CN118418034B - Method and related apparatus for optical endpoint detection for chemical mechanical polishing - Google Patents

Abstract

The embodiment of the application discloses an optical end point detection method and related equipment for chemical mechanical polishing, wherein the method comprises the following steps: in the chemical mechanical polishing process, acquiring collected reflection coherent spectrum data, and obtaining a reflection coherent spectrum curve according to the reflection coherent spectrum data; selecting a wavelength region on the reflection coherent spectrum curve, calibrating a characteristic extremum of the reflection coherent spectrum curve in the wavelength region, and acquiring a target coherent level; comparing whether the target coherence level is equal to the standard coherence level of a preset standard endpoint spectrum curve; if the two are equal, obtaining the difference between the reflection coherent spectrum curve and the standard endpoint spectrum curve, and carrying out optical endpoint detection according to the difference; the application can accurately and rapidly analyze whether the polishing end point is really reached, is not influenced by factors such as power supply voltage fluctuation, temperature change and the like, and has higher detection accuracy and high detection efficiency.

Description

Optical endpoint detection method and related equipment for chemical mechanical polishing

Technical Field

The present application relates to the field of equipment for manufacturing semiconductor integrated circuit chips, and in particular to an optical endpoint detection method for chemical mechanical polishing and related equipment.

Background Art

Chemical Mechanical Polishing (CMP) technology is a key technology to achieve comprehensive flattening of the silicon wafer surface, integrating online measurement, online endpoint detection, cleaning and other technologies. During the CMP process, effective endpoint detection (EPD) technology can accurately control the change in the thickness of the film on the wafer surface, avoid excessive or insufficient removal of surface materials, and ensure the flatness of the wafer surface. At present, online endpoint detection technology is mainly based on optical, electrical, acoustic or vibration, thermal, friction, chemical or electrochemical principles, and is achieved by detecting changes in parameters such as changes in the intensity of reflected light on the wafer surface, changes in the current of the drive motor, acoustic emission signals, polishing pad temperature, and ion concentration in the polishing liquid.

A common optical endpoint detection technology on the market is the monochromatic laser detection method, which mainly determines the polishing endpoint of the wafer surface by monitoring the reflected light intensity change curve of the polished surface. However, in the monochromatic laser detection method, other factors such as power supply voltage fluctuations, equipment vibrations, and temperature fluctuations will cause fluctuations in the reflected light intensity value, reducing the accuracy of optical endpoint detection.

Summary of the invention

The embodiments of the present application disclose a method and related equipment for optical endpoint detection of chemical mechanical polishing, which can eliminate the influence of factors such as power supply voltage fluctuation and temperature change on the detection results in optical endpoint detection, and improve the accuracy and efficiency of optical endpoint detection.

In a first aspect, an embodiment of the present application discloses an optical endpoint detection method for chemical mechanical polishing, the method comprising:

During the chemical mechanical polishing process, acquiring collected reflection coherence spectrum data, and obtaining a reflection coherence spectrum curve according to the reflection coherence spectrum data;

Selecting a wavelength region on the reflection coherence spectrum curve, calibrating a characteristic extreme value of the reflection coherence spectrum curve within the wavelength region, and obtaining a target coherence order, wherein the characteristic extreme value indicates an interference peak and an interference trough of the reflection coherence spectrum curve, and the target coherence order is a maximum coherence order within the wavelength region;

Comparing the target coherence level with a standard coherence level of a preset standard endpoint spectrum curve to see whether they are equal, the standard coherence level being the highest coherence level of the standard endpoint spectrum curve in the wavelength region;

If the two are equal, the difference between the reflection coherence spectrum curve and the standard endpoint spectrum curve is obtained, and optical endpoint detection is performed according to the difference.

As an optional implementation manner, in the first aspect of the embodiment of the present application, the acquiring a target coherence level includes:

Determining at least two characteristic extreme values among the characteristic extreme values calibrated within the wavelength region;

The target coherence order is obtained according to the wavelengths respectively corresponding to the at least two characteristic extreme values.

As an optional implementation manner, in the first aspect of the embodiment of the present application, obtaining the target coherence order according to the wavelengths corresponding to the at least two characteristic extreme values respectively includes:

Select any two characteristic extreme values from the at least two characteristic extreme values as first characteristic extreme values, and obtain the wavelength and calibration number corresponding to the first characteristic extreme values, where the calibration number is a serial number assigned to each characteristic extreme value in sequence;

The target coherence order is obtained according to the wavelength and calibration number corresponding to the first characteristic extreme value.

As an optional implementation manner, in the first aspect of the embodiment of the present application, the obtaining of the wavelength and calibration number corresponding to the first characteristic extreme value includes:

Obtaining the wavelength corresponding to the first characteristic extreme value;

The calibration numbers are assigned to the calibrated interference peaks and interference troughs in sequence starting from the first interference peak on the right side of the minimum wavelength in the wavelength region, and the calibration numbers corresponding to the first characteristic extreme values are obtained.

As an optional implementation manner, in the first aspect of the embodiment of the present application, obtaining the target coherence order according to the wavelengths corresponding to the at least two characteristic extreme values respectively includes:

Select two characteristic extreme values from the at least two characteristic extreme values as second characteristic extreme values, and obtain the wavelengths corresponding to the second characteristic extreme values and the number of characteristic extreme values between the second characteristic extreme values;

The target coherence order is obtained according to the wavelength and the number corresponding to the second characteristic extreme value.

As an optional implementation manner, in the first aspect of the embodiment of the present application, selecting two characteristic extreme values from the at least two characteristic extreme values as second characteristic extreme values, and obtaining the wavelengths corresponding to the second characteristic extreme values and the number of characteristic extreme values between the second characteristic extreme values, includes:

Select two characteristic extreme values indicating interference peaks from the at least two characteristic extreme values as second characteristic extreme values, and obtain the wavelengths corresponding to the second characteristic extreme values and the number of characteristic extreme values indicating interference troughs between the second characteristic extreme values; or

Select two characteristic extreme values indicating interference wave valleys from the at least two characteristic extreme values as the second characteristic extreme values, and obtain the wavelengths corresponding to the second characteristic extreme values and the number of characteristic extreme values indicating interference wave peaks between the second characteristic extreme values; or

A characteristic extreme value indicating an interference peak and a characteristic extreme value indicating an interference trough are selected from the at least two characteristic extreme values as the second characteristic extreme value, and the wavelength of the second characteristic extreme value indicating the interference peak, the wavelength of the second characteristic extreme value indicating the interference trough, and the number of characteristic extreme values indicating the interference peak between the second characteristic extreme values or the number of characteristic extreme values indicating the interference trough between the second characteristic extreme values are obtained.

As an optional implementation, in the first aspect of the embodiments of the present application, the difference is at least one of the mean square error MSE, the root mean square error RMSE, the mean absolute error MAE or the goodness of fit GOF.

As an optional implementation, in the first aspect of the embodiment of the present application, performing optical endpoint detection according to the difference includes:

If the difference meets a preset condition, the chemical mechanical polishing is stopped.

A second aspect of the embodiments of the present application discloses a chemical mechanical polishing method, wherein a sample is polished according to the optical endpoint detection method for chemical mechanical polishing described in any one of the first aspects of the embodiments of the present application until the polishing endpoint is reached.

The third aspect of the embodiment of the present application discloses an optical endpoint detection device for chemical mechanical polishing, comprising:

An acquisition module, used for acquiring the collected reflection coherence spectrum data during the chemical mechanical polishing process, and obtaining a reflection coherence spectrum curve according to the reflection coherence spectrum data;

A calibration module, used for selecting a wavelength region on the reflection coherence spectrum curve, calibrating a characteristic extreme value of the reflection coherence spectrum curve in the wavelength region, and obtaining a target coherence order, wherein the characteristic extreme value indicates an interference peak and an interference trough of the reflection coherence spectrum curve, and the target coherence order is a maximum coherence order in the wavelength region;

A judgment module, used for comparing whether the target coherence level is equal to a standard coherence level of a preset standard endpoint spectrum curve, wherein the standard coherence level is the highest coherence level of the standard endpoint spectrum curve in the wavelength region;

The detection module is used to obtain the difference between the reflection coherence spectrum curve and the standard endpoint spectrum curve when the judgment result of the judgment module is that the two are equal, and detect whether the optical endpoint of the polishing is reached according to the difference.

A fourth aspect of the embodiments of the present application discloses an electronic device, which may include:

A memory storing executable program code;

a processor coupled to the memory;

The processor calls the executable program code stored in the memory to execute an optical endpoint detection method for chemical mechanical polishing disclosed in the first aspect of the embodiment of the present application.

The fifth aspect of the embodiments of the present application discloses a computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute the optical endpoint detection method for chemical mechanical polishing disclosed in the first aspect of the embodiments of the present application.

Compared with the prior art, the embodiments of the present application have the following beneficial effects:

In an embodiment of the present application, during the chemical mechanical polishing process, collected reflection coherence spectrum data is obtained, and a reflection coherence spectrum curve is obtained based on the reflection coherence spectrum data. Furthermore, a wavelength region is selected on the reflection coherence spectrum curve, and within the wavelength region, the characteristic extreme value of the reflection coherence spectrum curve is calibrated, and a target coherence order is obtained, the characteristic extreme value indicates the interference peak and interference trough of the reflection coherence spectrum curve, and the target coherence order is the highest coherence order within the wavelength region; finally, the target coherence order is compared with the standard coherence order of the preset standard endpoint spectrum curve to see whether they are equal, and the standard coherence order is obtained. is the highest coherence order of the standard endpoint spectral curve in this wavelength region; when the two are equal, the difference between the reflection coherence spectral curve and the standard endpoint spectral curve is obtained, and optical endpoint detection is performed according to the difference; by implementing the embodiment of the present application, it is possible to analyze the reflection coherence spectral curve and determine whether the chemical mechanical polishing is close to the polishing endpoint according to the target coherence order, and then compare and analyze the difference between the reflection coherence spectral curve and the standard endpoint spectral curve to accurately and quickly analyze whether the polishing endpoint has been reached, without being affected by factors such as power supply voltage fluctuations and temperature changes, and the detection accuracy is high, and the efficiency of polishing endpoint detection is high.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

FIG1 is a schematic diagram of the application principle of an optical endpoint detection system for chemical mechanical polishing disclosed in an embodiment of the present application;

FIG2 is a schematic diagram of an application of an illumination area formed by an incident light beam on a polished surface of a wafer disclosed in an embodiment of the present application;

3 is a schematic diagram of the interference of reflected light between the oxide light-transmitting layer and the silicon substrate layer of the wafer disclosed in the embodiment of the present application;

FIG4 is a schematic diagram of a reflection coherence spectrum curve disclosed in an embodiment of the present application;

FIG5 is a schematic diagram showing the principle of an optical endpoint detection method for chemical mechanical polishing disclosed in an embodiment of the present application;

FIG6 is a schematic flow chart of an optical endpoint detection method for chemical mechanical polishing disclosed in Example 1 of the present application;

FIG7 is a schematic flow chart of an optical endpoint detection method for chemical mechanical polishing disclosed in Example 2 of the present application;

FIG8 is a schematic diagram of the determination principle of optical endpoint detection disclosed in an embodiment of the present application;

FIG9 is a schematic flow chart of an optical endpoint detection method for chemical mechanical polishing disclosed in Example 3 of the present application;

FIG10 is a schematic diagram of determining the target coherence level disclosed in an embodiment of the present application;

FIG11 is a schematic flow chart of an optical endpoint detection method for chemical mechanical polishing disclosed in Example 4 of the present application;

FIG12 is a schematic structural diagram of an optical endpoint detection device for chemical mechanical polishing disclosed in Example 1 of the present application;

FIG. 13 is a schematic diagram of the structure of an electronic device disclosed in an embodiment of the present application.

DETAILED DESCRIPTION

The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

It should be noted that the terms "first", "second", "third" and "fourth" in the specification and claims of the present application are used to distinguish different objects rather than to describe a specific order. The terms "including" and "having" in the embodiments of the present application and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device including a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

For example, please refer to Figure 1, which is a schematic diagram of the application principle of the optical endpoint detection system for chemical mechanical polishing disclosed in an embodiment of the present application. The optical endpoint detection system 10 for chemical mechanical polishing (hereinafter referred to as the optical endpoint detection system 10) shown in Figure 1 is used in conjunction with a chemical mechanical polishing platform 20.

In FIG1 , a polishing pad 40 is laid on the chemical mechanical polishing platform 20, and a wafer 30 is placed on the polishing pad 40. During the chemical mechanical polishing process, the chemical mechanical polishing platform 20 and the wafer 30 rotate separately to achieve polishing of the wafer 30 by the polishing pad 40. In addition, a detection window 201 is provided on the chemical mechanical polishing platform 20 and the polishing pad 40, and the detection window 201 is provided for use by the optical endpoint detection system 10 for optical endpoint detection.

The optical endpoint detection system 10 comprises: an incident light source 101 , an optical channel 103 , an optical analysis unit 104 , and a data acquisition and analysis unit 105 .

The incident light source 101 is used to emit a stable and continuous incident light beam 102, and the incident light beam 102 has a certain wavelength range.

The optical channel 103 is used to guide the incident light beam 102 into the detection window 201 , irradiate the polished surface of the wafer 30 to generate coherent reflected light, and the reflected light is then received by the optical channel 103 and sent to the optical analysis unit 104 .

The optical analysis unit 104 includes a reflection grating 1041 and a charge coupled device (CCD) detector 1042. The reflection grating 1041 has a strong dispersion function and performs spectroscopic processing on the reflected light, which is then received by the CCD detector.

The data acquisition and analysis unit 105 is used to complete data acquisition from the optical analysis unit 104, and can also perform optical endpoint detection of polishing and transmit the detection results to the computer, or transmit the collected data to the computer after completing data acquisition.

Further, please refer to FIG. 2 , which is a schematic diagram of the application of the illumination area formed by the incident light beam on the polished surface of the wafer disclosed in the embodiment of the present application. In conjunction with FIG. 2 , the incident light beam 102 emitted by the incident light source 101 is irradiated to the polished surface of the wafer 30 through the optical channel 103, forming an illumination area 102a on the wafer 30. Among them, the incident light source 101 can be selected but not limited to the incident form of line, point, dot matrix, etc., so that illumination areas 102a of different shapes will be obtained. In FIG. 2 , taking the incident light source 101 as a straight parallel light beam as an example, the wafer 30 rotates during polishing, and the illumination area 102a will form an annular detection area 301.

Furthermore, the reflected light reflected by the polished surface of the wafer 30 is a dynamically changing coherent characteristic spectrum, which is incident into the optical analysis unit 104 through the optical channel 103. The reflection grating 1041 in the optical analysis unit 104 is an interference grating with a strong dispersion function, which performs spectroscopic processing on the reflected light. Among them, taking the incident light source 101 as a straight parallel beam as an example, the reflected light reflected by the polished surface of the wafer 30 is directly irradiated on the reflection grating 1041 in parallel. At this time, the CCD detector 1042 can be preferably a surface array detector, which receives the linear array light signal of the coherent spectrum after the spectroscopic analysis; and for any form of incident light source 101 (including a straight incident form), if the reflected light reflected by the polished surface of the wafer 30 is first focused on the reflection grating 1041 by the focusing unit, at this time, the CCD detector 1042 can be preferably a linear array detector. The present application does not specifically limit the detection form of the CCD detector.

Furthermore, the polished surface of the wafer 30 includes an oxide light-transmitting layer (Ox light-transmitting layer) and a silicon substrate layer (Si substrate layer), please refer to FIG3, which is a schematic diagram of the interference of reflected light from the oxide light-transmitting layer and the silicon substrate layer of the wafer disclosed in the embodiment of the present application. As shown in FIG3, when the incident light beam 102 is irradiated on the polished surface of the wafer 30, two beams of reflected light will be generated. Among them, one beam is the reflected light generated at the interface between the air and the oxide light-transmitting layer, and the intensity is _IA ; the other beam is the reflected light generated at the interface between the oxide light-transmitting layer and the silicon substrate layer, and the intensity is _IB . Since the reflected light with an intensity of _{IB passes through the oxide light-transmitting layer and is reflected by the silicon substrate layer, there is an optical path difference between the reflected light with an intensity of IA} and the reflected light with an intensity of _IA . , the two reflected lights will interfere with each other, thus forming a reflection coherent spectrum. Assuming the refractive index of the oxide light-transmitting layer is n, the optical path difference formula is , then the coherence formula (1) of the reflection coherence spectrum curve satisfies:

(Formula 1)

Where d is the thickness of the oxide light-transmitting layer, is the wavelength in the reflectance coherence spectrum curve.

Among them, in the reflection coherence spectrum, when the optical path difference satisfies When the optical path difference satisfies When , the interference is destructive; where N is the coherence series, which can be integers 1, 2, 3, etc. is the wavelength of the interference peak, is the wavelength of the interference trough, and the coherence order decreases from left to right as the wavelength increases, which will be described in detail below in conjunction with FIG. 4 .

Please refer to FIG. 4, which is a schematic diagram of the reflection coherence spectrum curve disclosed in the embodiment of the present application; the reflection coherence spectrum curve shown in FIG. 4 can be obtained by fitting the acquisition and analysis unit 105 based on the collected reflection coherence spectrum data, and the collected reflection coherence spectrum data includes the intensity of the reflected light, etc. Then, according to the above-mentioned coherence formula (1), the reflection coherence spectrum curve at the start of polishing and the reflection coherence spectrum curve at the end of polishing as shown in FIG. 4 are obtained respectively. In FIG. 4, the horizontal axis is the wavelength, the unit is (nm), the vertical axis is the intensity of light (Intensity), a and b are the upper and lower limits of the wavelength range of the incident light beam 102, respectively, as shown , , ...corresponds to the wavelengths of the interference peaks when the coherence order N takes the values of 1, 2, 3, etc. Combined with the optical path difference equation of the interference peak , since the refractive index changes little with wavelength, it can be assumed to be a constant. As chemical mechanical polishing proceeds, the thickness d of the oxide light-transmitting layer decreases, and the wavelength of the interference peak of the same order N decreases synchronously, resulting in the formula , the reflected coherence spectrum curve drifts to the left, and the drift value of each interference peak is Variation of polishing thickness There is an approximate proportional relationship.

Therefore, assuming that the initial thickness of the oxide light-transmitting layer of the wafer to be polished is d ₀ , and the thickness becomes d ₂ when the polishing end point is reached, the wavelength range of the incident light beam 102 used for detection is [a, b]. During the polishing process, as the thickness decreases, the reflection coherence spectrum curve obtained by detection drifts toward the a side, as shown in FIG4 . As the spectrum curve drifts, the characteristic extreme value (peak or trough) of the reflection coherence spectrum curve on the a side will disappear. Therefore, during the polishing process, for any reflection coherence spectrum curve collected in real time, according to the coherence optical path difference formula , it can be obtained that the wavelength on the a side is smaller, the coherence order N corresponding to the interference peak closest to the a side is larger, and it decreases gradually as the wavelength increases from a to b.

As the reflection coherence spectrum curve continues to drift toward the a side, the interference peaks of higher coherence orders will continue to drift out of the wavelength range [a, b] and disappear from the a side. Therefore, according to the disappearance of the interference peaks during the polishing process, the polishing endpoint detection interval is divided into a preliminary judgment area and a precise judgment area.

Then, the initial thickness of the oxide light-transmitting layer of the wafer 30 is d ₀ , and the thickness is d ₂ when it reaches the polishing end point. When the thickness is d ₂ , the corresponding reflection coherence spectrum curve of the polished surface of the wafer 30 is the standard end point spectrum curve, and the highest coherence order of the standard end point spectrum curve in the wavelength range [a, b] is m. During the chemical mechanical polishing process, as the thickness decreases, the highest coherence order of the reflection coherence spectrum curve of the polished surface of the wafer 30 detected in real time in the wavelength range [a, b] continues to decrease. If the highest coherence order corresponding to the reflection coherence spectrum curve of the polished surface of the wafer 30 detected in real time in the wavelength range [a, b] is greater than m, it is determined that the polishing point is in the preliminary judgment area. If the highest coherence order corresponding to the reflection coherence spectrum curve of the polished surface of the wafer 30 detected in real time in the wavelength range [a, b] is equal to m, it is determined that the polishing point is in the precise judgment area.

Please further refer to Figure 5, which is a schematic diagram of the principle of the optical endpoint detection method of chemical mechanical polishing disclosed in an embodiment of the present application; in the schematic diagram taken as an example in Figure 5, the initial thickness of the oxide light-transmitting layer of the wafer 30 is _d0 , and the thickness is _d2 when reaching the polishing endpoint, and _d1 is a thickness between _d0 and _d2 . During the chemical mechanical polishing (CMP polishing) process, assuming that the highest coherence level corresponding to the polishing endpoint is 3, if the highest coherence level of the reflected coherence spectrum curve is detected to be 4 in real time, it is determined that the polishing point is in the preliminary judgment area; when the highest coherence level of the reflected coherence spectrum curve is detected to be 3 in real time, it is determined that the polishing point is in the precise judgment area. In combination with the right sub-figure of Figure 5, it can be seen that as the polishing thickness continues to decrease, the interference peak continues to drift to the left out of the range of wavelength a, where the wavelength a in Figure 5 , , , They are the wavelengths of the interference peaks corresponding to the coherence order of 1, coherence order of 2, coherence order of 3, and coherence order of 4 respectively.

In combination with the above introduction, the embodiment of the present application discloses an optical endpoint detection method and related equipment for chemical mechanical polishing, which can improve the accuracy and efficiency of optical endpoint detection of polishing. The present application will be described in detail below through specific embodiments.

Embodiment 1

Please refer to FIG. 6 , which is a schematic flow chart of an optical endpoint detection method for chemical mechanical polishing disclosed in Example 1 of the present application; as shown in FIG. 6 , the optical endpoint detection method for chemical mechanical polishing may include:

601. During the chemical mechanical polishing process, the collected reflection coherence spectrum data is obtained, and a reflection coherence spectrum curve is obtained according to the reflection coherence spectrum data.

The execution subject of the embodiment of the present application may be an optical endpoint detection device for chemical mechanical polishing (such as the data acquisition and analysis unit 105 in FIG. 1 or a device independent of the optical endpoint detection system for chemical mechanical polishing), or an independent electronic device (such as a terminal device), which is not specifically limited in the present application. The above-mentioned emission coherent spectrum data includes but is not limited to the intensity and wavelength of the reflected light.

The embodiment of the present application can be implemented in the process of chemical mechanical polishing based on the cooperation of the optical endpoint detection system 10 and the chemical mechanical polishing platform 20 as shown in FIG1. In conjunction with the optical endpoint detection system 10 of chemical mechanical polishing as shown in FIG1, in the process of polishing the wafer using the CMP technology, the incident light beam 102 emitted by the incident light source 101 first irradiates the oxide light-transmitting layer of the wafer, and generates a beam of reflected light with an intensity of _IA through the interface between the air and the oxide light-transmitting layer, and the incident light beam 102 continues to irradiate the silicon substrate layer of the wafer through the oxide light-transmitting layer, and generates another beam of reflected light with an intensity of _IB at the interface between the oxide light-transmitting layer and the silicon substrate layer. Since there is an optical path difference between the reflected light with an intensity of _IA and the reflected light with an intensity of _IB , the two reflected light beams will interfere with each other, thereby forming a reflection coherent spectrum. Further, the reflection coherent spectrum is first subjected to spectroscopic processing by the reflection grating 1041, and then received by the CCD detector, and then the data acquisition and analysis unit 105 is used to collect the reflection coherent spectrum data from the CCD detector in real time. Therefore, in the embodiment of the present application, the optical endpoint detection device or electronic device obtains the reflection coherent spectrum data collected in real time from the data collection and analysis unit 105 .

602. Select a wavelength region on the reflection coherence spectrum curve, calibrate the characteristic extreme value of the reflection coherence spectrum curve in the wavelength region, and obtain a target coherence order, wherein the characteristic extreme value indicates the interference peak and interference trough of the reflection coherence spectrum curve, and the target coherence order is the highest coherence order in the wavelength region.

In step 602, a wavelength region is selected on the reflection coherence spectrum curve, and within the wavelength region, the characteristic extreme value is calibrated, and the target coherence order within the wavelength region is obtained, wherein the characteristic extreme value indicates the interference peak and interference trough of the reflection coherence spectrum curve, and the target coherence order is the highest coherence order within the wavelength region. The wavelength region may be the wavelength range [a, b] of the incident light beam 102 emitted by the light source 101 as described above, or may be a part of the wavelength range [a, b] of the incident light beam 102. Furthermore, the wavelength region may also be the wavelength interval corresponding to the sampling data with a higher signal-to-noise ratio in the reflection coherence spectrum curve collected in real time.

Among them, the coherence order corresponding to the interference peak and/or interference trough with a larger wavelength is smaller, and the coherence order corresponding to the interference peak and/or interference trough with a smaller wavelength is larger. For details, please refer to the detailed description corresponding to the above-mentioned Figure 4.

603. Compare the target coherence level with the standard coherence level of the preset standard endpoint spectrum curve to see if they are equal, where the standard coherence level is the highest coherence level of the standard endpoint spectrum curve within a wavelength region; if the two are equal, execute step 604; if the two are not equal, end this process.

In the embodiment of the present application, the preset standard endpoint spectrum curve can be obtained by measuring a standard wafer sample in advance, or can be obtained in advance by theoretical calculation. In which, the same wavelength region is also selected on the preset standard endpoint spectrum curve, and the highest coherence order in the wavelength region is obtained as the standard coherence order.

In addition, as shown in FIG5 , a preliminary judgment area and an accurate judgment area are obtained according to the thickness of the standard wafer sample. In step 603 , the target coherence level and the standard coherence level are compared. When the two are equal, it is considered that the chemical mechanical polishing has entered the accurate judgment area, and step 604 is further executed; otherwise, the two are not equal, the process can be terminated and step 601 is executed.

604. Obtain a difference between the reflection coherence spectrum curve and the standard endpoint spectrum curve, and perform optical endpoint detection according to the difference.

The above difference is at least one of mean square error MSE, root mean square error RMSE, mean absolute error MAE or goodness of fit GOF.

Optionally, the optical endpoint detection based on the difference may include:

If the difference meets the preset condition, the chemical mechanical polishing is stopped.

The preset condition can be pre-set. When the preset condition is met, it is determined that the chemical mechanical polishing has reached the polishing endpoint, the chemical mechanical polishing is stopped, and the chemical mechanical polishing of the current wafer is completed (more implementation methods will be described in detail later). When the preset condition is not met, it is determined that the chemical mechanical polishing has not reached the polishing endpoint, and the step 601 is executed until it is detected that the polishing endpoint has been reached, and the chemical mechanical polishing is stopped.

It can be seen that in the chemical mechanical polishing process, the collected reflection coherence spectrum data is obtained, and a reflection coherence spectrum curve is obtained according to the reflection coherence spectrum data. Further, a wavelength region is selected on the reflection coherence spectrum curve, and the characteristic extreme value of the reflection coherence spectrum curve is calibrated within the wavelength region, and the target coherence order is obtained. The characteristic extreme value indicates the interference peak and interference trough of the reflection coherence spectrum curve, and the target coherence order is the highest coherence order within the wavelength region; finally, the target coherence order is compared with the standard coherence order of the preset standard endpoint spectrum curve to see whether they are equal. The standard is The coherence level is the highest coherence level of the standard endpoint spectral curve within the wavelength region; when the two are equal, the difference between the reflected coherence spectral curve and the standard endpoint spectral curve is obtained, and optical endpoint detection is performed based on the difference; by implementing the embodiments of the present application, the polishing position of the chemical mechanical polishing can be determined according to the target coherence level by analyzing the reflected coherence spectral curve, and then the difference between the reflected coherence spectral curve and the standard endpoint spectral curve can be compared and analyzed to accurately and quickly analyze whether the polishing endpoint has been reached, without being affected by factors such as power supply voltage fluctuations and temperature changes, and the detection accuracy and efficiency are high.

Embodiment 2

Please refer to FIG. 7 , which is a schematic flow chart of an optical endpoint detection method for chemical mechanical polishing disclosed in Example 2 of the present application; as shown in FIG. 7 , the optical endpoint detection method for chemical mechanical polishing may include:

701. During the chemical mechanical polishing process, the collected reflection coherence spectrum data is obtained.

The execution subject of the embodiment of the present application may be an optical endpoint detection device for chemical mechanical polishing (such as the data acquisition and analysis unit 105 in FIG. 1 or a device independent of the optical endpoint detection system for chemical mechanical polishing), or an independent electronic device (such as a terminal device), which is not specifically limited in the present application. The above-mentioned emission coherent spectrum data includes but is not limited to the intensity and wavelength of the reflected light.

The embodiment of the present application can be implemented in a chemical mechanical polishing process based on the cooperation of the optical endpoint detection system 10 and the chemical mechanical polishing platform 20 as shown in Figure 1. For more details, please refer to the description in step 601, which will not be repeated here.

702. Fit a corresponding reflection coherence spectrum curve according to the reflection coherence spectrum data.

In step 702, a corresponding reflection coherence spectrum curve is fitted according to the collected reflection coherence spectrum data.

For example, the thickness d of the oxide light-transmitting layer of the wafer and the refractive index n of the oxide light-transmitting layer are obtained, and the reflection coherence spectrum curve is obtained by fitting the above coherence formula (1).

703. Select a wavelength region on the reflection coherence spectrum curve, and calibrate a characteristic extreme value of the reflection coherence spectrum curve in the wavelength region, wherein the characteristic extreme value indicates an interference peak and an interference trough of the reflection coherence spectrum curve.

In step 703, a wavelength region is selected on the reflection coherent spectrum curve, and the characteristic extreme value is calibrated in the wavelength region. After the characteristic extreme value is calibrated, the interference peaks and interference troughs in the wavelength region can be clearly identified, and the wavelengths corresponding to the interference peaks and interference troughs, the wavelength corresponding to the first characteristic extreme value (interference peak or interference trough) in the wavelength region, and the wavelength corresponding to the last characteristic extreme value (interference peak or interference trough) can be obtained. Then, according to the wavelength corresponding to each characteristic extreme value, the number of interference peaks, the number of interference troughs, the total number of interference peaks and interference troughs, and the positional relationship between the interference peaks and interference troughs in the wavelength region can be determined.

704. Determine at least two characteristic extreme values among the characteristic extreme values calibrated within a wavelength region.

It can be understood that, in conjunction with Figure 4, after calibrating the characteristic extreme values, any characteristic extreme value can be determined from the wavelength region. Therefore, determining at least two characteristic extreme values specifically means: being able to determine at least two characteristic extreme values to be selected from all characteristic extreme values within the wavelength region, being able to obtain the wavelengths corresponding to the at least two characteristic extreme values, and clarifying the positional relationship of the at least two characteristic extreme values within the wavelength region.

705. Obtain a target coherence level according to wavelengths corresponding to at least two characteristic extreme values; the target coherence level is a maximum coherence level within a wavelength region.

By determining at least two characteristic extreme values in step 704, the wavelengths corresponding to the at least two characteristic extreme values can be obtained, and the positional relationship of the at least two characteristic extreme values among all characteristic extreme values within the wavelength region can be clarified. Then, based on the obtained information, the target coherence order within the wavelength region is obtained, that is, the highest coherence order within the wavelength region.

706. Compare the target coherence level with the standard coherence level of the preset standard endpoint spectrum curve to see if they are equal, where the standard coherence level is the highest coherence level of the standard endpoint spectrum curve in a wavelength region; if the two are equal, execute step 707; if not, go to step 701.

As shown in Figure 5, in step 706, the target coherence level and the standard coherence level are compared. When the two are equal, it is considered that the chemical mechanical polishing has entered the precise judgment area, and step 707 is further executed; otherwise, if the two are not equal, it can be considered that the chemical mechanical polishing is still in the preliminary judgment area, and the current process can be terminated and the execution of step 701 can be turned.

707. Obtain the difference between the reflection coherence spectrum curve and the standard endpoint spectrum curve. If the difference meets a preset condition, stop chemical mechanical polishing.

The above difference is at least one of mean square error MSE, root mean square error RMSE, mean absolute error MAE or goodness of fit GOF.

Optionally, if the difference meets a preset condition, stopping the chemical mechanical polishing includes:

When the difference meets the preset determination value range, it is determined that the polishing end point has been reached and the chemical mechanical polishing is stopped.

Among them, the judgment value interval can be calculated in advance based on the standard endpoint spectrum curve, or obtained through theoretical calculation, and can accurately detect whether the polishing endpoint is reached through the intensity difference value obtained in real time, so as to stop the chemical mechanical polishing in time when the polishing endpoint is reached, thereby improving the accuracy of chemical mechanical polishing.

When the difference is the mean square error MSE, for example, please refer to FIG. 8 , which is a schematic diagram of the determination principle of the optical endpoint detection disclosed in the embodiment of the present application; in FIG. 8 , the solid line is the standard endpoint spectrum curve, and the black dots are the target data points actually measured from the reflection coherence spectrum curve ( ）、（ ）、（ ）、…、（ )…; correspondingly, select the standard data point from the standard endpoint spectrum curve ( ）、（ ）、（ ）、…、（ )….for The target data point and standard data point corresponding to _i are ,get indivual , then according to , calculate the intensity difference value S, that is, the mean square error MSE, is the number of target data points, and i is a positive integer.

Through the above implementation, the MSE of the reflection coherence spectrum curve and the standard endpoint spectrum curve is calculated to improve the accuracy of optical endpoint detection. If the MSE meets the preset judgment value interval, it is determined that the polishing endpoint has been reached and the chemical mechanical polishing is stopped.

Alternatively, in another embodiment of the present application, when the difference is the goodness of fit GOF, the difference between the reflection coherence spectrum curve and the standard endpoint spectrum curve is obtained, including:

Collecting at least one spectral curve segment corresponding to the target sub-range in the wavelength region of the reflectance coherence spectral curve, and collecting at least one standard curve segment corresponding to the standard sub-range in the wavelength region of the standard endpoint spectral curve;

Each spectral curve segment is fitted with the corresponding standard curve segment to obtain the goodness of fit value.

In the above embodiment, one or more spectral curve segments corresponding to the target sub-range can be taken in the selected wavelength region. Similarly, a corresponding standard curve segment is also taken in the standard endpoint spectral curve. The spectral curve segment and the corresponding standard curve segment are fitted to obtain a goodness of fit value. Based on at least one spectral curve segment, at least one goodness of fit value is obtained.

Furthermore, if the difference meets the preset condition, the chemical mechanical polishing is stopped, including:

It is determined whether all goodness of fit values are greater than a threshold value. If all are greater than the threshold value, it is determined that the polishing end point has been reached and the chemical mechanical polishing is stopped; if at least one goodness of fit value is less than or equal to the threshold value, it is determined that the polishing end point has not been reached and the process turns to step 701. The threshold value can be set in advance based on an empirical value.

Furthermore, when the number of spectral curve segments is 2 or more, the average of all goodness of fit values is calculated to determine whether the average is greater than a threshold. If it is greater than the threshold, it is determined that the polishing end point has been reached and the chemical mechanical polishing is stopped; if it is less than or equal to the threshold, it is determined that the polishing end point has not been reached.

In the above implementation, no matter what kind of difference, one or more spectral curve segments corresponding to the target sub-range can be taken in the wavelength region. Similarly, a corresponding standard curve segment is also taken in the standard endpoint spectral curve to obtain the difference of each target sub-range respectively. Then, two methods can be used to detect whether the polishing endpoint is reached. The first method is: determine whether each difference meets the preset conditions. If each difference meets the preset conditions, it is determined that the polishing endpoint is reached; the second method is: first calculate the average value of all differences. If the average value meets the preset conditions, it is determined that the polishing endpoint is reached. Through the above two implementations, the detection accuracy of the polishing endpoint is improved.

By implementing the embodiments of the present application, it is possible to determine at least two characteristic extreme values from the reflection coherence spectrum curve by analyzing the reflection coherence spectrum curve, calculate the target coherence level based on the at least two characteristic extreme values, and determine whether the precise judgment area has been reached based on the target coherence level. When polishing enters the precise judgment area, the difference between the reflection coherence spectrum curve and the standard endpoint spectrum curve is further compared and analyzed, so that it is possible to accurately and quickly analyze whether the polishing endpoint has been reached, with simple operation and high accuracy.

Embodiment 3

Further, in one embodiment, please refer to FIG. 9 , which is a flow chart of the optical endpoint detection method of chemical mechanical polishing disclosed in Example 3 of the present application; in FIG. 9 , the step 705 of obtaining the target coherence order according to the wavelengths corresponding to at least two characteristic extreme values includes:

901. Select any two characteristic extreme values from at least two characteristic extreme values as first characteristic extreme values, and obtain the wavelength and calibration number corresponding to the first characteristic extreme value, where the calibration number is a serial number assigned to each characteristic extreme value in sequence.

902. Obtain a target coherence order according to the wavelength and the calibration number corresponding to the first characteristic extreme value.

Furthermore, the above-mentioned obtaining of the wavelength and calibration number corresponding to the first characteristic extreme value includes:

Obtain the wavelength corresponding to the first characteristic extreme value;

Starting from the first interference peak on the right side of the minimum wavelength in the selected wavelength region, calibration numbers are assigned to the calibrated interference peaks and interference troughs in sequence to obtain the calibration number corresponding to the first characteristic extreme value.

In the above implementation, calibration numbers are assigned to the calibrated interference peaks and interference troughs in sequence starting from the first interference peak on the right side of the minimum wavelength in the wavelength region, and then the wavelength and calibration number corresponding to the first characteristic extreme value are obtained.

For example, the first interference peak on the right side of the minimum wavelength in the wavelength range is taken as the first characteristic extreme value, and the first characteristic extreme value includes the kth characteristic extreme value and the pth characteristic extreme value in the wavelength range, and the wavelength corresponding to the kth characteristic extreme value is obtained. and calibration number k, and wavelength corresponding to the pth characteristic extreme value and calibration number p, k is less than p, and k and p are positive integers. Then, the target coherence level is calculated by formula (2): , formula (2) is as follows:

(Formula 2)

Calculated by the above formula (2) value, thereby obtaining the target coherence level. is a non-integer, we can further The values are rounded off and the target coherence level is finally obtained.

For example, please refer to Figure 10, which is a schematic diagram of determining the target coherence level disclosed in an embodiment of the present application; in Figure 10, the horizontal axis is wavelength (unit: nm), the vertical axis is light intensity (Intensity), and this wavelength region is the wavelength range [a, b]. Starting from the right side of the minimum wavelength a, the first interference peak is determined as the first characteristic extreme value, and the calibration number is 1, followed by the first interference trough being determined as the second characteristic extreme value, and the calibration number is 2, the second interference peak is determined as the third characteristic extreme value, and the calibration number is 3, the second interference trough is determined as the fourth characteristic extreme value, and the calibration number is 4, and so on, and other characteristic extreme values in the wavelength range [a, b] are assigned calibration numbers in sequence.

After the calibration numbers are assigned in FIG10 , the calibration numbers corresponding to the first characteristic extreme values (such as the kth characteristic extreme value and the pth characteristic extreme value introduced above) are obtained, and then the target coherence order can be calculated using the above formula (2) according to the wavelength and calibration number corresponding to the first characteristic extreme value.

Through the above implementation, the characteristic extreme values within a wavelength region are assigned calibration numbers in sequence. According to the wavelengths and calibration numbers corresponding to the two selected characteristic extreme values, the target coherence order can be accurately calculated using formula (2), so as to help accurately detect whether the polishing end point has been reached.

Embodiment 4

In another embodiment, please refer to FIG. 11 , which is a flow chart of an optical endpoint detection method for chemical mechanical polishing disclosed in Example 4 of the present application; in FIG. 11 , the step 705 of obtaining the target coherence order according to the wavelengths corresponding to at least two characteristic extreme values includes:

1101. Select two characteristic extreme values from at least two characteristic extreme values as second characteristic extreme values, and obtain wavelengths corresponding to the second characteristic extreme values and the number of characteristic extreme values between the second characteristic extreme values.

1102. Obtain a target coherence order according to the wavelength and number corresponding to the second characteristic extreme value.

In the above embodiment, two characteristic extreme values are selected, and the accurate target coherence order can be calculated based on the wavelength of the two characteristic extreme values and the number of characteristic extreme values between the two characteristic extreme values, so as to further quickly detect whether the polishing end point is reached based on the target coherence order, and it will not be affected by factors such as power supply voltage fluctuations and temperature changes, and the detection accuracy is high.

Combined with the above introduction, the optical path difference formula , in the case of constructive interference , in the case of destructive interference , it can be obtained that, in the first possible implementation, in the above step 1101, two characteristic extreme values are selected from at least two characteristic extreme values as second characteristic extreme values, and the wavelength corresponding to the second characteristic extreme values and the number of characteristic extreme values between the second characteristic extreme values are obtained, including:

Two characteristic extrema indicating interference peaks are selected from at least two characteristic extrema as second characteristic extrema, and the wavelengths corresponding to the second characteristic extrema and the number of characteristic extrema indicating interference troughs between the second characteristic extrema are obtained.

Among them, based on the calibrated characteristic extreme value, the wavelength corresponding to the second characteristic extreme value can be obtained, and since the second characteristic extreme value indicates the interference peak, combined with the formula for constructive interference , it can be seen that ,and and The absolute value of the difference, that is represents the number of characteristic extreme values indicating the interference trough between two second characteristic extreme values, so that and The value of and are the coherence orders corresponding to the second characteristic extreme value, and the wavelength of the second characteristic extreme value with coherence order N ₁ is ₁ , the wavelength of the second characteristic extremum of the coherence order _N2 is ₂ .

Alternatively, in a second possible implementation, in step 1101, two characteristic extreme values are selected from at least two characteristic extreme values as second characteristic extreme values, and the wavelengths corresponding to the second characteristic extreme values and the number of characteristic extreme values between the second characteristic extreme values are obtained, including:

Two characteristic extrema indicating interference troughs are selected from at least two characteristic extrema as second characteristic extrema, and the wavelengths corresponding to the second characteristic extrema and the number of characteristic extrema indicating interference peaks between the second characteristic extrema are obtained.

Among them, based on the calibrated characteristic extreme value, the wavelength corresponding to the second characteristic extreme value can be obtained, and since the second characteristic extreme value indicates the interference trough, combined with the formula for interference destructive , it can be seen that ,and and The absolute value of the difference, that is The number of characteristic extreme values indicating the interference peak between the two second characteristic extreme values can be obtained. and The value of and are the coherence series corresponding to the second eigenvalue, and the coherence series is The wavelength of the second characteristic extreme value of , coherence series The wavelength of the second characteristic extreme value of .

Alternatively, in a third possible implementation, in step 1101, selecting two characteristic extreme values from at least two characteristic extreme values as second characteristic extreme values, and obtaining the wavelengths corresponding to the second characteristic extreme values and the number of characteristic extreme values between the second characteristic extreme values include:

A characteristic extreme value indicating an interference peak and a characteristic extreme value indicating an interference trough are selected from at least two characteristic extreme values as the second characteristic extreme value, and the wavelength of the second characteristic extreme value indicating the interference peak, the wavelength of the second characteristic extreme value indicating the interference trough, and the number of characteristic extreme values indicating the interference peak between the second characteristic extreme values or the number of characteristic extreme values indicating the interference trough between the second characteristic extreme values are obtained.

Among them, based on the calibrated characteristic extreme value, the wavelength corresponding to the second characteristic extreme value can be obtained. From the optical path difference formulas in the two cases of constructive interference and destructive interference, it can be known that: ,and The number of characteristic extreme values indicating interference peaks or interference troughs between two second characteristic extreme values can be obtained. and The value of is the coherence order corresponding to the interference trough in the second eigenvalue, is the coherence order indicating the interference peak in the second characteristic extremum, and the coherence order is The wavelength of the second characteristic extreme value of , coherence series The wavelength of the second characteristic extremum .

Through any of the above implementations, the target coherence order can be accurately calculated to help accurately detect whether the polishing end point has been reached.

Embodiment 5

Please refer to FIG. 12 , which is a schematic diagram of the structure of an optical endpoint detection device for chemical mechanical polishing disclosed in Example 1 of the present application; as shown in FIG. 12 , the optical endpoint detection device for chemical mechanical polishing may include:

The acquisition module 1201 is used to acquire the collected reflection coherence spectrum data during the chemical mechanical polishing process, and obtain a reflection coherence spectrum curve according to the reflection coherence spectrum data;

The calibration module 1202 is used to select a wavelength region on the reflection coherence spectrum curve, calibrate the characteristic extreme value of the reflection coherence spectrum curve in the wavelength region, and obtain the target coherence order, wherein the characteristic extreme value indicates the interference peak and interference trough of the reflection coherence spectrum curve, and the target coherence order is the highest coherence order in the wavelength region;

The judgment module 1203 is used to compare whether the target coherence level is equal to the standard coherence level of the preset standard endpoint spectrum curve, where the standard coherence level is the highest coherence level of the standard endpoint spectrum curve in a wavelength region;

The detection module 1204 is used to obtain the difference between the reflection coherence spectrum curve and the standard endpoint spectrum curve when the determination result of the judgment module 1203 is that the two are equal, and detect whether the optical endpoint of polishing is reached according to the difference.

Optionally, the difference is at least one of mean square error MSE, root mean square error RMSE, mean absolute error MAE or goodness of fit GOF.

The above device is implemented, and during the chemical mechanical polishing process, the collected reflection coherence spectrum data is obtained, and a reflection coherence spectrum curve is obtained according to the reflection coherence spectrum data. Further, a wavelength region is selected on the reflection coherence spectrum curve, and within the wavelength region, the characteristic extreme value of the reflection coherence spectrum curve is calibrated, and the target coherence order is obtained, the characteristic extreme value indicates the interference peak and interference trough of the reflection coherence spectrum curve, and the target coherence order is the highest coherence order within the wavelength region; finally, the target coherence order is compared with the standard coherence order of the preset standard endpoint spectrum curve to see whether they are equal, and the target coherence order is obtained. The quasi-coherence order is the highest coherence order of the standard endpoint spectral curve within the wavelength region; when the two are equal, the difference between the reflected coherence spectral curve and the standard endpoint spectral curve is obtained, and optical endpoint detection is performed based on the difference; by implementing the embodiments of the present application, the polishing position of the chemical mechanical polishing can be determined according to the target coherence order by analyzing the reflected coherence spectral curve, and then the difference between the reflected coherence spectral curve and the standard endpoint spectral curve can be compared and analyzed to accurately and quickly analyze whether the polishing endpoint has been reached, without being affected by factors such as power supply voltage fluctuations and temperature changes, and the detection accuracy and efficiency are high.

The calibration module 1202 is specifically used to obtain the target coherence level in the following manner:

At least two characteristic extreme values are determined among the characteristic extreme values calibrated within a wavelength region; and a target coherence order is obtained according to the wavelengths respectively corresponding to the at least two characteristic extreme values.

Optionally, the calibration module 1202 is used to obtain the target coherence order according to the wavelengths corresponding to the at least two characteristic extreme values in the following manner:

Select any two characteristic extreme values from at least two characteristic extreme values as the first characteristic extreme values, obtain the wavelength and calibration number corresponding to the first characteristic extreme value, and the calibration number is the serial number assigned to each characteristic extreme value in sequence; obtain the target coherence order according to the wavelength and calibration number corresponding to the first characteristic extreme value.

Furthermore, the calibration module 1202 is used to obtain the wavelength and calibration number corresponding to the first characteristic extreme value in the following manner:

The wavelength corresponding to the first characteristic extreme value is obtained; calibration numbers are assigned to the calibrated interference peaks and interference troughs in sequence starting from the first interference peak on the right side of the minimum wavelength in a wavelength region, and the calibration number corresponding to the first characteristic extreme value is obtained.

Alternatively, the calibration module 1202 is used to obtain the target coherence level according to the wavelengths corresponding to the at least two characteristic extreme values in the following manner:

Select two characteristic extreme values from at least two characteristic extreme values as second characteristic extreme values, and obtain the wavelengths corresponding to the second characteristic extreme values and the number of characteristic extreme values between the second characteristic extreme values;

The target coherence order is obtained according to the wavelength and the number corresponding to the second characteristic extreme value.

Furthermore, the calibration module 1202 is used to select two characteristic extreme values from at least two characteristic extreme values as second characteristic extreme values, and obtain the wavelengths corresponding to the second characteristic extreme values and the number of characteristic extreme values between the second characteristic extreme values in the following manner:

Select two characteristic extreme values indicating interference peaks from at least two characteristic extreme values as second characteristic extreme values, obtain the wavelengths corresponding to the second characteristic extreme values and the number of characteristic extreme values indicating interference troughs between the second characteristic extreme values; or,

Select two characteristic extreme values indicating interference wave valleys from at least two characteristic extreme values as second characteristic extreme values, obtain the wavelengths corresponding to the second characteristic extreme values and the number of characteristic extreme values indicating interference wave peaks between the second characteristic extreme values; or,

A characteristic extreme value indicating an interference peak and a characteristic extreme value indicating an interference trough are selected from at least two characteristic extreme values as the second characteristic extreme value, and the wavelength of the second characteristic extreme value indicating the interference peak, the wavelength of the second characteristic extreme value indicating the interference trough, and the number of characteristic extreme values indicating the interference peak between the second characteristic extreme values or the number of characteristic extreme values indicating the interference trough between the second characteristic extreme values are obtained.

Optionally, the detection module 1204 is used to perform optical endpoint detection according to the difference in the following specific manners:

If the difference meets the preset condition, the chemical mechanical polishing is stopped.

For more information about the device, please refer to the corresponding description of the above method embodiment, which will not be repeated here.

Embodiment 7

The embodiment of the present application also discloses a chemical mechanical polishing method, which can utilize the optical endpoint detection method of the chemical mechanical polishing to polish the sample until the polishing endpoint is reached.

Please refer to the above introduction for the specific implementation method, which will not be repeated here.

Embodiment 8

The embodiment of the present application also discloses an optical endpoint detection system for chemical mechanical polishing, including an incident light source, an optical channel, an optical analysis unit, and a data acquisition and analysis unit. The data acquisition and analysis unit may be the optical endpoint detection device for chemical mechanical polishing described above.

Alternatively, the optical endpoint detection system for chemical mechanical polishing includes an incident light source, an optical channel, an optical analysis unit, a data acquisition and analysis unit, and the optical endpoint detection device for chemical mechanical polishing introduced above or the electronic device introduced in this application.

For more information, please refer to the above introduction, which will not be repeated here.

Embodiment 9

Please refer to FIG. 13 , which is a schematic diagram of the structure of an electronic device disclosed in an embodiment of the present application; the electronic device shown in FIG. 13 may include:

A memory 1301 storing executable program code; a processor 1302 coupled to the memory 1301; wherein the processor 1302 calls the executable program code stored in the memory 1301 to execute part or all of the steps of any one of the optical endpoint detection methods for chemical mechanical polishing of Figures 6, 7, 9 or 11.

An embodiment of the present application further discloses a computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute an optical endpoint detection method for chemical mechanical polishing disclosed in FIG. 6 , FIG. 7 , FIG. 9 or FIG. 11 .

The embodiment of the present application further discloses a computer program product. When the computer program product is run on a computer, the computer is enabled to execute part or all of the steps of any one of the methods disclosed in FIG. 6 , FIG. 7 , FIG. 9 or FIG. 11 .

An embodiment of the present application also discloses an application publishing platform, which is used to publish a computer program product. When the computer program product runs on a computer, the computer executes part or all of the steps of any one of the methods disclosed in Figures 6, 7, 9 or 11.

A person skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

The above is a detailed introduction to the optical endpoint detection method and related equipment for chemical mechanical polishing disclosed in the embodiment of the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for general technical personnel in this field, according to the idea of the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims (10)

1. An optical endpoint detection method for chemical mechanical polishing, characterized in that the method comprises: During the chemical mechanical polishing process, acquiring collected reflection coherence spectrum data, and obtaining a reflection coherence spectrum curve according to the reflection coherence spectrum data; Selecting a wavelength region on the reflection coherence spectrum curve, calibrating a characteristic extreme value of the reflection coherence spectrum curve within the wavelength region, and obtaining a target coherence order, wherein the characteristic extreme value indicates an interference peak and an interference trough of the reflection coherence spectrum curve, and the target coherence order is a maximum coherence order within the wavelength region; Comparing the target coherence level with a standard coherence level of a preset standard endpoint spectrum curve to see whether they are equal, the standard coherence level being the highest coherence level of the standard endpoint spectrum curve in the wavelength region; If the two are equal, obtaining the difference between the reflection coherence spectrum curve and the standard endpoint spectrum curve, and performing optical endpoint detection according to the difference; The obtaining of the target coherence level comprises: Determining at least two characteristic extreme values among the characteristic extreme values calibrated within the wavelength region; Acquire the target coherence order according to the wavelengths respectively corresponding to the at least two characteristic extreme values; The optical endpoint detection is performed according to the difference, comprising: If the difference meets a preset condition, the chemical mechanical polishing is stopped. 2. The method according to claim 1, characterized in that the step of obtaining the target coherence order according to the wavelengths corresponding to the at least two characteristic extreme values comprises: Select any two characteristic extreme values from the at least two characteristic extreme values as first characteristic extreme values, and obtain the wavelength and calibration number corresponding to the first characteristic extreme values, where the calibration number is a serial number assigned to each characteristic extreme value in sequence; The target coherence order is obtained according to the wavelength and calibration number corresponding to the first characteristic extreme value. 3. The method according to claim 2, characterized in that the step of obtaining the wavelength and calibration number corresponding to the first characteristic extreme value comprises: Obtaining the wavelength corresponding to the first characteristic extreme value; The calibration numbers are assigned to the calibrated interference peaks and interference troughs in sequence starting from the first interference peak on the right side of the minimum wavelength in the wavelength region, and the calibration numbers corresponding to the first characteristic extreme values are obtained. 4. The method according to claim 1, characterized in that the step of obtaining the target coherence order according to the wavelengths corresponding to the at least two characteristic extreme values comprises: Select two characteristic extreme values from the at least two characteristic extreme values as second characteristic extreme values, and obtain the wavelengths corresponding to the second characteristic extreme values and the number of characteristic extreme values between the second characteristic extreme values; The target coherence order is obtained according to the wavelength and the number corresponding to the second characteristic extreme value. 5. The method according to claim 4, characterized in that the selecting two characteristic extreme values from the at least two characteristic extreme values as second characteristic extreme values, and obtaining the wavelengths corresponding to the second characteristic extreme values and the number of characteristic extreme values between the second characteristic extreme values, comprises: Select two characteristic extreme values indicating interference peaks from the at least two characteristic extreme values as second characteristic extreme values, and obtain the wavelengths corresponding to the second characteristic extreme values and the number of characteristic extreme values indicating interference troughs between the second characteristic extreme values; or Select two characteristic extreme values indicating interference wave valleys from the at least two characteristic extreme values as the second characteristic extreme values, and obtain the wavelengths corresponding to the second characteristic extreme values and the number of characteristic extreme values indicating interference wave peaks between the second characteristic extreme values; or A characteristic extreme value indicating an interference peak and a characteristic extreme value indicating an interference trough are selected from the at least two characteristic extreme values as the second characteristic extreme value, and the wavelength of the second characteristic extreme value indicating the interference peak, the wavelength of the second characteristic extreme value indicating the interference trough, and the number of characteristic extreme values indicating the interference peak between the second characteristic extreme values or the number of characteristic extreme values indicating the interference trough between the second characteristic extreme values are obtained. 6. The method according to any one of claims 1 to 5, characterized in that the difference is at least one of mean square error (MSE), root mean square error (RMSE), mean absolute error (MAE) or goodness of fit (GOF). 7. A chemical mechanical polishing method, characterized in that the sample is polished according to the optical endpoint detection method of chemical mechanical polishing according to any one of claims 1 to 6 until the polishing endpoint is reached. 8. An optical endpoint detection device for chemical mechanical polishing, using the optical endpoint detection method for chemical mechanical polishing according to any one of claims 1 to 6, characterized in that it comprises: An acquisition module, used for acquiring the collected reflection coherence spectrum data during the chemical mechanical polishing process, and obtaining a reflection coherence spectrum curve according to the reflection coherence spectrum data; A calibration module, used for selecting a wavelength region on the reflection coherence spectrum curve, calibrating a characteristic extreme value of the reflection coherence spectrum curve in the wavelength region, and obtaining a target coherence order, wherein the characteristic extreme value indicates an interference peak and an interference trough of the reflection coherence spectrum curve, and the target coherence order is a maximum coherence order in the wavelength region; A judgment module, used for comparing whether the target coherence level is equal to a standard coherence level of a preset standard endpoint spectrum curve, wherein the standard coherence level is the highest coherence level of the standard endpoint spectrum curve in the wavelength region; The detection module is used to obtain the difference between the reflection coherence spectrum curve and the standard endpoint spectrum curve when the judgment result of the judgment module is that the two are equal, and detect whether the optical endpoint of the polishing is reached according to the difference. 9. An electronic device, characterized in that the electronic device comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the method according to any one of claims 1 to 6 or the method according to claim 7. 10. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program, wherein the computer program enables a computer to execute the method according to any one of claims 1 to 6 or the method according to claim 7.