Polishing pad conditioning method, apparatus, device, controller, and storage medium

CN121104903BActive Publication Date: 2026-08-21HWATSING TECHNOLOGY CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202511588496.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-10-31
Publication Date
2026-08-21
Estimated Expiration
2044-10-31

AI Technical Summary

Technical Problem

但目前的修整方法主要依赖于经验公式,缺乏全面的理论模型,存在修整时间长、抛光垫寿命短、去除效率不均匀、精确度差等缺点

Benefits of technology

[0040]本发明提供的抛光垫修整方法、装置、设备、控制器及存储介质,该修整方法在初次修整周期内对抛光垫进行修整时采用初始修整配方,在后续修整周期内进行修整时采用经自学习调整方法调整的修整配方。本发明通过抛光垫修整方法对抛光垫实现可控修整,根据抛光垫的磨损状态调整修整参数,优化修整效果,延长抛光垫的使用寿命,降低抛光垫的更换频率和成本。

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121104903B_ABST
    Figure CN121104903B_ABST
Patent Text Reader

Abstract

The present application relates to a polishing pad conditioning method, device, equipment, controller and storage medium, wherein the initial conditioning formula is used when conditioning the polishing pad in the initial conditioning cycle, and the conditioning formula adjusted by the self-learning adjustment method is used when conditioning in the subsequent conditioning cycle. The present application realizes controllable conditioning of the polishing pad through the polishing pad conditioning method, adjusts the conditioning parameters according to the wear state of the polishing pad, optimizes the conditioning parameters, prolongs the service life of the polishing pad, and reduces the replacement frequency and cost of the polishing pad.
Need to check novelty before this filing date? Find Prior Art

Description

[0001] This application is a divisional application of the invention patent filed on October 31, 2024, with application number 2024115382892 and title "Polishing Pad Trimming Method, Apparatus, Equipment, Controller and Storage Medium". Technical Field

[0002] This invention relates to the field of chemical mechanical planarization (CMP) of semiconductor wafers, and more specifically, to a method, apparatus, device, controller, and storage medium for dressing polishing pads in a chemical mechanical polishing apparatus. Background Technology

[0003] Integrated circuits (ICs) are the core and lifeline of the information technology industry. ICs are generally formed by successively depositing conductive layers, semiconductor layers, or insulating layers on a silicon wafer. This results in a thin film of filler layers deposited on the wafer surface. During the manufacturing process, the filler layers need to be continuously planarized until a patterned top surface is exposed, in order to form conductive paths between the raised patterns.

[0004] Chemical mechanical polishing (CMP) is the preferred planarization process in IC manufacturing. However, in CMP, over-polishing can damage material properties and affect device performance, while under-polishing may require additional processing steps, increasing costs.

[0005] Polishing pads play a crucial role in chemical mechanical polishing (CMP), responsible for storing and transporting polishing fluid, removing processing residues, transmitting mechanical loads, and maintaining the polishing environment. During polishing, the surface of the polishing pad gradually wears down, reducing polishing efficiency and quality; therefore, polishing pads need to be dressed during use. Currently, common dressing methods include diamond dressers. Diamond has extremely high hardness and provides good mechanical grinding effects when dressing polishing pads. However, current dressing methods mainly rely on empirical formulas and lack comprehensive theoretical models, resulting in drawbacks such as long dressing times, short polishing pad life, uneven removal efficiency, and poor accuracy. Summary of the Invention

[0006] The purpose of this invention is to provide a polishing pad dressing method, apparatus, equipment, controller, and storage medium, which utilizes intelligent adjustment of dressing parameters to improve dressing accuracy and efficiency, and also improve the uniformity of the polishing pad and extend its service life.

[0007] To achieve the above objectives, a first aspect of the present invention provides a polishing pad conditioning method, the method comprising the following steps:

[0008] Determine the dressing cycle for the polishing pad;

[0009] The dressing formula of the polishing pad is updated once after each dressing cycle of polishing operation.

[0010] Specifically, an initial dressing formula is used when dressing the polishing pad during the initial dressing cycle, and a dressing formula adjusted by a self-learning adjustment method is used when dressing in each subsequent dressing cycle.

[0011] In the polishing pad dressing method described above, optionally, the dressing cycle is an integer multiple of the time required to complete the polishing of each box of wafers to be polished, and the parameters of the dressing formula include dressing time, dressing pressure, and the movement speed of the dressing device relative to the polishing pad.

[0012] In the polishing pad conditioning method described above, optionally, the self-learning adjustment method includes the following steps:

[0013] Step 1: Use the morphology detection component to collect the current morphology data of the polishing pad after it has been repaired;

[0014] Step II: Calculate the actual amount of polishing pad removed based on the current morphology data and the historical morphology data of the previous cycle of the polishing pad, and collect the polishing time used for this polishing of the polishing pad. Fit the polishing pad removal amount calculation formula according to the actual amount of polishing removal and the polishing time.

[0015] Step III: Calculate the predicted trimming removal amount for the next cycle based on the trimming removal amount calculation formula, and calculate the predicted morphology data of the polishing pad after trimming in the next cycle based on the predicted trimming removal amount for the next cycle and the current morphology data. Calculate the expected trimming adjustment coefficient of the trimming formula parameters based on the deviation between the predicted morphology data and the target trimming morphology data.

[0016] Step IV: Determine the weights of the system's random influence and hardware parameter influence based on historical adjustment coefficients, and perform weighted processing on the expected adjustment coefficients;

[0017] Step V: Output the weighted expected trimming adjustment coefficients, and based on the weighted expected trimming adjustment coefficients, output a new trimming formula for the polishing pad trimming in the next cycle.

[0018] In the polishing pad conditioning method described above, optionally, steps I to V are performed on the polishing pad in sections.

[0019] In the polishing pad dressing method described above, optionally, the morphology detection component in step I acquires the current morphology data through a contact sensor and / or a non-contact sensor.

[0020] In the polishing pad dressing method described above, optionally, the step of fitting the formula for calculating the amount of polishing pad dressing removal includes:

[0021] Establish a linear regression model y = kx + b, where y is the amount of polishing pad removed, x is the polishing time of the polishing pad, k is the polishing rate of the polishing pad, and b is the intercept term of the linear model.

[0022] The actual trimming and removal amount y of data points in each area of ​​the polishing pad is collected. i and rest time x i ;

[0023] Define the error function, and select initial k and b values, and set the tolerance;

[0024] An iterative method is used to approximate k and b. When the value of the error function is less than or equal to the tolerance, the iteration stops and k and b are output.

[0025] In the polishing pad dressing method described above, optionally, the error function is a mean square error function:

[0026]

[0027] Where m is the number of data points.

[0028] In the polishing pad conditioning method described above, optionally, the iterative equation is:

[0029]

[0030] in, It is the partial derivative of the error function M with respect to k. It is the partial derivative of the error function with respect to b, and α is the step size, used to control the size of each update step.

[0031] In the polishing pad dressing method described above, optionally, the initial k value and b value are randomly selected values, or the initial k value and b value are calculated using the following formula:

[0032]

[0033] ∑x represents all x i The sum of , ∑y represents all y i The sum of ∑xy is all x i y i The sum of ∑x 2 For all x i 2 The sum of and n is the number of regions of the polishing pad.

[0034] To achieve the above objectives, a second aspect of the present invention provides a polishing pad dressing device for a chemical mechanical polishing apparatus. The polishing pad dressing device includes a fixed base, a swing arm, and a dressing device. The dressing device is connected to the fixed base via the swing arm. The swing arm is oscillating around the fixed base. The dressing device located at the end of the swing arm dresses the polishing pad using the dressing method as described in any one of the preceding first aspects.

[0035] To achieve the above objectives, a third aspect of the present invention provides a chemical mechanical polishing apparatus, the chemical mechanical polishing apparatus including the polishing pad dressing device as described in the second aspect above.

[0036] To achieve the above objectives, a fourth aspect of the present invention provides an intelligent controller, the intelligent controller comprising:

[0037] The memory is used to store computer-executable instructions or computer programs;

[0038] A processor, when executing computer-executable instructions or computer programs stored in the memory, implements the dressing method as described in any one of the preceding first aspects to dress the polishing pad for the chemical mechanical polishing apparatus.

[0039] To achieve the above objectives, a fifth aspect of the present invention provides a computer-readable storage medium storing computer-executable instructions or a computer program that, when executed by a processor, implements the trimming method as described in any one of the preceding first aspects to trim the polishing pad used in the chemical mechanical polishing apparatus.

[0040] This invention provides a polishing pad dressing method, apparatus, device, controller, and storage medium. The dressing method uses an initial dressing formula during the initial dressing cycle, and uses a dressing formula adjusted by a self-learning method during subsequent dressing cycles. This invention achieves controllable dressing of polishing pads by adjusting dressing parameters according to the wear state of the polishing pad, optimizing the dressing effect, extending the service life of the polishing pad, and reducing the replacement frequency and cost.

[0041] Other features and advantages of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0042] 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.

[0043] Figure 1 This is a flowchart of the polishing pad trimming method provided in the embodiments of the present invention;

[0044] Figure 2 This is a schematic diagram illustrating the implementation effect of the traditional trimming method provided in this embodiment of the invention;

[0045] Figure 3 This is a schematic diagram illustrating the implementation effect of an example of the trimming method provided in an embodiment of the present invention;

[0046] Figure 4 This is a schematic diagram of the dressing device for chemical mechanical polishing provided in an embodiment of the present invention;

[0047] Figure 5 This is a schematic diagram of the structure of the chemical mechanical polishing equipment provided in an embodiment of the present invention. Detailed Implementation

[0048] 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.

[0049] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0050] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiments or examples. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Relevant definitions of other terms will be given in the description below.

[0051] During actual use, polishing pads experience wear and deformation. The abrasive debris generated during polishing fills the micropores on the pad surface, causing glazing. This reduces the pad's ability to store and transport abrasive, leading to a decrease in material removal rate. Therefore, it's necessary to remove the glaze layer from the polishing pad surface to increase surface roughness. Polishing pads also deform during use, requiring trimming to flatten them. Proper trimming can restore the pad's surface from excessive wear and glazing, extending its lifespan. However, trimming reduces the pad's thickness, and excessive trimming may lead to excessive wear and shorten its lifespan. If the polishing pad wears too quickly, optimizing the trimming formula parameters can reduce wear.

[0052] Figure 1 This is a flowchart of the polishing pad trimming method provided in an embodiment of the present invention.

[0053] In this embodiment, the dressing cycle of the polishing pad is determined. Based on the usage and polishing effect of the polishing pad, a dressing formula is needed to dress the polishing pad. The time interval for periodically adjusting the dressing formula is the dressing cycle. The dressing formula of the polishing pad is updated once for each polishing operation completed within one dressing cycle.

[0054] In an optional embodiment, the frequency of polishing pad dressing within an appropriate dressing cycle can be selected according to the type of polishing pad and the requirements of the polishing process. For example, the polishing pad can be dressed once every time a polishing operation is performed using the polishing pad.

[0055] In this embodiment, the dressing cycle is an integer multiple of the time required to polish each box of wafers to be polished. The advantage of this implementation is that it ensures the same dressing formula is used on the polishing pads during the polishing of each batch of wafers, thereby maintaining stable polishing efficiency.

[0056] In some embodiments, the parameters of the polishing pad dressing formula include dressing time, dressing pressure, and the movement speed of the dresser relative to the polishing pad. Dressing time refers to the length of time the dresser dresses the polishing pad. It is understood that an excessively long dressing time may lead to excessive wear of the polishing pad, while an excessively short dressing time may not adequately dress the polishing pad to achieve suitable performance. The movement speed of the dresser relative to the polishing pad affects the uniformity and efficiency of the dressing effect. Dressing pressure, i.e., the pressure applied by the dresser to the polishing pad, ensures that the polishing pad surface is effectively dressed. In optional embodiments, the parameters of the polishing pad dressing formula may also include other parameters affecting the dressing effect of the polishing pad, such as dressing depth and dressing temperature.

[0057] In the polishing pad dressing method of this invention, an initial dressing formula needs to be pre-set based on historical data and experience. The initial dressing formula refers to the dressing formula used initially in the polishing process. Dressing parameters in the initial dressing formula, such as dressing pressure, speed, and dressing time, are set based on the type of polishing pad and the requirements of the polishing process. For more precise calculations, the surface of the polishing pad is divided into multiple regions according to its shape and size, and initial polishing parameters are set for each region.

[0058] During subsequent adjustments, a self-learning adjustment formula is used, and the adjustment parameters are automatically optimized through an algorithm. The self-learning adjustment method for the adjustment formula includes the following steps:

[0059] Step 1: Use the morphology inspection component to collect the current morphology data of the trimmed polishing pad;

[0060] Step II: Calculate the actual amount of polishing pad removed by dressing based on the current morphology data and the historical morphology data of the previous cycle of the polishing pad, and collect the dressing time used for this dressing of the polishing pad. Fit the formula for calculating the amount of polishing pad removal by dressing based on the actual amount of dressing removed and the dressing time.

[0061] Step 3: Calculate the predicted trimming removal amount for the next cycle based on the trimming removal amount calculation formula, and calculate the predicted morphology data of the polishing pad after trimming in the next cycle based on the predicted trimming removal amount for the next cycle and the current morphology data. Calculate the expected trimming adjustment coefficient of the trimming formula parameters by the deviation between the predicted morphology data and the target morphology data.

[0062] Step IV: Determine the weights of the system's random effects and hardware parameter effects based on historical adjustment coefficients, and perform weighted processing on the expected adjustment coefficients;

[0063] Step V: Output the weighted expected trimming adjustment coefficients, and based on the weighted expected trimming adjustment coefficients, output a new trimming formula for the polishing pad trimming in the next cycle.

[0064] In step I, the morphology data of the polishing pad refers to the profile of the polishing pad, that is, the collection of height information of various regions on the surface of the polishing pad. In different embodiments, the morphology detection component can be a contact sensor, which measures the surface morphology of the polishing pad by direct contact with it. Other detection devices, such as acoustic sensors and optical sensors, and other non-contact sensors, are also within the scope of protection of this invention. In some embodiments, a contact sensor can be used to measure the roughness and pressure distribution of the polishing pad, while a non-contact sensor can be used to measure the surface profile and thickness of the polishing pad. In an optional embodiment, the morphology of the polishing pad can be obtained by measuring the distance between the sensor and the polishing disk below the polishing pad, by contacting the surface of the polishing pad and following its undulations.

[0065] In step II, the actual amount removed by the polishing pad is the amount of material removed from the polishing pad during the polishing process. The value of the actual amount removed by the polishing pad is calculated based on the difference between the current morphology data of the polishing pad and the morphology data of the polishing pad in the previous cycle.

[0066] In an optional embodiment, the formula for calculating the amount of trimming and removal of the fitting polishing pad may include:

[0067] A linear regression model is established: y = kx + b, where y is the amount of material removed by the polishing pad, x is the polishing time of the polishing pad, k is the polishing rate of the polishing pad, and b is the intercept term of the linear model. Then, the actual amount of material removed (y) is collected from data points in each region of the polishing pad. i and rest time x i The linear regression model assumes a linear relationship between the dressing removal amount y of the polishing pad and the dressing time x of the polishing pad. By finding the optimal values ​​of k and b, the model minimizes the difference between the model-predicted dressing removal amount and the actual dressing removal amount. By establishing a linear regression model, the best fit of the function for the dressing removal amount in each region of the polishing pad is found, ensuring that the data points in each region are as close as possible to the calculation formula for the dressing removal amount in that region. Alternatively, other linear or nonlinear regression models can also be used in optional implementations.

[0068] In an optional embodiment, an error function is defined to measure the difference between the predicted and actual polishing pad trimming removal amounts, and initial k and b values ​​are selected, while a tolerance is set. It is understood that the tolerance is the maximum acceptable error range for the model.

[0069] In some embodiments, the error function can be the average error function, whose equation is:

[0070]

[0071] Where m is the number of data points.

[0072] It is understood that the mean squared error function can be used to evaluate the accuracy of model predictions. During the model training process, the mean squared error function can be used as a loss function. By minimizing the value of the mean squared error function, the model parameters can be adjusted to improve the model's prediction performance.

[0073] In some optional embodiments, k and b can be approximated using an iterative method, stopping the iteration when the value of the error function is less than or equal to the tolerance. In some embodiments, the iterative equations for k and b are:

[0074]

[0075] in, It is the partial derivative of the error function M with respect to k. It is the partial derivative of the error function with respect to b, and α is the step size, used to control the size of each update step.

[0076] In some examples, the tolerance is often set to a small value to ensure that the model can find the most suitable solution during the iteration process.

[0077] In an optional embodiment, the initial k and b values ​​can be random values ​​or estimated values.

[0078] As a feasible implementation method, the initial values ​​of k and b can be calculated using the least squares method, where...

[0079] The equation for calculating k is:

[0080]

[0081] The equation for calculating b is:

[0082]

[0083] Where ∑x represents all x i The sum of y, ∑y represents all y i The sum of ∑xy is all x i y i The sum of ∑x 2 For all x i 2 The sum of , where n is the number of data points in the dataset.

[0084] It is understood that estimating the initial k and b values ​​using the least squares method helps the above iterative algorithm converge to the optimal solution more quickly.

[0085] In step III, the predicted trimming removal amount for the next cycle is calculated according to the trimming removal amount calculation formula. Based on the predicted trimming removal amount for the next cycle and the current morphology data, the predicted morphology data of the polishing pad after trimming in the next cycle can be preliminarily calculated. By comparing the predicted morphology data with the target trimmed morphology data, the deviation is calculated, which is the expected trimming removal amount of the polishing pad in the next cycle. The expected trimming adjustment coefficient of the trimming formula parameters is calculated based on the expected trimming removal amount of the polishing pad in the next cycle. The trimming adjustment coefficient is used to modify various parameters in the trimming formula, such as trimming time, trimming pressure, and the movement speed of the trimmer relative to the polishing pad. Adjusting the trimming adjustment coefficient refers to adjusting the coefficients that modify various parameters. The trimming adjustment coefficient is affected by various factors, such as hardware parameters and the random influence of the system. By analyzing these factors that affect the accuracy of the trimming adjustment coefficient and adjusting the weights of these factors, the trimming adjustment coefficient can be adjusted to achieve the expected trimming effect.

[0086] In step IV, hardware parameters, such as the hardness, elasticity, and grain size of the polishing pad, will have an impact during the polishing pad dressing process. Adjusting the desired dressing adjustment coefficient in a weighted manner can improve the accuracy of the model.

[0087] It is understood that the random influence of the system is caused by a variety of unpredictable factors, such as temperature changes, material inhomogeneity, and random system vibrations. This random influence may affect the accuracy of the data. By adjusting the weight of the random influence of the system, the expected dressing adjustment coefficient of the polishing pad can be optimized.

[0088] In an optional embodiment, the desired adjustment coefficient can be adjusted more accurately by comparing the adjustment coefficients in several recent adjustment processes, and the impact of outliers can be reduced by analyzing similar data points.

[0089] In step V, a new dressing formula is formulated based on the expected dressing adjustment coefficient of the weighted polishing pad, and the new dressing formula is implemented in the dressing of the polishing pad in the next cycle.

[0090] Figure 2 and Figure 3 This is a schematic diagram illustrating the effects of traditional trimming methods and the polishing pad trimming method provided in this embodiment of the invention.

[0091] exist Figure 2 and Figure 3 In the figure, the horizontal axis represents the radius of the polishing pad, in inches; the vertical axis represents the groove depth of the polishing pad, in mils (thousandths of an inch). The trend lines in the figure show the different polishing effects of the polishing pad under traditional polishing methods and the polishing pad polishing method provided by this invention. The different trend lines are marked 0H, 5H, 10H, 15H, and 20H, with H representing hours, indicating the time period in the polishing pad polishing process. Figure 2 and / or Figure 3 It can be seen that during the dressing process, the longer the time, the worse the dressing effect of the polishing pad becomes, and the less stable the trend line becomes.

[0092] like Figure 2 and Figure 3 As shown, compared with traditional trimming methods, the polishing pad trimming method provided by this invention has a more stable trend line, which may mean that the polishing pad trimming method provided by this invention, by adjusting the trimming formula, makes the polishing pad trimmed more flat, enabling the polishing pad to maintain better uniformity during trimming, and improving the stability and repeatability of the trimming process.

[0093] Figure 4 This is a schematic diagram of the dressing device for chemical mechanical polishing provided in an embodiment of the present invention.

[0094] like Figure 4 As shown, the dressing device 100 includes a fixed base 101, a swing arm 102, and a dressing device 103. The dressing device 103 includes a dressing head 104 and a morphology detection device (not shown in the figure). In an optional embodiment, diamond particles are embedded in the bottom of the dressing device 103. Diamond particles have high hardness and corrosion resistance. The dressing head 104 reciprocates on the surface of the polishing pad, and the diamond particles cut and dress the polishing pad. The morphology detection device can measure the morphology of the polishing pad. In an optional embodiment, the type of morphology detection device is not limited. For example, non-contact sensors such as eddy current sensors and contact sensors such as contact displacement sensors are all within the scope of protection of this invention. In some embodiments, the morphology detection device and the dressing head 104 are arranged adjacent to each other in the horizontal direction. The advantage of this arrangement is that it avoids interference from the dressing head 104 to the measurement of the morphology detection device during the dressing of the polishing pad.

[0095] In an optional embodiment, the trimmer 103 is connected to the fixed base 101 via a swing arm 102. By swinging the swing arm 102 around the fixed base 101, the trimmer 103 located at the end of the swing arm 102 can perform reciprocating motion to trim the polishing pad.

[0096] Figure 5 This is a schematic diagram of the structure of the chemical mechanical polishing equipment provided in an embodiment of the present invention.

[0097] Figure 5 In the chemical mechanical polishing equipment, a dressing device 100 is included, the structure of which is as follows: Figure 4 As shown. Figure 5 In the illustrated embodiment, the chemical mechanical polishing apparatus further includes a polishing disc 200, a support head 300, and a liquid supply device 400. Figure 5 In the example, polishing disk 200 covers a polishing pad for polishing the wafer. A support head 300 presses the wafer onto the polishing pad on the polishing disk 200. A liquid supply device 400 sprays polishing liquid onto the surface of the polishing pad. Under the chemical action of the polishing liquid, the relative movement between the support head 300 and the polishing disk 200 causes friction between the wafer and the polishing pad, thus achieving polishing. During the polishing process, a dressing device 100 dresses the polishing pad. Specifically, the dressing device 100 can correct the surface deformation of the polishing pad during wafer polishing, ensuring the morphology of the polishing pad meets process requirements. This stabilizes the wafer polishing removal rate, achieves global wafer planarization, and extends the lifespan of the polishing pad.

[0098] The present invention also provides an intelligent controller, including a memory and a processor. The memory is used to store computer-executable instructions or computer programs. When the processor executes the computer-executable instructions or computer programs stored in the memory, it implements the dressing method as described in any of the foregoing embodiments for dressing the polishing pad of a chemical mechanical polishing apparatus.

[0099] Furthermore, the present invention also provides a computer-readable storage medium storing computer-executable instructions or a computer program, which, when executed by a processor, implement the trimming method as described in any of the foregoing embodiments for trimming the polishing pad of a chemical mechanical polishing device.

[0100] It should be noted that the computer-readable storage medium described above in this invention can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this invention, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0101] 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 chemical mechanical polishing device, characterized in that, The device includes a polishing pad, a support head, and a dressing device. The support head presses the wafer onto a polishing pad on the polishing pad. The relative movement between the support head and the polishing pad causes the wafer to rub against the polishing pad for polishing. The dressing device is configured to perform the following steps to dress the polishing pad: Determine the dressing cycle for the polishing pad; The polishing pad's dressing formula should be updated after each dressing cycle of polishing work. The parameters of the dressing formula include dressing time, dressing pressure, and the movement speed of the dressinger relative to the polishing pad. The initial dressing formula is used when dressing the polishing pad in the first dressing cycle, and the dressing formula adjusted by the self-learning adjustment method is used when dressing in each subsequent dressing cycle. Self-learning adjustment methods include: Step 1: Use the morphology inspection component to collect the current morphology data of the trimmed polishing pad; Step II: Calculate the actual amount of polishing pad removed by dressing based on the current morphology data and the historical morphology data of the previous cycle of the polishing pad, and collect the dressing time used for this dressing of the polishing pad. Fit the formula for calculating the amount of polishing pad removal by dressing based on the actual amount of dressing removed and the dressing time. Step 3: Calculate the predicted trimming removal amount for the next cycle based on the trimming removal amount calculation formula, and calculate the predicted morphology data of the polishing pad after trimming in the next cycle based on the predicted trimming removal amount for the next cycle and the current morphology data. Calculate the expected trimming adjustment coefficient of the trimming formula parameters by the deviation between the predicted morphology data and the target trimming morphology data. Step IV: Determine the weights of the system's random effects and hardware parameter effects based on historical adjustment coefficients, and perform weighted processing on the expected adjustment coefficients; Step V: Output the weighted expected trimming adjustment coefficients, and based on the weighted expected trimming adjustment coefficients, output a new trimming formula for the polishing pad trimming in the next cycle.

2. The chemical mechanical polishing equipment as described in claim 1, characterized in that, The chemical mechanical polishing equipment also includes a liquid supply device that sprays polishing liquid onto the surface of the polishing pad.

3. The chemical mechanical polishing equipment as described in claim 1, characterized in that, The dressing device includes a fixed base, a swing arm, and a dressing device. The bottom of the dressing device is inlaid with diamonds. The dressing head reciprocates on the surface of the polishing pad, and the diamond particles cut and dress the polishing pad.

4. The chemical mechanical polishing equipment as described in claim 1, characterized in that, The finishing cycle is an integer multiple of the time required to complete the polishing of each box of wafers to be polished.

5. The chemical mechanical polishing equipment as described in claim 1, characterized in that, Steps 1 through 5 all involve dividing the polishing pad into regions.

6. The chemical mechanical polishing equipment as described in claim 5, characterized in that, In step I, the morphology detection component acquires the current morphology data through contact sensors and / or non-contact sensors.

7. The chemical mechanical polishing equipment as described in claim 5, characterized in that, The steps for fitting the formula for calculating the amount of polishing pad trimming and removal include: Establish a linear regression model y = kx + b, where y is the amount of polishing pad removed, x is the polishing time of the polishing pad, k is the polishing rate of the polishing pad, and b is the intercept term of the linear model. The actual trimming and removal amount y of data points in each area of ​​the polishing pad is collected. i and rest time x i ; Define the error function, and select initial k and b values, and set the tolerance; An iterative method is used to approximate k and b. The iteration stops when the value of the error function is less than or equal to the tolerance. Output k and b.

8. The chemical mechanical polishing equipment as described in claim 7, characterized in that, The error function is the mean square error. function: Where m is the number of data points.

9. The chemical mechanical polishing equipment as described in claim 7, characterized in that, The iterative equation is: in, It is the partial derivative of the error function M with respect to k. It is the partial derivative of the error function with respect to b, and α is the step size, used to control the size of each update step.

10. The chemical mechanical polishing apparatus as described in claim 7, characterized in that, The initial k and b values ​​are randomly selected, or the initial k and b values ​​are calculated using the following formula: ∑x represents all x i The sum of y, ∑y represents all y i The sum of ∑xy is all x i y i The sum of ∑x 2 For all x i 2 The sum of and n is the number of regions of the polishing pad.

Citation Information

Patent Citations

  • Method for controlling planarization process and chemical mechanical planarization device

    CN110561201A

  • Polishing method and polishing apparatus

    CN112936090A