2, the light source 22 receives multiple trigger signals and emits light multiple times. Therefore, each time the polishing table 3 rotates, light is irradiated onto multiple film thickness measurement points on the workpiece W, including the center point.

[0029] Light emitted by the light source 22 is transmitted to the optical sensor head 25. That is, the light is transmitted to the optical sensor head 25 through the light-emitting optical fiber cable 31 and emitted from the optical sensor head 25. The light enters the workpiece W on the polishing pad 2 through the through hole 2b of the polishing pad 2. The light reflected from the workpiece W passes through the through hole 2b of the polishing pad 2 again and is received by the optical sensor head 25. The reflected light from the workpiece W is transmitted to the spectrometer 27 through the light-receiving optical fiber cable 32.

[0030] The spectrometer 27 is configured to decompose reflected light according to its wavelength and measure the intensity of the reflected light at each wavelength over a predetermined wavelength range. That is, the spectrometer 27 decomposes the reflected light from the workpiece W according to its wavelength and measures the intensity of the reflected light at each wavelength over a predetermined wavelength range to generate light intensity measurement data. The intensity of the reflected light at each wavelength can also be expressed as a relative value such as reflectance or relative reflectance. The light intensity measurement data is sent to the processing system 30.

[0031] The processing system 30 generates a reflected light spectrum as shown in Figure 3 from the light intensity measurement data. In the following description, the reflected light spectrum from the workpiece W is referred to as the measured spectrum. The measured spectrum of reflected light from the workpiece W contains information about the film thickness of the workpiece W. In other words, the measured spectrum of reflected light changes depending on the film thickness of the workpiece W. The processing system 30 is configured to determine (estimate) the film thickness of the workpiece W based on the measured spectrum of reflected light. More specifically, the processing system 30 determines a reference spectrum from the reference spectrum library 38 (see Figure 1) that is closest in shape to the measured spectrum of reflected light, and determines the film thickness pre-associated with this determined reference spectrum. The reference spectrum library 38 is created in advance before polishing the workpiece W and stored in the storage device 30a of the processing system 30.

[0032] The following describes one embodiment of the method for creating a reference spectral library 38. Figure 4 is a schematic diagram showing one embodiment of a workpiece processing system including a polishing apparatus 100 and a film thickness measuring apparatus 101 used in the method for creating a reference spectral library 38.

[0033] First, a reference workpiece RW having the same surface structure as workpiece W is prepared. More specifically, reference workpiece RW has an exposed surface made of the same material as workpiece W and has the same laminated structure. Next, reference workpiece RW is transported to the film thickness measuring device 101 by the transport device 103, and the initial film thickness, which is the film thickness of reference workpiece RW before polishing, is measured by the film thickness measuring device 101.

[0034] The film thickness measuring device 101 irradiates a stationary reference workpiece RW with light, generates a spectrum of reflected light from the reference spectrum, and determines the film thickness of the reference workpiece RW by analyzing the spectrum. The basic configuration of this film thickness measuring device 101 is the same as that of the optical film thickness measuring device 20, but it differs from the optical film thickness measuring device 20 in that it measures the film thickness of a stationary reference workpiece RW. The measured initial film thickness of the reference workpiece RW is transmitted from the film thickness measuring device 101 to the processing system 30.

[0035] After measuring the initial film thickness, the reference workpiece RW is transported to the polishing device 100 by the transport device 103 and polished by the polishing device 100. Polishing of the reference workpiece RW is performed in the same way as polishing of workpiece W. That is, the table motor 6 and the polishing head rotating device 15 rotate the polishing table 3 and the polishing head 1 in the direction indicated by the arrows in Figure 1, while polishing fluid is supplied from the polishing fluid supply nozzle 5 to the polishing surface 2a of the polishing pad 2 on the polishing table 3. As the reference workpiece RW is rotated by the polishing head 1, the reference workpiece RW is pressed against the polishing surface 2a of the polishing pad 2 by the polishing head 1 while polishing fluid is present on the polishing pad 2, thereby polishing the surface of the reference workpiece RW.

[0036] During the polishing of the reference workpiece RW, light is shone from the optical sensor head 25 onto the reference workpiece RW, similar to the polishing of the workpiece W, and a spectrum of reflected light from the reference workpiece RW is generated. In the following description, the spectrum of reflected light from the reference workpiece RW obtained during the polishing of the reference workpiece RW is referred to as the actual reference spectrum. Each time the polishing table 3 rotates, light is shone from the optical sensor head 25 onto multiple film thickness measurement points on the reference workpiece RW, including the center point. Each time the polishing table 3 rotates, the processing system 30 generates the actual reference spectrum from the light intensity measurement data generated by the spectrometer 27.

[0037] Figure 5 shows an example of multiple real reference spectra generated at different polishing times during the polishing of a reference workpiece RW. In Figure 5, the vertical axis represents the intensity of reflected light from the reference workpiece RW, and the horizontal axis represents the wavelength of the reflected light. As shown in Figure 5, since the real reference spectrum is generated from the reflected light from the reference workpiece RW being polished, the real reference spectrum changes gradually as the film thickness of the reference workpiece RW decreases (i.e., with polishing time). Therefore, the shapes of the multiple real reference spectra generated from reflected light at different polishing times during the polishing of the reference workpiece RW are slightly different.

[0038] In this way, during the polishing of the reference workpiece RW, multiple actual reference spectra corresponding to multiple polishing times are generated as the film thickness of the reference workpiece RW decreases. In the following description, the multiple polishing times at which these multiple actual reference spectra are generated may be referred to as multiple actual polishing times. After polishing the reference workpiece RW, the reference workpiece RW is transported to the film thickness measuring device 101 by the transport device 103, and the final film thickness, which is the film thickness of the reference workpiece RW after polishing, is measured by the film thickness measuring device 101. The measured value of the final film thickness is transmitted from the film thickness measuring device 101 to the processing system 30. The processing system 30 calculates the polishing rate of the reference workpiece RW by dividing the difference between the initial film thickness and the final film thickness by the time from the start to the end of polishing the reference workpiece RW.

[0039] Figure 6 is a graph showing an example of the decrease in film thickness of a reference workpiece RW with respect to polishing time. The film thickness decreases linearly with polishing time from the initial film thickness Tini to the final film thickness Tfin. In other words, the film thickness can be expressed as a linear function with polishing time as a variable. The polishing rate is calculated by dividing the difference between the initial film thickness Tini and the final film thickness Tfin by the time from the start to the end of polishing of the reference workpiece RW. The time from the start to the end of polishing of the reference workpiece RW corresponds to the difference between the polishing time tn at the final film thickness Tfin and the polishing time tm at the initial film thickness Tini. The polishing rate is expressed by the following formula. Polishing rate = [Tini - Tfin] / [tn - tm]

[0040] The processing system 30 is configured to generate a pre-polishing reference spectrum corresponding to the pre-polishing time before the start of polishing of the reference workpiece RW (e.g., tm-1, tm-2, etc.) and a post-polishing reference spectrum corresponding to the post-polishing time after the end of polishing of the reference workpiece RW (e.g., tn+1, tn+2, etc.) by extrapolation. As shown in Figure 1, the processing system 30 has an extrapolation model 40 pre-stored in its storage device 30a, and generates the pre-polishing reference spectrum and the post-polishing reference spectrum using the extrapolation model 40. Since the polishing time is proportional to the number of rotations of the polishing table 3, the pre-polishing time, the post-polishing time, and multiple actual polishing times (tm~tn) during polishing of the reference workpiece RW may be represented by the number of rotations of the polishing table 3.

[0041] In one embodiment, the extrapolation model 40 is a trained model constructed by machine learning using training data. The training data includes multiple real reference spectra (see Figure 5) obtained from polishing a reference workpiece RW, and multiple real polishing times (tm~tn) corresponding to the multiple real reference spectra. Each of the multiple real reference spectra included in the training data is associated with its corresponding multiple real polishing times. The processing system 30 constructs the extrapolation model 40 by performing machine learning according to instructions contained in a program stored in its storage device 30a. The machine learning is supervised machine learning using the above training data.

[0042] Each real reference spectrum represents multiple intensities of reflected light corresponding to multiple wavelengths, as explained with reference to Figure 5. These multiple intensities of reflected light corresponding to multiple wavelengths can be treated as numerical data representing the characteristics of the real reference spectrum. The extrapolation model 40 learns the characteristics of the multiple real reference spectra included in the training data and generates unknown spectra, namely the pre-polishing and post-polishing reference spectra, according to the learned characteristics. Since each real reference spectrum included in the training data is associated with a corresponding real polishing time, the extrapolation model 40 learns the correlation between the real reference spectrum and the real polishing time. Therefore, as a trained model, the extrapolation model 40 can generate reference spectra corresponding to polishing times outside the range of real polishing times tm~tn of the reference workpiece RW.

[0043] In this embodiment, the extrapolation model 40 is constructed using machine learning. Examples of machine learning applicable to this embodiment include long-short-term memory (LSTM), convolutional neural networks (CNN), and recurrent neural networks (RNN). In this embodiment, long-short-term memory (LSTM) is used as the machine learning method.

[0044] In other embodiments, the extrapolation model 40 may be a statistical model constructed according to a statistical algorithm. The statistical model is configured to analyze multiple real reference spectra and multiple real polishing times according to a statistical algorithm and to generate a pre-polishing reference spectrum and a post-polishing reference spectrum based on the analysis results. Examples of statistical algorithms used to construct the statistical model include autoregression, moving averages, autoregressive moving averages, and state spaces.

[0045] However, the specific configuration of the extrapolation model 40 is not particularly limited, as long as the extrapolation model 40 can generate a pre-polishing reference spectrum corresponding to the time before polishing and a post-polishing reference spectrum corresponding to the time after polishing.

[0046] In the example shown in Figure 6, the processing system 30 inputs the pre-polishing time tm-1, which is before the start of polishing of the reference workpiece RW, into the extrapolation model 40, and outputs a pre-polishing reference spectrum corresponding to the pre-polishing time tm-1 from the extrapolation model 40. Furthermore, the processing system 30 inputs the post-polishing time tn+1, which is after the end of polishing of the reference workpiece RW, into the extrapolation model 40, and outputs a post-polishing reference spectrum corresponding to the post-polishing time tn+1 from the extrapolation model 40. Similarly, the processing system 30 can generate reference spectra corresponding to polishing times outside the range of actual polishing times tm~tn, such as the pre-polishing time tm-2 and the post-polishing time tn+2.

[0047] The reference spectrum generated by the extrapolation model 40 is a virtual spectrum. The number of pre-polishing reference spectra and post-polishing reference spectra to be generated by the extrapolation model 40 is not particularly limited. Three or more pre-polishing reference spectra and three or more post-polishing reference spectra may be generated, or one pre-polishing reference spectrum and one post-polishing reference spectrum may be generated.

[0048] Figure 7 is a graph showing an example of the pre-polishing reference spectrum corresponding to pre-polishing times tm-1 and tm-2, and multiple actual reference spectra generated during the polishing of the reference workpiece RW. Figure 8 is a graph showing an example of the post-polishing reference spectrum corresponding to post-polishing times tn+1 and tn+2, and multiple actual reference spectra generated during the polishing of the reference workpiece RW. Figure 9 is a graph showing an example of multiple reference spectra, including the pre-polishing and post-polishing reference spectra shown in Figures 7 and 8, and multiple actual reference spectra generated during the polishing of the reference workpiece RW.

[0049] The pre-polishing reference spectrum is a virtual reference spectrum corresponding to a film thickness greater than the initial film thickness of the reference workpiece RW, and the post-polishing reference spectrum is a virtual reference spectrum corresponding to a film thickness less than the final film thickness of the reference workpiece RW. The processing system 30 adds the pre-polishing and post-polishing reference spectra shown in Figures 7 and 8 to multiple actual reference spectra generated during the polishing of the reference workpiece RW, thereby obtaining multiple reference spectra including the pre-polishing reference spectrum, the post-polishing reference spectrum, and multiple actual reference spectra, as shown in Figure 9.

[0050] Next, the processing system 30 calculates multiple reference film thicknesses at multiple time intervals tm-2 to tn+2 corresponding to multiple reference spectra (including pre-polishing reference spectra, post-polishing reference spectra, and multiple actual reference spectra) shown in Figure 9. Figure 10 is a graph showing an example of multiple reference film thicknesses at multiple time intervals corresponding to multiple reference spectra shown in Figure 9. The processing system 30 calculates the reference film thicknesses at multiple time intervals tm-2 to tn+2 from the polishing rate and multiple time intervals tm-2 to tn+2 corresponding to the multiple reference spectra. The polishing rate corresponds to the slope of the graph in Figure 10. The time intervals tm-2 to tn+2 used to calculate the reference film thickness include pre-polishing times tm-1, tm-2, post-polishing times tn+1, tn+2, and actual polishing times tm to tn.

[0051] For example, the reference film thickness at time tm+1 can be calculated using the following formula. Tini-[[Tini-Tfin] / [tn-tm]]·[tm+1-tm] Similarly, the processing system 30 can calculate the reference film thicknesses T-1, T-2, T+1, and T+2 corresponding to the pre-polishing times tm-1, tm-2 and the post-polishing times tn+1, tn+2. Since the time is proportional to the number of rotations of the polishing table 3, the time shown on the horizontal axis of Figure 10 may be represented by the number of rotations of the polishing table 3.

[0052] The processing system 30 creates a reference spectrum library 38 by associating each of the multiple reference film thicknesses shown in Figure 10 with the multiple reference spectra shown in Figure 9 (including a pre-polishing reference spectrum, multiple actual reference spectra, and a post-polishing reference spectrum). Each of the multiple reference film thicknesses uniquely corresponds to one of the multiple reference spectra. The reference spectrum library 38 is stored in the storage device 30a of the processing system 30.

[0053] During the polishing of the workpiece W shown in Figure 1, the film thickness of the workpiece W is estimated using a reference spectrum library 38. Specifically, the optical sensor head 25 irradiates the workpiece W with light and receives reflected light from the workpiece W. The spectrometer 27 decomposes the reflected light from the workpiece W according to wavelength and measures the intensity of the reflected light at each wavelength over a predetermined wavelength range to generate light intensity measurement data. The processing system 30 generates a measurement spectrum of the reflected light from the workpiece W from the light intensity measurement data. The processing system 30 determines the reference spectrum from the reference spectrum library 38 that is closest in shape to the measurement spectrum of the reflected light, and determines the reference film thickness associated with this determined reference spectrum. This reference film thickness is the estimated film thickness of the workpiece W.

[0054] The reference spectrum library 38 includes multiple real reference spectra obtained during the polishing of the reference workpiece RW, as well as virtual spectra generated by extrapolation, namely the pre-polishing and post-polishing reference spectra. Therefore, the range of reference spectra included in the reference spectrum library 38 is broadened, and the range of film thickness that can be estimated during the polishing of the workpiece W is broadened.

[0055] Figure 11 is a flowchart illustrating one embodiment of a method for creating a reference spectral library 38. In step S101, a reference workpiece RW having the same surface structure as workpiece W is prepared. Reference workpiece RW has an exposed surface made of the same material as workpiece W and has the same laminated structure. In step S102, the initial film thickness, which is the film thickness of the reference workpiece RW before polishing, is measured by the film thickness measuring device 101 (see Figure 4).

[0056] In step S103, the reference workpiece RW is chemically and mechanically polished using the polishing apparatus 100 shown in Figure 1. In step S104, the reference workpiece RW is illuminated with light at different actual polishing times during the polishing process. The processing system 30 generates a real reference spectrum of the reflected light from the reference workpiece RW. In step S105, after polishing the reference workpiece RW, the final film thickness of the polished reference workpiece RW is measured by the film thickness measuring device 101 (see Figure 4). In step S106, the processing system 30 calculates the polishing rate of the reference workpiece RW by dividing the difference between the initial film thickness and the final film thickness by the time from the start to the end of polishing the reference workpiece RW. The polishing rate of the reference workpiece RW corresponds to the slope of the graph shown in Figure 6.

[0057] In step S107, the processing system 30 extrapolates a pre-polishing reference spectrum corresponding to the pre-polishing time before the start of polishing of the reference workpiece RW, and a post-polishing reference spectrum corresponding to the post-polishing time after the end of polishing of the reference workpiece RW (see Figures 7 and 8). In one embodiment, the processing system 30 inputs the pre-polishing time into the extrapolation model 40, and outputs the pre-polishing reference spectrum corresponding to that pre-polishing time from the extrapolation model 40. Furthermore, the processing system 30 inputs the post-polishing time into the extrapolation model 40, and outputs the post-polishing reference spectrum corresponding to that post-polishing time from the extrapolation model 40. The order of steps S106 and S107 may be reversed, or steps S106 and S107 may be executed simultaneously.

[0058] In step S108, the processing system 30 adds the pre-polishing reference spectrum and the post-polishing reference spectrum to the actual reference spectrum to create multiple reference spectra that constitute the reference spectrum library 38 (see Figure 9). In step S109, the processing system 30 calculates the reference film thickness corresponding to each of the multiple times from the polishing rate and the multiple times corresponding to the multiple reference spectra (including a pre-polishing reference spectrum, multiple actual reference spectra, and a post-polishing reference spectrum) (see Figure 10). The multiple times include the pre-polishing time corresponding to the pre-polishing reference spectrum, the multiple actual polishing times corresponding to the multiple actual reference spectra, and the post-polishing time corresponding to the post-polishing reference spectrum. In step S110, the processing system 30 creates a reference spectrum library 38 by associating multiple reference film thicknesses with multiple reference spectra (including a pre-polishing reference spectrum, multiple actual reference spectra, and a post-polishing reference spectrum).

[0059] The embodiments described above are intended to enable persons with ordinary skill in the art to implement the present invention. Various modifications of the above embodiments can be made naturally by those skilled in the art, and the technical idea of ​​the present invention can be applied to other embodiments as well. Therefore, the present invention is not limited to the embodiments described, but is to be interpreted in the broadest sense according to the technical idea defined by the claims. [Explanation of Symbols]

[0060] 1 Polishing head 2 polishing pads 2a Polished surface 3 Polishing Table 5. Polishing fluid supply nozzle 6 Table motors 9. Operation Control Unit 10 Head Shaft 15. Polishing head rotation device 20 Optical film thickness measuring device 22 Light source 25 Optical sensor heads 27 Spectrometer 30 Processing Systems 31. Floodlight Fiber Optic Cable 32 Optical fiber receiving cable 38 Reference Spectral Library 40 Extrapolation Models 100 Polishing equipment 101 Film Thickness Measuring Device 103 Conveying device W Workpiece RW Reference Workpiece

Claims

1. A method for creating a reference spectral library used for estimating the film thickness of a workpiece, Before polishing the reference workpiece, measure the initial film thickness of the reference workpiece. The aforementioned reference workpiece is polished using a polishing device. Multiple actual reference spectra of reflected light from the reference workpiece are generated at multiple actual polishing times during the polishing of the reference workpiece. After polishing the reference workpiece, the final film thickness of the reference workpiece is measured. The polishing rate is calculated by dividing the difference between the initial film thickness and the final film thickness by the time from the start to the end of polishing of the reference workpiece. A pre-polishing reference spectrum corresponding to the time before polishing begins on the reference workpiece, and a post-polishing reference spectrum corresponding to the time after polishing ends on the reference workpiece, are generated by extrapolation. Based on the polishing rate, the following multiple reference film thicknesses corresponding to the pre-polishing reference spectrum, the multiple actual reference spectra, and the post-polishing reference spectrum are calculated. A method for creating a reference spectrum library, comprising associating the multiple reference film thicknesses with the pre-polishing reference spectrum, the multiple actual reference spectra, and the post-polishing reference spectrum, respectively, to create a reference spectrum library.

2. The method for creating a reference spectrum library according to claim 1, wherein the pre-polishing reference spectrum and the post-polishing reference spectrum are generated by an extrapolation model.

3. The pre-polishing reference spectrum is output from the extrapolation model by inputting the pre-polishing time into the extrapolation model. The method for creating a reference spectrum library according to claim 2, wherein the post-polishing reference spectrum is output from the extrapolation model by inputting the post-polishing time into the extrapolation model.

4. The method for creating a reference spectrum library according to claim 2, wherein the extrapolation model is a trained model constructed by machine learning using training data including the plurality of actual reference spectra and the plurality of actual polishing times.

5. The method for creating a reference spectral library according to claim 4, wherein the machine learning is long-term short-term memory (LSTM).

6. The method for creating a reference spectrum library according to claim 2, wherein the extrapolation model is a statistical model that analyzes the plurality of actual reference spectra and the plurality of actual polishing times according to a statistical algorithm and generates the pre-polishing reference spectrum and the post-polishing reference spectrum based on the analysis results.

Images