Wafer film thickness measuring method and device and chemical mechanical polishing equipment

Abstract

The invention provides a wafer film thickness measuring method and device and chemical mechanical polishing equipment, the wafer film thickness measuring method comprises the following steps: a measuring spectrum and an interference spectrum are obtained, the measuring spectrum is formed by measuring light, and the interference spectrum is formed by reflected light after the measuring light is reflected by a wafer; calculating a correction coefficient according to the measured spectrum, and correcting the interference spectrum according to the correction coefficient; calculating a reflectivity curve of the wafer according to the corrected interference spectrum; and determining the film thickness of the wafer according to the reflectivity curve. According to the wafer film thickness measurement method, the influence of measurement noise on an interference spectrum is eliminated, the control precision of the wafer surface quality is improved, and the chip manufacturing yield is further improved.

Description

Technical Field

[0001] This invention belongs to the field of chemical mechanical polishing technology and is used for processing semiconductor chips. Specifically, it relates to a wafer film thickness measurement method, apparatus, and chemical mechanical polishing equipment. Background Technology

[0002] Wafer manufacturing is a crucial link in the development of the ultra-large-scale integrated circuit (IC) industry. As Moore's Law continues to shrink, the feature size of integrated circuits continues to approach its theoretical limit, leading to increasingly stringent requirements for wafer surface quality. Consequently, the control of defect size and quantity in the wafer manufacturing process is becoming increasingly rigorous. Chemical mechanical polishing (CMP) is a global surface planarization technique used in semiconductor manufacturing to reduce the impact of wafer thickness variations and surface morphology. Because CMP can precisely and uniformly planarize wafers to the required thickness and flatness, it has become the most widely used surface planarization technique in semiconductor manufacturing.

[0003] Chemical mechanical polishing (CMP) involves pressing a wafer against a polishing pad using a support head. Polishing is achieved through the relative motion between the wafer and the pad, aided by abrasive particles in the polishing slurry. Some polishing processes require the removal of non-metallic layers. For controlling the polishing process of non-metallic layers, optical metrology equipment is often used to measure the thickness of the non-metallic layer. Specifically, the reflectivity curve of the surface is calculated by measuring the light intensity received by an optical fiber. This curve is then compared with the theoretical reflectivity curve (calculated using Fresnel's formula) to deduce the material thickness during the polishing process.

[0004] Controlling the polishing endpoint is crucial in chemical mechanical polishing (CMP) processes. This involves determining whether the film has been planarized to the desired flatness or thickness, or identifying when the required amount of material has been removed. However, the intensity of the light source and environmental conditions are inherently limited and cannot be kept constant. This can introduce systematic errors into optical measurement equipment, interfering with measurement results and affecting the control of the polishing endpoint. Ultimately, this can lead to wafer thickness and uniformity failing to meet the surface processing requirements of the next step, impacting the performance of semiconductor devices. Summary of the Invention

[0005] In view of the above, the present invention provides a wafer film thickness measurement method, apparatus and chemical mechanical polishing equipment, thereby solving or at least alleviating one or more of the above-mentioned problems and other problems existing in the prior art.

[0006] A first aspect of the present invention provides a method for measuring wafer film thickness, comprising:

[0007] The measurement spectrum and interference spectrum are acquired. The measurement spectrum is formed by the measurement light, and the interference spectrum is formed by the reflected light after the measurement light is reflected from the wafer.

[0008] Calculate the correction coefficient based on the measured spectrum, and correct the interference spectrum based on the correction coefficient;

[0009] The reflectance curve of the wafer is calculated based on the corrected interference spectrum;

[0010] The film thickness of the wafer is determined based on the reflectivity curve.

[0011] Optionally, the step of calculating the correction coefficient based on the measured spectrum and correcting the interference spectrum based on the correction coefficient includes:

[0012] Calculate the correction coefficient for each frame of the measured spectrum;

[0013] The interference spectrum is corrected frame by frame based on the correction coefficients for each frame of the measured spectrum.

[0014] Optionally, calculating the correction coefficient for each frame of the measured spectrum includes:

[0015] Calculate the light intensity baseline of the measured spectrum, where the light intensity baseline represents the mean light intensity of the measured spectrum;

[0016] Calculate the average light intensity for each frame of the measured spectrum;

[0017] The correction coefficient for each frame is determined based on the average light intensity and light intensity baseline of each frame.

[0018] Optionally, the step of calculating the correction coefficient based on the measured spectrum and correcting the interference spectrum based on the correction coefficient includes:

[0019] Calculate the correction coefficient for each wavelength point of the interference spectrum based on the measured spectrum;

[0020] The interference spectrum is corrected wavelength by wavelength using the correction coefficient at each wavelength point.

[0021] Optionally, calculating the correction coefficient for each wavelength point of the interference spectrum based on the measured spectrum includes:

[0022] The light intensity-wavelength function is obtained by performing a first-order or second-order fitting based on the light intensity value and wavelength point of the measured spectrum.

[0023] The fitted light intensity value corresponding to each wavelength point of the interference spectrum is determined based on the light intensity-wavelength function;

[0024] The correction coefficient for each wavelength point of the interference spectrum is determined based on the measured light intensity value and the fitted light intensity value corresponding to each wavelength point of the interference spectrum.

[0025] Optionally, acquiring the measurement spectrum and the interference spectrum includes:

[0026] Determine the wavelength offset of the measured spectrum;

[0027] The wavelength points of the measured spectrum and the wavelength points of the reflected spectrum are corrected according to the wavelength offset, respectively.

[0028] The measurement spectrum and interference spectrum after wavelength point correction are used as the measurement spectrum and interference spectrum, respectively.

[0029] Optionally, determining the wavelength shift of the measured spectrum includes:

[0030] At least two standard characteristic peaks are determined based on the optical properties of the measured light;

[0031] Gaussian fitting is performed on the measured spectrum to obtain at least two corresponding measurement characteristic peaks;

[0032] The wavelength offset of at least two measured characteristic peaks is determined based on at least two standard characteristic peaks and the corresponding at least two actual characteristic peaks.

[0033] Optionally, correcting the wavelength points of the measured spectrum and the reflection spectrum according to the wavelength offset includes:

[0034] The wavelength offset-wavelength function is obtained by performing a first-order fit based on the wavelength offset of the at least two measured characteristic peaks and the corresponding wavelengths.

[0035] The correction amount for each wavelength point of the measured spectrum is determined based on the wavelength offset-wavelength function, so as to correct each wavelength point of the measured spectrum.

[0036] The correction amount for each wavelength point of the interference spectrum is determined based on the wavelength offset-wavelength function, so as to correct each wavelength point of the interference spectrum.

[0037] Optionally, determining the wafer film thickness based on the reflectivity curve includes:

[0038] Based on the reflectance curve, the closest calibration spectrum is selected from the pre-calibrated calibration spectra corresponding to different film thicknesses;

[0039] The film thickness corresponding to the closest calibration spectrum is determined as the film thickness of the wafer.

[0040] A second aspect of the present invention provides a wafer film thickness measurement apparatus for performing the film thickness measurement method as described in the first aspect, comprising:

[0041] A light source for generating the measurement light;

[0042] The probe is used to emit the measuring light and receive the reflected light;

[0043] A first sensor is used to receive the measurement light and generate the measurement spectrum;

[0044] The second sensor is used to receive the reflected light and generate the interference spectrum;

[0045] The measured spectrum is used to calibrate the interference spectrum.

[0046] A third aspect of the present invention provides a chemical mechanical polishing apparatus, comprising: a wafer processing device, a controller, and a wafer film thickness measuring device as described in the second aspect;

[0047] While controlling the wafer processing apparatus to process the wafer surface, the controller controls the wafer film thickness measurement apparatus to perform the wafer film thickness measurement method as described in the first aspect to obtain the wafer film thickness information, and controls the wafer processing apparatus based on the wafer film thickness information.

[0048] The present invention has the following technical effects: The wafer film thickness measurement method of the present invention eliminates the influence of measurement noise such as light source fluctuation, temperature change and other external interference on the interference spectrum, improves the stability and accuracy of wafer film thickness measurement, thereby improving the control accuracy of wafer surface quality and thus improving the chip manufacturing yield. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0050] Figure 1 This is a schematic diagram of a chemical mechanical polishing (CMP) device.

[0051] Figure 2 This is a schematic diagram of a chemical mechanical polishing (CMP) device.

[0052] Figure 3 This is a schematic diagram of the wafer film thickness measuring device of the present invention.

[0053] Figure 4-6 This is a flowchart of the wafer film thickness measurement method of the present invention.

[0054] Figure 7 A schematic diagram of the spectral curve for frame-by-frame correction.

[0055] Figure 8 This is a schematic diagram of the reflectivity curve for wavelength-by-wavelength correction.

[0056] Figure 9 This is a schematic diagram showing the horizontal shift between the standard spectrum and the measured spectrum of a xenon lamp.

[0057] Figure 10 This is a schematic diagram of the standard spectrum and the corrected measured spectrum of a xenon lamp.

[0058] Reference numerals: bearing head 10; polishing disc 20; polishing pad 21; dressing device 30; liquid supply unit 40; optical measuring device 50; light source 51; probe 52; sensor 53; first sensor 531; second sensor 532; first optical path 541; second optical path 542. Detailed Implementation

[0059] To enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art should fall within the protection scope of the present invention.

[0060] In the description of this invention, it should be understood that the terms "longitudinal", "lateral", "up", "down", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0061] In addition, in the description of this invention, unless otherwise specified and limited, it should be noted that the terms "installation", "connection" and "linking" should be interpreted broadly. For example, they can refer to mechanical or electrical connections, or internal connections between two components. They can be direct connections or indirect connections through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms according to the specific circumstances.

[0062] In this invention, chemical mechanical polishing is also called chemical mechanical planarization, and wafer is also called wafer, silicon wafer, substrate, etc., with the same meaning and actual function.

[0063] This invention provides a solution for online measurement of non-metallic film thickness on a wafer surface during chemical mechanical polishing (CMP). This solution is suitable for monitoring the polishing endpoint using optical detection methods. Optical detection tracks the process progress by detecting changes in the reflectance spectrum of the wafer surface, thereby determining the endpoint of the polishing process. As a non-contact measurement method, optical detection can measure changes in the film material during the polishing process without damaging the wafer surface film.

[0064] It should be noted that, in this invention, wafer film thickness refers to the thickness of the non-metallic film layer on the wafer after thin film deposition, especially the real-time thickness of the non-metallic film layer during the removal of the non-metallic film layer, or it can refer to the thickness of the bare wafer.

[0065] like Figure 1 As shown, the chemical mechanical polishing equipment includes a support head 10 for holding and rotating the wafer, a polishing disk 20 covered with a polishing pad 21, a dresser 30 for dressing the polishing pad 21, and a liquid supply unit 40 for supplying polishing fluid.

[0066] During chemical mechanical polishing (CMP), the support head 10 presses the wafer onto the polishing pad 21 covering the surface of the polishing disk 20. The support head 10 rotates and reciprocates radially along the polishing disk 20, gradually removing impurities from the wafer surface in contact with the polishing pad 21. Simultaneously, the polishing disk 20 rotates, and the liquid supply unit 40 sprays polishing slurry onto the surface of the polishing pad 21. Under the chemical action of the polishing slurry, the relative movement between the support head 10 and the polishing disk 20 causes the wafer to rub against the polishing pad 21 for polishing. During polishing, a dresser 30 is used to trim and activate the surface morphology of the polishing pad 21. The dresser 30 can remove impurity particles remaining on the surface of the polishing pad 21, such as abrasive particles in the polishing slurry and waste material detached from the wafer surface, and can also smooth out surface deformations of the polishing pad 21 caused by abrasion.

[0067] like Figure 2 As shown, the chemical mechanical polishing (CMP) equipment also includes an optical measurement device 50. The optical measurement device 50 is positioned below the surface of the polishing disk 20 and rotates with the polishing disk 20 to achieve online measurement during polishing. The polishing pad 21 has a light-transmitting opening that penetrates the polishing pad 21, thus providing an optical path through it. The optical measurement device 50 includes a light source 51, a probe 52, and a sensor 53. Light generated by the light source 51 is emitted through the probe 52 and illuminates the wafer surface on the polishing pad 21 through the opening. The probe 52 simultaneously receives the reflected light from the wafer surface and transmits it to the sensor 53, thereby determining the non-metallic film thickness of the wafer based on the spectrum of the reflected light.

[0068] During wafer polishing, the carrier head 10 presses the wafer onto the polishing pad 21 and moves the carrier head 10 back and forth radially along the polishing disk 20. The optical measurement device 50 rotates with the polishing disk 20, so the position of the sampling point measured by the optical measurement device 50 on the wafer changes continuously, thereby enabling the acquisition of optical measurement signals at different radial positions of the wafer.

[0069] In CMP, xenon lamps are commonly used as the light source. However, the intensity of the xenon lamp changes between measurements. Factors such as temperature, ambient light, and sensor aging also introduce systematic errors, affecting the accuracy of reflectance spectrum film thickness calculation. This impact is becoming increasingly severe in affecting chip yield as semiconductor chip manufacturing processes improve.

[0070] To solve related technical problems, such as Figure 3 As shown, in one embodiment of the wafer film thickness measurement device of the present invention, the sensor 53 is configured to include a first sensor 531 and a second sensor 532. A first optical path 541 connects the light source 51 and the first sensor 531, as well as the light source 51 and the probe 52. A second optical path 542 is provided between the probe 52 and the second sensor 532. That is, the addition of the first sensor 531, which is connected to the light source 51, is used to calibrate the spectrum of the reflected light, thereby eliminating systematic errors introduced by intensity variations of the light source 51 and other factors.

[0071] During wafer polishing, light source 51 generates a light beam for measuring wafer film thickness. The light beam is simultaneously transmitted to probe 52 and first sensor 531 via first optical path 541. Probe 52 illuminates the wafer surface with the measurement light, and first sensor 531 generates a measurement spectrum based on the received measurement light. .

[0072] When measurement light shines on the wafer surface, part of the light is directly reflected from the upper surface of the wafer, while the other part is refracted by the upper surface and reflected from the lower surface of the wafer or the lower surface of the non-metallic film. The reflected light from the upper and lower surfaces of the wafer is received by the second sensor 532 via the probe 52 and the second optical path 542. The two parts of the reflected light interfere due to the optical path difference, and the second sensor 532 generates an interference spectrum based on the received reflected light. .

[0073] Optionally, the first optical path 541 and / or the second optical path 542 are integrated from optical components such as collimating lens group, focusing lens group and optical fiber to achieve efficient light transmission and meet measurement requirements.

[0074] Optionally, the probe 52 is mounted near the opening of the polishing pad 21. An ultraviolet fusion quartz window 22 can be provided at the opening to accommodate broadband beam transmission and resist contamination in the CMP environment (such as slurry splashes) to protect the internal optical components.

[0075] Optionally, sensor 53 can be one of a spectrometer, a photoelectric sensor, a power meter, or a camera. Preferably, the first sensor 531 and the second sensor 532 are different channels of the same spectrometer.

[0076] Optionally, the optical measurement device 50 also includes a control unit 55, which controls the light source 51 to emit light, controls the sensor 53 to receive optical signals, and performs functions such as uploading and synchronization.

[0077] Control unit 55 can independently perform tasks based on the measured spectrum Calibration Interference Spectrum This ensures that the interference spectrum uploaded to the host computer is the calibrated interference spectrum, and also allows for the measurement of the spectrum. and interference spectrum The data is uploaded to the host computer, which then performs the calibration operation.

[0078] On the other hand, based on the wafer film thickness measurement device of the present invention, one embodiment of the present invention also provides a wafer film thickness measurement method. The wafer film thickness measurement device of the present invention is used to execute the wafer film thickness measurement method to achieve a specific measurement spectrum-based method. Calibration Interference Spectrum The process.

[0079] like Figure 4 As shown, the methods for measuring the thickness of a circular film include:

[0080] S1. Obtain the measurement spectrum and interference spectrum Spectrum measurement Interference spectrum formed by measuring light It is formed by the reflected light after the measurement light is reflected by the wafer;

[0081] S2. Calculate the correction coefficient based on the measured spectrum, and correct the interference spectrum based on the correction coefficient. ;

[0082] S3. Based on the corrected interference spectrum Calculate the reflectivity curve R of the wafer;

[0083] S4. Determine the film thickness of the wafer based on the reflectivity curve R.

[0084] In step S1, the light beam emitted by the light source 51 is collimated, focused, and filtered by the first optical path 541 to form measurement light that illuminates the wafer surface. The first sensor 531 receives the measurement light and generates a measurement spectrum. The second sensor 532 receives the reflected light after it has been reflected by the wafer and generates an interference spectrum. Spectrum measurement and interference spectrum Complete synchronization is affected by measurement noise interference, which includes light source fluctuations, temperature changes, and other external interferences. Therefore, in step S2, the measurement spectrum can be used to determine the appropriate parameters. Correction coefficients were calculated to characterize the effect of measurement noise on the measurement spectrum during measurement. The influence of this, and its application to the interference spectrum Correction, thereby eliminating measurement noise during measurement of the interference spectrum. The impact.

[0085] In step S3, the reflectivity curve R of the wafer is calculated according to the following formula:

[0086] ,

[0087] in, It is the light intensity of the dark spectrum. and These are the standard light intensity and standard reflectance curves of the standard sample (calculated using Fresnel formula modeling).

[0088] In step S4, multiple pre-calibrated calibration spectra corresponding to different film thicknesses are used, and the calibration reflectance curve for each calibration spectrum is calculated. The calibration reflectance curve that is closest to the calculated reflectance curve R is selected from the multiple calibration reflectance curves, and the corresponding film thickness is determined as the wafer thickness.

[0089] In one embodiment, the correction coefficients in step S2 are correction coefficients related to the spectral frame, and the interference spectrum is corrected using a frame-by-frame correction method. S2 includes the following steps:

[0090] S21. Calculate the correction coefficient for each frame of the measured spectrum;

[0091] S22. Correct the interference spectrum frame by frame according to the correction coefficient of each frame of the measured spectrum.

[0092] Interference spectrum The light intensity of the i-th frame is represented as Corrected interference spectrum The light intensity of the i-th frame is represented as The correction coefficient for the i-th frame of the measured spectrum is The correction formula is:

[0093] .

[0094] Optional, such as Figure 5 As shown, step S21 specifically includes:

[0095] S211. Calculate the light intensity baseline bsl of the measured spectrum, where bsl represents the mean light intensity of the measured spectrum.

[0096] Optionally, the first 500 frames of data from the measured spectrum are used, and the light intensity baseline bsl is calculated by averaging the data. The calculation formula is as follows:

[0097] Where i represents the measured spectrum of the i-th frame, and j represents the light intensity value of the j-th frame of the spectrum. The value represents the intensity of the j-th light intensity in the i-th frame of the measured spectrum, and M represents the total number of light intensities in one frame.

[0098] S212. Calculate the average light intensity for each frame of the measured spectrum. The calculation formula is: .

[0099] S213. Determine the correction coefficient for each frame based on the average light intensity and light intensity baseline of each frame. The calculation formula is:

[0100] .

[0101] Figure 7 A schematic diagram of the spectral curves before and after frame-by-frame correction is shown. The gray curve represents the measured spectrum of a specific frame, and the green curve represents the reference spectrum of the previous 500 frames. It can be seen that the measured spectrum is shifted relative to the reference spectrum due to measurement noise. The blue curve represents the interference spectrum of the corresponding frame. The spectrum after calculating the correction coefficient for that frame and correcting it corresponds to the orange curve. It can be seen that the shift direction of the orange curve relative to the blue curve is consistent with the shift direction and degree of the green curve relative to the gray curve. This indicates that frame-by-frame correction effectively eliminates or reduces the influence of measurement noise on the interference spectrum.

[0102] In another embodiment, the correction coefficient in step S2 is a wavelength-dependent correction coefficient, and the interference spectrum is corrected using a wavelength-by-wavelength correction method. S2 includes the following steps:

[0103] S'21. Calculate the correction coefficient for each wavelength point of the interference spectrum based on the measured spectrum;

[0104] S'22. Correct the interference spectrum wavelength by wavelength according to the correction coefficient at each wavelength point.

[0105] Interference spectrum The wavelength and light intensity at the k-th wavelength point are respectively expressed as: and Corrected interference spectrum The light intensity at the k-th wavelength point is expressed as The correction coefficient for the k-th wavelength point is calculated based on the measured spectrum. The correction formula is:

[0106] .

[0107] Optional, such as Figure 6 As shown, step S'21 specifically includes:

[0108] S'211. Obtain the light intensity-wavelength function by performing a first-order or second-order fitting based on the light intensity value and wavelength point of the measured spectrum.

[0109] The wavelengths of the measured spectrum and the interference spectrum typically do not correspond, therefore the correction factor for each wavelength of the interference spectrum cannot be directly calculated based on the wavelengths of the measured spectrum and the light intensity value. Assuming the measured spectrum and the interference spectrum have m and n wavelength points respectively, the wavelength points and light intensity values ​​of the measured spectrum can be expressed as follows: and The wavelength points and light intensity values ​​of the interference spectrum can be expressed as follows: and .

[0110] The set of coordinate points consisting of each wavelength point and light intensity value in the measured spectrum ( By performing a first-order or second-order fit, the light intensity-wavelength function can be obtained.

[0111] S'212. Determine the fitted light intensity value corresponding to the wavelength point of each interference spectrum based on the light intensity-wavelength function.

[0112] The independent variable of the intensity-wavelength function is the wavelength point; substitute this into the wavelength point of the interference spectrum. This allows us to calculate the fitted light intensity value corresponding to each wavelength point in the interference spectrum. The fitted light intensity value is expressed as:

[0113]

[0114] S'213. Determine the correction coefficient for each wavelength point of the interference spectrum based on the measured light intensity value and the fitted light intensity value corresponding to each wavelength point of the measured spectrum.

[0115] For each of the n wavelength points in the interference spectrum, the measured spectrum also has n corresponding fitted light intensity values, from which the correction coefficient for the k-th wavelength point can be calculated. The calculation formula is:

[0116] ,in Determined by measuring the spectrum and / or by the first sensor 531.

[0117] Figure 8The diagram illustrates the reflectance curves calculated without wavelength-wise correction and after wavelength-wise correction. The red dashed line represents the wafer's reflectance curve R calculated from the interference spectrum without wavelength-wise correction, while the blue solid line represents the wafer's reflectance curve R calculated from the interference spectrum after wavelength-wise correction. It can be seen that the red dashed line exhibits "sawtooth" fluctuations due to measurement noise, which affects data analysis. The blue solid line is smoother, indicating that frame-wise correction effectively eliminates or reduces the impact of measurement noise on the interference spectrum, thus facilitating subsequent analysis.

[0118] In the two embodiments described above, interference spectra are corrected by measuring the spectrum through frame-by-frame correction and wavelength-by-wavelength correction, respectively. In another embodiment, the correction methods of both embodiments can be used simultaneously to correct the interference spectrum. Optionally, one of the correction methods can be selected based on the correction requirements, such as different requirements for correction response time or correction accuracy. Specifically, frame-by-frame correction is a faster correction method and can be used in scenarios requiring short response times; wavelength-by-wavelength correction is a more accurate correction method, but requires a longer correction time and needs to continuously acquire the measured spectrum as a correction reference, and can be used in scenarios requiring high correction accuracy.

[0119] The embodiments of the present invention achieve the technical effect of eliminating the influence of measurement noise such as light source fluctuations, temperature changes and other external interferences on the interference spectrum through frame-by-frame correction and / or wavelength-by-wavelength correction, thereby improving the stability and accuracy of wafer film thickness measurement, thus improving the control accuracy of wafer surface quality, and further improving chip manufacturing yield.

[0120] In the actual polishing process, the interference spectrum is not only affected by measurement noise, but also by horizontal shift. That is, the spectrum determined by the optical characteristics of the light source will shift in the horizontal direction, which also leads to inaccurate film thickness measurement results. Figure 9 The figure shows a comparison between the standard spectrum and the measured spectrum of a xenon lamp. The blue dots represent the standard spectrum, and the yellow dots represent the measured spectrum. There is a significant horizontal shift between the two, but the spectral trends are basically the same when viewed from a local perspective.

[0121] In one embodiment, step S1 further includes calibrating the horizontal offset of the measured spectrum and the interference spectrum, including the following specific steps:

[0122] S11. Determine the wavelength offset of the measured spectrum. The wavelength offset can be determined using conventional methods such as statistical values ​​and empirical values.

[0123] S12. Correct the wavelength points of the measured spectrum and the reflection spectrum respectively according to the wavelength offset.

[0124] S13. The measurement spectrum and interference spectrum after wavelength point correction are used as the measurement spectrum and interference spectrum, respectively.

[0125] Figure 10 The figure shows a comparison between the standard spectrum of a xenon lamp and the measured spectrum after fixed-value compensation correction. The wavelength shift was determined empirically to be -0.97 nm. The yellow dots represent the corrected measured spectrum. Figure 9 The measured spectrum is closer to the standard spectrum, which improves the accuracy of film thickness measurement results.

[0126] Optionally, to improve the accuracy of the correction for the wavelength points of the measured spectrum and the reflection spectrum, step S11 further includes the following specific steps:

[0127] S111. Determine at least two standard characteristic peaks based on the optical properties of the measured light, i.e., the optical properties of the light source.

[0128] S112. Perform Gaussian fitting on the measured spectrum according to the Gaussian peak-finding formula to obtain at least two corresponding characteristic peaks. The Gaussian peak-finding formula is as follows: ,in, For amplitude, The center of the peak The standard deviation is denoted as .

[0129] S113. Determine the wavelength offset of at least two measured characteristic peaks based on at least two standard characteristic peaks and the corresponding at least two actual characteristic peaks.

[0130] Taking a xenon lamp as a specific example, the spectrum of a xenon lamp has a wavelength of and The wavelengths of the two standard characteristic peaks in the interference spectrum are obtained by using the Gaussian peak-finding formula. and Thus, the wavelength shifts of the two measured characteristic peaks in the interference spectrum were calculated as follows: and .

[0131] Compared to wavelength offsets determined based on empirical values, wavelength offsets determined based on measured characteristic peaks are more precise and can provide higher correction accuracy, thus making film thickness measurement results more accurate.

[0132] Preferably, in order to further improve the correction accuracy of the wavelength points of the measured spectrum and the reflection spectrum, step S12 further includes the following specific steps:

[0133] S121. A wavelength offset-wavelength function is obtained by performing a first-order fit based on the wavelength offsets and corresponding wavelengths of at least two measured characteristic peaks, expressed as follows: Continuing with the xenon lamp as a specific example, based on the wavelength shift of the two measured characteristic peaks... and and the corresponding wavelength and A first-order fit is performed to obtain the wavelength offset-wavelength function.

[0134] S122. Determine the correction amount for each wavelength point of the measured spectrum based on the wavelength offset-wavelength function, so as to correct each wavelength point of the measured spectrum. That is, correct the wavelength value of the i-th wavelength point of the measured spectrum. Substitute the wavelength offset into the wavelength function to calculate the wavelength offset at that wavelength point. .

[0135] S123. Determine the correction amount for each wavelength point of the interference spectrum based on the wavelength offset-wavelength function, so as to correct each wavelength point of the interference spectrum. That is, correct the wavelength value of each wavelength point of the interference spectrum. Substitute the wavelength offset into the wavelength function to calculate the wavelength offset at that wavelength point. .

[0136] By making targeted corrections to each wavelength point of the measured spectrum and the interference spectrum, the correction accuracy can be further improved, thereby further improving the accuracy of the film thickness measurement results.

[0137] On the other hand, based on the wafer film thickness measuring device and wafer film thickness measuring method of the present invention, an embodiment of the present invention also provides a chemical mechanical polishing device, including: a wafer processing device, a controller, and the wafer film thickness measuring device provided in the embodiment of the present invention.

[0138] The wafer processing apparatus is used to perform a chemical mechanical polishing process on a wafer, including a carrier head 10 for holding and rotating the wafer, a polishing disk 20 covered with a polishing pad 21, a dresser 30 for dressing the polishing pad 21, and a liquid supply unit 40 for supplying polishing fluid.

[0139] The wafer thickness measurement device is positioned below the surface of the polishing disk 20. A controller is electrically connected to both the wafer thickness measurement device and the wafer processing device. While controlling the wafer processing device to process the wafer surface, the controller also controls the wafer thickness measurement device to execute the wafer thickness measurement method provided in this embodiment of the invention to obtain wafer thickness information, and further controls the wafer processing device based on the wafer thickness information. Control of the wafer processing device includes, but is not limited to, continuing to process the wafer according to the original polishing formula, adjusting the polishing formula, and / or ending polishing.

[0140] The above embodiments are only used to illustrate the embodiments of the present invention, and are not intended to limit the embodiments of the present invention. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the embodiments of the present invention. Therefore, all equivalent technical solutions also fall within the scope of the embodiments of the present invention, and the patent protection scope of the embodiments of the present invention should be defined by the claims.

Claims

1. A method for measuring wafer film thickness, characterized in that, include: The measurement spectrum and interference spectrum are acquired. The measurement spectrum is formed by the measurement light, and the interference spectrum is formed by the reflected light after the measurement light is reflected from the wafer. Calculate the correction coefficient based on the measured spectrum, and correct the interference spectrum based on the correction coefficient; The reflectance curve of the wafer is calculated based on the corrected interference spectrum; The film thickness of the wafer is determined based on the reflectivity curve.

2. The film thickness measurement method as described in claim 1, characterized in that, The step of calculating the correction coefficient based on the measured spectrum and correcting the interference spectrum based on the correction coefficient includes: Calculate the correction coefficient for each frame of the measured spectrum; The interference spectrum is corrected frame by frame based on the correction coefficients for each frame of the measured spectrum.

3. The film thickness measurement method as described in claim 2, characterized in that, The calculation of the correction coefficient for each frame of the measured spectrum includes: Calculate the light intensity baseline of the measured spectrum, where the light intensity baseline represents the mean light intensity of the measured spectrum; Calculate the average light intensity for each frame of the measured spectrum; The correction coefficient for each frame is determined based on the average light intensity and light intensity baseline of each frame.

4. The film thickness measurement method as described in claim 1, characterized in that, The step of calculating the correction coefficient based on the measured spectrum and correcting the interference spectrum based on the correction coefficient includes: Calculate the correction coefficient for each wavelength point of the interference spectrum based on the measured spectrum; The interference spectrum is corrected wavelength by wavelength based on the correction coefficient at each wavelength point of the interference spectrum.

5. The film thickness measurement method as described in claim 4, characterized in that, The calculation of the correction coefficient for each wavelength point of the interference spectrum based on the measured spectrum includes: The light intensity-wavelength function is obtained by performing a first-order or second-order fitting based on the light intensity value and wavelength point of the measured spectrum. The fitted light intensity value corresponding to each wavelength point of the interference spectrum is determined based on the light intensity-wavelength function; The correction coefficient for each wavelength point of the interference spectrum is determined based on the measured light intensity value and the fitted light intensity value corresponding to each wavelength point of the interference spectrum.

6. The film thickness measurement method according to any one of claims 1-5, characterized in that, The acquisition of the measurement spectrum and interference spectrum includes: Determine the wavelength offset of the measured spectrum; The wavelength points of the measured spectrum and the wavelength points of the reflected spectrum are corrected according to the wavelength offset, respectively. The measurement spectrum and interference spectrum after wavelength point correction are used as the measurement spectrum and interference spectrum, respectively.

7. The film thickness measurement method as described in claim 6, characterized in that, Determining the wavelength shift of the measured spectrum includes: At least two standard characteristic peaks are determined based on the optical properties of the measured light; Gaussian fitting is performed on the measured spectrum to obtain at least two corresponding measurement characteristic peaks; The wavelength offset of at least two measured characteristic peaks is determined based on at least two standard characteristic peaks and the corresponding at least two actual characteristic peaks.

8. The film thickness measurement method as described in claim 7, characterized in that, Correcting the wavelength points of the measured spectrum and the reflected spectrum based on the wavelength offset includes: The wavelength offset-wavelength function is obtained by performing a first-order fit based on the wavelength offset of the at least two measured characteristic peaks and the corresponding wavelengths. The correction amount for each wavelength point of the measured spectrum is determined based on the wavelength offset-wavelength function, so as to correct each wavelength point of the measured spectrum. The correction amount for each wavelength point of the interference spectrum is determined based on the wavelength offset-wavelength function, so as to correct each wavelength point of the interference spectrum.

9. The film thickness measurement method as described in claim 1, characterized in that, The process of determining the wafer film thickness based on the reflectivity curve includes: Based on the reflectance curve, the closest calibration spectrum is selected from the pre-calibrated calibration spectra corresponding to different film thicknesses; The film thickness corresponding to the closest calibration spectrum is determined as the film thickness of the wafer.

10. A wafer film thickness measuring apparatus for performing the film thickness measuring method as described in any one of claims 1-9, characterized in that, include: A light source for generating the measurement light; The probe is used to emit the measuring light and receive the reflected light; A first sensor is used to receive the measurement light and generate the measurement spectrum; The second sensor is used to receive the reflected light and generate the interference spectrum; The measured spectrum is used to calibrate the interference spectrum.

11. A chemical mechanical polishing apparatus, characterized in that, include: Wafer processing apparatus, controller, and wafer film thickness measuring apparatus as described in claim 10; While controlling the wafer processing apparatus to process the wafer surface, the controller controls the wafer film thickness measurement apparatus to perform the wafer film thickness measurement method as described in any one of claims 1-9 to obtain wafer film thickness information, and controls the wafer processing apparatus based on the wafer film thickness information.

Images