Semiconductor process temperature compensation method, temperature compensation system and computer equipment

Abstract

The invention provides a semiconductor process temperature compensation method, a temperature compensation system and computer equipment, and the temperature compensation method comprises the steps: obtaining a region sensitivity matrix of a wafer monitoring piece, so that the region sensitivity matrix can be used for representing the influence on the process value of each region when the temperature of any region is adjusted. And on the basis, according to the compensation coefficient, the overall sensitivity, the target process value, the regional sensitivity matrix and the initial average process value, obtaining a regional temperature compensation value matrix. Therefore, the acquisition of the regional temperature compensation value matrix considers the influence of mutual thermal action of the lamp tubes between different regions, so that the accuracy of the temperature compensation value in the temperature compensation matrix is high.

Description

Technical Field

[0001] This invention relates to the field of semiconductor processes, and more particularly to a semiconductor process temperature compensation method, temperature compensation system, and computer equipment. Background Technology

[0002] In semiconductor manufacturing, rapid thermal annealing is primarily used to repair lattice damage caused by ion implantation, activate impurity ions, and optimize device performance through interface engineering. Since temperature deviations in rapid thermal annealing equipment directly affect process effectiveness, product consistency and reliability, and production economics, adjusting these deviations is a core aspect of routine maintenance and process development for rapid thermal annealing equipment. Specifically, after changes in the chamber environment, the rapid thermal annealing equipment requires temperature compensation and uniformity adjustments based on the wafer monitoring data.

[0003] Please refer to the following: Figure 1 , Figure 1 This is a flowchart of a method for adjusting the temperature compensation of a rapid thermal annealing process. The procedure for adjusting the process temperature compensation is as follows: First, the process temperature compensation value for each region is obtained by directly calculating the relationship between the wafer monitoring wafer process value and the temperature. This is used to adjust the uniformity of the wafer monitoring wafer process value. Then, after the uniformity meets the standard, the overall process value is adjusted to reach the target value.

[0004] However, when directly calculating the process temperature compensation value for each region by using the relationship between wafer monitoring chip process values ​​and temperature, the mutual thermal interaction between lamps in different regions will cause the process values ​​of other regions to change as well. This makes the process temperature compensation value for each region calculated directly by using the relationship between wafer monitoring chip process values ​​and temperature inaccurate and unreliable. Summary of the Invention

[0005] This invention provides a temperature compensation method, a temperature compensation system, and a computer device to improve the accuracy of temperature compensation values ​​in the obtained temperature compensation matrix.

[0006] According to a first aspect of the present invention, a method for temperature compensation in a semiconductor process is provided, comprising: A wafer monitoring chip is provided, wherein the wafer monitoring chip is divided into N regions, where N is a positive integer greater than 1; The overall sensitivity and target process value of the wafer monitoring chip are obtained, wherein the overall sensitivity is used to characterize the relationship between the process value of the wafer monitoring chip and the overall process temperature. At the initial temperature, obtain N initial average process values ​​corresponding to the N regions; Obtain the region sensitivity matrix of the wafer monitoring wafer. The region sensitivity matrix has N rows and N columns. The element in the i-th row and j-th column of the region sensitivity matrix is ​​used to characterize the region sensitivity of the process value of the i-th region to the temperature change of the j-th region, 1≤i≤N, 1≤j≤N. A compensation coefficient is obtained, and a regional temperature compensation value matrix is ​​obtained based on the compensation coefficient, the overall sensitivity, the target process value, the regional sensitivity matrix and N initial average process values. The regional temperature compensation value matrix includes regional temperature compensation values ​​corresponding to N regions. Based on the regional temperature compensation value matrix, the initial temperature of each region is compensated accordingly so that the process value of each region meets the target process value.

[0007] Optionally, the method for obtaining the region sensitivity matrix of the wafer monitoring chip includes: A preset temperature variation is set for each of the N regions, and at least two first process values ​​are collected in the N regions at equal temperature intervals on both sides of the initial temperature within the preset temperature variation range. The preset temperature range includes the initial temperature. Under the condition of preset temperature change in each region, at least two first process values ​​corresponding to the same region and the initial average process value of the region are linearly fitted to obtain N region sensitivities corresponding to N regions under the temperature change condition of each region. Based on the N region sensitivities corresponding to N regions under preset temperature change conditions for each region, the region sensitivity matrix is ​​obtained. ,in, Used to characterize the regional sensitivity of the i-th region under a preset temperature change in the j-th region.

[0008] Optionally, the region sensitivity is used to characterize the slope of a straight line linearly fitted based on the initial average process value and the first process value corresponding to the same region.

[0009] Optionally, the method for obtaining the compensation coefficient and the regional temperature compensation value matrix includes: The N initial average process values ​​corresponding to the N regions are used as the initial process value matrix for the first time. Based on a number of preset initial compensation coefficients, the region sensitivity matrix and the initial process value matrix, iterative calculations are performed until the iteration stop condition is met. The initial compensation coefficient with the fewest iterations is determined as the compensation coefficient. The element in the i-th row of the first initial process value matrix is ​​the region's initial average process value for the i-th region. Substitute the compensation coefficient into the iterative calculation and sum the first temperature compensation value matrix generated in each iteration to obtain the regional temperature compensation value matrix.

[0010] Optionally, the method of determining the initial compensation coefficient as the compensation coefficient by iteratively calculating based on a preset number of initial compensation coefficients, the region sensitivity matrix, and the initial process value matrix until the iteration stopping condition is met includes: Set the iteration number n=1, and the current initial process value matrix is ​​the initial process value matrix of the first iteration; Several initial compensation coefficients are set, and the first temperature compensation value matrix for the nth time corresponding to each initial compensation coefficient is calculated according to a first preset formula, wherein the first preset formula is: ,in, This is the matrix of the first temperature compensation value for the nth time. For the overall sensitivity, The target process value, Let n be the initial process value matrix for the nth iteration. The initial compensation coefficient is... Calculate the expected process value matrix for each of the initial compensation coefficients at the nth time according to the second preset formula, whereby the second preset formula is: ,in, Let n be the matrix of the expected process values. This is the sensitivity matrix of the region; Determine whether the iteration stopping condition is met, wherein the iteration stopping condition is that the mean of all elements in the expected process value matrix is ​​equal to the target process value; If the mean of all elements in the expected process value matrix of the nth iteration is not equal to the target process value, the expected process value matrix of the nth iteration is used as the initial process value matrix of the (n+1)th iteration, and the (n+1)th iteration is performed. If the mean of all elements in the expected process value matrix in the nth iteration is equal to the target process value, stop the iteration and determine the iteration number corresponding to each initial compensation coefficient; compare the iteration numbers corresponding to all initial compensation coefficients, and determine the initial compensation coefficient with the fewest iterations as the compensation coefficient.

[0011] Optionally, all of the initial compensation coefficients are greater than 0 and less than 1, and any two of the initial compensation coefficients are not equal.

[0012] Optionally, the method for obtaining the regional temperature compensation value matrix includes: During the iterative calculation, the first temperature compensation value matrix generated in each iterative calculation is recorded. Substitute the confirmed compensation coefficients into the iterative process and re-execute the process, or directly call the first temperature compensation value matrix of the previous iterations that has been recorded. The temperature compensation value matrix of the region is obtained by summing the first temperature compensation value matrices corresponding to all iterations.

[0013] Optionally, the process value can be either the resistance value or the film thickness value.

[0014] According to a second aspect of the present invention, a temperature compensation system is provided, comprising: The information module is used to acquire the overall sensitivity and target process value of the wafer monitoring wafer, wherein the overall sensitivity is used to characterize the relationship between the process value of the wafer monitoring wafer and the overall process temperature. The sensitivity acquisition module is used to acquire the region sensitivity matrix of the wafer monitoring wafer. The region sensitivity matrix has N rows and N columns. The element located in the i-th row and j-th column of the region sensitivity matrix is ​​used to characterize the region sensitivity of the process value of the i-th region to the temperature change of the j-th region, 1≤i≤N, 1≤j≤N. The process value acquisition module is used to acquire N initial average process values ​​corresponding to N regions; A temperature compensation module is used to obtain a compensation coefficient, and further used to obtain a regional temperature compensation value matrix based on the compensation coefficient, the overall sensitivity, the target process value, the regional sensitivity matrix, and the initial average process value. The regional temperature compensation value matrix includes regional temperature compensation values ​​corresponding to N regions. Based on the regional temperature compensation value matrix, the initial temperature of each region is compensated accordingly so that the process value of each region meets the target process value.

[0015] According to a third aspect of the present invention, a computing device is provided, comprising: processor; A memory coupled to the processor is configured to store instructions that, when executed by the processor, cause the processor to perform the aforementioned semiconductor process temperature compensation method.

[0016] Compared with the prior art, the technical solution of the present invention has the following beneficial effects: In a temperature compensation method provided by this invention, the method includes acquiring a region sensitivity matrix of a wafer monitoring wafer. This region sensitivity matrix has N rows and N columns. The element in the i-th row and j-th column of the region sensitivity matrix characterizes the sensitivity of the process value of the i-th region to the temperature change of the j-th region, where 1 ≤ i ≤ N and 1 ≤ j ≤ N. Therefore, the region sensitivity matrix can be used to characterize the impact of adjusting the temperature of any region on the process value of each region. Based on this, a region temperature compensation value matrix is ​​obtained according to the compensation coefficient, overall sensitivity, target process value, region sensitivity matrix, and N initial average process values. The region temperature compensation value matrix includes N region temperature compensation values ​​corresponding to the N regions. Therefore, the acquisition of the region temperature compensation value matrix considers the influence of mutual thermal effects between lamps in different regions, resulting in high accuracy of the temperature compensation values ​​in the temperature compensation matrix.

[0017] Furthermore, since the initial compensation coefficient with the fewest iterations is determined as the compensation coefficient, the obtained compensation coefficient is more reliable. Based on this, iterative calculations are performed using several preset initial compensation coefficients, the sensitivity matrix, and the initial process value matrix until the iteration stop condition is met. The initial compensation coefficient with the fewest iterations is then determined as the compensation coefficient. Next, the compensation coefficient is substituted into the iterative calculation, and the first temperature compensation value matrix generated in each iteration is accumulated to obtain the regional temperature compensation value matrix. Therefore, after determining the compensation coefficient, overall sensitivity, and target process value, the regional temperature compensation value matrix obtained through iterative methods has high reliability, resulting in good uniformity of the adjusted wafer monitoring wafer. Moreover, during debugging, only the initial average process value of each region of the wafer monitoring wafer needs to be input to obtain the regional temperature compensation values ​​for each region required for debugging. This allows debugging to be completed using only one wafer monitoring wafer, effectively reducing the number of wafer monitoring wafers required during debugging, thereby reducing machine downtime waiting time and trial-and-error costs for annealing machines. Attached Figure Description

[0018] Figure 1 This is a flowchart of a method for adjusting temperature compensation in a rapid thermal annealing process; Figure 2 This is a flowchart of a semiconductor process temperature compensation method provided in an embodiment of the present invention; Figure 3 This is a flowchart of obtaining the region sensitivity matrix provided in an embodiment of the present invention; Figure 4 This is a flowchart of a method for obtaining compensation coefficients and a regional temperature compensation value matrix provided in an embodiment of the present invention; Figure 5 This is a flowchart of a method for obtaining compensation coefficients provided in an embodiment of the present invention; Figure 6 This is a flowchart of a method for further obtaining a regional temperature compensation value matrix provided in an embodiment of the present invention; Figure 7 This is a pull sensitivity test diagram provided by an embodiment of the present invention under a preset temperature change in the first region; Figures 8-11 This is a data graph showing the average value of the elements in the expected resistance value matrix during the iterative process provided in this embodiment of the invention. Detailed Implementation

[0019] As described in the background art, the mutual thermal interaction between lamps in different areas will cause changes in the process values ​​of other areas as well. This makes the process temperature compensation value of each area, which is directly calculated by monitoring the relationship between the process values ​​and temperature of the wafer, inaccurate and unreliable.

[0020] In view of this, the present invention creatively proposes a method for temperature compensation in semiconductor processes, comprising: A wafer monitoring chip is provided, wherein the wafer monitoring chip is divided into N regions, where N is a positive integer greater than 1; The overall sensitivity and target process value of the wafer monitoring chip are obtained, wherein the overall sensitivity is used to characterize the relationship between the process value of the wafer monitoring chip and the overall process temperature. At the initial temperature, obtain N initial average process values ​​corresponding to the N regions; Obtain the region sensitivity matrix of the wafer monitoring wafer. The region sensitivity matrix has N rows and N columns. The element in the i-th row and j-th column of the region sensitivity matrix is ​​used to characterize the region sensitivity of the process value of the i-th region to the temperature change of the j-th region, 1≤i≤N, 1≤j≤N. A compensation coefficient is obtained, and a regional temperature compensation value matrix is ​​obtained based on the compensation coefficient, the overall sensitivity, the target process value, the regional sensitivity matrix and N initial average process values. The regional temperature compensation value matrix includes regional temperature compensation values ​​corresponding to N regions. Based on the regional temperature compensation value matrix, the initial temperature of each region is compensated accordingly so that the process value of each region meets the target process value.

[0021] In this application, the regional temperature compensation value matrix takes into account the mutual thermal effects of lamps in different regions, resulting in high accuracy of temperature compensation values ​​in the matrix. Furthermore, by iteratively calculating appropriate compensation coefficients and accumulating the first temperature compensation value matrix generated in each iteration, the obtained regional temperature compensation value matrix exhibits high reliability, leading to good uniformity of the adjusted wafer monitoring wafer. Moreover, during debugging, only the initial average process values ​​of each region of the wafer monitoring wafer need to be input to obtain the required regional temperature compensation values ​​for each region. This allows debugging to be completed using only one wafer monitoring wafer, effectively reducing the number of wafer monitoring wafers required during debugging and consequently lowering the downtime waiting time for debugging and the trial-and-error costs of the annealing machine.

[0022] To make the above-mentioned objects, features, and beneficial effects of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. The terms "first," "second," "third," etc., in the specification, claims, and accompanying drawings of the present invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the present invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0023] This invention provides a temperature compensation method for adjusting the temperature compensation in a rapid thermal annealing process. Please refer to... Figure 2 The method includes: Step S1: Provide a wafer monitoring chip, which is divided into N regions, where N is a positive integer greater than 1; Step S2: Obtain the overall sensitivity and target process value of the wafer monitoring wafer. The overall sensitivity is used to characterize the relationship between the process value of the wafer monitoring wafer and the overall process temperature. Step S3: At the initial temperature, obtain N initial average process values ​​for the corresponding N regions; Step S4: Obtain the region sensitivity matrix of the wafer monitoring wafer. The region sensitivity matrix has N rows and N columns. The elements in the i-th row and j-th column of the region sensitivity matrix are used to characterize the region sensitivity of the process value of the i-th region to the temperature change of the j-th region, 1≤i≤N, 1≤j≤N. Step S5: Obtain a compensation coefficient. Based on the compensation coefficient, overall sensitivity, target process value, regional sensitivity matrix, and N initial average process values, obtain a regional temperature compensation value matrix. The regional temperature compensation value matrix includes the regional temperature compensation values ​​corresponding to the N regions. Step S6: Based on the regional temperature compensation value matrix, perform corresponding temperature compensation on the initial temperature of each region so that the process value of each region meets the target process value.

[0024] The steps in the above method are explained in detail below.

[0025] Regarding step S1, the wafer monitoring wafer is a test wafer used to monitor temperature. It helps determine whether the actual annealing temperature of the rapid thermal annealing equipment meets the set temperature. That is, the actual annealing temperature can be accurately determined by the correspondence between the process values ​​of the wafer monitoring wafer and the annealing temperature. Among them, the process values ​​can be, for example, the resistance value of a specific structure on the wafer monitoring wafer (such as a metal silicide layer or a doped layer) or the thickness of a specific film layer on the wafer monitoring wafer (such as a silicon dioxide layer) (hereinafter referred to as the film thickness value).

[0026] Regarding step S2, the overall sensitivity of the wafer monitoring chip can be directly obtained from the rapid thermal annealing equipment.

[0027] For step S3, the method for obtaining the N initial average process values ​​corresponding to the N regions includes: Step S31: Set multiple sampling points in each area; Step S32: Obtain the process values ​​for all sampling points; Step S33: Obtain the average value of the process values ​​of the sampling points in each region, and use them as the N initial average process values ​​for the corresponding N regions.

[0028] For step S4, please refer to... Figure 3 Methods for obtaining the region sensitivity matrix of a wafer monitoring chip include: Step S41: Set a preset temperature change for each of the N regions, and collect at least two first process values ​​in the N regions at equal temperature intervals on both sides of the initial temperature within the preset temperature change range. The preset temperature change range includes the initial temperature. Specifically, the first process value is the average process value of the region under the sampling temperature. The preset temperature variation range includes the initial temperature and the sampling temperature when the first process value is collected. The number of sampling temperatures is greater than or equal to two. The absolute value of the difference between the upper and lower limits of the preset temperature range and the initial temperature is greater than 0℃ and less than or equal to 10℃. Preferably, the preset temperature variation is ±5℃. The sampling temperature when the first process value is collected can be located on both sides of the initial temperature; that is, some sampling temperatures when the first process value is collected are lower than the initial temperature, and the remaining sampling temperatures are higher than the initial temperature. Correspondingly, taking an initial temperature of 0℃ and a preset temperature variation of ±5℃ as an example, the sampling temperatures when collecting the first process value can be -5℃ and 5℃.

[0029] Regarding step S41, it can be understood that any acquisition process in this step involves changing the temperature of any region to any acquisition temperature, and then acquiring the average process value of each region as the corresponding first process value. Furthermore, each region needs to have a preset temperature change set, and the acquisition temperature set for each region is the same. Therefore, if two acquisition temperatures are set, then 2N first process values ​​are obtained under the condition of temperature change in any region, and 2N values ​​can be obtained after step S41. 2 The first process value.

[0030] Step S42: Under the preset temperature change conditions in each region, at least two first process values ​​and the initial average process value corresponding to the same region are linearly fitted to obtain the sensitivity of N regions corresponding to N regions under the preset temperature change conditions in each region. In the process of obtaining the regional sensitivity of the j-th region to the preset temperature change of the i-th region, the first process value for linear fitting includes at least two first process values ​​of the j-th region collected when the preset temperature of the i-th region changes.

[0031] Step S43: Obtain the region sensitivity matrix based on the N region sensitivities corresponding to the N regions under the temperature change conditions of each region. ,in, Used to characterize the regional sensitivity of the i-th region under a preset temperature change in the j-th region.

[0032] In this embodiment, region sensitivity is used to characterize the slope of a straight line linearly fitted based on the initial average process value and the first process value corresponding to the same region. For example, please refer to... Figure 7 The wafer monitoring chip has four areas. Figure 7 To generate a sensitivity test chart for the pull-off under a preset temperature change in the first region, Figure 7 The horizontal axis represents the temperature of the first region, and the vertical axis represents the average process value of each region. Figure 7It can be seen that when the initial temperature of the first region is set to 0°C, and the preset temperature variation is ±5°C, at least two first process values ​​are collected in each of the four regions at equal temperature intervals. The collected temperatures are -5°C and 5°C. In each region, the slope of the straight line obtained by linearly fitting the initial process value and the first process value of the same region is the regional sensitivity of the corresponding region. That is, the regional sensitivity of the first region is... =-3.788, the regional sensitivity of the second region is =-2.019, the regional sensitivity of the third region is =-0.218, the regional sensitivity of the fourth region is =-2.602. Simultaneously, a preset temperature change is set for each region, and at least two initial process values ​​from all four regions are collected. These values ​​are then combined with the initial process values ​​to obtain the region sensitivity for each region, thus yielding 4. 2 Sensitivity of each region, corresponding to the region sensitivity matrix. .

[0033] For step S5, please refer to Figure 4 Methods for obtaining a compensation coefficient and a regional temperature compensation value matrix include: Step S51: Use the N initial average process values ​​corresponding to the N regions as the initial process value matrix for the first time. Based on several preset initial compensation coefficients, regional sensitivity matrix and initial process value matrix, iterative calculation is performed until the iteration stop condition is met. The initial compensation coefficient with the fewest iterations is determined as the compensation coefficient. Among them, the element in the i-th row of the initial process value matrix in the first iteration is the regional initial average process value of the i-th region. Step S52: Substitute the compensation coefficients into the iterative calculation and accumulate the first temperature compensation value matrix generated in each iteration to obtain the regional temperature compensation value matrix. In this embodiment, please refer to... Figure 5 Step S51 specifically includes: Step S511: Set the iteration number n=1, and the current initial process value matrix is ​​the initial process value matrix of the first iteration; Step S512: Set several initial compensation coefficients; In this embodiment, several initial compensation coefficients are all greater than 0 and less than 1, and any two initial compensation coefficients are not equal.

[0034] Step S513: Calculate the first temperature compensation value matrix for the nth time corresponding to each initial compensation coefficient according to the first preset formula. The first preset formula is: ,in, This is the matrix of the first temperature compensation value for the nth time. For overall sensitivity, For the target process value, This is the initial process value matrix for the nth iteration. The initial compensation coefficient; Step S514: Calculate the expected process value matrix for the nth iteration corresponding to each initial compensation coefficient according to the second preset formula. The second preset formula is: ,in, This is the expected process value matrix for the nth iteration. This is the region sensitivity matrix; Step S515 determines whether the iteration stopping condition is met, including: if the iteration stopping condition is not met, proceed to step S516; if the iteration stopping condition is met, proceed to steps S517 and S518 sequentially. The iteration stopping condition is that the mean of the elements in the expected process value matrix equals the target process value.

[0035] Step S516: If the mean of all elements in the expected process value matrix of the nth iteration is not equal to the target process value, use the expected process value matrix of the nth iteration as the initial process value matrix of the (n+1)th iteration and perform the (n+1)th iteration. Step S517: If the mean of all elements in the expected process value matrix of the nth iteration is equal to the target process value, stop the iteration and determine the iteration number corresponding to each initial compensation coefficient. Step S518: Compare the number of iterations corresponding to all initial compensation coefficients, and determine the initial compensation coefficient with the fewest iterations as the compensation coefficient.

[0036] As an example, the initial compensation coefficients are set to 0.8, 0.6, 0.5, and 0.3. Please refer to [the documentation / reference]. Figures 8-11 , Figures 8-11 This is a data graph showing the average value of the elements in the expected process value matrix during the iteration process. The horizontal axis represents the number of iterations, and the vertical axis represents the average value of the elements in the expected process value matrix. Figure 8 The data graph shows the initial compensation coefficient as 0.8. Figure 9 The data graph shows the initial compensation coefficient as 0.6. Figure 10 The data graph shows the initial compensation coefficient as 0.5. Figure 11 This is a data graph with an initial compensation coefficient of 0.3. (Reference) Figures 8-11 The target process value is 178. Figure 8 The data in the middle is divergent. Figure 9 The data converged to 178 after 60 iterations. Figure 10 The data converged to 178 after 30 iterations. Figure 11 The data converged to 178 after 40 iterations. Therefore, since 30 < 40 < 60, the compensation coefficient was determined to be 0.5.

[0037] In this embodiment, please refer to Figure 6 Step S52 specifically includes: Step S521: During the iterative calculation, record the first temperature compensation value matrix generated in each iterative calculation. Step S522: Substitute the confirmed compensation coefficients into the iterative process and re-execute the process, or directly call the first temperature compensation value matrix of each of the previous iterations that has been recorded. Step S523: Summate the first temperature compensation value matrices corresponding to all iterations to obtain the region temperature compensation value matrix.

[0038] It should be noted that after determining the compensation coefficient, when using this method in the future, the determined compensation coefficient can be directly used to replace the initial temperature compensation coefficient to execute steps S511 to S517 to obtain the regional temperature compensation value matrix.

[0039] In summary, in the temperature compensation method provided in this embodiment of the invention, the method includes obtaining a region sensitivity matrix of the wafer monitoring wafer. This region sensitivity matrix has N rows and N columns. The element in the i-th row and j-th column of the region sensitivity matrix is ​​used to characterize the sensitivity of the process value of the i-th region to the temperature change of the j-th region, where 1 ≤ i ≤ N and 1 ≤ j ≤ N. Therefore, the region sensitivity matrix can be used to characterize the impact of adjusting the temperature of any region on the process value of each region. Based on this, a region temperature compensation value matrix is ​​obtained according to the compensation coefficient, overall sensitivity, target process value, region sensitivity matrix, and N initial average process values. The region temperature compensation value matrix includes N region temperature compensation values ​​corresponding to the N regions. Therefore, the acquisition of the region temperature compensation value matrix considers the influence of mutual thermal effects between lamps in different regions, resulting in high accuracy of the temperature compensation values ​​in the region temperature compensation matrix.

[0040] Furthermore, since the initial compensation coefficient with the fewest iterations is determined as the compensation coefficient, the obtained compensation coefficient is more reliable. Based on this, iterative calculations are performed using several preset initial compensation coefficients, sensitivity matrices, and initial process value matrices until the iteration stop condition is met. The initial compensation coefficient with the fewest iterations is then determined as the compensation coefficient. Next, the compensation coefficient is substituted into the iterative calculation, and the first temperature compensation value matrix generated in each iteration is accumulated to obtain the regional temperature compensation value matrix. Therefore, after determining the compensation coefficient, overall sensitivity, and target process value, the regional temperature compensation value matrix obtained through the iterative method has high reliability, resulting in good uniformity of the adjusted wafer monitoring wafer. Moreover, during debugging, only the initial average process value of each region of the wafer monitoring wafer needs to be input to obtain the regional temperature compensation values ​​for each region required for debugging. This allows debugging to be completed using only one wafer monitoring wafer, effectively reducing the number of wafer monitoring wafers required during debugging, thereby reducing machine downtime waiting time and trial-and-error costs for annealing machines.

[0041] This invention also provides a temperature compensation system, characterized in that it includes: an information module, a process value acquisition module, a sensitivity acquisition module, and a temperature compensation module.

[0042] The system comprises the following modules: an information module for acquiring the overall sensitivity and target process value of the wafer monitoring wafer, whereby the overall sensitivity characterizes the relationship between the process value and the overall process temperature; a sensitivity acquisition module for acquiring the regional sensitivity matrix of the wafer monitoring wafer, which has N rows and N columns, where the element in the i-th row and j-th column characterizes the regional sensitivity of the process value of the i-th region to the temperature change of the j-th region, where 1 ≤ i ≤ N and 1 ≤ j ≤ N; a process value acquisition module for acquiring N initial average process values ​​corresponding to the N regions; and a temperature compensation module for acquiring a compensation coefficient, which, based on the compensation coefficient, overall sensitivity, target process value, regional sensitivity matrix, and initial average process value, obtains a regional temperature compensation value matrix, which includes the regional temperature compensation values ​​for the corresponding N regions. Based on the regional temperature compensation value matrix, the initial temperature of each region is compensated accordingly to ensure that the process value of each region meets the target process value.

[0043] This invention also provides a computing device, including a processor and a memory.

[0044] The memory is coupled to the processor and configured to store instructions that, when executed by the processor, cause the processor to perform the aforementioned semiconductor process temperature compensation method.

[0045] While the present invention has been disclosed above, it is not limited thereto. Any person skilled in the art can make various modifications and alterations without departing from the spirit and scope of the invention; therefore, the scope of protection of the present invention should be determined by the scope defined in the claims.

Claims

1. A method of semiconductor process temperature compensation, comprising: The method comprises: providing a wafer monitoring sheet, the wafer monitoring sheet being divided into N regions, N being a positive integer greater than 1; obtaining the overall sensitivity of the wafer monitoring sheet and a target process value, the overall sensitivity being used to represent the relationship between the process value of the wafer monitoring sheet and the overall process temperature; obtaining N initial average process values corresponding to the N regions at an initial temperature; obtaining a region sensitivity matrix of the wafer monitoring sheet, the region sensitivity matrix having N rows and N columns, an element located at the i-th row and the j-th column in the region sensitivity matrix being used to represent the region sensitivity of the process value of the i-th region to the temperature change of the j-th region, 1≤i≤N, 1≤j≤N; obtaining a compensation coefficient, and obtaining a region temperature compensation value matrix according to the compensation coefficient, the overall sensitivity, the target process value, the region sensitivity matrix, and the N initial average process values, the region temperature compensation value matrix comprising region temperature compensation values corresponding to the N regions; performing corresponding temperature compensation on the initial temperature of each region according to the region temperature compensation value matrix, so that the process value of each region meets the target process value.

2. The method of claim 1, wherein, The method for obtaining the region sensitivity matrix of the wafer monitoring sheet comprises: setting a preset temperature change for each of the N regions, and collecting at least two first process values in each region at an equal temperature interval on both sides of the initial temperature within the preset temperature change range, the preset temperature range comprising the initial temperature; under the condition of the preset temperature change of each region, performing linear fitting on at least two first process values corresponding to the same region and the initial average process value to obtain N region sensitivities corresponding to the N regions under the condition of the preset temperature change of each region. According to the preset temperature change condition of each region, the N region sensitivities corresponding to the N regions under the preset temperature change condition of each region are obtained to obtain the region sensitivity matrix wherein, is used to represent the region sensitivity of the ith region under the preset temperature change condition of the jth region.

3. The method of claim 2, wherein the semiconductor process temperature compensation is determined by: The region sensitivity is used to represent the slope of a straight line linearly fitted according to the initial average process value and the first process value of the same region.

4. The method of claim 1, wherein, The method for obtaining the compensation coefficient and the region temperature compensation value matrix comprises: N initial average process values of the N regions are taken as an initial process value matrix for the first time , based on a plurality of preset initial compensation coefficients, the region sensitivity matrix and the initial process value matrix, iterative calculation is performed until an iteration stopping condition is met, and the initial compensation coefficient with the least iteration times is determined as the compensation coefficient, wherein an element in the ith row of the initial process value matrix for the first time is a region initial average process value of the ith region; substituting the compensation coefficient into the iterative calculation, and accumulating the first temperature compensation value matrix generated in each iteration to obtain the region temperature compensation value matrix.

5. The method of claim 4, wherein the temperature of the semiconductor process is determined by a temperature sensor. The method for performing iterative calculation based on a plurality of preset initial compensation coefficients, the region sensitivity matrix, and the initial process value matrix until the iteration stopping condition is met, and the initial compensation coefficient with the least iteration number is determined as the compensation coefficient comprises: setting the iteration number n = 1, and the current initial process value matrix being the initial process value matrix in the first iteration; A plurality of initial compensation coefficients are set, and a first temperature compensation value matrix corresponding to each initial compensation coefficient is calculated according to a first preset formula wherein, is the first temperature compensation value matrix, is the overall sensitivity, is the target process value, is the initial process value matrix, is the initial compensation coefficient; calculating an nth-time expected process value matrix corresponding to each of the initial compensation coefficients according to a second preset formula, the second preset formula being wherein, is the nth-time expected process value matrix, is the area sensitivity matrix; determining whether the iteration stopping condition is met, wherein the iteration stopping condition is that the average value of all elements in the expected process value matrix is equal to the target process value; if the average value of all elements in the expected process value matrix in the n-th iteration is not equal to the target process value, taking the expected process value matrix in the n-th iteration as the initial process value matrix in the n+1-th iteration to perform the n+1-th iteration; If the average of all elements in the nth iteration of the expected process value matrix is equal to the target process value, stop the iteration, determine the iteration number corresponding to each initial compensation coefficient; compare the iteration numbers corresponding to all initial compensation coefficients, and determine the initial compensation coefficient with the least iteration number as the compensation coefficient.

6. The method of claim 4, wherein, If the initial compensation coefficients are all greater than 0 and less than 1, and any two of the initial compensation coefficients are not equal.

7. The method of claim 4, wherein the temperature of the semiconductor process is determined by a temperature sensor. The method for obtaining the area temperature compensation value matrix comprises: During the iteration calculation, record the first temperature compensation value matrix generated by each iteration calculation; Substitute the confirmed compensation coefficient into the iteration process and re-execute the iteration process, or directly call the recorded first temperature compensation value matrix of the previous iteration; Sum the corresponding first temperature compensation value matrices in all iteration numbers to obtain the area temperature compensation value matrix.

8. The method of semiconductor process temperature compensation according to any one of claims 1 to 7, wherein, The process value is a resistance value or a film thickness value.

9. A temperature compensation system, characterized by Comprise: An information module for obtaining an overall sensitivity of a wafer monitoring piece and a target process value, the overall sensitivity being used to represent the relationship between the process value of the wafer monitoring piece and the overall process temperature; A sensitivity acquisition module for obtaining an area sensitivity matrix of the wafer monitoring piece, the area sensitivity matrix having N rows and N columns, and the element located at the ith row and the jth column in the area sensitivity matrix being used to represent the area sensitivity of the ith area to the temperature change of the jth area, 1≤i≤N, 1≤j≤N; A process value acquisition module for obtaining N initial average process values corresponding to N areas; A temperature compensation module for obtaining a compensation coefficient, and the temperature compensation module is further used to obtain an area temperature compensation value matrix according to the compensation coefficient, the overall sensitivity, the target process value, the area sensitivity matrix and the initial average process value, the area temperature compensation value matrix comprising area temperature compensation values corresponding to N areas; According to the area temperature compensation value matrix, compensate the initial temperature of each area correspondingly, so that the process value of each area meets the target process value.

10. A computing device, comprising: Comprise: A processor; A memory coupled to the processor and configured to store instructions which, when executed by the processor, cause the processor to perform the method for semiconductor process temperature compensation according to any one of claims 1 to 8.

Images