Wafer state monitoring method, device, equipment and medium for RTP process

Abstract

This invention discloses a wafer status monitoring method, apparatus, device, and medium for RTP (Real-Time Trial) processes, belonging to the field of semiconductor manufacturing technology. The wafer status monitoring method includes the following steps: S1: acquiring real-time emissivity data of the wafer at the process position; S2: calculating the emissivity change rate based on the real-time emissivity data acquired in step S1; S3: comparing the emissivity change rate acquired in step S2 with a preset first threshold to obtain the wafer status. When the wafer status is abnormal, the process is stopped and an alarm is issued. This invention acquires emissivity and calculates the emissivity change rate, monitors the wafer status using the emissivity change rate and a first threshold, and captures transient signal changes caused by slight wafer skipping through the emissivity change rate. Intervention and alarm are initiated at the initial stage of warpage to prevent the accident from escalating, thereby improving wafer quality.

Description

Technical Field

[0001] This invention belongs to the field of semiconductor manufacturing technology, and particularly relates to a method, apparatus, equipment and medium for wafer condition monitoring in RTP process. Background Technology

[0002] In semiconductor chip manufacturing, rapid thermal processing (RTP) is a critical process for rapidly heating and cooling wafers. During high-temperature processes, wafers are highly susceptible to warping due to uneven thermal stress. In severe cases, warping can cause the wafer to bounce off the edge ring supporting it, a phenomenon known as "wafer skipping" or "wafer flying." This not only causes the wafer to break upon impact, contaminating the process chamber and damaging equipment, but more dangerously, sometimes the skipping is small enough that the wafer falls back onto the edge ring, allowing the process to continue undetected and impacting product yield.

[0003] Existing RTP equipment typically includes multiple infrared probes for temperature control, and some also have emissivity monitoring capabilities, but only set fixed upper and lower limits for emissivity to trigger alarms. The drawback of this approach is that slight wafer warping may not cause the absolute value of emissivity to exceed the fixed limit, but it is enough to disrupt process uniformity, leading to product yield degradation. Furthermore, this problem is often insidious and difficult to trace in subsequent stages. Therefore, there is an urgent need for a method that can detect instantaneous wafer warping or micro-wafer skipping in real time with high sensitivity.

[0004] The existing technology has the following drawbacks:

[0005] 1. It is impossible to monitor the process in real time;

[0006] Second: Additional hardware (laser, camera) is required, resulting in high costs;

[0007] Third: It usually only alarms after the fault occurs, which is a "post-fault" detection;

[0008] Fourth: A multi-channel thermometer capable of measuring emissivity is required, resulting in high hardware costs;

[0009] 5. When a slight wafer skipping occurs, it cannot provide an accurate early warning. Summary of the Invention

[0010] Based on the technical problems existing in the prior art, the present invention provides a method, apparatus, equipment and medium for wafer status monitoring in RTP process.

[0011] According to a first aspect of the present invention, a wafer condition monitoring method for RTP process is provided, comprising the following steps:

[0012] S1: Obtain real-time emissivity data of the wafer when it is in the process position;

[0013] S2: Calculate the rate of change of emissivity based on the real-time emissivity data obtained in step S1;

[0014] S3: Compare the rate of change of emissivity obtained in step S2 with a preset first threshold to obtain the wafer status. When the wafer status is abnormal, stop the process and issue an alarm.

[0015] A further improvement of the present invention is that step S1 includes the following steps:

[0016] S11: Obtain raw emissivity data;

[0017] S12: Use a filtering algorithm to remove noise from the original emissivity data obtained in step S11 to obtain denoised data;

[0018] S13: Use a machine learning algorithm to remove outliers from the denoised data obtained in step S12 to obtain real-time emissivity data.

[0019] A further improvement of the present invention is that, in S2, the emissivity change rate is calculated using one of the following methods: the first-order difference method, the window standard deviation calculation method, or the absolute value deviation calculation method based on a preset process baseline value.

[0020] A further improvement of the present invention is that calculating the preset first threshold includes the following steps:

[0021] S31A: Obtain historical normal process data;

[0022] S32A: Calculate the historical maximum emissivity change rate based on the historical normal process data obtained in step S31A;

[0023] S33A: The preset first threshold is calculated based on the historical maximum rate of change of radiation and the preset first safety factor.

[0024] A further improvement of the present invention is that calculating the preset first threshold includes the following steps:

[0025] S31B: Obtain the current process stage, which includes a rapid heating stage, a constant temperature stage, and a slow heating stage;

[0026] S32B: Based on the process stage obtained in step S31B, a preset second safety factor is obtained;

[0027] S33B: The preset first threshold is calculated based on the emissivity change rate obtained in step S2 and the second safety factor.

[0028] A further improvement of the present invention is that calculating the preset first threshold includes the following steps:

[0029] S31C: Obtain the rate of change of emissivity during the initial time period;

[0030] S32C: The mean and standard deviation of the rate of change of emissivity calculated based on the rate of change of emissivity during the initial time period;

[0031] S33C: The initial threshold is calculated based on the preset dynamic threshold coefficient, the mean of the rate of change of emissivity, and the standard deviation of the rate of change of emissivity;

[0032] S34: The rate of change of radiance in subsequent time periods is obtained by using a sliding window algorithm to update the initial threshold in real time, thereby obtaining a preset first threshold.

[0033] A further improvement of the present invention is that S3 includes the following steps:

[0034] S31: Obtain the preset first threshold;

[0035] S32: Compare the rate of change of emissivity with the first threshold. When the rate of change of emissivity is greater than or equal to the first threshold, the wafer is in an abnormal state. When the rate of change of emissivity is less than the first threshold, the wafer is in a normal state.

[0036] S33: When the wafer status is abnormal, a control command is generated to stop the current process and an alarm is issued.

[0037] According to a second aspect of the present invention, a wafer condition monitoring device for RTP process is provided, comprising the following steps:

[0038] The data acquisition module is used to acquire real-time emissivity data of the wafer when it is in the process position;

[0039] The rate of change calculation module is used to calculate the rate of change of emissivity based on the real-time emissivity data.

[0040] The status judgment module is used to compare the rate of change of emissivity with a preset first threshold to obtain the wafer status. When the wafer status is abnormal, the process is stopped and an alarm is issued.

[0041] According to a third aspect of the present invention, an electronic device is provided, comprising:

[0042] One or more processors;

[0043] Storage device for storing one or more computer programs.

[0044] When the one or more computer programs are executed by the one or more processors, the one or more processors implement the above-described wafer condition monitoring method for RTP process.

[0045] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, on which a computer program is stored, wherein the computer program, when executed by a processor, implements the above-described wafer state monitoring method for RTP process.

[0046] Compared with the prior art, the above-mentioned technical solution of the present invention has the following beneficial technical effects:

[0047] This invention calculates the rate of change of emissivity by acquiring emissivity, monitors the wafer state by using the rate of change of emissivity and a first threshold, and captures instantaneous signal changes caused by slight wafer skipping by capturing the rate of change of emissivity. It intervenes and alarms at the early stage of warping to prevent the accident from escalating, thereby improving wafer quality. Attached Figure Description

[0048] The accompanying drawings are provided to better understand the invention and are not intended to unduly limit the scope of the invention. Wherein:

[0049] Figure 1 This is a flowchart of a wafer condition monitoring method for RTP process according to the present invention;

[0050] Figure 2 This is a structural block diagram of a wafer condition monitoring device for RTP process according to the present invention;

[0051] Figure 3 This is a schematic diagram of emissivity comparison in Embodiment 1 of the present invention;

[0052] Figure 4 This is a comparison chart of the response time of the solution in Embodiment 1 of the present invention and the existing solution.

[0053] Figure 5 This is a schematic diagram of the computer system in an embodiment of the present invention. Detailed Implementation

[0054] The following description, in conjunction with the accompanying drawings, illustrates exemplary embodiments of the present invention, including various details to aid understanding. These details should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0055] This invention discloses a wafer status monitoring method, apparatus, device, and medium for RTP (Real-Time Trial) processes, belonging to the field of semiconductor manufacturing technology. The wafer status monitoring method includes the following steps: S1: acquiring real-time emissivity data of the wafer at the process position; S2: calculating the emissivity change rate based on the real-time emissivity data acquired in step S1; S3: comparing the emissivity change rate acquired in step S2 with a preset first threshold to obtain the wafer status. When the wafer status is abnormal, the process is stopped and an alarm is issued. This invention acquires emissivity and calculates the emissivity change rate, monitors the wafer status using the emissivity change rate and a first threshold, and captures transient signal changes caused by slight wafer skipping through the emissivity change rate. Intervention and alarm are initiated at the initial stage of warpage to prevent the accident from escalating, thereby improving wafer quality.

[0056] The following description, in conjunction with specific embodiments and accompanying drawings, further illustrates a wafer condition monitoring method, apparatus, device, and medium for RTP process according to the present invention.

[0057] Example 1

[0058] like Figure 1 As shown, this invention provides a wafer condition monitoring method for RTP (Rapid Thermal Processing) technology, which includes the following steps:

[0059] S1: Obtain real-time emissivity data of the wafer when it is in the process position;

[0060] S2: Calculate the rate of change of emissivity based on the real-time emissivity data obtained in step S1;

[0061] S3: Compare the rate of change of emissivity obtained in step S2 with a preset first threshold to obtain the wafer status. When the wafer status is abnormal, stop the process and issue an alarm.

[0062] Specifically, in S1, the process position is located inside the chamber. The wafer is placed on the lifting pin of the rapid thermal processing equipment, and the height of the wafer is controlled by the lifting pin until it reaches the process position, so that the wafer is stably placed on the edge ring.

[0063] Specifically, step S1 includes the following steps:

[0064] S11: Obtain raw emissivity data;

[0065] S12: Use a filtering algorithm to remove noise from the original emissivity data obtained in step S11 to obtain denoised data;

[0066] S13: Use a machine learning algorithm to remove outliers from the denoised data obtained in step S12 to obtain real-time emissivity data.

[0067] Specifically, in S11, raw emissivity data is acquired through several infrared probes, which are installed on a rapid thermal processing device to acquire the temperature and emissivity of the wafer.

[0068] Preferably, there are 7 infrared probes, of which 6 are infrared temperature probes, and 1 of them measures both temperature and emissivity.

[0069] Specifically, in S12, the filtering algorithm includes one of wavelet transform, Kalman filtering, or median filtering. The wavelet transform is used to remove high-frequency noise while preserving abrupt changes, improving data smoothness. The Kalman filter is used for real-time data stream processing, predicting the current value based on historical data and correcting for noise effects, thus improving data accuracy. The median filter is used to handle sudden outliers (e.g., sudden changes in conductivity), reducing the impact of impulse noise on data stability.

[0070] Specifically, the outliers mentioned in S13 are caused by sensor malfunctions or other environmental factors. Outlier processing is performed using one of the following methods: machine learning algorithm (e.g., Isolation Forest algorithm or Local Outlier Algorithm), interpolation method, or elimination method. The machine learning algorithm is used to detect outliers in complex data patterns. The interpolation method is used to correct slightly outlier data by linear interpolation or mean difference. For severely outlier data, elimination method is used to directly delete the outlier data to prevent individual outliers from affecting the overall analysis results.

[0071] Specifically, the rate of change of emissivity is calculated using one of the following methods: the first-order difference method, the window standard deviation method, or the absolute value of the deviation based on a preset process baseline value.

[0072] Specifically, when calculating the rate of change of emissivity using the first-order difference method, the following formula is used:

[0073] r(t) = [ε(t) - ε(t - Δt)] / Δt;

[0074] In the formula, r(t) is the rate of change of radiance at time t calculated using the first-order difference method; ε(t) is the real-time radiance data; ε(t-Δt) is the real-time radiance data at the previous sampling time; Δt is the sampling time interval, which is usually 0.005-0.1 seconds, corresponding to a sampling rate of 10-200Hz.

[0075] Specifically, when calculating the rate of change of emissivity using the window standard deviation method, the following formula is used:

[0076]

[0077] In the formula, n is the window size, for example, a window that includes five sampling points; This is the real-time emissivity data for the k-th sampling point within the window. The arithmetic mean of n real-time emissivity data points within the window. The rate of change of emissivity is calculated using the window standard deviation method.

[0078] Specifically, when calculating the rate of change of emissivity using the absolute value of the deviation based on a preset process baseline value, the following formula is used:

[0079] |Δε|=|ε i -ε baseline |;

[0080] In the formula, |Δε| represents the rate of change of emissivity when using the absolute value of the deviation calculation method based on the preset process baseline value; ε i Let ε be the real-time emissivity data at time i. baseline This is a preset process baseline value. The absolute deviation is the absolute value of the difference between the current emissivity value and a preset "baseline emissivity" representing a stable process. It reflects the "static" offset of the overall wafer orientation.

[0081] Specifically, after calculating the rate of change of emissivity using the first-order difference method, calculating the preset first threshold includes the following steps:

[0082] S31A: Obtain historical normal process data;

[0083] S32A: Calculate the historical maximum emissivity change rate based on the historical normal process data obtained in step S31A;

[0084] S33A: The preset first threshold is calculated based on the historical maximum rate of change of radiation and the preset first safety factor.

[0085] Specifically, in S31A, the historical normal process data refers to the real-time emissivity data when no abnormalities such as wafer skipping occur.

[0086] Specifically, in S32A, the first-order difference method is used to calculate the historical emissivity change rate corresponding to the historical normal process data, and then the largest historical emissivity change rate is selected from several historical emissivity change rates as the historical maximum emissivity change rate.

[0087] Specifically, in S33A, the first threshold is calculated using the following formula:

[0088] K static =max(|r(t)|)×k safe1 ;

[0089] In the formula, k safe1The first safety factor is set based on experience, preferably 1.5-2; K static The first threshold is calculated according to steps S31-S33.

[0090] Specifically, calculating the preset first threshold includes the following steps:

[0091] S31B: Obtain the current process stage, which includes a rapid heating stage, a constant temperature stage, and a slow heating stage;

[0092] S32B: Based on the process stage obtained in step S31B, obtain the preset second safety factor k. safe2 ;

[0093] S33B: The preset first threshold is calculated based on the emissivity change rate obtained in step S2 and the second safety factor.

[0094] Preferably, in S32B, the second safety factor for the rapid heating stage is 1.5-2, the second safety factor for the constant temperature stage is 0.8-1, and the second safety factor for the slow heating stage is 1-1.2.

[0095] Specifically, in S33B, the preset first threshold is calculated using the following formula:

[0096] K auto =(μ+3σ)×k safe2 ;

[0097] In the formula, K auto k is the first threshold calculated according to steps S31B-S33B. safe2 The second safety factor is μ, which is the average of the maximum absolute values ​​of the rate of change of emissivity for each process curve. The standard deviation is the rate of change of emissivity calculated using the window standard deviation method.

[0098] Specifically, calculating the preset first threshold includes the following steps:

[0099] S31C: Obtain the rate of change of emissivity during the initial time period;

[0100] S32C: The mean and standard deviation of the rate of change of emissivity calculated based on the rate of change of emissivity during the initial time period;

[0101] S33C: The initial threshold is calculated based on the preset dynamic threshold coefficient, the mean of the rate of change of emissivity, and the standard deviation of the rate of change of emissivity;

[0102] S34: The rate of change of radiance in subsequent time periods is obtained by using a sliding window algorithm to update the initial threshold in real time, thereby obtaining a preset first threshold.

[0103] Specifically, in S31C, the initial time period is the time period between the start of the process and the preset time, such as the time period from the start of the process to 100ms.

[0104] Specifically, in S32C, the mean rate of change of emissivity μ over 100 ms is calculated. init and the standard deviation of the rate of change of emissivity σ init ;

[0105] Specifically, the initial threshold is calculated using the following formula:

[0106] K initial =μ init +A×σ init ;

[0107] In the formula, K initial is the initial threshold, and A is the preset dynamic threshold coefficient.

[0108] Specifically, in S34, the real-time update of the initial threshold is based on the following formula:

[0109] window data =r(t-N+1-m)tor(tm); to represents the window range from t-N+1-m to tm;

[0110] In the formula, N is the window length and m is the window lag time;

[0111] current μ =mean(window data );

[0112] In the formula, current μ This represents the average rate of change of the current window emissivity.

[0113] current σ =std(window data );

[0114] In the formula, current σ This represents the standard deviation of the current window emissivity change rate.

[0115] K dynamic =current μ +A×current σ ;

[0116] In the formula, K dynamicThe first threshold is calculated according to steps S31C-S34C, and A is a preset dynamic threshold coefficient.

[0117] Specifically, S3 includes the following steps:

[0118] S31: Obtain the preset first threshold;

[0119] S32: Compare the rate of change of emissivity with the first threshold. When the rate of change of emissivity is greater than or equal to the first threshold, the wafer is in an abnormal state. When the rate of change of emissivity is less than the first threshold, the wafer is in a normal state.

[0120] S33: When the wafer status is abnormal, a control command is generated to stop the current process and an alarm is issued.

[0121] Specifically, stopping the current process includes stopping heating and recording the current process stage, the current real-time emissivity, and the current rate of change of emissivity as historical data for subsequent monitoring.

[0122] Specifically, the emissivity during normal manufacturing processes and the emissivity during wafer skipping are compared to, for example... Figure 3 As shown. This embodiment compares the response time of a wafer condition monitoring method for RTP process with that of existing technologies when facing wafer skipping, as shown below. Figure 4 As shown, the entire process is monitored, with a response time of less than 50ms.

[0123] Example 2

[0124] like Figure 2 As shown, a wafer condition monitoring device for RTP process includes:

[0125] The data acquisition module is used to acquire real-time emissivity data of the wafer when it is in the process position;

[0126] The rate of change calculation module is used to calculate the rate of change of emissivity based on the real-time emissivity data.

[0127] The status judgment module is used to compare the rate of change of emissivity with a preset first threshold to obtain the wafer status. When the wafer status is abnormal, the process is stopped and an alarm is issued.

[0128] Specifically, in the data acquisition module, the process position is located inside the chamber. By placing the wafer on the lifting pin of the rapid thermal processing equipment, the height of the wafer is controlled by the lifting pin until it reaches the process position, so that the wafer is stably placed on the edge ring.

[0129] Specifically, the data acquisition module includes:

[0130] The initial data acquisition unit is used to acquire raw emissivity data;

[0131] A denoising unit is used to remove noise from the original emissivity data using a filtering algorithm to obtain denoised data.

[0132] The outlier removal unit is used to remove outliers from the denoised data using a machine learning algorithm to obtain real-time emissivity data.

[0133] Specifically, in the initial data acquisition unit, raw emissivity data is acquired through several infrared probes, which are set on a rapid thermal processing device to acquire the temperature and emissivity of the wafer.

[0134] Preferably, there are 7 infrared probes, of which 6 are infrared temperature probes, and 1 of them measures both temperature and emissivity.

[0135] Specifically, in the denoising unit, the filtering algorithm includes one of wavelet transform, Kalman filtering, or median filtering. Wavelet transform is used to remove high-frequency noise while preserving abrupt changes, improving data smoothness. Kalman filtering is used for real-time data stream processing, predicting current values ​​based on historical data and correcting for noise effects, thus improving data accuracy. Median filtering is used to address sudden outliers (e.g., sudden changes in conductivity), reducing the impact of impulse noise on data stability.

[0136] Specifically, in the outlier removal unit, the outliers are caused by sensor malfunctions or other environmental factors. Outlier processing is performed using one of the following methods: machine learning algorithm (e.g., Isolation Forest algorithm or Local Outlier Algorithm), interpolation method, or elimination method. The machine learning algorithm is used to detect outliers in complex data patterns. The interpolation method is used to correct slightly outlier data by linear interpolation or mean difference. For severely outlier data, the elimination method is used to directly delete the outlier data to prevent individual outliers from affecting the overall analysis results.

[0137] Specifically, the rate of change of emissivity is calculated using one of the following methods: the first-order difference method, the window standard deviation method, or the absolute value of the deviation based on a preset process baseline value.

[0138] Specifically, when calculating the rate of change of emissivity using the first-order difference method, the following formula is used:

[0139] r(t) = [ε(t) - ε(t - Δt)] / Δt;

[0140] In the formula, r(t) is the rate of change of radiance at time t calculated using the first-order difference method; ε(t) is the real-time radiance data; ε(t-Δt) is the real-time radiance data at the previous sampling time; Δt is the sampling time interval, which is usually 0.005-0.1 seconds, corresponding to a sampling rate of 10-200Hz.

[0141] Specifically, when calculating the rate of change of emissivity using the window standard deviation method, the following formula is used:

[0142]

[0143] In the formula, n is the window size, for example, a window that includes five sampling points; This is the real-time emissivity data for the k-th sampling point within the window. The arithmetic mean of n real-time emissivity data points within the window. This represents the rate of change of emissivity calculated using the window standard deviation method.

[0144] Specifically, when calculating the rate of change of emissivity using the absolute value of the deviation based on a preset process baseline value, the following formula is used:

[0145] |Δε|=|ε i -ε baseline |;

[0146] In the formula, |Δε| represents the rate of change of emissivity when using the absolute value of the deviation calculation method based on the preset process baseline value; ε i Let ε be the real-time emissivity data at time i. baseline This is a preset process baseline value. The absolute deviation is the absolute value of the difference between the current emissivity value and a preset "baseline emissivity" representing a stable process. It reflects the "static" offset of the overall wafer orientation.

[0147] Specifically, the loading judgment module includes several calculation units, which are used to calculate the first threshold. The calculation units include a first calculation unit, a second calculation unit, and a third calculation unit.

[0148] Specifically, the first computing unit includes:

[0149] The historical data acquisition subunit is used to acquire historical normal process data;

[0150] The historical calculation subunit is used to calculate the historical maximum emissivity change rate based on the historical normal process data.

[0151] The first calculation subunit is used to calculate a preset first threshold based on the historical maximum emissivity change rate and a preset first safety factor.

[0152] Specifically, in the historical data acquisition subunit, the historical normal process data is the real-time emissivity data when no abnormalities such as wafer skipping occur.

[0153] Specifically, in the historical calculation subunit, the first-order difference method is used to calculate the historical emissivity change rate corresponding to the historical normal process data, and then the largest historical emissivity change rate is selected from several historical emissivity change rates as the historical maximum emissivity change rate.

[0154] Specifically, in the first calculation subunit, the first threshold is calculated using the following formula:

[0155] K static =max(|r(t)|)×k safe1 ;

[0156] In the formula, k safe1 The first safety factor is set based on experience, preferably 1.5-2; K static The first threshold is calculated for the first computing unit.

[0157] Specifically, the second computing unit includes:

[0158] A stage acquisition subunit is used to acquire the current process stage, which includes a rapid heating stage, a constant temperature stage, and a slow heating stage.

[0159] The coefficient determination subunit is used to obtain a preset second safety factor based on the process stage.

[0160] The second calculation subunit is used to calculate a preset first threshold based on the emissivity change rate and the second safety factor.

[0161] Preferably, in the coefficient determination subunit, the second safety factor for the rapid heating stage is 1.5-2, the second safety factor for the constant temperature stage is 0.8-1, and the second safety factor for the slow heating stage is 1-1.2.

[0162] Specifically, in the second calculation subunit, the preset first threshold is calculated using the following formula:

[0163] K auto =(μ+3σ)×k safe2 ;

[0164] In the formula, K auto The first threshold is calculated by the second calculation unit, and μ is the average of the maximum absolute value of the rate of change of emissivity of each process curve. The standard deviation is the rate of change of emissivity calculated using the window standard deviation method.

[0165] Specifically, the third computing unit includes:

[0166] The initial data acquisition subunit is used to acquire the rate of change of emissivity during the initial time period;

[0167] The basic data calculation subunit is used to calculate the mean and standard deviation of the rate of change of radiance based on the rate of change of radiance during the initial time period.

[0168] The initial threshold calculation subunit is used to calculate the initial threshold based on the preset dynamic threshold coefficient, the mean of the rate of change of emissivity, and the standard deviation of the rate of change of emissivity.

[0169] The third calculation subunit is used to obtain the rate of change of radiance in subsequent time periods. The initial threshold is updated in real time using a sliding window algorithm to obtain a preset first threshold.

[0170] Specifically, in the initial data acquisition subunit, the initial time period is the time period between the start time of the process and the preset time, such as the time period from the start of the process to 100ms.

[0171] Specifically, in the basic data calculation subunit, the average rate of change of emissivity μ within 100ms is calculated. init and the standard deviation of the rate of change of emissivity σ init ;

[0172] Specifically, the initial threshold is calculated using the following formula:

[0173] K initial =μ init +A×σ init ;

[0174] In the formula, K initial is the initial threshold, and A is the preset dynamic threshold coefficient.

[0175] Specifically, in the third calculation subunit, the real-time update of the initial threshold is based on the following formula:

[0176] window data =r(t-N+1-m)tor(tm);

[0177] In the formula, N is the window length and m is the window lag time;

[0178] current μ =mean(window data );

[0179] In the formula, current μ This represents the average rate of change of the current window emissivity.

[0180] current σ =std(window data );

[0181] In the formula, current σ This represents the standard deviation of the current window emissivity change rate.

[0182] K dynamic =current μ +A×current σ ;

[0183] In the formula, K dynamic The first threshold is calculated based on the first calculation unit.

[0184] Specifically, the state determination module includes:

[0185] A threshold acquisition unit is used to acquire a preset first threshold.

[0186] The judgment unit is used to compare the rate of change of emissivity with the first threshold. When the rate of change of emissivity is greater than or equal to the first threshold, the wafer status is abnormal. When the rate of change of emissivity is less than the first threshold, the wafer status is normal.

[0187] An alarm unit is used to generate a control command to stop the current process and issue an alarm when the wafer is in an abnormal state.

[0188] Specifically, stopping the current process includes stopping heating and recording the current process stage, the current real-time emissivity, and the current rate of change of emissivity as historical data for subsequent monitoring.

[0189] Specifically, the emissivity during normal manufacturing processes and the emissivity during wafer skipping are compared to, for example... Figure 3 As shown. This embodiment compares the response time of a wafer condition monitoring method for RTP process with that of existing technologies when facing wafer skipping, as shown below. Figure 4 As shown, the entire process is monitored, with a response time of less than 50ms.

[0190] This invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements any of the wafer state monitoring methods for RTP processes described above.

[0191] The present invention also provides an electronic device. The electronic device of this invention includes: one or more processors; and a storage device for storing one or more computer programs, which, when executed by the one or more processors, cause the one or more processors to implement a wafer state monitoring method for RTP process provided by the present invention. References are made below. Figure 5 This illustrates a schematic diagram of the structure of a computer system 800 suitable for implementing embodiments of the present invention in an electronic device. For example... Figure 5 As shown, the computer system 800 includes a central processing unit (CPU) 801, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage section 808 into a random access memory (RAM) 803. The RAM 803 also stores various computer programs and data required for the operation of the computer system 800. The CPU 801, ROM 802, and RAM 803 are interconnected via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0192] The following components are connected to I / O interface 805: an input section 806 including a keyboard, mouse, etc.; an output section 807 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 808 including a hard disk, etc.; and a communication section 809 including a network interface card such as a LAN card, modem, etc. The communication section 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to I / O interface 805 as needed. A removable medium 811, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 810 as needed so that computer programs read from it can be installed into storage section 808 as needed.

[0193] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A wafer condition monitoring method for RTP process, characterized in that, Includes the following steps: S1: Obtain real-time emissivity data of the wafer when it is in the process position; S2: Calculate the rate of change of emissivity based on the real-time emissivity data obtained in step S1; S3: Compare the rate of change of emissivity obtained in step S2 with a preset first threshold to obtain the wafer status. When the wafer status is abnormal, stop the process and issue an alarm. Step S3 specifically includes the following steps: S31: Obtain a preset first threshold; this specifically includes the following steps: S31C: Obtain the rate of change of emissivity during the initial time period; S32C: The mean and standard deviation of the rate of change of emissivity calculated based on the rate of change of emissivity during the initial time period; S33C: The initial threshold is calculated based on the preset dynamic threshold coefficient, the mean of the rate of change of emissivity, and the standard deviation of the rate of change of emissivity; S32: Compare the rate of change of emissivity with the first threshold. When the rate of change of emissivity is greater than or equal to the first threshold, the wafer is in an abnormal state. When the rate of change of emissivity is less than the first threshold, the wafer is in a normal state. S33: When the wafer status is abnormal, a control command is generated to stop the current process and an alarm is issued; S34: The rate of change of radiance in subsequent time periods is obtained by using a sliding window algorithm to update the initial threshold in real time, thereby obtaining a preset first threshold.

2. The wafer condition monitoring method for RTP process according to claim 1, characterized in that, Step S1 specifically includes the following steps: S11: Obtain raw emissivity data; S12: Use a filtering algorithm to remove noise from the original emissivity data obtained in step S11 to obtain denoised data; S13: Use a machine learning algorithm to remove outliers from the denoised data obtained in step S12 to obtain real-time emissivity data.

3. The wafer condition monitoring method for RTP process according to claim 1, characterized in that, In S2, the rate of change of emissivity is calculated using one of the following methods: first-order difference method, window standard deviation calculation method, or absolute value deviation calculation method based on preset process baseline value.

4. The wafer condition monitoring method for RTP process according to claim 1, characterized in that, Calculating the preset first threshold includes the following steps: S31A: Obtain historical normal process data; S32A: Calculate the historical maximum emissivity change rate based on the historical normal process data obtained in step S31A; S33A: The preset first threshold is calculated based on the historical maximum rate of change of radiation and the preset first safety factor.

5. A wafer condition monitoring method for RTP process according to claim 1, characterized in that, Calculating the preset first threshold includes the following steps: S31B: Obtain the current process stage, which includes a rapid heating stage, a constant temperature stage, and a slow heating stage; S32B: Based on the process stage obtained in step S31B, a preset second safety factor is obtained; S33B: The preset first threshold is calculated based on the emissivity change rate obtained in step S2 and the second safety factor.

6. A wafer condition monitoring device for RTP process, used to implement the wafer condition monitoring method for RTP process as described in any one of claims 1-5, characterized in that, Includes the following steps: The data acquisition module is used to acquire real-time emissivity data of the wafer when it is in the process position; The rate of change calculation module is used to calculate the rate of change of emissivity based on the real-time emissivity data. The status judgment module is used to compare the rate of change of emissivity with a preset first threshold to obtain the wafer status. When the wafer status is abnormal, the process is stopped and an alarm is issued.

7. An electronic device, characterized in that, include: One or more processors; Storage device for storing one or more computer programs. When the one or more computer programs are executed by the one or more processors, the one or more processors implement a wafer condition monitoring method for RTP process as described in any one of claims 1-5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements a wafer condition monitoring method for RTP process as described in any one of claims 1-5.

Images